Subsurface Ventilation Engineering

Subsurface Ventilation Engineering.
Dedication: This work has been undertaken in fulfillment of

a long-standing promise to my former teacher, mentor and
dear friend,
Professor Frederick Baden Hinsley
The book is dedicated to his memory.
Subsurface Ventilation Engineering

Malcolm J. McPherson *
B.Sc., Ph.D., C.Eng., FIMinE, FIMM, Mem.AIME, Mem.ASHRAE
Emeritus Massey Professor of Mining Engineering and former
Associate Dean for Research and Graduate Studies,
College of Engineering,
Virginia Polytechnic Institute and State University.
Chairman, Mine Ventilation Services, Incorporated
*Formerly of the University of Nottingham, England

and the University of California, Berkeley.
Foreword to the Web version of the book.

In preparing this updated version of the textbook originally published in 1993 and under the title of
Subsurface Ventilation and Environmental Engineering, the opportunity has been taken to update
material that has seen significant change since the first version was written. The word Environmental
has been removed from the title as this term has increasingly come to refer to the surface environment.
Several of the major changes are particularly applicable to developments in software that have taken
place for planning the ventilation and condition of the underground environment.
The book remains a reference and text for professional engineers, researchers and teachers as well as
for undergraduate and graduate students who have an interest in the planning and control of the
environment in underground mines or other subsurface openings. Accordingly, the theoretical
underpinnings of the subject are dealt with in some detail.
While the practical techniques of topics such as ventilation surveys or the use of computer software are
included, this text is not intended specifically to be a ventilation officers handbook. Such handbooks are
best prepared within individual countries or institutions to reflect local methodologies, conditions and
legislation.
In addition to printed copies, this updated text is available at no cost for downloading from the Internet
address of Mine Ventilation Services, Inc. at: http://www.mvsengineering.com/
ii
Contents
Preface
Acknowledgments
Unit conversion table
Chapter 1. BACKGROUND TO SUBSURFACE VENTILATION AND ENVIRONMENTAL ENGINEERING

1.1. Introduction
1.2. A brief history of mine ventilation
1.3. The relationships between ventilation and other subsurface systems
PART 1. Basic principles of fluid mechanics and physical thermodynamics
Chapter 2. INTRODUCTION TO FLUID MECHANICS
2.1. Introduction
2.2. Fluid pressure
2.3. Fluids in motion
Chapter 3. FUNDAMENTALS OF STEADY FLOW THERMODYNAMICS
3.1. Introduction
3.2. Properties of state, work and heat
3.3. Some basic relationships
3.4. Frictional flow
3.5. Thermodynamic diagrams
PART 2. Subsurface ventilation engineering
Chapter 4. SUBSURFACE VENTILATION SYSTEMS
4.1. Introduction
4.2. Mine systems
4.3. District systems
4.4. Auxiliary systems
4.5. Controlled partial recirculation
4.6. Underground repositories
Chapter 5. INCOMPRESSIBLE FLOW RELATIONSHIPS
5.1. Introduction
5.2. Atkinson's equation, the Square Law
5.3. Determination of friction factor
5.4. Airway resistance
5.5. Airpower
A5 Shock loss factors.
Chapter 6. VENTILATION SURVEYS
6.1. Purpose and scope of surveys
6.2. Air quantity surveys
6.3. Pressure surveys
6.4. Organization of pressure-volume surveys
6.5. Air quality surveys
iii
Chapter 7. VENTILATION NETWORK ANALYSIS

7.1. Introduction
7.2. Fundamentals of ventilation network analysis
7.3. Methods of solving ventilation networks
7.4. Ventilation network simulation packages
Chapter 8. MINE VENTILATION THERMODYNAMICS
8.1. Introduction
8.2. Components of the mine cycle
8.3. The complete mine cycle
Chapter 9. VENTILATION PLANNING
9.1. Systems analysis of the planning procedure
9.2. Establishment of the basic network
9.3. Airflow requirements and velocity limits
9.4. Planning exercises and time phases
9.5. Ventilation economics and airway sizing
9.6 Traditional method of ventilation planning
Chapter 10. FANS
10.1. Introduction
10.2. Fan pressures
10.3. Impeller theory and fan characteristic curves
10.4. Fan laws
10.5. Fans in combination
10.6. Fan performance
10.7. Booster fans
PART 3. Gases in the subsurface
Chapter 11. GASES IN SUBSURFACE OPENINGS
11.1. Introduction
11.2. Classification of subsurface gases
11.3. Gas mixtures
11.4. Gas detection and monitoring
Chapter 12. METHANE
12.1. Overview and additional properties of methane
12.2. The retention and release of methane in coal
12.3. Migration of methane
12.4. Emission patterns into mine workings
12.5. Methane drainage
Chapter 13. RADIATION AND RADON GAS
13.1. Introduction
13.2. The uranium series and radioactive decay
13.3. Radon and its daughters
13.4. Prediction of levels of radiation
13.5. Methods of monitoring for radiation
13.6. Control of radiation in subsurface openings
iv
PART 4. Heat and humidity

Chapter 14. PSYCHROMETRY: THE STUDY OF MOISTURE IN AIR
14.1. Introduction
14.2. Basic relationships
14.3. The measurement of water vapour in air
14.4. Theory of the wet bulb thermometer
14.5. Further psychrometric relationships
14.6. Summary of psychrometric relationships
14.7. Deviations from classical theory
14.8. Psychrometric Charts
A14 Derivation of the Clausius-Clapeyron equation
Chapter 15. HEAT FLOW INTO SUBSURFACE OPENINGS
15.1 Introduction
15.2. Strata heat
15.3. Other sources of heat
A15.1 Carslaw and Jaeger solution to the heat conduction equation
A15.2 Gibson's algorithm
A15.3 Background theory of the heat transfer coefficient
A15.4 Derivation of latent heat of evaporation at a wet surface
Chapter 16. SIMULATION OF CLIMATIC CONDITIONS IN THE SUBSURFACE
16.1. Background
16.2. Elements of mine climate simulation programs
16.3. Using a mine climate simulator.
Chapter 17. PHYSIOLOGICAL REACTIONS TO CLIMATIC CONDITIONS
17.1. Introduction
17.2. Thermoregulation of the human body
17.3. Physiological heat transfer
17.4. Indices of heat stress
17.5. Heat illnesses
17.6. Cold environments
17.7. Heat tolerance and
17.8. Effect of mine climate on productivity.
AI7 Listing of a thermoregulation computer model
Chapter 18. REFRIGERATION PLANT AND MINE AIR CONDITIONING SYSTEMS
18.1. Introduction
18.2. The vapour compression cycle
18.3. Components and design of mine cooling systems
18.4. Air heating
PART 5. Dust
Chapter 19. THE HAZARDOUS NATURE OF DUSTS
19.1. Introduction
19.2. Classifications of dust
19.3. Dust in the human body
19.4. The assessment of airborne dust concentrations
Chapter 20. THE AERODYNAMICS, SOURCES AND CONTROL OF AIRBORNE DUST
20.1. Introduction
20.2. The aerodynamic behaviour of dust particles
20.3. The production of dust in underground openings
20.4. Control of dust in mines

PART 6. Fires and explosions
Chapter 21. SUBSURFACE FIRES AND EXPLOSIONS
21.1. Introduction
21.2. Causes of ignitions
21.3. Open fires
21.4. Spontaneous combustion
21.5. Stoppings, seals and section pressure balances
21.6. The use of inert gases
21.7. Fire gases and their interpretation.
21.8. Explosions
21.9. Protection of personnel
21.10.Emergency procedures and disaster management.
vi
PREFACE TO THE ORIGINAL TEXT

This book has been written as a reference and text for engineers, researchers, teachers and students
who have an interest in the planning and control of the environment in underground openings. While
directed primarily toward underground mining operations, the design procedures are also applicable to
other complex developments of subsurface space such as nuclear waste repositories, commercial
accommodation or vehicular networks. The book will, therefore, be useful for mining, civil, mechanical,
and heating, ventilating and air-conditioning engineers involved in such enterprises. The chapters on
airborne pollutants highlight means of measurement and control as well as physiological reaction. These
topics will be of particular interest to industrial hygienists and students of industrial medicine.
One of the first technical applications of digital computers in the world's mining industries was for
ventilation network analysis. This occurred during the early nineteen sixties. However, it was not until lowcost but powerful personal computers proliferated in engineering offices during the 'eighties that the full
impact of the computer revolution was realized in the day-to-day work of most mine ventilation engineers.
This book reflects the changes in approach and design procedures that have been brought about by that
revolution.
While the book is organized into six parts, it encompasses three broad areas. Following an introductory
background to the subject, chapters 2 and 3 provide the fundamentals of fluid mechanics and
thermodynamics that are necessary for a complete understanding of large three-dimensional ventilation
systems. Chapters 4 to 10, inclusive, offer a comprehensive treatment of subsurface airflow systems
while chapters 11 to 21 deal with the airborne hazards that are encountered in underground openings.
Each chapter is self-contained as far as is practicable. The inter-related features of the topics are
maintained by means of copious cross-references. These are included in order that practicing engineers
may progress through a design project and be reminded of the wider repercussions of decisions that
might be made. However, numerous cross-references can be a little distracting. The student is advised to
ignore them during an initial reading and unless additional information is sought.
Many of the chapters are subdivided into theoretical and descriptive sections. Again, these can be read
separately although a full understanding of the purpose and range of application of design procedures
can be gained only through a knowledge of both. When used as a refresher or text by practicing
engineers, it is suggested that the relevant descriptive section be consulted first and reference made back
to the corresponding analysis or derivation when necessary.
The use of the book as an aid to teaching and learning can be moulded to suit any given curriculum. For
the full education of a subsurface ventilation engineer, chapters 1 to 10 may be employed during a course
on ventilation, i.e. airflow processes, leaving the chapters on gases, heat, dust, and fires and explosions
for further courses. Alternatively, undergraduate courses my concentrate on the practical aspects of the
subject, leaving the more theoretical analyses to graduate school. In any event the teacher may compile
his or her own syllabus at any given level by choosing relevant sections from selected chapters.
In most countries, mining activities are regulated by specific state or national legislation. This book has
been written for an international audience and reflects the author's experience of teaching and practice in
a number of countries. While guideline threshold limit values are given, the reader is frequently reminded
to consult the relevant local regulations for specific mandatory requirements and limitations on practical
procedures. To reflect the international readership, Systme Internationale (SI) units are employed and a
comprehensive list of conversion factors is provided.
vii
ACKNOWLEDGMENTS
There are many people without whose contributions the original edition of this book could not have been
written. First, I thank Shirley, my wife, for her patience and understanding not only through the long hours
of midnight oil burning that took place during the writing but, more particularly, for the extended periods,
stretching over many years, when she was left alone to look after the home and family while I was deep
under the surface of some faraway country.
I am grateful to former colleagues in the Department of Mining Engineering, University of Nottingham,
England, for sowing seeds of ideas that later produced practical designs and procedures, many of which
are reflected in this book; especially Ian Longson with whom I rediscovered the fascinations of
thermodynamic logic, Leslie H. Morris, Dr. Jim R. Brown and, most of all, Professor F. Baden Hinsley to
whom the book is dedicated. I am also privileged in having worked with excellent students from whom I
learned a great deal, at Nottingham, the University of California, Berkeley, and at Virginia Tech.
Despite having been involved in numerous research investigations, my knowledge of subsurface
ventilation and environmental engineering has been advanced primarily by working on feasibility studies
and practical projects with mining engineers in many countries. Most of the case studies and examples in
the book originated in such work. In particular, I am truly grateful for having had the opportunity of
interacting with dedicated professional engineers in the United Kingdom, the countries of East and West
Europe, South Africa, Australasia, India, South America, the United States of America and Canada.
I am indebted to the two ladies who shared typing the original manuscript. First, my daughter Alison D.
McPherson who took great delight in correcting some of my mathematics, and Lucy Musante, my
Secretarial assistant at Mine Ventilation Services, Inc. the most skilled and dedicated secretary with
whom I have ever worked. Most of the initial reviews of chapters were undertaken by staff of Mine
Ventilation Services, namely Daniel J. Brunner, Justus Deen, Martha O'Leary and, most particularly, Keith
G. Wallace who willingly volunteered far more than his fair share of the work. Several chapters were
reviewed by Dr. Felipe Calizaya, formerly at Berkeley and now at the University of Utah.
Some of the analyses described in the book arose directly out of funded research. The physiological
model in chapter 17 was developed for the U.S. Department of Energy via Sandia National Laboratories
as part of an investigation into climatic conditions in a deep geological repository for nuclear waste. Some
of the heat transfer and climatic simulation studies in chapters 15 and 16, and investigations into the
installation of booster fans outlined in chapter 9 were assisted by funding from the Generic Mineral
Technology Center in Mine Systems Design and Ground Control, Office of Mineral Institutes, U.S. Bureau
of Mines under Grant No. G1125151. I am indebted to those organizations for financing the work.
Finally, but also foremost, I thank the Good Lord for guiding my career to a point when I could prepare
this book.
viii
TABLE OF CONVERSION FACTORS BETWEEN IMPERIAL AND SI UNITS
Quantity
Imperial to SI
SI to Imperial
Length
1 ft
1 yd
1 in
= 0.304 8
= 0.914 4
= 0.025 4
m
m
m
1m
= 3.280 8
= 1.093 6
= 39.370 1
ft
yd
in
Area
1 ft2
1 in2
= 0.092 9
= 0.000 645
m2
m2
1 m2
= 10.763 9
= 1550.003
ft2
in2
Acceleration
1 ft/s2
= 0.304 8
m/s2
1 m/s2
= 3.280 8
ft/s2
Force
1 lbf
1 imp.ton f
= 4.448 2
= 9964.02
N
N
1N
=0.2248
lbf
Velocity
1 ft/s
1 ft/min
= 0.304 8
= 0.005 08
m/s
m/s
1 m/s
= 3.2808
= 196.85
ft/s
ft/min
Volume
1 ft3
= 0.028 32
m3
1 m3
= 35.315
ft3
1 yd3
1 imp. gal
1 U.S. gal
= 0.764 56
= 4.545
= 3.785
m3
litre
litre
1 litre
(0.001 m3)
= 1.308
= 0.2200
= 0.2642
yd3
imp. gal
U.S. gal
1 ft3/s
1 ft3/min
1 imp gal/h
1 imp gal/min
= 0.028 32
= 0.000 472
= 0.004 55
= 0.004 55
= 4.545
= 0.075 75
m3/s
m3/s
m3/h
m3/min
litre/min
litre/s
= 35.315
= 2118.9
= 220.0
= 220.0
= 0.220
= 13.20
= 15.84
ft3/s
ft3/min
imp. gal/h
imp. gal/min
imp. gal/min
imp. gal/min
U.S. gal/min
= 0.06313
litre/s
= 2.204 62
= 0.984 20
lb
imp. ton
= 1.1023
short ton
Volume Flow
1 U.S.
gal/min
Mass
1 lb
1 imp. ton
(2240 lb)
1 short ton
(2000 lb)
= 0.453 592
= 1.016 05
kg
t
= 0.907 18
ix
1 m3/s
1 m3/h
1 m3/min
1 litre/min
1 litre/s
1 kg
1 t = 1000 kg
Quantity
Pressure,
stress
Imperial to SI
1 lbf/ft2
= 47.880
1 lbf/in2
1 in w.g.
1 ft w.g.
1 mm w.g.
1 in Hg
1 mm Hg
= 6894.76
= 249.089
= 2989.07
= 9.807
= 3386.39
= 133.32
= 1.333 2
SI to Imperial
N/m2 =
Pa
N/m2
N/m2
N/m2
N/m2
N/m2
N/m2
mb
1 N/m2 = Pa
= 0.020 88
lbf/ft2
0.000 145
= 0.004 015
= 0.000 3346
= 0.101 97
= 0.000 2953
= 0.007 501
= 0.01
lbf/in2
in w.g.
ft w.g.
mm w.g.
in Hg
mm Hg
mb
Note: The millibar (1 mb = 100 N/m2) is included here as it is a familiar metric

unit of pressure. It is not, however, an SI unit.
1 Atk
= 0.059 71
1 PU
= 1.118 3
Airway
specific
resistance
1 in w.g.
per 10 000
ft3 /min
= 22.366
Ns2/m
8
Friction
Factor
1 lbf min2/ft4
= 1.8554 x l06
kg/m3
Density
1 lb/ft3
1 imp.
ton/yd3
1 short
ton/yd3
= 16.018 5
= 1328.94
kg/m3
kg/m3
= 1186.55
kg/m3
1 ft lbf
= 1.355 82
1Btu
1 cal
1 therm
1 kWh
= 1055.06
= 4.186 8
= 105.506
= 3.6
J
J
MJ
MJ
Power
1 hp
= 745.700
Heatflow
1 ft lbf/min
1 Btu/min
1 RT
Refrigeration
(imp.) ton
= 0.0226
= 17.584
= 3517
W
W
W
Airway
resistance
Energy,
work, heat
Ns2/m
8
N
s2/m8
1 Ns2/m8
= 16.747
Atk
= 0.894 2
PU
= 0.044 7
in w.g.
per 10 000
ft3/min
1 kg/m3
= 539.0 x 109
lbf min2/ft4
1 kg/m3
= 0.062 43
= 0.000 753
lb/ft3
imp. ton/yd3
= 0.000843
short ton/yd3
= 0.737 56
ft/lbf
= 0.000 948
= 0.238 89
= 0.009 478
= 0.000 278
Btu
cal
therm
Wh
1W
= 0.001 341
hp
1W
= 44.254
= 0.056 87
= 0.000 2843
ft lbf/min
Btu/min
RT
1J
Quantity
Imperial to SI
SI to Imperial
Specific
energy
1 ft lbf/lb
= 2.989
J/kg
1 J/kg
= 0.3345
ft lbf/lb
Calorific
value
1 Btu/lb
= 2326
J/kg
1 J/kg
= 0.000 430
Btu/lb
1 therm/imp.
ton
1 therm/short
ton
= 0.103 8
MJ/kg
= 9.634
= 0.116 3
MJ/kg
= 8.602
therm/imp
. ton
therm/sho
rt ton
Gas
constants
1 ft lbf/lb R
= 5.380 3
J/kg K
1 J/kg K
Specific heat
Specific
entropy
1 Btu/lb R
= 4186.8
J/kg K
1 J/kg K
= 0.000 2388
Btu/lb R
Specific
volume
1 ft3 /lb
= 0.062 43
m3 /kg
1 m3 /kg
= 16.018
ft3 /lb
1 ft3 /imp. ton
= 0.027 87
m3 /t
= 35.881
1 ft3 /short ton
= 0.031 21
m3 /t
= 32.037
ft3 /imp.
ton
ft3 /short
ton
= 0.185 9
ft lbf/lb R
Note: 1 metric tonne (t) = 1 000 kg

Dynamic
viscosity
Kinematic
viscosity
Permeability
1 lb/ft s
= 1.488 16
Ns/m2
1 poise
= 0.1
Ns/m2
1 ft2/s
= 0.092 903
m2/s
1 stokes
= 0.000 1
1 Darcy
1 md
1 Ns/m2
= 0.671 97
lb/ft s
= 10
poise
= 10.763 9
ft2/s
m2/s
= 10 000
stokes
= 0.98693x10l2
= 0.98693x10l5
m2
= 1.01324-12
Darcy
m2
= 1.01324-15
md
1 m2/s
Thermal
conductivity
1 Btu ft/ft2 h
R
= 1.730 73
W/m K
1 W/mK
= 0.577 79
Btu ft/ft2 h R
Thermal
diffusivity
1 ft2/s
= 0.092 303
m2/s
1 m2/s
= 10.764
ft2/s
1 ft2/h
= 2.5806x10-5
m2/s
1 m2/s
= 38 750
ft2/h
xi
Quantity
Imperial to SI
SI to Imperial
Thermal
gradient
1 F/ft
= 1.822 7
C/m
1 C/m
= 0.548 6
F/ft
Moisture
content
1 lb/lb
=1
kg/kg
1 kg/kg
=1
lb/lb
1 gr/lb
1 rad
1 Curie
1 rem
1 Roentgen
= 0.000 1429
= 0.01
= 37 x 109
= 0.01
= 2.58 x 10-4
kg/kg
Gray
Bq
Sv
C/kg
1 Gray
1 Bq
1 Sv
1 C/kg
= 7000
= 100
= 27 x 10-12
= 100
= 3876
gr/lb
rad
Curie
rem
Roentgen
Radiation
Notes:
Temperature
1 Gray
= 1 J/kg
1 Sievert (Sv)
= 1 J/kg
K
R
= C + 273.15
= F + 459.67
=1 disintegration/s
1 Becqueral
(Bq)
1 Coulomb (C)
= 1 amp.s
For differential temperatures, 1 Centigrade degree = 1.8 Fahrenheit degrees.

For actual temperature, 1.8 x t(C) + 32 = F
and
t(F) - 32
1.8
= C
xii
Chapter 1. Background
Chapter 1.
MALCOLM J. McPHERSON
Background to subsurface ventilation

and environmental engineering
1.1 INTRODUCTION
1.2 A BRIEF HISTORY OF MINE VENTILATION
1.3. THE RELATIONSHIPS BETWEEN VENTILATION AND OTHER SUBSURFACE SYSTEMS6

1.3.1. The objectives of subsurface ventilation
6
1.3.2. Factors that affect the underground environment
6
1.3.3. The integration of ventilation planning into overall system design
8
_____________________________________________________________________________
1.1
INTRODUCTION
Ventilation is sometimes described as the lifeblood of a mine, the intake airways being arteries that
carry oxygen to the working areas and the returns veins that conduct pollutants away to be expelled
to the outside atmosphere. Without an effective ventilation system, no underground facility that
requires personnel to enter it can operate safely.
The slaughter of men, women and children that took place in the coal mines of Britain during the
eighteenth and nineteenth centuries resulted in the theory and art of ventilation becoming the
primary mining science. The success of research in this area has produced tremendous
improvements in underground environmental conditions. Loss of life attributable to inadequate
ventilation is now, thankfully, a relatively infrequent occurrence. Falls of ground rather than
ventilation-related factors have become the most common cause of fatalities and injuries in
underground mines. Improvements in ventilation have also allowed the productivity of mines to be
greatly improved. Neither the very first nor the very latest powered machines could have been
introduced underground without an adequate supply of air. Subsurface ventilation engineers are
caught up in a continuing cycle. Their work allows rock to be broken in ever larger quantities and at
greater depths. This, in turn, produces more dust, gases and heat, resulting in a demand for yet
better environmental control.
This opening chapter takes a necessarily cursory look at the long history of mine ventilation and
discusses the interactions between ventilation and the other systems that, jointly, comprise a
complete mine or underground facility.
1.2
A BRIEF HISTORY OF MINE VENTILATION
Observations of the movements of air in underground passages have a long and fascinating
history. Between 4000 and 1200 B.C., European miners dug tunnels into chalk deposits searching
for flint. Archaeological investigations at Grimes Graves in the South of England have shown that
these early flint miners built brushwood fires at the working faces presumably to weaken the rock.
However, those Neolithic miners could hardly have failed to observe the currents of air induced by
the fire. Indeed, the ability of fire to promote airflow was rediscovered by the Greeks, the Romans,
in medieval Europe and during the Industrial Revolution in Britain.
The Laurium silver mines of Greece, operating in 600 B.C. have layouts which reveal that the
Greek miners were conscious of the need for a connected ventilating circuit. At least two airways
served each major section of the mine and there is evidence that divided shafts were used to
provide separate air intake and return connections to surface. Underground mines of the
Roman Empire often had twin shafts, and Pliny (23-79 A.D.) describes how slaves used palm
fronds to waft air along tunnels.
Although metal mines were worked in Europe during the first 1500 years A.D., there remain little
documented descriptions of their operations. The first great textbook on mining was written in Latin
by Georgius Agricola, a physician in a thriving iron ore mining and smelting community of
Bohemia in Central Europe. Agricola's "De Re Metallica", produced in 1556, is profusely illustrated.
A number of the prints show ventilating methods that include diverting surface winds into the
mouths of shafts, wooden centrifugal fans powered by men and horses, bellows for auxiliary
ventilation and air doors. An example of one of Agricola's prints is reproduced in Figure 1.1.
Agricola was also well aware of the dangers of blackdamp, air that has suffered from a reduction in
oxygen content, - 'miners are sometimes killed by the pestilential air that they breathe," and of the
explosive power of "firedamp", a mixture of methane and air "likened to the fiery blast of a dragon's
breath." De Re Metallica was translated into English in 1912 by Herbert C. Hoover and his wife,
Lou. Hoover was a young American mining engineer who graduated from Stanford University and
subsequently served as President of the United States during the term 1929-1933.
From the 17th Century onwards, papers began to be presented to the Royal Society of the United
Kingdom on the explosive and poisonous nature of mine atmospheres. The Industrial Revolution
brought a rapid increase in the demand for coal. Conditions in many coal mines were quite horrific
for the men, women, and children who were employed in them during the 18th and 19th centuries.
Ventilation was induced either by purely natural effects, stagnating when air temperatures on the
surface and underground were near equal, or by fire. The first ventilating furnaces of that era were
built on surface but it was soon realized that burning coals suspended in a wire basket within the
upcast shaft gave improved ventilation. Furthermore, the lower the basket, the better the effect.
This quickly led to the construction of shaft bottom furnaces.
The only form of illumination until the early eighteen hundred's was the candle. With historical
hindsight we can see the conjunction of circumstances that caused the ensuring carnage: a
seemingly insatiable demand for coal to fuel the steam engines of the Industrial Revolution, the
working of seams rich in methane gas, inadequate ventilation, furnaces located in methane laden
return air and the open flames of candles. There are many graphic descriptions of methane and
coal dust explosions, the suffering of mining communities, the heroism of rescue attempts and the
strenuous efforts of mining engineers and scientists to find means of improving ventilation and
providing illumination without the accompanying danger of igniting methane gas. Seemingly
oblivious to the extent of the danger, miners would sometimes ignite pockets of methane
intentionally for amusement and to watch the blue flames flickering above their heads. Even the
renowned engineer, George Stephenson, admitted to this practice during the inquiries of a
government select committee on mine explosions in 1835. A common method of removing methane
was to send a "fireman'' in before each shift, covered in sackcloths dowsed in water and carrying a
candle on the end of a long rod. It was his task to burn out the methane before the miners went into
the working faces.
John Buddle (1773-1843), an eminent mining engineer in the north of England produced two
significant improvements. First, he introduced "dumb drifts" which bled sufficient fresh air from the
base of a downcast shaft to feed the furnace. The return air, laden with methane, bypassed the
furnace. The products of combustion entering the upcast shaft from the furnace were too cool to
ignite the methane but still gave a good chimney effect in the shaft, thus inducing airflow around the
mine. Buddle's second innovation was "panel (or split) ventilation". Until that time, air flowed
sequentially through work areas, one after the other, continually increasing in methane
concentration. Buddle originally divided the mine layout into discrete panels, with intervening barrier
pillars, to counteract excessive floor heave. However, he found that by providing an intake and
A-MACHINE FIRST DESCRIBED. B-THIS WORKMAN, TREADING WITH HIS FEET, IS COMPRESSING THE BELLOWS. C-BELLOWS WITHOUT NOZZLES. D-HOLE BY WHICH HEAVY
VAPOURS OR BLASTS ARE BLOWN OUT. E-CONDUITS. F-TUNNEL. G-SECOND
MACHINE DESCRIBED. H-WOODEN WHEEL. I-ITS STEPS. K-BARS. L-HOLE IN
SAME WHEEL. M-POLE. N-THIRD MACHINE DESCRIBED. O-UPRIGHT AXLE.
P-ITS TOOTHED DRUM. Q-HORIZONTAL AXLE. R-ITS DRUM WHICH IS MADE OF RUNDLES.
Figure 1.1 A print from Agricola's " De Re Metallica"

(Reproduced by permission of Dover Publications)
return separately to each panel the ventilating quantities improved markedly and methane
concentrations decreased. He had discovered, almost by accident, the advantages of parallel
layouts over series circuits. The mathematical proof of this did not come until Atkinson's theoretical
analyses several decades later.
The quest for a safe form of illumination went on through the eighteenth century. Some of the
earlier suggestions made by scientists of the time, such as using very thin candles, appear quite
ludicrous to us today. One of the more serious attempts was the steel flint mill invented in 1733 by
Carlisle Spedding, a well known mining engineer, again, in the north of England (Figure 1.2). This
device relied upon a piece of flint being held against a rapidly revolving steel wheel. The latter was
driven through a gear mechanism by a manually rotated handle. The complete device was strapped
to the chest of a boy whose job was to produce a continuous shower of sparks in order to provide
some illumination for the work place of a miner. The instrument was deemed safer than a candle
but the light it produced was poor, intermittent, and still capable of igniting methane.
Figure 1.2 Spedding's Flint Mill (Reproduced by permission of Virtue and Co., Ltd.)
A crisis point was reached in 1812 when a horrific explosion at Felling, Gateshead killed 92 miners.
With the help of local clergymen, a society was formed to look into ways of preventing such
disasters. Contact was made with Sir Humphrey Davy, President of the Royal Society, for
assistance in developing a safe lamp. Davy visited John Buddle to learn more of conditions in the
mines. As this was well before the days of electricity, he was limited to some form of flame lamp.
Within a short period of experimentation he found that the flame of burning methane would not
readily pass through a closely woven wire mesh. The Davy Lamp had arrived (Figure 1.3).
Buddle's reaction is best expressed in a letter he wrote to

Davy.
"I first tried it in a explosive mixture on the surface,
and then took it into the mine... it is impossible for
me to express my feelings at the time when I first
suspended the lamp in the mine and saw it red
hot...I said to those around me, 'We have at last
subdued this monster."
The lamp glowed 'red hot' because of the methane burning
vigourously within it, yet the flames could not pass through
the wire mesh to ignite the surrounding firedamp.
Davy lamps were introduced into British mines, then
spread to other countries. Nevertheless, in the absence of
effective legislation, candles remained in widespread use
through the nineteenth century because of the better light
that they produced.
Perhaps the greatest classical paper on mine ventilation
was one entitled "On the Theory of the Ventilation of
Mines", presented by John Job Atkinson to the North of
England Institute of Mining Engineers in December, 1854.
Figure 1.3 The original appearance
Atkinson was a Mining Agent, an intermediary between
of the Davy safety lamp. (reproduced
management and the mine owners. He later became one
by permission of Virtue and Co., Ltd.).
of the first Inspectors of Mines. Atkinson appears to have
been well educated in mathematics and languages, and
was clearly influenced by the earlier work of French hydraulic engineers (Chapter 5). He seems to
have had some difficulty in having his paper accepted. Officers of the Institute decided, perhaps
understandably, that the 154 page paper was too long to be presented at a meeting. It was,
however, published and a meeting of the Institute arranged to discuss it. Despite publicity referring
to the importance of the subject, attendance at the meeting was poor and there was little
discussion. In this paper, Atkinson proposed and expanded upon the principles on which most
modern mine ventilation planning is still based. However, the analytical reasoning and
mathematical analyses that he developed in great detail were simply too much for engineers of the
day. The paper was consigned to the archives and it was some sixty years after Atkinson's death
that his work was "rediscovered" and put into practice.
During Atkinson's productive years the first power driven ventilators began to appear. These varied
from enormous steam-driven piston and cylinder devices to elementary centrifugal fans.
The years around the turn of the century saw working conditions in mines coming under legislative
control. Persons responsible for underground mining operations were required to obtain minimum
statutory qualifications. Mine manager's examination papers concentrated heavily on ventilation
matters until well into the twentieth century.
The nineteen twenties saw further accelerated research in several countries. Improved
instrumentation allowed organized ventilation surveys to be carried out to measure airflows and
pressure drops for the purposes of ventilation planning, although there was no practical means of
predicting airflows in other than simple circuits at that time. Atkinson's theory was confirmed in
practice. The first successful axial fans were introduced in about 1930.
In 1943, Professor F.Baden Hinsley produced another classical paper advancing understanding
of the behaviour of airflow by using thermodynamic analyses. Hinsley also supervised the work at
Nottingham University that led to the first practical use of analog computers in 1952 to facilitate
ventilation planning. This technique was employed widely and successfully for over a decade. The
development of ventilation network analysis programs for digital computers in the early sixties
rendered the analog devices obsolete. Initially, the network programs were written for, and required
the power of mainframe computers. These were employed throughout the seventies. However, the
nineteen eighties saw a shift to desktop computers and corresponding programs were developed.
This is now the dominant method used for ventilation planning (Chapter 7).
The discipline of mine ventilation is an addictive subject for researchers of industrial history, full of
lost discoveries and rediscoveries, excitement and despair, achievement and tragedy. It has been
the subject of many papers and books. An excellent place to commence further reading is the text
by Saxton serialized in Volume 146 of the Mining Engineer (Institution of Mining Engineers, U.K.),
(1986-'87).
1.3. THE RELATIONSHIPS BETWEEN VENTILATION AND OTHER

SUBSURFACE SYSTEMS
1.3.1. The objectives of subsurface ventilation
The basic objective of an underground ventilation system is clear and simple. It is to provide
airflows in sufficient quantity and quality to dilute contaminants to safe concentrations in all parts of
the facility where personnel are required to work or travel. This basic requirement is incorporated
into mining law in those countries that have such legislation. The manner in which "quantity and
quality" are defined varies from country to country depending upon their mining history, the
pollutants of greatest concern, the perceived dangers associated with those hazards and the
political and social structure of the country. The overall requirement is that all persons must be able
to work and travel within an environment that is safe and which provides reasonable comfort. An
interpretation of the latter phrase depends greatly on the geographical location of the mine and the
background and expectations of the workforce. Personnel in a permafrost mine work in conditions
that would be unacceptable to miners from an equatorial region, and vice versa, and neither set of
conditions would be tolerated by factory or office workers. This perception of "reasonable comfort"
sometimes causes misunderstandings between subsurface ventilation engineers and those
associated with the heating and ventilating industry for buildings.
While maintaining the essential objectives related to safety and health, subsurface environmental
engineering has, increasingly, developed a wider purpose. In some circumstances, atmospheric
pressure and temperature may be allowed to exceed the ranges that are acceptable for human
tolerance. For example, in an underground repository for high level nuclear waste, a containment
drift may be sealed against personnel access after emplacement of the waste canisters has been
completed. However, the environment within the drift must still be maintained such that rock wall
temperatures are controlled. This is necessary to enable the drift to be reopened relatively quickly
for retrieval of the nuclear waste at any subsequent time during the active life of the repository.
Other forms of underground storage often require environmental control of pressure, temperature
and humidity for the preservation of the stored material. Yet another trend is towards automated
(manless) working faces and the possible use of underground space for in-situ mineral processing.
In such zones of future mines, environmental control will be required for the efficient operation of
machines and processes, but not necessarily with an atmosphere acceptable to the unprotected
human physiology.
1.3.2. Factors that affect the underground environment
During the development and operation of a mine or other underground facility, potential hazards
arise from dust, gas emissions, heat and humidity, fires, explosions and radiation. Table 1.1 shows
the factors that may contribute towards those hazards. These divide into features that are imposed
by nature and those that are generated by design decisions on how to develop and operate the
facility.
Table 1.1 Factors that feature in the creation and control of hazards in the subsurface environment.
Factors that contribute to

NATURAL
FACTORS
DESIGN
FACTORS
Depth
below
surface
Method of
mining
Surface
climate
Layout of mine
or facility
Geology
Physical
and
chemical
properties
of rock
Rate of rock
fragmentation
HAZARD
DUST
GAS
EMISSIONS
ANCILLARY
CONTROL
AIRFLOW
CONTROL
Dust
suppression
Main fans
Gas
drainage
Mineral
clearance
Method of
working
Gas content
of strata
Type, size
and siting of
equipment
Subsurface
liquids
Vehicular
traffic
Age of
openings
Methods of control
Stored
materials
HEAT AND
HUMIDITY
Refrigeration
systems
Natural
ventilation
Airlocks,
stoppings, air
crossings,
regulators
FIRES AND
EXPLOSIONS
Monitoring
systems
RADIATION
Booster fans,
auxiliary
ventilation
Number, size
and layout of
openings
The major method of controlling atmospheric conditions in the subsurface is by airflow. This is
produced, primarily, by main fans that are usually, but not necessarily, located on surface. National
or state mining law may insist that main fans are sited on surface for gassy mines. While the main
fan, or combination of main fans, handles all of the air that circulates through the underground
network of airways, underground booster fans serve specific districts only. Auxiliary fans are used
to pass air through ducts to ventilate blind headings. The distribution of airflow may further be
controlled by ventilation doors, stoppings, air crossings and regulators.
It is often the case that it becomes impracticable or impossible to contend with all environmental
hazards by ventilation alone. For example, increases in air temperature caused by compression of
the air in the downcast shafts of deep mines may result in that air being too hot for personnel even
before it enters the workings. No practical amount of increased airflow will solve that problem. Table
1.1 includes the ancillary control measures that may be advisable or necessary to supplement the
ventilation system in order to maintain acceptable conditions underground.
1.3.3. The integration of ventilation planning into overall system design

The design of a major underground ventilation and environmental control system is a complex
process with many interacting features. The principles of systems analyses should be applied to
ensure that the consequences of such interaction are not overlooked (Chapter 9). However,
ventilation and the underground environment should not be treated in isolation during planning
exercises. They are, themselves, an integral part of the overall design of the mine or subsurface
facility.
It has often been the case that the types, numbers and sizes of machines, the required rate of
mineral production and questions of ground stability have dictated the layout of a mine without,
initially, taking the demands of ventilation into account. This will result in a ventilation system that
may lack effectiveness and, at best, will be more expensive in both operating and capital costs than
would otherwise have been the case. A common error has been to size shafts that are appropriate
for the hoisting duties but inadequate for the long term ventilation requirement of the mine. Another
frequent, related problem is a ventilation infrastructure that was adequate for an initial layout but
lacks the flexibility to handle fluctuating market demands for the mineral. Again, this can be very
expensive to correct. The results of inadequate ventilation planning and system design are
premature cessation of production, high costs of reconstruction, poor environmental conditions and,
still too often, tragic consequences to the health and safety of the workforce. It is, therefore, most
important that ventilation engineers should be incorporated as an integral part of a design team
from the initial stages of planning a new mine or other underground facility.
REFERENCES
Agricola, G. (1556). De Re Metallica. Translated from Latin by H.C. Hoover and L.H. Hoover.
Modern publication by Dover Pub., Inc., New York (1950).
Atkinson, J.J. (1854). On the theory of the ventilation of mines. North of England Institute of Mining
Engineers, No. 3 , p 118.
Hartley, Sir Harold. "Humphrey Davy". S.R. Publishers Ltd.
Hinsley, F.B. (1967). The control of atmospheric conditions in mines. 11th Cadman Memorial
Lecture. Mining Engineer No. 77, pp 289.
Hinsley, F.B. (1970). The development of the mechanical ventilation of coal mines, 1870-1940.
Univ. of Nottingham Mining Magazine Vol. 22.
Mason, E. (1954). Practical Coal Mining. Virtue and Co. Ltd. London.
McPherson, M.J. (1964). Mine ventilation network problems, solution by digital computer. Colliery
Guardian Vol. 209 No. 5392, pp 253-259.
Saxton, I. (1986-1987). Coal mine ventilation from Agricola to the 1980's. Serialized in the Mining
Engineer by the Inst. of Mining Engineers Vol. 146.
Scott, D.R., Hinsley, F.B. and Hudson, R.F. (1953). A calculator for the solution of ventilation
network problems. Trans. Inst. Min. Engrs. Vol. 112, pp 623.
Wang, Y.J. and Hartman, H.L. (1967). Computer solution of three-dimensional mine ventilation
networks with multiple fans and natural ventilation. Int. J. of Rock Mech. and Min. SC. Vol. 2, No. 2,
pp 129-154.
Part 1
Basic principles
of fluid mechanics
and physical
thermodynamics.
Introduction to Fluid Mechanics
Malcolm J. McPherson
Chapter 2. Introduction to Fluid Mechanics

2.1 INTRODUCTION ............................................................................................. 1
2.1.1 The concept of a fluid............................................................................................................. 1
2.1.2 Volume flow, Mass flow and the Continuity Equation............................................................ 3
2.2 FLUID PRESSURE.......................................................................................... 3

2.2.1 The cause of fluid pressure.................................................................................................... 3
2.2.2 Pressure head........................................................................................................................ 4
2.2.3 Atmospheric pressure and gauge pressure........................................................................... 5
2.2.4. Measurement of air pressure................................................................................................ 5
2.2.4.1. Barometers...................................................................................................................... 5
2.2.4.2. Differential pressure instruments .................................................................................... 6
2.3 FLUIDS IN MOTION ........................................................................................ 8

2.3.1. Bernoulli's equation for ideal fluids ....................................................................................... 8
Kinetic energy............................................................................................................................. 8
Potential energy ......................................................................................................................... 9
Flow work ................................................................................................................................... 9
2.3.2. Static, total and velocity pressures. .................................................................................... 11
2.3.3. Viscosity .............................................................................................................................. 12
2.3.4. Laminar and turbulent flow. Reynolds Number................................................................... 14
2.3.5. Frictional losses in laminar flow, Poiseuille's Equation....................................................... 16
2.3.6. Frictional losses in turbulent flow........................................................................................ 22
2.3.6.1. The Chzy-Darcy Equation ........................................................................................... 22
2.3.6.2. The coefficient of friction, f. ........................................................................................... 25
2.3.6.3. Equations describing f - Re relationships...................................................................... 27
Laminar Flow ............................................................................................................................ 27
Smooth pipe turbulent curve .................................................................................................... 28
Rough pipes ............................................................................................................................. 29
Bibliography ....................................................................................................... 31
2.1 INTRODUCTION
2.1.1 The concept of a fluid
A fluid is a substance in which the constituent molecules are free to move relative to each other.
Conversely, in a solid, the relative positions of molecules remain essentially fixed under nondestructive conditions of temperature and pressure. While these definitions classify matter into fluids
and solids, the fluids sub-divide further into liquid and gases.
Molecules of any substance exhibit at least two types of forces; an attractive force that diminishes
with the square of the distance between molecules, and a force of repulsion that becomes strong
when molecules come very close together. In solids, the force of attraction is so dominant that the
molecules remain essentially fixed in position while the resisting force of repulsion prevents them
from collapsing into each other. However, if heat is supplied to the solid, the energy is absorbed
internally causing the molecules to vibrate with increasing amplitude. If that vibration becomes
sufficiently violent, then the bonds of attraction will be broken. Molecules will then be free to move in
relation to each other - the solid melts to become a liquid.
2-1
When two moving molecules in a fluid converge on each other, actual collision is averted (at normal
temperatures and velocities) because of the strong force of repulsion at short distances. The
molecules behave as near perfectly elastic spheres, rebounding from each other or from the walls of
the vessel. Nevertheless, in a liquid, the molecules remain sufficiently close together that the force of
attraction maintains some coherence within the substance. Water poured into a vessel will assume
the shape of that vessel but may not fill it. There will be a distinct interface (surface) between the
water and the air or vapour above it. The mutual attraction between the water molecules is greater
than that between a water molecule and molecules of the adjacent gas. Hence, the water remains in
the vessel except for a few exceptional molecules that momentarily gain sufficient kinetic energy to
escape through the interface (slow evaporation).
However, if heat continues to be supplied to the liquid then that energy is absorbed as an increase in
the velocity of the molecules. The rising temperature of the liquid is, in fact, a measure of the
internal kinetic energy of the molecules. At some critical temperature, depending upon the applied
pressure, the velocity of the molecules becomes so great that the forces of attraction are no longer
sufficient to hold those molecules together as a discrete liquid. They separate to much greater
distances apart, form bubbles of vapour and burst through the surface to mix with the air or other
gases above. This is, of course, the common phenomenon of boiling or rapid evaporation. The liquid
is converted into gas.
The molecules of a gas are identical to those of the liquid from which it evaporated. However, those
molecules are now so far apart, and moving with such high velocity, that the forces of attraction are
relatively small. The fluid can no longer maintain the coherence of a liquid. A gas will expand to fill
any closed vessel within which it is contained.
The molecular spacing gives rise to distinct differences between the properties of liquids and gases.
Three of these are, first, that the volume of gas with its large intermolecular spacing will be much
greater than the same mass of liquid from which it evaporated. Hence, the density of gases
(mass/volume) is much lower than that of liquids. Second, if pressure is applied to a liquid, then the
strong forces of repulsion at small intermolecular distances offer such a high resistance that the
volume of the liquid changes very little. For practical purposes most liquids (but not all) may be
regarded as incompressible. On the other hand, the far greater distances between molecules in a
gas allow the molecules to be more easily pushed closer together when subjected to compression.
Gases, then, are compressible fluids.
A third difference is that when liquids of differing densities are mixed in a vessel, they will separate
out into discrete layers by gravitational settlement with the densest liquid at the bottom. This is not
true of gases. In this case, layering of the gases will take place only while the constituent gases
remain unmixed (for example, see Methane Layering, Section 12.4.2). If, however, the gases
become mixed into a homogenous blend, then the relatively high molecular velocities and large
intermolecular distances prevent the gases from separating out by gravitational settlement. The
internal molecular energy provides an effective continuous mixing process.
Subsurface ventilation engineers need to be aware of the properties of both liquids and gases. In
this chapter, we shall confine ourselves to incompressible fluids. Why is this useful when we are well
aware that a ventilation system is concerned primarily with air, a mixture of gases and, therefore,
compressible? The answer is that in a majority of mines and other subsurface facilities, the ranges of
temperature and pressure are such that the variation in air density is fairly limited. Airflow
measurements in mines are normally made to within 5 per cent accuracy. A 5 per cent change in air
density occurs by moving through a vertical elevation of some 500 metres in the gravitational field at
the surface of the earth. Hence, the assumption of incompressible flow with its simpler analytical
relationships gives acceptable accuracy in most cases. For the deeper and (usually) hotter facilities,
the effects of pressure and temperature on air density should be taken into account through
thermodynamic analyses if a good standard of accuracy is to be attained. The principles of physical
steady-flow thermodynamics are introduced in Chapter 3.
2-2
2.1.2 Volume flow, Mass flow and the Continuity Equation

Most measurements of airflow in ventilation systems are based on the volume of air (m3) that passes
through a given cross section of a duct or airway in unit time (1 second). The units of volume flow, Q,
are, therefore, m3/s. However, for accurate analyses when density variations are to be taken into
account, it is preferable to work in terms of mass flow - that is, the mass of air (kg) passing through
the cross section in 1 second. The units of mass flow, M, are then kg/s.
The relationship between volume flow and mass flow follows directly from the definition of density, ,
kg
mass
=
(2.1)
volume
m3
and
mass flow M
kg s
=
=
volume flow Q
s m3
giving M = Q kg/s
(2.2)
In any continuous duct or airway, the mass flows passing through all cross sections along its length
are equal, provided that the system is at steady state and there are no inflows or outflows of air or
other gases between the two ends. If these conditions are met then
M =Q = constant
kg/s
(2.3)
This is the simplest form of the Continuity Equation. It can, however, be written in other ways. A
common method of measuring volume flow is to determine the mean velocity of air, u, over a given
cross section, then multiply by the area of that cross-section, A, (Chapter 6):
m3
m
m 2 or
s
s
Then the continuity equation becomes
Q = uA
(2.4)
M = u A = constant kg/s
(2.5)
As indicated in the preceding subsection, we can achieve acceptable accuracy in most situations
within ventilation systems by assuming a constant density. The continuity equation then simplifies
back to
Q= uA
= constant
m3/s
(2.6)
This shows that for steady-state and constant density airflow in a continuous airway, the velocity of
the air varies inversely with cross sectional area.
2.2 FLUID PRESSURE

2.2.1 The cause of fluid pressure
Section 2.1.1 described the dynamic behaviour of molecules in a liquid or gas. When a molecule
rebounds from any confining boundary, a force equal to the rate of change of momentum of that
molecule is exerted upon the boundary. If the area of the solid/fluid boundary is large compared to
the average distance between molecular collisions then the statistical effect will be to give a uniform
force distributed over that boundary. This is the case in most situations of importance in subsurface
ventilation engineering.
2-3
Two further consequences arise from the bombardment of a very large number of molecules on a
surface, each molecule behaving essentially as a perfectly elastic sphere. First, the force exerted by
a static fluid will always be normal to the surface. We shall discover later that the situation is rather
different when the dynamic forces of a moving fluid stream are considered (Section 2.3). Secondly,
at any point within a static fluid, the pressure is the same in all directions. Hence, static pressure is a
scalar rather than a vector quantity.
Pressure is sometimes carelessly confused with force or thrust. The quantitative definition of
pressure, P, is clear and simple
P =
Force
Area
(2.7)
m2
In the SI system of units, force is measured in Newtons (N) and area in square metres. The resulting
unit of pressure, the N/m2, is usually called a Pascal (Pa) after the French philosopher, Blaise
Pascal (1623-1662).
2.2.2 Pressure head
If a liquid of density is poured into a vertical tube of cross-sectional area, A, until the level reaches
a height h, the volume of liquid is
volume = h A
m3
Then from the definition of density (mass/volume), the mass of the liquid is
mass = volume x density
= hA
kg
The weight of the liquid will exert a force, F, on the base of the tube equal to
mass x gravitational acceleration (g)
F = hAg
But as pressure = force/area, the pressure on the base of the tube is

P =
F
= gh
A
N
m2
or Pa
(2.8)
Hence, if the density of the liquid is known, and assuming a constant value for g, then the pressure
may be quoted in terms of h, the head of liquid. This concept is used in liquid type manometers
(Section 2.2.4) which, although in declining use, are likely to be retained for many purposes owing to
their simplicity.
Equation (2.8) can also be used for air and other gases. In this case, it should be remembered that
the density will vary with height. A mean value may be used with little loss in accuracy for most mine
shafts. However, here again, it is recommended that the more precise methodologies of
thermodynamics be employed for elevation differences of more than 500 m.
2-4
2.2.3 Atmospheric pressure and gauge pressure

The blanket of air that shrouds the earth extends to approximately 40 km above the surface. At that
height, its pressure and density tend towards zero. As we descend towards the earth, the number of
molecules per unit volume increases, compressed by the weight of the air above. Hence, the
pressure of the atmosphere also increases. However, the pressure at any point in the lower
atmosphere is influenced not only by the column of air above it but also by the action of convection,
wind currents and variations in temperature and water vapour content. Atmospheric pressure near
the surface, therefore, varies with both place and time. At the surface of the earth, atmospheric
pressure is of the order of 100 000 Pa. For practical reference this is often translated into 100 kPa
although the basic SI units should always be used in calculations. Older units used in meteorology
for atmospheric pressure are the bar (105 Pa) and the millibar (100 Pa).
For comparative purposes, reference is often made to standard atmospheric pressure. This is the
pressure that will support a 0.760 m column of mercury having a density of 13.5951 x 103 kg/m3 in a
standard earth gravitational field of 9.8066 m/s2.
Then from equation (2.8)
One Standard Atmosphere = x g x h
= 13.5951 x 103 x 9.8066 x 0.760
= 101.324 x 103
or
101.324
Pa
kPa.
The measurement of variations in atmospheric pressure is important during ventilation surveys

(Chapter 6), for psychrometric measurements (Chapter 14), and also for predicting the emission of
stored gases into a subsurface ventilation system (Chapter 12). However, for many purposes, it is
necessary to measure differences in pressure. One common example is the difference between the
pressure within a system such as a duct and the exterior atmosphere pressure. This is referred to as
gauge pressure..
Absolute pressure = Atmospheric pressure + gauge pressure
(2.9)
If the pressure within the system is below that of the local ambient atmospheric pressure then the
negative gauge pressure is often termed the suction pressure or vacuum and the sign ignored.
Care should be taken when using equation 2.9 as the gauge pressure may be positive or negative.
However, the absolute pressure is always positive. Although many quoted measurements are
pressure differences, it is the absolute pressures that are used in thermodynamic calculations. We
must not forget to convert when necessary.
2.2.4. Measurement of air pressure.
2.2.4.1. Barometers
Equation (2.8) showed that the pressure at the bottom of a column of liquid is equal to the product of
the head (height) of the liquid, its density and the local value of gravitational acceleration. This
principle was employed by Evangelista Torricelli (1608-1647), the Italian who invented the mercury
barometer in 1643.. Torricelli poured mercury into a glass tube, about one metre in length, closed at
one end, and upturned the tube so that the open end dipped into a bowl of mercury. The level in the
tube would then fall until the column of mercury, h, produced a pressure at the base that just
balanced the atmospheric pressure acting on the open surface of mercury in the bowl.
2-5
The atmospheric pressure could then be calculated as (see equation (2.8) )

P=gh
Pa
where, in this case, is the density of mercury.

Modern versions of the Torricelli instrument are still used as standards against which other types of
barometer may be calibrated. Barometric (atmospheric) pressures are commonly quoted in
millimetres (or inches) of mercury. However, for precise work, equation (2.8) should be employed
using the density of mercury corresponding to its current temperature. Accurate mercury barometers
have a thermometer attached to the stem of the instrument for this purpose and a sliding micrometer
to assist in reading the precise height of the column. Furthermore, and again for accurate work, the
local value of gravitational acceleration should be ascertained as this depends upon latitude and
altitude. The space above the mercury in the barometer will not be a perfect vacuum as it contains
mercury vapour. However, this exerts a pressure of less than 0.00016 kPa at 20 C and is quite
negligible compared with the surface atmospheric pressure of near 100 kPa. This, coupled with the
fact that the high density of mercury produces a barometer of reasonable length, explains why
mercury rather than any other liquid is used. A water barometer would need to be about 10.5m in
height.
Owing to their fragility and slowness in reacting to temperature changes, mercury barometers are
unsuitable for underground surveys . An aneroid barometer consists of a closed vessel which has
been evacuated to a near perfect vacuum. One or more elements of the vessel are flexible. These
may take the form of a flexing diaphragm, or the vessel itself may be shaped as a helical or spiral
spring. The near zero pressure within the vessel remains constant. However, as the surrounding
atmospheric pressure varies, the appropriate element of the vessel will flex. The movement may be
transmitted mechanically, magnetically or electrically to an indicator and/or recorder.
Low cost aneroid barometers may be purchased for domestic or sporting use. Most altimeters are, in
fact, aneroid barometers calibrated in metres (or feet) head of air. For the high accuracy required in
ventilation surveys (Chapter 6) precision aneroid barometers are available.
Another principle that can be employed in pressure transducers, including barometers, is the
piezoelectric property of quartz. The natural frequency of a quartz beam varies with the applied
pressure. As electrical frequency can be measured with great precision, this allows the pressure to
be determined with good accuracy.
2.2.4.2. Differential pressure instruments
Differences in air pressure that need to be measured frequently in subsurface ventilation
engineering rarely exceed 7 or 8 kPa and are often of the order of only a few Pascals. The traditional
instrument for such low pressure differences is the manometer. This relies upon the displacement of
liquid to produce a column, or head, that balances the differential pressure being measured. The
most rudimentary manometer is the simple glass U tube containing water, mercury or other liquid. A
pressure difference applied across the ends of the tube causes the liquid levels in the two limbs to
be displaced in opposite directions. A scale is used to measure the vertical distance between the
levels and equation (2.8) used to calculate the required pressure differential. Owing to the past
widespread use of water manometers, the millimetre (or inch) of water column came to be used
commonly as a measure of small pressure differentials, much as a head of mercury has been used
for atmospheric pressures. However, it suffers from the same disadvantages in that it is not a
primary unit but depends upon the liquid density and local gravitational acceleration.
When a liquid other than water is used, the linear scale may be increased or decreased, dependent
upon the density of the liquid, so that it still reads directly in head of water. A pressure head in one
2-6
fluid can be converted to a head in any other fluid provided that the ratio of the two densities is
known.
p = 1 g h1
or
h2 =
1
h
2 1
2 g h2
m
Pa
(2.10)
For high precision, the temperature of the liquid in a manometer should be obtained and the
corresponding density determined. Equation (2.10) is then used to correct the reading, h1 where 1 is
the actual liquid density and 2 is the density at which the scale is calibrated.
Many variations of the manometer have been produced. Inclining one limb of the U tube shortens its
practicable range but gives greater accuracy of reading. Careful levelling of inclined manometers is
required and they are no longer used in subsurface pressure surveys. Some models have one limb
of the U tube enlarged into a water reservoir. The liquid level in the reservoir changes only slightly
compared with the balancing narrow tube. In the direct lift manometer, the reservoir is connected by
flexible tubing to a short sight-glass of variable inclination which may be raised or lowered against a
graduated scale. This manipulation enables the meniscus to be adjusted to a fixed mark on the
sight-glass. Hence the level in the reservoir remains unchanged. The addition of a micrometer scale
gives this instrument both a good range and high accuracy.
One of the problems in some water manometers is a misformed meniscus, particularly if the
inclination of the tube is less than 5 degrees from the horizontal. This difficulty may be overcome by
employing a light oil, or other liquid that has good wetting properties on glass. Alternatively, the two
limbs may be made large enough in diameter to give horizontal liquid surfaces whose position can
be sensed electronically or by touch probes adjusted through micrometers.
U tube manometers, or water gauges as they are commonly known, may feature as part of the
permanent instrumentation of main and booster fans. Provided that the connections are kept firm
and clean, there is little that can go wrong with these devices. Compact and portable inclined gauges
are available for rapid readings of pressure differences across doors and stoppings in underground
ventilation systems. However, in modern pressure surveying (Chapter 6) manometers have been
replaced by the diaphragm gauge. This instrument consists essentially of a flexible diaphragm,
across which is applied the differential pressure. The strain induced in the diaphragm is sensed
electrically, mechanically or by magnetic means and transmitted to a visual indicator or recorder.
In addition to its portability and rapid reaction, the diaphragm gauge has many advantages for the
subsurface ventilation engineer. First, it reflects directly a true pressure (force/area) rather than
indirectly through a liquid medium. Secondly, it reacts relatively quickly to changes in temperature
and does not require precise levelling. Thirdly, diaphragm gauges can be manufactured over a wide
variety of ranges. A ventilation survey team may typically carry gauges ranging from 0 - 100 Pa to 0 5 kPa (or to encompass the value of the highest fan pressure in the system). One disadvantage of
the diaphragm gauge is that its calibration may change with time and usage. Re-calibration against a
laboratory precision manometer is recommended prior to an important survey.
Other appliances are used occasionally for differential pressures in subsurface pressure surveys.
Piezoelectric instruments are likely to increase in popularity. The aerostat principle eliminates the
need for tubing between the two measurement points and leads to a type of differential barometer. In
this instrument, a closed and rigid air vessel is maintained at a constant temperature and is
connected to the outside atmospheres via a manometer or diaphragm gauge. As the inside of the
vessel remains at near constant pressure, any variations in atmospheric pressure cause a reaction
on the manometer or gauge. Instruments based on this principle require independent calibration as
slight movements of the diaphragm or liquid in the manometer result in the inside pressure not
remaining truly constant.
2-7
2.3 FLUIDS IN MOTION

2.3.1. Bernoulli's equation for ideal fluids
As a fluid stream passes through a pipe, duct or other continuous opening, there will, in general, be
changes in its velocity, elevation and pressure. In order to follow such changes it is useful to identify
the differing forms of energy contained within a given mass of the fluid. For the time being, we will
consider that the fluid is ideal; that is, it has no viscosity and proceeds along the pipe with no shear
forces and no frictional losses. Secondly, we will ignore any thermal effects and consider mechanical
energy only.
Suppose we have a mass, m, of fluid moving at velocity, u, at an elevation, Z, and a barometric
pressure P. There are three forms of mechanical energy that we need to consider. In each case, we
shall quantify the relevant term by assessing how much work we would have to do in order to raise
that energy quantity from zero to its actual value in the pipe, duct or airway.
Kinetic energy
If we commence with the mass, m, at rest and accelerate it to velocity u in t seconds by applying a
constant force F, then the acceleration will be uniform and the mean velocity is
0 +u
u
=
2
2
m
s
Then
distance travelled = mean velocity x time
=
u
t
2
Furthermore, the acceleration is defined as

increase in velocity
u
=
time
t
m / s2
The force is given by

F = mass x acceleration
= m
u
t
and the work done to accelerate from rest to velocity u is

WD = force x distance
u
u
x t
2
t
2
u
=m
2
Nm
=m
Nm or J
(2.11)
The kinetic energy of the mass m is, therefore, m u2/2 Joules.
2-8
Potential energy
Any base elevation may be used as the datum for potential energy. In most circumstances of
underground ventilation engineering, it is differences in elevation that are important. If our mass m is
located on the base datum then it will have a potential energy of zero relative to that datum. We then
exert an upward force, F, sufficient to counteract the effect of gravity.
F = mass x acceleration
=mg
where g is the gravitational acceleration.

In moving upward to the final elevation of Z metres above the datum, the work done is
WD = Force x distance
= mgZ
Joules
(2.12)
This gives the potential energy of the mass at elevation Z.

Flow work
Suppose we have a horizontal pipe, open at both ends and of cross sectional area A as shown in
Figure 2.1. We wish to insert a plug of fluid, volume v and mass m into the pipe. However, even in
the absence of friction, there is a resistance due to the pressure of the fluid, P, that already exists in
the pipe. Hence, we must exert a force, F, on the plug of fluid to overcome that resisting pressure.
Our intent is to find the work done on the plug of fluid in order to move it a distance s into the pipe.
s
F
Figure 2.1 Flow work done on a fluid entering a pipe

The force, F, must balance the pressure, P, which is distributed over the area, A.
F=PA
Work done = force x distance

=
PAs
J or Joules
However, the product As is the swept volume v, giving

WD = P v
Now, by definition, the density is
m
=
v
or
m
v =
kg
m3
2-9
Hence, the work done in moving the plug of fluid into the pipe is
WD =
or
Pm
(2.13)
P/ Joules per kilogram.
As fluid continues to be inserted into the pipe to produce a continuous flow, then each individual
plug must have this amount of work done on it. That energy is retained within the fluid stream and is
known as the flow work. The appearance of pressure, P, within the expression for flow work has
resulted in the term sometimes being labelled "pressure energy". This is very misleading as flow
work is entirely different to the "elastic energy" stored when a closed vessel of fluid is compressed.
Some authorities also object to the term "flow work" and have suggested "convected energy" or,
simply, the "Pv work". Note that in Figure 2.1 the pipe is open at both ends. Hence the pressure, P,
inside the pipe does not change with time (the fluid is not compressed) when plugs of fluid continue
to be inserted in a frictionless manner. When the fluid exits the system, it will carry kinetic and
potential energy, and the corresponding flow work with it.
Now we are in a position to quantify the total mechanical energy of our mass of fluid, m. From
expressions (2.11, 2.12 and 2.13)
total mechanical
energy
kinetic
energy
mu 2
2
potential
energy
mZ g
flow
+ work
+ m
(2.14)
If no mechanical energy is added to or subtracted from the fluid during its traverse through the pipe,
duct or airway, and in the absence of frictional effects, the total mechanical energy must remain
constant throughout the airway. Then equation (2.14) becomes
u 2
P
m
+ Zg +
= constant
(2. 15)
Another way of expressing this equation is to consider two stations, 1 and 2 along the pipe, duct or
airway. Then
u 2
u 2
P
P
m 1 + Z1 g + 1 = m 2 + Z 2 g + 2
1
2
2
2
Now as we are still considering the fluid to be incompressible (constant density),

giving
1 = 2 =
(say)
u12 u 22
P P
+ (Z 1 Z 2 )g + 1 2
2
= 0
J
kg
(2.16)
Note that dividing by m on both sides has changed the units of each term from J to J/kg.
Furthermore, if we multiplied throughout by then each term would take the units of pressure.
Bernoulli's equation has, traditionally, been expressed in this form for incompressible flow.
2 - 10
Equation (2.16) is of fundamental importance in the study of fluid flow. It was first derived by Daniel
Bernoulli (1700-1782), a Swiss mathematician, and is known throughout the world by his name.
As fluid flows along any closed system, Bernoulli's equation allows us to track the inter-relationships
between the variables. Velocity u, elevation Z, and pressure P may all vary, but their combination as
expressed in Bernoulli's equation remains true. It must be remembered, however, that it has been
derived here on the assumptions of ideal (frictionless) conditions, constant density and steady-state
flow. We shall see later how the equation must be amended for the real flow of compressible fluids.
2.3.2. Static, total and velocity pressures.
Consider the level duct shown on Figure 2.2. Three gauge pressures are measured. To facilitate
visualization, the pressures are indicated as liquid heads on U tube manometers. However, the
analysis will be conducted in terms of true pressure (N/m2) rather than head of fluid.
u1
(a)
(c)
(b)
ps
pt
pv
Figure 2.2 (a) static, (b) total and (c) velocity pressures
In position (a), one limb of the U tube is connected perpendicular through the wall of the duct. Any
drilling burrs on the inside have been smoothed out so that the pressure indicated is not influenced
by the local kinetic energy of the air. The other limb of the manometer is open to the ambient
atmosphere. The gauge pressure indicated is known as the static pressure, ps.
In position (b) the left tube has been extended into the duct and its open end turned so that it faces
directly into the fluid stream. As the fluid impacts against the open end of the tube, it is brought to
rest and the loss of its kinetic energy results in a local increase in pressure. The pressure within the
tube then reflects the sum of the static pressure and the kinetic effect. Hence the manometer
indicates a higher reading than in position (a).The corresponding pressure, pt, is termed the total
pressure. The increase in pressure caused by the kinetic energy can be quantified by using
Bernoulli's equation (2.16). In this case Zl = Z2, and u2 = 0. Then
P2 P1 u12
=
2
The local increase in pressure caused by bringing the fluid to rest is then
pv = P2 P1 =
u12
2
Pa
2 - 11
This is known as the velocity pressure and can be measured directly by connecting the manometer
as shown in position (c). The left connecting tube of the manometer is at gauge pressure pt and the
right tube at gauge pressure ps. It follows that
pv = pt - p s
or
p t = ps + p v
Pa
(2.18)
In applying this equation, care should be taken with regard to sign as the static pressure, ps, will be
negative if the barometric pressure inside the duct is less than that of the outside atmosphere.
If measurements are actually made using a liquid in glass manometer as shown on Figure 2.2 then
the reading registered on the instrument is influenced by the head of fluid in the manometer tubes
above the liquid level. If the manometer liquid has a density 1, and the superincumbent fluid in both
tubes has a density d, then the head indicated by the manometer, h, should be converted to true
pressure by the equation
p = ( 1 d ) g h
(2.19)
Pa
Reflecting back on equation (2.8) shows that this is the usual equation relating fluid head and
pressure with the density replaced by the difference in the two fluid densities. In ventilation
engineering, the superincumbent fluid is air, having a very low density compared with liquids. Hence,
the d term in equation (2.19) is usually neglected. However, if the duct or pipe contains a liquid
rather than a gas then the full form of equation (2.19) should be employed.
A further situation arises when the fluid in the duct has a density, d, that is significantly different to
that of the air (or other fluid), a, which exists above the liquid in the right hand tube of the
manometer in Fig. 2.2(a). Then
p = ( 1 d ) g h ( d a ) g h2
Pa
(2.20)
where h2 is the vertical distance between the liquid level in the right side of the manometer and the
connection into the duct.
Equations (2.19) and (2.20) can be derived by considering a pressure balance on the two sides of
the U tube above the lower of the two liquid levels.
2.3.3. Viscosity
Bernoulli's equation was derived in Section 2.3.1. on the assumption of an ideal fluid; i.e. that flow
could take place without frictional resistance. In subsurface ventilation engineering almost all of the
work input by fans (or other ventilating devices) is utilized against frictional effects within the airways.
Hence, we must find a way of amending Bernoulli's equation for the frictional flow of real fluids.
The starting point in an examination of 'frictional flow' is the concept of viscosity. Consider two
parallel sheets of fluid a very small distance, dy, apart but moving at different velocities u and
u + du (Figure 2.3). An equal but opposite force, F, will act upon each layer, the higher velocity sheet
tending to pull its slower neighbour along and, conversely, the slower sheet tending to act as a brake
on the higher velocity layer.
2 - 12
u + du
F
dy
F
Figure 2.3 Viscosity causes equal but opposite forces to be exerted

on adjacent laminae of fluid.
If the area of each of the two sheets in near contact is A, then the shear stress is defined as
(Greek 'tau') where
F
A
(2.21)
m2
Among his many accomplishments, Isaac Newton (1642-1727) proposed that for parallel motion of
streamlines in a moving fluid, the shear stress transmitted across the fluid in a direction
perpendicular to the flow is proportional to the rate of change of velocity, du/dy (velocity gradient)
F
du
=
A
dy
(2.22)
m2
where the constant of proportionality, , is known as the coefficient of dynamic viscosity (usually
referred to simply as dynamic viscosity). The dynamic viscosity of a fluid varies with its temperature.
For air, it may be determined from
Ns
(17.0 + 0.045 t) x 10-6
air =
m2
and for water

64.72
0.2455 x 10 3
water =
t + 31.766
Ns
m2
where t = temperature (C) in the range 0 - 60 C

The units of viscosity are derived by transposing equation (2.22)
=
dy
du
N
m
s
m
or
Ns
m2
A term which commonly occurs in fluid mechanics is the ratio of dynamic viscosity to fluid density.
This is called the kinematic viscosity, (Greek 'nu')
Ns m 3
s
or N m
2 kg
kg
m
As 1 N = 1 kg x 1 m/s2, these units become

m ms
m2
=
kg 2
s
s kg
2 - 13
It is the transmission of shear stress that produces frictional resistance to motion in a fluid stream.
Indeed, a definition of an 'ideal fluid' is one that has zero viscosity. Following from our earlier
discussion on the molecular behaviour of fluids (Section 2.1.1.), there would appear to be at least
two effects that produce the phenomenon of viscosity. One is the attractive forces that exist between
molecules - particularly those of liquids. This will result in the movement of some molecules tending
to drag others along, and for the slower molecules to inhibit motion of faster neighbours. The second
effect may be visualized by glancing again at Figure 2.3. If molecules from the faster moving layer
stray sideways into the slower layer then the inertia that they carry will impart kinetic energy to that
layer. Conversely, migration of molecules from the slower to the faster layer will tend to retard its
motion.
In liquids, the molecular attraction effect is dominant. Heating a liquid increases the internal kinetic
energy of the molecules and also increases the average inter-molecular spacing. Hence, as the
attractive forces diminish with distance, the viscosity of a liquid decreases with respect to
temperature. In a gas, the molecular attractive force is negligible. The viscosity of gases is much
less than that of liquids and is caused by the molecular inertia effect. In this case, the increased
velocity of molecules caused by heating will tend to enhance their ability to transmit inertia across
streamlines and, hence, we may expect the viscosity of gases to increase with respect to
temperature. This is, in fact, the situation observed in practice.
In both of these explanations of viscosity, the effect works between consecutive layers equally well
in both directions. Hence, dynamic equilibrium is achieved with both the higher and lower velocity
layers maintaining their net energy levels. Unfortunately, no real process is perfect in fluid
mechanics. Some of the useful mechanical energy will be transformed into the much less useful heat
energy. In a level duct, pipe or airway, the loss of mechanical energy is reflected in an observable
drop in pressure. This is often termed the 'frictional pressure drop'
Recalling that Bernoulli's equation was derived for mechanical energy terms only in Section 2.3.1, it
follows that for the flow of real fluids, the equation must take account of the frictional loss of
mechanical energy. We may rewrite equation (2.16) as
u12
P
u2
P
+ Z 1 g + 1 = 2 + Z 2 g + 2 + F12
2
2
J
kg
(2.23)
where Fl2 = energy converted from the mechanical form to heat (J/kg).
The problem now turns to one of quantifying the frictional term F12. For that, we must first examine
the nature of fluid flow.
2.3.4. Laminar and turbulent flow. Reynolds Number
In our everyday world, we can observe many examples of the fact that there are two basic kinds of
fluid flow. A stream of oil poured out of a can flows smoothly and in a controlled manner while water,
poured out at the same rate, would break up into cascading rivulets and droplets. This example
seems to suggest that the type of flow depends upon the fluid. However, a light flow of water falling
from a circular outlet has a steady and controlled appearance, but if the flowrate is increased the
stream will assume a much more chaotic form. The type of flow seems to depend upon the flowrate
as well as the type of fluid.
Throughout the nineteenth century, it was realized that these two types of flow existed. The German
engineer G.H.L. Hagen (1797-1884) found that the type of flow depended upon the velocity and
viscosity of the fluid. However, it was not until the 1880's that Professor Osborne Reynolds of
Manchester University in England established a means of characterizing the type of flow regime
2 - 14
through a combination of experiments and logical reasoning. Reynolds' laboratory tests consisted of
injecting a filament of colored dye into the bell mouth of a horizontal glass tube that was submerged
in still water within a large glass-walled tank. The other end of the tube passed through the end of
the tank to a valve which was used to control the velocity of water within the tube. At low flow rates,
the filament of dye formed an unbroken line in the tube without mixing with the water. At higher flow
rates the filament of dye began to waver. As the velocity in the tube continued to be increased the
wavering filament suddenly broke up to mix almost completely with the water.
In the initial type of flow, the water appeared to move smoothly along streamlines, layers or laminae,
parallel to the axis of the tube. We call this laminar flow. Appropriately, we refer to the completely
mixing type of behavior as turbulent flow. Reynolds' experiments had, in fact, identified a third
regime - the wavering filament indicated a transitional region between fully laminar and fully turbulent
flow. Another observation made by Reynolds was that the break-up of the filament always occurred,
not at the entrance, but about thirty diameters along the tube.
The essential difference between laminar and turbulent flow is that in the former, movement across
streamlines is limited to the molecular scale, as described in Section 2.3.3. However, in turbulent
flow, swirling packets of fluid move sideways in small turbulent eddies. These should not be
confused with the larger and more predictable oscillations that can occur with respect to time and
position such as the vortex action caused by fans, pumps or obstructions in the airflow. The turbulent
eddies appear random in the complexity of their motion. However, as with all "random" phenomena,
the term is used generically to describe a process that is too complex to be characterized by current
mathematical knowledge. Computer simulation packages using techniques known generically as
computational fluid dynamics (CFD) have produced powerful means of analysis and predictive
models of turbulent flow. At the present time, however, many practical calculations involving
turbulent flow still depend upon empirical factors.
The flow of air in the vast majority of 'ventilated' places underground is turbulent in nature. However,
the sluggish movement of air or other fluids in zones behind stoppings or through fragmented strata
may be laminar. It is, therefore, important that the subsurface ventilation engineer be familiar with
both types of flow. Returning to Osborne Reynolds, he found that the development of full turbulence
depended not only upon velocity, but also upon the diameter of the tube. He reasoned that if we
were to compare the flow regimes between differing geometrical configurations and for various fluids
we must have some combination of geometric and fluid properties that quantified the degree of
similitude between any two systems. Reynolds was also familiar with the concepts of "inertial
(kinetic) force", u2/2 (Newtons per square metre of cross section) and "viscous force",
= du / dy (Newtons per square metre of shear surface). Reynolds argued that the dimensionless
ratio of "inertial forces" to "viscous forces" would provide a basis of comparing fluid systems
inertial force
u2
=
viscous force
2
1 dy
du
(2.24)
Now, for similitude to exist, all steady state velocities, u, or differences in velocity between locations,
du, within a given system are proportional to each other. Furthermore, all lengths are proportional to
any chosen characteristic length, L. Hence, in equation (2.24) we can replace du by u, and dy by L.
The constant, 2, can also be dropped as we are simply looking for a combination of variables that
characterize the system. That combination now becomes
u2
1 L
u
or
u L
= Re
(2.25)
2 - 15
As equation (2.24) is dimensionless then so, also, must this latter expression be dimensionless. This
can easily be confirmed by writing down the units of the component variables. The result we have
reached here is of fundamental importance to the study of fluid flow. The dimensionless group uL/
is known universally as Reynolds Number, Re. In subsurface ventilation engineering, the
characteristic length is normally taken to be the hydraulic mean diameter of an airway, d, and the
characteristic velocity is usually the mean velocity of the airflow. Then
Re =
u d
At Reynolds Numbers of less than 2 000 in fluid flow systems, viscous forces prevail and the flow will
be laminar. The Reynolds Number over which fully developed turbulence exists is less well defined.
The onset of turbulence will occur at Reynolds Numbers of 2 500 to 3 000 assisted by any vibration,
roughness of the walls of the pipe or any momentary perturbation in the flow.
Example
A ventilation shaft of diameter 5m passes an airflow of 200 m3/s at a mean density of 1.2 kg/m3 and
an average temperature of 18 C. Determine the Reynolds Number for the shaft.
Solution
For air at 18 C
= (17.0 + 0.045 x 18) x 10-6
= 17.81 x 10-6 Ns/m2
Air velocity, u =
Q
200
=
A 52 / 4
Re =
= 10.186 m/s
1.2 10.186 5
u d
=
= 3.432 10 6
6
17.81 10
This Reynolds Number indicates that the flow will be turbulent.

2.3.5. Frictional losses in laminar flow, Poiseuille's Equation.
Now that we have a little background on the characteristics of laminar and turbulent flow, we can
return to Bernoulli's equation corrected for friction (equation (2.23)) and attempt to find expressions
for the work done against friction, F12. First, let us deal with the case of laminar flow.
Consider a pipe of radius R as shown in Figure 2.4. As the flow is laminar, we can imagine
concentric cylinders of fluid telescoping along the pipe with zero velocity at the walls and maximum
velocity in the center. Two of these cylinders of length L and radii r and r + dr are shown. The
velocities of the cylinders are u and u - du respectively.
2 - 16
u - du
dr
P-p
dr
L
Figure 2.4 Viscous drag opposes the motive effect of applied pressure difference
The force propagating the inner cylinder forward is produced by the pressure difference across its
two ends, p, multiplied by its cross sectional area, r 2 . This force is resisted by the viscous drag of
the outer cylinder, , acting on the 'contact' area 2 rL. As these forces must be equal at steady
state conditions,
2 r L = r 2 p
However, =
du
dr
(equation (2.22) with a negative du)
giving
or
du
r p
=
dr
2 L
du =
p r
dr
L 2
m
s
(2.26)
For a constant diameter tube, the pressure gradient along the tube p/L is constant. So, also, is for
the Newtonian fluids that we are considering. (A Newtonian fluid is defined as one in which viscosity
is independent of velocity). Equation (2.26) can, therefore, be integrated to give
p 1 r2
u =
+C
(2.27)
L 2 2
At the wall of the tube, r = R and u = 0. This gives the constant of integration to be
C =
p R2
L 4
Substituting back into equation (2.27) gives

u =
1 p 2
(R r 2 )
4 L
m
s
(2.28)
Equation (2.28) is a general equation for the velocity of the fluid at any radius and shows that the
velocity profile across the tube is parabolic (Figure 2.5). Along the centre line of the tube, r = 0 and
the velocity reaches a maximum of
2 - 17
1 p 2
R
4 L
u max =
m
s
(2.29)
umax
r
Figure 2.5 The velocity profile for laminar flow is parabolic

The velocity terms in the Bernoulli equation are mean velocities across the relevant cross-sections. It
is, therefore, preferable that the work done against viscous friction should also be expressed in
terms of a mean velocity, um. We must be careful how we define mean velocity in this context. Our
convention is to determine it as
m
s
Q
A
um =
(2.30)
where Q = volume airflow (m3/s) and A = cross sectional area (m2)

We could define another mean velocity by integrating the parabolic equation (2.28) with respect to r
and dividing the result by R. However, this would not take account of the fact that the volume of fluid
in each concentric shell of thickness dr increases with radius. In order to determine the true mean
velocity, consider the elemental flow dQ through the annulus of cross sectional area 2 r dr at
radius r and having a velocity of u (Figure 2.4)
dQ = u 2r dr
Substituting for u from equation (2.28) gives

2 p
(R 2 r 2 ) r dr
4 L
dQ =
Q =
2
4
p
L
(R 2 r r 3 ) dr
Integrating gives
Q =
R4 p
8 L
m
s
(2.31)
This is known as the Poiseuille Equation or, sometimes, the Hagen-Poiseuille Equation. J.L.M.
Poiseuille (1799-1869) was a French physician who studied the flow of blood in capillary tubes.
2 - 18
For engineering use, where the dimensions of a given pipe and the viscosity of fluid are known,
Poiseuille's equation may be written as a pressure drop - quantity relationship.
8L
p =
R4
or
p = RL Q
where
RL =
8L
Ns
m5
(2.32)
Pa
and is known as the laminar resistance of the pipe.
Equation (2.32) shows clearly that in laminar flow the frictional pressure drop is proportional to the
volume flow for any given pipe and fluid. Combining equations (2.30) and (2.31) gives the required
mean velocity
um =
R4 p
1
8 L R2
R2 p
8 L
m
s
(2.33)
or
p =
8 um
R2
(2.34)
Pa
This latter form gives another expression for the frictional pressure drop in laminar flow.
To see how we can use this equation in practice, let us return the frictional form of Bernoulli's
equation
(P P2 )
u12 u 22
+ (Z1 Z 2 )g + 1
= F12
2
J
(see equation (2.23) )
kg
Now for incompressible flow along a level pipe of constant cross-sectional area,
Z1 = Z2 and u1 = u2 = um
then
(P1 P2 )
= F12
J
kg
(2.35)
However, (P1 - P2) is the same pressure difference as p in equation (2.34).

Hence the work done against friction is
F12 =
8 um
J
kg
(2.36)
Bernoulli's equation for incompressible laminar frictional flow now becomes

(P P2 )
8 um
u12 u 22
+ (Z1 Z 2 )g + 1
=
L
2
R2
J
kg
(2.37)
2 - 19
If the pipe is of constant cross sectional area, then u1 = u2 = um and the kinetic energy term
disappears. On the other hand, if the cross-sectional area and, hence, the velocity varies along the
pipe then um may be established as a weighted mean. For large changes in cross-sectional area, the
full length of pipe may be subdivided into increments for analysis.
Example.
A pipe of diameter 2 cm rises through a vertical distance of 5m over the total pipe length of 2 000 m.
Water of mean temperature 15C flows up the tube to exit at atmospheric pressure of 100 kPa. If the
required flowrate is 1.6 litres per minute, find the resistance of the pipe, the work done against
friction and the head of water that must be applied at the pipe entrance.
Solution.
It is often the case that measurements made in engineering are not in SI units. We must be careful
to make the necessary conversions before commencing any calculations.
Flowrate Q = 1.6 litres/min
1. 6
=
= 2.667 10 5
1000 60
m3
s
Cross sectional area of pipe A = d 2 / 4 = (0.02 2 ) / 4 = 3.142 10 4 m 2

Mean velocity, u =
2.667 10 5
Q
=
= 0.084 88
A
3.142 10 4
m/s
(We have dropped the subscript m. For simplicity, the term u from this point on will refer to the mean
velocity defined as Q/A)
Viscosity of water at 15 C (from Section 2.3.3.)
64.72
=
0.2455 10 3
15 + 31.766
= 1.138 10 3
Ns
m2
Before we can begin to assess frictional effects we must check whether the flow is laminar or
turbulent. We do this by calculating the Reynolds Number
ud
Re =
where = density of water (taken as 1 000 kg/m3)

Re =
1000 0.084 88 0.02

1.138 10 3
= 1 491
(dimensionless)
As Re is below 2 000, the flow is laminar and we should use the equations based on viscous friction.
Laminar resistance of pipe (from equation (2.32))
RL =
8L
8 1.1384 10 3 2 000
(0.01)
= 580 10 6
Ns
m5
2 - 20
Frictional pressure drop in the pipe (equation (2.32))

p = R L Q = 580 10 6 2.667 10 5
= 15 461
Pa
Work done against friction (equation (2.36))

F12 =
8 uL
R
8 1.1384 10 3 0.084 88 2000

1000 (0.01)
= 15.461
J
kg
This is the amount of mechanical energy transformed to heat in Joules per kilogram of water. Note
the similarity between the statements for frictional pressure drop, p, and work done against friction,
Fl2. We have illustrated, by this example, a relationship between p and Fl2 that will be of particular
significance in comprehending the behaviour of airflows in ventilation systems, namely
p
= F12
In fact, having calculated p as 15 461 Pa, the value of F12 may be quickly evaluated as
15 461
J
= 15.461
1000
kg
To find the pressure at the pipe inlet we may use Bernoulli's equation corrected for frictional effects
u12 u 22
P P2
+ (Z 1 Z 2 )g + 1
= F12
2
In this example
u1 = u 2
J
kg
Z1 Z 2 = 5 m
and
P2 = 100 kPa = 100 000 Pa
giving
F12 = 5 9.81 +
P1 100 000
= 15.461
1000
J
kg
This yields the absolute pressure at the pipe entry as

P1 = 164.5 10 3
or
164.5
Pa
kPa
If the atmospheric pressure at the location of the bottom of the pipe is also 100 kPa, then the gauge
pressure, pg, within the pipe at that same location
pg = 164.5 - 100 = 64.5 kPa
This can be converted into a head of water, h1, from equation (2.8)
p g = g h1
2 - 21
h1 =
64.5 10 3
= 6.576
1000 9.81
m of water
Thus, a header tank with a water surface maintained 6.576 m above the pipe entrance will produce
the required flow of 1.6 litres/minute along the pipe.
The experienced engineer would have determined this result quickly and directly after calculating the
frictional pressure drop to be 15 461 Pa. The frictional head loss
h
15 461
p
=
= 1.576
g
1000 9.81
m of water
The head of water at the pipe entrance must overcome the frictional head loss as well as the vertical
lift of 5 m. (An intuitive use of Bernoulli's equation). Then
h1 = 5 + 1.576 = 6.576
m of water
2.3.6. Frictional losses in turbulent flow

The previous section showed that the parallel streamlines of laminar flow and Newton's perception of
viscosity enabled us to produce quantitative relationships through purely analytical means.
Unfortunately, the highly convoluted streamlines of turbulent flow, caused by the interactions
between both localized and propagating eddies have so far proved resistive to completely analytical
techniques. Numerical methods using the memory capacities and speeds of supercomputers allow
the flow to be simulated as a large number of small packets of fluids, each one influencing the
behaviour of those around it. These mathematical models, using numerical techniques known
collectively as computational fluid dynamics (CFD), may be used to simulate turbulent flow in given
geometrical systems, or to produce statistical trends. However, the majority of engineering
applications involving turbulent flow still rely on a combination of analysis and empirical factors. The
construction of physical models for observation in wind tunnels or other fluid flow test facilities
remains a common means of predicting the behaviour and effects of turbulent flow.
2.3.6.1. The Chzy-Darcy Equation
The discipline of hydraulics was studied by philosophers of the ancient civilizations. However, the
beginnings of our present treatment of fluid flow owe much to the hydraulic engineers of eighteenth
and nineteenth century France. During his reign, Napolean Bonaparte encouraged the research and
development necessary for the construction of water distribution and drainage systems in Paris.
Antoine de Chzy (1719-1798) carried out a series of experiments on the river Seine and on canals
in about 1769. He found that the mean velocity of water in open ducts was proportional to the square
root of the channel gradient, cross sectional area of flow and inverse of the wetted perimeter.
u
where
A h
per L
h = vertical distance dropped by the channel in a length L (h/L = hydraulic gradient)

per = wetted perimeter (m)
and
means 'proportional to'
2 - 22
Inserting a constant of proportionality, c, gives

=
A h
per L
m
s
(2.38)
where c is known as the Chzy coefficient.

Equation (2.38) has become known as Chzy's equation for channel flow. Subsequent analysis shed
further light on the significance of the Chzy coefficient. When a fluid flows along a channel, a mean
shear stress is set up at the fluid/solid boundaries. The drag on the channel walls is then
per L
where per is the "wetted" perimeter

This must equal the pressure force causing the fluid to move, pA, where p is the difference in
pressure along length L.
per L = A p
(2.39)
(A similar equation was used in Section 2.3.5. for a circular pipe).

But
p = g h
giving
(equation (2.8))
Pa
A
h
g
per
L
(2.40)
m2
If the flow is fully turbulent, the shear stress or skin friction drag, , exerted on the channel walls is
also proportional to the inertial (kinetic) energy of the flow expressed in Joules per cubic metre.

or
u2
2
= f
J
m3
u2
2
Nm
m
or
N
m
(2.41)
m2
where f is a dimensionless coefficient which, for fully developed turbulence, depends only upon the
roughness of the channel walls.
Equating (2.40) and (2.41) gives
or
u2
A
h
g
=
per L
2
u =
2g
f
A h
per L
m
s
(2.42)
2 - 23
Comparing this with equation (2.38) shows that Chzy's coefficient, c, is related to the roughness of
the channel.
2g
f
c =
m2
s
(2.43)
The development of flow relationships was continued by Henri Darcy (1803-1858), another French
engineer, who was interested in the turbulent flow of water in pipes. He adapted Chzy 's work to the
case of circular pipes and ducts running full. Then A = d 2 / 4 , per = d and the fall in elevation
of Chzy's channel became the head loss, h (metres of fluid) along the pipe length L. Equation
(2.42) now becomes
2g d 2 1 h
f
4 d L
u2 =
or
h =
4f Lu2
2g d
(2.44)
metres of fluid
This is the well known Chzy-Darcy equation, sometimes also known simply as the Darcy equation
or the Darcy-Weisbach equation. The head loss, h, can be converted to a frictional pressure drop, p,
by the now familiar relationship, p = g h to give
p =
4f L u2
d
2
(2.45)
Pa
or a frictional work term

F12 =
p
4fL u 2
=
d 2
J
kg
(2.46)
The Bernoulli equation for frictional and turbulent flow becomes

u12 u 22
(P P2 )
4fL u 2
+ (Z 1 Z 2 )g + 1
=
d 2
2
J
kg
(2.47)
where u is the mean velocity.

The most common form of the Chzy-Darcy equation is that given as (2.44). Leaving the constant 2
uncancelled provides a reminder that the pressure loss due to friction is a function of kinetic energy
u2/2. However, some authorities have combined the 4 and the f into a different coefficient of friction
( = 4f) while others, presumably disliking Greek letters, then replaced the symbol by (would you
believe it?) , f. We now have a confused situation in the literature of fluid mechanics where f may
mean the original Chzy-Darcy coefficient of friction, or four times that value. When reading the
literature, care should be taken to confirm the nomenclature used by the relevant author. Throughout
this book, f is used to mean the original Chzy-Darcy coefficient as used in equation (2.44).
2 - 24
In order to generalize our results to ducts or airways of non-circular cross section, we may define a
hydraulic radius as
rh =
=
A
per
(2.48)
d 2
d
=
4d
4
Reference to the "hydraulic mean diameter" denotes 4A/per. This device works well for turbulent
flow but must not be applied to laminar flow where the resistance to flow is caused by viscous action
throughout the body of the fluid rather than concentrated around the perimeter of the walls.
Substituting for d in equation (2.45) gives
p = fL
per u 2
A 2
Pa
(2.49)
This can also be expressed as a relationship between frictional pressure drop, p, and volume flow,
Q. Replacing u by Q/A in equation (2.49) gives
or
fL per
Q 2
2 A3
p = Rt Q 2
where
Rt =
p =
fL per
2 A3
Pa
Pa
(2.50)
m 4
(2.51)
This is known as the rational turbulent resistance of the pipe, duct or airway and is a function only of
the geometry and roughness of the opening.
2.3.6.2. The coefficient of friction, f.
It is usually the case that a significant advance in research opens up new avenues of investigation
and produces a flurry of further activity. So it was following the work of Osborne Reynolds. During
the first decade of this century, fluid flow through pipes was investigated in great detail by engineers
such as Thomas E. Stanton (1865-1931) and J.R. Pannel in the United Kingdom, and Ludwig
Prandtl (1875-1953) in Germany. A major cause for concern was the coefficient of friction, f.
There were two problems. First, how could one predict the value of f for any given pipe without
actually constructing the pipe and conducting a pressure-flow test on it. Secondly, it was found that f
was not a true constant but varied with Reynolds Number for very smooth pipes and, particularly, at
low values of Reynolds Number. The latter is not too surprising as f was introduced initially as a
constant of proportionality between shear stress at the walls and inertial force of the fluid (equation
(2.41)) for fully developed turbulence. At the lower Reynolds Numbers we may enter the transitional
or even laminar regimes.
Figure 2.6 illustrates the type of results that were obtained. A very smooth pipe exhibited a
continually decreasing value of f. This is labelled as the turbulent smooth pipe curve. However, for
rougher pipes, the values of f broke away from the smooth pipe curve at some point and, after a
transitional region, settled down to a constant value, independent of Reynolds Number. This
phenomenon was quantified empirically through a series of classical experiments conducted in
Germany by Johann Nikuradse (1894-1979), a former student of Prandtl. Nikuradse took a number
of smooth pipes of diameter 2.5, 5 and 10 cm, and coated the inside walls uniformly with grains of
2 - 25
graded sand. The roughness of each tube was then defined as e/d where e was the diameter of the
sand grains and d the diameter of the tube. The advantages of dimensionless numbers had been
well learned from Reynolds. The corresponding f - Re relationships are illustrated on Figure 2.6.
Figure 2.6 Variation of f with respect to Re as found by Nikuradse

Figure 2.6 Variation of f with respect to Re as found by Nikuradse
Figure 2.6 variation of f with respect to Re as found by Nikuradse
The investigators of the time were then faced with an intriguing question. How could a pipe of given
roughness and passing a turbulent flow be "smooth" (i.e. follow the smooth pipe curve) at certain
Reynolds Numbers but become "rough" (constant f) at higher Reynolds Numbers? The answer lies
in our initial concept of turbulence - the formation and maintenance of small, interacting and
propagating eddies within the fluid stream. These necessitate the existence of cross velocities with
vector components perpendicular to the longitudinal axis of the tube. At the walls there can be no
cross velocities except on a molecular scale. Hence, there must be a thin layer close to each wall
through which the velocity increases from zero (actually at the wall) to some finite velocity sufficiently
far away from the wall for an eddy to exist. Within that thin layer the streamlines remain parallel to
each other and to the wall, i.e. laminar flow.
Although this laminar sublayer is very thin, it has a marked effect on the behaviour of the total flow in
the pipe. All real surfaces (even polished ones) have some degree of roughness. If the peaks of the
roughness, or asperities, do not protrude through the laminar sublayer then the surface may be
described as "hydraulically smooth" and the wall resistance is limited to that caused by viscous
shear within the fluid. On the other hand, if the asperities protrude well beyond the laminar sublayer
then the disturbance to flow that they produce will cause additional eddies to be formed, consuming
mechanical energy and resulting in a higher resistance to flow. Furthermore, as the velocity and,
2 - 26
hence, the Reynolds Number increases, the thickness of the laminar sublayer decreases. Any given
pipe will then be hydraulically smooth if the asperities are submerged within the laminar sublayer
and hydraulically rough if the asperities project beyond the laminar sublayer. Between the two
conditions there will be a transition zone where some, but not all, of the asperities protrude through
the laminar sublayer. The hypothesis of the existence of a laminar sublayer explains the behaviour
of the curves in Figure 2.6. The recognition and early study of boundary layers owe a great deal to
the work of Ludwig Prandtl and the students who started their careers under his guidance.
Nikuradse's work marked a significant step forward in that it promised a means of predicting the
coefficient of friction and, hence, the resistance of any given pipe passing turbulent flow. However,
there continued to be difficulties. In real pipes, ducts or underground airways, the wall asperities are
not all of the same size, nor are they uniformly dispersed. In particular, mine airways show great
variation in their roughness. Concrete lining in ventilation shafts may have a uniform e/d value as low
as 0.001. On the other hand, where shaft tubbing or regularly spaced airway supports are used, the
turbulent wakes on the downstream side of the supports create a dependence of airway resistance
on their distance apart. Furthermore, the immediate wall roughness may be superimposed upon
larger scale sinuosity of the airways and, perhaps, the existence of cross-cuts or other junctions. The
larger scale vortices produced by these macro effects may be more energy demanding than the
smaller eddies of normal turbulent flow and, hence, produce a much higher value of f. Many airways
also have wall roughnesses that exhibit a directional bias, produced by the mechanized or drill and
blast methods of driving the airway, or the natural cleavage of the rock.
For all of these reasons, there may be a significant divergence between Nikuradse's curves and
results obtained in practice, particularly in the transitional zone. Further experiments and analytical
investigations were carried out in the late 1930's by C.F. Colebrook in England. The equations that
were developed were somewhat awkward to use. However, the concept of "equivalent sand grain
roughness" was further developed by the American engineer Lewis F. Moody in 1944. The ensuing
chart, shown on Figure 2.7, is known as the Moody diagram and is now widely employed by
practicing engineers to determine coefficients of friction.
2.3.6.3. Equations describing f - Re relationships
The literature is replete with relationships that have been derived through combinations of analysis
and empiricism to describe the behavior of the coefficient of friction, f, with respect to Reynolds'
Number on the Moody Chart. No attempt is made here at a comprehensive discussion of the merits
and demerits of the various relationships. Rather, a simple summary is given of those equations that
have been found to be most useful in ventilation engineering.
Laminar Flow
The straight line that describes laminar flow on the log-log plot of Figure 2.7 is included in the Moody
Chart for completeness. However, Poiseuille's equation (2.31) can be used directly to establish
frictional pressure losses for laminar flow without using the chart. The corresponding f-Re
relationship is easily established. Combining equations (2.34) and (2.45) gives
p =
8 uL
R2
4fL u 2
d 2
Pa
2 - 27
Figure 2.7 Type of chart developed by Moody.

Substituting R = d/2 gives
f = 16
or
f =
16
Re
ud
dimensionless
(2.52)
Smooth pipe turbulent curve

Perhaps the most widely accepted equation for the smooth pipe turbulent curve is that produced by
both Nikuradse and the Hungarian engineer Theodore Von Krmn (1881-1963).
1
f
= 4 log10 (Re f ) 0.4
2 - 28
This suffers from the disadvantage that f appears on both sides of the equation. Paul R.H. Blasius
(1873-1970), one of Prandtl's earlier students, suggested the approximation for Reynolds Numbers
in the range 3 000 to 105.
0.0791
(2.54)
f =
Re 0.25
while a better fit to the smooth pipe curve for Reynolds Numbers between 20 000 to 107 is given as
0.046
f =
Re 0.2
Rough pipes
When fully developed rough pipe turbulence has been established, the viscous forces are negligible
compared with inertial forces. The latter are proportional to the shear stress at the walls (equation
(2.41)). Hence, in this condition f becomes independent of Reynolds Number and varies only with
e/d. A useful equation for this situation was suggested by Von Krmn.
f =
(2.55)
4 (2 log10 (d / e ) + 1.14) 2
The most general of the f - Re relationships in common use is the Colebrook White equation. This
has been expressed in a variety of ways, including
e
18.7
= 1.74 2 log10 2 +
Re 4f
4f
d
(2.56)
and
e/d
1.255
= 4 log10
+
f
Re f
3 .7
(2.57)
Here again, f, appears on both sides making these equations awkward to use in practice. It was, in
fact, this difficulty that led Moody into devising his chart.
The advantage of the Colebrook White equation is that it is applicable to both rough and smooth
pipe flow and for the transitional region as well as fully developed turbulence. For hydraulically
smooth pipes, e/d =0, and the Colebrook White equation simplifies to the Nikuradse relationship of
equation (2.53). On the other hand, for high Reynolds Numbers, the term involving Re in equation
(2.57) may be ignored. The equation then simplifies to
e / d
f = 4 log10
3 .7
(2.58)
This gives the same results as Von Krmn's rough pipe equation (2.55) for fully developed
turbulence.
Example
A vertical shaft is 400 m deep, 5 m diameter and has wall roughenings of height 5 mm. An airflow of
150 m3/s passes at a mean density of 1.2 kg/m3. Taking the viscosity of the air to be 17.9 x 10-6
Ns/m2 and ignoring changes in kinetic energy, determine:
(i) the coefficient of friction, f
(ii) the turbulent resistance, Rt (m-4)
(iii) the frictional pressure drop p (Pa)
(iv) the work done against friction, F12 (J/kg)
(v) the barometric pressure at the shaft bottom if the shaft top pressure is 100 kPa.
2 - 29
Solution
For a 400 m deep shaft, we can assume incompressible flow (Section 2.1.1.)
Perimeter,
52
= 19.635
4
per = 5 = 15.708 m
Air velocity,
u =
m2
Cross-sectional area, A =
150
Q
=
19.635
A
= 7.639 m/s
In order to determine the regime of flow, we must first find the Reynolds Number
Re =
1.2 7.639 5
ud
=
= 2.561 10 6
6
17.9 10
(i) Coefficient of friction, f:

At this value of Re, the flow is fully turbulent (Section 2.3.4.). We may then use the Moody Chart to
find the coefficient of friction, f. However, for this we need the equivalent roughness
5 10 3
e
=
= 0.001
d
5
Hence at e/d = 0.001 and Re = 2.561 x l06 on Figure 2.7 we can estimate f = 0.0049. (Iterating
equation (2.57) gives f = 0.00494. As the friction coefficient is near constant at this Reynolds
Number, we could use equation (2.55) to give f = 0.00490 or equation (2.58) which gives f =
0.00491). .
(ii) Turbulent resistance, Rt: (equation (2.51))
Rt =
f L per
2A
0.0049 400 15.708

2 (19.635 )
= 0.002 036 m 4
(iii) Frictional pressure drop, p: (equation (2.50))

p = R t Q 2 = 0.002 036 1.2 (150 ) 2 = 54.91
Pa
(iv) Work done against friction, F12 : (equation (2.46))

F12 =
p
54.91
=
= 45.76
1 .2
J
kg
(v) Barometric pressure at shaft bottom, P2 : This is obtained from Bernoulli's equation (2.47) with
no change in kinetic energy.
(Z1 Z 2 )g +
giving
P1 P2
= F12
P2 = (Z 1 Z 2 )g F12 + P1
= ( 400 9.81 1.2) 54.91 + 100 000
P2 = 104 654 Pa
or
104.654 kPa
2 - 30
Bibliography
Blasius, H. (1913) Ds ahnlichtkeitgesetz bei Reibungsvorgngen in flussigkeiten. Forsch. Gebiete
Ingenieus. Vol. 131
Colebrook, C.F. and White, C.M. (1937) Experiments with fluid friction in roughened pipes. Proc.
Royal Soc. (U.K.)(A), vol 161, 367
Colebrook ,C.F. and White, C.M. (1939). Turbulent flow in pipes with particular reference to the
transition region between the smooth and rough pipe laws. Proc. Inst. Civ. Eng. (U.K.) Vol II, 133
Daugherty, R.L. and Franzini, J.B. (1977). Fluid mechanics, with Engineering Applications (7th ed)
McGraw Hill.
Lewitt, E.H. (1959) Hydraulics and fluid mechanics. (10th ed.) Pitman, London
Massey, B.S. (1968) Mechanics of fluids. Van Nostrand , New York.
Moody, L.F. (1944) Friction factors for pipe flow. Trans. Am. Soc. Mech. Engr. Vol. 66 p.671-84
Nikuradse, J. (1933) Strmungsgesetze in rauhen Rohren. VDI -Forschungshaft, Vol 361
Prandtl, L. (1933) Neuere Ergebnisse der Turbulenz-forschung. Zeitschrift des VDI. No. 77, 105
Reynolds, 0. (1883) The motion of water and the law of resistance in parallel channels. Proc Royal
Soc. London, 35
Rohsenhow, W.M and Choi H. (1961) Heat, Mass and Momentum Transfer. International Series in
Eng. Prentice-Hall
Von Krmn, T. (1939) Trans ASME Vol. 61, 705
2 - 31
Fundamentals of Steady Flow thermodynamics
Chapter 3. Fundamentals of steady flow thermodynamics
3.1. INTRODUCTION ...........................................................................................1

3.2. PROPERTIES OF STATE, WORK AND HEAT .............................................2
3.2.1 Thermodynamic properties. State of a system....................................................................... 2
3.2.2 Work and heat ........................................................................................................................ 3
3.3. SOME BASIC RELATIONSHIPS ...................................................................4

3.3.1. Gas laws and gas constants.................................................................................................. 4
3.3.2. Internal Energy and the First Law of Thermodynamics......................................................... 7
3.3.3. Enthalpy and the Steady Flow Energy Equation ................................................................... 8
3.3.4. Specific heats and their relationship to gas constant .......................................................... 10
3.3.5. The Second Law of Thermodynamics................................................................................. 13
3.4. FRICTIONAL FLOW ....................................................................................14

3.4.1. The effects of friction in flow processes .............................................................................. 14
3.4.2. Entropy ................................................................................................................................ 16
3.4.3. The adiabatic and isentropic processes .............................................................................. 19
3.4.4. Availability............................................................................................................................ 20
3.5 THERMODYNAMIC DIAGRAMS..................................................................23

3.5.1. Ideal isothermal (constant temperature) compression. ....................................................... 24
3.5.2. Isentropic (constant entropy) compression ......................................................................... 26
3.5.3. Polytropic compression ....................................................................................................... 28
Further Reading ................................................................................................34
____________________________________________________________________________
3.1. INTRODUCTION
The previous chapter emphasized the behaviour of incompressible fluids in motion. Accordingly, the
analyses were based on the mechanisms of fluid dynamics. In expanding these to encompass
compressible fluids and to take account of thermal effects, we enter the world of thermodynamics.
This subject divides into two major areas. Chemical and statistical thermodynamics are concerned
with reactions involving mass and energy exchanges at a molecular or atomic level, while physical
thermodynamics takes a macroscopic view of the behaviour of matter subjected to changes of
pressure, temperature and volume but not involving chemical reactions. Physical thermodynamics
subdivides further into the study of "closed" systems within each of which remains a fixed mass of
material such as a gas compressed within a cylinder, and "open" systems through which material
flows. A subsurface ventilation system is, of course, an open system with air continuously entering
and leaving the facility. In this chapter, we shall concentrate on open systems with one further
restriction - that the mass flow of air at any point in the system does not change with time. We may,
then, define our particular interest as one of steady flow physical thermodynamics.
Thermodynamics began to be developed as an engineering discipline after the invention of a
practicable steam engine by Thomas Newcomen in 1712. At that time, heat was conceived to be a
massless fluid named 'caloric' that had the ability to flow from a hotter to a cooler body.
Improvements in the design and efficient operation of steam engines, made particularly by James
Watt (1736-1819) in Scotland, highlighted deficiencies in this concept. During the middle of the 19th
century the caloric theory was demolished by the work of James P. Joule (1818-1889) in England,
3-
H.L.F. Helmholtz (1821-1894) and Rudolph J.E. Clausius (1822-1888) in Germany, and Lord
Kelvin (1824-1907) and J.C. Maxwell (1831-1879) of Scotland.
The application of thermodynamics to mine ventilation systems was heralded by the publication of a
watershed paper in 1943 by Frederick B. Hinsley (1900-1988). His work was motivated by
consistent deviations that were observed when mine ventilation surveys were analyzed using
incompressible flow theory, and by Hinsley's recognition of the similarity between plots of pressure
against specific volume constructed from measurements made in mine downcast and upcast shafts,
and indicator diagrams produced by compressed air or heat engines. The new thermodynamic
theory was particularly applicable to the deep and hot mines of South Africa. Mine ventilation
engineers of that country have contributed greatly to theoretical advances and practical utilization of
the more exact thermodynamic methods.
3.2. PROPERTIES OF STATE, WORK AND HEAT

3.2.1 Thermodynamic properties. State of a system.
In Chapter 2 we introduced the concepts of fluid density and pressure. In this chapter we shall
consider the further properties of temperature, internal energy, enthalpy and entropy. These will be
introduced in turn and where appropriate. For the moment, let us confine ourselves to temperature.
Reference to the temperature of substances is such an everyday occurrence that we seldom give
conscious thought to the foundations upon which we make such measurements. The most common
basis has been to take two fixed temperatures such as those of melting ice and boiling water at
standard atmospheric pressure, ascribe numerical values to those temperatures and to define a
scale between the two fixed points. Anders Celsius (1701-1744), a Swedish astronomer, chose to
give values of 0 and 100 to the temperatures of melting ice and boiling water respectively, and to
select a linear scale between the two. The choice of a linear scale is, in fact, quite arbitrary but leads
to a convenience of measurement and simpler relationships between temperature and other
thermodynamic properties. The scale thus defined is known as the Celsius (or Centigrade) scale .
The older Fahrenheit scale was named after Gabriel Fahrenheit (1686-1736), the German scientist
who first used a mercury-in-glass thermometer. Fahrenheit's two "fixed" but rather inexact points
were 0 for a mixture of salt, ice and water, and 96 for the average temperature of the human body. A
linear scale was then found to give Fahrenheit temperatures of 32 for melting ice and 212 for boiling
water. These were later chosen as the two fixed points for the Fahrenheit scale but, unfortunately,
the somewhat irrational numeric values were retained. In the SI system of units, temperatures are
most often related to degrees Celsius.
However, through a thermodynamic analysis, another scale of temperature can be defined that does
not depend upon the melting or boiling points of any substance. This is called the absolute or
thermodynamic temperature scale. N.L. Sadi Carnot (1796-1832), a French military engineer,
showed that a theoretical heat engine operating between fixed inlet and outlet temperatures
becomes more efficient as the difference between those two temperatures increases. Absolute zero
on the thermodynamic temperature scale is defined theoretically as that outlet temperature at which
an ideal heat engine operating between two fixed temperature reservoirs would become 100 per
cent efficient, i.e. operate without producing any reject heat. Absolute zero temperature is a
theoretical datum that can be approached but never quite attained. We can then choose any other
fixed point and interval to define a unit or degree on the absolute temperature scale. The SI system
of units employs the Celsius degree as the unit of temperature and retains 0 C and 100 C for
melting ice and boiling water. This gives absolute zero as -273.15 C. Thermodynamic temperatures
quoted on the basis of absolute zero are always positive numbers and are measured in degrees
Kelvin (after Lord Kelvin). A difference of one degree Kelvin is equivalent to a difference of one
Celsius degree. Throughout this book, absolute temperatures are identified by the symbol T and
temperatures shown as t or denote degrees Celsius.
T = t(C) + 273.15
(3.1)
3-
The Kelvin units, K, are normally shown without a degree () sign. Thermodynamic calculations
should always be conducted using the absolute temperature, T, in degrees Kelvin. However,
as a degree Kelvin is identical to a degree Celsius, temperature differences may be quoted in either
unit.
The state of any point within a system is defined by the thermodynamic properties of the fluid at that
point. If air is considered to be a pure substance of fixed composition then any two independent
properties are sufficient to define its thermodynamic state. In practice, the two properties are often
pressure and temperature as these can be measured directly. If the air is not of fixed composition
as, for example, in airways where evaporation or condensation of water occurs, then at least one
more property is required to define its thermodynamic state (Chapter 14).
The intensive or specific properties of state (quoted on the basis of unit mass) define completely the
thermodynamic state of any point within a system or subsystem and are independent of the
processes which led to the establishment of that state.
3.2.2 Work and heat
Both work and heat involve the transfer of energy. In SI units, the fundamental numerical
equivalence of the two is recognized by their being given the same units, Joules, where
1 Joule = 1 Newton x 1 metre.
Nm
Work usually (but not necessarily) involves mechanical movement against a resisting force.
An equation used repeatedly in Chapter 2 was
Work done = force x distance
or
dW = F dL
Nm or J
(3.2)
and is the basis for the definition of a Joule. Work may be added as mechanical energy from an
external source such as a fan or pump. Additionally, it was shown in Section 2.3.1. that "flow work",
Pv (J) must be done to introduce a plug of fluid into an open system. However, it is only at entry (or
exit) of the system that the flow work can be conceived as a measure of force x distance. Elsewhere
within the system the flow work is a point function. It is for this reason that some engineers prefer not
to describe it as a work term.
Heat is transferred when an energy exchange takes place due to a temperature difference. When
two bodies of differing temperatures are placed in contact then heat will "flow" from the hotter to the
cooler body. (In fact, heat can be transferred by convection or radiation without physical contact.) It
was this concept of heat flowing that gave rise to the caloric theory. Our modern hypothesis is that
heat transfer involves the excitation of molecules in the receiving substance, increasing their internal
kinetic energy at the expense of those in the emitting substance.
Equation (3.2) showed that work can be described as the product of a driving potential (force) and
distance. It might be expected that there is an analogous relationship for heat, dq, involving the
driving potential of temperature and some other property. Such a relationship does, in fact, exist and
is quantified as
dq = T ds
(3.3)
The variable, s, is named entropy and is a property that will be discussed in more detail in Section
3.4.2.
3-
It is important to realize that neither work nor heat is a property of a system. Contrary to popular
phraseology which still retains reminders of the old caloric theory, no system "contains" either heat
or work. The terms become meaningful only in the context of energy transfer across the boundaries
of the system. Furthermore, the magnitude of the transfer depends upon the process path or
particular circumstances existing at that time and place on the boundary. Hence, to be precise, the
quantities dW and dq should actually be denoted as inexact differentials W and q.
The rate at which energy transfers take place is commonly expressed in one of two ways. First, on
the basis of unit mass of the fluid, i.e. Joules per kilogram. This was the method used to dimension
the terms of the Bernoulli equation in Section 2.3.1. Secondly, an energy transfer may be described
with reference to time, Joules per second. This latter method produces the definition of Power.
Power =
dq
dW
or
time
time
J
s
(3.4)
where the unit, J/s is given the name Watt after the Scots engineer, James Watt.
Before embarking upon any analyses involving energy transfers, it is important to define a sign
convention. Many textbooks on thermodynamics have used the rather confusing convention that
heat transferred to a system is positive, but work transferred to the system is negative. This strange
irrationality has arisen from the historical development of physical thermodynamics being motivated
by the study of heat engines, these consuming heat energy (supplied in the form of a hot vapour or
burning fuel) and producing a mechanical work output. In subsurface ventilation engineering, work
input from fans is mechanical energy transferred to the air and, in most cases, heat is transferred
from the surrounding strata or machines - also to the air. Hence, in this engineering discipline it is
convenient as well as being mathematically consistent to regard all energy transfers to the air as
being positive, whether those energy transfers are work or heat. That is the sign convention utilized
throughout this book.
3.3. SOME BASIC RELATIONSHIPS

3.3.1. Gas laws and gas constants
An ideal gas is one in which the volume of the constituent molecules is zero and where there are no
inter-molecular forces. Although no real gas conforms exactly to that definition, the mixture of gases
that comprise air behaves in a manner that differs negligibly from an ideal gas within the ranges of
temperature and pressure found in subsurface ventilation engineering. Hence, the thermodynamic
analyses outlined in this chapter will assume ideal gas behavior.
Some twenty years before Isaac Newton's major works, Robert Boyle (1627-1691) developed a
vacuum pump and found, experimentally, that gas pressure, P, varied inversely with the volume, v,
of a closed system at constant temperature.
1
P
(Boyles law)
(3.5)
v
where means 'proportional to'.
In the following century and on the other side of the English Channel in France, Jacques A.C.
Charles (1746-1823) discovered, also experimentally, that
v T
(Charles' law)
(3.6)
for constant pressure, where T = absolute temperature.
3-
Combining Boyle's and Charles' laws gives

Pv T
where both temperature and pressure vary. Or, inserting a constant of proportionality, R',
Pv = R ' T
(3.7)
To make the equation more generally applicable, we can replace R' by mR where m is the mass of
gas, giving
Pv = m RT
or
v
= RT
m
(3.8)
J/kg
But v/m is the volume of 1 kg, i.e. the specific volume of the gas, V (m3 per kg). Hence,
PV = RT
J/kg
(3.9)
This is known as the General Gas Law and R is the gas constant for that particular gas or mixture of
gases, having dimensions of J/(kg K). The specific volume, V, is simply the reciprocal of density,
1
V =
m3/kg
(3.10)
Hence, the general gas law can also be written as

P
= RT
J/kg
or
P
RT
kg
m3
(3.11)
giving an expression for the density of an ideal gas.
As R is a constant for any perfect gas, it follows from equation (3.9) that the two end states of any
process involving an ideal gas are related by the equation
P1V1
PV
= 2 2 = R
T1
T2
J
kg K
(3.12)
Another feature of the gas constant, R, is that although it takes a different value for each gas, there
is a useful and simple relationship between the gas constants of all ideal gases. Avogadro's law
states that equal volumes of ideal gases at the same temperature and pressure contain the same
number of molecules. Applying these conditions to equation (3.8) for all ideal gases gives
mR = constant
Furthermore, if the same volume of each gas contains an equal number of molecules, it follows that
the mass, m, is proportional to the weight of a single molecule, that is, the molecular weight of the
gas, M. Then
MR = constant
(3.13)
The product MR is a constant for all ideal gases and is called the universal gas constant, Ru.
In SI units, the value of Ru is 8314.36 J /K. The dimensions are sometimes defined as J/(kg mole K)
where one mole is the amount of gas contained in M kg, i.e. its molecular weight expressed in
kilograms. The gas constant for any ideal gas can now be found provided that its molecular weight is
known
3-
R =
8314.36
M
J
kg K
(3.14)
For example, the equivalent molecular weight of dry air is 28.966, giving its gas constant as
8314.36
J
(3.15)
R =
= 287.04
28.966
kg K
Example.
At the top of a mine downcast shaft the barometric pressure is 100 kPa and the air temperature is
18.0 C. At the shaft bottom, the corresponding measurements are 110 kPa and 27.4 C
respectively. The airflow measured at the shaft top is 200 m3/s. If the shaft is dry, determine
(a) the air densities at the shaft top and shaft bottom,
(b) the mass flow of air
(c) the volume flow of air at the shaft bottom.
Solution.
Using subscripts 1 and 2 for the top and bottom of the shaft respectively:(a)
P1
100 000
kg
=
= 1.1966
RT1
287.04 (273.15 + 18)
m3
1 =
In any calculation, the units of measurement must be converted to the basic SI unless ratios are
involved. Hence 100 kPa = 100 000 Pa.
P2
110 000
=
= 1.2751
RT2
287.04 (273.15 + 27.4)
2 =
(b) Mass flow M = Q1 1 = 200 1.1966 = 239.3
(c)
Q2 =
kg
m3
kg
s
239.3
m3
= 187.7
1.2751
s
Example.
Calculate the volume of 100 kg of methane at a pressure of 75 kPa and a temperature of 42 C
Solution.
The molecular weight of methane (CH4) is
12.01 + ( 4 1.008 ) = 16.04
R (methane) =
8314.36
= 518.4
16.04
J
kg K
from equation (3.14)
Volume of 100 kg
v =
mR T
100 518.4 (273.15 + 42)
=
= 217.8
P
75 000
m3
from equation (3.8)
3-
3.3.2. Internal Energy and the First Law of Thermodynamics

Suppose we have 1 kg of gas in a closed container as shown in Figure 3.1. For simplicity, we shall
assume that the vessel is at rest with respect to the earth and is located on a base horizon. The gas
in the vessel has neither macro kinetic energy nor potential energy. However, the molecules of the
gas are in motion and possess a molecular or 'internal' kinetic energy. The term is usually shortened
to internal energy. In the fluid mechanics analyses of Chapter 2 we dealt only with mechanical
energy and there was no need to involve internal energy. However, if we are to study thermal effects
then we can no longer ignore this form of energy. We shall denote the specific (per kg) internal
energy as U J/kg.
Now suppose that by rotation of an impeller within the vessel, we add work W to the closed system
and we also introduce an amount of heat q. The gas in the vessel still has zero macro kinetic
energy and zero potential energy. The energy that has been added has simply caused an increase
in the internal energy.
J
(3.16)
kg
The change in internal energy is determined only by the net energy that has been transferred
across the boundary and is independent of the form of that energy (work or heat) or the process path
of the energy transfer. Internal energy is, therefore, a thermodynamic property of state. Equation
(3.16) is sometimes known as the non-flow energy equation and is a statement of the First Law of
Thermodynamics. This equation also illustrates that the First Law is simply a quantified restatement
of the general law of Conservation of Energy
dU = W + Q
1 kg
Internal
Energy
U
q
Figure 3.1. Added work and heat raise the internal energy of a closed system
3-
3.3.3. Enthalpy and the Steady Flow Energy Equation

Let us now return to steady flow through an open system. Bernoulli's equation (2.15) for a frictionless
system included mechanical energy terms only and took the form
u2
+ Zg + PV = constant
2
J
kg
(3.17)
(where specific volume, V = 1/).

In order to expand this equation to include all energy terms, we must add internal energy, giving
u2
J
(3.18)
+ Zg + PV + U = constant
2
kg
One of the features of thermodynamics is the regularity with which certain groupings of variables
appear. So it is with the group PV + U. This sum of the PV product and internal energy is particularly
important in flow systems. It is given the name enthalpy and the symbol H.
=
PV + U
J/kg
(3.19)
Now pressure, P, specific volume, V, and specific internal energy, U, are all thermodynamic
properties of state. It follows, therefore, that enthalpy, H, must also be a thermodynamic property of
state and is independent of any previous process path.
The energy equation (3.18) can now be written as
J
u2
(3.20)
+ Zg + H = constant
kg
2
Consider the continuous airway shown on Figure 3.2. If there were no energy additions between
stations 1 and 2 then the energy equation (3.20) would give
2
u1
2
+ Z1 g
+ H1
u2
2
+ Z2g
J
kg
+ H2
However, Figure 3.2 shows that a fan adds W12 Joules of mechanical energy to each kilogram of air
and that strata heat transfers q12 Joules of thermal energy to each kilogram of air. These terms must
be added to the total energy at station 1 in order to give the total energy at station 2
2
u1
2
+ Z1 g
+ H 1 + W12 + q12
u2
2
+ Z2g
+ H2
J
kg
(3.21)
This is usually rearranged as follows:
u12 u22
+ (Z1 Z2 )g + W12 = (H2 H1 ) q12
2
J
kg
(3.22)
Equation (3.22) is of fundamental importance in steady flow processes and is given a special name,
the Steady Flow Energy Equation.
3-
q12
W12
.2.
.1.
Figure 3.2. Heat and work added to a steady flow system.
Note that the steady flow energy equation does not contain a term for friction, nor does it require any
such term. It is a total energy balance. Any frictional effects will reduce the mechanical energy
terms and increase the internal energy but will have no influence on the overall energy balance. It
follows that equation (3.22) is applicable both to ideal (frictionless) processes and also to real
processes that involve viscous resistance and turbulence. Such generality is admirable but does,
however, give rise to a problem for the ventilation engineer. In airflow systems, mechanical energy is
expended against frictional resistance only - we do not normally use the airflow to produce a work
output through a turbine or any other device. It is, therefore, important that we are able to quantify
work done against friction. This seems to be precluded by equation (3.22) as no term for friction
appears in that equation:
To resolve the difficulty, let us recall Bernoulli's mechanical energy equation (2.23) corrected for
friction
u12 u22
(P1 P2 )
+ (Z1 Z2 )g +
= F12
2
J
kg
If we, again, add work, Wl2, between stations 1 and 2 then the equation becomes
(P1 P2 )
u12 u22
+ (Z1 Z2 )g +
+ W12 = F12
2
J
kg
(3.23)
(Although any heat, q12, that may also be added does not appear explicitly in a mechanical energy
equation it will affect the balance of mechanical energy values arriving at station 2).
We have one further step to take. As we are now dealing with gases, we should no longer assume
that density, , remains constant. If we consider the complete process from station 1 to station 2 to
be made up of a series of infinitesimally small steps then the pressure difference across each step
becomes dP and the flow work term becomes dP/ or VdP (as V = 1/ ). Summing up, or integrating
all such terms for the complete process gives
3-
2
u12 u22
+ (Z1 Z2 )g + W12 = VdP + F12
2
1
J
kg
(3.24)
This is sometimes called the steady flow mechanical energy equation. It is, in fact, simply a
statement of Bernoulli's equation made applicable to compressible flow.
Finally, compare equations (3.22) and (3.24). Their left sides are identical. Hence, we can write an
expanded version of the steady flow energy equation:
2
u12 u 22
+ (Z1 Z 2 )g + W12 = VdP + F12 = (H 2 H1 ) q12
2
1
J
kg
(3.25)
or, in differential form,

u du g dZ + dW
= V dP + dF = dH dq
Equation (3.25) is the starting point for the application of thermodynamics to subsurface
ventilation systems and, indeed, to virtually all steady flow compressible systems. It is
absolutely fundamental to a complete understanding of the behaviour of airflow in
subsurface ventilation systems.
Students of the subject and ventilation engineers dealing with underground networks of airways
should become completely familiarized with equation (3.25). In most applications within mechanical
engineering the potential energy term (Z1 - Z2)g is negligible and can be dropped to simplify the
equation. However, the large elevation differences that may be traversed by subsurface airflows
often result in this term being dominant and it must be retained for mine ventilation systems.
3.3.4. Specific heats and their relationship to gas constant
When heat is supplied to any unconstrained substance, there will, in general, be two effects. First,
the temperature of the material will rise and, secondly, if the material is not completely constrained,
there will be an increase in volume. (An exception is the contraction of water when heated between
0 and 4C). Hence, the heat is utilized both in raising the temperature of the material and in doing
work against the surroundings as it expands.
The specific heat of a substance has been described in elementary texts as that amount of heat
which must be applied to 1 kg of the substance in order to raise its temperature through one Celsius
degree. There are some difficulties with this simplistic definition. First, the temperature of any solid,
liquid or gas can be increased by doing work on it and without applying any heat - for example by
rubbing friction or by rotating an impeller in a liquid or gas. Secondly, if the substance expands
against a constant resisting pressure then the work that is done on the surroundings necessarily
requires that more energy must be applied to raise the 1 kg through the same 1 C than if the
system were confined at constant volume. The specific heat must have a higher value if the system
is at constant pressure than if it is held at constant volume. If both pressure and volume are allowed
to vary in a partially constrained system, then yet another specific heat will be found. As there are an
infinite number of variations in pressure and volume between the conditions of constant pressure
and constant volume, it follows that there are an infinite number of specific heats for any given substance. This does not seem to be a particularly useful conclusion. However, we can apply certain
restrictions which make the concept of specific heat a valuable tool in thermodynamic analyses.
First, extremely high stresses are produced if a liquid or solid is prevented from expanding during
heating. For this reason, the specific heats quoted for liquids and solids are normally those
pertaining to constant pressure, i.e. allowing free expansion. In the case of gases we can, indeed,
3 - 10
have any combination of changes in the pressure and volume. However, we confine ourselves to the
two simple cases of constant volume and constant pressure.
Let us take the example of a vessel of fixed volume containing 1 kg of gas. We add heat, q, until
the temperature has risen through T. Then, from the First Law (equation (3.16)) with W = 0 for
fixed volume,
q = dU
From our definition of specific heat,
q
U
Cv = =
T v
T v
J
kg K
As this is the particular specific heat at constant volume, we use the subscript v. At any point during
the process, the specific heat takes the value pertaining to the corresponding temperature and
pressure. Hence, we should define Cv, more generally as
Cv
U
=
T v
(3.26)
However, for an ideal gas the specific heat is independent of either pressure or temperature and we
may write
Cv =
dU
dT
J
kg K
(3.27)
In section 3.3.2, the concept of internal energy was introduced as a function of the internal kinetic
energy of the gas molecules. As it is the molecular kinetic energy that governs the temperature of
the gas, it is reasonable to deduce that internal energy is a function of temperature only. This can, in
fact, be proved mathematically through an analysis of Maxwell's equations assuming an ideal gas.
Furthermore, as any defined specific heat remains constant, again, for an ideal gas, equation (3.27)
can be integrated directly between any two end points to give
U 2 U1 = C v (T 2 T1 )
J
kg
(3.28)
Now let us examine the case of adding heat, q, while keeping the pressure constant, i.e. allowing
the gas to expand. In this case, from equation (3.19)
J
kg
Therefore, the specific heat becomes
q = PV + U = H
q
H
CP =
=
T P
T P
(3.29)
J
kg K
(3.30)
J
kg K
(3.31)
In the limit, this defines Cp as

H
CP =
T P
3 - 11
Here again, for a perfect gas, specific heats are independent of pressure and temperature and we
can write
CP =
or
dH
dT
(H 2 H1 ) = C P (T 2 T1 )
J
kg K
(3.32)
J
kg
(3.33)
We shall use this latter equation extensively in Chapter 8 when we apply thermodynamic theory to
subsurface ventilation systems. The same equation also shows that for an ideal gas, enthalpy, H, is
a function of temperature only, Cp being a constant.
There is another feature of Cp that often causes conceptual difficulty. We introduced equations (3.29
and 3.30) by way of a constant pressure process. However, we did not actually enforce the condition
of constant pressure in those equations. Furthermore, we have twice stated that for an ideal gas the
specific heats are independent of either pressure or temperature. It follows that Cp can be used in
equation (3.33) for ideal gases even when the pressure varies. For flow processes, the term specific
heat at constant pressure can be rather misleading.
A useful relationship between the specific heats and gas constant is revealed if we substitute the
general gas law PV = RT (equation (3.9)) into equation (3.29)
H = RT + U
J/kg
Differentiating with respect to T gives

dH
dT
dU
dT
J
kg K
dH
= R + Cv
dT
J
kg K
= R +
i.e
CP =
or
R = C P Cv
J
kg K
(3.34)
The names "heat capacity" or "thermal capacity" are sometimes used in place of specific heat.
These terms are relics remaining from the days of the caloric theory when heat was thought to be a
fluid without mass that could be "contained" within a substance. We have also shown that
temperatures can be changed by work as well as heat transfer so that the term specific heat is itself
open to challenge.
Two groups of variables involving specific heats and gas constant that frequently occur are the ratio
of specific heats (isentropic index)
and
R
Cp
CP
Cv
(dimensionless)
(3.35)
(dimensionless)
3 - 12
Using equation (3.34), it can easily be shown that

R
CP
( 1)
(dimensionless)
(3.36)
Table 3.1 gives data for those gases that may be encountered in subsurface ventilation engineering.
As no real gases follow ideal gas behavior exactly, the values of the "constants" in the table vary
slightly with pressure and temperature. For this reason, there may be minor differences in published
values. Those given in Table 3.1 are referred to low (atmospheric) pressures and a temperature of
26.7C
Gas
Molecular
weight
M
R=8314.36/M
28.966
18.016
28.015
32.000
44.003
28.01
16.04
4.003
2.016
39.94
287.04
461.5
296.8
259.8
188.9
296.8
518.4
2077
4124
208.2
air (dry)
water vapour
nitrogen
oxygen
carbon dioxide
carbon monoxide
methane
helium
hydrogen
argon
Gas
Constant
J/kg K
Specific heats
Cp
Cv=Cp-R
J/(kgK)
1005
718.0
1884
1422
1038
741.2
916.9
657.1
849.9
661.0
1043
746.2
2219
1700
5236
3159
14361
10237
524.6
316.4
Isentropic
index
= CP
Cv
R/Cp
= ( 1) /
1.400
1.324
1.400
1.395
1.286
1.398
1.305
1.658
1.403
1.658
0.2856
0.2450
0.2859
0.2833
0.2223
0.2846
0.2336
0.3967
0.2872
0.3968
Table 3.1. Thermodynamic properties of gases at atmospheric pressures and a temperature of 26.7C.
3.3.5. The Second Law of Thermodynamics

Heat and work are mutually convertible. Each Joule of thermal energy that is converted to
mechanical energy in a heat engine produces one Joule of work. Similarly each Joule of work
expended against friction produces one Joule of heat. This is another statement of the First Law of
Thermodynamics. When equation (3.16) is applied throughout a closed cycle of processes then the
final state is the same as the initial state, i.e.
W
where
= q
dU
= 0 , and
J
kg
(3.37)
indicates integration around a closed cycle .
However, our everyday experience indicates that the First Law, by itself, is incapable of explaining
many phenomena. All mechanical work can be converted to a numerically equivalent amount of heat
through frictional processes, impact, compression or other means such as electrical devices.
However, when we convert heat into mechanical energy, we invariably find that the conversion is
possible only to a limited extent, the remainder of the heat having to be rejected. An internal
combustion engine is supplied with heat from burning fuel. Some of that heat produces a mechanical
work output but, unfortunately, the majority is rejected in the exhaust gases.
3 - 13
Although work and heat are numerically equivalent, work is a superior form of energy. There are
many common examples of the limited value of heat energy. A sea-going liner cannot propel itself by
utilizing any of the vast amount of thermal energy held within the ocean. Similarly, a power station
cannot draw on the heat energy held within the atmosphere. It is this constraint on the usefulness of
heat energy that gives rise to the Second Law of Thermodynamics.
Perhaps the simplest statement of the Second Law is that heat will always pass from a highertemperature body to a lower-temperature body and can never, spontaneously, pass in the opposite
direction. There are many other ways of stating the Second Law and the numerous corollaries that
result from it. The Second Law is, to be precise, a statement of probabilities. The molecules in a
sample of fluid move with varying speeds in apparently random directions and with a mean velocity
approximating that of the speed of sound in the fluid. It is extremely improbable, although not
impossible, for a chance occurrence in which all the molecules move in the same direction. In the
event of such a rare condition, a cup of coffee could become "hotter" when placed in cool
surroundings. Such an observation has never yet been made in practice and would constitute a
contravention of the Second Law of Thermodynamics.
The question arises, just how much of a given amount of heat energy can be converted into work?
This is of interest to the sub-surface ventilation engineer as some of the heat energy added to the
airstream may be converted into mechanical energy in order to help promote movement of the air.
To begin an answer to this question, consider a volume of gas within an uninsulated cylinder and
piston arrangement. When placed in warmer surroundings, the gas will gain heat through the walls
of the cylinder and expand. The piston will move against the resisting pressure of the surrounding
atmosphere and produce mechanical work. This will continue until the temperature of the gas is the
same as that of the surroundings. Although the gas still contains heat energy, it is incapable of
causing further work to be done in those surroundings. However, if the cylinder were then placed in
yet warmer surroundings the gas would expand further and more work would be generated - again
until the temperatures inside and outside the cylinder were equal. From this imaginary experiment it
would seem that heat can produce mechanical work only while a temperature difference exists.
In 1824, while still in his twenties, Sadi Carnot produced a text "Reflections on the Motive Power of
Fire", in which he devised an ideal heat engine operating between a supply temperature T1 and a
lower reject temperature, T2. He showed that for a given amount of heat, q, supplied to the engine,
the maximum amount of work that could be produced is
(T1 T 2 )
q
J
(3.38)
T1
In real situations, no process is ideal and the real work output will be less than the Carnot work due
to friction or other irreversible effects. However, from equation (3.38) a maximum "Carnot Efficiency",
c can be devised:
W ideal
c =
W ideal
q
(T1 T 2 )
T1
(3.39)
3.4. FRICTIONAL FLOW

3.4.1. The effects of friction in flow processes
In the literature of thermodynamics, a great deal of attention is given to frictionless processes,
sometimes called ideal or reversible. The latter term arises from the concept that a reversible
process is one that having taken place can be reversed to leave no net change in the state of either
the system or surroundings.
3 - 14
Frictionless processes can never occur in practice. They are, however, convenient metrics against
which to measure the performance or efficiencies of real heat engines or other devices that operate
through exchanges of work and/or heat. In subsurface ventilation systems, work is added to the
airflow by means of fans and heat is added from the strata or other sources. Some of that heat may
be utilized in helping to promote airflow in systems that involve differences in vertical elevation. An
ideal system would be one in which an airflow, having been initiated, would continue indefinitely
without further degradation of energy. Although they cannot exist in practice, the concept of ideal
processes assists in gaining an understanding of the behaviour of actual airflow systems.
Let us consider what we mean by "friction" with respect to a fluid flow. The most common everyday
experience of friction is concerned with the contact of two surfaces - a brake on a wheel or a tyre on
the road. In fluid flow, the term "friction" or "frictional resistance" refers to the effects of viscous
forces that resist the motion of one layer of fluid over another, or with respect to a solid boundary
(Section 2.3.3). In turbulent flow, such forces exist not only at boundaries and between laminae of
fluid, but also between and within the very large number of vortices that characterize turbulent flow.
Hence, the effect of fluid friction is much greater in turbulent flow.
In the expanded version of the steady flow energy equation, the term denoting work done against
frictional effects was F12. To examine how this affects the other parameters, let us restate that
equation.
u12 u 22
+ (Z1 Z 2 )g + W12 =
2
VdP + F12 = (H 2 H1 ) q12
J
(from equation (3.25))
kg
But from equation (3.33) for a perfect gas

(H 2 H1 ) = C P (T 2 T1 )
J
kg
giving
2
u12 u 22
+ (Z1 Z 2 )g + W12 = VdP + F12 = C P (T 2 T1 ) q12
2
1
(3.40)
J
kg
(3.41)
We observe that the friction term, F12, appears in only the middle section of this three part equation.
This leads to two conclusions for zero net change in the sum of kinetic, potential and fan input
energy, i.e. each part of the equation remaining at the same total value. First, frictional effects, F12,
2
appear at the expense of VdP . As F12 increases then the flow work must decrease for the middle
1
part of the equation to maintain the same value. However, the conversion of mechanical energy into
heat through frictional effects will result in the specific volume, V, expanding to a value that is higher
than would be the case in a corresponding frictionless system. It follows that the appearance of
friction must result in a loss of pressure in a flow system. A real (frictional) process from station 1 to
station 2 will result in the pressure at station 2 being less than would be the case of a corresponding
ideal process. The difference is the "frictional pressure drop".
Secondly, and again for a zero net change in kinetic, potential and fan energies, the lack of a friction
term in the right hand part of equation (3.41) shows that the change in temperature (T2 - T1) is
independent of frictional effects. In other words, the actual change in temperature along an airway is
the same as it would be in a corresponding frictionless process. This might seem to contradict
everyday experience where we expect friction to result in a rise in temperature. In the case of the
steady flow of perfect gases, the frictional conversion of mechanical work to heat through viscous
shear produces a higher final specific volume and a lower pressure than for the ideal process, but
exactly the same temperature.
3 - 15
Returning to the concept of frictional pressure drop, the steady flow energy equation can also be
used to illustrate the real meaning of this term. If we accept the usual situation in which the variation
in density, (= 1/V), along the airway is near linear, we can multiply equation (3.25) by a mean value
of density, m to give
(u12 u22 )
+ m (Z1 Z2 )g + mW12 = (P2 P1) + mF12 = m (H2 H1) mq12
2
J kg J
=
kg m3 m3
(3.42)
(Note the new units of the terms in this equation. They still express variations in energy levels, (J),
but now with reference to a unit volume (m3) rather than a unit mass (kg). As it is the total mass flow
that remains constant the steady flow energy equation (in J/kg) is the preferred version of the
equation. (See also, Section 2.1.2). However, most of the terms in equation (3.42) do have a
physical meaning. We can re-express the units as
J
m
Nm
m
N
m2
= Pa
(3.43)
As this is the unit of pressure, equation (3.42) is sometimes called the steady flow pressure
equation. Furthermore,
(u12 u 22 )
2
is the change in velocity pressure (see equation (2.17))
m (Z1 Z2 )g is the change in static pressure due to the column of air between Z1 and Z2
mW12
is the increase in pressure across the fan
(P2 P1 )
is the change in barometric pressure, and
m F12
is the frictional pressure drop, p, (equation (2.46)).
The frictional pressure drop may now be recognized as the work done against friction per cubic
metre of air. As the mass in a cubic metre varies due to density changes, the disadvantage of a
relationship based on volume becomes clear.
3.4.2. Entropy
In section 3.3.5. we discussed work and heat as "first and second class" energy terms respectively.
All work can be transferred into heat but not all heat can be transferred into work. Why does this
preferential direction exist? It is, of course, not the only example of 'one-wayness' in nature. Two
liquids of the same density but different colours will mix readily to a uniform shade but cannot easily
be separated back to their original condition. A rubber ball dropped on the floor will bounce, but not
quite to the height from which it originated - and on each succeeding bounce it will lose more height.
Eventually, the ball will come to rest on the floor. All of its original potential energy has been
converted to heat through impact on the floor, but that heat cannot be used to raise the ball to its
initial height. Each time we engage in any non-ideal process, we finish up with a lower quality, or
less organized, state of the system. Another way of putting it is that the "disorder" or "randomness"
of the system has increased. It is a quantification of this disorder that we call entropy.
3 - 16
Suppose we build a symmetric tower out of toy building bricks. The system is well ordered and has a
low level of entropy. Now imagine your favorite infant taking a wild swipe at it. The bricks scatter all
over the floor. Their position is obviously now in a much greater degree of disorder. Energy has been
expended on the system and the entropy (disorder) has increased. This is entropy of position. Let us
carry out another imaginary experiment. Suppose we have a tray on which rest some marbles. We
vibrate the tray gently and the marbles move about in a random manner. Now let us vibrate the tray
violently. The marbles become much more agitated or disordered in their movement. We have done
work on the system and, again, the entropy level has increased. This is entropy of motion.
What has all this to do with heat transfer? Imagine that we have a perfect crystal at a temperature of
absolute zero. The molecules are arranged in a symmetric lattice and are quite motionless. We can
state that this is a system of perfect order or zero entropy. If we now add heat to the crystal, that
energy will be utilized in causing the molecules to vibrate. The more heat we add, the more agitated
the molecular vibration and the greater the entropy level. Can you see the analogy with the marbles?
The loss of order is often visible. For example, ice is an "ordered" form of water. Adding heat will
cause it to melt to the obviously less organized form of liquid water. Further heat will produce the
even less ordered form of water vapour.
How can we quantify this property that we call entropy? William J.M. Rankine (1820-1872), a Scots
professor of engineering showed in 1851 that during a reversible (frictionless) process, the ratio of
heat exchanged to the current value of temperature, q/T, remained constant. Clausius arrived at
the same conclusion independently in Germany during the following year. Clausius also recognized
that particular ratio to be a thermodynamic function of state and coined the name entropy, s (after the
Greek work for 'evolution'). Clausius further realized that although entropy remained constant for ideal
processes, the total entropy must increase for all real processes.
ds =
0
T
When testing Clausius' conclusion, it is important to take all parts of the system and surroundings
into account. A subsystem may be observed to experience a decrease in entropy if viewed in
isolation. For example, water in a container that is placed in sufficiently colder surroundings can be
seen to freeze - the ordered form of ice crystals growing, apparently spontaneously, on the surface
of the less ordered liquid. The entropy of the water is visibly decreasing. Suppose the temperature of
the water is Tw and that of the cooler air is Ta where Ta < Tw. Then for a transfer of heat q from the
water to the air, the corresponding changes in entropy for the air (sub-script a), and the water
(subscript w), are:
and
ds a
ds w
Ta
q
Tw
(negative as heat is leaving the water)
Then the total change in entropy for this system is

1
1
ds a + ds w = q
Tw
T a
3 - 17
Using the thermodynamic property, entropy, we can now express the combined heat increase of a
system, by both heat transfer and frictional effects, as dqc
J
kg
= T ds
dq c
(3.44)
(see, also, equation (3.3)).

The symbol qc, is used to denote the combination of added heat and the internally generated
frictional heat.
Now to make the concept really useful, we must be able to relate entropy to other thermodynamic
properties. Let us take the differential form of the steady flow energy equation (3.25).
J
kg
VdP + dF = dH dq
Then
(3.45)
dF + dq = dH VdP
But the combined effect of friction and added heat is

dq c = dF + dq
Therefore, equations (3.44) and (3.45) give

J
Tds = dH VdP
(3.46)
kg
This is another equation that will be important to us in the thermodynamic analysis of subsurface
ventilation circuits.
In order to derive an expression that will allow us to calculate a change in entropy directly from
measurements of pressure and temperature, we can continue our analysis from equation (3.46).
ds =
dH
VdP
T
T
But
dH = C P dT
and
V
T
giving
ds = C P
R
P
J
kg K
(equation (3.32))
(from the General Gas Law PV = RT)
dT
dP
R
T
P
J
kg K
(3.47)
Integrating between end stations 1 and 2 gives

(s 2
s1 ) = C P ln T 2 R ln P2
T1
P1
J
kg K
(3.48)
This is known as the steady flow entropy equation.
3 - 18
3.4.3. The adiabatic and isentropic processes

An important thermodynamic process with which the ventilation engineer must deal is one in which
there is no heat transfer between the air and the strata or any other potential source. This can be
approached closely in practice, particularly in older return airways that contain no equipment and
where the temperatures of the air and the surrounding rock have reached near equilibrium.
The steady flow energy equation for adiabatic flow is given by setting q12 = 0.
Then
VdP + F12
J
kg
(H 2 H 1 )
(3.49)
The ideal, frictionless, or reversible adiabatic process is a particularly useful concept against which
to compare real adiabatic processes. This is given simply by eliminating the friction term
2
VdP
(H 2 H 1 )
or
J
kg
(3.50)
VdP = dH
In Section 3.4.2, we defined a change in entropy, ds, by equation (3.46) i.e.

J
(3.51)
kg
where qc is the combined effect of added heat and heat that is generated internally by frictional
effects. It follows that during a frictionless adiabatic where both q12 and F12 are zero, then dqc is also
zero and the entropy remains constant, i.e. an isentropic process.
dq c
= Tds
The governing equations for an isentropic process follow from setting (s2 s1) = 0 in equation (3.48).
Then
J
C P ln T 2 = R ln P2
T
P
1
1
kg K
Taking antilogs gives

T2
T1
P
= 2
P1
CP
(3.52)
The index R/Cp was shown by equation (3.36) to be related to the ratio of specific heats Cp/Cv =
by the equation
R
CP
( 1)
Then
T2
T1
P
= 2
P1
(3.53)
Values of the constants are given in Table 3.1. For dry air, = 1.400, giving the isentropic
relationship between temperature and pressure as
3 - 19
T2
T1
P
= 2
P1
0.2856
(3.54)
This is a particularly useful equation to relate pressures and temperatures in dry mine shafts where
adiabatic conditions may be approached.
The isentropic relationship between pressure and specific volume follows from equation (3.53)
T2
T1
P
= 2
P1
1 1
(3.55).
But from the General Gas Law, (equation (3.12))

T2
T1
P2 V 2
P1 V1
(3.56)
Combining equations (3.55 and 3.56) gives

P
= 2
P1
P2 V 2
P1 V1
P
= 2
P1
V2
V1
or
P1 V1
1 1
= P2 V 2
P
= 1
P2
= constant
(3.57)
3.4.4. Availability
In the context of conventional ventilation engineering, the energy content of a given airstream is
useful only if it can be employed in causing the air to move, i.e. if it can be converted to kinetic
energy. A more general concept is that of available energy. This is defined as the maximum amount
of work that can be done by a system until it comes to complete physical and chemical equilibrium
with the surroundings.
Suppose we have an airflow of total energy
u2
+
2
+ Zg + H
J
kg
(equation 3.20))
The kinetic and potential energy terms both represent mechanical energy and are fully available to
produce mechanical effects, i.e. to do work. This is not true for the enthalpy. Remember that
enthalpy is comprised of PV and internal energy terms, and that the Second Law allows only a
fraction of thermal energy to be converted into work. Suppose that the free atmosphere at the
surface of a mine has a specific enthalpy Ho (subscript o for outside atmosphere). Then when the
mine air is rejected at temperature T to the surface, it will cool at constant pressure until it reaches
the temperature of the ambient atmosphere To. Consequently, its enthalpy will decrease from H to
Ho.
3 - 20
J
kg
(H H o ) = C P (T To )
(3.58)
The process is shown on a temperature-entropy diagram in Figure 3.3. The heat rejected to the free
b
atmosphere is represented by the area under the process line
Tds
= area abcd . However, for an
isobar (constant pressure),
Tds = dH
a
= H Ho (see equation (3.46) with dP = 0. Hence, the
enthalpy change H - Ho gives the heat lost to the atmosphere. Equation (3.58) actually illustrates the
same fact, as heat lost is
mass (1 kg) x specific heat (Cp) x change in temperature (T - To)
for a constant pressure process.
Temperature T
Available
To
b
b
Tds
Unavailable
c
s
so
Entropy s
Figure 3.3. Temperature entropy diagram for constant pressure cooling
Now the Second Law insists that only a part of this heat can be used to do work. Furthermore, we
have illustrated earlier that when a parcel of any gas reaches the temperature of the surroundings
then it is no longer capable of doing further work within those surroundings. Hence, the part of the
3 - 21
heat energy that remains unavailable to do any work is represented by the Tds area under the
e
ambient temperature line To, i.e. To (s - so) or area bcde on Figure 3.3. The only part of the total
heat that remains available to do work is represented by area abe or (H - Ho) - To (s - so).
The available energy, , in any given airflow, with respect to a specified datum (subscript o) may
now be written as
u2
J
(3.59)
+ Zg + (H H o ) To (s s o )
2
kg
It should be made clear that the available energy represents the maximum amount of energy that is
theoretically capable of producing useful work. How much of this is actually used depends upon the
ensuing processes.
Let us now try to show that any real airway suffers from a loss of available energy because of
frictional effects. If we rewrite equation (3.59) in differential form for an adiabatic airway, then the
increase in available energy along a short length of the airway is
d = u du + g dZ + dH To ds
(Ho and so are constant for any given datum conditions).
This equation assumes that we do not provide any added heat or work. From the differential form of
the steady flow energy equation (3.25) for adiabatic conditions and no fan work (dq = dW = 0),
u du + g dZ + dH = 0
leaving d = To ds
J/kg
(3.60).
dq c
T
where qc is the combined effect of friction, F, and added heat, q. However, in this case, q is zero
giving
dF
ds =
T
and
T
J
d = o dF
(3.61)
kg
T
Now, from equation (3.51), ds =
As To, T and dF are positive, the change in available energy must always be negative in the absence
of any added work or heat. This equation also shows that the loss of available energy is a direct
consequence of frictional effects.
Available energy is a "consumable" item, unlike total energy which remains constant for adiabatic
flow with no work input. During any real airflow process, the available energy is continuously eroded
by the effects of viscous resistance in the laminar sublayer and within the turbulent eddies. That
energy reappears as "low grade" heat or unavailable energy and is irretrievably lost in its capacity to
do useful work. The process is illustrated in Figure 3.4.
3 - 22
Total energy
Available
Available
u12 /2
u22 /2
Z2 g
Z1 g
(H2 Ho) To(s2 - so)

(H1 Ho) To(s1 - so)
Ho + To(s1 so)
Unavailable
friction
Ho + To(s2 so)
Unavailable
Figure 3.4. Available energy decreases and unavailable energy increases both by
the amount To (s2 s1) when no heat or work are added.
3.5 THERMODYNAMIC DIAGRAMS.

During any thermodynamic process, there will be variations in the values of the fluid properties. The
equations derived in the preceding sections of this chapter may be used to quantify some of those
changes. However, plotting one property against another provides a powerful visual aid to
understanding the behaviour of any process path and can also give a graphical means of quantifying
work or heat transfers where the complexity of the process path precludes an analytical treatment.
These graphical plots are known as thermodynamic diagrams.
The two most useful diagrams in steady flow thermodynamics are those of pressure against specific
volume and temperature against specific entropy. These diagrams are particularly valuable, as areas
on the PV diagram represent work and areas on the Ts diagram represent heat. Remembering
that simple fact will greatly facilitate our understanding of the diagrams. In this section, we shall
introduce the use of these diagrams through three compression processes. In each case, the air will
be compressed from pressure P1 to a higher pressure P2. This might occur through a fan,
compressor or by air falling through a downcast shaft. The processes we shall consider are
isothermal, isentropic and polytropic compression. As these are important processes for the
ventilation engineer, the opportunity is taken to discuss the essential features of each, in addition to
giving illustrations of the visual power of thermodynamic diagrams.
3 - 23
3.5.1. Ideal isothermal (constant temperature) compression.
Pressure P
Suppose air is passed through a compressor so that its pressure is raised from P1 to P2. As work is
done on the air, the First Law of Thermodynamics tells us that the internal energy and, hence, the
temperature of the air will increase (equation (3.16)). However, in this particular compressor, we
have provided a water jacket through which flows a continuous supply of cooling water. Two
processes then occur. The air is compressed and, simultaneously, it is cooled at just the correct rate
to maintain its temperature constant. This is isothermal compression.
P2
V dP
P1
isotherm
Temperature
Specific Volume V
P2
P1
T1
isotherm
Tds
X
s2
s1
Specific Entropy s
Figure 3.5. PV and Ts diagrams for an isothermal compression.
3 - 24
Figure 3.5 shows the process on a pressure-specific volume (PV) and a temperature-entropy (Ts)
diagram. The first stage in constructing these diagrams is to draw the isobars representing P1 and
P2. On the PV diagram these are simply horizontal lines. Lines on the Ts diagram may be plotted
using equation (3.48). Isobars curve slightly. However, over the range of pressures and
temperatures of interest to the ventilation engineer, the curvature is small.
The process path for the isothermal (constant temperature) compression is shown as line AB on
both diagrams. On the PV diagram, it follows the slightly curved path described by the equation
PV = RT1
or
P =
RT1
V
Pa
(3.62)
On the Ts diagram, the isotherm is, of course, simply a horizontal line at temperature T1.
Let us first concentrate on the PV diagram. From the steady flow energy equation for a frictionless
process, we have
(u12 u 22 )
2
+ (Z1 Z 2 )g + W12
J
kg
= VdP
1
In this case we may assume that the flow through the compressor is horizontal (Z1 = Z2) and that the
change in kinetic energy is negligible, leaving
=
W12
J
kg
VdP
(3.63)
The integral VdP is the area to the left of the process line on the PV diagram, i.e., area ABXY for
1
the ideal compressor is equal to the work, W12, that has been done by the compressor in raising the
air pressure from P1 to P2. We can evaluate the integral by substituting for V from equation (3.62).
W12
RT
J
kg
1
2
VdP = P dP = RT1 ln P
1
1
1
(3.64)
as RT1 is constant
This is positive as work is done on the air.
Now let us turn to the Ts diagram. Remember that the integral Tds represents the combined heat
from actual heat transfer and also that generated by internal friction. However, in this ideal case,
there is no friction. Hence the Tds area under the process line AB represents the heat that is
removed from the air by the cooling water during the compression process. The heat area is the
rectangle ABXY or -T1 (s1 - s2) on the Ts diagram. In order to quantify this heat, recall the entropy
equation (3.48) and apply the condition of constant temperature. This gives
P
T1(s1 s2 ) = RT1 ln 2
P1
The sign is negative as heat is removed from the air.
J
kg
(3.65)
3 - 25
Now compare the work input (equation (3.64)) with the heat removed (equation (3.65)). Apart from
the sign, they are numerically identical. This means that as work is done on the system to compress
the air, exactly the same quantity of energy is removed as heat during this isothermal process. [This
result is also shown directly by equations (3.16) and (3.27) with dU = W + q = CvdT = 0
giving W = - q for an isothermal process].
Despite the zero net increase in energy, the air has been pressurized and is certainly capable of
doing further useful work through a compressed air motor. How can that be? The answer lies in the,
discussion on availability given in Section 3.4.4. All of the work input is available energy capable of
producing mechanical effects. However, all of the heat removed is unavailable energy that already
existed in the ambient air and which could not be used to do useful work. The fact that this heat is,
indeed, completely unavailable is illustrated on the Ts diagram by the corresponding heat area
-T1(s1 - s2) lying completely below the ambient temperature line.
This process should be studied carefully until it is clearly understood that there is a very real
distinction between available energy and unavailable energy. The high pressure air that leaves the
compressor retains the work input as available energy but has suffered a loss of unavailable energy
relative to the ambient air. If the compressed air is subsequently passed through a compressed air
motor, the available energy is utilized in producing a mechanical work output, leaving the air with
only its depressed unavailable energy to be exhausted back to the atmosphere. This explains why
the air emitted from the exhaust ports of a compressed air motor is very cold and may give rise to
problems of condensation and freezing.
Maintaining the temperature constant during isothermal compression minimizes the work that must
be done on the air for any given increase in pressure. Truly continuous isothermal compression
cannot be attained in practice, as a temperature difference must exist between the air and the
cooling medium for heat transfer to occur. In large multi-stage compressors, the actual process path
on the Ts diagram proceeds from A to B along a zig-zag line with stages of adiabatic compression
alternating with isobaric cooling attained through interstage water coolers.
3.5.2. Isentropic (constant entropy) compression
During this process we shall, again, compress the air from P1 to P2 through a fan or compressor, or
perhaps by gravitational work input as the air falls through a downcast shaft. This time, however, we
shall assume that the system is not only frictionless but is also insulated so that no heat transfer can
take place. We have already introduced the frictionless adiabatic in Section 3.4.2 and shown that it
maintains constant entropy, i.e. an isentropic process.
The PV and Ts diagrams are shown on Figure 3.6. The corresponding process lines for the
isothermal case have been retained for comparison. The area to the left of the isentropic process
line, AC, on the PV diagram is, again, the work input during compression. It can be seen that this is
2
greater than for the isothermal case. This area, ACXY or VdP , is given by the steady flow energy
1
equation (3.25) with F12 = q12 = 0,

i.e.
V dP = H 2 H1 = C P (T 2 T1 )
J
kg
(3.66)
On the Ts diagram, the process line AC is vertical (constant entropy) and the temperature increases
from T1 to T2. The Tds area under the line is zero. This suggests that the Ts diagram is not very
useful in further evaluating an isentropic process. However, we are about to reveal a feature of Ts
diagrams that enhances their usefulness very considerably. Suppose that, having completed the
isentropic compression and arrived at point C on both diagrams, we now engage upon an imaginary
3 - 26
second process during which we cool the air at constant pressure until it re-attains its original
ambient temperature T1. The process path for this second process of isobaric cooling is CB on both
diagrams. The heat removed during the imaginary cooling is the area Tds under line CB on the Ts
diagram, i.e. area CBXY. But from the steady flow entropy equation (3.47)
ds = C P
dP
dT
R
P
T
J
kg K
Pressure P
and for an isobar, dP = 0, giving ds = C P
P2
dT
T
isotherm
V dP
P1
Tds = C P dT
or
isentrope
Specific Volume V
P2
T2
T1
isentrope
Temperature
P1
isotherm A
s1=s2
Specific Entropy s
Figure 3.6. PV and Ts diagrams for an isentropic compression
3 - 27
Then heat removed during the imaginary cooling becomes

1
Tds
= Tds = C p dT = C p (T 2 T1 )
J
kg
(3.67)
Now compare equation (3.66) and (3.67). It can be seen that the heat removed during the imaginary
cooling is numerically equivalent to the work input during the isentropic compression. Hence, the
work input is not only shown as area ACBXY on the PV diagram but also as the same area on the
Ts diagram. Using the device of imaginary isobaric cooling, the Ts diagram can be employed to
illustrate work done as well as heat transfer. However, the two must never be confused - the Ts
areas represent true heat energy and can differentiate between available and unavailable heat, while
work areas shown on the Ts diagram are simply convenient numerical equivalents with no other
physical meaning on that diagram.
3.5.3. Polytropic compression
The relationship between pressure and specific volume for an isentropic process has been shown to
be
PV = constant
(equation 3.57)
where the isentropic index is the ratio of specific heats Cp/Cv. Similarly, for an isothermal process,
PV 1 = constant .
These are, in fact, special cases of the more general equation
PV n = constant, C
(3.68)
where the index n remains constant for any given process but will take a different value for each
separate process path. This general equation defines a polytropic system and is the type of process
that occurs in practice within subsurface engineering. It encompasses the real situation of frictional
flow and the additional increases in entropy that arise from heat transfer to the air.
Figure 3.7 shows the PV and Ts process lines for a polytropic compression. Unlike the isothermal
and isentropic cases, the path line for the polytropic process is not rigidly defined but depends upon
the value of the polytropic index n. The polytropic curve shown on the diagrams indicates the most
common situation in underground ventilation involving both friction and added heat.
The flow work shown as area ADXY on the PV diagram may be evaluated by integrating
2
C
VdP where V =
P
Then
VdP =
or, as
1
n
= P
1
n
C

P
1
n
= C
1
n

1 1n 2
P
1
1 1
n
1
nV
3 - 28
VdP
n
(P2V2 P1V1 ) = n R (T2 T1 ) J
kg
n 1
n 1
(3.69)
using the gas law PV = RT
Pressure P
This enables the flow work to be determined if the polytropic index, n, is known. It now becomes
necessary to find a method of calculating n, preferably in terms of the measurable parameters,
pressure and temperature.
P2
B C
isotherm
polytrope
V dP
isentrope
P1
Specific Volume V
C
polytrope
isentrope
T
Temperature
T1
P2
T2
B
isotherm
P1
Tds
Y
s1
Z
s2
Specific Entropy s
Figure 3.7. PV and Ts diagrams for a polytropic compression.
3 - 29
From the polytropic law (3.68) P1V1
= P2V2
P
V1
or
= 2
V2
P1
1
n
(3.70)
and from the General Gas Law, (equation (3.12))

V1
P T
= 2 1
V2
P1 T 2
(3.71)
We have isolated V1 /V2 in equations (3.70) and (3.71) in order to leave us with the desired
parameters of pressure and temperature. Equating (3.70) and (3.71) gives
P2
P1
1
n
P T
= 2 1
P1 T 2
P
T2
= 2
T1
P1
or
1 1
n
Taking logarithms gives

lnT 2
n 1
T1
=
n
ln P2
P1
(3.72)
This enables the polytropic index, n, to be determined for known end pressures and temperatures.
However, we can substitute for n/(n-1), directly, into equation (3.69) to give the flow work as
2
VdP
= R (T 2 T1 )
ln P2
P1
lnT 2
T1
J
kg
(3.73)
This is an important relationship that we shall use in the analysis of mine ventilation thermodynamics
(Chapter 8).
Turning to the Ts diagram, the
Tds area under the polytrope AD, i.e. area ADZY, is the combined
heat increase arising from internal friction, F12, and added heat, q12 :
J
kg
But from the steady flow energy equation (3.25)
Tds
= F12 + q12
(3.74)
F12 + q12 = H 2 H 1
J
kg
VdP
1
(3.75)
giving the area under line AD as

2
Tds
1
= H 2 H1
VdP
1
J
kg
(3.76)
3 - 30
We arrived at the same result in differential form (equation (3.46)) during our earlier general
discussion on entropy.
As (H2 H1) = Cp (T2 - T1), and using equation (3.73) for the flow work, we can re-write equation
(3.76) as
2
Tds
= C p (T 2 T1 ) R (T 2 T1 )
= (T 2 T1 )
ln P2
P1
lnT 2
T1
ln P2
P1
C p R
2
ln
T1
J
kg
J
kg
(3.77)
Now, using the same logic as employed in Section 3.5.2, it can be shown that the area under the
isobar DB on the Ts diagram of Figure 3.7 is equal to the change in enthalpy (H2 H1). We can now
illustrate the steady flow energy equation as areas on the Ts diagram
2
H 2 H1
F12
+ q12 +
J
kg
VdP
1
Area DBXZ =
Area ADZY
(3.78)
Area ADBXY
Once again, this shows the power of the Ts diagram.

Example.
A dipping airway drops through a vertical elevation of 250m between stations 1 and 2. The following
observations are made.
Velocity u (m/s)
Station 1
Station 2
2.0
3.5
Pressure P (kPa)
Temperature t (C) Airflow Q (m3/s)
93.40
95.80
28.20
29.68
43
Assuming that the airway is dry and that the airflow follows a polytropic law, determine
(a) the polytropic index, n
(b) the flow work
(c) the work done against friction and the frictional pressure drop
(d) the change in enthalpy
(e) the change in entropy and
(f) the rate and direction of heat transfer with the strata, assuming no other sources of heat.
Solution.
It is convenient to commence the solution by calculating the end air densities and the mass flow of
air.
P1
1 =
(equation (3.11))
RT1
=
93 400
= 1.0798
287.04 (273.15 + 28.20 )
kg
m3
3 - 31
2 =
P2
95 800
=
= 1.1021
287.04 (273.15 + 29.68 )
RT2
Mass flow M = Q1 x 1 = 43.0 x 1.0798 = 46.43
kg
m3
kg/s
(a) Polytropic index, n:

From equation (3.72)
lnT 2
n 1
T1
=
n
ln P2
P1
where T1 = 273.15 + 28.20 = 301.35 and T2 = 273.15 + 29.68 = 302.83 K
ln 302.83
n 1
301.35
=
n
ln 95.80
93.40
= 0.1931
giving n = 1.239. This polytropic index is less than the isentropic index for dry air, 1.4, indicating
that heat is being lost from the air to the surroundings.
(b) Flow work:
2
VdP = R (T 2 T1 )
ln P2
ln 95.80
P1
93.40
= 287.04(29.68 28.20 )
= 2200 .0
302.83
T
ln
2
ln
301.35
T1
J
kg
Degrees Celsius can be used for a difference (T2 T1) but remember to employ degrees Kelvin in all
other circumstances.
(c) Friction:
From the steady flow energy equation with no fan
2
F12 =
u1 u 2
2
+ (Z1 Z 2 )g
VdP
1
2 2 3.5 2
+ 250 9.81 2200 .0
2
= 4 .1
+ 2452 .5
2200 .0 = 248.4 J/kg
Note how small is the change in kinetic energy compared with the potential energy and flow work.
In order to determine the frictional pressure drop, we use equation (2.46):
p = F12
3 - 32
For this to be meaningful, we must specify the value of density to which it is referred. At a mean
density (subscript m) of
m =
1 + 2
2
1.0798 + 1.1021
= 1.0909
2
p m = 1.0909 248.4 = 271
J
kg
Pa
or, for comparison with other pressure drops, we may choose to quote our frictional pressure drop
referred to a standard air density of 1.2 kg/m3, (subscript st) giving
pst = 1.2 x 248.4 = 298 Pa
(d) Change in enthalpy:

H 2 H 1 = C p (T2 T1 )
where Cp = 1005 J/(kg K) for dry air (Table 3.1)
= 1005 (29.68 - 28.20) = 1487.4 J/kg

(e) Change in entropy:
(s 2
T
s1 ) = C p ln 2
T1
P
R ln 2
P1
= 1005 In(302.83/301.35) - 287.04 In(95.8/93.4)

=
4.924
7.283
= - 2.359 J/(kgK)
The decrease in entropy confirms that heat is lost to the strata.

(f) Rate of heat transfer :
Again, from the steady flow energy equation (3.25)
2
q12 = H 2 H 1
u1 u 2
2
(Z1 Z 2 )g
(Each term has already been determined, giving
q12 = 1487.4 + 4.1 - 2452.5 = - 961.0 J/kg

To convert this to kilowatts, multiply by mass flow
q12 =
961.0 46.43
= 44.6
1000
kW
The negative sign shows again that heat is transferred from the air to the strata and at a rate of 44.6 kW.
3 - 33
Further Reading
Hinsley, F.B. (1943) Airflow in Mines: A Thermodynamic Analysis. Proc. South Wales Inst. of
Engineers Vol LIX, No. 2
Look, D.C. and Sauer, H.J. (1982)Thermodynamics. Published by Brooks /Cole Engineering
Division, Monterey, California
Rogers, G.F.C and Mayhew, Y.R. (1957). Engineering Thermodynamics, Work and Heat Transfer.
Published by Longmans, Green and Co.
Van Wylen, G. J. (1959). Thermodynamics. Published by John Wiley and Sons.
3 - 34
Chapter 4. Subsurface Ventilation Systems.
Chapter 4. Subsurface Ventilation Systems
4.1. INTRODUCTION ...........................................................................................2

4.2. MINE SYSTEMS............................................................................................2
4.2.1 General principles ................................................................................................................ 2
Fans ................................................................................................................................. 3
Stoppings and Seals ........................................................................................................ 3
Doors and airlocks ........................................................................................................... 3
Regulators........................................................................................................................ 4
Air crossings..................................................................................................................... 4
4.2.2 Location of main fans........................................................................................................... 5
Gas control....................................................................................................................... 5
Transportation .................................................................................................................. 6
Fan maintenance ............................................................................................................. 6
Fan performance.............................................................................................................. 6
4.2.3. Infrastructure of main ventilation routes.............................................................................. 7
Mine resistance................................................................................................................ 7
Leakage control ............................................................................................................... 8
Direction of airflow ........................................................................................................... 9
Airflow travel distance and use of old workings ............................................................. 10
4.3. DISTRICT SYSTEMS ..................................................................................11

4.3.1. Basics of district system design ........................................................................................ 11
4.3.2. Stratified deposits ............................................................................................................. 12
Longwall systems........................................................................................................... 13
Room (bord) and pillar systems ..................................................................................... 14
4.3.3. Orebody deposits.............................................................................................................. 17
4.4. AUXILIARY SYSTEMS ................................................................................21

4.4.1. Line brattices and duct systems ....................................................................................... 21
4.4.2. Forcing, exhausting, and overlap systems. ...................................................................... 23
4.4.3. Air movers ......................................................................................................................... 25
4.5. CONTROLLED PARTIAL RECIRCULATION ..............................................25

4.5.1. Background and principles of controlled partial recirculation ........................................... 25
4.5.2. Controlled recirculation in headings.................................................................................. 26
4.5.3. District systems................................................................................................................. 29
Positions of fans............................................................................................................. 29
Pollution levels ............................................................................................................... 31
4.6. UNDERGROUND REPOSITORIES ............................................................33

4.6.1. Types of repository ........................................................................................................... 33
4.6.2. Ventilation circuits in repositories for nuclear waste......................................................... 33
4.6.3. Additional safety features.................................................................................................. 35
BIBLIOGRAPHY .................................................................................................35
4.1. INTRODUCTION
Practically every underground opening is unique in its geometry, extent, geological surroundings,
environmental pollutants and reasons for its formation - natural or man-made. The corresponding
patterns of airflow through those openings are also highly variable. There are, however, certain
features that are sufficiently common to permit classifications of structured ventilation systems and
subsystems to be identified.
In this chapter, we shall discuss the essential characteristics of subsurface ventilation systems, first
on the basis of complete mines and primary airflow routes. The opportunity is taken to introduce
some of the technical terms used by ventilation engineers. The terms chosen are those that are in
common use throughout the English speaking mining countries. Secondly, we shall look at district
systems for more localized areas of a mine. These, in particular, vary considerably depending upon
the geometry of the geologic deposit being mined. Although reference will be made to given mining
methods, the treatment here will concentrate on principles rather than detailed layouts. In most
countries, state or national mining law impacts upon the ventilation layout. System designers must,
as a pre-requisite, become familiar with the governing legislation. In the absence of any relevant
legislation applicable to the location of the mine, or where the engineer perceives it to be
inadequate, then it is prudent to utilize the pertinent laws from another country that has a welldeveloped history of mining legislation.
Thirdly, auxiliary ventilation systems will be examined, these dealing with the ventilation of blind
headings. The chapter also deals with the principles of controlled partial recirculation and the
ventilation of underground repositories for nuclear waste or other stored material.
4.2. MINE SYSTEMS

4.2.1 General principles
Figure 4.1. depicts the essential elements of a ventilation system in an underground mine or other
subsurface facility.
main
fan
downcast
shaft
booster
fan
upcast
shaft
air
crossing
Figure 4.1. Typical elements of a main ventilation system
Fresh air enters the system through one or more downcast shafts, drifts (slopes, adits), or other
connections to surface. The air flows along intake airways to the working areas or places where the
majority of pollutants are added to the air. These include dust and a combination of many other
potential hazards including toxic or flammable gases, heat, humidity, and radiation. The
contaminated air passes back through the system along return airways. In most cases, the
concentration of contaminants is not allowed to exceed mandatory threshold limits imposed by law
and safe for the entry of personnel into all parts of the ventilation system including return airways.
The intake and return airways are often referred to simply as intakes and returns respectively. The
return air eventually passes back to the surface via one or more upcast shafts, or through inclined or
level drifts.
Fans
The primary means of producing and controlling the airflow are also illustrated on Figure 4.1. Main
fans, either singly or in combination, handle all of the air that passes through the entire system.
These are usually, but not necessarily, located on surface, either exhausting air through the system
as shown on Figure 4.1 or, alternatively, connected to downcast shafts or main intakes and forcing
air into and through the system. Because of the additional hazards of gases and dust that may both
be explosive, legislation governing the ventilation of coal mines is stricter than for most other
underground facilities. In many countries, the main ventilation fans for coal mines are required, by
law, to be placed on surface and may also be subject to other restrictions such as being located out
of line with the connected shaft or drift and equipped with "blow-out" panels to help protect the fan in
case of a mine explosion.
Stoppings and Seals
In developing a mine, connections are necessarily made between intakes and returns. When these
are no longer required for access or ventilation, they should be blocked by stoppings in order to
prevent short-circuiting of the airflow. Stoppings can be constructed from masonry, concrete blocks
or fireproofed timber blocks. Prefabricated steel stoppings may also be employed. Stoppings should
be well keyed into the roof, floor and sides, particularly if the strata are weak or in coal mines liable
to spontaneous combustion. Leakage can be reduced by coating the high pressure face of the
stopping with a sealant material and particular attention paid to the perimeter. Here again, in weak or
chemically active strata, such coatings may be extended to the rock surfaces for a few metres back
from the stopping. In cases where the airways are liable to convergence, precautions should be
taken to protect stoppings against premature failure or cracking. These measures can vary from
"crush pads" located at the top of the stopping to sliding or deformable panels on prefabricated
stoppings. In all cases, components of stoppings should be fireproof and should not produce toxic
fumes when heated.
As a short term measure, fire-resistant brattice curtains may be tacked to roof, sides and floor to
provide temporary stoppings where pressure differentials are low such as in locations close to the
working areas.
Where abandoned areas of a mine are to be isolated from the current ventilation infrastructure, seals
should be constructed at the entrances of the connecting airways. If required to be explosion-proof,
these consist of two or more stoppings, 5 to 10 metres apart, with the intervening space occupied by
sand, stone dust, compacted non-flammable rock waste, cement-based fill or other manufactured
material. Steel girders, laced between roof and floor add structural strength. Grouting the
surrounding strata adds to the integrity of the seal in weak ground. In coal mines, mining law or
prudent regard for safety may require seals to be explosion-proof.
Doors and airlocks
Where access must remain available between an intake and a return airway, a stopping may be
fitted with a ventilation door. In its simplest form, this is merely a wooden or steel door hinged such
that it opens towards the higher air pressure. This self-closing feature is supplemented by angling
the hinges so that the door lifts slightly when opened and closes under its own weight. It is also
advisable to fit doors with latches to prevent their opening in cases of emergency when the direction
of pressure differentials may be reversed. Contoured flexible strips attached along the bottom of the
door assist in reducing leakage, particularly when the airway is fitted with rail track.
Ventilation doors located between main intakes and returns are usually built as a set of two or more
to form an airlock. This prevents short-circuiting when one door is opened for passage of vehicles or
personnel. The distance between doors should be capable of accommodating the longest train of
vehicles required to pass through the airlock. For higher pressure differentials, multiple doors also
allow the pressure break to be shared between doors.
Mechanized doors, opened by pneumatic or electrical means are particularly convenient for the
passage of vehicular traffic or where the size of the door or air pressure would make manual
operation difficult. Mechanically operated doors may, again, be side-hinged or take the form of rollup or concertina devices. They may be activated manually by a pull-rope or automatic sensing of an
approaching vehicle or person. Large doors may be fitted with smaller hinged openings for access
by personnel. Man-doors exposed to the higher pressure differentials may be difficult to open
manually. In such cases, a sliding panel may be fitted in order to reduce that pressure differential
temporarily while the door is opened. Interlock devices can also be employed on an airlock to
prevent all doors from being opened simultaneously.
Regulators
A passive regulator is simply a door fitted with one or more adjustable orifices. Its purpose is to
reduce the airflow to a desired value in a given airway or section of the mine. The most elementary
passive regulator is a rectangular orifice cut in the door and partially closed by a sliding panel. The
airflow may be modified by adjusting the position of the sliding panel manually. Louvre regulators
can also be employed. Another form of regulator is a rigid duct passing through an airlock. This may
be fitted with a damper, louvres or butterfly valve to provide a passive regulator or a fan may be
located within the duct to produce an active regulator. Passive regulators may be actuated by
motors, either to facilitate their manual adjustment or to react automatically to monitored changes in
the quantity or quality of any given airflow.
When the airflow in a section of the mine must be increased to a magnitude beyond that obtainable
from the system then this may be achieved by active regulation. This implies the use of a booster fan
to enhance the airflow through that part of the mine. Section 9.6 deals with the subject of booster
fans in more detail. Where booster fans are employed, they should be designed into the system
such that they help control leakage without causing undesired recirculation in either normal or
emergency situations. In some countries, coal mine legislation prohibits the use of booster fans.
Air crossings
Where intake and return airways are required to cross over each other then leakage between the
two must be controlled by the use of an air crossing. The sturdiest form is a natural air crossing in
which the horizon of one of the airways is elevated above the other to leave a sill of strata between
the two, perhaps reinforced by roof bolts, girders or timber boards. A more usual method is to
intersect the two airways during construction, then to heighten the roof of one of them and/or
excavate additional material from the floor of the other. The two airstreams can then be separated by
horizontal girders and concrete blocks, or a steel structure with metal or timber shuttering. Sealants
may be applied on the high pressure side. Control of the airway gradients approaching the air
crossing reduces the shock losses caused by any sudden change of airflow direction. Man-doors
can be fitted into the air-crossing for access.
Completely fabricated air crossings may be purchased or manufactured locally. These can take the
form of a stiffened metal tunnel. Such devices may offer high resistance to airflow and should be
sized for the flow they are required to pass. They are often employed for conveyor crossings.
Another type of air crossing used mainly for lower airflows and which requires no additional
excavation is to course one of the airstreams through one or more ducts that intersect a stopping on
either side of the junction. An advantage of this technique is that the ducted airflow may be further
restricted by passive regulators or up-rated by fans in the ducts.
In all cases, the materials used in the construction of air crossings should be fireproof and capable of
maintaining their integrity in case of fire. Neither aluminium nor any other low melting-point or
combustible material should be employed in an air crossing.
4.2.2 Location of main fans
In the majority of the world's mines, main fans are sited on surface. In the case of coal mines, this
may be a mandatory requirement. A surface location facilitates installation, testing, access and
maintenance while allowing better protection of the fan during an emergency situation. Siting main
fans underground may be considered where fan noise is to be avoided on surface or when shafts
must be made available for hoisting and free of airlocks. A problem associated with underground
main fans arises from the additional doors, airlocks and leakage paths that then exist in the
subsurface.
In designing the main ventilation infrastructure of a mine, a primary decision is whether to connect
the main fans to the upcast shafts, i.e. an exhausting system or, alternatively, to connect the main
fans to the downcast shaft in order to provide a forcing or blowing system. These choices are
illustrated in Figure 4.2 (a and b).
From the time of the shaft bottom furnaces of the nineteenth century, the upcast shaft has,
traditionally, been regarded as associated with the means of producing ventilation. Most mines are
ventilated using the exhaust system. An examination of the alternatives continues to favour a
primary exhaust system in the majority of cases. The choice may be based on the following four
concerns.
Gas control
Figure 4.2 shows that air pressure in the subsurface is depressed by the operation of an exhausting
main fan but is increased by a forcing fan. The difference is seldom more than a few kilopascals. As
strata gases are, typically, held within the rock matrix at gauge pressures of 1000 kPa or more, it is
evident that the choice of an exhausting or forcing system producing a few kilopascals will have little
effect on the rate of gas production from the strata.
reduced
pressure
increased
pressure
neutral
pressure
(a) Exhausting system
(b) Forcing system
(c) Push-pull system
Figure 4.2 Possible locations of main fans

5
Unfortunately, much of the gas is not emitted directly into the ventilating airstream but collects in
worked out areas, relaxed strata or in voidage that is connected to, but not part of the main
ventilation system. Such accumulations of gas are at near equilibrium pressure with the adjacent
airways. Hence, any reduction in barometric pressure within the ventilation system will result in an
isothermal expansion of the accumulated gases and produce a transient emission of those gases
into the ventilation system. This occurs naturally during periods of falling barometric pressure on the
surface. In some countries, a mandatory record of surface barometric readings is updated at the
beginning of each shift at coal mines. During a period of falling barometric pressure, unusually heavy
emissions of methane or deoxygenated air may be expected. At such times, it is not wise to sit down
for lunch against a seal or stopping beyond which are old workings.
Steady state operation of either exhausting or forcing fans will produce no changes of air pressure in
the subsurface. However, consider the situation of the stoppage of a forcing fan. The barometric
pressure throughout the ventilation system will fall rapidly. Accumulations of voidage gas will expand
and flood into the workings at the worst possible time i.e. when the airflow is considerably
diminished, causing a peak concentration of gas. Conversely, when a main exhausting fan stops, air
pressure in the system increases, compressing accumulations of gas. Hence, no peak of general
body gas concentration occurs within the airstream. It is true that when the main exhaust fan
restarts, the sudden reduction in barometric pressure will then cause expansion and emission of
accumulated gas. However, this occurs at a time of full ventilation and the peak of gas concentration
will be much less than that caused by stoppage of a forcing fan. A consideration of strata gas control
favours a primary exhausting system.
Transportation
The choice between main forcing and exhausting systems should take into account the preferred
routes for the transportation of mineral, personnel and materials. Ideally, conveyors, locomotives or
other modes of moving broken rock should not be required to pass through airlocks. Hence, a mine
design that has mineral or rock transportation routes in main intakes and rock hoisting in a downcast
shaft will favour an exhausting system. Alternatively, if there are good reasons to transport mineral in
the returns and upcast shaft then a forcing system may be preferred. This could be necessary, for
example, in evaporite mines producing potash or halite where the hygroscopic nature of the mineral
could give ore handling problems if transported within the variable humidity of intake airways.
In American coal mines, conveyors are normally required to be located in "neutral" airways,
ventilated by air that will neither pass on to working faces nor is returning from work areas. This
system has the advantage that smoke and gases produced by any conveyor fire will not pollute the
working faces. The major disadvantages are the additional potential for leakage and difficulties in
controlling the air quantity along the conveyor routes.
Fan maintenance
An exhausting fan will pass air that carries dust, water vapour, perhaps liquid water droplets, and is
usually at a higher temperature than that of the air entering the mine. The combined effects of
impact and corrosion on impeller blades are much greater on main exhaust fans. Forcing fans
handle relatively clean air and require less maintenance for any given duty. On the other hand,
corrosive air passing up through the headgear of an upcast shaft can cause much damage. This can
be prevented by drawing the air out the shaft into a near surface fan drift by means of a low pressure
- high volume fan.
Fan performance
A forcing fan normally handles air that is cooler and denser than that passing through an exhausting
fan. For any given mass flow, the forcing fan will pass a lower volume flow at a reduced pressure.
The corresponding power requirement is, therefore, also lower for a forcing fan. However, the effect
is not great and unlikely to be of major significance.
A counteracting influence is that forcing fans must be fitted with inlet grilles to prevent the ingress of
birds or other solid objects. These grilles necessarily absorb available energy and result in an
additional frictional pressure drop. Furthermore, the expanding evasee fitted to a main exhaust fan
recovers some of the kinetic energy that would otherwise be lost to the surface atmosphere.
When air is compressed through a fan its temperature is increased. If the air contains no liquid
droplets and there is insignificant heat transfer through the fan casing then the temperature rise is
given as
0.286 T1
T =
P
C
(4.1)
P1
(see derivation in Section 10.6.1.2)
where = fan isentropic efficiency (fractional)
T1 = Absolute temperature at inlet (K)
P1 = Barometric pressure at inlet (Pa) and
P = Increase in absolute pressure across the fan (Pa)
The rise in temperature through a forcing fan will be reflected by an increase in average temperature
in the intake airways. However, heat exchange with the strata is likely to dampen the effect before
the air reaches the work areas (Section 15.2.2.).
Figure 4.2(c) shows a combination of main forcing and exhausting fans, known descriptively as a
push-pull system. A primary application of a main push-pull system is in metal mines practicing
caving techniques and where the zone of fragmented rock has penetrated through to the surface.
Maintaining a neutral pressure underground with respect to surface minimizes the degree of air
leakage between the workings and the surface. This is particularly important if the rubbelized rock is
subject to spontaneous combustion. In cold climates, drawing air through fragmented strata
intentionally can help to smooth out extremes of temperature of the intake air entering the workings
(Section 18.4.6.).
In the more general case of multi-shaft mines, the use of multiple main fans (whether exhausting,
forcing or push-pull) offers the potential for an improved distribution of airflow, better control of both
air pressures and leakage, greater flexibility and reduced operating costs. On the other hand, these
advantages may not always be realized as a multi-fan system requires particularly skilled
adjustment, balancing and planning.
4.2.3. Infrastructure of main ventilation routes.
Although the simplified sketch of Figure 4.1 depicts the main or trunk intakes and returns as single
airways, this is seldom the case in practice other than for small mines. In designing or examining the
underground layout that comprises a subsurface ventilation system, the following matters should be
addressed:
Mine resistance
For any given total airflow requirement, the operational cost of ventilation is proportional to the
resistance offered to the passage of air (Section 9.5.5.2.). This resistance, in turn, depends upon the
size and number of the openings and the manner in which they are interconnected. Problems of
ground stability, air velocity and economics limit the sizes of airways. Hence, multiple main intakes
and returns are widely employed.
The mine resistance is greatly reduced and environmental conditions improved by providing a
separate split of air to each working panel. The advantages of parallel circuits over series ventilation
were realized early in the nineteenth century (Section 1.2).
Leakage control
The volumetric efficiency of a mine is defined as
VE =
Airflow usefully employed

100 per cent
Total airflow through main fans
(4.2)
where the 'Airflow usefully employed' is the sum of the airflows reaching the working faces and those
used to ventilate equipment such as workshops, electrical gear, pumps or battery charging stations.
The volumetric efficiency of mines may vary from 75 down to less than 10 per cent. The latter value
indicates the large and, often, expensive amount of air leakage that can occur in a mine. It is,
therefore, important to design a subsurface ventilation system to minimize leakage potential and to
maintain the system in order to control that leakage. Whenever possible, intake and return airways,
or groups of airways, should be separated geographically or by barrier pillars with a minimum of
interconnections.
A prerequisite is that all doors, stoppings, seals and air crossings should be constructed and
maintained to a good standard. A stopping between a main intake and return that has been
carelessly holed in order to insert a pipe or cable, or one that has been subject to roadway
convergence without the necessary repairs may be a source of excessive leakage. Unfortunately, if
a large number of stoppings exist between an intake and adjacent return then the leakage may
become untenable even when each individual stopping is of good quality. This can occur in workings
that have been developed by room and pillar methods. The reason for this is the dramatic decrease
in effective resistance to airflow when the flow paths are connected in parallel. For n stoppings
constructed between two adjacent airways, their combined (effective) resistance becomes
Reff =
R
n2
(see section 7.3.1.1)
where R is the resistance of a single stopping.

Figure 4.3 shows the dramatic reduction in effective resistance that occurs as the number of
stoppings increases. In such cases, it becomes important not only to maintain good quality stoppings
but also to design the system such that pressure differentials between the airways are minimized.
Air pressure management is a powerful tool in controlling leakage and, hence, the effectiveness,
volumetric efficiency and costs of a ventilation system. It is particularly important for mines that are
liable to spontaneous combustion. Ideally, resistance to airflow should be distributed equitably
between intakes and returns. In practice, one often observes return airways of smaller cross section
than intakes and that have been allowed to deteriorate because they are less frequently used for
travelling or transportation. This will increase the pressure differentials between intakes and returns.
Similarly, local obstructions caused by falls of roof, stacked materials, equipment or parked vehicles
will affect the pressure distribution and may exacerbate leakage. The positions and settings of
booster fans or regulators also have a marked influence on leakage patterns and should be
investigated thoroughly by network analysis (Chapter 7) during design procedures.
Figure 4.3. Equivalent resistance of sets of stoppings between a pair of adjacent entries.
R = Resistance of one single stopping
Direction of airflow
There are two considerations regarding the direction of the airstream - first with respect to the
transportation of the mined material. An antitropal system is one in which the airflow and transported
rock move in opposite directions. Other than the "neutral" airways of American conveyor roads, this
implies mineral transportation in intake airways. Conversely, a homotropal system is one in which
the airflow and the fragmented rock move in the same direction. This implies mineral transportation
in the return airways and is often associated with a main forcing system. The homotropal system
ensures that any pollution generated from the fragmented rock along the transportation route passes
directly out of the mine without affecting working faces. Such pollution may include dust, heat,
humidity and gases issuing from the broken rock or equipment. The higher relative velocity between
conveyed material and the airflow in an antitropal system can result in a greater entrainment of dust
particles within the airstream. Furthermore, a homotropal system is preferable in the event of a fire
occurring along the mineral transportation route. On the other hand, siting electrical or other
equipment capable of igniting a methane-air mixture in a return airway may be inadvisable or,
indeed, illegal for gassy mines.
The second concern in the matter of airflow direction is the inclination of the airway. An ascentional
ventilation system implies that the airflow moves upwards through inclined workings. This takes
advantage of the natural ventilating effects caused by the addition of heat to the air (Section 8.3.1).
In open stoping or mining layouts that involve multiple connections in inclined workings, ascentional
ventilation may be the only technique capable of either controlling or utilising natural ventilating
effects. Descentional ventilation may be employed on more compact mining systems such as
longwall faces and normally then becomes also a homotropal system with both air and conveyed
mineral moving downhill. However, this may cause difficulties in controlling the natural buoyancy
effects of methane in waste areas. The advantage claimed for descentional ventilation is that
because the air enters the workings at a higher elevation it is then cooler and drier than if it were first
coursed to the lower end of the workings.
Escapeways
Except for blind headings, there should always be at least two means of egress from each working
place in an underground mine or facility. Preferably, there should be two separate intake routes
designated as escapeways in case of a fire or other emergency. Within this context the term
"separate" is taken to imply that those airways have different and identifiable sources of intake air
such that a source of pollution in one of them will not affect the other - either through leakage or
series ventilation. Nevertheless, at least one return air route must always remain open and
travellable without undue discomfort, to allow for an emergency situation where the working face
itself becomes impossible to traverse.
Escapeways should be marked clearly on maps and by signs underground. Personnel should be
made familiar with those routes through regular travel or organized escape drills. Mining legislation
may dictate minimum sizes for escapeways and the frequency of their inspection.
Airflow travel distance and use of old workings
The routes utilized for main intake and return airflows should be reviewed from the viewpoint of
travel distance and corresponding time taken for a complete traverse by the air. For high strata
temperatures, it is advantageous for intake air to reach the workings as quickly as possible in order
to minimize the gain of heat and humidity. However, this is tempered by air velocity constraints and
ventilation operating costs.
In mines located in cold climates, it may be preferable to encourage natural heating of the intake air
by allowing it to take a circuitous and slow route in order to maximize its exposure to rock surfaces.
Another situation occurs when variations in air humidity or temperature cause problems of slaking
(sloughing) of strata from the roof or sides of the airways or workings. Here again, a case may be
made for the natural air conditioning gained by passing the intake air through a network of older
airways prior to reaching the current work areas (Section 18.4.6.).
The employment of old workings as an integral part of a ventilation system can result in significant
reductions in mine resistance and, hence, the operating costs of ventilation. Furthermore, return air
passing through abandoned areas will help to prevent buildup of toxic, asphyxiating or flammable
gases. However, using old workings in this way must be treated with caution. It is inadvisable to rely
upon such routes as they may be subject to sudden closure from falls of roof. Secondly, travellable
intake and return airways must always be maintained for reasons of safety and, third, old workings
liable to spontaneous combustion must be sealed off and the pressure differentials across them
reduced to a minimum.
From a practical viewpoint where old workings can be employed safely for airflow then it is sensible
to use them. However, during system design exercises they should not be relied upon to provide
continuous airflow routes but, rather, as a bonus in reducing the costs of ventilation. In any event, as
a mine develops, it becomes advisable to seal off old areas that are remote from current workings.
Unless this is done, then overall management and control of the airflow distribution will become
increasingly difficult.
10
returns
intakes
workings
intakes
returns
Figure 4.4 U-tube ventilation
4.3. DISTRICT SYSTEMS

4.3.1. Basics of district system design
Underground ventilation layouts serving one or more districts of a mine may be divided into two
broad classifications, U-tube and through- flow ventilation. Each of these takes on a diversity of
physical configurations depending upon the type of mine and disposition of the local geology.
As illustrated on Figure 4.4 the basic feature of U-tube ventilation is that air flows towards and
through the workings, then returns along airways separated from the intakes by stoppings and
doors. Room and pillar layouts and advancing longwalls tend to be of this type.
Figure 4.5 illustrates the alternative through-flow ventilation system. In this layout, primary intakes
and returns are separated geographically. Adjacent airways are either all (or mainly) intakes or
returns and, hence, reducing the number of leakage paths. There are far fewer stoppings and air
crossings but additional regulation (regulators or booster fans) is required to control the flow of air
through the work area. Practical examples of through-flow ventilation are the parallel flows from
downcast to upcast shafts across the multilevels of a metal mine, or the back-bleeder system of a
retreating longwall.
The simplest possible application of the U-tube system is for a set of twin development headings.
Indeed, the U-tube method is the only one capable of ventilating pilot workings that are advancing
into an unmined area. Through-ventilation requires the prior establishment of one or more
connections between main intake and return airways. Once that has been accomplished then
through-ventilation has several significant advantages. First, leakage of air from intake to return is
greatly reduced. Hence, lower total airflows are required to provide any required ventilation at the
11
working face. Secondly, the parallel airways and, often, shorter total travel distance of the airstream
give a lower district resistance - particularly for workings distant from the main shafts. This permits
reduced ventilating pressures. The combination of lower total airflows and lower ventilating
pressures leads to large reductions in ventilation operational costs. Furthermore, the fan duties will
remain much more stable in a through-flow system than the escalating demands of an advancing Utube layout.
intakes
workings
returns
back returns
intakes
Figure 4.5 Through-flow ventilation system
4.3.2. Stratified deposits

The vast majority of underground mines extracting coal, evaporites or other tabular forms of mineral
deposits normally do so by one of two techniques, longwall or room and pillar (bord and pillar)
mining. While the actual layouts can vary quite significantly from country to country and according to
geological conditions, this Section highlights the corresponding modes of airflow distribution that
may be employed.
12
Longwall systems
The two major features of longwall mining that have influenced the design of their ventilation
systems are, first, the control of methane or other gases that accumulate in the waste (gob) areas
and, second, the high rate of rock breakage on heavily mechanized longwalls that exacerbates the
production of dust, gas, heat and humidity.
Figure 4.6 illustrates some of the ventilation layouts used on longwall districts. Single entry systems
are employed primarily in European coal mines. Figures 4.6 (a and b) show the application of the Utube principle to advancing and retreating longwalls respectively. With the advancing system,
leakage of some of the intake air occurs through the waste area, controlled by the resistance offered
by the roadside packing material and the distribution of resistance and, hence, air pressure around
the district. This can give rise to problems of gob fires in mines liable to spontaneous combustion.
(a) Single entry

advancing
(e) Y-system
(b) Single entry

retreating
(c) Single entry retreating

with back bleeder
(f) Double-Z
system
(d) Double entry retreating

with back bleeder
(g) W-system
Figure 4.6 Classifications of longwall district ventilation systems.
13
Gases from the waste may also flush onto the face leading to unacceptable concentrations toward
the return end. The same difficulty may arise to a lesser extent when the U-tube principle is applied
to a retreating face, the abandoned airways being stopped off as the face retreats.
Figure 4.6(c) shows a single entry longwall with the back (or bleeder) return held open in order to
constrain the gas fringe safely back in the waste area and, hence, prevent flushes of waste gas onto
the face. The system illustrated in (c) is a combination of U-tube and through-flow ventilation.
Figure 4.6(d) illustrates the longwall system more often used in coal mining countries that have a
tradition of room and pillar mining such as the United States, Australia or South Africa. Two or more
entries are driven initially using room and pillar mining, these serving as the lateral boundaries of
retreat longwall panels. Again, back bleeders are used to control waste gas.
Figures 4.6(e), (f) and (g) illustrate a classification of systems for longwall faces where the make of
gas from the face itself is particularly heavy. The Y system provides an additional feed of fresh air at
the return end of the face. This helps to maintain gas concentrations at safe levels along the back
return airway(s). Figure 4.5(d) is, in fact, a double entry through-flow Y system. The double-Z layout
is also a through-flow system and effectively halves the length of face ventilated by each airstream.
The W system accomplishes the same end but is based on the U-tube principle. Both the double-Z
and W systems may be applied to advancing or retreating faces, depending upon the ability of the
centre return to withstand front abutment and waste area strata stresses. Again, in both the double-Z
and W systems, the directions of airflow may be reversed to give a single intake and two returns (or
two sets of multiple returns). This may be preferred if heavy emissions of gas are experienced from
solid rib sides.
Room (bord) and pillar systems
Figure 4.7 shows two methods of ventilating a room and pillar development panel; (a) a bidirectional
or W system in which intake air passes through one or more central airways with return airways on
both sides, and (b) a unidirectional or U-tube system with intakes and returns on opposite sides of
the panel. In both cases the conveyor is shown to occupy the central roadway with a brattice curtain
to regulate the airflow through it. It is still common practice in room and pillar mines to course air
around the face ends by means of line brattices pinned to roof and floor but hung loosely in the
cross-cuts to allow the passage of equipment. An advantage of the bidirectional system is that the
air splits at the end of the panel with each airstream ventilating the operational rooms sequentially
over one half of the panel only. Conversely, in the unidirectional or U-tube system the air flows in
series around all of the faces in turn. A second advantage of the bidirectional system arises from the
fact that ribside gas emission is likely to be heavier in the outer airways. This can become the
dominant factor in gassy coal seams of relatively high permeability necessitating that the outer
airways be returns. In most coal mining countries, legislation requires that gas concentration in
intake airways be maintained at very low levels.
Unfortunately, the bidirectional system suffers from one significant disadvantage. The number of
stoppings required to be built, and the number of leakage paths created between intakes and
returns, are both doubled. In long development panels, the amount of leakage can become
excessive allowing insufficient air to reach the last open cross-cuts (Section 4.2.3). In such
circumstances, attempts to increase the pressure differential across the outbye ends of the panel
exacerbate the leakage and give a disappointing effect at the faces.
The unidirectional system has a higher volumetric efficiency because of the reduced number of
leakage paths. However, in both cases, the line brattices in the rooms offer a high resistance to
airflow compared with an open cross-cut. This is particularly so in the case of the unidirectional
system where the useful airflow is required to pass around all of these high resistance line brattices
in series.
14
(a) Bi-directional (W) system.
(b) Uni-directional (U-tube system).
Figure 4.7 Room and pillar development with line brattices. These result in a
high resistance for the face section.
The imposition of line brattice resistance at the most inbye areas of a mine ventilation system forces
more air to be lost to return airways at all leakage points throughout the entire system. An analogy
may be drawn with a leaky hosepipe. If the end of the pipe is unobstructed then water will flow out of
it freely and dribble from the leakage points. If, however, the end of the pipe is partially covered then
the flow from it will decrease but water will now spurt out of the leakage points.
The problem can be overcome by employing auxiliary fans and ducts either to force air into the
rooms or exhaust air from them (Section 4.4). Figure 4.8 illustrates a room and pillar panel equipped
with exhausting auxiliary ventilation. With such a system, the fans provide the energy to overcome
frictional resistance in the ducts. The effective resistance of the whole face area becomes zero.
Smaller pressure differentials are required between intakes and returns for any given face airflows
and, hence, there is a greatly reduced loss of air through leakage. The electrical power taken by the
auxiliary fans is more than offset by the savings in main fan duties.
A further advantage of employing auxiliary fans is that each room is supplied with its own separate
and controllable supply of air. However, the fans must be sized or ducts regulated such that no
undesired recirculation occurs.
15
The choice between auxiliary fan and duct systems and line brattices in room and pillar workings
should also take into account the height and width of the airways, the size and required mobility of
equipment, the placement of ducts or brattices, the extent of pollution from dust, gas and heat, fan
noise, and visibility within the workings.
Figure 4.8 U-tube room and pillar development panel with auxiliary fans (zero face resistance)..
The systems shown on Figure 4.6 for longwalls each have their counterparts in room and pillar
mining. An example of a retreating double-Z (through-flow) system applied to a room and pillar
section is shown on Figure 4.9. There are, however, significant differences in the ventilation strategy
between the two mining methods. The larger number of interconnected airways and higher leakage
result in room and pillar layouts having lower resistance to airflow than longwall mines. It follows that
room and pillar mines tend to require higher volume flows at lower fan pressures than longwall
systems. Similarly, because of the increased number of airways and leakage paths it is particularly
important to maintain control of airflow distribution paths as a room and pillar mine develops. It is
vital that barrier pillars be left between adjacent panels and to separate the panels from trunk airway
routes. Such barriers are important not only to protect the integrity of the mine in case of pillar failure
but also to provide ventilation control points and to allow sealing of the panel in cases of emergency
or when mining has been completed.
16
Figure 4.9 A retreating room and pillar district using a through-flow ventilation system
4.3.3. Orebody deposits
Metalliferous orebodies rarely occur in deposits of regular geometry. Zones of mineralization appear
naturally in forms varying from tortuous veins to massive irregularly shaped deposits of finely
disseminated metal and highly variable concentration. The mining layouts necessarily appear less
ordered than those for stratified deposits. Furthermore, the combination of grade variation and
fluctuating market prices results in mine development that often seems to be chaotic. The same
factors may also necessitate many more stopes or working places than would be usual in a modern
coal mine, with perhaps only a fraction of them operating in any one shift. Hence, the ventilation
system must be sufficiently flexible to allow airflow to be directed wherever it is needed on a day-byday basis.
17
Ventilation networks for metal mines, therefore, tend to be more complex than for stratified deposits
and are usually also three dimensional. Figure 4.10 illustrates the ventilation strategy of many metal
mines although, again, the actual geometry will vary widely. Air moves in a through-flow manner
from a downcast shaft or ramp, across the levels, sublevels and stopes towards return raises, ramps
or upcast shaft. Airflow across each of the levels is controlled by regulators or booster fans.
Movement of air from level to level, whether through stopes or by leakage through ore passes or old
workings tends to be ascentional in order to utilize natural ventilating effects and to avoid thermally
induced and uncontrolled recirculation.
Figure 4.10 Section showing the principle of through-flow ventilation

applied across the levels of a metal mine.
Airflow distribution systems for individual stopes are also subject to great variability depending upon
the geometry and grade variations of the orebody. There are, however, certain guiding principles.
These are illustrated in Figure 4.11 to 4.13 for three stoping methods. In the majority of cases, where
controlled vertical movement of the air is required, stope airflow systems employ ascentional
through-flow ventilation. Although auxiliary fans and ducts may be necessary at individual
drawpoints, every effort should be made to utilize the mine ventilation system to maintain continuous
18
airflow through the main infrastructure of the stope. Series ventilation between stopes should be
minimized in order that blasting fumes may be cleared quickly and efficiently.
Main return to upcast shaft
Vent
raise
stope
broken ore
or fill
from intake shaft
orepass
or service
raise
Vent
raise
Haulage level
Figure 4.11 Simple ventilation system for shrinkage or cut-and-fill stopes.
sill
sill
Vent. Raise
or ramp
Vent. Raise
or ramp
DrawpoInts
Figure 4.12 Ventilation system for sub-level open stopes
19
Undercut
level
pass
ore
From downcast
shaft or ramp
To upcast
shaft or
ramp
grizzly
level
pass
ore
Haulage
level
AIRFLOW
LEAKAGE
Figure 4.13 Typical ventilation system for a block caving operation.
Leakage through ore passes creates a problem in metal mines as the ore passes may often be
emptied allowing a direct connection between levels. Airflows emerging from ore passes can also
produce unacceptable dust concentrations. Closed ore chutes and instructions to maintain some
rock within the passes at all times are both beneficial but are difficult to enforce in the necessarily
production oriented activities of an operating mine. The design of the ventilating system and
operation of regulators and booster fans should attempt to avoid significant pressure differences
across ore passes. Maintaining an orepass at negative pressure by means of a filtered fan/duct
arrangement can help to control dust at dumping or draw points. Attrition on the sides of ore passes
often enlarges their cross-section and may produce fairly smooth surfaces. When no longer required
for rock transportation, such openings may usefully be employed as low resistance ventilation raises.
Figure 4.13 for a block caving operation illustrates another guideline. Wherever practicable, each
level or sublevel of a stope should be provided with its own through-flow of air between shafts,
ventilation raises or ramps. While vertical leakage paths must be taken into account during planning
exercises, maintaining an identifiable circuit on each level facilitates system design, ventilation
management and control in case of emergency.
20
4.4. AUXILIARY SYSTEMS

Auxiliary ventilation refers to the systems that are used to supply air to the working faces of blind
headings. Auxiliary ventilation may be classified into three basic types, line brattices, fan and duct
systems, and "ductless" air movers. Ideally, auxiliary systems should have no impact on the
distribution of airflows around the main ventilation infrastructure, allowing auxiliary ventilation to be
planned independently from the full mine ventilation network. Unfortunately, this ideal is not always
attained, particularly when line brattices are employed.
4.4.1. Line brattices and duct systems
The use of line brattices was introduced in Section 4.3.2 (Figure 4.7) in relation to room and pillar
workings where they are most commonly employed. It was shown that a major disadvantage of line
brattices is the resistance they add to the mine ventilation network at the most sensitive (inbye)
points, resulting in increased leakage throughout the system. This resistance depends primarily
upon the distance of the line brattice from the nearest side of the airway, and the condition of the
flow path behind the brattice. This is sometimes obstructed by debris from sloughed sides, indented
brattices or, even, items of equipment put out of sight and out of mind, despite legislative prohibitions
of such obstructions. In this section we shall examine the further advantages and disadvantages of
line brattices.
Figure 4.14 shows line brattices used in the (a) forcing and (b) exhausting modes. The flameresistant brattice cloth is pinned between roof and floor, and supported by a framework at a position
some one quarter to one third of the airway width from the nearest side. This allows access by
continuous miners and other equipment. Even with carefully erected line brattices, leakage is high
with often less than a third of the air that is available at the last open cross-cut actually reaching the
face. This limits the length of heading that can be ventilated by a line brattice. The need for line
brattices to be extended across the last "open" cross-cut inhibits visibility creating a hazard where
moving vehicles are involved. The advantages of line brattices are that the capital costs are low in
the short term, they require no power and produce no noise.
Forcing
Exhausting
Figure 4.14 Line brattices used for

auxiliary ventilation.
Forcing
Exhausting
Figure 4.15 Fan and duct systems

of auxiliary ventilation.
21
Figure 4.15 shows the corresponding forcing and exhausting systems using auxiliary fans and
ducting. In most cases, in-line axial fans are used although centrifugal fans are quieter and give
higher pressures for the longer headings. The advantages of an auxiliary fan and duct are that they
provide a more positive and controlled ventilating effect at the face, they cause no additional
resistance to the mine ventilation system nor any consequential leakage throughout the network,
and are much less liable to leakage in the heading itself. For headings longer than some 30 metres,
auxiliary fans are the only practicable means of producing the required airflows. An exhausting duct
also allows the air to be filtered, an advantage for dust control where series ventilation is practiced.
The disadvantages involve the initial capital cost, the need for electrical power at the fans, the space
required for ducts and the noise produced by the fans.
Care should be taken to ensure that the pressure-volume characteristics of the fan are
commensurate with the resistance offered by the duct and the airflow to be passed. The latter is
determined on the basis of the type and magnitude of pollutants to be removed (Chapter 9). The
duct resistance is established as a combination of the wall losses within the duct, shock losses at
any bend or change of cross section and at discharge. The equations employed are those derived
for airway resistance in Section 5.4.
Example.
An airflow of 15 m3/s is to be passed through a 0.9m diameter fibreglass duct, 200m long, with one sharp
right-angled bend. From manufacturer's literature, the friction factor for the duct is 0.0032 kg/m3. Calculate the
total pressure to be developed by the fan and the fan power, assuming a fan efficiency of 60 per cent and an air
density of 1.2 kg/m3.
Solution.
Duct area A = 0.9 2 / 4 = 0.636 m 2
Perimeter per = 0.9
= 2.827 m
Let us first determine the shock loss (X) factors for the system from Appendix A5 (Chapter 5).
Entry: (Section A5.4) In the absence of any inlet fitting, the shock loss factor is given as Xin = 1.0.
This is caused by turbulence as the air enters the duct and should not be confused with the conversion of
static pressure to velocity pressure at entry.
Bend: (Figure A5.1) For a sharp right-angled bend, Xb = 1.2
Exit: (Section A5.4) This is not really a shock loss but represents the kinetic energy of the air provided by the
fan and lost to the receiving atmosphere.
Xex = 1.0
Total shock loss factor:
Xsh = 1.0 + 1.2 + 1.0 = 3.2
Equivalent resistance of shock losses:

X sh
R sh =
2 A2
=
3.2 1.2
2 (0.636) 2
Ns 2
m8
= 4.744
(See equation (5.18) )

Ns 2
m8
Duct resistance:
Rd =
k L per
Ns 2
A3
m8
22
0.0032 200 2.827

(0.636)
Ns 2
= 7.028
m8
Total resistance:
Rtot = R sh + R d
Ns 2
= 4.744 + 7.028 = 11.772
m8
Required fan total pressure:

pt
= Rtot Q 2
= 11.772 15 2 = 2 649
The required fan power is

or
pt Q
Pa
where = fan efficiency
2649 15
= 66.2
1000 0.6
kW
4.4.2. Forcing, exhausting, and overlap systems.

Figures 4.14 and 4.15 illustrate forcing and exhausting systems of auxiliary ventilation for line
brattices and fan/duct systems respectively. The choice between forcing and exhausting
arrangements depends mainly upon the pollutants of greatest concern, dust, gases or heat.
The higher velocity airstream emerging from the face-end of a forcing duct or, to a lesser extent, a
forcing brattice gives a scouring effect as the air sweeps across the face. This assists in the
turbulent mixing of any methane that may be emitted from fragmented rock or newly exposed
surfaces. It also helps to prevent the formation of methane layers at roof level (Chapter 12). In hot
mines, the forcing system provides cooler air at the face, even having taken the energy added by the
fan into account. Furthermore, as the system is under positive gauge pressure, the cheaper type of
flexible ducting may be used. This is also easier to transport and enables leaks to be detected more
readily.
The major disadvantage of a forcing system is that pollutants added to the air at the face affect the
full length of the heading as the air passes back, relatively slowly, along it.
Where dust is the main hazard, an exhausting system is preferred. The polluted air is drawn directly
into the duct at the face-end allowing fresh air to flow through the length of the heading. However,
the lack of a jet effect results in poor mixing of the air. Indeed, unless the end of the duct or brattice
line is maintained close to the face then local pockets of sluggish and uncontrolled recirculation may
occur. In all cases, it is important that the ducting or brattice line be extended regularly so that it
remains within some three metres of the face. This distance may be prescribed by legislation.
A further advantage of a ducted exhaust system is that a dust filter may be included within the
system. In this case, the additional pressure drop across the filter must be taken into account in
choosing the fan, and the filter serviced regularly in order that its resistance does not become
excessive. Exhaust ducts must necessarily employ the more expensive rigid ducting or reinforced
flexible ducting. If the exit velocity from an exhaust duct is high then an induction effect (Section
4.3.3), akin to a low pressure booster fan, can result in unexpected consequences including
recirculation in the main entries.
23
For long headings, the resistance of the duct may become so great that multiple fans connected in
series must be employed. If these are grouped as a cluster at the outbye end of the ducting then the
high (positive or negative) gauge pressure will exacerbate leakage. It is preferable to space the fans
along the length of the ducting in order to avoid excessive gauge pressures. The use of hydraulic
gradient diagrams assists in the optimum location of fans and to prevent uncontrolled recirculation of
leakage air. Multiple fans must be interlinked electrically and airflow or pressure monitors employed
to detect accidental severing or blockage of the duct, in which case all fans inbye the point of
damage must be switched off - again, to prevent uncontrolled recirculation.
It is clear that forcing and exhausting systems both have their advantages and disadvantages. Twoway systems have been devised that can be switched from forcing mode to exhausting for cyclic
mining operations. These may employ a reversible axial fan or, alternatively, both a forcing and an
exhausting fan, only one of which is operated at any one time with an appropriate adjustment of
valves or shutter doors within the duct arrangement.
The more common methods of combining the advantages of forcing and exhausting ducts are
overlap systems. Examples are shown on Figure 4.16. The direction and mean velocity of the air in
the heading within the overlap zone clearly depends upon the airflows in each of the ducts. These
should be designed such that the general body airflow in this region does not become unacceptably
low. Where permitted by law, controlled recirculation may be used to advantage in overlap systems
(Section 4.5.2). Where continuous miners or tunnelling machines are employed, the overlap
ventilator may be mounted on the machine. In all cases, it is important that the fans are interlinked
so that the overlap system cannot operate when the primary duct fan is switched off.
(a) Forcing system with exhaust overlap
(b) Exhausting system with force overlap
Figure 4.16 Overlap systems of auxiliary ventilation
24
4.4.3. Air movers

In addition to conventional ducted systems of auxiliary ventilation, a number of other techniques may
be employed to enhance or control the movement of air within localized areas of a mine or tunnel.
Jet fans, sometimes known as ductless, vortex or induction fans are free standing units that produce
a relatively high velocity outlet airstream. The jet of air produces two effects. First, the reach and
integrity of the air vortex depends upon the velocity at the fan outlet, the size of the heading and
whether the airway is a blind heading or part of a throughflow system. Satisfactory ventilation and
turbulent mixing at the face of a large heading can be obtained from a jet fan sited 100 m outbye.
Secondly, an induction effect occurs at the outer boundaries of the expanding cone of projected air.
This entrains additional air from the surroundings producing a forward moving flow that is greater
than the airflow through the fan itself. The conversion of velocity pressure into static pressure as the
plume of air decelerates also generates a true ventilating pressure. This is seldom greater than
some 20 Pascals but is sufficient to create significant airflows in large, low resistance airways. The
induction effect inhibits excessive recirculation provided that incoming intake air is provided at the
fan inlet. Jet fans have particular application in large room and pillar operations and may also be
used in series to promote airflow through vehicular tunnels.
An airflow can also be generated by a spray of water giving rise to spray fans. Inertia from the
motion of the water droplets is transmitted to the air by viscous drag and turbulent induction. Spray
fans may be used very effectively to control the local movement of air around rock-winning machines
such as continuous miners or longwall shearers. This assists in the rapid dilution of methane and in
diverting dust-laden air away from operator positions. The effect depends upon the shape, velocity
and fineness of the spray. Although dust suppression sprays also cause air induction it is usually
necessary to add additional sprays if these are to be used for local airflow control. Provided that the
service water is chilled, spray fans are also an efficient means of cooling the air in a work area.
Compressed air injectors are also induction devices. The compressed air is supplied through one or
more forward pointing jets within a cylindrical or shaped tube. The best effect is obtained when the
compressed air is supplied at the throat of a venturi. These devices are noisy and of much lower
efficiency than fans. However, they have a role in areas where electrical power is unavailable.
Mining law may proscribe the use of compressed air for the promotion of airflow in gassy mines as
the high velocity flow of air through the jets can cause the build-up of an electrostatic charge at the
nozzle. This produces the possibility of sparks that could ignite a methane-air mixture.
4.5. CONTROLLED PARTIAL RECIRCULATION

4.5.1. Background and principles of controlled partial recirculation
The idea of recirculating air in any part of a gassy mine has, traditionally, been an anathema to many
mining engineers. Most legislation governing coal mines prohibits any ventilation system or device
that causes air to recirculate. The background to such legislation is the intuitive fear that recirculation
will cause concentrations of pollutants to rise to dangerous levels. A rational examination of
controlled recirculation was carried out by Leach, Slack and Bakke during the 1960's at the Safety
in Mines Research Establishment in England. Those investigators made a very simple and obvious
statement but one that had, to that time, apparently been denied or ignored within the context of air
recirculation. They argued that the general body gas concentration, C, leaving any ventilated region
of a mine is given by
C =
Flow of gas into the region

Flow of fresh air passing through the region
(4.3)
25
The value of C is quite independent of the flowpaths of the air within the region, including
recirculation. It is true, of course, that if the through-flow of fresh air falls while the gas emission
remains constant then the concentration of gas will rise. This would happen, for example, if the air
duct serving a long gassy heading was dislocated while an inbye fan continued to run. This example
illustrates a case of uncontrolled recirculation. The definition of a system of controlled partial
recirculation is one in which a controlled fraction of the air returning from a work area is passed back
into the intake while, at the same time, the volume flow of air passing through the region is
monitored to ensure that it remains greater than a predetermined minimum value.
The advantages of controlled partial recirculation lie in the improved environmental conditions it can
provide with respect to gases, dust and heat, as well as allowing mining to proceed in areas of a
mine that are too distant from surface connections to be ventilated economically by conventional
means.
As illustrated in Section 4.5.2., general body gas concentrations may actually be reduced by
controlled recirculation. Furthermore, the higher air velocities that occur within a recirculation zone
assist in the turbulent mixing of gas emissions, reducing the tendency to methane layering and
diminishing the probability of accumulations of explosive methane-air mixtures.
As with gas, concentrations of respirable dust reach predictable maximum levels in a system of
controlled recirculation and may be reduced significantly by the use of filters. The greater volume of
air being filtered results in more dust being removed. The effect of controlled recirculation on climatic
conditions is more difficult to predict. However, both simulation programming and practical
observations have indicated the improvements in the cooling power of partially recirculated air for
any given value of through-flow ventilation.
As workings proceed further away from shaft bottoms, the cost of passing air along the lengthening
primary intakes and returns necessarily increases. Where new surface connections closer to the
workings are impractical, perhaps because the workings lie beneath the sea or due to great depth
then the cost of conventional ventilation will eventually become prohibitive - even when booster fans
are employed. Using the air more efficiently through controlled partial recirculation then becomes an
attractive proposition. If, for example, the methane concentration returning from a conventionally
ventilated face is 0.3 per cent, and the safe mandatory limit is 1.0 per cent, then the through-flow
provided from the main airways might be reduced to one half giving a methane concentration of 0.6
per cent while maintaining or increasing the face velocities by controlled recirculation.
During the 1970's the concept of controlled partial recirculation gained respectability and is now
practiced by several of the world's mining industries operating, in some cases, by authorized
exemptions from existing legislation.
The greatest disincentive against the introduction of controlled recirculation has been the risk of
combustion gases from a fire being returned to working areas. Further potential problems arise from
a consideration of transient phenomena such as blasting, or rapid changes in barometric pressure
caused by the operation of doors or fans, and the possible resulting peak emissions of gases. The
introduction of failsafe monitoring systems with continuous computer surveillance has revolutionized
the situation (Section 9.6.3.). These self-checking systems involve monitoring the concentration of
gases, air pressure differentials and airflows at strategic locations as well as the operating conditions
of fans and other plant. Fans can be interlinked electrically to obviate the possibility of uncontrolled
recirculation. Should any monitored parameter fall outside prescribed limits then the system will
automatically revert to a conventional non-recirculating circuit. The introduction of reliable monitoring
technology has allowed the advantages of controlled partial recirculation to be realized safely.
4.5.2. Controlled recirculation in headings
The most widespread application of controlled recirculation has been in headings. One of the
disadvantages of the conventional overlap systems shown in Figure 4.16 is the reduction in general
26
body air velocity within the overlap zone. This can be overcome completely by arranging for the
overlap fan to pass an airflow that is greater than that available within the heading, i.e. a system of
controlled recirculation, accompanied by the corresponding monitoring system and electrical
interlocks. This is particularly advantageous when applied to the scheme depicted in Figure 4.16(a)
as filtered air is then available to machine operators as well as throughout the length of the heading.
Figure 4.17 shows two examples of a primary exhaust system configured for controlled recirculation.
In both cases, an airflow Qt (m3/s) is available at the last open cross-cut and contains a gas flow of
Gi (m3/s). An airflow of Qh passes up the heading where a gas emission of Gh is added.
Gh
Gh
Qh
Qh
Qh
Qh
Qh
Qt
Qt + Qh
(a) Exhausting back into the intake
Qt
Qt
Qh Qt
Qt
(b) Exhausting back into the return
Figure 4.17 Controlled recirculation systems for headings.
Let us try to find the maximum general body gas concentrations that will occur in the systems.
Referring to Figure 4.17(a) and using the locations A, B, C and D as identifying subscripts, the
fractional gas concentration, Cg, at position D (leaving the system) must be
CgD =
Gi + Gh
Qt
(In these relationships, it is assumed that the gas flow is much smaller than the airflow).
However, inspection of the figure shows that this must also be the gas concentration at locations A
and B. In particular,
CgB =
Gi + Gh
Qt
27
But gas flow, G = Gas concentration x airflow

(Gi + Gh )
Qh
Qt
GB =
The gas flow in the duct, GC = GB + Gh, must then be

GC
(Gi + Gh )
Qh
Qt
Gh
Hence, the gas concentration in the duct,

CgC =
GC
Qh
Gi + Gh
G
+ h
Qt
Qh
(4.4)
This is the highest general body gas concentration that can occur anywhere within the system
shown on Figure 4.17(a).
Examination of equation (4.4) shows that if the gas flows, Gi and Gh, are fixed and the fresh air
supply, Qt, remains unchanged then the maximum general body gas concentration, CgC, must fall as
Qh is increased - that is, as the degree of recirculation rises. In the limit, at very high Qh, the gas
concentration in the duct tends toward that leaving the system at position D.
In a conventional non-recirculating system, the airflow taken into a heading is often limited to no
more than half of that available at the last open cross-cut, i.e. Qh = 0.5 Qt. Applying these conditions
to equation (4.4) gives
CgC (conventional) =
=
(Gi + Gh )
Gh
+
Qt
0.5 Qt
Gi + 3 Gh
(4.5)
Qt
Recirculation commences when Qh = Qt, giving

GgC (maximum, recirculating) =
=
(Gi + Gh )
G
+ h
Qt
Qt
Gi + 2Gh
Qt
(4.6)
CgC must be less than this at all greater values of Qh, i.e. higher degrees of recirculation. Comparing
equations (4.5) and (4.6) shows that in this configuration the maximum general body gas
concentration is always less using controlled recirculation than with a conventional system. Turning
to Figure 4.17(b), the analysis is even simpler. In this case, the maximum gas concentration (in the
duct) must be the same as that leaving the system (Gi + Gh)/Qt provided that Qh is equal to or
greater than Qt, i.e. controlled recirculation must exist. Here again, this is always less than would be
attainable with a conventional non-recirculating system.
28
A similar analysis for dust concentration, Cd, on the system shown on Figure 4.17 but with no dust
filter gives an analogous expression to that for gas.
Di + Dh
Dh
mg
(4.7)
CdC =
+
Qt
Qh
m3
Again it can be seen that the maximum concentration falls as Qh is increased. For dust it is more
pertinent to state the concentration at position B, i.e. in the main length of the heading: This
becomes
CdB =
Di + Dh
Qt
mg
m3
(4.8)
and is completely independent of the degree of recirculation.

If the filters shown in Figure 4.17(a) remove a fraction, , of the dust in the duct then it can be
shown that the corresponding concentrations become
CdC =
Qh (Di + Dh ) + DhQt
Qh (Qt + Qh )
mg
(4.9)
m3
and
CdB =
Di + (1 ) Dh
Qt + Qh
mg
m3
(4.10)
These equations show that when filters are used, dust concentrations fall throughout the system.
It has been assumed in these analyses that there is no settlement of dust.
Similar relationships can be derived for other configurations of controlled recirculation in
headings.
4.5.3. District systems
The extension of controlled partial recirculation to complete areas of a mine has particular
benefits in decreasing the costs of heating or cooling the air and for workings distant from the
surface connections.
Positions of fans
Figure 4.18 shows simplified schematics illustrating three configurations of fan locations in a
district recirculation system. In each case, the throughflow ventilation in the mains is shown as Qm
with Qc passing from return to intake in the recirculation cross-cut, to give an enhanced airflow of
Qm+ Qc in the workings. The ratio F = Qc/(Qm + Qc) is known as the recirculation fraction. The fan
that creates the recirculation develops a pressure of pr while the pressure differentials applied
across the outbye ends for the three systems shown are po1, po2 and po3 respectively.
The simplest configuration is shown in Figure 4.18(a) with the recirculating fan sited in the crosscut. This maintains the intakes and returns free for travel and unobstructed by airlocks. Locating
the fan in this position will tend to decrease the throughflow, Qm. Hence, if the total flow is to be
maintained, the applied pressure differential must be increased from
2
2
po1 = RmQm
+ Rw Qm
(with no recirculation)
29
where Rm is the combined resistance of the intake and return mains

to
2
po1 = Rm Qm
+ Rw (Qm + Qc )2
(4.11)
with the recirculation shown on Figure 4.18(a). These equations are derived by summing the
frictional pressure drops, p, around the path of the mains (subscript m) and workings (subscript
w) - and by applying the square law p = RQ2, where R = airway resistance (Section 5.2).
workings
workings
Rw
Rw
Qm+ Qc
workings
2
Qm+ Qc
Qm+ Qc
pr
Qc
Qc
Qc
Rc
Rc
pr
Rc
pr
pb
Qm
Qm
Qm
Gi
Rm
Rm
po1
po2
(a) Cross-cut fan

Q = airflow
G = gasflow
(b) In-line fan

R = resistance
Rm
po3
(c) Combined cross-cut and
booster fans
p = ventilating pressure
Figure 4.18. Schematics of district recirculation systems
Similarly, in all three cases (a), (b) and (c) the pressure required of the recirculating fan is given
by summing the frictional pressure drops around the workings and cross-cut
pr = Rc Qc2 + Rw (Qm + Qc )2
(4.12)
In system (b), the fan is located within either the intake or return inby the recirculation cross-cut.
In this position, the fan acts as a district booster fan as well as creating the controlled
recirculation. Hence, the throughflow, Qm, will tend to increase. Alternatively, if the throughflow is
to remain constant then either the outby pressure differential may be reduced to
2
po 2 = Rm Qm
Rc Qc 2
(4.13)
or a regulator can be introduced into either of the mains.
30
System (c) in Figure 4.18 combines a booster fan with a cross-cut fan and is the preferred
configuration where recirculation is employed due to the workings being distant from the surface
connections. In this system, the ventilating pressure applied across the district may be reduced
by the magnitude of the booster fan pressure to maintain a constant Qm. i.e.
2
po 3 = Rm Qm
+ Rw (Qm + Qc )2 pb
(4.14)
The total airpower consumed within the systems is given as the sum of the pQ products for all
airways. If corresponding airflows and the resistances are the same in each of the three systems
then the required total airpowers must also be equal, irrespective of the locations of the fans. In
practice, differences in the efficiencies of the fans will cause variations in the required total
electrical input to the fan motors. This may, however, be of minor significance.
A major consideration in all designs of controlled partial recirculation is that in conditions of an
emergency or plant stoppage, the system must fail-safe and revert to a conventional nonrecirculating configuration. The airflows must then remain sufficient to allow personnel to
evacuate the area safely and for the necessary ameliorative measures to be taken. The detection
of such conditions is provided by monitoring the environmental parameters and the operation of
the fans.
If the recirculating cross-cut fan in system (a) fails, then doors in the cross-cut must close
automatically. The throughflow ventilation will increase, reducing the general body gas
concentration in the return airways. However, the reduced air velocity in the working area will
increase the probability of local accumulations of gas to that of a conventional non-recirculating
system.
Stoppage of the in-line fan of system (b) is more serious. Again, doors in the cross-cut must be
closed but the throughflow of air will decrease, resulting both in diminished airflows in the
workings and also higher general body gas concentrations in the returns. However, this system is
capable of better airflow control than the cross-cut fan. The degree of recirculation may be varied
by modifying the duty of the fan installation (fan speed, vane settings or number of fans
operating), regulating the airflow in either the intake or return inbye the cross-cut or by adjustment
of a bypass path around the fan.
System (c) gives the greatest degree of flexibility. Stoppage of the cross-cut fan and closure of
the corresponding doors will increase the throughflow, Qm, However, should the booster fan fail,
then electrical interlocks should close down the cross-cut fan. Reduced airflows throughout the
system are maintained by the outbye pressure differential. Adjustment of the two fans allows a
much greater degree of independent control of the airflow distribution than either of the systems
(a) or (b).
Pollution levels
Although the general body gas concentration leaving any zone is independent of airflow
distribution within the zone - recirculating or otherwise (Section 4.5.1), any airflow passed from a
return to an intake airway may affect the quality as well as the quantity of the air in that intake.
Referring to Figure 4.18(b) suppose that the incoming intake air contains a gas flow of Gi (m3/s)
and a constant gas emission of Gw occurs in the workings. Let us derive expressions for the
general body gas concentrations in the face return (position 2) and intake (position 1).
The return concentration must be the same as that leaving the complete district and in the crosscut, assuming no other sources of gas emission:
31
Cg 2 = GgC = Cg (main return ) =
Gi + Gw
Qm
(4.15)
(see equation (4.3))

This is independent of the degree of recirculation.
To determine the gas concentration in the intake at position 1, consider first the gas flow passing
through the cross-cut (subscript c)
Gc = Qc CgC
= Qc
(G i + Gw )
Qm
Now the gas flow at position 1 is

G1 = G i + Gc
= Gi +
Qc
(G i + Gw )
Qm
The corresponding concentration is given by dividing by the corresponding airflow, Qm + Qc,

giving
C g1 =
G1
1
=
Qm
(Q m + Qc )
Gi Qm
Qc
(G i + Gw )
+
(Q m + Qc ) (Q m + Qc )
(4.16)
However, if we define the recirculation fraction as

F =
Qc
(Q m + Qc )
where, also,
1 F =
Qm
(Q m + Qc )
Then equation (4.16) becomes

C g1 =
1
Qm
[ (1 F ) Gi
1
[ G i + F Gw
Qm
+ F (G i + Gw ) ]
(4.17)
This verifies the intuitive expectation that as the degree of recirculation, F, increases then the gas
concentration in the intake also increases. However, as F is never greater than 1, and comparing
with equation (4.15), we can see that the intake, or face, gas concentration can never be greater
than the return concentration. Hence, in a district recirculation system the general body gas
concentration at no place is greater than the return general body concentration with or without
recirculation. The maximum allowable methane concentrations in coal mine intakes may be
prescribed by law at a low value such as 0.25 per cent. The value of F should be chosen such
that this limit is not exceeded. Similar analyses may be carried out for dust concentrations. In this
case, drop-out and the use of filters can result in significant reductions in the concentrations of
dust in a system of controlled partial recirculation. However, the enhanced air velocities in the
work area should not exceed some 4 m/s as the re-entrainment of settled particles within the
airstream accelerates rapidly at greater velocities.
32
The climatic conditions within a system of controlled partial recirculation depend not only upon the
airflows and positions/duties of the fans but also upon the highly interactive nature of heat
transfer between the strata and the ventilating airstreams (Chapter 15). The locations, types and
powers of other mechanized equipment, and the presence of free water also have significant
effects. The only practicable means of handling the large numbers of variables is through a
computer program to simulate the interacting physical processes (Chapter 16). Such analyses,
together with practical observations, indicate that wet and dry bulb temperatures at any point may
either increase or decrease when controlled partial recirculation is initiated without air cooling.
The increased air velocities within the recirculation zone enhance the cooling power of the air on
the human body for any given temperature and humidity. However, when controlled recirculation
is practiced in hot mines, it is normally accompanied by cooling of the recirculated air. This is less
expensive and more effective than bulk cooling the intake air in a conventional ventilation system
and significant improvements in climatic conditions may be realized. Again, practical experience
has shown that the higher airflows within a recirculation zone improves the effectiveness of
existing refrigeration capacity.
In closing this section the reader is reminded, once again, that air recirculation may be prohibited
by the governing legislation. The relevant statutes should be read, and/or enforcement agencies
consulted before instituting a system of controlled partial recirculation.
4.6. UNDERGROUND REPOSITORIES

4.6.1. Types of repository
Underground space is increasingly being utilized for purposes other than the extraction of
minerals or for transportation. The high cost of land, overcrowding and aesthetic considerations
within urban areas encourages use of the subsurface for office accommodation, manufacturing ,
warehousing, entertainment facilities and many other purposes. The safety and stability of a well
chosen geologic formation makes underground space particularly suitable for the storage of
materials, varying from foodstuffs and liquid or gaseous fuels to toxic wastes. The design and
operation of environmental systems in such repositories require the combined skills of mine
ventilation engineers and HVAC (heating, ventilating and air-conditioning) personnel. The
repositories must be constructed and operated in a manner that preserves the integrity of the
stored material and also protects the public from hazardous emissions or effluents.
Perhaps the most demanding designs arise out of the perceived need to store radioactive waste
in deep underground repositories. There are basically two types of this waste. First, there is the
transuranic, or low level, radioactive waste such as contaminated clothing, cleaning materials or
other consumable items that are produced routinely by establishments that handle radioactive
materials. Such waste may be compressed into containers which may be stacked within
excavated chambers underground. Secondly, there is the concentrated and highly radioactive
waste produced from some defence establishments and as the plutonium-rich spent fuel rods
from nuclear power stations. This waste may be packed into heavily shielded and corrosion
resistant cylinders and emplaced within boreholes, about one metre in diameter, drilled from
underground airways into the surrounding rock.
4.6.2. Ventilation circuits in repositories for nuclear waste
Figure 4.19 depicts the primary ventilation structure of a high level nuclear waste repository. As in
the figures illustrating mining circuits shown earlier in the chapter, this sketch is conceptual in
nature and is not intended to represent all airways.
33
During the operation of an underground repository two activities must proceed in phase with each
other. One is the mining of the rooms or drifts where the material is to be placed, together with
the excavation of transportation routes, ventilation airways and the other entire infrastructure
required in an underground facility. This is referred to simply as the mining activity. Secondly, the
Figure 4.19 Example of primary ventilation circuits for an underground nuclear waste repository.
hazardous waste material must be transported through the relevant shafts and airways to the
selected rooms for emplacement. Accordingly, this is known as the emplacement activity.
For reasons of environmental safety, the ventilation circuits for mining and emplacement activities
in a nuclear waste repository must be kept separate. Furthermore, any leakage of air through
doors or bulkheads between the two systems must always leave the mining zone and flow into
the emplacement zone - even in the event of the failure of any fan. Figure 4.19 shows how this is
achieved. The mining circuit operates as a through flow forcing system with the main fan(s) sited
at the top of the mining downcast shaft(s). On the other hand, the emplacement circuit operates
as a throughflow exhaust system with the main fans located at the top of the waste upcast
shafts. It should be remembered that within the nomenclature of underground repositories, the
term waste refers to the hazardous waste to be emplaced and not waste rock produced by mining
activities. With this design, any accidental release of radionuclides into the underground
atmosphere is contained completely within the emplacement circuit and will not contaminate the
mining zones.
34
The shaft or surface-connecting ramp used for transporting the nuclear waste underground is not
shown on Figure 4.19. This shaft will normally not form part of the main ventilation system but will
have a limited downcasting airflow which passes directly into a waste main return. Similar
arrangements may be made for the waste transportation routes underground, thus limiting the
potential dispersion of radioactive contamination in the event of a waste container being damaged
during transportation. Separate maintenance and repair shops are provided in the mining and
emplacement circuits
When emplacement activities have been completed in any given room then the ends of that room
may be sealed. In the case of high level nuclear waste, this may result in the envelope of rock
surrounding the airway reaching temperatures in excess of 150 OC, depending upon the rate of
heat emission from the waste, the distance of the canisters from the airway, the thermal
properties of the rock, and thermal induction of water and vapour migration within the strata. If the
drift is to be reopened for retrieval or inspection of any canister then a considerable period of
cooldown by refrigerated air may be required before unprotected personnel can re-enter. To
reduce the time and expense of the cooldown period, the emplaced room may not be completely
sealed but allowed to pass a regulated airflow sufficient to maintain the rock surface temperature
at a controlled level.
4.6.3. Additional safety features
Before any repository for hazardous waste is commissioned it must conform to the strictest
standards of safety and quality assurance in order to protect both the workers and the general
public from chemical or radioactive contamination. In the case of an underground repository,
design safeguards commence with an extensive examination of the suitability of the geologic
formation to act as a natural containment medium. This will involve the physical and chemical
properties of the rock, the presence and natural migration rate of groundwater and the probability
of seismic activity. Other factors that influence the choice of site include population density and
public acceptance of the surface transportation of hazardous waste to the site.
In addition to continuous electronic surveillance of the quality of the atmosphere throughout the
main ventilation routes of a nuclear waste repository, the fans, bulkheads and regulators must be
monitored to ensure that they operate within design limits and that pressure differentials are
maintained in the correct direction at all times. Regulators and doors may be fitted with electrical
or pneumatic actuators suitable for both local and remote operation. Should airborne radioactivity
be detected at any time and any location in the circuit then the air emerging from the top of the
waste upcast shafts is diverted automatically through banks of high efficiency particulate (HEPA)
filters. Separate or additional precautions should also be taken to protect surface buildings and
shaft tops from tornadoes, floods or fall-out from volcanic activity.
Bibliography
Burton, R.C., et al (1984) Recirculation of air in the ventilation and cooling of deep gold mines.
Trans. 3rd. International Congress on Mine Ventilation. Harrogate, England.
Dhar, N. and McPherson M.J. (1987) The effect of controlled recirculation on mine climate. Proc.
5th Ann. Workshop. Generic Mineral Tech. Center, Mine Systems Design and Ground
Control Alabama, USA pp. 139 - 151.
Hardcastle, S.G. (1985) Computer predicted airborne respirable dust concentrations in mine air
recirculation systems. Proc. 2nd U.S. Mine Ventilation Symp. Reno. pp. 239-248.
35
Leach, S.J. and Slack A. (1969) Recirculation of Mine Ventilation Systems. The Mining Engineer
(U.K.) January, No. 100.
Lee, R.D. and Longson, I. (1987) Controlled recirculation of mine air in working districts. Jnl.
Mine Ventilation Soc. of South Africa Vol. 40 No. 2, Feb.
Longson, I, Lee, R.D., and Lowndes, I.S. (1985) The feasibility of controlled air recirculation
around operating longwall coal faces. Proc. 2nd U.S. Mine Ventilation Symp. Reno pp. 227237.
Robinson, R. (1989) Use of booster fans and recirculation systems for environmental control in
British coal mines. 4th U.S. Mine Ventilation Symposium. Berkeley, CA, USA pp. 235-242.
Wallace, K.G et al (1986) Impact of using auxiliary fans on mine ventilation efficiency and cost.
AIME meeting. New Orleans, March 1986.
Wallace, K.G. (1988) Ventilation strategy for a prospective U.S. nuclear waste repository. 4th
International Mine Ventilation Congress, Brisbane, Australia pp. 65-72. 4-33.
36
Chapter 5. Incompressible Flow Relationships.
Chapter 5. INCOMPRESSIBLE FLOW RELATIONSHIPS
5.1. INTRODUCTION ...........................................................................................2

5.2. THE ATKINSON EQUATION AND THE SQUARE LAW ...............................2
5.3. DETERMINATION OF FRICTION FACTOR..................................................4
By analogy with similar airways................................................................................................ 4
From design tables................................................................................................................... 5
From geometric data ................................................................................................................ 5
5.4. AIRWAY RESISTANCE .................................................................................7

5.4.1. Size of airway..................................................................................................................... 8
5.4.2. Shape of airway ................................................................................................................. 9
5.4.3. Airway lining..................................................................................................................... 10
5.4.4. Air density ........................................................................................................................ 10
5.4.5. Shock losses.................................................................................................................... 10
Shock loss factor .................................................................................................................... 10
Equivalent length.................................................................................................................... 11
5.4.6. Mine shafts....................................................................................................................... 13
5.4.6.1. Shaft walls ................................................................................................................. 14
5.4.6.2. Shaft fittings............................................................................................................... 14
Longitudinal Fittings............................................................................................................ 14
Buntons............................................................................................................................... 14
Form Drag and Resistance of Buntons .............................................................................. 14
Interference factor............................................................................................................... 16
5.4.6.3. Conveyances............................................................................................................. 17
Resistance of a Stationary Convevance............................................................................. 17
Dynamic Effects of a Moving Conveyance ......................................................................... 18
5.4.6.4. Entry and exit losses ................................................................................................. 22
5.4.6.5. Total shaft resistance ................................................................................................ 22
5.4.6.6. Methods of reducing shaft resistance ....................................................................... 23
Shaft walls........................................................................................................................... 23
Buntons............................................................................................................................... 23
Longitudinal Fittings............................................................................................................ 23
Cages and Skips................................................................................................................. 23
Intersections and loading/unloadinq stations...................................................................... 24
5.5. AIR POWER ................................................................................................24

BIBLIOGRAPHY .................................................................................................25
APPENDIX A5 ....................................................................................................26
Shock loss factors for airways and ducts. ....................................................................... 26
A.5.1 Bends................................................................................................................................ 27
A5.2 Changes in cross-section .................................................................................................. 29
A5.3 Junctions ........................................................................................................................... 29
A5.4 Entry and Exit .................................................................................................................... 34
A5.5 Obstructions ...................................................................................................................... 35
A5.6. Interaction between shock losses .................................................................................... 38
5- 1
5.1. INTRODUCTION
In Chapter 2 we introduced some of the basic relationships of incompressible fluid flow. With the
exception of shafts greater than 500m in vertical extent, changes in air density along individual
airways may be ignored for most practical ventilation planning. Furthermore, in areas that are
actively ventilated, airflows are turbulent in nature, other than in very large openings.
In this chapter, we shall confine ourselves to incompressible turbulent flow in order to develop
and illustrate the equations and concepts that are most commonly employed in the practice of
subsurface ventilation engineering.
Although the basic relationships that were derived in Chapter 2 may be employed directly for
ventilation planning, John J. Atkinson introduced certain simplifications in his classical paper of
1854. These simplifications facilitate practical application but were achieved at the expense of
precision. As the resulting "laws of airflow" remain in common use, they are introduced and
discussed in this chapter. The important concept of airway resistance is further expanded by an
examination of the factors that influence it.
5.2. THE ATKINSON EQUATION AND THE SQUARE LAW

In seeking to quantify the relationships that govern the behaviour of airflows in mines, Atkinson
utilized earlier work of the French hydraulic engineers and, in particular, the Chezy Darcy
relationship of the form expressed in equation (2.49)
p = fL
per
u2
A
2
Pa
It must be remembered that Atkinson's work was conducted in the middle of the nineteenth
century, some thirty years before Reynold's experiments and long before Stanton, Prandtl and
Nikuradse had investigated the variable nature of the coefficient of friction, f. As far as Atkinson
was aware, f was a true constant for any given airway. Furthermore, the mines of the time were
relatively shallow, allowing the air density, , to be regarded also as constant. Atkinson was then
able to collate the "constants" in the equation into a single factor:
k =
f
2
kg
(5.1)
m3
giving
p = kL
per 2
u
A
Pa
(5.2)
This has become known as Atkinson's Equation and k as the Atkinson friction factor. Notice that
unlike the dimensionless Chezy Darcy coefficient, f, Atkinson's friction factor is a function of air
density and, indeed, has the dimensions of density.
Atkinson's equation may be written in terms of airflow, Q = uA,
giving
per
p = k L 3 Q2
Pa
A
(5.3)
Now for any given airway, the length, L, perimeter, per, and cross-sectional area, A, are all
known. Ignoring its dependence upon density, the friction factor varies only with the roughness of
5- 2
the airway lining for fully developed turbulence. Hence we may collect all of those variables into a
single characteristic number, R, for that airway.
R = kL
per
Ns 2
or
kg
m7
(5.4)
giving
p = RQ2
Pa
(5.5)
This simple equation is known as the Square Law of mine ventilation and is probably the single
most widely used relationship in subsurface ventilation engineering.
The parameter R is called the Atkinson's resistance of the airway and, as shown by the Square
Law, is the factor that governs the amount of airflow, Q, which will pass when a given pressure
differential, p, is applied across the ends of an airway.
The simplicity of the Square Law has been achieved at the expense of precision and clarity. For
example, the resistance of an airway should, ideally, vary only with the geometry and roughness
of that airway. However, equations (5.1) and (5.4) show that R depends also upon the density of
the air. Hence, any variations in the temperature and/or pressure of the air in an airway will
produce a change in the Atkinson resistance of that airway.
Secondly, the frictional pressure drop, p, depends upon the air density as well as the geometry of
the airway for any given airflow Q. This fact is not explicit in the conventional statement of the
Square Law (5.5) as the density term is hidden within the definition of Atkinson resistance, R.
A clearer and more rational version of the Square Law was, in fact, derived as equation (2.50) in
Chapter 2, namely
p = Rt Q 2
Pa
(equation (2.50))
where Rt was termed the rational turbulent resistance, dependent only upon geometric factors
and having units of m-4.
Rt =
f L per
2A
m -4
It is interesting to reflect upon the influence of historical development and tradition within
engineering disciplines. The k factor and Atkinson resistance, R, introduced in 1854 have
remained in practical use to the present time, despite their weakness of being functions of air
density. With our increased understanding of the true coefficient of friction, f, and with many
mines now subject to significant variations in air density, it would be sensible to abandon the
Atkinson factors k and R, and to continue with the more fundamental coefficient of friction f, and
rational resistance, Rt. However, the relinquishment of concepts so deeply rooted in tradition and
practice does not come about readily, no matter how convincing the case for change. For these
reasons, we will continue to utilize the Atkinson friction factor and resistance in this text in
addition to their more rational equivalents. Fortunately, the relationship between the two is
straightforward.
Values of k are usually quoted on the basis of standard density, 1.2 kg/m3. Then equation (5.1)
gives
5- 3
kg
k 1.2 = 0.6 f
(5.6)
m3
and equations (5.5) and (2.50) give

Ns2
R1.2 = 1.2 Rt
(5.7)
m8
Again, on the premise that listed values of k and R are quoted at standard density (subscript 1.2),
equations (5.3) to (5.5) may be utilized to give the frictional pressure drop and resistance at any
other density, .
p = k 1.2 L
R = k1.2 L
and
per
A
per
A3
Q2
1 .2
1. 2
p = R Q2
Pa
Ns2
m8
Pa
(5.8)
(5.9)
(5.10)
5.3. DETERMINATION OF FRICTION FACTOR

The surface roughness of the lining of an underground opening has an important influence on
airway resistance and, hence, the cost of passing any given airflow. The roughness also has a
direct bearing on the rate of heat transfer between the rock and the airstream (Chapter 15).
The coefficients of friction, f, shown on the Moody diagram, Figure 2.7, remain based on the
concept of sand grain (i.e. uniformly distributed) roughness. Furthermore, as shown by equation
(5.6), the Atkinson friction factor, k, is directly related to the dimensionless coefficient of friction, f.
However, the k factor must be tolerant to wide deviations in the size and distribution of asperities
on any given surface.
The primary purpose of a coefficient of friction, f, or friction factor, k, is to facilitate the prediction
of the resistances of planned but yet unconstructed airways. There are three main methods of
determining an appropriate value of the friction factor.
By analogy with similar airways
During ventilation surveys, measurements of frictional pressure drops, p, and corresponding
airflows, Q, are made in a series of selected airways (Chapter 6). During major surveys, it is
pertinent to choose a few airways representative of, say, intakes, returns, conveyor roadways or
particular support systems, and to conduct additional tests in which the airway geometry and air
density are also measured. The corresponding values of the friction factor may then be
calculated, and referred to standard density from equations (5.9) and (5.10) as
k1.2 =
p
Q
A 3 1 .2
L per
kg
m3
(5.11)
Those values of k may subsequently be employed to predict the resistances of similar planned
airways and, if necessary, at different air densities. Additionally, where a large number of similar
airways exist, representative values of friction factor can be employed to reduce the number or
lengths of airways to be surveyed. In this case, care must be taken that unrepresentative
obstructions or blockages in those airways are not overlooked.
5- 4
In practice, where the k factor is to be used to calculate the resistances of airways at similar
depths and climatic conditions, the density correction 1.2/ is usually ignored.
Experience has shown that local determinations of friction factor lead to more accurate planning
predictions than those given in published tables (see following subsection). Mines may vary
considerably in their mechanized or drill-and-blast techniques of roadway development, as well
as in methods of support., Furthermore, many modes of roadway drivage or the influence of rock
cleavage leave roughenings on the surface that have a directional bias. In such cases, the value
of the friction factor will depend also upon the direction of airflow.
From design tables
Since the 1920's, measurements of the type discussed in the previous subsection have been
conducted in a wide variety of mines, countries and airway conditions. Table 5.1 has been
compiled from a combination of reported tests and the results of numerous observations made
during the conduct of unpublished ventilation surveys. It should be mentioned, again, that
empirical design data of this type should be used only as a guide and when locally determined
friction factors are unavailable.
From geometric data
The coefficient of friction, f, and, hence, the Atkinson friction factor, k, can be expressed as a
function of the ratio e/d, where e is the height of the roughenings or asperities and d is the
hydraulic mean diameter of the airway or duct (d = 4A/per). The functional relationships are given
in Section 2.3.6.3 and are illustrated graphically in Figure 5.1 (see, also, Figure 2.7).
For fully developed turbulent flow, the Von Krmn equation gives
f =
k1.2
1
=
0 .6
4 [2 log10 (d e ) +1.14]2
(see equation 2.55)
Here again, the equation was developed for uniformly sized and dispersed asperities on the
surface (sand grain roughness). For friction factors that are determined empirically for nonuniform surfaces, this equation can be transposed, or Figure 5.1 can be used, to find the
equivalent e/d value. For example, a k factor of 0.012 kg/m3 gives an equivalent e/d value of
0.063. In an airway of hydraulic mean diameter 3.5 m, this gives the effective height of asperities
to be 0.22 m.
The direct application of the e/d method is limited to those cases where the height of the
asperities can be measured or predicted. The technique is applicable for supports that project a
known distance into the airway.
Example.
The projection of the flanges, e, in a tubbed shaft is 0.152 m. The wall to wall diameter is 5.84 m. Calculate
the coefficient of friction and Atkinson friction factor.
Solution.
The e/d ratio is
0.152
= 0.026
5.84
From equation (2.55) or Figure 5.1, this gives
f = 0.0135 and k1.2 = 0.0081 kg/m3
5- 5
Chapter 5. Incompressible Flow Relationships
Friction
factor, k
kg/m3
Coefficient of friction, f
(dimensionless)
Rectangular Airways
Smooth concrete lined
Shotcrete
Unlined with minor irregularities only
Girders on masonry or concrete walls
Unlined, typical conditions no major irregularities
Unlined, irregular sides
Unlined, rough or irregular conditions
Girders on side props
Drift with rough sides, stepped floor, handrails
0.004
0.0055
0.009
0.0095
0.012
0.014
0.016
0.019
0.04
0. 0067
0.0092
0.015
0.0158
0.020
0.023
0.027
0.032
0.067
0.004
0.006
0.0075
0.009
0.012
0.016
0.0067
0.01
0.0125
0.015
0.020
0.027
0.010
0.014
0.013
0.005
0.014
0.0045
0.017
0.023
0.022
0.008
0.023
0.0075
Steel Arched Airways

Smooth concrete all round
Bricked between arches all round
Concrete slabs or timber lagging between flanges all round
Slabs or timber lagging between flanges to spring
Lagged behind arches
Arches poorly aligned, rough conditions
Metal Mines
Arch-shaped level drifts, rock bolts and mesh
Arch-shaped ramps, rock bolts and mesh
Rectangular raise, untimbered, rock bolts and mesh
Bored raise
Beltway
TBM drift
Coal Mines: Rectangular entries, roof-bolted

Intakes, clean conditions
Returns, some irregularities/ sloughing
Belt entries
Cribbed entries
0.009
0.01
0.005 to 0.011
0.05 to 0.14
0.015
0.017
0.0083 to 0.018
0.08 to 0.23
Shafts1
Smooth lined, unobstructed
Brick lined, unobstructed
Concrete lined, rope guides, pipe fittings
Brick lined, rope guides, pipe fittings
Unlined, well trimmed surface
Unlined, major irregularities removed
Unlined, mesh bolted
Tubbing lined, no fittings
Brick lined, two sides buntons
Two side buntons, each with a tie girder
0.003
0.004
0.0065
0.0075
0.01
0.012
0.0140
0.007 to 0.014
0.018
0.022
0.005
0.0067
0.0108
0.0125
0.0167
0.020
0.023
0.0012 to 0.023
0.030
0.037
0.035
0.05
0.065
0.058
0.083
0.108
0.0037
0.011
0.0024
0.0021
0.0062
0.018
0.0040
0.0035
Longwall faceline with steel conveyor and powered supports2

Good conditions, smooth wall
Typical conditions, coal on conveyor
Rough conditions, uneven faceline
Ventilation ducting3
Collapsible fabric ducting (forcing systems only)
Flexible ducting with fully stretched spiral spring reinforcement
Fibreglass
Spiral wound galvanized steel
Table 5.1 Average values of friction factors (referred to air density of 1.2 kg/m3) and
coefficients of friction (independent of air density).
Notes:1 See Section 5.4.6. for more accurate assessment of shaft resistance.
2. k factors in excess of 0.015 kg/m3 are likely to be caused by the aerodynamic drag of free
standing obstructions in addition to wall drag.
3. These are typical values for new ducting. Manufacturer's test data should be consulted for
specific ducting. It is prudent to add about 20 percent to allow for wear and tear.
4. To convert friction factor, k (kg/m3) to imperial units (lb. min2/ft4), multiply by 5.39 x 10-7.
5- 6
Figure 5.1 The coefficient of friction varies with the height of roughenings divided by the hydraulic
mean diameter, e/d.
It should be noted that for regularly spaced projections such as steel rings, the effective friction
factor becomes a function of the spacing between those projections. Immediately downstream
from each support, wakes of turbulent eddies are produced. If the projections are sufficiently far
apart for those vortices to have died out before reaching the next consecutive support, then the
projections act in isolation and independently of each other. As the spacing is reduced, the
number of projections per unit length of shaft increases. So, also, does the near-wall turbulence,
the coefficient of friction, f, and, hence, the friction factor, k. At that specific spacing where the
vortices just reach the next projection, the coefficient of friction reaches a maximum. It is this
value that is given by the e/d method. For wider spacings, the method will overestimate the
coefficient of friction and, hence, errs on the side of safe design.
The maximum coefficient of friction is reached at a spacing/diameter ratio of about 1/8.
Decreasing the spacing further will result in "wake interference" - the total degree of turbulence
will reduce and so, also, will the coefficient of friction and friction factor. However, in this
condition, the diameter available for effective flow is also being reduced towards the inner
dimensions of the projections.
5.4. AIRWAY RESISTANCE

The concept of airway resistance is of major importance in subsurface ventilation engineering.
The simple form of the square law p = R Q2 (see equation 5.5) shows the resistance to be a
constant of proportionality between frictional pressure drop, p, in a given airway and the square of
the airflow, Q, passing through it at a specified value of air density. The parabolic form of the
square law on a p, Q plot is known as the airway resistance curve. Examples are shown on
Figure 5.2.
5- 7
The cost of passing any given airflow through an airway varies directly with the resistance of that
airway. Hence, as the total operating cost of a complete network is the sum of the individual
airway costs, it is important that we become familiar with the factors that influence airway
resistance. Those factors are examined in the following subsections.
500
R = 1.0
Frictional pressure drop Pa
450
R = 0.5
R = 0.2
400
350
300
R = 0.1
250
200
150
R = 0.05
100
50
0
0
10
20
30
40
50
60
Airflow Q cu. m per second
Figure 5.2. Airway resistance curves.

5.4.1. Size of airway
Equation (5.4) showed that for a given length of airway, L, and friction factor, k,
R
per
A3
where means "proportional to".
(5.12)
However, for any given shape of cross-section,

per A
(5.13)
Substituting for per in equation (5.12) gives

R
1
A 2 .5
or, for a circular airway

1
R 5
d
(5.14)
(5.15)
5- 8
These two latter proportionalities show the tremendous effect of airway size on resistance.
Indeed, the cross-sectional area open for flow is the dominant factor in governing airway
resistance. Driving an airway at only half its design diameter will result in the resistance being 25
or 32 times greater. Hence, for any required airflow, the cost of passing that ventilation through
the airway will also increase by a factor of 32. It is clear that when sizing an underground opening
that will form part of a main ventilation route, the resistance and ventilation operating costs must
be taken into account. Airway sizing is considered further in Chapter 9.
5.4.2. Shape of airway
Again, for any given length of airway, L, and friction factor, k, the proportionality (5.12) can be rewritten as
R
per
12
(5.16)
52
But for any given shape of cross section,
per
is a constant (see proportionality (5.13)).

A1 2
We term this parameter the shape factor, SF for the airway. Then if all other parameters remain
constant, including cross-sectional area, A, the resistance of an airway varies with respect to its
shape factor.
The planar figure having the minimum possible shape factor is a circle.
SF (circle) =
per
A
12
d
d 4
= 3.5449
All other shapes have a greater shape factor than this value. For this reason, shape factors are
usually normalized with respect to a circle by dividing by 3.5449 and are then quoted as relative
shape factor (RSF) as shown in Table 5.2. The further we depart from the ideal circular shape
then the greater will be the RSF and, hence, the airway resistance.
Shape of Airway
Relative Shape Factor
Circular
1.00
Arched, upright legs

Arched, splayed legs
1.08
1.09
Square
1.13
Rectangular
width:height = 1.5 :1
2 :1
3 :1
4:1
1.15
1.20
1.30
1.41
Table 5.2. Relative Shape Factors

In addition to demonstrating the effect of airway shape on its resistance to airflow, the purpose of
relative shape factors in ventilation planning is now limited to little more than comparing the effect
of shape for proposed airways of given cross-sectional area. It may also be useful as a correction
factor for older nomograms relating airway resistance, area and k factor which were produced on
the basis of a circular cross-section.
5- 9
5.4.3. Airway lining

Equation (5.4) shows that airway resistance is proportional to the Atkinson friction factor, k, and,
hence, is also directly proportional to the more fundamental coefficient of friction, f. The latter
depends only upon the roughness of the airway lining for fully developed turbulent flow.
5.4.4. Air density
The Atkinson resistance, R, as used in the square law, p = RQ2 depends upon the friction factor,
k. However, equation (5.1) shows that k, itself, depends upon the density of the air. It follows that
the Atkinson resistance also varies with the density of the air. On the other hand, the rational
resistance, Rt, as used in the rational expression of the square law, p = Rt Q2 (equation (2.50))
is a function of the geometry and lining of the airway only and is independent of air density.
5.4.5. Shock losses
Whenever the airflow is required to change direction, additional vortices will be initiated. The
propagation of those large scale eddies consumes mechanical energy (shock losses) and, hence,
the resistance of the airway may increase significantly. This occurs at bends, junctions, changes
in cross-section, obstructions, regulators and at points of entry or exit from the system.
The effects of shock losses remain the most uncertain of all the factors that affect airway
resistance. This is because fairly minor modifications in geometry can cause significant changes
in the generation of vortices and, hence, the airway resistance. Analytical techniques may be
employed for simple and well defined geometries. For the more complex situations that arise in
practice, scale models or computational fluid dynamics (CFD) simulations may be employed to
investigate the flow patterns and shock losses.
There are two methods that may be used to assess the additional resistance caused by shock
losses.
Shock loss factor
In text books on fluid mechanics, shock losses are often referred to in terms of the head loss or
drop in total pressure, pshock caused by the shock loss. This, in turn, is expressed in terms of
'velocity heads'.
pshock = X
u2
2
(5.17)
Pa
where = air density (kg/m3)

u = mean velocity of air (m/s) and
X = shock loss factor (dimensionless)
The shock loss factor can be converted into an Atkinson type of resistance, Rshock, by re-writing
equation (5.17) as a square law
pshock =
where
Rshock =
X Q2
= Rshock Q 2
2 A2
Pa
Ns2
2A2
m8
(5.18)
5 - 10
If rational resistances (Rt) are employed then the density term is eliminated and the
corresponding shock resistance becomes simply
Rt ,shock =
X
2A
m- 4
(5.19)
The major cause of the additional resistance is the propagation of vortices downstream from the
cause of the shock loss. Accordingly, in most cases, it is the downstream branch to which the
shock resistance should be allocated. However, for junctions, the cross-sectional area used in
equations (5.18) and (5.19) is usually that of the main or common branch through which all of the
local air-flow passes.
One of the most comprehensive guides to the selection of X factors is contained within the
Fundamentals Handbook of the American Society of Heating, Refrigerating and Air Conditioning
Engineers (ASHRAE). Similar design information is produced by corresponding professional
societies in other countries.
An appendix is given at the end of this chapter which contains graphs and formulae relating to
shock loss factors commonly required in subsurface ventilation engineering.
Equivalent length
Suppose that in a subsurface airway of length L, there is a bend or other cause of a shock loss.
The resistance of the airway will be greater than if that same airway contained no shock loss. We
can express that additional resistance, Rshock, in terms of the length of corresponding straight
airway which would have that same value of shock resistance. This "equivalent length" of shock
loss, may be incorporated into equation (5.9) to give an Atkinson resistance of
per
3 1 .2
A
R = k (L + Leq )
Ns2
m8
(5.20)
The resistance due to the shock loss is

Rshock = k Leq
per
3 1. 2
A
Ns2
m8
(5.21)
Equation (5.20) gives a convenient and rapid method of incorporating shock losses directly into
the calculation of airway resistance.
The relationship between shock loss factor, X, and equivalent length, Leq is obtained by
comparing equations (5.18) and (5.21). This gives
Rshock =
or
Leq =
per
= k Leq 3
2
2A
A 1 .2
1 .2 X A
2 k per
Ns2
m8
m
The equivalent length can be expressed in terms of hydraulic mean diameters, d = 4A/per, giving
Leq =
1 .2 X
d
8k
(5.22)
5 - 11
leading to a very convenient expression for equivalent length,

Leq = 0.15
X
k
hydraulic mean diameters
(5.23)
Reference to Appendix A5 for X factors, together with a knowledge of the expected friction factor
and geometry of the planned airway, enables the equivalent length of shock losses to be included
in airway resistance calculation sheets, or during data preparation for computer exercises in
ventilation planning.
When working on particular projects, the ventilation engineer will soon acquire a knowledge of
equivalent lengths for recurring shock losses. For example, a common rule of thumb is to
estimate an equivalent length of 20 hydraulic mean diameters for a sharp right angled bend in a
clean airway.
Example.
A 4 m by 3 m rectangular tunnel is 450 m long and contains one right-angled bend with a centre-line radius
of curvature 2.5 m. The airway is unlined but is in good condition with major irregularities trimmed from
the sides. If the tunnel is to pass 60 m3/s of air at a mean density 1.1 kg/m3, calculate the Atkinson and
rational resistances at that density and the frictional pressure drop.
Solution.
Let us first state the geometric factors for the airway:
4A
4 12
=
= 3.429 m
per
14
H
3
=
= 0.75
W
4
2 .5
r
=
= 0.625
4
W
d =
hydraulic mean diameter,

height/width ratio,
radius of curvature/width ratio
From Table 5.1, we estimate that the friction factor for this airway will be k = 0.012 kg/m3 at standard
density (f = 0.02). At r/W = 0.625 and H/W = 0.75, Figure A5.2 (at end of chapter) gives the shock loss
factor for the bend to be 0.75.
We may now calculate the Atkinson resistance at the prevailing air density of 1.1 kg/m3 by two methods:
(i)
Calculate the resistance produced by the shock loss separately:
For the airway, equation (5.9) gives
R = kL
per
A3
0.012 450 14 1.1

= 0.0401
=
1. 2
123 1.2
Ns2
m8
referred to density of 1.1 kg/m3
For the bend, equation (5.18) gives

Rshock =
=
2 A2
0.75 1.1
2 12
= 0.00286
Ns2
8
referred to density of 1.1 kg/m3
Then for the full airway
5 - 12
R = Rlength + Rshock
Ns2
0.04010 + 0.00286 = 0.04296
m8
referred to density of 1.1 kg/ m3
(ii) Using the equivalent length of the bend:

Equation (5.23) gives
Leq = 0.15
X
0.15 0.75 3.429
d =
= 32.15 m
k
0.012
Then, from equation (5.20)

R = k (L + Leq )
per
A3
14
1 .1
= 0.012 ( 450 + 32.15) 3
= 0.04296
1 .2
1 .2
12
Ns2
m8
and agrees with the result given by method (i).

If the coefficient of friction, f, and the rational resistance, Rt, are employed, the density terms become
unnecessary and the equivalent length method gives the rational resistance as
Rt = f (L + Leq )
per
2A
= 0.02 ( 450 + 32.15)
14
3
2 12
= 0.03906
m- 4
.The Atkinson resistance at a density of 1.1 kg/m3 becomes

R = Rt = 1.1 0.03906 = 0.04296
Ns2
m8
At an airflow of 60 m3/s and a density of 1.1 kg/m3, the frictional pressure drop becomes
p = R1.1 Q 2 = 0.04296 602 = 155 Pa
or, using the rational resistance,

p = Rt Q 2 = 1.1 0.03906 602 = 155
Pa
5.4.6. Mine shafts

In mine ventilation planning exercises, the airways that create the greatest difficulties in survey
observations or in assessing predicted resistance are vertical and inclined shafts.
Shafts are quite different in their airflow characteristics to all other subsurface openings, not only
because of the higher air velocities that may be involved, but also because of the aerodynamic
effects of ropes, guide rails, buntons, pipes, cables, other shaft fittings, the fraction of cross
section filled by the largest conveyance (coefficient of fill, CF) and the relatively high velocity of
shaft conveyances. Despite such difficulties, it is important to achieve acceptable accuracy in the
estimation of the resistance of ventilation shafts. In most cases, the total airflow supplied
underground must pass through the restricted confines of the shafts. The resistance of shafts is
5 - 13
often greater than the combined effect of the rest of the underground layout. In a deep room and
pillar mine, the shafts may account for as much as 90 percent of the mine total resistance.
Coupled with a high airflow, Q, the frictional pressure drop, p, will absorb a significant part of the
fan total pressure. It follows that the operational cost of airflow (proportional to p x Q) is usually
greater in ventilation shafts than in any other airway.
At the stage of conceptual design, shaft resistances are normally estimated with the aid of
published lists of friction (k) factors (Table 5.1), that make allowance for shaft fittings. For an
advanced design, a more detailed and accurate analysis may be employed. An outline of one
such method is given in this section.
The resistance to airflow, R, offered by a mine shaft is comprised of the effects of four identifiable
components:
(a) shaft walls,
(b) shaft fittings (buntons, pipes etc.),
(c) conveyances (skips or cages), and
(d) insets, loading and unloading points.
5.4.6.1. Shaft walls
The component of resistance offered by the shaft walls, Rtw, may be determined from equations
(5.9 or 2.51) and using a friction factor, k, or coefficient of friction, f, determined by one of the
methods described in Section 5.3. Experience has shown that the relevant values in Table 5.1
are preferred either for smooth-lined walls or where the surface irregularities are randomly
dispersed. On the other hand, where projections from the walls are of known size as, for
example, in the case of tubbed lining, then the e/d method gives satisfactory results.
5.4.6.2. Shaft fittings
It is the permanent equipment in a shaft, particularly cross-members (buntons) that account, more
than any other factor, for the large variation in k values reported for mine shafts.
Longitudinal Fittings
These include ropes, guide rails, pipes and cables, situated longitudinally in the shaft and parallel
to the direction of airflow. Such fittings add very little to the coefficient of friction and,
indeed, may help to reduce swirl. If no correction is made for longitudinal fittings in the calculation
of cross-sectional area then their effect may be approximated during preliminary design by
increasing the k factor (see Table 5.1). However, measurements on model shafts have actually
shown a decrease in the true coefficient of friction when guide rails and pipes are added. It is,
therefore, better to account for longitudinal fittings simply by subtracting their cross-sectional area
from the full shaft area to give the "free area" available for airflow. Similarly, the rubbing surface is
calculated from the sum of the perimeters of the shaft and the longitudinal fittings.
Buntons
The term bunton is used here to mean any cross member in the shaft located perpendicular to
the direction of airflow. The usual purpose of buntons is to provide support for longitudinal fittings.
The resistance offered by buntons is often dominant.
Form Drag and Resistance of Buntons
While the resistance offered by the shaft walls and longitudinal fittings arises from skin friction
(shear) drag, through boundary layers close to the surface, the kinetic energy of the air causes
the pressure on the projected area of cross-members facing the airflow to be higher than that on
the downstream surfaces. This produces an inertial or form drag on the bunton. Furthermore,
the breakaway of boundary layers from the sides of trailing edges causes a series of vortices to
be propagated downstream from the bunton. The mechanical energy dissipated in the formation
and maintenance of these vortices is reflected by a significant increase in the shaft resistance.
5 - 14
The drag force on a bunton is given by the expression:

Drag = CD Ab
u 2
2
(5.24)
where CD = coefficient of drag (dimensionless) depending upon the shape of the bunton,
Ab = frontal or projected area facing into the airflow (m2),
u = velocity of approaching airstream (m/s), and
= density of air (kg/m3) .
The frictional pressure drop caused by the bunton is:
p =
A u2
Drag
= CD b
A
A 2
N
m2
(5.25)
where A = cross-sectional area available for free flow (m2).

However, the frictional pressure drop caused by the buntons is also given by the square law:
N
p = Rb Q 2 = Rtb Q 2 = Rtb u 2 A2
where
m2
(5.26)
Rb = Atkinson resistance offered by the bunton (Ns2/m8),

Rtb = rational resistance of the bunton (m-4), and
Q = airflow (m3/s).
Equating (5.25) and (5.26) gives:

Rtb = CD
Ab
m- 4
2 A3
If there are n buntons in the shaft and they are sufficiently widely spaced to be independent of
each other, then the combined resistance of the buntons becomes:
R t n = n R tb = n C D
Ab
2A
m -4
It is, however, more convenient to consider the spacing, S, between buntons:

S =
L
n
(5.29)
where L = length of shaft (m), giving,

Rt n =
L
S
CD
Ab
2A
m -4
(5.30)
The effect of the buntons can also be expressed as an additional coefficient of friction, f,
or friction factor, kb = 0.6fb, since
5 - 15
fb L per
Rt n =
m- 4
2 A3
[from equation (2.51)]
(5.31)
Equations (5.30) and (5.31) give:

fb =
C D Ab
(dimensionless)
S per
(5.32)
Interference factor
The preceding subsection assumed that the buntons were independent of each other, i.e. that
each wake of turbulent eddies caused by a bunton dies out before reaching the next bunton.
Unless the buntons are streamlined or are far apart, this is unlikely to be the case in practice.
Furthermore, the drag force may be somewhat less than linear with respect to the number of
buntons in a given cross-section, or to Ab at connection points between the buntons and the shaft
walls or guide rails. For these reasons, an 'interference factor', F, is introduced into equation 5.30
(Bromilow) giving a reduced value of resistance for buntons:
Rt n =
A
L
CD b3 F
S
2A
m -4
(5.33)
For simple wake interference, F is a function of the spacing ratio ,

=
S
W
(dimensionless)
(5.34)
where W = width of the bunton.

However, because of the difficulties involved in quantifying the F function analytically, and in
evaluating the other factors mentioned above, the relationship has been determined empirically,
(Bromilow) giving
F = 0.0035 + 0.44
(5.35)
for a range of from 10 to 40.

Combining equations (5.33), (5.34) and (5.35) gives:
Rt n =
L
A
CD b3
S
2A
[ 0.0035
S
W
+ 0.44
m -4
(5.36)
If expressed as a partial coefficient of friction, this becomes:

fb = CD
Ab
S per
[ 0.0035
S
W
+ 0.44
m -4
(5.37)
Values of the coefficient of drag, CD, measured for buntons in shafts are invariably much higher
than those reported for freestanding aerofoil sections. Figure 5.3 indicates drag coefficients for
the shapes normally used for buntons.
5 - 16
Figure 5.3. Coefficients of drag, CD, for buntons in a shaft.

(from information collated by Bromilow)
5.4.6.3. Conveyances
Resistance of a Stationary Conveyance
The main factors that govern the resistance of a stationary cage or skip in a mine shaft are:
(a) The percentage of the free area in the cross section of the shaft occupied by the conveyance.
This is sometimes termed the coefficient of fill", CF.
b) The area and shape of a plan-view of the conveyance and, to a lesser degree, its vertical
height.
5 - 17
The parameters of secondary practical importance include the shape of the shaft and the extent
to which the cage is totally enclosed.
An analysis of the resistances offered by cages or skips commences by considering the
conveyance to be stationary within the shaft with a free (approaching) air velocity of ua (m/s). The
conveyance is treated as an obstruction giving rise to a frictional pressure drop (shock loss) of X
velocity heads:
ua 2
N
2
m2
= frictional pressure drop due to the conveyance (N/m2) and
= velocity of the approaching airstream (m/s)
= air density (kg/m3)
pc = X
where
pc
ua
(5.38)
The effective rational resistance of the stationary cage, Rtc, can be expressed in terms of X from
equation (5.19)
R tc =
X
2A
m 4
(5.39)
where the free shaft area, A, is measured in square metres. The problem now becomes one of
evaluating X.
One of the most comprehensive experimental studies on cage resistance was carried out by A.
Stevenson in 1956. This involved the construction of model cages and measuring the pressure
drops across them at varying airflows in a circular wind tunnel. From the results produced in this
work, Figure 5.4 has been constructed.
All of the curves on Figure 5.4 refer to a cage whose length, L, is 1.5 times the width, W. A
multiplying correction factor may be applied to the value of X for other plan dimensions. Hence,
for a cage that is square in plan, L/W = 1 and the correction nomogram given also on Figure 5.4
shows that X should be increased by 12 per cent.
The shape of the curves indicates the dominant effect of the coefficient of fill, increasing rapidly
after the cage occupies more than 30 per cent of the shaft cross section. The major cause of the
shock loss is form drag and the resulting turbulent wake. Skin friction effects are relatively small.
Hence, so also is the influence of the vertical height of the cage.
The curves refer to cages with covered roof, floor and sides, but open ends. In general, the effect
of covering the ends is to reduce the shock loss. For totally enclosed skips, the value of X given
by the curves should be reduced by some 15 per cent.
Dynamic Effects of a Moving Conveyance
In addition to the presence of a conveyance in a shaft, its motion will also influence the frictional
pressure drop and effective resistance of the shaft. Furthermore, in systems in which two
conveyances pass in the shaft, their stability of motion will depend upon the velocities of the
airflow and the conveyances, the positions of the conveyances within the shaft cross section,
their shape, and the coefficient of fill.
When two synchronized skips or cages pass, they will normally do so at mid-shaft and at the
maximum hoist velocity. The coefficient of fill will, momentarily, be increased (doubled if the two
cages have the same plan area). There will be a very short-lived peak of pressure drop across
the passing point. However, the inertia of the columns of air above and below that point, coupled
with the compressibility of the air, dampen out that peak quickly and effectively.
5 - 18
Figure 5.4. Shock loss factors for a conveyance in a mine shaft.
More important is the effect of passing on the stability of the conveyances themselves. They pass
at a relative velocity of twice maximum hoisting speed, close to each other and usually without
any continuous intervening barrier. Two mechanisms then influence the stability of the cages.
First, thin boundary layers of air exist on the surfaces of each conveyance and moving at the
same velocity. Hence, there will be a very steep velocity gradient in the space between passing
conveyances. Shear resistance (skin friction drag) will apply a braking action on the inner sides of
the conveyances and produce a tendency for the skips or cages to turn in towards each other.
Secondly, the mean absolute velocity in the gap between the passing cages is unlikely to equal
that surrounding the rest of the cages. There exists a variation in static pressure and, hence, an
applied force on the four sides of each conveyance (the venturi effect). This again, will result in a
tendency for lateral movement.
Passengers in conveyances will be conscious not only of the pressure pulse when the cages
pass, but also of the lateral vibrations caused by the aerodynamic effects. The sideways motion
can be controlled in practice by the use of rigid guide rails or, if rope guides are employed, by
using a tensioned tail rope beneath the cages passing around a shaft-bottom pulley.
5 - 19
The aerodynamics of moving conveyances are more complex than for free bodies because of the
proximity of the shaft walls. The total drag on the conveyance will reach a maximum when the
vehicle is moving at its highest speed against the airflow. The apparent frictional pressure drop
and resistance of the shaft will also vary and reach a maximum value at this same time. The
opposite is also true. The drag, apparent frictional pressure drop, and apparent resistance
of the shaft will reach a minimum when the conveyance is moving at maximum speed in the same
direction as the airflow. The amplitude of the cyclic variation depends upon the resistance of the
stationary cage (which in turn varies with the coefficient of fill and other factors as described in
the previous subsection), and the respective velocities of the airflow and the cage.
In the following analysis, expressions are derived to approximate the effective pressure drop and
effective resistance of a moving conveyance.
Consider a conveyance moving at its maximum velocity uc against an approaching airflow of
velocity ua (Figure 5.5).
ua
area A
uc
Figure 5.5. Moving conveyance in a shaft

If the rational resistance of the cage, when stationary, is Rtc (determined using the methodology
of the previous section) then the corresponding pressure drop caused by the stationary cage will
be:
pc = Rtc Q 2 = Rtc ua 2 A2
N
m2
(5.40)
Now suppose that the cage were stationary in the shaft when the approaching airflow has a
velocity of uc. The corresponding pressure drop would be:
p'c = Rtc uc 2 A2
(5.41)
m2
This latter expression approximates the additional pressure drop caused by movement of the
conveyance.
Adding equations (5.40) and (5.41) gives the maximum pressure drop across the moving
conveyance.
5 - 20
p c max = p c + p' c = R tc (u a
+ uc ) A 2
u
2
= R tc u a A 2 1 + c 2
u a
= pc
u
1 + c2
ua
(5.42)
(5.43)
m2
Similarly, the minimum pressure drop, occurring when the cage and airflow are moving in the
same direction is given by:
2
u
pc min = pc 1 c 2
ua
(5.44)
m2
Hence, the cyclic variation in frictional pressure across the shaft is:
2
pc u c / u a
N/m 2
(5.45)
This variation occurs much more slowly than the pressure pulse caused by passing conveyances.
The pressure drop due to a conveyance will:
i.
rise to pc max as the conveyance accelerates to its highest speed against the airflow:
ii.
fall to pc when the conveyance decelerates to rest;
iii.
drop further to pc min as the conveyance accelerates to full speed in the same direction as
the airflow; and again
iv.
rise to pc when the cage comes to rest.
If two cages of equal dimensions are travelling in synchronization within the shaft, but in opposite
directions, then the effect cancels out. Otherwise, the cyclic variation in pressure may be
measurable within the ventilation network and, particularly, at locations close to the shaft. This is
most likely to occur in a single conveyance shaft with a large coefficient of fill. A further effect will
be to impose a fluctuating load on any fans that influence the airflow in that shaft.
The corresponding range of effective resistance for a conveyance moving against the airflow may
also be determined from equation (5.42).
pc max = Rtc (ua 2 + uc 2 ) A 2
However, if the effective resistance of the conveyance moving against the airflow is Rtc max then:
pc max = Rtc max Q 2 = Rtc max ua 2 A2
(5.46)
Equating these two relationships gives:

Rtc max = Rtc
[1 +
uc 2
ua
(5.47)
5 - 21
Similarly, when the airflow and conveyance are moving the same direction:
Rtc max = Rtc
[1 _
uc 2
ua
(5.48)
It may be noted from equations (5.44) and (5.48) that when the conveyance is moving more
rapidly than the airflow (uc > ua), and in the same direction, both the effective pressure drop and
effective resistance of the conveyance become negative. In this situation, the conveyance is
assisting rather than impeding the ventilation. It should be noted that the approximation inherent
in this derivation leads to unacceptable accuracy at low values of ua.
5.4.6.4. Entry and exit losses
In badly designed installations or for shallow shafts, the shock losses that occur at shaft stations
and points of air entry and exit may be greater than those due to the shaft itself. Such losses are,
again, normally quoted on the basis of X velocity heads, using the velocity in the free area of the
shaft. The shock losses at shaft stations may be converted into a rational resistance:
Rt , shock =
(5.49)
2 A2
The shock loss factors, X, may be estimated from the guidelines given in the Appendix to this
chapter.
5.4.6.5. Total shaft resistance
Although interference will exist between components of shaft resistance, it is conservative to
assume that they are additive. Hence, the rational resistance for the total shaft is:
2
u
Rt = Rtw + Rtn + Rtc 1 c 2 + Rt ,shock
ua
walls
fittings
conveyance(s)
m 4
(5.50)
entries/exits
This may be converted to an Atkinson resistance, R, referred to any given air density in the usual
manner:
R = Rt
Ns2
m8
Having determined the total resistance of a shaft, the corresponding effective coefficient of
friction, f, or friction factor k (at any given value of air density) may be determined
f = 2 Rt
and
k =
f
2
A3
L per
(dimensionless)
(5.51)
kg/m3
5 - 22
5.4.6.6. Methods of reducing shaft resistance

Shaft walls
A major shaft utilized for both hoisting and ventilation must often serve for the complete life of the
underground facility. Large savings in ventilation costs can be achieved by designing the shaft for
low aerodynamic resistance.
Modern concrete lining of shafts closely approaches an aerodynamically smooth surface and little
will be gained by giving a specially smooth finish to these walls. However, if tubbing lining is
employed, the wall resistance will increase by a factor of two or more.
Buntons
A great deal can be done to reduce the resistance of buntons or other cross members in a mine
shaft. First, thought should be given to eliminating them or reducing their number in the design.
Second, the shape of the buntons should be considered. An aerofoil section skin constructed
around a girder is the ideal configuration. However, Figure 5.3 shows that the coefficient of drag
can be reduced considerably by such relatively simple measures as attaching a rounded cap to
the upstream face of the bunton. A circular cross section has a coefficient of drag some 60
percent of that for a square.
Longitudinal Fittings
Ladderways and platforms are common in the shafts of metal mines. These produce high shock
losses and should be avoided in main ventilation shafts. If it is necessary to include such
encumbrances, then it is preferable to compartmentalize them behind a smooth wall partition.
Ropes, guides, pipes, and cables reduce the free area available for airflow and should be taken
into account in sizing the shaft; however, they have little effect on the true coefficient of friction.
Cages and Skips
Figure 5.4 shows that the resistance of a cage or skip increases rapidly when it occupies more
than 30 per cent of the shaft cross section. The plan area of a conveyance should be as small as
possible, consistent with its required hoisting duties. Furthermore, a long narrow cage or skip
offers less resistance than a square cage of the same plan area. Table 5.3 shows the effect of
attaching streamlined fairings to the upstream and downstream ends of a conveyance.
No. of
Cage
Decks
Shape of Fairing
Cage Without
Fairings
Upstream
Fairing Only
Downstream
Fairing Only
Fairings at
Both Ends
Aerofoil
2.05
1.53
1.56
0.95
Aerofoil
3.16
2.46
2.71
2.11
Triangular with
ends filled
2.13
1.58
1.59
1.13
Triangular without
ends filled
2.13
1.65
1.79
1.48
Triangular with
ends filled
3.29
2.96
2.92
2.70
Triangular without
ends filled
3.29
2.86
3.18
2.80
Table 5.3 Shock loss factor (X) for a caged fitted with end fairings in a shaft (after Stevenson).
5 - 23
This table was derived from a model of a shaft where the coefficient of fill for the cage was 43.5
per cent. The efficiency of such fairings is reduced by proximity to the shaft wall or to another
cage. There are considerable practical disadvantages to cage fairings. They interfere with cage
suspension gear and balance ropes which must be readily accessible for inspection and
maintenance. Furthermore, they must be removed easily for inspection and maintenance
personnel to travel on top of the cage or for transportation of long items of equipment. Another
problem is that streamlining on cages exacerbates the venturi effect when two conveyances
pass. For these reasons, fairings are seldom employed in practice. However, the simple
expedient of rounding the edges and corners of a conveyance to a radius of about 30 cm is
beneficial.
Intersections and loading/unloadinq stations
For shafts that are used for both hoisting and ventilation it is preferable to employ air bypasses at
main loading and unloading stations. At shaft bottom stations, the main airstream may be diverted
into one (or two) airways intersecting the shaft some 10 to 20m above or below the loading
station. Similarly, at the shaft top, the main airflow should enter or exit the shaft 10 to 20m below
the surface loading point. If a main fan is to be employed on the shaft then it will be situated in the
bypass (fan drift) and an airlock becomes necessary at the shaft top. If no main fan is required at
that location then a high-volume, low-pressure fan may be utilized simply to overcome the
resistance of the fan drift and to ensure that the shaft top remains free from high air velocities.
The advantages or such air bypasses are:
they avoid personnel being exposed to high air velocities and turbulence at
loading/unloading points.
they reduce dust problems in rock-hoisting shafts.
they eliminate the high shock losses that occur when skips or conveyances are stationary
at a heavily ventilated inset.
At the intersections between shafts and fan drifts or other main airways, the entrance should be
rounded and sharp corners avoided. If air bypasses are not employed then the shaft should be
enlarged and/or the underground inset heightened to ensure that there remains adequate freeflow area with a conveyance stationary at that location.
5.5. AIR POWER

In Chapters 2 and 3, we introduced the concept of mechanical energy within an airstream being
downgraded to the less useful heat energy by frictional effects. We quantified this as the term F,
the work done against friction in terms of Joules per kilogram of air. In Section 3.4.2. we also
showed that a measurable consequence of F was a frictional pressure drop, p, where
p
J
kg
= mean density of the air
(3.52)
F =
and
The airpower of a moving airstream is a measure of its mechanical energy content. Airpower may
be supplied to an airflow by a fan or other ventilating motivator but will diminish when the airflow
suffers a reduction in mechanical energy through the frictional effects of viscous action and
turbulence. The airpower loss, APL, may be quantified as
J kg
kg s
APL = F M
where M = Q
(kg/s)
or W
(5.53)
(mass flowrate)
5 - 24
Then
APL = FQ
But
F = p/
from equation (5.52), giving
APL = p Q
(5.54)
As both p and Q are measurable parameters, this gives a simple and very useful way of
expressing how much ventilating power is dissipated in a given airway.
Substituting for p from the rational form of the square law
p = R t Q2
gives
(equation (2.50))
APL = Rt Q3
(5.55)
This revealing relationship highlights the fact that the power dissipated in any given airway and,
hence, the cost of ventilating that airway depend upon
(a) the geometry and roughness of the airway
(b) the prevailing mean air density and
(c) the cube of the airflow.
The rate at which mechanical energy is delivered to an airstream by a fan impeller may also be
written as the approximation
Air power delivered = pft Q
(5.56)
where pft is the increase in total pressure across the fan. For fan pressures exceeding 2.5 kPa,
the compressibility of the air should be taken into account. This will be examined in more detail in
Chapter 10.
Bibliography
ASHRAE Handbook (Fundamentals) (1985) American Society of Heating, Refrigerating and
Air Conditioning Engineers 1791 Tullie Circle, N.E., Atlanta, GA 30329.
Atkinson, J.J. (1854) On the Theory of the Ventilation of Mines. Trans. N. of England, Inst. of
Mining Engineers. Vol. 3 PP. 73-222.
Bromilow, J.G. (1960) The Estimation and the Reduction of the Aerodynamic Resistance of Mine
Shafts. Trans. Inst. of Mining Engineers. Vol. 119 pp. 449-465.
McElroy, G.E. (1935) Engineering Factors in the Ventilation of Metal Mines. U.S. Bureau of
Mines Bull. No. 385.
McPherson, M.J. (1988) An Analysis of the Resistance and Airflow Characteristics of Mine
Shafts. 4th International Mine Ventilation Congress, Brisbane, Australia.
Prosser, B.S. and Wallace, K.G. (1999) Practical Values of Friction Factors. 8th U.S. Mine
Ventilation Symposium. Rolla, Missouri, pp. 691-696.
Stevenson, A (1956) Mine Ventilation Investigations: (c) Shaft pressure losses due to cages.
Thesis, Royal College of Science and Technology, Glasgow, Scotland.
5 - 25
Appendix A5
Shock loss factors for airways and ducts.
Shock loss (X) factors may be defined as the number of velocity heads that give the frictional
pressure loss due to turbulence at any bend, variation in cross sectional area or any other
configuration that causes a change in the general direction of airflow.
pshock = X
u2
2
Pa
[equation (5.17)]
where = air density (kg/m3) and

u = air velocity (m/s)
The shock loss factor is also related to an equivalent Atkinson resistance.
Rshock =
Ns2
2 A2
m8
[equation (5.18)]
where A = cross-sectional area of opening (m2)

or rational resistance
Rt ,shock =
X
2A
[equation (5.19)]
This appendix enables the X factor to be estimated for the more common configurations that
occur in subsurface environmental engineering. However, it should be borne in mind that small
variations in geometry can cause significant changes in the X factor. Hence, the values given in
this appendix should be regarded as approximations.
For a comprehensive range of shock loss factors, reference may be made to the Fundamentals
Handbook produced by the American Society of Heating, Refrigerating and Air-Conditioning
Engineers (ASHRAE).
Throughout this appendix, any subscript given to X refers to the branch in which the shock loss or
equivalent resistance should be applied. However, in the case of branching flows, the conversion
to an equivalent resistance Rsh = X / (2A2) should employ the cross sectional area of the main
or common branch. Unsubscripted X factors refer to the downstream airway.
5 - 26
A.5.1 Bends
Figure A5.1. Shock loss factor for right angled bends of circular cross section
Figure A5.2. Shock loss factor for right angled bends of rectangular cross section.
5 - 27
Figure A5.3. Correction to shock loss factor for bends of angles other than 90
X = X90 x k
Applicable for both round and rectangular cross sections
5 - 28
A5.2 Changes in cross-section

(a) Sudden enlargement:
A2
A1
A = Cross sectional area
u1
u2
u = Air velocity
X2
= 2 1
A1
X1
A
= 1 1
A2
where X2 is referred to Section 2

2
where X1 is referred to Section 1. (Useful if A2 is very large)
(b) Sudden contraction:

A1
A2
u1
u2
A
X 2 = 0 .5 1 2
A1
A5.3 Junctions
The formulae and graphs for shock loss factors at junctions necessarily involve the branch
velocities or airflows, neither of which may be known at the early stages of subsurface ventilation
planning. This requires that initial estimates of these values must be made by the planning
engineer. Should the ensuing network analysis show that those estimates were grossly in error
then the X values and corresponding branch resistances should be re-evaluated and the analysis
run again.
(a) Rectangular main to diverging circular branch (e.g. raises, winzes)
u1
rectangular
u2
round
u
X 2 = 0 .5 1 + 2 .5 2
u1
R2,shock =
X2
Ns2
m8
2 A1
5 - 29
(b) Figure A5.4. Circular main to circular branch (e.g. fan drift from an exhaust shaft).
A =cross-sectional area (m2): u = air velocity (m/s):

Condition: A1 = A3
R2,shock =
R3,shock =
X2
Ns2
2 A1
m8
X3
Ns2
m8
2 A1
5 - 30
(c)
Condition:
A1
R2,shock =
R3,shock =
Figure A5.5. Rectangular main to diverging rectangular branch
A 2 + A3
X2
Ns2
2 A1
m8
X3
Ns2
m8
2 A1
5 - 31
(d) Figure A5.6. Circular branch converging into circular main
A = cross sectional area (m2):

R2,shock =
R 3,shock =
Q = airflow (m3/s)
X2
Ns2
m8
2 A1
X3
Ns 2
m8
2 A1
5 - 32
(e) Figure A5.7. Converging Y junction, applicable to both rectangular

and circular cross-sections.
(For a diverging Y junction, use the branch data from Figure A5.5.)
Conditions:
A 3 = A1 + A 2
A1 = A2
R1,shock =
R 2, shock =
X1
Ns2
m8
X2
Ns 2
m8
2 A3
2 A3
5 - 33
A5.4 Entry and Exit

(a) Sharp-edged entry:
X = 0.5
(b) Entrance to duct or pipe:
X = 1.0
(This is a real loss caused by
turbulence at the inlet and should
not be confused with the
conversion of static pressure to
velocity pressure)
(c) Bell-mouth:
X = 0.03 for r/D 0.2
(d) Exit loss: X
= 1.0
(direct loss of kinetic energy).
5 - 34
A5.5 Obstructions
(a) Obstruction away from the walls (e.g. free-standing roof support).
X = CD
Ab
A
where CD = Coefficient of Drag for the obstruction (see Figure 5.3)

Ab = cross sectional area facing into the airflow
A = area of airway
(b) Elongated obstruction
A2
A1
A1
L
A = cross sectional area open to airflow
Ignoring interaction between constituent shock losses, the effect may be approximated as the
combination of the losses due to a sudden contraction, length of restricted airway and sudden
enlargement. This gives
X2
A
= 1 .5 1 2
A1
2 k L per2
A2
or, when referred to the full cross-section of the airway,

A
X1 = X 2 1
A2
where
k = Atkinson friction factor for restricted length (kg/m3)

per2 = perimeter in restricted length (m)
= air density (kg/m3).
5 - 35
(c) Circular and rectangular orifices (e.g. regulators, door frames)

X1 =
1
Cc
A1
1
A2
where Cc = orifice coefficient (Figure A5.8)
Figure A5.8. Orifice coefficients (of contraction) for sharp edged circular or rectangular orifices.
5 - 36
(d) Figure A5.9. Shock loss factor for screen in a duct.

A = full area of duct
As = open flow area of screen
5 - 37
A5.6. Interaction between shock losses

When two bends or other causes of shock loss in an airway are within some ten hydraulic mean
diameters of each other then the combined shock loss factor will usually not be the simple
addition of the two individual shock loss factors. Depending upon the geometrical
configuration and distance between the shock losses, the combined X value may be either
greater or less than the simple addition of the two. For example, a double bend normally has a
lower combined effect than the addition of two single bends, while a reverse bend gives an
elevated effect. Figure A5.10 illustrates the latter situation. This Figure assumes that the shock
loss furthest downstream attains its normal value while any further shock within the relevant
distance upstream will have its shock loss factor multiplied by an interference factor.
Figure A5.10 Correction (interference) factor for first of two interacting shock losses.
Combined shock loss Xcomb = C X1 + X2
where X1 and X2 are obtained from the relevant graphs.
5 - 38
Chapter 6. Ventilation Surveys
6.1. PURPOSE AND SCOPE OF VENTILATION SURVEYS...............................2

6.2 AIR QUANTITY SURVEYS...........................................................................3
6.2.1. Rotating vane anemometers ............................................................................................. 3
6.2.2. Moving Traverses.............................................................................................................. 3
The traverse ................................................................................................................................... 4
Booking ........................................................................................................................................... 5
6.2.3. Fixed point measurement.................................................................................................. 6
6.2.4. Density correction ............................................................................................................. 6
6.2.5. Swinging vane anemometer (velometer) .......................................................................... 6
6.2.6. Vortex-shedding anemometer........................................................................................... 7
6.2.7. Smoke tubes ..................................................................................................................... 7
6.2.8. Pitot-static tube.................................................................................................................. 7
6.2.9. Fixed point traverses......................................................................................................... 9
Method of equal areas .................................................................................................................. 9
Log-linear traverse ...................................................................................................................... 12
Velocity contours ......................................................................................................................... 12
6.2.10. Hot body anemometers................................................................................................. 13
6.2.11. Tracer gases ................................................................................................................. 14
6.2.12. Measurement of cross-sectional area........................................................................... 16
6.3. PRESSURE SURVEYS ...............................................................................17

6.3.1. Gauge and tube surveys ................................................................................................. 17
Theory ........................................................................................................................................... 17
Practical Procedure..................................................................................................................... 19
6.3.2. Barometer and altimeter surveys .................................................................................... 21
Theory ........................................................................................................................................... 22
Practical Procedure..................................................................................................................... 24
6.4 ORGANIZATION OF PRESSURE-VOLUME SURVEYS............................25

6.4.1. Initial planning ................................................................................................................. 25
6.4.2. Survey management ....................................................................................................... 26
6.4.3. Quality assurance ........................................................................................................... 26
6.5. AIR QUALITY SURVEYS...........................................................................27

BIBLIOGRAPHY ................................................................................................28
APPENDIX A6. ...................................................................................................29

Application of the gauge and tube method of measuring frictional pressure drops in cases of
significant differences in elevation................................................................................................. 29
6 -1
6.1. PURPOSE AND SCOPE OF VENTILATION SURVEYS

A ventilation survey is an organized procedure of acquiring data that quantify the distributions of
airflow, pressure and air quality throughout the main flowpaths of a ventilation system. The
required detail and precision of measurement, and the rigour of the ensuing data analysis depend
upon the purpose of the survey. Perhaps the most elementary observation in underground
ventilation is carried out by slapping one's clothing and watching the dust particles in order to
ascertain the direction of a sluggish airflow.
Measurements of airflow should be taken in all underground facilities at times and places that
may be prescribed by law. Even in the absence of mandatory requirements, a prudent regard for
safety indicates that sufficient routine measurements of airflow be taken
(a) to ensure that all working places in the mine receive their required airflows in an efficient and
effective manner,
(b) that ventilation plans are kept up to date and
(c) to verify that the directions, quantities and separate identity of airflows throughout the
ventilation infrastructure, including escapeways, are maintained.
Similarly, routine measurements of pressure differentials may be made across doors, stoppings
or bulkheads to ensure they are also maintained within prescribed limits and in the correct
direction. The latter is particularly important in underground repositories for toxic or nuclear
materials and where spontaneous combustion may occur.
One of the main differences between a mine ventilation system and ductwork in a building is that
the mine is a dynamic entity, changing continuously due to modifications to the structure of the
network and resistances of individual branches. Regular measurements of airflow and pressure
differentials underground are necessary as a basis for incremental adjustment of ventilation
controls.
During the working life of a mine or other underground facility, there will be occasions when major
modifications are required to be made to the ventilation system. These circumstances include
opening up new districts in the mine, closing off older ones, commissioning new fans or shafts, or
interconnecting main sections of the mine. The procedures of ventilation planning are detailed in
Chapter 9. It is important that planning the future ventilation system of any facility is based on
reliable and verified data. Ventilation surveys that are carried out in order to establish a data base
for planning purposes must necessarily be conducted with a higher degree of organization, detail
and precision than those conducted for routine monitoring and control. This chapter is directed
primarily towards those more accurate surveys.
A major objective of ventilation surveys is to obtain the frictional pressure drop, p, and the
corresponding airflow, Q, for each of the main branches of the ventilation network. From these
data, the following parameters may be calculated for the purposes of both planning and control:
distribution of airflows, pressure drops and leakage

airpower (p x Q) losses and, hence, distribution of ventilation operating costs throughout
the network (Section 9.5.4)
volumetric efficiency of the system (Section 4.2.3)
branch resistances (R = p/Q2)
natural ventilating effects
friction factors (Equation 5.11)
While observations of airflow and pressure differentials are concerned with the distribution and
magnitudes of air volume flow, other measurements may be taken either separately or as an
integral part of a pressure/volume survey in order to indicate the quality of the air. These
measurements may include wet and dry bulb temperatures, barometric pressures, dust levels and
concentrations of gaseous pollutants.
6 -2
6.2 AIR QUANTITY SURVEYS

The volume of air, Q, passing any fixed point in an airway or duct every second is normally
determined as the product of the mean velocity of the air, u, and the cross-sectional area of the
airway or duct, A
Q = u A
m 2
m3
m =
s
s
Most of the techniques of observing airflow are, therefore, combinations of the methods available
for measuring mean velocity and cross-sectional area.
Prior to the invention of anemometers in the nineteenth century, the only practicable means of
measuring rates of airflow in mines was to observe the velocity of visible dust or smoke particles
suspended in the air. An even cruder old method was to walk steadily in the same direction as the
airflow, varying one's pace, until a candle flame appeared to remain vertical. Modern instruments
for the measurement of airspeed in mines divide into three groups depending upon (i) mechanical
effects, (ii) dynamic (velocity) pressure of the airflow and (iii) thermal effects.
6.2.1. Rotating vane anemometers
The vast majority of airspeed measurements made manually underground are gained from a
rotating vane (windmill type) anemometer. When held in a moving airstream, the air passing
through the instrument exerts a force on the angled vanes, causing them to rotate with an angular
velocity that is closely proportional to the airspeed. A gearing mechanism and clutch arrangement
couple the vanes either to a pointer which rotates against a circular dial calibrated in meters (or
feet) or to a digital counter.
The instrument is used in conjunction with a stopwatch and actually indicates the number of
"metres of air" that have passed through the anemometer during a given time period. The clutch
device is employed to stop and start the pointer or digital counter while the vanes continue to
rotate. A zero reset lever is also incorporated into the instrument. Low range vane anemometers
will typically have eight vanes, jewelled bearings and give repeatable readings for velocities in the
range 0.25 to 15 m/s. High range instruments may have four vanes, low-friction roller or ball
bearings and can be capable of measuring air velocities as high as 50 m/s. Digital vane
anemometers indicate directly on an odometer counter, an illuminated screen, or feed an
electronic signal into a data gathering system. Modern handheld instruments may also be fitted
with a microprocessor to memorize readings, dampen out rapid variations in velocity or into which
can be entered the cross-sectional area for the calculation of volume flow. Two types of vane
anemometer are included in the selection of ventilation survey instruments shown on Plate 1.
In order to obtain a reliable measure of the mean air velocity in an underground airway, it is
important that a recommended technique of using the anemometer is employed. The following
procedure has evolved from a combination of experiment and practical experience.
6.2.2. Moving Traverses
The anemometer should be attached to a rod of at least 1.5m in length, or greater for high
airways. The attachment mechanism should permit the options of allowing the anemometer to
hang vertically or to be fixed at a constant angle with respect to the rod. A rotating vane
anemometer is fairly insensitive to yaw and will give results that do not vary by more than 5 per
cent for angles deviating by up to 30 from the direction of the airstream. Hence, for most
6 -3
Hot-wire anemometer
Pitot-static tube
Digital vane
anemometer
Micro-aneroid
barometer
Diaphragm
pressure
gauge
Sling hygrometer
Stopwatch
Dial vane
anemometer
Plate 1. Selection of instruments that may be used in ventilation surveys
underground airways, allowing the anemometer to hang freely at the end of the rod will give
acceptable results. For airways of inclination greater than 30, the anemometer should be
clamped in a fixed position relative to the rod and manipulated by turning the rod during the
traverse such that the instrument remains aligned with the longitudinal axis of the airway.
The traverse
The observer should face into the airflow holding the anemometer rod in front of him/her so that
the dial is visible and at least 1.5m upstream from his/her body. To commence the traverse, the
instrument should be held either in an upper or lower corner of the airway, with the pointer or
counter reset to zero, until the vanes have accelerated to a constant velocity. This seldom takes
more than a few seconds. The observer should reach forward to touch the clutch control lever
while a second observer with a stopwatch counts backwards from five to zero. On zero, the
anemometer clutch is activated releasing the pointer and, simultaneously, the stopwatch is
started.
The path of the traverse across the airway should be similar to that shown on Figure 6.1. The aim
should be to traverse the anemometer at a constant rate not greater than about 15 per cent of the
airspeed. Ideally, equal fractions of the airway cross sectional area should be covered in equal
times. This is facilitated by the stopwatch observer calling out the elapsed time at ten second
intervals. The complete traverse should take not less than 60 seconds and may be considerably
more for large or low velocity airways. The final five seconds should be counted down by the
stopwatch observer during which time the traverse person stretches forward to disconnect the
clutch at the end of the time period. The length indicated by the anemometer is immediately read
and booked, and the instrument reset to zero.
6 -4
Figure 6.1. Path of a moving anemometer traverse
The procedure is repeated, traversing in the opposite direction across the airway. Traverses
should be repeated until three readings are obtained that agree to within 5 per cent. In
favourable steady state conditions, experienced observers will often achieve repeatability to
within 2 per cent. Larger discrepancies may be expected in airways where there is a highly
asymmetric variation in velocity across the airway, where the floor conditions are unstable, or
when obstructions exist in the cross section. Measuring stations should be chosen to avoid such
difficulties wherever possible. Another annoying cause of discrepancy is the opening of a
ventilation door during the period of measurement. Two or more sets of traverses should be taken
at different locations within each airway. Where cross-cuts or other leakage paths affect the
airflow then a sufficient number of additional measurement points should be traversed in order to
quantify the rate and direction of leakage.
Booking
The anemometer field book should be waterproof and laid out such that each double page has
segments for
1.
2.
3.
4.
5.
6.
names of observers
the location of the measuring station, time, date,
anemometer readings and corrections
dimensioned sketch of cross-section
calculation of area
calculation of air volume flow.
6 -5
The bookings are normally made by the stopwatch person. For each traverse, the anemometer
reading is divided by the corresponding time to give the air velocity. The mean of these station
velocities, ignoring any values outside the 5 per cent tolerance, gives the observed mean
velocity. In most cases, the time of each traverse at a station is the same, allowing the
anemometer readings to be averaged before calculating the mean velocity. The observed mean
velocity must then be corrected according to the calibration chart or curve for the instrument
(Section 6.4) to give the actual mean velocity, u.
The cross sectional area, A, is determined using one of the methods discussed in section 6.2.12.
The calculation of airflow is then completed as
m 2
m3
m =
s
s
Q = u A
Anemometer traverses may also be employed at the ends of ducts. However, it is recommended
that the technique not be used for duct diameters less than six times that of the diameter of the
anemometer.
6.2.3. Fixed point measurement
An estimate of duct airflow may be obtained by holding the anemometer at the centre of the duct
and multiplying the corrected reading by a further correction factor of 0.8. A similar technique may
be employed for routine check readings taken at well established measuring stations in airways.
The reading obtained from a stationary anemometer at a known location within the cross-section
should, initially, be compared with that given from a series of traverses in order to obtain a "fixed
point" correction factor for that station. This is typically 0.75 to 0.8 for the fixed point located some
one half to two thirds the height of the airway. Subsequent routine readings may be obtained
simply by taking an anemometer reading at the fixed point and applying the appropriate
calibration and fixed point corrections. Provided that the measuring station is well downstream of
any bends or major obstructions and the airflow remains fully turbulent then the fixed point
correction factor will stay near constant as the airflow varies.
6.2.4. Density correction
For precise work, anemometer readings may be further corrected for variations in air density:
u = ui + Cc
c
m
(6.1)
where
u = corrected velocity
ui = indicated velocity
Cc = correction from instrument calibration curve or chart
c = air density at time of calibration
m = actual air density at time of measurement
Equation (6.1) shows that the density adjustment
c
is effectively applied only to the
m
calibration correction and is ignored in most cases.

6.2.5. Swinging vane anemometer (velometer)
In its most fundamental form, the swinging vane anemometer (velometer) is simply a hinged vane
which is displaced against a spring from its null position by a moving airstream. A connected
6 -6
pointer gives a direct reading of the air velocity. The air enters a port at the side of the instrument.
This port can be fitted with interchangeable orifices or probes to give a range of measurable
velocities. Oscillations of the vane may be reduced by the eddy current damping produced when
a metal strip connected to the vane moves between strong permanent magnets. The delicacy of
the velometer together with its pronounced directional bias has limited its use in underground
surveys. However, it can serve a useful purpose in giving spot readings as low as 0.15 m/s in
gassy mines where hot wire probes are prohibited.
6.2.6. Vortex-shedding anemometer
For continuous monitoring systems, both rotating vane and swinging vane instruments with
electrical outputs have been employed. However, they both require relatively frequent calibration
checks when used in mine atmospheres. For this type of application, the vortex-shedding
anemometer is preferred as it has no moving parts.
When any bluff object is placed in a stream of fluid, a series of oscillating vortices are formed
downstream by boundary layer breakaway, first from one side of the body then the other. The
propagation of the vortices is known as a Krmn street and can often be observed downstream
from projecting boulders in a river. The rate of vortex production depends upon the fluid velocity.
In the vortex-shedding anemometer, the vortices may be sensed by the pulsations of pressure or
variations in air density that they produce. One apparent disadvantage noticed in practice is that
when sited in a fixed location underground for monitoring purposes they require calibration for
that specific location. They may also require electronic damping to eliminate large but short lived
variations in signals caused by the passage of vehicles.
6.2.7. Smoke tubes
Smoke tubes are perhaps the simplest of the mechanical techniques employed for measuring
airflows and are used for very low velocities. A pulse of air forced by a rubber bulb through a
glass phial containing a granulated and porous medium soaked in titanium tetrachloride or
anhydrous tin will produce a dense white smoke. This is released upstream of two fixed marks in
the airway. An observer with a spot-beam cap lamp is located at each mark. The time taken for
the cloud of smoke to travel the length of airway between the marks gives an indication of the
centre-line velocity of the air. This must then be adjusted by a centre-line correction factor to give
the mean velocity. The correction factor is usually taken to be 0.8 although a more accurate value
can be calculated for known Reynold's numbers. The length of airway should be chosen such
that at least one minute elapses during the progression of the smoke between the two marks.
Dispersion of the smoke cloud often causes the downstream observer some difficulty in deciding
when to stop the stopwatch. Due to the uncertainties inherent in the technique, smoke tubes are
normally employed as a last resort in slow moving airstreams.
6.2.8. Pitot-static tube
In section 2.3.2 we discussed the concepts of total, static and velocity pressures of a moving
stream of fluid. A pitot-static tube, illustrated on Figure 6.2, can be used to measure all three. This
device consists essentially of two concentric tubes. When held facing directly into an airflow, the
inner tube is subjected to the total pressure of the moving airstream, pt. The outer tube is
perforated by a ring of small holes drilled at right angles to the shorter stem of the instrument and,
hence, perpendicular to the direction of air movement. This tube is, therefore, not influenced by
the kinetic energy of the airstream and registers the static pressure only, ps. A pressure gauge or
manometer connected across the two tappings will indicate the difference between the total and
static pressure, i.e. the velocity pressure:
pv = p t p s
Pa
[ from equation (2.18) ]
(6.2)
6 -7
Airflow
Figure 6.2 Pitot-static tube

Furthermore, the velocity head is related to the actual velocity of the air, u
u =
where
2 pv
m/s
[ from equation (2.17) ]
(6.3)
actual density of the air (kg/m3)

(see equation (14.52) for air density).
Pitot-static tubes vary widely in overall dimensions. For measuring air velocities in mine airways
or at main fans, the longer stem may be some 1.5m in length. Much smaller versions are
available for use in ducts or pipes.
Modern pitot-static tubes reflect the total, static and velocity pressures of the airflow to an
excellent degree of accuracy. Unfortunately, the precision of the measurement depends also
upon the manometer or pressure gauge connected to the tappings. This imposes a practical
restriction on the lower limit of air velocity that can be measured by a pitot-static tube in the
turbulent airflows of an underground system.
6 -8
Example
If a diaphragm pressure gauge can be read to the nearest 1 Pa, then the lowest pressure that will give 10
per cent accuracy in the pressure reading is 10 Pa. Calculate the air velocity corresponding to a velocity
pressure of 10 Pa, assuming an air density of 1.2 kg/m3.
Solution
u =
2 pv
2 10
= 4.08
1. 2
m/s
As the great majority of underground openings have air velocities of less than 4 m/s, it is clear
that the use of the pitot-static tube for the measurement of air velocity is limited to ventilation
ducting and a few high velocity airways; primarily fan drifts and evasees, ventilation shafts, some
longwall faces and trunk airways.
One of the difficulties of using a pitot-static tube for the spot measurement of pressures or
velocities in a turbulent airstream is the oscillation in the readings. A small wad of cotton wool
inserted into the flexible pressure tubing between the pitot-static tube and the pressure gauge
damps out the short term variations. However, the cotton wool should not be so tightly tamped
into the tubing that the gauge reaction becomes unduly slow. Electronic diaphragm gauges are
often fitted with an internal damping circuit.
6.2.9. Fixed point traverses
The rotating vane anemometer is an integrating device, accumulating the reading as it is
traversed continuously across an airway or duct. Most other instruments for the measurement of
air velocity, including the pitot-static tube, do not have this advantage but are confined to giving a
single spot reading at any one time. In order to find the mean velocity in an airway from pitotstatic tube readings it is, therefore, necessary to take spot measurements at a number of
locations over the cross-section. This procedure is known by the contradictory sounding term
"fixed point traverse". Differing techniques of conducting such traverses vary in the number of
observations, locations of the instrument and treatment of the data. Three of these techniques are
described here. In all cases, the fixed point traverse method assumes that the distribution of flow
over the cross-section does not vary with time. For permanent monitoring stations, a grid of
multiple pitot-static tubes may be left in place.
Method of equal areas
In this method, the cross-section of the duct or airway is divided into subsections each of equal
area. Figure 6.3 shows a rectangular opening divided into 25 equal subsections similar in shape
to the complete opening. Using a pitot-static tube or anemometer, the velocity at the centre of
each subsection is measured. The mean velocity is then simply the average of the subsection
velocities.
There are a few precautions that should be taken to ensure satisfactory results. First, if a pitotstatic tube is employed then the velocity at each subsection should be calculated. Averaging the
velocity pressure before employing equation (6.3) will not give the correct mean velocity.
Secondly, it will be recalled that the velocity gradient changes most rapidly near the walls. Hence,
accuracy will be improved if the velocities for the subsections adjacent to the walls and,
especially, in the corners are determined from a number of readings distributed within each of
those subsections. Third, the number of subsections should increase with respect to the size of
the airway in order to maintain accuracy.
6 -9
Figure 6.3 Measuring points for a fixed point traverse in a rectangular opening. The
velocity in each shaded subsection should be averaged from several
readings distributed over the subsection.
As a guide, the recommended number of points, n, for a rectangular opening may be estimated
as
n = 100 e
+ 23
(6.4)
where e is the exponential exponent, 2.7183 and

A is the cross-sectional area
(m)
The estimated number of points may then be rounded to a value that is convenient for subdividing
the cross-sectional area but should never be less than 24. Correct positioning of the measuring
instrument is facilitated by erecting a grid of fine wires in the airway to represent the subsections.
In the case of circular openings, the method of equal areas divides the circle into annuli, each of
the same area. Readings should be taken at points across two diameters and the corresponding
velocity profiles plotted. Should those profiles prove to be skewed then readings should be taken
across two additional diameters. The number of measuring points recommended on each
diameter is given in Table 6.1. Figure 6.4 illustrates an 8 point traverse on each of 4 diameters.
Diameter of duct (m)
No. of points
< 1.25
1.25 - 2.5
> 2.5
12
Table 6.1 Number of measuring points on each diameter of a circular opening.
6 - 10
Figure 6.4 Measuring positions for a 8 point traverse on 4 diameters of a circular duct.
The locations of the points are at the centre of area of the relevant annulus on each diameter and
may be calculated from
r = D
where
r
n
D
N
2n 1
4N
=
=
=
=
(6.5)
radius of point n from the centre

number of the point counted outwards from the centre
diameter of the duct (m)
number of points across the diameter
Table 6.2 gives locations of points for 6, 8, and 12 point traverses in terms of fractions of duct
diameter measured from one side.
Where a pitot-static tube traverse is to be conducted across a duct from the outside then a
clamping device should be attached to the outer surface of the duct to hold the pitot-static tube
firmly in place. The positions of measurement should be marked on the stem of the instrument
6 - 11
using Table 6.2 or 6.3. After each relocation of the measuring head, the pitot-static tube should
be yawed slightly from side to side until the orientation is found that gives the greatest reading of
total or velocity pressure. The head of the instrument is then aligned directly into the airstream.
No. of
measuring
points on
each
diameter
Fractions of one diameter measured from side of duct

centre
0.044 0.146 0.296 0.704 0.854 0.956
0.032 0.105 0.194 0.323 0.677 0.806 0.895 0.968
12
0.021 0.067 0.118 0.177 0.250 0.356 0.644 0.750 0.823 0.882 0.933 0.979
Table 6.2 Positions of measuring points in a circular duct using the method of equal areas.
Log-linear traverse
A more accurate method of positioning points of measurement along the diameters of a circular
duct has been derived from a consideration of the logarithmic law equations that describe the
velocity profile for turbulent flow. The effects of observational errors are minimized when the
points are located according to this method, known as the log-linear traverse. The corresponding
locations are given in Table 6.3.
No. of
measuring
points on
each
diameter
Fractions of one diameter measured from side of duct

centre
0.032
0.135 0.321 0.679 0.865 0.968
0.021 0.117
0.184 0.345 0.655 0.816 0.883 0.978
12
0.014 0.075 0.114 0.183
0.241 0.374 0.626 0.759 0.817 0.886
0.925 0.986
Table 6.3 Log-linear traverse positions of measuring points in a circular duct.

Velocity contours
One of the difficulties that besets ventilation engineers in measuring large scale airflows is that
conditions are often not conducive to good accuracy. "Textbook" advice is to choose measuring
stations well away from obstructions, bends or changes in cross-section. Unfortunately, this is not
always possible, especially when measuring airflows at the inlets or outlets of fans. It is not
uncommon to find that longitudinal swirl in a fan drift or re-entry in an evasee causes the air to
move in the wrong direction within one part of the cross-section. Similarly, in obstructed but high
velocity airways underground such as many longwall faces, complex airflow patterns may exist.
6 - 12
A useful rule of thumb is that the averaging of spot velocities from a pitot-static tube traverse is
acceptable if more than 75 per cent of the velocity pressures (pv) are greater than the maximum
pv divided by 10.
For difficult cases, the construction of velocity contours can provide both a visual depiction of the
flow pattern and also a means of quantifying airflow. A scale drawing of the measurement crosssection is made on graph paper. A grid of fine wires is constructed in the airway to define the
points of measurement. The number of points should be not less than that recommended in the
previous subsection. The greater the number of measurement points, the more accurate will be
the result. The velocities at the corresponding points of measurement are entered on the graph
paper and contour lines of equal velocity (isovels) are constructed. Figure 6.5 shows an example
of velocity contours obtained on a longwall face.
Figure 6.5 Example of velocity contours at an airflow measuring station on a longwall face
The area enclosed by each contour can be determined either by planimeter (if the scale drawing
is large enough for good planimeter accuracy) or by the rudimentary method of counting squares
on the graph paper. The construction of isovels and determination of the areas enclosed are
greatly facilitated by the use of a computer software contouring package. Fully automated
systems have been devised that scan the actual cross-section and produce quantified velocity
contour diagrams. However, the expense of such systems is seldom justified other than in
research and testing laboratories.
By difference, the area of the band between each contour is evaluated and may be multiplied by
the mean of the bounding velocities and the area scale factor to give the airflow for that band.
Provided that the outermost contour is close to the walls then the velocity at the walls may be
taken as zero. The sum of all band airflows gives the total flow for the airway.
6.2.10 Hot body anemometers
When any heated element is placed in a moving fluid, heat energy will be removed from it at a
rate that depends upon the rate of mass flow over the element.
In the hot wire anemometer a wire element is sited within a small open ended cylinder to give the
instrument a directional bias. The element forms one arm of a Wheatstone bridge circuit. In most
hot wire anemometers, the temperature of the element is maintained constant by varying the
6 - 13
electrical current passing through it as the air velocity changes. In other designs, the current is
kept constant and the temperature (and, hence, electrical resistance) of the element is monitored.
Modern hot wire anemometers are compensated for variations in ambient temperature and most
also indicate dry bulb temperature. For precise work, readings should be corrected for air density:
u = ui
where u =
ui =
c =
m =
c
m
(6.5)
true air velocity

indicated air velocity
air density at calibration (usually 1.2 kg/m3)
actual air density at time of measurement.
Hot wire anemometers are particularly useful for low velocities and are reliable down to about 0.1
m/s. They are convenient for fixed point traverses in slow moving airstreams. If a hot wire
anemometer is to be used in a gassy mine then a check should first be made on the permissibility
of the instrument for use in potentially explosive atmospheres.
The Kata thermometer, described in Section 17.4.3.2 as a means of measuring the cooling power
of an airstream, can also be used as a non-directional device to indicate low air velocities,
typically in the range 0.1 to 1 m/s. The main bulb of the Kata thermometer is heated until the
alcohol level is elevated above the higher of the two marks on the stem. When hung in an
airstream, the time taken for the alcohol level to fall between the two marks, coupled with the
Kata index for the instrument and the air temperature, may be used to determine the nondirectional air velocity. The Kata thermometer is seldom used for underground work (except in
South Africa) because of its fragility.
6.2.11. Tracer gases
The rate at which injected gases are diluted provides a means of measuring air volume flow
without the need for a cross-sectional area. The method is particularly useful for difficult situations
such as leakage flow through waste areas, main shafts and other regions of high velocity and
excessively turbulent flow, or total flow through composite networks of airways.
Hydrogen, nitrous oxide, carbon dioxide, ozone, radioactive krypton 85 and sulphur hexafluoride
have all been used with the latter particularly suitable for leakage or composite flows. The gas
chosen should be chemically inert with respect to the mineralization of the strata.
There are two techniques of using tracer gases for the measurement of airflow. For high velocity
airways, the tracer gas may be released at a monitored and steady rate Mg (kg/s). At a point
sufficiently far downstream for complete mixing to have occurred, samples of the air are taken to
establish the steady state concentration of the tracer gas. Then
C =
qg
Q
Mg
(6.6)
g Q
or
Q =
where qg
Q
C
g
=
=
=
=
Mg
C g
m3
s
(6.7)
volume flow of tracer gas (m3/s)

airflow (m3/s)
downstream concentration of tracer gas (fraction by volume)
density of tracer gas at ambient pressure and temperature (kg/m3)
It is assumed that the volume flow of tracer gas is negligible compared with the airflow.
6 - 14
In the case of sluggish or composite flows, a known mass of the tracer gas, M (kg), is released as
a pulse into the upstream airflow. At the downstream station, the concentration of tracer gas is
monitored and a concentration-time, (C v t), graph is plotted as shown on Figure 6.6. Now, the
concentration C is given as the ratio of the volume flow rate of gas, qg, and the airflow, Q.
C =
But
qg
qg
(fraction, by volume)
Q
=
Mg
m3
s
where Mg = mass flow rate of gas at the monitored downstream station (kg/s)
and
g = density of gas at the prevailing temperature and pressure kg/m3)
giving
C =
Mg
g Q
Figure 6.6 Concentration-time curve at a tracer gas monitoring station
6 - 15
Hence, the complete area under the curve of C against t,

I =
(This equals the total volume of tracer gas that passed the station)
C dt
1
g Q
M g dt
But the total mass of gas released, M, must also be equal to
M g dt (assuming that all of
the air passing the upstream station also passes through the downstream station).
M
giving I =
g Q
or
Q =
M
I g
m3
s
(6.8)
6.2.12 Measurement of cross-sectional area

As the vast majority of airflows are determined as the product of a mean velocity and a crosssectional area, the accuracy of the airflow depends equally upon the measured velocity and
cross-sectional area. There is little point in insisting upon meticulous procedures for the
measurement of mean velocity unless the same care is applied to finding the cross sectional
area.
By far the most common method of measuring airway area is by simple taping. This will give good
results where the opening is of regular geometric shape such as a rectangle or circle. Airflow
measuring stations should, wherever possible, be chosen where the airway profile is well defined.
The frames of removed ventilation control doors can provide excellent sites for airflow
measurement. Shapes such as arched profiles or trapeziums may be subdivided into simple
rectangles, triangles and segments of a circle, and appropriate taped measurements taken to
allow the area to be calculated.
Inevitably, there are many situations in which airflows must be determined in less well defined
cross-sections. Several techniques are available for determining the corresponding crosssectional area. For shapes that approximate to a rectangle, three or more heights and widths may
be taped to find mean values of each. Care should be taken in such circumstances to make
allowance for rounding at the corners. This often occurs due to spalled rock accumulating on the
floor at the sides of airways.
A more sophisticated technique is the offset method in which strings are erected that define a
regular shape within the airway. These strings are usually two vertical and two horizontal wires
encompassing a rectangle. Taping from the wires to the rock walls at frequent intervals around
the perimeter allows a plot of the airway profile to be constructed on graph paper.
The profilometer is a plane-table device. A vertical drawing board is attached to a tripod in the
middle of the airway cross-section. Taped measurements made from the centre of the board to
points around the rock walls may be scaled down mechanically or manually to reconstruct the
airway profile on the drawing board. An electronic version replaces the tape by an ultrasonic
distance measuring device although reflections of the beam can produce errors within the
confines and rough surfaces of an underground airway.
The photographic method entails painting a white line around the perimeter of the measuring
station. A linear scale such as a surveyor's levelling staff is fixed vertically within the defined
6 - 16
profile. A camera is located such that it is aligned along a longitudinal centre-line of the airway
and with its lens equidistant from all points on the painted line. These precautions reduce
perspective errors. The area within the white line may be determined by overlaying the resulting
photograph with transparent graph paper.
Such time consuming methods tend to be employed for permanent measuring stations rather
than for temporary survey stations. In all cases, the cross sectional area of conveyors, ducts or
other equipment should be determined and subtracted from the overall area of the airway.
6.3. PRESSURE SURVEYS
The primary purpose of conducting pressure surveys is to determine the frictional pressure drop,
p, that corresponds to the airflow, Q, measured in each branch of a survey route. There are
essentially two methods. The more accurate is the gauge and tube or trailing hose method, in
which the two end stations are connected by a length of pressure tubing and the frictional
pressure drop measured directly. The second method, of which there are several variations,
involves observing the absolute pressure on a barometer or altimeter at each station.
Although tradition within individual countries tends to favour one or other of the two methods, both
have preferred fields of application. In general, where foot travel is relatively easy between
measuring stations, the gauge and tube method can be employed. Where access is difficult as in
multi-level workings or in shafts then the barometer method becomes more practicable.
6.3.1. Gauge and tube surveys
Figure 6.7 illustrates the principles of gauge and tube surveying. A pressure gauge is connected
into a length of tubing whose other ends are attached to the total head tappings of pitot-static
tubes sited at the end stations. In practice, of course, the tubing and instrumentation are all within
the airway. Let us deal first with the essential theory of the method before discussing the practical
procedure of gauge and tube surveying.
Theory
From the steady-flow energy equation (3.25) for an airway between stations 1 and 2, and
containing no fan,
2
u1 u2
+ (Z1 Z2 )g =
2
2
1
V dP + F12
J
kg
(6.9)
where
and
u
Z
g
V
F12
=
=
=
=
=
air velocity (m/s)

height above mine datum (m)
gravitational acceleration (m/s2)
specific volume of air (= 1/) (m3/kg)
work done against friction (J/kg)
If we assume a linear variation in air density between stations 1 and 2 then we can adopt an
arithmetic mean value of density for the airway, a = 1/V. Furthermore, the frictional pressure
drop referred to that density is given by equation (2.46) as
p12 = a F12
Pa
(6.10)
6 - 17
Figure 6.7 Measuring the frictional pressure drop between two stations by the gauge and tube method.
Applying these conditions to equation (6.9) gives
p12 = a F12 = a
(u1 u 2 )
+ a (Z 1 Z 2 )g (P2 P1 )
2
Pa
(6.11)
(See section 3.4.1 for a fuller explanation of this equation.)

pv =
However, velocity pressure
a u 2
2
and static pressure p s = a Z g + P when referred to the mine datum for elevation.
Hence, equation (6.11) may be written as
p12 = ( pv 1 + p s1 ) ( pv 2 + p s 2 )
=
where pt
pt1
pt 2
Pa
Pa
(6.12)
total pressure (pv + ps) as sensed by the total head tapping of a pitot-static tube.
This shows that the frictional pressure drop, p12, referred to the mean density in the airway
between the pitot-static tubes is given simply as the pressure gauge reading shown as P in
Figure 6.7. If that measured frictional drop is to be referred to a standard value of air density, st,
in order to compare or compound it with frictional pressure drops measured in other airways then
the correction is given as
6 - 18
p12 (standardized) = p12
st
a
In the great majority of cases no further calculation is required. This explains why the gauge and
tube technique is termed a direct method of measuring the frictional pressure drop in an airway.
(But see Appendix A6 for situations where there is a significant difference in elevation between
the two end stations)
Practical Procedure
The procedure for conducting a gauge and tube survey commences by assembling the
equipment and calibrating the gauges. For convenience, a list of the required equipment is given
here, together with some explanatory comments:
2 pitot-static tubes, approximately 1.25m in length. Shorter instruments may be employed

for small airways or for use in ducts.
An assortment of diaphragm pressure gauges ranging from a full scale deflection of not
more than 100 Pa to the highest pressure developed by any fan in the system. The
gauges should be calibrated in the horizontal position against a primary manometer
immediately prior to an important survey. The use of diaphragm gauges rather than
inclined manometers has greatly improved the speed of gauge and tube surveying.
A continuous length of nylon or good quality plastic tubing between 100 and 200m in
length. The tubing should be mechanically strong so that it can withstand being run over
by rubber-tyred vehicles or being dragged under doors without permanent damage. An
internal diameter from 2 to 3 mm is convenient. Larger tubing may become difficult to
handle while the waiting time of transmission of a pressure wave may become unduly
long if the tube is too narrow. The tube should be pressure tested before and after the
survey.
Short lengths of flexible tubing to connect the pitot-static tubes and gauge to the main
tubing. Metal connectors and clamps should also be carried in case it becomes
necessary to repair damage to the main tubing.
2 or 3 cans of spray paint for station marking. Chalk or industrial type crayons can also
be used.
1 pocket barometer and 1 whirling hygrometer.
1 waterproof field book and pencils.
1 100m flexible measuring tape.
Tool kit containing screwdrivers, adjustable spanners (wrenches) and a sharp knife.
The route of the traverse and sites of main junction stations should have been established before
commencing the observations (Section 6.4). Two persons are required for a gauge and tube
survey. It is helpful to have the lowest range gauge fixed within a box with a transparent top, and
side holes for extended pressure tappings. Straps around the waist and neck of the observer hold
the gauge in a horizontal position. This facilitates travelling and making observations for
consecutive tube lengths along an airway.
6 - 19
The following procedure for making the observations is recommended:

1. At the starting station, the pressure tubing is unwound and laid out along the airway in the
direction of the second main station. At the forward position, the zero setting of the gauge
is checked and, if necessary, adjusted by connecting the high and low pressure tappings
by a short length of tubing. The gauge is then connected in-line between the main tube
and the total head tapping of the leading pitot-static tube as illustrated on Figure 6.7. At
the rear position, the second pitot-static tube is similarly connected to the pressure
tubing. The flexible tubing used for connections should be of an internal diameter that fits
snugly on to the main tube, the gauge tappings and the pitot-static tube without requiring
undue force.
2. To make the observation, both pitot-static tubes are held facing into the airflow, away
from the body of the observer and at a position between one half and two thirds the
height of the airway. The gauge is observed until the reading becomes constant. This
may take two to three minutes depending upon the length and diameter of the main tube.
Light tapping of the fingers may assist in overcoming any slight frictional resistance of the
diaphragm or linkages within the gauge. On completing the gauge reading, the leading
observer should indicate that fact to the trailing observer either by cap-lamp signals or by
a tug on the tube. The barometric pressure, wet and dry bulb temperatures are also read
and booked by the leading observer together with the distance between observers. In
most cases this is the known length of the main tube. For shorter distances, the
measuring tape or other means should be used to determine the actual length.
3. The final duty of the leading observer is to paint or chalk an indicator mark on the rail or
airway side. A second tug on the tube or a cap-lamp signal indicates that it is time to
move on. The leading observer walks forward, dragging the tube behind him. When the
trailing observer reaches the indicator mark he simply stops, grasping the main tube
firmly.
4. The procedure is repeated for each tube length until the next main (junction) station is
reached.
The leading observer is kept busy while the trailing observer has little to do other than holding a
pitot-static tube at each station or substation and walking forward. However, it is preferable that
the observers exchange positions only in alternate shifts rather than during any one day. An
experienced team can progress along a traverse route fairly quickly. Indeed, using modern
equipment, it is usually the measurement of airflows by the accompanying airflow team rather
than frictional pressure drops that dictates the overall speed of the survey (see section 6.4).
Each major junction of airways should be a main station within a gauge and tube traverse. At
each of those junctions, the pitot-static tube should be held at the centre of the junction. If high
turbulence causes excessive fluctuations on the gauge then the static tapping(s) on the pitotstatic tube(s) may be employed. In this case, an anemometer should be held at the position of the
pitot-static tube(s) to measure the local velocity. The corresponding velocity pressures should be
calculated and applied as a correction to the gauge reading in order to determine the frictional
drop in total pressure.
Care should be taken at all times to ensure that the pitot-static tubes do not become clogged by
dust or other debris. Similarly, in wet conditions, it is vital to take precautions against water
entering any tube. Pitot-static tubes or the open ends of pressure tubing should never be allowed
to fall on to the floor during a traverse.
During the course of a pressure traverse, check readings should be taken of the pressure
differences across doors between airways. It is convenient to carry a separate 10m length of
flexible tubing for this purpose. It takes only a few seconds to attach a gauge of the required
6 - 20
range. If in doubt concerning the range, a high pressure gauge should be used first to establish
an approximate pressure difference, then exchanged for a more appropriate instrument if
necessary. It is usually sufficient to measure the static pressure across a door. Hence, the two
ends of the tubing should be protected against the very local air velocities that sometimes occur
from leakage close to a door. A practical way of doing this is simply to insert the end of the tube
into one's pocket.
6.3.2. Barometer and altimeter surveys
If the absolute static pressures are measured on barometers at the two ends of a subsurface
airway then the difference between those two measured pressures will depend upon
the difference in elevation between the stations,

the air velocities, and
the frictional pressure drop between the two stations at the prevailing airflow.
As the elevations and velocities can be measured independently, it follows that the barometric
readings can be used to determine the frictional pressure drop.
The concept of barometers was introduced in section 2.2.4.1. The instruments used for mine
barometric surveys are temperature compensated microaneroid devices. The facia of the
instruments are normally calibrated in kilopascals or other units of pressure. However, it will be
recalled that pressure may be quoted in terms of the column of air (or any other fluid) above the
point of measurement.
P = gh
(6.22)
Pa
[see equation (2.8)]

If the air density, , and gravitational acceleration, g, are regarded as constant then the pressure
may be quoted in head (metres) of air, h. This relationship is utilized to inscribe the facia of some
aneroid barometers in terms of metres (or feet) of air column.
h =
(6.23)
The instrument then indicates an approximate elevation or altitude within the earth's atmosphere
relative to some datum and, accordingly, is then called an altimeter. For a more accurate
elevation, the reading should be corrected for the difference between the actual values of and
g, and the standard values to which the altimeter has been calibrated. Some sophisticated
altimeters have an inbuilt bias which compensates for changes in air density with respect to
height. This allows a linear scale for altitude. For density compensated altimeters, the conversion
takes the form:
barometric pressure = exp(a b indicated altitude )
where the constants a and b depend upon the units of pressure and altitude, and factory settings
of the altimeter.
In most mining countries, barometric pressure surveys are carried out using direct indicating
barometers. In the United States, altimeters are commonly employed. However, if the altimeter
readings are converted to pressure units then the two methods become identical.1
It would seem to have been the larger market (and, hence, readier availability) of altimeters together with
the concepts of a head of air and a head of water that led to the early use of altimeters in the United States.
6 - 21
Theory
Again, we commence with the steady flow energy equation for an airway between stations 1 and
2, and containing no fan. In the usual case of polytropic flow, the energy equation gives the work
done against friction as:
2
F12
u u2
= 1
2
ln P2
P1
+ (Z1 Z 2 )g R(T 2 T1 )
lnT 2
T1
J
kg
(6.24)

or, for isothermal flow where T1 = T2
2
F12 =
u1 u 2
2
+ (Z1 Z 2 )g R T1 ln P2
P1
J
kg
where
and
P
T
Z
u
R
=
=
=
=
=
barometric pressure (kPa)

absolute temperature (degrees Kelvin)
elevation of barometer location (m)
air velocity at the barometer (m/s)
mean gas constant (J/kg K) [from equation (14.14)]
As all parameters are measurable in these relationships, the work done against friction, F12 can
be determined. This, in turn, can be converted into a frictional pressure drop, p12, referred to any
given air density, a.
p12 = a F12
Pa [see, also, equation (6.10]
A complication arises if the barometric pressures at the two stations are not read simultaneously.
In this case, the surface atmospheric pressure may change during any time interval that occurs
between readings at successive stations. If the atmospheric pressure at a fixed control station is
observed to increase by Pc during the time elapsed while moving from station 1 to station 2,
then the initial value, P1, should be corrected to
P1 + Pc
(6.25)
The correction, Pc may, of course, be positive or negative. By assuming a series of polytropic

processes connecting the control barometer (subscript c) to the traverse barometer (subscript 1),
it can be shown that a more accurate value of the correction is given as
Pc
P1
Pc
(6.26)
6 - 22
Example
The following two lines are an excerpt from a barometer field book.
Station
No.
Time
Traverse
barometer
P kPa
Temperatures
td
tw
C
C
13:42
103.75
15.6
14:05
104.61
17.2
Elevation
Z
m
Velocity
u
m/s
Control
barometer
Pc kPa
13.0
2652
2.03
98.782
14.2
2573
1.52
98.800
Using the psychrometric equations given in section 14.6, the moisture contents of the air at stations 1 and 2
were calculated to be
X1 = 0.008035 kg/kg dry air
and
X2 = 0.008536 kg/kg dry air
Equation (14.14) then indicates the corresponding gas constants as

R =
287.04 + 461.5 X
(1 + X )
giving R1 = 288.431 J/kg C and R2 = 288.517 J/kg C with an arithmetic mean of 288.474 J/kg C.
During the time period between taking barometer readings P1 and P2 the control barometer registered an
increase in atmospheric pressure of
Pc = 98.800 - 98.782 = 0.018
kPa
Equation (6.26) gives the corrected reading at station 1 as

P1 + Pc
P1
Pc
103.75
98.782
103.75 + 0.018
103.75 + 0.019 = 103.769
kPa
The steady flow energy equation (equation (6.24) then gives

2
F12
where
and
T1
T2
u u2
= 1
2
=
=
ln P2
P1
+ (Z1 Z 2 )g R(T 2 T1 )
lnT 2
T1
273.15 + 15.6
273.15 + 17.2
=
=
J
kg
288.75 K
290.35 K
6 - 23
F12 =
(
(
)
)
ln 104.61
2.03 2 1.52 2
103.769
+ (2652 2573 )9.81 288.474(17.2 15.6)
290
.
35
2
ln
288.75
0.91
101.7
774.99
674.22
J
kg
J
kg
J
kg
and the frictional pressure drop referred to standard density becomes

p12 = F12 = 101.7 1.2
= 122
Pa
(See, also, sections 8.2.2. and 8.3.3. for further examples.)
Practical Procedure
A barometric survey can be conducted with one observer at each station although an additional
person at the traverse stations facilitates more rapid progress. The equipment required is as
follows:
2 microaneroid barometers (or altimeters) of equal precision

1 whirling or aspirated psychrometer
2 accurate watches
1 anemometer
1 2m measuring tape
waterproof field books and pencils
2 or 3 cans of spray paint to mark station numbers.
The microaneroids should be calibrated against a primary barometer prior to an important survey.
A calibration cabinet can be constructed with the internal pressure controlled by compressed air
feeds and outlet valves. In addition to pressure calibration, the instruments should be checked for
temperature compensation and creep characteristics. Modern instruments are stable over the
range of temperatures normally encountered in mines and adapt to a change in pressure within a
few minutes.
For an underground barometer traverse only the main junctions need be considered as
measurement stations. Intermediate substations, as required in the gauge and tube technique,
are normally unnecessary. However, for main ventilating shafts, the most accurate results are
obtained by taking readings at intervals down the shaft (Section 8.2.2).
There are essentially two methods of handling the natural variations in atmospheric pressure that
occur during the course of the survey. One technique is to maintain a barometer at a fixed control
station and to record or log the readings at intervals of about 5 minutes. The second method is
the leapfrog procedure in which both barometers are used to take simultaneous readings at
successive stations. After each set of readings, the trailing barometer is brought up to the forward
station where the two barometers are checked against each other and reset if necessary. The
trailing barometer is then moved on to assume the leading position at the next station.
6 - 24
The traverse procedure commences with the observers synchronizing their watches. If the control
station method is used, the control should be established in a location that is reasonably stable
with respect to temperature and not subject to pressure fluctuations from fans, ventilation
controls, hoists or other moving equipment. A location on surface near the top of a downcast
shaft and shaded from direct sunlight is usually satisfactory. A recording barometer may be
employed at the control station, but only if it provides a precision equivalent to that of the traverse
barometer.
At each traverse station the following readings are logged:
date, barometer identification and name of observer

number and location of station
time
barometer reading
wet and dry bulb temperatures
anemometer reading at the position of the traverse barometer.
The location of each station should be correlated with surveyors plans to determine the
corresponding elevation. The traverse barometer should be held at the same height above the
floor at each station in order that its elevation can be ascertained to within 0.5m.
The anemometer should be employed to measure the air velocity at the position of the barometer.
There is no need to conduct an anemometer traverse. As shown in the example given in the
previous subsection, the effect of air velocity is usually small compared with the other terms in the
steady-flow energy equation.
6.4 ORGANIZATION OF PRESSURE-VOLUME SURVEYS
The preceding two sections have discussed the techniques of measuring volume flows, Q, and
frictional pressure drops, p, separately. It will be recalled that the results of the two types of
survey will be combined to give the resistance, R = p/Q2, and airpower loss, pQ, of each branch.
As airflows and, hence, frictional pressure drops vary with time in an operating subsurface facility,
it follows that p and Q should, ideally, be measured simultaneously in any given airway. Typically,
there are two observers measuring airflows and another two involved in the pressure survey. The
two teams must liaise closely.
6.4.1. Initial planning
A pressure-volume survey should be well planned and managed. The practical work for a major
survey commences a week or two before the underground observations by assembling, checking
and calibrating the equipment. In particular, it is inadvisable to rely upon manufacturers' original
calibrations of vane anemometers or diaphragm pressure gauges. If the equipment required for
calibration is unavailable locally then the work may be carried out by a service organization or the
instruments returned to the manufacturers for customized calibration. The calibration is normally
produced as a table of corrections against indicated readings and taped to the side of the
instrument or carrying case. Interpolation from the table can be carried out at the time of
measurement so that the reading, correction and corrected observation can all be logged
immediately.
The mine plan should be studied carefully and the routes of the survey selected. A full mine
survey will include each ventilation connection to surface and the infrastructure of airways that
comprise the primary ventilation routes. Subsidiary survey routes may be appended to include
individual working districts or to extend a data bank that exists from previous surveys. The routes
should be chosen such that they can be formulated into closed traverse paths or loops within the
ventilation network of the mine. Branches that connect to the surface close through the pressure
6 - 25
sink of the surface atmosphere. A main loop in a large mine may take several days to survey.
However, each main loop should be divided into smaller subsidiary loops each of which can be
closed within a single day of surveying.
An initial reconnaissance of the mine should be carried out, travelling through all airways selected
for the primary traverses and establishing the locations of main stations. These are normally at
major junctions of the ventilation system. Where two or more airways are adjacent and in parallel
- and the gauge and tube method is employed for the pressure measurements - then it is
necessary to take those measurements in one of the airways only. However, to obtain the total
airflow in that composite branch of the network, it will be necessary to take flow measurements at
corresponding points in each of the parallel airways. Airflow measuring stations should be
selected and marked on the plan and, also, on the walls of the airway.
The subsequent employment of survey data for ventilation network analysis and forward planning
(Chapters 7 and 9) should be kept in mind during the management of ventilation surveys. The
identification number assigned to each network junction should give an indication of the location
of the junction within the mine. In multi-level workings, for example, the first integer of station
numbers may be used to indicate the level.
A pre-survey briefing meeting should be held with all observers present. Each observer should be
fully trained in survey procedures, use of the instruments and techniques of observation. The
traverse routes and system of station identification should be discussed, together with an outline
schedule covering the days or weeks required to complete the survey.
6.4.2. Survey management
During production shifts, the airflows and pressure drops in an underground mine are subject to
considerable variation due to movement of equipment, changes in resistance in the workings and
opening of ventilation doors. Hence, the best time for ventilation surveys is when the mine is
relatively quiescent with few people underground. During the duration of a survey, the observers
should be prepared to work at weekends and on night shifts.
Although the frictional pressure drop and corresponding airflow should, ideally, be measured
simultaneously in each leg of the traverse, this is often not practicable. Nevertheless, the teams
should stay fairly near to each other so that there is a minimum delay between the two sets of
measurements in a given branch. With experienced observers the teams maintain close liaison,
assisting one another and always being conscious of the activity of the other team. This avoids
the infuriating situation of the pressure team opening doors to take a check reading while the
airflow team is in the middle of an anemometer traverse. Friendships have been known to suffer
on such occasions.
Immediately following each shift the two teams should check all calculations carried out
underground, transcribe the results of that shift's work from the field books to clean log sheets
and, also, to a large scale copy of a mine map. Positions of measured airflows and pressure
drops should be reviewed by both teams to ensure compatibility of measurement locations and to
correlate identification of station numbers. Any difficulties encountered during the shift should be
discussed. The final half hour or so of each working day may be spent in reviewing the ground to
be covered in the following shift and the allocation of individual duties.
6.4.3. Quality assurance
It is most important that control is maintained over the quality of all aspects of an important
ventilation survey, from initial calibration of the instruments through to the production of final
results. Field books or booking sheets should be laid out clearly such that persons other than the
observers can follow the recording of observations and calculations carried out underground. All
calculations should be checked by someone other than the originator. Most of the calculations
6 - 26
involved in ventilation surveys are quite simple and may be carried out on a pocket calculator.
The exception is for barometric surveys where a verified program for a personal computer is very
helpful. Commercially available spreadsheet software can readily be adapted for this purpose.
Adherence to Kirchhoff's Laws should be checked both at the time of observations wherever
practicable and, also, during the data transposition at the end of each shift. These laws are
discussed fully in Chapter 7. Briefly, Kirchhoff I requires that the algebraic sum of airflows
entering any junction is zero. Kirchhoff II states that the algebraic sum of standardized pressure
drops around any closed loop must also be zero, having taken fans and natural ventilation
pressures into account. In a level or near level circuit, the closing error of a pressure loop may be
expressed as the actual closure divided by the sum of the absolute values of the measured
frictional pressure drops around the loop. This should not exceed 5 per cent. The check
measurements of pressure differentials across doors are invaluable in tracing or distributing
observational errors. In the case of loops involving significant changes in elevation such as shaft
circuits, the sum of standardized pressure drops will be a combination of observational errors and
natural ventilating effects. The latter may be determined independently from temperature and
pressure measurements as discussed in section 8.3.
It is vital that good records be kept of each phase of a survey. The survey team leader should
maintain a detailed journal of the activities and achievements of each working day. This should
include the clean log sheets of results transcribed from the field books at the end of each shift.
The conclusion of a major survey should see the establishment of a spreadsheet type of data
bank or the extension of an existing data bank, holding the frictional pressure drop and
corresponding airflow for every branch included in the survey. Other details such as the dates of
observations, names of observers, instrument identifications and dimensions of airways may be
included. The data bank may then be used to calculate airway resistances, airpower losses and
friction factors, and also provides a data base from which a computer model of the mine
ventilation network can be generated (Chapter 9). Additionally, the resistances, resistance per
unit length and airpower losses may be shown on a colour-coded map in order to highlight
sections of airways that are particularly expensive to ventilate.
6.5. AIR QUALITY SURVEYS
While pressure-volume surveys are concerned with the distribution of airflow around a ventilation
system, the subsurface environmental engineer must also maintain control of the quality of that
air, i.e. the concentrations of gaseous or particulate pollutants, and the temperature and humidity
of the air. Such measurements should be made at specified times and places to ensure
compliance with mandatory standards and with a regard for the safety and health of the
workforce.
Details of the techniques of measuring and quantifying levels of dust, gas concentrations and
climatic conditions are given in Chapter 23, 11 and 14 respectively. In addition to mandatory
measurements a set of such observations made in a systematic manner around a continuous
path is known as an air quality survey. This procedure should be employed for two reasons. First,
it provides a means of tracking and quantifying the variation in pollutant levels and, secondly, it
enables zones of emission of gases, dust, heat and humidity to be identified. Measurements of
gas concentration are often made as a normal part of a pressure-volume survey in a gassy mine.
Similarly the observations of barometric pressure, wet and dry bulb temperature made during a
pressure survey may be used to compute and plot the variations in psychrometric conditions
throughout the traverse paths.
6 - 27
BIBLIOGRAPHY
Air Movement and Control Assoc. Inc. (1976). AMCA Application Manual (Part 3). A guide to
the Measurement of Fan System Performance in the Field.
American National Standard (1985). Laboratory Methods of Testing Fans for Rating. AMCA,
30 West University Drive, Arlington Heights, IL 60004.
ASHRAE Handbook (1985). Fundamentals Volume. ASHRAE, 1791 Tullie Circle, N.E. Atlanta,
GA. 30329, USA.
British Standard 848 Part 1 (1963). Methods of Testing Fans for General Purposes. British
Standards Institution, 2 Park St., London W.1.
Hartman H.L., Mutmansky, J. and Wang, W.J., (1982). Mine Ventilation and Air Conditioning.
Wiley Interscience.
Hinsley, F.B., (1962). The assessment of energy and pressure losses due to air-flow in shafts,
airways and mine circuits The Mining Engineer, Vol 121. No. 23, pp. 761-777.
Hinsley, F.B., (1964). An enquiry into the principles and mutual interaction of natural and fan
ventilation. The Mining Engineer, Vol 124. No. 49, pp. 63-78.
McPherson, M.J., (1969). A new treatment of mine barometer surveys. The Mining Engineer,
Vol 129 No. 109, pp. 23-34.
McPherson, M. J and Robinson, G. (1980). Barometric survey of shafts at Boulby Mine,
Cleveland Potash Ltd., Trans. Inst of Min. and Met. Vol 89 - reproduced Jnl. Mine
Ventilation Society of S. Africa, Vol 9, No.3, Sept. 1980.
McPherson, M.J., (1985). The resistance to airflow on a longwall face. 2nd U.S. Mine ventilation
Symposium, Reno, NV. pp. 531-542.
Mine Ventilation Society of South Africa, (1982). Environmental Engineering in South African
Mines. Mine Ventilation Society of S. Africa, Kelvin House, 2 Hollard St., Johannesburg, S.
Africa.
National Coal Board, U.K. (1979). Ventilation in Coal Mines: A Handbook for Colliery
Ventilation Engineers.
6 - 28
Appendix A6.
Application of the gauge and tube method of measuring frictional pressure drops
in cases of significant differences in elevation.
To this time, the gauge and tube technique has seldom been used for vertical shafts or highly inclined
airways although the often difficult task of measuring shaft resistance would be greatly facilitated by
leaving a length of small bore pressure tubing permanently in the shaft2. Where the ends of the tube are
at significantly different elevations a complication does, however, arise. It is found that in these
circumstances, the reading depends upon the location of the gauge and increases as the elevation of
the gauge decreases within the airway. This phenomenon occurs because the air within the tubing is
stationary and, hence, not affected by friction. The pressure at all points within the tube differs from that
outside the tube at the same elevation. For example, if the gauge is located at the bottom of a downcast
shaft, as shown in Figure A6.1 the pressure in the tube will be higher than at a corresponding elevation
outside the tube. If the temperature and moisture content inside and outside the tube are the same at
corresponding points it follows that the mean density in the tube must be a little higher than that in the
airway.
If the gauge is to indicate directly the frictional pressure

drop referred to the mean density in the airway, a then it
must be located at the position of mean density. In shafts
or other airways of constant slope and resistance, this is
very close to the midpoint. However, it is usually difficult
to take measurements at this position and it is more
practicable to site the pressure gauge either at the top or
bottom of the shaft. A correction must then be applied to
the reading in order to arrive at the frictional pressure
drop referred to the mean air density.
We shall continue this analysis assuming the gauge to
be at the base of a downcast shaft (Figure A6.1). The
pressure in the tubing at the high pressure tapping is Pc
and that at the low pressure tapping is Pb. The gauge
reads P = Pc - Pb. However, the pressure Pc must equal
the total pressure at station 1 plus the pressure due to
the head of static air within the tube
Pc = P1 +
Pc
Pb
P
where P
u
Z
and t
1 u 12
+ Z g t
= barometric (static) pressure

= air velocity
= depth of shaft
= mean air density within the tube
And, at the low pressure tapping of the gauge

Pb = P2 +
Figure A6.1 Gauge at the bottom of a downcast

shaft. The gauge is depicted as a manometer to
show the direction of the pressure difference.
2 u22
2
Where such a tube is left in place permanently, it should be tested for leaks and internal condensate before
being used for a frictional pressure drop observation. A disconnected compressed air pipe may be used as a
temporary pressure tube.
6 - 29
Then
1u12 2 u 2 2
P = Pc Pb =
+ Z g t + (P1 P2 )
(6.13)
Now, from equation (6.11), the frictional pressure drop in the airway, referred to the airway mean
density, a is given as
p12 = a
u1 u 2
2
+ Z g a + (P1 P2 )
(6.14)
Hence, the "error" in the gauge reading becomes

2
( 1u1 2 u2 ) (u1 u2 ) a
= P p12 =
+ Z g ( t a )
(6.15)
We can substitute for
a =
t =
where
1 + 2
(Mean density in airway)
1 + c
(Mean density in tube)
c = density in tube at position c.

Pc
RT2
Now
c =
giving
c 2 =
or
c = 2 +
and
Pc P2
RT 2
P
P2
2 =
= P
P2
RT2
(ignoring the small effect of velocity pressure on the air

density at position b.)
P2
Then, after some algebraic simplification, equation (6.15) becomes
Z g P 2
2 P2
(u1 + u 2 )
( 2 1 )
4
(6.16)
giving
Z g 2
p12 = P = P 1
+
2 P2
(u1 + u 2 )
( 2 1 )
4
(6.17)
This is the full form of the equation that allows the reading on the gauge at the base of the shaft
or slope to be corrected to mean density for the airway.
6 - 30
Z g 2
The term
2 P2
arises from the difference in mean air density between the airway and the tubing (equation (6.15),
2
(u1 + u 2 )
( 2 1 )
4
is the result of converting the velocity pressures at 1 and 2 to the mean density a. To be
precise, this latter term should be applied even when the airway is level. However, it is normally
insignificant and may be ignored for practical purposes, giving the frictional pressure drop
referred to the mean air density in the shaft as
while
Z g 2
p12 = P 1
(6.18)
2 P2
Equation (6.18) illustrates that the uncorrected reading, P, on the shaft bottom gauge
overestimates the actual frictional pressure drop.
Similar reasoning leads to the following equations for other configurations:

Gauge at the top of a downcast shaft
Z g 1
p12 = P (top ) 1 +
+
2 P1
Gauge at the base of an upcast shaft
Z g 3
p 34 = P (top ) 1
+
2 P3
(u1 + u 2 )
( 2 1 )
4
(u 3
(6.19)
+ u4 )
(3 4 )
4
(6.20)
(Subscripts 3 and 4 refer to the bottom and top respectively of an upcast shaft.)
Gauge at the top of an upcast shaft
Z g 4
p 34 = P (top ) 1 +
+
2 P4
(u 3
+ u4 )
(3 4 )
4
(6.21)
Again, the kinetic energy term involving u values can usually be neglected.
Hinsley, in 1962 showed that for both downcast and upcast shafts the frictional pressure drops
are given to a good approximation by the arithmetic average of the gauge readings at the top and
bottom of the shaft.
6 - 31
Calculated Example
The following data3 were established for a downcast shaft fitted with a tube throughout its length.
A pressure gauge can be connected to either the top or the bottom of the tube. Ignoring the small
effects of air velocity, determine the frictional pressure drop, p12, referred to the mean air density
in the shaft.
Depth of shaft = 1219.2 m
Barometric pressure, kPa
Top of shaft
Bottom of shaft
101.591
117.015
1.275
1.408
Air density, kg/m3

(from psychrometric measurements
ref. Chapter 14, Section 6)
Pressure gauge when

connected to top of tube, Pa
566
Pressure gauge when

connected to base of tube, Pa
655
Solution
(a) Using the pressure gauge reading at the top of the shaft, equation (6.19) applies. Ignoring the
kinetic energy terms:
Z g 1
p12 = P (top ) 1 +
2 P1
1219.2 9.8066 1.275

= 566 1 +
= 608
2 101591
Pa
(b) Using the pressure gauge reading at the bottom of the shaft equation (6.18) applies.
Z g 2
p12 = P 1
2 P2
1219.2 9.8066 1.408

= 655 1
= 608
2 117 015
Pa
In both cases, p12 is referred to the mean air density in the shaft and, hence, give the same result.
(In practice, it is to be expected that there will not be exact agreement between the results of
shaft top and shaft bottom gauge locations because of observational errors.)
The given data are taken from Hinsley, F.B. (1962) on the basis of a frictional adiabatic process in a dry shaft. [The
Assessment of Energy and Pressure Losses due to Airflow in Shafts, Airways and Mine Circuits (Section IV). The
Mining Engineer, Vol.121, No. 23, August 1962.]
6 - 32
The arithmetic mean density in the shaft is
mean =
1.408 + 1.275
2
= 1.341
kg
m3
If corrected to standard air density (st = 1.2 kg/m3) for comparison with other airways, the
standardized frictional pressure drop becomes
p st
= p12
st
mean
= 608
1 .2
= 544 Pa
1.341
To check the approximate method of establishing the frictional pressure drop from the mean of
the top and bottom gauge readings:The arithmetic mean of the pressure gauge readings at the top and bottom of the shaft is
566 + 655
= 610.5
2
Pa
showing that, in this case, the approximation is in excellent agreement with the calculated 608
Pa.
6 - 33
Chapter 7 Ventilation Network Analysis
Chapter 7. Ventilation Network Analysis

7.1. INTRODUCTION ...........................................................................................1
7.2 FUNDAMENTALS OF VENTILATION NETWORK ANALYSIS.......................3
7.2.1 Kirchhoff's Laws .................................................................................................................. 3
7.2.2 Compressible or incompressible flow in ventilation network analysis ................................ 4
7.2.3 Deviations from the Square Law......................................................................................... 6
7.3. METHODS OF SOLVING VENTILATION NETWORKS ................................6

7.3.1 Analytical Methods.............................................................................................................. 6
7.3.1.1 Equivalent Resistances ................................................................................................ 6
7.3.1.2 Direct Application of Kirchhoff's Laws ........................................................................ 10
7.3.2 Numerical Methods ........................................................................................................... 12
The Hardy Cross Technique............................................................................................... 13
(a) Mesh selection: ............................................................................................................. 18
(b) Estimate initial airflows: ................................................................................................. 18
(c) Calculate mesh correction factors: ................................................................................ 18
(d) Apply the mesh correction factor:.................................................................................. 19
(e) Completion of mesh corrections.................................................................................... 20
7.4. VENTILATION NETWORK SIMULATION PACKAGES...............................24

7.4.1 Concept of a mathematical model .................................................................................... 24
7.4.2 Structure of a ventilation simulation package ................................................................... 25
7.4.3 Operating system for a VNET package ............................................................................ 26
7.4.4 Incorporation of air quality into a network simulation package ......................................... 27
7.4.5 Obtaining a ventilation simulation package ...................................................................... 28
Hardware requirements and size of network ...................................................................... 28
Cost of software.................................................................................................................. 28
Speed.................................................................................................................................. 28
Scope and ease of use ....................................................................................................... 28
User's manual and back-up assistance .............................................................................. 28
7.4.6 Example of a computed network ...................................................................................... 29
REFERENCES ...................................................................................................39
7.1. INTRODUCTION
A vital component in the design of a new underground mine or other subsurface facility is the
quantified planning of the distribution of airflows, together with the locations and duties of fans
and other ventilation controls required to achieve acceptable environmental conditions throughout
the system. Similarly, throughout the life of an underground operation, it is necessary to plan
ahead in order that new fans, shafts or other airways are available in a timely manner for the
efficient ventilation of extensions to the workings. As any operating mine is a dynamic system
with new workings continually being developed and older ones coming to the end of their
productive life, ventilation planning should be a continuous and routine process.
The preceding chapters have discussed the behaviour of air or other fluid within an individual
airway, duct or pipe. Ventilation network analysis is concerned with the interactive behaviour of
airflows within the connected branches of a complete and integrated network. The questions
addressed by ventilation network analysis may be formulated quite simply. If we know the
resistances of the branches of a ventilation network and the manner in which those branches are
interconnected then how can we predict, quantitatively, the distribution of airflow for given
locations and duties of fans?
7- 1
Alternatively, if we know the airflows that we need in specific branches of the network then how
can we determine an efficient combination of fans and structure of the network that will provide
those required airflows? Ventilation network analysis is a generic term for a family of techniques
that enable us to address such questions.
In a given network there are a large number of combinations of airway resistances, fans, and
regulators that will give any desired distribution of flow. Practical considerations limit the number
of acceptable alternatives. However, the techniques of network analysis that are useful for
modern industrial application must remain easy to use, and sufficiently rapid and flexible to allow
multiple alternative solutions to be investigated.
Before the mid 1950's there were no practicable means of conducting detailed and quantitative
ventilation network analysis for complete mine systems. Ventilation planning was carried out
either using hydraulic gradient diagrams formulated from assumed airflows or, simply, based on
the experience and intuition of the ventilation engineer. Attempts to produce physical models of
complete mine ventilation systems using air or water as the medium met with very limited
success because of difficulties from scale effects. Following earlier research in Holland (Maas,
1950), the first viable electrical analogue computers to simulate ventilation networks were
produced in the United Kingdom (Scott et al, 1953) followed rapidly in the United States (McElroy,
1954). These analogues employed electrical current passing through rheostats to simulate
airflows. Successive adjustment of the resistances of the rheostats enabled the linear Ohm's law
for electrical conductors to emulate the square law of ventilation networks.
Linear resistance analogues became the main automated means of analyzing mine ventilation
networks in the nineteen fifties and early 'sixties. Rapid advances in the electronics industries
during that time resulted in the development of direct reading fluid network analogues that
replaced the linear resistors with electronic components. These followed a logarithmic
relationship between applied voltage and current, producing analogues that were easier to use
and much faster in operation (Williams, 1964). However, by the time they became available,
ventilation simulation programs for mainframe digital computers had begun to appear
(McPherson 1964, Hartman and Wang 1967). These proved to be much more versatile, rapid
and accurate, and their employment soon dominated ventilation planning procedures in major
mining countries. Coupled with continued improvements in ventilation survey techniques to
provide the data, ventilation network analysis programs resulted in hitherto unprecedented levels
of flexibility, precision and economics in the planning, design and implementation of mine
ventilation systems.
Throughout the 1970's network programs were developed for large centralized mainframe
computers. Their initial use by industry tended to be inhibited by the often pedantic procedures of
data preparation together with the costs and delays of batch processing. In the 1980's, the
enhanced power and reduced cost of microcomputers led to the evolution of self-contained
software packages that allowed very easy interaction between the user and the computer. These
incorporated the use of graphics. Ventilation engineers could, for the first time, conduct multiple
planning exercises on large networks entirely within the confines of their own offices. The
complete processing of data from survey observations through to the production of plotted
ventilation plans became automated. Personal computers, printers and plotters proliferated in
mine planning offices. Together with the ready availability of software, these led to a revolution in
the methodologies, speed and accuracy of subsurface ventilation planning.
In this chapter, we shall introduce the basic laws that govern the distribution of airflow within a
network of interconnected branches. The analytical and numerical methods of predicting airflows
will be examined before proceeding to a discussion of network simulation packages.
7- 2
7.2 FUNDAMENTALS OF VENTILATION NETWORK ANALYSIS

Any integrated ventilation system can be represented as a schematic diagram in which each line
(branch) denotes either a single airway or a group of openings that are connected such that they
behave effectively as a single airway (Section 7.3.1.1). Only those airways that contribute to the
flow of air through the system appear on the network schematic. Hence, sealed off areas of
insignificant leakage, stagnant dead ends and headings that produce no induction effects
(Section 4.4.3) on the main airflow need not be represented in the network. On the other hand,
the tops of shafts or other openings to surface are connected to each other through the pressure
sink of the surface atmosphere. The points at which branches connect are known simply as
junctions or nodes.
7.2.1 Kirchhoff's Laws
Gustav R. Kirchhoff (1824-87) was a German physicist who first recognized the fundamental
relationships that govern the behaviour of electrical current in a network of conductors. The same
basic relationships, now known as Kirchhoff's Laws, are also applicable to fluid networks
including closed ventilation systems at steady state.
Kirchhoff's first law states that the mass flow entering a junction equals the mass flow leaving that
junction or, mathematically,
j M = 0
(7.1)
where M are the mass flows, positive and negative, entering junction j.
However, it will be recalled that
kg
s
M = Q
(7.2)
where Q = volume flow (m3/s) and

Hence Q = 0
(7.3)
In subsurface ventilation systems, the variation in air density around any single junction is
negligible, giving
j Q = 0
(7.4)
This provides a means of checking the accuracy of airflow measurements taken around a junction
(Section 6.4.3).
The simplest statement of Kirchhoff's second law applied to ventilation networks is that the
algebraic sum of all pressure drops around a closed path, or mesh, in the network must be zero,
having taken into account the effects of fans and ventilating pressures. This can be quantified by
writing down the steady flow energy equation (3.25), initially for a single airway.
u 2
2
+ Z g + W
V dP
+ F
J
kg
(7.5)
7- 3
where u = air velocity (m/s)

Z = height above datum (m)
W = work input from fan (J/kg)
V = specific volume (m3/kg)
P = barometric pressure (Pa) and
F = work done against friction (J/kg)
If we consider a number of such branches forming a closed loop or mesh within the network then
the algebraic sum of all Z must be zero and the sum of the changes in kinetic energy, u2/2, is
negligible. Summing each of the remaining terms around the mesh, m, gives
V dP + [F W ] = 0
m
J
kg
(7.6)
Now the summation of V dP terms is the natural ventilating energy, NVE, that originates from
thermal additions to the air (Section 8.3.1). Hence, we can write
[F W ] NVE = 0
J
kg
(7.7)
This may now be converted to pressure units by multiplying throughout by a single value of air
density .
[F W ] NVE = 0
However,
F
=
W
=
NVE =
(7.8)
Pa
p (frictional pressure drop, equation (2.46))

pf (rise in total pressure across a fan) and
NVP (natural ventilating pressure, equation (8.32)),
each of these three terms being referred to the same (standard) density. The equation then
becomes recognizable as Kirchhoff's second law:
(p - p f ) NVP = 0
Pa
(7.9)
This is the relationship that is employed as a quality assurance check on a pressure survey
(Section 6.4.3), or as a means of determining a value for the natural ventilating pressure
(equation 8.51)).
7.2.2 Compressible or incompressible flow in ventilation network analysis
Kirchhoff's laws can be applied to fluid networks that conduct either compressible or
incompressible fluids. In the former case, the analysis is carried out on the basis of mass flow.
Equations (7.1) or (7.3) are employed for the application of Kirchhoff's first law and equation (7.7)
for Kirchhoff's second law.
There is another set of equations that must be obeyed within ventilation networks. Those have
already been introduced as the square law for each individual branch. When the flow is deemed
to be compressible, then the rational form of the square law should be utilized
p = Rt Q 2
Pa
7- 4
where Rt =
rational turbulent resistance (m-4)
But, as p = F (equation (2.46)),

p = F = R t Q 2
giving
J
kg
F = Rt Q 2
(7.10)
Kirchhoff's second law for compressible flow (equation (7.7)) becomes
R t Q 2 W NVE = 0
m
J
kg
(7.11)
As we progress from branch to branch around a closed mesh in a network then it is the algebraic
values of the pressure drops, p, or losses of mechanical energy, F, that must be summed. These
are both positive in the direction of flow and negative if the flow is moving against the direction of
traverse around the mesh. Hence, equation (7.11) may, more appropriately, be written as
R t Q Q W NVE = 0
m
J
kg
(7.12)
where Q = airflow with due account taken of sign ( m3/s)

and
IQI = absolute value of airflow and is always positive (+m3/s).
This device ensures that the frictional pressure drop or loss of mechanical energy always have
the same sign as airflow.
In the case of incompressible flow, the application of Kirchhoff's laws becomes more
straightforward. Equations (7.4) and (7.9) give
Q = 0
(Kirchhoff I)
RQ Q p f ) NVP = 0
(Kirchhoff II)
and
(7.13)
where R = Atkinson resistance (Ns2/m8)

It will be recalled that the three terms of this latter equation should each be referred to the same
(standard) value of air density, normally 1.2 kg/m3.
Computer programs have been developed for compressible flow networks. These require input
data (pressures, temperatures, elevations and air quality parameters) from which variations in air
density and natural ventilation effects may be calculated. On the other hand, where
compressibility and natural ventilating effects need to be taken into account, there are means by
which these can be simulated to an acceptable accuracy by an incompressible flow network
program. For these reasons, the great majority of subsurface ventilation planning employ the
simpler and faster incompressible flow programs. The more sophisticated and demanding
compressible flow programs are required for compressed air (or gas) networks or for specialized
applications in subsurface ventilation systems.
The remainder of this chapter will concentrate on incompressible flow network analysis.
7- 5
7.2.3 Deviations from the Square Law

The square law was derived in Section 2.3.6 from the basic Chezy Darcy relationship and further
developed in Section 5.2. Both the rational form of the square law, p = RtQ2, and the traditional
form, p = RQ2, apply for the condition of fully developed turbulence. Furthermore, both the
rational resistance, Rt, and the Atkinson resistance, R, are functions of the coefficient of friction, f,
for the duct or airway (equations (2.51), (5.4), and (5.1)). Hence, if the flow falls into the
transitional or laminar regimes of the Moody chart (Figure 2.7) then the value of f becomes a
function of Reynolds Number. The corresponding values of resistance, Rt, or R, for that airway
then vary with the airflow. If the form of the square law is to be retained then the values of
resistance must be computed for the relevant values of Reynolds Number.
However, many experimental researchers have found that plotting In(p) against In(Q) may give a
slope that deviates slightly from Chezy Darcys theoretical prediction of 2, even for fully
developed turbulence. The relationship between frictional pressure drop, p, and volume flowrate,
Q, may be better expressed as
p = RQn
Pa
(7.14)
where values of the index, n, have been reported in the range of 1.8 to 2.05 for a variety of pipes,
ducts and fluids. Similar tests in mine airways have shown that n lies very close to 2 for routes
along the main ventilation system but may be lower for leakage flows through stoppings or old
workings. This effect is caused primarily by flows that enter the transitional or, even, laminar
regimes. In the latter case, n takes the value of 1.0 and the pressure drop flow relationship
becomes
p = RL Q
where RL = laminar resistance

In the interests of generality, we shall employ equation (7.14) in the derivations that follow.
7.3. METHODS OF SOLVING VENTILATION NETWORKS

There are essentially two means of approach to the analysis of fluid networks. The analytical
methods involve formulating the governing laws into sets of equations that can be solved
analytically to give exact solutions. The numerical methods that have come to the fore with the
availability of electronic digital computers solve the equations through iterative procedures of
successive approximation until a solution is found to within a specified accuracy.
In both cases, the primary processes of solution may be based on the distribution of pressures
throughout the network or, alternatively, on the distribution of flows. The former may be preferred
for networks that involve many outlet points such as a water distribution network. On the other
hand, for networks that form closed systems such as subsurface ventilation layouts it is more
convenient to base the analysis on flows.
7.3.1 Analytical Methods
7.3.1.1 Equivalent Resistances
This is the most elementary of the methods of analyzing ventilation networks. If two or more
airways are connected either in series or in parallel then each of those sets of resistances may be
combined into a single equivalent resistance. Although of fairly limited value in the analysis of
7- 6
complete networks, the method of equivalent resistances allows considerable simplification of the
schematic representation of actual subsurface ventilation systems.
In order to determine an expression for a series circuit, consider Figure 7.1 (a). The frictional
pressure drops are given by equation (7.14) as
p1 = R1 Q n ,
p2 = R 2 Q n ,
p3 = R 3 Q n
Then for the combined series circuit,

p = ( p1 + p 2 + p 3 ) = (R1 + R 2 + R 3 ) Q n
or
p = R ser Q n where R ser = R1 + R 2 + R 3
Rser is the equivalent resistance of the series circuit.

In general, for a series circuit
R ser
= R
(7.15)
R1
R2
R3
p1
p2
p3
(a) Resistances connected in series
R1
Q1
R2
Q2
R3
Q3
p
(b) Resistances connected in parallel.
Figure 7.1 Series and parallel circuits.
In the case of a parallel circuit, Figure 7.1 (b) shows that each branch suffers from the same
frictional pressure drop, p, between the common start and end" junctions but passes differing
airflows. Then
7- 7
p = R1Q1 = R2Q2
giving
Q1 = ( p / R1)
= R3Q3
1
n
Q2 = ( p / R2 )
Q3 = ( p / R3 )
1
n
1
n
The three airflows combine to give

Q = Q1 + Q2 + Q3 = p
1
n
1
1
1
+
1 +
1
1
R1 n
R2 n
R3 n
We may write this as

p
Q =
1
n
Rpar
1
n
where Rpar is the equivalent resistance of the parallel circuit. It follows that
1
1
1
= 1 +
+
1
1
R1 n
R2 n
R3 n
Rpar
1
n
or in general,
1
R par
1
n
(7.16)
1
n
In the usual case of n = 2 for subsurface ventilation systems,

1
R par
(7.17)
Example
Figure 7.2 illustrates nine airways that form part of a ventilation network. Find the equivalent
resistance of the system.
Solution
Airways 1, 2, and 3 are connected in series and have an equivalent resistance of
Ra = R1 + R2 + R3
= 0.6 + 0.1 + 0.1 = 0.8
(equation (7.15))
Ns2
m8
This equivalent resistance is connected across the same two junctions, A and B, as airway 4 and,
hence, is in parallel with that airway The equivalent resistance of the combination, Rb, is given by
1
Rb
1
0. 8
1
0 .3
(equation (7.16)
7- 8
Figure 7.2
Example of a network segment that can be

resolved into a single equivalent resistance
from which
Ns2
Rb = 0.1154
m8
Notice that the equivalent resistance of the parallel circuit is less than that of either component
(ref. Buddle, Section 1.2). Equivalent resistance Rb is now connected in series with airways 5 and
6 giving a new equivalent resistance of all paths inby junctions C and D as
Rc = 0.1154 + 0.15 + 0.17 = 0.4354
Ns2
m8
Combining this in parallel with airway 7 gives

1
Rd
which yields
1
0.4354
Rd = 0.298
1
10
Ns2
m8
The equivalent resistance of all nine airways, Requ is then given by adding the resistances of
airways 8 and 9 in series
Requ = 0.298 + 0.12 + 0.13 = 0.548
Ns2
m8
7- 9
If it is simply the total flow into this part of a larger network that is required from a network
analysis exercise then the nine individual airways may be replaced by the single equivalent
resistance. This will simplify the full network schematic and reduce the amount of computation to
be undertaken during the analysis. On the other hand, if it is necessary to determine the airflow in
each individual branch then the nine airways should remain as separate resistances for the
network analysis.
The concept of equivalent resistances is particularly useful to combine two or more airways that
run adjacent to each other. Similarly, although a line of stoppings between two adjacent airways
are not truly connected in parallel, they may be grouped into sets of five to ten, and each set
represented as an equivalent parallel resistance, provided that the resistance of the stoppings is
large compared with that of the airways. Through such means, the several thousand actual
branches that may exist in a mine can be reduced to a few hundred, simplifying the schematic,
reducing the amount of data that must be handled, and minimizing the time and cost of running
network simulation packages.
7.3.1.2 Direct Application of Kirchhoff's Laws
Kirchhoff's laws allow us to write down equation (7.4) for each independent junction in the
network and equation (7.9) for each independent mesh. Solving these two sets of equations will
give the branch airflows, Q, that satisfy both laws simultaneously.
If there are b branches in the network then there are b airflows to be determined and, hence, we
need b independent equations. Now, if the network contains j junctions, then we may write down
Kirchhoff I (equation (7.4)) for each of them in turn. However, as each branch is assumed to be
continuous with no intervening junctions, a branch airflow denoted as Qi entering a junction will
automatically imply that Qi leaves at the junction at the other end of that same branch. It follows
that when we reach the last junction, all airflows will already have been symbolized. The number
of independent equations arising from Kirchhoff I is (j -1).
This leaves b - (j - 1) or (b - j + 1) further equations to be established from Kirchhoff II (equation
(7.9)). We need to choose (b - j + 1) independent closed meshes around which to sum the
frictional pressure drops.
The direct application of Kirchhoff's laws to full mine circuits may result in several hundred
equations to be solved simultaneously. This requires computer assistance. Manual solutions are
limited to very small networks or sections of networks. Nevertheless, a manual example is useful
both to illustrate the technique and assists in understanding the numerical procedures described
in the following section.
Example
Figure 7.3 shows a simplified ventilation network served by a downcast and an upcast shaft, each
passing 100 m3/s. The resistance of each subsurface branch is shown. A fan boosts the airflow in
the central branch to 40 m3/s. Determine the distribution of airflow and the total pressure, pb
developed by the booster fan.
Solution
Inspection of the network shows that it cannot be resolved into a series/parallel configuration and,
hence, the method of equivalent resistances is not applicable. We shall, therefore, attempt to find
a solution by direct application of Kirchhoff's laws.
The given airflows are also shown on the figure with the flow from A to B denoted as Q1. By
applying Kirchhoff I (equation (7.4)) to each junction, the airflows in other branches may be
expressed in terms of Q1 and have been added to the figure.
7- 10
Figure 7.3. Example network showing two meshes and Kirchhoff I applied at each junction.
The shafts can be eliminated from the problem as their airflows are already known. In the
subsurface structure we have 5 branches and 4 junctions. Accordingly, we shall require
bj+1
54+1
2 independent meshes
In order to apply Kirchhoff II (equation (7.9)) to the first mesh, consider the pressure drops around
that mesh, branch by branch, remembering that frictional pressure drops are positive and fan
pressures negative in the direction of flow. Figure 7.3 shows that we may choose meshes along
the closed paths ABCA and BDCB.
AB
BC
CA
Frictional pressure drop (RQ2)
Fan
0.2 Q12
1.0 x 402
- 0.1 (100 Q1)2
----------- pb
-----------
Summing up these terms according to Kirchhoff II gives

2
0.2 Q1 + 1600 pb 0.1 (100 Q1 )2
= 0
Pa
(7.18)
Collecting like terms gives
7- 11
pb = 0.1 Q1 + 20 Q1 + 600
(7.19)
Pa
Inspection of Figure 7.3 and commencing from junction B, shows that the summation of pressure
drops applied to the second mesh may be written as
0.5 (Q1 40)2 0.25 (140 Q1)2 1 40 2 + pb = 0
Pa
(7.20)
Notice that in this case, the direction of traverse is such that the pressure falls on passing through
the fan, i.e. a positive pressure drop. Substituting for pb from equation (7.19), expanding the
bracketed terms and simplifying leads to
2
0.35 Q1
+ 50 Q1 5100 = 0
This quadratic equation may be solved to give

Q1 =
50
50 2 + ( 4 0.35 5100 )
2 0.35
= 68.83 or 211.69
m3
s
The only practical solution is Q1 = 68.83 m3/s. The flows in all branches are then given from
inspection of Figure 7.3 as
AB
BD
BC
AC
CD
68.83 m3/s
28.83
40.00
31.17
71.17
The required booster fan pressure, pb, is given from equation (7.19) as
pb = 0.1 (68.83)2 + 20 (68.83) + 600 = 2450
Pa
7.3.2 Numerical Methods

Although the analytical methods of ventilation network analysis were given considerable attention
before the development of analogue computers in the 1950's, the multiple simultaneous
equations that they produced could not readily be solved by manual means. As shown in the
previous example, two mesh problems involve quadratic equations. A little further thought
indicates that three mesh problems would involve polynomial equations in Q of order 4. Four
meshes would produce powers of 8, and so on (i.e. powers of 2m-1, where m = number of
meshes).
Atkinson recognized this problem in his paper of 1854 and suggested a method of successive
approximation in which the airflows were initially estimated, then adjusted towards their true value
through a series of corrections. Although no longer employed, Atkinson's approach anticipated
current numerical methods.
7- 12
Since the mid 1960's, considerable research has been carried out to find more efficient numerical
methods of analyzing ventilation networks. Y.J. Wang in the United States developed a number of
algorithms based on matrix algebra and the techniques of operational research.
The method that is most widely used in computer programs for ventilation network analysis was
originally devised for water distribution systems by Professor Hardy Cross at the University of
Illinois in 1936. This was modified and further developed for mine ventilation systems by D.R.
Scott and F.B. Hinsley at the University of Nottingham in 1951. However, it was not until digital
computers became more widely available for engineering work in the 1960's that numerical
methods became truly practicable.
The Hardy Cross Technique.
Figure 7.4 shows the system resistance curve for one single representative branch in a ventilation
network. If the airflow, Q, is reversed then the frictional pressure drop, p, also becomes negative.
10000
n
p=RQ
p = RQn
p = RaQa
pa=RaQan
Qa
-100
-Q
Qa
Q
Q
100
p = RQn
p = R Qn
-p
-10000
Figure 7.4 In the system resistance curve for an airway, p and Q always have the same sign. The p,Q
slope remains non-negative.
7- 13
Recalling that a primary purpose of ventilation network analysis is to establish the distribution of
flow, the true airflow in our representative branch, Q, will initially be unknown. However, let us
assume an airflow, Qa, that is less than the true value by an amount Q.
Q = Qa + Q
The problem now turns to one of finding the value of Q. We may write the square law as
p = R (Qa + Q )2
(7.21)
or, more generally,

p = R (Qa + Q )n
(section 7.2.3)
(7.22)
Confining ourselves, for the moment, to the square law, equation (7.21) expands to
p = RQa
+ 2 RQa Q + R ( Q )2
(7.23)
Furthermore, the frictional pressure drop corresponding to the assumed airflow is

pa
= R Qa
(7.24)
where the error in p is
p = p pa
Substituting from equations (7.22) and (7.23) gives
p = 2 RQa Q + R ( Q )2
(7.25)
If we can assume that (Q)2 is small compared to 2 Q Q then we may write the approximation
as
p = 2 RQa Q
(7.26)
Note that differentiating the square law,

p = RQ 2
gives the limiting case when Qa Q. Then
p
Q
dp
dQ
= 2 RQ
(7.27)
and corresponds with the difference equation (7.26)

Then
Q =
p
2 RQa
(7.28)
7- 14
The technique of dividing a function by its first derivative in order to estimate an incremental
correction factor is a standard numerical method for finding roots of functions. It is usually known
as the Newton or Newton-Raphson method.
If the general logarithmic law is employed, equation (7.22) gives
Q
n
p = RQa 1 +
Qa
Expanding by the Binomial Theorem and ignoring terms of order 2 and higher gives
Q
n
p = RQa 1 + n
Qa
then
p = p pa
= RQa n
Q
Qa
( as pa = R Qa
giving
Q =
p
n RQa
(7.29)
n 1
If we refer back to Figure 7.4, we are reminded that p is the error in frictional pressure drop that
was incurred by choosing the assumed airflow, Qa.
p = RQ n RQa n
(7.30)
Hence, we can write equation (7.29) as
Q =
RQ n RQa
n RQa
(7.31)
n 1
The denominator in this expression is the slope of the p, Q curve in the vicinity of Qa.
So far, this analysis has concentrated upon one single airway. However, suppose that we now
move from branch to branch in a consistent direction around a closed mesh, summing up both
the p values and the p, Q slopes.
n
p = (RQ n RQa )
= RQ n RQa
and sum of p, Q slopes = (nRQa
n 1
We can now write down a composite value of Q for the complete mesh in the form of equation
(7.31)
7- 15
Qm
RQ n RQa
(n RQa
n 1
(7.32)
Note that Qm is no longer the correction to be made to the assumed airflow in any one branch
but, rather, is the composite value given by dividing the mean p by the mean of the p, Q slopes.
Equation (7.32) suffers from the disadvantage that it contains the unknown true values of airflow,
Q. However, the term RQ n is the sum of the corresponding true values of frictional pressure
drop around a closed mesh and Kirchhoff's second law insists that this must be zero. Equation
(7.32) becomes
Qm
RQa
n RQa
(7.33)
n 1
Another glance at Figure 7.4 reminds us that the pressure term in the numerator must always
have the same sign as the airflow, while the p, Q slope given in the denominator can never be
negative. Hence, equation (7.33) may be expressed as
Qm
where Qa
n 1
RQa Qa
n R Qa
n 1
(7.34)
n 1
means the absolute value of Qan-1.
There are a few further observations that will help in applying and understanding the Hardy Cross
process. First, the derivation of equation (7.34) has ignored both fans and natural ventilating
pressures. Like airways, these are elements that follow a defined p, Q relationship. Equation
(7.34) may be expanded to include the corresponding effects
Qm
(RQa Qa
(n R Qa
n 1
n 1
pf nvp)
+ Sf + Snv )
(7.35)
where pf and nvp are the fan pressures and natural ventilating pressures, respectively, that exist
within the mesh,
and
Sf and Snv are the slopes of the p, Q characteristic curves for the fans and natural
ventilating effects.
In practice, Snv is usually taken to be zero, i.e. it is assumed that natural ventilating effects are
independent of airflow.
For most subsurface ventilation systems, acceptable accuracy is achieved from the simple
square law where n = 2, giving
Qm
(RQa Qa pf nvp)
(2 R Qa + Sf + Snv )
(7.36)
The Hardy Cross procedure may now be summarized as follows:

(a) Establish a network schematic and choose at least (b-j+l) closed meshes such that all
branches are represented (section 7.3.1.2). [b = number of branches and j = number of junctions
7- 16
in the network.] Convergence to a balanced solution will be improved by ensuring that each high
resistance branch is included in only one mesh.
(b) Make an initial estimate of airflow, Qa, for each branch, ensuring that Kirchhoff I is obeyed.
(c) Traverse one mesh and calculate the mesh correction factor Qm from equation (7.35) or
(7.36).
(d) Traverse the same mesh, in the same direction, and adjust each of the contained airflows by
the amount Qm.
(e)Repeat steps (c) and (d) for every mesh in the network.
(f)Repeat steps (c), (d) and (e) until Kirchhoff II is satisfied to an acceptable degree of accuracy,
i.e. until all values of (RQi Qi
n 1
pf nvp) are close to zero, where Qi are the current
values of airflow.
A powerful feature of the method is that it is remarkably tolerant to poorly estimated initial
airflows, Qa. This requires some explanation as equation (7.28) and the more general equation
(7.29) were both derived on the assumption that Q was small compared to terms involving Qa.
In practice, it is observed that the procedure converges towards a balanced solution even when
early values of Q are large. There are two reasons for this. First, the compilation of a composite
mesh Q, tends to dampen out the effect of large Q values in individual branches. Secondly,
any tendency towards an unstable divergence is inhibited by the airway p, Q curves always
having a consistent (non-negative) slope.
Another advantage of the Hardy Cross technique is its flexibility. Using equation (7.35) for the
mesh correction factor allows a different value of the logarithmic index, n, to be used for each
airway if required. For subsurface ventilation circuits, this is seldom necessary although special
purpose programs have been written that allow either laminar (n = 1) or turbulent (n = 2) airflow in
each branch.
A more general observation is that Kirchhoff's first and second laws are not interdependent.
Kirchhoff I (Q = 0) may be obeyed at every junction in a network that is unbalanced with respect
to Kirchhoff II. Indeed, this is usually the case at the stage of selecting assumed airflows, Qa.
Furthermore, if one or more branches are to have their airflows fixed by regulators or booster
fans, then those airways may be omitted from the Hardy Cross analysis as they are not to be
subjected to Qm adjustments. Such omissions will result in Kirchhoff I remaining apparently
violated at the corresponding junctions throughout the analysis. This does not affect the ability of
the system to attain a balanced pressure distribution that obeys Kirchhoff II.
Example
Figure 7.5 (a) is the schematic of a simple network and gives the resistance of each branch. A fan
produces a constant total pressure of 2000 Pa and the airflow in branch 3 is to be regulated to a
fixed airflow of 10m3/s. Determine the distributions of airflows and frictional pressure drops, and
the resistance of the regulator required in branch 3.
Solution
Although the Hardy Cross procedure is widely utilized in computer programs, its tedious
arithmetic iterations ensures that it is seldom employed for manual application. For the purposes
of illustration, in this example we shall follow the procedures that illustrate those employed in a
typical ventilation network analysis software package.
7- 17
(a) Mesh selection:

There are 10 branches labelled in the network. However, branches 1 and 5 are connected in
series and may be considered as a single branch for the purposes of mesh selection.
Furthermore, branch 3 has a fixed (regulated) airflow and, hence, need not enter into the Hardy
Cross analysis. The effective number of branches is, therefore, b = 8 while there are j = 6
junctions, giving the required minimum number of meshes to be b - j + 1 = 3. An arrow is
indicated to give each branch a positive direction. This is usually chosen as the direction in which
the airflow is expected to move although either direction is acceptable. The branch arrow is
merely a convenience to assist in traversing closed meshes within the network.
If we write down a list of branch resistances in order of decreasing value, then the three at the top
of this list will correspond to branches 9, 10 and 7. We shall choose our first mesh commencing
on branch 9 and close the mesh through a convenient route but without using branches 10, 7 or
the fixed quantity, branch 3. Figure 7.5 (a) shows the loop starting on branch 9 and traversed by
mesh 1. The second mesh commences on branch 10 and closes without traversing branches 9,
7, or 3. Similarly, the third mesh commences on branch 7 and involves no other high resistance
or fixed quantity branch. The routes of the meshes so chosen are indicated on Figure 7.5 (a).
This technique of mesh selection is sometimes known as the branch tree method and leads to
efficient convergence towards a balanced network as each high resistance branch appears in one
mesh only.
A fourth mesh is selected, commencing on the fixed quantity (branch 3) and closing through any
convenient route that does not include other fixed quantities, and irrespective of branch
resistance values. This extra mesh will not be employed during the Hardy Cross analysis but will
prove useful for the final calculation of the regulator resistance.
(b) Estimate initial airflows:
This is accomplished in two stages commencing from zero flow throughout the network. First,
mesh 4 (the fixed quantity mesh) is traversed in a direction defined by the fixed airflow, adding a
Qm, of 10 m3/s (the required fixed airflow) to each branch around that mesh. As in all
subsequent applications of mesh correction factors, Qm, the actual correction to each branch
flow is Qm multiplied by +l if the branch direction arrow coincides with the path of the traverse, or
-1 if the branch direction arrow opposes the path of the traverse. This device gives the required
airflow in the fixed quantity branch while ensuring that Kirchhoff I remains true at each junction. A
similar procedure is followed in each of the three normal meshes, but employing a value of
Qm = 1. This merely ensures that the denominator of the mesh correction equation (7.35 or
7.36) will be non-zero during the first iteration. The airflows at this stage are indicated on Figure
7.5(a) and show the distribution at the commencement of the Hardy Cross procedure. It should
be noted that if the method were to be applied manually then the number of iterations could be
reduced by estimating a more realistic initial distribution of airflow.
(c) Calculate mesh correction factors:
Assuming the flow to be turbulent in each branch, we may utilize the square law and, hence,
equation (7.36) to calculate the mesh correction factors. In this example, there are no natural
ventilating pressures (nvp = Snv = 0) and the fan is assumed to remain at a fixed pressure
(pf = 2000 Pa and Sf = 0).
Applying equation (7.36) for mesh 1 gives
Qm1 =
=
(0.9 12 0.3 10 2 0.12 02 )

2 (0.9 1 + 0.3 10 + 0.12 0)
29.1
= 3.73 m3/s
7.8
7- 18
R=0.8
10
R=0.25
R=0.9
8
R=0.6
3
1
R=0.2
9
R=0.12
R=0.05
R=0.3
10
10
11
11
R=0.1
2000 Pa
Branch number
R=0.15
Mesh number
Figure 7.5(a). Example network showing airway resistances, initial assumed airflows
and the meshes chosen.
(d) Apply the mesh correction factor:
Mesh 1 is traversed once again, correcting the flows algebraically by adding Qm1 = 3.73. The
relevant branch flows then become:
Branch 9:
Branch 4:
Branch 8:
1 + (3.73 x 1) = 4.73 m3/s

10 + (3.73 x -1) = 6.27 m3/s
0 + (3.73 x -1) = -3.73 m3/s
Applying these corrections around mesh 1 gives the airflow distribution shown in Figure 7.5(b).
7- 19
R=0.8
10
R=0.25
R=0.9
8
R=0.6
3
1
R=0.2
9
R=0.12
4.73
3.73
R=0.05
2
10
R=0.3
6.27
11
11
R=0.1
Branch number
2000 Pa
R=0.15
Mesh number
Figure 7.5(b). Example network showing airflows after Qm1 = 3.73 m3/s has been applied to
mesh 1.
(e) Completion of mesh corrections
Equation (7.36) for mesh 2 now gives
Q m 2 =
(0.8 12 0.12 3.73 2 + 0.3 6.27 2 + 0.25 112 + 0.25 2 2 2000 )

3
= 153.47 m /s
2(0.8 1 + 0.12 3.73 + 0.3 6.27 + 0.25 11 + 0.25 2)
and applying these corrections around mesh 2 gives the airflow distribution shown in Figure
7.5(c).
7- 20
R=0.8
10
R=0.25
R=0.9
154.47
8
R=0.6
R=0.12
4.73
155.47
149.74
R=0.2
9
R=0.05
2
10
R=0.3
159.74
164.47
164.47
R=0.1
Branch number
2000 Pa
R=0.15
Mesh number
Figure 7.5(c). Example network showing airflows after Qm2 = 153.47 m3/s has been applied to
mesh 2.
Repeating the process for mesh 3 gives
(0.6 12 0.2 9 2 + 0.25 155.47 2 )
m3/s
= 73.03
2(0.6 1 + 0.2 9 + 0.25 155.47 )
and applying this correction around mesh 3 yields the airflow distribution shown on Figure 7.5(d),
completing the first iteration for the network.
Q m 3 =
7- 21
R=0.8
10
R=0.25
R=0.9
154.47
8
R=0.6
3
72.03
82.44
R=0.2
82.03
R=0.12
4.73
149.74
R=0.05
2
10
R=0.3
159.74
164.47
164.47
R=0.1
Branch number
2000 Pa
R=0.15
Mesh number
Figure 7.5(d). Example network showing airflows after Qm3 = -73.03 m3/s has been applied to
mesh 3 and completing the first iteration.
(f) The processes involved in (c), (d) and (e) are repeated iteratively until Kirchhoff's second law
(numerator of equation (7.35)) balances to within a small closing error (say 1 Pa) for each mesh.
The airflow correction factors and sums of pressure drops for each mesh are shown in Table 7.1.
Notice how the initial large oscillations are followed by a controlled convergence towards zero
despite the large opening values of Qm.
The final flow pattern arrived at after these nine iterations is given on Figure 7.5(e). The
corresponding frictional pressure drops are each calculated as p = RQ2 and shown in
parentheses against each airflow.
7- 22
Iteration
1
2
3
4
5
6
7
8
9
Sum of Pressure Drops Around

Mesh
p (Pa)
mesh1
mesh2
mesh3
-29
-1958
6027
-10326
28473
-4379
5367
8810
-1231
1579
2524
-358
485
613
- 91
128
114
-15
23
16
-2
3.2
2.0
0.2
0.4
0.2
-0.03
Mesh Correction Factor

Qm(m3/s)
mesh1
3.73
73.60
-34.26
-16.49
-7.18
-2.25
-0.434
-0.059
- 0.007
mesh2
153.47
-64.63
-33.36
-14.71
-4.69
-0.97
-0.143
-0.017
-0.002
mesh3
-73.03
34.17
16.66
7.43
2.40
0.45
0.056
0.007
0.001
Table 7.1 Mesh correction factors and mesh pressure drops converge towards zero as the
iterations proceed.
R=0.25
R=0.9
R=0.8
35.95
(1034)
R=0.12
R=0.6
25.10
(158)
10.85
(71)
R=0.2
20.85
(87)
18.30
(40)
R=0.05
R=0.3
10
(5)
28.30
(240)
(1140)
17.65
(280)
45.95
(317)
45.95
(211)
R=0.1
2000 Pa
R=0.1
Figure 7.5(e). Final airflows and frictional pressure

drops after nine iterations.
7- 23
The pressure drop across the regulator, preg, in branch 3 can be determined from Kirchhoff's
second law by summing the known pressure drops around mesh 4
preg = 2000 - (211 + 87 + 5 + 240 + 317) = 1140 Pa
As the airflow through the regulator is known to be 10 m3/s, the square law gives the regulator
resistance as
Rreg
preg/102 = 11.4
Ns2/m8
The actual dimensions of regulator orifice that will give this resistance may be determined using
section A5.5 given at the end of Chapter 5.
7.4. VENTILATION NETWORK SIMULATION PACKAGES

Between the 1960s and the 1990s, the development and increasing sophistication of ventilation
network simulation packages revolutionized the methodologies of mine ventilation planning.
Ventilation engineers wishing to use modern software need give little conscious thought to what
happens within the computer when they enter data or execute a simulation. Nevertheless, a
grasp of the structure of a network package provides an understanding of the scope and
limitations of the software, and promotes efficiency in its use. This section of the text provides an
outline of how a ventilation network simulation package operates.
7.4.1 Concept of a mathematical model
Suppose that a vehicle leaves its starting station at time t1 and travels for a distance d at an
average velocity of u. The time at which the vehicle arrives at its destination, t2 is given by
d
u
This equation is a trivial example of a mathematical model. It simulates the journey in
sufficient detail to achieve the objective of determining the arrival time. Furthermore, it is general
in that it performs the simulation for any moving object and may be used for any given values of
t1, d and u. Different data may be employed without changing the model.
t 2 = t1 +
More sophisticated mathematical models may require many equations to be traversed in a logical
sequence so that the result of any one calculation may be used in a later relationship. This logical
sequence will often involve feed-back loops for iterative processes such as the Hardy Cross
procedure.
A simulation program is a mathematical model written to conform with one of the computer
languages. The basic ambition of a simulation program is to produce numerical results that
approximate those given by the real system. There are three considerations that govern the
accuracy of a simulation program; first, the adequacy to which each individual process is
represented by its corresponding equation (e.g. p = RQ2); secondly, the precision of the data
used to characterize the actual system (e.g. airway resistances) and thirdly, the accuracy of the
numerical procedure (e.g. the cutoff criterion to terminate cyclic iterations).
Many simulation programs represent not one but a number of interacting features comprising
both physical systems and organizational procedures. In this case, several computer programs
and data banks may be involved, interacting with each other. In such cases, the computer
software may properly be referred to as a simulation package rather than a single program. This
7- 24
is the situation for current methods of ventilation network analysis on desktop personal
computers.
7.4.2 Structure of a ventilation simulation package
Figure 7.6 shows the essential structure of a ventilation network analysis package for a desktop
or laptop computer. It includes a series of input files with which the user can communicate
through a keyboard and screen monitor, either for new data entry or for editing information that is
already held in a file. All files are normally maintained on a magnetic disk or other media that may
be accessed rapidly by the computer.
Screen/Keyboard
Interaction with user
User Input
files
Fan data
bank
Co-ord data
files
Screen/printed
graphical
output
Network
Input file
VNET PROGRAM
Network
Output file
Screen/printed
tabular output
Output data
files
Figure 7.6 Data flow through a ventilation network analysis software package.
7- 25
The User Input Files shown on Figure 7.6 hold data relating to the geometric structure of the
network, the location of each branch being defined by a from and to junction number.
This same file contains the type of each branch (e.g. fixed resistance or fixed [regulated] airflow)
and information relating to branch resistances. The latter may be given (a) directly by the user,
(b) defined as a measured airflow/pressure drop or (c) implied by a friction factor, airway
geometry and an indication of shock losses. Additionally, the user may enter a value of resistance
per unit length of entry that has been established from the mine records as typical for that type of
entry (e.g. intake, return or conveyor). Although convenient, this facility should be used with
caution as it is tempting to overlook the large effects of variations in the sizes of entries or shock
losses arising from bends or obstructions. The User Input File allows entry of resistance data in
whatever form they are available. User input files also allow the operator to enter the location and
corresponding operating pressure of each fan in the network as well as natural ventilating
pressures.
The Fan Data Bank is a convenient alternative means of storing the pressure/volume coordinates
for a number of fan characteristic curves. Data from fan tests and manufacturers' catalogues can
be held in the Fan Data Bank and recalled for use in the subsequent analysis of any network.
The ventilation network analysis program (denoted as VNET on Figure 7.6) in most commercial
packages is based on the mesh selection process and Hardy Cross technique described and
illustrated by example in Section 7.3.2. However, those procedures have been modified to
maximize the rate of information flow, the speed of mesh selection and to accelerate
convergence of the iterative processes. In order to achieve a high efficiency of data transfer, the
VNET program may require that the network input data be specified in a closely defined format.
However, this is contrary to the need for flexibility in communicating with the user. Hence, Figure
7.6 shows a Network Input File acting as a buffer between the user-interactive files and the VNET
program. Subsidiary management programs, not shown on Figure 7.6, carry out the required
conversions of data format and control the flow of information between files. Similarly, the output
from the VNET program may not be in a form that is immediately acceptable for the needs of the
user. Hence, that output is dumped temporarily in a Network Output File. Under user control it
may then be (a) copied into an output data file for longer term storage, (b) displayed in tabular
form on the screen or printer, (c) produced as a plotted network schematic on a screen or
hardcopy printer or plotter and (d) transferred to other graphical software for incorporation into
mine maps.
If graphical output is required then the user-interactive Coordinate Data File shown on Figure 7.6
must first be established. This contains the x, y, z coordinates of every junction in the network. It
is convenient to enter such data directly from a drawn schematic by means of a digitizing pad or
plotter, or from CAD (computer assisted design) files that have already been established for mine
mapping. Alternatively, the schematic may be developed or modified directly via the monitor
screen. To edit an established Coordinate Data File it is often quicker to enter numerical values
for the amended or additional x,y,z coordinates.
7.4.3 Operating system for a VNET package
When running a ventilation network analysis package, the user wishes to focus his/her attention
on ventilation planning and not to be diverted by matters relating to the transfer of data within the
machine. Keyboard entries are necessary for control of the computer operations. However, these
should be kept sufficiently simple that they allow a near automatic response from the experienced
user. Messages on the screen should provide a continuous guide on operations in progress or
alternative choices for the next operation. For these reasons, a simulation package intended for
industrial use must include a management or executive program that controls all internal data
transfers.
The user is conscious of the executive program only through the appearance of a hierarchy of
toolbars, menus and windows on the screen. Each menu lists a series of next steps from which
7- 26
the user may choose. These include progression to other toolbars, menus or windows. Hence,
with few keyboard entries, and with very little conscious effort, the trained user may progress
flexibly and rapidly through the network exercise.
Figure 7.7 illustrates the executive level (master menu) of one ventilation network package.
Choosing the "execute ventilation simulation option will initiate operation of the VNET program
(Figure 7.6) using data currently held in the Network Input File. The exit option simply
terminates operation of the network simulator. Each of the other four options leads to other
menus and submenus for detailed manipulation of the operating system.
Execute
ventilation
simulation
View/print
listed output
Manage
Network files
Save/exit
EXECUTIVE
PROGRAM
Gas
Distribution
analysis
View/plot
schematic
Figure 7.7. Structure of an internal management system for a ventilation network software package.
7.4.4 Incorporation of air quality into a network simulation package

The primary purpose of a ventilation network analysis package is to predict the airflows and
pressure differentials throughout the network. However, the quality of the air with respect to
gaseous or particulate pollutants, or its psychrometric condition also depend upon the distribution
of airflow. Separate programs designed to predict individual aspects of air quality may, therefore,
be incorporated into a more general ventilation package. Branch airflows and/or other data
produced as output from the network analysis program may then be utilized as input to air quality
simulators.
As an example, the operating system depicted by Figure 7.7 incorporates a Gas Distribution
Analysis. This requests the user to identify the emitting locations and rates of production of any
7- 27
gas, smoke or fumes. It then retrieves the airflow distribution contained in the Network Output File
or any named output data file (Figure 7.6) in order to compute the distribution and concentrations
of that pollutant throughout the network.
It is important that users be cognizant of the limitations of such adjunct programs. For instance,
computer simulated gas distributions assume a steady state mixing of the pollutant into the
airflow and, hence, that the pollutant is removed from the system at the same rate as it is emitted.
This is acceptable for normal ventilation planning but will give misleading results if the airflow in
any area is so low, or the rate of emission so high that local accumulations (e.g. gas layering)
would occur in the real system. Such conditions might exist, for example, where ventilation
controls have been disrupted by a mine explosion.
7.4.5 Obtaining a ventilation simulation package
The first step in establishing a ventilation network analysis package at any mine or other site is to
review the availability of both hardware and software. There are several ventilation network
packages available for personal computers. The potential user should consider the following
factors:
Hardware requirements and size of network
The internal memory requirements of the computer will depend upon the scope of the software
package and the largest network it is capable of handling. With modern personal computers,
memory should seldom be a limiting factor. In addition to the microprocessor, keyboard and
screen monitor, the type of package described in Section 7.4.2 and 7.4.3 will require a printer
capable of handling graphical output. Again, most modern printers have this facility. For greater
flexibility, an x,y plotter is useful and is normally available in mine surveying offices. Exporting
network results to CAD/CAM systems allows results to be configured on mine maps.
Cost of software
This can vary very considerably depending upon the sophistication of the system and whether it
was developed by a commercial company or in the more public domain of a university or national
agency.
Speed
The speed of running a network depends upon the type of microprocessor, operating electrical
frequency, and the size and configuration of the network. Typical run times should be sought from
software developers for a specified configuration of hardware. For the majority of mine networks,
computing speed is unlikely to be a limiting factor with current or future generations of personal
computers.
Scope and ease of use
When reviewing simulation packages it is important that the user chooses a system that will
provide all of the features that he/she will require in the foreseeable future. However, there is little
point in expending money on features that are unlikely to be used. Furthermore, it is usually the
case that the more sophisticated packages demand a greater degree of user expertise.
Demonstration versions may be available to indicate the scope and ease of use of a program
package. Other users may be approached for opinions on any given system. Software developers
should be requested to give the names of several current users of their packages.
User's manual and back-up assistance
It is vital that any simulation package should be accompanied by a comprehensive user's manual
for installation and operation of the system. A user-friendly help facility should be part of the
software package. Additionally, the producer of the software must be willing and anxious to
provide backup assistance to all users on request. This can vary from troubleshooting on the
telephone or internet to updating the complete software system.
7- 28
7.4.6 Example of a computed network

The ventilation networks of actual mines or other subsurface facilities may contain several
hundred branches, even after compounding relevant airways into equivalent resistances (Section
7.3.1.1). However, in the interests of brevity and clarity, this example is restricted to the simple
two-level schematic shown on Figure 7.8(a). The software employed is VNETPC, a commercially
available ventilation network package.
In response to screen prompts, the user keys in each item of data shown in Table 7.2. This table
is, in fact, a printout of the User Input File (Section 7.4.2). The pressure/volume coordinates for
the fan are either read manually from the appropriate fan characteristic or copied over from a
characteristic that has been entered previously into a fan data bank. Figure 7.8(b) shows the fan
characteristic curve and corresponding pQ points entered into the computer for this example. The
branch data section of the User Input File illustrates that airway resistances may be entered
directly, from survey (p,Q) observations or from airway geometry and friction factor. The software
package employed for this example requires shock losses to be entered as equivalent lengths
(Section 5.4.5).
A request by the user to execute the simulation causes the User Input File to be translated to the
Network Input File (Figure 7.6) and the computed results to be deposited in the Network Output
File. These may be inspected on the screen or listed on a printer as shown on Tables 7.3, 7.4
and 7.5.
In order to produce plotted results, the x,y, z coordinates of each junction must be specified. This
is usually accomplished by transferring the coordinates from a computer aided design (CAD) file.
Alternatively, the coordinates can be entered manually or by placing a mine map or schematic on
a digitizing tablet or plotter and a sight-glass positioned over each junction in turn as directed by
screen prompts. The resulting Coordinate Data File is then combined with the computed results,
on request by the user, to produce the graphical outputs shown on Figures 7.9, 7.10 and 7.11.*
Similar plots may be generated for frictional pressure drops, resistance, airpower loss, and
operating cost. Some programs allow the user to select band widths for colour coding of the plots.
Similar plots may be generated for air quality parameters such as gas concentrations.
Following the relatively time consuming process of entering the initial data for a new network, and
having inspected the corresponding output, amendments can be made to the user-interactive files
or directly to schematics shown on the computer monitor, and the network re-run rapidly. In this
way, errors in data entry can be corrected and further network exercises pursued with only a few
keystrokes. It is this efficiency, rapidity and ease of testing many alternatives that has
revolutionized the business of ventilation system design. Indeed, it is the skill of the engineer in
interrogating computer output and deciding upon the next exercise that governs the speed of
ventilation planning rather than the mathematical mechanics of older methods of network
analysis.
INPUT DATA
VnetPC Model Data and Summary
TITLE: Textbook Example.
SETTINGS (chosen by the User):
Avg. Air Density: 1.20 kg/m3
Avg. Fan Efficiency: 65.0 %
*
Cost of Power: 8.00 c/kWh

Reference Junction: 1 - (Using junction 1 for a point
on the mine surface is a common convention)
Units: SI (The User may choose between SI and
British Imperial units)
The graphical outputs given directly by the VNETPC package show fine detail and clarity. The translations
necessary for incorporating into this textbook have resulted in greatly reduced quality.
7- 29
Main fan
2
Downcast
shaft
Downcast and
Hoisting shaft
109
Work area
Upcast
shaft
110
106
100
107
108
101
204
Level 1
102
103
104
200
105
211
Level 2
Work area
201
207
208
205
206
209
203
Figure 7.8(a) Layout of the example network. Junction numbers are shown in red.
7- 30
210
Table 7.2 Branch input data for the example network (two pages). This input table illustrates
alternative ways of entering airway resistance, R: (a) directly, (b) by friction [k] factor and airway
geometry, or (c) by measured airflow and corresponding frictional pressure drop.
Junctions
From
To
1
100
2
101
210
105
100
100
106
107
108
101
102
103
104
106
106
109
110
200
200
204
205
206
203
209
201
201
207
208
200
211
207
205
3
100
200
101
201
105
3
101
106
107
108
104
102
103
104
105
102
109
110
107
201
204
205
206
203
209
210
203
207
208
209
211
208
103
201
4
Q (fixed)
F (fan)
I (inject/
reject)
Surface
connections
Surf. Intake
Neither
Surf. Intake
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Neither
Surf. Exhaust
Mode of
Measured Measured
setting
Resistance airflow Q press. drop
resistance
Ns2/m8
m3/s
Pa
k Factor
k Factor
k Factor
k Factor
k Factor
k Factor
R
p/Q
R
p/Q
k Factor
R
k Factor
k Factor
k Factor
k Factor
k Factor
R
k Factor
R
k Factor
R
k Factor
k Factor
k Factor
R
R
k Factor
R/L
k Factor
R/L
k Factor
R
R
R
0.00844
0.0031
0.03797
0.01044
0.0016
0.00544
50
0.0608
0
0.08333
0.07504
0
0.08478
0.13688
0.00591
0.51852
0.36667
0.9
0.43981
200
0.084
1.2
0.06455
0.16406
0.66445
0
0.023
0.0336
1.49994
0.14954
0.28
0.1696
0.35
0.164
0.0001
k factor
kg/m3
Resistance
per km
Ns2/m per km
0.004
0.004
0.0075
0.0075
0.0035
0.0035
100
608
60
300
0.008
0.008
0.008
0.012
0.016
0.016
0.019
0.012
0.017
0.014
0.014
0.012
0.8333
0.019
0.14
0.018
Spreadsheet continued on next page
7- 31
Table 7.2 continued

Junctions
From
To
Length
m
1
100
2
101
210
105
100
100
106
107
108
101
102
103
104
106
106
109
110
200
200
204
205
206
203
100
200
101
201
105
3
101
106
107
108
104
102
103
104
105
102
109
110
107
201
204
205
206
203
209
209
201
201
207
208
200
211
207
205
3
210
203
207
208
209
211
208
103
201
4
600
220
600
220
220
600
Equivalent
length
m
0
0
200
0
0
150
Area
Perimeter
sq.m
15.9
15.9
12.57
12.57
19.64
19.64
Fixed
Quantity
cu.m/s
14.14
14.14
12.57
12.57
15.71
15.71
20
670
10
14
707
710
21
700
495
50
20
0
0
0
10
8
8
6
6
14
12
12
10
10
490
10
10
500
10
14
150
500
2000
12
0
25
8
8
8
12
12
12
20
140
200
1800
150
2000
640
0
0
20
0
33
10
14
10
10
14
15
Description
Colour
code
Symbol
Main Downcast Shaft

Main downcast shaft
Hoisting shaft
Hoisting shaft
Upcast shaft
Upcast shaft
doors
Level 1 intake
Level 1 regulator
Level 1 return
Level 1 return
Level 1 regulator
Level 1conveyor
Level 1 travel way
Level 1 travel way
Level 1 conveyor
Level 1 loader road
Level 1 Work Area
Level 1 return
Level 2 shaft bottom
Level 2 intake
Level 2 work area
Level 2 return
Level 2 return
Level 2 return
Level 2 Main
regulator
Level 2 cross-cut
Level 2 conveyor
Level 2 travel way
Level 2 cross-cut
Level 2 travel way
Level 2 travel way
Intake
Intake
Intake
Intake
Return
Return
Default
Intake
Intake
Return
Return
Default
Intake
Intake
Return
Intake
Intake
Default
Return
Default
Intake
Default
Default
Return
Return
None
None
None
None
None
None
DD
None
None
None
None
None
None
None
None
None
None
None
None
DD
None
None
None
None
None
Default
Default
Default
Default
Default
Default
Default
Intake
Return
Return
R
None
None
None
None
None
None
None
None
None
7- 32
Fan total pressure kPa
6
5
4
3
2
1
Airflow
Fan pressure
m3/s
kPa
0
100
200
250
300
350
5.88
5.52
4.82
4.33
3.73
2.89
455
0
0
100
200
300
400
500
Airflow cu. m. per second
Figure 7.8(b) Fan characteristic curve for the example network. The tabulated coordinate points are
entered into the computer. These should encompass the curve but need not be at
even intervals of airflow.
7- 33
Table 7.3 Output listing of branch results. Note that the computer has inserted regulators in three
of the four branches that were specified as fixed quantity airflows in the input data. The computer
has found it necessary to insert a booster fan (active regulation) in order to obtain the specified
airflow in the other fixed quantity branch.
From
1
100
2
101
210
105
100
100
106
107
108
101
102
103
104
106
106
109
110
200
200
204
205
206
203
209
201
201
207
208
200
211
207
205
3
To
100
200
101
201
105
3
101
106
107
108
104
102
103
104
105
102
109
110
107
201
204
205
206
203
209
210
203
207
208
209
211
208
103
201
4
(R)egulator
(B)ooster
or (F)an
Total
Resistance
Ns2/m8
0.00844
0.0031
0.03797
0.01044
0.0016
0.00544
50
0.0608
5.03444
0.08333
0.07504
1.66824
0.08478
0.13688
0.00591
0.51852
0.36667
0.9
0.43981
200
0.084
1.2
0.06455
0.16406
0.66445
0
0.023
0.0336
1.49994
0.14954
0.28
19.48297
0.35
0.164
0.0001
Airflow
m3/s
141.55
43.37
151.88
135.64
140
293.43
3.76
94.41
20
54.35
54.35
20
60.06
99.08
153.43
40.06
34.35
34.35
34.35
2.11
26.26
26.26
19.46
19.46
78.06
140
58.6
85.96
46.94
61.94
15
15
39.02
6.8
293.43
Pressure
drop
Pa
169.1
5.8
875.9
192.1
31.4
468.4
706.8
542
2013.8
246.2
221.7
667.3
305.8
1343.7
139.1
832.2
432.7
1062.1
519
893.1
57.9
827.5
24.4
62.1
4048.4
0
79
248.3
3305.4
573.8
63
4383.7
532.8
7.6
8.6
Airpower
loss
kW
23.94
0.25
133.03
26.06
4.4
137.44
2.66
51.17
40.28
13.38
12.05
13.35
18.37
133.13
21.34
33.34
14.86
36.48
17.83
1.88
1.52
21.73
0.47
1.21
316.02
0
4.63
21.34
155.16
35.54
0.95
65.76
20.79
0.05
2.52
Operating
cost
$/year
25807
271
143429
28093
4740
148184
2865
55169
43424
14427
12991
14389
19802
143539
23010
35943
16025
39334
19221
2032
1639
23428
512
1303
340716
0
4991
23012
167281
38319
1019
70895
22415
56
2721
Main Downcast Shaft

Main downcast shaft
Hoisting shaft
Hoisting shaft
Upcast shaft
Upcast shaft
doors
Level 1 intake
Level 1 regulator
Level 1 return
Level 1 return
Level 1 regulator
Level 1conveyor
Level 1 travel way
Level 1 travel way
Level 1 conveyor
Level 1 loader road
Level 1 Work Area
Level 1 return
Level 2 shaft bottom
Level 2 intake
Level 2 work area
Level 2 return
Level 2 return
Level 2 return
Level 2 Main regulator
Level 2 cross-cut
Level 2 conveyor
Level 2 travel way
Level 2 cross-cut
Level 2 travel way
Level 2 travel way
Ramp
Level 2 cross-cut
Surface fan drift
7- 34
Table 7.4 Output listing of fan operating (pQ) points, airpower generated and annual operating cost.
Fan
No.
From
To
Fan
Pressure
kPa
Fan
Airflow
m3/s
Fan
curve
Fan
configuration
Airpower
kW
Operating
Cost
$/year
Fan
name
3.809
293.43
On
1 in Parallel
1117.68
1 205 027
Main Fan
Table 7.5 Output listing of fixed quantity branches.

From
106
101
209
211
To
107
102
210
208
Fixed
quantity
m3/s
20
20
140
15
Booster fan
pressure
Pa
Regulator
resistance
Ns2/m8
5.03444
1.66824
1895
19.31337
Branch
resistance
Ns2/m8
0
0
0
0.1696
Total
resistance
Ns2/m8
Orifice
area
m2
5.03444
1.66824
0.59
1.02
19.48297
0.3
Branch
description
Level 1 regulator
Level 1 regulator
Level 2 Main regulator
Level 2 travel way
7- 35
Figure 7.9. Three dimensional output schematic showing airflows (m3/s). 3-D depictions can be rotated on the computer monitor
a particularly useful facility for metal mines.
7- 36
Figure 7.10 Output schematic showing airflows (m3/s) in Level 1
7- 37
Figure 7.11 Output schematic showing airflows (m3/s) in Level 2.
7- 38
References
Atkinson, J.J. (1854). On the Theory of the Ventilation of Mines. North of England Institute
of Mining Engineers. No. 3, p.118.
Cross, H. (1936). Analysis of Flow in Networks of conduits or Conductors. Bull. Illinois
University Eng. Exp. Station. No. 286.
Hartman, H.L. and Wang, Y.J. (1967). Computer Solution of Three Dimensional Mine
Ventilation Networks with Multiple Fans and Natural Ventilation. Int. J. Rock Mech. Sc.
Vol.4.
Maas, W. (1950). An Electrical Analogue for Mine Ventilation and Its Application to
Ventilation Planning. Geologie en Mijnbouw, 12, April.
McElroy, G.W. (1954). A Network Analyzer for Solving Mine Ventilation Distribution
Problems. U.S. Bureau of Mines Inf. Circ. 7704. 13 pp.
McPherson, M.J. (1964). Mine Ventilation Network Problems (Solution by Digital
Computer). Colliery Guardian Aug. 21. pp.253-254.
McPherson, M.J. (1966). Ventilation Network Analysis by Digital Computer. The Mining
Engineer. Vol. 126, No. 73. Oct. pp. 12-28.
McPherson, M.J. (1984). Mine Ventilation Planning in the 1980's. International Journal of
Mining Engineering. Vol. 2. pp. 185-227.
Scott, D.R. and Hinsley, F.B. (1951/52). Ventilation Network Theory. Parts 1 to 5. Colliery
Eng. Vol. 28, 1951; Vol. 29, 1952.
Scott, D.R., Hinsley, F.B., and Hudson, R.F. (1953). A Calculator for the Solution of
Ventilation Network Problems. Trans. Inst. Mine. Eng. Vol. 112. p. 623.
Wang, Y.J. and Saperstein, L.W. (1970). Computer-aided Solution of Complex Ventilation
Networks. Soc. Min. Engrs. A.I.M.E. Vol. 247.
Williams, R.W. (1964). A Direct Analogue Equipment for the Study of Water Distribution
Networks. Industrial Electronics, Vol. 2. pp. 457-9.
7- 39
Chapter 8. Mine Ventilation Thermodynamics

8.1. INTRODUCTION
...................................................................................1
8.2. COMPONENTS OF THE MINE CYCLE ........................................................2

8.2.1. Elements of the system...................................................................................................... 2
8.2.2. The downcast shaft............................................................................................................ 2
Example. .................................................................................................................................. 8
8.2.3. Level workings ................................................................................................................. 14
8.2.4. Upcast shaft. .................................................................................................................... 17
Example. ................................................................................................................................ 17
8.3. THE COMPLETE MINE CYCLE ..................................................................21

8.3.1. Natural ventilation ............................................................................................................ 21
8.3.2. Combined fan and natural ventilation .............................................................................. 26
8.3.3. Case Study ...................................................................................................................... 30
General observations on the table of results ............................................................................. 32
Polytropic indices ........................................................................................................................... 32
Fan performance............................................................................................................................ 33
Natural ventilating energy, NVE .................................................................................................. 33
Natural ventilating pressure. NVP ............................................................................................... 34
Heat additions ................................................................................................................................ 35
8.3.4. Inclined workings ............................................................................................................. 36
Dip circuit................................................................................................................................ 36
Rise Circuit. ............................................................................................................................ 40
8.3.5. The effect of moisture ...................................................................................................... 40
BIBLIOGRAPHY .................................................................................................41
APPENDIX A8.1 .................................................................................................42
Survey data from the survey of the shafts used in the examples given in Sections 8.2.2 and
8.2.4........................................................................................................................................ 42
APPENDIX A8.2 .................................................................................................44

A general steady flow analysis for airstreams that carry water vapour and liquid water
droplets................................................................................................................................... 44
8.1. INTRODUCTION
Many of the world's practising mine ventilation engineers - perhaps, even, the majority perform
their duties very successfully on the basis of relationships that assume incompressible flow.
Some of those engineers may question the need to concern themselves with the more detailed
concepts and analyses of thermodynamics. There are, at least, two responses to that query.
First, if the ranges of temperature and pressure caused by variations in elevation and heat
transfer produce changes in air density that are in excess of 5 per cent, then analyses that ignore
those changes will produce consistent errors that impact significantly on the accuracy of planned
ventilation systems. In practical terms, this means that for underground facilities extending more
than 500 m below the highest surface connection, methods of analysis that ignore the
compressibility of air may be incapable of producing results that lie within observational
tolerances of accuracy.
However, there is a second and even more fundamental reason why all students of the subject,
including practicing engineers, should have a knowledge of mine ventilation thermodynamics.
Although the incompressible flow relationships are simple to apply, they are necessarily based on
an approximation. Air is, indeed, highly compressible. It follows, therefore, that if we are truly to
comprehend the characteristics of subsurface ventilation systems, and if we are really to
understand the behaviour of large-scale airflow systems then this can be accomplished only if we
have a grasp of steady-flow thermodynamics.
In this chapter, thermodynamic analyses are carried out on a downcast shaft, underground
airways and an upcast shaft. The three are then combined to produce a thermodynamic cycle for
a complete mine, first, with natural ventilation only, then for the more usual situation of a
combination of fans and natural ventilation. In all cases, the analyses utilize pressure-volume
(PV) and temperature-entropy (Ts) diagrams. This chapter assumes familiarity with the basic
concepts and relationships that are developed in Chapter 3. It is suggested that the reader study
that chapter before proceeding with this one.
One further point - the temperatures and, to a lesser extent, pressures of the air are affected by
variations of water vapour in the air. The final section of this chapter outlines the modifications
that can be made while a more detailed treatment of moisture in air is given in Chapter 14.
However, for the time being, we shall assume that the airways are dry and not affected by
evaporative or condensation processes.
8.2. COMPONENTS OF THE MINE CYCLE

8.2.1. Elements of the system.
A subsurface ventilation system follows a closed cycle of thermodynamic processes and can be
illustrated by visual representations of those processes on thermodynamic diagrams that are
analogous to the indicator diagrams of heat engines. It was this resemblance that led Baden
Hinsley (1900-1988) into the realization that a mine ventilation system is, indeed, a gigantic heat
engine. Air enters the system and is compressed and heated by gravitational energy as it
descends in downcasting shafts or slopes. More heat is added to the air from the strata,
machines and other sources. Work is done by the air as it expands during its ascent through
upcast shafts or slopes. Some of the added heat is converted temporarily to mechanical energy
and assists in promoting airflow. In the great majority of mines, this "natural ventilating energy" is
supplemented by input fan work.
When the exhaust air re-enters the pressure sink of the surface atmosphere, it cools to the
original entry conditions, closing the cycle. Figure 8.1 illustrates the descending flow through a
downcast shaft between stations 1 and 2, level workings 2 to 3, and returning to surface through
an upcast shaft 3 to 4. We shall analyze each of the three processes separately before adding
the isobaric (constant pressure) cooling in the surface atmosphere to complete the cycle.
8.2.2. The downcast shaft
Air enters a downcast shaft at the pressure and temperature of the atmosphere existing at the
top of the shaft. As the air falls down the shaft, its pressure increases, just as the pressure on a
diver increases as he descends into the ocean. The rise in pressure is caused by the increasing
weight of the overlying column of fluid as we plunge deeper into the fluid. However, that simple
observation belies the common belief that air always flows from a high pressure region to a
connected lower pressure region. In a downcast shaft, exactly the opposite occurs showing how
easily we can be misled by simplistic conceptions.
The process of gravitational compression, or autocompression, in the downcast shaft produces
an increase in temperature of the air. This is independent of any frictional effects and will be
superimposed upon the influence of any heat transfer with the surrounding strata that may occur
across the shaft walls. The rate of that heat transfer depends upon the thermal properties of the
rock and the difference between the rock temperature and air temperature at any given horizon.
2
3
Figure 8.1. Elements of the system
Now, while the temperature of the mass of rock surrounding the shaft may change relatively
slowly with time, the air temperature at the shaft entrance can change from hour to hour and,
especially, between day and night. Because of these surface variations, it is common for the walls
and rock surrounding a downcast shaft to absorb heat during the day and to emit heat during the
night. The phenomenon continues along the intake airways and tends to dampen out the effects
of surface temperature variation as we follow the air into a subsurface facility. This is sometimes
called the "thermal flywheel". For the purposes of this chapter, we shall assume that any heat
exchange that takes place in the downcast shaft is distributed equitably so that the process
approximates closely to a polytropic law.
If we descend a downcast shaft, stopping every hundred metres or so to take measurements of
pressure and temperature, we can plot the PV and Ts diagrams using equation (3.9):
V =
RT
P
m3
kg
(Chapter 14 shows how this relationship should be amended for the presence of water vapour).
Equation (3.48) allows us to determine the variation of entropy
T
P
(s b s a ) = C p ln b R ln b
Ta
Pa
J
kg K
The starting point for the latter equation may be any defined temperature and pressure datum. It
is differences in entropy rather than absolute values that are important.
The diagrams that emerge have the appearance of those given in Figure 8.2. In practice, there is
often a scatter of points near the end extremities of the shaft due to high turbulence and uneven
mixing of thermally stratified airstreams.
The first step in the analysis of the data is to conduct a curve fitting analysis to the points on the
PV diagram. The form of the equation to be fitted is the polytropic law
P =
Vn
where C is a constant and n the polytropic index. Least squares regression analyses are widely
available for personal computers and hand calculators. The quality of the curve fit should be
reviewed. It is sometimes apparent that some points, particularly at shaft ends, should be rejected
or down-weighted.
If the value of the polytropic index is higher than the isentropic index, 1.4, for dry air, then the
combined effect of added heat and internal friction is positive. If, however, the index is less than
1.4 then sensible heat is being lost from the air to the strata or through conversion into latent heat
by evaporation of water. If the measurements have, unwisely, been taken shortly after a
significant change in surface temperature, then a variation in n may become apparent by
comparing the plotted points with the best-fit polytropic line.
In the majority of cases, satisfactory representation by a polytropic law can be obtained. Where
2
significant deviation occurs then the flow work
VdP can be obtained graphically as the area to

1
the left of the curve on the PV diagram.

A thermodynamic analysis of a polytropic compression is given in section 3.5.3. The work done
against friction, F12, in the shaft is given by combining the steady flow energy equation (3.25) with
the evaluation of the flow work term shown by equation (3.73). This gives
ln P2
u12 u 22
J
P1
(8.1)
F12 =
+ (Z 1 Z 2 )g R (T 2 T1 )
2
kg
lnT 2
T1
or, alternatively
u 2 u 22
J
n
F12 = 1
+ (Z 1 Z 2 )g
R (T 2 T1 )
2
kg
n 1
While most terms in these equations are measurable and, in principle, the values at the top
(station 1) and the bottom (station 2) of the shaft can be used to determine F12, there are some
practical pointers that contribute greatly to the achievement of good results. First, it is often
difficult to measure the mean velocity of air in a shaft. It may be preferable to measure the
airflow(s) in underground airways and to carry out the necessary summations and corrections for
air density to give u1 and u2. Fortunately, the magnitude of the kinetic energy term is usually very
small compared to the other factors.
The term (Z1 - Z2) is simply the depth of the shaft connecting stations 1 and 2. The local value of
gravitational acceleration, g, should be ascertained, taking into account mean elevation and
latitude. For the pressures and temperatures P1, P2, T1 and T2, it is inadvisable to use the raw
measurements made at the shaft extremities because of the unstable conditions that may exist at
those locations. It is preferable to construct separate plots of pressure and temperature against
depth and to use best fit lines to establish representative pressures and temperatures at depths
Z1 and Z2.
P2
2
2
V dP
1
P1
Vmd
V
T2
P2
P1
T1
Tds =
1
V dP
F12
q12
S1
S2
Figure 8.2. PV and Ts diagrams for a downcast shaft.
Equation (8.1) gives the amount of mechanical energy that is converted to heat through frictional
processes within the shaft airflow. This can be employed to determine the Chezy Darcy
coefficient of friction, f, for the shaft. From equation (2.46)
f
2d
4Lu 2
F12
(dimensionless)
(8.2)
where d = hydraulic mean diameter of the shaft (m)

L = length of shaft (m) and
u = mean air velocity (m/s)
Although the F term is the primary factor quantifying the work done against friction, it has one
major drawback. It cannot be observed directly but must be computed from measurements
including those of pressure and temperature. On the other hand, we know that the loss of
mechanical energy by friction produces a frictional pressure drop, p, which can be measured
directly (Chapter 6). It is, therefore, easier to conceptualize the notion of a frictional pressure
drop, p, than the fundamental frictional work, F, which produces that pressure drop. Furthermore,
recalling that ventilation planning for facilities of less than 500 m in depth can be conducted using
the simpler relationships of incompressible flow, it is useful to remind ourselves of the relationship
between p and F. This was introduced in Chapter 2.
p = F
(8.3)
Pa
Hence, the work done against friction determined from equation (8.1) can be converted to a
corresponding frictional pressure drop
p12 = F12
Pa
Now we have another problem. The density, , varies throughout the shaft, so what value do we
select? One option is to employ a mean value, md, where the subscript md denotes "mean
downcast." For polytropic flow in a dry shaft the variations in both temperature and pressure are
near linear with respect to depth. It follows that air density, = P/RT, also increases in a near
linear manner. Hence the arithmetic mean of densities measured at equal intervals throughout
the shaft may be employed. Again, observers should be cognizant of the unrepresentative values
that may be measured at shaft extremities. More generally,
2
md
dZ
kg
(Z1 Z 2 )
m3
(8.4)
Using the mean density, and ignoring the small kinetic energy term, the steady flow energy
equation can be written as
(Z1 Z 2 )g = Vmd (P2 P1 ) + F12
where Vmd = 1/md (mean specific volume)

or
p12 = F12 md = md (Z1 Z2 )g (P2 P1)
Pa
(8.5)
This is simply a form of Bernoulli's equation for incompressible flow. We have, however, taken
account of compressibility by using a mean value of air density. Equation (8.5) is highlighted as
we shall return to it later as a component in an approximation of natural ventilation pressure.
One further problem remains. If we wish to compare the frictional work done in a series of
different airways then the F terms should be employed as these depend only upon the physical
characteristics of the airway and the air velocity (equation (8.2)) - and are independent of the
thermodynamic state of the air. On the other hand, if we compare frictional pressure drops, using
equation (8.5), then these depend also upon the corresponding values of air density. As the latter
may vary considerably around a mine circuit, frictional pressure drops are not a good basis for
comparison. (This was the reason for consistent deviations that were noticed in the results of
some pressure surveys prior to the development of mine ventilation thermodynamics). The
problem can be met by choosing a standard" value of air density, st, to relate and F in
equation (8.3). If the same value is used for all airways, then the comparison between airways
becomes a constant multiple of F. The standard value of air density is normally taken as 1.2
kg/m3 and the corresponding "frictional pressure drop", st, should then be identified as being
'standardized', or referred to standard density. It should be understood that standardized pressure
drops have no direct physical significance other than the fact that they are a constant multiple of
work done against friction.
2
Let us turn now to the Ts diagram on Figure 8.2. The area under the process line
Tds
1
represents the combination of added heat and the internal generation of friction heat q12 + F12.
Fortunately, the three-part steady flow energy equation (3.25) allows us to separate the two
terms. Equation (8.1) gives F12, and
q12 = (H 2 H 1 )
u12 u 22
(Z1 Z 2 )g
2
J
kg
(8.6)
If the shaft is dry, then (H2 H1) = Cp (T2 T1) and by ignoring the change in kinetic energy, we
have
J
kg
q12 = C p (T 2 T1 ) (Z 1 Z 2 )g
(8.7)
In many shafts that have reached an effective equilibrium with respect to the temperature of the
surrounding strata there is very little heat transfer, i.e. adiabatic conditions with q12 = 0. Equation
(8.7) then gives
(T 2 T1 ) =
(Z1 Z 2 )
Cp
(8.8)
This well known equation allows the increase in temperature with respect to depth to be
calculated for a dry shaft in the absence of heat transfer. Inserting the constants 9.81 m/s2 for g
and 1020 J/kgK for Cp (value of specific heat of air with a moisture content of 0.0174 kg water
vapour per kg dry air) gives
(T 2 T1 ) = 0.00962 (Z1 Z 2 )
(8.9)
This result is usually rounded off and quoted as a dry bulb temperature adiabatic lapse rate of 1
C per 100 m depth. The increase in temperature is a result of potential energy being converted
to internal energy as the air falls through the shaft. The effect is also referred to as
autocompression.
In practice, the change in temperature often varies considerably from 1C/100 m due to both heat
transfer and reductions in dry bulb temperature caused by evaporation of water.
In section 3.5.3, it was shown that in addition to areas on the Ts diagram illustrating heat, the
same diagram may be used to indicate numerical equivalents of flow work. In particular, on the Ts
diagram in Figure 8.2, the total area under the isobar P2 between temperatures T2 and T1
represents the enthalpy change (H2 H1).
However, the flow work is given by the steady flow energy equation as
2
VdP
= (H 2 H 1 ) (q12 + F12 )
J/kg
(8.10)
As the areas representing (H2 H1) and (q12 + F12) have already been identified on the Ts
diagram, their difference gives a representation of the flow work.
Example.
This example is taken from an actual survey of an 1100 m deep downcast shaft of diameter
5.5m.1 At the time of the survey, the mass flow of air was 353 kg/s. Figures 8.3 (a), (b) and (c)
show the barometric pressures, wet and dry bulb temperatures and air density measured at
approximately 100 m intervals throughout the shaft. The measurements taken at the extremities
of the shaft have been omitted from these graphs as the turbulence at these points was not
conducive to stable or precise readings. The raw data for the survey are tabulated in Appendix
A8.1.
Variations in pressure, temperature and density.

The increase in pressure with respect to depth shown in Figure 8.3(a) is seen to be near linear
with a slope of 1.22 kPa per 100m, The mean air density in the shaft was 1.26 kg/m3 illustrating
the convenient rule of thumb that the rate of pressure increase in kPa/100m is approximately
equal to the mean density (in SI units). This follows from
or
gh
981
0.981
Pa
kPa
where g = 9.81 m/s2 and h = 100m

The temperature plots shown in Figure 8.3(b) should, theoretically, also be linear with respect to
depth for adiabatic conditions or where heat transfer is distributed uniformly throughout the length
of the shaft. In this case, the actual variation in dry bulb temperature shows an interesting feature.
The dry bulb temperature plot appears to be divided into discrete regions above and below 700m.
Visual observation indicated the existence of free water in the shaft. The lower part passed
through an aquifer. In this section tubbing lining was employed and an increased water make
produced falling droplets. The effects of enhanced evaporative cooling are shown clearly as a
reduction in the slope of the dry bulb temperature-depth line in the lower part of the shaft. The
mean dry bulb temperature lapse rate is 0.82 C/100 m down to 700m and 0.67 C/100 m at
lower elevations, both less than the adiabatic lapse rate of 0.962 C/100 m (equation (8.9)2. This
shows the important influence of water in the airway.
The wet bulb temperature observations exhibit slightly greater scatter but still follow a well defined
linear relationship with respect to depth, irrespective of evaporation, and having a lapse rate of
0.44 C/100 m. This is a little less than the wet bulb adiabatic lapse rate of 0.456 C/100 m given
by Figure 15.8 in Chapter 15. Hence, at the time of the survey there was a small heat loss from
the air into the surrounding strata.
1
Full details of the survey on which these examples were based are given in the paper listed in the
Bibliography for this Chapter: Barometric Survey of Shafts at Boulby Mine, Cleveland Potash, Ltd.,
Cleveland, McPherson M.J. and Robinson, G.
2
The mean moisture content in the shaft was 0.0108 kg/kg dry air. Inserting this value into equation (15.42)
of Chapter 15 gives a corrected dry bulb adiabatic lapse rate of 0.967 C per 100 m.
Barometric Pressure kPa
116
114
112
110
y = 0.0122x + 100.52
108
106
104
102
100
98
0
200
400
600
800
1000
1200
Depth m
Figure 8.3(a). Measurements of barometric pressure in the downcast shaft with a best-fit
straight line.
28
Temperature deg. C
26
y = 0.0067x + 18.064
24
Dry bulb temperature
22
y = 0.0082x + 17.029
20
y = 0.0044x + 15.571
18
Wet bulb temperature
16
14
0
200
400
600
Depth
800
1000
1200
Figure 8.3(b). Increase in wet and dry bulb temperatures with respect to depth in the downcast shaft.
Air Density kg/cu.m
1.34
1.32
1.3
1.28
1.26
1.24
1.22
1.2
1.18
0
200
400
600
800
1000
1200
Depth m
Figure 8.3(c). Increase in air density with respect to depth in the downcast shaft
with a best-fit straight line.
Figure 8.3(c) illustrates that the air density increases linearly with dept.
PV diagram, work dissipated against friction and frictional pressure drop.

Figure 8.4 shows the PV diagram for the shaft. Despite the non-linearity of the temperature, a
curve fitting exercise produced a good fit with the polytropic equation
PV 1.3004
= 79.451
Again, the non-representative points at the shaft ends were omitted for the curve fit. Figure 8.4
shows the curve of this equation superimposed upon the actual observations. The fact that the
polytropic index was less than the isentropic 1.4 showed that sensible heat was being removed
from the air - again, the combined effect of evaporative cooling and heat transfer.
The measurements may be used to determine the work done against friction within the airstream.
Equation (8.1) gives:ln P2
P1
+ (Z1 Z 2 )g R(T 2 T1 )
F12
lnT 2
T1
where subscripts 1 and 2 refer to the top and bottom of the shaft respectively.
2
u u2
= 1
2
10
116
114
112
Pressure kPa
110
108
106
y = 79.451x -1.3004
104
102
100
98
96
0.7
0.75
0.8
0.85
Specific Volume cu.m/kg
Measured
Curve fit
PV1.3004 = 79.45
where P is in kPa
Figure 8.4. PV diagram for the downcast shaft.
Independent measurements and corrections for density gave u1 = 12.78 and u2 = 11.63 m/s. The
shaft depth, (Z1 - Z2), was 1100 m and the value of g was 9.807 m/s2. If the equation is to be used
for the complete shaft then careful consideration must be given to the end values of pressure and
temperature. The lack of uniform conditions at the shaft extremities indicates that single
measurements of pressure and temperature at those locations are likely to produce erroneous
results. The pressure and temperature plots should be examined carefully in order to select
values of those parameters that are representative of an extrapolation of trends within the shaft.
In this case, the values chosen from the relevant plots were:
11
P1 = 100.79 kPa.
T1 = 17.2 + 273.15 = 290.35 K
and
P2 = 113.90 kPa
T2 = 25.5 + 273.15 = 298.65 K
Equation (8.1) then gives
)
)
ln 113.9
12.78 2 11.63 2
100.79
+ (1100 9.807 ) 287.04(25.5 17.2)
298
.
65
2
ln
290.35
=
14
+
10 788
10 336
J/kg
F12
466
J/kg
This calculation illustrates that the kinetic energy term is, indeed, small compared with the other
terms. More importantly, the potential energy and flow work terms are both large but similar
numbers. Hence, small errors in the parameters that contribute to each of those terms can have a
significant impact on the calculated value of F12. For this reason, it is preferable to apply equation
(8.1) for each increment of the shaft and sum the individual values of F rather than to use the
complete shaft. This approach also accounts for variations in the polytropic index throughout the
shaft. When this was done for the current example, the cumulative F12 was 493 J/kg.
The flow work may also be calculated from the alternative expression (equation (8.1))
n
1.3004
R (T 2 T1 ) =
287.04 (25.5 17.2) = 10 313
0.3004
(n 1)
J/kg
giving
= 14 + 10 788 10 313 = 489 J/kg
F12
We shall continue the analysis using the value of F12 = 489 J/kg obtained from this approach.
The frictional work can now be converted to a frictional pressure drop referred to any chosen air
density. The mean density in the shaft was 1.260 kg/m3 giving a corresponding frictional pressure
drop of
p md
= md F12
= 1.26 489 = 616
Pa
Pa
For comparison with other airways, the frictional pressure drop referred to standard air density,
1.2 kg/m3, becomes
p = st F12
Pa
= 1.2 489 = 587
Pa
Coefficient of friction and airway resistance:

The Chezy Darcy coefficient of friction is given by equation (8.2)
f
2d
4 Lu 2
F12
12
Using the mean air velocity of 12.20 m/s gives

f
2 5 .5
4 1100 (12.20 ) 2
489
= 0.00821 (dimensionless)
The Atkinson friction factor referred to standard density of 1.2 kg/m3 is
k =
st f
2
1.2
0.00821 = 0.00493
2
kg/m3
(The values of f and k are mean values as the type of lining changed through the aquifer).
The rational resistance of the shaft is given from equation (2.51) as
Rt
f perimeter L
m 4
2 A3
where perimeter = x 5.5 = 17.28 m

and
area, A = (5.5)2/4 = 23.76 m2
giving
Rt
0.00821
17.28
1100 = 0.00582
2
(23.76) 3
m 4
The Atkinson resistance referred to standard density becomes

R = st R t
= 1.2 0.00582 = 0.00698
Ns 2 /m 8
These resistance values refer to the 1100 m length of open shaft. The shock losses caused by
conveyances, inlet and exist losses must be assessed separately and added to give the total
shaft resistance.
Heat exchange.
The steady flow energy equation gives the heat transfer to the air as
2
(u1 u 2 )
(Z1 Z 2 )g
2
The change in enthalpy, taking evaporation into account, may be determined from the methods
given in Section 14.5.1. If no evaporation or condensation occurred then the change in enthalpy
would be given by equation (3.33).
q12 = (H 2 H 1 )
(H 2 H 1 ) = C p (T 2 T1 )
The specific heat of the air, Cp, would be 1005 J/kgK if the air were perfectly dry. Again, Chapter
14 indicates how this can be corrected for the presence of water vapour. The actual value based
on mean moisture content for this shaft was 1025 J/kgK.
13
If the shaft had been dry then

(12.78 2 11.63 2 )
(1100 9.807
2
14
10 788
q12 = 1025 (25.5 17.2)

=
8 507
2 294
J/kg
The physical interpretation of this calculation is that for each kilogram of air:
14 J of heat accrue at the expense of kinetic energy as compression decelerates the air;
10 788 J of thermal energy arise from the loss of potential energy; but
the total increase in thermal energy of the air is only 8 507 J.
hence, 2 294 J of sensible heat must be transferred from the air.
We have conducted this latter calculation on the assumption of a dry shaft. This was not the
situation in the actual case study. However, the result we have obtained does have a real
meaning. The 2 294 J/kg heat loss reflects the reduction in the sensible heat of the air. Most of
this was, in fact, utilized to evaporate water and, hence, was returned to the air as latent heat.
Chapter 14 elaborates on the meanings of these terms.
In order to convert the heat exchange into kilowatts, we simply multiply by the mass flow of air
2 294
353 = 810
1000
kg
kJ
kJ
or kW
=
kg
s
s
8.2.3. Level workings

Figure 8.5 shows the PV and Ts diagrams for flow along level workings between stations 2 and 3.
The diagrams reflect a decrease in pressure but increases in both specific volume and entropy,
indicating heat additions to the air. In practice, a large part of the heat exchange with the strata
takes place in the working areas where rock surfaces are freshly exposed.
Applying the steady flow energy equation to level workings gives
2
(u 2 u 3 )
+ (Z 2 Z 3 )g =
2
VdP
+ F23 = (H 3 H 2 ) q 23
J/kg
(8.11)
However, in this case (Z2 - Z3) = 0, giving

3
VdP = F23
2
Remember that the flow work
(u 2
u3 )
2
J/kg
(8.12)
VdP is the area to the left of the process curve on the PV
diagram and, here, dP is negative (falling pressure). Hence, equation (8.12) shows that work is
done by the air and, in level airways, is utilized entirely against friction and accelerating the air.
Here again, the change in kinetic energy is usually negligible, leaving the simple relationship
3
VdP = F23
J/kg
(8.13)
14
The heat flow into level workings or level airways may also be calculated by applying the
condition (Z2 - Z3) = 0 to the steady flow energy equation.
(u 2
u3 )
2
+ q 23
(H 3 H 2 )
J/kg
(8.14)
Again, changes in kinetic energy are normally negligible and, if we apply the condition of neither
evaporation nor condensation, then
= C p (T3 T 2 )
q 23
J/kg
(8.15)
An interesting situation often arises in well established return airways where equilibrium has
effectively been established between the temperatures of the air and the surrounding rock. Heat
transfer ceases giving adiabatic flow, q23 = 0. However, equation (8.15) shows that if q23 is zero
then T3 = T2. The process becomes simultaneously adiabatic and isothermal. This is similar to a
throttle in a gas stream, an example favoured by authors of textbooks on thermodynamics. In
Section 3.4.1 it was shown analytically that the degree of friction has no influence on the
temperature variation of the air. The adiabatic/ isothermal airway suggests a simple way of
comprehending this phenomenon. An ideal adiabatic (isentropic) decompression would produce a
fall in temperature. However, in a real (non-ideal) adiabatic decompression, frictional heat is
generated within the airflow at exactly the correct rate to counteract that fall in temperature.
3
On the Ts diagram of Figure 8.5, the
Tds heat area under the process line must be equal to the
2
combined effects of added heat and internally generated friction, q23 + F23. However, the area
under the P3 isobar between temperatures T3 and T2 is equal to the change in enthalpy (H3 - H2)
or Cp(T3 - T2) for a dry airway. We have already shown that q23 = Cp(T3 - T2). Hence, we can
identify separate areas on the Ts diagram that represent F23 and q23.
Unlike shafts of depths greater than 500m, most underground airflow paths involve relatively
small changes in density. This is certainly the case for level workings or airways. While such
airways play an important role in the mine thermodynamic cycle, the treatment of individual level
airways is normally based on incompressible flow. The steady flow energy relationship then
reduces to Bernoulli's equation with (Z2 - Z3) = 0
(u 2
u3 )
2
(P3 P2 )
mw
+ F23
J/kg
(8.16)
where the subscript mw denotes 'mean workings'. Neglecting the kinetic energy term, the
frictional pressure drop referred to the mean density becomes simply
p 23
= mw F23
= P2 P3
Pa
(8.17)
15
P2
V dP
2
q23
2
= F 23
P3
P2
P3
T3
T2
T ds
F23
S2
q23
S3
Figure 8.5. PV and Ts diagrams for level workings.
16
8.2.4. Upcast shaft.

Figure 8.6 shows the PV and Ts diagrams for an upcast shaft. As the air ascends the shaft,
decompression results in an increase in specific volume despite a decrease in temperature. The
latter is shown on the Ts diagram.
As the air returning from most underground facilities remains at a fairly constant temperature (the
'thermal flywheel'), upcast shafts are much less susceptible to variations in heat exchange than
downcast shafts. It is common to find that an upcast shaft is operating at near adiabatic
conditions. The path line 3 to 4 on the Ts diagram indicates a typical situation, descending from
P3, T3 to P4, T4 and diverted to the right of an isentrope by the effects of friction.
A thermodynamic analysis of the downcast shaft was given in Section 8.2.2. The corresponding
analysis for the upcast shaft follows the same logic and, indeed, the equations derived for the
downcast shaft can also be used, with suitable changes of subscripts, for the upcast shaft. For
that reason, details of the analysis will not be repeated. The reader may care to carry out a self
test by attempting to derive the following results and to prove the annotation of areas shown on
the Ts diagram.
For polytropic flow
F34
ln 3
u 32 u 42
P4
=
(Z 4 Z 3 )g + R (T3 T 4 )
2
T
ln 3
T4
q 34
J
kg
(8.18)
and
(Z 4 Z 3 )g
C p (T3 T 4 )
J
kg
(8.19)
Pa
(8.20)
Assuming constant (mean) density, mu ,

p 34
= mu (Z 3 Z 4 ) (P4 P3 )
Example.
The earlier example dealt with the downcast shaft of an 1100 m deep mine. Figures 8.6 (a), (b)
and (c) illustrate the conditions in the upcast shaft for the same mine and surveyed on the same
day. The plots of pressure and temperature against depth show little scatter and stable conditions
throughout the main length of the shaft. Again, measurements taken at the extremities of the
shaft have been ignored in generating the graphs.
The dry bulb temperature increases at a rate of 0.97 C/100 m indicating near adiabatic
conditions with essentially no evaporation or condensation. This is a confirmation of the
theoretical adiabatic lapse rate for dry bulb temperature in a dry shaft (ref. Section 15.3.1).
The adiabatic conditions are further verified by the wet bulb temperature lapse rate of 0.42
C/100 m (ref. Figure 15.8).
The PV diagram for the upcast shaft is shown on Figure 8.6(c). The polytropic curve fit gives the
PV relationship to be PV1.356 = constant. The polytropic index of 1.356 is close to the isentropic
value of 1.4 for dry air (ref. Section 3.4.3.). Comparing Figures 8.4 and 8.6(c) show that the
barometric pressure at the base of the upcast shaft was greater than that at the base of the
downcast shaft. This was because the main fans were sited underground at this mine in order to
mitigate noise at the surface.
17
P3
V dP
4
P4
V
T3
P3
P4
T4
4
3
V dP
F34
H 3 H4
q34
S3
S4
Figure 8.6. PV and Ts diagrams for an upcast shaft.
18
Upcast Shaft
Barometric Pressure
kPa
118
116
114
112
110
108
106
104
102
100
y = 0.0131x + 100.68
200
400
600
800
1000
1200
Depth m
Temperature deg. C
Figure 8.6(a) Measurements of barometric pressure in the upcast shaft with a best-fit straight line.
34
y = 0.0097x + 24.63
Dry bulb
Dry temperature
bulb temperature
29
24
y = 0.0042x + 19.575
19
Wet bulb temperature
14
0
200
400
600
800
1000
1200
Depth m
Figure 8.6(b). Increase in wet and dry bulb temperatures with respect to depth in the upcast shaft.
19
116
114
112
110
-1.3558
y = 81.447x
108
106
104
102
100
98
96
0.7
0.75
0.8
0.85
0.9
Specific Volume cu.m/kg
measured
Curve fit
PV1.3558 = 81.45
where P is in kPa
Figure 8.6(c). PV diagram for the upcast shaft.
20
8.3. THE COMPLETE MINE CYCLE

In the interests of clear explanation, it is convenient to investigate the phenomenon of natural
ventilation first, before going on to the more usual situation of a combination of fan and natural
ventilation.
8.3.1. Natural ventilation
Suppose that we have a vertical U tube with fluids of differing densities but the same height in the
two limbs. This is an unstable situation. The heavier fluid will displace the lighter and the fluids will
move. Motion will be maintained for as long as a difference exists between the mean densities of
the fluids filling each limb of the U tube. It would be difficult to keep this experiment going with two
completely different fluids. However, let us choose air as our fluid and apply heat at the base of
the U tube. If any slight perturbation then caused a movement of air so that one limb contained
warmer and, therefore, less dense air than the other, then the motion would accelerate until a
state of dynamic equilibrium was reached, dependent upon the rate of heat addition. This is the
process that causes natural ventilation in mines or any other flow system that involves heat
transfer and differences in elevation. It is the same phenomenon that causes smoke to rise up a
chimney or convective circulation in a closed cycle, and also explains the effectiveness of the old
shaft bottom furnaces described in Chapter 1.
In most cases, heat is added to the air in underground facilities. The air in an upcast shaft is then
warmer and less dense than in the downcast shaft. When the shaft bottoms are connected
through workings, this constitutes an enormous out-of-balance U tube. Airflow is promoted and
maintained at a rate that is governed by the difference between the mean densities of the air in
the two shafts and the depths of those shafts.
If we take the simple case of two shafts of equal depth (Z1 - Z2), one a downcast containing air of
mean density md, and the other an upcast of mean density mu, then the pressure at the base of
the downcast due to the column of air in the shaft will be md g(Z1 - Z2) (equation (2.8)) and the
pressure at the base of the upcast due to its column of air will be mu g(Z1 - Z2). Hence, the
pressure difference across the shaft bottoms available for promoting airflow through the workings
will be
NVP = md g (Z1 Z 2 ) mu g (Z1 Z 2 ) = g (Z1 Z 2 ) ( md mu )
Pa
(8.21)
where NVP = Natural ventilating pressure (Pa)

This simple analysis provides the traditional expression for natural ventilating pressure. However,
the following thermodynamic analysis indicates the limitations of equation (8.21) as well as
providing a fuller understanding of the mechanisms of natural ventilating effects.
Natural ventilation will occur whenever heat transfer occurs in the subsurface. If the rock is cooler
than the air then a natural ventilating pressure will apply in the reverse direction inhibiting the
airflow. In the case of a geographical region where the diurnal or seasonal ranges of surface air
temperature encompass the mean temperature of the strata in a naturally ventilated mine, then
reversals of airflow can occur. Because of such instabilities, very few medium size or large mines
are now ventilated purely by natural means.
The PV and Ts diagrams for a naturally ventilated mine are illustrated by combining Figures 8.2,
8.5 and 8.6 for the downcast shaft, level workings and upcast shaft respectively. The result is
shown on Figure 8.7 with the slight curvature of each of the process lines ignored for simplicity.
21
P2
P3
V dP
=
F
1 4
= NVE
P1
P2 P3
T3
T2
T4
P4
P1
Heat used
=
1 4
T
T1
Heat rejected to
atmosphere
=
q
1 4
S
Figure 8.7. PV and Ts diagrams for a naturally ventilated mine.
22
In the absence of a main surface exhaust fan, the airflow is emitted from the top of the upcast
shaft (station 4) at surface atmospheric pressure. We can now close the cycle back to station 1
(top of downcast shaft) by considering the process that takes place within the pressure sink of the
surface atmosphere. The exhaust air is emitted from the top of the upcast shaft with some finite
velocity. This causes turbulence within the atmosphere. In the majority of cases, the exhaust air is
at a temperature higher than that of the surface atmosphere. It cools at constant pressure,
rejecting heat to the atmosphere, contracts and follows the process lines 4 to 1 on both the PV
and Ts diagrams of Figure 8.7.
Just as we have combined the separate processes into a closed cycle on the PV and Ts
diagrams, so, also, can we combine the corresponding versions of the steady flow energy
equation:
2
Downcast:
u1 u 2
2
+ (Z1 Z 2 )g =
Workings:
u2 u3
2
2
Upcast:
u3 u 4
2
2
Atmosphere:
SUM:
u 4 u1
2
0
VdP + F
12
= (H 2 H 1 ) q12
(8.22)
1
3
+ (Z 2 Z 3 )g =
VdP + F
23
= (H 3 H 2 ) q 23
(8.23)
= (H 4 H 3 ) q 34
(8.24)
2
4
+ (Z 3 Z 4 )g =
VdP + F
34
3
1
+ (Z 4 Z 1 )g =
VdP + F
41
= (H 1 H 4 ) q 41
VdP + F =
12341
(8.25)
(8.26)
12341
Note that when we sum the individual terms, the kinetic energy, potential energy and enthalpy all
cancel out.
Let us examine some of the features of the summation equation (8.26). Recalling that each
VdP term is represented by the area to the left of the curve on the relevant PV
diagram, when we sum those areas algebraically, the cyclic integral VdP for the complete loop
individual
becomes the enclosed area, 1234, on the PV diagram of Figure 8.7. Furthermore, the
VdP areas for the workings and upcast shaft are negative (falling pressures). Inspection of the
PV diagram shows that the enclosed area is also negative, indicating that net flow work is done
by the air. This is the mechanical energy that produces and maintains motion of the air. It has, as
its source, an available part of the heat energy that is added to the air and is called the Natural
Ventilating Energy, NVE.
Further examination of the PV diagram shows that the NVE area depends primarily upon two
factors. First, the extent to which heat additions in the subsurface cause expansion of the specific
volume between stations 2 and 3, and secondly, the difference in barometric pressures caused
by the depths of the shafts, 1 to 2 and 3 to 4. The greater the heat additions or the depth of the
mine then the greater will be the NVE.
23
Equation (8.26) for the complete cycle gives
VdP =
J/kg
(8.27)
12341
where the symbol
denotes summation around the closed cycle. This shows that the
12341
natural flow work done by the air is utilized entirely against friction. The enclosed area, 1234, on
the PV diagram also represents the work that is done against friction. Unlike a manufactured heat
engine, a mine thermodynamic cycle produces no external work, although it would be perfectly
possible to achieve this by introducing a high volume - low pressure turbine into the airstream.
The mechanical energy thus produced would be at the expense of flow energy and the rate of
airflow would then diminish.
Let us take a closer look at equation (8.25) for the isobaric cooling of the exhaust air. We can
make three simplifications to this equation. First, if the tops of both shafts are at the same
elevation then Z4 = Z1. Secondly, we can assume that the surface atmosphere is at rest, u1 = 0
1
and, thirdly, as the free atmosphere is at constant pressure,
VdP = 0. Applying these

4
conditions gives
u4
2
= F41
(8.28)
J/kg
This confirms that the kinetic energy of the air issuing from the top of the upcast shaft is utilized
against the frictional effects of creating turbulence in the free atmosphere. Equation (8.25) also
quantifies the heat that is rejected to the surface atmosphere:
q 41 = (H 4 H 1 )
J/kg
(8.29)
or, if no evaporation or condensation occurs during the isobaric cooling,

q 41 =
C p (T 4 T1 )
J/kg
(8.30)
However, the summation equation (8.26) gives:
= 0
J/kg
(8.31)
12341
It follows that the total heat added in the subsurface,
q = q
12
+ q 23 + q 34 , is equal to the
1234
heat ultimately rejected to the surface atmosphere, q 41 , despite the fact that some of that added
heat has, temporarily, been converted to mechanical (kinetic) energy in order to create movement
of the air.
Turning to the Ts diagram, the heat area below each of the process lines 1-2, 2-3 and 3-4
represents the combination of added heat and frictional heat for the downcast shaft, workings and
upcast shaft respectively. It follows that the total area under lines 1-2-3-4 represents
F + q for the complete mine. However, the area under the isobar 4-1 represents the heat
1234
1234
24
rejected to the atmosphere, - q41, which has already been shown to equal
q . It follows that
1234
the 1-4 enclosed area 1234 on the Ts diagram represents the work done against friction in the
shafts and workings,
F . As the frictional loss within the atmosphere does not appear on the
1234
PV diagram, we have the revealing situation that areas 1234 on both diagrams represent
frictional effects in the subsurface. The PV area represents the work done against friction while
the Ts area shows that same work, now downgraded by friction, appearing as heat energy.
The area 41A on the Ts diagram lies above the atmospheric temperature line and, therefore,
represents rejected heat that, theoretically, is available energy. It could be used, for example, to
assist in the heating of surface facilities and can be significant in deep mines. However, that
energy is contained in a large volume of air, often contaminated by solid, liquid and gaseous
contaminants, and with a relatively small temperature differential with respect to the ambient
atmosphere. For these reasons, very few mines attempt to take advantage of the available part of
the reject heat.3
There are two main methods of quantifying the natural ventilating energy. First, if each of the
subsurface processes approximates to a polytropic law then the work done against friction in
each segment may be calculated from the form of equation (8.1).
Fi ,i +1 =
u i2 u i2,i +1
2
ln Pi +1
Pi
+ (Z i Z i ,i +1 )g R (T i +1 T i )
lnT i +1
Ti
J
kg
where subscript i takes the relevant value 1, 2 or 3. The sum of the resulting F terms then gives
the natural ventilation energy, NVE.
Secondly, if polytropic curve fitting shows significant deviations from observed data then each
segment can be further subdivided, or the enclosed NVE area on the PV diagram can be
determined graphically.
As most modern underground facilities are ventilated by a combination of fans and natural
ventilation, it is useful to be able to compare the ventilating potential of each of them. The
theoretically correct way of doing this is to express the mechanical work input from the fan in J/kg.
This is then directly comparable with the NVE. The fan work can be determined from
measurements of pressure, temperature and air velocity across the fan (Section 8.3.2.). However,
the more usual convention in practice is to quote fan duties in terms of fan pressure and volume
flow. A traditional and convenient device is to convert the natural ventilating energy, NVE, into a
corresponding natural ventilating pressure, NVP. It should be remembered, however, that unlike a
fan, an NVP is not located at a single location but is generated around a closed thermodynamic
cycle.
NVP = NVE x
Pa
(8.32)
Here, again, we have the same difficulty of choosing a value of air density. If frictional pressure
drops within individual branches have all been referred to standard density, 1.2 kg/m3, then it is
appropriate also to refer both natural and fan ventilating pressures to the same standard density.
This is, in effect, equivalent to comparing ventilating energies.
See section 18.4.2 for use of reject heat in cold climates.
25
Another method, although imprecise, is to sum equations (8.5, 8.17 and 8.20) that gave the
frictional pressure drops for the downcast shaft, workings and upcast shaft, respectively, on the
basis of the corresponding mean densities:
Downcast:
Workings:
Upcast:
SUM:
=
md (Z1 Z 2 )g (P2 P1 )
=
(P2 P3 )
= mu (Z 4 Z 3 )g (P4 P3 )
p12
p 23
p 34
p12 + p 23 + p 34 = (Z 1 Z 2 ) g ( md mu )
Pa
Pa
Pa
see (8.5)
see (8.17)
see (8.20)
Pa
(8.33)
In arriving at this summation, we have assumed that the depths of the two shafts are equal,
(Z1 - Z2) = (Z4 - Z3), and that the atmospheric pressure at the tops of both shafts is the same,
P1 = P4 . This is identical to the equation that we arrived at earlier (equation (8.21)) from a
consideration of the pressures at the bottoms of the two shafts and which were produced by their
corresponding columns of air. Calculation of natural ventilation pressure as
NVP = p12 + p 23 + p 34 = (Z 1 Z 2 ) g ( md mu )
Pa
(8.34)
is sometimes known as the mean density method. While equation (8.34) is simple to use and
much quicker than the two methods of determining natural ventilating effects introduced earlier, it
should not be regarded as an exact measure of the thermal energy that assists in the promotion of
airflow. The difficulty is that the summation of pressure drops is not referred to a single defined
value of air density. Each individual pressure drop is based on a different mean density. Their
summation cannot, therefore, be justified as a definitive measure of natural ventilation but remains
as a summation of measured frictional pressure drops. This can, in fact, be shown from the
differential form of the steady flow energy equation and ignoring the kinetic energy and fan terms
g dZ = VdP + dF
J/kg (see equation (3.25))
or, as V = 1/
J/m3 = Nm/m3 = N/m2 = Pa
g dZ = dP + dF
Integrating around a closed cycle dP = 0 gives
dZ
dF
J/m3 = Pa
The right side of this equation is the sum of the measured frictional pressure drops, p = F, while
the left side is the algebraic sum of the pressure heads exerted by the columns of air in the
system. It is, therefore, an exact form of equations (8.21) and (8.33).
8.3.2. Combined fan and natural ventilation
Main ventilating fans might be placed at a number of strategic sites in a subsurface system,
subject to any restrictions imposed by the governing legislation. However, to examine the
combination of fan and natural ventilation, we shall use the most common situation - that of a
main exhausting fan located at the top of the upcast shaft.
Figure 8.8 gives the corresponding PV and Ts diagrams. The difference between these and the
corresponding diagrams for the purely naturally ventilated mine is that station 4, the top of the
upcast shaft, is also the inlet to the fan and is at a sub-atmospheric pressure. On passing through
the fan from station 4 to station 5, the air is compressed back to atmospheric pressure.
26
P2
P3
natural
work
P1 C
P4
G 5
Fan flow work
P2 P3
T3
T5
T2
T4
Heat used
P1
P5
P4
Friction
in fan
T1
Heat rejected to
atmosphere
Strata heat
=
q
1 4
E
B
Figure 8.8. PV and Ts diagrams for a mine ventilated by both natural means
and a main exhaust fan.
27
Concentrating, for the moment, on the segments of the PV diagram that represent the
subsurface, the areas to the left of lines 1-2, 2-3 and 3-4 represent work done on or by the air as
it progresses through the downcast shaft, workings and upcast shaft respectively. The algebraic
summation of those work areas is the negative area 1234DC. This represents the net flow work
done by the air between stations 1 and 4. The area to the left of the fan compression curve 4-5
represents the flow work done on the air by the fan. In practice, both the upcast shaft and fan
processes are often near adiabatic, i.e. position 5 on the PV diagram is very close to position G.
Hence, we can divide the total work done by the air into two identifiable zones; area 123G
represents the natural ventilating energy and area C154D represents the fan work supplied to the
air while compressing it back to surface atmospheric pressure. The combination of the two
provides the total ventilating energy for the facility.
The process lines on the Ts diagram from station 1 to station 4 follow paths similar to those for
the naturally ventilated mine except that station 4 is now at a sub-atmospheric pressure. The fan
curve, 4 to 5 is angled slightly to the right of the ideal isentrope. Heat transfer is normally quite
small across the fan casing in terms of J/kg because of the high airspeed. Hence, the fan process
is essentially a frictional adiabatic. The derivation of the areas quantified on the Ts diagram
representing the fan work and associated frictional heat is given in Section 3.5.3.
The quantitative analysis of the cycle commences, once again, by stating the component steady
flow energy equations and summing them to determine the cycle equation.
2
Downcast:
u1 u 2
2
+ (Z1 Z 2 )g
Workings:
u4 u5
2
2
Atmosphere:
u 5 u1
2
(8.35)
VdP + F
23
= (H 3 H 2 ) q 23
(8.36)
VdP + F
34
= (H 4 H 3 ) q 34
(8.37)
VdP + F
45
= (H 5 H 4 ) q 45
(8.38)
= (H 1 H 5 ) q 51
(8.39)
2
4
+ (Z 3 Z 4 )g
3
5
+ (Z 4 Z 5 )g + W 45
4
1
+ (Z 5 Z1 )g
VdP + F
51
SUM:
= (H 2 H 1 ) q12
12
1
3
+ (Z 2 Z 3 )g
u3 u 4
2
Fan:
VdP + F
u2 u3
2
2
Upcast:
W45 =
VdP + F =
123451
(8.40)
123451
Consideration of equations (8.35) to (8.40) reveals a number of significant points:

(a) The work areas annotated on the PV diagram are confirmed by a rearrangement of
equation (8.40)
W45
123451
Work done
against friction =
Area 1234DC
VdP
J/kg
(8.41)
Work supplied
Natural flow
+ work done by air
by fan
Area 45CD
Area 123G
28
(b) The changes in kinetic energy and potential energy through the fan are both very small.
Ignoring these terms simplifies equation (8.38) to
5
W 45
VdP
+ F45
J/kg
(8.42)
The fan input produces flow work and overcomes frictional losses within the fan unit. The fan flow
work shows on the PV diagram. However, both flow work and the frictional losses appear on the
Ts diagram.
(c) The equation for isobaric cooling (dP = 0) in the atmosphere (equation 8.39) gives
2
u5
2
= F51
J/kg
(8.43)
assuming that the surface atmosphere is at rest (u1 = 0). The kinetic energy of the air exhausting
from the fan is dissipated against friction within the atmosphere. This is similar to the result
obtained for the naturally ventilated mine (equation (8.28)). However, it now becomes more
significant due to the higher velocity of the air issuing from the fan outlet, u5. The kinetic energy at
outlet is a direct loss of available energy from the system. It is, therefore, advantageous to
decelerate the airflow in order to reduce that loss. This is the reason that main fans are fitted with
expanding evasees at outlet.
(d) Summing equations (8.35) to (8.38) for the subsurface and the fan, and taking u1 = 0, gives
2
u5
2
+ W 45
= (H 5 H 1 )
12345
or
(H 5 H 1 ) = W 45 +
12345
u5
2
J/kg
(8.44)
showing that the increase in enthalpy between the inlet and outlet airflows is equal to the total
energy (work and heat) added, less the kinetic energy of discharge. However, if we inspect
equation (8.39) for cooling in the atmosphere, we have
2
(H 5 H 1 ) = q 51
u5
2
J/kg
(8.45)
J/kg
(8.46)
where - q51 is the heat rejected to the atmosphere.

Comparing equations (8.44) and (8.45) gives
q 51 = W 45 +
12345
This latter equation reveals the sad but unavoidable fact that all of the expensive power that we
supply to the fans of a ventilation system, plus all of the heat energy that is added to the air, are
ultimately rejected as waste heat to the surface atmosphere. It is possible to recover a little of the
exhaust heat in cold climates (Section 18.4.2.).
29
8.3.3. Case Study

The mine described in this case study was ventilated primarily by a main exhausting fan
connected to the top of the upcast shaft and passing an airflow of 127.4 m3/s at the fan inlet. Both
the downcast and the upcast shafts were 1219 m deep. Measurements of temperature and
pressure allowed Table 8.1 to be established. Throughout the analysis, it was assumed that the
air remained dry.
P
Location
Station Pressure Temperature

No.
kPa
V
Sp.
Vol.
m3/kg
VdP
101.59
283.15
0.8000
Tds
Entropy
F+q
J/kg
J/kgK
J/kg
(ii)
(iii)
(iv)
Density Flow work

3
kg/m
(i)
Top downcast
1.250
31.61
11 356
Bottom
downcast
Bottom upcast
116.49
295.06
0.7270
1.375
614
33.73
-1 923
113.99
322.84
0.8129
1.230
29 842
130.39
-12 621
Top upcast
99.22
310.93
0.8995
1.112
651
132.44
2 117
Fan outlet
101.59
314.05
0.8873
1.127
1018
135.70
0
Top downcast
101.59
283.15
0.8000
1.250
-31 055
31.61
Table 8.1. Thermodynamic parameters based on representative values of pressure and temperature
(i.e. taken far enough into the shafts to avoid end effects). In the compilation of the table, the
following equations were used. The equation reference numbers indicate the location of
their derivations.
(i)
V =
RT
P
and
1
V
where R = 287.04
ln(Pb / Pa )
ln (T b / Ta )
(ii)
VdP = R (T
(iii)
s = sa so = C p ln(Ta / To ) R ln(Pa / Po )
Ta )
J/kg
J/kg
(3.73)
J/kg
(3.48)
where Cp = 1005 J/kgK

Subscript a refers to the station number and subscript o refers to the pressure and temperature
datums of 100 kPa and 273.15 K respectively.
b
(iv)
ln(Pb / Pa )
Tds = (T b Ta ) C p R
ln(T b / Ta )
J/kg
(3.77)
30
120
2
115
110
105
1
100
95
0.6
0.65
0.7
0.75
0.8
0.85
0.9
0.95
Specific Volume V
v cu. m/kg
Figure 8.9(a) PV diagram for Case Study.
330
Absolute Temperature T (K)
320
G
310
300
290
friction
in fan
280
270
q1-4
260
D E
50
100
150
200
Specific Entropy s J/kg K

Figure 8.9(b). Ts diagram for Case Study
31
The PV and Ts diagrams were plotted as shown on Figures 8.9 (a) and (b) respectively. A great
deal of information can be extracted from the table and visualized on the diagrams. It is useful to
organize the analysis as follows:
General observations on the table of results
The
VdP column shows the large amount of flow work that is done on the air by gravity in the
downcast shaft. However, this is more than offset by the work done by the air against gravity as it
ascends the upcast shaft.
The addition of heat in the mine is concentrated in the workings, i.e. as the air progressed around
the circuit from the downcast to the upcast shaft bottom. This is shown by both the
Tds and
entropy columns. In this mine, the entropy increased continuously as the air passed through the
shafts and workings. However, this is not always the case. In a mine where heat lost by cooling in
any segment exceeds the heat generated by friction then that segment will exhibit a decrease in
entropy.
It is interesting to observe that the difference in barometric pressure across the shaft bottoms,
116.49 - 113.99 = 2.50
kPa
is greater than that across the main fan

101.59 - 99.22 = 2.37
kPa
This occurs quite frequently in deep mines and has often caused observers some puzzlement.
The phenomenon arises from natural ventilating effects in the shaft circuit.
Polytropic indices
For each segment of the circuit, a to b, the polytropic index, n, may be calculated from
ln(Pb / Pa )
n
=
n 1
ln(T b / Ta )
see (equation 3.72)
Segment
Downcast shaft
Workings
Upcast shaft
Fan
Surface atmosphere
Polytropic index n
1.4307
0.1943
1.3715
1.7330
0
The effects of friction and added heat cause a PV curve to swing to the right, i.e. finish with a
higher value of V than the corresponding isentropic process. This means that compression curves
such as those for the downcast shaft and fan will become steeper and adapt an increased value
of n, while decompression curves (workings and upcast shaft) will become shallower and take a
lower value of n. The greatly reduced value of n for the workings is simply a reflection of the heat
added in that segment. It should be noted, however, that the actual flow through the workings will
occur through a network of airways, each with its own series of thermodynamic processes. The
value of n calculated here for the workings is, therefore, a composite appraisal.
32
Fan performance
Table 8.1 shows that the fan produces 2 117 J/kg of useful flow work (area 45CD on the PV
diagram) and also adds 1 018 J/kg of heat (area 45ED on the Ts diagram) to the air. Hence, by
ignoring changes in kinetic energy and assuming no heat transfer across the fan casing, the
energy input from the fan,
5
W 45
VdP +
Tds
= 2117 + 1018 = 3135
J/kg
The same result is given by the change in enthalpy across the fan
W 45
= C p (T5 T 4 ) = 1005 (314.05 310.93) = 3136
While the input from the fan impeller is mechanical energy, the
J/kg
Tds part arises from the
frictional losses incurred almost immediately by the generation of turbulence. The volume flow of
127.4 m3/s was measured at the fan inlet where the air density was 1.112 kg/m3, giving a mass
flow of
M = 127.4 1.112 = 141.6
kg/s
Then useful flow work =
2117
141.6 = 300
1000
kJ kg
kJ
or kW
=
kg s
s
and fan frictional losses =
1018
141.6 = 144
1000
kW
The polytropic efficiency of the fan may now be expressed as

5
poly
VdP
4
(8.47)
VdP + Tds
or
300
100 = 67.6
300 + 144
per cent
Natural ventilating energy, NVE

The net flow work done by the air against friction in the shafts and workings is the area 1234DC
on the PV diagram. This is quantified by summing the tabulated
VdP values from station 1 to
station 4.
1234
VdP
= (11356 1923 12621) = 3188
1234
J
kg
or
3188
141.6 = 451
1000
kW
33
However, the fan adds flow work to the air of only 2117 J/kg, or 300 kW. Hence, the balance is
provided by natural ventilating energy,
NVE = 3188 + 2117 = 1071
J/kg
(area 123G on both the PV and Ts diagrams)

or
- 451 + 300 = - 151
kW
The negative signs show that this is work done by the air.
Natural ventilating pressure. NVP
The natural ventilating energy can be converted to a natural ventilating pressure by multiplying
by a specified value of air density. Choosing the standard value of 1.2 kg/m gives
NVP = NVE
=
1071
1.2 = 1.285
1000
kPa
For comparison with the fan pressure, we must refer that also to standard density, st
i
pi
st
p st
where subscript 'i' refers to fan inlet conditions, giving the standardized fan pressure as
1.112
2.37 = 2.196
1 .2
p st
kPa
The mean density method (Section 8.3.1) gives a value of

NVP = ( md mu ) g (Z1 Z 2 )
Pa
see (equation 8.21)
where
md
mu =
g
1.250 + 1.375
= 1.3125
2
kg/m 3
1.230 + 1.112
= 1.171
2
kg/m 3
9.81
m/s2
and
(Z1 Z 2 ) = 1219
giving
NVP = (1.3125 1.171) 9.81 1219

= 1692 Pa
or
1.692
kPa
This calculation illustrates that a natural ventilating pressure approximated by the mean density
method is not the same as that determined through NVE or a summation of F terms (Section
8.3.1.).
34
Heat additions
The total energy addition to the air is given by W 45 +
q and is equal to the increase in
12345
enthalpy across the system, (H5 H1) = Cp(T5 T1). Hence the heat transferred to the air is
T1 ) W 45
C p ( T5
12345
= 1005 (314.05 283.15) 3135

=
31055
3135
27 920
J/kg
or
27 920
141.6 = 3 953
1000
kW
The total heat rejected to the atmosphere is
q + W
q 51 =
45
= 27 920 + 3 135
12345
= 31 055
or
31 055
141.6 = 4 397
1000
kW
This agrees with the value in Table 8.1 for

The sum of the
J/kg [also given directly by Cp(T5 T1) ]
Tds within the surface atmosphere.
Tds values from station 1 to station 4 (in Table 8.1) gives the combined effect of
added heat and friction within the shafts and workings. The frictional component can be isolated
as
1234
Tds q
1234
= 31 107 27 920 = 3 187
or
3187
141.6 = 451
1000
J/kg
kW
This is in agreement with the value obtained earlier as
VdP
1234
Finally, the NVE may be checked by summing the complete
Tds column. This gives 1070 J/kg,
also in agreement with the value determined earlier.

This case study has illustrated the power of thermodynamic analysis. From five pairs of specified
temperatures and pressures, the following information has been elicited:
35
Pressure across main fan
2.37 kPa
(or 2.196 kPa at standard density)
Pressure across shaft bottoms
2.50 kPa
Fan flow work
300 kW
Fan losses
144 kW
Total energy input from fan (impeller work)
444 kW
Fan polytropic efficiency
67.6 per cent
Work done against friction in the shafts and workings:
451 kW
Natural ventilating energy
151 kW
Natural ventilating pressure at standard density
1.285 kPa
Heat added to the air in shafts and workings
3953 kW
Heat rejected to surface atmosphere
4397 kW
The case study also demonstrates the intrinsic coherence of the thermodynamic method;
alternative procedures have been employed to cross-check several of the results. Students of the
subject, having mastered the basic relationships, should be encouraged to experiment with the
analyses and to develop their own procedures. Each exercise will provide new and intriguing
insights into thermodynamic logic.
8.3.4. Inclined workings
So far, we have confined ourselves to a level connection between shaft bottoms and drawn a
single line on the thermodynamic diagrams to represent that connection. However, the PV and Ts
diagrams take on a different appearance when the connecting workings between shaft bottoms or
any two points on a main intake and main return, are inclined.
A steeply dipping circuit might be ventilated adequately with little or no ventilating pressure
applied across it. During a visit underground, one mine manager was quite disconcerted when a
ventilation survey team leader pointed out that opening an expensively constructed control door
in an outby cross-cut made no difference to the airflow in an adjoining dip circuit. Just as natural
ventilation plays a role in a shaft circuit, it also influences airflow in any other subsurface circuits
that involve changes in elevation.
Dip circuit.
The PV and TS diagrams for a dip circuit are shown on Figure 8.10. As air flows down the
descending intake airway(s) the pressure and temperature increase because of gravitational work
done on the air. Expansion occurs through the workings and as the air ascends the rising return
airways. If a survey team follows a route from station 1, an outby position in the main intake or
downcast shaft bottom, to station 2, a corresponding outby position in the main return or upcast
shaft bottom, then the resulting PV diagram will follow a loop from 1 to 2 whose shape is
determined by the three dimensional geometry of the circuit and the degree of heat transfer.
2
Figure 8.10 (a) shows that the
VdP flow work can be divided into two sections, the applied
1
ventilating energy (AVE) generated by the pressure differential applied across the ends of the
circuit, and the natural ventilating energy (NVE) generated by thermal effects within the district.
2
VdP = AVE + NVE = F12
J/kg
(8.48)
On traversing around the circuit from station 1 to station 2, the elementary VdP strips to the left of
2
the PV curve are dominant for falling pressure (- dP), giving a negative net value of
VdP .
1
36
NVE =
- VdP
Natural
ventilating
energy
Applied ventilating energy (AVE)

2
(a)
V
P1
NVE
(heat used)
Tds
P2
2
T
AVE
(b)
s
Figure 8.10. In dip workings, Natural Ventilating Energy (NVE) from added heat
assists in producing ventilation.
37
The analysis can also be conducted in terms of frictional pressure drops. Suppose we divide the
traverse route into a large number of small segments, then the frictional pressure drop along any
one segment, ignoring the small changes in kinetic energy, is given as
dp = dF = g dZ dP
In order to compare frictional pressure drops in differing segments, we refer the pressure drops to
standard density st
dpst =
st
dP
dp = st dF = st g dZ st
Integrating around the traverse route gives

2
dpst = pst = st g dZ st
dP
But V = 1/ and, if stations 1 and 2 are at the same level,
dZ = 0 , giving
1
p st
= st
VdP
1

p st
= st ( AVE + NVE )
Pa
(8.49)
p st
= AVP + NVP
Pa
(8.50)
or
where AVP = Applied ventilating pressure across the circuit (Pa)

and
NVP = natural ventilating pressure generated within the circuit, (Pa)
both being referred to standard density.
Equation (8.50) explains how a dip circuit may be ventilated by natural effects even when the
applied ventilation pressure is zero.
The NVP for the circuit may be determined by measuring the NVE area on the PV diagram,
NVP = NVE x st
or from the summation of frictional pressure drops, pst, measured during a survey, each referred
to standard density.
NVP
pst - AVP
Pa
(8.51)
The total area under the traverse loop 1 to 2 on the Ts diagram of Figure 8.10 (b) represents the
sum of frictional heat and added heat, F12 + q12. However, provided the stations 1 and 2 are at the
same level and ignoring changes in kinetic energy, the heat taken out of the district by the airstream
is (H2 H1), or Cp (T2 - T1) for dry conditions. This is identified as the heat rejected area on the Ts
diagram.
The remaining areas represent
F12 = AVE + NVE
J/kg
(8.52)
38
Applied ventilating energy (AVE)

Natural
ventilating
energy
(NVE)
(a)
P2
P1
2
1
NVE
AVE
(b)
s
Figure 8.11. In rise workings, Natural Ventilating Energy (NVE) from added
heat impedes ventilation.
39
Rise Circuit.
The PV and Ts diagrams for a rise circuit are illustrated on Figure 8.11. In this case the NVE has
an opposite sign to the AVE. Hence natural ventilating effects oppose the applied ventilating
energy. The analysis follows the same logic as for a dip circuit leading to the following results:
2
VdP = AVE NVE = F12
J/kg
(8.53)
Pa
(8.54)
and
pst
AVP - NVP
where, again, the applied and natural ventilating pressures are referred to standard density.
These equations show that if the natural ventilating effects become sufficiently high then they will
balance the applied ventilating energy or pressure. The flow work that can be done against
friction, F12, and the effective ventilating pressure, pst, then both become zero and airflow ceases.
8.3.5. The effect of moisture
In order to highlight underlying principles, this chapter has concentrated on the flow of dry air
throughout the system. This is seldom the case in practice. The air entering any facility from the
free atmosphere invariably contains water vapour. If this is maintained throughout the system
without further evaporation or condensation then the dry air thermodynamic relationships
developed in this chapter will continue to give satisfactory results. However, for more precise
results or if free water exists in the subsurface, as it often does, then evaporation and
condensation processes can have a major impact on the dry bulb temperature (see, for example,
Figure 8.3(b) ). In particular, the enthalpy change (H2 - H1) is no longer given simply by Cp(T2 - T1).
Equation (14.40), reproduced below, must be employed.
In order to take account of variations in moisture content the equations given below may be
employed. These equations are derived in Chapter 14. The parameters that vary with respect to
moisture content require to be redefined in terms of X kg of water vapour associated with each 1
kg of "dry" air, [i.e. in (1 + X) kg of air/vapour mixture]. Subscript m is used to denote moist air.
Gas constant:
Rm =
(287.04 + 461.5 X )
(1+ X )
(1005 + 1884 X ) J
(1+ X )
kg K
Specific heat at
constant pressure:
C pm =
Specific volume:
Vm =
(287.04 + 461.5 X ) T
(1 + X )
P
Density:
m =
1
Vm
J
kg K
m3/kg of moist air [see equation (14.18)]
kg moist air/m3 [see equation (14.21)]
40
An apparent specific volume and density based on 1 kg of dry air are also useful:
Vm (apparent ) = 461.5
m (apparent ) =
Enthalpy
T
(0.622 + X )
P
1
Vm (apparent )
H = 1005 t d + X [ 4187 t w + L + 1884(t d t w )]
m3/kg dry air
kg dry air/m3
J/kg dry air [see equation (14.40)]
where tw and td are wet and dry bulb temperatures respectively (degrees Celsius)
and
L = Latent heat of evaporation at wet bulb temperature
= (2502.5 2.386 t w ) 1000
J/kg
The reference numbers of these equations indicate the location of their derivation in Chapter 14.
Appendix A8.2 gives a more general analysis of the effects of moisture on steady flow
relationships.
Bibliography
Professor F. Baden Hinsley of Nottingham University, England, was the primary
architect of the development of mine ventilation thermodynamics. He wrote many papers
on the subject. Only the first of his major papers is listed here.
Hinsley, F.B. (1943). Airflow in Mines: A Thermodynamic Analysis. Proc. S. Wales Inst. of
Engineers, Vol LIX, No. 2.
Hall, C.J. (1981). Mine Ventilation Engineering. Published by the Society of Mining Engineers of
AIME.
Hemp, R. and Whillier A. (1982). Environmental Engineering in South African Mines. Chapters 2
and 16. Published by the Mine Ventilation Society of South Africa.
McPherson, M.J. (1967). Mine Ventilation Engineering. The Entropy approach. University of
Nottingham Mining Magazine, Vol. 19.
McPherson, M.J. and Robinson, G. (1980). Barometric Survey of Shafts at Boulby Mine,
Cleveland Potash, Ltd., Cleveland. Transactions of the Institution of Mining and Metallurgy,
Volume 89, 1980, ppA18 A28.
41
Appendix A8.1
Survey data from the survey of the shafts used in the examples given in Sections 8.2.2 and 8.2.4.
The examples given in Sections 8.2.2 and 8.2.4 of the text refer to the Boulby Potash Mine which
has the deepest mine shafts in the United Kingdom. It is only rarely that detailed ventilation
survey observations can be made throughout the lengths of major vertical shafts. The following
tables give the observations and corresponding thermodynamic parameters obtained from that
survey. They provide the basis for a useful case study by students and researchers in
thermodynamic processes in mines.
Observations in Downcast Shaft

True Moisture Relative Enthalpy Sigma
Depth Absolute Temperature,
heat
C
specific content humidity
Static
volume
pressure
m
kPa
Wet
Dry
m3/kg g/kg of
%
kJ/kg kJ /kg
bulb
bulb
dry air
0
100.770 15.78 17.17 0.8326
10.7
86.9
44.3
43.6
4.0 100.797 15.78 17.78 0.8340
10.4
81.7
44.3
43.6
10.7 100.831 15.67 17.22 0.8322
10.5
85.5
44.0
43.3
109.0 101.819 16.00 17.89 0.8260
10.5
82.6
44.6
43.9
207.4 103.016 16.39 18.61 0.8184
10.5
80.0
45.4
44.7
305.7 104.193 16.78 19.56 0.8118
10.5
75.6
46.2
45.5
404.1 105.391 17.33 20.28 0.8046
10.7
74.7
47.5
46.7
502.4 106.575 18.06 21.11 0.7981
11.0
74.3
49.3
48.4
600.8 107.769 18.22 21.94 0.7914
10.8
69.6
49.4
48.6
699.1 109.179 18.44 22.78 0.7832
10.5
65.5
49.6
48.8
797.5 110.349 19.39 23.44 0.7770
11.2
68.0
52.2
51.3
895.8 111.523 19.44 24.00 0.7701
10.9
64.7
52.0
51.1
994.2 112.727 19.72 24.72 0.7637
10.8
62.1
52.5
51.6
1092.5 113.793 20.28 25.44 0.7585
11.1
61.5
53.9
52.9
1108.5 114.070 20.50 25.83 0.7577
11.2
60.6
54.5
53.5
True
density
Comments
kg/m3
1.201
1.199
1.201
1.211
1.222
1.232
1.243
1.253
1.264
1.277
1.287
1.298
1.309
1.318
1.320
Surface
Above -30 level
Below -30 level
Manset
42
Observations in Upcast Shaft

True Moisture Relative Enthalpy Sigma True Comments
Depth Absolute Temperature,
heat density
C
specific content humidity
Static
volume
pressure
m
kPa
Wet
Dry
m3/kg g/kg of
%
kJ/kg kJ/kg kg/m3
bulb
bulb
dry air
1110.6 115.321 24.17
32.89 0.7680
13.0
47.3
66.4
65.1
1.302 Manset
1105.5 115.281 24.17
33.89 0.7706
12.6
43.3
66.3
65.1
1.298 Below ventset
1092.2 115.055 24.00
35.00 0.7746
11.9
38.5
65.8
64.6
1.291 Above ventset
1000.7 113.901 23.78
34.44 0.7810
12.0
39.6
65.4
64.2
1.280
909.2 112.521 23.33
33.33 0.7878
12.0
41.7
64.3
63.2
1.269
817.7 111.331 23.22
32.50 0.7942
12.4
44.6
64.4
63.2
1.259
726.2 110.147 22.72
31.67 0.8005
12.2
45.5
63.1
62.0
1.249
634.7 108.970 22.33
30.83 0.8069
12.2
47.2
62.2
61.1
1.239
543.2 107.718 21.78
30.00 0.8139
12.0
48.0
60.8
59.7
1.229 Passed cage
451.7 106.538 21.56
29.00 0.8204
12.3
51.5
60.5
59.4
1.219
360.2 105.357 21.11
28.22 0.8274
12.2
52.8
59.4
58.3
1.209
268.7 104.187 20.67
27.33 0.8342
12.1
54.7
58.3
57.3
1.199
177.2 103.013 20.28
26.22 0.8406
12.2
58.3
57.5
56.4
1.190
85.7
101.853 20.00
25.56 0.8483
12.3
60.2
57.0
55.9
1.179
10.1
100.939 19.50
24.50 0.8529
12.2
63.1
55.7
54.7
1.172 Below -30 level
4.9
100.922 19.33
23.89 0.8514
12.2
65.6
55.1
54.1
1.175 Above -30 level
100.874 19.44
24.17 0.8526
12.3
64.7
55.5
54.5
1.173 Surface
43
Appendix A8.2
A general steady flow analysis for airstreams that carry water vapour and liquid
water droplets.
The differential form of the steady flow energy equation (3.25) for 1 kg of air is:
u du g dZ + dW
= VdP + dF = dH dq
J/kg
(3.25)
Now consider an incremental length of airway within which every kg of air has associated with it X
kg of water vapour and w kg of liquid water droplets, each of these increasing by dX and dw
respectively through the increment. The vapour moves at the same velocity as the air, u.
However, the droplets may move at a different velocity, uw, and perhaps even in the opposite
direction to the air in the case of an upcast shaft. The terms in the steady flow energy equation
then become:-
Kinetic energy (J/kg dry air, i.e. Joules of energy associated with each kilogram of dry air)
u du :
: 1 kg of dry air :
X u du
: X kg of water vapour :
w u w du w
: w kg of droplets
u2
dX
2
: energy to accelerate dX of water vapour to u m/s :
uw
dw
2
: energy to accelerate dw of liquid to uw m/s
Potential energy (J/kg dry air)

g dZ
: 1 kg of dry air
X g dZ
: X kg of water vapour
w g dZ
: w kg of liquid water
Fan energy (J/kg dry air)

The energy input terms are now based upon 1 kg of dry air. The fan energy input, therefore,
remains unchanged at dW but with redefined units of J/kg dry air.
Flow work (J/kg dry air)
Provided that the specific volume is the volume of the mixture that contains 1 kg dry air,
Vm (apparent), then the flow work stays at V dP J/kg dry air.
Work done against friction (J/kg dry air)
Again this remains simply as F but defined in units of J/kg dry air.
44
Enthalpy (J/kg dry air)

dH = C p dT
: 1 kg of dry air
: X kg of water vapour of enthalpy Hx
where H x = C w t w + L + C pv (t d t w )
X dH x
Cw = specific heat of water

Cpv = specific heat of water vapour
w dH w
: w kg of liquid water of enthalpy Hw,

where Hw = Cw tw, for droplets at wet bulb temperature
H x dX
: energy content of dX kg of added vapour
H w dw
: energy content of dw kg of added droplets
Added heat (J/kg dry air)

This term remains at q defined in terms of J/kg dry air.
All of these terms may be collated in the form of equation (3.25) i.e. a three part steady flow
energy equation. In practice, the kinetic energy terms are normally very small and may be
neglected. The steady flow energy equation for airflow of varying moisture content then becomes
-[ g dZ + Xg dZ + wg dZ ] + dW = VdP + dF = [ dH + X dHx + w dHw + Hx dx + Hw dw ] - dq

J/kg dry air (8.55)
This equation is cumbersome for manual use. However, it provides a basis for computer
programs developed to simulate airflow processes involving phase changes of water. The
enthalpy terms encompass the energy content of the air, water vapour, liquid water and any
additions of vapour and liquid.
If we apply the mechanical energy terms around a closed cycle dZ = 0 then we obtain
J
W VdP =
F + g X dZ + w dZ
(8.56)
kg
total
Comparing this to the corresponding equation for dry conditions (equation 8.41), we see that the
fan and natural ventilating energies are now responsible not only for providing the necessary
work against friction but also for removing water vapour and airborne water droplets from the
mine.
45
Chapter 9. Ventilation Planning

9.1. SYSTEMS ANALYSIS OF THE PLANNING PROCEDURE ..........................1
9.2. ESTABLISHMENT OF THE BASIC NETWORK ............................................3
9.2.1. New mines or other subsurface facilities ........................................................................... 3
9.2.2. Existing mines.................................................................................................................... 3
9.2.3. Correlation study................................................................................................................ 5
9.3. AIRFLOW REQUIREMENTS.........................................................................6

9.3.1. Strata gas........................................................................................................................... 7
9.3.2. Diesel exhaust fumes......................................................................................................... 7
9.3.3. Dust.................................................................................................................................... 8
9.3.4. Heat.................................................................................................................................... 9
9.3.5. Workshops and other ancillary areas .............................................................................. 12
9.3.6. Air velocity limits .............................................................................................................. 13
9.4. PLANNING EXERCISES AND TIME PHASES............................................13

9.4.1. Network planning exercises ............................................................................................. 13
9.4.2. Time phases .................................................................................................................... 14
9.4.3. Selection of main fans...................................................................................................... 15
9.4.4. Optimization of airflow systems ....................................................................................... 15
9.4.5. Short term planning and updating the basic network ...................................................... 15
9 .5. VENTILATION ECONOMICS AND AIRWAY SIZING.................................16

9.5.1. Interest payments ............................................................................................................ 16
9.5.2. Time value of money, present value................................................................................ 18
9.5.2.1. Present value of a lump sum..................................................................................... 18
9.5.2.2. Present value of regular payments. .......................................................................... 19
9.5.3. Equivalent annual cost..................................................................................................... 21
9.5.4. Ventilation operating costs............................................................................................... 22
9.5.5. Optimum size of airway.................................................................................................... 23
9.5.5.1. Capital cost function .................................................................................................. 23
9.5.5.2. Operating cost function ............................................................................................. 24
9.5.5.3. Case Study................................................................................................................ 25
Task 1: Establish the capital cost function:......................................................................... 25
Task 2: Establish the operating cost function: .................................................................... 26
Task 3: Establish the total cost function: ............................................................................ 26
Task 4: Determine the optimum diameter: ......................................................................... 26
9.5.6. Incorporation of shaft design into network planning exercises. ....................................... 27
9.6. TRADITIONAL METHOD OF VENTILATION PLANNING ...........................29

Bibliography ........................................................................................................30
9.1. SYSTEMS ANALYSIS OF THE PLANNING PROCEDURE

In Section 1.3.3, we emphasized the importance of integrating ventilation planning with
production objectives and overall mine design during the early stages of planning a new mine or
other subsurface facility. The team approach should continue for ongoing forward planning
throughout the active life of the mine. Compromises or alternatives should be sought which
satisfy conflicting demands. For example, the ventilation planner may require a major airway that
is too large from the viewpoint of the rock mechanics engineer. An alternative might be to drive
two smaller airways but large enough for any equipment that is to move through them. Regular
9- 1
liaison between engineering departments and exchanges during inter-disciplinary planning

meetings can avoid the need for expensive re-designs and should promote an optimized layout.
Figure 9.1 illustrates an organized system of ventilation planning for an underground mine. The
procedure assumes the availability of computer assistance including ventilation simulation
software, and eliminates most of the manual techniques and intuitive estimates of older planning
methodologies (Section 9.6). The initial step is to establish a data base in a Basic Network File.
For an existing facility this requires information gained from ventilation surveys. The latter are
described in detail in Chapter 6. For a completely new mine, initial layouts should be discussed at
the cross- disciplinary planning meetings. It is usual, in this case, that several alternative layouts
are required to be investigated by the ventilation design team. The data base for a new mine is
established from initial estimates of required airway geometry, mining method, roughness of
airway surfaces and network layout. Such data will be revised and refined as the design
progresses.
In this chapter, we shall discuss the major facets of the planning procedure illustrated on Figure 9.1.
Ventilation
Surveys
Establish
BASIC
NETWORK
amend
basic
network
no
VNET
SIMULATION
good correlation?
yes
PLANNING
EXERCISES
Predicted
distributions of
airflows, pressures,
gas concentrations,
operating costs
Establish
required
airflows,
allowable
velocities and
permitted gas
concentrations
Ventilation criteria
satisfied?
Update network
no
yes
Sizes of
airways/fans
optimized?
no
Optimize
major
fans/airways
yes
Cycle through climate
simulations if required
proceed to
next time
phase
Figure 9.1. Systems analysis of subsurface ventilation planning.
9- 2
9.2. ESTABLISHMENT OF THE BASIC NETWORK

9.2.1. New mines or other subsurface facilities
It will be recalled from Chapter 7 that a ventilation network consists of a line schematic, each
branch representing one or more airways or leakage paths. In general, branches must be of
known resistance or designated to pass a specified airflow. For a planned but yet unconstructed
facility there are, of course, no airways or ventilation infrastructure that can be surveyed in order
to establish actual values of resistance. Recourse must then be made to the methods of
estimating airway resistance described in Section 5.4, incorporating airway geometry, types of
lining (friction factors) and shock losses. Experience and data from other mines that utilize
comparable layouts and methods of airway drivage, and operating in similar geological conditions
can be most helpful.
A variety of alternative plans will often require investigation during the initial design of a new
mine. Ventilation network analyses should be carried out in order to determine the efficiency and
effectiveness of each layout. The results of those analyses may then be considered, together with
the other aspects of mine design, at the general planning meetings.
9.2.2. Existing mines
The routes of surveys in the main ventilation infrastructure of an existing mine provide a skeleton
network, often referred to as a ventilation schematic. Additional branches can then be added or
incorporated as equivalent resistances to represent airways or leakage paths that were not
included in the primary surveys. As illustrated in the computed network example of Section 7.4.6,
branch resistances may be input to a network simulation program in several ways. This allows
considerable flexibility in assembling a representative basic network, commencing from surveyed
routes. Essentially, four methods are used to indicate resistance values in the first compilation of
a basic network file or to update a basic network that has previously been established:
1. For surveyed airways, the measured values of frictional pressure drop, p (Pa), and
corresponding airflow, Q (m3/s), may be input to the basic network file. The computer can
then calculate the resistance from the square law.
p
Ns 2
R =
Q2
m8
2. There are situations in which it is difficult or impossible to determine resistance values for
existing airways. If, for example, the airflow in an unrestricted branch is low at the time of
measurement then so, also, will be the frictional pressure drop. Low frictional pressure drops will
also occur in airways of large cross-sectional area even at substantial airflows. While any
pressure difference shown by the gauge should be logged, those lower than 5 Pa may yield
resistances of doubtful accuracy. Modern diaphragm gauges will indicate to within 0.5 Pa. At 5
Pa, the corresponding error is 0.1 or 10 per cent. Airflow measurements taken by approved
methods (Chapter 6) will normally be accurate to within 5%. Compounding these possible errors
gives a maximum corresponding error in airway resistance of
dp
dR
dQ
=
2
R
p
Q
(differentiation of the square law R = p/Q2)
giving
dR/R
(9.1)
= 0.1 (2 x 0.05)
= 0.2 or 20 per cent for the accuracy of the resistance of an airway
that has a frictional pressure drop of 5 Pa.
9- 3
If, because of low values, the airflow and/or pressure drop cannot be measured to the required
accuracies, the airway resistance should be calculated from the Atkinson equation
R = kL
per
A
1. 2
Ns2/m8
where k = Atkinson friction factor at standard density (kg/m3)

L = length of airway (m)
per = perimeter (m)
A = cross sectional area (m2)
and
Representative values of friction factor, k, should, preferentially, be determined from similar
airways that have been surveyed in that same mine. The density correction,
/1.2 is then often ignored. Otherwise, values of k may be estimated using one of the
methods discussed in Section 5.3. In all cases, care should be taken to allow for shock losses
and unrepresentative obstructions (Appendix A5, Chapter 5).
In many situations within a given mine, groups of airways are driven to the same dimensions
and with a similar friction factor. Typical values of resistance per metre length can be
established from measurements taken in, say, intakes, returns and conveyor routes. This
facilitates the addition of new or unsurveyed branches to the network. However, here again
care should be taken not to overlook obstructions or any other cause of shock losses.
3. It is often difficult to measure the resistances of doors, stoppings or seals directly, either
because of low airflows or inaccessibility (Kissell, 1978). In practice, the resistances of doors
or stoppings in mines vary from several thousand to, literally, 1 or 2 Ns2/m8. The reason for
this wide range is the highly non-linear relationship between area available for flow and
resistance
R 1/ d 5
where d = hydraulic mean diameter of opening ( 4A/per ).

Hence, the appearance of fractures or apertures in, or around, a previously tight stopping will
result in dramatic reductions in its resistance.
The following ranges are suggested for planning purposes and for cross-sectional areas up to
25 m2.
Single doors
Stoppings
Seals
10 to 50 Ns2/m8 (typical value, 25 Ns2/m8)

50 to 2000 Ns2/m8 (typical value 500 Ns2/m8
1000 to 5000 Ns2/m8
4. The resistances of leakage paths including worked out areas and caved zones are the most
difficult to assess for a network schematic. Leakage airflows can often be measured or
estimated from survey results. These may be entered as "fixed quantity" leakage airflows,
leaving the computer to evaluate the corresponding resistances. This 'inversion' method
remains valid while the number of fixed quantity branches is sufficiently small for unique
values of resistance to be calculated. A network that is over-restricted by too many fixed
quantity branches will produce a warning message from the VNET simulation package
(Section 7.4.6). Care should be taken to remove "fixed quantity" airways and to replace them
with branches set at the corresponding computed resistances prior to progressing with
planning exercises.
9- 4
9.2.3. Correlation study

Figure 9.1 shows the data pertaining to the initial network schematic stored in the Basic Network
File within computer memory. In order to ensure that the basic network is a true representation of
the mine as it stood at the time of the survey, and before that basic network is used to launch
planning exercises, it is important that it is subjected to verification through a correlation study.
This involves running a ventilation network simulation package (Section 7.4) on the data
contained in the basic network file and comparing the computed airflows with those that were
measured during the surveys. If all airways had been surveyed instantaneously with perfect
observers and perfect instruments, and if no errors had been made in compiling the basic
network, then the computed and measured airflows would show perfect agreement. This utopian
situation is never attained in practice. Despite the internal consistency checks on airflow and
pressure drop measurements made during a survey (Section 6.4.3) there will, inevitably, remain
some residual errors from observational and instrumental sources.
A larger potential source of difference between computed and measured airflows is the fact that a
VNET airflow distribution is mathematically balanced to a close tolerance throughout the network
and represents a steady-state 'snapshot' of the ventilation system. On the other hand, the
surveys may have taken several weeks or even months to complete. In a large active mine,
updating survey data should be a routine and continuous activity of the mine ventilation
department. During the time taken for a survey the airflow distribution may have changed due to
variations in natural ventilating pressures or the resistances offered by the ever-changing work
areas.
Lastly, errors made in transcribing survey data into the basic network or in key-punching the
information into the basic network file will, again, result in disagreement between computed and
measured airflows.
The correlation itself is normally carried out by a subsidiary program that lists the computed and
observed airflows together with their actual and percentage differences. Leakage airflows of less
than 3 m3/s may be ignored in this comparison as large percentage errors in such low airflows will
usually have little influence on the overall accuracy of the network.
The overall network correlation is quantified as
Absolute value of (differences between computed and actual airflows for surveyed branches) x 100
Absolute values of (measured airflows for surveyed branches)
(9.3)
A correlation is accepted as being satisfactory provided that:
(i) no significant airflow branch shows a difference of more than 10 per cent between computed
and measured airflows; and
(ii) the overall correlation is also within 10 per cent.1
However, because of the highly variable conditions that exist along mine traverses, the survey
team will often be able to weight certain measurements as being more, or less, reliable than
others. Such pragmatic considerations may be taken into account during the correlation study.
If the initial results of a correlation study indicate unacceptable deviations between computed and
measured airflows then the basic network must be improved before it can be used as a basis for
future planning exercises. The transcription of survey data into the basic network file should be
checked. In the case of computed airflows being consistently higher or lower than the
corresponding measurements then it is possible that an error has been made in the pressure
1
Experienced teams can often achieve overall correlation within 5 percent.
9- 5
ascribed to a main fan or may indicate that the resistances of primary airways such as shafts
should be verified. Additional or check measurements of surveyed loops may be required.
It is fairly common that disagreements will occur in localized areas of the network. Again, this may
indicate that the extent of the surveys had been insufficient to give an adequate representation of
the mine network and that further measurements are necessary. More often, however, the
problem can be resolved by adjustment of the estimated resistances or fixed airflows allocated to
unsurveyed flowpaths. Provided that the main infrastructure of the surveyed network is sound
then amendments to those estimated values will direct the simulated network towards an
improved correlation. However, if measurements made in the surveyed loops are significantly out
of balance then no amount of adjustment to subsidiary airways will produce a well-correlated
basic network.
9.3. AIRFLOW REQUIREMENTS

The estimation of airflows required within the work areas of a mine ventilation network is the most
empirical aspect of modern ventilation planning. The majority of such assessments remains
based on local experience of gas emissions, dust, or heat load and is still often quoted in the
somewhat irrational terms of m3/s per ton of mineral output, particularly for non-coal mines.
Corrections can be applied for variations in the age of the mine, the extent of old workings,
distances from shaft bottoms, depth and rates of production. However, as in all empirical
techniques, the method remains valid only whilst the proposed mining methods, machinery, and
geological conditions remain similar to those from which the empirical data were evolved.
Attempts to extrapolate beyond those circumstances may lead to serious errors in determining
required airflows. Fortunately, simulation techniques are available to assist in assessing airflow
requirements for both gassy and hot mines.
The characteristics of strata gas emissions, heat flow and dust production are discussed in
Chapters 12, 15 and 20, respectively. In this section, we shall confine ourselves to an
examination of the methods used to determine the airflows required to deal with given emission
rates of airborne pollutants.
The overall necessity is that in all places where personnel are required to work or travel, airflows
must be provided in such quantities that will safeguard safety and health, comply with statutory
requirements, and that will also furnish reasonable comfort (Section 1.3.1).
The quantity of air required for the purposes of respiration of personnel is governed primarily by
the concentration of exhaled carbon dioxide that can be allowed in the mine atmosphere rather
than the consumption of oxygen. For vigorous manual work, this demands about 0.01 m3/s of air
for each person and is negligible compared with the quantities of air needed to dilute the other
gases, dust, heat and humidity that are emitted into the subsurface atmosphere.
It is a legal requirement that air volume flows must meet the governing state or national mining
laws. Such legislation will normally define minimum airflows that should be provided at specified
times and places in addition to threshold limit values for airborne pollutants or the psychrometric
condition of the air. The ventilation planner must be familiar with the relevant legislation. In
countries where mining law is absent or does not cover the particular circumstances of the project
then it is prudent to follow well established regulations of a major mining country.
9- 6
9.3.1. Strata gas

The various methods used to predict the emission rates of methane are described in Chapter 12.
Whichever technique is employed, the final calculation of airflow requirement is
100 E g
Q =
Cg
m3/s
(9.4)
where Q = required airflow (m3/s)

Eg = gas emission rate (m3/s)
Cg = general body concentration to which gas is to be diluted (percentage by volume).
The value of Cg is often taken to be one half of the concentration at which the law requires action
to be taken.
Example
It has been predicted that during a 7 hour working shift, 2500 m3 of methane will be emitted into a
working face in a coal mine. If electrical power must be switched off at a methane concentration
of 1 per cent, determine a recommended airflow for the face.
Solution
The average rate of gas emission during the working shift is
Eg
2500
= 0.0992
7 60 60
m 3 /s
Let us take the allowable concentration for design purposes to be one half the legal limit. Then
Cg = 0.5 per cent.
Equation (9.4) gives the required airflow as
Q =
100 0.0992
= 19.84
0 .5
say 20 m3/s.
9.3.2. Diesel exhaust fumes

There are wide variations in the manner in which different countries calculate the ventilation
requirements of mines in which diesels are used. The basic stipulation is that there should be
sufficient ventilation to dilute exhaust gases and particulates to below each of their respective
threshold limit values. One technique, based on engine tests, is to calculate the airflow required
to dilute the mass emission of each pollutant to one half the corresponding TLV (threshold limit
value). The maximum of those calculated airflows is then deemed to be the required air quantity.
Some countries require analyses of the raw exhaust gases in addition to general body air
sampling downstream from diesel equipment. Distinctions may be drawn between short-term
exposure and time-weighted averages over an 8 hour shift (Section 11.2.1.).
In addition to exhaust gases, national enforcement agencies may require that other factors be
taken into account in the determination of airflow requirements for diesel equipment. These
include rated engine power (kW or bhp), number of personnel in the mine or area of the mine,
rate of mineral production, engine tests, number of vehicles in the ventilation split and forced
dilution of the exhaust gases before emission into the general airstream.
9- 7
There is, as yet, no method of determining airflow requirements for diesels that will guarantee
compliance with legislation in all countries. Here again, regulations pertinent to the location of the
project must be perused for specific installations.
The generic criterion that is used most widely for initial estimates of required airflow is based on
rated output power of the diesel equipment. For design purposes, many ventilation planners
employ 6 to 8 m3/s of airflow over the machine for each 100 kW of rated diesel power, all
equipment being cumulative in any one air split. However, it should be borne in mind that the
actual magnitude and toxicity of exhaust gases depend upon the type of engine, conditions of
operation and quality of maintenance in addition to rated mechanical power. It is for this reason
that mining law refers to gas concentrations rather than power of the diesel equipment. Some
manufacturers and government agencies give recommended airflows for specific diesel engines.
In hot mines, it may be the heat produced by diesel equipment that sets a limit on its use.
9.3.3. Dust
Pneumoconiosis has been one of the greatest problems of occupational health in the mining
industries of the world. Concentrated and long-term research efforts have led to greatly improved
understanding of the physiological effects of dusts, methods of sampling and analysis, mandatory
standards and dust control measures (Chapters 19 and 20).
There are many techniques of reducing dust concentrations in mines, ranging from water infusion
of the solid mineral through to dust suppression by water sprays and air filtration systems.
Nevertheless, with current mining methods, it is inevitable that dust particles will be dispersed into
the air at all places wherever rock fragmentation or comminution occurs - at the rock-winning
workplace and throughout the mineral transportation route. Dilution of airborne particles by
ventilation remains the primary means of controlling dust concentrations in underground mines.
Dispersed dust particles in the respirable range (less than 5 m diameter) will settle out at a
negligible rate. For the determination of dilution by airflow, respirable dust may be treated as a
gas. However, in this case it is realistic to estimate dust makes at working faces in terms of grams
(or milligrams) per tonne of mineral mined. This will, or course, depend upon the method of
working and means of rock fragmentation.
The required airflow is given as
Q =
Ed
P
Cd
3600
m3/s
(9.5)
where
Ed = the emission rate of respirable dust (mg/tonne)
P = rate of mineral production (tonnes/h) and
Cd = allowable increase in the concentration of respirable dust (mg/m3).
Example.
The intake air entering a working area of a mine carries a mass concentration of 0.5 mg/m3
respirable dust. Face operations produce 1000 tonnes of mineral over an 8 hour shift and add
respirable dust particles to the airflow at a rate of 1300 mg per tonne of mineral mined. If the
concentration of respirable dust in the return air is not to exceed 2 mg/m3, determine the required
airflow.
Solution.
Average rate of mineral production
1000
= 0.0347
8 3600
Average rate of emission of respirable dust
0.0347 1300
t/s
t mg
= 45.14 mg/s
s t
9- 8
Dust removal capacity of air, Cd = (allowable dust concentration - intake dust concentration)
=
(2 - 0.5)
= 1.5 mg/m3.
Required airflow
45.14
1 .5
mg m 3
s mg
30.1 m3/s
For larger particles of dust, the treatment is rather different. In this case, it is primarily the velocity
rather than the quantity of the airflow that is important. In any airway the distance over which dust
particles are carried by the airstream depends upon the air velocity and the settling rate of the
individual dust particles. The latter depends, in turn, upon the density, size and shape of the
particle as well as the psychrometric condition of the air. The heavier and more spherical particles
will settle more rapidly while the particles that are smaller or of greater aspect ratio will tend to
remain airborne for a longer period of time.
An effect of non-uniform dispersal characteristics is that the mineralogical composition of the
airborne and settled dust is likely to vary with distance downstream from the source. The sorting
effect of airflow on dust particles explains the abnormally high quartz content of dust that may be
found in the return airways of some coal mines (Section 20.3.8.).
The larger particles will not be diluted proportionally by increased airflow and, hence, higher
velocities. As the airspeed increases, these particles will remain airborne for longer distances
before settling out. Furthermore, as the air velocity continues to rise, previously settled particles
may be entrained by turbulent eddies into the airstream. The dust concentration will increase.
The effects of these mechanisms are illustrated on Fig. 9.2. A minimum total dust concentration is
obtained at a velocity of a little over 2 m/s. Fortunately, the total dust concentration passes
through a fairly broad based curve and air velocities in the range 1 to 4 are acceptable. Above 4
m/s the problem is not so much a health hazard as it is the physical discomfort of large particles
striking the skin. In addition to the question of ventilation economics, this limits the volume flow
that can be passed through any workplace for cooling or the dilution of other pollutants.
9.3.4. Heat
Most of the pollutants that affect the quality of air in an underground environment enter the
ventilating airstreams as gaseous or particulate matter and, hence, may be diluted by an
adequate supply of air. The problem of heat is quite different, involving changes in the molecular
behaviour of the air itself and its thermodynamic and psychrometric properties. Furthermore, it is
normally the case that air entering mine workings from the base of downcast shafts, slopes or
intake adits is relatively free of gaseous or particulate pollutants. On the other hand, the condition
of the intake air with regard to temperature and humidity depends upon the surface climate, and
the depth of the workings. Whilst the removal of heat by ventilation remains the dominant method
of maintaining acceptable temperatures in the majority of the worlds mines, quantifying the
effects of varying airflow is by no means as straightforward as for gases or dust.
The complication is that air velocity and, by implication, airflow is only one of several variables
that affect the ability of a ventilating airstream to produce a physiologically acceptable mine
climate. Calculations of the climatic effect of variations in airflow are necessarily more involved
than for gases or particulates.
The heat energy content of air, defined in terms of kilojoules of heat associated with each
kilogram of dry air, is known as Sigma Heat, S. This concept is discussed fully in Section 14.5.3.
Sigma heat depends only upon the wet bulb temperature of the air for any given barometric
pressure.
9- 9
Relative Dust Concentration
12
10
total
6
+10 microns
2
-5 microns
0
Air Velocity m/s
Figure 9.2 Variation of dust concentration with respect to air velocity.
The stages of determining the airflow required to remove heat from a mine or section of a mine
are as follows:
(a) Evaluate the sigma heat of the air at inlet, S1, using equations (14.44) to (14.47).
(b) Evaluate the highest value of sigma heat, S2, that can be accepted in the air leaving the
mine or section of the mine. This threshold limit value may be specified in terms of one of
the indices of heat stress (Chapter 17) or simply as a maximum acceptable (cut-off) value
of wet bulb temperature.
(c) Estimate the total heat flux, q12 (kW), into the air from all sources between inlet and outlet
(Chapter 15). This may involve simulation studies for additions of strata heat.
9 - 10
(d) The required airflow, Q, is then given as

Q =
m3
s
q12
(S 2 S1 )
(9.6)
where = mean density of the air (kg/m3)

The factor Q (S2 S1) is sometimes known as the Heat Removal Capacity (HRC) of a given
airflow, Q.
It should be noted that as S1 approaches S2, or for high values of added heat, q12, the required
airflow may become excessive. In this case, either q12 must be reduced or air cooling plant
installed (Chapter 18).
Example.
Air enters a section of a mine at a wet bulb temperature, tw of 20 C and a density, , of 1.276
kg/m3. The mean barometric pressure is 110 kPa. It has been determined that 2 MW of heat are
added to the air in the section. If the wet bulb temperature of the air leaving the section is not to
exceed 28 C and no air coolers are to be used, determine the required airflow.
Solution.
From equations (14.44 to 14.47, Chapter 14 on Psychrometry) sigma heat is given as
0.622 e sw
(2502.5 2.386 t w ) + 1.005 t w
(P e sw )
S =
kJ
kg
(9.7)
where saturation vapour pressure at wet bulb temperature is calculated as

17.27 t w
e sw = 0.6106 exp
237.3 + t w
kPa
At intake tw = 20C, giving

17.27 20
e sw 1 = 0.6106 exp
= 2.3375
237.3 + 20
kPa
and
S1 =
0.622 2.3375
(2502.5 2.386 20) + (1.005 20) = 53.25
(110 2.3375 )
kJ
kg
At return Using the same equations with tw set at the cut-off value of 28C gives
S 2 = 82.03
kJ/kg.
The required airflow at the intake density is given by equation (9.6)

Q1 =
=
q12
2000
=
(S 2 S1 )
1.276 (82.03 53.25 )
54.5
kJ m 3 kg
s kg kJ
m3 / s
9 - 11
In this section, we have assumed that the heat added to the air can be determined. Although this
may be straightforward for machine heat and other controlled sources (Section 15.3), it is more
difficult to quantify heat emitted from the strata. For deep mines, the design of combined
ventilation and air conditioning systems has been facilitated by the development of climatic
simulation programs. These may be incorporated into the planning procedure as shown at the
bottom of Figure 9.1. A more detailed discussion of the interaction between ventilation network
and climatic simulation programs is given in Section 16.3.5.
9.3.5. Workshops and other ancillary areas
In addition to areas of rock fragmentation, there are many additional locations in mines or other
subsurface facilities that require the environment to be controlled. These include workshops,
stationary equipment such as pumps or electrical gear, battery charging and fuel stations, or
storage areas. For long term storage of some materials in repository rooms, it may not be
necessary or even desirable to maintain a respirable environment. However, temperatures and
humidities will often require control.
In mines, it is usual for workshops and similar areas to be supplied with fresh intake air and to
regulate the exit flow into a return airway either directly or through a duct.
The airflow requirement should, initially, be determined on the basis of pollution from gases, dust
and heat as described in Sections 9.3.1 to 9.3.4. However, in the case of large excavations, this
may give rise to excessively low velocities with zones of internal and uncontrolled recirculation. In
this situation, it is preferable to employ the technique of specifying a number of air changes per
hour.
Example
An underground workshop is 40m long, 15m wide and 15m high. Estimate the airflow required
(a) to dilute exhaust fumes from diesel engines of 200 kW mechanical output and (b) on the basis
of 10 air changes per hour.
Solution
(a) Employing an estimated airflow requirement of 8 m3/s per 100 kW of rated diesel power
(Section 9.3.2) gives
200
8 = 16
100
m 3 /s
(b) Volume of room = 40 x 15 x 15 = 9000 m3

At 10 air changes per hour,
9000 10
= 25
m 3 /s
60 60
Hence, the larger value of 25 m3/s should be employed.
Airflow =
9 - 12
9.3.6. Air velocity limits

The primary consideration in the dilution of most pollutants is the volume flow of air. However, as
indicated on Figure 9.1, the velocity of the air (flowrate/cross-sectional area) should also be
determined. Excessive velocities not only exacerbate problems of dust but may also cause
additional discomfort to personnel and result in unacceptable ventilation operating costs.
Legislation may mandate both maximum and minimum limits on air velocities in prescribed
airways. A common lower threshold limit value for airways where personnel work or travel is 0.3
m/s. At this velocity, movement of the air is barely perceptible. A more typical value for mineral winning faces is 1 to 3 m/s. Discomfort will be experienced by face personnel at velocities in
excess of 4 m/s (Section 9.3.3) because of impact by large dust particles and, particularly, in cool
conditions. Table 9.1 gives a guide to upper threshold limit values recommended for air velocity.
Area
Working faces
Conveyor drifts
Main haulage routes
Smooth lined main airways
Hoisting shafts
Ventilation shafts
Velocity (m/s)
4
5
6
8
10
20
Table 9.1 Recommended maximum air velocities

In wet upcast shafts where condensation or water emissions result in airborne droplets, the air
velocity should not lie between 7 to 12 m/s. Water blanketing may occur in this range of
velocities. The resulting variations in shaft resistance cause an oscillating load on main fans and
can produce large intermittent cascades of water falling to the shaft bottom.
9.4. PLANNING EXERCISES AND TIME PHASES
The dynamic nature of a mine or any other evolving network of subsurface airways requires that
the infrastructure of the ventilation system be designed such that it can accommodate major
changes during the life of the undertaking. For a completely new mine, the early ventilation
arrangements will provide for the sinking of shafts, drifts or adits, together with the initial
development of the primary underground access routes. When the procedures
illustrated on Figure 9.1 are first applied to an existing mine, the establishment and correlation of
a basic network file will often reveal weaknesses and inefficiencies in the prevailing ventilation
system. The initial planning exercises should then be directed towards correcting those
deficiencies while, at the same time, considering the future development of the mine.
Following the establishment of a new or revised ventilation design, further planning investigations
should be carried out for selected stages of future development. Although this is commonly
termed "time phasing", it might more accurately be regarded as representing phases of physical
development rather than definitive periods of time. Schedules have a habit of slipping or are
subject to considerable revision during the course of mining.
9.4.1. Network planning exercises
Referring once again to Figure 9.1, the first step in a ventilation network planning exercise is to
establish the airflows required in all places of work, travel or plant location (Section 9.3). Air
velocity limits should also be set for the main ventilation flowpaths (Section 9.3.6). Additional
9 - 13
constraints on the direction and magnitudes of pressure differentials across doors, bulkheads or
stoppings, or between adjoining areas of the facility, may also be imposed. This is of particular
importance in nuclear waste repositories.
The ventilation engineers should review the existing network (correlated basic network or
previous time phase) and agree upon a series of alternative designs that will satisfy revised
airflow objectives and meet constraints on velocity and pressure differentials. The network
schematic should then be modified to represent each of those alternative designs in turn. The
network modifications may include
sealing worked-out areas and opening up new districts

adding main airways and lengthening existing airways
adding and/or removing air crossings, doors and stoppings
adding, removing or amending fixed quantity branches
adding or sealing shafts and other surface connections
adding, removing or relocating main or booster fans, or changing to different fan
characteristic curves in order to represent adjustments to fan speed or vane settings
amending natural ventilation pressures.
It is worth reminding ourselves, at this stage, that current network simulation programs do not
perform any creative design work. They simply predict the airflow and pressure distributions for
layouts that are specified by the ventilation planner. While computed output will usually suggest
further or alternative amendments, it is left to the engineer to interpret the results, ensure that
design objectives and constraints are met, compare the efficiency and cost effectiveness of each
alternative design and to weigh the practicality and legality of each proposed scheme.
Each alternative layout chosen by the design engineers should be subjected to a series of VNET
simulations and network amendments until either all the prescribed criteria are met or it becomes
clear that the system envisaged is impractical. Tables, histograms and spreadsheets should be
prepared in order to compare fan duties (pressures and air quantities), operating costs, capital
costs and the practical advantages/disadvantages of the alternative layouts.
9.4.2. Time phases
During the operating life of an underground mine, repository or other subsurface facility, there will
occur periods when substantial changes to the ventilation system must be undertaken. This will
occur, for example, when:
a major new area of the mine is to be opened up or an old one sealed

workings become sufficiently remote from surface connections that the existing fans or
network infrastructure are incapable of providing the required face airflows at an
acceptable operating cost
two main areas of the mine (or two adjoining mines) are to be interconnected.
"Time phase" studies should be carried out on alternative schematics to represent stages before
and after each of these changes. Additional scenarios may be chosen at time intervals between
major changes or to represent transitional stages. The latter often impose particularly heavy
duties on the ventilation system.
Time phase exercises should be conducted to cover the life of the mine or as far into the future as
can reasonably be predicted, assuming a continued market for the mined product.
It is, perhaps, intuitive to conduct time phase exercises in an order that emulates the actual
planned chronological sequence. This may not always be the most sensible order for the network
9 - 14
exercises. For example, if it appears inevitable that a new shaft will be required at some time in
the future, it may be prudent to first investigate the time phase when that shaft has become
necessary and then to examine earlier time phases. This assists the planners in deciding whether
it would be advantageous to sink the new shaft at a prior time rather than to wait until it becomes
absolutely necessary.
The time phase exercises will produce a series of quantified ventilation networks for each time
period investigated. These should be reviewed from the viewpoint of continuity between time
phases and to ensure that major additions to the ventilation infrastructure have an acceptable and
cost-effective life. Indeed, the need for such continuity should remain in the minds of the planners
throughout the complete investigation, although not to the extent of stifling viable alternative
designs at later stages of the time phase exercises.
The result of such a review should be a series of selected layouts that represent the continuous
development of the ventilation system throughout the projected life of the mine. Again, tables and
histograms should be drawn up to show the variations of fan duties and costs, this time with
respect to chronological order.
9.4.3. Selection of main fans
Care should be taken that any main fan purchased should be capable of producing the range of
pressure-volume duties necessary throughout its projected life. The variation in required fan
duties resulting from time phase exercises should be considered, particularly where a fan is to be
relocated at some time during its life.
For a new purchase, fan manufacturer's catalogues should be perused and, if necessary,
discussions held with those manufacturers, in order to ensure that the fan selected is capable of
providing the required range of duties. Fortunately, the availability of variable pitch axial fans and
inlet vane control on centrifugal fans allows a single fan to provide a wide range of duties with a
fixed speed motor. A more detailed discussion of fan specifications is given in Section 10.6.2.
9.4.4. Optimization of airflow systems
The network exercises carried out for any one time phase should incorporate a degree of
optimization concerning the layout of airways, and the locations and duties of new shafts and
fans. For a single major flowpath such as a proposed new shaft, a detailed optimization study can
be carried out (Section 9.5.5). However, the larger questions that can arise involve a balance
between capital and operating costs and might include a choice between additional airways and
fans of greater power, or the minimization of total power costs incurred by alternative
combinations of main and booster fans. Special purpose optimization programs have been written
to assist in the resolution of such questions. (Calizaya et al, 1988)
9.4.5. Short term planning and updating the basic network
This chapter is concerned primarily with the major planning of a subsurface ventilation system.
The time phases are chosen to encompass periods of significant change in the ventilation layout.
However, much of the daily work of mine ventilation engineers involves planning on a much
shorter time scale.
As working districts advance or retreat, the lengths of the airways that serve them change. Also,
strata stresses can result in reductions of the cross-sectional areas of those airways. Additional
roof support may be required resulting in increased friction factors. Stoppings, regulators, doors
and air-crossings may deteriorate and allow greater leakage. All of these matters cause changes
in the resistances of individual airways and, hence, of the complete network. Additionally, wear of
9 - 15
fan impellers can result in variations in the corresponding pressure - volume characteristic curves.
For these reasons, a basic network file that correlated well with survey data at the start of any
given time phase may become less representative of the changing real system.
In addition to regular maintenance of all ventilation plant and controls, further surveys should be
carried out on a routine basis and, in particular, whenever measured airflows begin to deviate
significantly from those predicted for that time phase. Visual inspections and frequent liaison with
shift supervisors are also invaluable for the early identification of causes of deviation from
planned airflows. Relationships with production personnel will remain more congenial if the latter
are kept well advised of the practical consequences of blocking main airflow routes with supplies,
equipment or waste material.
As new survey data become available, the basic network file should be updated accordingly. If
the basic network file is maintained as a good representation of the current mine then it will prove
invaluable in the case of any emergency that involves the ventilation system.
In the event that such updates indicate permanent and significant impact on the longer term plans
then the network investigations for the future time phases should be re-run and the plans
amended accordingly. In the majority of cases this may become necessary because of changes
in mine production plans rather than from any unpredicted difficulties with ventilation.
9 .5. VENTILATION ECONOMICS AND AIRWAY SIZING
The subsurface ventilation engineer must be capable of dealing with two types of costs:
(a) Capital costs that require substantial funds at a moment in time or distributed over a short time
period. A main fan installation or sinking a new shaft comes into this category.
(b) Working or operating costs, these representing the expenditure of funds on an ongoing basis
in order to keep a system operating. Although consumable items, maintenance and even small
items of equipment may be regarded as working costs, ventilation system designers often
confine the term operating costs to the price of providing electrical power to the fans.
The problem that typically arises concerns the combination of capital and operating costs that will
minimize the real total cost to the company or mining organization; for example, whether to
purchase an expensive but efficient fan to give low operating costs or, alternatively, a less
expensive fan that will necessitate higher operating costs. Another question may be whether the
money saved initially by sinking a small diameter ventilation shaft is worth the higher ongoing
costs of passing a required airflow through that shaft. These are examples of the types of
questions that we shall address in this section.
In the following examples and illustrations, we shall employ the dollar ($) sign to indicate a unit of
money. However, in all cases, this may be replaced by any other national unit of currency.
9.5.1. Interest payments
Money can be regarded as a commodity that may be circulated to purchase goods or services. If
we borrow money to buy a certain item then we are, in effect, renting the use of that money. We
must expect, therefore, that in addition to returning the borrowed sum, we must also pay a rental
fee. The latter is termed the interest payment. Hence, if we borrow $100 for a year at an annual
interest rate of 9 per cent, then we must repay $109 one year from now. Even if we owned the
money and did not have to borrow, spending the $100 now would mean that we are prevented
from earning interest on that sum by investing or lending it to an individual, company, bank or
other financial institution. Using money always involves an interest penalty.
9 - 16
If we borrow an amount of money, P (principal) at a fractional interest rate i (e.g. at 9 per cent, i =
0.09) then we will owe the sum S where
Now
S = P
after 1 year
S = P (1+i)
after 2 years
S = P (1 + i) (1 + i) = P(1 + i)2
after 3 years
S = P(1 + i)3
after n years
S = P(1 + i)n
(9.8)
Figure 9.3 gives a visual indication of the effects of compounding interest each year.
100
Value of $1 after n years
Annual rate of interest = 20 %
18
16
14
12
10
8
10
6
4
2
1
0
10
15
20
25
30
35
Years n
Figure 9.3 Effect of compounding interest.
Example 1
$250 000 is borrowed for 8 years at an annual interest rate of 10 per cent. Determine the total
sum that must be repaid at the end of the period, assuming that no intervening payments have
been made.
Solution
S
250 000 (1 + 0.1)8
$535 897
9 - 17
Example 2
How long does it take for an investment to double in dollar value at an annual interest rate of 12
per cent?
Solution
In this example, the final sum is twice the principal, S = 2P. Hence, equation (9.8) becomes
2P = P (1 + 0.12)n
giving
ln (2)
n =
= 6.12
years
ln (1.12)
9.5.2. Time value of money, present value
It is clear from a consideration of interest that the value of a set sum of money will vary with time.
At 9 per cent interest, $100 now is worth $109 in one year and $118.81 two years from now. The
effect of national inflation will, of course, partially erode growth in the value of capital. To
determine real purchasing power, rates of interest should be corrected for inflation. In this section
we are assuming that this has been done and quoted interest rates have taken projected inflation
into account.
9.5.2.1. Present value of a lump sum.
The time variation of the value of money makes it inequitable to compare two sums that are
borrowed or spent at different times. We need a common basis on which both sums can be fairly
evaluated. One method of doing this is to determine what principal or capital, P, we need to invest
now in order that it will grow to a desired sum, S, in a specified number of years. This is given by
a simple transposition of equation (9.8)
P =
(9.9)
(1 + i ) n
When determined in this way, the value of P is known as the present value of the future sum S.
If all future investments or expenditures are reduced to present values then they can correctly be
compared.
Example
Three fans are to be installed; one immediately at a price of $260 000, one in 5 years at an
estimated cost of $310 000 and the third 8 years from now at $480 000. Determine the total
expenditure as a present value if the annual interest rate is 10 per cent.
Solution
From equation (9.9), P =
S
(1.1) n
Fan
Time
n
Purchase price
S
Present value
P
now
$ 260 000
$260 000
5 years
$ 310 000
$ 192 486
8 years
$ 480 000
$ 223 924
Total
$ 676 410
In this example, the first fan is actually the most expensive.
9 - 18
9.5.2.2. Present value of regular payments.

In the case of operating costs, payments must be made each year. Such future payments may
also be expressed as present values in order to compare and compound them with other
expenditures.
We assume operating costs, So, to be constant and paid at the end of each year. Then at the end
of the first year (n = 1), equation (9.9) gives
So
(1+ i )
= present value of the first year's annual operating cost So: at an interest rate of i.
Po,1 =
where Po,1
Similarly, the present value of the same operating cost, So, in the second year (n = 2) is
So
Po,2 =
(1 + i ) 2
It follows that the total present value, Po, of operating costs, So paid at each year-end for n years
becomes
1
1
1
1
Po = S o
+
+
+ ..........
2
3
(1+ i ) n
(1 + i )
(1 + i ) (1+ i )
(9.10)
The bracketed term is a geometric progression which can be summed to give

Po
So
i
1
1
(1 + i ) n
(9.11)
Figure 9.4 gives a graphical representation of this equation for So = $1
$14
Annual rate of interest = 6%
Present value of $1
paid each year for n years
$12
$10
10
$8
12
14
16
$6
18
20
$4
$2
$0
0
10
12
14
16
18
20
22
24
26
28
Number of Years n
Figure 9.4 Present value of annual operating costs.

9 - 19
Example 1
Electrical power costs at a mine are estimated to be $850 000 in each of the next 12 years.
Determine the present value of this expenditure at 11 per cent interest.
Solution
With So = $850 000, i = 0.11 and n = 12 years, equation (9.11) gives
Po
850 000
1
1
0.11
(1 + 0.11)12
$5 518 503
Example 2
Tenders received from manufacturers indicate that a fan priced at $170 000 will cost $220 000
per year to run, while a fan with a purchase price of $265 000 will require $190 000 per year in
operating costs. If the costing period is 5 years at an annual interest of 7 per cent, determine
which fan is most economical.
Solution
Using equation (9.11) with n = 5 years, i = 0.07 and So set at $220 000 and $190 000 respectively
allows the present value of the operating costs to be determined.
Present values
Fan
Current
purchase
price ($)
Annual operating
cost($/year)
Purchase
price ($)
Operating cost
over 5 yr ($)
Total ($)
170 000
220 000
170 000
902 043
1 072 043
265 000
190 000
265 000
779 038
1 044 038
In this example, fan 2 is shown to be the more economic choice.

During the analysis of time phase exercises (Section 9.4.2) it may be necessary to establish the
present value of operating costs for periods that do not commence at the present time. By
applying equation (9.11) to periods of m years and n years respectively, and taking the difference,
the present value of operating costs for the future period n to m years becomes
Po,n m =
So
1
i (1 + i )n
1
m
(1 + i )
(9.12)
9 - 20
Example
This example shows an excerpt from a spreadsheet used to calculate the present values of
capital and operating expenditures. Present values are calculated from equation (9.9) for capital
(lump sum) expenditures and equation (9.12) for operating costs. The annual rate of interest is 10
per cent.
Period
Capital
expenditure($)
1. now
Operating cost
($/year)
500 000
500 000
620 000
2. now to end of 4 years.
Present value($)
1 965 317
(n = 0, m = 4)
3. at end of 4 yrs
235 000
4. end of 4 yrs to end of

10 yrs (n = 4, m = 10)
160 508
680 000
2 022 797
Total present value= $4 648 622

Spreadsheets of this type can be set up on a personal computer to give a rapid means of costing
the results of alternative network exercises.
9.5.3. Equivalent annual cost
When capital is borrowed, it is often more convenient to repay it, including interest, in equal
installments each year, rather than as a lump sum paid at the end of the complete time period.
This is a common method for domestic purchases such as a house (mortgage payments) or an
automobile. In industry, this equivalent annual cost facilitates budgetary planning and enables
major items of capital expenditure, such as a new shaft, to be spread evenly over a number of
years. The method of equivalent annual cost allows capital expenditures to be compared with
operating costs on a year by year basis and, for producing mines, also enables those capital
expenditures to be expressed in terms of cost per tonne of mineral.
The equivalent annual cost is given by a re-interpretation of equation (9.11). P is the capital that
must be invested now (present value) in order to be able to make a regular payment, S, each
year. The statement may also be made in reverse, i.e. S is the regular payment, or annual
equivalent cost, to be met each year in order to pay off the capital and interest on a borrowed
amount, P.
Transposing equation (9.11) gives the equivalent annual cost, EAC, as
EAC ( = S ) =
Pi
[1 1 /(1+ i ) n ]
(9.13)
9 - 21
Example
A mine shaft is to be sunk at a cost of $5.6 million. The life of the shaft is estimated to be 15
years during which time the average planned rate of mineral production is 1.6 million tonnes per
year. If the annual interest rate is 8.75 per cent, determine (a) the equivalent annual cost of the
shaft in $ per year and (b) corresponding production cost in $ per tonne mined.
Solution
(a) Using equation (9.13) with P = $5.6 million, i = 0.0875 and n = 15 years, gives the equivalent
annual cost as
EAC =
5 600 000 0.087 5

[ 1 1 /(1.0875 )15 ]
$684 509
(b) The corresponding production cost is

684 509
1600 000
$0.428 per tonne
9.5.4. Ventilation operating costs

A fan unit, comprising an electric motor, transmission and impeller converts electrical energy into
air power. The latter is reflected as kinetic energy of the air and a rise in total pressure across the
fan. Air power delivered by a fan was quantified in Chapter 5 as
pft x Q
where pft = rise in total pressure across the fan (Pa) and
Q = airflow (m3/s)
(See, also, Chapter 10 for the effects of compressibility).
However, the electrical power taken by the fan motor will be greater than this as losses occur
inevitably in the motor, transmission and impeller. If the overall fractional efficiency of the unit is ,
then the electrical input power to the motor will be
p ft Q
Electrical power charges are normally quoted in cost per kilowatt-hour. Hence, the cost of
operating a fan for 24 hours per day over the 365 days in a year is
So
p ft Q
e 24 365
1000
$/year
(9.14)
where e is the cost of power ($ per kwh).

In practice, the fan pressure, pft is often quoted in kPa, obviating the need for the 1000.
Equation (9.14) also applies for the annual cost of ventilating an individual airway. In this case, pft
simply becomes the frictional pressure drop across the airway, p, at the corresponding airflow, Q,
and the overall efficiency of the fan primarily responsible for ventilating that airway. In cases of
multiple fans, a weighted mean average of fan efficiency may be employed.
9 - 22
Example
An underground airway is driven at a capital cost of $1.15 million dollars. During its life of 8 years,
it is planned to pass an airflow, Q, of 120 m3/s at a frictional pressure drop, p, of 720 Pa. The
main fans operate at an overall efficiency, , of 72 per cent. If the annual rate of interest is 9.5 per
cent and the average cost of electrical power is $0.06 per kWh, determine the annual total cost of
owning and ventilating the airway.
Solution
The equivalent annual cost of owning the airway is given by equation (9.13) with P = $1.15
million, i = 0.095 and n = 8 years
EAC =
1150 000 0.095

[ 1 1 /(1.095 ) 8 ]
$211 652 per year
The annual operating cost is given by equation (9.14) with

pft = 720 Pa, Q = 120 m3/s, = 0.72 and e = $0.06 per kWh
So
720 120
0.06 24 365
1000 0.72
= $63 072 per year
The total yearly cost is given by adding the cost of owning the airway and the annual operating
cost of ventilation:
total annual cost C = EAC + So = 211 652 + 63 072
= $274 724 per year
9.5.5. Optimum size of airway

There are several factors that influence the optimum size of a subsurface airway, including
the airflow to be passed through it
the cost of excavation
limitations on air velocity (Section 9.3.5)
the span that can be supported adequately and
the size of equipment required to travel through the airway.
This section considers the first two of these matters. However, the size of a planned new shaft or
airway must satisfy the other constraints.
As the size of an airway is increased then its resistance and, hence, ventilation operating costs
will decrease for any given airflow (Section 5.4.1). However, the capital cost of excavating the
airway increases with size. The combination of capital and operating costs will be the total cost of
owning and ventilating the airway. The most economic or optimum size of the airway occurs when
that total cost is a minimum. The costs may be expressed either as present values or in terms of
annual (equivalent) costs.
In order to quantify the optimum size, it is necessary to establish cost functions for both capital
and operating costs that relate those expenditures to airway size.
9.5.5.1. Capital cost function
The business of arriving at a cost for excavating a mine shaft or major airway often involves
protracted negotiations between a mining company and a contractor. Even when the task is to be
undertaken in-house, the costing exercise may still be extensive.
9 - 23
There are, essentially, two components - fixed costs and variable costs. The fixed costs may
include setting up and removal of equipment such as temporary headgear, hoisting facilities or a
shaft drilling rig. Other items such as pipe ranges, electrical conduits, air ducts and track lines are
usually independent of airway size but are a function of airway length.
The variable costs are those that can be expressed as functions of the length and cross-sectional
area or diameter of the shaft or airway. These include the actual cost of excavation and supports.
The capital cost, Pc, may then be expressed as the cost function
=
Pc
Cf + f (A,L)
(9.15)
where Cf = fixed costs

and
f (A,L) = function of cross-sectional area, A and length, L.
In a simple case, the capital cost function may take the form
=
Pc
Cf + aV + bL
(9.16)
(m3)
where V = volume excavated, A x L

a and b are constants.
For mechanized excavations, there may be discontinuities in the capital cost function due to step
increases in the size and sophistication of the equipment for differing sizes of airway.
9.5.5.2. Operating cost function
Equation (9.14) gave the annual ventilation cost of an airway to be
= pQ
So
e
24 365
1000
$/year
(9.17)
However, from equation (2.50),

p = Rt Q 2
= rational resistance of the airway (m-4)

where Rt
and
giving
= Rt Q 3
So
24 365
1000
This demonstrates that the operating cost varies with airway resistance, Rt, air density, , and the
cube of the airflow, Q.
Substituting for Rt from equation (2.51) gives
So
where f
L
per
and
A
=
=
=
=
=
f L per
2A
Q3
24 365
1000
$/year
(9.18)
coefficient of friction (dimensionless)

length of airway (m)
perimeter (m)
cross-sectional area (m2)
9 - 24
In order to achieve the corresponding result with the Atkinson friction factor, k1.2, substituting for p
from equation (5.8) allows the operating cost function of equation (9.17) to be written as
So
= k 1.2
L per
A
1 .2
Q3
24 365
1000
$/year
(9.19)
9.5.5.3. Case Study

During the design of a proposed circular shaft, the following data were generated:
Shaft sinking:
Equipment set up and decommissioning costs
Excavation cost
Fittings and lining
Physical data:
Depth of shaft,
Effective coefficient of friction,
(friction factor k1.2 = 0.6f = 0.006 kg/m3)
Mean air density,
Airflow,
Fan efficiency,
Life of shaft,
Additional financial data:
Annual rate of interest,
Average cost of electrical power
= $650 000
= $290/m3
= $460/m
L
f
= 700 m
= 0.01
=
=
=
=
i
e
= 10 per cent
= $0 .075 per kWh
1.12 kg/m3
285 m3/s
0.65
15 years
Determine the optimum diameter of the shaft.

Solution
Task 1: Establish the capital cost function:
From the data given, the capital cost function can be expressed in the form of equation (9.16)
PC
650 000 + 290 V + 460 L
and
giving
D2
3
700 (m )
4
D = shaft diameter (m)
where V = AL =
Pc
D2
= 650 000 + 290 700
+ ( 460 700 )
4
The capital cost function then simplifies to

Pc
= 972 000 + 159 436 D 2
(9.20)
This is the present value of the capital cost of sinking the shaft. The complete analysis may be
carried out either in terms of present values or in annual (equivalent) costs. The latter method is
employed for this case study. Hence, the capital cost of shaft sinking must be spread over the 15
year life using equation (9.13) with i = 0.1 and n = 15.
9 - 25
Equivalent annual cost of shaft sinking:

(972 000 + 159 436 D 2 ) 0.1
EAC =
[1 1 /(1.1)15 ]
= 127 793 + 20 962 D2
$/year
(9.21)
Task 2: Establish the operating cost function:

This is accomplished by substituting the given data into equations (9.18) or (9.19)
Annual operating cost
0.01 700 D
0.075 24 365
So =
1.12 285 3
2
3
0.65
1000
2 (D / 4)
= 594.775 x 106/D5
$/year
$/year
(9.22)
Task 3: Establish the total cost function:

Addition of the capital and operating cost functions from equations (9.21) and (9.22) respectively
gives the total annual cost function as
C = 127 793 + 20 962 D 2 +
594.775 10 6
$/year
D5
(9.23)
Task 4: Determine the optimum diameter:

This can be accomplished in either of two ways.
(a) By graphical means.
The three cost functions have been plotted on Figure 9.5. The minimum point on the total
annual cost curve occurs at an optimum diameter of approximately 4.9m.
An advantage of constructing the cost curves is that it gives a visual indication of the
behaviour of the cost functions. In particular, the operating and total costs rise rapidly when
the airway diameter is reduced significantly below the optimum point. However, the total cost
curve remains fairly shallow above the optimum point. This is a typical result and illustrates
that the shaft size may be escalated to a standard diameter above the optimum to facilitate
shaft sinking. Indeed, experience has shown the wisdom of sizing a shaft above the optimum.
This allows subsequent flexibility for modifications to the airflow or mine production.
(b) By analytical means.
If the total cost function, C, has been expressed as a convenient function of cross-sectional
area, A, or diameter, D, then it can be differentiated to find the optimum location when the
slope dC/dA or dC/dD becomes zero.
In this case study, differentiation of equation (9.23) gives
5 594.775 10 6
dC
= 2 20 962 D
dD
D6
At dC/dD = 0
41924 D 7
= 2973 .875 10 6
from which
D = 4.93 m.
9 - 26
Annual cost $million per year
3
2.5
2
total cost
1.5
1
operating
cost
capital cost
(EAC)
0.5
0
3
3.5
4.5
5.5
Shaft diameter m
Figure 9.5 Annual cost functions for the circular shaft used in the Case Study
This is an accurate value of the optimum diameter previously read from Figure 9.5 as the
approximation 4.9m. In practice, the recommended diameter would be rounded up to at least
5m.
The corresponding total annual equivalent cost can be estimated from Figure 9.5 or calculated
by substituting D = 9.3 into equation (9.23):
C = 127 793 + 20 962 4.93 2 +
594.775 10 6
4.93 5
$ 841 502 per year
9.5.6. Incorporation of shaft design into network planning exercises.

Figure 9.1 shows the optimization of major airways and fans to be a part of the ventilation
planning process. However, it may be impractical to carry out such optimization for every
investigatory cycle of a ventilation network exercise. The sizing of a shaft is normally conducted
only after network analysis has established a satisfactory distribution of airflows underground and
can be combined with the shaft design procedures described in Section 5.4.6.
9 - 27
The demands made upon a mine shaft may include access for personnel and equipment, the
hoisting of planned tonnages of rock and the passing of specified airflows. We can now define a
set of guidelines to assist in the management of a combined network analysis and shaft design
exercise.
1. Assess the duties required for rock hoisting (tonnes per hour), number of personnel to be
transported, time allowed at shift changes and the size, weight and frequency of hoisting
materials and equipment.
2. Determine alternative combinations of conveyance sizes and hoisting speeds.
3. Conduct ventilation network investigations, initially on the basis of an estimated shaft
resistance until a satisfactory distribution of airflow is achieved. This will establish the
required airflow in the shaft.
4. Assess the dimensions of proposed shaft fittings including pipes, cables, guides and
buntons.
5. Conduct an optimization exercise to find the size of shaft that will pass the required
airflow at the minimum combination of operating costs and the capital expense of shaft
construction (Section 9.5.5.).
6. Check the free area air velocity. If this exceeds 10 m/s in a hoisting shaft or 20 m/s in a
shaft used only for ventilation then the cost of enlarging the shaft should be reviewed. It
will be recalled that the total cost vs. diameter curve is usually fairly flat above its
minimum point (Section 9.5.5.3).
7. Determine the coefficient of fill, CF, for the largest of the proposed conveyances (Section
5.4.6.3). If this exceeds 30 per cent for two or more conveyances, or 50 per cent for a
single conveyance shaft, then the dimensions of the skip or cage should be reviewed or
the size of the shaft, again, be re-examined.
8. Calculate the maximum relative velocity between the airflow and the largest conveyance,
(ua + uc)/(1 CF) (See section 5.4.6.3 for nomenclature.)
If this exceeds 30 m/s then additional precautions should be taken to ensure
aerodynamic stability of the moving conveyances. In any event, the relative velocity
should not exceed 50 m/s.
9. Assess the air velocities at all loading/unloading stations and, if necessary, redesign the
excavations to include air bypasses or enlarged shaft stations.
10. Determine the total resistance of the shaft using the methods described in Section 5.4.6.
Examine all feasible means of reducing the resistance including the streamlining of
buntons and the aerodynamic design of intersections. For shafts of major importance,
construct and test physical or CFD 2 models of representative lengths of the shaft and
main intersections.
11. Rerun ventilation network analyses with the established value(s) of shaft resistance in
order to determine the final fan pressures required.
It is clear that there is a considerable amount of work involved in the design, sizing and costing of
a proposed mine shaft or major airway. Fortunately, microcomputer program packages are
available that reduce the effort to little more than assembling the essential data. These are
particularly valuable for carrying out sensitivity studies in order to assess the effects of changes in
design or financial constraints.
Computational Fluid Dynamics
9 - 28
9.6. TRADITIONAL METHOD OF VENTILATION PLANNING

The methods of subsurface ventilation planning described in the previous sections of this chapter
rely upon the availability of computers and appropriate software. A question that may arise is how
such planning can be carried out without high speed computational aid. For this, we may revert
back to the traditional methodology that was well developed prior to the computer revolution.
Although the older techniques cannot begin to match the speed, versatility and detail of computer
assisted planning, they do retain a role in estimating generic values for airflows, fan pressures
and air quality at the early conceptual design stage of a proposed new mine or extension to
an existing facility.
The traditional approach proceeds along the following sequence:
1. Determine air volume flows required in working areas. This can employ the methods
discussed in Section 9.3. However, at a preliminary stage of planning, empirical values of
airflow based on rate of tonnage might be used. In this case, care must be taken to
ensure that the basis of the empirical guidelines is compatible with the intended mining
method and geology.
2. Assess the airflow requirements for development areas, mechanical or electrical plant
and workshops, and estimate the volume flows that pass through abandoned workings,
stoppings and other leakage paths. The estimation of leakage flows relies strongly upon
the experience and intuition of the ventilation engineer. Unfortunately, in many mines the
volumetric efficiency (Section 4.2.3) is fairly low. Inaccuracies in the estimates of airflow
through individual leakage paths, while perhaps of little consequence when considered
individually, will accumulate into major errors in the main ventilation routes. The square
law, p = RQ2, then produces twice the corresponding percentage error in the frictional
pressure drop. The treatment of leakage flows is probably the single greatest cause of
imprecision in the traditional method of ventilation planning.
3. Indicate the estimated airflows on a mine plan and compound them to show air flowrates,
Q, through every major airway.
4. Using the given airflows and proposed sizes of airways, determine the corresponding air
velocities. If these exceed limiting values (Section 9.3.6), the need for larger or additional
airways is indicated.
5. Assess the resistance, R, of each branch along the main ventilation routes, either from
estimated friction factors and airway geometries or on the basis of local empirical data.
6. Using the square law, p = RQ2, determine the frictional pressure drop, p, for each main
branch and indicate these on the mine plan.
7. Starting at the entrance of a downcasting surface connection, trace a path along intake
airways to the most distant workings, through those workings, and back to the surface via
return airways. Sum the frictional pressure drops around the complete traverse. This
exercise is repeated for a number of such traverses to incorporate various working areas.
The loop showing the greatest summation of frictional pressure drops then gives an
approximation of the main fan pressure required to ventilate the mine. Subsidiary circuits
may be controlled by regulators or upgraded by booster fans3. Pressure gradient
diagrams can be employed to give a visual indication of the cumulative pressure drops.
The traditional approach is similar, in principle, to that used for the design of duct systems in
buildings. It is simple in concept and requires little computational aid. Unfortunately, it suffers from
some severe drawbacks:
1. It relies strongly upon the experience of the engineer and his/her empirical knowledge of
the airflow distribution patterns. A mine is very different to a duct system in a building, not
3
where booster fans are permitted by the governing legislation.
9 - 29
only in scale but also because of the dynamic nature of mining operations, and the
tremendous variability in the geometry of airflow paths with respect to time as well as
location.
2. The highly interactive and nonlinear relationships that exist between ventilation
parameters are largely ignored. Leakage airflows through caved strata, old workings or
across stoppings, doors and air crossings are dependent upon the geometry of the flow
paths, the pressure differential and the degree of turbulence. Ascribing fixed values or
even fixed proportions of available airflow to leakage can do no more than achieve rough
approximations. In many mines, the majority of total airflow passes through leakage
paths. Errors in estimated leakages will accumulate and be reflected in the corresponding
main airflow routes and, because of the nonlinearity of the laws of airflow, can result in
large errors in the cumulative pressure drops.
3. There is a basic lack of reality in a procedure that estimates an airflow pattern and then
backtracks to find a fan pressure that will produce airflows capable of being manipulated
into the required distribution. In practice, when a fan is switched on, a complex
configuration of interdependent pressure drops and airflows is set up throughout the
network, the ventilating pressures producing airflows and the airflows, in turn, producing
frictional pressure drops. In effect, the airflows shown by the traditional method are
simply the initial estimates based on desired airflows in the workplaces and assumed
leakages. No attempt is made to simulate the actual airflow distribution that will occur
when the fans are switched on.
4. There is very little opportunity to study alternative options in order to optimize the
effectiveness and operating economics of the ventilation system.
5. Assuming the airflow distribution allows little flexibility in investigating the effects of fan
duty/position, or the adjustment or re-siting of regulators, doors and booster fans.
Bibliography
Aldridge, M.D. and Nutter, R.S. (1980). An experimental ventilation control system. 2nd. Int.
Mine Ventilation Congress, Reno, NV. pp.230-238.
Calizaya, F. et al. (1989) Guidelines for implementation of a booster fan monitoring and control
system in coal mines. 4th U.S. Mine Ventilation Symposium. Berkeley, Ca pp 382-391
Calizaya, F et al. (1988) A computer program for selecting the optimum combination of fans and
regulators in underground mines. Trans. 4th Int. Mine Ventilation Congress. Brisbane, Australia,
pp 141-150.
Dunmore, R. (1980) Towards a method of prediction of firedamp emission for British coal mines.
Proc. 2nd Int. Mine Ventilation Congress, Reno, Nevada, USA (AIME publication) pp 351-64.
Kissell, F.N. (1978) Some new developments in mine ventilation from the U.S. Bureau of Mines.
July, Mine Ventilation Soc. of S. Africa. Vol 31. pp 85-89.
Lambrechts, J. de V. and Howe, M.J. (1982) Mine Ventilation Economics. Chapter 33.
Environmental Engineering in South African Mines. Mine Ventilation Soc. of S. Africa.
Mahdi, A. and McPherson, M.J. (1971). An introduction to automatic control of mine ventilation
systems. Mining Technology. May.
McPherson, M.J. (1988) An analysis of the resistance and airflow characteristics of mine shafts.
4th International Mine Ventilation Congress. July, Brisbane, Australia
McPherson, M.J. (1984) Mine ventilation planning in the 1980's. International Journal of Mining
Engineering. Vol. 2 pp 185-227.
Rustan, A. and Stockel, I. (1980). Review of developments in monitoring and control of mine
ventilation systems. 2nd Int. Mine Ventilation Congress. Rena, NV. pp. 223-229.
9 - 30
Chapter 10. Fans
CHAPTER 10
FANS
10.1. INTRODUCTION .........................................................................................2

10.2. FAN PRESSURES.......................................................................................2
10.3. IMPELLER THEORY AND FAN CHARACTERISTIC CURVES ..................5
10.3.1. The centrifugal impeller.................................................................................................... 5
10.3.1.1. Theoretical pressure developed by a centrifugal impeller......................................... 6
10.3.1.2. Theoretical characteristic curves for a centrifugal impeller ....................................... 9
10.3.1.3. Actual characteristic curves for a centrifugal impeller............................................. 11
10.3.2. The axial impeller........................................................................................................... 14
10.3.2.1. Aerofoils .................................................................................................................. 14
10.3.2.2. Theoretical pressure developed by an axial fan ..................................................... 15
10.3.2.3. Actual characteristic curves for an axial fan............................................................ 20
10.4 FAN LAWS..................................................................................................22

10.4.1. Derivation of the fan laws............................................................................................... 22
10.4.2. Summary of fan laws ..................................................................................................... 24
10.5. FANS IN COMBINATION...........................................................................25

10.5.1. Fan in series .................................................................................................................. 26
10.5.2. Fans in parallel............................................................................................................... 27
10.6. FAN PERFORMANCE ...............................................................................28

10.6.1. Compressibility, fan efficiency, and fan testing.............................................................. 29
10.6.1.1. The pressure-volume method ................................................................................. 30
10.6.1.2. The thermometric method ....................................................................................... 34
10.6.2. Purchasing a main fan. .................................................................................................. 39
10.7. BOOSTER FANS.......................................................................................40

10.7.1. The application of booster fans...................................................................................... 40
10.7.2. Initial planning and location ........................................................................................... 40
Practical constraints on booster fan location ...................................................................... 41
Leakage and recirculation................................................................................................... 41
Steady state effects of stopping the booster fan ................................................................ 42
Economic considerations.................................................................................................... 42
10.7.3. Monitoring and other safety features ............................................................................. 42
10.7.4. Booster fan control policy............................................................................................... 43
Bibliography ........................................................................................................45
APPENDIX A10.1 ...............................................................................................46
Comparisons of exhausting and forcing fan pressures to use in pressure surveys.... 46
APPENDIX A10.2 ...............................................................................................49

Derivation of the isentropic temperature-pressure relationship for a mixture of air,
water vapour and liquid water droplets. ............................................................................ 49
Chapter 10. Fans
10.1. INTRODUCTION
A fan is a device that utilizes the mechanical energy of a rotating impeller to produce both
movement of the air and an increase in its total pressure. The great majority of fans used in
mining are driven by electric motors, although internal combustion engines may be employed,
particularly as a standby on surface fans. Compressed air or water turbines may be used to drive
small fans in abnormally gassy or hot conditions, or where an electrical power supply is
unavailable.
In Chapter 4, mine fans were classified in terms of their location, main fans handling all of the air
passing through the system, booster fans assisting the through-flow of air in discrete areas of the
mine and auxiliary fans to overcome the resistance of ducts in blind headings. Fans may also be
classified into two major types with reference to their mechanical design.
A centrifugal fan resembles a paddle wheel. Air enters near the centre of the wheel, turns through
a right angle and moves radially outward by centrifugal action between the blades of the rotating
impeller. Those blades may be straight or curved either backwards or forwards with respect to the
direction of rotation. Each of these designs produces a distinctive performance characteristic.
Inlet and/or outlet guide vanes may be fitted to vary the performance of a centrifugal fan.
An axial fan relies on the same principle as an aircraft propeller, although usually with many more
blades for mine applications. Air passes through the fan along flowpaths that are essentially
aligned with the axis of rotation of the impeller and without changing their macro-direction.
However, later in the chapter we shall see that significant vortex action may be imparted to the
air. The particular characteristics of an axial fan depend largely on the aerodynamic design and
number of the impeller blades together with the angle they present to the approaching airstream.
Some designs of axial impellers allow the angle of the blades to be adjusted either while
stationary or in motion. This enables a single speed axial fan to be capable of a wide range of
duties. Axial fan impellers rotate at a higher blade tip speed than a centrifugal fan of similar
performance and, hence, tend to be noisier. They also suffer from a pronounced stall
characteristic at high resistance. However, they are more compact, can easily be combined into
series and parallel configurations and can produce reversal of the airflow by changing the
direction of impeller rotation, although at greatly reduced performance. Both types of fan are used
as main fans for mine ventilation systems while the axial type is favoured for underground
locations.
In this chapter, we shall define fan pressures and examine some of the basic theory of fan
design, the results of combining fans in series and parallel configurations, the theory of fan testing
and booster fan installations.
10.2. FAN PRESSURES

A matter that has often led to confusion is the way in which fan pressures are defined. In Section
2.3.2. we discussed the concepts of total, static and velocity pressures as applied to a moving
fluid. That section should be revised, if necessary, before reading on. While we use those
concepts in the definitions of fan pressures, the relationships between the two are not
immediately obvious. The following definitions should be studied with reference to Figure 10.1(a)
until they are clearly understood.
Fan total pressure, FTP, is the increase in total pressure, pt, (measured by facing pitot
tubes) across the fan,
FTP = pt2 pt1
(10.1)
Fan velocity pressure, FVP, is the average velocity pressure at the fan outlet only,
pv2 = pt2 - ps2
2
Chapter 10. Fans
Fan static pressure, FSP, is the difference between the fan total pressure and fan velocity
pressure, or
FSP = FTP - FVP
= pt2 - pt1 - (pt2 - ps2) = ps2 pt1
(10.2)
(10.3)
The reason for defining fan velocity pressure in this way is that the kinetic energy imparted by the
fan and represented by the velocity pressure at outlet has, traditionally, been assumed to be a
loss of useful energy. For a fan discharging directly to atmosphere this is, indeed, the case. As
the fan total pressure, FTP, reflects the full increase in mechanical energy imparted by the fan,
the difference between the two, i.e. fan static pressure has been regarded as representative of
the useful mechanical energy applied to the system.
The interpretations of fan pressures that are most convenient for network planning are further
illustrated on Figure 10.1. In the case of a fan located within an airway or ducted at both inlet and
outlet (Figure 10.1(a)), the fan static pressure, FSP, can be measured directly between a total
(facing) tube at inlet and a static (side) tapping at outlet. A study of the diagram and equation
(10.3) reveals that this is, indeed, the difference between FTP and FVP.
Figure 10.1 (b) shows the situation of a forcing fan drawing air from atmosphere into the system.
A question that arises is where to locate station 1, i.e. the fan inlet. It may be considered to be
(i) immediately in front of the fan
(ii) at the entrance to the inlet cone, or
(iii) in the still external atmosphere
These three positions are labelled on Figure 10.1 (b). If location(i) is chosen then the frictional
and shock losses incurred as the air enters and passes through the cone must be assessed
separately. At location (ii) the fan and inlet cone are considered as a unit and only the shock loss
at entry requires additional treatment. However, if location (iii) is selected then the fan, inlet cone
and inlet shock losses are all taken into account. It is for this reason that location (iii) is preferred
for the purposes of ventilation planning. Figure 10.1 (b) shows the connection of gauges to
indicate the fan pressures in this configuration.
The same arguments apply for a fan that exhausts to atmosphere (Figure 10.1 (c)). If the outlet
station is taken to be in the still external atmosphere then the fan velocity pressure is zero and the
fan total and fan static pressures become equal. In this configuration the fan total (or static)
pressure takes into account the net effects of the fan, frictional losses in the outlet cone and the
kinetic energy loss at exit.
During practical measurements, it is often found that turbulence causes excessive fluctuations on
the pressure gauge when total pressures are measured directly using a facing pitot tube. In such
cases, it is preferable to measure static pressure from side tappings and to add, algebraically, the
velocity pressure in order to obtain the total pressure. The mean velocity can be obtained as
flowrate divided by the appropriate cross-sectional area. Particular care should be taken with
regard to sign. In the case of an exhausting fan, (Figure 10.1 (c)) the static and velocity pressures
at the fan inlet have opposite signs.
Chapter 10. Fans
FSP
pt1
ps2
fan
pt1
ps2
pt2
FTP
FVP = pv2
(a) Fan with inlet and outlet ducts
FSP
(ii)
(i)
fan
1 (iii)
pv1 = 0
FTP
FVP
(b) Fan and inlet unit in a forcing system
fan
2
FTP and
FSP
FVP = pv2 = 0
(c) Fan and outlet unit in an exhausting system
Figure 10.1 Illustrations of fan pressures.
Chapter 10. Fans
Fan manufacturers usually publish characteristic curves in terms of fan static pressure rather than
the fan total pressure. In addition to being more useful for ventilation planning, this is
understandable as manufacturers may have no control over the types of inlet and outlet duct
fittings or the conditions at entry or exit to inlet/outlet cones. Where fan velocity pressures are
quoted then they are normally referred to a specific outlet location, usually either at the fan hub or
at the mouth of an evasee.
The question remains on which fan pressure should be used when a fan is included in the route
of a pressure survey. The simple answer is that for main surface fans it is fan static pressure
(FSP) that should be employed. For a main forcing fan the FSP is given by the gauge static
pressure at the inby side of the fan (Figure 10.1(b)). But for a main exhausting fan the FSP is
given by the gauge total pressure at the inby side of the fan (Figure 10.1(c)). This apparent
anomaly arises from the way in which fan pressures are defined and is further explained in
Appendix A10.1.
10.3. IMPELLER THEORY AND FAN CHARACTERISTIC CURVES
An important aspect of subsurface ventilation planning is the specification of pressure-volume
duties required of proposed main or booster fans. The actual choice of particular fans is usually
made through a process of perusing manufacturer's catalogues of fan characteristic curves,
negotiation of prices and costing exercises (Section 9.5). The theory of impeller design that
underlies the characteristic behaviour of differing fan types is seldom of direct practical
consequence to the underground ventilation planner. However, a knowledge of the basics of that
theory is particularly helpful in discussions with fan manufacturers and in comprehending why
fans behave in the way that they do.
This section of the book requires an elementary understanding of vector diagrams. Initially, we
shall assume incompressible flow but will take compressibility of the air into account in the later
section on fan performance (Section 10.6.1).
10.3.1. The centrifugal impeller
Figure 10.2 illustrates a rotating backward bladed centrifugal impeller. The fluid enters at the
centre of the wheel, turns through a right angle and, as it moves outwards radially, is subjected to
centrifugal force resulting in an increase in its static pressure. The dotted lines represent
flowpaths of the fluid relative to the moving blades. Rotational and radial components of velocity
are imparted to the fluid. The corresponding outlet velocity pressure may then be partially
converted into static pressure within the surrounding volute or fan casing.
At any point on a flowpath, the velocity may be represented by vector components with respect to
either the moving impeller or to the fan casing. The vector diagram on Figure 10.2 is for a particle
of fluid leaving the outlet tip of an impeller blade. The velocity of the fluid relative to the blade is W
and has a vector direction that is tangential to the blade at its tip. The fluid velocity also has a
vector component in the direction of rotation. This is equal to the tip (peripheral) velocity and is
shown as u. The vector addition of the two, C, is the actual or absolute velocity.), The radial (or
meridianal) component of velocity, Cm is also shown on the vector diagram.
Chapter 10. Fans
evasee
volute
Figure 10.2 Idealized flow through a backward bladed centrifugal impeller.
10.3.1.1. Theoretical pressure developed by a centrifugal impeller.

A more detailed depiction of the inlet and outlet vector diagrams for a centrifugal impeller is given
on Figure 10.3. It is suggested that the reader spend a few moments examining the key on Figure
10.3 and identifying corresponding elements on the diagram.
Chapter 10. Fans
W2
C2
W2
Cm2
Cu2
u2
W1
C1
u1
e
a
W1
Ff
pf
c
g
Fb
pb
b
Subscript 1: Inlet
Subscript 2: Outlet
:
u:
Cm :
:
pf :
Ff :
C:
W:
Cu :
Absolute fluid velocity (m/s)

Fluid velocity relative to vane )m/s)
Peripheral component of fluid velocity (m/s)
pb :
Fb :
Pressure on back of vane (Pa)

Shear resistance on back of vane (N/m2)
Angular velocity (radians/s)

Peripheral speed of blade tip (m/s)
Radial component of fluid velocity (m/s)
Vane angle
Pressure on front of vane (Pa)
Shear resistance on front of vane (N/m2)
Figure 10.3 Velocities and forces on a centrifugal impeller.
Chapter 10. Fans
In order to develop an expression for the theoretical pressure developed by the impeller, we apply
the principle of angular momentum to the mass of fluid moving through it.
If a mass, m, rotates about an axis at a radius, r, and at a tangential velocity, v, then it has an
angular momentum of mrv. Furthermore, if the mass is a fluid that is continuously being replaced
then it becomes a mass flow, dm/dt, and a torque, T, must be maintained that is equal to the
corresponding continuous rate of change of momentum
T =
dm
(rv )
dt
(10.4)
Nm or J
In the case of the centrifugal impeller depicted in Figure 10.3, the peripheral component of fluid
velocity is Cu. Hence the torque becomes
T =
dm
(r C u )
dt
(10.5)
Nm or J
Consider the mass of fluid filling the space between two vanes and represented as abcd on
Figure 10.3. At a moment, dt, later it has moved to position efgh. The element abfe leaving the
impeller has mass dm and is equal to the mass of the element cdhg entering the impeller during
the same time. The volume represented by abgh has effectively remained in the same position
and has not, therefore, changed its angular momentum. The increase in angular momentum is
that due to the elements abfe and cdhg. Then, from equation (10.5) applied across the inlet and
outlet locations,
T =
dm
[r 2 C u 2 r1 C u1 ]
dt
(10.6)
Extending the flow to the whole impeller instead of merely between two vanes gives dm/dt as the
total mass flow, or
dm
dt
where Q
and
giving
kg
s
= Q
volume flow (m3/s)

fluid density (kg/m3)
=
=
T = Q [r 2 C u 2 r1 C u1 ]
(10.7)
Now the power consumed by the impeller, Pow is equal to the rate of doing mechanical work,
Pow
(10.8)
where
speed of rotation (radians/s)
giving Pow
Q [ r2 Cu2 r1 Cu1 ]
But
and
=
=
u2
u1
r2
r1
=
=
Hence Pow = Q [u 2 C u 2 u1 C u1 ]
(10.9)
tangential velocity at outlet

tangential velocity at inlet.
W
(10.10)
Chapter 10. Fans
The power imparted by a fan impeller to the air was given by equation (5.56) as pft Q
where pft = rise in total pressure across the fan.
In the absence of frictional or shock losses, pft Q must equal the power consumed by the impeller,
Pow. Hence
p ft = (u 2 C u 2 u1 C u1 )
(10.11)
Pa
This relationship gives the theoretical fan total pressure and is known as Euler's equation.
The inlet flow is often assumed to be radial for an ideal centrifugal impeller, i.e. Cu1 = 0, giving
p ft = u 2 C u 2
(10.12)
Pa
Euler's equation can be re-expressed in a manner that is more amenable to physical

interpretation. From the outlet vector diagram
W2
= C m2
+ (u 2 C u 2 ) 2
= C m2
+ u2
2u 2 C u 2 + C u 2
or
2 u 2 Cu 2 = u 2
= u2
W2
W2
+ (C m 2
+ C2
+ Cu 2 )
(Pythagorus)
Similarly for the inlet,

2
2 u 1 C u1 = u 1 W 1 + C 1
Euler's equation (10.11) then becomes

u 2 u12
p ft = 2
2
centrifugal
effect
W2
W1
2
effect of
relative
velocity
Gain in static pressure
C2
C1
2
Pa
(10.13)
change in
kinetic energy
+ Gain in velocity
pressure
10.3.1.2. Theoretical characteristic curves for a centrifugal impeller

Euler's equation may be employed to develop pressure-volume relationships for a centrifugal
impeller. Again, we must first eliminate the Cu term. From the outlet vector diagram on Figure
10.3,
tan 2 =
C m2
u 2 Cu 2
Chapter 10. Fans
giving
Cu 2 = u 2
C m2
tan 2
For radial inlet conditions, Euler's equation (10.12) then gives
Cm2
p ft = u 2 C u 2 = u 2 u 2
tan 2
But
Cm2 =
Pa
(10.14)
Pa
(10.15)
volume flowrate
Q
=
flow area at impeller outlet
a2
Equation (10.14) becomes

p ft = u 2
u2 Q
tan 2 a 2
For a given impeller rotating at a fixed speed and passing a fluid of known density, , u, a and
are all constant, giving
p ft = A B Q
Pa
where constants
A =
and
B =
(10.16)
u22
u2
a 2 tan 2
The flowrate, Q, and, hence, the pressure developed vary with the resistance against which the
fan acts. Equation (10.16) shows that if frictional and shock losses are ignored, then fan pressure
varies linearly with respect to the airflow.
We may apply this relationship to the three types of centrifugal impeller:
Radial bladed
2 = 90 and tan 2 = infinity
giving B = 0
Then
pft = constant A = u22
i.e. theoretically, the pressure remains constant at all flows [Figure 10.4 (a)]
Backward bladed
2 < 90 ,
tan 2 > 0
pft = A - BQ
i.e. theoretically, the pressure falls with increasing flow.
Forward bladed
2 > 90,
tan 2 < 0
pft = A + BQ
i.e. theoretically, pressure rises linearly with increasing flow.
The latter and rather surprising result occurs because the absolute velocity, C2 is greater than the
impeller peripheral velocity, u2, in a forward bladed impeller. This gives an impulse to the fluid
which increases with greater flowrates. (In an actual impeller, friction and shock losses more than
counteract the effect.).
10
Chapter 10. Fans
The theoretical pressure-volume characteristic curves are shown on Figure 10.4 (a).
The theoretical relationship between impeller power and airflow may also be gained from
equation (10.16).
Pow = pft Q = AQ - BQ2
(10.17)
The three power-volume relationships then become:

Radial bladed
B = 0
and
Pow = AQ (linear)
Backward bladed
B > 0
and
Pow = AQ - BQ2 (falling parabola)
Forward bladed
B < 0
and
Pow = AQ + BQ2 (rising parabola).
The theoretical power-volume curves are shown on Figure 10.4 (b). Forward bladed fans are
capable of delivering high flowrates at fairly low running speeds. However, their high power
demand leads to reduced efficiencies. Conversely, the relatively low power requirement and high
efficiencies of backward bladed impellers make these the preferred type for large centrifugal fans.
10.3.1.3. Actual characteristic curves for a centrifugal impeller
The theoretical treatment of the preceding subsections lead to linear pressure-volume
relationships for radial, backward and forward bladed centrifugal impellers. In an actual fan, there
are, inevitably, losses which result in the real pressure-volume curves lying below their theoretical
counterparts. In all cases, friction and shock losses produce pressure-volume curves that tend
toward zero pressure when the fan runs on open circuit, that is, with no external resistance.
Figure 10.5 shows a typical pressure-volume characteristic curve for a backward bladed
centrifugal fan. Frictional losses occur due to the viscous drag of the fluid on the faces of the
vanes. These are denoted as Ff and Fb on Figure 10.3. A diffuser effect occurs in the diverging
area available for flow as the fluid moves through the impeller. This results in a further loss of
available energy. In order to transmit mechanical work, the pressure on the front face of a vane,
pf is necessarily greater than that on the back, pb. A result of this is that the fluid velocity close to
the trailing face is higher than that near the front face. These effects result in an asymmetric
distribution of fluid velocity between two successive vanes at any given radius and produce an
eddy loss. It may also be noted that at the outlet tip, the two pressures pf and pb must become
equal. Hence, although the tip is most important in its influence on the outlet vector diagram, it
does not actually contribute to the transfer of mechanical energy. The transmission of power is
not uniform along the length of the blade.
The shock (or separation) losses occur particularly at inlet and reflect the sudden turn of near 90
as the fluid enters the eye of the impeller. In practice, wall effects impart a vortex to the fluid as it
approaches the inlet. By a suitable choice of inlet blade angle, 1, (Figure 10.3) the shock losses
may be small at the optimum design flow. An inlet cone at the eye of the impeller or fixed inlet
and outlet guide vanes can be fitted to reduce shock losses.
In the development of the theoretical pressure and power characteristics, we assumed radial inlet
conditions. When the fluid has some degree of pre-rotation, the flow is no longer radial at the inlet
to the impeller. The second term in Euler's equation (10.11) takes a finite value and, again,
results in a reduced fan pressure at any given speed of rotation.
11
Chapter 10. Fans
Fan total pressure pft
forward
radial
backward
0
0
Airflow Q
10
12
(a) Pressure-volume theoretical characteristics

2 00
Power Pow
forward
radial
backward
0
0
1 0
1 2
1 4
1 6
1 8
2 0
Airflow Q
(b) Power-volume theoretical characteristics
Figure 10.4 Theoretical characteristic curves for a centrifugal fan
12
Chapter 10. Fans
Figure 10.5 Effect of losses on the pressure-volume characteristic of a

backward bladed centrifugal fan.
The combined effect of these losses on the three types of centrifugal impeller is to produce the
characteristic curves shown on Figure 10.6. The non-overloading power characteristic, together
with the steepness of the pressure curve at the higher flows, are major factors in preferring the
backward impeller for large installations.
backward
radial
forward
forward
radial
backward
Figure 10.6 Actual pressure and shaft power characteristics for centrifugal impellers.
13
Chapter 10. Fans
10.3.2. The axial impeller

Axial fans of acceptable performance did not appear until the 1930's. This was because of a lack
of understanding of the behaviour of airflow over axial fan blades. The early axial fans had simple
flat angled blades and produced very poor efficiencies. The growth of the aircraft industry led to
intensive studies of aerofoils for the wings of aeroplanes. In this section, we shall discuss briefly
the characteristics of aerofoils. This facilitates our comprehension of the behaviour of axial fans.
The theoretical treatment of axial impellers may be undertaken from the viewpoint of a series of
aerofoils or by employing either vortex or momentum theory. We shall use the latter in order to
remain consistent with our earlier analysis of centrifugal impeller although aerofoil theory is
normally also applied in the detailed design of axial impellers.
10.3.2.1. Aerofoils
When a flat stationary plate is immersed in a moving fluid such that it lies parallel with the
direction of flow then it will be subjected to a small drag force in the direction of flow. If it is then
inclined upward at a small angle, , to the direction of flow then that drag force will increase.
However, the deflection of streamlines will cause an increase in pressure on the underside and a
decrease in pressure on the top surface. This pressure differential results in a lifting force. On an
aircraft wing and in an axial fan impeller, it is required to achieve a high lift, L, without unduly
increasing the drag, D. The dimensionless coefficients of lift, CL, and drag, CD, for any given
section are defined in terms of velocity heads
L =
C2
A CL
2
(10.18)
D =
C2
A CD
2
(10.19)
where C
A
=
=
fluid velocity (m/s) and

a characteristic area usually taken as the underside of the plate or aerofoil (m2)
The ratio of CL /CD is considerably enhanced if the flat plate is replaced by an aerofoil. Figure 10.7
illustrates an aerofoil section. Selective evolution in the world of nature suggests that it is no
lift
resultant
drag
angle of attack
Figure 10.7 An aerofoil section.
14
Chapter 10. Fans
coincidence that the aerofoil shape is decidedly fish-like. The line joining the facing and trailing
edges is known as the chord while the angle of attack, , is defined as that angle between the
chord and the direction of the approaching fluid.
A typical behaviour of the coefficients of lift and drag for an aerofoil with respect to angle of attack
is illustrated on Figure 10.8. Note that the aerofoil produces lift and a high CL/CD ratio at zero
angle of attack. The coefficient of lift increases in a near-linear manner. However, at an angle of
attack usually between 12 and 18, breakaway of the boundary layer occurs on the upper
surface. This causes a sudden loss of lift and an increase in drag. In this stall condition, the
formation and propagation of turbulent vortices causes the fan to vibrate excessively and to
produce additional low frequency noise. A fan should never be allowed to run continuously in this
condition as it can cause failure of the blades and excessive wear in bearings and other
transmission components.
Figure 10.8 Typical behaviour of lift and drag coefficients for an aerofoil.
10.3.2.2. Theoretical pressure developed by an axial fan

Consider an imaginary cylinder coaxial with the drive shaft and cutting through the impeller
blades at a constant radius. Depicting an impression of the cut blades on two dimensional paper
produces a drawing similar to that of Figure 10.9 (a). For simplicity each blade is shown simply as
a curved vane rather than an aerofoil section.
15
Chapter 10. Fans
Axis of
rotation
W1
C1
p1s
u
rotation
u
p2s
(a) Inlet and outlet velocity

diagrams.
C2
C1
W2
u
Cu2
C2
W1
C1
u
Cu2
(b) Inlet and outlet velocity

diagrams combined.
W2
u
Cu2
Figure 10.9 Velocity diagrams for an axial impeller without guide vanes.
The air approaches the moving impeller axially at a velocity, C1. At the optimum design point, the
C1 vector combines with the blade velocity vector, u, to produce a velocity relative to the blade
W1, and which is tangential to the leading edge of the blade. At the trailing edge, the air leaves at
a relative velocity W2,, which also combines with the blade velocity to produce the outlet absolute
velocity, C2. This has a rotational component, Cu2, imparted by the rotation of the impeller. The
initial axial velocity, C1, has remained unchanged as the impeller has no component of axial
velocity in an axial fan.
16
Chapter 10. Fans
Figure 10.9 (b) shows how the inlet and outlet velocity diagrams can be combined. As W1 and W2
are both related to the same common blade velocity, u, the vector difference between the two,
W1 - W2, must be equal to the final rotational component Cu2.
As the axial velocity, C1 is the same at outlet as inlet, it follows that the increase in total pressure
across the impeller is equal to the rise in static pressure.
p ft = p 2s p1s
Now let us consider again the relative velocities, W1 and W2. Imagine for a moment that the
impeller is standing still. It would impart no energy and Bernouilli's equation (2.16) tells us that the
increase in static pressure must equal the decrease in velocity pressure (in the absence of
frictional losses). Hence
2
W 2
W2
p ft = p 2s p1s = 1
2
2
Pa
(10.20)
Applying Pythagorus' Theorem to Figure 10.9 (b) gives

W1
= C1 + u 2
= C1
+ (u C u 2 ) 2
and
W2
giving
p ft
i.e.
p ft
= u Cu 2
[2C
u2
u Cu 2
Cu 2
2
(10.21)
Comparison with equation (10.12) shows that this is the pressure given by Euler's equation, less
the velocity head due to the final rotational velocity.
The vector diagram for an axial impeller with inlet guide vanes is given on Figure 10.10. In order
to retain subscripts 1 and 2 for the inlet and outlet sides of the moving impeller, subscript 0 is
employed for the air entering the inlet guide vanes. At the optimum design point the absolute
velocities, Co and C2 should be equal and axial, i.e. there should be no rotational components of
velocity at either inlet or outlet of the guide vane/impeller combination. This means that any vortex
action imparted by the guide vanes must be removed by an equal but opposite vortex action
imparted by the impeller.
We can follow the process on the vector diagram on Figure 10.10 using the labelled points a, b, c,
d, and e.
(a) The air arrives at the entrance to the guide vanes with an axial velocity, Co, and no rotational
component (point a)
(b) The turn on the inlet guide vanes gives a rotational component, Cu1 to the air. The axial
component remains at Co. Hence, the vector addition of the two results in the absolute velocity
shown as C1 (point b).
17
Chapter 10. Fans
C0
inlet guide vanes
1
impeller
u
2
C2
C0
and
C2
C1
e
a
Cu1
W2
W1
u
u
Cu2
Figure 10.10 Velocity diagrams for an axial impeller with inlet guide vanes.
18
Chapter 10. Fans
(c) In order to determine the velocity of the air relative to the moving impeller at location (1), we
must subtract the velocity vector of the impeller, u. This brings us to position c on the vector
diagram and an air velocity of W1 relative to the impeller.
(d) The turn on the impeller imparts a rotational component Cu2 to the air. The velocity of the air
relative to the impeller is thus reduced to W2 and we arrive at point d.
(e) In order to determine the final absolute velocity of the air, we must add the impeller velocity, u.
This will take us to point e on the vector diagram (coinciding with point a), with no remaining
rotational component provided that Cu1 = Cu2.
It follows that the absolute velocities at inlet and outlet, Co and C2, must be both axial and equal.
To determine the total theoretical pressure developed by the system, consider first the stationary
inlet guide vanes (subscript g). From Bernoulli's equation with no potential energy term
pg
C 2 C12
= 0
Applying Pythagorus' theorem to the vector diagram of Figure 10.10 gives this to be
pg
C u1
2
(10.22)
and represents a pressure loss caused by rotational acceleration across the guide vanes.
The gain in total pressure across the impeller (subscript i) is given as
pi
(W
2
1
W2
(10.23)
(see, also, equation (10.20))
Using Pythagorus on Figure 10.10 again, gives this to be
pi
(C
2
(C
u2
+ (u + C u 2 ) 2 (C o + u 2 )
2
+ 2 Cu 2 u
But, as Cu1 = Cu2, this can also be written as

pi =
(C
2
2
u1
+ 2 Cu 2 u
Pa
(10.24)
Now the total theoretical pressure developed by the system, pft must be the combination pg + pi.
Equations (10.22) and (10.24) give
p ft
= pg + pi
= Cu 2 u
Pa
(10.25)
19
Chapter 10. Fans
Once again, we have found that the theoretical pressure developed by a fan is given by Euler's
equation (10.12). Furthermore, comparison with equation (10.21) shows that elimination of the
residual rotational component of velocity at outlet (by balancing Cu1 and Cu2) results in an
increased fan pressure when guide vanes are employed.
The reader might wish to repeat the analysis for outlet guide vanes and for a combination of inlet
and outlet guide vanes. In these cases, Euler's equation is also found to remain true. Hence,
wherever the guide vanes are located, the total pressure developed by an axial fan operating at
its design point depends only upon the rotational component imparted by the impeller, Cu2, and
the peripheral velocity of the impeller, u.
It is obvious that the value of u will increase along the length of the blade from root to tip. In order
to maintain a uniform pressure rise and to inhibit undesirable cross flows, the value of Cu2 in
equation (10.25) must balance the variation in u. This is the reason for the "twist" that can be
observed along the blades of a well-designed axial impeller.
10.3.2.3. Actual characteristic curves for an axial fan
The losses in an axial fan may be divided into recoverable and non-recoverable groups. The
recoverable losses include the vortices or rotational components of velocity that exist in the
airflow leaving the fan. We have seen that these losses can be recovered when operating at the
design point by the use of guide vanes. However, as we depart from the design point, swirling of
the outlet air will build up.
The non-recoverable losses include friction at the bearings and drag on the fan casing, the hub of
the impeller, supporting beams and the fan blades themselves. These losses result in a transfer
from mechanical energy to heat which is irretrievably lost in its capacity for doing useful work.
Figure 10.11 is an example of the actual characteristic curves for an axial fan. The design point,
C, coincides with the maximum efficiency. At this point the losses are at a minimum. In practice,
the region A to B on the pressure curve would be acceptable. Operating at low resistance, i.e. to
the right of point B, would not draw excessive power from the motor as the shaft power curve
shows a non-overloading characteristic. However, the efficiency decreases rapidly in this region.
The disadvantage of operating at too high a resistance, i.e. to the left of point A is, again, a
decreasing efficiency but, more importantly, the danger of approaching the stall point, D. There is
a definite discontinuity in the pressure curve at the stall point although this is often displayed as a
smoothed curve. Indeed, manufacturers' catalogues usually show characteristic curves to the
right of point A only. In the region E to D, the flow is severely restricted. Boundary layer
breakaway takes place on the blades (Section 10.3.2.1) and centrifugal action occurs producing
recirculation around the blades.
A fixed bladed axial fan of constant speed has a rather limited useful range and will maintain
good efficiency only when the system resistance remains sensibly constant.. This can seldom be
guaranteed over the full life of a main mine fan. Fortunately, there are a number of ways in which
the range of an axial fan can be extended:(a) The angle of the blades may be varied. Many modern axial fans allow blade angles to be
changed, either when the rotor is stationary or while in motion. The latter is useful if the
fan is to be incorporated into an automatic ventilation control system. Figure 10.12 gives
an example of the characteristic curves for an axial fan of variable blade angle. The
versatility of such fans gives them considerable advantage over centrifugal fans.
20
Chapter 10. Fans
(b) The angle of the inlet and/or outlet guide vanes may also be varied, with or without
modification to the impeller blade angle. The effect is similar to that illustrated by Figure
10.12.
(c) The pitch of the impeller may be changed by adding or removing blades. The impeller
must, of course, remain dynamically balanced. This technique can result in substantial
savings in power during time periods of relatively light load.
(d) The speed of the impeller may be changed either by employing a variable speed motor or
by changing the gearing between the motor and the fan shaft. The majority of fans are
driven by A.C. induction motors at a fixed speed. Variable speed motors are more
expensive although they may produce substantial savings in operating costs. Axial fans
may be connected to the motor via flexible couplings which allow a limited degree of
angular or linear misalignment. Speed control may be achieved by hydraulic couplings,
or, in the case of smaller fans, by V-belt drives with a range of pulley sizes.
stall
C
pressure
B
efficiency
shaft power
Airflow
Figure 10.11 Typical characteristic curves for an axial fan.
21
Chapter 10. Fans
Figure 10.12 Example of a set of characteristic curves for an axial fan with variable blade angle.
10.4 FAN LAWS

The performance of fans is normally specified as a series of pressure, efficiency and shaft power
characteristic curves plotted against airflow for specified values of rotational speed, air density and
fan dimensions. It is, however, convenient to be able to determine the operating characteristic of
the fan at other speeds and air densities. It is also useful to be able to use test results gained from
smaller prototypes to predict the performance of larger fans that are geometrically similar.
Euler's equation and other relationships introduced in Section 10.3 can be employed to establish
a useful set of proportionalities known as the fan laws.
10.4.1. Derivation of the fan laws
Fan pressure
From Euler's equation (10.12 and 10.25)
p ft
= u Cu 2
22
Chapter 10. Fans
However, both the peripheral speed, u, and the rotational component of outlet velocity, Cu2
(Figures 10.3 and 10.10) vary with the rotational speed, n, and the impeller diameter, d. Hence,
pft (nd ) (nd )
or
p ft n 2 d 2
(10.26)
where means 'proportional to.'

Airflow
For a centrifugal fan, the radial flow at the impeller outlet is given as
Q
where d
and Cm2
=
=
area of flow at impeller outlet x Cm2

d impeller width Cm2
= diameter of impeller
= radial velocity at outlet (Figure 10.3)
However, for geometric similarity between any two fans

impeller width d
giving
Q
d2 Cm2
Again, for geometric similarity, all vectors are proportional to each other. Hence
Cm2 u n d
giving
Q
nd
(10.27)
A similar argument applies to the axial impeller (Figure 10.10). At outlet,

Q =
d2
4
C2
where C2 = axial velocity

As before, vectors are proportional to each other for geometric similarity,
C 2 u nd
giving, once again,
Q n d3
Density
From Euler's equation (10.12 or 10.25) it is clear that fan pressure varies directly with air density
p ft
(10.28)
However, we normally accept volume flow, Q, rather than mass flow as the basis of flow
measurement in fans. In other words, if the density changes we still compare operating points at
corresponding values of volume flow.
23
Chapter 10. Fans
Air Power
For incompressible flow, the mechanical power transmitted from the impeller to the air is given as
Pow
pft Q
Employing proportionalities (10.26) and (10.27) gives

n3 d5
Pow
(10.29)
10.4.2. Summary of fan laws

In the practical utilization or design of fans, we are normally interested in varying only one of the
independent variables (speed, air density, impeller diameter) at any given time while keeping the
other two constant. The fan laws may then be summarized as follows:
Variable speed (n)
Variable diameter (d)
n2
d2
d3
Pow
n3
Pow
Variable air density ()
d5
Q fixed
Pow
These laws can be applied to compare the performance of a given fan at changed speeds or air
densities, or to compare the performance of different sized fans provided that those two fans are
geometrically similar.
If the two sets of operating conditions, or the two geometrically similar fans are identified by
subscripts a and b, then the fan laws may be written, more generally, as the following equations
p ft ,a
p ft ,b
na
nb
da
db
a
b
(10.30)
Qa
n d
= a a3
Qb
nb d b
Pow ,a
Pow ,b
where pft
Q
Pow
n
d
and
=
=
=
=
=
=
na
nb
(10.31)
da
db
a
b
(10.32)
fan pressure (applies for both total and static pressure)

airflow
airpower
rotational speed
impeller diameter
air density
24
Chapter 10. Fans
Example.
Characteristic curves are available for a fan running at 850 rpm and passing air of inlet density
1.2 kg/m3. Readings from the curves indicate that at an airflow of 150 m3/s, the fan pressure is
2.2 kPa and the shaft power is 440 kW. Assuming that the efficiency remains unchanged,
calculate the corresponding points if the fan is run at 1100 rpm in air of density 1.1 kg/m3.
Solution.
As it is the same fan that is to be used in differing conditions, there is no change in impeller
diameter or overall geometry.
Equation (10.30) gives the new pressure to be
1100
p f = 2 .2
850
1. 1
= 3.377
1 .2
kPa
From equation (10.31), the new volume flow is
Q = 150
1100
= 194.1
850
m 3 /s
Equation (10.32) refers to the airpower rather than the shaft delivered to the impeller. However, if
we assume that the impeller efficiency remains the same at corresponding points on the
characteristic curves then we may also use that equation for shaft power.
3
Pow
1 .1
1100
= 440
= 874 kW

1 .2
850
By treating a series of points from the original curves in this way, a second set of characteristic
curves can be produced that is applicable to the new conditions.
10.5. FANS IN COMBINATION
Many mines or other subsurface facilities have main and, perhaps, booster fans sited at differing
locations; for example, there may be two or more upcast shafts, each with its own surface
exhausting fan. With the advent of simulation programs for ventilation network analysis (Chapters
7 and 9), the relevant pressure-volume characteristic data may be entered separately for each
fan unit. The system resistance offered to each of those fans becomes a function not only of the
network geometry but also the location and operating characteristics of other fans in the system.
That resistance is sometimes termed the effective resistance 'seen' by the fan.
There are situations in which it is advantageous to combine fans either in series or in parallel at a
single location. Such combinations enable a wide spectrum of pressure-volume duties to be
attained with only a limited range of fan sizes. In general, fans may be connected in series in
order to pass a given airflow against an increased resistance, while a parallel combination allows
the flow to be increased for any given resistance. Although ventilation network programs can
allow each fan to be entered separately, it is sometimes more convenient to produce a pressurevolume characteristic curve that represents the combined unit.
25
Chapter 10. Fans
10.5.1. Fans in series

Figure 10.13 shows two fans, a and b, located in series within a single duct or airway. The
corresponding pressure-volume characteristics and the effective resistance curve are also shown.
The characteristic curve for the combination is obtained simply by adding the individual fan
pressures for each value of airflow.
12
b
Q
10
pa
Fan Pressure p
combined fans
pb
pc
fan b
fan a
pc = pa + pb
pb
pa
effective
resistance curve
0
0
Airflow Q
Q8
10
12
Figure 10.13 Characteristic curves for two fans connected in series.
The effective operating point is located at C, where the resistance curve intersects the combined
characteristic. Fans a and b both pass the same airflow, Q, but develop pressures pa and pb
respectively The individual operating points are shown as A and B. For three or more fans, the
process of adding fan pressures remains the same. However, if the change in density through the
26
Chapter 10. Fans
combination becomes significant then the fan laws (Section 10.4.2) should be employed to
correct the individual characteristic curves.
As shown in Figure 10.13, the individual fans need not have identical characteristic curves.
However, if one fan is considerably more powerful than the other, or if the system resistance falls
to a low level, then the impeller of the weaker unit may be driven in turbine fashion by its stronger
companion. The weaker fan then becomes an additional resistance on the system. It is usual to
employ identical fans in combination.
10.5.2. Fans in parallel
For fans that are combined in parallel, the airflows are added for any given fan pressure in order
to obtain the combined characteristic curve. As shown in Figure 10.14, fans a and b pass airflows
Qa and Qb, respectively, but at the same common pressure, p.
11
10
Qc
Qa
Qb
Fan Pressure p
fan b
fan a
7
combined
fans
0
0
Qa
Qb
10
11
Airflow Q
12
Qc= Qa+Qb
13
14
15
16
17
18
19
20
Figure 10.14 Characteristic curves for two fans connected in parallel.
The operating point for the complete unit occurs at C with the individual operating points for fans
a and b at A and B respectively. Here again, the fans need not necessarily be identical. However,
care must be taken to ensure that the operating points A and B do not move too far up their
respective curves. This is particularly important in the case of axial fans because of their
pronounced stall characteristic. In practice, before this condition is reached, the fans may exhibit
27
Chapter 10. Fans
a noticeable "hunting" effect. For these reasons, the maximum variation in system resistance that
is likely to occur should be investigated before installing fans in parallel.
Three or more fans may be combined in parallel, adding airflows to obtain the combined
characteristic curve. Although Figure 10.14 shows two fans with differing characteristics, it is
prudent to employ identical fans when connected in parallel. This will reduce the tendency for one
of them to approach stall conditions before the other. However, variations in the immediate
surroundings of ductwork or airway geometry often results in the fans operating against slightly
different effective resistances. Hence, even when identical fans are employed, it is usual for
measurements to indicate that they are producing slightly different pressure-volume duties. In the
case of fans located in separate ducts or airways that are connected in parallel, the resistance of
those ducts or airways may be taken into account by subtracting the frictional pressure losses in
each branch from the corresponding fan pressures. In these circumstances, a better approach is
to consider the fans as separate units for the purposes of network analysis.
An advantage of employing fans in parallel is that if one of them fails then the remaining fan(s)
continue to supply a significant proportion of the original flow. In the example shown on Figure
10.14, if fan a ceases to operate then the operating point for fan b will fall to position E, giving
some 70 per cent of the original airflow. The latter value depends upon the number of fans
employed, the shape of their pressure-volume characteristic curves and the provision of nonreturn baffles at the fan outlets.
Fans may be connected in any series/parallel configuration, adding pressures and airflows
respectively to obtain the combined characteristic curve. This is particularly useful for booster fan
locations. A mine may maintain an inventory of standard fans, combining them in series/parallel
combinations to achieve any desired operating characteristic.
10.6. FAN PERFORMANCE
Power is delivered to the drive shaft of a fan impeller from a motor (usually electric) and via a
transmission assembly. Losses occur in both the motor and transmission. For a properly
maintained electrical motor and transmission, some 95 per cent of the input electrical power may
be expected to appear as mechanical energy in the impeller drive shaft. The impeller, in turn,
converts most of that energy into useful airpower to produce both movement of the air and an
increase in pressure. The remainder is consumed by irreversible losses across the impeller and
in the fan casing (Sections 10.3.1.3 and 10.3.2.3) producing an additional increase in the
temperature of the air.
Impeller efficiency may be defined as
Airpower
Shaft power
(10.33)
while the overall efficiency of the complete motor/transmission/impeller unit is given as

Airpower
Motor input power
(10.34)
In the following subsection, we shall define other measures of fan efficiency. As there are several
different forms of fan efficiency it is prudent, when perusing manufacturers' literature, to ascertain
the basis of any quoted values of efficiencies. It is also important to use the same measure of
efficiency when comparing one fan with another. However, the one parameter that really matters
is the input power required to achieve the specified pressure-volume duty, as this is the factor
that dictates the operating cost of the fan.
28
Chapter 10. Fans
10.6.1. Compressibility, fan efficiency, and fan testing

In our previous analyses, we have defined airpower as the product pQ on the basis of
incompressible flow. That simplification will give an error of less than one percent when testing
fans that develop pressures up to 2.8 kPa. Unfortunately, fan operating costs at many large
mines are such that one per cent may represent a significant expenditure. Furthermore, mine fan
pressures exceeding 6 kPa are not uncommon. For these reasons, we should take air
compressibility into account when measuring fan performance.
Applying the steady-flow energy equation (3.25) to a fan gives
2
u1 u 2
2
+ (Z1 Z 2 ) g + W =
VdP + F
12
= (H 2 H 1 ) q12
J
kg
(10.35)
where subscripts 1 and 2 refer to the fan inlet and outlet respectively
W
=
Impeller shaft work (J/kg of air)
V
=
Specific volume of air (m3/kg)
P
=
Absolute (barometric) pressure (Pa)
F12
=
Frictional losses (J/kg)
H
=
Enthalpy (J/kg) and
q12
=
Heat added through fan casing (J/kg)
For the purposes of this analysis we shall assume that the change in elevation through the fan is
negligible, Z1 - Z2 = 0. Furthermore, we shall assume that the change in air velocity across the
fan is also negligible compared to other terms, u1 = u2. This latter assumption implies that the fan
total pressure, referred to here simply as pf, is equal to the increase in barometric pressure
across the fan, P2 P1. Then
2
W =
VdP + F
12
J
kg
= (H 2 H 1 ) q12
(10.36)
It is also reasonable to assume that the heat transferred from the surroundings through the fan
casing is small compared to the shaft work. The steady-flow energy equation then simplifies to
the adiabatic equation
2
W =
VdP + F
12
= (H 2 H 1 )
J
kg
(10.37)
In order to define a thermodynamic efficiency for the fan impeller, we must first designate a
"perfect" fan against which we can compare the performance of the real fan. In the perfect fan,
there are no losses, F12 = 0. As we now have a frictionless adiabatic, i.e. an isentropic
compression we can write:
2
W isen =
VdP
1
= (H 2,isen H 1 )
J
kg
(10.38)
where the subscript isen denotes isentropic conditions.
29
Chapter 10. Fans
10.6.1.1. The pressure-volume method

There are two methods of determining the efficiency of a fan. The technique that is accepted as
standard by most authorities relies upon measurement of the fan pressure, airflow and shaft
power, and is known as the pressure-volume method.
2
The first step is to evaluate the integral
VdP . Here, we have a choice. We can use the

1
isentropic relationship
PV
= Constant C
(10.39)
where = the isentropic index (i.e. the ratio of specific heats Cp/Cv =1.4 for dry air). This will lead
to the isentropic efficiency of the fan.
Alternatively, we could employ the actual polytrope produced by the fan
PV n
Constant
and assume that it is a reversible (ideal) process. This will lead to the polytropic efficiency of the
fan. Both measures of efficiency are acceptable provided that the choice is stated clearly. While
the polytropic efficiency is closer to a true measure of output/input, and takes any heat transfer,
q12, into account, the isentropic efficiency has the advantage that the index is defined for any
given gas. For this reason, we shall continue the analysis on the basis of isentropic compression
within the ideal fan. In practice, polytropic and isentropic efficiencies are near equal for most fan
installations.*
Substituting V = ( C / P )
W isen =
=P
W isen =
Now
V2
V1
into equation (10.38) gives
dP = C
but as C
1 1
P
1
1 1
[P2 V2
P
= 1
P2
P1 V1 ] =
P V
P1 V1 2 2 1
1
P1 V1
Fan efficiencies can be defined in terms of areas on the Ts diagram of Figure 3.7:
isen =
Area ACBXY
Area DBXZ
poly
Area ADBXY
Area DBXZ
30
Chapter 10. Fans
giving
W isen
1
P
P1 V1 2
=
1
P1
J
kg
In order to convert the shaft work from J/kg to shaft power, Pow, (J/s or Watts), we must multiply
by the mass flow of air
M = Q1 1 =
Q1
V1
kg/s
where = air density ( kg/m3)
Then Pow ,isen
P 1
2
P1 Q1
=
1
P1
J
1
kg
(10.40)
Now let us multiply both the numerator and denominator by fan pressure, pf.
Pow , isen
1
P1 P2
= p f Q1

1 p f P1
We can rewrite this equation as

Pow ,isen
= p f Q1 K p
1
P1 P2
where Kp =

1 pf P1
(10.41)
1
(10.42)
and is known as the Compressibility Coefficient. By substituting p f = (P2 P1 ) we obtain
Kp
( 1)
P2
P

1
=
1 P2
1
P1
(10.43)
We can now see that our earlier assumption of incompressible flow ( Pow = pf Q ) involved an
error represented by the deviation of Kp from unity. Figure 10.15 shows the variation in
compressibility coefficient with respect to the pressure ratio P2/P1. For unsaturated air the value
= 1.4 may be used.
31
Chapter 10. Fans
The isentropic efficiency of the fan can now be defined as
isen =
Pow ,isen
Shaft power
p f Q1 K p
(10.44)
Shaft power
Employment of this equation is a standard technique of determining fan efficiency and is known
commonly as the pressure-volume method.
Compressibility Coefficient Kp
0.995
0.99
0.985
0.98
= 1.7
1.6
1.5
0.975
1.4
1.3
0.97
1
1.01
1.02
1.03
1.04
1.05
1.06
1.07
1.2 1.08
Pressure Ratio P 2 /P 1
Figure 10.15 Variation of Compressibility Coefficient with respect to pressure ratio.

( =1.4 for dry air.)
Two other terms used frequently in the literature are static efficiency and total efficiency. These
are obtained simply by using fan static pressure and fan total pressure, respectively, for pf in
equation (10.44).
All of these measures of efficiency are matters of definition rather than precision. However, care
should be taken to ensure that the same measure is employed when comparing fan
performances.
32
Chapter 10. Fans
Equation (10.44) indicates that the pressure-volume method of fan testing requires the
measurement of pressures, air volume flow and the impeller shaft power. It is frequently the case
that a large fan installed at a mine site does not meet the specifications indicated by factory tests.
Part of the problem may be the less than ideal inlet or outlet conditions that often exist in field
installations. In particular, uneven distribution of airflow approaching the fan can produce a
diminished performance and may even result in premature blade failure.
Another problem is that turbulence and an asymmetric velocity profile may make it difficult to
obtain good accuracy in the measurement of airflow (Section 6.2). Water droplets in the airstream
can also result in erroneous readings from both anemometers and pitot tubes.
Impeller shaft power can be obtained accurately in the laboratory or factory test rig by means of
torque meters or, in the case of smaller fans, swinging carcass (dynamometer) motors. However,
for in-situ tests, it may be necessary to resort to the measurement of input electrical power and to
rely upon manufacturer's data for the efficiencies of the motor and transmission.
Despite these difficulties, it is always advisable to conduct an in-situ test on a new main fan in
order to verify, or modify, the fan characteristic data that are to be used for subsequent ventilation
network exercises.
Example.
A fan passes an airflow of 300 m3/s at the inlet, and develops a pressure of 2.5 kPa. The
barometric pressure at the fan inlet is 97 kPa. The motor consumes an electrical power of 1100
kW. Assuming a combined motor/transmission efficiency of 95 per cent, determine the isentropic
efficiency of the impeller and, also, of the total unit.
Solution.
P1 = 97 kPa,
P2 = 97 + 2.5
99.5 kPa
Using a value of = 1.4 for air, equation (10.40) gives the air power (or isentropic shaft power)
as
Pow ,isen
P 1
2
=
P1 Q1
1
P1
J
1
kg
99.5
3.5 97 000 300
97
743.9 10 3
or
0.286
743.9
kW
Actual shaft power,

Pow = 1100 x 0.95 = 1045 kW
Isentropic efficiency of impeller
isen =
Pow ,isen
Pow
743.9
= 0.712 or 71.2 per cent
1045
33
Chapter 10. Fans
The overall isentropic efficiency of the unit is
isen (overall) =
743.9
= 0.676 or 67.3 per cent
1100
Note that if compressibility had been ignored, the isentropic shaft power would have been
= p f Q = 2.5 300 = 750 kW
Pow ,isen
involving an error of 0.8 per cent from the true value of 743.9 kW.
10.6.1.2. The thermometric method
Unsaturated Conditions
The problems associated with in-situ pressure-volume tests on mine fans have led researchers to
seek another method that did not require the measurement of either airflow or shaft power. The
earliest such work appears to have been carried out on the 1920's (Whitaker) although later work
was required to make it a practical proposition (McPherson 1971, Drummond 1972).
The thermometric method of fan testing is based on the enthalpy terms in the steady flow energy
equation. The shaft work for the ideal isentropic fan is given by equation (10.38)
W isen
= (H 2,isen H 1 )
If we assume that the air contains no free water droplets and that neither evaporation nor
condensation occurs within the fan then equation (3.33) gives
W isen
= (H 2,isen H 1 ) = C p (T 2,isen T1 )
J /kg
(10.45)
where Cp = Specific heat at constant pressure of the air (1005 J/kgK for dry air).
Similarly, for the real fan (no secondary subscript)
W
C p (T 2 T1 )
(10.46)
J/kg
The isentropic efficiency is then given as the ratio of the shaft work for the isentropic fan to that
for the actual fan
or
isen =
W isen
W
isen =
T isen
T
where Tisen
(T 2,isen T1 )
(10.47)
(T 2 T1 )
(T2,isen T1) C and
(10.48)
T
= (T2 - T1)
Tisen is, therefore, the increase in dry bulb temperature that would occur as air passes through
an isentropic fan, while T is the temperature rise that actually occurs in the real fan. These two
parameters are illustrated on the Ts diagram of Figure 3.7.
34
Chapter 10. Fans
As T can be measured, it remains to find an expression for the isentropic temperature rise. This
was derived in Chapter 3 as equation (3.53)
P
= 2
P1
T 2,isen
T1
( 1)
(10.49)
where = Ratio of specific heats, Cp/Cv (1.4 for dry air).

Then
Tisen = (T2,isen
giving
isen
Tisen
=
T
( 1)
P2
T1) = T1
1
P1
( 1)
T1 P2
1
=

T P1
(10.50)
(10.51)
This equation may be employed directly as all of its variables are measurable. It can, however, be
simplified by employing the compressibility coefficient, Kp, from equation (10.42)
( 1)
p ( 1)
P2
1 = K p f

P1
P1
Substituting in equation (10.51) gives
isen =
( 1) p f T1
Kp
P1 T
(10.52)
Using the value of = 1.4 for dry air gives
isen = 0.286
p f T1
Kp
P1 T
(10.53)
Furthermore, for fan pressures not exceeding 2.8 kPa, the compressibility coefficient may be
ignored for one per cent accuracy, leaving the simple equation:
isen = 0.286
p f T1
P1 T
(10.54)
This analysis has assumed that the airflow contains no liquid droplets of water. The presence of
water vapour has very little effect on the accuracy of equations (10.53 or 10.54) provided that the
air remains unsaturated. (See equations A10.15 and A10.16 in Appendix A10.2 following this
chapter.)
35
Chapter 10. Fans
Saturated Conditions
In many deep and hot mines, the reduction in temperature as the return air ascends an upcast
shaft may cause condensation. Exhaust fans operating at the surface will then pass fogged air - a
mixture of air, water vapour and droplets of liquid water. Equation (10.53) no longer holds due to
the cooling effect of evaporation within the fan. A more complex analysis, taking into account the
air, water vapour, liquid droplets, and the phase change of evaporation produces the following
differential equation which quantifies the temperature-pressure relationship for the isentropic
behaviour of fogged air:
L Xs
T
+
(R + R v X s )
P
(P e s )
dT
=
dP
Xs
L2 P
C p + X C w B X s +
(P e s ) Rv T 2
C/Pa
(10.55)
A derivation of equation (10.55) and the full definition of the symbols are given in Appendix A10.2
at the end of this chapter.
Inserting the values of the constants gives
L Xs
T
0.286 (1 + 1.6078 X s )
+
P
287.04 (P e s )
dT
=
dP
L2 P X s
1 + 4.1662 X 2.3741 X s +
463.81 10 3 (P e s ) T 2
C/Pa (10.56)
where T = Absolute temperature (K)

P = Absolute pressure (Pa)
L = Latent heat of evaporation (J/kg)
X = Total moisture content (kg/kg dry air)
Xs = Water vapour content at saturation (kg/kg dry air) and
es = Saturation vapour pressure at temperature T (Pa).
This equation can be programmed into a calculator or microprocessor for the rapid evaluation of dT/dP.
For the relatively small pressures and temperatures developed by a fan, we can write to a good
approximation
T isen = p f
dT
dP
(10.57)
To improve accuracy, the values used for T and P should be the mean temperature and pressure
of the air as it passes through the fan.
Although the thermometric method eliminates the need for airflow and shaft power in the
determination of fan efficiency, it does introduce other practical difficulties. The temperature of the
air may vary with both time and position over the measurement cross-sections due to vortex
action and thermal stratification. The method of measuring the temperature rise across the fan
must give an instantaneous reading of the difference between the mean temperatures at inlet and
outlet measuring stations. This may be accomplished by thermocouples, connected in series, with
the hot and cold junctions distributed over supporting grids at the outlet and inlet measuring
stations respectively (Drummond).
36
Chapter 10. Fans
Poorly designed evasees on exhausting centrifugal fans may suffer from a re-entry down-draught
of external air on one side. A test should be made for this condition before positioning the exit
thermocouple heads.
Although the thermometric technique does not require an airflow in order to calculate efficiency,
the airflow is nevertheless still needed if that efficiency is to be compared with a manufacturer's
characteristic curve. If the site conditions are such that greater confidence can be placed in a
determination of shaft power than airflow, then the latter may be computed as
Q =
m3
s
Shaft power
1 (H 2 H 1 )
(10.58)
where the shaft power is in Watts,

1 = air density at inlet (kg/m3)
and
H = enthalpy (J/kg)
If the air is unsaturated then
(H2 H1) = Cp (T2 - T1)
In the case of fogged air, the enthalpies must be determined from equation (A10.3) in Appendix
A10.2 in order to account for water vapour and liquid droplets..
Example 1.
A temperature rise of 5.96 C is measured across a fan developing an increase in barometric
pressure of 5 kPa and passing unsaturated air. The inlet temperature and pressure are 25.20 C
and 101.2 kPa respectively. Determine the isentropic temperature rise and, hence, the isentropic
efficiency of the impeller.
Solution.
101.2 + 5 0.286
T isen = (273.15 + 25.2)
101.2
Then
isen =
Tisen
T
4.143
= 0.695
5.96
1 = 4.143
or 69.5 per cent
Ignoring the effects of compressibility allows equation (10.54) to be applied, giving
isen = 0.286
(273.15 + 25.2)
5
101.2
5.96
= 0.707
or
70.7 per cent
Hence, in this example, ignoring compressibility causes the percentage fan efficiency to be
overestimated by 1.2.
37
Chapter 10. Fans
Example 2.
A surface exhausting fan passes fogged air of total moisture content X = 0.025 kg/kg dry air. The
pressure and temperature at the fan inlet are 98.606 kPa and 18.80 C respectively. If the
temperature rise across the fan is 2.84 C when the increase in pressure is 5.8 kPa, calculate the
isentropic efficiency of the impeller.
Solution.
The mean barometric pressure in the fan is
P
At inlet, T1
=
=
98.606 + 5.8/2
101.506 kPa
273.15 + 18.8
291.95 K
and at outlet, T2 =
291.95 + 2.84
294.79 K
Hence, the mean temperature is

T = 293.37 K or t = 20.22 C
The psychrometric equations in Section 14.6 allow the following parameters to be calculated. For
saturation conditions, the wet and dry bulb temperatures are, of course, equal.
Saturation vapour pressure:
es
17.27 t
= 610.6 exp
237.3 + t
17.27 20.22
= 610.6 exp
= 2369.5 Pa
237.3 + 20.22
Saturation vapour content:

es
X s = 0.622
(P e s )
= 0.622
2369.5
= 0.01487
(101506 2369.5)
kg/kg
Latent heat of evaporation:

L = (2502.5 2.386 t ) 1000
= (2502 .5 2.386 20.22) 1000 = 2454 .26 10 3
J/kg
Substituting the known values into equation (10.56) gives

0.286 [0.002959 + 0.001282 ]
dT
=
[1 + 0.10416 0.03530 + 2.29693
dP
= 3.604 10 4 C per Pa
This example shows that the term involving the total moisture content, X, is relatively weak
(0.10416). Hence, no stringent efforts need be made to measure this factor with high accuracy.
Then
dT
(equation (10.57))
dP
= 5800 3.604 10 4 = 2.09
T isen = pf
38
Chapter 10. Fans
The isentropic efficiency is
isen =
Tisen
T
2.09
2.84
0.736
or 73.6 per cent
10.6.2. Purchasing a main fan.

The results of ventilation network planning exercises will produce a range of pressure-volume
duties required of any new major fan that is to be installed. The process of finding and ordering
the fan often commences with the ventilation engineer perusing the catalogues of fan
characteristics produced by fan manufacturers. Several of those companies should be invited to
submit tenders for the manufacture and, if required, installation of the fan. However, in order for
those tenders to be complete, information in addition to the required pressure-volume range
should be provided by the purchasing organization.
(a) The mean temperature, barometric pressure, humidity and, hence, air density at the fan
inlet should be given. This allows data based on standard density to be corrected to the
psychrometric conditions expected in the field.
(b) In many cases of both surface and underground fans, noise restrictions need to be
applied. These restrictions should be quantified in terms of noise level and, if necessary,
with respect to direction.
(c) A plan and sections of the site should be provided showing the proposed fan location
and, in particular, highlighting any restrictions on space.
(d) The request for a tender must identify and, wherever possible, quantify the
concentrations and types of pollutants to be handled by the fan. These include dusts,
gases, water vapour and liquid water droplets. In particular, any given agents of a
corrosive nature should be stressed. The purchaser should further indicate any
preference for the materials to be used in the manufacture of the fan impeller and casing.
Specifications on paints or other protective coatings might also be necessary.
(e) Any preference for the type of fan should be indicated. Otherwise, the manufacturer
should be specifically invited to propose one or more fans that will meet the other
specifications.
(f) The scope of the required tender should be clearly defined. If the contractor is to be
responsible for providing, installing and commissioning the new fan, the individual items
should be specified for separate costing.
(g) The motor, transmission, electrical switchgear and monitoring devices may be acquired
and installed either by the provider of the fan or separately. In either case, the voltage
and any restrictions on power availability should be stated.
(h) Areas of responsibility for site preparation and the provision and installation of ducting
should be identified.
39
Chapter 10. Fans
10.7. BOOSTER FANS

10.7.1. The application of booster fans
In deep mines or where workings have become distant from surface connections the pressures
required to be developed by main fans may be very high in order to maintain acceptable face
airflows. This leads to practical difficulties at airlocks and during the transportation of personnel,
mined mineral and materials. More serious, however, is the fact that higher pressures at main
fans inevitably cause greater leakage throughout the entire system. Any required fractional
increase in face airflows will necessitate the same fractional increase in main fan volume flows for
any given system resistance. Hence, as both fan operating power and costs are proportional to
the product of fan pressure and airflow, those costs can rapidly become excessive as a large
mine continues to expand. In such circumstances, the employment of booster fans provides an
attractive alternative to the capital penalties of driving new airways, enlarging existing ones, or
providing additional surface connections. Legislation should be checked for any national or state
restrictions on the use of booster fans.
Unlike the main fans which, in combination, handle all of the mine air, a booster fan installation
deals with the airflow for a localized area of the mine only. The primary objectives of a booster fan
are
to enhance or maintain adequate airflow in areas of the mine that are difficult or
uneconomic to ventilate by main fans, and
to redistribute the pressure pattern such that air leakage is minimized.
A modern booster fan installation, properly located, monitored and maintained creates
considerable improvements in environmental conditions at the workplace, and can allow the
extraction of minerals from areas that would otherwise be uneconomic to mine. It has frequently
been the case that the installation of underground booster fans has resulted in improved
ventilation of a mine while, at the same time, producing significant reductions in total fan
operating costs. However, these benefits depend upon skilled system design and planning. An
inappropriate use of booster fans can actually raise operating costs if, for example, fans act in
partial opposition to each other. Furthermore, if booster fans are improperly located or sized then
they may result in undesired recirculation.
This section is directed towards the planning, monitoring and control of a booster fan installation.
10.7.2. Initial planning and location
An initial step in planning the incorporation of a booster fan into an existing subsurface system is
to obtain or update data by conducting ventilation surveys throughout the network (Chapter 6)
and to establish a correlated basic network (Section 9.2). Two sets of network exercises should
then be carried out.
First, a case must be established for adding a booster fan to the system. Network exercises
should investigate thoroughly all viable alternatives. These may include
adding or upgrading main fans

enlarging existing airways and/or driving new ones including shafts or other surface
connections
redesigning the underground layout to reduce leakage and system resistance - for
example, changing from a U-tube to a through-flow system (Section 4.3), and
reducing face resistance by a redesign of the face ventilation system or the replacement
of line brattices with auxiliary ducts and fans.
40
Chapter 10. Fans
If it is decided to progress with the planning of a booster fan installation then a second series of
network exercises should be carried out to study the preferred location and corresponding duty of
the proposed booster fan. Conventional VNET programs can be used for this purpose. Each of
the alternative feasible locations is simulated in a series of computer runs in order to compare
flow patterns and operating costs1. Flow patterns predicted by the simulations should be checked
against the constraints of required face airflows and velocity limits as for any other network
exercise (Figure 9.1). Additionally, there are several other checks that should be carried out for a
booster fan investigation:Practical constraints on booster fan location
Although there may be a considerable number of branches in the network within which a booster
fan may, theoretically, be sited in order to achieve the required flow enhancement, the majority of
those might be eliminated by practical considerations.
Booster fans should, wherever possible, avoid locations where airlocks would interfere with the
free movement of minerals, materials or personnel. Furthermore the availability of dedicated
electrical power and monitoring circuitry, or the cost of providing such facilities to each potential
site should be considered. If a booster fan is located remotely from frequently travelled airways,
then particular vigilance must be maintained to ensure that it receives regular visual inspections
in addition to continuous electronic surveillance (Section 10.7.3).
Leakage and recirculation
For any given pressure developed by a booster fan in a specified circuit, there exists a fan
position where the summation of leakages inby and outby the fan is a minimum without causing
undesired recirculation. In a U-tube circuit (Figure 4.4), this location occurs where the operation of
the fan achieves zero pressure differential between intake and return airways at the inlet or outlet
of the fan. (See, also Figure 21.7.) Conventional wisdom is to locate the booster fan at this
"neutral point." In practice, there are some difficulties. The neutral point is liable to move about
quite considerably because of
variations in resistance due to face operations and advance/ retreat of workings

changes in the pressures developed by the booster fan or any other fan(s) in the system
variations in the system resistance due to movement of vehicles or longer term changes
in airway resistances.
It is preferable to examine the complete leakage pattern predicted by ventilation network analysis
for each proposed booster fan location.
The question of recirculation must be examined most carefully. If the purpose of the booster fan is
to induce a system of controlled partial recirculation (Section 4.5), then the predicted airflow
pattern must conform to the design value of percentage recirculation. However, undesired or
uncontrolled recirculation must be avoided. At all times, legislative constraints must be observed.
Installing a booster fan in one area of a mine will normally cause a reduction in the airflows within
other districts of the mine. In extreme cases, reversals of airflow and unexpected recirculations
can occur. While examining network predictions involving booster fans, care should be taken to
check airflows in all parts of the network and not simply in the section affected most directly by
the booster fan.
In addition to modifying airflow patterns, booster fans are also a very effective means of
managing the pressure distribution within a network. It is this feature that enables booster fans to
influence the leakage characteristics of a subsurface ventilation system. A properly located and
1
Optimization programs have been developed (e.g. Calizaya, 1988) that determine the optimum
combination of main and booster fan duties, based on the minimization of total operating costs for each
booster fan location investigated by the engineers.
41
Chapter 10. Fans
sized booster fan can be far more effective in controlling leakage across worked out areas than
sealants used on airway sides or stoppings. Conversely, a badly positioned booster fan can
exacerbate leakage problems. These matters are of particular consequence in mines liable to
spontaneous combustion (Section 21.4.5). Hence, the pressure differences predicted across old
workings or relaxed strata should be checked carefully.
Steady state effects of stopping the booster fan
For each proposed location for a booster fan, network simulations should be run to investigate the
effect of stopping that fan. In general, the result will be that the airflow will fall in that area of the
mine and increase in others. (A cross-cut recirculation booster fan can cause the opposite effect
as discussed in Section 4.5.3). In all cases, the ventilation must remain sufficient, without the
booster fan, to ensure that all persons can evacuate the mine safely and without undue haste.
The implication is that main fans alone must always be able to provide sufficient airflows for safe
travel within the mine, and that booster fans merely provide additional airflow to working faces for
dilution of the pollutants caused by breaking and transporting rock. It is good practice to employ
two or more fans in parallel for a booster fan installation. This allows the majority of the airflow to
be maintained if one fan should fail (Section 10.5.2). Further, considerations regarding the
transient effects of stopping or starting fans are discussed in Section 10.7.3.
Economic considerations
The cost-benefits of a proposed booster fan should be analysed with respect to time. A major
booster fan installation should be assured of a reasonable life; otherwise it may be preferable to
tolerate short-lived higher operating costs. On the other hand, a cluster of standard axial fans
connected in a series/parallel configuration (Section 10.5) can provide an inexpensive booster fan
solution to a short term problem. The financial implications of such questions can be quantified by
combining total fan operating costs and the capital costs of purchasing and installing the booster
fans, using the methods described in Section 9.5.
10.7.3. Monitoring and other safety features
When a booster fan is installed in a subsurface ventilation system, it becomes an important
component in governing the behaviour of that system. However, unlike other components such as
stoppings, regulators or air crossings, the booster fan is actively powered and has a high speed
rotating impeller. Furthermore, it is usually less accessible than a surface fan in case of an
emergency condition. For these reasons a subsurface booster fan should be subjected to
continuous surveillance. The traditional way of doing this was to post a person at the booster fan
location at all times when it was running. This was one of life's more tedious jobs. Some
installations reverted to the use of television cameras. A bank of display monitors covering a
number of sites could be observed by a single observer at a central location. The problems
associated with such methods were that they depended upon the vigilance of human beings and
were limited to visual observations. The task may now be undertaken much more reliably and
efficiently by electronic surveillance, employing transducers to sense a variety of parameters.
These transducers transmit information in a near continuous manner to a central control station,
usually located on surface, where the signals are analysed by computer for display, recording and
audio-visual alarms when set limits are exceeded. Alarms and, if required, displays may be
located both at the control station and also at the fan site.
The monitors used at a booster fan may form part of a mine-wide environmental monitoring
system which can, itself, be integrated into a general communication network for monitoring and
controlling the condition and activities of conveyors, roof supports, excavation machinery and
other equipment. However, sufficient redundancy should be built into the booster fan surveillance
such that it remains operational in the event of failures in other parts of the mine monitoring
system.
42
Chapter 10. Fans
The parameters to be monitored at, or near, a booster fan installation include:
gas concentrations (e.g. methane) in the air approaching the fan

carbon monoxide and/or smoke both before and after the fan
pressure difference across the fan measured, preferably, at the adjoining airlock
an indication of airflow in the fan inlet or outlet
bearing temperatures and vibration on both the motor and the fan impeller and
positions of the airlock and anti-reversal doors.
All monitors should be fail-safe, i.e. the failure of any transducer should be detected by
interrogative signals initiated by the computer, resulting in automatic transfer to a companion
transducer, display of a warning message and printing of an accompanying hardcopy report. It
follows that all transducers must be provided with at least one level of back-up devices.
The monitoring function is particularly important where controlled partial recirculation is practiced
or at times of rapid transient change in the ventilation system. The latter may occur by an
imprudent operation of doors in other areas of the mine or at a time when either the booster fan
itself or other fans in the system are switched on or off. Any sudden reduction in air pressure
caused by such activities can cause gases held in old workings or relaxed strata to expand into
the ventilating airstream (Section 4.2.2).
A booster fan should be provided with electrical power that is independent of the main supply
system to the mine. However, electrical interlocks should be provided according to a
predetermined control policy (Section 10.7.4). This policy might include:
automatic stoppage of the booster fan if the main fan(s) cease to operate
stoppage of the booster fan if all of the local airlock doors are opened simultaneously
isolation of electrical power inby if the booster fan ceases to operate.
The second item on this list may also be required in reverse. Hence, for example, the booster fan
could not be started if the local airlock doors were all open. However, those doors should open
automatically when the booster fan stops, in order to re-establish conventional ventilation by the
main fans. The question of interlocks may be addressed by legislation and might also be
influenced by existing electrical systems. All electrical equipment should be subject to the
legislation pertaining to that mine.
If the fan is to operate in a return airway, then the design will be improved if the motor is located
out of the main airstream and ventilated by a split of fresh air taken from an intake airway. In this
regard, it is also sensible to ensure that there are no flammable components within the
transmission train.
Anti-reversal doors, flaps or louvres should be fitted to a booster fan in order to prevent reversal
of the airflow through the fan when it stops.
All materials used in the fan, fan site and within some 50m of the fan should be non-flammable.
Furthermore, no flammable materials should be used in the construction of the local airlock,
stoppings or air crossings. Finally an automatic fire suppression system should be provided at the
fan station.
10.7.4. Booster fan control policy
Continuous electronic monitoring systems provide a tremendous amount of data. During the
normal routine conditions that occur for the vast majority of the time, all that is required is that the
monitoring system responds to any manual requests to display monitored parameters, and to
43
Chapter 10. Fans
record information on electronic or magnetic media. However, the ability to generate and transmit
control signals raises the question of how best to respond to unusual deviations in the values of
monitored parameters.
The vision of complete automatic control of the mine environment produced a flurry of research
activity in the 1960's and '70's (Mahdi, Rustan, Aldridge). This work resulted in significant
advances in the development of transducers and data transmission systems suitable for
underground use. However, a completely closed loop form of environmental control has seldom
been implemented for practical use in producing mines. The difficulty is that there are a number
of variables which, although of vital importance in an emergency condition, may not be amenable
to direct measurement. These include the locations of personnel and any local actions those
personnel might have taken in an attempt to ameliorate the emergency situation.
If an automatic control policy is to be implemented for a booster fan, careful consideration should
be given to the action initiated by the system in response to deviations from normal conditions.
The control policy that is established should then be incorporated into the computer software
(Calizaya, 1989).
The simplest control policy is to activate audio-visual alarms and to allow control signals to be
generated manually. Manual over-ride of the system must be possible at all times. In order to
examine additional automatic responses, unusual deviations in each of the monitored parameters
listed in the previous subsection are considered here:
(a) Methane (or other gas of concern)
Four concentration levels may be chosen or mandated by law
(i)
(ii)
(iii)
(iv)
concentration at which personnel are to be withdrawn

concentration at which power inby the fan is to be cut off
concentration at which power to the booster fan should be cut
concentration below which no action is to be initiated.
Levels (i) and (ii) are usually fixed by law. Electrical power inby the fan should be isolated
automatically when level (ii) is reached. If the fan is powered by an electric motor which is located
within the airstream then the power to the fan should also be cut when level (ii) is reached.
However, this may cause even more dangerous conditions for any personnel that are located
inby. Having the motor located in a fresh air split allows greater flexibility. At the present time, it is
suggested that when level (iii) is exceeded manual control should be established.
Level (iv) is relevant only for those systems where control of the volume flow of air passing
through the fan is possible. This may be achieved by automatic adjustment of impeller blade
angle, guide vanes or motorized regulators. Level (iv) may be set at one half the concentration at
which power inby the fan must be cut. Variations in the monitored gas concentration that occur
below level (iv) are to be considered as normal and no action is taken. Variations between levels
(iv) and (iii) may result in a PID (proportional/integral/differential) response. This means that
control signals will be transmitted to increase the airflow to an extent defined by the level of the
monitored concentration, how long that level has existed and the rate at which it is increasing.
The objective is to prevent the gas concentration reaching level (iii). The maximum increase in
airflow must have been predetermined by network analysis such that the ventilation of the rest of
the mine is maintained at an acceptable level.
(b) Smoke or carbon monoxide
The system should be capable of distinguishing hazardous levels of smoke or carbon monoxide
from short-lived peaks caused by blasting or nearby diesel equipment. A monitoring system that
initiates alarms frequently and without good cause will rapidly lose its effectiveness on human
operators. There is no single answer to the response to be initiated if these alarms indicate an
actual fire. Similar alarms arising from detectors in other parts of the mine may allow the probable
44
Chapter 10. Fans
location of the fire to be determined. Permitting the booster fan to continue running may
accelerate the spread of the fire and products of combustion. On the other hand, stopping the fan
might allow toxic gases to spread into areas where personnel could have gathered, particularly if
damage to the integrity of air crossings or stoppings is suspected. The only realistic action in
these circumstances is to revert to manual control.
(c) Pressure differential and airflow
An increase in the pressure developed by the fan will usually indicate an increase in resistance
inby. This might be caused by normal face operations, partial blockage of an airway by a vehicle
or other means, or obstruction of the fan inlet grill. No action is required other than a warning
display and activation of an alarm if the fan pressure approaches stall conditions. A rapid fall in
fan pressure accompanied by an increase in airflow is indicative of a short circuit and local
recirculation. A warning display should be initiated together with checks on the positions of air
lock doors followed, if necessary, by a manual inspection of the area. A simultaneous rapid fall in
pressure and airflow suggests a problem with the fan itself. The power supply should be checked
and, again, the fan should be inspected manually. Such circumstances will usually be
accompanied by indications of the source of the difficulty from other transducers.
(d) Bearing temperatures, vibration
Indications of excessive (or rising) values of bearing temperatures or vibration at either the motor
or the fan impeller should result in isolation of power to the fan and all electrical equipment inby
the fan.
(e) Positions of airlock doors
An airlock door which is held open for more than a pre-determined time interval should initiate a
warning display and, if uncorrected, should warrant a manual inspection.
Bibliography
Aldridge, M.D. and Nutter, R.S. (1980). An experimental ventilation control system. 2nd. Int.
Mine Ventilation Congress, Reno, NV. pp.230-238.
ASHRAE (1985). Laboratory methods of testing fans for ratings. ANSI/AMCA. Atlanta, Georgia,
USA.
British Standard 848 (1963). Methods of testing fans for general purposes, Part 1. HMSO.
London.
De La Harpe J.H. (1982). Basic fan engineering. Chapter 7. Environmental Engineering in South
African Mines. Mine Ventilation Society of South Africa.
Drummond, J.A. (1972). Fan efficiency investigations on mines of Union Corp. Ltd. J. Mine
Ventilation Society of South Africa. Vol. 25, pp.180-195.
Mahdi, A. and McPherson, M.J. (1971). An introduction to automatic control of mine ventilation
systems. Mining Technology. May.
McFarlane, D. (1966). Ventilation Engineering. Davidson & Co. Ltd, Belfast, N. Ireland.
McPherson, M.J. (1971). The isentropic compression of moist air in fans. J. Mine Ventilation
Society of South Africa, Vol. 24, pp.74-89.
Rustan, A. and Stockel, I. (1980). Review of developments in monitoring and control of mine
ventilation systems. 2nd Int. Mine Ventilation Congress. Rena, NV. pp. 223-229.
Stepanoff, A.J. (1955). Turboblowers. Wiley & Sons Ltd.
Whitaker, J.W. (1926, 1928). The efficiency of a fan. Trans. Inst. of Mining Engineers. Vol. 72,
p.43. Vol. 74, p.93.
45
Chapter 10. Fans
APPENDIX A10.1
Comparisons of exhausting and forcing fan pressures to use in pressure surveys.
As stated in Section 10.2 of this chapter, it is fan static pressures that should be employed when
a main surface fan is included in the route of a pressure survey. It is not immediately obvious why
this should be so when total pressures are used at all other survey measuring points. The
apparent anomaly arises from the way in which fan pressures are defined.
In this Appendix, we will consider an identical main fan at a mine surface in each of three
configurations. In each case the fan operates at the same speed and passes the same airflow
against a constant mine resistance. In such circumstances the fan will deliver the same air power
and, hence, produce the same rise in total (facing tube) pressure. In this illustrative example we
will take the increase in total pressure across the fan to be pt = 4000 Pa while the air velocity at
the fan measuring station produces a velocity pressure pv = 500 Pa.
Configuration 1.
Mine Resistance
fan
pv,out = 500 Pa
ps,out = 0
pt,out = pv+ ps,out = 500 Pa
ps,in
-4000 Pa
pt,in
pt = pt,out - pt,in = 500 (-3500) = 4000 Pa

-3500 Pa
(FTP and FSP)
pv,in = pt,in - ps,in = -3500 (-4000) = 500 Pa

Figure A10.1 Main surface exhaust fan with no outlet cone.
Figure A10.1 shows the fan operating in an exhausting configuration with the same inlet and
exhaust cross sectional areas and no outlet cone. At the point of exit to the outside atmosphere
the velocity pressure remains at 500 Pa but the gauge static pressure is ps,out = 0, giving the total
pressure at this location to be
pt,out = pv,out + ps,out = 500 + 0 = 500 Pa
But as pt = pt,out - pt,in = 4000 Pa, it follows that the total pressure shown by a facing tube at the
fan inlet must be
pt,in = pt,out - pt = 500 4000 = -3500 Pa (negative with respect to the external
atmospheric pressure)
and the corresponding static pressure is ps,in = pt,in - pv,in = (-3500) 500 = -4000 Pa
As the kinetic energy of the outlet air is dissipated entirely within the external atmosphere, it does
not contribute to the ventilating pressure applied to the mine. Hence, that effective ventilating
pressure becomes 3500 Pa rather than 4000 Pa and is indicated by the reading on the facing
tube gauge, i.e. pt,in . Note, however that because of the way in which fan pressures are defined,
the total pressure at the fan inlet in a surface exhausting fan is equal to the Fan Static Pressure,
FSP (Figure 10.1(c)).
46
Chapter 10. Fans
Configuration 2.
Mine Resistance
fan
pv,out = 300 Pa
ps,out = 0
pt,out = pv,out+ ps,out = 300 Pa
ps,in
pt,in
pt = pt,out - pt,in = 300 (-3700) = 4000 Pa

-3700 Pa
-4200 Pa
(FTP and FSP)

pv,in = pt,in - ps,in = -3700 (-4200) = 500 Pa
Figure A10.2 Exhausting fan with an outlet cone, slowing discharge air
and reducing exit velocity pressure to 300 Pa.
Figure A10.2 illustrates the effect of adding an outlet cone to the previous example of a surface
exhausting fan. The mine resistance, airflow and fan power remain unchanged. Hence the
increase in total pressure across the fan remains at 4000 Pa. However, as shown in the figure,
the exit total pressure is now 300 Pa. Hence the gauge inlet total pressure becomes
pt ,in
= 300 4000 = 3700
Pa
Furthermore, as the inlet velocity pressure has remained unchanged at 500 Pa, the inlet static
pressure is now
p s,in
= p t ,in pv ,in
= 3700 500 = 4200
Pa
As in Configuration 1, the kinetic energy of the outlet air is dissipated entirely in the external
atmosphere and does not contribute to the ventilating pressure applied to the mine. However, the
corresponding exit velocity pressure has now been reduced from 500 Pa to 300 Pa, increasing
the effective ventilation pressure applied to the mine from the previous 3500 Pa to 3700 Pa,
measured as the total pressure at the fan inlet.
This example further illustrates the advantage of adding an outlet cone to an exhausting fan.
Note, however, that losses due to excessive turbulence and/or wall friction can more than
counteract this advantage if the cone is too long or diverges too rapidly.2
Angles of divergence (i.e. between opposite sides) are normally in the range 7 to 15 for mine fans.
47
Chapter 10. Fans
Configuration 3
Mine Resistance
Fan
pv,in = 0
ps,in = 0
pt,in = 0
pt = pt,out - pt,in
= 4000 - 0
= 4000 Pa
ps,out
pt,out
+3500 Pa
(FSP)
+4000 Pa
(FTP)
pv,out = pt,out - ps,out = 4000 3500 = 500 Pa

Figure A10.3 Forcing fan.
Figure A10.3 shows the same fan in a forcing mode with all other parameters as in the previous
two configurations. In this case we take the external station to be outside the duct entrance as part
of the fan power must be used to accelerate the air from zero to a velocity pressure of 500 Pa3.
As the total pressure at the external station, pt,in, is zero and the increase in total pressure across
the fan remains at 4000 Pa, it follows that the total pressure at the fan outlet, pt,out, must be 4000
Pa and the corresponding static pressure becomes
p s,out
= p t ,out pv ,out
= 4000 500 = 3500 Pa
This is the effective pressure that remains to ventilate the mine and, hence, the pressure to be
used if the forcing fan is included in a pressure survey. Once again, because of the way in which
fan pressures are defined, this is also the pressure designated as the Fan Static Pressure, FSP,
for a forcing fan (Figure 10.1(b)).
It is also the case that part of the airpower developed by an exhausting fan is utilized in accelerating the air
at the intake portals of the mine. However, in that case it is normally measured separately as shock losses
at those portals.
48
Chapter 10. Fans
APPENDIX A10.2
Derivation of the isentropic temperature-pressure relationship for a mixture of air, water
vapour and liquid water droplets.
It is suggested that the reader delays working on this appendix until Chapter 14 (Psychrometry)
has been studied.
During a flow process, the work done by expansion or compression of the air is given by VdP. In
the case of fogged air containing X kg of water (vapour plus liquid) per kg of dry air, this becomes
VsdP where the specific volume of the (1 + Xs) kg of the air/vapour mixture is
Vs
= (R + R v X s )
T
P
where
and
P
R
Rv
T
Xs
=
=
=
=
=
barometric pressure (Pa)

gas constant for dry air (287.04 J/kg K)
gas constant for water vapour (461.5 J/kg K)
Absolute temperature (K)
mass of water vapour associated with each 1 kg of dry air in the
saturated space (kg/kg dry air)
The volume of the (X - Xs) kg of liquid water is negligible compared to that of the gases. In a
fogged airstream, the wet and dry bulb temperatures are equal.
The steady flow energy equation (10.35) for an isentropic process (dF = dq = 0) then becomes
Vs dP = (R + R v X s )
T
dP = dH
P
J
kg
(A10.1)
The enthalpy of the (1 + X) kg of air/vapour /liquid mix, based on a datum of 0 C, is the

combination of
and
Cp t : heat required to raise 1 kg of dry air through t C

X Cw t : heat required to raise X kg of liquid water through t C
L Xs : heat required to evaporate Xs kg of water at t C
where Cp
Cw
L
Then
=
=
=
=
Specific heat of dry air (1005 J/kgK)

Specific heat of liquid water (4187 J/kgK)
Latent heat of liquid water at temperature t C
(2502.5 - 2.386 t)1000 J/kg
H = C p t + X Cw t + L X s
J
kg
(A10.2)
(A10.3)
Differentiating,
dH = (C p + X C w ) dt + X s dL + L dX s
J
kg
(A10.4)
It is assumed that no water is added or removed during the process. Hence, X remains constant.
We must now seek to formulate this expression in terms of temperature and pressure. First, we
evaluate dL and dXs as functions of temperature and pressure
49
Chapter 10. Fans
(a) Change of latent heat of evaporation, dL

From equation (A10.2),
dL = B dt
where B = 2386
(A10.5)
(b) Change of vapour content, dXs

From equation (14.4) for saturation conditions
Xs
= G
where G
es
and
P
es
(P e s )
=
=
=
(A10.6)
0.622
saturation vapour pressure (Pa) at t C
absolute (barometric) pressure (Pa).
By partial differentiation,
dX s
X s
X s
de s +
dP
e s
P
(A10.7)
From equation (A10.6)

X s
e s
es
GP
1
+
G
=
2
(P e s ) 2
(P e s ) (P e s )
(A10.8)
and
X s
P
G e s
(A10.9)
(P e s ) 2
Furthermore, the Clausius-Clapeyron equation (14.5) gives

de s
where Rv
and
T
L es
Rv T 2
=
=
dt
(A10.10)
Pa
gas constant for water vapour (461.5 J/kg K)

(273.15 + t)
(K)
giving dt = dT
Substituting from equations (A10.8, A10.9, and A10.10) into equation (A10.7) gives
dX s
L es
GP
(P e s )
Rv T
dT
G es
(P e s ) 2
dP
(A10.11)
50
Chapter 10. Fans
But as X s =
G es
(P e s )
from equation (A10.6),
equation (A10.11) becomes

dX s
Xs P
Xs
L
dT
dP
(P e s ) R v T 2
(P e s )
(A10.12)
Having evaluated dL (equation (A10.5)) and dXs (equation (A10.12)), we can substitute these
expressions into equation (A10.4) and, hence, (A10.1)
dH = (C p + X C w ) dt B X s dt +
Xs P
L Xs
L2
T
dt
dP = (R + R v X s ) dP
2
(P e s ) R v T
(P e s )
P
Collecting the dt and dP terms, and replacing dt by dT for consistency (the two are identical),
dT
dP
L Xs
T
+
(R + R v X s )
(P e s )
P
C
Pa
Xs
L P
C p + X Cw B X s +
(P e s ) R v T 2
(A10.13)
This is a general isentropic relationship describing the rate of change of temperature with respect
to pressure for a saturated mixture of air, water vapour and liquid water. All of the variables Xs, L,
and es depend only upon T and P.
For unsaturated air involving no liquid water, Xs becomes X and there is no change of phase.
Hence, the terms that include L become zero. Then
dT
dP
(R + R v X
T
C p + (Cw B ) X P
C
Pa
(see footnote*)
(A10.14)
Inserting values for the constants gives

dT
dP
R (1+ 1.6078 X ) T
C p (1+ 1.7920 X ) P
Expansion of (1 + 1.7920 X)-1 and ignoring terms of X2 and smaller leads to the approximation
dT
dP
R
T
(1 0.2 X )
Cp
P
= 0.286 (1 0.2 X )
T
P
(A10.15)
By considering an isothermal-isobaric change of phase, it can be shown that Cw - B = Cpv,

where Cpv = specific heat at constant pressure of water vapour. This gives a value for Cpv to be
4187 - 2386 = 1801 J/kgK. The 4.4 per cent deviation from the normally accepted value of
1884 J/kgK indicates the uncertainty of Cpv within the atmospheric range of temperatures.
51
Chapter 10. Fans
For dry air, X = 0 leaving

dT
T
= 0.286
(A10.16)
dP
P
(This correlates with equation (10.54) when dT = T, dP = pf and isen = 1).
Equation (A10.15) shows that if moist but unsaturated air is assumed to be dry then the fractional
error involved is approximately 0.2X. Hence, for an error of one per cent
or
0.2 X
X
=
=
0.01
0.05 kg/kg dry air
This vapour content represents fully saturated conditions at 40 C and 100 kPa. It is unlikely that
such moisture content will exist in mining circumstances without liquid water being present. Thus,
for unsaturated air in the normal atmospheric range, assuming dry air will give an accuracy of
within one per cent for the temperature/pressure gradient.
52
Gases in the Subsurface
PART 3
CHAPTER 11
GASES IN SUBSURFACE OPENINGS
11.1. INTRODUCTION .........................................................................................2

11.2. CLASSIFICATION OF SUBSURFACE GASES...........................................2
11.2.1. Threshold limit values (TLV) ............................................................................................ 2
11.2.2. Oxygen, O2....................................................................................................................... 5
11.2.3. Nitrogen, N2...................................................................................................................... 6
11.2.4. Methane, CH4................................................................................................................... 6
11.2.5. Carbon dioxide, CO2 ........................................................................................................ 9
11.2.6. Carbon monoxide, CO ..................................................................................................... 9
11.2.7. Sulphur dioxide, SO2 ...................................................................................................... 12
11.2.8. Oxides of nitrogen, NOx................................................................................................. 12
11.2.9. Hydrogen sulphide, H2S................................................................................................. 13
11.2.10. Hydrogen, H2................................................................................................................ 14
11.2.11. Radon, Rn.................................................................................................................... 14
11.3. GAS MIXTURES........................................................................................15

11.3.1. Threshold limit values for gas mixtures ......................................................................... 15
11.3.2. Diesel emissions ............................................................................................................ 17
11.3.3. Fires ............................................................................................................................... 18
11.3.4. Explosives...................................................................................................................... 19
11.3.5. Welding .......................................................................................................................... 19
11.4. GAS DETECTION AND MONITORING .....................................................20

11.4.1. Objectives and overview ................................................................................................ 20
11.4.2. Principles of gas detection............................................................................................. 20
11.4.2.1. Filament and catalytic oxidation (pellistor) detectors .............................................. 20
11.4.2.2. Flame safety lamps ................................................................................................. 22
11.4.2.3. Thermal conductivity and acoustic gas detectors ................................................... 23
11.4.2.4. Optical methods ...................................................................................................... 23
11.4.2.5. Electrochemical methods ........................................................................................ 24
11.4.2.6. Mass spectrometers ................................................................................................ 25
11.4.2.7. Paramagnetic analysers.......................................................................................... 25
11.4.2.8. Gas chromatography............................................................................................... 25
11.4.2.9. Semiconductor detectors ........................................................................................ 25
11.4.2.10. Stain tubes ............................................................................................................ 25
11.4.3. Methods of sampling...................................................................................................... 26
11.4.3.1. Manual methods...................................................................................................... 26
11.4.3.2. Automatic and remote monitors .............................................................................. 26
References .........................................................................................................27
11-1
11.1. INTRODUCTION
When air enters any mine or other subsurface structure, it has a volume composition of
approximately 78 percent nitrogen, 21 percent oxygen and 1 percent other gases on a moisture
free basis. A more precise analysis is given in Table 14.1. However, as the air progresses
through the network of underground openings, that composition changes. There are two primary
reasons for this. First, the mining of subsurface structures allows any gases that exist in the
surrounding strata to escape into the ventilating airstream. Such strata gases have been
produced over geological time and remain trapped within the pores and fracture networks of the
rock. Methane and carbon dioxide are commonly occurring strata gases.
Secondly, a large number of chemical reactions may cause changes in the composition of mine
air. Oxidation reduces the percentage of oxygen and often causes the evolution of carbon dioxide
or sulphur dioxide. The action of acid mine water on sulphide minerals may produce the
characteristic odour of hydrogen sulphide while the burning of fuels or the use of explosives
produce a range of gaseous pollutants. Most of the fatalities resulting from mine fires and
explosions have been caused by the toxic gases that are produced rapidly in such circumstances.
Several of the gases that appear in subsurface facilities are highly toxic and some are
dangerously flammable when mixed with air. Their rate of production is seldom constant.
Furthermore, their propagation through the multiple airways and often convoluted leakage paths
of a subsurface ventilation system is further modified by the effects of gas density differences,
diffusion and turbulent dispersion. These matters all influence the variations in concentrations of
mine gases that may be found at any time and location in an underground mine.
A primary requirement of a mine ventilation system is to dilute and remove airborne pollutants
(Section 1.3.1.). It is, therefore, necessary that the subsurface environmental engineer should be
familiar with the physical and chemical properties of mine gases, their physiological effects, how
they may be detected and preferred methods of control. In this chapter we shall classify and
discuss those gases that may appear in subsurface ventilation systems.
11.2. CLASSIFICATION OF SUBSURFACE GASES

Table 11.1 lists the gases that are most commonly encountered in underground openings. Each
of those gases is discussed in further detail within this section.
11.2.1. Threshold limit values (TLV)
Threshold limit values of airborne substances refer to those concentrations within which
personnel may be exposed without known adverse effects to their health or safety. The guideline
TLVs given on Table 11.1 have been arrived at through a combination of industrial experience
and from both animal and human studies. They are based primarily on recommendations of the
American Conference of Governmental Industrial Hygienists (ACGIH) and the U. S. National
Institute for Occupational Safety and Health (NIOSH). The values quoted should be regarded as
guidelines rather than clear demarcations between safe and dangerous concentrations. There are
two reasons for this. First, wide variations in personal response to given substances occur
between individuals, depending upon one's state of health, exposure history and personal habits
(e. g. smoking and use of alcohol). Secondly, TLVs are necessarily based on current scientific
knowledge and are subject to revision as new evidence becomes available.
To ensure compliance with statutory requirements, ventilation/environmental engineers and
industrial hygienists should familiarize themselves with any mandatory TLVs that have been
established within their particular country or state.
11-2
Table 11.1 Classification of gases most commonly found in subsurface openings

Name
Symbol Molecular Density at Density

Gas
Primary sources
Weight 20C and relative Constant
in mines
( based 100 kPa to dry air J/kg C
on C12 )
kg/m3
dry air
28.966
1.1884
287.04
Smell,
colour,
taste
Hazards
Guideline
TLVs
Methods of
detection*
Flammability
limits in air
per cent
none
oxygen
O2
31.999
1.3129
1.1047
259.83
air
none
oxygen
>19.5
deficiency;
may cause
explosive
mixtures
with
reactive
gases
inert
electrochemical,
paramagnetic,
flame lamp
nitrogen
N2
28.015
1.1493
0.9671
296.80
air, strata
none
methane
CH4
16.04
0.6581
0.554
518.35
strata
none
explosive;
layering
slight acid
taste and
smell
promotes TWA=0.5%
increased STEL=3.0%
rate of
respiration
catalytic
oxidation,
thermal
conductivity,
optical,
acoustic,
flame lamp
optical,
infra-red
carbon
dioxide
CO2
44.00
1.805
1.519
188.96
carbon
monoxide
CO
28.01
1.149
0.967
296.8
sulphur
dioxide
SO2
64.06
2.628
2.212
129.8
oxidation of
carbon, fires,
explosions, IC
engines, blasting,
respiration
fires, explosions,
IC engines,
blasting,
spontaneous or
incomplete
combustion of
carbon compounds
oxidation of
sulphides, acid
water on sulphide
ores, IC engines
none
highly
toxic;
explosive
acid taste,
suffocating
smell
very toxic; TWA=2ppm

irritant to
STEL=5ppm
eyes throat
and lungs
1%:isolate
electricity
2%:remove
personnel
TWA= 0.005%
STEL=0.04%
by difference
electrochemical,
catalytic
oxidation,
semiconductor,
infra-red
electrochemical,
infra-red
5 to 15
12.5 to 74.2
11-3
Name
Symbol Molecular Density Density

Gas
Primary sources
Weight at 20C relative Constant in mines
(based on and 100 to dry air J/kg C
C12)
kPa
kg/m3
nitric oxide
NO
30.01
1.231
1.036
277.1
nitrous
oxide
N2O
44.01
1.806
1.519
188.9
nitrogen
dioxide
NO2
46.01
1.888
1.588
180.7
hydrogen
sulphide
H2S
34.08
1.398
1.177
244.0
hydrogen
H2
2.016
0.0827
0.0696
4124.6
radon
Rn
222
9.108
7.66
37.45
water
vapour
H2O
18.016
0.739
0.622
461.5
IC engines,
blasting, welding
acid water on
sulphides,
stagnant water,
strata,
decomposition of
organic materials
battery charging,
strata, water on
incandescent
materials,
explosions
uranium minerals
in strata
Smell,
colour,
taste
irritant to
eyes,
nose and
throat
sweet
smell
Hazards
oxidizes
rapidly to
NO2
narcotic
(laughing
gas)
very toxic;
reddish
throat and
brown,
lung irritant;
acidic
pulmonary
smell
and taste infections
odour of highly toxic;
bad eggs irritant to
eyes and
respiratory
tracts;
explosive
Guideline
TLVs
Methods of
detection*
TWA=25ppm
electro-chemical,
infra-red
TWA=50ppm
electro-chemical
TWA=3ppm
ceiling: 5ppm
electro-chemical,
infra red
TWA=10ppm electro-chemical,
STEL=15ppm semi-conductor
none
highly
explosive
catalytic
oxidation
none
radioactive; 1 WL and
decays to
4 WL-months
radioactive
per year
particles
affects
climatic
environment
radiation
detectors
evaporation of
none
water, IC engines,
respiration,
spon.com. and
other fires
Flammability
limits in air
Per cent
4.3 to 45.5
4 to 74.2
psychrometers,
dielectric effects
*In addition to the methods listed, most of these gases can be detected by gas chromatography and stain tubes
11-4
Three types of threshold limit values are expressed in Table 11.1. The Time-Weighted Average
(TWA) is the average concentration to which nearly all workers may be exposed over an 8 hour
shift and a 40 hour work week without known adverse effects. However, many substances are
sufficiently toxic that short-term exposures at higher concentrations may prove harmful or even
fatal. The Short-Term Exposure Limit (STEL) is a time-weighted average concentration occurring
over a period of not more than 15 minutes. That is, concentrations above the TWA and up to the
STEL should not last for longer than 15 minutes. It is also recommended that such circumstances
should not occur more than four times per day nor at intervening intervals of less than one hour.
The Ceiling Limit is the concentration that should not be exceeded at any time. This is relevant for
the most toxic substances or those that produce an immediate irritant effect.
11.2.2. Oxygen, O2
Human beings and, indeed, the vast majority of the animal kingdom are completely dependent
upon the oxygen that comprises some 21 percent of fresh atmospheric air. Oxygen diffuses
through the walls of the alveoli in the lungs to form oxyhaemoglobin in the bloodstream. This
unstable substance breaks down quite readily to release the oxygen where required throughout
the body.
As muscular activity increases, so also does the rate of respiration and the volume of air
exchanged at each breath. However, the percentage of oxygen that is utilized decreases at
heavier rates of breathing. For low levels of physical activity, exhaled air contains approximately
16 percent oxygen, 79 percent nitrogen and 5 percent carbon dioxide. Table 11.2 indicates typical
rates of oxygen consumption and production of carbon dioxide. This table is helpful in estimating
the effects of respiration on the gas concentrations in a confined area.
Activity
Breaths/min
Inhalation rate
litres/s
Oxygen
consumption
litres/s
Carbon dioxide
produced
litres/s
At rest
12-18
0.08 - 0.2
0.005
0.004
Moderate work
30
0.8 - 1.0
0.03
0.027
Vigorous work
40
1.6
0.05
0.05
Table 11.2. Gas exchange during respiration (based on work by Forbes and Grove (1954)).
Example
Air is supplied at a rate of 5m3 /s to ten persons working at a moderate rate in a mine heading.
The intake air contains 20.6 percent oxygen and 0.1 percent carbon dioxide. Determine the
changes in concentrations of these gases caused by respiration.
Solution
For a moderate rate of activity, Table 11.2 indicates an individual consumption rate of some 0.03
litres/s for oxygen and a production rate of about 0.027 litres/s for carbon dioxide. Hence, for ten
persons:
oxygen depletion
carbon dioxide added
= 10 x 0.03 x 10-3 = 0.0003 m3/s

= 10 x 0.027 x 10-3 = 0.00027 m3/s
11-5
In the intake air:

oxygen flow
carbon dioxide flow
=
=
Therefore, in the exit air:

oxygen flow
carbon dioxide flow
=
=
5 x 0.206 = 1.03 m3/s

5 x 0.001 = 0.005 m3 /s
1.03 - 0.0003 = 1.0297 m3/s

0.005 + 0.00027 = 0.00527 m3/s
1.0297
=
100 = 20.594
oxygen concentration
5
0.00527
100 = 0.1054
carbon dioxide concentration =
5
percent
percent
This example illustrates the very limited effect of respiration on the concentrations of oxygen and
carbon dioxide in a mine ventilation system. Indeed, the ventilating air necessary for breathing is
negligible compared with other airflow requirements in the subsurface (Section 9.3.).
As air flows through an underground facility, it is probable that its oxygen content will decrease.
This occurs not only because of respiration but, more importantly, from the oxidation of minerals
(particularly coal and sulphide ores) and imported materials. The burning of fuels within internal
combustion engines and open fires also consume oxygen. The primary danger of this gas is,
therefore, that it may be depleted below a level that is necessary for the well being of the
workforce. In Table 11.1, oxygen is the only gas whose concentration should be maintained
above its recommended threshold limit value. The effects of oxygen depletion are as follows:
Percent oxygen in air
19.0
Effects
Flame height on a flame safety lamp reduced by 50 percent.
17
Noticeable increase in rate and depth of breathing - this effect will be

further enhanced by an increased concentration of carbon dioxide.
16
Flame lamp extinguished.
15
Dizziness, increased heartbeat.
13 to 9
7
below 6
Disorientation, fainting, nausea, headache, blue lips, coma.

Coma, convulsions and probable death.
Fatal.
Oxygen deficiency implies an increased concentration of one or more other gases. Hence,
although not listed explicitly in Table 11.1, even non-toxic gases will endanger life by asphyxiation
if they are present in sufficient concentration to cause a significant oxygen deficiency.
11.2.3. Nitrogen, N2
Nitrogen constitutes approximately 78 percent of air and is, therefore, the most abundant gas in a
ventilated system. It is fairly inert and occurs occasionally as a strata gas, usually mixed with
other gases such as methane and carbon dioxide.
11.2.4. Methane, CH4
Methane is produced by bacterial and chemical action on organic material. It is evolved during the
formation of both coal and petroleum, and is one of the most common strata gases. Methane is
11-6
not toxic but is particularly dangerous because it is flammable and can form an explosive mixture
with air. This has resulted in the deaths of many thousands of miners. A methane: air mixture is
sometimes referred to as firedamp.
Although methane is especially associated with coal mines, it is often found in other types of
subsurface openings that are underlain or overlain with carbonaceous or oil-bearing strata. The
methane is retained within fractures, voids and pores in the rock either as a compressed gas or
adsorbed on mineral (particularly carbon) surfaces. When the strata is pierced by boreholes or
mined openings, then the gas pressure gradient that is created induces migration of the methane
towards those openings through natural or mining-induced fracture patterns. The phenomena of
methane retention within the rock and the mechanisms of its release are discussed in Chapter 12.
Methane itself has no odour but is often accompanied by traces of heavier hydrocarbon gases in
the paraffin series that do have a characteristic oily smell. As indicated in Table 11.1, methane
has a density that is a little over half that of air. This gives rise to a dangerous behaviour pattern methane can form pools or layers along the roofs of underground openings. Any ignition of the
gas can then propagate along those layers to emission sources (Section 12.4.2). The buoyancy
of methane can also create problems in inclined workings.
Methane burns in air with a pale blue flame. This can be observed over the lowered flame of a
safety lamp at concentrations as small as 1 percent. In an abundant supply of air, the gas burns
to produce water vapour and carbon dioxide.
CH4 + 2O2
2H2O + CO2
(11.1)
Unfortunately, within the confines of mined openings and during fires or explosions, there may be
insufficient oxygen to sustain full combustion, leading to formation of the highly poisonous carbon
monoxide.
2CH4 + 3O2
4H2O + 2CO
(11.2)
The explosible range for methane in air is normally quoted as 5 to 15 percent1 by volume, with
the most explosive (stochastic) mixture occurring at about 9.8 percent. While the lower limit
remains fairly constant, the upper explosive limit reduces as the oxygen content of the air falls.
The flame will propagate through the mixture while it remains within the flammable range. Figure
11.1 illustrates a well known diagram first produced by H. F. Coward in 1928. This can be used to
track the flammability of air:methane mixtures as the composition varies. In zone A, the mixture is
not flammable but is likely to become so if further methane is added or that part of the mine is
sealed off. In zone B, the mixture is explosive and has a minimum nose value at 12.2 percent
oxygen. Zones C and D illustrate mixtures that may exist in sealed areas. A mixture in zone C will
become explosive if the seals are breached and the gases intermingle with incoming air.
However, dilution of mixtures in zone D can be accomplished without passing through an
explosive range.
The flammability limits (and particularly the UEL) for any flammable gas or vapour are not absolute
constants but vary with the conditions under which the ignition occurs. The UEL values of 14% and 15% that
are often quoted for methane encompass the range that has been observed for most tests involving upward
propagation of the flame. The type and energy of the igniting source (e.g. open flame or temperature and
area of heated surface also affect the results). Although ignition temperatures of sources are often quoted,
it is the temperature to which the flammable mixture is raised that really matters. The pressure of the mixture
has only a small effect. Values of explosibility limits, quoted without specifying the test conditions, should be
regarded as average values. It also prudent to assume the wider measured limits of flammability for
regulatory purposes and in guidelines for practice.
11-7
24
E : i mp ossib l e m ixt ur e in a ir
22
Oxygen per cent
20
A: may
become
explosive
18
B: explosive
16
14
12
C: may become explosive

when diluted by air
10
8
D: cannot become
explosive
6
4
2
0
0
10
12
14
16
18
20
Methane per cent

Figure 11.1 The Coward diagram for methane in air.
Mining law specifies actions that must be taken when certain fractions of the lower flammable
limit have been reached. For example, legislation may require that electrical power must be
switched off when the methane concentration exceeds 1 or 1 percent. Personnel other than
those concerned with the improvement of ventilation should leave any area when the methane
concentration exceeds 2 or 2 percent. The legislation of the relevant country or state must be
consulted for the precise values of limiting concentrations and required actions. Other regulations
specify the frequency at which measurements of methane concentration must be taken and
threshold limit values to be applied in specified locations.
In many countries, underground mines are classified as either gassy or non-gassy. These legal
terms relate to the potential for methane to be emitted into the workings. It is prudent that all
underground coal mines be designated as gassy. Any other mine may become legally gassy if
(a) methane emissions from the roof, floor or sides of openings have been observed or the
surrounding strata are deemed capable of producing such emissions,
(b) a methane ignition has occurred in the past, or
(c) the mine is connected underground to another mine that has already been classified as gassy.
Imposing a gassy designation on a mine can result in a significant increase in capital and
operating costs as all equipment and operating procedures must then be designed and
maintained to minimize the risk of igniting a methane:air mixture.
11-8
11.2.5. Carbon dioxide, CO2

Carbon dioxide appears in subsurface openings from a variety of sources including strata
emissions, oxidation of carbonaceous materials, internal combustion engines, blasting, fires,
explosions and respiration. Stagnant mixtures of air in sealed off areas often have an increased
concentration of carbon dioxide and decreased oxygen content. Such mixtures are sometimes
called blackdamp.
Table 11.1 indicates that carbon dioxide is more than fifty percent heavier than air and will,
therefore, tend to collect on the floors of mine workings. This was the reason for the small
animals depicted in some of Agricola's woodcuts of sixteenth century mine workings. It is
common to find emanations of blackdamp from the bottoms of seals behind which are abandoned
workings, particularly during a period of falling barometric pressure. Particular care must be taken
when holing through into a mined area that has not been ventilated for some time.
In addition to diluting oxygen in the air, carbon dioxide acts as a stimulant to the respiratory and
central nervous systems. The solubility of carbon dioxide is about twenty times that of oxygen.
Diffusion of the gas into the bloodstream is rapid and the effects on rate and depth of breathing
are soon noticed. Cylinders of oxygen used for resuscitation often contain some 4 percent of
carbon dioxide to act as a respiratory stimulant (carbogen gas).
The physiological effects of carbon dioxide have been listed by Strang and MacKenzie-Wood
(1985) as follows:
Percent carbon dioxide in air

2
Effects
0.037 0.038
None, normal concentration of carbon dioxide in air.
0.5
Lung ventilation increased by 5 percent
2.0
Lung ventilation increased by 50 percent.
3.0
Lung ventilation doubled, panting on exertion.
5 to 10
Violent panting leading to fatigue from exhaustion, headache.
10 to 15
Intolerable panting, severe headache, rapid exhaustion and

collapse.
Fortunately, the administration of oxygen accompanied by warmth and an avoidance of exertion

will usually lead to recovery with no known long term effects.
11.2.6. Carbon monoxide, CO
The high toxicity of carbon monoxide coupled with its lack of smell, taste or colour make this one
of the most dangerous and insidious of mine gases. It has a density very close to that of air and
mixes readily into an airstream unless it has been heated by involvement in a fire, in which case it
may layer with smoke along the roof.
Carbon monoxide is a product of the incomplete combustion of carbonaceous material. Although
colourless, it has the traditional name of whitedamp. The great majority of fires and explosions in
2
The percentage of carbon dioxide in air has been increasing over the past century, currently at a rate of
some 1.5 parts per million (0.00015 percent) each year. The rise is superimposed on an annual cyclic
variation of about 5 ppm (0.0005 percent) caused by seasonal reactions of plant life.
11-9
mines produce carbon monoxide. Indeed, most fatalities that have occurred during such incidents
have been a result of carbon monoxide poisoning. The mixture of gases, including carbon
monoxide, resulting from a mine explosion, is often referred to as afterdamp. Carbon monoxide is
also formed by internal combustion engines, blasting and spontaneous combustion in coal mines.
It can be generated as a component of water gas (carbon monoxide and hydrogen) when water is
applied to the base of incandescent coal during firefighting operations.
Carbon monoxide burns with a blue flame and is highly flammable, having a wide range of
flammability, 12.5 to 74.2 percent in air, with the maximum explosibility at 29 percent.
In order to understand the physiological effects of carbon monoxide, let us recall from Section
11.2.2. that oxygen passes through the walls of alveoli in the lungs and is absorbed by
haemoglobin (red cells)in the blood to form the fairly unstable oxyhaemoglobin. Unfortunately,
haemoglobin has an affinity for carbon monoxide that is about three hundred times greater than
that for oxygen. To compound the problem, the new substance formed in the bloodstream,
carboxyhaemoglobin (CO.Hb), is relatively stable and does not readily decompose. The
consequences are that very small concentrations of carbon monoxide cause the formation of
carboxyhaemoglobin which accumulates within the bloodstream. This leaves a reduced number
of red cells to carry oxygen molecules throughout the body. The physiological symptoms of
carbon monoxide arise because of oxygen starvation to vital organs, particularly the brain and
heart.
Physiological reactions to carbon monoxide depend upon the concentration of the gas, the time of
exposure and the rate of lung ventilation, the latter being governed primarily by physical activity.
In order to relate the symptoms of carbon monoxide poisoning to a single parameter, the degree
of saturation of the blood by carboxyhaemoglobin is employed. Although variations exist between
individuals, the following list provides a guideline of the progressive symptoms.
Blood saturation
percent CO.Hb
5-10
Symptoms
Possible slight loss of concentration
10-20
Sensation of tightness across forehead, slight headache
20-30
Throbbing headache, judgment impaired.
30-40
Severe headache, dizziness, disorientation, dimmed vision, nausea,

possible collapse.
40-60
Increased probability of collapse, rise in rates of pulse and respiration,

convulsions
60-70
Coma, depressed pulse and respiration, possible death.
70-80
Fatal.
Figure 11.2 gives a more practical guide of physiological reactions to carbon monoxide and takes
the level of physical activity and exposure time into account.
Due to the bright red colour of carboxyhaemoglobin, a victim of carbon monoxide poisoning may
have a flushed appearance, even after death. At high blood saturation levels, deterioration of the
body is often retarded after death due to the absence of internal oxygen.
11-10
70
65
1000 ppm
2000 ppm
death
Blood saturation (% Hb.CO)
60
750 ppm
55
50
500 ppm
45
400 ppm
unconsciousness
40
300 ppm
35
30
200 ppm
25
intoxication
20
100 ppm
15
10
50 ppm
no symptoms
0
0
2
1
3
2
9 10 11 12 13 14 15 16 17 18 19 20 21 22 resting
4
3
6
4
Time of exposure
7
5
10 moderate work
9
6
heavy work
(hours)
Figure 11.2 Physiological effects of carbon monoxide (developed from Strang and Wood, 1985).
[The three horizontal axes represent hours of exposure for each of three levels of activity.]
Persons suffering from carbon monoxide poisoning should be kept warm and removed from the
polluted atmosphere, preferably, on a stretcher. It may take more than twenty four hours for blood
saturation levels to return to normal. During this time, severe headaches may be experienced.
However, the return to normal blood saturation levels can be accelerated significantly by the
administration of pure oxygen. The rapidity with which carbon monoxide is absorbed into the
bloodstream and the slowness of its expurgation can result in dangerous blood saturation levels
occurring in firefighters who make short repeated expeditions into a polluted area.
In addition to the physical symptoms that have been listed for carbon monoxide poisoning,
experience of personnel involved in mine fires has indicated significant psychological reactions
that have had grave repercussions. Low levels of blood saturation can give an appearance of
intoxication including impairment of judgment and an unsteady gait preceding collapse. Victims
may become silent and morose, and may resist or fail to comprehend instructions that will lead
them to safety. The sense of time may be affected, a particularly significant symptom when selfrescuer devices are being worn. However, all of these reactions vary considerably between
individuals. In particular, total collapse will occur rapidly in high concentrations of carbon
monoxide.
11-11
A small degree of acclimatization has been observed in persons who are repeatedly exposed to
low levels of carbon monoxide as experienced, for example, by habitual smokers. This is thought
to occur because of an increase in the number of red cells within the bloodstream.
11.2.7. Sulphur dioxide, SO2
This is another highly toxic gas but one which, fortunately, can be detected at very low
concentrations both by its acidic taste and the intense burning sensation it causes to the eyes
and respiratory tracts. The latter is a result of the high solubility of the gas in water to form
sulphurous acid.
H2O + SO2
H2SO3
This, in turn, can oxidize to sulphuric acid, H2SO4.

When discussing very small concentrations, it is convenient to refer to parts per million (ppm)
rather than percentages. The conversion between the two is accomplished simply by moving the
decimal point four places in the appropriate direction. The following list of physiological reactions
to sulphur dioxide employs parts per million.
Concentration of
sulphur dioxide ppm
Effects
Acidic taste
Detectable by odour.
20
Irritation of eyes and respiratory system.
50
Severe burning sensation in eyes, nose and throat.
400
Immediately dangerous to life.
First aid for sulphur dioxide poisoning includes the administration of oxygen, immobility and
warmth. The longer term treatment is for acid corrosion of the eyes and respiratory system.
Sulphur dioxide is formed in internal combustion engines and by the oxidation of sulphide
ores, for example:
or
zinc blende
2ZnS + 3O2
iron pyrites
4FeS2 + 11O2
2ZnO + 2SO2
(11.3)
2Fe2O3 + 8SO2
(11.4)
These and other similar reactions can occur when sulphide ores are heated in a fire or by
spontaneous combustion. Although sulphur dioxide is colourless, white fumes may be seen due
to condensation of acidic water vapours or traces of sulphur trioxide, SO3.
11.2.8. Oxides of nitrogen, NOx
Three oxides of nitrogen are listed on Table 11.1. Nitric oxide, NO, nitrous oxide, N2O, and
nitrogen dioxide, NO2, are formed in internal combustion engines and by blasting. The proportion
of nitrous oxide is likely to be small. Furthermore, nitric oxide converts rapidly to nitrogen dioxide
in the presence of air and water vapour.
2NO + O2
2NO2
(11.5)
11-12
As nitrogen dioxide is the most toxic of these oxides of nitrogen, it is sensible to concentrate upon
the physiological effects of this gas.
At the temperatures found in underground openings, it is probable that nitrogen dioxide will be
mixed with a companion gas, nitrogen tetroxide, N2O4, having similar physiological effects. The
brown fumes of nitrogen dioxide dissolve readily in water to form both nitrous (HNO2) and nitric
(HNO3) acids.
2NO2 + H2O
HNO2 + HNO3
(11.6)
These acids cause irritation and, at higher concentrations, corrosive effects on the eyes and
respiratory system. The progressive symptoms are as follows:
Effects
Concentration of
nitrogen dioxide
ppm
40
May be detected by smell.
60
Minor throat irritation
100
Coughing may commence.
150
Severe discomfort, may cause pneumonia later.
200
Likely to be fatal
The immediate treatment for nitrogen dioxide poisoning is similar to that for sulphur dioxide,
namely, the administration of oxygen, immobility and warmth. An insidious effect of nitrogen
dioxide poisoning is that an apparent early recovery can be followed, soon afterwards, by the
development of acute bronchopneumonia.
11.2.9. Hydrogen sulphide, H2S
The presence of this highly toxic gas is readily detected by its characteristic smell of bad eggs.
This has given rise to the colloquial name stinkdamp. Unfortunately, hydrogen sulphide has a
narcotic effect on the nervous system including paralysis of the olfactory nerves. Hence, after a
short exposure, the sense of smell can no longer be relied upon.
Hydrogen sulphide is produced by acidic action or the effects of heating on sulphide ores. It is
formed naturally by bacterial or chemical decomposition of organic compounds and may often be
detected close to stagnant pools of water in underground mines. Hydrogen sulphide may occur in
natural gas or petroleum reserves and migrate through the strata in a weakly acidic water
solution. It can also be generated in gob fires. In such cases, free sulphur may be deposited by
partial oxidation of the gas.
2H2S + O2
2H2O + 2S
(11.7)
This can sometimes be seen as a yellow deposit in burned areas. However, in a plentiful supply
of air, hydrogen sulphide will burn with a bright blue flame to produce sulphur dioxide.
2H2S + 3O2
2SO2 + 2H2O
(11.8).
11-13
The physiological effects of hydrogen sulphide may be listed as follows:

Concentration of hydrogen sulphide ppm
Effects
0.1 to 1
Detectable by smell.
Beginning of toxicity
50 to 100
Slight irritation to eyes and respiratory tract,

headache, loss of odour after 15 minutes.
200
Intensified irritation of nose and throat.
500
Serious inflammation of eyes, nasal secretions,

coughing, palpitations, fainting.
600
Chest pains due to corrosion of respiratory

system, may be fatal.
700
Depression, coma, probable death.
1000
Paralysis of respiratory system, very rapid

death.
A victim who recovers from hydrogen sulphide poisoning may be left with longer term
conjunctivitis and bronchitis.
11.2.10. Hydrogen, H2
Although non-toxic, hydrogen is the most explosive of all the mine gases. It burns with a blue
flame and has the wide flammable range of 4 to 74.2 percent in air. Hydrogen can be ignited at a
temperature as low as 580 C, and with an ignition energy about half of that required by methane.
Hydrogen occasionally appears as a strata gas and may be present in afterdamp at about the
same concentrations as carbon monoxide. The action of water on hot coals can produce
hydrogen as a constituent of water gas (Section 11.2.6.). Dangerous accumulations of hydrogen
may occur at locations where battery charging is in progress. Hydrogen has a density only some
0.07 that of air. It will, therefore, tend to rise to the roof. Battery charging stations should be
located in intake air with a duct or opening at roof level that connects into a return airway.
11.2.11. Radon, Rn
This chemically inert gas is one of the elements formed during radioactive disintegration of the
uranium series. Although its presence is most serious in uranium mines, it may be found in many
other types of underground openings. Indeed, seepages of radon from the ground into the
basements of surface buildings have been known to create a serious health hazard.
Radon emanates from the rock matrix or from ground water that has passed over radioactive
minerals. It has a half life of 3.825 days and emits alpha radiation. The immediate products of the
radioactive decay of radon are minute solid particles known as the radon daughters. These
adhere to the surfaces of dust particles and emit alpha, beta and some gamma radiation.
Radon and the problems of radiation in mines are discussed further in Chapter 13.
11-14
11.3. GAS MIXTURES

Although the gases that occur most commonly underground have been discussed separately in
Section 11.2., it is more usual that several gaseous pollutants appear together as gas mixtures.
Furthermore, the importation of an ever-widening range of materials into subsurface facilities also
introduces the risk of additional gases being emitted into the mine environment.
11.3.1. Threshold limit values for gas mixtures
In order to widen the applicability of this section, the recommended threshold limit values of a
number of other gases and vapours are given in Table 11.3 and grouped in terms of probable
sources. Again, we are reminded that these TLVs are simply guidelines and that national or state
mandated limits should be consulted to ensure compliance with the relevant laws.
In any given atmosphere, if there are two or more airborne pollutants (gaseous or particulate) that
have adverse effects on the same part of the body, then the threshold limit value should be
assessed on the basis of their combined effect. This is calculated as the dimensionless sum.
C1
C
C
+ 2 + ...... n
T1
T2
Tn
(11.9)
where C = measured concentration

and
T = corresponding threshold limit value.
If the sum of the series exceeds unity, then the threshold limit value of the mixture is deemed to
be exceeded. (ACGIH, 1989)
Example
An analysis of air samples taken in a return airway indicates the following gas concentrations:
carbon dioxide: 0.2 percent,
hydrogen sulphide: 2 ppm,
carbon monoxide: 10 ppm

sulphur dioxide: 1 ppm
Determine the threshold limit values, TWA and STEL for the mixture.
Solution
Sulphur dioxide and hydrogen sulphide are both irritants to the eyes and respiratory system.
Furthermore, carbon dioxide is a stimulant to breathing, increasing the rate of ventilation of the
lungs. Hence, it may be regarded as being synergistic with the sulphur dioxide and hydrogen
sulphide. Carbon monoxide, however, affects the oxygen carrying capacity of the bloodstream
and need not be combined with the other gases in determining the threshold limit values for the
mixture.
From Table 11.1, the following time weighted averages (TWA) and short term exposure
limits (STEL) are determined:
Component
Threshold limit values

TWA
STEL
carbon dioxide
0.5 percent
3.0 percent
hydrogen sulphide
10 ppm
15 ppm
sulphur dioxide
2 ppm
5 ppm
carbon monoxide
50 ppm
400 ppm
11-15
Table 11.3 Threshold limit values for other gases and vapours that may be present underground
Substance
Guideline time weighted average limit

(ppm unless otherwise stated)
Cleaners and solvents

acetone
ammonia
toluene)
turpentine
750 (STEL = 1000)

25 (STEL = 35)
100 (STEL = 150)
100 (STEL= 150)
Refrigerants
ammonia
R11
R12
R22
R112
25 (STEL = 35)
1000 (ceiling)
1000
1000
500
Fuels
butane
gasoline vapour
liquid petroleum gas
naptha (coal tar))
pentane
propane
800
300 (STEL = 500)
1000
100
600 (STEL = 750)
1000
Welding and soldering

general welding fumes
iron oxide fumes
lead fumes
ozone (arc welding)
fluorides (fluxes)
5 mg/m3
5 mg/m3
0.15 mg/m3
0.1 ppm ceiling
2.5 mg/m3
Heated plastics
carbon dioxide
carbon monoxide
hydrogen chloride fumes
hydrogen cyanide
hydrogen fluoride
phenol (absorbed by skin)
0.5% (STEL = 3.0%)

50 (STEL = 400)
5 ppm ceiling
10
3 ppm ceiling
5
Explosives
carbon dioxide
carbon monoxide
oxides of nitrogen
ammonia
sulphur dioxide
nitrous acid fumes
nitric acid fumes
0.5% (STEL = 3.0%)

50 (STEL = 400)
25
25 (STEL = 35)
2 (STEL = 5)
2 (STEL = 4)
Others
chlorine (biocides
cresol (wood preservative)
mercury vapour
oil mist (mineral): vapour free
oil mist (vegetable): vapour free
sulphuric acid fumes (batteries)
0.5 (STEL = 1)
5
0.05 mg/m3
5 mg/m3
10 mg/m3
1 mg/m3
11-16
The carbon monoxide concentration measured at 10 ppm and treated separately is shown to be
less than the TWA of 50 ppm.
The equivalent threshold limit value of the remainder of the mixture is assessed from equation
(11.9) as:
TWA:
0.2
0 .5
carbon
dioxide
STEL:
0 .2
3
2
10
hydrogen
sulphide
+
2
15
1
= 1. 1
2
sulphur
dioxide
+
1
= 0 .4
5
The dimensionless TWA is greater than 1. Hence, the time weighted average limit is exceeded.
However, the short term exposure limit is less than unity. The implication is that personnel may
spend time periods not exceeding 15 minutes in this atmosphere, but not a complete 8 hour shift.
In some cases of gas mixtures, it may be practicable to monitor quantitatively for one pollutant
only, even though it is known that other gases or particulates are present. In such circumstances,
a pragmatic approach is to reduce the threshold limit value of the measured substance by a factor
that is assessed from the number, toxicity and estimated concentrations of other contaminants
known to be present.
11.3.2. Diesel emissions
The flexibility and reliability of diesel engines have resulted in a proliferation in their use for all
types of underground mines. This is, however, tempered by the emissions of exhaust gases, heat
and humidity that result from the employment of diesels. Guidance on the airflow requirements for
diesel equipment is given in Section 9.3.2.
The substances that are emitted from diesel exhausts include:
nitrogen
carbon monoxide
carbon dioxide
oxides of nitrogen
sulphur dioxide
diesel particulate matter
water vapour (Sections 15.3.2.2. and 16.2.3.)
The actual magnitude and composition of diesel exhausts are governed by the engine design,
quality of maintenance, exhaust treatment units, rating for altitude and skill of the operator.
Catalytic converters cause the exhaust gases to filter through granulated oxidizing agents. These
can successfully convert up to 90 percent of the carbon monoxide and 50 percent of unburned
fuel to carbon dioxide and water vapour. Sulphur dioxide may be partially converted to sulphur
trioxide and appear as vapour of sulphuric acid.
11-17
Water scrubbers give an improved removal of sulphur dioxide and particulate matter but do little
for carbon monoxide. Engine gas recirculation (EGR) systems help to reduce the oxides of
nitrogen.
The diesel particulate matter is a combination of soot, unburned fuel and aldehydes, and is
regarded as being the component of diesel exhaust that is most hazardous to health. There are
two reasons for this. First, the particles are typically less than one micron in diameter. They are
inhaled deeply into the lungs and have a high probability of being retained within the walls of the
alveoli. Secondly, the porous or fibrous nature of the soot particles enables them to adsorb a
range of polynuclear and aromatic hydrocarbons, giving diesel smoke its characteristic greasy
feel. There is evidence that these have carcinogenic properties (Waytulonis, 1988) although the
relationship between exposure to diesel exhaust emissions and the incidence of cancer remains
unclear (French, 1984). Exhaust filters assist greatly in the removal of particulate matter.
The operating advantages of diesel equipment are in direct conflict with the potentially hazardous
nature of exhaust emissions. This has resulted in a considerable research effort into the use of
diesels in mining during the 1980's (e. g. World Mining, 1982; Daniel, 1989; Haney, 1989;
Gunderson, 1990).
11.3.3. Fires
The predominant cause of loss of life associated with underground fires has been the gaseous
products of combustion. The particular gases depend upon the type of fire and the materials that
are being burned. Although carbon monoxide has resulted in the majority of such fatalities, the
importation of a widening variety of manufactured materials into mines has produced the potential
for other toxic gases to be emitted during a mine fire.
Coal mine fires that involve burning of the coal itself are likely to produce an atmosphere that is
deficient in oxygen and may contain
carbon dioxide
carbon monoxide
methane
water vapour
and smaller amounts of sulphur dioxide, hydrogen sulphide and hydrogen. Fires involving timber
emit the same gases. The actual concentration of each gas depends primarily upon the oxygen
content of the air within the fire zone (Chapter 21).
Metal mine fires arising from the spontaneous combustion of sulphide ores will emit products of
combustion that are rich in sulphur dioxide and, possibly, sulphuric acid vapour.
Fires involving diesel equipment will produce gases from burning diesel oil and also from any
plastic or rubber components that may become heated. The burning oil itself will emit
carbon monoxide
sulphur dioxide
hydrogen
carbon dioxide
oxides of nitrogen
hydrogen sulphide
Plastic materials are employed for an increasing variety of purposes in underground workings,
including brattice cloths, machine components, electrical cables and fittings, thermal insulation on
pipes or airway linings, the pipes themselves, instrumentation, containers and meshing. Some
plastic polymers will begin to disintegrate at about 250C and before actual combustion
11-18
commences. However, the evolution of gases increases rapidly when the material burns. While
all heated plastics give off
carbon monoxide
and
carbon dioxide
the most dangerous are the polyurethanes, nylon and polyvinyl chloride (PVC) which may also
produce
hydrogen cyanide
hydrofluoric acid fumes
hydrochloric acid fumes
The phenolic plastics evolve the same gases but require a higher temperature for pyrolysis.
Rubber based materials may also emit hydrogen sulphide when heated. Local or national
regulations may prohibit the use of certain plastics.
In addition to gases, fires will normally also produce copious amounts of the respirable
particulates that appear as visible smoke. These particles themselves may be toxic and are
irritating to the eyes and respiratory tracts.
11.3.4. Explosives
The gases that may be produced by blasting are listed in Table 11.3. The concentrations of gases
depend upon the type, quality and weight of explosive used, the means of detonation and the
psychrometric condition of the air. The degree of confinement also affects the concentrations and
time distribution of the gas emission - firing "on the solid" produces a sharper and higher peak of
blasting fumes than if a free face were available. Some of the gases and fumes may be adsorbed
onto mineral faces within the fragmented rock and be emitted over a longer time period (Rossi,
1971).
The oxides of nitrogen are formed mainly as nitric oxide during the detonation. The rate at which
this oxidizes to nitrogen dioxide depends upon the degree to which the blasting fumes are diluted
by a ventilating air current. Fumes that are purged from a mine during a re-entry period following
blasting may still contain a significant proportion of nitric oxide when the pulse of fumes exits the
mine.
11.3.5. Welding
The constituents of welding fumes can vary widely dependent upon the metals or alloys involved,
the welding process employed and any fluxes that may be used. The fumes contain particles of
amorphous slags and oxides of the metals being welded.
Arc welding in a local inert atmosphere such as argon will inhibit the formation of welding fumes
but can produce ozone. On the other hand, arc welding within the air can produce copious fumes
and carbon monoxide. Where stainless steel is involved, compounds of chromium and nickel can
appear as constituents of the welding fumes. The employment of welding fluxes may also result
in fluorides as well as a range of oxides being generated.
If no toxic compounds are present in the metals, welding rods or metal coatings, and the
conditions do not favour the formation of toxic gases, then the threshold limit of 5 mg/m3 given for
general welding fumes in Table 11.3 may be used as a guideline.
11-19
11.4. GAS DETECTION AND MONITORING

11.4.1. Objectives and overview
The primary purpose of monitoring the concentrations of airborne pollutants in a mine is to ensure
that the atmosphere provides a safe environment free from levels of toxicants that would create a
hazard to health. There are essentially three matters to consider. The first is the threshold limit
values deemed to be acceptable for each pollutant. Second is the choice of instrumentation best
suited for the detection and measurement of particular gases. Third, the question of where and
how frequently measurements are required must be addressed. Having discussed threshold limit
values in Sections 11.2.1. and 11.3.1., we now turn our attention to the operating principles of
instruments used to measure gas concentrations and methods of sampling.
11.4.2. Principles of gas detection
Advances in the fields of electronics, electrochemistry and micro-manufacturing have resulted in
significant improvements in the accuracy and reliability of equipment for the detection and
measurement of gas concentrations. Instruments now available are capable of measuring
concentrations of more than one gas and indicating fractions of a part per million for some toxic
gases. The same principles of detection method may be applied to more than one gas. Hence, in
this section we shall concentrate on those principles rather than discussing each gas in turn. The
methods of gas detection and monitoring used most frequently for subsurface gases are listed in
Table 11.1.
11.4.2.1. Filament and catalytic oxidation (pellistor) detectors
These devices are used primarily for the measurement of methane and other gases that will burn
in air such as carbon monoxide, hydrogen or the higher gaseous hydrocarbons. If an electrical
filament is heated to a sufficiently high temperature, then a combustible gas in the surrounding air
will burn and, hence, elevate the temperature of the filament even further. This will change the
electrical resistance of the filament. By arranging for the filament to act as one arm of a
Wheatstone bridge circuit, its change of resistance can be sensed as a change in voltage drop or
current that is proportional to the concentration of combustible gas. A second filament of the
Wheatstone bridge is also exposed to the air but inhibits oxidation of the gas either by a lower
operating temperature or catalytic poisoning. This acts as a balancing control against variations in
the temperature, moisture content and barometric pressure of the ambient air. The device is,
essentially, a resistance thermometer sensing the temperature of the active filament. Filament
detectors have been available since at least the 1950's.
Platinum has been used as the filament coil material as it has good resistance-temperature
characteristics at the 900 to 1000C temperatures required to promote the oxidation process in a
stable manner. There are, however, severe drawbacks to platinum filament detectors. First, the
response of the instrument is very sensitive to the geometry of the coil. Slight variations in the
pitch of the coil caused during manufacture or due to mechanical shock produce significant
deviations in the electrical output. Secondly, at the required high operating temperature,
evaporation of the metal causes the cross sectional area of the wire to be reduced and, hence,
increases its resistance. This is reflected by a slow but significant increase in bridge output for
any given gas concentration and a zero drift in fresh air.
During the 1960's, research was carried out in a number of countries to overcome the difficulties
inherent in filament methanometers. The platinum filament was reduced to less than 1 mm in
length and encased in a tiny bead of refractory material, such as alumina. The outside of the
bead was coated by a thin layer of catalytic material that produces stable oxidation of methane at
11-20
temperatures of some 500C. The catalysts are, typically, metal-salt combinations of palladium
and thorium or platinum and rhodium. These beads are called pellistors or pelements (Baker,
1969; Richards, 1970) and overcome most of the problems of the earlier filament detectors. First,
the encapsulation of the coil makes it almost immune to mechanical damage. Indeed, further
developments of electronic materials have enabled the coil to be eliminated. Secondly, there is no
evaporation of the coil and, hence, near elimination of zero drift. Furthermore, the life of the unit is
greatly increased while the reduced operating temperature requires less battery power and
improves the safety of the instrument. The balancing or reference pellistors to compensate for air
temperature, humidity and pressure are identical units except that the catalyst is poisoned by
dipping it in a hot solution of potassium hydroxide. The first commercial methanometers to utilize
pellistors appear to have been produced by Mine Safety Appliances (MSA).
Modern pellistor methanometers have a high degree of reliability, and operate satisfactorily and
continuously on mining machines within conditions of high humidity, dust and vibration. The
bridge output signals may be used to indicate on a dial or digital register, activate audio/visual
alarms, isolate electrical power to a machine, drive a recorder or transmit to distant devices
through a telemetering system. However, they still retain some disadvantages.
First, pellistors rely upon the oxidation process and, hence, the availability of oxygen. A good
quality, heavy duty, pellistor will indicate increasing concentrations of methane up to 9 or 10
percent. Beyond that, the decreasing concentration of oxygen will diminish the rate and
temperature of catalytic oxidation giving a reduced and false reading. Pellistor methanometers
are, in the main, manufactured for the range 0 to 5 percent. However, use in an oxygen-deficient
and methane-rich atmosphere could, again, give a false reading - apparently within the 0 to 5
percent scale. In some modern instruments, this promotes a warning signal or is combined with
another mode of detection for high concentrations of combustible gas. Another possible
consequence of repeated exposures to high concentrations of combustible gases is that
"cracking" of a hydrocarbon gas can deposit carbon within the catalytic layer resulting in scaling
and eventual destruction of the pellistor. The catalyst should contain an inhibitor to minimize the
effects of cracking. The choice of refractory material that forms the bead also influences the
degree of cracking.
A second disadvantage of pellistors is that they are subject to poisoning by some other gases and
vapours. Vaporized products of silicon compounds (greases, electrical components) or phosphate
esters (fluid couplings) will produce permanent poisoning of pellistors while halogens from
refrigerants or heated plastics can give a temporary reduction of the instrument output. The
readings of pellistor methanometers should be treated with caution when such instruments are
used downstream from a mine fire. Most of these instruments allow the ambient air to reach the
pellistor heads by diffusing through a layer of absorbent material such as activated charcoal. This
removes most potential poisons and also permits the instrument to be near independent of the air
velocities generally encountered underground.
A third feature of pellistor transducers is that they react to any combination of combustible gases
that pass through the absorbent filter. The instruments used in mines are primarily calibrated for
methane as this is the most common of the combustible gases found in the subsurface. However,
hydrogen, carbon monoxide and ethane will also pass through simple activated filters and
produce a response from the instrument. At 1 percent concentration of each gas, a catalytic
combustion methanometer will indicate the readings shown on Table 11.4. The readings remain
proportional to those given by methane up to the lower explosive limit of each gas.
11-21
Gas
Methanometer reading
percent
methane
1.00
carbon monoxide
0.39
hydrogen
1.24
ethane
1.61
propane
1.96
Table 11.4. Approximate readings on a catalytic combustion methanometer given by a 1 percent

concentration of each gas.
Fortunately, a pellistor methanometer is unlikely to be used in concentrations of carbon monoxide
that would be indicated on the instrument. Furthermore, hydrogen, ethane and propane all give a
reading above that for methane. Hence, the error is on the side of safety.
Example
A methanometer used in a mixture of air and ethane gives a reading of 3.4 percent. What is the
actual concentration of ethane? Assume that there are no other combustible gases present.
Solution
From Table 11.4, the correction factor for ethane is 1.61. Hence,
actual concentration of ethane =
3.4
= 2.1 per cent
1.61
A useful feature of oxidation methanometers is that any given reading indicates approximately the
same fraction of the lower flammable limit of the combustible gas being monitored. Hence, a
pellistor methanometer would indicate 5 per cent (approximately) if any one gas from Table 11.4
were present in a concentration equal to its lower flammable limit. The same is true of mixtures of
the gases, enabling a pellistor methanometer to be used as an indicator of how close a gaseous
mixture is to its lower flammable limit.
11.4.2.2. Flame safety lamps
These lamps were introduced early in the nineteenth century for the purposes of providing
illumination from an oil flame without igniting a methane-air mixture (Section 1.2). Their use for
illumination disappeared with the development of electric battery lamps. However, the devices
were retained for the purposes of testing for methane and oxygen deficiency.
A blue halo of burning gas over the lowered flame of the lamp becomes visible at about 1
percent methane. The size of the halo increases with methane concentration. At 2 percent, it
forms a very clear equilateral triangle. At 5 percent the flame spirals upwards into the bonnet of
the lamp and either continues burning or self extinguishes in a small contained explosion. In both
cases, the flame is prevented from propagating into the surrounding atmosphere by the tightly
woven wire gauzes (Section 1.2). The use of the flame safety lamp for methane testing has now
largely been replaced by pellistor methanometers as these are more accurate, reliable and safer.
Some coal mining industries have retained the flame safety lamp for its ability to indicate oxygen
deficiency (Section 11.2.2.). The height of the flame reduces progressively with oxygen content
and is extinguished at 16 percent oxygen. However, in a methane-rich atmosphere, the flame
may remain lit down to an oxygen concentration of 13 percent.
11-22
Various attempts were made to convert modified flame lamps into alarm and recording devices
(for example, Pritchard, 1961). However, some ignitions of methane have been attributed to flame
safety lamps that have been damaged or inadequately maintained (Strang, 1985). The flame
safety lamp has an honourable place in the history of mine ventilation but its role is almost over.
11.4.2.3. Thermal conductivity and acoustic gas detectors
At 20C and at normal atmospheric pressures, the thermal conductivity of methane is 0.0328
W/m C compared with 0.0257 W/m C for air. This difference is utilized in some high range
methanometers. Two heated sensors are employed, one exposed to the gas sample and the
other retained as a reference within a sealed air-filled chamber. A sample of the ambient air is
drawn through the instrument at a constant rate. The sample sensor cools at a greater rate due to
the higher thermal conductivity of the methane. The change in resistance of the sample sensor is
detected within an electrical bridge to give a deflection on the meter. Typical ranges for a thermal
conductivity methanometer are 2 or 5 percent to 100 percent. The principle is sometimes used in
conjunction with pellistors to provide a dual range instrument or to over-ride the falsely low
readings that may be given by catalytic combustion detectors at high concentration (Section
11.4.2.1.). To reduce interference from other gases, suitable filters should be employed. In
particular, carbon dioxide gives about half the response of methane but in a negative direction.
For use in mining, a soda lime filter is, therefore, advisable.
Acoustic gas detectors rely upon changes in the velocity of sound as the composition of a sample
varies. It is used, primarily, for high concentrations and has been employed in methane drainage
systems.
11.4.2.4. Optical methods
These subdivide into three groups. Interferometers utilize the refraction of light that occurs when
a parallel beam is split, one half passing through the sample and the other through a sealed
chamber containing pure air. The two beams are recombined and deflected through a mirror or
glass prism arrangement for viewing through a telescope. The optical interference between the
two beams causes a striped fringe pattern to appear in the field of view. This typically takes the
form of two black lines in the centre with red and green lines on both sides. The fringe moves
along a scale in proportion to the amount of gas present. Rotational adjustments of one of the
deflecting prisms can be read on a vernier and added to the optical scale in order to widen the
range of the instrument. Interferometers are sensitive to the presence of other gases and
appropriate filters should be used when necessary. Hydrogen gives a negative response with
respect to methane. At equal concentrations of the two gases, the reading is near zero. Carbon
dioxide and methane give similar responses. For air containing both carbon dioxide and methane
two readings are taken, one with a soda lime filter to remove the carbon dioxide and gives the
methane concentration only, and the second without that filter. The difference between the two
readings is an indication of the carbon dioxide concentration. The effects of other gases render
the interferometer unsuitable for situations where the composition of the sample is dramatically
different from that of normal air, for example, afterdamp, downstream from fires or samples taken
from behind seals.
The nondispersive infra-red gas analyser is one form of absorption spectrometer that is frequently
used for mine gas analysis. Identical beams of slowly pulsating infra-red radiation pass
through two parallel chambers, one containing a gas that does not absorb infra-red
(typically nitrogen) and the other fed by a stream of the sample. The pulsations of 3 to 4 Hertz
may be achieved electronically or by a rotating chopper arrangement. Beyond the sample and
reference chambers is a two-compartment sealed container (detector unit) filled with a pure
specimen of that particular gas which is to be detected. Detector units can be interchanged to
determine the concentrations of different gases. The two compartments of the detector unit are at
11-23
the same nominal pressure and are separated by a flexible diaphragm. The pulses of infra-red
radiation are directed sequentially into the two sides of the detector unit, heat the contained gas
and, hence, raise its pressure. However, the beam that has passed through the sample chamber
has already been partially absorbed at the relevant wavelength by molecules of the sought gas.
Hence, the pulses of pressure induced in that corresponding side of the detector unit are
weakened. The amplitude of the vibrating diaphragm is sensed by an electrical capacitor and
translated electronically to an output signal. Infra-red gas analysers are employed primarily at
fixed monitoring stations or laboratories. However, portable versions are available.
Laser spectroscopy is another means of air analysis that has considerable potential for
subsurface application. There are two systems that can be employed. One is the differential
absorption unit (DIAL system) in which two similar lasers are used, one tuned to the absorption
wavelength of the gas to be detected, and the other to a slightly different wavelength. The two
laser beams pass through the sample gas stream and are reflected back to a single receiver unit.
The difference in the two signals is processed to indicate the concentration of the gas being
monitored. The employment of a reference beam eliminates the effects of dust, humidity or other
gases. (Holmes and Byers, 1990).
The second laser technique for gas analysis is the light detection and ranging (LIDAR) method
which depends upon the Raman effect. When a gas is excited by monochromatic radiation from a
laser, a secondary scattered radiation is produced. The spectrum of this scattered radiation can
be analysed to indicate the concentrations of the gases that caused it.
The attraction of laser techniques for mine air sampling is that the laser beams may be directed
across or along subsurface openings to give continuous mean analyses for large volumes of air.
Furthermore, the lasers can also be used to monitor air velocity.
11.4.2.5. Electrochemical methods
Very small concentrations of many gases can be detected by their influence on the output from
an electrochemical cell. There are two primary types, both based upon oxidation or reduction of
the gas within a galvanic cell. The cell has at least two electrodes and an intervening electrolyte.
In the polarographic or voltametric analyser, a voltage from an external battery is applied across
the electrodes in order to induce further polarization (or retardation) of the electrodes. The gas
sample is supplied to the interface between the electrolyte and one of the electrodes (the
"sensing electrode). This may be accomplished by diffusion of the sample gas through a
hollowed and permeable sensing electrode. The electrochemical reaction at the
electrode/electrolyte interface changes the rate at which free electrons are released to flow
through the electrolyte and be collected by the "receiving electrode". For example, in a sulphur
dioxide electrochemical analyser, the oxidation is as follows:
SO2 + 2H2O
SO4 2 + 4H+ + 2e
The resulting change in electrical current is proportional to the concentration of sulphur dioxide in
the sample.
In amperometric cells, the gas reacts directly with the electrolyte and, hence, enhances or
reduces the current produced. Both types of cell are subject to interference by other gases. This
is minimized by appropriate selection of the materials employed for the electrodes and electrolyte,
the polarizing voltage applied to the polarographic cells and a suitable choice of filters.
Electrochemical cells may indicate the partial pressure rather than the concentration of a gas.
These require the zero to be reset when taken through a significant change in barometric
pressure - as will occur when travelling to different levels in a mine. Another disadvantage of
electrochemical analysers is that they can become temporarily saturated when exposed to high
concentrations of gas. The recovery period may be several minutes.
11-24
11.4.2.6. Mass spectrometers

In these instruments, the gas sample passes through a field of free electrons emitted from a
filament or other source. Collision of the electrons with the gas molecules produces ions, each
with a mass/charge ratio specific to that gas. The ions are accelerated by electromagnets and
then pass through a magnetic deflection field which separates them into discrete beams
according to their mass/charge ratios. The complete mass spectrum can be scanned and
displayed on an oscilloscope or the signals transmitted to recorders.
11.4.2.7. Paramagnetic analysers
Oxygen is one of the very few gases that are paramagnetic, i.e. it aligns itself as a magnetic
dipole in the presence of an applied magnetic field and, hence, creates a local anomaly within
that field. This property is utilized in a paramagnetic oxygen analyser. A weak permanent dumbbell magnet is suspended against a light applied torque within a non-uniform magnetic field. One
end of the dumb-bell is encapsulated within a bulb of nitrogen while the other end is exposed to
the gas sample. A rotation of the magnet is induced as the oxygen content of the sample varies.
The movement is amplified optically or electrically for display or recording.
11.4.2.8. Gas chromatography
Gas chromatographs are used widely for the laboratory analysis of sampled mixtures of gases.
Portable units are also manufactured. An inert carrier gas is pumped continuously through one or
more columns (or coils) which contain gas adsorbents. The latter may be granulated solids or
liquids. A small pulse of the sample gas mixture is injected into the line upstream from the
columns. The constituent gases are initially adsorbed by the column materials. However, the
continued flow of the carrier gas causes subsequent desorption of each gas at a time and rate
dependent upon its particular adsorption characteristics. The result is that the gases leave the
adsorbent columns as discrete and separated pulses. Their identification and measurement of
concentration is carried out further downstream by one or more of the detection techniques
described in this section.
11.4.2.9. Semiconductor detectors
One of the more recent techniques introduced for gas detection involves passing the sample over
the surface of a semi conducting material which is maintained at a constant temperature.
Adsorption of gas molecules on to the surface of the semiconductor modifies its electrical
conductance. Selectivity of the gas to be detected may be achieved by the choice of
semiconductor and the operating temperature. However, filters may be necessary to avoid
interference or poisoning from other airborne pollutants. Coupled with the development of thin film
technology, the semiconductor technique holds promise for increased future utilization.
11.4.2.10. Stain tubes
Stain tubes are used widely for a large variety of gases. They are simply glass phials containing a
chemical compound that changes colour in the presence of a specified gas. The phials are sealed
at both ends. To carry out a test, the ends are snipped open and a metered volume of sample air
pulled through at a constant rate, usually by means of a simple hand pump. The concentration of
gas is estimated either from the length of the stain or by comparing the colour of the stain with a
chart. Despite the lack of precision, stain tubes are used extensively in practice because of their
simplicity, portability and low cost.
11-25
11.4.3. Methods of sampling

The methods of sampling for subsurface gases vary from judicious positioning of the human nose
to sophisticated telemetering systems. The techniques most commonly applied can be divided
into two classes: (a) manual and (b) automatic or remote.
11.4.3.1. Manual methods
The locations and times at which hand-held equipment should be employed to measure methane
concentrations in gassy mines are normally mandated by the governing legislation. These include
pre-shift inspections by qualified persons prior to entry of the work force, and at intervals of some
twenty minutes throughout the shift on a working face.
The instruments that are most widely used for manual detection and measurement are catalytic
oxidation methanometers, oxygen deficiency meters and stain tubes. Multiple gas instruments are
commonly employed. Light gases, including methane, are most likely to collect at roof level while
carbon dioxide will tend to pool in low lying areas
areas. Hence, measurements should be taken in
those locations in addition to within the general body of air. Extension probes provide a means of
drawing a sample from a location that is unsafe or out of reach.
Grab samples are volumes of air or gas mixtures that are collected in sample containers
underground for subsequent laboratory analysis. Typically, the sample is drawn into a metal or
plastic container by means of a hand-operated pump or by water displacement. This is usually
the method employed for retrieving samples from behind seals or stoppings. The sample pipe
should extend far enough beyond the seal to prevent contamination by air that has breathed
through the seal during preceding periods of rising barometric pressure. Sufficient gas should be
drawn from the pipe prior to sampling in order to ensure that the captured gas is representative of
the atmosphere beyond the seal. Care should be taken to prevent loss or pollution of the sample
caused by pressure changes as the container transported to the surface. The gas seals on the
containers should be well maintained and samples should be sent to a laboratory for expeditious
analysis.
Personal samplers are devices that are worn continuously by personnel while at work. The
simplest are badges that change colour in the presence of selected gases or radiation. More
sophisticated samplers for dust or gases draw power from either an internal battery or a cap lamp
battery. Personal gas samplers may emit audio-visual alarm signals when a preselected
concentration of gas is exceeded.
11.4.3.2. Automatic and remote monitors
Permanent environmental monitors may be sited at strategic locations throughout a subsurface
ventilation system. The gases most frequently subject to this type of monitoring are methane,
carbon monoxide and carbon dioxide. Other gases may be monitored in workshops, storage
repositories and for special applications such as hydrogen at the roofs of battery charging
stations. Environmental monitors in an underground repository for nuclear waste must also be
provided to detect airborne radionuclides.
Permanent monitors should be mounted at strategic locations chosen on a site-specific basis.
However, it is prudent to site transducers at the intake and return ends of working areas and at
intervals along return routes. Carbon monoxide sensors mounted at intervals along conveyor
roads provide an earlier warning of fire than temperature-sensitive devices.
11-26
In addition to providing local audio-visual alarms, an environmental monitoring system operates

most effectively when it is integrated into a telemetering network. The signals from gas, airflow,
pressure and temperature transducers are transmitted to a remote computer and control station,
usually on surface, where those signals are analysed for trends, recorded on magnetic media,
operate alarms when appropriate and, possibly, generate feedback control signals to fans, doors
or motorized regulators. (See, also, Sections 9.6.3 and 9.6.4.) Monitors at critical locations such
as booster fans should be installed as dual or, even, triple units to safeguard against instrument
failure. Rechargeable batteries can provide backup in the event of a cut in electrical power. A
system of planned maintenance should be implemented in order to ensure continuous and
reliable operation. A computer-controlled telemetering system should be capable of detecting and
identifying failed monitors.
Machine mounted gas monitors should be provided on rock breaking equipment to test for
methane in coal and other gassy mines. These devices must be particularly rugged in order to
remain operational when subjected to dust, water sprays, vibration and impact. Machine-mounted
methanometers should not only provide audio-visual warnings but also be connected such that
the electrical power supply to the machine is isolated when the methane concentration reaches
one percent (or other concentration as required by the relevant mining legislation). Here again,
machine-mounted monitors may be integrated into a mine telemetering system.
Tube bundle systems are a slower alternative means of remote sampling. Air is drawn through
plastic tubing from chosen locations in the mine to monitoring stations which may be located
either underground or on surface. At a monitoring station, automatic valves on the tubes are
operated in a cyclic manner in order to draw samples from the tubes sequentially, and to pass
those samples into gas detection units. The latter are often infra-red analysers (Section 11.4.2.4.)
connected in series to monitor the concentrations of each of the required gases. A gas pump on
each tube maintains continuous flow. However, the main disadvantage of the tube bundle system
is the travel time of the gas between entering a tube and analysis. It is recommended that this
should not exceed one hour. While surface monitoring stations are preferred, workings distant
from shaft bottoms may necessitate local underground stations. Tube bundle systems are not
suitable for emergency situations that occur quickly such as an equipment fire. However, they
have proved valuable for the identification of longer term trends including incipient spontaneous
combustion.
The tubes should be fitted with dust filters and water traps. Variations of temperature and
pressure may cause condensation as the air passes through the tubes. Care must be taken
during the installation of a tube bundle system, particularly at joints, as leaks are difficult to detect
and locate. An important advantage of the system is that it requires no electrical power inby the
monitoring station. Hence, it remains operational when sampling from behind seals or when no
electrical supply is available.
References
ACGIH (1989). Threshold limit values and biological exposure indices for 1989-1990. American
Conference of Governmental Industrial Hygienists, Cincinnati, Ohio, USA.
Baker, A. R. (1969). A review of methanometry. The Mining Engineer, Vol. 128, No. 107, Aug.,
pp 643-653.
Baker, A. R. and Firth, J. G. (1969). The estimation of firedamp: Applications and limitations of
the pellistor. The Mining Engineer, Vol. 128, No. 100, Jan., pp 237-244.
Coward, H. F. and Jones, G. W. (1952). Limits of flammability of gases and vapours. Bull. US
Bureau of Mines, No. 503.
11-27
Coward, H. F. (1928). Explosibility of atmospheres behind stoppings. Trans. Inst. Mining

Engineers, Vol. 77, p 94.
Daniel, J. H. (1989). Diesels - backbone of a changing mining industry. 4th U. S. Mine Ventilation
Symposium, Berkeley, California USA, pp 561-568.
Forbes, J. J. and Grove, G. W. (1954). Mine gases and methods for detecting them. US Bureau
of Mines, M. C., No. 33.
French, I. W. and Assoc. Ltd. (1984). Health implications of exposure of underground mine
workers to diesel exhaust emissions - an update. Dept. of Energy, Mines and Resources, Ottawa,
Canada, p 607.
Gunderson, R. E. (1990). The advantages of electrically powered equipment for deep hot mines.
Journal of the Mine Ventilation Society of S. Africa, Vol. 43, No. 6, June, pp 102-111.
Haney, R. A. and Stolz, R. T. (1989). Evaluation of personal diesel particulate samplers. 4th US
Mine Ventilation Symposium, Berkeley, California, pp 579-587.
Holmes, J. M. and Byers, C. H. (1990). Countermeasures to airborne hazardous chemicals.
Published by Noyes Data Corp., New Jersey USA.
NIOSH (1987). Pocket guide to chemical hazards. National Institute for Occupational Safety and
Health; US Dept. of Health and Human Services, Washington, D. C. USA.
Pritchard, F. W. and Phelps, B. A. (1961). A recorder of atmospheric methane concentration
based on a butane flame lamp. Colliery Engineering, 120, Feb. 9th, pp 163-166.
Richards, G. O. and Jones, K. (1970). Some recent work on the detection and determination of
firedamp. The Mining Engineer, Vol. 130, No. 121, Oct., pp 31-40.
Rossi, B. D. (1971). Control of noxious gases in blasting work and new methods of testing
industrial explosives (translated from Russian). NTIS, Tech. translation no. TT70-50163.
Strang, J. and MacKenzie-Wood, P. (1985). A manual on mines rescue and gas detection.
Published by Weston and Co., Kiama, Australia.
US Bureau of Mines (1987). Diesels in underground mines. Proceedings of technology transfer
seminar, pp 1-165.
Waytulonis, R. W. (1988). Bureau of Mines diesel research high lights. 4th Int. Mine Ventilation
Congress, Brisbane, Australia, pp 627-633.
World Mining/World Coal (1982). Diesels in Mining. Special issue, Nov., pp 6-54
11-28
Chapter 12 Methane
CHAPTER 12
METHANE
12.1. OVERVIEW AND ADDITIONAL PROPERTIES OF METHANE ..................2

12.2. THE RETENTION AND RELEASE OF METHANE IN COAL ......................3
12.2.1. Gas retention in coal ........................................................................................................ 3
12.2.2. The release of methane from coal ................................................................................... 7
12.2.2.1. Diffusion and Darcy flow. .......................................................................................... 7
12.2.2.2. Desorption kinetics .................................................................................................... 8
12.2.3. Determination of gas content......................................................................................... 11
12.2.3.1. Indirect method (adsorption isotherms)................................................................... 11
12.2.3.2. Direct measurement method................................................................................... 12
12.3. MIGRATION OF METHANE ......................................................................13

12.3.1. Fluid Flow through a permeable medium. ..................................................................... 14
12.3.1.1. Incompressible flow................................................................................................. 14
12.3.1.2. Compressible flow ................................................................................................... 16
12.3.1.3. Radial flow of gas .................................................................................................... 17
12.3.1.4. Transient radial flow ................................................................................................ 18
12.3.2. The permeability of coal................................................................................................. 19
12.3.2.1. Effect of mechanical stress ..................................................................................... 19
12.3.2.2. Effect of gas pressure ............................................................................................. 21
12.3.2.3. Two phase flow ....................................................................................................... 22
12.4. EMISSION PATTERNS INTO MINE WORKINGS .....................................23

12.4.1. Sources of methane in coal mines................................................................................. 23
12.4.2. Methane layering ........................................................................................................... 25
12.4.3. Gas outbursts................................................................................................................. 28
12.4.3.1. In-seam outbursts.................................................................................................... 28
12.4.3.2. Outbursts from roof and floor .................................................................................. 29
12.4.4 Prediction of methane emissions into the ventilation system ......................................... 31
12.5. METHANE DRAINAGE..............................................................................32

12.5.1. In-seam drainage ........................................................................................................... 32
12.5.2. Gob drainage by surface boreholes............................................................................... 35
12.5.3. Cross-measures methane drainage .............................................................................. 36
12.5.4. Drainage from worked-out areas ................................................................................... 37
12.5.5. Components of a methane drainage system ................................................................. 38
12.5.5.1. Pipe ranges ............................................................................................................. 38
12.5.5.2. Monitors................................................................................................................... 39
12.5.5.3. Controls and safety devices .................................................................................... 40
12.5.5.4. Extractor Pumps...................................................................................................... 41
12.5.6. Planning a methane drainage system ........................................................................... 42
12.5.6.1. Data collection and estimation of gas capture ........................................................ 42
12.5.6.2. Design of the drainage network .............................................................................. 43
12.5.7. Utilization of drained gas ............................................................................................... 46
References ........................................................................................................47
12 -
Chapter 12 Methane
12.1. OVERVIEW AND ADDITIONAL PROPERTIES OF METHANE

The major properties and behavioural characteristics of methane were outlined in Section 11.2.4.
That Section should be reviewed as an introduction to the more detailed treatment given in this
chapter.
There are three primary reasons for giving particular attention to methane. First, it is the naturally
occurring gas that most commonly appears in mined underground openings. Secondly, it has
resulted in more explosions and related loss of life than any other cause throughout the recorded
history of mining. The flammability characteristics of this gas have been studied since, at least,
the time of Agricola in the 16th Century. The number and severity of coal dust explosions initiated
by methane ignitions declined after the development of electric cap lamps and the replacement of
shaft bottom furnaces with fans (Section 1.2.). However, the intensity and wider deployment of
mechanized rock-breaking equipment resulted in a renewed incidence of frictional ignitions of
methane following the 1960s (Richmond et al, 1983). Fortunately, modern standards of
ventilation and dust control prevent most of these from developing into the larger dust explosions
and disastrous loss of life.
The third reason for giving special attention to methane concerns the continued development of
methane drainage technology. Although the primary reason for extracting methane at high
concentration from the strata around mines continues to be the reduction of methane emissions
into these mines, a growing incentive has been the drainage of methane to provide a fuel source
in its own right. Methane drainage may now be undertaken for this purpose from strata where
there is no known intention of subsequent mining.
In this chapter we shall consider four broad areas; first, the manner in which methane is retained
within the strata and the mechanisms of its release when the rock is disturbed by mine workings
or boreholes. We shall also outline means of determining the gas content of carbonaceous strata.
Secondly, we shall examine the migration of the gas from its geologic sources towards workings
or boreholes. This will encompass an analysis of strata permeability and flow through fracture
networks. The third consideration is the dynamic pattern of methane emission into active mine
workings, varying from normal cyclic variations, through gas layering phenomena, to outburst
activities.
Finally, we shall classify the major techniques of methane drainage. No one of these has
universal application and the selection of drainage method must be made with great care
according to the geology of the area, the methods and layout of mining (if any) and the natural or
induced physical properties of the rocks.
First, however, let us list some further properties of methane in addition to those given in Table
11.1. The data are referred, wherever applicable, to standard temperature and pressure of 0C
and 101.3 kPa.
H
molecular structure :
H
melting point :
boiling point :
critical temperature and pressure :
latent heat of vaporization (at 111.3K) :
90.5 K
111.3 K
190.5 K and 4.63 MPa
508.2 kJ/kg
12 -
Chapter 12 Methane
latent heat of fusion (at 90.5 K) :
specific heat Cp :
specific heat Cv :
isentropic index :
solubilitv in water :
upper calorific value :
lower calorific value :
thermal conductivity :
dynamic viscosity :
58.8 kJ/kg
2184 J/(kg K)
1680 J/(kg K)
1.300
55.6 litres per m3 of water (33.1 litres/m3 at 20C)
55.67 MJ/kg
50.17 MJ/kg
0.0306 W/(m C) [0.0328 W/(m C) at 20C]
(10.26 + 0.0305t) x 10-6 Ns/m2
(temperature t = 0 to 100C)
12.2. THE RETENTION AND RELEASE OF METHANE IN COAL

In order to understand the manner in which methane is retained within coal and the mechanisms
of its release, it is first necessary to comprehend the internal structures of coal. The existence of
a large number of exceptionally small pores and the corresponding high internal surface area of
coal has been recognized for many years and numerous conceptual models have been
suggested. However, electron microscopy has revealed the actual structures that exist.
Coal is not a single material but a complex mix of fossilised organic compounds and minerals.
The composition and structure depend upon the nature of the original vegetation from which a
given coal seam formed, the timing and turbidity of the water flows involved in the sedimentary
processes of deposition and the metamorphic effects of pressure, temperature and tectonic
stressing over geological time. Electron micrographs have, indeed, confirmed the existence of
pores, many of which may not be interconnected. An example is shown on Plate 2. However,
other areas may indicate amorphous, granular, sponge-like or fibrous structures, even within the
same seam. Furthermore, the coal substance is intersected by a fracture network with apertures
varying from those that are comparable to pore diameters through to some large enough to be
seen with the naked eye.
In this Section we shall concentrate on the manner in which methane is-held within the coal, the
kinetics of its release when the geological equilibrium is disturbed by mining or drilling, and how
the gas content of a seam may be determined.
12.2.1. Gas retention in coal
Methane exists within coal in two distinct forms, generally referred to as free gas and adsorbed
gas. The free gas comprises molecules that are, indeed, free to move within the pores and
fracture network. Porosities of coals have been reported from 1 to over 20 percent. However,
those values depend upon the chosen definition of porosity and the manner in which it is
measured. Absolute (or total) porosity is the total internal voidage divided by the bulk volume of a
sample and may be difficult to measure accurately. Effective (or macroscopic) porosity is the ratio
of interconnected void space to the bulk volume. The latter definition is more useful in the
determination of recoverable gas from a seam. Unfortunately, the usual procedures of measuring
porosity depend upon saturating the internal voidage with some permeating fluid. The probability
of a gas molecule entering any pore or interconnection rises as its molecular diameter decreases
or its mean free path increases. (The mean free path is the statistical average distance between
collisions of the gaseous molecules.) Furthermore, any adsorptive bonding between the fluid and
the solid may obstruct the narrower interconnections. The measured effective porosity is,
therefore, dependent upon the permeating fluid. In order to handle the extremely small pores that
exist in coal, the maximum value of effective porosity is obtained by using helium as the
permeating fluid. Helium has a small molecular diameter (some 0.27 x 10-9 m), a relatively large
mean free path (about 270 x 10-9 m at 20C and atmospheric pressure) and is non-adsorptive
with respect to coal. For comparison, pore diameters of eastern U.S. coals vary from over
30x10-9 to less than 1x10-9 m (Gan, 1972).
12 -
Chapter 12 Methane
Plate 2. Electron micrograph of a bituminous coal (after Harpalani).
An attractive force exists between the surfaces of some solids and a variety of gases. Coal
surfaces attract molecules of methane, carbon dioxide, nitrogen, water vapour and several other
gases. Those molecules adhere or are adsorbed onto the coal surface. When the adsorptive
bond exceeds the short-distance repulsive force between gas molecules (Section 2.1.1.), then the
adsorbed molecules will become packed together as a monomolecular layer on the surface. At
very high gas pressures, a second layer will form with a weaker adsorptive bond (Jolly, 1968).
Figure 12.1 illustrates the variation of both total and adsorbed methane with respect to gas
pressure and at constant temperature. The curve illustrating adsorbed methane is known as the
adsorption isotherm. Some 95 percent of the total gas will, typically, be in the adsorbed form,
explaining the vast reserves of methane that are contained within many coal seams.
12 -
Chapter 12 Methane
Gas content cu.m/t at STP
35
total gas
dry
ash-free
20C
30
25
adsorbed gas
20
15
10
5
free gas
0
0
Gas pressure MPa
Figure 12.1 Examples of adsorbed, free and total gas isotherms for methane in coal.
The most widely used mathematical relationship to describe adsorption isotherms was developed
by I. Langmuir (1916 and 1918) for a monomolecular layer. This may be expressed as
q
bP
=
q max
1+ bP
where q
qmax
P
b
=
=
=
=
(12.1)
volume of gas adsorbed at any given pressure (m3/t at NTP)
the maximum amount of gas that can be adsorbed as a monomolecular
layer at the prevailing temperature (m3/t at NTP)
gas pressure (usually expressed in MPa)
Langmuirs constant (MPa-1 ), a function of the adsorptive bond
between the gas and the surface (1/b is the pressure at which q/qmax = 1/2)
The values of Langmuirs constant depend upon the carbon, moisture and ash contents of the
coal as well as the type of gas and prevailing temperature. Boxho et al (1980) reports values of b
decreasing from 1.2 MPa-1 at a volatile content of 5 percent to 0.5 MPa-1 at a volatile content of
40 percent. Figure 12.2(a) indicates the increased adsorptive capacities of the higher rank coals,
these having greater values of carbon content. Values of qmax are shown to vary from about 14 to
over 30 m3/t. The effect of the type of gas is illustrated on Figure 12.2(b).
Adsorption isotherms are normally quoted on a dry, ash-free basis. The amount of methane
adsorbed decreases markedly at small initial increases in moisture content of the coal. Most of
this natural moisture is adsorbed on to the coal surfaces. However, adsorptive saturation of water
molecules occurs at about 5 percent moisture above which there is little further decrease in
methane content. A widely used approximation is given by Ettinger (1958):
q moist
1
=
q dry
1+ 0.31h
where qmoist
qdry
h
=
=
=
(12.2)
gas content of moist coal (m3/t)

gas content of dried coal (m3/t)
moisture content (percent) in the range 0 to 5 percent.
(Assume 5 percent for greater moisture contents)
12 -
Adsorbed methane cu.m/t at STP
Chapter 12 Methane
40
30
anthracite
low volatile bituminous
20
high volatile bituminous
10
0
0
Gas pressure MPa
Adsorbed gas cu.m/t at STP
Figure 12.2(a) Examples of adsorption isotherms for methane in coal. The amount of
gas adsorbed increases with the carbon content of the coal.
40
35
carbon dioxide
30
25
methane
20
15
nitrogen
10
5
0
Gas pressure MPa

Figure 12.2(b) Adsorption isotherms for carbon dioxide, methane and nitrogen in a
bituminous coal at 25C.
12 -
Chapter 12 Methane
The mineral matter that comprises the ash constituent of coal is essentially non-adsorbing. Hence
the methane content decreases as the percentage of ash rises (Barker-Read, 1989). In order to
express gas content on an ash-free basis, a simple correction may be applied.
q actual = q ash free (1 0.01a )
where a
(12.3)
ash content (percent
12.2.2. The release of methane from coal

In the undisturbed state, equilibrium exists between free gas and adsorbed gas in the pores and
fracture networks of coal. If, however, the coal seam is intersected by a borehole, or disturbed by
mining, then the gas pressure gradient that is created will result in flow through natural or stressinduced fractures. The resulting reduced gas pressure in the pores will promote desorption. The
process will move from right to left along the appropriate adsorption isotherm. A glance at Figures
12.1 or 12.2 indicates that the rate of desorption with respect to gas pressure increases as the
pressure falls. The migration of methane from its original location is retarded by narrow and
tortuous interconnections between pores and microfractures, obstruction by adsorbed molecules
and the resistance offered by the fracture network.
12.2.2.1. Diffusion and Darcy flow.
Current thinking is that two types of flow take place. Within the micropore structure, diffusion flow
occurs while, in the fractures, laminar flow dominates. Let us first discuss diffusion flow. This
occurs whenever a difference in concentration of molecules of a given gas occurs. The simplest
means of quantifying diffusion flow is by Ficks Law:
ux = D
C
x
m/s
(12.4)
where ux
=
velocity of diffusion in the x direction (m/s)
D
=
coefficient of diffusion (m2/s but often quoted in cm2 /s)
C
=
concentration of the specific gas (m3/m3 of coal)
(The negative sign is necessary as movement occurs in the direction of decreasing concentration.)
Bulk diffusion is the normal gas-to-gas diffusion that occurs because of a concentration gradient
in free space (see Appendix A15.4). However, within the confines of a micropore structure, two
other forms of diffusion may become effective. Surface diffusion arises from lateral movement of
the adsorbed layer of gas on the coal surfaces while Knudson diffusion occurs due to transient
molecular interactions between the gas and the solid.
Laminar flow in the fracture network follows Darcys Law (1856) for permeable media.
u =
k P
x
where u
=
k
=
=
P/x =
m/s
(12.5)
gas velocity (m/s)

permeability (m2 )
dynamic viscosity (Ns/m2 )
pressure gradient (Pa/m).
(Note the analogy between Darcys Law and Fouriers Law of heat conduction, equation (15.4).)
Darcys Law is essentially empirical and assumes that the gas velocity at the surface is zero
(Section 2.3.3). For narrow passages, slippage at the walls (surface diffusion) may become
significant and can be taken into account by adjusting the value of permeability (Klinkenberg
effect, Section 12.3.2.2.).
12 -
Chapter 12 Methane
Controversy has existed on whether diffusion flow or Darcy flow predominates in coal.
Disagreements in the findings of differing researchers are probably a consequence of the wide
variations that exist in coal structure. It would appear that diffusion flow governs the rate of
degassing coals of low permeability while Darcy flow in the fracture network is the dominant effect
in high permeability coals.
12.2.2.2. Desorption kinetics
A newly exposed coal surface will emit methane at a rate that decays with time. Figure 12.3(a)
illustrates a desorption curve. This behaviour is analogous to the emission of heat (Section
15.2.2.). A number of equations have been suggested depending, primarily, on the conceptual
model adopted for the structure of coal. For a spherical and porous particle with single sized and
non-connecting capillaries, work by Wheeler (1951) and Carslaw (1959) leads to the series
q (t )
6
= 1 2
q max
where
fraction adsorbed q(t)/q max
q(t)
qmax
D
d
=
=
=
=
n
n =1
1
2
4 D n 2 2t
exp
d2
(12.6)
volume of gas (m3) emitted after time t (s)

total volume of gas in particle
coefficient of diffusion (m2/s)
equivalent diameter (m) of particle (= 6 x volume/surface area)
0.9
0.8
0.7
0.6
theoretical
actual
0.5
0.4
0.3
0.2
0.1
0
1000
2000
3000
4000
5000
6000
Time (s)
Figure 12.3(a) Example of a methane desorption curve plotted against time.

A more practicable approximation is given by Boxho et al (1980) as follows:
q (t )
q max
4 2 D t

1 exp
2
(12.7)
12 -
Chapter 12 Methane
Both of these equations approximate to:

q (t )
12
q max
d
Dt
12 D t
(12.8)
d2
while for q(t)/qmax < 0.25, the relationship can be simplified further to
q (t )
12
q max
d
Dt
(12.9)
This latter equation implies that during the early stages of desorption the emission rate is
proportional to
rate against
determined.
t . This is illustrated on Figure 12.3(b). Furthermore, plotting the initial emission
t allows the corresponding value of the coefficient of diffusion, D, to be
1
theoretical
initial
slope
0.9
actual
fraction desorbed q(t)/q max
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
50
100
time
150
200
Figure 12.3(b) A methane desorption curve plotted against the square root of time.
12 -
Chapter 12 Methane
Unfortunately, there is a major problem associated with all of these equations. The coefficient of
diffusion, D, is not a constant for any given sample. It is dependent upon
t he size range of individual particles
t he distribution of pore sizes
t he number, sizes and tortuosity of fractures (this, in turn, depending upon the stress history
of the sample)
temperature
pressure and current gas content.
The latter factors are particularly troublesome as these, and hence D, will vary as the degassing
proceeds. Lama (1987) has suggested the relationship
D = Do
where P
Do
=
=
m2
s
P
Po
(12.10)
current value of pressure (Pa) and

coefficient of diffusion at some original (seam)pressure, Po (m2/s)
The variations in D contribute to the changes in apparent permeability of a coal sample when
subjected to mechanical stress or variations in mean gas pressure (Section 12.3.2.). Considering
the sensitivity of D to a significant number of variables, it is not altogether surprizing that a wide
range of values have been reported, from 1 x 10-8 to 1 x 10-14 m2/s. In order to take the particle
size into account, it has become common to express the diffusivity as the ratio D/a2 (s-1) where
a = particle radius (m).
A further weakness is that the fundamental equation (12.6) is based on uniformly sized
capillaries. This leads to significant deviations between predicted and observed desorption rates
at the longer times (q(t)/qmax > 0.6) as indicated on Figure 12.4 (Smith and Williams, 1984).
The difficulties encountered in the derivation and general applicability of analytical desorption
equations led E. M. Airey (1968) to propose an empirical relationship:.
t
q (t )
= 1 exp
q max
t0
where t0
n
=
=
(12.11)
'time constant' = time for 63 percent of the gas to desorb (s) and
an index that varies from some 1/3 for bituminous coals to 1/2 for
anthracites and dependent, also, upon the degree of fracturing.
Both t0 and n are influenced by the range and magnitude of particle sizes and, hence, should be
quoted with reference to a specific size spectrum. For example, values of t0 for Welsh anthracite
lie, typically, between 1500 and 2100 seconds for the size range 0.05 to 0.30 mm at 25 C
(Barker-Read, 1989). Smaller particles will give reduced values of both t0 and n.
An improved desorption time parameter, , that is independent of the gas capacity of a sample
has been proposed by Radchenko (1981) and Ettinger et al (1986) (see, also, Barker-Read and
Radchenko (1989) and may be related to Airey's constants:
t n
= 60 1 exp 0
60
sec
(12.12)
12 - 10
Chapter 12 Methane
It can also be shown that
a2
sec
36 D
(12.13)
where a = radius of the particle (m)

The value of can be gained from the desorbed fraction v
1/(initial slope)2
sec
time curve (e. g. Figure 12.3(b)):
(12.14)
Equations (12.13 and 12.14) provide a ready means of determining D/a2 from a desorption test.
12.2.3. Determination of gas content
There are two reasons for measuring the methane contents of coal seams and associated strata.
First, such data are required in the assessment of methane emissions into mine workings and,
hence, the airflows required to dilute those emissions to concentrations that are safe and within
mandatory threshold limit values (Sections 9.3.1. and 11.2.1.). Secondly, the gas content of the
strata is a required input for computer models or other computational procedures to determine the
gas flows that may be obtained from methane drainage systems.
There are also two distinct approaches to the evaluation of gas content, the indirect method
which employs adsorption isotherms and the direct measurement method which relies upon
observations of gas release from newly obtained samples. Let us consider each of these in turn.
12.2.3.1. Indirect method (adsorption isotherms)
In this technique, representative chippings are obtained from the full thickness of the seam, mixed
and ground to a powder of known particle size range. The sample is dried at a temperature of not
more than 80C and evacuated to remove the gases that remain in the pore structure.
Unfortunately, the process of evacuation may alter the internal configuration of the pore structure
due to the liquefaction of tars (Harpalani, 1984). The phenomenon can be overcome by
evacuating at low temperature submerging the sample container into liquid nitrogen (-150C)
prior to and during the evacuation.
After returning to the desired ambient temperature, methane is admitted in stages and the gas
pressure within the sample container recorded at each increment. The volume of gas admitted
may be monitored at inlet (volumetric method) or by measuring the increase in weight of the
sample and container (gravimetric method). After correcting for free space in the sample
container, a plot of cumulative methane admitted against pressure provides the total gas
isotherm. If the porosity of the sample has been determined separately (Section 12.2.1.), then the
results can be divided into free gas and adsorbed gas as illustrated on Figure 12.1. The
isotherms should, ideally, be determined at, or close to, the virgin rock temperature of the actual
strata. If necessary, the curves can be corrected to seam temperature. At any given pressure, the
quantity of gas adsorbed falls as the temperature increases. Starting at 26C, the gas adsorbed
decreases by some 0.8 percent per C for bituminous coal and 0.6 percent per C for anthracite
(Boxho et al, 1980). The curves should also be corrected to the actual moisture content of the
seam using Ettinger's formula, equation (12.2).
In order to utilize the corrected gas isotherm, a borehole is drilled into the seam, either from the
surface or from an underground location. In the latter case, the hole must be sufficiently long (10
to 20 m) in order to penetrate beyond the zone of degassing into the mine openings. Seals are
12 - 11
Chapter 12 Methane
emplaced in the borehole to encapsulate a representative length within the seam. A tube from the
encapsulated length of borehole is attached to a pressure gauge and the rise in pressure is
monitored. The rate of pressure rise will be greater for coals of higher permeability. The initial rate
of pressure rise may be used in conjunction with open hole flowrates to determine in-situ
permeability. However, for the purposes of assessing seam gas content, it is the maximum (or
equilibrium) pressure that is required. Using this pressure, the gas content of the seam can be
read from the corrected isotherm curve.
An advantage of the indirect method is that it gives the total gas content of the seam. However,
this is not indicative of the actual gas that may be emitted into mine workings or recoverable by
methane drainage. Furthermore, the measured in-situ gas pressure will be influenced not only by
methane but also by other gases that may be present and, particularly, by the presence of water.
12.2.3.2. Direct measurement method.
This technique involves taking a sample of the seam from a borehole and placing it immediately
into a hermetically sealed container. The gas is bled off to atmosphere in stages and its volume
measured. The process is continued until further gas emissions are negligible.
An early method of direct measurement of gas content was developed in France (Bertrard et al,
1970). This utilized samples of small chippings from the borehole. Further research carried out by
the U. S. Bureau of Mines during the 1970's led to a procedure that used complete cores
(Diamond, 1981). Although developed primarily for surface boreholes, the technique is applicable
also to horizontal holes drilled from mine workings. The sampling personnel must be present at
the time the hole is drilled into the seam. A stopwatch is used to maintain an accurate record of
the elapsed times between which the sample length of strata is penetrated, start of core retrieval,
arrival of the sample at the mouth of the borehole and confinement within the sealed container.
Each container should be capable of holding about 2 kg of core and some 35 to 40 cm in length.
Longer cores should be subdivided as a precaution against major loss of data should one
container suffer from leakage. The seals on containers must be capable of holding a gas
pressure of 350 kPa without leaking. A pressure gauge should be fitted to each container.
Gas is bled off from the container at intervals of time commencing at 15 minutes (or less if a rapid
rise in container pressure is observed). The volume of gas emitted at each stage is measured,
usually by water displacement, in a burette. The time intervals are increased progressively and
the process allowed to continue until an average of not more than 10 cubic cm per day has been
maintained for one week.
There is, inevitably, some gas lost from the core between the time of seam penetration and its
confinement in a sample container. The volume of lost gas may be assessed by plotting the
cumulative gas emitted from the container against time . The initial straight line may be
extrapolated backwards through the recorded elapsed time of core retrieval in order to quantify
the lost gas. The retrieval time can be minimized by using wireline drilling. The technique is
illustrated on Figure 12.4. The analytical background to the procedure is embodied in equation
(12.9). [Should subsequent tests show that the rate of desorption follows a tn law where n
deviates significantly from 0.5 (Equation (12.11)), then the appropriate plot may be constructed as
gas evolved v tn in order to determine the lost gas.]
Another method of estimating the lost gas, developed by Smith and Williams (1984) makes
separate allowance for the elapsed times of drilling through the coal, core retrieval and period
spent at the mouth of the borehole before containment. This more sophisticated technique is
based on an analysis that takes account of the lack of uniformity in pore sizes (Smith and
Williams (1984a), Close and Erwin, (1989).
12 - 12
Chapter 12 Methane
500
Gas emitted cubic cm
400
300
200
100
0
-100
20
40
time
60
80
1
2
-200
-300
-400
Figure 12.4 Assessment of gas lost during core retrieval. In this example, elapsed time before
1
containment, t = 840 seconds (i.e. t = 29 s 2 ). Backward extrapolation of the

initial straight line gives the gas lost as 285 cm3.
The combination of measured gas and lost gas gives an estimate of the maximum amount of gas
that may be emitted into mine workings or captured by methane drainage. (It may be much less
than this.) However, following the termination of gas evolution measurements, the core will still
contain some residual gas. This can be quantified, if required, by crushing the coal and continuing
to measure the additional gas that is then liberated. The crushing process may take place in a
separate sealed ball mill within a nitrogen atmosphere. However, steel balls in the sample
container allow the crushing to take place without removing the sample from the container.
Another method is to activate a steel hammer by electromagnetic means within the container in
order to crush the sample through a fixed grid. In either case, the sample and evolved gas should
be cooled to the original ambient temperature before the gas is bled off.
A variation of the direct measurement method of gas content is the desorbmeter (Hucka, 1983)
in which desorbing methane from coal chippings in a sample vessel pushes a small plug of fluid
along a transparent spiral tube. An electronic version employing the same principle has been
developed in Germany (Janas, 1980).
12.3. MIGRATION OF METHANE
Following release from its long term geologic home, methane will migrate through rock under the
influence, primarily, of a gas pressure gradient. That movement will occur through the coal seam
and if the gas pressure gradient is transverse to the seam, also through adjoining fractured or
permeable strata. We assume that the flow paths and velocities are sufficiently small that laminar
flow exists. Hence, Darcy's Law applies. This was introduced in its simplest form as equation
(12.5). In this Section, we shall consider the further ramifications of Darcys Law.
12 - 13
Chapter 12 Methane
12.3.1. Fluid Flow through a permeable medium.

12.3.1.1. Incompressible flow
Let us consider, first, an incompressible fluid (liquid) passing through a permeable medium in the
direction x as illustrated on Figure 12.5. An elemental volume of the fluid within the medium has a
thickness, dx, and an orthogonal area A. Two forces act upon the element.
First, a force in the direction of flow is exerted because of the pressure differential, dP, across its
two faces. Let us call this Fp, where.
Fp = -A dP
(12.15)
(negative because P decreases in the x direction).

Secondly, a vertical downward force, Fv, exists due to gravitational pull:
Fv = m g
where the mass of the element (density , x volume Adx)
m = A dx
giving
Fv = g A dx
P1
x
dx
P2
Fp
Fv
Figure 12.5 Forces on an elemental volume of fluid, Adx, within a permeable medium.
Total force in the x direction, Ft = Fp + Fv cos .
12 - 14
Chapter 12 Methane
If the direction of flow is at an angle to the vertical then the component of Fv in the x direction is
Fv cos =
g A cos dx
(12.16)
(Note that is also the inclination of the permeable medium to the horizontal.)
The total force on the element in the direction of flow is then
Ft = Fp + Fv cos
= -AdP + g A cos dx
(12.17)
This can be expressed as force per unit volume

Ft
dP
=
+ g cos
Adx
dx
(12.18)
m3
This is a fuller version of the pressure gradient P / x in Darcy's Law, equation (12.5).
Hence, we can rewrite Darcy's Law for an incompressible fluid in a non-horizontal flow field as
ux =
k dP
g cos
dx
m
s
(12.19)
=
fluid velocity in the x direction (m/s)
where ux
k
=
permeability of medium (m2 )
=
dynamic viscosity of fluid (Ns/m2 )
and
dP is negative in the x direction.
In the case of horizontal flow (cos = 0) or where the fluid density is small, then equation (12.19)
reverts to the simple form of Darcy's Law.
u =
k dP
dx
m/s
(12.20)
As the fluid is incompressible, then u remains constant and we can integrate directly between the
bounding walls shown on Figure 12.5, giving
u =
k (P2 P1 )
X
k (P1 P2 )
X
m/s
(12.21)
or, flowrate, Q, across a given orthogonal area, A becomes

Q = uA =
(P1
P2 )
X
m3
s
(12.22)
12 - 15
Chapter 12 Methane
12.3.1.2. Compressible flow

For gases the fluid density is, indeed, low and the gravitational term may be neglected even for
non-horizontal flow. However, the gas will expand as it progresses along the flowpaths due to the
reduction in pressure. Hence the flowrate, Q, and gas velocity, u, will both increase.
Q = uA =
m3
s
dP
dx
(12.23)
where A = given orthogonal area across which the flow occurs (m2)
As Q is a variable, we can no longer integrate directly. However, if we write the equation in terms
of a steady-state (constant) mass flow, M (kg/s), then
M = Q =
kg
s
dP
dx
kg
P
RT
But density =
(12.24)
(General Gas Law, equation 3.11)
m3
where R = gas constant (J/(kg K)) and T = absolute temperature (K)

giving
M =
A P dP
RT dx
kg
s
(12.25)
As M is constant, we can integrate across the full thickness of the medium (Figure 12.5) to give
k A P1 P2
RT
2X
M =
kg
s
(12.26)
This can be converted to a volume flow, Q, at any given density. In particular, at the density, m,
corresponding to the arithmetic mean pressure, the flow becomes
m3
s
Qm =
(12.27)
However, equation (12.26) can be written as

M =
A (P1 + P2 ) (P1 P2 )
RT
2
X
k
Pm
RT
kg
s
(P1
P2 )
X
kg
s
(P1
P2 )
X
kg
s
(12.28)
(12.29)
12 - 16
Chapter 12 Methane

Qm
(P1
P2 )
X
kg
s
(12.30)
Comparing this with equation (12.22) shows that the volume flow of a gas at the position of mean
pressure is given by the same expression as for an incompressible fluid.
At any other density, , the volume flowrate, Q, is given by
Q = Qm
m3
s
(12.31)
If the flow is isothermal (constant temperature) then this correction can be expressed in terms of
pressure
Q = Qm
= Qm
Pm
P
(P1
+ P2 )
2P
m3
s
(12.32)
m3
s
(12.33)
12.3.1.3. Radial flow of gas

Figure 12.6 illustrates a borehole of radius rb intersecting a gas bearing horizon of thickness h.
The gas pressure in the borehole at the seam is Pb while, at some greater radius rs into the seam,
the gas pressure is Ps.
Ps
rs
P
Pb
rb
dr
Figure 12.6 Radial flow into a borehole.

12 - 17
Chapter 12 Methane
Consider the elemental cylinder of radius r and thickness dr, at which the gas pressure is P.
M =
RT
dP
dr
kg
s
m2
where A = 2 r h
giving
M =
k 2 h
r
P dP
RT
dr
or
M
dr
r
kg
s
k 2 h
P dP
RT
kg
s
Integrating between rs and rb gives

2
2
r
k 2 h Ps Pb
M ln s =
RT
2
rb
or
But as
(Ps + Pb )
2 RT
Qm
kg
s
k 2 h (Ps + Pb ) (Ps Pb )
2
RT
r
ln s
rb
(12.34)
kg
s
(12.35)
m3
s
(12.36)
Pm
= m
RT
=
2 h
(Ps Pb )
ln
rs
r b
Equations (12.31 and 12.33) can again be used to give the flowrate at any other density or
pressure.
12.3.1.4. Transient radial flow
The previous Section assumed a steady-state distribution in gas pressure throughout the
medium. This is not, of course, the real situation in practice. As a given source bed is drained, the
gas pressure will decline with time. An analysis analogous to that given for heat flow (Section
15.2.5.) leads to the time transient equation for a non-adsorbing medium:
k P 2 P
1 P
P
2 +
=
r
r r
t
where
and
t
=
=
Pa
s
(12.37)
rock porosity (dimensionless)

time (s)
For methane in coal, gas will be generated by desorption as the gas pressure falls (Figure 12.1)
and, hence, retard the rate of pressure decline very significantly. The gas desorption equations of
Section 12.2.2.2. must be coupled with equation (12.37) to track the combined effects of drainage
and desorption.
12 - 18
Chapter 12 Methane
12.3.2. The permeability of coal

In the previous section, it was assumed that the permeability, k, of the rock remained constant.
Unfortunately, in some cases including coal, this no longer holds. The anisotropy of the material
causes the natural permeability to vary with direction. Furthermore, the permeability changes with
respect to mechanical stress, gas pressure and the presence of liquids. We shall examine each
of these three effects in turn. First, however, let us clarify the concept of permeability and its
dimensions in the SI system of units.
Permeability, k, may be construed as the conductance of a given porous medium to a fluid of
known viscosity, . That conductance must depend only upon the geometry of the internal
flowpaths. In a rational system of units, the permeability must, therefore, be expressed in terms of
the length dimension. This can be illustrated by re-expressing Darcy's Law, equation (12.5) as
k = u
x
P
m Ns m 2
m
= m2
s m2
s
The units of permeability in the SI system are, therefore, m2 . However, the older units of Darcys
(or millidarcies, md) remain in common use. The Darcy was defined as "the permeability of a
medium that passes a single-phase fluid of dynamic viscosity 1 centipoise (0.01 Ns/m2 ) in
laminar flow at a rate of 1 cm3/s through each cm2 of cross-sectional area and under a pressure
gradient of 1 atmosphere (101 324 Pa) per cm". Such definitions make us grateful for the
simplicity of the SI system. The conversion between the unit systems is given as
1 md = 0.98693 x 10-15
m2
(12.38).
12.3.2.1. Effect of mechanical stress

Many researchers have reported tests on the response of coal permeability to applied loading
(e.g. Somerton, 1974; Gawuga, 1979; Harpalani, 1984). Such tests consist of delicate machining
of coal into sample cylinders of 30 to 50 mm in diameter and a length/diameter ratio of about 2
(Obert, 1967), then placing a sample into a triaxial permeameter. This device allows a liquid or
gas to be passed while the sample is subjected to axial loading (by a stiff compression machine)
and radial stressing (by oil pressure exerted on a synthetic rubber sleeve around the sample).
Such tests on coal have revealed the following phenomena during non-destructive loading.
Permeability reduces during loading and recovers during unloading. However, there may be
a significant hysteresis effect with lower permeabilities during unloading, particularly when
the axial and radial stresses are unequal.
Samples left under a constant applied stress exhibit a creep effect. The permeability
reduces with time down to a limiting value.
Following a loading test, a sample may not recover its initial permeability. Furthermore,
repeated loading tests will give progressively reduced permeabilities.
These observations suggest that coals exhibit a combination of elasticity and non-recoverable
strain. The latter is thought to be caused by permanent damage to the weaker bridges between
pores.
12 - 19
Chapter 12 Methane
shale
stressing
destressing
coal
test 2
test 1
Figure 12.7 The variation of permeability with hydrostatic

stress for two load/unload tests of a bituminous coal, After
the completion of test 1 the sample was left lightly loaded for
18 hours before commencing test 2.
k = A exp (B )
m2
Figure 12.7 shows the results

of hydrostatic loading (axial
and radial stresses remaining
equal) on the permeability of a
bituminous coal. Hydrostatic
stressing appears to minimize
the hysteresis effect. Two
load/unload cycles are shown.
After completion of the first
test, the sample was left for 18
hours in the triaxial cell at a
holding stress of slightly more
than 2 MPa. The creep
reduction in permeability
before commencement of the
second test shows clearly.
It appears from such tests that
the indicated permeability of a
coal sample measured during
a laboratory test depends
upon the stress history of the
sample. That is, the cycles of
loading and unloading caused
by mining close to the original
location of the sample, and the
method of recovering that
sample. However, for any
given laboratory test, the coal
permeability falls
logarithmically with respect to
applied hydrostatic stress, i. e.
the relationship is of the form
(12.39)
where = effective stress (MPa) = applied hydrostatic stress - pore (mean gas) pressure
A = constant (m2 )
B = constant (1/MPa).
The value of A is the theoretical permeability at zero stress and depends not only upon the coal
structure but also the stress history of the sample. Values of B from -1.5 to -2.5 (1/MPa) have
been reported for bituminous coal.
For comparison, Figure 12.7 includes a similar test on shale. The reduction in permeability with
applied stress is much less marked than that for coal.
12 - 20
Chapter 12 Methane
12.3.2.2. Effect of gas pressure

For most rocks, the permeability increases as the pore pressure exerted by a saturating gas
decreases. The effect was investigated by Klinkenberg (1941) who proposed the relationship
a
k = k liq 1 +
P
where kliq
and
m2
(12.40)
= permeability of the medium to a single phase liquid, or permeability at a very high

gas pressure (m2 )
= constant depending upon the gas (Pa)
The values of kliq and a may be determined by plotting experimentally determined values of k
against 1/P.
It is thought that the Klinkenberg effect is caused by surface diffusion, i. e. slippage of gas
molecules along the internal surfaces of the medium (Section 12.2.2.1.). However, in the case of
coal, a straight line relationship is not found when measured values of k are plotted against 1/P.
Hence, there is no longer any advantage in plotting the variables in that way.
Researchers have reported both rising and falling permeabilities with respect to mean gas
pressure during laboratory tests on coal. Figure 12.8 illustrates a typical set of results and shows
the effects of gas pressure and applied hydrostatic stress. Curve fitting exercises show that such
curves take the form
C2
+
exp (C 3 P C 4 )
P
Klinkenberg effect + dilation of flowpaths
k = C1 +
m2
(12.41)
where C1, C2, C3 and C4 are constants for the curve.

It will be observed that (C1 + C2/P) has the form of the Klinkenberg effect and is dominant at low
gas pressures. As the
gas pressure rises
through this phase,
surface diffusion
reduces and, adsorbed
molecules begin to
obstruct the flowpaths.
Hence, the permeability
falls. However, with
further increases in gas
pressure, the
exponential term in
equation (12.41)
becomes dominant
causing the
permeability to rise
again. One hypothesis
is that a sufficiently
high pore pressure
results in compression
of the coal substance
Figure 12.8 In addition to the effects of mechanical stress, the
and unconnected pores
permeability of coal is also a function of the internal gas pressure.
while the flowpaths
become dilated.
12 - 21
Chapter 12 Methane
12.3.2.3. Two phase flow

The pores and fracture networks of strata are often occupied by a mixture of fluids. In petroleum
reservoirs, three phase flow may occur with oil, gas and water as the occupying fluids. In most
other cases, including coal, two phase flow takes place as a mixture of gases and water.
The presence of water greatly inhibits the flow of gas and vice versa and, hence, reduces the
permeability of the rock to both phases. The effect is described quantitatively in terms of the
relative permeabilities kr,g and kr,w for gas and water respectively:
k r ,g =
and
k r ,w =
kg
k sg
kw
k sw
(dimensionless)
(12.42)
(dimensionless)
(12.43)
where kg and kw are the effective permeabilities of the rock to gas and water respectively, m2
(dependent on degree of saturation)
ksg = permeability of the rock to gas when saturated by gas, m2 (no water present)
and
ksw = permeability of the rock to water when saturated by water, m2 (no gas present)
In the petroleum industry, it is often assumed that ksg = ksw for the sandstones and limestones that
are typical of petroleum reservoir rocks. However, the effects of adsorption and gas pressure
indicate that this may not hold for coal.
Figure 12.9 illustrates a
typical behaviour of
relative permeabilities
with respect to
saturations of water and
gas. The actual loci of
the relative permeability
curves depend upon
kr,g
kr,w
kr,g
whether the coal
kr,w
substance is wetted
preferentially by the
water or the gas. This,
in turn, varies with the
proportion of coal
constituents, vitrain and
clarain tending to prefer
the gas while durain and
fusain are more easily
gas becomes
wetted by water.
mobile
The curves on Figure
12.9 suggest a net
hydrophobic coal, i.e.
the gas is the preferred
wetting phase. The
water will, therefore,
tend to reside in the
Figure 12.9 The relative permeabilities to gas, kr,g, and water kr,w
larger openings within
depend upon the degree of rock saturation by the two fluids.
the matrix and inhibit
12 - 22
Chapter 12 Methane
migration of the gas which exists in the smaller interstices. Hence, as indicated on Figure 12.9,
the gas will not become mobile until the water saturation has fallen significantly below 100
percent. This explains why considerable volumes of water may be produced from a borehole
before gas flows appear.
12.4. EMISSION PATTERNS INTO MINE WORKINGS
The rates at which methane are emitted into mine workings vary from near steady-state, through
cycles that mimic rates of mineral production, to the dangerous phenomena of gas outbursts or
"sudden large emissions". In general, the rate of gas emission into the mine ventilation system
depends upon:
initial gas content of the coal

degree of prior degassing by methane drainage or mine workings
method of mining
thickness of the worked seam and proximity of other seams
coal production rate
panel width (of longwalls) and depth below surface
conveyor speeds
the natural permeability of the strata and, in particular, the dynamic variations in
permeability caused by mining
comminution of the coal.
12.4.1. Sources of methane in coal mines

Variations in methane emissions into mines are influenced strongly by the dominant sources of
the gas. In room and pillar workings, gas will be produced from faces, ribsides and the pillars of
the seam being worked. While exposed pillars may be degassed fairly quickly (dependent upon
the coal permeability), ribside gas may continue to be troublesome for considerable periods of
time. In such cases, it is preferable for ribsides that border on virgin coal to be ventilated by return
air (Figure 4.7a). Peaks of gas emission will, in general, occur at the faces of the rooms due to
the high rate of comminution caused by mechanized coal winning. This will be moderated by the
degree of earlier degassing, either by methane drainage or by gas migration towards the
workings. The latter is enhanced by high coal permeability and a low rate of advance.
In addition to ribsides, the major sources of methane in longwall mines are the working faces and
roof and floor strata. Ribside gas tends to be a slow and near constant source. It is convenient to
classify the other sources of methane in a longwall mine into face (or coal front) gas and gob gas.
Peak emissions occur at coal-winning machines due to rapid fragmentation of the coal. It is this
emission that gives rise to frictional ignitions at the pick point (Section 12.1). The peak emission
moves along a longwall face with the machine. However, the freshly exposed coal front will also
emit methane - rapidly at first and decaying with time until the machine passes that point again.
Coal front gas is an immediate and direct load on the district ventilation system. Figure 12.10 is
an example of a smoothed record of methane concentration in air returning from a longwall face.
In this example, coal was produced on two out of three shifts for five days per week. The
correlation of methane make with face activity shows clearly, decaying down to a background
level over the weekend.
The gas flow into the gob area behind a longwall face originates from any roof or floor coal that
has not been mined from the worked seam, but more particularly from source beds within the roof
or floor strata. Any coal seams or carbonaceous bands within a range of some 200 m above to
100 m below the working horizon are liable to release methane that will migrate through the
relaxed strata into the gob area. If methane drainage is not practiced, then that methane will
subsequently be emitted into the mine ventilation system. Figure 12.11 illustrates the variations in
stress-induced permeability of roof and floor strata that create enhanced migration paths for the
gas.
12 - 23
Chapter 12 Methane
Figure 12.10 Recordings of airflow and methane flow in an airway returning from a
longwall face.
opening of
microfractures
and/or fissures
permeability
recompaction of strata
microfracturing
stress
stressed
zone
transition
coal face
stress relief
zone
roof source bed

recompaction
zone
working horizon
floor source bed
Figure 12.11 The migration of gas from roof and flow source beds toward the working
horizon occurs because of the large increase in permeability of adjacent
strata in the stress relief zone immediately behind a longwall face.
12 - 24
Chapter 12 Methane
As fragmented coal is transported out of the mine, it will continue to emit methane. The gas make
depends upon the degree of fragmentation, gas content of the cut coal on leaving the face, the
tonnage and the time spent during transportation. The desorption equations given in Section
12.2.2.2. allow estimates of the transport gas to be made. Precautions should continue to be
taken against accumulations of methane after the coal has left the mine. Gas explosions have
occurred in surface hoppers and in the holds of coal transport ships.
12.4.2. Methane layering
Methane emitted from the strata into a mine opening will often be at concentrations in excess of
90 percent. While being diluted down to safe general body concentrations, the methane will,
inevitably, pass through the 5 to 15 percent range during which time it is explosive. It is, therefore,
important that the time and space in which the explosive mixture exists are kept as small as
possible. This can be achieved by good mixing of the methane and air at the points of emission.
Unfortunately, the buoyancy of methane with respect to air (specific gravity 0.554) produces a
tendency for concentrated methane to collect in roof cavities and to layer along the roofs of
airways or working faces.
In level and ascentionally ventilated airways with inadequate airflow, the layer will stream along
the roof in the direction of airflow, increasing in thickness and decreasing in concentration as it
proceeds (Figure 12.12(a)),. Multiple feeders of gas will, of course, tend to maintain the
concentration at a high level close to the roof.
There are two main hazards associated with methane layers. First, they extend greatly the zones
within which ignitions of the gas can occur. Secondly, when such an ignition has taken place, a
methane layer acts very effectively as a fuse along which the flame can propagate -perhaps
leading to much larger accumulations in roof cavities or gob areas.
methane feeder
(a)
methane feeder
(b)
Figure 12.12 Methane layering in (a) a level airway

(b) a descentionally ventilated airway.
12 - 25
Chapter 12 Methane
Figure 12.12(b) indicates that in a descentionally ventilated airway, the buoyant methane layer
may stream uphill close to the roof and against the direction of the airflow. However, at the fringe
between gas and air, viscous drag and eddy action will cause the gas/air mixture to turn in the
same direction as the airflow. The result is that explosive mixtures may be drawn down into the
airway upstream from points of emission.
Although the layering phenomenum of methane in mines was all too obvious during the Industrial
Revolution (Section 1.2), systematic research on the topic seems not to have been well organized
until the 1930's (Coward, 1937). A combination of analytical and experimental work at the Safety
in Mines Research Establishment, England in the early 'sixties led to a quantification of the
important parameters (Bakke and Leach, 1962). These were
velocity of the ventilating airstream, u (m/s)

rate of gas emission, Qg (m3 /s)
width of airway W (m)
inclination of airway
relative densities of the air and gas
roughness of the roof above the layer.
Although a rough lining will promote better mixing than a smooth one (except for free-streaming
layers), the effect of roughness is fairly weak (Raine, 1960). Bakke and Leach found that the
characteristic behaviour of a gas layer was proportional to the dimensionless group
u
Q g
g
(12.44)
1
3
where / is the difference in relative densities of the two gases. (1 - 0.554 =0.446 for air and
methane)
Using a value of g = 9.81 m/s2 gives the dimensionless number for methane layers in air to be
L =
(9.81 0.446 Qg / W ) 13
u W
or L =
1.64 Q g
1
3
(12.45)
The dimensionless group, L, is known as the Layering Number and is of fundamental significance
in the behaviour of methane layers. Examination of equation (12.45) indicates that the air velocity
is the most sensitive parameter in governing the Layering Number and, hence, the length and
mixing characteristics of the layer. Although u is, theoretically, the air velocity immediately under
the layer, the mean value in the upper third of the airway may be used. For non-rectangular
airways, W may be taken as some three quarters of the roadway width. The power of 1/3 in
equation (12.45) reduces the effects of errors in estimated values of W and Qg.
Experimental data from Bakke and Leach have been employed to produce Figure 12.13 for level
airways. For any given gas make and roadway width, the horizontal axis may be scaled in terms
of air velocity. It can be seen from this graph that at low Layering Numbers (and, hence, low
velocity) the layer is affected very little by small additional increases in velocity. Indeed, in an
ascentional airway, the layer will actually lengthen as the airspeed increases from zero to the
free-streaming velocity of the methane. The mixing process is primarily due to turbulent eddies.
12 - 26
Chapter 12 Methane
The efficiency of mixing increases as the relative velocity between the methane and air rises.
However, as the air velocity continues to increase giving Layering Numbers of over 1.5 the layer
shortens rapidly. On the basis of such results, it can be recommended that Layering Numbers
should be not less than 5 in level airways.
Similar experiments in inclined airways have led to the recommended minimum layering numbers
shown in Table 12.1 in order to inhibit the formation of methane layers.
.
Angle to horizontal (deg.)
Ascentional
Descentional
10
15
20
25
30
35
40
45
5.7
3.3
6.2
3.7
6.6
4.1
6.9
4.3
7.2
4.5
7.4
4.6
7.6
4.7
7.7
4.8
7.8
4.9
Table 12.1. Recommended minimum Layering Numbers at various roadway inclinations.
Figure 12.13 Variation of layer length with Layering Number for level airways.
Methane layers can be detected by taking methanometer readings or siting monitors at roof
level. The most probable locations are in bleeder airways or return roadways close to a longwall
face. On detecting a methane layer, an immediate temporary remedy is to erect a hurdle cloth,
that is, a brattice cloth attached to the sides and floor but leaving a gap at the top. The size of the
opening should be such that the increased air velocity near the roof disperses the layer. The
hurdle cloth may need to extend some three quarters of the height of the airway. However,
12 - 27
Chapter 12 Methane
anemometer readings should be taken to ensure that the overall volume flow of air through the
area is not reduced significantly. Compressed air venturis or other forms of air movers can be
used to disperse methane layers provided that they are earthed against electrostatic sparking.
The longer term solutions are to (a) increase the airflow and, hence, the air velocities through the
affected panel; (b) reduce the rate of methane emission, or a combination of the two. Ventilation
network analyses should be employed to investigate means of increasing the airflow (Chapter 7).
These may include adjusting the settings of regulators or fans, and controlled partial recirculation
(Section 4.5.). The methane emissions can most effectively be reduced by installing a system of
methane drainage (Section 12.5.).
While methane layering was once a common occurrence in gassy coal mines, well designed
systems of ventilation and gas drainage are capable of eliminating this hazard from modern
mines. However, if stoppings or other ventilation controls are disrupted by any incidents in gassy
mines the reduction in airflows can result in the formation of methane layers and the possibility of
gas explosions.
12.4.3. Gas outbursts
The most dramatic mode of gas emission into mine workings is the release of an abnormally
large volume of gas from the strata in a short period of time. Such incidents have caused
considerable loss of life. In many cases, the rate of emission has been explosive in its violence,
fracturing the strata and ejecting large quantities of solid material into the workings. Gas outbursts
are quite different from rock bursts that are caused by high strata loadings. However, the
probability of a disruptive gas release from adjacent strata is enhanced in areas of abnormally
high stress such as a pillar edge in overlying or underlying workings.
There are two distinct types of gas outbursts, in-seam bursts and sudden large emissions from
roof and floor. Each of these will be discussed in turn.
12.4.3.1. In-seam outbursts
As the name implies, these are outbursts of gas and solids from the seam that is currently being
mined. They have occurred in many countries, particularly in coal and salt (or potash) mines. The
geologic conditions that lead to in-seam outbursts appear to be quite varied. The one common
feature is the existence of mechanically weakened pockets of mineral within the seam and which
also contain gas at high pressure. Methane, carbon dioxide and mixtures of the two have been
reported from in-seam outbursts in coal mines while nitrogen may be the major component in
potash mines (Robinson et al, 1981).
The genesis and mechanisms of in-seam gas outbursts have been a matter of some controversy.
A current hypothesis is that the structure of the material within outburst pockets has been altered
by tectonic stressing over geological time. Seam outbursts in coal mines have been known to
eject up to 5000 t of dust, much of it comprised of particles less than 10 m in diameter (Evans
and Brown, 1973). This can be accompanied by several hundred thousand cubic metres of gas
(Campoli, 1985). It is thought that pulverization of the coal has been caused by very high shear
forces. Laboratory tests have shown that a sample of normal anthracite can be reduced to the
fragile "outburst" anthracite by such means. Coal mines located in areas that have been subject
to thrust faulting are more prone to in-seam outbursts. A sample of "outburst" coal is, typically,
severely slickensided and friable to the extent that it may crumble to dust particles by squeezing it
in the hand. The seam is often heavily contorted in the vicinity of the outburst pocket - again,
evidence of the excessive tectonic stressing to which it has been subjected.
If the overlying caprock had maintained a low permeability during and since the pulverization of
the coal, then the entrained methane will remain within the zone. However, the small size of the
particles ensures that desorption of the gas can occur very quickly if the pressure is relieved
(equation (12.9)).
12 - 28
Chapter 12 Methane
When a heading or face approaches an outburst zone that contains highly compressed coal dust
and gas, stress increases on the narrowing barrier of normal coal that lies between the free face
and the hidden outburst pocket (Sheng and Otuonye, 1988). At some critical stage, the barrier will
begin to fracture. This causes audible noise that has been variously described as "cracking,
popping" or "like a two-stroke engine". When the force exerted by gas pressure in the outburst
pocket exceeds the resistance of the failing barrier, the coal front bursts outwards explosively.
The blast of expanding gas may initiate shock waves throughout the ventilation system of the
mine. A wave of decompression also passes back through the pulverized coal. The expansion of
the gas, reinforced by rapid desorption of large volumes of methane expels the mass of dust into
the airway. Although the outburst may last only a few seconds, the desorbing gas causes the dust
to behave as a fluidized bed. This can flow along the airway, almost filling it and engulfing
equipment and personnel for a distance that can exceed 100 m (Williams and Morris, 1972). The
cavities that remain in the seam following an outburst may, themselves, indicate slickensides.
Several cavities that are interconnected can contribute to a single outburst. The volume of dust
subsequently removed from the airway or face has often appeared too large for the cavity from
which it was apparently expelled.
The dangers associated with gas outbursts in mines are, first, the asphyxiation of miners by both
gas and dust. Compressed air 'lifelines' may be maintained on, or close to, faces that are prone to
in-seam outbursts. These are racks of flexible tubes connected to a compressed air pipe, and
with valves that open automatically when picked up. Such devices can save the lives of miners
who become engulfed in outburst dust. A second hazard is that the violence of the outburst may
damage equipment and cause sparking that can ignite the highly flammable gas/dust/air mixture.
Spontaneous ignitions of methane during outbursts in a salt mine have also been reported
(Schatzel and Dunsbier, 1989). Thirdly, the sudden expansion of a large volume of gas can cause
disruption of the ventilation system of a mine.
Precautionary measures against in-seam gas outbursts include forward drilling of exploratory
boreholes. However, pre-drainage of outburst pockets has met with limited success due to rapid
blockage of the boreholes. Similarly, testing samples of coal for strength or structure is uncertain
as normal coal may exist very close to an outburst pocket. Monitoring for unusual micro-seismic
activity is a preferred warning technique (Campoli, 1985). When it is suspected that a face or
heading is approaching an outburst pocket, then machine mining should be replaced by drill and
blast methods (volley firing), clearing the mine of personnel before each blast. This should be
continued until an outburst is induced or the dangerous area has been mined through.
12.4.3.2. Outbursts from roof and floor
These are most likely to occur in longwall mines. As discussed in Section 12.4.1. and illustrated
on Figure 12.11, methane will migrate from higher or lower source beds towards the working
horizon behind the working face. Strata in the stress relief zone normally exhibit a substantial
increase in permeability due to relaxation of the fracture network. The migration of gas then
proceeds through the overlying and underlying strata at a rate that follows cyclic face operations
and, under normal conditions, is quite controlled. However, a band of strong and low permeability
rock (caprock) existing in the strata sequence between the source beds and the working horizon
may inhibit the passage of gas. This can result in a reservoir of pressurized gas accumulating
beyond the caprock. Any sudden failure of this retention band will then produce a large and rapid
inundation of gas into the working horizon (Wolstenholme et al, 1969).
Figure 12.14 illustrates the development of a potential outburst from the floor. In Figure 12.14(a),
relaxation of the strata allows gas to be evolved from the source bed and migrate upwards.
A strong intervening bed of sufficiently high natural permeability or which contains induced
fractures allows the gas to pass through.
12 - 29
Chapter 12 Methane
working horizon
isobar
200 kPa
strong bed
400
(a)
600
800
1000
source bed
200 kPa
--
(b)
--- --- -- ---
--
--
--- --- -- -- caprock

gas reservoir
---
--
-----
-400 -600
800
-1000
Figure 12.14 Development of a gas outburst condition under a longwall panel.

(a) Safe condition: Isobars distributed; strong bed sufficiently permeable (or
fractured) allowing gas to migrate through in a controlled manner.
(b) Potential outburst condition: reduced permeability of the strong bed
resulting in an increased pressure gradient across this bed and high
pressure gas accumulations under it.
However, Figure 12.14(b) illustrates a condition in which a strong low permeability bed has failed
to fracture. Gas accumulates under this bed which is then subjected to an intensified gas
pressure gradient. This is the potential outburst situation. The rate at which gas accumulates in
the reservoir rock depends upon the relative rates of gas make from the source seam and gas
leakage through the caprock. Either an increase in the gas make from the source seam or a
decrease in flow through the caprock can produce the potential outburst condition.
Resistance to gas flow through the fracture network of the caprock may occur by increases in its
strength or thickness. Furthermore, a porous bed such as a fine-grained sandstone may suffer a
large decrease in relative permeability to gas if it becomes partially saturated by water (Figure
12.9).
There are considerable differences between the sudden large emissions that occur from roof and
floor strata. Strata above the worked seam are subjected to larger vertical movement due to
subsidence of the rock mass. This creates greater voidage in which gas can accumulate. Bed
12 - 30
Chapter 12 Methane
separation also promotes flowpaths parallel to the strata. Very large gas reservoirs may,
therefore, develop in the overlying rocks. However, fracturing of low permeability beds is more
probable and high pressures are less likely to develop in the gas reservoir rocks. For these
reasons, sudden emissions from the roof have usually been characterized by large flows that may
last from a few hours to several months (Morris, 1974). However, the initiation of the emission is
unlikely to be accompanied by violent dislocation of the immediate roof strata.
In contrast, gas outbursts from the floor are usually of shorter duration but more violent - even to
the extent of rupturing the floor strata upwards with ejection of solid material. The reduced
deformation in the floor sequence enables a caprock to resist induced enlargement of fractures.
This leads to high pressure accumulations of gas. However, the extent of the gas reservoir is
likely to be much less than one in roof strata. Morris (1974) reported methane emissions as large
as 140 000 m3 from floor outbursts, but over 8 x 106 m3 from sudden roof emissions.
Gas outbursts from roof or floor are often preceded by smaller increases in general body gas
concentrations, often intermittent in nature, during the hours before the major flow occurs. This is
probably caused by increasing strain within the caprocks and interconnection of bed separation
voidages before the main failure. Continuous gas monitors can detect such warnings.
Immediately before the burst, severe weighting on the roof supports and caving in the gob may
occur. Roof and floor outbursts are more likely to occur along planes of maximum shear stress in
the strata, i.e. in bleeders or airways bordering gob areas, or on the longwall face itself. Roof
outbursts may occur due to the initial failure of an overlying caprock within some two face lengths
of the initial starting-off line of the longwall. Similarly, passing under or over old pillar edges in
other seams can promote sudden failure of caprocks. However, previous workings may have
caused partially degassing of source beds. Most floor outbursts have occurred when no previous
mining has taken place at a lower level.
The most effective means of preventing roof and floor outbursts is regular drilling of methane
drainage holes wherever any potential caprock may exist. The holes should be angled over or
under the caved area and should penetrate beyond the caprock to the source bed(s). The
spacing between holes should be dictated by local conditions but may be as little as 10 m. It is
particularly important that the drilling pattern be maintained close up to the face. All holes should
be connected into a methane drainage system (Section 12.5.5.). In outburst prone areas, it is
essential that drilling takes place through a stuffing box in order that the flow can be diverted
immediately into the pipe range should a high pressure accumulation be penetrated.
It is often the case that routine floor boreholes produce very little gas. However, in areas liable to
floor outbursts, these holes should continue to be drilled as a precautionary measure. Tests
should be made of open-hole flow rates and rate of pressure build-up on the closed hole. When
both of these tests show higher than normal values, then it is probable that an outburst condition
is developing (Oldroyd et al, 1971). In the event of any indication that a gas reservoir is
accumulating in roof or floor strata, then additional methane drainage holes should be drilled
immediately to relieve that pressure and to capture the gas into the drainage pipe system.
12.4.4 Prediction of methane emissions into the ventilation system
As indicated earlier in Section 12.4 a considerable number of variables affect the rate at which
methane is emitted into the ventilation systems of coal mines or captured by methane drainage.
The non-linear interaction of the variables renders it difficult to apply analytical means to the
problem of predicting methane emissions. For this reason numerous numerical and empirical
models have been developed and utilized in several countries (e.g. Durucan et al (1992)).
Older empirical methods of predicting gas emissions into coal mines varied from curve fitting
procedures to pocket calculators (e.g. Creedy et al (1988); CEC (1988)). The continued
enhancement of computing power on desk and portable machines has promoted the utilization of
12 - 31
Chapter 12 Methane
sophisticated data processing software. This has resulted in improved reliability of predicting
methane emissions from data that are often difficult to correlate (e.g. Lunarzewski (1998). One
such procedure utilizes artificial neural networks (Karacan (2007)). This technique seeks patterns
and relationships between groups of input data that have appear to have obscure interrelationships from conventional mathematical or statistical methodologies.
A feature of artificial neural network software is that it can learn as more data is added, so
increasing its accuracy of prediction. Nevertheless, we are reminded that the reliability of all
empirical models is dependent on the range, quality and detail of of the measurements on which
the model is based. Hence, it is the responsibility of the user to ascertain that any given model is
applicable to the mine under consideration.
12.5. METHANE DRAINAGE
The organized extraction of methane from carboniferous strata may be practiced in order to (a)
produce a gaseous fuel, (b) reduce methane emissions into mine workings or (c) a combination of
the two. If the intent is to provide a fuel for sale or local consumption, then it is important that the
drained gas remain within prescribed ranges of purity and flowrates.
There is no single preferred technique of methane drainage. The major parameters that influence
the choice of method include
t he natural or induced permeability of the source seam(s) and associated strata
t he reason for draining the gas
t he method of mining (if any).
In this Section, we shall outline methods of methane drainage, the infrastructure of pipe ranges
and ancillary equipment, methods of predicting gas flows and conclude with a summary of the
procedure for planning a methane drainage system.
12.5.1. In-seam drainage
Drainage of methane by means of boreholes drilled into a coal seam is successful only if the coal
has a sufficiently high natural permeability or where a fracture network is induced in the seam by
artificial methods. Hence, for example, while in-seam drainage can be practiced in some North
American coalfields, it has met with very limited success in the low permeability coals of the
United Kingdom or Western Europe. A knowledge of coalbed permeability is necessary before inseam drainage can be contemplated (Section 12.3.2).
Advances in drilling technology have increased the performance potential of in-seam gas
drainage (Schwoebel, 1987). Using down-the-hole motors and steering mechanisms, boreholes
may be drilled to lengths exceeding 1600 m within the seam (Brunner (2006). Provided that the
coal permeability is sufficiently high, methane flows into mine workings can be reduced very
significantly by pre-draining the seam to be worked. Gas capture efficiencies up to 50 percent are
not uncommon, where
Gas capture efficiency =
Gas captured by methane drainage

100
Gas captured + gas emitted into ventilatio n
(12.46)
12 - 32
Chapter 12 Methane
In gassy and permeable seams, ribsides bordering on solid coal are prolific sources of gas.
Figure 12.15 illustrates flanking boreholes used to drain gas from the coal ahead of headings that
are advancing into a virgin area. On the other hand, previously driven headings or workings may
have degassed the area to a very considerable extent.
Figure 12.15 In-seam drainage boreholes to reduce methane flow into advancing
headings, applicable only where the coal is sufficiently permeable.
In-seam gas drainage can also be effective in permeable seams that are worked by the retreating
longwall system (Mills and Stevenson, 1989; Ely and Bethard, 1989). Figure 12.16 illustrates the
layout. Boreholes are drilled into the seam from a return airway and connect into the methane
drainage pipe system. The preferred spacing of the holes depends upon the permeability of the
seam and may vary from 10 to over 80 m. The distance from the end of each borehole and the
opposite airway should be about half the spacing between holes. The application of suction on
the boreholes is often unnecessary but may be required for coals of marginal permeability or to
increase the zone of influence of each borehole.
Figure 12.16. In-seam boreholes draining methane from a coal seam in a two-entry retreating longwall.
12 - 33
Chapter 12 Methane
The time allowed for drainage should be at least six months and, preferably, over one year.
Hence, the holes should be drilled during the development of what will become the tailgate of the
longwall.
Spalling of coal into the borehole can be a problem, especially in the more friable coals. This may
be reduced by employing smooth drill rods. Drill chippings can be removed by means of a
continuous water flush. Additionally, augers may be used to remove spalled coal from the
boreholes. The internal diameter of some sections of finished borehole may be considerably
larger than the drill bit. Perforated plastic liners can be inserted in order to maintain the holes
open, subject to the governing legislation. The first 5 to 10 m of each borehole are typically drilled
at 100 mm diameter. A standpipe is cemented into place and connected through a stuffing box
into the methane drainage pipeline (Section 12.5.5.1.). The remainder of the hole is drilled
through the standpipe at a diameter of some 75 mm.
The flowrate of gas from a gas drainage borehole will vary with time. Figure 12.17 illustrates a
typical life cycle for an in-seam borehole. A high initial flow occurs from the expansion and
desorption of gas in the immediate vicinity of the hole. This may diminish fairly rapidly but then
increase again as the zone of influence is dewatered, hence, increasing the relative permeability
of the coal to gas (Figure 12.9). This, in turn, is followed by a decay as the zone of influence is
depleted of gas.
In-seam boreholes drilled into outcrops or from surface mines may produce little methane. The
gas content of coal seams tends to increase with depth. It is probable that seams near the
surface have lost most of their mobile gas. However, multiple boreholes drilled with directional
control from a surface drill rig can be guided to follow a coal seam for distances that produce
acceptable flowrates.
Vertical holes drilled from the
surface to intersect coal
seams are likely to produce
very little gas because of the
short length of hole exposed
to any given seam. However,
hydraulic stimulation or
hydrofracturinq can be used
to enhance the flowrates. This
involves injecting water or
foam containing sand
particles into the seam. Other
sections of the borehole are
cased. The objective is to
dilate the fracture network of
the seam by hydraulic
pressure. The sand particles
are intended to maintain the
flowpaths open when injection
ceases. The success of
hydrofracturing depends upon
the natural fracture network
Figure 12.17 Typical life cycle of a gas drainage borehole in coal.
that exists within the seam
and the absence of clays that
swell when wetted. Friable coals are more likely to respond well to hydrofracturing. On the other
hand, in stronger coals, induced fractures may be concentrated along discrete bedding planes
12 - 34
Chapter 12 Methane
and give a poor recovery of gas. In the majority of cases, water must be pumped from surface
boreholes before gas flows can be attained.
Another means of in-seam gas drainage is to sink a small diameter shaft to intersect the coal bed
(U. S. Bureau of Mines, 1980). Long multiple boreholes are drilled radially outwards from the
shaft into the seam and connected into a methane drainage line that rises to the surface. This
method may be considered when a shaft is to be sunk at a later date at that location for mine
ventilation or service access. This enables degassing of an area for several years prior to mining.
12.5.2. Gob drainage by surface boreholes
The relaxation of strata above and below the caved zone in a longwall panel creates voidage
within which methane can accumulate at high concentration, particularly when other coal beds
exist within those strata. If this gas is not removed, then it will migrate towards the working
horizon and become a load on the ventilation system of the mine (Figure 12.11). Capture of this
"gob gas" may be accomplished either underground by cross-measures drainage (Section
12.5.3.) or by drilling boreholes from the surface.
to gas pump
Figure 12.18 Gob drainage of a longwall panel. Each hole becomes productive
after passage of the faceline.
Figure 12.18 depicts methane drainage from the gob of a longwall panel by surface boreholes.
This is a method that is favoured in the United States. Typically, three or four holes are drilled
from surface rigs at intervals of 500 to 600 m along the centreline of the panel and ahead of the
coal face. The holes may be 200 to 250 mm in diameter and drilled to within some 8 to 10 m of
the top of the coal seam (Mills and Stevenson, 1989). The holes should be cased from the
surface to a depth that is dictated by the local geology and, in particular, to extend below any
beds that are likely to act as bridging caprocks. A perforated liner can be employed in the rest of
the hole to inhibit closure from lateral shear.
12 - 35
Chapter 12 Methane
The initial gas made from the surface holes is likely to be small. However, as the face passes
under each borehole, the methane that accumulates in the caved area will be drawn towards that
borehole. Bed separation assists drainage from across the complete gob area. The first borehole
should be located far enough from the face start line (typically about 150 m) to ensure that it
connects into the caved zone. When a hole becomes active, the rate of gas production increases
sharply and may yield over 50 000 m3 /day of commercial quality methane for a period of several
months, depending upon the rate of mining. The more gradual decay is a result of reconsolidation
of the caved strata. Gas drainage pumps located on surface ensure that the gas flow remains in
the correct direction and may be employed to control both the rate of flow and gas purity. If the
applied suction is too great, then ventilating air will be drawn into the gob and may cause
excessive dilution of the drained methane.
In addition to longwall panels, gob drainage by surface boreholes can be utilized in pillar
extraction ares. In both cases, the technique can result in very significant reductions in emissions
of methane into mine workings.
12.5.3. Cross-measures methane drainage
Where the depth of coal workings precludes the drilling of methane drainage holes from surface,
the extraction of methane from relaxed strata can be accomplished by drilling from underground
airways. The principle of cross-measures drainage is illustrated on Figure 12.19. Capture
efficiencies in the range 20 to 70 percent have been reported for this technique, depending upon
the design and control of the system. Boreholes are drilled into the roof and, if necessary, also
the floor strata. The holes are normally drilled parallel to the plane of the coal face but inclined
over or under the waste. This is the dominant method of methane drainage in Europe and is
particularly applicable to advancing longwall panels. However, the technique can also be
employed for multiple-entry retreat systems.
Figure 12.19 Cross-measures methane drainage in an advancing longwall panel.
12 - 36
Chapter 12 Methane
The angle, length and spacing of boreholes must be decided on a site-specific basis. In general,
the holes should intersect the major gas emitting horizons. The spacing between holes should be
such that their zones of influence overlap slightly. Significant increases in general body methane
concentration along the airway is indicative of too great a distance between boreholes.
Conversely, if closing a borehole causes a rapid increase in the flowrate from neighbouring holes,
then the boreholes may be too close together. The hardness of the strata or other difficulties of
drilling will also influence the spacing. Typical spacings lie in the range 10 to 25 m between
boreholes.
Computer modelling procedures have been developed to assist in planning an optimum layout of
cross-measures boreholes (e. g. O'Shaughnessy, 1980). However, a practical study of the local
geology followed by a period of experimentation remains the most pragmatic means of
establishing a successful drilling pattern.
The boreholes are normally drilled from a return airway. However, in particularly gassy situations,
boreholes may be required on both sides of the panel. Although cross-measures methane
drainage is usually carried out from the current working horizon, a variation is to drill from airways
that exist in either overlying or underlying strata. Dewatering is facilitated by upward holes.
Routine underground drilling may require boreholes of smaller diameter than surface holes.
Cross-measures drainage holes are normally in the range 50 to 100 mm diameter. However, it is
necessary to employ standpipes in order to prevent excessive amounts of air being drawn into
the system. For any given degree of suction applied to a borehole, the purity of the extracted gas
will vary with the length of the standpipe. Hence, to ensure a high concentration, the standpipe
may need to be extended to within a few metres of the main gas-producing horizon. Standpipes
may be grouted into place with cement or resin, while rubber and sealing compounds have also
been used (Boxho et al, 1980).
The suction applied to gob drainage holes, whether drilled from the surface or underground, is a
sensitive means of controlling both the flowrate and the purity of the drained gas. The suction to
be applied on each individual borehole is a function of the gas pump duties, the geometry of the
network of pipes and settings of borehole valves. In general, increasing the suction applied to an
active borehole will increase the flow but decrease the purity. Borehole tests may be employed to
determine the optimum suction when the mass flow of methane attains a maximum value.
If the suction is too high, then the concentration of methane in the pipelines may fall below 15
percent when it becomes explosive. It is prudent to decrease the suction automatically when the
concentration falls to 30 percent. Much higher cut-off concentrations may be imposed by local
regulations or where the gas is utilized as a fuel. If the applied suction is too low, then excessive
emissions of methane can occur into the mine ventilation system. The optimum suction required
increases with respect to the rate of gas make from the source beds. A pragmatic approach is to
ensure that the interface between the air and methane remains in the strata above the level of the
airways. Although a borehole suction of some 10 kPa is fairly typical of cross-measures drainage
systems, large variations occur in practice. Indeed, as indicated in Section 12.4.3.2., gas may exit
the borehole at a positive pressure.
12.5.4. Drainage from worked-out areas
Although the rate of gas emission from any given source decays with time, an abandoned area
of a coal mine is likely to be surrounded by a large envelope of fragmented and relaxed strata.
The total make of gas may be considerable and continue for an extended period. In cases where
fragmentation has occurred across coal beds and between multiple seam workings, such
emissions of gas may go on for several years. In order to reduce pollution of a current ventilation
system, methane drainage pipes can be inserted through stoppings and gas drained from the old
workings. The pipes should extend beyond the stoppings to a point that is not affected by air
penetration during a period of rising barometer. The objective is to extract the gas at
12 - 37
Chapter 12 Methane
approximately the same rate as it is produced. However, additional suction may be applied when
the barometer is falling. Pressure balance chambers assist in reducing leakage across stoppings
(Section 21.5.5.).
Gas drainage can also be practiced when a gassy mine is abandoned. Again, pipes are left in the
seals on shafts or adits. This type of drainage may prevent seepages of methane escaping to the
surface and, possibly, creating a hazard in the basements of buildings. Drainage pipes that are
vented to the surface atmosphere should be elevated above ground level, well fenced against
unauthorised approach and provided with notices giving warning of inflammable gas. However,
as methane is a greenhouse gas discharging it to the surface atmosphere should be
discouraged.
12.5.5. Components of a methane drainage system
The methane that is produced from drainage boreholes must be transported safely to the required
point of delivery at surface level. The infrastructure that is necessary comprises five sets of
components:
pipe ranges
monitors
safety devices
controls
extractor pumps (if required)
12.5.5.1. Pipe ranges
The standpipe of each borehole connects into a branch pipe via a flexible hose, a valve and a
water trap. In the case of methane drainage in an underground mine, the branch pipes from each
district or section connect into main or trunk gas lines. Each main leads to a surface-connecting
pipe. Devices for monitoring, safety and control are located at strategic positions throughout the
system.
Both steel and high density polyethylene have been used as materials for methane drainage
pipes. Steel has the advantage of mechanical strength. However, it must be galvanized or
otherwise protected against corrosion on both the inner and outer surfaces. Where permitted by
legislation, heavy duty plastic piping may be preferred because of its lighter weight and ease of
installation (Thakur and Dahl, 1982). Polyethylene tubing up to some 75 mm in diameter may be
brought into the mine on reels, eliminating most of the joints that are required every five or six
metres with steel pipes. Pipes are sized according to the gas flows expected and vary from 75
mm to 600 mm in diameter. All pipes should be colour coded for easy recognition and in
correspondence with the relevant mine plan.
Pipes that are installed in underground airways should be suspended from the roof to reduce the
impact of falling objects. Where steel pipes are employed the joints should be capable of flexing
in order to accommodate airway convergence or other strata movements. Methane drainage
pipes should not be located in any airway where electrical sparking might occur such as a trolley
locomotive road. In any event, pipes should be earthed against a build-up of electrostatic
charges. Following the installation of a new length of pipe, it should be pressure tested either by
suction or by compressed air at a pressure of some 5000 kPa. All pipework should be inspected
regularly for corrosion and damage.
Pipes should be installed with a uniform gradient and water traps installed at unavoidable low
points including the base of a vertical or inclined surface connecting branch. Steel is preferred for
the main pipe. This may be located within an existing shaft. It is prudent to choose an upcast
shaft and this may, indeed, be enforced by law. Shaft pipes should be fitted with telescopic joints
to accommodate expansion and contraction. However, an alternative is to position a vertical pipe
in a dedicated surface connecting borehole. This method is attractive for the shallower mines and
12 - 38
Chapter 12 Methane
is the common technique within the United States. If necessary, surface pipes should be
protected against frost damage, either by thermal insulation or heated cladding.
12.5.5.2. Monitors
There are, essentially, three types of parameters that should be monitored in a gas drainage
system; pressure, flow and gas concentration. Pressure measurements are made at the mouths
of boreholes, across orifice plates to indicate flow and at strategic locations within the pipe
network. For boreholes that penetrate relaxed strata (Sections 12.5.2. and 12.5.3.) there may be
a sensitive relationship between borehole pressure, gas flowrate and concentration of the drained
gas, particularly where extractor pumps are used to maintain a negative gauge pressure at the
mouth of each borehole. Pressure differential indicators may vary from simple U tubes (Section
2.2.4.2.) containing water or mercury to sophisticated capsule or diaphragm gauges (Boxho et al,
1980). The latter may be capable of transmitting signals to remote recorders, alarms or control
centres. A simpler arrangement employs mechanical regulators to maintain set pressures at each
chosen location.
A variety of devices are used to indicate the rate of flow in gas pipes, including both swinging
vane or rotating vane anemometers and pitot tubes. However, the most popular method is to
monitor the pressure drop across an orifice plate (Figure A5.8). The gas flowrate, Q (m3/s) at pipe
pressure) and the pressure drop across the orifice, p, are related as shown by equation (2.50),
i.e.
p = Rt Q 2
where Rt
(Pa)
rational turbulent resistance of the orifice (m-4 ) and
density of the gas (kg/m3 ).
=
=
Furthermore, the rational turbulent resistance is given by equation (5.18) as

Rt =
where X
A
X
2 A2
=
=
shock loss factor for the orifice (dimensionless) and

cross-sectional area of the pipe (m2 ).
Combining these two equations gives

Q =
2p
m3 /s
(12.47)
where the pressure drop, p, is measured in Pascals.

An estimate for the value of X may be made from Figure A5.8 (Appendix to Chapter 5) for a clean
sharp-edged orifice plate, or from the corresponding equations:
X =
1
Cc
D 4
1
d
(12.48)
where D
=
diameter of pipe
d
=
diameter of orifice
and the orifice coefficient Cc is given by
C c = 0.48 (d / D )
4.25
+ 0 .6
(12.49)
12 - 39
Chapter 12 Methane
For greater precision, further multiplying factors may be incorporated into equation (12.47) to
account for variations in Reynolds Number, gas viscosity and, in particular, the roughness of the
edge of the orifice.
Spot measurements of gas concentrations in methane drainage networks may be made by
inserting probes through ports into the gas line and using high-range methanometers. Permanent
transducers are used to transmit signals of methane concentration at fixed locations to a remote
monitoring or control centre.
While the majority of methanometers described in Section 11.4.2 are intended for measurements
made in the subsurface airflow system, those employed in methane drainage pipes must be
capable of indicating gas concentrations up to 100 percent. The most popular instruments used
for this purpose are based on the variation of thermal conductivity with respect to gas
concentration (Section 11.4.2.3.). Interferometers and acoustic methanometers are also
employed while infrared gas analyzers may be used at permanent monitoring stations (Section
11.4.2.4.).
12.5.5.3. Controls and safety devices
The primary means of controlling an installed methane drainage system is by valves located in all
main and branch lines as well as at the mouth of each borehole. Manually operated isolating
valves are usually of the gate type in order to minimize pressure losses when fully open. These
are employed to facilitate extension of the system and pressure tests on specific lengths of
pipeline.
Diaphragm or other types of activated valves may be employed to react to control or alarm
signals. These might be used to vary the gas pressure at the mouth of each borehole in response
to changes in pipeline gas concentration. Abnormally high (or rapidly increasing) signals from a
general body methane monitor located in an airway that carries a gas pipeline may be used to cut
off gas flow into that branch. A simple but effective technique that is popular in the United States
is to attach a length of mechanically weak plastic tubing along the upper surface of each gas
pipe. The tubing is pressurized by compressed air or nitrogen which also holds open one or more
mechanically activated valves. Any fall of roof or other incident that damages the pipe is likely
also to fracture the monitoring tube. The resulting loss of air or nitrogen pressure causes
immediate closure of the corresponding valves.
The gas mixture that flows through methane drainage pipes is usually saturated with water
vapour. Liquid water is introduced into the system from the strata, by condensation and, perhaps,
following the connection of a borehole that has been drilled using a water flush. In order not to
impede the flow of gas, it is necessary to locate a water trap in the connection between each
borehole and the branch pipe, and also at all low points throughout the network. Water traps vary
from simple U-tube arrangements to automated devices. The latter consist essentially of a
compressed air or electric pump, activated by a float within a water reservoir. The size of the
reservoir and the settings of the high and low water levels are chosen according to the make of
water expected.
It is prudent to locate a lightening conductor close to any surface point of gas discharge.
Furthermore, flame traps should be inserted into the surface pipeline to prevent any ignition from
propagating into the subsurface system. Flame traps are devices that allow rapid absorption or
dissipation of heat without introducing an unacceptable pressure drop. Alternative designs include
closely spaced parallel plates, corrugated metal sheets or a series of wire gauze discs. A device
that is popular in Germany consists of a vessel containing glass spheres (Boxho et al.1980).
Flame traps are liable to be affected by dust deposits and should be designed for easy inspection
and maintenance. It is useful to locate a pressure differential transducer across the flame trap to
monitor for an unduly high resistance. If burning continues for any length of time at one end of a
flame trap, then the device can lose its effectiveness. Flame extinguishers may be used in
12 - 40
Chapter 12 Methane
conjunction with flame traps. These are pressurized containers that pulse an extinguishing
powder into the pipeline when activated by a temperature sensor.
The methane drainage network of a mine can be incorporated into an electronic environmental
surveillance and control system (Section 9.6.3.). This has a number of significant advantages.
The characteristic behaviour of the system can be investigated very readily from a control centre
by varying valve openings at chosen boreholes and observing the effect on gas flow and
concentration. Optimum settings can be determined and site-specific control policies established
for automatic control under the supervision of a computer. A fully monitored system also allows
the rapid detection and location of problem areas such as blockages, abnormal flows, fractured
pipes or faulty components.
12.5.5.4. Extractor Pumps
If in-seam gas drainage is practiced (Section 12.5.1.) and the coal bed has a high natural
permeability, then a satisfactory degree of degassification of the coal may be attained from the insitu gas pressure and without the use of extractor pumps. However, for gob or cross-measures
drainage (Sections 12.5.2. and 12.5.3.), it is usually necessary to apply controlled suction at the
boreholes in order to achieve a satisfactory balance between flowrate, gas concentration and
methane emissions into the mine.
There are two major types of gas extractor pumps used in methane drainage systems. Water seal
extractors take the form of a curved, forward bladed centrifugal impeller located either
eccentrically within a circular casing or centrally within an elliptical casing. In both cases, one or
two zones occur around the impeller where the casing approaches closely, but does not touch,
the impeller. In the remaining zones the casing widens away from the impeller. During rotation,
the water is thrown radially outwards by centrifugal action in the wide zones, drawing in a pocket
of gas from a central inlet port; then, as the impeller continues to turn, the water is forced inwards
by the converging casing, causing the pocket of trapped gas to be discharged into a conveniently
located central discharge port. Figure 12.20 illustrates a typical pressure volume characteristic
curve for a water seal extractor.
Figure 12.20 A typical pressure-volume characteristic curve for a wet-seal extractor pump.
12 - 41
Chapter 12 Methane
The major advantage of water seal extractors is that a gas seal is maintained without any contact
between moving and stationary components. Hence, there is little risk of igniting the gas or
propagating a flame. Water seal extractors have proved to be reliable and robust in practice, and
may be employed in banks of parallel machines to increase capacity. The temperature rise
across each unit is small, enabling the machines to be used in continuous or automatic operation.
However, periodic maintenance is required to remove any scaling or to guard against corrosive
action of the water.
The second family of gas pumps is the dry extractor. These may be reciprocating machines or
take the form of two lemniscate (dumb bell) shaped cams that rotate against each other within an
elliptical casing. This action encapsulates pockets of gas from the inlet port and transports them
across to the discharge port. Dry extractors are compact and, for any given speed, produce a
flowrate that is near independent of pressure differential. However, these devices are subject to
wear, create a substantial temperature rise in the gas and may produce considerable noise.
In the majority of cases, extractor pumps are located on the surface of a mine and, indeed, this
may be mandated by legislation. In some situations where the amount of drained gas is small, it
may be possible to site the extractor pumps underground and to release the gas into a well
ventilated return airway (if permitted by law or special exemption). A section of the airway around
the point of emission should be caged in and free from equipment to ensure that the gas has
been diluted to permissible limits before it enters any areas of activity.
12.5.6. Planning a methane drainage system
The detailed design and planning of a methane drainage system consists of two major phases.
First, a variety of data must be collated and analysed in order to estimate the gas flows that will
pass through the network of pipes. These flows provide the necessary input for the second
phase, the design of the actual network itself, including the dimensions of pipes and the duties
and locations of all ancillary components.
12.5.6.1. Data collection and estimation of gas capture
The data requirements and corresponding analyses for in-seam methane drainage (Section
12.5.1.) differ from those involved in gob drainage or cross-measures systems (Sections 12.5.2.
and 12.5.3.). However, a common first requirement is to obtain samples of coal from all source
beds and to determine the corresponding gas contents (Section 12.2.3.).
If particularly heavy emissions of methane occur from the worked horizon in headings, faces and
ribsides in virgin areas, then in-seam drainage may be practicable. In this case, it is vital that
permeability tests be carried out. Laboratory investigations are helpful in tracking the variations in
permeability with respect to stress, gas pressure and two phase flow (Section 12.3.2.). However,
in-situ tests provide the most reliable estimates of the actual permeabilities that will be
encountered. These involve one or more boreholes fitted with packers that encapsulate a length
within the seam of interest. Pressure build-up and draw-down tests indicate the transient increase
in pressure when the hole is shut in and the subsequent flow behaviour when it is re-opened.
Analysis of that time transient data allows the in-situ permeability to be determined (Section
12.3.1.4.). A simpler method that can be employed in beds of higher permeability is the slug test
(GRT, 1988). In this technique, a surface borehole is drilled to intersect the coal bed. A packer is
inserted above the seam through which passes a column into which is introduced a head of
water. A pressure transducer monitors the time transient variation in pressure and, hence, water
flowrate as the water enters the seam and the level in the column falls. Again, the transient radial
flow equation given in Section 12.3.1.4. is used as the basis of a computer algorithm to indicate
in-situ permeability.
12 - 42
Chapter 12 Methane
The coal samples should also be used to conduct adsorption and desorption tests (Sections
12.2.3.1. and 12.2.2.2.). The adsorption isotherm (Figure 12.1) can be employed to determine the
volume of gas that will be produced from a given area of coal reserves as the seam gas pressure
declines from its initial value to an estimated final state. The desorption tests provide data on the
variation of diffusion characteristics with respect to gas pressure and particle size.
Computer simulation programs have been developed that accept these various items of data and
model the flow of methane through coal beds towards boreholes or mine workings (e.g. Mavor
and Schwoebel, (1991); Patton et al (1994)). The models are based on interactions between the
relationships for gas release and migration given in Sections 12.2 and 12.3. Such programs can
most helpful in estimating the amount of gas that may be captured with differing sizes and
locations of in-seam boreholes.
Where a methane drainage system is to be designed for an existing mine, the emission patterns
into the mine workings should be established for all working areas, bleeder airways, development
headings and old workings. This is particularly important where the method is to be gob or crossmeasures drainage. Let us now turn to data collection and estimation of gas capture that are
pertinent to these systems.
Historical records of monitored data are very helpful but should be related to the corresponding
stratigraphy, mining layout and methods of working. The gas captured by gob and crossmeasures methane drainage is released primarily because of the stress relaxation and
fragmentation of source beds above and below the working horizon (Figures 12.11, 12.18 and
12.19). The natural permeability of those seams prior to mining may, therefore, have little
relevance to the subsequent rates of gas emission. However, the permeability-stress tests
described in Section 12.3.2.1. give an indication of the increased permeabilities induced by stress
relaxation. These, coupled with gas content data, desorption kinetics and the gas migration
relationships allow predictions to be made of future gas makes. Here again, computer models
have been developed to assist in such analyses (King and Ertekin, 1988). Several methods have
also been produced that allow hand calculations of predicted methane emissions from relaxed
roof and floor source beds (Dunmore, 1980).
In predicting gas production from future workings, it is important to take into account any changes
in the mine layout and the method and rate of coal extraction. In multi-seam areas,
degassification by earlier workings within some 200 m above or below the current horizon is likely
to result in a significant decrease in the gas available for drainage. Strata sequences should also
be examined for gas outburst potential (Section 12.4.3.).
12.5.6.2. Design of the drainage network
The first task in designing the hardware configuration of a methane drainage system is to draw a
line schematic of the layout. The network should be shown at its future maximum extent. The
locations of boreholes at that time and the corresponding predicted flowrates should be indicated
for all branches of the pipe network.
First estimates of pipe diameters can be made at this time (Section 12.5.5.1.). Gas velocities in
the range 8 to 12 m/s are suggested but, in any event, should not be greater than 15 m/s. Pipes
that are undersized for a future enhanced system will reduce the amount of gas that can be
drained or necessitate higher duties of the extractor pumps.
A number of methods are available for assessing the pressure distribution throughout the
network. Slide rules to assist in pipe network design are helpful (Boxho, 1980). A number of
computer programs have been produced for this purpose varying from single branch assessment
to full network analysis employing compressible flow relationships (Section 7.2.2.). These
programs can also be used for compressed air networks and allow the consequences of
changing pipe sizes, network layout or other design parameters to be assessed very rapidly.
12 - 43
Chapter 12 Methane
Significant changes in pressure and, hence, density occur as gas proceeds through the pipes of a
methane drainage system. It is, indeed, necessary to employ compressible flow relationships for
all but the simplest systems in order to determine frictional pressure drops. A methodology for
hand calculations can be derived from equations (2.50) and (2.51)
p = rt Q 2
Pa
where p
=
frictional pressure drop (Pa)
=
gas density (kg/m3 )
Q
=
flowrate (m3/s)
and
rt
=
rational turbulent resistance of the pipe (m-4 )
(We are using lower case rt for rational resistance here to avoid confusion with gas constants).
Now from equation (2.51)
f L per
f L d
rt =
=
3
2A
2 d 2 / 4
where f
L
and
d
=
=
=
64 f L
2
2 d
m 4
(12.50)
coefficient of friction (dimensionless)

length (m)
diameter of pipe.
In order to take compressibility into account, let us replace volume flow, Q, (m3/s) by mass flow
M = Q (kg/s) in equation (2.50) giving
p = rt
M2
Pa
(12.51)
kg/m3
(12.52)
But
P
RT
where P
R
and
T
giving
=
=
=
absolute pressure (Pa)

gas constant (J/kg K)
absolute temperature (K) of the gas
P p = r t M 2 RT
(Pa)2
(12.53)
If we consider a very short length of horizontal pipe, dL, having a resistance (drt ) and pressure
drop dp, then equation (12.53) becomes
P dp = M 2 RT (dr t )
(Pa)2
(12.54)
However, for a horizontal pipe, dp = -dP, and we may rewrite equation (12.54) as
P dP = M 2 RT (dr t )
(Pa)2
(12.55)
Integrating over the complete length of pipe (stations 1 and 2) and assuming isothermal
conditions (T = constant), gives
2
P1 P2
2
= M 2 RT rt
(Pa)2
(12.56)
12 - 44
Chapter 12 Methane
Substituting for rt from equation (12.50),

2
P1 P2
64 f L
2
M 2 RT
(Pa)2
(12.57)
Hence, if the pressure at one end of the pipe is known, then the pressure at the other end can be
determined.
Example
A methane/air mixture containing 45 percent methane enters an extractor pump at a pressure of
P2 = 50 kPa. The connecting pipe is 1000 m long, 300 mm in diameter and has a coefficient of
friction, f = 0.004. If the volume flowrate is 1 m3/s at standard atmospheric pressure and
temperature while the actual mean temperature of the gas is 25 C, determine the gas pressure,
P1, entering the pipe.
Solution
First, let us establish the values of the parameters required by equation (12.57)
P2 = 50 000 Pa; f = 0.004;
d = 0.3 m;
L = 1000 m;
T = 273 +25 = 298 K
To establish the gas constant and mass flow of the mixture, we utilize the fractional components
of 0.45 (methane) and 0.55 (air)
For methane:
for air:
and for the mixture:
Rm = 518.35 J/(kg K) (Table 11.1);

Ra = 287.04 J/(kg K)
R = (0.45 x 518.35) + (0.55 x 287.04) = 391.13 J/(kg K).
The flowrate is given as Q = 1 m3/s at standard conditions. As the density of air at those same
conditions is 1.2 kg/m3 and the relative density of methane is 0.554 (dimensionless, Table 11.1)
then the corresponding density of the mixture is:
= (0.45 x 0.554 + 0.55 x 1) 1.2 = 0.959 kg/m3.
giving mass flow M = Q = 1 x 0.959 = 0.959 kg/s
Substituting for the known values in equation (12.57) gives
P1 (50 000 )
2
64
0.004
(0.3)5
1000 (0.959 ) 391.13 298

2
giving P1 = 60 368 Pa or 60.37 kPa

This technique can be used, branch by branch, to determine the effective pressures (producing
flow) throughout the network. If an extractor pump is used, then the starting point will be the
absolute pressure at the pump inlet. On the other hand, the analysis may commence from a given
borehole pressure and proceed through the network in the direction of flow. This indicates the
suction pressure required to generate any desired flowrate. Iterative exercises can be carried out
to determine optimum pipe sizes. Allowance must be made for the pressure drops that occur at
all bends, valves, monitors and other fittings. Such assessments can normally be made from
tables or charts, provided by manufacturers.
12 - 45
Chapter 12 Methane
The buoyancy pressure, pbuoy, that assists gas flow in vertical pipes may be estimated simply as
pbuoy = (a - g) g Z
where a
g
g
Z
=
=
=
=
Pa
(12.58)
mean density of air in shafts (kg/m3)

mean density of gas mixture in the vertical pipe (kg/m3)
gravitational acceleration (m/s2) and
length of vertical pipe (m)
Where extractor pumps are employed, the buoyancy pressure is often ignored for design
purposes. However, if no pumps are used, it becomes a significant factor affecting gas flowrate.
The selection of the network monitoring system, transducers and safety devices should be
considered most carefully. Such choices must be made in accord with any national or state
legislation appertaining to gas pipelines.
Finally, a colour coded mine plan should be prepared showing the positions, dimensions and
other specifications of all pipes, control points, monitors, safety devices and extractor pumps.
This plan should be kept up to date as the system is extended.
12.5.7. Utilization of drained gas
Much of the methane produced from mines is passed into the surface atmosphere either via the
mine ventilation system or by venting the gas directly to the atmosphere from methane drainage
systems. Not only is this a waste of a valuable fuel resource but may have long-term global
environmental repercussions.
Where coalbed methane drainage is practiced, the production of gas of a consistent quantity and
quality may enable it to be employed in gas distribution facilities for sale to industrial or domestic
users. For other methods of methane drainage, variations in purity are likely to restrict utilization
of the gas to more local purposes. The simplest use of drained gas is for heating. Gas-fired
boilers or furnaces can provide hot water, steam or heated air for low cost district heating
schemes. Surface buildings on the mine and local communities can benefit from such systems.
Mineral processing plants may utilize the heat for drying purposes or to facilitate chemical
reactions. Variations in the purity of the gas fed to burners can be monitored continuously and
automatic adjustment of air: fuel ratios employed to maintain the required heat output.
Gas-fired turbines can be employed to produce electricity. Units are available that will develop
mine site electrical power from drained gas (Boxho et al, 1980, Owen and Esbeck, 1988). These
turbines can act as excellent co-generators producing not only several megawatts of electrical
power but also useful heat from the exhaust gases. Gas is normally delivered to the site of the
engines at relatively low pressure but is compressed to a pressure of 1000 to 1400 kPa before
injection into the combustion chambers. A supply of some 0.7 m3/s (standard pressure and
temperature) at 50 percent methane and 50 percent air can produce up to 4 MW of electrical
power and 9 MW of useable heat from a mine site gas turbine. Dual units may accept either
methane or fuel oil.
Other uses that have been found for drained methane include the firing of furnaces in local
metallurgical plants, kilns, coke ovens and for the synthesis of chemicals. In very gassy mines,
the methane concentration in the ventilating air returned to surface may be as high as 0.8
percent. Where furnaces are used on the mine site, then fuel savings may be achieved by
feeding those furnaces with return air.
12 - 46
Chapter 12 Methane
References
Airey, E. M. (1968). Gas emission from broken coal: An experimental and theoretical
investigation. Int. J. Rock Mechanics and Min. Eng. Sc., Vol. 5, pp 475-494.
Bakke, P. and Leach, S. J. (1962). Principles of formation and dispersion of methane roof layers
and some remedial measures. The Mining Engineer. Vol. 121, No. 22, pp 645-658.
Barker-Read, G. R. and Radchenko, S. A. (1989). Methane emission from coal and associated
strata samples. Int. Journ. of Mining and Geological Engineering, Vol. 7, No. 2, pp 101- 126.
Bertrard, C., Bruyet, B. and Gunther, J. (1970). Determination of desorbable gas concentration
of coal (direct method). Int. Jnl. of Rock Mechanics and Mining Sc., Vol. 7, pp 43-65.
Boxho, J. et al (1980). Firedamp drainage handbook. Published by the Coal Directorate of the
Commission of the European Communities, pp. 1-415.
Brunner, D.J., Schwoebel, J.J. and Brinton, J.S. (2006). Modern CMM strategies. 11th North
American Mine Ventilation Symp., Editors Mutmansky and Ramani, Penn State University.
Campoli, A. A. et al (1985). An overview of coal and gas outbursts. 2nd U. S. Mine Ventilation
Symp. Reno, NV, pp. 345-351.
CEC (1988). Investigation of firedamp and its emissions in coal seams. Final Report on ECSC
Contract 7220-AC/819, Report EUR 1147, Commission of the European Communities.
Close, J. C. and Erwin, T. M. (1989). Significance and determination of gas content data as
related to coalbed methane reservoir evaluation and production implications. Proc. of the 1989
Coalbed Methane Symp., Univ. of Alabama, Tuscaloosa, 19 pp.
Creedy, D.P. and Kershaw S. (1988). Firedamp prediction a pocket calculator solution. The
Mining Engineer. Inst of Mining Engineers, Vol. 147, No. 317, Feb., pp.377-379.
Coward, H. L. (1937). Movement of firedamp in air. Trans. Inst. of Mining Engrs., Vol. 94, p. 446.
Darcy, Henri (1856). Les fontaines publique de la ville de Dijon, Dalmant. Paris.
Diamond, W. P. and Levine, J. R. (1981). Direct method determination of the gas content of
coal: Procedures and results. U. S. Bureau of Mines Report No. 8515, pp. 1-36.
Dunmore, R. (1980). Towards a method of prediction of firedamp emission for British coal mines.
2nd Int. Mine Ventilation Congress. Reno, Nevada, pp 351-364.
Durucan, S., Creedy, D.P. and Edwards, J.S. (1992). Geological and geotechnical factors
controlling the occurrence and flow of methane in coal seams. Symp. on coalbed methane
research and devevelopment in Australia, Townsville, Australia.
Ely, K. W. and Bethard, R. C. (1989). Controlling coal mine methane safety hazards through
vertical and horizontal degasification operations. Proc. 4th U. S. Mine Ventilation Symp.,
Berkeley, CA, pp. 500-506.
Ettinger, I. L. et al (1986). Specification of gas dynamics in coal seams with the application of
methane diffusion in pieces, in "Urgent problems of mine aerogas dynamics." Izd-vo IPKON AN
SSSR, Academy of Sciences, Moscow, pp. 88-93.
12 - 47
Chapter 12 Methane
Ettinger, I. L. et al (1958). Systematic handbook for the determination of the methane content of
coal seams from the seam pressure of the gas and the methane capacity of the coal. Inst. of
Mining Academy, Moscow. NCB translation A1606/S.E.H.
Evans, H. and Brown, K. M. (1973). Coal structures in outbursts of coal and firedamp conditions.
The Mining Engineer. Vol. 173, No. 148, pp. 171-179.
Gan, H., Nandi, S. P. and Walker, P. L. (1972). Nature of the porosity in American coals. Fuel.
Vol. 51, pp. 272-277.
Gawuga, J. (1979). Flow of gas through stressed carboniferous strata. Ph.D. thesis. Univ. of
Nottingham, U. K.
GRI (1988). Technology Profile, Gas Research Institute, Chicago, Ill.
Harpalani, S. (1984). Gas flow through stressed coal. Ph. D. thesis. Univ. of California, Berkeley.
Harpalani, S. and McPherson, M. J. (1984). The effect of gas evacuation on coal permeability
test specimens. Int. Jnl. of Rock Mechanics and Mining Sc., Vol. 21, No. 3, pp. 161-164.
Hucka, V. J. and Lisner, U. W. (1983). In-situ determination of methane gas in Utah coal mines
(a case history). SME-AIME meeting, Atlanta, Georgia, USA, pp. 1-9.
James, T. E. and Purdy, J. L. (1962). Experiments with methane layers in a mine roadway. The
Mining Engineer. Vol. 121, No. 21, pp. 561-569.
Janas, H. (1980). Bestimmungen des desorbierbaren methaninhaltes an kohlenproben aus
exploratonsbohrungen. Unpublished report.
Jolly, D. C., Morris, L. H. and Hinsley, F. B. (1968). An investigation into the relationship
between the methane sorption capacity of coal and gas pressure. The Mining Engineer. Vol. 127,
No. 94, pp. 539-548.
Karacan, C. (2007). Modeling and prediction of ventilation methane emissions of U.S. longwall
mines using supervised artificial neural networks. International Journal of Coal Geology,
doi:10.1016/j.coal.2007.09.003.
King, G. R. and Ertekin, T. (1988). Comparative evaluation of vertical and horizontal drainage
wells for the degasification of coal seams. SPE Reservoir Engineering. May, pp 720-734.
King, G. R. and Ertekin, T. (1989). State of the art in modeling of unconventional gas recovery.
SPE (18947) Symposium. Denver, Colorado.
Klinkenberg, L. J. (1941). The permeability of porous media to liquids and gases. API Drill Prod.
Pract., pp. 200-213.
Lama, R. D. and Nguyen, V. U. (1987). A model for determination of methane flow parameters in
coal from desorption tests. APCOM '87. Proc. 20th Int. Symp. on the appl. of computers and
mathematics in the Minerals Ind. Johannesburg, pp. 275-282.
Langmuir, I. (1916). The constitution and fundamental properties of solids and liquids. J. Amer.
Chem. Soc., Vol. 38, pp. 2221-2295.
Langmuir, I. (1918). The adsorption of gases on plane surfaces of glass, mica and platinum. J.
Amer. Chem. Soc., Vol. 40, pp. 1361-1403.
12 - 48
Chapter 12 Methane
Lunarzewski, L. (1998). Gas emission prediction and recovery in underground coal mines.
International Journal of Coal Geology 35, pp. 117-145.
Mavor, M. J. and Schwoebel, J. J. (1991). Simulation based selection of underground coal mine
degasification methods. 5th U. S. Mine Ventilation Symposium, West Virginia, pp. 144-155.
Mills, R. A. and Stevenson, J. W. (1989). Improved mine safety and productivity through a
methane drainage system. Proc. 4th U. S. Mine Ventilation Symp., Berkeley, CA, pp. 477-483.
Morris, I. H. (1974). Substantial spontaneous firedamp emissions. The Mining Engineer. Vol.
133, No. 163, pp. 407-421.
Obert, L. and Duvall, W. I. (1967). Rock mechanics and design of structures in rock. Wiley and
Sons, New York, p. 333.
Oldroyd, G. C. et al (1971). Investigations into sudden abnormal emissions of firedamp from the
floor strata of the Silkstone Seam at Cortonwood Colliery. The Mining Engineer. Vol. 130, No.
129, pp. 577-591.
O'Shaughnessy, S. M. (1980). The computer simulation of methane flow through strata adjacent
to a working longwall coalface. Ph.D. Thesis, Univ. of Nottingham, U. K.
Owen, W. L. and Esbeck, D. W. (1988). Industrial gas turbine systems for gob gas to energy
projects. Solar Turbines Inc., San Diego, CA, pp. 1-14..
Patton S. B., Fan H., Novak T., Johnson P. W. and Sanford R L (1994), Simulator for
degasification, methane emission prediction and mine ventilation, Mining Engineering
(Littleton, Colorado), April. Vol 46, part 4, pp. 341 345.
Radchenko, S. A. (1981). The time of diffusion relaxation as the characteristics of coal
disturbance, in "The main questions of the complex development of the deposits of solid mineral
resources." Izd-vo IPKON AN SSSR, Academy of Sciences, Moscow, pp. 43-118.
Raine, E. J. (1960). Layering of firedamp in longwall workings. The Mining Engineer. Vol. 119,
No. 10, pp. 579-591.
Richmond, J. K. et al (1983). Historical summary of coal mine explosions in the United States,
1959-81. U. S. Bureau of Mines Information Circular No. 8909, pp. 1-51.
Robinson, G. et al (1981). Underground environmental planning at Boulby Mine, Cleveland
Potash, Ltd., England. Trans. Inst. of Min. Met., July. Also published in Jnl. Mine Ventilation Soc.
of S. Africa. Vol. 35, No. 9, Sept. 1982, pp. 73-88.
Schatzel, S. J. and Dunsbier, M. S. (1989). Roof outbursting at a Canadian bedded salt mine.
4th U. S. Mine Ventilation Symp., Berkeley, CA, pp. 491-499.
Schwoebel, J. J. (1987). Coalbed methane: Conversion of a liability to an asset. Mining
Engineering, April, pp. 270-274.
Sheng, J. and Otuonye, F. (1988). A model of gas flow during outbursts. 4th Int. Mine Ventilation
Congress, Brisbane, Australia, pp. 191-197.
Smith, D. M. and Williams, F. L. (1984a). Diffusional effects in the recovery of methane from
coalbeds. Jnl. of Soc. of Petroleum Engrs. Vol. 24, No. 5, Oct., pp. 529-535.
12 - 49
Chapter 12 Methane
Smith, D. M. and Williams, F. L. (1984b). Diffusion models for gas production from coal application to methane content determination. Fuel, Vol. 63, pp. 251-255.
Somerton, W. et al (1974). Effect of stress on permeability of coal, Berkeley, CA. U. S. Bureau of
Mines Contract H0122027.
Thakur, P. C. and Dahl, H. D. (1982). Mine Ventilation and Air Conditioning (ed. H. L. Hartman),
Chapter 4, "Methane Drainage", pp. 69-83.
U. S. Bureau of Mines (1980). Creating a safer environment in U. S. coal mines. The Bureau of
Mines Methane Control Program. U. S. Govt. Printing Office, Washington, D. C.
Williams R. and Morris, I. H. (1972). Emissions and outbursts in coal mining. Proc. Symp.
Environmental Eng. in Coal Mining. Inst. of Mining Engrs. Harrogate, U. K., pp. 101- 116.
Wolstenholme, E. F. et al (1969). Movement of firedamp within the floor strata of a coal seam
liable to outbursts. The Mining Engineer. Vol. 128, No. 105, pp. 525-539.
12 - 50
Radiation and radon gas
CHAPTER 13. RADIATION AND RADON GAS

13.1. INTRODUCTION .........................................................................................1
13.2. THE URANIUM SERIES AND RADIOACTIVE DECAY...............................2
13.2.1. Atomic structure; alpha, beta and gamma radiation ........................................................ 2
13.2.2. Radioactive decay and half-life........................................................................................ 4
13.3. RADON AND ITS DAUGHTERS .................................................................7

13.3.1. Emanation of radon.......................................................................................................... 8
13.3.2. Growth of radon daughters ............................................................................................ 11
13.3.3. Threshold limit values .................................................................................................... 12
13.4. PREDICTION OF LEVELS OF RADIATION ..............................................14

13.4.1. Emanation rate............................................................................................................... 14
13.4.2. Changes in working levels of radon daughters.............................................................. 14
13.5. METHODS OF MONITORING FOR RADIATION......................................18

13.5.1. Measurement of radon daughters.................................................................................. 19
13.5.2. Measurement of radon concentration ............................................................................ 20
13.5.3. Personal dosemeters ..................................................................................................... 20
13.6. CONTROL OF RADIATION IN SUBSURFACE OPENINGS .....................21

13.6.1. Ventilation systems for uranium mines .......................................................................... 21
13.6.2. Dilution and mixing processes ....................................................................................... 21
13.6.3. Radiation surveys .......................................................................................................... 23
13.6.4. Mining methods, mineral clearance and backfill............................................................ 24
13.6.5. Contamination from abandoned workings ..................................................................... 26
13.6.6. The influence of water.................................................................................................... 26
13.6.7. Air filters and rock surface liners.................................................................................... 27
13.6.8. Education and training ................................................................................................... 28
References ........................................................................................................28
13.1. INTRODUCTION
The element uranium is widely distributed within the crust and oceans of the earth. It has been
estimated that crustal rocks contain an average of some 4 grams of uranium per tonne. The
structure of the uranium atom is unstable. Emission of subatomic particles from the nucleus
causes uranium to change or decay into a new element, thorium. The process of radioactive
decay continues down through a series of elements until it reaches a stable form of lead. This
process has existed on earth from before the crust was formed. All forms of life on earth have
evolved and exist within a constant bombardment of natural radiation including that from the
uranium series of elements.
Although most of the uranium series of elements are solids, one known as radon, Rn, is a gas.
This facilitates escape, or emanation of radon from mineral crystals into the pore structure of the
rock from where it may migrate via a fracture network or interconnected interstices towards the
free atmosphere. However, the liberated radon itself decays into microscopic solid particles
known as radon daughters which may adhere to dust or other aerosol particulates, or remain
suspended as free ions in the air. If the rate of radon emanation into mine workings is high or the
13 -
space inadequately ventilated, then the radioactivity caused by continued decay of radon and its
daughters will reach levels that are hazardous to health. This is likely to occur within closed
environments in close proximity to rocks that contain traces of uranium ores. Radon is a potential
problem in any mine and tests for its presence should be carried out early in the development of a
new mineral deposit. The hazard can also exist in the basements of surface buildings.
As would be expected, the problem of radon is greatest in uranium mines and particular
precautions have to be taken in order to protect personnel from the development of lung cancers
caused by the inhalation and possible alveolar retention of radon daughters.
In this chapter, we shall outline the uranium decay series and quantify the mechanisms of radon
emanation and the growth of radon daughters within the atmospheres of subsurface openings.
Methods of measurement will be discussed while the final section is devoted to pragmatic
guidelines to be followed in designing a ventilation system for a mine with high radon emanations.
Readers who are interested only in the control of radiation in subsurface openings should
turn directly to Section 13.6.
13.2. THE URANIUM SERIES AND RADIOACTIVE DECAY

13.2.1. Atomic structure; alpha, beta and gamma radiation
The current concept of the structure of an atom is that of a nucleus consisting of protons of net
positive electric charge and neutrons of net zero charge, orbited by negatively charged electrons.
The electrical charge on the nucleus may be expressed in terms of the unit charge of an electron.
The atomic number of any given element indicates the level of net positive charge on the nucleus.
Figure 13.1 gives the atomic numbers of the elements in the uranium series. The atomic mass is
the nearest integer to the actual mass of the atom expressed relative to the mass of an individual
proton or neutron (atomic mass unit).
There are many classes of ionizing radiation. However, three are particularly relevant to decay
through the uranium series. An alpha ( ) particle has a positive charge of 2 units and a mass of
4 (helium nucleus). Hence, when an alpha particle is emitted from the nucleus, the atomic
number of that element reduces by 2 and the atomic mass reduces by 4. The next element in the
decay series is formed. A glance at Figure 13.1 will reveal the stages at which alpha particles are
emitted in the uranium decay chain. Alpha particles have relatively low energy levels. They travel
no more than a few centimeters in normal atmospheres and are halted by human skin.
However, if alpha particles are emitted within the lung, they can cause cellular alteration within
the alveolar walls, leading to possible lung cancer.
A beta particle has a negative charge of 1 and a negligible mass (electron). When this is
emitted, the atomic mass of the element remains effectively unchanged but the net positive
charge on the nucleus (atomic number) increases by one. This can be observed by movement
from left to right at atomic masses of 234, 214 and 210 on Figure 13.1. Beta particles not only
cause damage to the lung but can also penetrate human skin and may produce alteration of cell
tissue.
Both alpha and beta radiation involve the emission of subatomic particulates. Gamma ( )
radiation, however, is a very high frequency electromagnetic wave form (like x-rays)and is deeply
penetrating with respect to the human body.
13 -
83
82
85
86
87
88
89
90
238
91
Th
234
4.49 b y
Pa
21.4 day 1.17 min 0.248 my
230
226
222
218
214
92
Th
80 000 y
Ra
1 622 y
Rn
3.82 day
RaA
3.05 min
RaB RaC RaC'

26.8 min 19.7 min
Radon
daughters of
concern
164 s
210
206
RaD
22 y
RaE
Pb
stable
RaF
5.02 day 138.3 day
Figure 13.1 The uranium decay series and corresponding half lives. Radon gas, disintegrating
through the radon daughters RaA, RaB, RaC and RaC' give rise to alpha, beta and gamma radiation.
13 -
13.2.2. Radioactive decay and half-life

The rate at which atoms disintegrate, I, is known as the radioactivity and depends upon the
number of unaltered atoms and the probability of disintegration.
I
= N
where
and
dis/s
(13.1)
the decay constant (s-1 ) for that material and is a measure of the
probability of disintegration of any one atom.
For radon, = 2.1 x 10-6 disintegrations (dis) per second.
number of unaltered atoms remaining
The activity may also be expressed as

I
dN
dt
dis/s
(13.2)
If we commence with a fresh sample of a radioactive substance, then all of the atoms are capable
of disintegrating to the next lower element in the chain. However, as this process continues, there
are progressively fewer atoms remaining of the original substance. Hence, the rate of decay of
that substance will decline exponentially. This may be expressed in terms of the reducing number
of atoms that remain unaltered.
N = No exp (- t )
No
t
where
=
=
(13.3)
original number of atoms
time (s)
or, from equations (13.2) and (13.3),

I
d
{ No exp(- t ) }
dt
But the initial activity I 0

Hence,
I
= Io exp (- t)
N0
= No exp (- t)
from equation (13.1).

dis/s
(13.4)
showing that the radioactivity also decays exponentially with time. Figure 13.2 illustrates a decay
curve. By measuring the activity given by a radioactive substance while removing the decay
products, the corresponding curve can be plotted and the value of determined.
13 -
Activity, I or remaining number of atoms, N
2.500000
Io , No
2.000000
1.500000
I = Io e-t = N
Io/2 ,
1.000000
N /2
o
0.500000
0.000000
0
th400000
200000
600000
800000
1000000
Time t
Figure 13.2 The activity, I, and number of atoms, N, of the original substance both decay
exponentially with time. The half life, th, occurs when I and N are each half of their
original values.
The decay curve approaches zero activity exponentially; hence, the theoretical full lifespan of a
radioactive element is infinite. A more useful indicator of the aging process is the half-life of the
element. This is defined as the time taken for half of the original atoms to decay. As I = N, the
activity also reduces to half its original value at this same time. A simple relationship exists
between the half-life, th, and the decay constant, :
Initially,
I = I0
and at the half-life
= Io/2
Io exp(- th)
Hence,
exp(- t h ) =
1
2
or
- th
1
= ln
2
= -0.6931
13 -
giving
th
For radon,
0.6931
seconds
(13.5)
= 2.1 x 10-6 dis/s, giving
th (radon)
0.6931
2.1 x 10 6
0.33 x 106 seconds or 3.82 days.
The half-lives of the other elements in the uranium decay series are given in Figure 13.1 and vary
from U238 (4.49 billion years) to RaC' (164 microseconds).
The problems of radon in mines occur not only because it is a gas and can be emitted into the
ventilating airstream, but also from a consideration of half-lives. Radon gas is the decay product
of radium, Ra226 . This has a half-life of 1622 years, i.e. extremely long relative to the time taken
for a ventilating airstream to traverse through a mine. Hence, the source of radon is effectively
infinite. With a half-life of only 3.82 days, some of the radon will decay before it leaves the mine.
More importantly, the short half-lives of the radon daughters, RaA, RaB, RaC and RaC', indicate
that they will disintegrate readily, emitting alpha, beta and gamma radiation. These observations
have an important bearing on the design of ventilation systems for uranium mines.
13.2.3. Units of radioactivity
In the previous section we referred to the level of radioactivity in terms of atomic
disintegrations per second, dis/s. This unit is sometimes called the Becquerel (Bq) after the
French physicist who found in 1896 that uranium salts are radioactive.
1 Bq = 1 dis/s
(13.6)
A less rational but widely used unit of radioactivity is the Curie named after Marie (1867-1934)
and Pierre (1859-1906) Curie, who, in France, first separated and identified radium and a number
of other radioactive elements.
One Curie, Ci, approximates to the activity of 1 gram of radium or, more precisely, a source that
is disintegrating at a rate of 3.700 x 1010 atoms per second. Owing to the magnitude of this latter
value, sub-multiples of the unit are normally employed.
1 microcurie
1 picocurie
=
=
1 Ci
1 pCi
=
=
10-6 Ci = 37000 dis/s

10-12 Ci =
0.037 dis/s
(13.7)
Hence,
1 pCi = 0.037 Bq
(13.8)
or
1 Bq = 27 pCi
(13.9)
Analyses on radioactivity may be conducted in terms of either Becquerels or Curies.

Prior to the establishment of current knowledge on the effects of radioactivity on the human body,
it was thought that an average level of 100 pCi/litre from radon daughters was safe. This has
since been reduced to a third of that value. However, 100 pCi/litre became known as an
acceptable working level. The term was truncated to Working Level, WL, and is now used widely
with respect to radon daughters1.
1
The term Working Level is also defined as that concentration of short-lived radon daughters which
represents 1.3 x 105 MeV of potential particle energy while decaying to the stable Pb210.
13 -
Ionizing electromagnetic radiation produced by gamma emissions is measured in Rentgens

(after Wilhelm Rentgen of Germany who discovered x-rays in 1895). The Rentgen is defined
formally as the level of x or gamma radiation that produces 1 electrostatic unit of charge2 per
0.001293 g of air (1 cc at 101.324 kPa and 0C). In order to apply this in terms of the effect on
human bodies, the Rem (Rentgen equivalent man) is employed. This is the amount of ionizing
radiation that will cause the same biological effect as 1 Rentgen of x or gamma rays. Dose rates
are quoted in mRem/hour. Hence, a dose rate of 50 mRem/hour would produce a dose
equivalent of 25 mRem after half an hour.
The Rentgen and the Rem are the most commonly used units for ionizing radiation and
biological dosage. They are not, however, SI units. The rational SI equivalents are:
1 C/kg (Coulomb per kilogram) =
1 Sv (Sievert)
=
3876 Roentgens
1 J/kg = 100 Rems
As a Sievert is a very large unit, radiation doses are quoted in milli-Sieverts (mSv). One chest xray is equivalent to about 0.2 mSv.
13.3. RADON AND ITS DAUGHTERS
Radon gas emanates from the crystalline structure of minerals into the pores and fracture
networks of rocks. It migrates through the strata by a combination of diffusion and pressure
gradient towards mine openings. Radioactive decay produces the solid particulates of the radon
daughters as given in Figure 13.1, namely
half life
polonium,
lead,
bismuth,
polonium,
and lead
Po218 or radium A,
Pb214 or radium B,
Bi214 or radium C,
Po214 or radium C',
Pb210 or radium D,
RaA
RaB
RaC
RaC'
RaD
3.05 min
26.8 min
19.7 min
164 s
22 years
As the half-life of RaD is 22 years, we need concern ourselves only with the radon series down to
RaC'. Furthermore, RaC' has the extremely short half-life of 164 microseconds. Hence, the effect
of its decay is normally coupled with that of RaC.
The solid particulates of radon daughters that form during migration of the gas through strata are
likely to plate on to the mineral surfaces and be retained within the rock. However, the remaining
radon will continue to decay after it has been emitted into a mine opening. The radon daughters
will then adhere to aerosol particles or remain as free ions within the airstream.
In this Section, we shall examine the migration of radon through the rock and the growth of radon
daughters within a ventilating airflow.
1 electrostatic unit of charge is equivalent to 2.083 x 109 ion pairs.
13 -
13.3.1. Emanation of radon

When an alpha particle is projected from an atom of radium, the resulting atom of radon recoils
through a distance of some 3 x 10-8 m in minerals and 6 x 10-5 m in air (Thompkins, 1982).
Furthermore, the diffusion coefficient for radon within mineral crystals is very small. Hence,
although movements of radon atoms will occur within the crystals, the distances of individual
motion are small in comparison to most mineral grain sizes; the probability of any one radon atom
escaping into a pore is, therefore, also small. Nevertheless, a sufficient number do escape to give
rise to radon problems in uranium mines.
The migration of radon through the pore and fracture network of the strata may be analysed on
the basis of Fick's laws of diffusion modified for radon production and decay (Bates and Edwards,
1980). This leads to an approximate equation that describes the radon concentration, C, within
the pores with respect to distance into the rock
C

= C 1 exp x
pCi/m3 of space
(13.10)
where
C
x
and
radon concentration at infinite distance into rock (pCi/m3)

distance into rock from free surface (m)
radon decay constant (2.1 x 10-6 Bq)
rock porosity (fraction)
diffusion coefficient for rock (m2/s)
=
=
=
=
=
The units of radon concentration require a little explanation. The normal volumetric concentration
commonly used for other gases would give excessively low values for radon. It is more
convenient to express radon concentration in terms of the level of radioactivity (pCi or Bq) emitted
by each m3 of the radon:air (or radon:water) mixture.
Equation (13.10) is based on an assumed radon concentration of zero at the open rock surface.
The actual emanation at the rock surface can be measured directly or calculated as
J = C D
pCi
m2s
(Bates and Edwards, 1980)
(13.11)
The maximum value of radon concentration in the rock, C may be determined from
C
pCi
m3
(13.12)
where B is the rate of emanation from unit volume of rock (pCi/m3s) (sometimes known as
emanating power) and can be measured from samples of the rock. Values of both B and J for
differing rocks vary by several orders of magnitude.
To reiterate, J indicates the radon emanation from a solid rock surface while B refers to radon
emanation from fragmented rock.
J = B
pCi
m2s
(13.13)
13 -
A guide to coefficients of diffusion for radon in a range of materials is given in Table 13.2. Figure
13.3 may be used as a guideline to estimate a coefficient of diffusion for radon in rocks of known
porosity. However, it should be noted that the coefficient of diffusion varies widely with the type
and condition of pore fluid.
Medium
Coefficient of diffusion D m2/s
Rocks:
dense rock
porosity 6.2 percent
porosity 25 percent
0.05 x 10-6
0.2 x 10-6
0.27 x 10-6
0.5 x 10-6
3 x 10-6
Air
Water
Alluvial soil
Concrete
(10 to 12) x 10-6

0.0113 x 10-6
(3.6 to 4.5) x 10-6
(0.0017 to 0.003) x 10-6
Table 13.2. Coefficients of diffusion, D, for radon in various porous media (after Thompkins, 1985).
Figure 13.3 Guideline to the approximate coefficient of diffusion, D, for radon in rocks of
known porosity. Note that actual coefficients of diffusion depend upon the
type and state of pore fluids.
13 -
16
Concentration of radon
microCi/cu.m
(a)
14
12
10
8
6
4
2
0
0
10
Concentration of radon microCi/cu.m
Distance into rock m

16
14
(b)
12
10
8
6
4
2
0
0.01
0.1
10
Distance into rock m

Curve
1
2
3
4
Porosity
per cent
6.2
7.4
12.5
25
Coefficient of
Diffusion D m2/s
0.2 x 10-6
0.27 x 10-6
0.5 x 10-6
3.0 x 10-6
Figure 13.4 (a) Examples of variation of

radon concentration with distance into rock,
constructed from equations (13.10) and
(13.12) with B = 2 pCi/m3s and = 2.1 x 10-6
becquerels.
(b) The same curves plotted on a log scale
for distance shows better accuracy at
shorter distances.
13 -
10
Figure 13.4 illustrates the variation of radon concentration within the pores for the four rocks of
specified porosities and corresponding coefficients of diffusion given in Table 13.2. These curves
indicate that peak emanations of radon are likely to occur when rock is broken from the orebody
and reduced to fragments of about 10 cm in size. However, below that dimension there is
relatively little change in radon concentration within the pores, particularly for the higher porosity
rocks.
13.3.2. Growth of radon daughters
Radon is emitted into mine openings not only directly from the surrounding rock surfaces but also
from old workings and other zones of voidage. Hence, the total emanations will consist of a
mixture of the gas and its particulate daughters. In order to analyse the growth of the daughters,
consider an imaginary experiment. We commence with one litre of filtered air that contains a
radon concentration of 100 pCi/litre. The radon will immediately begin to suffer disintegration into
RaA. This has a half-life of only 3.05 minutes. Hence, the second radon daughter, RaB soon
appears. However, this has a half-life of 26.8 minutes and concentrations of the RaC (and RaC')
do not become significant until some 20 minutes from the start of the experiment. We now have a
situation in which the initial amount of radon gas is diminishing slowly (half-life of 3.82 days) but
each of the daughters is simultaneously being generated and decaying on a shorter time scale.
With its relatively small half-life, the concentration of RaA reaches a state of dynamic equilibrium
within a few minutes. Figure 13.5 illustrates the growth of the radon daughters from an initial
radon concentration of 100 pCi/litre, i.e. 1 Working Level.
Full (secular) equilibrium is reached in some 30 hours when the air is said to have fully aged.
However, 80 percent of the aging occurs within 90 minutes (5400 seconds). The concept of the
age of the air is most important in understanding the radioactive decay of radon. At secular
equilibrium, the number of atoms in each radon daughter is proportional to its half-life. The age of
the air can be estimated from the growth curve of Figure 13.5 for a given working level of radon
daughters and having corrected for the actual concentration of radon gas present. The latter may
be eliminated by using Figure 13.9 which gives the growth in total radon daughters for a range of
initial radon gas concentrations.
With a half-life of 3.82 days, the radioactivity caused by the decay of radon during the short time it
spends within the human lung is very limited. However, the shorter half-lives of the radon
daughters cause them to be much more prolific emitters within the respiratory system.
Furthermore, as they are particulates, they may adhere to mucous membranes and be retained
within the lung tissue. It is, therefore, the daughters of radon rather than the gas itself that are
dangerous to health.
As it is the decay of the radon daughters that causes the health hazard, it is clear from Figure
13.5 that airflows should be sufficiently vigorous to remove radon from mine workings as rapidly
as possible and before significant aging has occurred. It is also clear that near stagnant air in old
workings will be fully aged, i.e. contain the maximum concentrations of radon daughters and may
cause serious escalations in working levels if leakage occurs from those old workings into active
airways.
Although Figure 13.5 illustrates the growth of each radon daughter, it will be recalled that each
daughter is simultaneously suffering disintegration. Hence, a mirror image of Figure 13.5 would
indicate the corresponding decay curves. At full secular equilibrium, the growth and decay curves
for each radon daughter would be horizontal lines, equal in magnitude but opposite in sign. For
clarity, the decay curve for total radon daughters only is shown on Figure 13.5.
13 -
11
1.0
0.9
0.8
WL = 0.00159 t 0.74
0.7
0.6
0.5
WL = 0.00102 t 0.8
0.4
0.3
0.2
0.1
0
600 1200 1800 2400 3000 3600 4200 4800
Figure 13.5 The growth curve shows the development of radon daughters from an initial
radon concentration of 100 pCi/litre and with no replenishing radon. The decay curve
shows the fall in total working level due to radon and radon daughters and is a mirror
image of the growth curve.
13.3.3. Threshold limit values

Lung cancers occur throughout the general population but are exacerbated by the inhalation of
some types of dust particles and products of combustion. However, the higher incidence of lung
cancer in uranium mine workers is attributed to radon daughters. The current method of dosage
assessment is based on cumulative exposure. Any unit of time may be employed. However, the
most widespread measure of cumulative exposure is the Working Level Month, WLM, defined as
an exposure of 1 WL for a period of 1 month.
13 -
12
Assuming that an average of 170 hours are spent in the workplace each month, the cumulative
exposure of any individual may be calculated as
(WL hours of exposure)

170
WLM
Example
During the 40 hours a miner spends underground during a week, radiological monitoring indicates
the following levels and periods of exposure to radon daughters:
20 hours at 0.1 WL
15 hours at 0.2 WL
5 hours at 0.4 WL
Determine the cumulative exposure in WLM for that week.
Solution
Cumulative exposure =
(20 0.1) + (15 0.2) + (5 0.4)

170
= 0.041
WLM
The most commonly accepted TLV (threshold limit value) for mine workers is that no person
should be subjected to a cumulative exposure exceeding 4 WLM in one year. Legislation may
also require that there should be an upper limit of 2 WLM in any consecutive three months and an
instantaneous ceiling limit level of 1.0 WL.
The TLV of 4 WLM per year implies an average of 4/12 = 0.33 WLM per month, i.e. an average
radiation level of 0.33 WL. Legislation may require mandatory reporting and increased rates of
sampling at higher levels. However, the aim should be for lower actual levels in the active work
areas of a mine as leakage from old workings can cause rapid escalation of working levels in
return airways. [National or state legislation may also mandate notification to inspection agencies
at very low levels of radiation (Howes, 1990).]
There remains considerable doubt on the long term effects of exposure to low levels of
radioactivity. It should be expected that threshold limit values will be decreased yet further. In the
meantime, the International Commission on Radiological Protection has recommended that in
addition to remaining within specified threshold limit values, radiation levels should be maintained
"as low as reasonably achievable. This is often termed the ALARA principle. The practical
application of this rather vague phrase may involve mine management being required to notify the
inspectorate when a specified fraction of the exposure TLV is exceeded for any worker; and,
furthermore, to demonstrate that all reasonable steps have been taken to alleviate the problem.
The threshold limit values for gamma (ionizing) radiation are, typically, 5 Rems per year. Where
average gamma radiation measurements are in excess of 2 milli-Rentgens per hour, personal
gamma radiation dosemeters may be mandated for all personnel and records of cumulative
exposure maintained for each individual (e.g. Title 30 US Code of Federal Regulations,
paragraph 57.5047).
13 -
13
13.4. PREDICTION OF LEVELS OF RADIATION

13.4.1. Emanation rate
In order to predict the levels of radiation to be expected in any active area of a mine, the
emanation rates, J (pCi/m2s), from exposed surfaces and/or emanating powers, B (pCi/m3s), from
drill cores or other samples must first be measured. The relationship between the two is given by
equation (13.13). Surface emanation rates can be measured in-situ by attaching steel collection
receptacles tightly to a rock face (Archibald, et al, 1980) or by excavating chambers into the rock.
Multiple measurements made on a variety of ores and waste rock allow tables of surface
emanation rates, J, and emanating powers, B, to be assembled for a given mine or rock type.
In order to determine the total rate of emanation in a specified zone, the surface area of all
exposed rock faces, A (m2), and the volumes of broken ore, V (m3), must be assessed and
multiplied by the relevant values of J and B respectively.
Example
In the walls of a given stope, 350 m2 of ore and 300 m2 of waste rock surface are exposed. The
stope contains 320 t of fragmented ore. If the density of the ore is 2000 kg/m3, determine the rate
of emanation into the stope given that
and
Solution
Radon from:
J (ore)
=
J (waste) =
500 pCi/m2s;
100 pCi/m2s
Ore wall surface

=
Waste wall surface =
Total wall surface =
Volume of ore fragments =
350 x 500 = 175 000 pCi/s

300 x 100 = 30 000
205 000 pCi/s
320 000
2000
Radon from fragmented ore = 160 x 600

Total emanation of radon
B (ore) = 600 pCi/m3s
kg m 3
kg
160 m3
96 000 pCi/s
= 205 000 + 96000 =
301 000 pCi/s
13.4.2. Changes in working levels of radon daughters

A commonly accepted method of calculating the variation in natural radioactivity due to radon
daughters along an airway was developed by Schroeder and Evans at the Massachusetts
Institute of Technology in 1969. Consider the length of airway shown in Figure 13.6. The working
level of radon daughters at the exit from the airway, distance x (m) from the entrance, arises from
three factors
(i) decay of the radon gas already in the air at the entry point,
(ii) decay of the radon that emanates from the walls or fragmented rock within the airway itself,
and
(iii) continued aging of radon daughters that already exist in the airway at the entry point.
13 -
14
Each of these three behaves independently of the others. They may, therefore, be determined
separately and summed to give the working level of radon daughters at exit. Let us consider each
component in turn.
Q = uA m3/s
dx
Figure 13.6 Radon emanates from rock surfaces at a rate of J pCi/m2s.

(i) Decay of radon existing at entry
This can be determined from Figure 13.5. Curve fitting to the total working level (growth)
curve on this Figure gives
WL
where
and
102 x 10-5 t0.8 Working Levels for t < 2400 s (40 minutes)
WL
t
=
=
(13.14)
Working level due to decay of 100 pCi/litre of initial radon

travel time for air to traverse the airway (s)
(the coarser equation: WL = 159 x 10-5 t0.74 may be used for t < 4200 s or 70 minutes).
For other concentrations of radon, C (pCi/litre), equation (13.14) becomes
WL1 =
C
x 102 x 10-5 t0.8
100
(13.15)
This equation is the basis of Figure 13.9.

Example 1
An airflow of 24 m3/s with a radon concentration of 20 pCi/litre enters an 800 m long
rectangular airway of dimensions 4m by 3m.Calculate the working level of radon daughters at
exit caused by the initial radon.
Solution 1
Cross-sectional area, A = 4 x 3 = 12 m2
Mean velocity =
Residence time,
Airflow
24
=
= 2
A
12
tr =
m/s
distance
800
=
= 400 seconds
velocity
2
13 -
15
Then equation (13.15) gives

20
WL1 = 100 x 102 x 10-5 (400)0.8 = 0.025 Working Levels
This result may also be read directly from Figure 13.9.

The example will be continued for the remaining sources of radon daughters.
(ii) Decay of radon emanated from rock surfaces into the airway
To facilitate the analysis, it is assumed that radon emanates at a uniform rate of J
pCi/m2s from all solid surfaces around the opening. Consider the airway shown in Figure
13.6. The air enters at position 0 and leaves after traversing a distance of X (m). Over
the short length dx, the wall area is per dx (m2 ), where per = perimeter (m).
The radon emitted over this increment is, therefore,
J per dx
pCi/s.
If the airflow is Q = uA m3/s

where
u = mean velocity (m/s) and
A = cross sectional area (m2 )
then the radon emanation may be expressed in terms of cubic meters of air, i. e. an
increase in concentration:
Concentration from radon emitted in increment dx =
J per dx
Q
pCi
J per dx
uA
m3
(13.16)
The time taken for the air to travel from x to X is

t
( X x)
u
seconds
Equation (13.14) states that from a radon concentration of 1 Working Level, i.e.100 pCi/litre
(or 100 000 pCi/m3) the working level of radon daughters that will develop in this time is
102 10-5
( X x ) 0.8
u 0.8
WL
Hence, from a radon concentration of

J per dx
uA
pCi
3
m
or
J per dx
= Working Levels
100 000 uA
the working level of radon daughters will grow to

( X x )0.8
J per dx
-5
102 x 10
u 0 .8
100 000 uA
WL at exit from the airway.
13 -
16
To find the working level of radon daughters at exit due to radon emanations throughout the
complete length of airway, WL2, we must integrate from 0 to X:
WL2
102 x 10
-10
J per
Au
1 .8
( X x)
0. 8
dx
The integration is accomplished by substitution to give

WL2 = 102 x 10
-10
J per X 1.8
A u 1. 8 1 . 8
But the total residence time is tr = X/u, giving

WL2 = 56.7 x 10-10
J per 1.8
tr
A
(13.17)
Example 2
Continuing from Example 1, the known values are
u = 2 m/s
tr = 400 seconds
per = 14 m
Determine the activity of radon daughters at exit due to radon being emitted from the airway
surfaces at a rate of J = 250 pCi/(m2s).
Solution 2
=
WL2
56.7 x 10-10
250 14
(400)1.8 =
12
0.080 WL
(iii) Continued aging of radon daughters that existed in the air at the entry point
In general, the air that enters at location 0 on Figure 13.6 will already contain radon
daughters. These will continue to age and, hence, decay to a reduced working level by the
time they reach the exit. The fact that the daughters are continuously being replenished from
the decay of radon can be ignored as those effects have been considered separately in (i)
and (ii).
The reduction in working levels may be read from the decay curve on Figure 13.5 and
adjusted for the initial working level, WLin or, as the decay curve is a mirror image of the
growth curve, calculated as
WL3
WLin (1 - 102 x 10-5 tr 0.8 )
WL
Example 3
If the activity of radon daughters at entry into the airway of Example 2 is 0.05 WL, then after
the residence time of 400 s, this will have decayed to
WL3
0.05 {1 102 x 10-5 (400)0.8} = 0.0438
WL
There is, however, a weakness in this technique. It assumes that the radon daughters are in
equilibrium with each other and with their parent radon gas. This is not usually the situation in
13 -
17
ventilated areas underground. Further analysis of a suggestion made by Schroeder and

Evans (1969) leads to an improved estimate:
(t + t r ) 0.8 (t r ) 0.8
WL3 = WLin in
(t in ) 0.8
(13.18)
where tin = the age of the air at entry.

Let us repeat this previous example using the Schroeder and Evans technique, given that the
radon concentration at entry is C = 20 pCi/litre.
An estimate of the age of the air at entry, tin, is given by an inversion of equation (13.15).
1
tin
0.8
100
1
WL in
C 102 x 10 5
(s)
(13.19)
0.8
100
1
0.05
5
20 102 10
= 970 seconds
(This result can be estimated from Figure 13.9 at WL = 0.05 and radon concentration = 20 pCi/litre.)
WL3
(970 + 400 ) 0.8 (400 )0.8

0.05
(970 ) 0.8
0.0413
WL
To complete the series of examples given in this Section, the total working level of radon
daughters exiting the airway is
WLout
= WL1 + WL2 + WL3

= 0.025 + 0.080 + 0.041
0.146 WL
13.5. METHODS OF MONITORING FOR RADIATION

When radioactive emissions strike the atoms of other substances, they produce effects that may
include increases in temperature or secondary radiation. These effects can be measured and are
a function of the level of the primary emission. In general, there are two types of radiation
instrumentation, the thermal and photosensitive detectors.
In thermal detectors, the radiation is directed on to the hot junctions of a series of thermocouples
(thermopile) or a resistance thermometer within an evacuated chamber. The increase in
temperature of these sensors produces an electrical output that is representative of the level of
radiation. Another type employs a sensitive gas thermometer. Thermal detectors tend to be
fragile and are less suitable for portable instruments.
The most widely used principle of radiation detection in subsurface openings is photosensitivity.
The radiation is aimed at a material that emits photons (quanta of light) when irradiated. Zinc
13 -
18
sulphide is commonly employed. The photons are directed towards a photomultiplier tube (PMT)
where the light is amplified and converted to electrical pulses for counting and display.
13.5.1. Measurement of radon daughters
A number of instruments have been devised to measure the working levels of radon daughters
(Williamson, 1988). In the Kusnetz (1956) method, the air is pumped for 5 minutes at a steady
rate of 2 to 10 litres/minute through a filter of pore size less than 0.8 microns. The radon
daughters collected on the filter are allowed to age for a further 40 to 90 minutes and are then
exposed to a photomultiplier tube. The pulses of output energy are counted over a period that
depends upon the level of activity but should be small compared with the delay period, typically 1
to 2 minutes (Calizaya, 1991). The concentration of radon daughters in working levels is then
given as
WL =
where C
CE
TF
and
C CE
TF V
=
=
=
(13.20)
measured count rate (counts/min)

counter efficiency (instrument factor)
time factor corresponding to the 40 to 90 minute delay (Figure 13.7)
between the end of sampling and the midpoint of the counting interval
sample volume (litres)
A disadvantage of the Kusnetz and similar methods is the delay between sampling and
measurement. This limits the number of samples that can be taken in any one shift. Instant
radiation meters reduce the sampling and measurement cycle to a few minutes but may suffer
from reduced accuracy, particularly at low levels of activity (Williamson, 1988). These instruments
also involve the collection of radon daughters on filters and employ photomultiplier tubes and
display units to indicate count rates. The Instant Working Level Meter (IWLM) gives gamma
radiation in mRem/h as well as separate counts for alpha and beta radiation.
150
140
Kusnetz Time Facto r TF
130
120
110
100
90
80
70
60
30
40
50
60
70
80
90
100
Time (minutes)
Figure 13,7 Time factors for the Kusnetz method.

13 -
19
13.5.2. Measurement of radon concentration

An early grab sample technique for measuring radon concentration is the Lucas flask. This is a
lucite container whose sides and top are coated with zinc sulphide. To measure radon
concentration the flask is first evacuated by a vacuum pump. Sample air is then admitted via a
valve and filtered to remove radon daughters. The sample is allowed to age for about 3 hours in
order to achieve secular equilibrium. A flash of light (photon emission) occurs when an alpha
particle strikes the wall of the container. A window at the bottom of the flask is attached to a
photomultiplier unit and a reading taken of the output count rate.
The radon concentration is given as
(C b ) TF
FE 2.22 3 CE V
pCi
litre
(13.21)
where C = counts per minute

b = counts per minute due to background radiation from the ambient surroundings
TF = time factor (varies from 0.9762 at 2.5 hours to 1.0051 at 3.5 hours of elapsed time)
FE = flask efficiency (given by flask manufacturer)
2.2 = conversion of counts per minute to pCi (60 x 0.037 from equation (13.8))
CE = counter efficiency (instrument factor)
V = volume of flask (litres)
and
The remaining constant of 3 arises from the fact that alpha emission occurs at three levels during
decay from radon to RaD. This is shown on Figure 13.1. At secular equilibrium, the alpha
emissions at each level are equal. As the instrument detects total alpha activity the result must be
divided by 3 to give the concentration of radon alone (Calizaya, 1985).
Further developments have produced instruments suitable for continuous and integrating
sampling over longer periods. These are useful in establishing variations and average values of
radon concentrations at fixed locations.
13.5.3. Personal dosemeters
Radiation badges or personal dosemeters are designed to be attached to the clothing and
provide a measure of the cumulative radiation dosage to which the wearer has been subjected.
Thermoluminescent dosemeters (TLD) consist of four luminescent phosphors. At periods of one
to three months each badge is processed by heating it to a given temperature. The amount of
light emitted from each phosphor indicates the average levels and types of radiation to which the
badge has been exposed. Another type of radon detector employs an element consisting of polyalyl diglycol carbonate (PADC detector). A difficulty with personal detectors is that they may be
incapable of distinguishing between the radon daughters which are the main radiation hazard in
mines and other less harmful forms of radiation including the radon gas itself (Howes, 1990).
They are also susceptible to changes in atmospheric pressure, temperature and humidity that are
characteristic of subsurface environments.
Miniature versions of the pump and filtration units may prove to be more reliable than current
radiation badges. However, these are likely to be cumbersome as well as expensive for routine
use. Electronic dosemeters have also been developed for personal use ( Bartlett, 1993). To this
time, personal dosemeters have not found widespread use in subsurface workings.
13 -
20
13.6. CONTROL OF RADIATION IN SUBSURFACE OPENINGS

The control of radon and its daughters in underground mines should be addressed during the
design of the mine layout, choice of mining method and in selecting mineral transportation routes
as well as in planning the ventilation system. In this Section, we shall discuss the measures that
may be taken to reduce the hazard of radiation in mines. The guidelines that are suggested have
been established through a combination of practical observations and theoretical analyses.
Although these guidelines can be followed without regard to theoretical background, their success
is better assured if the ventilation engineer is familiar with the earlier sections in this chapter. The
concepts of aging and residence time are particularly important. It follows that the need for rapid
removal of radon and its daughters results in higher airflow requirements in uranium mines than
for most other subsurface openings. The high operating costs that can ensue make it particularly
important to design the ventilation system with a view to high efficiency and to employ the
techniques of computer-assisted network analysis (Chapter 7). A further prerequisite is to obtain
data on the geology of the area, rock surface emanation (J) rates and emanating powers of
fragmented ore (B) (Section 13.3.1.).
13.6.1. Ventilation systems for uranium mines
Particular regard should be paid to the locations and dimensions of intake airways in uranium
mines. The purpose is to deliver the intake air to stoping areas as free as practicable from radon
or its daughters. There are three methods of achieving this objective. First, the intakes should be
driven in the native rock and, as far as possible, not within the orebody. Such airways will be less
subject to emanations of radon from the rock surfaces.
Secondly, the residence times of air in intake airways should be kept to a minimum. Air velocity
limits normally accepted for the purposes of economics and dust control (Section 9.3.6.) are
frequently exceeded in the intakes of uranium mines. Ventilation requirements for uranium mines
are usually much higher than for other mines. Thirdly, where long intake airways are unavoidable,
then even higher velocities may be necessary and consideration may be given to the use of
airway liners. These are discussed further in Section 13.6.7.
Within the stoping areas, emphasis should, again, be placed on rapid air changes. Series
ventilation should be avoided and when booster fans are used, particular care should be taken in
the choice of their locations and duties in order to minimize recirculation. Uranium mines are a
case in which systems of controlled partial recirculation should not be employed.
Pressure differentials across sealed old workings should be in a direction such that any leakage
will pass into return airways and not into intakes. The "dirty pipe" principle may be used to
advantage in the design of ventilation systems for uranium mines (Section 18.3.1.). The number
of personnel required to work or travel in the return airways of uranium mines should be kept to a
minimum.
In order to provide young air to the faces of headings in uranium mines, it is preferable to
employ forcing systems of auxiliary ventilation. An exhaust overlap duct and filter may be added
to deal with dust problems (Section 4.4.2.).
13.6.2. Dilution and mixing processes
At a constant rate of emission, the rise in concentration of non-radioactive gases is inversely
proportional to the rate of through flow of fresh air (Section 9.3.1.). This is not the case for radon
daughters because of the ongoing effects of radioactive decay. Equation (13.17) indicates that if
an airway or stope is supplied with uncontaminated air and the rate of radon emanation remains
13 -
21
constant, then the exit working level of radon daughters is proportional to the residence time
raised to the power 1.8, i. e.,
WL (tr)1.8
(13.22)
where means 'proportional to'.

This shows that if the airflow is halved and, hence, the residence time doubled, then the exit
working level of radon daughters will increase by a factor of 21.8 = 3.48.
A more general relationship is gained by substituting
Q
1
for a given airway geometry, where Q = airflow (m3/s)
tr
giving
WL
(13.23)
Q 1.8
Then
WL 1
WL 2
Q
= 2
Q1
1.8
(13.24)
Example 1
A mine opening is ventilated by an airflow of 10 m3/s. The exit concentration of radon daughters
is 0.9 WL. If this is to be reduced to 0.33 WL, determine the required airflow.
Solution
Q2
WL 1
Q1
WL 2
1. 8
0 . 9 1. 8
10
0.33
= 17.46 m3 /s
Example 2
The radon daughter concentration leaving a mine section is 0.3 WL when the airflow is 15 m3/s. A
temporary obstruction caused by stocked materials reduces the airflow to 5 m3/s. Determine the
effect on the radon daughter concentration.
Solution
WL2
Q
= WL1 1
Q2
1. 8
15
= 0.3
5
1.8
2.17 WL
This is a dangerous concentration of radon daughters and illustrates the importance of

maintaining adequate airflows at all times in a uranium mine.
13 -
22
It should be recalled that equation (13.24), upon which this method of estimating the effects of
airflow is based, assumes that the air at entry to the opening is uncontaminated. This may not be
the situation in practice and, indeed, experience has shown that the formula often underestimates
the amount of air required (Rock and Walker, 1970).
When two airstreams of differing concentrations of radon daughters are mixed, then the resulting
concentration is given simply as the weighted mean:
WLmixture =
(Q WL )
(13.25)
Example 3
An airflow of 10 m3/s and radon daughter concentrations of 0.25 WL passes a seal from which
issues a leakage flow of 0.3 m3/s at 150 WL. (Near stagnant air in sealed areas can reach very
high concentrations of radon and its daughters.) Determine the radon daughter concentration in
the downstream airflow.
Solution
WLmixture
(10 0.25) + (0.3 150 )

10.3
= 4.61
WL
This example illustrates the dangerous levels of radiation that can arise from small leakages
through abandoned areas.
13.6.3. Radiation surveys
The preceding section makes it clear that modifications to the airflow distribution and small
leakages from old workings can have very significant effects on levels of radioactivity in mines
subject to radon emanations. In order to ensure the continuity of acceptable conditions and to
locate sources of contamination, it is useful to conduct radiation surveys. These involve taking
measurements of radon and radon daughters, commencing at points of fresh air entry and tracing
the primary ventilation routes through to mine exits. Figure 13.8 illustrates the type of results
produced by a radiation survey.
Sampling Control stations may be selected at strategic locations in order to establish timetransient trends or to correlate radiation levels with mining activities. Permanent monitoring
stations with recording and alarm facilities provide an even greater degree of Control (Bates and
Franklin, 1977).
The degree of equilibrium between the measured concentrations of radon and radon daughters
enables the age of the air to be established (Figures 13.5, 13.9 or equation (13.19)). High
equilibrium levels (old air) measured along the airflow paths are indicative of inadequate
ventilation, recirculation or leakage from old workings. Low equilibrium levels (young air) but
with elevated concentrations of radon and radon daughters imply that high rates of radon
emanation are occurring.
13 -
23
footwall
hanging
wall
orebody
1.0
upcast
shaft
0.5
stope
A
downcast
shaft
from
stope C
from
stope B
Figure 13.8 Example of a radiation profile produced from a radiation survey along
a main ventilation route in a uranium mine. (Airways adding or
subtracting air from the main route are not shown.)
13.6.4. Mining methods, mineral clearance and backfill

The choices of mine planning, stoping methods and operational procedures have a large influence on
the severity of a mine radiation problem that must, subsequently, be handled by the ventilation
system. In planning the extraction sequence, stoping areas should progress from the main exhaust
airways towards the trunk intake zones. This strategy of retreat mining ensures that worked out areas
and zones of fragmentation and voidage do not contribute towards the radioactive contamination of
current workings. Leakages from abandoned areas pass directly into return airways.
Stoping methods should avoid systems that involve large areas of exposed ore, sluggish ventilation
or high tonnages of fragmented rock. Hence, open or shrinkage stoping and caving techniques are
not advisable in uranium mines. Peak emanations of radon occur during and after blasting. It is
particularly important that adequate re-entry periods are employed for the clearance of blasting fumes
and the associated radon daughters before personnel are allowed to return to the workings.
13 -
24
Radon daughters (Working Levels)
10
0.1
0.01
0.001
0.0001
10
100
1000
10000
Time seconds
Figure 13.9 Growth of radon daughters as a function of time and radon concentration.
Piles of fragmented ore may produce high radiation levels, especially when they are disturbed by
mucking operations. This can result in very large variations in radon daughter concentrations
during a mining cycle. The broken ore should be transported from the mine as rapidly as
practicable. Where some aeration of fragmented ore is unavoidable such as in orepasses or at
transfer points, then consideration should be given to the use of exhaust hoods or air bypasses to
route the contaminated air directly into a return airway. The number of ore handling operations
within the mine should be kept as low as possible and out of main intake airways.
Haulage airways should be well maintained in order to avoid unnecessary comminution of the ore
or spillage during transportation. Furthermore, the production of dust particles from uraniumbearing rocks will cause increased levels of radiation (Bigu and Grenier, 1985). The settlement of
13 -
25
such dust particles within subsurface airways produces an escalating source of radon. Dust
suppression by water sprays or filters is particularly important in uranium mines (but refer to
Section 13.6.6 for the effects of water vapour on emanating surfaces).
Localized peaks of radon emanation may occur during drilling operations either for blasting or for
orebody exploration. Consideration should be given to the location of machine operators. Long
exploration holes should be sealed. Blasting patterns for developments should be selected to
minimize the overbreak envelope of fractured rock around the opening. The induced fracture
network within this envelope produces additional surface area for radon emanation as well as
enhancing the inflow of radon-contaminated groundwater. Good strata control techniques
including rock bolting and other support methods will help to minimize radon emanations.
In addition to controlling ground movement, the employment of backfill material will reduce
leakage flows through old workings. Both of these features are particularly important in uranium
mines. However, tests should be carried out on the radon emanation characteristics of the fill
material itself, particularly when it contains mill tailings. Freshly placed wet fill may emanate radon
gas at about twice the rate of the consolidated fill (Bates and Franklin, 1977; Thompkins, 1982).
The addition of cement to the fill material can result in a reduction of the radon emanation rate. If
backfilling operations produce a significant amount of radon, then care should be taken to ensure
that the contaminated air is exhausted into return airways.
13.6.5. Contamination from abandoned workings
Example 3 in Section 13.6.2. illustrates the high level of radioactive contamination that can occur
in uranium mine ventilation systems when slight leakage occurs from abandoned areas or
unventilated blind headings. The near stagnant air in such zones achieves a high concentration of
radon at secular equilibrium with the radon daughters (fully aged). Radon concentrations of many
thousands of pCi/litre may occur behind seals in uranium mines. It becomes particularly important
that barrier seals in uranium mines should be constructed and maintained to a high standard. The
faces of stoppings and adjoining rock walls may be coated with additional sealant material
(Section 13.6.7.). However, minor transients in atmospheric pressure can cause breathing
through seals and stoppings, resulting in peak emanations of radon and radon daughters during
periods of falling barometric pressure (Section 4.2.2.).
Such effects occur not only from sealed areas of the mine but also from fracture networks and
other voidages within the strata. Attempts have been made to modify mine atmospheric
pressures by fan control in uranium mines such that air pressures are elevated during working
shifts and depressed when few or no persons are underground (Schroeder et al, 1966; Bates and
Franklin, 1977). Leakages from sealed areas can be controlled by pressure balance chambers
(Section 21.5.5.) which maintain pressure differentials across seals at near zero. Another
technique is to employ a bleed pipe that connects the sealed area to a main return airway or
through a vertical borehole to surface. The sealed zone can be maintained at sub-atmospheric
pressure by employing a low capacity extractor pump or fan within the bleed pipe. Leakage into
the sealed area then remains safely in an inward direction. In the absence of these pressure
control techniques, a small pressure differential should be maintained across the sealed area
such that any leakage that occurs will be into return airways.
13.6.6. The influence of water
The emanation of radon from mineral crystals into the pores of a rock will be essentially the same
whether the interstices are filled with air or water. However, any migration of the groundwater can
provide a transport mechanism for the dissolved radon that is more efficient than diffusion of the
gas through a dry rock. When the water reaches a mine opening it will yield up its dissolved radon
very readily, particularly if the water is aerated by spraying or dripping into the airway. It is
13 -
26
prudent to capture such water into pipes as soon as possible in uranium mines and minimize its
exposure to the air.
In permeable wet strata, the emanation of radon can be reduced significantly by pre-draining the
area. This may be achieved by pumping from a ring of drainage boreholes. Better results can be
obtained by driving drainage levels below the stopping areas prior to mining. The effectiveness of
this technique can be further enhanced by boreholes drilled into the strata from the drainage
levels. In severe cases, the flow of water into development headings can be reduced by grouting.
Radioactive decay of the radon occurs whether it is contained within air or water. Hence, the
concentration of radon in groundwater depends upon the elapsed time since the gas was emitted
from mineral crystals into the water-filled interstices. Mine water that contains dissolved radon
should not be used for dust suppression sprays. However, if it is first brought to the mine surface
and aerated, then the radon content may be diminished to a level that renders the water suitable
for dust suppression.
Experimental observations indicate that commencing with dry rock, radon emanations can
increase dramatically as the moisture content of the rock increases (Bates and Franklin, 1977). A
similar effect occurs from raising the moisture content of the air that is in proximity to rock
surfaces. However, as the rock or air approaches saturation, the effect diminishes and radon
emanations fall when liquid water appears on the rock surface. The mechanisms that produce
these phenomena appear not to be clearly understood. It is thought that displacement of
adsorbed radon by water molecules on mineral surfaces may explain the initial increase in
emanation rates, while the later inhibition of radon release may be due to the interstices near the
surface becoming filled with liquid water.
13.6.7. Air filters and rock surface liners
As radon daughters are particulates, a large proportion of them can be removed by passing the
air through high efficiency dust filters. These should be capable of removing at least 95 percent of
particles 0.3 microns in size. As such filters are relatively expensive and can rapidly become
blocked in mining conditions, fibreglass prefilters may be used to remove the coarser particles
and, hence, improve the life of the high efficiency filters (Rock and Walker, 1970). The latter are
also affected adversely by humid conditions. The pressure drop across filters can be monitored to
indicate when renewal or cleaning has become necessary.
The major drawback to filters is that they do not remove the radon gas that continues to replenish
radon daughters. The filtered air must be supplied quickly to the personnel who are to be
protected. Figure 13.9 or equation (13.15) show that even if perfect filtration of particulates is
achieved, a radon concentration of 100 pCi/litre will generate 0.3 WL of radon daughters in 1218
seconds (20.3 minutes). However, if the radon concentration is 500 pCi/litre, then 0.3 WL of
radon daughters will appear in only 163 seconds (2.7 minutes) after filtration. Filters for radon
daughters are perhaps most effective for forcing duct systems supplying rejuvenated air to
headings.
Activated charcoal can remove radon gas from air. At the present time, large scale applications
appear to be impractical. However, gas masks with activated charcoal filters can be used to
protect personnel who are required to venture into high radon and radon daughter concentrations
for a short time.
A number of trials have been carried out into the use of sprayed coatings and film membranes to
reduce radon emanations into mine openings. These are unlikely to replace good ventilation as
the primary means of combatting the radon problem. However, they can have an application in
long intakes driven in high grade ore or permanent work places such as workshops. Liquid sprays
are suitable for application on rock surfaces while membranes may be attached to the faces of
13 -
27
stoppings. Grouting of the rock envelope with or without rock bolting can also be effective in
reducing inflows of both radon and water.
Tests conducted by the U. S. Bureau of Mines investigated the ability of a range of polymer
sealants to resist the passage of radon (Bates and Franklin, 1977). The permeability of a material
with respect to air has little bearing on its resistance to radon gas. The latter is a monatomic gas
that will diffuse through most substances. Furthermore, any lining material intended for use in
mines should have low toxicity and flammability both during application and after curing has been
completed. Additionally, it should be convenient to apply, inexpensive and not productive of
smoke or toxic fumes when heated. These stringent requirements limit the use of radon barriers
in practice.
The U. S. Bureau of Mines tests indicated that water-based epoxies were suitable for application
to rock surfaces. However, two spray applications are recommended, the first with a low viscosity
liquid to penetrate surface interstices, and the second with a thicker fluid to seal visible fractures.
Using differing colours for the two applications assists in achieving full coverage.
Further tests in Canada indicated that a polysulphide copolymer spray and aluminized mylar
sheeting were both capable of reducing radon diffusion through a bulkhead by more than 95
percent (Archibald and Hackwood, 1985). Additional data on the permeability of membranes to
radon is given by Jha et al (1982).
13.6.8. Education and training
Radon gas and radon daughters are a particularly insidious hazard. They are invisible, odourless
and can be detected only by specialized instruments. Furthermore, they have no short term
observable effects on the human body. It becomes particularly important that all workers in
affected environments should be made aware of the health hazards associated with radon and
the steps that can be taken to alleviate the problem. Booklets, videotapes and classroom teachins are particularly effective. These should emphasize the importance of a brisk throughflow of air
and the hidden dangers that may arise from recirculation, damaged stoppings or ventilation doors
left open.
References
Archibald, J.F. et al (1980). Determination of radiation levels to be encountered in underground
and open-pit uranium mines. 2nd Int. Congress on Mine Ventilation. Reno, Nevada pp 399-404.
Archibald, J. F. and Hackwood, H. J. (1985). Membrane barriers for radon gas flow restriction,
2nd U. S. Mine Ventilation Symposium, Reno, Nevada, pp. 251-257
Bartlett, D.T. (1993). Electronic Dosemeters: Use in personal dosimetry. Radiation protection
Dosimetry, Vol. 47, No. 1, pp 335-339. Oxford University Press.
Bates, R. C. and Edwards, J. C. (1980). Mathematical modeling of time dependent radon flux
problems, 2nd Int. Congress on Mine Ventilation, Reno, Nevada, pp 412-419.
Bates, R. C. and Franklin, J. C. (1977). U.S. Bureau of Mines Radiation Control Research, Conf.
on Uranium Mining Technology, Reno, Nevada.
Bigu, J. and Grenier, M. G. (1985). Characterization of radioactive dust in Canadian
underground uranium mines, 2nd U. S. Mine Ventilation Symp., Reno, Nevada, pp 269-277.
Calizaya, F. (1991). Private communication.
13 -
28
Calizaya, F. (1985). Control study of the evolution of radon and its decay products in radioactive
mine environments, Ph. D. thesis, Colorado School of Mines.
CFR. (1990) US Code of Federal Regulations, Vol. 30 (Mineral Resources), Part 57, Metallic
and non-metallic underground mines, US Govt. Printing Office, Washington, D. C.
Howes, M. J. (1990). Exposure to radon daughters in Cornish tin mines, Trans. Inst. of Mining &
Metallurgy, U. K., Vol. 99, pp A85-A90.
Jha, G. et al (1982). Radon permeability of some membranes. Health Physics, Vol. 42, 5, pp
723-725.
Kusnetz, H. L. (1956). Radon daughters in mine atmospheres -a field method for determining
concentrations, Ind. Hygiene Quart., March, pp 85-88.
Rock, R. L. and Walker, D. K. (1970). Controlling employee exposure to alpha radiation in
underground uranium mines, U. S. Bureau of Mines, U. S. Government Printing Office,
Washington, D. C.
Schroeder, G. L. and Evans, R. D. (1969). Some basic concepts in uranium mine ventilation,
Trans. AIME, Vol. 244, pp 301-307.
Schroeder, D. L. et al (1966). Effect of applied pressure on the radon characteristic of an
underground mine environment, Trans. AIME, Vol. 235, pp 91-98.
Thompkins, R. W. (1985). The safe design of a uranium mine, 2nd US Mine Ventilation Symp.,
Reno, Nevada, pp 289-294.
Thompkins, R. W. (1982). Radiation in uranium mines, CIM Bulletin, Vol. 75, Nos. 845, 846, 847.
Williamson, M. J. (1988). The exposure of mining personnel to ionizing radiations in Cornish tin
mines, 4th Inst. Mine Ventilation Congress, Brisbane, Australia, pp 585-592.
13 -
29
Part 4
Heat and Humidity
Psychrometry: The study of moisture in air
CHAPTER 14. PSYCHROMETRY: THE STUDY OF MOISTURE IN AIR
14.1. INTRODUCTION .........................................................................................2

14.2. BASIC RELATIONSHIPS.............................................................................3
14.2.1. Basis of measurement ..................................................................................................... 3
14.2.2. Moisture content (specific humidity) of air ....................................................................... 4
14.2.3. Saturation vapour pressure ............................................................................................. 5
14.2.4. Gas constant and specific heat (thermal capacity) of unsaturated air............................. 7
14.2.5. Specific volume and density of unsaturated air ............................................................... 8
14.2.6. Relative humidity and percentage humidity..................................................................... 9
14.3 THE MEASUREMENT OF WATER VAPOUR IN AIR ................................11

14.3.1. Chemical methods ......................................................................................................... 11
14.3.2. Electrical methods (electronic psychrometers or humidity meters) ............................... 11
14.3.3. Hair hygrometers ........................................................................................................... 11
14.3.4. Dew point hygrometers .................................................................................................. 12
14.3.5. Wet and dry bulb hygrometers (psychrometers) ........................................................... 12
14.4. THEORY OF THE WET BULB THERMOMETER......................................13

14.4.1. Heat balance on a wet bulb ........................................................................................... 14
14.4.2. Determination of moisture content and vapour pressure from psychrometer readings 15
14.5. FURTHER PSYCHROMETRIC RELATIONSHIPS....................................17

14.5.1. Enthalpy of moist air ...................................................................................................... 17
14.5.2. The adiabatic saturation process................................................................................... 19
14.5.3. Sigma Heat, S................................................................................................................ 20
14.6. SUMMARY OF PSYCHROMETRIC EQUATIONS ....................................21

14.7. DEVIATIONS FROM CLASSICAL THEORY .............................................22
14.7.1. Fogged air...................................................................................................................... 22
14.7.2. Imperfect gas behaviour ................................................................................................ 23
14.8. PSYCHROMETRIC CHARTS....................................................................24

References ........................................................................................................28
APPENDIX A14 ..................................................................................................29
Derivation of the Clausius-Clapeyron equation...................................................................... 29
14 -
14.1. INTRODUCTION
Around the surface of the earth, air that is not affected by any local source of pollution has a
composition that is remarkably constant. Air analyses are usually carried out on samples from
which all traces of water vapour have been removed. The moisture-free composition of air is
given on both a volume and mass basis in Table 14.1. Argon forms the largest fraction, by far, of
the monatomic gases. The value of the molecular weight given is, in fact, that for argon.
Gas
Nitrogen
Oxygen
Carbon Dioxide
Hydrogen
Monatomic gases
Volume
per cent
Mass
per cent
Molecular
weight
78.03
20.99
0.0381
0.01
0.94
75.46
23.19
0.05
0.0007
1.30
28.015
32.000
44.003
2.016
39.943
Equivalent molecular weight of dry air
28.966
Table 14.1 Composition of dry air

However, there is another gas present in the free atmosphere, water vapour, that is rather
different to the others in that its concentration varies widely from place to place and with time.
This is because the pressures and temperatures that exist within the blanket of air that shrouds
our planet also encompass the ranges over which water may exist in the gaseous, liquid or solid
forms - hence the appearance of clouds, rain, snow and ice. Evaporation of water, mainly from
the oceans, coupled with wind action, produce and transport water vapour through the
atmosphere. Increases in pressure or, more effectively, decreases in temperature may result in
condensation of the water vapour to form clouds which, in turn, can produce droplets large
enough to be precipitated as rain, snow or hail.
On other planets with very different atmospheres and gravitational fields, similar phase changes
occur in other gases. Because of its variable concentration within the earth's atmosphere,
airborne water has become the subject of a special study, psychrometry.
Changes of phase are particularly important within the confines of closed environments, including
subsurface ventilation systems. Ice to liquid and ice to vapour phase changes occur in mines
located in cold climates and, particularly, if situated in permafrost. However, the vast majority of
humidity variations that occur in underground airflows are caused by the evaporation of liquid
water or the condensation of water vapour. This chapter concentrates on the phase changes
between liquid water and water vapour. Virtually all mines produce water from the strata and/or
dust suppression techniques. Even with the hygroscopic minerals of evaporite mines, the water
vapour content in return airways is normally higher than that in the intakes.
Subsurface environmental engineers have a particular interest in psychrometry for two reasons.
First, if we are to comprehend fully the thermodynamic processes that occur in ventilation circuits
then variations in humidity must be taken into account. For example, strata heat may be emitted
into a wet airway without there being a corresponding increase in air temperature. This could
occur if all the added heat were utilized in exciting some of the water molecules until their kinetic
energy exceeded the attractive forces of other molecules in the liquid water. They would then
escape through the liquid/air surface and become airborne as a gas. The process of evaporation
1
The concentration of atmospheric carbon dioxide varies with both location and time. At the time of writing,
the average global concentration is increasing at about 0.00016 percent per year
14 -
increases the energy content of the air/vapour mixture. This may be termed a latent (or hidden)
rise in the heat content of the air as there is no commensurate increase in temperature and,
hence, no indication on an ordinary thermometer.
Alternatively, if there were no liquid water present, then the strata heat would be directed
immediately to the airstream, causing a temperature rise of the air that would be sensed by a
thermometer. This is an increase in the sensible heat of the air.
These examples illustrate that if we are to predict quantitatively the climatic effects of strata heat,
water inflows, machines or air coolers, then we need to have methods of analysis that take
humidity into account.
The second reason for the study of psychrometry is the effect of heat and humidity on the human
body. This is examined in detail in Chapter 18. However, for the time being, we will concentrate
on developing means of quantifying the psychrometric relationships that enable predictions to be
made of temperature and other climatic variables in the environment.
14.2. BASIC RELATIONSHIPS

Most of the psychrometric equations that are used in practice are based on the premise that air is
a mixture of perfect gases and that the air itself behaves as a perfect gas. Within the ranges of
temperatures and pressures that are reasonable for human tolerance, this assumption gives rise
to acceptable accuracy. The majority of this chapter assumes perfect gas laws. However, in
certain areas of some underground facilities, (including possible future mining scenarios) the
atmosphere will require control, but not necessarily within physiologically acceptable ranges. For
this reason, more accurate relationships are included that take some account of deviations from
the perfect gas laws.
14.2.1. Basis of measurement
A question that should be settled before embarking on a quest for psychrometric relationships is
how best to express the quantity of water vapour contained within a given airstream. As we saw
in Chapter 3, it is preferable to conduct our analyses on a mass (kg) basis rather than volume
(m3), as variations in pressure and temperature cause the volume of the air to change as it
progresses through a ventilation system. That choice relied upon the assumption that the mass
flow of air remained constant along a single airway. Now we are faced with a different situation.
The addition of water vapour to an airstream through evaporative processes, or its removal by
condensation, result in the mass flow of the air/vapour mixture no longer remaining constant.
Within the mixture, molecules of water vapour coexist with, and occupy the same volume, as the
nitrogen, oxygen and other gases that comprise the air. We can assume, however, that in the
absence of chemical reactions or the addition of other gases, it is only the concentration of water
vapour that varies, due to evaporation and condensation. The mass flow of the rest of the air
remains constant.
It is a convenient, although somewhat artificial, device to consider the air to be divided into a fixed
mass of "dry air" and an associated but variable mass of water vapour. For most purposes, we
can then refer to the moisture content in terms of grams or kilograms of water vapour per
kilogram of "dry air". Occasionally, we may use the alternative measure of grams or kilograms of
vapour per kilogram of the real mixture of air and vapour. Throughout this chapter we shall use
the term "air" to mean the actual mixture of air and water vapour, and "dry air" for that fraction
which does not include the water vapour. Hence, a moisture content of 0.02 kg/kg dry air means
that in each 1.02 kilograms of air, 0.02 kilograms are water vapour and 1 kilogram is "dry air".
14 -
14.2.2. Moisture content (specific humidity) of air

In order to quantify the mass of water vapour associated with each kilogram of "dry air", let us
conduct an imaginary experiment.
Suppose we have a closed vessel of volume V m3 containing 1 kg of perfectly dry air at a
pressure Pa Pascals and temperature T degrees Kelvin. If we inject X kg of water vapour at the
same temperature, the pressure within the vessel will increase to:
P = Pa + e
Pa
where e is the partial pressure exerted by the water vapour (Dalton's law of partial pressures for
perfect gases). Both the air and the water vapour occupy the same volume, V, and are at the
same temperature, T.
The problem is to determine the mass of water vapour, X, if we can measure nothing more than
the initial and final pressures, Pa and P.
From the general gas law:
(i) for the X kg of water vapour:
eV = X Rv T
where Rv
Joules (J)
(14.1)
gas constant for water vapour (461.50 J/kg K), and
(ii) for the original 1 kg of dry air,

Pa V = 1 x Ra T
where Ra
(14.2)
gas constant for dry air (287.04 J/kg K)
Dividing equation (14.1) by (14.2) gives

R
e
= v X
Pa
Ra
However, the absolute (barometric) pressure in the vessel is

P
Pa + e
Hence,
X =
Ra
e
R v (P e )
(14.3)
Inserting the values of the gas constants gives

X
0.622
e
(P e )
kg
kg dry air
(14.4)
This gives the moisture content of the vessel in kg per kg of dry air, provided that the partial
pressure of the water vapour, e, can be evaluated. This is, of course, simply the difference
between the initial and final absolute pressures. However, e can be determined independently, as
we shall see a little later in the chapter.
14 -
14.2.3. Saturation vapour pressure

Returning to our experiment, suppose we continue to inject water vapour. The partial pressure of
water vapour (and, hence, the absolute pressure in the vessel) will continue to rise - but only to a
certain limiting value. The partial pressure of the dry air fraction, Pa, will remain constant. If we
insist upon forcing yet more vapour into the system while keeping the temperature constant, then
the excess will condense and collect as liquid water on the sides and bottom of the vessel. (It is,
in fact, possible to achieve a condition of supersaturation in the laboratory but that will not occur
in natural atmospheres and is not considered here).
When the system refuses to accept any more water vapour then we say, rather loosely, that the
air has become saturated. In fact, it is not the air but the space that has become saturated. If we
were to repeat the experiment starting with the vessel evacuated and containing no air then
exactly the same amount of vapour could be injected before condensation commenced, provided
that the temperature remained the same.
The pressure, es, exerted by the water vapour at saturation conditions depends only upon
temperature and not upon the presence of any other gases. The relationship between saturation
vapour pressure and temperature for water has been determined not only experimentally but also
through thermodynamic reasoning by a number of authorities. The tables of Goff and Gratch
produced in 1945 are still regarded as a standard. The simplest analytical equation is known as
the Clausius-Clapeyron equation.
1 de s
L
=
e s dT
Rv T 2
(14.5)
where the latent heat of evaporation, L (J/kg) is the heat required to evaporate 1 kg of water.
The derivation of the Clausius-Clapeyron equation is given in the Appendix A14 at the end of the
chapter. This equation is based on the perfect gas laws and must, therefore, be regarded as an
approximation. Furthermore, the latent heat of evaporation, L, is not constant. The higher the
initial temperature of the liquid water then the less will be the additional heat required to
evaporate it. The relationship between latent heat of evaporation and temperature is near linear.
The equation
L = (2502.5 - 2.386 t) 1000 J/kg
(where temperature, t, is in degrees Centigrade)
(14.6)
is accurate to within 0.02 per cent over the range 0 to 60 C.

The equation may also be written as
L
(3154.2 - 2.386 T) 1000
J/kg
(14.7)
where the temperature T is in degrees Kelvin.

The Clausius Clapeyron equation can be integrated over any given interval
2
de s
es
1
Rv
(a bT )
T2
dT
where a = 3 154 200 J/kg and b = 2386 J/kg K from equation (14.7).
14 -
Integrating gives
e
ln s 2
e s1
T1
1 1
1
=
a
+ b ln
Rv T1
T2
T 2
giving
1
e s 2 = e s1 exp
R v
(T2 T1 )
T
+ b ln 1
a
T1T2
T2
Pa
(14.8)
If the saturation vapour pressure, es1, at any given temperature, T1, is known, then equation
(14.8) allows the saturation vapour pressure, es2, at any other temperature, T2, to be calculated.
At 100C, the saturation vapour pressure is one standard atmosphere (101.324 kPa) by definition
(Section 2.2.3). Hence, this may be used as a starting point for the integration.
There are two problems with equation (14.8). First, it is cumbersome for rapid calculation and,
secondly, the assumption of perfect gas behaviour in its derivation introduces some uncertainty.
The latter is not serious. Integrating down from a vapour pressure of 19.925 kPa at 60 C and
using equation (14.8) gives a maximum error of only 0.52 per cent over the range of 0 to 60 C.
Nevertheless, we can do much better and simplify the format of the relationship at the same time.
Let us take temperature T1 to be the freezing point of water, 273.15 K or 0C. Then temperature
T2 may be expressed in degrees Centigrade as
t
or
= ( T2 - 273.15)
T2 = (t + 273.15)
Furthermore, the logarithmic term in equation (14.8) can be rewritten as

T
ln 1
T2
(T T2
= ln 1 + 1
T2
t
= ln 1
T2
Expanding from the logarithmic series for the condition where t is much less than T2 gives the
term as approximately ( -t/T2 ).
Substituting into equation (14.8), simplifying and gathering constants together gives the form
Bt
es 2 = A exp
C + t
Pa
(14.9)
where A, B and C are constants.

Curve fitting from the standard tables of Goff and Gratch over the range 0 to 60C gives the
widely used equation:
17.27t
es 2 = 610.6 exp
237.3 + t
Pa
(14.10)
This is accurate to within 0.06 percent over the given range.
14 -
For the more usual mining range of 10 to 40C the curve fitting procedure gives
17.291t
es 2 = 610.162 exp
237.481 + t
Pa
(14.11)
having an excellent accuracy of within 0.01 per cent for that range.
Example
Find the saturation vapour pressure at a temperature of 30C.
Solution
(i) Inserting t = 30 C into equation (4.10) gives es = 4241.7 Pa or 4.2417 kPa
This is in error by 0.026 percent when compared with the value of 4.2428 kPa given by the
Goff and Gratch tables.
(ii) Using equation (4.11) gives es = 4.2431kPa, having an error of only 0.007 percent.
14.2.4. Gas constant and specific heat (thermal capacity) of unsaturated air
From the gas laws for 1 kg of dry air and X kg of associated water vapour, we have
and
Pa V
eV
=
=
1 Ra T
X Rv T
Adding these two equations gives

(Pa +e) V = (Ra + X Rv) T
or
PV
(Ra +X Rv) T
J/kg
(14.12)
Alternatively, we can treat the (1 + X) kg of air/vapour mixture as a perfect gas having an

equivalent gas constant of Rm. Then
PV
(1 +X) Rm T
J/kg
(14.13)
J/(kg K)
(14.14)
Equating (14.12) and (14.13) gives

Rm
(R a + X R v )
(1 + X )
Hence, we have shown that the gas constant for the moist air is given simply by adding the gas
constants for dry air, Ra, and water vapour, Rv, in proportion to the relative masses of the two
components.
The equivalent gas constant for moist air can also be expressed in terms of pressures by
substituting for X from equation (14.4), leading to
Rm
287.04
P
(P 0.378e )
J/(kg K)
(14.15)
14 -
Similarly, the equivalent specific heat (thermal capacity) of moist unsaturated air can be found by
adding the specific heats of the two components in proportion to their masses. The specific heat
at constant pressure becomes
C pm =
(C pa + XC pv )
=
=
where Cpa
and
Cpv
J/(kg K)
(1 + X )
(14.16)
Specific heat of dry air at constant pressure (1005 J/kg K)

Specific heat at constant pressure for water vapour (1884 J/kg K).
Also, the specific heat of moist air at constant volume is given as

(C va + X C vv )
(1 + X )
Cvm =
where Cva
and Cvv
=
=
J/(kg K)
(14.17)
Specific heat of dry air at constant volume (718 J/(kg K) )

Specific heat of water vapour at constant volume (1422 J/(kg K) ).
14.2.5. Specific volume and density of unsaturated air

The actual specific volume of moist unsaturated air, Vm, can be calculated from the general gas
law for 1 kg of the air/vapour mixture:
Vm
= Rm
T
P
Substituting for Rm from equations (14.14) and (14.15) gives

Vm =
and
(R a + X R v ) T
(1 + X )
P
Vm = 287.04
T
(P 0.378e )
m3/kg of moist air
(14.18)
m3 /kg of moist air
(14.19)
In these equations, T is in degrees Kelvin and P in Pascals.

As we saw in section 14.2.1 it is more convenient, for most purposes, to conduct our analyses on
the basis of 1 kg of dry air rather than a kilogram of the true mixture. The apparent specific
volume, based on 1 kg of dry air is simply
Vm (apparent) = 287.04
T
(P e )
m3/kg of dry air
(14.20)
The actual density of the moist air, m, is the reciprocal of the actual specific volume
or
(1 + X
P
(R a + X R v ) T
kg moist air/m3
(14.21)
(P 0.378e )
287.04T
kg moist air/m3
(14.22)
14 -
A useful approximate formula to calculate the effect of moisture on air density can be derived
from equation (14.21) giving
P
(1 0.608 X )
R aT
kg moist air/m m3
(14.23)
This equation makes clear that the density of air decreases as its moisture content rises.
The apparent density, based on 1 kg of dry air, is the reciprocal of equation (14.20).
m (apparent)
(P e )
287.04T
kg of dry air/ m3
(14.24)
14.2.6. Relative humidity and percentage humidity

Relative humidity is widely used by heating and ventilating engineers as an important factor
governing comfort of personnel in surface buildings. In open surface areas or in through-flow
subsurface facilities, its use as a physiological parameter is very limited and can be misleading,
owing to the wider range of air temperatures that may be encountered (see, also, Chapter 18).
On the other hand, the concept of relative humidity, rh, is a convenient way of expressing the
degree of saturation of a space. It is defined as
rh
e
100
e sd
percent
(14.25)
where esd = saturation vapour pressure at the air (dry bulb) temperature.
This definition indicates that relative humidity is the ratio of the prevailing vapour pressure to that
which would exist if the space were saturated at the same temperature. A similar concept is that
of percentage humidity, ph, defined as
ph
X
100
X sd
percent
(14.26)
where Xsd is the moisture content (kg/kg dry air) that would exist if the space were saturated at
the same dry bulb temperature.
Percentage humidity is approximately equal to relative humidity over the normal atmospheric
range. Substituting from equation (14.4) gives
ph
(P esd )
X
e
x 100
=
X sd
(P e ) esd
rh
(P e sd )
(P e )
As P is much larger than e or esd over the normal atmospheric range, the bracketed terms are
near equal.
14 -
Example
An airstream of temperature 20C and barometric pressure 100 kPa is found to have an actual
vapour pressure of 1.5 kPa.
Determine
(i) the moisture content of the air, X
(ii) the gas constant, Rm
(iii) the specific heat, Cpm
(iv) the actual and apparent specific volumes
(v) the actual and apparent densities
(vi) the relative humidity, rh
and
(vii) the percentage humidity, ph
Solution
(i) moisture content, X:
X
=
=
1.5
(100 1.5)
0.009472
kg vapour/kg dry air
0.622
or 9.472 g/kg dry air.
(ii) gas constant:

Rm
287.04 100
100 (0.378 1.5)
(iii) specific heat, Cpm

(1005 + 0.009472 1884 )
Cpm
=
(1 + 0.009472 )
288.68 J/kg K
= 1013 J/kg K
(iv) specific volumes:

From equation (14.19), Actual specific volume:
287.04(273.15 + 20)
Vm
=
1000(100 0.378 1.5)
= 0.8463
m3/kg moist air
(the 1000 in the denominator is necessary to convert the pressures from kPa to Pa
Apparent specific volume:
Vm (apparent) =
287.04(273.15 + 20)
1000(100 1.5)
0.8543
m3/kg dry air
(v) air densities:

Actual density is given by equation (14.22) or the reciprocal of actual specific volume
1
1
=
m
=
= 1.1816 kg moist air/ m3
Vm
0.8463
Similarly, the apparent density is given as
1
1
=
m (apparent) =
Vm (apparent )
0.8543
= 1.1705
kg dry air/m3
14 - 10
(vi) relative humidity, rh:

From equation (14.10), the saturation vapour pressure at a temperature of 20C is
esd
17.27 20
610.6 exp
(237.3 + 20)

e
rh =
100
=
e sd
1 .5
x 100
2.3375
2337.5 Pa or 2.3375 kPa
64.17
percent
(vii) percentage humidity, ph

If the space were saturated at 20C, then the saturation vapour pressure would be 2.3375 kPa.
Equation (14.4) gives the corresponding saturation moisture content to be
Xsd
0.622
2.3375
(100 2.3375 )
and equation (14.26) gives

X
ph =
x 100
X sd
0.014887
kg/kg dry air
0.009472
100 = 63.62 percent
0.014887
14.3 THE MEASUREMENT OF WATER VAPOUR IN AIR

There are a number of methods of measuring the humidity of air. Numerous instruments are
available commercially. These may be divided into five types.
14.3.1. Chemical methods
This involves passing a metered volume of air through a hygroscopic compound such as calcium
chloride, silica gel or sulphuric acid and observing the increase in weight. The method is slow and
somewhat cumbersome. However, it gives a direct measure of the total moisture content of an
airstream and, with care, will give results of high accuracy.
14.3.2. Electrical methods (electronic psychrometers or humidity meters)
The changes in the electrical properties (resistivity, dielectric constants) of some compounds in
the presence of water vapour are used in a variety of instruments to give a rapid indication of
humidity. Such equipment needs to be compensated against the effect of variations in ambient
temperature. Modern devices of this type may employ semi-conductors and electronic micro
processors to improve the versatility, reliability and stability of the instrument. Nevertheless, for
accurate work, it is advisable to check the calibrations of electronic psychrometers against a more
direct method.
14.3.3. Hair hygrometers
Many organic compounds such as bone, hair or other fibrous materials exhibit changes in their
volume and elasticity when exposed to water vapour. An example is the opening and closing of fir
cones as the weather changes. Human hair is particularly reactive.
Most inexpensive humidity meters sold for domestic display incorporate strands of hair
maintained under spring tension. Variations in humidity cause small changes in the length of the
14 - 11
hair. These are amplified mechanically through a lever arrangement. Recording instruments of
this type are used to move a pen across a revolving paper chart.
14.3.4. Dew point hygrometers
If an air/vapour mixture is cooled at constant barometric pressure then the actual vapour pressure
remains unchanged (see equation (14.4)). However, the falling temperature will cause the
corresponding saturation vapour pressure to decrease (equation (14.10)). At some point, the
saturation vapour pressure will become equal to the actual vapour pressure. According to
classical theory, condensation will then commence. (See also Figure 14.4).
The temperature at which saturation conditions are attained is known as the dew point
temperature. An inversion of equation (14.10) gives
e
237.3 ln
610.6
t (dew point) =
C
(14.27)
e
17.27 ln
610.6
In a dew point hygrometer, a sample of air is drawn over a mirrored surface. The rear of the
mirror is cooled, often by evaporation of a volatile liquid. A thermometer attached to the mirror
gives the temperature of its surface. This is read at the moment when misting first appears on the
mirror to find the dew point temperature. While dew point apparatus gives a good demonstration
of the effect of condensation by cooling, the visual detection of the initial film of condensate on
the mirror is somewhat subjective. Such instruments are used little in practice.
14.3.5. Wet and dry bulb hygrometers (psychrometers)
These are the most widely used type of hygrometers in subsurface ventilation engineering. They
give reliable results when employed by competent personnel and are simple to use. Because of
their widespread employment, the theory of the wet bulb thermometer is developed in Section
(14.4).
A wet and dry bulb hygrometer is simply a pair of balanced thermometers, one of which has its
bulb shrouded in a water-saturated muslin jacket. Air passing over the two bulbs will cause the
dry bulb thermometer to register the ordinary temperature of the air. However, the cooling effect
of evaporation will result in the wet bulb thermometer registering a lower temperature. Knowledge
of the wet and dry bulb temperatures, together with the barometric pressure, allow all other
psychrometric parameters to be calculated.
Instruments vary in (a) the precision of the thermometers, (b) the manner in which water is
supplied to the wet bulb and (c) the means of providing the required airflow over the thermometer
bulbs. The term "psychrometer", rather than "hygrometer", tends to be used for the more accurate
instruments.
The accuracy of the thermometers depends upon the quality of their manufacture and calibration,
and also upon their length. The better psychrometers can be read to the nearest 0.1C.
Variations considerably greater than this may occur over the cross section of an airway,
particularly near shaft stations. In order to maintain the instrument at a portable size, some
models achieve accuracy at the expense of range. The approximate extremes of temperature
should be known prior to conducting an important psychrometric survey, and an appropriate
instrument selected. Replacement thermometers should be purchased as balanced pairs. The
difference between the wet and dry bulb temperatures is more important than the absolute
temperature. Significant errors in both the wet and dry bulb temperature readings may occur by
14 - 12
radiation from surrounding surfaces or between the bulbs. This can be reduced by increasing the
air velocity over the thermometer bulbs. However, for precise results a polished radiation shield
should be provided separately around each of the bulbs.
Many hygrometers are fitted with a small reservoir of water into which dips one end of the muslin
wick. Capillary action draws water through the muslin to feed the wet bulb. In such devices, it is
important that the water reaching the wet bulb is already at wet bulb temperature. The water
within the reservoir will be at dry bulb temperature and, if supplied to the wet bulb at too liberal a
rate, will give a falsely elevated wet bulb temperature. Conversely, if the water supply is
insufficient then drying out of the wick will occur and, again, the wet bulb thermometer will read
too high. Because of these difficulties, the more precise psychrometers have no integrated water
reservoir and must have their wet bulbs wetted manually with distilled water.
If a wet bulb thermometer is located in a wind tunnel and air of fixed psychrometric condition
passed over it at increasing velocity, the indicated wet bulb temperature will decrease initially,
then level off. The reason for this phenomenon is that the envelope of saturated air surrounding
the wet bulb must be removed efficiently to maintain the evaporative process. The steady state
reading observed is essentially identical to the true thermodynamic wet bulb temperature attained
during an adiabatic saturation process (see section 14.5.2). The minimum air velocity which gives
the true wet bulb temperature is a function of the size and shape of the wet bulb, and its
orientation with respect to the direction of the airflow. For most commercially available
instruments, an air velocity of at least 3 m/s over the bulbs is suggested.
The static hygrometer (sometimes known as the Mason hygrometer) has no intrinsic means of
creating the required air velocity and relies on being hung in a location where the thermometer
bulbs are exposed to an external air velocity. The "whirling" or "sling" psychrometer has its
thermometers and water reservoir mounted on a frame which may be rotated manually about its
handle. Whirling the instrument at about 200 rpm will give the required air velocity over the bulbs.
After 30 seconds of rotation, the whirling should be terminated (but not by clamping one's hands
around the bulbs), the instrument held such that the observer is not breathing on it, and the wet
bulb temperature read immediately. The process of whirling should be repeated until the reading
becomes constant.
Aspirated psychrometers have small fans driven by clockwork or batteries. Air is drawn through
the radiation shield surrounding each bulb. Again, the wet bulb temperature should be observed
until it becomes constant. While the whirling hygrometers are, by far, the most widely used
instruments for routine measurements, aspirated psychrometers are recommended for important
surveys.
14.4. THEORY OF THE WET BULB THERMOMETER

The wet bulb temperature is a most important parameter in hot climatic conditions for two
separate but inter-related purposes. The first lies in its vital importance in evaluating the ability of
the air to remove metabolic heat from personnel. The second is the use of the wet bulb
temperature in quantifying the humidity of the air.
14 - 13
14.4.1. Heat balance on a wet bulb

Figure 14.1 illustrates (1 + X) kg of moist
unsaturated air approaching and flowing
closely over the surface of a wet bulb. On
leaving the wet surface, the 1 kg of air remains
the same but the mass of associated water
vapour has increased from X kg to Xs kg and
the air has become saturated. Thus a mass of
(Xs -X) kg of water has evaporated from the
wet surface for each kilogram of "dry" air
passing. The reason for employing 1 kg of dry
air as the basis of measurement is, again,
apparent - this remains constant, while the
mass of water vapour varies.
saturated
tw
(1+Xs) kg
unsaturated
td
(1+X) kg
latent
heat
Figure 14.1 Heat balance on a wet bulb.
As the molecules of water leave the wet

surface they take energy with them. This
"latent heat" energy is transferred to the air,
leaving the wet bulb at a reduced energy level.
This is reflected by a drop in temperature (the
evaporative cooling effect) and the wet bulb
thermometer gives a reading depressed below
that of the dry bulb. The greater the rate of
evaporation, the greater the wet bulb
depression.
As there is now a difference between the air
temperature and the cooler wet surface, a
transfer of sensible heat will occur by
convection from the air to the wet bulb. A
dynamic equilibrium will be established at
which the heat loss from the wet bulb by
evaporation is balanced by the convective
sensible heat gain:
Latent heat loss from wet bulb = Sensible heat gain by wet bulb
(14.28)
Each of these terms can be quantified. Latent heat transfer from the wet bulb to the air
q
L (Xs X)
J per kg dry air
(14.29)
where L = latent heat of evaporation (J/kg evaporated) at wet bulb temperature.

The air/vapour mixture approaches the wet surface at dry bulb temperature, td, and leaves at wet
bulb temperature, tw. The sensible heat transfer from the air to the wet bulb is then
q
= mass x specific heat x change in temperature

= (1 + X)
Cpm
(td - tw)
J/kg dry air
(14.30)
At equilibrium, latent heat transfer equals sensible heat transfer. Equations (14.29) and (14.30)
quantify the heat balance that exists on the wet bulb
L (Xs X)
(1 + X) Cpm (td - tw)
J/kg dry air
(14.31)
14 - 14
14.4.2. Determination of moisture content and vapour pressure from psychrometer readings
If the barometric pressure, P, and the wet and dry bulb temperatures, tw and td, are known then
the moisture content and, indeed, all other psychrometric parameters can be determined.
Commencing from the heat balance for the wet bulb, and using the shorthand notation t for the
wet bulb depression, td - tw, we have
L (Xs - X)
(1 + X) C pm t
But
Cpm =
then
C pa + X C pv
(1 + X )
LXs = (Cpa + X Cpv) t + LX
= C pa t + X ( C pv t + L)
giving the moisture content, X, as

X
LX s C pa t
kg/kg dry air
C pv t + L
(14.32)
All variables on the right side of this equation are known, or can be calculated easily for the given
P, td and tw: Hence, the moisture content, X, is defined.
The actual vapour pressure then follows from a transposition of equation (14.4)
e
PX
0.622 + X
Pa
(14.33)
Example
Determine the moisture content and vapour pressure for a barometric pressure of 100 kPa and
wet and dry bulb temperatures of 20 and 30C respectively.
Solution
At the wet bulb temperature, the variables required by equation (14.32) are calculated as follows:
L
=
=
esw =
Xs
(2502.5 - 2.386 x 20) 1000

2454.78 x 103 J/kg
or 2454.78 kJ/kg
17.27 20
610.6 exp
(equation (14.10)) =
237.3 + 20
= 0.622
2.3375
(100 2.3375 )
= 0.014887
2.3375 kPa
(equation (14.4))
kg/kg dry air
14 - 15

X
2454.78 10 3 0.014887 1005 (30 20)

1884(30 20) + 2454.78 10
= 0.01071
kg/kg dry air
36544 10 050
18 840 + 245 4780
The actual vapour pressure is then given by equation (14.33)

e
100 0.01071
0.622 + 0.01071
1.6927 kPa
This example illustrates that the term involving Cpv (1884 J/kg K) in equation (14.32) is very small
compared with L. A more traditional equation for actual vapour pressure follows from ignoring the
smaller term. Then equation (14.32) simplifies to
X
= X s
C pa t
L
Substituting for X and Xs from equation (14.4) gives

0.622
or
C pa t
e sw
e
= 0.622
L
(P e )
(P e sw )
= e sw
C pa t
(P e )
(P e )
(P e sw ) 0.622L
If we now assume that e and esw are small compared with P, the equation simplifies further to
e
= e sw
or
C pa
0.622L
e sw APt
Pt
Pa
(14.34)
Pa
(14.35)
The parameter A is known as the psychrometric "constant". However, because of the

simplifications made in the derivation, it is not a true constant; neither is it given precisely as
Cpa / (0.622L) but varies non-linearly with temperature and pressure between 0.00058 and
0.000648 C
over the atmospheric range. This variation of the psychrometric "constant" is a
weakness in the relationship. However, a value of 0.000644 C

most practical purposes.
gives acceptable results for
Example
Determine the actual vapour pressure from equation (14.35) for P = 100 kPa, tw = 20 C
and td = 30 C.
Solution
These are the same conditions as specified in the previous example. This gave
esw = 2.3375
kPa
e = 2.3375 - 0.000644 x 100 x (30 - 20)
= 1.6935 kPa
(a difference of less than 0.05 per cent from the earlier solution).
14 - 16
14.5. FURTHER PSYCHROMETRIC RELATIONSHIPS

14.5.1. Enthalpy of moist air
In Chapter 3, the steady flow energy equation was derived as
1
u12 u 22
+ (Z1 Z 2 )g + W in = VdP + F12 = (H 2 H 1 ) q in
2
J/kg
(14.36)
where
u
Z
g
Win
V
F
H
qin
And
=
=
=
=
=
=
=
=
air velocity, m/s

height above datum, m
gravitational acceleration, m/s2
mechanical (fan) work input, J/kg
specific volume, m3/kg
frictional conversion of mechanical to heat energy, J/kg
enthalpy, J/kg
heat input from external sources, J/kg
In its application to mine ventilation thermodynamics, (Chapter 8) the change in enthalpy between
end stations 1 and 2 was earlier assumed to be
H2 H 1
Cpa (t2 t1)
J/kg
(14.37)
where t1 and t2 were the dry bulb temperatures. This assumption was based on the premise that
the airway was dry - neither evaporation nor condensation was considered. For moist, but
unsaturated air, the specific heat term may be replaced by Cpm for the actual air/vapour mixture.
However, if evaporation or condensation does take place then the exchange of latent heat will
have a large additional effect on the dry bulb temperature. Equation (14.37) no longer applies. As
this is the situation in most underground airways, we must seek a method of evaluating enthalpy
that takes the moisture content of the air into consideration.
Let us carry out another imaginary experiment. Suppose we have 1 kg of dry air at 0C and, in a
separate container, a small amount, X kg, of liquid water. Now let us heat the air until it reaches a
dry bulb temperature of td. Similarly, we add heat to the water until it vaporizes and continue
heating until the vapour also reaches a dry bulb temperature of td. Finally, we mix the two
components to obtain (1 + X) kg of moist unsaturated air at a dry bulb temperature of td and some
lower value of wet bulb temperature, tw.
The total amount of heat that we have added to the air and water during this experiment
represents the increase in enthalpy of the system over its starting condition at 0C. If we choose
0C as our enthalpy datum, that same added heat represents the enthalpy level of the final
(1 + X) kg of mixture.
In order to quantify the added heat, consider the air and water separately:
(i)
The heat required to raise the temperature of 1 kg of dry air from 0C to td is simply
mass x specific heat x change in temperature
i.e.
qair
1 x
Cpa
td
14 - 17
(ii) We can convert the water at 0C to vapour at td in any of three ways, depending upon the
pressure that we maintain in the vessel:
(a) We could consider evaporating the X kg at water at 0C then raising the temperature
of the vapour to t d .
qwater
where L(o)
L(o) X + X Cpv td
Latent heat of evaporation at 0C.
(b) Alternatively, we might raise the temperature of the liquid water to td then evaporate at
that temperature.
qwater
where Cw
and
L(td)
=
=
X Cw td + L(td) X
specific heat of liquid water (4187 J/kg K)

Latent heat of evaporation at td C
or
(c) We could raise the temperature of the liquid to any intermediate value, t, evaporate at
that temperature, then continue to superheat the vapour until it reaches the required
temperature, td.
qwater
where L(t)
X Cw t + L(t) X + X Cpv (td-t)
Latent heat of evaporation at t C
(14.38)
In fact, as the two end conditions are precisely defined, it does not matter which method is used.
They all give the same result provided that account is taken of the variations of latent heat and
specific heat with respect to temperature.
For a reason that will become apparent later, let us choose method (c) with the intermediate
value of temperature chosen to be tw, the eventual wet bulb temperature of the air/vapour
mixture.
We have now found the total amount of heat added to the system and, hence, its enthalpy
relative to a 0C datum
H
qair + qwater
Cpa td + X [Cw tw+ L + Cpv (td tw)]
J per kg dry air
(14.39)
Here, the symbol L has reverted to its earlier meaning of latent heat of evaporation at wet bulb
temperature.
All of the parameters on the right side of the equation can either be measured, are constant or
can be calculated from equations (14.6) and (14.32). Hence, the enthalpy of an air/vapour mixture
can be determined from psychrometric measurements of air pressure and wet and dry bulb
temperatures. If such values of enthalpy are employed in the steady-flow energy equation (14.36)
then that equation may continue to be used for airways that involve evaporation or condensation
processes.
14 - 18
14.5.2. The adiabatic saturation process

Consider a long level airway with no heat additions from any source and free standing water
covering the floor (Figure 14.2). Unsaturated air enters at one end and moves sufficiently slowly
to allow full saturation to occur. The air exits at saturation conditions.
no heat additions
C O N S T A N T
unsaturated
S I G M A
H E A T
saturated
latent heat
sensible heat
water
1.2
temperature
1
0.8
dry bulb
0.6
0.4
constant wet bulb
0.2
0
10
30
50
70
90
110
130
150
Distance
Figure 14.2 An adiabatic saturation process.
The surface of the liquid water reacts in the same way as that of a wet bulb thermometer (Section
14.4.1) and, indeed, will be at wet bulb temperature. At inlet, the dry bulb temperature will be at a
higher value. Hence, sensible heat transfer will occur from the air to the water resulting in a
reduction in dry bulb temperature. Simultaneously, heat and mass transfer from the water to the
air will take place as water molecules escape from the liquid surface - a latent heat gain by the
air. These exchanges will continue until saturation when wet bulb, dry bulb and water temperature
all become equal (thermodynamic wet bulb temperature).
This is known as an adiabatic saturation process. The sensible heat lost by the air is balanced by
the latent heat gained by the air and the process involves no net addition or loss of heat. This
latter statement is true for the combination of air and water, but not quite true for the air alone.
Mass has been added to the airstream in the form of water molecules, and those molecules
already contained sensible heat before they were evaporated. Hence, the enthalpy of the air does
not quite remain constant. However, if we were to move along the airway taking psychrometric
readings, calculating enthalpy from equation (14.40) but omitting the term for the sensible heat of
liquid water, X Cw tw that has evaporated then the result would be a property value that remained
truly constant throughout the adiabatic saturation process. The significance of this property
seems first to have been recognized by Carrier and was named Sigma Heat, S, to distinguish it
from its near neighbor, enthalpy.
14 - 19
14.5.3. Sigma Heat, S

Sigma heat is much more than simply an interesting property of an adiabatic saturation process.
Indeed, it features in the majority of analyses concerning subsurface climatic changes.
Ignoring the sensible heat of the liquid water in equation (14.40) gives
S
H X C w tw
Cpa td + X [L +Cpv (td - tw)]
and
J/kg dry air
(14.40)
J/kg dry air
(14.41)
But from equation (14.32), and replacing t with (td - tw)

X [L + Cpv(td tw)]
LXs Cpa (td tw)
Notice that the left side of this relationship appears in equation (14.42). Substituting for
X [L + C pv (t d t w )] in that equation
gives
S
Cpa td + LXs - Cpa (td - tw)
LXs +Cpa tw
or
J/kg dry air
(14.42)
The td term cancels to produce a really neat equation for sigma heat. Now we are about to
discover a phenomenon of major significance in psychrometric processes. Recalling that Xs is the
saturation moisture content dependent only on wet bulb temperature and pressure (equations
(14.4) and (14.10)), and that L is the latent heat of evaporation, also at wet bulb temperature, it
follows that for any given barometric pressure, sigma heat, S, is a function of wet bulb
temperature only. Furthermore, as sigma heat remains constant during an adiabatic saturation
process so must also the wet bulb temperature remain constant. This is why the wet bulb
temperature appears as a horizontal line in Figure 14.2.
Using the concept of sigma heat, thermal additions or losses from an airstream can readily be
quantified from psychrometric observations. Furthermore, the behaviour of the wet bulb
temperature, a directly measurable parameter, is an immediate indication of heat transfer from
the strata, machines, coolers, potential energy or any other source. Contrasting this with dry bulb
temperature which varies during evaporation or condensation, we see the fundamental
importance of the wet bulb temperature in psychrometric processes.
14 - 20
14.6. SUMMARY OF PSYCHROMETRIC EQUATIONS

This is a convenient point to make a reference list of the more important equations that have
been derived in the previous sections, and to re-order them in the sequence that they are
required for most psychrometric calculations where P, tw and td are the measured variables. All
temperatures are expressed in degrees Centigrade and pressures in Pascals.
esw
Xs
17.27 t w
610.6 exp
237.3 + t w
e sw
= 0.622
(P e sw )
Pa
(14.43)
kg/kg dry air
(14.44)
Lw
(2502.5 - 2.386 tw) 1000
J/kg
(14.45)
Lw Xs + 1005 tw
J/kg dry air
(14.46)
[Lw
kg/kg dry air
(14.47)
kg/kg dry air
(14.48)
S 1005 t d
+ 1884 (t d t w )]
or
Lw X s 1005 (t d t w )
[Lw + 1884 (t d t w )]
PX
(0.622 + X )
(apparent)
(actual)
(rh) =
Pa (when P is in Pascals)
(P e )
+ 273.15)]
[287.04(t d
(P 0.378e )
[287.04(t d + 273.15)]
S + (4187 tw X)
e
x 100
e sd
(14.49)
kg dry air/m3
(14.50)
kg moist air/m3
(14.51)
J/kg dry air
(14.52)
percent
(14.53)
(Equation (14.44) is used for esd with tw replaced by td). Note that in this list of important
relationships, the need for the old troublesome psychrometric "constant" (Equation (14.35)) has
been entirely eliminated. This summary of equations can be programmed into spreadsheet
software or a programmable calculator to provide a rapid and accurate means of determining
values of psychrometric parameters from given barometric pressure and wet and dry bulb
temperatures.
Example
At entry to a level continuous underground airway of constant cross-section, psychrometer
readings give P1 = 110.130 kPa, tw1 = 23 C and td1 = 28 C. The corresponding readings at exit
are P2 = 109.850 kPa, tw2 = 26 C and td2 = 32C. If the volume flow of air at inlet is 25 m3/s,
calculate the heat and moisture added to the airstream during its passage through the airway.
Solution
The following table was produced in a few seconds from a spreadsheet programmed with the
psychrometric equations:
14 - 21

quation
Parameter
measured
Inlet
Outlet
Units
110.13
109.85
kPa
measured
tw
23
26
measured
td
28
32
(14.43)
esw
2808.5
3360.3
Pa
(14.10) using td
esd
3778.7
4753.2
Pa
(14.44)
Xs
0.016277
(14.45)
Lw
(14.46)
0.019628
kg/kg dry air
2440.5 x 10
J/kg water
74.032 x 10
J/kg dry air
2447.6 x 10
62.955 x 10
(14.47)
0.01417
0.017078
kg/kg dry air
(14.49)
2453.0
2935.0
Pa
(14.50)
(app)
1.2457
1.2206
kg dry air/m3
(14.51)
(act)
1.2633
1.2415
kg moist air/m3
(14.52)
64 320.1
75 889.43
J/kg dry air
(14.53)
rh
64.91599
61.8
percent
As the airflow was measured as 25 m3/s at inlet conditions, the mass flow of dry air
M
=
=
Q (inlet, apparent)
m 3 kg dryair
25 x 1.2457
s
m3
Increase in sigma heat along airway

S2 S1 =
74.032 - 62.955 = 11.077
Heat added
qin
M (S2 S1)
31.14 x 11.077
344.9
kJ/kg dry air
kg dry air
kJ
s
kg dry air
kJ/s or kW
Increase in moisture content along airway

X2 X1 = 0.017078 - 0.01417
Rate of evaporation =
=
= 31.14 kg dry air/s
0.002908 kg/kg dry air
M (X2 X1)
31.14 x 0.002908 = 0.0906
kg/s or litres/s
14.7. DEVIATIONS FROM CLASSICAL THEORY

14.7.1. Fogged air
The classical theory of condensation makes the premise that condensation cannot commence
until saturation conditions are reached. This is not quite in agreement with observable
phenomena. The formation of fog is likely to commence before 100 per cent relative humidity is
reached although the process of condensation accelerates rapidly as saturation is approached.
Smoke fogs (smogs) may occur over cities at relative humidities of less than 90 per cent.
14 - 22
The physical mechanism of condensation is complex. Condensation can occur only in the
presence of hygroscopic and microscopic nuclei on which the process can commence. It has
been estimated (Brunt) that there are between 2 000 and 50 000 hygroscopic nuclei in each cubic
centimeter of the atmosphere, having radii of 10-6 to 10-5 cm. The most productive sources of
these hygroscopic particles are the oceans and sulphurous smokes from burning hydrocarbon
fuels.
The natural formation of fog would appear to be a continuous process of condensation
commencing at comparatively low relative humidities when the more strongly hygroscopic nuclei
begin to attract water. This can often result in a noticeable haze in the atmosphere. As the
relative humidity approaches 100 per cent, the rate of condensation on all hygroscopic nuclei
increases; the droplets of water so formed increase rapidly in size and the haze develops into fog.
Fogging in subsurface ventilation systems occurs in two situations. First, when the strata are
cooler than the dew point temperature of the incoming air and, secondly, due to decompressive
cooling of humid return air. Hence, fogging in ascending return airways and, especially, upcast
shafts is not uncommon.
The formation of fogs in underground structures is undesirable for a number of reasons. The
reduction in visibility may produce a safety hazard, particularly where moving vehicles are
involved. High humidity can result in physical and chemical reactions between the airborne water
and hygroscopic minerals within the strata, and may produce falls of roof and spalling of the sides
of airways. Such problems can reach serious proportions in certain shales and evaporite mines.
Another difficulty that may arise in salt mines is absorbance of water vapour by the mined rock. If
the air is humid then the material may become "sticky" and difficult to handle. The problem may
be alleviated by transporting the ore in the return airways. One further problem of fogging in
upcast shafts is that it creates very wet conditions throughout the shaft, headgear and/or exhaust
fans.
If the air velocity in an upcast shaft lies within the range 7 to 12 m/s then droplets will tend to
remain in suspension, encouraging water blankets to form. The resulting increase in shaft
resistance causes a fluctuating load to be imposed on main fans, even to the extent of stalling
them. The resulting sudden reduction in airflow results in water cascading to the bottom of the
shaft and the whole process will be repeated in a cyclic manner. In extreme cases this may lead
to failure of the fan blades.
If it becomes necessary to dehumidify the air because of any of these problems then there are
two possibilities. The less expensive option, where it can be used, is to divert the air through old
stopes or workings that contain water-absorbent minerals in the broken strata. Alternatively the
air temperature may be reduced to below dew point in air coolers and the condensed water
removed from the system. All, or part of the reject heat produced by the refrigeration plant may be
returned to the downstream airflow to control its temperature. Chemical dehumidification is
seldom practicable in mining circumstances because of the large volumes of air involved.
14.7.2. Imperfect gas behaviour
The equations given for saturation vapour pressure ((14.10) and (14.11)) took deviations from
perfect gas behaviour into account. However, the other relationships derived in this chapter
assume that dry air, water vapour, and air/vapour mixtures all obey the perfect gas laws. These
laws assume that the molecules of the air are perfectly elastic spheres of zero volume and that
they exhibit no attractive or repulsive forces.
For the ranges of pressures and temperatures encountered in atmospheric psychrometrics,
assumption of the perfect gas laws gives results of acceptable accuracy. Where additional
14 - 23
precision is required, Hemp (1982) has suggested the following two corrections based on the
earlier work of Goff and Gratch.
Equation (14.3) for moisture content
X =
Ra
Ra
e
1.0048e
becomes X =
R v (P e )
Rv (P 1.0048e )
kg/kg dry air
(14.55)
and gas constants, R, are corrected for pressure and temperature as follows:
R(corrected) = R {1 [ (5.307x10-9 P + 9.49x10-6) (8.115x10-11 P + 2.794x10-6) td ] }
(14.56)
where
P is in Pa and td in C.
This gives a maximum change in R of 0.06 percent over the ranges 80 to 120 kPa and 0 to 60C.
At the mid range value of 100 kPa and 30C, the difference is 0.02 percent.
14.8. PSYCHROMETRIC CHARTS

Prior to the advent of electronic computation, employment of psychrometric equations was an
onerous process. In order to minimize the calculations, graphical depictions of psychrometric
processes were devised. These are known as Psychrometric Charts. Although the use of such
charts has largely been supplanted by programming psychrometric equations into spreadsheet
software and programmable calculators they still provide a powerful means of visualizing
psychrometric processes.
Most psychrometric charts show a representation of the psychrometric parameters for a fixed
barometric pressure. Figure 14.3 gives an example of a 100 kPa chart. This is one of a series of
such charts ranging from 80 to 130 kPa at intervals of 2.5 kPa produced by A.W.T, Barenbrug
(1974). As pressure is the least sensitive of the independent variables, interpolation between
charts is unnecessary for most practicable purposes. More sophisticated types of chart have
been produced that allow for pressure variations (Whillier, 1971). However, these seem not to
have found widespread use.
Returning to Figure 14.3 it is clear that knowledge of any two of the variables will fix a location on
the chart, indicating the climatic condition of the air, and enabling other psychrometric variables to
be read directly or determined with a minimum of calculation. The usual situation is one in which
the wet and dry bulb temperatures are known and can be plotted on the relevant chart. Engineers
who are adept at using psychrometric charts are better able to visualize variations in climatic
conditions.
Figure 14.4 illustrates the ease with which processes can be followed. Movement along a
constant wet bulb line (i.e. constant sigma heat) indicates adiabatic conditions. Any movement to
a higher or lower wet bulb temperature indicates that the air has been heated or cooled
respectively. A path to a higher moisture content (labelled as apparent specific humidity on Figure
14.3) indicates evaporation while movement along a constant moisture line denotes sensible
heating or cooling. Similarly, pure latent heating or cooling is indicated by a constant dry bulb
temperature.
14 - 24
Figure 14.3. Example of a psychrometric chart, at 100kPa.

(after Barenburg A.W.T. Psychrometry and Psychrometric Charts(1) )
Reproduced by courtesy of the Chamber of Mines of South Africa
14 - 25
Figure 14.4 Process lines on a psychrometric chart
Example
Air passes through a cooler at a mean pressure of 100 kPa. It enters at wet and dry bulb
temperatures of 30 and 40 C respectively, and exits at 25 C saturated. If the volume flow at
inlet, Q, is 20 m3/s, use the psychrometric chart to determine the rates of heat exchange and
production of condensate.
Solution
On the 100 kPa psychrometric chart, plot the two points 30/40 C and 25/25 C (Figure 14.5). The
latter lies on the saturation line (100 per cent relative humidity). It is immediately obvious that
during the flow of air from inlet to outlet both the moisture content (specific humidity) and sigma
heat have decreased. The following values are read from the chart. The corresponding values
calculated from the equations in section 14.6 are given in parentheses for comparison).
Inlet
Outlet
30/40 C
25/25 C
X g/kg dry air
23.2 (23.24)
20.3 (20.34)
S kJ/kg dry air
97.0 (97.1)
74.7 (74.8)
0.932 (0.9324)
0.884 (0.8838)
3.6 (3.601)
3.16 (3.167)
tw/td C
V (app) m /kg dry air

eact
kPa
14 - 26
S = 97
kJ/kg dry air
V(app) = 0.932
m3/kg dry air
e = 3.6 kPa
X = 23.2
g/kg dry air
Figure 14.5 Process line for the example. Wet and dry bulb temperatures at inlet and outlet
are plotted. The inlet point is used to illustrate how to read other psychrometric parameters.
14 - 27
Mass flow of dry air, M
Heat exchange
1
=
0.932
Q (app)
(S1 - S2) M
= (97 - 74.7) 21.5
Rate of condensation
= 0.0623
(X1 - X2) M
20 x
21.5 kg dry air/s

kJ kg
kg s
= 479.5 kW
= (0.0232 - 0.0203) 21.5
kg water/s or 0.0623 x 60 = 3.74 litres per minute.
References
Barenbrug, A. W. T. (1974). Psychrometry and Psychrometric Charts 3rd Edition, Chamber of
Mines of South Africa.
Brunt, D. (1939). Physical and Dynamical Meteorology. Cambridge Press.
Goff, J. A. and Gratch, S. (1945). Thermodynamic Properties of Moist Air. ASHVE Trans. Vol.
51, pp. 125-158.
Hemp, R. (1982). Psychrometry. Chapter 18. Environmental Engineering in South African Mines.
Mine Ventilation Society of South Africa pp. 435-463.
Van der Walt, N. T. and Hemp, R. (1982). Thermometry and Temperature Measurements.
Chapter 17. Environmental Engineering in South African Mines. Mine Ventilation Society of South
Africa, pp. 413-433.
Whillier, A. (1971). Psychrometric charts for all barometric pressures. J. Mine Ventilation Society
of South Africa, Vol. 24, pp. 138-143.
14 - 28
APPENDIX A14
Derivation of the Clausius-Clapeyron equation
When the enthalpy, or overall energy, of a system is increased by doing mechanical work on it,
then this work represents available energy all of which can theoretically be utilized. On the other
hand, if the enthalpy increase is brought about by the addition of heat then the Second Law of
Thermodynamics insists that only a part of this heat is available energy. The unavailable fraction
is utilized in increasing the internal energy of the system. Thus if we have 1 kg of liquid water of
volume V, internal energy U and pressure es and add an increment of heat energy the
corresponding increase in enthalpy will be
dH
dU + es dV
(ref. equation (3.19))
(14.A1)
Now a change of phase (evaporation from liquid to vapour) occurs at constant temperature and
constant pressure. During such a process the enthalpy change is also related to the increase in
entropy of the system
dH
where s
=
=
Tds
(see equation (3.46) with dP = 0)
(14.A2)
(14.A3)
entropy
Hence (14.A1) and (14.A2) give

dH
dU + es dV
Tds
If the 1 kg of water is completely evaporated, then the total heat added to accomplish this is the
latent heat of evaporation, L.
Equation (14.A3) can be integrated over the change of phase from liquid water (subscript L) to
vapour (subscript s)
L
H S - HL
Us -UL + es (Vs - VL)
Us + es VS Tss
UL + es VL - TsL
or
T (ss - sL)
(14.A4)
(14.A5)
This relationship defines a function of state, d, known as the thermodynamic potential which
remains constant during any isobaric - isothermal change of phase:
=
U + es V - Ts
(14.A6)
Now consider the situation at a slightly different temperature T+dT and pressure es+ des. The
thermodynamic potential that will apply under these conditions will be + d. Differentiating
equation (14.A6) to find d gives
d
dU + esdV + Vdes - Tds sdT
(14.A7)
but from equation (14.A3)

dU + esdV - Tds
Hence
d
Vdes - sdT
(14.A8)
14 - 29
As d is constant during evaporation at conditions T and es, and d+ d, remains constant

during evaporation at conditions T + dT and es + des, it follows that d remains constant as T
approaches T + dT and es approaches es + des. Therefore from equation (14.A8)
VLdes -sLdT
Vsdes - ssdT
(14.A9)
or
de s
dT
sL ss
VL V s
(14.A10)
This equation was first derived by Clapeyron in 1832.

The volume of liquid water, VL, is negligible compared with that of the vapour, VS, (At normal
atmospheric conditions VS /VL is of the order of 2500). Furthermore, from equation (14.A4)
- (sL ss)
L
T
and (14.A10) becomes

de s
L 1
=
dT
T Vs
(14.A11)
but from the general gas law for 1 kg of water vapour

VS
Rv T
es
Hence,
1 de s
e s dT
L
Rv T 2
the Clausius-Clapeyron equation
(14.A12)
14 - 30
Heat flow into subsurface openings
CHAPTER 15
HEAT FLOW INTO SUBSURFACE OPENINGS
15.1. INTRODUCTION .........................................................................................2

15.2. STRATA HEAT ............................................................................................2
15.2.1. Methods of determining strata heat load ......................................................................... 2
15.2.2. Qualitative observations................................................................................................... 4
15.2.3. Fourier's law of heat conduction ...................................................................................... 5
15.2.4. Geothermic gradient, geothermal step and thermal conductivity .................................... 6
15.2.5. An analysis of three dimensional heat conduction. ......................................................... 7
15.2.6. Solution of the radial heat conduction equation............................................................. 12
15.2.7. Heat transfer coefficient for airways .............................................................................. 14
15.2.7.1. Convective heat transfer ......................................................................................... 15
15.2.7.2. Radiative heat transfer ............................................................................................ 16
15.2.8. Summary of procedure for calculating heat flux at dry surfaces.................................... 19
15.2.9. Heat transfer at wet surfaces......................................................................................... 22
15.2.10. In-situ measurement of rock thermal conductivity ....................................................... 27
15.3. OTHER SOURCES OF HEAT ...................................................................29

15.3.1. Autocompression ........................................................................................................... 29
15.3.2. Mechanized equipment.................................................................................................. 32
15.3.2.1. Electrical equipment ................................................................................................ 34
15.3.2.2. Diesel equipment..................................................................................................... 35
15.3.2.3. Compressed air ....................................................................................................... 36
15.3.3. Fissure water and channel flow ..................................................................................... 38
15.3.4. Oxidation ........................................................................................................................ 41
15.3.5. Explosives...................................................................................................................... 42
15.3.6. Falling rock..................................................................................................................... 42
15.3.7. Fragmented rock............................................................................................................ 43
15.3.8. Metabolic heat................................................................................................................ 44
BIBLIOGRAPHY ................................................................................................45
APPENDIX A15.1 ...............................................................................................46
Analytical solution of the three dimensional transient heat conduction equation (15.13) as
obtained by Carslaw and Jaeger (1956)............................................................................. 46
APPENDIX A15.2 ...............................................................................................46

Gibsons algorithm for the determination of dimensionless temperature gradient, G. ....... 46
APPENDIX A15.3. ..............................................................................................47

Background to equations for the heat transfer coefficient, h. ............................................. 47
APPENDIX A15.4 ...............................................................................................53

Derivation of the equation for latent heat of evaporation at a wet surface. ........................ 53
15 - 1
15.1. INTRODUCTION
Heat is emitted into subsurface ventilation systems from a variety of sources. In the majority of
the world's coal mines, the airstream itself is sufficient to remove the heat that is produced. In
deep metal mines, heat is usually the dominant environmental problem and may necessitate the
use of large-scale refrigeration plant (Chapter 18). Conversely, in cold climates, the intake air
may require artificial heating in order to create conditions that are tolerable for both personnel and
equipment.
In Section 9.3.4, quantification of the heat emitted into a mine or section of a mine was required in
order to assess the airflow needed to remove that heat. Hence, a sensible place to commence
the study of heat flow into mine openings is to classify, analyze and attempt to quantify the
various sources of heat. The three major heat sources in mines are the conversion of potential
energy to thermal energy as air falls through downcasting shafts or slopes (autocompression),
machinery and geothermal heat from the strata. The latter is, by far, the most complex to analyze
in a quantitative manner. We shall deal with this separately in Section 15.2, then quantify other
sources of heat in Section 15.3.
15.2. STRATA HEAT

15.2.1. Methods of determining strata heat load
In reviewing the literature for means of determining the amount of heat that will be emitted from
the strata, the ventilation engineer is faced with a bewildering array of methods varying from the
completely empirical, through analytical and numerical, to computer simulation techniques. The
basic difficulty is the large number of variables, often interacting with each other, which govern
the flow of strata heat into mine airways. These include:
the length and geometry of the opening

depth below surface and inclination of the airway
method of mining
wetness of the airway surfaces
roughness of the airway surfaces
rate of mineral production or rock breaking
time elapsed since the airway was driven
volume flow of air
barometric pressure, and wet and dry bulb temperatures
virgin (natural) rock temperature
distance of the workings from downcast shafts or slopes
geothermic step or geothermic gradient
thermal properties of the rock
other sources of heating or cooling such as machines and cooling plant.
With such a variety of parameters, it is hardly surprising that traditional methods of predicting
strata heat loads have been empirical. Perhaps the simplest and most common of these has
been to quote strata heat flux in terms of heat load per unit rate of mineral production; for
example, kW per tonne per day. This can provide a useful and rapid guideline provided that all
conditions are similar to those on which the value was based. However, as rate of production is
only one of the several variables listed above, it is obvious that this technique can lead to gross
errors if it is applied where the value of any one of those variables is significantly different from
the original sets of measurements used to establish the kW/tonne/day value.
The more sophisticated empirical techniques extend their range of application by incorporating
estimated corrections for depth, distance, age, inlet conditions or, indeed, any of the listed
variables considered to be of local importance.
15 - 2
The purely analytical methods of quantifying heat flow from the strata are somewhat limited for
direct practical application because of the complexity of the equations that describe threedimensional, time-transient heat conduction. Indeed, they can be downright frightening. However,
the theory that has evolved from analytical research has provided the basis for numerical
modelling which, in turn, has resulted in the development of pragmatic computer simulation
packages for detailed prediction of variations in the mine climate.
A hybrid method has grown out of experience in running climatic simulation packages. It is often
the case that, for particular conditions, some of the input variables have a very limited effect on
the results. By ignoring those weaker parameters it is then sometimes possible to develop simple
equations that give an approximation of the heat flow.
In view of these alternative methodologies, what is the mine environmental engineer to do when
faced with the practical problems of system design? Experience gained from major planning
projects has indicated the following recommended guidelines:
1. If the objective is to plan the further development of an existing mine, or if there are
neighbouring mines working similar deposits at equivalent depths and employing the same
methods of working, then the empirical approach (kW per tonne per day) may be adopted for
the overall strata heat load on the whole mine or major sections of the mine. This
presupposes the existence of data that allow acceptable empirical relationships to be
established and verified.
Employing past and relevant experience in this way provides a valuable and simple means of
arriving at an approximate heat load which, when combined with the methodology of Section
9.3.4, will give an indication whether the heat can be removed by the airflow alone, or if
refrigeration is required.
However, let the user beware. If the proposed mine project deviates in any significant manner
from the conditions in which the empirical data were compiled (check the list of variables at
the beginning of this section), then the results may be misleading. In particular, great caution
should be exercised when employing empirical relationships established in other
geographical regions. A phrase commonly heard at mine ventilation conferences is "what
works there, doesn't work here."
2. The hybrid equations are very useful for rapid approximations of heat flow into specified types
of openings. Dr. Austin Whillier of the Chamber of Mines of South Africa produced many
hybrid equations for easy manual application, including:
(a)
Radial heat flow into established tunnels

q = 3.35 L k 0.854 ( VRT d )
where
q =
L
k
VRT
d
=
=
=
=
(15.1)
heat flux from strata (W)

length of tunnel (m)
thermal conductivity of rock (W/mC)
virgin (natural) rock temperature (C)
mean dry bulb temperature (C)
Throughout this chapter we shall use the symbol k for thermal conductivity and for temperature.
15 - 3
(b)
Advancing end of a heading

q = 6 k (L + 4 DFA ) ( VRT d )
W
where
L
= length of the advancing end of the heading (m). This should be not greater than
the length advanced in the last month. Equation (15.1) may be used for the
older sections of the heading. .
DFA = daily face advance (m).
(c)
One dimensional heat flow towards planar surfaces

Assuming good convective or evaporative cooling of the surfaces,
q =
where
A
=
=
C
=
t
=
A (k C ) 0.5 ( VRT d )
t 0. 5
(15.3)
area of surface (m2)

rock density (kg/m3)
specific heat of rock ( J/(kgC) )and
time since the surface was exposed (s)
A weakness of these equations is that they each contain the mean dry bulb temperature, d.
As this is initially unknown, it must be estimated. If, when the resulting value of q is used to
determine the temperature rise, the initial estimate of d is found to have been significantly in
error then the process may be repeated. However, as the hybrid equations promise nothing
more than rough approximations there may be little point in progressing beyond a single
iteration.
3. For accurate and detailed planning, a mine climate simulation package should be employed.
These are computer programs that have been developed to take all of the relevant variables
into account. They may be used for single airways or combined into a total underground
layout. Climate simulation programs go beyond the calculation of heat load by predicting the
effects of that heat on the psychrometric conditions in the mine. The principles of a climate
simulation program are outlined in Chapter 16.
For major projects, estimates of heat loads may be based on empirical and hybrid
methodologies for initial conceptual layouts, progressing to simulation techniques for detailed
planning.
15.2.2. Qualitative observations
Before embarking upon a quantitative analysis of strata heat flow, it will be useful to introduce
some of the observable phenomena in a purely qualitative manner.
First, when cool air passes through a level airway, its temperature usually increases. This is
caused by natural geothermal heat being conducted through the rock towards the airway, then
passing through the boundary layers that exist in the air close to the rock surface. In working
areas, the newly exposed rock surfaces are often perceptibly warmer than the air. However,
those surfaces cool with time until they may be only a fraction of a degree C higher than the
temperature of the air.
If the airway is wet then the increase in dry bulb temperature is less noticeable. Indeed, that
temperature may even fall. This is a result of the cooling effect of evaporation. Heat may still
emanate from the strata but all, or much of it, is utilized in exciting water molecules to the extent
15 - 4
that they leave the liquid phase and form water vapour. The heat content and wet bulb
temperature of the air/vapour mixture then rise because of the internal energy of the added water
vapour.
Another observation that can be made in practice is that although the air temperature in main
intake arteries rises and falls in sympathy with the surface climate, the temperatures in main
returns remain remarkably constant throughout the year. This is because cool air will encourage
heat to flow from the rock. However, as the temperature of the air approaches the natural
temperature of the rock such heat transfer will diminish. It can, of course, work in reverse. For
example, the temperature of the air leaving an intensively mechanized working area may be
greater than the local strata temperature. In that case, heat will pass from the air to the rock. The
air will cool and, again, approach equilibrium when its temperature equals that of the strata. A
mine is an excellent thermostat.
The envelope of rock immediately surrounding a newly driven airway will cool fairly rapidly at first.
There will, accordingly, be a relatively high rate of initial heat release into the air. This will decline
with time. A well-established return airway may have reached near thermal equilibrium with the
surrounding strata. However, the linings and envelope of rock around downcast shafts or main
intakes will emit heat during the night when the incoming air is cool and, conversely, absorb heat
during the day if the air temperature becomes greater than that of the surrounding envelope of
rock. This cyclic phenomenon, sometimes known as the thermal flywheel" (Stroh, 1979) is
superimposed upon longer term cooling of the larger mass of rock around the opening.
The boundary layers that exist within the airflow close to the rock surface act as insulating layers
and, hence, tend to inhibit heat transfer between the rock and the main airstream. It follows that
any thinning or disturbance of those boundary layers will increase the rate at which heat transfer
takes place. This can occur either through a rise in air velocity or because of a greater degree of
roughness on the surface (Section 2.3.6).
Finally, although there may be significant rises in the air temperature along intake airways, the
most noticeable increase usually occurs in the mineral winning areas. This is because, first, the
newly exposed and warm surfaces of both the solid and broken rock give up their heat readily
and, secondly, the mechanized equipment that may be concentrated in stopes or working faces.
Having introduced these concepts in a purely subjective manner, we now have a better intuitive
understanding from which to progress into a quantified and analytical approach.
15.2.3. Fourier's law of heat conduction
+d
q
x
dx
Figure 15.1 Linear heat flow.
When a steady heat flux, q, passes through a

slab of homogeneous material, the temperature
will fall from 1 at entry to 2 at exit (Figure 15.1).
Planes of constant temperature, or isotherms, will
exist within the material. Figure 15.1 also shows
two isotherms at a short distance, dx, apart and
with temperatures and
+ d.
The heat flux, q, is proportional to both the
orthogonal area, A, through which the heat
travels and the temperature difference, d,
between isotherms. It is also inversely
proportional to the distance, dx, between those
isotherms.
15 - 5
Hence,
q is proportional to A
where q
A
=
=
=
=
d
dx
heat flux (W)

area through which q passes (m2),
temperature (C) and
distance (m)
The negative sign is necessary since reduces in the direction of heat flow, i.e. d is negative.
To convert this relationship into an equation, a constant of proportionality, k, is introduced, giving
q = k A
d
dx
(Fourier's Law)
(15.4)
k is termed the thermal conductivity of the material and has units of W/(mC).
To be precise, k is a slowly changing function of temperature. It may also vary with the
mechanical stress applied to the material. In the strata around mine openings, the effective
thermal conductivity of the strata can be significantly different from those given by samples of the
rock when measured in a laboratory test (Section 15.2.10). The reasons for such differences
include natural or induced fractures in the strata, variations in mineralogy that may be direction
dependent, movements of groundwater, radioactive decay and local geothermal anomalies.
15.2.4. Geothermic gradient, geothermal step and thermal conductivity
The crustal plates upon which the continents drift over geological time are relatively thin
compared to the diameter of the earth. Furthermore, it is only in the upper skin of those plates
that mining takes place at the present time. The geothermal flow of heat emanating from the
earth's core and passing through that skin has an average value of 0.05 to 0.06 W/m2. It can, of
course, be much higher in regions of anomalous geothermal activity
In Fourier's law, equation (15.4), if we use the value of 0.06 W for each square metre of land
surface to give the variation of temperature, , with respect to depth D (= -x), then
d
0.06
=
dD
k
C
m
(15.5)
The increase of strata temperature with respect to depth is known as the geothermic gradient. In
practical utilization, it is often inverted to give integer values and is then referred to as the
Geothermal step = dD/d
m/C
(15.6)
It is clear from equations (15.5) and (15.6) that as we progress downwards through a succession
of strata, the geothermal gradient and geothermal step will vary according to the thermal
conductivity of the local material.
Table 15.1 has been assembled from several sources as a guide to the thermal conductivities
and corresponding geothermal steps that may be expected for a range of rock types. However, it
should be remembered that these parameters are subject to significant local variations. Site
specific values should be obtained, preferably from in-situ tests, for any important planning work.
15 - 6
Rock type
Copper orebody (Montana)
Copper orebody (Arizona)
Carboniferous
Clays
Limestone
Sandstone
Dolerite
Quartzite
Potash (low grade)
(high grade)
Halite (low grade)
(high grade)
Thermal conductivity
W/(mC)
0.8 to 1.1
1.3
1.2 to 3.0
1.8
3.3
2.0 to 3.6
2.0
4.0 to 7.0
3.5 to 5.0
5.0 to 7.0
1.5 to 4.0
4.0 to 6.0
Geothermal step
m/C
13 to 18
22
20 to 50
30
55
30 to 60
33
65 to 120
60 to 80
80 to 120
25 to 70
70 to 100
Table 15.1. Typical values of thermal conductivity and geothermal step for a range of rock types.
For air and water in the range 0 to 60C (Hemp, 1985)
ka
= 2.2438 10 4 T 0.8353
W/(m o C)
kw
= 0.2083 + 1.335 10 3 T
W/(m o C)
where T = Absolute temperature (K)

15.2.5. An analysis of three dimensional heat conduction.
During the practical application of empirical, hybrid or simulation techniques of assessing strata
heat loads (Section 15.2.1) the mine environmental engineer need give little conscious thought to
the theory of heat conduction or the derivations given in this Section. These analyses are
included here because they provide the essential relationships for the development and internal
operation of numerical models and, hence, mine climate simulation packages. Readers who are
not concerned with the analytical treatment should skip directly to Section 15.3.
Within the envelope of rock that surrounds an underground airway, the temperature varies with
both position and time. Our first task is to derive a general relationship that defines the
temperature of the rock as a function of location and time. For simplicity, we shall assume that the
thermal properties of the rock remain constant with respect to both position and time.
The general equations for heat conduction are readily derived in Cartesian co-ordinates.
However, for mine airways, it is more practical to work in cylindrical polar co-ordinates. Figure
15.2 represents a mass of strata surrounding an underground airway. The z axis represents an
airway simply as a line. The position of any point in the strata can be defined by co-ordinates z, r
and where r is the radial distance from the centreline and is the angle from the horizontal
measured in radians.
Consider a small trapezoidal element lying in a thin annulus of rock at a distance r from the centre
line. The length of the element is dz, and its radial height is dr. The inner width is r d increasing
to (r + dr)d at its outer limit.
15 - 7
qr
qz
qr
Figure 15.2 Three dimensional heat flow.

We shall analyse the heat flux passing through the element shown shaded. This can be
accomplished by examining, in turn, the partial heat flows in the directions of increasing
r, and z, and in that order.
15 - 8
1. The base of the element has an area r d dz, the curvature being negligible over such short
distances. The flux passing through this face is given by Fouriers Equation (15.4)
W
r
The opposite face has an area of (r + dr ) d dz and passes a heat flow of
dq1,r = k r d dz

dr
+
r
r
Heat retained by the element in the r direction is

dq2,r = k (r + dr ) d dz
r 2
2
+
=
dr
+
k
dr
d
dz
dr
k
dr
d
dz
2 +
r
r r 2
r
r 2
as the term involving (dr)2 is insignificant.
2
dq1,r dq 2,r = k r d dz
(15.7)
2. The heat flux in the direction of increasing passes through opposite faces, each of area
dr dz. The direction is actually r. Hence, Fouriers equation for heat entering the element is
dq1, = k dr dz
= k dr dz
(r )
r
while the heat leaving the element is

dq 2, = k dr dz
2
r d = k dr dz
d
+
+
2
r
r
r
Heat gain in the r direction is

dq 1, dq 2,
= k dr dz d
r 2
(15.8)
3. Repeating the exercise for the z direction, the opposite faces are trapeziums of height dr and
opposite edges of length r d and (r + dr)d, the curvature of the lines being insignificant
over such small distances. The latter have an average length of
r d + (r + dr ) d
2
The product of the differentials dr d is negligible compared with r d, giving the width to be
r d and face area of r d dr.
The heat entering the element in the z direction is
z
and leaving the opposite face,
dq 1,z
= k r d dr
dq 2,z
= k r d dr
dz
+
z
z
15 - 9
Then the rate of heat accumulated in the z direction is

dq 1,z dq 2,z
= k r d dr dz
(15.9)
z 2
The total rate of heat gain by the element, dq, is given by summing equations (15.7,15.8 and 15.9)
2
2
1 2
dq = k dr d dz r 2 +
+
+
r
r
r 2
r
z 2
(15.10)
The heat gain by the element in time t (where t = time in seconds) can also be expressed as
W
t
where m = mass of element and C = specific heat of the material.
dq = mC
But m = volume x density () = dr dz rd

Then
d
dq = dr dz rd C
t
(15.11)
Hence, from Equations (15.10 and 15.11)

r 2
1 2
2
k 2 +
r
= r C
+
+
2
2
r
r
t
W
m2
(15.12)
or
k
C
2
2
1
1 2
+
+ 2
2 +
=
2
2
r r
t
r
r
z
C
s
This fundamental relationship is the general three-dimensional equation for unsteady heat
conduction expressed in cylindrical polar co-ordinates.
For most purposes in strata heat conduction towards airways, we can assume that
2
z 2
= 0
and that
2
2
= 0
These simplifications are based on the premise that the natural geothermic gradient is small
compared to the radial variation in temperature around the incremental length of airway, and that
the heat conduction is radial along the full length.
15 - 10
Then
k
C
2
1
2 +
r r
r
=
t
C
s
The term k/C is a constant for the material and is called the thermal diffusivity, (m2/s) giving
2
1
+
2
r r
r
=
t
C
s
(15.13)
This is the form of the equation normally quoted for radial heat conduction and is the basis on
which strata heat flow is determined in mine climate simulation programs. Figure 15.3 gives a
graphical illustration of the time transient heat conduction equation.
60
1 day
10 days
50 days
55
100 days
1 year
rock temperature (deg. C)
50
5 years
45
10 years
40
35
30
25
20
0
10
distance into rock
12
14
16
18
(m)
Figure 15.3 An illustration of the variation of rock temperature with respect to

time and distance into the rock.
15 - 11
15.2.6. Solution of the radial heat conduction equation

Having derived equation (15.13) for the time-space variation in temperature around an
underground airway, we must now attempt to transform it into a practical procedure to determine
strata heat emission into the airway.
In order to facilitate further analysis and generality, it is convenient to express radial distances
and time as dimensionless numbers.
Dimensionless radius rd =
r
ra
where ra = effective radius of the airway (= perimeter / 2)

The actual shape of the cross section has little effect on the influx of strata heat into airways.
Dimensionless time Fo =
ra
(this is called the Fourier number)
Equation (15.13) then becomes

2
( r d r a )
1
1
=
r d r a ( r d r a )
(Fo r a 2 / )
As ra and are constants

2
2
r a ( r d )
1
rd ra
=
2
( r d )
r a / (Fo )
The ra2 and terms cancel, giving the radial heat conduction equation as
2
( r d )
1
=
r d ( r d )
(Fo )
(15.14)
The determination of heat flux into the airway commences by applying Fourier's law (equation
(15.4)) to each square metre of rock surface :

q = k
r s
W
m2
(15.15)

where
is the temperature gradient in the rock but at the rock/air interface (subscript s for
r s
surface).
This same heat flux, q, passes from each square metre of surface through the boundary layers
into the main airstream. However, for any given type of surface and flow conditions, the heat
transferred through the boundary layers is proportional to the temperature difference across those
layers,
i.e.
q = h ( s d )
W/m2
(15.16)
15 - 12
where s =
d =
h =
temperature of the rock surface (C)

dry bulb temperature in the main airstream (C) and
a heat transfer coefficient ( W/(m2C) ) that is a function mainly of the air velocity
and the characteristics of the rock surface.
We shall discuss the heat transfer coefficient further in Section 15.2.7. However, for the moment
let us accept it simply as a constant of proportionality between q and ( s d ) .
Assuming that we know the values of h and d, then we only need the rock surface temperature,
s, for equation (15.16) to give us the required heat flux, q. The problem turns to one of finding s.
As the heat flux from the strata must be the same as that passing through the rock/air interface,
we can combine equations (15.15) and (15.16) to give

q = k
r s
= h ( s d )
(15.17)
m2
Again, for generality, the temperature gradient at the surface may be expressed in dimensionless
form, G, where
G =
ra

( VRT d ) r s
(15.18)
where VRT = virgin (natural) rock temperature (C)

Combining with equation (15.17) gives
G ( VRT d ) =
h ra
( s d )
k
(15.19)
The group h ra/k is known as the Biot Number, B, or dimensionless heat transfer coefficient.
Then
or
G ( VRT d ) = B ( s d )
s =
G
(VRT d ) + d
B
C
C
(15.20)
We have now found an expression for s. However, it contains the dimensionless but yet
unknown temperature gradient G. This may be obtained from a solution of the general radial heat
conduction equation (15.13) using Laplace transforms. The mathematics of the solution process
can be found in Carslaw and Jaeger (1956). The result involves a series of Bessel functions and
is reproduced in Appendix A15.1 at the end of this chapter. Unfortunately, the solution appears
even more disconcerting for practical use than the original differential equation (15.13). However,
it is now more amenable to numerical integration. The results are shown on Figure 15.4 as a
series of curves from which the dimensionless temperature gradient, G, can be read for given
ranges of Fourier Number, Fo , and Biot number, B.
An algorithm produced by Gibson (1975) allows G to be determined much more easily than the
full numerical integration of the Carslaw and Jaeger solution. Gibson's algorithm (Appendix
A15.2) is suitable for programming into a personal computer and gives an accuracy of within 2
per cent over the majority of the ranges covered in Figure 15.4.
15 - 13
Older solutions to the general equation (15.13) (Carrier, 1940; Goch and Patterson, 1940)
assumed that the rock surface temperature was equal to the dry bulb temperature of the air. This
ignored the insulating effect of the boundary layers or, put another way, inferred an infinite heat
transfer coefficient and, hence, an infinite Biot Number. The uppermost curve on Figure 15.4
represents this bounding condition and gives the same results as the tables produced by Carrier,
and Goch and Patterson.
Having established a value of G, equations (15.17) and (15.20) then combine to give the required
heat flux:
W
G
q = h (VRT d )
(15.21)
B
m2
1
B = infinity
0.9
10
10
0.8
Dimensionles temperature gradient G
= Biot Number,
0.7
2
0.6
0.5
0.8
0.4
0.6
0.5
0.3
0.4
0.2
0.2
0.1
0.1
10
100
1000
Fourier Number Fo
Figure 15.4 These curves enable the dimensionless temperature gradient, G, in the rock at its surface to
be estimated for known Fourier and Biot Numbers.
15.2.7. Heat transfer coefficient for airways
In section 15.2.6, we introduced the heat transfer coefficient, h, for the rock surface as the
constant' of proportionality between heat flux across a boundary layer and the corresponding
temperature difference between the rock surface and the general airstream:
q = h ( s d )
W/m2
(15.16)
15 - 14
It is clear from this equation that for a well established airway where the rock surface
temperature, s, is very close to the air dry bulb temperature, d, the strata heat flux q will be
small. The value of the heat transfer coefficient will then have little effect on climatic variations in
the airway. Conversely, for newly exposed surfaces ( s d ) will be relatively high and the heat
transfer coefficient will be a significant factor in controlling the flow of heat from the strata.
It can be seen from Figure 15.4 that for Biot Numbers (B = hra/k) of 10 or more, errors in the heat
transfer coefficient and, hence, Biot Number will have little effect on the dimensionless
temperature gradient, G, which in turn controls heat flux into the airway. For typical sizes of
modern mine airways, the Biot Number is generally in excess of 5 where its accuracy (and,
hence, the accuracy of the heat transfer coefficient) remains of limited consequence. It is,
nevertheless, prudent to establish relationships that will allow us to evaluate the heat transfer
coefficient.
The heat transfer coefficient is a constant only within the confines of equation (15.16). The overall
heat transfer coefficient, h, for mine airways is made up of two parts, the convective heat transfer
coefficient, hc , and a component of the radiative heat transfer coefficient, hr.
15.2.7.1. Convective heat transfer
In practice the convective heat transfer coefficient changes with those factors that cause
variations in the thickness of the boundary layers, i.e. the air velocity and roughness on the
surface that produce near-wall turbulence. Hence, the convective heat transfer coefficient
depends upon the coefficient of friction, f, (or Atkinson friction factor) and the Reynolds Number,
Re. These parameters were introduced in Section 2.3 as factors that influence airway resistance
and losses of mechanical energy.
A rise in near-wall turbulence causes not only increased cross-flow momentum transfer close to
the laminar sublayer but also helps transport heat across the turbulent boundary layer into the
mainstream. There is a close analogy between heat and momentum transfer across boundary
layers. This has engaged the attention of researchers since the time of Osborne Reynolds and
many equations relating heat transfer coefficients to fluid properties and flow regimes have been
proposed for both hydraulically smooth and rough surfaces.
A relationship that has shown itself to agree with practical observations both underground
(Mousset-Jones et al, 1987; Danko et al, 1988) and in scale models (Deen, 1988) is based on
earlier work by Nunner (1956). For application in mine ventilation, it may be stated as
Nu =
0.35 f Re
1 + 1.592 15.217 f Re 0.2 1 / Re 0.125
dimensionless
(15.22)
The convective heat transfer coefficient, hc, is then given as

hc =
where
and
Nu k a
d
W/(mC)
(15.23)
ka = thermal conductivity of air (2.2348 x 10-4 T 0.8353 W/(mC) ), T being the

absolute temperature in degrees Kelvin)
d = hydraulic mean diameter (m)
The Nusselt Number (Nu = hcd/ka) is simply a means of expressing the heat transfer coefficient as
a dimensionless number in order to extend the generality of equation (15.22). It is analogous to the
Biot Number but is referred to the air and the characteristics of the airway rather than the rock.
15 - 15
100000
f
0.06
0.03
0.02
0.015
0.01
Nusselt Number Nu
0.005
10000
1000
100
10000
100000
1000000
10000000
Reynolds Number Re
Figure 15.5 Variation of Nusselt Number, Nu, with coefficient of friction, f, and Reynolds
Number, Re. (Based on air at 20 C)
Here again, equations (15.22) and (15.23) may be programmed into a calculator or personal
computer for a rapid determination of the convective heat transfer coefficient. For manual
estimation, the Nusselt Number may be read directly from Figure 15.5 for any given values of f
and Re. The background to equation (15.22) is given in Appendix A15.3.
A further phenomenon that has been observed experimentally and subjected to theoretical
analyses is a tendency for the convective heat transfer coefficient to decrease as the wall
temperature rises for any given value of Reynolds Number. This is caused by a thickening of the
temperature or thermal boundary layer and, hence, giving an enhanced insulating effect against
heat transfer between the wall and the main airstream. However, the actual heat transfer will
normally increase because of the greater value of ( s d ) in equation (15.16).
15.2.7.2. Radiative heat transfer
In addition to convective effects, heat may be lost from a rock surface by radiation. Oscillations in
atomic energy levels at the surface produce electromagnetic radiation. This propagates through
space until it reaches another solid surface where some of it is reflected, the remainder returning
15 - 16
to thermal energy by excitation of atoms on the receiving surface. Additionally, some gases,
including water vapour and carbon dioxide, will absorb a fraction of the thermal radiation causing
a rise in temperature of the gas. Elemental gases such as oxygen and nitrogen, the major
constituents of air, are not affected in this way.
According to the Stefan-Boltzmann equation, the net heat transfer by radiation from a hotter to a
cooler surface is given as
qr
= 5.67 10 8 (T1 T 2 ) A Fev

4
(15.24)
where T1 and T2 are the absolute temperatures of the hotter and cooler surfaces respectively (K)
A = the smaller of the two surfaces (m)
and
5.67 x 10-8 = the Stefan-Boltzmann constant. (Wm-2 K-4)
(15.25)
The parameter Fev, combines the thermal emissivity of the surfaces and the view factor which
quantifies the degree to which the surfaces can see each other (Whillier, 1982).
In subsurface environmental engineering, a more practical relationship for radiant heat transfer (in
Watts per square meter) from airway surfaces is
qr
W/m2
= hr ( s d ) Fev
(15.26)
where hr = radiative heat transfer coefficient [W/(m2C)]

The emissivity of polished metal surfaces is fairly low. However, for rough natural surfaces it is
usually more than 0.95. Furthermore, the view factor of each unit area of surface to water vapour
in the passing airstream, or to much of the nearby rock surface, also approaches unity. Hence,
we may approximate Fev to 1.
The radiative heat transfer coefficient varies from 5 to 7 W/(m2C) in the range of surface
temperatures 10 to 40 C. More precisely, it may be estimated as
hr
= 4 5.67 10 8 Tav
W/(m2C)
(15.27)
where Tav = Average absolute temperature of the two surfaces (K) In practice, this may be taken
as the dry bulb temperature of the air.
For a dry airway, the temperature of the rock surface will remain the same around any given
perimeter. Hence, T1 = T2 in equation (15.24) and there will be no net transfer of radiant heat
between rock surfaces. If part of the surface is wet, however, radiant heat will pass from the dry
areas to the cooler wet surface. This will cause a slight increase in strata heat transfer from dry
areas. However, the small rise in temperature of the wet surfaces will result in a slightly
diminished flow of strata heat to those surfaces. Although the two may not balance, the net effect
is small. Radiative heat transfer between rock surfaces is usually ignored.
The amount of thermal radiation absorbed by water vapour in the air varies exponentially with the
product of the vapour content and the distance travelled by the radiation through the air/vapour
mixture. At a vapour content of 0.019 kg/kg dry air, 22 per cent of the radiant heat will be
absorbed within 3 metres and 47 per cent within 30 metres. The remaining radiation that is not
absorbed will be received on other rock surfaces and, again, will have little impact on the mine
climate.
The radiative heat transfer coefficient, hr, is usually considerably lower than the convective heat
transfer coefficient, hc, in mine airways and may, indeed, be of the same order as the uncertainty
15 - 17
in hc. An estimate of the fraction ab of thermal radiation absorbed by water vapour in the air may
be obtained from
ab
= 0.104 ln (147 X L )
(15.28)
where In = natural logarithm

X = water vapour content of air (kg/kg dry air) (Section 14.2.2.)
and
L = distance travelled by the radiation through the air (path length, m)
This equation is based on curve fitting to empirical data.
The mean path of the electromagnetic waves propagating from a given point on a rock surface
and before they strike another rock surface will depend upon the geometry of the mine opening.
For a mine airway a typical mean path length may be of the order of three times the hydraulic
mean diameter.
The effective radiant heat transfer coefficient for thermal radiation between the rock and the air
becomes ab hr.
The overall heat transfer coefficient is then
W/(m2 C)
h = hc + a b h r
Analyses can also be carried out for the absorption of thermal radiation by carbon dioxide and
dust particles. However, under normal conditions the corresponding rates of thermal absorption
are very small.
Example.
The dry bulb temperature of the air in a 4m by 3m underground opening is 26 C. The
corresponding moisture content is 0.015 kg/kg dry air. Estimate the effective radiant heat transfer
coefficient.
Solution.
Hydraulic mean diameter, d h =
4 Area
4 12
=
= 3.43
perimeter
14
Assume the mean path length of the radiation to be

3 dh = 3 x 3.43 = 10.3 m
Fraction of radiation absorbed (from equation (15.28))
ab
= 0.104 ln (147 0.015 10.3) = 0.32

hr
= 4 5.67 10 8 (273.15 + 26 ) 3
= 6. 1
W/(m2 C)
Then the effective radiant heat transfer coefficient is

a b hr
= 0.32 6.1 = 1.95
W/(m2 C)
15 - 18
15.2.8. Summary of procedure for calculating heat flux at dry surfaces

Sections 15.2.5 to 15.2.7 have detailed the derivation of relationships that describe the radial flow
of strata heat through the rock towards a mine opening and across a dry surface into the main
airstream. As often occurs in engineering, the application of those relationships is straightforward
compared with the theoretical analyses that have produced them.
Before moving on to consider heat exchange at wet surfaces, it is convenient to summarize the
procedure for calculating the emission of strata heat across a dry surface and to illustrate that
procedure by a case study.
Calculation procedure for dry surface
1. Assemble the data:
airway dimensions (m) .
coefficient of friction, f (= Atkinson friction factor/0.6)
age of airway, t (seconds) .
airflow Q (m3/s) .
mean dry bulb temperature of air, d (C)
wet bulb temperature, w (C )
barometric pressure, P (Pa) .
rock thermal properties
thermal conductivity, kr (W/mC)
density, r (kg/m3)
specific heat, Cr (J/kg C)
kr
(m2/s)
diffusivity, r =
r Cr
virgin rock temperature, VRT (C)
2. Determine derived parameters:
cross sectional area, A (m2)
perimeter of airway, per (m)
hydraulic mean diameter, dh = 4A/per (m)
effective radius, ra = per /(2) (m)
Reynolds' Number, Re. For the purposes of this procedure, Re may be calculated from the
approximation Re = 268 000 Q /per
moisture content of air, X (kg/kg dry air) (from Section 14.6)
mean radiation path length, L (m)
3. Determine the Nusselt Number, Nu, either from Figure 15.5 or from equation (15.22): i.e.
Nu =
0.35 f Re
1 + 1.592 15.217 f Re 0.2 1 / Re 0.125
4. Determine the overall heat transfer coefficient, h:

(a) Convective heat transfer coefficient, hc
Nu
W/(m2C)
dh
where 0.026 W/(mC) = thermal conductivity of air
hc = 0.026
15 - 19
(b)
Effective radiative heat transfer coefficient, ab hr

hr = 22.68 x 10-8 (273.15 + d)3
W/(m2C)
Absorption fraction ab = 0.104 ln(147 X L)

Effective radiative heat transfer coefficient = ab hr
(c)
Overall heat transfer coefficient,

(W/m2C)
h = hc + ab hr
5 Calculate Biot Number, B:

B =
h ra
kr
(dimensionless)
6 . Calculate Fourier Number, Fo:

Fo =
r t
ra
(dimensionless)
7. Determine dimensionless temperature gradient in the rock but at the surface, G, either from
Figure 15.4 or from Gibson's algorithm (Appendix A15.2).
8 Determine heat flux, q:
q = h
G
(VRT d )
B
(W/m2)
9 . Calculate heat emission into airway:

q per length of airway
1000
(kW)
Case Study
This case study illustrates not only the calculation procedure but also typical magnitudes of the
variables. The purpose of the exercise is to determine the strata heat that will flow into an
incremental length of dry airway. The stages of calculation are numbered to coincide with the
steps of the procedure given above.
1. Given data:
airway dimensions
width = 3.5 m, height = 2.5 m, length = 20 m
Atkinson friction factor (at a = 1.2 kg/m3) = 0.014 kg/m3
i.e. coefficient of friction, f = 0.014/0.6 = 0.0233 (dimensionless) .
airway age = 3 months
365
t =
24 3600 = 7.884 10 6 seconds
4
airflow, Q = 30 m3/s .
dry bulb temperature in airway, d = 25 C
from psychrometric data, the moisture content of the air has been determined to be
X = 0.01 kg/kg dry air .
15 - 20
rock thermal properties:

conductivity,
density,
specific heat,
kr = 4.5 (W/mC)
r = 2200 kg/m3
Cr = 950 J/(kgC)
4. 5
r =
= 2.153 10 6
diffusivity,
2200 950
virgin rock temperature, VRT = 42 C
m 2 /s
2. Further derived parameters:

cross-sectional area, A = 3.5.x 2.5 = 8.75 m2
perimeter, per = 2(3.5 + 2.5) = 12 m
4A
4 8.75
hydraulic mean diameter, d h =
=
= 2.917 m
per
12
12
per
effective radius, r a =
=
= 1.910 m
2
2
30
= 670 000
12
mean radiation path length, L, is taken as 3 dh = 3 x 2.917 = 8.751 m
Reynolds' Number, Re = 268 000
3.
The Nusselt Number, Nu, at Re = 670 000 and f = 0.0223 is estimated from Figure 13.5 to be
2400. Alternatively, it may be calculated as
Nu =
0.35 0.0233 670 000
1 + 1.592 15.217 0.0233 670 000 0.2 1 / 670 000 0.125
4. Overall heat transfer coefficient, h

(a) convective heat transfer coefficient, hc,
N
2 433
hc = 0.026 u = 0.026
= 21.69
dh
2.917
= 2 433 (dimensionl ess)
W
m
2 o
(b) effective radiative heat transfer coefficient, ab hr

o
hr = 22.68 10 8 (273 .15 + 25 ) 3 = 6.01 W/(m 2 C)

absorption fraction
a b = 0.104 ln(147 0.01 8.751) = 0.266
a b hr = 0.266 6.01 = 1.60
(c) overall heat transfer coefficient,

o
h = hc + a b hr = 21.69 + 1.60 = 23.29

W/(m 2 C)
This result illustrates the limited effect of thermal radiation.
5.
Biot Number, B
B =
h ra
kr
23.29 1.91
= 9.885
4 .5
(dimensionl ess)
6. Fourier Number, Fo
t
2.153 10 6 7.884 10 6
Fo = r 2 =
(1.91) 2
ra
= 4.653
(dimensionl ess)
15 - 21
7. Dimensionless rock temperature at the surface, G, for B = 9.885 and Fo = 4.653.

This may be read from Figure 15.4 or computed from Gibson's algorithm (Appendix A15.2) as
0.60. It is clear from the graphical method that non-excessive errors in B (and, hence, heat
transfer coefficient) will have a very limited effect on G for the conditions cited in this case study.
8. Heat flux
q = h
G
(VRT d ) = 23.29 0.60 (42 25 ) = 24
B
9.885
W
m2
9. Heat emission into the 20m length of airway

= q x rock surface area W
or
24.0 12 20
= 5.76 kW
1000
15.2.9. Heat transfer at wet surfaces

The majority of underground openings have some degree of water evaporation or condensation
occurring on rock surfaces even if no liquid water is visible. The cooling effect of evaporation or
the heating effect of condensation on the rock surface will then result in a reduction or increase,
respectively, of the wall surface temperature. This, in turn, will modify the strata heat flow toward
that surface.
Figure 15.6 shows the heat balance that must exist for thermal equilibrium at the rock/air
interface.
ws
q = q L + q sen W/m 2
where
q = strata heat flux
qL = latent heat transfer
qsen = sensible heat transfer
Rock
qsen
q
airflow
qL
Figure 15.6 The heat

balance at a wet surface.
(15.29)
(W/m2)
(W/m2)
(W/m2)
In practice, any of these heat flows may be

positive or negative. Indeed, a common
situation occurs when the wet surface
temperature, ws, is less than the dry bulb
temperature of the air, d. In this case, sensible
heat will pass from the air to water on the rock
surface; qL must then accommodate the
combined values of q and qsen.
In order to prevent unnecessary repetition, we
shall assume evaporation in the following
analysis. The same logic applies for
condensation except that qL is negative.
15 - 22
Much of the theory pertaining to heat transfer at dry surfaces applies, also, to wet surfaces. In
particular, let us re-examine equation (15.21) which was derived for a dry surface.
q = hc
G
(VRT d )
B
ref (15.21)
m2
(We shall ignore the small effect of radiative heat transfer. Hence, the overall heat transfer
coefficient, h, becomes the convective heat transfer coefficient, hc).
Consider the functional dependence of each parameter in this equation in turn:
hc depends upon f, Re, ka and d (equations (15.22 and 15.23))
B depends upon hc, ra and kr (equation (15.19))
t
G depends upon the Fourier Number, Fo = r 2 and the Biot Number, B
ra
and VRT, the virgin rock temperature, is constant for any fixed location.
(Subscript r refers to the rock and a to the air or airway).
Notice that none of these parameters depends upon time transient temperatures within the rock
nor whether the surface is wet or dry. They have exactly the same values for both wet and dry
surface conditions. It is particularly significant that the dimensionless temperature gradient in the
rock at the surface, G, is independent of the surface temperature - another illustration of the value
of dimensionless numbers. Hence, the relationships derived earlier for hc, B and G for a dry
surface apply equally well for a wet surface. However, the evaporative cooling at a wet surface
has not been taken into account and this must cause the strata heat flux, q, to increase.
In order to compensate for evaporation while continuing to use equation (15.21) we could select a
value of d reduced sufficiently such that the value of q given by equation (15.21) becomes equal
to the heat flux actually passing through the wet surface. Let us name that reduced temperature
the effective dry bulb temperature, ef.
Then for the wet surface:
q = hc
G
(VRT ef
B
(15.30)
m2
Provided that we replace d by ef then the equations for heat transfer at a dry surface become
applicable also for a wet surface. The problem now is to find the value of ef.
In order to evaluate qL and qsen, we shall require the wet surface temperature, ws. The form of
equation (15.20) gives a relationship between ws and ef.
ws =
G
(VRT ef ) + ef
B
(15.31)
This transposes to
ef
B
= ws VRT
B
B G
(15.32)
15 - 23
Substituting for ef in equation (15.30) and simplifying gives

q = hc
(B G )
(VRT ws )
m2
(15.33)
We now have an equation for the strata heat flux where everything is known except the wet
surface temperature, ws. Let us establish similar equations for qL and qsen, the other two terms in
the heat balance of equation (15.29).
The sensible heat is given simply as
qsen = hc (ws - d)
W/m2
(15.34)
W/m2
(15.35)
The latent heat term, qL, is given by

qL
= 0.0007 hc Lws
=
=
=
=
where Lws
ews
e
P
(ews
e)
'
P
latent heat of evaporation at the wet surface temperature (J/kg)

saturated vapour pressure at the wet surface temperature (Pa)
actual vapour pressure in the mainstream (Pa) and
barometric pressure Pa)
The derivation of equation (15.35) is given in Appendix A15.4. Values of Lws, ews and e may be
determined from the relationships given in Section 14.6. In particular Lws and ews are functions of
ws only while e depends upon the wet and dry bulb temperatures of the mainstream, w and d,
and the barometric pressure, P.
Equations (15.33, 15.34, and 15.35) allow q, qsen and qL to be determined on the basis of known
mainstream conditions and an assumed wet surface temperature, ws. If the assumed value of ws
were correct, then the heat balance
q
= qsen + qL (from equation (15.29)
would hold. However, an error in ws will cause a corresponding error E in the heat balance:
E
q- (qsen + qL)
(15.36)
The wet surface temperature, ws, may now be corrected according to the value of E and the
process repeated iteratively until E becomes negligible. All heat flows are then defined.
Case Study
In order to illustrate the effect of water on heat flow across a rock surface, we shall use the same
data given for the case study in Section 15.2.8 but now assuming that the surface is fully wetted.
To recap on the data:
1. Constants:
Fourier Number,
Convective heat transfer coefficient,
r
Biot Number, B = hc a = 21.69
kr
Fo
hc
= 21.69 (W/m2C)
1.91
4.5
15 - 24
The value of B has changed by some 7 percent as we are now ignoring radiative effects i.e.
using hc instead of the overall heat transfer coefficient, h. However, when we employ Fo and
the modified B on Figure 15.4 or in Gibson's algorithm (Appendix A15.2), the value of
dimensionless temperature gradient G is not significantly different from the 0.6 obtained
previously.
2
Psychrometric Data
The psychrometric relationships used here are listed in Section 14.6. In addition to the dry
bulb temperature of 25C, we shall also require the wet bulb temperature, w, and the
barometric pressure, P, of the mainstream. For this case study we shall take these to be
17.9C and 100 kPa respectively. Additionally, we need to assume a value for the
temperature of the wet surface. As a first estimate, let us assume ws to be the mean of the
wet and dry bulb temperatures in the airstream, i.e. ws = (25.0 + 17.9)/2 = 21.45 C.
(a) For the main airstream:
17.27 17.9
= 2050 .4 Pa
Saturated vapour pressure, e sat ,air = 610.6 exp
237.3 + 17.9
Saturated moisture content, X sat ,air =
0.622 2050 .4
= 0.01302 kg/kg dry air
100 000 2050 .4
Latent heat of evaporation at wet bulb temperature, Lw ,air = (2502.5 2.386 17.9 )10 3
= 2459.79 10 3 J/kg
Sigma heat,
S air = 2459.79 10 3 0.01302 + (1005 17.9 ) = 50 016 J/kg
Actual moisture content
X air =
Actual vapour pressure,
e =
50 016 (1005 25)
2459.79 10 3 + 1884(25 17.9 )
= 0.01006 kg/kg dry air
100 000 0.01006

= 1592 Pa
(0.622 + 0.01006 )
(b) For the wet surface (at temperature ws)

17.27 21.45
= 2 556 Pa
Saturated vapour pressure, ews = 610.6 exp
237.3 + 21.45
Latent heat of evaporation, Lws = (2502.5 2.386 21.45 )10 3 = 2451.3 10 3

3
J/kg
Heat flows:
hc G
(VRT ws ) = 21.69 0.6 (42 21.45 ) = 31.1 W/m2
(9.207 0.6)
B G
Strata heat,
q =
Sensible heat,
q sen = hc ( ws d ) = 21.69(21.45 25 ) = 77.0
W/m2
15 - 25
q L = 0.0007 hc Lws (ews e ) / P
Latent heat,
= 0.0007 21.69 2451.3 10 3
(2556 1592 )
100 000
2
= 358.8 W/m
Correction of ws:
The error in the heat balance is
E = q (q sen + q L ) = 31.1 (77.0 + 358.8 ) = 251
W/m2
Corrected value of ws = 21.45 + (- 251) x 0.01 = 18.94 C

The correction function used here is 0.01 E. The constant 0.01 is chosen to give a reasonable
rate of convergence. While mathematical techniques are available to find optimum correction
functions for given iterative procedures, a more pragmatic approach (assuming computer
availability) is simply to experiment with a set of typical data until satisfactory convergence is
achieved.
The procedure from steps 2b to 4 is repeated iteratively until the absolute value of E becomes
less than 0.01 W/m2. This is clearly a task for a personal computer or programmable calculator.
Programming the equations and logic into an algorithm is a useful student exercise. The results
of running such an algorithm on this particular example are shown on Table 15.2. Balance was
achieved to within 0.01 W/m2 after nine iterations. The first line of the table gives values that are
in sensible agreement with the hand calculations given above.
Iteration
Wet surface
temperature
ws C
Strata heat
Latent heat
Sensible heat
q W/m2
qL W/m2
qsen W/m2
Error
E
W/m2
21.45
31.1
358.6
- 77.0
- 250.49
18.95
34.9
222.7
- 131.4
- 56.43
18.38
35.8
194.5
- 143.6
- 15.11
18.23
36.0
187.1
- 146.9
- 4.19
18.19
36.1
185.0
- 147.8
- 1.17
18.18
36.1
184.4
- 148.0
- 0.33
18.17
36.1
184.3
- 148.1
- 0.09
18.17
36.1
184.2
- 148.1
- 0.03
18.17
36.1
184.2
- 148.1
- 0.01
Table 15.2. Successive values computed during the iterative process for a wet surface.
At balance, 36.1 W of strata heat arrive at each square metre of wet surface. Simultaneously,
148.1 W of sensible heat pass from the air to the wet surface (qsen is negative). Hence, the sum of
the two, 184.2 W/m2 is emitted as latent heat to the air. The temperature of the wet surface is
18.17 C.
The effect of water on the surface has now been quantified. The net heat emission of 36.1 W/m2
compares with 24.0 W/m2 for the corresponding (dry) surface of the earlier case study. For the 20 m
15 - 26
36.1 20 12
= 8.66 kW
1000
compared with 5.76 kW when the surface was dry. However, whereas the 5.76 kW of heat from the
formerly dry surface was sensible heat, the net 8.66 kW emitted from the wet surface now takes the
form of latent heat. This results in significant increases in the sigma heat, wet bulb temperature and
moisture content of the air.
length of airway at a perimeter of 12 m, this translates into a heat load of
This Case Study has illustrated that the determination of heat flux from a wet airway surface involves
lengthy calculations including an iterative procedure. The integrity of those calculations can be
checked through the effective dry bulb temperature, ef , which was introduced as a device to assist
in the analysis. Equation (15.32) gave:
G
B
= ws VRT
B
B G
Using the value of wet surface temperature ws = 18.17C, determined through the iterative
procedure this becomes
ef
ef
0.6
9.207
= 18.17
42
= 16.51 oC
9
.
207
9
.
207
0
.
6
Recall that ef is what the value of the dry bulb temperature in a completely dry airway would
have to be in order to give the same value of heat flux as the actual wet airway. Equation (15.30)
or (15.21) for this hypothetical dry airway give a heat flux of
q = hc
G
(VRT ef
B
= 21.69
0 .6
(42 16.51) = 36
9.207
W
m2
and is in good agreement with the 36.1 W/m2 obtained for the actual wet airway.
It is clear from these case studies that the analytical determination of strata heat flux into
subsurface openings does not lend itself to manual calculation. Chapter 16 discusses simulation
programs which can readily be employed for rapid and detailed predictions of the underground
climate. However, the engineer who uses such programs is utilizing the theory and procedures
that have been described in this Section.
15.2.10. In-situ measurement of rock thermal conductivity
The experimental determination of the thermal conductivity of solids depends essentially upon
Fourier's law (equation (15.4)). A source of heat is applied to one side of a sample. The heat flux
across a known area and the corresponding temperature difference across the sample are
measured leaving the thermal conductivity as the only unknown.
A number of laboratory test rigs have been developed for the determination of the thermal
conductivities of rock samples. However, for subsurface environmental engineering, it is
preferable to measure rock thermal conductivity in-situ. The reason for this is the significant
difference that may be found between an in-situ value and the mean of numerous samples
measured in the laboratory. If the rock is homogeneous and unfractured around mine openings,
and if workings are above the water table then good correlation may be achieved between
laboratory tests and in-situ determinations. However, in the more usual situation of mining
induced fractures in addition to the existence of natural fracture patterns, and particularly in the
presence of groundwater migration, the effective thermal conductivity may be more than double
that indicated by small and unfractured laboratory samples (Mousset-Jones, 1984).
Water has a thermal conductivity of 0.62 W/(mC) which is considerably lower than that of most
rocks. Hence, the presence of still connate water in fissured or porous media will tend to inhibit
15 - 27
heat flow. Unfortunately, water in the strata surrounding subsurface openings is seldom stationary
and, dependent upon the rate of water migration, the transmission of heat by moving water may
be far greater than that resulting from conduction through the rock (see, also, section 15.3.3.).
This can result in considerable increases in the effective thermal conductivity of the strata.
The simplest method of determining rock thermal conductivity in-situ is to use natural geothermal
energy as the heat source. If possible, an airway should be found that is fairly well established,
but sufficiently long to give a readily measured rise in wet bulb temperature of the air over its
length. A thermocouple circuit to indicate the rise in wet bulb temperature over the selected length
of airway will give better accuracy than mercury in glass thermometers. From a cross section
within the chosen length, four boreholes should be drilled radially outward to a depth of at least
10 m. These should each be fitted with strings of thermocouples or other type of temperature
transducers, positioned to indicate the variation in temperature along the hole. Precautions
should be taken to ensure good thermal contact between the transducers and the side of the
borehole, and to prevent convection currents of air or water within the hole.
Returning to Figure 15.2, consider an elongated annulus of rock at radius r and length Y. The
heat flux, q, passes radially through the annulus which has an orthogonal area 2rY and
thickness dr. Fourier's law (equation (15.4)) gives
q = k r 2 rY
(d )
( dr )
(dr is negative as r is measured outward from the airway while q is usually directed toward the
airway).
Then
dr
r
2 k r Y
d
q
If we assume steady state conditions, q is constant for all values of r. We can then integrate to
give
2 k r Y
(15.37)
ln(r ) =
+ constant
q
From the measurements taken in a borehole, can be plotted against In(r) to give a straight line
of slope
q /( 2 k r Y )
(15.38)
Now, values of wet and dry bulb temperature, together with barometric pressures measured at
the two ends a and b of the airway allow the corresponding values of sigma heat, S, to be
established (Section 14.6). Furthermore, anemometer traverses give the airflow, Q, (Section 6.2)
and, hence, the mass flow of air, M
M = Q
kg/s
where = actual density of air (kg/m3 )
Then
q = M (S b S a )
(15.39)
Substituting for q in the expression for the measured slope of the v In(r) graph (15.38) allows kr
to be determined. An average value of rock thermal conductivity will be given by the data
obtained from several boreholes.
Although simple in principle, this in-situ method of measuring effective thermal conductivity of
strata may involve several difficulties. First, the airflow in the airway and the psychrometric
conditions at the inlet should, ideally, remain constant. Accuracy will be improved if all
15 - 28
measurements are made electronically and data logged every few minutes for several days. The
plots of against In(r) are not always straight lines. A sudden change in slope along a given set
of borehole data indicates that the borehole has intersected a change in rock type. A more
gradual curvature suggests that the strata temperatures have not reached equilibrium. Deviations
towards the mouth of the hole indicate non-steady conditions in the airway. In such situations, the
slope of the line at the rock surface should be used in the determination of kr. Differences in the
slopes of the lines are often found between boreholes drilled in different directions from the same
cross-section of the airway. Such deviations may indicate differences in rock type or, often, the
effect of water distribution within the strata.
A major advantage of this method is that it gives the effective thermal conductivity of the complete
envelope of rock around the chosen cross-section. Other techniques have been developed that
produce an artificial source of heat within a borehole with instrumentation to monitor the resulting
temperature field at other locations along the borehole (Danko, 1987). Such equipment can
produce results within a few hours for more limited representative volumes of rock.
15.3. OTHER SOURCES OF HEAT
15.3.1. Autocompression
When air or any other fluid flows downward, some of its potential energy is converted to enthalpy
(H = PV + U, equation (3.19)) producing increases in pressure, internal energy and, hence,
temperature. Actual examples are shown on Figure 8.3.
The rise in temperature as air falls through a downcast shaft or other descentional airway is
independent of any frictional effects (Section 3.4.1). Ignoring the small change in kinetic energy,
the steady flow energy equation (3.22) gives
(H 2 H 1 ) =
(Z1 Z 2 )g
+ q12
(15.40)
where H
=
enthalpy J/kg
Z
=
height above datum (m)
g
=
gravitational acceleration (m/s2) and
q
=
heat added from surroundings (J/kg)
(subscripts 1 and 2 refer to the inlet and outlet ends of the airway respectively).
This equation shows that the increase in enthalpy is, in fact, due to two components,
(a) the heat actually added, q12, and
(b) the conversion of potential energy (Z1 Z2)g.
The heat added, q12, may be positive, negative or zero but the increase in temperature due to
depth alone is definitive for any given dry shaft and moisture content of the air.
The effects of autocompression are virtually independent of airflow. In deep mines, the intake air
leaving the bottoms of downcast shafts may already be at a temperature that necessitates air
cooling. This is the inevitable result of autocompression.
The reverse effect, auto-decompression, occurs in upcast shafts or ascentional airways. This is
usually of less concern as upcast air will have no effect on conditions in the workings. However,
the reduction in temperature due to auto-decompression in upcast shafts can result in
condensation and fogging (Section 9.3.6 ). The mixture of air and water droplets may then reduce
the life of the impellers of exhaust fans sited at the shaft top.
The increase in temperature due to depth is sometimes known as the adiabatic lapse rate. For
completely dry airways, the change in enthalpy is given by equation (3.33)
15 - 29
H 2 H 1 = C pm (T 2 T1 )
where T
and
Cpm
=
=
J/kg
dry bulb temperature (C or K)

specific heat of the actual(moist) air (J/kgC).
(We are reverting to the symbol T for temperature in this section in order that we may use it to
denote absolute temperatures). Hence, equation (15.40) gives the adiabatic (q12 = 0) rise in dry
bulb temperature to be
(T2 T1 )
(Z1 Z 2 ) g
C pm
(15.41)
Substituting for Cpm from equation (14.16) gives
(T2 T1 )
(1 + X )
(Z1 Z 2 )g
(1005 + 1884 X )
where X = mass of water vapour (kg/kg dry air).
=
(15.42)
For a completely dry shaft or slope and at a value of g = 9.81 m/s2, this gives a dry bulb adiabatic
temperature lapse rate varying from 0.976 C per 100 m depth for dry air to 0.952 C per 100 m
depth at a water vapour content of 0.03 kg/kg dry air.
However, in the majority of cases, the shaft or airway surfaces are not completely dry and the
rate of increase in dry bulb temperature is eroded by the cooling effect of evaporation. If the
increase in vapour content of the air due to evaporation is X kg/kg dry air, then this will result in
a conversion of sensible heat to latent heat, L X (J/kg) where L = latent heat of evaporation.
Equation (15.41) then becomes
(T2 T1 )
(Z1 Z 2 )g L X
C pm
(15.43)
Using mid-range values of Cpm = 1010 J/kgK and L = 2453 x 103 J/kg gives the approximation
(T2 T1 )
= 0.00971 (Z1 Z 2 ) 2428.7 X
(15.44)
This equation is shown graphically in Figure 15.7 and is useful only if the increase in vapour
content is known. In practical design exercises, X is normally unknown but an estimate may be
made of the wetness of the shafts. Recourse must then be made to computer simulation
techniques (Chapter 16). It must be emphasized that equations (15.41 to 15.44) refer to the dry
bulb temperature only and while the air remains unsaturated. Should saturation occur then the
dry bulb temperature will follow the wet bulb temperature adiabatic lapse rate.
The behaviour of the wet bulb temperature due to adiabatic compression is more predictable than
dry bulb temperature. The enthalpy (and sigma heat) of the air increase linearly with depth in the
absence of any heat transfer with the surroundings. Furthermore, the increase in air pressure is
also near linear with respect to depth. Hence, as wet bulb temperature varies only with sigma
heat and pressure, the wet bulb adiabatic lapse rate in a shaft will be near constant irrespective of
shaft wetness. Any significant deviation in the wet bulb temperature can be due only to heat
transfer with the walls or shaft fittings.
15 - 30
0.6
0.4
td deg.C per 100 m
Adiabatic rise in dry bulb temperature
0.8
0.2
Rise in moisture content X g/kg dry air per 100 m
0
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
-0.2
-0.4
-0.6
-0.8
-1
Figure 15.7 The dry bulb adiabatic lapse rate (variation of air temperature with depth)
depends upon the rate at which water is evaporated.
Unfortunately, calculation of the wet bulb adiabatic lapse rate is somewhat more involved than for
the dry bulb.
We can write the variation in wet bulb temperature, Tw, with respect to depth, Z, as
dTw
dZ
dTw dP
dP dZ
(15.45)
From the steady flow energy equation for isentropic conditions (F = q = 0) and negligible change
in kinetic energy (equation (3.25)),
g dZ = V dP
or
g
dP
=
dZ
V
N/m3
or
Pa/m
(15.46)
where V = specific volume (m3/kg) = 287.04 Tw /(P - 0.378 es) for saturation conditions (equation
(14.19))
15 - 31
Furthermore, we have already derived the isentropic pressure/temperature relationship for an

air/vapour/liquid water mixture for fans as equation (10.56). The same equation applies to any
other isentropic airflow process involving air/vapour/liquid water mixtures. In order to track the
adiabatic wet bulb temperature lapse rate, we apply the constraint that critical saturation is
maintained throughout the shaft, i.e. X = Xs. Equation (10.56) then becomes
dTw
dP
Lw X s
T
0.286 (1 + 1.6078 X s ) w +
P
287.04 (P e s )
=
2
Lw P X s
1 + 1.7921 X s +
2
3
463.81 10 (P e s ) Tw
C/Pa
(15.47)
In this equation Tw is the absolute wet bulb temperature (K) and barometric pressure, P, is in Pa.
Equations (15.45 to 15.47) allow the wet bulb lapse rate, dTw /dZ to be determined for any given
P and Tw as Xs and es are functions of pressure and temperature only (Section 14.6). Tracking
the behaviour of dTw /dZ for a downcast shaft from equation (15.47) and commencing from known
shaft top pressure and wet bulb temperature shows that the adiabatic lapse rate for wet bulb
temperature is not exactly constant but decreases slowly with depth. For manual application,
Figure 15.8 has been constructed from values calculated for a depth of 500 m below the shaft
top. This gives an accuracy within 4 per cent for a 1000m shaft over the ranges of pressure and
starting temperatures shown1.
It should be recalled that the equations derived for autocompression effects give the changes in
temperature due to depth only. Actual measurements will reflect, also, heat transfer with the shaft
walls or other sources. These are likely to be most noticeable in downcast shafts due to
transients in the surface conditions.
15.3.2. Mechanized equipment
The operation of all mechanized equipment results in one, or both, of two effects; work is done
against gravity and/or heat is produced. A conveyor transporting material up an incline, a shaft
hoist and a pump are examples of equipment that work, primarily, against gravity. Vehicles
operating in level airways, rock breaking machinery, transformers, lights and fans are all devices
that convert an input power, via a useful effect, into heat.
With the exceptions of compressed air motors and devices such as liquid nitrogen engines, all
other forms of power including electricity and chemical fuels produce thermal pollution that must
be removed by the environmental control system.
Increasing utilization and power of mechanization in mines and other subsurface facilities has
resulted in such equipment joining geothermal effects and autocompression as a major source of
heat. With machines consuming about 2 MW of electrical power on some highly mechanized
longwall faces, a number of coal mines in the United Kingdom and Europe had to resort to
refrigeration equipment at depths below surface where it was not previously required. The
calculation of equipment heat is straightforward compared to that for strata heat.
The error depends primarily on the starting wet bulb temperature and is approximately 4 per cent at an
initial wet bulb temperature of 0C. At a starting wet bulb temperature of 20C the error is about 2 per cent.
15 - 32
Change in wet bulb temperature deg C per 100m depth
0.7
0.65
0.6
0.55
0.5
0.45
Shaft top
pressure
0.4
130 kPa
120 kPa
110 kPa
100 kPa
90 kPa
0.35
80 kPa
0.3
0
10
15
20
25
30
Shaft top wet bulb temperature deg. C

Figure 15.8 The wet bulb adiabatic lapse rate depends upon the initial pressure and
wet bulb temperature but is independent of evaporation within the shaft.
The following algorithm gives a good approximation to these curves:
Enter shaft top wet bulb temperature (tw) C
Enter shaft top barometric pressure (P) kPa
a = -7.6286E-07*P + 1.7593E-04
b = -1.2554E-07*P^2 + 2.238E-05*P + 1.1012E-02
c = -9.4107E-06*P^2 + 3.8054E-03*P + 0.33327
Wet bulb temp. adiabatic lapse rate = a*tw^2 - b*tw + c
15 - 33
35
15.3.2.1. Electrical equipment

Figure 15.9 illustrates the manner in which the power taken by an electrical machine is utilized.
The machine efficiency is relevant, within this context, in two ways. First, the total amount of heat
produced can be reduced only if the machine is replaced by another of greater efficiency to give
the same mechanical power output at lower power consumption. For any given machine, the total
heat produced is simply the rate at which power is supplied, less any work done against gravity.
Secondly, the efficiency of the machine determines the distribution of the heat produced. The
higher the efficiency, the lower the heat produced at the motor and transmission, and the greater
is the percentage of heat produced at the pick-point, conveyor rollers, along the machine run or
by any other frictional effects caused by the operation of the device.
Direct heat
losses from
machine body
Pow(1-E) (kW)
Motor Drive
Efficiency
E (kW)
Pow (kW)
Power taken
by machine
Useful energy
U = Pow x E kW
Wg
Work done
against gravity
Wg (kW)
U - Wg
Heat dissipated along

working length of
machine
U Wg (kW)
Figure 15.9 Heat produced by electrical machines.
15 - 34
Example.
A 2000 m long conveyor transports 500 t/hr through a vertical lift of 200 m. If the conveyor motor
consumes 1000 kW at a combined motor/transmission efficiency of 90%, calculate the heat
emitted
(1) at the gearhead and
(2) along the length of the conveyor.
Solution.
Work done against gravity = mass flow x g x vertical lift
500 1000
=
9.81 200 = 272.5 10 3 W or 272.5 kW
60 60
kg
m
m
s
s2
1 . Heat generated at gearhead = (100 - 90)% of 1000 kW = 100 kW
2 . Heat generated along length of conveyor = 1000 - 272.5 100 = 627.5 kW
15.3.2.2. Diesel equipment
The internal combustion engines of diesel equipment have an overall efficiency only about one
third of that achieved by electrical units. Hence, diesels will produce approximately three times as
much heat as electrical equipment for the same mechanical work output. This can be
demonstrated by taking a typical rate of fuel consumption to be 0.3 litres per rated kW per hour.
At a calorific value of 34 000 kJ/litre for the diesel fuel, the heat produced is
0 .3
(litres)
34000
60 60 (kW output. second)
(kJ heat )
(litre)
= 2.83 kJ/s (or kilowatts) of heat emitted for each kilowatt of mechanical output.
This heat appears in three ways each of which may be of roughly the same magnitude. One third
appears as heat from the radiator and machine body, one third as heat in the exhaust gases and
the remaining third as useful shaft power which is then converted to heat (less work done against
gravity) by frictional processes as the machine performs its task.
As with other types of heat emitting equipment there is little need, in most cases, to consider peak
loads. It is sufficient to base design calculations on an average rate of machine utilization. The
most accurate method of predicting the heat load is on the basis of average fuel consumption
during a shift. However, in many mines, records of fuel consumption by individual machines or,
even, in separate sections of the mine, seem not to be maintained. The ventilation planner often
must resort to the type of calculation shown above and using an estimated value for machine
utilization. The latter is defined as the fraction of full load which, if maintained continuously, would
use the same amount of fuel as the actual intermittent load on the machine. (Section 16.2.3)
A difference between diesel and electrical equipment is that the former produces part of its heat
output in the form of latent heat. Each litre of diesel fuel that is consumed produces approximately
1.1 litres of water (liquid equivalent) in the exhaust gases (Kibble, 1978). This may be multiplied
several times over by the evaporation of water from engine cooling systems, some emission
control units and enhanced evaporation from airway surfaces. In-situ tests have shown that the
factor can vary from 3 to 10 litres of water per litre of fuel (McPherson, 1986) depending upon the
power and design of the engine, exhaust treatment system and the proficiency of maintenance.
15 - 35
Example.
Two load-haul-dump vehicles consume 600 litres of diesel fuel in an 8 hour shift. Tests have
shown that water vapour is produced at a rate of 5 litres (liquid equivalent) per litre of fuel. If the
combustion efficiency is 95 per cent and the total calorific value of the fuel is 34 000 kJ/litre,
calculate the sensible and latent heat loads on the stope ventilation system.
Solution.
Total amount of heat produced from burning 600 x 0.95 litres of fuel
=
=
=
600 x 0.95 x 34 000

19.38 x 106 kJ in 8 hours
19.38 10 6
= 673 kW
8 60 60
(This is equivalent to the continuous running of diesel machinery of rated output = 673/2.83 = 238 kW)
Amount of water produced as vapour
= 600 x 0.95 x 5 = 2 850 litres (liquid equivalent)
i.e. 2 850 kg of water
Latent heat emitted in 8 hours
= 2450 x 2850 = 6.982 x 106 kJ
where 2450 kJ/kg is an average value for the latent heat of evaporation of water.
6.982 10 6
i.e.
= 242 kJ/s or kW
8 3600
Then, sensible heat produced = total heat - latent heat
= 673 242
=
431 kW
In summary, operation of the diesels results in heat emissions at an average rate of

431 kW sensible heat
242 kW latent heat
673 kW total heat
In addition to the particulate and gaseous pollutants emitted in diesel exhausts, the heat
produced by these machines mitigates against their use in hot mines. Nevertheless, the flexibility
and reliability of diesel units has tended to perpetuate their continued widespread employment
underground.
15.3.2.3. Compressed air
When compressed air is used for drilling or any other purpose then there are two opposing effects
that govern the heat load. First, the work output of the machine will result in frictional heat at the
pick point or other places as the machine performs its task. Second, the removal of energy from
the compressed air will result in a reduction of the temperature of that air at the exhaust ports of
the machine.
If we assume that the change in potential energy of the air is negligible and that there is no
significant heat transfer across the machine casing then the steady flow energy for the
compressed air motor is
15 - 36
u1 u 2
2
+W
= C p ( 2 1 )
J/kg
(15.49)
where u1 and u2 are the velocities of the air in the supply pipe and the exhaust ports, respectively,
W is the mechanical work done on the air (this is numerically negative as mechanical
energy is leaving the system; J/kg),
Cp is the specific heat of air, (1005 J/kgC) for dry air,
1 is the temperature of the supply air (C) and
2 is the temperature of air at the outlet ports (C).
In the majority of cases the temperature of the compressed air supplied to the motor is equal to
the ambient dry bulb temperature, amb. Furthermore, the velocity of the compressed air in the
supply pipe, u1, is small compared with that at the exhaust ports, u2.
Hence,
u2
2
+W
= C p ( 2 amb )
J/kg
(15.50)
Now consider the subsequent process during which the cold exhaust air mixes with the ambient
air, increasing in temperature, extracting heat from the general ventilating airstream and, hence,
cooling that airstream. Again, the velocity of the ambient airstream will be small compared with
that at the outlet ports, u2. The steady flow energy equation for this process becomes
2
u2
2
= C p ( amb 2 ) q
J/kg
(15.51)
where q is the heat added to the machine exhaust from the ambient air.
The overall effect may be quantified by adding equations (15.50 and 15.51) giving
W
= q
J/kg
Hence, the mechanical work output of the machine (which is subsequently transferred by friction
into heat) is balanced precisely by the cooling effect on the ventilating airstream. If the moisture
content of the compressed air is similar to that of the ambient airstream, then there is no net
heating or cooling arising from the expansion of the air through the compressed air motor.
Despite this analysis, a cooling effect is certainly noticeable when standing immediately
downstream from a compressed air motor. There are two reasons for this. First, the cooling of the
ambient airstream, q, by a motor exhausting at sub-zero temperatures is immediate and direct,
whilst part of the balancing heat produced by friction may be stored temporarily within the body of
the machine, in the solid or broken rock, and distributed over the working length of the machine.
Secondly, the compressed air supply will normally be drier and have a much lower sigma heat
than the ventilating air. The increase in sigma heat of the exhaust air will then exceed the
numerical equivalent of the work output and, hence, result in a true net cooling of the ambient air.
This will be of consequence only if the mass flow of compressed air is significant compared with
the mass flow of the ventilating airstream. The dryness of compressed air, coupled with a very
local kinetic energy effect [u2/2 = Cp(2 1)] also explains why air issuing from a leaky
compressed air pipe appears cool even though no energy has been extracted from it.
In calculating the heat load for a complete airway, district or mine, it errs on the side of safety to
assume that there is no net cooling effect from compressed air motors. Only if a detailed analysis
in the immediate vicinity of compressed air devices is required need the local cooling effects be
considered.
15 - 37
Air compressors are large sources of heat. Indeed, the local cooling effect at the outlet ports of
the compressed air motors are reflected by an equivalent amount of heat produced at the
compressors, even if the latter are 100 % efficient. This need be of little import if the compressors
are sited on surface and the compressed air is cooled before entering the shaft pipes. If,
however, compressors are sited in intake airways close to hot workings, then it is beneficial to
investigate the circulation of hot water from the compressor coolers through heat exchangers in
return airways.
15.3.3. Fissure water and channel flow
Groundwater migrating through strata toward a subsurface opening can add very considerably to
the transfer of geothermal heat through the rock (Section 15.2.10). Such water may continue to
add heat to the ventilating airstream after it has entered the mined opening.
Fissure water is often emitted at a temperature close to the virgin rock temperature (VRT). In
circumstances of local geothermal activity or radioactive decay, it may even be higher. The total
heat load on the mine environment can be calculated from the flowrate and the drop in
temperature of the water between the points of emission and effective exit from the mine
ventilation system.
Example.
A mine produces 5 million litres of water per day, emitted at an average temperature of 42 C.
When the water is delivered into the shaft sump for pumping to the surface, the water
temperature is 32 C. Determine the heat load from the water on the mine ventilation system.
Solution.
Heat load
=
=
Mass flow x specific heat x drop in temperature

5 000 000
24 60 60
kg
s
4187
1000
kJ
kg o C
(42 32)
o
= 2 423 kW
The rate at which heat is emitted into the airstream depends upon the difference between the
temperature of the air and the water, and whether the water is piped or in open channels. In the
latter case, cooling by evaporation will be the major mode of heat transfer and will continue while
the temperature of the water exceeds the wet bulb temperature of the air. Hot fissure water
should be allowed no more than minimum direct contact with intake air. The most effective means
of dealing with this problem are
(a) to transport the water in closed pipes and
(b) to restrict water flow routes to return airways.
A less effective but expedient measure close to the points of water emission is to restrict air/water
contact by covering drainage channels by boarding or other materials.
Where chilled service water is employed, condensate run-off may be at a temperature below that
of the local wet bulb temperature. While this continues, the condensate will continue to absorb
both sensible and latent heat, thus providing a cooling effect on the airflow.
In, or close to the working areas of many mines, it is often inevitable that open drainage channels
will be used. Similarly, the footwalls or floors may have areas that are covered with standing or
slowly moving water.
15 - 38
The transfer of sensible and latent heat between the surface of the water and the air may be
determined from equations (15.34 and 15.35).
i.e.
q sen
= hc ( ws d )
and
qL
= 0.0007 hc Lws
(ews
W/m 2
e)
W/m2
P
where subscript ws means water surface and d means dry bulb.
The heat transfer coefficient for the water surface, hc, is the least certain of the variables,
depending upon the geometry of the airway, the condition of the airflow (in particular, the local air
velocity) and the degree of turbulence existing in the flowing water. An approximation may be
obtained from equations (15.22 and 15.23) using the Reynolds' Number for the airway and a
value of coefficient of friction, f, that represents the degree of water turbulence on the liquid
surface. This may vary from 0.002 for a calm flat surface to 0.02 for a highly turbulent surface.
Example.
Water flows at a rate of 60 litres per second along an open drainage channel that is 0.5 m wide
and 0.25 m deep. The channel is located in an airway of cross-sectional area 12 m2 and
perimeter 14 m, and which passes an airflow of 50 m3/s. Over a 10 m length of channel, the
following mean values are given:
water temperature,
ws
air dry bulb temperature, d
air wet bulb temperature, w
barometric pressure,
P
=
=
=
=
32 C
30 C
25 C
112 kPa
Determine the sensible and latent heat emitted from the water to the air and the corresponding
drop in water temperature over the 10 m length. Assume that no heat is added to the water from
the strata.
Solution.
(a) Determine the heat transfer coefficient for the water surface.
60
= 0.06 m3/s
1000
0.06
Velocity of water =
= 0.48 m/s
0.5 0.25
Volume flow of water =
At this velocity, it is estimated that ripples on the water surface will have an equivalent coefficient
of friction, f = 0.008.
The Reynolds' Number for the airway
Re = 268 000
Q
50
= 957 140
= 268 000
per
14
Equation (15.22) gives the Nusselt Number to be Nu = 2127 (this can be estimated directly from
Figure 15.5)
Hydraulic mean diameter of the airway,
d =
4A
4 12
=
per
14
3.429 m
15 - 39
The convective heat transfer coefficient is then given by equation (15.23).

Using ka = 0.026 W/(mC) (reference Table 15.1),
2127 0.026
3.429
hc =
= 16.1 W/(m2C)
(b) Determine the psychrometric conditions.

Using the psychrometric relationships given in Section 14.6, the following values are determined:
for the water surface at 32 C
=
2426 x 103 J/kg
Lws
and
ews
=
4753 Pa
for the air at
w = 25 C, d = 30 C and P = 112 kPa,

e = 2805 Pa
[The advantages of keeping the psychrometric equations of Section 14.6 programmed into a
pocket calculator should be very obvious by now.]
(c) Determine the heat transfers from the water to the air:
Sensible heat, qsen = hc (ws - d) = 16.1 (32 - 30) = 32.2 W/m2
q L = 0.0007 hc Lws
Latent heat,
(ews
e)
(4753 2805 ) = 475.5 W/m2
= 0.0007 16.1 2426 10 3
P
112 000
As the width of the drainage channel in the 10 m length is 0.5 m, the heat flows may be stated as:
32.2 10 0.5
1000
475.5 10 0.5
1000
Sensible heat
Latent heat
0.161 kW
2.378 kW
These values illustrate the dominant effect of latent heat transfer and why it is important to
prevent direct contact between the air and the water.
The total heat gained by the air from the water is the sum of the sensible and latent heat
transfers,
q = 2.378 + 0.161 = 2.539 kW.
(d) Determine the drop in temperature of the water.
As 1 litre of water has a mass of 1 kg, the mass flow of water, m = 60 kg/s. The total heat lost by
the water, then becomes
q
m Cw
where
and
Cw
=
=
giving
change in water temperature (C)

specific heat of water ( 4187 J/(kgC) )
q
2.539 10 3
= 0.0101 C in the 10 m length.
=
m Cw
60 4187
15 - 40
15.3.4. Oxidation
Coal and sulphide ore mines are particularly liable to suffer from a heat load arising from
oxidation of fractured rock. An estimate of the heat produced can be determined from the rate of
oxygen depletion.
Assuming complete combustion of carbon and using the atomic weights of the elements,
C + O2
12 kg + 32 kg
CO2
44 kg
(15.52)
Each 1 kg of oxygen consumed oxidizes 12/32 kg of carbon. Taking the calorific value of carbon
as 33 800 kJ/kg gives a corresponding heat production of (12/32) x 33 800 = 12 675 kJ of heat
per kg of oxygen used.
Similarly, for complete oxidation of sulphur:
S
+ O2
SO2
32 kg + 32 kg
64 kg
(15.53)
Each 1 kg of oxygen consumed will oxidize 1 kg of sulphur (calorific value 9304 kJ/kg) to produce
9304 kJ of heat.
Example.
10 m3/s of air enters a working district in a coal mine at an oxygen content of 21% by volume, and
leaves at 20.8%. Calculate the heat generated by oxidation assuming complete combustion.
Solution.
The equivalent volume flow of oxygen at entry = 0.21 x 10 = 2.1 m3/s
The equivalent volume flow of oxygen at exit = 0.208 x 10 = 2.08 m3/s
Taking the density of oxygen to be 1.3 kg/m3 gives the oxygen depletion rate to be
(2.1 - 2.08) x 1.3 = 0.026 kg/s
Heat produced
0.026 x 12 675
330 kW
The type of calculation illustrated by this example is useful in determining the heat produced by
oxidation in an existing mine or section of a mine where air samples can be taken for analysis in
intake and return airways. The extent to which oxidation takes place depends upon the
mineralogical content of the material being oxidized, the psychrometric condition of the air and
the surface area exposed. These factors make it difficult to predict heat loads from oxidation by
other than empirical means.
In the case of open surfaces in operating airways and working places, the heat of oxidation will be
an immediate and direct load on the mine ventilation system. On the other hand the heat that is
produced by oxidation in waste areas, old workings, orepasses or in caved stopes will initially be
removed only partially by leakage air. The remainder will be retained causing a rise in the
temperature of the rock. This, in turn, enhances the rate of heat transfer to the leakage air and, in
most cases, an equilibrium is reached when the heat removed by the air balances the heat of
oxidation. However, in minerals liable to spontaneous combustion the increased temperature of
the rock will accelerate the oxidation process to an extent that the heat removed may not reach
equilibrium with the heat produced. In such cases the temperature will continue to rise until the
rock becomes incandescent resulting in a spontaneous fire (Section 21.4).
15 - 41
15.3.5. Explosives
The heat that is produced during blasting varies with the type and amount of explosives used
(charge density). The amount of heat released by most explosives employed in mining falls within
the range of 3700 kJ/kg for ANFO2 to 5800 kJ/kg for nitro-glycerine (Whillier, 1981). This heat is
dispersed in two ways. First, a fraction of it will appear in the blasting fumes and cause a peak
heat load on the ventilation system following blasting. In mines where a re-entry period is
enforced this peak load should have cleared prior to personnel being readmitted to the area.
Secondly, the remainder of the heat will be stored in the broken rock. The magnitude of this will
depend upon the mining method. If the rock is blasted into a free space through which the
ventilating airstream passes, such as an open stope or on a longwall face, and the fragmentation
is high, then as much as 40 - 50 % of the heat produced by the explosive may be removed rapidly
as a peak load with the blasting fumes. On the other hand, if the blast occurs within a region
through which there is little or no ventilation such as in sublevel or forced caving techniques, or if
the fragmentation is poor, then a much larger proportion of the heat will be retained in the rock.
Example.
In a 2000 tonne blast, the charge density of ANFO is 0.8 kg/t. It is estimated that 20% of the blast
heat will be removed within 1 h with the blasting fumes.
(1) Calculate the mean value of the rate of heat removed by the airflow during this hour.
(2) If the specific heat of the rock is 950 J/kg C determine the average increase in temperature of
the rock due to blasting.
Solution:
1. Mass of explosive used = 2000 x 0.8 = 1 600 kg ANFO
Heat produced by ANFO = 3700 x 1600 = 5 920 000 kJ
Twenty per cent of this heat is removed in the blasting fumes over 1 hour.
Average rate of heat removal with blasting fumes
=
5 920 000
kJ
0 .2
3600
s
329 kW
2 . Heat retained in rock = 5 920 000 x 0.8 = 4 736 000 kJ
Rise in rock temperature =
Heat retained
4 736 000
=
Mass Specific heat
2000 1000 0.950
= 2.49
kJ kg o
C
kg kJ
15.3.6. Falling rock

When strata or fractured rock moves downward under gravitational effects, the loss of potential
energy will ultimately produce an increase in temperature through fragmentation, impact, braking,
or other frictional effects. The fraction of the resulting heat that produces a load on the ventilation
system again depends upon the exposure of the broken rock to a ventilating airstream. Hence, for
example, a large amount of heat is produced by the mass of superincumbent strata subsiding
through to the surface over a period of time. Fortunately, the vast majority of this is retained within
the rock mass, raising its temperature very slightly, and little enters the mine ventilation system.
2
Ammonium nitrate, fuel oil mixture.

15 - 42
On the other hand, mineral descending through an ore pass or vertical bunker may immediately,
or subsequently, be exposed to a ventilating airstream. The loss of potential energy of the rock
will then appear as a heat load on the ventilation system. This can be significant.
Assuming that there is negligible heat transfer between the fragmented material and the
surrounding strata, the rise in temperature of falling rock, r, is given by the expression
z g
C
where z is the distance fallen (m),

g is the gravitational acceleration (9.81 m/s2),
C is the specific heat of rock (J/kgC)
or
r = 981/C
C per 100 m fall
(15.54)
(15.55)
15.3.7. Fragmented rock

When fragmented rock is exposed to a ventilating airstream and there is a temperature difference
between the rock and the air, then heat transfer will take place. This will occur on working faces,
drawpoints, conveyors and along other elements of a mineral transportation system.
The heat load from broken rock is given by
m C ( 1 2 )
kW
(15.56)
where
m is the mass flow of rock (kg/s),
C is the specific heat of rock (kJ/kgC),
1 is the temperature of the broken rock immediately after fragmentation (C) and
2 is the temperature of the rock at exit from the ventilation system (C).
In most mining methods the temperature of the solid rock immediately prior to fragmentation will
be less than the virgin rock temperature (VRT) due to cooling of the rock surface. On the other
hand, the process of fragmentation, whether by blasting or mechanized techniques, will raise the
temperature of the broken rock. As shown in earlier sections, estimates may be made of these
effects. However, in practice, the procedure may be simplified by assuming that the temperature
of the newly broken rock, 1, is equal to VRT.
The mean temperature of the broken rock at exit from the system, 2, will depend upon the
degree of fragmentation, the exposure of rock surfaces to the airstream, and the velocity and
psychrometric condition of the air. Mineral transported along a conveyor system will cool to a
much greater extent than in a locomotive system. Furthermore, wetted material will yield up its
heat more readily due to evaporative cooling. Although the rate of heat transfer from a given
particle size of rock to a known quantity and quality of airflow can be calculated, the many
variables in any actual transportation system enforce the use of empirical measurements of
temperature along the mineral transportation system. Such measurements are facilitated by the
use of infra-red thermometry.
15 - 43
Example.
Fragmented ore of specific heat 900 J/(kgC) enters the top of an ore pass at a temperature of
35C and at a rate of 500 tonnes per hour. At an elevation 200 m below, the ore is discharged
onto a conveyor and reaches the shaft bottom at a temperature of 32C. Calculate the heat
transferred from the broken ore to the ventilation system over the length of the conveyor.
Solution.
From equation (15.54), temperature rise of the rock in the ore pass
=
200 9.81
= 2.18 C
900
(assuming no heat transfer between the

fragmented ore and the surrounding strata)
Temperature of rock at bottom of ore pass = 35 + 2.18 = 37.18 C

500 1000
= 138.9 kg/s
3600
Heat transfer to the airflow from ore on the conveyor (see equation (15.56))
Mass flow of ore, m =
m C ( 1 2 ) = 138.9 0.900 (37.18 32) = 647 kW
15.3.8. Metabolic heat

The rate at which the human body produces metabolic heat depends upon a number of factors
including rate of manual work, physical fitness and level of mental stress. The question of heat
transfer between the human body and the surrounding environment is of great importance in
ascertaining the risk of heat stress in mine workers (Chapter 17). However, metabolic heat makes
only a small contribution to mine heat load. It may, nevertheless, become significant where labour
intensive activities take place in a location of limited throughflow ventilation such as a poorly
ventilated heading or a barricaded refuge chamber.
Like other heat engines, the human body emits heat by three mechanisms. The most significant
of these in physiological processes is heat loss from the body surface. Secondly, the large, wet
surface area of the lungs provides an effective heat exchanger and exhaust heat is emitted
through respiration. Third, any mechanical work done by the individual on the external
surroundings will produce frictional heat, unless that work is done against gravity. As the human
body is an inefficient heat engine, this latter mechanism is the least significant and is often
omitted in calculations of physiological heat transfer.
The metabolic heat produced by a fit worker who is acclimatized to the environment varies from
about 100 W for sedentary work, through 400 W for a medium level of activity such as walking, to
over 600 W during intermittent periods of strenuous manual work.(See Table 17.1)
15 - 44
BIBLIOGRAPHY
ASHRAE (1985). American Society of Heating, Ventilating and Air Conditioning Engineers.
Fundamentals Handbook in SI units. Chapters 3 and 5.
Bluhm, S.J. (1987). Chamber of Mines of South Africa. Private communication.
Chilton, T.H. and Colburn, A.P. (1934). Mass transfer (absorption) coefficients - prediction from
data on heat transfer and fluid friction. Industrial and Engineering Chemistry, Nov. p. 1183.
Carrier, W.H. (1940). Air cooling in the gold mines on the Rand. Trans. AIME, Vol. 141.
pp. 176-287.
Carslaw, H. and Jaeger, J.C. (1956). Conduction of heat in solids. Oxford University Press.
Danko. G. et al (1987). Development of an improved method to measure in-situ thermal rock
properties in a single drill hole. Proc. 3rd U.S. Mine Ventilation Symposium. Penn State.
pp. 33-52.
Danko, G. et al (1988). Heat, mass and impulse transport for underground airways. Trans. 4th
Int. Mine Ventilation Congress. Brisbane, Australia, pp 237-247.
Deen J.B. (1988). Laboratory verification of heat transfer analogies. M.S. Dissertation. University
of California, Berkeley.
Gibson, K.L. (1976). The computer simulation of climatic conditions in underground mines. Ph.D.
Thesis, University of Nottingham.
Goch,. D.C. and Patterson, H.S. (1940). Heat flow into tunnels. Jnl. Chem., Metal., and Mining
Soc. of S. Africa. Vol. 41. pp. 117-28.
Hartman, H.L. (1982). Mine Ventilation and Air Conditioning, Ch. 20. Wiley.
Hemp, R. (1965). Air temperature increases in airways. Journal of the Mine Ventilation Society of
S. Africa. Vol. 38, No. 2, Jan, Feb.
Hemp, R. (1982). Environmental Engineering in South African Mines. Ch. 22. Mine Vent. Soc. of
S. Africa.
Kibble, J.D. (1978). Some notes on mining diesels. Mining Technology. October, pp. 393-400.
McPherson, M.J. (1976). Refrigeration in South African Gold Mines. Mining Engineer (Feb.)
pp 245-258.
McPherson, M.J. (1986). The analysis and simulation of heat flow into underground airways.
International Journal of Mining and Geological Engineering, Vol. 4. pp.165-196.
Mousset-Jones, P. and McPherson, M.J. (1984). Measurement of in-situ thermal parameters in
an underground mine. Proc. 2nd Annual Workshop, Generic Mineral Technology Center (Mine
System Design and Ground Control) USA. pp.113-131.
Mousset-Jones, P. et al (1987). Heat transfer in mine airways with natural roughness. 3rd US
Mine Ventilation Symposium. Penn State. pp. 42-52.
Nunner, W. (1956). Z ver. Deut. Forschungshaft, 455.
15 - 45
Robinson, G. et al (1981). Underground environmental planning at Boulby Mine, Cleveland

Potash Ltd. Trans. Inst. of Mining and Met. London, July.
Stewart, J. (1982). Environmental Engineering in South African mines. Ch. 20. Fundamentals of
heat stress. pp. 495-533.
Stroh. R. (1979). A note on the downcast shaft as a thermal flywheel. Jnl. of the Mine Vent. Soc.
of S. Africa, 32. pp. 77-80.
Verma, Y.K. (1981). Studies in virgin strata temperatures, with special reference to the NCB'S
Western Area Mines. The Mining Engineer Vol. 140. No. 234 (March). pp. 655-663.
Whillier, A. (1981). Predicting cooling requirements for caving and sublevel stoping in hot rock.
Int. Conf. on Caving and Sublevel Stoping. AIME Denver.
Whillier, A. (1982). Environmental Engineering in South African Mines Ch. 19.
APPENDIX A15.1
Analytical solution of the three dimensional transient heat conduction equation (15.13) as
obtained by Carslaw and Jaeger (1956).
G =
e V
dV
[I 0 (V ) + V / DI1 (V )] 2 + [ Y0 (V ) + V / DY1 (V )] 2 V
where I0, I1, Y0 and Y1 are Bessel functions and V is the variable of integration.
APPENDIX A15.2
Gibsons algorithm for the numeric determination of dimensionless temperature gradient, G.
(rock diffusivity, m2/s)
Enter
t
ra
k
Then
F
B
x
y
c
c
d
(age of airway, seconds)

(effective radius of airway = perimeter/(2), metres)
(thermal conductivity of rock, W/(mC) ).
(Fourier Number )
hr a / k (Biot Number )
log10 (F )
log10 (B )
x (0.000104 x + 0.000997 ) 0.001419
{x [x (xc 0.046223 ) + 0.315553 ] + 0.006003}
y (x (4 x 34 ) 5 ) / 120
= t / r a
=
=
=
=
=
=
d = 0.949 + 0.1exp 2.69035 d 2
(216 + 5 x ) 0.0725 + 0.01tan 1 x

2
m = sqr (y - c ) +
70
0.7048
n = (y + c m ) / 2
G = 10 n / d
15 - 46
APPENDIX A15.3.
Background to equations for the heat transfer coefficient, h.
Consider the transfer of heat, q (W/m2) across the boundary layers shown on Figure A15.1. In the
laminar sublayer, there are no cross velocities and Fourier's law of heat conduction applies:
s
b
laminar sublayer
turbulent boundary layer
turbulent core
velocity
profile
Figure A15.1 Heat is transported by conduction and molecular diffusion across the laminar
sublayers, and by eddy diffusion across the turbulent boundary layers.
d
W/m2
dy
where ka = thermal conductivity of air W/(mC)
= fluid temperature (C) and
y = distance from the wall (m)
q = ka
(A15.1)
In the turbulent boundary layer (and within the mainstream), heat transfer is assisted by eddy
action and the equation becomes:
q = (k a + a C p E h )
d
dy
W/m2
(A15.2)
where a = air density ( kg/m3)

Cp = specific heat (1005 J/kgC for dry air) and
Eh = eddy diffusivity of heat (m2/s)
The term a Cp (J/m3C) is the amount of heat transported in each m3 for each Centigrade degree
of temperature difference while Eh (m2/s) represents the rate at which this heat is transported by
eddy action. The product a Cp Eh is much larger than ka.
15 - 47
Equation (A15.2) can be re-written as

k
d
q
moC
= a + Eh
s
aC p
a C p
dy
d
= { + E h }
dy
ka
as =
= thermal diffusivity of still air (m2/s)
a Cp
(A15.2a)
Figure A15.1 illustrates two boundary layers only. In fact, a third transitional or buffer layer exists
between the laminar and turbulent layers, within which both conduction and eddy diffusion are
significant. This is similar to the transitional region between laminar flow and fully developed
turbulence shown on Figure 2.6 for pipes.
Let us now consider the transfer of momentum across the boundary layers. For the laminar
sublayer where the flow is viscous, Newton's equation (2.22) applies:
du
dy
N/m2
(A15.3)
where = shear stress transmitted across each lamina of fluid ( N/m2 or Pa)
= coefficient of dynamic viscosity Ns/m2
u = fluid velocity (m/s)
Note that the units of shear stress may be written also as Ns/(m2s)
Therefore, as Ns are the units of momentum, also represents the transfer of momentum across
each square metre per second.
For the turbulent boundary layer, the equation becomes
= ( + a E m )
du
dy
N/m2
(A15.4)
where Em
= eddy diffusivity of momentum (m2/s) and
aEm = momentum transfer across each m2 by eddy action (Ns/m2)
Again, aEm is much larger than .
Then

du
=
+ Em
a
a
dy
= ( + E m )
where =
du
dy
J
kg
(A15.5)
= kinematic viscosity or momentum diffusivity (m2/s)

a
15 - 48
In order to combine heat transfer and momentum transfer, Reynolds divided equations (A15.2a)
and (A15.5), giving,
( + E h ) d dy
q a
=
( + E m ) dy du
aC p
or
( + E h ) d
q
=
( + E m ) du
C p
(A15.5a)
Reynolds argued that as the eddy components predominate and the eddy diffusivity for heat, Eh ,
must be closely allied to the eddy diffusivity for momentum, Em, we may equate those terms,
leaving
q
d
=
C p
du
(A15.6)
Reynolds integrated this equation directly, making the approximation that u remained linear with
respect to y across the composite boundary layers (see the velocity profile on Figure A15.1).
However, both Taylor and Prandtl, working independently, later realized that the integration had
to be carried out separately for the laminar sublayer and the turbulent boundary layer.
Laminar sublayer
From equation (A15.1)
q
dy
0
= ka
q = k a ( s b )
(A15.7)
where
and
s
b
=
=
=
thickness of laminar sublayer (m)

temperature at the surface (C)
temperature at the edge of the laminar sublayer (C)
Similarly, integrating equation (A15.3) from zero velocity at the wall to ub at the edge of the
laminar sublayer
ub
(A15.8)
Dividing (A15.7) by (A15.8) gives

q
k a ( s b )
ub
(A15.9)
Turbulent boundary layer

Here we employ equation (A15.6)
q
Cp
du = d
and integrate from the edge of the laminar sublayer (subscript b) to the mainstream (subscript m)
(see Figure A15.1):
15 - 49
q
Cp
giving
um
du

q
Cp
ub
( b m )
(u m u b )
(A15.10)
A difficulty here is that neither b nor ub are amenable to accurate location or measurement. We
can eliminate b by rewriting (A15.9) as
b = s
q
ub
ka
(A15.11)
Substituting for b in equation (A15.10) and engaging in some algebraic manipulation leads to
m )
1
Cp
um
u b C p
1
1 +
u m k a
Now the combination Cp/ka is another dimensionless number known as the Prandtl Number Pr
q
q
Cp
( s
( s
m )
1
um
ub
(Pr 1)
1+
um
(A15.12)
Furthermore, the shear stress, , is related to the coefficient of friction, f, and the mainstream
velocity, um, by equation (2.41)
= f a
um
2
Also,
q = h(s m) from equation (15.16) where
m = d = dry bulb temperature of the mainstream
Substituting for and q in equation (A15.12),
( s m )
h( s m )
1
=
f
[1+ (Pr 1)u b / u m ]
um
2
Cp a um
2
h
f
1
=
Cp a um
2 [1 + (Pr 1)u b / u m ]
The left hand side of this equation can be separated into three dimensionless groups
hd
ka
Nusselt No. Nu
a um d
ka
Cp
1
Reynolds No. (Re )
1
Prandtl No. (Pr )
Nu
R e Pr
15 - 50
Hence
Nu =
f
1
Re Pr
[1+ (Pr 1)u b / u m ]
2
(A15.13)
This is sometimes known as the Taylor or Taylor-Prandtl equation and is the basic relationship
that has been used by numerous other workers. These varied in the manner in which they treated
the ub/um ratio as ub remains elusive to measure. Taylor himself used an empirical value of
ub/um = 0.56. Several other relationships are listed in Table A15.1. A more comprehensive listing
is given by Danko (1988)
Value of ub/um
Authority
Taylor
0.56
Von Krmn
9.77 f / 2
Rogers and Mayhew
1.99 Re 0.125
Table A15.1 Some reported values of the ub/um ratio

A comprehensive theoretical and experimental study of flow over rough surfaces by Nunner
(1956) produced a more sophisticated equation
Nu =
1
f
Re Pr
2
1 .5
1 +
1/ 6
1/ 8
Re
Pr
(A15.14)
f
Pr
f 1
0
where
f0 = friction coefficient for a smooth tube at the same value of Reynolds' Number.
This is known as Nunner's equation.
An analysis of hydraulically smooth pipes by Colburn gave
Nu = 0.023 Re 0.8 Pr
0.4
(A15.15)
For air in the atmospheric range, the Prandtl Number is near constant. Using the values at 20C,
= (17 + 0.045 20 ) 10 6 = 17.9 10 6

k a = 2.2348 10 4 (273.15 + 20 )
0.8353
C p = 1005
J/ kg C
Ns/m 2
= 0.0257
(Section 2.3.3)
( )
W/ m o C
(Table 15.1)
gives
Pr
Cp
ka
1005 17.9 10 6
0.0257
= 0.700
(A15.16)
In order to illustrate the differences between the relationships quoted, Figure A15.2 has been
produced for values of Pr = 0.7 and f = 0.02. There is reasonable agreement within the Von
Krmn, Taylor and Rogers equations. The Nunner relationship produces Nusselt Numbers a
little less than one half those given by the other authorities. As might be expected, the Colburn
smooth tube equation gives much lower Nusselt Numbers and, hence, heat transfer coefficients.
15 - 51
Experimental evidence in mine airways (Mousset-Jones et al, 1987 and Danko et al, 1988) and
also in scale models under controlled conditions (Deen, 1988) has indicated a better correlation
with the Nunner equation than the others. Hence, this is the relationship suggested for use in
underground openings at the present time. Ongoing research will no doubt produce further
equations or algorithms for Nu.
One further problem remains with equation (A15.14), the evaluation of f0, the coefficient of friction
for a hydraulically smooth tube at the same Reynolds Number. However, the approximation given
in Section 2.3.6.3 gives acceptable results for the range of Reynolds' Numbers common in mine
airways:
f0
= 0.046 Re 0.2
(A15.17)
Substituting for f0 in the Nunner equation and using Pr = 0.7 gives

Nu =
0.35 f Re
(A15.18)
1.592
15.217 f Re 0.2 1
1 +
0.125
Re
1,000,000
von Krmn
Nusselt Number Nu
Taylor
Rogers and
Mayhew
100,000
Nunner
Colborn
(smooth tube)
10,000
1,000
100
10,000
100,000
1,000,000
10,000,000
100,000,000
Reynolds Number Re
Figure A15.2 Relationships between Nusselt Number and Reynolds Number proposed by
various authorities for Pr = 0.7 and f = 0.02.
15 - 52
APPENDIX A15.4
Derivation of the equation for latent heat of evaporation at a wet surface.
When any two gases, a and b, are brought into contact then molecules of each gas will diffuse
into the other at the interface. The fundamental law governing the rate of diffusion is
m = Dv
where
a
x
kg
(A15.19)
m2s
m = rate of diffusion across unit area of gas a into gas b (kg/(m2s)

a = density of gas a ( kg/m3)
x = distance beyond the interface (m) and
Dv = diffusion coefficient of gas a into gas b (m2/s)
(experimental results give this as 25.5 x 10-6 m2/s for the diffusion of water vapour into air
(ASHRAE, 1985))
This is known as Fick's Law and is analogous to Fourier's law for heat conduction:
q = k
W/m2
At a wet surface with unsaturated air flowing over it, water vapour is produced continuously by
evaporation and will be transported by molecular diffusion through the laminar sublayer then,
additionally, by eddy action through the transitional (buffer) and turbulent layers. The transfer of
heat, momentum and mass are all affected in the same way by eddies at any given surface.
Hence, the governing equations have the same form for all three. Just as we had
q = h ( s d )
W/m2
for heat flow across the boundary layers, so we can write an analogous equation for the migration
of water molecules:
m = hm ( vws v )
where
m
hm
vws
v
=
=
=
=
kg/(m2s)
(A15.20)
mass of water vapour migrating from each square metre per second ( kg/(m2s) )
mass transfer coefficient (m/s)
density of water vapour at the wet surface (kg/m3)
density of water vapour in the mainstream (kg/m3)
The density difference (vws - v) provides the potential or driving mechanism for the diffusion of
mass.
Now from the general gas law
e
where
v Rv T
Pa
e = partial pressure of water vapour (Pa)

Rv = gas constant for water vapour (461.5 J/kg K) and
T = absolute temperature (K)
Hence, equation (A15.20) becomes
m = hm
1
(ews e )
Rv T
kg
m2s
(A15.21)
15 - 53
e is the vapour pressure within the main airstream and ews the vapour pressure at the wet
surface. It is reasonable to assume that immediately adjacent to the liquid surface, the space will
be saturated. ews is then also the saturated vapour pressure at the wet surface temperature, ws.
Now, the general gas law for a mixture of air and water vapour gives
P
m =
where
Pa
Rm T
m = density of moist air ( kg/m3)

P
= barometric pressure (Pa) and
Rm = gas constant for moist air (J/(kg K)
However, for the atmospheric range in mines, Rm differs only slightly from the gas constant for dry
air, Ra = 287.04 J/(kg K) [Section 14.2.4]. Hence, we can express equation (A15.21) as
m = hm
= hm
m 1
(ews e )
Rv T m
m
Ra T
(ews e )
Rv T P
= hm m 0.622
(ews
e)
P
kg
(A15.22)
m2s
as
Ra
Rv
287.04
= 0.622
461.5
Now let us turn our attention to the mass transfer coefficient, hm. In 1934, Chilton and Colburn,
investigating the analogy between heat transfer and mass diffusion, produced a relationship
between the convective heat transfer coefficient, hc, and the mass transfer coefficient, hm. In its
most compact form (ASHRAE, 1985), this can be expressed as
hm
hc
Cp m
Pr
Sc
2/3
m/s
(A15.23)
where
Cp = specific heat of air (J/kg K)
a Cp
Pr =
is the Prandtl Number for air (Appendix A15.3)
ka
a = dynamic viscosity of air (Ns/m2
ka = thermal conductivity of air ( W/(mC) )
and
Sc is yet another dimensionless number (Schmidt Number) =
a
a Dv
Both the Prandtl and Schmidt dimensionless numbers remain reasonably constant over the range
of air pressures and temperatures found in subsurface facilities.
Using the values for air at 20C,
a = 17.9 x 10-6 Ns/m2;
ka = 0.0257 W/(mC)
Cp = 1010 J/(kgK) (mid-range value for moist air):
a = 1.2 kg/m3 (standard air density) and
Dv = 25.5 x 10-6 m2/s for the diffusion of water vapour into air,
15 - 54

gives
and
Pr = 0.703
Sc = 0.585
Combining equations (A15.22 and A15.23) gives

m =
hc
Cp m
Pr

Sc
2/3
m 0.622
(ews
e)
P
and, inserting the numerical values for Pr and Sc,

m = hc 1.130
0.622 (ews e )
Cp
P
kg
m2s
(A15.24)
Now the latent heat emitted in the vapour when m kg of surface water is evaporated is
kg
J
= W/m 2
kg
m s
q L = m Lws
where
Lws =
=
(A15.25)
latent heat of evaporation at the temperature of the wet surface (see equation (14.6))
(2502.5 - 2.386 ws)103
J/kg
Substituting for m from equation (A15.24) gives

q L = hc 1.130
0.622 Lws
Cp
(ews
e)
P
W
m2
(A15.26)
or, using the value Cp = 1010 J/kgK for moist air,

q L = 0.0007 hc Lws
(ews
e)
P
W
m2
(A15.27)
Equation (A15.27) may also be written as

q L = he
(ews
e)
W
m2
(A15.28)
where the evaporative heat transfer coefficient, he, is given as

he
= 0.0007 hc
Lws
P
W
m 2 Pa
(A15.29)
The ratio he/hc is known as the Lewis Ratio (LR) .

Using the latent heat of evaporation of water at 20C (2455 x 103 J/kg) and a barometric
pressure of 100 x 103 Pa gives
LR =
he
hc
= 0.017
C
Pa
(A15.30)
The Lewis ratio changes slowly through the atmospheric range of pressures and temperatures.
A correlation with psychrometric relationships occurs in equation (A15.26). The group
0.622 Lws /Cp is the reciprocal of the psychrometric constant introduced in Section (14.4.2).
15 - 55
Chapter 16 Simulation of Climatic Conditions in the Subsurface
CHAPTER 16. SIMULATION OF CLIMATIC CONDITIONS IN THE SUBSURFACE

16.1. BACKGROUND ................................................................................................ 1
16.2. ELEMENTS OF A MINE CLIMATE SIMULATION PROGRAM......................... 2
16.2.1. Organization of the programs ................................................................................................. 2
16.2.2. Incrementation of airway length.............................................................................................. 2
16.2.3. Heat additions ......................................................................................................................... 2
Strata heat ................................................................................................................................... 2
Machines and other sources of heat ........................................................................................... 3
16.2.4. Change in moisture content.................................................................................................... 5
16.2.5. Change in dry bulb temperature and autocompression.......................................................... 5
16.2.6. Change in barometric pressure .............................................................................................. 7
16.2.7. Change in wet bulb temperature............................................................................................. 7
16.2.8. Relative humidity and saturation conditions ........................................................................... 8
16.2.9 Indices of heat stress ............................................................................................................... 8
16.3. USING A MINE CLIMATE SIMULATOR ........................................................... 9

6.3.1. Data preparation ....................................................................................................................... 9
16.3.1.1. Physical description of airway .......................................................................................... 9
16.3.1.2. Condition of airflow at inlet.............................................................................................. 10
16.3.1.3. Thermal parameters and other heat sources ................................................................. 10
16.3.1.4. Physiological parameters (Worker data) ........................................................................ 11
16.3.2. Correlation tests.................................................................................................................... 11
16.3.3. Case Study ........................................................................................................................... 12
16.3.4. Organization of mine climate simulation exercises............................................................... 18
16.3.5. Interaction with Ventilation Network Analysis packages....................................................... 18
Bibliography........................................................................................................... 20
16.1. BACKGROUND
The complexities of the relationships that govern heat flow from the strata into ventilated underground
openings are illustrated by the analyses given in Section 15.2. Indeed, the routine use of those
relationships became practical only through the availability of computer assistance. The first computer
programs to simulate heat flow into mine workings were developed in South Africa (Starfield, 1966).
The early programs estimated strata heat flow from the Goch and Patterson tables (Section 15.2.6.),
either by interpolation or from regression fitted equations that approximated those tables. Since that
time, simulation programs of increasing sophistication have been developed in a number of countries.
Current programs recognize the influence of boundary layers close to the rock/air interface, allow for
heat sources other than the strata (Section 15.3) and predict the psychrometric effects of heat and
moisture additions on the mine climate and physiological effects on workers.
A common feature of mine climate simulation models is that they are based on solutions of the
fundamental equation for heat conduction (equation (15.13)), and on utilization of the dimensionless
Fourier and Biot Numbers. However, the programs may vary in the manner in which they determine
rock surfaces temperatures, heat transfer coefficients, and in the characterization and treatment of
wet surfaces (Mousset-Jones, 1988).
Individual programs may be constrained to particular geometries of stopes or working faces, while
others have been written for airways or headings. Some involve empirical relationships that are
applicable only to specified layouts or methods of working. Again, some program packages allow
combinations of airways, headings and working faces within a network structure while others are
essentially "single airway" simulators that must either be run separately for each branch or used in
conjunction with a ventilation network analysis package (Section 16.3.5).
This chapter is divided into two main sections. Section 16.2 outlines the logical operations of the
CLIMSIM (climatic simulation) program package and is written for researchers involved in further
program development or those who seek an understanding of the numerical procedures used in the
simulation. Readers who are interested only in the practical utilization of a mine climate
simulation package should turn directly to Section 16.3.
16.2. ELEMENTS OF A MINE CLIMATE SIMULATION PROGRAM

16.2.1. Organization of the programs
All mine climate simulations commence with the initial psychrometric condition of the air at the inlet
end of the airway (or face) being defined by the user. This is normally accomplished by specifying the
inlet wet bulb temperature, dry bulb temperature and barometric pressure. Under user control, the
program divides the airway into incremental lengths, each of which is sufficiently short that wet and
dry bulb temperatures may be assumed to be constant within the increment for the calculation of
strata heat flows.
Each increment is traversed in the direction of airflow and the following parameters are calculated:
sensible and latent heat flows from the strata and other sources
change in moisture content of the air
change in dry bulb temperature
conversions between potential and thermal energies for shafts or inclined openings
(autocompression)
change in barometric pressure
change in wet bulb temperature
other psychrometric parameters and indices of heat stress at the exit end of the increment
The conditions for the start of the next successive increment are then defined. Each incremental
length is treated in this way until the complete airway has been traversed. The following subsections
outline the computational procedures involved in each of the steps listed.
16.2.2. Incrementation of airway length
The length of airway increment, Yi, over which changes in temperature have no significant impact on
strata heat flux will vary according to the magnitude of heat additions, the airflow and inclination of the
airway. The value of Yi may be (i) fixed at some small value (say 2 to 20 m) within the program, (ii) a
fixed fraction of the total length or (iii) chosen by the user.
16.2.3. Heat additions
In any incremental length of airway there will, in general, be transfers of both sensible and latent heat
from one or more sources. In order to determine the corresponding psychrometric changes in the
airflow, the sensible and latent heat components are each accumulated separately.
Strata heat
Let us consider the rock surface to be divided into wet and dry areas. We can then define a wetness
fraction, w, as that fraction of total surface area that is covered or coated with liquid water. The
concept of wetness fraction is discussed further in Section 16.3.1.1.
The area of wet surface within the incremental length Yi now becomes
Aw
m2
= 2 r a Yi w
(16.1)
where ra = the effective radius of the airway (perimeter/ 2)

The remaining dry surface area is
Ad
m2
= 2 r a Yi (1 w )
(16.2)
Sensible heat transfer will take place on both the dry and wet surfaces while latent heat transfer
occurs at the wet surface only. The normalized strata heat flows (per square metre) may be
calculated using the methods described in Sections 15.2.8 and 15.2.9.
Let us denote these heat flows as follows:
sensible heat from dry surface:
sensible heat from wet surface:
latent heat from wet surface:
qsen,d
qsen,w
qL,w
(W/m2)
(W/m2)
(W/m2)
The strata heat flowing into the increment, Yi, is then the combination of:
dry surface: qsen,d Ad
2 r a q sen,d Yi (1 w ) W
(16.3)
wet surface: qsen,w Aw
2 r a q sen,w Yi w
(16.4)
wet surface: qL,w Aw
2 r a q L,w Yi w
(16.5)
where A = surface area (m)

and
subscripts w and d refer to wet and dry areas respectively.
Machines and other sources of heat
Here again, heat additions from operating equipment must be separated into components of sensible
and latent heat. In the case of electrical plant, all of the power supplied may be considered as a
sensible heat addition to the airflow, excepting any work that is done against gravity (Section
15.3.2.1). The user must supply two values representing
(a) the full power rating (kW) of the motor or device, FPR, reduced, if necessary, by the rate of
work done against gravity, and
(b) the machine utilization factor, MUF, defined as that fraction of time over which, if the machine
were running at full load, would consume the same amount of energy as the actual
intermittent operation of the device. Hence, the machine utilization factor for a motor running
continuously at full load would be 1.0.
The value of sensible heat produced by the electrical device is then given as
FPR x MUF
(16.6)
Diesel engines produce both sensible and latent heat (Section 15.3.2.2.). Here again, the user must
supply a full power rating, FPR and a machine utilization factor, MUF, for each piece of diesel
equipment. The average machine load, given as the product FPR x MUF (kW), is then converted to
fuel consumption (litres/hour) using a mean empirical value (typically 0.3 litres of fuel per kW engine
rating per hour). The total heat produced is then simply the fuel consumption multiplied by the calorific
value of the fuel (typical value 34 000 kJ/litre).
Using the values quoted, the fuel consumed, FC, becomes

FC = FPR MUF
0. 3
3600
(16.7)
litres/s
and the total heat produced = FC x 34 000 x 103
(16.8)
The user may, of course, specify the rate of fuel consumption directly rather than through the factors
on the right side of equation (16.7). However, practical experience has indicated that most mine
ventilation engineers find it more convenient to assess machine rating and utilization data than to
acquire fuel consumption for individual diesel units.
In order to separate out the amount of water vapour produced by the diesel engine, the user must
supply a third item of data, namely, the water/fuel ratio, WFR, defined as the litres of water (liquid
equivalent) produced for each litre of fuel consumed. The combustion of one litre of diesel fuel will
produce between 1.1 and 1.5 litres of water. However, this may be multiplied several times by engine
cooling systems and exhaust scrubbers in addition to enhanced evaporation from rock surfaces in the
immediate vicinity of the machine. Values as high as 9 litres of water per litre of fuel have been
reported (Mousset-Jones, 1987).
Water vapour is then added to the airstream at a rate of
FC WFR
i.e.
litres of fuel litres of water

s
litre of fuel
(16.9)
kg of water vapour per second
The equivalent value of latent heat is given as

FC x WFR x L
(16.10)
where L = latent heat of evaporation of water in J/kg

while the sensible heat produced by the diesel becomes (total diesel heat - latent diesel heat)
from equations (16.8 and 16.10) respectively.
Compressed air equipment adds no overall heat to a ventilating airstream and, indeed, can produce a
small net cooling effect (Section 15.3.2.3). Such equipment is usually ignored in mine climate
simulations.
Other sources of heat (Sections 15.3.3 to 15.3.8) may be specified by the user. In the general
purpose climate simulators, it is left to the user to select the sensible and latent heat components.
However, special purpose programs have been developed to handle cases such as ducts, pipes,
cables, water channels or spray chambers located within the airway. Air coolers may be specified
simply as negative heat sources.
A climate simulation program will compute the components of strata heat for each incremental length
of airway. In the case of machines or other sources of heat, the user must identify the locations of
those machines or other sources. This can be accomplished by specifying either (a) a spot source, or
(b) a distributed source.
A spot source will produce its heat at a fixed location and, hence, will appear in a single element of
airway length. This is the type of heat source chosen for stationary pieces of equipment such as
transformers, conveyor gearheads or mobile devices that move over short distances only. A
distributed source, as the term implies, will spread its heat load over a distance specified by the user
(e.g. length of a conveyor, hot or cold pipes, drainage channels). In a simple case, the heat will be
distributed linearly over the selected distance. A more sophisticated program may allow the heat to be
distributed according to a function specified by the user.
For each increment of airway length, the simulation program will determine and sum both the sensible
and latent heat components from all sources relevant to that increment. We can now refer to the
corresponding summations as qsen and qL, respectively, for each increment.
16.2.4. Change in moisture content
The apparent density, app, of the air at inlet to the airway is calculated from the psychrometric
condition of the air specified by the user for that starting location
app =
where P
e
d
(P e )
287.04 ( d + 273.15)
=
=
=
kg of dry air
m 3 of air

actual vapour pressure (Pa) and
dry bulb temperature (C)
The mass flow, M, of the dry air component of the air/vapour mixture is then given as
kg dry air
s
M = Q app
(16.11)
where Q is the known rate of airflow at inlet (m3/s)

The value of M remains constant along any given airway.
The rate at which water vapour is added to an increment of airway length is simply
kg
J kg
=
s J
s
where L = latent heat of evaporation (J/kg)
The increase in moisture content of the air becomes
X =
kg/kg dry air
LM
(16.12)
Hence, if we denote subscripts 1 and 2 for entry and exit from the incremental length, then
X 2 = X1 + X
kg/kg dry air
(16.13)
16.2.5. Change in dry bulb temperature and autocompression

In the absence of any fan work the steady flow energy equation (3.25) gives
2
(H 2 H 1 ) q12 =
u1 u 2
2
+ (Z1 Z 2 ) g
J
kg
(16.14)

where H
q12
u
Z
=
=
=
=
enthalpy (J/kg)
heat added (J/kg)
air velocity (m/s) and
However, equation (14.40) shows that the enthalpy term comprises a sensible heat component,
Cpad, and a latent heat component, just as q12 involves both sensible and latent heat. As it is only the
sensible heat component that affects the dry bulb temperature we can subtract the latent heat
component from both the H and q terms. Then equation (16.14) can be re-written as
(C pa ( d ,2 d ,1 ) q12,sen =
u1 u 2
2
+ (Z1 Z 2 ) g
J/kg
where Cpa = specific heat for dry air (1005 J/kg C)

We must remember that q12,sen is the sensible heat addition in J/kg while our earlier sensible heat
summation for the increment, qsen, was in Watts.
=
Hence, q12,sen
J
kg dry air
sen
(16.15)
The increase in dry bulb temperature then becomes
d = ( d ,2 d ,1 ) =
2
2
u1 u 2 + (Z Z ) g +
1
2
sen
1
C pa
C (16.16)
All terms on the right hand side of equation (16.16) are known except the change in kinetic energy.
This is caused only by the change in density in an airway of fixed cross-sectional area and, hence, is
very small. However, it can be estimated as follows:
2
u1 u 2
2
where Q
A
=
=
1 (Q1 Q 2 )
2
A2
volume flowrate (m3/s) and
cross-sectional area (m2)
From equation (16.11), Q = M/app

giving
2
2
u1 u 2
1 M2
1
=
2
2
2
2 A 1,app
2,app
J
kg
(16.17)
A difficulty arises here in that 2,app requires a knowledge of the psychrometric conditions at the exit
end of the increment. These are still unknown. However, as the change in kinetic energy is a very
weak parameter, satisfactory results are obtained by using the change in air density established for
the previous increment in equation (16.17). For the first increment in the airway, the kinetic energy
correction may be ignored without significant loss of accuracy.
Equation (16.16) then enables the change in dry bulb temperature, d, to be determined, having taken
into account both the sensible heat added and the change in elevation. The dry bulb temperature of
the air leaving the increment is simply
d ,2 = d ,1 + d
(16.18)
16.2.6. Change in barometric pressure

In the absence of a fan the variation in absolute pressure of the air, P, is caused by
(a) the conversion of mechanical energy into heat causing a frictional pressure drop, p, and
(b) changes in elevation (Z1 - Z2).
The effect of changes in moisture content on barometric pressure through any single element
remains small provided that the incremental length is not excessive.
The frictional pressure drop is given by Atkinson's equation (5.2)
per 2
Pa
u
A
=
Atkinson's friction factor ( (f)/2 kg/m3) and
=
airway perimeter (m)
p = k Yi
where k
per
(16.19)
This can be converted into the work done against friction

F12
= p / m
(16.20)
J/kg
where m = mean air density in the increment ( kg/m3). Here again, 1 may be used for the first
increment then, subsequently, m extrapolated from the previous increment.
Assuming polytropic flow through the incremental length of the airway, the steady flow energy
equation written in the form of equation (8.1) gives
2
R m (T 2 T1 )
ln(P2 / P1 )
u u2
= 1
ln(T 2 / T1 )
2
+ (Z1 Z 2 ) g F12
J/kg
(16.21)
The gas constant is given as

Rm =
287.04 + 461.5 X
J/kg K
1+ X
[see equation (14.14) ]
where T = absolute dry bulb temperature (d + 273.15 K)

and
X and d are the mean values for the increment.
The outlet pressure, P2, remains the only unknown in equation (16.21) and, hence, may be
calculated.
16.2.7. Change in wet bulb temperature
The psychrometric condition of air is completely defined if any three of its psychrometric properties
are specified. At the exit from the incremental length of airway, the moisture content, X2, the dry bulb
temperature, d,2, and the barometric pressure, P2, have now all been determined. All other
psychrometric parameters can then be calculated from the equations given in Section 14.6.
In particular, the wet bulb temperature, w,2, may be found from the following equations.
X2 =
S 2 1005 d ,2
[Lw ,2 + 1884 ( d ,2 w ,2 ]
kg
kg dry air

where S
= Lw,2 Xs,2 + 1005 w,2
Lw ,2 = (2502.5 2.386 w ,2 )1000

X s,2 = 0.622
and
e sw ,2
(P esw ,2 )
J/kg dry air

J/kg
kg
kg dry air
17.27 w ,2
e sw ,2 = 610.6 exp
237.3 + w ,2
Pa
In this sequence of equations, w,2 is the only unknown independent parameter. Hence, by assuming
an initial value, the equations may be cycled iteratively until a value of w,2 is found that satisfies all
the equations simultaneously.
16.2.8. Relative humidity and saturation conditions
One of the psychrometric variables that may be calculated for each incremental length of airway is
the relative humidity, defined as
rh =
where e
esd
e
e sd
=
=

actual vapour pressure (Pa) and
saturated vapour pressure at dry bulb temperature (Pa)
If the result of this calculation is a value exceeding unity (supersaturation) then this indicates that
condensation will take place. Such condensation will occur on all surfaces with a temperature less
than that of the air wet bulb temperature and also as a fog within the airstream. Indeed, the former will
occur even if the relative humidity is less than 1.0. If the computed psychrometric variables indicate
supersaturation, then the heat of condensation released causes the dry bulb temperature to rise until
saturation is attained at the same level of enthalpy as the original supersaturation condition. This is
approximated closely by setting the revised dry bulb temperature equal to the current value of wet
bulb temperature.
16.2.9 Indices of heat stress
Some mine climate simulators go beyond predicting variations in psychrometric conditions to include
one or more of the indices of heat stress. These fall into two main groups:
(a) indices that can be determined from measured or predicted climatic parameters (e.g. effective
temperature, ET, or wet bulb globe temperature, WBGT). These are sometimes referred to as
empirical indices.
(b) indices that are based on physiological responses of the healthy human body to an actual or
predicted environment for a specified work rate and type of clothing (e.g. Mean Skin
Temperature, MST, sweat rate, or Air Cooling Power, ACP). These are often called rational
indices.
The methods used within mine climate simulators to determine most of these indices are described in
Section 17.4.
16.3. USING A MINE CLIMATE SIMULATOR

The first stage of commissioning a mine climate simulation system is to obtain a program package
that provides the required features. The pointers given in Section 7.4.5 for network simulation
packages apply equally well here. Secondly, for any given airway or network of airways, the data
must be assembled. There is a much greater variety of information required for the prediction of
psychrometric conditions along each airway than for the ventilation network analyses described in
Chapter 7. Furthermore, some of those data may, initially, be of uncertain precision. Hence, the
correlation exercises that must precede employment of a climate simulator for long term planning will
often involve modifications of the initial data.
This section deals with data preparation, correlation trials, the procedures involved in running a mine
climate simulator, and design exercises that can be carried out with the assistance of such program
packages.
6.3.1. Data preparation
The parameters that influence heat flow into a subsurface airway were introduced in Section 15.2.1.
The following subsections deal more specifically with the data that must be quantified for each airway
in preparation for a mine climate simulation.
16.3.1.1. Physical description of airway
Geometry. The parameters required are airway length, cross-sectional area, perimeter, and levels
below a surface datum of both the inlet and exit ends of the airway. It is normally assumed that the
cross-sectional area, perimeter and gradient each remain constant throughout the length of the
airway. Should there be significant deviations in any of these parameters then the airway should be
divided into two or more sub-lengths for simulation.
Friction factor. The roughness of the airway lining is normally quantified as the Atkinson friction factor
(Section 5.2).
Age. The program package may allow the age of each end of the airway to be specified separately.
The age of each simulated increment of airway length is then computed assuming a uniform rate of
drivage. This feature takes into account the time taken to develop the airway and is particularly useful
for advancing headings.
Wetness fraction. This parameter was introduced in Section 16.2.3 as the fraction of airway surface
that is covered or coated with liquid water. However, we have not yet discussed the actual distribution
of water over the surface. Furthermore, although the strata heat flow analyses of Section 15.2
considered both dry and wet surfaces, the solution of the basic equation for heat conduction assumed
radial heat flow. In a simple case of a square airway with a completely wetted floor but dry sides and
roof, the wetness fraction would be 0.25. However, the heat flow through the rock would no longer be
radially symmetric.
The concept of wetness fraction as originally introduced by Starfield in 1966, envisaged the wetness
to be spread uniformly over the surface. With a surface that is damp but not completely covered in
water, asperities on the rough surface may be dry on their peaks but with liquid water in the troughs.
This water may be emitted through capillaries in the rock matrix or from discrete fractures before
spreading out by a combination of surface capillarity and gravitational forces. The temperature of this
type of surface will then take a value lying between those of the completely dry and completely wet
surfaces analyzed in Section 15.2 and, hence, be dependent upon the degree of wetting.
In practice, the wetting of airway surfaces is usually far from uniform. Within a single airway there
may be surface areas varying from completely wetted to visually dry. Mack and Starfield (1983)
proposed an "equivalent wetness factor" for such circumstances, this being defined as the wetness
fraction over a uniformly wetted surface that would give the same rates of heat and moisture transfer
as the actual surface of non-uniform wetness.
Values of wetness fraction can be established for existing airways through correlation exercises
(Section 16.3.2). Furthermore, ventilation engineers who are experienced in running a climate
simulation package become adept at estimating wetness fractions by visual inspection of airways in
much the same way that they can estimate friction factors from the appearance of surface
roughnesses. Wetness fractions seldom fall below 0.04, even in airways that appear quite dry.
For planned but yet unconstructed airways, the wetness fraction may be estimated from previous
experience of mining within that same geologic formation and at similar depths. If no such data are
available then hydrologic studies and the projected use of service water can provide an indication of
the potential average wetness of future workings. For particularly important projects including
underground repositories for hazardous wastes, test drivages within the relevant horizon(s) will
provide invaluable data regarding the migration of moisture from the strata.
16.3.1.2. Condition of airflow at inlet
Most mine climate simulators assume that the airflow and psychrometric condition of the air at inlet
have remained constant since the airway was driven. In the majority of cases this gives acceptable
results because of the thermal flywheel effect (Section 15.2.2). Variations in the surface climate are
usually well damped by the time the air reaches the working areas. If the transients that occur along
intake airways are of concern then methods that employ the principles of superposition are available
for such analyses (Hemp, 1982).
The inlet data required for conventional non-transient simulators are simply airflow, barometric
pressure, and wet and dry bulb temperatures. Satisfactory results for hot mines can be obtained by
assuming mean summer (not extremes) atmospheric conditions on surface.
16.3.1.3. Thermal parameters and other heat sources
The data in this category comprise:
Rock thermal conductivity:
It is the effective thermal conductivity of the strata that should be specified. It is preferable that this
parameter should have been established through in-situ tests as described in Section 15.2.10. If
values are available from laboratory samples only, then corrections may be necessary to obtain the
effective thermal conductivity that actually pertains underground (Mousset-Jones, 1988). Here again,
correlation trials with a climate simulator are valuable in establishing values of effective thermal
conductivity.
Rock thermal diffusivity:
This can be measured directly in-situ as well as by laboratory methods. However, it is normally
satisfactory to determine thermal diffusivity indirectly as
=
where kr
r
Cr
kr
r Cr
=
=
=
m 2 /s
effective thermal conductivity W/(mC)

rock density kg/m3 and
specific heat of the rock J/(kgC)
Virqin rock temperature and qeothermic step (Section 15.2.4):

The combination of VRT at the inlet of the airway and the rate at which rock temperature increases
with depth allows the VRT to be determined for each increment of length along a non-horizontal
10
airway. These parameters may be determined from temperature logs taken from boreholes. In the
case of a downcast shaft, the inlet VRT should actually be the rock temperature sufficiently far below
the surface to be unaffected by variations in the surface climate. This is usually between 10 and 20 m.
Heat sources other than the strata:
The location of every machine or other source of heating or cooling and the distance over which it
extends must be specified. The magnitudes of both the sensible and latent heat components should
be quantified either directly or indirectly by identifying a machine type (Section 16.2.3).
16.3.1.4. Physiological parameters (Worker data)
For mine climate simulators that predict a physiological index of heat stress (see Section 16.2.9), two
more sets of data are requested. One of these includes the metabolic work rate in Watts per square
meter of body surface and the typical posture (body view factor) of a worker. The second set of
physiological factors involves the thermal resistance of the clothing ensemble and the increase in
body surface area associated with that ensemble. Fortunately, all of these parameters can be entered
in a user-friendly fashion simply by choosing from lists of work activities and types of clothing (see
Table 16.1).
16.3.2. Correlation tests
All simulation packages should be subject to correlation tests in order to verify that they do, indeed,
simulate the real system to the required accuracy prior to their being employed for planning purposes.
The accuracy of data must also be verified by site-specific correlation trials. This is true for ventilation
network analysis (Section 9.2.3) and is also the case for mine climate simulation studies.
Due to the uncertainty that may be associated with some of the data, correlation trials of mine climate
simulators should include sensitivity runs involving those input parameters of uncertain accuracy.
Experience has shown that such correlations conducted for a few airways in any mine can not only
highlight previously unidentified sources of heat and humidity but also provide a range of typical data
values that may be tested against other airways until sufficient confidence has been established to
embark upon planning studies.
There is a definite procedure to follow in conducting sensitivity studies and correlations of climatic
simulation output with actual conditions in a mine. The airways chosen for initial correlation should
each be well established and continuous with no intermediate additions or losses of airflow. Any
gradient from horizontal to vertical is acceptable provided that it remains uniform along the length of
the airway. The primary trials should seek to provide correlation between computed results and the
effects of strata heat and, perhaps, autocompression. Additional sources of heat that can be
quantified easily, such as metered electrical equipment, may also be included. However, the initial
trials should avoid airways that contain diesel equipment or open drainage channels. Such sources of
heat should be subject to secondary correlation runs.
Careful measurements of airflow, barometric pressure, and wet and dry bulb temperatures should be
made at the intake end of the correlation airway. Additional wet and dry bulb temperatures should
also be taken at about 100 m intervals along the airway. All of the other parameters required as input
(Section 16.3.1) should be ascertained or ascribed initial estimated values.
In comparing the computed output with observed temperatures, attention should be focused first on
the wet bulb temperatures. If there is a consistent divergent trend between the computed and
measured values then it is probable that a continuous heat source has been over or underestimated,
or perhaps even omitted entirely. A check should be carried out on the depths of the airway ends and
their corresponding ages (if less than two years). Sensitivity runs should also be made to test the
effect of thermal conductivity.
11
If the observed wet bulb temperatures do not show a smooth trend then the reasons for
discontinuities should be investigated. There may be occurrences such as leakage of air or other
fluids from old workings, machine heat, or increased inflow of fissure water. Correlation exercises
provide a valuable educational experience in tracing sources of heat in the mine.
When reasonable correlation (+ 1 C) has been obtained for the wet bulb temperature, attention
should be turned to the dry bulb temperature. Any remaining deviations of this parameter will almost
certainly be due to the evaporation or condensation of water. If the deviation shows a consistent trend
then it is likely that the wetness factor has been wrongly assessed. Sensitivity runs on wetness factor
will test for such a condition. More localised deviations may be caused by inaccurate assessment of
the water vapour produced by diesel equipment, or the effects of dust suppression sprays.
Having followed this correlation procedure over a series of airways, the ventilation engineer will have
built up a store of information on the values and dispersion of heat sources in his mine. He will also
have determined a range of in situ values for thermal conductivities, wetness factors and
contributions of heat and humidity from mechanized equipment. At that stage, forward planning
studies may be initiated with confidence levels that have been established through the correlation
procedures. Climatic simulations then provide an invaluable tool of unprecedented detail in planning
environmental control for hot underground facilities.
16.3.3. Case Study
For clarity of explanation, this case study involves a single airway rather than a sequence of
branches. Figure 16.1 illustrates a 500 m inclined airway descending from 450 to 530 m below a
surface datum. The airway contains a diesel unit of output rating 175 kW working at a mean rate of 60
per cent full load, an air cooler of 200 kW cooling capacity and a conveyor expending 230 kW along
its length from the midpoint to the exit of the airway. The complete data for the airway have been
assembled under the categories given in Section 16.3.1 and are shown in Table 16.1. The order and
format of the input data depend upon the particular software package being used.
530m
450m
Q = 30 m/s
tw = 26C
td = 35C
P = 106 kPa
100m
175m
250m
500m
diesel
175 kW
60%
utilization
air
cooler
200 kW
conveyor
230 kW
Figure 16.1 Ramp used in case study of mine climate simulation.
12
Airway and Rock data

Branch
Name
Dry bulb
at inlet
Wet bulb
at inlet
Pressure
at inlet
Airflow
Length
Depth in
Depth out
c-s Area
Perimeter
C
30
C
24
kPa
106
m3/s
30
m
500
m
450
m
530
m2
8.75
m
12
Textbook
Friction
factor
Wetness
factor
kg/m
0.014
0.2
Age in
Age out
days
days
350
70
Branch heat sources

Spot source
1
Linear
1
Worker data
Clothing
Type of work
C
42
4.5
Sensible heat
(kW)
-200
Rated Power
(kW)
175
Start Location
(m)
250
W/(mC)
40
175
Location
(m)
100
Rock
conductivity
m/C
Location (m)
1
Diesel
Geothermal
step
VRT in
Length (m)
250
Utilization (%)
60
Sensible heat
(kW)
230
Rock
diffusivity
2
-6
Print
interval
m /sx10
2.153
Description
conveyor
slope
Latent heat
(kW)
0
Description
Water output
to fuel ratio
7.0
Description
Latent heat
(kW)
0
Cooler
Diesel No.1
Description
Conveyor
Thin trousers, short-sleeved shirt

Light to moderate manual work
Table 16.1 Input data for the case study as requested by the CLIMSIM mine climate simulator.
The shaded boxes are screen prompts. The values are entered by the user.
The tabulated results of the simulation are shown in Table 16.2. Here again, the form of the output
depends upon the program employed. The CLIMSIM package used for this case study allows any of
the computed variables to be produced in graphical form. Figures 16.2 and 16.3 are computer plots of
the variations in air temperature and relative humidity respectively. The effects of the diesel unit,
cooler and conveyor are shown clearly. Despite the influence of autocompression and strata heat,
evaporation in this rather wet airway results in an initial decrease of dry bulb temperature.
Figures 16.4 and 16.5 show the variation in mean skin temperature of the workforce and the wet bulb
globe temperature respectively. The mean skin temperature is seen to exceed the limit (for the
prevailing climatic conditions) at which workers may begin exhibiting the effects of heat stress. This
occurs at about 440 m from the beginning of the airway. An interesting feature of the mean skin
temperature graph is the increase in its slope at 368 m from the entrance, even though there are no
additional heat sources at or after this location. This is the point at which the average healthy worker
becomes fully covered in perspiration. As his body can no longer increase its cooling by further
wetting of body surface area it attempts to maintain thermal equilibrium by additional increases in skin
temperature (ref. Section 17.3 and Figure 17.6).This effect is not reflected by empirical indices of heat
stress such as wet bulb globe temperature or effective temperature as these are determined solely by
the condition of the airflow and do not indicate physiological response.
The actual reaction of workers on approaching the onset of heat stress or a full coating of perspiration
will be behavioural, i.e. reducing the work rate via less strenuous activity or more frequent rest
periods, and/or discarding clothing. Figures 16.6 and 16.7 show that the mean skin temperature can
be maintained below the relevant limit throughout the airway either by reducing the metabolic work
rate from 200 to 185 W/m2 or by wearing shorts rather than thin trousers.
13

Distance
dry bulb
wet bulb
air
moisture
air
relative
temp.
temp.
pressure
content
density
humidity
kPA
g/kg
kg/m
enthalpy
sigma
dry wall
virgin rock
mean skin
effective
wet bulb
heat
temp.
temp.
temp.
temp.
globe
kJ/kg
Strata heat to air

sensible
transfer
kW
kW
W/(m C)
temp.
m
kJ/kg
heat
latent
coef.
2
30.00
24.00
106.000
15.41
1.207
60.42
69.55
68.01
30.5
42.0
34.04
21.03
25.80
23.619
20
29.91
24.04
106.033
15.50
1.208
61.07
69.69
68.13
30.5
42.1
34.04
20.97
25.80
0.76
-0.42
23.602
40
29.83
24.07
106.067
15.58
1.208
61.71
69.83
68.26
30.4
42.2
34.03
20.92
25.80
0.75
-0.40
23.591
60
29.76
24.11
106.100
15.67
1.209
62.32
69.97
68.39
30.3
42.2
34.03
20.86
25.81
0.73
-0.38
23.581
80
29.69
24.15
106.134
15.75
1.210
62.92
70.11
68.51
30.3
42.3
34.03
20.81
25.81
0.72
-0.36
23.572
100
29.62
24.19
106.167
15.83
1.210
63.50
70.25
68.65
30.2
42.4
34.02
20.77
25.82
0.71
-0.35
23.562
120
33.48
26.36
106.200
17.66
1.194
56.76
78.90
76.95
33.9
42.5
34.78
24.96
28.49
0.92
-0.64
24.112
140
33.34
26.39
106.233
17.76
1.195
57.55
79.03
77.06
33.8
42.6
34.79
24.88
28.47
0.90
-0.62
24.096
160
33.21
26.42
106.266
17.86
1.196
58.32
79.15
77.18
33.6
42.6
34.79
24.80
28.45
0.88
-0.59
24.081
180
27.71
25.02
106.299
17.95
1.218
80.24
73.68
71.80
28.4
42.7
33.93
19.13
25.83
0.39
0.03
23.478
200
27.75
25.07
106.333
18.00
1.218
80.27
73.83
71.94
28.5
42.8
33.94
19.19
25.87
0.39
0.04
23.477
220
27.80
25.11
106.366
18.04
1.218
80.30
73.99
72.09
28.5
42.9
33.95
19.25
25.91
0.39
0.04
23.476
240
27.84
25.15
106.400
18.09
1.218
80.31
74.15
72.24
28.6
43.0
33.96
19.32
25.96
0.39
0.05
23.475
260
28.13
25.26
106.434
18.13
1.218
79.18
74.57
72.65
28.9
43.0
33.99
19.69
26.12
0.41
0.03
23.502
280
28.66
25.44
106.467
18.18
1.216
77.02
75.24
73.30
29.4
43.1
34.06
20.34
26.40
0.45
-0.02
23.554
300
29.17
25.61
106.501
18.23
1.214
75.00
75.92
73.96
29.9
43.2
34.12
20.95
26.68
0.49
-0.06
23.606
320
29.68
25.79
106.535
18.29
1.212
73.11
76.59
74.61
30.4
43.3
34.18
21.53
26.95
0.53
-0.11
23.657
340
30.17
25.96
106.568
18.36
1.211
71.33
77.26
75.27
30.9
43.4
34.23
22.08
27.22
0.57
-0.15
23.707
360
30.65
26.13
106.601
18.43
1.209
69.68
77.94
75.92
31.3
43.4
34.29
22.61
27.48
0.61
-0.19
23.756
380
31.12
26.30
106.635
18.50
1.208
68.12
78.61
76.57
31.8
43.5
34.44
23.11
27.74
0.65
-0.23
23.804
400
31.58
26.47
106.668
18.58
1.206
66.66
79.28
77.22
32.2
43.6
34.62
23.58
28.00
0.69
-0.27
23.851
420
32.02
26.63
106.701
18.66
1.205
65.29
79.96
77.88
32.7
43.7
34.80
24.04
28.25
0.72
-0.31
23.898
440
32.46
26.80
106.735
18.74
1.203
63.99
80.63
78.53
33.1
43.8
34.97
24.48
28.50
0.76
-0.34
23.943
460
32.89
26.96
106.768
18.83
1.202
62.78
81.31
79.18
33.5
43.8
35.14
24.90
28.74
0.79
-0.38
23.989
480
33.31
27.13
106.801
18.92
1.201
61.63
81.98
79.83
34.0
43.9
35.32
25.30
28.98
0.83
-0.41
24.033
500
33.72
27.29
106.834
19.02
1.199
60.54
82.66
80.49
34.4
44.0
35.49
25.69
29.22
0.86
-0.44
24.077
Diesel heat source at 100 m, sensible heat = 147.41 kW, latent heat = 149.74 kW
Spot heat source at 175 m, sensible heat = -200.00 kW, latent heat = 0.00 kW
Linear heat source starts at 250 m, length = 250 m, sensible heat = 230.00 kW, latent heat = 0.00 kW
Strata heat totals: sensible heat = -68.03 kW, latent heat = 164.87 kW
Table 16.2 Tabulated CLIMSIM output for the case study. This is shown
Other heat totals: sensible heat = 177.41 kW, latent heat = 149.74 kW
at 20m intervals for conciseness but is computed at 2m intervals as
Metabolic rate = 200 W/m, Skin temperature limit value = 34.98 C
requested in the input data.
14
36
Temperature deg. C
34
dry bulb
32
30
28
26
wet bulb
24
22
Diesel
Cooler
Conveyor
20
0
50
100
150
200
250
300
350
400
450
500
450
500
Distance m
Figure 16.2 Graphical output for wet and dry bulb temperatures along the airway.
Relative humidity per cent
90
85
80
75
70
65
60
55
50
0
50
100
150
200
250
300
350
400
Distance m
Figure 16.3 Graphical output for relative humidity along the airway.
15
Mean skin temperature deg. C
35.6
35.4
35.2
Limit of mean skin temperature to avoid heat stress
35.0
34.8
34.6
34.4
34.2
Workers become
fully coated in
perspiration
34.0
33.8
0
50
100
150
200
250
300
350
400
450
500
Distance m
Figure 16.4 Graphical output for average mean skin temperature of workers along the airway.
Wet bulb globe temperature deg. C
29.5
29.0
28.5
28.0
27.5
27.0
26.5
26.0
25.5
0
50
100
150
200
250
300
350
400
450
500
Distance m
Figure 16.5 Graphical output for Wet Bulb Globe Temperature along the airway.
16
35.2
35.0
34.8
34.6
34.4
34.2
34.0
33.8
33.6
33.4
0
50
100
150
200
250
300
350
400
450
500
Distance m
Figure 16.6 Mean skin temperature with metabolic work rate reduced to 185 W/m2.
35.2
35.0
34.8
34.6
34.4
34.2
34.0
33.8
33.6
0
50
100
150
200
250
300
350
400
450
500
Distance m
Figure 16.7 Mean skin temperature with workers wearing shorts and short-sleeved shirt.
17
16.3.4. Organization of mine climate simulation exercises

The flow chart given on Figure 16.8 illustrates the procedure for investigating variations in the air
temperatures, humidities and indices of heat stress throughout a ventilation network. From the
results of a ventilation network analysis, one or more routes should be selected through the
system. These may commence at an air intake point on surface, proceed through shafts, intake
airways, work areas and return passages back to surface. Routes should be chosen that may be
expected to produce the severest psychrometric conditions. Climatic simulations may then be
conducted along successive branches throughout the selected routes.
The input shown on Figure 16.8 may be supplied by manual interaction with the climate
simulation package as illustrated in the case study of Section 16.3.3. Alternatively, the input may
be drawn from data files that have been assembled over a period of time, again, either manually
or as a result of previous simulations. Such files may also be used or amended by other program
packages. For example, the airflow for each individual branch may have been determined from a
ventilation network analysis package (Sections 7.4 and 9.1). Similarly, branch inlet temperatures
and pressures will have been produced by climate simulations on upstream airways.
Figure 16.8 indicates that the computed results from a mine climate simulation should be
reviewed with respect to the selected heat stress indices (Chapter 17) or other climatic criteria
relevant to the particular study. Should these criteria not be met then the program should be rerun
successively in order to investigate means of reaching the desired standards. Parameters that
may typically be amended are airflow, airway size or lining, reduction in machine power (for
example, replacing diesels by electrical equipment), the installation, siting and cooling capacities
of heat exchangers or the attire and workrate of personnel. Sensitivity studies should also be run
to test the influence of any input parameters that may be of questionable accuracy.
The convenient editing features of modern climate simulators allow individual items of input data
to be updated with very few keystrokes and the program re-run within a matter of seconds.
There will usually be several ways of meeting the climatic criteria. Again, the procedure may be
cycled successively to investigate alternative arrangements. This sometimes necessitates going
back to re-evaluate an upstream airway.
16.3.5. Interaction with Ventilation Network Analysis packages
Figure 9.1, introduced in Chapter 9 to illustrate the systems analysis of subsurface ventilation
planning, shows the relationships of climate simulations to ventilation network exercises. While
the initial airflows used in climate simulations may have been produced by VNET exercises, it
might have been necessary to amend those airflows during the climate investigations in order to
meet set criteria. The sizes and linings of some airways may also have been altered, affecting
their resistances. In those cases, it becomes necessary to re-run the network analysis package in
order to revise fan duties and other consequential effects throughout the ventilation system.
Where the installation of cooling plant has proved to be necessary during the climate simulation
exercises, the airflow requirements should be reviewed most carefully in order to approach an
optimized combination of ventilating and cooling costs while, at the same time, providing airflows
capable of diluting other airborne contaminants (Section 9.3) (Anderson, 1986).
Programs have been developed that combine the functions of ventilation network analysis and
climate simulation. However the volume and variety of both data and output can become daunting
to the user. Practical experience of ventilation engineers leads to a preference for separate
program packages for network analysis and climate simulation while, at the same time,
18
maintaining the efficiencies of rapid transfer of information between data files. This also allows
manual modification of those files whenever required and establishes firmer control over the
design process by the planning engineers.
Select routes
for climatic
investigations
INPUT
inlet conditions
airway geometries
rock thermal properties
airflows
airway ages (inlet and outlet)
machines
cooling plant
CLIMATE
PROGRAM
Predicted climatic
variations along
selected routes
Establish climatic
criteria: e.g.
WBT, ACP, ET,
WBGT, Skin temp.
climatic
criteria
satisfied?
Amend one or
more variables
NO
YES
further
investigations
required?
YES
NO
Recycle
VNET
exercises
YES
have
airways, airflows
changed?
NO
SYSTEM
DESIGN
Figure 16.8 Climatic planning procedure.
19
Bibliography
Amano, K. et al (1980). Variation of temperature and humidity of air current passing through a
partially wetted airway. Journal Min. Metall, Inst. Japan. Vol.96 pp. 13-17
Anderson, J. and Longson, I. (1986). The optimisation of ventilation and refrigeration in British
coal mines. The Mining Engineer (U.K.). No.146 pp. 115-120
Banerjee, S.P. and Chernyak, V.P. (1985). A simplified method for prediction of temperature
and moisture content changes of air during its passage in a mine airway. Int. Bureau of Mining
Thermophysics, U.K. May Vol.1
Boldizsar, T. (1960). Geothermal investigations concerning the heating of air in deep and hot
mines. Mine and Quarry Eng. June pp. 253-258
Cifka, I. and Eszto, M. (1978). Investigation of the method of determining in-situ characteristics
of underground climatic conditions and the thermal properties of rocks. Report No. 16-5/78.K.
Mining Research Inst. Budapest, Hungary
Gibson, K.L. (1977). The computer simulation of climatic conditions in mines. 15th APCOM
Symp. Brisbane, Australia, pp. 349-354
Hall, A.E. (1975). Computer techniques for calculating temperature increases in stopes and mine
airways. Journal Mine Vent. Soc. of S. Africa. Vol.28. No.4. pp. 55-59
Hall, A.E. (1976). The design and analysis of ventilation systems in deep metal and non-metal
mines. MSHA Contract No. SO231095
Hemp, R. (1982). Sources of heat in mines. Chapter 22 of Environmental engineering in South
African mines. Mine Vent. Soc. of S. Africa.
Hemp, R. (1985). Air temperature increases in airways. Jour. Mine Vent. Soc. of S. Africa.
Lambrechts, J de V. (1967). Prediction of wet bulb temperature gradients in mine airways.
Journal Mine Vent. Soc. of S. Africa. Vol.67. pp. 595-610
Longson, I. and Tuck, M.A. (1985). The computer simulation of mine climate on a longwall coal
face. 2nd U.S. Mine Vent. Symp., Reno, Nevada. pp. 439-448
Mack, M.G. and Starfield, A.M. (1985). The computation of heat loads in mine airways using the
concept of equivalent wetness. 2nd U.S. Mine Vent. Symp. Reno, Nevada pp. 421-428
McPherson, M.J. (1975). The simulation of airflow and temperature in the stopes of S. African
gold mines. Proc. 1st Int. Mine Vent. Congress. Johannesburg pp. 35-51
McPherson, M.J. (1986). The analysis and simulation of heat flow into underground airways. Int.
Journal of Mining and Geological Eng. Vol.4. pp. 165-196.
Mine Ventilation Services. (2006). CLIMSIM for Windows 1.2. Climatic simulation
program user's manual..
Mousset-Jones, P. (1988). Determination of in-situ rock thermal properties and the improved
simulation of the underground climate. Ph.D thesis. Royal School of Mines. University of London,
pp. 553
20
Starfield, A.M. (1966). The flow of heat into the advancing stope. Journal Mine Ventilation Soc.
of South Africa. Vol.19 pp. 13-28.
Starfield, A.M. (1966). The computation of air temperature increases
in advancing stopes. Journal Mine Vent. Soc. of South Africa. Vol.19. pp. 189-199
Voss, J. (1975). Control of the mine climate in deep coal mines. Proc. 1st Int. Mine Ventilation
Congress. Johannesburg, p. 331
Vost. K.R. (1982). The prediction of air temperatures in intake haulages in mines. Journal Mine
Vent. Soc. of S. Africa. Nov. pp. 316-328
Whillier, A. and Thorpe, S.A. (1971). Air temperature increases in development ends. Research
Report 18/71. Chamber of Mines of S. Africa.
21
Chapter 17 Physiological reactions to Climatic Conditions
CHAPTER 17. PHYSIOLOGICAL REACTIONS TO CLIMATIC CONDITIONS

17.1 INTRODUCTION ..........................................................................................2
17.2 THERMOREGULATION OF THE HUMAN BODY........................................2
17.3 PHYSIOLOGICAL HEAT TRANSFER ..........................................................4
17.3.1 The metabolic heat balance.............................................................................................. 4
17.3.2 Respiratory heat transfer .................................................................................................. 7
17.3.3 Convective heat transfer .................................................................................................. 8
17.3.3.1 The convective heat transfer equation ....................................................................... 8
17.3.3.2 Clothing factors .......................................................................................................... 9
17.3.3.3 Convective heat transfer coefficient, hc ...................................................................... 9
17.3.3.4 Mean skin temperature, tsk ....................................................................................... 10
17.3.4 Radiant heat transfer ...................................................................................................... 13
17.3.5 Evaporative heat transfer................................................................................................ 14
17.3.5.1 Evaporation from unclothed and clothed bodies...................................................... 14
17.3.5.2 Skin wetness fraction ............................................................................................... 15
17.3.6 A thermoregulation model............................................................................................... 18
17.4 INDICES OF HEAT STRESS......................................................................22

17.4.1 Purpose and types of heat stress indices. ...................................................................... 22
17.4.2 Single measurements ..................................................................................................... 23
17.4.3 Empirical methods .......................................................................................................... 23
17.4.3.1 Effective temperature, ET ........................................................................................ 23
17.4.3.2 Wet Kata thermometer ............................................................................................. 24
17.4.3.3 Wet bulb globe temperature and wet globe temperature......................................... 25
17.4.4 Rational indices............................................................................................................... 27
17.4.4.1 Air Cooling Power (M scale), ACPM ........................................................................ 27
17.4.4.2 Specific Cooling Power (A scale), ACPA ................................................................. 29
17.4.5 Choice of heat stress index. ........................................................................................... 30
17.5 HEAT ILLNESSES......................................................................................30

17.5.1 Heat fainting.................................................................................................................... 31
17.5.2 Heat exhaustion .............................................................................................................. 31
17.5.3 Heat cramps.................................................................................................................... 31
17.5.4 Heat rash ........................................................................................................................ 31
17.5.5 Heat stroke...................................................................................................................... 32
17.5.6 Precautions against heat illnesses ................................................................................. 32
17.6 COLD ENVIRONMENTS ............................................................................33

17.6.1 Physiological reactions to cold environments ................................................................. 33
17.6.2 Protection against cold environments............................................................................. 34
17.6.3 Indices of cold stress ...................................................................................................... 35
17.7 HEAT TOLERANCE AND ACCLIMATIZATION..........................................36

17.7.1 Physiological tests .......................................................................................................... 36
17.7.1.1 Initial screening ........................................................................................................ 36
17.7.1.2 Heat Tolerance Screening ....................................................................................... 37
17.7.2 Acclimatization ................................................................................................................ 38
17.7.2.1 In-situ acclimatization ............................................................................................... 38
17.7.2.2 Physiological changes during acclimatization.......................................................... 38
17.8 EFFECT OF MINE CLIMATE ON PRODUCTIVITY....................................39

References ........................................................................................................41
APPENDIX A17.1 ...............................................................................................43
Listing of the thermoregulation model developed in Section 17.3.6 .............................. 43
APPENDIX A17.2 ...............................................................................................46

Acclimatization of workers in South African gold mines prior to 1990.......................... 46
17.1 INTRODUCTION
In conventional mining operations, the need to control air temperatures and humidities arises
primarily because of the relatively narrow range of climatic conditions within which the
unprotected human body can operate efficiently. In some areas of underground repositories or
where remotely controlled equipment is in use, a different spectrum of atmospheric criteria may
be allowed or even required. However, in this chapter, we shall concentrate on the effects of
climatic variations on the human body, and alternative means of quantifying the ability of a given
environment to maintain an acceptable balance between metabolic heat generation and body
cooling.
Section 17.3 on Physiological Heat Transfer is fairly comprehensive as this provides the basis of
computer models that predict the physiological reactions of the average acclimatized worker in
hot environments. Readers are encouraged to consult the chapter contents in order to select the
topics of his/her particular interest.
17.2 THERMOREGULATION OF THE HUMAN BODY

Within the human body, chemical and biological processes act upon consumed nutrients to
produce metabolic energy. A little of this is in the form of mechanical energy expended against
external forces. However, the majority of metabolic energy results in the internal generation of
heat. If the body is to remain in thermal equilibrium then this metabolic heat must be transferred
to the ambient surroundings at a rate equal to that at which it is produced. Section 17.3 analyses
and attempts to quantify the various elements of physiological heat exchange, namely, by
respiration, convection, radiation and evaporation, whose net sum must equal the rate of
metabolic heat generation for thermal equilibrium to exist. First, however, let us discuss in general
terms the mechanisms employed by the human body in adjusting to variations in climatic
conditions.
A conceptual model that is commonly envisioned is one in which the body is made up of a central
core of average temperature, tc, surrounded by an outer layer of tissue that has an average
temperature tsk. The latter is referred to simply as the mean skin temperature or MST. Contrary to
popular belief, the core temperature is not constant but varies normally between 36 and 38C
with respect to both location within the body and muscular activity for a healthy individual
experiencing no heat strain., The skin temperature is even more variable depending upon
position on the body surface, clothing, temperature and velocity of the ambient air, rate of
perspiration and, to a lesser extent, metabolic activity, The average temperature and wettedness
of the skin are parameters of major importance in physiological heat exchange. For a resting
person located within an environment that gives no sensation of either heat or cold, mean skin
temperature approximates to 34C. This may be termed the neutral skin temperature.
When the surrounding environment removes heat from the human body at a rate that is not equal
to the metabolic heat generation then one or both of two types of response will occur. A
behavioural response consists of conscious steps taken to alleviate the situation. These might
consist of shedding or donning clothing, and by decreasing or increasing physical activity. The
second type of response comprises the involuntary actions taken by the body to re-establish a
stable heat balance with the surroundings. This is known as body thermoregulation.
Temperature sensitive receptors exist throughout both the core and skin elements of the body.
These react to departures of tc and tsk away from their neutral values. The predominant mode of
heat conductance from the core to skin tissues is by the flow of blood. A 'high" signal from the
core receptors will produce vasodilation, the relevant blood vessels expand to allow a greater flow
of blood and, hence, heat from the core to the skin. Further increases in the high signals from
core receptors will result in perspiration and evaporative cooling, the most powerful mode of
surface heat transfer from the body.
When the skin temperature exceeds its neutral value of some 34C the skin receptors react to
excite the sweat glands into additional activity. However, some areas are more efficient at
producing perspiration than others to the extent that unevaporated run-off may occur in those
zones before other areas are fully wetted. There is also considerable variation in sweat rates
between individuals.
Low signals from the skin receptors (cold conditions) cause vasoconstriction. The blood vessels
that supply skin tissues contract. The vital organs in the core of the body continue to be supplied
by warm blood and are thus protected preferentially while the skin is allowed to cool. This can
cause extremities such as fingers, toes or ears to suffer tissue damage by frostbite in very cold
conditions. Reduced skin temperature may also produce the involuntary muscular movement of
shivering in an attempt to generate additional metabolic heat.
In order to follow the thermoregulatory reactions of the human body to hot conditions, let us
conduct an imaginary experiment in which a volunteer undertakes physical work at a steady but
unhurried rate within an environment that is initially fairly cool, say at 10C. In this situation, heat
loss from the body will occur, mainly, by convection with lesser amounts of cooling by respiration
and radiation. The sweat glands are effectively inactive and evaporation of water by direct
diffusion through the skin contributes only a little evaporative cooling. The procedure of the
experiment is to increase the air temperature in increments, maintaining all other parameters
constant, including the rate of work and, hence, metabolic heat generation. The incremental
changes in air temperature are made sufficiently slowly to allow the body to maintain itself in or
near thermal equilibrium. The purpose of the experiment is to observe the changes that take
place in the modes of heat transfer.
As the air temperature rises from its initial low value, the skin temperature, tsk, also increases in
an attempt to retain the differential between tsk and the temperature of the surrounding
environment and, hence, to maintain convective and radiant cooling. However, the influence of
respiratory cooling decreases because the heat content of the expired air rises at a lesser rate
than that of the inspired air.
At some point, as the air temperature continues to climb and depending upon the rate of work,
signals from the core receptors cause sweating to commence, resulting in partial wetting of the
body surface. This initiates some significant changes in the heat transfer processes. First,
throughout the period of increased sweating and until full surface wetness occurs, evaporative
cooling moderates the rate at which skin temperature continues to rise with respect to ambient
temperature. Also, throughout this stage, the combined effects of convective and radiative cooling
are progressively replaced by evaporative heat transfer. When the skin temperature approaches
and exceeds some 34C, the skin receptors react to produce further enhanced sweating. The
increased area that is coated in perspiration and the corresponding effectiveness of evaporative
cooling both grow rapidly. However, when the skin becomes fully wetted, further cooling can be
gained only by additional rises in skin temperature. Hence, the skin temperature resumes a
higher rate of increase (see Figure 17.6).
If the dry bulb temperature of the surroundings exceeds skin temperature then the convective and
radiant heat transfers will reverse. Evaporative cooling must then balance the heat gain from
convection and radiation in addition to metabolic heat.
When mean skin temperature exceeds about 36C the subject is likely to begin exhibiting the
effects of heat strain (Section 17.5) unless he has been well acclimatized to hot working
environments. At mean skin temperatures above 37C the person is in danger of excessive and,
perhaps, even fatal core temperatures. This might be a good time to terminate the experiment.
17.3 PHYSIOLOGICAL HEAT TRANSFER

A great deal of research has been devoted to quantification of the various modes of heat
exchange associated with the human body and the corresponding physiological reactions. Such
work has involved laboratory and field experiments, analytical studies and numerical simulations.
The development of a numerical thermoregulation model utilizes a combination of relationships
that approximate the modes of heat transfer between the human body and the surrounding
environment. The purpose of such a model is to predict one or more of the parameters that are
deemed to represent physiological reaction such as mean skin temperature, core temperature or
sweat rate. One such model is described in Section 17.3.6. First, however, we must develop
relationships that quantify the individual modes of physiological heat transfer.
17.3.1 The metabolic heat balance
In common with many other living organisms, the human body is a biological heat engine of low
mechanical efficiency. Fuel is consumed in the form of nutrients and combines with oxygen to
produce
metabolic heat
mechanical work and
changes in mass (body growth)
The rate of body growth is negligible compared to other physiological changes and can be
ignored for the purposes of this analysis. The mechanical work output, shown as W on Figure
17.1, is seldom more than 20 per cent of the total metabolic energy even for strenuous efforts by
trained athletes and is usually very much less than this. It is considered only when the work
involves significant lifting or lowering of a mass, including the human body itself.
Metabolic energy = metabolic heat + work done against gravity (W)
[proportional to oxygen consumption]
For example, a person of mass 70 kg walking up an incline of 4 in 100 at a speed of 2 m/s would
do work against gravity at a rate of
70 9.81 2
4
= 55
100
kg
m m
m
= N
= W
2 s
s
s
The metabolic energy produced by the person should be reduced by this amount to give the
corresponding metabolic heat generation. However, in walking uphill, the rate of oxygen
consumption and, hence, metabolic energy is itself increased. In practice, it is usual to ignore
mechanical work in analyses of physiological response to climatic conditions.
At equilibrium, the generation of metabolic heat, M, is balanced by heat transfer from the body to
the surroundings. However, a fraction of the metabolic heat, Ac, may be accumulated within the
body resulting in a transient rise in core temperature, tc, and (usually) also skin temperature. By
assuming that 80 per cent of the body mass is at core temperature, tc, and 20 per cent at skin
temperature, tsk, the rate of heat accumulation may be approximated as
Ac =
m b C b (0.8 t c + 0.2 t sk )
time (seconds )
(17.1)
where mb = mass of body (kg)

and
Cb = average specific heat of the body ( 3470 J/(kgK) ) (Stewart, 1982)
For sedentary persons at thermal equilibrium with the surroundings, average oral temperature is
approximately 36.9C. Skin temperature varies over the body surface and with the ambient air
temperature but has an average value of 34C for non-stressed personnel experiencing a
sensation of thermal comfort. The effect of heat storage, Ac, is to increase core temperature, tc,
and, hence, skin temperature This normally results in enhanced body cooling in order to attain a
new equilibrium at which the rate of heat accumulation declines to zero.
Figure 17.1 shows that heat loss from the body occurs through a combination of heat transfer
processes:
respiratory heat exchange (breathing), Br .

convection, Con .
radiation, Rad, and
evaporation, Evap
Br (breathing)
Fuel
Work
W
Con
(convection)
Evap
Ac
(evaporation
Rad
(radiation)
Figure 17.1 The human body is a biological heat engine producing both work and
metabolic heat, M:
M = Con + Rad + Evap + Br + Ac
Conductive heat transfer may also take place at any areas of contact with solid surroundings.
Here again, this is usually small and is neglected in the following sections. In this analysis we will
also ignore any heating or cooling caused by the ingestion of food or drink. We can now state the
metabolic heat generation, M, to be
M =Br + Rad + Con + Evap + Ac
(17.2)
At thermal equilibrium, the rate of heat accumulation, Ac, is zero.

Then
M = Br + Rad + Con + Evap
(17.3)
This equation is known as the metabolic heat balance. Each of the terms Br, Rad, Con, and Evap
may, theoretically, be negative. However, their net sum must be equal to the rate of internal heat
generation, M, if equilibrium is to be maintained. In this condition, evaporative and respiratory
heat transfer from the body must remain positive while both Rad and Con may be negative.
Physiological heat exchange and metabolic heat generation are usually quoted in terms of Watts
per square meter of body surface (W/m2) where the skin area, Ask, may be estimated from the
DuBois relationship
Ask
= 0.202 m b
=
=
where hb
and
mb
0.425
hb
0.725
m2
(17.4)
height of body (m)

body mass (kg)
For an average sized man of mass 70 kg and height 1.73 m, Ask = 1.83 m2.
Metabolic heat generation depends primarily upon muscular activity but also varies with respect
to the condition of individual health, physical fitness and emotional state. Table 17.1 gives
average metabolic rates for a series of activities both in terms of total energy production (W) and
normalized (W/ m2) on the basis of Ask = 1.83 m2. From this point on we shall refer to the modes
of heat transfer in terms of W/m2.
Table 17.1 Metabolic rates for various activities
Activity
Metabolic heat production

W/m2 (M)
40
58.5*
70
Sleeping
Seated
Standing but relaxed
W
73
107
128
Walking on the level at:

1m/s
1.4 m/s
1.8 m/s
238
320
403
130
175
220
Manual work:
very light
light
moderate
heavy
174
265
448
622
95
145
245
340
* 1 met = 58.5 W/m2 and is sometimes used as a unit of metabolic rate
17.3.2 Respiratory heat transfer

The large internal area and wetness of human lungs give a good efficiency of heat exchange.
The actual degree of respiratory cooling depends upon the rate of breathing and the wet bulb
temperature of the ambient air.
Br =
where mresp
Sout
Sin
m resp (S out S in )
Ask
=
=
=
W/m 2
(17.5)
mass rate of breathing (kg air/s)

sigma heat of exhaled air (J/kg)
sigma heat of inhaled (ambient) air (J/kg)
Experimental evidence (Fanger, 1970) has indicated that the rate of respiration, mresp is
proportional to the total metabolic energy production
m resp
= 1.7 10 6 metabolic energy = 1.7 10 6 M Ask kg/s
(17.6)
Hence, equation (17.5) becomes

Br = 1.7 10 6 M (S out S in )
W/m 2
(17.7)
where the metabolic rate, M (W/m2) may be estimated from Table 17.1.
The sigma heat terms Sout and Sin are calculated from the equations given in Section 14.6 at the
corresponding wet bulb temperatures for exhaled air, tex, and ambient inhaled air, tw, respectively.
The latter can be measured directly. The wet bulb temperature of the near-saturated exhaled air
depends upon body temperature and the psychrometric condition of the ambient air. Again,
empirical data (Fanger, 1970) have produced the relationship
tex = 32.6 + 0.066 td + 0.0002 e
where td
and
e
=
=
(17.8)
ambient dry bulb temperature (C)

actual vapour pressure of the ambient air (Pa)
Example
Estimate the rate of respiratory cooling for a man of average size (Ask = 1.83 m2) walking at 1.8
m/s in an atmosphere of wet bulb temperature, tw = 18C, dry bulb temperature, td = 25C and
barometric pressure, P = 110 kPa.
Solution
From Table 17.1 the metabolic rate, M, is selected as 220 W/m2. Applying the given climatic
conditions to the psychrometric equations of Section 14.6 leads to:
and
Sigma heat, Sin = 47 334 J/kg

actual vapour pressure, e = 1566 Pa.
Then, temperature of exhaled air (equation (17.8))

tex = 32.6 + (0.066 x 25) + (0.0002 x 1566) =
34.563C
Applying this temperature of saturated air to the same equations of Section 14.6 gives the sigma
heat of the expired air to be
Sout = 113 756 J/kg
Equation (17.7) then gives the heat removed by respiration as

Br = 1.7 x 10-6 x 220 (113 756 - 47 334) = 24.8
W/m2
This represents some 11 per cent of the metabolic rate of 220 W/m2. We shall continue with this
example later as we introduce the other modes of heat transfer.
The influence of respiratory cooling decreases as the ambient wet bulb temperature rises towards
that of the expired air. At ambient wet bulb temperatures in excess of tex, condensation will occur
within the lungs and respiratory tracts, accompanied by a rapid onset of heat strain.
17.3.3 Convective heat transfer
17.3.3.1 The convective heat transfer equation
The convective, radiative and evaporative modes of heat transfer from the human body are all
affected by the degree of covering and the thermal properties of clothing. In the case of an
unclothed body, the convective heat transfer is given by equation (15.16)
Con = hc (t sk t d )
where hc
and tsk
W/m 2
(17.8)
convective heat transfer coefficient (W/(m2 C))

skin temperature (C)
=
=
To take the effect of clothing into account, let us consider the heat transfer in two parts:
(i)
(ii)
from the skin at temperature tsk to the outside of the clothing at temperature tcl, and
from the clothing to the surrounding air of dry bulb temperature td.
For each square metre of skin surface, we can write

Con = hcl (tsk - tcl)
where hcl
=
and
Con =
W/m2
(17.9)
effective heat transfer coefficient through the clothing (W/(m2 C))

W/m2
(17.10)
hc fcl (tcl - td)
hc remains the surface convective heat transfer coefficient from the (now clothed) body to the
surroundings (W/(m2 C)) while
f cl
surface area of clothed body

surface area of unclothed body
(17.10A)
The latter factor accounts for the increase in overall surface area caused by clothing.
At equilibrium, equations (17.9) and (17.10) represent the same value of Con.
As tcl is particularly difficult to establish independently, we can eliminate it by re-writing equation
(17.9) as
t cl
= t sk
Con
hcl
(17.11)
For practical application, 1/hcl may be expressed as

R cl
m 2 C/W
= 1 / hcl
(17.12)
and is known as the thermal resistance of the clothing. Substituting for tcl in equation (17.10) and
re-arranging gives
Con =
(t sk
R cl
td )
+ 1 / (f cl hc )
W/m 2
(17.13)
Note that for an unclothed body, Rcl = 0 and fcl = 1. The relationship then simplifies to equation (17.8).
The only variable in equation (17.13) that is amenable to direct measurement is the ambient dry
bulb temperature. The remaining parameters may be estimated from the following empirical data
and relationships.
17.3.3.2 Clothing factors
The effective thermal resistances, Rcl, of clothing ensembles commonly worn in underground
workings are given in Table 17.2 together with typical corresponding area factors, fcl. This table
has been derived from data reported in the ASHRAE Fundamentals Handbook (1985).
Ensemble*
Thermal resistance+
Rcl
C m2/W
Area
factor fcl
hcl = 1/Rcl
W/(m2C)
Unclothed
Walking shorts
Shorts & short sleeved shirt
Thin trousers & short sleeved shirt
Thick trousers & long sleeved shirt
Long sleeved coveralls & long sleeved shirt
0
0.051
0.076
0.085
0.110
0.143
1.00
1.05
1.11
1.14
1.28
1.28
infinite
19.6
13.2
11.8
9.1
7.0
*All ensembles include hardhat, boots and stockings

+
2
Thermal resistance of clothing is sometimes stated in units of clo, where 1 clo = 0.155 C m /W
Table 17.2 Clothing factors for ensembles commonly worn in mining.

The insulation provided by clothing is primarily due to the small pockets of air trapped within it. If
the clothing is fully wetted then heat transfer through it may be estimated as for conduction
through still water in a porous medium. The heat transfer coefficient may then be the order of
some 50 W/(m2 C) and Rcl of 0.02 (m2C)/W.
17.3.3.3 Convective heat transfer coefficient, hc
A number of authorities have reported experimental data and proposed empirical relationships for
the convective heat transfer coefficient applicable to the human body (Stewart, 1982; ASHRAE,
1989). This parameter depends upon the psychrometric condition of the air, the localized
(microclimate) relative velocity, and the size and shape of the body. A simplified relationship that
gives acceptable agreement with experimental data for the velocity range 0.2 to 5 m/s is
hc = 0.00878 P 0.6 u 0.5
where
and
P
u
=
=
W/(m 2 C)
(17.14)

local (effective) relative velocity (m/s)
The local (effective) relative velocity is not the same as the difference between the velocities of
the body and the general airstream. Migration of heat and water vapour creates an envelope
around the body within which the velocity and psychrometric properties of the air differ from those
in the general airstream. This envelope is known as the body microclimate. The velocity within the
microclimate is affected by the corresponding temperature gradient and body movements as well
as the mainstream air velocity. However, a practical approximation may be obtained for the range
0 to 3 m/s as
u = 0.8 ur + 0.6
m/s
(17.15)
where ur = relative velocity between the main airstream and the general motion of the body.
For values of ur in excess of 3 m/s, the correction becomes negligible and we may then accept
that u = ur.
17.3.3.4 Mean skin temperature, tsk
The only parameter in equation (17.13) that remains to be investigated is the mean skin
temperature. This is a factor of prime importance, controlling convective, radiant and evaporative
modes of heat transfer. While neutral skin temperature is of the order of 34C (Section 17.2) it
may rise to 36C in hot conditions without the subject necessarily suffering from heat strain. Well
acclimatized individuals may be able to tolerate skin temperatures of up to 37C without harm.
However, the probability of progressive symptoms of heat strain increases rapidly after skin
temperatures have reached 36C. Skin temperatures of more than 37C are likely to be indicative
of dangerously high core temperatures. A close correlation has been observed between mean
skin temperature and heat strain. Hence, mean skin temperature is one of the physiological
response parameters that may be used as an indicator of heat strain.
Skin temperature may vary considerably over the surface of the body depending upon the
temperature and velocity of the ambient airstream, orientation of the body, the degree of covering
and thermal resistance of clothing, and wetness of the skin.
Figure 17.2 shows an experimental curve obtained by Gagge et al (1969) for unclothed subjects
within an airflow of 40 per cent relative humidity and low velocity. Fanger (1970) and other
authorities have reported some indication that skin temperature reduces as metabolic rate
increases. However, those experiments were related to the measurement of thermal "comfort" in
surface buildings rather than for hot industrial conditions.
The relationship obtained by regression analysis on the curve shown on Figure 17.2 is
tsk = 24.85 + 0.322 td - 0.00165 td2
(17.16)
where td = dry bulb temperature of the ambient air (C).

The value given by this curve or equation has been known as a Base Skin Temperature1.
Base skin temperature is included here as background interest. It is not, however, a required parameter in
the development of a human thermoregulation model (Section 17.3.6).
10
40
Base skin temperature deg C
38
36
34
32
30
28
26
24
22
20
0
10
15
20
25
30
35
40
45
50
Dry bulb temperature deg C

Figure 17.2 Mean skin temperature for unclothed subjects as a function of dry
bulb temperature in air of low velocity and relative humidity of 40
per cent (after Gagge, 1969).
tsk = 24.85 + 0.322 td 0.00165 td2
The reducing slope of this curve as td and tsk increase indicates the moderating effect of
evaporative cooling.
The average skin temperature obtained from equation (17.16) should be corrected for air velocity
and the effects of clothing. As the velocity increases the base skin temperature will, in general,
fall provided that the air temperature is less than that of the skin. The effect is mitigated by the
presence of clothing. It is assumed here that heavy clothing having a thermal resistance, Rcl, of
0.15 (C m2/W) will provide full protection against the effect of air velocity on skin temperature.
The correction to be applied to the low velocity" skin temperature, tsk, of equation (17.16) is
t sk ,vel = 0.009 hc (t sk t d )
(0.15 R cl )
0.15
(17.17)
The heat transfer coefficient, hc, is calculated as a function of air velocity from equation (17.14).
The constant 0.009 has been determined by curve fitting to results reported for unclothed and
fully wetted personnel (Stewart, 1982).
The base skin temperature should be increased beyond the values given on Figure 17.2 because
of the insulation provided by clothing. Gagge (1969) estimated that the skin temperature for fully
clothed bodies was approximately 1C higher than for unclothed subjects. In order to take
account of the degree of thermal insulation, the clothing correction to base skin temperature may
be approximated as
t sk ,cl = +
R cl
0.15
(17.18)
11
Despite the corrections for air velocity and clothing, the skin temperature obtained in this way
remains based upon measurements that were made in a low air velocity and a relative humidity of
40 per cent. This is the reason it is called an initial or base skin temperature that will subsequently
be amended to reflect the response of the human thermoregulation system to the actual
conditions.
Example
In the previous example, a man walking at a speed of 1.8 m/s was found to produce respiratory
cooling of 24.8 W/m2. Let us continue that example to determine the convective heat transfer, Con.
The psychrometric condition of the air given in the earlier example was:
tw = 18C,
td = 25C,
P = 110 kPa
The additional data now supplied are:

(i) The man is clothed in thick trousers, long sleeved shirt and boots. From Table 17.2, the
corresponding values are obtained for
Rcl = 0.110 C m2/W and fcl = 1.28
(ii) The ambient air velocity is 0.5 m/s and the man is walking at 1.8 m/s in the same direction as
the airflow. The relative velocity between the two is, therefore, ur = 1.3 m/s.
Solution
The microclimate velocity is given by equation (17.15)
u = (0.8 1.3 ) + 0.6 = 1.64
m/s
Equation (17.14) then gives the convective heat transfer coefficient to be

hc
= 0.00878 (110 000 )
0 .6
(1.64)0.5
= 11.91 W/(m 2 o C)
In the absence of any skin wetness data (for the moment), we use equation (17.16) to determine
a base skin temperature, i.e. for an unclothed subject in a low velocity air stream
t sk = 24.85 + (0.322 25 ) 0.00165 25 2
= 31.869
Equation (17.17) gives the correction for air velocity to be
t sk ,vel = 0.009 11.91(31.87 25 )
(0.15 0.11)
0.15
= 0.196 o C
The clothing correction is given by equation (17.18) as
t sk ,cl
0.11
= 0.733
0.15
The corrected base skin temperature now becomes
12
t sk = 31.869 0.196 + 0.733
= 32.406 o C
The convective cooling is then given by equation (17.13)

Con =
(32.406 25 )
(0.11+ 1 / (1.28 11.91))
W/m 2
= 42.2
17.3.4 Radiant heat transfer

The exchange of radiant heat takes place between the outer surface of the body or clothing and
any surrounding surfaces or ambient water vapour. The temperature of the outer surface of
clothing is given by equations (17.11 and 17.12) as:
t cl
= t sk Con R cl
(17.19)
This reduces to tsk if the body is unclothed, i.e. Rcl = 0. If the radiant temperature of the
surroundings is tr then the absolute average radiant temperature becomes
Tav
(t cl
+ tr
2
+ 273.15
(17.20)
The Stefan Boltzman relationship (equation (15.27)) then gives the radiant heat transfer
coefficient to be:
hr
= 4 5.67 10 8 Tav
W/ m 2 o C
(17.21)
For most mining circumstances, hr lies within the range 5 to 7 W/(m2 C). The radiant heat
transfer equation (15.26) gives
Rad = hr f r (t cl t r
W/m 2
(17.22)
The emissivity of the body may be taken as unity unless especially reflective clothing is worn.
The view (or posture) factor, fr , has been reported in the range 0.7 to 0.75 for seated, standing or
crouching personnel (ASHRAE, 1989; Stewart, 1982).
Example
Continuing again with the previous example, the man is walking along an airway where the
temperature of the rock walls is 26C. Calculate the radiant heat exchange.
Solution
From equation (17.19) the temperature of the outer surface of the clothing is
t cl
= 32.406 42.2 0.11 = 27.76 C
As tr = 26C, the absolute average radiant temperature (equation (17.20)) is

Tav
(27.76 + 26 ) + 273.15
2
= 300.03
13
Equation (17.21) then gives the radiant heat transfer coefficient to be

hr = 4 5.67 10 8 (300.03 )
W/(m2C)
= 6.13
Taking the posture factor, fr, to be 0.73 for a walking person, equation (17.22) gives the radiant
cooling as
Rad = 6.13 x 0.73 x (27.76 - 26) = 7.9 W/m2
17.3.5 Evaporative heat transfer
17.3.5.1 Evaporation from unclothed and clothed bodies
An analysis of the heat and mass transfers at a wet surface was given in Appendix A15.4
following Chapter 15. Equation (A15.28) gave the evaporative or latent heat transfer to be
Evap = q L
where he =
=
and
ews =
e =
Lsk =
= he (ews e )
W/m2
evaporative heat transfer coefficient at the outer surface

L
W
0.0007 hc sk
2
P
m Pa
(17.23)
(17.24)
saturated vapour pressure at the wet surface temperature (Pa)

actual vapour pressure in the main airstream (Pa)
latent heat of evaporation of water at wet surface temperature (J/kg)
The ratio he/hc ( known as the Lewis Ratio) is then

he
hc
= 0.0007
Lsk
P
C /Pa
(17.25)
Employing mid-range values of Lsk = 2455 x 103 J/kg and P = 100 000 Pa gives a Lewis Ratio of
0.017C/Pa.
In the case of physiological heat transfer, ews becomes esk, the saturated vapour pressure at the
wet skin temperature. Furthermore, if the skin is only partially wetted then equation (17.23)
becomes
Evap = w he (esk - e)
W/m2
(17.26)
where w = fraction of the surface that is wet (see, also, Wetness fraction, Section 16.3.1.1)
In order to take the effect of clothing into account, we employ the analogy between heat and
mass transfers by expanding equation (17.26) to the same form as equation (17.13)
i.e.
Evap =
where Re,cl =
and
fcl
w (e sk e )
R e,cl + 1 / (f cl he )
W/m2
17.27)
the resistance to latent heat transfer through the clothing, (m2 Pa/W), and is
analogous to the thermal resistance, Rcl
the clothing area ratio [Equation (17.10A)].
14
In order to evaluate Re,cl, let us recall from equation (17.12) that Rcl = 1/hcl. We might expect,
therefore, that the heat and mass transfer analogy would give Re,cl = 1/he,cl where he,cl is the
evaporative heat transfer coefficient through the clothing. Practical observations on sweating
mannikins have determined that Re,cl is proportional but not equal to 1/he,cl for any given clothing
ensemble. We can write
R e,cl
i cl
1
he,cl
m2 Pa/W
(17.28)
where icl may be called the vapour permeation efficiency of the clothing. This varies from 0 for a
completely vapour proof garment to 1.0 for an unclothed person. Oohori (1984) reported that
most indoor clothing ensembles have icl values of approximately 0.45. Direct experimentation with
sweating mannikins or other means have been employed to determine icl values for a range of
clothing types. Most of the icl values published by ASHRAE (1993) are in the range 0.35 to 0.41.
The evaporative heat transfer coefficient for the outer surface of the body or clothing was given
by equation (17.24). By replacing the coefficients he and hc by their counterparts for evaporative
heat transfer through the clothing, we obtain
he,cl
= 0.0007 hcl
Hence, as hcl =
or
1
R cl
R e,cl
R e,cl
Lsk
P
and he,cl =
W/(m2 Pa)
(17.29)
1
R e,cl i cl
R cl P
0.0007 Lsk i cl
hc R cl
he i cl
m 2 Pa
W
(17.30)
Substituting into equation 17.27 (from equation (17.30)) in order to eliminate Re,cl gives
Evap =
w he (e sk e )
h R cl + 1
c
i cl f cl
W/m2
(17.31)
17.3.5.2 Skin wetness fraction

The skin wetness fraction, w, or "sweat fraction is a measure of the proportion of skin surface
area from which evaporation of sweat takes place. Diffusion of water vapour through the skin
occurs at a rate normally equivalent to about w = 0.06 and this may be taken as a base value.
(Individuals who are dehydrated or who are acclimatized to living in arid regions may reduce this
base value to about 0.03).
Signals from temperature receptors within the body activate sweat glands in order to increase the
size of the wetted area. The sweat rate is closely related to the core temperature and mean skin
temperature. Each of these three may be employed as an indicator of potential heat strain.
Furthermore, as core temperature is affected strongly by metabolic energy, M, it follows that
sweat rate must also be influenced by the metabolic rate.
15
An analysis of the effective limits of evaporative cooling given in the ASHRAE Handbook on
Fundamentals (1985) leads to the estimated curves shown on Figure 17.3. These provide a base
wetness fraction for any given metabolic heat generation and ambient wet bulb temperature.
Figure 17.3 Base wetness fraction as a function of ambient wet bulb temperature and
metabolic rate.
The curves converge to give a fully wetted body surface (w = 1) at a wet bulb temperature of
33C, irrespective of the metabolic heat production. The curves are described numerically by the
algorithm contained within the Sweat Fraction subroutine of the computer program listed in
Appendix A17.1 at the end of this chapter.
There is, however, an additional effect when the temperature receptors indicate a high skin
temperature. Figure 17.4 illustrates the value to be added to the base wetness fraction when skin
temperature exceeds 32.5C. This curve has been obtained from a consideration of sweat
production as a function of the thermoregulatory signals generated by body temperature
receptors. The curve is represented adequately by the equation
16
sin ((31.85 t sk 762) 0.01745 )

+ 0. 5
(17.32)
2
(where the sine function is computed in radians) for skin temperatures between 32.5 and 39.0C.
w sk
The actual value of skin wetness fraction is then calculated as

w = base wetness fraction + wsk
with the maximum value set at w = 1
(17.33)
There is considerable variation in sweat rate between individuals and according to the degree of
acclimatization. Furthermore, the production of perspiration is by no means constant over the
body surface. A run-off of sweat may occur in some zones before others are fully wetted. Light
absorbent clothing may assist in holding and distributing perspiration that might otherwise be lost
as run-off, so improving the evaporative cooling process. This can result in a distinct chilling effect
on wetted skin during rest periods or when entering a cooler or drier atmosphere.
tsk
Figure 17.4 Additional wetness (sweat) fraction caused by high skin temperature.
Example
Let us continue once again with the earlier example to find the evaporative heat transfer. As a
reminder,
metabolic heat generation,
base skin temperature,
ambient wet bulb temperature,
ambient dry bulb temperature,
convective heat transfer coefficient,
thermal resistance of the clothing,
and clothing area factor,
M
tsk
tw
td
P
hc
Rcl
fcl
=
=
=
=
=
=
=
=
220
W/m2
32.406 C
18
C
25
C
110 000 Pa
11.91
W/(m2 C)
0.11
C m2/W
1.28
17
The only additional item of information that we now require is the vapour permeation efficiency of
the clothing, icl. This is taken as 0.45.
Solution
The evaporative heat transfer is given by equation (17.31), i.e.
Evap =
w he (e sk e )
h R cl
c
i cl
+ 1
f
cl
We must determine the values of those parameters in this equation that are not already known.
From the psychrometric equations given in Section 14.6 the saturation vapour pressure at a skin
temperature of 32.406C is calculated as
esk = 4863 Pa
with a corresponding latent heat of evaporation
Lsk = 2425.2 x 103 J/kg
The actual vapour pressure in the ambient atmosphere has already been calculated (in the
respirable cooling example) as
e = 1566 Pa
Equation (17.24) gives the evaporative heat transfer coefficient to be
he = 0.0007 11.91
2425.2 10 3
110 000
= 0.1838
W/ m 2 Pa
The base wetness fraction, w, at a metabolic rate of M = 220 W/m2 and a wet bulb temperature
tw = 18C may be estimated from Figure 17.3 or computed from the Sweat Factor subroutine
given within the program listed in Appendix 17.1 as w = 0.268. Had the base skin temperature
been greater than 32.5C then an addition to the wetness fraction would have been required
from Figure 17.4.
Evap =
0.268 0.1838 (4863 1566 )

= 44.0
(11.91 0.11 / 0.45 ) + (1 / 1.28 )
W/m 2
17.3.6 A thermoregulation model

The analyses in the previous subsections have produced a set of equations that describe the
heat transfers from a human body. A numerical model of the human thermoregulation system
involves organizing those equations into a logical sequence: then, after comparing the net heat
transfer with a known metabolic heat generation, adjust one or more of the physiological
parameters in a manner that simulates the thermoregulatory response of an average, healthy
person. The process is repeated iteratively until either thermal equilibrium is reached (when net
heat transfer equals metabolic heat) or the model indicates untenable conditions of heat strain.
18
It must be appreciated that considerable variation occurs between individuals in their reaction to
climatic variations. Empirical data gained from physiological tests represent average response.
Combined with the approximations that are inherent in the analyses of surface heat and mass
transfer theory, this indicates that the results of any thermoregulation model provide guidelines to
average reaction rather than precise predictions for individual workers.
The thermoregulation model developed here employs mean skin temperature as the physiological
indicator of heat strain. The model calculates an initial or base skin temperature on the basis of
ambient dry bulb temperature from Gagge's equation (17.16) or from Figure 17.2.2 The heat
transfer from respiration, Br, convection, Con, radiation, Rad, and evaporation, Evap, are
calculated from equations (17.7, 17.13, 17.22 and 17.31) respectively. The difference between
net heat transfer and metabolic heat then indicates the heat accumulated, Ac (see equation
(17.2)).
W/m2
Ac = M - (Br + Con + Rad + Evap)
(17.34)
If Ac is positive then heat is temporarily accumulated within the body. The reaction of an actual
person may be to rest, reduce the rate of work, or discard clothing (the behavioural response).
However, the simulation model assumes that the metabolic rate, M, is maintained. The skin
temperature, tsk, and, consequently, wetness fraction are increased as a function of Ac. In the
thermoregulation model given as a program listing in Appendix A17.1, the correction to tsk may be
taken as 0.01 Ac or 0.02 Ac.
The heat transfer equations are re-evaluated iteratively until Ac is within 1 W/m2 of zero. If the
final equilibrium skin temperature is over 36C then the worker may exhibit the progressive
symptoms of heat strain, while skin temperatures in excess of 37C should be considered to be
dangerous.
Example
The example that was built up sequentially through Sections 17.3.2 to 17.3.5 is continued here to
illustrate the reaction of the model and iterative convergence towards thermal equilibrium.
Recalling the results of the earlier calculations, the initial or base skin temperature2 was
calculated to be 32.406C and the heat transfers associated with this skin temperature were
calculated to be
Respiration, Br,
Convection, Con,
Radiation, Rad,
Evaporation, Evap,
Total (net cooling)
=
=
=
=
=
24.8
42.2
7.9
44.0
118.9
W/m2
W/m2
W/m2
W/m2
W/m2
and
The metabolic rate was, however, 220 W/m2. Hence, the initial rate of heat accumulation is
Ac = 220 - 118.9 = 101.1 W/m2
This causes the simulated skin temperature to increase to
32.406 + (0.02 x 101.1) = 34.428C.
The heat transfer equations are then re-evaluated iteratively until Ac approaches zero.
Table 17.3 illustrates the convergence of the model (listed in Appendix A17.1) towards
equilibrium when net cooling equals metabolic rate.
Alternatively, the initial skin temperature may simply be assumed to be the neutral value of 34C
19
Skin
Respiration Convection
Temperature
Br
Con
C
W/m2
W/m2
1
2
3
4
5
32.41
34.43
34.90
34.73
34.80
25
25
25
25
25
42
54
56
55
56
Radiant
Rad
W/m2
8
11
12
12
12
Evaporation Wetness Net Cooling

Heat
Evap
fraction
W/m2
accumulated
W/m2
Ac W/m2
44
106
135
124
129
0.27
0.55
0.68
0.63
0.65
119
196
229
216
221
101
24
-8
4
-2
Table 17.3. Convergence of the thermoregulation model towards thermal equilibrium.

These results show that the person will be perspiring at a medium rate ( w = 0.65) and will attain
an equilibrium skin temperature of 34.8C if he maintains his work rate at 220 W/m2. This skin
temperature indicates acceptable conditions with negligible risk of heat strain.
By running the thermoregulation model listed in Appendix A17.1, the equilibrium skin temperature
for any given combination of psychrometric conditions, air velocity, type of clothing and metabolic
rate can be computed. This may then be used as a measure of potential heat strain. As indicated
earlier, the probability of unacceptable heat strain is small for skin temperatures up to 36C in
acclimatized workers. However, for even greater stringency, Figure 17.5 has been developed
from the work of Stewart (1982) and Wyndham (1974). This gives maximum allowable values of
equilibrium skin temperature for fully wetted and acclimatized persons after four hours of
continuous work if there is to be a probability of no more than 10-6 of dangerous core
temperatures.
Mean skin temperature deg C
37.0
36.5
36.0
35.5
35.0
34.5
34.0
33.5
33.0
80
100 120 140 160 180 200 220 240 260 280 300
Metabolic heat production W per sq. m
Figure 17.5 Limiting values of mean skin temperature corresponding to a one in a million risk of core
temperature rising above 40C in fully wetted acclimatized men after four hours of continuous work.
For the purposes of illustrating the behaviour of equilibrium skin temperature, cyclic runs of the
thermoregulation model produced the results that are depicted on Figures 17.6 (a, b, c) for three
degrees of clothing. In these runs the barometric pressure was maintained at 100 kPa, metabolic
20
rate at 100 W/m2, and the dry bulb temperature and radiant temperature both set at 5C above the
wet bulb temperature. For those specific conditions, the equilibrium skin temperature can be read
from Figure 17.6 for given values of wet bulb temperature, air velocity, and type of clothing.
Figure 17.6
For the specific values M = 100 W/m2, td = tr = tw + 5C and P = 100 kPa,
this figure illustrates the variation of mean skin temperature with ambient wet bulb
temperature on
(a) unclothed, (b) lightly clothed (c) heavily clothed personnel
21
These sets of curves show clearly:

the increases in skin temperature both below and above the zone of evaporative regulation
the levelling of skin temperatures during evaporative regulation (when body wetness is
increasing)
the increased skin temperatures caused by clothing, and
the reduced influence of air velocity as clothing is added.
17.4 INDICES OF HEAT STRESS
17.4.1 Purpose and types of heat stress indices.
It is clear from the preceding sections that the reaction of the human body to climatic variations
involves a complex mix of psychrometric parameters and physiological responses. It is due to
such complexity that heat stress indices rely, directly or indirectly, on data gained from
experimental tests on volunteer human subjects undergoing monitored work rates in a range of
environmental conditions. For more detailed and wider investigations, the availability of high
speed computing allows a thermoregulation model such as that described in Section 17.3.6 to be
employed. This requires a specification of all relevant data pertaining to work rate, clothing and
psychrometric condition of the air.
For less rigorous applications such as routine checking of the cooling power of an existing
environment or for initial conceptual designs, it is convenient to employ some measure of air
cooling power or physiological reaction that can be quoted as a single number or index of heat
stress. For manual use, such an index should be capable of direct measurement, or calculated
easily from only a few types of observation. Over ninety indices of heat stress were
developed during the twentieth century (Hanoff, 1970). This reflects the large number of variables
involved, the complexity of the human thermoregulation system and the range of climatic
conditions under which human beings may be required to work.
Even with similar climatic conditions, work rates and clothing ensembles, differing indices of heat
stress can exhibit substantial variations in their predicted limits of environmental acceptability
(Brake and Bates, 2002). The further reasons for such differences include the variety of
treatments, assumptions and numerical values used by researchers with respect to clothing and
physiological responses. It should also be borne in mind that most research into heat stress in
non-sedentary workers is based on young, healthy and acclimatized personnel. In practice, there
are significant variations between individuals within any given workforce. It is for this reason that
initial screening and heat tolerance tests are advisable for recruits or returnees who will work in
hot conditions (ref. Section 17.7).
We may classify heat stress indices into three types:
single measurements
empirical methods (dependent on measurements of temperature and air velocity)
rational indices (based on the metabolic heat balance [equation 17.2] and physiological
thermoregulation models).
Indices within each of these groups that have been recommended for subsurface ventilation
systems are introduced in the following subsections.
22
17.4.2 Single measurements

No single psychrometric parameter is, by itself, a reliable indicator of physiological reaction. In hot
and humid environments where the predominant mode of metabolic heat transfer is evaporation,
the wet bulb temperature of the ambient air is the most powerful variable affecting body cooling.
Indeed, some mines retain the wet bulb temperature as a sole indicator of climatic acceptability. A
psychrometric (aspirated) wet bulb temperature of 27 to 28C may be employed as a proximate
criterion above which work rates or shift hours should be reduced, while in former years 32C has
been regarded as an upper limit of acceptability for very hot mines.
The second most important variable in hot conditions is the air velocity. Although air velocity, by
itself, gives little indication of climatic acceptability, it can easily be combined with wet bulb
temperature. This may be achieved by measuring the natural wet bulb temperature, that is, the
temperature indicated by a non-aspirated wet bulb thermometer held stationary within the
prevailing airstream.
Dry bulb temperature alone has a very limited role on climatic acceptability in hot mine
environments. However, dry bulb temperatures in excess of 45C can give a burning sensation
on exposed skin facing the airstream. Hot dry environments may also result in skin ailments
(Section 17.5.4). Heating and ventilating engineers concerned with comfort conditions in surface
buildings often employ a combination of dry bulb temperature and relative humidity.
Unfortunately, workers in underground mines or involved in the development of other subsurface
facilities must often have to contend with much wider ranges of climatic acceptability.
For cold conditions where convection and radiation are the primary modes of heat exchange, dry
bulb temperature and air velocity become the dominant factors. These may be combined into a
wind chill index (Section 17.6).
17.4.3 Empirical methods
These techniques produce indices of heat stress that have either evolved from a statistical
treatment of observations made on volunteers working in a controlled climate or are based on
simplified relationships that utilize measurable parameters but have not been derived through a
rational or theoretical analysis.
17.4.3.1 Effective temperature, ET
Effective temperature is one of the older indices of heat stress. In the mid 1920's, F.C. Houghton
and C.P. Yaglou of the then American Society of Heating and Ventilating Engineers (ASHVE)
conducted a series of experiments in which three volunteers were asked to pass between two
adjoining rooms. One room was maintained at saturated conditions and negligible air velocity
during any given experiment while the wet bulb temperature, dry bulb temperature and air velocity
were varied in the second room. The instantaneous and subjective thermal sensation of each
volunteer on passing between rooms was recorded simply as "hotter, cooler or the same". The
experiments were conducted first on men stripped to the waist then repeated with the same
subjects wearing lightweight suits.
The effective temperature was defined as the temperature of still saturated air that would give the
same instantaneous thermal sensation as the actual environment under consideration. Charts
were produced that allowed the effective temperature to be read for given wet bulb temperature,
dry bulb temperature and air velocity. An example is shown on Figure 17.7.
The concept of effective temperature has been employed widely both as a comfort index for office
workers and also as a heat stress index for industrial and military occupations. The effect of
23
radiant heat exchange may be taken into account in the Corrected Effective Temperature in
which the dry bulb temperature is replaced by the globe temperature measured at the centre of a
blackened 150 mm diameter hollow copper sphere.
The use of effective temperature for the prediction of physiological strain has been shown to have
several shortcomings (Wyndham, 1974) and most mining industries have abandoned it. Where
effective temperature remains in use, it is recommended that periods of continuous work should
be reduced at effective temperatures in excess of 28C and work should be terminated if it rises
above 32C.
Figure 17.7 Effective temperature chart (normal scale) for lightly clothed personnel.
17.4.3.2 Wet Kata thermometer
The Kata thermometer appears to have been introduced by Dr. L. Hill in 1916. It consists of an
alcohol thermometer with an enlarged bulb at the base and a smaller bulb at the top of the stem.
In use, the enlarged bulb is submerged in warm water until the alcohol partially fills the upper
reservoir. The instrument is then hung freely in the prevailing airstream and a stopwatch
employed to determine the time taken for the alcohol level to fall between two marks inscribed on
24
the stem. This is combined with an instrument factor given with each individual thermometer to
produce a measure of the cooling power of the air in mcal/(cm2s).
When the main reservoir of the Kata thermometer is covered in wet muslin, the instrument gives
cooling powers that are representative of a 20 mm wetted bulb at 36.5C. These have been found
to correlate reasonably well with core temperatures of unclothed and acclimatized workers
subjected to hot and humid conditions (Stewart, 1982). Wet Kata readings may be approximated
from measurements of air velocity, u, and wet bulb temperature, tw,
Wet Kata cooling power = 0.7 + u 0.5 (36.5 t w )

mcal/(cm2s)
(17.35)
(South African Chamber of Mines, 1972)
The use of Wet Kata readings as an index of heat stress is limited to hot and humid
environments. Only the South African gold mining industry has utilized the Wet Kata reading in
mines as a heat stress index on a routine basis.
17.4.3.3 Wet bulb globe temperature and wet globe temperature
The wet bulb globe temperature (WBGT) relies upon two measurements only. The first is the
reading indicated by a wet bulb thermometer held stationary in the prevailing airstream. This is
sometimes termed the natural wet bulb temperature, tnw, and, at air velocities of less than about 3
m/s, will indicate a temperature greater than that of the normal psychrometric (aspirated) wet bulb
thermometer, tw (Section 14.3.5). Secondly, the temperature at the centre of a matte black hollow
sphere or globe temperature, tg, is measured. The two are combined to give
WBGT = 0.7 tnw + 0.3 tg
(17.36)
When a significant source of radiant heat is visible then

WBGT = 0.7 tnw + 0.2 tg + 0.1 td
where td
dry bulb temperature C
(17.37)
This latter form of the equation is normally employed only in sunlight.

The wet bulb globe temperature is a function of the major climatic parameters that affect
physiological reaction, i.e. wet and dry bulb temperatures, air velocity and radiant temperature.
However, it has the advantage that it does not require a separate measurement of air velocity.
The National Institute of Occupational Safety and Health of the United States (NIOSH, 1986) has
employed WBGT as a heat stress standard and recommended the exposure limits shown on
Figure 17.8. Heat stress levels based on the WBGT have also been subject to an International
Standard (ISO, 1982).
A simplified version of the wet bulb globe temperature was developed by J.H. Botsford in 1971
and is known as the wet globe temperature (WGT). The instrument, now called a Botsball,
consists of a 6 cm diameter blackened copper sphere covered by a wetted black fabric. A dial
thermometer indicates the temperature at the middle of the hollow sphere.
Regression equations may be employed to relate WGT to WBGT (Ciriello, 1977) for known
psychrometric conditions. For the particular range of dry bulb temperatures 20 to 35C, relative
humidities greater than 30 per cent and air velocities less than 7 m/s, Onkaram (1980) has
suggested the experimentally derived equation
WBGT = 1.044 WGT - 0.187
(17.38)
where WGT is measured in C.
25
Figure 17.8 Wet bulb globe temperature exposure for acclimatized workers at varying working
time per hour(recommended by the U.S. National Institute of Occupational Safety
and Health, Pub. 86-113).
The WBGT and its simplified companion, WGT are both well suited to the rapid assessment of
climatic conditions in existing locations where instrumentation may be located. They are, at
present, less useful for the prediction of air cooling power in underground openings that have not
yet been constructed, as neither natural wet bulb temperature nor wet or dry globe temperatures
are included as output in current subsurface climate simulators.
Improvements in instrumentation, microelectronics and computer software have resulted in the
development of hand-held devices that give an in-situ indication of air cooling power (e.g. Wu and
Gillies, 1997).
26
17.4.4 Rational indices

A rational index of heat stress is one that has been established on the basis of the physiological
heat balance equation (17.3) and that conforms to heat transfer relationships such as those
developed in Section 17.3. A thermoregulation model (Section 17.3.6) may be used for detailed
investigations of existing or proposed facilities. The establishment of such a model on a mine
office microcomputer enables it to be employed for day to day routine assessments. However, for
rapid manual assessment or where predictions of average cooling power for a work area are
required, a thermoregulation model can be simplified into charts or tables by establishing specific
values for the weaker parameters, or defining fixed relationships between those parameters and
the more dominant variables.
A choice must be made on the physiological response that is to be used as an indicator of
climatic acceptability. This is usually one of the linked parameters, core temperature, mean skin
temperature or sweat rate. In the thermoregulation model output illustrated in Figure 17.6, mean
skin temperature is employed in association with the limit values of Figure 17.5 as the indicator of
acceptability.
17.4.4.1 Air Cooling Power (M scale), ACPM
The illustrative example shown on Figure 17.6 (a) gives a series of curves relating equilibrium
skin temperature to wet bulb temperature and air velocity for unclothed personnel producing 100
W/m2 of metabolic heat, and for specified values of
P
100 kPa
and
td = tr = tw + 5C
(17.39)
Figure 17.5 indicates that at M = 100 W/m2 an equilibrium skin temperature of 35.8C is
associated with a probability of no more than one in a million that core temperature will exceed
40C, i.e. a negligible risk of unacceptable heat strain. Scanning across the tsk = 35.8C line on
Figure 17.6 (a) allows us to read the corresponding limiting wet bulb temperatures for air
velocities 0 to 5 m/s.
If the procedure were repeated with a sufficiently large and detailed number of charts of the type
illustrated on Figure 17.6 then tables relating metabolic heat, wet bulb temperature and air
velocity could be compiled for each clothing ensemble. This exercise has been carried out, not by
a graphical construction of charts, but by cyclic operation of the thermoregulation model
described in Section 17.3.6. The results are shown on Figure 17.9.
The lines on this chart have been established from equilibrium conditions when the ambient
environment removes metabolic heat at the same rate as it is generated. Hence, they may be
regarded as limit lines of climatic acceptability. For this reason, the vertical axis has been
renamed Air Cooling Power (M scale) as a reminder that the lines represent equilibrium with
metabolic heat, M, and to distinguish it from earlier versions of Air Cooling Power.
In order to use the chart as a rapid means of assessing climatic acceptability of any given
environment, the psychrometric wet bulb temperature, tw, and metabolic heat generation, M, are
plotted as a coordinate point. If this point lies above the relevant air velocity line for the chosen
clothing ensemble then the average cooling power of the air is greater than metabolic heat
generation and personnel will be able to attain thermal equilibrium with the environment at that
same rate of work. However, if the air cooling power is less than the relevant limit curve then the
workers will discard clothing, reduce their work rate or risk the onset of heat strain.
In addition to the specified relationships of td = tr = tw + 5C and P = 100 kPa, the ACPM chart is
based on a body posture factor of fr = 0.75 and the following specifications for clothing:
27
Unclothed
2
Light Clothing
Heavy Clothing
thermal resistance Rcl (Cm /W)
0.085
0.143
area factor, fcl
1.14
1.28
vapour permeation efficiency
0.45
0.45
It will be recalled, also, that the air velocity is the relative velocity between the workers body and
the main airstream.
Figure 17.9 Air Cooling Power (M scale) or ACPM chart.

[Based on td = tr = tw + 5C and P = 100 kPa]
As the effects of dry bulb temperature, radiant temperature and barometric pressure are relatively
weak in underground environments, the ACPM chart may be used without undue error for wet
bulb depressions from 2 to 8C and for the range of atmospheric pressures that normally exist in
28
mine workings. The computer model listed in Appendix A17.1 can be used to develop similar
charts for any other specified conditions. Air cooling charts have the advantage that they are
based on a thermoregulation model and can also be used to assess the acceptability of existing
workplace conditions.
Example 1
In a mine working, the wet bulb temperature is 28C and the air velocity is 1 m/s. Estimate the
limiting rates of continuous work for unclothed, lightly clothed and heavily clothed personnel.
Solution
From Figure 17.9 the three values of air cooling power at tw = 28C and u = 1 m/s are estimated
at:
lightly clothed: 168 W/m2
heavily clothed: 127 W/m2
unclothed: 313 W/m2
Table 17.1 may be used to relate these cooling powers to types of work activity.
Example 2
It is predicted that in a future subsurface development the air velocity will be 2 m/s. If personnel
are to sustain a metabolic rate of 220 W/m2 while lightly clothed, estimate the upper limit of
psychrometric wet bulb temperature.
Solution
From the light clothing curves on Figure 17.9, at an air velocity of 2 m/s and an air cooling power
of 220 W/m2, the limiting wet bulb temperature is estimated at 25.6C.
Example 3
Acclimatized mine employees are required to work in a location where the wet bulb temperature
is 29 C. The workrate will be moderate with occasional extra effort. Recommend the type of
clothing and the minimum air velocity to minimize the probability of unacceptable heat stress.
Solution
From Table 17.1 an average metabolic heat production rate is estimated at 260 W/m2.
Plotting a point at M = 260 W/m2 and tw = 29 C on the Air Cooling Power (M scale) chart of
Figure 17.9 shows that it lies close to an air velocity of 0.5 m/s in the unclothed region. Hence,
the recommendation is that these workers should wear no more than brief shorts, boots and
hardhats, and that the air velocity should not be less than 0.5 m/s.
17.4.4.2 Specific Cooling Power (A scale), ACPA
The original concept of air cooling power as employed in this chapter arose from pioneering work
conducted by the Chamber of Mines Research Organization of South Africa (Mitchell and Whillier,
1972; Wyndham, 1974; Stewart and Whillier, 1979; Stewart, 1982). This involved detailed
analytical investigations as well as some thousands of tests in a large environmentally controlled
and monitored wind tunnel. A thermoregulation model was developed on the basis of the
physiological heat transfer equation (17.3). That model was restricted to the hot and humid
conditions prevalent in the stopes of deep South African gold mines. In those conditions the
influence of respiratory cooling is relatively small and was ignored. The model was further
confined to fully wetted and unclothed personnel. An air cooling power chart was produced for the
specific conditions of td = tr = tw + 2C and P = 100 kPa and for wet bulb temperatures in the
range 25 to 35C (Stewart, 1982). This was termed the Specific Cooling Power (A Scale) chart
and yielded curves that are comparable to those given for unclothed personnel on Figure 17.9
within that range of wet bulb temperature.
29
17.4.5 Choice of heat stress index.

With such a variety of heat stress indices the question arises on when to employ an empirical or
rational index. It would be impractical to conduct measurements of physiological response (e.g.
body core temperature, sweat rate or mean skin temperature during normal mining operations.
Hence, for routine manual monitoring of the workplace empirical indices are usually preferred by
supervisors (Section 17.4.3)3.The values obtained for the preferred empirical index can then be
compared with limits of acceptability that have been estimated from controlled experiments on
volunteers (e.g. Figures 17.7 and 17.8).
As the reliability and costs of handheld cooling power instruments continue to improve we may
see these being more widely used for routine workplace monitoring. Similar devices might also be
incorporated into minewide monitoring systems for hot mines.
Where computer simulation programs are employed to predict climatic conditions in planned
future workings a wider choice is available. Provided that the relevant variables of environmental
temperatures and air velocities are produced by the simulation package then one or more of the
empirical indices of heat stress can again be shown. However, a thermoregulation model may
also be included in the simulation package in order to predict one or more of the physiological
responses. The latter can be used as a direct and more precise indicator of the level of heat
stress experienced by workers for specified types of clothing and rates of physical work. For
example, in the CLIMSIM package (Chapter 16) the empirical indices produced are effective
temperature and wet bulb globe temperature. Additionally, the thermoregulation model listed in
Appendix A17.1 is incorporated within CLIMSIM in order to predict mean skin temperature as an
indication of physiological reaction and to give warnings of impending heat stress (see Case
Study, Section 16.3.3).
17.5 HEAT ILLNESSES
The human thermoregulation system depends upon the efficient operation of the core and skin
temperature receptors, the flow of blood throughout the body but particularly between the core
and skin tissues, and the production of perspiration. If any of these mechanisms loses its
effectiveness then the body will progressively exhibit the symptoms of one or more of a series of
disorders known collectively as the heat illnesses. These may arise as separate and recognizable
ailments with identifiable causes. However, for workers in hot and humid environments they can
occur in combination.
A common initial symptom is a loss of interest in the task and difficulty in remaining alert. In any
adverse environment, the desire to seek more comfortable surroundings is a psychological
reaction that is just as much a part of the body's defence mechanism as thermoregulatory effects.
Suppression of such predilection may result in irritability or displays of anger. This may be
observed even in persons who are cognizant of the effect. The physical symptoms often reveal
themselves first as a loss of coordination and dexterity.
It follows, even from these initial symptoms of heat strain, that both manual and mental work
productivity will suffer, morale is likely to be low, absenteeism high and standards of safety will
decline in environments that are unduly hot. This is true for either a labour intensive method of
working or one that is heavily mechanized.
However, Air Cooling Power Charts (e.g. Figure 17.9) are also directly applicable to existing workplaces.
30
Heat illnesses are introduced in the following subsections and in order of increasing seriousness.
However, it should be remembered that symptoms may overlap and, in cases of doubt, it should
be assumed that the patient is a victim of heat stroke.
17.5.1 Heat fainting
This occurs most frequently when a person stands still for an extended period in a warm
environment. Blood tends to pool in the lower parts of the body causing a temporary reduction in
blood supply to the brain and, hence, a short term loss of consciousness. A common example
arises when soldiers in heavy uniform are required to remain motionless for long periods of time
during ceremonial parades.
The treatment is simply to restore an adequate supply of blood to the brain by allowing the patient
to lie flat in a cooler area and to loosen or remove clothing. Recovery is normally rapid and
complete. The probability of heat fainting is reduced by intermittent body movements.
17.5.2 Heat exhaustion
Conductance of heat within the body is facilitated primarily by the flow of blood. If the volume of
blood is insufficient then heat exhaustion may ensue. A decrease in blood volume may result
from dehydration caused either by an inadequate intake of fluids or by a salt deficient diet.
Alternatively, a combination of environmental heat stress and metabolic rate may cause the
heartbeat to exceed some 180 bpm. The inadequate time interval between contractions of heart
muscles may then be insufficient to maintain an adequate supply to the heart chambers - the rate
of blood flow drops. A significant increase in total volume of blood occurs during periods of heat
acclimatization.
The additional symptoms of heat exhaustion are
tiredness, thirst, dizziness
numbness or tingling in fingers and toes
breathlessness, palpitations, low blood pressure
blurred vision, headache, nausea and fainting
clammy skin that may be either pale or flushed.
On the spot treatment should include removal to a cool area and administration of moderate
amounts of drinking water. If the patient is unconscious then heat stroke should be assumed. In
any case, a medical examination should be carried out before the victim is allowed to return to
work.
17.5.3 Heat cramps
If the electrolytic balance of body fluids is sufficiently perturbed then painful muscular contractions
occur in the arms, legs and abdomen. This may occur by salt deficiency or drinking large
amounts of water following dehydration. Medical opinion no longer favours salt tablets but
recommends a diet that provides a more natural supply of salt. However, immediate treatment
may include the administration of fluid containing no more than 0.1 per cent salt.
17.5.4 Heat rash
This is sometimes known as prickly heat by residents of equatorial regions. It is caused by
unrelieved periods of constant perspiration. The continuous presence of unevaporated sweat
produces inflammation and blockage of the sweat ducts. The typical appearance of the ailment is
31
areas of tiny red blisters causing irritation and soreness. Although not serious in itself, heat rash
may lead to secondary infections of the skin and, if sweat rates are sufficiently inhibited, may
produce an increased susceptibility to heat stroke. Heat rash can be prevented by frequent
bathing and the provision of cool living quarters.
Skin irritations can also occur in the hot, dry environments of some evaporite mines, especially at
areas of skin contact with hat bands or protective devices.
17.5.5 Heat stroke
The most serious of the heat illnesses, heat stroke, may occur when the body core temperature
rises above 41C. Co-ordination of the involuntary nervous system, including body
thermoregulation, is achieved by means of proteins controlled by the hypothalamus, part of the
brainstem. At temperatures in excess of 41C, damage to those proteins results in the
transmission of confused signals to the muscles and organs that control involuntary reactions.
Furthermore, irreversible injury may be caused to the brain, kidneys and liver. Heat stroke carries
a high risk of fatality.
The initial symptoms of impending heat stroke are similar to those of the less serious heat
illnesses, i.e. headaches, dizziness, nausea, fatigue, thirst, breathlessness and palpitations. If the
patient loses consciousness, then heat stroke should be assumed. The additional symptoms
arise from disruption of the thermoregulatory and other nervous systems of the body:
perspiration ceases, the skin remains hot but is dry and may adopt a blotchy or bluish
colouration.
disorientation may become severe involving dilated pupils, a glassy stare and irrational or
aggressive behaviour.
shivering and other uncontrolled muscular contractions may occur.
there may be a loss of control of bodily functions.
Although treatment should commence as soon as possible, current medical opinion is that the
patient should not be moved until core temperature has been brought under control. The
administration of first aid should include:
Cooling by air movement and the application of water to the skin. The water temperature
should be not less than 15C in order to avoid vasoconstriction of the skin blood capillaries.
Massaging the skin also assists in promoting blood flow through the surface tissues.
The application of CPR (cardiopulmonary resuscitation) in cases where perceptible heart

beat or respiration are absent. This includes chest massage and mouth-to-mouth breathing.
Frequent sips of water if the patient is conscious.
Medical assistance should be sought and treatment continued until the core temperature falls
below 38.5C. The patient should then be covered by a single blanket and transported by
stretcher to hospital. Unfortunately, death or permanent disability may still occur if irreversible
damage has occurred in vital organs. Even after physical recovery, the heat tolerance of the
victim may be impaired, perhaps permanently.
17.5.6 Precautions against heat illnesses
There is a great deal that can be done to protect a workforce against unacceptable heat strain.
First, the design and engineering control of the ventilation and air quality systems of the mine or
facility should be directed toward providing a safe climatic environment within all places where
32
personnel work or travel routinely. If threshold limit values for climatic parameters have not been
set by legislative action, then they should be chosen and enforced by self-imposed action of the
mine or industry. Cool and potable drinking water should be available to workers in hot mines.
Clothing that is appropriate for the work environment and level of activity should be worn by
personnel. In hot environments, this may mean very little clothing or lightweight garments that are
loosely weaved, porous and absorbent. For unusual circumstances such as rescue operations
where personnel are occasionally required to enter hot areas that do not meet the criteria of
climatic acceptability, microclimate protective garments may be used as a short-term and
temporary measure. Such devices typically take the form of specialized jackets that contain
distributed pockets to hold solid carbon dioxide (dry ice), cooled gels or water ice. The
microclimate garments are worn over underclothing in order to avoid ice burns. Outer clothing
reduces direct heat transfer from the surroundings. It is important that this type of microclimate
jacket is changed before the cooling medium is exhausted, as the garment may then become an
effective insulator. More sophisticated microclimate suits, incorporating temperature controlled
fluids circulating through capillary tubing, have also been manufactured. However, these are
expensive and less applicable to mining environments.
A workforce that is required to work in hot environments should be trained to recognize the initial
symptoms of heat strain and to adopt sensible work habits. The latter include appropriate, but not
excessive, rest periods and adequate consumption of cool drinking water. A good dietary balance
should be sought that includes sufficient but, again, not excessive salt. Vitamin C
supplementation is also regarded as improving heat tolerance (Visagie, 1974).
Finally, workers should have been subjected to a process of natural or induced acclimatization
prior to being employed in particularly hot conditions (Section 17.7).
17.6 COLD ENVIRONMENTS
This chapter is concerned primarily with physiological reactions to heat. However, personnel at
mines or other facilities that are located in cold climates or at high altitudes may be subjected to
low temperatures for at least part of the year. Even in the more temperate zones, workers in main
intake airways may suffer from cold discomfort during winter seasons. It is, therefore, appropriate
that we include the effects of cold environments on the human body.
17.6.1 Physiological reactions to cold environments
Heat loss from the clothed body in cold surroundings is mainly by convection with lesser amounts
by radiation and respiration. The most important climatic parameters in these circumstances are
ambient (dry bulb) temperature and air velocity. These may be combined into an equivalent wind
chill temperature (Section 17.6.3).
As heat loss occurs, the initial behavioural response is to don additional clothing and to increase
metabolic heat production by conscious muscular activity. Involuntary physiological reaction is
initiated by reductions in either the body core temperature or mean skin temperature and consists
of an increase in muscular tension. In skin tissues, this causes the familiar "gooseflesh and
progresses into shivering within localized muscle groups. This may result in metabolic heat
generation being raised by up to 120 W/m2. With further cooling, the degree of shivering
increases to encompass the whole body, producing as much as 300 W/m2 of metabolic heat and
effectively incapacitating the person.
33
If the core temperature falls below 35C, the

body thermoregulation system will be affected.
Mean skin
Subjective response
Core temperatures of less than 28C can prove
temperature C
fatal, although successful recovery of
31
uncomfortably cold
individuals has been achieved from core
temperatures of less than 20C. The subjective
30
shivering
feeling of comfort depends upon the mean skin
temperature and also the surface temperature
29
extremely cold
of the extremities. Commencing from the
neutral mean skin temperature of 34C, this
25
limit of tolerance without numbing
table shows subjective reactions that have been
reported for sedentary persons. However, the
increased blood flow caused by a medium or hard work rate may enable reduced values of mean
skin temperature to be tolerated.
The large surface to volume ratio of fingers, toes, and ears maximizes heat loss while the
influence of vasoconstriction is particularly effective in reducing blood flow to those areas. The
latter is a natural reaction to protect, preferentially, the vital organs within the body core.
However, it can give rise to severe discomfort and tissue (frost) damage to the extremities. The
following reactions to hand surface temperatures have been reported:
20C
16C
15C
5C
uncomfortably cold
limit of finger dexterity
extremely cold
painful
Skin surface temperatures some 1.5 to 2.0C higher produce similar reactions at the feet.
17.6.2 Protection against cold environments
Human beings have been described as tropical animals inasmuch as our thermoregulation is
much more effective in warm environments. In cold conditions we rely more on the behavioural
response of donning extra clothing and/or engaging in additional muscular activity.
The first line of defence against any potentially adverse environment in the subsurface is the
initial design and subsequent control of the ventilation and air conditioning systems. Methods of
heating mine air are discussed in Section 18.4. These are subject to economic limitations and can
lead to problems of strata control if the mine is within permafrost. It may, therefore, be necessary
to provide means of personal protection against a cold working environment.
Personnel should be instructed on physiological reactions to cold ambient surroundings and,
particularly, on the use of protective clothing. The thermal insulating properties of garments
depend upon the entrapment of air rather than the fibres of the material. The incorporation of
radiation reflective layers can also be helpful. A compromise must be sought between the body
movements required to perform the work and the restraint offered by bulky insulated clothing. An
ambient temperature of -35C appears to be the practical limit for many activities while, at -50C
only short term exposure should be permitted. Protection to the hands and feet is particularly
important but does, of course, affect dexterity. Auxiliary heating of some 10W to each hand and
foot allows activity to continue at even lower temperatures provided that the rest of the body is
well insulated.
Facial tissues do not suffer unduly from vasoconstriction. Hence, face masks are usually
unnecessary. However, the effects of wind chill can result in freezing of exposed skin. Fur-lined
hoods should be used to avoid direct wind on the face at low ambient temperatures. Pre-heating
inspired air is not required for healthy individuals for air temperatures down to -45C.
A limited degree of cold acclimatization can occur in personnel who have been subjected to
frequent exposure. An improved blood flow to skin tissues gives increased circulation to the
34
extremities and reduces the risk of frostbite. Additional heat is also generated by the metabolism
of body reserves of adipose (fatty) tissue.
17.6.3 Indices of cold stress
A series of experiments were carried out by Siple and Passel (1945) in Antarctica involving rates
of cooling by convection and radiation from a plastic cylinder of 5.7 cm diameter whose surface
temperature was 33C. This led to an empirical formulation known as the Wind Chill Index, WCI,
which described the rate of heat loss from the cylinder. When expressed in SI units, this becomes
WCI = 1.163 10.45 + 10 u u (33 t d )
W/m2
(17.40)
where u = air velocity (m/s) and td = dry bulb temperature (C)

Although the Wind Chill Index is open to criticism arising from the differences in heat transfer
phenomena between a small plastic cylinder and a human body, it has been used as a basis for
the more commonly employed Equivalent Wind Chill Temperature, teq . This is defined as the
ambient temperature in an airstream moving at 1.79 m/s (4 mph) that gives the same WCI (rate of
cooling) as the actual combination of air temperature and velocity. That definition leads to the
equation
t eq = 33
(10.45 + 10
u u (33 t d )
22
(17.41)
Figure 17.10 Equivalent wind chill temperature (developed from data

reported by the U.S. Army Research Institute of Environmental
Medicine).
35
This relationship may be employed for air velocities not less than 2 m/s. Figure 17.10 gives a
graphical depiction of equivalent wind chill temperature. It is important to remember that teq is
indicative of a rate of cooling rather than a physical temperature. The ambient temperature, td,
controls the limit to which any body can cool and is independent of air velocity.
17.7 HEAT TOLERANCE AND ACCLIMATIZATION
In any mine or other subsurface facility where the wet bulb temperature exceeds 25C, it is
advisable to consider means of combatting potential heat strain within the workforce. The return
on such efforts is shown not only by a decreased risk of heat illnesses but also improvements in
safety statistics and productivity. This section considers procedures of physiological testing and
the acclimatization procedures that may be conducted on personnel prior to their employment in
hot conditions.
Much of the research and development on heat stress and acclimatization procedures for hot
mines has been carried out for the gold mining industry of South Africa. Most of the information
given in this Section originates from that work (e.g. South African Department of Minerals and
Energy, 2002; Bluhm and von Glehn, 2006; Burrows, 1989; Stewart, 1982).
Prior to 1990, the primary objective of heat stress procedures practiced in South Africa was to
reduce the incidence of heat stroke for persons who worked in deep gold mines. It was not
uncommon for wet bulb temperatures to reach 32 C in stoping areas. Following initial screening
and work/heat tolerance tests to identify the heat intolerant, new and returning mine recruits
underwent acclimatization procedures over several days in surface-located climatic chambers.
For historical interest those former procedures are outlined in Appendix A17.2.
Since the early 1990s the approach has changed. First, a determined effort was undertaken to
reduce wet bulb temperatures to not more than 28 C in active mining zones. Secondly, the
primary objective of reducing heatstroke has given way to a more general form of risk
assessment through an approach known as Heat Stress Management. This also includes revised
procedures of initial screening and heat tolerance tests but is followed by a natural progression of
supervised on-the-job heat acclimatization. The combination of reduced wet bulb temperatures
and the more holistic approach of Heat Stress Management has enabled the elimination of
several days spent in surface acclimatization chambers.
17.7.1 Physiological tests
17.7.1.1 Initial screening
The initial screening of new or returning recruits who will work in hot conditions in South Africas
gold mines consists of two parts. First, a general medical examination is carried out on each
candidate. In addition to health checks this examination includes the medical, occupational and
heat disorder history of each recruit being recorded within a database.
Secondly, a separate physical evaluation is carried out to identify features that would indicate
unsuitability for physical work in hot environments. These factors include:
Age
Tolerance to heat declines with age, a trend which accelerates rapidly over the age of 40 years. A
general recommendation is that persons of 50 years or more should not normally undertake
strenuous activities in hot environments.
Body dimensions.
Both body mass and Body Mass Index (BMI) should be recorded during the physical examination.
The latter is defined as:
36
BMI =
body mass
height
kg
m2
(17.42)
Table 17.4 gives recommendations of the Department of Minerals and Energy, Mine Health and
Safety Inspectorate of South Africa, 2002.
Employee status
Criterion/Standard
New or returning recruit
Body mass < 50 kg

Body mass 50 to 55 kg
BMI: 15 to 29
BMI: 30 to 35
BMI > 35
Unsuitable: reject
Suitable but not for strenuous work
Suitable
Suitable if no medical concerns
Unsuitable: reject
Existing worker
Body mass < 45 kg

Unsuitable: reject
Suitable if no medical concerns or history of
heat disorders
No allocation to strenuous work
(i.e. M > 160 W/m2)
Unsuitable
heat disorders
Suitable
heat disorders
Unsuitable
BMI < 15
BMI: 15 to 19
BMI: 20 to 29
BMI: 30 to 35
BMI > 35
Recommendation
Table 17.4. Body mass and Body Mass Index criteria for physical work in hot environments.
Source: Department of Minerals and Energy, Mine Health and Safety Inspectorate of South Africa, 2002.
http://www.dme.gov.za/pdfs/mhs/guidelines/thermal_stress.pdf
17.7.1.2 Heat Tolerance Screening

The aim of heat tolerance screening is to identify persons who are likely to suffer from heat stress
while working in hot environments. It is also indicative of the physical fitness of the individual.
The procedure commences with a rest period of 30 minutes in an environment that feels
comfortable to men wearing shorts only. Candidates who have oral temperatures greater than
37 C at the end of the rest period are excused from the heat tolerance screening and referred
for further medical evaluation.
The heat tolerance screening itself consists of a 30 minute period during which the test subjects
undertake a bench-stepping procedure at a rate of 24 steps per minute to a bench height of 30.5
cm. This approximates to a work rate of 80 W. One step consists of raising the upright body such
that both feet are placed on the bench then back to the floor. The climate should be maintained at
a dry bulb temperature of 29.5 C and a wet bulb temperature of 28 + 0.3 C while the air velocity
is maintained within the range 0.3 to 0.5 m/s.
37
Individuals whose oral temperature at the end of the test does not exceed 37.6 C are deemed to
be able to acclimatize successfully under supervision in the mine environment. Those with oral
temperatures higher than 37.6 C are recognized as heat intolerant and will not be required to
work in hot areas of the mine.
17.7.2 Acclimatization
17.7.2.1 In-situ acclimatization
For candidates who have successfully passed the medical, physical and heat tolerance
screenings, full acclimatization can be expected within some 12 successive work shifts in hot
conditions. During this time physiological changes take place which reduce susceptibility to heat
stress. Formal supervision must be provided during the period of acclimatization. Supervisors
should be capable of measuring wet and dry bulb temperatures, recognizing signs of ill health
and heat stress, ensuring that potable water is available and trained in emergency treatments.
Workers in hot conditions should be trained to recognize symptoms of heat stress, not only in
themselves but also in fellow workers. The training should also encompass good work practices
such as self-pacing of physical work and the use of rest periods. The latter may be imposed by
management.
It is important to remember that acclimatization does not provide immunity against heat illnesses,
including heat stroke, if good work practices are not maintained on a permanent basis. A loss of
full acclimatization can occur after missing only a few days of work in hot conditions. Similarly, if
exposure to those conditions occurs only intermittently full acclimatization will neither occur nor
be maintained. Hence, visits by persons such as senior officials or visitors should take place only
while accompanied by acclimatized supervisors.
17.7.2.2 Physiological changes during acclimatization
During the process of acclimatization, several significant physiological changes take place that
enhance the effectiveness of the body thermoregulation system. These are accompanied by a
noticeable psychological improvement. Acclimatization produces a dramatic increase in blood
volume, averaging about 25 per cent. This, in turn, results in several beneficial reactions. First,
there is an improvement in the response of blood flow conductance to increased metabolic heat
production or greater heat stress. Coupled with an enhanced elasticity in the walls of blood
vessels, this allows a greater flow of blood both to internal organs and to skin tissues. As a
consequence, both core temperature and heartbeat decrease for any given workrate and level of
environmental heat stress.
There is also a dramatic increase in sweat rate for any given core temperature during
acclimatization. Perspiration commences at a lower core temperature, increasing the range and
efficiency of evaporative cooling. The improved ability to produce sweat appears also to be
associated with the increased blood volume as well as enhanced performance of the sweat
glands.
Finally, acclimatization raises the mechanical efficiency of the human body, i.e. any given task is
achieved with a lower production of metabolic heat. This, again, results in a lower core
temperature and reduced thermal strain.
The combined effect of the changes obtained through acclimatization is that the worker can
perform appropriate tasks in hot conditions more efficiently, with a greatly reduced risk of heat
illness, and in a more stable psychological state.
38
17.8 EFFECT OF MINE CLIMATE ON PRODUCTIVITY

From a consideration of the psychological as well as physiological effects of hot working
conditions, it can be expected that as the cooling power of the ambient airstream decreases
below some 300 W/m2, personnel will become substantially less effective. There will be a
deterioration in the rate and quality of individual work performance. Unduly hot conditions may
result in poor morale and are likely to produce increased accident rates and problems of
absenteeism.
However, the same difficulties can arise from other causes including standards of supervision,
management style and labour relations, geological conditions and legislative requirements. For
this reason, few studies have been reported that give quantified relationships between mine
environmental conditions and those factors that contribute to overall mine productivity. It is,
however, relatively easy to demonstrate the debilitating influence of heat on individual work
performance. Figure 17.11(a) summarizes results obtained from monitoring the effectiveness of
acclimatized men engaged in manual work in a heading (Poulton, 1970). The work consisted of
shovelling fragmented rock into ore cars. The temperature of the near saturated air and the air
velocity at the workplace were varied to give the range of effective temperatures (Section
17.4.3.1) shown on the figure. The work performance at an effective temperature of 22C was
taken to represent 100 per cent. The results show that work efficiency begins to suffer at an
effective temperature of about 27C and declines rapidly when the effective temperature exceeds
30C. The latter approximates to an air cooling power of some 270 W/m2 for unclothed personnel.
100
worker performance %
90
80
70
60
50
40
30
20
10
0
20
22
24
26
28
30
32
34
36
Effective temperature, deg C

Figure 17.11(a) Influence of Effective Temperature on worker performance(after Poulton).
Performance is expressed as a percentage of the largest value reported.
39
Figure 17.11 (b) depicts results reported in a comprehensive study by Howes (1978) using
statistics collated by the Chamber of Mines for gold mines in South Africa. The air cooling power
used in this figure was an early version of the ACP (A scale).
Productivity per cen t
100
90
80
70
60
50
100
150
200
250
300
350
400
Air Cooling Power (ACP) W/sq. m

Figure 17.11(b) Effect of Air Cooling Power on mine productivity (after Howes).
Productivity is expressed as a percentage of the largest value reported.
The correlation between Figures 17.11 (a) and (b) is clear for a labour intensive operation.
However, any environment that causes physical discomfort will involve a decline in standards of
care and attention. The potential for accidents and loss of production is greater in a mechanized
operation where an error made by a single person can have serious repercussions. Low morale
and the irritability that often appear as an early symptom of heat strain can easily lead to strained
industrial relations. This can have disastrous effects on overall mine profitability. Furthermore, the
costs of the additional personnel that are required to compensate for absenteeism in hot mines
can be high in an expensive labour market.
Failure to create a reasonably comfortable working environment underground not only puts the
health and safety of the workforce at risk, it also results in lower productivity and profitability.
40
References
ASHRAE (1985, 1989 and 1993). American Society of Heating, Refrigerating and Air
Conditioning Engineers. Fundamentals Handbooks.
Bluhm, S. and von Glehn, F (2006). Private communication.
Botsford, J.H. (1971). A wet globe thermometer for environmental heat measurement. American
Industrial Hygiene Assoc. Journal: 38, 264
Brake, R. and Bates, G, (2002). A valid method for comparing and rational and empirical heat
stress indices. Ann. Occupational Hygiene, Vol. 46, No.2. pp. 165-174. Published by Oxford
University Press.
Burrows, J.H. (1989). Private communication.
Chamber of Mines of South Africa (1972). Routine mine ventilation measurements.
Johannesburg.
Ciriello, V.M. and S.H. Snook, 1977. The prediction of WBGT from the Botsball. American
Industrial Hygiene Association Journal, 38(June): p. 264-271.
Dreosti, A.O. (1935). Problems arising out of temperature and humidity in deep mines of the
Witwatersrand. J. Chem, Metall, Min. Soc. of S. Africa, Vol. 36, pp 102-129.
duBois, D. and duBois, E.F. (1916). A formula to estimate approximate surface area if height
and weight are known. Archives of Internal Medicine. 17: pp. 863-871.
Fanger, P.O. (1970). Thermal comfort. Danish Technical Press. Copenhagen.
Gagge, J. et al (1969). The prediction of thermal comfort when thermal equilibrium is maintained
by sweating. ASHRAE Trans. Vol. 75 Part 2 p.108.
Hanoff, J.S.T. (1970). Investigation of the limit of tolerable stresses to men subjected to climatic
and working conditions in coal mines. Glckauf-Forschung. Vol. 31 (4). pp. 182-195.
Hill, L. Griffith, O.W. and Flack, M. (1916). The measurement of the rate of heat loss on body
temperature by convection, radiation and evaporation. Phil. Trans. Royal Soc., Section B
(London) Vol. 207 pp 183-220.
Houghton, F.C. and Yaglou, C.P. (1923). Determining lines of equal comfort. ASHVE Trans.
Vol. 29 pp.163-176 and 361-384.
Howes, M.J. (1978). The development of a functional relationship between productivity and the
thermal environment. J. Mine Vent. Soc. of S. Africa, Vol. 31, pp 21-38.
International Organisation for Standardization (1982). Hot environments - Estimation of the
heat stress on working man, based on the WBGT (wet bulb globe temperature). IS0 Standard
7243 (See also Standard 7730, 1984).
Mitchell, D. And Whillier, A. (1972). The cooling power of underground environments. Journ.
Mine Ventilation Soc. of S.Africa. Vol 25, pp 140-151.
41
NIOSH (1986). Criteria for a recommended standard Occupational exposure to hot

environments, revised criteria. U.S. Dept. of Health and Human Services, USDHHS
(NIOSH) publication 86-113.
Onkaram, B. et al (1980). Three instruments for assessment of WBGT and a comparison with
WGT (Botsball). American Industrial Hygiene Assoc. 41: pp. 634-641.
Oohori, T. et al (1984). Comparison of current two-parameter indices of vapor permeation of
clothing as factors governing thermal equilibrium and human comfort. ASHRAE Trans. Vol. 90 (2).
Poulton, E.C. (1970). Environment and human efficiency. C.C. Thomas, Springfield, Illinois,
U.S.A.
Siple, P.A. and Passel, C.F. (1945). Measurements of dry atmospheric cooling in subfreezing
temperatures. Proc. Americ. Philosophical Soc., Vol. 89, p.177.
South African Department of Minerals and Energy, Mine Health and Safety Inspectorate
(2002). Guideline for the compilation of a Mandatory Code of Practice for an Occupational Health
Programme on Thermal Stress. 78pp.
http://www.dme.gov.za/pdfs/mhs/guidelines/thermal_stress.pdf
Stewart, J.M. (1982). Environmental engineering in South African mines, Chapters 20 and 21.
Mine Ventilation Soc. of S. Africa.
Stewart, J.M. and Whillier, A. (1979). A guide to the measurement and assessment of heat
stress in gold mines. Journ. Mine Ventilation Soc. of S.A. Vol. 32, No. 9. pp. 169-178.
Visagie, M.E. et al (1974). Changes in vitamin A and C levels in black mine workers. S. African
Med. Journ. Vol. 48, pp. 2502-2506.
Windham, C.H. (1974). The physiological and psychological effects of heat. Chapter 7 The
Ventilation of South African Gold Mines. Mine Vent. Soc. of S. Africa. pp 93-137.
Wu, H.W. and Gillies, A.D.S. (1997). The development of a personal response heat stress meter
calibrated to the range of physiology of the range of employees within the mine workplace. Proc.
6th International Mine Ventilation Congress, Pennsylvania, U.S. pp. 255-261. Published by the
Society of Mining Engineers.
Yaglou, C.P. and Miller, W.F. (1925). Effective temperatures with clothing. ASHVE Trans. Vol. 31
pp. 89-99.
42
APPENDIX A17.1
Listing of the thermoregulation model developed in Section 17.3.6
The program is listed in the original BASIC language in which it was developed and can be
translated into any other preferred language suitable for running on a personal computer. Alpha
characters are printed in upper case for clarity.
The data required are indicated in lines 20 to 100 and the computed parameters identified in lines
560 to 790. The program cycles iteratively, varying the mean skin temperature and other
dependent variables until the Air Cooling Power (M scale) equals the input metabolic rate. Mean
skin temperatures in excess of 36C generate a warning message.
10 OPEN "LPT1" FOR OUTPUT AS #1
20 INPUT TW,TD (deg C),P(kPa) ,TW,TD,P
30 INPUT "RAD TEMP OF SURROUNDINGS (deg C) ",TR
40 P=1000*P
50 INPUT RELATIVE AIR VELOCITY (m/s) ",UR
60 INPUT CLOTHING RESISTANCE, RCL (degC m^2/W) ,RCL
70 INPUT CLOTHING AREA RATIO,Fcl ,FCL
80 INPUT CLOTH VAPOUR PERMEABILITY EFFICIENCY, Icl ,ICL
90 INPUT METABOLlC RATE (W/m^2) ,M
100 INPUT "BODY VIEW FACTOR,Fr ",FR
110 IF UR<3 THEN 140
120 U=UR
130 GOT0 150
140 U=0.8*UR+0.6
150 T=TW
160 HC=.00878*P^.6*SQR(U)
170 GOSUB 930
180 SINT=S
190 GOSUB 1010
200 GOSUB 1050
210 TEX=32.6+(.066*TD)+(0.0002*E)
220 EA=E
230 GOSUB 1080
240 REM
250 REM RESPIRATORY HEAT, BR
260 T=TEX
270 GOSUB 930
280 SEX=S
290 BR=.0000017*M*(SEX-SINT)
300 REM
310 REM CONVECTIVE HEAT, C
320 CNV=(TSK-TD)/(RCL+1/(FCL*HC))
330 TCL=TSK-CNV*RCL
340 REM RADIATIVE HEAT, R
350 TCL=TSK-CNV*RCL
360 TAV=(TCL+TR)/2+273.15
370 HR=4*5.67E-08*(TAV)^3
380 R=HR*FR*(TCL-TR)
390 REM
400 REM EVAPORATIVE HEAT, EV
410 GOSUB 1190
420 IF TSK>32.5 THEN GOSUB 1410
43
430 T=TSK
440 GOSUB 850
450 HE=.0007*HC*L/P
460 GOSUB 810
470 RECL=RCL*P/(.0007*L*ICL)
480 EV=W*(ES-EA)/(RECL+1/(FCL*HE))
490 ACPM=BR+CNV+R+EV
500 PRINT TSK
510 IF ABS(ACPM-M)<1 THEN 560
520 TSK=TSK-.01*(ACPM-M)
530 W=W1
540 IF TSK>32.5 THEN GOSUB 1410
550 GOT0 310
560 PRINT #1,ACPM = ,:PRINT #1,USING ####;ACPM;:PRINT #1, W/m^2
570 PRINT #1,TCLOTH = ,:PRINT #1,USING ##.##;TCL;:PRINT #1, C
,
580 PRINT #1,SKIN TEMP = ;:PRINT #1,USING ##.##;TSK;:PRINT #1, C
590 IF TSK>=36 AND TSK<37 THEN PRINT #1,***ONSET OF HEAT STRAIN***:GOTO 610
600 IF TSK>=37 THEN PRINT #1 ,"*** DANGEROUS SKIN TEMPERATURE ***
610 PRINT #1,TEXHAUST = ,:PRINT #1,USING "##.##";TEX;:PRINT #1," C "
620 PRINT #1, "SWEAT = ,:PRINT #1,USING "##.##;W
630 PRINT #1,HC = ,:PRINT #1,USING ###.##;HC;:PRINT #1, W/m^2 C,
640 PRINT #1,"HR = ",:PRINT #1,USING ###.##;HR;:PRINT #1," W/m^2 C ,
650 PRINT #1,"HE = ",:PRINT #1,USING ###.##;HE;:PRINT #1," W/m^2 Pa",
660 PRINT #1,"FCL = , FCL
670 PRINT #1,"RECL = ,:PRINT #1,USING "###.###";RECL;:PRINT #1," m^2 Pa/W",
680 PRINT #1,"RCL = ,RCL
690 PRINT #1,"RESP = ",:PRINT #1,USING "###.##";BR;:PRINT #1," W/m^2 ,
700 PRINT #1,CONV = ",:PRINT #1,USING ###.##;CNV;:PRINT #1," W/m^2
710 PRINT #1,"RAD = ",:PRINT #1,USING ###.##;R;:PRINT #1," W/m^2 ,
720 PRINT #1,"EVAP = ",:PRINT #1,USING "###.##";EV;:PRINT #1," W/m^2"
730 PRINT #1,"SINT = ",:PRINT #1,USING "####.#";SINT*.001;:PRINT #1," kJ/kg ",
740 PRINT #1,"SEX = ",:PRINT #1,USING "####.#";SEX*.001;:PRINT #1," kJ/kg"
750 PRINT #1,"LSK = ",:PRINT #1,USING "#####.#";L*.001;:PRINT #1," kJ/kg ",
760 PRINT #1,"EAIR = ",:PRINT #1,USING ###.##;EA*.001;:PRINT #1, kPa"
770 PRINT #1,"ESK = ",:PRINT #1,USING ###.##;ES*.001;:PRINT #1," kPa ,
780 PRINT #1,"X = ,:PRINT #1,USING "##.####;X;:PRINT #1, kg/kg dry
790 PRINT #1,"MICRO VELOCITY = ";:PRINT #1,USING ##.##;U;:PRINT #1, m/s"
800 END
810 REM SAT VAP PRESS
820 ES=610.6*EXP(17.27*T/(237.3+T))
830 RETURN
840 REM
850 REM LATENT HEAT
860 L=(2502.5-2.386*T)*1000
870 RETURN
880 REM
890 REM MOISTURE CONTENT
900 X=.622*E/(P-E)
910 RETURN
920 REM
930 REM SIGMA HEAT
940 GOSUB 850
950 GOSUB 810
960 E=ES
970 GOSUB 890
980 S=L*X+1005*T
44
990 RETURN
1000 REM
1010 REM ACT. MOISTURE CONTENT
1020 X=(S-1005*TD)/(L+1884*(TD-TW))
1030 RETURN
1040 REM
1050 REM ACT VAP PRESS
1060 E=P*X/(.622+X)
1070 RETURN
1080 REM SKIN TEMP
1090 REM REF GAGGE, 1969
1100 TSK=24.85+.322*TD-.00165*TD*TD
1110 IF W<.9 THEN 1130
1120 TSK=24.49+.249*TD
1130 REM VELOCITY (HC) CORRECTION
1140 DTSK=8.999999E-03*HC*(TSK-TD)*(.15-RCL)/0.15
1150 TSK=TSK-DTSK
1160 REM CLOTHING CORRECTION
1170 TSK=TSK+(RCL/.15)
1180 RETURN
1190 REM SWEAT FRACTION
1200 IF TW> 27.194*EXP(-.004*M) THEN 1230
1210 W=0.06
1220 GOT0 1390
1230 IF TW<33 THEN GOT0 1260
1240 W=1
1250 GOT0 1390
1260 IF M>=200 THEN GOT0 1320
1270 A=-(1.781265E-04-2.544E-07*M+8.557743/M^2.218731)
1280 B=.012095-1.79735E-05*M+1063.187/M^2.31927
1290 C=-(.194294-3.279531E-04*M+46903/M^2.44229)
1300 D=.998117-1.816454E-03*M+92432l/M^2.651246
1310 GOT0 1360
1320 A=-(3.97379E-05+7.075456E-08*M+17.880231/M^2.218441)
1330 B=2.369825E-03+3.359426E-06*M+2211.1688/M^2.316119
1340 C=-(-5.682282E-03+5.86905E-05*M+95583.18/M^2.437565)
1350 D=-.321605+7.958265E-04*M+1875636/M^2.645912
1360 W=D+TW*(C+TW*(B+TW*A))
1370 IF W<.06 THEN W=.06
1380 IF W>1 THEN W=1
1390 W1=W
1400 RETURN
1410 REM INCREASE W FOR TSK>32.5
1420 IF TSK<38 THEN 1450
1430 W=1
1440 GOT0 1480
1450 DW=SIN((31.85*TSK-762)*.01745)*.5+.5
1460 W=W1+DW
1470 IF W>1 THEN W=1
1480 RETURN
45
APPENDIX A17.2
Acclimatization of workers in South African gold mines prior to 1990.
This appendix is included for historical interest and is taken from the earlier printed version of this
textbook. It reflects the philosophy and procedures of the South African gold mining industry
relating to heat stress in underground workers before 1990. The updated procedures at the time
of this current writing are outlined in Section 17.7 of the main text.
Physiological tests
Initial screening
An initial screening of recruits for work in wet bulb temperatures of more than 25C should
investigate both body mass and age. Any person with a body mass of less than 50kg is unlikely to
have a skin surface area that is sufficient to ensure efficient body cooling. Conversely, individuals
with a mass exceeding 100kg have a high and constant metabolic burden in supporting their own
weight. No persons in either of these groups should be asked to work in hot environments.
After the age of 40 years, the probability of heat stroke increases significantly. Although
considerable variation exists between individuals, it is prudent to relocate persons of more than
45 years of age to zones of reduced heat stress.
Work capacity tests
A simple test of work capacity consists of a ten minute period of constant and controlled activity
that requires an oxygen consumption rate of 1.25 litres per minute. At the conclusion of the
period, the heartbeat provides an indication of maximum rate of oxygen consumption and, hence,
work capacity. Persons with test heartbeats of over 140 bpm should be prohibited from tasks
requiring hard manual exertion or from working in conditions of heat stress. If the test results
indicate a heartbeat of less than 120 bpm then that individual should be capable of strenuous
work. If the wet bulb temperature in the workings does not exceed 27.5C the initial screening
and the work capacity tests are all that are required.
Heat tolerance test
This more rigorous test may be carried out on persons who have passed both the initial screening
and the work capacity tests, and who may be required to work in wet bulb temperatures in excess
of 27.5C. The test consists of requiring the participants to engage in a controlled activity
corresponding to an oxygen consumption of 1.25 litres/minute for a period of four hours in a
climatic chamber. The wet bulb temperature should be at least equal to the highest value that will
be encountered within work areas in the mine.
Oral temperatures are taken at intervals of not more than 30 minutes throughout the test and the
following criteria applied:
If oral temperature
remains below 37.5C:
varies between 37.5 and 38.6C:
varies between 38.6 and 40.0C:
exceeds 40.0C:
person eligible immediately for work in hot mine environment

eligible for short-term or micro-climate acclimatization
eligible for full acclimatization
test terminated and person classified as heat intolerant
If it is suspected that heat intolerance is caused by a temporary illness or body disorder then the
test may be repeated after treatment and recovery.
Heat acclimatization
The organized process of heat acclimatization may take place over a period of several days
during which the participants are subjected to gradually increasing work rates and/or levels of
heat stress. During this time, definite physiological changes occur that enable the human body to
better resist the adverse effects of a hot environment.
46
The first attempts at organized acclimatization for work in hot mines appear to have been carried
out in 1926 within the South African gold mining industry (Dreosti). Considerable effort was
devoted to the improvement of acclimatization procedures and South Africa seems to have been
the only country that provided formal acclimatization for the majority of recruits into deep and hot
mines. However, advances in the design of ventilating and air conditioning systems have led to
significant improvements in climatic conditions within those mines. The modern philosophy is to
promote working environments that do not necessitate full acclimatization procedures.
Methods of acclimatization
There are essentially three methods of acclimatization for workers in hot environments. The first
and, perhaps, most natural is for the individual to spend the first two weeks of underground work
in tasks of increasing physical effort (or less frequent rest periods) within the actual mine
environment. This may be combined with commencing in relatively cool areas of the mine and
progressively moving the participants to locations of increased heat stress through the two week
period.
Although such "on-the-job acclimatization appears to be convenient, it does have significant
drawbacks. If successful acclimatization is to be achieved, then the process must be monitored
and controlled by well-trained supervisors. Work rates, body temperatures and psychrometric
conditions must be logged and, if necessary, modifications made to the place or rate of work.
Individuals who exhibit signs of heat intolerance must be identified before they suffer unduly from
heat strain. Drinking water must be administered at frequent intervals. The difficulty is that such
procedures do not blend readily into the production-oriented activities of a working stope or face.
It is for this reason that on-the-job acclimatization has, in general, been less successful than the
more easily controlled procedures.
For individuals whose oral temperature does not exceed 38.6C during a heat tolerance test a
shorter period of acclimatization may provide the required degree of protection. The use of microclimate jackets during acclimatization may be deemed acceptable for candidates who are familiar
with underground operations. For the first week in hot conditions, the worker performs his
expected rate of work while wearing a micro-climate jacket. This allows the individual to be fully
productive during the acclimatization period while, at the same time, offering protection against
heat illness.
The use of surface or underground climatic chambers designed specifically for fully controlled
acclimatization procedures was developed for the South African gold mining industry (Stewart,
1982). The success of those procedures in reducing the incidence of heat stroke and improving
productivity led to the widespread use of the technique within that country. The environment
within the climatic chamber is maintained at a wet bulb temperature of about 31.7C and an air
velocity of 0.5 m/s. The work consists of stepping on and off a bench at a rate indicated by an
audible and visual metronome. The height of the bench is adjusted according to each man's body
weight such that every person works at the same rate. This correlates with an oxygen
consumption of some 1.4 litres/minute and, hence, is equivalent to hard manual work. The task is
carried out for four continuous hours per day over a period of several days. Measurements are
made of body temperature, and drinking water is administered every thirty minutes. During the
first day of acclimatization, the stepping task is conducted at only half the normal pace for most of
the four hours. The time spent on full work rate (oxygen consumption of 1.4 litres/minute) is
increased gradually until the final day is entirely at the full rate.
The period of acclimatization within a climatic chamber originally encompassed eight days.
However, the improvement of conditions underground, coupled with heat tolerance testing have
resulted in the acclimatization period being reduced to four days with far fewer men required to
undergo it. The ultimate objective is to reduce the wet bulb temperature to less than 27.5C in all
mines at which time acclimatization will no longer be required (Burrows, 1989).
47
Chapter 18
Refrigeration Plant and Mine Air Conditioning Systems
CHAPTER 18. REFRIGERATION PLANT AND MINE AIR CONDITIONING SYSTEMS

18.1. INTRODUCTION .............................................................................................. 2
18.2. THE VAPOUR COMPRESSION CYCLE .......................................................... 2
18.2.1. Basic principles ....................................................................................................................... 2
18.2.2. Refrigerant fluids..................................................................................................................... 4
18.2.3. Basic components of the vapour compression cycle.............................................................. 5
18.2.4. Performance of a refrigeration cycle....................................................................................... 6
18.2.4.1 The Carnot cycle ............................................................................................................... 6
Coefficient of Performance, COP, ................................................................................... 8
18.2.4.2. The actual cycle ............................................................................................................... 8
18.3. COMPONENTS AND DESIGN OF MINE COOLING SYSTEMS.................... 15

18.3.1. Overview of mine cooling systems ....................................................................................... 15
18.3.2. Heat exchangers................................................................................................................... 20
18.3.2.1. Indirect heat exchangers................................................................................................ 22
18.3.2.2. Performance calculations for indirect heat exchangers ................................................. 22
18.3.2.3. Direct heat exchangers .................................................................................................. 26
Chilled water spray chambers ....................................................................................... 28
18.3.2.4. Performance calculations for direct heat exchangers. ................................................... 31
18.3.2.5. Heat exchange across the walls of pipes and ducts ...................................................... 41
Air inside a duct ............................................................................................................. 42
Water inside a pipe ........................................................................................................ 42
Air outside a pipe ........................................................................................................... 43
Radiative heat transfer coefficient ................................................................................. 43
18.3.3. Water distribution systems................................................................................................... 45
18.3.3.1. Pipe sizing ...................................................................................................................... 46
18.3.3.2. Pipe insulation ................................................................................................................ 47
Table 18.3. Thermal conductivities of pipe materials. ................................................. 48
18.3.3.3. Energy and temperature changes within water systems ............................................... 48
18.3.4. Energy recovery devices....................................................................................................... 53
18.3.4.1. Water turbines ................................................................................................................ 54
18.3.4.2. Hydrolift systems ............................................................................................................ 55
18.3.5. Design of mine cooling systems ........................................................................................... 56
18.3.5.1. Location of main plant .................................................................................................... 56
Surface plant.................................................................................................................. 57
Underground plant ......................................................................................................... 57
Combinations of surface and underground plant .......................................................... 59
18.3.5.2. Service water cooling ..................................................................................................... 61
18.3.5.3. Hydropower .................................................................................................................... 62
18.3.5.4. Ice systems ................................................................................................................... 62
Manufacture of ice ......................................................................................................... 64
Transportation of ice ...................................................................................................... 65
Incorporation into the mine cooling system ................................................................... 66
Economics of ice systems ............................................................................................. 67
18.3.6. Summary of design process ................................................................................................. 67
18.4. AIR HEATING ................................................................................................ 68

18.4.1. Utilization of waste heat ........................................................................................................ 69
18.4.2. Heat recovery from exhaust air............................................................................................. 69
18.4.3. Direct heating........................................................................................................................ 71
Table 18.4. Gross calorific values of gases................................................................... 71
18.4.4. Indirect heating ..................................................................................................................... 72
18.4.5. Ice stopes.............................................................................................................................. 73
18.4.6. Geothermal and cycled storage heat.................................................................................... 74
References.................................................................................................. 76
18-
Chapter 18
18.1. INTRODUCTION
One of the earliest methods of temperature control in underground mines was the importation of
naturally produced ice from the surface. Blocks of ice were transported in ore cars to cool miners in
the Comstock Lode under Virginia City in Nevada, USA, during the 1860's. The vapour compression
refrigeration cycle, currently the most widespread method of artificial cooling, appears to have first
been used in mining during the 1920's. Examples included the famous Morro Velho Mine in Brazil
(1923) and experimental work in British coal mines (Hancock, 1926). Air cooling techniques in mining
gained further recognition in the 1930's including their utilization in the gold mines of South Africa and
in the Kolar Goldfields of India. However, it was the 1960's that saw the start of a real escalation of
installed mine cooling capacity.
Large centralized refrigeration plant, located underground, became popular in the South African gold
mines. Limitations on the heat rejection capacity of return air, combined with the development of
energy recovery devices for water pipelines in shafts and improved "coolth" distribution systems led
to a renewed preference for surface plant. (The term "coolth" is a decidedly unscientific but
descriptive word sometimes used in association with chilled fluid distribution systems.)
Although the traditional role of mine cooling has been to combat geothermal heat and the effects of
autocompression in deep metal mines, an additional influence has been the escalating amount of
mechanized power employed underground, particularly in longwall coal mines. This resulted in
smaller scale and more localized use of air cooling units in such mines at depths where, prior to
intensive mechanization, heat had not been a limiting environmental problem. Examples of these
installations appeared in the United Kingdom and Germany.
In this chapter, we shall examine the essential theory of the vapour compression refrigeration cycle in
addition to discussing the design of mine cooling systems and some of the methods of distributing
"coolth" to the working areas of a subsurface facility. The chapter concludes with a section on the
opposite problem, that of increasing the temperature of the intake air for mines in cold climates.
18.2. THE VAPOUR COMPRESSION CYCLE

There are many transient phenomena that are known to produce a cooling effect, varying from
endothermic chemical reactions to the sublimation of solid carbon dioxide (dry ice). Where a
continuous cooling effect is required, then a means must be employed by which a supply of
mechanical, electrical or thermal energy is utilized to remove heat from some source, and to transport
it to a thermal sink where it can be rejected. If the primary objective is to cool the source, then the
device is known as a refrigerator. If, however, the desired effect is to heat the sink, then it is called a
heat pump. In fact, both effects occur simultaneously. Hence, a domestic refrigerator cools the interior
of the container but heats the air in the kitchen.
Of the several devices that have been developed to achieve a continuous refrigeration or heat pump
effect, the most common is based on the vapour compression cycle. This may be used on small units
such as air conditioning equipment fitted to automobiles or for very large scale cooling of mine
workings where many megawatts of heat require to be transferred. In this section, we shall examine
the basic principles of the vapour compression cycle, how rates of heat exchange may be calculated,
and the essential components of a refrigeration plant.
18.2.1. Basic principles
When a liquid boils, it does so at constant temperature provided that the applied pressure remains
fixed. The heat added is utilized in increasing the internal kinetic energy of the molecules until they
18-
Chapter 18
can no longer remain in the liquid phase but burst free to form a vapour or gas (Section 2.1.1). If,
however, the applied pressure is raised to a higher value, then additional heat is required to vaporize
the liquid. The boiling temperature will increase. The relationship between pressure and boiling point
for any given liquid may be defined as the vapour pressure line on a pressure-temperature diagram
such as that shown on Figure 18.1. The fluid is a liquid on the left of the curve and a vapour (gas) on
the right of the curve. The liquid may be vaporized either by increasing the temperature or decreasing
the pressure. Similarly, condensation from vapour to liquid may occur either by decreasing the
temperature or increasing the pressure.
Figure 18.1 A pressure-temperature diagram showing limits of phase change.
The change in physical appearance of a fluid on crossing the vapour pressure line is quite distinct,
liquid to gas or vice versa. However, there is a critical pressure-temperature coordinate beyond which
the change of phase is gradual rather than sudden and there is no clearly defined moment of
evaporation or condensation. This is known as the critical point.
Although the vapour pressure curve appears as a single line on the pressure-temperature diagram, it
takes a finite amount of time and energy exchange to cross that line. During this time, part of the fluid
will be liquid and the remainder will be vapour. This is the situation that exists inside a boiling kettle.
The region within which the two phases coexist is more clearly shown on the PV (pressure against
specific volume) diagram of Figure 18.2(a). The corresponding temperature-entropy (Ts) and
pressure-enthalpy (PH) diagrams are shown on Figures 18.2(b) and (c) respectively. The latter two
diagrams are particularly useful in analyzing and quantifying both the work and heat transfer
processes in the vapour compression cycle.
18-
Chapter 18
In all three diagrams of Figure 18.2, a horizontal line within the two phase region indicates that the
pressure and temperature both remain constant during any given isobaric phase change. However,
the volume, entropy and enthalpy all increase
significantly during vaporization. Different horizontal
lines on anyone of these diagrams will indicate different
values of pressure and corresponding boiling (or
condensation) temperature (Figure 18.1).When a liquid
boils at a given value of applied pressure it will extract
heat from the surroundings or other available medium.
If the vapour thus produced is then transported to a
new location and compressed to a higher pressure,
then it can be condensed at a correspondingly higher
temperature, yielding up its heat of condensation to the
new surroundings or any cooling medium that may be
supplied. This is the basic principle underlying the
vapour compression refrigeration cycle.
18.2.2. Refrigerant fluids
The differing pressure-temperature relationships of
various fluids allow each of those fluids to act as a
refrigerant over specified temperature ranges. Carbon
dioxide and even water have been used as
refrigerants. For the ranges of pressures and
temperatures acceptable in refrigeration plant for air
conditioning, ammonia is a particularly efficient
refrigerant although its toxicity limits its use. The ideal
refrigerant for mining use is one that gives a high
efficiency of heat transfer, is non-corrosive to metal,
non-toxic and has a boiling temperature close to 0C at
a positive pressure with respect to the ambient
atmosphere. Fluorinated hydrocarbons have been
widely employed in many types of industrial plant as
well as domestic refrigeration equipment. Although
non-toxic, fluorinated hydrocarbons react adversely
with atmospheric ozone and are being replaced by
alternative refrigerants.
Figure 18.2 PV, Ts and PH diagrams for
change of phase.
Refrigerant fluids are commonly referred to, not by their

chemical names, but by R (refrigerant) number. Table
18.1 indicates some of those designations.
Refrigerant No.
Chemical name
Refrigerant No
Chemical Name
R11
R12
R22
R23
R30
R50
R110
R113
R120
trichlorofluoromethane
dichlorodifluoromethane
chlorodifluoromethane
trifluoromethane
dichloromethane
methane
hexachloroethane
trichlorotrifluoroethane
pentachloroethane
R122
R170
R290
R600
R717
R718
R744
R1150
R1270
trichlorodifluoroethane
ethane
propane
butane
ammonia
water
carbon dioxide
ethylene
propylene
Table 18.1 Refrigerant numbers of selected fluids.

18-
Chapter 18
18.2.3. Basic components of the vapour compression cycle

Figure 18.3 shows that there are four essential components of hardware in a vapour compression
refrigeration unit. The evaporator is a heat exchanger, typically of the shell-and-tube configuration in
mining refrigeration plant. In the larger units, the refrigerant liquid is on the outside of the tubes while
the medium to be cooled (for example, water, brine or glycol) passes through the tubes. Smaller units
employed for direct cooling of an airstream are sometimes called "direct evaporators" and contain the
refrigerant within the tubes while the air passes over their outer surface.
heated air or
water to heat
rejection
superheated
vapour
high pressure:high temperature
liquid
condenser
expansion
valve
compressor
warm air
or water
evaporator
low pressure: low temperature
chilled air or
water
Figure 18.3 Major components of a vapour compression cycle.
Within the evaporator, the refrigerant pressure is maintained at a relatively low level and boils at a
correspondingly low temperature. For example, refrigerant R12 will boil at 4C if the pressure is 351
kPa (approximately 3.5 atmospheres). The heat required to maintain the boiling is extracted from the
gas or liquid passing on the other side of the tube walls. Hence, that gas or liquid is cooled. The
refrigerant, now vaporized, collects at the top of the evaporator and is allowed to gain a few degrees
of superheat to ensure full vaporization before it passes on the compressor. Except for direct
evaporators, it may be necessary to insulate the external surface of the evaporator in order to prevent
excessive heat gain from the ambient atmosphere.
18-
Chapter 18
The compressor is the device where mechanical work is input to the system. Reciprocating, screw or
centrifugal compressors are all employed. The latter are favoured for the larger units and where the
required pressure ratio remains constant. However, the development of large diameter screw
compressors allows a good efficiency to be maintained under conditions of variable cooling load.
(Baker-Duly, 1989). Multi-stage compressors are employed to give high differentials of pressure and,
hence, large temperature differences between the evaporator and condenser. Electric motors are
normally employed to drive the compressors on mine refrigeration units although internal combustion
engines may be used on surface or as standby units. The duty of a refrigeration plant can be modified
by changing the speed of the compressor. The flow of vapour through the compressor and, hence,
rate of heat transfer can also be controlled by inlet guide vanes.
The refrigerant vapour leaves the compressor and passes into the condenser at a relatively high
pressure and temperature. The condenser itself may be of similar construction to the evaporator, that
is, a shell-and-tube heat exchanger. Heat is removed from the refrigerant by air, water or some other
fluid medium to the extent that the refrigerant cools and condenses back to a liquid. As the pressure
is high, this occurs at a relatively high temperature. At a pressure of 1217 kPa, refrigerant R12 will
condense at 50C. The latent heat of condensation is removed by the cooling fluid for subsequent
rejection in a cooling tower or other type of separate heat exchanger. As the vapour compression
cycle is a closed system (ignoring heat losses or gains from the surrounding atmosphere), the rate at
which heat is removed from the refrigerant in the condenser must equal the combined rates of heat
addition in the evaporator and work provided by the compressor.
The condensed refrigerant passes from the condenser to the fourth and final component of the cycle.
This is the expansion valve whose purpose is simply to reduce the pressure of the refrigerant back to
evaporator conditions. An expansion valve may be a simple orifice plate or can be controlled by a
float valve. At the exit from the expansion valve, the liquid is at low pressure and has a
correspondingly low boiling temperature. Provided that the pipework is insulated, the latent heat for
boiling can come only from the liquid refrigerant itself. Hence, the temperature of the refrigerant drops
rapidly as it passes from the expansion valve to the evaporator where it enters as a mixture of liquid
and vapour, thus closing the cycle.
18.2.4. Performance of a refrigeration cycle
18.2.4.1 The Carnot cycle
Electrical motors or heat engines are devices that convert one form of energy into another. Their
efficiency may, therefore, be defined as an energy output/input ratio. However, in the case of
refrigerators or heat pumps, the purpose is to remove heat from a given source and to reject it at a
higher temperature to a receiving sink. A different measure of performance is required.
Figure 18.4 shows the temperature-entropy diagram for the ideal (frictionless) Carnot cycle to which
an actual vapour compression cycle can aspire but never attain. We may follow the ideal cycle by
commencing at position 1, the entry of refrigerant vapour to the compressor.
The ideal compression process is isentropic (Section 10.6). The compressed refrigerant vapour
enters the condenser at position 2 and, in the ideal condenser, passes through to position 3 with
neither a pressure drop nor a fall in temperature. This assumes frictionless flow and perfect heat
transfer. The ideal expansion valve allows an isentropic fall in pressure and temperature to station 4,
the entrance to the evaporator. The cycle closes via frictionless flow and perfect heat transfer in the
evaporator.
18-
Chapter 18
12.5
temperature T
12
11.5
condenser
expansion
valve
11
compressor
10.5
evaporator
10
9.5
0
0.1
0.2
0.3
sb
0.4
0.5
0.6
sa
0.7
0.8
0.9
entropy s
Figure 18.4 Temperature-entropy diagram for an ideal vapour compression cycle
(Carnot cycle)
In order to quantify the ideal Carnot cycle, we apply the steady-flow energy equation (3.25). The
kinetic and potential energy terms are small and, in any event, cancel out in a closed cycle. Using the
numbered station points from Figure 18.4, the steady-flow energy equation gives:
compressor:
condenser:
expansion valve:
evaporator:
W12
q23
0
q41
cycle summation: W12 + q23 + q41
=
=
=
=
=
H2
H3
H4
H1
H1
H2
H3
H4
J/kg
J/kg
J/kg
J/kg
(18.1)
(18.2)
(18.3)
(18.4)
J/kg
(18.5)
where:
W12 = mechanical energy added by the compressor
q23 = heat exchange in the condenser (this is numerically negative as heat is leaving the refrigerant)
q41 = heat added to the refrigerant in the evaporator
H
= enthalpy
We can rewrite equation (18.5) as
W12 + q41 = - q23
J/kg
(18.6)
thus confirming our earlier statement that the heat rejected in the condenser is numerically equal to
the sum of the compressor work and the heat added to the refrigerant in the evaporator.
18-
Chapter 18
The measure of performance of a refrigeration cycle is known as the

Coefficient of Performance, COP,
and is defined as
COP =
q 41
Useful cooling effect (evaporato r heat transfer)
=
Work input from the compressor
W12
(18.7)
Using equation (18.5) once again, this may be rewritten as

COP =
q 41
q 23 q 41
(18.8)
Recalling that the area under a process line on a Ts diagram represents heat (Section 3.5), the terms
in equation (18.8) may be related to Figure 18.4 for the Carnot cycle:
q41
-q23
and -q23 - q41
Hence, W12 =
=
=
=
Area 41ab
Area 23ba
Area 1234 (remember that q23 is numerically negative)
Area 1234 and equation (18.8) becomes
Carnot COP =
T1 (s a s b )
Area 41ab
=
(T2 T1 )(s a s b )
Area 1234
=
T1
T4
or
(T2 T1 )
(T3 T 4 )
(18.9)
as T1 = T4 and T2 = T3
Hence, the ideal or Carnot coefficient of performance is given as the ratio:
Carnot COP =
Evaporator temperatur e (absolute)

Condenser temperatur e Evaporator temperatur e
(18.10)
Example
The evaporator and condenser of a refrigeration unit have temperatures of 4 and 50C respectively.
Determine the maximum possible coefficient of performance of this unit.
Solution
The Carnot or ideal coefficient of performance is given by equation (18.10)
Carnot COP =
(273.15 + 4)
(50 4)
= 6.025
18.2.4.2. The actual cycle

A real vapour compression cycle has a coefficient of performance that is necessarily lower than the
corresponding Carnot COP. There are two reasons for this. First, in actual compressors and
expansion valves, there are, inevitably, increases in entropy. Furthermore, as the refrigerant passes
through real condensers and evaporators there will be slight changes in pressure and temperature. In
18-
Chapter 18
a well designed unit, the latter are small compared with the differences in pressure or temperature
between the condenser and evaporator.
Secondly, it would be impractical to design a refrigeration unit that attempted to follow the Carnot
cycle. Referring again to Figure 18.4, it can be seen that station 1 (entry to the compressor) lies within
the two-phase region. The presence of liquid droplets would cause severe erosion of the compressor
impeller. It is for this reason that a few degrees of superheat are imparted to the vapour before it
leaves a real evaporator.
Figure 18.5 illustrates the temperature-entropy and pressure-enthalpy diagrams for a practical vapour
compression cycle. The major difference between Figure 18.5(a) and the Carnot cycle shown on
Figure 18.4 is that the compression commences slightly beyond the saturated vapour line (station 1)
and involves an increase in entropy to station 2. On entering the condenser, the superheated vapour
undergoes a near isobaric cooling process (stations 2 to 2'), losing sensible heat to the cooling
medium until condensing temperature is reached at station 2'.
Although the process through the expansion valve is no longer isentropic, the fluid remains at
constant enthalpy (no heat exchange takes place) and, hence, appears on the pressure-enthalpy
diagram as a vertical line 3 to 4.
Equations (18.1 to 18.4) arose from those parts of the steady-flow energy equation that do not involve
the friction term. They apply equally well to both the real and the ideal cases. Furthermore, the
definition of coefficient of performance given in equation (18.7) remains applicable. Substituting for
W12 and q41 from equations (18.1) and (18.4) respectively gives
Actual COP =
q 41
H H4
= 1
W12
H 2 H1
(18.11)
The reason for introducing the pressure-enthalpy diagram now emerges. The coordinate points for
real vapour compression cycles can be plotted on pressure-enthalpy-temperature charts that have
been derived through tests on many refrigerant fluids. An example is given for refrigerant R12 on
Figure 18.6. The values of enthalpy may be read from the relevant chart enabling the actual
coefficient of performance and other parameters to be determined. For more precise work, tables of
the thermodynamic behaviour of refrigerants are available. The values of enthalpy and entropy on the
charts and tables are based on a specified datum temperature. This is usually either absolute zero
(-273.15C) or -40C. The datum employed has little import as it is differences in values that are used
in practical calculations.
A measure of efficiency of the refrigeration unit may now be defined by comparing the actual COP
with the ideal Carnot COP.
Cycle efficiency =
Actual COP
Carnot COP
(18.12)
18-
Chapter 18
12.5
temperature T
12
isobar
11.5
expansion
valve
11
condenser
compressor
1
4
evaporator
10.5
10
9.5
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
entropy s
0.9
(a)
12.5
pressure P
12
11.5
expansion
valve
11
condenser
2
compressor
10.5
evaporator
10
9.5
0
0.1
0.2
0.3
0.4
0.5
0.6
enthalpy H
0.7
0.8
0.9
(b)
Figure 18.5 Temperature-entropy and pressure-enthalpy diagrams for an actual vapour

compression cycle
18- 10
Chapter 18
Figure 18.6 Pressure-enthalpy diagram for refrigerant R12 (based on zero degrees Kelvin).
Reprinted by permission from the 1985 ASHRAE Handbook Fundamentals.
18- 11
Chapter 18
Example
A plant employing refrigerant R12 is used to chill water for a mine distribution system. Water is also
used to remove heat from the condenser. The following measurements are made:
Evaporator:
Water flowrate:
Water inlet temperature:
Water outlet temperature:
Refrigerant pressure:
50 litres/s
20C
10C
363 kPa
Condenser:
Water flowrate:
Refrigerant pressure:
140 litres/s
1083 kPa
Compressor:
Refrigerant inlet temperature:
Refrigerant outlet temperature:
7C
65C
Using the pressure-enthalpy chart for R12, analyze the performance characteristics of the unit.
Solution
Step 1: On the pressure-enthalpy diagram, draw horizontal lines to represent the pressures in the
evaporator (363 kPa) and condenser (1083 kPa). For clarity, the relevant area of the R12
pressure-enthalpy chart has been enlarged and reproduced as Figure 18.7. The
corresponding temperatures may be estimated directly from the chart.
tevap
tcond
= t4 = 5C
= t3 = 45C
These temperatures can, of course, be verified by direct measurement.

Step 2: Establish the nodal points on the pressure-enthalpy chart and read the corresponding values
of enthalpy.
The temperature of the refrigerant entering the compressor (station 1) is given as 7C. This
represents 7 - 5 = 2 of superheat. Station 1 is, therefore, established on the evaporator
pressure line at a temperature of 7C. Note that the lines of constant temperature change
from horizontal to near vertical on passing the saturated vapour line. We have ignored the
small pressure drop between the evaporator and the compressor inlet.
Station 2 is established in a similar manner, i.e. at the measured compressor outlet
temperature of 65C and lying on the condenser pressure line.
Station 3, representing the refrigerant leaving the condenser, lies on the junction of the
evaporator pressure line and the saturated liquid curve.
The expansion process is represented by a vertical line dropped from station 3 until it
intersects the evaporator pressure line at station 4.
The four critical enthalpies may now be estimated from the chart:
compressor inlet
H1 = 356 kJ/kg
compressor outlet/condenser inlet
H2 = 386 kJ/kg
condenser outlet
H3 = 244 kJ/kg
evaporator inlet
H4 = 244 kJ/kg
18- 12
Chapter 18
Figure 18.7 Example vapour compression cycle on the Pressure-enthalpy chart.
Step 3: Establish the coefficients of performance and cycle efficiency:

Carnot COP =
T4
(273.15 + 5) = 6.954 equations (18.9 or 18.10)
=
(45 5)
(t 3 t 4 )
Actual COP =
(H 1 H 4 )
(H 2 H 1 )
Cycle efficiency =
356 244
= 3.733
386 356
equation (18.11)
Actual COP
3.733
=
1000 = 53.7 per cent
Carnot COP
6.954
equation (18.12)
Step 4: Determine the useful cooling effect, or evaporator duty, qevap. This can be established from
the known flowrate and temperature drop of the water being cooled as it passes through the
evaporator:
qevap
= mw,evap Cw Tw, evap
where: water mass flowrate

specific heat of water,
temperature drop of water
kW
mw,evap
Cw
Tw, evap
giving qevap = 50 x 4.187 x 10 kJ/s or kW
(18.13)
= 50 litres/s = 50 kg/s
= 4.187 kJ/kgC
= 20 -10 = 10C
=
2093.5 kW
18- 13
Chapter 18
Step 5: Determine the mass flowrate of refrigerant, mr, around the system. The evaporator duty,
already found to be 2093.5 kW, can also be expressed as
= m r (H 1 H 4 )
q evap
kW
(18.14)
Hence,
mr
2093 .5
(356 244 )
18.69
kg/s
Step 6: Determine the compressor duty. The work input from the compressor to the
refrigerant,W12, can be established in two ways. First,
W12
= m r (H 2 H 1 ) = 18.69 (386 356 ) = 560.7 kW
(18.15)
and, secondly, from the definition of coefficient of performance given in equation (18.7),
q evap
2093 .5
= 560.8 kW
Actual COP
3.733
The slight difference arises from rounding errors.
W12 =
Step 7: Evaluate the condenser duty. The heat transferred from the refrigerant in the condenser is
simply
q cond = q evap + W12 = 2093.5 + 560.7 = 2654.2
kW
(equation (18.6))
As the cooling water flowrate through the condenser is known to be mw,cond = 140 kg/s, the
temperature rise of this water, Tw cond, may be determined from
q cond
giving
Tw ,cond
= mw ,cond Cw Tw ,cond
=
2654.2
140 4.187
= 4.53 C
The performance of a refrigeration unit may be enhanced beyond that illustrated in the example. For
large units, it may be necessary to employ two or more stages of compression in order to achieve the
desired difference in temperature between evaporator and condenser. In that case, it is usual to
employ the same number of expansion valves as stages of compression. A fraction of the refrigerant
is vaporized on passing through an expansion valve. That fraction can be read from the
corresponding point on the pressure-enthalpy diagram or calculated from the isobaric enthalpies at
the saturated liquid and saturated vapour curves. This "flash gas" may be separated out and piped to
the corresponding intermediate point between compressor stages. This is known as interstage
cooling or, simply, intercooling and reduces the total mass of refrigerant that passes through all
stages of compression. The result is to reduce the required compressor power and/or increase the
useful refrigerating effect.
In some units, the hot liquid refrigerant leaving the condenser is partially cooled by a secondary
cooling circuit. This subcooling reduces the fraction of flash gas produced at the downstream side of
expansion valves and gives a small increase in the performance of the plant.
18- 14
Chapter 18
The coefficient of performance as defined by equation (18.7) takes into account the isentropic
efficiency of the compressor. However, it does not cater for the efficiency of the motor or other device
that drives the compressor. Hence, the W12 in equation (18.7) may be replaced by the power fed to
the compressor drive unit to give a more practical coefficient of performance. Furthermore, the
compressor is not the only device that consumes energy in a refrigeration unit. There will also be
water pumps or fans to promote the flow of the cooled medium at the evaporator and the cooling
medium at the condenser. An overall coefficient of performance may be defined as:
Overall COP =
Heat transfer in the evaporator

Total energy consumptio n of the refrigeration unit
(18.16)
Care should, therefore, be taken that any quoted values of COP are interpreted in the correct
manner.
18.3. COMPONENTS AND DESIGN OF MINE COOLING SYSTEMS

18.3.1. Overview of mine cooling systems
In the majority of large scale mine cooling systems, there are usually three sets of heat transfer
involved:
transfer of heat from the work areas to the evaporators of the refrigeration units
transfer of heat from the evaporators to the condensers in the refrigeration units (the vapour
compression cycle)
transfer of heat from the condensers to the free atmosphere on surface.
There exist tremendous variations in the duty, complexity and efficiency of the hardware involved in
these three phases, dependent upon the severity and dispersion of the heat problem in the mine. In
this section we shall discuss a range of systems varying from a simple spot cooler for very localized
applications to large integrated systems that may be required for deep and hot mines. This will serve
both as an overview and an introduction to a more detailed examination of individual components and
overall system design.
Figure 18.8 shows a simple application of a direct evaporator or "spot cooler" to a specific working
area in a mine. The evaporator of the refrigeration unit takes the form of copper or cupronickel tubular
coils located within an air duct. The refrigerant passes through the inside of the evaporator tubes and
cools the air flowing along the duct. The heat from the condenser is rejected directly to the return air.
This system has the advantage that the cooling effect produced by the refrigeration unit is utilized
directly and immediately at the place where cooling is required. There is no loss of efficiency
introduced by an intervening water reticulation system between the evaporator and working area.
Similarly, heat rejection from the condenser is direct and immediate. The spot cooler is merely an
industrial version of a domestic air conditioning unit where heat is rejected directly to the outside
atmosphere.
Unfortunately, the spot cooler has a major disadvantage that limits its application in subsurface
ventilation systems. Glancing, again, at the blind heading illustrated in Figure 18.8, it is clear that the
air emerging from the end of the duct will be at a reduced temperature when the refrigeration unit is
operating. Although this is the desired effect, a consequence will be that the flow of strata heat into
the heading will increase due to the lowered air temperature, a phenomenon that is examined in
Chapter 15. Yet more heat is added as the air returns over the condenser. The net effect is that the
return air leaving the area has a greater enthalpy (and, usually, temperature and humidity) than would
18- 15
Chapter 18
be the case if the refrigeration unit were not operating. The increase in enthalpy is the sum of the
additional strata heat and the energy taken by the compressor. The use of spot coolers is restricted
by the availability of return air and the debilitating effect on psychrometric conditions within local
return airways.
condenser
evaporator
fan
fan
Figure 18.8 The simplest application of a direct evaporator or spot cooler with local
heat rejection in a mine heading.
There are three ways of alleviating the situation. First, the condenser could take the form of a shell
and tube heat exchanger and be cooled by a water circuit (Figure 18.3). The hot water can then be
piped away and recirculated through a heat exchanger in a main return airway. Secondly, the
refrigeration unit itself may be located away from the working area and water that has been chilled by
the evaporator piped to a heat exchanger in the workings. Thirdly, the first two arrangements may be
combined, resulting in the system illustrated in Figure 18.9. Here, the refrigeration plant of, perhaps, 2
MW cooling duty, is sited in a stable location and provides chilled water for a number of work areas.
The chilled water flows through thermally insulated pipes to heat exchangers in the faces, stopes or
headings and returns to the plant via uninsulated pipes. Some of the chilled water may be used for
18- 16
Chapter 18
dust suppression purposes in which case additional make-up water must be supplied to the plant. Hot
water from the condenser is recirculated through heat exchangers that are located in a main return
airway. If that return airway is to remain open for persons to travel, either for routine purposes or as
an emergency escapeway, then the psychrometric conditions must remain physiologically acceptable
(Chapter 17). Again, this provides a limit on the degree of heat rejection and, hence, size of
refrigeration plant that can be utilized as a district cooler. If, however, a return route can be dedicated
fully to heat rejection, then physiological acceptability limits may be exceeded. In this case, the
dedicated return or "dirty pipe", as it is sometimes called, must be made inaccessible to inadvertent
entry by personnel. Inspection or maintenance can be carried out either when the plant is shut down
or by persons wearing protective clothing (Section 17.5.6). It may be necessary to seek exemption
from national or state legislation in order to utilize a "dirty pipe" arrangement.
air
coolers
chilled
water
intake
chiller
plant
hot
water
heat
rejection
return
Figure 18.9 A district chiller supplying cold water to air coolers in headings, stopes or faces and rejecting
heat into a return airway.
For mines that have a widespread heat problem, the economies of scale and the need for flexibility
indicate a requirement for large centralized cooling facilities. Banks of individual refrigeration units,
each producing a typical 3.5 MW of cooling capacity, may be assembled to give a total duty which
might exceed 100 MW for a large and deep metal mine. The "coolth" is normally distributed via chilled
water lines to provide any required combination of bulk air cooling, face or stope air cooling and
chilled service water.
18- 17
Chapter 18
Until the mid-nineteen seventies, centralized plant tended to be located underground in excavated
refrigeration chambers close to shaft bottoms. All of the mine return air could then be utilized for heat
rejection. In South Africa, it was common for large cooling towers to be situated in, or adjacent to, the
upcast shaft bottoms. Chilled water from the centralized underground plant could be transmitted to
other levels. However, at elevations greater than some 500 m below the plant, water pressures in the
pipe ranges become excessive. This can be counteracted either by pressure reducing valves
(adjustable orifices) or by employing water to water heat exchangers and secondary, low pressure,
cold water circuits. Unfortunately, there are difficulties associated with both of these types of device.
Pressure reducing valves necessitate the employment of high duty pumps to raise the heated water
back to the plant elevation. Maintaining the high pressure water in a closed circuit by means of water
to water heat exchangers balances the pressure heads in the supply and return shaft ranges.
However, the intermediate water to water heat exchangers produce an additional loss of heat transfer
efficiency between the work areas and the refrigeration plant. Furthermore, they are a further source
of potential corrosion and fouling of internal tubes.
In the mid 'seventies a number of factors coincided to promote a trend towards the location of
centralized refrigeration plant at the surface of deep and/or hot mines. First, there was a significant
drop in the wet bulb temperatures at which mining workforces would be expected to work. Cut-off wet
bulb temperatures have been reduced from 32C to 28C with an expectation of further improvements
to 27C or less. Secondly, the combination of greater depths of workings and more intensive
mechanization produced higher heat loads to be handled by the mine environmental control system.
These factors combined to give very significant increases in the required cooling capacities of mine
refrigeration plant, to the extent that untenable and uneconomic quantities of return air would be
required for heat rejection. There was no choice but to locate the larger new plant on surface where
heat rejection is relatively straightforward. Thirdly, the problem of high pressures developing in the
shaft pipes was combatted by the development of energy recovery devices including water turbines
(Section 18.3.4).
A simplified schematic of one arrangement involving surface refrigeration plant is indicated on Figure
18.10. The figure illustrates water being cooled by a surface refrigeration plant and collected
temporarily in a cold water dam. The hot water from the condensers is cycled around the condenser
cooling tower for heat rejection to the atmosphere. If a natural supply of sufficiently cool water is
available from a stream or river, then there may be no need for the capital and operating expense of
refrigeration plant. Alternatively, in dry climates, a modest required degree of cooling may be
achievable simply by spraying the water supply through a surface cooling tower. Furthermore, it may
be necessary to operate the refrigeration plant during the summer months only.
The water passes from the surface cold water dam to an insulated shaft pipe through which it falls to
the working levels. The water may then be passed through a turbine at one or more subsurface levels
before being stored temporarily in underground cold water dams. The turbine(s) achieve three
beneficial results. First, the pressure of the water is reduced to that of the mine atmosphere at the
corresponding level. Hence, no extensive high pressure water systems need exist in the workings.
Secondly, the mechanical output power produced by the turbines may be employed directly to assist
in driving the pumps that raise the return hot water to surface. However, because the demand for cold
water may be out of phase with the availability of hot return water, it is preferable to use the turbines
to drive generators for the production of electrical power. Third, the removal of energy from the water
by the turbines results in a reduced temperature rise of that water, improving the cooling efficiency of
the system.
From the underground cold water dams, the chilled water may be utilized for a variety of purposes
including water to air heat exchangers for cooling the air at the entrance to a stope or face, or for
recooling at intermediate points along the stope or face. Water to water heat exchangers may be
employed to cool secondary water circuits such as a supply of potable drinking water or dust
suppression service water. However, the modern trend is to supply chilled service water directly from
the cold water dam.
18- 18
Chapter 18
condenser
cooling
tower
pre-cooling
tower
refrigeration
plant
make-up
water
condenser
evaporator
hot water dam
cold water dam
turbine
LEVEL X
electric
power
chilled service
water
hot water dam
cold water dam
hot water return

from workings
chilled water to heat

exchangers in mine
workings
turbine
LEVEL Y
electric
power
chilled service
water
cold water dam

chilled water to heat
exchangers in mine
workings
hot water dam
hot water return

from workings
hot water pipe
pump
insulated chilled
water pipe
valve
Figure 18.10 Schematic of centralized plant on surface supplying chilled water for
underground heat exchangers and mine service water.
18- 19
Chapter 18
Warm water returning from the heat exchangers or via drainage channels is directed into hot water
dams adjacent to a shaft. The pumps that return this water to surface may be powered partly by the
energy recovered by the turbines. Where surface refrigeration plant is in use, the hot water from the
mine may be sprayed through pre-cooling towers prior to its return to the plant evaporators. This
gives an additional low cost supplement to the cooling capacity of the system. Any required make-up
water and anti-corrosion chemicals are added at this stage.
The purpose of the dams on both the cold and hot water sides of the layout is to provide capacitance
to the system. This permits short term fluctuations in demand for chilled water while using smaller
refrigeration plant than would otherwise be necessary. At times of low demand, the temperature of
the water in the surface dams can be further reduced by recycling that stored water through the plant.
This smoothes the variations in cooling load required of the refrigeration units.
Having introduced the broad concepts of subsurface cooling systems, we are now in a better position
to examine the component parts in greater detail.
18.3.2. Heat exchangers
In general, a heat exchanger may be defined as a device that facilitates the transfer of thermal energy
from one solid or fluid system to another. In subsurface air conditioning systems, there are two
classifications of heat exchanger in common use, both involving heat transfer between fluids.
An indirect heat exchanger promotes heat transfer across a solid medium that separates the two
fluids. There is no direct contact between the fluids. Examples include:
the shell and tube evaporators and condensers of refrigeration units (Figure 18.3)
shell and tube water to water heat exchangers employed to transfer heat between water
systems of differing pressure and/or water quality (Figure 18.11)
tubular coil heat exchangers to transfer heat from air to water (Figure 18.12).
As the term implies, direct heat exchangers involve direct contact between the two fluids. Cooling
towers and other types of spray chambers fall into this category. The objective may be to reject heat
from the water to the air as in a conventional cooling tower. Conversely, a spray chamber supplied
with chilled water provides a means of cooling an airstream.
The higher efficiency of heat transfer associated with direct heat exchangers caused a distinct trend
away from tube coil air coolers during the nineteen eighties. Good performance of large horizontal
spray chambers for bulk air cooling promoted the further development of compact and enclosed spray
chambers for more localized use.
In this section we shall examine the essential structure and operating principles of both indirect and
direct forms of heat exchanger.
18- 20
Chapter 18
primary chilled
water circuit
secondary
water circuit
Figure 18.11 A water-to-water heat exchanger
chilled water
Airflow
Figure18.12 An air-to-water heat exchanger showing one layer of tubes.
18- 21
Chapter 18
18.3.2.1. Indirect heat exchangers

A shell and tube heat exchanger may contain over 200 tubes. In a counter-flow heat exchanger, the
fluids inside and outside the tubes move in opposite directions, maximizing the total heat transfer.
However, the water to water heat exchanger illustrated in Figure 18.11 shows that this may be
sacrificed in the interest of compactness.
In an air to water indirect heat exchanger, a bank of tubes carrying chilled water is located within an
air duct. Figure 18.12 shows one layer in a bank of tubes. The tubes are often inclined to the direction
of airflow to improve the efficiency of the unit and to control the movement of condensate.
The two important features of a heat exchanger are that a good efficiency of heat transfer is obtained
and that the pressure drops suffered by both fluid streams should remain within acceptable limits.
Heat transfer is facilitated by choosing a tube material that has a high thermal conductivity and is also
resistant to corrosion and the build-up of scale deposits within the tubes. Chemical additives can be
used to control such fouling of the tubes. Copper tubes are commonly employed for air to water heat
exchangers. For water to water heat exchangers and in the evaporator and condensers of
refrigeration units, cupronickel and stainless steel are more resistant to corrosion. Even titanium may
be used within evaporators and condensers.
If tubes are not maintained in a clean condition then the efficiency of the unit can fall dramatically.
Periodic manual brushing of the tubes is time-consuming and may be replaced by a variety of
automatic tube cleaning devices including reverse flushing with brush inserts or ultrasonic vibration.
Air to water tube coils, as illustrated on Figure 18.12, are subject to caking by dust deposits in mine
atmospheres. This may be handled by periodic cleaning with a high pressure water jet. Alternatively,
sprays may be located permanently upstream from the coils. In humid atmospheres, the coils may be
made self-cleaning by arranging them into horizontal banks with air passing upwards over the tubes.
The droplets of condensate fall back through the coils giving a continuous and oscillating cleaning
action on the outer surfaces of the tubes.
A second major factor that controls the duty of a heat exchanger is the effective area available for
thermal transfer. Spiral fins welded to the tubes, as illustrated on Figure 18.12, may be used. Metal
plates welded between consecutive and partially flattened tubes have also been used (Mcke, 1984;
Weuthen, 1975). It is important that such means of area enlargement should have good thermal
contact with the tubes and that they are oriented such that they present minimum resistance to flow
over the tubes. In evaporators and condensers, the tube surfaces may be knurled or sintered. In
addition to enhancing the surface area, this assists in the promotion of boiling or condensation.
18.3.2.2. Performance calculations for indirect heat exchangers

At equilibrium, the heat gained by one fluid in a heat exchanger must equal the heat lost by the other
fluid. Hence, taking the example of an air to water heat exchanger, we can express the rate of heat
transfer to be
q = mw C w t w
where mw
Cw
tw
ma
S
=
=
=
=
=
= m a S
(18.17)
mass flow of water (kg/s)

specific heat of water (4187 J/(kg C)
rise in temperature of the water (C)
mass flow of air (kg/s)
fall in sigma heat of the air (J/kg) (ref. section 14.6).
18- 22
Chapter 18
As each of the factors in equation (18.17) can be calculated easily from measured temperatures and
flowrates, either the water or the air may be used to determine the rate of heat transfer, q. As
measurements on the water circuit can normally be made more accurately than those on the airflow,
the former are preferred for a determination of heat transfer. If condensation occurs on the outside of
the coils then the third part of equation 18.17 is slightly in error as it does not take into account the
heat removed from the system by the condensate. However, this is usually small and may be
neglected.
Another way of expressing the heat transfer is in terms of an overall heat transfer coefficient, U,
(W/m2C) for the coils and adjacent fluid boundary layers, and the difference between the mean
temperatures of the air and the water, (tma - tmw), giving
= UA (tma - tmw)
where
(18.18)
A = area available for heat transfer. (This is analogous to equation (15.16) for a rock surface.)
As the temperatures of both the air and water streams are likely to vary in a logarithmic rather than a
linear fashion through the heat exchanger, it is more accurate to employ logarithmic mean
temperature difference in equation (18.18), giving
UA (t 1 t 2 )
W
(18.19)
ln (t 1 t 2 )
where ln indicates natural logarithm and t1 and t2 are the temperature differences between the
fluids at each end of the heat exchanger.
q =
The UA product is usually quoted as a measure of the effectiveness of an indirect heat exchanger.
A performance check should be carried out at monthly or three monthly intervals in order to determine
any deterioration in the UA caused by corrosion, scaling, or other forms of deposition on the tube
surfaces.
Example
Measurements made on a counterflow air to water cooling coil give the following results.
Air:
Inlet wet bulb/dry bulb temperature

Outlet wet bulb/dry bulb temperature
Mass flowrate
Barometric pressure
Water: Inlet temperature

Outlet temperature
Mass flowrate
=
=
=
=
28/34C
22.7/22.7C
5.1 kg/s
105 kPa
= 17 C
= 23 C
= 4 kg/s
Determine the operating duty of the coil and the UA value.

Solution
From the psychrometric equations given in section 14.6, the sigma heats of the air at inlet and outlet
conditions of 28/34 and 22.7/22.7C, respectively, are found to be
Sin
and
Sout
= 84 696 J/kg
= 63 893 J/kg
18- 23
Chapter 18
From equation (18.17) the heat lost from the air is:
qa
= ma (Sin - Sout)
5.1 (84 696 - 63 893) = 106.1x103 W or 106.1 kW.
Also, from equation (18.17), the heat gained by the water is:
qw
mw Cw tw =
4 x 4187 x (23-17) = 100.5x103
W or 100.5 kW.
At equilibrium, qa and qw must be equal, showing that the errors in measurements have caused a
deviation of some 5.5 percent. Much larger discrepancies often occur due mainly to difficulties in
making measurements in highly turbulent and, often, saturated airflows. We shall continue the
calculation using the rate of heat transfer given by the water circuit, 100.5 kW.
In order to calculate the UA value from equation (18.19), we must first evaluate the temperature
differences between the water and air at the air inlet and air outlet ends (subscripts 1 and 2,
respectively).
t 1 = t a,in t w ,out = 34 23 = 11
t 2 = t a,out t w ,in = 22.7 17 = 5.7
Notice that we use the dry bulb temperature at the air inlet and before saturation conditions are
attained for the determination of t1. This is equivalent to taking the latent heat of condensation as a
factor contributing towards the UA value of the heat exchanger.
Equation (18.19) now gives
UA = q w
ln (t 1 / t 2 )
ln (11 / 5.7 )
= 100.5
= 12.47 kW/C
(11 5.7)
(t 1 t 2 )
The UA of a clean coil may lie between 10 and 25 kW/C depending upon the design of the heat
exchanger and the configuration of fluid flows. Records should be kept of the periodic performance
tests on each heat exchanger. Significant reductions in UA values indicate the need for cleaning or
replacement of the tubes.
Although UA values are normally determined by measurement as illustrated in the example, they may
also be defined by the following equation:
x
1
1
1
1
1
=
+
+
+
+
UA
h i Ai
hfi Ai
k t Am
h o Ao
hfo Ao
where
h
A
x
kt
=
=
=
=
(18.20)
heat transfer coefficients (W/(m2C)

area available for heat transfer (m2)
thickness of tube walls (m)
thermal conductivity of tube material (W/(mC)
and subscripts
i = inside surface of tubes
o = outside surface of tubes
m = mean of inner and outer surfaces.
18- 24
Chapter 18
hfi and hfo are the heat transfer coefficients associated with fouling (deposits) on the inside and
outside surfaces of the tubes respectively. On a clean tube there are no such deposits; hfi and hfo are
both then infinite.
Radiation terms have been left out of equation (18.20) as these are normally small in a heat
exchanger. They may, however, become significant in situations such as an uninsulated pipe
suspended in an airway.
Equation (18.20) can be further simplified for practical application. First, the term x / (ktAm) is very
small compared with the others. Secondly, the terms involving fouling of the tubes are often combined
and attributed to the inside surface only, giving
1
1
1
1
=
+
+
UA
h i Ai
h o Ao
hf Ai
C/W
(18.21)
As fouling of the tubes occurs, hf will decrease causing the UA value also to fall. Whillier (1982)
quotes a typical value of some 3000 W/(m2C) for hf in a mine refrigeration plant. For turbulent flow
inside smooth tubes, the heat transfer coefficient can be determined from equation (15.23) and the
Colburn equation (A15.15)
hi = 0.023
where k =
d =
Re =
=
u =
=
Pr =
and Cp =
k
Re0.8 Pr 0.4
d
W/(m2 C)
(18.22)
thermal conductivity of fluid ( W/(mC) )

internal diameter of tube (m)
Reynolds number = ud/ (dimensionless)
fluid density (kg/m3)
fluid velocity (m/s)
dynamic viscosity (Ns/m2)
Prandtl number = Cp /k (dimensionless, may be taken as 0.7 for air)
Specific heat at constant pressure ( J/(kgC) )
For air and water, the values of viscosity and thermal conductivity within the temperature range 0 to
60C are given in sections 2.3.3 and 15.2.4 respectively. Other expressions for the heat transfer
coefficient at rough surfaces are given in Table A15.1.
The values of ho for the outer surfaces of the tubes vary widely according to geometry and the
configuration of tubes. For turbulent cross flow of air over tubular surfaces, McAdams (1954) gives:
for a single tube:
ho
= 0.24
ka
Re 0.6
d
and for a bank of staggered tubes:

k
ho = 0.29 a Re 0.6
d
W
m
2 o
W
m
2 o
(18.23)
(18.24)
In these relationships, d is the outer diameter of the tubes and Reynolds' Numbers are determined on
the basis of the maximum velocity of the air as it flows between the tubes. The selection of cooling
coils is facilitated greatly by tables and graphs provided by manufacturers.
18- 25
Chapter 18
18.3.2.3. Direct heat exchangers

In direct heat exchangers, air is brought into contact with water surfaces. The purpose may be to cool
water from the condensers of a refrigeration plant. In this case the hot water is sprayed into a cooling
tower and descends as a shower of droplets through an ascending airstream. Heat is transferred from
the water droplets to the air by a combination of convection (sensible heat) and evaporation (latent
heat). The cooled water that collects at the base of the cooling tower is then returned to the
condenser (see the condenser cooling tower on Figure 18.10).
Alternatively, the objective may be to cool the air. In this case, chilled water is sprayed into a vertical
or horizontal spray chamber. Provided that the airflow enters with a wet bulb temperature that is
higher than the temperature of the water, then heat will be transferred from the air to the water by a
combination of convection and condensation.
Although there are significant differences between the designs of cooling towers and spray coolers,
there are several common factors that influence the amount and efficiency of heat exchange:
water mass flowrate

supply temperature of water
air mass flowrate
psychrometric condition of the air at inlet
duration and intimacy of contact between the air and the water droplets.
This latter factor depends upon the physical design of the heat exchanger. In particular,
(i)
and
(ii)
the relative velocity between the air and water droplets

the size and concentration of water droplets - governed by the flow and pressure of the
supply water, the type and configuration of spray nozzles, and the presence of packing
within the heat exchanger.
The traditional use of direct heat exchangers in mine air conditioning systems has been for cooling
towers either underground or on surface. Tube coil heat exchangers with closed circuit water systems
have been used extensively for air cooling in or close to mine workings. However, the greater
efficiency of direct heat exchangers led to the development of large permanent spray chambers for
bulk air cooling. Furthermore, through the nineteen eighties, smaller portable spray chambers began
to be employed for local cooling.
Cooling Towers
First, let us examine the essential features of cooling towers. Figure 18.13 illustrates a vertical forced
draught cooling tower of the type that may be used on the surface of a mine. Hot water from
refrigeration plant condensers is sprayed into the cooling tower and moves downward in counterflow
to the rising air current. The purpose of the packing is to distribute the water and airflow uniformly
over the cross section and to maximize both the time and total area of contact between the air and
water surfaces. The packing may take the form of simple splash bars or riffles arranged in staggered
rows, egg-crate geometries or wavy (film-type) surfaces located in vertical configurations. The
materials used for packing may be treated fir or redwood timber, galvanized steel, metals with plastic
coatings and injection molded PVC or polypropylene. Concrete is used primarily for casings,
structural reinforcements and water sumps or dams. Air velocities through counterflow packed cooling
towers lie typically in the range 1.5 to 3.6 m/s.
18- 26
Chapter 18
heated air
fan
mist
eliminator
sprays
hot water in
packing
airflow
water droplets
air in
cooled water out
Figure 18.13 Counterflow cooling tower. (The same
configuration may be used to cool air by
supplying the tower with chilled water.)
In underground cooling
towers, the packing is often
eliminated completely or
takes the form of one or two
screens arranged
horizontally across the tower.
Such heat exchangers are
essentially vertical spray
chambers. Other means of
maximizing contact time are
then employed. These
include designs that
introduce a swirl into the air
at entry, and directing the
sprays upward rather than
downward. Again, in
underground installations,
airflow is induced through the
cooling towers by the mine
ventilating pressure or by
booster fans in return
airways rather than by fans
connected directly to the
tower. The pressure drop
through underground cooling
towers may be further
reduced by replacing the
mist eliminator screens with
an enlarged cross-sectional
area; the lower air velocity in
this zone decreases the
carry-over of water droplets.
The optimum air velocity in
open spray towers lies in the
range 4 to 6 m/s with a
maximum of some 8 m/s.
Water loadings should not be
greater than 16 litres/s per
square metre of crosssectional area (Stroh, 1982).
Cooling towers of the type used for mine air conditioning are typically 10 to 20 m in height and some
3 to 8 m in diameter, depending upon the rate at which heat is to be exchanged. Heat loads may be
as high as 30 MW.
Natural draught cooling towers do not have fans but rely upon air flow induction by the action of the
sprays, or by density difference between the outside atmosphere and the hot moist air within the
tower. The hyperbola shaped cooling towers employed commonly for the high heat loads of power
stations are of this type and may be over 150 m high in order to accentuate the chimney effect.
The air leaving a cooling tower is usually saturated. This often results in the formation of a fog plume
in the surface atmosphere. The environmental effects of such a plume should be considered carefully
and may influence the siting of a surface cooling tower. In large installations with unacceptable fog
plumes, part of the hot condenser water may be cooled within a finned tube indirect heat exchanger.
This involves sensible heat exchange only and reduces the humidity of the air leaving the cooling
tower. However, such an arrangement detracts from the overall efficiency of heat transfer.
18- 27
Chapter 18
In a direct exchange cooling tower, some water is lost continuously from the circuit both by
evaporation and by drift (or carry-over) of small droplets. The evaporation loss approximates some
one percent for each 7C of water cooling and drift loss is usually less than 0.2 percent of the water
circulation rate (ASHRAE, 1988). However, evaporation can result in a rapid escalation in the
concentration of dissolved solids and other impurities leading to scaling, corrosion and sedimentation
within the system. In order to limit the buildup of such impurities, some water is continuously removed
from the system (bleedoff or blowdown). The bleedoff rate is controlled by monitoring the quality of
the water but may be of the order of one percent of the circulation rate. The combined losses from
evaporation, drift and bleedoff are made up by adding clean water to the circuit.
Further protection of metal components and, particularly the tubes of condensers is obtained by
administering anticorrosion compounds. These generally take the form of chromates, phosphates or
polyphosphonates of zinc, and promote the formation of a protective film on metal surfaces. Some of
these compounds are toxic and precautions may be necessary against accidental release into natural
drainage systems. Biocides such as chlorine are also added on a periodic basis to control the growth
of algae and other organic slimes.
Chilled water spray chambers
If the water supplied to a direct contact heat exchanger is at a temperature below that of the wet bulb
temperature of the air, then cooling and dehumidification of the air is achieved. Chilled water spray
chambers fall into two categories with respect to size. First, the larger installations are constructed at
fixed sites for bulk cooling of main airflows. Secondly, portable spray chambers for localized use
within working areas have many advantages over the more traditional tube coil (indirect) stope or face
coolers. We shall consider each of these two applications in turn.
For bulk air cooling, spray chambers may be designed in vertical or horizontal configuration. Indeed, if
the cooling tower shown on Figure 18.13 were supplied with chilled water, then it would act equally
well as a vertical spray air cooler. Such designs may be employed either on surface or underground
for bulk cooling of intake air and may have heat transfer duties up to 20 MW.
Horizontal spray chambers have more limited maximum capacities of some 3.5 MW. They are,
however, more convenient for underground use in that existing airways may be utilized without
additional excavation. Figure 18.14 illustrates a single stage horizontal spray chamber. The sprays
may be directed against or across the airflow. The nozzles can be distributed over the cross section
as shown in the sketch or, alternatively, located at the sides or near the base of the chamber.
Although the position of the nozzles appears not to be critical, it is important that both the sprays and
the airflow are distributed uniformly over the cross section. The spray density should lie within the
range 2 to 5 litres/second for each square metre of cross sectional area (Bluhm, 1983).
The area of liquid/air interface and efficiency of heat exchange increase for smaller sizes of droplets.
However, very small droplets result in excessive carry-over or necessitate highly constrictive mist
eliminators. Furthermore, higher water pressures and, therefore, pumping costs are required to
produce fine sprays. In practice, droplet diameters of some 0.5 mm and water pressures in the range
150 to 300 kPa give satisfactory results in horizontal spray chambers (Reuther, 1988).
At positions fairly close to the nozzles, the relative directions of the air and water droplets may be
counterflow or crossflow, dependent upon the orientation of the nozzles. However, aerodynamic drag
rapidly converts the spray to parallel flow, particularly for the smaller droplets. In order to regain the
higher efficiency of counterflow heat exchange, the spray chamber may be divided into two or three
stages so that the air leaves the chamber at the position of the coldest sprays. Chilled water should
be supplied at as low a temperature as practicable but, in any event, not higher than 12C.
18- 28
Chapter 18
chilled water inlet
mist
eliminator
cooled air
warm air
water return
water sump
Figure 18.14 A single stage , horizontal spray chamber.
Figure 18.15 illustrates a two stage spray chamber. An additional sump at each end assists in
balancing the pumping duties.
If the surrounding strata are unfractured and unaffected by water, then the spray chamber may utilize
the full cross section of a bypass airway. If, however, the rock must be protected against the effects of
water, then concrete lining or prefabricated sections may be employed to contain the spray chamber.
In cases where the strata is very sensitive to water (such as evaporite strata) then it is advisable to
protect the rock surfaces for 50 to 100 m downstream from the spray chamber.
The cross-sectional area of the spray chamber should be chosen to give a preferred air velocity of
some 4 to 6 m/s, but not more than 7 m/s. Higher air velocities will reduce the efficiency of heat
exchange and can result in excessive pressure drops in the airflow.
In addition to cooling and dehumidifying the air, spray coolers can also reduce dust concentrations.
However, the buildup of dust particles in the recirculating water system may cause fouling of the
pipes and other heat exchangers. This may require filters or sedimentation zones to be included in
the design.
Internal packing can also be employed to improve the efficiency of horizontal spray chambers (Stroh,
1980). This can increase the air frictional pressure drop across the cooler. However, the water supply
pressure (gauge) may be as low as 30 kPa since the spray nozzles can be replaced by low
resistance dribbler bars.
In order to extend the applicability of direct spray coolers, enclosed and portable units have been
developed (Thimons, 1980; Ramsden, 1984; Reuther, 1988). These may be mounted on wheels or
sleds and are often referred to as spray mesh coolers. Figure 18.16 illustrates the principle of
operation. In order to maintain the design to acceptable dimensions for portable application, the water
loading may be much higher than open spray chambers and it becomes even more important to
maximize the area and time of contact between the air and water. This occurs in three stages. First,
the air passes through the lower mesh of plastic or knitted stainless steel. Secondly, the airflow is
directed through the upward pointing and finely divided sprays. The final stage of cooling occurs
within the upper mesh which also acts as a demister. The distribution of droplets of water falling from
the upper to the lower mesh may be assisted by installing drip fins. These also help to prevent water
running down the walls (Heller, 1982). Between the two meshes, heat exchange occurs by a
combination of counterflow, crossflow and parallel flow. Another type of portable spray cooler
employs film packing and crossflow heat exchange (Reuther, 1988).
18- 29
Chapter 18

mist
eliminator
1st stage
2nd stage
cooled air
warm air
chilled water inlet

water return
high level
control valve
pump with low

level switch
pump
Figure 18.15 A two-stage horizontal spray chamber
cool air
demister
sprays
chilled water
mesh
warm air
guide vanes
lattice
return
water
Figure 18.16 A section through a portable mesh cooler.

18- 30
Chapter 18
The dry bulb temperature of air can be reduced by passing it through a spray chamber supplied with
unchilled water. The device is then known as an evaporative cooler. No heat is removed from the air.
Hence, if the water is supplied at wet bulb temperature, then the exhaust air may attain that same wet
bulb temperature while the moisture content will have increased. Evaporative coolers have an
application for surface structures in hot dry climates but are seldom used in underground mines. They
have been employed for very localized effects in hot evaporite mines. However, the reduced dry bulb
temperature encourages enhanced heat flow from the strata (Section 15.2.2) which, when combined
with the raised moisture content, reduces the cooling power of the air downstream from the cooled
area.
18.3.2.4. Performance calculations for direct heat exchangers.
A common theoretical analysis may be applied to direct heat exchangers irrespective of the direction
of heat transfer. The results of such analyses apply equally well to cooling towers and chilled water
spray chambers. In order to avoid unnecessary repetition, we will conduct the following analysis on
the assumption of a cooling tower (Figure 18.13). Water recycles continuously through the condenser
of a refrigeration plant (where it gains heat) and the cooling tower where it rejects that heat to the
atmosphere.
The first observation is that if we ignore the small heat losses from interconnecting pipes and the
equally small effects of make-up water, then, at steady state, the heat rejected in the cooling tower
must be equal to the heat gained in the condenser. This leads to the initially surprising statement that
the rate of heat rejection in the cooling tower depends only upon the heat load imposed by the
condenser and not at all upon the design of the cooling tower. However, if the cooling tower is
inefficient in transferring heat from the water to the air, then the temperature of the water throughout
the complete circuit will rise until balance is attained between heat gain in the condenser and heat
rejection in the cooling tower. This would be unfortunate as the coefficient of performance of the
refrigeration plant deteriorates as the condenser temperature increases (Section 18.2.4).
Figure 18.17 illustrates the decrease in temperature of the water as it falls through the cooling tower
and the corresponding increase in the wet bulb temperature of the ascending air.
Figure 18.17 also defines the meaning of two terms that are commonly employed:
The range is the change in temperature of the water
range
= tw,in - tw,out
(18.25)
while the approach is the difference between the temperatures of the water outflow and the wet bulb
temperature of the air inflow:
approach
= tw,out - ta,in
where
t
and subscripts w
a
in
and
out
=
=
=
=
=
(18.26)
temperature, C
liquid water
air
(wet bulb)
inflow
outflow
Manufacturers will usually accept requests for approach values down to 2C.
18- 31
Chapter 18
36
34
tw,in
30
ta,out
28
water
26
24
approach
Temperature
range
32
22
tw,out
20
ta,in
air
18
0
10
12
Distance through cooling tower
Figure 18.17 Variation of water temperature, tw, and wet bulb temperature of the air, ta,
through a cooling tower.
The quantitative analysis of the cooling tower commences by writing down the balance that must exist
between rates of heat gained by the air and heat lost from the water:
ma (Sout - Sin)
= mw Cw (tw,in tw,out)
(18.27)
where m = mass flow, kg/s

S = sigma heat of air, J/kg
and Cw = specific heat of water (4187 J/kg C)
This is, in fact, an approximation as evaporation changes the value of mw within the tower. However,
the error does not normally exceed 4 percent and may be neglected for practical purposes.
18- 32
Chapter 18
As in so many cases of assessing performance, it is useful to imagine the unattainable perfect cooling
tower. In such a device, the two curves on Figure 18.17 would coincide and, in particular, the water
would leave at inlet air wet bulb temperature while the air would leave at the temperature of the
incoming water, that is:
=
=
tw,out
ta,out
and
ta,in
tw,in
(18.28)
(18.29)
The concept of a perfect cooling tower allows us to devise efficiencies of heat transfer for both the
water and the air streams:
water efficiency
w =
=
actual heat lost from the water

theoretical maximum heat that could be lost from the water
mw C w (t w ,in t w ,out )
mw C w (t w ,in t a,in )
(t w ,in t w ,out )
(t w ,in t a,in )
range
range + approach
(18.30) (See Figure 18.17)
Similarly, air efficiency
a =
=
where
actual heat gained by the air

theoretica l maximum heat that could be gained by the air
(Sout S in )
m a (S out S in )
=
m a (Sw ,in S in )
(Sw ,in S in )
Sw,in
(18.31)
sigma heat of saturated air at a temperature equal to that of the inlet water, J/kg.
[The term "thermal capacity" is sometimes employed where we have used "theoretical maximum heat".]
Although the water and air efficiencies are useful indicators of the efficacy of heat transfer for each of
the two fluids, neither gives any clue to the overall quality of design for the complete cooling tower.
For example, if the water flow were low giving relatively few small droplets falling through a large
airflow, then the water efficiency would be very high. However, the air would be used to only a small
fraction of its thermal capacity and, hence, the air efficiency and overall heat transfer would be low.
To examine this problem let us consider, again, perfect heat transfer in a cooling tower. Then,
maximum thermal capacity of water
and
maximum thermal capacity of airstream
mw Cw (tw,in - ta,in)
=
ma(Sw,in - Sin) J
The ratio of these maximum thermal capacities defines the tower capacity factor, R
R =
(t w ,in t a,in )
mw
Cw
(Sw ,in S in )
ma
(dimensionless)
(18.32)
The liquid to gas ratio, mw/ma normally lies within the range 0.4 to 2.0.
The theoretical concept of tower capacity factor ignores the practical reality that in a real cooling
tower the rate of heat loss from the water must equal the rate of heat gain by the air. It may be
defined in words as follows:
18- 33
Chapter 18
'If both of the given fluid streams were used to their maximum thermal capacity, then
R would be the number of Watts lost by the water for each Watt gained by the air.
An important observation here is that R depends only upon the flowrates and the inlet conditions of
the water and air. It is completely independent of the construction of the cooling tower. Notice also
that we have been quite arbitrary in choosing to define R as the ratio of water thermal capacity
divided by air thermal capacity, rather than the other way round. The values of tower capacity factor
vary from 0.5 to 2.
The tower capacity factor is also related to the ratio of air and water efficiencies.
From equations (18.30 and 18.31)
a
w
(S out S in ) (t w , in t a, in )
(Sw ,in S in ) (t w , in t w , out )
and substituting for
(S out
S in ) =
a
w
mw
C w t w , in t w , out
ma
(
(
t w , in t a, in
mw
Cw
ma
Sw , in S in
from equation (18.27) gives
)
)
(18.33)
As this is identical to equation (18.32), then tower capacity factor

R =
a
w
(18.34)
indicating that the ratio of air and water efficiencies is also independent of the design of the cooling
tower or spray chamber.
The overall effectiveness of heat transfer within the cooling tower must be dependent on the fluid
having the smaller thermal capacity, i.e. the fluid having the larger efficiency. To quantify this concept,
we can say that the Effectiveness, E, is the larger of the values of a and w. Another way of
expressing this, using equation (18.34), is:
E = a if R > 1
and
E = w if R < 1
(18.35)
a and w are, of course, equal when R = 1.

Prior to 1977, there seems to have been no one single parameter that could be used to describe the
quality of design for any given cooling tower. The rate of heat transfer, air and water efficiencies, and
the tower capacity factor are all dependent on inlet temperatures. It is, therefore, somewhat
misleading to quote a single value of cooling duty to characterize a heat exchanger.
In the mid 1970's, Whillier of the South African Chamber of Mines conducted an analysis on results
published earlier by Hemp (1967 and 1972) relating to a series of tests during which a spray tower
was successively supplied with a range of inlet water temperatures. Whillier (1977) noticed that two
distinct curves were produced when (a) water efficiency and (b) air efficiency were plotted against
tower capacity factor. Furthermore, these two curves appeared to be mirror images of each other. In
a moment of inspiration he recalled that the order of the ratio that we choose to define R is quite
arbitrary (equation (18.32). Whillier took the values of tower effectiveness, E, as defined by equations
(18.35) and plotted them against R*, where
18- 34
Chapter 18
R* = R when R < 1
and
(18.36)
R* = 1/R when R > 1

(i.e. inverting the ratio when R exceeds unity).
To his delight, all the points for the spray tower then lay on a single curve within an acceptable
experimental scatter, and no matter whether the tower was used to cool water or air. The shape of
that curve is illustrated on Figure 18.18. The fact that all test results lay on a single curve held
promise that an equation or characteristic number for that curve could be used as a means of
quantifying the overall merit of the design and construction of the cooling tower or spray chamber,
and independent of the inlet temperatures or flowrates of the air and water. Whillier attempted several
curve fitting exercises on his results and found that a reasonable correlation was obtained by the
exponential relationship
E
F R*
(18.37)
where F was a constant for that particular spray tower. Different values of F in the range 0 to 1 would
be obtained for other cooling towers or spray chambers. Whillier termed this parameter the Factor of
Merit for the cooling tower. A means had been found by which the performance characteristic of a
direct heat exchanger could be quantified in a single number. A factor of merit of 1.0 would indicate a
perfect counterflow direct-contact heat exchanger. Practical values lie within the range 0.3 to 0.8.
Effectiveness
0.9
0.8
0.8
0.7
0.7
0.6
0.6
0.5
0.5
E = F R*
0.4
0.4
0.3
F = 0.3
0.2
0
0.2
0.4
0.6
0.8
Ratio of Thermal Capacities R*

Figure 18.18 Variation of E with respect to R* over a range of
Factors of Merit, F, for a direct heat exchanger.
Further work by Bluhm (1980, 1984) investigated the factor of merit for multi-stage spray chambers,
and produced an improved algorithm for spray coolers:
18- 35
Chapter 18
where N =

1 exp(N (1 R ))
1 R exp ( N (1 R ))
(18.38)
(18.39)
R 0.4 (1 F )
It is recommended that this algorithm is used for factors of merit below 0.4 or above 0.7.
Typical factors of merit suggested by Whillier (1977) and Bluhm (1980) are shown on Table 18.2.
Once the factor of merit has been established for any direct-contact heat exchanger,
either by testing or by estimation from Table 18.2, then the performance of the unit can be
established for any given flow rates and inlet conditions of the air and water streams.
Type of unit
Typical Values of Factor

of Merit, F
Vertical spray filled towers

No packing:
high water loading
low water loading
With packing:
high water loading
low water loading
Industrial packed tower
0.5 to 0.6
0.6 to 0.7
0.55 to 0.65
0.65 to 0.75
0.7 to 0.8
Horizontal spray chambers

single-stage
two-stage
three-stage
Packed air coolers
crossflow
counterflow
Mesh coolers
0.4 to 0.5
0.58 to 0.68
0.65 to 0.75
0.55 to 0.65
0.68 to 0.78
0.55 to 0.65
Table 18.2. Typical factors of merit for direct contact heat exchangers.
Example 1
A cooling tower operates at a barometric pressure of 95 kPa. The following temperature
measurements were obtained:
air inlet wet bulb temperature,
air outlet wet bulb temperature,
water inlet,
water outlet,
ta,in
ta,out
tw,in
tw,out
=
=
=
=
18.0 C
28.0 C
32.0 C
22.5 C
Determine the factor of merit for the cooling tower.
18- 36
Chapter 18
Solution
Using the psychrometric equations given in section 14.6 (or from a 95 kPa psychrometric chart),
sigma heats may be determined to be:
= 52.05 kJ/kg
at air inlet (18.0 C)
Sin
at air outlet (28.0 C),
Sout = 90.90 kJ/kg
at water inlet temperature (32.0 C), Sw, in = 111.64 kJ/kg
Air efficiency (from equation (18.31))
(S out S in )
(90.90 52.05) = 0.652
a =
=
(Sw ,in S in )
(111.64 52.05)
Water efficiency (from equation (18.30))
(t w ,in t w ,out ) (32.0 22.5)
w =
=
= 0.679
(t w ,in t a,in )
(32.0 18.0)
Tower capacity factor (from equation (18.34))
a
0.652
R =
=
= 0.961
w
0.679
Tower effectiveness(from equations (18.35))
As R < 1 then E = w = 0.679.
Also, from equations (18.36)
R* = R = 0.961
Factor of merit (from equation (18.37))
E = F R*
ln(E ) = R * ln(F )
or
ln(E )
ln(0.679 )
F = exp
= exp 0.961 = 0.668
*
R
Note that we have been able to calculate the factor of merit using measured temperatures and the
barometric pressure only. We have not required any fluid flowrates.
Bluhm's equations (18.38 and 18.39) give the factor of merit to be 0.659, a difference of some 1.3
percent in this example.
Example 2
Suppose the cooling tower of the previous example were to be converted into a vertical spray cooler
by supplying it with chilled water. Calculate the operating characteristics of the cooler given that:
water inlet temperature,
tw,in
air inlet wet bulb temperature, ta,in
water flowrate,
mw
air flowrate
or ma
P
=
=
=
=
=
=
5C
18C
100 litres/s or kg/s
120 m3/s (at an air density of 1.1 kg/m3),
120 x 1.1 = 132 kg/s
95 kPa
18- 37
Chapter 18
Solution
The wet bulb temperature of the inlet air is the same as in the previous example. Hence, the
corresponding sigma heat remains at
Sin =
52.05 kJ/kg
However, at the new inlet water temperature of 5C, the psychrometric equations of Section 14.6 give
Sw,in = 19.38 kJ/kg.
Tower capacity factor (from equation (18.32))
(t w ,in t a,in ) 100
mw
(5 18 )
R =
Cw
=
4187
(19.38 52.05 )1000
(Sw ,in S in ) 132
ma
Tower effectiveness
As R > 1 then R* = 1/R
= 1.262
(equation (18.36))
1
= 0.7923
1.262
R* =
Now the factor of merit for the tower, F, was calculated in the previous example as 0.668. This
remains the same despite the new inlet conditions. Equation (18.37) gives
E = F R*
= 0.668 0.7923
= 0.7264
Fluid efficiencies
As R > 1, E = a (equation (18.35))
Hence a = 0.7264
and from equation (18.34)
0.7264
= 0.5755
1.262
Now, equation (18.30) gives
(t w ,in t w ,out )
(t w ,in t a,in )
or
t w ,out
= t w ,in w (t w ,in t a,in ) = 5 0.5755 (5 18 ) = 12.48 C
and equation (18.31) gives
(Sout
a =
(Sw ,in
S in )
S in )
from which
S out
= a (Sw ,in S in ) + S in
= 0.7264 (19 380 52 050 ) + 52 050 = 28 319 J/kg or 28.3 kJ/kg
18- 38
Chapter 18
The air wet bulb temperature giving this sigma heat may be read from the 95 kPa psychrometric chart
or calculated by iterating equations (14.44 to 14.47), giving ta,out = 9.18C.
Rate of heat transfer, q
Using the water circuit,
q = mw C w (t w ,out t w ,in ) = 100 4187 (12.48 5 )
= 3.132 10 6 or 3.132 MW
As a cross-check, we may use the air circuit to give

q = m a (S out S in ) = 132 (28 319 52 050 )
= -3.132 x 106 W or -3.132 MW
The heat gained by the water is shown to balance the heat lost by the air.1
Example 3
A mesh spray cooler is to be supplied with water at a temperature of 10C. The unit is required to cool
12 kg/s of air from a wet bulb temperature of 29C to 23C and at a barometric pressure of 100 kPa.
Determine the rate of water flow that must be sprayed through the cooler, the temperature of the
return water and the cooling duty of the unit.
Solution
In the absence of further data concerning the device, we take a mean value of 0.6 as the factor of
merit for a mesh cooler (from Table 18.2).
The data may then be summarized as follows:
ta,in = 29C
tw,in = 10C
F
= 0.6
ta,out = 23C
tw,out = unknown,
ma = 12 kg/s
mw = unknown
The sigma heats can be calculated from the psychrometric equations given in section 14.6 or
approximated from the 100 kPa psychrometric chart:
Sin (at 29C wet bulb temperature) =
Sout (at 23C)
=
Sw,in (at 10C)
=
92.28 kJ/kg
67.11 kJ/kg
29.21 kJ/kg
Air efficiency (from equation (18.31))

(S out S in ) (67.11 92.28)
a =
=
= 0.3991
(29.21 92.28)
(Sw ,in S in )
Water efficiency (from equation (18.30)
(t w ,in t w ,out ) (10 t w ,out ) (t w ,out 10)
=
=
w =
(10 29 )
(t w ,in t a,in )
19
(18.40)
Note that because of measurement uncertainties, good agreement may not be obtained in actual operating
conditions.
18- 39
Chapter 18
Capacity factor (from equation (18.34))
0.3991 19
7.583
R = a =
=
(t w ,out 10) (t w ,out 10)
w
(18.41)
We now have a decision to make - whether to assume that R is greater or less than unity, allowing us
to ascribe values for E and R*. Let us assume that R > 1 (we shall see in a moment that this was the
wrong decision). Using equations (18.35 and 18.36), if R > 1, then E = a and R* = 1/R.
E = F R*
or
0.3991 = 0.6
giving
(t w ,out
(t w ,out 10 )
7.583
10 ) = 7.583
ln(0.3991)
= 13.64
ln(0.6 )
or
tw,out = 23.64C.
In order to check the consistency of the value of R, we can return to equation (18.41):
7.583
= 0.56
(23.64 10)
showing that our assumption of R > 1 was incorrect.
R =
Let us, therefore, repeat the analysis assuming that R < 1. Then equations (18.35 and 18.36) give
E = w and R* = R.
E = F R*
or, using w = (tw,out - 10)/19 from equation (18.40) and R = 7.583/(tw,out - 10) from equation (18.41)
(t w ,out
10 )
19
= 0 .6
{7.583 / (tw ,out
10 )}
To simplify the algebra, we may substitute

(tw,out - 10) = a
giving
a = 19 x 0.6(7.583/a)
This can easily be solved iteratively on a pocket calculator to give
a
= 14.56
tw,out
and
24.56C
18- 40
Chapter 18
Using equation (18.41), again, to check the value of R gives

R = 7.583/14.56 = 0.521
In this case, our assumption that R < 1 is shown to be consistent.
Water flowrate (from equation (18.32))
mw = R
m a (Sw ,in S in )
C w (t w ,in t a,in )
= 0.521
12 (29 210 92 280 )

= 4.955 kg/s
4187 (10 29 )
Cooling duty
This can be calculated directly from the air data:
q = m a (S out S in ) = 12 (67.11 92.28 ) = 302 kW
To check the heat balance, the heat gain by the water is given as
q = mw C w (t w ,out t w ,in ) = 4.955 4.187 (24.56 10 ) = 302 kW
18.3.2.5. Heat exchange across the walls of pipes and ducts

When a fluid flowing through a pipe or duct is at a temperature different to that of the ambient air,
then heat exchange will occur across the walls of the pipe and any insulating material that surrounds
it. The analysis of indirect heat exchangers given in Section 18.3.2.2 applies equally well for pipes
and ducts.
Consider the insulated pipe shown on Figure 18.19 carrying a fluid of temperature tin when the
temperature of the outside air is tout.
tout
kp
kins
r1
tin
r2
hin
hout
r3
Figure 18.19 Heat transfer across an insulated pipe or duct.
18- 41
Chapter 18
The rate of heat transfer from a short length of pipe over which neither tin nor tout change significantly
is given as:
q = UA (t in t out )
W (see equation
(18.18)
The UA value is given by

r r2
r r1
1
1
=
+ 2
+ 3
+
UA
hin Ain
k p A12
k ins A23
where hin
hout
hr
r
kp
kins
A
0.95
(hout
1
+ 0.95 hr ) Aout
C/W
(18.42)
= heat transfer coefficient at the inner surface ( W/(m2 C) )

= heat transfer coefficient at the outer surface ( W/(m2 C) )
= radiative heat transfer coefficient at the outer surface ( W/(m2 C) )
= radii (m)
= thermal conductivity of pipe ( W/(m C) )
= thermal conductivity of insulation ( W/(m C) )
= area (or mean area) across which heat transfer occurs (m)
= assumed value for the product of emissivity and view factor, (Whillier, 1982)
Values of thermal conductivity for pipe materials can be obtained from Table 18.3.
The heat transfer coefficients may be estimated from the following relationships which take into
account the variations of fluid viscosity and thermal conductivity with respect to temperature.
Air inside a duct
The Colburn equation (18.22) gives the heat transfer coefficient for turbulent flow inside smooth
tubes. For air, the Prandtl Number, Pr, remains essentially constant at 0.7 over a wide range of
temperatures and pressures (Appendix A15.3). Further analysis of the Colburn equation gives the
approximation
hin (air ) = 3.169
( a u a )0.8
d
0. 2
(1 + 9.4 10
t in 6.49 10 6 t in
W/(m2 C)
(18.43)
where a = air density (kg/m3)

ua = air velocity in duct (m/s)
d = internal diameter of duct (m)
The effect of temperature is fairly small over the range of 0 to 60C. Hence, taking a value of 20C
gives the simpler approximation
hin (air ) = 3.22
( a u a )0.8
d
0 .2
W/(m2 C)
(18.44)
The incorporation of air density allows this relationship to be employed for compressed air pipes as
well as ventilation ducts.
Water inside a pipe
A similar analysis for water inside a pipe leads to the approximation
hin (water ) = (1430 + 20.9 t in )
where
uw =
uw
d
0. 8
0.2
W/(m2 C)
(18.45)
water velocity (m/s)

18- 42
Chapter 18
In this case, the near constant value of water density allows it to be eliminated as a variable.
However, the viscosity of water varies significantly in the range 0 to 60C. Hence, the effect of
temperature should not be ignored in equation (18.45).
Air outside a pipe
This time, the relationship given by McAdams for the outside of a single tube (equation (18.23))
provides the basis for further analysis. This leads to an approximation that may be
used for both longitudinal flow and cross flow in subsurface airways.
hout (air ) = 4.24
( a u a )0.6
d
0 .4
(1 + 0.0015 t out )
W/(m2 C)
(18.46)
Radiative heat transfer coefficient

This can be calculated from the Stefan-Boltzman equation (15.27) using the dry bulb temperature of
the surrounding air, giving
hr = 22.68 10 8 (273.15 + t out )
W/(m2 C)
(18.47)
However, an adequate accuracy for most purposes is obtained simply by assuming a value of
hr = 6 W/(m2 C).
Over long lengths of pipe or ducting the inner and outer temperatures, tin and tout, are each likely to
change in a nonlinear manner. In this case the log. mean temperature difference should be employed
giving
q = UA
where
and
(t 1 t 2 )
ln (t 1 / t 2 )
(18.48)
t1 and t2 are the values of the temperature difference (tin - tout) at the two ends of the pipe
In = natural logarithm
Example
A chilled water pipeline is 1000 m long and comprises a 200 mm internal diameter steel pipe of
thickness 4 mm, inside a 30 mm sheath of phenol formaldehyde insulation. The water has a velocity
of 2 m/s and a mean temperature of 8C while the surrounding air has a velocity of 4 m/s, a mean dry
bulb temperature of 25.5C and a density of 1.12 kg/m3.
Determine
(a) the UA value per metre length of pipeline,
(b) the actual heat gain and temperature rise of the water
and (c) the temperature of the outer surface of the insulation.
Solution
(a) Let us first collate the data required by equation (18.42)
Thermal conductivities (from Table 18.3)
steel:
kp = 45 W/(m C)
phenol formaldehyde: kins = 0.032 W/(m C
Radii (with reference to Figure 18.19)
r1 = 0.1m;
r2 = 0.104m;
r3 = 0.134 m
18- 43
Chapter 18
Areas of heat transfer (per metre lenqth)

Ain
= 2 r1 = 2 x 0.1
= 0.6283 m2
Aout
= 2 r3
= 2 x 0.134
= 0.8419 m2
A12 = 2
A23 = 2
(r1
+ r2 )
(0.1 + 0.104 ) = 0.6409 m2
= 2
2
2
(r 2
+ r3 )
(0.104 + 0.134 ) = 0.7477 m2
= 2
2
2
Heat transfer coefficients

For water in the pipe (from equation (18.45)):
hin (water ) = (1430 + 20.9 t in )
=(1430 + 20.9 8 )
0. 8
uw
W/(m2 C)
0 .2
d1
2 0. 8
(0.2)
0 .2
= 3837 W/(m2 C)
For air outside the pipe (from equation (18.46)):

hout (air ) = 4.24
= 4.24
( a u a )0.6 (
1 + 0.0015 t out )
0 .4
d3
(1.12 4)0.6 (1 + 0.0015 25.5)

0 .4
0.268
= 18.33
W
m2 oC
For radiative heat transfer at outside of pipe (from equation (18.47))

hr = 22. 68 x 10- 8 (273.15 + 25. 5)3 = 6.04
W/(m2 C)
The large value of hin (water) shows the relative ease with which heat is transferred from the
water to the steel pipe.
The UA value per metre length can now be calculated from equation (18.42). Inserting the
numerical value gives:
1
1
0.004
0.03
1
=
+
+
+
(18.33 + 0.95 6.04) 0.8419
UA
3837 0.6283
45 0.6409
0.032 0.7477
water inside pipe
steel
insulation
air outside insulation
or
0.00041
0.00014
1.3037
UA
= 1/1.3037 = 0.7670
1.2538
0.0493
C/W
W/C
The dominant effect of the insulation is shown here. Indeed, the terms involving the water inside
the pipe and the steel pipe itself may be neglected without significant loss of accuracy.
(b) The average heat loss per metre length of pipe is then simply
q = UA (tin - tout) = 0.7670 (8 - 25.2) = -13.42 W per metre

(negative as heat is transferred from the air to the water).
18- 44
Chapter 18
Over the 1000 m pipeline, the total heat transferred to the water is then
qtot = 13.42 x 1000 = 13 420
In order to determine the actual increase in temperature of the water, tw, it is first necessary to
calculate the mass flow of water
mw = w u w area of flow = 1000 2
0.2 2
4
= 62.83 kg/s
where w = density of water (1000 kg/m3)

q tot
C w mw
where Cw = specific heat of water (4187 J/(kgC) giving
Then t w =
t w =
13 420
= 0.051 C
4187 62.83
This small rise in temperature justifies the assumption of a "constant" water temperature at a
mean value. In practice, the actual temperature rise may be significantly greater because of
damaged or imperfectly applied insulation, particularly at flange points.
(c) The heat flux remains the same at each surface boundary of the pipe and insulation. This allows
the temperatures at those boundaries to be determined. In particular, at the outer surface of the
insulation,
q = hout Aout (tsurf - tout)
where tsurf = surface temperature of the insulation.
Inserting the known values gives

t surf =
13.42
+ 25.5 = 0.87 + 25.5 = 24.63C
18.33 0.8419
No condensation will occur on the outside of the insulation provided that the wet bulb temperature
of the air remains below 24.63C. However, any imperfections in the insulation may allow the
ingress of water vapour and condensation on the steel tube. This can result in problems of
corrosion.
The techniques described in this section can be incorporated into mine climate simulation programs
(Chapter 16) in order to take into account heat transfers between pipes or ducts and the surrounding
airflow. If the fluid inside the pipe moves in the same direction as the ventilating airstream, then the
heat transfers and corresponding temperature changes can be computed sequentially along each
incremental length of simulated airway (Section 16.2.2.). In the case of the piped fluid and the
ventilating airstream moving in opposite directions, then an iterative procedure becomes necessary to
compute the equilibrium condition.
18.3.3. Water distribution systems
A significant item in the design and economic analysis of mine air conditioning systems is the choice
of piping that carries water between the refrigeration plant and heat exchangers or mining equipment.
18- 45
Chapter 18
There are, essentially, two matters of importance. First, the size of the pipe should be selected
carefully. If it is too small for the given flowrate, then the frictional pressure drops and pumping costs
will be high. On the other hand, if too large a pipe is chosen the costs of purchase and insulation will
be high while the larger surface area and low velocity will exacerbate unwanted heat transfer with the
surroundings.
This brings us to the second consideration, the thermal properties of the pipe material and the need
for additional insulation. We shall discuss each of these matters in turn.
18.3.3.1. Pipe sizing
The first step in determining the diameters of the various pipes that comprise a water network is to
assess the flowrates that are to be carried in each branch. The cooling duty, q, of a heat exchanger is
given as
q = mw Cw (tw,out - tw,in)
where
and
mw
Cw
tw,in
tw,out
=
=
=
=
(18.49)
water flowrate through the heat exchanger (litres/s or kg/s)

specific heat of liquid water (4187 J/kg C)
inlet water temperature (C)
outlet water temperature (C)
The heat transferred is dependent equally upon the water flowrate and the temperature rise of the
water through the unit. The latter, in turn, is governed by the construction of the heat exchanger and
the supply temperature of the water. Calculations of the type included in Section 18.3.2 allow
flowrates to be assessed. For service water lines, the flowrates must also satisfy the machine
requirements for dust suppression or hydraulic power.
Having established the water flowrate, the actual sizing of a pipe becomes a combination of simple
analysis and practical considerations. Taking into account capital, operating and installation costs,
van Vuuren (1975) suggested optimum water velocities of 1.8 to 2.0 m/s for 4000 kPa pipes and 2.0
to 2.2 m/s for 7000 kPa pipes. A common rule of thumb is to employ 2 m/s for an initial exercise in
pipe sizing. The size of pipe is then given as
d2
4
= A =
Q
u
m2
or
d =
where
and
4Q
Q
= 1.13
u
u
d
A
Q
u
=
=
=
=
pipe diameter (m)

pipe cross sectional area (m2)
flowrate (m3/s)
velocity (m/s)
The frictional pressure drop, p, along the pipe may then be determined from the Chezy Darcy
relationship (equation (2.45)):
p =
4 f L w u 2
2d
Pa
(18.51)
18- 46
Chapter 18
For practical calculations, water density, w, may be taken as constant at 1000 kg/m3. The length L is
often set at unity to give the frictional pressure loss in Pa per metre.
New steel pipes will have a coefficient of friction, f, approximating 0.0039. However, the aging
processes of corrosion and scaling may cause this to increase by 20 to 50 percent. Plastic pipes have
a coefficient of friction averaging about 0.0032. Manufacturers usually provide charts for their pipes
relating frictional pressure losses to pipe diameter and flowrate. The pressure loss that can be
tolerated depends upon the pumping cost function which is proportional to the product pQ (ref.
equation (9.17)), and the pressure required at the end of the pipeline. If the pumping costs are shown
to be high, then consideration should be given to increasing the pipe diameter and reducing the
velocity.
Pressure losses are also incurred at bends or pipe fittings. Again, manufacturer-specific data can be
employed to determine such losses. However, for mine water systems, it is common to cater for
bends and fittings by simply adding ten percent to the computed pressure drop.
Pipe diameters determined through this procedure should be rounded up to the next highest standard
size. Consideration should be given to any increases in water flow that may be required in the future.
Furthermore, pipe ranges that might be used for fire fighting should be sized to give adequate flow
and pressure at the hydrants. The sizes of pipes that are to be installed within existing airways or
shafts may also be constrained by the space available.
18.3.3.2. Pipe insulation
The performance of heat exchangers is dependent upon the temperature of the supply water. It is,
therefore, important that heat gain along chilled water pipelines is minimized. This is normally
accomplished by providing thermal insulation on those pipes. One exception is the special case
where the pipeline is located within an intake airway which, itself, requires air cooling. In such
circumstances, the insulation may be partially, or totally, omitted.
In addition to having a low thermal conductivity, the insulating materials for chilled water pipes in
mines should conform to several other criteria:
The insulation should provide a barrier against the ingress of water that condenses on the outer
surfaces. A porous and permeable insulation that becomes waterlogged will be much more
conductive to heat. Furthermore, water that reaches the surface of metal pipes will cause
problems of corrosion. The required vapour barrier may be achieved either by covering the
insulation with an impermeable sheath or by using an insulating material that is, itself, nonpermeable. An example is cellular glass in which the small air pockets are isolated from each
other.
The insulation or outside sheath should have sufficient mechanical strength to withstand the
rigours of transport and utilization in mine environments. Insulation on shaft pipelines must be
able to withstand impact from falling objects.
The flammability of insulating materials and the toxicity of gases that they may produce when
heated in a mine fire is of great importance. Polyurethane foams have been employed widely in
the past. However, they emit carbon monoxide and, perhaps, cyanide vapours at temperatures
that may be reached in mine fires. Phenolic foams in which phenols are bound into a resin base
are safer in that they have a higher ignition temperature, and a much lower rate of flame spread
(Rose, 1989).
The cost of purchase and ease of installation of insulating materials or pre-insulated piping
should also be taken into account.
18- 47
Chapter 18
A common practice has been to apply insulating materials to pipes on the mine site - either on the
surface or after the pipework has been installed in its underground location. The application of foam
insulators may be carried out by injecting the foam within a PVC vapour barrier sheath that has been
pulled over spacers on the outside of the pipe. Pre-cast semicircular sections of cellular glass or other
rigid insulators may be bound on to the pipes with adhesives or an outer sealing material. Expanded
polystyrene granules and fibreglass have also been used widely (Arkle, 1985; Ramsden, 1985). The
thickness of applied insulators varies from 25 to 40 mm.
The continued development of thermoplastics has allowed the manufacture of pipes that combine
mechanical strength with good thermal insulating properties. These may consist of concentric and
impermeable cylinders with the annulus containing a cellular infill. Typical values of thermal
conductivity are given in Table 18.3.
Material
Thermal conductivity
W/(m C)
Still air
0.028
Expanded polystyrene
0.038
Polyurethane and polystyrene
0.034
Polyisocyanurate
0.023
Phenol formaldehyde
0.032
Glass fibre
0.029
Unplasticised polyvinylchloride (UPVC)
0.14
Still water
0.62
Wood
0.17
Steel
45
Table 18.3. Thermal conductivities of pipe materials.
18.3.3.3. Energy and temperature changes within water systems
In Section 3.4.1 we saw that fluid frictional effects have no influence on temperature changes within a
steady-state airflow. This is not the case for water or other near incompressible fluids. Furthermore, in
Section 15.3.1, the autocompression of air in a mine shaft was shown to produce an increase in air
temperature that depended upon the depth of the shaft. Again, this is not true for water in a shaft
pipe.
In order to reassess the situation for water, let us return to the steady flow energy equation (3.25).
2
u1 u 2
2
+ (Z1 Z 2 )g + W12 =
V dP + F
12
= (H 2 H 1 ) q12
J
kg
(18.52)
where
and
u
Z
g
W
V
F
H
q
=
=
=
=
=
=
=
=
velocity (m/s)
gravitional acceleration (m/s2)
work input (J/kg)
specific volume of fluid (m3/kg) = 1/ where = fluid density (kg/m3)
work done against friction (J/kg)
enthalpy(J/kg).
heat added from surroundings (J/kg).
18- 48
Chapter 18
In applying this equation to water, we recall the basic definition of enthalpy from equation (3.19):
H = PV + U
where
J/kg
(18.53)
U = internal energy and by assuming the water to be incompressible, V = constant.
The right hand sides of equation (18.52) then give

V (P2 P1 ) + F12 = P2V P1V + U 2 U 1 q12
or
F12
= U 2 U 1 q12
(J/kg)
(18.54)
However, we may take a single value for the specific heat of water, Cw = 4187 J/(kg C).
(U 2
U 1 ) = Cw (T 2 T1 )
J/kg
(18.55)
J/kg
(18.56)
Substituting in equation (18.54),

F12
= Cw (T 2 T1 ) q12
Furthermore, the frictional pressure drop, p, is given by equation (2.46)

p = w F12 = w Cw (T 2 T1 ) w q12
where
Pa
(18.57)
= density of water (kg/m3)
In the particular case of adiabatic flow (q12 = 0), the increase in water temperature along a pipeline
becomes
(T2
T1 ) =
w Cw
(18.58)
The term 1/wCw is sometimes known as the Joule-Thompson coefficient and, for water, has a
standard value of 1/(1000 x 4187) = 2.39 x 10-7 C/Pa.
The adiabatic rise in temperature along a water pipe is shown by equation (18.58) to be directly
proportional to the frictional pressure drop and is independent of changes in elevation. The frictional
pressure drop, p, may occur by viscous or turbulent action within the length of the pipe or by shock
loss across a valve or other obstruction.
For flow through a water pump or turbine, we can apply the left and right sides of equation (18.52).
However, the change in kinetic energy and the difference in elevation across the device are both
negligible, giving
W12 = (H 2 H 1 ) q12
= V (P2 P1 ) + (U 2 U 1 ) q12
=
(P2
P1 )
+ Cw (T 2 T1 ) q12
J/kg
(18.59)
If there is negligible heat exchange across the casing of the pump or turbine, then we may assume
adiabatic flow (q12 = 0) giving the temperature rise as
(T2
T1 ) =
(P2 P1 )
1
W12
w
Cw
(18.60)
18- 49
Chapter 18
While W12 is the total work applied to a pump impeller, the term (P2 P1)/w represents the output of
useful mechanical energy. The difference between the two indicates the degradation of some of the
input power to thermal energy and produces a measurable temperature rise. In a turbine, it is the
other way round with (P1 - P2)/w representing the work input and -W12 being the output work.
For a pump, we can also write that the actual work output is given by
W12 =
(P2
P1 )
J
kg
(18.61)
and, for a turbine, the actual work output is

W12 =
(P2
P1 )
J
kg
(18.62)
where = efficiency of the device

Substituting for W12 in equation (18.60) gives the temperature rise
across a pump:
P1 ) 1
1
Cw w
and across a water turbine
(T2
T1 ) =
(T2
T1 ) =
(P2
(P2
P1 )
Cw w
1) =
(P1
P2 )
Cw w
(18.63)
(1 )
(18.64)
In these analyses, it has been assumed that water is incompressible. In fact, the slight compressibility
of water does give it a small isentropic temperature change with respect to pressure (Whillier, 1967).
This is given approximately as
t
= (0.759 t 0.2) 10 6
P
where
C per kPa
(18.65)
= initial temperature (C)
This effect may be neglected for most practical purposes. However, if equations (18.63) or (18.64)
are transposed in order to determine the efficiency of the device from temperature observations, then
the measured temperature rise should be corrected according to equation (18.65) (negatively for a
pump and positively for a turbine).
Example 1
The barometric pressures at the top and bottom of a 1500 m deep shaft are 90 and 107 kPa
respectively. Chilled water at 3C and ambient atmospheric pressure enters the top of a 150 mm
diameter steel pipe that is suspended in the shaft. A pressure reducing valve at the bottom of the pipe
controls the water flowrate to 35 litres/s and reduces the water pressure to that of the ambient
surroundings. Determine the pressures and temperatures of the water at the inlet and outlet of the
pressure reducing valve. Assume that no heat exchange occurs across the pipe wall.
18- 50
Chapter 18
Solution
The frictional pressure drop in the pipe is given by equation (18.51)
p =
4 f L w u 2
2d
Pa
The known values are

L = 1500 m
w = 1000 kg/m3
d = 0.15 m
where Q = flowrate = 35/1000 = 0.035 m3/s and A = c-s area = x (0.15)2/4 =0.01767 m2
giving u = Q/A u = 1.981 m/s
Allowing for aging effects, we shall assume the coefficient of friction for the steel pipe to be f = 0.005.
We shall also allow an additional pressure loss of 10 percent to allow for bends and fittings.
Then
p =
4 0.005 1500 1000 (1.981)

2 0.15
(1 + 0.1)
= 431.68 10 3
Pa
Water pressure at shaft top = 90 x 103 Pa

Water pressure at shaft bottom (inlet to pressure reducing valve)
= 90 + w g L - p
= 90 x 103 + (1000 x 9.81 x 1500) - 431.68 x 103
= 14 373 x 103 Pa
or
14 373 kPa
At outlet from the valve, the water pressure is reduced to the shaft bottom barometric pressure of 107 kPa.
Hence,
pressure drop across the valve = gauge pressure at valve inlet
= 14 373 - 107 = 14 266 kPa
The temperature rise in the shaft pipe is given by equation (18.58) as
t sh =
431.68 10 3
p
=
= 0.103
1000 4187
w Cw
Temperature at valve inlet
= water temperature at shaft top + 0.103 = 3.103 C
As no energy is added or extracted by the pressure reducing valve, the flow is purely frictional.
Hence, equation (18.58) gives
Pressure loss across valve
14 266 10 3
= 3.407C
t valve =
=
w Cw
1000 4187
giving the water temperature at the valve outlet to be
3.103 + 3.407 = 6.510
C.
18- 51
Chapter 18
The same total increase in water temperature of 3.51C from shaft top to outlet of the valve at the
shaft bottom would occur whatever the cause of the restriction on flow - a valve, rougher lining of the
pipe or a smaller diameter of pipe, provided that no energy of any kind is added or extracted from the
water and that the water is emitted at the shaft bottom barometric pressure.
As a check, the overall temperature rise of the water between the shaft top and the outlet of the shaft
bottom valve can be calculated directly. From the steady flow energy equation (18.52) with no work
input (W12 = 0), for an incompressible fluid V = 1/ = constant, and ignoring changes in kinetic energy
(Z1
or
Z 2 )g = V (P2 P1 ) + F12
p tot = w F12 = (Z1 Z 2 )g w (P2 P1 )
= 1500 x 9.81 x 1000 - (107 - 90) 1000

= 14.715 x 106 - 0.017 x 106
= 14.698 x 106
Pa
(The difference in barometric pressure is shown to have very little effect on ptot.)
Then
t tot =
=
p tot
w Cw
(equation (18.58))
14.698 10 6
1000 4187
= 3.51 C
This check calculation illustrates that when the water is emitted from the pipe at the underground
barometric pressure, the temperature rise is given approximately as
(Z1
Z 2 )g
= 3.51 o C over the 1500m depth of the shaft, or 2.34 C per 1000 m depth.
4187
In the past, this has mistakenly been referred to as an "autocompression temperature rise." The
misconception arose because in the special case of a pipe open at the bottom, the frictional pressure
drop in the pipe approximates the static pressure caused by a column of fluid in the shaft pipe.
Example 2
If, in Example 1, the pressure reducing valve at the shaft bottom were replaced by a water turbine of
efficiency 75 percent, what would be the temperature of the water leaving the turbine? Calculate,
also, the power produced by the turbine. Assume that all other parameters, including the flowrate of
35 litres/s, remain unchanged.
Solution
As the condition and flow of water in the pipe have not been altered, the increase in water
temperature down the pipe remains at tsh = 0.103C and the water enters the turbine at a pressure
of Pin = 14 373 kPa and a temperature of 3.103C.
18- 52
Chapter 18
At outlet from the turbine, the water pressure has fallen to the shaft bottom barometric pressure of
Pout = 107 kPa. The temperature rise across the turbine is given by equation (18.64).
(Pin
t turb =
Pout )
Cw w
(1 )
(14 373
107 )10 3
(1 0.75 ) = 0.852 C
4187 1000
Then the temperature of the water leaving the turbine is

3.103 + 0.852
= 3.955 C.
Comparing this with the temperature of water leaving the pressure reducing valve of Example 1, i.e.
6.510 C, shows that replacing the valve by a water turbine not only produces useful power but
reduces the temperature of the chilled water leaving the shaft station by 6.510 - 3.955 = 2.555C.
This represents an equivalent refrigeration capacity of
= 35 x 4187 x 2.555 = 374.5 x 103 W or 374.5 kW
mw Cw t
The output work from the turbine is given by equation (18.62)

W12 =
(Pout
Pin )
(107 14 373 )10 3

1000
0.75 = 10 700
J
kg
(negative, as work is leaving the water). To transform this to an output power, we simply multiply by
the mass flowrate, i.e.
10 700 x 35 = 374.5 X 103 W
or 374.5 kW
We have, therefore, demonstrated that the power output from the turbine is equal to the equivalent
refrigeration capacity that has been produced by the system. This suggests a rapid means of
approximating the overall temperature rise of the water from the shaft top to the turbine outlet at shaft
bottom barometric pressure:
t tot =
=
(Z1
Z 2 )g
turbine output power
Cw
mw C w
1500 9.81
374 500
= 3.514 - 2.555
4187
35 4187
= 0.959 C
This method remains acceptable provided that the frictional pressure losses are small compared with
the static pressure caused by the column of water.
18.3.4. Energy recovery devices
The trend towards siting main refrigeration plant on the surface of mines accelerated during the
1980's. This was caused in large part by the utilization of energy recovery devices, particularly water
turbines. Prior to that time, the provision of chilled water at the surface had three major drawbacks.
First, it caused high water pressures at shaft bottoms which necessitated a choice between high
pressure heat exchangers and associated pipework or, alternatively, to pass the water through
pressure reducing valves. The latter choice not only resulted in high operating costs of pumping to
return the water to surface, but also converted the potential energy of the water to thermal energy,
giving a rise in temperature and eroding the cooling capacity of the chilled water.
18- 53
Chapter 18
Energy recovery devices address each of these problems and divide into two major groups. First,
water turbines or hydraulic motors employ the potential energy of shaft water lines to drive electrical
generators or other devices. Secondly, hydrolift systems utilize the weight of descending chilled water
to help raise heated return water back to surface. We shall discuss each of these systems in turn.
18.3.4.1. Water turbines
As illustrated by Example 2 in section 18.3.3.3, shaft bottom turbines provide:
work output that may be used to drive pumps, generators or any other mechanical device; this
work may provide part of the energy required to raise return water back to surface;
reduced temperature of the subsurface supply of chilled water, and hence, savings in
refrigeration costs;
a low pressure water system underground.
Two types of turbines are in use. Figure 18.20 illustrates an impulse turbine, commonly known as a
Pelton Wheel. In this device, the pressure of the supply water is used to produce kinetic energy by
means of one or more nozzles. The high velocity jets are directed at cups attached to the periphery of
a rotating wheel. Inside the casing of the device, the jets operate at the ambient atmospheric
pressure. Hence, the water leaves at zero gauge pressure. Control of the mechanical output may be
achieved by a needle valve in the nozzle which changes the water flowrate. Alternatively, if the flow is
to be maintained, deflector plates may be used to direct part of the jets away from the cups.
The advantages of a Pelton Wheel are that it is simple and rugged in construction, reliable and may
reach efficiencies of over 80 percent. The speed control is effective and no high pressure seals are
required on the casing.
inlet
nozzle
needle valve
runner
cups
outlet
Figure 18.20 Diagrammatic representation of a Pelton Wheel.
18- 54
Chapter 18
The second type of water turbine is the reaction device and is often referred to as a Francis or Kaplan
turbine. This is essentially a centrifugal runner (Figure 10.2) with the impeller designed to operate in
reverse and incorporating guide vanes to minimize shock losses. Indeed, a reversed centrifugal pump
may be used as a reaction turbine although at a considerably reduced efficiency. Control of the
device is by valves located upstream and/or downstream. Unlike the Pelton Wheel, a reaction turbine
operates in a "flooded" mode giving rise to its primary advantage, i.e. the water pressure at exit can
be maintained above the ambient air pressure and, hence, may be utilized further downstream to
overcome pipe losses, provide the required pressure at spray chambers or to activate further
hydraulic devices. Pelton Wheels give better efficiencies at the higher duties, typically 2 MW. The
Francis turbines are better suited for lower duties2.
In addition to shaft bottom stations, water turbines may be used for more local applications. For
example, if the water pressure available at a spray chamber exceeds that required at the spray
nozzles, then a turbine can be sited in the water supply line and used to power local pumps or fans.
This is particularly advantageous for multi-stage spray chambers (Ramsden, 1985).
18.3.4.2. Hydrolift systems
The principle of operation of a hydrolift system is illustrated on Figure 18.21. The device is sometimes
also known as a hydrochamber or hydrotransformer. The chamber is excavated or constructed
underground and is alternately filled from opposite ends with chilled water delivered from surface,
then with hot water returned from the mine.
Let us assume that the chamber is initially filled with warm water. Both of the automated valves A are
opened and valves B are closed. The pump at the surface cold water dam causes chilled water to
flow down the shaft pipe and into the chamber. This, in turn, forces the hot water to exit the chamber
and ascend the return shaft pipe. As the two columns of water are nearly balanced, the duty of the
pump is simply to overcome pipe friction. When the chamber is filled with chilled water, valves A are
closed and valves B are opened. The underground hot water dam pump then forces the hot return
water into the chamber, expelling the chilled water to the underground cold water dam.
The cyclic nature of the system can be negated by having three hydrolift chambers with the sets of
valves phased to give a continuous flow into and out of the water dams. The size of chambers and
water dams should be determined on the basis of the cooling requirements of the mine. An inherent
disadvantage of the hydrolift system is that the rate of chilled water flow must be equal to that of the
heated return water. It is, therefore, advisable that the water dams should be sized to accommodate
several hours supply.
A loss of efficiency occurs because of heat exchange with the walls of the chamber and due to mixing
and thermal transfer at the hot and cold interface within the chamber. The latter may be minimized by
inclining the chamber such that the hot water end is at a higher elevation.
As a guide, if the pressure differential available across the turbine (in kPa) is less than 70 mw0.67 (where mw is
the flowrate in litres/s), then a Francis turbine will give a better performance.
18- 55
Chapter 18
refrigeration
plant
condenser
cold water
dam
evaporator
pre-cooling
tower
hydrolift chamber
cold water
dam
hot water
dam
pump
Figure 18.21 Operating principle of a hydrolift system.
18.3.5. Design of mine cooling systems

The demand on a subsurface cooling system is dictated by the magnitude and distribution of the heat
load and the surface climate. The availability of cold natural water at the surface can result in large
savings in cooling costs. Similarly, passing the water supply through a surface cooling tower either
prior to or instead of a refrigeration plant can produce low-cost cooling.
Section 18.3.1 gave an overview of mine cooling systems and introduced spot coolers for isolated
heat loads (Figure 18.8), water chiller plant for sections of a mine (Figure 18.9) and a large
centralized system (Figure 18.10). In this Section, we shall concentrate on the latter and discuss the
alternative arrangements by which the hardware components may be integrated into a unified mine
cooling system.
18.3.5.1. Location of main plant
Figures 18.22 to 18.24 give examples of system configurations using main refrigeration plant located
on surface, underground and a combination of the two. In these figures, the ancillary components of
pumps, valves and retention dams are not shown.
18- 56
Chapter 18
Surface plant
The ease of maintenance, construction and heat rejection, coupled with the utilization of energy
recovery devices have led to a preference for the main refrigeration units to be located on surface.
Figure 18.22(a) illustrates the simplest system that employs surface plant. Chilled water from the
refrigeration plant passes down an insulated pipe in the shaft to a high pressure water to water heat
exchanger at one or more shaft stations. This heat exchanger is a source of potential loss of system
efficiency and care should be taken with the design and choice of tube materials to minimize
problems of corrosion, scaling and erosion. Figure 18.22(b) shows the water to water heat exchanger
replaced by a turbine. This configuration requires the water flow in the shafts to approximate that
cycling through the mine, with capacitance provided by water dams.
In both cases, the heated water ascending to surface may be sprayed through a pre-cooling tower in
order to lower its temperature before returning to the refrigeration plant. This can be advantageous in
reducing the required capacity and operating costs of the refrigeration plant. However, the
effectiveness of a pre-cooling tower is dependent upon the difference between the temperature of the
return water and the wet bulb temperature of the surface atmosphere. An "approach" (see Figure
18.18) of 2C is typical. For this reason, a pre-cooling tower may be utilized only during the colder
seasons.
Figure 18.22(b) also shows part of the chilled water being used for bulk cooling of the air before it
enters the downcast shaft. The vertical packed spray chambers used for this purpose are normally of
the induced-draught type with the fan located between the spray chamber and the shaft. This helps to
prevent fogged air from entering the shaft. If the experience of any given mine is that a heat problem
exists during the summer months only, then all that may be necessary is to install a surface bulk air
cooling system and utilize it during the warm season. However, if the workings are distant from the
shaft bottoms and the heat production is concentrated heavily within those working zones, then bulk
air cooling on surface will have greatly reduced effectiveness.
Underground plant
Figure 18.23 illustrates the principle of the system favoured by the South African gold mines prior to
the 1980's. The refrigeration plant is located entirely underground. Hot water from the condensers is
cycled through open spray cooling towers situated at or near the base of upcast shafts and supplied
by air returning from the mine. The advantages of this system are that it eliminates the need for
surface-connecting pipe ranges and the associated pumping costs. It also avoids any environmental
problems that may arise from surface plant. The major disadvantage of the system is that its duty is
limited by the capacity of the return air to accept heat rejected in the underground cooling towers.
Furthermore, saturated air subjected to decompression in the upcast shaft can cause problems of
heavy fogging, water blanketing in the shaft (Section 9.3.6) and reduced life of main exhaust fans.
18- 57
Chapter 18
optional bulk
cooling of
downcast air
pre-cooling
tower
to
cooling
tower
pre-cooling
tower
to
cooling
tower
to
downcast
shaft
condenser
condenser
evaporator
evaporator
(b)
(a)
chilled water
return water
chilled water
turbine
water to water
heat exchanger
cold water
dam
high pressure
to
low pressure
workings
from
workings
to
workings
from
workings
hot water dam
Figure 18.22 Examples of system configurations using surface refrigeration plant.
18- 58
Chapter 18
upcast shaft
hot water from

condensor
cooling
tower
refrigeration
plant
return air
from mine
condenser
evaporator
chilled water
to mine
return water
from mine
Figure 18.23 System configuration for underground plant.
Combinations of surface and underground plant

As the depth of mining increases, so also does the severity of geothermal and autocompressive heat
problems. Unfortunately, increasing depth also exacerbates the costs of installing, maintaining and
operating pipelines in shafts and, hence, eroding the advantages of a surface plant. Figures 18.24 (a,
b and c) illustrate examples of alternative means of combining surface and underground refrigeration
plant into an integrated system. In each of these configurations, the essential role of the underground
plant is to concentrate the heat in the return mine water into a lower flow, higher temperature system
for transmission to surface.
This enables smaller shaft pipes to be selected and also reduces the pumping costs. Furthermore,
the increased temperature of the water returning to surface improves the effectiveness of the precooling tower, allowing it to be utilized throughout the year. Another advantage of the combined
system is that it adds considerably to the flexibility and future upgrading of the cooling duty. Surface
plant alone is ultimately limited by the flow capacity of the shaft pipes. Similarly, underground plant
alone is limited by the thermal capacity of the return air. However, with a combined system the
cooling duty can be more easily upgraded by adding to the capacity of both surface and underground
plant and, hence, increasing the temperature difference between the descending and ascending shaft
pipes.
18- 59
Chapter 18
pre-cooling
tower
pre-cooling
tower
to
cooling
tower
to
cooling
tower
pre-cooling
tower
to
cooling
tower
condenser
condenser
condenser
evaporator
evaporator
evaporator
(a)
refrigeration
plant
(b)
water-water
heat exchanger
(c)
refrigeration
plant
condenser
condenser
evaporator
evaporator
hot
dam
cold
dam
refrigeration
plant
condenser
evaporator
chilled water
to mine
return water
from mine
return water
from mine
chilled water
to mine
chilled water
to mine
return water
from mine
Figure 18.24 Examples of system configurations for combined surface and underground refrigeration plant.
18- 60
Chapter 18
The simplest combined system is shown on Figure 18.24(a) where the surface plant is used to cycle
water through the shaft pipes and to remove heat from the subsurface condensers. Figure 18.24(b)
shows a system of potentially improved efficiency in which the chilled water from surface first passes
through a water-to-water heat exchanger before entering the condensers. The warm return water
from the mine is pre-cooled in the water-to-water heat exchanger prior to passing through the
evaporators.
Figure 18.24(c) shows the water-to-water heat exchanger replaced by a turbine. In this case, the
chilled water emerging from the turbine passes on to the mine via a cold water dam. The return water
in the hot water dam is split. Part of it removes heat from the condenser and returns to surface. The
remainder is re-cooled through the subsurface evaporator and returned to the cold water dam.
In each of the three examples of combined plant shown on Figures 18.24 (a, b and c), some of the
chilled water produced on surface may be used for bulk cooling the downcast air. Similarly, in all
three cases, if mine return air is available for heat rejection, then some of the hot water from the
underground condenser may be cycled through a subsurface cooling tower before returning to the hot
water dam.
18.3.5.2. Service water cooling
Service water is supplied to mining equipment primarily for dust suppression purposes and perhaps,
also, as a means of transmitting hydraulic power (Section 18.3.5.3). Heat and dust problems are
generally greatest in the working areas and, particularly, where rock breaking equipment is in use.
The heat emitted by the machines combines with that from the newly exposed strata to cause rapid
increases in air temperature (Section 15.2.2). Uncooled water supplied at dry bulb temperature to
dust suppression sprays exacerbates the problem by increasing the humidity of the air. However, if
the service water is chilled, then it will both cool and dehumidify the air through which it is sprayed.
This effect will continue until the effluent water running into drains or on rock surfaces attains the
ambient wet bulb temperature.
The effect of chilled service water in a local workplace can be quite dramatic, depending upon the
temperature and flowrate of both the water and the ventilating airstream. Indeed, it is recommended
that the temperature of the service water should not be less than 12C to prevent cold discomfort of
personnel. A primary advantage of chilled service water is that it is applied where and when cooling is
most needed, i.e. where and when rock breaking or rock transportation is taking place.
The cooling efficiency of chilled service water is greater than that normally obtained from heat
exchangers because of the higher temperature of the return water. For example, water supplied to
sprays at 12C may eventually leave the district at the prevailing wet bulb temperature of, say, 27C,
giving a range of 15C. For comparison, typical temperature ranges of water across local heat
exchangers may be of the order of 6 to 10C. It follows that for the same cooling duty, lower water
flowrates can be used when the service water is chilled. This, in turn, reduces the required size of
pipes and pumping costs. Furthermore, the increased temperatures of return water enhance the
utilization of the low-cost pre-cooling towers on the mine surface.
The effectiveness of air cooling at the entrance to a stope or face is diminished by a large heat flux
from newly exposed or fragmented strata. However, such heat can be partially counteracted by the
use of chilled dust suppression water, enabling intake air cooling to be employed more effectively.
The benefits to be gained from chilled service water clearly depend upon how much water is required
by the equipment, and any adverse effects on the strata or travelling conditions that may arise from
too copious a supply of water. The practical approach is to determine the rate of water flow required
for dust suppression or other machine purposes and the expected temperature range. The fraction of
the total heat load removed by service water can then be estimated.
18- 61
Chapter 18
Example
Dust suppression water is supplied to a workplace at a rate of 100 litres per tonne of rock mined. If
the water is supplied to the machine at 12C and leaves the district at 26C, determine the cooling
provided by the service water at a mining rate of 5 tonnes per minute.
Solution
Water flowrate, mw =
100 5
= 8.333 litres or kg/s
60
Rate of heat removal = mw Cw t

= 8.333 x 4187 x (26 - 12)
= 488.5 x 103 W or 488.5 kW
The amount of service water used may vary from 20 to 3000 litres per tonne depending upon the
mining method and the mineralogical constituents of the dust particles. There is a corresponding
large range in the rate of possible heat removal by dust suppression water. However, due to the high
efficiency of the technique and the utilization of existing water lines, it may be regarded as the first
approach to the supply of "coolth" to a rock breaking area.
The fraction of total heat load that can be removed by chilled service water is eroded at greater
depths due to the increased effects of autocompression and geothermal heat, while the supply of
service water remains constant for any given mining method.
18.3.5.3. Hydropower
In section 18.3.4, we discussed the utilization of potential energy made available in shaft water pipes.
In addition to driving shaft station turbines or hydrolift systems, the water pressure available at spray
chambers can also be used to activate fans and pumps at those locations. The concept of
hydropower takes this idea further and combines it with service water cooling (Section 18.3.5.2).
The use of electrical or diesel power involves the production of heat. Expenditure is then incurred to
remove this heat from the mine by ventilation and/or cooling systems. The utilization of machine
power by such means involves both an environmental and an associated cost penalty. However, if
the machine is activated by chilled water under pressure provided by the shaft column, then the
hydraulic machine power and climatic control act in concert rather than in opposition. In a hot and
deep mine, chilled hydropower provides, simultaneously, a means of machine power, cooling and
dust suppression, all from a single system.
For relatively low powered units such as drills, hydraulic motors may be employed directly while, for
larger machines, turbine-electric combinations may be more practical. Furthermore, differing
combinations of water flow, pressure and temperature may be required for the demands of power,
cooling and dust suppression. A hydropower system should be designed and balanced to satisfy
those varying demands.
18.3.5.4. Ice systems
The limitations of heat rejection capability underground, together with the use of energy recovery
devices, have promoted the trend away from subsurface refrigeration plant towards surface units.
However, despite water turbines and hydrolift devices, pumping costs remain the factor which limits
the mass flow of water that can be circulated through the shaft pipelines of deep mines. Recalling the
18- 62
Chapter 18
relationship q = mw Cw t, heat removal from a mine can be enhanced by increasing the temperature
differential, t, between supply and return water. This can be accomplished by combinations of
surface and underground refrigeration plant (Section 18.3.5.1). Despite such advances in mine
cooling technology, heat remains the primary limitation on the depth at which mining can take place.
In order to proceed beyond the physical restraints imposed by the circulation of fluids, other means of
transmitting "coolth" have to be considered. One of these is to employ the latent heat of melting ice.
Although the use of ice in mines is not new (Section 18.1), the first pilot plant employing an ice
pipeline was constructed in 1982 at the East Rand Proprietary Mines in South Africa (Sheer, 1984).
Further full scale applications followed soon afterwards (Hemp, 1988).
Ice has a specific heat varying from 2040 J/(kg C) at 0C to 1950 J/(kg C) at -20C. It has a latent
heat of melting of 333.5 kJ/kg. The mass flow of water circulated around a mine for cooling purposes
can be reduced by a factor of over 5 if it is supplied in the form of ice.
Example
A mine requires 10 MW of cooling. Calculate the mass flow of water involved
(a) if the water is supplied at 3C and returns at 20C
(b) if the water is supplied as ice at -5C and returns at 20C.
Solution
(a) for the water system:
q = m w C w t w
or
mw =
C w t w
10 10 6
4187 (20 3 )
= 140.5 kg/s or litres/s
(b) for the ice supply:

Heat absorbed as the temperature of mi kg of ice increases from -5 to OC,
q ice = m i C i t i
Choosing the specific heat of ice as 2030 J/(kg C) gives

= mi x 2030 x 5 = 10 150 mi
qice
Latent heat absorbed as mi kg of ice melts at 0 C

qL = Lice mi
= 333 500 mi
Heat absorbed as mi kg of liquid water increases in temperature from 0 to 20C:

q w = m i C w t w = m i 4187 20 = 83 740m i
Then, total heat absorbed

q
q ice
qL
qw
10 10 6
mi (10 150 + 333 500 + 83 740) = 10 x 106 W

giving mi = 23.4 kg/s
18- 63
Chapter 18
Hence, in this example, the volume of cooling water is reduced by a factor of 6 when it is supplied as
ice at -5C. The calculation also illustrates the dominant amount of heat involved in the change of
phase from ice to liquid, qL.
There are four primary items to consider when investigating an ice system for mine cooling;
the large scale manufacture of ice,
the means of transporting it underground,
how it is best incorporated into a mine cooling system, and
the economics of the system.
Each of these factors will be considered in turn.
Manufacture of ice
The ice that is utilized in a mine cooling system may be supplied either as particulate ice at sub-zero
temperature or as a slurry of ice crystals within liquid water or brine.
Particulate ice can be manufactured as cubes, cylinders or tubes of clear ice, or as ice flakes that can
be compressed into pellets for transportation. Water is sprayed or caused to flow over the surfaces of
plates, tubes, concentric cylinders or drums that are cooled on their opposite sides by a refrigerant
fluid. The formation of ice creates an additional thermal resistance to heat transfer. This necessitates
an evaporator temperature of some -15 to -30C giving rise to coefficients of performance in the
range 2.3 to 3.5, lower than might be expected from a well designed water chiller.
The ice is usually removed from the cold surfaces by mechanical scrapers or by periodic cycling of
hot refrigerant through the freezer unit. In the production of flake ice, a thin film of ice is removed
continuously from the surface of a rotating drum or metal belt. The particles of ice fall into a hopper
that feeds into a screw or belt conveyor for removal from the unit.
As an example of large-scale particulate ice manufacture, a plant at East Rand Proprietary Mines
(South Africa) produced 6000 tonnes of ice per day from six 1000 tonne/day units. Each tubular unit
consisted of eighty double walled tubes with ice formation on both the inner (99 mm dia) and outer
(508 mm diameter) surfaces of the 4.5 m long tubes. The refrigerant passed through the annulus of
each tube. The 15 minute cycle consisted of 13.5 minutes of freezing and 1.5 minutes harvesting
(Hemp, 1988). The refrigerant was ammonia and ice harvesting was achieved by passing hot
ammonia liquid through the annuli in the icemaker tubes. This arrangement also improved the
efficiency of the unit by subcooling the liquid refrigerant before it returned to the evaporators.
Slurry icemakers promote the formation of microcrystals of ice within a stream of water or brine and
are at an earlier stage of development for mine cooling systems. Slurry ice has one significant
advantage over particulate ice. All industrial and domestic supplies of water contain many impurities
including dissolved salts of calcium, magnesium and sodium. The freezing process favours pure
water. Hence, if the freezing occurs sufficiently slowly or if the ice/water interface is continuously
washed with water, then the ice crystals will have a greater purity than the water in which they grow.
The dissolved salts remain in the liquid and, hence, become more concentrated. In contrast, the
formation of plate, tube or flake (particulate) ice occurs quickly and with a more limited washing action
by the water, giving rise to the entrapment of salts within the ice matrix. If the microcrystals, usually of
less than 1 mm in size, are removed from the ice slurry and washed, then they will be relatively free
from impurities. This process is known as freeze desalinization. The provision of purified water to a
mine cooling system assists greatly in minimizing corrosion and scaling within pipes and heat
exchangers.
There are at least three methods that have the potential for large-scale manufacture of slurry ice
(Shone, 1988). In the indirect process, water or brine is passed through tubes that are surrounded by
cold refrigerant. Microcrystals of ice are formed within the moving stream of water. Careful control of
the salinity, flowrate and temperature is required to prevent the formation of solid ice on the sides of
the tube. Scraper devices can also be used to prevent ice from accumulating on the walls.
18- 64
Chapter 18
The vacuum ice-making process involves the evacuation of water vapour from a vessel that contains
brine. The triple point of water is reached at 0.6 kPa and 0 C when boiling and freezing occur
simultaneously (ref. Figure 18.1). Ice, liquid water and vapour coexist at this point. The ice slurry that
is formed is kept in motion by an agitator until it is pumped from the vessel. About 1 kg of water
vapour must be removed for each 7.5 kg of nucleated ice. The vapour is compressed and condensed
for recycling. In the vacuum ice-making process, the water acts as its own refrigerant fluid (Section
18.2.2).
An even less developed method is the direct process in which a mixture of brine and immiscible liquid
refrigerant are sprayed through a nozzle into a receiving chamber. The evaporation of the refrigerant
cools the mixture and promotes the nucleation of ice crystals within the brine. The ice slurry collects
at the bottom of the vessel and is pumped out. The refrigerant vapour is evacuated from the top of the
vessel for compression, condensing and recycling.
In all three processes, the salinity of the brine has an important influence on the manufacture of ice
slurry. In general, a higher concentration of dissolved salts will give a smaller production of ice
crystals but at greater purity. The freezing temperature will also be lower, resulting in a reduced
coefficient of performance for the unit. For mine cooling systems, a salt concentration between 5 and
15 percent would appear to be appropriate. However, the optimum salinity for any given installation
should be determined from tests on the actual water to be used.
Future employment of ice slurries in mine cooling systems may be combined with energy recovery
devices. The duties of direct-contact heat exchangers are considerably enhanced when the sprays
are supplied with ice slurry of up to 20 percent ice (Gebler, 1988). Droplets in the spray chamber
remain at 0C until the ice crystals are melted. Tests have indicated that with an ice fraction of only 40
percent, pipeline pressure drops are about three times those for water (Shone, 1988).
Transportation of ice
The mass of ice required for a mine cooling system is such that it must be supplied continuously
rather than in a batch transportation system. While slurry ice can be pumped through pipes,
particulate ice must be transported from the ice-making plant to the shaft by conveying systems or by
a hydraulic or pneumatic pipeline. In the latter case, the supply air temperature should be not more
than 8C in order to prevent undue agglomeration of the ice particles.
The particulate ice falls through the shaft pipeline either as dilute flow in which the particles are
separated and retain their individual identity or as dense or slug flow. In the latter case, the ice
particles agglomerate into discrete slugs separated by air spaces. If the ratio of ice to air is too great,
then extrusion flow will develop accompanied by a greatly increased risk of pipe blockage. The
factors involved in the type of flow through an ice pipeline are the ice/air ratio, the shape, temperature
and size of the particles, and the type and size of pipe. Unplasticised (U) PVC has been found to be a
satisfactory pipe material for ice lines. However, changes in cross section or lips at expansion joints
should be avoided. At shaft stations, radii of curvature of 3 m or more allow the pipelines to be
extended horizontally for several hundred metres. To cite the East Rand Mines example again, four
uninsulated UPVC pipes of inside diameter 270 mm and wall thickness 22.5 mm were employed
(Hemp, 1988). The longest pipe extended through a vertical distance of approximately 2900 m. Each
pipe was capable of carrying 200 tonnes per hour (55.6 kg/s) allowing for considerable future
expansion of the mine cooling duty.
18- 65
Chapter 18
Incorporation into the mine cooling system

Figure 18.25 gives a simplified representation of one layout that utilizes particulate ice.
pre-cooling
tower
water
pre-chiller
ice-making
plant
condenser
evaporator
hot water dam
ice silo
cold water
dam 0C
to mine
from mine
Figure 18.25 Layout of an ice system. Ice silos may be located on several levels.
Return water pumped from the mine passes through a surface pre-cooling tower and is further cooled
in a water pre-chiller. This may be supplied by "coolth" from the ice-making refrigeration plant or,
alternatively, be a separate water chiller package. The particulate ice falls through the shaft pipeline
and is discharged into an ice/water mixing dam or a silo with a perforated base. Water from the hot
water dam is also sprayed into the ice silo. The discharge from the silo enters the cold water dam at
0C.
The bed of ice in the silo should be maintained at a thickness of 1 to 2 m. The supply of ice to each
silo must be matched to the required cooling duty at that level. Position transducers can be used to
monitor the ice level in each silo. The transmitted signals may be employed to control ice feed rates.
It is advisable to provide sufficient ice bunkerage to handle short term fluctuations in demand. Longer
18- 66
Chapter 18
term variations can be handled by activating an appropriate number of individual sets in the icemaking plant or by compressor control on the refrigeration units (Section 18.2.3).
Economics of ice systems
The benefits of ice as a medium of heat transfer in mines may be listed as follows:
greatly reduced water flow in shaft pipelines; therefore - smaller pipes and lower pumping costs
water is available at 0C underground rather than the 3 to 6C common with conventional
chilled water systems; therefore there is less flow and reduced pumping costs in the subsurface
(secondary) circuits, and improved performance of heat exchangers.
coolth" is stored in ice bunkers and silos to satisfy short term variations in demand
the system is more easily capable of being uprated for future increases in cooling load
the subsurface system is simpler having reduced or eliminated the need for turbines
the quality of the water is improved by ice-making.
The major disadvantage of the system is the considerable increase in both capital and operating
costs of the ice-making plant. As the depth of a mine increases and the heat load and associated
pumping costs both rise, then, at some point, it becomes economically advantageous to convert from
water chilling to an ice system. A study in South Africa showed that the crossover point occurs at a
depth of approximately 3000 m (Sheer, 1984). However, this will be influenced not only by the cost of
hardware but also power charges, surface climate, geothermic gradient, rock properties and heat
from equipment or stored materials (in the case of subsurface repositories). It is, therefore, prudent to
conduct a site-specific economic study on any proposed ice system.
18.3.6. Summary of design process
Although this chapter has concentrated on the hardware associated with mine air conditioning
systems, it is pertinent to remind ourselves that the design of such systems cannot be separated from
ventilation planning (Sections 9.1 and 16.3.5). In particular, the initial question is how much heat can,
or should, be removed by ventilation and how much by cooling systems.
This leads us into the first stages of the design procedure for a mine cooling system.
1.
Establish the expected heat load for the entire mine and for each identifiable district, level or
zone in the mine. The techniques for doing this are discussed in Chapters 15 and 16.
2.
Again, for the whole mine and each individual area, determine the airflows required to deal with
dust and polluting gases, taking into account air velocity limits (Section 9.3). Ventilation network
analyses should be conducted in order to establish airflow distribution and leakage patterns
(Chapter 7).
3.
Determine the heat removal capacities of the airflows (Section 9.3.4). If this is greater than the
corresponding heat loads for individual sections of the mine, then the heat can be removed
entirely by ventilation and there is no need to consider a cooling system.
4.
Conduct exercises to determine the feasibility of removing the excess heat by increased airflows
without exceeding velocity limits set by physiological, economic or legislative considerations
18- 67
Chapter 18
(Section 9.3.6). If the additional airflows are unacceptable, then the need for air conditioning is
established and the design of the cooling system can proceed.
5.
Determine the heat to be removed by cooling as the difference between the heat load and the
heat removal capacity of the air (Section 9.3.4). Preferably, a climatic simulation program should
be used in order to conduct a detailed analysis of the positions and duties of heat exchangers
and application of chilled service water (Chapter 16).
Interactive studies should also be conducted to optimize between ventilation and air conditioning
(Figures 9.1 and 16.4) (Anderson, 1986). These investigations will establish the required
distribution of air cooling devices. If this is limited to a few localized areas, then the feasibility of
employing spot or district coolers (Section 18.3.1) should be investigated. If, however, the heat
problem is more widespread, then the study should be widened to examine centralized
refrigeration plant.
6.
Investigate the alternative locations of refrigerating plant (Section 18.3.5.1) and the feasibility of
employing energy recovery devices (Section 18.3.4).
7.
By summing heat exchanger capacities and allowing for line losses, establish the duty of the
refrigeration plant.
8.
Estimate evaporator and condenser temperatures on the basis of the desired temperature of the
cooled medium and the heat rejection facilities respectively.
9.
Determine the required flowrates of chilled water and, hence, the sizes of pipelines (Section
18.3.3.1) and pumping duties.
10. Invite tenders from manufacturers of refrigeration plant and ancillary equipment, including
valves, pipelines, pumps, heat exchangers, instrumentation and controls.
11. Establish capital, installation and operating costs of the cooling system.
18.4. AIR HEATING

In the colder countries of the world, the temperature of intake air entering a mine may fall well below
0C for a major part of the year. This can result in severe problems along intake routes, and
particularly in surface connecting downcast shafts, slopes or adits. First, a build-up of ice on shaft
walls and fittings can cause a significant increase in resistance to airflow. Furthermore, large pieces
of ice dislodged by shaft operations or melting can present a hazard to personnel working in or near
the shaft. Repeated cycles of freezing and thawing occurring on a daily or seasonal basis can result
in severe damage to concrete linings in shafts or other intake airways, particularly when there is a
seepage of ground water through the lining.
Secondly, sub-zero temperatures can have adverse effects on the operation and maintenance of
equipment. This may result in poor productivity as well as a reduction in safety.
Thirdly, personnel who are inadequately protected against a cold environment will suffer the
symptoms described in Section 17.6. Here again, this will lead to diminished levels of productivity and
a deterioration in safety standards.
Winter temperatures can fall below -40C at the surfaces of mines in cold climates (Hall, 1988;
Moore, 1985). In these conditions, it becomes necessary to heat the intake air. This is the situation
that obtains at many Canadian mines (Hall, 1989).
18- 68
Chapter 18
The methods of air heating suitable for mine applications may be listed as follows:
utilization of waste heat produced from plant or processes on the mine surface
heat recovery from warm return air
direct heating
indirect heating
ice stopes
geothermal or cycled storage heating.
Each of these techniques is examined in the following subsections. To choose the method most
applicable for any given facility, it is necessary to consider technical feasibility, reliability, flexibility and
economics. Because of the large airflows that are usually involved, the costs of heating may far
exceed those of producing the ventilation (Hall, 1985). In very cold climates, it is often the aim simply
to increase the intake air temperature to about 0C and to allow autocompression, geothermal and
other sources of heat to produce more comfortable temperatures in the main working areas.
In order to minimize operating costs, air heating systems might be employed only during main
working shifts and when the surface temperature drops below a preset value. Furthermore, two or
more of the systems listed above may be combined to minimize energy demand or to provide
additional capacity for periods of abnormally low temperature.
There is, however, one situation in which air heating should be avoided or used with great caution.
This occurs when the mine workings are in permafrost. Allowing air temperatures to exceed 0C may
cause partial melting of the permafrost and give rise to problems of ground control in the mine.
Indeed, at the Polaris Mine, 1400 km from the North Pole, refrigeration techniques have been
employed during the short summer in order to maintain the intake temperatures below the melting
point of permafrost (Van der Walt, 1984).
18.4.1. Utilization of waste heat
The first matter to investigate for any proposed air heating project is whether heat is available from
other plant or processes on the surface of the mine. Secondly, the costs and technical feasibility of
using that thermal energy to raise the temperature of intake air should be studied. Common sources
of waste heat are compressor stations, generators and mineral processing operations (de Ruiter,
1989). If the waste heat is in the form of hot water or steam, then it may be pumped directly through
insulated pipes to heat exchanger coils located within all, or part, of the intake airstream. Ice
accumulations on the heat exchanger can be avoided by using a single array of coils only.
In water systems, an automatic procedure should be incorporated in order to empty the coils and
associated pipework during a shut-down period or if the water temperature approaches freezing point.
A further safeguard against frost damage is to employ a glycol-water mixture. A 50 percent glycol
mixture lowers the freezing point to -40C.
If the available heat source provides relatively low temperature thermal energy, then it may be
necessary to employ a heat pump (Section 18.2). Heat pumps may also allow the utilization of natural
heat sources such as a large lake or the ocean, provided that these remain unfrozen through the
winter months.
18.4.2. Heat recovery from exhaust air
This method may be considered if the downcast and upcast shafts are close to each other and there
is a significant difference between the intake and return air temperatures. Figure 18.26 shows that
during cold periods, the warm air returning from the mine is diverted through a direct contact spray
chamber, transferring useful heat to the descending water droplets. Water from the sump is pumped
18- 69
Chapter 18
through a heat exchanger located within the intake airstream. Again, it is advisable to employ a single
array of coils to minimize ice build-up, and to allow for automatic dumping of water from the heat
exchanger and pipework when the flow ceases.
cooled water
spray
chamber
intake
air
summer
heat
exchanger
external
heat
source
fan
winter
sump
heat
exchanger
warmed water
warm air from
upcast shaft
downcast
shaft
Figure 18.26 Heat recovery from return air augmented from an external heat source.
The spray chamber may be replaced by a tube coil heat exchanger. However, this reduces the overall
heat transfer efficiency of the system and also introduces the dangers of ice accumulation and
corrosion of the tubes, particularly if the return air is humid.
Heat exchange from upcast to downcast airflows is seldom practicable by itself. However, it may be
employed in combination with an external source of heat as shown on Figure 18.26. If that source is
waste heat from other plant, then it may be kept in continuous operation. However, if the external
source of heat is gas, oil or electricity, then it may be used only during working shifts and when the
surface air temperature falls below a set value.
A more direct means of recovering some of the heat from warm return air is by controlled partial
recirculation (Hall, 1988, 1989). This technique gives one hundred percent efficiency of heat transfer
from return to intake for that fraction of the air which is recirculated. This is limited by the rate at which
gaseous and particulate pollutants are added to the mine air. The system should be capable of
reverting to normal through-flow ventilation should monitors indicate a trend towards unacceptable
concentrations of pollutants or following blasting operations. The volume of air that is recirculated
may be controlled automatically by the pollutant concentrations in the return air or set at a fixed
18- 70
Chapter 18
fraction of the return airflow. The recirculating cross-cut should be located as close to the surface
connection as practicable in order to take advantage of the pollutant dilution arising from air leakage.
If the level of contamination by airborne pollutants is sufficiently low to allow a reduction in the
through-flow of fresh air, then the costs of heating the intake air will also be reduced. However, any
proposed system of controlled partial recirculation must incorporate all of the safeguards and
precautions discussed in Section 4.5.
18.4.3. Direct heating
The employment of electrically heated elements to raise the temperature of mine air is limited by high
operating costs. It is practical only for very light duties or where electrical power is particularly cheap.
Natural gas, propane or other light hydrocarbon fuels may be injected through nozzles and burned in
flame jets within the intake airstream. As no intermediate heat exchangers are involved, this is the
most efficient way of using the fuel. The products of combustion, carbon dioxide and water vapour,
are added to the airstream in relatively small amounts. However, the gas nozzles should be well
maintained to prevent the formation of carbon monoxide through incomplete combustion. A prudent
precaution is to locate a gas monitor downstream from the jets with alarms and automatic fuel cut-off
at a predetermined concentration of carbon monoxide.
The rate of fuel consumption for any given heating duty may be calculated from the calorific value of
the fuel. At the temperatures to which the products of combustion are cooled in mine air, the majority
of water vapour will condense. Hence, it is the gross calorific value that should be used. Table 18.4
gives the gross calorific values of several gases. For natural gas and other mixtures, an approximate
value may be obtained by summing the calorific values of the constituents weighted according to
percentage composition.
Gas
carbon monoxide
hydrogen
methane
ethylene
ethane
propylene
propane
butane
natural gas
(depends upon composition)
Gross calorific value

MJ/m3
11.8
11. 9
37.1
58.1
64.5
85.7
93.9
121. 8
26 to 56
Table 18.4. Gross calorific values of gases

based on a temperature and pressure of 15C and 101.3 kPa respectively.
For petroleum fuel oils, gross calorific values vary from 42 to 47 MJ/kg.
Example
An airflow of 100 m3/s and density of 1.4 kg/m3 is to be heated from -20C to 2C by propane jets.
(a) Determine the rate of fuel consumption in m3 per hour.
(b) If the air is supplied at 21 percent oxygen and 0.03 percent carbon dioxide, determine the
corresponding concentrations downstream from the jets.
18- 71
Chapter 18
Solution
(a) Mass flow of air ma
Heat required, qa
= Q
= 100 x 1.4
= 140 kg/s
= ma
x Cp
x t
= 140 x 1005 x (2 = 3.095 x 106
W
3.095
MW.
or
(-20))
[To be precise, we should take the presence of water vapour into account. However, the effect is
very small at these low temperatures and may be neglected.]
The gross calorific value of propane is selected from Table 18.4 as 93.9 MJ/m3. Hence, heat
produced at the jets
MJ m 3
m3 s
q p = 93.9 propane flowrate
or MW
As qp = qa = 3.095 MW, the required propane flowrate = 3.095/93.9

or
= 0.0330 m3/s
0.0330 x 3600 = 118.7 m3 per hour at standard temperature and pressure.
(b) The chemical balance for burning propane is

C3H8
1m3 propane +
502
5m3 oxygen
3 CO2
+ 4 H2O
3m3 carbon + 4m3 water

dioxide
vapour
Therefore 0.0330 m3/s of propane will consume

5 x 0.0330 = 0.165 m3/s oxygen to produce
3 x 0.0330 = 0.099 m3/s carbon dioxide
At entry, the airflow includes
100 x 0.21 = 21 m3/s oxygen and
100 x 0.0003 = 0.03 m3/s carbon dioxide.
Downstream from the jets,

oxygen flowrate
carbon dioxide flowrate
= 21 - 0.165
= 0.03 + 0.099
= 20.835 m3/s and

= 0.129 m3/s
As the airflow is 100 m3/s these same values give close approximations to the concentrations of
the gases downstream from the jets.
18.4.4. Indirect heating
In order to prevent products of combustion from entering the mine ventilation system, fuels may be
used in a separate burner to heat a glycol-water mixture. This is then recirculated through an indirect
heat exchanger located within the intake air. Again, a single array of coils will minimize the risk of ice
accumulation on the tubes. This technique of indirect or offline heating inevitably gives a lower
thermal efficiency than open flame jets within the intake air. However, the flowrate of the glycol
mixture provides an added measure of control and flexibility.
18- 72
Chapter 18
18.4.5. Ice stopes

Figure 18.27 illustrates a metal mining technique that allows air to be heated from sub-zero
temperatures to approaching 0C at very low cost. Water is sprayed into the top of an abandoned
open stope. Cold air coming directly from the downcast shaft or slope enters near the top and
circulates through the stope, freezing the water droplets, before exiting at the other end.
spray level
sub 0C air from

downcast shaft or
ramp
to mine
drainage level
Figure 18.27 Depiction of an ice stope.

The air is heated by three mechanisms:
(a) directly from the liquid water droplets,
(b) by latent heat of fusion (333.5 k/Jkg) as the water freezes, and
(c) by strata heat from the wall rock.
The ice particles fall and form an ice bed that gradually accumulates in the stope through the winter
months.
In summer, warm intake air may be cooled by circulating it through the same stope and melting the
ice. A drain level at the base of the stope carries water to a main sump. However, the air cooling may
be enhanced by circulating the liquid melt back through the sprays.
The ice stope technique has been used very successfully at the Stobie Mine of Inco Ltd. in Canada
(Stachulak, 1989). In this mine an airflow of 307 m3/s is heated in two ice stopes at a cost estimated
at about one-ninth that for equivalent gas heating. In that installation, water is supplied to the nozzles
18- 73
Chapter 18
at a pressure of between 550 and 760 kPa. Compressed air is added to the line shortly before the
nozzles. This provides a finely divided spray and prevents icing up of the jets. The compressed air is
also used to clear water from the pipes when the system is to be shut down.
The disadvantages of the ice stope technique are that it requires the availability of large open
excavations and also necessitates an increased pumping capacity for the mine drainage system.
Example
Water at 7C is sprayed at a rate of 30 litres per second into an ice stope. Air enters the stope at -22 C
and is required to be heated to -2 C. Calculate the maximum air flowrate through the stope that can
attain this 20 C increase in air temperature.
Solution
Mass flowrate of water, mw = 30 kg/s
Heat released by water in cooling from 7C to 0C
qw = mw Cw tw
=
30 4187 7
1000
= 879 kW
Heat of fusion released by ice formation (latent heat of fusion, Li = 333.5 kJ/kg)
qi = Li mw = 333.5 x 30 = 10 005 kW
Total heat transferred to air
qa = qw + qi = 879 + 10 005 = 10 884
kW
(This illustrates the dominant effect of the latent heat of ice formation.)
However, this is also equal to
giving
ma
ma Cp tair
10 884 1000
= 541
1005 20
kg/s
If the air density were 1.3 kg/m3, this would give an air volume flowrate of 541/1.3 = 417 m3/s.
18.4.6. Geothermal and cycled storage heat
Another low cost method of controlling the temperature and humidity of intake air is to course it
through one or more sets of old workings or fragmented strata before admitting it to current work
areas. This takes advantage of a combination of natural geothermal energy and the "thermal
flywheel" effect of heat storage within the envelope of rock surrounding a mine opening (Section
15.2.2).
The flow of geothermal heat into a mine airway is greatest at the moment of excavation and reduces
with time until near equilibrium is attained when temperatures in the surrounding envelope of rock no
longer change significantly with respect to time (Figure 15.3). If, however, the inlet air is cycled
between hot and cold, then the surrounding rock will act as a storage heater, absorbing heat during
the hot periods and emitting it during the cold periods. This cyclic behaviour is superimposed upon
the longer term flux of geothermal heat.
18- 74
Chapter 18
Let us assume a situation in which the virgin rock temperature (VRT) at a given level is higher than
the mean annual dry bulb temperature of air entering that level from the downcast shaft or slope. In
order to moderate the extremes of winter temperature, the air is passed through a set of old workings
that we may refer to as the control district. Figure 18.28 illustrates the seasonal variation of air
temperature at inlet and exit of a control district, and ignoring the shorter-term diurnal changes. The
moderating effect of the control district is shown clearly. The inlet peaks and troughs of summer and
winter temperatures are damped very considerably by the time the air reaches the exit.
summer peaks at inlet
winter troughs at inlet
Figure 18.28 Example of the variation in air temperature at the inlet and exit of a
climate control district
However, the figure also shows a longer term trend over which the average exit temperature falls until
a repeated cyclic variation is attained. If the winter temperatures of the air leaving the control district
are acceptable in this condition of dynamic equilibrium, then that district can be used indefinitely to
moderate the extremes of seasonal variations. If, however, after a few years the winter exit
temperatures fall below tolerable levels, then the intake air should be diverted through a younger
control district.
Control districts that are no longer effective may be allowed to regenerate by sealing them off. The
surrounding rock temperature will gradually tend towards the VRT. After a regeneration period which
may be several years, the panel can again be opened up as a control district. If several old districts
are available, then some of them may be used as control districts at any one time while others are
regenerating (Moore, 1985).
18- 75
Chapter 18
The actual form of Figure 18.28 for any given mine depends upon the extremes of surface climate,
the rate of airflow through the control district and the thermal properties of the surrounding strata.
Simulation programs (Chapter 16) are most helpful in predicting the cyclic behaviour of the air
temperatures (Moore, 1985; MVS, 1990). Manual methods of estimating the effects of varying inlet
temperatures on strata heat flow have also been developed (Hemp, 1982) .
Although the use of a control district can yield a useful degree of air conditioning, it does have some
disadvantages and limitations. First, the passage of air through old workings requires additional fan
energy. This can be minimized by arranging for the air to move fairly slowly along multiple flow paths,
requiring only a small applied pressure differential. Simulation tests coupled with site-specific
measurements allow the optimum layout to be determined.
Secondly, old workings are liable to changes in resistance due to strata closure and falls of roof.
Hence, in order to ensure continuity of ventilation, control districts should be capable of being
bypassed at short notice. Neither should they be relied upon as emergency escapeways.
Thirdly, the conditioning of intake airflow through old workings is untenable where there is a danger of
spontaneous combustion or significant emissions of strata gases.
In addition to moderating temperatures, control districts can also serve a useful purpose where
variations in intake humidity cause roof control problems. This can occur in deposits of salt or some
shales. The hygroscopic nature of these rocks results in absorption of water vapour and a loss of
mechanical strength (Robinson et al, 1981).
Where caving techniques in a metal mine have left a fragmented and permeable rock mass that
extends to the surface, then winter air may be heated and summer air cooled by drawing intake
ventilation through the broken rock. This technique can combine the phenomena of storage heating
and the latent heat of ice formation. One of the best known examples of this essentially free form of
air conditioning is at the Creighton Mine in the Sudbury Basin of Canada (Stachulak, 1989). Intake air
is drawn through a bed of broken rock that is approximately 244 m by 183 m in plan and 137 m high.
Precipitation and seepage results in ice formation during the winter and melting in the summer. The
air progresses through old drawpoints, slusher drifts and ore passes where it is regulated according
to its temperature. This technique results in intake air of near constant temperature reaching the
deeper current workings throughout the year.
References
Anderson, J. and Longson, I. (1986). The optimization of ventilation and refrigeration in British coal
mines. The Mining Engineer. Vol. 146, pp 115-120.
Arkle, K. et al (1985). Use of thermal insulation materials in mines. Jnl. Mine Ventilation Soc. of S.
Africa. Vol. 38, No.4, April, pp 43-45.
ASHRAE (1988). Handbook on Equipment. Ch. 20. Cooling towers. American Society of Heating,
Refrigerating and Air Conditioning Engineers.
Baker-Duly, C. et al (1988). Design of a large flexible underground refrigeration installation. 4th Int.
Mine Ventilation Congress, Brisbane, Australia, pp 443-449.
Bluhm, S.J. (1980). Predicting performance of spray chambers for cooling air. Heating, Refrigeration
and Air Conditioning, S. Africa, pp 27-39.
Bluhm, S.J. (1981). Performance of direct contact heat exchangers. Jnl. Mine Ventilation Society of
S. Africa. Vol. 34, Aug. and Sept., Nos. 8 and 9.
18- 76
Chapter 18
Bluhm, S.J. (1983). Spot cooling of air in direct contact heat exchangers. Report of Environmental
Engineering Lab. Chamber of Mines Research organization, South Africa.
Burrows, J. (1982). Refrigeration - theory and operation. Environmental engineering in S. African
mines. Chapter 23. Published by Cape and Transvaal Printers (Pty), Cape Town.
deRuiter, H. et al (1989). Geothermal heat recovery at Lac Minerals Ltd., Macassa Division. 91st
CIMM Annual General Meeting. May, Quebec, Canada.
Gebler, W.F. et al (1988). The design and evaluation of a large direct-contact ice-air cooler for Vaal
Reefs Gold Mine. 4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 425-432.
Hall A.E. (1985). The use of controlled recirculation ventilation to conserve energy. 2nd U.S. Mine
Ventilation Symposium, Reno, Nevada, pp. 207-215.
Hall, A.E. et al (1988). The use of controlled recirculation to reduce heating costs in Canada.
4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 301-307.
Hall, A.E. et al (1989). Controlled recirculation investigation at Central Canada Potash Division of
Noranda Minerals, Inc. 4th U.S. Mine Ventilation Symposium, Berkeley, CA, pp 226-233.
Hancock, W. (1926). Local air-conditioning underground by means of refrigeration. Trans. Inst.
Mining Engrs., Vol. 72, pp 342-366.
Heller, K. et al (1982). Development of a high performance, low maintenance, in-line water spray
cooler for mines. 1st US Mine Ventilation Symposium, Tuscaloosa, AL.
Hemp, R. (1967). The performance of a spray-filled counterflow cooling tower, Jnl. Mine Ventilation
Soc. of S. Africa, Vol. 25, pp 159-173.
Hemp, R. (1972). Contribution to the design of underground cooling towers by A. Whillier. Jnl. Mine
Ventilation Society of S. Africa, Vol. 25, No.6, June, pp 85-93.
Hemp, R. (1982). Environmental engineering in South African mines, Chapter 22. Sources of heat in
mines, pp 569-612. Published by Cape and Transvaal Printers (Pty), Cape Town.
Hemp, R. (1988). A 29 MW ice system for mine cooling. 4th Int. Mine Ventilation Congress, Brisbane,
Australia, pp 415-423.
McAdams, W.H. (1954). Heat transmission. McGraw-Hill, New York, 3rd. ed.
Moore, D.T. (1985). Geothermal air heating in a southwestern Wyoming trona mine. 2nd U.S. Mine
Ventilation Symposium, Reno, Nevada, pp 561-570.
Mcke, G. and Uhlig, H. (1984). Performance of finned coil and water spray coolers. 3rd Int. Mine
Ventilation Congress, Harrogate, England, pp 231-237.
MVS (1990). Investigation of geothermal air heating for General Chemical's Green River soda ash
operations. Mine Ventilation Services, Inc. Report to General Chemical Corp., May, pp 1-33.
Ramsden, R. and Bluhm, S.J. (1984). Air cooling equipment used in South African gold mines.
3rd Int. Mine Ventilation Congress, Harrogate, England, pp 243-251.
Ramsden, R. and Bluhm, S. (1985). Energy recovery turbines for use with underground air coolers.
2nd U.S. Mine Ventilation Symposium, Reno, Nevada, pp 571-580.
18- 77
Chapter 18
Ramsden, R. (1985). Insulation used on chilled water pipes in South African gold mines. Jnl. Mine
Ventilation Soc. of S. Africa, Vol. 38, No.5, May, pp 49-54.
Reuther, E.U. et al (1988). Optimization of spray coolers for cooling deep coal mines.
4th Int. Mine Ventilation Congress, Brisbane, pp 451-458.
Robinson, G., Weller, R.C and McPherson, M.J. (1981). Underground environmental planning at
Boulby Mine, Cleveland Potash Ltd. Trans. Inst. of Mining and Metallurgy (U.K.), July. Reprinted in
the Jnl. of S. African Mine Ventilation Soc., Vol. 35, No.9, Sept. 1982, pp 73-88.
Rose, H. (1989). Personality Profile. Jnl. Mine Ventilation Soc. of S. Africa, Vol. 42, No.6, June, pp
110-113.
Sheer, T.J. et al (1984). Research into the use of ice for cooling deep mines. 3rd Int. Mine Ventilation
Congress, Harrogate, U.K., pp 277-282.
Shone, R.D.C. and Sheer, T.J. (1988). An overview of research into the use of ice for cooling deep
mines. 4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 407-413.
Stachulak, J. (1989). Ventilation Strategy and Unique Air Conditioning at Inco Limited. 4th U.S. Mine
Ventilation Symposium, Berkeley, CA, pp 3-9.
Stroh, R.M. (1980). Refrigeration practice on Anglo American gold mines. Mine Ventilation Soc. of S.
Africa Symposium on Refrigeration and Air Conditioning. Published by Cape and Transvaal Printers
(Pty), Cape Town.
Stroh, R.M. (1982). Refrigeration practice, Chapter 24, and Chilled water reticulation, Chapter 25.
Environmental engineering in S. African mines. Published by Cape and Transvaal Printers (Pty),
Cape Town.
Thimons, E.D. et al (1980). Water spray vent. tube cooler for hot stopes. US Bureau of Mines Tech.
Progress Report TPR 107.
Van der Walt, J. (1984). Cooling the world's coldest mine. Jnl. S. African Mine Ventilation Society.
Vol. 37, No. 12, Dec., pp 138-141.
Van Vuuren, S.P.J. (1975). The optimization of pipe sizes in a refrigeration system. Jnl. Mine
Ventilation Soc. of S. Africa, Vol. 28, No. 6, pp. 86-90.
Weuthen, P. (1975). Air coolers in mines with moist and warm climatic conditions. 1st Int. Mine
Ventilation Congress, Johannesburg, S. Africa, pp 299-303.
Whillier, A. (1967). Pump efficiency determination from temperature measurements. South African
Mechanical Engineer, Oct., pp 153-160.
Whillier, A. (1977). Predicting the performance of forced draught cooling towers. Jnl. Mine Ventilation
Soc. of S. Africa, Vol. 30, No. 1, Jan., pp 2-25.
Whillier, A. (1982). Heat transfer: Environmental engineering in S. African mines, Chapter 19.
Published by Cape and Transvaal Printers (Pty), Cape Town.
18- 78
Part 5
Dust
The hazardous nature of dusts
CHAPTER 19. THE HAZARDOUS NATURE OF DUSTS

19.1. INTRODUCTION .........................................................................................1
19.2. CLASSIFICATIONS OF DUST ....................................................................2
19.2.1. Size of aerosol particles................................................................................................... 2
19.2.2. Physiological effects of dusts........................................................................................... 3
19.3. DUST IN THE HUMAN BODY .....................................................................4

19.3.1. The respiratory system .................................................................................................... 4
19.3.2. Mechanisms of dust deposition in the respiratory system ............................................... 6
19.3.2.1. Impaction................................................................................................................... 6
19.3.2.2. Sedimentation ........................................................................................................... 6
19.3.2.3. Diffusion .................................................................................................................... 6
19.3.2.4. Interception................................................................................................................ 6
19.3.2.5. Electrostatic precipitation .......................................................................................... 7
19.3.3. Dust diseases .................................................................................................................. 7
19.3.3.1. Fibrogenic effects ...................................................................................................... 7
19.3.3.2. Coal workers' pneumoconiosis (black lung).............................................................. 7
19.3.3.3. Silicosis ..................................................................................................................... 7
19.3.3.4. Asbestosis ................................................................................................................. 8
19.3.3.5. Precautions against dust diseases............................................................................ 8
19.3.4. Threshold limit values ...................................................................................................... 9
19.4. THE ASSESSMENT OF AIRBORNE DUST CONCENTRATIONS ...........10

19.4.1. Background.................................................................................................................... 10
19.4.2. Particle count methods .................................................................................................. 11
19.4.3. Gravimetric methods...................................................................................................... 12
19.4.4. Photometric (light-scattering) methods .......................................................................... 13
19.4.5. Personal samplers ......................................................................................................... 15
19.4.6. Other methods of airborne dust measurement.............................................................. 16
19.4.7. Discrimination of quartz particles................................................................................... 16
19.4.8. Sampling strategy .......................................................................................................... 16
References ........................................................................................................18
19.1. INTRODUCTION
The natural atmosphere that we breathe contains not only gaseous constituents but also large
numbers of liquid and solid particles. These are known by the generic name aerosols. They arise
from a combination of natural and industrial sources including condensation, smokes, volcanic
activity, soils and sands, and microflora. Most of the particles are small enough to be invisible to
the naked eye. Dust is the term we use in reference to the solid particles. The physiologies of airbreathing creatures have evolved to be able to deal efficiently with most of the aerosols that occur
naturally. However, within closed industrial environments, concentrations of airborne particulates
may reach levels that exceed the ability of the human respiratory system to expel them in a timely
manner. In particular, mineral dusts are formed whenever rock is broken by impact, abrasion,
crushing, cutting, grinding or explosives. The fragments that are formed are usually irregular in
shape. The large total surface area of dust particles may render them more active physically,
chemically and biologically than the parent material. This has an important bearing on the ability
of certain dusts to produce lung diseases.
Respiratory problems caused by dust are among the oldest of industrial ailments. The first
legislation for mine dust appears to have been formulated in 1912 when the Union of South Africa
19 -
introduced laws governing working conditions in the gold mines of the Witwatersrand. Other
countries introduced similar legislation in the 1920's and '30's. However, those laws were
concerned primarily with silicosis and required proof of employment in siliceous rock mining
(Carver, 1975). The early legislation reflected medical opinion of the time, namely, that hardrock
dust caused silicosis which led to tuberculosis and eventual death. At that time coal dust was not
regarded as particularly harmful. However, the number of recognized cases of coal workers'
pneumoconiosis (CWP) increased dramatically through the 1930's. The British Medical Research
Council initiated an investigation into respirable disease within the anthracite workers of South
Wales. In 1936, the need for protective legislation in the United States was acknowledged by a
National Silicosis Conference. Both in Europe and the United States the hazards of coal dust
were identified first in anthracite mines but by 1950 it was confirmed that workers in bituminous
coal mines were also subject to coal workers' pneumoconiosis (known, also, in America as black
lung).
It took many years for a quantitative and definitive link to be established between the "dustiness"
of mine atmospheres and respiratory disfunction. In retrospect, there were three reasons for this.
First, it may take several years of exposure before the victim becomes aware of breathing
impairment. Secondly, the lung reactions to dust are often similar to those of naturally occurring
ailments and, thirdly, the commonly used measure of dust concentration was the number of
particles in a unit volume of air. Correlations between dust concentration measured in this way
and the incidence of pneumoconiosis were not clear.
A turning point occurred in 1959 at the International Pneumoconiosis Conference held in
Johannesburg, South Africa (Orenstein, 1959). Recommendations were made at that conference
which resulted in a re-direction of pneumoconiosis studies, particularly with regard to the methods
and strategies of dust sampling. It was recognized that those particles of equivalent diameter less
than 5 x 10-6 m (5 microns) were the ones most likely to be retained within the lungs. These were
named respirable dust. It was further established that the mass concentration of respirable dust in
any given atmosphere was a much better measure of the potential health hazard than the earlier
particle count methods. Instruments began to be developed that mimicked the dust retention
selectivity of the human lung and, furthermore, could be used continuously to give the integrated
effect over an 8 hour period.
In this chapter we shall concentrate on the effects of mineralogical dusts on the human
respiratory system and the techniques that are now employed for the sampling and measurement
of airborne dust.
19.2. CLASSIFICATIONS OF DUST

There are a number of ways to classify aerosol particles depending upon the purpose of any
given study. Two such classifications are particularly relevant to the subsurface environmental
engineer, first with regard to size distributions of the particulates and, secondly, in terms of
physiological effects.
19.2.1. Size of aerosol particles
Mine dusts vary widely in shape, dependent largely upon the prevalent mineral constituents. The
simplest method of quantifying the size of a non-spherical particle is the projected area or
equivalent geometric diameter. This is the diameter of a sphere that has the same projected area
as the actual particle. Other measures of equivalent diameter are defined in Chapter 20.
Typical size ranges of some common aerosols are given in Table 19.1. In general, the size
distribution within each range follows a lognormal curve. Particles do not become visible to the
naked eye until they are more than 10 microns equivalent diameter. It follows that the harmful
19 -
respirable dusts are invisible. Nevertheless, it must be assumed that heavy visible concentrations
of dust in a mine atmosphere are accompanied by high levels of respirable dust (Section 20.3.1.).
Another unfortunate aspect of the small sizes of respirable dusts is that they have a very low
settling velocity and, indeed, can remain suspended in air indefinitely.
Type of aerosol
Respirable dust
Coal and other rock dusts
Normal atmospheric dusts
Diesel particulates
Viruses
Bacteria
Tobacco smoke
Pollens causing allergies
Fog
Mist
Light drizzle
Size range in microns

(10-6 m)
Lower
Upper
7
0.1
100
0.001
20
0.05
1
0.003
0.05
0.15
30
0.01
1
18
60
5
50
50
100
100
400
Table 19.1. Size ranges of common aerosols.

19.2.2. Physiological effects of dusts
A classification of dusts with respect to potential hazard to the health and safety of industrial
workers may be divided into five categories.
1. Toxic dusts. These can cause chemical reactions within the respiratory system or allow toxic
compounds to be absorbed into the bloodstream through the alveolar walls. They are poisonous
to body tissue or to specific organs. Some metalliferous ores fall into this category. The most
hazardous include compounds of arsenic, lead, uranium and other radioactive minerals, mercury,
cadmium, selenium, manganese, tungsten, silver and nickel (Walli, 1982).
2. Carcinogenic (cancer causing) dusts. The cell mutations that can be caused by alpha, beta and
gamma radiation from decay of the uranium series make radon daughters the most hazardous of
the carcinogenic particulates (Chapter 13). A combination of abrasion of lung tissue and surface
chemical action can result in tumour formation from asbestos fibres and, to a lesser extent,
freshly produced quartz particles. Exposure to arsenic dust can also cause cancers. Diesel
exhaust particulates are a causative factor in lung and other types of cancer.
3. Fibrogenic dusts The scouring action of many dusts causes microscopic scarring of lung
tissue. If continued over long periods this can produce a fibrous growth of tissue resulting in loss
of lung elasticity and a greatly reduced area for gas exchange. The silica (quartz, chert) and
some silicate (asbestos, mica, talc) dusts are the most hazardous of the fibrogenic dusts and may
also produce toxic and carcinogenic reactions. Welding fumes and some metalliferous ores
produce fibrogenic dusts. Long and excessive exposure to coal dusts also gives rise to fibrogenic
effects.
4. Explosive dusts. These are a concern of safety rather than health. Many organic materials,
including coals other than anthracite, become explosive when finely divided at high
concentrations in air. Sulphide ores and many metallic dusts are also explosive. Hazards
associated with the explosive nature of some dusts are discussed more fully in Chapter 21.
5. Nuisance dusts. Quite apart from adverse effects on the health of personnel, all dusts can be
irritating to the eyes (Gibson and Vincent, 1980), nose and throat and when in sufficiently high
concentration may cause reduced visibility. Some dusts have no well-defined effects on health
but remain in the category of a nuisance dust. These include the evaporites (halite, potash,
gypsum) and limestones. The soluble salts of halite (NaCl) and potash (KCl) can occasionally
cause skin irritations, particularly around hatbands or tightly fitting dust masks.
19 -
19.3. DUST IN THE HUMAN BODY

The lungs are the organs where the oxygen necessary for metabolic activity is introduced into the
body. Through repetitive inhalations and exhalations air is brought into very close proximity to
flows of blood, the two being separated by a very thin membrane of about 0.5 microns in
thickness. Oxygen diffuses across the membrane from the air to the blood while carbon dioxide
diffuses in the other direction. The exchange is maintained by a difference in concentration
across the membrane for each of the two gases.
The respiratory system is equipped with defence mechanisms against gaseous or aerosol
pollutants that may exist in the inspired air. However, these mechanisms can be defeated by toxic
or carcinogenic agents. Furthermore, after years of exposure to unnaturally high concentrations
of dust, the defence system can simply become overloaded allowing the lungs to become much
less efficient as gas exchangers and, also, more susceptible to bronchial infections and
pulmonary illnesses.
In this Section, we shall outline the structure and normal operation of the human respiratory
system, the mechanisms of dust deposition within that system, and the processes that lead
towards those ailments known collectively as dust diseases.
19.3.1. The respiratory system
nasopharynx
larynx
trachea
bronchi
bronchioles
clusters of alveoli
covered by blood
capillaries
Figure 19.1 Diagrammatic representation of the human

respiratory system.
Figure 19.1 is a simplified

illustration of the human respiratory
system. Air enters through the
nostrils where it passes through a
mat filter of hairs in order to enter
the nasopharynx. This filter is the
first line of defence and removes
the larger dust particles. Those
remain trapped until they are blown
out or pass back through the
nasopharynx to be swallowed.
Within the larger volume of the
nasopharynx, the air velocity is
reduced. In this zone and, indeed,
throughout all of the branched air
passages leading to the alveoli, the
walls are lined with ciliated and
mucous-secreting cells. The
hairlike cilia wave to and fro with a
directional bias that promotes
movement of the mucous towards
the throat where it can be
swallowed. Most dust particles
greater than 10 microns in size are
captured by the hair filter or
mucous before inhaled air reaches
the larynx. Air that is breathed
through the mouth bypasses the
protection offered by the nostrils
and nasopharynx.
19 -
The air then progresses through the trachea, a tube some 20 mm in diameter, 120 mm long and
comprised of a series of tough cartilage rings. This divides into the left and right bronchi
averaging about 12 mm in diameter and 48 mm long. There are, of course, considerable
variations in these dimensions between individuals. The air passages continue to subdivide
repeatedly into the bronchioles in a manner that is analogous to the roots of a tree. After sixteen
subdivisions the bronchiole diameter is about 0.6 mm. The total number of branched passages in
human lungs may be over 80 million.
Although the diameter of individual air passages decreases with branching, their greatly
increased number produces a significant rise in total cross sectional area available for flow.
Hence, the air velocity decreases to the extent that airflow is completely laminar in the smaller
bronchioles. All of these passages are coated with mucous which is continuously expelled
upwards towards the trachea by the motions of cilia. The mucous secretions and any captured
particulates are usually expelled by coughing or swallowing within one day. Mucous coatings
within the air passages are normally quite thin. However, bronchial ailments can cause an
increase in both the viscosity and thickness of mucous layers to the extent that they can restrict
air movement significantly. The higher air velocities through constricted passages can lead to
audible noise (wheezing).
The smallest bronchioles terminate in clusters of air sacs or alveoli some 0.2 to 0.6 mm in size.
The walls of the alveoli provide the 0.5 micron thick membrane, or epithelium, across which gas
exchange occurs. There are an estimated three to four hundred million alveoli in the average
human lungs with a total area of some 75 m2 available for gas exchange in a healthy adult.
Involuntary muscular movements of the diaphragm and ribcage induce rhythmic changes of
pressure differential between the alveoli and the external atmosphere giving rise to normal cycles
of inhalation and exhalation. Breathing rates vary from 12 to 40 inhalations per minute depending
upon the level of activity. Table 11.2 indicates corresponding gas exchange rates for oxygen and
carbon dioxide.
Few dust particles of sizes greater than 3 microns reach the alveoli. Numerous models have been
proposed to indicate the relationship between particle size and sites of deposition (e.g. Walli,
1982; Schrder, 1982; National Research Council, 1980). For very small captured particles,
clearance from the alveoli may be slow or non-existent. Particles of 0.2 micron diameter have a
25 percent probability of retention while, for 0.02 micron particles, the retention rate rises to about
55 percent.
Mucous producing cells within the alveoli lubricate the surfaces and facilitate their freedom to
dilate and contract. Removal of dust particles from the alveoli is undertaken by relatively large (10
to 50 micron) cells known as macrophages or phagocytes. These nomadic cells originate in some
white blood cells, are free to move over the alveolar walls and are capable of engulfing particles
up to 10 microns in diameter. A commonly accepted view is that the macrophages and any
encapsulated dust particles migrate out of an alveolar cluster into the bronchioles where they can
be expelled by ciliary motion. Another theory is that macrophages are transported from the alveoli
through fluid filled capillaries that lead to lymphoids at bronchiole junctions (Brundelet, 1965).
The average life expectancy of a macrophage is about one month. However, these scavenger
cells may die rapidly if they engulf toxic particles. The number of macrophages increases when
the lungs are subjected to increased entry of dust particles. However, there remains considerable
doubt about the relationship between respirable dust concentration and macrophage production.
It appears to depend upon the mineralogical composition of the particles. Furthermore, dust
loaded cells are cleared less rapidly than the normal turnover of dust-free cells (National
Research Council, 1980).
Soluble particles dissolve in mucous. The resulting ions are either removed in the mucous or
diffuse through the epithelium for elimination through the bloodstream.
19 -
19.3.2. Mechanisms of dust deposition in the respiratory system

Deposition of dust particles within the zones of the respiratory system varies with the size and
aerodynamic characteristics of the particles, the geometry of the air passages and the patterns of
airflow. The three most important mechanisms of deposition are impaction, sedimentation and
diffusion, while interception and electrostatic precipitation can become significant for particular
types of dust. These five modes of deposition are discussed in the following subsections.
19.3.2.1. Impaction
The density and, therefore, the momentum of dust particles are greater than that of a comparable
volume of air. At each bend of the tortuous paths followed by air as it pulses in and out of the
lungs, dust particles will tend to follow a straight line and impact into the mucous coated walls of
the air passages.
The effectiveness of deposition by impaction increases with the acuteness of the bend and the
velocity of the air. There appears to be considerable anatomical variation between individuals in
the tortuosity of the branching bronchioles. Velocities are greater in the upper air passages.
However, the latter also have a larger diameter and particles must travel further across the
streamlines before impact occurs. During exertion, breathing rates and, hence, air velocities
increase throughout the system. The subject takes air deeper into the lungs and, furthermore, the
resistance of the nostrils and hair filter will promote breathing through the mouth. For all of these
reasons, heavy physical labour will increase deposition by impaction in the respiratory system.
Constriction of air passages by thickening of the mucous layer, bronchial infections or lung
damage will also result in higher air velocities and, hence, increased deposition by impaction.
19.3.2.2. Sedimentation
This term refers to the gravitational settlement of particles and is effective at low air velocities for
dust particles greater than 0.5 microns. Smaller particles become subject to Brownian motion and
diffusion effects. Sedimentation assists in the deposition of larger particles in the nasopharynx
during the reversal points of the breathing cycle. More importantly, however, sedimentation is an
effective mechanism of deposition in the low velocity laminar flows within the finer bronchioles
and the alveoli.
Another factor is that the full capacity of the lungs is seldom used. During normal quiescent
breathing the tidal volume may utilize some 65 to 75 percent of lung capacity. Sedimentation of
dust particles will occur in the near stagnant air of the unused dead-space. A phase of heavy
breathing followed by a quiescent period will first draw dust particles into the deeper recesses of
the lung and then encourage deposition by sedimentation in the dead-space as breathing
becomes more shallow.
19.3.2.3. Diffusion
Submicron particles are subject to random displacement by bombardment from gas molecules
(Brownian motion). The effect increases as the size of the particles decreases and becomes
significant for particle diameters of less than 0.5 microns. Although Brownian motion occurs
throughout the respiratory system, it becomes an effective mode of dust deposition only when the
mean displacement becomes comparable with the size of the air passage. Hence, it is particularly
important in the alveoli and finer bronchioles.
19.3.2.4. Interception
Interception becomes significant for fibrous particles. A dust fibre is often defined as a particle
where the length to diameter ratio exceeds 3. Such particles tend to align themselves with the
direction of airflow and fibres 200 microns long can penetrate deeply into the lung. Nevertheless,
the ends of fibres are likely to contact the walls of air passages, particularly at bends and
bifurcations, and accumulations of fibres can occur at these locations. This is the mechanism of
19 -
interception. The effect is accentuated by curly fibres such as those of chrysotile asbestos while
the straight fibres of amphibole asbestos have a greater probability of penetrating to the alveoli.
19.3.2.5. Electrostatic precipitation
Within the working areas of a mine, newly produced particles of mineral dust may carry a
substantial electrostatic charge. The moving electromagnetic fields that surround such particles
can induce charges of opposite sign on the walls of air passages in the respiratory system. This
results in the electrostatic precipitation of particles on to the walls and capture by the film of
mucous.
19.3.3. Dust diseases
Pneumonoconiosis, now more usually shortened to pneumoconiosis, is a generic term for
damage to cardio-respiratory organs caused by the inhalation of dust. More specific names have
been given to illnesses caused by particular types of dust. Some of those are discussed in this
Section and are often known as the dust diseases.
While the toxic and carcinogenic properties of some dusts cause specific physiological reactions
(Section 19.2.2.), most dusts that occur in mines are not in themselves fatal. However, through
progressive incapacitance of the lungs, they render the victim more susceptible to respiratory
ailments such as colds or influenza as well as tuberculosis, chronic bronchitis and emphysema
which may result in hastened death. Chronic bronchitis is an inflammation of the linings of the
respiratory system and is accompanied by periodic or constant coughing. Emphysema involves
the rupture of inter-alveolar walls caused by excessive pressure within the alveoli. This is often a
result of constrictions in the bronchioles. The rupture frequently occurs during bouts of coughing.
Adjacent alveoli break through into each other. This leads to progressive abnormal
breathlessness, particularly on even light exertion. Respiratory difficulties may also cause
excessive strain on the heart with resultant cardiac complications
19.3.3.1. Fibrogenic effects
The fibrogenic dusts introduced in Section 19.2.2. promote the abnormal development of fibrous
tissue within the alveolar clusters. They may commence at discrete foci with fibrous tissue that
may then radiate outwards to form fibroblasts. These, in turn, can merge into nodules and
conglomerates. The permanent scarring and change in the alveolar structure can have severe
secondary effects. Gas exchange across the alveolar walls is inhibited and the loss of natural
elasticity can cause a significant reduction in tidal volume. Furthermore, the production of
macrophages is reduced allowing uncontrolled accumulations of dust particles to occur in the
alveoli. Large fibrous masses can also distort or damage blood vessels, causing functional
impairment of the heart (National Academy of Sciences, 1976).
19.3.3.2. Coal workers' pneumoconiosis (black lung)
Many dusts, including coal, produce a low biological response. However, over sufficiently long
periods of exposure a build-up of retained dust occurs in the form of soft plaques within the lung
tissue. These can be observed as small black spots on chest x-rays. Similar early diagnoses can
be made for other mineral dusts including ores of iron (siderosis), tin (stannosis) and aluminium
(aluminosis). In the case of coal, such indication of coal workers' pneumoconiosis may not be
revealed for some 10 to 15 years after initial employment in coal mines. Furthermore, the
subjects may not be aware of any incapacitance during that time. In more advanced cases, the
opacities grow in size and number until they coalesce. This is likely to be accompanied by fibrosis
and all of the consequences described in the previous subsection.
19.3.3.3. Silicosis
This is one of the most dangerous of the dust diseases and is caused by particles of free
crystalline silica (quartz, sandstones, flint) but not by the silicates in clays or fireclay. The hazard
is greatest from freshly produced dust in operations involving mining or comminution of silica
19 -
bearing rocks, or from sandblasting. It is suspected that the more severe cases of coal workers'
pneumoconiosis may be associated with quartz particles mixed with coal dust.
The early stages of the disease produce, again, local foci of dust accumulations that may be
observed on x-ray films. However, there appears to be a level of silica accumulation above which
progressive massive fibrosis occurs. A number of theories have been advanced to explain this
behaviour (Schrder, 1982). It is probable that the initial accumulations cause simple microscopic
abrasion from the hard cutting edges of the particles. The initiation of progressive fibrosis may be
the result of a toxic reaction to silicic acid or electro-chemical surface energy on newly cut and
charged particles. A further suggestion is that the breakdown of poisoned macrophages may
invoke an auto-immune reaction that produces fibrous antibody structures.
19.3.3.4. Asbestosis
Asbestos is an inorganic mineral fibre comprised mainly of silicate chains. The two common
forms of asbestos are chrysotile containing tough curly fibres, and amphibole with long, straight
and brittle fibres. Asbestos fibres are captured in the respiratory system primarily by interception
(Section 19.3.2.4.) and accumulations are most likely to occur at bends and bifurcations.
However, the aerodynamic characteristics of fibres are determined by their diameter rather than
length and long fibres may reach the alveoli.
Asbestosis is associated with fibrosis, but of a different type from that given by advanced cases of
silicosis or coal workers' pneumoconiosis. The initial plaques are more brittle and contain sharp
raised ridges that may become calcified. During further progression of the disease, fibrous bands
radiate throughout the lungs and cause a significant loss of elasticity. This reduces the tidal
volume and, together with the fibrotic loss of gas exchange area, leads to abnormal
breathlessness. The reduction in oxygen transfer causes blood pressure in the pulmonary artery
to rise putting the right ventricle of the heart under strain. Cardiac failure may follow.
Cancers of the bronchial system, lung tissue and abdominal organs have been linked with
excessive exposure to asbestos fibres. However, it is considered that it is not the silicate chains
themselves that cause these cancers but rather carcinogens that are adsorbed on the fibre
surfaces prior to inhalation.
19.3.3.5. Precautions against dust diseases
The problems of dust diseases have probably attracted more research funding than any other
environmental hazard in underground mining. Three distinct areas of research have been
undertaken in countries with major mining industries, namelymedical studies into the
development, treatment and diagnosis of dust diseases, the techniques of sampling and
measurement of airborne dust, and dust suppression and control in mines.
The primary precautions against dust diseases mirror those same three areas. Experience has
shown that although highly skilled interpretation of x-ray films is required, this remains the most
important tool in discovering the onset of a dust disease and in monitoring its progress. Personnel
who are required to work in atmospheres that contain any of the toxic, carcinogenic or fibrogenic
dusts should be given free access to chest x-ray examinations. An examination should be given
before the commencement of employment in order to identify any existing condition. Further
chest x-rays are recommended at intervals of not more than two years. Workers should be
assured that early identification of coal workers' pneumoconiosis or silicosis does not necessarily
imply physical impairment nor loss of employment but should be regarded as an indication that
reassignment to work in less dusty conditions would be prudent. Legislation may guard against
significant reductions in financial remuneration.
The second group of precautions concerns organized and routine procedures of dust sampling,
preferably by means of instruments that measure mass concentration of respirable dust. This
area is discussed further in Section 19.4. and is intended to test compliance with set threshold
limit values. In most countries the latter are mandatory values enforced by law.
19 -
The most effective means of protecting mine personnel against dust diseases is control of the
dust. This involves minimizing the production of dust, suppressing it at the source, removing
airborne dust and separating workers from dusty areas. Great improvements have been made in
these areas since the 1950's. The subject is handled in greater detail in Chapter 20.
19.3.4. Threshold limit values
National experience of mining systems, geological conditions, legislation, litigation, labour
organizations, research and social consciousness in various countries has resulted in wider
variations in recommended or mandated threshold limit values for dusts than for other airborne
contaminants in mines. The threshold limit values given in Table 19.2 are based primarily on
recommendations of the American Conference of Governmental Industrial Hygienists (ACGIH)
and are simply a guide to airborne concentrations that are currently thought to produce no
adverse effects from a daily working time exposure.
Aerosol
Aluminium oxide
Arsenic
Asbestos
Borates
decahydrate
others
Calcium carbonate
Calcium sulphate
Carbon black
Coal
TLV (TWA) mg/m3 unless

otherwise stated*
10
0.01
0.1 fibres/ml
5
1
10
10
3.5
2
10/(% quartz) for >5% quartz)
10
2.5
2
5.0
2
10 (total), 5(respirable)
3
10
Comments
carcinogen
carcinogen,
for fibres longer than 5 microns
marble, limestones
gypsum
carcinogen
respirable (<5% quartz)
Fibrous glass
Fluorides
Graphite (natural)
respirable fraction
Iron oxide fumes
(>1% quartz)
Kaolin
respirable, no asbestos, <1% silica
Magnesite (magnesium oxide)
Mica
respirable fraction
"Nuisance" dusts
non-hazardous materials
Oil mist
mineral
5
excluding vapour
vegetable
10
Perlite
10 (total), 3 (respirable)
Portland cement
10(total), 5 (respirable)
Radon daughters
4 WLM/year
see Section 13.3.3
Quartz
0.05
respirable fraction
Silicon Carbide
10
respirable, no asbestos, <1%silica
Soapstone
3
respirable fraction
Talc
2
respirable fraction
Welding fumes
5
Table 19.2. Guideline threshold limit values for selected aerosols given in milligrams per cubic
metre of air (based primarily on recommendations of the American Conference of
Governmental Industrial Hygienists). Considerable variations occur in the laws of
other countries.
See Section 11.2.1. for definitions of Threshold Limit Value and Time Weighted Average.
19 -
The comments made in Section 11.2.1. on threshold limit values for gases apply equally well
here. In particular, subsurface ventilation engineers must become familiar with limiting
concentrations mandated by their own national or state laws.
There is a further reservation that applies to the use of threshold limit values for dust
concentrations. Despite the move from particle count to gravimetric methods of measurement,
there remain considerable variations in the results given by differing instruments (e.g. Phillips,
1984). These arise from differences in the design of instruments, the particle size distribution
curves they are intended to follow and the efficiency of dust capture. Even a single instrument
can indicate a different result if the rate of sample airflow through the device is altered. For these
reasons, individual countries or mining industries have tended to "adopt" a particular instrument
as a standard and to employ empirical conversion factors in order to compare data obtained by
other devices (Rogers, 1991). It follows that while such variations in instrument performance
exist, the application of dust threshold limit values for purposes of legislation or enforcement
should be referred to a particular instrument used according to a specified procedure.
Another factor should be borne in mind when comparing threshold limit values imposed by
differing legislative authorities. Laws governing dust concentrations usually specify not only the
limiting concentrations of respirable dust but also the sampling locations. Hence, a measurement
required to be taken in a return airway should not be compared to one that is obtained at the
position of a machine operator.
For coal mines, legislation may require reduced threshold limit values in certain areas. For
example, the 2 mg/m3 TLV for respirable dust in American coal mines is reduced to 1 mg/m3 for
intake airways within 61 m (200 ft) of a working face. Further restrictions are applied when quartz
particles are present in the coal dust. One method is to reduce the TLV to 10 divided by the
quartz percentage when the quartz content in the dust exceeds 5 percent. This is based on a
quartz TLV of 0.1 mg/m3.
19.4. THE ASSESSMENT OF AIRBORNE DUST CONCENTRATIONS

19.4.1. Background
The accurate assessment of dust concentrations in relation to the health of personnel in mines is
beset with difficulties. First, the fact that physiological consequences develop very slowly is
compounded by dust concentrations varying across wide limits with respect to both time and
place in a mine. Readings of dust concentrations measured over a short time interval have very
little relevance to the long term health of the workforce. It was this difficulty that masked
correlations between most particle count data and incidence of pneumoconiosis until the 1959
Johannesburg Conference. Secondly, as indicated in the previous Section, it is not only the
concentration of particles that matters but also their size distribution and mineralogical
composition. It is not surprising, therefore, that even modern instruments may vary quite
significantly in the quality and type of data they produce.
In choosing dust instrumentation it is important to define the primary purpose of the intended
measurements. These may be part of long term investigations with the aim of establishing
environmental standards. Other research surveys may be aimed at the spatial and temporal
variations of dust concentration to which specific groups of personnel are exposed, or to
investigate the effects on airborne dust of particular mining equipment or dust suppression
techniques. Routine measurements in a mine are made to protect the health of the workforce and
to check or ensure compliance with mandatory standards.
The earliest attempts to measure airborne dust quantitatively are reported to have occurred in the
eighteenth century using observations of dust deposition on a polished surface (Walli, 1982).
19 - 10
However, the major developments in this field have taken place since the introduction of the
konimeter, a hand held jet-impact device invented by Kotz, a South African government mining
engineer, in 1916. Modern versions of the konimeter remain in use in that country although its
remaining life must be very limited.
Since the Johannesburg Conference of 1959, the vast majority of new instruments measure the
mass concentration of respirable dust (<5 microns). Most of these instruments pass the dust in a
continuous air sample through combinations of elutriators (settling chambers), cyclones, or jetimpactors and filters. Furthermore, in order to meet the need for longer term readings, most
gravimetric instruments have been designed to operate over several hours. Such data are
necessary from the viewpoint of health protection but mitigate against rapid readout, continuous
monitoring and short term control. Alternative methods of dust assessment combined with
modern electronic integrating circuitry allow short term and long term data collection to proceed
simultaneously. Many such instruments have utilized the principle of light beam scattering by dust
particles (the Tyndall effect).
A recognition of the differing objectives of health protection and more immediate concerns such
as dust explosibility, visibility and nuisance effects have led to two distinct applications of dust
monitors. First, lightweight personal samplers are increasingly being worn by underground
workers. These give a record of the respirable dust concentrations to which the person has been
exposed, on a shift by shift basis. Secondly, heavier and more sophisticated instruments may be
set up to measure dust concentrations either for several hours or, on a more permanent basis, to
allow continuous monitoring for local indication and recording or transmission to a central control
station.
Modern trends in dust instrumentation are toward:
an increase in the use of personal samplers,
cascade devices or other means of selection to indicate mass concentrations in each of a
number of particle size ranges,
an immediate indication of mineral content (particularly quartz) in addition to mass
concentration and
an increase in light scattering methods for continuous monitoring.
In the following subsections, we shall introduce the main principles employed in instruments for
the assessment of airborne dust. However, because of the rapid evolution and number of new
devices, we shall not attempt detailed descriptions of individual monitors. For these, the reader is
referred to trade journals and manufacturers' literature.
19.4.2. Particle count methods
Although particle count methods were used for many years and models were produced in a
number of countries, they are now nearing obsolescence for use in mines. There were essentially
two families of particle count instruments. One of these relied on the impact principle in which a
short-lived but high velocity pulse of air was induced through a jet and directed at a receiving
surface of treated glass, film or liquid. The most enduring of this type of instrument is the
konimeter (Section 19.4.1.).
The second type of particle count instruments was the thermal precipitator. Molecular
bombardment from a heated wire diverted dust particles from a moving sample stream on to an
adjacent glass slide (Section 20.2.4.). The slides, films or liquids produced by particle count
instruments were subjected to microscope analysis for counting the number of particles in each
size range. For many years this was carried out manually and was somewhat subject to individual
bias. Latterly, it has been conducted by "light assessors" that really measure surface area
(Martinson, 1982). Computer controlled particle analyzers are now available that can perform
such a task much more rapidly and efficiently.
19 - 11
19.4.3. Gravimetric methods

Figure 19.2 illustrates the operation of one of the earlier and very successful gravimetric
samplers, the MRE, developed at the Mining Research Establishment of the then National Coal
Board in the United Kingdom (Dunmore et al, 1964). This instrument continues to be a standard
in several countries. Air passes through the instrument at a rate of 2.5 litres/min under the action
of a diaphragm pump. The air velocity between the parallel plates of the elutriator drops to the
extent that particles larger than 7 microns settle out. The remaining finer particles pass on to be
collected on a 5 micron membrane filter. The filter is weighed before and after a sampling period
which may, typically, be eight hours.
elutriator
filter
paper
motor
counter
pulsation
damper
flow
meter
diaphragm
pump
Figure 19.2 Schematic of an MRE Gravimetric Dust Sampler

The specifications of the elutriator are such that it gives a particle removal characteristic which
resembles the BMRC curve on Figure 19.3. This curve was defined by the British Medical
Research Council as representing the penetration rates of respirable dust to the alveoli of the
human lungs. An alternative model adopted by the American Conference of Governmental
Industrial Hygienists is illustrated for comparison on the same figure. The MRE instrument
achieved the rare distinction of being specified, by name, as a standard in American mine
legislation.
Many other gravimetric samplers have been produced since the 1960's. Some have replaced the
elutriator with a nylon or metal cyclone as illustrated on Figure 19.4. The small dimensions and
light weight of these cyclones make them particularly suitable for personal samplers.
Other types and sizes of filters are also employed. While the filter should be efficient in trapping
respirable dust, it should not restrict the airflow to the extent that it inhibits the required constant
airflow throughout the sampling period. Silver metal filters have been shown to be useful if the
sample is subsequently to be heated in order to determine combustible content or if x-ray
19 - 12
diffraction is used to measure the amount of quartz present (Knight and Cochrane, 1975). The
maximum period of sampling through any given filter is reduced because of premature blockage
in the presence of diesel fumes or mists of water or oil (Gardiner, 1988).
100
90
penetration (per cent)
80
BMRC
70
60
50
40
ACGIH
30
20
10
0
0
10
Particle aerodynamic diameter (microns)

Figure 19.3 Lung penetration curves for respirable dust as defined by the
British Medical Research Council (BMRC) and the American
Conference of Governmental Industrial Hygenists (ACGIH).
19.4.4. Photometric (light-scattering) methods

It was the English physicist John Tyndall (1820-'93) who first explained the scattering of sunlight
by dust particles in the atmosphere. The blue component of white light is scattered most which
explains why the sky appears blue. The same effect is utilized in the photometric methods of
measuring dust concentration. The corresponding instrumental techniques have been developed,
particularly in Germany, although devices using the same principle are now also produced in the
United Kingdom and the United States of America.
The early Tyndalloscope split the beam from a white light source. Half proceeded through
polarizing filters to the eyepiece. The other half was diverted through a sample chamber where
light reflected by the dust particles was collected and directed to the eyepiece. The filters were
rotated until the two half beams that were visible simultaneously in the eyepiece appeared to
have the same intensity. The angle of filter rotation was employed as an indication of the dust
concentration. Modern Tyndallometers use photosensors to detect the deflected light and
produce an electrical output for visual display, recording or transmission. Preferred angles of
forward light scatter can be chosen to minimize the effects of particle shape or index of refraction.
19 - 13
to pump
filter
fine dust
air
There are two techniques of particle size

discrimination used in these devices. Some
photometric dust instruments eliminate the nonrespirable particles by passing the sample
through a pre-classifier comprised of an elutriator
or a cyclone. A number of cyclones in cascade
can be used to give several size ranges. The
other method of particle size discrimination is to
use a laser or monochromatic light source. The
choice of wavelength gives a means of
discriminating in favour of the desired size range
of particles. The angle between the light source
and the photodetector is another method of
selecting a preferred size range (Breuer et al,
1973; Thaer, 1975).
Although photometric dust instruments really
respond to light reflected from particle surfaces,
using them for selected size ranges enables the
readings to be interpreted in terms of total
volume of those particles and, hence, mass
concentration (assuming constant density of the
particle material). Figure 19.5 illustrates an
instrument that combines a forward scattering
laser unit with an 8 hour filtration system. This
permits direct calibration of the light unit with
respect to mass concentration of respirable dust.
coarse dust
Figure 19.4 Cyclone head of a personal dust filter.
The major advantage of modern light scattering

instruments is that the addition of electronic
circuitry permits a combination of immediate
readout and integration over any chosen time
interval. Hence, they can be employed for both
short term and long term sampling. Coupled with
cascade (sequential) interrogation of a sample
stream they can indicate, record and transmit
information for each of a series of particle size
ranges (Breuer, 1975; Oberhalzer, 1987).
handle
exhaust
memory unit
elutriator
monitor
detector
laser diode
signal
detector
battery
light trap
filter
pump
Figure 19.5 Schematic of a SIMSLIN II dust sampler.

19 - 14
19.4.5. Personal samplers

Stationary gravimetric dust monitors provide a means of sampling over an extended time period
(typically an eight hour shift) and are useful for assessing average dust concentrations at
specified locations. However, they do not measure the actual dust inhaled by individual workers.
Neither are they capable of showing temporal variations of dust concentrations to which the
worker is subjected. For many years attempts were made to develop practical and wearable dust
monitors. Personal monitors should be unobstrusive to the worker and capable of giving reliable
indications of dust level either instantaneously or over short time intervals. Furthermore, they
should also be capable of recording full shift average dust concentrations for subsequent entry
into a data log for the individual worker.
Early versions of personal dust monitors included a small lightweight cyclone (Figure 19.4) that
was pinned to the lapel of the workers clothing. The unit was powered by a battery worn on the
belt. The position of the cyclone inlet could result in non-representative readings caused by dust
from clothing. The device was also incapable of giving short-term readings.
Figure 19.6. The NIOSH 2004 tapered-element oscillating microbalance personal dust sampler.
Reproduced from Volkwein, J.C. et al (2004). Performance of a new personal
respirable dust monitor for mine use. Pittsburgh Research Laboratory, Report of
Investigations 9663. NIOSH Publication No. 2004-151.
A determined multi-year effort funded by the U.S. National Institute for Occupational Safety and
Health (NIOSH)1 produced a personal dust monitor that overcame many of the drawbacks of
earlier models (Volkwein et al (2004), Volkwein et al (2005)). The heart of this monitor is a
tapered-element oscillating microbalance. Air is drawn into the unit via an inlet located close to
the lens of the wearers caplamp, i.e. representative of the air being inhaled. A continuous sample
passes through a 48mm diameter tube alongside the caplamp cord down to the sampler which is
incorporated into a casing that also holds the caplamp battery (Figure 19.6).
1
Development work was conducted by Rupprecht and Patashnick Co., Inc of Albany, New York.
19 - 15
Within the sampler the air passes through a cyclone to separate out the respirable fraction of
dust. The latter passes along the hollow tapered element and through an exchangeable filter
which is located on the narrow end of the element. Electronic components are employed to
oscillate the element and filter at resonant frequency. As the mass of dust collected by the filter
increases, the resonant frequency decreases. From that change the monitor evaluates the mass
captured by the filter over a given time period. The readings given by a screen display on top of
the beltworn unit include the average respirable dust concentration since the start of the shift and
a running average over the past 15 or 30 minutes. Data from the sampler can be downloaded into
a personal computer after the end of a shift.
19.4.6. Other methods of airborne dust measurement
In addition to the principles of dust measuring units outlined in the preceding subsections, there
are several other types of devices that are available or under development. The impact principles
employed in some of the old particle count devices still have a role as classifiers to remove nonrespirable dust. This may further reduce the weight of personal samplers. A series of jet orifice
sizes can be used to simulate a lung penetration curve (National Research Council, 1980).
Another type of device utilizes a carbon-14 radioactive source to pass beta rays through a mass
of dust particles on a filter or impact collection plate. A detector measures the attenuation of the
beta rays. In use, the device is run for a predetermined time. The difference between the flux of
beta rays before and after sampling is processed by internal circuitry to indicate the mass of
particles collected (National Research Council, 1980). A similar device developed in Poland
employs strontium-90 as the radioactive source and measures reflected back-scatter of beta rays
from the surface of the sample (Krzystolik et al, 1985).
A technique that may find further application is to measure the change in resonant frequency of a
piezo-electric quartz crystal as a mass of sampled dust collects on it. Here again, such units
arranged in cascade with interstage cyclones or impactors can be used to indicate a number of
particle size ranges (Sem et al, 1977).
19.4.7. Discrimination of quartz particles
Mine dust is seldom composed of a single mineral. As quartz particles occur commonly and are
particularly dangerous (Section 19.3.3.3.), efforts have been directed at quantifying the quartz
content of dust samples. This is of concern not only in hardrock mines but also because of the
increased use of mechanized extraction and roof bolting in coal mines.
The employment of x-ray diffraction gives well defined peaks for quartz on the output spectrum.
This method has been used for a number of years as a means of mineralogical analysis of mine
dust (Knight and Cochrane, 1975; Bradley, 1975). A further development is to incorporate the
principle into dust monitoring equipment in order to display the quartz content concurrently with
respirable dust concentration.
19.4.8. Sampling strategy
Throughout the history of organized dust measurements in mines, investigators have been faced
with problems of large variations in the observed results. Furthermore, the significance of these
results has often been subject to debate and interpretation. The very real variations in the
concentration and mineralogical content of dusts that exist with respect to time and location in
any mine are compounded by the differing efficiencies with which alternative instruments simulate
the dust retention characteristics of the human lung. These difficulties have led to coining of the
phrase of sampling strategy which really means the "why, where, when and how" dust samples
are taken.
19 - 16
The "why", i.e. the objectives of dust sampling, were outlined in Section 19.4.1. It is, indeed,
important to have a clearly defined purpose for any given set of dust measurements, particularly
where the setting or checking of mandatory standards is involved. The objectives of a dust survey
will then usually influence the "where, when and how."
When checking compliance with threshold limit values, the locations of area measuring points are
normally dictated by law - for example at specific points in intake and return airways or with
respect to the positions of machine operators. Research into dust minimization techniques in
various countries has been greatly influenced by the wording of the relevant national law. For
example, spray fan or other air diversion techniques can produce significant reductions in dust
concentrations at the operator position of a continuous miner but may have little or no effect on
respirable dust counts in the return airway. If, on the other hand, the purpose of a measurement
is to check the dust production of a particular piece of equipment or operation, then the
instruments should be located at specified distances upwind and downwind from that equipment
or operation.
The times and durations of area sampling will, again, usually be specified by law, often at
intervals of two or three months at each mandatory sampling point. The samples may be required
to be submitted for analysis to specified laboratories operated or authorized by government
agencies. Furthermore, check samples may be taken by government inspectors.
For short term sampling, it is all too easy to introduce conscious or inadvertent bias into dust
measurements. The maximum dust make from a machine will be obtained when the equipment is
running at full load. However, the measured concentration of non- toxic dust should not
necessarily be interpreted as an epidemiological (health) hazard. It becomes so only if personnel
are working in or downstream from the dust make and are exposed to concentrations that exceed
the set standards on the basis of an 8 hour time weighted average.
The importance of adhering to agreed types of instruments and sampling procedures was
emphasized in Section 19.3.4., particularly when checking compliance with statutory
requirements. Indeed, the instruments and measurement procedures may be specified within the
regulations. A factor that is often overlooked is the effect of air velocity on the sampling ports of
the instrument. At the higher air velocities, dust particles can be diverted around the sampling
ports (Baskhar and Ramani, 1987). Isokinetic devices can be fitted to the instrumentation in order
to match the inlet port sampling velocity with that of the approaching airstream.
The variability of measured dust concentrations coupled with the intrinsic uncertainties of
instrumentation and sampling procedures should dictate that those measurements be subjected
to statistical analysis. In practice this is often relegated to a rejection of "unrepresentative or
suspect" samples and the straight averaging of the remainder (National Academy of Sciences,
1976; Martinson, 1982). Such a procedure is clearly open to bias and is not to be recommended.
Where any given sampling practice or location consistently yields a significant number of suspect
samples, then an investigation should be carried out in order to determine whether the
unexpected variations are, indeed, real and, if not, the weaknesses of the sampling technique.
Statistical examinations can be carried out to test whether the number of samples is sufficiently
large to be representative of the particular type of work, time or location. The data produced by
wider use of personal samplers, together with computer based records and statistical analyses,
facilitate such tasks.
This returns us to the utilization of personal samplers. The question of sampling strategy became
of particular importance with the advent of 8 hour gravimetric sampling units and the relationship
between the results given by these instruments and the earlier, short term, particle count units.
Personal samplers have revolutionized the philosophy of dust sampling strategy. Provided that a
correlation has been established between the results given by specified types of personal
sampler and corresponding levels of epidemiological hazard, then there appears to be little
19 - 17
reason why those types of personal sampler should not be used for both individual exposure
monitoring and compliance with appropriately worded regulations. This separates sampling for
health reasons from those measurements that may be taken for purposes of control or planning.
The question arises on how many underground personnel should be asked to wear personal
samplers. This would appear to depend upon the degree of perceived exposure to dust and the
type of mining. In heavily mechanized workings, there may be relatively few persons within the
more dusty areas. Furthermore, the mobility and differing occupations of some workers may
subject them to wide variations of dust exposure. In such circumstances, it would be in the
interests of those workers to wear personal samplers. The initiation of the practice should be
accompanied by suitable explanations of the benefits of the devices, how they should be treated,
and the manner in which the data will be used.
In labour intensive mines with much larger numbers of personnel, many of them engaged in
similar occupations, it may be unnecessary to ask all persons to wear personal samplers. A
minimum number that is deemed to be representative of each work group should be decided,
including the more mobile supervisory staff (Quilliam, 1975). The actual persons wearing the
samplers may be changed on a rota basis. However, after each shift the exposure record of each
individual should be updated according to the relevant representative value for that group and
location. Provided that the instruments are compatible, the data obtained can be used to compare
dust indices for differing occupations, locations and mines (van Sittert, 1988).
The modern approach is to combine full-shift personal samplers with both long and short term
dust monitoring that is incorporated into mine environmental electronic surveillance systems. The
data obtained from such means will help to safeguard the health of workers and be valuable in
the planning and design of future mining operations.
References
Baskhar, R. and Ramani, R.V. (1987). A comparison of the performance of impactors and
gravimetric dust samplers in mine airflow conditions. 3rd U.S. Mine Ventilation Symposium, Penn
State, pp 502-508.
Bradley, A.A. (1975). The determination of the quartz content of gravimetric mass samples of
airborne dust by an x-ray technique. 1st Int. Mine Ventilation Congress, Johannesburg, South
Africa, pp 455-458.
Breuer, H. (1975). TBF 50 and Tyndallometer TM digital - two instruments supplementing each
other for the occupational hygienic and technical assessment of dust conditions. 1st Int. Mine
Ventilation Congress, Johannesburg, South Africa, pp 445-452.
Breuer, H., Gebhart, J. and Robock, K. (1973). Photoelectric measuring apparatus for
determination of the fine dust concentration. Staub Reinhaltung der Luft (in English), Vol 33.
Brundelet, P.J. (1965). Experimental study of the dust-clearance mechanism of the lung. Acta
Pathol. Microbiol. Scand. Suppl. 175, pp 1-141.
Carver, J. (1975). Respirable dust regulations in the United Kingdom. 1st Int. Mine Ventilation
Congress, Johannesburg, South Africa, pp 399-405.
Dunmore, J.H., Hamilton, R.J. and Smith, D.S.G. (1964). An instrument for the sampling of
respirable dust for subsequent gravimetric assessment. Journ. Sc. Instruments, Vol. 41, pp 69672.
19 - 18
Gardiner, L.R. (1988). Personal gravimetric dust sampling for the South African gold mining
industry. 4th International Mine Ventilation Congress, Brisbane, Australia, pp 507-515.
Gibson, H. and Vincent, J.H. (1980). An investigation of "fly-dust" nuisance in mines. 2nd Int.
Mine Ventilation Congress, Reno, USA, pp 620-622.
Knight, G. and Cochrane, T.S. (1975). Gravimetric dust sampling with quartz analysis and its
use in metal and mineral mines. 1st Int. Mine Ventilation Congress, Johannesburg, S. Africa, pp
407-414.
Krzystolik, P.A. et al (1985). Portable coal dust/stone dust analyzer. 2nd U.S. Mine Ventilation
Symposium, Reno, USA, pp 171-179.
Martinson, M.J. (1982). Sampling pathogenic airborne particulates. Chap. 14, Environmental
engineering in South African mines. Mine Ventilation Society of South Africa, pp 357-378.
National Academy of Sciences (1976). Mineral resources and the environment: Coal workers'
pneumoconiosis - medical considerations, some social implications. Washington, D.C.
National Research Council (1980). Measurement and control of respirable dust in mines. U.S.
National Academy of Sciences, Washington, D.C., NMAB-363.
Oberhalzer, J.W. (1987). Assessment of colliery dust levels using a computerized dust
measuring system. 3rd U.S. Mine Ventilation Symp. Penn State, pp 617-624.
Orenstein, A.J. (editor) (1959). Proc. of the International Pneumoconiosis Conference,
Johannesburg, South Africa. Published by Churchill, London, 1960, pp 632.
Phillips, H.R. (1984). New methods and standards for respirable dust monitoring in the New
South Wales coal industry. 3rd Int. Mine Ventilation Congress, Harrogate, U.K., pp 203-208.
Quilliam, J.H. (1975). A review of dust sampling techniques in South African gold mines. 1st Int.
Mine Ventilation Congress, Johannesburg, S. Africa, pp 419-421.
Rogers, W.G. (1991). Silica; Crystalline quartz exposure standards... What is the most
appropriate standard to use? Journ. of the Mine Ventilation Society of South Africa, March, pp 3843.
Schrder, H.H.E. (1982). The properties and effects of dust. Chapter 12, Environmental
Engineering in South African Mines, Mine Ventilation Soc. of S. Africa, pp 313-336.
Sem, G.J., Tsurubayashi, K. and Homma, K. (1977). Performance of the piezo-electric
microbalance respirable aerosol monitor. Am. Ind. Hyg. Assoc. Jour., Vol. 38, p 580.
Thaer, A. (1975). The Leitz Tyndallometer TM digital as applied to the determination of fine dust
concentration under the auspices of Occupational and Environmental Medicine. Presented at the
Kungl Arbetarskyddsstyrelsen Seminar, Stockholm, April 23.
Van Sittert, J.M.O. (1988). An overview of the proposed guidelines to standardize the respirable
dust sampling strategy for risk determination in collieries. 4th Int. Mine Ventilation Congress,
Brisbane, Australia, pp 527-531.
Volkwein, J.C. et al (2004). Performance of a new personal respirable dust monitor for mine use.
Pittsburgh Research Laboratory, Report of Investigations 9663. NIOSH Publication No. 2004-151.
19 - 19
Volkwein, J.C. et al (2005). Implementing a new personal dust monitor as an engineering tool.
International Occupational Hygiene Association, 6th International Scientific Conference,
Pilanesberg, South Africa.
Walli, R.A. (1982). Mine dusts. Mine ventilation and air conditioning (ed. Hartman), Chap. 5, pp
84-130.
19 - 20
The aerodynamics, sources and control of airborne dust
CHAPTER 20.
THE AERODYNAMICS, SOURCES AND

CONTROL OF AIRBORNE DUST
20.1. INTRODUCTION ................................................................................................... 2

20.2. THE AERODYNAMIC BEHAVIOUR OF DUST PARTICLES................................. 2
20.2.1. Gravitational settlement ................................................................................................................ 2
20.2.1.1. Stokes' Law and terminal velocities. ...................................................................................... 3
20.2.1.2. Slip flow .................................................................................................................................. 5
20.2.2 Brownian motion ............................................................................................................................ 9
20.2.2.1. Brownian displacements ........................................................................................................ 9
20.2.2.2. Brownian diffusivity .............................................................................................................. 10
20.2.3. Eddy diffusion ............................................................................................................................. 11
20.2.4. Other forms of dust transportation .............................................................................................. 12
20.2.5. Coagulation................................................................................................................................. 13
20.2.6. Impingement and re-entrainment................................................................................................ 15
20.2.7. Computer models of dust transport ............................................................................................ 16
20.3. THE PRODUCTION OF DUST IN UNDERGROUND OPENINGS ...................... 16

20.3.1. The comminution process........................................................................................................... 16
20.3.2. Mechanised mining ..................................................................................................................... 19
20.3.3. Supports...................................................................................................................................... 20
20.3.4. Blasting ....................................................................................................................................... 20
20.3.5. Loading operations ..................................................................................................................... 21
20.3.6. Transportation and crushing ....................................................................................................... 21
20.3.7. Workshops .................................................................................................................................. 22
20.3.8. Quartz dust in coal mines ........................................................................................................... 22
20.4. CONTROL OF DUST IN MINES.......................................................................... 23

20.4.1. Dust suppression ........................................................................................................................ 23
20.4.1.1. Pick face flushing and jet-assisted cutting ........................................................................... 23
20.4.1.2. Water infusion ...................................................................................................................... 24
20.4.1.3. Wetting agents, foams and roadway consolidation.............................................................. 25
20.4.2. Removal of dust from air............................................................................................................. 25
20.4.2.1. Water sprays ........................................................................................................................ 26
20.4.2.2. Wet scrubbers ...................................................................................................................... 30
20.4.2.3. Dry filters and separators ..................................................................................................... 33
20.4.2.4. Personal respirators ............................................................................................................. 37
20.4.3. Dilution and layout of the ventilation system .............................................................................. 37
20.4.4. Separation of personnel and dust............................................................................................... 38
References .................................................................................................................. 38
20 -
20.1. INTRODUCTION
The physical characteristics of aerosols have been subjected to intensive study for the free
surface atmosphere. This is an important area in meteorology and investigations of the behaviour
of contaminant plumes in the atmosphere. Somewhat less attention has been paid to the
aerodynamic characteristics of dust when the carrying airstream is confined within the boundaries
of ducts or tunnels.
The first main section in this chapter outlines the several phenomena that govern the manner in
which airborne dust is transported through the branches of a ventilation network and the
deposition of dust particles on the roof, floor and sides of mine airways.
A prerequisite to the successful control of airborne dust in a mine is an understanding of the
potential sources of the dust. These are discussed in the second main part of the chapter. While
some sources are obvious such as a power loader or tunneling machine, others are less so
including the crushing of immediate roof strata by modern powered roof supports. The final
section outlines the methods of dust control in mining operations. These include prevention of the
formation of dust, suppression and removal of dust particles from the air, isolating personnel from
concentrations of dust and the diluting effects of airflow. The latter was introduced in Section
9.3.3.
Readers who are interested only in the practical aspects of the topic are advised to
concentrate on Sections 20.3.2 to 20.4.1 and 20.4.2.2 to 20.4.4.
20.2. THE AERODYNAMIC BEHAVIOUR OF DUST PARTICLES

The very large size range of dust particles that exist in the ventilation system of an active mine
results in a variety of differing phenomena influencing the behaviour of the particles. The smallest
particles act almost as a gas and react to molecular forces while the larger particles are
influenced primarily by inertial and gravitational effects. In this section we shall consider the
influence of gravitational settlement, molecular diffusion, turbulent or eddy diffusion, coagulation,
impingement, re-entrainment and computer simulations.
20.2.1. Gravitational settlement
The rate at which a particle falls through air under the action of gravity depends not only upon the
size and density of the particle but also its shape. In Section 19.2.1., the concept of an equivalent
geometric diameter based on projected area was introduced. This is the diameter of a sphere that
has the same projected area as the actual particle.
The majority of analyses in this subject assume that each particle is a homogeneous sphere. In
the study of particle aerodynamics this has given rise to further alternative definitions of
equivalent diameter including:
Stokes' diameter:
the diameter of a sphere that has same density as the actual particle
and falls through air at the same rate
aerodynamic diameter: the diameter of a sphere of density 1 g/cm3 that falls through air at
the same rate as the actual particle.
Despite these additional definitions, the geometric diameter remains the one that is most
commonly used in practice.
20 -
weight
drag
1
s d 3 g
6
3 d u
upthrust
20.2.1.1. Stokes' Law and terminal velocities.

When any body is suspended in a fluid, at
least two forces act upon it (Figure 20.1).
One is the weight of the body within the
prevailing gravitational field. The volume of a
1
sphere of diameter d is d 3 m3. If this has
6
a density of s (kg/m3 ) then its weight
becomes
1
s d 3 g
N
(20.1)
6
where g = gravitational acceleration (m/s2 )
1
a d3 g
6
Figure 20.1 Forces on a particle falling in air.
1
d 3 g ( s a ) = 3 d u
At equilibrium
6
However, the sphere displaces its own

volume of fluid and will experience an
opposing upthrust equal to the weight of
fluid displaced, i.e.
1
a d 3 g
6
where a =
(20.2)
density of the fluid (kg/m3)
The net force causing downward movement is the combination of the two:
1
d 3 g ( s a )
6
(20.3)
If the particle is moving relative to the fluid then it will experience a further resistance or drag
because of viscous shear and conversion of some of its kinetic energy into turbulent eddies within
the fluid. A general expression for drag was given in Section 5.4.6.2. as
Drag = C D Ab a
where CD
u
and
Ab
u2
2
(20.4)
= coefficient of drag (dimensionless)

= relative velocity between the particle and the fluid, m/s
= projected area (= d 2/4), m2
Many investigators have investigated relationships between CD and Reynolds' Number, Re, for
fully submerged bodies (e.g. Prandtl 1923). In the case of spheres, the diameter is used as the
characteristic length in the calculation of Reynolds' Number. For the particular case of laminar
flow around a particle, Sir George G. Stokes (1819-1903), the Cambridge physicist, showed that
CD =
24
Re
(20.5)
20 -
Now, as Re =
a u d
a
CD =
24 a
a u d
Substituting for Ab and CD

Drag =
where a = dynamic viscosity of the fluid (Ns/m2 ),
equation (20.4) gives
24 a d 2
u2
a
a u d 4
2
= 3 a d u
(20.6)
As the particle accelerates downwards, its velocity, u, increases until the drag equals the
downward force quantified in equation (20.3
1
d 3 g ( s a ) = 3 a d u
6
(20.7)
At that point of dynamic equilibrium, the velocity of fall becomes constant and is renamed the
terminal velocity, ut. Equation (20.7) may now be rearranged as
ut =
d 2 g ( s a )
18 a
m/s
(20.8)
Equations (20.6 to 20.8) have all been referred to as Stokes' Law.

Stokes' Law applies with good accuracy to particles that are above the respirable range (5
microns). Smaller particles become sensitive to slippage and molecular forces. Stokes' Law is
based on the assumption of laminar flow. If the terminal velocity is sufficiently high to cause the
onset of a turbulent wake then the transfer of kinetic energy from the particle to the fluid (inertial
effects) can no longer be ignored. The upper limit of Stokes' Law occurs at a Reynolds Number,
Re, of about 0.1 which, for many mineral particles falling at their terminal velocity through air, is
equivalent to geometric diameters of approximately 20 microns.
For larger particles at their terminal velocity, ut , we may balance equations (20.3) and 20.4):
2
u
d2
1
d 3 g ( s a ) = C D
a t
6
4
2
giving
ut
4 d g (s a )
CD a
3
m/s
(20.9)
For dust particles in air, s >> a and the term (s a) is usually truncated to s. Flagan and
Seinfeld (1988) suggest the approximations for coefficients of drag, CD given in Table 20.1.
20 -
Reynolds No., Re
CD
< 0.1
24
(Stokes Law)
Re
0.1 < Re < 2
24
3
9
Re +
Re 2 ln(2 Re)
1 +
Re 16
160
2 < Re < 500
24
1 + 0.15 Re 0.687
Re
500 < Re < (2 x 105)
0.44
Table 20.1 Approximations for coefficients of drag for spherical particles (after Flagan
and Senfeld, 1988)
20.2.1.2. Slip flow
Stokes' Law applies to dust particles that are large in comparison to the mean free path of the
gas molecules. Hence, those particles see the gas as a continuum. As the particle size
approaches the mean free path of the gas molecules this no longer holds. Two effects are then
observable; first the jerky dislocations caused by molecular bombardment, known as Brownian
motion and discussed in Section 20.2.2., and secondly, the drag force reduces as the small
particle becomes more able to move or "slip" through intermolecular voids.
In order to quantify the very small distances now being considered, let us recall that the mean
free path of a gas molecule is defined as the average distance it moves between collisions with
other gas molecules. Although air is a mixture of gases, it is convenient to treat it as a single gas
of equivalent molecular weight 28.966 and gas constant 287.04 J/kgK.
From the kinetic theory of gases it can be shown that the mean free path, , is given by
=
where
P
R
and
T
1
2
(20.10)
0.499 P (8 / RT )
=
dynamic viscosity of fluid (Ns/m2 )
=
pressure (N/m2 )
=
gas constant (J/kgK)
=
absolute temperature (K)
For air at
P = 100 kPa,
T = 293 K (20 C),
R = 287.04 J/kgK and a = 17.9 x 10-6 Ns/m 2 (Section 2.3.3.),
17.9 10 6
0.499 10 5 {8 / ( 287.04 293 )}
1
2
= 6.52 10 8 m or 0.0652 microns.
When particle diameters fall below 5 microns, the effect of slippage becomes significant. In order
to extend the applicability of Stokes' Law, a correction factor, Cc, can be introduced to reduce the
calculated value of drag. Thus, for small particles equation (20.6) is corrected to
Drag =
3 a d u
Cc
(20.11).
20 -
A number of relationships between Cc and d have been suggested (e.g. Allen and Raabe, 1982),
based mainly on a series of classical experiments on liquid aerosols carried out by Millikan
between 1909 and 1923. Values of the slip correction factor for air at 25 C and 101 kPa are
given in Table 20.2.
d (microns)
0.01
0.05
0.1
0.5
1.0
5.0
10.0
Cc
22.7
5.06
2.91
1.337
1.168
1.034
1.017
Table 20.2. Slip correction factor for dust particles in air (after Flagan and Seinfeld, 1988).
The equation C c =
9.56 10 8
d 1.045
+ 0.99 , with d expressed in metres, gives Cc within an accuracy
of 2 percent.
Incorporating the slip correction factor into Stokes' Law for terminal velocity, equation (20.8) gives
a relationship that can now be extended down to a particle size of 0.01 microns:
ut =
d 2 g ( s a )C c
18 a
m/s
(20.12)
Figure 20.2 gives a graphical representation of this equation for particles of varying diameter and
density falling through air of temperature 20C. The curvature of the lines on this log-log plot is
due to the effects of slippage.
Example
Determine the terminal velocities and time taken for particles of geometric equivalent diameter
0.1, 1, 10 and 100 microns to fall a distance of 2m through air of density a = 1.1 kg/m3 and
dynamic viscosity, a = 18 x 10-6 Ns/m2 . The density of the dust material is 2000 kg/m3 .
Solution
The terminal velocities for the 0.1, 1, and 10 micron particles can be estimated from the s = 2000
kg/m3 curve on Figure 20.2. For more precise values the slip corrected Stokes' equation (20.12)
gives
9.81(2000 1.1) d 2 C c
d 2 g ( s a )C c
= 6.052 10 7 d 2 C c
=
ut =
6
18 a
18 18 10
Applying this relationship to each of the given particle diameters and reading corresponding
values of Cc from Table 20.2 (remembering to multiply microns by 10-6 to convert diameters to
metres) gives
d (microns)
0.1
10
Slip correction, Cc
2.91
1.168
1.761 x 10-6
7.069 x 10-5
100
1.017
ut (m/s)
calculated
estimated from Figure 20.2
1.75 x 10
-6
7.1 x 10
-5
6.155 x 10-3
6.0 x 10
-3
0.605
out of range
20 -
1.0E-02
terminal velocity m/s

and Brownian displacement m in each sec
1.0E-03
1.0E-04
9000
1.0E-05
7000
5000
s = 800 kg/m3
4000
1200
3000
1600
1.0E-06
2000
2500
1.0E-07
0.01
0.1
10
Particle geometric diameter microns

Figure 20.2 Slip corrected terminal velocities and Brownian displacements of dust particles
falling through still air of viscosity 17.9 x 10-6 Ns/m2 (20C). Based on equations
(20.12) and (20.17 with barometric pressure of 100kPa and g = 9.81 m/s2.
20 -
These terminal velocities for the 0.1, 1.0 and 10 micron particles are acceptable as the diameters
fall into the range of applicability of the slip-corrected Stokes' equation. The 100 micron particle,
however, is well above the 20 micron limit for laminar flow and we must revert to the more
general equation (20.9). This requires a value of coefficient of drag, CD. Table 20.1 would allow
us to calculate CD if we knew the Reynolds' Number. Unfortunately, that depends upon the
terminal velocity which we are trying to find. The problem can be solved iteratively, starting from
the approximation ut = 0.605 m/s given by the Stokes' equation, (20.8).
Re =
1.1 100 10 6 u t
a d ut
=
= 6.111u t
a
18 10 6
(20.13)
= 6.111 0.605 = 3.7
Table 20.1 gives the appropriate expression for coefficient of drag as

CD =
24
1 + 0.15 Re 0.687
Re
)=
24
1 + 0.15 3.7 0.687
3 .7
= 8.882
(20.14)
Equation (20.9) now gives an improved value of ut

ut =
4 d g ( s a )
=
3
CD a
=
4 100 10 6 9.81(2000 1.1)

3
1.1 C D
1.5417
CD
1.5417
8.882
= 0.517 m/s
(20.15)
Equations (20.13, 20.14 and 20.15) can readily be entered into a programmable calculator or
spreadsheet software for iterative solution. The values of the variables over eight iterations are as
follows.
u (m/s)
0.605
0.517
0.485
0.472
0.467
0.465
0.464
0.463
Re
3.70
3.16
2.96
2.89
2.85
2.84
2.84
CD
8.88
10.10
10.66
10.90
11.00
11.05
11.06
The procedure converges to ut = 0.463 m/s

The time taken for each of the particles to fall through
2 m can now be determined as t = 2/ut .
diameter
(microns)
0.1
1.0
10
100
ut (m/s)
1.761 x 10-6
7.069 x 10-5
6.155 x 10-3
0.463
t
315 hours
7.9 hours
5.41 minutes
4.32 seconds
It is clear from this example that little gravitational settlement of respirable dust (< 5 microns) can
be expected within the retention times of ventilated areas underground. Coupled with the effects
of Brownian motion, submicron particles can be considered to remain in permanent suspension.
Indeed, Figure 20.2 indicates that for the 0.1 micron particle Brownian displacement is the
dominant effect.
20 -
20.2.2 Brownian motion

For very small particles, the bombardment by fluid molecules is no longer balanced on all sides.
The result is that the particles undergo random and jerky displacements. This is known as
Brownian motion and can be seen under an optical microscope.
20.2.2.1. Brownian displacements
As Brownian movements are random, it is necessary to analyze their effect statistically on a
complete population of particles. If we consider a vertical plane in still air of uniform dust
concentration and with only Brownian motion causing horizontal movement of the particles, then
the average displacement of particles moving through the plane in one direction (+ x ) will be
equal to the average displacement of particles in the opposite direction (- x ). Hence, the net
displacement is zero - not a very useful result. However, if we square the displacements (positive
or negative) then the sum is always a positive number. We can then quantify Brownian motion in
terms of mean-square displacement, ( x )2
A relationship for mean-square displacement was first derived by Einstein in 19051 (see, also,
Seinfeld, 1986) and has been verified by numerous observers:
(x)2 = 2
where M
R
=
=
A
t
=
=
and
(The ratio
Cc
MR
T
t
3 d
A
m2
(20.16)
molecular weight of gas

gas constant (J/kgK)
(note that MR = Ru = universal gas constant, 8314 J/K kg-mole, Section 3.3.1.)
Avagadro's constant (6022 x 1023 molecules in each kg-mole)
time over which the displacement takes place (s)
Ru
8314
J
=
= 1.381 10 23
is known as Boltzmann's constant.)
23
molecule K
A
6022 10
For air at 20C (T = 293 K), the viscosity is 17.9 x 10-6 Ns/m2. Equation (20.16) can then be
simplified to
(x )
2 1.381 10 23 293 C c t
d
3 17.9 10 6
or
x = 6.925 10 9
Cc t
d
(20.17)
Using the values of Cc given in Table 20.2 and Cc = 1 for d > 10 microns, equation (20.17) has
been superimposed on Figure 20.2 with t set at 1 second. This allows the Brownian displacement
to be compared with the terminal velocity curves. Inspection of the Figure indicates that at some
point, as particle diameter decreases, Brownian displacement becomes predominant. This occurs
within the range 0.2 to 0.6 microns, dependent upon the density of the material. At all lower
diameters gravitational settlement is effectively nullified.
Historical note: In 1905 Albert Einstein completed his doctoral thesis. It described his studies into the existence and behaviour of
atoms and molecules. He rapidly applied this work to explain the phenomenon of Brownian motion. Later that same year he
published his groundbreaking Theory of Special Relativity.
20 -
20.2.2.2. Brownian diffusivity

A consequence of random Brownian displacements is that migration of particles will occur from
regions of higher to lower dust concentrations. We can describe the process as a form of diffusion
and obeying Fick's Law:
N b = Db
where
Nb
and
c
x
Db
dc
dx
(20.18)
is the flux of particles through an area of 1 m2 in one second [particles/(m2 s)]

by Brownian diffusion
is the concentration (particles/m3 )
is distance (m) in the direction considered
is a coefficient known as the Brownian diffusivity (m2/s).
Let us now attempt to find a relationship
that will allow us to quantify the Brownian
diffusivity. Consider the 1 metre cube shown
on Figure 20.3. It contains c particles, i.e. a
concentration of c particles/m3 . In time t, a
net number of those particles, c, will diffuse
across a 1 m2 plane by Brownian
dislocations and as a consequence of the
concentration gradient dc/dx.
concentration
= c particles in 1 m3
Let us take the average Brownian

dislocation, x, to be the distance through
which the particles move in time t. Hence,
their average velocity in the x direction will
be x/t. Furthermore, the flux across the 1
m2 plane will be the number of particles
involved multiplied by their average velocity,
i.e.
N b = c
Figure 20.3 c particles diffuse from a concentration of

c particles/m3 through an orthogonal area of
1 m2 and over a Brownian dislocation
distance of x metres.
x
t
particles
m2 s
(20.19)
Combining equations (20.18) and (20.19)

dc
x
= Db
gives c
t
dx
Now, over the very small distance of a Brownian dislocation, x (= x ), we can state that c = dc
x dx = Db dt
and t = dt, giving
Integrating both sides between corresponding boundary limits
x2
= Db t
2
m2
(20.20)
As we chose the distance x to be the average Brownian dislocation, x , we can combine with
equation (20.16) to give
20 -
10
( x ) 2 = 2 Db t =
2M R T C c
t
A
3 d
m2
(20.21)
from which
Db
M R T Cc
A
3 d
m2/s
Again, for air at 20 C, inserting the values

MR/A =
1.381 x 10-23 J/(molecule K),
T
=
293 K
=
17.9 x 10-6 Ns/m2
and
gives
C
m2/s
D b = 2.398 10 17 c
d
(20.22)
(20.23)
Note, also, from equation (20.20) that the mean dislocation is related to the Brownian coefficient
of diffusion
x =
2 Db t
(20.24)
20.2.3. Eddy diffusion

The previous two sections have considered the effects of gravity and molecular bombardment on
dust particles. In ventilated areas, a larger influence is exerted on dust particles by the turbulent
nature of the airflow. The transport of dust particles by eddies can also be described by a
diffusion equation
dc
particles
Ne =
(20.25)
dx
m2 s
where
and
= eddy diffusivity (m2/s)

2
2
Ne = flux of particles through an area of 1 m in one second (particles/m s)
by eddy diffusion.
The total rate of diffusion by both Brownian action and eddies is given by combining equations
(20.18) and 20.25). Then
dc
particles
N = N b + Ne = (Db + )
(20.26)
dx
m2 s
[A glance back at equations (A15.2) and (A15.5) reveals the analogy with diffusion for both heat
and momentum.]
The flux of particles passing from the turbulent core of an airflow through the buffer boundary
layer to the laminar sublayer is of particular interest as these are particles that have a high
probability of being deposited on the solid surfaces. Gravitational settlement will, of course, add to
such deposition on floors or other upward facing surfaces. Eddy action can impart sufficient
inertia to a dust particle to carry it into the laminar sublayer. Within the sublayer there are no such
eddies. Hence only Brownian bombardment can superimpose further transverse forces. Ignoring
any effects of re-entrainment, Brownian dislocations at the surface and away from the surface will
be zero. Hence there will be a Brownian concentration gradient towards the surface. Coupled with
the initial transverse inertia, this will tend to produce deposition of particles that enter the laminar
20 -
11
sublayer. As may be expected, this phenomenon is influenced by the same factors that affect the
boundary layers of fluid flow through rough ducts, i.e. fluid density, viscosity and velocity
(Reynolds' Number) as well as the roughness of the surface.
The average size of eddies grows from zero at the edge of the laminar sublayer to a maximum
within the turbulent core and, hence, varies with distance, y, from the surface. In order to take the
other variables into account, a dimensionless distance, y* is defined as
u
y= y
(20.27)
2f
where y
u
f
and
=
=
=
=
=
actual distance from the surface (m)

average velocity of fluid (m/s)
coefficient of friction for the surface (dimensionless
fluid density (kg/m3 )
dynamic viscosity (Ns/m2)
Note that yu / has the form of a Reynolds' Number. (The group u/2f is sometimes referred to
as the friction velocity.) Values of eddy diffusivity are suggested in Table 20.3.
Dimensionless distance from
surface
0 < y* < 5 (laminar sublayer)

5 < y* < 20 (buffer layer)
y*
> 20 (turbulent core)
Eddy diffusivity (m2/s)

0.001
0.012
0. 4
3
y
[y 1.6]2
(y 10 )
Table 20.3. Expressions for eddy diffusivity, , as a function of dimensionless distance from a
surface, y* (after Owen, 1969).
In order to track the combined Brownian and eddy transverse transportation of dust particles
across an airway, it is necessary to carry out integrations of equation (20.26) across each of the
zones specified in Table 20.3 (Bhaskar and Ramani, 1988). This is accomplished in a manner
similar to that used for convective heat transfer in Appendix A15.3.
20.2.4. Other forms of dust transportation
The processes of sedimentation, Brownian and eddy diffusion, coupled with coagulation, are the
predominant mechanisms leading to the deposition of dust particles. There are, however, other
phenomena that play a secondary role in governing the behaviour of airborne dust.
Many particles gain an electrical charge during formation. The effects of frictional flow as air
moves through a duct or airway can also induce electrical charges on dust particles. Even
particles that are initially uncharged may gain dipole characteristics due to Van de Waal's forces.
The primary effect of electrostatic forces is to increase rates of coagulation (Section 20.2.5).
20 -
12
Suppose a dust particle of charge, q, moves through an electrical field of strength E, then it will
experience an electrostatic force, qE. This may occur particularly around electrical equipment. At
equilibrium velocity, this force is balanced by fluid drag (equation (20.11) for laminar flow around
the particle), giving
3 a d
ue = qE
Cc
(20.28)
where ue = the electrical migration velocity relative to the air (m/s).

The induction of an electrical charge on dust particles to assist in deposition is utilized in
electrostatic precipitators (Section 20.4.2.3.) and in the control of paint or powder sprays.
However, the high voltages that are required impose a limit on the use of such devices in
underground openings.
Phoretic effects refer to phenomena that impart a preferential direction to Brownian motion.
Thermophoresis is the migration of particles from a hotter to a cooler region of gas and is caused
by the enhancement of Brownian displacement at higher temperatures (equation (20.16)). The
dust particles are subjected to greater molecular bombardment from the side of higher
temperature. The temperature gradient must be considerable to produce a significant effect and
the phenomenon has little influence on dust deposition in mine airways. However, it is utilized in
instruments such as the thermal precipitator (Section 19.4.2.).
Photophoresis occurs when an intense light beam or laser is employed in a dusty atmosphere.
The absorption of light by the particle causes an uneven temperature field to exist around that
particle. The resulting excitation of nearby gas molecules causes thermophoresis to occur in a
direction that depends upon the induced temperature field around the surface of the particle.
An effect that encourages dust deposition on wet surfaces is diffusiophoresis. The migration of
water vapour molecules away from an evaporating surface will result in a replacing flux of the
more massive air molecules towards the surface. The result will be a net Brownian force on dust
particles also towards the surface.
20.2.5. Coagulation
In any concentration of dust particles, collisions between the particles will occur due to Brownian
motion, eddy action or differential sedimentation. Dependent upon the surface properties of any
two such particles, they may adhere together to form a larger single particle. As the process
continues, some particles will grow to the extent that their terminal velocity becomes significant
and they will flocculate out of suspension. This phenomenon of coagulation is influenced by the
number and size distribution of the particles (large particles are more likely to be struck by other
particles), temperature and pressure of the air (governing Brownian displacements) and electrical
charge distributions. The shape of the particles and the presence of adsorbed vapours on their
surfaces also affect the probability of their adhering upon collision.
Analysis of coagulation is, again, an exercise in statistics. Consider, first, a concentration of n
3
3
particles in 1m . The average frequency of collisions (dn/dt particles involved in collisions per m
per second) clearly depends upon the number of particles in that space. We can write
dn
= an
dt
particles
s m3
(20.29)
where a = the probability of any two particles colliding. (Negative as the number of discrete
particles is decreasing with time.)
20 -
13
However, the probability of collision is itself proportional to the number of particles

a = Kn
giving
1/s
(20.30)
dn
= Kn 2
dt
particles
(20.31)
sm 3
3
K is known as the coagulation coefficient or collision frequency function (m /(particles.s)).

Equation (20.31) can be integrated readily:
dn
1
n
K dt
= Kt + constant
At t = 0, n = no = original concentration of particles, giving

constant =
1
no
so that
particles
m3
1
1
= Kt +
n
no
(20.31a)
i.e. at any given time, t, the particle concentration is given as

n =
no
no K t + 1
particles
(20.31b)
m3
Values of the coagulation constant can be found for any given dust cloud by plotting the variation
of particle concentration with respect to time. For Brownian coagulation of equal sized particles in
a continuum, K is given by
K =
8 MR T
3 A a
m3
particles.s
Hence for air at 20C ,

MR
= 8314 J/K kg-mole,
T
= 293 K and
giving K
0.6 x 10
-15
(Flagan and Seinfeld, 1988)
m /(particle. s)
=
=
(20.32)
23
6022x10 ,
-6
2
17.9 x 10 Ns/m
(20.33)
Ranges of size distribution and the other matters that influence coagulation result in considerable
variations being found in observed values of the coagulation coefficient.
There is a further problem that limits the applicability of this analysis; not only has it taken no
account of the differing sizes of particles, K changes as the agglomerates grow larger. A
somewhat more sophisticated approach concentrates on one size range at a time and considers
the appearance of particles of that size by agglomeration of smaller particles. Additionally, their
progression out of the size range as they continue to grow should be taken into account. Let us
assume, for the sake of explanation, that diameters are additive. (Actually, we should use particle
volume rather than diameter.) Then, for example, particles of size 10 microns can appear by
coagulation of smaller particles. If we employ subscripts to denote the size of particles, then
20 -
14

61
and
69
610
i.e. 6 (1 micron particles) agglomerating with 6 (9 micron particles) yields 6 (10 micron particles).
Similar examples are
Totals:
32
53
34
25
38
and
38
and
57
and
36
and
25
particles collide to yield
310
510
310
210
19 particles of size 10 microns.
In each of these groups, the collisions result in the number of particles being halved. Using the
concept of coagulation coefficient and the form of equation (20.31), we can write that the rate of
formation of particle size k (10 microns in our example) is:
dn k (formation ) 1
=
2
dt
k 1
ij
ni n j
i =1
particles formed
(20.34)
m3 s
where Kij is the particular coagulation coefficient for colliding particles of size i and j
and
nk is the number of particles of size k that are formed from the collisions of ni particles
(size i) and an equal number of nj particles (size j).
However, while all of this is going on, particles of size k are disappearing because further
coagulation causes them to grow out of that size range. This can occur by each particle size k
agglomerating with another particle of any size. In this case, we count the number of k size
particles that are disappearing rather than being formed. Hence, we no longer require the factor
of 1/2 and can write:
max
dn k (disappearance)
K km n m n k
=
dt
m =1
particles lost
m3 s
(20.35)
where max = largest size of particle to be considered relevant to the processes of coagulation
As nk has a single value at any given time, t, it can be brought outside the summation sign.
Combining equations (20.34) and 20.35) gives the overall rate of change of concentration of
particle size k:
dn k
1
=
2
dt
k 1
i=1
max
K ij n i n j n k
km n m
m =1
particles
m3 s
(20.36)
This result was reported by Chung (1981) but attributed to Smoluchowski. Even more complex
analyses have been conducted for liquid aerosols involving not only particle size changes by
coagulation but also by evaporation. These are of relevance in meteorology and surface
atmospheric pollution.
20.2.6. Impingement and re-entrainment
The phenomena of impingement and re-entrainment become significant only in situations of high
velocity or excessive turbulence such as may occur in and around ventilation shafts or fan drifts.
In such cases, the momentum gained by some dust particles may cause them to be ejected from
20 -
15
the curved streamlines of eddies and impinge on the walls or other solid objects. Deposition by
impaction of the particles on the walls can then occur. This is the principle employed in impact
dust samplers such as the konimeter (Section 19.4.2.).
Impact deposition in mine airways is counteracted to some degree by re-entrainment in those
same conditions of high velocity and turbulence. A particle on any surface and submerged within
the laminar sublayer can be made to roll over the surface by viscous drag of the air when a
sufficiently high velocity gradient exists across the sublayer. An accelerated rolling action may
cause the particle to bounce until it momentarily escapes beyond the sublayer where capture by
eddies can re-entrain it into the main airstream. Chaotic turbulence can have the same effect by
transient thinning of the sublayer. The phenomena associated with these boundary layer effects
are, again, influenced by Reynolds' Number and surface roughness. Re-entrainment can be
analyzed by considering the drag and frictional forces on particles on or very close to solid
surfaces (Ramani and Bhaskar, 1984).
20.2.7. Computer models of dust transport
The earlier mathematical models developed to describe dust transport in mine airways were
empirical in nature (e.g. Hamilton and Walton, 1961). The growing availability of digital computers
since the 1960's, combined with a better understanding of aerosol behaviour, led to the
development of mathematical models to simulate the behaviour of dust particles in mine
ventilation systems (Bhaskar and Ramani, 1988). Such a model may be based on a form of the
convective diffusion equation
dc
2c
dc
= Ex
u
+ sources sinks
2
dx
dt
x
partic les
m3 s
(20.37)
where c = concentration (particles/m )

t = time (s)
x = distance along the airway (m)
u = air velocity (m/s)
2
and
Ex = turbulent dispersion coefficient in the x direction (m /s)
This can be solved numerically between given boundary limits of time and distance
(Bandopadhyay, 1982) to track the temporal variations of dust concentration along a mine airway.
The "sinks" term is determined from the relationships given in the preceding subsections and, in
particular, the effects of gravitational settlement, Brownian motion, eddy diffusion and
coagulation. The "sources" must be defined as a dust production - time curve or histogram that
characterizes the make of dust from all significant sources along the length of airway considered.
20.3. THE PRODUCTION OF DUST IN UNDERGROUND OPENINGS

The majority of dust particles in mines are composed of mineral fragments. Oil aerosols may
become significant when drilling operations are in progress. Diesel exhaust particulates can also
form a measurable fraction of airborne dust in those mines that utilize internal combustion
engines. However, in this Section we shall concentrate on the manner and processes through
which mineral dusts are formed. Although the primary means of controlling mine dusts are
discussed in detail in Section 20.4 we shall introduce some of these, for particular operations, in
this Section.
20.3.1. The comminution process
Mineral dusts are formed whenever any rock is broken by impact, abrasion, crushing, cutting,
grinding or explosives. For any given material, the energy input required to break the rock is
20 -
16
proportional to the new surface area produced. As dust particles have a large surface area
relative to their mass, it follows that any fragmentation process which produces an excessive
amount of dust involves an inefficient use of energy. Before discussing specific operations that
produce dust, a valuable insight into particles size distribution can be gained from a brief analysis
of the comminution process.
Suppose a given brittle material is broken into fragments and the particles classified into a series
of size ranges. Commencing with the mass of finest particles and progressively adding on the
mass of each next coarser range, a table of cumulative "mass finer than" can be assembled. If
this is plotted against particle diameter on a log-log basis (Figure 20.4) then a straight line is
obtained for the smaller particles and curving over at larger sizes. The curve of Figure 20.4
m
r

x
follows an equation of the form M = 1 1

kg
(20.38)

xo
where x
= particle diameter, (m) - we use x here, temporarily, in order not to confuse
diameter with the differential operator, d
xo
= diameter of the initial fragment (m)
M
= cumulative mass finer than size x (kg)
r
is a constant that depends upon the particular comminution process and
m
is a characteristic of the material having values in the range 0.5 to 1 and varying
only slightly with the method of comminution. (This is known as the Gaudin
Meloy Schuhmann equation (Marshall, 1974; Gaudin and Meloy, 1962).)
If the term [1 x/xo]r is

expanded by the
binomial theorem, then
for x << xo
Cumulative mass M
(fraction of initial mass)
x
M = r
xo
0.1
0.01
0.001
0.001
0.01
0.1
kg
(20.39)
For dust particles, x is

certainly very much
smaller than xo. Equation
(20.39) quantifies the
straight line portion of
Figure 20.4 and has
been shown to hold for
particle sizes down to
0.01 microns (National
Research Council, 1980).
Particle diameter as fraction of original fragment
Figure 20.4 Typical size distribution graph of cumulative

mass finer than against particle size.
Let us now try to find a means of determining (i) the mass and (ii) the number of particles in each
size range:
20 -
17
(i) mass
Consider the mass, dM, of particles contained within the incremental range x to x + dx.
Differentiating equation (20.39) gives
rm
dM = m x m 1 dx = C x m 1 dx
kg
(20.40)
xo
where C = constant for that particular material, process and initial size.
Now let us take a finite size range from, say, D/10 to D (e.g. 0.5 to 5 microns). Then integrating
equation (20.40) between those limits gives the corresponding mass for that range.
M (D / 10 to D ) =
or
[ ]
C m
x
m
D
D / 10
C m
1
D 1 m
m
10
M (D / 10 to D ) = constant D m
(20.41)
As m is always positive this equation shows that the mass in each size range increases with
particle diameter. In practice this means that only a small part of the total rock broken will be
produced as dust particles. For coal, values in the range 5 to 9 kg per tonne (0.5 to 0.9 percent)
of particles less than 7 microns have been reported (Qin and Ramani, 1989). However, only a tiny
fraction of this will become airborne as respirable dust.
(ii) number of particles
Returning to our infinitely small increment of particle size range, x to x + dx, the volume of each
particle is x 3 / 6 . If the material is of density, , then the mass of each particle becomes
x 3 / 6 . For dn particles in that range, the total mass becomes
x3
6
dn = dM = C x m 1dx
giving
dn = C ' x m 4dx
where
C' = constant for that material, process and initial size.
particles
Integrating over the finite size range D/10 to D gives

D
C'
C'
1
n(D / 10 to D ) =
x m 3 D / 10 =
D m 3 1 m 3
m3
m 3
10
i.e
constant
n(D / 10 to D ) =
particles
D 3m
(20.42)
(20.43)
As m lies in the range 0.5 to 1.0, this shows that the number of particles rises logarithmically as
the particle diameter decreases.
Equations (20.41) and (20.43) indicate that in any rock breaking process, the bulk of mass will
appear as larger fragments. However, the number of fine dust particles produced may be
enormous. Fortunately, most of those particles remain attached to the surfaces of larger
fragments. The degree to which dust particles are dispersed into the air would seem to depend
upon the nature of the rock as well as the comminution process. For brittle materials, the
fragmentation becomes more 'explosive' in nature; the resulting surface vibration causes an
enhanced dispersion of dust particles into the air. Hence, although comminution of softer
materials may generate more dust particles, a greater proportion of those will remain adherent to
the surfaces of larger particles and will not become airborne. The production of airborne
respirable dust has been reported in the range 0.2 to 3.0 grams per tonne (Qin, 1989; Knight,
1985).
20 -
18
20.3.2. Mechanised mining

abraded dust
crushing
breaking off
= rake angle
= clearance angle
(a) Shallow cut
(b) Deep cut
(c) Blunt pick
Figure 20.5 A zone of pulverized rock forms

ahead of a cutter pick.
Machines that break rock from the solid have the

potential to be prolific sources of dust. These include
longwall power loaders, continuous miners,
roadheaders, tunnelling machines, raise borers and
drills. Figure 20.5(a) illustrates a pick point acting against
a rock face. Compressive forces induce a zone of
pulverized material immediately ahead of the pick point.
As the pick moves forward into that zone, the resultant
wedging action produces tensile failure along a curved
plane - a chip is broken away. The process is repeated
continuously as the pick advances. The majority of the
pulverized material is abraded onto the surfaces of the
rock face and the chip. The amount of dust produced at
the tensile failure plane itself may be quite small in
homogeneous brittle material and is influenced by the
presence of preformed dust in natural cleavage planes.
However, the explosive nature of that tensile failure is a
major factor in determining the amount of dust that is
projected into the air.
A machine that takes a greater depth of cut will require
higher torque and may be subject to greater vibration
and bit breakage. However, a comparison of Figures
20.5(a) and (b) indicates that more of the broken
material will be in the form of chips and, hence, the
amount of dust produced in terms of grams per tonne will
be reduced. The specific energy (per tonne mined) will
also fall. Figure 20.5(c) shows that the greater area of
contact given by a blunt pick will create additional dust in
the pulverized zone. If such wear causes a significant
reduction in the rake angle (Figure 20.5(a)) then the
back of the bit will rub against the newly formed face,
absorbing additional energy and producing further
pulverized rock. Furthermore, as the clearance angle
reduces, the chip may not be ejected efficiently but
remain in place to be crushed against the unbroken rock.
The design of a rock cutting bit is a compromise between
the efficiency of cutting (energy absorbed per tonne),
wear characteristics and dust production. Considerable
diversity of opinion exists on preferred bit geometries for
given machines and rock types.
Another factor that influences the proportion of dust which
becomes airborne is the speed at which the pick moves.
For any given depth of cut, an increased speed results in
greater rate of comminution and, hence, dust production.
Additionally, movement of the cutter drum causes a higher
relative velocity to be induced between the local airstream
and the material on the face (or fragments broken from the
face). This assists in entrainment of dust particles into the
air (Section 20.2.5.). The effect of pick speed on airborne
dust is illustrated on Figure 20.6.
20 -
19
20.3.3. Supports
10
Pick speed
3 m/s
8
2
6
4
0.25
2
0
0
10
15
Figure 20.6 An illustration of the effects of pick speed and pick

depth of cut. Actual dust makes also depend on other factors
including sharpness of picks, effectiveness of dust suppression and
air velocities around the cutting head.
Crushing of roof and floor

strata by roof supports
may liberate significant
amounts of dust when the
support is moved. This
can be a particular
problem on mechanized
longwall faces that are
equipped with powered
hydraulic supports. As
setting and yield loads of
the supports increase so,
also, does the amount of
dust produced. The
repeated lowering and
raising of these supports
can give a near
continuous source of dust
on longwall faces. Unless
roof coal is left, this may
be high in quartz content.
The effect can be
minimized by using wideweb roof beams or
cushioning materials.
Sheets of flexible material
linking adjacent canopies
have also been used to
mitigate against roof dust.
20.3.4. Blasting
Drill and blast remains the predominant method of mining in metal (hardrock) mines. The peak
concentrations of dust and gases (Section 11.3.4.) that are produced by the larger blasts are
usually too high to be diluted effectively by the normal ventilating airflow. This necessitates the
mine, or part of the mine, being evacuated of personnel for a re-entry period during and after the
blast. The length of the re-entry period can vary from half an hour to several hours for stoping
areas, dependent upon the layout of the ventilation network and the velocities of the air. This is a
classical example of isolating personnel from the dust.
The amount of dust produced depends upon a number of factors including
the mining method

the type of rock
the choice of explosive
the charge density and drilling pattern and
the type of stemming.
Blasts that eject the fragmented material into an air space (e.g. open stoping) will tend to produce
sharper but shorter lived peaks of dust than caving techniques. However, the latter may result in
20 -
20
more pulverized material capable of being entrained into the airstream during subsequent loading
and transportation operations. Water ampoules have been employed as stemming in an attempt
to reduce dust emissions from blasting operations.
Another technique is to place very fine but high capacity water sprays (fog machines) upwind of
the blast before and during the re-entry period. The combination of increased humidity and fine
water droplets assists in the agglomeration and sedimentation of dust particles. Spraying the
muckpiles produced by blasting is advisable before loading commences.
Secondary blasting also produces short peaks of dust concentration. This is yet one further
reason for employing methods of mining that minimize the need for secondary blasting.
20.3.5. Loading operations
This is another part of some mining cycles that can produce a great deal of dust whether the
loading operations are carried out by slushers, load-haul-dump (LHD) vehicles or loading
machines in headings. The dust arises from a combination of particles produced previously from
the mining process and held within the muckpile, and those that are generated by further
comminution during loading.
In addition to adequate (but not excessive) airflows, the primary means of combatting dust from
loading operations are water sprays and ensuring as little disturbance as possible to the loaded
material. The air velocity should not be less than 0.5 m/s at loading points. Abrasion of the floor
by heavy slusher buckets should be minimized. It is preferable to employ lighter buckets in
tandem operating at a speed of some 0.6 m/s (Sandys and Quilliam, 1982). Spray bars should be
located at intervals along slusher paths and, particularly, at points of transfer between buckets.
Muckpiles in headings should be sprayed with water continuously or frequently during mucking
operations except where hygroscopic minerals inhibit the copious use of water. In hot mines, prechilling of this water produces cooling as well as dust suppression (Section 18.3.5.2.). Steam
injection into muckpiles and the addition of wetting agents into the water has also been found to
be beneficial in some cases (Knight, 1985). Exhaust auxiliary ventilation is preferred for dusty
operations in headings, employing a force overlap, if necessary, to deal with gas emissions at the
face (Section 4.4.2.).
The skill of the driver of an LHD can have considerable influence on dust production. Choosing
the best point to insert the bucket into the muckpile will result in filling the bucket with a minimum
number of thrusts and with least disturbance to the material. Similarly, at the dump point, the
muck should be tipped gently and not dropped from a height. This should also be borne in mind
during the design of tipping operations from rail-mounted dump cars. Cones and chutes at dump
points should be designed to minimize impact forces on tipped material.
20.3.6. Transportation and crushing
Dust is produced throughout most mineral transportation routes, including conveyors, transfer
points, bunkers, skips, airlocks and vehicular traffic. Dust on the surfaces of conveyors may be
re-entrained into the air due to vibration of the belt as it passes over rollers. Spillage returning on
the bottom belt, if not cleared, will generate dust as the material is crushed against rollers.
Similarly, an excessive use of water can result in dust adhering to the belt surface. This may
subsequently be deposited under the conveyor during the return journey of the bottom belt. Belt
scraper devices or brushes at the drive heads should be properly maintained and all
accumulations of debris or dust should regularly be cleaned from under the conveyor and at
return rollers. Conveyor structure should be inspected routinely and attention paid to damaged
idlers and centering devices.
20 -
21
Vehicle arrestors on rail transportation systems should incorporate deceleration devices in order
to avoid impact loads on either the vehicles or the transported material. Tracks should be
adequately maintained and not allowed to develop sudden changes in direction or gradient.
The mineral transportation routes and mine ventilation system should be planned together in
order to avoid, wherever possible, minerals being transported through an airlock. The high
velocities that can occur over belt conveyors at airlock leakage points can cause excessive
production of dust. This can be minimized by employing side plates and attaching a length of
flexible material (such as old belting) on the conveyor discharge side of the airlock so that it drags
over the surface of the conveyed material.
Unless the mineral is hygroscopic, it should be kept damp throughout its transportation through
the mine. Bunkers and, wherever possible, conveyor transfer points and stage loaders should be
shrouded and fitted with internal sprays. It is also useful to duct the air from such shrouds directly
into return airways. Sprays or dribbler bars onto conveyors some 5 to 10 m before a transfer point
are often more effective than sprays actually at the transfer point itself.
Ore passes in metal mines should avoid lengthy segments of free fall. Air leakage at dump and
draw points should be into the ore pass and, hence, pull dust laden air away from personnel. This
can be arranged by an opening into the ore pass and connected either directly or via ducting to a
return airway. If this is not practicable then dusty air drawn by a fan from an intermediate point in
an ore pass can be filtered and returned to the intake system.
Crushers in any mine are prolific sources of dust. Here again, sprays may be used on the
material before, during and after the crushing process. This is another situation where it is
particularly valuable to draw air from the crusher enclosure and filter it.
20.3.7. Workshops
Aerosols produced in underground workshops are likely to occur as oil mists, diesel particulate
matter and welding fumes. The latter may be handled by exhaust hoods extracting air from
welding bays and directing it into a return airway. Indeed, all of the airflows through workshops
should, preferably, pass into return airways. The general arrangements for diluting and removing
airborne contaminants from workshops are discussed in Section 9.3.5.
20.3.8. Quartz dust in coal mines
The availability of instrumentation that can discern the quartz content of mine dusts within each of
a range of particle sizes (Section 19.4.7.) has led to the observation that airborne dust in coal
mines often has a quartz content that is significantly higher than that of the coal seam being
worked. Furthermore, the percentage of quartz becomes particularly high in the finer sizes
including the respirable range (Ramani et al, 1988; Padmanabhan and Mutmansky, 1989).
Coupled with the special danger to health of quartz dust, this has led to research aimed at
discovering the causes of such anomalous appearances of quartz in airborne dusts of coal mines.
There would appear to be at least two explanations. First, roof and floor strata usually have a
higher quartz content than the coal seam. Hence any fragmentation of those strata will cause
emissions of quartz dust. This can occur by rock-winning machines cutting into the roof or floor,
cross-measures drilling for roof-bolting or other purposes, development drivages out of the seam
or exceeding the height of the seam, hydraulic roof supports and fracturing of roof or floor strata.
A second, less obvious, cause of the apparently anomalous percentages of quartz in the dust of
coal mines is hypothesized to be the different comminution characteristics of coal and quartz
(Section 20.3.1.). Fragmentation of the stronger and more brittle quartz minerals may result in a
greater proportion of that dust being ejected into the air than is the case for coal. The greater
degree of entrainment would favour the finer particles.
20 -
22
20.4. CONTROL OF DUST IN MINES

The initial decisions that affect the severity of dust problems are made during the stages of
design and planning for the mining of any geological deposit. The methods of working, rate of
mineral production and equipment chosen all influence the amount of dust that is generated and
becomes airborne. The layout of the mine, sizes and numbers of airways, and the efficiency of
the ventilation system dictate the rate at which airborne contaminants, including dust, are diluted
and removed from the mine.
For an existing mine, there are four main methods of controlling the production, concentration and
hazards of airborne dust:
Suppression - the prevention of dust becoming airborne

Filtration and scrubbing - the removal of dust from the air
Dilution by airflow , and
Isolation - separation of personnel from the higher concentrations of dust.
In general, good management and housekeeping at a mine assist greatly in maintaining control of
the dust problem. These measures include planned maintenance schemes for equipment,
quantitative ventilation planning, cleaning up spillage, rock debris and local accumulations of
dust, and adequate supervision of work practices.
20.4.1. Dust suppression
It is difficult and often expensive to remove respirable dust from the air. Hence, every attempt
should be made to prevent it from becoming airborne in the first place. Methods of achieving this
are known collectively as dust suppression and are discussed in this section.
20.4.1.1. Pick face flushing and jet-assisted cutting
Figure 20.5 gives a visual impression of how a rock face is pulverized in advance of a moving
cutter pick. Pick face flushing involves directing a jet of water at the pick point during the cutting
process. This has been found to give markedly improved dust suppression when compared to
conventional water sprays on the drums of shearers, continuous miners or tunnelling machines.
The water that feeds each jet can be channeled through conduits drilled in the bit holder and via a
phasing valve that activates the jet only while the bit is cutting rock. Water filters are required to
prevent blockage of the nozzles. A further advantage of pick face flushing is that the streak of
incendiary sparks that often appears behind the pick in dry cutting is quenched. Hence, the
incidence of frictional ignitions of methane is reduced greatly. Interlock switches may be
employed to ensure that the machine cannot operate without the dust suppression water being
activated.
A number of researchers have investigated the extension of pick face flushing to much higher
water pressures, not only to further improve dust suppression but also in an attempt to produce a
higher efficiency of rock cutting. The use of high pressure water jets alone, with or without the
addition of abrasive particles, has had only limited success as a practical means of mining.
However, combining the mechanism of cutter picks with high pressure water jets directed at the
pick point has led to significant improvements in machine performance and the extension of
mechanized mining to much harder material that, previously, could be mined only by drill and
blast techniques. This technique is known as jet assisted cutting.
In addition to environmental enhancements, jet assisted cutting permits the same rate of
comminution with reduced loading on the cutter pick. This results in a significant reduction in wear
and, hence, less production time lost because of picks having to be changed. Furthermore, the
total specific power (per tonne mined) required by the combination of a high pressure water pump
and the cutting machine can be less than that of a conventional machine.
20 -
23
The benefits of jet assisted cutting are attainable by increasing the water pressure but reducing
the nozzle size in order to keep the flow rate no greater than that employed in conventional pick
face flushing. This can be important in hot mines or where floor strata react adversely to water.
However, it has been reported that there is little apparent improvement in levels of airborne dust
until the water pressure attains some critical value (Taylor et al, 1988). This would appear to be in
the range 10 to 15 MPa for cutting coal. After the critical water pressure is attained, a dramatic
reduction in airborne dust can be expected. However, this levels out again at water pressures in
excess of 20 MPa. Indeed, if the velocities of the jet and resulting spray are too high then reentrainment can exacerbate dust concentrations. Work continues on the preferred location of the
jet. Distances as small as 2 mm between the nozzle and the pick point have been suggested
(Hood et al, 1991).
The environmental and operational benefits of jet assisted rock cutting arise from at least seven
mechanisms (Hood, 1991).
(a) The pulverized rock immediately ahead of the pick point is wetted before it has an opportunity
to become airborne.
(b) The cooling action of the jet reduces wear: the bits remain sharp for significantly longer
periods of time and bit breakage is less frequent.
(c) Impact of the high velocity jet will produce an aerosol of very fine water droplets around the
cutting head, thus enhancing the agglomeration and capture of airborne dust particles.
(d) The washing action of the high-energy jet removes the cushion of pulverized material quite
efficiently. This allows the pick point to act on a much cleaner surface. The effect of a
cushion of pulverized rock is to distribute the force exerted by the pick over a broader front,
i.e. similar to that of a blunt pick, Figure 20.5(c). It is to be expected that the total amount of
finely crushed rock would be reduced.
(e) Penetration of the water into natural cleavage planes in the material and ahead of the
mechanical effect of the bit assists in pre-wetting dust particles that already exist within
those planes.
(f) Frictional ignitions of methane are virtually eliminated.
(g) The total specific energy required for the rock cutting process may be reduced.
20.4.1.2. Water infusion
A technique of dust suppression that has been employed by some coal mining industries since
the 1950's is pre-infusion of the seam by water, steam or foam. One or more boreholes are drilled
into the seam in advance of the workings through which the fluid is injected. The migration of
water through the natural fracture network of the coal results in pre-wetting of included dust
particles. The success of the method is dependent upon the permeability of the seam and the
type of coal-winning equipment employed. Good results have been reported where coal ploughs
are used - these relying more upon coal breakage along natural cleavage than the cutting and
grinding action of shearers or continuous miners (Heising and Becker, 1980).
In practice, some in-situ experimentation is usually necessary to determine the optimum injection
pressure and flowrate, and the time period of injection. Water pressures in the range 2 to 34 MPa
have been reported with water volumes of 7 to 20 litres per tonne in South African coal mines
(Sandys and Quilliam, 1982). Best results are obtained at fairly modest pressures but applied
over as long a period as possible. British experience in coals of limited permeability indicated
water pressures of 1.5 to 2.5 MPa and flowrates of 0.2 to 2 litres/min. If too high a pressure is
used then the water flows preferentially along major planes of weakness. Hydrofracturing may
20 -
24
occur, resulting in weakened roof conditions during mining and, possibly, backflow along bed
separation routes to give water inflows at the current working faces. Water infusion is not
recommended in areas of weak roof/floor strata or in the proximity of faults or other geological
anomalies. Steam and wetting agents have been employed in attempts to improve pre-saturation
of the zone. Water infusion must also be expected to influence the migration of strata gas
(Section 12.3.2.3.). Holes drilled initially for in-seam methane drainage may subsequently be
used for water infusion (Stricklin, 1987).
20.4.1.3. Wetting agents, foams and roadway consolidation
Worldwide experience of surfactants used as wetting agents in dust suppression water has been
highly variable. The technique has been employed since at least 1940 (Hartman, 1940). In
addition to the use of wetting agents to enhance the effects of water infusion, they may be
employed to improve the performance of sprays and also, at sufficiently high concentration, to
produce a foam around a rock fragmentation process.
Rocks vary considerably in their wettability characteristics. If surfactants added to muckpile
sprays are to be effective then they must be at a high enough concentration to cause penetration
of the fragmented material within an acceptable time period (Knight, 1985). The potential effects
of such concentrations on mineral processing should be considered carefully. Wetting agents
added to sprays are considered to have three beneficial effects. First, the reduced surface
tension allows greater atomization of the water - the droplets are smaller and greater in number,
hence, improving the probability of capturing dust particles (Section 20.4.2.1.). Secondly, the
existence of a liquid coating on dust particles will improve the chances of coagulation when two
particles collide. Third, the molecular structure of surfactants tends to counteract electrostatic
forces that may keep particles apart (Wang, 1991).
If a wetting agent is in sufficient concentration within a spray directed at a rock cutting device then
a foam can be formed that enshrouds the comminution process. This assists in coating the
fragments with a wetting fluid and in inhibiting entrainment of the dust into the air. Again, this
approach has met with mixed success (Bhaskar, 1991). It also interferes with ventilation of the
cutting head and should be used with caution in gassy conditions.
Accumulations of dust on roadway floors used for travelling in both underground and surface
mines can become airborne when disturbed by traffic. Roadway consolidation involves the use of
water, hygroscopic salts and binders to encapsulate the dust and maintain the floor in a firm but
moist state. Flakes of calcium chloride or magnesium chloride may be employed with lignin
sulphonate as a binder. The process involves raking and levelling the surface dust, and spraying
it lightly with water until it is wetted to a depth of some 2 to 3 cm. The addition of a wetting agent
may be necessary. The total amount of water required can be of the order of 40 litres per m2.
Free-standing pools of water should be avoided. The hygroscopic salt should be spread evenly at
a rate that depends upon the mean humidity of the air. For flake calcium chloride this will vary
from about 3.8 kg/m2 at a relative humidity of 40 percent down to 0.1 kg/m2 for a relative humidity
of 90 percent. It is advisable to apply three quarters of the salt during the initial application and
the remainder about one week later. The treatment will normally last for about six months
although re-spraying with water may be required after three months. Sodium chloride (common
salt) will be effective while the relative humidity remains above 75 per cent. In all cases, care
should be taken against corrosion of equipment and, in particular, within the vicinity of electrical
apparatus.
20.4.2. Removal of dust from air
The larger dust particles will settle out by gravitational sedimentation in the air velocities typical of
most branches in a mine ventilation system. Unfortunately, the more dangerous respirable
particles will effectively remain in suspension. Removing these from the air for large flowrates can
be expensive. The choice of a dust removal system is dictated by the size distribution and
20 -
25
concentration of particles to be removed, the air flowrate and the allowable dust concentration at
outlet. The size of any unit is governed primarily by the air volume flow to be filtered. Operational
costs can be determined from the product of the pressure drop and air flowrate through the unit,
pQ (Section 5.5.), and the means of supplying and filtering water in the case of wet scrubbers.
Where high efficiency is required for large flowrates over a wide range of particle sizes such as
the emergency filters needed on nuclear waste repositories (Section 4.6.), two or more types of
filters may be arranged in series, each taking out progressively smaller particles. This prevents
the finer filters from becoming clogged quickly and, hence, prolongs the life of the system before
cleaning or renewal of filters becomes necessary.
The efficiency of any dust removal facility, , may be expressed either in terms of number of
particles per m3 of air:
No. of particles in /m 3 No. of particles out /m 3
p =
No. of particles in /m 3 )
or in terms of mass of particles:
Mass of particles in /m 3 Mass of particles out /m 3
m =
Mass of particles in /m 3
(20.44)
(20.45)
In both cases, it is usual to further restrict the count of particles or mass to a specified size range.
Hence, for protection against pneumonconiosis, it is preferable to employ equation (20.45) for
respirable particles only, i.e. less than 5 microns equivalent diameter.
Devices to remove dust from air may be fitted to other pieces of equipment such as rock cutting
machinery, along transportation routes, within ventilation ducting or as free-standing units to filter
dust from the general airstream. In this section, we shall discuss principles of the devices that are
most commonly employed to reduce concentrations of airborne dust in mine atmospheres,
namely, water sprays, wet scrubbers and dry filters or separators.
20.4.2.1. Water sprays
Water is by far, the most widely used medium for conditioning mine air, whether it be for cooling
(Section 18.3), dust suppression or dust filtration. Open sprays can also be employed to direct,
control or induce airflows in order to protect machine operators from unacceptable concentrations
of dust (Section 20.4.4.).
The important parameters governing the efficiency of a spray can be highlighted through an
analysis of the capture of dust particles by water droplets. Consider Figure 20.7 which illustrates
air passing over a water droplet with a velocity relative to the droplet of ur. The streamlines of air
bend around the droplet. However, the inertia of dust particles causes them to cross those
streamlines. Particles that lie closer to the centre line of motion will impact into the droplet and be
captured by it.
We can conceive a flow tube of diameter y from which all particles are captured while particles
that are further from the tube centreline will be diverted around the droplet. The efficiency of
capture by a single droplet, E, can be defined as the ratio of the cross-sectional areas of the
capture tube to the facing area of the droplet:
E =
y2
Dw2
where Dw = droplet diameter (m)
(20.46)
20 -
26
If there is a uniform dust
concentration of n
particles/m3 then the rate
of capture of particles by
one droplet of water is
n u r area of capture tube
Air streamlines
Path of dust particles
particles m 2
m
s
m3
or particles/s; that is:
Dw
particles collected per droplet per

D2
second = E n u r w
(20.47)
4
Figure 20.7 A water droplet of diameter Dw captures dust particles from a

tube of air of diameter y. The relative velocity between the
droplet and the air is ur (m/s).
In order to maintain consistency

with the definition of dust
concentration that we are using
here (particles/m3), it is
preferable to restate this latter
expression in terms of particles
collected per cubic metre of air
rather than particles captured
per second. We can do this by
dividing by the air flowrate Q
(m3/s).
Then rate of capture by one droplet (dn/dt = rate of change of dust concentration) becomes
Dw2 1
dn
(one droplet ) = E n u r
dt
4 Q
particles s
droplet.s m 3
or
particles
(20.48)
droplet.m 3
where
t
time (s)
(negative as concentration is falling).
Now if water is dispersed in the spray at a volume flowrate of W (m3/s) and the volume of each
droplet is Dw3 / 6 , m3 , then the rate at which droplets are formed and pass through the spray is
W
Dw3 / 6
6W
Dw3
m 3 droplet droplets
=
s
s
m3
(20.49)
Multiplying by the particle capture for one particle, equation (20.48), gives the total rate at which
particles are captured per cubic metre of air:
Dw2 1 6W
dn
(all droplets ) = Enu r
dt
4 Q Dw3
W 1
3
= Enu r
Dw Q
2
particles
droplets
s
droplet m
3
particles
(20.50)
m3 s
20 -
27
Now consider Figure 20.8. Dust particles and air pass each other effectively in counterflow with a
relative velocity of ur such that they move through a separation distance dx in time dt, i.e.
ur
dx
dt
m
s
dx
Figure 20.8 Section through a water spray. Dust particles and water droplets pass each
other in counterflow with a relative velocity such that they move through a
separation distance dx in time dt, over which the dust concentration falls by
dn particles per m3 of air.
During that time, the dust concentration changes from n to n - dn, i.e. the rate of change of dust
concentration is -dn/dt (particles/m3s).
But
dn
dn dx
=
dt
dx dt
dn
ur
dx
particles
m3 s
Combining with equation (20.50) gives
dn
dn
ur
=
dt
dx
3
n
W
E
ur
2 Dw
Q
or
dn =
3
n W
E
dx
2 Dw Q
particles
m3
Integrating over the complete effective length of the spray, L (distance moved by particles plus
distance moved by droplets (in counterflow) in the x direction), gives the total number of particles
removed between the inlet concentration, Nin, and outlet concentration, Nout (particles/m3 )
N out
N in
dn
n
3 E W
=
2 Dw Q
ln{N out / N in } =
dx
0
3 E W
L
2 Dw Q
20 -
28
N out
N in
3 E W
= exp
L
2 Dw Q
(20.51)
Reference to equation (20.44) shows that the particle removal efficiency of the spray is given by
p =
i.e.
N in N out
N
= 1 out
N in
N in
3 E W
L
2 Dw Q
p = 1 exp
(20.52)
Examination of this equation is most instructive in understanding the performance of sprays. The
dust removal capacity of the spray increases with E, the capture efficiency of each droplet. To be
precise, this depends upon the nature of the flow and the relative sizes of dust particles and water
droplets. However, a coarse approximation for fully developed turbulence, based on work
reported by Jones (1978), can be assessed as
E = 0.266 ln(K ) + 0.59
(20.53)
over the range 0.2 < K < 4 where ln means natural logarithm and the dimensionless parameter K
is:
u r D p2
(20.54)
K =
9 Dw
where
Dp
=
=
particle density (kg/m3 )

particle diameter (m)
and
kinematic viscosity of the air (Ns/m2).
In particular, the capture efficiency increases with the relative velocity between the dust particles
and droplets (ur), the diameter (Dp) and density () of the particles (these three governing particle
inertia) and increases further as the water droplets become smaller (Dw).
Returning to equation (20.52) reinforces the fact that the overall efficiency of the spray improves
with smaller water droplets. A coarse spray of large water droplets will have very little effect on
airborne respirable dust.
A parameter of basic importance in equation (20.52) is the water to air ratio (W/Q). Values in the
range 0.1 to over 2 litres of water per cubic metre of air have been reported. The lower values
produce poor efficiency of dust capture. However, if too high a value is attempted then the
concentration of droplets may become so large that coalescence occurs. The larger droplets then
lead to decreased efficiency. A practical range of W/Q for sprays and wet scrubbers in mines is
0.3 to 0.6 litres/m3 . Tests on compressed air-powered atomizing nozzles have indicated an
optimum W/Q value of 0.45 litres/m3 (Booth- Jones et al, 1984).
The last point to be gleaned from equation (20.52) is confirmation of the intuitive expectation that
the spray efficiency is improved as the length (L) and, hence, time of contact between the air and
the water droplets is increased.
In order to produce the finely divided sprays necessary to affect respirable dust, a number of
methods are employed. The simplest technique is to supply high pressure water to the nozzles.
Pressures of some 3000 to 4000 kPa applied across suitable nozzles give smaller droplets at
20 -
29
spray velocities high enough to cause air induction - surrounding dust laden air is drawn into the
spray and thus improves the dust removal capacity of the unit. A variety of nozzle designs are
available commercially. These control the shape as well as influencing the atomization of the
spray. Full cone and hollow cone sprays have good air induction characteristics while fan shaped
sprays are excellent at confining the dust clouds produced by shearers and continuous miners.
Atomization is further improved in some nozzles by impinging the high velocity jet against an
impact surface located facing and close to the orifice. Another arrangement causes the water to
rotate rapidly around an orifice before ejection. In all cases, it is particularly important in mining
that nozzle designs should mitigate against blockage from particles either in the water supply or
(in the case of machine- mounted sprays), thrown forcibly against the jet from an external source.
Compressed air-assisted sprays can produce fine atomization with droplets in the respirable size
range. The water feed is connected into the compressed air supply close to the nozzles. The
water enters the compressed airstream either by its own applied pressure or by venturi action. It
is advisable to insert non-return valves into the water line. The combination of very high
turbulence at the nozzles and expansion of the compressed air into the ambient atmosphere
produces fine droplets.
Compressed air-assisted sprays can be further enhanced by the addition of a sonic device to the
nozzle (Schrder et al, 1985). Air expands through the nozzle into a facing resonator cup where it
is reflected back to complement and amplify the initial shock wave at the mouth of the orifice. An
intense field of sonic energy is focused in the gap between the nozzle and the resonator cup.
Water droplets issuing from the nozzle and passing through the sonic field are further broken
down to respirable sizes and, indeed, to submicron diameters. Similar effects can be achieved by
high frequency oscillation of pairs of piezo-electric crystals.
A high degree of atomization can be achieved without high pipeline pressures through
impingement devices. A free-standing "fog machine" of this type may consist of a stainless steel
disc spinning at about 3000 rpm. A low pressure water supply is fed to the centre of the disc.
Centrifugal action causes the water to flow outwards over the surface of the disc to impact at high
velocity on a ring of stationary and closely spaced vanes around the perimeter. A fan impeller
located on the upstream side of the disc projects the fog-laden airstream forward. The same
principle is employed in wetted fan scrubbers and in some industrial humidifiers.
20.4.2.2. Wet scrubbers
As the name suggests, these are devices that also employ water to achieve dust removal.
However, in this case the water streams (or sprays) and the airflow are controlled within an
enclosure designed to maximize the parameters that improve the efficiency of dust capture
(equation (20.52)). Wet scrubbers bring dust particles into intimate contact with wet surfaces and
within a highly turbulent mixture of air, water droplets and dust. They have become popular for
mining applications as they require less maintenance than most other dust filters and can achieve
respirable dust capture efficiencies exceeding 90 per cent.
Here again, we shall restrict our discussion to the operating principles employed in the most
common wet scrubbers. Many competing devices are marketed and manufacturers' literature
should be consulted to match performance with required duties, and to compare capital and
operating costs.
20 -
30

filtered
water
water
eliminator
fibrous bed
reject
water
fan
Figure 20.9 Fibrous (flooded) bed filter.
The fibrous (or flooded) bed) scrubber illustrated on Figure 20.9 is one of the most widely used
devices employed in mine dust collectors. Stainless steel or other non-corrosive material is used
as the fibre material. Water is either admitted along the top of the fibrous bed and allowed to
trickle downwards through it or, preferentially, the water sprayed directly into the air upstream
from the fibrous bed. The air follows a tortuous path through the bed while the inertia of the dust
particles causes them to strike and adhere to the wet fibres. The efficiency of dust removal
increases with the fineness of the fibres, the thickness of the bed and the velocity of the air. This
must be balanced by the resistance of the unit to airflow and, hence, the operating cost.
Efficiencies exceeding 90 percent for respirable dust can be attained.
The dust laden water collects at the bottom of the fibrous bed from where it is drained, filtered
and recycled. Arrangements must be made to remove the effluent sludge and to supply make-up
water. In all cases, wet scrubbers can be supplied with chilled water to achieve simultaneous
cooling and dust collection. Again, filtration within the chilled water cycle is necessary.
A water eliminator is required by most designs of wet scrubber in order to remove residual
droplets of water. Several different systems of water elimination are available in practice including
a second fibrous mat, a series of wavy or inclined plates, turning vanes to induce swirl into the air
and, hence, throwing droplets outwards towards the duct walls, or an egg-tray arrangement. Here
again, droplet removal is achieved by impingement.
Figure 20.10 illustrates the principle of the wetted fan scrubber. Sprays upstream and/or at the
facing boss of a fan produce droplets that are mixed intimately and at high velocity with air across
and around the fan impeller blades. The polluted water collects around the internal surface of the
fan casing for removal and recycling. The addition of a fibrous bed downstream from the fan gives
a powerful combination of dust collection devices. A disadvantage of wetted fan scrubbers is the
pitting that may occur on the impeller blades and requiring additional fan maintenance. Designs
employing centrifugal as well as axial fans have been developed. Wetted fan scrubbers are well
suited to lower airflows and have an application as in-line dust collectors in auxiliary ventilation
ducting.
The venturi scrubber, depicted on Figure 20.11, has no moving parts. Sprays are located
upstream and/or at the throttled section of a venturi arrangement. Air velocities through the throat
are typically in the range 60 to 120 m/s with a high degree of turbulent mixing. This encourages
the impaction of dust particles into water droplets. Venturi scrubbers are compact, simple and
rugged, and can reach efficiencies of more than 90 percent. However, it is costly in operating
power and is suitable for limited airflows only.
20 -
31

filtered
water
water
eliminator
reject
water
fan
Figure 20.10 Wetted fan scrubber.
clean
water
additional
(optional)
water supply
water
eliminator
Figure 20.11 Venturi scrubber

The flooded orifice scrubber, illustrated on Figure 20.12, also has no moving parts and has the
additional advantage that there are no nozzles that might become clogged. Air from the inlet duct
flows outwards beneath a lip that is submerged in water. Movement of the air causes extreme
agitation of the water and entrainment of droplets. Collection efficiencies of more than 80 percent
can be achieved with this system.
The preferred location for a dust collection device is as close as practicable to the source of the
dust. The types of wet scrubbers outlined in the previous paragraphs are suitable as freestanding units or within ventilation ducts. However, attempts to attach them to coal or rock
winning machines have shown them to be somewhat bulky for that application and insufficiently
robust to withstand the rigours of a working face. A device that met increasing favour for shearers
and continuous miners through the 1980's was the simple high pressure spray fan or induction
tube (Jones, 1978; Sartaine, 1985; James and Browning, 1988; Jayaraman et al, 1989). This is
illustrated by Figure 20.13 and consists of a water jet spraying into a tube of some 100 mm
diameter. The water is supplied at pressures in the range of 6 to 12 MPa through a nozzle of
about 1.5 mm diameter. The momentum of the fine droplets induces an airflow through the tube
and is very effective in removing dust. A single spray within a relatively small tube appears to be
more effective than multiple nozzles within a larger induction tube. Furthermore, hollow cone
sprays give a better performance than solid cone sprays.
20 -
32
airflow
airflow
water
outlet
water
inlet
Figure 20.12 Principle of the flooded (or wet) orifice scrubber.
A series of 9 to 12 high pressure spray fans

built into a longwall shearer drum is capable of
promoting an airflow of up to 2 m3 /s around
the drum and can give reductions in airborne
dust concentrations of 80 percent compared to
conventional pick face flushing using the same
amount of water (James and Browning, 1988).
The direction of induced airflow is away from
the coal face and towards the travelling track.
Hence, dust laden air is drawn around the
cutter picks and down the face side to the tube
high
inlets. At the outlet of the tubes, the dust-laden
pressure
droplets are discharged against deflector
water
plates and fall on to the conveyor. Similarly, a
number of induction tubes can be mounted in
Figure 20.13 The high pressure spray
parallel on the boom of a continuous miner
fan or induction tube.
(Jayaraman, 1989). When employed in
headings, the air induction may cause local
recirculation and greatly improved ventilation of the cutter heads. Provided that adequate airflow
is supplied to the face end of the heading, this will enhance the overall safety of the environment.
However, legislative enforcement agencies should be consulted in industries where recirculation
is prohibited.
The advantages of the high pressure spray induction tubes are that:
they are simple, robust and have no moving parts

they can be built into the machine structure
they promote ventilation of the cutter heads as well as removing dust
they give a good efficiency of dust capture and
provided that the water pressure is maintained, there is little chance of blockage.
20.4.2.3. Dry filters and separators

There are many situations in subsurface ventilation systems where increasing the humidity of the
air by the use of wet scrubbers is inadvisable. These include mines where heat and humidity is
already a problem although cycling chilled water through wet scrubbers will reduce temperature,
humidity and dust concentration simultaneously. Other difficulties that can arise from increases in
humidity include clogging of hygroscopic minerals during transportation, roof control where the
overlying strata is subject to rapid weathering, and where the mineral is subject to spontaneous
combustion. In such circumstances, it may be preferable to employ dry filters to remove airborne
dust.
20 -
33
dust
bag
clean air
dust
cake
dusty air
hopper
clean-out gate
Figure 20.14 Principle of a fabric filter.
R = R(fabric ) + R(cake ) =
where
and
Figure 20.14 illustrates a fabric

filter. The air passes through a
fabric leaving dust particles
adhering to the material or to the
dust cake that builds up on the
high pressure side. Air may flow
from the inside to the outside of
the bags as illustrated or in the
opposite direction, in which case
the bags are supported on an
internal frame. The dust cake is
dislodged at intervals of time by
mechanical agitation or a
reversed air pulse and falls to be
collected in the hopper. The
collected dust is either removed
dry and bagged by a mechanical
cleaning system or as a slurry
produced by addition of water to
the hopper.
The dust cake itself accomplishes
most of the filtration and
efficiencies of over 99 percent can
be achieved in the submicron
range. Airflow through the fine
apertures of the dust cake is
laminar. Hence, fabric dust
collectors tend to follow a linear
(p = RQ) relationship. The overall
resistance of the unit arises from
the combined effects of the dustimpregnated fabric and the dust
cake. It is a straightforward matter
to show that
m
+
A kf
k c A
dynamic viscosity of air (Ns/m2)
A
x
kf
=
=
=
kc
m
=
=
=
surface area of filter (m2)

thickness of fabric (m)
permeability of dust-impregnated fabric (m2)
permeability of dust cake (m2)
Ns
m5
(20.55)
mass of dust in the dust cake (kg)

density of the dust cake (kg/m3)
[The definition and units of permeability are explained in Section 12.3.2. However, manufacturers
may assume a standard value of air viscosity and quote filter permeabilities in terms of m3/s of
airflow through each (m2 of area) for unit pressure gradient through the material (Pa/m), i.e. m3/s
per m2 per Pa/m.]
20 -
34
Equation (20.55) quantifies the increase in resistance as the thickness and, hence, mass of the
dust cake builds up. This also increases the capture efficiency of the device. The pressure
developed by the unit fan will rise and the air quantity will fall. An excessive pressure may cause
rupturing of the filter fabric. The cake must, in any case, be dislodged before the airflow drops to
an unacceptably low value. A backward curved (non-overloading) centrifugal fan operating on a
steep pressure-quantity portion of its characteristic curve is advisable.
The simplest type of fabric cleaning mechanism is an electro-mechanical agitator. If operation of
the unit can be interrupted every few hours (dependent upon dust loading) then the fan can
automatically be switched off and the bags shaken by the agitator. More sophisticated units allow
continuous operation by cycling the filtration and cleaning around several separate
compartments.
Reverse flow cleaning involves a temporary reversal of air direction. This eliminates the
mechanical linkages of the agitator system and is preferred for some types of fabric such as glass
cloth where the severe flexing action of mechanical shaking may break the fibres. Pulsed jet
reverse flow increases the efficiency of cleaning. Acoustic methods have also been employed to
dislodge filter cakes.
The choice of fabric material usually lies between cotton weaves, felted fabrics or a synthetic
such as polypropylene. The felted fabrics give an initially higher efficiency but synthetics are
preferable where moist conditions or hygroscopic minerals may tend to produce a sticky dust
cake. A newly installed bag will have a relatively low resistance. The initial mechanism is that dust
particles will become lodged within the material. This increases both the resistance to flow and
capture efficiency. Subsequent cleaning cycles will remove the dust cake but will have little effect
on dust that has become impregnated in the material (more is dislodged from the smoother fibres
of synthetic material). It follows that performance tests on a fabric dust collector should be
delayed until dust impregnation of the material has reached steady state.
Two types of cyclone have been developed for dust removal, both of which can be operated dry
or with the addition of water to improve capture efficiency. The conical cyclone operates by the
dusty air being constrained into a helical vortex of reducing radius. Figure 19.4 was drawn to
illustrate the conical cyclones used in dust samplers. Larger versions can be used as dust
collectors. Dust particles are subjected to two opposing forces in a cyclone; the centrifugal force
that tends to throw the particles out toward the wall, and drag of the air which tends to pull them
inward toward the central air outlet tube. The greater the mass of the particle and the rotational
velocity the more efficient the cyclone will become. Hence, the performance is enhanced for
larger particles and as the physical size of the cyclone decreases. Cyclones are normally
employed in groups for air cleaning. The centrifugal action is improved by arranging for the air to
enter tangentially. It is essential to remove the dust from the base continuously in order to avoid
re-entrainment. The finer particles that escape in the outlet air may be removed by a second
cyclone or other filtration device connected in series.
The cylindrical cyclone imparts helical vortices to the airflow by means of turning vanes in a duct.
The dust which concentrates and moves in helical fashion along the walls is collected and
removed through an annulus formed by a second inner duct. The capital cost of cyclones is
relatively low. They have no moving parts and are easy to maintain. However, the power
requirements are such that they are constrained to applications of low airflow.
Electrostatic precipitators are used widely as air cleaners in buildings and for surface industrial
applications such as the removal of fly ash from power station stacks or capturing aerosols in the
chemical and metallurgical industries. Although a well designed electrostatic precipitator can
reach capture efficiencies of over 99 percent in the submicron range, their need for high voltages
prohibits their use in gassy mines and mitigates against their employment in other underground
facilities.
20 -
35
The principle of operation of an electrostatic precipitator is that when an aerosol is passed

through an electric field produced by a pair of electrodes then the particles will become charged
and migrate towards one of those electrodes. For industrial applications, the active electrodes are
charged to voltages between 20 and 60 kV while the dust collecting electrodes are earthed. The
electric field is considerably enhanced in regions of sharp curvature on the electrode surfaces.
For this reason, the active electrodes are often wires hanging vertically downward. The wires are
usually charged negatively as this gives a more stable performance for heavy duty performance
although ozone can be formed. High energy electrons are emitted from the negatively charged
wires. Each electron collision with a gas molecule causes the ejection of two further electrons
which go on to repeat the process. This escalating process produces an electron avalanche and
is often accompanied by a visible glow; hence, the phenomenon is termed a corona. The gas
molecules that have lost electrons become positive ions and migrate towards the negatively
charged wires. However, further away from the active electrodes, the free electrons lose their
kinetic energy to the extent that they are no longer capable of dislodging further electrons from
gas molecules but are, instead, absorbed into those molecules. The electron avalanche ceases
and the edge of the corona is reached.
The gas molecules are then negatively charged, i.e. negative ions, and migrate towards an
earthed electrode. During that migration they become attached to dust particles which are also,
therefore, drawn towards an earthed electrode and adhere by electrostatic attraction to the
surface of that electrode. Upon contact, the particles begin to leak their charge to the earthed
electrode. Other layers of charged particles arrive and build up progressively. They too will
gradually give up their charge. However, the outermost layer of dust is always the most heavily
charged and will be analogous to a skin compressing the underlying particles and causing the
build-up of a dust cake. The dust can be dislodged into a lower hopper by rapping the earthed
electrodes.
In tube electrostatic precipitators, a single wire forms the active electrode suspended in a metal
cylinder which acts as the grounded electrode. However, for the larger flows found in industrial
applications, the plate precipitator has become more common. This is illustrated in Figure 20.15.
Air passes over the charged wire electrodes which are suspended between a series of grounded
plates. The dust collects on the surfaces of the plates. For some applications, the mechanisms of
dislodgment by rapping may be replaced by running a film of liquid down the plate surfaces or by
periodically dipping the plates into a liquid bath.
The efficiency of an electrostatic precipitator can be determined by an equation first derived by W.
Deutsch in 1922:
A ue
= 1 exp
where A
Q
and
ue
=
=
=
(20.56)
area of plates (m2)

airflow (m3/s)
electrical (ion) migration velocity (m/s) (see equation (20.28))
The electrical migration velocity depends upon the type of dust and varies between 0.02 m/s for
fly ash to 0.2 m/s for gypsum. Although theoretical procedures have been derived for
quantification of the electrical migration velocity, tables of empirical values have, to this time,
proved to be more reliable (ASHRAE, 1988).
20 -
36
grounded plates
charged wire
electrodes
Figure 20.15 A plate electrostatic precipitator.

20.4.2.4. Personal respirators
Every effort should be made to maintain dust concentrations in subsurface workings within
mandatory threshold limits and safe for the health of the workforce. A final line of defence is the
personal respirator used to filter inhaled air. Two types are available. The first of these is a mask
that fits around the nose and mouth. The filter is necessarily a compromise between dust removal
efficiency and resistance. A respirator that requires more than about 150 Pa of pressure
difference at normal breathing rates is unlikely to be tolerated by personnel. Furthermore, contact
of the mask on the face can be irritating, especially in hot conditions. An improved version,
sometimes called an airstream helmet, utilizes a belt-mounted battery to power a small fan. This
passes air through a filter and up a tube to the helmet. The cleaned air flows downwards between
a transparent visor and the face of the wearer. This device does not rely on breathing effort nor is
there any direct face contact with the visor. It also provides eye protection with less visual
impedance than that given by goggles or safety glasses.
20.4.3. Dilution and layout of the ventilation system
Despite the availability of dust collectors, dilution of mine dust by the mine ventilation system
remains the primary method of controlling this hazard. The effects of airflow and air velocity have
already been discussed in earlier chapters. Section 9.3.3 deals with airflow requirements for
respirable and non-respirable dust while recommended air velocity limits are listed in Section
9.3.6. Exhaust systems of auxiliary ventilation are preferred for dust problems in headings
(Section 4.4.) while overlap arrangements can also handle gas emissions. Furthermore, it is
relatively straightforward to install in-line filters or dust collectors within ventilation ducts.
20 -
37
Controlled partial recirculation, where allowed by legislative authorities, coupled with dust filtration
systems, can result in very significant reductions in general body dust concentrations (Section
4.5). The district ventilation systems discussed in Section 4.3 and designed to facilitate the
dilution and removal of airborne pollutants in working zones apply equally well to respirable dust.
Consideration might also be given to homotropal ventilation in which the airflow and mineral flow
are in the same direction (Section 4.2.3.). As conveyors or other mineral transportation systems
are then in return airways, any respirable dust they produce does not pass on to a working area.
Furthermore, on a longwall face with uni-directional coal winning, few personnel need be on the
downwind side of the machine. Despite these advantages, homotropal ventilation does have
some drawbacks, particularly in mines with heavy gas emissions (Stevenson, 1985).
20.4.4. Separation of personnel and dust
In Section 20.3.4 we described the re-entry period after blasting in metal mines as a classical
example of the separation of personnel from dust concentrations. Several other methods are
available to reduce the exposure of individuals or groups to dust. The United States Bureau of
Mines was active in developing this approach, particularly for the protection of the operators of
longwall face equipment and continuous miners in room and pillar workings.
Airflow diverters of two types have been fitted to such machines. First, barriers have been added
to shearers in order to divide the face airflow before it reaches the location of the cutting drum.
This is positioned such that it provides a split of relatively clean air to the shearer operator. A
great deal of research has been conducted into the use of spray fans to control the direction and
flow of air at continuous miners and longwall shearers. Appropriate location and design of these
triangular or cone shaped sprays not only assists in dust suppression but also ensures that
airborne dust is diverted away from operators' positions (National Research Council, 1980).
Air curtains have also been employed to prevent dust clouds from reaching operators' positions,
as well as assisting in the ventilation of cutter heads (Ford and Hole, 1984; Froger et al, 1984;
James and Browning, 1988). The air curtains may be directed across the top, bottom and sides of
the cutting zone. They are produced from tubes of about 10 mm diameter maintained at an air
pressure of approximately 1.5 kPa. A 2.5 mm slot runs along the length of the tube with an
attached guide plate angled such that air leaves the tube tangentially, clinging to the guide plate
(the Coanda effect) until it is deflected into the required direction by a splitter. Entrainment of
additional air assists in both the ventilating and dust control effects.
Another development that reduces dust exposure to machine operators has been the use of
remote controls. These allow personnel to stand some distance from the mineral- winning
machines while maintaining control by hand-held wireless units. Finally, studies leading to the
reorganization of work practices have also promoted reduced dust exposure of face personnel
(Tomb et al, 1990).
References
Allen, M.D. and Raabe, C.G. (1982). Re-evaluation of Milliken's oil drop data for the motion of small
particles in air. J. Aerosol Sci. 13, pp 537-547.
ASHRAE (1988). American Society for Heating, Refrigerating and Air-conditioning Engineers, Equipment
Handbook, Chapter 11, pp 11-12.
Bandopadhyay, S. (1982). Planning with diesel powered equipment in underground mines. Ph.D. thesis.
The Pennsylvania State University.
20 -
38
Bhaskar, R., Gong, R., and Jankowski, R.A. (1991). Studies in underboom dust control to reduce
operator exposure to dust. 5th U.S. Mine Ventilation Symp., W. Virginia, pp 197-206.
Bhaskar, R. and Ramani, R.V. (1986). Behavior of dust clouds in mine airways. Transactions of AIME,
Vol. 280, pp 2051-2059.
Booth-Jones, P.A., Annegarn, H.J. and Bluhm, S.J. (1984). Filtration of underground ventilation air by
wet dust-scrubbing, 3rd Int. Mine Ventilation Congress, Harrogate, U.K., pp 209-217.
Chung, H.S. (1981). Coagulation processes for fine particles. Ph.D. Thesis, The Pennsylvania State
University.
Flagan, R.C. and Seinfeld, J.H. (1988). Fundamentals of air pollution engineering. Published by Prentice
Hall, 542 pp.
Ford, V.H.W. and Hole, B.J. (1984). Air curtains for reducing exposure of heading machine operators to
dust in coal mines, Ann. Occup. Hyg., Vol. 28, pp 93-106.
Froger, C., Courbon, P. and Koniuta, A. (1984). Dust-laden airflow control applied to worker protection:
scale model study of a steep seam working, 3rd Int. Mine Ventilation Congress, Harrogate, U.K., pp 215217.
Gaudin, A.M. and Meloy, T.P. (1962). Model and a comminution distribution equation for repeated
fracture, Trans. AIME 223, pp 243-50.
Hamilton, R.J. and Walton, W.H. (1961). The selective sampling of respirable dust in inhaled particles
and vapours (editor Davies), Pergamon Press, Oxford.
Hartman, I. and Grenwald, H.P. (1940). Use of wetting agents for allaying coal dust in mines, U.S.
Bureau of Mines, IC-7131, 12 pp.
Heising, C. and Becker, H. (1980). Dust control in longwall workings, 2nd Int. Mine ventilation Congress,
Reno, Nevada, pp 603-611.
Hood, M. (1991). Private communication on jet-assisted rock cutting.
Hood, M., Knight, G.C. and Thimons, E.D. (1991). A review of jet-assisted rock cutting. Trans. of
ASME, Journal of Engineering for Industry.
James, G.C. and Browning, E.J. (1988). Extraction techniques for airborne dust control, 4th Int. Mine
Ventilation Congress, Brisbane, Australia, pp 539-546.
Jayaraman, N.I. et al (1989). High pressure water-powered scrubbers for continuous miner dust control,
4th U.S. Mine Ventilation Symp., Berkeley, CA, pp 437-443.
Jones, A.D. (1978). Experimental and theoretical work on the use of a high pressure water spray to
induce airflow in a tube and capture airborne dust. MRDE Report No. 73, National Coal Board, U.K.
Jones, A.D. (1978). Optimal design for water-powered dust extraction, M. Phil. Thesis, University of
Nottingham, U.K.
Knight, G. (1985). Generation and control of mine airborne dust, 2nd U.S. Mine Ventilation Symp., Reno,
NV, pp 139-150.
Marshall., V.C. (1974). Comminution, Chameleon Press Ltd., London.
20 -
39
National Research Council (1980). Measurement and control of respirable dust in mines, U.S. National
Academy of Sciences, Washington, D.C., NMAB-363.
Owen, P.R. (1969). Dust deposition from a turbulent airstream (Aerodynamic capture of particles: editor,
Richardson).
Padmanabhan, S. and Mutmansky, J.M. (1989). An analysis of quartz occurrence patterns in airborne
coal mine dusts, 4th U.S. Mine Ventilation Symp., Berkeley, CA, pp 463-474.
Prandtl, L. (1923). Ergebnisse der aerodynamischen versuchtsanstalt zu Gottingen; Oldenbourg, Munich
and Berlin, p 29.
Qin, J. and Ramani, R.V. (1989). Generation and entrainment of coal dust in underground mines, 4th
U.S. Ventilation Symp., Berkeley, CA, pp 454-462.
Ramani, R.V. et al (1988). On the relationship between quartz in the coal seam and quartz in the
respirable airborne coal dust, 4th Int. Mine ventilation Congress, Brisbane, pp 519-526.
Ramani, R.V. and Bhaskar, R. (1984). Dust transport in mine airways, Proc. Coal Mine Dust
Conference, W. Virginia University (ed. Peng), pp 198-205.
Sandys, M.P.J., and Quilliam, J.H. (1982). Sources and methods of dust control, Chap. 15,
Environmental Engineering in South African Mines, Mine Ventilation Society of S. Africa.
Sartaine, J.J. (1985). The use of water-powered scrubbers on NMS Marietta drum miners, 2nd U.S. Mine
Ventilation Symp., Reno, Nevada, pp 733-740.
Schrder, H.H.E., Runggas, F.M. and Krss, J.A.L. (1984). Characterization of sonically atomized
water-spray plumes, 3rd Int. Mine Ventilation Congress, Harrogate, U.K., pp 219-228.
Seinfeld, J.H. (1986). Atmospheric chemistry and physics of air pollution, Wiley, New York.
Stevenson, J.W. (1985). An operator's experience using antitropal and homotropal longwall face
ventilation systems, 2nd U.S. Mine Ventilation Symp., Reno, Nevada, pp 551-557.
Stricklin, J.H. (1987). Longwall dust control at Jim Walters Resources, 3rd U.S. Mine Ventilation Symp.,
Penn. State, pp 558-563.
Taylor, C.D., Kovscek, P.D. and Thimons, E.D. (1988). Dust control on longwall shearers using water
jet-assisted cutting, 4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 547-553.
Tomb, T.F. et al (1990). Evaluation of longwall dust control on longwall mining operations, SME Annual
Meeting, Salt Lake City, Utah, Feb., pp 1-10.
Wang, Y.P. et al (1991). Use of surfactants for dust control in mines, 5th U.S. Mine Ventilation Symp., W.
Virginia, pp 263-270.
20 -
40
Part 5
Fires and Explosions
Subsurface fires and explosions
CHAPTER 21.
SUBSURFACE FIRES AND EXPLOSIONS
21.1 INTRODUCTION
21.1.1. The fire triangle and the combustion process

21.1.2. Classification of mine fires
3
4
21.2 CAUSES OF IGNITIONS

21.2.2. Electrical apparatus
21.2.3. Conveyors
21.2.4. Other frictional ignitions
21.2.5. Explosives
21.2.6. Welding
21.2.7. Smoking and flame safety lamps
4
5
5
6
7
7
7
21.3. OPEN FIRES
21.3.1. Oxygen-rich and fuel-rich fires

21.3.2. Effects of fires on ventilation
21.3.2.1. The choke effect
21.3.2.2. The buoyancy (natural draft) effect
21.3.3. Methods of fighting open fires
21.3.3.1. Firefighting with water
21.3.3.2. High expansion foam
21.3.4. Control by ventilation
21.3.4.1. Pressure control
21.3.4.2. Airflow reversal
8
9
9
10
11
12
13
14
14
15
21.4. SPONTANEOUS COMBUSTION
17
21.4.1. The mechanisms of spontaneous combustion in minerals

21.4.1.1. The phases of oxidation
21.4.1.2. The effects of water vapour
21.4.1.3. The path of a spontaneous heating.
21.4.2. Susceptibility to spontaneous combustion
21.4.3. Precautions against spontaneous combustion
21.4.4. Detection of a spontaneous heating
21.4.5. Dealing with a spontaneous heating
21.4.5.1. Excavating the fire
21.4.5.2. Burying the fire
21.4.5.3. Sealants
21.4.5.4. Localized pressure balancing
21.4.5.5. Flooding and sealing off
17
17
18
19
19
20
23
24
24
24
25
25
26
21.5. STOPPINGS, SEALS AND SECTION PRESSURE BALANCES FOR

EMERGENCY SITUATIONS
27
21.5.1. Site selection of seals

21.5.2. Sequence of building seals
21.5.3. Construction of seals and stoppings
21.5.4. Re-opening a sealed area
21.5.5. Section pressure balances
28
28
28
30
31
21.6. THE USE OF INERT GASES
32
21.6.1. Carbon dioxide

21.6.2. Combustion gases
21.6.3. Nitrogen
21.6.4. Methods of application and control
32
33
33
33
21 -
21.7. FIRE GASES AND THEIR INTERPRETATION
35
21.7.1. The processes of burning and the gases produced

21.7.2. The detection and trend analysis of fire gases
21.7.3. Explosibility diagrams
35
36
39
21.8. EXPLOSIONS
45
21.8.1. Initiation of explosions

21.8.2. Mechanisms of explosions
21.8.2.1. Gas explosions
21.8.2.2. Coal dust explosions
21.8.2.3. Sulphide dust explosions
21.8.2.4. Dust explosibility tests
21.8.3. Suppression of mine explosions
21.8.3.1. Stonedust and water barriers
21.8.3.2. Triggered barriers and explosion detectors
21.8.4 Explosions in sealed areas.
21.8.4.1 Causes of explosions in abandoned areas.
21.8.4.2 Recognition of the hazard.
21.8.4.3 Approaches to the problem and the initial NIOSH Report.
21.8.4.4 Requirements of seals
45
46
48
48
54
54
54
54
55
56
56
56
56
58
21.9 PROTECTION OF PERSONNEL
61
21.9.1. Training and preparedness

21.9.2. Methods of warning and locating personnel
21.9.2.1. Stench warning systems
21.9.2.2. Ultra-low frequency radio signals.
21.9.2.3. More recent systems
21.9.2.4. Locating trapped persons
21.9.3. Self-rescuers
21.9.3.1. Filter self-rescuers (FSR's)
21.9.3.2. Self-contained self-rescuers (SCSR's)
21.9.4. Escapeways
21.9.5. Refuge chambers
61
62
62
63
63
64
64
65
65
66
68
21.10 EMERGENCY PROCEDURE AND DISASTER MANAGEMENT
68
21.10.1. Immediate response

21.10.2. Command centre
21.10.3. Disaster management
69
69
69
References
70
21.1 INTRODUCTION
The most feared of hazards in underground mines or other subsurface facilities are those of fires and
explosions. Like airplane crashes, these do not occur often but, when they do, have the potential of
causing disastrous loss of life and property as well as a temporary or permanent sterilization of mineral
reserves. Furthermore, "near-misses" occur all too frequently. The incidence of mine fires appears not to
be declining despite greatly improved methods of mine environmental design and hazard control. This is a
consequence of several matters; first the growing variety of materials that are imported into modern mine
workings, varying from resins and plastics to liquid fuels and hydraulic fluids. A second factor is the
continuous increase in the employment of mechanized procedures, many of the machines involving
flammable liquids and materials that can produce toxic fumes when over-heated.
21 -
Although this chapter concentrates on technical considerations the incidence of mine disasters is greatly
influenced by the priority afforded to safety and health by individual companies, states and countries. Mine
disasters are more prevalent at times when market prices for minerals and fuels are high. Attention paid to
mine safety is greatest in those countries that have active and well-funded research activities in this area
and which maintain industry awareness through an ongoing output of reports and papers. Such work also
assists in the promulgation of new or updated legislation.
These activities are often inter-related. A period of declining statistics of mine fatalities coupled with a
period of limited research tends to create a laissez-faire attitude in both industry and legislative bodies or
agencies. A boom market is then more likely to produce a spate of serious mine accidents. Public disquiet
results in (often hasty) legislative action to promote new research investigations and the imposition of
stricter standards and penalties on mining companies. Unfortunately, public memory tends to be short and
history has shown the repetitive nature of this cycle. It is within that context that this chapter has been
written.
The enormous loss of life due to mine fires and explosions during the eighteenth and nineteenth centuries
preoccupied the minds of mining engineers and scientists of the time (Chapter 1). The majority of deaths
arising from mine fires and explosions are caused, not by burning or blast effects, but by the inhalation of
toxic gases, in particular, carbon monoxide. There are two major differences between underground fires
and those that occur in surface structures. The first concerns the long distances, often several kilometres,
that personnel might be required to travel in passageways that may be smoke-filled. Secondly, the
ventilation routes are bounded by the confines of the airways and workings, causing closely coupled
interactions between the ventilation and behaviour of the fire.
It is difficult for anyone who has not had the experience, to comprehend the sensations of complete
isolation and disorientation involved in feeling one's way through a long smoke-filled mine airway in zero
visibility. It is a cogent exercise to turn off one's caplamp in an unilluminated return airway and to walk just
a few steps, even without the trauma of a highly polluted atmosphere.
It is, therefore, a matter of ongoing importance that all personnel involved in the design and operation of
underground openings should have some knowledge pertaining to the prevention and detection of
subsurface fires and explosions, as well as procedures of personnel warning systems, escapeways,
firefighting, toxic gases, training, fire drills and the vital need for prompt response to an emergency
situation. These are some of the topics that are discussed in this chapter.
21.1.1. The fire triangle and the combustion
process
Perhaps the most basic precept in firefighter
training is the fire triangle shown on Figure
21.1. This illustrates that the combustion
process which we term "fire" requires three
components: fuel, heat and oxygen.
Remove any one of these and the fire will be
extinguished. The fuel may be solids, liquids
or gases. The liquids and gases might be
introduced into the mine environment by
natural or mining processes, or may be
produced by heating solid materials.
Whenever a combustible solid or liquid is
heated to a sufficiently high temperature
(flashpoint), it will produce a vapour that is
capable of being ignited by a flame, spark or
hot surface which has the required
21 Figure 21.1 The fire triangle.
concentration and duration of thermal energy. Gasoline has a flashpoint of -45C while most commonly
available solids require the application of a flame for them to reach flashpoint. The ignition temperature of
any given substance is the lowest temperature at which sustained combustion is initiated. Flaming is the
process of rapid oxidation of the vapours accompanied, usually, by the emission of heat and light. In the
case of self-sustained burning, that heat is sufficient to raise the temperature of the newly exposed or
surrounding areas of surface to flashpoint. However, combustion can continue at a slower rate without
flaming through the process we know as smouldering. In this case, the oxidation process continues on the
surface of the material and produces sufficient heat to be self sustaining, but not enough to cause the
emission of vapours in the quantity required for flaming combustion.
The oxygen which forms the third side of the fire triangle is normally provided by the air. Flammable liquids
such as the oil of a flame safety lamp will cease to burn when the oxygen content of the air is reduced to
some 16 per cent (Section 11.2.2.). Flaming combustion of all kinds is extinguished at oxygen contents of
10 to 12 per cent while smouldering is usually terminated at oxygen concentrations below 2 per cent.
However, some materials may contain sufficient inherent oxygen for slow combustion to continue at even
further reduced levels of atmospheric oxygen. Coupled with the low values of thermal conductivity of
crushed material, this can result in "hot spots" lying dormant in abandoned areas for long periods of time,
but capable of re-ignition if a renewed air supply is admitted subsequently.
21.1.2. Classification of mine fires
Fires underground can be classified into two broad groups, open and concealed fires. Open fires occur in
airways, faces and other openings that form part of the active ventilation system of the mine and, hence,
affect the quality of the mine airflows quickly and directly. As the term might imply, open fires are often
accompanied by flaming combustion because of the availability of oxygen and offer the possibility of direct
attack by firefighting teams. Conversely, concealed fires occur in areas that are difficult or impossible to
access such as caved or abandoned zones. These are usually, but not necessarily, initiated as a result of
spontaneous combustion and can occur in both coal and sulphide ore minerals as well as within any
imported organic matter such as paper, discarded fabrics (e.g. oily rags) or timbering in abandoned areas.
The degree to which concealed fires propagate and pollute the mine atmosphere depends upon the rate at
which air leaks through the areas affected. The matter of spontaneous combustion is discussed in further
detail in Section 21.4.
21.2 CAUSES OF IGNITIONS

The variety of procedures, processes and materials used in modern mining provides many opportunities
for the ignition of flammable materials. However, the most commonly reported causes of fires and
explosions in mines are listed in the following subsections.
Machines intended for use underground should be designed to operate with a high degree of safety in a
harsh physical environment, and are subject to legal requirements and conditions in most mining
countries. It is no surprise, therefore, that the majority of fires attributable to machines arise out of:
misuse
lack of proper maintenance
removal or bypassing of safety features such as diagnostic devices, environmental monitors or
thermal trip switches and
running unattended for long periods of time.
21 -
Exhaust systems on diesel equipment should be fitted with partculate filters or water scrubbers that not
only reduce airborne pollution (Section 11.3.2.), but also prevent the emission of incandescent particles.
Furthermore, hoses, transmission or brake fluids and a variety of components made from synthetic
materials on modern diesels may be capable of producing toxic gases when ignited. All vehicles or other
diesel equipment should be fitted with on-board fire extinguishers.
It is particularly important that equipment containing significant quantities of oil, such as large transformers
or air compressors, should be safeguarded by thermal trips, pressure relief valves and other devices
necessary for automatic cut-off in the event of any abnormal condition. Such devices should be subjected
to routine testing and maintenance. Wherever possible (and as may be required by law) non-mobile
equipment should be located within enclosures with fire-resistant roof, floor and walls, and which are
ventilated to a return airway. Again, fire extinguishers and, preferably, an automatic fire suppression
system should be available within the chamber. In coal mines, the surroundings in adjacent airways should
routinely be coated with stonedust (Section 21.8.3.1.).
21.2.2. Electrical apparatus
In addition to the general comments on mechanized equipment made in the preceding subsection,
electrical gear can give rise to incendiary hazards from sparking and overheating. Switchgear or battery
charging stations should be sited such that they are not affected by convergence or falls of roof. This is
most liable to occur close to mineral winning areas. Furthermore, start switches should be protected
against accidental operation by glancing blows from falling debris or passing traffic. Electrical sub-stations
and battery charging chambers should be equipped with non-aqueous fire extinguishers.
Cables in airways should be hung in catenary fashion on cradles suspended from the roof. They should be
located such that they will not be pinched by convergence or the yielding of roof supports, nor be impacted
by vehicles. The insulation and type of sheathing must be suitable for the electrical load and rigours of the
underground environment. All such cables should be inspected routinely for evidence of physical damage.
Electrical failures should result in immediate isolation of the power by means of overload and earth
leakage protective devices. In gassy mines, all electrical motors and heavy current devices should be
enclosed within flameproof casings so that any ignition of methane is contained within the equipment.
Signalling or other light current apparatus should be certified as intrinsically safe, i.e. incapable of
producing sparks of sufficient energy to ignite a methane:air mixture.
During non-working shifts, the electrical power supplied to each area of the mine should be isolated at the
appropriate highest level control centre or substation. Precautions should be taken against power surges
caused by lightning strikes on surface power lines, transformers, substations or any other lines that may
conduct the surge to underground excavations. Similarly, particular care should be taken against electrical
leakage in the vicinity of explosives or fuel storage areas.
21.2.3. Conveyors
Conveyor fires have been subjected to particular study because of the rapidity of fire propagation along
the early rubber-based types of conveyor belting. Modern conveyor belting for underground use must be
subjected to fire propagation tests (e.g. Verakis, 1991; Mutmansky et al, 2007). Three types of materials
are used for mine conveyor belts, namely, styrene-butadiene rubber (SBR), neoprene (NP) and polyvinyl
chloride (PVC). Composites of these materials are also employed. Following ignition of the belt material
and removal of the igniting source, the fire should preferably fail to propagate or, if it does, move at a slow
rate. However, it should be noted that heated belt material may produce hazardous fumes.
Numerous tests have indicated that fire propagation rates along conveyor belting are influenced by
airspeed (e.g. Hwang et al, 1991). At a relative velocity of some 1.5 m/s between the belt surface and the
21 -
adjacent airstream, a phenomenon known as flash over attains its maximum effect. Flash over occurs
when a flame front from the burning belt reaches forward over an unburned surface with an optimum angle
and length such that the radiant effect on that surface reaches a maximum. This can cause flaming of the
top layer of belting and a significant increase in flame propagation rate along the surface of the belt.
Deeper layers in the weave of the material may or may not be ignited. The effect has been observed in
belts of compositions including SBR, PVC, and SBRneoprene combination (Verakis and Dalzell, 1988).
Flashover involves a serious hazard as belt surface propagation rates may reach some 10 m/min. The
spread of fire along mine conveyors is influenced strongly by the turbulence of the airflow. Hence,
laboratory tests of small samples of belting can give misleading results. Large scale gallery tests are more
reliable.
Conveyor fires are most likely to be initiated by friction. If the belt becomes staked (immobilized) at any
point along its length and the drive rollers continue to turn, then high temperatures will be generated at the
drive head. Temperature monitors or belt tension transducers can sense this condition. Such devices
should be wired to isolate electrical power from the conveyor drive when an alarm condition is detected.
Similarly, a seized idler or return roller can become red-hot from the friction of a belt moving over or
around it. Conveyors should be patrolled regularly during operation in order to detect the development of
faulty rollers. Worn bearings will often be noisy and may also be detected by the smell of heated surfaces.
A further frictional hazard can occur if the conveyor becomes misaligned to the extent that the belt rubs
against surrounding surfaces such as the conveyor structure or airway sides.
In all of these cases, a fire may be initiated when lubricants, coal dust or flammable debris reach their
ignition points. It follows that dust or spillage should not be allowed to accumulate around and, particularly,
underneath conveyors. A clean conveyor road is much more likely to be a safe one. In coal mines,
conveyor entries should be well rock dusted
21.2.4. Other frictional ignitions
The main cause of methane ignitions on the working faces of coal mines is frictional sparking at the pick
points of coal winning machinery. This occurs particularly when the machine cuts through sandstone or
pyritic material. Two approaches have been taken to reduce this hazard. One is to ensure that there is
sufficient ventilation around the cutting drum to provide rapid dilution of the methane as soon as it is
emitted. It is, of course, important that the overall airflow at the working face is adequate to prevent
methane layering (Section 12.4.2.) and that the layout of the system minimizes flushes of methane from
worked-out areas (Section 4.3.2.). A number of devices have been employed to enhance air movement
across the pick points of shearers and continuous miners (e.g. Browning, 1988). Unfortunately, these may
exacerbate the dust problem unless combined with a wet scrubber (Section 20.4.2.2.).
The second approach to the incendiary streak of sparks that sometimes trails behind a cutter pick is to
quench it with water. This technique combines the suppression of both dust and methane ignitions. It is
achieved by pick face flushing and, even more efficiently, by jet assisted cutting (Section 20.4.1.1.).
Rope haulage systems have been the cause of some mine fires. Care should be taken that all pulleys and
return wheels are routinely serviced and lubricated. Ropes should not be allowed to rub against solid
surfaces such as the roof, sides or floor of airways and, particularly, timber supports. If haulage ropes
must pass through holes in stoppings then, again, the ropes should not contact the sides of the orifices.
Fluid couplings and enclosed gearings or direct drives are preferred to mechanical clutches, belts or Vdrives for the transmissions of mining machinery. However, where the latter are employed then, again,
regular inspections and maintenance are required to ensure their continued safe operation. Similarly,
mechanical braking systems should be well looked after.
21 -
21.2.5. Explosives
The initiation of fires from explosives or igniter cord remains a danger in non-gassy mines. Incandescent
particles from blasting operations may contain sufficient heat energy to ignite dry wood or combustible
waste material. Igniter cord should never be hung on timber supports. A strict record should be maintained
on all explosives and detonating devices at the times of issue and return to the stores. The relevant
national or state legislation should be consulted for the conditions under which explosives may be stored
or transported underground.
21.2.6. Welding
All welding operations that are permitted underground should be carried out under well-controlled
conditions. Where there is any possibility of methane or other flammable gases being present then testing
for those gases should be carried out before and, at intervals, during the welding operations. Hot slag and
sparks from welding are easily capable of igniting combustible materials such as coal, wood, paper and
waste rags. Wherever possible, such materials should be removed from the vicinity of welding operations
and the remainder wetted down or coated by stonedust. Molten metal should not be allowed to drop on the
floor. Slag pans should be used to capture hot run-off. This is particularly important in coal mines, in shafts
and near timber supports. Fire extinguishers must be available at the sites of all welding operations.
Gas containers employed in oxy-acetylene cutting should be stored and used in a secure upright position.
Gas bottles must never be stored or used in the vicinity of explosives or concentrations of flammable
liquids.
21.2.7. Smoking and flame safety lamps
It is a sad fact that the use of smoking materials has been suspected as the cause of some fires and
explosions in mines. In those mines that have been classified as gassy, carrying such materials (often
known as contraband) into the subsurface is illegal. This law should be enforced with the utmost rigour.
Through well chosen examples during training and refresher classes a workforce will, themselves, ensure
compliance with non-smoking regulations.
In subsurface openings where smoking is permitted then, again, education, posters and warning signs
should be employed as ongoing reminders of the possible disastrous consequences of careless disposal
of smoking materials.
Damaged flame safety lamps have also been suspected of igniting a methane:air mixture. In the few
places where these devices remain in use, they should be treated with care and subjected to inspection
after each shift. When a high concentration of methane is detected by a blue flame spiralling rapidly within
a flame safety lamp (Section 11.4.2.2.) then the lamp should be lowered gently and, if necessary,
smothered inside one's clothing. Familiarity with the procedure should be gained through training and will
counter the natural reaction of the untrained person to drop the lamp or to throw it away in panic. (Section
11.4.2.2)
21.3. OPEN FIRES

Fires that occur in mine airways usually commence from a single point of ignition. The initial fire is often
quite small and, indeed, most fires are extinguished rapidly by prompt local action. Speed is of the
essence. An energetic ignition that remains undetected, even for only a few minutes, can develop into a
conflagration that becomes difficult or impossible to deal with. Sealing off the district or mine may then
become inevitable.
21 -
The rate at which an open fire develops depends, initially, upon the heat produced from the igniting
source. A fine spray of burning oil from a damaged air compressor can be like a flame thrower and ignite
nearby combustibles within seconds. On the other hand, an earth leakage from a faulty cable may cause
several hours of smouldering before flames appear. The further propagation of the fire depends upon the
availability of fuel and oxygen (Figure 21.1). A machine fire in an untimbered metal mine airway will remain
localized if there is little else to burn in the vicinity. Conversely, an airway that is heavily timbered or with
coal surfaces in the roof, floor or sides will provide a ready path for speedy development and propagation
of a fire.
When an open fire has developed to the extent of causing a measurable change in the temperature of the
airflow then it can affect the magnitudes and distributions of flow within the mine ventilation system.
Conversely, the availability of oxygen to the fire site controls the development of the fire. This Section
discusses the coupled interaction between fire propagation and ventilation, and the means by which open
fires in mines may be fought.
21.3.1. Oxygen-rich and fuel-rich fires
At the start of most open fires in ventilated areas, there is a plentiful supply of oxygen - more than
sufficient for combustion of the burning material. Indeed, if the air velocity is brisk then heat may be
removed at a rate greater than that at which it is produced. The heat side of the fire triangle is removed
and the fire is "blown out". These are examples of oxygen-rich fires. Assuming that the fire continues to
proliferate, it will consume increasing amounts of oxygen and, at the same time, produce greater volumes
of distilled gases and vapours. The point may be reached when the heat of combustion produces
temperatures that continue to remain high enough to distill gases and vapours from the coal, timber or
other available fuels but with insufficient oxygen to burn those gases and vapours completely. The fire has
then become fuel-rich.
The development of an oxygen-rich into a fuel-rich fire is a serious progression and produces a much
more dangerous situation for firefighters. When flammable gases at temperatures exceeding their ignition
point meet relatively fresh air then they will ignite along the gas:air interfaces. The added turbulence may
produce intimate mixing of air and unburned gases resulting in explosions. These phenomena can occur
downstream from an open fire if air leaks into the firepath from adjacent airways. Firefighters are then
faced with a difficult decision. Leakage of air from adjacent airways must be into the firepath in order to
prevent spread of the fire into those adjacent airways, yet the admittance of that air may cause explosions
and propagation of the fire at a rate much greater than that allowed by burning of the solid material itself.
A similar effect occurs when buoyancy of the hot gases causes roll-back of smoke at roof level against the
ventilating current (Section 21.3.2.2.). This can occur over the heads of workers who are fighting the fire
from an upstream position. Again, burning of the gases along the air interface can occur, igniting coal or
timber in the roof and producing the danger of explosion. Personnel involved in fighting a fuel-rich fire may
become aware of pressure pulses or rapid fluctuations in the movement of the air. These are caused by
rolling flames and "soft" explosions as gases ignite along gas:air mixing zones. The same phenomena can
be observed following ignitions of methane (Section 1.2.). Such pulsations may be a precursor to a larger
and more violent explosion.
It follows that every attempt should be made to prevent an oxygen-rich fire from developing into a fuel-rich
fire. This underlines the need for early detection and prompt action. An intuitive reaction to a fire may be to
restrict the air supply and, hence, remove the oxygen leg of the fire triangle. This can be accomplished by
building stoppings or erecting brattice cloths upstream from an airway fire. However, consideration of the
dangers inherent in fuel-rich fires indicates that restricting the airflow might be inadvisable. Analyses of
gases downstream from fires can be interpreted to indicate whether a fire is oxygen-rich or fuel-rich
(Section 21.7).
21 -
21.3.2. Effects of fires on ventilation

An open fire causes a sharp increase in the temperature of the air. The resulting expansion of the air
produces two distinct effects. First the expansion attempts to take place in both directions along the
airway. The tendency to expand against the prevailing direction produces a reduction in the airflow. This is
known as the choke or throttle effect. Secondly, the decreased density results in the heated air becoming
more buoyant causing local effects as well as changes in the magnitudes of natural ventilating energy.
21.3.2.1. The choke effect
Consider an airway before it is affected by a fire. Air flows along it at a mass flowrate of M kg/s and doing
work against friction at a rate of F J/kg. The airpower dissipated against friction, Pow, is the product of the
two
J kg
kg s
Pow = F M
or
Watts
(21.1)
The effect of a fire in the airway upon Pow depends upon the reactions of fans, natural ventilating
pressures and ventilation controls throughout the system. However, if no deliberate action is taken to
change these factors, it is reasonable to estimate that it remains sensibly constant.
Equation (7.10) gave us
p
F =
= Rt Q 2
J
kg
where p
and
Rt
Q
=
=
=
=
(21.2)
frictional pressure drop (Pa)

mean density of air (kg/m3)
rational turbulent resistance of the airway (m-4 )
mean value of airflow (m3/s)
Combining equations (21.1) and 21.2) gives

Pow = F M = R t MQ 2
or
Pow = R t
as
Q = M
M3
(21.3)
Equation (21.3) may be written as

P
M = ow
Rt
13
kg
s
(21.4)
As Pow and Rt are constants,

M
(21.5)
where means 'proportional to'

Hence, as the fire causes the density of the air to decrease, the mass flow of air will also decrease for the
same energy dissipation. This phenomenon produces the choke effect. It should be noted, however, that
the volume flow exiting the airway has increased.
As M = Q , proportionality (21.5) can be written as
21 -
m3
s
1
3
(21.6)
Note, also, from Pow = FM = constant, that as M decreases, the work done against friction per kg of air, F,
must increase - a result of the increased volume flow and, hence, turbulence.
The choke effect is analogous to increasing the resistance of the airway. For the purposes of ventilation
network analyses based on a standard value of air density, the raised value of this "pseudo resistance",
Rt', can be estimated in terms of the air temperature as follows:
R t = Pow
M3
m -4
Hence, for any standard (fixed) value of density and constant air power loss,
1
Rt '
M3
But combining with proportionality (21.5) which represents the actual reduction in mass flow,
Rt '
For a given barometric pressure, the general gas law (Section 3.3.1.) gives
1
T
where T = absolute temperature (K)
Hence,
Rt ' T 2
(21.7)
The value of the "pseudo-resistance" Rt', increases with the square of the absolute temperature. However,
it should be recalled that this somewhat artificial device is required only to represent the choke effect in an
incompressible flow analysis.
Litton et al (1987) have also produced an estimate of the increased resistance in terms of the carbon
dioxide evolved from a fire.
21.3.2.2. The buoyancy (natural draft) effect
The most immediate effect of heat on the ventilating air stream is a very local one. The reduced density
causes the mixture of hot air and products of combustion to rise and flow preferentially along the roof of
the airway. The pronounced buoyancy effect causes smoke and hot gases to form a layer along the roof
and, in a level or descentional airway, will back up against the direction of airflow. The layering effect can
be estimated using the method given in Section 12.4.2.
This phenomenon of roll-back creates considerable difficulties for firefighters upstream from the fire,
particularly if the conflagration has become fuel-rich. The roll-back is visually obvious because of the
smoke. However, it is likely to contain hidden but high concentrations of carbon monoxide. Furthermore,
the temperatures of the roll-back may initiate roof fires of any combustible material above the heads of
firefighters. The most critical danger is that tidal flames or a local explosion may occur throughout the rollback, engulfing firefighters in burning gases.
21 -
10
One method of reducing roll-back is to increase the airflow in the airway. This, however, will increase the
rate of propagation of the fire. Another method is to advance with hurdle cloths covering the lower 60 to 80
per cent of the airway (Section 12.4.2.). The increased air velocity at roof level will help to control the rollback and allow firefighters to approach closer to the fire. However, this technique may also cause the rollback gases to mix with the air and produce an explosive mixture on the forward side of the hurdle cloth.
Furthermore, the added resistance of the hurdle cloth might reduce the total airflow to the extent that a
fuel-rich situation is promoted. The behaviour of open fires is very sensitive to modifications to the airflow.
Hence, any such changes should be made slowly, in small increments, and the effects observed carefully.
A third method of combatting roll-back is to direct fog sprays towards the roof. In addition to wetting roof
material, the air induction effects of the sprays will assist in promoting airflow in the correct direction at roof
level.
A more widespread effect of reductions in air density is the influence they exert in shafts or inclined
airways. This was handled in detail under the name of natural ventilation in Section 8.3.1. The effect is
most pronounced when the fire itself is in the shaft or inclined airway, promoting airflow if the ventilation is
ascentional and opposing the flow in descentional airways. Indeed, in the latter case, the flow may be
reversed and can result in uncontrolled recirculation of toxic atmospheres.
If the air temperatures can be estimated for paths downstream of the fire then the methods given in
Section 8.3.1. may be employed to determine the modified natural ventilating pressures. Those
temperatures vary with respect to
size and intensity of the fire

distance from the fire
time
leakage of cool air into the airways affected and
heat transfer characteristics between the air and the surrounding strata.
At any given time, air temperatures tend to fall exponentially with respect to distance downstream from a
fire. Climatic simulation models (Chapter 16) may also be employed to track the time transient behaviour
of air temperatures downstream from a fire. However, in that case, two matters should be checked. One is
that the limits of application of the program may be exceeded for the high temperatures that are involved.
Secondly, the transient heat flux between the air and strata will be much quicker than for normal climatic
variations. Hence, the virgin rock temperature (VRT) in the simulation input should be replaced by a
"surrounding rock temperature" (SRT), this being an estimate of the mean temperature of the immediate
envelope of rock around the airway before the fire occurs.
Having determined air temperatures in all paths downstream from the fire, the revised natural ventilation
pressures for the mine can be determined. These may then be utilized in network analysis exercises to
predict the changes in flow and direction that will be caused by a fire of given thermal output. A number of
fire simulation packages have been developed to allow numerical modelling of mine fires (e.g. Trutwin et al
(1992). Greuer, (1984); Greuer, 1988; Dziurzynski et al, (1988); Deliac et al, (1985); Stefanov et al, (1984);
Wala (1998); Gillies et al (1995). Gillies A.D.S., et al (2004).
21.3.3. Methods of fighting open fires
The majority of open fires can be extinguished quickly if prompt action is taken. This underlines the
importance of fire detection systems, training, a well-designed firefighting system and the ready availability
of fully operational firefighting equipment. Fire extinguishers of an appropriate type should be available on
vehicles and on the upstream side of all zones of increased fire hazard. These include storage areas and
fixed locations of equipment such as electrical or compressor stations and conveyor gearheads.
21 -
11
Neither water nor foam should be used where electricity is involved until it is certain that the power has
been switched off. Fire extinguishers that employ carbon dioxide or dry powders are suitable for electrical
fires or those involving flammable liquids.
Deluge and sprinkler systems can be very effective in areas of fixed equipment, stores and over
conveyors. These should be activated by thermal sensors rather than smoke or gas detectors in order to
ensure that they are operated only when open combustion occurs in the near vicinity.
The two direct methods of firefighting introduced in this Section involve the application of water and high
expansion foam. The additional or complementary means of fire management by adjustment of ventilation
controls and the injection of an inert gas are discussed in Sections 21.3.4. and 21.6. respectively.
21.3.3.1. Firefighting with water
Except where electricity or flammable liquids are involved, water is the most common medium of
firefighting. When applied to a burning surface, water helps to remove two sides of the fire triangle. The
latent heat of the water as it vapourises and the subsequent thermal capacity of the water vapour assist in
removing heat from the burning material. Furthermore, the displacement of air by water vapour and the
liquid coating on cooler surfaces help to isolate oxygen from the fire.
Water is normally applied by hosepipes upstream from the fire. A difficulty in subsurface firefighting is the
limited reach of water jets imposed by the height of the airway. This underlines the vital need for water to
be available at adequate pressure and quantity in the firefighting range. In order for a water jet to reach
some 30 m in a typical coal mine entry, water pressures should be in the range 800 to 1400 kPa (Mitchell,
1990) and capable of supplying up to five hoses from a manifold connected to a single hydrant. In practice,
the range of water jets in mine airways may often be no greater than 10 m. The nozzles should
preferentially be of the adjustable type to give either a jet or a fog spray.
Hard won lessons indicate the need for careful forethought in designing a mine firefighting water network.
The air and the water should flow in the same direction so that firefighters do not become dependent on a
water supply that passes through the fire before it reaches them. Hydrants should be located at strategic
points with respect to areas of increased fire hazard, at intervals along airways and at cross-cuts with
access doors. All fittings for hydrants and range components should be standardized throughout any given
mine. Non-metallic caps should be used at hydrant outlets to minimize corrosion. However, these caps
must always be removable by hand and without undue force. Supplies at firefighting stations should be
inspected at set intervals to ensure their operational efficiency at all times. Range fittings should include
tee-pieces, blank-off caps and manifolds. It is particularly important that hosepipes be unrolled and
examined for deterioration on a planned maintenance schedule and that they should be stored according
to manufacturers' recommendations.
If access can be gained to an airway that runs parallel to a fire then fog sprays can be directed through
doors or holed stoppings into the path of the fire. This can be effective if the sprays are employed at an
early stage and immediately downstream from the fire front. (see Fig 21.2). However, for a large
conflagration or where the fire has become fuel-rich, it is likely to lose its effectiveness.
The locations of pumps and configuration of their power supplies should be considered carefully with
respect to the layout of the mine. The pumps and routes of their cables should be chosen such that they
are least likely to be disrupted by a fire. Dual power supplies via alternative routes may be considered.
Furthermore, power for firefighting pumps should be capable of being maintained when electricity to
working sections of the mine has to be isolated. Underground sumps can provide valuable water capacity.
However, the firefighting system should also allow water to be supplied in adequate quantities from
surface locations.
21 -
12
21.3.3.2. High expansion foam

Large volumes of water-based foam provide a valuable tool for fighting fires in enclosed spaces such as
the basements of buildings or in the holds of ships. It has been employed for mine fires since at least 1956
(Eisner). The method is employed on large fires and, although it has had somewhat limited success in
extinguishing mine fires, it can play a valuable role in cooling and quenching an area to an extent that
allows firefighters with hoses to approach closer to the firefront. Even when sealing an area has become
inevitable, valuable time for rescue operations can be bought by employing high expansion foam.
The bubbles are generated by a fan which blows air through a fabric net stretched across a diffuser. The
net is sprayed continuously with a mixture of water and foaming agent. Bubbles can be produced at a rate
of several cubic metres per second (Strang and MacKenzie-Wood, 1985). Compounds such as
ammonium lauryl sulphate may be employed as the foaming agent while the addition of
carboxymethylcellulose improves the stability of the bubbles (Grieg et al, 1975).
The objective is to form a plug of high expansion foam which fills the airway and is advanced on to the fire
by the ventilating pressure. The ratio of air to water within the foam may be in the range 100:1 to 1000:1.
As the foam advances, bubbles break around the perimeter of the airway when they touch a dry surface.
However, the liquid that is released wets that surface and allows advancement of the following bubbles.
Shrinkage of the foam occurs continuously at the leading edges and accelerates because of radiant
effects as it approaches the burning material.
Control of the combustion process is achieved by two primary mechanisms. First, vapourization of the
water removes heat from the site and, secondly, the increased concentration of water vapour may produce
an extinguishing atmosphere. As the air within the bubbles is heated to 100C it will expand by some 30
per cent. However, the vapourization of liquid water to a gas involves an expansion of about 1700:1.
Assuming an air:water mix in the foam of 1000:1 a thousand litres of air expands to 1300 litres while 1 litre
of water evaporates to become 1700 litres of water vapour giving 3000 litres of mixture. If the air originally
had an oxygen content of 21 per cent then the evaporation of water will reduce that to
1300
21
= 9.1 percent which will extinguish flaming combustion.
3000
Despite these mechanisms, high expansion foam does have some drawbacks. First, it may be quite
difficult to generate a foam plug that fills the airway completely. As the plug builds up, the air velocity will
increase through the narrowing channel between the plug and the roof, tending to maintain the gap open.
Judicious employment of brattice cloths may assist in forming a complete plug of foam. It is important to
control the path of the foam and, in multi-entry systems, this can be problematic. The natural direction of
movement of the foam is dictated by the ventilating pressure. Here again, brattice cloths or stoppings in
cross-cuts to adjacent parallel entries can assist in controlling the direction of the foam. Major obstructions
caused by roof falls are quite liable to occur during a large underground fire. A foam plug may not be able
to climb over such obstructions with the ventilating pressure available.
The greatest danger of foam plugs is that the reduction in airflow may promote a fuel-rich fire with the
attendant danger of explosion. Downstream gases should be monitored for the development of this
condition. Both increases and decreases in combustible gases have been reported in differing fires when
high expansion foam has been employed. The reduction in airflow will tend to raise the concentration of
combustible gases. However, as the inert mixture of air and water progresses downstream, condensation
of the water occurs, allowing the air fraction to increase and, hence, modifying the combustible gas
concentrations.
After the application of high expansion foam has been initiated, it is important to maintain it in operation
during fire fighting as intermittent production of foam can exacerbate the development of an explosive
atmosphere. This underlines the need for good training so that operators are familiar with the equipment
and procedure. Furthermore, care should be taken that sufficient supplies of foaming agent are available
before the operation is started (Timko et al, 1988).
21 -
13
21.3.4. Control by ventilation

When contemplating changes to airflows and applied pressure differentials during a fire emergency, there
are four types of effect that must be considered most carefully.
(a)The effect on the combustion process: The importance of avoiding the progression of an oxygen-rich
fire into a fuel-rich fire has already been stressed in Section 21.3.1.
(b) The effect on direction and rate of propagation of the fire: Every attempt should normally be made to
prevent an open fire from spreading into other airways. However, exceptions from this general rule may
become necessary to guide products of combustion away from trapped personnel. An example may be the
deliberate destruction of a stopping or air crossing to divert or short-circuit a fire path from an intake airway
into an adjacent return. Again, any modifications of the airflow passing through the fire zone must seek to
achieve a balance between speed of propagation and control of the combustion process.
(c) Effects on the distribution of products of combustion: This becomes a critical issue when personnel
have become trapped inby the fire, particularly if their exact whereabouts are unknown. However, any
steps that will improve atmospheric conditions in escapeways require to be investigated.
(d) Effects on airflow distributions in other parts of the mine: While the consequences of ventilation
changes in the zone affected by the fire are of immediate concern, the effects of such changes throughout
the rest of the mine should not be overlooked, particularly in a gassy mine or when personnel may still be
evacuating other areas.
If a computer model of the mine ventilation network has been maintained up to date then this will prove
invaluable in investigating the predicted effects of proposed changes to the ventilation system. With a
modern network analysis package (Section 7.4.), a personal computer or terminal in the emergency
control centre can produce such predictions within seconds. Nevertheless, the uncertainties inherent in a
fire situation demand that actual changes to the airflow system be made incrementally while observing the
reactions on distributions and gas concentrations. The following subsections discuss the practical
strategies that may be employed to control a fire by ventilation.
21.3.4.1. Pressure control
Airways that are parallel and adjacent to the fire path will remain unpolluted provided they are maintained
at a higher atmospheric pressure. These allow access for escape, firefighting, building or strengthening of
stoppings in cross-cuts, or to apply water sprays into the fire path. In multi-entry workings, control of such
pressure differentials can be achieved by the erection of brattice cloths in the adjacent airway as illustrated
in Figure 21.2. Even if the pressure differential in the desired direction is not completely achieved, the
reduced rate of toxic leakage may allow time for personnel to escape. If necessary, the brattice cloths may
be advanced pillar by pillar to remove smoke sequentially from the adjacent airway. Devices such as the
"parachute stopping" or "inflatable seal" have been developed to replace brattice cloths in such
circumstances. These can be erected quickly and give improved seals around the perimeter of the airway
(Kissell and Timko, 1991).
A consequence of this technique is that the airflow over the fire will be increased to an extent that
depends upon the configuration and resistances of the local airways. Pressure differentials between
airways can also be modified by the use of a temporary fan instead of a restriction in the adjacent airway.
In this case, airflow over the fire will be reduced. The location and pressure developed by the fan must be
selected with care in order to avoid recirculation of products of combustion. Where pressure differentials
are small, even the few Pascals developed by a free-standing auxiliary fan can induce the desired effect
(Section 4.4.3.).
21 -
14
fire
smoke
brattice
Figure 21.2 A brattice sheet in the adjoining airway clears smoke from that airway by
promoting leakage into the firepath.
21.3.4.2. Airflow reversal

Many mines operate under a legislative requirement that the airflow provided by main fans must be
capable of being reversed promptly. The background to such laws is the fear of a fire or other inundation
of airborne pollutants occurring within a downcast shaft or main intake airway. Noxious and, possibly,
flammable gases could then contaminate all, or most, of the ventilation system including working areas
and return escapeways. If the fire is detected at an early stage then pollution of the complete system may
be prevented by prompt reversal of the airflow. Even where contamination of the total network has
occurred, air reversal may allow clearance of pollutants from return airways to the extent that a fresh air
route may be established between surface and refuge chambers where personnel may be trapped. The
decision to reverse a complete mine ventilation system is fraught with difficulties and has very seldom
been taken in practice.
There are essentially three methods of achieving the reversal of airflow in a ventilation system. Where
axial impellers are used on the main fans, then changing the direction of rotation will cause reversal of the
airflow. This can be implemented electrically at the fan motor. However, axial fans operate efficiently in
one direction only. The aerofoil section of each impeller blade is designed to give aerodynamic stability of
flow through the fan. When operating in reverse, breakaway of the boundary layers over the blades
occurs, resulting in high shock losses. The "lift" of the blades and, hence, the throughflow of air is greatly
reduced (Section 10.3.2.). Similarly, fixed guide vanes, fan casings and evasees, all designed for a
forward direction, will produce high shock losses when the airflow is reversed. The reversed air quantity
may be reduced to less than 50 per cent of the normal forward flow (Dunn, 1982).
In the case of centrifugal fans, airflow reversal can be achieved only by means of reversal doors. The flow
direction through the fan itself remains unchanged. For an exhausting centrifugal fan located at the mine
surface, hydraulic or pneumatic activation of the reversal doors opens the fan inlet to the outside
atmosphere and, simultaneously, diverts the fan exhaust into the mine shaft or slope. The opposite occurs
for a forcing centrifugal fan. Where reversal doors are fitted as part of a surface fan installation, their
operation should be checked routinely as part of a planned maintenance procedure.
Although the flow direction through a centrifugal fan remains unchanged, the shock losses incurred when
air reversal doors are activated results in a reduced flow. The amount of the reduction is site specific and
depends entirely upon the design and siting of the reversal doors, and the configuration of the fan with
respect to the inlet and outlet duct arrangements. The possibility of requiring air reversal should be
considered when designing the layout of airways and ventilation doors around an underground main fan.
Such reversal should be attainable rapidly by opening or closing those doors.
During the course of ventilation network planning exercises (Chapter 9), it is often possible to design
systems that allow rapid reversal of airflow in one section of the mine, or in a single airway, without total
reversal at the main fans. This may be achieved by the strategic location of doors that can be opened or
closed to permit airflow in either direction. Means of such local reversal might be considered, for example,
for a conveyor route that is also to serve as an intake during normal operations.
21 -
15
Mandating the provision of air reversal facilities for an underground mine appears to be a reasonable
safeguard. However, except in a clear-cut case such as a fire in or very close to a downcast shaft, the
potential risks associated with reversing the airflow may be greater than those of maintaining the normal
direction of flow, particularly in the short time period that may be available for making critical decisions.
The reasons that mine managements have very rarely decided to reverse ventilation during an emergency
are both practical and also because of possible litigation should lives be lost as a consequence of the
reversal. During the trauma of a major emergency involving changing conditions in air quality and possible
disruptions of ventilation structures and communications, it may be impossible to know with certainty the
locations, movements and dispersal of the workforce. Reversal of the airflow could then result in smoke
and toxic gases being drawn over personnel who had assembled in a previously unpolluted zone. It may
be expected that people who work routinely in a section of the mine will be familiar with the local
ventilation system and, in case of an emergency, will act in accordance with that knowledge. Reversing
the airflow could create additional uncertainty and confusion in their actions.
The majority of doors in airlocks or access paths between intakes and returns are designed to be self
closing, assisted by the mine ventilating pressure. This may be a legislative requirement. In the event of
airflow reversal, those doors will be blown open and create short circuits unless they are provided with
self-locking devices. Even in the latter situation or where powered doors are employed, the proportion of
air leakage must be expected to increase when the pressure differential across the door reverses. Hence,
coupled with the diminution in overall flow caused by the reversal procedures, the reversed ventilation
reaching the working areas must be expected to be much lower than the normal forward flow.
High temperatures usually preclude firefighting rescue teams from approaching a fire from the downstream
side. If the airflow is reversed over a fire in an intake then firefighting teams must transport their equipment
and materials to a fresh air base inby the fire.
The expansion of gases held in old workings or other voidage, and resulting from a drop in barometric
pressure, is discussed in Section 4.2.2. In the case of a forcing fan being reversed to create an exhausting
system, the rapid fall in barometric pressure throughout the system may cause large emissions of voidage
gas. If these contain high concentrations of methane, passing it into the fire zone could result in a series of
explosions propagating far back into the mine. Another possibility is that during the actual process of
reversal, flammable gases from the strata or produced by incomplete combustion or volatilization of
hydrocarbons may be drawn back over the fire, again, leading to the possibility of explosions.
Although there may be several hundred tonnes of moving air in a major subsurface structure, the braking
effects of viscous shear and turbulence causes it to be a well-damped system. Hence, when a main fan
stops, the effect is noticeable almost immediately at all places underground. In the majority of cases,
natural (thermal) ventilating effects will maintain movement in the normal direction.
In the situation of forced reversal, the transient effects will exist for much longer time periods than for a
simple stoppage of a main fan. This is because redistributions of the natural ventilating effects will not be
completed until a new equilibrium of heat transfer has been established between the strata and the airflow.
This may take several hours or even days. Furthermore, the fire itself will create thermally induced airflows
with, perhaps, local reversals and recirculations. There is, therefore, some uncertainty concerning the
speed at which reversal can be attained throughout the system and the stability of the reversed airflows.
Research in Poland has indicated the efficacy of accelerating reversal by the use of intensive water sprays
directed into the top of an upcast shaft (Trutwin, 1975).
21 -
16
21.4. SPONTANEOUS COMBUSTION

When air is allowed to percolate through many organic materials including coal then there will be a
measurable rise in temperature. The same phenomenon can be observed in crushed sulphide ores and is
caused by a progressive series of adsorptive, absorptive and chemical processes. These produce heat
and an observable elevation in temperature. The percolating airflow will, therefore, remove that heat
increasingly as the temperature of the material rises. If the leakage airflow is sufficiently high then a
balanced equilibrium will be reached at which the rate of heat removal is equal to the rate at which heat is
produced; the temperature will stabilize. The process will also reach an air-constrained equilibrium if the
airflow is sufficiently low to inhibit the oxidation processes. However, between these two limits there is a
dangerous range of percolating airflows that will encourage spontaneous heating.
Each material that is liable to spontaneous combustion has a critical temperature known as the minimum
self-heating temperature (SHT). This is the lowest temperature that will produce a sustained exothermic
reaction or thermal runaway. Hence, if the temperature reaches the SHT before thermal equilibrium is
attained then the oxidation process will accelerate. The temperature will escalate rapidly, encouraging
even higher rates of oxidation until the material becomes incandescent. At this stage, smoke and gaseous
products of combustion appear in the subsurface ventilation system. The mine then has a concealed fire.
The primary dangers of such occurrences are the evolution of carbon monoxide, the ignition of methane
and combustion progressing into airways to produce open fires.
The phenomenon of spontaneous combustion has been recognized since at least the seventeenth century
(PD-NCB, 1978). An early theory postulated that oxidation of pyritic material within coal provided centres
of enhanced activity. Bacterial action is a factor in the initial natural heating of hay and other foodstocks.
This can play a part in the spontaneous combustion of timber or organic waste material underground but is
unlikely to contribute significantly to the self-heating of coal or other minerals. Similarly, while increases in
the temperature of materials can be observed in ore passes (Section 15.3.6.), the gravitational energy of
collapsing waste areas in mines produces temperature rises that are insufficient to promote spontaneous
combustion. Current concepts of the initiation of self-heatings are discussed in the following subsection.
21.4.1. The mechanisms of spontaneous combustion in minerals
Although spontaneous combustion can occur in crushed or caved sulphide minerals and in heavily
timbered areas within metal mines, the problem is most common in coal mines. Research in this area has
concentrated on the spontaneous combustion of coals.
The development of self-heating requires a large surface area of crushed material combined with a slow
migration of air through that material. Hence, the problem arises in goaf (gob) areas, caved zones,
crushed pillar edges, fractured coal bands in roof or floor strata, stockpiles and tips on surface, and within
abandoned sections of mines. The progressive stages of spontaneous combustion appear to be complex
and not yet fully understood. Here, we shall examine the effects of oxygen, elapsed time and water.
21.4.1.1. The phases of oxidation
The oxidation processes of coal occur in four stages (Banerjee, 1985).
(a) Physical adsorption (Section 12.2.1.) of oxygen on coal commences at a temperature of
about -80C and is reversible but diminishes rapidly as the temperature increases to
become negligible beyond 30 to 50C. The process of adsorption produces heat as a byproduct of the modified surface energy of the material. This causes the initial rise in
temperature.
(b) Chemical absorption (known also as chemisorption or activated sorption) becomes
significant at about 5C. This progressively causes the formation of unstable compounds of
hydrocarbons and oxygen known as peroxy-complexes.
21 -
17
(c) At a temperature which appears to approximate the self-heating temperature (SHT) of the
coal, the peroxy-complexes decompose at an accelerating rate to provide additional oxygen
for the further stages of oxidation. This occurs within the range of approximately 50 to 120C
with a typical value of 70C. At higher temperatures, the peroxy-complexes decompose at a
greater rate than they are formed (Chakravorty, 1960) and the gaseous products of chemical
reaction appear - in particular, carbon monoxide, carbon dioxide, water vapour, and the
oxalic acids, aromatic acids and unsaturated hydrocarbons that give the characteristic odour
of "gobstink" (Section 21.4.4.).
(d) When the temperature exceeds some 150C, the combustion process accelerates rapidly.
Incineration of the coal occurs with escalating emissions of the gaseous products of
combustion.
The rate at which oxygen is consumed
varies both with time and the phase of
oxidation as illustrated in Figure 21.3.
Oxygen is taken up rapidly in the earlier
stages of chemisorption and as the
pyroxy-complexes are formed.
However, this reduces with time and as
the surfaces become coated
(weathered) with those oxygen
compounds. The temperature curve
tends to level off and may reach
equilibrium. However, if the SHT is
reached, as illustrated in Figure 21.3,
then both the rate of oxygen
consumption and the temperature
escalate rapidly.
14
12
10
8
6
4
2
0
100
80
60
40
20
T I M E
Figure 21.3 Examples of changes in oxidation rate and
temperature with respect to time (developed from
Muzyczuk, ref. Banerjee, 1985).
The time scale is omitted as this can vary from a few
hours to many days depending upon the type and
fineness of the coal, and the flowrate and psychrometric
condition of the air.
21.4.1.2. The effects of water vapour

There are two processes involving water that
act in opposite directions. First, the moisture
content of the coal is driven off by evaporation
during the early stages of heating. Hence,
some of the heat is removed in the water
vapour as latent heat of evaporation, tending
to inhibit the temperature rise of the coal. The
second process involves adsorption of water
vapour from the air by the coal (Hodges and
Hinsley, 1964). The heat of adsorption
(sometimes called heat of wetting) produces
an increase in the temperature of the material.
It follows that the net effect depends on which
of the two processes is dominant. Coal which
is completely saturated with water is unlikely
to increase in temperature due to heat of
wetting. However, a dry crushed coal that is
infiltrated by moist air can exhibit an initial
rapid rise in temperature from this
phenomenon.
21 -
18
Adsorption of water vapour adds considerably to the early stages of spontaneous heating of coal. It is
significant that coal mines in dry climates tend to be less troubled by spontaneous combustion. However, it
has been observed that re-ignition of coal frequently occurs on re-opening a mining area that has been
flooded to extinguish a fire. This is thought to occur for two reasons. First, the flooding and subsequent
drainage may produce further disintegration of the coal and the creation of new surfaces; and, secondly, it
is probable that many points of higher elevation or entrapped gases may not be fully flooded and remain at
a sufficiently high temperature (hot spots) to inhibit adsorption. When cooler air is admitted subsequently it
will pick up water vapour as it progresses through the wet conditions. An initial short-lived cooling of the
hot spots may be followed by rapid adsorption of water vapour and a renewed escalation of temperature.
21.4.1.3. The path of a spontaneous heating.
The processes of oxidation and adsorption do not occur uniformly throughout a mass of crushed
combustible material. The rate and directions of air migration and the air:surface contact area depend
upon the geometry of the zone, compaction from overlying strata and the fineness of the crushed material.
Hence, the rapid escalation of temperature that characterizes the development of a concealed fire occurs
first at discrete foci or "hot spots". The synergistic effects of iron pyrites are now thought to be caused by
differential rates of expansion during the early stages of heating. The coal around the pyrites becomes
more finely crushed and produces additional area for oxidation. Furthermore, the pyrite itself becomes
oxidized, adding to the escalation in temperature. Green crystals of ferrous sulphate are formed which
oxidize further to the yellow hydrated ferric oxide, a characteristic feature of zones in coal mines that have
been involved in spontaneous combustion.
The migration of the heating again depends upon the rate and direction of air leakage. However, in
contrast to open fires, the tendency is for a spontaneous heating to propagate through crushed
material against the airflow, i.e. towards the intake airways.
21.4.2. Susceptibility to spontaneous combustion
A large number of tests and indices have been devised as suggested measures of the liability of
differing coals and other materials to spontaneous combustion. These have involved:
coal petrology and rank, the younger and lower ranks of coal being more susceptible
rates of oxygen consumption or temperature rise at specified phases of the oxidation process
self-heating temperatures (SHT) or other temperatures at specified stages of the heating process
rates of heat production during isothermal or adiabatic tests.
No such test or index has been found to have universal application. The difficulty is that susceptibility to
spontaneous combustion depends not only upon the material but also its physical state as well as the
psychrometric condition and migration paths of the leakage airflows, the latter depending, in turn, upon the
mining methods and layout. The matters of additional relevance are:
amount and degree of comminution of crushed material left in the goaf (gob) area, these
depending upon:
friability of coal
type of coal winning machine
efficiency of coal clearance (removal)
roof, floor or pillar coal left (percentage extraction)
thickness of seams and need for multi-lift or caving mining systems
depth and pre-stressing (microfracturing) of coal
geological disturbances
21 -
19
methods of stowing and sealing of roadsides

gradient of the seam and proximity of other seams
length of face
rate of face advance or retreat
roof/floor stability and strata stresses in the vicinity of pillar edges, stoppings, air crossings
and ventilation doors
the air pressure differential across the affected area
the layout and resistances of surrounding airways and faces, including obstructions
degree of consolidation and, hence, resistance of the caved areas
the relative moisture contents of the coal and air
the reduction of oxygen content in the goaf (gob) areas by the emission of methane or other gases.
With this number and variety of factors it is not surprising that efforts to characterize the susceptibility of
coals to spontaneous combustion purely on the basis of laboratory tests have met with rather limited
success. An improved approach involves allocating weighted credit to each of the factors listed in an
attempt to develop a site-specific indication of liability to spontaneous combustion (e.g. Banerjee, 1985;
PD-NCB, 1978).
21.4.3. Precautions against spontaneous combustion
As with most potential hazards in the subsurface environment, precautionary measures against
spontaneous commence at the time of planning and design of the mine. Core samples of the seam or ore
should be subjected to susceptibility tests discussed in the previous section. The layout of the ventilation
network should be designed to minimize pressure differentials between adjoining airways and across
caved areas. This might be arranged, for example, by favouring through-flow rather than U-tube
arrangements (Section 4.3.1.). While design airflows must be sufficient to deal with gases or other
airborne pollutants, consideration should be given to means of reducing those airflows such as methane
drainage. Booster fans, where allowed by law, provide a powerful means of air pressure management and,
coupled with the techniques of network analysis to investigate locations and fan duties, are most valuable
in reducing incidences of spontaneous combustion. Branch resistances in the surrounding ventilation
network should be kept as low as practicable by means of larger cross-sections or driving parallel entries.
Furthermore, obstructions in those airways should be avoided.
The probability of spontaneous combustion can be reduced by minimizing the amount of coal, timber,
paper, oily rags or other combustible materials that are left in gob areas. This may be inevitable if top coal
must be left for the purposes of roof control. Nevertheless, efficient clearance of the fragmented coal from
the face and good housekeeping should be practiced in mines that have a history of spontaneous
combustion.
It is important for the mine ventilation engineer to be conscious of the zones in which spontaneous
combustion is most likely to occur. Recalling that some leakage takes place through the strata around
stoppings and doors, spontaneous heating may occur in coal which exists within that strata, whether it be
in the roof, floor or sides. Good strata control and the liberal application of roadway sealants can help in
such circumstances.
Pillars in coal mines should be designed large enough to minimize crushing at the edges and corners.
Side bolts can help to maintain the integrity of pillars while the injection of low viscosity grouts might be
used as a last resort. Here again, the application of surface sealants assists in preventing ingress of air.
However, the most difficult types of spontaneous fires occur within caved zones and, in particular, in the
goaf (gob) areas of coal mines.
21 -
20
Figure 21.4 illustrates the air migration paths and the zones most liable to spontaneous heating in the
caved area of an advancing longwall. Those critical zones occur where the leakage airflows lie within that
range which provides sufficient oxygen to promote continued oxidation of combustible material, but not
enough to remove heat at the rate at which it is generated. There are two distinct zones. One of these lies
along the original starting line of the face where incomplete consolidation allows a leakage path between
intake and return. Within the central portion of the gob, consolidation allows little leakage. However,
another critical zone occurs between the fully consolidated central core and the advancing face. This zone
is not stationary but advances with the face. Recalling the time factor involved in the development of a
spontaneous heating (Figure 21.3), it is clear that if the face is advanced continuously and at a sufficient
rate then any potential heating may be "buried" by consolidation of the cave before it has had time to
develop into a spontaneous fire. The most hazardous times occur at weekends and over holiday periods
when additional precautions may be required. These include the application of roadway sealants, grout
injections into roadside packs, and lowering the pressure differential across the gob by pressure balance
techniques (Section 21.4.5.4.) In severe cases, injection of an inert gas into the gob may be considered
(Section 21.6).
Figure 21.5 gives a similar illustration for a back-bleeder retreating longwall where the gob immediately
behind the face is ventilated deliberately in order to prevent flushes of methane on to the faceline. Here
again, the face start-up line and a moving zone trailing behind the face provide the critical zones most
liable to spontaneous heatings. Despite the fact that the caved area is ventilated, less incidences of
spontaneous heatings have been reported using this system than for advancing longwalls. As advancing
or retreating systems tend to be favoured in separate geographical regions with differing coal seams and
climates, there may be many reasons for this, as discussed in Section 21.4.2. Nevertheless, there are at
least two features that may tend to mitigate against spontaneous heatings in back-bleeder layouts. First,
there is likely to be a smaller pressure differential applied laterally across the gob area. This is particularly
the case at the location of the starting line. Secondly, the fact that the area behind the face is actively
ventilated may cause the critical zone between the ventilated and consolidated areas to be narrower and
more quickly buried by the caving roof strata.
In mines with a history of spontaneous combustion, it is necessary to seal all abandoned workings. This is
particularly important when those areas are adjacent to current workings and even more so when they
exist within overlying strata. Spontaneous fires have resulted from undetected leakage airflows between
current gob areas and workings that had existed in higher seams many years previously.
Following the completion and withdrawal of equipment from a section of a mine liable to spontaneous
combustion, all entries into the section should be sealed and the atmospheric pressures applied to those
seals balanced as far as it is practicable to do so. This may be accomplished simply by a re-arrangement
of doors or by the dismantling of stoppings and air crossings. Active pressure balancing techniques may
be employed as discussed in Section 21.5.5. The sites of seals should be prepared at strategic control
points during the development of a section and should involve channels being excavated into the sides,
and providing a nearby stock of building materials. This will facilitate sealing the district rapidly in the event
of an uncontrollable fire.
21 -
21
advancing face
critical zone
start-off line
critical zone
Sufficient air to
remove heat as it is
produced.
Critical zones.
Sufficient oxygen to
promote oxidation but
not enough to remove
heat at the rate at which
it is produced. The
self-heating
temperature may be
reached.
critical zone
critical zone
Consolidation of
material prevents
enough oxygen to
cause combustion that
can reach selfsustaining temperature.
retreating face
start-off line
Figure 21.4 Zones showing liability to

spontaneous combustion in the caved area
behind a single entry advancing longwall face.
Figure 21.5 Zones showing liability to

spontaneous combustion in a multiple entry
retreating longwall with back bleeders.
21 -
22
21.4.4. Detection of a spontaneous heating

There are essentially three classes of detection for incipient or active spontaneous heatings. The oldest
and, many would argue, still the most reliable is through the human senses. The aromatics and
unsaturated hydrocarbon gases that are produced during the oxidation phases (Section 21.4.1.) give
rise to an odour known colloquially as gobstink. This has been described variously as like "petroleum,
the oil used in a flame-safety lamp and sterilizing liquid." Although, as with all odours, it can be described
only by analogy it is, nevertheless, very distinctive and unlikely to be forgotten. Figure 21.6 illustrates the
increasing strength of the smell with respect to gas concentrations and temperature of the oxidizing coal.
The odour is likely to be the first indication of a heating that can be detected by the human senses. The
output of water vapour may also be observed either as a haze in the air or from beads of condensation
on steel supports or other surfaces in locations where such condensation is unusual. At a later stage,
the matter is put beyond doubt by the appearance of smoke.
The second class of detectors are thermal
devices used to determine increases in
temperature. Infra-red scans of roadway
sides have been employed to identify the
emission points of warm gases into airways
and are useful for localized heatings in
pillars or around stoppings (Chakravorty and
Woolf, 1980). They are less useful as a
permanent and continuous means of
warning. Thermocouples or thermistors have
been left in gob areas. However, they too
have met with little success to the present
time. First, their wiring is unlikely to
withstand the mechanical stresses of an
active caving zone, even when sheathed.
Secondly, the thermal conductivity of
crushed rock is low. Hence, the temperature
even within a metre of an active centre of
heating may indicate no abnormal condition.
300
250
200
150
100
50
0
50
100
150
200
250
The most widespread method of detecting

the onset and development of a
spontaneous heating is by monitoring gas
concentrations in return airways (ref. Section
11.4.3). The gases that are evolved are
indicated on Figure 21.6. The interpretation
of the relative values and trends of these
concentrations is discussed in detail in
Section 21.7.
Figure 21.6 Example of the development of odour and

gases from a medium-volatile bituminous coal. The relative
positions of the gas curves will remain the same for other
coals but the actual concentrations vary with the rank of coal
and magnitude of leakage airflows.
21 -
23
21.4.5. Dealing with a spontaneous heating

There is a procedure that should be followed when detection systems indicate that an active
spontaneous heating is developing in a mine. First, a gas monitoring station should be set up
downstream from the affected area and air samples taken at intervals of not more than 30 minutes. If an
air monitoring system already exists in the mine then even greater detail of gas concentration trends can
be followed. Personnel should be evacuated from all return airways affected and, if the condition is
developing rapidly, also from the rest of the mine except for those involved in dealing with the situation.
Simultaneously, steps should be undertaken to identify the location of the fire. This may be obvious if
smoke appears from discrete places in a close vicinity such as from a fire in leakage paths around a
stopping or in the crushed corner of a pillar. The location of the fire is more difficult to detect if it occurs
in a caved zone or inaccessible old workings. Whether or not smoke is present, it is useful to conduct a
carbon monoxide survey. Measurements can be made by hand-held instruments or by stain tubes
(Section 11.4.2.). Indicating the results on a mine map can assist in selecting the most probable sites of
the fire.
Having located the fire, the next step is to decide how to control and, if possible, extinguish it. A variety
of methods exist for this purpose. The injection of an inert gas as a practical and powerful method of
dealing with both open and concealed fires in mines was developed through the 1980's and is discussed
separately in Section 21.6. Other techniques of fighting concealed fires are introduced in the following
subsections.
21.4.5.1. Excavating the fire
If the fire is known to be within a few metres of an airway then it may be possible to dig out, cool and
remove the incandescent material. This can be the case for fires in pillars, around stoppings or air
crossings, or for some gob fires. It may be necessary to drill holes into the zone in order to more
accurately determine the site and extent of the fire. The excavation should commence from the upwind
side of the fire to minimize the exposure of personnel to smoke and carbon monoxide. The airway
should be wetted and/or coated in stonedust for a distance of some 10 m on either side of the
excavation. Care should be taken to ensure control of the roof which may have become weakened by
heat. Readings of methane and carbon monoxide concentrations should be taken frequently
downstream from the site. As the fire is approached, it may be necessary to spray water on the workers
to cool them. When the seat of the heating is exposed, it should be cooled by water jets applied around
the periphery. Spraying water into the heart of a glowing carboniferous mass can result in the formation
of water gas (Section 11.2.6.). The hot material should be loaded into metal conveyances and
dampened thoroughly before transporting it out of the mine. When all of the incinerated material has
been removed, the void should be cooled and, after some 24 hours to ensure no re-ignition, filled with an
inert material such as limestone dust or gypsum-based wet fillers.
If it is impracticable to remove the fire physically then it becomes necessary to prevent ingress of air to
the fire location. This leads us into the remaining techniques of dealing with a concealed fire.
21.4.5.2. Burying the fire
In some cases, it is possible to prevent or reduce access of air to the fire location by burying it under
collapsed roof strata. Localized leakage through a roadside pack can be reduced by bringing the roof
down in the airway, employing shotfiring if necessary, removing only a portion of the debris and
compacting the remainder over the area to be sealed. The excavated roof must, of course, be well
supported leaving the airway with an anticline.
If the heating occurs in the critical moving zone behind a longwall face (Figures 21.4 and 21.5) and is
detected sufficiently early, then it is sometimes possible to bury it under consolidated caved material by
a temporary increase in the rate of face advance (or retreat). This is likely to be successful only if the
incipient heating has been detected by an early warning gas detection system and well before smoke
appears.
21 -
24
21.4.5.3. Sealants
A variety of sealants have been employed on stoppings, airway surfaces, roadside packs and pillars in
order to increase their resistance and, hence, reduce the access of leakage air. These may be applied
on external surfaces or injected as grouts into the strata or packed material. While sealants that include
resins or gels produce the lowest permeabilities and allow a degree of flexibility, the choice must often
be made quickly and on the basis of local availability. Concrete and gypsum plasters can be sprayed
quickly and effectively on to airway surfaces, as well as being useful for grouting and seal infills between
stoppings. Water based slurries using mill tailings or other waste material have also been employed as
grouts or for injecting into fire zones. These may be applied through boreholes drilled either from an
underground location or from the surface. The injection of sodium silicate into coal pillars has been
found to be effective (Banerjee, 1985).
While all known leakage paths connecting to the fire zone through roadsides should be sealed, it
appears to be particularly advantageous to seal on the inlet or high pressure side. It is, however, often
difficult to locate the relevant points of inward leakage on the inlet side. One way of doing this, where
allowed by law, is to employ the techniques of pressure management (ref. following subsection) to
intentionally and temporarily reverse the direction of leakage across the fire area. The appearance of
smoke or elevated concentrations of carbon monoxide in the intake will identify the normal inlet points
after which the flow reversal should be terminated or, better still, the leakage reduced to zero. This
method should be employed with caution and only when adequate gas monitoring facilities are available
in order to avoid unknown recirculation. The approval of governmental agencies should be sought if
necessary.
21.4.5.4. Localized pressure balancing
If no differential pressure exists across a level permeable zone then there can be no air leakage through
it. Pressure balancing involves raising the pressure on the return side or decreasing the pressure in the
intake until the leakage flow is reduced to near zero. This principle can be applied to complete sections
of abandoned workings as described in Section 21.5.5. or, in a more localized manner, to gain control of
a gob fire without sealing the district.
Figure 21.7 shows the principle of this technique which can be applied to a variety of situations. In the
example illustrated, a fire has commenced along the starting line of an advancing longwall. The pressure
differential between intake and return in that vicinity has been reduced to near zero by the installation of
a fan and regulator in the return airway. In cases of low normal pressure differentials, an induction fan
without any surrounding brattice may be sufficient (Section 4.4.3.). The hydraulic gradients shown on the
same figure illustrate the pressure differentials with and without the pressure balance. Where applicable,
the method can be applied quickly and at low cost to arrest the development of a spontaneous heating
and bring about an immediate reduction in carbon monoxide emissions. A pressure survey (Section 6.3.)
may be run around the district to determine the effective fan pressure required. If there is convenient
access, then a length of pressure tubing can be installed between the relevant points in the intake and
return airways including an in-line pressure gauge. The regulator in the return airway can be adjusted
until the pressure balance is achieved. Furthermore, an adjustable orifice within the short length of fan
ducting can be used to modify the effective fan pressure. The locations of the fan and regulator may be
changed in order to achieve finer control of the zone in which the pressure balance is applied.
The method of localized pressure balancing is very flexible and can achieve spectacular and speedy
success in many scenarios. However, it requires skilled personnel to devise and control each
installation. If applied inexpertly, it can result in partial recirculation of products of combustion. In any
case, it is a prudent precaution to employ a carbon monoxide monitor in the intake to detect any reversal
of leakage flows.
21 -
25
fire
zone
goaf
(a)
p
pressure differential
across fire zone
fire
zone
regulator
Normal hydraulic gradient
goaf
fan
(b)
p
pressure differential near

zero across fire zone
Hydraulic gradient with fan

and regulator in place
Figure 21.7 Example of localized pressure balancing to control a spontaneous heating

in a goaf (gob) area.
(a) without pressure balance
(b) with pressure balance
21.4.5.5. Flooding and sealing off
Flooding can be used in two ways to extinguish concealed fires. First, if the affected zone lies to the dip of
current workings then the fire itself can be flooded. This should be done in a controlled manner in order to be
able to handle the products of combustion and other gases displaced by the water. Furthermore, if a
previously flooded area is re-opened then it may rapidly re-ignite (Section 21.4.1.2.). The second use of
flooding is to provide very effective seals in airways that have low-lying sections. The water penetrates
fissures in the surrounding strata and provides a near perfect barrier against air leakage.
21 -
26
This brings us to the larger matter of building seals to control concealed fires in inby areas. While sealing
a section of a mine is often considered as a last resort, it is important to recognize those situations in
which it will become inevitable and then to seal quickly. This allows an inert atmosphere to build up and
to extinguish active combustion. Re-entry and a resumption of mining may then be possible within a
relatively short time. If, however, sealing-off has been delayed unduly, then the fire may have developed
into such a large conflagration that the area could be sterilized indefinitely. The one situation in which
sealing-off must be delayed is while there is the slightest possibility of rescuing trapped personnel. The
procedures involved in constructing seals during a fire emergency are discussed in Section 21.5.
21.5. STOPPINGS, SEALS AND SECTION PRESSURE BALANCES FOR

EMERGENCY SITUATIONS
This Section concentrates on the role of stoppings and seals in emergency situations involving
fires or explosions. The routine use of these devices during normal mining operations is described in
Section 4.2.1.2.
There is considerable divergence in differing geographical areas on the meaning of the terms stoppings
and seals. In some regions, the two are regarded as synonymous while, in others, they are interpreted
as quite different types of structures. To prevent confusion, we shall use the terms here according to the
following definitions:
Temporary stopping: A light structure erected from brattice cloths, other fabrics or boarding but
which will not withstand any significant physical loading.
Stopping: A single or double walled structure constructed from blocks, bricks, sandbags,
prefabricated steel panels or from substantial boarding attached to tight roof supports. Stoppings
are routinely employed as ventilation controls to minimize leakage between adjoining intake and
return airways in active mining areas but are not intended to be explosion proof. It is to be
expected, therefore, that stoppings will be destroyed or severely damaged in areas through
which an explosion has passed.
Seal: A barrier designed to withstand mine explosions.
Temporary stoppings, often in the form of brattice cloths or quickly erected alternatives are employed
during the fighting of open fires in order to regulate the airflow over the fire, change the air pressure in
nearby airways (Section 21.3.4.1) or re-route airflows to assist in rescue operations.
Stoppings may also be used as a more substantial means of modifying and/or controlling the airflow
paths in the event of a mine fire and to re-establish controlled ventilation during rescue operations.
Seals are used when alternative means of controlling a fire or spontaneous heating have been
exhausted and inby rescue operations have been terminated. The purpose is to stop the flow of air and
allow an inert atmosphere to build up within the affected zone. The decision on when to erect seals
should be contingent on the particular circumstances. In the case of a deeply seated concealed fire
seals may be erected in all entries to the area at a fairly early stage and before the heating reaches an
open airway. The district may subsequently be re-entered for salvage of equipment or resumption of
mining either after a cool-down period or when further arrangements have been made to control the
heating. Sealing an airway fire should be carried out as soon as it becomes clear that the fire is out of
control and rescue operations have been terminated.
21 -
27
21.5.1. Site selection of seals

It is important that all openings into an uncontrolled fire zone be sealed except for sampling pipes. Such
openings may include boreholes and connections into old workings. In the shallower mines, there may be
fractures extending through to the surface. The potential effects on surface structures and their inhabitants
should be considered.
The sites should be selected to minimize the number of seals required. These are matters that should be
taken into account during planning of the mine layout in order to facilitate rapid sealing of a section should
that subsequently prove to be necessary. The precise locations should be in well supported areas with
solid roof, floor and sides.
A well organized mine will have pre-prepared sites for seals at control points in each district and with a
stock of the appropriate materials nearby (IME, 1985). In any case, those sites should allow ready access
for the supply of further materials and, also, for the provision of ventilation to personnel involved in building
the seal. The latter may necessitate a temporary duct and auxiliary fan. If the site is polluted by products of
combustion then the construction must be undertaken by teams fitted with breathing apparatus. Where
there is danger of an explosion or outrush of gases during construction, sites should be chosen that permit
rapid escape of the personnel.
It is unlikely that locations will be found which satisfy all of these requirements. Site selection for stoppings
and seals invariably requires compromises between optimum positions and practical considerations.
21.5.2. Sequence of building seals
In the case of concealed fires and where there is no imminent danger of explosion, it is preferable to
complete intake seals first in order to terminate airflow into the fire zone, then to follow up quickly with
closure of the return airways. However, in the case of an open fire, or where monitored gas concentrations
indicate that an explosive atmosphere may develop then all stoppings or seals should be completed
simultaneously and personnel evacuated from the mine for a period of 24 hours.
21.5.3. Construction of seals and stoppings
It is unlikely that explosions arising from fires in mines will generate pressure peaks of more than some
350 kPa gauge pressure1 (50 psig2) on the faces of seals although considerably higher pressures have
been measured in experimental explosions (Strang, 1985) and in explosions initiated by an ignition of
methane. Hence, seals which (by our definition) should be explosion-proof must be able to withstand such
dynamic pressure pulses. Figure 21.8 illustrates a typical seal. The end walls may be constructed from
sandbags, masonry or any form of non-flammable blocks. The strength of the seal should be provided by
the infill and any girders or other supports that may be employed in the intervening space. The length of
Airway width + height
+ 0.6 with a minimum length
seal is shown as 3 to 5 m but may be calculated as
2
of 3 m (Hornsby et al, 1985) and should be chosen with reference to the condition of the surrounding
strata and the type of infill available.
Construction of a seal in coal mines should commence by applying stonedust liberally to the airway inby
the seal. Stonedust barriers may also be erected inby the seal to arrest an explosion before it reaches the
face of the seal (Section 21.8.3.1.). Figure 21.8 illustrates the end walls recessed into the roof, floor and
sides. This is considered to be good practice although it may not be necessary provided that the strata is
competent, the length of seal adequate and a modern (quick-setting) wet infill is available. Conveyor
1
Sometimes known as overpressure

psig means pounds force per square inch gauge pressure. British Imperial units have been added as sections in
this chapter are of particular interest in the United States following the Sago and Darby Mine explosions.
2
21 -
28
structure should be dismantled and cleared from the area. Similarly, pipes and cables should be
dislocated and removed.
pressure tube
sealed
area
sampling
tube
infill
access
tube
3 to 5 m
Figure 21.8 Example of an explosion proof seal where an active fire
exists in the sealed area.
While dry material such as sand, stonedust or fly ash may be used for the infill, a better seal is obtained
by employing a gypsum based plaster. This is pumped in as a liquid which penetrates immediate
fractures in the surrounding strata then sets to a compact solid material. It is important that the infill
completely blocks the seal to the roof.
The gas sampling pipe should extend at least 30 m inby the seal to allow for 'breathing' induced by
fluctuations in barometric pressure. Three flexible tubes may be threaded through the metal sampling
pipe in order to draw independent samples from roof, floor and middle of the sealed airway.
The steel access tube shown on Figure 21.8 has a number of purposes. First, it allows ventilation to be
supplied to the fire while construction of the seal is in process. This tends to retard the development of a
fuel-rich situation and, hence, reduces the risk of explosion whilst workers are still at the site. Despite the
access tube, workers may be subjected to roll-back of smoke during construction of a seal. When each
of the seals has been completed then all access tubes should all be blanked off simultaneously by
strong steel plates both ends. All personnel must leave the mine for a period of 24 hours or until
monitored signals of gas concentration indicate that an inert atmosphere has been attained.
Following the sudden cessation of airflow, significant changes take place within the sealed zone. The
concentration of combustible gases increases while the oxygen content decreases. In the case of open
fires, it is probable that the mixture of gases will pass through an explosive range (Section 21.7.).
Meanwhile, convection effects will create movements of the changing atmosphere and, perhaps, rolling
or 'tidal' flames. The combination of these effects may produce a series of explosions. These can be
monitored by pressure transducers connected to the pressure tubes on the seals. An older but simpler
method is to leave a glass U-tube containing mercury attached to a pressure tube. If the mercury is
subsequently found to have been blown out then an explosion has occurred within the sealed zone.
A second reason for the access tube is to allow subsequent re-entry and inspection by rescue teams
equipped with breathing apparatus. The outby end of the tube should be at a convenient height for this
purpose. Should it become necessary, the access tube can be filled with infill material. This is the reason
for the slight downward inclination of the tube towards the sealed area.
21 -
29
There is one major disadvantage to the type of seal we have been discussing - the time taken for its
construction. In the case of an open fire in a timbered or coal-lined airway, rapid action is vital. It may be
necessary to isolate the zone in a much shorter time period than that required to build an explosion proof
seal. An alternative is to erect a stopping that contains a pressure-relief flap. This is a technique that has
met with some success in the United States (Mitchell, 1990). Figure 21.9 illustrates such a vented
stopping. Boards are attached firmly to two or three tight chocks (cribs) on both upstream and
downstream sides except for the top quarter or third of the airway. A sealant can be sprayed over the
boards. A weighted flap of conveyor belting is attached across the top of the outby side and held open
by a cord which is tensioned by the weight of a canister of water. At the appropriate moment the canister
at each site is punctured and all personnel leave the mine. The flaps close as the canisters empty.
Except for violent explosions, the structure of a vented stopping will remain intact. Lesser pulses simply
blow open the flap which then falls back into place leaving the integrity of the stopping secure. Although
the vented stopping does not have the structural strength or resistance of an explosion- proof seal, it
provides a temporary expedient until permanent seals can be established, should those prove to be
necessary.
weighted belt
pulley
punctured
water vessel
Figure 21.9 A vented emergency stopping.
21.5.4. Re-opening a sealed area
The simple method of breaching stoppings or seals and re-establishing an air circuit suddenly is not one
that can be recommended when a mine or section of a mine has been sealed because of a fire. Even
when gas samples indicate that the fire is no longer active, hot spots may still remain that could result in
re-ignition when supplied with oxygen. The preferred technique is first to send in rescue teams equipped
with breathing apparatus to make a thorough inspection of the area and to take additional air samples. If
the sealed area is extensive and, particularly, if the atmosphere will pass through an explosive range
when mixed with fresh air, it may be necessary to re- ventilate in stages, building additional seals or
stoppings further inby and closer to the original fire zone.
21 -
30
21.5.5. Section pressure balances

No stopping or seal in the subsurface has infinite resistance. Even if the stopping or seal itself were
perfect, the potential for leakage still occurs in the surrounding strata. Such leakage can delay the
extinction of a fire or, if serious, may maintain the fire indefinitely. Section 21.4.5.4. described how air
pressure management can be employed to control the leakage airflows feeding a concealed fire. The
same principle can be applied to complete sections of a mine.
The most rudimentary form of pressure balancing involves the re-arrangement of doors and aircrossings to ensure that each seal or stopping is exposed to return airway pressure. This reduces the
pressure differential applied across the sealed district to the pressure drop over the length(s) of airway
between the stoppings. Figure 21.10(a) illustrates an improved version. An additional wall is constructed
some 3 to 5 metres in front of each main stopping or seal to form a pressure chamber. A duct of about
0.5 m diameter connects the chambers and, hence, equalizes their pressures. There is then no pressure
differential applied across the sealed district, nor can there be any continuous flow around it. These are
examples of passive pressure balancing.
ducting
pressure chambers
(a) A passive pressure balance.
infill
pressure
pressure duct
pipe
gauge
sampling
pipe
balancing duct
pressure balance chamber
(b) An active pressure balance chamber.

Figure 21.10 Passive and active pressure balance chambers.
The drawback to passive pressure balances is that they do not prevent "breathing" of the seals or
stoppings during periods of changing barometric pressure. This can be overcome by employing an
active (or powered) pressure balance system. This provides a means of equalizing the air pressures on
the inner and outer faces of the same seal or stopping and is illustrated on Figure 21.10(b). Two ducts
pass through the outer wall into the pressure chamber. One of these (the pressure duct) is supplied by a
ventilating pressure either by a small fan or by laying the duct through a nearby door or stopping in order
to utilize the mine ventilating pressure. The pressure chamber will be pressurized positively if the seal is
on the return side of the main ventilation network and negatively if it faces an intake airway. The second
21 -
31
(balancing) duct simply passes through the outer wall. Both ducts are fitted with variable dampers to give
a flexible means of fine adjustment to the pressure within the chamber.
Two tubes are employed to monitor air pressure, one passing completely through the seal and the other
into the chamber. A gauge across these two tubes will indicate zero when there is no pressure
differential between the chamber and the inby side of the seal. The duct dampers can be adjusted
manually when the pressure gauge deviates from zero. The arrangement can be made automatic by
employing an electronic pressure gauge which transmits an amplified signal to servo-motors on one or
both of the duct dampers.
A number of special purpose adaptations to the principle of pressure chambers have been devised. One
example may occur in a nuclear waste repository to ensure that leakage through an air lock takes place
in one consistent direction at all times, irrespective of the direction of pressure differential across the
airlock (Brunner et al, 1991). Another example, where permitted by legislation, is the use of compressed
air to provide positive pressurization of the pressure chamber.
21.6. THE USE OF INERT GASES

The injection of inert gases to assist in the control of subsurface fires has been undertaken since, at
least, the 1950's (Herbert, 1988). However, from 1974, significant developments in the deployment of
nitrogen took place in Germany. The technique has become commonplace in coal mining areas where
spontaneous combustion occurs frequently (Both, 1981). The overall purpose of injecting an inert gas is
to reduce the oxygen content in order to prevent or inhibit combustion. The objectives may further be
classified as follows:
To prevent concealed heatings in zones that are highly susceptible to spontaneous combustion.
To reduce the risk of explosions during sealing or stopping-off procedures.
To accelerate the development of an inert atmosphere in a newly sealed zone and to prevent
the creation of an explosive mixture when it is re-opened.
To control the propagation of an open fire during rescue, firefighting and sealing operations.
To prevent an explosive mixture forming due to breathing of a seal.
Three types of gases have been used in the procedure for which the term inertization has been coined;
carbon dioxide, products of combustion and nitrogen. In this Section we shall discuss the employment of
these gases in addition to methods of application and control.
21.6.1. Carbon dioxide
Carbon dioxide has a density of 1.52 relative to air (Table 11.1). This makes it particularly useful for the
treatment of fires in low-lying areas such as dip workings or inclined drifts (Froger, 1985). However, the
same property can render it difficult to control in horizontal workings. A 20 t tanker of liquid carbon
dioxide will produce some 9000 m3 of cool gas. The liquid form may be piped into the area where it is
required and expanded through an orifice or, indeed, injected directly into a localized heating. In both
cases, the gas removes heat from the fire as well as promoting an inert atmosphere. However, piping
the liquid carbon dioxide can give rise to freezing problems as well as difficulties in handling the pipes.
The use of carbon dioxide as an inerting gas has several other disadvantages. It is quite soluble in water
and can suffer some loss in wet conditions. More significant perhaps, is the fact that it adsorbs readily on
to coal and coked surfaces, even more so than methane (Figure 12.2(b)). When exposed to
incandescent carboniferous surfaces it may be reduced to carbon monoxide. Furthermore, it is
considerably more expensive than nitrogen.
21 -
32
21.6.2. Combustion gases

Following the sealing of a fire zone, gases produced by the combustion processes, combined with the
consumption of oxygen, will produce an extinguishing atmosphere. However, it may be rich in
combustible gases and become explosive if air is subsequently re-admitted (Section 21.7.3.). The
products of full combustion, primarily mixtures of carbon dioxide, nitrogen and water vapour, have been
employed as an injected inert gas. Flue gas from burning coal has been used in China (Sun, 1963) while
modified jet engines have been employed in several countries including Poland, Russia and
Czechoslovakia (Strang, 1985). The latter method involves burning kerosene at rates of some 0.7 kg/s
to produce 30 m3/s of inert exhaust gases. These are cooled by large quantities of water and admitted
into the fire zone. The engine produces a power output of about 30 MW which can be usefully employed.
Where the law allows the underground use of a jet engine or where it can be employed on surface for a
drift mine, then the large output of inert exhaust gases makes it attractive. However, it nullifies the
employment of gas analysis as a means of following the progression of the fire, the capital cost is high
and a highly specialized team is required to operate and maintain it.
21.6.3. Nitrogen
Liquid nitrogen is the basis for the majority of inertization schemes now employed for subsurface fires.
Again, the liquid gas is supplied in tankers of, typically, 20 t capacity giving about 16500 m3 of gas. For
continuous operation throughout a period of gas injection, the tankers may unload into a bulk storage
vessel of up to 40 t capacity and which has been brought to the mine site.
Due to the low boiling temperature of nitrogen, the liquid must be evaporated before piping it into the
mine. Figure 21.11 indicates the principle of a mobile evaporator which, again, has been brought to the
mine for the emergency period. Typically, two water circuits are employed; a primary circuit using
atmospheric heat and secondary heaters powered by electricity or liquid/gas fuels. The gaseous
nitrogen passes through a bank of controllers before entering the mine pipeline. A subsidiary nitrogen
line provides a feedback to maintain a constant pressure in the storage vessel. The maximum gas feed
rate into the mine depends upon the duty of the evaporator but may, typically, be within the range 1 to 6
m3/s.
Liquid nitrogen is a by-product of the commercial production of oxygen and is much less expensive than
liquid carbon dioxide. Furthermore, it is not as soluble as the latter, does not adsorb so readily on carbon
surfaces and with a density approximating that of air, mixes readily without stratification.
21.6.4. Methods of application and control
In order to assess the volume flow of inert gas required, the rate of oxygen supply to the fire may be
determined from the inlet oxygen concentration and a measured or estimated airflow. It is then a
straightforward calculation to determine the flow rate of inert gas required to dilute the oxygen down to
10 per cent in order to extinguish flaming combustion, or to less than 2 per cent to suppress
smouldering.
The inert gas may pass into the mine via water pipes or compressed air pipes commandeered for the
purpose. Alternatively, gas feed boreholes may be drilled from the surface to intersect a fire zone (e.g.
Zabrosky and Klinefelter, 1988). Perhaps the most difficult aspect of inertization is controlling dilution of
the inert gas by air leakage. If leakage air enters in significant quantity between the gas injection point(s)
and the fire, then the technique may fail. It follows that inertization is most likely to succeed where the
fire is in a single entry with no leaking crosscuts or, in the case of a concealed fire, where the air inlet
points have been well defined. Conversely, multi-entry systems offer a greater opportunity for
21 -
33
liquid nitrogen bulk storage vessel
fan
radiator
nitrogen
heat
exchangers
water
electric or
gas heaters
controllers
(flow, pressure, temperature)
nitrogen gas
Figure 21.11 Simplified schematic of a mobile nitrogen evaporator.
dilution of the inert gas. Although the employment of inert gases can create difficulties in interpreting the
analyses of gas samples taken downstream from the fire, it is usually possible to detect whether the fire
is being suppressed. If there is no noticeable effect on an open fire within an hour or two then potential
air leakage points should be investigated. Additional stoppings or pressure management techniques
may be required to reduce the inward leakage.
For the control of spontaneous heatings in the goaf (gob) areas of longwall mines, the location of the fire
and air entry points should first be established (Section 21.4.5.4.). Furthermore, a knowledge of air
migration paths in goaf areas is invaluable (Figures 21.4 and 21.5). Injection pipes should be inserted
from the airways or working face into the inlet zones, using boreholes if necessary. The volume
flowrates of inert gas required are usually much lower for spontaneous heatings than for open fires.
However, the reaction may be slower. Indeed, the monitored concentrations of carbon monoxide and
methane may increase for up to 36 hours as the inert gas displaces those gases from the fire zone. A
steady nerve is useful at such times. When the carbon monoxide concentration begins to fall, that can
be used as a controlling guide to the required injection rate of the inert gas. Oxygen concentration in the
return airways downstream from goaf (gob) inertization should be monitored to ensure that it remains
above the relevant mandatory limits (19 to 19.5 per cent).
For coal mines with a history of recurring spontaneous combustion, the trend is towards establishing a
permanent nitrogen "fixing" plant on the surface (e.g. BHP Billiton Sustainability Report, San Juan Coal
Company , 2005). The lower grade of nitrogen that may be produced by this method is of little
consequence for inertization. A permanent network of nitrogen pipelines throughout the working sections
of the mine allows the gas to be fed at a relatively low rate but continuously into the caved zone.
21 -
34
Properly designed, this creates an inert atmosphere throughout the critical zones (Figures 21.4 and
21.5). As a further projection to future developments, the infrastructure is then in place for complete
inertization of the working face if and when the techniques of automation and remote control make that
cost effective.
Where inertization is having a beneficial effect it is important that it be maintained for as long as
required. While a deeply seated fire can be controlled by an inert gas, it will seldom be cooled sufficiently
to extinguish it. Hence, premature cessation of the operation may result in a rapid escalation of the fire.
Similarly, injection into a sealed area should be continued until the oxygen content falls below 2 per cent.
21.7. FIRE GASES AND THEIR INTERPRETATION

In Section 11.3.3. we introduced the gases that are produced in the majority of underground fires. Other
than use of the human senses, monitoring the quality of the air in a mine is the dominant method of
detecting a fire or spontaneous heating. Sampling the air downstream from a fire or from within a newly
sealed area and plotting the trends is the primary method of tracking the behaviour of the fire and the
development of atmospheres that are, or may become, explosive. However, as the gases emitted vary
with the phases of oxidation, time and temperature, it is necessary to employ skilled interpretation of
those trends.
21.7.1. The processes of burning and the gases produced
When coal or timber are burning, three processes are in progress.
distillation of gases from the solid material
oxidation of the solid material on its surface with the emission of heat and light (this is why
the surface glows more brightly when fanned with fresh air) and
flaming combustion - the burning of combustible gases produced by the first two
processes. Again heat and light are produced. Some of the heat passes back to the
surface by radiation and convection to assist in the promotion of further distillation.
The fire gases and their relative proportions depend upon the contributions of each of these three
processes. The gases of distillation from coal are carbon monoxide, carbon dioxide, hydrogen and water
vapour. Methane and other hydrocarbon gases are also produced. Timber distills the same gases
although the amount of hydrogen may be negligible. When flaming combustion occurs, the combustible
gases burn to a degree that is governed by the availability of oxygen in the air. The final mixture leaving
the fire zone is, therefore, a result of the gases of distillation and the extent to which the fire has become
fuel-rich.
The complexity of the processes involved can be illustrated by considering just a few of the ways in
which methane can burn:
2O 2
2H 2 O + CO 2
2CH 4 + 3O 2
4H 2 O + 2CO
3CH 4 + 5O 2
6H 2 O
CH 4
+ 2CO + CO 2
Furthermore, secondary reactions may produce water gas and further reduction of carbon dioxide to
carbon monoxide.
21 -
35
21.7.2. The detection and trend analysis of fire gases

For many purposes of analysis the sampled atmosphere is considered as comprised of air, combustibles
and inerts (excess nitrogen and carbon dioxide). Figure 21.6 illustrates the combustible gases emitted
as coal is heated in a limited air supply such as occurs in a spontaneous heating. It is clear that carbon
monoxide is a leading indicator of the early stages of such a fire. However, the saturated hydrocarbons,
in particular, ethylene, are useful indicators of burning coal. In general, as the fire develops into open
combustion, the major gaseous product, carbon dioxide, forms an increasing proportion of the pollutant
emission.
During any fire incident, running graphs should be maintained of the concentrations of the gases that are
monitored or gained from the analysis of samples. Transmitting transducers installed downstream from
the fire give a stream of near continuous data. However, in many cases, samples must still be obtained
manually by means of evacuated chambers or hand pumps attached to sample containers. The sample
vessels should be clean, dry and free from any lubricants that may contaminate the gas sample. A
mobile laboratory should be available at the mine surface during any major incident to provide rapid
analysis of samples. The rate of sampling must be dictated by the urgency of the situation and the speed
at which conditions are changing.
Since the beginning of the twentieth century a number of ratios and composites of gas concentrations
have been suggested to assist in the interpretation of fire gases. Table 21.1 indicates some of these.
Ratio
CO / O2
CO2 / O2
CO /(Excess N2 + CO2 + combustibles )
(CO2 + 0.75CO 0.25H2 ) / O2

CO / CO 2
Name
Graham's Ratio or Index for Carbon

Monoxide (ICO)
Young's Ratio
Willett's Ratio
Jones and Trickett Ratio
Oxides of Carbon Ratio
Table 21.1. Gas ratios used in interpreting trends of gas concentrations produced by mine fires.
A feature of several of the ratios is the oxygen deficiency, 2 . This is a measure of the oxygen that
has been consumed and is based on two assumptions; first, that the air has been supplied with 20.93
per cent oxygen and 79.04 per cent inert gases (excepting 0.03 per cent carbon dioxide). That 79.04 per
cent contains traces of other gases but is referred to simply as nitrogen. Secondly, it is assumed that no
nitrogen has been consumed or added (except from the air) through the area under consideration. If no
oxygen is consumed, then the O 2 /N 2 ratio would remain at 20.93/79.04 = 0.2648 irrespective of the
addition of other gases. For any measured values of O2 and N2, the concentration of oxygen that was
originally in place can be calculated as
20.93
N2
79.04
Hence, the amount of oxygen that has been consumed, or oxygen deficiency is given as
20.93
O 2 =
N 2 O2
per cent
(21.8)
79.04
21 -
36
The oxidation of coal and the corresponding production of gases were studied by Dr. J.S. Haldane and
others at the beginning of the 20th Century (e.g. Haldane and Meachen, 1898). By 1914, Ivon Graham
had recognized the importance of carbon monoxide as an early indicator of the spontaneous heating of
coal and the equally vital influence of the oxygen that was consumed. He first suggested using the index
CO/O 2 , now known as Graham's Ratio or the Index for Carbon Monoxide (ICO).
Graham's Ratio is the most widely used indicator of an incipient heating in coal mines and has often
given warnings several weeks before any odour could be detected. It has the significant advantage that
it is almost independent of dilution by leakage of air as this affects both numerator and denominator
equally. However, none of the indices listed in Table 21.1 is infallible and Graham's Ratio does have
some drawbacks. First, its accuracy becomes suspect if very little oxygen has been consumed, i.e.
Graham's Ratio is unreliable if the oxygen deficiency, O 2 , is less than 0.3 per cent. This is a weakness
shared by the other indices that involve oxygen deficiency. Secondly, it will be affected by sources of
carbon monoxide other than the fire including the use of diesel equipment, or if the air supplied to the fire
is not fresh. The latter can occur if the fire is fed, partially, by air that has migrated through old workings
and contains blackdamp (de-oxygenated air). Again, like other trace gases or indices, a normal range of
Graham's Ratio should be established for any given mine. This will usually be less than 0.5 per cent.
Any consistently rising values in excess of 0.5 per cent is indicative of a heating.
Example
An air sample taken from a return airway yields the following analysis:
nitrogen,
oxygen,
carbon monoxide,
N2
O2
CO
=
=
=
Oxygen deficiency,
O2
Grahams Ratio
CO
O 2
79.22 per cent

20.05 per cent
18 ppm = 0.0018 per cent
20.93
79.22 20.05
79.04
=
20.98 - 20.05
0.0018
100
0.93
= 0.93 per cent
0.19 per cent
The CO2 /O 2 , or Young's Ratio is, again, nearly independent of dilution by fresh air. Carbon dioxide is
the most prolific of the gases produced in mine fires. Hence, the values of CO2 /O 2 will be much higher
than CO/O 2 . As a fire progresses from smouldering to open flame, the burning of carbon monoxide
will produce an increase in carbon dioxide. Hence a simultaneous rise in CO2 /O 2 and fall in CO/O 2
indicates further development of the fire. However, as both ratios have the same denominator, the
straightforward plots of carbon dioxide and carbon monoxide show the same trends. Young's Ratio
suffers from similar limitations to Graham's Ratio. Additionally, the concentration of carbon dioxide may
have been influenced by adsorption, its solubility in water, strata emissions of the gas and other
chemical reactions.
Willett's Ratio was introduced by Dr. H.L. Willett in 1951 with specific reference to situations where there
is a higher than usual evolution of carbon monoxide by ongoing low temperature oxidation. In these
cases, the gradual extinction of a fire in a sealed area may not be reflected well by the carbon monoxide
trend alone but as a percentage of the air-free content of the sample.
The Trickett or Jones-Trickett Ratio is used as a measure of reliability of sample analysis and also as an
indicator of the type of fuel involved. It can be used for the gaseous products of both fires and
explosions. Typical values are shown on Table 21.2. Dilution by fresh air has no effect on the JonesTrickett Ratio. However, it is subject to the limitations of oxygen deficiency.
21 -
37
The Oxides of Carbon Ratio, CO/CO2 is a useful pointer to the progression of the fire, rising during the
early stages and tending to remain constant during flaming combustion. However, the CO/CO2 rises
rapidly again as a fire becomes fuel-rich and is an excellent indicator of this condition. This ratio may
also be favoured because it is unaffected by inflows of air, methane or injected nitrogen (Mitchell, 1990).
It is, however, subject to variations in carbon monoxide and carbon dioxide that are not caused by the
fire.
Fuel
Jones-Trickett Ratio
Fires
methane
coal, oil, conveyor belting,
insulation and polyurethanes
timber
0.4 to 0.5
0.5 to 1.0
0.5 to 1.0
0.9 to 1.6
Explosions
methane
coal dust
methane and coal dust
0.5
0.87
0.5 to 0.87
no combustion process
< 0.4
impossible mixture (reliability check)
> 1.6
Table 21.2. Typical values of the Jones-Trickett Ratio (after Strang and MacKenzie-Wood, 1985).
It is clear that each gas and composite ratio has both strengths and weaknesses as a warning of
impending fire or indicator of fire development. During any fire, it is recommended that spreadsheet and
graphics software should be utilized to store all monitored data and sample analyses on a desk
computer. This should be located in, or close to, the emergency control centre. All gas concentrations
and preferred ratios should be made available as screen and hard-copy graphs plotted against time.
These should be reviewed repeatedly as new data or analyses become available. The trends are more
informative than the absolute values. Hence, repeated samples from a few set stations are preferred to
samples taken at many different locations. The injection of inert gases will affect the gas analyses. The
one exception is that the CO/CO2 ratio is not influenced by the injection of nitrogen.
Gas concentrations and indices give pointers to the average intensity but not the size of the fire. In some
cases, where there is limited leakage between the fire and sampling points, an estimate of the extent of
the fire may be made from the flowrate of each of the gases. This is given as the product of the air
flowrate and the relevant gas concentration.
The principles of gas detection and methods of sampling are discussed in Section 11.4. Ionization
smoke detectors draw smoke particles through a chamber where they are charged by a radioactive
source such as americium 241 or krypton 85. An ion collecting grid and amplifier produce an electrical
output that is a function of the smoke concentration (e.g. Pomroy, 1988). Versions of this have been
produced that can distinguish between smoke from a fire and diesel particulate matter (Lytton, 1988).
The improvements that have been made in electrochemical methods of detecting carbon monoxide
(Section 11.4.2.5.) have led to this being a common method of fire detection as part of a mine
environmental monitoring system. The employment of computer analysis allows filtering of false alarms
from diesels or blasting operations that gave earlier versions a doubtful reputation (e.g. Eicker and
Kartenberg, 1984). The filter program may be fairly simple such as ignoring short term peaks and giving
audio visual alarms only when a significant upward trend is indicated. More sophisticated systems take
21 -
38
into account the time-variant generation of carbon monoxide between successive sensors distributed
along an airflow path (Boulton, 1991).
The combination of computer controlled monitoring systems and ventilation network analysis programs
allows the possibility of not only detecting the existence of a fire at an early stage, but also its probable
location. This is facilitated by strategic location of the sensors (Pomroy and Laage, 1988). The
employment of tube bundle sampling is particularly useful for detecting the early and slow development
stages of a spontaneous heating. This technique is described in Section 11.4.3.2.
Temperature monitors, often known rather loosely as "heat sensors" have an application when mounted
above fixed equipment and, particularly, when used to activate deluge or sprinkler systems (Section
21.3.3.). They are of somewhat limited use as fire detectors in airways as the air temperature drops
rapidly downstream from a fire. Such sensors may be subjected to greater variations in air temperature
from normal operations (such as passage of a diesel vehicle) than from the early stages of a fire. If
made sufficiently sensitive, temperature sensors may lose credibility because of an excessive
recurrence of false alarms.
21.7.3. Explosibility diagrams
When a fire becomes fuel-rich, there is a danger that explosive mixtures of gases will propagate away
from the immediate fire zone. Furthermore, following sealing of a fire area, it frequently occurs that the
rising concentration of combustible gases and falling concentration of oxygen produce mixtures that
pass through an explosive range. Similarly, the dilution of combustible gases that occurs when a sealed
area is re-opened may, again, result in passing through an explosive range. In order to be able to predict
and control such circumstances, it is necessary to have an understanding of the flammability limits of
gases and gas mixtures.
This subject was introduced in Section 11.2.4. and illustrated by the Coward diagram for methane shown
in Figure 11.1. That diagram should be reviewed, if necessary, as a reminder of the concepts of upper
and lower flammability and how these converge to a nose-limit as inert gases are added.
In many situations, the large majority of combustible gas is methane and Figure 11.1 may be employed
to determine whether the mixture of oxygen, methane and inerts lies within the explosive triangle or is
likely to do so in the near future. However, in other cases, the presence of carbon monoxide and/or
hydrogen may cause significant changes in the Coward diagram (Coward, 1952, Strang J. and
Mackenzie-Wood, 1990). Figure 21.12 shows the individual explosive triangles for all three of these
combustible gases. The corresponding coordinate points are given in Table 21.3.
Gas
Flammable
lower
limits
upper
gas
Nose
limits
oxygen
methane
5.0
14.0
5.9
12.2
carbon monoxide
12.5
74.2
13.8
6.1
hydrogen
4.0
74.2
4.3
5.1
Table 21.3. Vertices of explosive triangles (percentages).
21 -
39
22
20
18
16
impossible mixture
14
12
10
8
6
4
not explosive
2
0
C
10
20
30
40
50
60
70
80
90
100
Figure 21.12 Coward explosive triangles for methane, carbon monoxide and hydrogen.
If the method of the Coward diagram is to be useful for mixtures involving more than one combustible
gas then we must be able to quantify the explosive triangle for those composite mixtures. Let us attempt
to do that.
A basic precept in the world of science and engineering is that if anything is done to upset the
equilibrium of a system then that system will react in an attempt to reach a new equilibrium with the
minimum adjustment of the component parts (Le Chatelier's Principle). Let us adopt subscripts 1, 2 and
3 for the three combustible gases and call their percentage concentrations p1, p2 and p3 respectively. If
they do not react chemically with each other then the mixture will have a total combustible concentration
of
p t = p1 + p 2 + p 3
percent
(21.9)
Furthermore, Le Chatelier's Principle leads to the prediction that for gas flammability limits, L1, L2 and L3
(where these can be upper, lower or nose limits), the corresponding gas flammability limit of the mixture,
Lmix, will be given by:
pt
p
p
p
= 1 + 2 + 3
Lmix
L1
L2
L3
(21.10)
21 -
40
Example
An air sample produces the following analysis:
CH4
8 per cent
CO
5 per cent
H2
3 per cent
Determine the lower flammability limit of this mixture.
Solution
The total percentage of combustible gases is
pt = 8 + 5 + 3 = 16
percent.
Using the individual gas lower flammability limits given in Table 21.3, together with equation (21.10)
gives
16
8
5
3
=
+
+
Lmix
5.0 12.5
4 .0
giving Lmix = 5.82 percent as the lower flammability limit of the mixture. A similar calculation may be
used to determine the upper and nose flammability limits of the combustible content.
The upper and lower limits lie on the line AB on Figure 21.12 and, hence, are defined completely.
However, the oxygen content at the nose limit remains to be found before we can construct the
explosive triangle for the mixture. To find the oxygen content at the nose limit, we must first determine
the excess inert gas (let us call it nitrogen) that has to be added in order to reach that nose limit.
Consider the situation for methane. If we start from any point on the line AB and add nitrogen then we
shall move in a straight line towards the origin, O. The mixture will become extinctive when we cross the
line AC. At that moment we shall have added an amount of nitrogen which, when expressed per unit
volume of methane, is a constant. (This follows from the fact that both AB and AC are straight lines.) As
we can commence at any position on AB, let us choose point B. On adding nitrogen and moving towards
O, we shall cross the extinction line at point C where the methane concentration is 14.14 per cent. The
remaining (100 - 14.14) = 85.86 percent is nitrogen. We have, therefore, added 85.86/14.14 = 6.07m3
of nitrogen for each m3 of methane. A similar exercise can be carried out for carbon monoxide and
hydrogen to give the values in Table 21.4.
Nitrogen to be added to make mixture
extinctive:
(N+ m3 of nitrogen per m3 of
combustible gas)
6.07
4.13
16.59
Combustible gas
methane
carbon monoxide
hydrogen
Table 21.4. Volumes of excess nitrogen to be added, N+, in order to make flammable gases extinctive.
This table gives the excess nitrogen to be added, N+, if the combustible content consisted of one
gas only. For a mixture of combustible gases the excess nitrogen required, Nex, is given as
N ex
Ln
N1+ p1 + N 2+ p 2 + N 3+ p 3
pt
percent
(21..11)
where Ln = percentage of combustible mixture at the nose [equation (21.10 with Lmix = Ln ].
21 -
41
The required oxygen content at the mixture nose limit is then simply 20.93 per cent of the air
fraction,
i.e.
Oxygen (nose limit) = 0.2093 (100 - Nex - Ln) percent
(21.12)
Example
A sample taken from a sealed area yields the following analysis.
methane
carbon monoxide
hydrogen
8 per cent
5 per cent
3 per cent
oxygen
inerts
6 per cent
78 per cent
pt = 16 per cent
Construct the Coward diagram for this condition.

Solution
(a) Using equation (21.10) and Table 21.3:
(i) lower flammability limit, Llow:
16
8
5
3
=
+
+
Llow
5
12.5
4
gives Llow = 5.82 percent combustible
(ii) upper flammability limit, Lup:
16
8
5
3
=
+
+
Lup
14
74.2
74.2
gives Lup = 23.56 percent combustible
(iii) nose flammability limit, Ln:
16
8
5
3
=
+
+
Ln
5 .9
13.8
4.3
gives Ln = 6.62 per cent combustible
(b) Equation (21.11) and Table 21.4 give the excess nitrogen required for an extinctive atmosphere
to be
6.62
{(6.07 8) + ( 4.13 5) + (16.59 3)}
N ex =
= 49.24 percent
16
and, finally, oxygen content at the nose limit is given by equation (21.12) as
0.2093 (100 - 49.24 6.62 ) = 9.24 percent
The explosive triangle has now been defined completely and has been constructed as Figure 21.13. The
actual mixture point has also been entered on the diagram and illustrates that although the mixture is not
explosive, it will become so if air is allowed to enter the area.
Coward diagrams are most useful in tracking trend directions of gas mixtures. They do, however, have a
drawback. As each new sample analysis becomes available, the mixture point on the Coward diagram
moves - but the explosive triangle also changes its shape and position. It is analogous to shooting at a
moving target. Fortunately, the calculations are simple and a new triangle can be developed for each
sample in order to follow actual trends. The complete process can readily be programmed for automatic
21 -
42
appearance of the updated Coward diagram and mixture point on a computer screen. Running through
a set of consecutive data for any given sampling location produces a dynamic picture of explosive
triangles and mixture points moving over the screen. This creates a strong visual impact of timetransient trends.
22
Llow
20
18
16
explosive
Lup
methane
8 percent
carbon
monoxide
5 percent
hydrogen
3 percent
14
12
10
Ln
8
actual
mixture
explosive if mixed
with air
4
2
0
10
20
30
40
50
60
70
80
90
100
Figure 21.13 Coward diagram for mixture of gases in the example.

An older method of dealing with variations in gas mixtures is the U.S. Bureau of Mines composite
diagram illustrated on Figure 21.14 and founded on earlier work by Zabetakis et al (1959). In this
diagram, the y axis is an "effective combustible" defined by a weighted combination of the volumetric
percentages of the three combustible gases. The weighting takes account of the explosibility of carbon
monoxide and hydrogen compared with that of methane. The x axis is a combination of the excess
nitrogen and 1.5 times the concentration of carbon dioxide, the 1.5 allowing for the greater extinctive
power of carbon dioxide.
The "excess nitrogen" required here is the percentage of nitrogen in excess of that justified by the
oxygen present.
79.04
Excess nitrogen = actual N 2 O 2
20.93
The set of explosive triangles are approximations based on methane but adjusted for incremental
combined additions of carbon monoxide and hydrogen.
21 -
43
22
20
18
explosive when
mixed with air
16
measured
point
14
12
10
explosive
triangles
8
6
4
non-explosive
2
0
0
10
20
30
40
50
60
70
80
90
100
Figure 24.14 U.S. Bureau of Mines diagram.

Total combustible, pt = methane + carbon monoxide + hydrogen
R = methane/pt
x = excess nitrogen + 1.5 x carbon dioxide
y = methane + 0.4 x carbon monoxide + 1.25 x hydrogen
All gases expressed in percentage by volume.
Example
Using a more detailed analysis of the sample used in the previous example gives
and
methane
carbon monoxide
hydrogen
oxygen
nitrogen
carbon dioxide
8 per cent
5 per cent
3 per cent
6 per cent
68 per cent
10 per cent
21 -
44
Total combustibles, pt = 8 + 5 + 3 = 16 per cent

methane
8
=
= 0. 5
pt
16
methane ratio R
excess nitrogen
= 68
effective inerts (x axis) =
{ 6 79.04 }
= 45.3 percent
20.93
45.3 + (1.5 10) = 60.3 per cent
effective combustibles (y axis) = 8 + (0.4 5) + (1.25 3) = 13.75 percent

Plotting this point on Figure 21.14 shows that it lies above the R = 0.5 triangle, thus agreeing with the
Coward diagram that this mixture is not explosive but will become so if air is added.
21.8. EXPLOSIONS
There has scarcely been a major mining industry that has not been traumatised by underground
explosions of gases, dusts and mixtures of the two. The potential for disastrous loss of life when such an
explosion takes place is very high. Fatality counts have too often been in the hundreds from a single
incident. Fatalities and injuries produced by explosions arise from blast effects, burning and, primarily,
from the carbon monoxide content of afterdamp - the mixture of gases produced by the explosion. In the
dust explosion at Courrieres coal mine in France (1906), 1099 men lost their lives. One of the most
catastrophic explosions on record occurred at Honkieko, Manchuria (1942) when over 1500 miners died.
The carnage that took place in coal mines during the Industrial Revolution (Section 1.2.) was caused
primarily by underground explosions and resulted in the start of legislation governing the operation of
mines.
21.8.1. Initiation of explosions
For an explosion to occur, the same three components of the fire triangle must exist simultaneously;
namely, a combustible material, oxygen and a source of ignition. However, there is a further condition for
mine explosions; the combustible material must be a gas or finely divided dust mixed intimately with the
air and in concentrations that lie between lower and upper flammability limits.
The majority of explosions in mines have been initiated by the ignition of methane. This, in itself, is a
very dangerous occurrence. However, it becomes much worse when the shock wave raises combustible
dust into the air such that it can be ignited by the flame of the burning methane. The resulting dust
explosion is likely to be much more violent than the initial methane blast. Indeed, the majority of methane
ignitions result in blue flames flickering backwards and forwards along the gas:air interface without
developing into an explosion. It is only when the turbulence of the airflow or thermal effects produce a
mixture within the explosive triangle (Section 21.7.3.) that the combustion accelerates into an explosion.
Finely divided particles of any combustible solid can become explosive, including metallic dusts,
sulphide ores and most organic materials. Precautions must be taken against explosions in the
manufacturing, processing and silo storage of many food stuffs including grain. However, in this Section,
we shall confine ourselves to the igniting sources, mechanisms and suppression of explosions that
occur in subsurface ventilation systems.
In Section 21.2. we classified the major igniting sources of fires and explosions in mines. Unfortunately,
explosible mixtures of gases can be ignited by electrical sparks of energy levels as low as 0.3 millijoules,
as illustrated by the tiny spark that ignites gas in a piezo-electric or induction coil cigarette
21 -
45
lighter. It is, therefore, prudent to give a little further attention to the initiation of explosions by sparking
phenomena in addition to those sources of ignition discussed in Section 21.2.
Incendiary sparking arises from heat produced from a chemical reaction. The thermite process involves
the reaction of aluminium powder with iron oxide.
2 Al + Fe2 O3 Al2 O3 + 2 Fe
(21.13)
This process produces so much heat that it has been used on small scale welding operations. A similar
effect occurs when a surface consisting of aluminium, magnesium or their alloys collides with a rusted
steel or iron surface. Incendiary sparks are produced that are well capable of igniting a methane:air
mixture. It is prudent to prohibit the importation of any light alloys into gassy mines and, indeed, this is
enforced by law in a number of countries.
Frictional sparking has been the cause of the increased incidence of methane ignitions on coal faces
and which have accompanied the proliferation of mechanized mineral winning. This is discussed in
Section 21.2.4.
In addition to sparking caused by the misuse or damage to electrical equipment, electrostatic sparks
may also be capable of igniting a flammable mixture of gases. Electrostatic charges are built up on nonconducting (or poorly conducting) surfaces as a regular feature of many everyday operations and,
particularly, at pointed or sharply curved regions on those surfaces. Electrical potentials of 10 000 volts
are commonly generated. This phenomenon may occur, for example, where belts run over pulleys, at
the nozzles of compressed air jets (particularly, if liquid or solid particles are entrained) and within nonconducting ventilation ducts. Even the charge that builds up on the human body in dry conditions can
produce dangerous sparks (Strang and McKenzie- Wood, 1985). In such conditions, workers should not
wear rubber- soled footwear.
All machines with moving parts should be adequately earthed against the build-up of electrostatic
charges. Other devices that are liable to this phenomenon should be similarly protected. "Anti-static"
materials are available for ducting, belts and pipes.
21.8.2. Mechanisms of explosions
An explosion may be defined as a process in which the rates of heat generation, temperature rise and
pressure increase become very great due to the rapidity of combustion through the mixture. In a typical
methane:air explosion, the temperature rises to some 2000 C, i.e. by a factor of about seven. Even
higher temperatures may be reached if the explosion is contained within a sealed volume. The speed of
the process is so great that it is essentially adiabatic. The result is that the pressure in the immediate
vicinity increases to a peak value very rapidly and is relieved by expansion of the air. This produces a
shock wave that propagates in all available directions.
In a mine opening the expansion is constrained by the airway surfaces giving rise to high velocities of
propagation. Initially, the flame front travels more slowly than the shock wave and the explosion is
known as a deflagration. However, if the unburned zone ahead of the flame front becomes more
conducive to combustion by approaching closer to stochastic condition ( approximately 9.6 percent
methane) then the flame front will accelerate and the peak pressure at the shock wave rises. Although
the shock wave velocity also increases, the distance between the two narrows. This may continue until
the flame front catches up with the shock wave (Figure 21.16). The explosion is then described as a
detonation. The speed of the explosion and the peak pressure at the shock wave then escalate
significantly. Conversely, if the unburned zone between the flame front and the shock wave moves away
from stochastic conditions (e.g. by lack of fuel or the presence of stonedust) then the explosion will
weaken.
21 -
46
At the actual site of the initial ignition the flame speed and rate of pressure rise are at their lowest. It is
only as the shock wave and following flame front progress along the airway that their intensity increases,
given the availability of both fuel and oxygen within the explosive range. It is for this reason that victims
at the initiation site of an explosion often exhibit little blunt trauma but have succumbed to carbon
monoxide poisoning. An ignition of a 10 per cent methane:air mixture in a heading may produce flame
speeds of 660 m/s (twice the speed of sound) as it accelerates away from the face of the heading.
Figure 21.14A illustrates the passage of a deflagrating explosion that has been initiated at the face of a
heading. Expansion of the gaseous products of combustion can proceed only away from the face of the
heading. However, that expansion is resisted by the inertia of the outby atmosphere. The result is the
development of the shock wave which is essentially the moving boundary between the normal mine
atmosphere and the elevated pressure of the expanding explosion. Hence, there is a highly
concentrated pressure gradient across the shock wave. It is this shock wave that causes "blast" effects
and can inflict extremely high dynamic loads on objects that lie in the path of the explosion. The shock
wave produces another effect. It causes intimate mixing of any accumulations of methane as it passes
and will produce a continuous explosive path provided there is sufficient gas and oxygen to remain in the
explosive range.. Even worse, it raises dust into the air to produce concentrations which, if the dust is
combustible, will themselves be explosive - resulting in even more favourable conditions for enriching
the fuel path of the explosion.
profile of gauge pressure
h i g h
t u r b u l e n c e
unburned gas
and dust
shock
wave
combustion
products
flame
front
Figure 21.14A In an explosive deflagration the flame front is preceded by a shock wave.
The mixing process is further exacerbated by the highly turbulent conditions that exist between the flame
front and the shock wave. Such turbulence, coupled with the high speed of the unburned mixture can
result in scouring of the sides of the airway, especially if the coal is friable. The impact of flying debris
can also result in significant erosion of the airway as well as dispersion of dust into the air.
If there is any obstruction that prevents the air being projected forward in front of the shock wave the
peak pressure becomes greatly escalated . This occurs when the shock wave encounters a facing seal,
stopping or any other ventilation control. Similarly, high pressures will be generated when a developed
explosion enters a blind heading. In such circumstances victims are typically found to have suffered from
blunt trauma in addition to severe burning. The latter is thought to be caused by swirling flames that exist
until extinguished by lack of oxygen.
These are the basic mechanisms that drive explosions through subsurface ventilation networks. Bends,
junctions, obstructions and the availability of fuel and oxygen all affect the rate of propagation. A mine
21 -
47
explosion is a highly dynamic phenomenon and seldom reaches steady-state even for a few
milliseconds. For most of its life it is either proliferating or subsiding.
21.8.2.1. Gas explosions
Explosions of methane from coal are by no means confined to mines. Ignitions have too frequently
occurred in storage silos and in coal cargo ships, causing great damage and loss of life. Methane
concentrations of 40 per cent have been measured in such facilities. Layering phenomena and the
production of explosive mixtures of methane and air make unloading operations particularly hazardous.
Precautions against ignitions in these circumstances include good ventilation of the facility, monitoring
for methane and the gaseous products of spontaneous combustion, ready availability of nitrogen or
carbon dioxide cylinders for inertization, strict control of ignition sources and prohibition of smoking in the
vicinity. Bacteria have also been used to convert the methane into carbon dioxide (Kolada and
Chakravorty, 1987).
Incidences of hydrogen explosions at battery charging stations and ignitions of oil vapours from
machines have been reported. However, the vast majority of gas explosions in mines have involved
methane. Modern systems and standards of mine ventilation are successful in maintaining general body
concentrations of methane at safe levels throughout the ventilated areas. However, strata emissions of
gas are often at methane concentrations of over 90 per cent. It follows that between the points of
emission and the general body of the airstream, zones of an explosive mixture must exist. The
ventilation arrangements should ensure that these zones are as small as possible and, wherever
practicable, maintained free from igniting sources. For example, the danger of methane, coal dust, air
and sparks existing simultaneously at the pick-point of a coal winning machine can be reduced by pickface flushing or jet-assisted cutting (Section 20.4.1.1.).
Dangerous accumulations of methane in roof cavities should be avoided by filling those cavities with an
inert material. Air velocities should be sufficient to prevent methane layering (Section 12.4.2.) and
longwall systems should be designed to maintain the gas:air interface within the goaf (gob) and without
flushing on to the faceline. In appropriate circumstances, methane drainage will reduce the amount of
gas that enters the mine ventilation system (Section 12.5.).
21.8.2.2. Coal dust explosions
The danger of coal dust being explosible in air depends upon a number of factors including:
(a) Concentration of the dust and presence of methane: The flame of a dust deflagration is
propagated by the combustion of each discrete particle in turn. If the particles are sufficiently
far apart then the heat produced by a burning particle will be insufficient to ignite its closest
neighbour. The flame will not propagate. The lower flammable limit is attained when the
particles become sufficiently close to allow sequential ignition. At the other extreme, if the
dust particles are too close together, there will be insufficient oxygen to allow complete
combustion. The system will become internally fuel-rich and the rate of heat generation will
fall. The upper flammability limit is attained when the temperature is reduced below the
ignition point of the dust.
The lower flammability limit may be of the order of 50 g/m3 with maximum explosibility at
150 to 350 g/m3 depending upon the volatile content of the coal (Holding, 1982). The upper
flammability limit might be as high as 5000 g/m3 (Strang and McKenzie-Wood, 1985) but is
dependent on the availability of oxygen. To put these flammability limits into perspective, it
should be recalled that threshold limit values for respirable coal dust concentrations can be
as low as 2 milligrams per cubic metre (Table 19.2). A concentration of 50 g/m3 produces a
suffocating atmosphere. It is unlikely that such concentrations of airborne dust will exist
under normal conditions in the active branches of a mine ventilation network. Nevertheless,
in the absence of dust suppression, explosible dust concentrations might occur around the
cutting heads of rock-breaking machines. Furthermore, a very thin coating of dust on the
surfaces of airways or a conveyor can produce an explosive atmosphere when disturbed by
a shock wave or highly turbulent conditions.
21 -
48
Figure 24.14B. The flame of a coal dust explosion emerging from a 2.5 m (8ft) high test gallery at the
Barbara Mine, Poland
Coal dust and methane are synergistic in their explosibility characteristics. The flammability
limits of each of them are widened in the presence of the other. The relationship between the
two varies with the volatile content of the coal. A typical example, based on results reported for
an Australian coal, is shown on Figure 21.15. It should be appreciated that when heated, coal
emits its volatile content commencing with methane. Hence, a coal dust explosion involves the
combustion of gaseous fuels as well as carbonaceous solid particles. It is common for partially
de-volatilized coal dust to impact and adhere to solid surfaces during a coal dust explosion. This
leaves coked material and carbon filaments on supports, equipment and airway surfaces which
assist investigators tracing the role and path of coal dust in an explosion. On the other hand, a
high velocity flame may leave little trace of coking.
(b) Fineness of the dust: There appears to have been little work done in establishing definitive
relationships between explosibility and the size of dust particles. However, the increased surface
area and more intimate mixing with the air given by the finer particles suggest that explosibility
will increase as the average particle size decreases. An accepted assumption is that any
combustible dust which can be raised into the air is explosible. This encompasses all particles of
less than some 250 microns diameter.
(c) Type of coal: There is considerable evidence to show that coal dust explosibility increases
with volatile content (e.g. Holden, 1982). Hence, low rank coals are more prone to dust ignitions.
Anthracite dust is normally considered to be non-explosible in mining conditions. However, with
a sufficiently high energy of initiation, low volatile coal dusts can produce violent explosions. The
disastrous explosion in Manchuria (1942) occurred while mining a coal of volatile content 15 to
19 per cent. The explosibility of coal dust decreases with respect to ash and moisture content as
well as with age. The latter factor is considered to be due to the coating
21 -
49
of a partially oxidized layer on older dust particles. It follows that coal dust on, or close to, a
working face is more liable to ignitions.
(d) Strength of initiating source: The energy level of the initial ignition plays a large role in
governing the rate of propagation and growth in power level of a dust explosion. The majority of
coal dust explosions in mines have been initiated by a methane ignition, this providing the
starting shock wave and flame front. However, any other igniting source of sufficient power can
result in a dust explosion in the absence of a flammable gas. The Courrieres dust explosion
(1906) is thought to have been caused in a methane-free atmosphere by a case of prohibited
explosives (Cybulska, 1981).
To comprehend the power levels that can be attained by coal dust explosions, let us return to the matter
of a spearheading shock wave driven by a more slowly advancing deflagrating flame front. A flame
velocity of 50 m/s will produce a shock wave velocity of some 375 m/s. (A lower flame velocity is unlikely
to propagate). However, Figure 21.16 shows that if the dust concentration and methane favour the
growth of the explosion then the flame velocity will increase at a greater rate than that of the shock wave
until both exceed some 1100 m/s. Beyond that, the process may develop further into a detonation when
adiabatic compression in the shock wave can produces temperatures that exceed the ignition point of
the fuel. Gas and dust particles no longer have to wait until they are ignited by a following flame. They
ignite spontaneously. The flame front and shock wave then advance in unison and may reach speeds of
over 2000 m/s (six times the speed of sound). It is little wonder that coal dust explosions are so feared
throughout the world of mining.
Figure 21.16 also illustrates the dynamic pressure developed across the shock wave. Even at the lower
end of this curve the pressures are sufficient to disrupt doors, stoppings and air-crossings, while the
upper end explains the devastation that can be caused by a well-developed explosion.
During post-explosion inspections, the dominant path of the explosion may be deduced from the
twisting of steel supports, the distortion of track rails, the direction of failure of strong stoppings
and the dislocation of heavy equipment. The displacement of less sturdy objects may be
misleading and can be caused by secondary deflections of the shock wave, particularly at bends
or junctions. Furthermore, an explosion is often followed rapidly by an implosion of fresh air due to
rapid cooling and contraction of the afterdamp gases. If open burning is in progress then the fresh
air can initiate repeated explosions. These phenomena can also leave a confused picture for
subsequent investigators.
The passage of an explosive flame is unlikely to ignite solid coal. However, paper, clothing and,
especially, feeders of methane may be left burning if there is sufficient oxygen remaining in the area.
This is most probable in the fringe areas of the explosion and can initiate further explosions of any
methane that continues to be emitted into the area. The probability of such subsequent explosions is
greatly increased if the initial explosion(s) have disrupted the ventilation system.
21 -
50
Figure 21.14C Photographs of post-explosion No.1 Surface Portal, dislodged and damaged steel arches,
and remnants of ducting. Westray Mine Public Inquiry, Exhibit No. 59 (Richard, 1997).
21 -
51
Figure 21.14D Photographs of roof falls, dislodged steel arches, and damaged equipment.
Westray Mine Public Inquiry, Exhibit No. 59 (Richard, 1997).
21 -
52
40
30
100
100
80
80
60
60
40
40
20
20
20
10
4.8
5.0
5.2
5.4
5.6
5.8
Figure 21.15 Example of the reduction in the lower flammability limit

of coal in the presence of methane. Other curves vary
according to the rank of the coal.
(Based on work reported by Jensen et al (1989))
0
0
20
40
60
80
100
Figure 21.16 Variation of shock wave velocity and dynamic pressure

with respect to flame velocity.
(Developed from data reported by Strang and MacKenzie-Wood, 1985)
21 -
53
21.8.2.3. Sulphide dust explosions

Although explosions of sulphide ore dust are a hazard in some metal mines, they do not have the destructive
power of coal dust explosions and have attracted much less attention by researchers (Holding, 1975). These
explosions are initiated, primarily, during blasting operations of ores that contain more than 40 per cent
sulphur. The main danger to life is production of the highly toxic sulphur dioxide gas which may be retained
for several hours, particularly if ventilation controls have been disrupted. Furthermore, the gas reacts with
water vapour causing corrosion problems from the sulphuric acid fumes that are produced. A number of
precautionary measures have been suggested including hosing dust depositions from the walls of stopes
prior to blasting and the use of water atomizers before and during blasting (Section 20.3.4.). Other techniques
involve stemming the blast holes with water ampoules or powdered limestone and inserting explosive
charges with zero-time detonators into bags of hydrated limestone powder suspended in the stopes. The
latter technique fills the air with inert and suppressant dust particles immediately prior to detonation of the
main round (Hall et al, 1989).
21.8.2.4. Dust explosibility tests
Explosibility tests for mine dusts vary considerably in the equipment and procedures employed. The essential
sequence consists of producing a dust:air or dust:oxygen mixture and applying an igniting source, usually an
electrical spark of constant energy level. The parameters that are varied include the dust concentration, the
type of dust, added amounts of stonedust (usually powdered limestone) and methane concentration. The
results may be recorded simply as "ignition" or "no ignition", or in the more sophisticated tests, the rates of
increase of pressure and temperature and the peak values of those variables (Jensen et al, 1989).
Considerable variations occur in the reported explosibilities of similar dusts. These may be attributable, in
part, to differences in test conditions. The results are influenced, for example, by variations in the uniformity of
the dispersed dust, means of measuring dust concentration, strength of the igniting source and turbulence
within the test chamber. This latter factor also indicates that whilst laboratory test results may provide
valuable comparisons between differing dusts, they are not necessarily indicative of actual behaviour within
the mine environment. Full scale galleries have been constructed in several countries for investigations into
mine fires and explosions.
Indices of dust explosibility have been defined and referred to specific equipment and procedures in differing
countries (e.g. Holding, 1982). While these produce useful relative information, caution should be applied
when comparing one index with another.
21.8.3. Suppression of mine explosions
The primary safeguards against mine explosions are, once again, a well designed and operated ventilation
system, planned maintenance of equipment and good housekeeping to control accumulations of combustible
dust and to ensure adequate applications of stonedust. This brings us to the most widely used method of
suppressing the propagation of mine explosions.
21.8.3.1. Stonedust and water barriers
Finely divided limestone dust serves at least two purposes when mixed intimately with coal dust in a mine
airway. First, it will act as an inert dilutent when the mixture becomes airborne, serving to increase the
distance between combustible dust particles. Secondly, particles of stonedust will absorb heat and reduce the
ability of airborne coal dust to propagate a flame. Provided that there is sufficient stonedust present, this
technique provides an efficient means of suppressing the full development of an explosion.
It is important that the stonedust be mixed well with the coal dust. Tests have shown that strips of coal dust
interspersed with strips of stonedust will sustain an explosion. Furthermore, combustible dust on a conveyor
may propagate an explosion even when the airway is adequately stonedusted. Precautions should, therefore,
be taken to ensure that the stonedust is spread on roof, floor and sides often and consistently. An efficient
21 -
54
means of doing this is to employ trickle-dusters. These are devices that emit stonedust into the air at a
controlled rate. The stonedust mixes with the airborne coal dust and settles with it to produce a uniformly
mixed deposition on all surfaces.
An additional means of using stonedust, particularly applicable to single entry longwall systems, involves
stonedust barriers. These are boards supported on pivots across the airway, usually near the roof, and
loaded with stonedust. Dust loadings vary from 30 to 60 kg per metre length of board. A number of boards
are located close together within a short length of airway to form the complete stonedust barrier. The intention
is that the boards and their contents will be dislodged by the shock wave of an explosion to produce a high
concentration of airborne stonedust at the time the flame front arrives and, hence, prevent its further
propagation.
The locations of stonedust barriers should be considered carefully. If they are too close to the seat of the
explosion then the flame front may have passed before the stonedust has become adequately distributed; too
far away then the stonedust will become too dispersed by the time the flame front arrives. A sensible
arrangement is to have a lightly loaded (30 kg/m) barrier, that can more easily be displaced, located within
some 200 m of face operations and more heavily loaded barriers further outby. While stonedust barriers may
be sited in all district intakes and returns, conveyor entries are particularly recommended to be fitted with
these devices.
Stonedust barriers do have some disadvantages. They must be inspected regularly to ensure that they
remain capable of being dislodged - yet this renders them prone to accidental or mischievous disturbance. In
damp conditions, moisture-proofed stonedust should be employed to ensure that the dust will be dispersed
adequately by a shock wave. Stonedust barriers may fail to suppress methane explosions. Furthermore, if an
explosion is allowed to develop into a detonation then it is unlikely to be halted by a stonedust barrier. Indeed,
the objective of barriers is to suppress an explosion at an early stage of its development.
A water barrier is intended to serve the same purpose as a stonedust barrier. In this case, troughs holding
some 40 to 90 litres of water are positioned across the airway and attached to roof supports. The troughs
themselves are constructed from polystyrene foam or a similarly weak material so that they will disintegrate
when subjected to a shock wave travelling at 100 m/s. The cross-section becomes filled with water droplets.
Suppression of the flame is achieved by the cooling effect of evaporation and displacement of oxygen.
21.8.3.2. Triggered barriers and explosion detectors
Stonedust and normal water barriers both suffer from the disadvantage that they rely upon the shock wave to
disperse the stonedust or water. Triggered barriers are designed to incorporate an internal power source. A
typical design consists of an enclosed water tank connected to a concentrated array of nozzles along a short
length of airway. A containment diaphragm prevents flow from the tank to the nozzles under normal operating
conditions. A bottle of compressed nitrogen or carbon dioxide is located inside the water tank. The gas
container is also fitted with a rupture disk.
Upon activation, a heater within the gas bottle causes the gas to expand and puncture the rupture disk. This
causes very rapid pressurization of the water, breakage of the containment diaphragm and delivery of water
at high pressure to the nozzles. Several hundred litres of water can be dispersed in less than one second.
Some designs utilize a powdered flame suppressant in place of the water and a soft explosive instead of the
gas container.
Activation of a triggered barrier is initiated by an electrical signal from a detector device located closer to the
working area, where an explosion is more likely to commence. Infra-red, ultra-violet, temperature and
pressure pulse sensors have all been employed as detection devices.
In addition to the active power source to disperse the fire suppressant, an advantage of the triggered barrier
is that an optimum distance can be selected between the sensor and barrier. This ensures that the barrier will
be activated at the correct moment with respect to the approaching flame front.
21 -
55
21.8.4 Explosions in sealed areas.

21.8.4.1 Causes of explosions in abandoned areas.
The majority of mine explosions are initiated by operations in the active zones of the mine, particularly arising
from ignitions of methane at the pick points of mining machines. These explosions have sometimes caused
damage to seals separating active mining zones from abandoned workings and allowing methane to flood
into the current ventilation network. The fuel provided by this methane can add to the proliferation and
violence of the explosion (for example, the Westray Mine explosion in Nova Scotia, 1992, Richard, K. P.,
1997). It was for this reason that the designs of seals to unventilated old workings sought to be capable of
withstanding explosion forces. It has now been accepted that explosions can occur within abandoned areas
(example Sago Mine, West Virginia, 2006). It is difficult to determine with certainty the causes of such
explosions. The possibilities include:
friction
The collapse of strata that contains quartz (particularly sandstones) is capable of producing incendiary sparks
by friction of rock against rock or rock on steel (Powell,1975, Nagy, 1960)
lightening
Lightening strikes on surface may be conducted for discharge in mine workings, including abandoned areas,
via borehole casings, pipes, cables or any other conductive path (Sacks, H.K. and Novak, T., 2005).
adiabatic compression
The atmospheric pressures developed by sudden collapses of strata into sealed areas can cause rises in
temperature to the ignition point of explosive mixtures (McPherson, 1995; Lin, W. 1997).
spontaneous combustion which reaches open areas. (See Section 21.4)
21.8.4.2 Recognition of the hazard.

Following the Sago Mine explosion (West Virginia, 2006, 12 fatalities)) and Darby Mine explosion (Kentucky,
2006, five fatalities) urgent new research was initiated into safeguarding miners against the effects of
explosions occurring in abandoned areas of coal mines. The National Institute for Occupational Safety and
Health (NIOSH) in the United States reported that during the period 1986 to 2006 there were 12 incidents of
explosions in sealed areas of mines in the United States (Zipf et al, NIOSH, 2007). In ten of those explosions
seals were destroyed indicating that the U.S. design standard then in use 140 kPa (20 psig) was far from
adequate. This contrasted with standards of the 345kPa (50psig) in the UK or 500kPa (73 psig) in Poland and
Germany) where no seal failures due to explosions in old workings were reported in the same period.
One effect of the Sago Mine explosion was the issuance of a Mine Safety and Health Administration (MSHA)
Emergency Temporary Standard that seals should be designed and constructed to withstand an explosion
pressure of 345 kPa (50 psig) and that new research on mine seals be initiated (see Section 21.1). This
provided the impetus for NIOSH to embark on such research. The first of their reports was issued in 2007
(Zipf et al) with the promise of further investigations to be conducted.
21.8.4.3 Approaches to the problem and the initial NIOSH Report.
An approach to the safety of abandoned workings has been either to ventilate those old workings
continuously in order to remove methane or to seal them off and allow an inert atmosphere to be established.
Both of those techniques have disadvantages. With regard to continuous ventilation, even when air passes
slowly through caved areas it is still possible for pockets of methane and explosive mixtures to exist within
those areas, particularly in inclined workings. Secondly, it is most unwise to pass air through old workings
where the coal is prone to spontaneous combustion. And thirdly, recalling that ventilating fans are often the
largest consumer of electrical power at a mine, the associated costs can become very high if airflows are
used to ventilate increasing areas of abandoned workings.
With regard to sealing off an abandoned area, the intention is that an inert atmosphere be established in
sealed areas by the combined effect of oxidation (replacing oxygen by carbon dioxide) and emissions of
21 -
56
methane. However, this can be effective only if the seals are designed to minimize leakage. Furthermore,
recalling from Section 4.2.3 that the effective combined resistance, Reff, of seals constructed essentially in
parallel falls dramatically as their number increases it becomes important that the number of seals into the
area is also minimized.
R eff
R
n2
where R is the resistance of a single seal and n = number of seals.
An added complication is that seals will breathe as the external atmospheric pressure varies. (Section
4.2.2). The effect is that air will leak through or around seals during periods of a rising or falling barometric
pressure. For these reasons an inert atmosphere may never be attained in all parts of an abandoned and
sealed area. In particular, the breathing of seals can produce a time variant zone of explosive mixture
immediately inby each seal.
The NIOSH Report of 2007 (Zipf et al) described a rigorous investigation involving laboratory (closed volume)
tests, full scale gallery tests and the utilization of mathematical models. These tests examined
the dependence of pressures developed by an explosion on the volume of explosive mixture,

the case of products of combustion expanding into inert voidage and caved areas
the case of gases resulting from the explosion venting into other active or inactive parts of the mine,
the run-up distances through which an explosion can proliferate up to becoming a detonation (Section
21.8.2.2).
On the basis of their investigation, NIOSH suggested that if abandoned working are to be sealed rather than
ventilated then mine management decide on a case-by case basis whether to
(a) build unmonitored seals that will withstand most explosions within old workings, or
(b) utilize lighter seals but to monitor continuously the atmosphere in a zone inby the seals.
Table 21.5 gives a synopsis of the applications of unmonitored and monitored seals that are described in
more detailed in the NIOSH 2007 Report.
At the time of writing, NIOSH is continuing with its investigations. In particular, further work should provide
guidance on the construction of seals.
Australia has already adopted regulations that consider the conditions under which unmonitored
and monitored seals can be used (Oberholzer and Lyne, 2002). Monitoring is normally carried out
by means of tube bundle systems (Section 21.7.2). Furthermore, in addition to monitoring, many
Australian mines utilize inertization techniques to maintain an inert atmosphere inby the seals
(Section 21.6).
21 -
57
Unmonitored
Design
pressure
Monitored
Design
pressur
e
345 kPa
Panel and district seals.

Length of sealed volume >50m.
100% filled with explosive mix,
completely confined, unvented
4400 kPa
Panel and district seals.

Length of sealed volume >50m in any direction .
Monitoring ensures that that maximum length of
explosive mix behind seal < 5m and volume of
explosive mix < 40% of total sealed volume. Some
venting into inby inert atmosphere. (typical
situation with breathing seals)
(50 psig)
Panel, district and crosscut seals.

Length of sealed volume <50m in
any direction. Sealed area
completely filled with explosive mix.
Confined but can vent somewhat
into gob
800 kPa
(640 psig)
345 kPa
(120 psig)
Panel , district and crosscut. seals.

Length of sealed volume <50m in any direction
Monitoring ensures that maximum length of
explosive mix behind seal < 5m
Volume of explosive mix < 40%total sealed
volume.
(50 psig)
Table 21.5 A simplified synopsis of the design pressures recommended for unmonitored
and monitored seals.
(Derived from Zipf et al. Explosion pressure design criteria for new seals in U.S. coal mines. NIOSH IC 9500, 2007)
Figures 21.17 and 21.18 show layouts that illustrate the possible locations for district, panel and cross-cut seals.
21.8.4.4 Requirements of seals

Due to the many uncertainties associated with atmospheric conditions and changing geometries
within sealed areas it is difficult to know with any certainty the distributions and concentrations of
gas mixtures much beyond the seal locations. Hence the conservative approach is to assume
that well developed explosions can occur within the sealed area and to construct seals that can
withstand such explosions . The requirements of such seals are:
they be as leakproof as practicable, taking into account the number of seals to be built
capable of withstanding repeated explosions
explosion proof in both directions taking into account the shear strength of strata in the roof,
floor and sides, and
that they can withstand strata convergence over the long term.
It may prove difficult to design practicable and cost-effective single-wall seals that can meet all of
these requirements. A wise objective of further investigations would be aimed at designing mine
layouts that minimize the number of seals required and the utilization of double walled seals with
an infill that is economically and readily available. Figure 21.19, with or without the sampling
arrangements, illustrates the type of seal that has proved successful in meeting these listed
requirements.
It is a prudent precaution to pre-prepare the sites of future seals as soon as practicable after
excavation of the relevant airways. Such preparation should include cut-outs for the seal walls in
airway sides and the storage of materials for construction of the seals. Such pre-preparation will
prove invaluable in cases of emergency. Furthermore, airways should be liberally stone dusted
both inby and outby the sites of seals.
21 -
58
panel
seals
preprepared
sites of
district
seals
cross
-cut
seals
cross
-cut
seals
preprepared
sites of
district
seals
panel
seals
Figure 21.17 Panel extraction with panel and cross-cut seals.

Multiple cross-cut seals are required.
21 -
59
panel
seals
preprepared
sites of
district
seals
preprepared
sites of
district
seals
panel
seals
Figure 21.18 Barrier pillars left between panels (subject to rock mechanics considerations.)
No cross-cut seals required.
21 -
60
sampling
tube
sealed
area
infill
3m or (Height + Width) + 0.6

2
Figure 21.19 Example of an explosion proof seal.
21.9 PROTECTION OF PERSONNEL

21.9.1. Training and preparedness
In all underground emergencies, the first priority is the safety of personnel. The surest protection
against the initiation and hazards of mine fires and explosions is training and practice of safety
procedures. Classes and practical sessions should be held, not only for new recruits, but at regular
intervals of time for all employees. These sessions should include discussions on the causes of
fires and explosions and how they propagate. The elements of subsurface firefighting and
operation of fire-suppression equipment should be outlined. It is prudent to ensure that supervisors
and other selected members of the workforce should receive additional instruction and practice in
firefighting. As in mine rescue work, competitions between teams can do much to foster interest.
This further training may include the use of temporary stoppings, airflow control and air pressure
management. However, training of all underground personnel should concentrate particularly on
warning systems and location of trapped persons

self rescuers
escapeways and
refuge chambers.
These four items are discussed in the following subsections.

Mine management plays a crucial role in minimizing the risks of fires and explosions, and in
responding correctly when they do occur. In addition to matters of training, the planned
arrangement of services dictates the efficiency of emergency response (Section 21.10.).
Communications, routes of water pipes, cables, airflows and escapeways, availability and
effectiveness of firefighting equipment, reliable environmental monitors and an integrated fireprotection policy (Sheer, 1988) can make the difference between a localized ignition that is quickly
extinguished and disastrous loss of life. Ventilation and safety engineers should engage in mock
scenarios and computer simulations of emergency situations. Such exercises assist in raising
questions related to response strategies and mine preparedness such as establishing distances
from work locations to refuge chambers and escapeways.
21-
61
All underground workers should be required to engage in fire drills and emergency evacuation
procedures at least once in each three months. These should involve leaving the mine by
escapeways other than the normal travelling routes. Such drills should be accompanied by
discussions on choices of escape routes and ventilation systems, and should also encourage
familiarity with the layout of airways within the mine.
It is a useful exercise, as part of emergency preparedness, for mine management to establish prior
friendly relationships with local police, fire, hospital and news agencies. This can be further
facilitated by arranging occasional tours of the mine for such people.
21.9.2. Methods of warning and locating personnel
The rates at which mine fires can propagate and products of combustion spread throughout an
underground network of airways make it vital that all personnel in the mine be warned of the
emergency as quickly as possible. This can be difficult because of the distances involved and wide
dispersal of the workforce, particularly in metal mines. Video cameras may be located at strategic
locations although these might be obscured by smoke.
Most mining industries employ "firebosses, firemasters" (or similar titles) whose task is to patrol the
mine before and, perhaps, during each working shift and following blasting operations to search for
any indications of a fire. To ensure adequate and timely coverage of the mine, tags are left at
strategic points showing the time and date of the last visit. Fire patrols should follow the direction of
airflow in order to more easily locate the probable source of any smoke or fire odour.
Modern paging-type mine telephone systems enable voice warnings to be transmitted to main
work areas and along major transport routes. These are particularly suitable for concentrated
mining methods such as the longwall system. However, the extension of paging techniques along
the several hundred kilometres of active airways in a large metal mine would be cost prohibitive.
The methods that have been devised to transmit fire warnings to a widespread workforce in
underground openings include:
stench warning systems,
radio signals and
more recent systems
Following receipt of any form of fire warning, mine workers should immediately engage in the
emergency evacuation or personnel protection procedures in which they should have been trained
and practiced (Section 21.9.1.).
21.9.2.1. Stench warning systems
Versions of this technique have been employed in large metal mines since the 1930's. The
procedure involves releasing a gas having a very distinctive odour into intake airways or
compressed air systems. Having been exposed to low concentrations of the stench gas during
training sessions, miners recognize the odour and can initiate emergency procedures. The gases
used resemble those introduced into natural gas distribution systems. Ethyl mercapton mixed into
freon has been commonly employed. However, this can be toxic, corrosive and produce an
unbearable stench at high concentrations. Furthermore, it may lose its potency when transported
through steel pipes. A safer and healthier alternative is tetrahydrothiaphene (Ouderkirk et al,
1985).
A basic disadvantage of the stench warning technique is that it relies on the mine ventilation
system. Transmission times of the stench depend upon the large range of air velocities that exist
within a mine ventilation network. In openings of large cross-sectional area, travel times can
21-
62
become dangerously high to serve as a fire warning system. Furthermore, the concentration of the
odour will also vary from very high, close to the emission point, to barely detectable in remote
areas of a complex ventilation network. To add to the difficulties, the ventilation system may have
been modified or disrupted by the fire (Section 21.3.2.).
These drawbacks can be minimized by selecting multiple injection points in locations reasonably
close to major work areas, in addition to downcast shafts or other primary inlets. Stench gas
injector mechanisms should incorporate pressure balance arrangements and controlled release
rates to prolong the warning and to avoid short peaks of excessively high concentration. Inby
injectors should be capable of remote activation. The choice of injector sites and release rates
should be made carefully. The objective is to carry a clearly detectable odour as rapidly as
possible to all parts of the mine where personnel may be working. Ventilation network analysis
packages that incorporate gas distribution modules are very helpful in selecting optimum locations
and release rates (Section 7.4.4.).
21.9.2.2. Ultra-low frequency radio signals.
Many investigations have been conducted into the use of radio waves for warning mine personnel
of an emergency condition and, also, for locating persons who have become trapped. High or
medium frequency radio signals are limited to line-of-sight transmission or depend upon the
existence of metal conductors such as pipes, cables or conveyor structure. For personnel warning
or location, the radio signals must be capable of transmission through rock and of being detected
by lightweight and low-powered personal receiver units.
Older developments in this area have involved ultra-low frequency devices. At frequencies in the
range 630 to 2000 Hz and employing transmitting powers of 1 kW, rock penetration distances of
over 1600 m have been achieved (Hjelmstad and Pomroy, 1990). This indicates that transmitters
located on surface will be suitable for most mines, while underground transmission sites will serve
for the deeper workings. The signal strength decays inversely with the cube of the distance
through the rock.
The transmitter consisted of a radio (electromagnetic) signal generator, amplifier and antenna. The
latter may be made from ten coils of insulated copper wire coiled into a loop of some 30 m
diameter. The personal receiver units are small, intrinsically safe and may be powered by the
wearer's caplamp battery. The pencil size receiver antenna is formed from a high-permeability
ferrite core wound with copper wire and is very efficient in detecting electromagnetic radiation.
More highly powered units can be fitted to vehicles or other equipment. The transmitter and
receivers are tuned to a common resonant frequency to discriminate against electromagnetic noise
or stray signals. The voltage produced by a receiver unit is amplified and may then be used to
generate an audio-visual warning. Incorporation into the caplamp unit can produce on-off blinking
of the light until nullified by operation of a switch. Research continues into the use of rockpenetrating radio waves for voice communication. Again, this will serve for both warning and
personnel location purposes.
21.9.2.3. More recent systems
Following the Sago Mine explosion ( West Virginia, 2006) a plethora of wireless tracking systems
have been developed to locate persons in underground mines, particularly in Australia and the
United States. These systems have included hand-held radios, leaky feeder and radio frequency
devices. The first of these to be approved by the U.S. Mine Safety and Health Administration is
normally energized via d.c. power cables with back-up battery power in case of failure of the cable
system. (MSHA, 2008)3. The units can be used to track miners both before and after an
emergency situation. Text messaging as well as gas detection can be added to the system to
maintain communications between equipped miners and a central location where all such
3
Mine Tracer Miner Location System, Venture Design Services. Inc.
21-
63
communications are logged . At the time of writing other tracking devices are under investigation
for approval in U.S. mines.
21.9.2.4. Locating trapped persons
The situation has all too often arisen where miners have become trapped in locations that are
inaccessible because of accumulations of gases, products of combustion, water or falls of roof. To
facilitate rescue operations, it is necessary to identify the locations of the trapped personnel.
A very old method that has often led to life-saving operations is the seismic technique. An
explosive charge or heavy hammer (pile-driver) blow on the mine surface may be heard or felt by
personnel in relatively shallow mines. In deeper operations, similar seismic signals may be
generated from underground locations. On hearing these signals, the trapped personnel respond
by hammering on pipes, rails or, simply, the floor or walls. If this response is detected by the ears
or geophones of rescue personnel then mine maps are consulted, if necessary, to select the
probable location(s) of the trapped persons. Pilot boreholes may be drilled to those locations to
supply food and air while rescue operations proceed.
21.9.3. Self-rescuers
An estimated 80 to 90 per cent of fatalities in mine fires or explosions are caused by carbon
monoxide poisoning. Self- rescuers are compact versions of breathing apparatus and are intended
to allow the wearers to pass through atmospheres that are contaminated by smoke or carbon
monoxide. In many countries, it is mandatory for all persons who enter an underground mine to
carry a self-rescuer. The lighter versions can be worn on the belt of each person who ventures
underground while caches of the heavier self rescuers may be kept close to work areas and on
vehicles. The purpose of self-rescuers is to allow the wearers to escape from or through
contaminated atmospheres. Indeed, they are often known as escape breathing apparatus (EBA's).
They are not intended for rescue operations or other type of work and should not be confused with
the more specialized breathing equipment used by trained rescue teams.
The two essential features of self-rescuers are the types of atmosphere in which they are effective
and the time they allow for escape. Their operational duration depends not only upon the type of
self-rescuer but also:
manual effort (e.g. speed of walking)
breathing habits, experience, physical condition and mental state of the wearer, and
environmental conditions.
Although many lives have been saved by self-rescuers, others have been lost even when such
devices were available. In many of these cases, the error has been in failing to don the apparatus
at the appropriate time. This matter should be stressed heavily during training sessions. In
particular, self-rescuers should be put on as soon as it is suspected that the air is, or may become,
contaminated, i.e. by means of telephone messages, personnel warning systems (Section 21.9.2.),
shock waves or strong pulses in the ventilation, unusual odours or the appearance of smoke.
Furthermore, during training, personnel should be required to practice unpacking and donning the
apparatus. Training versions of the units are available for this purpose. Experience has shown that
classroom instruction on the opening and donning of a self-rescuer is insufficient to emulate the
physical and psychological conditions under which it may be used in an emergency. It is
recommended that the procedure be practiced in a smoke-filled atmosphere or, at the very least,
with closed eyes.
There are basically two types of self-rescuer units, the filter self-rescuer (FSR) and the selfcontained self-rescuer (SCSR).
21-
64
21.9.3.1. Filter self-rescuers (FSR's)

Each of these small and compact units fits inside a hermetically sealed plastic or stainless steel
case which is worn on the belt. To use the device, the seal is broken by a lever arrangement, the
mouthpiece inserted inside the lips and gripped by the teeth, and nose clips put on. The complete
unit is held close to the face and chin by head straps. Air is drawn through three types of filter
before passing through a heat exchanger to the mouthpiece. Exhaled air also traverses the heat
exchanger and through a one-way valve to the external atmosphere.
The first level of filtration simply removes dust particles. The second filter is a drying agent typically 9 per cent lithium chloride and 91 per cent calcium bromide impregnated into activated
charcoal (Strang and MacKenzie-Wood, 1985). The activated charcoal assists in removing sulphur
dioxide, hydrogen sulphide and oxides of nitrogen. The third level of filtration contains a catalytic
mixture of granulated manganese dioxide, copper oxide and a little silver oxide. The mixture is
known widely as hopcalite. This catalyst converts carbon monoxide into carbon dioxide with close
to 100 per cent efficiency. It is, however, poisoned by water vapour. Indeed, the duration of the
device is governed by the drying agent in the second level of filtration. The period of operation is,
typically, about one hour for saturated conditions and carbon monoxide concentrations of up to 1.5
per cent.
Filter self-rescuers suffer from two disadvantages. First, they rely on there being at least 16 per
cent oxygen in the ambient atmosphere (Section 11.2.2.). Secondly, both the drying filter and the
hopcalite involve exothermic reactions. Indeed, the temperature of the inhaled air is an indication
of the level of carbon monoxide present. At high concentrations, the temperature of the filtered air
may exceed 90C and cause great discomfort including blistering of the lips and mouth. However,
removing the mouthpiece for temporary relief in such circumstances is likely to be fatal. The small
heat exchanger close to the mouth piece absorbs heat from the filtered air and rejects it to the
exhaled air. Because of these limitations it is prudent to replace filter self-rescuers by selfcontained self-rescuers with the possible exception of small mines with ready access to the
surface.
21.9.3.2. Self-contained self-rescuers (SCSR's)
As the name implies, these supply all the respiratory needs of the wearer and are independent of
the gaseous constituents of the ambient atmosphere. A considerable variety of self-contained selfrescuers have been produced. The main disadvantage is their weight and bulk. Current
generations of SCSR's vary from 2 to 5 kg and may not be sufficiently convenient for personal
wear - hence, requiring caches of the units to be available close to active working areas. Research
continues in attempts to develop a self-contained self-rescuer that is sufficiently light and compact
to be worn on the belt while giving the 60 or 90 minute operating life that may be mandated by law.
National and state legislation should be consulted for regulations that govern the requirement,
types and operating durations of self-rescuers.
As with the FSR's, the self-contained versions are hermetically sealed in polypropylene or metal
containers to ensure a long shelf life. However, if the seal is broken or damaged, the unit must be
replaced. Periodic water bath tests should be carried out to ensure that the casings remain airtight.
The units include noseclips, fitting straps and goggles. Two types of self-contained self-rescuers
are in common use.
The compressed oxygen SCSR is a recirculating system. Expired air passes through a soda lime
filter which removes carbon dioxide, then into a flexible "breathing bag" where it is enriched with
oxygen from a compact gas cylinder. The oxygen may be supplied at a base rate of some 1.2 to 2
litres per minute but will be increased automatically if deflation of the breathing bag indicates a
rising demand for air. The air passes to and from the mouthpiece via a flexible breathing tube. The
unit is worn on the chest and supported by neck and waist straps.
21-
65
The chemical SCSRs are lighter devices and develop oxygen from the reaction that occurs when
water vapour and carbon dioxide from exhaled breath pass through potassium superoxide (KO2).
The oxygen in this compound contains an additional electron compared with free gaseous oxygen.
This gives a fairly weak bond and free oxygen is released readily according to the following
reactions:
4KO2 + CO2 + H2O K2CO3 + 2KOH + 3O2
and
(21.14)
KOH + CO2 KHCO3
Exhaled air from the mouthpiece passes down the flexible breathing tube and through a bed of
granulated potassium superoxide where oxygen is added and carbon dioxide removed. It is then
collected in a breathing bag for re-inhalation. A heat exchanger can be incorporated to maintain
the temperature of the inhaled air to no more than 45C in most situations.
The chemical production of oxygen can be initiated simply by exhaling vigorously into the unit.
However, there may be little or no time to accomplish this safely. Chlorate candles are
incorporated into most chemical SCSR's to overcome this difficulty. The chlorate candle is started
by percussion or other device when a ripcord is pulled. This provides some three to four litres of
oxygen during the following few minutes and until the potassium superoxide reaction becomes
effective.
Because of their vital role in cases of emergency, it is important that self-rescuers be inspected
according to a timetable that may be imposed by regulations but should occur at intervals of no
more than 3 months or after an SCSR has been carried or worn on the person of any individual. In
particular the seal of each unit should be examined to ensure its integrity. A log should be
maintained on inspections of each individual unit.
The bulk and weight of most current self-contained self rescuers may render them unsuitable for
continuous wearing by a miner. Caches of SCSRs may be stored at underground locations. Such
caches should be within a travel time of five to ten minutes, depending on the duties of any worker,
preferably in a level intake airway but accessible from more than one entry. However, caches of
SCSRs should be available at both the main gate and tail gate ends of a longwall panel and in
designated escapeways. Caches of SCSRs should be checked as part of pre-shift inspections.
These should be at least the number of SCSRs stored underground as the maximum number of
personnel underground at any one time. Training should ensure that each miner knows the
locations of SCSR caches. It is prudent to have self rescuer units on personnel transportation
vehicles.
21.9.4. Escapeways
In any subsurface facility, certain paths should be selected as preferred evacuation routes in the
event of an emergency. These escapeways should be highlighted during training sessions and fire
drills. There are three matters of importance:
selection of airways as escapeways
preparation and maintenance, and
protection and use of escapeways during an emergency.
The choice of escapeways commences by a pragmatic examination of the airflow routes, travel
distances, geographical layout of airways, the physical state of those airways, directions of leakage
through stoppings, and the locations of air crossings, doors and other ventilation controls
that may be dislocated by a fire or explosion. The primary requirement is that escapeways should
be maintained free from products of combustion for as long as possible following the outbreak of a
fire.
21-
66
At least two escapeways must be available from all areas of routine mineral extraction. This is
usually mandated by law. In the case of single development headings or other special
circumstances, legislation may limit the number of persons who may be allowed into those areas at
any one time. For single entry longwall systems, the section intake and return airways must serve
as the escapeways. However, even in multi-entry systems it is prudent to maintain at least one
return as a escapeway. Incidents in the intakes or mineral extraction zone may prevent evacuation
along intake escapeways. Depending upon air velocities and the efficacy of any barricades that
may be erected, it might be possible to outpace the airflow and stay ahead of products of
combustion along a well maintained return escapeway.
At least one (and preferably more) intake airway for each section of a mine should be maintained
as an escapeway. Where two or more intakes are designated as escapeways then they should be
truly independent ventilation routes as far back to the primary mine inlets as practicable. Two
parallel airways, where one is ventilated by air that has passed through the other, are not
independent ventilation routes and cannot be regarded as separate escapeways (MSHA, 1984).
Strata stresses should also be considered in selecting escapeways. Routes that are subject to
crushing will become excessively expensive to maintain in a good travellable condition and should
not be chosen as escapeways.
Network analysis programs with pollutant distribution modules (Section 7.4.4.) can be employed to
check and improve the selection of escapeways for any given fire scenario. Computer packages
that simulate fire situations are particularly valuable for this purpose (Section 21.3.2.2). Even more
specialized programs have been developed specifically for the identification of preferred
escapeways (Barker-Read and Li, 1989). The reliability of escapeways can be subjected to
objective tests through the techniques of fault-tree analysis through which the causes and
consequences of events may be interlinked, tracked and analyzed (Goodman and Kissell, 1989).
The preparation and maintenance of escapeways should be carried out with careful consideration
given to the potential conditions in which those escapeways may be used, i.e. in zero visibility by
persons wearing self-rescuers and who are in a state of anxiety. Obstructions, patches of poor
roof, uneven or tracked floors and unfenced junctions may create little difficulty during routine
travel but can make the difference between life and death when the airway is filled with smoke.
Escapeways must be maintained in a condition suitable for unimpeded foot travel at all times.
Signs and coloured reflectors should be employed to indicate directions of escape. However, here
again, these visual indicators may be rendered invisible by smoke. Reflectors should, preferably,
be located on the sidewalls rather than at the roof of airways as smoke is liable to be thickest at
roof level. Some form of lifeline should extend throughout the length of an escapeway. This may
simply be a water or compressed air pipe, a cable, wire line or part of the structure of a conveyor,
but placed at a convenient height to be followed by hand contact. Such lifelines should also have
cones or other devices at intervals to give a tactile indication of direction. It is prudent to deposit
extra caches of self- rescuers in well-identified locations within escapeways.
Persons who work in any section of a mine should be completely familiar with the relevant
escapeways. This can be promoted by using alternative exit routes during fire drills and displaying
escapeway maps within the section and at other locations, such as shaft stations, where miners
may congregate.
During an actual emergency, every attempt should be made to maintain escapeways free from
products of combustion for as long as possible. This might be accomplished by the placement of
brattice cloths or other devices to control pressure differentials between airways (Section
21.3.4.1.). Again, training classes and fire drills give the opportunity of imparting familiarity with
procedures involved in evacuation through escapeways. These include the advisability of travelling
in groups for mutual assistance and the use of lifelines.
21-
67
21.9.5. Refuge chambers

When personnel have become trapped within underground workings and all escape routes have
become inaccessible then refuge chambers provide a last resort to preserve life. These are
lengths of airway or fire-resistant prefabricated chambers which might be portable and either rigid
or inflatable within which miners may wait until they can be reached by rescue personnel. Blind
headings or other single ended zones are the preferred locations. However, totally enclosed
prefabricated chambers may be sited in through-flow airways or cross-cuts. The requirement,
number and locations of refuge chambers may be mandated by legislation. Where used, they
should be within reasonable access from active working sections (NIOSH, 2007).
Refuge chambers should be sized according to the maximum number of persons who work in that
area of the mine and for an occupancy period of not less than two days. Assuming that each
resting person has need of 0.15 litres of air per second (Table 11.2) gives an air requirement of 13
m3/day. However, this should be multiplied by a safety factor of 3 to allow for increased
concentrations of exhaled carbon dioxide. Hence, if twenty persons are to be accommodated, then
the chamber should have a minimum volume of 13 x 2 x 3 x 20 = 1560m3 . An opening of crosssectional area of 15 m2 would, therefore, require a length of 1560/15 = 104 m if there is no
additional air feed from a compressed air line or surface borehole.
Two types of refuge chamber have bought trapped miners vital waiting time. Pre-prepared and fully
equipped refuge chambers should be considered for workings that are distant from surface
connections, and in long development or exploration headings. They should ideally be equipped
with water, compressed air, communication with the surface, gas and temperature monitors,
sanitary arrangements, extra self-rescuers, food in sealed containers, first aid supplies, oxygen
cylinders, emergency supplies of breathable air, and reading/writing materials (Halasz, 1985). The
entrances to pre-prepared refuge chambers should be well marked. Routes leading to refuge
chambers may be fitted with directional blinking lights ( similar to Christmas tree lights) to assist
miners traversing polluted airways. All underground workers and management should be trained
in the maintenance and use of refuge chambers and be aware of the existence and locations of
such chambers. Prefabricated chambers should be constructed from non-combustible and heatresistant materials.
As a last resort, trapped personnel may be able to construct improvised refuge chambers provided
that brattice cloths, timber, hammers, saws and nails are available. Leakage points may be sealed
by clothing, rags, boards or, indeed, anything else that is available. In such cases, an active
compressed air line into the zone is invaluable in maintaining the pressure above that of the
connecting and, perhaps, polluted airways. If such barricades have to be erected within a
continuous airway then attempts should be made to minimize the pressure differential applied
across the ends. This may entail holing stoppings or opening doors in the immediate vicinity.
Following the NIOSH Report of Dec. 2007 the Mine Safety and Health Safety Administration of the
United States published for public comment a Proposed Rule on Refuge Alternatives in American
coal mines (MSHA, 2008). At the time of writing, it is intended that a Final Rule will be established
by the end of 2008. This follows a period since 1969 when federal law allowed the regulatory
agency to require that refuge chambers be installed in coal mines. Until 2008, however this
authority was never exercised.
21.10 EMERGENCY PROCEDURE AND DISASTER MANAGEMENT

The manner in which a major fire or explosion in a mine is handled depends largely on the
forethought and planning that has been expended on such an eventuality. The early stages of an
emergency are often fraught with uncertainty, and chaos can easily occur. It is, therefore, vital that
those in charge of subsurface operations should have established and documented a definitive
21-
68
procedure to be followed when a dangerous condition is discovered. All key personnel should be
very familiar with that procedure. This final Section discusses the three areas that should be
addressed.
21.10.1. Immediate response
There are two sets of actions that should occur immediately and simultaneously when a major fire
or an explosion takes place. First, all available warning systems for underground personnel should
be activated; the numbers of persons underground and their work locations should be established,
return airways should be evacuated, and responsible persons should be sent to inspect the
affected area (as far as is possible). While this is in progress, telephone messages should be
transmitted to key persons or agency representatives not already at the site. These include
mine management and senior supervisory staff

heads of mine specialist departments (e.g. ventilation, safety, electrical, mechanical)
regional mine rescue centres and rescue stations at neighbouring mines
government inspectorate or enforcement agencies
union officials
local police, fire and medical facilities.
A list of these persons or agencies and their telephone numbers should be posted permanently in
a prominent position in the surface operations centre of the mine and updated whenever
necessary. At all times, when people are in the mine, there must be a designated person on
surface who has the authority and responsibility to initiate these immediate actions.
21.10.2. Command centre
An emergency command centre should be established in a surface office as rapidly as possible.
This should be manned by a few key officials, normally the senior management of the mine and
others with detailed and expert knowledge of firefighting and rescue operations. Those persons
should have the authority to take part in decision making. However, they should be supported and
advised by specialist engineers, other government officials and union representatives. Office
accommodations equipped with desks, telephones, mine maps and stationery should be made
available for these supplementary groups adjoining the command centre. Provision should also be
made for rescue personnel and those involved in the analysis of gas samples.
It is important that the chain of command is firmly established. There should be one person in
overall control. Decisions are usually arrived at after consultation and discussion. However, the
final word lies with that one person. Instructions relating to firefighting, rescue and evacuation must
be issued only from the command centre. Direct communication should be maintained between the
command centre and the mine telephone system as well as to surface fans, power centres and
pumping stations.
21.10.3. Disaster management
In addition to the operations that are conducted underground during an emergency, there are
numerous other facets that require detailed organization and management. It is to be expected that
many other people will arrive at the mine site throughout the emergency, including the news
media, family and friends of underground workers and volunteers varying from willing but unskilled
lay persons to specialist consultants. Police authorities should be asked to assume control of traffic
and all access roads to the mine property including surface buildings and mine entrances. Medical
facilities may require to be set up on the mine surface and vehicles made
21-
69
readily available for transportation to hospitals. Medical personnel should be established on site
and who must be able and willing to be escorted underground if necessary.
Accommodation may need to be found for immediate family members of trapped miners. This
should, preferably, be close by but off the mine site. The presence of clergy and specialists in
trauma reactions can be of tremendous assistance in those facilities. Responsible persons should
be put in charge of continuous catering facilities and janitorial services.
Every effort should be made to establish good liaison with media personnel and to accommodate
them with arrangements for interviews and telephones. One spokesperson for the mining company
should be responsible for all statements to the press. As part of disaster preparedness, that person
should have been trained in communicating through the media; in particular, making accurate and
relevant statements and responses to questions within a 30 or 60 second television interview. It
assists in maintaining good relationships with media personnel to keep them informed of the time
of forthcoming statements.
Both company personnel and reporters should be aware of their mutual responsibilities. The latter
have a duty to inform the public while, at the same time, they should not engage in any activities
that will interfere with rescue or firefighting operations. As televised interviews can be broadcast
immediately or within seconds, it is of great importance that information relating to individuals be
given to concerned family members before being made available to the media. While the great
majority of news people act in a responsible manner, there is the occasional renegade whose
enthusiasm exceeds his/her common sense. In one such incident, a young reporter managed to
evade the check-in procedures at the top of a mine shaft and reached an underground fresh-air
base during rescue operations. He was brought out on a stretcher, having fainted at the scene.
References
Banerjee, S.C. (1985). Spontaneous combustion of coal and mine fires. CMRS, Dhanbad, India.
Published by Balkema.
Barker-Read, G.R. and Li, H. (1989). Automatic selection of safe egress routes away from
underground fires. Mining Science and Technology. Vol. 9, pp 289-308.
Bhowmick, B.C. et al (1991). Dynamic balancing of pressure - its relevance for control of fire in
Indian coal mines. Indo- Polish Workshop on Mining Research and Mining Methods. Central
Mining Research Station, Dhanbad, India, pp 1-15.
BHP Billiton Sustainability Report (2005). Safety Case Studies. Bleederless longwall ventilation
at San Juan Underground Mine manages risk of fire and explosion.
http://sustainability.bhpbilliton.com/2005/repository/safety/caseStudies/caseStudies14.asp
Both, W. (1981). Fighting mine fires with nitrogen in the German coal industry. The Mining
Engineer, U.K., May.
Boulton, J.R. (1991). Wave trend crossing - a new tool for detecting fires in a mine employing
diesel equipment. 5th U.S. Mine Ventilation Symposium, W. Virginia, pp 9-17.
Browning, E.J. (1988). Frictional ignitions. 4th Int. Ventilation Congress, Brisbane, Australia, pp
319-326.
Brunner, D.J., Wallace, K.G. and Deen, J.B. (1991). The effects of natural ventilation pressure
on the underground ventilation system at the Waste Isolation Pilot Plant. 5th U.S. Mine Ventilation
Symp., W. Virginia, pp 593-604.
21-
70
Chakravorty, R.N. and Woolf, R.L. (1980). Evaluation of systems for early detection of
spontaneous combustion in coal mines. 2nd Int. Mine Ventilation Congress, Reno, Nevada, pp
429-436
Chakravorty, S.L. (1960). Auto-oxidation of Indian coals. Jnl. Min. Met. Fuels, Nos. 8, 9 and 11,
pp 1 and 10.
Coward, H.F. and Jones, G.W. (1952). Limits of Flammability of Gases and Vapors. U.S. Bureau
of Mines, Bulletin 503, US Government Printing Office, 153 pp.
Cybulska, R. (1981). Examples of coal dust explosions. Ignitions, Explosions and Fires (ed. A.J.
Hargraves). Australian Inst. of Mining and Metallurgy, Illiwarra Symposium, pp 7.1-7.8.
Deliac, E.P. et al (1985). Development of ventilation software on personal computers in France
and the application to the simulation of mine fires. 2nd U.S. Mine Ventilation Symposium, Reno,
Nevada, pp 19-27.
Dunn, M.F. et al (1982). Main mine reverse performance characteristics. 1st U.S. Mine Ventilation
Symp., Tuscaloosa, Alabama, pp 23-28.
Dziurzyski, W. Kruczkowski, J. (2007), Validation of the Mathematical Model Used in the Ventgraph
Programme on the Example of the Introduction of New Headings to the Ventilation Network of Mine
ARCHIVES OF MINING SCIENCES Poland VOL 52; PART 2, pages 155-170
Dziurzyski, W., Tracz, J. and Trutwin, W. (1988). Simulation of mine fires. 4th Int. Mine
Ventilation Congress, Brisbane, Australia, pp 357-363.
Eicker, H. and Kartenburg, H.J. (1984). Investigation of methods for early detection of fires due
to spontaneous combustion. 3rd Int. Ventilation Congress, Harrogate, U.K., pp 411-416.
Eisner, H.S. and Smith, P.B. (1956). Firefighting in underground roadways: experiments with
foam plugs. S.M.R.E. Res. Rep. No. 130, U.K.
Froger, C.E. (1985). Firefighting expertise in French underground mines. 2nd U.S. Mine
Ventilation Symp., Reno, Nevada, pp 3-10.
Gillies, A.D.S., Wala ,A.M., Wu,H.W (1995). Validation Study of the Mine Fire Simulation Model.
Proc. 7th US Mine Vent. Symp. pp.199-206.10th U.S. Mine Ventilation Symp, Anchorage, Alaska.
pp 445-455.
Gillies A.D.S., Wala A.M. and Wu H.W., (May 2004). Case Studies from Application of Numerical
Simulation Software to examining the effects of fires on mine ventilation systems, 10th US Mine
Ventilation Symposium, Anchorage, Alaska, pp 445-455.
Goodman, G.V.R. and Kissell, F.N. (1989). Fault tree analysis of miner escape during mine fires.
4th U.S. Mine Ventilation Symp., Berkeley, CA, pp 57-65
Graham, J.I. (1914). Absorption of oxygen by coal. Trans. Inst. of Mining Engrs., U.K., Vol. 48, p
521.
Greuer, R.E. (1988). Computer models of underground mine ventilation and fires. U.S. Bureau of
Mines, IC 9206 (Recent developments in metal and non-metal mine fire protection), pp 6-14.
21-
71
Greuer, R.E. (1984). Transient-state simulation of ventilation systems in fire conditions. 3rd Int.
Mine Ventilation Congress, Harrogate, U.K., pp 407-410.
Grieg, J.D., Lloyd, P.J.D. and Quail, R.W. (1975). Some aspects of the use of high-expansion
foams in underground firefighting. 1st Int. Mine Ventilation Congress, Johannesburg, S. Africa, pp
239-245.
Halasz, L. (1985). Establishment and use of refuge bays at the Western Areas Gold Mining
Company, Ltd., Chamber of Mines of S. Africa, Mine Safety Div., Vol. 4, No. 2.
Haldane, J.S. and Meachen, F. (1898). Oxidation and spontaneous combustion. Trans. Inst. of
Mining Engrs., U.K., Vol. 16, p 457
Hall, A.E. et al (1989). Sulphide dust explosion studies at H-W mine of Westmin Resources, Ltd.
4th U.S. Mine Ventilation Symp., Berkeley, CA, pp 532-539.
Herbert, M.J. (1988). The use of nitrogen inertization in the British Isles. 4th Int. Mine Ventilation
Hjelmstad, K.E. and Pomroy, W.G. (1990). Novel fire warning system for underground mines.
Mining Engineering, U.S.A., Jan., pp 107-112.
Hodges, D.J. and Hinsley, F.B. (1964). The influence of moisture on spontaneous heating of
coal. Trans. Inst. of Mining Engineers, U.K., Vol. 123, No. 40, pp 211-224.
Holding, W. (1982). Explosible dusts. Chapter 28, Environmental Engineering in South African
Mines. Mine Ventilation Society of South Africa, pp 763-771.
Holding, W. (1975). An approach to the potential problem of sulphide dust ignitions at Prieska
Copper Mine. 1st International Mine Ventilation Congress, Johannesburg, pp 207-211.
Hornsby C.D. et al. (1985). Memorandum of the Committee of Mining Engineers. United Kingdom.
Hwang, C.C. et al (1991). Modeling the flow-assisted flame spread along conveyor belt surfaces.
5th U.S. Mine Ventilation Symp., West Virginia, pp 39-45.
IME (1985). Sealing-off fires underground. Memorandum prepared by the Institution of Mine
Engineers, Doncaster, England, pp 1-47.
Jensen, B. et al (1989). An experimental approach to the determination of explosible lean limits
for coal dust and methane mixtures. 4th U.S. Mine Ventilation Symp., Berkeley, CA, pp 549-557.
Kissell, F.N. and Timko, R.J. (1991). Pressurization of intake escapeways with parachute
stoppings to reduce infiltration of smoke. 5th U.S. Mine Ventilation Symp., West Virginia, pp 28-34.
Kolada, R.J. and Chakravorty, R.N. (1987). Controlling the hazard of methane explosions in coal
storage facilities. 3rd U.S. Mine Ventilation Symp., Penn State, pp 334-339.
Lin, W. and McPherson, M.J. (1997).The Ignition of Methane and Coal Dust by Air Compression - The
Experimental Proof. SME Annual Meeting, Denver, March, 1997.
Litton, C.D. (1988). Complexities of mine fire detection due to the use of diesel-powered
equipment. 4th U.S. Mine Ventilation Symposium, Berkeley, California, pp 162-166.
21-
72
Litton, C.D. et al (1987). Calculating fire-throttling of mine ventilation airflow. U.S. Bureau of Mines
Report, RI 9076.
McPherson, M.J. (1995). The Adiabatic Compression of Air by Large Falls of Roof, 7th US Mine
Ventilation Symposium, Lexington KY, pp 257-262.
Mitchell, D.W. (1990). Mine fires. Maclean Hunter, Chicago, Illinois, 167 pp.
MSHA (U.S. Mine Safety and Health Administration) (1984). Report of Investigation, Underground
Coal Fire, Wilberg Mine, Utah.
MSHA (2008). News Release No. 08-126-NAT. Feb 1
http://www.msha.gov/MEDIA/PRESS/2008/NR080131.asp
MSHA, 2008. Proposed Rule on Refuge Chambers. U.S. Federal Register / Vol. 73, No. 116 /
Monday, June 16, 2008 / Proposed Rules.
Nagy, J. and Kawenski, E.M. (1960). Frictional Ignition of Gas during a roof fall. U.S. Bureau of
Mines, ROI 5548. pp 1-11.
NIOSH, (Dec, 2007). Research Report on refuge alternatives for underground coal mines.
http://www.cdc.gov/niosh/mining/pubs/pubreference/outputid2643.htm.
Mutmansky,J.M. et al (Dec. 2007). Report to the U.S. Dept. of Labor. Final Report of the
Technical Study Panel on the utilization of belt air and the composition and fire retardant properties
of belt materials in underground coal mining.
http://www.msha.gov/BeltAir/BeltAirFinalReport122007.pdf
Oberholzer J.W. and Lyne B.J. (2002). A review of the requirements for testing of the strength of
ventilation structures to be used in Queensland mines. Proc. Queensland Mining Industry Health
and Safety Conference, pp 105-112
Ouderkirk, S.J. et al (1985). Mine stench fire warning computer model development and in-mine
validation testing. 2nd U.S. Mine Ventilation Symp., Reno, Nevada, pp 29-35.
PD-NCB Consultants (later British Mining Consultants) (1978). A review of spontaneous
combustion problems and controls with application to U.S. coal mines. Report to the U.S. Dept. of
Energy, ET-77-C-01-8965
Pomroy, W.H. (1988). Fire detection systems for non-coal underground mines. U.S. Bureau of
Mines Information Circular IC 9206, pp 21-27.
Pomroy, W.H. and Laage, L.W. (1988). Real-time monitoring and simulation analysis of mine
atmospheres to locate and characterize underground mine fires. 4th Int. Mine Ventilation
Powell,F. and Billinge, K. (1975). The frictional ignition hazard associated with colliery rocks, The
Mining Engineer, Institution of Mining Engineers, Vol. 134, No 174 pp 527-533.
Richard, K. P.1997, The Westray Story, A Predictable Path to Disaster, Four volumes, published
by the Province of Nova Scotia, Canada.
Sacks,
H.K.
and
Novak,
T.(2005).
Corona-discharge-initiated
mine
Industry Applications, IEEE Transactions, Volume 41, Issue 5, pp 1316 1322.
explosions
21-
73
Sheer,T.J. (1988). Safety precautions for the protection of miners against underground fires. 4th
Int. Mine Ventilation Congress, Brisbane, Australia, pp 1-12.
Stefanov, T.P. et al (1984). Unsteady-state processes during an open fire in a ventilation network.
3rd Int. Mine Ventilation Congress, Harrogate, U.K., pp 417-420.
Strang, J. and MacKenzie-Wood, P. (1985). Mines Rescue, safety and gas detection. Weston
Publishers, Kiama, Australia, 366 pp.
Strang J. and Mackenzie-Wood (1990). A Manual on Mines rescue, Safety and Gas Detection,
Colorado School of Mines Press, 473 pp.
Sun, W.T. (1963). First industrial trial of fighting underground fires by combustion gases at
Tchegan Colliery. Int. Mining Congress, Saltzburg, Germany.
Timko, R.J., Derick, R.L. and Thimons, E.D. (1988). Analysis of a fire in a Colorado coal mine.
4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 345-353.
Trutwin, W. (1975). Reversal of airflow by means of water sprays in upcast shafts. Polish
Academy of Sciences, Mining Archives Quarterly, Vol. 20, No. 4 (Published in Polish).
Trutwin, W., Dziurzyski W. and Tracz J., ((1992). Computer simulation of transients in mine
ventilation. 5th Int. Mine Ventilation Congress, Johannesburg, South Africa, pp 193-200.
Verakis, H.C. (1991). Reducing the fire hazard of mine conveyor belts. 5th U.S. Mine Ventilation
Symp., West Virginia, pp 69-73.
Verakis, H.C. and Dalzell, R.W. (1988). Impact of entry air velocity on the fire hazard of conveyor
belts. 4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 375-381.
Wala, A.M. (1998). Teaching the principles of mine fire using computer-aided instruction. Computer
Applications in Engineering Education, Vol. 5, Issue 4 pp 249-255.
Wu H.W. and Gillies A.D.S. (2006). Queensland mine emergency level exercises assisted by fire
simulation. 11th U.S./North American Mine Ventilation Symposium. pp.351-358.
Willett, H.L. (1951). Gas analysis behind stoppings. Trans. Inst. of Mining Engrs., Vol. 111, p 629.
Zabetakis, M.G. et al (1959). Determining the explosibility of mine atmospheres. U.S. Bureau of
Mines, IC 7901, pp 11.
Zabrosky, C.E. and Klinefelter, G. (1988). Microcomputers assist in extinguishing a bituminous
coal mine fire. 4th Int. Mine Ventilation Congress, Brisbane, Australia, pp 337-344.
Zipf, R.K., Sapko, M.J. and Brune, J.F. (2007). NIOSH Informational Circular 9500, Explosion
pressure design criteria for new seals in U.S. coal mines. 76 pp.
http://www.cdc.gov/niosh/mining/pubs/pubreference/outputid2477.
21-
74

Subsurface Ventilation Engineering - McPHERSON

Uploaded by

Copyright:

Available Formats

Subsurface Ventilation Engineering - McPHERSON

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Subsurface Ventilation Engineering - McPHERSON

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What topics are covered in Part 1?

What topics are covered in Part 1?

Subsurface Ventilation Engineering.

Dedication: This work has been undertaken in fulfillment of

*Formerly of the University of Nottingham, England

Foreword to the Web version of the book.

Chapter 1. BACKGROUND TO SUBSURFACE VENTILATION AND ENVIRONMENTAL ENGINEERING

Chapter 7. VENTILATION NETWORK ANALYSIS

PART 4. Heat and humidity

20.4. Control of dust in mines

PREFACE TO THE ORIGINAL TEXT

TABLE OF CONVERSION FACTORS BETWEEN IMPERIAL AND SI UNITS

Note: The millibar (1 mb = 100 N/m2) is included here as it is a familiar metric

1 ft3 /imp. ton

1 ft3 /short ton

Note: 1 metric tonne (t) = 1 000 kg

For differential temperatures, 1 Centigrade degree = 1.8 Fahrenheit degrees.

Background to subsurface ventilation

1.2 A BRIEF HISTORY OF MINE VENTILATION

1.3. THE RELATIONSHIPS BETWEEN VENTILATION AND OTHER SUBSURFACE SYSTEMS6

A BRIEF HISTORY OF MINE VENTILATION

Figure 1.1 A print from Agricola's " De Re Metallica"

Buddle's reaction is best expressed in a letter he wrote to

1.3. THE RELATIONSHIPS BETWEEN VENTILATION AND OTHER

Factors that contribute to

1.3.3. The integration of ventilation planning into overall system design

Introduction to Fluid Mechanics

Chapter 2. Introduction to Fluid Mechanics

2.2 FLUID PRESSURE.......................................................................................... 3

2.3 FLUIDS IN MOTION ........................................................................................ 8

Introduction to Fluid Mechanics

Introduction to Fluid Mechanics

2.1.2 Volume flow, Mass flow and the Continuity Equation

2.2 FLUID PRESSURE

Introduction to Fluid Mechanics

But as pressure = force/area, the pressure on the base of the tube is

Introduction to Fluid Mechanics

2.2.3 Atmospheric pressure and gauge pressure

The measurement of variations in atmospheric pressure is important during ventilation surveys

Introduction to Fluid Mechanics

The atmospheric pressure could then be calculated as (see equation (2.8) )

where, in this case, is the density of mercury.

Introduction to Fluid Mechanics

Introduction to Fluid Mechanics

2.3 FLUIDS IN MOTION

Furthermore, the acceleration is defined as

The force is given by

and the work done to accelerate from rest to velocity u is

The kinetic energy of the mass m is, therefore, m u2/2 Joules.

Introduction to Fluid Mechanics

where g is the gravitational acceleration.

This gives the potential energy of the mass at elevation Z.

Figure 2.1 Flow work done on a fluid entering a pipe

Work done = force x distance

However, the product As is the swept volume v, giving

Introduction to Fluid Mechanics