Process Engineering Training

Program
MODULE 2
Introduction to Kiln Processes
Section CONTENTS
1 Glossary of Terms in Chemical Processes
2 Wet Process Kilns (New CTC 14)
3 Semi Wet Process (New CTC 15)
4 Lepol Process (New CTC 16)
5 Long Dry Process Kilns (New CTC 17)
6 Suspension Preheaters (New CTC 18)
7 Precalciner Kilns (New CTC 19)
8 Precalciner Kilns- (Lecture 24)
9 Kiln Systems- FLS
10 A New Perspective on the Cement Making Process Through LINKman
11 Investigation of Dust Generation Mechanisms in Rotary Cement Kilns, 1984
12 Choosing a Kiln Shell Cooling System
13 CETIC Workshop- Kiln Shell Cooling
14 Criteria Governing the Use of Short Rotary Kilns for the Production of
Cement
15 Instruction for Kiln Operation
16 Take These Chains

HBM Process Engineering Conference

Hearth Design for Preheater and Precalciner Kilns
Preheater Cleaning Methods and Successes

Presentations
Suspension Preheaters
Precalciner Kilns
Kiln Systems
Bowmanville Cement Technology-Kiln Systems
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 1

Glossary of Terms in Chemical Processes

 GLOSSARY OF TERMS

Introduction

The following section is a selection of terms listed under alphabetical headings.. Most
of the descriptions/definitions are of commonly used terms in the cement making
process or chemistry areas. it is not intended that this should be a comprehensive
‘dictionary’ of cement manufacture but more as a reference for those unfamiliar with
some of the more ‘industry specific’ jargon.
 GLOSSARY OF ‘lERB,Q

Ash Residue from burning coal, contains mostly silica,

alumina and iron oxides. Most ash combines with
clinker and leaves the rotary kiln as part of the
product.

Ash Heterogeneity Poor mixing of ash with the kiln feed having adverse
effect on cement quality.

Air Swept Type of grinding mill where all material is carried

out of the mill by a fast moving gas stream.

Alignment (Most commonly ‘kiln alignment’). Process of

ensuring that each tyre station on a particular kiln
is in line, to minimise any distortion of the shell
during rotation. See also kiln.

Anhydrite Manufactured or dehydrated from of gypsum. Less

reactive than gypsum and is used as a partial
substitute for gypsum to improve and optimise
cement workability.

Alkalis General term for sodium and potassium.

ASR Reaction between certain forms of silica found in

Alkali Silica Reaction aggregates and alkalis present in concrete. A gel is
formed which takes up water, expands and cracks the
concrete.

Alkali Cycle Circulating load of alkalis and alkali compounds

which are evaporated in the kiln burning zone and
recondense in other parts of the system. This is a
cause of build-up in the back end of dry process kiln
systems.

Alumina Ratio (AR) Chemical control ratio obtained from the formula:

A o r Alumina j6A1203
F Iron Oxide %Fe203
Values of about 1.4 - 1.5 are optimum for ease of
clinker combination.

Air Setting Slow process of water absorption into bagged

cement, causes lumpy cement. Can be more rapid if
potassium sulphate in cement is 1% or more.
A - continued

Alite The general term for C$, the main strength forming
compound present in Portland cement.

Air Separate (AS) Type of precalciner kiln where air is drawn from
the kiln hood in a separate duct to provide
combustion in the precalciner vessel - see also
‘tertiary air’.

Air Through (AT) Type of precalciner kiln where the air for
combustion of fuel in the precalciner vessel passes
up the kiln tube. Air through kilns have a lower
output than similar sized AS kilns.

Auxiliary Firing Also called:

-- riser pipe firing

- secondary firing
- partial precalcination
- back-end firing

Process adopted on conventional preheater kilns

where part of the fuel is injected into the kiln back-
end riser. Can improve both kiln output and
stability. Partial precalcination refers more to
situations where tbe riser is extended or enlarged to
be more like a ‘precalciner vessel’.
B

Chemist who developed a method for calculating the

proportion of the compounds C$, Cfi, C3A and
C,#W present in Portland cement.

Burning Process of clinker formation in the cement kiln.

Also - burning zone, area of the kiln where these

final reactions take place.

Belite General term for C& the secondary strength giving

component in cement.

Bauxite Impure form of aluminium oxide (A1203) sometimes

used to correct raw materials deficient in Alumina.

Blending Process of mixing variable materials to produce a

less variable feed to the next stage in the process.

Blow Tank Device using one or more vessels to transport

powdered material. The vessels are alternately
filled, pressurised and then exhausted into a
conveying line.
 C
Clinker Partly fused (melted) nodular product of cement
rotary kiln.

Cooler See Rotary Cooler

Planetary Cooler
Grate Cooler

Cyclone Device used to cause gases carrying material to form

a vortex. Material separates from the gas stream by
centrifugal forces.

Also hydrocyclone uses a similar process to split

slurry into two streams, one containing mostly fine
particles, the other containing coarse particles for
re-processing.

Cyclone Preheater Series of cyclones used to separate material from

hot gases after preheating stages.

Closed circuit Energy efficient grinding process where the mill

product is not ground fine enough in a single pass
through the mill to be final product. It is separated
into two streams - a product stream and a reject
stream which is returned to the mill.

Calcium Carbonate Primary raw material for cement manufacture,

occurs as chalk, limestone, marble, coral or mixed
with clays (marl)

Clay Secondary raw material for cement manufacture,

containing mainly silica alumina and iron oxides.

Centrifugal Pump Used for transporting slurry or water, develops

pressure by drawing slurry in at the centre of a
rotating impeller and ‘throwing’ it outwards towards
the pump outlet (low to medium pressure duties).

Calorific Value (CV) Measurement of the heat available from combustion

of a fuel.
(Higher Heating Value) CV of coal 7000 kcal/kg = 13,000 Btu/lb
(Lower Heating Value) CV of oil 10000 kcal/kg = 18,000 Btu/lb
See also - kilocalorie.
Diaphragm Slotted dividing wall between chambers or at outlet
(Division Head) end of a ball mill. Can be double plate with scoops
between to control flow of material, or single plate.

Decarbonation Process by which limestone or chalk (calcium

(Calcination) carbonate) is converted into reactive lime before
clinker reactions can occur:

CaC03 -> CaO + CO2 (gas)

(lime)

Dicalcium Silicate C2S also known as Belite.

Ettringite Product of reaction between C3A, gypsum and

water. It is insoluble and forms on the surface of
CQA crystals preventing rapid reaction of C3A and
allowing concrete to be placed before setting occurs.
Formation of ettringite once concrete has hardened
(sulphate attack) causes expansion and cracking.

Endothermic Term usually applied to chemical reactions, it

implies that heat is required for the reaction to
proceed, i.e. heat is absorbed during the reaction,
e.g. decarbonation reaction.

Exothermic Opposite of Endothermic, i.e. once the reaction

starts, it is self-sustaining, heat is released. E.g.
combustion reactions, clinker formation reactions.
 Free Lime Residual lime from the limestone or chalk raw
material which has not combined with the other
oxides to form Portland cement compounds. All
cement contains some free lime. Excess free lime
can cause unsoundness in hardening concrete.

False Set A stiffening of the concrete mix due to the

precipitation of gypsum crystals, sometimes referred
to as plaster set. Can be broken down by further
mixing.

Flash Set Rapid irreversible hardening due to C3A hydration,

usually signifies insufficient gypsum in the cement.

Ferrite Phase C4AF

FK Pump Fuller Kinyon Pump, also known as a screw pump,
used for transporting powders, operates by screwing
the bulk powder into a fast moving air stream which
then transports the material to silos, feed bins, etc.

Gypsum Calcium sulphate dihydrate. Occurs naturally and

when interground with cement clinker will control its
setting rate. The natural mineral can contain
anhydrite, clay, quartz and calcite as impurities.

Gas Conditioning Tower Chamber with water spray equipment designed to

cool and increase humidity in exhaust gases to
improve electrostatic precipitator performance.

Grate Cooler Horizontal grate consisting of alternate rows of

fixed and moving plates. The plates have holes
through which air is blown to cool the clinker on top
of the grate. The moving plates move backwards
and forwards to move clinker along the grate.

Grindability ‘Or grindability test’. This is a measurement of the

behaviour of a material in a grinding process. The
test result can be used to give an absolute or
comparative measurement of power requirements for
grinding.
H

Hydration General term covering chemical reactions between

a material and water. The setting of Portland
cement in the presence of water is a hydration
reaction.

High Efficiency Separators See ‘separators’. Modified design of separator using

a variable speed rotor with vertical blades and a
controlled gas flow to obtain a cleaner cut between
coarse and fine material.

Hydraulic In this definition ‘hydraulic properties’. Materials

which exhibit setting behaviour when mixed with
water are said to have hydraulic properties.
Examples are:

(a) Portland cement

(b) Some blast furnace slags
(c) Some power station coal ashes
(d) Volcanic ashes
 Liners Steel plates (now usually high chrome) used to
protect grinding mill shells and encourage correct
grinding media action.

Also - Nihard liners, abrasion resistant material used

in, e.g. centrifugal pump internal parts.

Also - Manganese steel liners, moderate abrasion

resistant material used where impact resistance is
also important.

Linkman Commercial name for an automated computer

control system using ‘rule based’ logic for plant
control. i.e., if temperature ‘high’ then reduce fuel.

Lime Saturation Factor Chemical control factor derived from the formula
(for clinker)

LSF(%) = CaO
2.8 ( S i + 1.6 (A1203) + 0.7 (Fe&

It is an empirical relationship, a value of over 100%

indicating excess lime in the mix.

Normal operating values are in the range 93-98%.

An increase in LSF of 1% increases cement C3S
content by ~ 2%

Loss of Ignition (LOI) Test to determine loss in weight of a sample when

heated. Used to detect excess hydration of stock
cement and levels of ‘decarbonation’ in
preheater/precalciner systems.
Media Or ‘grinding media’, steel spherical (grinding balls) or
cylindrical (cylpebs) pieces used in ball mills.
(Grinding Balls)
Sizes: balls 3.5 inch - 0.6 inch
cylpebs 1 inch - 0.4 inch

Materials, various but most common 12-25% chrome

steel/cast iron alloy, forged and/or heat treated for
maximum abrasion resistance and strength.

NO, Compounds of nitrogen and oxygen, formed in the

high temperature regions of the rotary kiln. Their
quantity in the exhaust gases can be a very useful
indication of temperature conditions in the burning
zone.

Open Circuit Grinding process where the material goes through

the mill in a single pass to emerge as final product.
See also Closed Circuit.
Precipitator Or electrostatic precipitator, gas cleaning device
which operates by electrostatically charging dust
particles to attract them out of a gas stream before
discharge to atmosphere.

Planetary Cooler Tubes similar to a ‘Rotary Cooler’ but attached

circumferentially to the kiln shell (8-10) on one kiln).

Portland Cement Material formed by grinding Portland cement clinker

with approximately 5-8% gypsum. When mixed with
water, hydrates are formed which result in a dense
durable material. Name derives from resemblance
of hardened product to Portland stone.

Portland Cement Clinker The material formed by heating to partial fusion (at
about 2640 F) a carefully proportioned and intimate
mixture of limestone and shale. Typically, it
contains four crystalline compounds in the following
proportions :-

C3S (alite) SO-65

C2S (belite) 10-20
$A (aluminate) 8-12
C&F (ferrite) l-10

Portland Cement Types Type 1 (OPC)

Ordinary Portland Cement

Type 2
Moderate sulphate resisting or moderate heat of
hydration Portland Cement

Type 3 (RHPC)
Rapid Hardening Portland Cement
A finely ground cement for rapid early strength
growth, e.g. for use in precast sections.

Type 4 (LHPC)
Low Heat Portland Cement

Type 5 (SRPC)
Sulphate Resisting Portland Cement

Walcrete
Masonry cement (mortars etc.)
O

Quality Assurance All those planned and systematic actions necessary

to provide adequate confidence that a product or
service will satisfy given requirements for quality.

Quality Control Process whereby the quality of raw materials,

processes and products is monitored and tested to
ensure it complies with Company requirements,
ASTM and customer requirements.

Residue A measure of the size of particles or fineness of a

powder (or slurry) by passing a sample through a sieve
(Expressed as Fineness) of known aperture and calculating the weight
fractions retained on the sieve (the residue).

Usually expressed as wt% + x microns (wt% plus)

Typical sieve sizes in common use are 325 and 200
mesh.

Roller Mill See Vertical Spindle Mill.

Rotary Cooler Rotating steel tube with a heat resisting refractory

brick liner for part of its length. The remaining
length is lined with steel lifters. Clinker passes
from the kiln into the cooler and is cooled by the
airflow induced by the main kiln fan.
 S
Silica Ratio Chemical control ratio obtained from the formula:

-S - or Silica
A+F Alumina + Iron Oxide

Increasing silica ratio generally improves concrete

strengths but the mix becomes progressively more
difficult to burn. For OPC an optimum of about 2.5
is preferred.

Shale Sedimentary rock containing silica alumina and iron

oxides secondary raw material for cement
manufacture.

Sedimentation General term for the layering process at the bottom

(Sedimentary Rocks) of seas or lakes which controlled the formation of
most cement raw materials.

Sieve Bend (DSM Screen) Slotted screen formed into an almost circular arc,
slurry is forced under pressure to flow inside the arc.
Fine slurry passes. through the screen, coarse is
retained.

Secondary Air This air originates from the clinker cooler where it
has been used to recover heat from clinker. It is the
second source of combustion air, hence, ‘secondary’.
See ‘Primary Air’.

Stacker/Reclaimer System used for feeding material on to a stockpile

and recovering it for process use. A stacking
process which layers the material and has a mixing
or 'homogenising' effect. Sometimes referred to as
a blending stockpile.

Separator Also known as a ‘classifier’ device which uses gravity

or gravity and centrifugal forces opposed by an
upward moving gas stream to separate a powdered
material into two streams, one containing mainly
coarse particles, the other containing mainly fine
particles.
T

Tertiary Air This term has been adopted to refer to the air
removed from the kiln hood via a ‘tertiary air duct’
to provide combustion air for fuel in the 'precalciner
vessel’.

Tri-calcium Silicate C3S - See Alite.

Tetra Calcium Alumino Ferrite C&F - See‘ Ferrite

Tri-calcium Aluminate C3A - Highly reactive component of Portland

cement Its reaction with water is controlled by
adding ‘Gypsum’ to the mix. See also 'Ettringite'
and ‘Flash Set’;
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 2

Wet Process Kilns

(CTC 14)
 WET PROCESS ROTARY KILNS

CONTENTS

1. INTRODUCTION

2. ZONES OF KILN

2.1 Drying Zone of the Wet Process Rotary Kiln

2.2 Heat Exchangers

3. EXTERNAL HEAT EXCHANGERS

3.1 Davis Preheater

4. INTERNAL HEAT EXCHANGERS

4.1 Chain Systems

4.2 Heat Transfer Mechanisms Within the Chain System
4.3 Chain System Design

5. SPIRAL CHAIN SYSTEMS - OPERATION

5.1 Slurry Moisture Effect on Kiln Performances of

Chain Design
5.2 Dust Loss
-5.3 Dust Insuf f lation

6. MAINTENANCE OF SPIRAL CHAIN SYSTEMS

6.1 Chain Weight Monitoring

6.2 Wear Rates and Materials
6.3 Shell Lining in the Chain Area

APPENDIX I CHAIN DESIGN - WORKED EXAMPLE

APPENDIX II BARNSTONE CHAIN TRIAL
1. INTRODUCTION

This paper deals with those kiln systems having slurry as a raw material.

The systems that will be dealt with in some detail are shown in Figure 1.

Obviously, the water must be removed and the material preheated before the calcining
and burning processes can commence. The removal of the water can be achieved either
by evaporation in the kiln or by an external device utilising thermal or filtration
techniques.

In this paper the various types of heat exchangers are discussed and in particular the
spiral chain system.

2. ZONES OF THE KILN

The simple wet process kiln is considered to have a number of recognisable zones whilst
the dry has one fewer viz, the drying zone.

These zones are best shown diagrammatically in relation to the gas and material
temperature. The general form of these temperatures within a typical simple wet
process kiln is shown in Figure 2.

The points of inflection of the material temperature curve indicate the various zones
of the burning process:

a) Drying Zone - This will be dealt with in detail in the following section.

b) Preheating Zone - Completion of the feed drying takes place together

with the elevation of the feed temperature to the start of decarbonation.

c) Calcining Zone - Decarbonation of the feed takes place together with the
dehydration of clay minerals and the formation of new phases.

d) Sintering Zone - Exothermic reactions of clinker formation.

e) Cooling Zone.

Material temperatures of significance are:

a) 110 C drying commencement of decarbonation.

b) 550 - 600 C commencement of decarbonation.

c) 850 - 900 C completion of decarbonation.

 A.kiln with chain system

1
B. kiln with desiccator

C.kiln with Davis Preheater

FIGURE 1: ROTARY KILNS - WET PROCESSES

2
 . . . .’.. . . -...
-. . .
! x.
... .
*5
. ..
k... Gas
....

\
-. -. . .-.
. . . .* . .
I.* -. .
I. -. . . . . . . .
. ... ...

,’
.. . .

+---COOLING ZONE---+SINTERING ZONE +------CALUNING ZONES PREHEATING ZONE+-DRYING ZONE4

FIGURE 2: KILN - TEMPERATURE PROFILE

For wet process kilns the proportion of the kiln lengths required for the various zones
are as follows:-

a) Drying Zone 30 - 35% kiln length

b) Preheating Zone 10 - 15% kiln length

c) Calcining Zone 30 - 35% kiln length

d) Sintering Zone 15 - 20% kiln length

e) Cooling Zone 2 - 3% kiln length

2.1 Drying Zone of the Wet Process Rotarv Kiln

Of the various zones mentioned previously, the drying zone is the one peculiar to the
wet process rotary kiln. Therefore the following sections will concentrate on the
features of rotary kilns relevant to the drying process.

2.1.1 Theory of Drying

The drying process is one of simultaneous heat and mass transfer. Heat from the exit
gases is transferred to the material and mass in the form of water vapour is evaporated
to the gas stream from the material. The process may be summarised into three
sections, as shown in Figure 3.

a) Slurry Preheating

b) Evaporation of free water from the surface of the material

c) Evaporation of moisture held by adsorption and capillary action from the

material

The drying does not begin until the material has reached a temperature above the dew-
point of the gas (70-75 C) So long as the surface of the material contains ‘free’ water,
conditions of saturated vapour pressure will exist in the gas boundary layer above the
material surface. An equilibrium will be established whereby the amount of heat and
vapour transferred per unit time will remain constant together with the surface
temperature of the material (70-100 C)

4
When the surface becomes dry (4% moisture) the water held by adsorption and capillary
action is evaporated to an increasingly greater extend. The vapour pressure over such
water is lower than it is over a free water surface and hence the rate of drying will
decrease. According to the laws governing simultaneous heat and mass transfer a
reduction in the drying rate will be accompanied by a similar reduction in the rate of
heat transfer.

The latter will be achieved by a rise in the temperature of the material with the result
that the temperature difference between the gas and material is reduced.

2.1.2 Factors Affecting tbe Rate of Drying

The rate of drying is dependent on various factors such as the gas temperature,
saturation pressure of the gas, mass transfer coefficient, surface area and heat transfer
coefficient. The factors that can be controlled are summarised below:-

a) Gas Temperature

b) Partial pressure of the water vapour in the gas

c) Gas velocity

d) Nature and design of kiln inserts or preheaters

The first three are all related to the gas temperature and an increase in gas
temperature will increase the rate of drying. However, increasing gas temperatures will
inevitably increase kiln exit temperatures and hence heat losses. It is therefore more
economically sound to increase the effectiveness of the kiln inserts.

2.2 Heat Exchangers

The various methods employed for ensuring adequate heat exchange between the gases
and material within a kiln system may be divided into two distinct types, namely
external and internal heat exchangers.

5
 -.
\.
\
‘\
(a)

MOISTURE CONTENT
% HzO--

FIGURE 3: THE DRYING PROCESS

3. EXTERNAL HEAT EXCHANGER

The type of external heat exchanger now in operation within the Group is the Davis
Preheater.

3.1 Davis Preheater

The Davis Preheater, two examples of which are installed at Ewekoro Works in Nigeria,
is shown in Figure 4. Hot kiln gases pass up the vertical shaft into the annular heat
exchanger. Extruded filter cake is fed into the preheater in the form of nodules via a
rotary air lock valve, thus forming a bed through which kiln gases are diverted by the
dome. The cake is formed by filtering the slurry to approximately 18-20% moisture and
passing the resulting cake between two opposingly rotating drums possessing holes in
their circumferences.

The preheater consists of a bowl and a hearth which rotate around different axes at
similar speeds, the offset of the two axes causes nodules to be swept off the hearth and
into the shaft to-the kiln. Feed control is achieved by either varying the rotational
speed or the offset of the two axes. Effective seals must be maintained between
stationary parts and the hearth and bowl.

The preheater did not enjoy widespread employment due to a number of problems which
are inherent in the system. Alkalies in the raw materials can cause build-up on the
underside of the dome and in the kiln chute, leading to a high restriction. Further
restrictions arise from the bed depth (280-350 mm wg) and deposition of hot dust within
the bed. The latter also can affect the flow properties of the nodule bed, leading to
irregular feed into the kiln. The successful operation of the Davis Preheater is very
dependent on the nature of the raw materials.

The plastic raw materials at Ewekoro are favourable whereas the less plastic raw
materials at Dunstable were unsuitable, since the nodules broke down, causing blinding
of the nodule bed.

7
 Exhaust /

Extruded filter cake

I
!I.
Y
I

FIGURE 4: DAVIS PREHEATER

8
4. INTERNAL HEAT EXCHANGERS

Chain systems are the only internal heat exchangers within the Group.

4.1 Chain Systems

Chain systems can be categorised into two general types: Garlands or Festoons and
Curtains. Garlands and Festoons consist of long chains which are looped and hung at
both ends in the system, often involving complicated weaving of the chains. The
complications of these types of chain arrangement makes installation and maintenance
difficult with the result that the systems tend to become neglected and consequently
lose their efficiency.

Curtain chains are shorter and are hung only from one end thus making the installation
and maintenance easier. Problems with material flow through chained systems led to
the development of curtain chain systems arranged in a spiral, thereby promoting
material flow.

The conversion of UK kilns to the spiral chain system commenced in the 60's Most
Blue Circle wet process kilns now have spiral chain systems and through our consultancy
agreements many other kiln chain systems throughout the world have been, or are being,
converted to spirals.

A considerable amount of the development work on spiral chain systems was carried out
at Northfleet Works in the lat 1960’s and early 1970's Northfleet with its six identical
kilns presented the ideal opportunity for optimising the spiral chain system. Many
different chain lengths and weights were tried and from the records kept on Northfleet
together with the feedback on other installation, a standard spiral chain system design
was derived.

4.2 Heat Transfer Mechanisms Within the Chain System

Figure 5 illustrates the various mechanisms whereby heat is transferred from the gas
stream to the slurry feed by means of the chain system.

As it will be seen, the process of heat transfer is complicated.

9
 Heat transfer from:-
1. Hot gas to chains
2. Hot gas to feed surface
3. Hot gas to kiln shell lining
4. Hot gas to material adhering to the chains
5. Hot chains to feed
6. Hot shell lining to feed
7 . Internal bulk mixing of feed
8. Kiln shell to atmosphere

FIGURE 5: HEAT TRANSFER MECHANISMS WITHIN THE CHAIN SYSTEM

10
The hot gases transfer heat to the chains, the kiln lining and the upper feed surface by
radiation and convection. As the hot chains pass through the feed bed they cool and
transfer heat by a complicated process of conduction, radiation and solid convection.
A similar process heats up the feed particles adjacent to the hot lining as it passes
beneath the bed. The hot particles from the lower surface and around the chains and
also from the upper surface exposed to the gas are mixed with the bulk of the feed bed
as it tumbles when the kiln rotates. This mixing of the feed transfers heat from the
surface to the bulk and exposes fresh particles at the upper and lower surfaces. A small
amount of heat is lost by conduction through the shell and radiation to atmosphere. The
internal mechanisms are further complicated by re-radiation between the hot lining and
the chains and the feed surface. In the initial section of the chains heat transfer via
the chains is more complex since wt slurry can coat the chains, be dried and carried by
the gas to a point where it re-enters the feed. Hot dust is also deposited into the cool
slurry and this heats up the feed as well as reducing the moisture content without any
real loss of water.

4.3 Chain System Design

A spiral chain system consists of chains hung in curtain formation from multi-start
spiral attachments. The attachments advance in the opposite direction to the kiln
rotation when viewed from the closure plate i.e. they tend to screw the feed down the
kiln.

Over the years of development of the Blue Circle spiral chain system a common basis
for design has been established. This allows for simplification and direct comparison’
of kiln performances. The following constraints apply to all wet,process kilns:-

1. The individual chain lengths are two-thirds of the kiln diameter.

2. The spiral pitch is one kiln diameter
3. The number of spiral starts is three
4. The chain distribution is even
5. The weight of chain per unit length has been check weighed
6. The attachment holes are 2.5 inches apart
7. The length of kiln chained is determined by the kiln shape
8. The length of start gas is normally one kiln diameter

The variable factors in the design of a chain system are the total weight installed and
the initial starting point.

4.3.1 Individual Chain Lengths

The present standard of 2/3D has been arrived at by practical experience. It may or
may not be the ideal length from the point of view of heat transfer, but it does satisfy
the equally important aspect of free material flow.

11
At the cold end of the system, the chain length is particularly critical as regards
material flow. A few inches length extra in this zone has, in a number of instances,
impeded the flow of slurry and caused slurry spillage over the kiln back end. With a
chain length greater than 2/3D the end of the chain can catch on the attachments, on
the next helix down the kiln causing tangles which prevent steady material flow through
the chains.

4.3.2 Spiral Pitch

The recommended pitch is one kiln diameter. It might be argued that this is rather an
arbitrary figure and could be altered to our advantage. However, since its main
influence is in respect of material transport, with which ‘there is no problem using the
standard pitch, there seems little virtue in further experimentation.

In addition, the ID pitch has been adopted as standard in all kilns and would prove a
costly exercise to change, with little advantage. A shorter pitch would not be
recommended as it would create a mound of chain on the kiln floor and hinder material
flow.

4.3.3 Spiral Starts

The standard number of spiral starts is three. One possible advantage if increasing the
number of starts is that it would give a more truly even density i.e. longitudinally and
circumferentially and it was to establish the effects of this, that trials with six starts
were carried out on Shoreham, Westbury and Oxford.

The following is a summary of the results of the trial on the Shoreham kilns (Table 1).

Initially the six start section was chained to the equivalent density of the three system
in order to keep an even density and produce a comparable situation. This was achieved
by doubling the individual chain spacing which gas up to 9" gaps between chains. Before
and after kiln performances are summarised below"-

12
 TABLE 1:
RESULTS OF TRIALS AT SHOREHAM USING 6 START SPIRALS

3 Start System 3 & 6 Start Modified 3 & 6

System Start System

Output tph 20.4 20.2 20.3

Back-end temp. C 195 232 196

Slurry moisture, % 38.9 38.4 38.9

Exit draught mm wg 142 116 127

Dust loss % clinker 13.7 18.6 14.2

Tonnes chain 39.0 41.0 50.0

For a similar clinker and slurry moisture output the exit gas temperature increased by
approximately 40 C whilst the exit draught was reduced by 26 mm wg and the dust loss
increased by 5% as would be expected with the higher exit gas temperature. In order
to reduce the level of exit gas temperature, the total chain weight was increased to 50
tonnes by adding 9 tonnes in the six start section, which doubled the density and halved
the chain spacing. These results were also found on the Westbury, Wouldham and
Oxford kilns.

The conclusion is that the circumferential chain spacing plays a more significant part
in the heat transfer from gas to chain than the longitudinal spacing. This is attributed
to the fact that the wider the chain spacing the less will be the number of times a unit
of gas will come into head-on contact with chain as it passes through the system.

4.3.4 Chain Distribution

An even distribution of chain weight throughout the system is the recommended

practice.

Formerly there was a strong body of opinion which favoured the use of a high density
section at the wet end to provide a large surface area to catch dust. The disadvantages
of such a high density section are that it can impede the free flow of slurry owing to:-

(i) its purely physical effect as a dam

(ii) the potentially high rate of heat transfer causing rapid changes in the
physical nature of the slurry i.e ‘balling-up’

13
The philosophy behind the present chain system design is to reduce the quantity of dust
produced through slow drying of the slurry rather than to devise a means of collecting
it once it has been produced.

The principle is now applied to all the Blue Circle spiral chain designs.

4.3.5 Individual Chain Weight

It is vitally important that the chain to be used is weighed to ensure that the designed
weight is in fact installed. In the case of second hand chain, which can vary
considerably in weight per unit lengths, check weighing is best done in parcels of cut
lengths.

The standard chain recommended is 23 mm oval link with a weight of 10.4 kg/m. This
chain has, in our experience the optimum weight/surface area ratio. A lighter chain
will not give the same useful life in the kiln and for a given total weight will occupy
more holes, so reducing the flexibility of the system.

Heavier chain has been used on occasions in areas prone to mud ring formation. Fewer
chains are required to give a similar density.

4.3.6 Attachment Mole Spacing

The minimum spacing between chain attachments holes is determined by the space
required to accommodate the recommended method of fixing shown in Figure 6. This
is 2.5 inches.

It is essential that the attachment plates be drilled at 2.5 inch centres regardless of the
fact that in the initial design all the holes are not occupied by a chain. The vacant
holes leave some flexibility for subsequent adjustments to the chain weight. With
current downward trend in slurry moistures and so the need t o increase weight of chain,
it is even more important that there are available holes in the system to allow for the
addition of the extra chains.

4.3.7 Length of Kiln Chained

There was, in the past, a tendency to squash the chain system with the idea o f leaving
as much of the kiln as possible available for decarbonation and burning. Works were
also attracted to short chain systems since it meant less downtime for the initial
installation of the spiral attachments.

Experience, initially at Northfleet and confirmed at other Works, showed that an

increase in clinker output could be achieved by increasing the chain system length and
decreasing the density. The effect this had on the Northfleet kilns is shown in Figure 7.

14
I X II
L Chains bolted alternately

V i e w ‘Xfq’

FIGURE 6: METHOD OF FIXING SPIRAL CHAINS

Following on from this work, the question was raised as to how far down a kiln the chain
system could be extended to give an improvement in performance. It was also
important to establish how the improvements made could be related to other kilns.

One obvious way of translating the Northfleet chained length is to express it as a

percentage of the total kiln length and apply this percentage to other kilns. But it will
be appreciated that all kilns vary considerably in shape and proportions. For example,
some have no enlargement at the wet end, whilst other have enlarged wet zones of
widely different lengths. The same situation, of course, applies to burning zones.
Under these circumstances it would be unrealistic to say all kilns should be chained over
the same percentage kiln lengths.

The surface area used in existing kilns for burning plus calcining was calculated and
related to the best outputs. This gas a specific surface of C Zone/B Zone expressed at
tpd/m2

From data collected covering a wide number of kilns, a value of 1 tpd/m2 has
subsequently been used to represent the best practice for the present. Attempts to
extend the chain system into the calcining zone at Northfleet resulted in a reduced
output from the kiln due to the reduced CZ/BZ It was found that the reduced kiln
throughout restored the loading of the CZ/BZ to 1 tpd/m2 thus confirming this figure
to be acceptable in the design of a chain system.

However, this situation should not deter us from trying to achieve higher BZ/CZ ratings
than 1 tpdim2 in the future. Already a number of kilns have exceeded this figure. The
answer may well lie in improving the heat transfer in the burning and calcining zones
by establishing better flame conditions. The latter would be greatly assisted by
improved heat recuperation within the coolers.

Using the above standard, the required surface area for the CZ/BZ at the
Watson/Walker rated output for the kiln can be calculated. This area is then deducted
from the total kiln internal surface area to give the area available for chaining and
hence the chained length. -An example of how the chained, length of a kiln depends upon
its shape is given by a comparison between Holborough 1 and Kirton 3/4 These kilns
have similar ratings based on surface area/velocity considerations, but have different
L/D ratios i.e. Holborough 22, Kirton 44.

Holborough 1 has an enlarged zone at both the wet and burning zones. Further, the
enlarged burning zone at Kirton accounted for 36% of the total kiln surface whereas for
Holborough 1 the value is only 18%. The combined effect of these factors is that the
recommended maximum chained lengths were Holborough 22.7% and Kirton 37.6%.

4.3.8 Starting Point of the System

The chains should start as near the cold end consistent with there being neither slurry
in droplet entrainment from the wet end chains nor slurry spillage due to damming.

16
CHAINED LENGTH (FEET) -

x x

X
3-5

1.5
I I I 1
20 30 40 50
LID RATIO

FIGURE 8: TOTAL, CHAIN WEIGHT

A distance of 1.0 kiln diameter between the closure plate and start of the ‘chains is a
safe initial figure.

However, the starting point can vary depending on the L/D ratio of the kiln. For kilns
with a high L/D (e.g. over 40) the starting point may be as high as 3.5D whilst in a low
L/D (e.g. under 25) it may be possible to reduce the starting point distance.

An example of chain design using the above method is given in Appendix 1.

4.3.9 Total Chain Weight

The relationship between actual chain weight and clinker output is variable, but from
our operating experience it appears to be related to the kiln shape as shown in Figure
8. This figure can be used to give an indication to the likely tonnage of chain required
for a particular kiln.

However, since the graph does not take into account the effects of different slurry
moisture contents or exit temperature, it is necessary in practice to tune each system
according to its performance.

The main guide in this respect is the exit temperature since if this is too high then the
system requires additional chain weight. When tuning chain systems, care must be taken
to assess the weight in the existing system accurately. Performance data must be
related to the actual weight in the system and not the installed weight, since the effect
of chain wear can be significant.

19
5. SPIRAL CHAIN SYSTEMS - OPERATION

The following section covers the effect that changes in the chain system have on kiln
performance. The effects of lower slurry feed moistures dust loss and insufflation on
chain design are discussed.

5.1 Effect of Lower Slurry Moistures on Kiln Performance

and Chain Design

The effects of lowering slurry moistures on kiln performance are two fold;

a) Lower moistures lead to reduced kiln feed consumption and hence

temperature differences between the gas and the feed.

b) The reduction in gas velocity reduces the heat transfer rate.

The rate at which heat transfer takes place in a chain system can be
expressed by the following general expression:

Q U*A* TLM

where Q = heat required to evaporate slurry moisture

u = overall heat transfer coefficient
A = contact surface area
TLM = log mean temperature difference between gas and
material

for a decrease in slurry moisture from 40% to 30%, making some very broad
assumptions it can be calculated that:

i) Q decreases by 21%
ii) U decreases by 17%
iii) TLM decreases by 27%

In order to compensate for the above, the contact surface area A must increase by 32%.
Although the work to be done by the chain system is reduced, the ability of the chains
to transfer heat to the slurry is reduced to a greater extent. It is therefore necessary
to increase A by increasing the weight of chain in the system.

Thus, the programme of moisture reduction carried out on a wet process kilns over
recent years has been accompanied by corresponding increases in the overall chain
weight in many installations. Figure 9 illustrates the above principle for Norman kiln
over a period of 10 years. Over this period the weight of chain has approximately
doubled.

20
Another example of the effect of adding extra chain to a wet process kiln is shown in
Figure 10 for Kanthan kiln. It will be seen that the kiln output was raised during the
period 1967 to 1972 at the expense of back end temperature. From 1972 the weight of
chain was increased as the slurry moisture was reduced. This resulted in higher output
at reduced fuel consumption and back end temperature. By careful design of the chain
system initially it is possible to achieve all these benefits simultaneously without
elevating the back end temperature.

5.2 Dust Loss

When designing and tuning chain systems it is still our aim to minimise dust creation and
to this end a slow even drying rate with some moisture in the nodules leaving the chain
system is preferred. The drier the material leaving the chains then the smaller the size
grading of that material and the greater the likelihood of dust pick up. Some factors
which lead to fast drying rates, which are to be avoided in order to achieve lower dust
losses are:-

1. Increased throughput for a given chain system

2. Operating the kiln with high levels of excess air
3. Insufficient chain installed for the desired level of output
4. Insufficient chain installed for the feed moisture level
5. Insufflation of dust to the kiln with a given chain system

All these factors manifest themselves as high exit gas temperature. These findings and
others relating to kiln performance were confirmed by the chain trial carried out at
Barnstone which is described in Appendix 2.

5.3 Dust Insufflation

There can be no argument against dust insufflation, except on chemical grounds, when
yearly nett dust loss figures of less than 3% have been achieved. This represents a
substantial fuel saving over kilns operating under ‘open circuit’ conditions.

However, the kiln dust loss may increase with insufflation; the figures for Norman and
Oxford Works indicate that it doubles as a percentage of clinker output. The best
approach is, therefore, to try to minimise the initial dust creation and return 100% of
the dust to the kiln.

In order to do this and also reduce fuel consumptions the exit gas temperature must be
reduced to the lowest possible level accompanied by consistent kiln performance. This
can be achieved by increasing and maintaining chain weight. Any increase in weight
should comply with the recommended even density distribution.

21
i I

I I I I
1968 1978
YEAR-

FIGURE 9: NORMAN WORKS KILN 22

200

20

40

35

30

1967 1972 1978

Y E A R -

FIGURE 10: KANTHAN WORKS KILN 23

6. MAINTENANCE OF SPIRAL CHAIN SYSTEMS

It is to be expected that over a period of time a chain system will wear and lose weight.
Evidence of wear is usually seen in the first instance as an increase in kiln exit
temperature and secondly, when the loss in weight has become appreciable, as a drop
in clinker output. Figure 11 shows the relationship between these parameters for
Northfleet Works. For optimum kiln performance it is therefore important to maintain
chain weights by regular maintenance.

Accurate records of all changes made to any chain system as well as chain weights
within that system should be kept in order to assist planned maintenance of the system.

Figure 12 shows a typical chain diagram for Northfleet Works in which the position,
weight and type of chain are recorded together with any alterations and renewals made
to the system. An important part of record keeping is accurate weighing of the chains
in order to establish how much wear has taken place and what chain addition is needed.

6.1 Chain Weight Monitoring

Figure 13 shows some methods of weighing chains. Figures A, B and C show methods
applicable outside the kiln and Figure D shows the method used inside a kiln.

Figure A shows the whole chain suspended from a spring balance. This gives a direct
reading of chain weight.

Figure B shows the chain attached at one end and the spring balance held in a position
such that the end link is now at the same level as the attachment link. This method
gives half the chain weight and is useful when the total chain weight is outside the
range of the available spring balance.

Figure C shows the chain attached at one end and the spring balance attached at any
convenient point near to the chain. The free end link has been offered up to the same
level as the link at the attachment point and then the chain has been hooked onto the
spring balance using the link that is at the same level as the spring balance hook.

The actual chain weight is then twice the indicated weight as half the length of the
chain is supported by the balance.

The above method was tried out in the open several times in order to prove to our
satisfaction that method C gives half the weight of chain. It is suggested that anyone
using this method should do the same.

24
TOTAL CHAIN WEIGHT (TONNES) TONNES OF CHAIN TPH. B.E.T “C

/ -
/
+ /
/
+ /
/
/
l
/

++ /
 FEET 0 10 20 30 40 50 60 70 60 90 100 HO 120 130 140 150 160 170 180 190 200 210 220 230 240 250 260

SAMPLF PORT POSN 1 2 4 5 6 7

TOTAL

REMARKS - 1'4 SPIRALS RENEWED 25'0" - 52' 9" &

SUMMARY - 1 SPIRAL RENEWED 71'3" 89'9" USING S/H "TRAWLER CHAIN"
SPIRAL RENEWED 173'0" 182'3" USING MS 0 CHAIN

Kiln length 650’. Diam. 18’ 6” % of length chained 29.88 Attachments available 10 5 spirals

FIGURE 12: NORTHFLEET WORKS CHAIN SYSTEM KILN NO. 3 MODIFICATION NO. 28 - DATE 10.1.80
 FIGURE C
FIGURE B

FIGURE A METHOD APPLIED IN KILN

FIGURE 13: KILN CHAIN WEIGHTS 27

Figure D shows the method most widely used in a kiln. This is similar to method C and
again, half the chain weight may be measured. However, a greater degree of accuracy
may be obtained by suspending more than half of the chain from the balance. The
actual weight of the chain can then be found by relating the total number of links in the
chain, not forgetting the attachment link or shackle, to the number of links suspended
by the balance and the indicated weight.

In this matter, any discrepancy due to uneven wear along the length of the chain may
be minimised. The simplest method of weighing the chain is to hook the top of the
spring balance to the highest accessible link and then hang as much of the remaining
chain as possible from the spring balance hook.

To maximise the use of the information gained from chain weighing, it is essential that
chain is check weighed before installation and that records are kept of the chain weight,
both at installation and subsequently, so that over the years a picture of the wear
patterns will emerge.

It is obviously not feasible to weigh all of the chains in a kiln (there are over 3000 in
each of Northfleet’s 6 kilns) and a representative sample is therefore measured.

To maintain a degree of statistical significance, it is important that the same group of

chains is weighed each time. Usually, one chain in each turn of the helix is marked with
an extra washer (on the attachment bolt) and two adjacent chains, on either side,
weighed with the marked chain such that 5 chains are weighed in each helix revolution.

6.2 Wear Rates and Materials

The chain wear pattern in any system is essentially the same.

At the ‘wet’ end of the system the chains are worn down by the abrasive slurry acting
as a grinding paste. At the ‘dry’ end of the system the chains are worn mainly by
oxidation and chemical attack.

Trials are being carried out particularly at Northfleet to find an abrasion resistant chain
for the wet end of the system and a chain less prone to chemical attack for the dry end
of the chains.

Reduced wear in the wet end was achieved by installing chain manufactured to
withstand the abrasive conditions encountered in the trawling industry. The analysis
shows this to be EN 14 B "T" mild steel, which differs from the standard EN 5 mild steel
by an addition of approximately 1.6% manganese.

28
The cast round heat resisting chain of composition 25% Cr 12% Ni has given the best
performance to date. At Northfleet this type FMR 56 gave a life of some 4-5 years
before cracking of the links appeared with the resultant loss of chain. The trouble was
attributed to the material reverting to a ‘sigma’ phase and becoming brittle. A trial
of FMR 56 N will take place shortly. This is a heat resisting 25% Cr, 12% Ni chain with
a higher percentage carbon addition to try to overcome the reversion to ‘sigma’ phase.

22% Cr, 10% Ni heat resisting chain FMR has shown to have only about half the life of
the FMR 56. It has been noted that in the wet zone of the chain system a round link
chain wears faster than an oval link.

Figure 14 shows the results of trials with different chains at Northfleet and the results
of continuous monitoring of the chain systems to determine their wear pattern. It is
of the greatest importance to keep accurate records on the wear in the chain systems
in order to facilitate a planned maintenance operation to maintain the system at its
optimum performance.

6.3 Shell Lining in the Chain Area

Various methods have been tried for protecting the shell in the cold end of the chain
system. Bricking and refractory concrete can cause maintenance problems in this area
whilst shell protection by means of lining plates can prove unsuccessful due to
distortion.

An alternative method now more generally adopted is to weld 1 inch square steel bars
to the shell.

These are placed 6 inches to 9 inches apart axial to the kiln. Staggered welding of 3
inch runs alternating with 3 inch gaps, on both sides of the bar, have been found to be
satisfactory. Should the bar spacing prove to be too wide then additional bars can be
inserted at a later date. Kiln shells protected in this way have shown no wear after
several years use. The advantage of this method is that it is simpler and can be
installed on a piecemeal basis.

In the hot zone the best results in the UK have been obtained with brick as opposed to
refractory concrete. The latter has, in most cases, given a poor life and been generally
unpredictable.

Our recommendation would, therefore, be that brick is used in the hotter section of the
chains. Care should be taken in particular to ensure that tyre sections are adequately
protected by refractories.

29
 -- --
1
\
MILD STEEL
RING CHAIN i
1’
//

WEAR RESISTING
RING CHAIN
/AT’-
MILD STEEL OVAL CHAIN . /’
BEFORE USING HEAT
( RESISTING /
/ 1’
1’
1’ MILD STEEL
RING CHAIN
1’

/’/
p

DVAL CHAIN -1’

I \

HEAT RESISTING RING

CHAIN - FMR 53

- - - -I - -
1’ OVAL CHAIN 1 ~Y&ISTING R IN G
- 1’ OVAL CHAIN CHAIN - FHR 56

0 50 100
% O F C H A I N S’ISTEM-

FIGURE 14: NORTHFLEET - CHAIN WEAR RATES

 Square bars

Feed end

TAN:= D !
izl
Bricks Protection bars
I

+----I--+
FIGURE 15: SHELL LINING DEVELOPMENT
Normally, one would brick approximately 50% of the chained area of the kiln. However,
the proportion of the system lined with bricks will depend on the configuration of the
kiln with regard to tyre position. For example, if 50% of the designed bricked area
extended up to one of the tyre positions then it is normal to extend the bricking just
beyond the tyre position.

Figure 15 shows a development of a kiln shell which has been lined in the recommended
manner.

32
 APPENDIX 1

Chain Design - Worked Example

Kiln = 122 metres long

3.4 metres inside shell diameter
3.0 metres inside brickwork

Kiln Output Rating = 780 tpd

Surface Area Inside Brickwork = 1150 m2
L/D ratio = 40.7

Surface area required for calcining zone/burning zone = 780 m2

Area available for chains = 1150 - 780 = 370 m2

Kiln length available for chain system = 39.3m (i.e. 3.4m start gap and
35.9m chained length)
(32.2% of kiln length)

Weight of chain required, from Figure 21.9

Chain weight/clinker tph ratio = 2.1

Weight of chain = 2.1 x 780 = 68.25 tonnes

24

Length of spiral attachment with reference to Figure 16 below:

Figure 16: .- .

D = DIAMETER INSIDE SHELL (metre)

D1 = DIAMETER ATTACHMENT HOLE CENTRES
Equal t o D - 0 . 1 3 ( m e t r e s ) 33
 APPENDIX 2

Barnstone Chain Trial

Investigations into the design of kiln chain systems were carried out by Engineering R
& D Department of Barnstone. As part of these investigations two ten day trials took
place on Barnstone No. 2 kiln. The purpose of the trials was to provide accurate
operating data relating to a known chain system. From this data it has been possible
to show that there are relationships between variables such as BET or clinker output and
conditions within the chain system.

Kiln operating conditions were altered to provide five sets of results. This involved
changes in clinker output, BET and slurry moisture as indicated below:-

Clinker Output B.E.T. Slurry B.E.O2

Tonne/Hour C Moisture % %

1 3.42 209 39 3.4

2 4.37 214 39 2.1
3 4.20 305 39 5.7
4 3.94 260 45 3.3
5 4.64 274 40 3.7

Figures 17 and 18 show the main correlations found from the Barnstone trial. It must
be appreciated the extend to which the back end temperature relates to the
performance within the chains. For example high back end temperatures lead to faster
drying and higher dust losses caused mainly by the higher velocities and the breakdown
of material at the dry end of the system (indicated by the sample size gradings). It
must be appreciated that the conditions shown only relate to the kiln at Barnstone,
however, the trends observed have in practice been substantiated on all wet kilns.

35
 I=
CONDITION SLURRY
c%!%Eu:! BET MOISTURE BEO
TEST t/hr *I- ‘I. /.l

200 39 3 4

214 39 2.1
4 2 305 39 5 7
3 94 260 45 3 3
4 64 274 40 37 .
I 1 : .
Extent of chain system cl
I - I
I

DISTANCE FROM CLOSURE (ft)

FIGURE 17: VARIATION OF FEED MOISTURE CONTENT THROUGH CHAINS

 E
f 0
200 250 300 3 4 5
KILN EXIT GAS TEMPERATURE ( C) CLINKER OUTPUT t TONNE/ H R I

A. Feed Moisture Ex Chains versus B.E.T. B. Feed Moisture Ex Chains versus Output

t I I
0 2 4 6 8 10 12 14 0 2 i 6 8 10 12 14
FEED MOISTURE EX CHAINS I %) FEED MOISTURE EX CHAINS (%)

C. Nodule Size versus Feed Moisture Content D. Dust Loss versus Feed Moisture
Ex Chains

FIGURE 18: BARNSTONE CHAIN TRIALS

37
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 3

Semi Wet Process

(CTC 15)
 SEMI-WET PROCESS KILNS

CONTENTS

1. INTRODUCTION

2. THE CASE FOR CONVERTING

3. CHOICE OF PROCESS

3.1 Long Chained Kiln

3.2 Dryer + Preheater/Precalciner

4. FILTER PRESSING

4.1 Northfleet Conversion - Background

4.2 The Rittershaus & Blecher Filter Press
4.3 Other Types of Filter Press

5. SEMI-WET PROCESS WITHOUT FILTER PRESSES

1. INTRODUCTION

There are a large number of wet process cement kilns still operating around the world
today. Most of these had no alternative at the time they were built but to employ the
conventional wet process technology of feeding slurry direct to a long kiln. These
plants have high fuel consumption by today’s standards and this generally leads to high
fuel costs.

The prime reason for converting a wet process kiln to a different type of process is to
make savings on the fuel costs. Other good reasons can include :

- general modernisation
- increased output
l reduced production costs

It is sometimes possible to convert a wet process kiln directly to a modern dry process
system. However, it is more usual to convert to a process which is part way between
the wet and dry processes - this is known as the semi-wet process.

This paper will consider :

- the advantages and disadvantages of converting

- selection of the kiln system
- design considerations
2. THE CASE FOR CONVERTING

It was stated above that the main reason for converting a wet process kiln is the
reduction in fuel consumption achieved. Looking at a typical heat balance for the wet,
process the breakdown of heat consumption would be as shown in Table 1.

Table 1

Clearly the largest consumption of heat is in evaporating the slurry moisture. This is
also the one item which can be influenced significantly by choice of process. I
I
Modern kiln systems can produce clinker with a heat consumption of less than 700
Kcal/Kg clinker by using efficient preheater-s with up to 6 stages, calciners and more
efficient coolers. Table 2 compares the heat consumption of various processes and kiln
systems.

Table 2

It should be noted that some of these kiln systems have to be fed with dry raw meal in
order to achieve the stated heat consumption.

2
The potential for fuel savings is considerable. However, the kiln system can not be
considered in isolation as no advantage would be gained by putting in the most efficient
preheater if additional fuel then had to be used to dry the raw materials. The most
efficient design therefore is one which feeds the raw materials to the kiln as dry as
possible but which also utilises all the waste heat available from the kiln to achieve this
level of moisture content.

There are a number of different options for converting a wet process. One which may
be considered, if the raw materials are suitable, is to convert to a modern dry process
plant. However, this would mean replacing the entire slurry preparation plant with a
new dry grinding facility with blending and storage. In addition the kiln would have to
be speeded up, the slope possibly increased and, because it is likely that an increase in
output would occur, the cooler would also be inadequate. The capital cost of such a
conversion would be close to that of a new plant and is unlikely to be justified.

The most cost-effective solution will probably be the one which utilises as much of the
existing plant as possible particularly if the raw materials have a high inherent moisture
content.

If the existing slurry preparation stage is to remain in use the beat option is to
mechanically de-water the slurry by means of filter presses. There still remains a
choice of kiln system to which the filter cake is fed.

- long chained kiln

- drier + preheater/precalciner

A further option, which is dependent on the raw materials being suitable, is to feed a
mixture of slurry and dry raw material direct to a dryer/preheater combination, thus
avoiding the need for a filter press plant. This will be discussed later.

3.1 Long Chained Kiln

This involves simply feeding the filter cake to the existing kiln with a modified chain
system. The advantages are :

- minimum capital cost

- minimum downtime
- fuel consumption of 11 So-1250 kcal/kg clinker

The main disadvantage of this system is the increase in gross dust loss from the kiln.
This is caused partly by the fact that the kiln feed material dries out in the chains much
quicker than a slurry and this generates more dust. The dust must be returned to the

3
process and in doing so can increase the dust cycle significantly. It is not unusual to
have a 40-50 % gross dust loss from the kiln.

The basic flowsheet is as shown in Figure 1.

3.2 Dryer + Preheater/Precalciner

In this system the filter cake is fed to a crusher/dryer unit which uses kiln exhaust gases
directly to dry the material. One of the main considerations in the design of this
system Is to ensure that the choice of number of preheater stages will give the required
amount of waste heat for drying the filter cake. Generally speaking a 2-stage preheater
can supply sufficient heat for drying material up to about 28% moisture content. Below
19% moisture a 3-stage preheater may be sufficient.

A typical flowsheet Is shown in Figure 2.

3.2.1 Crusher Drier

The main piece of machinery in this system is the crusher drier. Its function is to dry
and disagglomerate the filter cake in one operation to provide a kiln feed material
suitable for a preheater system.

Although Blue Circle does not have any direct experience of this type of unit on filter
cake the F.L.Smidth design has been considered in the past. It is essentially a hammer
mill swept by hot gases with material fed in by a box feeder or screw. A separation
chamber on the outlet is designed such that only material of the required fineness is
carried through whereas coarse material falls back for ‘further pulverisation.

A static or dynamic separator may also be installed if better control of the particle size
distribution is required. However, this is not normally required with filter cake
prepared from slurry.

From the separation chamber the material passes into a dust collection cyclone before
being fed to the preheater. At this point the moisture content is about 1% and the
waste gas temperature 120-150 C

4
 -
FILTER PRESS

, c>
DUST INSUFFLATED DUST TO SCOOPS DOUBLE FLAP I -

DUST RETURNED DUST DUMPED

TO PROCESS
\
CLINKER

FIG. 1 Long chained kiln with filter cake feed

 SEPARATING
CYCLONE
I DUST

FILTER PRESS

Em

2 STAGE
PREHEATER

WATER
’ SPRAY

CRUSHER/DRIER
CLINKER ‘\

FIG. 2 2-Stage preheater and crusher/drier with filter cake feed

4. FILTER PRESSING

The heart of the conversion to semi-wet is the filter press plant. Its function is to
remove as much of the slurry moisture as possible and provide a continuous feed of
filter cake to the kiln.

Blue Circle has had direct experience of two types of modern filter press within the last
ten years. These are the Rittershaus & Blecher (R & B) unit and the Edwards & Jones
(E & J) unit. At the time of installation the R & B unit was a well proven but expensive
design from Germany whereas the E & J unit from a UK manufacturer was less
expensive but also less well proved. The two units are similar in concept with the major
difference being the press closing mechanism. The E & J press was installed at
Shoreham works which has since been closed down. However, Northfleet works in Kent
still has eight of the Rittershaus & Blecher presses in operation.

The Northfleet installation will be described here to illustrate some of the design
concepts and parameters. A flowsheet of the filter pressing facility is shown in Figure
3.

4.1 Northfleet Conversion - Background

Northfleet works has a unique combination of raw materials, technical and site
constraints which determined that it was originally built as a wet works and that the
conversion to semi-wet would have to be based on the filter cake route. These are
summarised below :

- The secondary raw material (clay) is situated on the, north side of the
River Thames whereas the works is on the south side. Clay slurry is
transported 11 km to the chalk quarry - also on the south side of the
river.

- The chalk quarry is 5 km from the works and chalk is transported to the
works by pumping as a finished slurry.

- The clay contains dissolved salts which prevent the effective use of slurry
viscosity reducing additives. The minimum slurry moisture is about 36 %.

- Both the chalk and clay contain about 20 % moisture as dug.

- The chalk also contains a significant amount of hard flint which can be
easily removed in a wet milling system but would be far harder to remove
in a dry grinding process.

The use of filter presses as a means of process conversion was therefore a foregone
conclusion.

7 i
 LIME

I.
DO!3 NG

.
-.-.B.-.-.-.w.m. ““‘;
AIR RECEIVER
I .
. I
I .
’ I -.-.-.-.-.-.-.-.q
I . I
I .

.I
,
I
I

.
SLURRY I .
1 PRESSURE .
I
.
I
.

J VESSEL I
.
FILTER PRESS ;
.
I UNDERFLOW TO
.
,.a SLURRY SURGE
TANK

1 1
SLURRY 1

@-
SLURRY
SURGE
I

TANK

FIG. 3 Schematic’ of filter press facility w KILN

00
4.2 The Rittershaus and Blecher Filter Press

The R * B filter press is a “plate and frame” press of the “overhead suspension” type.
It is a pressure filter and can operate at up to 25 bar although the Northfleet
installation was only designed for 15 bar. It incorporates a variable number of plates
which are covered with filter cloths and which form chambers between each pair of
plates when fitted together. The plate size is 2.0 m x 2.0 m and up to 132 plates can
be arranged together to form a bundle. ..

Operation of the filter press is on a batch process. It begins when the plate bundle is
compressed by a hydraulic closing device and the slurry is introduced into the chambers
via a central core. When the press is full of slurry pressure can be applied either by a
high pressure pump or compressed air (as at Northfleet) The filtrate drains along the
ribs of the plate and is collected in a common launder. An individual filter plate is
shown in Figure 4.

When the filtrate flow has dropped to a minimum the filtration process is considered to
be over and the pressure can be released before discharging the filter cakes. Some of
the parameters for the Northfleet presses are given below :

Chamber capacity = 0.343 Tonnes filter cake

No. of chambers = 125 per press

Press output = 43 Tonnes filter cake per cycle ’

Cake moisture (avge) = 19 %

Slurry requirement = 35.3 m3 per cycle

Filtrate produced = 13.6 m3 per cycle

4.2.1 Press Cycle Time

In order to calculate the required number of presses of a given size to feed the kiln it
is necessary to know the press cycle time. A typical range of values is given in Table 3.

Much of the cycle time is fixed and cannot be influenced significantly. However, the
largest single component is the actual filtration time and this may be quite variable..
Some of the factors which may affect filtration times are discussed in section 4.2.2.

9 I
Central core
for filling [
11111 support
- bosses
/

..

Recessed I‘
chamber
Filtrate
drainage
ribs

Fig. 4 Filter Plate

““CM*

I Flocculent Storaae I

Concentrate Feed Floe’

Feed
Beltwash Water

I
3~---
@____ .-

Mlxmg Drum
1
@ I Q

41

u
+6
....

I t —
-1
I
I Beltwash Water

// .

Fig. 5 Schematic of Continuous Belt Press

/
10
 I

Table 3

OPERATION MINIMUM

4. Pressurisation

4.2.2 Press Performance

If a filter press installation is being considered it is essential to carry out tests to fully
investigate the factors which affect filtration times and cloth life. At Northfleet a 10
plate 2m square press (identical to those eventually installed) was purchased in order
to determine the technical factors and economics of filtration of the Northfleet slurry.

Some of the factors which should be considered are :

i) What is the range of pressing times ?

If the kiln output is to be maintained the presses must be able to supply

sufficient filter cake when the pressing times are at their longest.

Pressing times can be affected by

- Clay type and content in the raw mix. A clay with a high Montmorillinite
content will lead to long pressing times as will one with low Silica Ratio
(less sand).

- Cloth blinding. If the filter cloth material blinds i.e. the spaces between
the fibres of the cloth fill up with particles of slurry then the cloth will
become impervious to water and pressing times will increase significantly.

1 1
 ● Slurry additives Hydrated lime has been found to be an effective
flocculent and an addition rate of O.1% (on dry raw meal) can reduce
pressing times by 20%. This is usually accompanied by a slight increase
in cake moisture.

The full range of pressing times can only be determined if the full range of possible raw
mixes can be teated. This is not very practical and it is likely that at some time in the
future an increase in pressing times may occur. The slurry additive should be used at
this time and will prove more cost effective than oversizing the filter press facility to
cope with unforseen raw material mixes.

ii) What is the average cloth life ?

The economics of the process can be significantly affected by cloth life and this
is determined by the following.

● Cloth blinding. Obviously if a cloth blinds to such an extent that the

filtration time is increased to an unacceptable level then the cloth must
be replaced.

● Mechanical failure. This is generally the main reason for cloth

replacement. It is caused by cloth flexing and tends to occur round the
/ edges of the plate and at the support bosses. At Northfleet flint damage
can occur if small pie- of flint find their way into the final slurry and
are then trapped between the sealing edges of the plates. The cloth is cut
and rapidly tears further unless replaced.

iii) Which cloth material should be sekcted ?

Different cloth materials must be evaluated in order to optimise press cycle

times and cloth life.

Modem filter cloths are woven from man-made monofilament fibrea. These give
improved cake release and resistance to blinding compared to the older type of
multifilament cloth. A material such as Trevi.ra B (used at Northfleet) is made
from a modified polyester which has good deformation properties and will
tolerate high pH (necessary if Hydralime is used). Some ordinary polyesters will
break down under conditions of high PH. The only other cloth which will tolerate
high pH is Nylon but this does not have such good deformation properties
although it resists abrasion better than polyester. At Northfleet Trevira B cloths
have a lifetime of about 12,000 cycles.

The other main factor to be considered in cloth selection is its cake release
properties. An R & B press automatically releases one cake every 4 sea. If a
cake fails to drop from the chamber then manual intervention is necessary. This
can significantly increase the press discharge time (and thus the overall cycle
time).

12
 It is important to note that cloth performance is very dependent on raw
materials. A cloth which works well at one plant may have little potential at
another.

4.3 Other Tw)es of Filter Press

Although the plate and frame type of prcsure filter is the only one currently being
operated by Blue Circle there are a number of other types available on the market.
Some of these have bees tested in the past on cement slurries and a brief description
is included.

4.3.1 Continuous Belt Presses

Two types of continuous belt filter press have been evaluated by Blue Circle. These are
the “Osprey” press, made in the UK by William Jones (Chemical Engineers) Ltd under
licence from Klein GmbH of Germany, and the “Andritz” prwss which is an Austrian
design.

The principle of operation of these presses is as follows.

● Flocculation. The slurry is first mixed with a flocculent solution in either a

pipeline or rotating drum. Preliminary dilution of the slurry with w~ter may help
this process.

● 1st stage water separation. The flocculated slurry is fed onto a porous belt
where filtrate water drains through under gravity.

o 2nd stage water separation. The drained material is fed between two polyester
filter belts which pass around a series of rollers of success ively decreasing
diameter. The effect is to apply pressure and shear at the same time and this
causes release of the water.

Good operation of these presses depends very much on the successful initial flocculation
of the feed slurry.

A schematic of the Osprey press is shown in Figure 5.

The major advantage of this type of press is that it is a continuous process thus
removing the need for intermediate storage of the cake. The size of individual units
is relatively small giving an output of about 13-15 Tph of cake at 22% moisture each.

13
4.3.2 Vacuum filters

The two main types of vacuum filter are the rotating drum and the rotating disc filter.
The principle of operation of the two is very similar.

Advantages of the vacuum filter are that it is a fully automatic and continuous process.
Cake thickness can be adjusted by varying the speed of rotation of the drum or disc.

The major disadvantage is that because it is a vacuum filter the maximum pressure
differential is vexy limited. It cannot therefore be used for materials which form
relatively impermeable filter cakes. The output of these units is also relatively low
compared to an R & B press for example.

5. SEMI-WI?’I’ WITHO~ FILTER P-

It was mentioned earlier that if the raw materials were suitable it was possible to feed
slurry direct to a crusher/drier unit without the need for a filtration stage. This is only
feasible if an additional@ raw material or additive is fed to the crusher drier at the
same time. The average moisture content of the combined feed may not exceed about
30%.

The main advantages of this process are that both the capital and operating co~ts of the
~filter press plant are avoided. This may have to be offset against a somewhat higher
overall feed moisture content.

Blue Circle has experience of operating such a plant at Rordal in Denmark. Some of
the factors which contributed to the decision to select this process are described below.

Historically the Rordal plant had always been wet process due mainly to the fact that
the chalk component is dug from below the water table and has a moisture content of
24-29%. It contains 1-5% flint. The clay also was dug from beneath the water table
and was slurried before being pumped some 4 miles to the plant. After adding pyrite
ash the resultant slurry moisture was still only 30-32%. The composition was based on
the subsequent addition of 10% fly ash to the burning zone and these factors accounted
for the low fuel consumption of about 1300 Kcal/Kg clinker.

Conversion to a completely dry process system was considered but quickly dismissed on
economic grounds as it would have entailed scrapping the existing slurry preparation
system and investing in a dryer, raw meal grinding plant, raw meal storage and blending
as weII as a new kiln system. A conventional semi-wet process with slurry filtration
was also considered as this would have retained the existing slurry preparation system.
However, it would only have been possible to reduce the slurry moisture from 30-32%
to 21%, with the addition of Hydralime, and this would have reduced the fuel
consumption to about 890 Kcal/Kg clinker. (This takes account of the fact that fly ash
contains 2-5% residual Carbon which reduces the consumption of “purchased fueltt). As
the capital and operating costs of the filtration plant would be quite high it was decided
to investigate whether the filtration step could be omitted totally. The estimated fuel

14
consumption was 135 Kcal/Kg higher at 1025 Kcal/Kg clinker but it was felt that this
was a small price to pay for avoiding the costs of filtration. It could always be added
in the future if the economics changed.

This is the process which was eventually selected, as shown in Figure 6. Initially
provision was made for either direct injection of slurry to the crusher/drier or premixing
in a screw with fly ash and recycled raw meal. However, the premixing screw was
found to be unnecessary as direct injection worked well throughout the range of feed
rates. (This had previously only been tried on a pilot plant).

The crusher/dryers were verys uccessful from a process point of view but the wear on
the hammers exceeded the expected 5 gin/t of clinker by a factor of ten. Various types
of alloys and hard facing were tried but none gave significant improvement. Eventually
it was determined that an accumulation of fine flint particlea in the crusher/separator
system was concentrating in the dry slurry coating on the internal surface if the
crusher/drier and was grinding the hammers away. The separatom were by-passed and
the wear rates dropped to the expected value without any significant change in clinker
burnability or quality.

15
 “.

SEPARATING
CYCLONE OUST

2 STAGE
PREHEATER

PRECALCINER ~ <1--’’’’”’’-”’
Y
1
* WATER
f SPRAY

KILN FLY ASH

t=
COOLER

CLINKER
o 0

CRUSHER/DRIER
?!!7--
%
SLURRY

FIG. precalciner and crusher/drier with slurry feed

 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 4

Lepol Process
(CTC 16)
 LEPOL PROCESS

CONTENTS

1. INTRODUCTION

2. PROCESS DESCRIPTION

2.1 Preparation of Nodules

2.2 Lepol Grate
2.3 Lepol Kiln
2.4 Clinker Coolers

3. CHAIUCTERISTICS OF LEPOL PROCESS

3.1 The Effect of Raw Materials Physical Properties on Plant

Performance
3.2 Kiln and Grate Chamber Coating
3.3 Grate Preheater Time Constant
3.4 Dust Loss

4. LEPOL PLANT PERFORMANCE

4.1 Output Ratings

4.2 Increasing the Output of Lepol Plant
4.3 Fuel Consumption

5. LEPOL PLANT CONTROL

5.1 Essential Controls

5.2 Instrumentation

6. FUTURE DEVELOPMENTS

APPENDIX I Properties of Raw Materials

II Nodule Porosity Test
III Determmlg the Strength and Grindability of Nodules
IV General Data for Lepol Kilns Worldwide
1. INTRODUCTION

The Lepol system was invented by Dr Otto Lellep and commercially developed by
Polysius in the early 1920’s. The first commercial Lepol unit was sold in 1928. The
early systems were single pass i.e. kiln gases pass through the bed of nodules on the
Lepol grate once. In 1952, the present day double pass system, where the kiln gases are
passed through the bed of nodules twice, was introduced. This development made the
system more efficient thermally and by 1973 approximately 65 million tonnes/year of
Lepol cement kiln capacity had been installed out of a total world cement making
capacity of around 650 million tonnes per year.

The Lepol process has now been superseded, for dry process operation, by the suspension
preheater.

The first Group Lepol plant was installed at Cauldon in 1957 and this was followed by
several other plants both in the UK and overseas. Currently, the Group operates only
two Lepol plants in the UK, at Cookstown and Weardale however, the Companyfs
knowledge base includes experience from a number of other Lepol sites. Lepol plants
similar to those used in the cement industry are also used to burn lime, magn=ia,
dolomite and in the heat treatment of iron ores.

2. PROCESS DESCRIPTION

Figure 1 shows a typical Lepol plant flowsheet. The dry meal is nodulised with about
12% water in the presetting screw and noduliser. These nodules are fed to the Lepol
grate where they are dried, preheated and partly calcined before passing into the rotary ~
kiln where the rest of the clinkering process takes place. Finally, the hot clinker is
cooled in a conventional cooler.

2.1 Preparation of Nodules

Although originally the raw materials were wetted and nodulised in the noduliser, on
some sites, it has been found that by ‘pre-wettingl the materials before they enter the
noduliser it is possible to achieve nodules of greater strength and more even size which
has given rise to more efficient kiln operation and increased clinker production.

Hence, the preparation of nodules now takes place in two stages. In the first stage, the
water and dry raw material are added to the pre-wetting mixers, while in the second
stage the material from the ‘pre-wettingt mixers is formed into nodules in the disk
nodulisers. Each kiln generally has two pre-wet mixers f ceding two nodulisers.

Figure 2 shows the raw material feed and nodulising system.

 ,Lw;~wa,&F
PtiE WETTING SCREW
NODULISER

I @ A EXHAUST AIR TO PRECIPITATORS

J,~y 1
‘TACK
COAL FROM COAL MILL
1

I
I m
,
\
KILN –. - –-. .

m
HOT AIR TO COAL MILL
I
INTERMEIJfiTE
CYCLONES
I CLINKER COOLER

II
5%’=5
I J

COOL
COOLING AIR CLINKER

D
PRECIPITATOR

Fiq. 1. Typical Lepol Plant Flow Sheet

RAW MEAL FEED SET POINT ( FROM KILN PLATFORM)
FEED CONTROJ
SCREW
RATIO CONTROL
,.. WEIGH
FEEDER

.. ........... ............... ...... . .. ........ ..

+
/=~

PRE-

DISH NODULISER

Q ,@e
;,0
,- ‘;e”.O*t*.~.* . . . .

I
+ ““ LEPOL GRATE
TRAVERSER BELT , ;,t
‘
\ .
4
b
●

*.C . ... . . . . . . . . . . . . . . . . . . . .

+ “-
FJ1. 2. Raw Material Feed and Nodulising .—System

3
The proportions of water added to the pre-wet mixers and nodulisers varies from works
to works and is dependent on the properties of the raw materials and the type of sprays
employed. Sites are known where all water is added to the nodulising dish and at the
other extreme where 90% of water addition occurs in the pre-wetter.

In recent years a number of conversions from wet to Lepol process have taken place
where a slurry is filtered to produce a filter cake. This cake is then fed into a Lepol
grate, set in front of a shortened kiln.

2.1.1 Pre-Wetting

The dry, raw material is weighed to the pre-wetting screws and the water addition rate
is controlled by the weighfeeder signal, thus ensuring constant nodule moisture. The
feed rate of raw material which represents the kiln feed is controlled from the kiln
control room.

The mixers are double shaft contra-rotating paddle type with tungsten carbide wear
tips, having a typical length of 2.5 metres, shaft speed of 20-50 rpm and a feed rate of
50 raw meal tonnes per hour.

The water is either added through an open ended pipe or a simple spray system along
the central axis of the mixer according to the raw material characteristics.

2.1.2 Nodulising

Figure 3 shows a disk noduliser (Polysius). The disk is a shallow rotating drum mounted
so that its angle to the vertical can be varied. In the UK, the nodulisers have the
following typical dimensions:-

Diameter (m) Depth (m) Speed (RPM) Nominal Tonnes per

Hour of Raw Meal

3.5 1.0 14.8 45

4.0 1.0 14.4 50

The angle of the tilt is adjustable and is normally in the range of 25° - 35° to the
vertical. Two driven scrapers keep the sides and bottom of the dish clear of build-up.
 All dlmcnsmrrs m mm.

Type D H B G HB GH GL
L6 4600 900 1700 L900 2600 f+aso 5700
L& fAoo 90@ 1660 L800 2600 L7C0 5500
L2 L200 900 1620 4750 2600 L550 5270
Lo Looo 900 1580 4650 I 26C0 [ UOO 5080
38 I 3800 900 15L0 L600 I 2600 ‘ 4250 L880
36 ! 3600 900 I Ifboo L1OO 2200 i LOOO i ULO
34 3LO0 850 1350 Looo I 2200 3950 k2fbo
32 3200 800 1300 LOOO I 2350 3650 Lofbo
30 3000 750 1250 3000 2350 3L50 ! 3840
20 2800 700 , 1180 2600 I 2100 3200 3600
2!5 2500 650 1120 3500 2100 3000 33L0
23 2300 600 10LO 3300 2050 2800 3120
21 2100 5L0 ] 1000 ~ 2650 1L50 2500 2700
18 1800 f$70 8L0 I 2500 1L50 2300 2Uo
15 1500 400 I 770 2000 1150 1900 I 21L0
10’) \ 1000 250 I 750 1500 90C 1100 150C
I
7’ ) 1 700 180 620 1100 I 650 I lLOO 1200
,
Note : The construction of types 7 and 10 does not entwsly agree wtih the

dimension sheet. Both typss havs an intmltely vurlablc drwc and

no scraper,

~g. 3. Polysius Nodu(iser

2.1.3 Desired Nodule Prqxxties

Nodule properties significantly affect kiln performance and so it is essential to produce

nodules to as high a standard as possible. Ideally, nodules should, when fed onto the
Lepol grate, give a porous bed having a low resistance to gas flow i.e. good
permeability. In order to achieve this, nodules with the following properties are
needed:-

a) Uniform size (around 12 mm diameter).

b) Correct moisture level; excessive water will cause the nodules to stick
together, while too dry nodules will have insufficient strength and
therefore be likely to crumble. The optimum water content which is
dependent upon raw material characteristics is usually in the range of 11
to 15%.

c) Sufficient ‘green’ strength to withstand both, handling and the pressure of

the overlying nodules in the hopper and on the grate.

d) Sufficient strength to withstand rapid heating as the water and C02 are
liberated from the nodule. Nodules on top of the bed have to withstand
the thermal shock of being placed directly into the Lepol drier chamber
at 320°C.

e) Sufficient dry strength to withstand the transfer from the grate to kiln
and abrasion in the kiln.

2.1.4 Factors Affecting Nodule Properties

These are summarised in Table 1.

It can be seen that some of the factors involved are incompatible. For instance, as the
residence time decreases, the nodules become smaller, but have a weaker green
strength. In practice a compromise has to be accepted where uniform small size is the
major consideration. This is particularly so, as nodule bursting due to lack of porosity
becomes negligible with small sized nodules of 10 mm and less.

2.1.5 Future Developments in Nodulising

Although it is generally accepted that the use of pre-wet mixers has improved nodule
quality considerably on some works, there still remains some work to be done to
optimise the energy input at the pre-wet stage.

Techniques used by industries which use very poor materials for nodulising such as
screening and recycling of undersized nodules would be extremely difficult to employ
because of the large quantities of nodules involved.

6
 TABLE 1:- FACTORS AFFECTING NODULE PROPERTIES

DESIRED PROPERTIES CONTROLLING FACTORS COURSES OF ACTION

Uniform Size 1. Constant meal and water 1. Automatic control of meal and water

2. Residence time 2. Adjustment of pan speed, angle and depth

3. Water distribution 3. Positioning and type of water sprays in prewet

mixer and noduliser
Correct Moisture 1. Constant meal and water 1. Automatic control of meal and water
Green Strength 1. Moisture content 1. Automatic control of meal and water ratio

2. Increases with size 2. Long residence time in noduliser

3. Work expanded on nodules 3. Long residence time in prewet mixer, also long
residence time in noduliser (low angle, high depth)
4. Increased clay content or weathered shale
4. Plasticity of raw materials
Bursting Resistance 1. Large nodules more prone to 1. Small nodules require minimum residence time
bursting

2. Porosity, dense nodules have low 2. Maximum porosity, with low residence time
bursting resistance
Abrasion Resistance 1. Dense nodules i.e. low porosity 1. As for ‘Green Strength’

2. Raw material characteristics e.g. 2. If possible, modify raw mix to include more plastic
friability index clay

4
Although the use of binders to improve nodule properti= has been generally
uneconomic, rapidly changing fuel prices could lead to an increase in their viability.

2.2 IAXXI1Grate

The Lepol grate is shown schematically in Figure 4. The grate itself consists of slotted
plates each 337 mm by 241 mm inches carried by an endless chain. It is driven at the
discharge end, the linear speed being in the region of 0.9 to 2.0 metres per minute.

The grate is totally enclosed. Above the grate there is a refractory lined chamber with
access doors for maintenance. Below, there is a system of hoppers and screw feeders
to remove the fine material, known as ‘riddlings’, which fails through the grate plate
slots.

The chamber above the grate is sub-divided by a vertically adjustable dividing wall
across the width of the grate. This wall is set to give a gas space above the bed of
around 50 to 150 mm. Below the grate, there is a similar sub-division, except that in
this case a complete seal is desirable. It is not always possible to achieve the complete
seal since the return strand of the grate has to pass through the partition wall. In
effect, the chamber is divided into four:-

Above Drier
Above Calciner
Below Drier
Below Calciner

as shown in Figure 4.

In the more modern and larger Lepols, which are mainly filter cake fed, the trend is to
sub-divide the grates into three upper, two or three lower chambers and where a grate
cooler is installed, to use the waste heat from the cooler exhaust to predry the
nodulised feed.

On some smaller Lepols, if nodule breakage in the drying compartment is a significant

problem, an extra fan is added to recirculate some of the below drier exhaust back to
cold end of the above drier compartment in order to reduce the gas temperature in this
zone and hence reduce the initial drying rate.

2.2.1 Nodule Flow Patt.exn

The nodulisers discharge directly onto a traversing belt ~hich evenly distributes them
over the area of the hopper. The level of material in the hopper must be such that it
is deep enough to provide an air seal, but not so deep as to cause compaction of the
nodules. A material height of 0.6 to 1m above the grate plates is the best compromise.

8
 AUXILIARY STACK u { I

BELOW CALCINER BELOW DRIER

PRECIPITATORS ~
(205°C TO 320”C-70 TO-80 nn-n.wg.) (115”C,-60 T0-85mm.wg.)

CYCLONES

Fig. w 4. Diagrammatic Layout of Lepol Grate

——

CQ
The nodules are carried on the Lepol grate beneath the cut off plate, through the above
drier and above calciner chambers and onto the kiln chute for entry into the kilnby
which timetheyhavebeen dried, preheated andpartly calcined (or decarbonated). The
depth of the nodule bedis usually between 170and220 mm.

2.2.2 Gas Path

Referring to Figure 4, the main gas flow enters the grate from the kiln, via the chute,
into the above calciner chamber and is then pulled through the nodule bed by the
‘intermediate’ fan into the below calciner chamber. From here, the gas stream is
cleaned by the bank of cyclones, circulated by the fan and returned to the above drier
chamber. The ‘final fan’ then pulls the gas through the drier bed. Thus, the gas stream
passed through the nodule bed twice, the so-called ‘double-passing’ action. The gas
stream is finally dedusted by the electrostatic precipitators before being drawn through
the final fan and being exhausted to the stack and to atmosphere.

The majority of the gases follow this flow pattern, however, a limited amount of gas
is allowed to pass under the wall above the grate from the above calciner chamber to
the above drier chamber. This operation is known as single passing and is allowed to
reduce the thermal shock as nodules pass from the above drier chamber to the above
calciner chamber.

Typical grate suctions for UK Lepol plants are shown in Table 2.

2.2.3 Fans and Cyclones

The gas flow through the grate is induced by two fans in series known as the
‘intermediate’ and ‘final’ (or ID) fans. A group of high efficiency cyclones are installed
downstream of the intermediate fan to:-

a) Protect the fan from dust abrasion and build up on the blades.
b) To avoid blinding the nodule bed above the drier with dust.

Details of the cyclone systems on UK Works are shown in Table 3.

2.2.4 Auxiliary Stack

A natural draught chimney with a movable lid is mounted above the calciner. It is
generally known as the auxiliary stack. This stack is used while the kiln is being lit.
Also when the grate stops, it is opened to release heat quickly in order to protect the
metal parts from being heat soaked. In the past, this has simply vented to atmosphere,
however, the modern trend is to route this to the main precipitator, simply by-passing
the Lepol grate.

10
 TABLE 2:- UK LEPOL KILNS SYSTEM PRESSURE LOSSES (mm wg)

Historical Data:
Currently Operating Works Works Now Converted to Precalciners

WORKS COOKSTOWN WEARDALE CAULDON DUNBAR

Kiln 1 12 123 123

Above Calciner 10 94 855 10 10 5

Below Calciner 67 70 70 65 60 55 55 55 70

Above Drier 13 10 5 25 15 10 20 20 15

Below Drier 102 90 95 65 55 50 70 70 80

PD Across Cyclones not available 135 135 115 115 115 70 70 70

Across Precipitator 38 60 38 25 25 20 50 50 115

Across I D Fan 140 203 175 203 203 203 165 165 230

Nodule bed 170 mm Nodule bed depth Nodule bed 165 mm Nodule bed 190 mm thick,
thick, 110 mm gas No. 1, 190 mmhio. thick, 150 mm gap 100 mm gap between bed
between bed and 2, 225 mm, 50 mm between bed and dividing and dividing wall
dividing wall gap between wall
dividing wall
 TABLE 3:- DUST AND RIDDLINGS SYSTEMS -UK WORKS

Historical Data:
Currently Operatiig Works Works Now Converted to Precalciners

WORKS COOKSTOWN WEARDALE CAULDON DUNBAR

Kiln 1 1 2 123 123

Cyclones
Four sets of two on each kiln Three sets of two on each kiln
Number Four X 2 Threex2 Fourx2

Diameter mm 1829 1829 1829 1829 1829 1829 1829

Vortex Finder Diameter

mm 915 915 915 915 915 915 915

Spigot Diameter mm
229 229 229 229 229 229 229

Dust Loading Not available 1 ton per hour per kiln 0.6 tph per kdn 0.8 0.8 1.2

Outlet Dust Burden 700 mglm3 average

Seal Amangement Double Flap Valves Double Flap Valves Double Flap Valves Double Flap Valves

Riddlinas Svstem

Loading tph 1.0 3.0 4.5- 6.0 2.5 2.5 3.5

Riddlings System Not Returned Via slide valve into screw Drag/screw/screw 400 mm screw to elevator to grate
conveyorto cross screw to inlet
elevator to kiln inlet

Riddlings and cyclone dust returned

to noduliser

Precipitator Dust Dumped

2.2.5 Cyclone Dust, Riddlings and Precipitator Dust Handling Systems

The combined weight of these materials is usually 5’%0 to 15% of the clinker output.

Precipitator dust, usually about 1 to 2% of clinker output, is dumped. The riddlings and cyclone
dust handling systems are under review since there are several methods of disposal being
developed on the UK plants.

The methods currently used on UK plants are shown in Table 3. Most methods are an attempt
to improve on the original system of returning the combined dust, via an elevator, to the side of
the kiln/grate chute as shown in Figure 5. This is not a very ‘clean’ method since most of the
material is immediately re-entrained and recycled.

Cauldon Works was unique in the sense that the material could not be returned directly to the
grate/kiln system without upsetting the grate operation. It was, therefore, returned to the
noduliser feed system and, as suc~ it gave rise to a slightly unstable noduliser operation. This,
however, was more acceptable than the alternative grate instability. Since the material had to be
re-wette~ addhional water and, therefore, fiel was required which was equivalent to at least 14
kcal/kg clinker gross.

The UK Works have developed more towards either:-

a) Feeding the returns through the back of the kihdgrate chute and so returning the
material directly into the kiln feed.

b) Dumping in order to reduce the alkali recycling load on the kiln. Although this
gives a waste materi~ the volatile bleed effkct has ve~ significant beneficial effect
on kiln stability and production rate.

2.3 Le~ol Kiln

The Lepol kiln is much smaller than the equivalent output wet kil~ as the materials entering the
kiln are already partly decarbonated. Atypical Lepol kiln length/diameter ratio (L/D) would be
in the region of 16 to 20/1 as compared with a preheater kiln L/D of 15/1 and a typical wet kiln
L/D of 35/1. Lepol kilns normally have a maximum speed of 1.5 rpm.

2.4 Clinker Coolers

Akhoug~ generally, Lepol kiln plants are equipped with either Recupol or Fuller grate coolers,
there is no technical reason why satellite coolers, as used at Rugby South Ferriby UK or even
rotary coolers, cannot be used.

One development that has taken place recently, within the Group, is the use of waste heat from
cooler exhaust gases to augment the drying of the raw material during milling. This has reduced
the fbel consumption for material drying considerably.

A development on large titer cake fkd Lepol plants is to use cooler exhaust gas for the predrying
of nodules in a third Lepol grate chamber.

13
 5-150/0 CLINKER ,
L
) OUTPUT

NODULE FEED

)
)
J
\
FINAL
FAN
/

\
,.
\
‘1 ‘1

\
A PRECIP”S I
+
\
-1 HHI
lt~~ t HA TO TIP (-lo/o CLINKER
OUTPUT )
EL 1
////i///~/’ ‘-///’’-/’~,’
—-—— . RIDDLINGS
-1 --
-1 [OR
~ INTERMEDIATE
> t FAN
I
> I
//////////////
4–~ CYCLONES
m , ;~R

t To Tlp

Fiq. s . Diagrammatic Layout of Lepol Grate Dust Handlingfiy@.

3. CHARACTERISTICS OF LEPOL PROCESS

3.1 The Effect of Raw Material Phvsical Properties on Plant Performance

The effeet of nodule properties, which are largely determined by the types of raw material used,
on plant outputs is shown in Table 4. The test used to determine these nodule properties is
described in Appendix 1, together with the Pfefferkorn Plasticity test and a nodule porosity test.
These various tests are very usefbl tools when comparing the performance of different plants.

As can be seen from Table 4, the raw materials used in the Group Lepols are not ideal for the
production of strong nodules and so nodulising has to be carried out as precisely as possible.
Techniques such as automatic water control and pre-wetting are used to ensure that the nodules
are as near to their optimum properties as possible.

3.2 Kiln and Grate Chamber Coating

Since the alkahs in the raw material tend to be concentrated on the surface of the nodules, there
is a tendency for build-up to form on the grate chamber walls, chute to the kiln and on the kiln
lining. In the latter region the build-up takes the form of clinker rings. When excessive, this
build-up tends to restrict gas and material flow and hence reduces plant output.

The clinker rings were, until recently, removed by ‘shooting’ the offending ring with anything
between 500 and 5000 rounds of lead headed shells. More recently, a water jetting system using
a very high pressure, low volume, waterjet (7000 psi, 16 galhnin) has been used to thermally
shock and remove the ring. Water jetting is now used on a regular basis, not only to remove
clinker rings, but also to cut back build-up in the grate chamber and chute areas.

The heavy eoatin~ however, does have the advantage of reducing the refractory consumption to
around 0.35 kghome of clinker compared with typical consumption of 1.0 to 1.5 kghonne of
clinker for wet and dry process kilns.

Volatiles can also build-upon the material bed, restricting air flow through the bed. In the past,
this has been countered by opening doors in the hot end of the above grate compartment to chill
the volatiles and so reduce their ‘stickiness’. Whilst this is effective in the short term, it should not
be used as a long term solution. Effective control of kiln conditions with in extreme cases,
selective dumping of cycle dust will be a better all round solution.
 TABLE 4: THE EFFECT OF NODULE COMPFU2HVE STRENGTH AND FRIABILITY ON KILN OUTPUTS

NODULE PROPERTIES OUTPUTS

OVEN DRIED 150°C FURNACE HEATED 800°C BEST WEEK %

(+1 56 HRS) WATSON/WALKER
PLANT COMP. STRENGTH COMP. STRENGTH OUTPUT TPD% PLANT RATING
LB-LOAD % FRIABILITY LB/LOAD % FRIABILITY

Villaluenga 100 0.05 74 0.02 1510 98.7

Cordoba 82 0.04 73 0.09 1350 88.2

Cookstown 71 0.16 70 0.14 1403 92.0

Golden Bay 63 0.61 48 1.03 785 67.1

Cauldon 43 0.095 47 0.46 787 69.0

Dunbar 30 7.96 26 1.58 1286 84.3

Weardale 31 3.27 24 3.34 1200 78.7

Mixcoac 29 6.10 18 4.90 837 59.7

3.3 Grate Preheater Time Constant

The process will take a considerable time (2-3 hours) tore-settle, following the changes
in parameters such as raw material feed rate, gas flow rate or gas temperatures. This
is because the quantities of materials which recycle such as dust and alkalis, take a long
time to reach equilibrium. This problem is accentuated by the comparatively long grate
residence time (i.e. the time the nodules take to pass through the preheater) of 20-25
minutes).

3.4 Dust Loss

Since the nodule bed tends to act as a filter, the dust loss from the system is relatively
low. Precipitator dust loss tends to be around 1.5% of clinker output, while cyclone
dust loss varies between 2% and 3%.

3.5 Exit Gas Conditioning

Since the gases leaving through the grate are humidified by the water from the nodules,
no further gas conditioning is required for precipitation.

3.6 Clinker Nodules

The nodules tend to pass through the kiln intact, giving a uniform sized clinker with only
a few large accretions and some fines.

3.7 Inleakiruz Air

One of the major weaknesses of the Lepol preheater is the high level of inleaking air
into the system. These inleaks are induced by the grate suction of between 2.5 and
100 mm w.g. and the main areas are though the nodule hopper and through the various
seals e.g. inspection doors, flap valves, rotating shaft seals, casing joints, etc. The
grate exit oxygen level is approximately 9% equivalent to 38% of the gases being
inleaking air and at the final fan, levels of up to 50% inleaking air (100% dilution of the
kiln gases) are not uncommon.
4. LEPOL PLANT PERFORMANCE

4.1 Outt)ut Ratius

In principle, the output of a Lepol Plant is determined by 4 main factors:-

a) Grate Size - there is a point where the grate loading exceeds the grate’s
capabilities, either in terms of heat transfer or pressure drop.

b) Gas Velocity in the kiln - this will limit kiln output to the point where
dust pick-up from the kiln blinds the grate.

c) Kiln Size - there is a limit to the amount of heat that can be transferred
within a kiln and this is mainly dependent on its surface area.

d) Material Quality and Nodule Preparation - the more friable the nodules,
the lower the limiting gas velocity at which dust pick-up blinds the grate.
Grate blinding can also be caused by weak nodules breaking down on the
grate and by alkali condensation on the grate nodules.

In general, the first and last of these factors are normally considered to be the most
significant.

4.1.1 Grate Size

Grate loadings are conveniently expressed in terms of tonnes per day clinker m2 (grate
area). A best perf orrnance target of 14.5 tpd/m2 has been set. From Table 5, it can
ben seen that with the exception of Cookstown, the outputs of the Group Lepol plants
fall well below this level. Following major maintenance to the grate, improved
operating techniques - including the installation of LINKman - and modification of the
front-end of grate cooler to an air beam type, Cookstown performance now approaches
the best known as defined in grate loadings.

4.1.2 Gas Veloci~

The gas velocity can be equated to tonnes per day clinker/m2 (kiln cross sectional area).
A best performance target of 150 tpd/m2 has been set. Table 6 shows the group
performances expr=ed in this way and once again it can be seen that they fall well
below the target.

18
 TABLE 5:- GROUP LEPOL GRATE DIMENSIONS AND PERFORMANCE

GRATE DIMENSIONS (M)

BEST WEEK
WORKS KILN LENGTH GRATE BEST WEEK OUTPUT
AREA OUTPUT TPD/m2
DRYER CALCINER TOTAL WIDTH m2 TPD GRATE AREA

COOKSTOWN 9.2 18.4 27.7 3.9 110 1570 14.3

WEARDALE 1 9.2 18.4 27.7 3.9 110 1200 10.9

2 9.2 18.4 27.7 3.9 110 1147 10.4

CAULDON 1 9.0 11.9 20.9 3.9 81 814 10.0

2 9.0 11.9 20.9 3.9 81 768 9.5
3 9.0 11.9 20.9 3.9 81 787 9.7

DUNBAR 1 9.0 11.9 20.9 3.9 81 912 11.3

2 9.0 11.9 20.9 3.9 81 919 11.3
3 9.2 18.4 27.7 3.9 110 1286 11.6

30LDEN BAY 2 21.1 3.0 65 449 6.9

3 20.9 3.9 81 772 8.9

rlIXCOAC 1 8.7 17.4 26.1 3.7 97 837 8.6

 TABLE 6:- GROUP LEPOL GRATE DIMENSIONS AND PERFOHNCE

BACK-END INTERNAL SA OUTPUT SPECIFIC OUTPUT

KILN DIMENSIONS (M)
X-SECTIONAL BEST WEEK BEST WEEK
(M’)
AREA (M) (+156 HRS)
tpd/m2
~N DIAMETER LENGTH TPD
WORKS BE. X- tpd/m2
SECTION INT SA
(Gas Velocity) (Kiln Size)

672.6 1570 128.2 2.33

COOKSTOWN 4.35/3.93 60 12.25

672.6 1200 98.0 1.78

WEARDALE 1 4.35/3.92 60 12.25
672.6 1147 93.6 1.70
2 4.35/3.92 60 12.25

515.9 814 89.2 1.58

1 3.81 48.16 9.13
CAULDON 87.2 1,52
48.12 8.81 506.4 768
2 3.75
506.9 787 89.3 1.55
3 3.75 48.16 8.81

672.9 912 100.4 1.36

1 3.80 63 9.08
DUNBAR 101.2 1.37
63 9.08 672.9 919
2 3.80
672.6 1286 105.0 1.91
3 4.35/3.92 60 12.25

419.0 449 97.4 1.07

GOLDEN 2 2.74 57 4.30
512.7 722 79.5 1.41
3 3.8 48 9.08
BAY

756 837 68.3 1.11

1 4.35/3.92 65 12.25
MIxCOAC
 .

TABLE 7- GROUP LEPOL GRATE DIMENSIONS AND PERFORMANCE

LEPOL GRATE DIMENSIONS KILN DIMENSIONS WATSON / WALKER LIMIT

TPD

KILN GAS KILN FACTOR *

WORKS
LENGTH WIDTH AREA DIAMETER LENGTH BE X-SEC INT SA GRATE VELOCITY SA OUTPUT
(M) (M) (M) (M) (M) (M*) (M*)

COOKSTOWN 27.7 3.9 110 4.35/3.92 60.0 12.25 672.6 1595 1525 1525 VELIKS 1575

WEARDALE 1 27.7 3.9 110 4.3513.92 60 12.25 672.6 1595 1525 1525 VELIKS 1525
2 27.7 3.9 110 4.35/3.92 60 12.25 672.6 1595 1525 1525 VELIKS 1525

CAULDON 1 20.9 3.9 81 3.81 48.16 9.13 515.9 1170 1410 1170 GSIKS 1170
2 20.9 3.9 81 3.75 48.12 8.81 506.4 1170 1310 1130 KILN SIZE 1130
3 20.9 3.9 81 3.75 48.16 8.81 506.9 1170 1310 1130 KILN SIZE 1130

DUNBAR 1 20.9 3.9 81 3.80 63 9.08 672.9 1170 1310 1485 GRATE 1170
2 20.9 3.9 81 3.80 63 9.08 672.9 1170 1310 1485 GRATE 1170
3 27.7 3.9 110 4.35/3.92 60 12.25 672.6 1595 1525 1525 VELIKS 1525

GOLDEN BAY 2 21.1 3.0 65 2.72 57 4.30 419.0 840 630 990 VEL/KS 630
3 20.9 3.9 81 3.81 48.16 9.13 515.9 1170 1440 1170 GSIKS 1170

MIXCOAC 1 25.6 3.8 97 4.35/3.92 65 12.25 756 1400 1590 1700 GRATE S. 1400

M
4.1.3 Kiln Size

The most significant factor in relation to kiln size is the internal surface area of the
kiln and accordingly a best performance target of 2.25 tpd/m2 (kiln internal surface
area) has been set. The Group performances are shown in Table 6. The fact that these
fall below this level can be attributed to the effect of raw materials on gas velocity
rating and grate rating.

4.1.4 Best Performance Outpulxi

It will be noted in Table 7 that the limiting values have been given for Lepol plants in
the Group. This was found necessary as the maximum output achievable on one count
can exceed the limit on another. For instance, we have some kilns which are capable
of more output than their respective grates could be expected to handle.

4.1.5 Ancillary Equipment

It is appreciated that the above limits do not take account of the limits of the ancillary
plant such as Coolers, Coal Mills, Fans, etc. These must be considered separately for
each Lepol plant.

4.2 Increasimz the Oulmut of Lewd Plants

When considering uprating the outputs of operating Lepol plants consideration must first
of all be given to the limits to increased production. These limits may take the form
of ancillary plant items, such as coolers, coal mills, fans or may be the nature of the
raw materials.

The mechanical condition of the Lepol grate is also likely to be a limiting factor.

4.2.1 Use of Additives to Improve Raw Material Propextiea

Since the raw material properties are of such great importance to Lepol kiln
performance, the use of additives to increase nodule strengths and to reduce nodule
friability has been considered many times.

These additives generally take the form of montmorillonite clays, such as Bentonite
which tend to be sticky and hence, bind the nodules more strongly.

In the past, the use of these additives has generally proved uneconomic. However, with
changing fuel costs, it is possible that their use could become viable.

22
4.2.2 Use of Overgrate Burner

Another means of increasing the output of Lepol plants is the use of an auxiliary overgrate burner.
Theoretically, the burner which is situated in the above calciner region of the grate uses inleaking
air for combustion and changes the temperature profiles within the grate/kiln complex. The
operation of the overgrate burner is similar in principle to precalciners used on suspension
preheater plants and leads to a reduction in kiln gas velocity per tonne of clinker. It is claimed
that in addition to a significant increase in output (5- 10OA)kiln stability is improved and that there
is possibly a reduction in kiln build-up. The latter is almost certainly due to improved control of
the volatile cycles through operating the kiln with higher oxygen levels.

4.3 Fuel Consumption

Reported fbel consumption vary between 790 and 980 kcalkg clinker (gross) according to the
carbon content of the raw meal, level of dust recycling cooler efficiency, percentage running time,
etc. A typical figure is 840 kcal/kg gross. The semi-wet route using filter cake logs produced
from slurries by filtration has a fbel consumption of around 1050 kcalkg (gross). This higher fiel
consumption is a consequence of the higher moisture level of the feed.

In addition to the 840 kcal/kg gross, the semi-dry route requires extra energy to dry the raw
materials. Typical raw materials at 5°Amoisture as dug, for example, would require a iiuther 120
kcal/kg gross heat equivalent for drying the feed assuming a 50?40drying efficiency.

However, on those plants with grate coolers, the waste exhaust gas from the cooler can be used
to reduce the fuel requirements for drying of the raw material.

Table 8 shows typical heat balances for Lepol, Humboldt and wet kilns. Comparing the Lepol
Process with the wet process, it can be seen that the lower fbel consumption of the Lepol process
carI be attributed mainly to the reduction in free water loss associated with the low feed material
moisture. The dust loss is also significantly lower than in the wet process, whilst the overall fiel
efficiency of the Lepol process means that the combustion gas loss is significantly, reduced. The
radiation loss is lower with the Lepol process because of a smaller size kiln for the same output.
However, the lower fhel consumption does lead to some reduction in cooler efficiency as shown
by the increased cooler exhaust gas loss. The excess air loss on the Lepol process is greater than
on the wet process. Since some of this excess air merely dilutes the kiln exhaust gases, reduction
of irdeaking air will not necessarily directly reduce fiel consumption, as this will depend on the
location of the inleak.

The comparison of the Lepol Process with the Humboldt shows that although the raw material
moisture is much lower with the Humboldt plant, the need to inject water into the preheater on
this particular kiln means that the ii-ee water losses are similar on the two processes. If a
conditioning tower were to be used on the Humboldt kiln there would be a reduction in this loss
by about 25 kcaVkg. The excess air loss is also much lower with the Humboldt process than the
Lepol. However, the combustion gas loss is lower with the Lepol process because of the lower
exit gas temperature. In the case of the Humboldt process the exit gas is normally used for drying
raw materials.

23
TABLE 8:- TYPICAL HEAT BALANCES OF LEPOL, HUMBOmT * _ KILNS
,,

LEIXIL HUMBOLDT WIZI,

ATA

omea per hour 44.7 83.18 81.74

~raw coal feed H20 5.55 3.57 12.35
1raw coal consumption (WD3) 12.19 10.91 25.71
J standard coal consumption 11.99 11.96 22.03
J nodule/raw meal/slurry 13.0 0.2 38.8
loisture nil 17.8 . .
keheater water injection rate 78.99 78.90 7;.:
n3/hr) 3.23 4.00 3.04
h raw meai CaC03 + MgC03 0.10 0.40 0.40
h raw meal aiumina 3.6 10.0 7.60
6 raw meal organic carbon 99 93 53 ,,
6 dust on clinker 123 215 200
;linker temperature ‘C 18.3 22.9
hit gas temperature ‘C 10.2 3.60 1.66
Ixit gas analysis % C02 nil 0.15 0.05
kit gas analysis % 02
lxit gas analysis % CO

iEAT BALANCE kcal/kg

Gross)
396 396 387
theoretical heat 19 18 10
clinker 10ss 15 27 24
?aw meal C02 baa 67 96 166
~ombustion gas loss 164 151 697
Free water loss 41 14 10
Excess air ioss nil 5 4
co loss 5 ,, 5 45
Dust loss 94 99 48
Cooler exhaust gas loss ,,

TOTAL 801 811 1391

Less temperature correction 35 4 23

Total accounted for 766 807 1368

Standard coal consumption 389 837 1542

Heat ex slurry carbon 13 52 52
Total fuel consumption 852 889 1592
Heat unaccounted for 86 82 224
(Radiation etc)

24
5. LEPOL PLANT CONTROL

5.1 Essential Controls

The efficient functioning of the process depends very much on the co-ordinated
operation of the nodulisers, Lepol grate, kiln and clinker cooler. This inevitably means
that an appreciable amount of remote control and instrumentation is necessary. Almost
all the controls are operated by the kiln burner from a control/instrument panel situated
in the kiln control room.

5.1.1 Auxiliary Stack Damper

This is a vital control system and its purpose is to avoid overheating of the grate whilst
it is not in motion or when a fan drive fails. Movement of the damper is effected
manually by a wire rope and winch system so that it is independent of any power or
instrument failures.

However, due to environmental pressure, there are schemes to dilute the hot gases with
cold air and then exhaust the mixed, cooled gases through the precipitators to the kiln
stack. These systems have to be properly engineered and controlled to ensure that no
local overheating occurs. There have been considerable problems in the introduction
of satisfactory by-pass systems within the Group and, at present, none are operating in
a completely satisfactory manner.

5.1.2 Intermediate and Final Fan Dampem

The gas flow through these fans is manually controlled from the kiln control room by
power eyeliner operated multi-leaf dampers.

5.1.3 Kiln Feed

The dry raw meal feed rate to the kilns is controlled by automatic weight-feeding to
the pre-wetting screws. In those plants where pre-wetting screws are not used, the raw
materials are weigh fed to the noduliser dishes. The water addition control is a ratio
controller with the kiln feed rate as the set point.

5.1.4 Grate Speed and Bed Depth

Grate speed is adjusted through a variable speed motor. The bed depth is determined
by a variable height cut off knife above the grate piates which extends the full width
of the grate and is mounted at the bottom of the nodule hopper. This knife is
automatically adjusted to keep the hopper level constant. As the hopper is fed at a
constant rate from the nodulisers, constant feed rate to the grate is ensured.

25
A side benefit from this method of control is that the depth of nodules acting as an air seal is
constant and that variations in inleaking air from this source are much reduced.

5.1.5 Coal Feed Control

Both the Company’s coal fired Lepol kilns are ‘direct fired’. The raw coal is fed to the tills by
Besta Feeders. This feeder is basically a drag scraper conveyor driven through a variable speed
gearbox. The volume of coal delivered is controlled by the feeder speed.

A recent development is the use of the ‘loss in weight’ system for the control of coal feed to mill.
This method ensures that a constant weight is fed to the kiln rather than a constant volume, as
obtained horn the Besta or similar feeders. Unfortunately, this method has not yet been applied
to any of the Company’s Lepol kilns.

5.1.6 Coal Mill Controls

These consist, essentially, of airflow control of hot air supply to the mill. This is discussed in
more detail in the paper on Coal Firing.

5.1.7 Cooler Controls

These are discussed more filly in the paper on Clinker Coolers, but for grate coolers consist
essentially of grate speed and airflow control.

5.1.8 Kiln Speed Control

Kiln speed is adjustable through a variable speed motor. The normal range of speed is from 0.5
to 1.5 rpm.

5.2 Instrumentation

As indicated earlier in 5.1, the burner has to co-ordinate the operation of the different sections of
the system and, hence, it is necessary to provide him with sufficient itiormation i.e. adequate
instruments. The basic instruments (on a plant with a grate cooler) are:-

5.2.1 Temperature

Above and below drier (Lepol Grate)

Above and below calciner (Lepol Grate)
Cooler exhaust air
Clinker output
Coal mill inlet and outlet

26
5.2.2 Pressures and Suctions

Above and below drier

Above and below calciner
Below cooler grate (separate indications of each chamber)
Kiln hood
Coal mill inlet arid outlet
Coal mill outlet also serves as firing pipe pressure

5.2.3 Oxygen Level

Kiln exit
Grate exit

5.2.4 Speed

Lepol Grate
Kiln
Cooler Grate

5.2.5 Feed Rate

Raw meal tonnage

Nodulising water

5.2.6 Damper Positions

Final fan
Intermediate fa
Cooler exhaust stack

27
6. FUTURE DEVELOPMENTS

6.1 Umatti of Intermediate and Final Fans

On those plants where production is limited by inte~ediate and final fan capacity,
consideration is being given to uprating through increased intermediate and final fan
capacity. However, other factors, such as dust pick- up in the kiln will eventually
impose limits to the increase in output than can be gained in this way. Care must also
be taken to ensure that use of larger fans does not simply incre=e in leak around the
system, that incorrect relative operation of the two fans do= not create excessive
single passing.

6.2 Overmate Burner

There are now some 20 plants operating in Europe using the overgrate burner. The
operation of plants in this way reduces kiln loading and claimed benefits include
increased output, reduced fuel consumption and even reduced build-up.

6.3 Wet to Semi-wet Conversions

Conversion to semi-wet Lepol grate/kiln process with production of nodules, through

filter pressing is an attractive avenue for increasing the fuel economy of wet process
works.

6.4 Use of Additives to ‘Irmrove’ Raw Materials

The use of additives, such as montmorillonite clays and lignosulphonates to increase

nodule strength and to reduce friability has generally proved uneconomic in the past.
However, because of rapidly changing fuel costs, these additiv= may become viable ,and
their use is being reviewed in the UK.

28
PROPERTIES OF RAW MATERIALS

Pfefferkorn Plasticity Test

Specimen Preparation

1500 grams of the air dried sample are weighed out, enough distilled water is added
(about 14% by weight to start) to make it plastic and it is thorowzhly wedged. If the
sample crumbles when worked, it is too dry; if it sticks to the hands and working
surface, it too wet. The sample is placed in a plastic bag.

The lightly oiled mould is placed on a firm smooth surface and a small proportion of the
sample worked into the mould finishing flush with the top to give a specimen 40 mm
high x 33 mm diameter. Care must be taken to remove all voids in the specimen. The
mould is then split and the specimen carefully removed.

Plastici~ Test

The base plate and the plunger face are lightly oiled. The plunger is raised and held by
means of the lock and the specimen is placed in the centre of the base plate. Check
that the index line on the stem of the plunger will appear in the scale window. Release
lock and raise the plunger to its highest position and let it fall freely. Read the height
of the deformed specimen from the scale. The procedure is repeated with a new
specimen. The two values for deformed length should agree to within 0.5 cm, if not,
further specimens must be prepared until satisfactory agreement is reached.

After each specimen is deformed, it is weighed quickly and dried overnight at 110°C.
After cooling in a desiccator, the specimen is re-weighed.

A little water (10 ml) is added to the remaining sample and thorou~hlv wedged, a second
pair of specimens are moulded and the procedure carried out again. This is repeated
until the deformed height reaches approximately 0.8 ems. 5 or 6 pairs of specimens are
usually required to cover the range of moistures, bearing in mind if it crumbles when
worked it is too dry, if it sticks to the hands and working surface, it is too wet.

29
Calculation

The water content is expressed as a percentage of the weight of the wet specimen.

% H20 G1 . weight of wet specimen

G2 = weight of dry specimen

G1 - G2
% Water = x 100

Deformed Ratio:

~ original height (i.e. 40 mm)

hl = deformed height in mm

The points corresponding to the above values are entered on a rectangular co-ordinate
graph in which the ordinate represents the percentage of moisture and the abscissa the
deformation ratio h

The point of intersection of the ratio b = 3.3 with the curve gives a % water reading.
hl
This value is the Pfefferkorn Plasticity Index.

Example

A clay has the following characteristics:-

~ G1 - G2
hl x 100 % water
G1

6.70 30.5
3.06 27.8
1.90 25.8
1.40 24.0

These values are entered on a graph and the curve is traced.

30
The value of ~ on the curve gives a value of 28.1% water.
hl

This is the Pfefferkorn Plasticity Index.

The curve constructed from the data given above is Figure 6 and Figure 7 gives similar
curves for the raw materials of the UK Lepol plants.

31
 +-
.-
.-U
n
t I I I I I I
0 0 0 0 0 0 0
G .4 Ik d 1+ Al A 0
#-
32
 A

.. —. ———__—_ —_______ ——. ———————_—_— —___ __ —____ —___________

I

A - Cookstown Works

B - Cauldon Works

C - Weardale Works
D - Dunbar Works

00: I ! I I I I f
15 16 n 10 19 al 21 22
WATER ?4.

fjg+~ 7. Pfef ferkorn Plasticity Index - U.K. Lepol Works

Nodule Porositv Test

Sampling Samples from the two disha are taken and graded -
9rnm + 3mrn.

Sampling Blending - The nodule from both dishes are well mixed.

Sample Weighing - 200 gramrnes are weighed out in the laboratow and
placed in a tray.

Sample Drying The tray is placed in an oven where it remains for 4

hours for moisture content determination and sample
preparation.

Sampiing for Porosity - 100 grammes are weighed out from the dried samplea
after moisture determination and placed inside a wire
basket, which in turn, is inserted in a Buchner Flask.

Vacuum A vacuum is maintained in the Buchner Flask for 15

minutes via a mercury leg system.

Filling the Pores - The vacuum is maintained in the flask while the nodules
are saturated with paraf f in, drop by drop, until they are
almost covered.

When the nodules are saturated, the paraffin entry valve

is closed and the vacuum maintained for 5 minutes
more.

Weighing of Saturated
Nodules The nodules are withdrawn from the Buchner flask.

They are then lightly dried with filter paper.

The nodules are weighed.

Porosity Determination - In accordance with the following formulae:-

v~ = Weight of dried nodules = 100 m-ammes = 38.16

Density of the SIUITY 2.62

34
V.2 = Weight of ~araffin saturated nodules - weight of dried nodules
Density of the paraffin

= X - 100 marnmes
0.776

% porosity = v~ x 100
VI + V.2

35
 APPENDm m

~AL PROCEDURE FOR D~G

~ STRENGTH AND FWAMLH’Y OF NODULES PREP~ FROM KILN FEEDS

1. Introduction

In a Lepol type process one of the factors affecting the performance of the grate
is the physical properties of the nodules. If the raw materials used produce strong
nodules with a low friability, few problems are usually experienced in operating
such a process. If, on the other hand, the nodules are weak and very friable, the
bed can become ‘blinded’. Similarly, in a wet process, slurries producing good
nodules are likely to lead to small dust losses , whilst slurries producing weak,
friable nodules, can result in high dust losses.

Blue Circle has developed two set procedures for assessing the nodule properti=
of a mix - one assesses the crushing strength of individual nodules; the other
measures the friability (wear) of a sample of nodules after rumbling in a closed
cylinder. Both tests are normally carried out on oven dried and furnace heated
samples. This note presents details of these test methods.

2. Materials

For the above tests at least 10 kg (dry basis) of finished mix is required. This can
be prepared at Research Division from the individual raw materials or it can be
an actual sample of an existing works slurry or feed.

3. Nodule Premration

Initially all the materials are subjected to full chemical analysis. Mixes can then
be proportioned in the laboratory by grinding either dry or wet in a laborato~ ball
mill. In the case if of an actual sample of a Works feed or slurry, chemical
analysis a wise precaution to check that the sample is representative.

The next stage is to adjust the moisture content of the mix to give a plastic mass.
Preferably, the required moisture content should be determined from the
Pfefferkorn test, but a subjected assessment of the desired consistency is possible.
Starting with a dry feed, this can be achieved by mixing batches of approximately
2 kg of dry feed and about 340 ml of water in a Hobard mixer. Starting with a
sluny, the moisture content can be reduced either by filter pressing or by slowly
drying on a hot plate. If the latter method is used, it is essential that the sample
is continuously stirred.

36
A small cylinder about 22 mm across and 15 mm deep is then used to cut cakes from the sample
for nodule preparation. The cakes are ejected from the cylinder by a spring-loaded plunger. The
cylinder dimensions have been chosen to give, by rolling the cakes between the palms of hands,
spherical nodules of about 20 mm diameter. The friability and breakage load tests require
together the production of about 160 nodules.

The nodules are placed in a siigle layer on enamelled dishes and lefl for 12 hours in a drying oven
at 150°C. They are then removed to a desiccator, allowed to cool and divided into two batches.

One batch is used for testing as dried, whilst the other batch is weighed and heated in an electric
fiumace at 800”C for 75 minutes. The sample is allowed to cool in a desiccator, re-weighed and
the loss in weight used to calculate a figure for loss on ignition. This sample is then tested as for
the oven dried sample.

For special investigations, tests may also be done on nodules heated to temperatures other than
800°C.

4. Test Procedure and Presentation of Results

4.1 Friability

For this test a sample of nodules is taken as near as possible in weight to 380 g. These are placed
in a rumbler which consists of a 6 inch diameter x 6 inch long (1 5.2 x 15.2’ cm) steel cylinder with
smooth surfaces. One end-plate of the cylinder can be unscrewed for insertion and removal of
a test sample. The cylinder is driven through a gear box and vee-belt, giving a speed of 20.7
rev.hnin.

The sample is rumbled for 10 minutes, then carefidly removed from the cylinder onto a 14 BS
sieve [1180 micro-metres) on a sieve pan and the cylinder and end plates are also carefid brushed
out over the sieve. The fines are sieved out, care being taken not to cause fhrther breakdown of
the nodules by the sieving process. The undersize is then weighed and the weight multiplied by
6 (to bring it to an hourly basis).

The weight must also be multiplied by the ratio 380 to bring it to the
sample weight
cotion basis of380 g sample.

37
The friability is then expressed as a percentage viz:-

Friability = Wt. Dassin~ 14 BS sieve x 6 x 380 x 100%

380 sample wt.

Friability = Wt. uassin~ 14 BS sieve x 6 x 100%

sample weight

4.2 Nodule Strength

For this test some 30 nodules are selected at random and the load required to
break each nodule under compression is recorded using a 1000 kgf capacity
compression testing machine. The loading rate is approximately 0.025 mm/see and
the machine is fitted with load recording and automatic stop facilities. The result
is expressed as a mean of the individual breakage loads.

4.3 Use of Resuita

Both tests are to be regarded as comparative and the results have to be assessed
relative to the properties of nodul= made from other raw mixes. By experience
gained in examining nodules made from feeds used in very large number of kilns
for which the operating parameters are known, satisfactory predictions can be made
of the performance of a new kiln feed in either a Lepol process or a wet process.

38 I
 APPENDIX IV

GENERAL W4TAFORLEIW KILNSWORLBJIDE

LERILKILNAND GRATESIZESWITH PR(MUCTILN

L(MDINGS

A) cat ic Data Cases

DIAMETEI

GRATE
KILN KILN GRATE LOADI
KILN TPD Y
LENGTH BACK SURFCE TPD/F? KILN/ TPD/M2 DIMENSIONS GRAT~ TPD/M
CCHPANY KILN ND OUTPUT M FRONT MIDDLE END t+’ SURFACE CSA/ ? B/E CSA KCAL/KG W (M) 1 (M) 5A M AREA

Lafarge Frangey 1 886 54 3.6 3.6 3.8 554.18 1.5988 9.08 97. 5s 869 3.9 21 81.9 10.82

Dyckerhoff Lengerich 5 960 60 3.8 3.8 3.8 640.88 1.4947 9.08 105.74 862 3.9 24 93.6 10.26

8(X Cookstown 1 1570 61 3.92 3.92 4.35 672.6 2.3300 12.25 128.12 846 3.9 27.7 110.0 14.3

Bcc Weardale 1 1265 60 3.92 3.92 4.35 690.52 1.8319 12.25 103.23 811 3.9 27.7 108.03 11.71

Bcc Weardale 2 1265 60 3.92 3.92 4.35 690.52 1.8319 12.25 103.23 811 3.9 27.7 108.03 11.71

.afarge La Mane 1 1003 53 3.92 3.92 3.92 586.10 1.7113 9.73 103.07 775 3.9 24 93.6 10.72

.afarqe La Mal le 2 1051 54.5 4 4 4 616.38 1.7051 10.18 103.25 775 3.9 24 93.6 11.23

iolderbank Rochefort 1 1245 60 4 4 4.4 703.72 1.7692 12.57 99.07 789 3.9 26 101.4 12.28

;F Calcia 8effes 1 1800 68 4.2 4.2 4.6 840.27 2.1422 13.85 129.92 846 3.9 34 132.6 13.57

[talclmenti Castrovi11 ari 1 1649 62.5 4.2 4.2 4.6 772.31 2.1352 13.85 119.02 816 3.9 31 120.9 13.64

)yckerhoff Lengeriche 7 1600 65 4.2 3.8 4.6 775.97 2.0619 13.85 115.49 862 3.9 29 113.1 14.15

.afarge V D’Azergue 1 1305 65 4 4 4 735.13 1.7752 10.18 128.21 814 3.9 29 113.1 11.54

Iolderbank Merone 5 1530 65 4.4 4.4 4.8 844.04 1.8127 15.21 100.62 840 4.8 36 172.8 8.85

:F Calcia Ranville 1 1108 57 4 4 4 644.65 1.7186 10.18 108.85 783 3.9 24 93.6 11.84

F Calcia AirVault 3 613 38 3.34 3.34 3.4 353.37 1.7347 7.07 86.72 828 3.9 29 113.1 5.42

F Calcia AirVault 4 1391 65 4 4 4.4 762.36 1.8246 12.57 110.69 804 3.9 29 113.1 12.30

F Calcia hirvault 5 1336 65 4 4 4.4 762.36 1.7525 12.57 106.32 808 3.9 29 113.1 11.81

talcimenti ~atera 1 1730 562.5 4.2 4.2 4.8 785.40 2.2027 15.21 113.78 849 3.9 31 120.9 14.31

talcimenti [sola DF 3 1800 70 4.6 4.6 952.95 1.8889 16.62 108.31 826 4.8 39 187.2 9.62
— - —
B) General Data Base - sane sites no longer in production

-
OIAMETERS I

GRATE
KILN KILN GRATE LOADI
}IMENSIONS Y
KILN TPD LENGTH BACK SURF$CE TPD/M2 KILN/ TPO/M2 GRATE TPD/M
KILN NO. CMJTPUT H FRONT MIDDLE ENO M SURFACE CSA t.? B/E CSA KCAL/KG W (M) 1 (M) SA M AREA

Cauldon 814 48.2 3.81 3.81 3.81 516.36 1.5674 9.13 89.13 3.9 20.9 81.51 9.99
I
Dunbar

Dunbar
919

1286
63

60
3.8

3.92
3.8

3.92
3.8

4.35
672.93

690.52
1.3657

1.8624
9.08

12.25
101.22

104.94
3.9

3.9
20.9

27.7
I 81.51

108.03
11.27

11.90
I
Cwdoba 1377 65 4 4 4.4 762.36 1.8062 12.57 109.58 3.9 27 105.3 13.08

Vasilico 1546 60 4 4 4 678. % 2.2783 10.18 151.88 890 3.9 29 113.1 13.67

Vasilico 1393 60 4 4 4 678.58 2.0528 10.18 136.85 890 3.9 29 113.1 12.32

Villaluenga 1485 65 4 4 4.4 762.36 1.9479 12.57 718.17 3.9 27 105.3 14.10

Vi 1laluenga 1560 65 4 4 4.4 762.36 2.0463 12.57 124.14 3.9 30.3 118.17 13.20

Mixcnac

Golden Bay
839

504
65

57
3.96

2.74
3.96

2.74
4.42

3.2
768.28

446.48
1.1065
1.1288
12.69

6.16
66.10

81.85
3.8

3
25.6

21
I 97.28

63
8.62

8.00
1.5045
Golden Bay 785 48 3.86 3.86 3.86 521.76 9.40 83.49 3.9 21.7 84.63 9.28
1.2631
Sal isbury 400 36 3.2 3.2 3.2 316.67 6.16 64.96 3.2 21.6 69.12 5.79
2.5371
Cassis 35 2.73 1 2.73 I 2.73 ! 256.20 ! 4.26 152.44 2.4
18.5 l_4!L. 14.64
 1
——.—
)
Hm--”---”--–”-~”----’----”—~
1
1
,
!
!
1
I
I
co m I
CD I
–m–—----”--
s k
a)
z
b
0 ml
Q E
c1) n 1
-1 I
c)
I
I
I
I
I
1— — --------+-— ‘-
<
( I
1
I
41
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 5

Long Dry Process Kiln

(CTC 17)
 LONG DRY PROCESS KILNS

CONTENTS

1. INTRODUCTION

2. HISI’ORY

3. DESIGN ASPECTS

3.1 Sizing
3.2 Kiln Volume Load
3.3 Kiln Power

4. PROCESS AND OPERATION

4.1 Kiln Zones

4.2 Kiln Internal Heat Exchangers
4.3 Pressure Drop
4.4 Dust LOSS
4.5 Heat Balance of Fuel Consumption
4.6 Kiln Control

5. PROCESS OPTIMISATION

5.1 General
5.2 Refractories
5.3 Preheater Stage Addition

6. CHEMICAL FACI’ORS

6.1 Feed Chemical & Physical Factors

6.2 Alkali Cycles

7. ANCILLARY PLANT

7.1 Kiln Exhaust Dust Collection

7.2 Raw Material Drying
1. INTRODUCTION

The development of the long dry process kiln can be directly traced to the first rotary
kilns used for clinker burning. The long dry kiln retains the advantage over other
processes of being the simplest system with potentially the lowest manpower
requirements. Because of its simplicity, running times of over 90% should be achievable
and maintenance costs should be relatively low. Against these factors the long dry kiln
gives the poorest production rates per unit volume of kiln of any of the clinker burning
processes and although considerable fuel savings are available compared with the wet
process kiln, fuel consumption does not equal the levels obtained with preheater and
precalciner kilns.

This paper will examine the history and present design of long dry kilns, describe the
process aspects and identify the important areas for process optimisation.

2. HISTORY

The original UK patent for a rotary kiln process for clinker burning dates from 1885 and
a diagram of the process is shown in Figure 1. It can be seen that the design is
essentially a short dry process kiln with lifters to improve heat exchange. However, the
length to diameter ratio of the kiln proved to be insufficient for adequate exchange and
the kiln system was, initially, not further developed in the UK. The original design was
lengthened and enlarged in the United States and the forerunner of the short dry process
kiln emerged. Typical dimensions of these early kilns were 6-7ft diameter and 60-80ft
long, with productions of 35-55 stpd (30-50 tpd).

From this point kiln development diverged in three directions. In areas where materials
were too wet to be &ied economically the wet process became dominant. But in many
areas of the US dry raw materials were available and the dry kilns were developed in
two forms. It was recognised that the high exit gas temperatures from the short dry
kilns was a waste of fuel and that these gases could be used in a waste heat boiler to
generate steam and hence electric power. In remoter areas, with unreliable and
expensive external power supplies, this option was especially attractive and its use
became widespread. The other option was to lengthen the kiln from the original 12-15:1
length to diameter ratio to 30-40:1, offering a much greater area for heat transfer and
thereby reducing kiln exit temperatures and fuel consumption. The vast majority of
non-preheater dry process kilns in the world are now of the long dry type. Cheaper,
more reliable electric power supplies have made the waste heat boiler kilns uneconomic.

A further development which took place with long dry kilns was the use of multiple
diameters. Many theories were evolved by US and European kiln manufacturers about
the desirability of increasing retention time and hence kiln diameter in the preheater,
calcination and burning zones. All were tried in various kilns. The most common
survivor in long dry kilns has been the expanded chain/preheating zone. This offers the
dual advantage of longer retention and thus heat transfer in this zone and also lower gas
velocities to generate less dust entrainment. As larger kiln diameters in the burning

—.
.__.__ ..... _____ .. ... . .... _
zone became associated refractory life, these types were less frequently
supplied.

View of section

w.

:;,
;;
;:
..:
1“
m’
4 .,

z-

L
/
,..
,..
,.
>:
:.,.
..
,.
...-.

FIGURE 1: RANSOMES PATENT 13MWING OF THE CEMENT ROTARY KILN

2
3. DESIGN ASPECTS

This section on kiln design is intended to act as a guide to the general principles of
design and the importance of factors such as length, diameter, speed and volume load
on kiln production and fuel efficiency.

3.1 -

The most difficult aspect of kiln design is the clinker capacity which can be expected
from a long dry process kiln. This is treated in more detail in the paper titled “Output
Targets”, but based on current knowledge a rating of 9.2 stpd/ft2 (90 t/d/m2) of back-
end kiln cross sectional area would be a target figure. The cross sectional area referred
to is usually the kiln back end area inside brick. An example of this means of selection
would be:-
Required production = 1500 stpd (1360 t/d)

Expected target stpd/ft2 = 9.2, but use 8.5 as design

Therefore area required = 163 ft2

Diameter (inside brick) = 14.4 ft

Diameter (inside shell) = 15.7 ft (4.8m)

Having selected the nearest standard kiln diameter, a conventional length/diameter

ratio of about 35:1 (shell dimensions) would normally be selected as this has been shown
to give the required surface area for efficient heat transfer within reasonable
dimensions.

3.2 Kiln Volume Load. S1o~e and Smeed

The factors volume load, kiln slope and speed are all
interconnected. The volume load of a kiln is defined
as the percentage of the kiln cross sectional area
which is occupied by material. It is conventional
that this is expressed using raw meal bulk density
and feed rate. Figure 2 gives a summary of volume
load and the angle bisected at the axis of the kiln.
Centic angle a 0/0 of kiln filling
The normal range of volume load selected is between 110° 15.65 0/0
7 and 13%. As the principal limit to heat transfer in 105° 13.75 0/()
a kiln is the passage of heat from the surface of the 100° 1210 w)
material to the bulk, at higher volume loads the bed 95° 10.?0 0/0
depth of material in the kiln becomes so great that 90° 9.09 0/()
the overall heat transfer from the gas to the 85° 7.75 w)
material is reduced and kiln output and fuel
80° 6.52 o/o
?5° 5.40 W-J
70° 4.50 9’()

FIGURE 2: KILN LOADING

efficiency begin to suffer. At lower than 7% load, it is felt that more refractory is
exposed to direct flame radiation than necessary and kiln brick life is reduced.

If a volume is selected then the kiln average residence time can be simply calculated
ie:-

If kiln dimension is 540 x 14.4 ft (inside brick)

Total volume = 87944 ft2

Thus material volume = 8794 ft3 (10% volume load)

If kiln production is 1500 x 2000 x 1.55 = 4,650,000 lb/day where 1.55 is the raw
meal factor (assuming for simplicity
no internal dust c<,cle) and raw meal
density of 60 lb/ft J

Then kiln residence time = 8794 X 60 X 24

4,650,000

= 2.72 hours

Kiln residence time is also calculable by a formula first given by the US Bureau of
Mines. The formula is equally applicable to all rotary kilns i.e. driers, calciners.

T= 1.77 XL XAXF
PxDx N

where T = Kiln residence time (minutes)

L= Kiln length (ft)

D= Kiln diameter (ft)

N= Kiln speed (rPm)

P= Kiln slope (degrees)

A= Square root of the natural angle of repose of dry material (degrees)

6.325 is normally used for raw meal
F= A factor associated with zones of enlargement in kiln. For parallel
tube F = 1

By applying the factors already known.

163 = 1.77 X 540 X 6325 X 1

Px14.4x N

PxN= 2.575

4
 It can be seen that the slope of the kiln and the speed required are connected and one
must be selected before the other is specified. Early kiln development traditionally
used low speed ranges and higher kiln slopes, but it became recognised that since higher
kiln speed turned over the material in the kiln more frequently and improved heat
transfer so the trend to higher speeds and lower slopes proceeded. Modern kiln design
uses kiln slopes of 2-4% or 1.2-2.4°.

If 3.5% is selected for the example kiln then:-

P = 2.0 degrees

and since P x N = 2.575

then the speed required = 1.29 rpm

This speed will, of course, be averaged and a range of 0.3 to 1.5 rpm would be normally
selected in this case in order that lower and higher than average feed rates could be
accommodated.

3.3 Kiln Power

The power requirement for the kiln drive can be calculated on the basis of two
components. The first is the power needed to overcome the rolling resistance of the
kiln given by:-

HP1 = Wxbdxtd xNx F x0.000092

rd

where W = Total kiln vertical load on all supports (lb)

bd = Diameter of the roller shaft (inches)

td = Diameter of the kiln tyres (inches)

N= Kiln speed (rPm)

F= Coefficients of friction of the roller bearings

= 0.018 for oil lubrication
= 0.06 for grease lubrication
rd = Roller diameter (inches)

A second component is the power needed to maintain the movement of charge in the
kiln.

HP2 = (Dxsin 0)3x NxLx K

_... —.—.. ...—

where sin e is determined from volume load (see Figure 3).

D Kiln diameter inside lining (ft)

L = Kiln length (ft)

K = 0.00076 for normal cement kiln

0.0092 for rotary lime kiln
0.0018 for cooler or dryer with lifters

N Kiln speed (rPm)

The sum of the two components HP] and HP2 will give a minimum size for kiln drive,
but an overdesign is usual to allow for higher kiln production, speed or use of kiln
internals.

20
18
.n
16
14
12
10
8
6
4
2
0
0.4 0.s 0.6 0.7 0s 0.9
Sinus
Sine

FIGURE 3: DEIXRMIN ATION OF SIN e FOR DIFFERENT KILN LOADING

6
4. PROCESS AND OPERATION

4.1 Kiln Zones

In common with all kiln processes the long dry kiln can be divided into regimes of
different temperature which occupy certain zones of the kiln. Typically these would
be:-

% kiln lenpth Material temperature

Preheating zone 30 100- 1250°F (50 - 650°C)

Decarbonation zone 45 1250- 2000°F (650 - 1100°C)

Burning zone 20 2000- 2650°F (1 100- 1450°C)

Cooling zone 5 2650- 2400°F (1450 - 1300°C)

Figure 4 shows the gas and material temperatures graphically. In the division of zones
it is not possible to be specific as the zones and temperatures will vary with kiln
internals, quality of kiln flame, kiln feed composition, etc.

4.2 Kiln Internal Heat Exchangers

The principle of improving kiln heat transfer and thus increasing kiln output and
reducing fuel consumption has long been recognised. The first patent on the rotary
cement kiln (Figure 1) included a set of rudimentary lifters for that purpose and
development of efficient kiln internals continues today. All designs have common
features; all internals aim to increase the transfer of heat by exposing greater surface
of the kiln charge to gas, and all internals are heated in the gas stream and take this
heat into the material as the rotation of the kiln moves them under the kiln charge.
The efficiency with which each design succeeds in transferring heat whilst avoiding the
stirring up of excessive quantities of dust has been the subject of much debate. Some
typical designs are:-

4.2.1 Ceramic Cruciform

Two designs are shown in Figures 5 and 6. The designs largely rely for their efficiency
on the division of the flow of a greater surface area per unit weight of material to the
gas stream, improving heat transfer. These systems have been used in conjunction with
chain systems and are usually installed over 5-1 2% of kiln length in the hotter zones of
the kiln where wear of metallic elements would be rapid. Their disadvantages are the
cost of installation and maintenance and the higher gas velocities through the decreased
free area of the kiln, possibly increasing dust loss ex-kiln.

7
 o 0 o
0 0 0 o
u) o Ua 0
elm u) 0
,0.
4’
4
48
0
●
8°
;
9’
6°
0
u) :“ n
z
~ :“ <
1
0 0 000000 0
0 0 0000 00
0 m 0 0 m
*mm %
do mmlwwdlml
FIGURE 5:
FIGURE 6:
CERAMIC CRUCIFORM
CERAMIC CRUCIFORM
~
ICROSS SHAPED)

FIGURE 7:
CERAMIC LIFrERs FIGURE &
KILN STIRRERS

9
4.2.2 Ceramic Lifters/Stirrers

Again generally used in the hotter zones of the kiln the various designs of lifters (Fi~e
7) and stirrers (Figure 8) attempt to expose more s~f ace area of material to the gas
stream. The danger with the lifters in Figure 7 is that by @opping material into the gas
stream, dust loss from the kiln may be greatly increased. Stirrers attempt to increase
material surface area without entraining dust.

4.2.3 Metallic Lifters/Stirrers

At gas temperatures less than 2000°F (1 100°) heat resistant steel shapes may be used
rather than ceramic. Designs are very similar to the ceramic equivalents. Although
corrosion of the special steels can be a significant cost, the risk of breakage and
premature failure sometimes found with ceramic materials can be avoided.

4.2.4 Metallic Cruciform

Various designs (e.g. Figures 9 and 10) have been used and have the advantage of over
ceramic systems of more complex arrangements and a smaller impediment to gas flow.
However, they are limited to gas temperatures less than 1800°F (l OOO°C) and have been
replaced by more easily maintained chains which can be used in this zone.

4.2.5 Chains

Without doubt, the most effective form of heat exchanger applied to long dry kilns has
been the chain system. Historically, the use of chains in long dry kilns was restricted
by the quality of heat resistant chains which were available, but as improved qualities
of cast or fabricated chrome/nickel alloys have become available the chain system has
been extended towards the hotter zones of the kiln and greater efficiency of heat
transfer obtained.

Table 1 shows the various grades of special steels available together with normal
application.

Many designs of chain system exist for long dry kilns. One common feature is that the
garland system employed in some wet process kilns is virtually absent in normal
worldwide practices for long dry kilns and curtain chains are almost universally
employed. The hanging arrangement varies from system to system with circumferential
chain hangers still a common arrangement. Blue Circle has specified spiral curtain
systems for long dry kilns since 1982.

The BCI chain design for long dry kilns is related to the system used in wet process
kilns. In current design the total chain weight selected corresponds to approximately
2.5t of chain per t/h of clinker production. The spiral arrangement and method of chain
hanging is exactly as the system used in wet process kilns and explained in detail in that

10
. .

-
FIGURE 9:
METALLIC CRUCIFORM

FIGURE 10:
METALLIC CRUCIFORM
 TABLE 1

ALLOYS AND APPLICATIONS

BASIC COMPOSITION TEMPEI#iTURE DESIGNATIONS

GRADE FUNGE (METAL) AEX ACI

FMR-50 18% Cr - 8% Ni 1200- 1500°F 304

FMR-53 22% Cr - 10% Ni 1550- 1750°F 302 HF

FMR-56 25% Cr - 12% Ni 1650- 1850°F 309 HH

FMR-62 25% Cr - 20% Ni 1650- 2000°F 310 HK

FMR-153 22% Cr - 4% Ni 9% Mn Ambient 1750°F

FMR-350 28% Cr - 50% Co Ambient 2250°F

GENEIWL DATA

SURFACE SURFACE SURFACE AREA

NO. LINKS AREA PER AREA PER WEIGHT AREA TO
SIZE PER FOOT LINK FOOT PER FOOT WEIGHT RATIO

5/8”x2 1/2” 4.8 19.3 in2 92.64 in2

4“0 27”8I ‘11”2 I

3/4” x 3“ 6.48 16.8
I
7/8” X 3“ 4.0
I I
33.5 134.0 8.27
I 15.9

1!! ~ JJv 4.0 I 39.5 I 158.0 11.0 I 14.4

12
paper. However, there are some major differences between the wet and dry chain
design, including the length of the kiln occupied by the chain system. If a similar length
of system to the wet was employed in long dry kilns, the hotter end of the chains in the
dry kiln would be exposed to much higher gas temperatures than for wet kilns. In order
to compensate for this difference the BCI system is conventionally hung over 17-24%
of kiln length. One to two kiln diameters of chain free zone are allowed within this
length at the kiln feed end. Normally, if greater than 1 diameter of free zone is left
spiral hangers are located in this zone for any future increase in chain weight required.

The range of kiln length for the chained zone may depend upon such factors as kiln
tapers, etc., but more importantly should attempt to allow free holes for locating
additional new chain when existing chain wear is detected. In this matter total chain
weight can be maintained without removing chain which has only suffered 20-40% wear.
This gives benefits not only in optimizing production and fuel efficiency, but also
reduces the costs of maintenance. Figure 11 gives a typical simple spiral curtain chain
system design for a long dry kiln.

The original Blue Circle system for long dry kilns used the same general rules as for wet
process kilns i.e.

a) 3 spiral starts for hangers.

b) Constant chain density.
c) Chain lengths % of kiln diameter (inside shell).
d) 7/8” chain size - this is chain only, no allowances made for hanger.
e) Spiral pitch 1 kiln diameter.

Design changes have been made to accommodate differences between wet and dry
kilns:-

● As space is relatively limited for the long dry kiln chain system compared with
wet process kilns, the use of a greater number of spiral starts has become
common and 4 to 6 starts may be specified.

● Variable chain density may be employed with higher chain density at the hot end
of the chains.

● A variable chain length may be specified, with longer chains at the system hot
end for dust capture and shorter at the cold end to encourage dust disengagement
from the gas stream.

● Y4W chain may be employed.

Only e) of the above list remains unchanged.

13
 Kiln Section

Elevation showing ex:en: of spiraling

FIG.11 SPIRAL CHAIN SYSTEM

FIGURE 11: SPIRAL CHAIN SYSTEM

14
4.2.6 Effectiveness

The effectiveness of any system of heat exchangers must be monitored regularly, both
for reduction of kiln exit temperature and dust. A factor which can confuse the
underlying exit temperature is the water spray used to cool the kiln exhaust gases. It
is essential that the quantity of water used is accurately monitored and used to back
calculate the underlying temperature of kiln exit gases. An example of this calculation
is shown in Table 2.

TABLE 2

CALCULATION OF “REAL” BACK END TEMPERATURE

Measured kiln exit temperature 500UF

Kiln production 860 stpd

Measured water flow 22 USgpm

Kiln exit oxygen 2.0%
Kiln fuel consumption 4.2 mBTU/ton
Feed calcium carbonate 78.0%

1. Calculation of gas quantities:-

lbs/lb clinker

C“2 ex feed 0.5485

ex fuel 0.4408

TOTAL 0.9893

N2 ex combustion gas 1.2547

ex excess air 0.1611

TOTAL 1.4158

02 ex excess air 0.0612

H20 ex coal 0.0180

ex feed 0.0032

TOTAL 0.0212

15

.—
 TABLE 2 (cent’d)

2. Total heat content of gases to kiln exit:

Heat content = (0.9893 X 0.245)

above O°F + 1.4158 X 0.250
+ 0.0612 X 0.232) X T
+ (0.0212 X 0.460) X (T - 212) +
0.0212 X 1184

= 0.6203T = 23.0

3. Total heat content gases ex kiln (including water):

22 usgpm = 176 lb/rein

Kiln production = 1194 lb/rein

.“. water spray = 0.1474 lb water/lb clinker

Heat content = (0.9893 X 230)

+ 1.4158 X 0.249
+ 0.612 X 0.227) X 500
+ 0.1686 X 1184 + 0.1686 X 0.450 X 288

= 297.0 + 221.1
= 518.1

4. “Real” kiln exit. temperature:

0.6203T + 23.0 = 518.1

.“. T= 798°F

Note: This calculation has, for simplicity, not included dust as part of the
calculation. Ifan estimateof dust quantity is available this should alsobe
included.

16
4.3 Pressure hOD

The differential pressure across the kiln is determined by the gas velocity through the
kiln. Gas volume is determined by clinker production and kiln fuel consumption. Gas
velocity will then be determined by the kiln cross sectional area and any kiln internals.
As such the range of kiln pressure differentials is large i.e. 1-6” wg (25-150 mm wg, but
in general the high length to diameter kilns with extensive kiln internals will have
pressure drops at the upper end of this range.

4.4 Dust Loss

No general rules can be applied to establish the absolute level of dust loss of any kiln.
This will depend on the characteristics of the raw meal and good design of the kiln
internals to avoid dust generation. An example of poor internal design would be a
bucket type design which would cascade the raw meal into the gas stream. The
mechanisms of dust pick-up and subsequent deposition in lower velocity areas of the kiln
are not well understood, but for a given raw meal and design of kilns internals the gross
dust loss from a long dry kiln will largely depend on gas velocity through the kiln. The
approximate relationship is:-

Dust loss = gas velocity 4

but the numbers which have been proposed for the exponent vary between 3.5 and 5.
The most obvious drawback to high dust loss is the overloading of the dust collection
system used for kiln gases. In the case of a bag filter system, the result will be an
increased differential pressure across the unit which will restrict the flow to the kiln
fan and reduce production. The case of electrostatic precipitators the dust loss to stack
will increase until emission control limits are exceeded and reduced production will be
necessary.

A less obvious effect of dust loss is the heat loss from the kiln represented by the dust.
Although the majority of long dry kilns have a direct return of dust to feed, heat loss
will comprise two components.

a) Loss of sensible heat as the dust leaving the kiln is reduced from kiln exit
temperature to the temperature of the dust collecting unit. A typical example
would be:-

17
Dust loss 20% on clinker

Kiln exit temperature (before cooling) 900°F (Tke)

Meal temperature ex-bag filter temperature 400°F (Tbf)

Heat loss = dust loss x specific heat x (Tke - Tbf)

= 0.2 x 0.2 x (900 - 400)

=. 20 BTU/lb clinker

= 0.04 mBTU/ton clinker

b) A less obvious heat loss is caused by the effective transfer of heat in the dust
towards the kiln back-end. An example would be that duct picked up at a
material temperature 1200”F would be quickly heated to the gas temperature at
that point. If this dust is deposited in the cooler zones of the kiln then the
overall temperature of the material in this area would be increased, reducing the
differential temperature between gas and feed and reducing heat transfer from
gas to feed. The result is higher kiln gas exit temperature and loss of fuel
efficiency.

4.5 Heat Balance and Fuel Consunmtion

A typical heat balance for a long dry kiln is shown in Table 3. Other than theoretical
heat the major constituents are the heat in the kiln exhaust gases, kiln shell loss and
cooler exhaust losses. If water injection were not used the kiln exit temperature would
be approximately 1000”F (540”C). A breakdown of the heat balance is shown in Table 4
compared with the performance of a suspension preheater system. Two major
differences emerge:-

a) Lower heat loss in the kiln preheater exhaust gases due to the superior heat
transfer characteristics of the suspension preheater.

b) The lower shell losses from the kiln as the surface area of the suspension
preheater kiln, per unit of clinker production, is only some 30% of the area of
the long dry kiln.
 TABLE 3

LONG DRY KILN - TYPICAL HEAT BALANCE

INPUT CONDITIONS

Kiln outputs t/h 40.0

% Raw coal H20 8.0
% Raw coal consumption 18.0
Kiln fuel consumption mBTU/ton 3.78
% Raw feed moisture 0.5 + water injection 200 lb/rein
% Feed CaC03 + MgC03 78.0
% Feed A12203 4.0
% Feed organic carbon 0.1
% Dust on clinker (gross) 25.0
(net) 0.0
Clinker temperature ex cooler ‘F 250
Kiln exit gas temperature “F 520
Exit gas analysis % C02 28.0
% 02 2.2
% co 0.0
Cooler exhaust temperature ‘F 500
Cooler exhaust quantity lb air/lb clinker 2.2

HEAT BALANCE FROM NOMOGIWh4S

mBIXJ

1. Theoretical heat 1.49

2. Clinker loss 0.085
3. Raw meal C02 loss 0.12 )
4. Combustion gas loss 0.44 )
5. Free water loss 0.34 ) Total exit gases 0.955
6. Excessive air loss 0.055)
7. co loss 0.0 )
8. Dust loss 0.035
9. Cooler exhaust loss 0.50
10. Kiln shell loss 0.79
11. Cooler shell loss 0.035
Less temperature correction 0.018 (t-32) 0.125

Heat ex organic carbon 0.045

TOTAL 3.81

19
 TABLE 4

COMPARATIVE HEAT OUTPUTS - LONG DRY AND PREHEATER KILNS

Long Dry Preheater

Actual Actual
% of total mBTU/lb % of total mBTU/lb

Theoretical heat 39.4 1.49 50.9 1.49

Clinker loss 2.2 0.085 2.9 0.085

Exit gas loss 25.3 0.955 16.7 0.49

Dust 10SS 0.9 0.035 0.3 0.01

Cooler exhaust loss 13.2 0.50 18.4 0.54

Kiln shell loss 20.9 0.79 9.6 0.28

Cooler shell loss 0.9 0.035 1.1 0.035

100% 100%

Actual total 3.78 mBTU/t 2.93 mBTU/t

20
4.6 Kiln Control

The normal instrumentation for the long dry process kiln is similar to the wet process
kiln and includes:-

Kiln exit oxygen

Kiln exit carbon monoxide or combustibles
Kiln exit temperature
Feed rate

Fuel rate

Kiln exit suction

Kiln speed
Kiln drive power or amps

In addition, other instruments may include:-

Kiln exit NOX, SOX

Burning zone temperature radiation pryometer

Secondary air temperature

Intermediate chain calcining zone temperature measured at the hot end of the
chain system

Control of the long dry kiln is similar to other kiln processes, It is usual to operate long
dry kilns in a fixed feed/kiln speed ratio. If, for example, the burning zone of the kiln
was cooling one possible action would be to slow down the kiln temporarily whilst
leaving the feed at normal levels. Although this would cure the burning zone problem
and kiln speed could be increased, a deeper bed of material will have been formed at
the kiln cold end. As this passes through the kiln it will suffer from reduced heat
transfer and will be under-prepared as it enters the burning zone. A kiln flush may
occur.

The control of burning zone temperature has traditionally used a combination of the kiln
burner’s eye, kiln drive power and possibly a burning zone pyrometer. Kiln exit N“x has
enabled the development of expert control systems, such as LINKrnan and this is more
fully explained in the paper on that subject.

Kiln exit oxygen should normally be maintained at the lowest possible level without
generation of carbon monoxide or excessive sulphate cycles causing kiln rings 1-2%
oxygen is often achieved in practice. Higher levels of oxygen have the effect of
decreasing flame temperature and thus heat transfer in the burning zone. This will

21
increase kiln exit temperature and when combined with the larger quantities of kiln exit
gas will give a significant increase in heat lost in kiln exit gases.

Chain gas temperature is a further control measure and maintenance of a given chain
gas temperature will assist in ensuring even feed preparation to the burning zone.

5. PROCESS OPTIMISATION

5.1 General

As detailed in the “Output Targets” paper the maximum output obtainable from a long
dry kiln is determined by either:-

a) Heat transfer - excessive kiln chain or exit temperatures.

b) Gas velocity - excessive dust loss ex-kiln.

Provided no other limits are dictated by ancillary plant such as coal mill, raw meal
availability etc.

The output limit for heat transfer has been defined in terms of the kiln internal surface
area, inside brickwork and without heat exchangers. The best output achieved on long
dry kilns has been found to be 0.092 stpd/ft (0.9 t/d/m2). The limit for gas velocity has
been established as a function of kiln cross sectional area at the kiln exit again inside
brickwork and without heat exchangers. The best obtained is 9.2 stpd/ft (90 t/d/m2).
The lower of the two outputs established by this calculation would establish a target
output for a given kiln size.

It will generally be the case that normal long dry kilns dust loss is the limit to kiln
output. On normal length/diameter kilns 35:1) the level of gas temperature into the
chain has not yet proved to give very high chain wear rates.

Dust loss from long dry kilns is at a higher level than for wet kilns, losses of 20-50% on
feed not being uncommon. Unless clinker alkali constraints dictate that some fine, high
alkali fraction dust must be dumped, both cyclone and dust filter dusts are returned to
the kiln back end. Measurements of dust loss on ignition indicate that the vast majority
of this is generated in the chain zone.

As kiln production rises towards this target, increased heat transfer is required from gas
to material. For a given kiln, this increase can only be achieved by higher gas
temperatures and this must mean increased chain gas and kiln gas exit temperatures and
fuel consumption for the kiln will worsen. The rate at which this will occur is shown
schematically in Figure 12 and will also depend on the quantity and quality of kiln
internals.

22
1400

1300

1200

1100

1000

900

800

700

I I I I
85 90 95 100

YO of Max Production

FIG.12 KILN EXIT VS KILN OUTPUT

Additionally, as output rises, the fuel rate to the kiln will need to be increased and
additional combustion will be generated. Higher volumes of combustion products
increase gas velocity in the kiln and dust loss will increase.

In order to optimise kiln production the following are important areas.

5. 1s1 Cooler Efficiency

The highest possible grate cooler efficiency must be obtained by operating the cooler
under “deep bed” conditions, balancing the air flow between the cooler compartments
and avoiding unstable cooler control. More heat will then be recuperated in the
secondary air and less fuel input will be needed in the kiln. As a result gas velocity
through the kiln will be lower and dust loss lower. Additionally, the higher secondary
air temperature will increase flame temperature, improving heat transfer in the kiln.

Examrde

10% increase in cooler efficiency is equivalent to approximately 0.125 mBTU/t

(33 KCal/Kg) fuel saving, a reduction in kiln exit velocity of 2.5% and an approximate
reduction of 10% in dust loss or 2.5% increased production.

5.1.2 Kiln Seals

Both kiln seals, at hot and cold ends, must be maintained in good condition. Inleaking
air through the hot end seal effectively replaces secondary air from the cooler!
increasing kiln fuel consumption and dust loss.

Examrde

Decreasing inleaking air through the hood and hot end seal from 10% to 5% will reduce
fuel consumption by approximately 0.04 mBTU/t (11 KCal/Kg) and reduce kiln exit
velocity by 0.9% and an approximate reduction of 4% in dust loss or 0.9% increased
production.

5.1.3 Flame

Optimizing the kiln flame by ensuring the correct pulverised fuel residue, primary air
quantity and velocity, will improve heat transfer in the burning zone. Kiln exit
temperature will be reduced, as will fuel consumption.

5.1.4 Excess Air

Kiln exit oxygen must be maintained at the minimum level which can be achieved
without carbon monoxide being present at significant levels. Kiln flame temperature

24
 will be optimised, improving heat transfer and the volume of the kiln exhaust gases
reduced.

Exarnde

Reducing average kiln exit oxygen when possible from 2.5% to 1.5% will reduce fuel
consumption by approximately 0.075 mBTU/t (20 KCal/Kg) and reduce kiln exit velocity
by 10% and dust losses by 35 or increase production by 10%.

The reverse of excess air i.e. reducing conditions must equally be avoided as this lowers
flame temperature and effectively lengthens the flame. Kiln exit temperature will be
increased and unburnt carbon monoxide also represents unburnt fuel.

5.1.5 Kiln Internals

The principal area for tuning of kiln internals will be the amount, and distribution of
chains in the kiln. Although the BCI system initially uses 2.5t of chain/t/h of
production, greater weights of chain can undoubtedly be installed to give additional
benefit if kiln mechanical condition permits. Studies in wet kilns indicated that
lengthening the chain zone towards the hotter end of the kiln gave higher production.
However, the extension of the chains will be limited by the temperature of the kiln
gases. As the chain wear, due to corrosion varies with kiln atmosphere as well as
temperature, no rule has been derived to decide the limit of chain zone and
experimentation on plant is necessary to determine this factor. In wet kilns a maximum
amual wear of 30% in any single spiral is considered acceptable and similar targets
should be applied to long dry kilns.

Within the confines of the chain zone determined as above, greater weight of chain can
be located and the density of the chain system (expressed as chain tons/1000 ft3 of kiln
volume) increased. Mechanical aspects of kiln design must be checked before greater
weight of chain is installed. As chain density rises, although a “knock-out” effect may
be collecting dust at the same time, a greater restriction is imposed on gas flow and
dust loss from the kiln may also rise. Again, as the mechanisms of dust generation
within chain systems cannot accurately be predicted, careful observation of dust loss
is essential when optimizing chain systems.

When installing lifters or stirrers in a kiln, similar rules must be observed. Refractory
internals can be used in hotter areas, but the material must still be appropriate for the
conditions in the kiln zone concerned. Badly designed lifters may generate dust; dust
loss must be monitored before and after installation.

In overall terms, optimizing kiln internals is the most effective way of h-nproving kiln
performance, but needs careful observation of the changing kiln conditions at each
modification.

25

——..
ExamDle

Reducing kiln exit temperature by 100”F (55°C) (ignoring water injection) will reduce
fuel consumption by approximately 0.10 mBTU/t (27 KCal/Kg) and reduce kiln exit
velocity by 2.0% and dust losses by 8.0% unless chain/lifter design is unsuitable.
Alternatively 2.0% additional production may be achieved.

5.2 Refractories

Comparing the heat losses from preheater kilns, it is clear that one major difference
is in the shell losses from the kiln and preheater system against the long dry kiln. Use
of insulating refractory has long been a standard in preheater and this use has now
become widespread in long Q kilns. Materials such as Orylex 125, a low density
alumina refractory, can be used for up to 60% of the kiln length from the cold end, but
as their abrasion resistance is poor, use in the chain zone is not possible. Average shell
temperatures in the insulated zone are 300° (140°C) or less and fuel savings of O.1-
0.12 mBTU/t (27-32 Kcal/Kg) are possible.

5.3 Preheater Stage Addition

As an up-rating procedure the addition of a 1 or 2 stage preheater can be an attractive

investment. A typical layout is shown in Figure 13. ,,

The increased output from a given kiln will depend on the original kiln performance, but
it is estimated that at least 12% extra production can be expected from a single stage
addition. Greater gains will occur with kilns which were formerly under maximum
target outputs. A two stage addition would yield at least 25% extra production. Some
shortening of the kiln may also be advised depending on length to diameter ratio.

The savings in fuel consumption are principally in the areas of kiln exit gases and shell
losses. A fuel consumption of 3.5 mBTU/t (930 Kcal/Kg) can be expected for a one
stage system and 3.30 mBTU/t (880 Kcal/Kg) for two stages.

An example of the benefits of using a cyclone as a preheater rather than for de-dusting
can be seen in a test carried out at Sharjah Works. At this plant, a one stage preheater
is installed, but the option exists for direct feed to the kiln. The comparison in Table
5 does not specify effects on fuel consumption and output, but the former is estimated
to have risen by 0.5 mBTU/t (130 Kcal/Kg) with feed direct to kiln.

26
 T

FIGURE 13: ONE STAGE SUSPENSION PFZEHEATER

TABLE 5

sHARIAHWORKS- TEsTs WITH ALTERNATIvE FEED POSITIONS

Kiln feed position To riser To kiln hearth

Air bleed % open 5-15% 75-85%
Preheater exhaust temperature “F 670 670
Oxygen in preheater exhaust % 6.5 12.5
Calculated cyclone exhaust temp “F 840
1290

The air bleed enters at the exit from the cyclone and controls preheater exhaust
temperature.
6. CHEMICAL FACTORS

6.1 Kiln Feed Physical and Chemical Factors

The requirements for a consistent long dry kiln feed both physically and chemically are
similar to other processes. Since the system has a long residence time, but no effective
nodulisation of the feed prior to the burning zone, its sensitivity to chemical change
falls somewhere between wet and preheater kilns. As such a target kiln feed chemical
variation of less than 0.2V0 CaC03 or 1.0?40LSF standard deviation is usual, with higher
variations leading to progressively less stable kiln operation, more loss of burning zone
coating and poorer refractory life.

Kiln feed residue is normally controlled by combinability considerations and the long dry
process does not have any special requirements. However, it is possible that
overgrinding of the raw meal will generate more dust loss from the kiln and should be
avoided.

6.2 Alkalis CYcles

As with all other kiln processes, alkalis contained in the raw feed and fuel are
volatilised in the burning zone of the kiln and are carried in the gas stream towards the
cooler areas of the kiln.

As gas temperatures decrease a proportion of the alkalis condense on colder surfaces,

such as kiln refractory and kiln internals. In extreme cases these deposits can cause
rings in the kiln or severe build up on kiln internals. The remainder either condenses
in the gas stream or on the dust in the stream.

Although some of this alkali is retained within the kiln, much of the finer dust is rich
in alkali and leaves the kiln to be captured in the dust collection system. In areas
where low alkali cement is required the ability to reject alkali by discarding the dust
is useful. However, since kiln dust has been pre-heated and partially decarbonated, the
amount of dust discarded needs to be minimised.

As a high proportion of alkali is normally contained in the finer dust, a system capable
of collecting dust in a manner which separates the coarse and fine fractions is
preferable. Tables 6a and 6b show typical fresh feed, cyclone and bag house dust psd
and chemistry for Tulsa Works. The coarse, low alkali fraction can then be returned to
the kiln, while the fine fractions is discarded. Pre-cycloning or multiple series-fielded
electrostatic precipitators offer this facility. Figures 14a and 14b show two alkali
cycles, with and without discarding the dust collected in a bag filter.

28
 (a)

9
B
o
B
i
:
:

9
CLINKER 100 9
;
a 100 FEED
o ●

a
(b)

KILN

FIG.14 - LONG DRY KILN ALKALI CYCLES

Example (a) With dust discarded

(b) With dust return to feed

AH figures represent mass flow of K 20 in this example.

 TABLE 6a

CHEMICAL ANALYSES OF THE NO. 1 AND 2 KILNS SAMPLES - TULSA WORKS

Samples No. 1 Kiln No. 2 Kiln

Cyclone Bag House Cyclone Bag House
Feed Dust Dust Feed Dust Dust
Sample Code
(T9-No. M94) 752 753 754 760 761 762
Chemical Analysis
%
13.6 14.9 14.3 13.7 15.0 14.4
:;:63 3.2 3.5 3.9 3.2 3.6 4.0
Fe OS 1.2 1.3
Ca6 42.8 4::+ 4::: 4::: 4::: 43.8
LOI 35.6 26.2 29.3 35.6 24.1 28.1
MgO 2.0 2.1 1.4
Mn 03 :::4 0.05 ;:;6 :::4 0.05 0.04
Pd
#05 0.09 0.09 0.10 0.09 0,09 0.11
2 0.13 0.15 0.19 0.14 0.15 0.17

Total 0.40 3.8 3.0 0.46 4.8

K20 0.62 1.3 0.63 ;::
Na90 0.22 0.34 :::2 0.22 :::2 0.44
Total 99.6 99.4 99.7 99.7 99.5 100.1
%. LSF 100.2 97.6 96.6 99.3 98.4 95.3
S/(A+F) 3.09 3.04 2.70 3.11 3.00 2.71
A/F 2.67 2.50 2.79 2.67 2.57 3.08
 TABLE 6b
PARTICLE SIZE DISTRIBUTION OF NO. 1 AND 2 KILNS SAMPLES- TULSA WORKS

Samples No. 1 Kiln No. 2 Kiln

Cyclone Bag House Cyclone Bag House
Feed Dust Dust Feed Dust Dust
Sample Code
(T9-CiO. M94) 752 753 754 760 761 762
Apparent Parti le
Density (kg m5 ) 2710 2780 2770 2710 2800 2770
~ % Finer (pm)
4000 100
2000 99.6
1000 100 99.5 100 100
500 99.98 100 99.5 99.98 99.99 100
250 99.5 99.0 99.4 99.5 99.91 99.99
125 94.5 97.8 99.4 94.8 97.7 99.98

90 87.1 91.6 99.3 88.4 91.6 99.97

82.8 86.7 99.2 84.2 87.1 99.9
:: 78.7 81.4 99.2 80.2 81.7 99.8
45 68.2 66.5 98.6 69.7 65.6 99.5

65
:: 62 u x % ;: %
20 91
:: :: 83 :: :: :2
;; 24 68 22 67
# 11 38 :2 9 38
: 11 3 10 14 3 10
 7. ANCILLARY PLANT

7.1 Kiln Exhaust Dust Collection

Normal temperatures of the exhaust gases from a long dry process kiln would vary from
800- 1300”F (430-700”C). These temperatures exceed those which can be handled by
conventional dust collection equipment. A glass-fibre bag filter is limited to 480”F
(250”C) and an electrostatic precipitator 750”C (400”C). Cooling of the exhaust gases
is either by cold air bleed or, more commonly, by a water spray either located inside
the kiln at the back end, or in a large duct or conditioning tower outside the kiln.

The bag filter systems are usually equipped with pre-cyclones to remove a proportion
of kiln dust prior too the bag filter. A pressure drop of 2-4 wg (50-100 mrnwg) is normal
across the cyclone(s). The removal of a large proportion of dust in the pre-cyclones
reduces overloading of the bag filter and protects against the possibility of “blinding”
the bags causing increased pressure drop.

The bag filters can operate under pressure or suction. Both have some disadvantages.
The bag filter is a large chamber prone to inleaking air when under suction. This will
reduce capacity of the bag filter and main fan unless kept under control. Pressure
filters are not susceptible to inleak, but the absence of a stack can be a problem with
emissions of sulphur dioxide at a low level near urban areas.

Electrostatic precipitators are widely used on kiln exhaust gases outside North America
and have the advantage of lower pressure drops than bag filters. However, they suffer
from a greater sensitivity to process conditions and can be complex to maintain. Again
some gas conditioning is required, but it is normal to operate long dry kiln precipitators
at 540-720”C (280-380°C).

7.2 Raw Material Drving

Adequate heat is available in the kiln exhaust gases to dry up to 12% moisture in the
raw materials and a variety of raw milling systems have been coupled to long dry kilns.
Current raw mill designs usually require that the kiln exhaust gas is cooled before use
in the mill.

Note: Where general data has been supplied, both Imperial and Metric units have been
included. For calculation only, Imperial system units have been used.

32

..—.
.—.— . ._—..
—......—
——
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 6

Suspension Preheaters
(CTC 18)
 SUSPENSION PREHEATER KILN SYSTEMS

CONTENTS

1. INTRODUCTION

2. INITIAL DEVELOPMENT OF PREHEATER KILNS

3. PRINCIPLES OF OPERATION

3.1 General Overview

3.2 Particle Size
3.3 Dispersion
3.4 Heat Exchange
3.5 Separation
3.6 System Design

4. TYPES OF PREHEATER

4.1 Humbolt
4.2 FLS
4.3 Dopol
4.4 Dopo190
4.5 Wedag and Kawasaki
4.6 Buhler-Miag
4.7 Gepol (Krupp-Polysius)
4.9 SKET/ZAB
4.10 Prerov

5. PREHEATER KILN OPERATION

5.1 Available Control Parameters

5.2 Assessment of System Conditions
5.3 Reviewing Preheater Performance

6. OTHER PROCESS CONSIDERATIONS

6.1 Optimum Oxygen Levels

6.2 Poor Heat exchange
6.3 Volatile cycles
6.4 Auxiliary (Backend) Firing
6.5 Cyclone Modernisation

7. HEAT AND MASS BALANCE DATA

8. suMMARY
1. INTRODUCTION

This paper describes in general terms the principles behind the kiln system comprising
of a short kiln connected to a preheater system in which the meal is heated by intimate
contact with the hot gases emerging from the kiln. These are known as suspension
preheater (or SP) kilns. Variations on this type of system are produced by all the major
equipment suppliers and the SP kiln has formed the base from which the latest
generation of kiln systems - precalciner kilns - have been developed. Precalciner kilns
are discussed in a later paper, but many of the operational and design considerations are
common to both types of process and consequently on going development work in
precalciner kiln systems is also having a knock-on effect on the potential design and
optimisation of preheater kilns. SP and Precalciner kiln systems make up the vast
majority of the new kiln systems which have been installed in the last twenty years.

The paper briefly covers the history of the SP system, before discussing in general
terms the principles under which the systems operate, the major design variations that
have been produced by the suppliers and covering a number of significant operational
considerations.

Generally Blue Circle has installed preheater kilns where raw materials have permitted
and information concerning a number of SP kilns in the UK and abroad are included in
this paper.

1
2. INITIAL DEVELOPMENT OF PREHEATER KILNS

It was in 1885 that Frederick Ransome took out a patent in England for the first use of
a rotary kiln to produce cement clinker. This kiln was gas fired as coal dust firing had
not been developed at that time. Later oil firing was used, to be followed lastly by coal
firing. The first rotary kilns were about 1.8 to 2.0 m in diameter and 20 to 25m long
with capacities of 30 to 50 t/d. Interestingly the original RanSome design included the
use of lifter bricks to improve heat transfer and mixing although it was many years
later that lifters were first regularly used.

The rotary kilns were initially used for the Wet process M, this process gave a consistent
blend of raw materials with the available preparation technology. In England the high
natural moisture of the raw materials also encouraged the use of the Wet proces. In
other areas, particularly the USA, rotary kilns were also used with dry powder feed and
a significant number of long Wet and long Dry process kilns are still operational today.
However all long kilns suffer from the relatively poor heat transfer between the
material and gas flows in the kiln. Efforts to improve this initially lead to the
introduction of the Lepol process from 1928 and then in 1934 to the issue of the first
patent concerning the concept of dry raw mix suspension preheater.

This patent was originally submitted to the Patent Office in Prague, Czechoslovakia in
June 1932, by M. Vogel-Jorgensen, an engineer from F.L. Schmith Co. The patent
description contains all the characteristic features of a four stage cyclone based raw
mix suspension preheater utilising excess heat for raw material drying, which is still the
most common preheater in the cement industry. Nevertheless it was about 20 years
before the concept became a reality, when between 1951 and 1953 a rotary kiln at the
cement works of Bombe & Bleckman in Beckum, Germany was fitted with a preheater
system designed by the Humbolt (now KHD) Company. This four stage preheater was
similar in design to those used today and was fitted to a kiln 3.2m in diameter by 40m
long producing 300 t/d. Despite the advantages of the new process - low fuel
consumption and use of the waste gases in raw material drying and grinding - further
development of the ‘new’ process was slow with only nine further preheater plants being
built up to 1959, all by KHD. Part of the reason for the slow spread of SP kilns was
that operational problems were encountered in a number of the early installations due
to the development of internal volatile cycles, leading to blockages and poor run times.
However most of these problems were overcome as an improved understanding of kiln
operation developed, and by 1966 all the leading manufacturers of cement production
equipment included suspension preheater in their range of products.

Blue Circle commissioned its first Humbolt preheater kiln at Plymstock in 1961, with
a second kiln of this type being added in 1966. Further SP kilns are operated by the
group at Hope and Aberthaw in the UK, whilst a number of kilns of this type are
operated by associated companies (present and past) around the world. Details of some
of these are given in Table 1.
 TABLE 1

PRESENT AND PAST B(H PREHEATER KILNS

NORKS KILN KILN TYPE OF WATSON/ BEST AVE. % w/w FUEL PREHEATER OUST
DIMENSIONS PREHEATER WALKER OUTPUT TPD CONSUMPTION EXIT LOSS%
(m) TARGET Kcal/Kg TEMPERATURE
OUTPUT TPD

HOPE 70X4.77 Humbolt 2260 2158 95.5 Best 805 203 9

1 70X4. 77 Humbolt 2260 2190 88.4 Typical 223 8
850

PLYMSTOCK 54.4x3.81 Humimlt 1380 58.3 301 6

1 54.3x3.81 Humbolt 1380 1080 64.1 1000 320 5.5

WAURN PONOS 1 % X 3.96 Humklt 1510 1118 74.1 1072 340 5

PEDRO OE 1 67 X 4.57 Fuller/ 1910 1756 91,9 823 330

SINO Humbolt

BAMBURI 64 X 3.80 Humbolt 1350 935 69.2 906 350 15

; 60 X 4.20 Humba 1t 1600 1363 85.2 937 386

ATOTONILCO 1 53.4X3.51 Humbolt 1260 619 49.2 1028 280 6

2 53.4x3. 51 Humbolt 1260 544 43.2 1254 290 6.5
3 70 x 4.40 Dopol 2030 1092 53.8 Ilea 350 13.5
4 75 x 5.00 Dopol 2540 2046 80.6 895 330 1.5

EL FUERTE 2 50.3x3. 66 Humbolt 1225 643 52.5 1009 340 3

FIANRESA 2 55/60x3. 8 DopQl 1255 023 65.6 941 325

LICHTEN8URG 4 53.7x3. 75 Humbolt 1360 959 70.6 898 299 3.1

5 68.OX4. 80 Humblt 2180 2042 93.7 987 398 2.1

ABERTHAW 6 65 X 4.2 Humbolt 1824 1780 97.6 805

ASHAKA 1&2 70 x 4.0 Gepol 1150 64.0 -

 TABLE 2

TYPICAL SP KILN SIZES

DIAMEIER (i/s) LENGTH DESIGN

(m) (m) omPuT (TPD)

SITE

LORUENS 2.8 36 300

MONCADA 3.05 48 480

EL FUERTE 3.7 50.3 600

PLYMSTOCK 3.8 54 600

NO.1

BAMBURI 3.81 54 750

LICHTENBURG 3.75 53.7 800

NO. 4

PLYMSTOCK 4.0 54.4 800

NO. 2

WAURN PONDS 4.0 58 850

ATOTONILCO 4.4 70 980

ABERTHAW 4.2 64 1150

NO. 5

ABERTHAW 4.2 65 1300 “

NO. 6

HARLEYVILLE 4.6 67.1 1500

LICHTENBURG 4.8 68 1700

NO. 5

HOPE 4.77 70 1850

BECKUM 5.3 80 3000

GORADZE 5.75 92 3500

ALLMENDINGEN 6.0 100 4000

TSUKUMI 5.8 110 4100

Output refersto design. Significantly higher ’best outputs’ have been achievedon
the BCI Kilns

4
Suspension preheater kilns have been built with design production capacities of from as
low as 300 tpd, to as high as 4500tpd. A range of typical kiln sizes, covering BCI and
non BCI units, with the design production capacities are shown in Table 2. The most
common size for this type of kiln is between 1000 and 2000 tonnes per day, and
suspension preheater kilns of this size are still being built. Above about 2000 tpd the
precalciner system is now normally the preferred option.

The most common form of suspension preheater comprises of four stagesi of preheating.
In this form it can have a fuel consumption of as low as 750 kcals/kg, although more
commonly the long term average fuel consumption will fall between 770 and 830
kcals/kg.
FIGURE 1: RATE OF HEATING OF RAM) MEAL PARTICLES

100

80

t- 60
Cr
if size

b
Id
40
!3
t-
2
LIJ
%
LLJ
l-- 20
u
(J
i5
!%
U-I r
I 1
0 I

.05 .1 .15
SECONDS
 3. PRINCIPLES OF OPERATION

3.1 General Oveniew

On a long dry process kiln, powdered raw meal is fed to the kilns in piace of the slurry
used in the wet process. Commonly the long dry kiln uses chains to assist the heat
transfer sometimes in association with other internal fittings, such as crosses or lifting
bricks. All of these fittings are intended to improve heat transfer, and so improve the
rate at which the temperature of the material is raised.

Any solid will heat up when exposed to a higher temperature, but the rate at which the
solid takes up this available heat will be governed by a number of physical properties

a) The “specific heat” of the solid

b) The temperature difference between solid and gas
c) The particle size of the particle to be heated
d) The time the solid is exposed to the temperature
e) The degree of contact between solid and gas

3.2 Particle size

When an individual particle is heated at a particular temperature, the time taken for
that particle to heat up depends on the size of the particle. This is illustrated in
Figure 1. This Figure also shows that in general for the range of particle sizes that
exist in raw meal the time periods involved are low - in the order of 0.1 seconds.
Consequently as long as the system has a suitable residence - say around 1 second - it
should be possible to design a system in which efficient heat transfer can be achieved
between finely ground raw meal and a gas stream.

3.3 Dispersion

Normally the major heat transfer problem within the cement kiln is the degree of
contact between solid and gas, as the individual particles of the raw meal exist in a
slumped bed in the lower portion of the kiln whilst the gas occupies the rest of the
cylinder. Consequently the heat transfer surface of the material bed is very small.
.
As gas to solid heat exchange is difficult to accomplish from just a simple heat soak -
as performed within the cement kiln - a method which produces improved contact
between solid and gas would be beneficial. Obtaining the bmt results would require:

a) The meal to be intimately mixed with the gas

b) The meal to be finely divided.

7
 GAS INLET

U
!$
1-
S
L1.1
n MEAL OUT
z
P

LENGTH LENGTH
PARALLEL FLOW COUNTER CURRENT FLOW

TEMPERATURE PROFILES FOR PARALLEL AND COUNTER FIJOW HEAT EXCHANGE

Significant improvements in heat transfer could therefore be obtained if the meal were dispersed
in the gas flow as this would increase the solid phase surface area by several orders of magnitude.
The obvious solution to this is to throw the material into the gas stre~ distributing the solid as
evenly as possible across the gas stream. In order to achieve this, the distribution and mixing have
to be peflormed in a separate vessel to the cement kiln. As the intention is to intimately mix gas
and soli~ and move the solid through the unit operation on a continuous basis it is preferable for
the vessel to include vertical movement of the solid.

It should also be noted that the cement kiln cannot be completely replaced as, in the finai stages
of burning, the reaction vessel must be capable of intimately mixing the individual particles and
of handling the ‘sticky’ product.

3.4 Heat Exchange

There are several methods whereby the relative flows of two heat exchange mediums - cold raw
meal and hot gases in our case - can be presented to one another. However in general all can be
considered to be variations on two main themes:

a) Co-current or Parallel Heat Transfer

Here the cold raw meal and hot gases are introduced at one end of a system and then
travel through the system in the same direction before being discharged at the other end.
l%is system can be thought of as a tube or duct. In passing through the tube the meal is
carried by the gas strew the gas cools and the meal heats up. With good mixing and
sutllcient reaction time meal and gas will emerge at the discharge end at almost the same
temperature.

b) Counter-current Heat Transfer

In this case the raw meal travels in the opposite direction to the gas stream. In the case
of a vertical duct the cold meal will enter at the top of the duct and fall down through h,
whilst the hot gas will enter at the bottom and be drawn up to the top. The powder is
gradually heated on its way down the duct and under good conditions can achieve a
temperature close to that of the incoming hot gas, whilst the temperature of the gas on
leaving the top of the duct can approximate that of the incoming raw meal.

The principles of these two methods of heat exchange are shown in Figure 2. In co-current flow
a large temperature difference exists between the gas and the powder at the start of the heat
exchanger, which reduces asymptotically through the heat exchanger. This means that the system
is initially much more efficient for each unit of heat exchange area than towards the end of the
exchanger. The counter current system should be superior in that for given inlet temperatures,
and flow rates the meal should emerge at a higher temperature than in co-current flow, but this
is more difficult to achieve in practice, largely because of problems associated with achieving

9
FIGURE 3: TYPICAL GAS AND FEED TEMPERATURE ?ROFi LES FOR
HOPE PREHEATER

1100

1000

900

800

----- ____ —.
700

600

500 i

L
‘u
a
$ 300
u
t-

———- —— __ __ .-— —— ____ _

200

100
5 I
I
I
I
● ❉ I

MEAL

J
suitable dispersion and the required direction of flow for each component over the range of
potential operating conditions.

Consequently, the majority of suspension preheater operate on the co-current flow principle, with
the thermal efficiency of the system being improved by repeating the process a number of times,
as indicated in Figure 3. The overall effect of combining several stages of co-current heat
exchange in series is to produce a final result that strongly resembles a single counter current
stage in terms of heat exchange.

3.5 Separation

At the end of an individual stage of co-current heat exchange the material stream will be
suspended in a gas stream. It is now necessary to separate this combined stream of material and
gas into two separate streams. This is done in a cyclone.

A typical example of a cyclone is shown in Figure 4. Separation is by centrifugal force and as

such the dimensions of the cyclone body, and inlet and outlet ducts are important for maximum
efficiency. If the cyclone had 100°A efficiency then all the meal in the inlet stream would be
removed from the gas stream, however this is not a practical situation although a cyclone with a
very high efficiency could be designed where the dust is a single sized material and actual gas flow
rates are known in advance and would not vary.

Cyclone inefficiencies allow preheated meal to return to colder areas of the system carrying the
extra heat with it. This creates a heat transfer inefficiency and can cause the fhel consumption of
the overall kiln system to increase. However designing for improved cyclone efficiency can
require either:

a) Larger size or greater number of cyclones tier a heat exchange stage,

or

b) Greater pressure drop - and hence an improved fan specification and increased
power consumption.

As, in gener~ the methods available for improving cyclone efficiency are costly in terms of either
capital or operating costs, the final cyclone design requires efficiency to be balanced against cost
and in a single system consisting of a series of stages of heat exchange - each of which will consist
of a vessel or duct for heat exchange and a separation cyclone - the cyclones will normally be
designed to be moderately efficient in the lower stages (about 75 to 80%), and very efficient in
the top stage (93°A to 950A). Efficient cyclones are required for the top stage in order to reduce
the load on ancillary equipment for such operations as gas cleaning, dust handling, and kiln feed.

11
 Cleaned Alr tiiht

Annular Eddy
.

Vortox Finder

—oc—

JC “ arbtrory,
usualfy Oc/4

- Jc —
FIGURE 4
oust out
CYCLONE C OMFO- I
It should be noted that since the mid ’80s equipment suppliers have put considerable
effort into the development of improved cyclones: better efficiency with lower pressure
drop, and these are discussed in Section 3.6.7.

3.6 Overall General Svstem Design

3.6.1 General Ovemiew

In a typical full suspension preheater system (as shown in Figure 5), the hot combustion
gases are drawn by the waste gas fan through a series of stages of heat transfer, with
each stage consisting of a vessel or duct for heat transfer followed by a separation
cyclone. The waste gas fan is sited after all of the heat exchange stages, which are
arranged vertically. The hot gases from the kiln enter the lowest stage and then are
drawn upwards through the series of heat transfer vessek, each of which is known as
a riser duct. The meal is introduced close to the top of the system into the highest
riser duct, which connects the topmost cyclones to the next highest cyclone. The gas
flow carries the meal up to the top cyclones allowing heat transfer to take place.

The top cyclones separate the meal and the gas streams, allowing the gases to leave the
preheater system and pass the meai downwards into the next highest riser duct where
the meal is again picked up by a gas stream. This pattern is repeated down through the
preheater until the meal is transferred from the lowest suspension preheater cyclone
directly into the kiln.

In each stage of preheating the meal is introduced near to the bottom of a vertical gas
riser duct, at which point a large temperature difference exists between the meal and
the gas. The meal is carried up the duct by the gas and by the time it enters the next
cyclone, meal and gas have almost reached thermal equilibrium. This rapid heat
exchange occurs because of the large temperature driving force and because of intimate
mixing between gas and individual particl~ which have a large surface area - typically
400 to 500 m2/kg - for heat transfer.

Very little heat exchange occurs within the cyclones, which are primarily a means for
separating meal and gas flows. The gas leaves via the top of the cyclone whilst the
meal is falls down the cyclone wall and out through a material feed pipe in the base of
the unit. This pipe drops as near vertically as possible to the next stage in the process:
a further riser duct or to the kiln. The material feed pipe should be equipped with a
flap valve to prevent material being drawn from the lower riser duct up to the base of
the cyclone by the pressure differential between these two areas (this is known as gas
short-circuiting) as this reduces the overall thermal efficiency of the system.

The upper preheater stages simply preheat the meal, but within the lowest two stages
decarbonation begins to occur. As decarbonation requires a significant heat input it is
obviously beneficial to obtain as high a degree of decarbonation as possible, and in
general suspension preheater kiln systems manage to obtain about 25 to 30%
decarbonation before the meal enters the kiln.

13
 GAS AND RAW MEAL FLOWS IN THE PREHEATER

33o b 360°L
I
—--- —+
,I 1 I I .0)7

STAGE I
Wlil’ ‘~-RA..EALNLET

60040
Xo’c

STAGE III

----- +’

rbr
J
/
\
,\
1’ I STAG E IV
\J
1

--–+ GAS FLOW

v
~ MEAL FLOW

IOK) +0 llotl~ ‘-l-T //

14

v I/
 Each preheating stage consists of a heat exchange unit and a gas/material separation
unit. Most suppliers number the stages from the top down - BCI normally use this
convention, and it is used throughout this paper - however some suppliers, including
Polysius, number the stages from the lowest stage upwards.

Gases leaving the top of a preheater have a significant heat and dust content. The heat
is normally utilised for drying in raw milling, so tying raw mill operation in to that of
the kiln. The waste gases from the kiln are normally split after the kiln induced draught
fan to go either to the raw mill or to the gas cleaning equipment. Exhaust gases from
the raw mill circuit will rejoin the kiln gases before dedusting. The exhaust gases from
the kiln are too hot for most dedusting equipment and so those gases not being used in
the raw mill system will normally pass through a conditioning tower, where water sprays
are used to control the temperature to the required levei. Electrostatic precipitators
or bag filters have been used for de-dusting, with EP’s being the more common
application at present. Nevertheless as environmental constraints become tighter the
use of bag filters is likely to increase. In some cases where water is in short supply, hot
EP’s have been utilised, but prediction of dust collection efficiencies is difficult in these
units.

On some older SP kiln systems - including Hope and Plymstock - a conditioning tower
was not included in the original design. On these sites temperatures are controlled by
spraying water into the riser duct between the two top preheater stages. Whilst this
offered a capital cost saving it down grades the preheater systems in practical terms
from four to three and a bit stages.

3.6.2 Kiln Size

As a significant amount of heat exchange is performed in the preheater the kiln unit can
be much smaller than is required for the long wet or dry processes. Commonly the L/D
ratio for a preheater kiln is about 16:1, where the diameter refers to that inside the
lining. This value compares to an L/D ratio of over 30:1 on a Long Dry kiln.

3.6.3 Kiln Hearth

The kiln hearth, where the raw meal from the lowest preheating stage enters the kiln,
is a critical area where careful design is important. The meal at this point is partially
“’ decarbonated and so high recirculation of material to the preheater will have a
significant thermal penalty. Further this transition point from almost horizontal kiln
to vertical duct will have a smailer cross-sectional area than either the ductwork before
it or the kiln after it (following the material flow). Consequently the gas velocity will
be relatively high so increasing the potential for material pick-up. In severe cases the
material pick-up at the hearth can generate excessive pressure loss in the riser duct and
become a limiting factor to kiln output.

The hearth must be designed therefore to keep gas velocities as low as possible in order
to prevent entrainment of material, without being so iow as to permit material to short

15
 ..

I l%+ // I I ll/XE.EA,ER-
11
:R

I “1
I
MILL l._. -.. ~

L
\

FIGURE 6

G KILN AND RAW MILL CONFIGURATION

circuit the final preheat stage and drop straight into the kiln. The velocity is critical, and it is
known that where this velocity is high - above about 30m/sec - kiln operation will suffer.

Even for a well designed hearth, volatile cycles can cause build-up around the hearth, so
constricting the area and increasing velocities above the critical level. These cycles are discussed
in Section 6.3.

3.6.4 Preheater streams

Normally a suspension preheater kiln will have a single preheater stream attached to it. For larger
systems of over 2500 tpd some suppliers may provide two streams, although there are isolated
cases of twin stream preheater being attached to smaller units. For outputs of 3500 tpd and over,
a twin stream preheater will always be provided (this would also normally be a calciner kiln now).

3.6.5 Number of Preheater Stages

Initially the most common system incorporated four stages of preheating; a typical system has
previously been shown in Figure 5, which also includes representative temperatures for meal and
gas after each preheating stage and an indication of the directions of flow of gas and material.
This shows that the gases still have a significant heat content on leaving the preheater tower.
Further stages of preheating would potentially make use of some of this heat giving extra fiel
savings in clinker production, but this was not initially pursued for three significant reasons:

a) the heat associated with the exhaust gases from the preheater are normally utilised within
the raw milling system to dry the raw materials. This means that the raw milling system
becomes integrated with the kiln system as shown in Figure 6. Typically the preheater
exhaust gases from a four stage preheater have contained sufficient heat to dry raw
materials with moisture contents of up to 100/o. This will be discussed fin-ther in the paper
on Raw Milling.

b) The original cyclone designs that were developed, utilising a four stage system, typically
operated with a pressure drop of 750 to 900mm WG. This was close to the design
limitations for large scale dirty (dust laden) fm operation at that time, and also any further
increase in pressure drop would have a significant cost penalty associated with increased
power consumption which would have to be off-set against potential fiel savings.

c) The extra height of tower required for further preheating stages would increase the capital
costs.

More recently the equipment suppliers have put considerable effort into improving cyclone design.
This effort has produced more compact designs and @eatly reduced the pressure drop developed
by a system of suspension preheater, as indicated in Table 3.

17
 TABLE 3

TYPICAL PREHEATER PRESSURE DROPS -

4 STAGES

OLD STYLE MODERN

CYCLONES CYCLONES

?X stage 1 860 500

3X stage 2 540 340

5X stage 3 370 230

ex stage 4 220 130

hearth 50 40

Pressure drop data in mm WG

Both sets of data represent kilns operating at or slightly above design

throughput.

Modern systems can be designed with pressure drop of down to 300mmWG but
400 to 500 mmWG is more common.

In recent years with the advent of precalcining vessels, larger output systems, and lower
pressure drop preheater designs, five stage preheater have become more common and
where appropriate six stage preheater have occasionally, been supplied, although to date
the latter have only been installed on precaiciner systems. Where raw material
moistures increase for a limited part of the year, it is possible to design the preheater
for the normal material moisture level feeding the raw materials to the riser duct
between the top two stages, but when moistures rise feed the raw materials into the
preheater one stage further down. The riser to the top stage is then basically treated
as a transport duct, whilst the top stage cyclones are still used for preliminary gas
cleaning. A higher gas exit temperature results and this can be utilised to supply extra
dxying capacity in the raw mill.

In general terms, addition of a fifth stage to a reasonably efficient four stage system
should save about 25 kcals/kg when considering the kiln and preheater system in
isolation. The equivalent figures for mpving from five to’ six stages is 15 to 20 kcals/kg,
whilst reducing from four to three stages would cost approximately 35 kcals/kg.

18
3.6.6 System Pressure Loss

The preheater (waste gas) fan draws gas through the preheater overcoming the flow
resistance of the system. This is generated by the volume of gases being drawn through
the preheater, the mass of material being carried by the gases at each stage, and as a
consequence of the original design (sizing) of the individual ducts and cyclones. The
resistance of the system is an important consideration and on many plants becomes the
limiting factor on kiln output.

The preheater must be designed with gas ducts of a suitably sized cross-sectional area
to ensure that as the gas travels through the duct it will have sufficient veiocity to
transport the meal upwards. If this does not happen some of the meal will drop
downwards into the cyclone below (known as meal shortcircuiting), without being fully
preheated. Obviously this situation will be most critical during startup conditions when
the gas flows are low. This could easily be covered by sizing the ducts for these low
flow conditions, however the ducts must not be so small that high velocities are
generated under normal operating conditions as this would generate a high pressure
drop.

Raw meai contains particles of a range of sizes. These do fall within a relatively
narrow band, however it is difficult to define the precise velocity at which a particle
will ‘fly’ with the gas stream. This is partly because each case is a non-ideal situation
which is also affected by the downward momentum of the material stream, the
dispersion of the material into the riser duct, and potential interference from other
particles and from gas eddies. However it has been shown that nearly all meal particles
will ‘fly’ at a gas velocity of 24m/s. Most of the modern preheater designs are based
on the considerably lower riser gas velocities of 13 to 15 m/s. Whilst this normally
gives effective lift, it does require both good material dispersion into the duct and
smooth gas profiling to ensure trouble-free operation.

Cyclone design is discussed in the next Section (3.6.7), and this also has a significant
effect on the overall system pressure drop. Many of the design considerations that
reduce pressure drop within a cyclone will also reduce the separation efficiency, which
will in turn lead to increased recirculation of raw meal to the lower temperature
regions of the preheater. This will have a thermal penalty for the overall system and
may increase the pressure drop in the riser duct above the cyclone due to the increased
material loading at that point. Cyclone design therefore remains a compromise between
cyclone collection efficiency and system pressure loss, both of which will effect fan
power requirements. Nevertheless measurements have shown that most of the pressure
drop through the preheater occurs across the cyclones and it is in this area that the
greatest potential for reducing pressure drop exists, especially on older systems.

As pressure drop is related to gas volume, one ‘obvious’ way to minimise the pressure
drop is to reduce the excess air levels. This is valid but only within certain constraints,
which are further discussed in Section 6.1. Certainly every effort must be made to
minimise the inleak of air across the preheater system and the kiln seals.

19
 z
m --
1- -
I
r-
I
--
I
I-_

1,
I
,
J:
@ --
!
I
-— .,— _
I ‘1
/ /
0
■
r- ‘1
-->~----____
,m
I
+.

\
L+ .- —___
I ‘1
? 1.-
20
 3.6.7 Cyclone design

Cyclone design changes somewhat across the height of the preheater tower. The top
stage has to have a high separation efficiency in order to minimise the consequences of
dust loss from the preheater on the other process operations:such as gas cleaning, raw
milling, kiln feed system. Separation efficiencies in the top stage cyclones are normally
at least 94% and preferably higher. However in general high cyclone efficiency will
equate with high pressure drop and consequently about 30% of the total system pressure
drop will occur across the top stage cyclones. The top stage will also normally consist
of at least two cyclones, to which the gas flow has to be split.

For the lower cyclones a reduced efficiency can be accepted, although any reduction
in efficiency will have a thermal penalty associated with it, as reduced separation
efficiency will result in greater quantities of heated meal being carried up the system
to cooler areas so tending to eventually increase the waste gas temperature. The
further down the system we progress, the harsher the environment in which to maintain
any internais, and so cyclone design becomes a three way compromise between
separation efficiency, pressure drop that will develop and potential maintenance
requirement. The variation in cyclone shape down the preheater, as a consequence of
the above considerations, are illustrated as an example from one supplier in Figure 7.
Equipment suppliers have put considerable effort in recent years into achieving
improved separation efficiencies with minimal pressure drop. Consequently in assessing
cyclones, conventional (old style) and modern generation units must be considered.

Typical cyclone efficiencies for cyclones of each generation are set out in Table 4. As
can be seen for both generations the efficiency tends to increase progressively up the
tower.

TABLE 4

TYPICAL CYCLONE EFFICIENCIES

OLD STYLE MODERN MODERN MODERN

FLS KHD POLYSIUS

TOP

STAGE 1 95 93 96 96.5
.
STAGE 2 85 87 90 85

STAGE 3 85 85 82 to 85 85

STAGE 4 70 83 75 to 80 85

STAGE 5 NIA 78 70 to 75 80

Data suDDlied by Manufacturer

21
FIGURE 8 MATERIAL FLOW THROUGH PREHEATER

8tph

?
135tph
I I
164
>

29

r,,
191
4 156
<

Ii \

Y
2

35

162 -r- [
233
>

71

/
I

Q
3C30Zl

E KILN
127
22
The effect of cyclone efficiency on the material streams in the preheater is shown in
Figure 8.

Traditionally cyclone efficiency is increased by application of any of the following:

a) Increasing the inlet velocity (refractory deflectors)

b) Gas flow smoothing at the inlet

c) Increasing the separation velocity (increase dip tube diameter)

d) Increasing the outlet velocity (cone the dip tube)

e) Increasing the dip tube (vortex finder) length (providing it is not deeper than the
existing inlet depth)

f) Installation of flap valve on the material down pipe from the cyclone: if this
doesn’t exist

g) Installation of guide vanes

h) Installation of a sloping ‘goat track’ inlet

i) Increasing the inlet ‘wrap around’ from 180 to 270 degrees

The new generation of low pressure drop cyclones generally incorporate many of these
features, although actual application varies somewhat from one supplier to another.
Examples of traditional and modern generation cyclones are shown in Figures 9 and 10.
Most suppliers have developed ‘improved’ cyclones in a similar form, although Polysius
have produced a more novel concept which includes a sloped gas inlet and roof, and very
small vortex finders (Table 5).

23
 PG”Z
mla
3
Ill
z
. . . -----
a
. . . . ..J -------- . .. . . . . . . . . . . . . . . . . . . . . . . . . .-
1
[
-.....:
,*
G
9
, c
03
.-
m:
al tn
... . ..-.
:*
r
a
s

. . . .. .. .. .. .. .. .. . . . ----- ,. . ----- . . - $1 ----

9
. ...4
>
24
 TABLE 5

TYPICAL VORTEX LENGTHS

FL!WFULLER KHD POLYSIUS

STAGE 1 0.61 1.18 0.75

STAGE 2 0.42 0.45 0.20

STAGE 3 0.43 0.45 0.20

STAGE 4 0.43 0.45 0.20

STAGE 5 0.38 0.45 0.17

Vortex length as a proportion of inlet depth

The length of dip tubes varies across the height of the tower and from supplier to supplier.
Typical values are given in Figure 11. Older systems have sometimes had the lower stage dip
tubes removed due to mechanical unreliability. This is a retrograde step which is discussed firther
in Section 6.2.

3.6.8 Kiln Output/Pressure Drop Relationships

The pressure drop developed through the preheater system is related to the gas flow and hence
to the output, as a relationship can also be developed between output and exhaust gas volume.
The relationship between output and pressure drop will be a power law expression of the form:

output (tonnes/hr) cc Pressure drop’

with the value of the power being site, and even stream, specific. Examples of this relationship
from a number of BCI sites are shown in Figure 12, with values of the exponent varying between
1.45 and 1.8. Each suspension preheater works should collate operational data from their site to
be aware of the relationship on that particular stream but should ensure that the kiln is operating
at similar oxygen conditions at each feed rate as variations in oxygen level will have a significant
effect on gas volume. Knowledge of the relationship is needed in any up-rating exercises, whilst
deviations from an established relationship can give clues to why a system is operating differently
on occasions. Where a relationship has not been established a value of 1.8 should be used.

25
m
Lu
.
 FIGURE 12:

OUTPUT V PRESSURE DROP

90(

80

70 I I

I
60 I ABER17dAW

I !
/

t
1
40(
PLYMSTOCK
(HUMBOLDT)
LIP- T1”L5
I
3a I

20(

,,

10(
i 30 — 4 50

OUTPUT TONNES/HR

27
 FIGURE 13: HUMBOLDT PREHEATER

n’ /
AU AI LIHIII I 1 ~ \
1
STACK
I PREHEATER FAN

[
PREHEATER
STAGE —
1i i\

/
111
FEED

I
/ I I
Ui(

~!
Y,
STACK
DAMPER C
4

F
++ \

\
\

.-—----
\

n--l !1 A
\
\
I 5TAGE IV

~ NT--Y
m
CQ

Q , –VACUUM
- ‘DUST
4. TYPES OF PREHEATER

As indicated in Section 2, all the major cement plant suppliers offer suspension preheater systems,
many of which differ slightly in design. The major variations are described in the following
sections, where it should be noted that the general layouts are being considered rather than the
more detailed design aspects of individual components, such as cyclones.

4.1 Humbolt (KHD~

This is the original type of systerq and arguably the simplest and most widely applied. The basic
layout is illustrated in Figure 13, with twin top stage cyclones and a single cyclone for each stage
below this. The Hurnbolt system was also made under licence by Fuller horn the USA and by IHI
from Japan. Depending on the required output a single line or double line (two strings operating
in parallel) preheater system could be installed.

In general double line systems were considered for outputs of above 2500 tonnes per day,
although installations on smaller kiln systems are known.

4.2 F.L. Smidth

F.L. Smidth were comparatively late in joining in the development of the suspension preheater
system and at first built kiln preheater systems with only two cyclone stages. The first four stage
preheater was ordered from FLS in 1966. Their present preheater layout is very similar to that
to KHD, and is, therefore, not illustrated (it is to be noted, however, that the cyclone geometry
as used by KHD and FLS varies between these companies).

4.3 DoDol (Polysius)

The Polysius Dopol preheater is significantly difllerent from other designs. The name derives from
the words ‘Doppelstrom’ (meaning dual current) and Polysius.

The preheater as shown in Figure 14 is a system of double cyclones but with one common
cyclone(stage 3 in a four stage system) which is identified as a ‘turbulence shaft’. At high outputs
the top stage consists of four cyclones in parallel whilst at lower outputs two are utilised.
Splitting the gas stream allows the use of smaller cyclones for the same gas volume whilst
maintaining a high degree of separation. The turbulence shaft was originally designed to act as
a counter-current exchanger and as an intensive mixing stage for the raw meal in order to prevent
any irregular preheating patterns between the two streams. The system was designed to achieve
very large (for the time) kiln capacities without the need for fimdamentai change in the design of
the system and without having to resort to more than one preheater line.

29
FIGURE 14: 00P0L PREHEATER

B
lrt

I !

11

III

A ROTARY KILN

1
Iv )) B RAW, MEAL FEED
C EXIT GAS FOR
UTILIZATION
AND DEDUSTING

30
4.4 Do~oi 90 (PolYsius)

More recently Polysius have introduced the Dopo190 preheater. Here the configuration,
as shown in Figure 15, is similar that of the basic Hurnbolt preheater.

4.5 Weda~ and Kawasaki

This preheater has been supplied by Hurnbolt-Wedag and under licence by Kawasaki
Heavy Industries of Japan. The layout is similar to that of the Dopol preheater of
Polysius but with vortex chambers inserted between the sets of double cyclones as
shown by Figure 16. A series of constrictions are added in the riser duct from the kiln
to the Stage IV cyclone with the purpose of creating a turbulent gas stream and
preventing build-up at this position.

4.6 Buhler Miag

The Miag preheater, Figure 17, also consists of double cyclone working in parallel.
Stage 4 however consists of a large conical shaped shaft acting as a counter-current
stage.

Material from Stage II is fed to the top of the shaft and hence to the Stage 111cyclones
from whence it is fed to the centre of the shaft. Some of the material returns to Stage
111and hence a high recirculation is set up. When the capacity of the gas to carry the
raw meal is exceeded it drops through the chute and into the kiln. The large cross
section of the shaft and the recirculation tend to minimise alkali build-up on the wails.
This type of preheater has been made under Iicence in the USA by Allis Chalmers Co
and in Spain by ATEINSA Company of Madrid. Fuel consumption is quoted at 750 to 800
kcals/kg, but BCI have no practical experience with this type of preheater.

4.7 ~DOl (KruDD- Polysius)

The suspension preheater described so far basically rely on stepwise concurrent flow
with perhaps limited counter current operation. This preheater, developed by Krupp,
is an almost complete counter current heat exchange unit. The feed enters the system
into a riser duct near the top of the tower and flows in parallel with the gas stream into
the top stage which has double cyclones for de-dusting.

As can be seen in Figure 18, the tower below the cyclones is divided into four chambers.
The meal drops down the tower in counterflow to the gas stream and at each
constriction a conical structure distributes the descending material stream. When
passing the annular throats the accelerated gas keeps the feed particles momentarily
in suspension until the point is reached where the lifting capacity of the gas is exceeded
and material drops into the chamber below. The same process then takes place in each
chamber until the material enters the kiln. The preheater can therefore be considered
to consist of five stages; one co-current and four counter current.

31
..

FIGURE 15

POL90 SIX STAGE PREHEATER ON A CALCTNER KILN

32
 /
\ \\ I I r
\
d ~1

+
YY
Iv
//
\ 1/
‘)-/ ‘
//
\\ (/J’ ) // I

33

FIGURE 16:
FIGURE In MIAG PREHEATER

-1
II “

! \

ION PREHEATER SHAFT

1. I I 1 1

-%t’-l\ I

KILN TRANSITION

34
 FIGURE 18:

GEPOL PREHEATER

CYCLONE STAGE

I/l Fw

CHAMBER I
w
JIII!
AI
1

7T
RAW MEAL
INLET

CHAMBERII

CHAMBER III

CHAMBER IV

OUTLET To KILN

35
 -‘--
i’. I II
.7.3
kgxm~ ~lng INOICATU
urn
?oua cYcLollca

ra9c 2

?WNOAT$ON

KILN ?[18 CMUTC

FIGURE 19

MODERN GEPOL INSTALLATION

36
Despite the counter flow principle the fuel consumption of the Gepol is generally higher
(at 800 to 900 kcals/kg) than that of the conventional preheater, although on a total
system energy basis the lower thermal efficiency is compensated by the lower pressure
drop across the preheater - especially when considering the older cyclone designs -and
hence lower power requirements for the preheater fan.

In recent years the topmost chamber has been replaced by a more conventional co-
current arrangement in order to improve heat recuperation. This layout is shown in
Figure 19.

One advantage of this preheater is in capital cost. It is a self supporting structure

which does not need to be housed in a building, giving reduced construction time. This
is particularly useful in the conversion of old wet and long dry kilns to this process as
the minimum of down time is achieved by erecting the tower over the kiln whilst it is
still in operation. Only a short down time is then necessary to shorten the kiln and
connect it to the preheater. It is also claimed to less susceptible to problems associated
with volatile recirculation.

4.8 Sket/Zab meheater

This is another form of shaft preheater, and has been designed by the SKET/ZAB
Company of Dessau in the former Eastern Germany. The shaft sections are oval in
cross-section and are off-set to encourage material and gas recirculation. The system
pressure drop is low, typically being 250 to 400rnm WG. The system is shown in
Figure 20.

4.9 Prerov Preheater

Produced by the Prerov Engineering Works of Prerov, Czechoslovakia, this design

consists of a vertical shaft with counter-current heat exchange, a lower cyclone stage
with two cyclones and an upper multi-cyclone stage for dust collection (Figure 21).
False air in-leaks are minimal due to the joint-free construction, and moderate system
pressure drop of 350mm WG.Gas exit temperature is quoted to be about 360 C with a
dust concentration of 30 to 50 g/Nm3. Fuel consumption is typically indicated to be
between 750 and 800 kcals/kg.

37
 FIGURE 20 GENERAL ARRANGEMENT OF SKET PREHEATER

I
I
I
I /
STAGE I
/
/

STAGE II

STAGE III

STAGE IV

STAGE v

.
 FIGURE 21: PREROV PREHEATER

preheater shtit CIOSue

Ob- Zyklonstufe — En~tiubungssyklone

UPper ~clone stage — dust collection cyclones

Untem Zyklonstufe
Lower ~cione stage

pendelklappe
flap Wives

RoheM-Streukegel
‘w mti dispersion cone

11, ~ ‘“” ‘ 10

M preheater exit g= duct

‘ ~: Raw mix feed pipe

Counter current preheater shaft

Untetieif des WT-Schachss

km? pert of preheater shaft

Ofen-Abgasieitung
~ exit gas duct

Ofen.Edatik~er KiIn feed compment

WTGabl&se Preheater fan

Drehofen Roq &

Druckgefi&Rohmehlpuw
Raw mix pressure tank conveyor

39
 Figure 22

PREHEATER INSTRUMENTATION

o- 1200”C
AIR FLOW ------w::::::?’
I I I
Gaa volume reoorder o-loo%
STAGE 1 Prehaatar 02 remrder o-lo%
CO alarm o-1%
NOX remrdar 0 to 2000 ppm
Fan UL Indloator O-1000 mmWG

Temperature Indicator O-120fYC - alarm 54(YC

Suction Indioator O-1000mm WG

Suction alarm only 10 mm

Temperature indkabr O-120WC

Suction Indloator O-500 mmWG
t

STAGE 3
Suction alarm only 10mm
+

Temperature Indloetor O-1200”C

Suction Indloetor O-500 mmWG
!

Temperature remrder 0- 1200”C

STAGE 4

Suction reoordar D-200 mm WG

.. 6,
9 Suction alarm only 10mm

Temperature remrdaf 0- 1200”C alarm 820/360”C

\ ~

4 ~sucti”n’nd’mtor O-100 mmWG

o-1%
o-5%
! NC)X recorder 0-3000 ppm 40
~ ‘lLNHEARTH
 5. PREHEATER KILN 0PEIU4TION

5.1 Available control mrameters.

As in any other cement clinkering process, when operating a suspension preheater kiln,
the kiln supervisor is expected to produce the maximum potential tonnage of good
quality clinker at the lowest practical energy consumption. In addition by operating the
kiln under stable conditions mechanical stresses will be reduced and the brickwork life
will be extended. This in turn will contribute to maximum running hours being achieved
and annual output being maintained at a high level. For a large SP kiln (2000 to 2500
tpd) an annual run factor of 85°L would be targeted, with a slightly higher figure for
smaller kilns especially if used in combination with a planetary cooler rather than a
grate type.

In running the kiln, four variables exist with which the operator can control the internal
conditions; fuel rate, feed rate, kiln speed, and fan setting. Many companies link feed
rate and kiln speed so that only one of these is altered by the operator with the other
adjusting automatically under a predefined ratio, however this loop is not normally used
on BCI works. Ideally the kiln speed and fan setting are quickly brought up to values
close to the normal maximum levels at a reasonable feed rate. As the heat distribution
within the kiln and cooler system is optimised the feed rate is maximised, and then the
kiln can be controlled mainly through the use of variations in fuel rate; however the
practical situation is rarely quite so simple.

5.2 Assessment of Svstem Conditions

In order to monitor conditions in the preheater a series of continuous temperature and

pressure readings must be available to the operator. High pr-ure differential across
a particular stage suggests that build up has developed in the ducts, or that meal is
being re-entrained and carried round and round, whilst loss of suction at a cyclone base,
or rapidly rising material or gas temperatures indicate that a blockage has deveioped
in that cycione. Gas analysis is normally installed at the preheater exit and at the kiln
exit. The gases are normally monitored for 02, CO, and NOX content, whilst in a
limited number of cases SOX is also measured. Normal measurement points are shown
in Figure 22.

There are a number of useful guides potentially available for the kiln operator to use
in estimating the internal condition of the kiln. The items normally used are kiln drive
‘ power, burning zone temperature, and gas NOX analysis, whilst at least one case is
known where secondary air temperature is used as a guide to kiln, internal conditions-
although reliance on this signal would not be recommended by BCI.

5.2.1 Kiln Drive Power

The kiln drive power is monitored by measuring either the power drawn directly or the
amps pulled by the drive. The signal gives early warning of the condition inside the kiln
.

41
 F’IGURE 23: KILN AMPS TRACE
i“
I*O
! ,
1.0
0
em
,1 ,
1000 i, ,0 0
; .
!
i , ,0 .-
!. .
0
pm
I ●
,
&
-

. 00. 0 N
8.
.,.,H-
!0 Do I1OO o
I
!1I Ii pm
Ill
..- ~
1,
,000
I
. . . 0
II 0 0
II
., .,
6
19
t-
1
4
I
m
16 ‘1
400 0
,, ~~
!0,0 ,,00 ,
II
;---
.; 4
.,
i
T
0
\
pm
I I* ~ ~
‘/
!000 ,,0. 0
II
,--” 2
!,
... --
!0.
t-
0
P
.,
I 1’
..-.
!. i
# ’11
,... , , >*a a
“1
,,.;
,,, -1--
12
--
.-..-.
c
‘1
Ill am
i
,,, -
!000 ,,”0
0
1 ~
1
I‘“1
,L’
1
10
!,0
am
I ‘, -1
I ,000 )200 o
,, ,,
..Je
. . .
t. e t-
0
Im,
“,. ,,
,1, !
,! II,,,~~olll~ I /.
11
1oo. ,,., .
iL
9-8
0 -Dn

.4,
“nfiv’FF~OOi 7!o !’”+
,,1’
J-Ill
..,”
,!
b
42
 as a whole unit, an example -from Hope works - being given in Figure 23. The signal
supplies information in three ways:

a) Width of Band

This is an indication the depth of feed and of the evenness of coating within the
kiln. With a thin bed of material and even distribution of coating around the
circumference the power drawn by the kiln over a revolution will be relatively
consistent. The power drawn by the kiln drive will still vary a little over a
revolution but the resulting signal will be a relatively thin band. Where coating
is uneven, the kiln drive will require more power when the thicker sections of
coating are rising and less when they are falling during the normal turning of the
kiln, and hence a wider band will remit in the amps signal. Where a ring exists
feed will dam up behind it and this will also tend to widen the band width of this
signal, although the extent of the widening will depend on the size and position
of the ring. Basically as the material bed depth builds up this is positioned off
the axis of the kiln and a greater eccentric load results.

Consequently, a wide trace on the kiln drive power signal generally

indicates uneven coating, ring formation and/or a greater bed depth of
feed, as long as no mechanical factors are relevant.

b) Change in Band Width

A change in band width indicates a thinning or a thickening of coating or

material bed within the kiln. A change from wide to thin band would
indicate a thinning down of build-up and a loss of damming effect. Where
a step change occurs this would indicate the sudden break away of all or
part of a ring which results in an instantaneous change in the eccentric
load. This suggests that there is likely to be a surge of material coming
down the kiln, and allows the operator to react by increasing the heat
input to the kiln before the resultant cooling becomes obvious at the front
end, so reducing the extent of the disturbance.

Very occasionally a situation develops where kiln amps have ciimbed very
high and suddenly dropped back significantly. The operator must be aware
that this could potentially signify a ‘melt out’ of coating near the front
end. In this situation the kiln must be cooled as quickly as possible and
.
extra air must be put to the front compartments of the cooler, or the
stream of material from the kiln will severely damage the cooler.

c) Trend in direction of the signal

As the material passing through the kiln becomes hotter it becomes

stickier. Because of this, as the kiln turns the material tends to rise
higher up the wails and thicker coating will also form. These factors in
turn give a greater eccentric load and so the power drawn signal tends to
rise. Therefore a rising kiln drive power signal indicates that the kiln is

43
 warming up, and conversely a falling signal suggests that the kiln is
cooling.
Modifications to fuel and feed rate can be made to increase or reduce the
temperatures within the kiln.

It should be noted that this power signal can also be affected by changes in kiln
speed and by major variations in the volatile cycle.

5.2.2 Burning Zone Temperature

A two colour pyrometer can be installed to view a fixed point in the kiln. This will then
give an indication of whether that particular portion of the kiln is getting warmer or
cooler. These signals were frequently used in the late ’70s and early ’80s as an
indication of kiln conditions, but were never considered to be ideal for the following
reasons:

a) Indicated temperature significantly affected by dust cycle from cooler.

b) Indicated signal affected by movement of the burning zone. This

is often significant but can give misleading results.

c) Systems often found to be maintenance intensive.

Consequently, direct measurement of the burning zone temperature was replaced by the
use of the NOX signal (see Section 5.2.3), however this temperature signal is still used
on a number of sites world-wide, and with improved technology and the need to reduce
NOX emissions is beginning to make a comeback as a major control signal.

5.2.3 NOX Measurement

The measurement of the quantities of oxides of nitrogen generated in the kiln flame has
become widely used as a means of assessing the temperatures within the burning zone.
A number of forms of oxide of nitrogen can be formed within the cement kiln, hence
the term NOX to cover all the permutations. However in practical terms over 90% of
, NOX generated in the kiln is made up of NO and this is the component normally
monitored.

Within a flame NOX can be formed by three routes, two of which generate a base level
of NOX which largeiy depends on the type of fuel and firing system, whilst the third is
strongly positively dependant on the temperatures within the flame. Provided the flame
is reasonably consistent, the flame temperate will reflect the temperature of the
materiai in the burning zone and so NOX levels in the gases will be a measure of burning
zone temperature: the hotter the burning zone the higher the NOX level. Obviously NOX
levels cannot be measured at the end of the flame as no sampling system exists, and so
the normal monitoring positions are either the kiln backend -sampled inside the kiln-or

44
 at the preheater exit. On most Blue Circle SP systems NOX leveis are monitored at
both positions.

The kiln backend is the preferred position but this suffers from the harsh local
environment and as such is maintenance intensive. 1ne prenea~er
-I- I --- allaly~la
.--I ...:. pUIIIC IS
--:-. ,- ., men
used as a backup unit.

A good correlation can be obtained between NOX generation and clinker free iime -
Figure 24- although there is no absolute NOX level at which a kiln should be operated.
The optimum value will be site and time specific and so the target needs to be
continuously adjusted slightly in order to get the best performance out of the kiln
system.

BLUE CIRCLE EXPERIMENTAL WORKS

f\ KILN TRIAL, MAY 1981
N‘aitl~
ox Typical NOX /Free Lime Relationship
(lfprn)l/
600 I

I
I
500 t
I
I
400 I
I
I
300 ~
‘ree (%)
‘i me1 10

- 8

1
,,
,!, , I I ,,, ,,, ( #,, , , 1
I -o
o1234!jij7 89 10 11 12 13 14 15 16 17 18 19 20 21
Time(hrs)

FIGURE 24

45
 NOX generation will be greatly reduced by the on-set of reducing conditions, but such
conditions will also cause other operational problems within a suspension kiln due to
increased volatile cycles, and so should be avoided.

5.2.4 Secondary Air Temperature

Some (non-BCI) sites have used secondary air temperattire as a guide to burning zone
temperature. These generally use a thermocouple sited well up in the cooler throat or
hood, and consequently this is affected by heat radiated at the front end of the kiln.
Therefore much of what the thermocouple ‘sees’ reflects movement of the burning zone.

This is not a BCI approved measurement for assessment of burning zone temperature
as it is affected by too many factors. However it can be used ‘intelligently’ by an
operator to the confirm the data supplied by other signals under many conditions.

5.3 Reviewing Preheater Performance

The following sections discuss briefly the means of analysing and interpreting preheater
performance from the physical measurements and simply mathematical simulations that
can be carried out on any plant. In interpreting the reasons for poor performance the
most important aid of all, is comparison data collected during those periods when the
plant is running well. It must be emphasised that the tim~ when the plant is running
well do not offer the engineer a chance to relax, but rather present the opportunity to
fully identify the process conditions that are necessary to optimise plant operation.

5.3.1 Veloci~ Distribution

In order to analysis gas velocities through the preheater it is obviously necessary to

know the gas volumes in each area. These cannot be measured directly but can be
inferred from a combination of measured information and calculated heat and mass
balance data. This information ailows areas of potential low veiocity (where short
circuiting may occur) and high velocity (where there will a pressure drop penalty) to be
identified.

‘ 5.3.2 Pressure Distribution

Although in general this gives similar information to the velocity distribution data, it
has the advantage of allowing continuous measurements to be made. An unusual
pressure profile can be indicative of build-up on the duct walls, meal re-circulation,
poor dispersion, or short circuiting of meal.

46
5.3.3 Gas and Feed Temperature Distribution

Typical temperatures for each stage have previously been shown in Figure 5. For an
efficiently operating system similar temperatures should be obtained for the gas at the
cyclone inlet and outlet, and for the meal in the feed pipe from the cyclone. Significant
differences can be due to:

poor heat transfer in the riser (inefficient meai distribution, or meal inlet too far
up the duct)

gas short circuiting up the feed pipe (damaged or non-existent flap valve)

or

meal dropping down from the inlet above (short circuiting due to inefficient meal
distribution/poorly sited material inlet, or low gas velocities)

A detailed temperature survey across a duct cross-section will also indicate the
efficiency of meal distribution across the duct. Localised high temperatures are
indicative of poor meal distribution into those areas.

5.3.4 Stage by Stage Gas Analysis

Gas analysis surveys across each stage of the preheater for 02, C02 and CO are useful
in order to :

a) Identify areas of inleaking air

b) Confirm where decarbonation is occurring

Where this is performed, analyses should be completed at more than one point on the
cross-sectional area in order to minimise the potential effects of streamlined gas flow.

5.3.5 Stage by Stage Mat~al Analysis

This gives useful information concerning the extent of any volatile cycles. In many
cases variation in LOI levels in the upper preheater stages can be used to give an initial
estimate of cyclone efficiency using mass balance techniques: as where temperatures are
too low for decarbonation, changes in LOI between the feed and product points on an
individual stage will be largely a consequence of cyclone inefficiencies in the stage
below. In some cases where the meal contains significant quantities of low temperature
volatiles this technique cannot be applied usefully, although it may be possible to ‘trace’
another component. Typical material analyses for each stage of a four stage preheater
are set out in Table 6.

47
Fig.25

Kiln Inlet Gas Analysis

C02 at constant 02 vs Raw Meal Decarbonation

100

....... .. .. ...
+ .... ....... ... .... .. ... . ..... .---. --. -..--., ---------------- --------------- --------------- -.
90 I
s --- .-s---- .-. .

o
m-
u
c
80 . . . . . . . . . . . ..-. . . -------------- k
....... ...... .. .. .... .. .-------------- ...... .... . .. .. --------------- ..

.. . . . . . . . . . . . . . .--------------- . . . . . . . . . . . ..- ............... --------------- .. . . . . . . . . . . ..- ..

70 ---------------

cl) 60 --------------- x I.--.-.------.-, ..

n
50
16 17 18 19 20 21 22 23 24 25
C02 at Kiln Inlet

&i
I
Decarbonation at 2.5 % 02
-.x.s:.>>.\Y.,,x. x.:.x.,,.,:.
x*.:.
M*+w*.>,..,.m* ~-.sw xv.~ .—- .,.$+
Decarbonation at 4 % 02
. *i.,x.,. ..: .>..:.:.,>:.,..:.. x.>,M.:.>. S,L.
v
... .. s.. k *.*., vi.: c.. ..... .. . :<,.:...!...
>!.!
I ~
i
s>:.... x.>
!.>.:.!.!.
>.. %
 TABLE 6

MATERIAL ANALYSES IN PREHEATER

EX SAGE EX STAGE EX STAGE EX STAGE

1 2 3 4
Si02 13.4 13.3 13.2 13.9 15.7

f%l~o~ 3.6 3.7 3.8 3.9 4.2

Fe203 1.5 1.5 1.5 1.5 1.7

CaO 43.8 44 44.9 47.9 50.8

LOI 35 34.5 33.2 28.2 22

sasso~ 1.3 1.3 1.2 2.2 33.0

K20 0.41 0.46 0.54 0.86 1.20

Na20 0.11 0.12 0.12 0.15 0.19

c1 0.025 0.056 0.13 0.35 0.49

Carbon 0.48 0.44 0.20 0.05 0.06

5.3.6 Kiln Inlet Gas Analysis

Carbon Dioxide is created in the first place by fuel burn out, with the balance of C02
generation coming from the decarbonation of meal. Meal samples collected at the
bottom of a preheater and analysed for LC)I generally give a good picture of the average
carbonation levels achieved in the preheater, but on occasions can be deceiving. As a
useful check a series of measurements of C02, corrected to constant oxygen, can give
eariy warning of meal short circuiting to the kiln. Figure 25 presents an example of
C02 verses decarbonation level for two oxygen contents.

5.3.7 Time Dependant Process Study Using Data processors

~ In the past, the engineer has had to complete process measurements by eye. It is now
possibly to use small electronic data proc~sors to store process measurements - such
as temperature and pressure - over a significant time period but sample, for instance,
on a one second time base, and then analyse the results away from the measurement
point. This is useful in the identification of cyclic behaviour in the process
measurement.

49
6. OTHER PROCESS CONSIDEIWTIONS

6.1 ODtirmlm Oxmzen Levels

Excess air must be available for combustion. This must be pulled through the kiln by
the waste gas fan, and so the capacity of this fan is a potential constraint on the
system. Originally excess air was maintained to ensure that reducing conditions did not
exist in areas within the kiln where clinker quality could be affected and to ensure that
potentially explosive conditions did not arise in the waste gas cleaning equipment -
commonly electrostatic precipitators. To do this the level of oxygen and carbon
monoxide at the backend of the kiln and at the preheater exit would be monitored and
maintained at minimal levels through controi of the fan setting to vary the oxygen level
(e.g. for increasing CO, raise oxygen level).

It should also be noted that the minimum level of CO is site specific, as the value at
the kiln back end can vary slightly with the type of fuel or performance of the fuel
preparation and firing plant, whilst the presence of organic material in the preheater
feed is likely to create traces of CO at the preheater exit.

A fan can pull a given volume of gases against a particular pressure drop, and it is
known that within any one system the pressure drop will depend on the physical design
of the system, the density of the gas stream and the velocity of the gas stream. ‘The
first of these is largely fixed when the plant is built, the second the operator feels he
has no control other, and so the third was used as the control parameter. In the early
days of preheater operation the obvious preferred method of operation was to minimise
oxygen levels to the point where CO was just present on the basis that this minimised
excess air and therefore gas volume which in turn should minimise the pressure drop
through the preheater tower. At this time preheater kilns were often run with backend
oxygen levels of between 1.0 and 1.5%. However this is not the optimum situation for
the following reasons:

a) Before increases in CO become evident in the backend gas analyses

the proportion of CO in each area of the flame increases
significantly. This leads to longer, cooler flames, with lower peak
flame temperatures and the peak temperature is likely to move
slightly further down the kiln. This has two potential
consequences. Firstly the lower flame temperature will reduce the
rate of heat transfer, and so kiln exit gases could have a higher
volume even if the mass was lower. Secondly there may not be
sufficient heat in flame to control any perturbations in the
system. Where extra coal was added the flame could then become
more reducing.

50
 b) Volatile cycles within a kiln (which are considered in Section 6.3)
increase greatly as the flame becomes more reducing. These
involve minor components which can evaporate in the burning zone
and then condense in the cooler zones of the kiln and preheater
either depositing onto the surface of meal or the sides of the
preheater. In either case the result can be material building up on
the sides of the riser ducts and so reducing the area available for
gas flow.

This creates higher gas velocities and increased pressure

drop. It is known that many preheater kiln are significantly
effected by volatile cycles, and that operation at higher
oxygen levels will help to contain these effects. The
optimum oxygen level for any particular system will depend
on the relative proportions of potentially volatile
components present in the raw materials (meai and fuel)
and on the efficiency of operation of individual components
of the system; such as firing system, cooler, meal
distribution.

The overall result of the above considerations is that most preheater kilns will now
operate with a backend oxygen level of between 2.0% and 2.5%. Preheater exit oxygen
levels will normally be 1.0% to 1.5% higher than the backend oxygen level. These
values will change significantly when auxiliary firing (Section 6.4. ) is applied. The
backend level will rise by about 0.5% (to 2.5% to 3.0%) whilst the preheater exit level
will fall, commonly running between 3.0% and 3.5%.

Where fan capacity has become a limiting factor, fan uprating should be considered.
However gas velocities in key areas such as the kiln hearth should also be considered
before finalizing on this as a solution.

6.2 Poor Heat Exchange

Where poor heat exchange has been identified over a particular stage, this will normally
be due to material short circuiting, to poor material distribution in the duct or to a
combination of both.
..

The commonest causes of poor heat exchange are inappropriate siting of the material
inlet or poor material distribution. Where the inlet is positioned too high on the duct,
the heat transfer time is reduced, while if it is sited too low on the riser duct there is
the possibility for material to short circuit, especially if the cyclone has no or little
vortex tube.

Material distribution is achieved by material !bouncing off? a splash plate or distribution

box. The splash plate was the original device whilst distribution boxes are now more
common, especially on the lower stages. Splash plates are installed within the duct and

51
 Splash Box Operation

?...

.-

‘,

52
potentially suffer from attack caused by the local conditions, but are adjustable,
whereas the splash box is attached to the outside of the duct but cannot be adjusted.
Where poor distribution is identified as a regular problem a splash plate should be
adjusted or replaced with a distribution box at the next major repair. On the 10WV
stages where material may be sticky it is advisable to attach some means of externally
cleaning the box, such as compressed air. Operation of a splash box is shown in Fi~e
26. Where material short circuiting is a problem, this is nonnaily seen to be more
severe when the system is operating on reduced throughput, with the fan held back, and
may often give an unstable kiln. It is expected that theoretical calculations would
indicate a low gas velocity at this point.

The simple cure is to locally increase the gas velocity just below the meal inlet point
for the stage by building in a refractory constriction in order to reduce the gas flow
area. Care should be taken to avoid excessive turbulence by angling the constriction.

Where heat transfer within each stage appears to be satisfactom, but fuel consumption
is high then the condition of the cyclone vortexes and material pipe flap valves should
be examined. Particularly in oIder systems it is not uncommon for the vortex to have
been removed from the lowest stage. This is because these vortexes suffered from
failure and would collapse into the cyclone body causing a blockage. Removal and
replacement would then cause a lengthy stoppage. In order to avoid this the penalty of
reduced fuel efficiency was accepted. Removal of a vortex in either of the bottom two
stages is considered to reduce cyclone efficiency by 10 to 15%, and have a fuel penalty
of about 10 kcals/kg clinker. Over the last ten years novel vortex designs have been
developed for the lower stage cyclones in which the vortex is formed from a number of
small interlined pieces which can be removed or replaced much more quickly than the
old form which had to be weided.

These new vortexes are available in high temperature steels or refractory materials, and
in combination with improved hanging techniques are proving sufficiently reliable to last
for a normai kiln campaign (one year). Example of this is given in Figure 27. The
suppliers are still improving the designs and so any vortex replacement should consider
the latest market position, and the possible volatile cycie within the kiln system, as the
extent of this could have a bearing on the preferred composition of a metal vortex. In
general the metal vortexes are currently proving to be more reliable than the ceramic
versions.

53
 1.., I I all
4
Zx b
*
a
oIn I
..... .........
...............
...............
...............
...............I
I
k d
u
d r I a=kx 1 1
54
 6.3 Volatile Cvcles

6.3.1 Cause

A limited number of minor components in the raw mix and/or fuel can become highly
concentrated within the cement kiln system and then create operational problems. The
components that are considered to potentially cause difficulties are compounds of
Chlorine, Fluorine, Sulphur, and the Alkalies (Sodium and Potassium). These substances
are present in the raw materials in low proportions, but are likely to evaporate or
decompose under the temperature regimes found in the burning zone. Once this
happens they become associated with the gas stream and cool aS this looses its heat to
the material bed, until they either condetie or react to form compounds that will
condense. At this point they are present in the gqs stream as potentially sticky liquids
and will adhere to any surface with which contact is made; this can be a particle
surface or a vessel side wall. Once on a particle this can still stick to any wall with
which contact is made until the temperature drops to a sufficiently low level for the
liquid to solidify. The temperatures at which the potential compounds are liquid
coincide with those found at the kiln hearth and in the lower preheater stages, and will
lead to material building up on the walls in these areas. At best this reduces the duct
dimensions and so causes increased pressure drop, whilst at worst it reduces output and
causes instability or cause blockages to develop in the preheater system, leading to
significant down time.

6.3.2 Chemistry Based ~hltiOtlS

From a Process Engineer’s point of view, the easiest way to minimise the volatile cycles
is to push the problem to the Works chemist and expect the raw material mix to be
changed to minimise the inputs of volatile materials. Some effort can also be put to
balancing the inputs in order to minimise the potential for evaporation at any particular
kiln condition (ie increase the potential for the compound to come out in the clinker:
Alkali Sulphates are less likely to evaporate than Calcium Sulphate, but will increase
the alkali level in clinker). Another chemical approach is to control the type of build-
up, as alkali sulphate based build-up will be harder and more difficult to remove than
Calcium Sulphate based deposits.

..
6.3.3 ~ Based sOhltiOOS

The magnitude of the volatile cycles can also be reduced by careful control of kiln
conditions. In general volatile cycles increase slowly with rising temperature or length
of burning zone. There will also be a significant increase in cycle if reducing conditions
develop ciose to the material bed. Consequently, the magnitude of these cycles can be
reduced by operation at lower temperatures or, on occasions, higher kiln oxygen levels
although this will depend on the starting point.

The areas over which serious build-up occurs can also be minimised by ensuring that the
material feed to the lowest stage riser is:

55
a) Well distributed across the duct

b) Sited as low as possible whilst still avoiding material short circuiting.

These two factors ensure that the temperature zone in which deposition is likely, is
minimised by reducing the gas temperature quickly and efficiently.

Once it has been accepted that some build-up will occur,’ a method of removing it has
to be devised. Two basic methods exist (manual or automatic), and normally a
combination of the two is required to maintain efficient operation. For manual cleaning
‘poke holes’ are usually placed at strategic points around the hearth, the riser ,ducts to
the lower preheater stages, and in the cyclone bodies. Soft build-up can be removed by
compressed air lances, whilst hard deposits require the use of either steel bars,,, or high
pressure (5000 psi) water lances. Particularly when using high pressure water, care must
be taken not to damage the surrounding refractory. Manual cleaning requirements are
site specific and can vary from occasional -as dictated by preheater suctions - up to on
a daily basis.

If buildup is persistent and always occurs in the same part of the preheater tower then
the use of air blasters is to be recommended. These consist of a small pressure tank,
fed off a ring main, which is attached to a nozzle fitted into the riser duct. This fires
off the contents of the tank at a selected time interval so that the air blast removes
material building up around the area of the open nozzle. A number of nozzles will be
necessary to clean an area, and as the nozzle has to be imbedded in refractory it is ,not
easy to relocate a blaster if its effectiveness is found to be poor. Blasters are intended
to move a slab of material, not just the area of the nozzle. For this to be achieved the
time cycle of the blast has to allow material to build-up; too short a blast cycle simply
moves material over the area of the nozzle.

Extreme care must be taken when removing build-ups, particularly when a cyclone
blockage has deveioped. Hot raw meal flows very freely and can cover a floor or run
down through the preheater and kiln in seconds. Basic rules are:

● Locate probable area of build-up or blockage from control room

instrumentation

● Use all available protective clothing, including heat resistant items for
head, eye and hands as a minimum. Towards the lower stages full heat
resistant body covers are required.

● Do not work close to the build-up; ie use long lances or poles

● Never work over the pole or lance.

● If blasters are also in use, these should be turned off and discharged
.
before working in that riser duct.

56
 ● Work from the top of the preheater down.

● Work on a blockage from above. If possible attack it as quickly as

possible, as it will become harder as it cools down.

● When working on a blockage, personnel should not be allowed access to

any floor or area below the blockage or in the potential direction of
material flow. This includes the kiln inlet, kiln hood and cooler outlet
areas.

6.3.4 Design Concept Based Solutions (Gas Bleeds)

Volatile cycles develop because the recirculating components become trapped within the
kiln/preheater system, with only proportions of the component in the feed to the burning
zone passing through this area and leaving in the clinker, and very small quantities
working leaving the system completely via the stack. Volatile cycles can also be
controlled by deliberately extracting a high concentration of the volatiles from a
suitable point in the system. This can be achieved by extracting some of the gas stream
from the system at the base of the riser duct immediately above the kiln backend, as
at this point the concentration of volatiles in the gas stream is high.

Any gas bleed will have a fuel penalty associated with it, as the high temperature gases
would normally be used for preheating meal. The gases will also have a significant dust
content which has to be collected and dumped. The loss of this dust increases the raw
material requirement per tonne of clinker. For a standard Suspension Preheater, it is
generally assumed that a bleed level of about 30% prepresents the maximum
economically justifiable level. Bleeds can also be utilised in precalciner systems and
in an air separate calciner, in which the kiln exit gas volume is 30 to 45°A of that of
an SP kiln, 100% kiln gas bleeds are justifiable under certain conditions. In considering
this, control of cycles for improved kiln operation is easier on a Calciner kiln than on
a simple Suspension Preheater, however bleeds are also used to control alkali levels in
clinker, and this is more easily done on an SP kiln than a Calciner, as the latter
normally generates lower magnitude volatile cycles.

The gas/dust stream is pulled out of the preheater by an induced draught fan, and then
quickly mixed with cooiing air in order to reduce the temperature to below the level at
which the volatiles solidify. A typical detail from this area is shown in Figure 28. Most
systems are designed to reduce this temperature to between 300 and 400 C initially.
After this the gases will normaily be further cooled using air and/or water (conditioning
tower) to a temperature at which the gases can be dedusted. Precipitators or bag
filters have both been used for de-dusting kiln gas bleed systems although at present
EPs are preferred as some difficuitiw have been encountered with bag cleaning. In
some circumstances where a very low bleed (5 to 10%) is required the bleed stream can
be passed through a cyclone and then the largely de-dusted gas stream can be returned
to the main kiln dust collection system for final de-dusting. This bleed will not be as
efficient as one with separate total dedusting - especially where the bleed is designed
for Chloride control - and an enlarged main precipitator will be required, but this can

57
1’. ftotary kiln
2. Gas duct I
3. Bypass-connection
4. Fresh air-supply
5. 6ypass mixing chamber
6. Diluted bypass gas I
@
7. Tertiary air Inlet .

w
pk
8. Meal feed in calciner
9. Calciner

.—.—
o1
“ - .
. .

FIGURE 28

DETAILED LAYOUT OFGASBLEED EXTRACTION

be a cheaper capital cost option than installation of a separate de-dusting system.
Where water is in short supply a hot precipitator can be used and the gases can be
treated at 300 to 350 C. In this case the relative costs of a hot precipitator have to be
balanced against those of a standard EP, conditioning tower, and water
treatment/preparation. Typical bleed system flow sheets, including gas cleaning and
material processing, are presented in Figures 29 to 31.

The cost penaities associated with a bleed system arise from:

● loss of heat content in bleed gases

● extra material processing up to and including kiln feed

● operating cost of bleed system, including fans and dust collection

The fuel penalty is variable as it depends on the dust loss as well as the gas
temperature. The suppliers have put considerable effect into improved hearth design
over recent years in order to minimise dust generation at the hearth and the consequent
potential for material recirculation to the preheater. The fuel penalty is likely to be
2.0 to 2.5 kcals for each percentage of kiln gases bypassed, and this range will be valid
for simple preheater and calciner systems with the higher value applying to older
systems with higher dust losses and to calciner kilns with higher gas exit temperatures.
Figure 32 can be used to estimate the fuel penalty associated with bleed systems.

The other two iterns cannot be quantified as they will be site specific, however
comment can be made concerning the quantity of dust loss associated with a bleed
system. The old BCI standard is to assume that raw material processing, up to and
including the kiln feed system, should be increased in size by 1.5%, for each anticipated
10% bleed of kiln gases. Dust loss itself on older systems with high hearth gas
velocities has been found to be around 400 grrns per Nm3 of gases extracted. On more
modern systems, with lower gas velocities, this can be reduced to around 200 grins per
Nm3 removed from the riser. The suppliers claim that the standard hearth designs -
especially on large calciner kilns - will now generate significantly less dust than this,
but that in some of these cases the dust generation has to be increased deliberately in
order to control the effects of the gas bourne volatiles by increasing the quantity of
dust available for absorption of these volatiles. Consequently some suppliers are now
designing a variable splash plate into the base of the kiln feed pipe in order to optimise
dust recirculation between the kiln and riser duct during the commissioning phase.

6.4 Auxilliarv (Backend) Firing

Two significant limitations to production on preheater kilns are the quantity of heat
available in the kiln exit gases for transfer into the meal and the gas velocity at the
hearth. In some kilns where output has Steadily been improved since the original
installation, coal mill capacity may also have become a limiting factor, or alternatively
high coal firing rates may be having a detrimental effect on refractory life.

59
 I
I
b
60
 ..

090.!3

y
.-” , ,’u I
bi -63 I -6> i II /lLll\

T’

“
OIL-PREPLRAI I (m

07
— [

( b Jlacf
J

23 &?f4u

-100 12
m
129xlcm
4ca il 100.14 L Gzck
I 94
100.16 I?O@
I — ~ .*AJ
-!-00.17
10-1 la44&a
~+o eo

6,. %
l!ud .
100.24
100.20
UAW* * .~

!00.03
= I & ‘“’k
—

CmLid
1-.*-/k.
+ d&
— l!-?!-?
z .,025-BKH
I -.,

mGURE 30

POLYSIUS DESIGN GAS BLEED AND DUS’I’ HANDLING

61
 FIGURE 31

KHD DESIGN GAS BLEED AND DUST HANDLING

- USING HIGH TEMPEWTURE PRECIPITATOR

. —. .— ._l—.—— *1,*

l-=

Iru%.

i 3-
i
I

8
L.—. —.— .—. _
.—. — -—-— .—-— -—-— am
._. _ . A’

M8.m

UnJln
. •n

?’T’
lm.M

Ala,m

, AI*, n
mum
I

@===@l
&
●
410

s 11*

At

I
110 a?
m

m
maw.I
I 110.J6
!
 Figure 32

ESTIMATED HEAT CONSUMPTION VS % GAS BLEED OFF

AT THE KILN BACK END
950

.
/
925
●

0
/
.

●
@ “ ;0% DU ST LOSS
0
.
.
# . #“
. &
x 0 #
●

.
.
4’ . H“
0 e
875 4“ 10% DUST LOSS
30% DUST LOSS 4 &
I ‘“ . M“
10% DUST LOSS . z “ . #
d~
4 ● d“
. 0 “
850 .
4
d
1- / 4“
a
z #
●
#
.
0“
● ●

— SUSPENSION PREHEATER KILN

. M“ (100% FUEL INJECTION INTO THE KILN)
&
.
825 0. 0.- PRECALCINER KILN
(40% FUEL INJECTION INTO THE KILN)

I DUST LOSS E)( KILN % ON CLINKER

800 -
1, I
I I I I I
o 10 20 30 40 50 Go 70 80 go ‘
100
%GAS BLEED OFF AT THE KILN BACK END
In any of these situations auxiliary firing offers potential advantages. In this process
a limited quantity of fuel is fired into the riser duct from the kiln shortly above the
hearth area. Due to the intimate gas/soIid contact the heat generated by combustion
is absorbed efficiently, and the combustion gases are generated in the riser and so do
not contribute to gas velocity over the hearth. The kiln h~ to be operated with a
higher backend oxygen level in order to provide the oxygen for combustion, but some
of the back end seal inleak will also take part in the combustion reactions. It is known
that this inleak tends to stream up the back wall of the lowest preheater stage (the wall
on the kiln side of the preheater riser) giving significantly increased oxygen
concentrations, and so BCI normally position the auxiliary fuel injection point to fire
into this area, but away from the wall. Overall preheater exit temperatures and CO
levels will rise slightly, but NOX emissions would be expected to drop as combustion in
this area will be below the temperature at which thermal NOX will be generated.

In addition to increasing output, auxiliary firing is frequently credited with increasing

kiln stability and reducing the magnitude of volatile cycles. The later arises from
increased oxygen levels in the kiln, and from easing the duty on the front end burner so
making the existence of reducing conditions less likely.

The quantity of fuel that can burnt in the back end riser duct will vv depending on the
size of the system and the type of fuel to be burnt, but in general, with a suitably
prepared fuei, auxiliary firing rates of up to 20% of the totai fuel requirements have
been achieved. Higher levels tend to give rise to excessive temperature and CO
increases at the preheater exit.

Auxiliary firing can be divided into three forms. In the first a normal high grade fuei
is used. The target here is simply to increase production and benefit from increased
stability. The second type uses well prepared cheaper fuels or combustible wastes.In
this case the production benefits are reduced marginally, in favour of reduced fuel
costs. The third approach is to use poorly prepared combustible wastes. In this last
form the production advantages will be significantly reduced but substantial cost savings
are likely as the waste is likely to be delivered free or at a negative cost (payment
received for disposal).

Two potential problems exist with large sized materials (such as complete tyms):

a) Input to the system is a batch (even if one unit at a time) process.This

will lead to reduced control and potential cycling of items such as
preheater exit CO level. It will also require the use of higher excess air
levels.

b) The material will drop into the kiln and combustion will initiate
there. Depending on the oxygen levels combustion or pyrolysis will
occur, with the products being carried in the gas stream to the
preheater where the combustion reactions will continue to
completion.

64
 In these cases there wiil be limited benefit in terms of gas velocity across the hearth.

On suspension preheater kilns BCI have direct experience with the use of oils, coals, and
tyres covering ten kiln systems, and information from a number of other sites as a
consequence of our participation in European Cement Makers technical forurn of CETIC.

Utilisation of auxiliary firing with prime fuels has produced output increases of upto
10°!4. However the reliability of this secondary firing system will have a major impact
on the achievable benefits. The most reliable systems have been that have been
supplied from a kiln indirect firing system with spare capacity. Where this has not been
possible a small mill - known as the TASS mill (Figure 33) - has been supplied to
directly fire to the backend area. Whilst these are allow cost capital option and
initially gave good results, in the longer term the reliability has not been acceptable.
Currently further development work is on-going concerning improved feeding systems
and alternative mill types.

6.5 Cyclone Modernisation

On older systems where pressure drop/fan capacity is a limiting factor, cyclone

modernisation can be considered as method for improving output, although the potential
effects on gas velocities in the riser ducts and -especially - across the kiln hearth shouid
be considered. Cyclone design is discussed briefly earlier in this paper and in the
Precalciner Kilns paper, whilst the variations available in cyclone modernisation have
been discussed in Sections 3.6.6 and 3.6.7 and so will not be repeated here.

As the major area of pressure drop is across the top stage cyclone the biggest potential
benefit can be gained in this area, but it is also the area in which the trade off in
cyclone efficiency against pressure drop can have the greatest detrimental effect on
the overall process. Consequently great care must be taken in the redesign and careful
application of the redesign.

Whilst improvement to the lower stages individually offer lower potential the overall
sum of improvements is still likely to be significant. In addition there are the extra
advantages that:

a) The redesign does not have to be as precise as for the top stages.
.. b,
Much of the necessary ‘metal bending’ can be completed whilst the kiln
is operational, so minimizing the down time required.

This approach has been taken at Harleyville, where the lower cyclone inlets have been
modified one by one during the planned major repairs, and the majority of the
replacement ductwork has been welded in place whilst the kiln was running.

65
 STORDY - TAS-COAL MICRONISING MILL

COAL INLET
AIR
INLET
●

++

AMA-
1
L
I

-1
I
8
. . . . .— . . ..——
r---- “
. . . . . .— . . . . . ..——.—

/
●

r-- “---- ‘..”--..----” --------

. . . .. . . . . . . . . . . . . . . . . . . . . . . . . ..u.

/
●

P1
iK -----

. . . .
.---—-.

. . . . . . . ——...

b
~ ,.-. ..-. ..-. -— —....
-.. .. . . . . . . ..- —..

I /’

‘I--J m MICROCOAL
4K ~ IN AIR
f ounm
1 I
I

.,.:::1:%!:::::
:Mm.tw.fi: ::::::
‘.:::
:,:,:
;
........y.~.;,~
‘...,,,...,.. ..,.w.::::,
.....
..: ..,.::::wf
....V
+..,
,?..#x<
.:.:s.~:: ...>.% ...w,....:.:.:.,...,,,,,,,,
..,..,.,..
....
.......... ......,,.,,,,,,,,.,.,.,.
...,...,,,
......,.,.,.
..,...,.,.,.,.,,,,,,,
....,:.:.:,:,:,:,>::,:,:.
.......... .x.
..,*+:.:.:.:.:.:
.
....,......iv .
.. ... .
.. . ........:~
.
....................... :::;:
::::::.:.:. .... ..:.:,:::,:,:.:.:,,,,
..>....,.,.,.,......,..
.......... ...,
..,.,
....... ..
,:.:.:,:.:.,.:,:.:.:,:,:.~,~...,.
7. HEAT AND MASS BALANCE DATA

A four stage preheater kiln can operate with a net fiel consumption of about 750 kcals/kg,
although in general the long term average value is somewhat higher, at 770 to 830 kcalskg.
Optimum fiel consumption does, however, need all aspects of operation to be carefblly controlled
as even minor inefficiencies can lead to significant increases in fiel consumption. Heat and mass
balance data for atypical four stage preheater kiln is shown in Table 7 and compared to the older
processes.

TABLE 7

HEAT AND MASS BALANCE DATA -1

WET LONG DRY SUSPENSION GEPOL

PROCESS PROCESS HEATER PROCESS

Theoretical 420 420 420 420

heat

Water 579 5 5 5
Evaporation

Dust 10SS 20 85

Exit gas loss 125 290 150 250

Cooler exhaust 75 100 110 100

Heat with 13 20 27 25
clinker

Shell loss 148 160 60 70

-kiln/cooler

Shell loss 20 40
-other

TOTAL 1380 1080 792 910

All data in kcals/kg (net)

This shows that the major differences in fhel consumption between an SP kiln and the long kiln
systems arise from the kiln waste gas losses and fi-om kiln shell radiation. Table 8 sets out a
comparison of the typical SP kiln with a Hope SP kiln operating under two conditions: good
operation and poor operatio% and this emphasises the importance of efficient operation to the SP
kiln.

67
 TABLE 8

HEAT AND MASS BALANCE DATA -2

Typical SP Process Hope - goqd Hope - poor

operation operation

HEAT IN

Combustion-net 764 807 872

other 30 71 71

TOTAL 794 878 943

,.. ::::..:..:
:.:.:.:.;
:,::
... :: :....:., :,..: :y,:.. .....;+,:x..:,,.:,. ;.,:.:.:.:.:, :.:..
:.,:.::.. ,.;,,,,::,. ,,,..x:.
:~.:
.’.
,“..
.:,:.;’
.,:::::..:
.:..:::.::.
.:::,
:i:,
!.;
:::.
..:::::
,,,,,...,.i.i.?
:::,...;.:.;.:
.,,,,,,,::.:,
:,,;:.:...
., .,..,.,.,.,.,’
.,.,,. .:.,,:,
.,,,.:, ,,’;::,
,:,:,,:::,. /..:..,:.!.:,:::.:..
,,:, ...... ,.:.,:.::.
: :,’.:.:,
.. :::::X
,.,.. .,. . :,.:!$.
....3’.: ‘.,,:.
,, ”.’
.,
,:... ‘,.,.,.,
.:,...:
.’.!.::
,..:,.,,,,,.,
. L:.:W.,.,
,,:,.,..: ,. :::.,,,.’:.:.‘..:.:::,.,
,,::,
y.,..,:..:.:.::.,:,:,,
,..,. ,,..:.:...
.,.., .,,,.,, ,.:.: ,:,:
.“,’,
: .,, :.+: .
,.,
. .
,.
.... . ..:
,,,..:...,..,
,,,:
.,....?., ,,. .:’:
:’.’:”.”.
: .,.
...
, :“::.”.
,.
........ :,.,.:.,:,, .,,,..
.,. ,..,.:...
.,, ..-.
.. :,, ~ .:.::.::: ;., : :.. : .,
, : .::. ,,:,
,,:; ,.~,:
,. .. :. ,.,.:.:.:.:.,: ............
,,,,<,:<.
. ....:,: . ....,.,:..
,...: ... ,.:,.
::.:
,,. ;,~,;
....,,.:..,.,
...,x.,:’::
...::,.: , ,.....:,.....:.:.:,
..,,.:,..,. .. .,:
. ,.., ,....:: .,,,..,:,,,:,
,{.:...::.....:;
,,,: :’:.,
::...:;:~
.
,}.:::;
:.,.., ..:.
. .::
,.,..:..:.
. :.”
.‘.:.,’”:.:.
“.:,,,,.:
:.;:
:.,..?, ..:,’
,,. ::;
,:.’
:...,:’
:.:.’:,
.:
. :’:‘,’
:. : “.,:”
.’,’:
,,:::
....,’.’
. “..’:
.},:
::., ::,,.: ..,,,.,,,
?..::.,,.:.: .. ,., ,,, ..
. ..”.
,.....:. . . ..:..,’
,,..:,.:.:....
,.....’,.
.,., ,.,
,..:., ,,:..::,, ,..
.::::..... ::
:.: , .,:,:.:
,.:.,:.:,.,.
,:,.,..:,,::,,::::,
:.:
. .,,: ,:,
~::i,;,:
:i
.,. .. . ...,. ,.,:..,:.. ..+:’ ‘ .,:
‘ ,,,.
. ., ,:,:,:,;.:,
+: .,.’..‘.’-’.::.
!.M :.,.::..,.,,:.:,
,!-:::.,:. .. . ,,:::;,::::
fi:::.:.
,.: .. ,,, ,: .,
, ,.,:,:,,, .:,.,
.,,
{:,::.
:. ,. ,.. ........:,... . .’’.’.. ::.. ..,.:.;. .... :.,,:...., .
HEAT OUT

Theoretical heat 420 420 420

,,
Dust 10SS 6 7 5

Exit gas loss 174 259 310

Cooler exhaust 76 79 98

Heat with clinker 28 37 24

Shell loss- 65 61 71
kiln/cooler

Shell loss-other 25 15 15”

TOTAL 794 878 943

Although— low fuel consumption can be obtained from SP kilns any deviation from
optimum operation is likely to have a significant effect. “General guidelines for some
significant examples of this are listed below, although in practise each case should be
, examined closely:

● increase in kiln BET by 10 C

cost is 7 to 10 kcals/kg

● increase BE02 BY 1%
kiln
cost is 5 to 6 kcals/kg

● increase cooler efficiency by 10%

saving is 45 to 50 kcals/kg

68
● increase front seal inleak by 0.1 kg aidkg
cost is 11 to 14 kcaldkg

● increase backend seal inleak by O.lkg air/kg

cost is 17 to 20 kcaldkg

● bypass 1% kiln gases

cost is approximately 1.5 to 2.5 kcaldkg

● increase clinker LSF by 10/0

cost is 6 to 8 kcaldkg (low LSF)
cost is 15 to 20 kcaldkg(high LSF)

s’! increase clinker SR by O.10/0

cost will be between 6 and 18 kcaldkg

● add fifth preheater stage

saving will be approx 25 kcaldkg

● add sixth preheater stage

saving will be 10 to 15 kcaldkg

● remove fourth preheater stage

cost will be approx 35 kcaldkg

69
8. SUMMARY

The suspension preheater kiln represented a major step forward in system design at the time of
its inception and by the 1970’s had become the preferred production option. Using this
technology kiln systems were built with production capacities of up’to approximately 4500 tonnes
per day, although the more typical size is 1500 to 3000 tpd. In the 1980’s precalciner kiln systems
were developed as an extension of SP technology with allows greater output to be realised from
a given kiln tube, although with a marginally higher burner fiel requirement. Consequently
precalciners are now fi-equently preferred to simple SP kilns, although an SP kiln is arguably still
a more suitable installation for a production requirement of 1000 to 2000 tonnes/day.

The technology of suspension preheating is still developing, being applied to both SP and
precalciner kilns, so there is the potential for on-going improvement to existing kiln systems for
many years to come.

Where the market is increasing significantly there is the potential for modernisation of an SP kiln
into a calciner system to meet the extra production requirement, whilst in a static market attention
to detail in the day to day operation of an SP kiln gives the potential to minimise production costs.
In either case the SP kiln will be around for many years yet.

70
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 7

Precalciner Kilns
(CTC 19)
 PRECALCINER KILNS

CONTENTS

1. INTRODUCTION

2. PRECALCINER KILN PROCESSES - MAIN FEATURES

3. ADVANTAGES OF THE PRECALCINER KILN PROCESS

4. TYPES OF PRECALCINER SYSTEM

4.1 Polysius Prepol Precalciner

4.2 FLS - ILC and SLC Precalciner
4.3 KHD Pyroclon Precalciner
4.4 IHI - NSF and SF Precalciners
4.5 RSP Precalciner Process
4.6 Mitsubishi Fluidised Bed Precalciner, MFC and N-MFC

5. AUTOMATIC CONTROL OF THE PRECALCINER VESSEL

6. FACTORS AFFMHTNG THE OUTPUT OF A-S PRECALCINER

SYSTEMS

6.1 General Considerations

6.2 Comparison of FLS Precalciner Kiln Systems at Cauldon
and Cemento Melon

7. BCI OUTPUT WTINGS FOR PRECALCINER KILNS

7.1 Introduction
7.2 Kiln Shell Output Ratings - Air Separate Precalciner Kilns
7.3 Kiln Shell Output Ratings - Air Through Precalciner Kilns
7.4 Kiln Hearth Gas Velocity Considerations
7.5 Precalciner Vessel Ratings
7.6 The Effect of Fuel Quality upon Precaiciner Performance
 8. THE “BEST” PRECALCINER KILN SYSTEM

9. PRECALCINER KILN UPRATINGS - CASE STUDIES

9.1 General
9.2 Low Cost Uprating - Cemento Melon No. 9 Kiln
9.3 Conventional Plant Uprating - Kanthan No. 3 Kiln

9.3.1 Introduction
9.3.2 Preconversion - Plant Limitations
9.3.3 Kiln Uprating Modifications
9.3.4 Technical Evaluation of Uprating Proposals
9.3.5 Results Obtained from Kanthan No. 3 Kdn Uprating

9.4 Conclusions Drawn From Recent Precalciner Kiln Uprating Exercises

10. OPERATIONAL PROBLEMS WITH PRECALCINER KILNS:

EXAMPLES AND SOLUTIONS

10.1 General
10.2 Practical Examples

11. TYPICAL OPERATING CONDITIONS AND HEAT BALANCES

FOR A MODERN A-S PRECALCINER KILN

11.1 Number of Cyclone Stages

11.2 Typical Operating Conditions and Heat Balance
11.3 Gas Velocity Profiles
11.4 Preheater Cyclone Loadings
11.5 Preheater Cyclone Geometry
11.6 Summary

APPENDIX I CARBON DIOXIDE CONTENT OF THE GASES

LEAVING A KILN AND THE DECARBONATION OF
THE MEAL ENTERING THE KILN
1. INTRODUCTION

Demands for higher outputs from kilns and the need to utilise raw materials which are
less suited to traditional preheater kilns, have resulted in the development of new
cement manufacturing processes. The main area of development has been focused upon
the Precalciner kiln process.

During the 1980’s standard unit kiln size was typically around 3000 tonnes per day
clinker output. The traditional preheater process is less suitable for higher outputs due
to the large kiln diameter required and the resultant problems with bricklife. Since this
time it has become more common to build kilns in the “standard” size range of 5,000 to
7,000 TPD clinker which gives the benefits of scale i.e. a lower capital cost per tonne
clinker. Precalciner kilns have been made with capacities up to 10,000 TPD.

This paper will examine the following aspects of Precalciner kilns:-

(i) Description of the main features of the Precalciner kiln process.

(ii) Types of Precalciner design currently available.

(iii) Factors affecting the performance of Precalciner kiln systems i.e. output,
fuel consumption and process stability. Some typical precalciner kiln
operating problems and their solutions are described in order to assist
engineers with kiln optimisation.

(iv) An examination of the output ratings which have been achieved in

particular, the kiln shell and Precalciner vessel loadings and their
interrelationship.

(v) Precalciner kiln uprating - two recent exarnpl= of cost effective kiln
upratings are described in order to demonstrate practical ways of
overcoming plant bottlenecks.

The main aim of this paper is to concentrate on how to get the best performance from
a Precalciner kiln process. Whilst some of the commercially available systems are
briefly described, this paper is not intended as a description of all the systems currently
available or being developed.

We will now consider the features and categorisation of the various Precalciner kiln
processes.
2. PRECALCINER KILN PROCESS - MAIN FEATURES

Precalciner kiln processes have the following features:-

(i) A proportion of the total fuel input is introduced at a point between the
kiln and the preheater system. A purpose’ designed precalciner vessel
allows the fuel to combust and the decarbonation process to take place.

(ii) Combustion air required for the Precalciner vessel is provided by the
following methods according to the type of Precalciner system used. The
two basic systems are described in greater detail later on in this paper,
but are briefly as follows:-

Air-seuarate (A-S) Precalciner - Combustion air for the Precalciner is

taken from the kiln hood or a take-off from the clinker cooler via the
Tertiary air duct. Typical fuel input i&- 40% to kiln: 60% to
Precalciner.

Air-through (AT) Precalciner - Combustion air for the Precalciner is

obtained by running the kiln with higher levels of excess air. Typical fuel
split is:- 65% to kiln: 35% to Precalciner.

(iii) Material from the preheater cyclone(s) undergoes a greater degree of

calcination than in the conventional preheater kiln process. This is
achieved by exposing the meal to high temperatures for a short period
within the Precalciner vessel. Typically, in an Air separate precalciner
system, some 60% of the total fuel is in combusted in the Precalciner
vessel. The gas residence time is typically 2 seconds and the meal is
exposed to gas temperatures around 860-900°C (typically 880°C). This
residence time is critical to the performance of the precalciner as is the
vessel design. The de-carbonation level of the meal is typically increased
to approximately 92% within the Precalciner vessel.

(iv) Because of the high decarbonation level of the meal leaving the
Precalciner/Preheater system, the thermai load requirements of the kiln
tube are significantly reduced. In the Air separate Precalciner system
some 40% of the fuel is used to:-

complete the decarbonation process

raise the meal to the sintering zone temperature to produce

clinker

(v) Due to the reduced thermal load required in the kiln tube, a higher
specific output is possible from the Precalciner kiln process especially
with the A-S system.
 For example, comparing the similar size kilns at Hope and Rawang Works:-

Hope Rawanq

Kiln System Preheater with A-S Precalciner

riser duct firing

Kiln shell size (m) 4.77 x 70 4.7 x 74

Kiln inside lining

surface area (M2) 961 1000

Peak Clinker
output (TPD) 2350 5070

Kiln shell loading

(TPDiM2) 2.45 5.07

Hence, the Rawang kiln achieved 240% of the loading of the Hope preheater kilns. Whilst
this comparison is not complete and other factors tiect the loadings achieved, it does
illustrate how the Precalciner kiln process can allow large kiln capacities to be achieved.

(vi) It will be noted that the commercially available precalciner kiln systems do not
attempt to achieve 100°/0decarbonation within the precalciner vessel. The reasons
for this are as follows:-

An exponential increase in the gas/meal residence time would be needed

to achieve higher decarbonation levels. Alternatively, the higher operating
temperature would make system less fiel efficient.

Greater risk of build-ups between the precalciner and the kiln inlet.

3
3. ADVANTAGES OF THE PRECALCINER KILN PROCESS

(i) With conventional preheater kiln systems, outputs in excess of 3000 TPD
require large kiln diameters i.e. above 5.Om shell diameter. A feature of
these large kiins is the increase in kiln bricking problems. Hence, the
higher specific kiln shell output achieved by the Precalciner process can
help to overcome these limitations.

(ii) Kiln stability - a correctly designed Precalciner kiln system can have
several advantages over conventional preheater systems by virtue of the
following:-

(a) The major proportion of the meal decarbonation process is

controlled outside of the kiln shell.

(b) Simple process control loops e.g. the automatic control of the
calciner fuel input based upon the exit gas temperature allow a
tight control over the level of meal decarbonation to be achieved.

Both these factors can help to make the Precalciner process generally
more stable than the Preheater process. If correctly designed, controlled
and operated, there is generally less risk of kiln fhishing associated with
the Precalciner kiln process. However, reference is drawn to Section 10
which describes some typical problems associated with Precalciner kilns
operation and the techniques used to analyse and overcome’ these
probhxns.

(iii) The precalciner system is more suitable for materials containing higher
levels of sulphur, chloride and alkalies than are normally acceptable for
stable preheater kiln operation. Where the raw material chemistry
dictates the need for a kiln bleed system, this can be more efficiently
achieved in the Precalciner procm. Figure 1 shows the relationship
between the overall fuel consumption for varying degrees of gas bleed
from the kiln back end. In the Precalciner kiln, the quantity of gas per
Kg clinker leaving the kiln is significantly lower than in a Preheater kiln.
Hence, the concentration of alkalies etc, is higher and approximately only
half of the gas volume needs to be bled off for the same alkali reduction
level.

The flexibility of the Precalciner process allows it to be considered a

more economical process for materials which are borderline on chemistry
for a conventional preheater process.
 Fig. / EST I MATED HEAT CONSUMPTION VS “h GAS BLEED OFF
AT THE KILN BACK END
HEAT CONSUMPTION
(k callkg NETT)
950

./
925 /
/“
,“ 30”10DUST LOSS
●
/ A
900 .
./ ‘
/ ‘
●
/
. 10”/. DUST LOSS
875 .-
30°1. DUST LOSS /“
.
6
,/”
850
—SUSPENSION PREHEATER KILN
(100% FUEL INJECTION INTO THE KILI

,<
—.— PRECALCINER KILN
825 (60”1. FUEL INJECTION INTO THE KILN)

800 Z 20 30
% GAS
60 50
DUST LOSS EX KILN
I
60
BLEED OFF J4TTHE KILN BACK END
I
70
“1. ON CLINKER
1
,
00
1
90 100

ul
(iv) Capital cost - the capital cost differences between Preheater and
Precalciner kiln systems are not clearly defined for the following
reasons:-

(a) New kiln lines tend to be built in standard sizes with unit sizes
tending to increase in recent years. Hence, the upward trend in
outputs from 3000 to 6000 TPD (and above) new kiln lines tends to
automatically dictate the choice of a precalciner system. This is
to permit the lowest capital investment per tonne clinker
produced.

(b) for smaller kiln outputs i.e. below 2000 TPD, the Precalciner kiln
system may have a slight disadvantage over the preheater kiln in
capital cost terms.

Notwithstanding this, the new kiln line for Athi River is an A-S
Precalciner kiln of 1600 TPD. The 1290 TPD No. 9 kiln at Cemento
Melon is an A-S Precalciner kiln though this choice was dictated by the
requirements of a 30% bleed for sulphur. This kiln currently operates at
around 1850 TPD without use of the Bleed. Hence, there are many cases
of small kilns using the Precalciner process in preference to the preheater
process.

(v) The reduction of NOX emission levels through use of precalciner kiln
technology is widely published. The use of the so called De-NOx burners
is a feature of some precalciners, such as the DD furnace and IHI-NSF
designs. Their purpose is to produce local reducing conditions in part of
the precalciner vessel in order to reduce the levels of NOX produced. The
reduction in NOX emission is becoming an increasingly important
environmental requirement.

6
4. TYPES OF PRECALCINER

There are many different designs of Precalciner in operation and classification by type
is difficult. This is due to several designs using various combinations of the different
features. A simplified classification is as follows:

(i) Extended Riser Duct systems such as the KHD PYROCLON and
POLYSIUS PREPOL designs.

(ii) Spouted Bed types such as the FLS+LC and KAWASAKI NKSV types.

(iii) Precalciners with a separate combustion chamber such as the

ONODA/KAWASAKI RSP precalciner.

(iv) Fluidised bed precalciner - MITSUBISHI N-MFC system.

(v) Precalciner vessels employing a swirling gas action, e.g. IHI-NSF and CSF
types.

The above classification is complicated by the numerous design options and modification
possibilities. Hence, separate combustion chambers are now available with the PREPOL
design and ductwork extensions have been added to earlier IHI-SF systems such as
Rawang No.3 Kiln.

In the following section, some of the common designs available will be briefly described.
For further details please refer to the relevant manufacturers 1 sales literature.

4.1 Polvsius Preuol Precalciner

This system is available in A.S. and AT versions and is currently used with the DOPOL-
90 preheater system. The design shown in Figure 2A is of the Extended Riser Duct
type.

A modified version of the PREPOL-AS is the Combustion Chamber (C. C.) type which
is designed primarily for low grade fuel. This uses a separate combustion chamber
which allows combustion to take place in pure tertiary air. Figure 2B also shows this
design.

The PREPOL-AS design is installed at Dunbar and Lichtenburg Works. Both plants
suffered problems with poor kiln stability initially and contributory reasons for this were
as follows:

(i) The meal entering the PREPOL was not well dispersed in the gas stream
and some meal could bypass the PREPOL and enter the kiln directly.
 FIGURE 2 A

PREPOL@-AS

PREPOL”-AS

Thecombustionair for the PREPOL@-AS

(Air Separate) calaner is passed by the
rota~ kiln separately via the tertiary air
duct. The advanta e is that a smaller-
sized kiln can then \ e used.

in the PREPOL@-AS,the fuel feed points

on the calciner are iacated “nthe
transition pieces between the tertiary air
dud and the caianer. The gnition and
combustion condiions are ideal at this
“*
. . . -

:Rli%i%’Ei2!ElfTh;2idad
the kii gases have the effect of
stabilizing the combustion cycle.

lhe raw meai from cyclone stage II is

entrained by the gas stream, separated
‘mstoge I of the preheater and passed
‘do the kiln inlet.

PREPOL”-AS-LC

The PREPOL”-AS- LC aiiows economical

use of high-ash, iow-caiorie fuels and
blends.

In contrast to the PREF!OL”-AS, the iniets

of the two tertia~ air flows of the
PREPOL”-AS- LC (Low-grade Com-
bustible) into the caianer shaft are
iateraiiy staggered. This aiiows optimum
controi of the mixture of fuei partides
and combustion air with a minimum
pressure drop. The combustion speed is
increased and compiete combustion
guaranteed.

~_.._ %J
1
‘J
 ..— . -..-.—— .—

PREPOL@-AS-CC for economical

use of low-grade fuels

FIGURE 2 B

PREPOLe-AS-CC

Po$miusdeveloped the PREPOL%S-CC

(combustion chamber) diner system
to improve combustion when us’ low-
grade fuek This asldnef can also
%!
used to advantage when:

● handlingraw material components

with hardly combustible camtkmh
and
● la e internal alkali cycles- whd
“h% cornbustian- are to be antid-
pateddueto the fuelandrow
material Wnpositkm

‘Theseparate cmbshan
“ chamber,
connectdta thecaldner loop byogas
duct, isthemainfeature of the
PREPOL”-AS-CC system

Cambustianstarts inpuretertia~airat
the -e of the combustionchamber
wimfeahatcare farms inthezaneof
low meal concentration $alii carbon
whddoes not bumatthecoreofthe
chamber isfed “fro the h kilngases
via a chute with the raw meal, where it
reacts with the residual oxygen.

The combustion chamber is fitted with

one tangential and one central tertia~
air inlet. hhaust gases from the kiln and
from the combustion chamber enter the
calaner loop.

The raw meal from cydane stage II

enters the combustion chamber near the
tangential tertia~ air connection.

lb fuel is iniected at the middle of the

combustion chamber. The tip of the
burner is located below the top edge of
the tangential tertiafy air inlet. The
psibitii of controlling the combustion
temperature ensures good calciner
efficiency when using solid fuels.

9
L
a
c
.-
tj
0
u
a)
k
I_
I
10
 (ii) Overdrafting of the preheater system was nec=ary to achieve adequate
meal pick up velocities. At Lichtenburg the ki~ tended to be run with
around 70% fuel spiit to the kiln to ensure a higher gas velocity in the
lower riser. This was prior to the system being modified.

(iii) The design has been modified to inciude more efficient gas accelerators
and dispersion boxes to improve meal pick up and to ensure more even
mixing in the gas stream.

Since these kilns were commissioned in the mid 19801s, Polysius have improved their
design. The Extended Riser Duct design is relatively simpie and can give good r~ulta,
provided that the above design details are resolved. This design is also very suitable for
retrofitting to existing kilns as part of an uprating exercise. FiWe 2C shows the latest
Prepoi design, which overcomes the above problem areas.

PREPOL-AT systems have been installed with capacities in the range of 800 to 4750
TPD. The PREPOL-AS system has been instailed with outputs from 1500 to 7S00 TPD.

4.2 FLS - TLC and SLC Precdciners

F L Smidth supply severai prectdciner designs in A-T and A+ configurations. Their

design features a spouted bed and the designations refer to:-

LL.C. - In Line Calciner

S.L.C. - Separate Line Calciner

See Figures 3A and 3B for details.

The ILC system is installed at Cauidon and Cenmnto Melon works. The SLC system
features a twin stream preheater with different cyclone sizes. The precaiciner vessel
is suppiied with combustion air via the Terthry air duct and is comected to the
Calciner stream cyclones. Material is also fed to the kiln stream cyciones which
feature smaller cyclones than the Calciner stream. Materiai leaving the lowest stage
of the kiln stream is also fed into the precaiciner vessel. This design has a high turn
down rate of capacitiu since it can operate as either single or twin stream system.

The SLC Precalciner thus operates in an oxygen rich environment since it receives air
from the cooler only. Hence, this system is well suited for use with low grade fuels and
gas firing.

In the ILC system, the Tertiary air duct joins the kiln back end riser duct immediately
below the bottom of the spouted bed. The gas velocity reduces from 20-25 m/s down
to around 5-6 nds in the spouted bed. This tends to make the meal recirculated around
the calciner vessel, giving it a longer residence time than the gas.

11
 FIGURE 3A
ILC-E: In - line Calciner
ILC-E: Separate Line Calciner

Double-string cyclone preheater kilnwith a precalcinerplaced parallelto the kiln riser duct Single-string cyclone preheater kiln with precalciner built into the kiln
Thus the combustion In the calcinertakes place In heatedatmosphericair drawn from the rtser duct Combustion air for the precalclner Is drawn through a
cooler through a separate tertiary air duct The exhaustgas from the calciner and the kiln
separate tertiary air duct between the cooler and the calciner.
are conducted through the two independentpreheaterstringswithout being mixed.

Features:

. Recommend4 capacity range 3000-5500 tpd Recommendedcapacity range: 1500-4000 tpd

. Fifing In calclneK 55- 6(Y% Ratio of firing in calclrreE55- 65?4.
. Bypass of kiln gas: 0- 3(M Bypassof kilngas: 0- 100?!
-320mmkW -470 mmVW -430mm WG

7
290”C
I30!FC

(
-1

480”C

670”C

740”C
I b’
‘“”u
-$s ‘i
4
620”C 890”C
900”C 880”C
1. Calclner 1. Calciner
2. Dividinggate 2. Damper
3. Changeovergate .-
3. try-pass .. I
4. By-pass 7 I II

780”C
4. w

2*C

T
3rc
The ILC-AS system has given good operational results at Cauldon and Cemento Melon. Two
kilns were also installed at Hermosillo (A-S) and Zapotiltic (AT) works in Mexico, formerly
within the BCI group. Details of the Air Through design (ILCE) using a modem 5 stage LP
cyclone preheater are shown in Figure 3C. General details of the spouted bed design are shown
in Figure 3D.

4.3 KHB Pvroclon Precalciner

The KHD Pyroclon design is an Extended Riser Duct design which is installed at Kanthan works
No. 3 kiln (3350 TPD). A 7000 TPD unit is currently being installed at Yanbu Cement with
BCTC consultancy input.

For outputs of around 5000-7000 TPD, KHD use two Pyroclon Precalciners with twin Tertiary
air ducts and twin preheater streams.

The Pyroclon design is shown in Figure 4. KHD have also developed their Pyrorapid short
Precalciner kiln system which features a two tyred kiln of 10-12 length to diameter ratio. These
kilns have developed very high output loadings and this aspect is discussed in Section 6.

The I?yroclon design is seen as a simple design and has given good service without the complexity
of some other systems.

4.4 IHI - NSF and SF Precalciners

The SF process (see Figure 5A) was developed by IHI and over 30 installations were made before
the process was upgraded to the NSF process. Rawang No. 3 kiln was one of the last SF
processes installed in 1980. This was converted to NSF process with a Trunk Duct precalciner
extension in 1992. Figure 5B shows the NSF process and Figure 5C shows the Rawang
conversion. The Trunk Duct design was developed at Chichibu cement and systems of this type
are also designated the CSF process. The earlier SF design suffered several problems with poor
combustio~ low gadmeal residence time, build ups in the vortex duct, raw meal short circuiting
the Flash Furnace, etc. In the NSF precalciner, many of these fhilings have been resolved by
redesign.

In the NSF procesq some of the meal enters the precalciner vessel directly (see point A in Figure
5C) whilst the remainder enters the riser duct (point B). The latter material quenches the hot
gases leaving the kiln and helps reduce build-ups. D-IIdo not tend to put all the meal into the riser
in order to reduce pressure losses. The Tertiary air enters the base of the precalciner or Flash
Furnace (FF). This sets up a swirling action. The precalciner burners are located either in the
Vortex Chamber roof or the base of the FF.

The new 6100 TPD Precalciner kiln line being installed at Kanthan Works (KK4) is also an IHI-
NSF type.

13
 FIGURE 3C
ILC-E: In - line Calciner usinq Excess air

Single-string cyclone preheater kiln with a small precalciner built into

the kiln riser duct. Combustion air for precalciner is drawn through
the kiln.

Features:

. Recommended capacity range: 900-3300 tpd

- Ratio of firing in calcinec 15 -25%
- Bypass of kiln gas: O - 25°A

- 375 mm WG
1 285°C

‘co
460”C

615°C

746°c

840°C
1. Calciner /
2. By-pass

1.

‘2. ~
1
m
1
200”C
I
 — ---

FIGuRE 3D

Preheated
rawmeal Praheatedraw meal

!-
. ‘k!
Q“
;, TOkilninlet
Gasfromkiln

Gasfromkiln
----- .
ILL--E mkz-na ILC cukinez

Exit gas with

prccaicincd raw meal

Preheatedrawmeal fromC.strirrg

0
Air fromcooler

\ll t *U u
SLC dim
f)
..

15
 _FiGURE 4 KHD PYROCLON PRECALCINER WITH PYRORAPID KILN

PymRAPiD@ -
rapid burning process realized

P\
in an extremely short rota~ kiln
) u
\w with 2 miief stations

I l\ HI Thelength: diameterratii
I l\E/ of a rotarykilnin the PYRO-
RAPID’systemis usually

ll\Yn no morethan 101. This

wows rapidtxeheatin9and
sinteri~ whichis knownto
barseliidlnker fxOPSftleS

● Iiile ~ requirement
castof refractory
● raduoed
metanal
.no morethan W kiln sup
@s tiich reeultain a
statidly definedsystem
/0 lowheat lossesthrough
W8iis

PYROCLON’S@8?71RP wlh
separate kiln gas he WId som-
M qdone sepamor for kdn
ad twr?craW waare 9as

,.6
 FIGURE 5A-IHI SF PROCESS

SF ~CESS. <ECTION THROUGM FURNA~

FIGURE 5B- IHI NSF PROCESS

T = 320°C

5=2=
P = -550 mmWG
preheater blowar .::.W ~
.,

r“”
‘.’,.::
. ...
+
..:-.. ..~’”.”
..-.
. .- / \ .,
.... ,—
,.-

.. ..‘1
,“

.f

‘:”./””

1-
L.P.L.
-t ,
JI
-\ cyclone
bb ,<.
—-CX Ol#et
t
I
I
; I — —,
@-------~/
u

+) L.P.L.
cyclona
IHI NSF-PROCE~s-

Clinker throughput
7200 t.p.d.
Fuel consumption
720 K&l/Kg
FIaeh

Qa,.
Furnaca Type of fuel
Coal only
------ -*
\ Nos. of cyclone stage
I@, -, -\ I
\<-, ~~i j 5-stage
“\ Nos. of L.P. L. cyclone
\ %-~ /’ 2nd, 3rd & 4th stage cyclones
\

\ t ‘WOn~av air ~Uct

\- ““’””” EzEiEi
:

yk
~
\
\ \. ;t

v
\ ~QG c n,,. ”.?.. -

T = 830”C \
\ Kiln burner
13 = 85% \\
\\
--
ROWY kiln
T m 1050”C ‘ : OL
.— ‘%L A 7
u, = Z.mk]
I
~~
Clinkw coolar C1inkor
17
l-l
.
!3
..—
I
The Trunk Duct design relies upon high gas velocities to reduce dust deposition problems
i.e. around 25 m/s velocity. However, this area can be prone to build up if the kiln is
operated at low throughputs for any period when the gas velocities are lower. The
Trunk Duct conversion at Rawang resulted in an increase in gas residence time from 1.1
to 2 seconds.

For new installation, such as the new IHI 1600 TPD kiln plant for Athi River, the
conventional NSF Precalciner process will be installed without any Trunk Duct. The
latter design is more appropriate for upratings or conversions.

4.5 RSP Precalciner Process

The RSP suspension preheater (abbreviation for Reinforced Suspension Preheater) has
been jointly developed by the Onoda Cement Company and Kawasaki Heavy Industries
in Japan.

It consists essentially of a two compartment calciner i.e. the heating shaft supplied with
a swirl burner and the calcining shaft arranged about parallel to it. The combustion
gases and raw material from the calciner are mixed in a mixing chamber before passing
to the lowest stage of the preheater. A typical flow sheet for the system is shown in
Figure 6A.

A more detailed representation of the calciner and mixing chamber is given in Figure
6B where it can be seen that hot air from the cooler enters both the swirl burner and
swirl calciner. The swirl burner burns only a small portion of the fuel used at this point,
but plays the main role in stabilised burning with a relatively large air to fuel ratio.
The swirl calciner is the most important part in the burning process of the RSP. It
enables efficient combustion near to stoichiometric fuel-air ratio and the raw meaI
which enters the system here is rapidly calcined in the swirling flames. The mixing
chamber is designed to mix effectively the gas/meal mixture from the swirl calciner
with the kiln gas to ensure effective heat exchange.

The original installation with a capacity of 240 tpd was located at Onoda’s Tahara Plant
and was basically a test plant. Subsequently, the design as been built under licence by
Allis-Chalmers and CLE and installed world-wide as with other precalciner systems, the
RSP design is suitable for retrofitting to existing kiln plants and an example of one such
retrofit is shown in Figure 6C - the CLE conversion at Lexos, France

4.6 Mitsubishi Fluidised Bed Precalciners. MFC and N-MC

The Mitsubishi Fluidised Calciner (MFC) is a joint development of Mitsubishi Mining and
Cement and Mitsubishi Heavy Industries. A separately heated fluid bed reactor located
between the kiln and preheater is used to effect calcination. The fluid bed can be
arranged in two alternative ways; the so called ‘overflow system’ is illustrated in Figure
7A and in this system the calcined raw material in the bed overflows directly into the
kiln. In the carry over system material is carried out of the bed by the upward flow of

19
FLOW SHEET FOR RSP SYSTEM ONODA

RAw MEAL

----+ MATERIAL FLOW

— GAS F LOW

: SWIRLBURNER
b

swiRL CALCINER

MIXING/
,’
CHAMBER

ROTIRY KILN

FIGURE 6B RSP_CALCINER

INER

ATURE

11-p[ ● : MATERIAL FLOW

..
TO ROTARY KIL~ FROM KILN HOOD 20
 RSP KILN CONVERSION BY CLE

LEXOS cement plant

1977 1982 1986

Q!@’!l tpd 1500 1700 2400

Rotary kiln m 3,6 X 70

Slope !4 3.5

Power Kw 330

Rotation speed rpm 3

RSP Dia m 3.050

Gas residential time s 1.8 1.6 2.9

Heat consumption kcallkg 795 805 767

FUEL oil coal coal mixture

LHV kcallkg 9600 4800-5300 4900

VM % 15-25 23

Ashes 540 28-34 35-40

Fineness 70 11.5(1 OOmu) 18(90 mu)

Sulfur ?fO 1.8-2 2

Combustion % 96-97 96
efficiency

Figure 6C

21
 fVIFC PRECALCINER

FLOWSHEET OF MITSUBISHI FLU I DISED CALCINER SYSTEM

—.
.
a) Overflow b) Carry Over

. .

T
/?

EEa
1Q) I PrcciDitator
I 01 fluidization blower ~
I

Vll !3

,,

r
* > (7) Waste ~S duct

(8) Secondary
-w f
air
(From clinker ccdcr)
inlet
1 r YI
=(6) Raw
inlet

fluid bed calcalcincr

s(4) Fluidizea

(5) Burners

ation air inkt

(3) Air chamber (From clinker cooler)
+
(2) Fluidization nazzles 22
gases through the bed and into the kiln back end chamber from which it is carried to the
lowest stage cyclone of the preheater where it is separated and fed to the kiln.

Hot air from the cooler is utilised in the fluidised calciner, part being diluted with cold
air to a temperature of 350°C and used to fluidise the contents of the vessel via
fluidisation nozzles in the base. The remainder having a temperature of 500-600°C is
used as secondary air. Effective calcination of the raw meal is achieved using the
additional heat from fuel directly injected into the fluidised bed. One of the main
features claimed for the MFC is the stability of the calcining operation due to the
uniform temperature distribution, usually ~lO°C throughout the bed.

The fuel used in the MFC may be oil or coal or a variety of waste and low grade fuels,
in the case of coal it is claimed that relatively coarse particles can be completely burnt
due to the long retention time on the bed.

The earlier MFC was superseded by the N-MFC during the early 1980’s Figure 7B. The
fluidised bed was redesigned and installed at Higashidani No. 1 kiln and operating data
is shown for this kiln.

The N-MFC has an extended Precalciner body with, in the case of Higashidani, a loading
of approximately 10.9 TPD/m3. This redesign is an attempt to overcome some of the
problems associated with the earlier MFC. In the early fluidised bed design, the control
of the pressure drop across the bed is critical and overfilling of the bed must be
avoided. Whilst overfilling can be protected against by having overflow facilities, this
increases the risk of under-prepared meal entering the kiln. Although the N-MFC
precalciner vessel is extended, it is not certain how much of this additional volume
contributes to the meal residence time. However, this system is claimed to give a high
meal residence time of around 6 seconds. In general terms, the N-MFC system is seen
as a more complex design which does not have the advantages of simplicity of control
compared with other precalciner systems.

23
 FIGURE-7B V/aste gas &
aw material
tR

N-MFC PRECALCINER

Cooler
extraction
Raw
meal
Coal
,$

Fluidized
bed

mwlnedfaadllm

(1412Udl
Cabificvaluedal (N@
6260kaUkgbk#n
5,760Iu#lq iorIritn
-245 nwnA4
SP-IDF
96.1%NrnJ/h
,/\/l1622w -170-

U
M 63.1dNrnJI / 33910 NnWh

N-k. -,,,

Cal
-----
3.15un
‘“w’ ‘3——
Illustration shows the operating conditions of Higashidani kiln No. 1,
24
5. AUTOMATIC CONTROL OF THE PRECALCINER VESSEL

The stability of the Precalciner process is dependent upon the automatic control loops
which adjust the fuel input. The purpose of these loops is as follows:

(i) To maintain constant operating temperature conditions to ensure an even

degree of meal decarbonation.

(ii) To avoid temperature variations induced by changes in the kiln feed rate
and variations in the clinker cooler operation.

(iii) To ensure complete combustion of the fuel.

The most important control loop is the automatic control of the fuel input according to
the precalciner exit gas temperature. For this loop to be stable, it is essential that
there are adequate excess air levels in the precalciner vessel. If incomplete combustion
took place, the exit gas temperature could be low due to Carbon Monoxide production.

A dangerous situation could arise if the falling exit gas temperature automatically
resulted in more fuel being added without sufficient combustion air. Hence, the need
to protect the system against incomplete combustion and incorporate CO alarms and
trip limits within the control system.

Changes in the Tertiary air temperature leaving the cooler will also affect the heat
input to the precalciner. Similarly, as the feed rate to the kiln is increased, this would
initially tend to reduce the precalciner exit gas temperature until the fuel rate
increases automatically to compensate. It is possible to compensate for both these
changes in the control philosophy. However, in general terms, the latter control actions
may only contribute approximately 10% of the total control action. Hence, the most
important loop, with approximately 90% contribution, is the fuel input/exit temperature
control.

Some precalciner kiln systems have based their control philosophy on control of the
meal temperature, leaving the following cyclone stage (i.e. Stage 5 in a 5 stage
preheater). However, this approach is not recommended for the following reasons:

a) The meal temperature may be affected by hot gas short circuiting from
the kiln back end. The risk of this increases at low feed rates or if the
flap valves do not function properly.

b) If poor gas/meal dispersion arises, the meal temperature my be reduced

by mixing with some cooler meal which has dropped through from an
upper stage. Here again, this is more likely at low kiln feed rates when
the gas velocities are lower than normal.

25
6. FACTORS AFFECTING THE OUTPUT OF A-S PRECALCINER KILN SYSTEMS

6.1 General Considerations

It is important to appreciate that a complete A-S precalciner kiln system consists of

several process units which, together, dictate the overall performance which can be
achieved. Although similar kiln designs may be used at different plants, their age and
design requirements may dictate that different limits apply. This is illustrated by the
A-S Precalciner kiln plants at Cauldon and Cemento Melon. Both plants feature single
stream FLS. In Line Calciner (ILC) designs of a similar period:-

6.2 Commrison of FLS Precalciner Kiln Svsterns at Cauldon and Cemento Melon

Works Cauldon Cemento Melon

Kiln No:- 1 9

Date commissioned 1985 1982

Kiln peak daily

output TPD 3100 1850 (stable)

1900 (absolute max)

Kiln shell size 3.95 X 58 3.6 X 49

Kiln shell surface

area inside lining
(M2) 646.9 492.6

Kiln shell loading

based on peak daily
output 4.81 3.86 max

26
Hence the above kiln shell loadings suggest that the Cauldon kiln is achieving
higher loadings than the Cemento Melon kiln. However, this simple approach
overlooks other key process limitations. For example, comparing the Precalciner
vessel loadings shows the following:-

Cauldon Cemento Melon

Precalciner Vessel
Effective volume (M3) 535 190

Precalciner loadin
(TPD clinker/(M f )) 5.81 10.0 max

Gas residence time in

seconds in Precalciner 2.44 1.64

Stage 4 meal - %
decarbonation 93 85

Hence, the Cemento Melon Precalciner vessel is very heavily loaded and does
well to achieve an 85% decarbonation level with a gas residue time of only 1.64
seconds.

Further comparison of the clinker cooler and preheater cyclone loadings shows
the follows:-

Cauldon Cemento Melon

Clinker Cooler
grate area (M2) 74 37.9

Cooler Loading
(TPD clinker/M2) 42.0 50.1 max

Cyclone Stage 2-4

loading (TPD
clinker/M2) 87.1 86.4

The preheater cyclone loadings are high at both plants and this is shown up by
the high preheater pressure losses measured during site surveys:-

Cauldon = 897 rnmwg at 3075 TPD

Cemento Melon = 842 mmwg at 1817 TPD

27
The Cemento Melon clinker cooler is very heavily loaded at 50.1 TPD/M 2 . Despite this
very high loading, the kiln system achieves a high run factor of 95°A and the cooler
efficiency is good at 68%. Other plant limitation with Cemento Melon No. 9 kiln
include the following:-

(a) The Preheater and Electrostatic precipitator f am are heavily overloaded

(the preheater fan runs at 27% over its normal installed power).

(b) Cemento Melon raw materials feature high alkali and sulphur levels and
the preheater system is prone to very heavy build-ups.

(c) The gas circuit system is heavily loaded, but has been uprated by means
of sonic water sprays in the Gas Conditioning Tower.

(d) The kiln feed system etc is operating at its maximum limit. Hence, if we
are seeking to establish realistic output targets for both plants, it would
be unwise to consider the kiln shell loading alone! For example, a simple
summary of the main output limitations would be:-

Cauldon - The high kiln shell loading is partially compensated for by the
generously sized precalciner vessel. Due to the small kiln inlet area,
operation with high kiln excess air levels will tend to increase dust pick
and recirculation around the kiln hearth. Hence improvements to the kiln
burner operation to allow reduced excess air levels without high CO levels
proved beneficial to achieving the maximum kiln output.

Cemento Melon - The very tightly sized precalciner vessel is partially

compensated by the reasonably well sized kiln shell. However, although
the kiln shell target rating is 2150-2265 TPD, this output cannot be
achieved without some major changes, i.e.

(a) Replace/uprate the preheater and electrostatic precipitator (ESP) f am.

(b) Extend the precalciner vessel and modify the Tertiary air duct entry into
the riser duct.

(c) Uprate the kiln feed weigher and replace the airlift by a bucket elevator
to reduce the specific gas volume per kg clinker leaving the preheater by
4%.

(d) Modify the preheater cyclones to avoid excessive gas pressure losses and
higher inleaking air levels over the preheater.

(e) Uprate the main ESP to avoid high dust emissions.

(f) Uprate the clinker cooler using latest cooler technology.

(g) Uprate the ancillary plant e.g. raw milling, coal milling, etc.

28
Hence, the kiln shell limitation is of secondary importance and significance to the above
major plant limitations. As described in the following section, it is proposed to rate
precalciner kiln plants on the basis of the following unit ratings:-

kiln shell
precalciner vessel
-.
preheater cyclone
Clinker cooler (see relevant CTC papers for further details).

7. BCI OUTPUT RATINGS FOR PRECALCINER KILNS

7.1 Introduction

This section considers the target kiln ratings for A-S and AT Precalciner kilns. As
demonstrated earlier, the achievable output from a Precalciner kiln is dependent upon
many factors. In this section, we shall first examine the output targets for the kiln
shell and later consider the Precalciner vessel rating.

It should be appreciated that these ratings are not independent variables. Hence a kiln
operating with a very high Precalciner vessel loading, is likely to achieve a lower level
of decarbonation (e.g. Cemento Melon No. 9 kiln with 85% decarbonation instead of a
target 92% decarbonation) in the meal entering the kiln. Hence, the kiln is less likely
to achieve the maximum kiln shell target rating as it has to do more work on completing
the meal decarbonation.

7.2 Kiln Shell Outmut Ratirms - Air se~arate Precalciner Kilns

The conventional BCI kiln shell rating is based upon the kiln surface area inside the
lining (i.e. kiln shell diameter less 0.4m for refractory) and is as follows:

Air separate precalciner (A-S) = 4.6 TPD/m2

There are several limitations with usin a single rating based upon the kiln shell surface
area, i.e. (i) the rating of 4.6 TPD/m % for A-S systems is exceeded by several kilns
including Rawang, Kanthan and Cauldon. This rating is very cautious for kilns above 4m
shell diameter.

Table 1 and Figure 8 show the relationship between the kiln shell rating (TPD clinker/m2
versus kiln shell diameter (inside steel in this case).

29
 TABLE 1

AIR-SEPARATE PRECALCINER KILN RATINGS-1

Kiln Shell Kiln Shell Kiln Shell Kiln Shall

works TPD Clinker Kiln Diameter Kiln Length Kiln tJD Surface Area Surface Rating Internal Volume Volume Rating
-Kiln No PC Tym ~e* ~ ~ !@@ UB [m2] I!?Q!EU l/B (m3] TPDIM3

Rawang No.3 NSF 5070 4.7 74 17.21 999.7 5.07 1074.6 4.72
Kedah No.1 NSF 5000 4.7 74 17.21 999.7 5.00 1074.6 4.65
Cauldon ILC 3110 3.95 58 16.34 646.0 4.8f 574.1 5.42
Kanthan No.3 PYROCLON 3350 4.2 57 15.00 680.5 4.92 646,4 5.18
Melon No.9 ILC 1900 3.6 49 15.31 492.6 3.86 394.1 402
/%albor&Rordal SLGS 5500 4.75 74 17,01 1011.3 5.44 1099.8 5.00
Tonasa-Sulawesi SLC-S 7500 5.5 87 17.06 1393.9 5.38 1777.3 4.22
Hoang Thach SLC-S 3300 4.15 64 17.07 754.0 4.38 706.9 467
Siarn-Tabkwan~No.5 SLC 9000 6.0 96 17,14 1688.9 5.33 2364.5 3.81
Siam-Khao Wong SLC 10000 6.0 106 18.93 1864.8 5.36 2610.8 3.83
Gujamt SLC 6000 4.75 75 17.24 1024.9 5.85 1114.6 538
Siam-Tabfa 2 SLC 5500 5.0 80 17.39 1156.1 4.76 1329.5 4.14
Adelaide Brighton SLC 4150 4.57 67 10.07 877.7 4.73 915.0 4.54
Sikka SLC 2300 3.75 75 22.39 789.3 2.91 661.1 348
Ibuki NSF 3045 3.8 79 23.24 843.8 3.81 717.3 4.25
Tochlgi NSF 2640 3.5 66 21.29 842.8 4.4i 498.1 530
Akoh No.2 SF 7600 5.6 94 18.08 1535.6 4.95 1996.3 3.81
Chlchibu No.1 CSF 2400 3.9 54.8 15.66 602.6 3.%8 527.2 4.55
Hachlnode Scs 4500 4.7 75 17.44 1013.2 4.44 1089.2 413
Akoh No.1 Scs 5000 4.7 81 10.84 1094.2 4.57 1176.3 4.25
Gifu KSV 2724 4.1 60 18.22 697.4 3.$1 645.1 4.22
Isa KSV 9900 6.2 105 lB,IO 1913.2 5.17 2774.2 3.57
Saitama DD 5500 5.0 83.1 l&07 1200.9 4.58 1381.0 3.98
Kawara MFC 7200 5.4 95 10.00 1492.3 4.82 1865.3 3.86
Railton PYROCLON 3000 4.0 58 lB.11 056,0 4.57 590.4 5.08
Sangwon PYROCLON 3300 4.8 52 11.82 718.8 4.50 790.7 4.17
RudersdorfN0,5 PYROCLON 5000 5.2 61 12.71 919.9 5.44 1103,8 4.53
Perils PYROCLON 2700 4.0 56 15.56 833,3 4.26 570.0 4.74
Spenner PYROCLON 2136 4.08 40 10.86 462,7 4.62 425.9 5.02
Gtasim PREPOL 3300 4.4 65 15.25 816,8 4.04 816.8 4.04
Merone PREPOL 2510 4.0 54 1:5,00 610.7 4.i’t 549.7 4.57
Danyang PREPOL-CC 6000 5.2 78 113.25 1176.2 5.10 1411.5 4.25
Tong Yang PREPOL-CC 7200 5.6 87 lm73 1421.3 5.07 1847.6 3.90
TPI Saraburi PREPOL-CC 7500 5.6 87 10.73 1421.3 5.28 1847.6 4.06
Lexos CLE-RSP 2400 3.6 60.75 21.80 701.2 3.42 561.0 4,28
 EiQL!BE_B
.—..
—
_——. ——
———
—.
~! KILN SHELL TPD/M2 VERSUS DIAMETER1
71m

TPD/M2=2.6+0.556* Diameter

/
5

:B
❑
❑
❑
4 ❑

El

—A— —
3
3.25 3.75 4.25 4.75 5.25 5.75 6.25 6.75

d
Kiln Shell Diameter (1/S) m

__—. — __— —— _. —.——————

-——
The following should be noted:

(i) Several of the kilns listed exceed the rating of 4.6 TPD/m2. The larger diameter
kilns tend to have the highest shell ratings.

(ii) Plant suppliers tend to size the kiln shell on a volumetric basis (i.e. TPD/m3 of
internal volume inside refractory). Hence, the larger kilns tend to be designed
with a higher shell surface area rating.

(iii) Kilns such as the KHD Pyrorapid design use a short kiln concept with a typical
length/diameter ratio of 10-13. These kilns tend to have high shell loadi~gs, i.e.
RudersdorfNo.5 = 5.44 TPD/m2

The Hualien No.3 kiln (5.0x 55m) with an output of 5300 TPD peak has achieved
a shell loading of 6.67 TPD/m2. This is the highest known shell loading. Such
high figures tend to imply that kiln shell heat transfer limitations may not be the
most important factor affecting AS kiln outputs and gas velocity limits must also
be considered.

Whilst the short kiln concept can achieve very high shell loadings, it would be
unwise to apply the above ratings on conventional A-S Precalciner Kilns with an
LID ratio of 15-18. The short kiln concept can have some limitations when used
with materials with dficuh combinability characteristics. Build ups and high kiln
exit temperatures can also be problem areas.

(iv) Hence, for conventional A-S Precalciner Kilns, a simplified relationship could be
used which relates the kiln output to kiln diameter, i.e.

Kiln Shell Output rating :

TPD/m2 = 2.6 + 0.556 D

Where D = kiln diameter inside steel

TPD/m2 = Rating applied to the kiln surface area inside refractory.

(v) There are some limitations with applying the above output ratings:

a) Whh small diameter kilns, the effect of kiln coating upon the total sutiace
area for heat transfer will be more pronounced than on a larger kiln.

b) If the plant suppliers design the kiln shells on a volumetric loading basis,
the smaller dhmeter kilns may not have been “pushed” to their maximum
output potential. Hence, the limits on kiln output may be dictated by
ancillary plant limits (see example of Cemento Melon No.9 Kiln
comparison in Section 6.2)

32
 c) There is a risk with using a single rating factor such as 4.6 TPD/m2 for all
kiln sizes when larger kilns can clearly exceed this rating. Similarly, this
target rating may not be achievable with smaller diameter kilns due to
coating, as mentioned above.

(vi) The higher the kiln shell rating achieved by a kilm the greater the risk is of the kiln
being limited by gas velocity constraints on the kiln hearth. This aspect is
reviewed in Section 7.4 and it has a bearing on the higher loadings achieved by
larger diameter kilns.

(vii) Hence in summary, it is recommended that both of the above rating figures should
be compared for a given kiln system. The kiln shell ratings should then be
compared with the ratings for the precalciner vessel, the preheater and clinker
cooler before setting any targets.

For small diameter kilns (i.e. 3.6 to 4.Om inside shell diameter), beware of using
both the formula and 4.6 TPD/m2 if the kiln has heavy coating characteristics. In
additio~ check the kiln hearth area and gas velocity limitations as discussed in
Section 7.4.

For kilns above 4.Om diameter inside shell, the “safe” loading is 4.6 TPD/m2.
However, this will give an underestimate of the potential kiln capacity as given by
the formula. Hence the formula will tend to set a maximum figure for the kiln
shell rating.

7.3 Kiln Shell Outwt Ratinm - Air Through Precalciner Kilns

Table 2 and Figure 9 summarise the output and kiln dimensions of several FLS and Polysius A.T.
systems. The following points are noted:

(i) The A.T. output ratings are generally lower than the equivalent A-S kiln shell
ratings for a similar sized kiln.

(ii) Here agai~ the highest shell loadings tend to be achieved by the larger diameter
kilns which reflects the plant suppliers tendency to size on kiln internal volume and
not kiln internal surface area.

(iii) The highest rating is achieved by Ssang Yong Cement Donghae plant at 3.98
TPD/m2 for a 5.2m diameter kiln.

(iv) The Table does not indicate other plant limitations such as preheater size, clinker
cooler size, ancillary plant sizing, etc.

33
 IAIBLE-2

I AIR-THROUGH PRECALG,AER KILN RATII=

Kiln Shell Kiln Shen Kiln Sheii Kiin Sheii

works TPD Clinker Kiln Diameter Kiln Length Kiln LID SurfaceArea Surface Rating intemai Voium~ Voiume Rating
Kih!!kL ec TvRe &* m ❑ IeJ2!! !/B (m3] Ifww

El Atto ILC-E 3500 5,25 86 17.73 1310.4 2.67 1588.8 220

Setubal ILC-E 3500 5.25 83 17.11 1264.8 2,77 1533.4 2.28
Pedro Leopoldo ILC-E 2000 5.25 83 17.11 1264.6 2.29 1533.4 1.89
Zapotlitic iLC-E 2500 4.75 75 17.24 1024.9 2.44 1114.6 224
Castiilejo de Anover ILC-E 2200 4.55 69 16.83 899.6 2.45 933.3 2.36
Merida iLC-E 2200 4.55 68 16.39 886,6 2.48 919.8 2.39
Limerick ILC-E 2100 4,75 17.24 1024.0 2.05 1114.6 1.88
Guangzhou ILC-E 2000 4.55 ;: 17.11 925.7 2.16 960.4 2.08
Pioneer ILC-E 2000 4.35 16.96 831.4 2.41 821.0 2.44
Rlo Bijao iLC-E 2000 4.55 :: 16.38 880.6 2.26 919.8 2.17
La Ccmronne PREPOL-AT 3000 4.6 75 17.86 989.6 3.03 1039.1 2.89
Beocin PREPOL-AT 3000 4.8 60 13,84 820.4 3.62 012.3 3.29
Hanll PREPOL-AT 3000 4.8 60 13.64 829.4 3.82 912.3 3.29
Atien PREPOL-AT 2200 4.0 65 18.06 735.1 2.99 681.6 3.33
Dartca PREPOL-AT 3000 4,8 70 16.67 623.6 3,25 069.8 3.09
SSang Yon*Seoui PREPOL-AT 4200 5.2 70 14.58 1055.6 3.08 1266.7 3.32
Partizan PREPOL-AT 3200 4.6 16.67 923.0 3.48 969.8 3.30
Sllte PREPOL-AT 4700 5.2 i% 16.67 1206.4 3.00 1447.6 325
Tabligbo PREPOL-AT 2400 4.8 85 19.32 1175.0 2.04 1292.5 1.86
Dodam PREPOL-AT 3000 4.8 60 13.64 829.4 3.62 912.3 3.29
Luong PREPOL-AT 3000 4.8 60 13,64 829.4 3.82 9123 3.29
Recife PREPOL-AT 1500 4.0 60 10.67 878.6 2.21 610.7 2.46
Palmeiras de Golas PREPOL-AT 800 3,6 16.88 542.9 4.47 434.3 1.84
Monay PREPOL-AT 1800 4.2 : ‘ 17.11 776.0 2.32 737.2 2.44
SSang Yong-Donghae PREPOL-AT 4200 5.2 70 14.58 1055.6 3;98 1266.7 3.32
Souseias PREPOL-AT 3500 5.0 75 16.30 1083.8 3.23 1246.4 2.81
Loma Negra-Catamarca PREPOL-AT 1600 4.4 70 17.50 870.6. 1.82 8706 182
Balconea PREPOL-AT 2500 4.0 74 17.62 070.4 2.56 10252 244
Cantagalo PREPOL-AT 1500 4.0 60 16.67 678.6 2.21 610.7 2.40
Onaba PREPOL-AT 2400 4.4 65 16.25 810.8 2.94 816.8 2.94
LlchtenburgNo 6 (Phase 1) PREPOL-AT 1300 4,2 63 16,58 752.1 1.73 714.5 182
Pedro Leopoldo PREPOL-AT 2200 4,4 65 16.25 816.8 2.69 816.8 2.69
Perkasa PREPOL-AT 4750 5.4 78 15.60 1225.2 3.88 1531.5 310
Jura-Wlldegg PREPOL-AT 1800 3.8 56 16.47 598.2 3.01 508.4 354
Safi PREPOL-AT 4.2 60 15.79 716.3 2.23 680.5 2.35

L-A-l
4=
 u)
\
\
m
Ui
1/)1
131 \\
u)
K I ‘\
\
■
99
\
\
99
\
\ m
\
❑ ■
\
m
m
35
 (v) The approximate relationship between kiln output rating and kiln diameter
is as follows:

Kiln shell output rating:

TPD/m2 = 0.9 + 0.556D

Where D = Kiln diameter inside steel

TPD/mz = Rating applied to the kiln surface area inside the

refractory.

(vi) The equivalent BCI rating based upon a fixed factor would be
approximately 3.5 TPD/m2 for kilns in the mid size range of around 4.7m
diameter.

(vii) As in the case of the A-S systems, both of these ratings have limitations
and caution should be exercised when applying them to small and large
diameter kilns.

(viii) Gas velocities at the kiln hearth will have a greater influence on the
perf orrnance of AT Precalciner Kilns. This is due to the higher specific
gasflow at the kiln inlet due to the need to run with higher kiln excess air
levels. This aspect is reviewed next.

(ix) A further ef feet of operation with high excess air levels, is the reduction
of the flame temperature and hence the rate of heat transfer in the
burning zone. This limits the maximum output rating of AT systems.

7.4 Kiln Hearth Gas Velocitv Considerations

In the conventional Preheater kiln process, the gas velocity at the Kiln Hearth often
represents a major limitation to increased output. Higher gas velocities at the hearth
tend to result in increased dust pick up and recirculation around the lower cyclone
stages. In addition, preheater gas temperatures can increase as the dust carries
additional heat into the preheater system.

Normally, gas velocities at the kiln hearth are less of a limiting factor in the A-S
Precalciner kiln system due to the reduced fuel burned (40% of total) in the kiln.
However, the comparison of the Cauldon and Cemento Melon kilns, in Section 5.2,
indicated that the Cauldon kiln output can be limited by high gas velocities at the
hearth. In certain cases (e.g. Lichtenburg No.6 and Rawang No.3 kilns), there has been
a tendency to run the kiln hearth at higher gas velocities in order to ensure adequate
meal pick up in the riser ducts, etc. However, this method of operation can limit the
maximum output and cause higher gas pressure losses.

36
There is no fixed limit to the gas velocity at the kiln hearth. The critical velocity will depend upon
several factors, including

Kiln hearth design - e.g. chute and down pipe angles and the transition from
downpipe to hearth and hence kiln shell.

Raw meal fineness and gas temperatures profile.

Degree of buildup present, kiln volatile cycles and cleaning facilities available.

As a guideline, kiln hearth gas velocities in the region of 25-30 m/s are critical to the degree of
dust recirculation.

It is worth looking at the area of the hearth in relation to the kiln diameter by comparing the data
for the Cauldo~ Cemento Melo% Kanthan and Rawang Kilns. This is shown in Table 3 and
Figure 10.

At peak outputs the gas velocities are approaching 30 m/s at Cauldon. The Rawang kiln has the
largest availableflow area in relation to the kiln cross sectional area. However, this kiln tends to
operate at higher back end oxygen levels and a non typical kiln/precalciner fiel split, which
complicates direct comparison. Minimizingthe kiln hearth velocity is a critical area if the Cauldon
kiln output is to be maximised. Kiln capacity was increased to 3100 TPD after improving the fine
coal conveying system to the kiln to achieve adequate pickup velocity. Running with lower kiln ~
excess air levels was then possible without high CO levels. This in turn helped to minimise hearth
velocities and maximise output.

For rating purposes, with A.S. Precalciner Kilns a practical limit of 30 nds is recommended. It
is possible to operate at higher velocities thought this will tend to result in higher gas pressure
losses and higher preheater temperature profiles. The highest reported kiln hearth velocity is
around 40 nis for a preheater kiln with high recirculation around the kiln hearth. Please refer to
the CTC paper on kiln output targets concerning hearth velocities for AS and AT Precalciner
Kilns.

The above illustrates the range of kiln hearth velocities found with A-S systems. Obviously kiln
hearth velocity constraints are more relevant to the A.T. systems, which rely upon overdrafling
the kiln to provide combustion air for the precalciner fbel. Table 4 compares the output versus
gas velocity limitations for similar sized A-S and AT Precalciner Kilns

7.5 Precalciner Vessel Ratinm

Table 5 shows the sizes of Precalciner vessels used with different kiln systems. This shows
precalciner loadings within the range of 16.5 to 4.7 TPD clinker per m3 of effective Precalciner
volume. The effkctive volume is defined as the volume between the meal entry and the exit to the
next cyclone stage. The highest loaded Precalciner vessel listed in Table 5 is the R. S.P. No.3 kiln
at Omi, Japan. However, some of the effective volume data given is as reported by the plant
suppliers. This may not comply with the definition given above.

37
 1[
]TABLE3-AIR sEpARATE PRECALCINER KILNS-KILNHEARTHAREAs

CAULIXM!I l@I!41Mr!4

Kiln Diameter inside 4.7

3.6 3.95 4.2
shell at back end (m)

Cross sectional area of 11.34 14.52

8.04 9.90
Kiln inside lining (mz)

Kiln Hearth area inside 5.08

2.31 2.94 3.64
lining (m2)

Kiln Hearth area as a

percentage of the Kiln 34.98
28.72 29.70 32.10
cross sectional area (“A)

3110 3350 5070

Peak Clinker output TPD 1900

Actual gas velocity at- 30.7

22.1 29.S 28.7
Kiln Hearth (m/s)

Kiln Target Peak output 5212

2267 3103 3358
using formula (TPD)
Note:- this is Kiln Shell
Target only and ignores
other plant limits

w Output as ‘A of peak kiln

m 97.3
‘83.8 100.2 99.7
shell target rating
 FIGURE 10

Kiln Diameter and Hearth Area

6

5 .-m

,/ /“
-,/.
/“”
/
I

4 \

m’-’
3

2 1 I I 1 I

3.4 3.6 3.8 4 4.2 4.4 4.6 4.8

I

Kiln Diameter at Back End (l/S)-m

I

L
Kiln Hearth Area as% of Kiln C.S.Afl
J
36

/m
34
/
//”
,/”

/ ,/”
,/
,/ /
32
/m
//-’

30

28
[-----’’””””
I

3.6
1

3.95 4.2 4.7

Kiln Diameter at Back End (l/S)-m
20
 TABLE4 AIR SEPARATEAND AIR THROUGH PRECALCINER COMPARISON

Kiln Size:- 4.7m Diameter *74 m]Long Surface Area (mZ)= 999.66
Maximum Kiln Hearth Area is assumed= 6.19 m2 (minimum area is = S.08 m2 )
Assume common 720 net Kcal/Kg Clinker fuel consumption for simplicity

Precalciner AIR AIR

nm IIW!!&W PARAIE

Kiln Output Rating (TPD) 3500 5200

Fuel Split (Yo)

Kiln Fuel 65 40
Precalciner Fuel 36 60

Kiln Fuel-net Kcal/Kg Clinker 468 288

Kiln Back End % Oxygen 7 2

Kiln Back End Gas Flow

Nm3/Kg Clinker 0.934 0.434

Kiln Back End Gas Velocity

mls 31 20

NOTE: -Kiln Hearth velocity is more critical to A.T. Kiln output

Excess air levels with A.T. Kilns has adverse effect upon flame
temperature and heat transfer in burning zone.

4=
I o
 Precalciner Precalciner
Precalciner Peak Vessel Effective Vessel Loading
Kilnplan! W@Xss GIMIJQXIJ?D W~a@3) IPQ/lm3

Rawang No.3 SF 4150 350 11.9

Rawang No.3- Uprated NSF 5070 740 6.9
Kedah NSF 5000 476 10.5
Ibuki NSF 3000 580 5.2
Nanyou No.5 NSF 7100 780 9.1
Chichibu No.2 Plant K DI SF-Oil fired 5000 350 14.3
SF-Coal fired 4600 350 13.1
Edmonton SF 2575 270 9.5
Chichibu No.2 Plant CSF 5300 450 11.8
Chichibu No.1 Plant K6 CSF 2400 260 9.2
Lichtenburg No.6 PREPOL 3200 470 6.8
Dunbar PREPOL 3000 445 6.7
Isa No.2 KSV 9900 890 11.1
Higashidani N-MFC 2050 170 12.1
Omi No.3 RSP-Oil fired 3886 236 16.5
RSP-Coal fired 3849 236 16.3
Banda Aceh DD 3300 346 9.5
Kamiiso No.6 DD 3960 355 11.2
Saitama DD 5500 730 7.5
Rashadiya DD 3400 483 7.0
Kanthan No.3 PYROCLON 2700 478 5.6
Kanthan No.3- Uprated PYROCLON 3350 478 7.0
Malkhed PYROCLON 1500 320 4.7
Davenport ILC 2360 440 5.4
Tagawa-No.l ILC 4360 520 8.4
Cauldon ILC 3110 535 5.8
Melon No.9 ILC 1900 190 10.0
$= Hermosillo ILC 2760 415 6.7
Of interest is the lower ratings achieved at Omi and Chichibu No.2 plant, Kiln Dl, on the
changeover from oil to coal firing.

Whilst the effective Precalciner volume can give an indication of the gas residence time in the
Precalciner, it does not indicate the meal residence time. For example, in a fluidised bed system
such as the Mitsubishi N-MFC Precalciner, the gas residence time may be around 1-2 seconds.
However, the quoted meal residence time is higher at around 6 seconds. Similarly, in the Spouted
bed designs, the meal residence time will tend to exceed the mean gas residence time.

In the SF process such as Rawang No.3 Kilq a Precalciner loading of 11.9 TPD/m3 corresponded
to approximately 1.1 seconds gas residence time, with a decarbonation level of 75-80Y0. In this
syste~ the meal residence time will be close to the gas residence time and this high loading
contributed to the lower meal decarbonation levels achieved. Hence, the Precalciner volume
loading (TPD/m3) and gas residence time parameters are not an absolute guide to a Precalciner
vessel’s performance. Since the meal residence time will vary between the different designs, it is
obvious that certain designs will be able to achieve higher specific loadings than others.

AS a guideline to the sizing of Precalciner vessels, the following figures are recommended for new
lines and upratings:

a) Maximum design Precalciner vessel loading:

6 TPD/m3 effective volume for Extended Riser Duct type Precalciners

7 TPD/m3 effective volume for Spouted Bed type Precalciner.

b) These ratings are based upon the Peak Clinker output. They are based upon an
acceptable loading criteria.

c) Ideally, the Precalciner vessel should be more generously sized provided that this
does not result in too low a gas velocity profile and meal entrainment problems.
A minimum gas residence time of 2 seconds is recommended. This allows some
design margins for

Uprating at a later date

Possible use of low grade fhels

d) It is considered a ftdse economy to undersize the Precalciner vessel in order to

reduce capital. This can prove costly to modifi at a later date if it is planned to
uprate the kiln capacity, e.g. Rawang No.3 Kiln example.

e) The Precalciner size must be considered in conjunction with the kiln shell sizing.
A tightly sized Precalciner vessel with a tightly sized kiln shell may reduce the
capital cost but is less likely to achieve the rated output on a consistent basis.

42
The above loadings assume that the Precalciner vessel operates at a typical gas temperature of
around 870-890 average 880”C. It is possible to compensate for an undersized vessel by
increasing the operating temperature to 900°C or slightly above. However, whilst higher
operating temperatures can partly offset the effects of low gas/meal residence time, this results
in less efficient operation. The higher precalciner exit gas temperature will increase the
temperature profde across the preheater systerq leading to higher exit gas heat losses. The higher
operating temperature also increases the risk of build ups and blockages.

A grett many claims have been made for different Precalciner systems allowing operation at lower
temperature% e.g. 860”Cinstead of 880”C. In practise, too low an operating temperature reduces
the decarbonation level unless the system design allows a high meal residence time. Hence, some
of the claims for low operating temperature Precalciners have not been proven in practical
operation.

In certain cases, the lower precalciner operating temperature maybe possible if the system is
matched with an oversized kiln shell. The generously sized kiln shell can then compensate for the
lower decarbonation level leaving the Precalciner vessel. This feature has been noted with some
of the Japanese Precalciner kilns which result from conversions of Wet/Lepol processes.
Alternatively, being earlier installations, the kiln shell was generous sized as the current loadings
were not foreseen. This fbrther emphasises the need to look at the Precalciner vessel and kiln
shell loadings together and not as separate, independent parameters.

7.6 The Effect of Fuel Oualitv U~on Precalciner Performance

Much of the earlier Precalciner kiln system developments took place in Japan and many of these
earlier systems were designed primarily for oil firing. The changeover from oil to coal firing
showed up several limitations with the earlier designs i.e. inadequate residence time to allow
complete ‘burn out of the fiel. The use of lower &ades of fiel &o demonstrated the need to
increase the size of the precalciner vessel to improve gas/mea.l/fiel residence times.

43
This relationship was demonstrated by the experiences with Rawang No.3 Kiln system.
This kiln used a conventional twin S.F. precalciner vessel design by I.H.I. (see Section
3.4). The original design basis was as follows:

Clinker guaranteed output = 4000 TPD

Fuels used: Kiln - Oil

Precaiciner - Shale and oil

Precaiciner vessel effective volume = 350 m3

Precalciner loading = 4000 = 11043 TPD/m3

350

This loading was high but had been exceeded by other oil fired SF processes.

However, the Rawang process fires Shale containing wound 6% Carbon into the
precalciner vessel. This low grade fuel dictates the precalciner performance and there
were many problems with poor fuel burn out, inadequate combustion, poor ‘air/fuel
mixing, low and variable decarbonation levels, etc. Whilst designed for at least 85%
meal decarbonation, the SF system gave oniy 75% with a high degree of variability.

Kiln output was further adversely affected when the kiln firing system was coqverted
from oil to coal firing.

Figure 11 shows the Precalciner ratings of thr~’ IHI-SF processba and illustrates the
lower loading which can be achieved with lower quality fuels. This figure is a
simplification of the data and other kiln limitations did affect on the overall kiln
performance. As noted earlier, it is possible to compensate for poor precalciner ve&el
performance if a generously sized kiln shell is used, as in the case of the kilns quoted
in Figure 11.

In order to overcome the limitations affecting Rawang No.3 Kiln, as well as allow the
output to be increased to 5000 TPD, the following changes were made:

The twin SF precalciner vessels were modified using a “Trunk Duct” to give a
single precalciner vessel of 740 m3 volume. (see Figure 5)

This reduced the design precalciner loading to

5000 = 6.76 TPD/m3

740

Two new Preheater fans were installed.

Preheater modifications were made to the Stage 2-5 cyclones and four
new Stage 1 cyclones were installed.

44
 WUBE 11-THE EFFECT OF FUEL QUA~Y UPON SF PR ECALCllY~!NGs

. FLASH FURNACE LOADING (TPD CLINKERfm3 FURNACE vQL~EJ

VERSUS NETT FUEL CALORIFIC VALUE “

15

CHICHIBU 011-

14

CHICHIBU COAL
13

12

11
./ RAWANG OIL/SHALE

u a I
I
10 [ I 6 t

7 8 9 10
3 4 5 6
1 2
NETT FUEL CALORIFIC
VALUE (X 10? KCALS/KG)
 The coal firing system was converted from Direct to Indirect firing and provisions ~
were made to burn coal in the precalciner vessel.

Environmental improvements included a new Electrostatic Precipitator,

modifications to the Gas Condhioning Tower, additional Gravel Bed filter units.

The Clinker Cooler was uprated with later conversion to Coolax design.

Whilst this conversion has had some problem areas, it has been improved gradually and the kiln
is currently producing record outputs at high specific loadlngs, i.e. up to 5,200 TPD.

The Rawang uprating exercise demonstrates the following points:

(i) Iflow grade fhels are likely to be used in the process - ensure that the PreCalCiner,
vessel is generously sized.

(ii) The use of a single precalciner vessel is preferable to twin vessels in order to avoid
gas/meal/fiel mixing and balancing problems.

8. ~E “BEST” PRECALCINER KILN SYSTEM

A frequently asked question is “which is the best precalciner kiln system?” To answer this, the
“best” system available is the one which combines the following characteristics:-

(i) When selecting anew Precalciner kiln plant from several alternative tenders, the
“best” system is that which is the most economical in terms of capital cost per
tonne clinker produced. Differences between fiel and power consumption are
usually of secondary importance to rninimising investment costs. Hence, the
“best” technical solution may not be the “best” commercial solution.

(ii) The preferred system should feature the following design requirements, as
described within this paper:-

The gas and raw meal residence time must be sufficient to achieve a high
degree of meal decarbonation i.e. around 92%. A minimum gas residence
time of 2 seconds is recommended.

If the process utilises low grade fiels, then the gas residence time maybe
increased to ensure complete bum-out without CO.

46
 Avoid designs which give insufficient meal dispersion in the gas
stream due to too low a gas velocity or the lack of a dispersion
device (e.g. splash plates, gas accelerators, dispersion boxes etc).

Avoid designs which require a high operating temperature to

achieve adequate fuel burn-out (i.e. 900° plus)

(iv) The design should be of a proven reliable type - check the tenderers
experience base for similar plant references.

(v) The preheater system must use a proven cyclone design. Check the
preheater gas velocity profiles and predict cyclone pressure losses. Check
for areas of low gas velocity without dispersion devices which could allow
raw meal to bypass critical areas.

(vi) The precalciner must feature full automatic control of the precalciner
exit gas temperature. This should control the precalciner fuel input(s) and
include the features described in Section 4.

(vii) Current BCTC preference is for the Spouted Bed design instead of the
Extended Riser Duct design. The Spouted Bed design generally tends to
achieve a higher output loading due to the meal recirculation
characteristics. Notwithstanding this, the Extended Riser Duct design can
be equally effective provided attention is paid to adequate meal, gas and
fuel distribution/mixing.
(vii) Preference is given to these designs using a single precalciner vessel
instead of twin units. This is to avoid uneven fuel, gas, meal distribution
in twin stream precalciner vessels with outputs above 3,400 TPD clinker.

(ix) Evaluate the precalciner kiln process as a whole with due reference to the
BCTC output ratings described in this paper. It is considered to be a false
economy to undersize either the kiln shell of precalciner vessel size.
These can prove costly or impractical to rectify at a future date if kiln
uprating is required.

47
9. PR.ECALCINER KILN UPRATINGS - CASE STUDIES

9.1 General

The following sections describe two examples of precalciner kiln upratings carried out
in recent years. The first example; Cemento Melon No. 9 kiln, demonstrates how
uprating can be achieved, in certain cases, without high capital expenditure. This
represents one of the lowest cost kiln upratings carried out within the BCI group.

The second example, Kanthan No. 3 kiln, is a more typical example where more
extensive modifications were needed to overcome plant bottlenecks.

9.2 Low Cost Um-atinE - Cemento Melon No. 9 Kiln

No. 9 kiln was commissioned in 1982 as a 1290 TPD as a precalciner kiln with ‘a 30%
Bleed for sulphur. Commissioning of the kiln subsequently showed that the Bleed was
not required as some 45% of the sulphur was present in the form cf sulphide and was
lost in the preheater exhaust gases.

The main limitations to achieving higher clinker outputs were found to be as follows:-

(i) Inadequate raw meal conveying capacity from the blending silos to the
kiln feed weigher.

(ii) Higher outputs produced problems with high dust emissions and blockages
of the Gas Conditioning Tower (GCT).

(iii) Blockages of the Stage 4 cyclone downpipe and kiln inlet.

(iv) Poor clinker cooler performance and kiln/cooler dust recycle.

(v) Poor kiln burner flame characteristics.

To overcome these limitations, the following changes were made (see figure 12):

(i) The pneumatic conveying line between the silos and the kiln feed weigher
was uprated by removing bends, modifying pipework etc. The kiln feed
airlift was also uprated.

(ii) The spill back water spray nozzles in the GCT were replaced by Sonic
Water Sprays. This allowed the GCT exit temperature to be reduced to
around 140°C. The reduction in operating temperature has allowed the
very tightly sized electrostatic precipitator (ESP) to achieve acceptable
outlet dust emissions without the need for extensive upgrading.

48
 w
E!
f
.--— -— h
I
I
I
-J
49
f
t
 (iii) The Stage 4 cyclone downpipe was modified by removing the bend in it. Air
cannons were installed at critical areas of the kiln back end riser duct. Additional
access points were provided to improve the preheater cleaning.

(iv) An additional cooler fan was installed and the cooler quench fan was uprated. The
cooler roof was raised in order to reduce the cooler throat air velocity.

(v) The kiln burner pipe was modified to increase the flame momentum. The primary
air fan was uprated to avoid the previous unstable operating point which caused
an unsuitable flame.

These were the main areas of modifications. Minor improvements were made as and where
necessary to the ancillary plant. In addhiou there is a high motivation to “push” output at
Cemento Melon. Similarly, great emphasis has been placed upon ma.ximisingkiln run times up
to 95.8% currently (1994 annual average).

The benefits of this uprating and the other operational improvements areas follows:- ~

(a) For simplicity, it has been assumed that a common kiln run time of 95% applied
before and after. In practise this high run time has been achieved by ongoing
improvements:-

BEFORE AFTER

Clinker Peak Output (TPD) = 1,300 1,850

Clinker annual average

output at 950/0run time
and 0.95 average/peak
output ratio = 428,240 609,410

Capital cost in US$ = 418,000

Increased output TPA clinker = 181,170

Output increase = 42.3%

Capital cost per Tonne extra
clinker = 2.31 US$/tonne

It should be appreciated that the ancillary plant at Cemento Melon was reasonably well sized to
support the additional output. However, on-going improvements to the raw milling and coal
milling plant allowed these units to keep up with higher clinker outputs from No. 9 and No. 8
kilns. Hence, raw mill outputs were increased from 65 to 80 tph.

50
The capital cost per tonne for this uprating is extremely low and compares with
investments costs of around 130-170 US$/tome clinker for a similar sized new kiln line
with all ancillary plant. The uprating was possible because the key plant items were
reasonably well sized. New kiln plants which are sized on the basis of minimum
investment cost/tonne may not have a similar design margin for such large output
increases.

A more typical uprating exercise is shown in the following example.

9.3 Conventional Plant Umatin~ - Kanthan No. 3 Kiln

9.3.1 I.nlmduction

Kanthan No. 3 kiln was commissioned in 1984-5 and achieved an output of 2,500
TPD clinker. The A-S Precalciner kiln system included a Pyroclon Precalciner
of KHD d~ign. During 1990, feasibility studies were carried out which examined
how the kiln capacity could be uprated. The BCI target rating for the 4.2x 57m
kiln at this time was 3250 TPD and this was used as the basis for the Tender
enquiry. The Tender enquiry specified the areas of the kiln system where
modifications were required to improve output/fuel consumption. Surveys of the
kiln system were carried out by APMC/BCI and the following limitations were
noted:-

9.3.2 Preconversion - Plant Limitations

(i) Kiln operation - the poor performance of the kiln enlarged hood and
tertiary air duct system resulted in the duct being partially blocked. This
meant that the kiln had to operate with higher excess air levels to provide
combustion air in the precalciner. Thus, the kiln was operating between
A-S and AT modes of operation. For maximum output, it was essential
to achieve full Air Separate Precalciner operation.

(ii) The preheater pressure drop survey showed the following:-

There was a high gas pressure loss over the duct from the Stage 1
cyclones to the preheater fan.

The pressure loss here was around 8.5 mbar (87 mmwg) compared
with an expected 3 to 4 mbar.

The pressure loss over the Precalciner (Pyroclon) exit and Stage 4
cyclone inlet was comparatively high.

51
 Inleaking air levels over the preheater system, as indicated by the
gas oxygen content was comparatively low. However, the kiln exit
oxygen was high (5%). Due to the Tertiary air duct problems, the
kiln had to be run with higher excess air levels to provide
combustion air to the precalciner.

The total gas pressure losses over the Stage 1 cyclone and Stage
2 cyclone outlet was high and would increase rapidly if output was
increased.

(iii) Other plant bottlenecks included the following areas:-

Clinker cooler fan capacity.

Cooler exhaust system - Gravel Bed Filter size.
Raw milling capacity.
Gas Conditioning Tower and Main Electrostatic precipitator
capacity.

These areas were investigated and provisions were made for their uprating
within the overall uprating proposals.

9.3.3. Kiln Uprating Modifications

(i)

To overcome the Tertiary air duct problems, a new dedusting system was
installed. This allowed the clinker dust to be settled out before entering
the Tertiary air duct.

(ii) The preheater was modified as follows:

Sta~e 1 cyclones - Two new 5.Om diameter Low Pressure drop cyclones
of KHD design were installed. These replaced the existing twin 4.4m
diarnetw cyclones of conventional design.

Duct to Preheater Fan

A new preheater of 2250 kw replaced the existing 1400 kw fan. To avoid

high gas pressure losses, new ductwork was installed between the Stage
1 cyclones and the fan. The new fan was also located at ground level
instead of in the tower smcture. This allowed a smoother ductwork
arrangement into and out of the new fan and avoided the high pressure
losses previously found.

Staize 4 Cvclone

The upper part of the Stage 4 cyclone was removed and an enlarged new
section was installed. This converted the cyclone to a low pressure drop
design with a sloping inlet shelf.

52
 Previously the area where the Pyrochm precaiciner duct joined Stage 4
inlet was prone to build ups and this resulted in the higher gas pressure
losses. By modifying the Pyroclon duct to marry the Stage 4 sloping inlet
shelf, it was possible to reduce the gas pressure losses.

Kiln Back End Riser

The tertiary air duct used to enter the kiln back end riser duct after a
series of bends. This was modified to ailow a smoother gas entry.

Other Modifications

New segmented Vortex finders were fitted together with other

modifications to permit the new equipment to be installed.

The above modifications are illustrated in Figures 13-15.

9.3.4 Technical Evaluation of Uprating Proposals

The proposals for uprating the plant were evaluated using the f ollowing approach:

(a) Heat balances were constructed for the before and after conditions.

(b) Cyclone pressure drop and gas velocities are predicted from (a) for the
following cases:

existing cyclone data

modified cyclone data

(c) The preheater fan capacity was checked with allowances for the
following:

Higher gas voiumes resulting from less efficient kiln operation

Allowances for preheater and riser duct build ups

These checks r~ulted in the preheater fan size being increased from 2100
to 2250 kw.

53
 FIGURE 13 KANTHAN PREHEATER MODIFICATIONS
HIGHLIGHTED AREAS SHOWN BOLD DENOTE CHANGESTO ORIGINAL PREHEATERARRANGEMENT

..

I I I
I
I

\
1 t i
+’+ r ,

I I
I n --1 --Y

I \
I JI dldtl

R
t I t
1.1 1,1 I , I

:
.W 1 , ! , 1

54
 TERTIARY AIR DEmJSTiNG MODIFICATIONS
-.

-----
(

+
I
I
I

TERTLARY AIR

55
 FIGURE 15 KANTHAN NO 03 KILN-RISER DUCT MODIFICATIONS
STAGE 3 STAGE 3
HEAL HEAL

> I 4 I

BE I
\ \

Bll
GAS
r I
1.1u
I

I \L II I l\ 1 a

{I’: I Ill\’. I II { ffi I

BUI;D UP TERTIARY
OF tlATERIAL AIR

1 I
1

ORIGINAL OESIGN INITIAL REDESIGN FINAL DESIGN

HIGH PRESSURELOSS OF TERTIARY AIR TERTIARY AIR ENTERS ABOVE

DUE TO BENOS ENTRY TO RISER STAGE 3 IIEAL LEVEL
56
 (d) It was essential that the kiln could achieve the uprated capacity even if the process
did not achieve its guaranteed fiel consumption i.e.

Fuel consumption by heat balance for

No.3 Kiln prior to conversion = 789 net kcal/kg clinker

Guaranteed fiel consumption for

conversion (maximum) = 755 net kcal/kg clinker

Hence the cyclone sizing gas velocities and pressure losses were checked for 789
rather than 755 net kcallkg clinker.

(e) The Tenderer proposed to extend the Pyroclon precalciner by 11 metres.

However, this proposal was rejected by APMCIBCTC on the following grounds:

The existing precalciner volume was adequately sized according to BCTC

ratings.

It involved additional costs and was not practical from civil engineering
considerations.

(0 APMC/13CTC checked the proposals for modi@ing the kiln back-end riser duct
and tertiary air. It was noted that there was a risk of Stage 3 material short
circuiting into the Tertiary air duct at low kiln tonnage. Hence, it was agreed to
modifi the design of this area as shown in Figure 15 (see final design).

(d Several other proposals for uprating the cooler and cooler exhaust system were
rejected on the grounds of cost, of unproven design or as non-essential. It is
essential to minimisekiln conversion costs to the minimurn essential requirements.
All too oflen costs are escalated to include items which are desirable, but are not
essential to the uprating requirements. Tenders should be critically reviewed in
order to establish the essential and non-essential items.

9.3.5 Results Obtained from Kanthan No. 3 Kiln Upgrading

Following commissioning and optimisationwork No. 3 kiln has achieved peak outputs of around
3370 TPD compared with APMC/BCI expectation of 3250 TPD peak output. A comparison of
the preheater operating conditions and gas pressure losses is shown in Table 6. This shows that
the overall pressure drop was 97 mbar at 3370 TPD.

57
 393 97 374
Stage 1 cyclone exit 69

631 80.4 557

Stage 2 cyclone exit 50

768 54 728
Stage 3 cyclone exit 40

888 33.5 900

Slage 4 cyclone exit 28

10.2 15.2
Stage 4 cyclone inlet

19 16.6
Pressure loss across Stage 1 26.4
Pressure loss across Stage 2 10
12 20.5
Pressure loss across Stage 3
17.8 18.3
Pressure loss across Stage 4
10.2 15.2
Pressure loss across KilnlPrecalciner
69 97
Total

2500 3370
CLINKER OUTPUT (TPD)

ul
03
Prior to conversion, the overall preheater pressure drop was around 69 mbar at 2500
TPD. If the preheater had not been modified, the pressure 10SS would have increased
significantly as the gas velocities approached 30-32 m/s. The predicted gas pressure
loss at 3370 TPD would have been at least 115 mbar. Hence, the preheater
modifications saved at least:-

Minimum savings = 115-97 = 18 mbar pressure loss.

If the kiln had continued to operate with partial Air Through operation, the pressure loss
at 3370 TPD would have been higher due to higher dust recirculation at the kiln hearth,
etc. In this case, the overall losses may have risen to around 121 mbar making the
savings worth:-

Maximurn savings = 121-97 = 24 mbar pressure loss.

Operating the preheater at high pressure losses of around 121 mbar would have resulted
in higher inleaking air and higher power costs.

9.4 Conclusions Drawn From Recent Precalciner Kiln Urratin~ Exercises

The following is a summary of some of the conclusions which can be drawn from these
and other Precalciner uprating exercises:-

(i) Always carry out accurate surveys of the existing plant performance
before considering modifying the system. The surveys should include the
following:-

Full preheater/precalciner swey including gas temperatures,

pressures, oxygen, CO and C02 analyses, meal temperatures and
loss on ignition results to estimate decarbonation and dust recycle
levels.

(ii) Construct accurate heat balances for the present and likely uprated cases.
Take realistic targets for any improvements made to, say, clinker cooler
efficiency by modernizing the cooler design.

(iii) Identify those areas of the preheater, which will give the most benefits.
As in the case of Kanthan No. 3 kiln, these may include some “untypical”
areas, such as the precalciner exit duct and the ductwork between the
Stage 1 cyclones and the preheater fan.

(iv) Check gas pressure losses at the inlet and outlet of each cyclone stage.

(v) Do not accept, at f ace value, that Tenderers proposals will automatically
produce the savings claimed. Check these independently and consult
BCTC for advice.

59
 (vi) Avoid unproven cyclone design modifications which risk poor collection
efficiencies, etc. The Tender evaluation process should examine
parameters such as:-

The ratios of the cyclone inlet and outlet g= ~e~ comp~ed with,
the cyclone body areas.

Gas velocity profiles into and out of each cyclone stage.

Inlet duct design e.g. degree of wrap around and shelf design
(sloping or horizontal).

Depth of vortex finder compared with inlet duct depth.

Meal downpipe sizes and loadings.

Reference lists of similar installations.

(vii) Allowance should be made in the design of the cyclones and fans for less
efficient operation such as:-

Higher than expected fuel consumption.

Higher gas pressure losses due to build-up.

Remember that the uprated kiln has to operate under adverse, as well as
theoretical, efficiencies.

(viii) Avoid modifications which are desirable, costly and non-essential. Avoid
designing to the “wish list”. Subject desirable items to a financial analysis
which examines their justification on an incremental benefit basis.

(ix) Always examine the low cost approach to uprating f irst before considering
full phmt uprating schemes. Exampl= such as Cemento Melon No. 9 kiln
and Aberthaw No. 6 kiln uprating show that high capital expenditure is not
always necessary if basic plant sizings are adequate.

The above list is not intended as full guide to all the areas which have to be examined
when considering a major kiln uprating. BCTC can provide guidance and advice on these
and other aspects of kiln uprating exercis=. The establishment of realistic output
targets and identification of the major plant limitations is an essential part of this
process.

60
n
J
z
1
v
s
62
10. OPERATIONAL PROBLEMS WITH PRECALCINER KILNS AND SOLUTIONS

10.1 General

Since 1980, several precalciner kiln plants have been commissioned within the Blue
Circle Group, its associated companies and through its consultancy services. The
following examples are taken from several of these commissioning to serve as
examples. The problems were real ones, but have been simplified here to illustrate the
main problem areas and their solutions.

10.2 Practical ExamDles

10.2.1 Build-up Problems in Kiln Back-end Riser

Problem:- During operation of a 2500 TPD, 4 stage A-S precalciner kiln,

repeated blockages of the riser duct occurred when the material
was introduced into the upper riser duct (see Figure 16).

Solution:- The raw meal from the Stage 3 cyclone was introduced into the
kiln back-end riser duct at a lower point. This meal effectively
quenches the hot gases (1050-1 10°C) leaving the kiln and helps to
significantly reduce the level of build-up.

Notes- The introduction of meal at a lower point in the riser is a feature

of several precalciner processes such as the FLS-ILC and IHI-NSF.
In the latter process, the meal tends to be split to the vessel and
the riser. If all the meal enters the lower riser, it can cause
higher pressure losses and there is also more risk of underprepared
meal entering the kiln directly. Hence, a flow splitter device is
often used.

10.2.2 Overdrafting of the Kiln and High Fuel Input for the Maximum Kiln Output

Problem:- In order to achieve the maximum output froma3150 TPD, 4 Stage,

A-S precalciner kiln system, the operators found it necessary to
operate with 70% fuel to the kiln and 30% fuel to the precalciner.
Kiln stability remained poor with frequent kiln flushing.

Testwork:- Carry out full preheater survey gas pressure, temperature, oxygen,
CO and C02 analysis tests. Measure raw meal loss on ignition
levels at each stage. Obtain dimensions of riser ducts and cyclone
inlets and outlets. Estimate the meal decarbonation levels
(apparent) at each stage. Construct heat and mass balances for
kiln system.

63
 Results:- The meal entering the kiln appeared to be well prepared with a
high apparent decarbonation level of 94%. However, the C02
leaving the kiln was found to be 22%. Theoretically, from heat and
mass balance estimates, the C02 leaving the kiln should have only
been around 17%. It was suspected that the Stage 4 cyclones were
not collecting all of the meal entering the precalciner vessel.
Some material was short circuiting these cyclones by entering the
kiln directly.

The gas velocity checks showed that the pick-up velocity in the
riser duct was only 14 m/s. Hence, some of the Stage 3 meal was
not being entrained in the riser and was entering the riser directly.
This unprepared meal was contributing to the kiln stability
problems and its decarbonation resulted in the higher than
expected C02 levels in the kiln exit gases.

Solution:- The riser duct was modified by fitting a refractory restriction (i.e.
gas accelerator, see Figure 17) below the level where the Stage 3
meal entered. This accelerated the gases to approximately 25 m/s
before the entry of the Stage 3 meal. A BCI dispersion box was
also fitted in order to disperse the meal across the full cross
section of the riser duct.

By operating the system with 70°A fuel input to the kiln, the
operators were artificially increasing the gas velocities in the kiln
back-end riser. However, this measure was not fully effective
since some unprepared meal could still short-circuit into the kiln.
Hence, the need to design the gas accelerator and dispersion box.
Please refer to Appendix I for an example of how the C02 in the
kiln exit gases can be related to the meal decarbonation level.

10.2.3 Kiln is Unable to Produce a satisfactory Clinker Quality

Following a Major Uprating

Problem:- Following major plant modifications to a 2500 TPD, 4 Stage, A-S

precalciner kiln plant, difficulties were found in achieving a
satisfactory free lime. If the kiln feed was increased above 2000
TPD, the kiln flushed. At lower outputs the free limes were
variable and high.

Testwork:- The key area where modifications were made were the riser duct
and Stage 4 cyclone. Hence, these areas were targeted first by
testing the following:-

(a) Measure the gas temperature entering and leaving Stage 4 cyclone.

(b) Measure the meal temperature leaving Stage 4 cyclone.

64
 3
—--
tn -
/“
q \
0 w
4 a
g
I
w \\ f
m --~,
[
4-.
——— _ ,.
-=-J \
●—— — A.\ .’
F)
>
*0
/
y +E \-
\
65
 (c) Carry out a gas temperature and pressure survey of the riser above
and below the point where Stage 3 meal enters the riser.

Results and Conclusions:

Stage 4 Temperature Results

Before (°C) After (°C)

Adjustments Adjustments

Gas entering 918 876

Gas leaving 797 875

Meal leaving 850 856-868

Difference between Gas

Temperature Entering
and Meal Temperature
Leaving Stage 4 68 20-8

The precalciner was operating at a high exit temperature in order to improve the Stage
4 meal decarbonation levels. However, this high temperature was ineffective, as some
Stage 2 meal was not being picked up in the riser duct between Stage 4 exit and Stage
3 inlet. This was indicated by:-

The local low temperature of 797°C. This was due to cooler Stage 2 meal
at 585°C dropping downwards into the cyclone exit. This area of the
preheater was enlarged so gas velocities here were thus lower for a given
clinker output. (See Figure 18)

The Stage 4 meal temperature was 68°C lower than the inlet gas
temperature. This indicated that the Stage 4 temperature was reduced
by mixing with cooler meal from Stage 2. To overcome this problem, the
following changes were made:-

(a) The splash plate for Stage 2 meal dispersion was fully inserted into
the gas duct leaving Stage 4 cyclone. Later results showed that
the temperature difference between the gas entering, leaving and
in the body of Stage 4 became closer to the meal temperature
leaving Stage 4. This situation was improved as the kiln capacity
was increased and the gas velocities increased.

(b) Similar tests showed that a similar problem was arising with the
Stage 3 meal entering the riser duct. Hence, some unprepared
meal was entering the kiln directly giving a lower overall
decarbonation level. Hence, the Stage 3 meal dispersion plate was

66
Meal from Sta e 2 cycione

1 After 876 C

918 C

87$ C
..
.
.
“.
,
.
u Before_
Stage 2
u

%
i Meal shortclrcu~s to kiln

STAGE 4 CYCLONE

STAGE 4 Before 850 C

Meal i After -c

67
 fully inserted into the duct. This was prone to heat distortion so
a thicker heat resistant plate was used. Alternatively, a dispersion
box design could be considered.

10.2.4 Kiln Outputs Restricted Due to High Gas Pressure Losses and Other Limitations

Case 1
Problem:- A 3000 TPD, 4 Stage, A-S precalciner kiln suffered reduced output
for a period. Inspection of the preheater system does not show up
any obvious areas of unusual build-up. However, the preheater
pressure loss was higher than normal.

Testwork:- A stage by stage preheater gas pressure and temperature survey

was carried out.

Results and Conclusions:-

The pressure drop results pointed to the higher restriction being
between the kiln exit and Stage 4 cyclone exit.

Static Pressure (mmwd Surveys

Kiln Output Maximum

Case Restricted OUtDUt

Clinker output TPD 2300 3075

Kiln Shell & Hearth 11 33

Lower Riser Duct

Below Precalciner 127 63

Precalciner Base 173 9’1

Stage 4 Cyclone Roof

Before Exit 208 133

Hence, the overall pressure drop over the kiln, precalciner and
Stage 4 cyclone inlet plus body was 208-133 = 75 mmwg more than
normal. The loss in clinker output was 3075-2300 = 775 TPD below
the maximum potential.

The area of high pressure loss was in the upper riser and base of
the precalciner vessel. Subsequently, a heavy build-up was found
in this area. Once this had been cleaned, the pressure losses
returned to normal. This exercise pointed to the need to improve

68
 the automatic and manual cleaning facilities in the riser-
precalciner base area. Hence, additional air cannons and
inspection ports were provided. Due to the lack of suitable
inspection ports, the area of build-up could not be detected or
cleaned. However, the pressure drop survey pinpointed the
problem area. It is recommended that all preheater/precalciner
kiln plants should have regular surveys carried out since this helps
early identification of potential problem areas.

Case 2
Problem:- A 2600 TPD, 4 Stage, A-S precalciner kiln could not achieve
outputs above 2440 TPD due to preheater fan capacity limitations.
Checks upon the preheater fan showed that it was performing in
accordance with its design fan curve and there were no obvious
mechanical defects such as too high an impeller to inlet cone
clearance etc.

Testwork:- Initially check fan to confirm actual and design operation. Check
the overall preheater pressure loss and gas flow at fan against
design. Check cooler performance. This showed that the gas
volume was higher due to the clinker cooler efficiency being only
50% instead of the designed 63%. Separate studies were carried
out to optimise the cooler. The overall preheater pressure loss was
75 mbar compared with an expected 68 mbar. An oxygen survey
showed that the preheater inleaking air was very low.

A stage by stage pressure loss survey was carried out. This showed
that at outputs above 2100 TPD, the pressure losses increased
rapidly, especially over the Stage 2 and 3 cyclones (see Figure 19).

Preheater Inspection:-
A full preheater inspection was carried out at the major kiln stop.
It was found that the cyclone inlets at Stages 2 and 3 had been
incorrectly constructed. A refractory spoiler was fitted at each
cyclone inlet which was designed to reduce the width from 2.625m
to 2.2m. Unfortunately, these had been cast incorrectly and the
inlet ducts were only 2.0 m wide. The spoilers were modified and
the ef feet upon the preheater pressure loss is shown in Figure 19.
The kiln achieved 2620 TPD clinker. Subsequent optimisation of
the clinker cooler gave a further increase to 2700 TPD.

69
 FIGURE 19 CYCLONE PRESSURE DROP REDUCTION

Width increased from 2.0 to 2.2 m

/
2[
/{
Cast “spoiler”
I / /
at cyclone inlet ‘ ~
/

1!

$/’
&
f
/

10

,
12EEEEl ‘––

lzEEicl —
100 120 t40 160 180

KILN FEED RATE (TPH MEAL) 70

11. TYPICAL OPEIWTING CONDITIONS AND HEAT BALANCES
FOR A MODERN A-S PRECALCINER KILN

11.1 Number of Cvclone Stares

The overall fuel consumption and gas pressure loss characteristics of a precalciner kiln
process will depend upon several factors. Foremost amongst these is the selection of
the number of preheater stages. During the 1970’s and early 1980’s, it was more
common to install 4 stage single stream preheaters for capacities up to 3000 TPD.
Exceptions to this with the group were:-

Dunbar:- 5 stage preheater selected due to carbon in raw materials.

Additional stage required to avoid high exit temperature and
recover heat.

Rawang:- 5 stage twin preheater selected since shale is not preheated and
meets limestone plus sand in precalciner v~el. Additional stage
required to avoid very high exit temperature.

More recently, the adoption of 5 stage preheater has become more common. The
development of LP (low pressure drop cyclones) has allowed the use of 5 or 6 cyclone
stages without excessive pressure losses. For a new kiln line, a 5 stage preheater would
normally be specified, except in the following areas:-

At new plants such as Lakshmi in India, a 6 stage preheater was specified

in order to minimise the waste gas handling and fuel consumption
requirements. Availability of water for gas quenching, etc, is a problem
area due to the plants location.

The limit on drying capacity with a modern 5 stage preheater system is

around 5 to 6%. Hence, if the raw materials are higher (7-80A) 4 stage
operation may be required. However, there are ways round this problem
as follows:-

(a) Operate 5 stage preheater as a 4 stage system, using kiln feed

entry before Stage 2. This allows higher drying capacity to be used
for periods of high moisture (e.g. winter or monsoon conditions).

(b) Incorporate a hot gas generator in the raw material drying and mill
circuits.

(c) Use cooler exhaust gases ducted to the raw milling circuit (cf. the
ATOX boost system installed at Cauldon Works).

Hence, for simplicity, the following data relates to a modern 5 stage

precalciner system.

71
11.2 @eratinr Conditions and Heat Balance

Figure 20 shows some typical operating conditions and heat balance for a 5 stage A-S
precalciner of 6000 TPD output. These conditions are based on the following:-

Raw materials of average combinability characteristics.

Modern clinker cooler design used to give a cooler efficiency of around

72-75%.
,,

Kiln fuel consumption is typically 720 net Kcal/Kg clinker for 5 stage
operation. This would increase to around 750-760 net Kcal/Kg clinker for
a modern 4 stage precalciner process.

11.3 Gas Velocitv Profiles

Assuming a preheater exit gas oxygen content of 3.5%, tie gas volumes at each stage
are typically as follows (subject to the plant suppliers d~ign):-

Gas F1OW
Stage Exit Nm3/K~ Clinker

5 1.44
4 1.41
3 1.38
2 1.35
1 1.31
Kiln 0.43

For modern LP cyclones, the gas velocity profiles are typically as follows:-

Cyclone inlets = 12-18 m/s

Cyclone outlets = 10-16 m/s

11.4 Preheater Cyclone Loadimzs

The cyclone loading is defined as the TPD clinker per m2 cyclone body cross sectional
area (inside refractory). This empirical figure gives a rough indicator to the relative
size of the cyclones for a given clinker output. The higher the loading, the higher the
gas velocity profile is likely to be. Typical figures for new LP preheater supplied by
KHD, Polysius, UBE and IHI are:-

Preheater Cyclone Loading

(TPD/m2) for stages 2-5 = 70-80 TPD/m2

72
FIGURE 20 FIVE STAGE A.S. PRECALCINERt_KILN
tiPICAL PREHEATER GAS TEMPERATIJRE AND pRESSLJRE PROFILE

Gos- stat.
Temp. press.
expected
‘c mbar
waste gas 310 47
after Dopo L
pre CI+ Lone S

gosduct 5R

gosauct s 490 38

pre cycLone 4

gOSducL 4R

gOSducL 4 660 30

pre cycLone 3

gosduct 3R

gasaucL 3 7e5 23

pfe cycLone 2

gosduct 2R

gasd~c: 2 880 16

pee cycLone I 10

caLclnator 880

caLcleatoF- ~l~o 4
Inl-ec

Left .alr

~lLninle:

k~Ln 1200 2

73
FLS LP cyclones feature a generally smaller cyclone body area with a greater
enlargement of the upper cyclone body area. Hence, the FLS loadings are higher i.e.

FLS Loading for Stages 2-5 = 105-115 TPD/m2

The cyclone loading is estimated as follows:-

Stage 2-5 cyclone diameter = 7.90m

Lining thickness (mm) = 215

Cyclone diameter inside lining = 7.9-0.43

= 7.47m

Cross sectional area inside

lining = m J7.47)2
4
= 43.83 m2

Clinker output = 3500 TPD

Cyclone loading stages 2-5 = 3500

43.83

= 79.9 TPD/m2

Since this cyclone design is not of the FLS geometry, it complies with the BCI guideline
of 75-80 TPD/m2. This loading can be exceeded, but at the cost of higher gas pressure
losses.

When checking a preheater design, the heat and mass balances are used to arrive at the
actual gas volumes through each cyclone stage. From this the cyclone pressure loss can
be roughly calculated. The cyclone gas inlet and outlet duct dimensions as well as the
expected operating gas temperatures and pressures are used to estimate the gas velocity
profile. These must be checked to ensure @at adequate pick-up velocities are achieved.
Whilst raw meal can be carried over at low velocities (5-6 rids) through spouted bed
precalciners, such low velocities would cause short circuiting problems in the riser ducts
since this velocity is too low to pick-up the meal. Hence; the riser duct design should
be closely examined to ensure adequate dispersion devices and gas accelerators are
used. Minimum preheater velocities are around 10 m/s at the cyclone outlets before
meal enters the riser duct. Higher gas velocities are usually used to pick-up the meal.

74
Care has to be taken that preheater are correctly sized since:-

Undersized cyclones - Pressure loss= are high and fan capacity may
be insufficient.

Oversized cyclones - Pressure losses are low, high risk of short

circuiting and poor heat exchange causing
unstable operation. System may have to be
“overdrafted” to function correctly. This
costs fuel and fan power.

11.5 Preheater Cyclone Geome~

In addition to the above constraints, it is essential to check that preheater cyclones are
of a proven design. It is possible to produce a low cost preheater by using small cyclone
bodies with very large cyclone inlet and outlet ducts. This will give low pressure losses,
but will be inefficient by having a high dust loss. There have been several examples
where cyclones have been modified to reduce dust loss due to incorrect sizing criteria
being used.

As a simple check, the cyclone inlet and outlet areas are compared with the cyclone
body area (inside lining). For example cyclone exarnple:-

Cyclone body area = 43.83m2

Cyclone inlet area = 10.96m2
Cyclone outlet area = 12.79m2

The critical ratios are thus:-

Cyclone Inlet area/

Cyclone body area = 10.96
43.83

= 0.25

Cyclone Outlet area/

Cyclone body area = 12.79
43.83

= 0.295

75
These ratios are acceptable and compare with the following guidelines for cyclone inlets
for Stages 2-5.

Inlet/Body Ratio

Older preheater designs

(1970’s) high pressure
loss 0.18- 0.21

1980’s medium to low

pressure loss cyclones 0.20-0.27

1990’s LP cyclones 0.21 -0.34

In the case of current Stage 1 cyclones, the above ratios for Stage 1 cyclone inlets are
lower i.e. around 0.21-0.25.

FLS LP cyclones with their narrower cyclone body size tend to have higher inlet/body
ratios i.e.

Stages 2-5 - 0.42-0.44

Stage 1 - 0.43

11.6 Summarv

The above guidelines are provided as a cross reference when checking new and existing
preheater cyclone designs. Whilst the use of empirical data is no absolute guarantee as
to whether or not a cyclone will function efficiently, these guidelines can help to
identify suspect or unproven designs.

In one recent case, at Stage 1 cyclone modification was found to give an inlet/body
ratio of around 0.33 which was too high for a non-FLS cyclone geometry. Subsequent
tests showed preheater dust loss levels of over 15% on clinker i.e. twice the expected
level. The cyclone inlet area was reduced at the expense of pressure loss to improve
collection efficiency and half the dust loss. Fortunately, the preheater fan was large
enough to cope with the higher pressure loss.

When assessing the performance of a precalciner system, it is thus important to examine

carefully the limitations of its component parts which include the preheater cyclone,
precalciner, kiln, clinker cooler and ancillary plant. Always ensure that preheater fans
have a generous design margin to cope with any adverse pressure losses.

76
 APPENDIX I

CARBON DIOXIDE CONTENT OF THE GASES LEAVING A KILN

AND THE DECARBONATION OF THE MEAL ENTERIN G THE KILN

Carbon Dioxide in the gases leaving the precalciner kiln are generated by the following-

(a) C02 from fuel combustion

(b) C02 from decarbonating the meal from approximately 85-92% to 100%
decarbonation before sintering.

If the precalciner vessei is inefficient and the meal do= not receive adequate
preparation/decarbonation prior to the kiln, this will be indicated by higher kiln exit
C02 levels. The C02 generated from the meal decarbonation in the kiln is greater than
the C02 from fuel combustion.

The attached graphs show the low and variable decarbonation level of the Stage 5 meal
entering Rawang No. 3 kiln. These results were taken after the kiln was uprated to
5000 TPD. This was attributed to material short circuiting the enlarged precalciner
vessel by two possible routes:-

(a) Stage 4 meal not being picked up in the riser duct.

(b) Stage 3 meal short circuiting through the Stage 5 cyclones.

Overdraf ting the kiln appeared to help (a) but the highly cohesive nature of the raw
meal (limestone mainly up to stage 4) can make it difficult to disperse efficiently in the
riser ducts.

Please note that the C02 graph is derived from Heat Balance data specific to each kiln.
Hence, this graph is a guideline only. If similar problems are suspected for any
precalciner kiln system, then this graph should be constructed from Heat Balances using
the fuel characteristics, fuel consumption and gas analysis data.

77
 APPENDIx I
——

Rawang Works Kiln inlet C02 content - vs Decarbonation

(at different 02 content)
0/
/0 decarbonation
95

90

\ “.., ‘.
85 \* “(J ‘N~
\ .. \\
\ “.., \2%
“..

\ “’”.,, “’,
\ .,.. \.
.=\
80 \ ..: *. \. . . . . .
\ ,, .
\\ ...ti
..*m
*.= \
.. .
\ .. \
\ \
.
75
2%02 3%02 4%02
“k== “k’ ,.. ‘L\
\
5.3%02 \ “. x\
-~ -- A -- .....~..-. —+— \ .. .
.. .
.. .
\ ,. .
,.
\ \
70 I I I I I ,.Y \

14 16 18 20 22 24 26 20 30

% c02 (dry basis - Orsat)

VARIATION OF CALCINATION %
INSTHSTACECYQLONEIFRM7 AM-245?W
28,
9s -
94
93
%2 -
m
-
-
II
90 -
89
88 -
87 -
08
85
84 -
83
02 -
81 -
80 -
x
T!31 I 1
7.ml l’I’I’I’ l’I’I’I’ +ti’l’l’l’l’~
1
2PM
Date: 16/12/92
TIME
Cllnker produced : 1512 ~ (Ist. Shift)
Kiln Feed : 345 tph a A cyaom + B cya-
5th. Stage temp. : EIso”c
78
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 8

Precalciner Kilns
(Lecture 24)
 24.1

1 INTRODUCTION

In the last few years demands for higher outputs from kilns and the use of raw materials not
well suited to traditional preheater kilns, have resulted in advances in the development of
new cement manufacturing processes.
The major development has been in the field of precalcination and this paper deals with the
various designs currently available of this type of kiln system.

2 PRECALCINATION

2.1 INTRODUCTION
A development of recent years has been the introduction of precalcination in the dry
process manufacture of cement, Precalcination is the term used to describe processes in
which the fuel input is split between the kiln and some earlier point before the feed
material enters the kiln. such as between kiln and preheater, whereby the kiln receives
material with a greater degree of calcination than normal,
There are a variety of systems on offer from the main plant suppliers and developments
are still taking place and claims being made many of which are relatively untried.

2.2 PURPOSE
The main purpose behind this development has been to reduce the thermal load on the kiln
or permit the use of a smaller sized kiln for a given output. This is especially important for
outputs in excess of 3000 t.p.d where conventional kiln sizes lead to ever increasing
bricking problems as the capacity increases. Also the smaller kilns result in decreased
capital costs, up to about 10% for kilns and preheater,
 24.2

A second approach has been in connection with the uprating of existing plant where major changes
have been made to coolers, preheater cyclones, fans, etc. Again the object is to relieve the thermal
load on the kiln, reduce build-up problems and give steadier operation at the higher output. The
techniques used for precalcination are however somewhat simpler in this case.

The third possibility arising from precalcination is to take advantage of the reduced volume of kiln gas
where alkali bleed-off is required. The resulting greater concentration of alkalis in the kiln gas means
that a given percentage bleed is thermally less wasteful than with a normal system,

2.3 PRECALCINATION SYSTEMS

Classification is difficult but to try to simplify the range of the systems available
they can be divided into two main categories:

a) Those with a separate furnace inserted between the kiln and the stage 4
preheater cyclone in which the auxiliary fuel is burned and the precal cination takes
place.

b) Those where the auxiliary fuel is simply injected into the riser duct (possibly
enlongated) from the kiln to the stage 4 preheater cyclone.

As a further variation the air for the combustion of the auxiliary fuel may pass either
from the cooler through a separate duct to the burner or through the kiln. Generally
the (a) systems have separate ducts whilst the simpler (b) systems use the kiln
although there is sometimes a choice with these.

The main stream of development has been in Japan as a result of an expanding market
and the idea originated from alumina kiln practice. This resulted in the principle of a
separate furnace unit or flash furnace in the I.H.I. (Fuller) and R.S.P. (Onoda) systems
and a separate fluidised bed calciner with Mitsubishi. F L Smidth have also
independently been developing a vessel for the combustion of auxiliary solid fuels
which might be similarly classified.
24.4

Precalcination takes place in the 'Flash Furnace' where hot gases both directly from the kiln and from a grate
cooler via a separate duct enter from the bottom in strong turbulent flow, having passed through a vortex
chamber. Raw meal enters from the top of the furnace and three burners located in the roof of the chamber
create a downward motion of raw mix and hot combustion gases which meet the ascending gases creating a
strong blending effect and effecting rapid heat transfer.

The flow pattern is illustrated in Fig. 24.2.

The flash calciner diameter of a 3500 t.p.d. preheater plant is 7.4m and for 4000 t.p.d. is 8.2m. The height
of a flash calciner equals the cyclone height of the lowest preheater stage.

As can be seen from the flowsheet the precalcined raw material is separated from the gas stream in the fourth
stage of the preheater and thenceforth gas and material progress through the system as in a normal four stage
suspension preheater.

According to I.H.I. only 40% of the total heat required is supplied to the rotary kiln, the remainder being
supplied into the flash calciner. Up to 90% calcination of the raw material is achieved before entering the kiln
compared with about 10 - 15% in a normal suspension preheater.

Table 24.1 shows a typical heat balance for the process as claimed by I.H.I.
24.5
 24.6

4 THE KAWASAKI/ONODA - RSP PRECALCINING PROCESS

The RSP suspension preheater (abbreviation for Reinforced Suspension Preheater) has been jointly developed
by the Onoda Cement Company and Kawasaki Heavy Industries in Japan.

It consists essentially of a two compartment calciner i.e. the heating shaft supplied with a swirl burner and the
calcining shaft arranged about parallel to it. The combustion gases and raw material from the calciner are mixed
in a mixing chamber before passing to the fourth stage of the preheater. A typical flow sheet for the system is
shown in Fig. 24.3.

Fig. 24. 3

FLOWSHEET FOR RSP SYSTEM ONODA

 24.7

A more detailed representation of the calciner and mixing chamber is given in Fig. 24.4 where it can be seen
that hot air from the cooler enters both the swirl burner and swirl calciner. The swirl burner burns only a small
portion of the fuel used at this point but plays the main role in stabilized burning with a relatively large air to fuel
ratio. The swirl calciner is the most important part in the burning process of the RSP. It enables efficient
combustion near to stoichiometric fuel-air ratio, and the raw meal which enters the system here is rapidly
calcined in the swirling flames. The mixing chamber is designed to effectively mix the gas/meal mixture from
the swirl calciner with the kiln gas to ensure effective heat exchange.

The original installation with a capacity of 240 t.p.d. was located at Onoda's Tahara Plant and was basically a
test plant. Subsequently a 3000 t.p.d. installation was built by Onoda at Ofunato and at September 1977 seven
RSP plants were in operation or under construction.

The RSP furnace is supplied with 55 to 70% of the total heat requirement with the balance supplied to the kiln
which has only the duty of sintering the feed, 80 to 90% of the calcination having been carried out in the furnace.

A typical heat balance for the RSP process as given by Onoda is reproduced in Table 24.2.

Table 24.2

HEAT BALANCE OF AN RSP PREHEATER SYSTEM

HEAT INPUT kcal/kg HEAT OUTPUT k cal/kg

From fuel 740.6 Theoretical heat 429.2

Sensible heat of fuel 2.9 Sensible heat of clinker 28.7
Sensible heat of raw mix 11.3 Sensible heat of cooler 82.4
exhaust
Sensible heat of exit gases 143.2
Sensible heat of dust 5.6
Radiation losses 59.8
Balance 7.3
TOTAL 756.2 TOTAL 756.2
24.8

To date all installations have been oil fired. Test work burning coal on the 240
t.p.d. plant has been carried out, though in this case it was not found possible to
substantially complete combustion in the swirl calciner and this was completed
in the mixing chamber.

Fig. 24. 4

RSP CALCINER
 24.9

5 THE MITSUBISHI FLUIDISED CALCINER

The Mitsubishi Fluidised Calciner (M.F.C.) is a joint development of Mitsubishi Minina and Cement and
Mitsubishi Heavy Industries. A separately heated fluid bed reactor located between the kiln and preheater
is used to effect ca lcination. The fluid bed can be arranged in two alternative ways; the so called
‘overflow system' is illustrated in Fig. 24.5(a) and in this system the calcined raw material in the bed
overflows directly into the kiln. In the carry over system, material is carried out of the bed by the upward
flow of gases through the bed and into the kiln back end chamber from which it is carried to the fourth
stage cyclone of the preheater where it is separated and fed to the kiln.

Fig 24.5 FLOWSHEET OF MITSUBISHI FLUIDISED CALCINER SYSTEM

The fluid bed calcalciner is illustrated in Fig 24 6

 24.10

Hot air from the cooler is utilised in the fluidised calciner, part being diluted with cold air to a
temperature of 3500C and used to fluidise the contents of the vessel via fluidisation nozzles in the
base. The remainder having a temperature of 500 - 6000 C is used as secondary air. Effective
calcination of the raw meal is achieved using the additional heat from fuel directly injected into the
fluidised bed. One of the main features claimed for the MFC is the stability of the calcining
operation due to the uniform temperature distribution, usually + 100C throughout the bed.
 24.11

The fuel used in the MFC may be oil or coal or a variety of waste and low grade fuels, in the case of
coal it is claimed that relatively coarse particles can be completely burnt due to the long retention
time on the bed.

For a 2,200 t.p.d. kiln 4.3m x 65m the MFC fluid bed reactor had a diameter of 4.Om and height of
4.5m and required a fluidisation fan handling 650m3 /min at 3500C and 1,800 mm w.g. and having
a 180 kW motor. The heat consumption was quoted to be 781 kcal/kg. The heat balance given in
Table 24.3 below is for an oil fired kiln with a coal fired MFC.

Table 24.3
HEAT BALANCE OF AN MFC PREHEATER SYSTEM
HEAT INPUT kcal/kg HEAT OUTPUT kcal/kg
From fuel kiln 619 Theoretical heat 435
M.F.C. 152 Sensible heat of 17
clinker

Sensible heat of fuel 3 Sensible heat of cooler 107

exhaust
Sensible heat of raw mix 13
Evaporation of raw meal 1
Sensible heat of primary air 9 moisture

Sensible heat of exit 155

gases
Sensible heat of dust 8

Losses 73
TOTAL 796 TOTAL 796
 6 THE F L SMIDTH PRECALCINING PROCESS

The F L Smidth precalciner is illustrated in Fig. 24.7 and it can be seen to be of simpler construction than
those of Japanese manufacture. One fuel injection point adjacent to the raw material inlet is used and the gas
flow and shape of the precalcining chamber produces a dense vortex in which the fuel and meal are
intimately mixed before combustion air is available so when the hot air is available combustion takes place
and calcination is achieved, though there is no flame as such.

Fig. 24.7 FLS PRECALCINER IN DOUBLE STRING PREHEATER

The chamber as it has now developed is simpler than the original concept which was to have a fluid bed of
material in the bottom cone through which the gas riser would project. There were originally two feed inlet
points and three oil burners were used. Experience proved that fluidisation was not necessary and only one
feed pipe and burner were required.
24.14

7.2 THE POLYSIUS PREPOL

The system is illustrated by Fig. 24.10 schematically and by Fig. 24.9 in projection. The design results from
Polysius's efforts to simplify the process and achieve the same precalcining effect no matter what kind of
cooler is utilised. The combustion of the fuel necessary to effect precalcination of the raw meal takes place
in the connecting duct between the kiln and preheater and the required combustion air is drawn through the
kiln. Raw meal from the vortex vessel of the Dopol preheater enters the ascending duct where heat from the
auxiliary burners causes extensive calcining. After separation in the lowest cyclone stage the raw meal
enters the kiln. The ascension duct is sized to ensure complete combustion of the fuel before it reaches the
Stage 4 cyclones.

The Prepol system was developed at the works of Rohrbach, Dotternhausen and is often referred to
by these names.

Fig. 24.9 POLYSIUS PREPOL SYSTEM

24.16

8 GENERAL COMMENTS ON PRECALCINATION SYSTEMS

8.1 PROVISION OF COMBUSTION AIR FOR PRECALCINER

In the various systems described above, the air required for combustion of the fuel added between
kiln and preheater has been supplied in two ways; either by a duct from the cooler or by operating
the kiln with a high level of excess air so that there is sufficient oxygen in the kiln exit gases to
allow complete combustion of the fuel added after the kiln.

There are advantages and disadvantages to both these methods but where a standard planetary
cooler is employed, the air must pass through the kiln.

When the additional air is passed through the kiln the increased level of oxygen results in a shorter
flame, and consequently a lower gas temperature at the kiln feed end, which again reduces the risk of
coating in the preheater and build up in the kiln. There were, initially, doubts as to whether the flame
produced when running with high excess air levels would have adequate heat to properly sinter the
clinker, but operating results for several kilns over the last year have proved these doubts unfounded.

It is obvious that when the precalciner combustion air is supplied via the kiln the velocity of the kiln
gases will be higher. In order to maintain this velocity at the permissible level, (21 m/sec according to
Polysius) the diameter ratio between kilns operated this way and those with a separate air duct can be
calculated to be 1.2 : 1. The disadvantage of requiring a kiln of large diameter for this method of
operation is offset to some extent by being able to dispense with the separate air duct from the cooler
which must be as long as the kiln and in the case of a 3000 t.p.d. kiln have an internal diameter of 2.2m.
This duct with refractory insulation and supporting structure can be very expensive.

The air through kiln system has the additional advantage of being easier to control; because in the
separate duct system the ratio of gas flow between kiln and duct must be maintained constant, and due
to variations in the kiln such as degree of coating etc., a complicated control system is required.
 24.17

A disadvantage of the air through kiln method is that the oxygen required for combustion of the fuel added
at the precalciner is relatively thinly dispersed in the kiln exit gas compared to the oxygen rich air from the
cooler supplied in the separate duct system. This makes the effective combustion of the fuel added at this
point and precalcination of the raw material more difficult.

8.2 EFFICIENCY

It is too early to be certain of the relative efficiency of the various plants but in general the separate
flash furnaces tend to have the same or marginally lower fuel consumption than a conventional
suspension preheater kiln.

On the other hand some of the simpler systems have a slightly higher fuel consumption (10 - 20
kcal/kg) as a result of higher exhaust gas temperature and CO. Both tend to consume slightly more
electric power.

From an operational point of view the flash furnaces appear to be the more complex of the two categories
and we would expect more problems than with conventional plants although the suppliers claim otherwise.
In this respect the F.L.S. arrangement with completely separate preheater strings for kiln gas and calciner
gas, including separate fans, may be advantageous.

8.3 OUTPUT RATING OF KILNS WITH PRECALCINERS

As discussed in a previous paper B.C.I.'s rating of the potential production capacity of a kiln, assuming
ancillaries of adequate size, is based on two parameters, gas velocity expressed in t/day m2 of kiln
cross section and kiln surface area expressed in t/day m2 of kiln internal surface. For kilns with
precalcination, however, because of the limited information available to date our ratings are expressed
only in terms of kiln surface area, though regard is paid to what we consider to be acceptable
velocities.
24.18

For simple precalciner systems, i.e. those in which all the air is passed through the kiln and
precalcination is achieved in the riser duct from the kiln, such as the Prepol and Pyroclon, an average
figure of 3.6 t.p.d/m2 (kiln internal surface) has been derived with 10% more for best day
performance.

For the precalcination systems with separate air ducts a figure of 4.0 t.p.d/m2 (kiln internal surface)
is proposed with again 10% more for best day performance.

8.4 ALKALIS

A secondary but important consideration is the question of alkalis. With a full flash calciner system where
the tertiary air bypasses the kiln, the kiln gas quantity is much less than normal yet contains the same
amount of alkali. Thus if this gas is partly bled off the losses due to sensible and latent heat in the associated
gas and dust are less than normal. In fact they are about halved for the same alkali reduction. The graph in
Fig. 24.11 shows the general order of magnitude of the fuel consumption for flash calciner and normal four
stage preheater kilns with various amounts of bleed-off and at two levels of kiln dust loss expressed as % of
clinker.

If the chemistry is such that a bleed-off above 20% were needed we would consider using a precalciner and
should it be necessary to make a very low alkali clinker, the F.L.S. Dania system has merit. In this all kiln
gas is bled-off and dedusted.

Precalcination and normal suspension preheater systems with alkali bleeds both permit the removal of
alkali sulphate but the effects of chlorides, particularly KCL, are less well known with precalciners.

There is a possibility of blockages in the preheater above the calciner and research on alkali volatility
rates is in progress. In this respect temperatures in the 800 - 9000C ranges are critical.
24.20

9 CONCLUSION

As mentioned at the beginning of this paper the development of precalciner process is still
relatively new. Many developments are still being made and operational experience is limited
but whilst more knowledge of the performance and potential of the various systems is
becoming available, the effectiveness of many of the systems now offered remains to be
proved, On the basis of our present knowledge we would envisage the use of precalcination as
indicated below but every potential application needs very careful examination.

1. For new large plants above 3000 t.p.d. consideration should be given to a separate
precalciner system such as those offered by F L Smidth or I.H.I.

2. For a new plant with a need to produce low alkali clinker from materials which dictate a
substantial bleed-off, consider a precalciner system with alkali bleed, such as FLS or I.H.I.

3. For an existing large suspension preheater plant being uprated, i.e. modifications to cooler,
cyclones, fan, etc., consider auxiliary firing to ease the thermal load on the kiln and give steadier
running. Typical systems from Polysius, Humboldt, Miag, as appropriate to existing plant.
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 9

Kiln Systems
(FLS Lecture)
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

5.1.A

Wet-Process Kiln Systems

An important part of the world production of cement is still being made according to the
wet-process.

During the last decades the inquiries for wet plants have decreased dramatically and for the last
decade the wetprocess projects only represent a few percent of the total implemented projects
calculated on basis of tons
clinker produced.

Today wet kilns are being built in special cases:

- a cheap and simple expansion of an existing wet plant

- raw materials with high moisture content
- in areas with cheap fuel - typically a waste fuel.

Examples of kiln systems based on slurry are shown in figure 1. This sketch also indicates the
relative kiln length required for the various process types. The long wet-process kiln, A, with
different types of internal fittings is, of course, the most common type and we will revert to that
later.

Spraying of slurry into a chain system has been used in some cases, but has generally been
abandoned. The more recent version is the use of a spray drier as a separate unit in front of the
kiln. The spray drier consists of a fast rotating centrifugal atomizer. All the water from the slurry
still has to be evaporated, but the advantage is that the exit temperature is somewhat lower,
1500C, than for a long wet kiln.

On the other hand, the system involves an extra power consumption of 5-6 kWh/t, and the capital
cost of the drying equipment is normally higher than the lower costs of a smaller kiln.

Other types of slurry driers installed at the kiln inlet have also been used, for instance the
Miag-calcinator, but they have generally been abandoned.

The remaining three examples in figure 1 show various arrangements for reducing the moisture
content of the slurry by filtration, the aim being to reduce the heat required for evaporation of the
moisture. Complicated pressure filters are normally required for the plastic slurry typically found
in wet plants, but for an unplastic slurry a disc filter might give sufficient reduction of the water
content.

The filter cake can be fed directly to a kiln fitted with chains, or it can be nodulized and fed to a
kiln equipped with a long cross system. The filter cake can also be dried before entering the kiln
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
in a drier crusher, a sort of hammermill, or in a rotary drier. Finally, the filter cake can be
extruded and used on a grate preheater, a Lepol grate for instance.

Slurry filtration is mainly of interest for converting from a wet to a dry process.

The classic, economic wet-process kiln may have internal fittings as shown in figure 2, consisting
of slurry preheater, garland chain system and a cross system. Such kilns are designed for a low
dust load in the kiln gases, and hence are suitable for plastic raw materials which form nodules of
good strength, and when the kiln gas velocity is relatively low.

Such nodule kilns can operate economically with a low exit gas temperature, 1700C, and for
instance a fuel consumption of 1350 kcal/kg clinker at 38% moisture in slurry. Further, they have
the advantage of a high alkali concentration in the dust, making alkali elimination easy. The
disadvantages are that even with suitable raw materials, a large kiln size for a given output is
required and for kilns larger than say 600-700 t/day, such installations would hardly be
considered.

For larger kilns it would not be economical to maintain the same gas velocity in the kilns, as in
fact the output is approx. proportional to L x D1.5 and not to D2.

Larger kilns are, therefore, better operated as "powder" kilns with the chain system extended so
far down in the kiln that the material leaves the chain system as dust at 3000-4000C.

The chain system for such a kiln is shown in figure 3. The system consists of perpendicular
chains suspended on spirals. The density of the, chain system is quite important for the transport
capacity; normally about 20 m chain per m3 kiln volume are used in the inlet end, and say 40
m/m3 in the hot end where material transport is no problem. The total amount of chains will, in
general, be between 10 and 15 m chain per t/24 h output.

It is not always an advantage to increase the amount of chains, because it may increase the
internal dust circulation and lead to higher exit gas temperatures.

A more recent development in chain suspension is also shown in figure 4. The idea is to hang the
chains very close on the spirals and space the spirals sufficiently apart so that the chains in the
bottom can move without sliding over the chains of the neighbouring spiral. Improved transport
capacity and reduced wear of the chains can be obtained, while the lining, of course, is more
exposed.

Garland chains would, from an operational point of view, also in the large kilns be efficient in
the middle zone of a chain system, but the wear of the first chain link can be severe, especially
when corrosion is a problem. As repairs are troublesome and time consuming, the tendency is to
use only vertical chains arranged in a simple and uniform pattern
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
In figure 5, operational data are given for two kilns at the same plant, one operated as a nodule
kiln and the large kiln as a powder kiln. For the large kiln is also shown the important reduction
in heat consumption obtained by reducing the slurry moisture using slurry thinners and at the
same time reducing the output and insufflating 8% fly ash.

It has been mentioned that with wet kilns, especially nodule operated kilns, an electrostatic
pricipitator is effective at separating the alkalies and the dust.

The alkali-rich dust is discarded and the remainder is returned to the kiln. Dust return to the kiln
feed is normally not advisable, as it often means that more water per ton of clinker has to be
evaporated, and it may also cause toublesome mud rings in the chain system.
In small kilns dust return with scoops on the kiln shell below the chain system can be very
successful but in large kilns, say over 1000 t/d, this system is not advisable. Problems with the
seals in combination with underpressure and gas velocity in such kilns may lead to reduced
output if scoops were used for dust return.

Therefore, in large kilns dust insufflation in the burning zone will be the best way to dispose of
the dust, and if the amount does not exceed 5-10% of the dry feed, there are generally no
problems. On the contrary, an improvement of the lining life is often experienced.

Although wet kilns of capacities up to 4000 t/d are in operation, it is now generally accepted that
the optimum size of a wet kiln is about 1500 t/d with kiln dimensions of approximately 6∅ x 5∅
x 170 m.

One disadvantage of wet kilns is the specific power consumption of the kiln drive motor which
increases with the size, while, for instance, a 300 t/d kiln requires 4-6 kWh/t, and the
consumption for a 2500 t/d kiln is 12-16 kWh/t.

The most modern wet kiln is the Rordal solution shown in figure 5A. This solution combines
modern calciner kiln technology with traditional slurry preparation.

The system is attractive in regions where the natural moisture content in the raw materials is high
or in cases where slurry preparation already exists and a dry component can be added to reduce
the overall moisture content in the feed.

A paper giving a description of the kiln is enclosed in chapter 5.1.C: Modification of kilns.

As mentioned earlier, today's inquiries for wet kilns are related to special cases. For this reason
the development work on this kiln type has not been given special attention for the past 10 years.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
5. 1. B

Dry-Process Kiln Systems

Introduction

For more than a century FLS has designed rotary kilns for the cement industry. For the last 20
years the main development efforts have been concentrated on dry kiln systems.

The development has had three different trends, viz.

A) lower heat consumption

B) lower power consumption
C) larger units

but all three resulting in lower production costs.

In modern FLS dry-process kiln systems preheating of the raw meal up to calcination temperature
takes place in a cyclone preheater and a major part of the calcination process may take place in a
separately fired stationary calciner. The remaining part of the raw meal calcination as well as the
final heating and clinkering process is carried out in a rotary kiln without internals.

Cooling of the clinker is usually performed in a grate or planetary cooler of the well-proven FLS
Folax and Unax design.

The present FLS programme of dry-process kiln systems includes five main types' each of which
with its own special advantages and field of application.

This makes it possible to select the most suitable type of equipment for any given set of
conditions and requirements.

The main features and advantages of the five systems are described on the following pages.

Kiln Systems

The five FLS dry-process kiln systems are called:

a)Suspension Preheater Kiln SP

b)In-Line-Calciner with Excess air ILC-E
c)In-Line-Calciner Kiln ILC
d)Separate Line Calciner Kiln SLC
e)Separate Line Calciner - Special SLC-S
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
a) SP-Kiln

The well-known suspension preheater kiln was introduced more than 25 years ago. An example
of an SP-kiln is shown in figure 6.

An SP-kiln can be equipped with a 1-6 stage preheater, single or double string.

The cooler type can be chosen between the planetary and the grate cooler.

The normal capacity range is 700-2800 tpd. A typical kiln with 2800 tpd capacity is equipped
with a single string preheater with Low-Pressure cyclones and 4 or 5 stages.

In an SP-kiln with 4 or more stages the material temperature in the preheater reaches the
calcination temperature. The degree of calcination of the material introduced into the kiln system
can reach 50%.

In some cases a minor firing in the riser pipe, 5-10% of the total fuel consumption, can be
advantageous for the following reasons:

- reduces excess air in exit gases from top stage

- improves heat economy
- increases 02 -content in burning zone
- reduces evaporation of SO2, Na20, K20 in the burning zone.

This kiln type can accept a higher concentration of sulphur, sodium, potassium, and chloride
without requiring a by-pass. Especially in cases, when the Cl-concentration in the raw mix is
high (0.015% - 0.020%) the SP-type can, depending on an easy burnability, be offered without a
by-pass.

An SP-kiln equipped with planetary Unax coolers gives a very simple operation and with the
latest innovations within cooler and preheater technology it can give excellent results.

An example of a 5-stage SP-kiln with planetary coolers is shown in figure 7.

c) ILC-E-Kiln

The in-line calciner kiln with excess air is a further development of the SP-kiln with a small
precalciner built into the riser pipe. An example is shown in figure 8.

The ILC-E-kiln is generally equipped with 4-6 stage preheater, single or double string.

In connection with semi-dry systems the preheater can be equipped with 1-3 stages of cyclones.

The cooler type can be chosen between the planetary and the grate cooler.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
The normal capacity is 900-3300 tpd; a typical kiln with 3000 tpd will have the dimensions 5.0 x
78 m equipped with Unax coolers 9 x 2.55 m x 25.2 m. The kiln diameter and the cooler size
will limit the upper capacity.

In an ILC-E kiln the material in the preheater reaches the calcination temperature, and the degree
of calcination of the material introduced into the kiln can reach 60%. The calciner is built into the
riser pipe. The gas velocity in the calciner is reduced to ≈ 10 m/s ensuring an extension of
the retention time of gas, fuel, and material.

This kiln type is normally designed for a 15-25% firing in the riser pipe. In practice 10-15% have
proven to be advantageous, ensuring the lowest overall heat consumption and a good and stable
kiln operation.

With 85-90% firing in the kiln the excess air ratio in the burning zone is high, giving

- an easy operation
- low coating tendency in kiln outlet and riser pipe
- long kiln lining lifetime

The ILC-E-kiln as well as the SP-kiln can accept a higher concentration of minor components
(S, Na20, K20, Cl) without requiring a by-pass.

c) ILC-Kiln

The in-line-calciner kiln system was introduced in 1976 and a considerable number of systems in
operation have proven the reliability of the configuration.

An example of an ILC-kiln is shown in figure 9.

An ILC-kiln can be equipped with 4-6 stage preheater, single or double string. The calciner
vessel is built into the kiln riser pipe. The air for combustion in the calciner is drawn from the
cooler through a separate tertiary air duct between the grate cooler and the calciner, and mixed
with the gases from the kiln at the inlet of the calciner.

The usual capacity range is 1500-4000 tpd. There are no limitations as such for the maximum
capacity of this kiln type but due to the configuration of the system with a built-in calciner the
tower height willincrease considerably, making the ILC-kiln too expensive.

An ILC-kiln is a true calciner kiln with normal gas temperature in the calciner and the lower
cyclone stage in the range of 870-9000C. With this temperature the decarbonisation of CaCO3
occurs rapidly. The normal degree of calcination of the material introduced into the kiln is
90-95%.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
The hot air from the cooler is mixed with the kiln gases prior to being used for combustion; this
means that combustion in the calciner is sustained by a gas mixture relatively low in oxygen
(11-12%). The draught in both the kiln and the calciner is controlled by the same fan. This
necessitates a variable damper in the hot air duct in order to have some means of draught control.

In calciner kilns the normal firing in the kiln is 310-320 kcal/kg clinker and the rest is introduced
into the calciner. This gives the following fuel distribution between the kiln and the calciner
vessel:

Heat Consumption To Kiln To Calciner

790 kcal/kg 315 475
100 % 40% 60%
700 kcal/kg 315 385
100 % 45% 55%

The ILC-calciner vessel is built into the riser pipe. The retention time of the gas in the unit
should be minimum 3 seconds to ensure a safe outburning of the fuel in the vessel. The gas
velocities in the calciners are normally kept within 5-7 m/s.

An example of a calciner vessel is shown in figure 10.

The in-line system offers a flexibility for a by-pass between 0 and 100% of the kiln gases. The
penalty in the heat consumption for by-pass is 1.6-2.0 kcal/kg clinker/% by-pass compared to
4.5-5.5 kcal/kg clinker/ % by-pass in an SP- or ILC-E-kiln.

A calciner kiln can be dimensioned for low-grade fuels.

With high ash content in a coal there is a certain limit for the lower heat value (Hi) in the burning
zone due to high temperature requirements. But for the calciner vessel there is no limit as the
temperature is low (≈ 9000C) and the ash is well-mixed with the raw meal and used as
clay-component in the clinker.

An example of a 4-stage ILC-kiln with Folax grate cooler is shown in figure 11.

d) SLC-Kiln

The separate-line calciner kiln system was introduced in 1977 in Japan. Since then quite a
number of systems have been put into operation with capacities up to 6000 tpd.

An example of an SLC-kiln is shown in figure 12.

An SLC-kiln can be equipped with 4-6 stages in the double string preheater. This kiln type is
mainly used for large units above 3000 tpd where 2 strings would be used in any case. By using
separate strings for kiln and calciner some very important operational advantages are obtained.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
The double string cyclone preheater has the precalciner placed parallel to the kiln riser duct. Thus
the combustion in the calciner takes place in heated atmospheric air drawn from the cooler
through a separate tertiary air duct. The exhaust gas from the calciner and the kiln are conducted
through the two independent strings without being mixed.

The kiln is started on the kiln string operating as a normal SP-kiln.

The material from the bottom stage is led, via a two-way damper, to the kiln. Towards the end of
this preparatory period the calciner string is preheated with hot air from the cooler/calciner
start-up burner. When the output reaches 35-40% of the nominal kiln capacity, the 2-way damper
is switched to direct the raw meal to the calciner, feed to the calciner string is started as well as
the burner in the calciner; this takes a few minutes.

Within a few hours full kiln capacity is obtained.

The normal capacity range is 3000 tpd and upwards. There are no limitations for the maximum
capacity of this kiln type.

Like the ILC and SLC-S-kilns the SLC-kiln is a true calciner kiln. The degree of calcination of
the material introduced into the kiln is generally 90-95%.

The fuel distribution between the kiln and calciner is the same as mentioned for an ILC-kiln -
40-45% in kiln (315 kcal/kg) and the rest in the calciner.

The calciner vessel is placed parallel to the kiln riser duct. The vessel is very well suited for all
fuel types even low-volatile fuels as the combustion in the calciner takes place in hot
atmospheric air and (as an option) the combustion temperature in the calciner can be
indepedently controlled.

The gas retention time in the calciner unit should be minimum 2.7 seconds to ensure a safe
outburning of the fuel in the vessel. The gas velocities in the calciner are usually kept within
4.5-6.5 m/s.

The SLC-system offers an independent and accurate draught control for the kiln and calciner
strings, adjusting speed of individual fans without damper in the tertiary air duct.

A by-pass can be fitted to the kiln string but only a range of 0-30% of the kiln gas is
recommended.

An example of a 5-stage SLC-kiln with grate cooler is shown in figure 13.

e) SLC-S Kiln
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

The separate-line calciner special kiln system was introduced in 1988 at the ROrdal cement plant.

An exmple of an SLC-S-kiln is shown in figure 14.

An SLC-S-kiln can be equipped with 4-6 stages in a single or double string preheater with the
precalciner placed in parallel to the kiln riser duct.

In connection with semi-dry systems the preheater can be equipped with 1-3 stages of cyclones.

The combustion in the calciner takes place in hot atmospheric air drawn from the grate cooler
through a separate tertiary air duct. The exhaust gas from the kiln and the calciner are mixed
before being introduced into the cyclone preheater.

The draught in both the kiln and the calciner is controlled by the same fan. This necessitates a
variable damper in the riser pipe in order to have some draught control.

The system offers the very advantageous feature that the temperature in the calciner vessel and
the riser pipe can be adjusted by means of the material dividing gate below the next lower
preheater stage.

With difficult burnable coal (f.inst. petcoke, anthracite) the temperature in the calciner can be
increased to 1000-10500C by sending more material to the riser pipe and less to the calciner.
owing to this feature optimum control of coating in riser pipe, outburning of fuel, stability of
preheater etc. is obtained without changes in preheater temperature profile.

The gas temperature of the lower preheater stage is in the normal range of 870-9000C; with this
temperature level the degree of calcination of the material is 90-95%.

The normal capacity range is 1500 tpd and upwards. There are no limitations for the maximum
capacity of this kiln type.

The calciner vessel is placed parallel to the kiln riser duct. The vessel has the above-mentioned
feature of temperature flexibility, making it very well suited for all fuel types even low-volatile
fuels.
The gas retention time in the calciner unit and duct should be minimum 2.7 seconds to ensure
safe outburning of the fuel in the vessel. By operation with temperatures of 10000C a retention
time of 1.5 seconds is sufficient.

The gas velocities in the calciner is usually 6.5-8.0 m/s.

A by-pass can be fitted to the kiln string but only a range of 0-30% of the kiln gas is
recommended.

An example of a 5-stage SLC-kiln with grate cooler is shown in figure 15.

F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

Components

This section will be dealing with the major components, i.e.

a) kiln, b)preheater, c) calciner and d) cooler.

a) Rotary Kiln

The FLS progrmme of rotary kilns includes 8 standard diameters, ranging from 3.45 m to 5.0 m.

The kiln is supplied with three supports as well as an extra support behind the cooler if the kiln is
fitted with a planetary cooler.

The standard kiln length is selected to 17 times the inner diameter of the kiln. The length is a
chosen standard length and can naturally be changed. The reason for a length/diameter ratio of 17
is very good operational experience with this ratio or even longer. It is well known that a slightly
longer kiln reduces the temperature in the riser pipe by 25-500C, which normally has a good
effect on the coating in the riser pipe, contrary to a shorter kiln, which will increase the
temperature and give a negative effect on the coating, resulting in more cleaning work in the riser
duct.

In relation to modernisation of existing kilns the ratio if often chosen to l/d ≈ 20-24 with
excellent result.

As an alternative a shorter l/d-ratio can be supplied for ILC, SLC and SLC-S-kiln systems, but it
is only recommendable in projects with extremely easy burnability.

The best argument for a short kiln on two supports is the investment. The investment for the kiln
tube can be reduced by 12-15%. The kiln tube represents in relation to a complete cement line
approximately 3% in price. The saving of a short kiln will thus be 0.4-0.5% of the total line.

This saving shall be carefully compared to a reduction in production factor due to coating
problems in the kiln riser pipe duct. This reduction in production due to operational problems can
easily be a question of percent.

The advantages/disadvantages can be listed as follows:

F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

Advantages

• Lower investment cost

• Space saving

• Reduction in specific fuel consumption due to lower radiation loss from kiln shell.

Disadvantages

• Increased kiln inlet temperature causes increased coating in kiln inlet box and riser pipe

• Increased danger of back-spillage from kiln

• Increased danger of kiln crank and consequential damage of kiln seals.

Be careful when selecting a new kiln.

For the different kiln types the thermal load is calculated in two ways:

a) The kiln load expressed as t clinker/m3 kiln volume/day

b) The burning zone load as Gcal/m2 kiln area/day.

The normal ranges are as follows:

SP ILC-E ILC SLC SLC-S

3
t/m /d 1.8-2.3 1.8-2.5 3.6-4.8 3.6-4.8 3.6-4.8
Gcal/m2/h 2.8-5.5 2.8-6.0 2.4-4.8 2.4-4.8 3.6-4.8

The burning zone load is calculated based on the following parameters:

Kiln diameter lower value 3.45 m

upper value 5.00 m
Heat consumption 730 kcal/kg clinker

Figure 16 is showing the kiln diameter as a function of the production and the burning zone load
as a function of the clinker production.
The figure shows that the burning zone load is considerably lower than for SP and ILC-E-kilns.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
The lifetime of the burning zone lining in a calciner kiln is expected to be higher than for an
SP/ILC-E-kiln calculated in kg bricks/tons clinker.

Normal values for calciner kilns is a brick consumption of 0.1-0.5 kg/tons clinker compared to
SP/ILC-E figures of 0.5 kg/tons under favourable conditions.

Preheaters LP-design

General Features and Advantages

The FLS-LP cyclones are the designation of a complete new line of preheater cyclones
introduced on the market in 1984.

The LP cyclones can be considered as traditional reversed flow cyclones with vertical axes with a
geometry which has been optimised for the application as preheater cyclones in 4 to 6- stage
dry-process cement kiln systems with low specific heat consumption (680-850 kcal/kg cl.)

Thus the design of the cyclones has been chosen in order to obtain the best compromise between
the fulfilment of the three basic requirements for a preheater cyclone, viz.:

1) Low pressure drop

2) High separating efficiency
3) Small physical dimensions.

Further, the LP cyclone design allows for two unique features:

1) An almost complete elimination of horizontal shelves inside the preheater, which eliminates
accumulation of raw meal and resulting occasional dust rushes, and

2) The omission of a central pipe (dip tube) in the bottom stage cyclone, in which such a pipe
usually has a very short lifetime owing to the high temperature level and the aggressive gases to
which it is exposed.

Finally, the LP cyclones for the two lowest preheater stages have as standard equipment the
special steep outlet cone part which (on FLS-preheaters with the old type of cyclones) to a great
extent has proven to eliminate cyclone plugging, even when operating with difficult raw
materials.

The design of the LP cyclones is based on the results of a large number of pilot plant tests made
at the FLS Research Center at Dania in Denmark.

The operational results from the first four preheaters with the LP cyclone in operation have
confirmed the expected low pressure drop coefficient of the LP cyclones, and at the same time it
has been shown that the regulating range for the preheaters is wider than anticipated, allowing for
stable operation at down to 60-70% of nominal capacity.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

Besides, the temperature profile in the preheaters indicates satisfactory separating efficiencies in
all the cyclones. Thus, with the LP cyclones the following separating efficiencies can be expected
in the individual stages in a 5-stage preheater (without calciner) working with a "normal" raw
meal:
Stage 1: 95%
Stage 2: 90%
Stage 3: 88%
Stage 4: 88%
Stage 5: 85%

The above-mentioned features for the LP cyclone preheaters can be summarized as follows:

Advantages of FLS-LP Cyclone Preheaters

1) Cyclone design allows for a low nominal pressure loss across preheater (down to 210 mm WG
for a 4-stage preheater and 280 mm WG for a 5-stage preheater).

2) Wide regulation range even for preheaters designed for minimum nominal pressure loss.

3) Low exit gas temperature due to high separating efficiency in all cyclone stages.

4) Special design of bottom stage cyclones eliminates the central pipe, which usually has a very
limited lifetime.

5) Cyclone design allows for a compact preheater tower.

6) No horizontal surfaces inside the cyclones on which material can accumulate.

7) Minimum number of cyclones in the preheater (up to 3500 tpd only one cyclone in each stage)
thus minimizing installation and maintenance costs.

8) Special cyclone outlet cone design minimizes plugging possibility in the two lowest preheater
stages.

LP Cyclone Types

The LP cyclone series consist of three main types, namely the top stage type, the intermediate
stage type and the bottom stage type. The design of the three main types are compared to
each other in figure 17.

Each of the cyclones are available in a number of standard diameters ranging up to 7.2 m for the
lined versions and to 7.05 m for the unlined versions, cf. section 3.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
Usually, a single string 4-stage preheater will contain four cyclones (one in each stage) all with
the same diameter but of different shape as one upper stage type, two intermediate types (with
different lining thickness and with and without steep outlet cone, respectively), and one bottom
stage type are included.

Correspondingly, a single-string 5-stage preheater will normally contain cyclones of two

diameters as the two upper cyclones usually have a somewhat smaller diameter than the three
lower cyclones. Also in this case all the cyclones in the preheater will be different, including one
upper stage type, two intermediate stage type (of different diameter) with normal cone, and one
intermediate type with steep outlet cone part and, finally, one bottom stage type.

As an alternative to the single cyclone in the first stage, single string LP cyclone preheaters can
be equipped with two parallel cyclones in the top stage. This will reduce the overall height
somewhat, but the installation costs will increase and in some situations it becomes difficult to
avoid an almost horizontal piece of exhaust gas duct after one of the cyclones, in which dust
accumulation might take place.

Some clients would be of the opinion that two parallel cyclones of relatively small diameter
would have better separating efficiency than a single top stage cyclone with a larger diameter.
Traditional cyclone theory indicates that the separating efficiency decreases with increasing
cyclone diameter. This is, however, not the case for LP cyclones working with cement raw meal.

Thus, for new preheaters it is recommended to apply a single top stage cyclone in each preheater
string except in the case the client specifically asks for two parallel top cyclones.

Calciners

All FLS calciners consist of a cylindrical vessel with a conical bottom. This design allows for
ample internal calciner volume while keeping the calciner weight and surface heat loss at a
minimum.

The different FLS-calciners are shown in figures 18, 19, 20, and 21.

A calciner is dimensioned in each case from the following parameters:

Kiln type
Fuel type to calciner
Kcal/kg to calciner
Retention time of gas/material
Gas velocity in the vessel
Gas velocity at calciner inlet.

In the ILC-E and the ILC system kiln exhaust gas enters the calciner axially in the bottom cone
and the calciner exhaust gas leaves the calciner through a side outlet at the top. Due to the rapid
increase in cross sectional area in the conical calciner bottom a strong vortex is created, ensuring
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
an effective mixing of fuel, raw meal and gas.

In the ILC calciner the mixing of fuel, raw meal and gas is further enhanced by the introduction
of the tertiary air duct tangentially on the calciner bottom cone. The resulting moderate swirl in
the calciner further increases theparticle retention time in the calciner.

In the SLC and SLC-S systems the hot tertiary air from the cooler enters the calciner through the
central inlet in the bottom cone and the calciner exhaust gas leaves the calciner either through a
side outlet or - in case high temperature operation is anticipated - through an outlet cone
connected to a central outlet pipe.

For the same production capacity, the calciners of the SLC and SLC-S systems are smaller in
dimensions than the ILC calciner as no kiln exhaust air is led through the calciner.
Also the calciners of the SLC and SLC-S systems are charac terised by a strong vortex formation
in the bottom cone, ensuring an effective mixing of raw meal, coal and tertiary air and a high
ratio between particle and gas retention time (measured at approximately 4 in an industrial-scale
calciner).

All FLS calciners can be fired with liquid, gaseous or solid fuels, the calciner burners being
placed so-as to ensure good distribution of fuel across the calciner cross section and rapid
ignition of the fuel.

A unique feature of the SLC-S system is that the temperature level inside the calciner is
independently controlled by the position of the dividing gate effecting the distribution of raw
meal conducted to the calciner and kiln riser duct, respectively.

Thus by leading a relatively larger amount of raw meal to the kiln riser duct, keeping the fuel
input to the calciner constant, the mean temperature in the calciner vessel can be brought up to
950-l0500C. The temperature of the exhaust gas and the degree of calcination of the raw meal
leaving the calciner will also increase.

However, on mixing with the kiln exhaust gas still containing uncalcined raw meal before
entering the calciner cyclone, the temperature of the gas/particle suspension is reduced to approx.
9000C. Thus, a normal temperature level is maintained in the calciner cyclone. Similarly, a
normal degree of calcination of 90-95% is maintained for the raw meal supplied to the kiln.

The increased temperature in the calciner ensures effective combustion even of fuels with low
reactivity, such as lowvolatile coals and anthracite.

A similar control of the temperature level in the calciner for the SLC system can be effected by a
dividing gate controlling the distribution of raw meal from the second lowest cyclone stage in the
calciner string between the calciner bottom and the calciner outlet pipe.
Coolers

FLS offer three different cooler types for the kiln system:
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

Planetary cooler UNAX

Grate cooler FOLAX
Rotary cooler DUAX

In connection with the different kilns the following cooler types can be applied:

Kiln Type SP ILC-E ILC SLC SLC-S

UNAX YES YES NO NO NO
FOLAX YES YES YES YES YES
DUAX YES YES YES YES YES

The coolers are described in a separate chapter.

Selection of Kiln Systems

When selecting a new kiln system there are many factors with
influence on the choice, such as

Production t/h
Clinker quality
Fuels to kiln/calciner
Burnability of raw mix
Composition of raw mix
Minor Components (by-pass)
Drying requirements of materials
Environmental requirements
Total flexibility
Production cost
Investment cost

Generally, the choice is made in close collaboration with the plant.

F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
• New Modern Cement Factory

• few months ago FLS signed a big contract with an Asean company for a complete new
production line. A flow sheet for the plant is shown in Appendix 22.

The plant is designed for

High reliability
Low fuel costs
Low power consumption

and includes

2 vertical raw mills ATOX with external material circulation

1 CF-silo for prehomogenisation of raw meal

A kiln feed system type Loss of Weight

An SLC-kiln, 5-stage preheater

A grate cooler equipped with double pass

A heat recovery system, including a preheater boiler, an air quench cooler boiler, a
turbine and a generator.

2 vertical coal mills ATOX for lignite grinding

1 Roller press Kbppern with material circulation, ball mill UMS and a high-efficiency
separator SEPAX for cement grinding.

This is in fact a modern cement factory satisfying the three points

High reliability
Low fuel costs
Low power consumption.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
5.1.C

Modernisation of Kiln Systems

Kiln systems which are in good maintenance condition can often be modernised with great
success.

A good deal of possibilities exist for a modernisation and a careful feasability study should be
carried out in each individual case to determine the optimal solution.

In the enclosure three conversions are shown, which have been carried out within the last few
years, viz.:

a) The Sanson plant

Modernized from a long dry kiln to a 4-stage SP-kiln.

b) The Vaiderrivas plant

Modernized from a 2-stage SP-kiln to a 5-stage SP-kiln.

c) The Rordal plant

Modernized from a wet-process plant to a semi-wet system, comprising a 2-stage SLC-S
calciner kiln with drier crushers.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
 F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR

SP: Suspension Preheater kiln

Conventional single-string cyclone preheater kiln without precalciner.

Features:

- Recommended capacity range: 700 - 2800 tpd

- Ratio of firing in riser duct: 0-151
- Bypass of kiln gas: 0 – 30% max
- 5.5m / 6.0m diameter

Special advantages:

- Planetary cooler can be employed

- Cheapest solution for small capacities
- Very low specific power consumption
- Simple operation - well suited for manual control
- Accepts higher input of chlorides than precalcining
systems with tertiary air duct (without bypass).

Flow sheet of 5-stage SP kiln system. Typical temperatures in the system are
indicated together with the negative pressure in the exhaust gas exit based on a
system designed for minimum overall pressure drop. Instead of the planetary Unax
cooler, the kiln may be provided with a Duax or a Folax cooler, especially if
subsequent conversion to the SLC precalcining system (by adding an extra calcining
string) is to be considered.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
ILC-E: In-line Calciner using Excess air
Single-string cyclone preheater kiln with a small precalciner built into the kiln riser duct.
Combustion air for precalciner is drawn through the kiln.

Features:
- Recommended capacity range: 900 - 3300 tpd
- Ratio of firing in calciner: 15-25%
- Bypass of kiln gas: 0 - 25% max

Special advantages:
- Planetary cooler can be employed
- Cheapest solution for medium capacities
- Low specific power consumption
- Easy operation due to high excess air percentage in kiln
- Low coating tendency in kiln inlet and riser duct
- Long kiln lining life due to stable kiln coating
- Accepts higher input of chlorides and sulphur than precalcining sy
stems with tertiary air duct (without bypass)

Flow sheet of ILC-E kiln system with 5-stage preheater and Unax cooler. Typical temperatures
in the system are indicated together with the negative pressure in
the exhaust gas exit based an a system designed for minimum overall pressure drop.
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
F.L.SMIDTH INTERNATIONAL CEMENT PRODUCTION SEMINAR
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 10

A New Perspective on the Cement Making Process

Through LINKman
A New Perspective on the Cement Making Process through LINKman

David Haspel
Ross Taylor
John Brooks
*Rai Shekber

Image Automation Limited

London UK

* Enexco consultants Ltd, Delhi

Abstract:

As a result of working with LINKman over the last decade, commissioning personnel have had the opportunity
to learn a great deal about the cement-making process. This is because a consistent approach to kiln control can
be implemented for many hours and days through the application of LINKman This means that the various
complex process interactions can be studied over long periods of time and relevant conclusions drawn. This
paper covers the key topics to arise out of this learning. They range from; NOx and its interpretation; process
dynamics, both short and long-term; combustion systems and their associated flame characteristics; together
with the introduction of the new concepts of flame sensitivity and robustness.

A New Perspective an the Cement Making Process through LINKman

1. Introduction

In 1980, Blue Circle Industries instigated a research project to look into the possible application of
microprocessors to the problem of cement kiln optimization. This project developed into the LINKman
technology which is a simple, multi-variable rule-based controller based on the concept of mimicking the
actions of the best operator in order to achieve optimum operation of the plants. This technology was launched
as LINKman into the commercial world in 1987, and has since progressed from being a Blue Circle Industries
proprietary technology to a product that is being successfully marketed throughout the world through the sira
subsidiary, image Automation Ltd with the full support of Blue Circle Industries. (Ref. 1)

The reference list has now reached 25 system (Table 1) and the installation sites span eleven countries,
including India. Indeed, LINKman operation now exceeds 400,000 hours of on-line kiln control. LINKman is
also being evaluated in the petrochemical, glass and pigment industries. The technology is well established and
recognised as being both practical and successful. For example Blue Circle Industries freely publish that fuel
based savings in excess of £4,000,000/year over seven manufacturing units are being achieved as a direct result
of the LINKman technology and its associated projects (Ref. 2)

It is through the application of this technology during the last ten years over the whole spectrum of cement
making processes (Lepol, Wet, Dry, Long Dry and Precalciner - Table 2) together with the experience of a wide
range of combustion system and fuels (coal, petroleum coke and general waste fuels) that has given us an
in-depth understanding of the processes that we believe is unavailable to those not fortunate to work with
LINKman. Some of the findings are outlined in the following sections.
2. NOx and Its Interpretation

The early LINK-man work in 1980 involved looking for a new on-line indicator of product quality. It was this
work within Blue Circle that proved the true value of NOx as a kiln control parameter by it being used as the
key input into the LINKman based high level control strategies. Fig. 1 shows the basic NOx burning zone
temperature relationship. However, many other workers have had difficulties in replicating these results.
Retrospectively, this is almost certainly due to us having had the good fortune to carry out our original work on
good recirculatory direct fired burners using standard steam coal. We were able to understand how NOx was
formed and behaved in a cement making kiln without the added complication of combustion induced problems.
Subsequently, as a result of using NOx monitors, we were able to obtain a much greater understanding of the
cement making process dynamics. For example Fig. 2 shows some of the relationship between NOx and kiln
power under a number of extreme kiln conditions.

However, as we have increasingly had to work with non-recirculatory, low primary air, multi-channel burners
around the world, we rapidly learnt that NOx can no longer be assumed to be a reliable sensor and hence we can
no longer use NOx as the key kiln control parameter for LINKman based process optimization on all plants.
Consequently, we have to use whichever indicators the process operator employs. This uncertainty in the
behaviour of NOx with respect to cement kiln dynamics is still a major topic for discussion around the world.
Image Automation has therefore been forced to use a range of other key control parameters whilst at the same
time having to develop strategies capable of controlling kilns with very poor unpredictable and sensitive flames.
We have discussed the topic of NOx under these conditions in a previous paper (Ref. 3). The key messages
were:

* NOx generation mainly arises from thermal NOx;

* NOx levels are very
dependent upon;

i) flame shape and kiln atmosphere; eg. NOx is degraded by CO > 3000ppm;

ii) burning zone material temperature; eg. NOx is a good indicatorof peak material temperature in the burning
zone under normal kiln operation, and hence product quality;

iii) insufflation conveying air; eg. conveying air has a significant impact on the level of NOx generated within
the kiln.

iv) process dynamics; eg. NOx and kiln power do not necessarily follow each other when the kiln is highly
fluxed due to excess alkalis being present in the burning zone;

We now need to add to this list the impact of secondary air temperature, fuel type and degree of excess air when
dealing with non-robust flames (see section 5).
3. Process Cycles (Short-Term)

If the cement kiln is fired by a well designed optimal burner, then we can learn a great deal about that particular
kiln by assuming that the NOx level is a good indicator of the peak material temperature in the burning zone
when operating under normal oxidizing conditions. Also as a result of studying NOx and kiln power on a four
stage dry process kiln under LINKman based control, we rapidly learnt to respect the power of the volatile
cycles within kilns. Although the thermal balance of the vaporization/condensation of the various alkali salts
and the decomposition and formulation of calcium sulphate are neutral when the kiln is -in equilibrium, the
thermal inertia of these processes when they are not in equilibrium, ie, the process is disturbed, prove to be very
large indeed. We can show that up to 10% of the total energy required to burn clinker is locked up in the volatile
cycle of a typical dry process kiln and that the magnitude of this energy is very temperature sensitive, ie, the
level of alkali recycling with the kiln/preheater system increases from 300% to 500%.

with less than a +500C change in the burning zone temperature once equilibrium has been reached, see Fig. 3. In
addition, the level of recycle can increase very dramatically under reducing conditions, for example, typically by
up to tenfold.

There is also a continuous but variable, building up and degradation of the kiln coating and this in turn modifies
the material flow rate entering the burning zone in spite of the constant feed rate into the system. This process is
therefore very rarely in equilibrium and hence the thermal inertia associated with the volatile cycle often tends
to be the largest single dynamic force to be reconciled when looking to optimum control of the plant.

Similar volatile cycle related effects have been seen on the whole spectrum of cement making processes,
ranging from Lepol kiln grates, right through to the wet process kilns.

4. Process Cycles (Long-Term)

It is usually assued by most people that a cement kiln, because of its low total feed in the system, will have a
half life of a matter of hairs for achieving a new equilibrium state, and that there is little or no carry over from
one days operation to the next. However, this is not what the plant operator knows to be true since a kiln can run
with high outputs for very many days and then slowly degrade for no apparent reason to a below average
performance. Sometime later it then tends to drift back to a good performance mode again.

In our opinion, the reason for this is mainly due to cement kiln dynamics having a half life of not hours but
many days due to the influence of the kiln coating. This means that the history of the kiln is contained within the
coating itself and that this history will continue to influence the burning zone dynamics for many days after it
has been formed. For example, it can take up to twenty days for a dry process kiln to rebuild the coating of a
kiln to enable the output to reach its maximum. Fig. 4 shows this effect through the impact of LINKman on the
output of a dry kiln process over a period of 20 days. Any Periods Of excess overburning, however short, will
tend to build a high volatile content coating. These volatiles when recycled with the fresh feed as the coating is
being reformed, will have a significant impact on the burning zone thermo-dynamics.
5. Combustion and Associated Flame Characteristics

This area is of increasing concern from the point of view of being in the business of commissioning process
optimising system. The control system has to work with whatever combustion system has been installed
together with the range of fuels the client has chosen to burn. With the advent of low primary air, or
multi-channel, burners chasing a theoretical 2 to 2.5 kcal/kg of clinker fuel saving per percentage reduction in
primary air* and the demise of the 20% simple direct fired system, we see a host of burner associated kiln
control problems that can, in certain circumstances, be very extreme indeed.

The main problem is that the majority of these new burners appear to be inherently, (i) non-recirculatory and,
(ii) very sensitive to kiln conditions. In the first case, this means that the clinker in the burning zone mist be
subject to partial reducing conditions which will result in a reduction of the clinker reactivity and an increase in
the magnitude of the volatile cycle. In the latter case, we look for a robust, or an insensitive flame which has the
ability to maintain a consistent heat release profile over the normal operating range of secondary air
temperatures and a practical range of excess oxygen.

Many of the combustion systems on which we have worked have shown poor NOx characterisation and/or low
NOx levels, high CO levels with "high" oxygen levels (> 2%) and exhibit flame sensitivity to a greater or lesser
degree. On some sites NOx remains very low at, 200 ppm to 300 ppm ie, the fuel NOx level only. In our opinion
it is much better for control purposes to have a good, tight robust flame, initially with a high NOx level. This
can later be used through LINKman based high level control to operate the kiln at substantially lower NOx
levels (often less then 50% of pre-control period), lower burning zone temperatures, and with a lower fuel
consumption.

* Equivalent to a theoretical maximum of 0.3% fuel saving/ 1% reduction in primary air on a

typical dry process plant.

This type of control then normally results in higher kiln outputs, more reactive clinker and increased refractory
life. Fig. 5 shows that the optimum kiln exit oxygen should be in the region of 0.5% to 1. 5% oxygen. At these
levels, CO should be in the region of 200 to 300 ppm. However, we often see OD levels in the region of 1,000
ppm to 10,000 ppm at 2% to 3% oxygen at the kiln inlet on kilns with multi-channel or low momentum burners.
This must not only greatly or even totally negate the apparent fuel savings, but also makes for very difficult
control of the kiln. In an extreme case NOx and oxygen mimic each other. Fig.6. shows NOx correlating with
oxygen on Works A (low momentary, non-robust. flame) and NOx correlating with amps on Works B (optimal
burner, robust flame). Appendix 1 discusses the characteristics of low primary air burners in more depth.

The type of problem that can occur is that when the controller is looking to increase the heat in the front end of
the kiln, the obvious call for an increase in fuel rate with a poor burner can result in the heat available in the
burning zone actually being reduced and the excess heat being either released at the back of the kiln, or resulting
an a large increase in unburnt CC). In addition, sensitive burners are prone to close coupling from the cooler in
that an increase in secondary air temperature can greatly increase the heat transfer within the burning zone
(especially with petroleum coke). This coupling mechanism in extreme cases also tends to be self-driven, until
again, some other factor takes over the burning zone dynamics (ie. material hold up in the burning zone). The
reverse can also be true in that a small reduction in secondary air temperature can cause the burning zone
temperature to enter a long-term dive; again until again some other mechanism breaks the cycle.

It has been shown by Fuel & Combustion Technology Ltd (FCT), Ref. 4, that the effect of a 3400C change in
secondary air temperature on the heat release profile of a high petroleum coke direct fired kiln can change the
heat release in the burning zone by over 30% (see Fig. 7a). The effect on low momentum multi-channel burners
is likely to be very similar or even greater. The kiln tends to be very difficult to handle with this type of fuel and
the particular design and operation of burner. We find in the practical world, that typically in excess of 10% of
fuel has to be taken off the coal mill once the burning zone/secondary air temperature cycle has started to build
up on these petroleum coke fired kilns. This can occur in a matter of minutes and fuel changes must be made
very rapidly in order to stop the burning zone from overheating, the coating being melted out and the NOx
becoming very high indeed.

In addition, sophisticated burners in our opinion tend to wear or distort after a period of time so that the quality
of combustion becomes degraded and this can be relatively rapid, le, in the order of two to three months. This is
normally manifested as high oxygen, high CC) conditions and kiln outputs normally end up being significantly
reduced.

It is our intention to study the new concepts of flame sensitivity and robustness in greater depth. Figs. 7a, 7b, 7c,
show the results of the initial FCT study which supports the concept that secondary air temperature has a very
significant impact (up to 30%) on the heat release pattern of a petroleum coke based flame. With the good
mixing of a more optimum burner, the effect can be seen to be much less significant (less than 5%). The
conclusions arising out of this and further work will be presented in due course.

6. Conclusion

The application of LINkman to a range of cement making processes has made, and we believe, will continue to
make a major impact on the fundamental understanding of the cement making process. For example:

(a) the early work proved the value of NOx as a good indicator of burning zone temperature and product
quality. This has however, been tempered by the fact that NOx is not always a good indicator of burning
zone and this is almost always due to poor combustion of the fuels being used.

(b) the application of LINKman has also confirmed the importance of volatile cycles on the cement making
process and much work has subsequently been carried out to study this further;

(c) the process dynamics of the cement making process have been shown to have a half life of many days (ten
days plus), and not a few hours as was previously thought.

(d) the present day design of burners appears to be far from perfect from a kiln operation point of view. With the
general adoption by the cement industry of the low primary air multi-channel burner has, since the 1974 oil
crises, probably generated more operational problem than it has solved, especially with respect to kiln control.
The new concept of flame sensitivity has been introduced to help address this problem where flame insensitivity
or flame robustness has been defined as the ability to maintain a consistent heat release profile over the normal
operating range of secondary air temperatures, and a practical range of excess oxygen together with fuel
changes. This robustness is particularly important with a process plant that can be as highly disturbed as the
cement making process, since it is often necessary to change the energy input into the process by at least ± 5%,
and sometimes by ± 10%. Under these circumstances, it is very important to be able to have these changes in
heat input reflected in the burning zone, and not reflected elsewhere in the process.

November 1990
APPENDIX I

Low Primary Air Burners

Low primary air burners are increasingly being specified because of the theoretical improvement in the energy
efficiency of the kiln due to using less low temperature primary air and more high temperature secondary air.
However, the danger in using low primary air burners is that these often result in a degradation of the quality of
combustion due to poor mixing between the primary air, fuel and the secondary air from the cooler. This results
in less efficient heat transfer within the kiln system as a whole, higher back-end temperatures and often
increased back-end Co. 'The poor quality combustion will also tend to reduce the quantity of NOx at the kiln
back-end due to the local reducing conditions within the kiln causing the NOx to be destroyed by the CO (2NO
+ 2CO - N2 + 2CO2).

The lower temperatures in the kiln will also lead to less thermal NOx being formed in its own right. These
conditions are normally related to low NOx with high CO and high levels of back-end oxygen ( >1%) . With a
good burner we would expect CO to be in the order of 200 ppm to 300 ppm and to increase rapidly when the
back-end oxygen drops below 0.7%. However, this is rarely seen and therefore one can reasonably conclude
that the present family of low NOx burners are the result of designing an "inefficient" burner. The increase in
back-end temperature and higher average CO levels will lead to an increase in heat losses and will counter the
theoretical energy efficiency savings since two-thirds of the heat of combustion of carbon is due to the
contribution of CO:

2C + 02 - 2CO (Ist stage contribution + 26.417 kcals/kg mol)

2CO + 02 - 2CO2 (2nd stage combustion + 67.635 kcals/kg mol)

The main reason for the poor combustion of the majority of low primary air burners lies in the fact that, when
designing these burners, the primary air velocity is generally not increased sufficiently to compensate for the
reduced primary air/volume.

Under these circumstances, the primary air and fuel does not have sufficient momentum. flux to entrain the
secondary air and create external recirculation of primary air and thus create the desired good mixing between
primary air, fuel and secondary air (Fig. 8) (Ref. 5, 6). With good mixing, ie with external recirculation, the
burner has good combustion burn out and hence low 00 with low oxygen levels, together with high heat transfer
in the burning zone. This, in turn, leads to lower back-end temperatures and hence lower exhaust gas losses and
greater thermal efficiency.

To reiterate, good mixing promotes efficient combustion of fuel and air, and generates less Co. NOx levels are
higher and the energy losses are smaller due to lower temperature levels, gas volumes and CO in the exhaust
gases- The external recirculation gases associated with good mixing also help protect the bricks from direct
impingement by the flame and the clinker feed from reducing conditions.

For a fixed primary air level, as expressed as a percentage of the stoichiometric requirement, increasing the
velocity reduces the amount of CO at the kiln back-end for constant level of back-end oxygen (see Fig. 9). Fig.
10 shows the inverse ratio between the optimum primary air requirement for good mixing and the primary air
percentage so that for very low primary air burners the velocity requirements approach sonic velocities (330
m/s). However, if the primary air volume exceeds the optimum value required for good mixing, then clearly
excessive heat losses will also result due to the low temperature of the excess primary air relative to the
secondary air -it displaces.

Dr Lowes et al have shown that it is the primary air momentum (mass x velocity) which needs to be optimised
(Ref 5). Hence, for low primary air burners, high velocities will be required which, in turn, call for higher nozzle
pressures and therefore more powerful blowers. However, the correct design of low primary air burners, given
sufficient velocity, can overcome the problem of poor mixing with the one provison that the high coal residue
particles will tend to be propelled to the back of the kiln before they are burnt generating high cc) conditions in
the back of the kiln. However, introducing swirl into the design of the burner will help increase particle
residence tire and promote good mixing.

The conclusion from these observations is that one must not only consider the simple heat balance effects of
reducing the primary air volume when designing low primary air burners. The effect of primary air velocity on
flame quality and its effects on mixing, back-end temperature, CO and NOx must also be considered. This is
especially Important when considering using NOx as a key control parameter for automatic kiln optimising
system such as LINKman.

List of References

1. An Expert System for Kiln Control"; M Hislop, A Lorimer

Second NCB International Seminar on Cement and Building
Materials, India 1989.

2. Department of Trade and Industry, Expert System Opportunities, Case Study 1, HMSO 1990.

3. "Practical Problems in the Use and Interpretation of NOx for

Kiln Control"; D W Haspel, World Cement, May 1989.

4. Private Communication with Fuel & Combustion Technology Ltd.

5. "Optimisation of the Design and operation of Coal Flames in

Cement Kilns"; T M Lowes, L T Evans. J. Inst. Energy, Dec 1989.

6. "NOx Generation and Control in Boiler and Furnace Plant"; T M

Lowes, C P Kerton. Inst. of Energy Symposium, Sep 8 1988.
 LINKman Applications Summary
Systems:

3 LINKman non-cement applications: Oil, Glass, Titanium dioxide

22 LINKman cement applications (32 kilns):

11 Wet process
10 Dry
4 Precalciner
3 Long Dry
2 Semi-wet
2 - Lepol

Protocols/Plant interface

11 OPTO
4 GPP
3 Allen Bradley
3 TCS
1 Gould
1 Bailey
1 Micon
1 Siemens
1 ABB

OPTO - Enables LINKman to communicate with plant via analogue signals

GPP- General Purpose Protocol - Enables LINKman to talk to a slave computer e.g a Siemens protocol is
available to talk to LINKman GPP

*- DEC MicroVAX computer rather than a DEC PDP

312/3
 LINKman REFERENCE LIST

SITE PROCESS IO

1. 1985 Aberthaw, 1 Dry process OPTO

UK kiln

2. 1986 Hope, 2 Dry process OPTO

UK kilns

3. 1986 Northfleet, 2 Semi-wet kilns OPTO

UK 1 Wet process kiln

4. 1987 Westbury, 2 Wet process OPTO

UK kilns

5. 1987 Cookstown, 1 Lepol process OPTO

N.Ireland kiln

6. 1987 Llandarcy, Lube Oil plant TCS

UK

7. 1988 Masons, 1 Wet process TCS

UK kiln

8. 1988 Lebec, 1 Long dry kiln OPTO

USA & cooler

9. 1988 Obourg, 2 Wet process OPTO

Belgium kilns /GPP
(Polab)

10. 1988 St. Helens, Float Glass process GPP(PDP)

UK

11. 1989 Artesia, 1 Wet process kiln GOULD

USA & cement mill/cooler

12. 1989, Le Havre, 1 Dry process kiln GPP

France (Modcomp)

3. 1989 Canakkale, 1 Precalciner kiln OPTO

Turkey & cooler 312/1
SITE PROCESS 10

14. 1990 Swanscombe, 2 Wet process OPTO

UK kilns

15. 1990 Tioxide, Titanium dioxide *OPTO

UK kiln

16. 1990 Ravena, 2 Wet process BAILEY

USA kilns

17. 1990 Atlanta, 2 Long dry *MICON

USA process kilns

18. 1990 Orissa, 1 Precalciner kiln/*ALLEN

India cooler/raw mill BRADLEY

19. 1990 Meron, 1 Lepol kiln *GPP

Italy 1 Precalciner kiln (Siemens)

20. 1990 Untervaz, 1 Dry process * ALLEN

Switzerland kiln BRADLEY

21. 1991 Dalmia, 1 Precalciner kiln/ *ALLEN

India cooler/raw mill BRADLEY

22. 1991 Plymstock, 1 Dry process *TCS

UK kiln

23. 1991 Jakarta, 2 Dry process OPTO

Indonesia kilns

24. 1991 Lichtenburg, 1 Dry process *Siemens

S. Africa kiln

25. 1991 Dong Hae, 1 Dry process ABB

S. Korea kiln

312/2
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 11

Investigation of Dust Generation Mechanisms in

Rotary Cement Kilns, 1984
tolerable level of dust loss and hence achieve a kiln fuel consumption in
the region of 1100 kcal/kg It is believed that 50% of the improvement
results from the reduced water content and 50% from the improved chains.

A desire to maximise the temperature difference between gas and solids

throughout the kiln has led to the decisions: (1) to abandon dust insuffla-
tion at the burner (the high-ash coal flame is already relatively cool),
(2) not to use scoops, (3) to include fuel in the slurry, which is reported
to yield heat in a region spanning the start of the calcining zone, (4) to
develop the Schugi Flexomix system of dust return and (5) to appropriately
adapt the chain system.

3. CHAIN SYSTEM DEVELOPMENT

The original kiln chains at O b u r g consisted of slurry lifters and

buckets followed by curtains. Initial attempts to feed lower moisture con-

tent slurries provided increased levels of dust emission and hack end tem-
perature. The first change made was to reduce the length of the chain
system to attempt to maintain a 6 to 7% moisture content in the material
leaving the chains. This arrangement yielded a back end temperature of
about 350°C and a high dust loss (30 to 40%). Observations indicated a
relatively low level of heat transfer in the nodule zone, a rapid drying of
the slurry (leaving a relatively short plastic zone near the back end),
poor dust capture e in the chain system

Further modifications were made to increase heat transfer and reduce

B E T without breaking nodules and to maintain a reasonably long plastic

zone (28 to 22% H2O to aid dust capture. T h e chains used are called
"thermochains" made up of garlands which exactly follow the curve of the
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 12

Choosing a Kiln Shell Cooling System

Ciments Calcia
Italcementi Group

M. Dancoisne
23 April 1997

CETIC
WORKSHOP:
“KILN SHELL COOLING”

SUMMARY OF WORK CARRIED OUT IN 1996-97

OF RESULTS OBTAINED

A. ADMINISTRATIVE REPORT

B. TECHNICAL REPORT
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 13

CETIC WORKSHP-"Kiln Shell Cooling"

work 1996-1997
 B. TECHNICAL REPORT

CONTENTS

INTRODUCTION

OBJECTIVES FOR THE WORKING GROUP

I. EXCHANGE OF INFORMATION AND EXPERIENCES

I.1 Life span of refractory bricks

I.2 Coating and heat loss through the shell

II. VARIOUS ASPECTS OF A FAN SYSTEM FOR SHELL COOLING

II.1 Design
II.2 Sizing
II.3 Control
II.4 Energy consumption
II.5 Economic considerations
II.6 Acoustics

III. STANDARD SIZING PARAMETERS

III.1 List of standard parameters

III.2 Notion of efficiency - Methods of optimisation

CONCLUSION
INTRODUCTION

66 of the 91 kilns belonging to the member companies of this working group, i.e. almost 3/4, are equipped
with a fan system for shell cooling.

Semi-dry process kilns are least likely to be fitted with such a system; dry process kilns, in particular those
with a precalciner, and wet process kilns, on the other hand, are most likely to be fitted with such a system,
since they are subject to greater thermal stress.

The coating created by the effect of forced cooling has a dual purpose:

- physical and chemical protection of the bricks, reducing the wear rate and increasing life span.

- thermal insulation on the outer part of the kiln, shell and tyre, reducing thermomechanical and
thermochemical stress.

OBJECTIVES FOR THE WORKING GROUP

The objectives defmed for the working group were as follows:

1.1 To exchange information and experiences in relation to:

- life span of refractory bricks

- coating and heat loss through the shell

1.2 To analyse the following aspects:

- design
- sizing
- control
- energy consumption
- economic considerations
- ecological considerations

1.3 To determine standard sizing or evaluation parameters:

- specific air flow rates

- air speeds
- others
- efficiency coefficient
1. EXCHANGE OF INFORMATION AND EXPERIENCES

1.1 Refractory life

Extending brick life is one of the main objectives of a shell cooling system.
But it is not often possible to achieve comparable situations with and without cooling, making it
possible to define the benefits in statistical terms: original cooling equipment, other modifications
carried out at the same time, etc.

Where a comparison is possible, the benefits in terms of specific consumption of refractory bricks
due to extended service life can be as much as 30 or even 50% in the area covered by the fans,
which can amount to a reduction in specific consumption for the complete kiln of up to 20%.

Many plants report kiln run extensions of several weeks up to the planned stops, as a result of air
blowing to cool the shell.

I.2 Coating and heat loss through the shell

The loss of heat through the kiln shell can be broken down into two phases:

heat flow through the walls of the kiln, i.e. through the coating, the refractory lining and the
steel shell.
heat flow dissipated at the surface of the shell by radiation and forced convection created by
the fans.

The heat flow is the same, of course, but transferred in two different ways.

1.2.1 Heat flow throuvh the walls of the kiln

Heat flow from the inside of the coating to the outside of the shell is governed by heat
exchange laws due to conduction through a multi-walled cylinder.
It depends on the difference in temperature between the two zones and the resulting
thermal resistance created by the series of walls.

This thermal resistance is expressed in “C/W using the formula:

R = ( )*1 ~(Lnh h-1,)

24 n An
where: L = length of wall
m/m-l = ratio of external/internal radii of layer n of the wall
Fur = thermal conductivity of layer n

Orders of magnitude of thermal conductivity X for the materials concerned are, as

W/m”C:

- coating : 1.2 to 1.7 (references: < 2)

- brick at 1000°C : 2103
(aluminium 75% < dolomite < magnesium-chromium < magnesium-spinel)
- steel shell : 50 to 60

This shows that the coating is almost twice as insulating as the brick, itself 20 times
more insulating than the steel.
Il. VARIOUS ASPECTS OF A FAN SYSTEM FOR SHELL COOLING

II.1 De&n

The following aspects are taken into consideration when designing a system.

ILL.1 Length of shell to he cooled

Initially, shell cooling systems were fitted to the hottest areas, in order to limit the
temperature from a thermomechanical point of view, and thus to prevent:
- deformation phenomena, of a cyclic or permanent nature
- thermochemical corrosion
- problems involving movement and slide between the shell and the tyre.

Work by the Group has demonstrated the benefit of using a shell fan system for the
coating itself, which tends to extend the length of the area cooled, particularly upstream,
and on the bricks which tend not to form a good coating, for example.

This extension is obviously linked to the inherent characteristics of each plant: raw meal,
process, kiln geometry, bricking pattern, . . .
The length of the area cooled by fans in relation to the total length of the kiln, on the
units examined, was on average 30 % , but ranged from 10 to 50 % .
For a three-support kiln, for example, the length cooled by fans can run from the
downstream tyre to the middle tyre, including the shell under the tyres.

11.1.2 Tvues of fan

Three types of fan have been reported:

- axial helicoidal fans, which are the most common

- fmed speed centrifugal fans with direct transmission
- double speed centrifugal fans with belt transmission

Diagrams are included in the appendices.

The choice is made on the basis of criteria applicable to each specific site (layout, size
. . . .>.

At this stage, it should be mentioned that:

- most fans are fixed speed; some units have double speed fans, but none have
variable speed fans, which do not appear to be beneficial.
- the fitting of a silencer, when this is necessary due to the regulations applicable to
the works or environmental considerations, is a major part of the cost of the project
but must be included at the outset.

11.1.3 Possible methods of installation

The following methods are possible:

- a bank of fans side by side, with diffusers, which can be fitted:

. either with one diffuser per fan
. or with 2 fans per diffuser, which offers additional control
options in terms of output

- air-box fed by 1 or 2 fans with ventilation holes (Obourg).

II.3 Control

All the kilns examined are equipped with an infrared kiln shell thermography system. However,
mere is not always an operational link between thermography and the cooling system.

A distinction can be made between:

interlocks with the status of the kiln (operating or down) and thermography
operation of the cooling system according to kiln temperature.

11.3.1 Kiln onerating/kiln down interlock

The system on the ENCI kiln is a good example:

- start-up and shut-down of the shell cooling fans with the main kiln motor (with shut-
down to avoid deformation of the shell due to cooling of a generator if the kiln is
stationary even on the barring gear).

11.3.2 Oneration of the cooling fan svstem

11.3.2.1 Automatic oneration

Operation of the fans is interfaced with shell thermography:

- start-up and shut-down of the fans at temperature thresholds, programmed into the
system
- forced operation of the fans if the scanner is not operating (breakdown or other
reason), which is also a safety interlock.

11.3.2.2 Manual - local oneration

The fans can be controlled:

- manually from the control room, regardless of the shell temperature

- locally regardless of the kiln status, in particular for maintenance purposes.

This example of the ENCI kiln is not, however, the most frequent case.
It requires:

- a reliable thermography system

- a method of matching the position of the thermographic image with the fans, which
may be in groups of 2’s or 3’s.

Some Italcementi units have double-speed fans, or fans fitted in pairs with one diffuser.
But the majority of plants include fmed speed fans which operate either permanently or
are operated “manually” by the operator.
 11.5.2 Oueratine costs

The main operating cost is electricity consumption by the fans.

According to the same sources, this is: Average Minimum Maximum

in $ / 1000 m3/h cooling air 0.013 0.003 0.03

in $ / t clinker 0.06 0.01 0.12

The cost of maintenance is minimal.

This must be off-set by savings made on other operating costs:

- cost of the refractory lining: 1 kWh is approximately equal to

100 g dolomite brick ’
50 g magnesium brick

- improved kiln productivity:

. regular operation due to stability of the coating: reduced coating falls,
avalanches, and consequences on the clinker which may result from the same.
. reduced number of stoppages: production capacity, heating costs.

- costs associated with the thermomechanical problems avoided, replacement of shell,

tyres , etc.

These savings cannot be quantified on a general basis.

II.6 Ecolokal aspects - acoustics

The replacement of a water spraying device by a fan system involves an ecological aspect.
The essential aspect is that of acoustics.

11.6.1 Summarv of acoustic data

- Total acoustic power Lwt of N fans with an acoustic power of Lw is, in dB(A):
Lwt = Lw + 10 log N

- At a distance R (m) from a group of fans with a total acoustic power of Lwt,
the acoustic pressure level (“noise”) Lpt is then:
Lpt = Lwt - 8 - 20 log R

e.g.: A bank of 15 fans at 80 dB(A) each produces noise at 100 m expressed as:
Lpt = 80 + 10 log 15 - 8 - 20 log 100 = 44 dB(A) .

11.6.2 Acoustic Dower index

This also involves:

- output (Q) from the fan
- length (D) of shell cooled by fans

One can define Qs = Q/D in m3/hour per m.

The acoustic power index is thus defined as: Lws = Lw - 10 log Qs

e.g. for a 15 000 d/h fan cooling I m of shell, this gives:

Lws = 80 - IO log 15 000 = 38 dB(A)
CONCLUSION

The proposal for continued work by the workshop put forward by Mr. Kerton includes the conclusions of the
work completed to date.

Reporting a “positive influence towards the formation and retention of an internal coating. The advantages
of this cooling operation will not become apparent if the coating does not form or if it is greatly influenced
by internal parameters, which today are not necessarily the same as they were in the past”,

“The members are unanimous in proposing a logical extension to their work, in order to codify the other
parameters which influence protection of kiln shells”.

- Characteristics of the raw meal and the fuel

- Selection of refractories
- Selection of steel for shell.

Shell protection is also the main aim of the “STM system to assist in the diagnosis of thermomechanical risks
in rotary kilns”, developed by Lafarge and F.C.B.
A presentation on this system was given by the “Maintenance” working group to the CETIC Technical
Commission in June 1996.

In fact, although cooling of the shell by fans is the main method of protection, it is not always enough and
action on other parameters, such as those mentioned above, may in fact prove to be necessary.
 LIST OF REFERENCES

Itakementi - SoA BerPamo - CETIC - “Kiln shell cooling” Workshop

l Salerno, 14-15 March 1996 - Situation at Italcementi - Objectives of kiln shell cooling - General aspects
- Acoustic aspect.

l Cauldon, 26-27 September 1996 - Experience of cooling the kiln burning zone at Salerno - Heat losses
with and without cooling - Estimating the thickness of the coating.

ENCI - Maastricht - CETIC - “Kiln shell cooling” Workshop

l Control of cooling fans in the burning zone of kiln 8 - F. Erens, September 96.

l Analysis of the refractory lining of kiln 8 - F. Erens, September 96.

l Experiments with water injection in planetary cooler of kiln 8 - F. Erens, September 96.

l Explosion in kiln 8 - Investigation of causes - Recommendations - Plan of approach.

F. Erens. March 97.

Minutes of meetings of the Workshou

l Meetings of 14 and 15 March 1996 in Salerno - J.Ph. Naudin. July 96.

l Meeting of the Technical Commission - Maastricht. 5 June 1996 - J.Ph. Naudin

l Meeting in Cauldon on 26 September 1996 - J.Ph. Naudin. February 1997

l Meeting of 7 March 1997 at Lafarge Ciments in Saint-Cloud - M. Dancoisne. March 1997.

l Summary of the Group’s work for the 1996-97 year - M. Dancoisne. April 1997
 LIST OF APPENDICES

OUESTIONNAIRE - SURVEY ON SHELL COOLING EOUIPMENT

NUMBER OF KILNS WITH A SHELL COOLING SYSTEM

MAIN PARAMETERS FOR SHELL COOLING FAN SYSTEMS
GENERAL CHARACTERISTICS OF BURNING LINES STUDIED
GENERAL CHARACTERISTICS OF COOLING SYSTEMS STUDIED
OPERATING CHARACTERISTICS OF FAN COOLING SYSTEMS
HISTOGRAM FOR KILNS WITH COOLING EQUIPMENT BY PROCESS
HISTOGRAM SHOWING DENSITY OF COOLING AIR
HISTOGRAM SHOWING AIR SPEEDS AT THE DIFFUSER OUTLET
HISTOGRAM SHOWING SPECIFIC INSTALLED POWER

DIAGRAMS - TYPES OF SHELL COOLING UNITS

ITALCEMENTI - SALERNO
ENCI - MAASTRICHT

DIAGRAMS - TYPES OF KILN SHELL FANS

AXIAL FAN
CENTRIFUGAL FAN WITH BELT TRANSMISSION
ASSEMBLY WITH TWO CENTRIFUGAL FANS IN ONE DIFFUSER

DIFFUSER GEOMETRY - POSITION AND CHARACTERISTIC DIMENSIONS

HEAT LOSS THROUGH CYLINDRICAL KILNS

HEAT LOSS COEFFICIENT FOR A ROTARY KILN SHELL

HEAT LOSS AS A FUNCTION OF TEMPERATURE, FOR DIFFERENT KILN DIAMETERS AND
AIR SPEEDS
HEAT LOSS COEFFICIENT FOR DIFFERENT KILN DIAMETERS AS A FUNCTION OF AIR
SPEED

CRITERIA FOR EVALUATING LOSSES THROUGH SHELLS

HEAT LOSSES WITH AND WITHOUT FORCED VENTILATION

DIAGRAM SHOWING FORCED CONVECTION MEASUREMENT DEVICE

LAFARGE - F.C.B. “STM” SYSTEM

SUMMARY SCREEN FOR THE SYSTEM

APPLICATION EXAMPLE
RESOURCES REQUIRED
 total total total

JJE CIRCLE UK 2 0 5
USA 1 2
-.-

CALCIA 40% 2 0 0

I
0

=I=
C.B.R. 83% 2 2 -I 2 I 2

ENCI 100% 1 1

NTERMOSELLE 100% 1 1

ITALCEMENTI 22 65% 6 5 22 16 3 / 1

_AFARGE F r 100% 2 2 6 6
I
-_ ___--..__-

OBOURG 100%
2;2-
TOTAL 73% 14 12 14 12
CC fL&xs Lyi*
73% 86% 63% 86%

TIC - Refroidissement Viroles de Fours M.Oancoisne - Avril 1997

 -
Ifl
ii!
:: 0
G
-’
2

N
SH
1760
W
H
1015
W
H
101:
Mas
vs
285C
!
-
Rum
VSF
33OC
Is0 M a t -P o r-
-
v s s ~__
v s s vs:
2000 1780 1201
Sar
VSP VSF
900
__
VIII

?OO(
-
7 9 0 1 8 8 0 112951 1 2 8 C 1520 1370 137(
__-.-
100
111
5
1,2
100 100
57,9
5
1
57,s
5
1 2,6
-__ -
1
1
-_
1
i

.-A - ,2,6,
836

100
99
-
830

75
86
1
- I-
810
__.
75
50,6
840

75
46,7
89C
-
1oc
44,:
8X
-~
75
80C 900

-100 -
33,6
800

50
33,:

l--t--t
1,2,7 1,2 I,2 -
1,2 106 2.6.3 4.3.6.7 1.5
_---
90110 90110 5oi50 60/40 60140 100 1W 75-25 50150 72/X! 2158/9/l 1W
-
3,95 5,30 3 . 9 03,753,75 4,40 5,50 3,K 3,60
_ _ 4,77
_- 4,42
- - 6,lO 6,lO 4,42 4,4i 5.4JJ -
4,oc
- __-- ~.-.-
99 150 127 255 25.5 12,7 12,7 18,9 9 . 6
18.9
----I.I 8 8 1 12.6
8.8 2 6 20.4 196
--L 9.i 8.0 10.2
ALL-2
385 1 308 1 4.7 1 483 4.4 4,6 4,6 60 5,6
586 9,l
%I 43 4,4 3,3 3,3
- .-
196 137 152 165
185 67 1 2 0 667.8
120 7.8 180 90 40 48 58
-
17,8 20,3 20,3 17,7 16,8 16.5 16.7 19,5
-11.3
11,3
- -
M TM TM B TB M B M T I, 1
____- __.
_._- ---
_----
M B’ 8’ M MM T MTM M M B M
8’ - B’ B’ B TB
TB TTMM MM - B B
M 00 T MTM M M M M
=ZZL= - - -

. _

M Dancome AwlI 19’ii‘

-
 .:
L.-L.-

T
-
C H N N W W Ant Har MOk Mi3S Cal Cal is0 Mat POI 5 10
_. - _.
500 1725 2070 176 2900 560 320t 2850 1201 1800 2000 1780 120‘0 GC IO 300( 1
-- -
1 5 13 13 13 1 1 15 9 4 22 27 18 16 15 -T 2+2
144 43 10 -- 15 80 If 18 20.2 23,4 21 23,i 5 101 i f 54/i! >
__- --.
20 12 40 3( 30 42 48 35 48
--.

39E03
0 0 0 0 0 0 0.5 0.5 0 0,s I_. 0 0
130 225 792 545 421 336 35:3 2% 3 258
-
19 a 30 38 75
I --z 19 29 18 -. - -
550 --~
- -.. 535 1170 1311 1242 992 1395 - -3
3% 350 847 254 164<
--_
%a 7.5 20 15 29 15 19 -- -13
_- 3,6 29
_-
75
_- -...__ _ 102 9( -75
- 105 68 68 : 80 68 105

5-
-..~- - IPC ZM
_. 1 mars
-~. --l
__-. 66.2 a9 60.7 -64:
- _-.. -~ 65 j9 90
" " n YIlLrl- " -4 In 3( 46
I,5 10 4 1.5 22 4 2
--L2 4
-d.- 5.5 5,5 6 5,5 6 22
20 - 1 4 23 214 42 121 95 94 72 _- 13: i i i1 %i-
-
0.15 - O.lC 0.27 0,11 0,31 3. 0.31 0.31 0.3: -0,21- ,,5. 2 034
L 0,27 0-! io_- 0,6
02 -0.6 0,2 1.8 0.3 2,4 2,2 3.2 1,9 1.6 IcI 0.7 .-z
-J-- - 0.08 3,2
1.35 yz 5.5 5,5 5.5 506 5,5
_- -~-. --.
20.3 346 5 5 49,6 49.E
L 121 1oc
0,os 0,OG 53-l 3 0.17 0.16 0.2' 0,24
-. - .- ~ -.- -0.31 -
0,09
v
).OSI 0,261 0.141 1,2 I(- 02 0.25 2,4 1.2 1.7 1 .a 1.2 1,l 3,05 2.4
. . .-
1 1 1 1 = 1 1 1 -i- 1
- -. ~.- 1 1 1 1 00::
0 0 1 --. 0
_~ 1 1 1 0 0 0 1 i’l L
1 1 0 0 1 1 0 1 1 0 1 1 57”.‘
A+tF4
~_- seuil seui seuil 2v 2v AJM 2v
-O-L- min cc 30 30 50 79 xi- so 70 55 40 50
120 79 50 60 70 55 50 70 - 74 i 50 i
10,7 20 30.8 STi 15.9 26,9 22.4 28.; iii ! 7
4s 185
__I- 21,9 16 29 ia 20 -.- _ 15 3
280 380 350 250 200
----I -~
360 ;uis: :uiss cuiss cuiss CUISS :uis: CUIS: -- -
0.03 0.005 007 i-s D.08 0,12 0.03 I0 Of i..- 0,08 0.11 012
- --A--.- -L- G L o,oE _.

4$El%
,030 0.009 0.009 0,003 0,01 &ii& 0.01 I 0,024 0.007 0 . 0 1 t 0.011 0.010 0.02 I 0.014
_..
nini mini mini mini mini mini mini
- mini
- mini mini 1 mini mini mini
- mini
- --.-
sur "wiS+r&du, c
P.th - rcc(*,.bruil I -
gm P” :ONO w Pal
pfr& .I 5% compar Eomplr =-_=
3 $
TIC - Refroidissement Viroles de Fours h4 Dancoisne - Awl 1997
_- 1
:,.‘ .,:
( :,‘ : .; .‘.-: ..;;,,, :,. . . ,:; :,. .,: .: : . . .’ :
I
/ ,
0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
co co e N
--
8
C5
w
7
0
N
w
7
I
L
 I
/
/
E:
01
0:
s
I
L
 GEOMETRIE DES DIFFUSEURS
-_------------_------------

POSITION ET DIMENSIONS
CARACTERISTIQUES
i fcr tT,z:rJ r+N 0 CtiA.=plcx @S\ST7c 31,u E. rJi,ah,;

Ventiladur
F aA
 1
- - - tlo = 7 m I I I
I
0 , =5m .I-

- - - 0 , =3m

= 2d”c wL in m/s :
%
EW = 0.9 I Fl

200 300 coo 500

Wandoberflkhentemperatur 9, i n OC

Temperatura di parete
 A - PAR VENTILATION FORCEE
a-1 -;Coewh -cc h)i

I rayorlnenient

s,. /la-b.J-ct( Cazu’ecfi*

&;:?,.
,$$
<.:.:, ..J , convection naturelle

q1= Q,;,*(tu-te)*Cp,;r
0,0082) . t ‘*333 kcal./h.m2.YI
q2 = a * At (viroies-air)
QG, te (=hcUr-- &.I

qtoc = q1+ qz

B-SANSVENTILATIONFORCEE

I , rayonnement

At = (tviroles-t air)
t ~kut - e CL

a =(8,2+8,5)+(0,0073+ 0.0082).t'.3" kcaYh.m2."C

q tot =a*At
 225

150 ~&&$a$
A . . . ..’
100 I
Ir 28.88 ,

L -26.80
I
 u i7-iaLa-r -I- 1-l 4X2.4 C4RTlfw~
INSTALLATION NON EQUIPEE
DUNE THERMOGW.PHLE INSTALLATION EQUIPEE
DUNE THERMOGFWPHIE

8
I hESURE D E S TE!.WEPATURES
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 14

Criteria Governing the Use of Short Rotary Kilns

for the Production of Cement
Criteria governing the use of
short rotary kilns for the production
of cement
Authors: H. Herchenbach and A. Wolter, Cologne
Translation from the journal ZEMENT-KALK-GIPS", No. 9/1982, P. 450-455
Bauverlag GmbH, Wiesbaden

Criteria governing the use of short rotary kilns

for the production of cement *)
Authors: H. Herchenbach and A. Wolter, Cologne

Abstract:
Subject to complete pre-calcining taking place in the cyclone diameter. Restrictions rather exist for thermal cross-sectional
preheater with PYROCLON* firing system, a shorter rotary loads and the gas velocity inside the kiln tube. ,
kiln like the PYRORAP09 kiln will be better matched to the For finding an answer to the query about the actually required
requirements of process engineering because the resulting. kiln length for heating the raw meal up to sintering temper
faster clinker burning enables profiting of advantages of ature and sintering reaction, a kiln of &5 m diameter and
reaction kinetics. Moreover, the PYRORAPID3 kiln is not only 47 rn length, fed with meal precalcined at a rate of 60-70%,
characterized by somewhat more favourable capital expend has been used for determining what tube length Vill be
iture but also associated with significant saving of operating needed for attaining the remaining 30-35% calcining. The
costs compared with conventional rotary kiln plants. Fears as results obtained have been shown in illustration 1. Residual
regards impaired operational safety of short rotary kilns have decarbonation has definitely been completed not earlier than
not been confirmed but on the contrary, the specific load on after 15-20 m tube length 2). Provided residual decarbonation
the sintering zone has been markedly lowered in spite of a would take place in a PYROCLONO calciner, a tube length of
higher kiln space load. This phenomenon had, in turn, approx. 30 m would be sufficient. This means that specificed
positive effects on the service life of refractory linings and kiln kiln space loads of 5.5 to 6.0 t/M3 can be attained.
operation. Provided, special conditions are being fulfilled,
these results have opened up new opportunities for optimizing
clinker production and reducing the costs involved.
2. Considerations as to reaction kinetics and
process engineering
1. Introduction Thermal decomposition of the calcium carbonate, of the clay
Upon watching the advances made with regard to cement minerals and other suspended raw-meal components results
clinker calcining during the last decades, it becomes obvious in the formation of chemically active disintegration products).
that continuous attempts have been made to achieve a Their surfaces display greater chemical reactivity than
specific reduction of the rotary kiln dimensions. This is a materials heat treated over long periods. Prior to the
logical consequence of the fact that although this equipment commencement of the most important clinker reaction, i.e.
meets the specific process requirements, it is a source of
high operating costs. In this respect reference should be
made to poor heat transfer conditions known of endothermal
processes, high capital expenditure and cost of wear for
mechanical units and refractory lining, as well as the resulting
plant availability.
Achievements gradually made in the past towards this goal
have been raw-material preheating, dewatering of the clay
components and decarbonation, all of which have been
implemented within other machines. Sign posts along this
line have been the grate preheater, the cyclone preheater
with calciner, as e.g., for the PYROCLONO system. This
system enables an almost complete raw-meal decarbonation
accompanied by heating up to about 900oC.)
For lack of exact methods for calculating the conditions of
heat transfer within a rotary kiln, its dimensions can up to
now be determined only with the aid of experiments made
empirically. HUMBOLDT WEDAG base standard designs on a
diameter/length ratio of about 1:14.0 for kilns with cyclone
preheater but without PYROCLONO system. An interesting
aspect in this respect is that for none of the more than 500
preheater/rotary kiln systems known to us being operated, kiln inlet Tyre I
the rotary kiln can be considered too short related to its
Fig. 1: Empirical determination of the calcining zone as a function of
Enlarged version of the lecture read on the occasion of the VDZ conference held in
the temperature of the material discharged from the kiln (precalcined
DOsseldorf in February 1982. at a rate of 65 to 70 01o).
2
 the formation of alites (C+C2S - C3S) the heating process is 3) Vice versa, it is analogously possible to lower the lime
unavoidably accompanied by the reactions forming calcium standard at quick burning while maintaining the strengtt
aluminates, calcium aluminate ferrite and, in particular belite development pattern.
(C2S). The longer the heating towards alite stability, the more
will the belite already formed and the remaining free lime 4) The not yet completely examined influences of secondat
have an opportunity for recrystallization. This will hinder the components, such as alkalies, sulfates, fluorides or heav
delicate process of nucleus formation and complicate further metals may be more pronounced for quick burning as
growth of alite crystaIS5). Consequently, for raw material of they frequently affect, in particular the formation of alite
standard burning properties, quick heating up will significantly nuclei5).
shorten the time required for finishing burning. Apart from In addition, quick burning combined with rapid cooling result
that, any acceleration of lime bonding will have an over in structural activation of the clinker phases' 1)12) . The cryst.
proportional effect because the velocity of lime bonding lattice may be characterized by disarrangement and foreign
steadily decreases with the duration of burning (illustration 2)6) elements be incorporated beyond equilibrium.
This aspect is of specific importance for the development of
a belite cement because the required activation of clinker o
low lime content can, if at all, probably be realized only
during a rapid burning process.

A direct consequence of increased foreign element incorpo

ration is the decreased tendency to form circuits of trace
elements. Doubtlessly, circuit phenomena of sulfates and
chlorides known of conventional preheater/kiln systems ha,
to be corrected downward for the rapid burning process. Th
shorter retention time within the high-temperature zone, the
trend to lower absolute temperatures within the sintering
zone explained earlier and the improved conditions for
incorporation of the above substances in the crystal lattice
the clinker, complicate evaporation and thermal recirculatior
to the calciner and/or the preheater. These facts enable
drawing the following conclusions:
1) The PYRORAPIDO kiln is of particularly little hazard to
theenvironment and discharges smaller amounts of
pollutant with the waste gas. This is enhanced by the fifth
stage of the cyclone preheater.
Problems of accretions and clogging within the calciner
and the preheater, caused by circuit processes, occur
Fig. 2: Influence of heating gradient on the duration of finishing less frequently and in most cases do not require the
burning (schematic).
Because of the logarithmic time dependence of lime bonding the installation of a partial-gas withdrawing system which
heating gradient has an overproportional effect. later results in higher operating costs.
Rapid clinker burning will be unsuitable whenever harmft
matter, such as sulfates and alkalies, must not be
Apart from the fact that the kiln tube portion needed for incorporated in the cement clinker for attaining specific
calcining is superfluous at complete precalcining, changing cement properties, and therefore have to be removed
over to a shorter and quickly revolving kiln enhances from the system over a bypass.
profiting of the reaction advantage just described and, there 4) The same applies if the raw meal or the hot meal contain
fore, is has to be considered a correct adaptation measure as mineralogically more heterogenic matter which is known
regards process engineering. to have very poor reactivity.
The fact of clinker mineralization being quicker in a
PYRORAPID'9 kiln than during conventional clinker burning, 5) Similarly, the PYRORAPID9 kiln providing for complete
influences clinker and cement properties: decarbonation in the PYROCLONO will not be of advantag
1) High reactivity.of the disintegration products originating when firing lumpy (waste) fuel. The combustion heat of
from the raw-meal components reduces the necessity of these fuels should preferably be utilized for residual
pronounced sintering of the clinker granules which, in calcining in the kiln tube 13).
turn, means improved grindability)). These statements are basically applicable to kilns subject to
2) In spite of the lower energy required for grinding, early specific high load and characterized by extensive precal
compressive strength of the cement is increased at cining. On the one part the withdrawal.of waste-gas portions
identical lime standard, compared with clinker burned too is increased in these plants because of the reduced specific
slowiy9)10).
waste-gas volume and, on the other part the evaporation -
coefficients of the volatile secondary elements are lower thar
for preheater kilns without second firing system or even long
dry-process kilns without preheater. Therefore, higher
3 -
portions of otherwise volatile secondary components are preheater and precalciner respectively are more dis
bonded in the clinker for this type of kiln, too. advantageous and entail, in particular a significantly
An argument frequently brought forward against the short higher gas temperature in the inlet chamber 14) . The
kiln is the high temperature in the inlet chamber with the reason is the limit existing for the fuel quantity that can
be burned in the second firing system, i.e. the increase in
allegedly accompanying accretion problems.
the fuel quantity burned in the rotary kiln resulting in
Such thesis being: subject to identical specific kiln space increased dust circuits.
load, the inlet chamber temperature is said to be lower at
increased length/diameter ratio (UD - 13). We would like to Moreover, the flame in a rotary kiln of larger diameter can
be made shorter at the advantage of rapid burning than in
comment as follows on this statement:
a longer kiln of smaller diameter.
1) An identical specific kiln space load (expressed in UM3 d) 4) Even the higher inlet chamber temperatures resulting from
calls for an almost identical degree of precalcining for the substantial precalcining (see under 1 above) during the
raw meal at the kiln inlet. Proof has been fumished that PYRORAPID" process, do not necessarily call for more
the inlet chamber temperature is closely associated with cleaning expenditure because the circulating factors of
the degree of precalcining. This statement is also valid for harmful matter decrease (see statements made above) and
standard preheater kilns without PYROCLON" system. because condensation is shifted towards the preheater/
Heat radiation over the outer kiln wall is of certain precalciner system. However, condensation is less critical
importance for the temperature in the inlet chamber Le' there because it occurs rather in the dispersed meal flow
for kilns of identical specific volume load. The surface but to a smaller extent at the walls.
being by about 14 % smaller at UD = 10 against a ratio
of UID - 13, however, results in a temperature increase of 3 ' Comparison of capital expenditure and
no more than approx. 15 K for the kiln waste gas which is
an order of magnitude of 1 %. operation costs '
The rough cost comparison has been based on a combination
3) In practice a narrower IdIn of comparatively higher gas of the 5-stage preheater with short kiln carried on 2
velocities entails more pronounced dust circuits. However, supporting roller assemblies and the other well-known
for exergetic reasons dust circuits between rotary kiln and systems (illustration 3).

.
 At an assumed (always identical) plant capacity of 1500 t/day Table 2
and a fuel portion burned in the second firing system which Comparison of capital expenditure based on
has been matched to the specific operating conditions, the PYRORAPID kiln
characteristic data listed in table I will result.
Version A% B% C% D%
Table 1
Basic process engineering data for a comparison of Preheater 100 89 81 71
capital expenditure and operating costs Refract. material 100 83 80 70
Plant version Unity A B C D for preheater
Rated clinker t/d 1500 1500 1500 1500 Rotary kiln + kiln head 100 122 138 171
production Tertiary air line 100 120 0 0
Kiln dimensions m 3.6 x 3.6 x 3.8 x 4.2 x 60 Bricklining material
38 50 58 (Tertiary air line, kiln, 100 ill 98 119
(diam. x length) kiln head)
Preheater PR396 PR396 PS396 3962 Waste-gas fan 100 100 102 100
2 2 Foundations (for kiln) 100 126 130 156
5stage 4stage 4stage 4stage Preheater building 100 86 86 59
Fuel portion burnt Erection costs 100 94 84 86
in second firing % 60 50 30 - (for complete brickwork)
system Erection costs 100 100 100 100
Inside kiln cross m2 8.0 8.0 9.1 11.3 (machinery and equipment)
section Total capital expenditure 100 104.5 101.7 108.7
Inside kiln volume m3 300 400 490 680
Space required m2 750 950 930 1000 It implies approx. 3.5% savings because of a reduced
(preheater + kiln) machine weight, approx. 13 % savings for the refractory
brickwork and up to 30 % savings for foundations (non
Spec. heat kJ/kg 3245 3350 3350 3350
existence of the 3rd roller assembly) compared with a kiln
requirement of an L/D ratio of 13.
Spec. waste-gas M'/kg 1.60 1.62 1.62 1.62
volume *)
The comparison of operating costs has been based on an
Pressure loss in mbar 72.5 63.0 61.0 60.0 annual production of 480,000 t and includes the expenses of
preheater **) electrical energy, fuel and refractory material for the different
Gas velocity at m/s 6.0 7.0 14.0 11.0 types of rotary kiln with preheater, kiln head and bumer. The
kiln inlet **) approx. cooler has not been considered. The results have been
Spec. sintering 10 8.87 10.76 15.49 19.67 enumerated in table 3. The expenditure of secondary
zone load kJ/M2 h operating media, wear parts and operating personnel has
Spec. kiln t/M3 d 5.33 4,00 3.27 2.35 been considered identical for all four alternatives.
spaceload -)
Table 4
*) at standard conditions, **) at max. production Costs of clinker production based on PYRORAPID" kiln

Version Unit A B CD
The cost comparison includes all machines, with foundation
Spec. manufac- % 100 102.7 102.6 103.8
and building expenses, starting with the preheater blower
up to the kiln head inclusive. Neither the clinker cooler nor tudng costs
the electrical installations have been incorporated in the Spec. DM/t - 0.74 0.71 2.04
investigation. Table 2 shows the comparison of capital extra costs clinker
expenditure for versions A to D. The price advantage of the Absolute extra DM/ - 355.00 340.0001500.00
PYRORAPID"' kiln with 5-stage PYROCLOW1 preheater 0 0
(process A) is obvious equalling approx. 5% against the costs (1980) year
PYROCLONO 'R' kiln (process B), approx. 2 % against the
PYROCLOW 'S' kiln (process C) and approx. 9% against the The PYRORAPIDO kiln turns out particularly favourable with
standard preheater kiln system (process D). respect to the wear of refractory lining. This is opposed to a
- The comparison is based on the price level prevailing in higher consumption of electrical energy because of the
Germany in 1980 and on a one-4eg preheater. If a two-leg pressure loss in the 5th preheater stage. The improved heat
preheater were chosen, which will be the case for plants of consumption is due to lower wall radiation and waste-gas
higher production capacity, the capital expenditure for the losses.
quick burning kiln would become slightly less favourable.
The cost comparison established for clinker production
(table 4) includes the operating costs and capital charges
At identical rotary kiln volume, but varying UD ratio (see (depreciation and payment of interest for the capital input)
chapter 2) the PYRORAPID'5 kiln - assessed by itself - will which stand for the clinker production costs. A straightline
at any rate offer advantages as to capital expenditure.

5
depreciation over 8 years and an interest rate of 12% have
been entered in the calculation. Related to the German cost
level of 1980 and subject to using brown coal as primary
energy carrier, the production costs of methods B to D are
by approx. 0.7 to 1.0 DM/t of clinker higher than of the quick - ?I
P&A
burning kiln with five-stage preheater. This results in annual
savings of OM 340,000.-- up to DM 500,000.-- for the
example chosen.

Table 3
Comparison of operating costs based on PYRORAPID'
Ki’

Version A B C D
Electrical energy 100 94 91 91
Fuel 100 103 103 103
Refract. material 100 108 120 132
Other operating media 100 100 100 100
Mechanical wear parts 100 100 100 100
Operating staff 100 100 100 100
Total of operating cost 100 102.5 102.8 103.2

4. Concluding remarks
Chapter 2. explained why the short kiln is a step towards
quick burning and may, therefore, yield a further improved Fig. 4: First short kiln plant with kiln dimensions of m dia. x 40 m
length. The illustration shows the rotary kiln,the parallel mounted
clinker burning. This statement naturally relates to comparable tertiary air line (with dust separator) and the tube cooler arranged
other plant configurations at otherwise identical conditions underneath and providing for opposite material flow.
as to raw materials, fuel and cooling. There is no doubt that
the negative effect of excessive raw-meal heterogeneity can
eliminate the positive effect of rapid calcining and heating up.
This is also true of the influences originating from the cooling
conditions. It has been mentioned earlier that the
PYRORAPID' kiln will not always be the most appropriate
equipment. However, it frequently offers advantages that
should be made use of. The essential advantage of a short
rotary kiln with respect to process engineering is the reduced
stress on the sintering zone at increased specific kiln space
load. This will have a particularly favourable influence on
service life, kiln operation and formation of accretions.
These features have also come to the fore in the first short
kiln plant which started operation in a German cement plant
in March 198116)17). This facility (40 rn long kiln of 4.0 rn
diameter) (illustrations 4 to 6) has proved to be capable of
burning the material decarbonated in the PYROCILON" up to
99 % to Portland cement clinker of top quality. The clinker can
be easily ground and yields any desired cement quality. At a
specific kiln space load of approx. 5.5 Um'3 d, utilization of the
( I
kiln volume is improved by about 50% compared with kilns of
standard length and precalcining system. Further investiga
tions are yet to be carried out for this plant and the results
obtained are to be supported by long-term observations.

Fig. 5: Partial view of the short kiln plant with the common kiln head.

6
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 15

Instructions for Kiln Operations

 INSTRUCTION FOR KILN OPERATION

The object of this instruction is to give a summary of the most important factors to be
observed in operation of the rotary cement kiln.

This is not a recipe, which, if followed exactly, would give exact results; but a summary of
experiences and observations which should help the operator to draw the -right conclusion from
his/her own observations and to react properly.

All rules for proper kiln operation can be described by one word: CONSISTENCY. It is
agreed that it is not easy to achieve this, but we believe that by analyzing the various factors
which influence kiln operation, we can make recommendations which will be helpful.

First, we realize that there are two groups of variables which influence kiln operation.
These variables are:

A. Variables over which the kiln operator has no influence

1. Chemical and physical properties of the raw materials.

2. Quality of the fuel used, i.e., calorific value, ash content, volatile matter, and
moisture content.

3. The accuracy of the feeders and the drives for machinery.

B. Variables which the kiln operator can control

1. Raw material feed to the kiln.

2. Firing rate of the kiln.
3. Kiln speed.
4. Gas temperature gradient along the kiln.
5. Draft through the kiln.
6. Progress of the material through the kiln.
7. Combustion air.
8. Temperature of combustion air.
9. Dust feed to the kilns.

It is our intention in these instructions to discuss the variables in the latter group and show
what the kiln operator can do to reach optimum results in terms of clinker production, quality, and
economy of operation.
1. RAW MATERIAL FEED

The feed rate determined by the fresh feed flowmeter should be kept absolutely constant,
i.e., the production rate for the kiln under normal operating conditions should be constant and
uniform. The fresh feed rate for the raw materials should be coupled with the motor driving the
kiln, so that the same proportion of materials enters the kiln at any speed. On long dry kilns, the
general rule should be feed/speed should control the burning zone, and feed/draft should control
CGT and backend.

A. Changes Resulting From Material Characteristics

If a change occurs in the raw material feed rate (including dust) it will always affect the
load in the kiln and its temperature. Such a change may be due to variations of chemical or
physical properties of the feed. Whatever the cause, it is important to initiate corrective action as
soon as the deviation is discovered. If this is not done, "cycling" will undoubtedly occur
producing underburned material alternating with overburned.

This condition can develop as follows: if a bigger than usual quantity of material reaches
the burning zone, the fuel feed must be increased to obtain good clinker. Often, this results in the
succeeding material being better prepared than usual. If this quantity of material is followed by a
smaller than usual amount (i.e., a zone with little material which owing to the preceding increase
in the fuel feed, is better prepared as well), this material will generally be overburned. In order to
restore the rapidly rising temperature of the burning zone to the normal value, the fuel feed must
be severely throttled and as a result, the next batch of material will be underburned; thereby,
providing the foundation for the next disturbance. The result is that the output alternates between
under and overburned clinker, leading to serious wear of the kiln lining.

To prevent such cycling, it may be preferable to allow this underburned material to run
through for a time; but only when it is evident from the temperatures that the next batch of
materials is sufficiently prepared and the present variation in the material feed is only of a
temporary nature.

Marked changes in the chemical and physical properties of the material can also lead to
cycling because the flow characteristics of the material change. This becomes apparent if one
observes the progress of material through the kiln. It may also happen that due to the changes in
the composition of the raw material, the deposit formed in the kiln works loose, resulting in
accumulation of material.

B. Changes Initiated By the Operator

Any change in the raw material feed rate is followed by a corresponding change in the
output, with all its consequences (fuel consumption, draft, load of the kiln, temperature
distribution, amount of dust produced, etc.) at constant kiln speed.
 When the feed of raw material is increased, first the temperature of the exhaust gas drops,
followed by a progressive drop in the temperature as the material passes through the kiln, if the
firing rate and draft are not corrected accordingly. Due to the increased output, the temperature
characteristics of the kiln (flue-gas, chain gas, etc.) must be raised slightly in order to re-establish
the equilibrium.

When the material feed is decreased the gas temperature rises, beginning at the feed end; if
the firing rate and draft are not corrected properly. With reduced output, it is generally
permissible to run the kiln at lower gas exit temperature and less dust will be generated.
Disturbances associated with changes in the temperature of material and/or gas in the kiln always
occur when the feed rate is changed rapidly.

Correction to the material feed rate, with the object of regulating the kiln in normal
service; should therefore be avoided under all circumstances, as it will eventually lead to irregular
operation.

If the material feed is to be changed, in order to change the output of the kiln, it must take
place slowly and gradually, so the kiln is always in equilibrium. The instant at which such
changes are made must always be chosen in such a manner that the kiln can withstand the change.
If, for instance, the output is to be increased, this step should be taken when the kiln is hot, so that
little change has to be made to the fuel feed and the draft.

RULE 1: The feed of material to the kiln should not fluctuate.

2. FIRING RATE

If you have been successful in maintaining a steady feed rate a outlined in the previous
section, the next objective for good kiln operation is to maintain an absolutely constant firing rate.
Again, there are inherent characteristics in any fuel which changes (moisture, grindability, heat
value) and for which we have to compensate. It is important to detect these variations as soon as
possible so that the necessary corrective steps are kept as small as possible (not more than 5% of
normal value) in order to minimize the changes in the kiln. If coal is used, it should be as dry as
possible (maximum I % water) and sufficiently finely pulverized to a minimum of 85 % passing
200 mesh.

Changing The Firing Rate

Increasing the firing rate with an adequate supply of combustion air produces a hot burning
zone and a gradually increasing temperature in the kiln. If, at the same time, the draft is
intensified, i.e., more secondary air drawn through the kiln, the temperature may drop in the
burning zone because the heat is drawn too strongly to the back end. Increasing the quantity of
secondary air lengthens the flame. Combustion becomes less intensive and the flame temperature
drops.
 Under these circumstances, the temperature of the exhaust gases rises abruptly. If there is a
shortage of combustion air, the temperature in the kiln drops due to incomplete combustion in
spite of the larger amount of fuel introduced to the kiln. In that case a high proportion of CO will
be detected in the flue gas and the kiln dust becomes dark.

CO detected in the flue gas means poor combustion and therefore also poor economy;
which must be avoided at all times.

Reducing the firing rate causes the burning zone to cool down gradually followed by
cooling of the whole kiln.

There is one condition where a decrease in firing rate can actually cause an increase in
burning zone temperature. This occurs if there was insufficient combustion air before the change
was made. In this case, reducing the firing rate causes better combustion of the fuel and CO will
disappear from the exhaust gas.

Rule 2: Keep corrections to the firing rate as small as possible, but react quickly.

3. KILN SPEED

When the speed of the kiln is changed, the flow rate of the material and consequently the
time during which the material remains in the kiln is changed. At the same time, it causes a
change in the load of the kiln. This, of course, affects the heat transfer to the material and the
temperature gradient of the kiln.

Increasing the kiln speed while maintaining constant raw material feed first causes a
temperature reduction in the kiln because the material -advances faster and enters the burning
zone less well prepared. There is a risk of the material rushing, and of underburned clinker being
produced if the operator does not react quickly enough and temporarily increases the firing rate.

Due to the increased firing rate required in the burning zone, the exhaust temperature also
increases because there is less material per unit length of the feed section than before. The heat
transfer becomes less favorable, the drying point of the material moves upwards, and the rushing
of material occurs (possibly the coating may work loose). After one to two hours (depending on
the transit time) a lighter load suddenly enters the calcining zone, the firing rate has to be reduced,
and cycling may result.

If the material feed is changed at the same time and in the same proportion as the kiln
speed, the load in the kiln remains relatively constant. An observant kiln operator can then keep
the kiln in equilibrium if he watches the instruments carefully (gas temperatures) and corrects the
firing rate (and perhaps the draft) quickly enough.

Reduction of the speed first results in a temperature rise in the burning zone because the
material already in the kiln is exposed to the flame longer and is better prepared. The exhaust
temperature may even drop to begin with because the flame becomes shorter and because the load
at the inlet end increases. To prevent the burning zone from overheating, the fuel feed must be
reduced. As a result, the succeeding feed of material will be less well prepared and, after a short
time (when the material has passed -through the kiln), it will become more difficult to obtain
good clinker. For this reason, variation of the kiln speed in normal operation should be avoided if
possible.

Rule 3: In normal operation, the speed of the kiln should be kept constant and not
continually changed to control the kiln.

4. GAS TEMPERATURE GRADIENT

Fluctuation of the gas temperature at the chains or at the gas outlet, due to overcorrective
actions in the fuel feed and/or the draft, results in the drying point being displaced along the kiln.
The material tends to move faster after the point where final moisture is driven off.

If the gas temperature rises rapidly, the critical transition point shifts towards the feed end.
Thus, more material is temporarily converted into the dry state per unit of time than under normal
conditions, and after about one to two hours (depending upon transit time) material accumulates
in the burning zone.

The same phenomenon, but much sooner, can occur if the temperature of the calcining
zone rises rapidly. As a result, the point at which the material begins calcination is displaced
upwards. This can cause a large amount of material to flow towards the burning zone for a short
time and produce bad clinker, apparently as rushes, which can be most unpleasant. This
phenomenon is often observed after the kiln was briefly stopped, e.g., for inspection. The kiln
lining exposed to the flame radiation becomes hotter when the kiln is at rest. On again starting up
the kiln, the material comes into contact with very hot walls and the Co2 escapes violently,
thereby fluidizing the material bed (as in a pneumatic conveyor). This then causes the material to
rush through the burning zone.

If conversely, the gas temperature drops rapidly, the critical points (transition from moist
to dry state, and the beginning of the calcining zone) move down the kiln. For a short time, this
gives rise to gaps in the material bed at these points. After a while the material level drops in the
burning zone. The product is then overburned, or, if the fuel feed is reduced, the succeeding
material is inadequately prepared, resulting in underburned clinker some time after.

Hence, every such change in the temperature is the source of the next disturbance. With
overburned material the fuel feed is reduced, resulting somewhat later in the underburned clinker,
which is counteracted by forcing (increasing the fuel feed and temperature); this is followed by a
rush of underburned material and the entire sequence is repeated until the temperature balance can
be restored.

It is therefore important to ensure that the kiln is run with the temperature as constant as
possible. At all costs, violent changes in temperature gradient must be avoided.
 If accumulation of material takes place in the kiln, counteraction should be performed as
soon as possible, preferably without the reduction of kiln speed, for the reasons given in Section
3. Special attention must be paid to the temperature at the end of the kiln and below the chains, in
order that these temperatures may change as little as possible. If the disturbance is very sever, it is
preferable to change the kiln speed for a short time and heat the material.

As a rule, it is better not to reduce the firing rate much when there is temporarily a low
material level in the burning zone, but to remain slightly overheated, thereby adequately
preparing the succeeding material without producing undue temperature fluctuation in the kiln. In
most cases the equilibrium in the kiln can be restored in this way.

If temperature changes are necessary in the kiln (i.e., adapt to changed production rate,
dust feed, chemical composition of the raw material, etc.), the change should always be very slow
and spread over several hours. Displacement of the critical points in the kiln then become
apparent slowly, and it is not unduly difficult to maintain the equilibrium in the kiln.

RULE 4: The kiln gas temperatures including NOx, chain gas temperatures, and back
end temperatures should be kept as nearly constant as possible.

5. DRAFT

Increasing the draft beyond the quantity necessary for proper combustion produces a larger
amount of flue gas and always results - at least temporarily in a temperature increase of the
exhaust gas and a decrease in that of the burning zone. This is due to the fact that the flame is
extended, while its temperature decreases. If, on the other hand, combustion was incomplete (due
to shortage of air), increasing the draft will result in a temperature increase of the burning zone
since the heat formerly lost in the CO is now utilized in the kiln.

Throttling the draft shortens the flame, resulting in a drop in the temperature of the flue gas
and an increase in that of the burning zone (provided there is no shortage of air and no CO is
produced).

RULE 5: Corrections to the draft should be small and should only be made to keep the
gas
temperature constant and when combustion necessitates (too much CO or 02 in
the flue gas). Adjust the draft until there is no CO in the flue gas in normal
operation.

6. PROGRESS OF THE MATERIAL THROUGH THE KELN

The changes in material flow are caused by changes in kiln speed, material feed,
temperature of gases, and draft through the kiln.

These causes and effects were discussed already under points 1 to 5.

 Disturbances may occur in the operation of the cooler if the secondary air flowing through
the cooler changes in quantity or temperature. The cause of this may be:

A. Changed granulation of the clinker (i.e., very fine or underburned clinker plugs the
cooler, whereas very coarse clinker reduces the resistance of the grate to the air
flowing through it. In both cases, a change in secondary air temperature will occur).

B. Plugging or gaps in the grate due to defective plates.

C. Changing the speed of the grate - resulting in a change in the thickness of the
material bed. This affects the amount and temperature of the air through the cooler.

Every effort should be made to run with the quantity and temperature of the air as nearly
constant as possible, so that the flame in the kiln remains constant. Indeed, if the amount of air to
the cooler is increased, the excess air in the kiln will also increase; the temperature of the burning
zone drops; that of the exhaust gases rises; and the temperature distribution in the kiln is upset
with all the consequences which gave already been described.

If the amount of air to the cooler decreases, the opposite is true, also upsetting the
temperature.

If the air temperature rises, but the quantity remains constant, then the flame will become
hotter in the burning zone; the heat transfer to the material will become more intense and the
temperature of the exhaust gas will drop resulting in an overall disturbance to the temperature
distribution in the kiln.

Therefore, it is important to determine the amount of secondary air and to keep it as nearly
constant as possible by keeping the cooler in fully automatic control and following airflow set
points determined by the plant technical department.

The only variation to be made by the kiln burner will usually be the undergrate pressure set
point determined by the clinker size and density.

RULE 6: The secondary air flowing into the kiln should be kept at constant
temperature and flow.
7. COMBUSTION AIR

There are three sources of combustion air:

• Primary air, introduced into the kiln with the fuel; in this case, it serves as a carrier of dust;

• Secondary air, which is preheated in the clinker cooler and pulled into the kiln by the
induced draft fan; and
• Leakage, which should be as small as possible.

Under normal operating conditions, the rate of flow of primary air should not be changed.
However, if the production rate is appreciably altered, the amount of air should be adapted
accordingly.

Increasing the quantity, speed, or pressure of the primary air changes the flame shape in
the kiln.

EXCESS AIR

The diameter of the burner pipe should be sized to suit the amount, speed and pressure of
the primary air in order that the desired short, hot flame may be obtained with complete
combustion. The smaller the proportion of primary air, the higher its speed should be so that
proper and sufficient pressure and turbulence are created to achieve good combustion.

All leakage of air into the kiln should be prevented wherever possible (by closing kiln
doors, openings around burner pipe, doors, valves, even in the smoke chamber and dust
collector).

8. TEMPERATURE OF COMBUSTION AIR

It is of utmost importance to keep the temperature of the combustion air as constant and
high a possible. This will result in better fuel economy due to the heat recuperated from the
clinker; and secondly the hotter the secondary air, the faster the combustion of fuel occurs, thus
giving better flame control and higher kiln production.

Since the temperature and quantity of secondary air is influenced by the clinker cooler, its
operation is an integral part of the kiln operation.
 Blue Circle Cement

PROCESS
ENGINEERING
TRAINING PROGRAM

MODULE 2

Section 16

Take these Chains

International Cement Review
 Because many wet-process kilns remain in operation worldwide, the
often-neglected internal kiln chain system still plays a vitally important role
in both the recuperation of heat, evaporation, raw material transport and
the prevention of mud rings. ICR takes guidance from specialist who design
and fit cement kiln chain systems.
A good chain system must handle three basic functions: (a) the effective
collection of the dust carried by the hot gases into the chain system, (b) effective
transport of the sticky slurry in the so called "Plastic" zone, and (c) preheating of the raw
materials. According to one of the leading experts on chain system design and operations,
Anthony de Buss: "What can be said right away is that item (b) i.e. the transportation of
the sticky slurry is the most important one. Without an effective transport system that
moves a maximum tonnage of sticky slurry without causing spillbacks, mud rings or
build-ups - you have nothing- except for a very inefficient kiln, operating at best at
around 75-80 per cent of its maximum capacity.

"Based on my own kiln-chain work over the past 30 years, I believe I have come
up with the ultimate slurry transport concept and indeed, I have applied this approach in a
number of differing kilns with great success. In essence, I have been able to solve the
transport problem associated with sticky slurry.

"Some clients tell me that their slurry is the worst in the world - despite having
looked at various solutions, sometimes for over 20 years or more - and they have not
found a satisfactory solution. The kiln has too high a fuel consumption the kiln capacity
is restricted or the kiln suffers from a combination of both effects.

"When you first talk to them you hear that they have tried oval chains, ring
chains, heavy chains, light chains, high densities, low densities, garland designs, curtain
chain designs, concentric spiral systems ... and on it goes ... spiral systems with three
spirals, six spirals, twelve spirals and all manner of other combinations of design and
chain size.

"It is only logical that the answer has come from people working on kiln chain
systems and there are not too many of these specialists left. True, some clients have their
chain systems designed by very large equipment companies and while such companies
may have excellent skills in the engineering and building of cement plants, they are not
necessarily fully attuned in the design of chain systems.
 "Let's face it, many of the real chain experts that such major engineering groups
may have employed are either now retired or dead and while some of the well-known
names in cement projecting may have experience in building kiln lines etc, they
invariably have no in-house expertise from those who 'eat, breathe and live' kiln chains.

"Recently, I contracted to replace a brand new chain system that had been
designed and fitted by one of these large engineering groups. From the records the chain
system had only been fitted three years previously and the plant technicians told me that
the kiln had never performed well, operating at about 15 per cent under the contract
guarantees.

In my opinion, the problems I have encountered at this particular plant are not too
difficult to solve. In fact some of the incorrect aspects of design and operation center
around a few recognizable trends ranging from the 'plastic' zone is too short, the
suspension system is badly designed, the spirals are wrong, the chains are not right for
the job to more specific items such as the free height under the chains is too low or too
high. The failure to optimize on the details of a correctly designed chain system can often
make a big difference in operation.

"The danger for the client in selecting an expert to provide a chain system is that
not only does he need to be an expert in the design but also he has to have knowledge of
installation supervision and actual start-up procedures as well as being able to guarantee
performance and chain lifetime utilization. One can sympathize with those clients who
opt for the security of the larger company but a good chain designer should be able to
offer such practical experience and performance guarantees."
 Mechanical dewatering
Although not mechanical dewatering devices in the true sense, since they perform
a substantial amount of mechanical work, in addition to the drying of the slurry, chains
are often classed as mechanical dewatering devices.

Walter Duda's excellent Cement Data Book highlights two different types of chain
systems:

• The chain curtain (fig 1)

• Garland chains (fig 2)

Chain curtain arrangement

 In chain curtains, Duda notes that the chains are attached only on one end. The
length of chains should be approximately 0.7 of the kiln's internal diameter (D), while the
distance of the attachments of the particular chains along the circumference is normally
about 30cm, although in less dense chain curtains, a distance of 4045cm can be used. The
chain curtain is normally arranged vertically but as Duda notes, in the former Soviet
Union, the chain curtain is often attached along a helical line.

The lowest point of the garland chain loop should be 0.4D above the kiln's lining
(fig 2). Garland chains are attached at both ends, along a helical line, which with the
kiln's longitudinal axis forms an angle of 45-60°. In kilns with diameters up to 4m, this
angle should be 45-60° while a 60° angle should be applied in kilns with a diameter
above 4m.

It is recognized that most wet cement kilns outside the former Eastern Block only
employ curtain chains. Theoretically, garland chain systems are slightly more efficient
but only if such systems are properly maintained. If they are not they can fall into
disrepair and become less efficient than a curtain chain system.

The total chain weight should be about 15 per cent of the kiln's daily capacity. As an
example, if the kiln capacity is 1800tpd, the total weight of chain should be about 270t.
Chain densities can normally vary between 6-10m2 of chain per cubic meter of internal
kiln volume.

Cast chains or not?

A cast chain is said to offer better flexibility in its manufacture in that the carbon
content of the steel can be modified to the specific use of the chains, with other
components heat resistant alloys, such as chrome and nickel – being chosen for the right
application. Cast links will also cool to their definite shape and are not usually subject to
any deformation. In contrast, the manufacture of welded links can result in deformation
during the shaping process. This will create surface micro-cracks that invariably lead to
corrosion. Forged stainless steel chains will have every link welded, which may lead to
regions of weakness on every chain when operating under high temperature conditions.
Cast chains usually cost slightly more than forged chains, however, most kiln operators
seem to prefer cast chains in the "hot zones".
 In 1995 Magotteaux developed a new chain design based on the simple premise
that 'the more surface you obtain for the same chain weight, the more efficient the heat
exchange system becomes. Using this maxim, Magotteaux's new chain was developed to
include an increased heat exchange surface of 25 per cent for the same weight of chain.

Magotteaux also developed and tested various new alloys and introduced chains
with up to 20 per cent chrome and 10 per cent nickel as a means of improving chain
quality. Magotteaux also claimed to be able to link specific kiln conditions to its chains,
enabling the introduction of tungsten and rare earth elements plus a balancing of the
chrome/nickel contents in order to provide a chain that is resistant to reducing conditions
at high temperatures and attack from sulfur, chlorides and other chemicals.

With this modern chain design, allowing for 25 per cent more surface area per
unit weight and with custom made alloy selections backed by extensive practical tests,
Magotteaux claims to have developed an effective solution for reducing fuel consumption
and increasing kiln chain wear life.

But it's not just the chains

Apparently, Magotteaux has developed a very interesting lining to be used
concerning an effective chain system. It is said to be a good alternative to a castable
refractory lining. A castable refractory lining is also very good, but it is obvious that such
an approach requires great care in selecting the proper mixture of aggregates and even
greater care in installation. As far as ICR is aware, the new Magotteaux lining has been
installed in a number of plants owned by the major cement groups.

Furthermore - says de Beus - the chain system is in itself not complete and to be
totally effective one should install additional heat exchangers. For example, refractory
tumblers and dams have been successfully installed. "I personally like the combination of
dams and tumblers because in operation the raw materials cascade from one refractory
lined compartment into another, surrounded by heat sources from the lining, the tumblers,
the dams and the hot gases.
 "Some people would argue that the tumblers will cause more dust. This is true to
some extent and for kiln with a normal dust loss of 8-11 per cent of feed rate (dry basis)
any extra dust created by tumblers is insignificant and, in any case, will be caught by an
effective dust filtration system at the beginning of the chain system. In my view - says cle
Beus - The tumblers with dams will improve kiln efficiency by at least three per cent.

With an effective dam and tumbler system, the efficiency of the heat exchange
can be almost doubled. In effect, the tumbler turns over the load and exposes the "cold"
core to the heat of the gases. The dams stop raw materials and dust flushing down the kiln
and control the flow of material. Both dams and tumblers are made reinforced inside with
a steel fabric in order to help withstand the different temperature profiles and to
strengthen the raised profile refractory material.

"Dust loss from the kiln is primarily a function of gas velocity inside the chain
system and at the exit of the kiln. To have low dust loss, gas velocities must be kept as
low as possible and this can be achieved in the following three ways:

• Low fuel consumption and hence lower gas volumes

• Lower exit gas temperatures and hence a second reduction in gas volume
• No kiln lining in the low temperature zones of the chain system, thereby
increasing the effective kiln diameter.

A properly designed system should offer the following main flow stages: a
properly designed /calculated free, empty space at the feed end of the kiln followed by an
effective dust filtration system, which then is followed by an effective mud transportation
system, followed by a large surface area for effective operation of the raw materials
preheating chains, followed by a combination of refractory dams and tumblers.
Furthermore, to ensure efficient operation of the system, the temperatures of the raw
materials and the gases at the hot end of the chain system should be able to be measured
and properly recorded.

When properly designed, heat exchangers could be in the first 35-40 per cent of
the kiln length and thus for a kiln of 185m in length the heat exchange/chain system
could in the region of 65-70m in length.

A good heat exchange system could narrow the difference between a wet process
kiln and a 5-stage preheater dry process kiln to 500Kcal/kg clinker and the difference
between a wet process kiln and a semi-wet process kiln to 250Kcal/kg clinker.
 "Considering today's cost of fuel, the price of cement and the high cost of
converting to a dry process, many wet process kilns can easily survive if equipped with a
good kiln chain system. Further substantial cost savings can be gained if waste fuels can
be used for which the wet process kiln is very suitable.

Don't forget the cooler

The clinker cooler must invariably be adapted to handle the higher kiln capacity,
resulting from an improved chain system says de Beus. This can mean multiple cooling
fans, improved under-grate sealing and larger cooler exhaust fans. If the existing cooler is
too large, which can often be the case in some of the cement kilns operating in the CIS
and former East European countries, the first one-third length of the cooler must be
narrowed, to get better clinker distribution over the width of the cooler. Other options
designed to improve cooler performance could also include fixed cooling air beams.
However, each cooler must be individually studied. There is no mystique and one
certainly should not go overboard on expense. In fact, most plants can make the
necessary cooler modifications themselves.

Having a better heat exchange system in the kiln provides for a steadier kiln
operation and hence fewer problems with the cooler. Poor cooler operation might cost
several hundred thousand dollars per year in downtime and cooler repairs.

Furthermore, the kiln must have a good discharge end seal. The combustion air
must come from the preheated primary air and from the cooler as secondary air - not from
any other source.

It is claimed that a good chain system fitted to a wet process kiln could save as
much as US$500,000 per year compared to a kiln fitted with a badly performing system.
As such, the payback for a good chain refit, employing the latest materials and someone
fully knowledgeable as to the intricacies of fitting a complete chain/heat exchange system
could therefore be within 18 months to two years.
 As Mr. de Beus concludes: It is possible to gain some notable successes in this
specialist field of chain design. Increased wet kiln capacities by 20 per cent are possible
as is fuel consumption lowered by 15 per cent. At the same time an effective chain
system should offer reduced dust losses by as much as 6-10 per cent of kiln feed, with all
dust returned to the kiln. Other gains include the elimination of spillbacks and mudrings.
Elimination of kiln cycling and - not least - increased life of the burning zone bricks
because of a much stabled kiln operation.

"A rotary kiln must turn steadily and continuously to optimize on performance. Speed
changes, slow-down, stoppages are all bad news for both chains and the refractory
system. In my humble opinion, the possibly savings accrued by expecting the plant or
engineering office to design and fit a chain system will invariably be wiped out in a short
space of time by a poorly operating kiln. Chain systems are a job for experts!"

Wickede
Below: A typical oval link curtain system as supplied by
Kettenwerk (Heko)
Ruhr. Germany. Heko is a specialist in
supplying welded chains which it claims haie
have many advantages
over cast chains, provided they are welded In fact, notes
Heko, mechanical properties of cast chain demonstrate a
substantially lower tensile strength and much lower elongation
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

HBM PROCESS ENGINEERS

CONFERENCE

• Hearth Design for Preheater and

Precalciner Kilns
• Preheater Cleaning Methods and
Successes
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

HBM PROCESS ENGINEERS

CONFERENCE

• Hearth Design for Preheater

and Precalciner Kilns
HEARTH DESIGN FOR

PREHEATER AND

PRECALCINER KILNS
 SYSTEM G-AS VELOCITIES
i
(a3, TEMPERATURE

KILN BBDY 6 - 10 m/see @ lk50°C

HEARTH 20-30 m/xc @ 115OOC

RISER 15-25 m/ax @ 850°C

CYCILBNE INLET 15-25 m/set @ 85OOC

 EFFECTS OF HEARTH VELOCITY

HEARTH PREHEATER PREHEATER

VELOCITY PRESS. DROP EXIT TEMP.

20 m/s 480 mm w.g. 380°C

35 m/s 871 mm w.g. 4oo”c

 CHARACTERISTICS OF MEAL CYCLING

e UNSTABLE SUCTIONS,
MORE PROMINENT IN LOWER STAGES

e VISUAL OBSERVATION IN AREA JUST ABOVE

HEARTH SHOWS BRICHTi’DARK CYCLING
 CRITICAIi ZONE FOR
HEARTH VELOCITY

HEARTH IN GOOD CONDITION

30 - 35 m/s

HEARTH IN POOR CONDITION

20 - 30 m/s
 HEARTH PROBLEMS

8) PROTRUDING OR MISSING BRICKS

Q) HEARTH BRICKS “HEAVING”
a DA1MAGED CONCRETE
B) BUILD UP IN MEAL PATH
B, HEARTH REFRACTORY TOO THICK
0 ARCH REFRACTORY TOO THICK
@ STEP BETWEEN MEAL PIPE AND
HEARTH SURFACE
 CONCLUSIONS

e LOOK AT THEORETICAL HEARTH VELOCITY

a MEASURE HEARTH AREA AND COMPARE WITH
“NQRMlAL” AREA
o IF HEARTH VELOCITY IS ~30 m/s AND SYMPTOMS
OF RECIRCULATION EXIST, CORRECT BY
CLEANING OR REPAIRING REFRACTORY
B) IF HEARTH VELOCITY IS >30 m/s THEN CONSIDER
MODIFYING TO TWIN SLOPE OR GOLDEN BAY
HEARTH
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

HBM PROCESS ENGINEERS

CONFERENCE

• Preheater Cleaning Methods

and Successes
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

PRESENTATIONS

Suspension Preheaters

Precalciner Kilns

Kiln Processes

Bowmanville Cement Technology-

Kiln Systems
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

PRESENTATION

Suspension Preheaters
SUSPENSION PREHEATER
 What are we trying to do

Rotary cement kiln

 Transfer Heat

• Conduction

• Convection

• Radiation
 Radiation

Temp 1 Temp 2 driving force

2 1 16 - 1 = 15

3 1 81 - 1 = 80

3 2 81 - 16 = 65
 oF oC
Kiln Temperature Zones
3500
-1800
3000

Gas Temperature -1500

2500

-1200
2000

Feed Temperature
-900
1500

-600
1000
-300
500

0 0
High
Preheating Zone Low Temperature Burning Cooling
Calcining Zone Temperature
Zone Zone
Calcining
Zone

Blue Circle Technical Training

Heat transfer in kiln
How can heat transfer be improved
Kiln rotation
8

3
1
Gas

5
7
6
 Suspension Preheater

• Development of Suspension Preheater

• Suspension Preheater description

• Preheater Control Parameters

• Preheater Operation
 Suspension Preheater

• Rotary kiln developed around 1885 by Ransome

• Long wet and dry kilns used but high fuel
consumption
• SP Patent issued 1934 but 1st installed in 1950’s
• Initial problems with operation - volatiles etc.
• By 1960’s, all plant suppliers offering SP Process
• Units installed from 300 to 4500 tpd
• Currently SP for 1000 - 2000 tpd
• SP fuel consumption - 770 - 830 kcal/kg
• (N.B. Wet process 1400 - 1500 kcal/kg)
 Suspension Preheater

• Production of clinker involves:

• Heat transfer e.g.
– FROM HOT GAS TO COLDER MEAL
– CLINKERING REACTION
• Mass Transfer e.g.
– COMBUSTION OF COAL
– DECARBONATION OF RAW MEAL

• Heat transfer can be affected by process

choice
• Mass transfer is affected by choice of fuel
and raw materials
Suspension Preheater
 Suspension Preheater

THIS REPLACED BY THIS

Hot Gas

Raw Meal - surface

400 - 500 m2/kg

Raw Meal in

Raw Meal - exposed surface

around 2.7 m2/kg Hot Gas

 Suspension Preheater
Cleaned gas

Cyclone inlet

Gas and meal

Vortex finder
Directs the cleaned gas
towards the outlet and
prevents short circuit of
gas/meal through cyclone
Meal to next riser

Increase velocity = Increased efficiency Increase velocity = Increased pressure drop

Suspension Preheater
Feeding into a riser
Feeding into a riser
 Suspension Preheater

• Poor Heat exchange

– Generally due to poor dispersion or short
circuit of material
• Dispersion achieved by either
– Splash plate (adjustable but prone to wear)
 Suspension Preheater
• Splash box (outside riser but not
adjustable)
– Short circuiting can be a problem at start-up
when gas velocity is low
– Long term solution is to fit a constriction below
the box to accelerate the gas
Feeding into a riser
Gas flow into cyclone
 Suspension Preheater
• Deficient heat exchange due to vortex:
– If vortex finder is not of sufficient length
then gas can bypass cyclone from inlet to
outlet, carrying heated meal to colder part of
process.
 Suspension Preheater

• Inefficient cyclones lead to hot meal being

carried back to a colder part of the process
• Higher pressure drop = bigger fan, higher
power costs, higher inleak
• COMPROMISE SITUATION
• Lower stage efficiencies - 75 - 80 %
• Top stage - 90 - 95 %
• Higher efficiency to prevent high dust load for
gas cleaning / dust handling / kiln feed
 Suspension Preheater

• Heat transfer depends upon:

• Specific heat of solid
• Gas and solid temperature difference
• Particle size of particle to be heated
• Time the solid is exposed
• Degree of contact between gas and solid
• Kiln tube is still required after
decarbonation as material becomes sticky
and elements require intimate mixing.
• Need to replace drying/heating/calcining
zone.
 Suspension Preheater

• Heat exchange carried out in the riser -

dispersed and mixed in gas flow
• Meal and gas require separation before
further heat exchange at next preheater
stage
• Separation achieved by centrifugal forces
in cyclones
• Cyclone design is a compromise between
separation efficiency and pressure drop.
 Suspension Preheater
330 - 360 °C, 760 mmwg

530 - 560 °C, 540 mmwg

680 - 720 °C, 370 mmwg

800 - 840 °C, 220 mmwg

1010 - 1100 °C, 50 mmwg

Conventional Temperature / Pressure drop (GB)

Suspension Preheater
Low pressure drop cyclones
1. Widened inlet to increase area
and decrease velocity.
2. Sloping shelf on inlet to reduce
build-up of meal in the inlet.
3. Increased wrap-around to
encourage cyclonic action
4. Shorter dip tube required due to 3
5. Changed geometry gives reduced
capital cost
 Suspension Preheater
Meal ex. Stage 3

(d) Kiln Hearth

Gas + (a) Gas flow over bed - dust pick-up α velocity3.5
(b) Hearth area is smaller than kiln or riser
raw meal
- velocity increases - potential for dust pick up.
(c) (c) Recirculation leads to heat penalty
and pressure drop - need to limit velocity
BUT (d) Must have sufficient velocity to lift
Stage 3 meal to Stage 4 or result is fall-through
(b)
(a)
Combustion gas

Meal ex. Stage 4 Design velocity34m/s

 Suspension Preheater
High efficiency
top stage
Preheating

Lower efficiency
stages Calcining - 25 to 30 %
decarbonation

Remaining decarb. plus clinkering

Kiln hearth
 Suspension Preheater

• Preheater kiln control

• Raw meal feed rate / kiln speed
• Fan - related to oxygen level
• Fuel feed rate

• Need to assess the effects of changes

made
• Kiln drive power - either kW or Amperes
• Burning zone temperature - pyrometer
• NOx - back end and / or preheater exit
• Combination of the above?
 Suspension Preheater

• Kiln drive power variations indicate

condition of feed and coating in kiln -
• Wide variations indicate uneven coating / ring formation
(higher power when lifting than falling)
• Change in level of power indicates thinning / thickening of
coating. Sudden change indicates coating / ring breakaway.
Operator prepared for rush of material through kiln
• Direction of kiln amps:
• Hotter kiln has stickier material which rises higher and
forms thicker coating - increasing kiln power required
• BUT Kiln power can be affected by kiln speed and volatile
cycles.
 Suspension Preheater

• Burning zone temperature

• Colour pyrometer used in 70/80’s to indicate
changes in BZT but affected by:
– Dust from the cooler
– Burning zone ‘movement’ can give
misleading results
– Systems maintenance intensive
• More recent cameras show more potential for
control
 Suspension Preheater
• NOx - oxides of nitrogen - principally NO - created by
chemical reaction and temperature in flame.

• Assume that coal / firing system are constant then

change in NOx means change in BZT.

• Back-end mainly used but preheater exit measured as a

back-up.

• NOx level is site specific - no absolute level specified.

• Warning - NOx is destroyed by carbon monoxide (CO)

so decrease in NOx could lead to increase in fuel to heat
BZ - result is more CO and cooling of BZ - flush!
 Suspension Preheater

• Process Considerations

• Oxygen Levels - what is the optimum level

• Preheater inefficiency
– Heat Exchange
– Gas short circuit
• Volatiles - is there a problem?
 Suspension Preheater
• Optimum Oxygen level
– Require excess oxygen for complete combustion
i.e. when CO appears at the back end - 1 to 1.5%
O2
– If CO at back-end is evident then excessive CO in
flame results in longer, cooler flame.
– Less heat given out when carbon burns to CO
rather than CO2
– Reduced gas temperatures so decreased heat
transfer.
– Increase in volatile cycles - build-up in hearth /
riser / cyclones.
 Suspension Preheater
• Evaporation of volatiles takes energy from
burning zone and releases it at the back of the
kiln - increased fuel consumption.

• Decreased area for gas flow due to build-up

• Decreased area gives higher velocity which

gives higher pressure drop - change in
operation of the fan - reduced volume

• Decreased oxygen available giving a reduction

in output.
 Suspension Preheater

• Best solution:
– Don’t get into a cycle in the first place!
• Most preheaters with low inleak run 2 -
2.5 % back end, 3.5 - 4 % at preheater
exit.
 Suspension Preheater

• Volatiles - how do we control them

• Compounds such as sulphates, chlorine,
fluorine and alkalis - sodium and potassium.
• Solution 1:
– Get the chemist to sort it out!
– Balance compounds ex: sulphates and alkalis so
that they become part of the clinker.
• Potassium and Sodium Sulphate
• Excess sulphate will react to give Calcium Sulphate
 Suspension Preheater
• Solution 2:
– Operate process to minimise problems -
HOW?
• Sufficient oxygen to limit the initiation of a
cycle
• Area most suitable for sulphate build-up is
riser / stage 4 due to geometry and
temperature.
– Good dispersion of meal across the riser to
stage 4
– Low entry point (avoiding fall through) to
riser
 Suspension Preheater

• Quench gases as quickly as possible to

minimise build-up on riser - deposit on
meal to combine into the clinker

• Remove volatiles from the process

instead of recirculating around the
system
Suspension Preheater
Summary

770 - 830 kcal/kg

Pressure drop through system relates

to output and production cost

If volatiles are balanced then

minimum build-up - use sufficient oxygen

Reducing conditions affect clinker quality

and refractory life
THE END
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

PRESENTATION

Precalciner Kilns
PRECALCINERS

1
 KILNS WITH
PRECALCINERS

1.- WHAT ARE THEY

2.- HOW DO THEY WORK
3.- WHY INSTALL THEM

2
 Calciner vessel
Back-end Fuel

7-8% 02

AIR THROUGH PRECALCINER

AIR SEPARATE PRECALCINER

Calciner vessel
Back-end Fuel
Tertiary Air duct

2 % - 02

3
 DECARBONATION

CaCO3 + HEAT

CaO + CO2
LIME CARBON
DIOXIDE

4
 DECARBONATION VS TEMPERATURE

100

80

LEVEL OF 60
DECARBONATION
40

20

0
800 820 840 860 880 900

TEMPERATURE IN CYCLONE AFTER CALCINER (ºC)

5

5
 RELATIONSHIP BETWEEN DECARBONATION OF CaC03
TEMPERATURE AND CO2 ATMOSPHERE

5 3 SEC

RATE OF 4
DECARBONATION
(mg/min - cm²) 950ºC
3
7 SEC
2
850ºC 900ºC
1 800ºC 40 SEC

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

PARTIAL PRESSURE OF CO2 (atm)

6
 PREHEATER KILN
RAW MATERIAL TO KILN - 30%
DECARBONATED

PRECALCINER KILN

RAW MATERIAL TO KILN - 85-95%

DECARBONATED
7

7
BY-PASS GAS REMOVAL
PREHEATER KILN
P.H.
PH
VOLATILES
+ALL COMBUSTION GASES
+ 70% OF CO2

HORNO
KILN

AIR SEPARATE PRECALCINER

P.C, VOLATILES HIGHER

PH + 40% OF COMBUSTION GASES CONCENTRATION
+ 15% OF CO2 OF VOLATILES

BYP/CPASS

8
HORNOKILN

8
 FUEL CONSUMPTION VS GAS WITHDRAWN FROM RISER DUCT

950

30% DUST LOSS

925

FUEL 900
CONSUMPTION
(KCAL/KG NET)
30% DUST LOSS
875
10% DUST LOSS
10% DUST LOSS
850

825

800
0 10 20 30 40 50 60 70 80 90 100

% OF KILN GAS TAKEN FROM RISER DUCT

Precalciner kiln Preheater kiln 9

9
ADVANTAGES OF PRECALCINATION
• Higher tonnage from given kiln tube
• More stable operation
– Lower refractory usage
– Better average t/d on annual basis
– Lower variability in clinker quality
• Better with raw materials with high
sulphur/chloride etc
• Better for NOx reduction

10

10
Air Separate Pre-Calciner
(Prepol)

Calciner vessel

Back-end Fuel

Tertiary Air duct

11

11
 ILC-E : In line calciner - air through system

-375 mm WG
Kiln with precalciner - single cyclone stream, 285ºC

small calciner in the riser duct. Air for combustion

in the calciner passes through the kiln
460ºC

Characteristics :
- Recommended Range of capacities : 900 - 3300 tpd 615ºC

- Calciner fuel : 15 - 25% of total

- Possible by pass quantities : 0 - 25% 745ºC

840ºC
1. Calciner
2. By Pass

200ºC 12

12
 RAW MATERIAL FROM CYCLONE 3

CYCLONE 4

Calciner vessel

Back-end Fuel
Tertiary Air duct

AIR SEPARATE CALCINER - INLINE VESSEL

13

13
 FLS PRECALCINER

FEED

FUEL

PREHEATED COMBUSTION AIR 14

FROM CLINKER COOLER
14
15

15
16

16
 DESIGN OF PRECALCINERS
• Residence time for fuel combustion
• Avoid raw feed short circuit
• Good mixing :
– Oxygen
– Fuel
– Raw feed
• Residence time for decarbonation

18

18
Yanbu
KHD twin
string
Precalciner
with chloride
bleed

19

19
 Review - Kiln Processes
Process L/D Speed rph Capacity TPD Fuel
Consumption
(kCal/kg)
Wet 35 - 40/1 50 - 80 500 - 3,000 1,200 - 1,400

Long Dry 35-40/1 60 - 90 300 - 2,000 950 - 1,100

Lepol 16 - 20 / 1 60 - 90 500 - 1,800 750 - 900

Preheater 14 - 17/1 120 - 180 500 - 2,500 750 - 850

AT PC 14 - 17/1 180 - 240 1,000 - 4,500 750 - 850

AS PC 15 - 18/1 180 - 240 2,400 - 10,000 700 - 850

20

20
 CONVERSION N°3
KILN - RAWANG

TRUNK DUCT

CYCLONE 5

OUTLET

FLASH
FURNACE

VORTEX
CHAMBER

21

21
MODIFICATIONS TO DEDUSTING TERTIARY AIR

NEW DEDUSTING CYCLONE

TERTIARY AIR

NEW DEDUSTING CYCLONE

TERTIARY AIR

HORNO

22
Nº 3 KILN IMPROVEMENTS KANTHAN
22
 PROBLEMS OF BLOCKAGES IN KILN RISER AND KILN BACK-END

UPPER STAGE 3
MEAL FEED

EFFECT OF LOWER
ENTRY FOR STAGE
3 MATERIAL

BUILD UP AREAS

23

23
 MODIFICATIONS TO KILN BACK END

MATERIAL FROM CYCLONE 3

MATERIAL FROM CYCLONE 3

DISPERSION BOX
BEFORE
AFTER
MATERIAL

MATERIAL

ACCELERATOR FOR
TERTIARY AIR GASES IN RISER DUCT
FUEL FUEL

SHORT CIRCUIT

24

24
SHORT CIRCUITING OF RAW MEALIN 4TH STAGE CYCLONE

MATERIAL FROM CYCLONE 2

BEFORE 797 C

DISPERSION PLATE AFTER 875 C

BEFORE 918 C

BEFORE AFTER 878 C

CYCLONE 2
SHORT CIRCUIT
TO KILN

CYCLONE 4

CYCLONE 4 BEFORE 850 C

MATERIAL 25
AFTER 858 - 868 C
25
 REDUCTION OF CYCLONE PRESSURE DROP

REFRACTORY LIP
RESTRICTS INLET

20

CYCLONE
PRESSURE
DROP 15

(MBAR) CYCLONE 3
BEFORE
AFTER

10
CYCLONE 2
BEFORE

AFTER

BEFORE
AFTER
100 120 140 160 180
KILN FEED RATE TONNES/H 26

26
 VIEW FROM CALCINER

AJUSTABLE ORIFICE ASSEMBLY 27

27
 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

PRESENTATION

Kiln Systems
 Kiln Systems
An Introduction

Blue Circle Technical Training

 Kiln Process Overview
• Clinker making processes
• Wet kiln
– Process
– Equipment
• Suspension preheater
• Precalciner
• Conclusion

Blue Circle Technical Training

 Clinker making process
• (Drying)
• Preheating
• Decarbonation/Calcining
– CaCO3 ⇒ CaO + CO2
• Sintering
• Cooling

Blue Circle Technical Training

RANSOMES PATENT DRAWING OF THE CEMENT ROTARY KILN

View of the longitudinal Section

Cross Section

Blue Circle Technical Training

Drying/heating

Blue Circle Technical Training

Drying/heating
Decarbonation

Blue Circle Technical Training

Drying/heating
Decarbonation
Heat/melts

Blue Circle Technical Training

Drying/heating
Decarbonation
Heat/melts
Clinker

Blue Circle Technical Training

Drying/heating

Decarbonation
Heat/melts
Clinker
Cooling

Blue Circle Technical Training

 oF oC
Kiln Temperature Zones
3500
-1800
3000

Gas Temperature -1500

2500

-1200
2000

Feed Temperature
-900
1500

-600
1000
-300
500

0 0
High
Preheating Zone Low Temperature Burning Cooling
Calcining Zone Temperature
Zone Zone
Calcining
Zone

BlueBlue Circle Technical Training

Circle Technical Training
 Rotary Kiln

Blue Circle Technical Training

Feed System
Rotary Kiln

Blue Circle Technical Training

 Firing
Feed System System
Rotary Kiln

Blue Circle Technical Training

Feeding System
Firing System
Rotary Kiln

Cooler

Blue Circle Technical Training

 Wet kiln
• Length Diameter ratio: 40:1

• Drying zone 30 to 35 % of length

• Preheating 10 to 15 % of length
• Calcining 30 to 35 % of length
• Sintering 15 to 20 % of length
• Cooling 2 to 3 % of length

Blue Circle Technical Training

Largest BC wet kiln: ~ 6 m by 200 m - 2300 tpd
BCI current experience:
108 to 7,250 tpd - (9,980 tpd exist)
Blue Circle Technical Training
 Suspension preheater kiln
• Typical length to diameter ratio: 17:1
– Golden Bay Cement: 3.8 x 60 15.8:1
• Capacity Golden Bay: ~ 1600 tpd
• Wet kiln 1600 tpd: 4.7 x 188 m

• Decarbonation of material entering the kiln:

– generally 40 % GB: 70 %
– wet kiln: 40 % decarbonation @ 45 % length

Blue Circle Technical Training

 Suspension Preheater
Feeding
Suspension Preheater

Decarbonation: 40 %

Rotary kiln: 17: 1 or ~ 40 % of wet kiln

Firing System

Cooler
Blue Circle Technical Training
 Typical
high efficiency
cyclone

Blue Circle Technical Training

Feeding

Air Separate Pre-Calciner

(Prepol)

Calciner vessel

Back-end Fuel

Tertiary Air duct

Decarbonation: 95 % Rotary kiln: 17:1 L:D

Blue Circle Technical Training

 Summary
• L:D ratio
– Wet kiln: ~ 40:1
– Preheater/Precalciner kiln: ~ 17:1
• Decarbonation rate
• Today’s standard size: 5000 tpd
– Triple string precalciners up to 10,000 tpd
• Preheater kiln < 2000 tpd

Blue Circle Technical Training

 Blue Circle Cement

PROCESS ENGINEERING TRAINING

PROGRAM

PRESENTATION

Bowmanville Cement
Technology- Kiln Systems
Bowmanville Cement Technology -
Kiln Systems
 Kiln Systems
An Introduction

Production level
Limitations
 Clinker making process
● (Drying)
● Preheating
● Decarbonation/Calcining
– CaCO3 ⇒ CaO + CO2
● Sintering
● Cooling
Drying/heating
 Kiln Chemistry
Drying and preheating
Dry process kilns have very little free water
associated with the kiln feed, a small amount of
water is also usually “bound” inside the shale or clay
minerals present. This requires energy to release. In
a pre-heater or pre-calciner kiln, this occurs in the
upper stages of the tower.
Drying/heating
Decarbonation (800C)
 Kiln Chemistry
Decarbonation
Decarbonation or calcination as it is sometimes
termed, is the primary energy intensive reaction of
cement kilns. During this process, Calcium Carbonate
is dissociated into Calcium Oxide (lime) and Carbon
Dioxide according to the reaction:
CaCO3 ⇒ CaO + CO2 ↑
This reaction absorbs around 400 kcal/kg of heat. The
degree to which this reaction occurs within the pre-
heater tower is the principal difference between pre-
heaters and pre-calciners.
Drying/heating
Decarbonation
Heat/melts (1250 C)
Drying/heating
Decarbonation
Heat/melts
Clinker (1450C)
Drying/heating

Decarbonation
Heat/melts
Clinker
Cooling
 Kiln Chemistry
Reactions in the Kiln system
Raw meal/slurry undergoes a series of chemical changes as
it passes from the kiln inlet to the cooler. These changes or
reactions occur in approximately the following order:
Drying and pre-heating.
• De-carbonation.
• The formation of low temperature melts and fluxes.
• C2S formation
• C2S converting to C3S
Dry and heat
Decarbonate- 800 C
Heat and Melts-1250 C Cool
Clinker- 1450 C
 Temperature profiles of gas and material in a rotary kiln

2000
System boundary
Gas commencement of
decarbonation
1500
Material

1000

500

cooling burning decarbonation preheating drying

Main thermal Subsidiary thermal

system system
Cooler Kiln
Precal Suspension Preheater
 Suspension Preheater

Material temp 850C Gas temp: 1150C

 Kiln Feed

Suspension Preheater
and kiln

Gas handling

Firing System
Cooler
Calciner Kiln Plant
 Kiln Process Overview

● Chemistry of the clinker making processes

● Suspension preheater
● Precalciner
● Clinker coolers
● Flames
 Large Plant ?

Kiln 1: 3500t/day

Kiln 2: 4000t/day

Kiln 3: 5500t/day
Kiln 4: 5500t/day
Kiln 5: 10,000t/day
Kiln 6: 10,000t/day

Total: 11.9million t/year