Design With Energy

CAMBRIDGE URBAN AND ARCHITECTURAL STUDIES
DESIGN WITH ENERGY

THE CONSERVATION AND USE OF ENERGY IN BUILDINGS
CAMBRIDGE URBAN AND ARCHITECTURAL STUDIES
General Editors
LESLIE MARTIN
Emeritus Professor of Architecture, University of Cambridge
LIONEL MARCH
Rector and Vice-Provost, Royal College of Art
VOLUMES IN THIS SERIES

1. Urban Space and Structures, edited by Leslie Martin and
Lionel March
2. Energy, Environment and Building, by Philip Steadman
3. Urban Modelling, by Michael Batty
4. The Architecture of Form, edited by Lionel March
5. The Evolution of Design, by Philip Steadman
6. Bruno Taut and the Architecture of Activism, by Iain Boyd
Whyte
7. Incidence and Symmetry in Design and Architecture, by Jenny
A. Baglivo and Jack E. Graver
8. Design with Energy: The Conservation and Use of Energy in
Buildings, by John Littler and Randall Thomas
Design with energy
The conservation and use of energy
in buildings
JOHN LITTLER
and
RANDALL THOMAS
The right of the

University of Cambridge
to print and sell
all manner of books
was granted by
Henry VIII in 1534.
The University has printed
and published continuously
since 1584.
CAMBRIDGE UNIVERSITY PRESS

CAMBRIDGE
LONDON NEW YORK NEW ROCHELLE
MELBOURNE SYDNEY
Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 1RP
32 East 57th Street, New York, NY 10022, USA
296 Beaconsfield Parade, Middle Park, Melbourne 3206, Australia
© Cambridge University Press 1984
First published 1984
Library of Congress catalogue card number: 83—7537
British Library cataloguing in publication data

Littler, John
Design with energy, the conservation and use of
energy in buildings. — (Cambridge urban and
architectural studies)
1. Architecture and energy conservation
I. Title II. Thomas, Randall
696 NA2542.3
ISBN 0 521 24562 1 hard covers
ISBN 0 521 28787 1 paperback
Transferred to digital printing 2003
WD
For David Bullett and Michele Thomas
Contents
Units, symbols, abbreviations, conventions and conversion

factors page xi
Preface xv
1 Energy and buildings 1

1.1 Introduction 1
1.2 Energy and the built environment - past and present 2
1.3 Energy and the built environment - the future 5
2 Site planning and analysis 14

2.1 Introduction 14
2.2 Solar radiation 16
2.3 Wind 24
2.4 Soil 33
3 Building design 42
3.1 Introduction 42
3.2 Energy demand and thermal response 42
3.3 The internal environment 47
3.3.1 Comfort 47
3.3.2 Temperature 48
3.3.3 Room air movement, ventilation and relative 52
humidity
3.4 Size and type 59
3.5 Form and orientation 60
3.6 External and internal layout 67
3.7 Construction - general 70
3.8 Foundations and walls 76
3.9 Floors 83
3.10 Windows 84
3.11 Doors 87
3.12 Ceilings and roofs 88
4 Passive solar design 97

4.1 Introduction 97
4.2 Elements of passive solar systems 100
4.2.1 Glazing 100
4.2.2 Single and multiple glazing 103
vii
viii Contents
4.2.3 Insulating blinds and shutters 109
4.2.4 Shading 116
4.2.5 Radiation enhancement using reflectors 117
4.2.6 Thermal storage 118
4.3 Passive solar heated buildings 123
4.3.1 Direct gain 123
4.3.2 Attached sun spaces 128
4.3.3 Thermal storage walls 137
4.3.4 Roof ponds 144
4.3.5 Roof-space collectors 146
4.3.6 Convective loops 147
4.4 Design methods for passive solar buildings 148
4.4.1 Introduction 148
4.4.2 Characteristics of a perfect design model 150
4.4.3 Models approaching the desired degree of 150
flexibility
5 Active solar heating 159

5.1 Introduction 159
5.1.1 Definitions 159
5.1.2 Popularity of solar heating 159
5.1.3 Outline of the active systems 159
5.1.4 Approximate energy available 160
5.1.5 Hot water compared to hot water + space heating 161
5.2 Solar air heating 161
5.2.1 Collectors 162
5.2.2 Air to water heat exchangers 163
5.2.3 Dampers 163
5.2.4 Auxiliary energy 163
5.2.5 Pebble beds 163
5.2.6 Controls 165
5.2.7 Results of monitored systems 170
5.3 Solar water heating 171
5.3.1 Outline of systems 171
5.3.2 Collectors 173
5.3.3 Pumps 175
5.3.4 Heat exchangers 175
5.3.6 Drain-down valves 176
5.3.7 Transfer fluid 176
5.3.8 Controllers 177
5.4 New types of system 177
5.4.1 Evacuated tubes 177
5.4.2 Plastic collectors 177
5.4.3 Zeolite heating and cooling collectors 178
5.4.4 Refrigerant-charged loop 178
5.5 Installed systems 179
5.6 Multifamily installations 182
Contents IX
5.7 Predictive methods 182
6 Space heating and ventilation 189

6.2 Solid-fuel heating 196
6.3 Water-distribution systems 205
6.4 Forced-air systems 217
6.5 Heating with electricity 225
6.6 Heat pumps 228
6.7 Group schemes 234
6.8 Conclusion 236
7 Thermal storage 241

7.2 Sensible-heat storage in water 243
7.3 Sensible-heat storage in rock 249
7.4 Phase-change energy storage 252
7.5 Developments in storage 255
7.6 Conclusion 258
8 Wind energy 262

8.2 Power extracted by turbines 264
8.2.1 Power in the wind 264
8.2.2 Variation of wind speed with location 264
8.2.3 Variation of wind speed with height 267
8.3 Power extracted by real turbines 268
8.4 Types of turbine 270
8.5 Electrical generation 278
8.5.1 Electrical generators 278
8.5.2 Gearboxes 280
8.5.3 Electricity storage 280
8.6 Use of wind-generated electricity 282
8.6.1 AC or DC? 282
8.6.2 Inverters 283
8.7 Economies of scale 284
8.8 Installed systems 284
8.8.1 Conservation house at NCAT 284
8.8.2 Wind-powered cottage at NCAT 287
9 Water-supply systems 289

9.2 The Cambridge Autarkic House water-supply system 290
10 Waste disposal and utilization 293

10.2 Aerobic systems 294
x Contents
10.3 Anaerobic systems 295
10.4 Methane digestion 296
11 Domestic-energy saving 304

11.2 Domestic hot water 304
11.3 Electrical power 306
11.4 Cooking 307
11.5 Space heating 308
11.6 Transportation 308
12 Housing case studies 310

12.2 New houses — three solar air-heated houses in 310
Peterborough
12.3 New houses - energy-efficient houses in Newnham, 315
Cambridge
12.4 Rehabilitated houses — renovation of a farmhouse to 317
include a sun space (contributed by Peter Clegg)
12.5 Rehabilitated houses - house conversion incorporating 319
a roof-space collector (contributed by Peter Clegg)
12.6 Rehabilitated houses - a nineteenth-century terraced 322
house
13 Non-domestic case studies 326

13.2 The swimming pool, Sheiling Schools 326
13.3 The new BRS office building 328
13.4 Agricultural buildings 329
13.5 School buildings (contributed by Nick Baker) 330
Appendix 1: Weather data 342

A 1.1 Introduction 342
A 1.2 Solar spectrum 342
A 1.3 Global solar radiation data for the UK - Kew and
Bracknell 343
A 1.4 Global solar radiation data on the horizontal for
three UK stations 343
A 1.5 Direct and diffuse solar radiation for Kew, UK 343
A 1.6 Temperature data for the UK 344
A 1.7 Wind data 344
Appendix 2: Thermal performance 347
Appendix 3: Interstitial condensation 358
Index 363
Units, symbols, abbreviations,
conventions and conversion factors
(1) Principal units

(In some cases a base unit is combined with a multiplier because this is its
most common form.)
°C degree Celsius
d day
g gram
h hour
ha hectare
J joule
K degree kelvin
kcal kilocalorie
1 litre
MTCE megatonnes of coal equivalent
m metre
mb millibar
N newton
p person
Pa pascal (N/m 2 )
ppm parts per million
s second
t tonne
V volt
W watt
The principal units are frequently used with the following multiples and
submultiples:
10"1 deci d 10 deca da
10"2 centi c 102 hecto h
10"3 milli m 103 kilo k
6
10" micro JJL 10 6 mega M
10~9 nano n 10 9 giga G
XI
xii Units, symbols and abbreviations
(2) Symbols
(Where two meanings are given, the correct one should be evident from the
context of the chapter.)
A area
Cv ventilation heat loss
cp specific heat; power coefficient of wind turbine
E energy
G mean rate of moisture emission; gust ratio
gx mean inside moisture content
go mean outside moisture content
H effective height
/ insolation
k extinction coefficient
N number of room air changes per hour
AP indoor—outdoor vapour pressure difference
Ap pressure drop between points (of an element of construc-
tion)
q energy transferred to the water in a solar collector
R thermal resistance
rt individual thermal resistance
ry vapour resistance
T temperature
tai inside air temperature
tT mean radiant temperature
tres dry resultant temperature
At temperature drop
U thermal transmittance
V wind speed, volume
v indoor air speed, volumetric rate of air change
Y admittance
p density
0 tilt angle
/i absolute viscosity
(3) Abbreviations
AIA American Institute of Architects
BRE Building Research Establishment
BRS Building Research Station
BSS British Standard Specification
CIBS Chartered Institution of Building Services (formerly IHVE)
CSTC Centre Scientifique et Technique de la Construction
IHVE Institute of Heating and Ventilation Engineers
PCL Polytechnic of Central London
RIBA Royal Institute of British Architects
Units, symbols and abbreviations xiii
(4) Conventions
V(3X) denotes the square root of 3X.
(5) Conversion factors

Length
1 cm = 0.394 in.
1 m =3.281 ft
Area
I m 2 = 10.76 ft2
1 ha =2.471 acre
Volume
1 m3 =35.31 ft3
1 1 = 0.2642 gallons (US) = 0.220 gallons (UK)
Mass
1 kg = 2.205 lb
Density
1 kg/m3 = 0.062 lb/ft3
Force
1 N = 0.2248 1b (force)
Pressure
1 Pa = 0.004 in. H2O
1 kPa = 0.145 psi
Energy, work, heat

1 kJ = 0.948 Btu
1 kWh = 3414 Btu
1 GJ = 278 kWh
1 therm = 105.5 MJ
1 tonne coal equivalent (TCE) = 26.4 GJ
1 tonne oil equivalent = 44.7 GJ
Power
l k W = 1.341 hp
Heat flux
lW/m 2 = 0.317 Btu/(ft2 h)
Thermal conductivity
1 W/(m K) = 0.578 Btu/(ft2 h °F)
xiv Units, symbols and abbreviations
Heat transfer coefficient
1 W/(m2 K) = 0.176 Btu/(ft2 h °F)
Specific heat
1 kJ/(kg K) = 0.239 Btu/(lb °F)
Temperature
°C = (5/9)(°F-32)
Temperature intervals
1°C=1.8°F
Preface
The authors have rendered a valuable service to the building industry by

the preparation of this volume with its wealth of actually built, rather than
projected, examples. The book is well and soundly written with its empha-
sis maintained from start to finish on their basic theme, the necessity for
the use in buildings of energy in many forms and the increasing desirability
of learning how to minimize the use of non-renewable forms of energy in
meeting those essential needs. Directed primarily towards architects,
builders and owners in the United Kingdom, and consequently written
within the scope of SI units, the book will be helpful to all who plan to
build in latitudes north of the 50th parallel, where winter heating is more
important than summer cooling.
Proper ventilation and the admission of outdoor air under suitable
circumstances are not neglected, however. Building design features which
are recommended by the authors are liberally illustrated by photographs
and drawings of residential, institutional and commercial structures which
actually exist. The soundness of their recommendations has been verified
in most cases by the first-hand knowledge of the authors.
Many of the 'Energy' books which have appeared in recent years have
dealt almost exclusively with non-renewable forms such as the wind and
the sun. These are not neglected here but their applicability is subjected to
careful and objective analysis, with the intent of giving guidance which is
based upon knowledge and experience rather than upon enthusiasm alone.
The reader may well divide the book into three sections. The first,
Chapters 1 through 3, deals with energy and the built environment, the
building site and its energy attributes and with a general discussion of
building design principles. The second section, Chapters 4 through 7, is
devoted to space heating by passive and active solar systems, to conven-
tional space heating and to thermal storage as it can be accomplished with
the three currently available materials, water, rocks and phase-change sub-
stances. The third section, Chapters 8 through 13, discusses wind energy,
water supply and waste disposal, energy conservation in the home and case
studies for built and operating residential and institutional structures.
The appendices are filled with useful data pertaining to the atmospheric
temperatures and solar intensities which prevail throughout the UK. The
concluding appendix is particularly pertinent, since it deals quantitatively
with the amount of moisture which is added to the indoor environment by
the occupants. This is an aspect of design which is all too often overlooked
xv
xvi Preface
until it is encountered by the individuals whose cooking, cleaning, bathing
and breathing are responsible for the rise in indoor vapor pressure.
The literary styles of the two authors are so similar that the book is a
first-rate example of collaboration, with each writer contributing the
material with which he is most familiar. A convenient list of conversion
factors is included for the assistance of readers to whom SI units are still
unfamiliar. Design with energy is destined to become a valuable addition
to the libraries of both professionals and owners who are concerned with
the use and conservation of energy in buildings.
JOHN I. YELLOTT, M.B.E., P.E.

College of Architecture, Arizona State University
Tempe, AZ 85287 USA
Acknowledgements
The authors wish to thank: Paul Ruyssevelt for drawing many of the
diagrams; Christopher Martin for helpful comments on the chapter on
active solar heating; and Prof. John Yellott and Dr David Bullett for their
helpful comments on the whole of the text.
1
Energy and buildings
1.1 Introduction
Forecasting is an art in which all are likely to be wrong but some will be
more wrong than others. We expect the next few years to be a time for
reflection. A variety of opinions on the conservation and use of energy in
buildings now exists and a limited amount of data is available. Economic
recession has increased the pressure on public and private funds; and the
lack of national, let alone global, strategies for resource development has
hindered development of energy conservation projects and the exploi-
tation of alternative sources of energy. The result is an atmosphere of
caution which at its worst could result in inactivity and resignation and at
its best could lead to significant, if somewhat restrained, progress in the
use and conservation of energy in the built environment. This may have to
be enough - building, after all, is another art of the possible.
Our approach in this book has been to provide designers with a system-
atic framework for considering the use of energy in buildings. We have
tried to cover the more important of the wide variety of topics that must
be considered and to provide sufficient references to allow the reader to
pursue points of particular interest in depth. A great deal of information
has been produced during the last few years but virtually no overviews
exist. Different research groups work independently of each other and
often work with different aims in view. The overall result is a jungle of
papers, data, evaluations and opinions for which we have tried to offer a
guide.
Because of our past work we have a strong bias towards integration —
energy use in buildings is a function of the site, form, method of construc-
tion, controls available, pattern of use by the occupants and the psycho-
logical feeling of comfort, as well as the material and energy flows through
the building.
The structure of the book reflects this by starting with broader con-
siderations of site and design, then examining specific topics such as solar
and wind energy and then finishing with a fairly detailed examination of a
number of applications in both domestic and non-domestic buildings.
Our preoccupations have been the areas most familiar to us and so the
book is centred on northern Europe, especially the United Kingdom. How-
ever, because much of the impetus for the development of ambient energy
sources comes from the United States and because of the extensive work
1
Fig. 1.1. Energy consumption tonnes oil
in European countries equivalent/p/y
(tonnes oil equivalent/p/y).[ 11 LEGEND
(Primary electricity is that PRIMARY
derived from nuclear power ELECTRICITY
and hydroelectric plants.)
done there we have tried to bridge the oceanic gap and draw extensively
on developments in the US. From another point of view, characteristics of
the European built environment such as higher densities and lower accept-
able space heating temperatures are of relevance to energy use across the
Atlantic.
We have not dealt with numerous topics ranging from alternative
sources of energy such as wave power to feed the national grid to the legal
aspects of alternative sources of energy, for example overshadowing of
solar collectors by neighbouring buildings. In certain cases we have pro-
vided references to specialist topics not included in the text.
1.2 Energy and the built environment - past and present

It is worthwhile recalling that, historically, oil production is negligible. On
a broad timescale the oil age presently coming to its conclusion is rather
insignificant, although all of us have been profoundly marked by its effects.
An alternative solution to energy supply and use must be found based on
renewable sources of energy.
Fig. 1.1 shows per capita consumption of energy in various European
countries and the energy sources that contribute to the figures.
Europe on the whole is a net fuel consumer but the UK is in the
unusual and fortunate position of being a net producer. For 1981, esti-
mated total consumption of primary energy was 317 MTCE compared to a
production of 350 MTCE.l2! (Primary or gross energy is the calorific value
of the raw fuel, for example oil, coal, natural gas, nuclear and hydroelec-
tricity. Delivered, or net, energy is the calorific value of the fuel actually
received by the consumer.)
Energy and the built environment - past and present
Total primary energy
Renewables
Nuclear (and hydro)
Natural gas
1970 1980 1990 2000 2010 2020

Fig. 1.2. Primary energy
projections for a low-energy
Coal is of major importance in the UK energy supply. Coal supplies are scenario. W
well distributed throughout the UK and estimates of technically recover-
able reserves are in the order of 45 000 MTCEJ31 By comparison, recover-
able UK North Sea oil reserves are estimated at about 6000 MTCEJ4!
During the 1960s, pressure from oil led to a drastic drop in production to
110 Mt/y. In 1974 after a miners' strike a plan was developed to increase
production to between 125 and 150 Mt/y by 1986. In 1981 annual output
was about 128 Mt/y. 12J
In 1979 (when UK energy consumption was 356 MTCE and production
only 329 MTCE), it was estimated that if a 'low-energy' strategy was
adopted, in the year 2000 the UK could be entirely self-sufficient on
North Sea oil and gas and indigenous coal supplies (renewable energy
sources were assumed to contribute less than 2% of the total). Fig. 1.2
shows the makeup of energy sources in such a scenario, which aims for a
decrease in energy consumption through a combination of conservation
and increased efficiency in industry, the domestic sector, the commercial
and institutional sector, transport and energy-supply industries.
Essential to any low-energy strategy is a change in the energy use in
buildings. Before considering energy, though, it is worthwhile taking a
quick look at population patterns and built forms. Europe is much more
densely populated than the US. To the average American there was a hint
of this in the absurdly small sizes of imported cars — but such cars are per-
fectly in scale with road widths and the pattern of buildings.
The UK has a population of about 60 million and an area of
245 000 km2 — a density of about ten times that of the US. As one
observer, commenting on the dangers of nuclear power plants and reflect-
ing on the accident at Harrisburg, said: 'Ours is a small, densely populated
island. It is not possible to run very far or very fast.'l5!
Housing densities reflect this with average residential developments in
the UK having about 125—175 p/haJ 6 l A recent scheme for energy-
conscious housing uses passive solar gain and a density of about 40 houses/
ha. Looking back a bit, one of the present authors has a home in an area of
nineteenth-century, urban two-storey terraced houses which has a present
density of about 300 p/ha. The approximate breakdown of the UK housing
Fig. 1.3. Estimated gross
energy consumption of
buildings in the public-
service and miscellaneous I
categories.!11! (BRE,
Crown Copyright, HMSO.)
1s
Jo
I so
z
<
stock is semidetached houses, 31%; terraced houses, 30%; detached houses,

16%; purpose-built flats, 14%; converted flats, 8%, and miscellaneous,
1%J7)
It has been estimated that building services account for 40-50% of
the national consumption of primary energy. By sector, industry uses 41%
of the energy input; households use 29%; transport, 16%; and other users,
14%.I81 Of the mean net energy consumption per household of 81 GJ/y
the breakdown has been estimated as space heating, 64%; water heating,
22%; cooking, 10%; and TV, lighting, etc., 4%. The comparable mean gross
energy-consumption figure is 138 GJ/y and the respective percentages 54,
22, 11 and 13 J 8 1 In the domestic sector in the UK and in much of the
rest of northern Europe there is no requirement for air conditioning. This
is, of course, in contrast to a large part of the US.
During the space heating season it has been estimated that 41—61% of
households are intermittently occupied.l9l It is now common for occupied
living rooms to be kept at about 20 °CJ1O1 This represents a rise of about
1 °C per decade from a post-World-War II level of about 17 °C. The most
likely reason for this trend was reported to be that, given the choice,
people now prefer to wear less clothing in a warmer room than more
clothing in a cooler one, although their thermal comfort would be the
same in both. (See Chapter 3.) In the interest of energy conservation for
society as a whole, a number of countries have taken steps to limit tem-
peratures in spaces. In Britain, for example, temperatures in non-domestic
buildings should not exceed 19 °C.
In non-domestic buildings the pattern of energy use varies greatly. Fig.
1.3 shows energy-consumption estimates.
In the non-domestic sector lighting, and hence electricity, plays a much
greater role. Cooling loads, often linked to lighting as an unwanted heat
source, are also often important.
Energy and the built environment - the future 5
Table 1.1. Approximate delivered energy demands (for all energy
requirements)
Delivered energy demand

Building type (MJ/m2 floor area/y)
(1) Commercial glass houses'* 2500

(2) Large office buildings and large
shops6 1700
(3) 1975 schools0 1280
(4) 2000 low-energy school target0 600
(5) 1982 house<* 610
(6) Extremely well-insulated house
of the future e 300
"Ref. [12]
*Ref. [4]
°Ref. [13]
d
Authors' calculation updating^14! for 1982 Building Regulations.
e
Based on the Wates House at the National Centre for Alternative Tech-
nology J 15 1
An approximate idea of the energy consumption in a range of building

types is given in Table 1.1.
1.3 Energy and the built environment - the future

The built environment changes slowly — a commonly cited figure for the
lifetime of buildings is 60 years although, of course, this varies greatly.
Office buildings are likely to last 30—70 years and houses 60 years or more.
Servicing systems change more rapidly, with innovation and replacement
occurring, say, every 10-30 years. In the last 40 years in the home of one
of the authors the heating system has changed from open coal-burning
fires to a mixture of gas room heaters and electric resistance heaters to gas
central heating with an open fireplace for amenity. Each change has rep-
resented an increase in comfort and, because of this, probably in primary
energy consumption, although the efficiency of use, particularly from
open coal-burning fires to central heating, has been improved.
Will the next change be to solar energy? Probably not or at least not in
the immediate future. Solar energy influenced both Greek and Roman
urban planning and house design and in the latter civilization by the end of
the first century the use of the sun for heating had become so important
that laws were enacted to protect a structure's exposure to the sunJ 16 !
For the next 2000 years solar energy applications were limited but diverse
and included lenses for heating copper (Leonardo da Vinci) and coupled
copper tanks partly filled with water and air to produce pressurized air to
play organ pipes (de Caus)J 17l
In the nineteenth century the foundations of widespread use of solar
energy were laid with experiments in steam generation to drive an engine
and pump (1876), generating electricity by focussing sunlight on
6 Energy and buildings
selenium-platinum junctions (1876), and theoretical studies on solar
radiation intensity and transmission through glass.! 17J In the early twen-
tieth century, flat-plate collectors were being used for domestic hot-water
heating in California. Since then, numerous attempts have been made to
use solar energy actively and passively in building design and services, but
until the 1970s the ready availability (for part of the world's population)
of oil delayed research and development for alternative sources of energy
and meant that those applications available, as well as a number of energy
conservation measures, were not cost-effective.
In the last decade, increasing prices and a realization that fossil-fuel
resources are limited encouraged solar energy work. One UK project on
which the present authors participated was that of the Cambridge Univer-
sity Autarkic House,!18' 19> 2°1 a plan for a house that would be com-
pletely independent of mains services, that would rely solely on solar and
wind energy for heating and power and that would recycle water and
wastes to make the best use of scarce resources. If ever a project was a
child of its time, this was it, but unfortunately it was a child of the world
and an orphan in the UK with this country's North Sea oil and gas and its
300-year supply of coal. The project was not continued but it did stimu-
late enormous interest. Unfortunately, many of the problems it had hoped
in part to alleviate remain.
For example, the use of solar energy for space heating is environmen-
tally sounder than the use of coal. Coal is a source of pollution and perhaps,
on a large scale, a major danger to our climate. One immediate problem
with coal (and oil and natural gas) is the so-called 'acid rain' produced
when sulphur dioxide and other contaminants from power stations com-
bine with vapour in the atmosphere to form sulphuric acid. Upon falling to
earth as rain it seeps into the soil, releasing aluminium and manganese, and
poisoning trees. It has been estimated that half of Germany's trees are
suffering from this pollution.!211 In Norway, the effect of Britain's power
stations has been to destroy fish life in more than half of the country's
riversJ 21 !
In the long term, fossil-fuel burning affects the carbon dioxide content
of the atmosphere and this is presently a subject of great concern. The
danger is that the increasing CO2 content could lead, via the 'greenhouse
effect', to a retention of heat that would otherwise escape from the lower
atmosphere and so warm the earth's surface and perhaps drastically affect
climate. More knowledge of the carbon cycle is required but, in the mean-
time, solar energy should not be dismissed as uneconomical or unnecessary
because of abundant coal. In both the short and long terms we should be
striving for environmentally safe solutions to energy supply and, indeed,
ultimately we must find them.
Some results of the research into the potential of solar energy have
been encouraging. For example, one recent study! 6 ' of passive solar heat-
ing of houses showed that a number of measures, including lean-to con-
servatories and glazing the south-facing aspect of a roof so that it may be
used as a warm-air plenum, are likely to be cost-effective in the not too
distant future. Reports on active solar heating systems for domestic hot
water vary. The Building Research Establishment (BRE) finds them
uneconomic at present but estimates that costs might be reduced to about
50% of the current figure for a 4.5 m2 system by attention to collector
design (with significant potential for high-performance collectors) and to
satisfactory installation on buildingsJ 22! However, at least one report has
indicated that domestic solar hot-water heating systems will soon become
economically viable.I23' One difficulty facing all ambient energy sources
is that they are judged by different economic criteria than are con-
ventional and nuclear sources. Similarly, they are not given the generous
tax allowances that these latter often receive. Thus, for example, a com-
parison between solar energy and energy from coal, where the UK Central
Electricity Generating Board is heavily subsidized, is necessarily biassed.
A general consensus seems to be that, given present trends, which is
another way of saying if we bumble along and continue to spend phenom-
enally more on nuclear power than alternative sources of energy, the
'alternatives' may be contributing something under 5% to the UK primary
energy requirement in the year 2000.1 4 » llj 241
The most likely development in the field of energy and buildings and
one which is to be welcomed is greater emphasis on energy conservation.
Ever since cow dung was applied to the walls of dwellings by our ancestors,
the benefits of increased insulation and greater thermal comfort have been
appreciated. Perhaps because it is one of the crudest and easiest ways to
conserve energy, as well as one of the most cost-effective, greater insu-
lation has been widely adopted. The dramatic trend in insulation of
domestic buildings is shown in Table 3.11.
There are, of course, numerous other ways to conserve energy. Good
management can be exceptionally effective — particularly in existing build-
ings that have undergone numerous transformations; over the years the
scope for conservation is enormous. Techniques now available such as heat
recovery on extract air and activated carbon filters, which permit a greater
proportion of recirculated to fresh air in controlled ventilation systems,
can be expected to reduce energy consumption in the near future. Other
means are only in their infancy — controls are likely to be one of the
greatest areas for development. On the domestic scale, for example, one of
the present authors is currently developing a home-energy management
system linked to a microcomputer which will sense temperatures through-
out the home and outdoors, sense solar radiation and wind speed, allow
for the thermal response time of the house and provide desired comfort
conditions according to as complex a schedule as the occupants desire —
all they will have to do is vary the program at the computer console which
may well be linked to their own television for display. Additional features
include control of the electrical loads, an energy-accounting system for
recording consumption and paying bills and a home-security system.
The BRE has estimated that a combination of energy-conservation
measures and the use of ambient energy sources could save 15% of the
primary national energy consumption.!11 J The programme would include
combined generation of heat and electricity in which the waste heat of
electrical generation would be used in district heating schemes, widespread
employment of heat pumps, increased thermal insulation, controlled
mechanical ventilation, solar collectors to supplement the hot-water
supply of existing dwellings and a limited use of aero generators for space
and water heating.
In the following chapters, we have tried to present a systematic
approach to energy conservation and the use of ambient energy in build-
ings and to provide information on a variety of techniques, not all of
which are presently cost-effective, but which we expect to be in the short-
to medium-range future.
The question of costs is a vexed one, in part because of its relation to
the assumed future real cost of energy. The uncertainty of energy costs
affects everyone and large organizations such as the Central Electricity
Generating Board are attempting to develop flexible, step-by-step strat-
egies which try to ensure that, whatever the future vagaries of energy
supply and demand, the requirements of customers for economical and
reliable supplies of electricity are met. Predictably enough, one of the
central strategies is the continued development of nuclear power. Another
is to explore better ways of burning coal and a third is to identify and
develop renewable sources of energy which may be useful for electricity
generation J 25 1
An approach to cost-calculations, which is gaining in favour, is lifecycle
costing in which assumptions about the future, such as those for fuel costs,
are made explicit. When energy-conservation measures are assessed on this
basis they are often much more attractive than on the simple payback
method commonly used. Nevertheless, there is no clear consensus as to
what is to be done and in what order. As a preface to the work which
follows we present the results of a number of studies of the comparative
value of energy-saving measures. One of the present authors made one of
the first detailed studies and the results are given in Table 1.2. (Because of
the context of the study, passive solar measures were not included.)
Fig. 1.4 shows the Electricity Council Research Centre's results for a
number of conservation measures based on dividing the potential annual
energy savings by the additional capital cost (above that of an electric
warm-air heating system).
Fig. 1.5 shows the results of a study of energy conservation in the
Felmore Housing Scheme at Basildon, a group of 430 houses supplied by a
district heating scheme.
Of course, there are many considerations other than cost involved in
energy conservation - these vary from providing resources for future
generations and achieving self-sufficiency in energy in order to escape
political dependence on countries controlling the energy supply, to indi-
vidual decisions about comfort and a desire to have the most modern tech-
nology available in one's home.
We believe that we are moving towards a community which uses energy
and resources both more efficiently and in closer harmony with nature -
but progress is, and will continue to be, slow. There is a great need for
research into materials, equipment, techniques and attitudes. The vicious
self-fulfilling argument that 'since the renewable sources of energy have
Energy and the built environment - the future
Fig. 1.4. Ranking of energy-
saving measures.! 2 7 1
(Internal temperatures are
20 °C/16 °C.)
Insulation, existing
and new designs
32.2-82.7 (MJ/(£ y)
Heat pump, proposed
new designs
36.0-64.7 MJ/(£ y)
Ventilation
heat recovery
Solar
10.8-18.0 MJ/(£ y) collector
5.4-14.4 MJ/(£y)
Saving per item Saving per group
A Avoid waste
§ Improve layout and orientation

Improve shape (surface area)
Reduce glazing area JO]
4 Double glaze
(|) Improve insulation (med-low U) 25
© Reduce ventilation (1 1 / 2 -1 ac/h) 15
7 Use shutters on windows !5^T5$$$$$^$^^
B Reduce standards
1 Reduce int. temp. 18-15.5 °C Jo]

2 Reduce int. temp. 18-11 °C
C Increase efficiency of plant

© Improve responsiveness 15
(2) I mprove boiler efficiency ^20M$$S$$SS$S$$^^ 10 25|
D Use most efficient fuel Assuming a cumulative saving in A, B, C of 75%
© Coal conversion efficiency 50% = 50 units total

2 Gas conversion efficiency 42% = 60 units total
Fig. 1.5. Aspects of energy

conservation at the Felmore
Housing Estate in
not been proven to be reliable and capable of contributing significantly to Basildon.l 2 8 ! (The main
our energy supplies, the capital allocated to their research and develop- design effort was placed on
items marked by a circle.)
ment must be severely limited' will have to be attacked.
It seems reasonable to expect the following for the near future in the
field of energy and buildings:
(1) The building stock will not have changed greatly; design pro-
fessionals will be working largely on existing buildings; both new and
existing buildings will be vital to improving the national use of energy.
Table 1.2. Ranking of energy cost-effectiveness for measures of conservation and alternative energy
supplies^26]
i Annual GJ/£*'
Measure ]Break-even 9% dividend
Roof insulation 50 mm DIY :5.8 1.2

Roof insulation 100 mm DIY :IA 0.7
Roof insulation 150 mm DIY ]1.8 0.5
Roof insulation 50 mm Installed ]1.4 0.4
Double glazing — DIY ]1.5 0.5
Roof insulation 200 mm DIY 11.4 0.4
Roof insulation 250 mm DIY 11.2 0.3
Wall insulation (cavity) 100 mm Installed 0.9 0.3
Roof insulation 200 mm Installed ().8 0.3
Double glazing — Installed ().8 0.2
Double glazing with heat-reflecting coatings — Installed 0.8 0.3
Heat recovery from hot water — Installed ().7 0.2
Wall insulation (cavity) 150 mm Installed ().7 0.2
Roof insulation 250 mm Installed 0.7 0.2
Double glazing with heat-reflecting coatings
and krypton — Installed 0.7 0.2
Single glazing with foam blinds — Installed 0.6 0.1
Wall insulation (cavity) 200 mm Installed 0.6 0.2
Electric heat pump in new house — Installed 0.5 0.1
Wall insulation (solid) 50 mm Installed 0.4 0.1
Conversion from oil to electric heat pump _ Installed 0.4 0.1
Conversion from electric resistance heating
to electric heat pump — Installed 0.3 0.1
Heat recovery from warm air — Installed 0.3 0.1
Solar collectors for water heating on wall
of:
New house ().3 0.1
Existing house ().3 0.1
Solar collectors for water heating on roof
of:
New house C).3 0.1
Existing house C).3 0.1
Solar collectors for space and water heating
without long-term storage on wall of:
New house c).2 0.1
Existing house c).2 0.1
without long-term storage on roof of:
New house c).3 0.1
Existing house c).2 0.1
with long-term storage on a house of floor
area 100 m2 c).l 0.0
Aerogenerator c).l 0.0
(2) Energy sources will be varied; there will be greater reliance on solid
fuel; renewable energy sources are still in their infancy; energy conser-
vation will change the demand profile.
(3) The construction industry is unlikely to have evolved greatly and,
consequently, the widespread use of techniques of energy conservation
requiring a high degree of skill on-site is Likely to prove difficult. However,
the introduction of components such as prefabricated and preinsulated
panels for timber-frame housing, which are common in Scandinavia, may
challenge traditional building methods on cost grounds.
(4) The preference for houses rather than flats is unlikely to change;
more one- and two-person households may be expected; dwelling size will
continue to be small although smaller household sizes will mean more
space per person.I 29 1
(5) Houses will be better insulated and will be more airtight.
(6) Heating systems will become more sophisticated with microelec-
tronic-based control systems contributing to energy conservation.
(7) Occupants will become more interested in energy conservation and
more adept at controlling their environment.
In many ways the situation is analogous to that of buildings in New
York City just before the widespread introduction of lifts. The more far-
sighted designers realized the potential of lifts, and at relatively little cost
to their clients provided a lift shaft which could be used later according to
the capital available. We shall argue for the provision of a large number of
'shafts' and a smaller number of lifts'.
We believe that today's buildings should be considered by their
designers and users to be worth preserving for at least the next 50 years
and hope that in the remaining chapters the reader will find advice that
will help to make them so.
Notes to Table 1.2

a „ . . Annual energy
£bZ (GJ) saved by the measure
Break-even value = -^—- -
Annual cost of installing the measure (i.e. mortgage interest plus capital repayment)
9°/ d' 'd d 1 - Annual energy (GJ) saved by the measure
Mortgage interest + capital repayment + 9% interest on capital
The 9% dividend figure is included for comparison since there is a school of thought which says that
when deciding, for example, to buy insulation one should compare the savings with the return one
could expect if, instead, one invested the money. This approach is rather stringent since energy con-
servation measures will normally increase the value of the property.
b
Figures in GJ/£ have been converted from the original figures which were in kWh/£ rounded to the
nearest 10.
c
Capital and the following fuel costs, have been left at 1976 levels. The GJ of fuel which could be pur-
chased for £1 are as follows, and they are quoted as mean costs over a 20-year period assuming either
0% or 10% inflation of fuel price per year.
Electricity (0% inflation) = 0.18 GJ/£
Electricity (10% inflation) = 0.07 GJ/£
Gas (0% inflation) = 0.50 GJ/£
Gas (10% inflation) = 0.18 GJ/£
Thus an occupant who uses gas and who assumes that the price will rise by 10%/y should carry out all
the measures in Table 1.2 with an index of more than 0.50 GJ/£. This implies all the measures above
'electric heat pump in a new house'. If the more stringent criterion of a 9% return is demanded, then all
measures above 'DIY double glazing' qualify.
References
[ 1 ] Anon. (1979). 'France Bottom of Energy Supply League'. Building
Services and Environmental Engineer, April, p. 22.
[2] Anon. (1982). Department of Energy Press Notice. Reference No.
51, March, p. 31.
[ 3 ] Anon. (1976). Coal for the Future Progress with Plan for Coal and
Prospects for Coal to the Year 2000. Department of Energy.
[4] Leach, G. (1979). A Low Energy Strategy for the United Kingdom.
London: IEED.
[5] Shayer, B. (1979). 'Harrisburg here?' Original Equipment Manu-
facture Design, May, p. 3.
[6] Turrent, D., Doggart, J. & Ferraro, R. (1980). Passive Solar Housing
in the UK. London: Energy Conscious Design.
[7] Romig, F. & Leach, G. (1977). Energy Conservation in UK Dwell-
ings: Domestic Sector Survey and Insulation. London: IIED.
[8] Anon. (1976). Energy Consumption and Conservation in Buildings.
BRE Digest 191. Garston, Watford: BRS.
[9] Desson, R.A. (1976). Energy Conservation: The Intermittent
Occupation of Dwellings and Domestic Energy Consumption. BRE
CP 37. Garston, Watford: BRS.
[10] Hunt, D.R.G. & Steele, M.R. (1980). 'Domestic temperature
trends'. The Heating and Ventilating Engineer, April, pp. 5 — 15.
[11] Bakke, P. (Chairman) (1975). Energy Conservation: A Study of
Energy Consumption in Buildings and Possible Means of Saving
Energy in Housing. BRE CP 56/75. Garston, Watford: BRS.
[12] Winspear, K. (1980). 'Energy and UK glasshouse crop production'.
In: Energy Conservation and the Use of Renewable Energies in the
Bio-Industries, Vogt, F. (ed.). Oxford: Pergamon.
[13] Romig, F. & O'Sullivan, P. (1979). 'Buildings and the energy
future'. In: Buildings: the Key to Energy Conservation, Kasabov, G.
(ed.). London: RIBA.
[14] Anon. (1976). Heat Losses from Dwellings. BRE Digest 190.
Garston, Watford: BRS.
[15] Todd, R. (1977). 'Low energy housing at the National Centre for
Alternative Technology'. University of Nottingham Conference on
Ambient Energy and Building Design. Welwyn: Construction Indus-
try Conference Centre.
[16] Butts, K. & Perlin, J. (1978). 'The middle age of solar energy use'.
Solar Age, 3(3), 24-5.
[17] Kemper, J.P. (undated). Pictorial History of Solar Energy Use.
University of Capetown, South Africa.
[18] Pike, A. (1972). 'Cambridge studies'. Architectural Design, 42 (7),
441-5.
[19] Pike, A. (1974). 'The autonomous house'. Architectural Design, 44
(11), 681-9.
[20] Littler, J.G.F. & Thomas, R.B. (1977). 'Solar energy use in the
Autarkic House'. Transactions of the Martin Centre for Architec-
tural and Urban Studies, 2, 93-110.
[21 ] Girling, R. (1982). 'Ten billion dying trees'. The Sunday Times,
4 April, p. 20.
References 13
[22] Wozniak, S.J. (1981). Cost and Performance of Solar Water Heating
Systems in the UK. BRE IP 14/81. Garston, Watford: BRS.
[23] Soldatos, P.G. (1975). UK-ISES - Midlands Branch Conference
on Solar Energy in the 80s - Technical and Economic Viability.
[24] Long, G. (1975). 'Solar energy: its potential contribution within
the United Kingdom'. Department of Energy Paper No. 16.
London: HMSO.
[25] England, G. (1980). 'Planning for uncertainty'. Lecture, 4
December. Central Electricity Generating Board.
[26] Littler, J.G.F. (1976). 'Priority in selecting energy conservation
methods and solar or wind heating in houses'. Autarkic Housing
Project Working Paper No. 27. Cambridge: University of
Cambridge, Department of Architecture.
[27] Siviour, J.B. (1979). 'Prospects for energy savings in home heating'.
CIBS Annual Conference.
[28] Anon. (1979). 'Felmore Housing, Basildon'. In: Buildings: the Key
to Energy Conservation, Kasabov, G. (ed.), London: RIB A Energy
Group.
[29] Hitchin, E.R. & Cheyney, B.S. (1980). 'Building trends and their
effect on appliance and heating-system design'. Building Services
and Environmental Engineer, 2 (10), 6-9.
Site planning and analysis
2.1 Introduction
Traditional site planning includes evaluation of the aesthetics of a site,
population densities, land-use patterns, slope, drainage, soil characteristics,
incident solar radiation, daylighting, exposure to wind and numerous other
considerations which are treated in standard works.!1* 2] j n ^ s chapter
these subjects will be discussed only when specifically applicable to the use
of ambient energy sources or the opportunity for energy conservation in
buildings. In the more recent past the attitude of many designers has been
one of ignoring both the natural characteristics of the site and the poten-
tial of solar and wind energy. Instead, they concentrated merely on avoid-
ing potentially deleterious effects such as summertime overheating.
Important exceptions to this way of thinking include Olygay & Givoni
who wrote classic works on climate and architecture J3» 4J In an age of
rapidly dwindling fossil-fuel reserves, though, it is important to use a site
to best advantage. Fortunately, much can be done to conserve energy
merely through good design, on both the large and small scale. In the
former category especially, the possibilities depend on social conceptions
of work, home and leisure but in the future we may see a closer inte-
gration of places of work and residence to reduce transportation energy.
This could be encouraged by a gradual renovation of cities, resulting in
their increasing attractiveness as places of residence and thus reducing the
tendency towards suburban sprawl. For example, near the city home of
one of the present authors, a former warehouse is being converted into
flats. Among the results are a higher density and, for the occupants, less
dependence on vehicular transport.
If urban densities increase in such a socially acceptable way it becomes
easier and more economical to implement technical measures such as
energy and materials recovery on sewage and solid waste.
Quantitative work!5' 61 has begun on some of these questions, particu-
larly for new developments, and several years hence we can expect to dis-
cuss issues such as whether an upper limit should be imposed on densities
to permit extensive exploitation of solar energy for space and water heat-
ing or whether such a limit might be too low to permit combined heating
and power to be provided economically J 7 1 Or perhaps there is a range of
options. The real difficulty, as suggested previously, is attempting to
anticipate what our sources of energy will be and what they will cost
during the next 50-100 years.
14
Introduction 15
In assessing a site, the meteorological scale of interest must be deter-
mined. A convenient classification^8! is
(1) macroscale including the regional scale;
(2) local scale (town and site conditions);
(3) microscale (around, on and within buildings).
Lacy I81 has also provided a very useful discussion of the way in which 25
principal meteorological elements (solar irradiance, temperature, humidity,
pollution content, wind speed, etc.) may be used in dealing with problems
of town planning, design (of individual structures), construction and build-
ing maintenance and running costs.
The macroscale is a 'given' element and on some of the colder, wetter,
greyer days in the UK, one is reminded of the apophthegm that the only
problem is architects' insistence on continuing to build out-of-doors. On
the local scale the designer must take account of:l 5 l
(1) The natural drainage and transpiration - these often differ from
that of the surrounding countryside and so can affect humidity.
(2) Smoke and waste gases produced by factories — gases and sus-
pended solids can screen localities from sunlight; in the past Manchester
and London lost nearly an hour of daylight because of this.
(3) Reduction of wind speed due to the increased surface roughness
over cities (see Section 2.3) — surface wind speeds also tend to be variable
and unpredictable due to the channelling effect of the buildings.
(4) More-effective absorption of solar radiation in city areas - this com-
bined with combustion of fuels and reduced wind speeds creates an area of
higher temperature, a so-called 'heat island'. The temperature of a large
industrial city can be 1-2 K higher than that of its surroundings. The
Swedish Council of Building Research, in its reporting format for solar
energy buildings, requires information on atmospheric clarity and for one
house at Linkoping, for example, has called for an investigation of radi-
ation losses due to local cloud formation produced by condensation trails
from intensive air traffic overhead.^9!
It is the microscale especially that interests us here because it lends
itself best to artificial changes and hence to an element of control.
Judicious selection of a site, based on a knowledge of existing or potential
variations in microclimates, may also permit the designer to overcome
climatic factors, as well as to exploit to the utmost the use of ambient
energy sources. The principal elements to consider are solar energy, the
wind, ambient temperature, soil and vegetation.
Fig. 2.1 summarizes the typical energy gains and losses in winter and
summer for any structure. The sun will, of course, determine the solar
radiation loads and the wind will affect air movement through the struc-
ture, as well as changing the external surface resistance (see Appendix 2).
Siviour, in a study of the effects of weather on house-heating require-
ments^ 11 1 listed the following parameters as those expected to be of
importance:
outside air temperature;
inside air temperature;
outside radiant temperature;
16 Site planning and analysis
Fig. 2.1. Typical energy (a) WINTER
gains and losses in winter Ambient Weather
and summer for any
structured 10] Energy Output Parameters Energy Input Parameters
Thermal conduction —- Positive

and convection • Internal occupancy loads
Air exfiltration Residential or Internal appliance loads
Internal lighting loads
Radiative losses commercial Solar radiation loads
Negative
Water drains structure Infiltration loads
Heating system
input
(b) SUMMER
Ambient Weather
Energy Output Parameters Energy Input Parameters
Positive
Internal occupancy loads
Radiative losses •*— Internal appliance loads
Internal lighting loads
Air exfiltration —*. Solar radiation loads
Infiltration loads
Water drains -*— Thermal conduction and convection
Hot-water load
Air conditioning
(heat extraction)
wind speed;
cloud cover;
rainfall;
solar radiation;
vapour-pressure difference;
orientation.
For the unoccupied semidetached houses he studied he found statistically
that outside air temperature and solar radiation were the most important
weather variables governing heating requirements. Other variables such as
wind, cloud and rain had only a small effect. It is important to note, how-
ever, that the test houses were unoccupied - occupied ones will have all
too human complications such as people opening windows with the heat-
ing system on - and that the results are a function of the building. A well-
insulated home designed with large areas of south-facing glazing for passive
solar gain will show a marked performance difference if its orientation is
changed by 180°.
2.2 Solar radiation

The evaluation of the solar energy incident on building facades has been
the subject of considerable scientific and practical world 12 ' 13> 14> 151
because of its importance for interior comfort conditions for daylighting
Solar radiation 17
Fig. 2.2. Stereographic sun-
paths for latitude 53° north
with shading mask super-
imposed.! 17 1
4.00
and, commercially, for air-conditioning loads. Appendix 1 provides data

on solar energy in the UK — here we will examine some practical conse-
quences and applications.
Suppose we have a house with a south-facing bay window (P) and wish
to determine whether a neighbouring building to the south-east will signifi-
cantly reduce the direct passive solar gain into the window. We start with a
stereographic sunpath as shown in Fig. 2.2. Then, as in Fig. 2.3, we draw a
section and plan through point P in order to calculate the vertical and
horizontal angles subtended at Pby the obstruction.
These are marked on a shading mask and an outline of the mask is
superimposed on the stereographic sunpath diagram as shown in Fig. 2.2.
From this it is possible to see, for example, that point P will be shaded
from the sun's direct radiation between approximately 8.45 a.m. and
10.15 a.m. on 15 October. (A more impressive and immediate method is
to use a device^16! which is similar to a sextant in which a tiny telescope
mounted on a level platform views the obstructions through a grid appro-
priate to the latitude. The observer stands at the proposed site of the
building and can see through the grid an indication of the times at which
any object casts a shadow on the site. There is no need to measure the
heights of obstructions and there is the further advantage that the image
may be photographed.)
Then, with data for the average distribution of direct solar radiation
throughout the day in question,! 13] we can assess the importance of
Fig. 2.3. Determination of
overshadowing by neigh-
bouring buildings, (a) Sec-
tion, (b) Plan, (c) Construc-
tion of a shading mask.
(a) Section
Axis for azimuths

of points 1—3
Axis for elevations of points 1—3
(c) Construction of a shading mask
shading. A more refined calculation would also take diffuse radiation into
account. A simple model of diffuse radiation would probably assume an
isotropic sky — more-complex models^15! will refine this assumption to
allow for a circumsolar diffuse component and an isotropic diffuse com-
ponent.
Similar calculations for other days during the heating season facilitate
the estimation of the loss of passive solar gain. (Chapter 4 discusses passive
solar buildings in detail.)
The obvious next step is to consider how houses should be grouped to
best use the sun's energy. Pennyland, a local-authority scheme for 177
houses in Milton Keynes, was designed to take advantage of passive solar
gain. The planning density is, very roughly, about 100 p/ha of the total
site area, a figure which gives considerable scope for selective orientation.
Ideally, for maximum passive solar gain, houses should be orientated
towards the south (see Chapter 3) but this is, of course, not always
Solar radiation 19
possible and, indeed at times, not desirable for aesthetic or other reasons.
In Pennylandl18! it was necessary to consider, in addition to passive solar
gain,
(1) the need to establish a movement network with special emphasis on
footpaths and cycleways to reduce mobility energy requirements;
(2) the shape and contours of the site, existing hedgerows and trees,
land use, density and tenure policies; and
(3) the need to relate to the development in adjoining areas.
One of several preliminary studies investigated orientating all houses
due south, as shown in Fig. 2.4, but this resulted in a conflict between the Fig. 2.4. South-facing
housing groups and the footpath network. Similar conflicts occurred as the housing option for
Pennyland.flSJ
orientation constraint was relaxed further until the eventual solution,
which has now been constructed, of 75% of the houses within ± 30° of
due south, 20% within ± 45° of due south and 5% aligned east-west (see
Fig. 2.5). The final effect is incontestably more pleasing and should appeal
to designers concerned about energy constraints encouraging rigid and
monotonous developments.
Fig. 2.5. Final layout for

Pennyland. [ 1 9 1 (Courtesy
of Milton Keynes Develop-
ment Corporation Building
Department.)
The mean distance between approximately parallel rows of south-facing
reduction houses is about 23 m. The extent of overshadowing of one group of houses
by another for the design options was estimated using specially developed
solar shading calculators (see Fig. 2.6).
Similar considerations influenced the Northlands 1 Housing at
Basildon.I20 '211 All houses face within ± 22° of south and overshadowing
is avoided by a spacing of about 20-30 m between adjacent rows of two-
Fig. 2.6. Use of a solar
shading calculator. (Building storey buildings (see Figs. 2.7 and 2.18).
group A reduces radiation Turrent et al. 118J provide the following rules of thumb for the spacing
on B by less than 10% and distances between two-storey terraces assuming a flat site:
on C by 10-20%.) (Courtesy
of B.Everett.) daylighting 6—10 m;
sunlighting 10—16 m;
privacy 15—18 m;
passive solar 20—23 m;
A separation of 20-23 m between buildings of a height of 6.2 m
corresponds to a vertical shadow angle of 14.6-16.7° and since the solar
altitude falls to 17° at latitude 50° N on 22 December at noon there will
still be some overshadowing of ground-floor windows throughout part of
the day. However, solar radiation in the depth of winter is not very signifi-
cant (see Appendix 1) and so a degree of overshadowing may be accept-
Fig. 2.7. View of Basildon able. At Basildon, for the particular design of two-storey terraces houses
Northlands 1 Housing used, and taking a ventilation rate of 1 air change per hour and an internal
showing insolation of south
facades over lower temperature of 21 °C, it was shown that substantial overshadowing result-
buildings.!211 (Courtesy of
Antoine Rafaul.)
Solar radiation 21
2
Table 2.1. Density, plot width and energy lossl ^
Density Number of Plot width* 10% 20%

(p/ha) storeys (m) energy loss energy loss
100 1 5 ?b P
6 P P
7 P P
125 2 5 P P
6 ? ?
7 F F
180 3 5 F P
6 F F
7 F F
a
The cases shown are for orientations of due south.
b
P (pass) indicates that at the given density, a plot width is compatible
with no more than the energy loss shown; F (fail) represents incompat-
ibility and (?) represents an inability to make confident predictions with
the mathematical model used for the analysis.
ing from a spacing between rows of about 10 m increased the energy

demand by 6% over that of the unshaded house.l 22 ' 23J
Plot width is another element in the solar energy—land-use matrix. 6
Cathain and Jessop at the Martin Centre in Cambridge have studied some
of the theoretical aspects of layout for the Milton Keynes Development
Corporation. Using several proposed density figures and the solar shadow
prints of Fig. 2.6 they were able to analyse which combinations of plot
width, building height and density were consistent with 10 and 20%
energy losses due to overshadowing. Table 2.1 summarizes their results.
Narrow plot widths (with correspondingly deeper plots) reduce energy
losses, and at higher densities it becomes more difficult to make maximum
use of passive solar gain. Turrent et al. ,118J however, believe that with
'some design ingenuity' it should be possible to increase the maximum
compatible density up to 180-200 p/ha. To put this figure in context
they cite 1978 public-sector housing-scheme figures of 125 p/ha (38%),
150-200 p/ha (38%), and above 200 p/ha (24%); densities in the private
sector are generally lower at 100-150 p/ha.
Vegetation has been associated with housing for millennia by both
native and professional designers attempting to create an aesthetically
attractive and a functionally well-tempered environment. The influence of
natural vegetation on the microclimate near the house can include reduced
solar radiation, lower wind speeds, cleaner air due to the filtering of pol-
lution products, reduced noise levels and lower temperatures caused by
e vapot ranspiration.
The reduction in solar radiation due to deciduous trees is well known
and designers often include them in plans without perhaps always con-
sidering the question in adequate depth (or, more appropriately, height).
There seems to be a certain belief that once the leaves fall, so does the
tree, or at the least that the tree turns transparent. Careful observation
Fig. 2.8. A horse chestnut

tree in Cambridge in July
and December.
Fig. 2.9. Relative light

penetration at five sampling
sites for North American
barren trees.t 2 5 ! (Heights
indicated are approxi-
mations.)
Young ash Native pecan Sassafrass
5m 14m 8m
Black oaks
10 and 12 m
quickly reveals, however, that this is not so. Fig. 2.8 shows a horse chest-
nut tree in July and December and, obviously, even such a graceful young
tree is not to be ignored in the winter.
Fig. 2.9 shows the results of one of the rare scientific studies of sunlight
penetrating barren trees (in the US). The variability among species is par-
ticularly noteable.
Other variable aspects are the times of leaf initiation and leaf fall. While
these will vary from year to year it would be interesting to see a study
comparing the periods when certain trees are bare with a measure of the
space heating requirement, such as degree days.
Mature height also depends greatly on species. To the south of houses
we would suggest smaller trees. Depending on the distance from the house,
trees in the 5—10 m range are recommended if passive solar gain is to be
maximized. The following are just a few of the suitable species: maple
Solar radiation 23
Transpiration Fig. 2.10. Main water
movements and factors for
consideration when
CLIMATE investigating subsidence
damage.!27 J (Courtesy of
P.G. Biddle.)
BUILDING
MOVEMENT
(LEVEL AND
CRACK PATTERN) Rajnfa,,
Run-off
I LOAD I
Inflow
(including leaking
y drains)
SOIL
PERMEABILITY Outflow
O
Drainage
(Acer capillipes), dogwood (Cornus kousa chinensis), magnolia (Magnolia

spp), weeping elm (Ulmus glabrd) and fruit trees.
Numerous other considerations, both aesthetic and technical, may enter
when siting trees, and a good standard workl 26 ! will deal with many of
these. Suffice it to say here, because of the possibility of underground
thermal storage being used in the house (see Chapters 5—7), attention
must be paid to possible foundation damage. Fig. 2.10 shows the main
water movements and factors for consideration when evaluating possible
subsidence damage.
For buildings in non-shrinkable soils, that is, chalks, gravels, sands and
loams, there is no threat of damage from trees and hence no need to
specify a minimum distance of trees from buildings, according to Flora.l 27l
However, on certain shrinkable clay soils and with the rooting systems of
some trees, damage is possible^28! although experts disagree on the extent
of the problem and its implications for tree siting: on the one hand, there
is a rule-of-thumb, based on the idea that the radius of root spread of a
tree generally equals its height, that single trees should be located a dis-
tance equal to the height of the tree at maturity away from a building.^27!
On the other hand, Flora, who disagrees with previous recommendations,
maintains that 'any building which is adequately founded to cope with all
problems normally associated with building in shrinkable or expansive clay
soils, should present no problems when trees are planted' J 27 1 The thumb
rule cited corresponds to an angle of 45° if the tree is in line with the sun
at noon and a collecting surface such as a window, thus it would normally
have to be moved back (to about 3.8 times the mature height) anyway to
completely avoid interfering with passive solar gain. Nonetheless, attention
should be paid to both loss of solar radiation and possible structural
damage in shrinkable clay soils when locating trees, particularly those dis-
placed to the east or west of the line due south from the window.
2.3 Wind
In temperate climates and particularly where summer cooling problems are
not significant, as is generally the case in northern Europe, site design
usually centres on reducing the wind speed around the building to mini-
mize heat loss by infiltration and conduction—convection at the surfaces
(see Fig. 2.1). The potential conflict between such a situation and one
where a windmill is sited near the building is evident; additionally the
means of reducing the wind speed, which are usually vegetation or fences,
may interfere with passive solar gain. We will first examine the more
typical case where there is no windmill.
Avoiding exposed sites is a commonsense first rule — annual mean wind
speeds at the summits of hills can be from 30 to 100% higher than the
annual mean wind speeds in the surrounding districts.^29! The BRE have
devised a system for assessing wind loads, which accounts for local topo-
graphic influences and for the surface roughness of the environment 130l
(see Chapter 8). For the latter the categories used are
(1) open country with no shelter;
(2) open country with scattered windbreaks;
(3) country with many windbreaks, small towns, outskirts of large cities;
(4) surface with large and numerous obstructions, for example, city
centres.
The effect on the wind speed varies with height: at 3 m or less the BRE
factor used for determining the design wind speed for the first category is
about 50% higher than for category 4; at a height of 50 m the comparable
figure is 10%.
The second rule is to lay out houses to reduce wind speed. Fig. 2.11
shows some simple schematics of air flow around buildings.
On the windward wall there is a build-up of pressure and on the leeward
walls a reduction of pressure, that is, suction; the roof can experience
either pressure effect depending on such factors as slope and angle relative
to the wind direction.!30' 321 The wind-affected infiltration is generally an
air flow through the structure from windward to leeward wall and roof
surfaces.
Reducing the wind velocity at low level cuts infiltration losses and pro-
vides a more comfortable environment with fewer draughts, as well as a
building that consumes less energy and, although this has long been
known, little quantitative work has been undertaken to assess how to
Wind 25
Fig. 2.11. Schematics of air
flow around build-
ings.[30, 31]
(a) (b)
(c)
reduce the wind speed most effectively. At the Princeton Centre for
Environmental Studies, research has included an analysis of the relative
benefits of fences, houses and trees as wind barriers.!321 Fig. 2.12 shows
the wind flow, as represented by smoke patterns generated in a wind
tunnel, for an isolated house and for houses sheltered by fences, upwind
houses and simulated trees.
In Fig. 2.12{d) the wind flow is basically that of an oncoming stream
stagnating against the windward wall of the house and then flowing up and
over the windward roof to the peak from which it separates. Where the
external pressure is higher as on the windward wall, flow tends to be into
the house through poor seals around windows, cracks around doors and so
forth; where the external pressure is lower, as in the visible wake on the
leeward side, air flows from inside to out through any available conduit.
The infiltration loss can be reduced by blocking the paths available for
flow - through better construction, draught sealing, etc., as described in
Chapter 3 — and by sheltering the house from the deleterious effects of
the wind. Interestingly, a series of tests on the unsheltered house indicated
that infiltration was worst when the wind blew towards the house at a 45°
angle. This was due mainly to changes in the pressure produced by the
wind angle on the roof and leeward wall surfaces.
Fig. 2A2(b) shows how a solid fence in particular affects the flow
pattern on the windward side. Experimental results indicated that a fence
one-fifth the height of the house and approximately one house height to
the windward side reduces air infiltration by about 30% when the wind is
perpendicular to the house (this falls to about 15% when the wind angle is
45°). It should be noted that this was in unheated models. In heated build-
ings infiltration will also depend greatly on the stack effect (see Chapter 3).
In Fig. 2.12(c) the much more balanced pressure distribution produced
over the sheltered house is apparent. As it was not the object of the par-
ticular study to assess spacing effects, quantitative results were not pro-
vided but the authors did indicate that house sheltering should be far more
effective than fence sheltering. For the windward boundaries of com-
Fig. 2.12. Flow patterns over
houses.!321 (a) Isolated
housing - no sheltering.
(b) Solid fence provides
sheltering, (c) Upwind house
provides sheltering, (d)
Simulated tree provides
sheltering. (Courtesy of
G.E. Mattingly.)
munities of such houses they suggested that both deciduous and evergreen
trees might be used.
Fig. 2A2(d) shows the flow pattern resulting from sheltering by an
evergreen tree. Of particular note is the relatively calm wake similar to the
effect produced by an upwind house. Wind-tunnel results indicated that a
single row of trees approximately the height of the house and planted
1.5-2.5 house heights upwind from the house can reduce the air infil-
tration by 40%. A combination of fence and trees results in a reduction of
about 60% compared to the case of an unsheltered house.
It was then estimated that, given that air infiltration was responsible for
about one-third of a winter heating bill, fence and tree sheltering could
result in a 20% reduction in total fuel consumption, depending upon the
prevailing wind direction and the outside temperature. This estimate may
be compared with a fuel reduction of 40% in a house in South Dakota
Wind 27
with plants on three sides, compared to an identical one with no wind-

breaksJ 33 !
Landscape elements such as trees and hedges are thought to be more
effective than solid barriers, such as fences, by several authorities J32> 34J
Branches, leaves and evergreen needles tend to 'filter' the wind and break
down the organized eddy patterns which prevail in the wakes of solid
barriers. This sifting effect reduces wake pressures, giving a more balanced
pressure distribution on windward houses.
The suggested separation of trees from house is of particular interest. A
gap of 2.5 house heights corresponds to an angle of about 22° and so
means that on a clear day at noon in December there will be some over-
shadowing of the ground floor and so some loss of direct passive solar gain.
There will be, of course, a continuous obstruction of a small amount of
diffuse radiation over the course of the heating season - the effect should
Fig. 2.13. Reduction of wind Free wind speed (m/s)
byashelterbelt.[35]
30% - 17.6
(Courtesy of G.E. Sheard.)
13.2
8.8
- 20% -
10% -
10% 20% 30% 40%

Reduction in wind by shelter
Fig. 2.14. The effect of Measurements 3.7 m above level ground

shelter on greenhouse fuel
consumption (measurements + 1 0 % ) ^ Belt is 4.6 m high — moderately permeable
3.7 m above ground level).
(Courtesy of G.E. Sheard.)
Wind
- 20% iK J \ l\
-10% I
V^-30%' ^
^ - _ _ - 20%'
v\ _- 10%-^^
0 Up to 30% reduction
+ 10% in wind over glasshouse
^ in this position
46 46 92 184
metres
not be very great but it may nevertheless be worthwhile to quantify the

loss for a specific site. It should be noted too that a shelterbelt at even 1.5
house heights separation affords little shading in the summer and if over-
heating is a problem another means of protection must be sought.
Greenhouse researchers have also been interested in the use of natural
or artificial shelters to reduce windspeed. Sheardl35) cites a rule-of-thumb
that windbreaks should have a minimum height equal to half the height of
the ridges of the glasshouses and be sited about ten times their height away
from the glasshouses. He gives a diagram, reproduced as Fig. 2.13, showing
the reduction of wind by a shelterbelt, as well as an estimate of the
resultant fuel savings (Fig. 2.14). As for homes, the loss of passive solar
Wind 29
Fig. 2.15. Wind diagram for
Malmslatt, Sweden.l 9 ! (The
diagram gives the relative
frequency (%) of winds from
NE varying directions - each
ring represents 5%. Obser-
vations were over a 24 h
period for 12 months and
were made during all
weathers. The solid line is
for wind speeds of 1 - 9 9
knots (17 knots = 8 m/s)
and the broken line for wind
speeds of 1 7 - 9 9 knots.
Thus, for example, approxi-
mately 7% of the time,
winds in the range 1 - 9 9
knots come from the direc-
tion of due north to 15°
east of north.
SE
1975 AUG SEP Fig. 2.16. Typical monthly

(in 1975) wind roses for
NW Malmslatt, Sweden (mean
wind speed in m/s)J 9 l
WPA
gain must be remembered when evaluating the fuel savings. (It is also
possible that the crop will be of lower value due to lack of light for photo
synthesis and thus slower growth.)
To aid the siting of windbreaks for any particular application, wind
diagrams are used. Fig. 2.15 shows the relative frequency of wind direction
in Malmslatt, Sweden, used in describing the ambient environment of a
solar house.l9^ Monthly wind roses (see Fig. 2.16), if available, are ideal
because they allow design for a specific heating season. Wind roses for a
number of locations may be found in the Chartered Institution of Building
Services (CIBS) Guide.
Wind speed and direction can only be obtained accurately by measure-
ment but often, on first examining a site, a very rough quantitative idea of
wind conditions is adequate and useful. For this reason we include here
the Beaufort scale (Table 2.2).
At least one UK project has investigated and employed shelterbelts to
reduce heat losses. The development at Basildon, mentioned previously,
incorporated shelterbelts (see Fig. 2.17) which run predominantly north-
Table 22. The Beaufort
Beaufort Wind speed

number Description (m/s)
0 Calm; smoke rises vertically 0.3

1 Direction of wind shown by smoke drift but
not by wind vanes 0.3-1.5
2 Wind felt on face; leaves rustle; ordinary vane
moved by wind 1.6—3.3
3 Leaves and small twigs in constant motion;
wind extends light flag 3.4—5.4
4 Raises dust and loose paper; small branches
are moved 5.5-7.9
5 Small trees in leaf begin to sway; crested
wavelets form on inland waters 8.0—10.7
6 Large branches in motion; whistling heard in
telegraph wires; umbrellas used with difficulty 10.8—13.8
7 Whole trees in motion; inconvenience felt
when walking against wind 13.9—17.1
Fig. 2.17. Shelterbelt

pattern and recommended
spacing.! 23 J (Courtesy of
Ahxends, Burton & Koralek.)
Section through planting
south (thus effectively countering the prevailing wind), while the housing
runs east-west (see Fig. 2.18), so conflict between passive solar gain and
vegetation for shelter is fortunately minimized. Ash, grey alder, red oak
and wild cherry are among the principal trees in the windbreaks and the
mature height is expected, on average, to be about 25—30 m. With the
trees in leaf the estimated porosity is 30%J 3? l
Lest it be thought that consideration of solar energy and wind patterns
on site is relatively new, consider Henry Niccolls Wright's 'heliothermic'
site planning for the American suburbs in the 1930sJ 38 l Fig. 2.19 shows
Wind 31
Fig. 2.18. Site plan for

Felmore Housing at
Basildon.l 22 ! (Courtesy of
his plans for a development designed to be cool in summer and warm in Ahrends, Burton & Koralek.)
winter. The plan of the houses, their orientation and location on site all
contribute to using winter insolation to best advantage and to reducing
summer insolation. Similarly, winter winds are minimized to reduce heat
losses and summer breezes maximized for cooling.
Now let us examine the case where a windmill is to be located on-site
to provide either electricity or space heating. Chapter 8 deals with wind-
mills in detail but, for the moment, we must introduce some idea of the
approximate size of the structure under consideration. In a relatively
sheltered area such as Kew, London, a wind turbine with a swept area of
24 m2 (4 m blades on a 6 m cross-member, see Fig. 8.18) is likely to pro-
vide (at the alternator output) approximately 15.8 GJ/y and so could
make a significant contribution to the space heating load or could meet
the entire electrical demand of a home (excluding the domestic hot-water
demand).!39 !
Normal practice would place such a wind turbine on a tower at least
10 m high to take advantage of the increased wind speed (see Chapter 8).
As for siting, precise information is unfortunately scarce. Manufac-
turers' guidelines, which tend to be based on experience, are often quali-
tative exhortations, for example to avoid areas where obstructions can
either prevent the wind from actually reaching the turbine or where they
create turbulence which decreases the efficiency of the machine and
increases vibration. Most manufacturers of small- to medium-sized turbines
(say 1-5 m in diameter) suggest towers of at least 5-6 m (Dunlite mills
were sold with 12—18 m towers, and Goldingl 29' recommends at least
Fig. 2.19. Heliothermic site "Heliothermic" site
planning.! 3 8 1 planning, from Henry
Niccols Wright, 'Solar
Radiation as Related to
Summer Cooling and
Winter Radiation in
Residences', New York
1936
composite
daily
shadow:
winter
solstice
broadside
V " to greatest
winter
insolation
^ narrow end
" to prevai I ing
winter wind
) \ \
9-12 m) and, furthermore, add that the centre of the turbine should be
4—5 m above buildings and clear of vegetation within 200-300 m.
It is obviously quite difficult in many situations to meet such con-
straints and it is to be regretted that, at present, the loss of available
energy as a function of distance of a small windmill from a configuration
of obstructions is not known. One step towards relaxation is to site the
turbine where it at least has access to the dominant winds, using a wind
rose for this purpose, and to estimate the loss of energy.
Sheltering a house from the wind and siting a windmill which will be
Soil 33
used to supply energy to the house will thus pose a complex problem. If
transmission losses from the turbine to the energy store or point-of-use
were negligible the situation would be simplified but, as Chapter 8
explains, this is not necessarily so. As Fig. 2.12(c) and (d) suggests, if a
shelterbelt, be it vegetation or other houses, is used, a wind turbine must
be sited either at a considerable distance from it or on a tower sufficiently
high to avoid the effects of turbulence.
One possibility that was considered by the Cambridge Autarkic Housing
Project, as well as other groups, was to site the wind turbine on the roof of
the house. This has a number of disadvantages: Goldingl29! cites the
possibility of the usually turbulent wind flow over the roof producing
enough thrust on the turbine to damage the roof in high winds, but there
may also be adverse reactions from the occupants due to noise and
vibration. The Autarkic Housing Project viewed the mounting of a turbine
on the roof as a research project and hoped that by using the house struc-
ture to support the turbine, the cost of the tower could be avoided; the
savings were expected to outweigh the added costs incurred in strengthen-
ing the house structure to support the turbine. Precise expenses were not
determined but a preliminary cost-analysis indicated that the cost of
incorporating the tower would in fact be quite high. This is not surprising
given the size of the wind turbine compared to the house as shown in
proper scale in Fig. 12.3.
2.4 Soil
Aside from the traditional considerations of foundations, thermal energy
storage and the use of the soil to reduce heat losses are the two principal
points of interest when evaluating the soil of a site for a low-energy build-
ing. The effect of soil type on the choice of a waste-disposal system, which
may also be important, is covered in Chapter 10. Soil as a building material
is unlikely to play a major role in low-energy European homes of the near
future and is not discussed here. (For those who wish to pursue the sub-
ject, a number of people, especially architects interested in the use of
natural materials, have looked at the possibilities available.)140> 41> 421
The pros and cons of aboveground and underground storage of energy
are presented in Chapter 7. In a variety of situations, for reasons of space
especially, underground storage will be the sole viable solution and if the
water table is low this need not be more than a routine difficulty.
However, in much of the UK and many areas of northern Europe the
water table is quite high. Fig. 2.20, developed for agricultural purposes,
shows how large an area is seasonally or permanently wet as judged by
examining the top 40 cm of soiU43! - a deeper sampling depth, as would
be needed for underground storage of energy, would of course increase the
wet acreage.
For interseasonal storage of 10-20 GJ solar energy for a house, it
would not be unusual to have to excavate, say, 100 m3 to allow for both a
water store and its insulation.l 44! A concrete basement is the likely struc-
tural solution (see Chapter 7). It will have to resist the lateral pressure of
g. 2.20. Types of soil
iter regime in southeast
lgland. After ( 42 1
Soil Survey of England and Wales

Types of soil water regime
] Soils with permeable substrata,remote from groundwater. Rarely or seldom wet.
L'l'lvJ Soils affected by groundwater, with generally good permeability.
$/fty/A Soils affected by groundwater, with slow or variable permeability. Seasonally or
permanently wet
ioils with impermeable substrata, in drier lowlands with
significant moisture deficit in most years.
50 km
I I Urban areas
the earth and, additionally, if the basement is submerged partly or wholly

in water, the hydrostatic pressure of the water. For dry soils the lateral
pressure is about 4.7 Pa/m of depth; the hydrostatic pressure of water is
9.8 Pa/m of depth.I451 An example of the loading for a basement at a
depth of 3 m with a water table at 1 m is shown in Fig. 2.21.
If, as is likely, water is used as the storage material in the heat store in
the basement, its weight cannot be relied upon to keep the structure from
rising majestically out of the earth. This is because allowance must be
made for the basement being empty initially, during inspections and in the
event of leaks. Furthermore, a high water table may mean that even with
backfilling, cohesion between the soil and the basement walls is negligible
Soil 35
Fig.
1 2.21. Basement loading.
1 ;: 1 mt '
Basement •
3m
1
i 4 4 444444J 14.1 kPa 19.6 kPa 33.7 k Pa

19.6 k Pa Earth Hydrostatic Combined
pressure pressure pressure
and so may not contribute to resisting flotation. The costly consequences

are likely to be a need for additional concrete in the form of a 'lip' for the
basement slab or thickened walls and slab, or some combination of these.
Finally, the tendency of the clay soils common to England to shrink
and expand with changes in moisture content (which affects the siting of
vegetation as mentioned earlier) can result in differential movements and
lead to cracking. If moisture were able to penetrate and soak the thermal
insulation in the store the effect on the thermal performance would be
catastrophic. Consultation with a structural engineer is recommended to
determine the appropriate foundation to minimize such risks.
More optimistically, but in more favourable circumstances, a number of
full-scale experiments are underway to evaluate storage of solar energy in
the soil itself. In France, near Grenoble, at a site with a water table 9 m
below the surface, heat from solar collectors is being stored in a clay
soil J 46 1 The boundary sides of the store are insulated and on top the soil
is covered with a l 5 m X 10m slab of 20 cm thick polyurethane. On
Prince Edward Island, Canada, water storage is expensive and rocks are
scarce but an abundant supply of clay has encouraged work on the use of
wet dirt or 'mud' as a storage medium.!47!
Soil may be used to reduce the heat loss from a building by delving to
an extreme and locating the structure in the ground or by creating a berm
around the ground-floor storey. Sterling and his colleagues^33! have
covered the topic in depth. The principal advantage claimed is substantial
energy savings due to reduced conduction and ventilation heat losses. Fig.
2.22 shows the soil temperature compared with the air temperature at
Cambridge (UK). At greater depths the amplitude of the mean tempera-
ture fluctuation decreases. Sterling^33! notes that in the Minneapolis-St
Paul area, at 5—8 m the temperature is almost constant at 10 °C.
The relatively high soil temperatures during the heating season afford
an opportunity to reduce the heat losses, and sheltering the structure with
the earth both minimizes the effect of the wind and provides insulation.
Fig. 2.23 shows the Ecology House by architect John Barnard^50! at
Osterville, Massachusetts, which is completely underground and features
an internal courtyard; solar collectors are combined with a forced-air
furnace for heating. A reduction of 75% in the energy demand for heating
and cooling has been claimed.
A subsequent development, likely to be more practical in northern
Fig. 2.22. Average tempera- 18 Air temperature
tures at Cambridge.!48. 49 ]
16
14 Soil temperature at 1 m
12
oc10
8
6
4
2
0
Jan Apr Jul Oct Dec
Europe (by the same architect), combines an earth surround and covering
with an exposed south facade to take advantage of passive solar gain. Fig.
2.24 shows the Ecology House in Stow, Massachusetts. Heating is by
forced air with a heat pump and energy use is 50-60% of that of a normal
house of the same size.
Additional loading due to earth, groundwater and frost; avoidance of
moisture penetration (from surface run-off, a high water table, during
heavy rainfalls, etc.) and provision of adequate ventilation and light are
some of the factors that must be evaluated carefully before opting for an
underground or semiunderground structure. (Metz^51! provides a useful
discussion of some of the severe problems that arise from the need to
waterproof the underground structure.) That these potential difficulties
can be overcome is being illustrated in the US, but for northern Europe, at
least for the time being, there is insufficient information on such buildings
to recommend their use in other than exceptional circumstances.
An effective partial measure which may easily be taken is the use of
built-up earth berms. These avoid, in particular, the problems associated
with a high water table, while lowering the heat loss due to conduction
and reducing air infiltration; the energy savings resulting will, of course,
depend on the amount of wall area covered, among other factors. In
houses incorporating underground thermal storage, berms may be the most
suitable and inexpensive way of using the excavated soil. It must be
remembered, however, that berms are not a panacea. Earth is not a very
good insulator and the soil temperature will normally be below that of the
house so there is always a case for including insulation between the soil
and the house. This is particularly true when the berm is exposed to the
weather.
An additional consideration when examining the soil of a site is its
potential for producing a source of fuel. In most soils of northern Europe
there should be little difficulty in growing some variety of tree which, as
wood, could contribute somewhat to the fuel supply, as well as providing
a windbreak. For example, by coppicing certain species of hardwoods
(such as sycamore, poplar, alder), it is possible to get higher yields than by
traditional methods. The UK Department of Energy!52) cites yields in
Soil 37
Fig. 2.23. Ecology House
at Osterville, Massachu-
Fireplace Kitchen setts. [33] (Courtesy of
rn Dining J.E. Barnard Jr.)
Floor plan
Living room Atrium Bedroom
Section
Fig. 2.24. Ecology House
at Stow, Massachusetts.!321
(Courtesy of J.E. Barnard
Jr.) Study
Glass Bedroom Mechanical

Bedroom 2 and |
1 roof equipmen
atrium room / I
Garage
Livin
itchen® 9 Bedroom
room
Dining area 3
Entrance
Floor plan
Glass covered Study

atrium
Section
References 39
2 2
Pennsylvania of up to 1 kg/(m y), or approximately 18 MJ/(m y) (see
Table 6.5). If we use a combustion efficiency of 65%, an area of 50 m2
would provide about 1.6% of an annual space heating load of 36 GJ.
References
[ 1 ] Lynch, K. (1971). Site Planning. Cambridge, Mass.: MIT.
[2] McHarg, I. (1971). Design with Nature. New York: Doubleday.
[3] Olgyay, V. (1963). Design with Climate. Princeton: University
Press.
[4] Givoni, B. (1976). Man, Climate and Architecture. London:
Applied Science.
[ 5 ] van Straaten, J. (1967). Thermal Performance of Buildings.
Barking: Elsevier.
[6] (1979). Built Environment, 5 (4).
[7] Owens, S.E. (1979). 'Energy and settlement patterns'. Built
Environment, 5 (4), 282-6.
[8] Lacy, R.E. (1972). Survey of Meteorological Information for
Architecture and Building. CP 5/72. Garston: Building Research
Station.
[9] Ahlstrom, B., Hedeman, E. & Dattermark, B. (1977). 'Solar energy
house in Linkoping'. Document D9. Swedish Council for Building
Research.
[10] Kovach, E.G. (ed.) (1973). Technology of Efficient Energy Utiliz-
ation. Brussels: NATO.
[11] Siviour, J.B. (1974). The Effects of Weather on House Heating
Requirements — An Interim Report on Unoccupied Houses.
ECRC/M710. Capenhurst, Chester: Electricity Council.
[12] (1971). Sunlight and Daylight Planning Criteria and Design of
Building. Department of the Environment Welsh Office. London:
HMSO.
[13] CIBS Guide (1975). Section A6, 'Solar data'. London: CIBS.
[14] Page, J.K. (1978). 'Methods for the estimation of solar energy on
vertical and inclined surfaces'. International Report BS46. Depart-
ment of Building Science, University of Sheffield.
[15] (1979). 'Meteorology for solar energy applications'. UK — ISES
Conference C18.
[16] Solar Site Selector, Lewis and Associates, Grass Valley, CA.
[17] Petherbridge, P. (1965). Sunpath Diagrams and Overlays for Solar
Heat Gain Calculation. BRE CP 39. Garston, Watford: BRS.
[18] Turrent, D., Doggart, J. & Ferraro, R. (1980). Passive Solar
Housing in the UK. London: Energy Conscious Design.
[19] Seed, J.L. (1982). 'Milton Keynes Development Corporation'.
Private communication.
[20] Berry, J., Emerson, R., Harrison, J.W. & Kasabov, G. (1977).
'Conservation of energy in housing'. Building Services Engineer,
2(45), 288-93.
[21 ] Anon. (1978). 'Energy conscious housing layout'. RIBA Journal,
85(6), 229-30.
[22] Hermsen, J., Nelson, G. & Witham, D. (1979). 'Energy-conscious
housing at Basildon, Essex', AJ, 41 (170), 758-72.
[23] Anon. (1975). Northlands-Felmore Report. Northlands-Felmore
Design Team.
[24] 6 Cathain, C. & Jessop, M. (1978). 'Density and block spacing for
passive solar housing'. Transactions of the Martin Centre, 3, 137—
63.
[25] Holzberlein, T.M. (1979). 'Don't let the trees make a monkey of
you'. Proceedings of the Fourth National Passive Solar Conference.
Newark, Delaware: ISES — American Section.
[26] Marlowe, O.C. (1977). Outdoor Design. London: Crosby, Lock-
wood and Staples.
[27] Flora, T. (1977). Trees and building foundations'. In: Trees and
Buildings, Hall, T.H.R. (ed.). Westbury, Wiltshire: Arboricultural
Association.
[28] Biddle, P.G. (1977). Tree-root damage to buildings - an arbori-
culturalist's experience'. In: Trees and Buildings, Hall, T.H.R. (ed.).
Westbury, Wiltshire: Arboricultural Association.
[29] Golding, E.W. (1977). The Generation of Electricity by Wind
Power. London: Spon.
[30] Anon. (1970). The Assessment of Wind Loads. BRE Digest 119.
Garston, Watford: BRE.
[31] Arens, E.A. & Williams, P.B. The effect of wind on energy con-
sumption in buildings'. Energy and Buildings, 1, 77—84.
[32] Mattingly, G.E. & Peters, E.F. (1977). 'Wind and trees: air infil-
tration effects on energy in housing'. Journal of Industrial Aero-
dynamics, 2, 1 — 19.
[33] Sterling, R. (undated). Earth Sheltered Housing Design. University
of Minnesota: The Underground Space Center.
[34] Woodruff, N. (1954). 'Shelterbelt and surface barrier effects on
wind velocities, evaporation, house heating and snow drifting'.
Technical Bulletin No. 77. Agricultural and Engineering Station,
Kansas State College of Agriculture and Applied Science.
[35] Sheard, G.F. (1975). The Effects of Wind on Glasshouse Pro-
duction. Littlehampton Glasshouse Crops Research Institute.
[36] Anon. (1939). The Meteorological Glossary. Littlehampton.
London: HMSO.
[37] Price, C, Ahrends, Burton & Koralek. (1981). Private communi-
cation.
[38] Wright, H.N. (1936). Solar Radiation as Related to Summer Cool-
ing and Winter Radiation in Residences. New York: J.B. Pierce
Foundation, 40 W 40th St, NYC.
[39] Littler, J.G.F. & Thomas, R.B. (1979). 'Wind power for domestic
use in the United Kingdom'. Proceedings IEE International Con-
ference on Future Energy Concepts. London.
[40] Vale, B. (1973). 'A review of the Ministry of Agriculture's earth
houses'. Autonomous Housing Study Working Paper 17. University
of Cambridge, Department of Architecture.
[41] Weeks, C.T. (1978). 'Method in madness'. Building Design, p. 19.
[42] Fitzmaurice, R. (1958). Manual on Stabilized Soil Construction for
Housing. New York: UN Technical Assistance Programme.
[43] Thomasson, A.J. (1975). Soils and Field Drainage. Soil Survey
Technical Monograph No. 7. Harpenden: Soil Survey.
[44] Littler, J.G.F. & Thomas, R.B. (1980). Thermal storage in the
References 41
Autarkic House'. University of Cambridge, Transactions of the
Martin Centre, 4, 139-56.
[45 ] Salvadori, M. & Levy, M. (1976). Structural Design in Architecture.
New York: Prentice-Hall.
[46] de la Casiniere, A., Person, J.-P. & Vachaud, G. (1980). 'Field
experimentation of the "soil therm" interseasonal storage system
of solar energy in the subsoil application to bio-industries'. In:
Energy Conservation and Use of Renewable Energies in the Bio-
Industries. Vogt, F. (ed.). Oxford: Pergamon.
[47] Caffell, A. & MacKay, K.T. (1980). 'Mud storage: a new concept in
greenhouse heat storage'. In: Energy Conservation and Use of
Renewable Energies in the Bio-Industries. Vogt, F. (ed.). Oxford:
Pergamon.
[48 ] Anon. (1976). A verages of Temperatures for the United Kingdom.
Meteorological Office, London: HMSO.
[49] Data from the University Botanic Garden, Cambridge.
[50] Barnard, J.E., AIA, Marston Mills, Mass.
[51] Metz, D. (1982). 'Keeping dry underground'. Solar Age, 7 (5),
24-31.
the United Kingdom'. Department of Energy. Energy Paper No.
16. London: HMSO.
Building design
3.1 Introduction
Le Corbusierl11 said that a house was a machine for living in - in this
chapter we attempt to describe some of the machine's components. Unfor-
tunately, the state of the art is such that, for a given set of inputs, say
environmental and material, one isn't quite certain how the machine or
building will perform. The heart of our present dilemma is that we cannot
precisely say even how individual building components respond in situ-
ations outside laboratories. To take but one example, infiltration losses are
difficult to determine because of the effect of workmanship, condition of
seals, extent of exposure to wind and so forth. And, remembering the old
adage: 'The best laid plans of mice and men . . .', if we consider that, even
in a well-designed low-energy house, the occupants will undoubtedly use
the house in ways that would amaze the designers, the description of a
building's performance is a hazardous task indeed. What is even more
difficult though is to make accurate recommendations about how to
design different, more-energy-efficient buildings. Of course there are the
banal (but worthwhile) solutions such as increasing the insulation level and
decreasing the ventilation rate but, as we will see, the latter can only be
done to a certain point which is well short of the one where occupants and
designer breathe uneasily. For more-complex questions such as choosing
between lightweight and heavyweight buildings our information is imper-
fect. And as for other areas which may become the technology of
tomorrow, evacuated walls, say, we know almost nothing.
We start with a brief introduction to some of the broader physical
principles of design, such as materials and environmental conditions,
before examining the detailed questions of form, layout and construc-
tional elements. The reader will note that we have ignored some aspects
entirely. For example, we have left aesthetics to designers and clients but,
or perhaps, since, we feel that there need be no conflict between low-
energy buildings and attractive ones.
3.2 Energy demand and thermal response

Fig. 3.1 shows the components of the seasonal heat loss in a typical house
built to the standard of the 1982 Building Regulations^2' 3J ((/-values
(thermal transmittances) in W/(m2 K): windows 5.6; walls 0.6; floor 0.6
42
Energy demand and thermal response 43
Fig. 3.1. Percentage com-
Windows
position of heat loss in a
Ventilation typical semidetached house
buHt to 1982 Building
Regulations Standards.
Roof
External walls
Ground floor
(assumed); roof 0.35; ventilation rate = 1 air change/h = 225 m 3 /h

(assumed); average internal to external temperature difference (assumed)
7 K; heating season of 5544 duration (assumed)).
Such a house has an estimated annual gross space heating requirement
of 36.3 GJ; Siviour suggests that this would be reduced to about 13 GJ
(assuming 70% overall utilization of 'free heat' from occupancy, solar heat-
ing and losses from the hot-water system).l4l (In related work, Siviour has
shown that due to the 'free heat' from the sun, typical houses function as
solar collectors with efficiencies of 6-7% for space heating.)!51
Siviour noted that problems arise in using 100% of the free heat
because of the variability of input (some of course occurs when there is no
space heating demand) and its association with steam and smells. His esti-
mate of 70% utilization followed his detailed analysis shown in Table 3.1.
Table 3.2 shows the heat emission from the adult male human body —
for women and children correction factors of 85 and 75%, respectively,
may be used.
Siviour has compared a number of published estimates of internal free-
heat gain. Table 3.3 summarizes his results.
The CIBS estimates that the radiation heat gain during the heating
season per square metre of glass will be approximately as follows: south-
facing window, 670 MJ; east- and west-facing windows, 400 MJ; north-
facing window 240 MJ.
Obviously, utilization of free heat will vary from building to building.
One of the objects of energy efficient design is to use as much of it as is
feasible and economical. (Appendix 2 discusses how the design peak and
annual heating requirements may be calculated.)
An as yet unresolved point but one of some interest is whether light-
weight or heavyweight buildings use energy more efficiently. The terms
themselves are relative. The old Institute of Heating and Ventilating
Engineers (IHVE) Guidef81 described as heavyweight, buildings with a
construction of solid internal walls and partitions and solid floors and
ceilings. Lightweight buildings were characterized by lightweight demount-
able partitions with suspended ceilings and floors which were either solid
with a carpet or wood block finish or were of the suspended type. For
houses it may be convenient to think of a heavyweight house as one with
a double wall of, say, brick on the outside and concrete block inside and a
lightweight house to be of single-brick construction with, say, expanded
44 Building design
Table 3.1. Analysis of free-heat gai
Degree of
Source availability Comments
(1) Occupancy
a. People Good Heat from people is where it is required
b. Appliances Fair Heat is sometimes where required, e.g.
television
c. Lights Good Heat is where required
d. Cooking Poor Concentrated in time and space and
accompanied by steam and smells
e. Drawing curtains Good Heat saving can be throughout the
house
(2) Domestic hot water
a. Storage vessel Fair Concentrated but continuous
b. Pipe losses Fair Dispersed but intermittent
c. Point of use Poor Concentrated and accompanied by
moisture
(3) Solar heating Fair Varies from day to day; during the day
from room to room
Table 3.2. Heat emission from the adult male human body at 20°Cdry-
bulb temperature^
Heat emission (W)

Degree of activity Sensible Latent Total
Seated at rest 90 25 115
Light work 100 40 140
Walking slowly 110 50 160
Light bench work 130 105 235
Medium work 140 125 265
Heavy work 190 250 440
polystyrene adjacent to the brick or to be of timber-frame construction

(with insulation between the studs).!91
The most recent CIBS Guidel 10! provides a more rigorous interpret-
ation of heavyweight and lightweight in terms of the response factor given
by
2(AY) + (NV/3)
/r
Z (AU) + (NV/3)
where
A = area of a structural element (m 2 );
Y = admittance of a structural element (see Appendix 2)
(W/(m2 K));
TV = number of room air changes per hour;
V — room volume (m 3 );
U= thermal transmittance of a structural element (W/(m2 K)).
A thermally lightweight building will have a response factor of about 2.5
and a heavyweight building one of about 6.
Energy demand and thermal response 45
Table 3.3. Assessments of internal free heat gains in
domestic premises during the winter^11
Source MJ/day
Occupants5 14.4_21.6
Lighting, appliances, cooking 29.5-63.3
Hot water* 12.2-54.0
a
Family size varies from 2.9 to 4; note that only sensible-
heat production should be accounted for.
b
Assumed to be 50% of total energy input to water heat-
ing.
Table 3.4. Specific heats of selected materials^11]
Specific heat Specific heat

Item (J/(kgK)) Item (J/(kg K))
Granite 330 Polystyrene 1250

Copper 376 Standard hardboard 1250
Mild steel 502 Insulating fibreboard 1400
Glass 830 Perspex 1460
Concrete 880-1040 Timber 1500
Aluminium 920 Polythene 2300
PVC 1040 Water 4187
Heavyweight buildings will have high thermal capacities. Thermal

capacity is the ability of a body to store heat and is given by the product
of its mass and its specific heat. Specific heat is the quantity of heat
required to raise a unit mass of a material through a unit degree of tem-
perature. Table 3.4 shows the specific heat of some common materials (see
also Table A 2.3 of Appendix 2).
In steady-state conditions the same thermal transmittance, or U-value,
can be provided by a thick heavyweight wall, say that of a stone castle, or
a thin lightweight wall, say a timber-frame house with insulation. However,
when sudden temperature changes occur, as they do continually, a thick
wall of high thermal capacity will only warm up, or cool down, slowly; the
converse is true for thin lightweight walls. This response will affect the
energy consumption of buildings.
For large buildings, Burberry^12 J has stated that, if there is any inter-
mittency of use or if close control of temperature is appropriate, light-
weight structures offer better economy of heating. (On the other hand,
heavyweight construction, in part because it resists summertime peak
temperatures better, may provide a better economy of air condition-
ing.)! 13]
For small buildings such as houses, opinion is divided on the relative
merits of lightweight, rapid-response structures and heavyweight ones with
longer response times. One important point to consider is the occupancy
pattern. Although this is often difficult to predict and the house may, in
46 Building design
fact, be designed to accommodate a variety of patterns during its life, if it
is known that the house will be continuously occupied, for example
housing for the elderly as opposed to that for young, single people, a
heavyweight structure is likely to offer advantages. Where use is inter-
mittent a lightweight structure which warms up quickly will tend to save
energy. Another consideration is the importance of passive solar gain in
the design but even here opinion varies. At Basildon, in the estate
described in Chapter 2, the houses were designed to take advantage of
passive solar gain by attention to orientation and larger glazed areas on the
south walls. The design study recommended a lightweight structure
because of cheaper installation costs and, with an intermittent heating
regime, lower heating costs.! 141 The subsequent report to the client, the
Basildon Development Corporation, incorporated the study and went on
to discuss the importance of heavyweight solid party walls in terraced
housing. Because of such walls, the difference between heavyweight and
lightweight external constructions is less distinct. Similarly, such
unknowns as the occupants' choice of furnishings will affect thermal
response. Heavy sofas and chairs may store a fair amount of energy, thus
tending to make the house more heavyweight. On the other hand, a carpet
over a solid floor will reduce the response time, just as tapestries on
mediaeval walls did. The final construction, shown in Fig. 3.21, was a
compromise with cavity brickwork on the ground floor and timber-frame
construction on the first.
Another hybrid approach was used in a development of 29 two- and
three-person solar heated flats in the London Borough of LewishamJ15 !
There, a heavyweight floor with embedded coils (see Chapter 6) to provide
background heating was used in conjunction with highly insulated walls
which also had a low admittance. The construction technique was dry
lining on plaster dabs, 150 mm aerated concrete blocks, 50 mm cavity and
115 mm brick, giving a (/-value of 0.76 W/(m2 K) and an admittance of
2.0 W/(m2 K) (compared to, say, 4 for a more traditional construc-
tion).l 16l
(In addition to the greater insulation afforded by 150 mm blocks, they
also add strength to the construction and help overcome the problem,
often encountered with 100 mm lightweight blocks, of their being too
weak to carry the structure's weight.)
An example of heavyweight construction deliberately used in conjunc-
tion with passive solar gain may be found in a group of 14 houses for the
elderly in Bebington (see also Chapter 4)J 1 7 ' 181 Occupancy there is con-
tinuous and to increase the effect of the thermal capacity of the structure
the heavyweight walls are insulated externally. The principle may be
extended even further and special storage elements incorporated in the
building to 'soak up' excess solar energy for later use - examples are given
in Chapter 4. The approach depends on whether the designer thinks there
will be solar energy available for storage and whether his or her design can
both take advantage of such heat and avoid overheating. If overheating
occurs, for example when the building fabric and furnishings cannot
absorb enough energy and if the heating system cannot be turned down
The internal environment 47
quickly enough during a period of intense passive solar gain, there is no
energy saving but there will be discomfort. The interrelations among types
of construction, varying casual gains and heating systems are not well
understood at the moment and this is reflected in the variety of systems
being tried. It may turn out that the differences in energy consumption
among the various solutions are not great but more research will be
required to establish this. We may also see a range of optimum solutions,
even within one country, depending on the weather conditions, notably
incident solar radiation and temperature.
Theoretical work, based on the admittance method (see Appendix 2)
by British Gas,l19l has found that, for an insulated terrace house, a heavy
structure requires a larger peak heat input than a light structure of the
same level of insulation and that the slower overnight cooling of a heavy
building means that its average internal temperature is higher than that of
a light building. The consequence of the second point is that the heat
losses over a 24 h period will be higher and to balance these the heavy-
weight building will require a larger total heat input.
3.3 The internal environment

3.3.1 Comfort
The philosopher Karl Popper has suggested that it is much easier to
work towards a society that minimizes unhappiness rather than one that
maximizes happiness and much the same might be said of comfort. We
have a fairly good idea of what makes most people uncomfortable but the
provision of an ideal environment for all of the people all of the time has
proven to be an illusive, and perhaps futile, goal.
Comfort is a subjective experience, dependent on individuals and such
factors as people's past experience of heating and whether or not they
themselves are to pay the forthcoming heating bill. In practice a designer
tends to provide conditions that are acceptable to a majority of the per-
sons occupying a space or building and not so unacceptable to the others
that complaints are too vociferous.
Comfort conditions depend mainly on the air temperature, the mean
radiant temperature of the surroundings, the air speed and the relative
humidity. It is interesting to keep in mind the similarity of the body's
temperature control system to mechanical control systems for heating and
air-conditioning installations. The important temperature-sensing system in
the hypothalamus at the base of the brain operates like a thermostat,
monitoring changes in blood temperature caused by the body's internal
metabolic changes and by temperature gradients across the skin. The sys-
tem has a set point of about 37.0 °C which it tries to maintain. If the body
temperature is sensed at less than this value, physiological responses such
as shivering occur to increase the metabolic rate so that more heat is
generated. If the contrary occurs, the body starts to sweat and the subse-
quent evaporation of this moisture from the skin provides cooling.
Between the extremes of shivering and sweating the body controls its
48 Building design
temperature by altering the blood flow to the surface. The goal of these
responses is the maintenance of thermal equilibrium.
Seated at rest, adult males produce approximately 115 W (Table 3.2)
of which about 80% is lost from the body by convection and radiation
(and to a much lesser degree conduction) and some 20% by evaporation
(mainly from the lungs but also from the skin). Thermal comfort depends
upon achieving a balance between the heat being produced by the body
and the loss of heat to the surroundings. Since the body is at a higher
temperature than the surroundings and since we sense skin temperature
rather than room temperature, the function of heating and air-conditioning
systems is in fact to control personal cooling.
The principal relationships between physical factors of the environment
and heat loss from the body are the following:
(1) The air temperature affects heat loss by convection and evaporation.
(2) The mean radiant temperature affects heat loss by radiation.
(3) The air speed affects heat loss by convection and evaporation.
(4) The vapour pressure of the air affects heat loss by evaporation.
In the 1950s and 1960s, in the midst of the 'building boom', interest in
comfort was related to office environments, the effect of large glazed sur-
faces, overheating and air conditioning. There was a search for uniformity.
Now it might be said that, with increasing concern about energy consump-
tion and the recognition that passive solar gain can usefully reduce the
space heating demand of buildings, the ethos of comfort research must
also change. The importance of this is readily apparent if one considers
that a 1 K lowering of the internal temperature reduces the energy
required for heating by approximately 5—10%.l2°l One encouraging sign
in comfort research is Humphreys' suggestion, which admittedly is not
shared by all, that since the indoor comfort temperature depends on the
outdoor mean temperature the internal temperature be allowed to follow
the external and that continuous change at a rate of 2 K per week would
be virtually unnoticeableJ 21 J Another is work on comfort in passive solar
heated buildings where the thermal environment is likely to be less uni-
form. l 22 l We may also expect to see studies of the effects on comfort of
the lower air-change rates so often used to conserve energy.
3.3.2 Temperature
The temperature index presently used by the CIBSl23! is the dry
resultant temperature (there is also a wet resultant temperature which
takes into account humidity) given by the formula
t -'r
res
I
where
t^ = inside air temperature;
tv = mean radiant temperature;
rres = dry resultant temperature; and
v — indoor air speed.
23
Table 3.5. UK design conditions' ^
Resultant Relative
Occupancy/ temperature humidity
Season category (°C) (%)
Summer Continuous 20-22 50
Transient 23 50
Winter Continuous 19-20 50
Transient 16-18 50
At 'still' indoor air speeds of v — 0.1 m/s the formula is simply

'res = V<U + 'ai).
The dry resultant temperature, more commonly abbreviated to the result-
ant temperature, is that temperature recorded by a thermometer at the
centre of a blackened globe 100 mm in diameter.
The formula shows that both the air temperature and the radiant con-
ditions are important for comfort and this must be kept in mind when
selecting a heating system. Ideally, all surfaces should be at the same tem-
perature but, of course, glazed areas will tend to have a lower temperature
during the winter and a higher one in summer, especially if solar absorbing
glass is used. (One of the advantages of double glazing is to improve com-
fort conditions by raising the mean radiant temperature of the internal
surfaces.) In the past it was at times suggested that a higher mean radiant
temperature than air temperature was preferred but research has shown
that in fact this is not so.l 24 !
The temperature gradient, which depends on both the heating and
ventilating system and the thermal insulation of the space, is also of
importance for comfort. A guideline is that in a room at 22 °C the tem-
perature at ankle level should not be less than 19-20 °C and the head-
level temperature should not be more than 3 °C higher than the ankle-
level temperature.!24 ' 251 Floor-heating systems tend to satisfy this
comfort criterion well by minimizing the vertical temperature gradient but
care should be taken to ensure that the surface temperature is not greater
than about 27 °CJ 25 1 Similarly, but on a higher plane, discomfort may be
experienced if the head is subjected to radiation from a heated ceiling. The
CIBS gives no specific temperature but cites the dependence of the maxi-
mum permissible temperature on room height and area and reports that
temperatures of up to 40 °C have been found acceptable in rooms of up
to 5 m square and of usual height J 25 1
Table 3.5 gives the CIBS recommendations for design conditions and
Table 3.6 for resultant temperatures in a variety of spaces.
Humphreys^ 2°1 provides an excellent discussion of desirable tempera-
tures in dwellings. Fig. 3.2, from his work, shows approximate tempera-
tures for thermal neutrality, a state in which the occupants experience no
heat stress or thermal strain. In constructing the chart it was assumed for
simplicity that the combined effect of the temperature of the air and the
50 Building design
Table 3.6. Recommended design values for dry resultant temperature^23^
'res 'res
Type of building (°C) Type of building (°C)
Art galleries and museums 20 Hotels:
Bedrooms (standard) 22
Assembly halls, lecture halls 18 Bedrooms (luxury) 24
Public rooms 21
Banking halls: Staircases and corridors 18
Large (height > 4 m) 20 Entrance halls and foyers 18
Small (height < 4 m) 20
Laboratories 20
Bars 18
Law courts 20
Canteens and dining rooms 20
Libraries:
Churches and chapels: Reading rooms (height > 4 m) 20
Up to 7000 m3 18 (height < 4 m) 20
> 7000 m3 18 Stack rooms 18
Vestries 20 Store rooms 15
Dining and banqueting halls 21 Offices:
General 20
Exhibition halls: Private 20
Large (height > 4 m) 18 Stores 15
Small (height < 4 m) 18
Police stations:
Factories: Cells 18
Sedentary work 19
Light work 16 Restaurants and tea shops 18
Heavy work 13
Schools and colleges:
Fire stations; ambulance stations: Classrooms 18
Appliance rooms 15 Lecture rooms 18
Watch rooms 20 Studios 18
Recreation rooms 18
Shops and showrooms:
Flats, residences, and hostels: Small 18
Living rooms 21 Large 18
Bedrooms 18 Department store 18
Bed-sitting rooms 21 Fitting rooms 21
Bathrooms 22 Store rooms 15
Lavatories and cloakrooms 18
Service rooms 16 Sports pavilions:
Staircases and corridors 16 Dressing rooms 21
Entrance halls and foyers 16
Public rooms 21 Swimming baths:
Changing rooms 22
Gymnasia 16 Bath hall 26
Hospitals: Warehouses:
Corridors 16 Working and packing spaces 16
Offices 20 Storage space 13
Operating-theatre suite 18-21
Stores 15
Wards and patient areas 18
Waiting rooms 18
(As, for example, Table A2.2.) (Reproduced from Section Al of The CIBS Guide, by
permission of the Chartered Institution of Building Services.)
Fig. 3.2. Approximate
temperatures for thermal
neutrality for various
weights of clothing (air
velocity 0.1 m/s).[20]
(BRE, Crown Copyright,
HMSO.)
40 50 60 70 80 90 100 110 120
room surfaces is sufficiently accurately given by the room temperature of

a globe thermometer, that the air movement is generally slight and that the
humidity of the air has only a small effect on perceived warmth. (The last
assumption is supported by Mclntyre who, in investigating whether an
increase in humidity would permit lowering air temperatures, concluded
that the effect of humidity on warmth was so small that it might safely be
neglected and thus that the air temperature could not be safely
lowered.)!261 The metabolic rate in W/m2 of body surface is a reflection
of the energy expenditure of the occupants. 67 W/m2 is an average daily
value for men and women. Durnin & Passmorel27! give a complete range
of energy expenditure in daily life.
The 'Clo values' refer to the amount of clothing worn. One Clo
(0.155 m2 °C/W) is the insulation afforded by a suit as normally
wornJ 28 ! Half a Clo corresponds to a skirt and blouse or trousers and
shirt and one and a half Clo to a suit plus thicker underwear, a waistcoat
and/or cardigan and thicker socks. A change to 1.5 Clo permits the accept-
able room temperature to be decreased by almost 4 K at the average meta-
bolic rate!
Lightweight indoor clothing could be developed so that higher insu-
lation could be achieved without reverting to the bulky apparel of 50 years
ago and Humphreys, although noting this, goes on to base his conclusions
on the more common value of 1.0 Clo. He suggests that in winter the nor-
mal daytime temperature for a dwelling might be 19 °C with provision for
raising or lowering the temperature by perhaps 4 K in a particular room
for a particular occasion. To allow for different probable usages, normal
settings of 21 °C for living rooms, 17 °C for kitchens and, say, 15 °C for
passages and stairs are suggested. (In well-insulated houses these tempera-
tures will tend to be closer to each other.)
52 Building design
Fig. 3.3. Corrections to the
resultant temperature to
take account of air move-
ment.! 2 3 ] (Reproduced
from Section Al of the CIBS 2
Guide, by permission of the O
Chartered Institution of
Building Services.)
3
0.2 0.4 0.6 0.8 1.0
AIR MOVEMENT/(m/s)
Bedrooms present a special case. With sufficient insulation a sleeping

person can be warm enough in a bedroom at below 0 °C and it is common
for many people in the UK to sleep at a temperature which is not far from
ambient. However, Humphreys recommends 15 °C to allow in part for
people reading in bed, for example with their shoulders exposed, and to
permit moisture produced by sweating during sleep to evaporate from the
bedding. Should the bedroom be used for study or as a sitting room it is
suggested that provision be made for raising the temperature to 22 °C.
These results serve as an excellent point of departure for an actual
design and indicate the subjective nature of some of the assumptions that
must be made when recommending comfort temperatures.
It must also be remembered that specific local conditions can be of
great importance. Sunshine streaming in a window, an open solid-fuel fire
and an electric-bar fire are examples of radiant sources of heat which by
heating the person permit lower air temperatures. The psychological effect
of seeing the source of heat may also have a bearing on comfort. Finally,
the needs of special groups must be considered and so, for example for
elderly people, care must be taken to provide higher temperatures to com-
pensate for the reduction in physical activity.
3.3.3 Room air movement, ventilation and relative humidity

A certain amount of air movement is desirable to avoid feelings of
stuffiness and areas of stagnation. However, because of the cooling effect
of air, its excessive movement should be avoided so that room occupants
do not complain of draughts. One CIBS report^25! indicates that seden-
tary persons prefer air speeds of 0.1-0.2 m/s if the air is at ordinary room
temperature - at lower temperatures only the lower part of the range is
acceptable and, correspondingly, at higher temperatures (for example, dis-
charge from warm-air grilles) the upper range or somewhat higher is satis-
factory.
The CIBS guide recommends raising the resultant temperature when
room air velocities are greater than 0.1 m/s. Fig. 3.3 shows the suggested
corrections.
Local sources of discomfort may be dominant in determining whether
Fig. 3.4. Air supply
\ Occupant Outsider required for odour removal
\ 1 Good Acceptable - a d u l t s . [ 3 2 ] (BRE, Crown
7 io 2 Intermediate Copyright, HMSO.)
c 3 Fair Objectionable
o
CD
Q.
CO
O.
CD
1
Uddns
| " •• ""• 2
- 3
n i i i
5 10 15
Occupation density — m3/person
individuals consider themselves to be uncomfortable. Low-level draughts

caused by infiltration and cold convective currents can be particularly
troublesome^29! and due care should be taken to avoid them; if this is not
done, higher internal temperatures (with correspondingly greater energy
consumption) may be set by the occupants to achieve a similar overall
standard of comfort.
Ventilation in buildings is required to provide a continuous supply of
oxygen for breathing, to remove the products of respiration and occu-
pation and to remove artificial contaminants produced by cooking, process
work and so forth; in the latter category water vapour and heat may be
included. Additionally, ventilation may be needed to provide air for a
heating system.
The oxygen supply needed for breathing is about 0.009 g/(s p) (at an
energy expenditure rate of 2500 kcal/d),l30J giving an air requirement of
0.03 l/(s p), but since this is significantly less than that required for other
purposes it never serves as a basis for design. Similarly, to prevent the CO2
produced by respiration from exceeding the maximum permissible level
(assumed to be 0.5%), the fresh-air supply need only be 1.25 l/(s p)J 31 1
To keep the atmosphere fresh and reasonably free from odour, how-
ever, requires considerably more fresh air, as can be seen in Fig. 3.4. If we
consider a closed living area of, say, 50 m 3 with four people in it, the
'intermediate' situation, curve 2 of Fig. 3.4, calls for about 2 l/(s p) or
roughly one-half an air change per hour.
The subjective impression of fresh air has important implications for
energy conservation - the main reasons for opening windows in winter
are to freshen the air, to avoid condensation and to avoid stuffiness.1331
Odours from heating appliances, particularly those burning paraffin, can
also lead to a sensation of unpleasantness. (Additionally, large amounts of
water vapour may be produced by such devices.)
Detailed air-supply rates are given by the CIBSf23! according to the
type of space.
Less-commonly appreciated contaminants than those such as CO 2,
54 Building design
38
Table 3.7 .Range of open areas measured in six UK dwellings^ ]
Effective open area (cm2 )
Before After
Adventitious opening weatherstripping weatherstripping
Door 38-210 3-45
Window 6-110 3-14
Background (room with
suspended floor) 52-150 52-150
Background (room with
solid floor) 25-62 25-62
combustion products, tobacco fumes and body odours have been the sub-
ject of recent research. For example, formaldehyde (HCHO) is found in
many products used by the building industry and studies have shown that
in houses with low ventilation rates contaminant levels can be too
highJ 34 ! A second case is radon which is one of the products in the decay
chain of uranium. Uranium is found naturally in most rocks, soils and
common building materials. Radon and its daughter products emit alpha
particles which, if inhaled, can damage lung tissue. Long-term low-level
exposure levels have not been established but work is underway on the
physiological effects of radon and on the modelling of alpha-particle
activity levels at various air-change rates.l34 l Brundrett has stated that
'present evidence suggests that to avoid danger the minimum ventilation
rate in Britain should be 0.2 air changes an hour'.! 35 ! Fuller, in the US,
believes that 0.5 air changes per hour can keep indoor pollutants below
critical levels.l34l
Natural ventilation is due to the pressure differences resulting from the
action of the wind on open areas (wind effect) and the variation in the
buoyancy of the air inside and outside (stack effect). Recent research indi-
cates that the stack effect is important in two-storey buildings as well as
taller ones. (Monitoring of a full-scale test structure! 36! 4.9 m long by
2.4 m wide and 2.4 m high has shown that the stack-effect component of
the infiltration is a function of the effective leakage area and its distri-
bution, the degree to which the house is shielded from the wind, several
geometric parameters and the square of the indoor—outdoor temperature
difference.)
The open areas include purpose-provided openings such as vents, flues,
chimneys and openable windows, cracks in and around room components
(that is, doors and windows) and 'background leakage areas' which are left
when the first two groups of openings are sealed - examples of these
include joints between ceilings and walls and the porosity of room sur-
faces. l 37 l Table 3.7 shows the effective open areas in some UK houses and
the importance of the background openings.
To make the ventilation rate less dependent on wind speed and external
temperature it is necessary to reduce the total effective open area to a
small value. It then becomes possible to use mechanical ventilation systems
-•- Natural ventilation-House unsealed Fig. 3.5. Natural and mech-
anical ventilation rates in
O Mechanical ventilation-House unsealed British Gas Corporation test
3-0
• Mechanical ventilation-Windows and house.f 3 8 !
doors sealed
o
#
• o
o •
° o
v +
• + %
5 + O
2 3 4 5
Wind speed(m/s)
(see Chapter 6) to control the air-change rate in the house because they are
thus able to generate a relatively large pressure. If the total effective open
area is not reduced a mechanical ventilation system simply results in a
higher air-change rate as can be seen in Fig. 3.5.
To effectively reduce infiltration rates it is necessary not only to
weatherstrip windows and doors but also to ensure that care is paid in the
construction process to eliminate openings such as cracks between window
frames and masonry walls, around light fittings and where pipes pass
through the structure (services in general are a major source of infil-
tration), in floors and around skirting boards. This would reduce the total
effective open area by attacking the problem of the background areas.
Greater control on-site would be required and inspection procedures
would have to be developed. By employing these measures and others such
as external lobbies, higher glazing levels and the closing of chimneys when
not in use, it should be possible to reduce the hourly air-change rate to a
range of about one-half to one. A great deal of research is presently being
undertaken to assess the cost-effectiveness of such measures and to deter-
mine how best to ventilate homes. The two principal options are mech-
anical ventilation in a very 'tight' house or natural ventilation with fewer
adventitious openings but with purpose-provided openings which give
some control over ventilation to deal with, for example, problems of con-
densation (see below) and stuffiness. At present, ventilation rates are not
specified by the Building Regulations and ventilation is one of the least
well-understood aspects of design. The situation may be similar to the
lightweight v. heavyweight argument and research may prove both
solutions to be suitable.
56 Building design
Fig. 3.6. Ventilation heat 1976 Building Wei I-insulated
loss as a percentage of the Regulations
total design heat loss (for a
typical semidetached house
of 90 m2 floor area for three
insulation levels; the specific
fabric heat losses are: tra-
ditional, 1.4 W/(m3 K);
Building Regulations
1.0W/(m3 K); well-insulated
house 0.5 W/(m3 K) (after Design heat loss 7.6 kW 5.8 kW 3.6 kW
Ref. (381). (BRE, Crown
Copyright, HMSO.)
In the past, little attention was paid to the heat loss that accompanies
natural ventilation because it was a minor component of the total loss.
However, as insulation standards increase it becomes more important, as
can be seen in Fig. 3.6.
In traditional English houses, hourly air-change rates were approxi-
mately in the range of one-half to three (see Fig. 3.5)J 38 ' 391 European
construction, with its different techniques and sometimes higher stan-
dards, was able to reduce this considerably.
Fig. 3.7 shows some of the results from a study of ventilation rates in a
group of modern Belgian houses conducted by the BRE and its Belgian
equivalent (the Centre Scientifique et Technique de la Const ruction) J 40 J
The Belgian houses were single storey with three bedrooms and were con-
structed of large prefabricated concrete panels with a central core of poly-
styrene. Joints between the panels were sealed. Windows were wooden
framed and single glazed. The method of construction and the higher
quality of the windows were the principal reasons for the relatively low
air-change rates found.
Without a doubt, the initial effect of reducing infiltration is advan-
tageous because less energy is wasted and because lowering the air speed in
the room permits lower temperatures while maintaining the same degree of
comfort.
Secondary effects may, however, be harmful. Care must be taken to
ensure that the quality of the room air does not suffer, that sufficient air
is supplied for heating appliances that are not room sealed and, as men-
tioned before, that local sources of discomfort are avoided. Nevrala &
Etheridgel37! have discussed complaints of discomfort due to low-level
draughts arising in certain situations such as a well-insulated room which
retains a single-glazed window. They also point out that, with increasing
insulation, heat emitters have become smaller and, in addition, the econ-
omic pressure to save capital cost has resulted in a curtailment of the
former practice of oversizing heating-system components. The result is
increasing sensitivity to the ventilation load. Care must be taken to allow
for the 'background leakage areas' which become correspondingly more
important as draught stripping is employed, and to take into account the
magnitude and variability of the air-change rate which can arise from
changes in wind speed and direction.
The greatest secondary effect is the danger of condensation and mould
growth which even now are common problems in dwellings. As is well
ventilation rate Fig. 3.7. Measurement of
(air changes/hour) whole house natural
ventilation rates.!40 1
BRE measurements
O with heating
9 no heating G
09 G
• ventilators shut
CSTC measurements
X with heating G
08 -
G
07 - G
G
G
06 -
G 3
G
0 5 - G
X
G
0 A - G o^ •
G (1^
G °
G 0
0-3 -
0-2 - X
0-1 -
U 5 6
wind speed (m/s)
known, the amount of water vapour that air can hold is limited and at this
limiting point the air is said to be saturated. The higher the temperature of
the air the greater the amount of water vapour it can contain. The water
vapour contributes to the vapour pressure of the air and the ratio of the
vapour pressure of a given mixture of water vapour and air to the vapour
pressure of a saturated mixture at the same temperature is called the
relative humidity. Fig. A 3.2 shows how these variables are related. At a
given temperature, as water vapour is added to the air the relative
humidity will rise to the saturation point (100% rh) and any further
vapour will be deposited as condensation. Alternatively, if the temperature
of a given amount of air is lowered and the amount of water vapour
remains constant the relative humidity rises because the cooler air can
support less moisture. At a relative humidity of 100%, any further cooling
causes water vapour to condense. For example, at point A in Fig. A 3.2
the dry-bulb temperature is 20 °C, the mixing ratio is 10.4 g/kg and the
relative humidity is 70%. If this air is cooled to the dew-point temperature
of about 14.7 °C where the relative humidity is 100%, point B, it can hold
no more moisture.
The latter causes problems in buildings when, for example, warm, moist
58 Building design
air reaches a cold single-glazed window or even the wall of an unheated
bedroom. An important point to note is that condensation, particularly in
dwellings, does not necessarily occur in the room where the water vapour
is produced - any cooler surface in the house may provide a suitable site.
Condensation can also occur in the building fabric since nearly all building
materials are permeable to water vapour. The BRE has developed a
method^411 of assessing the risk of this interstitial condensation which is
outlined in Appendix 3, and the problem is further discussed in Section
3.8.
Mould growth can occur when the average relative humidity exceeds
70% for extended periodsJ 42 ! This situation may arise, depending on the
standard of insulation, particularly in bedrooms which are normally
unheated or only intermittently heated.
The figure of 70% is also an approximate upper limit for human com-
fort, with 40% being a lower bound. Low relative humidities can lead to
electrostatic shocks, while physiological factors such as dry throats and
cracking skin have been found to be related to low absolute water-vapour
pressures.!231
The effect of lower ventilation rates is to increase the moisture content
of the air as can be seen in the equation
where
G = mean rate of moisture emission;
p = density of air;
v = volumetric rate of air exchange;
gi,go = mean inside and outside moisture contents (mass of water
per unit mass of dry air).
Typical moisture emission rates are shown in Table 3.8.
Cooking and drying clothes are the major moisture producers shown in
the table. One way of combatting condensation and mould growth is to
isolate moisture-producing activities by, for example, washing and drying
clothes in a utility room separated from the rest of the house. Other
suggestions include an extract fan in the kitchen (this may pose problems
if the kitchen contains a boiler whose flue is not sealed from the
room),I 42l and increased insulation and additional heating in bed-
rooms. f43J In general, there is no ideal solution to the problem, short of
moving to a more hospitable climate, and so care must be taken to allow
some flexibility in the design by, for example, permitting higher ventilation
rates than would be desirable from the point of view of energy conser-
vation alone. Table 3.9 summarizes the advice to designers of one compre-
hensive guide to condensation in dwellings.
If there is no ideal solution there is at least a practical radical one which
is a combination of the ideas discussed above. It starts from a reasonable
position - that people are people — then moves to one less easily defens-
ible — that people won't change and will therefore continue, for example,
boiling vegetables until the kitchen is like a steam bath. The idea is that
Size and type 59
Table 3.8. Moisture emission
Moisture
emission
Sources where moisture per day
emission is low (kg) Comments
5 persons asleep for 8 h 1.5 Typical moisture emission rates
2 persons active for 16 h 1.7
Cooking 3.0 3 h cooking by gas cooker
Bathing, dish washing, etc. 1.0 Estimated
Daily total 7.2
Additional sources where
moisture emission is high
Washing clothes 0.5 Estimated
Drying clothes 5.0 Measured, for 3 kg dry weight
of clothes, spun dry
Paraffin heater during
evening 1.7 1.7 1 paraffin, e.g. 4 kW for 5 h
Daily total 14.4
ventilation rates should be kept high, as in the past, for personal comfort,
to avoid condensation and to ensure that the building fabric is free of
danger. To avoid a profligate use of energy not only would a mechanical
ventilation system be used, but heat recovery on the extract air would be
incorporated (see Chapter 6). Such systems have been recommended by
many, tried by some and are commercially available for the domestic scale.
The only problems are the cost of sealing the house (reducing the total
effective open area), as we have seen it is necessary to do, and the cost of
the equipment.
3.4 Size and type

Reducing the size of a building to a minimum will lower the energy costs
of construction, as well as those of heating. Given the human need for
space, which is albeit culturally conditioned, this tends to be a fairly
unpopular solution to energy conservation; however, where the density of
occupation (volume per person) is naturally high as in, say, schools com-
pared to homes, the scope for economy may be greater.
A change in building type may prove more acceptable. The appearance
of grouped condominium housing in the US may hearken a development
towards the terraced housing of the UK.
The relative advantages of terraced homes, which of course have fewer
walls exposed to the outside air, are evident in Fig. 3.8 (the absolute
seasonal heat requirements shown have decreased in all cases due to the
introduction of higher insulation standards). Work at the BRE has given
similar results.!451
For the Basildon housing discussed previously, the effect of varying the
terrace length was considered and it was concluded that very short terraces
60 Building design
44
Table 3.9. Precautions for avoiding condensation 1 '
Key aims
1
1.1 Assure reasonable bedroom temperatures. Target: maintain a
temperature of at least 10°C at all times
1.2 Provide fan extraction to kitchens
1.3 Provide ventilated clothes-drying space
2 Make sure structural £/-value is adequate in all places, that the
insulation is positioned to the best advantage and is not broken
by 'cold bridges'
3 If vapour barriers are necessary, ensure that they are properly
positioned
4 If interstitial condensation is likely to occur, make certain that
it can escape and that vulnerable materials are protected
5 Insulate cold-water pipes
6 Make certain that flat- or pitched-roof construction has been
properly considered in the light of local climatic conditions
7 Consider carefully the method of paying for heating in relation
to the occupiers' probable economic circumstances
8 Ensure that the heating system has adequate controls so that it
will be used to the best advantage in combining efficiency and
economy
9 Ensure that secondary air supply to warm-air heating is
correctly positioned to avoid recirculation of moist air
10 Avoid dangerous 'dead-end' positions where heating or venti-
lation may be inadequate
11 In window design allow for easily reached and easily controlled
ventilation without rain penetration
12 Ventilate larders to the outside
13 Ventilate other cupboards as necessary to the interior
14 Make arrangements for occupiers to be issued with simple advice
on the avoidance of condensation, especially on the method of
operating the heating
increase the overall energy demand by up to 10% (see Fig. 3.9). However,
increasing the terrace length above eight or so houses gave little overall
advantage.
3.5 Form and orientation

Dear to the heart of all designers is the question of form. Fig. 3.10 shows
11-year-old Judith Cornick's winning entry for a rotating solar house.
Fortunately or unfortunately, according to one's point of view, the
building industry is unlikely to adapt to such novel structures in the near
Form and orientation 61
Fig. 3.8. Analysis of com-
middle of terrace parative space heating
requirements for typical
terraced and detached
houses. (The dots indicate
specific results; the dashed
line shows the trend of
middle of terrace energy demand.)! *21
staggered
end of terrace
detached house
0-6 0-7 0-8 0-9

seasonal heat requirements GJ/m2
Low U valua Fig. 3.9. Effect of varying

Paak demand 5 SkW
terrace length.! 46 !
No Dec Jan Feb Mar Apr

HEATING SEASON
future (Buckminster Fuller's Dome and Alex Pike's Autarkic House

encountered similar problems) and so we are forced to consider more-
conventional solutions.
For large buildings, the constraints imposed by the site and intended
use may be very important and both heating and cooling loads will have
to be considered. The examples selected by Kasabov,!48! ranging from
three-storey rectangles to sprawling structures with courtyards, indicate
62 Building design
Fig. 3.10. A rotating solar
house.f47! (Courtesy of
J. Cornick.)
Base
that a wide variety of forms are currently claiming, perhaps justifiably, to

be energy efficient. There is a strong, easily understood hope on the part
of architects that considerations of energy use will not greatly restrict the
forms available to designers and hence the scope for creativity. It is a hope
we share but, at present, insufficient data are available to decide whether
it is one that is justified. Hawkes & MacCormac, in arguing for office build-
ings with interior courtyards, have presented theoretical calculations show-
ing the possibility of significantly reduced energy demand.I491
For fairly simple forms, computer programs are available, which will
evaluate the heating and cooling requirements J 5 0 ' 51 I Results tend to
indicate that lower buildings of more compact shape are more efficient,
that is, have lower running costs, but one must be wary of the assumptions
made when studying such analyses. In particular, the results can be sensi-
tive to the amount and distribution of the glazing.
Most studies have found that if passive solar gain is not accounted for,
the aspect ratio (the ratio of the plan length to breadth of the rectangle
that has the same area and perimeter of the proposed form of the building)
has relatively little effect on energy consumption (see Fig. 3.11). Nonethe-
Fig. 3.11. Variation of
1.01 primary energy consumption
for heating and lighting, with
aspect ratio and building
area.! 52 1 (No allowance has
been made for solar gains
or for orientation.) (Crown
Copyright.)
0.94
O
Q.
O
o
0.86
0.79
500 1000 1500 2000 2500 3000
Area of Building—m2
less, for small buildings a first guideline should be to avoid the temptation
to use complex forms since they will tend to have an increased surface area
and may result in self-shadowing.
If solar gains are considered, as is preferable, we must first consider the
energy balance at windows. Fig. 3.12 provides the necessary data.
With south-facing glazing and no curtains, over the heating season the
energy balance is negative but, as Burberry has shown by curtaining such
windows, there is neither a loss nor gain (Fig. 3.13). By going to south-
facing double glazing, even without curtains, the energy balance over the
heating season becomes positive and if an insulating shutter is drawn at
night, at no time during the year is the energy balance negative. The advan-
tage of south-facing glazing compared to other orientations is also evident
from Fig. 3.12.
It might seem then that there is a clear-cut case for all glass south-facing
walls but unfortunately this is not so. Designers agree that, whenever
possible, glazing should be to the south rather than other orientations but
the amount of glazing is a point of contention, largely because of disagree-
ment as to how much of the incoming radiation is useful. This is related to
the thermal performance of the structure, the heating system, the desired
internal temperature and so forth and thus will require further study in
monitored buildings. At Basildon (see Fig. 2.7) the (single-glazed) windows
64 Building design-
Fig. 3.12. Energy balance for
one square metre of glazing
in the Bracknell area
(assumed internal tempera- \
ture: 18°C).[ 5 3] (a) Daily
average conduction loss. S 3 \ /
(b) Daily average solar heat \

V
gain through uncurtained ^ 4
single glazing, (c) Daily \ /
energy balance through
south-facing glazing. \ /
id) Daily energy balance
through west-facing glazing.
Sep Oct Nov Dec Jan Feb Mar Apr May
(e) Daily energy balance
through north-facing glazing. (a)
16
/
\
\
/
I 5
14
/
\
>
/
•? 3
Sep Oct Nov Dec Jan Feb Mar Apr May

(b)
occupied 40% of the south wall area,!14! and at Pennyland, Milton

Keynes, the comparable figure was 35%J 541
In the meantime, if we assume conservatively that a curtained south-
facing double-glazed window experiences no net heat loss during the heat-
ing season, the question of the optimum length to breadth ratio changes
and elongated south-facing structures show savings. Recent developments
of terraced houses orientated to the south and with glazing concentrated
on the south, when viewed as units are very energy-efficient forms.
As we have seen (Chapter 2) with an actual site, departures from an
orientation of due south are often required by planning and aesthetic con-
siderations. For terraced houses, computer studies for Basildon indicated
that deviations from the south of up to 22.5° had no significant effect on
the energy demand ;l 46l similar work for Pennyland found that deviations
from the south of up to 30° reduced the potential energy savings due to
passive solar gain by only 2—5%J54J
For semidetached housing with double glazing, a study of the effect of
orientation has shown that groups with the long axis east-west and
glazing concentrated on the south are likely to receive approximately
6.1 GJ of solar gain (4.7 GJ from the glazing; 1.4 GJ through the struc-
ture) during the period from October to March, compared to about
4.7 GJ (3.2 GJ through the glazing; 1.5 GJ through the structure) for
+6-
V\ /f,
nee (MJ/rn^
+2-
n
\ >v double, shutter / / /
Sep ' Oct \ PslovN^Dec1 Jan ' Feb / M a r ' Apr
1
May
Energy i
\
>. double, no /
^v curtains y
^ ^
-6- single, no curtains
(c)
Apr May
single No curtains
id)
+6-
Sep ' \ O c t ' Nov ' Dec ' Jan ' Feb ' M a r ^ / A p r / May
-6- No curtains
single
(e)
groups with the long axis running north-south and glazing concentrated
on the east and west wallsJ 5!
For individual houses, a slight variation in form, which incorporates
large south-facing glazing, is the trapezoidal plan of the Ouroborus
Housel 55! in Minnesota (Fig. 3.14). To reduce the heat loss the north wall
is shorter than the south and is bermed.
As to height, compact two- and three-storey buildings tend to be
efficient because of their reduced surface area in relation to volume.
Bungalows (one-storey detached dwellings) are particularly wasteful of
energy.
66 Building design
Fig. 3.13. Net heat balance
through one square metre of
OILJ J
south-facing single glazing
(assumed internal tempera-
is and losses—MJ pe
+400 - Assumed duration of heating season
ture: 17.1 °C; windows are
curtained at night).[ *21 +350 -
+300 -
+250 -
co
+200 - O)
CD
CO
I
+ 150 -
+ 100 -
+50 -
-50 -
pi
u —
h x l
Ullllll lllllt l/
.-4
-100 -
-150 -
|USt
pril
ary
ber
bru ary
ber
rch
ber
cto ber
> <D
O)
D
<f CD
E
CD
E
CD
ZJ CD < I I
Q. O o CD
o LL
CD QCD
if) z
Fig. 3.14. Ouroborus House

in Minnesota.
External and internal layout 67
5
15-20 m
2 m max.
5-10 m
Stream
(c)
Dependent buildings
— Less used spaces
Most used spaces
(d)
Fig. 3.15. Internal and

external layout for a
detached house.!57!
(a) Ground floor.
(b) First floor, (c) Site
plan, (d) Exterior zone
concept.
For larger buildings Markus & Morris^56! provide a very useful dis-
cussion and summary of the effect of form on energy consumption.
3.6 External and internal layout

Wherever possible, non-heated spaces should be used to shelter heated
ones. Fig. 3.15 shows how a garage and berm form part of an exterior zone
which is organized on a north—south axis.
If possible, the structure should be arranged to avoid self-shadowing
68 Building design
Fig. 3.16. Solar heated flats

at Lewisham, London.!* 5 1
(Courtesy of Max Fordham
and Partners.)
where passive or active solar gain may be used. The advantage of increased
architectural interest which results from more-complex forms must be
weighed against increased heat loss and potential loss of solar energy.
South-facing bay windows, for example, will have a less favourable energy
balance than a plane window because of the increased surface area and
because all of the glazing is not at an optimum angle for solar gain.
Additionally, by creating pockets of air pressure, bay windows and similar
features can cause greater infiltration lossesJ 58!
Of course, on-site a compromise is inevitable. Fig. 3.16 shows the
arrangement of the Lewisham Flats where the site lay between a row of
terraced houses and a park. The site sloped to the south but faced north
over the park, thus the new buildings were bound to overlook the terraced
houses and would potentially block off their view of the park. To over-
come this difficulty (and to receive planning permission) the development
consists of a series of blocks with gaps in between to maintain visual access
to the park. The south-facing roofs of the flats are much higher than those
of the terraced houses and so are free from this form of overshadowing,
but some shading will be caused by adjacent blocks and by the extended
walls required by building regulations to retard the spread of fire.
Fig. 3.15 illustrates many of the guidelines for internal layout.
Obviously, these will vary with house type, client preferences, choice of
heating system and so forth but it should be possible to accommodate
some energy-conserving measures in any design. A principal objective, well
understood by the designers and occupants of nineteenth-century terraced
housing, is to divide the internal plan into areas which can be treated dif-
External and internal layout 69
Table 3.10. Temperature distribution upstairs and downstairs in a highly
insulated semidetached house^60^
Downstairs
(lounge, diner/kitchen) Upstairs (3 bedrooms)
Energy Average Energy Average
consumption temperature^ consumption temperature^
(MJ/d) (°C) (MJ/a) (°C)
Internal doors
closed 59.4 19 10.8 15
Internal doors
open 93.5 19 3.6 16.5
a
Control temperatures were 19 °C downstairs and 15 °C upstairs. The
house was unoccupied and had an open staircase.
ferently according to need. The north part of the plan consists principally
of task-orientated spaces which are used relatively infrequently, while to
the south are the major habitable rooms where the sun provides essential
warmth and light. In general, services are concentrated in the north-east
corner of the house to reduce pipe runs and to facilitate future installation
of, for example, a heat pump to recover heat from the bath water.
Adequate lighting in the house is provided by the mainly south-facing
glazing. The need for electrical lighting should be minimized and, as a rule,
a small heat loss should be accepted in preference to a requirement for
electrical lighting because of the primary energy demand of electricity (see
Chapter 6). For this reason, the stairwell is lighted during the day by a
small north-facing window. Where a choice exists, windows in walls are
preferable to skylights because the thermal coefficient is lower (single
glazing in a roof has a {/-value of about 6.6 W/(m2 K) compared to
5.6 W/(m2 K) for glazing in a wall (see Table A 2.7) and, additionally, the
temperature will be higher at the roof than the wall); any skylight installed
should be provided with a shutter to cover it at night. (In commercial situ-
ations, however, where a certain level of illumination is required continu-
ously, the question of horizontal v. vertical daylighting is more complex.
It will depend on the illumination required and tables of frequency of day-
light levels must be consulted.)!59J
The staircase is sealed to separate the bedroom temperature zone from
the ground floor and to reduce heat losses due to the 'stack effect'. Sealing
the staircase may also help to prevent warm, moist air rising from the
kitchen—dining areas below to the cooler areas upstairs and causing con-
densation. Provision of a mechanical extract fan in the kitchen, perhaps
operated by a relative-humidity sensor, is a useful precaution and will also
remove objectionable odours. (Mechanical ventilation of bathrooms and
toilets should also be considered.)
An alternative approach which is sometimes adopted is to leave the
staircase open and use the ground floor to heat the first floor. If this is
done, experimental evidence presented in Table 3.10 has shown that a
significant saving in energy consumption (70.2 MJ/day compared to
70 Building design
91A MJ/day) may be achieved by closing internal doors and accepting a
somewhat lower temperature in the upstairs bedrooms.
We would recommend that thermal zoning be used vertically and hori-
zontally when possible.
Free heat from the sun can be used more effectively than that from the
kitchen. The essence of passive solar design is to use the winter heat gain
and avoid overheating which is perhaps most likely to occur in the spring
and autumn but, in some instances, may also occur in winter^611 (see
Chapter 4). Utilization of passive solar gain is increased when it can be
redistributed among a number of spaces from areas where there is an
excess. Cross-currents should be avoided except when specifically needed
to prevent overheating. If for a particular design it seems likely that over-
heating may occur, it is best to plan the interior spaces so that warm air
can reach, and then exit at, the top of the building.
The entrance hall acts as a draught lobby, a feature which should
always be included. In designs where the entrance is from the south, the
lobby can be a small glazed conservatory and thus increase the passive
solar gain.
A woodburning stove for amenity purposes and to make use of any
wood produced on site is provided and located centrally to take advantage
of heat losses. (The flue must be equipped with a well-fitting damper to
prevent high heat losses when the fireplace is not in use.) Linen cupboards
with heaters and hot-water storage devices should also be located away
from external walls to make the best use of their heat.
Fig. 3.17 shows the application of many of these ideas to a small
terraced house, Ambient Energy Design House 1, which was designed by
the authors and Lucy Krall (see also Chapters 6 and 12).
A point of note is the kitchen position. In a small house, options are
restricted and in the design of Fig. 3.17, the north-east corner seemed the
best place for the kitchen even though it meant that the window looked
out on the street. It has been argued that kitchens should not be
positioned on the south side in houses designed for passive solar gain as
there is a risk of overheating caused by the combination of solar gains and
heat generated by cooking, but this would seem to vary too greatly with
specific designs to be accepted as a rule. As to whether the kitchen and
dining room should be separated, at Basildon it was thought that if the
kitchen had a mechanical extract there was little difference whether they
were separated or combined, but if there were no such device there was
probably a slight advantage in combining the two since the greater amount
of air would aid in absorbing any moisture produced.l46l
3.7 Construction - general
For a number of years it had been recognized that UK insulation standards

were dreadfully low compared, in particular, to northern European stan-
dards. I 63 ! Fortunately, the recent trend has been towards more and more
insulation, as can be seen in Table 3.11 which compares the Building
Regulations requirements with a number of other recommendations.
71
Wl NDOWS South racing windows ar«

GARDENS are longer than wifc m the traditional
manner which allows the family to enjoy a lawn larger than rho» 0*1 the north side, Th* north
and flowers close to the house while still having window* are double glazed and have heat
space to grow vegetables at ine 1ar end. reflecting gins internally to reduce heat ion.
AH windows have Timber framei to reduce heat
GARDEN CONSERVATORY/LOBBY: To reduce transmission The two south bedroom window!
ventilation heal loss while still providing garden arecJose together 10 reduce the construction
access and ipxe to keep boots and IOOIS. cost of the sotar air heater on this pan of the
WALL CONSTRUCTION: 250 brick block cavity

construction nvith cavitv fill plus 100 mm SOLAR COLLECTOR is fixed by timber c o m
po^y tire thane insulation fined internally onto the southern upper part of Ehe facade as
U = 0 . 2 W / m ' "C well as on the south facing roof. [ S e l e c t i o n ] .
HEATING: Upstairs all room* are warmed Irom

one air outlet at the top ol the Mn, AH doors
upstairs have transfer gnlh to permit (he warm
air to circulate The warm ait is forced through
duct between
the floor joisK and under ihe top step to emerg*
in the centre of the house f ram where the air
can circulate freely and is extracted from the
bathroom.
LfVJNQ AREA, Closing ibese doors an certain

days when healing •£ required, the system permits
the main living area lo be healed ••> • In >,IVI ,- by
Mlar gain.
THE BOILER is houted in a separate Space with
access Irom (he hall and with a grill to the living
area to allow the recovery o1 any 'wasie' running
heat.
STORE e.g. FOR FUEL: Sealed olf thermally by
floor to ceiling panition-the fower pan consist!
ol inflating blocks whirjh can be removed lor
access. Where solid fuel boilerl are TO be used, the
floor slope? at 30° to fjciliute movement of coa!
chutes cou Id make powinie compleieiy automatic

boiler feedintjl.
UNATAP to -diver water and energy
THERMAL ROCK STORE: This is2 m J — !

constructed of hlockwork and filled with fast sized
rocks.
ENTRANCE LO86Y: To remove heat Ion

through ventilation. Both doors are ', single
gltwd.
PARKING: O u t p a c e per house is allowed
GROUNDFLOOS
Fig. 3.17. Ambient Energy

Design House.! 62] (fl)
Ground-floor plan, (b) First-
floor plan.
72 Building design
2
Table 3.11. U-values (W/(m K)) for dwellings
Ground External
Source floor wall5 Roof Glazing
(1) 1976 Building
Regulations! 641 1.0 0.6 single6
(2) 1981 Building
Regulations^ 21 -c 0.6 0.35 single d
(3) DOE better insulated
housing! 651 0.5 0.3 single—double
(4) British Gas well-
insulated house!91 0.45 0.4 0.3 double
(5) BRS low-energy
houses (terrace)!66! 0.35 0.35 double optional
(6) Electricity Council
Research Centre target!67J 0.3 0.3 0.3 double
a
Excluding glazing.
b
Walls plus windows were required to give an average {/-value not exceed-
ing 1.8 W/(m2 K).
c
If the floor is between a dwelling and the external air or between a dwell-
ing and a ventilated space the maximum {/-value is 0.6 W/(m2 K).
d
Single-glazed window openings can be up to 12% of the perimeter wall
area. If double or triple glazing is used the 12% may be similarly increased.
There is thus an encouraging trend towards substantially reducing heat

losses. New buildings, other than dwellings, at present have a maximum
{/-value of 0.6 W/(m2 K) for walls, floors and roofs.! 68! In the relatively
near future we are likely to see all buildings, including dwellings, with
{/-values limited to 0.2-0.3 W/(m2 K). Double and triple glazing will
become common. In Scandinavia, which is admittedly considerably colder
than Britain, such insulation levels are often already attained.
One reason why insulation will be increased in new houses is that it is a
very easy thing to do and one which the occupants should not need to
worry about for the life of the house. Other ways of saving energy, such as
heat-recovery devices, are often mechanical and fraught with the usual
problems of maintenance, replacement of parts, and so forth. The diffi-
culties of increasing the insulation level are mainly constructional and the
BRE has recently started a study of how the highly insulated houses of the
future will be built.! 69 ! Part of the intention is to compare design details
shown on the drawings and those achieved on-site and to investigate the
additional work required to install the insulation.
A detail of some concern in the design of low-energy buildings is the
avoidance of cold bridges. The term refers to the rapid passage of heat
from inside to out if certain parts of the construction have a much lower
thermal resistance than adjacent parts, as can occur when a metal frame-
work is exposed both inside and outside. An example we are familiar with
was the steel windmill tower of the Autarkic House, which passed through
a highly insulated roof. The solution adopted was to adequately insulate
all interior parts of the tower with sprayed polyurethane foam.
Construction - general 73
Fig. 3.18. Incorrect and
Cold bridge correct roof construction
techniques.! 70) (j) Cold
bridge, (b) Inadequate
ventilation, (c) Acceptable:
continuity of insulation at
ceiling to waJl junction
while retaining ventilation
path.
Ventilation path blocked

by insulation
Sag in underlay
between rafters
Board between
rafters
Another example is given in Fig. 3.18(#) which shows how a cold bridge
may be formed in certain forms of roof construction. Fig. 3.18(Z>) indi-
cates that other forms avoid cold bridges but instead block the ventilation
path; and Fig. 3.18(c) shows one way of avoiding both problems.
A wide variety of insulating materials is now available to the
designer.^71' 721 Plasterboard-backed polystyrene slabs (available with a
vapour barrier), sprayed ureaformaldehyde foams, mineral-fibre mats and
cellular glass blocks are but some of the materials on the market. The
correct choice will depend principally on the application and cost but con-
sideration should also be given to how the insulation level could be
increased in the future since some materials lend themselves better to this
than others.
74 Building design
Fig. 3.19. Penetration of
formaldehyde vapour into
a building.! 73 ) Formaldehyde
vapour
Classroom
Formaldehyde
vapour
Safety of insulation materials is a raging issue at the moment and is

likely to remain one for some time. Greater choice of insulants has not
been without its drawbacks - manufacturers have rushed to put new
products on the market before failsafe ways of incorporating them in
buildings have been developed. To take but one notorious case, the use of
ureaformaldehyde foam cavity insulation is being seriously questioned in
the UK because it has proved difficult, if not impossible, to ensure that the
cavity is completely sealed on top, bottom and sides to prevent toxic
formaldehyde gas from entering the building. Fig. 3.19 shows a problem
case in a school where gaps under window sills allowed air to be blown
into the cavity pushing formaldehyde gas through openings in the inner
leaf of the wall and into a classroom. Numerous complaints of headaches
and respiratory problems resulted and the school has been closed pending
remedial action.
Fire safety is also being scrutinized. The increasing number of plastic
products in buildings and the large volumes of plastic insulants which are
often used have given some cause for concern!74> 75» 761 and due caution
should be exercised when using such insulants. To cite one example, the
Agrement Board has insisted that where the cavity of a house is insulated
with polystyrene beads the top of the cavity must be sealed with a non-
combustible material to prevent the beads blowing out and into the eaves
where they could constitute a fire hazard.!711
Most insulation materials are, however, safe (government testing pro-
cedures are likely to become more stringent in the future) and the ques-
tion the designer faces is how much of a material to use. In determining
the insulation level it must be remembered that the law of diminishing
returns applies since the thermal transmittance of a layer of material is
given by the thermal conductivity of the material divided by its thickness
(see Appendix 2). Thus to use an extreme example, if a building with
600 m2 of surface area is relatively uninsulated ((/-value of 4.0 W/(m2 K))
and has a fabric heat loss of 48 kW, if 100 mm of expanded polystyrene is
added to the walls the heat loss falls dramatically to about 3.8 kW. If a
second 100 mm of insulation is added, the heat loss falls by only another
1.8 to 2.0 kW.
Construction - general 75
Increasing insulation also has a hidden cost at times. If, for example,
one insulates an existing wall internally, living space is lost. On the other
hand, with new houses especially (but in certain cases, such as loft insu-
lation, in existing ones), increased insulation may affect only the material
costs with the increase in labour charge being minimal and in such cases
greater insulation can be very worthwhile.
A number of methods have been proposed to determine the economical
thickness and cost-effectiveness of insulation and the interested reader is
referred to the original workJ45> 77> 78> 79» 8°1 It is noteworthy that the
intellectual acceptance of higher insulation levels (as evidenced by the
figures in Table 3.11) is such that in many homes and buildings, more
insulation is being used than is either required by law or calculated to be
cost-effective. It would seem that many people feel that the wolf is at the
door, that past recommendations have not been sufficiently stringent and
so have squandered energy unnecessarily. If increased insulation is to be
provided as insurance for the future, attention should be given first to the
floor, walls and flat roofs since these are the building components which
are not easily altered later.
A major design decision is the form of construction. Interest in timber-
frame construction, for which techniques are well developed, is growing, in
part because it lends itself to high levels of insulation. Other advantages
claimed are facility in obtaining continuity of the insulating envelope and
reduction of cold bridging J 81 1 Air infiltration can be lower and it may be
easier to include an internal vapour barrier than in a masonry construction.
It must be said, however, that the internal vapour barrier is also more
necessary in timber-frame construction to guard against rotting of the
wood.
Nonetheless, traditional masonry construction is likely to remain domi-
nant in the UK for some time. With care it can incorporate relatively high
levels of insulation and, additionally, is more suitable if thermal storage
capacity is an important consideration. Traditional masonry construction,
improved by insulation and vapour barriers, and timber-frame construction
are thus the two most likely possibilities for houses in the near future and
examples of both are given below.
Proper selection of materials and building type can also save
energy.^82' 83J In the UK, typical two-storey houses have been found to
require between 95 and 180 GJ of primary energy (for major building
components only); for large blocks of flats, constructed in reinforced con-
crete the primary energy requirement for construction has been estimated
as 230-265 GJ per flat J 8 4 ' Thus, in a house built to present standards,
the primary energy required for space heating alone over a few years
equals the primary energy required for construction. However, as the
'running' energy demands fall, we may expect more attention to be paid to
construction energy costs. One sign in this direction is the use of recycled
materials in building construction. In the US at least one house has been
made almost entirely in this way with secondhand newspapers being trans-
formed into insulation and wall coverings, old glass becoming bricks and
insulation and recycled aluminium used for roof trusses and gutters.
76 Building design
Fig. 3.20. Wall construc- 115 50 110 16
tions.! 2 ' 9> 7 8 J Contempor-
ary, C/=0.89 W/(m2 K).
(b) Heavyweight insulated,
£/=0.40W/(m 2 K).
(c) Lightweight insulated, - Plywood wall board
£/=0.40W/(m 2 K).
Vapour barrier
(d) Timber-frame wall, -Lightweight block - Mineral-fibre mat (between studs)
£/=0.44W/(m 2 K). - A i r cavity - Waterproof building paper
-Brick - Weatherboard boarding
(e) Timber-frame wall - -Plaster
(d) - B r i c k or
brick veneer,
£/=0.28W/(m 2 K). -Breather p<
. 115 50 110 .. 1 - Cladding
if) Highly insulated,
( / « 0 . 3 W / ( m 2 K).
(g) Highly insulated,
t / ~ 0 . 2 5 W / ( m 2 K). 115 . 100 , 110 , 16
- Foil backed plasterboard

- Glass fibre
- Brick
- Cavity
plasterboard
ightweight block
Cavity fill
3.8 Foundations and walls

In the US, in colder areas, consideration is being given to insulating foun-
dations. In the UK, less concern has been voiced about such losses but at
least one experiment with cellular glass insulation of the edge of the floor
slab (basements are much less common in Europe than in the US) is under-
way. I651 In the Peterborough Houses discussed in Chapter 12, insulation
is used under the whole floor. In the future such protection may become
common.
Fig. 3.20 shows a number of wall constructions of low {/-value and Fig.
3.21 gives two detailed sections used in actual houses (Appendix 2
explains how the {/-values may be calculated).
Many of the designs incorporate a vapour check to reduce the possi-
bility of interstitial condensation. During the winter the higher vapour
pressure inside buildings causes moisture to move towards the outside
(where the vapour pressure is lower). In traditional masonry construction
it seems probable that the problem has been lessened by the ability of the
masonry to harmlessly soak up a large amount of condensate until it can
distil outwards or evaporate from the surface during favourable
periods.I421 However, now, with more-heavily insulated buildings and new
materials, it is not uncommon for the dew point of the moisture laden out-
going air to be reached within the building element. At this point inter-
Foundations and walls 77
stitial condensation occurs and this can lead to structural failure (in
timber-framed homes the increase in moisture content could allow fungal
decay)!85! and greatly reduced thermal performance (for example, if the
insulation becomes waterlogged). (See Appendix 3 for the calculation
procedure.)
The vapour check must be placed on the warm side of the construction.
Commonly this is achieved by using aluminium foil-backed plasterboard or
heavy-gauge polythene sheet just under the interior wall finish. Continuity
of the protection is important but difficult to achieve in practice and
because of this the term Vapour check', rather than 'vapour barrier', is
being used more and more to reflect what happens in actual construction
— a vapour check, even if not completely effective, is still useful.!441 With
aluminium foil-backed plasterboard it is desirable to seal the joints with a
polythene strip (which will have to be put in place, for example, by nailing
to battens, before the plasterboard is fixed). With polythene sheet, seams
may be caulked or taped and then clamped between the various construc-
tion components. Achieving a continuous vapour seal around switch drops,
socket outlets, flues, pipework and cables is also very important but is,
needless to say, a meticulous and time-consuming task. Because of the
likelihood of some moisture penetration into the building fabric, venti-
lation on the cold side of the insulation will be allowed for, in a good
design, if possible, particularly behind areas where achieving a seal is
arduous. One construction technique that bears consideration is to incor-
porate a 100 mm cavity in a brick-block wall but only insulate half of it,
for example by attaching polystyrene slabs to the inner face of the block.
While this raises the (/-value (to about 0.45 W/(m2 K)) it guards against
condensation problems. In almost all forms of construction, given the
difficulty of achieving an adequate vapour seal, it is important that the
exterior surface be permeable to water vapour to avoid moisture accumu-
lation within the construction.
Fig. 3.22 shows an example of a polythene ceiling vapour barrier
stapled to prefabricated roof trusses. Joints were taped to effect a con-
tinuous seal and the polythene was taped to the trap door in the ceiling.
Fig. 3.23 shows the opening for a switch drop in the same ceiling. With
conventionally taped joints it proved very difficult to produce an effective
seal at such points.
The correct position for insulation is related to the heating and occu-
pancy patterns. Where intermittent heating is possible, insulation on the
inner face of the wall will tend to warm up quickly when the room is
heated and will thus give a potential reduction in energy use, provided the
heating-system controls are suitably selected. For continuous heating, the
position of the insulation is less significant.!121 One computer simulation
study of internal and external insulation found that internally applied
insulation was marginally more effective than external.!86] p o r example,
for a semidetached house with 220 mm solid-brick external walls the
internal use of 65 mm thick thermal board (12.7 mm plasterboard-
52mm polystyrene) to give a {/-value of 0.51 W/(m2 K) reduced the
seasonal space heating requirement by 31%. Polystyrene insulation, 55 mm
78 Building design
\
100 mm insulation
13 mm plasterboard
9 mm spacers
50 mm deep f irestop blocking
polythene vapour barrier
75 mm insulation
125 mu(500g) poly-

thene construction d p c
50 mm deep f irestop blocking
vapour barrier turned behind

ceiling plasterboard
foil backed plasterboard
plasterboard lining
500 mu (2000g) polythene dpc
50 mm cellular
glass slab edge insulation
Fig. 3.21. Detailed sections

through two external
walls! 6 5 * 1 4 1 (a) DOE
HDD Better Insulated
Housing ((/-values in
W/(m2 K): wall 0.5, roof
0.3). (b) Basildon Housing
Estate ({/-values in
W/(m2 K): wall and roof,
0.55). (a) Crown Copyright,
(b) Courtesy of Ahrends,
Burton & Koralek.)
50 X 25 Press impregn;
tilting BD. spiked to tru
and to 25 X 34 SW fillet
also spiked to trusses
SW battens fixed
to studs through
ply sheathing ~
125 X 38 SW ledge
tched to joists
80 Building design
Fig. 3.22. Installation of a
polythene ceiling vapour
barrier.!65J (Courtesy of
PSA.)
Fig. 3.23. Opening for a

switch drop in a ceiling with
a vapour barrier.! 6 5 J
(Courtesy of PSA.)
thick with a 13 mm rendered finish, applied externally to give the same

£/-value gave a corresponding reduction of 30%. Our opinion is that the
figures are too close to draw any conclusion as to whether internal or
external insulation is preferable from an energy point of view. The same
study confirmed that the response to heat inputs was much quicker with
internally applied insulation than external.
Because heating systems and occupancy patterns will change during the
lifetime of the building and because few data are available on actual energy
consumption in varying building types, designers must make an informed
Front wall of living room Back wall of main bedroom Fig. 3.24. Sketches from
infra-red thermographs of
r r Abertridwr Project
Houses.[89]
Battens
Contours of
lower surface
temperatures
First floor •
Ground -
floor
i Cold spot,
could be
Cold spots, -I
could be
interstitial cavity bridge
condensation behind insulation
guess as to the best solution. Some consolation for the lack of more-
positive advice may be found in the growing awareness that, as insulation
levels rise, variation in thermal mass, occupancy patterns, heating systems
and position of insulation become, in a sense, less significant. Construction
methods for new houses tend to favour insulation inside or in the cavity. If
continuous heating is desired - and one argument for it has been the
avoidance of serious condensation^87! - an inner cavity wall of high
thermal capacity will give a structure that will warm up and cool down
slowly. For existing homes it is possible to insulate either inside or outside.
One approach to external insulation, which has the advantage of not
reducing the living space of the house, is that used in the Granada
House.l88! Semirigid slabs of mineral wool 100 mm thick were fixed to
existing solid-brick.walls and then covered with redwood weatherboarding.
Another possibility (simulated in the insulation study described above) is
fixing polystyrene boards with steel wire mesh reinforcement and water-
proof rendering.
An interesting way of examining the effectiveness of construction tech-
niques has recently come into more widespread use. Infra-red thermo-
graphs, which display temperature differences as varying colours (indicated
by different patterns in Fig. 3.24), can be used to detect features such as
cold bridges and the possibility of interstitial condensation. In practice,
infra-red thermographs have often shown that insulation is not properly
installed and so the prudent designer should make some allowance for his
or her calculated {/-values not being achieved on-site.
Colours and finishes will have an effect on the thermal behaviour of the
building. The texture of the outer wall will have an effect on the con-
vective heat transfer coefficient and the colour will control the reflection
and absorption of shortwave radiation. These factors become less import-
ant thermally for opaque walls as insulation is increased, although the use
of light-coloured walls to reflect radiation into the house through windows
remains valid of course. Table 3.12 gives the radiation properties of
selected materials.
Depending on the design of the house and factors such as the furniture,
82 Building design
Table 3.12. Emissivity and absorptivity values for various surfaces^90)
Absorptivity
Emissivity of solar
Surface at 10-38 C radiation
Black non-metallic surfaces 0.90-0.98 0.85-0.98
Red brick, concrete and stone, dark paints 0.85-0.95 0.65-0.80
Yellow brick and stone 0.85-0.95 0.50-0.70
White brick, tile, paint, whitewash 0.85-0.95 0.30-0.50
Window glass 0.90-0.95 transparent^
Bright aluminium, gilt or bronze paints 0.40-0.60 0.30-0.50
Dull copper, aluminium and galvanized steel 0.20—0.30 0.40—0.65
Polished copper 0.02-0.05 0.30-0.50
Highly polished aluminium 0.02-0.40 0.10-0.40
(With permission of the American Society of Heating, Refrigerating and

Air-Conditioning Engineers Inc., Atlanta, Georgia.)
fl
See Table 4.1 for more precise figures.
the use of lightweight finishes internally and lightweight partition walls

may lead to a rapid rise in air temperature during periods of high passive
solar gain. Materials with greater thermal capacity, on the other hand, will
have a moderating effect on temperature changes. The use of dark colours
internally will increase the absorption of solar energy at the expense of the
reflected component. If the consequence of this is a need for greater
artificial lighting such an approach should be avoided.
In the future we may expect to see some radical developments in wall
construction. Currently, the trend is to increase the wall thickness to allow
for greater insulation (the conservation house at the National Centre for
Alternative Technology, Machynlleth (see Chapter 8) carried this to the
point of a cavity wall incorporating 450 mm of glass fibre insulation),!91J
but this increases the plan area and raises construction costs. The use of
polished aluminium foil in (unfilled) cavities to reflect heat transferred by
radiation (it has no effect on conduction heat transfer) has been appreci-
ated and advocated for some time. The logical consequence of this has
been the development of high-performance insulation!921 or 'super-
insulation' composed of multiple radiation barriers in an evacuated slab.
Evacuation can prevent the conduction heat transfer. Such a material was
investigated for use in the Autarkic Housel93! but no proprietary product
of adequate thickness and sufficiently low cost could be found. Another
possibility considered for the Autarkic House was foam-filled glass fibre
reinforced plastic (GRP) modules which would be stacked on top of each
other and attached to a frame. Research is also being carried out on solar
'skins' for solid walls. MacGregor^94! has suggested that a translucent layer
of insulation be applied to the external surface of a solid wall so that the
heat loss from the wall is reduced while the solar gain is increased by being
transmitted and trapped by the translucent coating.
We may also expect to see insulation materials used structurally. The
Danish Zero Energy House, for example, used modified 300 mm rock-
Floors 83
Table 3.13. U-values for solid floors^
17-values (W/(m2 K))
Two exposed edges

Dimensions (m) Four exposed edges at right-angles
30 x 15 0.36 0.21
30 x 7.5 0.55 0.32
15 x 15 0.45 0.26
15 x 7.5 0.62 0.36
7.5 x 7.5 0.76 0.45
3x3 1.47 1.07
Fig. 3.25. The Frusher

Polyurethane Home.t96 1
(Courtesy of Building
Design.)
wool batts as the load-bearing wall element J 95 1 And for those who favour
'organic' architecture, the Frusher House in Wisconsin, shown in Fig. 3.25,
was constructed by foaming polyurethane around an inflated balloon used
as a mould. The 180 mm thick walls are covered with a fireproof rubber
skin.
3.9 Floors
For ground floors it is not possible to calculate {/-values in the same way
as for walls. In the case of solid floors in contact with the ground, the size
and edge conditions of the slab must be accounted for and in the case of
suspended floors allowance must be made for the higher temperatures
(compared to the outside air) in the enclosed air space under the floor,
which result from low ventilation rates in that space. Tables 3.13 and 3.14
give standard 17-values.
Fig. 3.26 shows a typical construction for a solid floor and Table 3.15
gives the effect on the U-value of increasing the insulation.
84 Building design
Table 3.14. U-values for suspended timber floors 1971
{/-values (W/(m:2 K))
Bare, or with With any surface

lino, plastics With carpet finish and 25 mm
Dimensions (m) or rubber or cork quilt over joists
30 x 15 0.39 0.38 0.30
30 x 7.5 0.57 0.55 0.39
15 x 15 0.45 0.44 0.33
15 x 7.5 0.61 0.59 0.40
7.5 x 7.5 0.68 0.65 0.43
3x3 1.05 0.99 0.56
Table 3.15. U-values for solid floors^
(/-values (W/(m2 K))

Thickness of mineral
fibre slab insulation Detached house Semidetached house
(mm) (8.9 m x 7.3 m) (5.5 mx 8.2m)
0 0.77 0.68
25 0.49 0.45
40 0.40 0.37
50 0.36 0.33
60 0.32 0.30
70 0.30 0.28
Similarly, Fig. 3.27 shows a typical construction for a suspended floor

and Table 3.16 gives the effect on the (/-value of increasing the insulation.
For suspended timber floors the BRE has stated that the safest course
is not to include a vapour barrier because of the possibility of spillage or
leakage from plumbing or heating installations.I971 If either of these
occurred, the limited possibilities for the water to escape might lead to
conditions conducive to fungal attack and eventual rotting.
Because of the difficulties of modifying the floor structure afterwards
we recommend designing for a {/-value of 0.2—0.3 W/(m2 K).
3.10 Windows
Glass, as has been noted, can 'roast you, freeze you, cut you and blind
you'. Nonetheless, its more attractive properties endear it to designers and
occupants alike. The characteristics of glass and other translucent materials
and their use in relation to passive solar gain are discussed in Chapter 4.
Here we will briefly address certain questions concerning windows.
The first is whether or not double glazing should be installed. In a
house built to 1981 Building Regulations Standards with single glazing,
approximately 28% of the total heat loss is through the windows (see Fig.
3.1) - with double glazing this is halved. Double glazing all the nation's
Windows 85
18
Table 3.16. U-values for suspended timber floors^ ' 1
(7-values (W/(m2 K))

Thickness of
mineral-fibre mat Detached house Semidetached house
(mm) (8.9 m x 7.3 m) (5.5 m X 8.2 m)
0 0.72 0.79
60 0.32 0.34
80 0.28 0.29
100 0.24 0.25
140 0.20 0.20
180 0.17 0.17
SOLID CONCRETE Fig. 3.26. Solid-floor con-

INSULATION DETACHED HOUSE SEMI DETACHED HOUSE
struction detail. 1 7 8 1
STANDARD 8.9m x 7.3m FLOOR AREA (65m2) 5.5m x 8.2m FLOOR AREA (45m3)
MINERAL FIBRE U VALUE U VALUE
SLABS W/m2 deg C W/m* deg C
FULL FLOOR
UNINSULATED 0.77 068
25 mm 0.49 0 45
40 mm 0.40 037
50 mm 0.36 0 33
60 mm 0.32 0.30
70 mm 0.30 0.28
25 mm PERIMETER
INSULATION
NOT LESS THAN
250 mm wide 0.72 0.64
500 mm wide 0.66 0.58
600 mm wide 0.64 0.56
750 mm wide 0.62 0.54
1000mmw.de 0.59 0.52
Semi detached house assumed to be narrow frontage
CONCRETE FLOOR INSULATION

MINERAL FIBRE SLAB
Cavity Brick
dwellings would save almost 1% of the national primary energy consump-

tion, according to BRE calculations J 45 1 But opinion varies as to whether
it is economical to do so. The BRE has found that only in the most favour-
able case of high future energy costs, combined with a new dwelling
heated to a high standard with good controls, is double glazing 'just about
cost -effective'. I451 It may be argued, however, that this is exactly what we
86 Building design
Fig. 3.27. Suspended-floor GROUND FLOORS
construction detail.!781 (1) SUSPENDED TrMBER
INSULATION STANDARD DETACHED HOUSE SEMI DETACHED HOUSE

MINERAL FIBRE 8.9m x 7 3m FLOOR AREA (65m 2 ) 5.5m x 8 2m FLOOR AREA (45m 2 )
MAT U VALUE U VALUE
W/m 2 deg C W/m 2 deg C
UNINSULATED 072 0.79

60 mm 0.32 0.34
80 mm 0.28 0.29
100 mm 0.24 0.25
140 mm 0.20 020
160mm 0 19 018
180 mm 0.17 0.17
200 mm 0.16 0 15
s 1) Insulation of Suspended Timber Floor assumed to use Plastic net technique and incorporat
2) Semi detached house assumed to be narrow frontage.
VENTILATED TIMBER GROUND FLOOR WITH MINERAL FIBRE MAT/PLASTIC NET
should be designing. Littler has found that double glazing is economical in

new houses heated with electricity and marginally so if gas or solid fuel is
used for heating.!981 At the Basildon homes described previously, double
glazing was not judged to be cost-effectiveJ43J
Most calculations on the cost-effectiveness of double glazing are based
only on the reduction in heat loss, but the numerous homeowners who
have purchased double glazing in recent years have often done so for
reasons of comfort, a less easily quantified aspect. Double glazing results
in fewer down draughts and higher surface temperatures of the glass facing
the room (approximately 4 - 7 K higher than that of single glazing for the
same internal and external conditions),! 31 thus increasing the comfort
level and the space which may be fully used in a room. In buildings where
windows are kept to a minimum area, comfort conditions are less affected
than in spaces where large glazed areas are used to take advantage of
passive solar gain. The choice is also related to how curtains are used. If
heavy, well-fitting curtains are drawn regularly at night the relative advan-
tage of double glazing is diminished. We recommend double glazing when-
ever circumstances permit and furthermore suggest that triple glazing be
considered in certain applications. Some of the triple-glazed windows com-
mercially available are of significantly higher quality without being much
more expensive than a number of double-glazed units.
Window quality is especially important in reducing infiltration (see
Table 4.5) but, unfortunately, it has often been sacrificed to cost.
Additionally, workmanship around windows often suffers because the
Doors 87
importance of weathertight construction is not appreciated. One solution
to the problem could be the introduction of kits for wall windows which
might include insulation, damp proofing, sills and lintels in order to bring
the quality of the window and its surrounding construction up to that of
some of the well-detailed roof windows now available. A first step towards
energy conservation for all windows, and one that has been found unani-
mously to be cost-effective is draught sealing with synthetic rubber strips,
felts, metal, plastics, wool or tufted polypropylene pile. It may also be
possible in some cases, for example when a home has a mechanical venti-
lation system, to reduce infiltration losses by opting for fixed windows,
but the possible energy savings must be carefully balanced against the loss
of flexibility. Window frames are commonly of timber, metal or plastic.
Table A 2.8 compares the overall (/-value of glass and frames for several
types (unfortunately, plastic frames were not available) and shows the
advantage of wooden frames.
Wooden frames are also somewhat less likely to serve as condensing
surfaces. In the past, designers often accepted that single glazing would act
as a condenser and the higher-quality frames incorporated a means of
draining out the condensate. Double glazing reduces the risk of conden-
sation on glass because, as has been noted, the temperature of the surface
exposed to the room is higher than with single glazing and is more likely
to be above the dew-point temperature of the room air. In insulated, well-
ventilated and heated buildings with proper consideration given to localiz-
ing moisture production it should be possible to avoid serious condensation
problems with double glazing, but this is not to say that they will not
occur, particularly if the ventilation rate is reduced to limit heat loss. The
BRE has developed a condensation prediction chart which helps to assess
the degree of risk.I 3'
Curtains and shutters for windows are discussed in Chapter 4.
3.11 Doors
As with windows, doors should be draught proofed. The use of glass in
external doors should generally be avoided because of the higher heat
losses entailed. Where glass is used, for example in patio doors, thermal
protection in the form of shutters or heavy curtains for night use should
be provided. Self-closing doors, provided they are acceptable to the occu-
pants, can help reduce ventilation losses and are becoming a feature of
low-energy homes.
A British Standard is in preparation which will aid designers to select
more-energy-efficient doors J " l It suggests three grades of resistance to
air penetration with the highest standard being a maximum air leakage of
12 m3 /(h m) of length of opening joint at a pressure differential of 150 Pa.
An additional concern with doors (particularly front doors) and to
some extent windows, too, is that warping may set up new ventilation
patterns which are unsatisfactory. In the Abertridwr Project referred to
previously some warping resulted in stagnant pools of air favouring con-
densation and mould growth.
Building design
Fig. 3.28 Pitched-roof ROOFS
construction detail.f781 PITCHED ROOF
UVALUE
INSULATION STANDARD W/m 2 deg C
MINERAL FIBRE MAT
UNINSULATED 2.01 1.97

60 mm 0.50 0.50
80 mm 0.40 0.40
100 mm 0.33 0.33
140 mm 0.25 0.25
160 mm 0.22 0.22
180 mm 0.20 0.20
200 mm 0.18 0.18
Note: Roof construction : tiles on battens on sarking felt v

roof space and horizontal plasterboard ceiling.
MINERAL FIBRE MAT IN PITCHED ROOFS
Mineral Fibre Mat
3.12 Ceilings and roofs

The roofs of both existing and new homes are often the easiest and most
cost-effective sites for insulation. Materials used are usually preformed
glass or mineral-fibre blankets and batts or loose-fill materials (often blown
into the loft to settle between the ceiling joists) made of glass fibre or
cellulose. Combustible products such as plastics and treated cellulose fibres
are being studied by the BRE in order to develop tests which will ensure
that such loft insulation materials meet fire safety performance require-
ments J 7 6 J
{/-values for insulated roofs are also being investigated by the BRE and
one result of the work is the realization that more information is needed
on air speeds in loft spaces.llo °l Fig. 3.28 shows a schematic of a typical
construction for a pitched roof (more care would be required in actual
detailing to avoid a cold bridge - see Fig. 3.18) and Table 3.17 gives the
effect on the {/-value of increasing insulation. (If 200 mm seems high, it
should be remembered that in Scandinavia it is not unusual to have
250 mm of mineral wool in the roof.)
Similarly, Fig. 3.29 shows a typical construction for a flat roof and
Table 3.18 gives the effect on the {/-value of increasing the insulation.
Vapour barriers are often used on the warm side of the insulation to
prevent moisture entering the wall from the interior of the building. The
risk of interstitial condensation may be assessed in a similar way to that
for walls (see Appendix 3) and, unfortunately, the difficulty of achieving
a satisfactory seal is also similar, so ventilating the roof space to the out-
Ceilings and roofs 89
Table 3.17. U-values for pitched roofs
Thickness of lvalues (W/(m2 K))
(mm) Pitch (30°) Pitch (22.5°)
0 2.01 1.97
60 0.50 0.50
80 0.40 0.40
100 0.33 0.33
140 0.25 0.25
180 0.20 0.20
200 0.18 0.18
a
Roof construction: tiles on battens on sarking felt with ventilated roof
space and horizontal plasterboard ceiling.
Table 3.18. U-values for flat roofsa^^
Thickness of
mineral fibre mat
(mm) 17-va
17-values (W/(m 2 K))
0 1.54
60 0.47
80 0.38
100 0.32
140 0.24
180 0.19
200 0.18
a
Roof construction: granite chips, three-layer felt, 25 mm
timber boarding, 9.5 mm plasterboard with a vapour
barrier.
side air, whenever possible, is wise. Ventilation through the soffit of pro-
jecting eaves on opposite sides of the building, as shown in Fig. 3.18, is
often a simple and effective solution. Special care is required with shallow
roof slopes, say of 25° or less since, here, any escaping vapour has a
smaller volume of air to mix with and high humidity levels may thus result
— purpose-made ventilation openings may be required.
A consequence of placing substantial insulation between the ceiling
joists will be a lowering of the loft temperature. Fig. 3.30 shows represen-
tative loft temperatures as a function of the landing temperature.
In severe weather it is apparent that the loft temperature could fall
below freezing, thus endangering cold-water tanks and pipes in the loft. If
lofts are sealed from the house (to reduce the ventilation heat loss) and
internal temperatures are lower (for example, when the occupants are
away) the risk of damage is greater. British Gas thus suggests that where
insulation levels are high the cold-water tank be located within the living
areaJ101!
The roof angle is particularly important if solar collectors are to be
90 Building design
Fig. 3.29. Flat-roof con- FLAT ROOF : INSULATION BELOW DECKING
struction detail.! 78 !
FLAT ROOF TIMBER JOIST
INSULATION STANDARO 25 mm 25 mm 50 mm
MINERAL FIBRE MAT TIMBER BOARDING CHIPBOARD WOOD WOOL
U VALUE U VALUE U VALUE
W/m 2 deg C W/m 2 degC W/m 2 deg C
UNINSULATED 1.54 1 60 0.93

60 mm 0.47 0.47 0.39
80 mm 0.38 038 0.32
100 mm 0.32 0.32 028
140 mm 0.24 0.24 0.22
160 mm 0.22 0.22 0.20
180 mm 0.19 0.20 0.18
200 mm 0.18 0.18 0.16
Roof construction : Granite chips, 3 layer fe timber boarding, chipboard or wood wool
and 9.5 mm plasterboard with a vapour barr
MINERAL FIBRE MAT UNDER BUILT UP ROOF
erboard vt
Vapour Barnf
Timber
Boarding
Fig. 3.30. The effect of

ambient and landing
temperatures on loft
temperatures (with 200 mm
glass fibre insulation in the
ceiling).U<>lJ
-12 -10 - 8 - 6 - 4 - 2 0 2 4 6 8 10 12
Ambient temperature°C (hourly average)
installed either at the time of construction or in the future. (We believe

that homes should be designed with potential collector area and roof slope
in mind. Normally, provision of a suitable south-facing area is not difficult
and the cost of increasing a shallow sloping roof to an acceptable angle at
the time of construction should not be very great - the increase in solar
References 91
Fig. 3.31. A 'green'
roof.I 1 0 4 ! Key: 1. local
turf; 2. peat; 3. plain faced
bituminous felt; 4. 16 mm
plywood decking; 5. firring
piece; 6. joists; 7. 100 mm
mineral-fibre insulation;
8. 50 mm mesh chicken
fifl M -8
-9
-10
wire; 9. 37 mm X 20 mm
softwood battens; 10. 500 g
polythene; 11. 9 mm
"11 medium density hardboard
(class 1).
collector efficiency can be significant. A secondary advantage of a greater

slope is that if appropriate rafters are used the loft space is more likely to
be available for extra living space at a later date.) Or if the loft space is to
be used as a source of preheated warm air a correctly sloped roof will
provide somewhat higher temperatures. In this case the roof colour should
be as dark as possible to increase the solar radiation absorbed. Experiments
are presently underway to evaluate the use of a heat pump combined with
very simple roof panels used as air collectors.! 1 0 2 l (See also Chapter 6.)
Another technique is the use of panels which serve both as solar collectors
and roofing elements. Active solar collection is discussed in detail in Chap-
ter 5.
A second aspect of the roof angle, which may affect the design of some
houses, is its effect on the pressure distribution around the house. For roof
slopes of less than 30° the wind-produced pressures are negative on the
windward surface, while for slopes exceeding 30° the pressures are
positive.!58!
Finally, two roof constructions to keep in mind are insulating ceilings
which require no roof cladding and 'green roofs'. The first are perhaps best
suited to industrial buildings. One example is a cold store in Yorkshire
which used structural insulation panels composed of polystyrene sand-
wiched between prestressed steel skins - a weather-resistant external
surface eliminated the need for cladding.!103!
'Green roofs' are those covered with vegetation, usually grass. Their
chief attraction is probably visual since present designs, as opposed to tra-
ditional ones, call for conventional insulation to be incorporated too to
reduce heat losses. Fig. 3.31 shows one design for a grass roof over hard-
wood joists.
References
[ 1 ] Le Corbusier (1972). Towards a New Architecture. London: The
Architectural Press.
[2] The Building (Second Amendment) Regulations 1981. Statutory
Instrument 1981 No. 1338. London: HMSO.
[3] Anon. (1972). Double Glazing and Double Windows. BRS Digest
140. Garston, Watford: BRS.
[4] Siviour, J.B. (1976). Designs for Low Energy Houses. ECRC/M922.
Chester: The Electricity Council.
92 Building design
[5] Siviour, J.B. (1977). Houses as Passive Solar Collectors. ECRC/
Ml070. Capenhurst, Chester: The Electricity Council Research
Centre.
[6] (1977). CIBS Guide, Section A 7. 'Casual Gains'. London:
Chartered Institution of Building Services.
[7] Siviour, J.B. (1977). Calculating Solar Heating and Free Heat and
Their Contribution to Space Heating in Buildings. Capenhurst,
Chester: The Electricity Council Research Centre.
[8] (1970). IHVE Guide Book A. London: Institute of Heating and
Ventilating Engineers.
[9] Nevrala, D.J. (1980). 'Heat services for housing - the insulated
house design requirements'. Studies in Energy Efficiency in Build-
ings, No. 2. British Gas, Watson House Research Station.
[10] (1980). CIBS Guide, Section A3. 'Thermal Properties of Building
Structures'. London: Chartered Institution of Building Services.
[11] Everett, A. (1970). Materials. London: Batsford.
[12] Burberry, P. (1974). 'The choices for designers'. Architects
Journal, 160 (37), 615-28.
[13] Anon. (1979). Glass and Offices. St Helens, Merseyside: Pilkington
Technical Advisory Service.
[14] Hermsen, J. (1979). 'Building study energy conscious design at
Basildon, Essex'. Architects Journal, 170 (41), 758-72.
[15] Anon. (1979). 'Brownhill R.oad Flats, Lewisham'. In: Buildings the
Key to Energy Conservation, Kasabov, G. (ed.). London: RIB A
Energy Group.
[16] Fordham, M. (1982). Max Fordham and Partners. Private com-
munication.
[17] Anon. (1979). 'Solar housing for the elderly'. In: Buildings, the Key
to Energy Conservation, Kasabov, G. (ed.). London: RIBA Energy
Group.
[18] Justin, B., Guy, A.G. & Shaw, G. (1980). 'Monitoring the thermal
performance of houses with solar walls'. Helios, 10, 1-5.
[19] Hitchin, E.R. (1980). 'The sizing of heating systems for well-
insulated houses'. Studies in Energy Efficiency in Buildings No. 2.
British Gas, Watson House Research Station.
[20] Humphreys, M.A. (1976). 'Desirable temperatures in dwellings'.
The Building Services Engineer, 44 (8), 176-80.
[21 ] Humphreys, M.A. (1979). 'The variation of comfortable tempera-
tures'. Energy Research, 3, 13-18.
[22] Cooper, I. (1981). University of Cambridge Martin Centre Con-
ference on Comfort and Energy Conservation in Buildings.
[23] (1978). CIBS Guide, Section Al. 'Environmental Criteria for
Design'. London: Chartered Institution of Building Services.
[24] Brundrett, G.W. (1973). Thermal Comfort in Buildings. Electricity
Council Research Centre, Capenhurst.
[25] Billington, N.S. (1974). Domestic Engineering Services. London:
IHVE.
[26] Mclntyre, D.A. (1977). 'Can humidification save on heating bills?'
Heating and Ventilating Engineer, 51 (596), 10-11.
[27] Durnin, J.V.G.A. & Passmore, R. (1976). Energy, Work and
Leisure. London: Heinemann.
[28] Gagge, A.P., Burton, A.C. & Bazett, H.C. (1941). 'A practical sys-
References 93
tern of units for the description of the heat exchange of man with
his environment'. Science, 94 (2445), 428-30.
[29] Nevrala, DJ. (1979). 'Some energy implications of thermal com-
fort criteria'. In: Studies in Energy Efficiency in Buildings.
London: British Gas.
[30] Mayo, A.M. & Nolan, J.P. (1964). 'Bioengineering and bioinstru-
mentation'. In: Bioastronautics, Schaefer, K.E. (ed.). New York:
Macmillan.
[31 ] Billington, N. (1978). 'Comfort'. Building Services, October, pp.
37-8.
[32] Anon. (1977). Ventilation Requirements. BRE Digest 206.
[33] Brundrett, G.W. (1976). Ventilation: A Behavioural Approach.
ECRC/M902. Chester: The Electricity Council.
[34] Fuller, W. (1981). 'What's in the air for tightly built houses'. Solar
Age, June, pp. 30-2.
[35] Brundrett, G. (1981). 'Air pollution'. Building Services, March, pp.
38-40.
[36] Modera, M.P., Sherman, M.H. & Grimsrud, D.T. (1981). 'Long-
term infiltration measurements in a full-scale test structure'. AIC
Conference on Building Design for Minimum Air Infiltration at the
Royal Institute of Technology, Stockholm, Sweden.
[37] Nevrala, DJ. & Etheridge, D.W. (1977). 'Natural ventilation in
well-insulated houses'. ICHMT Seminar. Heat Transfer in Buildings.
Dubrovnik.
[38] Etheridge, D.W. & Nevrala, D.J. (1979). 'Air infiltration and our
thermal environment'. Building Services and Environmental
Engineer, 1 (7), 10-13.
[39] Anon. (1956). Domestic Heating - Estimation of Seasonal Heat
Requirements and Fuel Consumption in Houses. BRS Digest 94.
London: HMSO.
[40] Guillaume, M., Ptacek, J., Warren, P.R. & Webb, B.C. (1978).
Measurement of Ventilation Rates in Houses with Natural and
Mechanical Ventilation Systems. Garston: Building Research
Establishment.
[41] Anon. (1969). Condensation. BRE Digest 110. Watford: Building
Research Station.
[42] Loudon, A.G. (1971). The Effects of Ventilation and Building
Design Factors on the Risk of Condensation and Mould Growth in
Dwellings. BRE CP 31/71. Watford: Building Research Station.
[43] Anon. (1979). 'Condensation and mould growth'. Domestic Energy
Note 4. Department of Energy.
[44] Anon. (1970). Condensation in Dwellings. Part 1. A Design Guide.
Part 2. Remedial Measures. Ministry of Public Buildings and Works.
London: HMSO.
[45] Bakke, P. (1975). Energy Conservation: A Study of Energy Con-
sumption in Buildings and Possible Means of Saving Energy in
Housing. BRE CP 56/75. Garston, Watford: BRE.
[46] Berry, J., Emerson, R., Harrison, J.W. & Kasabov. G. (1977). 'Con-
servation of energy in housing'. Building Services Engineer,45,288—93.
[47] Anon. (1981). 'Children's designs for a solar house'. What's New in
Building. March, 121.
94 Building design
[48] Kasabov, G. (ed.) (1979). Buildings, the Key to Energy Conser-
vation. London: RIB A Energy Group.
[49] Hawkes, D. & MacCormac, R. (1978). 'Office, farm, energy and
land use'. RIBA Journal, 85 (6), 246-8.
[50] Rodrigues, CM., Cole, RJ. & O'Sullivan, P.E. (1978). 'Computer-
ised thermal design models'. The Building Services Engineer, 2
(46), 19-27.
[51] Anon, (undated). Building Better with BEEP. The Electricity
Council-EC 3819.
[52] Anon. (1981). 'Guidelines for environmental design and fuel con-
servation in educational buildings'. Department of Education and
Science Architects and Building Branch Design, Note 17.
[53] Anon. (1979). How Windows Save Energy - Technical Appendix.
St Helens, Merseyside: Pilkington.
[54] Turrent, D., Doggart, J. & Ferraro, R. (1980). Passive Solar
Housing in the UK. London: Energy Conscious Design.
[55] University of Minnesota Project Ouroborus.
[56] Markus, T.A. & Morris, E.N. (1980). Buildings, Climate and
Energy. London: Pitman.
[57] Penz, F. & Thomas, R.B. (1979). Entry for the Competition 'The
Rational Use of Energy'. Unpublished.
[58] Mattingly, G.E. & Peters, E.F. (1977). 'Wind and trees. Air infil-
tration effects on energy in housing'. Journal of Industrial Aero-
dynamics, 2, 1 — 19.
[59] Hunt, D.R.G. (1979). Availability of Daylight. Garston, Watford:
BRE.
[60] Siviour, J.B. (1980). 'The effect of increasing house insulation on
electric heating equipment design'. IEE Conference on Effective
Use of Electricity in Buildings.
[61] Pickup, G.A. & Miles, A.J. (1979). 'The performance of domestic
wet heating systems in contemporary and future housing'. Studies
in Energy Efficiency in Buildings. British Gas Watson House
Research Station.
[62] Krall, L., Littler, J.G.F. & Thomas, R.B. (1979). 'Ambient Energy
Design House 1'. In: Ambient Energy Design. Cambridge: privately
published.
[63] Hayes, D.H. (1971). 'Higher standards of thermal and acoustic
insulation - but when?' IHVE Journal, 39, A39-A41.
[64] The Building Regulations 1976. Statutory Instrument 1970 No.
1681. London: HMSO.
[65] Campbell, P. & Bratt, J. (undated). Better Insulated Housing. DOE
Construction 21. Department of the Environment, Housing Devel-
opment Directorate.
[66] Anon. (1979). 'BRS low energy house laboratories'. In: Buildings,
the Key to Energy Conservation. Kasabov, G. (ed.). London: RIBA
Energy Group.
[67] Siviour, J.B. (1976). 'A new approach to space heating require-
ments in houses'. Electricity Council Research Centre. Energy
Management in Buildings Conference, Aston University.
[68] The Building (First Amendment) Regulations 1978. Statutory
Instrument 1978-723. Department of the Environment. London:
HMSO.
References 95
[69] Anon. (1980). 'Buildability of highly insulated housing'. BRE
News, 52, 17.
[70] Anon. (1982). 'Building Regulations Changes - Second Amend-
ment'. Architects Journal, 175 (13), 59-71.
[71 ] Barratt, P.V.L. (1979). 'Ways to meet the changed thermal insu-
lation regulations'. Building Services and Environmental Engineer,
1 (12), 28-31.
[72] Martin, D. (ed.) (1980). Specification 80. London: The Architec-
tural Press.
[73] Coates, J. & Sudjic, D. (1982). 'Toxic gas drives pupils out of class-
rooms'. The Sunday Times, 16 May, p. 4.
[74] Rogowski, B.F.W. (1976). Plastics in Buildings - Fire Problems
and Control BRE CP 39/76. Garston: Building Research Establish-
ment.
[75 ] Anon. (1979). Cellular Plastics for Building. BRE Digest 224.
Garston: Building Research Establishment.
[76] Rogowski, B. & Sutcliffe, R.J. (1980). Fire Performance of Loft
Insulating Materials. BRE IP 25/80. Garston: Building Research
Establishment.
[77] Venning, P. (1979). 'How to cost energy conservation in building'.
RIBA Journal, 86 (10), 446-7.
[78] Boys, J. (1979). 'Insulation'. The Architect. November, pp. 41-9.
[79] Potter, J.G. (1980). 'Assessing the economic priorities for improv-
ing building insulation'. Building Services and Environmental
Engineer, 3 (4), 28-30.
[80] Anon, (undated). 'Economic thickness of insulation for existing
industrial buildings'. Department of Energy Fuel Efficiency Book-
let 16.
[81 ] Ferraro, R. & Godoy, R. (1981). 'Thermal insulation'. What's New
in Buildings, June, pp. 47—8.
[82] Stein, R.G. & Serber, D. (1979). 'Energy required for building
construction'. In: Energy Conservation Through Building Design.
Watson, D. (ed.). New York: McGraw Hill.
[83] Herman, P.R. (1975). 'The energy cost of the construction and
habitation of timber frame housing'. Building Science Special
Supplement: Energy and Housing.
[84] Gartner, E.M. & Smith, M.A. (1976). Energy Costs of House Con-
struction. BRE CP 47/76. Garston: Building Research Establish-
ment.
[85] Thorogood, R.P. (1981). Vapour Diffusion Through Timber
Framed Walls. BRE IP 1/81. Garston: Building Research Establish-
ment.
[86] Farbon, J.W. (1981). 'Inside or out?' Building Services, 3 (6), 5 0 - 1 .
[87] Randell, J.E. & Boyle, J.M.A. (1979). 'Low energy housing -
Salford's answer'. Building Services and Environmental Engineer,
1 (10), 18-21.
[88] Wilson, D.R. (1977). 'Existing buildings - experience from the
"House for the Future" project of Granada TV. University of
Nottingham Conference on Ambient Energy and Building Design.
Welwyn: Construction Industries Conference Centre.
[89] Anon. (1979). 'Abertridwr Project (new)'. RIBA Journal, 86 (10),
450.
96 Building design
[90] ASHRAE Handbook (1977). Fundamentals. New York: American
Society of Heating, Refrigerating and Air Conditioning Engineers.
[91 ] Anon. (1978). 'Energy needs reduced to a minimum'. RIBA
Journal, 85 (6), 235-6.
[92] Hnilicka, M.P. (1961). 'Engineering aspects of heat transfer in
multilayer reflective insulation and performance of NRC insu-
lation'. In: Advances in Cryogenic Engineering. Timmerhaus, K.D.
(ed.). New York: Plenum Press.
[93] Littler, J.G.F. (1975). 'Multi-layer insulation evacuated or filled
with air or krypton'. Department of Architecture, University of
Cambridge. Autonomous Housing Study Working Paper 26.
[94] MacGregor, A.W.K. (1979). 'A solar skin for solid walls'. The
Passive Collection of Solar Energy in Buildings, UK-ISES Con-
ference C19.
[95] Esbensen, T.V. & Korsgaard, V. (1976). 'Dimensioning of the solar
heating system in the Zero Energy House in Denmark'. UK-ISES
Conference C12, pp. 64-77.
[96] Blythe, J. (1979). 'House of foam'. Building Design, February, pp.
26-7.
[97] Anon. (1972). Heat Losses Through Ground Floors. BRS Digest
145. Garston: Building Research Station.
[98] Littler, J.G.F. (1980). 'Low energy houses'. Ambient Energy, 1
(1), 57-62.
[99] Covington, S.A. (1981). The Selection of Doors and Doorsets by
Performance. BRE IP 2/81. Garston: Building Research Establish-
ment.
[100] Anderson, B.R. (1981). The Assessment of U-values for Insulated
Roofs. BRE IP 3/81. Garston: Building Research Establishment.
[ 101 ] Pickup, G.A. & Miles, A.J. (1979). 'Energy conservation - field
studies in domestic heating'. Studies in Energy Efficiency in Build-
ings. British Gas Watson House Research Station.
[102] Anon. (1981). 'Research in the UK - University of Aston in
Birmingham'. Helios, 11, 12-13.
[103] Anon. (1980). 'Insulation panels need no cladding'. What's New in
Building, October, p. 81.
[104] Ollis, J. (1977). 'Timber - available methods and systems'. The
Architect, December, pp. 59-62.
Passive solar design
4.1 Introduction
Most buildings receive sunlight via windows, and in the design of passively
heated buildings this warming effect is more fully exploited.
The five tasks to be accomplished in a successful design are as follows:
to increase the amount of sunlight entering the building by enlarg-
ing the glazed area;
to avoid causing excessive overheating or glare;
to reduce the large heat loss through the glazing;
to distribute the heat acquired;
to store the surplus for use when the sun stops shining.
There has been some sterile argument about terminology.' "Passive"
and "natural" solar designs' are terms coined to differentiate the use of
sunlight which provides warmth without the use of complicated controls,
pumps and fans, from 'active solar designs' which employ solar collectors
and fairly complex controls. The meaning of the terms will become more
clear whilst reading the descriptions of the built examples later in this and
the following chapters.
It is of interest to realize that active solar systems associated with very
large thermal storage could be used to provide all the space heating and
hot water in a building, even in the UK. On the other hand, passive designs
can never be expected to do this in northern Europe, and there they must
be regarded as fuel savers. In the US and other countries blessed with more-
intense and regular winter sunshine, it is possible in cold but clear regions
to rely entirely on passive solar space heating. Apart from the number of
sunny days in a particular region, there are other constraints acting on the
design team when dealing with passive solar buildings. For example, the
density in housing layout in northern Europe is higher than one might
find in many parts of the US and the elevation of the sun is lower, thus
overshadowing is a severe problem. This point is enhanced because passive
techniques usually rely on replacing walls rather than roofs with glazing
and, of course, the walls are more subject to shading by houses nearby.
Secondly, passive designs constrain the floor plan inside the building. This
feature should be compared with the effect of using active solar collection,
which can be regarded as an auxiliary boiler and which does not need to
influence internal layout. It may be imagined therefore that, in favourable
areas, a passive solar architecture will emerge, which sympathetically
97
98 Passive solar design
Fig. 4.1. Annexe to St.
George's School, Wallasey,
near Liverpool.! 2 1
resolves the problems introduced by the designs described later in this

chapter.
A building of great interest in this context is the Annexe to St George's
School at Wallasey (1961; Fig. 4.1) in the north-west of England. One of
the first in the world consciously to exploit passive gain, it illustrates
several of the tasks confronting the design team, led in this case by the
architect A.E. Morgan. The whole of the southerly wall is glazed (see Fig.
4.2 which shows the floor plan) - thus very large amounts of sunlight may
enter the fabric. The windows contain two panes of glass, thus roughly
halving the rate of heat loss through the enormous area of glazing (about
500 m2 J 1 * 2 ' For convenience, the opening windows are single glazed and,
of course, framing members take up roughly 10% of the facade. Overheat-
ing is reduced by the massive construction (ground and intermediate floors
are of 250 mm solid brick). The exposed surfaces of the heavy fabric
absorb energy and help to smooth out midday solar peaks. Overheating
and glare are reduced by the deep format of the window wall which
Introduction 99
Fig. 4.2. Ground floor and
section of Wallasey
School.(2]
stf*tf
-Double glazed
solar wall
metres
Double glazed
solar wall
— Outer glass
Inner glass
creates 1500 mm overhangs and, clea. y, these are more effective at the
appropriate times, when the sun is high in the sky. Heat distribution
occurs naturally since the glazing opens onto the heated spaces and a uni-
formly warm environment is achieved.!11 Teachers report that the stu-
dents, who sit facing parallel with the window wall, do not find glare a
problem.
In spite of the massive construction, the thermal mass contributes little
to heat storage over a 24 h period, and not at all over a period as long as a
month.
Auxiliary heat is supplied by the students and the lights, the latter
being switched on well before the school opens. It is notable that a build-
ing occupied from 9 a.m. until 4 p.m. is well suited to passive solar heat-
ing, although one in use from, say, 11 a.m. to 7 p.m. might be even more
successful.
The only complaints arising at Wallasey (where the back-up boiler sys-
tem has never been used) concern odours, which are a result of the low air-
change rate. The building is heavily insulated with 125 mm of polystyrene
applied externally (giving a {/-value of about 0.25 W/(m2 K) for the walls
and roof. Davies estimates the floor coefficient as 0.34 W/(m2 K)).
(U-values are defined in Appendix 2.)
It must be clear that the solar contribution expressed as a fraction of
the heating load of a building increases as more measures are taken to
reduce this load by insulation. It should also be clear that the absolute
solar contribution will actually fall with rising insulation, since the passive
energy (for a given glazed type and area) will more often exceed the
demand, resulting in a rejection of the surplus heat.
Having introduced the concept and outlined some of the problems
involved in buildings which deliberately use large amounts of passive solar
gain, the rest of this chapter will concentrate on features of the com-
ponents involved, on a description of the various building types which
have evolved and on built examples of these basic designs.
4.2 Elements of passive solar systems

4.2.1 Glazing
Exacting demands are made on the glazing systems used in passive
designs. Firstly, the skin must be weatherproof to driving rain and winds.
Secondly, it must be highly transparent to solar radiation but opaque to
long wavelengths characteristic of heat radiated from objects at 15—30 °C
(about 3—20 jum). Thirdly, it must have a reasonable lifetime (say ten
years) when warm and illuminated, and sufficient strength to withstand
wind loads and small blows. Clearly, in situations where people might fall
against the skin, the glazing must fail safely.
Finally the cost should be low, handling easy, and the appearance must
be pleasing since, almost by definition, the area will be large and the
impact great. Given these requirements it is often sensible to choose glass.
Table 4.1 lists some of the relevant properties.
The Glass and Glazing Federation^3] has issued an excellent book con-
cerning handling properties and installation, and Linsleyl4! gives some
very useful information concerning sealants and glazing compounds.
Passive-gain applications usually do not warrant the extra cost of the
more highly transparent water-white glass unless it is non-tempered,
especially in view of the higher dirt accumulation likely on vertical glazing
compared with sloping active solar collectors. The loss of transmission
from dust can be as high as 20%, but in areas which are not subject to large
amounts of industrial pollution, a more reasonable figure is 2-5%. Clearly,
it pays to wash down windows fairly frequently!
In the domestic situation, glass should not be continued down to
ground level. The Building Regulations indicate that below 800 mm above
Elements of passive solar systems 101
Table 4.1. Properties of glass
Rough cast, often
Material 'Window glass' Tempered6 Clear wired U-section Water-white
Examples of Pilkington Pilkington Brothers Pilkington e.g. Profilit by a,c

manufacturers Brothers Brothers Pilkington
Brothers
Thickness (mm) 3,4,6 3,4,6 6 6,7 3,4,5
Thermal expansion 6-9 6-9 6-9 6-9 6-9
coefficient
(10"6/K) (Steel 11-13, aluminium 22, timber 3-6)
Effect of heat Chipped edges Resists thermal shock.. May crack but does Cast edges less see tempered
may cause Breaks into small not disintegrate likely to be
cracks harmless pieces chipped
Ease of Little problem More difficult6 Little problem Fixed widths a
installation (262, 332 mm)
special glazing bars
% solar transmittance 3 mm 86 Depends on iron <80 80-85 89-92
at normal incidence 4 mm 84 content
6 mm 80
% transmittance of
heat radiation About 3 (all types)
Approximate price 9 21 15 14 ASG 5 mm,
in£/m 2 1981 $20/m2 in
(not installed) USA
a
Water-white glass by ASG (Kingsport, Tennessee, USA) is tempered.
b
Tempered and toughened are synonomous. Tempered glass must be ordered to the correct size and cannot subse-
quently be cut.
c
No European supply with a UK outlet has been found.
Table 4.2. Safe areas of low-level glazing
Area of glass Minimum thickness

(m 2 ) (mm)
0.2 3
0.5 4
0.8 5
2.5 6
the ground level (inside or outside the building) glazing becomes 'low-level'.
In the domestic situation, if annealed glass is used, it should have the
thickness indicated in Table 4.2. Toughened, wired or laminated panes
could be less substantial.
Thus, for a typical situation in which the glazing is set to match
600 mm stud spacing, a window 2 m high would have to be made with
6 mm glass. Toughened glass or plastic would not of course suffer this
restriction. Some passive collection designs benefit from a covering which
provides neither sharp reflections nor perfect vision. The appearance of
large areas of float glass over an opaque wall (see subsequent sections), or
Fig. 4.3.UseofKalwall
glazing in a roof. (Courtesy even an attached sun space, may be softened by a mildly textured surface.
of Kalwall Corporation, Cast glass provides such distortions.
Manchester, New Hampshire.)
Some designs go further and use diffusing glass, which has the advan-
tage of distributing the solar heat gain all over the thermally absorbing
room surfaces. The problem of glare can be made better or worse in this
situation since the whole of the window is now a source of high illumi-
nation. If the floor plan encourages occupants to sit and work so that side
illumination is provided, then daylighting is enhanced and ceiling reflec-
tion avoided, but if the users face the window wall, glare will be a problem.
Glass requires glazing compounds for its installation. Generally, in
passive applications the frame around the sheet will not rise above 50 °C.
But the temperature excursions over a Trombe wall may reach this value,
and putty will be unsatisfactory. Two-part rubberizing compounds, seal-
ants or preformed compression gaskets are especially suitable J 4 J Gener-
ally, cracking caused by thermal stress is not a problem in passive-gain
designs. As a matter of good practice, the depth of glass hidden by the
frame should be kept to the minimum for satisfactory mechanical per-
formance, to avoid a high temperature gradient at the edges - where
feathering cracks produced during cutting may develop into major cracks.
Table 4.3 illustrates some of the properties of plastics. Plastics have
some strong advantages and penalties. The cost (1981) of inexpensive
glazing bars for 3 mm glass in greenhouses is about £3/linear m. The maxi-
mum area of 3 mm glass recommended for vertical use (aspect ratio 1:3)
is about 1 m 2 , implying about £5/m2 for the glazing bars. However, some
plastics can be screwed down (Fig. 4.3) under a wooden batten placed
over the lapped sheets and a strip of glazing compound sealant.
A second benefit is that the plastic will not break into dangerous shards
either during handling or when installed. Additionally, its lightweight
nature (for example, ^ 1 kg/m2 for Kalwall) may allow cheaper structures
to be designed and certainly makes for easier installation. There are no
regulations concerning the size of plastic glazing in the domestic situation,
but some of the materials present a fire hazard.
There are other severe drawbacks to the use of plastics. In general, they
age, and although one might expect the rigid glass fibre or polycarbonates
to survive for 20 years, most of the other materials will either be totally
Fig. 4.4. Kalwall glazing to
the 'Ark' on Prince Edward
Island. (Courtesy of Kalwall
Corporation, Manchester,
New Hampshire.)
disfigured by scratches, or will have blown away! They also lose trans-
mission with age and thermal stress. The best product is probably rigid
glass fibre which at 150 °C loses 9% of its transmission after 300 hours.
Most plastics become yellow with age by ultra-violet action, although
Filon (glass fibre protected with proprietary films, which also increase
resistance to erosion) is believed not to suffer in this way. Those plastics
which really are rigid (for example, polycarbonate) must be glazed in with
great care - allowing a wide rebate (for example, 12 mm) to avoid panels
slipping out under wind flexing, and yet to allow for thermal expansion.
Even polycarbonate will bow in and out according to the humidity and
temperature, and flexible covers such as Kalwall tend to present a wavy
appearance which worsens as thermal cycling 'works' the anchoring points.
In this case, the material can be stretched onto an aluminium frame which
has roughly the same coefficient of expansion. Of course, this introduces
another expansion problem! Plastic films, which often have large expan-
sion coefficients, tend to sag when warm. Plastics have a different appear-
ance from glass. This point is strongest with polythene and glass fibre. The
latter contains both glass and resin and the different refractive indices lend
a bluish cast to the material, rendering it translucent not transparent. For
passive solar applications not involving areas of glazing which people look
through, this can be an advantage. Finally it is important to notice that the
plastics (and particularly the films) transmit more heat radiation than does
glass. Fig. 4.4 illustrates a large roof glazed with glass fibre, which is also
employed in the Trombe wall shown in Fig. 4.47.
4.2.2 Single and multiple glazing

Multiple glazing is used to reduce heat loss and noise transmission.
For the latter, a wide space is needed (for example, 100 mm); but, for the
former, gaps of 15-25 mm are suitable.
It is useful to consider why several layers of glazing can reduce heat
loss. Heat passing through a material is usually quoted in terms of a
{/•value, which has the units W/(m2 K).
This, for example, means that a single pane of glass 1 m2 with U —
Table 4.3. Properties of plastics
Ridged
Material Polycarbonate Fibreglass polyvinyl Acrylic
Example of manufacturer or trade Rohme & Haas: Kalwall: Filon British ICI: Perspex
name Makrolon Industrial
Plastics: Cob ex
Thickness (mm) 2-5 0.5-1.5 0.15-3.0 3,5,6
Surface spread of flame rating 1-2 1 1 3
(glass = 0)
Thermal expansion (lO^/K) 6.5 15 5-7 5-9
Effect of heat Maximum No mechanical Softens at Softens at
temperature degradation up 80°C 90°C,depoly-
130°C to 180°C merises at
200°C
Ease of installation Requires flexible Maximum span Can be solvent Can be solvent
sealant to 900 mm for welded. Corru- welded
glazing bar 1.5 mm. gated, can be
Nail down nailed down
% solar transmission 85 85-90 77 90
(3 mm) (2 mm) (2 mm)
% transmittance of heat radiation 3 5 Very little. 5
(0.5 mm) None beyond
Approximate price in £/m2 1981 30 8 18 17

(not installed) (3 mm) (3 mm) (3 mm)
Comments: all can be cut with Unbreakable. Standard width Very brittle Can be molded
scissors or a fine saw Appearance 'poor' rolls 2', 3', 4' below 0°C in an oven.
after 6 years or 5' by up to Appearance
exposure 50' maintained over
7 years
5.6 W/(m2 K), situated in a window with temperatures of 20 °C inside and

5 °C outside, loses heat at the rate of: (5.6 W/(m2 K)) (20 - 5) =
84 W/m2. If the glass has an area of 2 m2 then the rate of heat loss is:
(84W/m 2 )(2m 2 ) = 168 W.
Thus, over a period of 12 h, the amount of heat loss is:
(168 W) (12 h) = 2016 Wh or 2.016 kWh (= 3.60 MJ),
which is equivalent to a 'one-bar' electric fire burning for 2 hours!

The {/-value is essentially a measure of conductance rather than resist-
ance (R) and the two are related by:
clearly, the units of R are m2 K/W.

The resistance to heat flow of the window is composed of several com-
ponents illustrated in Fig. 4.5.
Film or flexible sheet
Polythene, PVC,UV Polyvinyl Rigid

UV inhibited inhibited Polyester fluoride polystyrene
- - Mylar Tedlar 3M: Amariplastics

Flexiguard
0.075-0.2 0.025-0.2 0.08 - 2,3,4
16-18 4 18-25
Collapses _ _ Retains mech- _
anical properties
from-35°C to
150°C
Does not shrink, Cracks easily on
thus easier than sawing
Tedlar
Staple-or-nail under framing strip
85-90 85-90 87 91
(0.13 mm)
up to 85 — 30 50 > 90 when —
(0.13 mm) anti-reflected
* 0.1-0.5 4 «l-3 *5-12 - 10

(0.04 mm) (0.04 mm) (3 mm)
Ages rapidly Softening Very strong Extremely Very clear and
temperatures strong non-distorting
low
Typical values of these resistances are 0.045 m2 K/W (rt), 0.005 m2

K/W (r2 for 3 mm glass), 0.13 m2 K/W (r 3 ) (see also Appendix 2). Thus
the total resistance to heat flow (rl 4- r2 + r 3 ) is about 0.18 m2 K/W, and
the (/-value is about 5.6 W/(m2 K).
When two panes of glass are used we have the situation shown in Fig.
4.6. We now have, instead of two film resistances, four plus the stationary
trapped air, giving a {/-value of 2.8 W/(m2 K). Values of the resistances
cannot be ascribed so simply in this case, since there is now radiation
interchange between the glazing layers, which short circuits the con-
ductive and convective resistances.!6J Similarly, three layers of glass pro-
vide a (/-value of about 1.9 W/(m2 K).
Clearly, the resistance to heat flow of a window is more or less indepen-
dent of the thickness of the glazing material since the resistance of the
glass or plastic is small compared with the film resistances. The other
important observation concerns the wind-cooling effect. Roughly speak-
ing, the film resistance R varies with the wind speed Kand the tempera-
Inside Outside
Room air cooled by

glass creates air
movement
Stagnant internal film r3 •:•
Glass or plastic with low resistance r2

Fig. 4.5. Thermal resistances
on a window.
ture
/F_V-0.8 ,T v, - 0 . 7
where To and Vo are the initial values. For example, if Ro is 0.045 at a

wind speed of 3 m/s, at V = 6 m/s it becomes:
(f) (£
/ 6 \0-8 / 7\
or
* 6 m/s = 0.026 m 2 K/W.
Care must be taken with the relationship, which applies only to turbulent
flow. Turbulence depends on the wind speed and window detailing and
will be strongly affected by the shelter from neighbouring trees and so on.
Roughly, one might expect laminar flow from 0 - 3 m/s and turbulent
flow above 3 m/s. In the laminar regime the film resistance will be higher,
decreasing rather suddenly at the transition to turbulent flow. Shelter is
thus significant!
The gap between the glazing layers must be large enough to prevent
short circuiting by conduction across a narrow air gap or through small
convection cells, but not so wide that gross circulation is set up. Fig. 4.7
illustrates this effect and shows that a gap of 15-25 mm is optimal. Of
course, any breaks in the seal between the layers, which allow air to pass
in and out of the interpane space, will reduce the effectiveness of the
multiple glazing.
Radiation was mentioned earlier. Two panes of glass or plastic at dif-
Inside Outside Fig. 4.6. Thermal resistances
on a double-glazed window.
U Fig. 4.7. Thermal resistance

W/(m2K) of a double-glazed window
3.5- as a function of glazing
separation.
3.3-
3.1-
2.9-
2.7-
2.5
11 13 15 100 mm
ferent temperatures exchange heat by heat radiation (infra-red). Coatings

can be applied which are more or less transparent to solar radiation (in
other words they cannot be readily 'seen') but which look like metallic
mirrors to infra-red. Thus these coatings reflect heat back into a room if a
window is so coated on the inside. By the same token, when facing out-
wards they radiate heat much less well than uncoated glass, thus reducing
heat loss. The optimal position on single glazing is facing inwards, and on
double glazing on the outer surface of the inner pane (or on plastic
between the panes). Since single glazing thus treated will be cooler, there
is an added risk of condensation. This can be particularly serious in green-
houses.
Coated glass gives the {/-values shown in Table 4.4.
Double glazing could be improved even further by filling the interpane
space with krypton (an inert gas) when the C/-value drops from 2.8 to 2.6
for the uncoated case, or from 1.8 to 1.0 in the coated case. (If double-
a 1
Table 4.4. U-values of windows *
Glazing material Coating */=W/(m2 K)

Single glass None 5.6
Single plastic None 6.6
Single glass Coated 3.3
Single plastic Coated 3.5
Double glass No coatings 2.8
Double plastic No coatings 4.3
Double glass Coated c 1.8
Double plastic Coated c 1.8
Double glass + interpane plastic No coatings 2.0
Double glass + interpane plastic Plastic coated 1.2
3M Quadpane^ Anti-re fleeted 1.5
a
Taken from Ref. [6].
b
NB: these values are approximate, depend on conditions and are open to
dispute.
c
Coated on surface 2 counting from the inside.
<*Data from M. Rubin, Lawrence Berkeley Labs., 1983.
Table A.S. Infiltration rates through windows or door cracks in ms/(h m)

of crack
Wind speed (km/h)
Element Condition 8 16 24 32
Double-hung-sash wood Average, not 0.7 2 3-6 5.5
windows weather stripped
Average, weather 0.4 1.2 2.2 3.3
stripped
Badly fitting 2.5 6.4 10 14
not weather stripped
Badly fitting 0.6 1.8 3.2 4.7
weather stripped
Between frame and wall Not caulked 0.3 0.7 1.3 1.9
Caulked 0.1 0.2 0.3 0.4
Metal casement Unstripped 1.3 3 5 7
glazed coated windows could be evacuated, the (/-value could drop to

0.5 W/(m2 K).)
In 1982 Pilkington Brothers commenced production of sealed double-
glazed units fitted with coated glass (called 'Kappafloat'). The unit has a
(/-value of 2.0 W/(m2 K).
Air leaks around the window closure or frame can provide a serious
heat loss. Typical values are shown in Table 4.5. Clearly, in passive solar
gain applications it is essential to avoid ill-fitting windows. Most should be
Table 4.6. Conditions used in calculating the results shown in Table 4.7
(a) [/-value single-glazed unshuttered window 5.6 W/(m2 K)

(b) [/-value double-glazed unshuttered window 2.8 W/(m2 K)
(c) [/-value coated double-glazed unshuttered window 1.2 W/(m2 K)
(d) [/-value single-glazed shuttered window 0.53 W/(m2 K)
(e) [/-value double-glazed shuttered window 0.49 W/(m2 K)
(f) [/-value single-glazed curtained window 3.9 W/(m2 K)
(g) [/-value coated (heat mirror) double-glazed shuttered window
0.40 W/(m2 K). Shutters consisting of 50 mm polystyrene 15 mm
from the glass, and used at night (13 h)
Ambient temperature 5°C (day), 2°C (night)
Internal temperature 20°C (day), 16°C (night)
Glazed area 30 m2
Shutters in place at night only.
Table 4.7. Heat loss with different window

treatments
Heat loss over 24 h period

Case (MJ)
(a) 210
(b) 105
(c) 47
(d) 112
(e) 61
(f) 175
(g) 29
fixed closed and, in general, the large area of glass is held in a frame which
only encounters the masonry at the perimeter of the window and not
around every pane of glass! However, a typical installation comprising
30 m2 of glazing might have a perimeter length of 23 m and opening
sections of 3 m2 with a length round the frame of 7 m. Thus the total
crack length could be 30 m and, where care has been taken, the infiltration
at a wind speed of 16 km/h (typical of the UK) would be 13 m 3 /h, corre-
sponding to a heat loss in the winter of about 70 W.
4.2.3 Insulating blinds and shutters

We have seen that the heat loss through glazing can be high — more
than 5.6 W/(m2 K) in exposed situations for single glazing. For example,
30 m2 of single glass would typically (in the UK) lose about 200 MJ/d in
winter if unprotected at night. Tightly fitting shutters 50 mm thick used
at night could reduce this loss to about 100 MJ/d.
Table 4.7 summarizes what happens on a particular day with different
window treatments. The conditions are listed in Table 4.6.
It is important to place these losses in context. A reasonably well-
insulated house of 100 m2 floor area might, on such a winter's day, endure
a heat loss of 270 MJ. The solar gain through an unobstructed south-facing
window of 30 m2 might be 108 MJ on a January day in the southern UK.
Reference to Table 4.7 indicates that large areas of glazing could be a
source of heat loss rather than heat gain, unless defensive measures are
taken.
Later sections of this chapter illustrate various designs adopted for
using passive gain. Some of these involve attaching to the building glazed
spaces which can be isolated under adverse conditions. In those cases, the
windows may need no further treatment;however, when spaces are occu-
pied under all weather conditions, it is essential that the glazing be pro-
tected at night. It is likely that in these cases curtains will be used. There is
very little information concerning the insulating effect of curtains because
they are ill defined and their manner of attachment is very important. If,
indeed, a stationary layer of air were trapped between the drape and the
glass, the extra resistance imposed would be about 0.2 m2 K/W. Con-
ventionally it is said that well-fitting curtains reduce the heat loss through
single glazing by 20%, corresponding to a resistance of 0.08 m2 K/W. The
energy saved (compare cases (a) and (f) in Table 4.7), is small but the cost
of curtains made to measure, plus the track, self-installed, is £10-20/m 2
(1980). The cost of effective shutters or blinds is £15-40/m 2 .
Types of window insulation

Glazing can be protected on the outside, or between the panes of
multiple units, or on the inside. External insulation implies systems which
can withstand winds, whose sliding tracks, operating mechanisms, or
hinges do not freeze up in winter, which can withstand ultra-violet radi-
ation from the sun and which do not ruin the external appearance. Inter-
pane systems can fail mechanically and inaccessibly, but they are pro-
tected from tampering, and can solve the 'storage problem'.
Internal shutters
The simplest form of insulation is a roller blind which when installed
to leave a gap at the window head of less than 25 mm, no gap at the sill,
and less than 5 mm at the sides, reduces the heat loss by 30%. Such a
blind, when aluminized, conserves 45% of the normal heat loss. Fig. 4.8
illustrates an installation suggested by Cuchanan and GresseJ7! The cost
of roller blinds is less than £10/m2 but minor modifications must be made
to the window reveal to incorporate the vertical track which provides the
edge seals. These are shown in Fig. 4.8. (Successful blinds can be made
from cotton onto which is stuck aluminized Mylar. The cotton should be
made more rigid with a fabric stiffener.)
The next step is to use quilted material for the blind.!8! For example,
the 'Window Quilt'I91 (Fig. 4.9), which consists internally of three layers
of polyesterfibre fill and externally of decorative colours, is said to have a
{/-value of 1.2 W/(m2 K), and should reduce the heat loss through single
glazing from 5.6 W/(m2 K) to 1 W/(m2 K). The reduction is thus
4.6 W/(m2 K); if the Window Quilt were installed over double glazing, the
drop would be from 2.8 W/(m2 K) to 0.8 W/(m2 K) or only 2 W/(m2 K).
Fig. 4.8. Sealing the blind to
the window frame. (Reprinted
from Popular Science, with
permission © 1979, Times
Mirror Magazines Inc.)
V ¥ *"
Section
through
track
Two reversible end brackets
1L, f made of 18 gauge galvanized steel,

house the roller system.
"^ Air seal

at
the sides
The seat roller The edge bead
presses the quilt slides through track
against the upper shaped to contain
window trim to seal. the bead snuggly.
A plastic edge guide snaps
into the bracket to guide edge
bead into the track.
Window Quilt has five ultrasonically

bonded layers: inner and outer decorative
polyester fabric; two layers of polyester
batting, plus an aluminized film vapour
barrier in center to prevent condensation
The quilt is a five- Soft foam seals and reflect heat. Ultrasonic bonding pro-
layer sandwich. against sill cess does not penetrate vapour barrier.
the outside layers bottom batten snaps
are decorative. open to remove quilt
for cleaning Fig. 4.9. Window quilt fitted
into extruded plastic
track.[81 (Reprinted from
This observation is of course general: insulation over double glazing has Popular Science, with
less effect than over single glazing. permission © 1979, Times
An even better roll shade is made from four layers of metallized Mirror Magazines Inc.)
fabrid 10 ) (Fig. 4.10) and has a lvalue of « 0.5-0.6 W/(m2 K). This
device shares with the 'High "R" shade' (Fig. 4.11) the ability to roll up
flat (saving storage space) and to open up its layers by admitting air when
covering the window. The High 'R' shade over a double-glazed window
End plug removable
for motor servicing
Curtain cover tube
6" O.D. for curtain drop to9'0"
S"O.D. for curtain drop to 18'0"
Self-inflating sandwich
Smooth, levet surface required

Curtain wall features automatic operation: at bottom, adjustable limit switch
Electric motor inside roller tube is con- n tube sets bottom of curtain drop
trolled by thermostat and Freon sensor.
Latter is enclosed in glass tube mounted
outside house. Four layers of aluminized
plastic film inflate and deflate via vents.
Price of 10-by-20 ft. shade: about $4.25
per ft 2 .
Multiple curtains
Fig. 4.10. Curtinwall.U 0 ! Curtains may be
mounted end to end
(Reprinted from Popular to achieve lengths over 20'0"
Science, with permission © up to 4 curtains may
weight tube be run from one controller
1979, Times Mirror
Magazine Inc.)
shows a measured thermal transmittance of 0.33 W/(m2 K), indicating a
transmittance for the shade of about 0.4 W/(m2 K). A two-layer low
emissivity blind (« £30/m 2 ), developed by the Energy Design Group,
reduces the {/-value of a single-glazed window to a measured 1.1 W/(m2 K)
and has very effective edge seals.l53!
Blinds such as these can readily be automated to respond either to sun-
light or temperature differences. Obviously, they can also help to keep
out sunshine if necessary.
The next type of internal insulation consists of a slab of expanded
plastic foam (Fig. 4.12), which can be covered with fabric and either
pushed into place in the window reveal, or stuck between fabric forming a
rigid sheet. This is hinged like a casement, attached to the ceiling, or
allowed to slide in horizontal track. To improve on the thermal perform-
ance of the blinds described above, the foam would need to be 70 mm
thick (expanded polystyrene) or 50 mm (polyurethane) giving an overall
thermal transmittance of about 0.3 W/(m2 K) when installed over a single-
glazed window. The success of the barrier will depend primarily on the
seals to the reveal, (inwards) particularly at the sides and sill and, second-
arily, on the seals (outwards) around the window itself.
Such large moveable panels (which should be sheathed in fire-resistant
material), are awkward to stow away. When hinged, the swing negates a
2-part extruded plastic Fig. 4.11. High 'R'
- head frame for easy
access. Made with
ShadeJlO]
integral head seals.
Compact single roll
design with strong
spring return
5 layers expand to
r form dead air spaces
that slow corrective
and conductive energy
transfer
- Radiant energy flow
greatly reduced with
low emmittance materials
- Permanently shaped
spacers conform tightly
when rolled up yet
separate the layers
when pulled down
- Extruded plastic jamb
frames with integral
jamb seal
Thermally effective
summer through winter
at windows and sliding
glass doors
Fig. 4.12. Fold-up casement

and push up internal
shutters.lH] (Reprinted
from Home Energy for the
80 's, with permission,
Garden Way Publishing.)
Passive solar design
Fig. 4.13. Sliding shutters
in Termorac House.! *2 1 L
Fig. 4.14. Thermoblind.[l3j

large amount of floor space. An alternative method (used in the Termoroc
House in Malmo, Sweden) used many parallel strips of polyurethane
measuring about 50 mm X 50 mm stuck between fabric, so forming a
flexible sheet. This sheet slides in track mounted on either side of the
window which extends in a curve onto the ceiling for 'parking' the shutter
when not in use (Fig. 4.13).
Thermoblindl13! solve the stowage problem by a concertina action (see
Fig. 4,14). The cost is about £30/m2 and the lvalue of the 20 mm
version over single glazing is about 1 W/(m2 K).
A still more convenient system is shown in Fig. 4.15 in which foam-
filled louvres act like a very deep Venetian blind. This type of system is
also used in Guardia shutters^10! (Fig. 4.16), consisting of air-filled
plastic extrusions which can be drawn back to the sides of the window.
The t/-value of a single-glazed window covered with such shutters is about
1.6 W/(m2 K) and their cost is about £40/m 2 . They also provide a deter-
rent to burglars!
Shurcliffel14! gives a very useful review of innovative shutter designs.
Fig. 4.15. Adjustable insulat-
ing louvres, f 11 1 (Reprinted External shutters
from Home Energy for the
80 's, with permission, Anderson^15 J has used huge panels of styrofoam hung from steel
Garden Way Publishing.) track outside patio windows. It is difficult to obtain a good seal to the
Fig. 4.16. Guaidia
WINDOW OR GLASS DOOR Shutters.! 13)
92mm
the existing recess is less,
Guardia provides a special fit-
ting service.
a) Minimum distance, edge of tra( ; to outmost projection on window

or door frame - 58mm.
b) Width of track - 34mm.
c) Inner length of louvre - 68mm. ) Total length of
d) Windowside length of louvre - 65mm. ) louvre - 133mm.
e) Thickness of louvre - 18mm maximum.
0 Louvre overlap - minimum 35mm.
maximum 50nm.
g) Number of louvres per installation - varies from 8 to 10 per metre
run.
h) Angle of rotation of louvres - approx. 150 degrees.
Fig. 4.17. Zero Energy

House 1975.
walls of the house or the window frames in this manner, and unless the
lower edge of the shutter has a deep-pile draught seal or slides in a second
track, the performance will not be very good.
The Zero Energy House!16! uses sliding shutters (Fig. 4.17) made of
glass fibre boxes containing plastic foam.
The heat loss is reduced from 3.1 W/(m2 K) (the value quoted by the
Danish team for double glazing) to 0.4 W/(m2 K). These shutters are
operated remotely from inside the house via a window mechanism which
adds to their cost.
External shutters have not been widely used for insulation in solar
houses; but 'rolladen' - rolling blinds (Fig. 4.18) consisting of narrow
overlapping strips of hollow plastic or aluminium — are common in West
Germany as a protection against forcible entry. In the Philips^17! Experi-
mental Solar House at Aachen, such shutters are used. Unless tightly fit-
ting they cannot be very effective and in order to achieve a reasonable
storage volume when rolled up at the head of the window, the thickness is
limited to 10-20 mm.
Interpane blinds
The third method for insulating glazing uses Venetian blinds between
panes of double glazing (for example, many Velux roof lights are so
equipped).
The blind can be held with the slats at any angle but does not usually
retract. The Massachusetts Institute of Technology (MIT) Solar V
Home!18! (Fig. 4.30) uses narrow lamellae about 15 mm wide with a
mirror finish on one surface. At night the louvres are closed to provide an
interpane metallized sheet which reflects back the infra-red. Work by
Berladl19' indicates that similar windows 100 mm wide achieve {/-values
of 1.15W/(m2 K).
4.2.4 Shading
Especially in the case of direct solar gain to occupied spaces, it may
be necessary to keep out the sun on occasions. Typical methods include
permanent overhangs, or fixed, angled slats over the windows and mobile
blinds or Venetian shutters.
The south-facing Terry House (in New Mexico) illustrated later in this
chapter (Fig. 4.31), uses wooden boards lined up east—west in echelon
over the windows, as shown in Fig. 4.19.
In northern latitudes it is probably sensible to use a mobile device since
admitting sunshine is usually desirable. The Zero Energy House employs
external awnings which wind down from the window head.
If the aim is to exclude glare, but not heat, then an internal blind may
Fig. 4.18. Rolladen rolling be satisfactory - once the sunlight has been absorbed by the blind it will
shutters.! 11 1
mostly appear in the room as heat. Only if the blind is white or shiny
metallic will a large part of the solar energy be reflected out again. For this
reason, external awnings are more successful in combatting overheating.
The metallized Venetian blinds in the MIT V Home referred to earlier
reflect the sunlight onto the ceiling of each room, thus avoiding glare.
When overhangs are preferred they tend to be rather large. If sun at an
elevation of 50° is to be allowed to shine on only 25% of the window indi-
cated in Fig. 4.20, then the overhang will be about 1 m.
An excellent survey of shading devices incorporated in aesthetically
pleasing ways is given by Olgyay & OlgyayJ 20 !
Finally, an interesting device called 'cloud gel' has been invented by
ChahroudU211 It consists of a gel sealed between double glazing and is
Fig. 4.19. Sun shading out-
side the Terry House, New
Mexico, US.
. Low sun dm, V

^^ winter
£ xs-8 /
w 5 i/
L Fig. 4.20. Overhangs in

summer.
Fig. 4.21. External reflec-

-hi
1m ""---. tion enhancement in winter
- A— •
)°
-Jl ^ 3 m X 3m
normally transparent. Above about 30 °C it becomes milky white and

opaque, thus excluding unwanted radiation.
4.2.5 Radiation enhancement using reflectors

If the horizontal surface outside a window is made reflective, the
amount of radiation entering the building is increased. However, for
passive solar gain applications, one is primarily interested in the heating
season when the sun is low in the sky. Fig. 4.21 indicates that rather large
areas of reflector would be needed.
Since much of the radiation is diffuse, and the reflectors are not per-
fectly specular, the situation is much more complex than indicated in Fig.
4.21. Work by Baker, McDaniels & Kaehnl 22! shows that, for a specular
reflector with a reflectance of 0.8, the optimum tilt of the reflector is
about 7° downwards (that is, sloping down away from the building) and
that increasing the length beyond about twice the window height is of
small benefit (Fig. 4.22). (This presupposes a configuration in which the
reflector starts at the same level as the window.) The width required is
shown in Fig. 4.23 and indicates little gain beyond a width of four times
that of the window. The maximum enhancement is shown in Fig. 4.24 for
various months and suggests that in the winter months about 50% more
radiation would be directed at the window using a reflector of four times
the width of the window with a length of twice the height of the window.
In the UK the enhancement will be reduced because of our predomi-
Fig. 4.22. Enhancement at i i 1 1 1 1 1 1
45° N with reflector .8 - -
length.!221 Enhancement in _
1.7 _ Bn- 10
light gathered for the
90 ^ — —"
reflector-collector system at
. 2
•
45° N latitude as a function 1.6 - —-

\
of reflector length. Calcu- \ r/
lations were performed for \ /
the month of January for ) / =5
three representative choices 4 - ' /

r— Bj = 80 -
of reflector and collector
orientation angles. 13 /// LATITUDE-45° N _
0 DOUBLE GLAZING
/ / = 80
12 P- 0.8
///
11 r -
i i | 1 1 J I |
2.0 2.5 3.0 3.5
R/L
Fig. 4.23. Enhancement with 1.6 1 1 i i i

reflector width.[22] . . QDEC
Enhancement in light ^^^°^" JAN -
gathered plotted as a func- 1.5 -
tion of total reflector width
for the months of December, A _
January and February at 1.4 -
45° N latitude.
1.3 -
2
<
I
LATITUDE-45°N
1.2 -
^T =80°, ^ R S 5 °
; / DOUBLE GLAZING
£• = 0.8
I I 1
2 4 6 8 10
W/L
nantly diffuse radiation. McDaniels et al. quote a figure of « 1.25—1.30

for enhancement of diffuse radiation with the reflector dimensions above.
Thus one would expect an overall gain in our climate, not of 50% but of
« 35-40%.
4.2.6 Thermal storage (see also Chapter 7)

The problem of thermal storage in passive solar design has bedevilled
workers in this area from the beginning. It is plain that in clear climates
such as the Pyrenees or New Mexico, frequent winter sun is available in
quantities large enough to supply the major part of the space heating load
of a building, and is regular enough to do this reliably. Thermal storage is
essential to carry over heat acquired during the day until the sun again
becomes powerful the following day, or, at worst, the next-but-one day.
The situation is very different in the UK. For example, the ratio of solar
radiation falling on south-facing vertical surfaces in the southern UK to
north-eastern US in December is roughly 1:3. The percentage of possible
sunny hours in the UK is about 35%, in New Mexico 80% and in Nice 60%.
Fig. 4.24. Enhancement as a
function of time of year.l 22 1
t = 0, noon
dj = collector tilt
0 R = reflector tilt
X = latitude
p = main reflectivity
R = reflector length
L = collector length
0.0
M A M J J A S O N D J
MONTH (DAY = 21)
Enhancement factor P plotted as a function of time

throughout the year. The abscissa labels time in terms of the
twenty-first day of each month. A collector angle of 6T = 80° and a
reflector angle of 0° were assumed, with A = 45°. The maximum
value of RIL was 3-94 for 21 Dec. at t = 3 hr. Most of the winter
months required R \L to be less than 2-0.
It is not yet clear how thermal mass should best be used in the UK in
conjunction with large areas of glazing. It is, however, obvious that spaces
which cool off during the night or on overcast days, and which contain
large amounts of thermal mass, must be heated up again when the sun
does shine. Thus heavyweight spaces opening directly to the sun may well
be cold for a long time after the sun is up. For this reason it seems sensible
in our climate to consider detached or controlled thermal storage.
Phase-change storage (see also Chapter 7)

Some materials, notably calcium chloride and a potassium salt of
phosphoric acid, have hydrates which melt at low temperatures. Essen-
tially, the salts dissolve in their own water of crystallization. For example,
the material chliarolithe, developed by CNRSl23! in Nice, consisting of
CaCl2 water vermiculite and a stabilizer, melts at 28 °C with a latent heat
of fusion of 55 kWh/m3. It has been tested over 6000 melting cycles. The
material has a conductivity of 0.5 W/(m2 K) compared with a value for
water of 0.6 W/(m2 K); the latter figure, however, neglects convection.
Because the conductivity is low, large thin slabs are needed to absorb heat.
To absorb 200 MJ in five hours would require a surface area of very
roughly 100 m2 if the temperature were allowed to rise 10 °C above the
melting point on the outside of 100 mm thick slabs.
The French system places the chliarolithe in 50 mm pvc drainpipes; but
these do not really provide the desired surface area to volume. Thermol 81
Rods also contain a mixture melting near 28 °C and are available com-
mercially. Another American system 124fll uses hollow polymer concrete
tiles filled with a eutectic salt, sodium sulphate decahydrate. The tiles
weigh about 50 kg/m 2 , the salt melts at about 32 °C and the tiles store
Fig. 4.25. MIT Solar V
House: window detail.! * 8 1
(Reprinted from Popular
Science, with permission ©
1978, Times Mirror
Magazine Inc.)
HEAT MIRROR
PREVENTS HEAT
FROM ESCAPING
DAYTIME NIGHTTIME
By day, solar radiation is reflected to Transparent Heat Mirror in window lets
ceiling by mirror-finish venetian-blind in solar radiation, but prevents room
louvers. Ceiling tiles with chemical cores heat from escaping.
store heat. At night, tiles slowly reradi-
ate heat as outside temperature drops.
about 0.6 kWh over the range 30-35 °C. The salt is in two layers, each
only 10 mm thick. Such a configuration is very suitable for absorbing heat
at the correct rate. These tiles were used in the ceiling of the MIT Solar V
Home mentioned in Section 4.3.1, and illustrated in Fig. 4.25 which indi-
cates how the incoming sun is redirected by the reflecting blinds onto the
ceiling tiles. For a room of 15 m 2 , the storage provided using Sol-Arl24al
tiles would be 36 MJ. Most phase-change systems available commercially
change phase and absorb most heat at temperatures which are high for
passive solar applications in the UK. A very good summary of products
marketed in the US can be found in Ref. [24b]. Calorl24cl in the UK
provides interesting materials, changing phase at 12.8, 31 and 35 °C.
Rock-bed storage
Occupied rooms warmed directly by sunlight may overheat; but the
tolerable temperature swing is low - say 28-18 °C. Thus, at best, the
temperature difference in the air available to charge the rock store is only
10 °C, and this is insufficient to drive much energy into the rocks. For
example, with rock of density 2250 kg/m3 and voids of 40% (that is, solid
60% and thus an effective density of 1350 kg/m 3 ), the heat capacity is
about 1.15 MJ/(m3 K). However, the rocks cannot all be raised to 28 °C,
otherwise the room space becomes hot. A reasonable mean value is half
the air temperature swing or 5 K. For a rock bed of 10 m3 the heat
capacity is thus (1.15 MJ/m3) (10 m 3 ) (5 K) or 58 MJ. This situation then
is totally different from that in which air-heating solar collectors charge a
rock bed up to, say, 80 °C.
The process of charging the rock store with surplus heat from a passive-
gain space, using air as the transfer medium, is more successful than revers-
ing the procedure and warming the space in the evening with warm air
from the store. This is because the air will exit from the store at a low
Fig. 4.26. Rock-bed
storage.!25 !
SUNSHINE
supply temperature (that is, 25-18 °C in the cited case, depending on the
state of discharge of the store).
To move appreciable quantities of energy at very low supply tempera-
tures implies very high air volumes, high air speeds and chilling effects to
the unfortunate occupants. The preferred method is to allow heat from a
low-temperature rock store under the floor to rise by conduction, natural
convection and radiation through the floor into the space requiring
warmth. This process is also quite slow. One cannot expect a heat transfer
rate from the floor of more than about 10 W/(m2 K). If the store is of
10 m3 and measures 3 m X 4 m X 0.8 m, the upper surface area is 12 m 2 .
Assuming a room temperature of 18 °C and a store at 25 °C, the best that
can be expected is a heat supply of (12 m 2 ) (7 K) (10 W/(m2 K)) or
0.84 kW.
Balcombl25! describes how a rock bed should be sized, and what fan
speeds and pressure drops should be expected. Kreith & Kreiderl26! also
discuss pressure drops across rock stores. In practice, it seems that a rock
bed calculated to store Z kWh between certain temperature limits actually
only stores and liberates about 0.75 Z kWh.
Following Balcomb one may, for example, design a system to store
72 MJ. Accepting that the temperature change in the rocks will be 10 K
then the volume will be given by:
72 MJ = (volume m 3 ) (5 K) (1.15 MJ/(m3 K)) or 12.5 m 3 .
Suppose the shape is 4 m X 4 m X 0.78 m, positioned under the floor, as
shown in Fig. 4.26, then the face area is 4 m X 0.78 m or 3.12 m 2 .
Next we calculate the air flow necessary to transfer the 72 MJ in a
suitable time interval. This rate depends on assumptions about the length
of time that the sun shines to produce the surplus 72 MJ. Suppose this
period is five hours, then the rate of heat transfer is 4000 W and suppose
that there is a temperature difference in the air stream of 10 K correspond-
ing to an air flow rate of 1100 m 3 /h. Then the face velocity of the air at
Fig. 4.27. Rock-bed Rock bed performance map
performancet25! Sea level
816
408.5
163.4
81.7
40.9
16.3
8.17
4.08
1.63
0.82
0.41
0.16
0.08
0.51 1.02 1.52 2.03 2.54
Length required to achieve 95% heat transfer (m)
the entry to this store is

1100 m 3 /h
3.12 m2
or 350 m/h.
Reference to Fig. 4.27 shows that, with 125 mm diameter rocks, the
distance the air travels into the bed before yielding up 95% of its heat is
less than 2 m and therefore a 4 m long bed will satisfactorily absorb the
heat. The pressure drop will be less than 2.5 Pa/m of bed or 9 Pa total. The
pressure drop in the ductwork to and from the rock bed would probably
be about 17 Pa giving a total of less than 30 Pa.
The amount of storage required will be reviewed when considering the
solar energy available in passive solar designs.
Thermal storage in the building fabric

In sunny climates, large amounts of thermal storage are recom-
mended in the building fabric. To quote Balcomb: 'It is not advisable to
design for transfer of more than approximately 1/3 of the net heat out of
a space to a rock bed. "Net heat" means the solar energy input minus the
daytime losses. To exceed 1/3 will result in either excessive air flow rates
or excessive (room and store) temperatures.'!251
It is thus common in New Mexico to use 350 mm thick internal adobe
brick walls. In the Unit 1 First Village House (Fig. 4.36), for example,
there is about 40 m3 of adobe walling between the attached sun space and
the rest of the house plus the mass of sun-space floor. The volumes of rock
beds are 12.5 m3 and 10 m 3 . Thus the ratio of structural storage (45 m 3 )
in the space heated directly by the sun, to the rock-bed storage: 12.5 m3
+ 10 m3 with 4% voids + floor = 24 m 3 , is about 2:1. However, in the
UK there are few days on which more solar energy is available during day-
light hours than is required to heat the building during these same hours.
Passive solar heated buildings 123
In other words, apart from smoothing out possible short-term surplus
towards early afternoon, thermal storage may be redundant. This point is
illustrated more clearly in the subsequent discussion concerning the Unit 1
House.
There is, of course, a certain amount of thermal mass in any construc-
tion. For example, even an insulated timber-framed room lined on the
inside with 15 mm plaster board has a heat capacity of about
24 kJ/(m2 K), that is, a room with 60 m2 of wall and ceiling has a heat
capacity of 1.4 MJ/K in its plaster alone. Furnishings might add a further
0.4 MJ/K. For the whole house this adds up to about 10 MJ/K of thermal
storage. For a house with an insulated brick and block cavity and solid
floors downstairs, the value might be about 70 MJ/K. If we assume that
the fabric of the house cools down overnight to 13 °C and requires warm-
ing up during the following day to 16 °C (thus allowing the walls to be
cooler than the room air), then 215 MJ of heat will be needed. Very
roughly, the solar energy available through a vertical surface in the UK in
winter is 3.6 MJ/(m2 d). Thus to supply this 215 MJ would require 60 m2
of glazing. This calculation is, of course, illustrative only — it is open to
great criticism, and is only intended to illustrate the pitfalls of drawing on
experience from sunny countries where winter solar radiation is likely to
be three times that experienced in the UK.
4.3 Passive solar heated buildings
For convenience in description, one can divide the designs used for passive
solar space heating into six types:
direct gain;
attached sun space;
convection loops;
Trombe or water walls;
roof ponds;
roof space collectors.
Obviously, all the methods will depend upon exposing a large area of
glazing to the sun and, apart from the roof ponds, the glazing will gener-
ally be vertical or steeply sloping.
4.3.1 Direct gain

Most buildings have some windows which admit the sun. In direct-
gain designs, this area is greatly increased. The advantages over other
passive forms lie in the conventional appearance and in the low cost which
arises because the builder is not involved in unfamiliar techniques, and
because no extra plan area is introduced to the building. Unfortunately,
the design also carries some drawbacks. Firstly, unless diffusing glass is
used the occupants may suffer from the glare of the direct sunlight, and
furnishings can fade in the shorter wavelengths of light penetrating the
glass. (Note that diffusing glass will distribute the light and heat to all
room surfaces.) Secondly, even if a lot of energy is entering via the win-
dow, on a cold day the glass will be considerably below room temperature,
21
Table 4.8. US passive buildings^ ^
Direct gain 23
Trombe wall*5 13
Water walls 12
Roof-space collector 4
Sun space 24
2
Fig. 4.28. Estimated yearly - UIO
consumption of artificially J/y

introduced heat as a func-
tion of glazed area and -0.8
design indoor temperature, Number of days when
0.5 room air changes per Number of days when heating is supplied
hour. -0 5 no heating is needed uneconomically
T =22°C 4
~34 •• • l24days/>
104
2
lOl
145
17
| |
0 0.2 0.4 06 0 8 10
Glazed fraction of wall
thus providing a source of radiant cooling to the occupants. When the sun
goes down the glazing causes a great deal of heat loss. Even multiple
glazing and heat mirrors are unlikely to reduce this loss below 1 W/(m2 K),
which is three to five times greater than a well-insulated wall (per m 2 ).
Night insulation may thus be mandatory unless the space is not to be used
until late the next morning. Finally, the temperature will swing widely
unless controlled thermal storage is present, leading to overheating in
summer.
The buildings described in the following section all try to moderate
these drawbacks without adding greatly to the initial cost.
There are very few documented examples of most of the six basic
passive designs in Europe, the most common being the use of Trombe
walls or direct gain. Experience in the US suggests that this position will
alter quickly, and Table 4.8 lists the number of passive designs which are
very briefly described in a survey by the American Institute of Architects
(1978)J 27 ' The number in 1983 is probably about 100 000.
The Wallasey school building in the UK has already been described, and
it remains only to add that lights are turned on from 6 a.m. to 8 p.m. to
supplement the heat provided by the students and the sun. Daviesf1! esti-
mates the energy supplied by students and lights, and Fig. 4.28 indicates
that, at an internal temperature of 20 °C, the number of days during a
year when no heating from any non-solar sources is needed varies from
three to 102 as the area for direct solar gain increases. The number of days
when the loss through the glass exceeds the gain is estimated as 104. Of
House floor plan.1 18 1
(Reprinted from Popular
Science, with permission ©
1978, Times Mirror
Magazine Inc.)
" PANELS OPEN STORAGE SOLAR TILES
FOR CABINETS ON TOP OF
SUMMER ( BELOW ) SETTEES
VENTILATION
ENTRANCE
FLOOR PLAN
course, this figure is about half the days in the heating session and in many
ways it is representative of a particular building - one which is not occu-
pied at night and which has no insulating shutters on the glass wall.
The second UK example was built near London in 1956 and designed
by EJ.W. CurtisJ 28 ! One reason for including a large solar window was to
add a feeling of spaciousness to a small ground plan o f l l . 2 m X 6.1 m
(total floor area 132 m 2 ). The unobstructed south front is entirely made in
ply-glass double-glazed units set in timber, with a total area of 61 m 2 , and
it was only on appeal that planning permission was granted — refusal
having been on the grounds that 'it was too modern for the area'.
The brick walls consist of two 38 mm cavities separated by heavy
aluminium foil, and the calculated (/-values are: east and west walls
1.1 W/(m2 K); roof 0.62 W/(m2 K); north wall 0.79 W/(m2 K) - these
values reflect fuel prices and attitudes towards insulation in 1956 and are
not satisfactory today. During one monitoring period from 23 March to
8 May 1957 the back-up heat pump was only 'required for short boosts as
it was possible to heat the entire house from direct solar gains via the
south and west window walls'.t28* From 8 May onwards, no back-up
heating was required. Average temperatures were 21 °C (day-time) and
22 °C (evenings). Electricity consumption for the only back-up - the heat
pump - was 24 GJ (day-time) and 5.4 GJ (night-time).
No condensation was found with the solar wall but, in summer, over-
heating was a problem - though not a serious one. It is notable that no
special attempt was made to introduce thermal mass — the east and west
walls were of cavity brickwork, the south wall of glass and the north wall
of resin-bonded plypanels - the ceilings of fibreboard on timber joists -
the ground floor was, however, in poured concrete.
One of the most exciting buildings using direct solar gain is the experi-
mental MIT Solar V Classroom.!18' 29J This project avoids glare by reflect-
ing incoming radiation to the ceiling with interpane Venetian blinds, it uses
phase-change tiles in the ceiling to moderate temperature swings and has a
heat mirror film in the windows to reduce heat loss via the glazing. The
floor plan is shown in Fig. 4.29 and the outside in Fig. 4.30.
House. (Reprinted from
Popular Science, with per-
mission © 1978, Times
Mirror Magazine Inc.)
r 1
p
1 i •
-
* wwi
g•• 1
:
1
The south-angled windows are fixed closed with a total area of 18 m 2 ,
and are triple glazed with two sheets of water-white glass and one coated
plastic sheet half-way between them. The two gaps are each 20 mm and
the solar transparency of the coated plastic is 70% implying a total win-
dow transparency of 57%. The heat loss through the treated window is
said to be one-quarter to one-third of that through conventional double
glazing over a 24 h period. The inner gap is occupied by a Venetian blind
whose horizontal slats are covered, on the upper slightly concave surface,
with aluminized mylar whose reflectance is initially 85%.
The slats are very closely spaced (10 mm) and are 8 mm wide. The tilt
can be varied through the season as the sun's elevation in the middle of the
Passive solar heated buildings Yll
day alters, in order to redirect the direct solar radiation towards the ceiling
tiles. The blind even in its 'open' position reflects out 5—10% of the solar
radiation.
The half of the ceiling adjacent to the windows is covered with 100 tiles
30 mm thick and 580 mm square containing pouches of Na2SO4.10H2O
and 9% NaCl to lower the melting point, borax to induce recrystallization
and the thickening agent Cab-O-Sil. The ensemble melts at 23 °C. The
total heat stored in the tiles is about 90 MJ for a 6 K temperature change.
The tiles are darkly coloured to absorb radiation. Seats under the windows
are also covered with tiles adding about 29 MJ/6 K of storage — they are
charged because the blinds do not reach the bottom of the windows.
The floor consists of 100 mm of concrete resting on gravel and, in
terms of a daily temperature cycle, its storage capacity is about 65 MJ for
a 5 K temperature change. The floor is further stabilized by coupling to a
large volume of earth which is partly enclosed by insulation. 100 mm
styrofoam board extends 450 mm down below the walls and 900 mm
beyond the corners of the building.
The walls contain 150 mm and the ceiling 280 mm of glass fibre batts
giving 17-values of 0.32 and 0.17 W/(m2 K). The heat loss of the occupied
building has been measured as 11.2 MJ/(K d) without a solar contribution.
Measurements indicate that the minimum night-time temperature is
normally 17 °C and the maximum day-time temperature is 22 °C in
winter.
This design offers features which may make it appealing in the UK cli-
mate. Glare is restricted by sacrificing some solar gain, much of the
thermal storage is only 'present' at high temperatures, and thus the space
is not kept cold unnecessarily and is of relatively fast response (it might be
better for that reason to use a timber floor). However, the low transmit -
tance of the windows means that if the building were in the UK, then
typically, in winter, only about 29 MJ/d of solar heat would reach the
interior space. Assuming a mean £/-value of 0.9 W/(m2 K) for the windows,
the heat lost over 24 h would be about 29 MJ. It is thus not clear that
insulating shutters used at night would not be a better solution in the UK.
Karen Terry's house is the last direct-gain design to be discussed. It is
situated in the hilly country around Santa Fe, New Mexico, at 2200 m
elevation and is located as far south as Gibraltar. The site experiences
3300 °C heating days per year. (1000 °C days means 100 days on which
the ambient temperature was 10 K below a fixed temperature, or 200 days
with a 5 K gap, etc. The number is usually assessed hourly and roughly
represents the severity of the heating load in a particular site. The data
here are expressed to a fixed temperature of 16.5 °C, and typical values
for the UK are Thames Valley 2100, east Anglia 2400 and east Scotland
2700 (see also Appendix 2).)
Karen Terry supervised the building and made many of the internal
fittings. David Wright was the architect, and this beautiful house (Fig.
4.31) cascades down a southerly hillside in a series of one-storey terraces.
The walls are in adobe 350 mm thick, which are insulated on the outside
with 50 mm of rigid urethane foam, rendered, and bermed on the north,
Fig. 4.31. Terry House, New
Mexico, US.t 52 J
east and west sides. The floors are also made of adobe brick.I
Each of the four terraces has a south-facing roof of tempered double
glazing 5 m X 2 m in four sections. These are insulated at night with
internal shutters and protected against summer sun by fixed horizontal
wooden slats on the outside (Fig. 4.19).
The incoming sunlight strikes the solid walls and floors and also hollow
internal partitions filled with drums containing water. The total water
volume is 0.35 m 3 with a heat capacity of 16 MJ/K. The 65 t of wall and
49 t of floor also add thermal mass whose heat capacity on a diurnal basis
is roughly 70 MJ/K.
Back-up heating is provided by two woodburning stoves; but these are
rarely needed. In the winter of 1975-76 only half a cord of wood - about
7.2 useful GJ - was burned.
Table 4.9 summarizes the performance of some of the houses discussed
in the preceding section.
4.3.2 Attached sun spaces

The attached sun space in the Autarkic House shown in Fig. 4.32 is
one of the very few examples of a solar space designed specifically to pro-
vide passive heating to the rest of the building in the UK climate. The per-
formance was explored using hourly weather data as an input to the com-
puter model which is described in Refs. [31] and [32].
Fig. 4.33(fl) illustrates the floor plan and (b) shows insulating shutters
between the conservatory and the house. In sunny weather these shutters
would normally be open, allowing the transfer of heat acquired by the
solar space to the rest of the house. In cold dull weather the shutters
would be closed, thus isolating the glazed space (which would be cold)
from the house. This facility to 'change the size of the inhabited volume'
is an attractive feature of this type of design. When the living volume is
diminished in this way, the amount of glazing to occupied areas is also
Table 4.9. Summary of direct-gain spaces for January and May (calculated)0
Solar Approximate6
radiation heat loss
penetrating Approximate through
appropriate South-facing daily solar glazing
glazing at area of input (with Mean daily without Hear
Site and floor area correct tilt glazing no shading) temperature shutters balance
(m 2 ) (MJ/(m2 d)) (m2) (MJ/d) (°C) (MJ/d) (MJ/d)
Terry House, Sante Fe Jan 14.4 36 518 -2 202 «+300

(85) May 16.9 608 +7 112 «+500
MIT V, Boston Jan 4.7 18 85 -1.6 27 « +50
(88) May 6.1 109 +0.5 25 «+100
Curtis, Rickmansworth Jan 3.2 61 202 +4.0 205 «0
(132) May 5.8 360 11.0 108 * +250
Wallasey Jan 3.2 520 1660 +4.0 1760 *-100
(2500) May 5.8 3020 + 11.0 900 * +2000
a
By the authors.
b
Mean internal temperature 18°
Fig. 4.32. The Autarkic

House, Cambridge. (Photo
by John Donat.)
reduced, the volume of the space is lowered, and the degree of wall insu-
lation is increased. It should be noted that current Building Regulations
require 10% glazing between a room which 'borrows' light and an adjacent
conservatory. In turn, the conservatory then requires windows equal to
10% of its own, plus 10% of the room-next-door's floor area.
The sun space provides, in theory, extra and inexpensive space, and it
can be fitted onto existing or new houses, thus giving one of the few
credible ways of attacking the problem of existing buildings. The space is
best considered as a general amenity rather than one for growing plants.
Large areas of soil or foliage imply high levels of humidity with possible
problems of odour and condensation. Further, there seems little basis in
Fig. 4.33. The Autarkic

House, Cambridge. Insulat-
ing shutters are marked 4 S\
(a) Ground-floor plan. fact for assuming that a few plants convert appreciable quantities of
(b) Central north-south carbon dioxide into oxygen or, indeed, act in any way to clean the air.
section. The major advantages then of this design of passive heating are its
potential low cost, the fact that it acts as a buffer between the outside and
the inside (when the insulating shutters are closed) and the pleasant space
created. Since the space is not required for occupation, its temperature
can be allowed to swing more widely than that of a direct-gain room, and
yet it moderates the temperature swing of adjacent rooms. Olgyayl 33'
illustrates the temperature excursions which are tolerable (Fig. 4.34). The
conservatory can then be allowed to vary from about 12-30 °C before it
becomes unusable in the UK, whereas an occupied space should preferably
be kept between about 15 and 22 °C. (However, the growth of many types
of plants would restrict the permissible temperature range in a conserva-
tory.) Inherently then, this wider temperature range can lead to more
energy storage and less rejection of surplus heat.
From the point of view of design, it is most difficult to predict the
performance of the attached sun space. For one thing, the cost and thus
the thermal integrity of the space can vary enormously and the perform-
ance will be very strongly influenced by the quality of construction.
Furthermore, since the space will be occupied by people who may open
windows or doors, the degree of human influence is high. Finally, if
manual means are used for transferring heat to the rest of the house via
switched fans or opening doors or moveable insulation, one cannot assume
that these operations will be carried out at appropriate times - or even
carried out at all!
During the design of the solar space in the Autarkic House, data were
used for 1962-63 (a 'bad' winter) and 1964—65 (an 'average' winter). It
Table 4.10. Number of days (expressed as a fraction of the number of days
of the indicated type) when the shutters between living area and sun space
would be opened^32]
Month Overcasta Intermittent Sunny Total
October 1962 9/10 11/11 10/10 30/31

November 1962 9/21 4/5 3/4 16/30
December 1962 0/16 0/3 4/12 4/31
January 1963 4/19 6/6 6/6 16/31
February 1963 2/15 5/6 6/7 13/28
March 1963 12/15 7/7 9/9 28/31
April 1963 10/11 8/8 11/11 29/30
a
Overcast days are defined as those when the hours-of-bright-sunshine
indicator is less than 0.1 (on a scale of ^ 0—1 for every day-time hour).
Sunny days are those which have an hours-of-bright-sunshine indicator
greater than or equal to 0.5 for more than half of the day-time hours.
Any other day is an intermittent one.
°c
50 Fig. 4.34. Schematic bio-
climatic index. (Reprinted
with permission, Princeton
43 University Press © 1963,
from Design with Climate:
38 Bioclimatic Approach to
Architectural Regionalism,
by Aladar Olgyay.)
32
27
21
16
10
-1 -
20° 1 1 1 i I 1 1 1 I I I
0 10 20 30 40 50 60 70 80 90 100%
Relative humidity
was predicted that the sun space would acquire sufficient solar energy to
raise its temperature to 18 °C on the number of days illustrated in Table
4.10. Thus, even in a bad year, the conservatory was sufficiently warm for
use on 136 days out of 212 in the heating season. For an average year the
corresponding number was 155 out of 212. Clearly, it was also warm
enough in the summer.
During the calculation, allowance was made for the radiation incident
Table 4.11. Statistics on energy collection in the attached sun space of the
Cambridge Autarkic House
MJ energy saved by storing

Number of days surplus day-time solar gain for
in the month heating in the:
Month an excess Evening Night Next day
May 1962 27 0.0 0.0 23.0
June 1962 30 0.0 9.0 0.0
July 1962 30 0.0 16.6 4.0
August 1962 31 0.0 13.7 0.0
September 1962 30 0.0 5.8 0.0
October 1962 30 0.0 1.4 11.9
November 1962 9 1.1 0.0 37.8
December 1962 0 0.0 0.0 0.0
January 1963 1 1.4 0.0 0.0
February 1963 5 16.9 11.5 7.9
March 1963 19 10.8 6.8 46.1
April 1963 26 0.7 0.4 33.8
Totals 238 30.9 65.2 164.5
on the sloping glazing, reflection and transmission losses and thermal

storage.
Surplus energy was stored for use later on the same or the following
day. In fact, the number of days when more energy was supplied by the
attached sun space than was needed to heat the house during the hours of
sunlight was remarkably small (obviously these remarks refer only to the
heating season). Table 4.11 illustrates the value or otherwise of thermal
storage in this case. It seems clear, then, that thermal storage is required to
smooth out the midday peak of radiation, but is not very gainfully
employed over longer periods. The decision was thus made to limit storage
to one day's excess solar input on a spring or autumn day, with due allow-
ance for the inefficiency of charging a store.
When the radiation (on the horizontal - which is the only measured
value available, and then not for all sites) has a value of 8 MJ/(m2 d),
typical of early March or early October, that incident on a south-facing
vertical surface is about 9 MJ/(m2 d). The radiation penetrating double
glazing (suggested for the Autarkic House sun space) is thus about
6.3 MJ/(m2 d). The space heating load of this particular low-energy house
was about 72 MJ on a typical autumn day (including heat loss from the
conservatory) and the glazed area was 30 m 2 . Thus it can be seen that the
amount of thermal storage chosen (36 MJ) was reasonable.
The principal choices to be made in the design of an attached sun space
(apart from considerations of overshadowing, structure, access and so on)
concern these points: Is the skin to be double or single glazed? How great
should be the thermal mass? How is the heat to be transferred to the rest
of the building and how is ventilation and/or shading to be effected for
Fig. 4.35. Thermal storage
for conservatories.
days when overheating might occur? To a great extent, the solutions to

these problems depend on the future occupants and their demands on the
space. If the space is to be used in the evening then single glazing with
insulating shutters is probably the correct solution, whereas if the space is
to be used predominantly as a solar collector it is probably better to use
double glazing, although the cost is so high that it cannot be justified on
energy grounds alone. In the view of the authors, controlled thermal mass
in the form of a small rock-bed unit charged with warm air drawn from the
top of the sun space and placed under an adjacent room in the house, or in
the form of phase-change materials, is the sensible choice in the UK. In the
latter case it is more difficult to use the stored heat for warming the rest of
the building after sundown and this method is more appropriate to a con-
servatory which is used in the evenings. Warm air can be transferred from
the space to the house by opening doors (which requires intervention) or
by fans in a communicating wall, one at high level blowing from sun space
to living room and possibly one at low level blowing in the reverse direc-
tion. A simple thermostat may control the fans, with an override if the
living room rises above a preset temperature. If a rock-bed store is used, it
may be possible to use the same fan system with a bypass either to the
store or to the living room (Fig. 4.35).
Either system is expensive - one fan costs about £90 to install (1980),
and simple high and low vents in the common wall, operated by mech-
anical greenhouse ventilator actuators may be more appropriate. The slot-
area required can be calculated approximately:
Rate of heat transfer (kW) = 0.174 (h)1/2 (Dt)3/2,
where
A = vent area in m 2 ;
h = height between vents in m;
Dt = temperature difference between upper and lower vents in K.
For example, a conservatory linked to an adjacent room by a high
opening of 1 m2 passing air at 25 °C in one direction, and a low opening
of 1 m2 passing air at 20 °C in the opposite direction should cause a heat-
transfer rate of about 2Vi kW. (In practice, the flow is likely to be less
Fig. 4.36. First Village
Unit 1, Santa Fe, New
Mexico, US (designed by
Susan and Wayne Nichols).
because of vent shape and the restricted size of the roof in an attached
conservatory, which tends to diminish the free flow of the column of air.)
Ventilation may also be required if the conservatory overheats and it is
customary to provide an area of openings equal to 20% of the floor area of
a greenhouse. However, this probably underestimates the requirement in
an attached sun space whose surfaces, apart from the southerly one, are
better insulated. In the height of summer 30-40% will be needed. It is
more desirable, however, to limit the access of summer sunshine by
shading, since windows and doors which can open are more expensive and
will increase undesirable air infiltration in the winter. Sections 4.2.3 and
4.2.4 deal with some of the shading methods available.
One of the very few dwellings for which experimental information is
available concerning the performance of an attached sun space is in New
Mexico and is illustrated in Figs. 4.36 and 4.37.
This house consists of living accommodation folded round and shelter-
ing a south-facing conservatory. In a terrace, a similar effect could be
obtained by sinking the sun space in each house behind the south facade,
rather than leaving it as an 'add-on' exposed piece. This concept is
exploited in a proposal for a Steiner School residential building shown in
Fig. 4.38. The glazed area need not be a conservatory and, as in the
Steiner case, can serve as circulation and play space.
Returning to Fig. 4.37, it may be seen that the rooms are disposed suit-
ably for morning and evening use.
The north wall of the conservatory, which leads to the other rooms in
the house, is made of adobe, 350 mm thick on the ground floor and
250 mm thick on the first floor. The external walls are also about 300 mm
thick and the floor is paved with adobe bricks. The space thus has a high
heat capacity (see Section 4.2.6). In addition, hot air from the roof of the
6 m high sun space is extracted and blown down into two rock beds
underneath the living room and dining room. The vertical (15 m 2 ) and
\.
Fig. 4.37. First Village,

Unit 1, floor plans.l 39 l
Fig. 4.38. Proposed new

residential building for
Steiner Schools.
(Architects, Feilden Clegg
Design.)
sloping south walls (25 m2 at 50°) of the conservatory are double glazed.
Fig. 4.39 illustrates the behaviour in winter. For example, on 28 December,
as the insolation (on the horizontal) rose to 1.8 MJ/(m2 h) at midday, the
upper air in the conservatory climbed to about 42 °C, the air in the centre
to 32 °C, yet the air outside was only 5 °C. The total insolation during the
day was about 8.5 MJ/m2 (horizontal) and, at that latitude and time of
year, the ratio of radiation on a south-facing vertical surface to that on the
horizontal!35J is about 1.2:1, and the ratio for a surface sloping at 50°
is 1.3:1 ,l 34 l thus the total radiation incident is about 424 MJ/d. The
transmitted valuel34! is thus about 306 MJ/d. Now, heat loss via the
Fig. 4.39. Representative
performance of First
Village Unit 1 House,
December-January 1978-
79. Prepared by Los
Alamos Laboratories.! 3 9 J
Upper rake: % inch from outside m °C 50
I inch from inside 4 38
27
16
27
21
16
Floor, above east edge of rock bed
100 Dining room air
80
60
40
20 Ambient
0
Total electric power Baseboard heaters
kW 5
26 28 1 3
Day of month
double glazing over a 24 h period with the temperature shown in Fig. 4.39
amounts to about 173 MJ (when the glazing is uninsulated at night), and
the ventilation heat loss from the conservatory is about 94 MJ/d. Thus the
conservatory gains a net 306 - (173 + 94), or 39 MJ/d. Some of this heat
is put into the rock stores as indicated by the 'fan-on' marks. The fan was
running for three to four hours on 28 December at a rate of 50 m 3 /(h m 2 )
of glazing, and the temperature difference between the warm air pumped
into the store and the store itself was about 6 °C. But only half of this
temperature difference is useful for charging a store and, assuming a useful
value of 3 K, one finds a heat transfer of 27 MJ. Of course, in reality the
fans only came on because the heavy fabric of the greenhouse was not able
to absorb heat fast enough in the middle of the day. The exposed surfaces
of adobe brick in the sun space in the walls and floor would only absorb
heat at a rate of about 7 MJ/h and at midday, this would be insufficient to
prevent overheating - the rate of heat transfer via the fans would add a
further 7 or more MJ/h.
The significant point to notice is the small back-up heat needed in the
morning. In other words, in the climate of New Mexico, the house works
beautifully. It is indeed a splendid example of sensitive design.
Our problem in the UK is not so simple.
Whereas the insolation in the New Mexico case was 8.5 MJ/(m2 d), that
on a good day in December or January in the average year, 1964—65 in
the UK, was 3.2 MJ/(m2 d) - by chance also on 28 December! (An aver-
age value would be 2.0 MJ/(m2 d) for this time of year.) The ratio of
34
radiation on the horizontal to south-vertical! J is a healthy 1.6:1 and the
same value is assumed for the 50° tilt. Thus the incident radiation in the
UK would be 209 MJ/d. The transmitted value would be about 144 MJ/d.
The heat loss via the glazing on that day would have been the same as in
Santa Fe (173 MJ), and via ventilation probably greater, since in our
humid climate an air change of 1/h might be inadequate — however, the
same value, 94 MJ, is accepted in this calculation. Thus the heat balance is
adverse (+ 144 - 173 - 94) or -123 MJ/d. The design would then require
modification, and to be successful would probably move towards the
isolable sun space of the Autarkic House. In this case, most of the heat loss
would occur only during the daylight hours (being reduced to about
90 MJ/d) and the conservatory would then supply a surplus of 36 MJ to
heat the house and the fabric of the greenhouse.
4.3.3 Thermal storage walls

Trombe and Michel built several houses, near the Solar Furnace at
Odeillo, latitude 42°, in the Pyrenees, which use dark heavyweight walls Warm air
Sunlight -
covered with glazing in which solar radiation is absorbed during the day.
Delayed heat
Slowly the heat penetrates the walls to appear several hours later in the transfer
occupied rooms to the north. If heat is required during the day, vents can Double glazing
be opened to encourage air circulation — see Fig. 4.40. (In fact, as
Balcomb has pointed out, this design was patented in the US in the last =L_Cool air
century but the description Trombe wall has remained.)
The advantages conferred by this method of passive solar heating are Fig. 4.40. Trombe wall.
the lack of human intervention necessary, the fact that the wall can be Double glazed Trombe wall
structural thus off-setting some costs, and the straightforward nature of showing convected heat
rising from the warmed
the system. However, it has several important drawbacks; possibly the space. In practice there
worst of these in the UK context is the degree to which winter sunlight is would be flaps on the slots
prevented from reaching occupied spaces. (Thus a mixture of direct gain to prevent cold air at night
entering the room.
and Trombe wall becomes more desirable.) Secondly, the cost is high -
the wall and foundations being far stronger than is needed for structural
reasons. Thirdly, the system lacks the control desirable for a country with
a solar regime which is intermittent. Finally, the appearance from the out-
side tends to be monolithic, and the nature of the blank wall influences
internal layout.
The system does however have possibly greater potential in the non-
domestic sector, and the Benedictine monastery at Pecos in New Mexico is
a good example of such an application. The office and warehouse attached
to the monastery are 85% passively solar heated by a mixture of direct
gain and a water wall. Table 4.12 indicates the heat storage capacities of
various materials (see also Appendix 2).
It is clear that water has about twice the heat capacity of the other
materials, and thus a hollow wall, or a series of steel drums, filled with
water, will absorb more heat for the same volume. In addition, because of
convection currents, the water will transfer the heat more rapidly, thus
keeping the hot surface exposed to the sun at a lower temperature during
the day. This reduces the heat loss outwards through the glazing.
The Pecos monastery shown in Fig. 4.41 contains a mixture of direct
Table 4.12. Thermal properties of some storage materials
Specific heat Thermal conductivity

(MJ/(m3 K)) (W/(m K))
Water 4.2 0.6*

Adobe brick 1.7 0.5
Common brick 1.7 0.7
Dense concrete 2.6 1.7
a
In a configuration which encourages convection, this figure will be much
greater, and a water wall will tend to have the same temperature through-
out its thickness.
Fig. 4.41. Pecos Monastery,

office warehouse.
gain (100 m 2 ) and water wall (40 m 2 ), all of which is double-glazed. The
wall consists of 140, 200 1 oil drums painted black, filled with water and
treated with corrosion inhibitors. Beyond the glass, thin aluminium
shutters provide some night insulation and serve to screen out the sun in
summer. The drums are in a container which can be opened to admit more
heat to the building and which is insulated with 100-150 mm of glass
fibre. The building is well suited to passive gain since the warehouse sec-
tion can tolerate wide temperature swings and low minima, although, in
fact, whilst the temperature can go down to —13 °C, the normal daily
swing is only ±3 K. The swing in the direct-gain space housing the offices
is greater, at ±4-5 K.
The first of the Odeillo Houses (Fig. 4.42), designed by Felix Trombe
and Jacques Michel and built by the Centre National des Recherches Scien-
tifiques (CNRS), was established in 1967 and has a floor area of 80 m2
with four main rooms. The walls have {/-values of 1.3 W/(m2 K) and the
windows (20 m 2 ) are single glazed with areas of 6 and 8 m2 on the south
and north sides respectively. The gap between the double glazing and the
Fig. 4.42. Trombe wall
house at Odeillo.
black painted Trombe wall (48 m 2 ) is 120 mm. Some of the collected heat
is passed into the house via air naturally circulating through slots at the
top and bottom of the wall. The delay in heat penetrating the wall is
14-16 h for a thickness of 600 mm.
The house is 60—70% solar heated, and it is rare for the rooms to be
too cold (occasionally this happens on winter mornings), or too warm
(some autumn afternoons). According to CNRSj 36 ! such buildings in the
north of France (Chauvency le Chateau, Meuse) achieve only 35-45%
solar space heating.
It is significant that the average duration of sunshine during the heating
season at Odeillo is 10 h/d.
A second series of (attached) houses was built as a terrace at Odeillo in
1974 (Fig. 4.43). Their superior insulation (G = 1 W/(m3 K) against
1.63 W/(m3 K) for the earlier houses) permits a reduction of collector area
from 0.16 m 2 /m 3 to 0.1 m 2 /m 3 of inhabited volume. This reduction
allows more ordinary windows to be used on the south facade. The wall
(/-values are about 0.4 W/(m2 K). (G-values are global heat-loss coefficients
from a space.) Table 4.13 indicates some of the parameters.
On a gloomy day (24 December 1974) when the radiation striking the
glazing was only 1.15 MJ/(m2 d) no temperature rise was observed on the
Trombe wall which in fact consistently declined in temperature. Fig. 4.44,
however, illustrates the behaviour of the 1974 houses on a better winter's
day. The volume of air flowing between the slots cut in the lower and
upper parts of the wall (see Table 4.13) is significant. On the day described
in Fig. 4.44 the maximum flow for a temperature difference between the
slots of 22 °C was 70 m 3 /h (see Section 4.3.2), representing a heat transfer
of about 550 W for each panel of the collector, each of area 3.9 m 2 . Many
Trombe walls are fitted with dampers to control this movement since the
air temperature may easily rise by about 30 °C in summer. Such dampers
fl
Table 4.13. Data for the Odeillo Houses
Houses built in 1967 Houses built in 1974
Number of houses 1 3
Floor area 76 m2 180-250 m2
Collector area m2/house 0.16 0.1
volume m3
Back-up heating Electricity Electricity
Trombe wall material 48 m2 45-65 m2
600 mm concrete 370 mm concrete
2200 kg/m3 2200 kg/m3
k = 1.75 k = 1.75W/(mK)
C = 0.6 C = 0.32 MJ/(m3 K)
Trombe wall surface Black acrylic paint Black acrylic paint
Trombe wall glazing 2 panes 3 mm glass 2 panes 4 mm glass
Air path Glass to wall 120 mm Glass to wall 120 mm
Circulation vents 3.5 m apart 2.20 m apart (vertically)
565 mm X 110 mm 840 mm X 95 mm
Each section of wall 4.4 m high X 1.3 m wide 2.5 m high X 1.6 m wide
a
Data from Ref. [36].
Fig. 4.43. Trombe wall

house at Odeillo.
W/m Fig. 4.44. Data concerning
1000- Trombe wall (1974) house
for 23 December 1974.136)
Solar radiation striking

500- glass of Trombe wall in
,one day
C
Black surface of wall -60
Air flowing from wall
inlet 18 °C) 40
- 20
-0
8 10121416 18202224 2 4 6 8
Time (hours)
should be designed (for example, from light flaps of polythene) to prevent

a reverse flow at night.
The time delay in heat transmission through the wall is large (14—16 h
for the thick and 9-10 h for the thinner wall). In fact, this rate and the
inner temperature depend on moisture content and the walls take at least
a year to dry out.
The long-term efficiency of solar radiation capture is between 20 and
40% in the heating season, and at that southerly latitude it diminishes to
less than 20% in summer. (These values refer to insolation on the hori-
zontal.) The ratio of heat transferred by conduction through the wall to
convection of the air is about 60:40 and the solar contribution to the
space heating load is from 60 to 70%.
It is interesting to speculate on what would happen if these houses were
built in the UK. When similar buildings are placed in Meuse, to the north-
east of Paris, they perform less well, contributing 35—45% of space heating.
En passant, remember that in New Mexico a good winter's day supplies
11 MJ/(m2 d) to a vertical surface; in the UK the value is only
5.5 MJ/(m2 d), which is very much less than the value of 20 found in
Odeillo.
Now, data from Odeillo indicate some days with radiation at the UK
level, and the data are reproduced in Fig. 4.45. The heat transferred on
that day in the warm-air stream was less than 108 MJ/m2 and that through
the wall less than 215 MJ/m2 of wall. The collection efficiency was
between 5 and 10%. However, in winter, even days with radiation levels of
5.5 MJ/m2 are rare in the UK.
One way to improve the performance of a Trombe or water wall is to
provide night insulation. In the Baer Housed38! the walls of water-filled
steel drums are covered at night with large insulating panels which lie on
the ground during the day, acting as reflectors (see Fig. 4.46). In the
142 Passive solar deisgn
Fig. 4.45. Data from Odeillo W/m 2
on Trombe wall house for 400 -
22 February 1975.
300-
/ \ Solar radiation on vertical
200 -
/ I
100- /Total
/5.0
0-
^^y MJ/m2 I
- 30 °C
•—-.._ / \Black side of wall
- 20
Room side -^
of wall
- 10
^^ Outside air temperature
-0
8 1012 141618202224 2 4 6 8
Time (hours)
Fig. 4.46. Baer House.! 52]
Tyrrell House (Fig. 4.47), polystyrene beads are blown upwards between
the panes of the double glazing covering the Trombe wall, with a small fan,
at night, in the so-called Beadwall system.
Fig. 4.48(a) illustrates a terrace of single-storey houses using Trombe
walls and built near Liverpool. Fig. 4AS(b) shows their floor-plan and
Table 4.14 summarizes some of their design parameters.
There is a large gap between the black wall and the glazing (about
600 mm), and the solar wall is pierced by the front door and lounge win-
dow. External canopy blinds provide summer shading.
The warm air from the solar wall rises to the roof space to be directed
through the house from the north perimeter back to the base of the col-
lector. Fans can be used to provide forced circulation.
31
Table 4.14. Acorn Gose Houses^ ^
Trombe wall Double glazed

Black-brick density 2200 kg/m3
Thickness 225 mm
(7= 1.21 W/(m2 K)
Width 6 m, height 3.8 m (approximately)
Collector area m2/house volume m3 =0.13
Building Double glazed
Cavity brick walls with 100 mm glass fibre,
internal brick of density 2000 kg/m3
£/=0.29W/(m 2 K)
Floor U= 0.32
Ceiling £7=0.32
Roof U= 0.31
Floor area 60 m2
Back-up heating Electricity
Fig. 4.47. Tyrrell House.
Table 4.15 illustrates the annual amounts of electricity used as auxiliary

heating.
Some interesting conclusions can be drawn. Firstly, the means for space
heating are 11.6 (Trombe) and 21.5 (conventional). The standard devi-
ations are 3.7 and 5.3 respectively. From such a small sample it is thus not
possible to indicate how much less fuel is likely to be consumed by such
passive houses. Secondly, in spite of their occupants (elderly people, some
on their own and with no children), heating hot water accounts for a large
fraction of the total energy use. Thus, not all our attention should be
focussed on reducing space heating loads.
A second terrace of low-cost maisonettes built with unvented mass
walls and using Maxsorb selective surfaces is shown in Fig. 4.49.
Table 4.15. Annual fuel use in the conventional and passive solar houses at
Acorn Gose, between 8 September 1980 and 7 September 1981 (GJ)
Space heating Water heating
Trombe houses
3 10.3 5.7
4 8.2 6.6
5 18.9 7.0
6 10.4 4.9
7 9.5 4.6
Conventional houses
8 15.8 8.6
9 24.6 5.7
10 15.2 6.8
11 23.2 7.8
12 28.8 3.9
Fig. 4.48. (a) View of the

solar terrace at Bebington
looking west, (b) Floor plan
of the sheltered housing at
Bebington.
Plan
4.3.4 Roof ponds

Roof ponds are used where nocturnal cooling is useful in the sum-
mer and the sun is high enough in the sky in the winter to provide heat on
a horizontal plane. They are not appropriate for homes in more-northern
latitudes, but the concept may be useful for those commercial buildings
which require cooling.
Harold Hay's House in Atascadero, California (outdoor temperatures
Fig. 4.49. Low cost
maisonettes in Bath using
mass walls with selective
surfaces (designed by
Feilden Clegg Associates).
8-28 °C) is completely heated and cooled using the system illustrated in
Fig. 4.50.
On the roof are 23 t of water in clear plastic bags with black polythene
underneath. Insulating panels can be drawn over the water bags, and extra
water can be sprayed on in summer. In hot weather, at night, the panels
are opened allowing the bags to cool by radiation to the cold night sky,
and on the following day the panels are closed permitting the cold bags to
cool the ceiling (which is of steel decking). In winter the panels are opened
during the day, allowing the bags to store solar energy.
The system has not been tried for cooling commercial buildings in the
UK and the scheme might not work so well because our skies are much less
clear than those over the desert areas of California.
Alternative methods of cooling buildings naturally are, of course, avail-
Fig. 4.50. Thermal storage
roofs, summer and winter
operation.t 35 !
Y V /"
Summer cooling
Winter heating
able and the wind-scoop towers of Persia and Egypt, which encourage
draughts to pervade buildings, are well known.
4.3.5 Roof-space collectors

When a traditional pitched roof requires replacement, the question
obviously arises 'is it worth glazing?'
Intuitively, a south-facing glass roof starts with several advantages over
other forms of passive solar heating. No intrusion is made into living space,
no extra space is added, the existing fabric of beams may be used and,
perhaps most important of all, the roof is less likely to be overshadowed
by neighbouring buildings.
The intuitive objection raised is that such shallow slopes (say 30°)
reflect a great deal of the solar radiation striking the glazing. However,
this is not true even for direct sunlight; loss in clear-sky radiation pen-
etrating double glazing at a slope of 30° compared with a vertical window
is less than 10% in the UK during a heating season.
In addition, much of our radiation, particularly during the heating
season, is diffuse, coming partly from the direction of the sun but also
from the whole sky-vault, which favours a sloping surface.
There are, of course, severe practical problems associated with trapping
warm air in roofs. The space is notoriously leaky, the timbers may suffer
from the high temperatures and the daily temperature cycling, and this
problem may be particularly acute with the trussed-rafter construction
common today.
A solar roof should be darkly coloured throughout on the inside and
the unglazed surfaces must be insulated. Since the space is of light thermal
Fig. 4.51. Convective loop.
weight it will respond quickly to radiation, and either the warm air must
be blown elsewhere, into the house or into a rock store, or thermal mass
must be introduced into the loft. There is virtually no experience with this
kind of system, but it shares with an attached sun space the advantage that
the occupants can tolerate wide swings in its temperature.
At night, unless the glass is covered, nocturnal cooling may well reduce
the temperature very drastically, thus the ceilings of the rooms below must
be insulated. If the glazing is left uncovered at night, thermal storage
should not be introduced into the roof space, otherwise it will become a
'coolth' storage.
An actual roof collector system is described in Chapter 12.
4.3.6 Convective loops

The final type of 'passive' solar energy system to be mentioned bears
similarities to active thermosiphoning designs for heating domestic hot
water. Fig. 4.51 illustrates a solar collector situated below the space to be
heated, which allows warm air to rise either into thermal storage or into
the space to be heated. Cool air may then return to the collector or be
vented.
This type of system shares many of the problems inherent in active col-
lectors - the possibility of very high temperatures in the unlikely event of
stagnation of the air, high absorber-plate temperatures following from
poor heat transfer to the air and thus a low collection efficiency and, from
the design point of view, the collector is an 'add-on' piece of hardware.
Since the air flow is not fan driven, it is below the desired rate for high
performance efficiency, which affects the economics of the situation. Any
rock storage system attached to a convective loop will introduce an extra
pressure drop which may drastically reduce the performance of the
collector.
Baerl38J recommends that the width of the flow path should be ^ of
the length from top to bottom of the collector — thus a 3 m length implies
a flow path 200 mm wide; in addition to which must be added spaces
between glazing and collector plate, and a thickness for insulation. The
return-duct cross-section should be 5% of the collector area.
In very sunny conditions, the air flow in such a collector can be
adequate; but in the UK weather it is doubtful whether such a system will
work very well. Balcomb et al. ,I391 for example, quote the very low value
of 2 Pa for the pressure driving air around such a system on a good day
(with an average collector temperature of 49 °C and a room (thus collec-
tor inlet) temperature of 21 °C). Of course, on the other hand, when there
is no sunlight, the collector ceases to work forwards, but precautions must
be taken to avoid it working backwards, when it produces a heat loss!
The principle of convective loops can of course be applied in many situ-
ations without formal acknowledgement. For example, when a greenhouse
is placed below a rock bed or fusion-storage system, or an attached sun
space transfers heat by convection into living rooms.
Ref. [39] presents an excellent review of the convective loop system;
however, it is probably not the most suitable passive system for northern
European conditions.
4.4 Design methods for passive solar buildings

4.4.1 Introduction
About 50 programs simulating energy use in buildings have been
reviewed by one of the present authors^40! in order to identify those
which could offer the possibility of use as comprehensive models for the
evaluation of passive solar designs (see Table 4.16).
The major attributes necessary are discussed below and three models
are identified as potentially the most suitable. An indication is provided of
current deficiencies in the design tools presently available.
There are several ways in which the success, or likely success, of a par-
ticular passive solar design for a building, may be assessed — other than by
intuition — and in terms of energy and cost.
The most expensive way, and the least flexible, involves constructing
each design in a particular climate and observing its annual performance -
for a variety of occupants and a variety of years with varying weather. The
BRE method^411 of simulated occupancy removes some of these variables.
A much less costly method would use reduced scale models;l42l but so
far, the models used have been of the Los Alamos Test Cell variety measur-
ing about 2 m X 2 m X 3 m (high), which do however allow side-by-side
comparison of various methods of passive solar capture in heated or free-
running mode.
The third method of assessment would involve 'perfect' analogue or
digital simulation models. Such a perfect model would be able to handle
changes of climate, location, occupants and design. Ideally it would be
interactive with the designer — it should be capable of quickly answering
all the 'what if?' questions of the type 'should I sacrifice 2 m of rear
northerly garden to reduce overshadowing by buildings on the other side
of the street?'; 'is it cost-effective to install the hi-tech windows on the
Design methods for passive solar buildings 149
Table 4.16. Possible passive models
Program Why the model is not a prime contender
name Location as a passive simulator, at time of printing
BLAST 3 CERL,* LBL 6

(US); Boeing CDCC
Time Share (UK)
DEROB 4 Univ. Austin
Texas (US)
DOE 2.1 LASL, d LBL 6
FREHEAT Colo. State Limited documentation, only one zone,
Univ. (US) only mass wall or direct gain
SOLAR 5 UCLA,* Archi- Primarily an excellent building descrip-
tecture (US) tion model not an 'energy' model
SUNCODE Ecotope Inc.,
Seattle (US)
SYRSOL Syracuse Univ. Uses only the sol-air method
(US)
TRNSYS Univ. Wisconsin Heavily weighted towards active solar
(US) systems
2 ZONE LBL (US)
UWENSOL Univ. Washington,
Seattle (US)
ESP Abacus, Strathclyde
(UK)
SUNSPOT LASL (US) Excellent research r . . ,
, *n j• , • out not orsan-
B
tool for direct gain 1 . ,
< lzed rfor
PASOLE LASL (US) Excellent research 1 general users
tool for mass walls
THERM Watson House,
London
TASS Cranfield (UK)
BEEP CEGB^(UK) Private code not accessible for changes
or inspection of methods. Incomplete
documentation, not user friendly.
BUILD Univ.
Nottingham (UK)
HOUSE ECRC* (UK)
NBSLD NBS* (US) Superseded by BLAST
UWIST UWIST1' Not documented.
BRISTOL Univ. Bristol,
Architecture (UK)
UMIST Building Dept.,
hybrid UMIST/
a
Construction Engineering Research Laboratories.
b
Lawrence Berkeley Laboratories.
c
Control Data Corporation.
^Los Alamos Laboratories.
e
University of California at Los Angeles.
•f Central Electricity Generating Board.
g
Electricity Council Research Centre.
^ National Bureau of Standards.
'" University of Wales, Institute of Science and Technology.
i University of Manchester, Institute of Science and Technology.
Table 4.17. Passive design strategies
1 1
1
Thenm o ] <2tOT*fl Ciiffopp {ft
1
f 1 Storage Insulated
Non- under isolated Selective
Capture strategy Isothermal isothermal pcm a floor storage Painted coating6
Direct gain 7 7 7 - 7 7 -
Attached sun space 7 7 7 7 7 -
Mass wall 7 7 J - - 7 7
Trombe wall 7 7 s/ - - 7 7
Thermosiphon - - v/ si 7 7 7
Double envelope - - si - - 7 -
Roof-space collection - - si 7 7 7
a
Phase-change material, for example melting at 25 °C.
b
Solar absorptive, but with low emissivity in the infra-red (for heating), or the converse (for cooling).
south side which only start to reject sunlight when the northerly rooms
are above 22 °C?\ and so on. Needless to say, such a perfect model does
not exist.
4.4.2 Characteristics of a perfect design model

A perfect model would answer all possible questions and thus the
problem of which criteria to use for the assessment of passive designs may
be left aside. It needs to be said though that the criteria are not clear -
does one use solar load ratios; annual back-up demand; annual solar energy
usefully used per m2 of floor area, per m2 of window, etc? This problem
has not been resolved.
The model would ideally be comprehensive, cheap to run and contain
fast input and readily understood output; but these requirements conflict,
and the only models which approach even one requirement - the desired
degree of flexibility - are of medium size (say 5000-10 000 lines of
FORTRAN code) running (at present) on main frames.
Table 4.17 indicates the range of strategies which a simulation model
should practically be expected to attack. The great scope of physical prob-
lems which is implied by these strategies is indicated in Table 4.1.8.
4.4.3 Models approaching the required degree of flexibility

In 1979 The Design Office Consortium (DOC)l43l produced a report
concerning 46 'Models for Energy in Buildings'. In 1980l44 l Lebens
assessed ESP, DEROB 4, HOUSE and THERM and Jamesl45 ! examined
ESP, DEROB 4 and PASOLE for use as passive simulators. Together with
the results of a SERlt 46 ! study, these reports provided a basis for drawing
up the list of models shown in Table 4.16.
Some of the models can be discounted for this particular application on
the grounds indicated in Table 4.16.
Table 4.18. Some of the physical properties to be modelled
Radiation behaviour at glazing (multiple glazing, glass, plastic, infra-red

and visible)
Radiation exchanges in spaces (on room and furniture surfaces)
Air movement causing heat movement (under the influence of tempera-
ture, gradients and infiltration) in zones and between zones and via doors,
corridors, vents and stairs
Insolation and overshadowing
Building geometry
Control systems
Back-up heating systems
Infiltration
Heat loss to the ground
Comfort models specifically addressing passive buildings
Occupancy
Appliance usage
Moveable window insulation
In Table 4.19 the residual models are presented in more detail. This
shortlist may then be reduced to Table 4.20. The simulation models, then,
which seem potentially capable of handling passive systems are these:
BLAST 3, DEROB 4, ESP (and whilst SUNCODE is specifically written
for passive applications, it is designed to fit on a mini, and thus suffers
restrictions). THERM is omitted because of its limited shading treatments.
Table 4.21 indicates which characteristics can be handled by the four
models.
It is probably fair to say that none of the models is yet in a state where
it could be used fairly to evaluate passive solar designs. ESP and DEROB 4
currently offer the frameworks necessary within which (with changes and
additions) most of the design questions could be answered. SUNCODE is a
very convenient package which fits on medium-sized machines. For the
smaller office it probably represents the optimal model available.
BLAST 3 has the framework which could accommodate all the routines
necessary but, being a response-factor method with a 'time history', has
problems in coping with rapid changes such as movable insulation. It seems
likely that within a few years the outstanding problems associated with the
models will have been solved, resulting in sophisticated design tools. The
major problem areas seem to be: air movement (natural, buoyant and
forced by infiltration, and fan forced); heat loss to the ground and via
floors; model validation; better information concerning the use of passive
buildings by occupants; and problems concerning comfort models specifi-
cally relating to passive designs.
Table 4.19. Some details concerning prime contenders for passive models
Program Research Architect Hourly Time period of Calculation Reasons why a model is less suitable
name oriented oriented time steps calculation method^ than others for passive simulation
BLAST 3 J J J Any rf
DEROB 4 J Any fd
DOE 2.1 J J J Any rf Has been forced to handle passive
and remains heavily weighted with
HVAC plant
SUNCODE J s/ Any fd
2 ZONE J Any rf Limited to 2 zones
UWENSOL J Any fd Not very well documented
ESP J J v/ Any fd
THERM v/ Any xfb
TASS J s/ Any iic Excellent input and output but less
detailed in its handling of other
problems
HOUSE •J Four-day slots fd Only deals with four-day slots
BRISTOL J v/ Any Hard wired Being analogue, presently difficult
Analogue to alter**
UMIST J Any Hard wired Being analogue, presently difficult
Analogue to alter**
UWIST J sj Any fd Essentially a research tool without
the detailed effort necessary at
input and output stages for ease of
use
a
finite difference = fd; response factor = rf.
b
But the boundary temperatures are the wall surfaces not air temperatures.
c
But an fd subset can be run for specific problems such as condensation.
^This position could alter with advances in analogue technique. The analogue method has great advantages of speed but
disadvantages of linear treatments of radiation.
e
Table 4.20. Shortlist of models for passive use
Does it have
shading Comparison ofa Machine
routines by: computing units space in Availability
Program Number of zones for superficially thousands of on a com-
name model can handle Other Self similar jobs words mercial basis Machine type
BLAST 3 10 Yes Yes 140 170 Public domain CDC
DEROB 4 Maximum 9 Yes6 Yes 600 120 £250 c CDC
in practice
ESP-f Multiple Yes Yes 500 120 £5000 DEC 10
Prime 450
Honeywell 6060
HP 3000
SUNCODE^ 10 No d Yes 20 50 £750 DEC 10 and
VAX
THERM Multiple Yes Yes 160 60 Time share only
but crude
a
These numbers predominantly reflect the difference between rf and fd treatments.
b
Some users have experienced software crashes if the other buildings are completely detached.
c
The excellent graphics are not included for this price.
d
A shading 'horizon' is used.
e
More models may be added in 1983 as documentation improves.
•^Chosen for development by the European Commission in its Passive Solar Programme.
# Chosen for use by the UK Department of Energy and the US Department of Energy, in their Passive Solar Programmes.
Table 4.21. 'Best'codes
Topic handled ESP* BLAST 3 DEROB 4 SUNCODE
Designs Direct gain 7 7 sj
Attached sun space 7 7 J 7
Thermosiphon — J j 7
Roof-space collector 7 7
Double envelope - — Difficult —
Mass walls vented 7 7
Mass walls unvented 7 J J 7
Under floor rock beds 7 7
Physical Air-heat movement by convection - by fans 7 J J 7
problems Infiltration / j i /
Solar radiation mapping round spaces0 7 k
Variable glass emissivity m - J 7
Variable room colour 7 J J 7
Effect of furnishings Difficult Difficult Difficult -
Air-temperature-stratification n _ j —
Moveable window insulation — Schedule Schedule Schedule
Isothermal and non-isothermal storage 7 7
Phase change walls — J — 7
Isolated storage - J seeRef. [47] 7
Adequate handling of beam radiation - 7 Ref. [47] 7
Adequate handling of diffuse fadiation 7 v/ v/ 7
Adequate treatment of sky temperature
Weather input Complete 'set' 7 J J 7
Daylighting 7 J J -
Surface temperature for providing comfort temperatures 7 Difficult J 7
Validation a d b c, d
Documentation v. good poor good v. good
Graphics output excellent - excellent P
Building input via digitizing tablet 7 - in principle -
Building input format Cartesian Cartesian Assemble /
coordinates coordinates standard
shapes
Internal tables of construction parameters such as 7 J - 7
insulation values
Is it continuing to be improved? 7 7
Comments f g e h
Notes to Table 4.21
a
Limited comparisons have been made between simulated (ESP) and observed data from conventional terraced houses
in Scotland.^ 0 1
b
DEROB 4 has been compared with results from Los Alamos Test Cells!511 with quite satisfactory results, and has
been compared^ 47 ! using runs on imaginary buildings with very large solar apertures, with BLAST 3, SUNCODE
and DOE 2.1, with poor results. These poor results were due to convergence errors and subsequent runs are believed
to be satisfactory.
c
SUNCODE has been validated against results from Los Alamos Test Cells and against a battery of analytic tests on
simple enclosures, where the final results can be calculated exactly.
d
Recently (1982) SUNCODE and BLAST 3 have been extensively compared with experimental data on a conventional
house at SERI with very good results.
e
(1) There may be problems with the convergence accelerators used and in some cases this may have led to simulation
results at odds with identical runs on other large models (see Ref. [47]). There are four degrees of convergence.
The first level corresponds to that achieved in DOE 2.1 ,i 4 8 l the third iterates on advection and the fourth on all
heat transfers. Instabilities have also crept in when thermostat dead bands were at odds with the time stepping
through the day. The CPU time can rise by a factor of eight at the highest convergence.
(2) DEROB 4 is exceptionally good at setting up the model equations for the user J 4 ^ 1
(3) The overshadowing routines may be confused by large windows since the centre of the window is used as the
viewing point for calculations.
(4) DEROB 4 is one of the longest established codes and in general ought for this reason to be bug free.
f It is possible that fairly detailed air-movement routines will be available to run with ESP in the near future.
8
Work is in progress on adding a description of air movement.
h
SUNCODE does not have a ray-tracing facility.
i
At present, ESP does not have routines called 'attached sun-space', 'Trombe wall', etc. but those ticked in the table
can be simulated by use of the multiple-zone facility. Named routines are likely to be available in 1983-84.
j The models can amend the infiltration value which is an input variable, according to wind speed and internal-
external temperature differences.
k
The fraction of solar radiation absorbed by the walls or the floor in each zone is an input variable.
' SUNCODE does not use a building plan. It accepts wall etc. dimensions.
m
A separate package called 'WINDOW' allows any window with multiple layers and coatings to be inserted into ESP.
n
Under implementation for 1983-84.
° Mapping could be handled by dividing surfaces into multiple zones. In all models this greatly increases computing
time.
P Under implementation 1983.
References
[ 1 ] Davies, M.G. (1975). 'Heating buildings by winter sunshine', Build-
ing Science Special Supplement on Energy and Housing, pp. 53—6.
[2] Pilkington Brothers (1979). How Windows Save Energy Technical
Appendix. Pilkington Brothers R & D Laboratories, Ormskirk,
Lancashire, UK.
[3 ] Pilkington Brothers (1978). Glass and Glazing Federation Glazing
Manual. Pilkington Brothers R & D Laboratories, Ormskirk,
Lancashire, UK.
[4] Linsley, G.F. (1977). 'Glazing flat plate solar collectors', Practical
Aspects of Domestic Solar Water Heaters. London, UK Branch of
the International Solar Energy Society, London.
[5] Berman, S.M. & Silverstein, S.D. (1975). Energy Conservation and
Window Systems. National Technical Information Service, Spring-
field, Virginia, US.
[6] Littler, J.G.F. (1979). Thermal balance at windows'. /. of Energy
Research, 2, 173.
[7] Balon, P. (1980). 'Save energy with ordinary roller shades'. US
Popular Science, January, p. 54.
[8] Stepler, R. (1979). 'Insulating shades'. US Popular Science,
October, p. 98.
[9] Information can be obtained from Appropriate Technology Cor-
poration, Box 975, Brattleboro, Vermont, US.
[10] Information can be obtained from Thermal Technology Corpor-
ation, Box 130, Snowmass, Colorado, US; Guardia Shutters,
Gallery House, Dunstable Street, Ampthill, Bedford, UK, and The
Insulating Shade Company, Box 282, Branford, Connecticut, US.
[11] Clegg, P. & Watkins, D. (1979). Home Energy for the Eighties.
Garden Way Publishing Co., Vermont, US.
[12] Rosengren, B. & Morawetz, E. (1976). 'The Termoroc house'.
European Solar Houses, London UK. Branch of the International
Solar Energy Society, London.
[13] Information can be obtained from Thermoblind Insulated Window
Shutters, Waldorf Way, Dale Road, Wakefield, UK.
[14] Shurcliffe, W.A. (1977). Thermal Shades and Shutters. Shurcliffe
Publishing, 19 Appleton Street, Cambridge, Massachusetts, US.
[15] Anderson, B. & Riordan, M. (1976). Solar Home Book. Cheshire
Books, Harrisville, New Hampshire, US.
[16] Esbensen, T.V. & Korsgaard, V. (1976). 'Dimensioning the solar
heating system in the Zero Energy house in Denmark'. European
Solar Houses, London, UK Branch of the International Solar
Energy Society, London.
[17] Bruno, R., Hermann, W., Horster, H., Kersten, R. & Mahdjuri, E.
(1976). 'The Philips experimental house'. European Solar Houses.
London, UK Branch of the International Solar Energy Society,
London.
[18] Stepler, R. (1978). 'The solar laboratory'. US Popular Science,
June, p. 96.
[19] Berlad, A., Jaung, R., Yoh, Y. & Tutu, N. (1978). 'Transport
control in window systems'. Mechanical Engineering Department
References 157
Report. State University of New York, Stony Brook, New York,
US.
[20] Olgyay, V. & Olgyay, J. (1957). Solar Control and Shading
Devices. Princeton University Press, Princeton, US.
[21 ] Information can be obtained from SunTek Associates, 500 Tamal
Vista Road, Corte Madera, California, US.
[22] McDaniels, D., Baker, S., Kaehn, E., Lowndes, D., Mathew, H.,
Reynolds, J. & Gray, R. (1975 and 1978). Time integrated cal-
culation of the insolation collected by a reflector/collector sys-
tem'. Solar Energy, 17.5, 277 and 20.5,415.
[23] Schneider, M. & Berger, X. (1978). 'Utilisation de parois a chaleur
latente'. Groupe d'Ecothermique Solaire du Centre National de
Recherches Scientifiques, Nice, France.
and
Bordeau, L. & Jaffrin, A. (1978). 'Phase change collector wall
versus water collector wall'. Groupe d'Ecothermique Solaire du
Centre National de Recherches Scientifiques, Nice, France.
[24] (a) Information can be obtained from the Architectural Research
Corporation, 40 Water Street, New York, US, and from the (c)
Calor Solar Group, Calor House, Windsor Road, Slough, UK.
Reference should also be made to (b) Eissenberg, D. & Wyman, C.
(1980) 'What's in store for phase change?' Solar Age, May, p. 12.
[25] Balcomb, J.D. (1979). 'Designing fan forced rock beds'. Solar Age,
November, p. 44.
[26] Kreith, F. & Kreider, J.F. (1975). Solar Heating and Cooling.
McGraw Hill, Washington, US.
[27] American Institute of Architects (1978). A Survey of Passive Solar
Buildings. American Institute of Architects Research Corporation,
Washington, US.
[28] Curtis, E.J.W. (1974). 'Solar energy applications in architecture'.
Low Temperature Solar Collection of Solar Energy in the UK.
London, UK Branch of the International Solar Energy Society,
London.
[29] Shurcliff, W.A. (1978). Solar Heated Buildings of North America.
Brick House Publishing Company, Andover, Massachusetts, US, pp.
105-10.
[30] Shurcliff, W.A., ibid., pp. 186-7.
[31] Littler, J.G.F. & Thomas, R.B. (1984). The Autarkic House, to be
published.
[32] Littler, J.G.F. & Thomas, R.B. (1979). 'Energy use in an autarkic
house'. Transactions of The Martin Centre, Cambridge University,
p. 2.
[33] Olgyay, V. (1963). Design with Climate. Princeton University
Press, Princeton, US.
[34] Basnett, P. (1975). 'Estimation of solar radiation falling on a
vertical surface from measurements on a horizontal plane'. Elec-
tricity Council Research Centre Report M 846, Capenhurst, UK.
[35] McFarland, R.D. & Stromberg, P. (1980). Passive Solar Design
Handbook, editors Balcomb, J.D. & Anderson, B., Department of
Energy, Washington, US.
[36] Trombe, F., Robert, J.F., Cabanat, M. & Sesolis, B. (1979). 'Some
performance characteristics of the CNRS house collectors'. Passive
Collection of Solar Energy in Buildings. London, UK Branch of
The International Solar Energy Society, London.
[37] Greenwood, P. & Ward, H. (1979). 'Solar homes for the elderly,
Acorn Close, Wirral'. Passive Collection of Solar Energy in Build-
ings. London, UK Branch of The International Solar Energy
Society, London.
[38] Baer, S. (1978). 'Solar design'. New Mexico Solar Energy Associ-
ation. Association Bulletin, 3.2.
[39] Balcomb, J.D., McFarland, R.D., Jones, R.W. & Wray, W.O., in a
series of papers, Los Alamos Laboratories, Los Alamos, New
Mexico, US.
[40] Littler, J.G.F. (1982). 'Overview of some available models for
passive solar design'. Computer Aided Design, 14, 15.
[41 ] Fisk, DJ. & Morrison, R.C. (1979). 'Energy conservation tests with
simulated occupancy'. Building Research and Practice, p. 148.
[42] Littler, J.G.F. & Watson, M. (1981). 'Passive solar design and the
use of reduced scale models and component tests'. Report to the
Department of Energy, Energy Technology Support Unit, AERE,
Harwell, Oxford, UK.
[43 ] Burgess, K.S. (1979). Computer Programs for Energy in Buildings.
Design Office Consortium, Cambridge, UK.
[44] Lebens, R. (1981). Private communication. Architecture and
Computer Aided Design, London.
[45] James, R. (1980). Private communication. CAP Scientific, London.
[46] SERI, 1980. Analysis Methods for Solar Heating and Cooling
Applications, Second Edition. Solar Energy Research Institute,
Golden, Colorado, US.
[47] Judkoff, R., Christensen, C, O'Doherty, B., Simms, D., Hannifar,
M. & Wortman, D. (1980). 'A comparative study of four passive
building energy simulations DOE-2.1, BLAST, SUNCAT-2,
DEROB-3'. Fifth National Passive Conference Proceedings. US
Branch of the International Energy Society, Boulder, Colorado,
US. A valuable series of internal and published papers is available
from Judkoff, A. and Wortman, D. at SERI, covering further code
validation carried out between 1980 and 1983.
[48] Arumi-Noe, F. (1981). Private communication. Department of
Architecture, University of Texas, Austin, Texas, US.
[49] Judkoff, R. (1981). Private communication. Solar Energy Research
Institute, Golden, Colorado, US.
[50] Clarke, J.A. & Forrest, I. (1978). 'Validation of the ESP thermal
simulation programme'. ABACUS Paper 61, Department of Archi-
tecture, University of Strathclyde, Scotland.
[51] Arumi-Noe, F. (1979). 'Field validation of the DEROB/PASOLE
system'. Third National Passive Conference Proceedings. US Branch
of the International Solar Energy Society, Boulder, Colorado, US.
[52] Further information may be obtained from the New Mexico Solar
Energy Society, Santa Fe, New Mexico, who produce excellent
colour slide sets concerning many aspects of solar design.
[53] Further information may be obtained from the Energy Design Group,
1 Canton Place, London Road, Bath.
Active solar heating
5.1 Introduction
5.1.1 Definitions
The somewhat arid discussion of the semantics of active hybrid and
passive systems has been alluded to in the previous chapter.
Here, an active system is really taken to mean a bolt-on arrangement,
which is usually not part of the building structure (for example, a solar
hot-water heating system) and which often involves pumps or fans. Of
course, some solar air heaters do form the roofs of buildings (see Section
5.2) and some hot-water systems flow naturally by a thermosiphon.
5.1.2 Popularity of solar heating

The UK Solar Trade Association^1! assesses how widely such sys-
tems are used, and the 1981 figures show that there are 60 firms manu-
facturing solar systems or components or installing them. From 1974 to
1981, 173 000 m2 of collector were produced. About 21 000 m2 of hot-
water system were installed in 1981, comprising about 5000 systems.
About 2400 swimming-pool systems have been installed from 1974 to
1981. The main complaints seem to involve misunderstandings either by
clients or manufacturers of the likely output or benefit of a given system.
There has been no 'Gallup Poll' of users' reactions in the UK, but in the
US a study for the Solar Energy Research Institute^2! shows these con-
clusions: two-thirds of houseowners strongly wish to see solar energy
developed over other sources; one-third of people feel solar is technically
and economically practical today for homes; two-thirds of people have
not considered investing in solar technology for their homes.
5.1.3 Outline of the active systems

A typical solar hot-water heating system is shown in Fig. 5.1. Cold
water is heated in the normal hot-water cylinder by a coil through which
hot water from the boiler circulates, or by an immersion heater. This coil
or heater is at the top (hottest) part of the tank. Expansion in the tank is
taken up through the vent. Hot water for taps is drawn off at the top.
Solar heating is introduced by adding a second coil at the bottom of the cold feed
hot-water cylinder, through which hot water from solar collectors is circu- Pumped single tank system
lated. This closed loop from the collectors is normally pumped, but if the Fig. 5.1.1 3 1 Pumped single-
collectors can be placed at least a metre below the tank, and large pipes tank solar hot water system.
159
160 Active solar heating
Fig. 5.2. Solar air-heating Storage charging mode
system.!4) —— Building heating mode (day)
Building heating mode (night or cloudy)
Daytirne heating
I
Hot region
To room
AAAA registers
Nighttime interface
position Auxiliary
Rock bed storage bin
furnace
Sunny-daytime
interface position
Cool region r c 0 0 |
Room-air
return
Blower
Filter
used, then the circulation will thermosiphon without a pump. Expansion

of the solar loop occurs to the small tank shown, which could be closed
with a flexible diaphragm.
Numerous variations are possible on this theme and will be briefly
described in Section 5.3.
A typical solar air-heating system is shown in Fig. 5.2. This system is
used to heat air for space heating but, as will be described in Section 5.2,
domestic hot water can be added to such a design. Return air from the
house is directed by a fan through the collector and back into the house
via room grilles. Should the collectors be capturing more heat than is
needed, the air is sent into a storage system which can be used for space
heating at night by reversing the air stream. Many variations are possible
around this theme, and thermosiphon air designs are quite often used
when the system can be planned with a very low pressure drop.
5.1.4 Approximate energy available

At this point, some notion of the energy captured may be helpful. At
the optimum angle, about 4 GJ of solar radiation falls per m2 of collector
surface per year in the UK. While collectors themselves might capture
energy under good conditions with an efficiency of 50% the more useful
figure is the ratio of solar energy delivered to taps or central-heating sys-
tems as a percentage of solar radiation striking the collector or radiation
striking the ground. Very roughly, these two figures are both likely to lie
between 10 and 20% on an annual basis. Thus a system of 5 m2 of col-
lectors will deliver the following amount of energy as hot water or hot air:
(0.15) (5 m 2 ) (4 GJ/y) = 3 GJ/y.
The cost of this energy provided, using peak electricity, is £45 or £9/m2
of collector (March 1982 price of 6 p/kWh including a typical proportion
of the standing charge). The energy harvest is roughly 0.6 GJ/(m2 y).
Solar air heating
Valves Fig. 5.3. Detailed solar ail-
Fan heating system.!51 PH =
preheat tank for domestic
hot water. (Reprinted with
permission, from Solar
Heating and Cooling,
Hemisphere Publishing
Corporation.)
Mode 3
5.1.5 Hot water compared to hot water + space heating

Undoubtedly, most consumers are more likely to reap early econ-
omic benefit from a solar hot-water system than from a design which also
supplies space heating. Families or institutions with a sustained and heavy
use of hot water are likely to benefit most; whereas for someone living on
their own, at the moment in the absence of tax incentives, the economics
are not favourable. This is not so true in the US where incentives are avail-
able, or for those who carry out the installation themselves.
The Milton Keynes Polytechnic of Central London (PCL) House, which
is described in Section 5.3, does use solar space heating and captured
15.3 GJ in the year April 1979-March 1980, for a collector area of
34.5 m2 (0.44 GJ/(m2 y)).
5.2 Solar air heating
Fig. 5.2 illustrates the basics of an air system. In Fig. 5.3 a more detailed
system is presented.
In Mode 1, when dampers 1 and 2 are in the positions shown, cool air
from the house is returned to the collector and passed back into the house
after being warmed by the sun and topped up, if necessary, by the
auxiliary heater.
In Mode 2, the house does not require heat and the solar energy is
stored in the pebble bed which is charged by injecting heat from the top.
This aids stratification.
In Mode 3, the house requires heat but the sun is not shining and
energy is drawn from storage. Air flows through the pebble bed in the
opposite direction from that in Mode 2, so that the hottest air is used to
heat the house.
Other modes are possible and may be more useful in regions of low
solar irradiance such as the UK. These will be discussed later but, before
talking about the modes, it will be useful to present some of the design
principles for the components of air systems.
5.2.1 Collectors
Commercial collectors are under system-testing at the Solar Pilot
Test Facility!6! and some designers use site-built systems. Solar Age lists
those commercially available in 1980 in Ref. [7].
There is again great disparity between the collector performance and
the system performance. For example, the Denver Solar House!8' has a
collector which transfers only 25% of the incident energy to storage or to
hot water. Air collectors, on the other hand, can have instantaneous
efficiencies comparable with any other collector of similar sophistication.
Usually, however, the efficiency of air collectors is degraded by a fluid-
mass-flow which is lower than that typically used in water collectors (the
values being in the order of 20-70 m3/(m 2 h) for air and 35-70 kg/(m2 h)
for water, where the data is expressed per m2 of collector. Air has a
density of roughly 1 kg/m 3 .).
The principal reason for the lower flow rate used is the size of fan and
the pumping energy. For space heating applications typical of an air sys-
tem, 30 m2 or more of collector would be used, implying air flowing at
600-2100 m 3 /h (350-1200' 3 /min). A fan to handle this flow at a
pressure drop around the system of 150 Pa (0.6" WG) implies a power
consumption of about 150 W. The parasitic electrical load is thus not
negligible.
The flow rate presents further problems. Velocity in duct work should
not exceed 5 m/s giving a duct size of 350 mm X 350 mm at 2100 m 3 /h,
to which must be added at least 50 mm of insulation to keep duct heat
losses below 10% of the collector output. Thus, the insulated ducts occupy
a large space. Secondly, in Mode 1, when the house is being heated, there
may often be a mismatch between the air volume desirable for efficient
collector operation and that required for comfortable heating. Assuming a
building heat demand of, say, 5 kW, then the air supplied at, say, 50 °C
must flow at about 500 m 3/h, equivalent only to the lower limit for the
collector. This limitation stems from one of the principal advantages of an
air system, namely that the same fluid is used for cooling the collector and
Solar air heating 163
warming the home, thus avoiding losses across heat exchangers or radiators.
A lower supply temperature (for example 35 °C) would raise the flow rate
necessary in the home to 1000 m 3 /h but the air leaving a typical grille
would then be travelling too fast and be too cool, leading to discomfort.
The second general comment about the collectors (and duct work) is
that leaks degrade the system performance to a large extent. It is essential
therefore that the collectors do not present too high a resistance to flow
(< 100 Pa) and that they be of high-quality construction.
5.2.2 Air to water heat exchangers

In general, water for domestic use is required at a higher temperature
than air for space heating and the heat exchanger for domestic use is
placed shortly after the collector outlet. If there is danger of freezing, the
loop should drain down or be filled with a fluid of low freezing point. The
loop will probably be pumped and is thus very similar to a solar hot-water
circuit. Adequate heat transfer is essential and, for example, to heat 100 1
of water from 10 to 40 °C in an 8 h sunny period implies a heat transfer
rate of 450 W. Since the air volume is high compared with the energy
transferred, the duty of the exchanger is abnormal and care must be taken
to check that the pressure drop is not too high.
5.2.3 Dampers
Three-way diverting dampers are generally used, and in the UK
such units are relatively unusual in the domestic situation. Since it is
important that the leakage past a closed damper should be low, it is prob-
ably advisable to use a packaged unit consisting of two dampers and associ-
ated duct work which is specifically designed for this purpose. Journals
such as Solar Age regularly advertize such units.
5.2.4 Auxiliary Energy

The top-up system for the warm air ('aux' in Fig. 5.3) may obviously
be of any kind, provided response is reasonably rapid in order to deal with
bouts of cloudiness. The control for the auxiliary heater can be placed in
the air stream to the house to achieve the desired delivery temperature.
5.2.5 Pebble beds

This is a difficult part of the system to design, both from the point
of view of size and internal layout.
The store in a French house near Paris (48° 40' north), which has
62 m2 of air-heating collector, has a volume of 40 m 3 . The volume of the
store adopted in houses with 32 m2 of collector at Peterborough (52°
north) is 4 m 3 , and that of the Wimpey Air House in London with 33 m2
collector is 8.5 m 3 .
Roughly speaking, the specific heat of rocks (or bricks or pebbles) is
1 kJ/(kg K). The density is about 1400 kg/m3 and the void factor about
0.45, thus a bed of 40 m3 has a heat capacity of 25 MJ/K. For a tempera-
ture rise of 20 °C throughout the bed, 0.5 GJ would be stored. Typically,
in northern latitudes the energy captured per day in spring or autumn
would rarely exceed 8 MJ/m , and for a collector area of 62 m2 the 40 m3
2
store would be very seldom fully charged. On the other hand, the daily
heating demand of the house in spring or autumn is probably about 50 MJ
and thus the store when fully charged could provide about ten days of
heating if it started off at, say, 60 °C.
Clearly, the optimization to be made is fairly complex but unless the
store is cheap it seems unlikely to pay off except in cases where it is quite
small, satisfying perhaps one day's heating. This point is emphasized by
the observation that, for a well-insulated house and its short heating
season, the number of days when there is excess solar energy beyond that
needed to heat the house and hot water during the day is quite small. It
was for these reasons that the homes at Peterborough were fitted with
small stores of 4 m 3 .
The second problem with pebble beds is their internal layout. Clearly,
the overall pressure drop must be kept low (~ 50 Pa). Owing to the large
surface area, there is not normally a problem with more or less total heat
transfer from pebbles to air, or vice versa.
The heat transfer can be represented b
where
h = volumetric heat transfer in W/(m3 K);
G = mass velocity of air (kg/(m2 s) = air mass flow rate/bed frontal
area;
D = pebble diameter (m).
For example, suppose 50 mm diameter pebbles are used in a bed 1 m X
2 m X 2 m (air flowing along a 2 m dimension); then with 30 m2 of col-
lector requiring air at 50 m3/(m 2 h):
T 0.7
50 X 30 \ kg/s /
= 650
a . l X3600X 2 / m2
= 1650 W/(m3 K).
Since the volume of the bed is 4 m 3 , for a temperature difference of 2 °C

the rate of heat transfer is 13 kW.
The pressure drop Ap is given,! 1O1 approximately, in Pascals by:
where
L = bed length (m);
GD = defined above;
p = air density (roughly 1.1 kg/m 3 );
/x = absolute viscosity of air (roughly 1.8 X 10"5 N(s m) or
kg/(m s)).
Solar Heating Fig. 5.4. Control system for
a solar air heating system.[51
Valves
Fan
Thus
50X30 , ,, 2 .
(2 m)
l . l X 3 6 0 0 X 2 k g / ( m S)
Ap -
(1.1 kg/m 3 ) (0.05 m)
(1.8X 10" 5 )
21 + 1750
GD
= 32Pa(~0.13"WG),
thus achieving a low pressure drop is not difficult either; the main prac-
tical problems seem to be:
- arranging the inlet and outlet and rock pattern so that the bed (or
stack of bricks, etc.) enjoys an even air flow which is well dis-
tributed;
- ensuring that the bed remains dry in the humid UK climate so that
there is no build up of organic material (in practice this problem
does not seem to arise);
- keeping out rodents, etc. during building operations;
- sealing the enclosure to retain the same integrity as the rest of the
air system;
- installing the control sensors at appropriate positions;
- timing the building process so that the installation of the store
does not interrupt the natural flow of site trades. For example,
wet trades involved in blockwork may be undesirable in a timber-
framed house.
5.2.6 Controls
Fig. 5.4 indicates the control lines for a typical installation. One
controller runs the domestic hot-water (DHW) system by turning on a
Table 5.1. Control modes in the solar air heated houses at Peterborough
Mode Mode characterization YX YS xe YZ Yd Fanl
1 Venting collector heats off

water (summer)
2 Collector heats water off

and circulates (summer)
3 Collector heats water Y>S satisfied on

and store or off X<
4 Store heats house and Y<S X>6 Y<e not off

DHW, top up by gas if satisfied
necessary
5A Collector heats water, Y>e not on
collector heats rooms, satisfied
gas top up available
5B Collector heats rooms, S>50 Y>e not on
collector heats water satisfied
6A Gas heats rooms X<6 Y<6 not off

satisfied
6B Collector preheats X<6 Y<e not off
make up air, gas tops satisfied
up
Fail safe Fail safe (power failure) off
Note: Fig. 5.5 identifies the location of dampers D145 and D23. Other dampers, valves and temperature sensors are
listed in the text.
pump to circulate water from the preheat tank, PH in the diagram,

through the air to the water heat exchanger (AWE), when sensors indicate
that the air at the exchanger is hotter than the water at the bottom of PH.
Care is needed in choosing the temperature differential and, typically, a
5—10 °C difference might be suitable for switching on. This loop needs
frost protection and boiling protection. Freezing can be avoided by anti-
freeze or a non-freezing fluid or a drain-down system. These are arrange-
ments typical of a 'conventional' water-filled collector system. Boiling may
occur in summer and, whilst a drain-down arrangement can cope with this
problem, an antifreeze system cannot. In that case, provision must be
made to ventilate the air collector to prevent the air stream reaching
100 °C.
The air circulation presents a more challenging problem of control and
is discussed in terms of the Peterborough system,! u 1 illustrated in Fig.
5.5, for the house in Fig. 5.6. See also Table 5.1.
Dampers:
DI admits outside fresh air to collector system
DS safety vent with fusible link to prevent collector overheating
Summer- DHW
CH winter DHW Boiler solar
Fan 2 switch switch switch pump pump V3 D145 D23 DO DS DI
off off s on on on if 1 2 1 open open open

off off /K7and if if if
A<60 r>ioo y>120 y>ioo
off off on on on if 1 2 2 closed closed closed
off off A<Y and
A<60
off on on on if 1 3 1 closed closed closed
off off A<Y and
A<60
on on on on if 2 1 3 closed closed closed
off off ,4<Xand
A<60
off on on on if 1 2 3 closed closed closed
off off A<Y and
A<60
off off on on on if 1 2 3 closed closed closed
off off ,4<rand
A<60
on on on on off - 3 1 closed closed closed
off off
on on w on on off - 3 4 closed closed closed
off off
off off off open open
DO summer vent to allow thermosiphoning to occur drawing air via

the collector, without fans, over the DHW coil
Valves:
VI cuts out the heating coil from the gas boiler to the warm air
stream
V2 cuts out the heating coil from the gas boiler to the domestic
hot-water storage tank
V3 3-way from heat-exchanger in DHW tank to either air stream
from collector position (1) or to heat-store exchanger (2)
Fans:
Fan 1 is collector fan; Fan 2 is warm-air heating fan
Sensors:
5 heat-store temperature at the bottom
X heat-store temperature at the top
Y collector outlet temperature
Z ambient temperature
6 room set point
A temperature at the bottom of DHW cylinder
Roof
collector
D23 Air flow between links

2&3
Air to house
1 &2&3
D145 2&3
1 &3
1 &2
Fig. 5.5. The solar air- B temperature in the middle of DHW cylinder
heating system for the c warm air heating supply temperature (« 35 °C)
Peterborough Houses.!x 11
(Courtesy of Peterborough Commentary concerning the modes and their control:
Development Corporation,
Chief Architect K. Maple-
stone.) Mode 1
In the summer, space heating is not required. In order to avoid
unnecessary use of electricity, the fans are turned off by manually switch-
ing the summer-winter control. Air then circulates naturally by a thermo-
siphon. If, for example, air enters the collector at 20 °C and exits at 70 °C
then the head of air is equivalent to about 4 Pa (O.Ol" WG).
Fig. 5.6. Section of a
Peterborough House.11 * 1
(Courtesy of Peterborough
Development Corporation,
Chief Architect, K. Maple-
Solar
collectors
stone.)
The pressure created is given roughly by :
i>=3400 [^--~r
where
P = pressure in Pa;
Ta = inlet air temperature in K;
Tm = mean air temperature in K;
and where Ta and Tm must be in degrees absolute, that is, °C + 273.
This low pressure is sufficient to drive roughly 20 m 3 /(h m 2 ) of col-
lector area through the ductwork from the outside entry point to the out-
side exit point.
The hot air passing over the DHW transfer coil heats the loop to the
DHW tank through which water circulates, driven by the solar hot-water
pump. The pump is on when the air temperature at the collector exit (Y)
is above the temperature of the water at the bottom of the DHW storage
tank 04). The pump is switched off if A rises above 60 °C to prevent
scaling. The small pressure across the collector is not enough to drive air
through the thermal store, and thus the control system ignores signals X
and S in this mode. Similarly, space heating is switched off, so that signals
from 6, Z are ignored.
Mode 2
This is also a summer mode, but the controller perceives that Yis
< 100 °C and thus adjusts the dampers to close all connections to ambient,
allowing circulation of air around the closed loop of the collectors and
DHW transfer coil.
Mode 3
In mode 3 fan 1 is used to circulate air through the store (which
measures 2 m X l m X 2 m high) and the DHW transfer coil. Space heating
Table 5.2. Details of two solar air houses
Design System
Floor heat Collector Storage Solar collection
Project area load aperture Collector Storage volume fraction efficiency
name Ref. (m 2 ) (GJ/y) (m 2 ) type type (m 3 ) (%) (%)
Dourdain 9 197 126 62.5 Single glazed Rocks 40 32 21
low e glass
black steel
Wimpey a 115 33.3 Matt black No fines6 8.5 40 c 16
a
Report to be available from the UK Department of Energy late 1982.
b
A Wimpey development consisting of coarse aggregate with no fine particles, bound by cement.
c
Including small amounts ( « 10%) of passive solar gain via windows.
is not needed. The store is charged both to provide space heating in the
evening and to preheat the domestic hot water, since a closed circulation
loop connects the heat store to the solar coil at the bottom o f the DHW
tank.
Other modes
The reader may follow the other controller operations by examining
Table 5.1 and Fig. 5.5; but mode 6B is worth emphasizing in the context
of northerly latitudes.
In this mode, the collector is used in conditions of low radiation
intensity, merely to preheat makeup air. Normally, makeup air is not
deliberately introduced in the Peterborough Houses, infiltration is assumed
to provide fresh air, or occupants open windows; but infiltration in winter
probably constitutes 40% of the load of poorly sealed but well-insulated
houses. Pressurizing the house by drawing in fresh air decreases infiltration.
The air volume required is small — perhaps 0.5 house air changes per hour
or about 100 m 3 /h. This represents a flow of only 3 m 3/(h m 2 ) of col-
lector which is greatly below that recommended for high collector
efficiency. The mode is as yet experimental.
The controller
Clearly, to cope with such a complex series of modes a sophisticated
controller is required, and the one used in the Peterborough Houses is built
around a microprocessor.
5.2.7 Results of monitored systems

Data concerning one 'active air house' has been gathered by Energy
Conscious Design within a Commission of the European Communities
(CEC) programme monitoring solar performance J 13 1 Table 5.2 lists some
of the parameters describing the house.
The Dourdain house is in an area with about 4.6 GJ/(m2 y) of irradi-
ation on the horizontal (compared with about 3.4 GJ/(m 2 y) in the south-
ern UK, see PCL/MK house in Section 5.3). The performance is rather
Solar water heating 171
Fig. 5.7. Solar water-heating
systems.!14 !
Drainback Closed Loop
Draindown
Open Loop
poor, and the report! 131 indicates that the storage is too large, the system
too leaky and the air distribution in the collection poor.
Wimpey has also built a test dwelling and some results are shown in
Table 5.2.
5.3 Solar water heating

5.3.1 Outline of systems
There are four main variants of systems for heating water, and they
are illustrated in Fig. 5.7.
In the drain-back system shown in Fig. 5.7(a), a closed loop carries
water (with no antifreeze) from the pump through a heat-exchanger coil in
the hot-water storage tank to the collectors and back to the pump. When
the sun is not shining, the pump is switched off and water drains back to
the water reservoir which is within the insulated volume of the building.
On start-up, the pump must raise the water back to the collectors.
Provided the pipes all slope correctly, no water is left to freeze or to
boil in the collectors under very cold or very hot conditions. Since the
loop is closed, scale deposition and corrosion are reduced in the collectors.
In the closed-loop method shown in Fig. 5.7(b), the same loop is used
but, being protected with antifreeze, it does not drain back. To take up
expansion there is a small tank which may contain a diaphragm to reduce
corrosion.
In Fig. 5.7(c) the heat exchanger is avoided, the water being heated is
the water used at the taps. On drain-down at the end of the day, or when
the sun goes in, a valve opens to discharge the water from the collectors
either to waste or into a sink. Direct heating systems expose the collectors
to mains pressure and to scale formation above about 65 °C; but the
absence of a heat exchanger can increase system efficiency by 10%. The
water circulation tends to destroy stratification in the storage tank, which
may impair performance by a similar percentage.
In Fig. 5.7(d) the same water is heated, and fed to the taps, but freezing
is avoided by circulating warm fluid from the storage tank, on demand by
the freezing sensor. It is thus not suitable for cool climates.
The parasitic power associated with pumped systems is small, the
pumps typically using less than 70 W (or about 0.4 GJ/y). If the vertical
distance from the dmin-back (Fig. 5.1 (a)) reservoir to the top of the col-
lectors is very large then a second pump may be needed to fill the loop in
drained systems, but often one pump can be used both to charge and
circulate.
It is important to note that mains pressure is used in drain-down sys-
tems to refill collectors, and can of course be reduced by a suitable valve
to prevent damage. Such a system is fail safe on power failure if the drain-
down valve is 'normally open', and the system has a vacuum breaker at the
highest point.
In retrofit situations it may be difficult either to introduce a second
heat exchanger into the DHW storage tank (below the boiler coil), for non-
electric heating, or to build in a second preheat storage tank with a heat
exchanger. In such cases the direct systems (drain-down and open-loop)
have obvious advantages, and northern climates probably exclude the
open-loop form.
Antifreeze can form sludges and its acidic break-down compounds can
accelerate corrosion. Checks should be made every year on the freeze pro-
tection afforded, and at the PCL Milton Keynes house it appears that
water containing antifreeze opens up small holes which otherwise might
seal themselves, causing leaks.
A good review of other advantages and disadvantages of these systems
may be found in Ref. [15].
Comparative figures for the various designs, (a) to (d), have not been
found, but variations of more than about 10-20% seem very unlikely for
strictly comparable set-ups designed according to Fig. 5.7. However, very
large variations are probable stemming from more-mundane considerations.
Table 5.3, for example, illustrates this point.
The cost per unit of energy (Table 5.3, published in 1981) varies by
almost 100%. In addition to these variations for collectors operating under
Table 53. From comparative performances estimated by the Tennessee
Valley Authority for typical domestic installations^14^
Estimated energy Cost of delivered

delivered by solar energy (1981)
System (GJ/y) (£/GJ)
Eagle Sun drain-back 15.7-18.9 3.8
Reynolds 1454 13.1-15.8 4.3
Exxon 3 11.7-14.0 5.0
Reynolds 1404 11.2-13.5 4.6
Rheem Ruud (insulated) 11.2-13.5 5.8
ITC 10.4-12.6 5.1
STCHWSG 72120 10.4-12.4 5.1
Sunmaster 10.1-12.2 5.6
Rheem Ruud (uninsulated) 9.2-11.0 7.0
Grumman 9.0-10.8 7.1
Exxon 9.0-10.8 5.9
Northrup 9.0-11.0 5.0
Solar one 5 8.8-10.6 5.6
Table 5.4. Components of an active liquid system
Collectors Heat exchanger

Pump Back-up heater
Controller Drain-down valve
Sensors Vacuum breakers
Thermal storage
Overflow tank Transfer liquid
identical conditions, the annual yields will depend on the draw-off

patterns by the users (and the amount of hot water demanded), over-
shadowing, pipe and tank insulation, settings on the control system and so
on.
It seems likely that 'good practice' at installation will do more to
achieve high performance than will an exhaustive choice of the best
'laboratory' systems or the best 'laboratory' collectors.
However the components are assembled, they may include the items in
Table 5.4. Several of these items are conventional and will not be discussed
further. The special components are dealt with briefly below.
5.3.2 Collectors
A large number of manufacturers make collector panels which are
very similar in principle. There are exotic versions such as evacuated tube
collectors and xeolite heating and cooling panels, which may well sweep
the market in the future. Traditionally, a solar hot-water panel operates by
circulating water through a web of pipes joined by a plate. The plate is
black, is insulated at the back and glazed at the front. Hutchinsl16 ! pre-
sents a useful set of measurements concerning the transmittance of glass,
low-iron glass and plastic cover sheets for collectors, plus measurements of
Fig. 5.8. Instantaneous
efficiency of collectors on
1.. ergy transferred to the water
the PCL/MK House (e.g.
(measured by PCL).[9, 17] .8 - A = panel area
.6 -
temperature difference
' .4 -
.2-
-i—r
.56 1.12 1.68 2.24 2.78 3.36
100 A T/l
Fig. 5.9. Performance of a AT/I (K/J h"1 m"2)

variety of collectors.! l 8 1 0.56 1.12 1.68 2.24 2.78 3.36 3.89 4.44
(Reprinted with permission, 100
I I I I I I I I
from Solar Engineering and (1) Double-glazed flat-plate collector
Thermal Processes, aluminium roll-bond, flat-black paint
(2) Double-glazed flat-plate collector
J. Wiley, NY.) 80 antireflective coating on three glass surfaces •
steel absorber, copper tubing, black-chrome
selective surface
(3) Single-glazed flat-plate collector
60 steel absorber, black-chrome reflective surface .
(4) Double-glazed air heater, flat-black paint
flow rate 0.015 m 3 (m 2 s)
(5) Same as 4, but flow rate 0.010 m 3 (m 2 s)
° 40
I
~o
o
20
I I I
0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
AT/I (K/Wm"2)
absorptance and emittance of coatings. The instantaneous performance of

a collector is described by diagrams such as Fig. 5.8. The graph shows that,
for example, when the solar radiation striking the collector is that preva-
lent on a clear summer day at noon (2500 J/(h m 2 ) (« 700 W/m2)) and
when the air temperature is 20 °C and the inlet and outlet water tempera-
tures are 30 and 60 °C, then the efficiency of transferring energy from the
solar radiation to the water is 55%, that is
AT
/2500 = 0.01
which, from Fig. 5.8, gives q/AIzs 0.55. At low insolation levels typical of
a winter day (for example 0.3 MJ/(h m 2 )) the efficiency, even for heating
water from 10 to 15 °C with an ambient of 8 °C, would be 0%. Typical
efficiencies are shown in Fig. 5.9 for a variety of collectors.
The band of performance of the water heaters is not enormous, and it
has been suggested already that system performance may not reflect col-
lector performance. However, manufacturers ought to be able to provide
this kind of graph and their product should not be markedly inferior to
the ones shown.
Johanssenl19! has surveyed manufacturers. He finds that most give a
five-year guarantee, many do not provide operating manuals, some do not
supply design drawings and specifications, and some do not give perform-
ance data. A panel with a selective surface (which reduces heat loss by
reducing radiation to the cover glazing) may reach 200 °C if the water
stops flowing. Clearly then, the manufacturer should use components,
including insulation, capable of withstanding such conditions. If the panel
is installed between rafters there is some fire risk, and finally the anchor
points must resist wind lift on the assembly.
MichelsonI20 J estimates the optimum size of a system and finds that
for DHW systems, 5 m2 of collectors and 250 1 of hot-water storage is
optimal. She assumed a demand of 250 1/d at 50 °C.
More information is available in Ref. [21], and there is now a British
Standard, BS5919, covering solar collector installations.
The cost of such an installation would be in the range £1000—1500.
5.3.3 Pumps
Flow rates of 35-70 kg/(m2 h) are recommended. Even at the
higher value, a closed system with 5 m2 of collector requires a flow of
only 350 kg/h, and the system is likely to have a pressure drop of less
than 40 kPa.
5.3.4 Heat exchangers

In a solar system, the heat exchanger transfers energy between the
solar loop and the hot-water storage tank. Collector efficiency falls off
very rapidly as the collector temperature rises, and thus it is desirable to
keep the temperature drop between the collector loop and the storage
tank as low as possible. This is done in two ways. Firstly, the heat
exchanger is positioned at the coolest point in the storage tank, at the
bottom. Secondly, the heat exchanger must have a large area. Unfortu-
nately, the transfer rate from exchangers diminishes as the temperature
difference between the metal and the water is reduced, at a rate faster than
unity, that is, the heat-exchange rate per 1 K temperature difference is
smaller the lower the value of the temperature difference. Thus values
quoted for heat exchangers typically used between a boiler and a tank are
higher than will be found if the same exchanger is driven from collectors.
The British Standard previously mentioned suggests that a 60% heat-
exchange performance will only degrade system performance by 3% com-
pared with a 100% efficient heat exchanger. Typically, such a calculation
implies an exchanger with an area equal to about 0.15—0.2 of the col-
lector area. Duffie & Beckman find!22! that, for every 1 K temperature
drop across the heat exchanger, the collection efficiency falls by 1-2%.
5.3.5 Thermal storage

For a domestic hot-water system, approximately 50 1 of storage per
m2 of collector is appropriate. Naturally, any optimization is difficult
since it depends on the family size, etc.
For space heating, a much larger area of collector is often used. The
Milton Keynes Solar Water-Heated House designed at the Polytechnic of
Table 5.5. Solar heating in homes and flats (from Ref. [13])
Solar
Collector System fraction
area Collector Storage efficiency of load*
Place (m 2 ) type (m 3 ) 1/m2 (%) (%)
South UK 34.5 Single-glazed, 4.2 122 11 57
black paint
France (mid) 153 Single-glazed, 10 65 27 43
black paint
France (mid) 340 Single-glazed, 30 88 24 25
black paint
Italy (NE) 129 Single-glazed, 3 23 14 18
black paint
Holland 52 Single-glazed, 4.1 79 35 26
selective
Holland 35 Single-glazed, 2 57 15 32
selective
a
Fraction of space and hot-water load supplied by solar energy.
Central London has 122 1/m2 of storage. Table 5.5 illustrates some other
results.
Great play has been made in the past about encouraging stratification,
so that cold water is returned to the collectors and hot water accumulates
at the top of the storage tank. In direct systems the water is stirred and
stratification is unlikely.
There is disagreement about the advantages of promoting a temperature
gradient. For example, the Shell simulation model^23! suggests a 15%
improvement in solar contribution with a stratified preheat tank.
Duffiel35! has suggested a much smaller effect.
Chapter 7 deals in more detail with interseasonal thermal storage.
5.3.6 Drain-down valves

To optimize system performance, drain back (Fig. 5.7(a)) should
occur at least once a day before nightfall, so that warm water is not left in
the collector to cool down overnight. The implication is that the water
reservoir should be insulated. In drain -down systems, where the water goes
to waste, it is customary for the system to open the valve when the col-
lector temperature drops to near freezing (see the freezing sensor in Fig.
5.7(c)). Drain-down valves are commercially available. Care should be
taken with open-loop systems in hard-water areas, since the drain-down
valve may not seal properly if scale builds up.
5.3.7 Transfer fluid

Water is most commonly used, with or without antifreeze. If a toxic
antifreeze is used, some water authorities insist on double-walled heat
exchangers. There are, however, non-aqueous fluids, for example 'Sun
Temp', which freezes below - 4 0 °C, boils at 355 °C, is non-toxic and, in
principle, non-corrosive. Some problems have been experienced using non-
New types of system 111
Fig. 5.10. The efficiencies,
based on aperture area, of
various evacuated collectors
60 are shown hatched com-
pared with the efficiency of
a well-engineered, double-
glazed, flat-plate collector
with a selective coating on
theabsorber.l 24 !
£ 40
20
0.04 0.07 0.11 0.14
AT/1 (K/Wm2)
aqueous liquids and thin aluminium roll bond collectors. When micro-
organisms growing in the cool part of a system are killed by heat and
swept into the collectors, traces of metal in the organisms may set up
galvanic cells causing corrosion.
5.3.8 Controllers
Controllers for all types of systems are available. Many contain
digital read-outs for temperatures in various parts of the system and,
clearly, if the owner is aware and sensitive to the behaviour of the system,
he may be able to make improvements. Reports from the PCL/MK
housel 21 ! show that by 'tinkering' with the system the annual solar con-
tribution to the space and water heating load was raised from about 15 to
about 50%.
5.4 New types of system

5.4.1 Evacuated tubes
Fig. 5.10 indicates the superior performance of evacuated tube col-
lectors particularly at low light levels or high exit water temperatures (that
is when AT/I is high).
Many variants have been produced (see Ref. [24]), one of the simplest
being shown in Fig. 5.11. At present, such collectors seem useful when
high outlet temperatures are required, for industrial purposes or for driving
absorption air conditioners.
5.4.2 Plastic collectors

High-temperature extrusions are available. The version illustrated in
Fig. 5.12 is very suitable for the 'Do it Yourself (DIY) market, since fixing
the individual tubes to the heaters is simple but tedious. The web comes as
a roll which makes access to roofs more convenient in some cases.
Plastic collectors are reviewed in Ref. [25], and their performance is
Fig. 5.11. Philips Mark
Four solar collectors.! 24 ]
Hi
exchanger
not markedly inferior to copper designs. Fig. 5.13 illustrates a glazed all-
plastic collector.
It should be remembered that whilst polycarbonates have a low thermal
(« 2.5—50 jum) transmittance of about 0.04,l16 l their solar transmittance
(« 0.3—2.5 fim) is also poor (0.76 for the twin-wall version). Some plastics
have very high solar transmittance (for example, Teflon 95) but often high
thermal transmittance too.
-COPPER OR
WA°TLPDR PLASTIC
WATER MANIFOLDS
5.4.3 Zeolite heating and cooling collectors
FLEX-LOOP RETURN '
Tchernev has produced a combined heat-pump collector, as shown in
Fig. 5.12. Typical EPDM Fig. 5.14.I27! The zeolite filling is in a sealed box containing water under
absorber layoutl 2 5 J (EPDM
= ethylene propylene diene
reduced pressure. When the zeolite (a yellow material) is hot it expels
monomer). water vapour which on condensation transfers heat to the water loop. At
night, when it cools off, the vapour pressure falls from its day-time value
of around 7 kPa, to about 0.7 kPa, and water distills from the heat-
exchanger area into the zeolite, thus withdrawing its latent heat of vapour-
ization from the water loop, producing chilled water.
5.4.4 Refrigerant-charged loop

An extremely interesting system has been developed by R.
French^281 and is shown in Fig. 5.15.
Freon Rl 14 (freezing point - 9 4 °C, boiling point 4 °C at atmospheric
pressure) is pressurized in the sealed loop to a maximum of 600 kPa. When
the sun shines, the freon boils and vapour passes up into the heat
exchanger (dashed in Fig. 5.15). Liquid freon runs back to the bottom of
the collector. As the temperature in the store rises, the pressure goes up
until, at 600 kPa (68 °C in the storage tank), the pressure limiter shown in
the diagram closes. No liquid may now enter the collector, which contains
only vapour. The liquid will all be in the heat exchanger.
Installed systems 179
Fig. 5.13. Temtech
• Glazing: transparent polycarbonate; UV
inhibited
collector.! 26]
• Ultraviolet filter: TEDLAR® film-0.025
i Absorber panel: polysulfone, with black
absorbtivity enhancer GLAZING PANEL—'EDGE STRIP—4M
Ml HP ZEZE7
» Insulation: 20 mm thick isocyanurate
foam-vinyl and aluminum-foil clad
• Panel weight: 6 kg/m2 dry; 11 kg/m2
filled with water (approximately)
> Recommended operating temperature:
/?VWS
INSULATION '
IS ^ FILM 8 FOIL
120°Cmax.
» Recommended operating pressure: 205 kPa SPACER <
y ROOF LINE
» Maximum stagnation temperature: 150 °C
i Width: 0.025 m, 58.8 mm
• Length: 2940 mm standard
• Thickness: 50 mm
ALUMINIUM FRAME Fig. 5.14. Integrated zeolite

LOW-IRON panel with flooded-type
DOUBLE GLAZING evaporator—condenser.! 27 ]
BLACK CHROME
ON COPPER COVER
WATER LOOP
LIQUID WATER
HEAT EXCHANGER
INSULATION (COPPER FIN
COIL IN
COPPER CONTAINER)
COPPER CONTAINER
FILLED WITH
ZEOLITE
HIGH POINT
Fig. 5.15. Freon-charged
VAPOR L I N E -
^ - - " ^ STAINLE
S S STEEL TANK collector.! 28]
RESERVOIR/
VAPOR SEPARATOR 1
A 10°-50° A
SIGHT GLASS
COLLECTORS
20
PRESSURE GAUGE/
ACCESS PORT
LIQUID LINE
PRESSURE LIMITER
^ — LOW POINT
The system thus operates without pumps or controls. The heights indi-
cated prevent reverse thermosiphoning at night.
5.5 Installed systems

The Performance Monitoring Group of the EEC has produced an interest-
ing report concerning solar heating.! 131
Fig. 5.16. Polytechnic of
Central London House at
Milton Keynes (MK/PCL
House) (from Ref. [9]).
. Heol Distribution _
jsj sacrificial element
electric immersion
""—1 heater
Fig. 5.17. Month by month Monthly Efficiency: Collection ( H System (—)

performance of MK/PCL ;
|
year [ 1980 1979 |
House.! 9]
— _ —;—-
i
i — 1 '
i I
"-• t i" — *
]
40 ,_ _ . L.- »-..-..i-—[--.-: i . . - ......4 -- .i —

I
ry
I i
,4-4-'--
L ! -t
! I
- • —
—L _ _ ; — u ..... '•••4 .;._*..

-
10 — ! • • • - -
_ ;
Jan Feb Mar Apr May Jon Jui Aug Sep Oct Nov Dec
The scheme shown in Fig. 5.16 and discussed below is very typical of
DHW and space heating plus DHW systems. The flow rate through the
collectors is only 16 l/(m2 h), which is considerably below that recom-
mended, and the annual collector performance is 22%. Fig. 5.17 shows the
performance month by month. Peak efficiency occurs with very low inlet
temperatures and low insolation in the winter. The very large area of col-
lector (32.5 m 2 ) is partly unused in the summer, when the four-person
family does not need all of the hot water available (which is why the mean
storage temperature from May to September is 57 °C).
Table 5.6. Energy summary for MK/PCL house
Gas and
People and Passive gain Active solar electricity
appliances via windows contribution for heating
(MJ) (MJ) (MJ) (MJ)
January 1400 680 1206 3388
February 1270 890 1664 1606
March 1380 1470 2148 2042
April 1310 1640 2329 688
May 930 1250 1364 72
June * * 659 7
July * * 637 0
August * * 651 0
September 300 240 803 4
October 1030 800 1325 162
November 1330 880 1371 1242
December 1400 610 1124 2506
Sub totals 10350 8460 15281 11717 Total = 45 808 MJ/y
% of total 23 18 33 26
* No space heating is needed in these months.
Monthly Load (Space Heating + Hot Water) Fig. 5.18. Monthly loads for
Auxiliary | | Solar £
MK/PCL House.
1000
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
I 1980 I 1979 I
The system efficiency is the sum of the energy delivered to the hot-
water system plus that delivered to the space heating system, divided by
the solar energy incident on the collectors. Energy losses from pipes and
tanks are only counted as useful when the room temperatures are below
the thermostat set point.
Fig. 5.18 illustrates the contributions to the heating load, the annual
figures being roughly 50% solar and 50% gas plus electricity.
Table 5.6 suggests, however, that these figures do not tell the whole
story. Firstly, it is important to realize that, even for this house, which is
not extremely well insulated or sealed, the heating season is shortened by
comparison with poorly insulated dwellings. Secondly, even though the
house has most of its windows to the north (10 m2 north, 8.5 m2
south), quite a large fraction of the total load (18%) is met passively. But
thirdly and most significantly, the occupants and appliances contribute
23% of the total load.
5.6 Multifamily installations

Sillmanl29! finds that economies of scale occur above about 50 dwelling
units with communal thermal storage. The cost of delivered energy to the
consumer is not a smooth function of storage volume. Clearly, if 100% of
the heat needed is available from stored solar energy, then capital may be
saved on auxiliary heating plant. Thus a '100%' solar scheme may be more
economical than a 50% solar scheme for large numbers of houses or flats.
The same is true at the one-house scale, but interseasonal storage is so
expensive (partly because the heat losses are a large fraction of the energy
stored) that such projects are unlikely to be economical until a different
method of storage is developed.
5.7 Predictive methods

Properly formulated, computer simulations can be a powerful design tool.
However, simulations require a significant amount of expertise and
expense which often precludes their use for the design of small systems. In
addition, the hourly meteorological data needed for the calculations are
not available in many locations. Simulations can be used in another way
which overcomes these difficulties. The results of modelling a particular
type of system for a range of practical design parameters and a variety of
climates can be used to generate a design method for this system type. A
design method incorporates the information gained from the simulations
in a correlation which relates the long-term performance of the system to
the important system design parameters and the weather. A design method
cannot provide as much or as accurate information as a simulation. How-
ever, the information from a design method is usually accurate enough for
design purposes! 351 and it can be obtained much more easily and with
significantly less expense than by simulations or experiments.
Such correlation methods have been developed by Klein, Duffle and
Beckman at Madison 13°1 in the US for active and passive systems and by
Balcomb etal. at the Los Alamos Laboratories for passive systems. In the
UK, Kennal 31 ! has produced a similar program.
Typically, such a design method requires the inputs shown in Table 5.7
which is the worksheet for input to f-chart, the 'Madison method'! 36J
(this correlation method uses the Hottel-Whillier equations). Tables 5.1 (a)
to 5.7(e) are included to give readers an idea of the detail required to use
such a method. Complete explanations of the symbols are not given.
Table 5.8 illustrates the output from f-chart 4, using input data for the
Peterborough Houses discussed in Section 5.2 and Chapter 12.
F-chart is probably the most common method in the US for sizing active
Predictive methods 183
Table 5.7(a). Collector input data to f-chart 4
o
o
-S *cl «S
® § .9 s
rj S^ 00 *£
u w £ ?N Collector parameters
x x x x x Cl Collector area m2
x x x x x C2 FR-ULC product W/(m2 K)
x x x x x C3 FR-TAU-ALVHA (normal incidence)
x x x x C4 Concentration ratio
x C5 CPC acceptance half-angle
x C6 Number of covers
x C7 Index of refraction
x C8 Extinction coefficient-length (KL)
x C9 Incidence angle codifier constant^
x x x x x ClOCoUectorflowrate-specific heat/area W/(m2 K)
x x x Cll Tracking axis (1 = t-W, 2 = N-S, 3 = 2-axis)
xx x Cl2 Collector slope6 . degrees
x x x Cl3 Collector azimuth6
x x x x x C14 Ground reflectance6
x x x C15 Incidence angle codifiers (10, 20, 30, 40, 50, 60, 70, 80 degrees)
a
If parameter C9 is specified, then parameters C6, C7, and C8 are not necessary.
6
Twelve values may be specified for parameters C12, C13, and C14. Only one value is necessary if these
parameters do not vary by month.
c
FR = Collector heat removal factor; UL = Overall collector energy loss coefficient.
systems. However, it is not sensitive to load profile, control strategy and

freeze protection technique. 1321 Mo re-complex hour by hour simulation is
necessary for an exploration of these more esoteric variables. TRNSYS in
the US or programs from Cardiff University,l33l PCU 341 or ABACUS^37!
in the UK are available for such an exercise.
Table 5J(b). Losses during heat transfer
! |
ffi Q < Transfer parameters
x Tl EPS-CMINb of collector store i/X c /collector area W/(m 2 K)

x T2 *X4a of collector inlet pipe or duct W/K
x T3 UA of collector outlet pipe or duct W/K
x T4 Collector duct leak rate (%)
x T5 Duct leak location (1 = inlet, 2 = outlet, 3 = both)
a
UA = £/-value x area.
b
EPSMIN = load heat exchanger effectiveness x smaller of the two capacitance rates in the heat
exchanger.
c
HX = heat exchanger.
Table 5.7(c). Thermal storage data
"g
Storage parameters
x SI Tank capacity/collector area KJ/(m 2 K)

x S2 Storage unit height/diameter ratio
x S3 Heat loss coefficient W/(m 2 K)
x S4 Environmental temperature °C
x x x S5 Hot-water auxiliary tank UAa W/°C
x x x S6 Hot-water auxiliary tank environmental temperature °C
x S7 Rock-bed capacity/collector area kJ/(m 2 K)
x S8 Phase-change volume/collector area m3/m2
x S9 Phase-change material density kg/m 3
x SI0 Void fraction
x SI 1 Solid-phase specific heat KJ/(kJ/kg)
x SI2 Liquid-phase specific heat KJ/(kJ/kg)
x S13 Heat of melting KJ/(kJ/kg)
x S14 Melting temperature °C
a
Parameters S5 and S6 are necessary only when domestic hot-water heating is considered.
Predictive methods 185
Table 5 .l(d). Data concerning the heat delivery
a
43
o
ffi ffi Delivery device parameters
x x D1 EPS-CMIN of load heat exchanger Hx W/K

x D2 Minimum temperature for Hx operation °C
x D3 Delivery heat-pump number
x D4 Minimum heat-pump absorber temperature °C
x D5 Heat-pump bypass temperature °C
Table 5.7(e). Input data concerned with loads

ing
60 60
cd
'-§
ess 1
ibine
43
to
Spac
Wat
i3
£ U Load parameters
X X LI Building UAa W/K

X X L2 Room temperature °C
X X L3 Hot-water use 1/d
X X L4 Hot-water set temperature6 °C
X X L5 Water-mains temperature6 °c
X X X L6 Total process or space heating load6 MJ/d
x L7 Hours per day
X L8 Load-return temperature °C
a
Note that the load for space heating is determined in a very approximate
way using a global value for heat loss.
b
Twelve values may be specified for parameters L3—L6. Only one value is
necessary if these parameters do not vary by month. If L6 is specified for
space heating systems, then LI and L2 are not necessary.
Table 5.8. Predicted performance of the Peterborough Houses
* * * F Chart analysis (Version 4.0) * * *

Kew, London, UK Latitude 51.5
Thermal performance
HT TA SHLOAD HWLOAD
(GJ) (DEG-C) (GJ) (GJ) FNP
JAN 4.82 4.3 4.64 0.97 0.05

FEB 5.86 4.9 3.88 0.87 0.19
MAR 9.50 7.5 4.24 0.97 0.39
APR 10.85 10.2 2.54 0.94 0.60
MAY 13.05 12.4 0.84 0.97 0.68
JUN 13.37 16.0 0.64 0.94 0.69
JUL 12.87 16.0 0.17 0.97 0.71
AUG 12.44 16.4 0.36 0.97 0.69
SEP 11.09 16.0 0.69 0.94 0.70
OCT 8.85 12.0 2.51 0.97 0.51
NOV 5.98 7.7 2.37 0.94 0.33
DEC 4.60 4.6 4.82 0.97 0.03
YR 113.28 10.7 27.70 11.42 0.35
HT = total solar energy incident on the collector (for Kew UK)
TA - air temperature
SHLOAD = space heating load
HWLOAD = hot water load
FNP = fraction of non-purchased energy (i.e. fraction of solar energy
in this case)
References
[ 1 ] Information may be obtained from the Solar Trade Association,
28 Store St, London.
[2] Holland, E. (1981). 'Living with the sun'. Solar Age. April, p. 22.
[3] McCartney, K. (1978). Practical Solar Heating. Prism Press,
Dorchester, UK.
[4] Kreider, J.F. & Kreith, F. (1975). Solar Heating and Cooling.
Hemisphere Press, Washington, DC, US. Readers are also advised to
consult the greatly enlarged, and invaluable 1982 edn.
[5] Duffie, J.A. & Beckman, W.A. (1980). Solar Engineering of
Thermal Processes. Wiley, New York.
[6] Information may be obtained from the Building Services Research
and Information Association, Bracknell, UK.
[7] Cuminsky, M. (1980). 'Directory of air collectors'. Solar Age.
November, p. 39.
[8] SeeRef. [5] above, p. 441.
[9] Information may be obtained from Energy Conscious Design, 44
Earlham St, London, in connection with their Reports to the
European Commission concerning Active Solar Space and Water
Heating Performance.
[10] SeeRef. [5J.
References 187
[11] Information may be obtained from Peterborough Development
Corporation, Peterborough, UK who designed the three solar air-
heated houses with assistance from the Building Unit at the Poly-
technic of Central London. The drawings shown here are repro-
duced by courtesy of the Development Corporation.
[12] Dodson, C. (1979). 'Passive ventilation alternative'. Passive Col-
lection of Solar Energy. London, UK Branch of the International
Solar Energy Society, London.
[13] Information may be obtained from Energy Conscious Design, see
Ref. [ 9 ] .
[14] Eagle Sun (1981). 'Advertisement'. Solar Age. August.
[15] Schiller, S.R. (1981). 'Freeze protection'. Solar Age. August.
[16] Hutchins, M.G. (1982). 'Optical properties of materials for flat
plate collectors'. Helios, 14, 7, published by the Solar Energy Unit,
University College, Cardiff, UK.
[17] Information may be obtained from the Built Environment
Research Group at the Polytechnic of Central London.
[18] See Ref. [ 5 ] , p. 272.
[19] Johannsen, M. (1979). 'Survey of UK collectors'. Helios, 7, 1 (see
Ref. [16]).
[20] Michelson, E. (1981). 'Optimum sizing of solar water heating sys-
tems'. Helios, 11,4 (see Ref. [ 16]).
[21 ] Horton, A. & Grove, S. (1979). Milton Keynes Solar House
Solar Heating System, 1975-1979, Performance and Cost
Analysis. Built Environment Research Group, Polytechnic of
Central London.
[22] See Ref. [ 5 ] .
[23] Gillet, W. & Rosenfeld, J. (1981). 'Review of British Standard BS
5918'. Sun at Work in Britain, 1213, 12. Published by the UK
Branch of the International Solar Energy Society, 21 Albermarle
St, London.
[24] Graham, BJ. (1979). 'Evacuated tube collectors'. Solar Age,
November, p. 15.
[25] Madsen, P. & Goss, K. (1981). 'Non-metallic solar collectors'.
Solar Age, January, p. 30.
[26] Best, D. (1982). 'New plastics head for higher temperatures'. Solar
Age, February, p. 51.
[27] Best, D. (1982). 'Innovations with real expectations'. Solar Age,
March, p. 37.
[28] Best, D. (1981). 'What you should know about phase-change water
heaters'. Solar Age, December, p. 22.
[29] Sillman, S. (1981). 'Performance and economics of annual-storage
solar-heating systems'. Solar Energy, 27.6, 513.
[30] Duffie, J.A. & Mitchell, J.W. (1982). 'F-chart predictions and
measurements'. US Annual Meeting of The Association of Mech-
anical and Electrical Engineers, April.
[31 ] Kenna, J.P. ( 1 9 7 2 - 8 3 ) , in various publications from the Solar
Energy Unit, University College, Cardiff, UK.
[32] Augustyn, J. & Schiller, S. (1982). 'A rating system will help con-
sumers and industry alike'. Solar Age, May, p. 48.
[33] Information may be obtained from the Solar Energy Unit, Univer-
sity College, Cardiff.
[34] Information may be obtained from the Built Environment
Research Group, Polytechnic of Central London.
[35] Duffie, J.A. (1982). Private communication.
[36] Information may be obtained from the Solar Laboratory, Univer-
sity of Wisconsin, Madison, Wisconsin, US.
[37] Information may be obtained from ABACUS, Department of
Architecture and Building Science, University of Strathclyde,
Glasgow, UK. FLARE is a comprehensive model for the analysis of
active solar systems including air, water and photovoltaic collectors.
Space heating and ventilation
6.1 Introduction
From the underfloor heating of the Romans to the microbore hydronic
installations of today may seem a small way but the path has been cir-
cuitous and the neglect of innovation has contributed to squandered
resources and lowered living standards. For example, Fig. 6.1 shows an
invention which was on the market in the early 1900s but failed to receive
the attention it merited.I1' The design is similar to the National Coal
Board's present series of room heaters with back boilers. (Current designs
include the safeguards of a vent behind the back boiler and a makeup
tank.)
Prior to the 1950s, domestic heating in the UK was provided principally
by open fires. Then, with rising prosperity, cheaper energy in the form of
imported oil and technical advances such as small glandless centrifugal
pumps designed to circulate small quantities of water against relatively
high heads, central heating began to be adopted. In the past 30 years the
percentage of the housing stock with central heating supplied by oil, gas or
other fuel, has increased from about three to 50.121 Electricity for space
heating also gained in popularity with the introduction of underfloor heat-
ing and storage heaters charged with cheaper night-time electricity.
This wide range of conventional heating systems developed for tra-
ditional houses is generally also applicable to low-energy buildings. Since
numerous authors have dealt with standard applications,l2j 3» 4» 5»6> 7 ' the
emphasis of the following discussion will be on a brief but comprehensive
presentation of the heating and ventilating systems commonly available
and an analysis of their use in buildings and, in particular, homes where
the energy demand is significantly reduced by the conservation and supply
measures discussed in the previous chapters.
Energy shortages and the consequences they impose on designers
should not, of course, be regarded separately from other concomitant
changes in society. The question of the choice of a heating system should
be intimately related to patterns of work and occupancy. As we have seen
(Chapter 1) the BRE estimates that 41-61% of households (that is both
occupied flats and houses) are intermittently occupied. Thus, for many
houses the heating system required is one that will provide a great deal of
heat quickly for several hours in the evening. Even this energy is not
required homogeneously. In the earlier part of the evening, heat is required
189
190 Space heating and ventilation
Fig. 6.1. Hydroradiant fire
and single-pipe gravity
system designed by A.H.
Barker.! * 1 (Courtesy of
cms.)
principally in the living and dining rooms and, in the latter part, in the
bedrooms.
The overall pattern of demand in a low-energy house is one of moderate
demand throughout a somewhat reduced heating season, with some
periods of heightened demand during cold spells. If the occupants are in
the house during most of the day the energy requirements will be fairly
uniform during the daylight hours, with due allowance for, for example,
solar gain during spells of bright sunshine; at night, demand will fall con-
siderably. If the occupants are away during the day a morning peak
demand from, say, 6.30 a.m. to 9.00 a.m. and an evening peak of, say,
6.00 p.m.-l 1.00 p.m. are likely.
Each situation must be analysed individually, bearing in mind that
during the lifetime of the house, varying groups of people are likely to use
it. Considerations of flexibility, in both the immediate and long-term
futures; reliability, given the impossibility of assuring any single energy
supply; and the difficulty of foreseeing which sources of energy will prove
economical argue for the use of various fuels.
Another consideration is that, in new construction or renovating old
buildings, the involvement of the users, or potential users, in the design
process is important for their satisfaction. It is also essential to energy
conservation because even sound design measures such as self-closing
doors, shutters for windows, solid-fuel heating systems, thermostats and
time switches which people are not willing to accept are worse than use-
less. For a house then, a meeting with a client to discuss heating-system
preferences, desired comfort levels, pattern of occupancy, acceptability of
control measures, and so forth, is of great benefit, particularly if the
designer remembers that during the lifetime of the house numerous other
people will live in it. The result of the meeting will hopefully be a fairly
precise determination of the appropriate heating system and an idea of the
maximum permissible variation in the heating and other environmental
Introduction 191
Table 6.1. Percentage contributions (in terms of delivered energy) to the
1974 UK domestic space heating requirement^ 1O1
Fuel Percentage Fuel Percentage
Coal 30 Electric —on peak 5

Other solid fuel 8 Electric — off peak 6
Oil 14 Electric — total 11
Gas 37
requirements, as well as the compromise possible in extreme climatic con-

ditions or emergencies.
Having selected a system, the owners or occupants must be able to con-
trol it and thus the energy consumption of the building, whether this be
by provision for 'task heating', part-load operation of systems, operation
at specific times, partial use of spaces or similar strategies. Additionally, on
at least a monthly basis, some way of determining the actual energy used
should be provided so that the value of conservation is apparent to those
using the space and so that corrective measures can be taken if required.
As one of the occupants of the first Milton Keynes Experimental House
said, regretting the designers' decision to provide only conventional con-
trols in the house: 'If we could glance at a dial and know how much
energy we were paying for . . . I think we could effect even greater savings'
(that is, in addition to those afforded by the solar collectors).^
The primary consequence of reducing the heat loss in a low-energy
house is that smaller heating components may be used. In some cases the
result has been to encourage development of appliances specifically
designed for the lower load and these are discussed below. A secondary
effect is that more flexibility may be possible in siting heat distributors.
For example, with double glazing reducing draught and condensation
problems at windows, radiators may be placed closer to the heat source
and back-to-back in adjacent rooms where possible, thus reducing costs.
For alternative sources of energy such as solar and wind power, the site
will be quite important in determining the contribution to the heating
demand but for conventional sources the effect is minor. Approximately
80% of the UK is presently served by gasf 91 and proximity to coal-mining
areas may influence the choice of heating systems, but, apart from these
considerations, fuels are generally homogeneously available.
Table 6.1 shows the approximate percentage contributions in terms of
delivered energy (this must be multiplied by the efficiency of the heating
appliance to determine the useful energy) of various fuels to the UK
domestic space heating requirement.
Table 6.2 shows the energy-conversion efficiencies.
Another way of viewing the primary energy (PE) to delivered energy
(DE) efficiency, for example electricity, is that 3.69 units of primary
energy input are required for one unit of delivered energy. The significance
of this is the often-made point, which we strongly endorse, that a rational
energy policy must look beyond economics to the primary energy require-
Table 6.2. Energy conversion efficiencies for primary to delivered and
delivered to useful energy for space heating fuels (1974) (after Ref [10])
PE-DE DE-UE Overall (PE-UE)

Type of fuel efficiency a efficiency6 efficiency
Coal 0.969 0.45 0.44
Other solid fuel 0.831 0.65 0.54
Gas 0.909 0.7 0.64
on 0.919 0.7 0.64
Electricity 0.271 \.0c 0.27
Domestic sector
(all fuels) 0.73 0.66 0.49
a
The efficiency of conversion of primary energy to delivered energy
allows for production, transportation, etc.
b
The efficiency of delivered energy to useful energy is necessarily approxi-
mate because of variation in published estimates of conversion efficiency
and because of the wide range of conversion devices.
c
This figure must be treated with some caution. If, for example, electric
heating is used and, because of the 'dead' band of the thermostat, over-
heating occurs, the DE—UE efficiency will be less than 1.0. Heat wasted
by electric immersion heaters in the summer can also be significant.
ments of not only fuels but all the material items (including food) of a
society to see how available primary energy resources may best be dis-
tributed in both space and time.
In the case of fuels, for numerous historical and technical reasons, there
is not always a correspondence between primary energy costs and econ-
omic costs. Table 6.3 shows C. Sutherland's valuable table of comparative
space heating costs and average thermal efficiencies.
The relative positions of the various fuels tend to change over the years
(or months!) according to economical and political realities. Anyone
actually choosing a heating system on cost grounds alone will have to
obtain current data on both capital and running costs and then estimate
how the running costs will change with time. We can only agree that it is
more easily said than done.
Returning to Tables 6.2 and 6.3, efficiency is a particularly vexed
question since it depends on the appliance and the system. For example,
the efficiency of solid-fuel heating varies with the position of the chimney
and with the pattern of use of the house since heat stored in the chimney
may or may not contribute to usefully warming a space. The BRE^12!
provides additional data on system efficiencies. The efficiency also
depends on the load and, since a central-heating boiler may spend the
majority of its operating time at 20-50% of its rated output, it is import-
ant to evaluate its performance under these conditions. Fig. 6.2 shows the
relationship between boiler efficiency and load.
British Gas claims that best current practice can provide overall
efficiencies in the heating season of 75—80%, depending on whether the
installation allows advantage to be taken of casing losses.^13! Modern
Introduction 193
Fig. 6.2. Boiler efficiency
as a function of load for five
current compact wall-hung
gas appliances.!13!
Appliance thermal output-kW
boilers generally have a low thermal capacity and quick response which
makes them especially suitable in a lightweight low-energy house designed
for intermittent heating.
A common word of caution when selecting any heating system is that
too much emphasis should not be placed on quoted full-load thermal
efficiencies obtained under laboratory conditions. A unit which is slightly
less efficient but which, for example, provides more uniform heating at
low level and which responds more quickly to controls, may give lower
overall running costs. Nevertheless, it is important to consider some
measure of efficiency when selecting any heating appliance since measured
efficiencies vary from about 20 to 80%.
Both the EEC and the UK Department of Energy are aware of the
enormous scope for energy conservation in this area and are working on
directives and legislation to encourage development and selection of high-
efficiency convertors. In the US, appliances must now be sold with an
efficiency rating and an indication of how the device compares with
analogous devices.
Domestic heating in the UK is dominated by gas, in contrast to Euro-
pean coun+~:es such as West Germany and Switzerland where oil is the
principal fuel. In 1978, of the new central heating systems installed in
Britain, over 80% were fuelled by gas.f14! With increasing oil prices the
additional demand placed on gas caused British Gas some concern. Former
chairman Sir Denis Rooke said: 'It is not that we are running out of gas, but
that we have to limit winter peak demand to our total supply capacity in
the transmission and distribution lines.'!151 One measure taken was to
cancel promotion of central-heating systems and gas fires, which would
have aggravated winter demand. Sales of oil-fired central heating systems
have been virtually static in the recent pastf16' and it is more likely that
they will fall than rise as customers shift to gas, solid fuel or wood.
Electricity for heating has profited from increased oil prices, and sales
of off-peak storage heaters in particular have risen considerably. Elec-
tricity, more than most other fuels, only makes sense economically in
houses insulated to a high level. The Electricity Council is encouraging
improved thermal standards and generally advocating storage heaters to
provide the base load, with any other heating needed being provided by
direct-acting space heaters in each roomJ 1 7 !
Solid fuel is receiving more and more consumer interest, although the
Table 6.3. Comparative space heating running costs for the UK (south-east area - winter 1982-
Cost per Heat content

Sold in unit sold per unit sold
Fuel units of in pence on MJ basis Application
Bituminous coal group 2 50 kg 494 1499 Open fire

Bituminous coal group 2 50 kg 494 1499 Open fire with backboiler
Sunbrite coke 50 kg 624 1400 Room heater
Anthracite Stove Nuts 50 kg 716 1685 Room heater with backboiler
Housewarm singles 50 kg 478 1499 Smoke consuming room heater
with backboiler
Anthracite grains 50 kg 633 1627 Gravity feed boiler or backboiler
1 Unit 4.90 3.6 Electric fire
Electricity general tariff 4.90 3.6 Electric radiators on day rate
Electricity white meter 1.90 3.6 Single storage heater
Economy 7 night rate 1 Unit Multiple storage heaters
1.90 3.6
Gas 33.5 105 Radiant-convector gas fire
1 Therm 33.5 105 Wall heater
(general domestic tariff) 33.5 105 Gas-fired boiler or backboiler
15 kg
Liquified petroleum gas cylinder 835 737 Butane heater
1 litre 14.40 26 LPG-fired boiler** or backboiler
Heating oil 1 litre 21.23 37 Oil-fired boilere or backboiler
28 sec. viscosity
Heating oil 1 litre 21.85 Oil-fired boiler*
38
35 sec. viscosity
a
Annual heating requirement of a large room. Maintenance and standing charges included when
applicable.
b
Annual heating requirement of an average sized house. Maintenance and standing charges included
when applicable.
c
Difference between white meter and general tariff annual standing charges.
d
Delivered to 1200 1 storage cylinder. Rental shown in standing charge column.
e
Delivered to 2725 1 storage tank.
(Courtesy of C.M.J. Sutherland.)
present domestic consumption of ten million tonnes is well below (and

likely to remain so) the 30-40 million tonnes used annually from 1947 to
1964 [18] Appliance sales are both for central-heating installations and
supplementary heating where, for example, an open fire may serve as a
'focal point' in a room while providing a comfortable heating source.
Of course, the explanation of these trends is complex, with costs being
interwoven with sociological and political factors. For some, central heat-
ing will always be preferable for reasons of habit and convenience, and
similar considerations will mitigate against the selection of solid fuel in any
form. The pattern of occupancy and the thermal response of the building
Introduction 195
Additional Average
annual annual Cost ofa Cost of*
Average Cost per standing maintenance 4.14 GJ for 79.9 GJ for
thermal useful MJ charge cost room heating central heating
efficiency in pence (£) (£) (£) (£)
30% 1.10 — 7 165 -
50% 0.66 - 7 - 481
65% 0.69 - 7 106 -
75% 0.57 - 7 - 415
70% 0.46 - 7 - 335
75% 0.52 — 7 — 381

100% 1.36 196
100% 1.36 980
95% 0.56 10.00* - 90
95% 0.56 10.00* 410
57% 0.56 39.60 120
73% 0.44 39.60 18 120
75% 0.52 39.60 25 369
92% 1.93 — — 177 -

75% 0.75 36.00 25 - 601
75% 0.77 - 22 - 578
75% 0.77 - 28 - 580
(discussed in Chapter 3) also influence the choice. To avoid discomfort

from overheating (or under heating), the heating-system response time
should be equal to or less than that of the building. If intermittent heating
is required, the response of the heating system should be quick both for
comfort and so that the economy in running cost is not lost by the long
preheating period of a slow-response system.
Table 6.4 shows the approximate order of response of heating-system
emitters; the ranking scale is not linear and in practical terms the differ-
ence between any two emitters may not be very great. Other authorities
might also choose a somewhat different order and, say, place panel radi-
ators above column radiators. Ultimately, the precise ranking will depend
on the exact appliances selected and how they are used in a space.
In general, we believe that intermittent heating and quick-response
systems are most suitable for any house designed to take advantage of the
highly random nature of passive solar gain in northern Europe.
Table 6.4. Quickness of response of emitters^
Order of
Type of emitter response
Fan convectors or ducted warm air 1 (fastest)
Individual natural convectors 2
Convective skirting heating 3
Column radiators 4
Ceiling heating (low thermal capacity system) 5
Panel radiators 6
Radiant skirting 7
Floor heating 8 (slowest)
To conclude this section on a somewhat pessimistic note, the problem

of choice under conditions of uncertainty is one we always have with us.
Even when there is an element of relative certainty, such as the availability
of coal, it is difficult to say whether its use for domestic heating should
take the form of solid fuel, substitute natural gas or electricity. Depending
on the grades of coal available and consumer preferences, perhaps all three
will be appropriate.
It is almost becoming the conventional wisdom to state that the options
must be kept open to permit a change over to another fuel or source of
energy, but to give specific recommendations is considerably more chal-
lenging. The Solid Fuel Advisory Service, not surprisingly, recommends
that, no matter what heating system is installed originally, the provision of
a chimney is sensible since it provides a long-term fuel option and affords
a choice to future occupants. However, on purely economical grounds,
there must be an assumption of the probability of the chimney being used
during the lifetime of the house since an unused chimney is undoubtedly
an investment without a return, as well as an added source of ventilation
heat loss.
Systems using hot water or warm air have the advantage of being con-
vertible to coal, gas, oil or electricity. Electric heating, on the other hand,
cannot of course be adapted to other fuels. Looking to the future, design-
ing a system for low supply temperatures is likely to facilitate conversion
to solar or other sources of ambient energy.
Heating systems may be discussed according to the type of heat source,
the extent of the system, that is, central heating or direct space heaters,
the medium of heat transfer, the type of emitter or a combination of
these. The following material has been divided principally into solid-fuel,
water-distribution, forced-air and electric sections with appropriate cross-
references made to avoid duplication of material.
6.2 Solid-fuel heating

Table 6.5 shows the characteristics of a number of solid fuels.
Solid fuels may be used for direct heating in fires or stoves, in a central-
Solid-fuel heating 197
Table 6.5. So lid-fuel characteristics
Calorific value
Density
Fuel (MJ/kg) (GJ/m 3 )* (kg/m 3 )*
Anthracite6 35 26 750
Bituminous coal6 30 22 720
Coke6 28 11 380
Woodc'<* 18 7 420
a
Allowance for air space made.
6
Refs. [3, 19].
c
12% moisture content; mean of red oak, white elm and chestnut.
<*Ref. [20].
Fig. 6.3. Wood- and coal-

burning stove.!211
(Courtesy of Norcem, UK
Ltd.)
heating system or in a combination of both. Open fires are highly

inefficient (see Table 6.3) since most of the convected heat emitted goes
up the flue. A stove brought out into a room or arranged so that the con-
vection from the back and sides is directed into the room is approximately
twice as efficient. Stoves capable of burning traditional solid fuels as well
as wood and peat are available in a range of sizes, but in choosing one for a
low-energy house care must be taken to ensure that it is not in fact over-
sized. Wood is generally seen to complement other forms of solid fuel but
its future availability is questionable since, on a national scale, the best use
of wood may be in construction. On the other hand, in both urban and
rural situations, there is, sad to say, wood which is presently wasted and a
certain quantity of wood may be grown on-site. Fig. 6.3 shows a small
free-standing stove (J0tul) capable of burning coal or wood.
Many fires or stoves either come with or can be provided with back
boilers to supply heat for radiators and domestic hot water. In a low-
energy house, especially, it may be difficult to balance the highly variable
Fig. 6.4. Openable stove
with a backboiler.l 2 J
Loading hatch
Door
Combustion chamber
Waterway Waterway
Fire bars
Entry for combustion air
Ash tray controlled by thermostatic
damper
demand for space heating with that for hot water. It may be more reason-
able to separate the space heating demand (as supplied by both the
appliance and the radiators it feeds) from the hot-water demand by
supplying the latter with an alternative source such as an immersion heater
(at least during the summer). In this way a situation where a room might
be overheated because hot water was required could be avoided. Fig. 6.4
shows a stove with a back boiler.
A number of multifuel cookers are also available, which will provide
domestic hot water and hot water for radiators. Similar considerations of
sizing and matching supply with demand apply.
Solid-fuel boilers for central-heating systems are now often designed to
operate on wood, wood products and peat, as well as on the more tra-
ditional solid fuels. New domestic installations usually opt for a gravity-
feed hopper which need only be filled about once a day, thus eliminating
the inconvenience of the frequent stoking associated with traditional
boilers. The clinker formed by combustion must be removed every several
days. As with other solid-fuel appliances, consideration must be given to
handling the fuel and clinker when positioning the boiler and there is an
evident advantage if the fuel store can be located above the hopper. Fig.
6.5 shows a representative solid-fuel boiler.
Typically, these boilers are used in conjunction with hydronic systems
(see Section 6.3) but, occasionally, air is used as the heat-transfer medium.
One of the principal reasons for discussing solid fuels in conjunction
with ambient energy sources is that we may expect to have coal, or wood,
and the sun for some time to come. Others include the economical cost
and the low primary energy cost (see Table 6.2). Nevertheless, the
marriage is not always ideal for reasons related more to control than
convenience. We shall, however, assume, and increasing consumer interest
Fig. 6.5. Solid-fuel boiler.J2]
Flue pipe
Hopper
Flue way
Anthracite beans
or grains
Waterway
Waterway -
Clinker tray H——— Clinker ejector ram
in solid fuel indicates that this is reasonable, that design and handling
problems can be minimized. Design considerations include provision of a
conventional flue for a solid-fuel boiler and storage facilities. The boiler
may draw its combustion air inlet either from inside the space or outside,
with the latter having the advantage of minimal interference from the
ventilation system in the house. In either case the products of combustion
go up the flue which should terminate above roof level.
Storage facilities required vary, with minimum recommendations of
2.7 m3 for coke and 1.3 m3 for other smokeless fuels and coal being
standard in the past for small houses,!221 but with lower energy demands
these might be reduced. On the other hand, the ability to store solid fuel is
in some ways a major advantage since it affords a measure of independence
from the vagaries of both the weather and national supply systems.
Handling the fuel has become a cleaner and simpler operation in
modern appliances but, nevertheless, it remains a dissuasive factor for
many individuals who are willing to pay more in exchange for greater con-
venience. For those who do choose solid fuel, conversion to gas, for
example, can be fairly simple as occupants become older and less able to
cope with solid fuel.
The control problem is related to the thermal response of the house,
the pattern of occupancy and the importance of using passive solar gain in
the house. Direct-heating appliances and some central-heating ones are
likely to have a slow thermal response because of the heat capacity of the
large quantity of metal in the appliance. As the heat demand in the house
varies with ambient conditions, the appliance may require frequent manual
attention. In the case of central heating, control of the boiler is normally
by a thermostat in the water controlling a forced-draught fan. The
response is basically similar to that of other hydronic systems using gas or
oil but care must be taken to choose a lightweight boiler if a quick
Fig. 6.6. Ground-floor plan
of the Vale House (after
Ref. [23]).
response is desired. Any system of solid-fuel heating which deliberately

or inadvertently stores heat in a chimney will, of course, have a slower
response.
Both the problems and potential of solid-fuel heating are apparent in
the houses discussed below.
Fig. 6.6 shows the ground-floor plan of the Vale Home, about 20 km
to the north of Cambridge. Upstairs there are three bedrooms and a bath-
room. The following comments draw heavily on the excellent descriptions
of the house by the Vales themselves.!23' 241
The Vales restored the house and insulated it to a high level (lvalue
of walls (including glazing) 0.42 W/(m2 K); roof = 0.25 W/(m2 K); floor
= 0.54 W/(m2 K); estimated ventilation rate = 0.5 air changes per hour;
insulation is inside the brick walls). They spend the greater part of each
day at home looking after their smallholding, doing associated research
work, taking care of their children and doing the usual housework. As part
of what they refer to as their 'alternative approach', they felt that the
heating system should be very simple and designed for long life. Conse-
quently, they chose an Aga coke-burning thermal-storage cooker, which
also provides domestic hot water, backed up by two solid-fuel stoves. Out-
side the heating season, cooking is done in a small electric cooker and hot
water is provided by an immersion heater.
The cooker is located centrally in the house and connected to a pre-
fabricated insulated chimney (which gives off some heat in an upstairs
bedroom). Heat produced is stored in the 500 kg of steel within the
cooker and this, combined with a simple thermostatic control which con-
trols the draught, allows the cooker to operate under the constant-load
Fig. 6.7. The Vale
House.! 2 4 1 (By permission
of Macmillan, London and
Basingstoke.)
conditions necessary to high efficiency which, in this case, is about 70%.

It is estimated that the Aga operates as a continuous 1 kW direct space
heater (no radiators are fed) and over the heating season provides 18 GJ in
addition to about 22 GJ for domestic hot water and cooking.
Additional heat is provided by J^tul and Ulefos stoves with maximum
heat outputs of about 8 and 4 kW, respectively, but commonly the J^tul
alone, operating at its minimum output of 3 kW, is adequate - both are
used by the Vales to burn wood. Average fuel consumption of the J0tul
is about 20 kg/d (for a 16 h day). Wood is stored under cover and coke for
the Aga is stored in a bin in an outbuilding.
No heating has been found to be needed upstairs since warm air rises up
the open staircase into the bedrooms whose doors effectively serve as
openable vents.
The designers admit that the system does not offer the potential for
control or maintenance of exact temperatures of more-conventional
systems and that it has the disadvantages of requiring coke to be carried,
wood to be sawn occasionally and carried and ashes raked out. On the
credit side, however, it makes good use of primary energy, requires little
maintenance, is of high reliability and is well suited to the family's
continuous-occupancy pattern.
Recently, a large single-glazed conservatory (see Fig. 6.7) was added to
the entire south side of the house to provide a sheltered area for plant
growth and to permit passive solar gain to play a greater role in space
heating.
Another example of an existing house using solid-fuel heating, but this
time on a grand scale, is the Bailey House in SurreyJ 25 ! This large, five-
bedroom timber-frame house uses a heating system, specified and planned
by the Solid Fuel Advisory Service, based on twin-gravity-feed anthracite
boilers with a combined maximum output of 53 kW. The boilers feed
radiators throughout the house and supply domestic hot water. A twin-
thermostat system allows individual control of the burning rate of each
fire so that only one boiler need be fired in the summer when just hot
Fig. 6.8. National Coal
Board experimental low-
energy consuming house
design.t 26 !
Jl fj
LJ
FIRST FLOOR PLAN
GROUND FLOOR PLAN

water is needed. A control unit, incorporating externally fitted frost stats
which will override the thermostats inside the house to compensate for
sudden drops in the outside temperature, operates the heating.
A number of experimental designs for low-energy houses using solid-
fuel heating exist. Probably the most considered example is the National
Coal Board's Terraced House shown in Fig. 6.8. The detailed discussion of
the project, given in Ref. [26], forms the basis of the following comments.
In essence the house is a well-insulated volume, designed so that the
internal environment remains fairly steady. The calculated design heat-loss
rate is only 2.8 kW. ((/-values in W/(m2 K) are: walls, 0.3; roof, 0.25;
windows, 2.8; floor, 0.4.) Well-sealed double-glazed windows and air-lock
lobbies to the front and rear doors are used to reduce the ventilation heat
loss. A structural chimney of high thermal capacity provides chimney heat
gain to the house (in this way a house-heating efficiency of about 80% is
claimed) and 100 mm of insulation is placed external to a 225 mm
masonry wall. The exposed surface of the insulation is rendered to provide
protection against the weather and possible damage.
Solar panels are used to supply part of the energy for domestic hot
water, with an estimated 1.2 kW contribution from the heating system.
For a total maximum heat output of 4 kW at an efficiency of 80% a burn-
ing rate of 0.7 kg/h is required and, although this is within the normal
range of a conventional appliance burning smokeless fuel, the equipment
could not operate satisfactorily for extended periods at the 0.1-0.2 kg/h
burning-rate levels required during milder weather conditions. It is also
difficult to maintain an attractive visual fire appearance over the operating
range of the appliance and, if smoke produced is to be kept below a satis-
factory level, the area of water-cooled surface within must be kept to a
minimum.
Consequently, heat transfer to the house must be principally by the
circulation of warm air rather than by direct radiation from the firebed or
from hot-water radiators. A coal-fired boiler (see Fig. 6.9) enclosed within
a brick convection chamber heats the air in the chamber which is dis-
tributed by natural convection (the stairwell is open) and around the
ground floor by a small axial fan located in the partition wall between the
kitchen and living room. A proportion of the air is recirculated and pro-
vision is made for a carbon filter unit to prevent odours and moisture
being circulated around the house. Highly volatile bituminous coal appears
to be the most suitable solid fuel because of the low burning rates required.
The boiler output is controlled from a room thermostat sited in the
kitchen and a second thermostat in the lounge switches the axial fan on
and provides filtered warm air from the kitchen when required. The boiler
is fitted with an override thermostat to limit the temperature of the warm
air leaving the combustion chamber. Additional heating is provided to a
bathroom radiator and two small bedroom radiators from the domestic
hot-water system.
Fig. 6.10 shows Ambient Energy Design House 1 (see Fig. 3.17 for the
floor plans) which was developed independently of the National Coal
Board work, by the authors and Ms L. Krall. How the heating system
Space heating and ventilation
Fig. 6.9. Experimental low-
energy coal-burning
appliance.!26]
LOAD
REFUEL
4 kW coal burning warm

air appliance for low energy building
evolved in practice and aspects of the house other than the heating are
described in detail in Chapter 12. A 3.7 kW solid-fuel boiler is used for
space heating and domestic hot-water heating. The fuel store is filled from
the street and has a sloped floor towards the adjacent boiler. Solar col-
lectors on the roof are used to heat air which is ducted to a rock-bin store
under the staircase, mixed with recirculated filtered air drawn from the
kitchen and then circulated throughout the house. The boiler, which draws
its combustion air supply from outside, is used to supply hot water to a
coil in the rock store (December to February) or a fan coil unit in the air
stream (September to November and March to May). Control is simply by
a room thermostat which turns the fan in the fan coil unit on or off and a
thermostat in the air stream, which opens or closes the valve to the appro-
priate heating coil, depending on the month. The combustion air supply
which comes from outside is separated from the air supply necessary for
ventilation and space heating to permit greater control and efficiency. The
rock-bin store is insulated and is not expected to behave like an
uninsulated chimney. Wall insulation is internal and the house in these and
other ways is designed to respond relatively quickly to passive solar gain.
Water-distribution systems 205
Fig. 6.10. Ambient Energy
Design House I.! 2 7 !
6.3 Water-distribution systems
Heat sources for water-distribution systems may be solid fuel, gas, oil or
electricity, with gas the most popular and electricity the least; Chapter 5
has shown that active solar energy is also easily integrated with a water
system. Just as for solid fuels, it is possible with gas to have direct space
heaters or a radiant fire with a backboiler but it is more common in
central-heating systems to have a floor-standing or wall-hung boiler.
Oil-fired boilers (see Fig. 6.11) use a vapourizing burner or a pressure
jet to atomize the fuel for efficient combustion. For smaller domestic
installations a vapourizing burner is normally adequate. Control of the
boiler is achieved with a valve (actuated by a thermostat in the water)
which varies the amount of fuel to the burner. A useful measure to ensure
that oil (and to a lesser extent gas) systems are operating efficiently is to
analyse the flue gases to ensure that complete combustion is achieved but
not at the cost of excessive air carrying the heat away. For smaller instal-
lations, periodic use of portable analysers, which give carbon-dioxide
content or, less commonly, oxygen content and temperature readings, may
be used.
Recent developments in boilers include the incorporation of fluidized
beds, provision for multifuels and heat centres in which the domestic hot-
water facility is combined with the boiler. In Holland the Stone—Platt
Fluidfire domestic boiler has been used in experimental low-energy
houses.l28l In the boiler, tubes carrying low-pressure water are passed
through a shallow fluidized bed of aluminium-oxide particles, thus giving
very rapid heat transfer - this results in greater efficiency and a reduced
boiler size. Present models are gas fired but the principle could be applied
to other fuels.
Multifuel burners are usually designed with gas or oil as the principal
fuel with provision to shift to solid fuels, wood, propane or straw depend-
ing on the model. Adaptor kits such as swing-arm pressure jet attachments
Fig. 6.11. Oil-fired boiler.l2)
Stabiliser
Thermostat
Flame detector
Protectorelay
Transformer
Oil control valve
to allow oil to be burned can effect the switch from one fuel to another.
Manufacturers claim high efficiencies with all fuels, and sales indicate that
many purchasers find the flexibility attractive as an insurance against
shortages of a particular fuel, but some criticism of the trend has been
voiced. A Swiss report states that dual-fuel boilers (gas or oil plus a solid-
fuel option) cannot be designed for maximum possible efficiency since
combustion space and heating surface are a compromise between the differ-
ent fuels J 29 J Over the life of the boiler, therefore, a great deal of energy
will be wasted and the solid-fuel option hardly ever used, it is claimed. The
alternative suggestions are a twin combustion chamber boiler or, better, an
independent solid-fuel boiler for those times when another fuel - usually
wood - is available.
Heat centres (see Fig. 6.12) are a combination of a boiler, an insulated
hot-water cylinder, a circulating pump, an electric programme control,
thermostats (boiler, high limit and cylinder) and a flowshare motorized
valve.!30 J Advantages of the system include reduced piping runs (and
associated heat losses), fast recovery on domestic hot water owing to the
higher efficiency, elimination of any need to oversize the boiler to ensure
adequate temperature since the flowshare valve allows full output from the
boiler to go to the central-heating system whenever necessary, and ease of
installation.
Standard boilers burning oil usually require a conventional flue,
Fig. 6.12. Interior of a heat
centre.!3°1 (Courtesy of
Harvey-Habridge Ltd.)
although some can use a balanced flue (see Fig. 6.13) as most gas boilers
do. The great advantage of balanced flues is that they need no chimney.
Current distribution techniques for hydronic systems in standard
domestic applications include traditional two-pipe, one pipe with single
entry valves, one pipe with conventional connections and microbore.
Typical water temperatures are 82 °C flow and 70 °C return.
Two-pipe systems (see Fig. 6.14), which have separate pipes for flow
and return, are simple to design, provide water at approximately the same
temperature to all emitters and employ either radiators or convectors.
Possible disadvantages are an increase in cost and a more obtrusive appear-
ance.
One-pipe systems (see Fig. 6.15) serve radiators progressively with the
main flow, not through the emitter but bypassing it. Connections to the
emitters are taken off in series and if there is flow through several
emitters the flow to the second in a group includes part of the return from
the first. Average temperatures in successive emitters are thus reduced and
the size must be increased to compensate for this. Also, since the flow
through emitters depends on the resistance of the emitter compared to the
resistance of the bypass pipe, high-resistance emitters such as convectors
may not be suitable. A principal advantage is greater simplicity of piping
and, hence, installation. In the past, one-pipe systems were considered
Fig. 6.13. Balanced flue
appliance.! 31 !
Combustion chamber
and heat exchanger
Fig. 6.14. Two-pipe system. Local control valve

Lockshield valve (for balancing system)
Fig. 6.15. One-pipe system Local control valve Lockshield valve (for balancing system)
(with conventional con-
nections).
n—a
most suitable for small installations and, with low-energy houses requiring
less heat and fewer emitters, they may be perfectly appropriate.
Local control valves for both two- and one-pipe systems may be either
manual or thermostatic (automatic). Special single-entry thermostatic
valves (see Fig. 6.16) exist for single-pipe systems which provide a perma-
nent bypass for most of the flow and direct the rest to the radiator.
Microbore systems (see Fig. 6.17) employ smaller pipe sizes (commonly
Fig. 6.16. One-pipe system 6, 8 and 10 mm diameter) and feed manifolds (with more-conventional
with single-entry thermo- 22 mm diameter pipe) rather than individual radiators. One of their main
static valve. potential attractions for low-energy homes is the inherent lower water
content which gives a faster response to both heat outputs and thermo-
static controls.
The most common piping material is, by far, copper in small instal-
lations, with steel sometimes used in larger projects. Cross-linked poly-
thene tube, recently introduced by British Steel, may become popular in
the future.
Panel radiators, usually of pressed steel, are the most common form of
emitter. In spite of their name about 70% of the heat emitted is convective.
Principal types available are simple single panel (see Fig. 6.18), double or
multiple panel and single panel with a finned back to increase the heat out-
put. In general, radiators respond poorly to control because of their high
water content but, of course, in a low-energy house the radiator size will
be reduced and thus so will the amount of water. A typical output from a
single-panel radiator 1.0 m X 0.7 m and of 40 mm depth, operating at a
Double-entry valve Fig. 6.17. Microbore system
(after Ref. [32]).
Manifolds
Fig. 6.18. Front and rear

view of single-panel pressed-
56 K differential between mean water and room air temperatures, is steel radiator.!33 1 (Courtesy
1.9 kW/m2 ;I331 radiators of increased thickness give higher outputs. of Hudevad, Gt. Britain.)
Natural convectors (sometimes known as convector radiators) (see Fig.
6.19) consist of a steel frame inside a casing whose design encourages
movement of air over the frame by the stack effect. Since they have both
less metal and less water than panel radiators they heat up more rapidly
and respond more quickly to control. Heat output is approximately
2.2 kW/m2 for a 1.0 m X 0.5 m convector of width 89 mm operating at a
60 K differential between room air and mean water temperatures^34! and
this can result in significant savings in both cost and space. In spite of
these advantages, public acceptance has been slow but, for low-energy
houses, particularly those that are intermittently heated, they would seem
to be well suited and so interest is likely to increase.
Fig. 6.19. Natural con-
vector. I 34 1
Clip-on
casing
Air
Lock bleeder
shield valve
valve
Fixing holes
Aluminium (chassis reversible
heat exchangers for left or right hand
Air warmed independently fixing)
by both banks of fins
Fig. 6.20. Radiator location rodiotion losses to window ore bo lanced by rodiation
in traditional houses, t 2 1 from rodiotor ^
il
no staininqof
wa||
free location
of furniture -v /down
/ drauqht
counteracted
\
- rodiator
Fan convectors, as their name implies, use a fan to blow air over the
heater battery rather than relying on natural convection and so are capable
of higher heat-transfer rates. Individual control is very easy since a time-
switch and thermostat can be used to govern the fan and hence the output
of the device. They have two disadvantages, however. The first is that an
electrical supply is required for each unit but this may be compensated for
by reduced pipe runs and the smaller number of appliances required due
to the flexibility and greater strength of these emitters. The second is that
they tend to be noisy.
Radiators and convectors have, in the past, commonly been positioned
under windows (see Fig. 6.20), but in low-energy houses with reduced
draughts, double glazing and smaller heat loads this is less necessary. Also,
some low-energy houses incorporate floor (or close to the floor) to ceiling
glazing on the south wall and here emitters should be brought back into
the room to prevent obstruction of incoming solar radiation.
For radiators which are located on external walls, research has shown
that reflecting materials behind radiators can significantly reduce the
amount of heat loss through the wall, the exact extent depending largely
on the {/-value of the walU35J The material used can vary from inexpen-
sive cooking foil to higher-quality aluminium foil.
Underfloor heating may be used exclusively or in conjunction with
other emitters as in the case of underfloor heating on the ground floor and
radiators in the first-floor rooms. Either metal or plastic piping may be laid
coiled in a screed whose details of construction will vary depending on
whether or not it is desired to take advantage of the thermal capacity of
the floor. Floor surface temperatures are usually limited to 30 °C to avoid
discomfort. Approximately half of the heat emitted may be as radiation
and the system has the advantage of maintaining an even-temperature
gradient. A number of low-energy houses have adopted underfloor heating
because the low temperatures employed (flow temperatures are commonly
less than 50 °C) are easily supplied by either conventional equipment or
alternative sources such as solar panels and heat pumps. It should be noted,
however, that it is also possible to operate radiators and convectors at
temperatures as low as 40 °C. Performance will of course be lower and the
size must be increased, but not necessarily to an unacceptable point, since
the heat load to be supplied in a low-energy house is less.
Other emitters for use with hydronic systems include radiant panels,
radiant strips and skirting heating (of three types: convective, radiant—
convective and radiant). Details of these and further information on the
emitters described above may be found in standard references.^2' 41
Water-distribution systems commonly use a small, glandless, electrically
driven centrifugal pump (see Fig. 6.21) which may or may not be supplied
with the boiler. Gravity systems, which require large-bore piping and low-
resistance emitters, but have the advantage of being independent of a
power supply, are now uncommon.
Allowance for expansion of water is normally by the provision of a cold
feed (for replacement of water if necessary) and expansion tank in an open
system (see Fig. 6.22). In a sealed system (see Fig. 6.23), which may be
required if height in the building is limited, a diaphragm or membrane tank
(see Fig. 6.24) together with a safety valve and pressure gauge are used. A
sealed system can operate at a pressure above atmospheric and at a tem-
perature higher than 82 °C, in which case, to avoid the possibility of
accidents as a result of contact with surfaces, radiators are not recom-
mended — convectors or skirting heating are more appropriate and may be
more economical. A detailed comparison of open and closed heating sys-
tems is given by RockhilU37 !
Almost all hydronic systems include provision for a domestic hot-water
supply. Fig. 6.25 shows a typical example where the hot-water supply
comes from an indirect hot-water cylinder fed from the central-heating
boiler. Here an immersion heater in the indirect cylinder heats the water
during the summer when the central heating is switched off but, depending
on the system, it may be more efficient to simply use the boiler.
Fig. 6.21. Domestic circulat-
ing pump (approximate
dimensions 130 mm diameter
byl30mmdepth).[36]
(a) Installed, (b) Cut-away.
Controls, for any heating system, are an integral, if at times under-

valued, part of the effort to save energy. Fisk has pointed out the complex
interaction between occupants and controls in domestic heating, which he
sees as probably the unique situation in which those who benefit from the
controls also pay for themJ 3 8 ! Consequently, a failure on the part of the
controls to maintain comfort conditions may be acceptable if the failure is
accompanied by reduced costs. The control system, to be successful, he
Fig. 6.22. Open distribution
system.l 2 !
Cold water main -
t}
Open vent from j Cold water feed to heating circuit

primary flow (no
valves or means
of isolation) r
r
Hot water primary •

connections J
OO—O—tXl ) Heating flow
—<x3—Q"~M—(-Heating return
Convectors Fig. 6.23. Sealed distribution

system (schematic represen-
Mi Mi F-h tation after Ref. I 3 ) ) .
Pump
Expansion
Safety vessel
Boiler
valve
^Diaphragm Fig. 6.24. Diaphragm expan-

sion tank.!37] (a) When
system is filled, no water
enters tank when cushion
and water pressure are in
equilibrium, (b) As tempera-
ture increases, diaphragm
moves to accept expanded
water, (c) When water rises
Closed expansion tank (with diaphragm) to maximum, full acceptance
of expansion is achieved.
adds, must be matched with the intended system of energy management

by the occupants.
In older hydronic central-heating systems it was common for the boiler
to remain under the control of its own boilerstat as long as a timeclock or
programmer indicated that a need for heat might arise. This involved
necessarily higher losses than when internal temperature sensors keep the
boiler closed down when there is no demand for heat, and so the next step
was to use a single room thermostat to control the system. This, in turn,
has the disadvantage of controlling the whole house from one source and
so the only room accurately controlled is the one which has the thermo-
Fig. 6.25. Domestic hot- vent for hot water system heoting
water supply using an ball valve feed t expansion astern
indirect hot-water cylinder cold water cistern vent for primary circulation /
/
fed from a central-heating warning pipe r ^l . «\ /
boiler.! 2 1 warnmq r
<"U
' b a l l valve
cold water feed for L—cold water

hot water supply I feed for
central
heating
I
> indirect
ma J
--« /I hw
cylinder j
I
i
central heating flow
bath*
H,
hot water distributing I J low 6 !

pipe "return I
i
boiler I sink
1"
Note; service pipe omitted for clarity central heating return
stat, the others either underheat or overheat. However, it has been

objected that good design and, especially, correct sizing of the heat
emitter, minimized this problem. The argument then went that even the
best design could not properly anticipate the effects of other heating
influences in the house, whether these be from passive solar gain or casual
gains from, say, equipment. To make the best use of these gains, individual
room controls are required.
In hydronic systems these commonly take the form of thermostatic
radiator valves (see Fig. 6.26) located on the flow pipe just before it enters
each radiator (that is, the local control valve of Fig. 6.15). Little infor-
mation exists on the precise amounts of energy to be saved by the use of
these valves (and by alternative control strategies in general)!39! but a
number of authorities view them as especially suitable for domestic com-
fort and economy.!2' 4°1 (Burberry & Aldersley-Williamsl2! provide a
valuable detailed guide to a number of control systems.) Because, as
mentioned previously, overheating by 1 K results in an approximately
Temperature Fig. 6.26. Thermostatic
radiator valve.t 41 1
sensor
Hand
wheel
Compression
pin
Mushroom valve
5-10% increase in energy consumption, accuracy of control is essential

and it has been suggested that room thermostats and thermostatic radiator
valves with differentials of less than 0.5 K be used J 4 °l Larger differentials
cause the occupants to set the thermostat higher in order to prevent the
space temperature falling below the desired level.
Thermostatic radiator valves are best used in conjunction with a time
control of the pump and boiler, which may be quite simple, giving, say,
one or two periods of operation for the heating and hot-water service per
day. A variation can be used to separate heating and hot water if desired.
In the future, microprocessors and other similar small processors are
likely to be used to augment the enormous potential of controls.I42' 431
As suggested in Chapter 1, it will be possible to provide considerably more-
complicated control programs. A simple application will be the use of the
boiler at the beginning of the day for heating only (rather than heating and
domestic hot water) to speed response. The memory facilities of 'micros'
will permit temperature regulation according to the period of the day to
allow, for example, energy savings when the occupants are more active.
Programs will also vary from day to day and season to season to allow for
weekends and holidays and any other changes.
'Intelligent' programmers will be used for optimum start control. They
will learn the heating curve of a house and then, on the basis of external
and internal temperatures, will switch the boiler on so that the desired
temperature will be achieved at the proper time.
Other capabilities include monitoring the systems for faults and planned
maintenance and energy accounting to provide homeowners with records
of energy consumption and costs.
The real challenge with microprocessors is with the associated hardware
of the control system, that is, the devices that interface with the micro-
Fig. 6.27. System diagram of
the BRE Solar Energy
House.! 44 ! (BRE, Crown
Copyright, HMSO.)
processor. For example, it would be very easy to incorporate in the micro-

processor program facilities for turning off a local heating source, such as a
radiator with a thermostatic valve, if a window in the space is open. How-
ever, the cost of the sensing equipment and the wiring back to the micro-
processor might discourage the inclusion of such a facility.
In low-energy houses and particularly those which are separated into
zones, intermittently heated and designed for passive solar gain, individual
space control is very important. For houses intended to maintain fairly
uniform environmental conditions throughout the house and during the
day, a single room thermostat (sited properly) may be adequate. Similarly,
in such houses, the importance of low thermal capacity heating systems is
less. The following examples illustrate the use of water-distribution sys-
tems for space heating, in some cases in conjunction with active solar
collection.
In the BRE Solar Energy House space heating is provided by hot-water
natural convectors equipped with thermostatic radiator valves and fed
from microbore piping (see Fig. 6.27).f44l The radiators are of larger than
normal size to permit lower water temperatures (from the solar collector
and heat-pump system) to be used. When the temperature of the main
35 m 3 storage tank is sufficiently high, heat for the radiators is taken from
it via a heat exchanger. At other times the supply is drawn from the 1 m3
well-insulated tank which is heated during off-peak hours by a small water
to water heat pump that uses the large tank as its heat source.
Underfloor heating is used in one of the Salford Strawberry Hill
Forced-air systems 217
Feed and expansion tank Fig. 6.28. Salford Straw-
berry Hill House with
underfloor heating, t 4 51
Heat pump | Exhaust air

compressors Heat pum to outside
liquid
receivers
Houses!451 (see also Section 6.6). Fig. 6.28 shows the approximately
180 m of flexible reinforced pvc hosepipe (of 12.5 mm bore) which was
set in the screeds over the concrete floors. The house was deliberately
designed to be heavyweight and heat is stored both in the floors and walls
(brick exterior, concrete block interior and cavity filled with blown poly-
urethane granules). Although the minimum lifetime of the pipes is not
known, it is expected to be long since the screed provides protection from
both mechanical damage and ultra-violet radiation and because the maxi-
mum anticipated temperature is 40 °C. The pipe was laid using 30 m loops
from two headers at each floor level with the geometry adopted permitting
standard pipe lengths to be used for the circulation of a large volume of
water at low pressure drop. No balancing valves were required.
6.4 Forced-air systems

Forced-air systems comprise the following three principal groups:
(1) ventilation systems;
(2) heating systems which may or may not be combined directly with
ventilation systems;
(3) combined heating and ventilating systems which incorporate heat-
recovery devices.
In the domestic context Europe has consistently been ahead of the UK in
developments in energy conservation in general, and forced-air systems in
particular. Over 30 years ago Scandinavia started a trend by equipping
blocks of flats with mechanical exhaust systems.l46l With an increasing
need to conserve energy and greater awareness of the importance of the
Fig. 6.29. Ducted air heat- Plenum
ing.l 4 ?] (Cavendish Ltd.) 90° Plenum take-off
/ ^ R e t u r n ^_^
air g r i f l e r ^ ^ Register box
— 2% in. stack
Angle boot
Reversible
elbow 45° elbow
,— Side elbow
_ Hat elbow
(90°) ^Depth
Plenum base
Saddle
Reducer -,
End boot
45°
adjustable 90 (
elbows adjustable
elbow
ventilation heat loss in well-insulated buildings these simple ventilation

systems are now being supplanted by balanced systems with both mech-
anical supply and exhaust. In France, in 1977, about two-thirds of all
dwellings built had some form of mechanical ventilation J 46 1 Systems
designed solely for ventilation are commonly used with radiators or con-
vectors or electric heating but are applicable to almost all forms of heating
systems.
The great advantage of a balanced mechanical ventilation system is
that, at least in theory, it permits precise control of the air entering and
leaving an enclosed volume and, consequently, the heat loss associated
with that air. But, as we saw in Chapter 3, it is necessary to reduce the
effective open area for natural ventilation to a small value.
Warm-air heating systems account for just under 10% of the market in
the UK for domestic central-heating systems, but this figure may grow in
the future as their potential for greater control over air temperature,
moisture content and ventilation becomes more valued. A traditional
objection to air systems is that they are generally bulkier than hydronic
systems and although this remains true in a low-energy house the duct
sizes will tend to be smaller than in the past and should pose no serious
problems if the duct runs are considered at an early stage in the design. An
important advantage is that thermal response is more rapid than in
hydronic systems. Fig. 6.29 shows the elements of a ducted warm-air
system with the combustion air for the heater control unit coming from
within the space heated.
The heat source may run on solid fuel, oil, gas or electricity, with gas
being most common in the UK. Warm-air units are divided into two broad
categories - direct fired, where heat is generated in the primary source and
transferred to the air stream by a heat exchanger that may be as simple as
refractory bricks or steel plate; and indirect systems that use water to
carry heat from a boiler which can also serve the hot-water system. The
Fig. 6.30. Central heating
by natural convection.! 49 ]
efficiency of some direct warm-air heaters recently approved by British

Gas is about 75%.l48l
Normally, a centrifugal fan is used to distribute fresh air mixed with
recirculated air through the ducts, but some air systems dispense with both
ducts and fan and rely simply on natural convection. Fig. 6.30 shows the
circulation pattern in a house heated by a gas-fired air heater that warms
a chamber which is an integral part of the house and is made of the same
materials as is used for the walls, that is, brick, load-bearing block or con-
crete. Air in the heat chamber is warmed and rises by convection to be dis-
charged by high-level grilles into adjoining rooms. Circulation is set up and
cooler air returns to the chamber via low-level grilles, thus continuing the
cycle. Upstairs rooms are warmed by conduction through the ceiling and
air movement up the stairway. Such a system is extremely simple and may
be suitable for an unzoned house of slow thermal response and fairly con-
tinuous occupancy.
In systems with fans, attention must be paid to selecting a fan which
will distribute sufficient air without creating excessive noise. There is a
certain amount of flexibility between fan power and duct size, with a
larger fan compensating for a smaller duct in cases where space is limited.
Ducts are commonly made from galvanized sheet steel, although glass fibre
and certain plastics are suitable, in a circular or rectangular section, and
may be fabricated with insulation such as polyurethane or may be
insulated afterwards. Duct sizes vary of course but in a small well-insulated
house will probably be no greater than 300 mm X 300 mm for the main
runs and considerably smaller for most of the others. Air is introduced into
spaces by grilles which are preferably located in the floor but may be in
the walls or ceiling. Ideally, extract grilles are sited in the diagonally
opposed position of the space. Supply grilles are often fitted with a
damper which may be controlled manually or automatically to regulate
the air flow to the room. This is a consideration of some importance in a
low-energy house, in which the maintenance of lower temperatures in
unoccupied areas can result in significant savings. Automatic control of the
Fig. 6.31. Warm-air heating
with control of the supply
air.150]
air supply and the temperature in a space may be achieved by motorized

dampers connected to a room thermostat, but for a domestic setting this
is likely to be both expensive and complicated and hence prone to failure.
Typical controls are much simpler (and cruder than in hydronic systems)
and consist most often of a room thermostat situated in the living room,
which controls the operation of the air heater which then controls the fan;
a timeswitch is also often used.
Standard references deal with air velocities in the ducts, the 'throw'
from registers and siting of supply, return and transfer grillesJ4> 7 ' As for
radiators in hydronic systems, the reduced heat requirement and more-
uniform thermal environment (due to double glazing, reduction of
draughts, etc.) of a low-energy house will tend to permit greater flexibility
in locating registers.
Fresh-air supply in traditional warmed-air heating systems is mainly
from infiltration and enters the warmed space directly. With more-tightly
sealed houses and greater control of ventilation and air supply in the
house, however, it is preferable to draw the fresh air from outside the
heated volume. Fig. 6.31 shows a proprietary system which takes a con-
trolled amount of fresh air from the roof space and mixes it with a varying
amount of recirculated air for heating and circulating to the rooms. Com-
bustion air exits by the heater flue and heated air is lost through exfil-
tration (in an uncontrolled manner).
The scope for energy conservation is considerably increased if a heat-
recovery device is incorporated in the system. This is easily done in a
balanced system which has inlet and exhaust streams in close proximity at
some point. Usually, a plate-type heat exchanger is used for individual
houses, with a thermal wheel or heat-pipe unit becoming practical for
blocks of flats. Run-around coils are often used in commercial office
buildings and heat pumps may be used for heat recovery in a variety of
applications. Plate heat-recovery units are passive devices. They act as
Fig. 6.32. Heat-recovery
devices.!53, 54] (j) piate-
type heat exchanger.
(b) Thermal wheel.
(Courtesy of the Electricity
Council).
(a)
counterflow heat exchangers by transferring heat from the exhaust stream

to, for example, thin aluminium sheets. Incoming air on the other side of
the sheets is simultaneously warmed and heat-recovery efficiencies of
60-70%, depending on the differences in temperature, are obtainable.!511
Fig. 6.32 shows a plate-type heat exchanger and a thermal wheel — the
latter normally requires an electrical input to turn the wheel but has the
advantage, at least in some designs, of facilitating latent-heat recovery in
addition to sensible-heat recovery. The efficiency of such devices can be as
highas80%J 52 l
In general, warm-air systems and hydronic systems can be equally suit-
able for low-energy houses, with warm air perhaps preferable in inter-
mittently occupied, fast thermal response dwellings. The scope for con-
trolled ventilation and heat recovery may, on grounds of simplicity of
incorporation, favour warm-air systems. Integration with solar heating,
whether this be mere preheating of incoming fresh air or an active solar
collection system with storage, is also feasible. Recent developments in
component design and control have led to a high degree of comfort (that
is, in particular, low vertical thermal gradients) and less noise. The follow-
ing examples illustrate the use of forced-air systems in dwellings.
In the Milton Keynes Solar House described previously (see Fig. 5.16)
space heating is provided by a single fan convector fed with water heated
by a combination of active solar collection and a conventional 6.5 kW gas-
fired boiler J 55 1 The open staircase of the house permits heated air from
the convector, which is located on the ground floor, to circulate through
the house. Water as low as 30 °C can supply the convector coil.
Fig. 6.33. Schematic of
forced-air heating and
ventilation layout in a
Cambridge house (scale
1:200). (a) Ground floor.
(b) First floor.
Fresh-air
inlet in
roof void
It is of interest to examine in detail the evolution of one component of

the above system, the fan convector, because it illustrates the common
problems of correctly specifying material and finding suitable equipment
commercially. The fan convector produced by the manufacturer contained
a belt-driven centrifugal fan for which a 180 W three-phase supply was not
available at the house (for houses in the UK, single-phase 240 V supply is
standard). Consequently, a 650 W single-phase motor was fitted as the
only readily available model; actual consumption was determined to be
500 W. Analysis of the performance of the house indicated that the
primary energy consumption of the fan was actually greater than the solar
space heating contribution.!56J In part because of this, a detailed study of
possible replacement of the unit was made and a direct-drive centrifugal
fan with a power consumption of 200 W (when suitably modified) was
selected; primary energy consumption with the new unit has been calcu-
lated to be less than the useful solar contribution.
Fig. 6.33 shows the forced-air proprietary heating system of Fig. 6.31
installed in a home in Newnham, Cambridge (see also Chapter 12).
The design heat load of the house is 7.5 kW and a gas-fired air heater
designed for a maximum output of 9.4 kW was selected. Domestic hot
water is provided by a circulatory water heater (rated at 4.4 kW) situated
inside the air heater, using the same flue and gas connections but operating
independently and under separate controls; if necessary, supplementary
hot water is from an electric immersion heater in the same cylinder. One-
half an air change per hour, or about 190 m 3 /h of fresh air, is drawn in by
a 300 W centrifugal direct-drive fan in the heater unit, mixed with an
amount of recirculating air which varies according to the heat load and
distributed to the spaces through ducts insulated with 50 mm of glass
fibre; the largest duct size is 275 mm X 175 mm. Downstairs, supply
registers were sited, on the perimeter system, in the floor to deliver air
reasonably uniformly throughout the space. Upstairs, air is delivered
similarly. Return grilles were placed on the opposite wall at ceiling level
and return ducts take the recirculated air back to the heater. Space was
left in the 'false loft' for a small heat exchanger to be installed at a later
date and the roof was designed specifically for future solar preheating of
the fresh air. This was achieved by deliberately leaving a path for air flow
from the eaves to the ridge between the insulation and tiles.
Fig. 6.34. Salford Straw-
berry Hill House with
forced-air heating.[451
Extract air from

bathroom and kitchen
Domestic hot
water pump
Preheat
cylinder
Heat pump
evaporator
Heat pump
evaporators
in water tanks
Meat pump Heat pump

compressors liquid receivers
Forced-air heating was adopted as an alternative choice to underfloor

heating in a second Strawberry Hill Housel45 ! (see also1 Section 6.6). Fig.
6.34 shows the basic features of the system which, instead of relying on
the quick-response capability of air systems, opts for a relatively uniform
environment and a heat store within the living area. The heat store is a
450 mm cavity in the wall dividing the living room and hall from the
dining room and kitchen, with the cavity filled with bricks separated by
air passages. The brick stack is heated at night by a closed circulation of
air. A second closed circuit is used to circulate air through a ground-floor
ceiling void when heat is required in the house, with control of the fan
being by a thermostat in the living room.
One of the first low-energy houses to use a forced-air system was the
Danish Zero Energy House (see Chapter 3). Because the house is entirely
heated by solar energy with low supply temperatures inevitable at the end
of the heating season, the radiator sizes required were large and it was
judged that this was not compatible with the desired architectural appear-
ance ;l53l additionally, it was felt that the relatively large volume of water
in such a system would make it slow in response and so waste energy.
Floor heating was thought to be ideal because of the low inlet temperature
and the general demand for a high degree of comfort, but the floor con-
struction did not permit its use and so a fan-coil system was selected
instead. Six identical units (three are located in the half of the house
shown in Fig. 6.35) consisting of a return grill, heating coil, fan and outlet
grille (see Fig. 6.36) are used for heating and ventilation in the house. Each
Fig. 6.35. Heating-system
schematic for one-half of
the Zero Energy HouseJ 53 !
Fan-coil unit
Fig. 6.36. Lengthwise sec- Air

tion through a fan-coil ' duct
unit.[53]
m
Return Heating Fan Outlet
grill coil grill
unit supplies approximately 700 W at a water temperature of 45 °C and a

room temperature of 23 °C. Air volume per unit is 140 m 3 /h under design
conditions and was deliberately kept low to reduce the noise level. The
water supply for the units comes from a water store charged by solar col-
lectors. Individual electric room thermostats, which allow automatic
temperature reduction at night, control starting and stopping of the fans.
Ventilation air is introduced to each half of the house through inlet
ducts in the utility racks by a separate supply fan and is distributed to the
individual spaces. Return air from the rooms flows through the utility
racks themselves to the bathrooms where it is expelled. A purpose-made
plate heat exchanger (see Fig. 6.32) is used to recover heat from the
exhaust air in each half of the house and transfer it to the incoming supply
of fresh air. By paying particular attention to pressure losses in the system
it was possible to use low-power supply and exhaust fans which together
consume only 26 W.
Heating with electricity 225
Storage radiator Fig. 6.37. Storage radi-
ator. [ 5 71 (Courtesy of the
Electricity Council; current
Control models are slimmer and have
different controls.)
Insulation
Element
Storage block
6.5 Heating with electricity

The advantages of electricity include convenience, low maintenance,
cleanliness, relatively inexpensive installation costs and an absence of
transportation problems. Direct heating of individual spaces can be by
electric fires, convectors, fan heaters and infra-red heaters, and indirect
heating can be provided by storage radiators (see Fig. 6.37) which take
advantage of the less-expensive rates for night-time electricity produced
when demand on the national grid is lowest and when the most efficient
stations are in operation.
Standard storage radiators have no moving parts and rely simply on
storage blocks of a brick-like material with a high thermal capacity being
charged and then discharging heat by radiation and convection from the
casing. This lack of control means that storage heaters tend to consume
more energy than is necessary, the most notorious case in a changeable
European climate being when a warm day follows a cold day and windows
are opened to dissipate excess heat. This problem is accentuated in a house
designed to take advantage of passive solar gain since, ideally, instan-
taneous shut-down of the conventional heating system is desired when
solar energy can meet the energy demand. The other principal objection to
storage heaters in any house is their tendency to be large, heavy, immov-
able objects.
Storage fan heaters overcome the control problem somewhat by incor-
porating better insulation and a fan which is controlled by a built-in or
remotely sited thermostat. This provides a quicker response but involves
greater bulk and higher costs.
A central-heating system called 'Electricaire', based on the storage
heater principle, has been developed by the UK Electricity Council (see
Fig. 6.38). The heart of the system is a thermal storage unit with a well-
Fig. 6.38. 'Electricaire'
warm-air storage heating
Outlet
system.! 57 ! (Courtesy of
register
the Electricity Council.)
Insulated
ducting
Electricaire unit
insulated core. Air passed through the core by a fan in the unit is heated,
proportionately mixed with cool air to maintain a desired temperature and
discharged through ducts to each heated area or room. Rising costs of
electricity have discouraged adoption of the system since its introduction.
An alternative approach which is now advocated by the Electricity
Council is to combine multiple storage heaters for the base load, with any
other heating needed being provided by direct-acting space heaters in each
room with extensive use of thermostatic control J 17 1 Such a choice may
raise the effective efficiency of the storage system from 70% (for storage
radiators producing about one-third of their heat when not required) to
close to 100%J39l It will be appreciated that adoption of any particular
system depends on an interplay of factors such as costs of night-time and
day-time electricity and meteorological information to correctly estimate
the base-load component and that all of these factors are only too variable.
What is certain, however, is that electricity is only a sensible choice if the
house is very well insulated.
Realizing this, the Electricity Council has instituted awards schemes for
homes which make the most economical use of energy for water and space
heating. For example, in the Medallion Awards Scheme for private builders
the following regulations (among others) must be complied with:I 58 ' 591
(1) Homes should be all-electric (sole source of energy for heat, light
and power).
(2) No fireplaces or flues are to be included in the design.
(3) A high standard of thermal insulation must be met (new dwellings,
for example, must have a {/-value of 1.4 W/(m2 K) for the walls
including windows, and roofs a [/-value of 0.3 W/(m2 K)).
(4) Design and control of domestic hot water and space heating must be
such that at least 50% of the energy required is consumed at the
cheaper night rate.
Heating with electricity 227
Fig. 6.39. Electric floor
Light-weight insulating warming.! 57 J (Courtesy of
^ inner leaf the Electricity Council.)
Thermal
insulation
Wood flooring > Floor joist Electrical element in Fig. 6.40. Electric ceiling
insulating envelope. heating.! 5 7 1 (Courtesy of
the Electricity Council.)
100 mm mineral
fibre insulation
Element $%?
connections
Ceiling finish Ceiling mateirial '
It is a measure of the confusion that reigns in the energy-supply industry

that two major campaigns with opposing goals are being waged simul-
taneously, namely to incorporate chimneys in houses and to exclude them.
Since, in general, the approach advocated by the present authors is one of
flexibility we find the first two conditions listed above indefensible.
Other ways of using electricity include floor warming and ceiling heat-
ing. In the case of floor warming, heating elements are embedded in a
concrete floor slab during construction (see Fig. 6.39). The concrete slab,
which is insulated, then acts as a storage radiator with control usually
either by a room thermostat which varies the degree of charge given to the
floor or by a combination of inside and outside sensors which controls the
degree of charge more precisely.
Ceiling heating consists of a number of flexible low-temperature radiant
elements which operate at about 40 °C. They are usually in sheet form
(see Fig. 6.40) and fixed between the ceiling joists and the final ceiling
finish which is most often plasterboard. To avoid excessive heat transfer
upwards, a high level of insulation is placed above the elements. Control is
usually by a wall-mounted air thermostat controlling the elements in the
room or space where it is sited.
Houses employing electric heating seem to be associated more with
energy conservation than the use of passive solar gain. For example, the
'Guildway Home', selected by the Electricity Council as illustrative of the
approach they would like to see furthered, has a floor area of 143 m2 and
an insulated timber construction which gives {/-values of about
0.40 W/(m2 K) for the walls (excluding glazing).f60^ Background tempera-
tures are maintained by storage heaters and additional heat is provided by
thermostatically controlled direct heaters. No specific reference is made to
designing for passive solar gain.
Another Medallion Award development of detached two-storey, four-
bedroom houses with a floor area of 116 m2 uses electric ceiling heating
which covers about 70% of the total ceiling area J 6 1 ' In ground-floor
rooms the sealed elements are stapled to the joists below 80 mm of glass
fibre and above the plasterboard; first-floor arrangements are similar but
with 150 mm of glass fibre between the joists in the roof space. Each room
is under individual thermostatic control. A 2.6 kW storage heater sited in
the hallway provides a base heating contribution.
6.6 Heat pumps

Heat pumps extract thermal energy from a low-temperature source, raise
the temperature of that energy and then exhaust it to a high-temperature
sink, thus increasing its usefulness by the input of work. The most com-
mon arrangement for this is an electric motor driving a compressor and in
the ordinary household refrigerator heat is removed from the food, raised
in temperature in the compressor and exhausted to warm up the kitchen.
In principle a heat pump is identical to a refrigerator - only the objectives
differ in that the heat pump is used to heat the high-temperature sink and
the refrigerator to cool the source. Typically, a refrigerator will deliver
about twice as much heat into the kitchen as is put in electrically to drive
the motor J 62 1 Heat pumps can improve upon the figure and therein lies
their relevance to energy conservation.
The most common cycle used in heat pumps is the vapour compression
cycle shown in Fig. 6.41. The system is sealed and filled with a refrigerant.
In the cycle, refrigerant vapour is compressed and pumped to the con-
denser where heat is given out. The condensed liquid then passes through
the throttle (or expansion valve) and cools before going to the evaporator
where it takes up heat, evaporates, and goes on to the compressor, so con-
tinuing the cycle.
Electricity is most commonly used to drive the compressor via a motor
and, although convenient, it has the usual disadvantage of the low ratio of
primary to delivered energy supplied by the national grid (see Table 6.2).
Alternative power units include internal combustion engines using oil or
gas as fuel, diesel engines and external combustion engines.
The source of heat to the evaporator is most often the atmospheric air
but it can also be the ground or water; Sumnerl62^ discusses the merits of
these options in detail. The ground and water have the advantage of being
less variable in temperature but their use as a heat source usually involves a
cost penalty. In the US work has also been undertaken on combining solar
energy and heat pumps with preliminary results indicating that, except in
Heat pumps 229
Fig. 6.41. Vapour com-
I Mechanical Circulation I pression cycle.!631
Heat injected
I into dwelling
CONDENSER
Energy input| COMPRESSOR! THROTTLE VALVE
Heat
EVAPORATOR extracted
from outside
source
special circumstances, the combination is not cost-effective although indi-

vidually both measures are likely to beJ 6 4 l
The ratio of the heat output to the work input is known as the
coefficient of performance (sometimes specified as the coefficient of per-
formance for heating) and is given for an ideal (Carnot cycle) machine by
COP = 'sink
•*sink -*sc
where r s i n k and Tsource are absolute temperatures. (Unfortunately there is

disagreement in terminology and this quantity is sometimes called the
'performance energy ratio'J 6 2 !) For practical heat pumps the COP is
usually in the region of half the ideal value. The BRE has found that heat
pumps in buildings operate with a COP between 2 and 3 averaged over the
year although it is quite possible that, in the future, heat-pump values of
greater than 3 may be obtained.!651 It should be noted that a disadvantage
of heat pumps, evident from the definition of the COP, is that perform-
ance falls as the temperature of the source falls - a consequence of this is
that supplementary heating is often used at the coldest times. A second
parameter of interest is the primary energy ratio (PER), which is defined
as the ratio of the heat delivered to the energy content of the fuel con-
sumed to provide the work input required.!661
Fig. 6.42 shows the vapour absorption cycle which is less commonly
used in heat pumps. The principal difference between it and the vapour
compression cycle is the substitution of a heat absorption unit for the
compressor. In the absorber part of the unit, the refrigerant vapour from
the evaporator is dissolved and condensed in a secondary fluid called the
absorbent (with the heat evolved being delivered to the building). The
refrigerant and absorbent are then separated by the application of heat in
the generator, with the evaporated refrigerant passing to the condenser,
heat being delivered to the building and the still-liquid absorbent returning
to the absorber. The rest of the cycle is the same as the vapour com-
pression cycle. Since no mechanical work is done in the absorption cycle
the coefficient of performance tends to be substantially lower but, on the
Fig. 6.42. Vapour • Refrigerant
absorption cycle.!63 J - Absorbent
, Thermal Circulation - Refrigerant/Absorbent
GENERATOR
Heat injected
into dwelling
CONDENSER
Burner
. Thermal energy
| input
I THROTTLE VALVE
j ABSORBER Heat injected

L. into dwelling Heat
extracted
EVAPORATOR
from outside
source
other hand, there is an opportunity for greater simplicity and reliability in

the design.
Most heat pumps available in the UK are based on US designs for signifi-
cantly different conditions. First, the heat requirement in UK houses is
often less because of reduced floor areas and more-compact house forms,
because internal temperatures tend to be lower (say 18—20 °C rather than
20-23 °C in the US) and because the British tend to wear more clothes
and use curtains over windows. Secondly, the UK climate is characterized
by moderate temperatures and relatively high humidity. In the US, both
heating and cooling are often needed, and partly because it is relatively
easy to add a heat-pump capability to an air-conditioning unit, heat pumps
have become quite popular there — it is estimated that 1.5 million units
have been installed J 63 1 Because the need in the UK is for a small heating-
only unit, though, it is not appropriate to simply import the technology
from across the Atlantic. For example, the COP of US heating—air-
conditioning units as heat pumps can be improved significantly by
doubling the compressor size and designing for heating onlyJ 66 !
Research in the UK is geared to best using primary energy, increasing
reliability and efficiency and lowering costs. British Gas has studied gas-
fuelled heat pumps and concluded that they may have significant advan-
tages over electrically powered versions. Table 6.6 compares the efficiency
of a variety of heat pumps and gas boilers.
The main points of note are that heat pumps make better use of pri-
mary energy than boilers and that, although the coefficient of perform-
ance of an electric heat pump is two-and-a-half times that of a natural-gas-
fired heat pump, the latter uses primary energy more efficiently. Electric
heat pumps use primary energy significantly better than other electric
heating systems but, compared to traditional domestic boilers, their slight
performance gain does not at present justify the increased cost.
Technical problems that remain to be overcome in developing a suitable
Heat pumps 231
Table 6.6. Efficiency of primary energy use of heat pumps and boilers
(after Ref [63])
Energy Energy
supplied from Energy
Primary to heat outside into
Type of system energy Losses pump source building
Electric heat pump 100 13a 27 54 81

(COP = 3)
Gas boiler 100 25* — — 75
(80% efficiency)
Substitute natural-gas 100 44 c 56
boiler
(85% efficiency)
Gas-fired absorption 100 6d 94 43 113
heat pump
(COP= 1.2)
Substitute natural-gas- 100 30 70 30 84
fired absorption heat
pump
(COP= 1.2)
a
Comprising generation and transmission losses.
b
Comprising 6% production loss and 19% flue loss.
c
Comprising 30% production loss and 14% flue loss.
d
Production loss.
heat pump for the UK include starting current and defrosting. In the UK
there is presently a limit on the use of single-phase motors in domestic
premises to prevent starting surges unduly interfering with supply to
neighbouring premises, and this limit will affect the size of heat pumps
which can be used practically. One estimate is that if the electric motor
output could be kept below 750 W there would be little need for electrical
reinforcement of the local network.I671 The use of three-phase devices
would overcome the problem and could increase efficiency but the extra
cost, particularly in the case of modifying the supply to an existing house,
would be a disadvantage. The UK Workshop on Heat Pumps concluded
that 'further investigations are required to extend the acceptability of
single-phase heat pumps' J 66 1
Defrosting principally affects air-source heat pumps of both the vapour
compression and vapour absorption types. Within a certain range of
ambient temperatures, water from the atmosphere, which condenses on
the outside coil, freezes on the evaporator surface and reduces the heat-
transfer efficiency. In some electric heat pumps the ice is removed by
operating in the reverse mode for a short time and it is envisaged that,
with certain gas heat pumps, defrosting could be achieved by stopping the
flow of central-heating water through the condenser, allowing hot refriger-
ant to melt the iceJ 63 ! Because of the relatively high humidities in the UK,
even at low temperatures, defrosting is a problem. Tests made by the Elec-
tricity Council have shown that, on average, 2.4% of the electricity
demand of a heat pump used in the UK would be expended on defrost-
ing. I631 Again, the Workshop on Heat Pumps, in suggesting several possible
solutions, recommended further research.
Cost was one of the principal reasons for the past lack of interest in
heat pumps. The capital cost of a heat pump is greater than that of a boiler
of comparable power, thus, the additional cost must be repaid in a limited
period by savings in running costs. The Workshop on Heat Pumps con-
cluded that, for devices in the 3—8 kW output range, if the additional
capital expenditure did not exceed about £200 (1976 prices) for a whole
house-heating system, heat pumps could be competitive and it saw this as
the basic question for heat-pump development in the domestic area. At
least one manufacturer has recently claimed that heat pumps are now
more economical than heating systems using boilers, but manufacturers
tend to be an optimistic breed. Long-term performance trials to verify
system reliability and seasonal COPs will be needed before firm claims can
be made. Any additional back-up heating provided must also be costed
carefully.
The Workshop also thought that an alternative approach of small,
inexpensive heat pumps for single-room heating merited further research
and the BRE has been investigating this possibility. It should be noted that
in some countries such as West Germany and France the government sub-
sidizes the initial cost of heat pumps.
The most straightforward domestic application of the heat pump is as a
heating device, either in a central-heating system or as a room heater. In
low-energy houses, and particularly those using ambient sources of energy,
where the complexity of both the space heating and domestic hot-water
heating systems is increased and where energy storage is often incorpor-
ated, heat pumps can play additional roles as 'energy-management' links
between inputs and outputs or between storage and supply.
In general, experts are optimistic about the prospects for the wide-
spread introduction of heat pumps for space heating during the period
1985-2000. It is thought that one of the major markets will be as a
replacement for the first generation of central-heating systems, which were
less efficient and which become oversized as insulation standards in houses
are increased. In the longer term, given the development of suitable
devices, it may be expected that every low-energy house will have a heat
pump for, say, either space heating or some waste-heat-recovery function.
The following examples give an idea of current developments in domestic
heat pumps.
The BRE Solar Energy House shown in Fig. 6.27 uses three heat pumps
as energy-management links in a system which owes part of its complexity
to the need to extract as much experimental data from a given physical
situation as possible. The first heat pump upgrades heat from the solar
collector into the main heat store when radiation is low. The second is a
small water—water heat pump used to upgrade heat from the main store
for use by the space heating system and the third upgrades heat from the
main store for use in the domestic hot-water system, whose required tern-
Heat pumps 233
Fig. 6.43. Schematic of a
heat-pump system for new
and renovated houses.! 6 8 1
perature is often higher than that of the store. Other BRE experimental
houses use heat pumps to supply all the space heating, to recover heat
from the outgoing ventilation air and to recover heat from waste domestic
hot water held in a catch tankJ 6 5 !
A novel application of heat pumps is found in the Salford Strawberry
Hill Houses shown in Figs. 6.28 and 6.34.I451 For these homes it was
decided, for reasons of cost and flexibility in the choice of an energy
source, to use heat pumps run on night-time electricity only. Exhaust air
from the house is used as the energy source. Because the heat pump runs
only at night and because it is desirable to store that energy to avoid
wasting it when there is no heat demand, cold-side heat storage was
employed. Thus, instead of extracting heat from the exhaust air directly,
heat is taken from a tank of water, forming ice during the night-time
operation of the heat pump. The ice formed is then melted during the day
by the warm extract air which passes over the tanks. Such a system which,
it will be appreciated from Fig. 6.28, is by no means simple, has nonethe-
less the important practical advantage of requiring no special provision for
defrosting. The expected coefficient of performance of the heat pump has
been taken to be 2.5. Heat pumps are also used in the houses for preheat-
ing domestic hot water from 10 to 35—40 °C with direct night-time elec-
tricity used to supply the additional energy required.
As a final example, J. Keable's use of a standard air-to-air heat pump
may be cited since he has demonstrated its applicability and reliability in
both a renovated house and a new one.l 68 ! The principles are the same in
both, with air being extracted from an air passage formed underneath a
roof covering of slate or tiles (see Fig. 6.43) before reaching the heat
Fig. 6.44. Coal-fired hot-
water boilers at Basildon
NewTowni 6 9 ) (Courtesy
of Ove Arup and Partners.)
pump. Heat from the heat pump is sent to the thermal store directly and,
since the machine senses the temperature of the store and the ambient
air rather than the indoor temperature, it is easy to avoid operation at
unfavourable times of low temperatures. Consequently, the coefficient of
performance is improved. Distribution of heat from the thermal store is by
hot water to low-temperature radiators in the case of the renovated house
and by air through ducts in the new one. The system may be considered
to be a combination of a low-cost solar collector and heat pump.
6.7 Group schemes

Combinations of traditional and ambient energy sources similar to those
discussed for individual houses in the preceding sections may be used for
group schemes.
In the Basildon Housing cited previously, coal-fired group heating was
selected. Three boiler houses located in a sunken site each serve 140
houses. Each boiler house has two hot-water boilers rated at 580 kW each
and fired by underfed stokers which mechanically extract bituminous
washed coal singles from a bunker and feed the fuel direct to the furnace
(see Fig. 6.44)J69 1 In this way, handling is eliminated and a clean atmos-
phere results in the boiler house. To avoid the high costs of buried mains,
the hot-water flow and return pipes are housed in overhead ducts running
along covered walkways and hidden beneath the eaves of the pitched roof
houses. In the homes, pressed-steel low-pressure radiators fitted with
thermostatic control valves are used and the occupants control their
(metered) heating supply through the valves and a time programmer.
A central gas-fired boiler is used in tandem with roof-mounted solar
collectors in the Lewisham Flats mentioned previously (see Fig. 6.45). The
solar collectors provide both preheating of the domestic hot water and
underfloor space heating in the lower flat of each block.
The active solar energy system consists of ten 20 m2 solar panels form-
Group schemes 235
KEY Fig. 6.45. Schematic of solar
0 solar collector prels collectors used with gas
central heating.I 70 !
(Courtesy of Max Fordham
® preheat cylinder
and Partners.)
© feed + expansion tank
©pump
(6) dom^tjc hot water
© floor coils
ing the complete south-facing pitch of each block of the building.I711

Solar heat is designed to be collected at the lowest possible temperature.
Thus, since the preheat cylinder is likely to be colder than the building, if
the solar collector is warmer than the preheat cylinder there is useful heat
to be gathered and the pump is switched on. This pump is set to switch on
at a 2 K differential and off at a shade above a 0 K differential.
The preheat cylinder consists of a standard indirect hot-water cylinder
of capacity 280 1 fitted with a heat-transfer coil able to absorb the output
of the solar panel. The preheat tank feeds three hot-water cylinders with
preheated water. The pipework from the preheat cylinder to each
domestic hot-water cylinder is effectively a common hot-water service cold
feed and incorporates the normal design details appropriate to that appli-
cation.
Each flat contains its own hot-water cylinder and electrical intake. The
bathroom and kitchen are not back to back which means that separate
plumbing stacks are required for each. However, this saves the need for
mechanical ventilation, which reduces both capital and maintenance costs.
The simplicity of this part of the installation contributes to the possibility
of providing a solar energy scheme within the standard housing costs.
A temperature detector embedded in the floor compares the floor tem-
perature with the preheat cylinder temperature. If the floor temperature is
lower than the preheat cylinder temperature then heat can more usefully
be employed to heat the floor. When the external thermostat detects an
outside temperature below 15 °C, which shows that heat for the building
can usefully be used, the water flow is diverted from the preheat cylinder
to the floor. The back-up gas-fired heating system is controlled by indi-
vidual thermostatic radiator valves which will close down when solar heat
is provided. The rate of fossil-fuel heat consumption is thus reduced by
any solar gain. The control of the radiators in each flat is arranged on a
one-pipe loop and a thermostatic valve ensures that there is no flow when
the building is up to temperature.
The floor panels consist of 15 mm welded-steel pipework to British
Standard 1387 laid in the floor finish. The floor finish consists of 100 mm
of screed laid on top of a mineral-wool sound-insulating quilt all laid above
precast concrete planks. The pipes are laid at 225 mm centres. The heat
output of the floor panels is 100 W/K temperature differential between
the water and the room air temperature.
6.8 Conclusion
In selecting a heating system, the resources available to the society, the
real needs of the occupants, initial costs, running costs and flexibility must
be considered, but, as is evident from the preceding discussion, designers
of low-energy houses have arrived at a variety of solutions - every readily
available form of heating from solid fuel to electricity has been used —
after considering some or all of these factors. In part, the explanation lies
in the varying relative importance given to individual factors - if, for
example, the predominant concern is efficient use of primary energy
resources, solid fuel is preferable to electricity consumed either directly or
in storage heaters. (However, if the electricity were to be used in heat
pumps or if it were provided by wind or wave power rather than con-
ventional generating plants, heating with electricity could be acceptable.)
Another reason lies in the effect of decreasing the energy demand,
which causes both initial and running costs to be lowered. The same per-
centage cost-difference between, say, electricity and solid fuel, is perceived
differently if in one case the annual heating costs are £900 and £600, and
in the other they are £225 and £150.
References 237
In addition, it is useful to distinguish between houses which are
designed primarily with a reduced energy demand in view and those which
attempt to combine a low energy requirement with the best use of passive
solar gain. In the case of the latter, the heating-system response (both
hardware and controls) is extremely important and quick-response systems
should be selected. The question is also related to the pattern of occu-
pancy and in houses which are used more or less continuously the choice
of heating devices which incorporate some thermal storage capacity, such
as Aga cookers or electric storage heaters, may be justified.
All of these factors influence the decision to select central heating or
individual space heating, or either an intermediate or combined solution.
In a compact, continuously occupied well-insulated house it may be poss-
ible to heat the first floor adequately from the ground floor simply by
incorporating an open plan and leaving the doors open (or using transfer
grilles in the doors). Or a parallel system of, say, gas central heating and a
solid-fuel 'focal-point' heater with its attractive amenity value may be
used to heat only those parts of the house which are actually occupied.
The choice rests with the occupants and the designer.
In any case, it is judicious to design for flexibility and to consider care-
fully such factors as allowing for low supply temperatures (and it may be
noted that both hydronic and air systems are capable of this), provide
space for thermal storage (in water, rock beds, eutectic salts, etc.), include
a chimney and allow for later conversion to a solar roof. Because indi-
viduals have varying criteria for comfort the heating system should also be
flexible enough to allow the occupants to easily adjust temperature set-
tings and, particularly, to lower them and thus save energy.
In part due to the lack of experience with low-energy homes, the
selection of a heating system remains an art — given the uncertainty sur-
rounding future energy supplies and costs it is likely to remain so for some
time.
References
[ 1 ] Field, A.A. (1977). 'Anticipations'. Building Services Engineer,
45(l),A22-6.
[2] Burberry, P. & Aldersley-Williams, A. (1977). 'Domestic heating'.
Architects Journal, 166 (33), 301-25; 166 (37), 453-68; 166
(39), 605-18.
[3] Burberry, P. (1975). Environment and Services. London: Batsford.
[4] Kell, J.R. & Martin, P.L. (1971). Heating and Air Conditioning of
Buildings. London: Architectural Press.
[ 5 ] Kut, D. (1968). Heating and Hot Water Services in Buildings.
Oxford: Pergamon.
[6] Barton, J.J. (1970). Small-Bore Heating and Hot Water Supply for
Small Dwellings. London: Newnes-Butterworth.
[ 7 ] Carrier, W.H. (1959). Modern Air Conditioning, Heating, and
Ventilating. New York: Pitman.
[8] Tulloch, J. (1978). 'Two-year solar energy experiment produces
poor results'. Building Design, 22, 7.
[9] Anon. (1980). 'Gas: more changes than problems'. Domestic Heat-
ing +, March, p. 5.
[10] Romig, F. & Leach, G. (1977). 'Energy conservation in UK dwell-
ings: domestic sector survey and insulation'. Working Paper.
London: IIED.
[11] Sutherland, C. (1982). Private communication. (Conversion to MJ
by R. Thomas: Copyright C. Sutherland.)
[12] Anon. (1956). Domestic Heating - Estimation of Seasonal Heat
Requirements and Fuel Consumption in Houses. BRS Digest 94.
[13] Green, M.B. & Pickup, G.A. (1978). 'Useful figures for the
efficiency of gas central heating'. Building Services and Environ-
mental Engineer, 1 (4), 8-9.
[ 14] Anon. (1979). 'Gas comes top in heating'. Building Services and
Environmental Engineer, 2 (1), 5.
[15] Anon. (1979). 'Gas: the heat is on'. Domestic Heating +,
September, p. 6.
[16] Anon. (1979). 'Oil over a barrel'. Domestic Heating +, September,
p. 16.
[17] Anon. (1980). 'A comeback for electricity?' Building Services and
Environmental Engineer, 2 (7), 17-19.
[18] National Coal Board, London (1980). Private communication.
[19] Solid Fuel Advisory Service, London (1980). Private communi-
cation.
[20] Baumeister, T. (ed.) (1967). Standard Handbook for Mechanical
Engineers. New York: McGraw Hill.
[21] Anon. (1977). Jjtul - Magic Warmth from Wood. Newcastle
Emlyn, Dyfed: Simon Thorpe.
[22] Anon, (undated). The Solid Advantages of Solid Fuel. London:
Solid Fuel Advisory Service.
[23] Vale, R. (1977). 'Low cost thermal upgrading of an existing house'.
Environment and Planning B, 4, 173-84.
[24] Vale, B. & Vale, R. (1980). The Self-Sufficient House. London:
Macmillan.
[25] Anon. (1979). 'Saving fuel in comfort'. Domestic Heating +, May,
p. 41.
[26] Anon. (1980). 'The development of solid-fuel heating'. Building
Services and Environmental Engineer, 2 (7), 18-20.
[27] Littler, J.G.F. & Thomas, R.B. (1979). Ambient Energy Design:
House Heating for the 1980s. Cambridge: Ambient Energy Design.
[28] Withers, M. (1977). 'Heating'. The Architect, April, pp. 51-4.
[29] Anon. (1979). 'Householder's choice criticised'. Building Services
and Environmental Engineer, 1 (6), 3.
[30] Anon, (undated). 'Combination gas boiler and cylinder for central
heating and domestic hot water'. Technical Data Sheet. Telford:
Harvey Habridge.
[31] Anon. (1975). Flues for Domestic Gas Appliances. Watson House,
London: British Gas.
[32] Anon. (1977). Microbore. Bilston: The Wednesbury Tube
Company.
[33] Anon, (undated). High Quality Output Radiators. Epsom: Hudevad
Britain.
References 239
[34] Anon, (undated) Boulter Cam-ray Natural Convector Radiator.
Norwich: Boulter Boilers.
[35] Pimbert, S.L. (1977). 'Reflecting materials behind radiators and
shelves save energy'. Building Services Engineer, 45 (5), A22—3.
[36] Anon, (undated). The Myson 'L' Series Circulating Pump. Ongar,
Essex: My son.
[37] Rockhill, H.P. (1967). 'Closed heating systems'. IHVE Journal, 35
(99), 1-7.
[38] Fisk, D. (1980). 'Controlling the heat output'. Building Services
and Environmental Engineer, 2 (7), 21.
[39] Bakke, P. (Chairman) (1975). Energy Conservation: A Study of
Energy Consumption in Buildings and Possible Means of Saving
Energy in Housing. BRE CP 56/75. Garston: BRE.
[40] Vint, M. (1979). 'Controls: the missing link'. Domestic Heating +,
December, p. 20.
[41] Anon, (undated). Taco-Constanta Thermostatic Radiator Valve.
Basingstoke: Tacotherm.
[42] Fisk, D J. (1979). Microelectronics in Building Services. BRE CP
12/79. Garston, Watford: BRE.
[43] Ullathorne, D.P. (1980). 'Microprocessors and their possible
applications to central heating systems'. Watson House Bulletin,
44(311), 39-45.
[44] Seymour-Walker, KJ. (undated). BRE Proposals for Experimental
Low-Energy Houses. Garston: BRE.
[45] Randell, J.E. & Boyle, J.M.A. (1979). 'Low energy housing -
Salford's answer'. Building Services and Environmental Engineer,
1(10), 18-21.
[46] Field, A. (1978). 'Controlled ventilation for dwellings'. Building
Services and Environmental Engineering, 1 (2), 6—7.
[47] Zanelli, L. (ed.) (1973). Central Heating, Plumbing, Electricity.
London: Marshall Cavendish.
[48] Dunstan, G.F. & Green, M.B. (1979). 'Warm air heating in low
energy housing'. 2nd International Symposium on Energy Con-
servation in the Built Environment, Copenhagen.
[49] Anon. (1976). Huskiheat Gas Fired Brick Central Heating.
Bishops Stortford: Heatpak.
[50] Anon. (1977). Controlled Ventilation for Dwellings Combined
with Modairflow Gas Warm Air Heating Systems. Northampton:
Johnson and Starley.
[51] de Labbeye, P. & Lebrun, H. (1976). 'Tests effected on a heat
exchanger to be utilized in a double flow'. Electricite de France
Research Centre Report HC 312 W 774. Ecuelles: EDF.
[52] Applegate, G. (1970). 'Heating a factory for nothing'. The Heating
and Ventilating Engineer, 44 (516), 21 —4.
[53] Esbensen, T.V. (undated). The Zero Energy House. Technical
University of Denmark.
[54] Anon. (1978). Heat Recovery for Industry. Technical Information,
IND 18. The Electricity Council.
[55] Horton, A., Grove, S. & Lewis, A. (1979). Milton Keynes Solar
House Performance and Cost Analysis of Solar Heating System.
Built Environment Research Group, Polytechnic of Central
London.
[56] Horton, A. (1978). Milton Keynes Solar House Supplement 1:
Alterations to the Fan Convector. Polytechnic of Central London,
BERG.
[57] Anon, (undated). A Guide to Electric Space and Water Heating for
Small Commercial Premises. EC 3487. The Electricity Council.
[58] Anon. (1979). The Medallion Award Specification. The Electricity
Council.
[59] Anon. (1977). Medallion Award Homes. EC 3697. The Electricity
Council.
[60] Anon. (1980). 'Attention to energy saving'. Building Services and
[61] Anon. (1979). 'Striking a medal for low-cost heating'. Domestic
Heating +, March, p. 28.
[62] Summer, J.A. (1976). Domestic Heat Pumps. Dorchester: Prism.
[63] Jessen, P.F. & Johnston, I.W. (1980). 'The domestic gas-fired heat
pump'. Watson House Bulletin, 44 (311), 47-56.
[64] Mitchell, J.W., Freeman, T.L. & Beckman, W.A. (1978). 'Heat
pumps'. Solar Age, 3 (7), 24-8.
[65] Freund, P., Leach, S.J. & Seymour-Walker, K. (1976). Heat Pumps
for Use in Buildings. BRE CP 19/76. Garston: BRE.
[66] Anon. (1976). UK Workshop on Heat Pumps. Didcot: Energy
Technology Support Unit.
[67] Anon. (1980). 'Domestic heat pumps'. Building Services and
Environmental Engineer, 2 (7), 22-3.
[68] Keable, J. (1977). 'The application and economics of heat pumps'.
In: Ambient Energy and ^Building Design Conference Proceedings,
University of Nottingham, pp. 69—82.
[69] Anon. (1979). 'Coal heating for estate'. Building Services and
[70] Anon. (1979). 'Brownhill Road Flats, Lewisham'. In: Buildings
the Key to Energy Conservation, Kasabov, G. (ed.). London:
RIBA Energy Group.
[71 ] N. Ryding, Max Fordham and Partners (1982). Private communi-
cation.
7
Thermal storage
7.1 Introduction
Storage has been discussed in a number of the preceding chapters and,
notably, in conjunction with passive solar heating. Here we shall cover the
topic more systematically but in some cases in less detail. Storage is often
the key to the successful and economical use of solar or wind energy for
space and hot-water heating. While it is possible, as we have seen in Chap-
ter 3, to drastically reduce the space heating demand of buildings, the
residual demand must either be met from a conventional source of fuel or
from energy stored during a period of greater availability. Most buildings
using solar or wind energy have been forced to limit storage capacity
because of cost and employ a fossil-fuel back-up. Research projects, such
as that of the Cambridge Autarkic House Project, are responsible for most
of the (rare) designs which rely only on ambient sources of energy. It is
nevertheless encouraging to know that even in the UK it is possible to
provide in an average year 100% of the space and domestic hot-water
heating requirements with solar energy alone J 1 !
The size of the store required for any application will depend on the
energy demand, the source of energy and the selection of the storage
medium. Because the cost per unit of storage falls as the size increases,
more and more attention is being given to projects which involve large
buildings or groups of dwellings. This in turn permits new equipment to
be used more economically and novel means of storage to be examined.
An example of the first point is that on a very small scale the additional
capital cost of a heat pump in a system may not be economical but for
larger schemes this cost becomes relatively small and heat-pump upgrading
of stored heat becomes much more attractive. Among the novel large-scale
stores being considered are uninsulated rock caverns (similar to those now
used for storing oil, for example, in Sweden), confined groundwater
aquifers^2' 3» 4J and clay soils.t5^ Work is also underway in a number of
countries on extracting, usually with a heat pump, the heat that is stored
by natural processes in earth, rock and water.
Storage, of course, helps to compensate for daily, monthly and seasonal
variations in the supply of energy. In a sense, the problem facing designers
is not one of the availability of energy since, even in the UK, for example,
the incident radiation is 80 times the present energy demand.^6! The
questions are how to economically capture and store ambient sources of
241
242 Thermal storage
Fig. 7.1. Average monthly 431.6
totals of output power from
a solar collector and wind 359.7
generator at Aberporth.l71
Curve (a): solar energy col-
lector on south-facing roof 287.8
at 30° elevation, efficiency =
35%. Curve (b): wind gener- 214.8
ator at 10 m height with
Vrated = 20 knots, V c u t . i n
= 10 knots, efficiency = 143.9
40%. Curve c: combination
of (a) and (b), with equal 71.9
areas. (Reproduced with
permission of the Controller
of Her Majesty's Stationery 0
J F M A M J J A S O N D
Office.)
energy. Fig. 7.1 shows the pattern of availability of solar and wind energy
at one site, Aberporth (52 08 N, 04 34 W, 133 m above sea level), in the
UK.
Solar energy is least available during the winter heating season, when it
is needed most. On the other hand, wind energy has a much more suitable
pattern and, additionally, is quantitatively more important (each site must,
of course, be considered individually). Consequently, it is likely that a
smaller store could be used with the wind system.
The scope for storage applications is vast and by no means limited to
on-site energy supply. Ryle, arguing against nuclear power stations, has
said that wind energy should be supplied to the national grid.l8! Because
much of the fluctuation in the demand placed on the grid is related to the
provision of low-grade space heating for domestic and other buildings and
because the central storage systems envisaged would involve multiple con-
version of energy and thus low efficiency, he suggests that each home be
provided with a small store. Energy would be stored as heat, perhaps by
using electric resistance heaters, and storage would be for a period of
about 150 h. Of course, a larger store could also be used to improve the
seasonal load factor of conventional power stations. Photovoltaic systems
also need an inexpensive effective means of storage before they will be
used widely. In the US it is predicted that, within this decade, photo-
voltaics will be competing with oil for power stations in the sunnier areas
of the country.I91 Storage will be required for daily and longer periods
and the full scope of possibilities from centralized stores to individual ones
in homes will need to be studied.
We would suggest that in order to make the best use of ambient energy,
space for storage be provided in every building. This would be somewhat
similar to the provision of adequate sanitary facilities (often neglected in
the past) or the inclusion of chimneys in new houses, as suggested by the
UK Solid Fuel Advisory Service. The space could change its use with time
— it might start as part of a utility room or garage, become a solid-fuel
store and then finish as a thermal store for energy from national wind-
turbine power stations. The concept has already taken hold in Sweden
Sensible-heat storage in water 243
Table 7.1. Thermal properties of storage media*
Density Specific heat

Media (kg/m3) (kJ/(kg K))
Water (liquid) 1000 4.19
Rock (typical) 2500-3500 0.88
Iron 7860 0.50
*Ref. [10].
where it is not unusual to incorporate 4—5 m3 of heavily insulated water

storage in the domestic heating system.
The rest of this chapter deals principally with the various thermal
storage media available for use in the 30-90 °C range and their appli-
cations to buildings.
7.2 Sensible-heat storage in water

For many applications, provision of, say, 20-80% of the space heating and
domestic hot-water energy demand will be the object of an ambient energy
collector and storage system.
Water has been the traditional choice for low-grade thermal storage
because of its high specific heat, minimal cost, safety in operation, con-
venient use as a heat-transfer agent and storage medium and compatibility
with existing heating and water systems. In most cases, thermal energy is
stored in water as sensible heat, which simply means that the temperature
of the material is raised and the energy is stored internally. The amount of
energy that can be stored as sensible heat is given by the formula
=
Energy stored ( ~~3 ) P ' cp ' A71,
where
p is the density of the material (kg/m3 );
cp is the specific heat (at constant pressure) (kJ/(kg K));
Aris the change in temperature (K).
As mentioned previously, thermal energy can also be stored as latent
heat by making use of the heat absorbed by certain substances when
undergoing a phase change. Tables 7.1—7.2 provide some data of relevance
to the selection of a thermal storage medium.
Although water has significant advantages over other media for sensible-
heat storage, it is not without its drawbacks. In particular, the cost of con-
tainment can be high (the cost of insulation may also be significant) and
there are risks of corrosion due principally to the presence of dissimilar
metals in many systems and the possibility of air entering the system. In
general, sensible-heat stores tend to be large and heavy and this can pose
architectural problems of space allocation and structural loading. When
using water for heat storage, attention must also be paid to the possibility
of damage due to freezing in cold climates, the risks of overheating in the
244 Thermal storage
Table 7.2. Energy storage in varying media for a 10 K temperature risea
Energy storage
Medium (MJ/m3)
Sensible-heat storage
Water 41.9
Iron (scrap, 20% voids) 31.4
Concrete (dense) 22.0
Rock (30% voids) 18.5
Brick (fairly dense) 16.7
Phase-change heat storage
Glaubers salt (Na2SO4 • 10 H2O) transition temperature 392
+ 10 K rise in bin
Ice (heat released) 307
"Refs. [6, 10, 11 and 12].
system if either the water reaches boiling point or mechanical failure inter-
rupts flow and the danger, for example, to insulation, if leaks occur.
Determining the size of the store can be as complex a problem as one
desires. In the design for the Cambridge Autarkic House a computer
simulation of energy supply and demand using hourly weather data from
Kew, London, suggested that 40 m 3 of water storage (surrounded by an
additional 70 m3 of polyurethane insulation) would be needed to supply
an average annual combined space heating and domestic hot-water load
of 7.2 GJ.l 1 ' 131 Provision of such a store is not, however, economical and
in most situations space and cost constraints will limit the storage size and
simplify the problem of arriving at a suitable volume; conventional fuels
then meet the additional energy demand.
For domestic hot-water heating systems in the UK the optimum value
of storage has been found to lie in the range 35-120 1 per m2 of solar
collector areal 14 ! (see also Chapter 5). Most such systems have 4—6 m2 of
collector and so if we use the two upper figures we have a volume of water
of 0.72 m 3 . In the US, for domestic hot-water and space heating systems,
a similar guideline is 40-80 1 per m2 of solar collector J 15 1 For space heat-
ing, both collector areas and storage volumes will, of course, be much
larger with similar ratios of 50-120 1 of storage per m2 of collector being
common (see Table 5.5).
Exactly what physical form a store should take has been a somewhat
neglected question. More work has gone into collector design than, for
example, into suitable stores. Numerous materials have been considered
by several authors.I13' 16J Mild-steel tanks are a fairly standard choice
because they are readily available at competitive prices and can be
designed to withstand pressure should the system be pressurized. A tend-
ency to corrode, particularly at high temperatures, is their principal draw-
back. At such temperatures galvanizing does not provide suitable pro-
tection. For example, in hard waters above 60 °C the zinc becomes
cathodic to the steel, thus negating the desired effect.^17! Even with
system precautions such as nitrogen rather than air covers to reduce
Table 7.3. Approximate costs of linings and coatings for sealed tanks^13^
Maximum
Cost per m2 a temperature Lifetime
Material (£) (°C) (y)
Aluminium 21 Satisfactory 10-15
at 95
Butyl rubber6 7-40 70-120 1-20
Soft rubber 6 27-35 80-100 10-20
Hard rubber 6 27 105 10-20
Epoxy with glass6 7-8 90 10
Bituminous paintc 10 95 1
Organic emulsion 1.50 270 20
a
1979 approximate costs.
6
Includes preliminary surface treatment.
c
Includes preliminary shotblasting of internal surfaces and allows for two
coats of paint.
atmospheric corrosion, protection is needed if the store is to have a reason-

able life. A common guideline for solar systems is that readily replaceable
components should last ten years and those which are not, 20 years.
Linings, coatings and water additives (which may be combined with the
first two items) are often used for protection but they are not without
their problems. For linings and coatings, cost and temperature resistance
are the principal difficulties.
Linings fall broadly into the categories of metals, glasses, rubbers and
plastics. One possibility is an aluminium lining, itself protected by a vinyl
coating, which must be inspected for peeling every three to four years. Its
cost and those of several other options are shown in Table 7.3.
Vitreous-glass linings are unsuitable for prolonged periods at tempera-
tures of 90—95 °C. Butyl rubbers can vary in their maximum temperature
from 70 to 120 °C with costs mounting at the higher limit. Lifetimes of
the cheaper linings are likely to be in the range of only one to five years.
General-purpose soft rubbers are suitable up to 80 °C and special formu-
lations will withstand 100 °C. Hard-rubber linings, which have a maximum
temperature of about 105 °C, have the disadvantage that cracking may
occur if the tank is made of too-thin metal sheet and is jarred or settles
unevenly. Probable lifetimes of higher-quality products are ten to 20 years.
Coatings include epoxy coatings, paints and organic emulsions. Manu-
facturers are reluctant to guarantee epoxy compounds at up to 90 °C with
cycling down to 40 °C. It is conceivable that an epoxy with glass system
would be suitable but the probable lifetime is no more than ten years.
Acceptable polyurethane paints are not presently marketed. There is a
bituminous paint which has the required temperature resistance but it
requires annual maintenance. A promising recent development is an
organic emulsion which, when painted on, combines chemically with the
steel surface.! 181 Its lifetime is claimed to be the same as that of the tank
and the temperature limit is a comfortable 270 °C.
246 Thermal storage
3
Table 7.4. Approximate costs of providing 40 m of
storage in sealed containers^13]
Approximate cost0
Material (£)
Mild steel 2200
Stainless steel 8000-26 200
Copper 15 000
GRP 15 000
Polypropylene 6800
a
1979 approximate prices.
Anti-corrosion additives are common in central-heating systems and

can be suitable for use in solar systems. (Wozniakl14! discusses additives
in detail.) Manufacturers are reluctant to reveal the composition of their
products but it is likely that many are based on mixtures of sodium nitrate
and either sodium benzoate or sodium tetraborate. The required strength
varies with the system but 1% is a common figure. At such low concen-
trations, heat-transfer characteristics are unlikely to be affected, nor will
the viscosity vary greatly and so the energy consumption for pumping
should not be increased. Component lifetimes in protected systems are
estimated at longer than 20 years. The cost of protecting 40 m3 of tank,
assuming an additive concentration of 1%, is about £720 (1979 prices). To
add a note of caution, however, the Greater London Council has studied
corrosion inhibitors for central-heating systems and has found numerous
problems ranging from cost to possible toxicity problems to excessive care
required in controlling the requisite concentration.! 19J
Numerous materials other than mild steel are available for tanks but
these tend to be much more costly, as is apparent from Table 7.4.
Stainless steel and copper have the advantage of being highly resistant
to corrosion but for large stores costs are excessively high. For smaller
applications, it is quite common to use copper cylinders designed originally
for domestic central-heating systems as stores. GRP tanks have excellent
corrosion resistance but most are not designed to cope simultaneously
with pressure and high temperatures. Instead of the normal polyester
resins which are not suitable at 90 °C because of degradation of the
molecular bond between the resin and glass, a variant such as epoxy must
be used. Cylindrical tanks with dished ends are favoured to reduce the
stresses induced by pressurization. Periodic inspection is required since, in
certain circumstances, water may enter the material itself and eventually
cause structural weakening. Cylindrical polypropylene tanks can be
pressurized and can withstand 90 °C temperatures; lifetimes are likely to
be of the order of 20 years.
Glass reinforced concrete (GRC) tanks are presently excluded from use
as major water-retaining structures. 12°1 Precast concrete can pose prob-
lems of porosity and cost, as well as requiring careful consideration of the
top and bottom seals for the precast sections. Exotic storage possibilities
abound but tend to be even more expensive. One example is the com-
mercially available evacuated tanks of stainless steel inside and carbon
steel outside, used for cryogenic applications. For the Autarkic House, the
authors found a more promising possibility in the use of two standard
carbon-steel tanks, one inside the other with a vacuum (0.1 Pa) maintained
between them. Since no insulation is needed and because the store is com-
pact, the size of any surrounding structure is also reduced. This solution
may be economically competitive with more-conventional ones. The great
uncertainty, however, is whether the vacuum can be maintained without
continuous pumping.
Two other storage possibilities merit mention. The first are the com-
bined solar collector and storage units which, in various forms, are in use
in many countries. The Japanese, for example, have developed 'water
pillows' made of vinyl (transparent on top and black for the bottom of the
pillow), which are used without a covering during the summer and with
one in winter.I211 Doubts have been expressed about the suitability of
this kind of device in the UK because of their relatively poor performance
and the structural problems of installing heavy units on roofs.l22 l How-
ever, their simplicity is in their favour and in situations where internal
space for storage cannot easily be found they may be both practical and
aesthetically attractive.
Development along different lines has led to a heat-storage module
made of polypropylene.I231 The module was designed to be completely
safe in a domestic environment and so is unpressurized and serves only for
storage. Heat exchangers link the store to the heating system. The basic
module has a storage capacity of 1000 1 and the insulation which sur-
rounds the inner vessel limits the heat loss to about 100 W at the maxi-
mum operating temperature of 95 °C.
We can now examine how thermal storage in water may be incorpor-
ated in buildings. The design for the Autarkic House stored 131 is an
extreme case but one which illustrates many of the problems involved. The
initial conception included a thermal store sited underneath the house for
protection from the elements and so that advantage could be taken of the
heat loss from the top of the store. In the early stages in the project it was
not realized, however, how large the store would have to be. The final
volume of 110 m 3 (40 m 3 of water 4- 70 m 3 insulation) could only be
incorporated underneath the house to avoid the aesthetic problem of a
massive aboveground object of little architectural interest. But, as we have
seen (Chapter 2), in many parts of the UK the water table is quite high.
(The site for the Autarkic House was in fact known in the past as Duck's
Dole, which gives an indication of the water table.) On such sites, to
resist hydrostatic pressures and to avoid moisture infiltration from the
surrounding earth, it is necessary to develop quite an elaborate structure.
The weight of the water in the tanks cannot be counted on because pro-
vision must be made for empty tanks to allow, for example, for inspection.
The design solution chosen was a basement made of reinforced concrete.
After excavation to a depth of about 4 m a preparatory layer of concrete
coated with a damp-proof course of bituthene was to be poured. Then
248 Thermal storage
Fig. 7.2. Insulation of the
Danish Zero Energy House
thermal store.
would follow a 100 mm concrete screed, a reinforced concrete base

350 mm thick, polyurethane insulation 700 mm thick, a steel tank pro-
tected with an organic emulsion to hold the water, 700 mm more of
insulation on top of the tank and, finally, polythene again. The polythene
was to function as a large impermeable bag to prevent any possibility of
water movement from the ground outside to the inside. The insulation
was to be foamed in place to provide a better thermal barrier than slabs.
On a similar scale, in the Danish Zero Energy House,!24! designed to be
self-sufficient in space heating and hot-water supply, a 30 m3 cylindrical
steel tank surrounded with 60 cm of mineral wool (which was thought to
be water repellent) was buried in the ground just outside the house. The
water table is well below the tank bottom and so no precautions were
taken against ground moisture but, to prevent rainwater from penetrating
the insulation, an earth-covered roof was built over it; a mechanically
ventilated air space separates the insulation on top of the tank from the
earth cover. Fig. 7.2 shows the tank being insulated with the mineral wool.
Not all designers will have the luxury of being able to consider the
incorporation of thermal storage at an early stage. Similarly, for the exist-
ing housing stock, provision of storage will often be a challenge. In the
case of the Milton Keynes Solar House shown schematically in Fig. 5.16, it
was decided to install a solar heating system without altering the basic
form or plans of one of the standard house types J 25 1 A combination of
37 m2 of solar collector (installed on the roof) and 4.5 m3 of water
storage, squeezed into the bedroom cupboards, was provided. The storage
consisted of two steel tanks of rectangular cross-section (so that the avail-
able space could be best used) arranged vertically. Inhibitors (to BSS
3150) were used to reduce the corrosion hazard and ethylene glycol was
added to alleviate the risk of frost damage.
In the more recent (1980) Solar Court Houses at Milton Keynes a com-
bination of 40 m2 of solar collector and 2.0 m3 of water storage in stan-
dard copper DHW cylinders was used.I26' 271 The cylinders, which are
insulated with 50 mm of proprietary board or quilt material, warm air for
Sensible-heat storage in rock 249
space heating. The indirect store is unpressurized but the collector circuit
is pressurized to a nominal 152 kN/m 2 . The collector circuit water is, by
volume, 25% proprietary antifreeze with a corrosion inhibitor. An
auxiliary gas-fired boiler is used as back-up.
For existing houses, finding space for a large store may be difficult, but
a simple small one (of the order of 500 1) for a domestic hot-water system
can often be located in a loft space. WozniakJ 14! in discussing the use of
commercial hot-water cylinders of copper in solar heating systems, has
pointed out that a common fault is inadequate insulation of the storage
vessel, which can result in excessive heat losses. As a guideline he suggests
that, for a 120 1 cylinder, 50-100 mm of glass fibre insulation or its
equivalent is economical.
Although metal is by far the most common material for storage con-
tainers in the range of 100 1 to, say, 50 000 1, other solutions have been
tried. To cite but one, in Australia a 20 m3 thermal store of 50 mm thick
precast reinforced concrete was incorporated in a solar installation.^28!
The system preheats water to warm soft-drink cans. Concrete was selected
in preference to mild steel because of lower cost. Space was no problem at
the factory and so the store could be located above ground. The unpress-
urized cylindrical tank was insulated with sprayed polyurethane 125 mm
thick and protected from the weather with an external sealer which was
sprayed on. Internally, the tank was coated with a synthetic rubberized
sealing compound.
Larger applications often rely on stores similar to those above but
increased in scale. In Madrid, a new large office building has a 500 m3
water-storage tank in the basement which serves as a heat accumulator.! 291
It is designed to bridge a period of 15 days without sunshine — Madrid's
longest sunless period on record is 11 days. In Canada, a 120 000 m 2
office building has a 6000 m3 thermal storage tank which provides water
for both heating and cooling.l30' Increasingly in that country, thermal
storage is being seen as a way of levelling the electricity demand and avoid-
ing high day-time tariffs. At Studsvik, Sweden, imaginative work is under-
way on the use of earth pit storage for annual heat storage. Fig. 7.3 shows
a demonstration plant with approximately 640 m 3 of water storage. If the
concept proves practical and economical, it is intended to design larger
stations which would serve up to 1000 houses.
7.3 Sensible-heat storage in ruck
Sensible-heat storage in rocks has been discussed in some detail previously

and so here we shall simply summarize the advantages and disadvantages
of this solution and discuss some applications.
Because the specific heat of rock is much lower than that of water and
because about one-third of the total volume of such stores consists of
voids, their energy storage capacity is lower than that of an equivalent
amount of water (Table 7.2). Consequently, a larger store will be needed
and where space is limited this will be a disadvantage. Also, in contrast to
water systems, heat cannot simultaneously be added to and removed from
250 Thermal storage
Fig. 7.3. Large-scale storage
of solar energy at Studsvik,
Sweden.[31J
/I
JJ Rotating lid Solar collector
j0
Ips'' ^ Drainage
Insulation
Water pipe
Solar collector Solar radiation
Reflector
the storage unit. The following advantages have, however, been notedJ 12 ^
(1) There are no problems of liquid leakage, pipe corrosion or freezing
(although there may be problems of air leakage and biological contami-
nation).
(2) Thermal diffusion through the store can be much less than that
associated with liquid stores. This is because normal practice in rock beds
is to choose stones of an appropriate shape and of uniform size to ensure
that minimal (point) contact is achieved and thus minimize conductive
heat transfer in the store. A marked division between the hot and cold
parts of the store results, with the thermal division moving one way or the
other, layer by layer, as heat is added or withdrawn.
(3) A heavy solid store can easily be located in the normal excavated
area under a building. Additionally, in areas with a high water table, it will
not usually be necessary to design (except at the time of construction) for
an empty store as must be done with a water system.
The container for the rocks is commonly concrete but sound wood-
frame boxes are also satisfactory and bricks would also seem to be suitable.
In the US one of the best-known rock stores is the Denver Solar House
described by Lof et al l 32 l There, 10 640 kg of 3 cm diameter rock was
used in a store composed of two cylindrical tubes 0.91 m in diameter and
5.5 m high. In Europe few houses have incorporated rock stores. One of
the rare examples is a house in Iioux, France, which has a 30 m 3 rock
store and a balanced air flow hypocaust (a hollow space under a floor used
for heat accumulation and transfer) circulation system (see Fig. 7.4)). A
UK example is the rock store in the Peterborough Houses (see Chapter 12)
which are being monitored by the EEC and so will provide much-needed
data on the performance of such stores in European conditions.
Sensible-heat storage in rock 251
Fig. 7.4. Rock store and
circulation system of a house
at Lioux, France.! 33 !
(Courtesy of Dominic
Michaelis Associates.)
Typical transverse section through ground floor and foundation

system, second version showing hypocaust air circulation system.
I Thermal store: rock. 2 Rough screed/insulation/damp-proof
membrane/site concrete/hardcore. 3 Extruded clay bricks laid side-
ways to allow passage of air. 4 Warm air delivery and collection
chamber to and from store. 5 Equalisation duct. 6 Adjustable floor
grill. 7 Space heating delivery chamber. 8 Reinforced concrete
foundation. 9 External wall of cellular brickwork. 10 Insulation.
II Insulated floor.
At least two other projects are also underway in the UK to provide

additional data. Again, as part of an EEC testing programme, the UK is
studying a half-size solar heating system consisting of 19 m2 of collector
and a 4 m3 rock-bed store.l 34l And at the Open University there is a
study underway of the potential for interseasonal storage of heat, which
involves the development of a prototype combined solar collector and
pebble-bed store. Fig. 7.5 shows a section through the intended store
design.
One application for the store, which has been studied, is to provide the
total annual heating and hot-water requirements for 100 houses.l36l The
store would be 280 m long, 10 m wide and 4 m deep and would have a
maximum temperature of 110 °C; pebble size is about 4 cm. The insu-
lation is likely to be 0.6 m of glass fibre.
252 Thermal storage
Fig. 7.5. Interseasonal solar
energy storage in a pebble- Garages
bed store.t 3 ^] (With the

permission of Dr B. Jones Solar air collector
and Mr T. Oreszczyn.)
7.4 Phase-change energy storage

As was mentioned in Chapter 4 some materials undergo a phase change,
for example from solid to liquid, accompanied by the absorption of a
large amount of heat (the heat of fusion) at a temperature which makes
them potential storage media. For a substance to be suitable, the phase
change must be reversible, that is, the heat must be capable of being
released over a large number of cycles without material degradation. To
render this abstract description more realistic, many readers will be
familiar with phase-change stores in the form of small refrigeration packs
for picnics. The pack is taken solid and cold from a refrigerator and put in
an insulated container with, if one is fortunate, avocado pears, lobster and
champagne. Heat from the food and drink is absorbed by the packs which
turn soft. After the picnic, the pack is placed in the refrigerator to which
it releases its heat, turning solid in the process. If such a pack could only
be used once it would be a rather expensive refrigerant (and heat store)
but its attractiveness to purchasers is that it can be used again and again.
Most important of all is that such materials can remove (and store) more
energy than is possible through sensible-heat storage (see Table 7.2).
Another significant advantage is that the process occurs at a specific tem-
perature (the melting point).
Hydrated salts are the phase-change materials that have been investi-
gated in greatest depth with the intention of using them for space heating
applications. Telkesl37' 38» 391 in over 30 years of work has described the
merits and problems of a number of salts, including Na2SO4 • 10 H2O
(Glauber's salt), Na2CO3 • 10 H2O and Na2HPO4 • 12 H2O. The character-
Phase-change energy storage 253
istic reaction for such materials is between the hydrated salt and the
anhydrous salt as shown in
Na2SO4 • 10 H2O + Energy t- Na2SO4 + 10 H2O.
Energy storage occurs by the reaction proceeding from left to right when
energy is added. The total energy stored depends on the complete process
since sensible heat is involved in heating the hydrated salt to the transition
temperature (32 °C in this case), heat of fusion to cause the phase change
(this is often referred to as latent-heat storage or, occasionally, as energy
storage using latent heat of phase change) and sensible heat, again to raise
the anhydrous salt and solution to the final temperature. Energy extrac-
tion from storage is the reverse procedure.
Since such salts are mainly water they tend to be inexpensive; addition-
ally, they generally pose no fire hazard. Those proposed for heat storage
tend to be non-toxic. Many occupy a smaller volume in the solid state
than the solution from which they form, thus there is not necessarily a
danger of fracturing a rigid container upon solidification!401 (as there is
with ice, for example). They are not, however, without their disadvantages
(which is also true for other storage media) and, unfortunately, these dis-
advantages have severely limited practical applications. The main obstacles
have tended to be physical. Glauber's salt, for example, does not actually
melt at 32 °C but has an incongruent transformation which is more pre-
cisely rendered by
Na2SO4 • 10 H2O (solid) (at 32°) + Energy £ x (Na2SO4) (solid)

4- [(1 - J C ) Na2SO4 + 10 H2O] (solution),
that is, both solid and liquid phases exist at temperatures above 32 °C.
Since the solid anhydrous salt is denser than the saturated solution, the
two components separate out with the solid settling at the bottom of the
container. On cooling it reacts only partially with the solution to reform
Na2SO4 • 10 H2O and, after a small number of cycles, phase separation
reduces the heat of fusion to a fraction of its initial value. The second
problem, which is also common to other salt hydrates, is that supercooling
can occur when heat is extracted from the store. If this happens the latent
heat of fusion may not be recovered or it may be recovered at a tempera-
ture well below the melting point J 10 1 These problems have been the sub-
ject of a great deal of research. To avoid phase separation, thickening
agents can be added, permitting extensive cycling, but the heat of fusion
of such mixtures is considerably lower than for pure salt hydrates.I411
Supercooling may be prevented by nucleating additives which have the
same crystal structure as the salt hydrate and low solubility in the salt
hydrate melt. Thus, the supercooling of melts of Na2SO4 -10 H2O and
CaCl2 • 6 H2O is reduced to less than 1 °C by the addition of 5% borax
(Na2B4O7 • 10 H2O) and 1% strontium chloride hexahydrate (SrCl2 •
6 H2O) respectively;!411 patents have in fact been granted to Telkes for
such techniques.^39' 42 1 Research continued during the past decades and
interest grew but technical problems and lack of economic viability meant
254 Thermal storage
that few buildings incorporated such storage. Recently, however, com-
plete commercial systems have been marketed in both the US and the UK
using Glauber's salt as the salt hydrate. The UK system, developed by
Calor, depends on the formation, during preparation, of a finely dis-
tributed network of an inert material (authors' note: not identified) which
is able to retain in suspension all nucleating crystals and solids associated
with incongruent transformation in the salt hydrateJ 43 J Since the solids
are dispersed, they remain available for recombination with the solution
and, so, the transformations remain completely reversible. An application
of this approach is described below.
The same group has also worked on Na2HPO4 • 12 H2O which has a
transformation temperature of 35.5 °C. In Sweden, studies of CaCl2 •
6 H2O with a melting point of 29.5 °C are underway I44J with the inten-
tion of application to the solar heating of buildings. The Swedes see par-
ticular promise in phase-change heat storage combined with heat
pumps^451 since the release of heat at a constant temperature permits
uniform evaporator temperatures to be maintained (see Chapter 6) - this
is, of course, not possible when an external air or ground source is used as
a source of heat.
Paraffin waxes are a second type of phase-change material presently
being studied, in spite of previous objections to these being a fire hazard in
buildingsJ391 These materials have relatively high specific heats for
organic materials but low thermal conductivities. The consequence of the
latter characteristic is that rates of heat transfer are low, thus putting heat
in (important if the goal is to charge a store under rapidly changing con-
ditions of solar energy availability) and taking it out (important if, for
example, supplying a fast-response heating system is the requirement) can
be a problem. Studies at Cardiff!461 concluded that 'it is unlikely that a
paraffin wax will supplant water as the storage medium for domestic hot-
water applications', but that the possibility of their use in space heating
systems was not excluded. Research is continuing on different configur-
ations of heat exchangers and a series of waxes with different melting
points (42 °C rather than the 52 °C of the waxes first studied).
Back in 1949, the Dover (Massachusetts) House, claimed to be the first
house heated entirely by solar energy, was built and occupied.I47' The
solar heating system incorporated a double-glazed air-heating collector and
about 13 m3 of storage bins containing Glauber's salt, and was designed,
not surprisingly, by Dr Telkes. Problems in operation of the kind discussed
previously led to a loss of efficiency and eventual removal. Nevertheless,
the experiment was judged to be of value and theoretical work continued.
In 1973, the University of Delaware, where Telkes was working, opened
their experimental Solar One House which again uses air collectors and
storage in salt hydrates, but this time in a more sophisticated configuration
which can be used for heating and cooling.I48' 49 J Two outer vertical
stacks of pans contain sodium thiosulphate pentahydrate (Na2S2O3 •
5 H2O) and constitute the 'hot' store. A central stack, containing a
eutectic, which is mainly sodium sulphate decahydrate, is the 'cold' store.
Developments in storage 255
3
The total volume is about 6 m and storage capacity is about three days'
winter heating load or one day's summer cooling.
Closer to home, the Calor solar domestic hot-water and space heating
system has been installed in houses in Edinburgh, Dublin and Milton
Keynes. The basic system consists of 20—40 m2 of solar collectors (avail-
able roof area has often proved to be an important constraint) used to heat
water whose energy can be stored in one or two 0.4 m3 units, each capable
of storing 216 MJl 5O l; storage is therefore not for prolonged periods. The
storage unit is an unpressurized plastic tank containing, principally, sodium
sulphate decahydrate.
Several home-heating systems have also been designed using what may
seem to be an illogical and surprising material — ice! The latent heat of
fusion of ice is high (333 kJ/kg) (see also Table 7.2) and its cost minimal,
although the cost of containment must be considered. Fracture of the con-
tainer on freezing can be avoided by adding a small quantity of antifreeze
(ethylene glycol) or an appropriate salt and by allowing a free space above
theliquidJ 40 !
In the US, where throughout much of the country there is a need for
heating in the winter and cooling in the summer, an Annual Cycle Energy
System (ACES) operating in the following manner has been proposed for
residential and commercial building: 15l1
(1) Heat obtained by freezing water to remove the heat of fusion is
pumped by a one-direction heat pump to the building.
(2) Ice produced during the heating season is stored and used to provide
air conditioning during the summertime.
Fig. 7.6 shows a schematic of the system.
In the UK the use of ice as a storage medium is found in the Salford
Houses described previously (Chapter 6). Because of the more temperate
climate, for a domestic situation provision need only be made for heating.
7.5 Developments in storage
Ambient energy sources, with their supply variations ranging from the
daily to the seasonal, increase an existing need for energy storage due to
the variation in energy demand associated with patterns of human activi-
ties and weather. There is enormous potential for storage systems, for
application at the points of both energy generation or collection and use,
which are less expensive, provide more-efficient access to the stored energy
and which have a larger thermal storage capacity. Suggestions for storage
(which do not necessarily fulfil all of the criteria given) vary from the
mundane to the exotic. It is encouraging that most may be viewed as
technically feasible and some are so for the very near future.
Starting with sensible heat, at the Ark Project in Canada small-scale
storage in a clay mud is being investigated for a greenhouse heating sys-
tem J 53 1 On a larger scale, research into solar ponds, which can function as
both collectors and stores, continues;!54J initial work in Israel by
Taborl 55 ! has provoked interest in less sunny climates and a proposal has
256 Thermal storage
Fig. 7.6. Schematic of the evaporator. As the

ACESsystemJ 52 ) (For refrigerant reverts to vapour
space heating of hot water: it extracts the heat from
the compressor pumps hot water sprayed on the outside
vapour refrigerant through of the evaporator, making
the heat exchanger. The ice. The ice is flushed to the
refrigerant continues water-ice storage bin. For
through the space heating space cooling: ice-water
coil where it condenses. from the bin is circulated
The now-liquid refrigerant through the chilled water
then moves through the coil.)
expansion valve and
Developments in storage 257
Fig. 1.1. A rolling-cylinder
heat store in a solar
system.! 57 !
been made for such a pond in no less a place than London,t56 l in spite of
the city's image of a capital with an umbrella-carrying populace. In a solar
pond, incident radiation penetrates the liquid in the pond and falls on the
blackened base which is thus heated. Normally, convection currents would
cause the heated liquid at the bottom to rise and transfer the heat to the
surface, but in a solar pond this is prevented by dissolving suitable salts
(for example, sodium chloride, NaCl) in the lower layers of the water to
increase their density while not significantly affecting the transparency.
The temperature at the bottom of the pool then increases and the energy
may be used, for example, for space heating of a large group of buildings.
Novel developments in phase-change storage include the rolling-cylinder
concept shown in Fig. 7.7. The cylinder contains Glauber's salt and by
continually turning at three revolutions per minute provides just enough
stirring action to prevent component separation.
It is also possible to store latent energy in adsorbent beds of materials
such as silica gel and activated aluminaJ 12 ' 581 The principle of operation
is that suitable materials, in addition to their capacity for sensible-heat
storage, remove water from an air stream flowing through them, with a
consequent latent-heat gain.
Mechanical storage devices are generally thought to hold little promise
for use in solar energy systems because of the losses involved in converting
heat to work, but a number of chemical (the term is used loosely since a
number of storage techniques, including phase-change stores, involve
chemical effects) systems for thermal storage are being actively investi-
gated (chemical stores for electricity are discussed in Chapter 8). One such
development is based on the heat of mixing. For example, the well-known
258 Thermal storage
reaction when sulphuric acid and water are mixed releases heat but suffers
from a relatively low energy density (about 100-200 MJ/m3 depending
on temperature). However, by using it as a heat pump, as in a system being
developed jointly by the Rutherford Laboratory and the Open University,
the store can be improved greatly.!591 Suggestions for the longer term
include metal hydrides, some of which can store more than ten times as
much energy as phase-change materials J 60^ Work is also underway on
photochemical reactions using both organic and inorganic materials. Again,
the attraction is that some of the materials involved, for example norbor-
nadiene, have higher storage capacities than phase-change materials.l6> 61J
7.6 Conclusion
As with heating systems, a number of storage techniques are available and
are being used in energy-efficient buildings. Sensible-heat storage in water
is by far the most common solution but phase-change materials are gaining
acceptance as their chemistry becomes understood and as physical prob-
lems are overcome. It may be that sensible-heat storage, on grounds of
simplicity and cost, will become the preferred store for large-scale appli-
cations such as community heating schemes, and phase-change storage will
prevail where space is limited as in individual houses.
The factors that will influence such choices are the same as those
governing the viability of new means of energy storage - progress in devel-
oping and selecting appropriate materials or combinations of materials,
material costs, containment costs, safety and the confidence that exists
among designers and consumers that the store will perform adequately for
a reasonable lifetime. There is a desperate need for more information on
storage media and performance over prolonged periods. The US Workshop
on Solar Energy Storage Subsystems^60] called for research on everything
from store geometries in relation to intermittent solar energy availability
to basic work on heat-transfer properties of materials.
Long-term storage for single houses, it is generally agreed, is not
presently economic by any standard but short-term storage for specific
applications such as domestic hot-water heating in the summer is nearly so.
In the future, what percentage of the domestic hot-water and space heat-
ing load should be met by an ambient source of energy combined with a
store, and what the material should be will depend on the factors cited
above, with cost being the most notable. We can say with confidence,
however, that provision of storage or, as a minimum, space for storage,
should be considered for every building being designed now — within the
lifetime of most of them it will prove its worth. At the very least the space
will be useful to store coal for the future.
References
[ 1 ] Littler, J.G.F. & Thomas, R.B. (1977). 'Solar energy use in the
References 259
[2] Alexandroff, J.-M. & Alexandroff, G. (1977). 'Stockage Longue-
Duree de l'Energie Solaire'. UNESCO International Conference and
Exhibition on Solar Building Technology. London.
[3] Andersson, O. & Gustafson, G. (1981). Thermal energy storage in
deep confined groundwater aquifers'. Swedish Council for Building
Research Summary SI: 1981. Stockholm.
[4] Givoni, B. (1977). 'Underground longterm storage of solar energy
- an overview'. Solar Energy, 19, 617-23.
[ 5 ] Anon. (1981). 'Storage of heat in clay in central Sweden'. Swedish
Building Research News, 1981: 1. Stockholm.
the United Kingdom'. Department of Energy Paper No. 16.
London: HMSO.
[7] Caton, P.G.F. & Smith, C.V. (1977). 'Wind and solar radiation -
availability in the United Kingdom'. In: Energy Conservation and
Use of Renewable Energies in the Bio-Industries. Vogt, F. (ed.).
Oxford: Pergamon.
[8] Ryle, M. (1977). 'Economics of alternative energy sources'. Nature,
267, 111-17.
[9] Tucker, A. (1981). 'Futures'. The Guardian, 21 May, p. 10.
[10] Duffie, J.A. & Beckman, W.A. (1974). Solar Energy Thermal Pro-
cesses. New York: Wiley-Interscience.
[11] (1977). ASHRAE Handbook - Fundamentals. New York:
American Society of Heating, Refrigerating and Air Conditioning
Engineers.
[12] International Solar Energy Society - UK Section. (1976). Solar
Energy -a UK Assessment. London: ISES-UK.
[13] Thomas, R.B. & Littler, J.G.F. (1979). 'Thermal storage in the
[14] Wozniak, S.J. (1979). Solar Heating Systems for the UK: Design,
Installation and Economic Aspects. London: HMSO.
[15] Anon. (1977). Heating and Air Conditioning Systems Installation
Standards for One and Two Family Dwellings and Multifamily
Housing Including Solar. Vienna, Virginia: Sheet Metal and Air
Conditioning Contractors National Association.
[16] Pickering, E. (1975). 'Residential hot water solar energy storage'.
Proceedings of the Workshop on Solar Energy Storage Subsystems
for the Heating and Cooling of Buildings. American Society of
Heating, Refrigerating and Air Conditioning Engineers.
[17] Anon. (1971). IHVE Guide Book B 1970. London: Institution of
Heating and Ventilating Engineers.
[18] Manufacturers data. (1977). Neutra-rust. Middlesex: Double H
International Marketing Ltd.
[19] Anon. (1978). 'Control of corrosion in central heating systems -
experience gained from a few installations'. Greater London
Council D and M Bulletin 118.
[20] Anon. (1978). GRC. BRE Digest 216. HMSO.
[21 ] Tanashita, I. (1970). 'Present situation of commercial solar water
heaters in Japan'. International Solar Energy Conference,
Melbourne.
[22] Wozniak, S. (1979). 'Solar heating systems for the UK'. Building
Services and Environmental Engineer, 2 (2), 16—18.
260 Thermal storage
[23] Field, A. (1981). 'Bridging the load gap from solar and heat
pumps'. Building Services and Environmental Engineer, 3 (5),
16-17.
[24] Esbensen, T.V. & Korsgaard, V. (1976). 'Dimensioning of the solar
heating system in the Zero Energy House in Denmark'. UK-ISES
Conference C8, pp. 39-52.
[25] Anon. (1975). Solar Heated House in Milton Keynes. Milton
Keynes: Milton Keynes Development Corporation.
[26] Anon. (1980). 'Solar heated houses, Milton Keynes'. What's New in
Building, July, p. 5.
[27] Mead, I.K. (1981) Calor Alternative Energy. Private communi-
cation.
[28] Morse, R.N. (1978). 'Solar industrial process heating for can warm-
ing at Queanbeyan, New South Wales, Australia'. CSIRO Solar
Energy Studies Report No. 10. East Melbourne: CSIRO.
[29] Richards, R.A.C. (1980). 'Extensive solar energy use in Spanish
building'. Building Services and Environmental Engineer, 2(9),
14-15.
[30] Anon. (1981). 'Energy consciousness shows at CIBS/ASHRAE
Joint Conference'. Building Services and Environmental Engineer,
3 (10), 9.
[31 ] Roseen, R. (1981). 'A combined solar collector array and heat
store for estate applications — assessment of a prototype'.
Summary S9:1981 of Report R59: 1981 (in Swedish). Swedish
Council for Building Research.
[32] Lof, G.O.G., El-Wakil, M.M. & Chiou, J.P. (1964). 'Design and
performance of domestic heating system employing solar heated
air — The Colorado House'. Proceedings of the UN Conference on
New Sources of Energy, 5.
[33] Anon. (1977). 'Solar energy developments'. RIBA Journal, 84 (3),
103-10.
[34] Taylor, G. (1981). 'Understanding solar heating'. Building Services,
3(1), 48-9.
[35] Jones, B.W. (1980). 'A solar power system (Prometheus) to provide
100 per cent of low-grade heat needs'. Applied Energy,
September-October, pp. 329-46.
[36] T. Oreszczyn. (1981). 'The Open University'. Private communi-
cation.
[37] Telkes, M. (1974). 'Solar house heating - a problem of heat stor-
age'. Heating and Ventilating, May, pp. 68-75.
[38] Telkes, M. (1949). 'Storing solar heat in chemicals'. Heating and
Ventilating, November, pp. 79-86.
[39] Telkes, M. (1974). 'Solar energy storage'. ASHRAE Journal,
September, pp. 38-43.
[40] Randell, J.E. (1977). 'Long term heat storage'. In: Ambient Energy
and Building Design Conference Proceedings, University of
Nottingham, pp. 135—51.
[41] Meisingset, K.K. & Gronvold, F. (1981). 'Latent heat storage in
salt hydrates'. Helios, 11,9-10.
[42] Telkes, M. (1954). US Patent 2 677 664. May 4.
[43] Anon. (1980). Calor Thermal Storage. Slough: Calor Alternative
Energy.
References 261
[44] Carlsson, B., Stymme, H. & Wettermark, G. (1979). 'Low-
temperature storage of heat in salt hydrate melts'. Appendices to
Research Report 750016-8. Stockholm: Swedish Council for
Building Research.
[45] Jonsson, A. (1980). 'Heat pump systems using chemical energy
storage'. Report R145. Stockholm: Swedish Council for Building
Research.
[46] Marshall, R.H. (1980). 'A prototype phase-change heat-storage
device'. Building Services and Environmental Engineer, 2(11),
12-13.
[47] Nemethy, A. (1949). 'Heated by the sun'. American Artisan,
Residential Air Conditioning Section, August.
[48] McVeigh, J.C. (1977). Sun Power. Oxford: Pergamon.
[49] Boer, K.W. (1973). 'The solar house and its portent'. Chemical
Technology, 3, July, pp. 394-9.
[50] Anon. (1980). The Solar Revolution. Slough: Calor Alternative
Energy.
[51] Fischer, H.C. (1975). 'Annual cycle energy system (ACES) for
residential and commercial buildings'. Proceedings of the Workshop
on Solar Energy Storage Subsystems for the Heating and Cooling
of Buildings. Charlottesville, Virginia, pp. 129-35.
[52] Arnold, B. (1979). 'House on ice'. Building Services and Environ-
mental Engineer, 1 (9), 12.
[53] Caffell, A. & MacKay, K.T. (1981). 'Mud storage: a new concept in
greenhouse heat storage'. In: Energy Conservation and Use of
Renewable Energies in the Bio-Industries, Vogt, F. (ed.). Oxford:
Pergamon.
[54] Colbeck, I. (1978). 'A review of solar ponds. Storage in solar
energy systems'. UK-ISES Conference C15, pp. 3 1 - 4 3 .
[55] Tabor, H. (1963). 'Large area solar collectors for power pro-
duction'. Solar Energy, 7, 189.
[56] Bryant, H.C. & Colbeck, I. (1977). 'A solar pond for London'.
Solar Energy, 19(3), 321.
[57] Anon. (1978). 'New heat-storage device'. International Power
Generation. April—May.
[58] Close, D.J. & Dunkle, R.V. (1970). 'Energy storage using desiccant
beds'. International Solar Energy Conference, Melbourne.
[59] Anon. (1979). 'Chemical heat pump and storage potential on test'.
Building Services and Environmental Engineer, 2 (1), 8 - 1 1 .
[60] Davison, R.R. (1975). 'Long-term and seasonal storage group
report'. Proceedings of the Workshop on Solar Energy Storage Sub-
systems for the Heating and Cooling of Buildings. American
Society of Heating, Refrigerating and Air Conditioning Engineers.
[61 ] Anon. (1978). 'A chemical way of heating the home'. New
Scientist, 4 May, p. 295.
8
Wind energy
8.1 Introduction
In the wake of costs for wave power, which have exceeded original esti-
mates, interest in the UK is tending to concentrate on wind power as a
means of central electricity generation using renewable sources of energy.
In such northern latitudes the use of concentrating solar collectors for
power generation is not viable.
For this reason a consortium of companies in conjunction with the
North of Scotland Hydroelectric Board is about to construct a 3 MW
turbine in the Orkneys, illustrated in Fig. 8.1. There are also tentative
plans to place large fields of vertical-axis turbines in shallow parts of the
North Sea off the east coast of England. These turbines would be of the
Musgrove type shown in Fig. 8.15. Studies^1! show that up to 30% of
electricity in use at a given time in the UK could be supplied from variable
inputs such as wind turbines, without upsetting the grid network. South-
ern California Edison expects a 30% contribution by 1991 \2^
However, this chapter seeks to deal with small, local wind turbines. It
has to be pointed out that the energy derived from wind-power devices is
not unending (the turbines probably should have a lifetime of 20 years or
so in the absence of freak weather), is not free (the devices are quite large
and thus costly) and that to some, the machines are not handsome, and to
others, represent a hazard.
It may help to provide a brief checklist of advantages and problems
presented by the home use of wind power, before embarking on a descrip-
tion of the energy and devices available.
Checklist of attributes and problems

- Whilst the free-stream wind speeds can be reasonably high, those
in urban areas and near obstructions can be greatly reduced.
- The power in the wind varies as wind speed cubed (K 3 ).
- To extract, say, 6 kW in a reasonably exposed place requires a
turbine of diameter about 5 m, or the width of a terraced house.
- To rise above the layer of more-slowly moving air near the ground
implies a tower and the visual intrusion may be unacceptable.
- Legal problems arise associated with damage if there is a mishap,
with using the grid as a storage battery and with planning regu-
lations.
262
Introduction 263
Fig. 8.1. 60 m diameter,
3 MW turbine to be built on
Burger Hill, Orkney, by the
Wind Energy Group (a joint
venture of Taylor Woodrow
Construction, GEC Power
Engineering and British
Aerospace).
- Aside from the initial capital cost, maintenance must be carried

out on an object which is up on a tower.
- Acoustic noise and TV interference can occur.
- Storage batteries produce direct current, whereas many household
appliances require AC. The inverter is quite expensive.
- Whilst it might be pleasing to be unaffected by black-outs, unless
the costs of electricity storage are drastically reduced, it is not
realistic to hope to become independent of the grid except in
unusual circumstances.
- In the US, tax credits are available (40% of the first $10 000 via
the Federal Government, and up to 55% of the cost via State
Government); but there is no such assistance in the UK.
- The credits for power returned to the grid when surplus electricity
is being generated are probably going to remain small, if they exist
at all.
264 Wind energy
- In the UK it is not legal to sell electricity to a neighbour across a
plot boundary.
8.2 Power extracted by turbines

8.2.1 Power in the wind
In 1972 Betz derived a relationship between the wind speed K(m/s),
the area swept out by the turbine A (m2 ) and the power P (W) which an
ideal device can extract:
P= 0.645 (,4 X V3)
More specifically, Betz found that:
,4X V3)
where p is the density of air. However, although the air density may
change by 15% during the year and is about 10% less at 1000 m height,
this effect is usually neglected.
This formula gives us four vital pieces of information.
— The power from a device suitable for home use of, say, 4 m
diameter, operating at a wind speed of 5 m/s (roughly the mean
UK wind speed in non-urban areas) is quite small:
P= 0.645 (f-)2 TTX5 3 W,
P=l kW.
- The power varies with the swept area or as the diameter squared.
- The power varies as the cube of the wind speed.
— Only 59% of the power in the wind can be extracted because
sufficient momentum must be left to carry the air away from the
turbine. Only an ideal device actually approaches this value of
59% and most turbines extract considerably less than this, typical
values being around 40%.
The dominant theme then is that the device must be placed in the
windiest regime available. If the wind speed is 5 m/s in the unobstructed
countryside then, in a neighbouring built-up area, it may well drop to
3 m/s. This drop has a drastic effect, the power falling from 1 kW, in the
example above, to 0.22 kW.
8.2.2 Variation of wind speed with location

The wind regimes over the UK have been observed for a long time
and we are therefore fortunate in having a large amount of information
available. The particular aspect of the wind that concerns us is the
duration of wind speeds of different magnitudes over the yearly period.
The raw material gathered at meteorological stations, from which such
information can be obtained, is the average wind speed for each hour of
the year, gathered over a number of years. These data are recorded on an
anemograph and can be used in a variety of ways. Perhaps the most useful
is to determine the Vs0 for a particular site. Vso is the hourly mean wind
speed equalled or exceeded for 50% of the time. To find V50 for a site, the
Power extracted by turbines 265
35.8 Fig. 8.2. Number of hours in
the year when the wind has
the indicated speed.!61
31.3
RHOSSIU DOWN A
STANNSK <EAD B
LEICESTE R C
26.8
22.4
17.9 \ \ \
\
\
\
\
13.4
V
X
8.9
\
" " * * • * ^
* - ^ B
4.5
c
"*—• —
1000 2000 3000 4000 5000 6000 7000 8000 9000

Hours in the year
Fig. 8.3. Cumulative fre-

\• 1 quency distribution of
V/V50 for 35 observing sites
intheUK.l 3 ]
•
-t
<
•
H V-
0-5 10 1-5 20 2-5
hourly mean wind speeds are plotted as a frequency graph, for example
Fig. 8.2. Thus, in Leicester, Vso is about 2.5 m/s, that is, the wind blows
faster than 2.5 m/s for 50% of the year.
Using the same hourly mean wind-speed data, the ratio of V/V50 can be
plotted as a cumulative frequency graph as shown in Fig. 8.3, Vbeing the
wind speed at any moment. This graph was derived from data for 35 sites
around the UKl3J and clearly indicates that once the Vso of a site is
known, the proportion of the time that wind speeds of different magni-
tudes will be equalled or exceeded can be determined. For example (see
266 Wind energy
Fig. 8.4. Hourly mean wind
speeds (m/s) exceeded for
50% of the time, 1 9 6 5 - 7 3 .
Valid for an effective height
of 10 m and a gust ratio of
1.60 and for altitudes
between 0 and 70 m above
mean sea leveU 4 ! (The
gust ratio is defined in the
text.)
(Reprinted with permission,
HMSO.)
Fig. 8.3), for a site with a V/Vso value of 5 m/s, windspeeds of 5 m/s will
be equalled or exceeded for 50% of the time since K/K50 = 1.0. Wind
speeds of 10 m/s, corresponding to a V/Vso value of 2.0, will be equalled
or exceeded only 7% of the time.
Ideally, the value of K50 should be determined by long-term obser-
vations at the site in question. However, a reasonably reliable short-cut to
finding the distribution of different wind speeds is available by using a
map of Vso (Fig. 8.4). This mapl 4 ! applies to open inland sites and refers
to the Vso at an effective height of 10 m. For sites and heights other than
the ones to which the map refers, modification factors must be applied.
For sites other than the open inland site to which the map refers, the
map value is modified by considering the gust ratio at the site in question.
Since January 1973, anemograph stations have measured for each clock
Power extracted by turbines 267
Table 8.1. Variation of gust ratio with terrain^
Gust ratio Modification factor

Type of site (G) for map value 1.6/G
City 2.10 0.76
Suburban 2.00 0.80
Open inland 1.60 1.00
Coastal 1.45 1.10
Hill 0.80 2.00
Table 8.2. Variation of wind speed with height (from Caton^ and
Rayment^)
Type of site Power-law exponent Modification of map value
City0 0.23 0.4476 H023

Suburban 0.22 0.4820 H022
Open inland 0.17 0.6761 H017
Coastal 0.13 0.8154 H013
Hill
a
Heights measured upwards from prevailing roofline.
hour the maximum gust speed of about 3 s duration and, simultaneously,

the hourly wind speed. The maximum gust speed for each hour has been
divided by the mean speed for each hour and the average value from
several years' data has been taken. This is the gust ratio (G). It has been
found that sites with similar terrains have similar gust ratios. For example,
the gust ratio for city sites is about 2.1:1 whereas the ratio for open inland
ones, to which the map applies, is about 1.6:1. To find the relevant V50,
one therefore has to multiply the map value by 1.6/G, where G is the gust
ratio characteristic of the site (see Table 8.1).
8.2.3 Variation of wind speed with height

For heights other than the standard effective height of 10 m, the
value can be modified using the power law:
V=cXHP,
where V is the wind speed, c is a constant, H is the effective height and p is
an exponent characteristic of the underlying terrain. For an open site,//
can be taken to be the height above ground. For a city site, where build-
ings are densely packed, it has been suggested that the effective height
should be the height above the general level of rooftops. For a suburban
site, the effective plane will be dependent on the extent of screening by
buildings and trees, but for a low-density, low-rise region, it seems reason-
able to assume the zero plane to be the ground level as long as the prevail-
ing wind is not blocked by nearby trees or buildings. Exponents and
constants for modifying the map values of Fig. 8.4 are shown in Table 8.2.
268 Wind energy
We are now in a position to judge V50 for any site in the UK.
Raymentl5! has shown that the total annual energy in the wind (E^) in
the UK is such that
Ev/i = 43.43 (K 50 ) 3 MJ/(m2 y),
when V50 is measured in m/s.
Using the wind map and the modification factors for VSQ , it is therefore
possible to determine the annual energy available in the wind for any site
and height. For example, suppose we wish to consider a site in the Mid-
lands in a non-urban open inland area with no major obstructions, a value
of V50 from Fig. 8.4 of 4.5, and a tower height of 25 m. The terrain
modification from Table 8.1 is 1.0. The height modification from Table
8.2 is:
(0.6761) (25 m)°- 17 = 1.17.
Thus, Vso for the site is
(4.5 m/s) (1.0) (1.17 ) = 5.26 m/s
and
E^ = 43 A3 (5.26)3 MJ/(m2 y)
= 6.32 GJ/(m2 y).
Thus, a turbine with a diameter of 5 m and a swept area of 20 m2
encounters an annual wind energy of 127 GJ/y.
As we have seen, even the most efficient device could only extract 59%
of this, or 75 GJ/y and, unfortunately, the energy actually available for
use will be very much less than this.
Before moving on to a discussion of further losses, a city example is
considered. If the site is in London, in a congested position with a mast at
25 m but rooftops at 8 m, the following data might apply:
Vso (Fig. 8.4) = 4.0 m/s,
Terrain modifier (Table 8.1) = 0.76,
Height modifier (Table 8.2) = 0.86,
E^ = 43.43 (4.0 X 0.76 X 0.86)3 MJ/(m2 y)
= 774 MJ/(m2 y).
Thus, even our ideal device of 20 m2 could only extract 9.14 GJ/y com-
pared with 75 in the country.
Fig. 8.5 illustrates the energy in the wind across the UK.
8.3 Power extracted by real turbines

The actual power extracted by a turbine depends on three factors, namely
the cut-in speed, the rated speed and the power coefficient.
The lowest wind speed which will overcome friction in the turbine bear-
Power extracted by real turbines 269
Fig. 8.5. Energy in the wLid
at an effective height of
10 m in open country-
side.! 5]
ings and actually cause the device to produce useful power is called the
cut-in speed and is typically 3-4.5 m/s (a survey of commercially available
devices with rated outputs of less than 50 kW gives a mean cut-in speed of
4.2 m/s).
Above the rated wind speed, the turbine, gearbox, generator and mast
would be overstressed unless wind is spilled, thus output is constant above
the rated speed until at even higher speeds the turbine is shut down to
avoid damage.
The power coefficient is the ratio of power extracted by the blade sys-
tem to the power available in the wind and thus has a maximum value of
59%.
Returning now to the rated speed, Fig. 8.6 illustrates data derived from
Fig. 8.3 and shows the value of (V/Vso)3 (which is proportional to the
power in the wind) against the number of hours in the year.
Suppose V50 is 4.5 m/s, then a high wind speed of 12 m/s would have a
270 Wind energy
Fig. 8.6. Power duration
curve typical of the UK.
20
Energy rejected by fixing
18 cut-in speed at V50
16
14 Energy rejected by fixing
cut-in wind speed at 12 m/s
12
10
o
a. 8
6
4
2
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0x8760
Hours in the year
value of (V/V50)3 of 19. But this speed is only available for 0.01 of the
time during the year. The area under the curve in Fig. 8.6 is proportional
to the energy per year in the wind. It is clear that the fraction of power for
wind speeds above 12 m/s is small, yet to be able to absorb these wind
speeds the structure of all the components must be adequate. A cheaper
and less-robust system could be built if the rated speed were only 10 m/s,
for example. Optimization of the performance suggests that rated speeds
should be about 2.25(K5O).
The cut-in speeds can then be decided. For example, if a cut-in speed of
Vso is chosen, then (V/V50)3 = 1.0. The value of (V/V50)3 at the rated
speed is 2.253 or 11.4, thus at Vso the device is only operating at 1.0/11.4
of its maximum capacity. The part load efficiency of components such as
gearboxes, bearings and generators can be very poor. Moreover, the power
available annually, which is rejected by not using winds below V50, is small
as can be seen in Fig. 8.6. Typically, then, cut-in speeds should be between
K 5 O andl.3(F 5 O ).
It should be noted that 80% of the wind energy normally lies between
speeds of 1.125 X Vso and 2.5 X Vso, and if our 'perfect' device of 20 m2
operating in the Midlands has a cut-in speed of 1.125 (4.5) or 5 m/s and a
rated speed of 2.5 (4.5) or 9 m/s it would capture 58 GJ/y.
8.4 Types of turbine

Earlier in this chapter the Betz relationship was quoted, which suggests
that a device can only extract 59% of the wind power. There is some dis-
agreement about the theoretical maximum and Sabinin has suggested a
figure of 69%J61
There is, however, little doubt that practically speaking the best figure
currently achieved is an extraction rate of about 45% of the energy in the
wind, by aerodynamically designed blades. (There is a little confusion
about power coefficients since some authors have quoted values as a per-
centage of the Betz maximum, for example 60% of the theoretical maxi-
Types of turbine 271
Fig. 8.7. Power coefficients
0-6 — for different types of
A
B
SLOW-SPEED MULTI-8LADED R OTOR '
rotor.t 6 !
4-BLADED SAIL-TYPE ROTOR
C HIGH-SPEED ROTOR
<~> O-5
^-—
/V V k
O 0 3
u
A
cc
£ O-2
A
/ \ 4 6
TIP SPEED RATIO /lQ
10 12
mum value of 59%. In this text the Cp is quoted as the fraction of power
in the wind, that is, 36% in the above case of 59 X 60%.) GoldingI6]
quotes the values in Fig. 8.7.
The power coefficients are plotted against the ratio of the speed of the
blade tip (that is, 2irrN) to the upwind air speed, where r is the radius and
TV the rate of revolution; for example, a large device such as Growian II in
Fig. 8.8 has a radius of 73 m and rotates at 17 rpm, giving a tip—speed
ratio of 6 at a wind speed of 22 m/s.
It is clear from Fig. 8.7 that the power coefficient, although fairly
broad for well-designed airofoil blades, falls off quite quickly on either
side of the peak. A good design would be one chosen so that, at a par-
ticular site, the wind speeds would be such that power output over the
year was maximized and this implies keeping the tip—speed ratios within
fairly narrow bounds to preserve a good power coefficient.
In the Midlands example quoted previously then, the output will fall
from 58 to about 36 GJ/y because of the actual value of the power
coefficient of even a well-designed rotor.
Types of rotor
A large number of forms have appeared over the course of time
including the vertical-axis Chinese windmill of Fig. 8.9 and the horizontal
one described by Hero of Alexandria in the third century BC, shown in
Fig. 8.10.
To extract power from the wind, the turbine must first intercept it as it
passes through the area swept out by the blades or vanes. To do this at low
wind speeds when the turbine is rotating slowly requires a large number of
vanes. Thus, the multibladed pumping mill is able to perform efficiently
when the tip—speed ratio is low.
However, as the wind speed increases, the large area of vanes exerts a
considerable rotational force on the air column and efficiency falls. This
can be seen in Fig. 8.11. If the turbine consists of two or three thin
aerofoil-section blades, at low wind speeds the blades do not rotate fast
272 Wind energy
7
Fig. 8.8. Growian turbine! 1
with a 125 m tower.
(Reprinted with permission
© 1981, Popular Science,
Times Mirror Magazines Inc.
Photograph by C. De Groote.)
enough to intercept all the wind passing through the swept area and the
efficiency is low. At higher wind speeds, the blade tips may be rotating at
about six times the wind speed and can extract power efficiently. This is
because the thin blades do not present a large area to the wind and so drag
is reduced and the air column is not caused to rotate. The power
coefficient of such turbines is in the region of 0.45 when the tip-speed
ratio is about 6 (Fig. 8.11). Since turbines with aerofoil blades work more
efficiently than other types, and do so at high rotational speeds, they are
ideal for combining with an electrical generator because the amount of
gearing-up will not need to be as great as with other types.
Figs. 8.12 and 8.13 illustrate the Savonius and multibladed devices also
shown in Fig. 8.11, both of which are more suitable for pumping irrigation
water than generating electricity.
The two categories of aerofoil turbines that can be used for electricity
Fig. 8.9. Primitive vertical-
axis Chinese windmill. 161
Fig. 8.10. Small horizontal-

type windmill blowing an
organ (from The Pneumatics
of Hero of Alexandria^})-
generation are the horizontal-axis propellor types and the vertical-axis

types. The horizontal-axis machines have a turbine with a rotating shaft
that is parallel to the ground and supported on a tower, for example Fig.
8.8. It is immediately apparent that the turbine must cope with changes in
wind direction. If the turbine is mounted downwind of the tower, as is
usually the case with large aerogenerators of this type, the turbine may be
self-directing. If it is designed to be upwind of the tower, as is often the
case with smaller aerogenerators, a tail vane is required. There are advan-
tages and disadvantages associated with each strategy.
If the turbine is downwind, the blades can be coned. This means that
the blades are inclined at a small angle so that their tips are downwind of
274 Wind energy
Fig. 8.11. Power coefficients
for various turbines.
Ideal efficiency
for propeller-type
generators
Dutch four-arm type
1 2 3 4 5 6 7
Ratio of blade tip to wind speed
Fig. 8.12. Savonius rotor

and air flow through the
blades.! 8]
J/Vind
the hub about which they rotate. This is done to reduce stresses. Since the
turbine reduces the axial momentum of the wind, it must experience an
axial force which the blades and tower have to resist. This axial force is
independent of the number of blades in the turbine and is solely deter-
mined by the power that the turbine extracts. Since aerogenerators have
'low-solidity' turbines (the ratio of the total frontal area of blade pre-
sented to the wind to the area swept by the blades has a low value,
typically 0.05), the stress that the axial force produces per unit area of
blade will be high. If the blades are coned, the centrifugal force at high
rotational speeds will make the blades attempt to assume an upright
position. This will tend to counter the effects of the downwind axial force.
The disadvantage of a downwind turbine is that once every revolution
each blade must pass through a patch of relatively calm air in the lee of the
tower. This will momentarily change the stresses in it and can have highly
undesirable effects. For example, the NASA 100 kW machine in Fig. 8.14
was found to have blade stresses twice as great as was predicted due to the
'shadow' cast by a staircase on the tower.
To avoid such problems, the supporting tower for any downwind-
Fig. 8.13. Multiblade turbine.
Fig. 8.14. ERDA/NASA

100 kW wind turbine at
Plum Brook, near Cleveland,
Ohio.
276 Wind energy
mounted device should be as 'transparent' as possible while still being
designed to withstand the axial force transmitted to it from the turbine.
If the windmill is mounted upwind of the tower, a tail vane will be
required, which will add to the expense of the machine. This vane can also
be used for feathering the blades. Although upwind mounting considerably
reduces the problem of transient stress changes in the blades it does not
eliminate it. There will be a patch of calmer air that builds up in front of
the tower. Moreover, it is now the tower that has to withstand fluctuating
stresses as each limb shields it from the wind once per revolution. If the
blades need to be coned, they will have to be arranged so that their hub is
some way upwind of the tower to prevent the blade tips hitting it. This
will lead to a considerable eccentric load on the tower.
Apart from the detailed aerodynamic characteristics of the blades,
which will not be considered here, the question of how many blades to
have in the turbine, and what the solidity should be, arises. For maximum
efficiency, drag and air-column rotation should be minimized so a low
solidity is desirable. However, as previously described, the axial force is
independent of the number of blades. It is therefore impractical to aim for
a very low solidity unless very strong blade materials can be used to with-
stand the stresses. The number of blades between which the solidity is
divided is usually two or three. Two blades mean that static and dynamic
balancing is simpler so tower vibrations and blade stresses are likely to be
reduced. Since the turbine will probably have to be balanced where it is
manufactured, a two-bladed turbine presents less of a transportation prob-
lem. However, the main difficulty with two-bladed turbines is their
behaviour when the wind changes direction. The moment of inertia of the
blades about the yaw axis changes every 90°, and so at every revolution
the blades and tower experience two periods when there is considerable
resistance to tracking the wind. This will cause unwanted stresses in the
system. With three blades this is no problem although the difficulties of
balancing and transportation apply.
With horizontal-axis aerogenerators, the turbine shaft must be con-
nected to the gearbox and alternator at the top of the tower. Although it
would be possible to use a right-angle drive shaft to connect the turbine
shaft to a gearbox and alternator at ground level, the torsion in such a
system would appear to be a problem. The tower then has to withstand
the considerable loads imposed on it by the turbine, the gearbox, and the
alternator, as well as that imposed by the axial force on the turbine and
the wind loads on the tower itself. With the alternator at the top of the
tower, some means must be found to lead the electricity generated to
ground level, while still allowing the turbine to track the wind. Clearly, a
simple cable would become twisted. This is normally achieved by the use
of slip rings which are liable to cause maintenance problems.
The second type of aerogenerator is the vertical-axis machine. In the
Savonius rotor (Fig. 8.12), two half-cylindrical blades are offset about a
vertical axis in such a way that the wind can pass between them. In
addition to the pressure on the concave surface facing the wind, the wind
is directed through the machine so that it produces an additional pressure,
assisting rotation, on the back side of the other half of the rotation.
Although this type of machine is very simple to construct, often from an
oil barrel cut in half, its performance characteristics (see Fig. 8.11) are not
suitable for the generation of electricity due to its inability to operate at
high speeds.
The major advance in vertical-axis machines was made by Darrieus in
1926, although it was not until the late 1960s that South and Rangi
rediscovered his patent when they tried to file their own for a similar
machine. The first type of Darrieus machines employ curved aerofoil-
section blades rotating about a vertical shaft. In order to reduce bending
stresses, the blades are usually shaped to a 'troposkein' curve, the configur-
ation that a skipping rope assumes. This means that when rotating at high
speed, the only major stresses in the blades are tensile ones.
The aerodynamic principle that applies to such machines is the same as
that for horizontal-axis aerofoil machines. With a horizontal-axis machine
a torque is produced continually as long as the turbine is facing the wind.
With the vertical-axis machine, however, power is extracted from the wind
when the blades are crossing the path of the wind. When travelling directly
up- or downwind, no torque is produced. However, the cyclical nature of
the power extraction is minimized by the momentum of the turbine. With
a three-bladed turbine, there will be no part of each revolution during
which the blades do not extract power. As with horizontal-axis machines,
the blades should be of the lowest possible solidity for maximum
efficiency. If this is done, the power coefficient of the turbine can be
comparable to a horizontal-axis machine when the tip—speed ratio is about
4 (see Fig. 8.11). The Darrieus-type machines have a large proportion of
the blades a considerable distance from the tower so the 'shadow' effect of
the tower is largely avoided. Moreover, because the shaft of the machine is
vertical, the alternator and gearbox can be at ground level which avoids the
need for a sturdy tower. Another advantage of the Darrieus-type is that
the machine can extract power from the wind without needing a direction-
ing system. This means that no tail vane or slip rings to carry current from
the alternator are required. However, troposkein machines do have several
drawbacks. The curved aerofoil blades are expensive to manufacture and it
is only the outermost portion of the blades that sweep an appreciable area
of wind thereby extracting power.
The Musgrove variable-geometry vertical-axis windmill has upright aero-
foil blades attached to a cross-member that rotates about a vertical axis.
For a given diameter and height, the cross-section of wind swept out by
the blades is therefore much greater than that swept by a Darrieus
machine. This leads to a greater power output. Although the blades have
to cope with bending stresses, being straight they are cheaper to make. To
reduce bending stresses in high winds and to reduce the power output, the
blades are allowed to incline towards the horizontal as centrifugal forces
overcome the restraining force of a wire attached between the vertical
shaft and the blades. This reduces the swept area and hence the power out-
put, as well as allowing the blades to take stresses increasingly in tension
rather than in bending. Fig. 8.15 illustrates a Musgrove turbine built at the
278 Wind energy
Fig. 8.15. Musgrove turbine,
diameter 6 m, blade length
4 m.
Rutherford High Energy Physics Laboratory and designed to provide elec-

tricity for the Cambridge Autarkic House.
8.5 Electrical generation

8.5.1 Electrical generators
Various kinds of generators can be attached to a wind turbine. The
choice of system will generally have been made by the manufacturer of the
wind-power system and a consumer will usually elect a system based on
the following questions:
— How large an electrical output is required?
- Is the output to be used only for heating?
- Is the output to be stored in batteries?
— Is the system to be connected to the electricity grid?
There are three classes of generator in common use for turbines, two pro-
ducing alternating current and one direct current. A survey of about 50
commercially available wind systems in the range 0-50 kW shows that
small systems (0-5 kW) tend to use DC generators. Intermediate systems
supply AC from alternators and large AC systems use induction generators.
A fourth type, the synchronous generator is as yet uncommon.
In the past, large DC systems have been proposed, as Goldingl6! points
out. Table 8.3 summarizes some of the properties of generators.
The first problem with designing the system lies in matching the power
output from a turbine which varies as Cp X V3, with the power converted
by a generator which varies at best as the (rotational speed)2. The second
problem concerns matching the load with the generator output.
Electrical generation 279
Table 8.3. Selection of generators
Efficiency
Relative at full load
cost/kW
DC generator
wound field 5 90
permanent magnet 4 93
AC alternator
wound field 3 90
permanent magnet 3 93
vehicle alternators 20-70
AC induction generator 1 90
Mechanical Fig. 8.16. Overall power

output coefficients of two 10 kW
Power in wind commercial systems
(diagram plotted from data
inMarierU°l).
Electrical
output
MJ
1.4-
0.7-
1 2 3 t4t 5 6 7 8 9 12 13 14 15
Cut-in speeds
Wind speed (m/s)
The wind turbine, if terminated by a fixed resistive load, such as an

immersion heater, will face a very large starting torque at low wind speeds
(thus delaying the onset of generation). It will tend to stall in fluctuating
winds, even when, for much of the time, generation would be possible.
The overall performance will thus be markedly improved with the use of
suitable control.
At low speeds a system of switching resistive loads in and out allows the
load to be very small when the turbine starts up; allows higher rotational
speeds, thus moving to a more efficient part of the Cp curve; and is able to
bring in more and more load as the wind rises to simulate a load varying as
V3. The field current may also be varied (in field-wound generators) which
over a limited speed range allows the electrical output roughly to vary as
V\
Power coefficients for the turbines were shown in Fig. 8.11 but the
concept can be extended as in Fig. 8.16 to cover the turbine plus gener-
280 Wind energy
ator. In this case then, C po represents the fraction of power in the wind
which is converted to electricity. Whereas power coefficients for the tur-
bines reached values of 0.45, overall power curves only reach values of
about 0.3. (The curves shown are for a Windworks 10 kW system, with
10 m diameter, three-b laded horizontal-axis rotor and a 240 V DC alter-
nator, and a Millville '10-3-IND' with 7.6 m diameter, three-bladed
horizontal-axis rotor and a 220 V AC induction generator.)
It is clear then that the generators and gearboxes cause additional
power loss. This loss appears as heat, and if the equipment is within the
heated volume of the house, such heat can be valuable.
8.5.2 Gearboxes
The electrical generators with small numbers of poles all require high
shaft speeds of 1000 rpm or higher. The turbine shaft is rotating typically
at 50-100 rpm for a 10 kW device and the gear ratio is about 15:1. Losses
in gear boxes or timing belts or chains to overcome friction are significant
particularly at low loads and Fig. 8.17 indicates typical values. Similarly,
except for the very highly developed rare earth permanent magnet gener-
ator! 16) mentioned in the previous section, major losses also occur in the
generator. Thus the overall power coefficients of Fig. 8.16 now look
reasonable. At the rated power of 36 MJ, the percentage recovery factors
are as follows:
Power in the wind 100(%)

(1) Turbine power coefficient 40-45
(2) Gears 92
(3) Generator 80-85
Overall power coefficient (product of 1, 2 & 3) 30-35
Most generators require a field current. At low wind speeds, residual

magnetism will be insufficient to allow electrical generation. Thus the
generator curve of Fig. 8.17 is depressed still further in many cases by the
provision of field power.
8.5.3 Electricity storage

During calm periods, users must rely for electricity on the grid, a
back-up supply such as a diesel generator, or batteries.
Estimates of the duration of calm periods vary rather widely, but when
the authors examined five years of hourly wind speeds recorded in East
Anglia, the longest calm period (defined as one incapable of generating
electricity from a 5 kW system with its attendant inefficiencies) was 42
days.
The average electricity consumption in the UK for needs other than
space and water heating is about 6.5 GJ/y plus about 6.5 GJ for cooking.
To store the energy for one day's use of lights and appliances (about
18 MJ) requires about 110 kg of lead acid batteries, at a cost of about
Electrical generation 281
1.0- Fig. 8.17. The part-load
efficiencies of gears, trans-
Gears + transmission mission and alternator,
0.9- excluding any field
current supplied to the
0.8 - generator.
0.7-
0.6 -
I °-5 -
LJJ
0.4 -
0.3 -
0.2-
0.1-
0 0.2 0.4 0.6 0.8 1.0

Fraction of full rated load
£350. To store 18 MJ/d for 42 days would cost almost £15 000 (at 1982
prices) for the batteries alone.
In most cases then, the turbine system will be used either to supply
only heat (in which case thermal storage can be used) or to replace grid
electricity supplies when the wind is blowing. In isolated cases connection
to the grid may be very expensive and batteries may be used. Car batteries
are unsuitable, being designed for short-term but heavy demands, but
industrial truck cells, whose discharge is normally less than 50% should last
more than ten years. Lead antimony units will be less expensive than lead
calcium ones but will require more frequent topping up. A large number of
cells requires suitable access space, ventilation and temperatures kept
between about 10 and 25 °C.
Exotic batteries
Very large amounts are being spent on battery research for load
handling in grid systems and vehicles. Useful reviews of new systems are to
be found in the Electrical Review and US Popular Science.
Table 8.4 illustrates some of the cells under evaluation. At present there
is no real alternative to lead acid cells for the storage of electricity on a
small scale.
The cells require a charging unit which tapers the energy supplied, to
prolong battery lifetime. An inverter is also needed in most cases, to
transform the DC supply from the battery to AC of the correct frequency.
Many motorized appliances will not work with a DC supply, televisions
282 Wind energy
Table 8.4. New cell systems (adapted from Ref. [11])
Ni/Fe Ni/Cd Zn/air Fe/air H 2 O2

Pb/PbO2
General data H2SO4 KOH
State of Proven Traction Small Lab. cells Prototypes Demonstration

development batteries batteries cells
Operating 30 30 30 « 40-50 ^40-50 « 100-200
temp (°C)
Overall 71 55 53 43 47 42
efficiency (%)
Cycle life 2000 4000 2000 1500 1500 1500
Cost (£/kWh) 25 65 250 75 23 -
and other electronic systems demand a supply close to 50 or 60 Hz. Since

motors call for a starting current of about five times their running current,
quite large inverters are needed and a 3 kW device will cost about £1500.
When the turbine system is used as a grid substitute in windy conditions,
an induction generator will provide a constant voltage and frequency out-
put at the values of the grid and a current which varies with the power
abstracted from the wind. Thus the output is suitable for direct con-
nection to the house system and when more power is produced than can
be absorbed, the energy could, with CEGB agreement, be fed back into the
grid. When the mill is not producing power, unless automatic disconnec-
tion is provided, the grid would supply the generator turning it into a
motor.
If a DC generator has been installed it may still be connected to the grid
(thus avoiding the use of batteries) via a Gemini synchronous inverter,
which normally converts the DC to AC for use by appliances in the home
but in time of surplus sends the excess down the line. The Gemini does not
produce a perfect sine wave and may cause some problems with sensitive
devices such as microcomputers.
8.6 Use of wind-generated electricity

8.6.1 AC or DC?
Many household appliances demand AC. Power devices generally use
induction motors, since they are robust and cheap, and these will not work
on DC. Electronic equipment demands AC with a closely regulated fre-
quency and a reasonably sinusoidal wave form. Transformers fail at
markedly low frequencies and synchronous clocks and phonographs will
turn at incorrect speeds if the wrong frequency is supplied.
There are always alternative pieces of equipment, often of lower power
demand, designed, for example, for boats and caravans; but in general it
will be highly inconvenient to have only DC power supplied to a house.
Furthermore, if low voltages are used, extra costs will be incurred due to
Use of wind-generated electricity 283
the heavier cables needed. DC demands extra robust switches and thermo-
stats. Normal fluorescent light fixtures should not be used on DC. Any
appliance which relies only on heating will work equally well on AC or
DC.
8.6.2 Inverters
Our recommendation then is that 240 V AC (UK) or 120 V AC (US)
be used. There is unlikely to be enough power to charge a thermal store
and, unless grid connection is made, cooking would have to be by other
means. To convert the DC power from batteries an inverter is required (or
a motor-generator set). At peak load an inverter has an efficiency of about
90%. Any losses appear as heat which may be useful. At part loads, since
power is required to run the inverter, the efficiency drops to, say, 80%.
When not in use the power drawn is typically 50 W, which over 24 hours
would represent a large power drain on the battery. Thus the device should
switch itself off when no power is demanded. A more extensive review of
problems associated with inverters can be found in Wind Power for the
Homeowner,l10l andCox^ 12 ! reviews the problems caused by present-day
inexpensive inverters which may feed reactive power into the grid if such
connection is made.
8.6.3 Thermal storage

In many cases, those using wind power will connect to the grid via a
Gemini^13! synchronous inverter or use the power for space heating or
domestic hot-water heating. Such uses avoid most of the control problems
associated with batteries.
Surplus energy may be stored as heat. In the UK, storage heaters oper-
ating on off-peak electricity are common and can be used to absorb power
produced by a turbine and generator. According to the type of generator,
the current, and possibly the voltage, will vary with the strength of the
wind; a simple control system which connects more and more heating
elements in the circuit will be necessary to match the power uptake to
power output. One great advantage of the heating mode of use of wind
power is that even short bursts of energy can be used and the system will
not be upset by a varying source of power.
Instead of heating the bricks (which, being made essentially of iron ore,
are of high thermal capacity) in a storage heater, water may be heated. As
discussed in Chapter 7, water cannot practically be used much above 80 °C.
This means that the heat capacities of the two kinds of store are rather
different, as shown in Table 7.2.
To reduce the surface temperature of a storage unit at 500 °C to a safe
value of 30 °C would require about 300 mm of high temperature insulation.
This degree of insulation would ensure that less than 15% of the heat
would escape via the insulation over the first five-day period if the store
were left unused. Hot water could be provided via an air loop to a heat
exchanger for domestic use. Watson discussed such storage further.f81
284 Wind energy
8.7 Economies of scale

Generally, the high costs of electrical or thermal energy storage militate
against sole reliance on wind power at the one-house scale for either elec-
trical or heating demands. For example, an electrical store to cover the
long calm periods found even in the UK would cost between £10 000 and
£20 000. A thermal store using high thermal mass bricks would have a
volume of about 10 m3 including insulation (6 m 3 ) and would cost about
£2000 including a control system for charging and an air loop for heat
extraction, on the assumption that all the space heating and domestic
water needs for 40 calm days would be stored.
If smaller storage units are used the savings on network energy to the
consumer are only the marginal ones of electricity, gas, etc. at the pre-
vailing rate, and not the metering or installation charges.
Watsonl8! has looked exhaustively at the economies of scale and found
that if electricity prices rise at 12%/y for 20 years, then for sites with wind
speeds such that Vso is > 4 m/s, a turbine heating system with grid back-up
is cheaper than an all-grid heating system. The fraction of heat supplied by
wind power is 70% for the optimally designed system. Watson indicates
how to size the system and in this particular case finds that a ten-day store
and a turbine of 60 m2 swept area (V50 = 4 m/s) is the cheapest solution.
(Areas vary according to V50, for example at Vso = 5.0 m/s the area is
31 m 2 .) However, such a system competes with gas only at sites for which
V50 > 6 m/s.
By contrast, Watson finds that even at the lower fuel-inflation rate of
8%/y for a group of 20 houses, a turbine system and central store with grid
back-up is less expensive than an electrical system solely supplied from the
grid, for all sites with Vso > 4 m/s;and is less expensive than a gas system,
for all sites with V50 > 5.5 m/s. Table 8.5f8l illustrates this point.
8.8 Installed systems

8.8.1 Conservation House at NCAT
Fig. 8.18 illustrates the Conservation House at the National Centre
for Alternative Technology (NCAT) (Machynlleth, Powys, Wales).
The calculated specific-heat loss for the house is 66 W/K. The peak
demand is 1.2 kW and the space heating requirement with incidental gains
taken into account is only an average of 29 MJ/d in December and January
and a total of 3.4 GJ over the heating season. This compares with 47 GJ
over the heating season for an equivalent sized conventional house.
Heating system
An air-to-air heat pump was chosen for space heating and the low-
power requirement enabled a small and cheap system to be employed. The
compressor is the same type as that used in a domestic deep freeze. The
evaporator heat exchanger in the roof space is fed with outside air mixed
with ventilation air from the kitchen extractor fan and the condenser is
mounted with the compressor in a cupboard on the upper floor. Warm air
6
Table 8.5 \ 1 Total* present values (PV) of wind-powered system for a
group of 20 houses. Discount rate 4%, fuel inflation rate
Present value of items in each house

(£)
Back-up energy bought 1728
Thermal store 243
Heat distribution 284
Radiators, etc. 1068
Present value of aerogenerator per house Total present value
F50(m/s) (£) (£)
3.5 7481 10804

4.0 5426 8749
4.5 4089 7412
5.0 3180 6503
5.5 2524 5847
6.0 2050 5373
Present value of electricity and heating system per house (no turbine sys-
tem) £9885
Present value of gas and heating system per house (no turbine system)
£6064
a
Prices at May 1979.
Fig. 8.18. The NCAT Con-

servation House.! 141
Wind
Central heating
driven
heat pump Water
generator
heating
heat pump
240 V AC I
240V
Cloakroom Kitchen l
Inverter
vent fan vent fan 11QV
Insulated
cooker JI Standby
BATT ERY
Small
power
Lighting
O O • STORE Voltage
regulator
generator
110V DC L A standby generator is incorporated

to boost the system during long
windless periods in cold weather and
when the wind generator is being
serviced.
286 Wind energy
is fed to all rooms via the underfloor spaces and returned via ceiling-level
vents. The refrigerant circuit can be reversed by a changeover valve for
automatic defrosting of the evaporator and for summer cooling should this
prove necessary. Experience of the hot summer of 1976 suggests that, in
this location, cooling will not be needed, as the long thermal time constant
of the house provides a very stable temperature with little diurnal variation.
In 1978 a propane-fired heating system was installed experimentally to
facilitate measurements of the building's performance. Preliminary results
show a heat loss close to the calculated value and suggest that heating costs
on low-priced fuels should average around £1 per week (at 1978 prices).
Hot water
A similar, small heat pump is used to heat the 200 1 hot-water tank.
A large volume of relatively low-temperature (50 °C) water is used to
maximize the COP while maintaining an adequate hot-water supply (for
some purposes cold water will not need to be mixed with it). The evap-
orator of this heat pump is immersed in a 275 1 waste-water tank below
the ground floor. By discharging water at a similar temperature to the
cold-water-feed temperature, or colder if necessary, sufficient energy can
be extracted to maintain the temperature of the hot-water cylinder. It has
been estimated that heat lost from the water to the house is approximately
balanced by the electrical energy supplied to the heat pump. A COP of 2
has been measured for this system. Electrical consumption is estimated at
16.2 MJ/d or 5.90 GJ/y.
Energy supply
The house described, so far, as having a total energy demand of about
one-fifth of a similar sized conventional house, is suitable for normal mains
electricity powering, but as the emphasis at the Centre is on ambient
energy use and the site is not connected to the national grid, a 2 kW
Dunlite aerogenerator was installed next to the house, together with 72 MJ
of lead acid battery storage. On a good site this would supply most of the
energy required, except for long windless periods combined with cold
weather. With the relatively low wind speed near the house, there is
insufficient wind energy for heating but adequate electricity for lighting
and cooking is supplied at 110 V DC. It cannot be claimed that the wind-
electric system is generally a cost-effective option, mainly because of
battery replacement cost. However, the initial cost of the system (about
£4000) is similar to the cost of connecting isolated premises to the
national grid, and in such situations a wind-power scheme is much more
attractive financially. However, a house designed specifically for wind-
power heating should incorporate thermal storage (rather than batteries)
equivalent to at least a week of winter load.
Problems experienced
— The 2 kW wind generator was poorly sited being next to the house
and in the shadow of surrounding ridges.
- As a result the energy output has been less than expected and
Fig. 8.19. Wind-powered
house at NCAT.11*1
D.H.W. cylinder
Thermostatic mixing Control unit

valve Auto shutdown^conrtroI
I* AC «n^99n\/ ^ . ^
Pump ph-ph
because the battery store (72 MJ) is too small, not all the output
is available. It has been estimated that by resiting the device on a
neighbouring hill top, the annual electricity output would rise to
9.7 GJ/y. Using the method outlined earlier in this chapter the
following data would be expected:
Rotor diameter for Dunlite 2000 = 4.11m
V50 from Fig. 8.4 = 4.5 m/s
Gust ratio, Table 8.1 = 1.6
Gust ratio, modification to Vso = 1.0
Correction for height from Table 8.2 = 0.6761 # 0 - 1 7
(Above 10 m quoted in Fig. 8.4) = 1.05
V50 for the site = 4.7 m/s
Ewt (wind energy/y encountered by
blades) = 61 GJ/y
Power coefficient of the Dunlitel10! = 0.284
Combined efficiency of gears and
generator = 70%
Power between cut-in and rated speed = 80%
Electrical power output/y = 61 (0.284X0.7X0.8)
= 9.7 GJ/y
- The system of a 110 V DC output driving a 240 V AC heat-pump
compressor via batteries and an inverter is very inefficient.
— The heat pump iced up very readily in cold weather thus impairing
its COP.
8.8.2 Wind-powered cottage at NCAT

Attempts were made at the NCAT to provide most of the power for
a small cottage, using a 5 kW Swiss Elektro machine.!151
Fig. 8.19 indicates the scheme, in which most of the energy is used via
immersion heaters to provide domestic hot water and space heating via a
seven-day thermal storage unit and underfloor heating. The heaters are
switched automatically in series and parallel to provide a suitable load for
the output voltage which varies between 60 and 220 V.
288 Wind energy
15
The major problem encountered^ ! has been the poor output which
has been only 50% of the rated output at the rated wind speed.
References
[ 1 ] Whittle, G.E., Bossanyi, E.A., Maclean, C, Dunn, P.D., Lipman,
N.H. & Musgrove, P.T. (1980). 'A simulation model of an elec-
tricity generating system incorporating wind turbine plant'. Third
International Symposium on Wind Energy Systems, Copenhagen,
BHRA Fluid Engineering Cranfield, Bedford, UK.
and
Bossanyi, E.A. (1982). 'Wind and tidal integration into an elec-
tricity network'. Cranfield Wind Energy Workshop, Cranfield,
Bedford, UK.
[2] Article in Electrical Review, 13 March 1981, Vol. 208, no. 10, p.
19. No title or author known.
[3] Rayment, R. (1976-77). 'Wind energy in the UK'. Building
Services Engineer, 44, 63.
[4] Caton, P.G.F. (1976). 'Maps of hourly mean wind speed over the
United Kingdom'. Meteorological Office Climatological Memor-
andum, no. 79.
[5] Rayment, R., see Ref. [3].
[6] Golding, E.W. (1976). The Generation of Electricity by Wind
Power. Wiley, London.
[7] Anon. (1981). US Popular Science, January 1983.
[8] Watson, M. (1979). Wind Turbines for the Provision of Domestic
Heating. Dissertation of the Department of Architecture, Univer-
sity of Cambridge; and (1983), 'Use of wind energy for space and
water heating'. Rural Power Sources, International Solar Energy
Society, May 1983, London.
[ 9 ] Bennett Woodcroft (1851). Pneumatics of Hero of Alexandria.
Charles Whittingham, London.
[10] Marier, D. (1981). Wind Power for the Homeowner. Rodale Press,
Emmaus, Pennsylvania, US.
[11] Talbot, J. (1981). 'Rechargeable cells'. Electrical Review, 208, 18.
[12] Cox, C.H. (1981). 'Making the utility connection'. Solar Age,
December, p. 39.
[13] Information on Gemini synchronous inverters is available from
Wind works, Box 329, Route 3, Mukwonago, Wisconsin, US and
from Upton Ltd, East Bergholt, UK.
[14] Willoughby, J. & Todd, R. (1981). Ambient Energy Utilization at
the National Centre for Alternative Technology. National Centre
for Alternative Technology, Powys, UK.
[15] Todd, R. (1980). 'Energy and buildings at the Centre for Alterna-
tive Technology'. Technical Information Report 1, National Centre
for Alternative Technology, Powys, UK.
[16] A very efficient mult-pole rare earth generator has been designed
by Bolton and Ferraris at Imperial College, London, and built at
the Rutherford Laboratory, Chilton, UK. It is driven at low rpm.
Water-supply systems
9.1 Introduction
Water-supply and waste-disposal systems are related in two ways. Firstly,
in the UK, WC (toilet) flushing consumes about 36.5 1/p d or roughly one-
third of the average household water use.I11 Secondly, the point of final
waste disposal may also be a source of fresh water. About one-third of the
public water supply is from rivers, yet 60% of the local-authority sewage
discharges to these rivers do not comply with the recommended 20/30
standard.I21 (The Royal Commission on Sewage Disposal (1898-1915)
made recommendations which were later adopted on the quality of sewage
effluents to be discharged to water courses. The maximum concentration
of suspended solids was limited at 30 mg/1 and Biochemical Oxygen
Demand (BOD) was limited to 20 mg/1 for a dilution of 1.8 with river
water; the Biochemical Oxygen Demand is a measure of the oxygen con-
sumed during the oxidation (stabilization) of organic matter by a mixed
microbial population and under aerobic conditions.)
In this chapter we shall briefly deal with water-supply systems and in
the following one with waste disposal. Chapter 11 gives some water (and
energy) conservation measures.
Virtually all homes in the UK are connected to mains water supplies
and are likely to be so in the foreseeable future since planning permission
is extremely difficult to obtain for sites that are not serviced. This tends to
concentrate development and one of the interests of a group like the
Autarkic Housing Project was to study potential changes in land-use
patterns if housing was not constrained by availability of mains servicing
systems.
In other countries, for example the US, private water supplies are often
much more important. Even in the UK, occasions arise where alternative
supplies must be considered - one of the present authors has acted as a
consultant for a group setting up a holiday centre for inner-city youngsters
on an off-shore island without any form of servicing.
The use of wells, infiltration galleries, springs and streams is covered in
standard works on water supply and will not be discussed here. Instead we
will outline the water-supply system we proposed for the Cambridge
Autarkic House, which was based on rainwater collection, recycling and
heat recovery on waste water.
289
290 Water-supply systems
Fig. 9.1. The Cambridge Two-compartment 10 I
Autarkic House water-supply chlorinating unit
system. Collection
flush
200 I
To
WC
Kitchen Washing
and machine,
bathroom shower, Kitchen and
drinking hand bathroom hot
taps basin water spray taps
Mesh
filter
Pump for
pressurization
3
2 m
Storm tank
Dirty Mixing valve to
supply water at
Slow 45 °C
Clean
sand Heat
filter exchanger 0.5 m3
"jPre-heat Reverse
tank osmosis unit
4 m 3 Storage
25 m 3 Domestic
tank
hot water
storage tank
9.2 The Cambridge Autarkic House water-supply system

Fig. 9.1 illustrates the scheme. Water is collected from the roof and sloping
walls of the house in a storm tank before transfer to a storage tank via a
slow sand filter.
A previous calculation^3! had shown that in Cambridge (which is in a
comparatively dry area of the UK) a storage tank of 25 000 1 and a col-
lection area of 145 m2 would be required to guarantee a daily water
supply of 100 1. This is considerably less than typical present water-
consumption rates for dwellings and so extreme conservation measures
would be required. Moreover, neither the required volume nor collection
area was available and so recycling of water was necessary.
The slow sand filter used is a standard item which consists of a vertical
cylinder containing fine sand at the top, coarser sand in the middle and
gravel below. As the water trickles through the filter the turbidity should
be reduced from 50 ppm (maximum permissible) to 5 ppm, the bacteria
content reduced by 85-99% depending on the initial count and the
colouration should become clearer.l3l Sufficient bacteria may remain to
cause problems and so it is safer to further purify the water. With time,
turbidity in the filter builds up and backwashing is required to clean it.
In this system, further purification is accomplished by then pumping
the water to a 10 1 chlorinating unit and two 100 1 storage tanks in the
roof. The chlorinating unit uses calcium hypochlorite tablets to disinfect
The Cambridge Autarkic House water-supply system 291
Table 9.1. Roof areas and storage volumes for water production
(101.7 l/d) with a solar stilfi3]
Size of storage reservoir m3

5 10 15 25
Fraction of roof facing
south and acting as a still Roof areas m2
0 174 163 155 145

Vi 160 97 92 88
1 150 69 66 63
the water. Disinfection refers to the adequate destruction of waterborne

pathogenic microorganisms including bacteria, viruses and protozoa.^4!
Disinfection varies from sterilization in that the latter also destroys certain
bacterial spores. One storage tank supplies the kitchen and bathroom
drinking taps. The other serves the washing machine, shower and hand
basin, as well as the hot-water-system feed.
The hot-water-system feed passes through a heat exchanger where it is
warmed by discharged domestic hot water, then through the small preheat
and large thermal storage tanks which are both solar heated. It is then
pressurized and supplied to spray taps in the kitchen and bathrooms. After
giving up its heat in the heat exchanger the waste water goes to a reverse
osmosis unit. This unit eliminates suspended solids and removes almost all
organic matter, bacteria and spores, as well as dissolved salts from, for
example, soap. It is placed after the heat exchanger so that the entering
temperature is not excessive. Reverse osmosis units use electrical energy to
reverse the normal osmosis pattern and increase the concentration gradient
to produce clean water from the grey water fed into them.
Clean water from the unit is pumped to the non-drinking-water storage
tank in the attic and dirty water is pumped to a 200 1WC flush tank.
Other 'autonomous' water systems are well described by Rump,! 1!
Smithl5! andtheValesJ 6 !
Use of a solar still was also considered for the Autarkic HouseJ3 ! An
inexpensive but durable form of still might consist of a series of self-
supporting polyurethane foam trays in a roof, each glazed with a single
pane of glass at as low an angle as 6.5-10°. The joints must be well sealed,
the inside of the still painted black on the bottom with, for example, a
black-pigmented silicone paint, and the internal sides painted white. A
shallow layer of water (9-10 mm) is maintained in the tray and when hot
it evaporates, condensing on the cover glass and running down into a
collecting gutter.
Table 9.1 shows the results of a computer program developed to esti-
mate the roof area and storage volume needed to satisfy the previously
cited daily water demand of 101.7 1. The table is based on a calculated
daily output, from March to September for a poor year, of 1.6 1/m2 still
area. In all three cases the entire roof area is used to collect rain. Thus, the
table states, for example, that if a storage reservoir of 10 m3 is chosen, the
292 Water-supply systems
roof area required, if one-half of the roof is a solar still (and all of it
collects rain) is 97 m 2 . In the event, the required collector area was not
available and so recycling with the reverse osmosis unit was selected.
References
[ 1 ] Rump, M.E. (1978). Potential Water Economy Measures in Dwell-
ings: Their Feasibility and Economics. BRE CP 65/78. Garston,
Watford: BRE.
[2] Anon. (1971). 'Out of sight out of mind'. Report of the Working
Party on Disposal of Sludge in Liverpool Bay. London: HMSO.
[3] Littler, J.G.F. (1975). 'Solar still and water supply calculation'.
Autarkic Housing Project Working Paper 20, University of
Cambridge, Department of Architecture.
[4] Leckie, J., Masters, G., Whitehouse, H. & Young, L. (1975). Other
Homes and Garbage. San Francisco: Sierra Club.
[5] Smith, G.E. (1973). 'Economics of water collection and waste
recycling'. Autarkic Housing Project Working Paper 6. University
of Cambridge, Department of Architecture.
[6] Vale, R. & Vale, B. (1975). The Autonomous House. London:
Thames and Hudson.
10
Waste disposal and utilization
10.1 Introduction
Surprisingly enough, several years ago one of the highlights in California
Governor Jerry Brown's politics was the introduction of a law forbidding
the sale of any lavatory which flushes more than 14 1 of water. Waste dis-
posal does not often receive such publicity although the problems it entails
in both industrialized and non-industrialized countries, albeit for different
reasons in the two groups, are impressive.
Here we shall deal with wastes in greater detail than we did with water
systems but not because the central network is less extensive — in the UK
94% of all households are connected to mains sewers^1! (in the US, on the
other hand, the comparable figure is about 67%l21). Rather, it is because
wastes can be a source of on-site energy if methane digesters are used.
Although this tends to be less practical at the level of a single home,
groupings of houses and other building types such as schools should not
ignore the energy potential of the wastes they produce.
Mains servicing in the UK, as elsewhere, consists of a cistern-flush toilet
connected to a network of underground sewers which transport sewage
and domestic waste water to a treatment or disposal facility. While to
those of us who use such systems almost nothing seems more natural, it is
of interest to note that the concept of using storm sewers for human
wastes is only about 140 years old and the first integrated system only
came into full use about 1870 in London. The first sewers merely con-
veyed the wastes to bodies of water where they were discharged with often
disastrous consequences — sadly this practice continues today in many
areas. Present waste-treatment plants vary in size and operation but in the
UK almost all incorporate a primary treatment which removes most of the
suspended solid matter, such as paper, by means of sedimentation, settling
or septic tanks; the sludge which accumulates is removed periodically and
often treated — 40% is applied to agricultural land, 20% is tipped and 40%
is dumped at seaJ 3 !
Secondary treatment involves bacterial digestion of the organic wastes
and leaves a residue containing nitrogen and phosphorus. When this residue
is discharged into lakes or streams it can result in eutrophication, with
simpler organisms like algae growing too quickly to be eaten by the more
slowly growing organisms higher up the food chain. Algal blooms remove
the oxygen from the water, causing it to decay and become stagnant.
293
294 Waste disposal and utilization
Additional, but expensive, tertiary treatment can reduce such dangers. In
the UK, one-quarter of the rivers and one-half of the canals have been
described as of 'poor', 'doubtful' or 'grossly polluted' quality f4* 5J
(sewage is not, of course, the only source of pollution). There have also
been indications that the common practice of dumping sludge at sea may
have adverse effects on phytoplankton communities and even fisheries.I6J
Water supplies might also be considered to be at risk since, as was
mentioned in the previous chapter, one-third of the public water supply is
from rivers.
The disposal techniques presently available (throughout the world) are
conveniently divided into aerobic and anaerobic systems.
10.2 Aerobic systems

The common aerobic systems vary from the primitive, such as the pit
latrine, to the fairly sophisticated, such as the Clivus Multrum.
Kalbermatten et alX1^ provide a useful introduction to and bibliography
for many of the techniques. Here we will consider only the Clivus Multrum
since it would seem to be the aerobic method most likely to be socially
acceptable in European areas of high population density.
The Clivus Multruml7J (see Fig. 10.1) is an aerobic decomposition
chamber designed, primarily, for single-family houses, to dispose of human
excreta and organic kitchen wastes. The prototype was constructed in
1939. The device is divided by vertical partitions into three chambers
which are connected at the bottom. A specially designed toilet bowl is
situated above the excreta chamber and the garbage chute leads to the
refuse chamber. The addition of kitchen wastes or other organic material
permits the correct carbon-nitrogen ratio for composting (the device is in
fact a continuous composting toilet) to be achieved. The decomposing
material travels slowly down the container to the storage chamber.
Aerobic decomposition (by the microorganisms in the wastes) is achieved
by sucking air through the wastes from channels running along the con-
tainer. A draught is created by the presence of a chimney which is some-
times equipped with a fan. Since the inside pressure is lower than the out-
side no odour can escape and in a properly operating Multrum it is claimed
COMPOSTING V
by the manufacturers that there are no odour control problems although
this does not seem to be a claim universally accepted by users.l8l
Fig. 10.1. The Clivus In particularly cold climates electric heaters may be needed to control
Multrum.f9! (Courtesy of the temperature so that decomposition, which is temperature dependent,
Clivus Multrum USA, Inc.) proceeds. Care should also be given in the design of the installation to
ensure that cold air is not drawn into the house.
The great advantages of the Clivus are that no water is needed and the
volume of wastes is only 5—10% of the original volume, the rest being
given off principally as non-noxious carbon dioxide and water vapour. The
end product is a sludge amounting to approximately 10—40 1 of humus-
like soil per person per year; on average the sludge is removed every
15-24 months. A bacteriological analysis of the sludge revealed that no
fecal conform bacteria {E. coli) were present in the samples studied nor
Anaerobic systems 295
Inspection ports Fig. 10.2. Typical septic-
tank design.!2!
Inlet
Outlet
were any species of pathogenic Clostridium found.I10! Only about 0.5%

of the bacteria present in the sludge were pathogenic and these are known
to be common in soils. The sludge contains nitrogen, phosphorus and
potassium plus a number of minor plant nutrients and so will be of some
value as a fertilizer and soil conditioner.
A major difficulty with the Clivus is that it does not provide a means of
disposing of waste water from the sink, shower and clotheswasher, the so-
called 'greywater'. Conventionally, this is done by discharging the water to
a septic tank followed by a drainage field or seepage pit, but Clivus has
proposed a number of simpler solutions such as leaching chambers with no
pretreatment, based on the less harmful nature (lower BOD, virtual
absence of pathogens if toilet wastes are voided to the Multrum) of grey-
water compared to typical sewerage.111!
10.3 Anaerobic systems

Septic tanks are the most common anaerobic systems. Conventionally, the
tank is a small closed chamber (or group of chambers) sited just below
ground level, which receives both excreta and waste water (see Fig. 10.2).
Solids settle to the bottom where they are digested anaerobically.
Digestion is reasonably good, as evidenced by a reduction of approxi-
mately 50% in the BOD but enough sludge still accumulates so that the
tanks must be desludged every one to five yearsJ 7 ! The continuous
effluent from the tank is low in suspended solids but has a high BOD; it is
essentially untreated and is not suitable for direct discharge to a water
course. Normally, the effluent is treated aerobically on a percolating filter,
by irrigation on land (which may result in odour problems) or by subsur-
face irrigation (which may pollute water supplies).
Septic tanks do not normally make any special provision for the pro-
duction or collection of the methane gas which results from anaerobic
digestion. Methane digesters, on the other hand, are designed specifically
with this in mind and so offer a potential source of energy for cooking,
heating or lighting in the form of methane. They also deal with all of the
wastes, giving a single final product, sludge, which can be used as a soil
conditioner and fertilizer. The smell of digested sludge is a bit like that of
tar and not at all similar to that of untreated sewage or septic-tank effluent.
Historically, digesters evolved from septic tanks and digestion has
proven itself to be of value in waste disposal worldwide - which is not to
say that problems have not been encountered. In the UK, anaerobic
Fig. 10.3. Chinese methane Gas outlet pipe
digester.!17]
Water level
Ground level
digestion is by far the most widely used single process for treating sewage
sludge in municipal plants and its controlled use for this purpose dates
back to the last century.! 121 In these plants it normally suffices to raise
the temperature of the sludge by 20 °C at all times of the year (very
roughly, from 7-27 °C in winter and 15-35 °C in summer) for adequate
performance.!13! The methane produced is often used to drive the plant's
machinery.
In Europe, agricultural applications of digesters are exciting interest and
it has been reported that at least a dozen systems are already operating
successfully in Britain and France, with payback periods (at 1978 energy
prices) of around six years J 1 4 '
In Asia, numerous countries have developed small-scale methane
digesters, or bio-gas plants as they are often called.! 7 ' 15 ' 16J In the Chinese
province of Szechwan, it was estimated in 1975 that 400 000 digesters
were built or under construction and that 17 000 000 people used the gas
produced for cooking and lighting.!17! Fig. 10.3 shows a typical 10 m3
digester. As the gas is generated, it pushes down on the fermentation liquid
which, to relieve the pressure, rises and flows through a small hole near the
top of the sludge chamber into the area on top of the device. As the
pressure drops, or as gas is drawn off through the outlet pipe, the liquid
level recedes.
10.4 Methane digestion

Methane digestion is complex and only a brief introduction can be given
here; numerous works,! 8 ' 18s 19> 2 0 ' 2 1 1 from which much of the following
discussion is drawn, treat the subject in greater detail.
Fig. 10.4 shows the basic process, Fig. 10.5 summarizes the biochemical
reactions and Fig. 10.6 shows the sequential mechanisms schematically.
In the first stage of anaerobic waste treatment the 'acid-producing'
bacteria, a heterogeneous group which includes facultative anaerobes, pre-
dominate and aerobes degrade the organic fraction of the raw waste to
substrates which can be used by the methane-producing bacteria. The
generation, or doubling times, of members of the acid-producing bacteria
Methane digestion 297
Digester gas Fig. 10.4. The methane
digestion process.
Supernatant
s
Settled solids
Supplementary heat (if separation
occurs)
Proteins Amino acids A variety of acids, Fig. 10.5. Biochemistry of

Carbohydrates Sugars especially acetic anaerobic waste treatment.
Fats Fatty acids CO 2 ,H 2 J 20% (Courtesy of C. Freeman.)
NH?, H2S Supernatant
Resistant material +
Residue
First stage: 'acid-forming' bacteria Second stage:

methane-producing
bacteria
are believed to be very much shorter than for the methanogenic group -
minutes or hours as opposed to days. Thus, if a 'shock' loading occurs, the
acid producers are able to respond with an increase in population size
much more rapidly than the methane formers, with a consequent common
result being an increase in the volatile acid concentration and possible
inhibition.
The methane-producing bacteria are strictly anaerobic — even small
amounts of oxygen are harmful to them. There are several different types,
with each being characterized by its ability to convert a relatively small
number of organic compounds into methane. Thus, for complete digestion,
several different types are needed.
Biologically, successful digestion depends upon achieving and (for
digesters which are loaded continuously) maintaining a balance between
acid-forming and methane-producing bacteria. This can be accomplished
by supplying raw wastes of a suitable composition and by proper main-
tenance of the pH and temperature.
The composition of the feed should have a carbon to nitrogen ratio
(C/N) of about 30:1 for digestion to proceed at an optimum rate;l 20 l
Freeman & Pylel211 give a broader range of between 10 and 30 for 'good
operation'. The 30:1 ratio is based on the digester bacteria using up carbon
(for energy) 30 times faster than nitrogen (which in part is used for build-
ing cell structures). If there is too much carbon in the raw wastes, nitrogen
will be used up first, carbon will be left over and the digester will slow
down. If there is too much nitrogen, on the other hand, the carbon soon
becomes exhausted and fermentation stops. The remaining nitrogen is lost
as ammonia gas and the reduction in nitrogen content decreases the
fertility of the effluent sludge. Table 10.1 shows the approximate carbon
and nitrogen values of some wastes. (The figures must be used cautiously
since chemical analyses do not necessarily represent the carbon and
Table 10.1. Typical carbon-nitrogen ratios ofsome feeds^21)
N
Material (% dry weight) C-N ratio
Night soil 6 6-10

Cow manure x.7 18
Chicken manure 6.3 7.3
Horse manure 2.3 25
Hay, grass 4 12
Hay, alfalfa 2.8 17
Seaweed 1.9 79
Oat straw 1.1 48
Wheat straw 0.5 150
Bagasse 0.3 150
Sawdust 0.1 200-500
Table 10.2. Typical reported yields from anaerobic digesters^211
Gas yield Gas

(m 3/kg volatile composition
Material matter fed) (% methane)
Cow dung 0.1-0.3 65

Chicken manure 0.3 60
Pig manure 0.4-0.5 65-70
Farm wastes 0.3-0.4 60-70
Elephant grass 0.4-0.6 60
Chicken manure—paper pulp 0.4-0.5 60
Chicken manure-grass clippings 0.4 68
Sewage sludge 0.6 68
nitrogen available to the bacteria and also because the carbon and nitrogen
contents can vary greatly with age and growing conditions of the plant,
diet and age of the animal, etc.)
Gas production is often given as a function of the volatile solids in the
wastes (see Table 10.2). (Volatile solids is that portion of a solid sample
which volatilizes when the sample is heated to 600 °C; it is a measure of
the organic portion of the sample since the ash remaining is the inorganic
or mineral portion. Total solids is the weight proportion of solids in a
sample; 10% total solids means that 100 g of a sample would yield 10 g
dry solids on evaporation at around 100 °C.) Table 10.3 gives a detailed
breakdown of the gas composition.
Another aspect of the feed is particle size. Feeds such as straw, news-
paper and refuse may require grinding to, say, a diameter of 10-15 mm if
they are to be pumped or piped.
Feed concentration (total solids percentage) is important because if too
much digestible material is fed in, acid accumulation and inhibition can
result. Experience with operating digesters has shown that a feed slurry
containing 7-9% solids is optimum for digestion J81
The concentration will also affect the ease of mixing and pumping the
Methane digestion 299
Table 10.3. Representative composition of digester gas
produced from farm wastes 1191
Component Percentage volume

Methane (CH4) 54-70
Carbon dioxide (CO2) 27-45
Nitrogen (N2) 0.5-3
Hydrogen (H2) 1-10
Carbon monoxide (CO) 0.1
Oxygen (O2) 0.1
Hydrogen sulphide (H2S) trace
digester material. Mixing is important because it promotes the interaction

of the microorganisms with their substrate (the digester contents) and
because it disperses any local pockets where acids might have concentrated.
Large units often aim at continuous mixing but the machinery and energy
consumption implied makes this less feasible for a smaller plant. Infre-
quent mixing (say once or twice a day) has proved adequate for many of
the Indian village plants (where a typical problem seems to be controlling
the scum which can form at the surface).
The term pH refers to the amount of acid or base present in solution (it
is defined as the logarithm to the base ten of the reciprocal of the
hydronium ion concentration expressed in moles per litre). McCartyl18!
has stated that anaerobic treatment can proceed quite well with a pH
varying from about 6.6-7.6 with an optimum range of about 7.0—7.2;
Freeman & Pylel211 give a wider operating range of 6.0—8.0. When start-
ing off a digester, a neutral pH of 7.0 should be aimed for. The initial acid
phase of digestion may cause the pH to drop but care must be taken that it
does not fall below approximately 6.2 because at this point efficiency
drops off readily and the acidic conditions can become quite toxic to the
methane bacteria; this control can be achieved either by decreasing the
waste feed to the digester or by adding neutralizing materials such as lime,
or by both. As digestion proceeds the pH rises (see Fig. 10.6) until it
Sequential mechanism
reaches a point when it is well buffered and the digester contents can of digestion
stabilize themselves even when large amounts of acid or alkali are added.
At this point, raw feed may be added periodically and a constant pro-
duction of gas and sludge achieved.
Temperature affects both the rate and course of digestion; the reactions
themselves are slightly exothermic overall. Two optimum temperature
levels for anaerobic treatment have been reported. The first is in the meso-
philic range from 29 to 38 °C and the second in the thermophilic range
from 49 to 57 °C. Although treatment proceeds much more rapidly in the
thermophilic region (with a consequently shorter detention time, that is,
the average period of retention of sludge in the digester), the additional Time
energy requirement needed to maintain the temperature tends to limit
Fig. 10.6. Sequential mech-
such applications to municipal digesters. At below 15 °C digestion pro- anism of digestion.
ceeds very slowly although the process can function at a temperature as (Courtesy of C. Freeman.)
Fig. 10.7. Effect of tempera-
ture on gas production.!221 0.5
^ 0.4
"8
T3
•D
0.3
0.2
A 100% raw sludge

0.1 • 97% refuse, 3% raw sludge
O 100% raw sludge
Detention time = 30 days
30 40 50 60
Temperature (°C)
low as 5 °C.l211 Maintenance of a constant temperature is important for

optimum gas production. Fig. 10.7 shows the effect of temperature on gas
production using selected feeds.
The problem that arises at this point is that almost all of what is known
scientifically about methane digestion is due to studies of large-scale pro-
cesses. Domestic-scale digesters with mixed feeds supplied at irregular
intervals have received very little attention and indeed the number of such
devices operating in western countries is probably very small indeed.
Consequently, we can only supply estimates based on our work and the
work of C. Freeman for the Autarkic Housing Project.
Fig. 10.8 shows a design (developed in conjunction with Farm Gas Ltd)
for a domestic-scale digester. The annual feed to such a digester in the
context of a four-person home and garden might consist of human wastes,
kitchen wastes, paper and garden wastes.
A digester which has only human wastes available in limited amounts
is not practical. The average weights of faeces and urine produced by a
mixed population are 0.086 and 1.055 kg/(d p)l 241 (wet solids basis) and
it was estimated that this would give 85 kg/y dry weight and might pro-
duce 40 m3 of digester gas per year.! 24 ' 25J If, as a conservative estimate,
we assume that only 50% of this is methane and using a calorific value of
37.4 MJ/m3 for pure methane, we have about 0.8 GJ. Since four-person
families using conventional cookers (cooking being the most likely house-
hold application for methane) annually use between 4.3 and 6.8 GJ (see
also Chapter 11), it is evident that other sources of waste are required if
methane is to contribute significantly to the energy demand of a house.
It was estimated that kitchen wastes and paper might contribute 310 kg
dry weight and supply 2.2 GJ/yJ 25 ' 26» 271 Vegetable garden wastes from a
References 301
Fig. 10.8. Schematic of an
anaerobic digester for the
Autarkic House.1231
Wast inlet
0.2 m3 mixing tank

Gas pipe
Flue
Chopper/mixer
Gas gap Digested waste
Digester inlet pipe
Digester outlet pipe
Insulation
1.8 m3 digester
Heat exchanger
Gas diffuser
400 m2 area were calculated to be about 90 kg dry weight and might pro-
vide 0.7 GJ/y.I 25 ' 2 7 ' 28> 291 Thus the total energy supply would be
approximately 3.7 GJ/y and would be capable of contributing significantly
to the energy requirement for cooking. (From this, one would have to
subtract any energy needed to maintain the digester at a working tempera-
ture but this will depend on such factors as design, location of the digester
in the house and climate.)
It must be emphasized, however, that work on domestic-scale digesters
is only in its infancy - both theory and practice need development. To
close on a note of caution: at a concentration of 5-14%, methane is
flammable and so poses a serious safety problem in the home. (To keep
this in perspective, though, North Sea gas currently supplying many UK
homes is 95% methane.) Consequently, burying a gas store in the garden
may be a reasonable safety precaution. Similarly, the danger of explosive
mixtures of gas and air also means that digestion tanks and the gas-
collection systems should always be kept under a positive methane
pressure.
References
[ 1 ] Anon. (1970). Taken for granted'. Report of the Working Party on
Sewage Disposal, Ministry of Housing and Local Government.
London: HMSO.
[2] Barley, J. & Wallman, H. (1971). 'A survey of household waste
treatment systems'. Journal Water Pollution Control Federation,
December, pp. 2349-60.
[3] Coker, E. & Davis, R. (1978). 'Sewage sludge - waste or agricul-
tural asset?' New Scientist, 78 (1101), 298-300.
[4] Anon. (1970). Report of River Pollution Survey of England and
Wales. Department of the Environment, London: HMSO.
[5] Anon. (1972). Report of River Pollution Survey of England and
Wales. Department of the Environment, London: HMSO.
[6] Anon. (1971). 'Out of sight out of mind'. Report of the Working
Party on Disposal of Sludge in Liverpool Bay. London: HMSO.
[7] Kalbermatten, J.M., De Anne, S.J. & Gunnerson, C.G. (1980).
Appropriate Technology for Water Supply and Sanitation.
Washington, DC: The World Bank, p. 18.
[8] Leckie, J., Masters, G., Whitehouse, H. & Young, L. (1975). Other
Homes and Garbage. San Francisco: Sierra Club, p. 215.
[9] Anon. (Undated). Clivus (Multrum). Tyreso, Sweden: A. B. Clivus.
[10] Anon. (1977). Bacteriological Studies of Clivus Multrum End-
Products. Cambridge, Mass: Clivus Multrum USA, Inc.
[11] Fagel, M. & Lindstrom, C. (1976). The Treatment of Household
Water in Homes Equipped with the Clivus Multrum Organic Waste
Treatment System. Cambridge, Mass: Clivus Multrum USA, Inc.
[12] Anon. (1974). 'Anaerobic treatment processes and methane pro-
duction'. Notes on Water Pollution No. 64. Department of the
Environment. London: HMSO.
[13] Escritt, L.B. (1971). Sewers and Sewage Works. London: George
Allen and Unwin.
[14] Martin, D. (ed.) (1980). Specification 80. London: Architectural
Press.
[15] Prasad, C.R., Krishna Prasad, K. & Reddy, A.K.N. (undated). Bio-
Gas Plants - Prospects, Problems and Tasks. Bangalore: Indian
Institute of Science; New Delhi: Management Development
Institute.
[16] Subramanian, S.K. (1977). Bio-Gas Systems in Asia. New Delhi:
Management Development Institute.
[17] Smil, V. (1977). 'Energy solution in China'. Environment, 19 (7),
27-31.
[18] McCarty, P.L. (1964). 'Anaerobic waste treatment fundamentals'.
Public Works, September, pp. 107-12; October, pp. 123-6;
November, pp. 91—4; December, pp. 95—9.
[19] Anon. (1973). Methane Digesters for Fuel Gas and Fertilizer.
Woods Hole, Massachusetts: New Alchemy Institute.
[20] Stafford, D.A. (1974). 'Methane production from waste'. Effluent
and Water Treatment Journal, February, pp. 73—9.
[21 ] Freeman, C. & Pyle, L. (1977). Methane Generation by Anaerobic
Fermentation - an Annotated Bibliography. London: Intermediate
Technology.
[22] Diaz, L.F., Kurz, F. & Trezek, GJ. (undated). Methane Gas Pro-
duction as Part of a Refuse Recycling System. University of
California, Berkeley. Department of Mechanical Engineering.
[23] Cheshire, M.J., Farm Gas Ltd. (1977). Private communication.
[24] Snell, RJ. (1943). 'Anaerobic digestion III. Anaerobic digestion of
undiluted human excreta'. Sewage Works Journal, 15 (4), 679-701.
[25] Imhoff, K.M., Muller, W.J. & Thistlethwaite, D.K.B. (1972). Dis-
posal of Sewage and Other Water Borne Wastes. London: Butter-
worths.
[26] Skitt, J. (1972). Disposal of Refuse and Other Wastes. London:
Charles Knight.
References 303
[27] Klein, S.A. (1972). 'Anaerobic digestion of solid wastes'. Compost
Science, January—February, pp. 6—11.
[28] Thomas, R.B. (1980). 'Biological aspects of an Autarkic House'.
Ph.D. Thesis, University of Cambridge.
[29] Rosenberg, G. (1952). 'Methane production from farm wastes as a
source of tractor fuel'. Agriculture, 58, 487—94.
11
Domestic-energy saving
11.1 Introduction
Results from the recent Better Insulated Housing Programmed are sum-
marized in Table 11.1 which shows the fractions of energy consumed on-
site for various purposes.
The experimental houses at Bebington (Fig. 4.48), which are electrically
heated, indicate that only about half the electricity supplied annually is
used directly for space heating.
Much of the energy provided to the lights, fridge and so on, contributes
usefully during the heating season to warming the space. Capper I21
suggests the fractions shown in Table 11.2.
One might wish to argue a little with the low figures (Leachl3! assumes
0.8 in all cases and Siviour's conclusions are shown in Table 3.1) but the
fact remains that a substantial fraction of the heating in houses arises from
electricity (or gas) used in appliances. Since this energy is wholly or partly
electrical (generated with an efficiency of roughly 30%), the cost to the
country in terms of primary fuel, or to the consumer in terms of cost,
constitutes a large part of the annual fuel bill for each house. Thus, one
would expect to find regulations concerning the efficiency of appliances.
There are almost none in the UK, although in the US some goods must be
marked with an efficiency indicator.
In the Bo'ness studyj 1 ! nearly all households have a clothes washer,
one-third a tumble drier, all have a fridge and 40% have a freezer. Most
have a colour television. Two-thirds of the households have electric fires,
one-third have calor gas and one-fifth have paraffin heaters as subsidiary
heating devices.
11.2 Domestic hot water

Typically, of the water used in houses, 25% is fed directly from the mains
(that is, cold water for drinking, cooking, etc.), 40% is cold water from the
storage tank and 35% is hot water. A reduction in demand is highly desir-
able nationally, for the water industry is extremely capital intensive (cost-
ing, even in 1967, £50000 to create one job compared with a national
average of £4000). Table 11.3 indicates how warm water is consumed.
The energy required to produce 80 m 3 of water at 45 °C from a mains
304
Domestic hot water 305
Table 11.1. On-site energy consumption
(%)
Space heating 50
Appliances 20
Immersion heater (DHW) 20
Cooking 7
Lighting 3
Table 112. Fraction of energy

which contributes to space
heating
Lighting 1.0
Water heating 0.5
Cooking 0.4
Appliances 0.4
Table 11.3. Use of warm water in households
a b
Use (1) (m 3 /y)
Washing hands 5 14 4
Washing hair 20 4 4
Washing face and hands 10 14 7
Bath 110 8 44
Shower 40
Dishes 15 14 11
Clothes washer 50-300 2 10
Total 80
*Ref. [4].
b
Estimated number of uses for a family of four per week.
c
Projected annual consumption for a family of four.
temperature of 10 °C is about 12 GJ, which corresponds well with the

mean use of the immersion heater in the Bo'ness houses (12 GJ/y)^ 1 !
Clearly, the principal activities for which economies may be made are
bathing and washing clothes.
Front-loading washing machines and cold-water washing powders are
mentioned below. Showers use less than half as much water as baths. Part
of the high load ascribed to washing hands arises from the dead-legs
between the basin and the tank (which typically hold 1-21). Insulating
such pipes only marginally affects the problem, since the water is likely to
cool off between each draw off. Smaller pipes help but may lead to block-
age with deposits. Careful design reduces the pipe run, of course. The care
lavished on water economy in the Minimum Energy Dwelling in Cali-
306 Domestic-energy saving
Table 11.4. Major domestic electrical consumers other than cookers and
immersion heaters
Typical rating UK annual US annual

in UK consumption consumption
Appliance (W) (GJ/y) (GJ/y)*
Lighting - 1.8* 7.2
0.9<*
Fridge 100 2.1a>d 2.7
Freezer 300 1.8* 4.3
3.6*
Tumble drier 2000 1.4* 3.6
Clothes washer 2500 0.7* 0.2
Dish washer 3000 3.1* 1.2
1.3*
Kettle 3000 0.9*
Colour TV 300 0.9* 1.8
Central-heating pump 80 0.5* —

Overblanket 100 0.4* -
Toaster 1200 0.2* -
* Data calculated from NEDDO Report on Energy HMSO1974 and the
Better Insulated Housing Programme.!^
b
New York Public Service Commission Report on Appliance Efficiency,
1973, published by the New York State Energy Commission.
* Estimated.
<* Leach.*3 J
fornia,l5l must represent the ultimate in energy reduction. This house

features the 'Ultraflow' system in which push-button control at each
point of use activates a solenoid valve and mixer at the hot-water tank,
allowing delivery through one flexible pipe of small diameter. There are
savings on installation and energy costs.
11.3 Electrical power

It commonly seems to be thought that a large array of electrical gadgets
produces a great consumption of electricity but, since the power of such
devices and the time for which they are switched on is low, the energy
used may be small.
The items which consume by far the largest amounts of energy are
shown in Table 11.4.
Manufacturers are responding only very slowly to the implications of
these figures. The most obvious economies which are available now con-
cern colour televisions (whose power rating has come down rapidly in the
last few years with solid-state techniques) and lights (where fluorescent
Cooking 307
tubes are now commonplace). There are also proximity switches (at
present for offices) which detect people and switch off lights in empty
rooms. It seems quite possible that the consumption for lights and TV
could be halved. The same is not so true for the other appliances.
Energy use in clothes washers is considerably lower for front-loaders
(which use less water), and great economies and added convenience can be
achieved using ultra high-speed spin-drying cycles (of at least 1200 rpm).
Clearly, the water spun out of clothes does not need to be removed by
heating in a tumble drier, or by evaporation into the house from clothes
strewn about radiators, thus reducing the possibilities of condensation.
The weight of clothes to be carted out to a clothes line is also ameliorated!
Economies can be made by insisting that the dish washer or clothes
washer accepts hot water from the hot storage tank without mixing in cold
water unless necessary (that is, avoiding unnecessary use of the water
heater in the appliance). Clearly, too, cold-water washing cycles and
powders, which are common in the US, should reach the UK ultimately.
The only other obvious areas to attack concern food preservation. Fortu-
nately, 'well-aired pantries' (adding considerably to the heat load of a
house in winter) are less common now. But fridges and freezers all too
frequently have skin condensing coils, which seems insane unless the coils
are mightily well insulated from the cooled space. The insulation, in prac-
tice, is poor, door seals quickly wear out and magnetic strips cease to close,
all adding unnecessarily to the electrical load.
In the absence of a rating system, consumers should perhaps consult the
Which Guides which have dealt with appliance efficiencies. Obviously,
front-opening freezers should be avoided, and chest freezers kept full.
Market penetration by freezers seems likely to be about 40% by the year
2000l3 l and thus, both in domestic and national terms, the energy con-
sumed is considerable. Even chest freezers are poorly insulated and, pro-
vided they do not have skin coils and if one can seal on added insulation in
such a way that interstitial condensation does not occur, energy may be
saved.
11.4 Cooking
Table 11.5 indicates the consumption by an electric cooker and hob. Gas
consumption would be higher by at least a factor of 1.3.
Obviously, microwave ovens can save energy but the sacrifice in quality
may be too great. Grilling relies on a fierce heat and free movement of air
to carry away water vapour. It is difficult to enclose a grill. At present,
ovens must merely be insulated to ensure that the user does not burn him-
self by touching the casing but, clearly, more could be done to reduce heat
loss. Gas ovens, in which the burning is inside the oven, require ventilation
whose cooling effect must be counteracted by using more gas. Induction
heating certainly directs heat very positively into the pans on a hob but
although induction hobs have been made, they remain very expensive.
Leach suggests that cooking efficiency will rise considerably (specific
308 Domestic-energy saving
Table 11.5. Energy consumption for cooking per household
UK annual US annual
Rated power consumption consumption
Function (W) (GJ/y) (GJ/y)
Oven 3000 1
Grill 2500 V 43a-5Ab 5.9C
Hob 1000-2000/ring J
a
Bo'ness Better Insulated Housing Study, Ref. [ 1 ] .
6
NEDDO, Report on Energy, HMSO, 1974.
c
New York Environment Research Centre Energy Index, 197r4.
fuel index: 1975 = 1, 2010 = 0.5). He also suggests that the amount of
cooking will fall (1975 index = 1, 2010 index = 0.86).
In 1974, about 55% of dwellings used gas for cooking, with the pro-
duction of a good deal of water vapour (about 1 kg per 0.04 GJ). The
vapour is in addition to that arising from the cooking itself (about
0.05 kg/h).l6l Since much of this water vapour, plus that from showers,
etc. (also about 0.05 kg/h) and people, must be removed by ventilation, it
is important to reduce the amount of water vapour created in order to
economize on the ventilation heat loss.
11.5 Space heating
Comfort levels are discussed in Chapter 3. Very often, people saddled with
electric heating in poorly insulated houses use portable devices such as
paraffin or liquified gas heaters. Inevitably, there are then complaints
about condensation, since each kilogramme of fuel burned produces
slightly more than 1 kg of water vapour. To remove the vapour, windows
must be opened! (Dew-point calculations are addressed in Appendix 3.)
In highly insulated houses where air-change rates are likely to be low, it
is highly undesirable to use such portable heaters. It may be worth point-
ing out that an electric overblanket rated at roughly 100 W consumes very
little energy, and retains comfort levels in bed, whilst allowing room tem-
peratures to sink to, say, 16 °C. Naturally, the room temperature ought to
be kept above the dew point at room surfaces, with the possible exception
of the window.
11.6 Transportation
The growth of energy use in transport stems mainly from the increasing
use of cars in the UK. (Currently the ratios are roughly as shown in Table
11.6.) The average car consumes about 45 GJ of fuel per year. There are
about 15 million cars compared with 20 million households. Thus the total
energy consumption for all cars is about 7 X 108 GJ, whilst that for space
heating in all households is about 10 X 108 GJ (total energy use in houses
being about 15 X 108 GJ.
References 309
Table 11.6. Ratios of passenger km (after
LeachW)
Bicycle 1
Motorcycle 2
Rail passenger 100
Bus and coach passenger 150
Car passenger 1000
Clearly, the incentive to save fuel for a household is as strong in trans-

port as in space heating and the methods are well known and excellently
summarized by LeachJ3J They obviously include selecting a car with a
good km/1 rating for the type of travel the household commonly enjoys,
the use of steel-core radial tyres at the correct pressure, proper main-
tenance and gentle driving. Simple precautions such as these, following the
choice of a car, may reduce consumption by anything up to 50%!
References
[ 1 ] Capper, G. (1981). Interim Report Number Three, Better Insulated
Housing Programme. Building Department, Heriot Watt University,
Edinburgh, UK.
[2] Capper, G., ibid.
[3] Leach, G. (1979). A Low Energy Strategy for the UK. Inter-
national Institute for Environment and Development, Percy Street,
London.
[4] Littler, J.G.F. (1975). 'Solar still and water supply computation'.
Working Paper 20, Martin Centre for Architectural and Urban
Studies, Cambridge, UK.
[ 5 ] Burt Hill Associates (1977). Minimum Energy Dwelling. Burt Hill
Associates, Butler, Pennsylvania, US.
[6] Burberry, P. (1975). Environment and Services. Mitchell's Building
Construction. London, Batsford.
12
Housing case studies
12.1 Introduction
This chapter takes the reader through some of the design processes which
occurred during the planning of several low-energy houses. The Peter-
borough Houses are complex, having large air-heating solar collectors, a
sophisticated ductwork system and microprocessor control. The Newnham
Houses are more conventionally heated, but highly insulated. These new
dwellings are discussed in Section 12.2, and in 12.3 reference is made to
the rehabilitation of existing houses, including one with a passive roof-
space solar collector.
12.2 New houses - three solar air-heated houses in Peterborough

The authors won a competition at the 1979 National Energy Show, for a
terraced low-energy house, designed with Lucy Krall, shown in Fig. 12.1.
Peterborough Development Corporation became interested after seeing the
model on television and suggested that the same sort of principles could be
applied to one of their houses.
In the end, a terrace of three was built, as shown in Fig. 12.2. Because
a lot of domestic hot water was likely to be produced by large collectors,
we felt that six-person houses should be built. Larger units also yielded a
greater roof area than would smaller terraced houses. In fact, in order to
provide a reasonably large area (32 m2 gross), collectors were also placed
on the first-storey walls.
In retrospect, this idea of dual slope collectors was probably influenced
by the Autarkic House (Fig. 12.3) in which the architectural design forced
the provision of collectors on three azimuths and two tilts. The whole of
the southerly facade in the Peterborough Houses is glazed, since the
ground floor consists of a conservatory opening into the living room. This
idea of wholly glazing a facade was turned into an extremely elegant sys-
tem by the Energy Design Group, who won an award for their rehabili-
tation scheme at the First European Passive Solar Design Competition. The
solar skin is illustrated in Fig. 12.4.
The severe problems faced by the architects at Peterborough were to
incorporate the proposed heating system into a workable house. This is
rather the converse of the normal procedure! It led to great compromises
on all sides.
310
Three solar air-heated houses in Peterborough 311
Fig. 12.1. The initial design
in 1979 for the Peterborough
houses.
Fig. 12.2. The homes at

Peterborough in 1983.
Fig. 12.5 illustrates the site plan, and indicates a major problem of
siting. Overshadowing was avoided by placing the three houses on the
perimeter of the development but at the cost of an orientation 40° east
of south.
It remains to be seen how drastic this poor orientation will be. Of
course, with the shallow roof pitch (30°) and the prevalence of diffuse
radiation in the UK, the harm may, comparatively, be less than in a region
with more beam radiation.
Inside the house, the architects solved the problem of providing lobbies
to exterior doors, as shown in Fig. 12.6. The 4 m3 heat store was placed
on a reinforced slab, forming part of a core in the centre of the house, con-
taining all the dampers, etc. needed to direct the air stream past, in or out,
312 Housing case studies
Fig. 12.3. The Cambridge
Autarkic House (architect:
Alex Pike; technical design
team: D. Forrest, J. Littler
and R. Thomas, 1977).
of the store. The slab is insulated throughout and thickened under the
store as shown in Fig. 12.7. The volume was chosen to provide only one
day's storage - larger stores would scarcely ever be completely filled dur-
ing the heating season.
Building the store itself constituted a problem, since the house is timber
framed and it was undesirable to bring in a wet trade just to build a box
for a pile of loose rocks. Thus, lateral pressure was avoided by using
stacked bricks. Calculation quickly shows that pressure drops are likely to
be small at sensible flow rates across stores of this size (1 m X 2 m X 2 m)
using bricks or even 50 mm stones. Nor is the rate of heat transfer a prob-
lem (see Chapter 4). The store will inevitably be stratified. The difficulty
arises in ensuring an even air flow through the pile.
On the skin of the house the architects had to incorporate windows
between collector panels to the fully glazed first floor, and to cope with
extensive thermal movement of the collector panels. The section in Fig.
12.8 shows that panels stretch from the ridge to eaves (about 5 m) and
under conditions of control failure, with no air flowing through the panels,
they will reach at least 150 °C. Even under low flow conditions, tempera-
tures above 130 °C have been observed (June 1983). Fig. 12.8 indicates
the solution adopted to cope with extensions up to 20 mm.
The architects encountered major problems in incorporating the duct-
work (which is very large, for example bottom feeder 300 mm X 250 mm
X 6000 mm), the collector panels and glazing, and the heat store. Com-
promise stretched only so far on the structure - in fact, to double-glazed
windows, slab insulation and the normal standard of wall insulation. This
normal standard at Peterborough (nominally 75 mm of glass wool) is fairly
high, corresponding to a wall {/-value of about 0.4 W/(m2 K) but, as an
advisor, one would have sought a value of, say, 0.3 W/(m2 K). However,
Fig. 12.4. Solar skin for a efforts were made to incorporate an internal vapour barrier to reduce
Georgian terrace in Bath,
by Peter Clegg and Richard infiltration; but it was clear during construction that many procedures had
Feilden (1980). to be rethought and design changes made in order that the barrier be easily
Three solar air-heated houses in Peterborough 313
Fig. 12.5. Site plan of the
Peterborough Houses.
(Figures 12.5-9 by Peter-
borough Development
Corporation, Chief Architect
K. Maplestone. Design and
drawings by M. Rodgers,
R. French and A. Newman.)
Fig. 12.6. Ground floor of

the Peterborough Houses.
Fig. 12.7. Slab thickened

and insulated under the heat
store.
Fig. 12.8. (a) Uniform size
of collector panels (2.4 m)
and their linkage by flexible
ducts, (b) Flexible link to
bottom feeder duct at the
level of the conservatory
roof.
installed and remain without breaks. The bookl11 concerning such design,
to be produced for the International Energy Agency, should help in this
regard.
The major problem for the Peterborough engineers was coping with a
totally alien method of heating or preheating the air. The collectors may
overheat and require fail-safe venting. The hot water is heated by a transfer
coil in the air stream (see Fig. 5.2) but this sort of transfer from air to
water, with a low pressure drop in the air stream and a strange set of con-
straints, often takes one off the end of a manufacturer's performance
curve. The three-way dampers (Fig. 5.2) had to be specially made,
although complete units with a fan and both dampers are available off-the-
shelf in the US. The desirable rate of air flow via the collectors (for high
collection efficiency) is about three times that needed for heating the
house. Thus, a two-fan system was adopted.
Finally, the design called for up to ten modes of operation (see Table
5.1). Such control complexity requires a microprocessor and it remains to
be seen how successfully the control scheme works. One of the main
advantages of a solar air-heating system is that the fluid heated in the col-
lectors subsequently heats the house, without the intervention of heat
exchangers. Since the solar heated air may require further warming by a
gas coil, the system is necessarily of the 'parallel' type. This is by dis-
tinction from a system in which solar heated water is further heated by gas
and then passed into radiators, in a 'series' heating arrangement. From the
point of view of collector efficiency, the latter is highly undesirable.
A second advantage which the designers saw for the system was the
potential use of very low-grade heat. Cold makeup air to the house can be
drawn in via the collectors (Mode 6B, Table 5.1), and even on fairly dull
days can be warmed before passing over the auxiliary heater. In this way
the house is pressurized against infiltration by cold air. It remains to be
Energy-efficient houses in Newnham, Cambridge 315
Fig. 12.9. Section through
H h single-glazed collector.
Standard profiled steel
sheets.
seen whether this mode is useful. In practice, solar air systems tend to leak
badly. An example of a Peterborough panel was pressure tested^2! and
found to be almost gas tight at 100 Pa. The original design called for an
optimum width of 30 mm between the two layers of metal (the front
being painted black and exposed to the sun, and the back being sealed to
insulation) and a fluted pattern (Fig. 12.9). The fluted shape achieved a
higher surface area for heat transfer but proved too expensive for imple-
mentation.
Computer simulation using 'f-chart' (see Chapter 5) suggests that the
active system should provide 30—40% of the space and hot-water heating
for the year. Monitoring is underway to confirm or refute this calculation.
However, a further boost to the solar fraction will arise from the single-
glazed conservatory linked to the living room by two pairs of double-
glazed doors. The doors will be opened during the heating season when the
conservatory temperature exceeds that in the living room. Fans may sub-
sequently be fitted if necessary. The contribution from the conservatory is
likely to be about 25% of the space heating load.
12.3 New houses — energy-efficient houses in Newnham, Cambridge
For two houses in Newnham, Cambridge, one problem the architects faced
was how to take advantage of passive solar gain on a restricted site where
the surrounding homes are orientated along a street running approximately
north-east—south-west. The solution adopted was to abandon the tra-
ditional rectangular shape by cutting off the south-facing corner of the
rectangle, thus giving a south-facing wall (see Fig. 6.33). The approach
introduced some complexity in space planning and construction but this
was more than outweighed by the spacious and visually attractive interior
achieved. Glazing was concentrated on the south, south-east and south-
west facades (see Fig. 12.10) and kept to a minimum elsewhere. In
addition to the compact form and the south-facing facade for passive solar
gain, energy conservation was achieved in a number of other ways includ-
ing high insulation levels and careful selection of the heating system (see
Fig. 12.10. View from the
south-east. (Architects:
Lyster, Grillet and Harding.)
Fig. 12.11. Fixing of

expanded polystyrene
insulation to the inner wall.
Chapter 6). Provision was also made for incorporating future developments
in energy conservation and the use of ambient energy.
Perhaps the most interesting aspect of the houses was that this was all
achieved by constant interaction between clients and architects and by
one of the present authors acting as energy consultant. For example, one
suggestion which was adopted was to alter the roof space to reduce the
heated volume and allow for the installation of an air-to-air heat-recovery
device at a later date. On the other hand, closing off the staircase to separ-
ate the house into zones of varying temperature was not accepted for
reasons of aesthetics and convenience. But what was important was that
the occupiers of the building had participated in its thermal design and so
were much more likely to use it properly and economically.
The wall construction from inside out was 12 mm plasterboard,
100 mm lightweight concrete block, 100 mm cavity with 50 mm of
expanded polystyrene on the inner half, and 100 mm brick, giving a
{/-value of 0.39 W/(m2 K). Fig. 12.11 shows the expanded polystyrene
boards being fixed to the concrete block with wall ties.
Renovation of a farmhouse to include a sun space 317
Roof construction was basically tiles on battens above 150 mm of glass
fibre insulation with 12 mm plasterboard, giving a {/-value of
0.22 W/(m2 K). The floor consisted of 22 mm floorboards over 80 mm of
glass fibre insulation suspended from the floor joists, giving a fZ-value of
0.27 W/(m2 K). For reasons of reduced heat loss and increased thermal
comfort it was decided to use multiple glazing. Triple glazing (with two
vacuum-sealed gaps) was selected (see Fig. 12.10) rather than double
glazing because of the better thermal performance and also because the
higher quality of the windows (wood, finish and ironmongery) was judged
to outweigh the disadvantage of the somewhat higher cost.
It was anticipated that high insulation levels, multiple glazing, provision
of a lobby and attention to quality of workmanship on-site would produce
a well-sealed house with considerably reduced ventilation heat losses. To
guard against the possibility of condensation on internal surfaces, extract
fans were fitted in the utility room, kitchen and bathrooms. The estimated
on-site energy consumption for space heating is about 300 and
400 MJ/(m2 y) for both space heating and domestic hot water, which
compares favourably with an extremely well-insulated house of the near
future (see Chapter 1).
In addition to allowing for future solar preheating of the inlet air and
incorporation of a small air-to-air heat exchanger, as mentioned in Chapter
6, the following measures were planned for:
(1) increased wall insulation by filling the cavity;
(2) use of approximately 2 m3 in the utility room as a thermal store;
(3) incorporation of a unit to recover heat from waste water by group-
ing services in one area of the house.
12.4 Rehabilitated houses — renovation of a farmhouse to include a sun

space (contributed by Peter Clegg)
In 1978, work began on the subdivision into four dwellings and renovation
of a farmhouse and outbuildings in the lower Cotswolds. One wing of the
main farmhouse faced due south and offered opportunities for the use of
passive solar design (see Fig. 12.12). About one-third of the house was
completely rebuilt using reclaimed natural materials and following the
pattern of vernacular details. This was done primarily as an exercise using
traditional construction but built to a high insulation standard. New roofs
contain 150 mm glass fibre, and the walls consist of 200 mm of rubble
stonework built up against 50 mm of 'Styrofoam' with an inner skin of
dense concrete blocks which also contain foam insulation. Internal thermal
mass of the building was therefore kept relatively high.
One of the other precepts of vernacular design was that window open-
ings were kept relatively small, and roofs had low eaves, with dormers to
illuminate the upper floors. Taking into account these restrictions, and the
owner's interest in plants, the obvious way to make a better solar contri-
bution to the building was to construct a conservatory.
A bronze anodized aluminium glazing system was used, in order to
match the colouring of the traditional material, to reduce maintenance
Fig. 12.12. The converted
farmhouse (Feilden Clegg
Design, Bath).
costs and for the provision of a light glazing bar. The system, manu-
factured under the name 'Florada' is of US origin and uses a curved bar
with polycarbonate glazing under the eaves to simplify construction and
reduce the structure. Using typical sections and lengths of bar from differ-
ent free-standing greenhouse models, a 13 m long structure was built; some
of it forms an entrance porch to the house, another section is used as a
greenhouse proper and a third section is intended as additional living area.
The glazing bars are tilted over at an angle, primarily for aesthetic
reasons, but also to admit more direct sunlight in winter. The foot of the
glazing bar rests on the lip of a glazed clay drainage channel to collect
rainwater, on top of a 750 mm high insulated retaining wall (the building
is partially underground).
It was intended that excess heat should be vented into the house via
windows, doors and clay pipes which form openings through the wall. A
return-air system was envisaged, and a duct was constructed to collect air
from the ridge of the house and discharge it under a suspended timber
floor over part of the ground floor of the building. The mechanical part of
this system has never been fully completed and is unlikely to be, in view of
the fact that dissipation of heat into the house seems to work satisfactorily
by simply opening doors and windows. A considerable amount of heat is
also stored in the masonry walls on the outside of the house, which helps
to keep night-time temperatures in the greenhouse above freezing for the
plants inside.
Typical of many houses constructed by architect-clients the project is
part of the continuing experiment and never seems to get finished! A num-
House conversion incorporating roof-space collector 319
ber of problems have been found with the greenhouse system in that,
whilst it is basically relatively cheap and simple it does not lend itself par-
ticularly well to being tilted away from the vertical on the south face and,
despite the avoidance of laps in the glazing details, the structure leaks air
and even driving rain, particularly around the roof vents. The thermal
performance of the conservatory is difficult to assess, though it did make a
noticeable contribution to heating the home in the first winter, but a sub-
stantial proportion of this may have been due to the reduction of cold-air
infiltration on a particularly windy site. In the exceptionally cold winter
of 1981—82, the conservatory temperature was maintained between 5 and
10 °C above the outside temperature, which was higher than expected.
The conservatory performed exceptionally well as a plant house, with
the first cucumbers being harvested early in May 1982. It appeared to
require little or no shading, and summertime temperatures in the green-
house and the house can be reduced to acceptable levels using through
natural ventilation.
12.5 Rehabilitated houses — house conversion incorporating a roof-space

collector (contributed by Peter Clegg)
Any passive solar measure incorporated into a building — particularly in
renovating an existing building - must either be very simple and econ-
omical to install, or must serve a dual function (for example, a conserva-
tory), in which case the entire cost of the passive solar measure need not
be justified in purely energy-saving terms. In our opinion, if neither of
these options is applicable, the measures may not be economical when
considered against simple high-insulation alternatives.
In the case of renovating existing buildings, it is often possible to justify
additional expense at a time when other work to the building is being
done, since the overcost of the passive solar measures will be much less.
This is the case in a recent conversion which incorporates a simple roof-
space collector into a former agricultural building.
The existing building is 25 m long and 6.0 m wide, single storey, and
faces south 30° east. The ratio of external envelope area to floor area is
very high, and the building form is thus bound to generate a higher heat
requirement than a more compact plan. The solar exposure was also very
good, and it therefore seemed appropriate to investigate various forms of
passive solar heating.
The amount of direct solar gain through windows was limited by a
desire to use the existing window openings and avoid excessive areas of
glazing directly into rooms. Conservatories were not regarded as appropri-
ate because of the low eaves height, and the desire to maintain the simple
form of the existing building. Trombe walls were rejected because of the
difficulty of adapting an existing masonry structure while retaining the
present windows.
The most appropriate option was to use the roof space itself as a col-
lector, and the house, with heavyweight masonry partitions and floors as
thermal storage, see Fig. 12.13(fl, b). The roof of the building needed strip-
Fig. 12.13. Conversion of an
agricultural building near Warm air distribution
Bristol for Glyn England. WINTER HEATING
(Architects: Feilden Clegg to living, dining and
Design. Energy Consultants: utility room SUMMER COOLING
Energy Design Group Ltd
with John Willoughby and
Cedric Green.)
Twin wall polycarbonate

roof-space collector
Return
air from
corridor
Summe
ventilation
out
House conversion incorporating roof-space collector 321
ping off, felting and replacing, and this was bound to result in rejecting a
certain percentage of the original tiles. The savings on replacing felt tiles
and battens could therefore realistically be taken into account and off-set
against the cost of the roof-space glazing.
The collector itself consists of 16 mm twin-wall Makrolan polycarbon-
ate sheeting supplied by Roehm Ltd, supported on battens at 1 m centres
over the existing rafters. This form of glazing is inexpensive but of long
life (see Chapter 4) and, unlike many plastics, does not have the drawback
of high infra-red transmittance. The horizontal area immediately under the
collector is insulated and lined with foil-faced building paper to reflect
solar radiation at high angles of incidence in summer, all other surfaces are
dark brown. Heat is distributed via a 400 mm diameter axial fan to a duct
with registers in the dining room, living and utility rooms with a return-air
fan at the end of the corridor leading to the bedrooms. The fans are con-
trolled by differential thermostats, and when the house can absorb no
more heat and the temperature rises above the upper limit of the differen-
tial controller (approximately 23 °C), the distribution fans shut down.
When the collector temperature then rises above 40 °C, the roof space is
ventilated to the outside air via roof vents controlled by heat-operated
greenhouse-type vent controls.
There is a further thermostatically controlled fan to exhaust warm air
from the collector should the heat-operated vents fail to operate or be
insufficient in area to cope with the build up of heat. In summer, manual
vents are open and the collection space is continuously ventilated to avoid
the build up of heat.
The building was studied using a thermal simulation model SCRIBE
devised by Cedric Green at the University of Sheffield, and various energy
strategies were compared. The results are given in Table 12.1, together
with the approximate costs of the various options.
The base case was simply to insulate the building to 1982 Building
Regulation Standards, which resulted in an annual heat loss of approxi-
mately 94 GJ. Increasing the insulation standards to halve the (/-values,
and allowing for the installation of triple-glazed high-performance win-
dows, reduces the seasonal heat loss by somewhat less than one-half (high
insulation case). The incorporation of the roof-space collector was shown
to further halve the seasonal heat requirements - a measure that was
calculated to be cost-effective over a period of approximately eight years,
based on a simple payback calculation assuming no increase in the real cost
of energy.
The proposed system is completely untested, but every indication is
that it illustrates an appropriate and straightforward use of passive solar
energy. The concept throughout the design of the installation has been to
use simple materials and off-the-shelf industrial components to avoid
excessive costs resulting from purpose-made designs. One of the key
elements was the design of the control system to allow for flexibility in
setting various temperature limits and fan-speed settings.
The proposal illustrates a general approach to passive solar design in the
UK, in that solar energy during the heating season (particularly the
a
Table 12.1. Results of computer simulation study using SCRIBE
Capital
cost No. of Fabric {/-values W/(m2 K)
(£) glazings walls roof floor
Base case 0 1 0.6 0.4 0.5

840 2 0.6 0.4 0.5
930 3 0.6 0.4 0.5
2900 3N/1S<* 0.3 0.2 0.3
High insulation 2600 3 0.3 0.2 0.3
d
High insulation, 4400 3N/lS 0.3 0.2 0.3
direct passive solar gain
High insulation + 4500 3 0.3 0.2 0.3
roof-space collector
Collector contribution 1900
only i
a
Available from Cedric Green, University of Sheffield, UK.
b
Night storage heaters.
c
The cashflow of course would become more favourable year by year assuming that the rises in fuel
costs outstrip those in mortgage costs.
d
Triple glazed with night insulation.
shortened heating season resulting from high insulation measures) is avail-
able in such limited quantities that it is not economical to introduce com-
plex and expensive storage systems. With masonry building there is often
sufficient storage capacity within the structure of the building itself to
absorb additional heat gains during the day. Glazing the roof, whilst
obviously causing potential problems of overheating during summer, does
allow for increased collection of diffuse solar energy in winter.*
12.6 Rehabilitated houses — a nineteenth-century terraced house

The rehabilitation of existing houses is of special importance to energy
conservation because it is here that spatial, constructional and financial
constraints are most clearly apparent.
Let us examine the recent renovation of the Cambridge home of one of
the present authors. Fig. 12.14 shows the original ground- and first-floor
plans, a fairly typical arrangement of 'two up, two down' with a kitchen
and bathroom tacked on at the back.
The garden end of the house faces 68° east of south and so offers only
a limited opportunity for the use of passive solar gain. Nonetheless, the
glazed area on this facade was made as large as was practicable (see Fig.
12.15) so that the house 'opened on' to the garden and so that deep
interior spaces were lighted naturally. Additionally, a roof light was
installed in the back bedroom to provide natural lighting.
*Note added in print: This rehabilitation is working extremely well. Reports are
available from the Energy Technology Support Unit of the Department of Energy.
A nineteenth-century terraced house 323
Vent rate Specific Auxiliary Energy6 Cost Annual savingc

air changes heat loss heat saving saving Capital cost
(per h) (W/K) (GJ/y) (GJ/y) (£/y) (£)
2 529 96 — — —
2 507 88 9 72 0.09
2 491 81 15 125 0.13
1 341 61 35 291 0.10
1 303 50 46 383 0.15
1 290 49 47 393 0.09
1 290 28 68 569 0.13
22 186 0.10
Bathroom
/WC
Kitchen
Bedroom Dining
room
Bedroom Living room Bedroom
(a) (b) Id)
Fig. 12.14. Original and

refurbished floor plans for a
Studies at Cambridge of the existing housing stock have shown that the terraced house (a) Original
ground floor, (b) Original
scope for the use of passive solar gain is significant. Approximately 60% of first floor, (c) New ground
all houses have either a facade and roof (55%) or a roof only (5%) which floor, (d) New first floor.
faces within 45° of due southJ 3 ' For terraced houses in this group,
analysis has shown that total glazing is 11 m2 on average with 5 m2 on the
'south' side (the terraced houses included a mixture of two- and three-
storey buildings). (For the refurbished house of Fig. 12.14, the 'south' side
window area is 7.8 m2 and the 'north' side 3.9 m 2 .) For all houses in the
Fig. 12.15. View of the
garden-facing facade of a
renovated terraced house in
Cambridge.
Cambridge study it was found that, on average, the solar contribution was
6% of the space heating energy consumption. For the 60% of all houses
facing within 45° of due south, the solar contribution is estimated to be
8%.
After years of living in what were beginning to feel like exceptionally
small rooms, it was decided to enlarge the house on both floors and to
make the ground floor open plan as shown in Fig. 12.14. Previously, the
living and dining rooms were each served by a gas fire (electric resistance
heaters were used upstairs) and effectively could be sealed off from each
other so that energy could be saved by just using heating where needed.
The open plan lacked this advantage but the desire for spaciousness pre-
vailed. The original staircase, however, which permitted sealing the ground
floor from the first floor, was retained.
A central-heating system, supplying both heating and domestic hot
water (thus replacing an electric immersion heater), was installed with a
gas-fired wall-hung boiler located in the kitchen, two radiators downstairs
and four upstairs. All radiators are equipped with thermostatic radiator
valves and a small standard programmer schedules the system for two
periods of operation each day. For amenity purposes or perhaps simply
because of a palaeolithic desire for an open fire, one of the fireplaces
housing a gas fire was opened up and a well-fitting chimney damper was
installed to reduce the ventilation heat loss when the fireplace was not in
use.
Consideration was given to the use of internal insulation on the street
side of the house to improve the thermal characteristics of the 225 mm
brick wall but was ruled out when the builders' estimated costs were
received. For the new addition, however, it proved relatively easy (some
problems were encountered with items such as lintel widths) to incorpor-
References 325
ate high levels of insulation. The external garden-facing wall consists of
12 mm plaster lining, 100 mm lightweight concrete block, a 50 mm
cavity lined with polystyrene slabs and 105 mm exterior brick, giving a
{/-value of 0.43 W/(m2 K). The sloping ceiling-roof in the bedroom
which looks onto the garden is of 12 mm plasterboard, a polythene
vapour barrier, 150 mm glass fibre insulation between the joists, and slate
tiles on roofing felt and battens, giving a {/-value of 0.22 W/(m2 K). The
new solid floor in the kitchen consists of quarry tiles, a 50 mm screed,
50 mm expanded polystyrene board, a damp-proof polythene membrane
and a 100 mm concrete slab over 100 mm hardcore; the estimated {/-value
is 0.35 W/(m2 K). The insulation in the existing loft has been increased
from 100 mm to 200 mm of glass fibre.
All windows and doors are weatherstripped. Double glazing was con-
sidered but delayed because of limited funds. When restoring a house it is
only too easy, and entirely understandable, to use the mortgage money
for the work that only the builder can do and to plan on the installation
of, say, secondary double glazing later by either a specialist or oneself.
Other cost sacrifices included a small conservatory and active solar
collectors mounted on the roof for domestic hot water. Needless to say, a
solar heated swimming pool in the garden was reluctantly axed.
Water and waste services were all concentrated in one corner to reduce
costs and length of runs. The hot-water cylinder and all hot-water piping
are insulated. In the future it might be possible to install a small heat-
recovery unit in the kitchen to take advantage of the waste heat from the
bath above and perhaps the sinks.
The house as described has a design heat loss of 7 kW. The calculation
used follows the method of Appendix 2 and assumes an hourly air-change
rate of 1.2 and an internal to external temperature difference of 20 K.
The heat loss could be reduced in the future by employing the insulation
measures which were not possible because of cost during the initial
renovation.
The overall result is a comfortable, pleasant home which is much more
open to the garden than the original house was. One of the charms of
nineteenth-century terraced housing is that it is not too difficult to con-
vert from the mean, cold and damp conditions the first occupants endured
to more spacious and enjoyable surroundings. The energy supply and con-
servation measures which should accompany that transition are in many
ways modest but they are significant and of immediate effect.
References
[ 1 ] Air Infiltration Centre (1983). Building Design for Minimum Air
Infiltration. Air Infiltration Centre, Bracknell, UK.
[2] Pressure testing carried out at the Polytechnic of Central London,
1982.
[3] Penz, F. (1982). Private communication based on work on passive
solar gain at the Martin Centre, Department of Architecture, Uni-
versity of Cambridge, UK.
13
Non-domestic case studies
13.1 Introduction
This chapter briefly presents some non-domestic buildings designed with
energy conservation in mind.
13.2 The swimming pool, Sheiling Schools

Given the opportunity to design a swimming pool for a school for handi-
capped children, it was automatically part of the brief to keep the running
costs to a minimum, despite the fact that water temperatures had to be
maintained at 27 °C for use throughout the year. The Energy Design
Group made an initial study which investigated the potential of heat
pumps and heat-reclaim systems, glazed and unglazed solar collectors,
variable speed ventilation control and the use of a pool cover.
Initially it was felt that a heat-pump system would be preferable, using
extract air as a primary source of heat. This, however, entailed expensive
duct work to return the air to the boiler room, and also added consider-
ably to the capital costs since an auxiliary gas-fired heater could not be
dispensed with. Other heat-reclaim systems were dependent on the use of
ozone or other very expensive purification systems which reduced their
overall cost-effectiveness. (Chlorine from the pool is corrosive over the
long term.)
The entire building was built to a very low budget, competitive with
quotations from design-and-build contractors who offered cheap standard
solutions. The over cost of energy-saving measures had thus to be kept to
a minimum.
It was therefore decided to opt for high thermal insulation (100 mm of
polystyrene on the roof and 80 mm of glass fibre to the walls with double
glazing to all windows), a minimum volume for the building, and a high
internal thermal mass. The greatest energy savings were effected by a four-
step ventilation-control system operated by a timeclock and humidistat
sensor. This, in conjunction with a simple pool cover was calculated to
halve the actual energy consumption of the pool for an overcost of
approximately £800 (1980 prices). When the pool is not in use, the pool
cover is unrolled by the caretaker and the ventilation rate is automatically
reduced, as is the throughput of warm air. At night the system shuts down
to the extent that it is operating on a 150 mm diameter Ventaxia wall fan.
326
The swimming pool, Sheiling Schools
Fig. 13.1. The Sheiling
Pool, Thornbury Schools
(Feilden and Clegg
Partnership).
With such drastic reduction in the waste heat from the building, other
ways of reclaiming the heat become uneconomical. A roof light with a
south-facing slope was incorporated for the time when solar collection
becomes economical in the future. Detailed predictions showed that
because of the high water temperature required and year-round use, pool
insulation would only be marginally cost-effective. Further reductions in
energy consumption were achieved by using a central thermostatic control
on the water supply to the showers, and percussion valves on the showers
themselves reduced water consumption.
The building is primarily of interest as an example of constructing a
low-energy building to a very low budget and of devising an approach
which first and foremost is concerned with reducing energy consumption
by simple means rather than being attracted to more exotic and expensive
methods of producing energy.
Fig. 13.1 illustrates the complete scheme.
328 Non-domestic case studies
Fig. 13.2. View from the
south-east of the new BRS
office building.! * 1 (Courtesy
of the Property Services
Administration.)
13.3 The new BRS office building

The new BRS office building at Garston, Watford (see Fig. 13.2) is one of
a handful of non-domestic buildings for which energy conservation was a
major consideration. In designing the engineering services the principal
objective was to minimize energy consumption over the life of the build-
ing, while maintaining a good standard of environmental comfort.
The building is rectangular in shape with the long axis running east—
west. By calculating the total primary energy consumption for lighting and
plant and taking into account passive solar gain through the double-glazed
windows which are used throughout, it was found that the optimum, that
is, minimum energy size window was at a 50% window/wall ratio (giving
an overall U-value of 1.8 W/(m2 K)). For the north facade the comparable
figure is 30%.
To guard against overheating, individually motorized external sun
blinds are provided for each 3.6 m wide structural bay. When raised the
blind fits neatly into a recess above the window and when lowered, the
blind is at eye level from standing position in the room. In this position it
prevents direct radiation entering the room when the sun's altitude is
greater than 30°.
The blinds lower automatically when required to reduce the risk of
uncomfortably high temperatures. The following conditions must be met
for operation:
(1) Solar radiation must be high enough to justify protection from the
sun.
(2) The room must be at or above 18 °C and space heating is not
required.
(3) Wind conditions must be sufficiently calm for use of the blinds.
Agricultural buildings 329
The room occupants may override the automatic blind controls if they so
wish.
Heating is from three atmospheric gas-fired boilers with flue-gas
dilution, which supply variable temperature hot water to natural and
forced-air convectors along perimeter walls. A recirculation mechanical
ventilation system is included with the rate of fresh air intake variable for
experimental purposes. Heat recovery on the extract air is effected by a
heat wheel of 75% total efficiency for latent and sensible heat combined.
For research purposes, a solar heating system providing water for wash
basins is included. During the winter windows are locked shut but in the
summer months the mechanical ventilation system is switched off and the
windows are unlocked so that they may provide natural ventilation.
Automatic controls are provided for the lighting in order to conserve
electrical energy and to provide BRE staff with a facility for future experi-
ments. Occupants can only switch on their lights when sky illuminance
does not afford adequate lighting. This is accomplished by allowing
manual switching of a back 'bank' of lights when the daylight level falls
below a level equivalent to 300 lux at a point 2 m into the room; at
200 lux, at the same point, the front 'bank' becomes available for switch-
ing.
13.4 Agricultural buildings

One of the most promising applications of solar energy is for crop drying,
particularly grain crops, since the period of greatest solar energy is almost
concurrent with the harvest season.
In many years in the UK a high proportion of the grain harvested must
be dried before it can be stored safely in bulk. When necessary, about one-
third of the grain is dried on farms using 'high'-temperature (40 °C
upwards) 'continuous' and batch driers, although relatively little use is
now made of the high-temperature batch driers. Most of the other two-
thirds of the grain is dried more slowly in low-temperature bulk driers,
often using only ambient air. Limited inputs of heat may be used at times
to raise the temperature of the incoming air to produce a relative humidity
low enough to enable drying of the grain to continue. Depending on the
incoming moisture of the grain, the air flow, volume of grain and design of
the plant, etc., drying to a safe moisture content may take from one to ten
days or even longer. (Safe moisture content for long-term bulk storage of
grain is 14%, and is lower for certain other seeds.) Unless heat is used for
long periods with this system, most energy is used in blowing the air. With
warm air available from solar collectors, the time taken for drying grain
could be much reduced, if some way of recirculating the air could be
found, and the need for additional heat much reduced or eliminated dur-
ing daylight hours.
Fig. 13.3 shows a design (patent pending) developed by Mr Richard
Wedgwood and one of the present authors for a solar roof for a typical
storage barn. The system is relatively inexpensive and has the advantage
Fig. 13.3. Barn using solar Outer skin
heating for crop drying. I 2 1 Direction of air flow fnsulating inner surface
(Courtesy of R. Wedgwood.) Air Manifold
intak
Duct
Feeder ducts Central /

ducting Fan
of being suitable for new and existing barns. No storage facilities are
required and maintenance is an absolute minimum.
Air is drawn laterally along a double-skin roof consisting of an outer
surface which absorbs solar radiation, and an inner insulating surface.
Suitable materials for the outer surface include sheet aluminium painted
black, black plastics, black-painted asbestos and corrugated sheet steel.
The insulating surface can be insulating fibreboard, fibreglass, expanded
polystyrene boards or a number of other materials.
The heat absorbed is transmitted through the outer surface to the air
stream which it warms. The air then passes through a manifold and duct
before distribution by a fan unit to central ducting and feeder ducts under-
neath the stored, harvested material.
Work elsewhere!3' 41 indicates that the efficiency of the double-skin
collector is likely to be in the range of 10—40% depending on such factors
as materials and air-flow rate.
The economic viability of the design depends principally on the costs of
collector materials and the electrical energy to run the fan. Preliminary
studies indicate that the proposal is economically attractive now and
monitoring studies are underway to confirm this.
13.5 School buildings (contributed by Nick Baker)

The average delivered energy consumption per school child is around
4.9 GJ/y and most of this goes to provide space heating.^5! Taking
account of the various prices of fuels used, this represents about £40/y.
This significant cost has prompted education authorities to make strenu-
ous efforts to reduce energy consumption in existing buildings, and to
make energy conservation a major factor in new school design, and it is
this area with which this section is concerned.
Before discussing the contemporary ideas in low-energy design, it is use-
ful to look at the environmental problems of schools in a recent historical
context.
The 1950s and 1960s saw a certain preoccupation with providing
adequate daylight. There existed a design rule for a minimum daylight
factor of 2% and, to achieve this, shallow-plan buildings with large areas of
School buildings 331
Fig. 13.4. Romsey School,

Hampshire.
glazing were necessary. This, together with contemporary tastes in archi-
tectural style, resulted in school buildings such as Romsey School, Hamp-
shire, which was completed in 1959 (see Fig. 13.4).
Then, system building was seen as a way to meet the growing demand,
by reducing building time and design costs. Many authorities formed con-
sortia in order to develop their own building systems which, although they
did not have to be, tended to be of lightweight construction type.
At no time during this period was insulation given much consideration.
Energy was cheap and an understanding of thermal comfort was not
applied to the school environment. The result of this combination of a
high level of glazing, large surface area and a lightweight and poorly insu-
lated structure was a very poor-quality environment with the discomfort
of winter underheating only being exceeded by that of summer overheat-
ing. Moreover, because of the inherent lack of thermal stability in such
buildings and the customary use of rudimentary control systems with a
single-boiler thermostat, winter underheating was sometimes combined
with winter overheating — in sunny conditions it was not unusual to have
occupants of south-facing rooms sweltering and opening windows while
those in north-facing rooms shivered.
It seemed to take some time to realize that an environment of this kind
might not be the best in which to learn and to teach, and various
studies I 6 ' 7J were carried out at the BRE in the late 1960s and early
1970s. Parallel work was also being carried out on the ideal office environ-
ment where similar problems had been experienced with highly glazed
office blocks.! 8'91
The notion of creating an 'artificial' or 'controlled environment' began
to be applied to school building.110» n l A mild reaction set in which was
strengthened by the conviction that a 'controlled environment' meant an
economical one. By 1974 the oil crisis - oil being a fuel on which the
majority of schools relied — had begun to make energy economy a serious
factor for the first time since the war.
A number of prestigious office blocks had been completed, boasting
'integrated environmental design' with deep plans, small windows demand-
ing high levels of artificial lighting, and full mechanical ventilation. Control
Fig. 13.5. Elmstead Primary
School, near Colchester,
Essex, (a) Exterior.
(b) Interior. (Courtesy of
Essex County Council.)
was taken away from the occupants who were informed that they were
now enjoying 'an optimum environment'.
The principles embodied in these office designs were at this time
adapted to a number of schools. Undoubtedly it was a reaction; a belief
that openable windows and daylight lead inevitably to energy wastage, but
it would be unfair to suggest that the early examples of controlled-
environment schools were as barren and sterile as their bigger commercial
office counterparts. Scale and educational requirements resulted in many
of these designs being very pleasant spaces and the improved environment
- thermal, lighting and acoustic — was appreciated by their occupants.
A good example is Elmstead Primary School, near Colchester, Essex
(see Fig. 13.5). Here, a novel constructional system of load-bearing con-
crete panels, with a heavyweight roof in conjunction with a small area of
glazing and deep plan, provided a stable thermal environment. Energy
Fig. 13.6. Heat-pump oper-
ation at Roach Vale Primary
School, (a) Winter operation.
(b) Summer operation.
(Courtesy of Essex County
Council.)
TEACHING AREA COURTYARD
(a)
fresh outside
TEACHING AREA
(b)
conservation was beginning to dominate considerations of comfort alone

and, in subsequent buildings of this type, more emphasis was placed on
energy conservation. Essex County Council, in particular, began to experi-
ment with heat pumps and heat recovery, as exemplified by the Roach
Vale Primary School at Colchester^11 J (see Fig. 13.6).
To summarize then, the school environment had deteriorated with the
development of highly glazed buildings and, subsequently, lightweight con-
struction, during a period when low energy costs did not draw attention to
the need to provide a more efficient envelope. Awareness of discomfort
problems, closely followed by the rising energy costs of the early 1970s
brought about a reaction which tended to reject the need for daylighting
and natural ventilation and adopt an 'engineered' or 'controlled-environ-
ment' approach. This brings us to the present day.
There is currently yet another swing in emphasis. The two main factors,
comfort and energy conservation, are as important as ever, particularly the
latter, but our understanding of them has developed. Firstly, we are now
beginning to understand that energy conservation is not a purely technical
problem in the sense that we simply have to increase insulation, reduce
surface area, seal up the windows and mechanically ventilate. Indeed,
studies! 121 have shown that some early controlled-environment schools
consumed more energy than their earlier counterparts, and certainly this
has been the case with many office examples. We now realize that the
occupants have to be taken into account. Field studies made by Haighl13!
Fig. 13.7. Yateley Newlands

School, Hampshire, (a)
Exterior, (b) Plan, (c) Con-
servatory. (Courtesy of R. have focussed attention on the need for better-designed controls if the
Bryant and The Architects occupants are expected to interact intelligently with the building and its
Journal.) systems.
Thus it might be said that the problem of designing a building and its
systems, however automated and controlled, and ensuring that it is econ-
omical, is much more difficult than building scientists thought a decade
ago. Related to this, we now realize that our notions of 'the optimum
environment', one which is held within a narrow band of environmental
parameters, was oversimplistic. A view that is gaining acceptance is that, in
fact, most occupants are happier in an environment which shows consider-
able variation in both time and space.!14!
This has led to the current position in school design where a much less
closely controlled environment is provided. A further factor is that a
'selective' approach is being made to the ambient environment as a source
of light, solar energy, and air movement, rather than an 'exclusive'
approach, where the external environment is seen as something to exclude
at all costs.! 151
A good example of this more relaxed approach is the primary school at
Ground floor plan.
Section AA.
Section BB.
Yateley Newlands, Hampshire,! 16J shown in Fig. 13.7. Although it was not
designed with energy conservation as the major criterion, it is well insu-
lated, and has a number of characteristics which result in low energy use.
One of these is that, although it is a daylit school, it is quite different from
the overglazed 2% daylight factor schools of the 1960s and 1970s. This is
a result of understanding that daylighting quality, as well as quantity, is of
major influence on the degree to which artificial lighting has to be used.
Rooms which have a large range of brightness due to the provision of large
glazed areas on one side only, without diffusing surfaces, will require
artificial lighting in the dimmer areas even when the average lighting level
is quite high. At Yateley, this problem has been solved by providing high-
j
ii 111
^ • • • • * • • ,
(0
level roof lights which illuminate a light-coloured wall, as well as side

windows.
Yateley, due in part to its need for daylight, has moved away from the
deep plan and has adopted an elongated form, compensating for increased
surface area by adopting high levels of insulation. This plan form immedi-
ately offers the possibility of much greater use of solar gains, provided of
course that the correct controls are installed, and that overheating dis-
comfort is prevented by shading the direct sun from the occupants and
providing sufficient ventilation.
Another interesting feature of Yateley is the atrium. This space, which
is unheated and is formed simply by glazing over an otherwise open court-
yard, has proved to be a valuable asset to the school and functions both as
part of the general circulation area and a useable space in its own right,
Fig. 13.8. St John's School,
Clacton on Sea, Essex.
(a) Exterior, (b) Plan.
(c) Schematic of energy
systems. (Courtesy of Essex
County Council.)
although its environmental standards are quite different from the fully
protected interior.
A recently completed project, which demonstrates a more explicit
move towards ambient energy sources, is St Johns School, at Clacton on
Sea, Essex,l17l shown in Fig. 13.8.
Fig. 13.9. Locksheath

School, Hampshire. The Essex County Architects' Department, which was responsible for
(a) Seasonal operation. the earlier controlled environment approaches typified by Elmstead, to
(b) Interior of conservatory. which reference has already been made, have adopted a complex range of
technical options - heat pumps using groundwater, thermosiphon air-
heating panels and conservatories. Whilst still showing a highly 'engineered
approach', it has moved some way from its earlier 'exclusive' predecessors.
Another Hampshire project at Locksheath (see Fig. 13.9), demonstrates
that even a simple concept can become complex in its realization.^18!
Having determined that in a well-insulated building it is ventilation which
accounts for the major loss (because of the statutory high fresh-air require-
ment), the designers use a conservatory running along the south side of the
elongated building, as both circulation space, and as a solar collector to
preheat ventilation air drawn into the class spaces via a warm-air heating
system. This has the advantage of always making use of the gains, even
when the conservatory temperature is below that of the heated spaces,
while not suffering from the problems of direct-gain spaces, where seated
occupants in direct sunlight experience overheating even at modest air
temperatures. The conservatory also provides a shared craft area, and
thermal analysis shows that temperatures will for the most part be accept-
able for non-sedentary activities, with no auxiliary heating.
What has been the result in terms of energy performance, of these
developments in school design? In a survey! 51 carried out in the mid
1970s of schools in Essex, most of which had not yet adopted any signifi-
cant energy-conservation measures, the average delivered energy was about
1 GJ/(m2 y), or 1.3 GJ/(m2 y), primary energy. However, there was wide
Fig. 13.10. Energy con-
sumption in schools, t 5 1
to
§ 15
I
o
L
r
0.5 1.0 1.5
2
Delivered energy (GJ/(m y)
divergence of individual schools' energy consumption, as shown in Fig.

13.10.
This variation was not attributable to a systematic variation of building
type, suggesting that occupant behaviour and management was the major
influence. Detailed surveys of schools often indicate this - for example, in
one school the caretaker will carefully vary switch-on time throughout the
heating season to prevent wasteful preheating, whereas in an otherwise
similar school, the time clock may simply be set for 6.00 a.m. for the
whole heating season. This, together with a few other extravagances such
as leaving windows open, can easily lead to a doubling of energy consump-
tion.
Thus it is difficult to establish a baseline for comparing the actual effect
of building design. Rather, we have to take the view that we are making
improvements in building design, system design, management and occu-
pant behaviour, all together. The degree to which these are interdependent
is an interesting issue.
There is still a shortage of monitored energy data from schools, and
subsequent comparison of figures relies upon calculated values. In 1980,
Page!17! showed a comparison between primary energy consumption for
controlled-environment schools in Essex, and a target value given in
Department of Education and Science Design Note 17 (1979). The com-
parison showed that whilst earlier controlled environment designs were
above the target, they were well below the standard SEAC (system-built)
school of that time. Furthermore, later improved designs were well below
the DES target, as shown in Fig. 13.11.
Primary energy consumption is sensitive to the use of electricity due to
the poor conversion efficiency at the power station. Thus, although heat-
ing energy is small, the lighting and fan power required in controlled-
environment buildings constitutes a relatively high primary energy
demand. The 'selective' approach, typified by the approach of the
Hampshire schools, may result in higher heating demand, but the overall
primary energy performance is likely to be as good or better than the
'exclusive' controlled-environment buildings. For example, the overall
Fig. 13.11. Comparison of
annual primary energy in 1.5 -
schools.l 1 7 l
[~~]SEAC (typical)
primary schools DES maximum current standard
1.0 - ELMSTEAD
ROACH VALE
2* WALTON THE GUNFLEET

I
CLACTON STJOHNS
> 0.5
CO
primary energy consumption of Locksheath school is calculated to be as

lowas0.4GJ/(m 2 y).
We must not delude ourselves that calculated and actual performance
are one and the same. As we have seen already, the performance of a build-
ing is very much dependent upon the behaviour of the occupants. The
'selective' buildings will require even more cooperation from the occupants
than the controlled-environment buildings. Nevertheless, a target has been
identified and the principles of how this target can be met are known.
To summarize, school building design is currently directing its attention
to the 'selective approach', producing buildings which in plan form and use
of glazing are not so different from the buildings of two decades earlier.
However, with a much greater understanding of thermal comfort require-
ments, and of the need to integrate the ambient gains with the auxiliary
systems, these contemporary buildings will perform much better. Further-
more, the educational values of these buildings, which clearly relate to the
climate, is being recognized. There is considerable enthusiasm and support
in the development of these designs, and it is likely that we shall see more
progress and innovation in selective (or passive) design in new schools than
in most other non-domestic sectors.
References
[ 1 ] Price, P. (1977). 'Energy conservation in new offices at BRS'.
Construction, 23, 25-8. This section is based entirely on this
article; we are greatly indebted to the author.
[2] British Patent Filing No. 7 941 163. November 1979. Improve-
ments in and Relating to Crop Drying Facilities. R.B. Wedgwood &
R.B. Thomas.
[3] Ekstrom, N. & Gustafsson, G. (1979). The Application of Solar
Collectors for Drying of Grain and Hay. Uppsala, Sweden: Swedish
Institute of Agricultural Engineering.
[4] Ferguson, W.E. & Bailey, P.H. (1979). 'Solar air heater for near-
ambient crop drying: description, testing methods and 1977
References 341
results'. Departmental Note No. SIN I275. Penicuik: Scottish Insti-
tute of Agricultural Engineering.
[5] Yannas, S. & Wilkenfield, G. (1978). 'Energy strategies for second-
ary schools in Essex'. Research Paper 1/78. London: Architectural
Association.
[6] Langdon, FJ. & Loudon, A.G. (1970). 'Discomfort in schools
from overheating in summer'. Journal of the Institute of Heating
and Ventilation Engineers, 38, 1 8 1 - 9 .
[7] Humphreys, M.A. (1974). Classroom Temperature, Clothing and
Thermal Comfort - A Study of Secondary School Children in
Summertime. BRE CP 22/74. Garston, Watford: BRE.
[8] Humphreys, M.A. & Nicol, J.F. (1970). 'An investigation into the
comfort of office workers'. Journal of the Institute of Heating and
Ventilation Engineers, 37, 2 6 5 - 7 9 .
[9] Hardy, A.C. (1971). 'Architecture and building science'. RIBA
Journal, November.
[10] (1971). Integrated Environmental Design: A Feasibility Study for
School Buildings. County Architects Department, Gloucester
County Council, December.
[11] Crowe, R.V. & Page, P.A. (1976). 'Integrated environmental design
of schools'. Proceedings of the Symposium on Energy Conservation
in the Built Environment. London: Construction Press—CIBS.
[12] O'Sullivan, P. & Austin, MJ. (1976). 'Energy targets (schools)'.
Proceedings of the Symposium on Energy Conservation in the Built
Environment. London: Construction Press—CIBS.
[13] Haigh, D. (1981). 'User response in environmental control'. In: The
Architecture of Energy. London: Construction Press.
[ 14] Cooper, I. (1981). 'Comfort theory and practice'. Applied Energy,
2 (4), 2 6 7 - 9 .
[15] Hawkes, D. (1981). 'Building shape and energy use'. In: The
Architecture of Energy. London: Construction Press.
[16] Smith, C.S. & Hawkes, D. )1981). 'Yateley Newlands Primary
School'. Architects Journal, 173 (25), 1199-214.
[17] Page, P.A. (1981). 'School buildings - Essex County Council
Architects Department'. In: The Architecture of Energy. London:
Construction Press.
[18] Baker, N.V. (1982). 'The influence of thermal comfort and user
control on the design of a passive solar school building'. Energy
and Building (in press).
Appendix 1: Weather data
A 1.1 Introduction
UK weather data should be covered in a fairly comprehensive way, in a

book currently under production, by Page & LebensJ1J In the meanwhile,
very restricted data is given here. Maps of global solar radiation are avail-
able from the Meteorological Office for the UK at Bracknell.
French data is presented in the Atlas Solaire Francais^ for 33 sites in
terms of solar energy penetrating glazing at various orientations. Details
are also given for overshadowing and degree days.
US data is well presented in the book by MazriaJ3!
Data for Germany can be found in Bossel's bookl 4 ' which contains
ambient temperature and global solar radiation, amongst other data, for
each month.
A 1.2 Solar spectrum

The spectral distribution of solar radiation varies with local conditions and
time of day, but Fig. A 1.1 provides representative values. The diagram
also indicates how glass affects the distribution of energy with wavelength.
Fig. A 1.1. The spectral dis-

tribution of transmitted
energy.! 5 1
spectral transmission of
6 mm heat absorbing glass
^transmitted by
eat absorbing glass
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2
Wavelength in micrometres
342
Weather data 343
Fig. A 1.2 gives the transmittance of float glass for comparison. It may be
seen that most of the ultra-violet, and all of the infra-red characteristic of
bodies at room temperature, is filtered out.
Fig. A 1.3 illustrates values on inclined surfaces at Kew, calculated by
Page.
A 1.3 Global solar radiation data for the UK — monthly means in

MJ/(m 2 d). Kew and Bracknell latitude, about 51.5° (tilt 0 = horizontal;
90 = vertical)
Kewl6 l
azimuth Bracknelll 7! Bracknelll 7 ] Bracknelll 7 ! Bracknelll 7!
Month tilt 0 south 90* east 90* west 90 a north 90fl
January 1.9 3.0 1.2 1.1 0.8

February 3.6 5.4 2.8 2.5 1.2
March 7.4 7.8 4.8 3.8 1.9
April 11.5 7.7 6.6 6.0 3.0
May 15.9 7.7 8.3 7.7 4.3
June 17.4 7.7 9.8 9.0 5.0
July 15.8 7.7 8.0 8.4 4.9
August 13.3 7.7 6.8 7.0 3.4
September 10.2 7.8 5.2 5.3 2.3
October 5.7 6.9 3.4 3.6 1.6
November 2.5 4.9 2.0 1.8 0.8
December 1.5 3.0 1.1 1.1 0.6
a
Excluding ground-reflected radiation.
A 1.4 Global solar radiation data on the horizontal for three UK stations
in MJ/(m 2 d)
Kew Eskdalemuir Lerwick
January 1.9 1.5 0.8

February 3.6 3.9 2.7
March 7.4 6.3 5.6
April 11.5 10.8 11.8
May 15.9 14.0 14.5
June 17.4 16.3 16.7
July 15.8 13.8 14.0
August 13.3 11.3 11.0
September 10.2 8.1 7.3
October 5.7 4.3 3.2
November 2.5 2.0 1.1
December 1.5 1.2 0.6
Latitude 51.5° 55° 61°
A 1.5 Direct and diffuse solar radiation for Kew, UK (Fig. A 1.4)
It is important for readers to note the very high fraction of diffuse radi-
ation from the overcast skies of the UK.
344 Appendix 1
Fig. A1.2. The spectral trans- 100
mission characteristics of
glasses.! 51 90
80 mrn f loat glass
S
c 70 /K \
> S
j \
o
1 60 \
f"
§ 50 \
3 mm heat absorbing glas 5
I 40 v ^—^
* 30
20
\ \
/
10
0
p p
u
p p POO-* -* N3
I
• N
-—-«.
B
CO
S §§8 8 8 CJ1
8 S §S8
CJI
o
s Wavelength in micrometres
o
A 1.6 Temperature data for the UK (monthly means °C)
Month Kew Eskdalemuir Aberporth

January 4.0 3.0 5.8
February 4.9 3.1 5.6
March 6.8 4.0 6.3
April 9.4 5.5 9.0
May 12.5 8.4 11.2
June 15.9 10.1 13.5
July 16.9 12.1 15.3
August 16.5 12.0 15.3
September 14.7 10.3 14.0
October 11.8 8.1 11.8
November 7.5 5.5 8.5
December 4.9 4.1 6.7
Latitude 51.5° 55° 52°
A 1.7 Wind data

Wind data is presented in Chapter 8.
References
[1] J. Page and R. Lebens are preparing comprehensive tables of data

for a publication in conjunction with the UK Department of
Energy. The book, entitled Data Handbook, will be published early
1984.
[2] Claux, P., Pesso, J. & Raoust, M. (1982). Atlas Solaire Francais -
Energie Transmise et Calcules de Laisser de Masque.
[3] Mazria, E. (1979). Passive Solar Energy Book, professional edn.
Emmaus, Pa: Rodale Press.
References 345
18 Fig. A 1.3. Mean monthly
daily values of total irradi-
16
-
4 45° tilt ation on slopes of 40, 60
and 90° at Kew, compared
with that on a horizontal
14 -
1 : Y
60 c tilt
surface. Estimated from
horizontal surface data,
1957-71.[7]
/( ' X
12
90 ° tilt
: 10 >
"O
8
I
6
I
I
4h
I
2 r -_ A Horizontal
i i i i i i i 1 1 1 1
Jan Mar May Jul Sep Nov
Fig. A1.4. Monthly mean

daily irradiation on a vertical
KEW Horizontal: measured Horizontal south-facing surface (esti-
15.0 Vertical south: estimated mated) and a horizontal
surface at Kew, measured
1955-70. Direct part clear,
diffuse stippled.! 7 1
Vertical south
10.0
5.0
J F M A M J J A S O N D J F M A M J J A S O N D
Month
[4] Bossel, U. (1980) 'Klimadaten Mitteleuropas 3'. Adelebsen, W.

Germany: Solentec GmbH.
[5] Turner, D.P. (ed.) (1969). Windows and Environment. Ormskirk:
Pilkington Brothers UK.
[6] Collingbourne, R.H. (1975). 'UK solar radiation network on the
availability of solar radation data from the Meteorological Office
for solar energy application'. International Solar Energy Society,
UK Section, Conference Proceedings, February 1975. London.
[7] 'Solar Energy UK Assessment'. International Solar Energy Society,
UK Section, 1976. London.
Appendix 2: Thermal performance
An estimate of the thermal performance of a building is vital for sizing the

heating plant (and the ventilation and refrigeration plant if there is one),
for thermal comfort, for the avoidance of condensation and for energy
conservation. The simplest way of estimating thermal performance is the
{/-value technique based on the steady-state heat-loss equation
q= (?AU+Cv)(ti-to),
where
q = heat requirement (W);
A = area of a structural element (m 2 );
U = thermal transmittance of a structural element (W/(m2 K));
Cv = ventilation loss (W/K);
t{ = inside temperature (°C);
tQ = outside temperature (°C).
The equation is valid in situations where temperatures do not change sig-
nificantly with time. It is more accurate when applied to buildings with
little fenestration in, say, winter conditions of low external temperature
and little or no solar gain. It is less suited to buildings designed for passive
solar gain on an autumn day when the sun appears intermittently through
the clouds. There are also important reservations about the value of a
technique based on steady-state, one-dimensional heat flow to deal with
the problems of, for example, corners in buildings, thermal energy storage
in the building fabric and changes in the (/-value due to, say, moisture
penetration.
However, this method has the great virtue of simplicity and is readily
understood by those familiar with only the simplest mathematics. The
(/-value, or thermal transmittance, of a structural element is given by
(/-value (W/(m2 K))=l/R,
where R(m2 K/W) is the sum of the resistances of surfaces and layers of
material. More explicitly,
=
^ ^ s i + ^ i +^?2 + • • • +^w +^so>
where
Rsi is the resistance of the inner surface (dependent on both radi-
ation determined by the emittance of the surface and convection
determined by the layer of air immediately adjacent to the
346
Thermal performance 347
x
Table A 2.1. Inside surface resistance, Rs^ ^
Surface resistance (m2 K/W)
High emissivity Low emissivity

Building element Heat flow factor factor
Walls Horizontal 0.12 0.30

Ceilings or roofs, Upward 0.10 0.22
flat or pitched,
floors
Ceilings and floors Downward 0.14 0.55
Notes:
(1) High emissivity factor assumes 6j = e2 = 0.9.
Low emissivity factor assumes ei=0.9,6? = 0.05.
(2) Surface temperature is assumed to be 20 C.
(3) Air speed at the surface is assumed to be not greater than 0.1 m/s.
(Reproduced from Section A3 of The CIBS Guide, by permission of the
Chartered Institution of Building Services.)
Table A 2.2. Outside surface resistance (Rso)for stated exposure^
Surface resistance for stated exposure

(m2 K/W)
Building Emissivity
element of surface Sheltered Normal Severe
Wall High 0.08 0.06 0.03

Low 0.11 0.07 0.03
Roof High 0.07 0.04 0.02
Low 0.09 0.05 0.02
Note: Form (shape) factor for radiative heat transfer is taken to be unity.
surface);
Rx,. . . , Rn are the resistances of the layers of material (or voids);
and
Rsa is the resistance of the outer surface.
Internal and external surface resistances are given in Tables A 2.1 and
A 2.2 respectively.
The thermal resistance of a layer of material is given by
R{ (m2 K/W) = thickness of layer (m)/k (W/(m K)),
where k is the thermal conductivity. Typical thermal conductivities and
other properties of common building materials are shown in Table A 2.3.
Table A 2.4 shows the resistance of a variety of air spaces.
Table A 2.5 shows a (/-value calculation for the wall of Fig. A 2.1.
(/-values have been calculated and compiled for a wide variety of con-
structions. Table A 2.6 gives both (/-values and admittance (Y-) values (dis-
348 Appendix 2
1 1
Table A 2.3. Properties of common building materials* 1 1
Thermal Specific heat

Density conductivity capacity
Material (kg/m 3) (W/(m K)) (J/(kg K))
Walls
(external and internal)
Brickwork (outer leaf) 1700 0.84 800
Brickwork (inner leaf) 1700 0.62 800
Concrete block 600 0.19 1000
(lightweight)
Plasterboard 950 0.16 840
Surface finishes
Plaster (dense) 1300 0.50 1000
Plaster (lightweight) 600 0.16 1000
Roofs
Asphalt 1700 0.50 1000
Felt — bitumen layers 1700 0.50 1000
Floors
Cast concrete 2000 1.13 1000
Timber flooring 650 0.14 1200
Insulation
Expanded polystyrene 25 0.035 1400
(EPS) slab
Glass fibre quilt 12 0.040 840
Polyurethane board 30 0.025 1400
Urea formaldehyde 10 0.040 1400
(UF) foam
a
Values used in the calculation of U- and 7-values by the CIBS.
cussed below) for several examples. Tables A 2.7 and A 2.8 give {/-values
for glazing and windows (for floors see Chapter 3).
Now let us consider the heat loss from the simple structure, shown in
Fig. A 2.2, which incorporates the wall in Fig. A 2.1. Table A 2.9 shows a
calculation for the fabric heat loss for the building, assuming an external
temperature of —1 °C and an internal temperature of 19 °C as the design
conditions. These could be considered as the basis for the 'maximum' heat
loss for which the heating system should be sized.
To the fabric heat loss must be added the ventilation heat loss, Cv,
which is given by the formula
where
N = the number of air changes per hour;
V = volume of the space (m 3 );
t{ = inside temperature (°C);
to = outside temperature (°C).
Table A 2.4. Standard thermal resistances for unventilated airspaces^
Thermal resistance / (m2 K/W) for heat

Type of air space flow in stated direction
Surface
Thickness emissivity Horizontal Upward Downward
5 mm High 0.10 0.10 0.10

Low 0.18 0.18 0.18
25 mm High 0.18 0.17 0.22
or more Low 0.35 0.35 1.06
High emissivity plane and 0.09 0.09 0.11
corrugated sheets in contact
Low emissivity multiple foil 0.62 0.62 1.76
insulation with airspace on
one side

Table A 2.5. Determination of the thermal transmittance (U-value) of a

wall
Thermal Thermal
Thickness conductivity resistance
Element (m) (W/(m K)) (m2 K/W)
Rsi = inner surface resistance — — 0.12
R 2 = plaster 0.013 0.16 0.08
R2 = concrete blocks 0.100 0.19 0.53
R3 = polystyrene 0.050 0.035 1.43
R4 = brickwork 0.105 0.84 0.13
Rso = outer surface resistance 0.06
i? = Total resistance = 2.35 m2 K/W
U= l/i? = 0.43W/(m2 K)
Fig. A2.1. Typical wall

construction.
Normal
exposure
r t \
50 mm 100 mm 13 mm
105 mm cavity
brickwork lightweight lightweight
filled concrete plaster
with blocks
350 Appendix 2
Table A 2.6. U-values and admittance (Y-) values for construction
elements^
Admittance
Construction [/-value 7-value
(outside to inside) (W/(m2 K)) (W/(m2 K))
(l)WaU:
105 mm brickwork, 50 mm mineral 0.53 4.3
fibre, 100 mm heavyweight concrete
block, 13 mm lightweight plaster
(2) Wall:
105 mm brickwork, 75 mm mineral 0.37 3.7
fibre, 105 mm brickwork, 13 mm
lightweight plaster
(3) Roof:
10 mm tile, loft space, 100 mm glass- 0.34 0.7
fibre quilt, 10 mm plasterboard

Table A 2.7. U-values for glazing (without frames)^
[/-valuea for stated exposure

(W/(m2 K))
Normal
Construction Sheltered (standard) Severe
Single window glazing 5.0 5.6 6.7
Double window glazing
with airspace
25 mm or more 2.8 2.9 3.2
12 mm 2.8 3.0 3.3
6 mm 3.2 3.4 3.8
3 mm 3.6 4.0 4.4
Triple window glazing
with airspace
25 mm or more 1.9 2.0 2.1
12 mm 2.0 2.1 2.2
6 mm 2.3 2.5 2.6
3 mm 2.8 3.0 3.3
Roof glazing skylight 5.7 6.6 7.9
Horizontal laylight with
skylight or lantern light over
Ventilated 3.5 3.8 4.2
Unventilated 2.8 3.0 3.3
a
In calculating these values, the thermal resistance of the glass has been
ignored.
1
Table A 2.8. U-valuesfor typical windows^ ^
Fraction of [/-value for stated exposure

area occupied (W/(m2 K))
by framea
Window type (%) Sheltered Normal Severe
Single glazing
Wood frame 10 4.7 5.3 6.3
20 4.5 5.0 5.9
30 4.2 4.7 5.5
Aluminium frame
(no thermal break) 10 5.3 6.0 7.1
20 5.6 6.4 7.5
30 5.9 6.7 7.9
Aluminium frame
(with thermal break) 10 5.1 5.7 6.7
20 5.2 5.8 6.8
30 5.2 5.8 6.8
Double glazing
Wood frame 10 2.8 3.0 3.2
20 2.7 2.9 3.2
30 2.7 2.9 3.1
Aluminium frame
(no thermal break) 10 3.3 3.6 4.1
20 3.9 4.3 4.8
30 4.4 4.9 5.6
Aluminium frame
(with thermal break) 10 3.1 3.3 3.7
20 3.4 3.7 4.0
30 3.7 4.0 4.4
a
Where the proportion of the frame differs appreciably from the above
values, particularly with wood or plastic, the [/-value should be calculated
(metal members have a [/-value similar to that of glass).
Fig. A2.2. Structure for
heat-loss calculations.
2.5 m
5.0 m
2.5 m
352 Appendix 2
Table A 2.9. Determination of the fabric heat loss for a simple structure
Temperature
{/-value Area difference a Fabric heat loss
Element (W/(m2 K)) (m 2 ) (K) (W)
South wall6 0.43 11.9 20 102

South window 2.9 0.6 20 35
East wall 0.43 12.5 20 108
North wall 0.43 12.5 20 108
West wall 0.43 12.5 20 108
RooF 0.34 25.0 20 170
Floor** 0.84 25.0 20 420
Sum of fabric heat loss 1051 W
a
Internal temperature 19°C; for simplicity the external temperature of
both the air and ground has been taken as — 1 °C.
b
Wall construction: 105 mm brickwork, 50 mm polystyrene, 100 mm con-
crete block, 13 mm plaster; unheated loft space therefore use 12.5 m2
area for east and west wall.
c
Roof construction: 10 mm tile, loft space, 100 mm glass fibre quilt,
10 mm plaster board;use plan area for calculation.
d
Floor construction: suspended timber floor with carpet.
(The factor of 0.36 is the volumetric specific heat of air, 1300 J/m 3 K,
divided by 3600, the number of seconds in an hour.) Alternatively, for
situations of high occupancy, the ventilation heat loss can be calculated
from
Cv (W) = Number of people X Air-change rate per person (m 3 /s)
X Volumetric specific heat of air X (t-x —to).
In the case of the simple structure of Fig. A 2.2, if we assume one-and-a-
half changes per hour the ventilation heat loss is 675 W, giving a total
design heat loss of 1.73 kW. (The ventilation heat loss can be estimated
more precisely using the crack estimation method.I21) Note that this cal-
culation makes no allowance for 'free-heat' gains, which is reasonable
since, in sizing the heating system, it is best to be cautious and not rely
on such a variable factor.
Of course, most buildings are considerably more complex, with several
floors, partitions and variable ventilation rates in, for example, kitchen and
living rooms. Burberry!3' 41 provides more-detailed computation pro-
cedures. Additional information on ^/-values may be found in a BRE pub-
lication!5! and for the whole subject of energy demand and thermal per-
formance, the CIBS GuideW and its US counterpart, the ASHRAE Hand-
book^1rl are invaluable.
To calculate the seasonal heat loss, the simple method devised by the
BRE for small houses shown in Table A 2.10 may be used. For the pur-
poses of this exercise we will assume one occupant but use figures for the
heat gains from electricity, cooking and water heating, which are in accord
with the greater values given in Chapter 3 for groups of several people.
Table A 2.10. Seasonal heat requirements (after Ref.
Heat losses
(1) Fabric heat losses

Element x Area x Seasonal x Average x Rate of
transmittance temperature heat loss
coefficient difference a
(m 2 ) (W/(m2 K)) (K) (W)
Roof 25.0 0.34 10 85

Window 0.6 2.9 10 17
External walls 49.4 0.43 10 212
Ground floor 25.0 0.84 10 210
Total 524
(2) Ventilation heat loss
V
= 0.36 x 1.5 x 62.5 m3 x 10 K
= 338W
(3) Total heat loss = 862 W x 0.02 b
= 17.2GJ
Heat gains
(1) Windows
South facing 0.6 m2 x 0.68 c GJ/m2 0.4 GJ
(2) Body heat (1 GJ/p) 1.0 GJ
(3) Electricity (based on estimated consumption) 3.3 GJ
(4) Cooking (gas) 6.0 GJ
(5) Water heating 2.0 GJ
Total seasonal gains 12.7 GJ
Net heat requirement 17.2 - 12.7 = 4.5 GJ

Gross heat requirement - Net heat requirement x 100 /(% efficiency of
heating system)
a
An average temperature difference of 10 K was found to adequately rep-
resent the actual variation in the difference between internal and external
temperature during the heating season. (Other sources sometimes use
varying assumptions. For example, in Fig. 3.1 a 7 K difference was used.)
b
0.02 is the length of the heating season, 20 x 106 s, divided by 109 to
convert the heat loss to GJ.
c
Heat gains from unobstructed east- and west-facing windows are given as
0.41 GJ/m2 and north-facing windows as 0.25 MJ/m2.
Obviously, the method can easily incorporate more-sophisticated refine-

ments. For example, in more-recent housing the average air-change rate
during the heating season is not as high as the figure of two originally
suggested by the BREl8! or even the one-and-a-half used above (see Chap-
ter 3) but a better estimate is easily substituted. Similarly, if temperature
354 Appendix 2
data is available, the calculation can be made monthly and summed over
the heating season and if solar radiation data is available a more precise
figure for solar gains can be used.
Seasonal heating requirements can also be estimated using the degree-
day methodJ 4 ' 9,10] ^ degree day, according to the 'generally accepted'
definition^9] is the product of one day (24 h) and 1 K difference of tem-
perature between the base temperature and the average outside air tem-
perature, (t m a x + t m i n )/2. The base temperature is 15.5 °C which allows
for a 2.8 °C lift due to incidental gains from such sources as people, lights,
cooking appliances and motors to bring the internal temperature up to
18.3 °C which is considered comfortable for normal domestic purposes.
Maps of degree-day areas and 20-year mean degree-day figures are avail-
able. I91 For example, for the period 1957—77 in the Thames Valley area
(which includes London) the average number of degree days from
September to May was 2030. Returning to the simple structure of Fig.
A 2.2 we can calculate that the total heat loss is 87 W/K. (Note that we
haven't considered heat gains since we will accept the 2.8 °C figure as
standard. In practice, however, this must be carefully checked because in
highly insulated buildings casual gains may keep the building sufficiently
warm down to outside temperatures of, say, 12-13 °C rather than the
15.5 °C base temperature.) The seasonal heat loss, assuming that the
building is kept at 18.3 °C throughout the heating season, is thus approxi-
mately
87 ^ X 24 T ^ - X 2030 °C - d a yJ = 4238 kWh = 15.3 GJ.

K day
The difference between this figure and the net heat requirement in Table
2.10 is due principally to the use of the 2.8 °C lift figure. In such a small
well-insulated structure the inclusion of large incidental gains would in fact
considerably increase the lift and consequently lower the effective number
of degree days.
Refinements in the method can easily be introduced - Page is develop-
ing a program which incorporates the concept of day and night degree
days so that the effects of curtaining windows at night can be properly
evaluated.^111
The degree-day method is perhaps of greatest use in evaluating the
efficiency of energy-conservation measures. For this application actual fuel
consumption before and after, for example, the installation of insulation
can be compared and by dividing by the observed number of degree days
for an equal period before and after, the effect of weather can be separ-
ated from that of the insulation. Steady-state heat loss calculations can be
carried out quite simply on advanced calculators and minicomputers. The
Royal Institute of British Architects (RIBA) has a number of such pro-
grams ranging from the design day heat loss for a building design to a
calculation which derives the heat loss and solar gain from monthly
weather data and then balances losses against solar and internal gains to
assess the need for additional space heating.!121
More-sophisticated approaches to thermal performance exist. In the UK
Table A 2.11. Admittance ( Y-values) for internal walls^11
Admittance (7-value)
Construction (W/(m2 K))
(1) 105 mm brickwork 4.1

(2) 105 mm brickwork with 13 mm lightweight 3.6
plaster on each side
(3) 105 mm brickwork, 25 mm air gap, 105 mm 4.9
brickwork
(4) 105 mm brickwork, 25 mm air gap, 105 mm 3.8
brickwork with 13 mm lightweight plaster
on each side
(5) 100 mm lightweight concrete block 2.0
(6) 100 mm lightweight concrete block with 2.3
13 mm lightweight plaster on each side
the increasing importance of summertime design for the prediction of air-

conditioning loads and internal temperatures led to the development of
the admittance procedure.! 13' 14] This technique has also proved useful
for investigating the performance of intermittently heated buildings and,
on a smaller scale, the comparison of different internal surfaces. This last
aspect, which is the only one that space allows to be even briefly treated
here, depends on a knowledge of the admittance of the surface. This is the
rate of heat flow between the internal surface of the construction and the
space temperature for each degree of swing in space temperature about its
mean value. The admittance is a function of the thickness, thermal con-
ductivity, density and specific heat capacity of the materials used in the
construction and of the frequency of the thermal input (usually a 24 h
cycle is assumed). For thin structures the admittance is equal to the
U-value, thus, for example, the admittance of single glazing is
5.6 W/(m2 K)J 15 1 With multilayer constructions the admittance is deter-
mined primarily by the characteristics of the materials in the layers next to
the internal surface. Table A 2.6 has shown the admittance of several
external wall constructions. Table A 2.11 gives figures for some internal
walls.
If one is designing a lecture room which is used intermittently and
heated with, say, electric fires, if the internal walls are of 105 mm brick,
each 1 K rise will require 4.1 W per square metre of wall. Change to
100 mm lightweight concrete block and the figure drops to 2.0 - other
things being equal, the audience will be more comfortable more quickly in
the second case.
Another approach is the response factor method whose basis has been
described^16! as follows:
The heat flux through the surfaces of a wall can be expressed in terms of
356 Appendix 2
the surface temperature and thermal characteristics of the wall using time
series techniques. Thus, all unknown heat fluxes at the surface can be
related to the surface temperatures and, if the temperature history of the
surface is known, the heat balance equations can be solved to find the
current value of surface temperature. When the heat transfer coefficient
depends on the surface temperature it is necessary to solve the equations
by an iterative procedure. Once the surface temperature has been found it
becomes part of the temperature history to be used for future calculations.
While the mathematics of this approach are considerably more difficult
and often beyond the range of all but specialists in heat transfer, the avail-
ability of computer programs^17! that can incorporate it should lead to
increased acceptance. As Kusuda said in the discussion following Mitalas'
paperj 1 6 ! such advanced techniques are needed to deal with the 'hour by
hour evaluation of instantaneous heat gain and heat loss, responding to the
changing climatic conditions and energy use schedules'.
Electrical analogue techniques which model thermal response also exist
but, in general, advances in digital computers have favoured methods
developed for them. Complex thermal prediction models are now common
in consulting engineers' firms and the research centres of the energy supply
industries. Further computer advances should soon make them available to
many architects' offices.
Among the more complicated models are those of Nottingham Univer-
sity (the BUILD program), the University of Strathclyde (the ESP pro-
gram) and Ecotope (Seattle, US) (the SUNCODE program). These could
well be used in the final design stages of a house. Even more elaborate pro-
grams are available, in particular for commercial buildings but they tend to
be difficult to use and too expensive to run repeatedly for design purposes.
References
[ 1 ] (1980). Thermal properties of building structures'. CIBS Guide.
Section A3. London: Chartered Institution of Building Services.
[2] (1976). 'Air infiltration'. CIBS Guide. Section A4. London:
Chartered Institution of Building Services.
[3] Burberry, P. (1975). Environment and Services. London: Batsford.
[4] Burberry, P. (1979). 'Predictions of thermal performance'. Archi-
tects Journal, 172 (43, 44, 45, 46), 893-905, 951-62, 993-4,
1055-62.
[5] Anon. (1975). Standard U-values. BRE Digest 108. Garston,
Watford: BRE.
[6] CIBS Guide. London: Chartered Institution of Building Services.
The guide is continually updated.
[7] ASHRAEHandbook, 1977, Fundamentals. New York: American
Society of Heating Refrigerating and Air Conditioning Engineers.
[ 8 ] Anon .(1956). Do m estic Hea ting - Estimatio n of Seaso nal Hea t
Requirement and Fuel Consumption in Houses. BRE Digest 94.
Garston, Watford: BRE.
[9] Anon, (undated). Degree Days. Department of Energy Fuel
Efficiency Booklet 7. London: HMSO.
References 357
[10] Peach, J. (1977). 'Degree days'. Building Services Engineer, 10
(44), 213-17.
[11] Page, J.K. (1978). 'The development of systematic climatological
design procedures for solar buildings'. Sun at Work in Britain, 6,
48-50.
[12] Royal Institute of British Architects Publications Ltd, London.
[13] Loudon, A.G. (1968). Summertime Temperatures in Buildings.
BRS Current Paper 47/68. BRE: Garston, Watford.
[14] Millbank, N.O. & Harrington-Lynn, J. (1974). Thermal Response
and the Admittance Procedure. BRS Current Paper 61/74. BRE:
Garston, Watford.
[15] (1976). IHVE Guide Book A. London: Institute of Heating and
Ventilating Engineers.
[16] Mitalas, G.P. (1968). Calculation of Transient Heat Flow Through
Walls and Roof. New York: ASHRAE.
[17] Kusuda, T. National Bureau of Standards. Washington, DC.
Appendix 3:
Interstitial condensation
Just as a temperature gradient exists within a structural element, a dew-

point gradient depending on the water vapour diffusion properties of the
element exists too. If at any point in the structure the actual temperature
is below the dew point then condensation will occur at that point.
Table A 3.1 gives some typical values of vapour resistance and thermal
and vapour resistivities (thermal resistivity is the reciprocal of thermal
conductivity — see Table A 2.3).
With more and more insulation being used, the designer must remember
to consider both the thermal and vapour properties he or she is specifying.
This is particularly true since some very good thermal insulants, for
example glass fibre, are also very permeable to water vapour.
Let us now return in Fig. A 3.1 to the wall construction of Fig. A 2.1
and using the standard BRE procedure!* 1 assess whether there is a risk of
interstitial condensation.
Procedure
(A) Assumed interior/exterior air temperature difference, AT, is 20° —0°
= 20 °C.
(B) Thermal resistance, rt (thermal resistance equals the thermal resistivity
times the thickness of the material).
(a) Inside air to point 1 (from Table A 2.1) = 0.12 m2 K/W;
(b) Point 1-213 mm lightweight plaster
7 - ^ X 0 . 0 1 3 = 0.08;
u.lo
(c) Point 2 - 3 100 mm lightweight concrete block
^ X 0.100 = 0.53;
(d) Point 3-4 50 mm expanded polystyrene
1
0.035 X 0.050 =1.43;
(e) Point 4 - 5 105 mm brickwork
1 X 0.105 =0.13;
0.84
(0 Point 5 to outside air (from Table A 2.2) = 0.06
Rt = 2>t = 2.35
358
Interstitial condensation 359
1
Table A 3.1. Selected thermal and vapour properties of materials^ 1
Vapour resistance (rv)

(MN s/g m)
Membranes
Average gloss paint film 7.5-40
Polythene sheet (0.06 mm) 110-120
Aluminium foil 4000
Thermal Vapour
resistivity resistivitya
(m K/W) (MN s/g m)
Materials
Brickwork 0.7-1.4 25-100
Concrete6 0.7 30-100
Rendering 0.8 100
Plaster 2 60
Timber 7 45-75
Plywood 7 1500-6000
Fibre building board 15-19 15-60
Hardboard 7 450-750
Plasterboard 6 45-60
Compressed strawboard 10-12 45-75
Wood—wool slab 9 15-40
Expanded polystyrene 30 100-600
Foamed urea-formaldehyde 26 20-30
Foamed polyurethane 40-50 30-1000
(open or closed cell)
Expanded ebonite 34 11000-60 000
Glass fibre 29 6C
a
Resistivity = 1/diffusivity.
b
An approximate figure (manufacturers' data depend on material density,
etc.) for the vapour resistivity of lightweight concrete block is 50 MN s/g.
c
Manufacturers' data. For glass fibre the vapour resistivity is typically
taken as that of fresh air.
105 mm 50 mm cavity filled 100 mm lightweight Fig. A3.1. Wall construction

brickwork with polystyrene concrete blocks with temperature plots.
- 13 mm lightweight
(a) Internal moisture contri-
1 ///
plaster bution 3.4 g/kg. (b) Internal
20 p
<//
20
U
i / / , moisture contribution
6.4 g/kg.
15 ^^ - Structural
15 "^Structural
10
1 10
Dewpoint
5 x
Dewpoint 5
/
^7
7
0 0 — *
Outs de Inside
4 3 21 5 / 4 3 21
(a) Zone of possible interstitial
condensation
(b)
360 Appendix 3
(C) Temperature drops between points At (°C) =
X r
- 235 * ~ 8'5 X
'«
Initial Temperature Resulting

temperature drop temperature
(a) Inside air 20.0

(b) Inside air to point 1 = 8.5 X 0.12=* 20.0 1.0 19.0
(c) Point 1-2 8.5 X 0.08 =* 19.0 0.7 18.3
(d) Point 2-3 8.5 X 0.53=* 18.3 4.5 13.8
(e) Point 3-4 8.5 X 1.43=* 13.8 12.2 1.6
(f) Point 4-5 8.5 X 0.13=* 1.6 1.1 0.5
(g) Point 5 to outside air = 8.5 X 0.06=* 0.5 0.5 0.0
(h) Outside air 0.0
(D) Plot the temperature profile as shown in Fig. A 3.1.

(E) Assume that the outside air is at 0 °C and is saturated at a mixing ratio
of 3.8 g/kg (see Fig. A 3.2). A mixing ratio of 3.8 g/kg corresponds to
an outdoor vapour pressure of 6 mb (read on the extreme right-hand
scale of Fig. A 3.2). The BRE states that at normal ventilation rates a
person not engaged in physical exertion adds roughly 45 g/h of mois-
ture to the air and that this results in an excess moisture content over
outdoor air of some 1.7 g of water vapour per kilogramme of dry air.
For dwellings they suggest a safer design value of 3.4 g/kg to allow for
the moisture gains from cooking and bathing (see Chapter 3 for
moisture emission rates) and the likelihood of restricted ventilation in
cold weather. We thus have a total moisture content of 7.2 g/kg which
corresponds to a vapour pressure of 11.4 mb. The indoor-outdoor
vapour pressure difference, AP, is therefore 11.4 — 6.0 = 5.4 mb.
(F) Vapour resistances, rv, equal the vapour resistivities (from Table
A 3.1) times the thickness of the component.
(a) Point 1-2 60 X 0.013 = 0.78 MN s/g

(b) Point 2-3 50 x 0.100 = 5.00 MN s/g
(c) Point 3-4 350 x 0.050 = 17.50 MN s/g
(d) Point 4-5 25x0.105 = 2.63 MN s/g
# v = 2r v = 25.91 MNs/g
(G) Pressure drops between points, Ap (mb):

AP
Ap = = • X ry
RV
Interstitial condensation 361
Fig. A3.2. Chart relating
moisture contents and
temperatures!11 (see Chapter
3 for an explanation of the
lettered points). (BRE,
Crown Copyright, HMSO.)
dry bulbtemperature °C
Corresponding
dew-point
Initial Vapour Resultant temperature
vapour pressure vapour from
pressure drop pressure Fig. A 3.2
(mb) (mb) (mb) (°C)
(a) Inside air 11.4 9.4
(b) Point 1-2 0.21 X 0.78 => 11.4 0.2 11.2 8.8
(c) Point 2-3 0.21 X 5.0 => 11.2 1.1 10.1 7.2
(d) Point 3-4 0.21 X 17.5 => 10.1 3.7 6.4 0.9
(e) Point 4-5 0.21 X 2.63=^ 6.4 0.6 5.8fl 0.0fl
fl
The difference between 6.0 and 5.8 is due to approximation errors. The dewpoint
temperature corresponding to 6.0 mb has been used.
362 Appendix 3
(H) Plot the dew-point temperature profile as shown in Fig. A 3A(a).
(I) Estimation of condensation risk: condensation can occur in the con-
ditions assumed at any point where the structural temperature is lower
than the dew-point temperature. Since this does not occur anywhere
for the case studied there is no computed risk of condensation.
If, however, we were to approximately double the moisture contri-
bution from internal activities to 6.4 g/kg and repeat the procedure we
would have the graph of Fig. A 3.1(Z>). In this case there is a risk of inter-
stitial condensation in the zone indicated. One possible way of reducing
this risk would be to use plasterboard with an aluminium-foil backing as
the internal surface, taking care to seal the joints of the plasterboard sheets.
Similar calculations can be made using a number of computer routines
including the RIBA programs!21 mentioned in Appendix 2.
In all cases, it must be remembered that the initial assumptions are both
quite important and subject to considerable uncertainty, particularly with
regard to internal moisture conditions.
References
[ 1 ] Anon. (1969). Condensation. BRE Digest 110. Garston: Building
Research Station.
[2] Royal Institute of British Architects Publications Ltd, London.
Index
AC/DC supply, 282 Glyn England House, 319

active solar heating, 6-7,159, 203-4, 232, Hay House, 144
234 HDD Better Insulated Houses, 78
active solar heating - collection angle, Lewisham flats, 46, 68, 234-5
89-91 Lioux House, 250-1
admittance, 44, 47, 347, 350, 355 Locksheath School, 388
adobe brick, 127,134 Milton Keynes Houses, 175,180,191,
agricultural buildings, 329-30 221,248
air flow in Trombe walls, 139 Minimum Energy Dwelling,
air movement in rooms, 52 California, 305
antifreeze, 176 MIT Solar 5 House, 116,120,
appliance energy, 304 125 et seq.
atria, 37-8, 336 National Coal Board House, 202-3
Newnham House, 222, 315-16
batteries - electric, 281 Odeillo Houses, 139,140
Beadwall, 142 Ourobouros House, 65-6
berms,35,65,127 Pennyland, 18-19
BLAST, 151 Peterborough Houses, 167, 250, 310
blinds, 109-12, 126 Roach Vale Primary School, 333
boilers, 194,198-201, 203-6, 221, 230-1, Romsey School, 331
234, 329 Salford Strawberry Hill Houses,
Buildings 216-17, 223, 233
Ambient Energy Design House, 70, Sheiling School,330
71,203,205 Solar One House, 245
Ark Project, 255 St. Johns School, Clacton on Sea, 337
Autarkic House, 6, 72, 82,128,132, Termoroc House, 114
241, 247, 278, 289-90, 310,312 Terry House, 117,127
Autarkic House Turbine, 277 Tyrell House, 142
Baer House, 142 Vale House, 200, 201
Balcomb House, 134 etseq. Wallassey School, 98,124
Basildon Northlands 1 House, 20, 29, Yately Newlands School, 334
59, 63-4, 78, 234 Building Regulations, 42, 72, 84
Bebington Houses, 143 buildings - heavyweight, 43, 46-7, 217
BRE Solar Energy Houses, 216,232-3 buildings - lightweight, 43, 46-7, 331
BRS Office Building, 328-9 buildings - thermal response, 42-7, 77,
Conservation House, Wales Turbine, 194
284
Curtis House, 124 clothing, 4, 51
Danish Zero Energy House, 115, coatings on glass, 107,126
223-4, 248 coefficient of performance of turbines,
Denver Solar House, 162, 250 229-30, 232, 234
Dover Massachusetts House, 254 cold bridges, 72, 73, 75, 81
Ecology Houses, 35-6 collectors - area solar hot water, 175
Elmstead Primary School, 332 collectors - evacuated tubes, 177
Energy Design Group Solar Skin, 310, collectors-panels, 173
312 collectors - plastic, 177
Feilden and Clegg Farmhouse, 318 collectors - refrigerant charged loop, 177
First Village House, 134 et seq. collectors - surface temperatures, 312
Frusher Polyurethane Home, 83 collectors - swim pool, 326
363
364 Index
collectors - thermal expansion, 312 glazing - heat mirrors, 107,126
comfort, 47, 48, 86, 331, 333 glazing-infiltration, 108
computer models, see: ESP, BLAST, glazing - Kalwall plastic glazing, 103
DEROB, FCHART, SCRIBE, glazing - plastic, 102-4
SERIRES, SUNCODE glazing - properties of glass, 101
condensation, 53, 56-60, 76-7, 81, 88, glazing-sealants, 102
308, 358-62 glazing - U values, 103 et seq.
condensation, see also moisture emission ground coupling, see also heat storage,
conservatories, 129 etseq., 200, 238 126
conservatory at Peterborough, 315 ground reflection, 117
conservatory retrofit, 318
contaminants in air, 53, 54, 74 heat centre, 212-13
control of air solar systems, 165 heat exchangers, 163, 175
control of solar hot water systems, 177 heat mirrors, 107, 126
controls-heating, 193,198, 201, 204, 208, heat pumps, 91,126, 216, 220, 228-34,
212-15, 220, 354 241,258,326,333,338
convective loops, 147 heat recovery, 59, 220-1, 317, 329, 333
convectors, 207, 210-11, 216,221-2, 329 heat storage, see thermal storage
cooking, 307 heat transfer - air movement, 133
cooling, 145 heat transfer - air to water, 314
heat transfer - via doors, 318
dampers, 163, 314 heat transfer - wall openings, 133,139
daylight, 330-1, 335 heating systems - ceiling, 49, 227
degree days, 127,354 heating systems - efficiency, 192-3
DEROB, 150 heating systems - electrical, 225-8
design methods - passive solar, 148 etseq. heating systems - forced air, 217-24, 311
design methods - solar hot water systems, heating systems - solid fuel, 70,196-205,
179 et seq. 324
diffuse solar data, 345 etseq. heating systems - temperature response,
domestic appliances, 304 195-6, 218
domestic hot water supply, 144, 211, 233, heating systems - underfloor, 211, 216,
244 227,236
drain back systems, 171,176 heating systems - water distribution,
drain down systems, 172,176 205-17, 324
ducts, 147,162 hot water system installed solar example,
179
economics of retrofit, 319 etseq. hot water usage, 305
efficiencies of collectors, 174 housing density, 3, 18, 21
electrical generators, 278
energy - consumption, 2,4, 330 infiltration at windows, 108
energy-delivered, 191-2, 330, 338 infiltration, see also ventilation, 24, 26,
energy-primary, 2,191, 222,. 230, 328, 53,55-6,68,75,86
338-40 insulation, 7, 58, 74-91, 203, 226, 315,
energy-useful, 191-2 324,331
energy collected in solar systems, 160
energy in the wind, 268 Kalwall plastic glazing, see glazing
energy saving in appliances, 304
ESP, 150 layout, 67-8
external shutters, 114, 328 lighting, 69, 329, 331
lights-heat gain, 123
F-chart4, 182
fan coil unit, 204, 223-4 mass, see heat storage
fans for air collectors, 162 mass walls, 137 etseq.
form, 60-5, 315, 328 meteorological data, 16-24, 130, 242,
free heat gains, 43-5, 70, 352-3 342-4
furnishings, thermal mass, 123 moisture emission,58-9, 360
glare, 99,102,127 occupancy patterns, 4, 45-6, 77, 80,189,
glazing, 100,331,333 237
glazing - coatings, 107 open loop system, 172,176
glazing-multiple, 84-7,103,317,328,350 orientation, 19, 64, 311
glazing - effect of dust, 100 overshadowing, 20, 68, 116
glazing - effect of wind chill, 106
Index 365
passive solar gain, 18, 46, 63-6, 70, temperature, dry resultant, 48-50
9A etseq., 123 etseq., 195,216, temperature, mean radiant, 48-9
237, 315, 323 thermal performance, 346-56
pebble beds, 120 et seq. ,133et seq., thermal storage, 33, 35, 99, 118 et seq.,
163 et seq. 241-58
plastic glazing, 102-4 thermal storage - adsorbent beds, 257
pool covers, 326 thermal storage - annual cycle energy
pump, 189, 206, 211-12, 235 system, 255-6
thermal storage - bricks and adobe, 127,
radiators, 207-8, 210-11, 234 134, 223
reflectors, 117,141 thermal storage - earth, 255
refrigerators and freezers, 307 thermal storage - effect of furnishings, 123
rehabilitation of buildings, 317, 319 thermal storage - for wind turbines, 283
response factor method, 355-6 thermal storage - ground coupling, 126
rock bed storage, 120 et seq., 133 et seq., thermal storage - Peterborough houses,
163 etseq. 311
roof ponds, 144 thermal storage-phase change, 119,127,
roof space hybrid collector, 146, 319 252-5
roofs, 88-91 thermal storage - photovoltaic, 242
room air movement, 52 thermal storage - rock and pebble beds,
120 et seq., 133 et seq., 163 et seq.,
Savonius Rotor, 272, 276 204, 249-51
school buildings, see buildings thermal storage - solar ponds, 255, 257
SCRIBE, 321 thermal storage - structural, 122
SERI-RES, 151 thermal storage - water, 128, 137,141
shutters, external, 114 thermal storage for solar hot water, 175,
shutters, internal, 109,112 182
simulation models, see ESP. BLAST, thermal storage walls, 137 et seq.
DEROB, FCHART, SCRIBE, thermosiphoning, 147, 160, 168
SERIRES, SUNCODE transfer fluids, 176
site planning, 14-39 transport energy use, 308
site planning - Beaufort scale, 30 Trombe walls, 137 et seq.
site planning - glasshouses, 28
site planning - soil, 33-6 urethane foam, 127
site planning - underground structures,
35-6 vapour check, 76, 77, 80, 84, 88
site planning - water table, 33 vegetation and plants, 21-3, 26-7, 30, 32,
site planning - wind, 24-33 36,91,129
site planning - wind turbine, 31-2 Venetian blinds, 116
site planning - windbreaks, 25-30 ventilation, 53-7, 77
skirting heating, 211 ventilation air preheat, 314
solar air heating, 160 etseq. ventilation control, 326
solar hot water - anti-freeze, 176 ventilation, mechanical, 54-5,69,217-18,
solar hot water - collector area, 175 329,331
solar hot water - control, 177
solar hot water - design method F-chart, wall insulation - retrofit, 317
182 waste disposal systems, 289, 293-301
solar hot water - economics, 161 waste disposal systems - aerobic, 295
solar hot water - hard water effects, 176 waste disposal systems, methane digester,
solar hot water - heating, 159,161, 293, 295-6, 300-1
111 etseq. waste disposal systems, septic tanks, 295
solar hot water - storage, 175,182 water heating energy, see also hot water,
solar hot water - stratification, 176 305
solar hot water — system size, 175 water supply systems, 289-92
solar hot water - transfer fluids, 176 wave power, 262
stack effect, 25,54,69 wind - gust ratio, 266
storage, see thermal storage wind - hourly speeds in the UK, 266
storage fan heaters, 225 wind - power in the, 242, 264
storage radiators, 225-6 wind - variation with height, 267
stratification, 176 wind - variation with location, 264
structural thermal storage, 122 wind energy, 262 et seq.
SUNCODE, 151 wind systems, real outputs, 279
sunspaces, 128 etseq. wind turbines - Autarkic House, 277
366 Index
wind turbines - Conservation House, wind turbines - rated speed, 268
Wales, 284 wind turbines - types of rotor, 271
wind turbines - cut in speed, 268 window insulation-Beadwall, 142
wind turbines - Darrieus, 277 window insulation - blinds, 109-12, 126
wind turbines - Wind Energy Group, window insulation - external shutters ,114
3 MW, 263 window insulation - internal shutters,
wind turbines - gears, 280 109,112
wind turbines - Musgrove, 277 window insulation - movable, 109 et seq.
wind turbines - NASA 100 kW, 274 window insulation - Venetian blinds, 116
wind turbines - power coefficient, 268, windows, 63, 351
270 wood stoves, 128
wind turbines - power extracted by, 268

Design With Energy - The Conservation and Use of Energy in Buildings

Uploaded by

Copyright:

Available Formats

Design With Energy - The Conservation and Use of Energy in Buildings

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Design With Energy - The Conservation and Use of Energy in Buildings

Uploaded by

Copyright:

Available Formats

CAMBRIDGE URBAN AND ARCHITECTURAL STUDIES

VOLUMES IN THIS SERIES

The right of the

CAMBRIDGE UNIVERSITY PRESS

© Cambridge University Press 1984

First published 1984

Library of Congress catalogue card number: 83—7537

British Library cataloguing in publication data

Transferred to digital printing 2003

Units, symbols, abbreviations, conventions and conversion

1 Energy and buildings 1

2 Site planning and analysis 14

4 Passive solar design 97

5 Active solar heating 159

5.7 Predictive methods 182

6 Space heating and ventilation 189

7 Thermal storage 241

8 Wind energy 262

9 Water-supply systems 289

10 Waste disposal and utilization 293

11 Domestic-energy saving 304

12 Housing case studies 310

13 Non-domestic case studies 326

Appendix 1: Weather data 342

Appendix 2: Thermal performance 347

Appendix 3: Interstitial condensation 358

(1) Principal units

(5) Conversion factors

Energy, work, heat

The authors have rendered a valuable service to the building industry by

JOHN I. YELLOTT, M.B.E., P.E.

Energy and buildings

1.2 Energy and the built environment - past and present

Total primary energy

1970 1980 1990 2000 2010 2020

stock is semidetached houses, 31%; terraced houses, 30%; detached houses,

Delivered energy demand

(1) Commercial glass houses'* 2500

An approximate idea of the energy consumption in a range of building

1.3 Energy and the built environment - the future

Saving per item Saving per group

§ Improve layout and orientation

7 Use shutters on windows !5^T5$$$$$^$^^

1 Reduce int. temp. 18-15.5 °C Jo]

C Increase efficiency of plant

D Use most efficient fuel Assuming a cumulative saving in A, B, C of 75%

© Coal conversion efficiency 50% = 50 units total

Fig. 1.5. Aspects of energy

Measure ]Break-even 9% dividend

Roof insulation 50 mm DIY :5.8 1.2

Notes to Table 1.2

Thermal conduction —- Positive

2.2 Solar radiation

and, commercially, for air-conditioning loads. Appendix 1 provides data

Axis for azimuths

(c) Construction of a shading mask

Fig. 2.5. Final layout for

Density Number of Plot width* 10% 20%