1.

included with MSExcel
Workbooks

Muneer
eir

N. Abodahab
.

i. Kubie

WINDOWS IN BUILDINGS
THERMAL,ACOUSTIC,VISUAL AND SOLAR PERFORMANCE

4

4

I

/

,

ArhitccturaI
Press

I

Windows in Buildings

.

w

Bodies like crystal, houses made of glass Sunlight spread across, yet shadow not to be found Urdu verse by Baqar Naqvi

Without a glass palace

Life becomes a burden. Glass opens up a new age Brick building only does harm. Scheerbart

Windows in Buildings: Thermal, Acoustic, Visual and Solar Performance
(CD-ROM with MS-Excel workbooksincluded)

T. Muneer, N. Abodahab, G. Weir and J. Kubie
Napier University, Edinburgh a chapter on Solar Radiation andDaylight Data D. Kinghorn by
with

Architectural Press
OXFORD AUCKLAND BOSTONJOHANNESBURGMELBOURNENEW DELHI

Animprint of Butterworth-Heinemann Linacre House, Jordan Hill, Oxford 0X2 8DP

Architectural Press

225 Wildwood Avenue, Woburn, MA 01801-2041

-

A division of Reed Educational and Professional Publishing Ltd

A member of the Reed Elsevierplc group

First published 2000

© T Muneer, N Abodahab and J Kubie 2000
All rights reserved.No part of this publication may be reproduced in any material form (includingphotocopying or storing in any medium by electronicmeans and whether or not transientlyor incidentally to some other use of this publication) without the written permission of the copyright holder except in accordance with the provisionsofthe Copyright,Designsand Patents Act 1988or under the terms ofa licence issued by the Copyright LicensingAgencyLtd, 90 Tottenham Court Road, London, England W1P OLP. Applications for the copyright holder's written permissionto reproduce any part ofthis publication should be addressed to the publishers

British Library Cataloguingin PublicationData A catalogue record for this book is available from the British Library Library of CongressCataloguingin Publication Data A catalogue record for this book is available from the Library of Congress ISBN 0 7506 4209 2

Printed and boundin Great Britainby Biddles Ltd. www.Biddles.co.uk

LANTA
FOR EVERY Trfl.ETHAT PEIRLISH RcTThRWORTh-HEINEIMANN WE WILLPAY FOR DTCVTOPLANT AND CAREFOR ThEE. A

TREE

3 1.5 References THERMAL PROPERTIES OF WINDOWS Thermal comfort Thermal transmission of multiple-glazedwindows Free convection analysis for an enclosure Temperature stratification in windows Frequency of condensation occurrence on double-glazed windows 66 76 81 V . Arup R&D PREFACE ACKNOWLEDGEMENTS MICROSOFT EXCEL WORKBOOKS ON COMPACT DISK INSTRUCTIONS ON THE USE OF WORKBOOKS AND EXECUTABLE FILES 1 1.9 1.1 vii ix xi xiii xvii 1 2 2 3 1.6 1.1 3.4 3.4 1.3 3.2 The functionality of Excel — an example-basedtour References THE MICROSOFT EXCEL COMPUTING ENVIRONMENT 34 37 38 40 51 3 3.1 Excel — A user-friendlyenvironment for number crunching 2.12 INTRODUCTION Windows and life Windows and energy conservation Windows as daylight providers Therapeutical aspects of daylight Windows as solar energy providers Windows as sound insulators Window technology Condensation in buildings Life cycle assessment of multi-glazed windows Legislation and regulations Use of computers in window design Scope of the present book 4 5 5 6 10 13 13 14 14 17 18 19 2 2.10 1.CONTENTS FOREWORD Professor MichaelHolmes.2 3.5 1.8 1.7 1.2 1.11 1.

1 References APPENDIX: Web-based and other software material INDEX 249 .4 Impact assessment 7.2 Planning 7.3 Solar radiation transmission 4.6 Selectionof the 'optimum' window type References 202 203 206 222 226 231 232 235 235 235 237 242 245 SOLAR RADIATION AND DAYLIGHT DATA International daylight measurement programme 8.1 Aural comfort 6.8 5.6 5.3 The acoustic environment 6.1 Definition of life cycle assessment 7.2 Slope data 8.5 Noise transmission through glazing References 7 WINDOWS AND THEIR LIFE CYCLE 7.5 Improvement analysis 7.vi CONTENTS 4 WINDOWS AND SOLAR HEAT 85 85 4.1 90 114 122 138 143 143 145 147 150 155 158 175 180 180 182 185 185 186 189 191 194 199 201 5.2 Noise sources 6.1 Windows as solar energy providers 4.3 5.3 Inventory analysis 7.7 5.2 5.4 Health and productivity effects 6.3 Sky scan data 8 8.9 References WINDOWS AND DAYLIGHT Daylighting fundamentals Windows as daylight providers Design of daylighting systems Daylight factor Glare Availability of daylight Luminance transmission Optimisation of glazing area Innovative developments 6 ACOUSTIC PROPERTIES OF WINDOWS 6.4 5.5 5.2 Solar radiation availability 4.4 Passive solar buildings References 5 5.

The value is further enhanced in that the authors have taken a holistic approach that extends intothe issuesofsustainabilityandwhole life costs. architecturally it defines the character of the building. Modern glazing systems using selective coatings and cavities containing inert gases assist. which is seen as very desirableby many. minimizing energy consumption and providing glare-free daylighting.FOREWORD The windowis more than a link between the inside and outside of a building. These will be of great value to the student. Professor Muneer and his co-authors. Additionally some of the latest research in these areas is included. Professor Michael Holmes ARUP R&D . offers greatchallenges to the engineer in terms of satisfying building regulations. It can be expected that this book will become a much quoted reference source. to take full advantage of these systemsit is essential to understand howwindows work. together with the techniques necessary for the calculation of external irradiance and illuminance. have brought together significant quantities of data related to the overall performance of windows. However. who between themcover a wide range of expertise in the field. practising engineer and researcher. The transparent building. A novel feature of the book is that calculations are backed up by software tools operating within the environment ofthe Excel spreadsheet.

Page blank in original .

new algorithms and tools have to be provided whichincorporate the emergingknowledge-base.PREFACE Ifdaylight is the 'form giver' of a building. Properly selected glazing can save energy. It may be madeto blend with the desired building architecture. To quote E.help reduce the environmental impact on heating and lighting without compromising acoustical protection. while powerful.The book is written with a view to provide easy-to-usecharts. Almost every engineering professional has a spreadsheet loaded on his or her computer. students. does not allow the user the .' The simplicity of data entry and its manipulation afforded by spreadsheets makes it an ideal design and analysis tool. Dueto recent engineeringdevelopmentsit is now possible to design and build buildings with glass walls which keep the weather out and warmthin. Forster. safety or comfort. Modern spreadsheet software offer very powerful numerical and decision-makingfacilities. In effect the large-scale use of glass in modern buildings implies its 'rediscovery'.e. andhavegone a little farther down our particular path than we have yet gone ourselves. Another advantage offered by the spreadsheet medium is that eachformula is readable. The object of this book is to present the state of the art of high-efficiency glazing systems and their energetic and environmental impact on modern buildings. The target audience are the architects. Glazing systemslie at the heart ofthe window design and it is therefore logical to explore the interaction of windows and buildings. This was the reason for the production of this book. Glass offers a high degree of functionality. educators and researchers working in the area of energy efficiency in buildings. the spreadsheet cell. This enables the user to easily follow the algorithmic chain. tables and software for the design and selection of glazing system for their appropriate incorporation in modern buildings. builders and building designers. and fits in its appropriate position. the window is thedaylight provider. building services engineers. On the other hand. M. Consequently. i. This book exploits this situation to fill the gap between the ability to perform involved calculations and the desire to do so without the tedium of writing one's own software. commerciallyavailable software. 'The only books that influence us are those for which we are ready. Another contributing factorto the present development is the emerging use of computer spreadsheets for engineering computation.

neither the author nor the publisher warrants the performance or the results that may be obtained by using the enclosed suite of computer spreadsheets or programs. we hope. However. due to its very design. the accompanying programs are licensed on an 'as-is' basis. each dealing with a given computational aspect. 'just the beginning'. The reader may then be able to take the book onward to a different plane! The following program-copying policy is to be maintained. . 'No matter how much progress one makes. This point will be demonstrated in Chapter 2 where more sophisticated features of Microsoft Excel are introduced. thereis always the thrill ofjustbeginning. The necessaryrigour for understanding the physical principles and analysis is backed up with a number of computer workbooks. daylight. Each individual reader may keep backup copies as required. Although prepared with great care and subject to rigorous quality control checks. In many cases such commercial or shareware software. To quote Robert Goddard. Accordingly. In due course there is bound to be much more activity for this type of presentation where the boundary between the author and the reader starts to become blurred. spreadsheets require only a few minutes' training before the user is fully in charge of the situation. The purchaser or user of this book assumes their performance or fitness for any particular purpose. may take several weeks and months of training. solar heat transmission. The book covers the main design aspects of windows — thermal insulation. sound insulation and life cycle assessment.' The hard—soft combination of material presented here is.x PREFACE facility to examine or alter the computer code. further distribution of the programs constitutes a violation of the copyright agreement and hence is illegal. Owing to their widespread use and user-friendly layout. The authors welcomesuggestions for additions or improvements. Copies are limited to a one person/one machine basis.

Koga: Solar radiation and daylight data for Fukuoka. John Mardaijevic. In this respect acknowledgementis given herein to the relevant sources of information.J. Thepresent text was undoubtedly enriched by the inclusion ofmaterial from a number of sources. the names of authors are followed by their publications. Mike Holmes. Meteorological Office (UK): Solar radiation data. Tregenza: Solar radiation and daylight datafor Manchester and Sheffield. Yasuko Koga. Edinburgh. Mike Cash and Marie Milmore is particularly appreciated. Butterworth Architecture. Berkeley. Kerr MacGregor. Royal Greenwich Observatory: Data for solar declinationand equation of time.C.ACKNOWLEDGEMENTS The present is the culmination of research and review undertaken by the text authors during the pastdecade. Nordan (UK).E. Norton & Company. Stephen Selkowitz and Peter Tregenza. Boltz and G. P. Their contributions are gratefully acknowledged. Glasgow. John Fulwood. Lanarkshire. Baolei Han. and P. Pilkington plc. CRC Press. Helen Rose. The help extended by the publishers Neil Warnock-Smith. Carmody et al. FraunhoferInstitute for Solar Energy Systems. Bracknell and the Lawrence Berkeley Laboratory. Pritchard: Lighting. Geoff Levermore. The following individuals are thanked for their support: Daruish Arasteh. Alistair Gilchrist. Paul Littlefair. D. Interpane. W. Longman. G. CIBSE: Energy efficiency in buildings.W. Living Design. T. Strathclyde University. Johnson: Low-e glazing design guide.L. Tuve: CRC Handbookoftablesfor applied engineering science. Many organisations have either sponsored or actively supported this programme of work. H. Glasgow.: Residential windows. Van Nostrand Reinhold. Ken Butcher. The authors are grateful to George Pringle for his assistance in artwork. The Royal Academy ofEngineering.R. Baiham. J.D. . Ander: Daylighting. Littlefair: Solar radiation and daylight data for Garston. Germany. Nakamura and Y.E. Noteworthy among them are Napier University.the UK Meteorological Office. London. Pilkington plc: Glass in building. the Chartered Institute of Building Services Engineers. Freiburg. Merseyside. R. The authors would like to extend special thanks to their partners for the patience and forbearance provided throughout the length of this project.

Page blank in original .

dll]) Fr206. Ar. sunset and day-length 4 Windows and solar heat IChap4l . Xe or SF6 infill Calc3-03 Calc3-04 Calc3-05 Data3-O1 infill (approximate solution) Prog72.for Calc3-Ola Optimisation of window aspect and tilt for maximising solar energy capture (Solver) FORTRAN-based DLL for Calc2-06 FORTRAN program for Calc2-06 U-value of double-glazedwindow using air. convection and radiation analysis Psychrometery — dew point temperature determination (FORTRAN-based DLL [Prog72.MICROSOFT EXCEL WORKBOOKS ON COMPACT DISK Folder names given within square brackets 2 The Microsoft ExcelcomputingenvironmentIChap2l Calc2-O1 Calc2-02 Calc2-03 Calc2-04 Calc2-05 Calc2-06 Calc2-07 U-value for single-glazing(sequential computation) Glazing U-value as a function of wind speed (dynamic graphs) Thermal properties of infill gases (lookup tables) Finding the window aspect for a desired solar energy budget (Goal Seek) Day of the week (Visual Basic for Applications) Solar geometry for any tilted window (FORTRAN-based DLL [Fr206. Ar. Kr. Kr.dll Progl6.dll Calc4-O1 Calc4-02 Calc4-03 Temperature stratification in a double-glazedwindow Combined conduction.dll]) Psychrometric chart FORTRAN-based DLL for Calc3-05 Equation of time and solar declination angle Time conversion: LCT to AST and vice versa Sunrise. Xe or SF6 3 Thermal properties of windows IChap3I Calc3-OlbAir U-value of double-glazed window using Air (precise solution) Calc3-OlbAr U-value of double-glazed window using Ar (precise solution) Calc3-O1bKr U-value of double-glazed window using Kr (precisesolution) Calc3-OlbXe U-value of double-glazedwindow using Xe (precise solution) Calc3-02 U-value of triple-glazed window using air.

required for Calc5-Ol .xiv MICROSOFT EXCELWORKBOOKS ON COMPACT DISK Calc4-04a Calc4-04b Calc4-05 Calc4-06 Calc4-07a Calc4-07b Calc4-08 Calc4-09 Calc4-1O Solar geometry based on LCT Solar geometry based on AST (24-hour cycle) Cloud radiation model Sunshine radiation model Page radiation model (FORTRAN-based DLL [PageRM.dll FORTRAN-based DLL for Calc4-09 5 Windowsand daylight [ChapSi Calc5-O1 Lookup table for sky component: CIE overcast sky (FORTRAN-based DLL {Fr305. UK PageRM.dll FORTRAN-based DLL for Calc4-07a Clear. UK (28 June 1992) Irradiance. illuminance and zenith luminancedata for Garston. included in CD Calc5-02 Luminous efficacy and zenith luminance (Perez) models (FORTRAN-based DLL [Fr309b. Japan (6 June 1994) FORTRAN-based DLL for Calc5-Ol FORTRAN-based DLL for Calc5-02 Comma separated variables file.dll]) Page clear-sky radiation model (FORTRAN-based DLL [Clear.dll 1n3-5.dll]).dll]) Calc4-11 Data4-Ol Data4-02 Diffuse radiation model Decomposition of averaged-dailyirradiation into hourly values (FORTRAN-based DLL [Fr308. illuminanceand zenith luminancedata for Fukuoka.csv'. UK (1 August 1991) Irradiance. Requires '1n3-5 .dll]) Slope irradiance (Muneer) model Irradiance and illuminance transmission for multi-glazed windows — generalisedprocedure Equation of time and solar declination data for a four-year cycle Hourly horizontal and slope irradiance and other meteorological data for Bracknell.dll]) Calc5-03 Luminous efficacy and zenith luminance (Muneer—Kinghorn) models Calc5-04 Calc5-05 Calc5-06 Data5-Ola Data5-Olb Data5-02 Fr305. illuminance and zenith luminance data for Garston.dll Fr309b.dll FORTRAN-based DLL for Calc4-07b Fr308.csv Slope illuminance (Muneer—Kinghorn) model Sky luminance distribution — absolute co-ordinates (Perez model) Sky luminance distribution — relative co-ordinates (Perez model) Irradiance.

Japan (1994) Total number of workbooks = 53.UK (199 1/2) Slope illuminance and irradiance data for Edinburgh (Heriot- Watt). Exe and other files = Total number offiles on CD 14 = 76 .for: Sky luminance distribution plotter — colour raster LD_Numbr.Exe and other filesi LD_Fill.exe: luminance Sky (UK system) luminance distribution plotter — luminance values LD_Vals.exe GINO files required for the execution of the above exe files: Ggrafiib.txt: Data file for Plot_XY.exe LD_Indat. Japan (1994) Slope illuminanceand irradiance data for Garston.dll.for: X-Y plotter Data XY.MICROSOFT EXCEL WORKBOOKS ON COMPACT DISK xv 6 Acoustic of properties windows IChap6I Calc6-O1 Sound-reduction calculation for a given facade 8 Solar radiation and daylight data D Kinghorn fChap8J Calc8-O1 Data8-Ol Data8-02 Data8-03 Data8-04 Data8-05 Data8-06 Data8-07 Data8-08 Data8-09 Data8-lO Sky luminance data transposition Slope illuminance and irradiance data for Fukuoka. UK (1993/4) Slope illuminance and irradiance data for Manchester.dll.dll and GINO. UK (1994) Sky luminance distribution data for Fukuoka. UK (1993) UK (1992/3) Slope illuminance and irradiance data for Edinburgh (Napier).exe: Sky luminance distribution plotter — colour raster distribution recordingscheme LD_Numbr. UK (1993) Sky radiance distribution data for Fukuoka. Ginlibmp.csv: Data file for LD_Fill.exe: X-Y plotter LD_Fill.exe: Sky Plot XY.for:Sky luminance distribution recording scheme (UK system) LD_Vals. Slope illuminanceand irradiance data for Sheffield.exe and LDVa1s.con Total number of FORTRAN.UK (1993) Sky luminance distribution data for Sheffield.for: Sky luminance distribution plotter — luminance values Plot_XY. Total number of DLLs and FORTRAN programs = 9 Windows 95198/NT4Executablefiles on CD (FORTRAN.Japan (1994) Sky luminance distribution data for Garston. Mwinerr.

Page blank in original .

two text files which provide data for the executable programs and four GINO graphics software files (three 'DLL' and one 'con' file). For all DLL-based workbooks the followingprocedure MUSTbe used: • Launch Excel software • Open the DLL-based workbook • Enter data in the relevant worksheet(s) • Simultaneouslypress ALT + F8 keys • Select the relevant macro from the dialog box • Either click the Run button or simultaneouslypress ALT+R keys . Chapters 1 and 7 contain no electronic material. Some of the workbooks use dynamic-linked libraries (DLLs). Step-by-stepinstructions are therefore provided in the box below. four directly executable files. it is recommended that that material be read before using the workbooks. The Chapter 2 provides detailed instructions on the use ofExcel workbooks and workbooks are categorised on a chapter basis and are therefore placed within the respective folders. The present authors developed the DLLs using FORTRANsoftware. Care must therefore be exercisednot to alter the cell addresses. The files may be copied on the user's PC and used directly. Care must be exercised when handling DLL-based workbooks. DLLs are powerful tools that take away the tedium of large. Note that some ofthe workbooks contain multiple worksheets with interlinking formulae. sequential number crunching using the power of programming languages. Ifany changes are madeto the workbooks then the user should use the 'Save As' command and use a different file name to save the altered version ofthe file(s). The output of such files is beyond the sophistication of the graphic facilitiesavailable within Excel. All the 76 files included in the CD-ROM are naturally 'read only' and thus protected from any accidental deletion. GINOis a versatile graphics environment which has been used to produce the above-mentioned executable files.INSTRUCTIONS ON THE USE OF WORKBOOKS AND EXECUTABLE FILES The CD accompanying this book contains 53 Microsoft Excelworkbooks.

care must be exercised in using the present electronic files.xviii THE USE OF WORKBOOKS AND EXECUTABLE FILES The exefiles enable graphical representation of large and varied data sets. The respective input data files are also included. use the MS-DOS command: ATTRIB -R *. If any changes are madewithin the files. The followingprocedure may be used to remove the READ only flag from all the files thus enabling any edits to be performed: • • Copy the file(s) from the CD onto the hard disk in any suitable directory (say in the 'C' drive) After changing over to the above directory. they must be resaved under different names.csv and Data_XY.txt. . Finally. In any case. the original set of files may always be retrieved from the CDROM. It is important that the user conform to the format of the given files ifthey wish to import theirown data in the given files LD_Indat. Note that the latter files maybe created from any of the popular spreadsheet software.

lower infiltration rates and larger areas ofglazed aperture have been required in thedesign of buildings. optically switchable glazings. psychologically as well as environmentally. termed cold bridging. anti-reflectivesurface treatments. However the edge-sealcreates a thermal bridge.higher levels of insulation. Various technologies have been introduced to double-glazed windows to further enhance their performance. Examples of these new advances include spectrally selective low-emissivity coatings on glass and on thinplastic films. Recently.solarcontrol infrared transparentglazings. These malfunctions have strengthened the demands for windows with higher thermal performance. the conventional window has become the weakest thermal fabric in a building. It causesenergy loss at the perimeter ofthe window hence reducing the window's overall thermal performance. . solar energy. Windows have very appropriately been described as being the eyes. thereby creating an enclosure suitable for nondurable coatings and/or substitute infihl gases. The use of double-glazed windows is still the most common method ofproviding a reasonable level ofthermal resistance. studies have shown that 6% of the United Kingdom's energy consumption is due to heat losses from domestic glazings alone. Cold bridging may also cause condensation occurrence on the bottom part ofthe window inner pane. lowcoatings. between internal and external environments. silica aerogels. Better window glazing calls for a parallel development in edge-seal and frame technology. noise. the ears and the nostrils of a building. polarised glazings and evacuated enclosures. multi-layer film suspension systems.In view ofthis increasing demand.The edge-seal isolates the cavity between the glazings. In addition to their poor thermal performance. Windows provide humans with a variety of functions which include the supply ofthe interior spaces of buildings with light. holographic coatings.1 INTRODUCTION Windows. are essential architecturally. as a key element of buildings. For example. honeycombs. Most double-glazedwindows currently manufactured consist of two panes of glass separated from each other by an edge-seal. conductivity infihl gases. Today more attention is being given to the performance ofwindows and they are being designedto do exactly what we require of them. conventional windows can create comfort problems and damaging accumulations of condensation. socially. air and view according to the desires of the occupants and to shield them from dust.

Humans. more of the sun's heat is trapped giving rise to what has been termed the greenhouse effect. and could increase the risk of storms. Humans spend 1. however rough and crude. At leastone-quarter ofthe domestic heating bill in these countries is due to the thermal energy loss through glazings because they are the weakest thermal component in a building. in seeking shelter. dust and smells and to protect the inside from the inclement conditions of the prevailing weather. The admittance of air. need to be able to remain in touch with the outside world. From the earliest have made structures which. the material and the external environment. some regulating device is needed to keep out excessive noise.2 Windows and energy conservation Most of the energy used today originates from fossil fuels (gas. Therefore. including carbon dioxide and gases that cause acidrain. and how much is admitted will depend on the climate. In a nutshell. Similarly. people must have windows in their buildings. rain. they meant to protect them from the dangers and discomforts of the external environment. to see or to be seen. 1. Some of the functions which are of importance to architects and buildingservicesengineersare considered briefly in the following sections. It is a window in a buildingthat allows humans to look out of or look into a building. As carbon dioxide and other gases build up in the atmosphere. heat and cold can thus be prevented or regulated. additionally. Windows also provide the passers-by with attractive views into buildings that can be used effectively in the design of commercial premises. but the dominant link is achieved through the sense of sight. the shelter must have some form of visual contact. This could potentially result in the earth becoming hotter. or simply to observe the outside world. Burning fossil fuels emits pollutants. The greatest problem is how to admitwhat is wanted and keep out what is not.1 Windows and life the greater part of their lives in shelters. were times. How much light is to be introduced depends on what the shelter is used for. oil and coal). Using energy more efficiently is one of the most cost-effectivemeans of reducing emissions of carbon dioxide and. coastal flooding and drought. the opening is also expected to admit fresh air. helps to conserve fossil fuels. to watch the activities of others. Modern technologyis able to overcome what seems to be an insurmountable problem. snow. The oldest way of bringing light into a shelter is to let the natural light enter through openings and at the same time. Buildingsaccount for the bulk of the energy budget ofthe OECD countries.2 WINDOWS IN BUILDINGS rain and excessive heat or cold. improving the thermal insulation of the . that is.

the better is its thermal insulation and the more energy it can conserve. However. by reducing reliance on artificial lighting. Detailed analysis of the mechanism of heat loss through single. Daylight. daylight can be an effective means of saving energy and reducing the environmental impact. Traditionally. for instance. while other conditions remain the same.Better U-values can be achieved by using modern technology. The inclusion of daylight in the workplace provides workers with social and physiological benefits. The use ofexcessive glazed areas in an attemptto admitmore . For example. our ability to read. Using a low-emissivitycoating to improve a double-glazedwindow's performance has little effect on daylight transmissivity. the U-value is reduced by about half. electrical lighting in the UK accounts for an estimated 5% of the total primary energy consumption and office buildings may consume up to 60% of their total energy in the form of electric lighting. Low-emissivity(low-e) coatings are invisible. the transmissivity becomes 75%.The daylight transmissivityof a doubleglazed window with 4 mm thick float glass is 80% and if one low-e coating is added. Examples of this technology are given later in this book.and doubleglazed windowswill be presented in Chapter 3. For example. However. Therefore. it may impair. may increasethe ability ofoffice workers to do deskwork. occupants may feel discomfort by glare due to the excess of illumination in their field of view. Daylight and artificial light may need to be integrated to avoid lack of proper lighting within a building. however. which will be discussed later. lighting cannot be provided by daylight alone for the entire working day throughout the year. be troublesome in some situations. Daylight can. Heat is lost through windows by conduction. The lower the window U-value. 1.3 Windows as daylight providers Daylight is one of the largest concerns for both architects and residents. This parameter. but they are remarkably well suited for increasing thermal comfort and controlling solar impact. for example. Interior objects such as furniture and paintwork may suffer visual degradation and fading if subjected to daylight for longer periods of time. It is convenient to combine all types ofheat loss in a single parameter that describesthe behaviour ofa glazed window. convection and radiation.INTRODUCTION 3 window will contributeto energy conservation and environmental improvement. as a consequence of using the low-e coating. Daylight is also required to enhance the appearance ofan interior andits contents by admitting areas oflightand shade that give shape anddetail for objects. If the level of glare becomes very high. the main purpose of a window is to admit daylight and many houses are still designed on the assumption that daylight will be the working light for a greaterpart of the day. is called the thermal transmittance or U-value.

These differences have farreaching consequencesfor the performance of windows. During the night. concern for the lack of illuminance data for the rest of the country. is diagnosed as seasonal affective disorder (SAD). Research has shown that daylight has an important bearing on the human brain's chemistry. there is lack of sufficient daylight indoors. The age ofthe data may not create any serious concern. This problem is exacerbated in high-latitude locations where. although the clean air Acts passed in major towns and cities across the country could possibly influence present daylight levels. This common disorder among people living near high latitudes. but also on daylight. can reflectsolar energy while admitting the worthwhile daylight. the above process is not halted completely and this causes drowsiness. a substance identified as a neurotransmitter. For example.4 WINDOWS IN BUILDINGS daylightmay giveriseto solar overheating. People react . As a consequence of the absence of adequate measured illuminance data. At daybreak the incident daylight affects the glands in such a way that they cease production of melatonin. Within Britain around half a million people have been reported to show symptoms of SAD. There are algorithms that allow the prediction of illuminance when solar irradiation is provided as an input parameter. Lack of serotonin is known to be a cause of depression. Light entering via eyes stimulates the nerve centres within the brain which controls daily rhythms and moods. More recently new building constructions have employed illuminance data from Kew. With the advent of superinsulated windows it is possible to provide much larger glazed areas thus exploiting daylight and passive solar heat gains. in the UK buildings have traditionally been designed using daylight data recorded from the National Physical Laboratory in Teddington near London between 1933 and 1939. Further research has indicated a link between human exposure to light and serotonin. about 10 miles from Teddington. If the receipt of the dosage of daylight is not high enough. which causes drowsiness and thus sleepiness. 1.4 Therapeutical aspects of daylight to changing seasons with altered moods and behaviour. thus reducing cooling loads without increasing the lighting load. It has been shown that values of average horizontal illuminance in the northern part of the UK are about two-thirds of those reported for Kew. building designers have to rely on predictive tools and models. There is. Some windows using special lowemissivitycoatings. Thusvalidated insolation models will provide information not only on the interception of irradiation. however. These models should be capable of accurately predicting illuminance values from meteorological parameters such as solar radiation. seasonally. especiallyin commercialbuildings. such as Northern Europe. the pineal glands produce melatonin.

the optimum air gap should only be around 20 mm. Examples of these sources are road traffic. influencesperformance of tasks. A well-designed double-glazedwindow can provide over 40 dB of sound insulation. ventilation and lighting.from noise sources elsewhere in the building and/or from noise sources outside it. often with large windows. Thus a compromise is needed between thermal and acoustic considerations. This scheme is being used by the UK government to implement the Kyoto Protocol. and results in permanent changes in the normal functioning of the human organism which causes mental and/or physical deterioration. 1. . Thus. The instrument of the Kyoto Protocol on global warming has been used to challenge building services engineers to become involved in a CO2 permit scheme. such as heating. The World Energy Council has estimated that the overall efficiency of some air-conditioning systems is as low as 5%. an overdesigned air-conditioning system imposes a serious penalty on the environment. Witheach doubling of glass thickness. Modern architecture using light-weightstructures. load-bearing walls of older buildings. It causes annoyance or dissatisfaction. Solar energy may be considered as a serious candidate to reduce the use offossil fuels for buildingenergyconsumption. Ideally. Humans do need contactwith the outside world through hearing desired sound. and the acoustical isolation ofthe absorbent material aroundthe edgesofthe air space. Sound insulation does not mean eliminating all sources of sound.5 Windows as solar energy providers During the past quarter-century many building air-conditioning systems were overdesigned. Sound insulation within a building is as importantas other building services. For example. Noise is communicated to rooms within a building via many different paths. railways. but for larger widths there is no serious degradation in the thermal performance. for best thermal insulation and lower cost. the resulting plant capacity in the UK building stock exceeds the true requirements by as much as 30%. the gap width of the window cavity. affects communication. air-tightness.6 Windows as sound insulators Noise is defined as unwanted sound and has many effects on human beings. aircraft.the corresponding sound insulation is increased by about4 dB. The best air gap for sound insulation has a width of at least 150 mm. Factorscontributing to thesound insulation ofa multipleglazed window are its mass. allows more sound to be transmitted into the building compared with the heavy. Even then the annualenergyincident onUK buildings (1614 TWh) exceedsthe country's oil production (1504 TWh). Withreferenceto theEuropeanUnionmemberstates thereceipt ofsolar energywithin theUnited Kingdom is towards the lower end. industry and building mechanical services.INTRODUCTION 5 1. leads to damage to hearing.

1. Recently. a German window company. therefore. the addition of extra glazing was the only option for improving window thermal performance. manufacture and installation is being developed. superinsulated windows • The improvement in comfort through the elimination of cold downdraughts and radiation exchange. A low-e glazing. • An increase in the total light admission in residential and other buildings by allowing greater window areas to be employed without increasing the overall losses. Some of the technologies utilising coating techniques are summarised 1.6 WINDOWS IN BUILDINGS 1.7.7. Research in all aspects of windows is very active and new technology in material.7 Window technology During the last decade window technology has seen more dramaticchanges than any other building technology. has brought onto the market a window with a U-value which is as good as an insulated wall. designers and users. a reduction in overheating problems in temperate and tropical zones in summer.1 Wavelength selectivecoatings Transparent wavelengthselective coatings on glazing substrates can be usedto below. Interpane. inert infill gases. For instance. Transmission in the longwave infrared (far infrared) is nearly zero for the commonly used materials. • Better noise-attenuation performance. However. In the past. acts as a selective reflector .1 Low-emissivity coatings A low-emissivity (low-e) coating is a low absorptive coating to suppress infrared radiation exchange. Windows having the combination of these technologies are referred to as superinsulated windows. Greater • A reduction in condensation problems at the window edge area. design. • need forflexibility and freedom for architects. their reflectivity for the shortwave infrared is low. also offer the following advantages: In addition to reducing energy loss from buildings. leading to a reduction in the indoor cooling. The followingare some of window technologies presently in the marketplace. energy • With appropriateselective coatings. reflect or absorbcertain bands of radiation while allowing the transmission of others. insulating edge spacers and low conductive frames in window design. Most of the solar energy is shortwave infrared while the energy radiated from warm objects is in the far infrared band. modern technology has introduced low-emissivity coatings. 1.

The spray reacts with the glass surface at about 600°C forming tin oxide in the air. Sixty per cent of the heat lost through ordinary windows is through longwave infrared radiation. . This is probably the most efficient process for glass coating. aqueous tin chloride. although silver is the most widely used. Metal films are approximately less than 10 nm thick and exhibit partial visibility and solar transparency.7. A low-e window essentially doubles a window's thermal resistance because the low-e coating nearly shuts down the infrared conduit. Low-emissivity films are generally of the type dielectric/metal/dielectricor dielectric/metal. These coatings have been improved by Pilkington by the use of a two-layer coating to reduce the iridescence of thicker tin oxide coatings. The wavelength selective properties of dielectric/metal/ dielectric films are derived from both the optical properties of the metal and dielectric layers and the interference effects caused by the film stack. The preferred metals are silver.1 shows the spectral transmittance of low-e coatings. solar reflectance is low and the reflectance of the far infrared emitted from internal objects is very high. Figure 1.7.INTRODUCTION Radiation of blackbody at300 K 7 100 90 80 70 50 40 30 20 10 60 0 • I Wavelength (i. Traditionally.tm) Figure 1. must exhibit high infrared transmittance in order to preserve the infrared reflectance of the metal.1 Transmittance of float and Iow-e coatedglass (Source: Pilkington plc) where solar transmission is quite high. when used to overcoat a metal. The dielectricfilm. gold and copper. These characteristics are appreciated by building occupants as in winter more daylight and solar energy are welcome and less heat loss from the interiors is required. tin organometallic liquid or powder is sprayed onto the surface of the glass as it leaves the float bath.

for a temperature difference of one Kelvin between the inner andouter environmentsseparated by the glazing system. gas cavity fillings and insulating frame and spacers. the overall system efficiency can be increased.7.0 W/m2K. The window U-value is defined as the rate of heat loss per square metre. Filling the gap between glazing layers with gases such as krypton or xenonincreases the insulating value of the unit sufficiently to make it possible to reduce the size of the gaps and to build units with an overall thickness of 25 mmor less. The visible portion could be used for photovoltaic conversion and the infrared portion could be used for photothermal conversion. The spectrum holographicallycoated film would be laminated between two glazing substrates for protection.8 WINDOWS IN BUILDINGS 1.7. A simple design utilising a conventional dielectric/metal/dielectric low-e coating might be used to separate the visible and the near infrared portions of the solar spectrum. the U-value is approximately 1. Spectral splitting coatings 1. By tailoring the solar energy to that of the photovoltaic response. Cold mirrors are generallymultilayer dielectricinterference films.7. These coatings are already used in head-up displays for cockpit and automobile windscreens where instrument information is projected optically onto the screen and reflected to the pilot or the driver. Also.1. with one or more low-emissivity coatings.2 Spectral splitting and cold mirrors can be used to divide the solar spectrum into different broadband regions. the solar spectrum could be partitioned from low to high energy as the low-c transition wavelengthsbecome shorter. . The cold mirror coating has the opposite spectral response to that of the transparent low-c film. under steady-state conditions. Superwindowsare commerciallyavailable now and the range will continue increasing. This allows the superwindowsto be used in conventional double-glazingsash designsand framing systemsand makes them viable for building retrofit.l.1. For triple-glazing with two low-c coatings of emissivity=0. Holographic coatings 1. The cold mirror exhibits a high reflectance in the visible region and transmits highly in the infrared. Spectral splitting coatings allow for multiple photovoltaic and combined photovoltaic—photothermal systems.3 Holographically coatedglazings can be tuned to reflect any waveband in the solar while allowing 75—80% transmittance in the visible waveband. If transparent low-e coatings with different spectral characteristics were used.2 Superwindows Superwindowsare glazing systemscomprising multiple panesofglass or plastic films. the photovoltaic will operate more efficiently if infrared heating is eliminated. In this way various glazings can be tailored to particular photovoltaic or photothermal needs.

These windows utilise a vacuum in combination with a low-emissivity coating on one of the the geometric internal surfaces. 1. 1. Conduction throughthe spacers then becomes the majorvehicle of heat transfer. The resulting heat loss coefficient for a 2 cm thick space of granules between glass panes is about 1 W/m2K depending on the size of the granules. It has a higher diffuse transmittance but also a higher thermal conductivity.7. 1. These materials are.2 Xerogel Xerogel is a material similar to aerogel resulting from attempts to avoid the expensive supercritical drying process involved in monolithic aerogel manufacture. translucent rather than transparent.4 Vacuum windows Evacuated windows circumvent the high cost of gas filling.7. Monolithic aerogel has excellent properties. the translucent natureofthis designmakes it less suited for normal windowsthan monolithic aerogel.37 W/m2K for a double-glazed window with a 2 cm layer ofaerogel. The capillary and honeycomb types are most commonly made of polycarbonate or acrylic plastic.3. can be poured into the space between double-glazing.1 Aerogel windows are materials which allow solar energy The air space between window panes can be filled with aerogel.7.3.3. on the other hand.an approachwhich eliminatescavity gas convection and most of the radiantheat transfer.3 Transparent insulationmaterials Transparent insulation material (TIM) to be readily transmitted. However. It is predicted that only a slight vacuum would be required to achieve a U-value of 0.3 Geometric media Geometric media is a term used to describe glazings in which composition of the material gives rise to advantageous insulating and solar transmission properties. but minimisethe heat loss through the surface.7.7. 1. however.INTRODUCTION 9 1. This technology is currently being developed to overcome the followingproblems: . Granular aerogel. being both highly transparent and having thermal properties approaching that of an opaque wall. a microporous silicate foam material which reduces thermal transmission. Two forms of aerogel exist for use in windows. monolithic (in the form of continuous slabs) and granular (in the form of granules).

In dwellings. Condensahigh-latitude tion is a natural phenomenon which occurs when warm moist air comes into contact with a cold surface. the temperature of the glass pane of the single-glazed window approaches the outside ambient temperature. causing its water vapour to condense. The pressure • • • within the cavity has to be maintained to extremelylow values. furniture and paintwork. Single-glazed windows are the weakest thermal elements in building.7. Table 1. e. producing thermal expansion that could overstress rigid edge seals. condensation on glazings may obstruct the view through windows and.1 Per cent gas loss through various sealants after 20 years Ar/SF6 Singleseal Polysulphide Ar 13 to 13 to 15 15 Ar Ar 6 to 13 SF6 Polyurethane Silicone 8 to 10 8 to 10 Butyl Polyisobutylene Pennapol P-2 Permapol LPM Thikol LP Dual seal Polysulphide 15 to 20 >45 5 to 7 4 to 5 5 to 10 1 to 4 3 to 9 33 to 45 0 to 1 3 to 5 5 to 6 Polyurethane Silicone 2 to 5 12 to 15 1.10 WINDOWS IN BUILDINGS • The seal must be able to maintain the vacuum required.8 Condensation in buildings be described as the modern disease of buildings. condensed water will run off causing damage to window frame. The indoor air which is in Condensation might . Table 1. which cools the air below its saturation point.g.7. The exceptionally low heat transfer across the glazing requires special attentionbe given to the frame of the glazing unit.1 shows the percentage loss for different gases/gas mixtures through various sealants after twenty years' service. test and developmanufacturing processes to commercialisethe concept. Spacers are needed in the evacuated cavity to keep the glass layers from collapsing. The temperature difference between the outer and inner panes will be large. In winter. More work is needed to optimise. if it becomes so excessive and no means for drainage is provided within the window frame. ofthe order of 1O_8 bar. In countries millions of households are affected by it.

internal warm air can carry more water vapour than cold air. However. Considerable amounts of water vapour are generated within dwellingsby the occupants andtheir activities. the temperature of the external air is usually so low that its moisture contentis also very low even if the air is saturated. therefore.1 The effect ofinfiltration In coldweather. Secondly. Heating of a building is importantin many respects. Condensation results from 1. Firstly. 1. washing and drying clothes.2 The effect of internal environment The internal humidity is the most important factor in condensation occurrence.5 kg of moisture for each m3 burned. External humidity in temperate climates (with the exception of warm weather condensation) has little influenceon internal condensation. It usually affects high thermal capacity structures which are very slow to respond to an increase in temperature. An average person engaged in sedentary activitieswill exhale more than a litre of water vapour in 24 hours. natural gas produces 1. forming condensation. Flueless heating appliances that depend on combustion to produce their heat are the bulk producers of water vapour. merely by the process of breathing. cooking. the 'warm weather condensation' phenomenon may become the cause for condensation. there will be a short period when condensation can form on cold structural surfaces until they have had a chance to warm up. Thus.ifheating of a building is supplied in consistent levels. produce about 12 litres of water vapour per day.INTRODUCTION 11 contact with the glass surface discards some of its water vapour. Its occurrence should. Therefore water vapourwhich coldair cannot carry will become suspended in the warm air instead of being condensed on building furniture and structure.5 kg of water vapourto the internal air. 1. e.8. Typical family activities. It has been estimated that four persons livingin a house for 12 hours will.g. The most susceptible parts of a dwelling to condensation are those areas which are not heated or are heated inadequately. add 2.g.1.1. physical The factors that effect the formation ofcondensation are as follows.1 The causesof surface condensation a series of relatively simple and well-understood factors. When there is a sudden change from cold to warm humid weather. be thoroughly predictable. The moisture-laden air can then be removed by ventilation. if cold air infiltrates inside the building. More energetic activities can increase this amount by up to four times. breathing. The amount of rainfall does not affect the risk of condensation.8. e. it . the moisture increase within the building will be insignificant.8.

the building shellhas to have reasonable levels of thermalinsulation. surface condensation may occur in areas where cold-bridging is possible. Opening windowsanddoors is being discouraged because of anticipated increased energy costs. flues or chimneys.8. it is theoretically possible to avoid all condensation by adequate ventilation. On the otherhand. on the one hand.12 WINDOWS IN BUILDINGS will keep theinternalsurfaces warm and above dewpoint. suspended floors.1. Also. Surface condensation will occur . the greater the ventilation. the internal surface temperature should be maintained within a few degrees below the internal air temperature. 1. the less the ventilation andthe heat inputto the building. themore likely the condensation occurrence.4 The effect of room surface temperature if the temperature of the surfaces of the internal space falls below the dew pointof the neighbouring air. consistent. However. However. In addition. In the past. 1.8. Adequate insulation can minimise the risk of surface condensation by maintaining the surface temperature at a reasonable level. standard levels of temperature throughout the building and the internal air should be maintained. i. Therefore. the greaterthe heat necessaryto replace that which is lost by ventilation.1. Insulation must always be complemented with a reasonable heating system if its effectiveness is to be realised. The paradox that faces building services engineers nowadays is that. Therefore. and consequentlythe greater the cost. Such ventilators provide a controlled amount of fresh air in a small but adequate quantity which circulates around the window area keeping condensation at a minimum. occupants are sensitive to the discomfort ofdraughts created especiallyby open windows. no amount of thermal insulation can make an unheated room warm. New construction techniquesgenerally reduce natural ventilation. To completely eliminate the risk of surface condensation. Intermittent heating most frequently obtained by switching the heating systems off during periods of unoccupancy and switching them on during habitationhours will increase the condensation rate. This is also a requirement for thermalcomfort. dwellingswere naturally ventilated by means of less well-fitted doors and windows. the outdoor air is likely to be at lower moisture content than the indoor air. for humans to be thermally comfortable surface temperature should never fall more than 5°C below the air temperature.e.3 The effect of ventilation In temperate climates. Ventilation should be consistent throughout the whole building and ventilators should be incorporated into windows.

in addition to the well-researchedthermal performance criteria. and the comfort conditions expected. much legislation and many regulations on energyconservation and otherareas have been imposed in . in kitchen. Moist air should be removed at source. bathroom or shower and should not be allowed to move to cooler areas such as bedrooms. both the aural and visual characteristics ofwindow installations become paramount. leaving room for judgement and is not intended to be prescriptive. To do this successfully requires that the demands of modern living. Global environmental impact optimisation is similar. By control ofhumidity.2 The avoidance of condensation to the way in which buildings are heated. Means for the removal of moisture from within the house by ventilation must therefore be provided.INTRODUCTION 13 1. Optimum window designincludes the impact of energy consumption. much of the inconvenience of condensation can be prevented. while minimising thermal losses and electric lighting demand.9 Life cycle assessment of multi-glazed windows Life cycle assessment (LCA) is a systematic approach to assessment of environmental impacts associated with a product. The air extraction can be done either by natural means through open windows or other vents. 1.This book addresses a continuing need to focus on sustainable development. global environmental impact and occupant comfort. temperature and air movement. Along with rising demands to improve efficiency and decrease energy consumption in buildings comes an expectation for continual improvement in building interiors. Each of the above objectives provides loose guidance. Optimisation of energy consumption incorporates embodied energy. Condensation problem should be seriously addressed during the initial design of a dwelling. using LCA as an assessment tool to develop a greater understanding of the window life cycle. andto highlightimprovements which are necessaryto lessen its environmental impact and make the processes involved more benign. process or activity. ventilated andinsulated. but evaluation is based on energy generation. Occupant comfort optimisation guides the user towards a glazing solution which offers sufficient noise attenuation.8.be incorporated into design criteria. or mechanicallyby fitting an extraction fan. To this end. while ensuring that the needs of future generations are not compromisedby today's activities.10 Legislation and regulations In order to ensure that a window functions properly. adopting a holistic or whole lifeapproachto designmethodologies. thermal performance and electric lighting demand over the life cycle ofa window. Condensation is very much related 1. and greenhouse gas production.

Computational fluid dynamics (CFD) can be described as the use of computers to model and evaluate the behaviour of fluids in given situations. For example. Canada. building services engineers. Computer modelling offers one way of assessing complex transportmechanisms and the interaction ofthese in buildings. maximum infiltration through windows.14 WINDOWS IN BUILDINGS many countries.12 Scope of the present book This book aims to provide building professionals with the necessary information to enable analysis and design of multiple-glazed windows. tables and software for the design and selection of glazing system for their appropriate incorporation into modern buildings. Thermal. 1. The CFD modelling results obtained in this book have been verified and validated by a variety of methods. minimum fire and safety requirements.The target audienceis architects. 1. aural and shortwave energetic transfer characteristics of windows are addressed here. Use of CFD and other microprocessor-intensive applications is bound to grow as a result of the developments taking place in the hardware and software technologies. minimum acoustic insulation. and they have to incorporate double-glazed or triple-glazed windows into their new houses to meet the requirements of the legislation. The regulations related to windows address the following issues: minimum window area for daylighting. the UK legislationnow precludes the use of single-glazed windows. Norway. maximum U-value for window thermal characteristics. This has led to the use and innovation of double-glazed windows and thus increased the quality of the indoor environment of buildings. the USA and Germany. students. educators and researchers working in the area of energy efficiency in buildings. In some countries these regulations are only recommended by the government but in otherstheyare mandatory. The book is written with a view to provide easy-to-use charts. maximum window area in combination with solar transmittance for summer conditions. This technique is being used increasinglyin the building services industry. The use of CFD modelling in window research areas is well documented. The strict legislationand regulations have made it impossible for house builders to use single-glazedwindows. especiallyin the northerncountries such as the UK. builders and building designers.11 Use Use of computers in windowdesign of computers to predict the performance of building elements is now widespread. The ever-falling price and increasing processing power of personal computers has led to a significantly increased sophistication in building energy .

the databases will be used to analyse the interaction of window and the building.INTRODUCTION 15 analysis. daylight illuminance and other weather parameters. Such computers would easily be able to handle detailed hourly databases of solar radiation. . The Microsoft Excel-basedapplications provided in the presenttext enable large amounts ofcomputation using desktop or laptop computers. integrated circuits of 256 MB RAM will be commonly available. In turn. A study by the US-based Semiconductor Industry Association has shown that by the year 2001.

Page blank in original .

faster time to work completion. This will no doubt transform the way in which engineeringdesign and analysis is undertaken. barrier to the use ofcomputer hardware would be the software requiring a long training period. (1997) and Liengme (1997). In general terms. will be available on a single chip. a few hours compared to several weeks that may be required for some of the above-mentioned applications. There are very many such texts available and those readers who have not had the opportunity of using Excel in the past may refer to Hallberg et al. study guides and training videos. reduced costs. The former text provides a basic drill in the use of Excel whereas the latter includes more focused applications directly to be used for engineeringapplications. many people would argue that it is the smaller software applications that have had the most immediate effect on their work environments. more complex engineeringapplication packages such as those used for undertaking finite element analysis (FEA) or computational fluid dynamic (CFD) work present the top end of the market.2 THE MICROSOFT EXCEL COMPUTING ENVIRONMENT Modern engineering designand analysis demands for increased productivity. and perhaps only. improved quality and worldwide access to information. 1998) predicts that by the year 2005 complete systems. information technology (IT) advances have been a catalyst to significantchanges in the way design practices are run. Help is available in the form of well-writtentexts. Undoubtedly the key drivers of this change have been the explosive breakthrough in the designanddevelopmentofintegrated circuit chips andthe emergence of computer aided design (CAD) software. The real. with the storage and processing power of a 1997 personal computer. These are: . In just ten years. The advantage of using a spreadsheet-basedcomputing environment suchas Microsoft Excel is that the training times are of the order of. there are some very compelling reasons to switch from a manual (calculator-based)to computer-based engineeringdesign. While larger. The Institution of Mechanical Engineering's Professional Engineering magazine (PE. It would be possible to use palm-sized computers and large numerical databases for undertaking numeric-intensivecalculations. Further. Computer processor technology is one of the most important areas of growth. at most. Chips with the respectivecapacities of a gigabit and one terabit memory will be commonly available by the years 2002 and 2010. the learning curve for Excel is very gentle and easy.

VisiCalcwasrapidly replaced by Lotus 1-2-3 in the 1980s. Use of graphics: Information can be displayed. Excel — A user-friendly environment for number crunching 2. without long waiting periods. which now extends well beyond calculating lists of figures or making colourful charts. • Ease of analysis: The ability to perform 'what-ifanalysis' is significantly • improved. lookup tables (for performing interpolations on large property . For most science and engineeringstudents it requires only a few hours' training to master the software at an intermediate level. organisations not using this technology would fall into the 'obsolete' class. Thepower of Excel for engineeringuse is exploitable through its multitude of functions. It is also the corporate standard for spreadsheets. Microsoft Excel.e. i. The chances are that most people who perform any work with numbers do haverecourse to a spreadsheet. Now with the wider adaptability ofcomputer-based applications in the industry andacademia. which was faster. Discussion can then take place for design scenarios. macros (to perform repeatedly a chain of commands or operations). Today with over 50 million users world-widethe sales of Excel are only second to the Web browser software. Faster: The results are available at applied. Prior to the launch of spreadsheets scientists and engineers had looked down on PCs. However. Mainstream activity: Although some organisations take pride in being 'early adopters'.18 WINDOWS IN BUILDINGS • design calculations and optimisation. copied and printed in an tested and • Cheaper: Very significant avings in the person-hours required for basic Anynew models madeavailable in the scientific journalscanbe rigorously • • • easy-to-understand manner thus making it easy to communicate with clients who may not have a scientific background. Some journals claim that without the spreadsheet the PC would never have developed so rapidly in the early 1980s. The average Excel user probably employs only a tenth of its power.1 The history of the development of computer spreadsheets may be traced to VisiCalc which in the early 1970s created a significant impact in the corporate field. Ease of communication: Design information may be transmitted through the World Wide Web easily and cheaply. VisiCalc changed that image. Excel comes with a range of templates and on-screen training tips. But Lotus were slow in transferring their product to the Microsoft Windows environment and consequentially lost ground to their nearest rival. most prefer to use only the mainstream technology.once. Ease ofupdating: Newer databases may be linked-up quiteeasily and thus the design process can be kept abreast of rapid technological changes. smarter and easier to use.

1 It was pointed out in Chapter 1 that the overall heat transfer coefficient (U-value)of windows and other building components is a convenient index to .Mathematica or Matlab face strong competition from spreadsheets.1 Sequential computation Sequential computation is the most fundamental and perhaps one of the most useful properties of Excel. This book reviews the current window technology and the associated knowledge base. As stated above.2 The functionalityof Excel — an example-basedtour This chapter is designed to give readers a very focused introduction to the potentialofExcelas a CAD tool.xls workbook contained in the CD accompanying this book. This example clearly demonstrates the position of Excel in the world market as the premier tool for analysing tabular and graphical information. 2. It may be worthy of note that all examples and workbooks available on the companion CD run in the Microsoft Excel 97 software and are equally effective within the Microsoft Windows95/98 or the Microsoft NT environments.However. 2. British Petroleum is one of the world's largest oil and petrochemical corporations. Thosewho would like to gain the basic skillsfor using Excelare referred to the texts mentioned above. In this respect new books are emerging which highlight the computing potential ofExcel. Demonstration of this feature is provided in the form of the followingexample using the Calc2-O1. Example 2. goal seek (for solving non-linear equations). this field is very virgin and a lot more development is bound to take place in the near future. solver (for solution ofa given set ofnon-linear equations and optimisation) and interlinked graphs (for performing what-if and other analyses). However.2. Due to theirvery nature engineering subjects easily lend themselves to the use of spreadsheets.2. In this respect packages such as Mathcad. tools andproceduresare included here which are of direct relevance to the present text. The information can be loaded as Microsoft Excel spreadsheets. since all examples and computational tools provided in the companion CD use Excel as the problem-solvingenvironment it was felt that a brief tour of the most essential features of Excel would be of benefit to the reader. Annually.Onlythose functions. whichin contrast require only a fraction ofthe training time.THEMICROSOFT EXCELCOMPUTING ENVIRONMENT tables 19 and data querying). The relevant material is presented in the followingsections. they publish their Statistical Review of WorldEnergy on their home page on the Internet. it is assumed that the reader has a basic familiarity with the Excel package.

or by opening the workbook once the Excel software has been launched.1) is the combined radiative—convective internal surface resistance R the thermal resistances of structural components and R0 the combined radiative—convective external surface resistance (m2K/W). If the global parameters are too numerous or where the computation demands such a structure they may be provided via an additional worksheet.R + R0)' (2.2. such as glass thickness and thermal conductivity.20 WINDOWS IN BUILDINGS evaluate theirthermal transmission characteristics. hro the longwave radiation heat transfer coefficient and h0 the convectivetransfer coefficient. . as will be shown in otherexampleworkbooks. R0 is defined as (m2K/W). The U-value of an element is the reciprocal of the sum of the individual resistances of the material layers which compose it. For a typical building component: U = (R1 + >. Independent variablessuchas wind speedmaybe provided in a column so that the effect of its variation on the overall heat-loss speed. windward or leeward. recently made available in the scientific journals. Compare the effect of these models on the U-value against calculations based on the CIBSE (1982) and ASHRAE (1993) recommended models. Open the Calc2-Ol.xls workbook eitherby double-clickingthe file icon from the Windows 95/98 or NT environment. At this stage it would be useful to explain the layout of all the workbooks provided with this text. A steady-state heat transfer condition is assumed for the determination of the respective U-valuesand these are used widely by design engineers to estimate fabric heat flows in buildings. The top and the left part of the sheet confirm the name and title of the workbook and any 'global' parameters. Develop an Excel worksheet to ease the following tasks: (a) Computation of U-value for varying glass thickness and external wind (b) Determination of the critical heat transfer paths (thermal resistance) which have the most significant effect on U-value. R0 = (Chro + h0Y' (2. on U-value. (d) Investigation into the influence of the wind direction.2) is the emissivity factor. (c) Investigation into the influence of the external heat transfer coefficient (h0) models. U-valuestake into account the conduction through the fabric together with the convectiveand radiative heat transfer processes taking place at the internal and external surface of the building element.2.

.2. expressed in rn/s. respectively for the summer and winter months.65v Thereader mayeasilyincorporate any ofthe above-rnentionedmodels intocell BlO ofthe 'Main' worksheet in Calc2-O1.THEMICROSOFT EXCEL COMPUTING ENVIRONMENT 21 coefficient may be explored.2. The most critical resistance in a singleglazed window is the internal heat transfer coefficient.1 + 3. based on an up-to-date review of all notable studies. Rather. if the wind speed is halved or the internal heat transfer coefficient is doubled.lv h0 = 7. is given by h0 = 16.9 + O.3) (2.7 + 6. h0 = 5.xls (Cells D9:D15 in Sheet 'Main'). (1972): h0 = 18. (2.8 +4.8) Cole and Sturrock (1977): Sharples (1982): = 1. Note that Excel has a 15-digitprecision and hence it surpasses the accuracy of the usual (single precision) FORTRANcodes. Tasks (a) and (b) above can be addressed by changing the input variables. Equation (2.2.4).Ov (2.2. Tasks (c) and (d) require further information. if theglass thicknessis doubled from thegiven 4 mm to 8 mman insignificantchange takes place in the U-value.4) v is the mean wind speed.xls. However. progressively significantchanges in the U-valueresult. This feature will be demonstrated via Calc2-02. The model proposed by Loveday and Taki.2.2. Loveday and Taki (1998) have undertaken field measurements of the heat loss coefficients of walls. (2.6v0605 h0 = 5.2. The effecton U-valuemay then be investigated for a number ofwind speed values. For example. (2.42v (2.7) (2.7 W/rn2K and 34 W/m2K.2. A red border divides the inputs from the outputs. (2.3)) and ASHRAE (Eq.7v°5 Other models are given below.5) Ito et al.4)) procedures result in lower U-value estimates. Note that ASHRAE (1993) does not explicitly provide an expression of the form given in Eq.6) (2.4) has been developed by Loveday and Taki (1998) to enable a comparison of R0 with other work. they provide design values of 22. Their research suggests that the CIBSE (Eq. Sequential computations of the individual thermal resistance are then undertaken leading up to the U-valueestimate.2.2.

(1972) for leeward direction of wind is: h0 = 8.22 WINDOWS IN BUILDINGS Research on the effect of wind direction on U-value of walls has been undertaken by Ito et al. may also be examined quite conveniently.1 models were presented for the external (exterior to the building or window surface) convective heat transfer coefficient. graphical outputs are also sought by designers to provide an insight into the overall picture. (1972) model for leeward direction Use the embedded graph facility to explore the variation of U-value as a function of the glass thickness. facility A lookup table consists of a column or row of ascending values. Example 2. from 3 to 12 mm. While spreadsheet tables are essential for querying information expressed in numerical format.2. Any change in these parameters will automatically change the graph.2 In Section 2. Using Calc201 . Calc2-02.xls as the starting point. Using Calc2-01 . 2.3 Lookup tables a very useful facility for creating lookup tables. Within Calc2-02.2.3 + 0. 2.2. The change in the window heat loss. resulting from a windward to a leeward condition.9) Once again.2. The model recommended by Ito et al.725v (2.2 will demonstrate this feature of Excel. say. this model may be incorporated within the worksheet as discussed above and its influence investigated. called Excel provides . (1972) and Sharples (1982).xlsthe minimum and maximum value of wind speed may be provided in the sheet entitled 'Main'. for a range of.xls compute the U-value of a single-glazed window for a variation of wind velocity from 0 to 3 rn/s and display your results graphically for the following models: CIBSE (1982) ASHRAE (1993) Loveday and Taki (1998) Ito et al.2 Dynamicgraphs An interesting feature of Excel is that it provides a dynamic link between numeric and graphic environments.2. Example 2. often. This type of analysis is presented in the followingsection.2.xlswas thus created to perform the above numerical experiments. Through this the user may find one piece of information that is based on another.

In the present context.161 0.x106 (m2/s) 4.96 5.76 5.161 0.16 5.674 0.673 0.885 6.6 222.Develop an Excel spreadsheet using the property matrix with a view to compute the relevant properties for a temperature of 25°C.672 0.674 0.303 5.161 0.9 233. and corresponding data for each compare value.594 40. The analysis ofany multi-glazedwindow heat loss involves calculation of the convective heat transfer within the window cavity.THEMICROSOFT EXCEL COMPUTING ENVIRONMENT Table 2.673 0. Example 2.2 211.298 6.6 241.6 kxlO3 (W/rn.6 203.8 199. Help is available for the design engineer in the form of Excel's Vlookupfunction.740 4.31 5.1 229.42 5.69 4.93 5.775 55.252 39.61 pxlO7 (kg/m. a large number of window glass selection tables are available from most manufacturers.0 214.8 218.161 0.809 33.161 0.672 0.161 0. The corresponding data are the thermophysical property values.560 4.916 38. In this illustration.58 5.52 5. argon.71 5.050 34.497 5. are available in the literature.161 0. as shown in Table 2. the first column (temperature of the gas) contains data for the compare values.K) 0.944 7.85 4.87 4. demonstrated via the Calc2-03.3 226.591 31. This in turn requires estimation of the thermophysicalproperties of the gas at the mean cavity temperature.40 6.4 207.696 52. . Lookup tables are encountered by design engineers very frequently.161 0.xls workbook.112 5. For example.2.71 5.28 6.671 0.671 0.05 5.49 5.82 5.161 0.5 255.385 43.673 0.779 Pr 0.924 5.161 0.673 0.9 263.710 49. demonstration is provided by including lookuptables for thermophysicalproperties for three inert gases.40 5.27 6.380 7.15 6.161 0.304 8..161 vxlO7 (m2/s) 30.23 5.189 r.695 5.04 5.161 0.090 6.s) 195.45 4.1 Therinophysical propertiesfor xenon (Hanleyet al.78 4.671 0.3 270. krypton and xenon.315 35.161 0.14 5.828 8.33 320 330 340 350 5.673 0.674 4. used as an infill material in multiglazed windows.671 0.10 6.93 6. i. 1974) 23 T (K) 250 255 260 265 270 275 280 285 290 295 300 310 p (kg/rn3) 6.00 4.947 59.1 which gives the thermophysical properties of xenon.2.161 0.603 36.673 0. Manual interpolation of these properties for any given value of temperature is time-consuming.e.671 0.161 0.57 'compare values'.797 46.1 248.2.510 6.977 42. Inevitably this requires interpolations for each ofthe desired thermophysicalproperty.K) Cp (kJ/kg. which is approximately the average of the corresponding cavity glazing temperature.3 Thermophysical properties of xenon gas.

These are then used along with properties obtained at 295 K to enable interpolations to be carried out. The Open the Calc2-03. the position of the upper bound for the given temperature is obtained. At this stage the reader is encouraged to try out computations at other temperatures and indeed to copy into the suite of cells discussed above the property data for other gases. Study the layout of the 'Gas Properties' worksheet. The function VLOOKUP(L5. There is.e. given in cells K6:K8 respectively. An atriumfacility. To avoid excessive overheating the window .2. however. The entire table of properties is produced in a temperature-descending manner (see cell range A24:H39). The Vlookup function performs its taskin the followingmanner.xls workbook corresponding values of kinematic viscosity (v).A24:A39.2.4 is obtained.2. 2. Excel varies the value in the specified cell untilthe result in the target cell Example 2. uses the arguments: lookupvalue.are then obtained in a likewise manner (see the worksheet under discussion for further details). The object is to provide a well-lit yet comfortable environment within the atrium.1 if is required). The properties for krypton and argon are respectivelygiven in cell ranges A44:H66 and A71:H103. but not the input values.4 Goal Seek Design engineers often encounter a situation where the result of a formula or the final result at the end of a series ofcomputations is known (or desired).1) function in cell 18. In this particular instance the values are obtained for a temperature of 295 K (see column K of the worksheet). The INDIRECT('Gas Properties'!18)function is then usedin cell MS to obtain the numerical value of this upper bound. Then using the MATCH('Gas Properties'!L5. 298. is to be designed for a location north ofEdinburgh.see cell K5.15 K) is known the corresponding values for the desired properties are obtained using the VL000KUP function mentioned above.15 K a more rigorous procedure is required and this is explained below. table array and column index to look in the first column to find the largest value less than or equal to 298. adjoined to a planned office building. Thefollowingexamplewill adequately demonstrate this procedure. i.'Gas Properties'!A7:A22. thermal conductivity (k) and Prandtlnumber(Pr).15 K. Once the upper bound of the temperature (300 K) bracketing the given temperature (298. To precisely obtain the properties at the given temperature.24 WINDOWS IN BUILDINGS in the usual manner (see Section 2. any help Provide a temperature value of 25°C in cell C4. For the given temperature of25°Cin cell C4. properties are obtained corresponding to the largest value less than or equal to the given temperature.-1) function in cell 17 and the ADDRESS(17+ 23. To solve such a problem the Goal Seek facility may be used.1). the corresponding absolute temperature is calculated in cell L5 (298.15). a problem with the above procedure.

The data. These basic data are provided in columns A to E while solar geometry calculations are performed in columns F to J. Choose the OK button. enter the value of 300. If for any reason the Goal Seek function was not originally installed with Excel software then run the setup program to install it. are essential. the Goal Seek and Solver facilities do not work with other user-defined macros or functions and hence the sequential computations provided in columns F to V manner.1 will appear.2.xls workbook. Use the Excel Goal Seek facility to find out the aspect for vertical windows which will result in the total energy receipt of no more than 300 Wh/m2. The basic data for the site. In this particular instance the cell address is V22 and the content of this cell shows that the total incident energy is 484 Wh/m2. (e) In the present context the Goal Seek procedure converges to the value of 261° from north for the window aspect.xls workbook. do represent the approximate climate of the east of Scotland. A much more compact version of the slope radiation workbook using the Visual Basic for Applications module is provided in Chapter 4. That procedure and thephysics ofslope radiationis explainedtherein. however.e.enterthe referenceofthe cell which contains the value for window aspect. C5.THEMICROSOFT EXCEL COMPUTING ENVIRONMENT 25 orientation is to be set to a value which will limit the incident solar radiation to a given design value. They have therefore been hidden. For any given instance oftime the incident solar radiation on any tilted window requires knowledge ofhorizontal diffuse and global irradiation. the search is not exhaustive. This quantity corresponds to the selected window aspect of south. Hourly horizontal diffuse and global irradiation data for a nearby location is available for several years. A dialog box such as the one shown in Figure 2. Of course.only one day's hourly data is included in the Calc2-04. further analysis related to the transmission of short-wave energy through the specified multi-glazed . i. The copyright of these data resides with the UK Meteorological Officeandhence the data have been suitably randomised to avoid any breach of the copyright. As an example. The slope irradiation is then computed in a sequential manner as shown in columns K to V. surface geometry and ground albedo (reflectanceto solar radiation) are provided in cells C4:C7. i. In the To value box. (b) In the Set cell box. Refer to the Calc2-04. enter the name of the target (or results) cell. Unfortunately. It must be pointed out that the Goal Seek facility works only 'locally'.e. (c) (d) In the By changing cell box. Note that columns 'F' to 'U' contain intermediate steps. The use of the Goal Seek procedure is now explained in a step-by-step (a) From the Tools menu select 'Goal Seek'.

That topic will be the subject of discussion in 2. which may be related to either a single or several independent variables.xls workbook. embedded within the Calc2-05. A function.1 The Goal Seek dialogbox windows would Chapter 4. Using this routine a Visual Basic program has been developed.26 WINDOWS IN BUILDINGS Iioal $eek . Progl-2.5 Visual Basic for Applications Microsoft Excel includes the Visual Basic for Applications (VBA) programming language which enables the user to write modules which may be either subroutines.xlsdemonstrates the use of VBA. This VBA program. is linked to the 'Main' worksheet. require knowledge of the day of the week for a given date.2.et cell: To value: Sy çhangingcell: OK JV2iJ2E ! icc' Cancel Figure 2. modern control algorithms for energy-efficient buildings. Example 2.For. A FORTRAN routine. The example workbook Calc2-05. returns thevalue to a cell (or a range of cells) of a user-defineddependent variable.5 In order to differentiate between weekdays and weekends. on the other hand.2. Open command if Excel has already been launched.xlseither viaa double-clickon the fileicon. A subroutine may perform repetitive tasks(macros) or enable the userto communicate with the workbook via a user form. Best available copy . or by using the File. To fully appreciate the mechanics ofthis workbook the reader is advised to undertake the followingprocedure: • Open theworkbook Calc2-05. has been presented by Muneer (1997). be required. thus launching Excel and opening the workbook in a single step.suchas the optimum start algorithm.2. functions or link-ups to user forms.

weekday numberand the day of the week. based on wellestablished programming languages such as FORTRAN or C. Figure 2. Theease with which input data maybe provided in Excel is second to none.THEMICROSOFT EXCEL COMPUTING ENVIRONMENT 27 • Simultaneouslypress theAlt and Fit keys thus launching the Visual Basic for Applications environment and the accompanying Visual Basic code. i. This reduces the overall amount of memory needed in the computer system. (c) Programs written in different languages can call the same DLL functions. E5 and F5. Furtherdetails ofthecomputation ofJulian daynumberanda number of other VBA routines used in this book are provided in Muneer (1997). IDN. but is usually used as a library for applications (Microsoft. • You may close down the VBA environment by simultaneouslypressing the Alt and Q keys.xlsis included to enable the user hands-on practice.2. 1995b). • Try out the workbook under discussion by inputting various dates (cell range A5:C5). nwkdayandnweekwhich respectivelycalculate the Julian day number.Harris (1997) and McBride (1997).6 FORTRAN or C-based DynamicLinked Libraries of the most powerful features of Microsoft Excel operating under the Microsoft Windows environment is its ability to link with Dynamic Linked Libraries (DLLs).e. . A DLL contains one or more functions that are compiled. A Dynamic Linked Library is an executable file. large numbercrunching modules may be developed within a FORTRAN DLL and then a link provided via the Excel-VBA front-end. linked and stored separately from applications using them. The power of such rich mediums can therefore be fully exploited.2. For thepresent example this codeis displayed within theCalc2O5 module. Further information on the Visual Basic environment and the development 2. as long as each program follows the functions' calling conventions. DLLs are in effect directly executable files. (b) General functions placed in the DLLs result in a small size of the applications that share these functions. Thus. It may be noted that there are three user-defined functions within the Calc2O5 module. These functions are respectivelyand sequentiallycalled uponin cells D5. resulting in improved machine performance. ofrelevant codes may be obtained from Microsoft (1995a).2shows the schematicofthe information flows in sucha schemeand the example workbook Calc2-06. The advantages of DLLs include: One (a) Multiple applications can accessthe same DLL.

The solar declination (DEC) may be defined as the angular position of the sun at noon with respect to the equatorial plane. Computation of EOT and DECis required for all solar energy-related analysis. LOT.For: High-precision algorithm for EOT. DEC and solar geometry provided by Muneer (1997). Develop an Excel workbook which links with a suitably formatted FORTRAN DLL. D. has been presented by Muneer (1997). Muneer's routine was also incorporated by a NASA contractor. the time required for one full rotation ofthe earth is less than a solar day by about 4 minutes. Gronbeck (1998).6 A solar day is defined to be the interval of time from the moment the sun crosses the local meridian to the next time it crosses the same meridian.2. the standard time (as recorded by clocks running at a constant speed) differs from the solar time.xlsis based on the FORTRANcode Progl-6. with the view to exchanging input/outputdata between an Excel worksheet and the FORTRAN DLL. Yallop. A high-precision algorithm and a FORTRAN routine.28 WINDOWS IN BUILDINGS Excel Memory space In liii L Out I Array I Figure 2. Due to the fact that the earth rotates in a diurnal cycle as well as moving forward in its orbit. Thus. based on the work undertaken by the British astronomer B.2 A schematic of information flow for execution of a Dynamic LinkedLibrary Example 2. (b) Simultaneously press the Alt and Fil keys thus launching the Visual Basic for Applications environment and the accompanyingVisual Basic code.2. to provide a web-based calculator for EOT. based on the above code. 31. Calc2-06. .xlsafter launching Excel.suchas the solar heat gain through any window system. The difference between the standard time and solar time is defined as the equation of time. A step-by-stepprocedure for usingthis workbook is detailed below: (a) Open the workbook Calc2-06. See text box on p. DEC and solar geometry.

Value Myanay(16) Next i EndSub Figure 2.3).CelIs(i.Cells(2.Value Myarray(15) Sheets(Compute).DLL .THEMICROSOFT EXCEL COMPUTING ENVIRONMENT (c) Activate the 'Compute' sheet by clicking its tab 29 at the bottom of the workbook.Cells(2.3 VBA code for linking CaIc2-06. The relevant data for solar geometry calculations are to be provided in the 'Site geography.Cells(i. (d) For each change made in the date and time columns run the macro thus. = SheetsC'Compute").Cells(i.Cells(i.2).For DeclareSub Progl6Lib'c:\Wi_BkMAK\Fort_Dll\Fr206. 2). $).5).4). window geometry' sheet.Cells(i. say X.Cells(i. Value MyarTay(3) Myarray(4) Myarray(5) — =Sheets('Site").Value Myarray(7)— Sheets("Compuie). and INC (solar inclination angle) will be introduced in Chapter 4.IO).Value Fori6To 17 Sheets('Compute"). Simultaneously press the Alt and F8 keys to reveal the macro 'CaliSolinc'. 7). SOLAZM (solarazimuth).Cells(2. Value Myarray(14) Sheets("Compute").Value Myarray(6) Myarray(9) Sheets("Compute"). Try out the workbook under discussionby inputting various date and time values (cell range A6:E17). Sheets("Compute").Cells(i.xls with Fr206.Value Myarray(13) — 9). Call Progl6(Myarray(l)) 6).Value 5). 1I). 4).CelIs(i.Value Myarray(8) Sheets(Compute'). Myarray(2) Sheets(Site").Value Myarray(l2) = Sbeets('Compute). Note that the terms AST (apparent solar time).Value— Myarray(l I) Sheets('Compute").Value Myarray(l) Sbeets(°Site"). l). 3).Value Sheets(Site").Value Myarray(lO) Sheets(Compute"). Run this macro by choosing the macro name and then click 'Run'.Cells(2.Value —Sheets("Site").2. SOLALT (solar altitude). For those readers wishingto producetheirown DLLswith a view to control the execution from Excel the followingprocedure is recommended: Step 1 Produce the required FORTRAN code and call it.Cells(i.Cells(i.dll" Alias'_PROGI6@4' (MyarrayAsDouble) Sub CallSolinc() Dim Myarray(l To 16) As Double l).Cells(i.Cells(2.

14159 NTIMES 2 YLAT=comArr (1) YLONG=comArr (2) YRLONG=comArr (3) AZI=comArr (4) TLT=comArr (5 iyr=comArr (6) imt=comArr (7 idy=comArr (8 ihr=comArr (9 ime=comArr (10) Computation of output variables eat. usedfor producing FR206.DLL command: Dumpbin /Exports Y. Say the resulting productis called (by compiling Y.DLL file in a suitable folder andthen change the file path address in the VBA code accordingly. solinc is undertaken within this sub-module. dec.DLL.This detail is further clarified by Figures 2.30 WINDOWS IN BUILDINGS Step 2 Using FORTRAN compiler and the code X.xls via the VBA code shown Figure . solalt.4 FORTRAN code Progl6. solazm.3 The DLL links with CaIc2-06.0 PROG16 pi=3.4 which respectivelyinclude the listings of SUBROUTINE PROG16 (comArr) !MS$ATTRIBUTES DLLEXPORT REAL(S) comArr(16) dtor=3 .2.2.For produce the DLL and then building).2.2. ast.14159/180. in 2.0 comArr(12)=dec comArr (13)=ast cornArr cornArr (14)=solalt (15)=solazm comArr (16) =solinc END SUBROUTINE Figure 2. comArr (11)=eot/15 .DLL Step 3 Find out the alias name of the DLL by using the MS-DOS Step 4 Place the Y.For.For' in Chap2 Folder for the full list.3 and 2. Refer to 'Progl6.

In the authors' experience this facility of Excelis probably the most powerful feature as will be demonstrated by Example 2. Also. It is possible for the user to select the solution time and the desired number of iterations. The Estimates option enables the selection of Tangent option suitable for near-linear problems or the Quadratic option which is more suitable for problems with a high nonlinearity. Derivativesand Search procedures. or simultaneouslypress ALT + R keys 2. It may therefore not give all the possible solutions. as mentioned above. However. To use the Solverfacility theuserdefinesthe problem that needs to be solved by identifying a target cell.minimisedor made to reach a specifiedvalue.2.2. the changing cells (containing the numerical values of the variables) and the constraints imposed upon the optimisation routine. precision of constraints.xls. Full control of the solution process is possible via the use of the Solver Option dialog box. Solver only works 'locally'. Microsoft Excel Solver can handle up to 200 independent variables with upper or lower boundswith an additional 100 constraints. multi-dimensional optimisation problems may also be easily handled for function maximisation or minimisation.THE MICROSOFT EXCEL COMPUTING ENVIRONMENT 31 the VBA andthe FORTRANsourcecode used for developingCalc206. For all DLL-based workbooks the following procedure MUST be used: • Launch Excel software • Open the DLL-based workbook • Enter data in the relevant worksheet(s) • Simultaneously press ALT+ F8 keys • Select the relevant macro from the dialog box • Either click the Run button.7 Microsoft ExcelSolver Microsoft Excel Solver is based on well-established numeric methods for equation solving and optimisation.2. In Section 2. Solver is similar to Goal Seek. These methods supply relevant numeric inputs to the candidate model and through an iterative procedure the results are generated. Within the choice of solution method three options are available: the Estimates. except that within Solver there is an ability to solve a family (or set) of linear or non-linear equations. automatic (independent-variable)scaling and the solution method used by the Solver. integer tolerance. Thetext box below provides a step-by-stepguide for the executionof all DLLbased workbooks. The targetcell is thecell whose value is to be maximised.7. The Derivativesoption enables a choicebetween Forward differencing .4 Excel's 'Goal Seek' facility was demonstrated.

Using the Excel Solver facility optimise the aspect and tilt of a given solar collector to maximise the interception of solar irradiation. Finally. If for any reason the Solver function was not originally installed with Excel software then run the setup program to 2 In the Set Target box. It may be possible for the Solver to reach other local maxima or minima. the Search option makes it possible to experiment with Newton(steepest descent routinewhich requires less processor work) and Conjugate gradient (routine which requires fewer iterations to obtain convergence) methods.5 will appear. Calc2-07.7 Refer to Example 2. If the 'time limit exceeded'messageis displayed. (faster and yet approximate computation) andCentral differencing (slowerbut A briefintroduction to the Solver facility was provided above.2.5). resulting in an energy yield of 660 Wh/m2. A dialog box such as the one shown in Figure 2. The Solver optimisesthe collector aspect and tilt to the respective values of 164° due north and 20° tilt. This quantity corresponds to the selected aspect of west and tilt 900.2. .2. Upon successful convergence the user has a choice of retaining the Solver solution or reverting to the original settings. i. install it. enter the reference of the cells which contain the values for collector aspect and tilt. 6 Choose the Solve button. Launch the Solver procedure. Example 2.e. enterthe nameofthe target (or results) cell. Answer reports are also generated and a sample report is shown in Table 2. Remove the constraints and thenproceed to solve. 4 In the By Changing Cells box. This may be demonstrated as follows. click the 'continue' button. Note that for a vertical collector with an aspect of due west the total energy received is 285 Wh/m2. as shown in the example (see Figure 2. Further information on the Solver facility is provided in Microsoft (1992) and Person (1992).4 wherein Excel's Goal Seek procedure was used to find the value of the window orientation for the atrium facility which would limit the incident solar radiation to a specified value. 5 Furnish any suitable constraints. 3 Select the Max option. In this particular instance the cell address is V22 and the content of this cell shows that the total incident energy is 285 Wh/m2. C5:C6.2. The relevant procedure is now explained in a step-by-stepmanner.32 WINDOWS IN BUILDINGS precise computation). I From the Toolsmenu select 'Solver'. Youmayfind that a local maximum value has been reached.xls may be used for this exercise.2.2.

Set Target Cell It ___________ 1° tJLI olve Close .J CYiueOf uess EqualTo: Max ChangingCells: y dd -.Solver Pararneers rT. ( C Ivila •( 0 Pelete It.' $$' $4 S Subject to the Constraints: tions iange o jelp Figure 2.5 The Solver dialog box .2.

(1997) Visual Basic for Applications5. (1974) The viscosity and thermal conductivitycoefficients for dense gases and liquid Argon.crest.34 WINDOWS INBUILDINGS Table 2. org/software-central/html/forum. It is worthmentioning that theSolverprocedure requires a longer time to set References ASHRAE (1993) Handbook of Fundamentals. McCarty. N. Atlanta. Cole. Kinkoph. (1997)Special edition: Using Microsoft Excel 97. 3rd edn. B. Gronbeck. J.. Refrigerating and Air-conditioning Engineers.0 Answer Report Worksheet:[Calc2-07. Building and Environment 12(4). variable scaling and the solution method. Que Corporation. Georgia. J. B. integer tolerance. Chem.2. Hanley. C. M. Phys. California. 3.. Santa Rosa. and Sturrock. Xenon. and Nielsen. USA. Indianapolis. shtml Hallberg..2 Answer report generatedby Microsoft Excel Solver procedure Microsoft Excel 8. . 979. Readers may experiment with these settings to get a better grasp of the mechanics of this procedure. R. Krypton. and Haynes. R. The iterations required to reach a suitable solution are dependent on the precision of constraints. W. Nitrogen and Oxygen. Ray. S. CIBSE (1982) CIBSE Guide. Harris. (1977) The convective heat exchange at the external surface of buildings. Sams Publishing. 207—214. (1998) http://solstice. London.Wh/m2 Adjustable cells Cell $C$5 $C$6 Name Aspect (duenorth) Tilt Originalvalue 180 Final value 164 20 90 Constraints Cell $C$6 $C$6 Name Cell value 20 20 Formula $C$6< =90 $C$6> = 10 Tilt Tilt up the problem than the time taken by iterations.xls]model Target Cell (Max) Cell $V$22 Name Original value 484 Final value 660 Totalincident energy. American Society of Heating. Chartered Institution of Building Services Engineers. H.

ASHRAE Semi-annual meeting. V. Butterworth-Heinemann. and Oka. Letts Educational. J. 23—29. Microsoft (1995b) Fortran Power Station Programmer's Guide. Muneer. New York. N. Santa Rosa.B. Proc. (1992) Using Excel4for Windows. A. (1982) Modelling the heat transfer through opaque building components for the detailed computation of thermal performance under varying climatic conditions. 7 October 1998. T. R. Wiley. K. Microsoft Corporation. H. Seattle. Microsoft Press. P. London. New Orleans. 2nd edn. Department of Building Science. Que Corporation. Seattle. L. (1972) A field experiment study on the convective heat transfer coefficient on exterior surface of a building. University of Sheffield. McBride. Sharples. (1997) A Guide to Microsoft Excel Scientistsand Engineers. Internal Report BS63. K. Redmond. Microsoft Corporation. California. London. (1997) Solar Radiation and Daylight Modelsfor Energy Efficient Design ofBuildings.IMechE. . D. BSER&T 19. Kimura. Washington. (1998) Outside surface resistance: Proposed new value for building design. Person. and Taki. Microsoft (1995a) Microsoft Excel/VisualBasic Programmer's Guide.THEMICROSOFT EXCELCOMPUTING ENVIRONMENT 35 Ito. (1997) Introductory Visual Basic. Microsoft (1992) Microsoft Excel User's Guide 2. Loveday. FE (1998) ProfessionalEngineering. Oxford. 39—40. for Liengme. S..

Page blank in original .

96)is being replaced by argon (M= 39. Table 3. air (molecularmass. For example. The percentage market share of these inert gases has increased considerably in all the countries and it is over one quarterof the total sales within Belgiumand Germany.3)in Germany. The most valuable advance in multiple glazed window technologiesis the low-emissivity coated (low-e)glass and the use ofheavier gases. by krypton 83.1 Market share of gas filled multiple-glazed windows in Europe (Han. 1996) (%) 1986 Sweden Finland 10 10 1987 12 12 1988 14 14 Denmark 15 1 GreatBritain Belgium Germany 20 20 20 3 25 25 20 4 25 25 . The use of the above new technologieshas dramatically changed the thermal performance of buildings and the way in which buildings utilise solar energy and daylight. and the Building Research Establishment in Watford are all engaged in window research. window research has developed very rapidly over the last decade. the Centre for Window and Cladding Technology in Bath. Over the last decade significant advances have been made in glazing technology and numerous new designs and new materials have been incorporated by the multiple glazing window industry.0. Due to these requirements. Within the United Kingdom the Glass and Glazing Federation based in London.3 THERMAL PROPERTIES OF WINDOWS Recently greatly improved performance has been demanded of the building fabric in general and of the window in particular.95) in the UK. (M Table 3. In the USA the Lawrence Berkeley Laboratoryat Berkley has undertaken pioneering research work related to window technology. M= 28. the Advanced Glazing IndustryClub in Oxford.1 shows the market share of gas-filled multiple-glazingwindows in some European countries.0.8) in the USA andSwitzerland and by xenon(M= 131.

1993). morale can fall and workers may refuse to work in such an environment. Humans respond to the interaction of these six parameters. 1985).38 3. For example. health and productivity can be affected. W RES=rate of heat exchange by respiration. W A heat balance is reached if the rate of heat storage S = 0. W W=rate of external work. When humans become thermally uncomfortable. 1993). The above equation then suggeststhe wayto establish equilibrium. The feelings of thermal comfort or discomfort reported by humans are complex and are not completely understood. The heat balance equation for the human body can be represented in the followingform (Fanger.1 WINDOWS IN BUILDINGS Thermal comfort is often defined as that condition of mind that expresses satisfaction with the prevailing thermal environment (ASHRAE. W RAD =rate of heat exchange by radiation. consideration is generallygiven to how to cool indoorenvironments to . 1973): Thermal comfort S=MET+W±RAD+C±K-E-RES where (3. Heat transfer into thebodyandgeneratedwithin it must be balanced by heat outputs from the body in order to maintain the body's internal temperature at around 37°C. it is not exclusively a function of air temperature as the other five parameters are also relevant. For this reason. there has been interest in research into the conditions which produce thermal comfort. radiant temperature. The human response to thermal environments is affected by four basic environmental parameters (air temperature. in fact. thermal comfort in offices or vehicles may be greatly affected by solar radiation and specifying comfort limits in terms of air temperature alone will be inadequate (Parsons. W E = rate of heat loss by evaporation. Thermal comfort is often assessed in terms of air temperature alone but.A human has to adjust his/her clothing and activity and regulate his/her position with regard to heat sources (artificial or natural) in order to meet their requirements. In hot climates. W MET =metabolic rate.1) C = rate of heat exchange by convection.1. This definition emphasises that comfort is a psychological phenomenon and not related justto physical environment or physiologicalstate(Parsons. humidity and air movement) and two personal parameters (the metabolic heat generated by human activity and clothing worn by a person). W S =rate of heat storage. W K = rate of heat exchange by conduction.

and by radiation exchange between occupant and the glass and other surroundings. excessive window area with inappropriate or non-functioning solar control systemswas to be avoided in order to improve thermal comfort and reduce heating plant capacity and running costs. Windows (of. The dry resultant temperature tres is commonly used to describe the environmental conditions. as key elements of building construction.1. In winter. Traditionally. 1993).5 (tair+ tr). the long wavelength radiation exchange between the cold surface of window glass and the occupant contributes to a sensation of cold discomfort. the higher the mean radiant temperature and the higher the dry resultant temperature and. the situation is reversed. the temperature of the internal glass surface influences thermal comfort as a result of heat loss produced by long wavelengthradiation exchange between the occupant and the window (Button and Pye. say. 15 often used.2) tr=mean radiant temperature of the surrounding surfaces. consideration is given to how to heat environments. usable blinds and control devices were preferred (CIBSE. which arises from the absorption of solar radiation and higher temperature of the outside air.THERMAL PROPERTIES OF WINDOWS 39 provide thermal comfort.the less comfortable the room. the air temperature tajr and the air velocity v (m/s). m/s tres =dry resultant 1989). °C v =air velocity. °C For indoor still air conditions a simpler formulation. influenceoccupant thermal comfort by heat gain or heat loss throughthe glass. In cooler climates. tresO. During summer. Wray (1980) has presented a simple procedure for determining the comfort criteria in passive solar buildings and has shown that the optimum uniform . The dry resultant temperaturecan be calculated from the followingformula tajr/l0V tres — where tajr= air temperature. °C temperature. For an occupant close to a window. A modern building is expected to provide a satisfactory thermal environment. + tr (3 . consequently. Windows. up to 30% of main facade area) with simple. the higher the window surface temperature. In summer. the internal surface of glass can be much hotter due to the absorption of solar radiation.which eitherraises or lowers the room air temperature. It represents the temperature at any pointin the room as influenced by the mean radiant temperature of the surrounding surfaces tr. The long wavelength radiation exchanged between the window glass and the occupant and also direct shortwavelength solar radiation receivedby the body through the glass contribute to a sensation of hot discomfort.

The steady-state heat transfer through the fabric of a building may be obtained through the analysis given below.2) AF =surface area offabric.40 WINDOWS IN BUILDINGS for comfort temperature an office environment lies between 18 and 26°C.45tr+ 0. This exchange can take place by conduction. He also has established the followingrelationship: tres 0. However. individuallyor in combination. Thus through the use of new technologies. winter conditions tend to govern building design.e. K The thermaltransmittance.1.1) where T=temperature difference betweenthe inside and outside air. m2 rate. In northerncountries.3) 32 a hundred years ago as an essential requirement for energy conservation and thermal comfort within buildings. Thermal performance is the key issue of a multiple-glazed window. i. if the double-glazed window is also coated with one low-e coating and the air is replaced with argon. Heatexchange betweenbuildingsandtheirsurroundings occurs in two ways. a double-glazed window with a 20 mm air gap will reduce heat loss by up to about half compared with a single-glazed window.55tajr Thermal transmission of multiple-glazed windows (3. W q =fabric heat losstransmittance U-value. convection or radiation. heat transfers to or from a buildingdue to air exchange throughthe gaps in the fabric. It is therefore usual to denote the rate of heat transfer from a building to the outside as positive. For example. Double-glazed windows were recognised about qF UFAFLXT (3. is given by (3. Firstly. UF. Secondly. 1994). the heat loss can be reduced to a quarter of the single-glazed window (Muneer and Han.2.2. All research so far undertaken on double-glazed windowshas aimed to improve theirthermalcharacteristics and hence reduce the overall heat loss from buildings. the indoor environment can be dramatically improved. =fabric thermal or W/m2K where . heat transfers to or from a building due to the temperature difference between the building and its surroundings.

In winter the air from outside is heated as it enters the space and then leavesthe space at the room temperature. Combining Eqs (3.6) . is the density of air entering the building space (kg/rn3) and J1 is the volumeflow rate of air entering the building space (m3/s). can be calculated as = p1i' (3. is qv = iic. m2K/W Thethermal transmittances for a wide range ofbuildingcomponents aregiven.5) and substituting into Eq.2. hour1 V= volume of the space.2. (3. In any building there is an infiltration and exfiltration of air due to unavoidable gaps in the construction. q. so thenet energy transfer rate is known as the ventilation loss.THERMAL PROPERTIES OF WINDOWS 41 = heat transfer coefficient for the fabric inside surface.2. m. The rate of air movement into and out of any space within a building depends on the pressure differences. The rate of energy transfer due to air movement.4) where c. comfortable conditions inside any occupied space. or the ventilation energy transfer rate (W). W/m2K R= sum of the resistances of the individual layers of the fabric. In steady state. T1 is the internal temperature (°C) and T0 is the external temperature (°C). W/m2K h0= heat transfercoefficientfor the fabric outside surface.2.5) n=air change rate per hour.3) where p. the mass flow rate of air into the space (kg/s).2.2.(T — T0) (3. which in turn are affected by wind direction and speed around the building. 1986..4) yields = — T0) (3. in CIBSE Guides (CIBSE. 1989).3) and (3. is the specific heatof air at constantpressure (J/kgK). m3 A certain number ofair changes are required to maintain fresh. for example. q..2. Thetermair change rate is frequently used in airmovement calculations and is expressed as n= where 36OO- (3.

The total thermal transquantified by mittance ofwindows.2.6) can be written as: q = nV(T1 — T0) (3.2. This heat is transferred in two ways: . and the spacer between panes.42 WINDOWS IN BUILDINGS For atmospheric over the normal inside temperature and humidity range. m2 Af= projected area of the entire fenestration. frame.1) and or calculated separately. Using area-weighted U-values for each contribution.8) for the calculation of the overall U-value of a fenestration: U0 (UAcg + UegAeg+ UfAf)/Af (3. W/m2K U= frame U-value. m2 Acg Af= projected area of frame. W/m2K Centreof glass thermalanalysis To understand heat loss mechanism through glazing. m2 = Ueg edge-of-glass 3.1. cold weather conditions govern the design of windows it is usual to assume that heat is transferred from the building to the outside atmosphere. consists ofthreecomponents arising from the glazing unit. Eq.8) where = projected area of glazing.1 U-value. In a single-glazedwindow. 1993) provides Eq.2.7) (3.2. (3. 3.2. Since. in general. m2 A8= projected area of edge-seal. U0. These components canbe measured In accordance with Eqs (3.2. it is worth while to the start by analysing the heat loss through a single-glazedwindow.2. W/m2K U=centre-glass =overall U-valueof the window. heat loss from windows is the thermal transmittance or U-value. the ASHRAE Handbook ofFundamentals(ASHRAE. (3.2). W/m2K U-value.1 Heat transfer to the internalglass surface Heat is lost from the room to the internal glass surface whenever the glass surface is at a lower temperature than the internal air temperature and the room surface temperature. air the mean value of the product PiCp is approximately equal to 1200 J/m3K.2. Hence as a good approximation for normal ventilation conditions. there are three identifiableregions associated with the heat loss through a window: these are conduction through the glass pane and combined convection and radiation on either side of the pane.

particularly on exposed sites.1) As with the inner glass surface.THERMAL PROPERTIES OF WINDOWS 43 • By exchangeof long wavelengthradiationbetween the glass surface and the room surfaces.1 illustrates the mechanism for heat loss through double-glazed units.0 W/mK 3. The long wavelength radiation exchange depends on the temperature of the surfaces of the outside surroundings and on the sky temperature. 1993). The additional pane also provides added thermalresistance to long wavelength radiation exchange.1. The heat transfer at this surface varies considerably and is climate dominated (Buttonand Bye. This effect is demonstrated by the formation of dew and frost on surfaces exposed to clear skies due to their cooling below the ambient air temperature.2. The sealed gap provides a significantthermal resistance due to the low thermalconductivityof the infill gas compared with glass. The greater heat loss is usually that due to radiation exchange unless the glass surface has a low-emissivitycoating.1. Wind-driven rain will further increase the heat loss due to contactcooling and evaporation.3 Heat transfer from the outer glass surface (3. The rate of heat transfer by convection/conduction is usually high due to the influence of wind.2. An effectivemethod for reducing window heat loss is to add a second pane of glass separated from the first pane by a sealed space. The resistance of glass to heat transferis relatively low and this can be calculated as follows: Rg where Rg=glass thermal resistance. m2K/W Lg=glass thickness.1. • By convection and/or conduction from the room air moving over the surface of the glass. m kg=glass thermalconductivity. the sky temperature can be extremely low.2. normally taken as 1. heat transfer from the outer glass surface is by long wavelength radiation exchange to the outside surroundings and by convection and/or conduction to the air moving over the exterior surface of the glass.2. . With clear skies. Figure 3. The gap and the additional pane increase the thermal resistance of the window. 3.2 Heat transfer through the glass In this case heat transfer is by conduction.

/Convection Radiation Figure 3.1 Mechanism for heat loss through double-glazed units .2.Conduction/convection .

W/m2K = hr radiation heat transfer coefficientwithin the cavity.4) h = h + hr where h=infill gas convection heat transfer coefficient. Such values are equivalent to those generally required by legislationfor wall installation (Button and Bye.4 W/m2K for windows of less than 30 mm total thickness.2.2. W/m2K = (3. m K/W rg h = cavity heat transfercoefficient. 32.1. will be discussedlater in The radiation conductance (hr) for each cavity is given as: hr4[i+_ l]T where (3.1) to calculate the heat loss through the window. (3. W/m2K J+Lgrg where (3.2.1.1. The U-valueobtained can then be substituted into Eq.5) = Stefan—Boltzmann constant.2.2. W/m2K The infill gas convection heat transfer coefficient (he) in this chapter more detail.3) N number of window cavities of Lg= total thickness glass panes. m =thermal resistivity of glass. 1989).THERMAL PROPERTIES OF WINDOWS 45 The thermaltransmittance. for multiple-glazed windows can be calculated using the procedure outlined in BS 6993 (British Standards Institution.1. = emissivity of the outer pane W/m2K4 . Present research work is targeted to the achievement of U-values ofless than 0.4 U-valuecomputation basedon BS 6993 U= where 1 (3.2) h=conductanceof multiple glazed unit.1. 1993).U-value.

2 shows the variation of surface convection coefficientagainst wind speed and direction.46 WINDOWS IN BUILDINGS emissivity of the inner pane Tf= absolute temperature corresponding to the average of the mean pane temperatures. Shelteredcondition corresponds to the third floor of buildingsin urban centres. the exterior heat transfer coefficient h0 is standardised to 100 c'J E C w 0 a) 0 C) C 0 C C) a) 0 C) C) Cl) Wind speed (mis) Figure 3. varies around 12. The external heat transfer coefficient.05 to 0. Theexternal heat transfercoefficient. K The effective emissivities of the surfaces bounding each enclosed gas space are required in order to calculate hr. (1972). Normally.845 for Uncoated glass surfaces (BS 6993) and range from 0. It also depends on wind direction as demonstrated experimentally by Sharples (1982) and Ito et a!. Figure 3.2 Variation ofsurface convection with wind speedand direction . 16.12 for low-emissivity coated glass surfaces.7 and 33. The effective emissivity is 0.2. normal and severe. Chapter 2 provides further information on models for obtaining h0. Normal condition is adopted for most suburban and rural buildings or the fourth to eighth floors of buildings in urban centres. Severe condition is applicable only for buildings on coastal or hill sites or above the fifth floor in suburban or rural areas or windows above the ninth floor in urban centres. h0.5.3 W/m2K corresponding respectively to the three conditions of exposure: sheltered.2. The latter values are frequently quoted by the respective manufacturers. h0 is a function ofmean wind speed.

2. the thermal performance of double-glazed windows has recently seen dramatic improvements. reduce the temperature in the vicinity of the edge-seal substantially. The internal heat transfer coefficient is given by the expression (3. frame and shading system.2. 3.3 W/m2K is usually assumed.0 W/m2K for natural convection at vertical surfaces.2. It is crucial that these improvements are accompanied by a parallel developmentin edge-seal (spacer) and frame technology. Once again this is a standardised value for the purpose of comparing glazing U-values. If the interior surface of the glazing has a low hemiis given by: spherical emissivity.1.6) = h1.THERMAL PROPERTIES OF WINDOWS 16. As mentioned previously. When a fan-blown heater is situated below or above a window.2 Edge ofglass and frame thermalanalysis Any glazed aperture (fenestration) in a building envelope usually consists of glazing material.7) The value of is 3. + where hrj = internal radiation heat transfer coefficient. this value will be larger. W/m2K The radiation heat transfercoefficientfor normal glass surfaces is 5. Thermallypoor edge-seals and frames may increase the heat loss through the window and impairits overall U-valueand. The total rate of heat transfer through a fenestration system can be calculated knowing the separate heat transfer contributions of the centre-glass. moreover.1. 1989). • aluminium spacer • steel spacer • metal spacer with thermal break .7 W/m2K for the purposes ofcomparison ofglazing U-value 47 for thevertical windows. Hence for ordinary glasssurfaces and naturalconvection the value of 8.3 x where (3.3 W/ m2K (BS 6993. thus increasing the possibility of condensation occurrence on the inner pane. The most widely used spacers in multiple-glazedunits fall into one of the following classes: h = hemisphericalemissivityof the coated surface.edge-sealand frame.the radiation coefficient h = 5. W/m2K =internalconvectionheat transfercoefficient..

B and C are correlation coefficients.266 0. Using more insulative spacer helps in increasing edge zone temperatures on the inside.2 W/m2K for most types of frames and glazing (Aschehoug and Baker.1 Coefficients for edge-of-glass U-value (ASHRAE. The thermalconductivity of the aluminium spacer is much higher than that of a foam spacer. 1995).2. Frame U-values for a variety of frame and spacer materials and glazing unit thickness are given in ASHRAE (1993).681 0.043 0. 1993). Estimating the rate of heat transfer through the frame is complicated by: • the variety of fenestration products and frame configurations • the different combinations of the materials used for frames • the different sizes available in residential and commercial applications • the glazing unit width and spacer type Table 3. B and C for metal.1 below demonstrates the calculation of the centre-glazing. degrades the thermal performance of the glazing unit locally. Table 3.027 0. The latter reference provides Eq.1 gives values for A.774 0.842 0. called cold-bridging. The U-value improvement for the foam spacer over the aluminium design is of the order of O.2. Example 3.xls workbook.033 .1—O. ASHRAE's procedure has been incorporated within the Calc3-Ola.2. thus significantly reducing the risk of condensation. Spacers in multipane units greatly increase conductive heat transfer between the contacted innerand outerglazing.706 C —0.897 0.010 0. (3.2. Laboratory measurements have shown that the conductive region ofedge-seal is limited to a 65 mmwide band around theperimeter of theglazing unit (ASHRAE.769 B 0.682 0.edge and frame heat loss. insulating (including wood) and fused glass spacers.3.2. 1993) Spacer Type Metal Insulating Glass Metal+ insulating A 1.1) where A. and a combination of insulating and metal spacers.2.1) for the calculation of the edge-ofglass U-value as a function of spacer type and centre-glass U-value: UegA+BU+CU2 (3. This phenomenon.48 WINDOWS IN BUILDINGS • fibreglass/plasticspacer • butyl spacer • foam spacer The products most commonly used in Europe are either hollow aluminium extrusions or roll-formed galvanised steel sections.

6 3.2. Such experimental results.2.2.9 3.70 1. krypton filled Centre-glassU-value (W/m2K) 2.2.6 Table 3./i.7 2.1 — 2.8 2.4 5.1 insulated 2.2.air filled Triple glazing.0 2. Frank andWakili (1995) used a two-dimensionalfinite difference model to determine the linear thermal transmittance coefficient.6 5. low-e.1 10.2 2.1 with thermal break Aluminiumclad wood.4 Single 10. for different frame and spacer bar types.2 2.THERMAL PROPERTIES OF WINDOWS 49 Aschehoug and Baker (1995) measured the frame U-valuesfor different window configurations.7 2.76 1.9 3.6 — 6.6 — 2.7 2. shown in Table 3. air filled Triple glazing.6 5.air filled Double glazing.argon filled Triple glazing. 1995) Glazing system Double glazing.4 insulated Aluminium 2.2 2.9 2.9 — 2.3 1.9 — 3.6 3.8 2.2 4.2 6.3. demonstrate that the frame U-values are generally greater than the centre-glass Uvalues shown in Table 3.2.2.1 6.99 1.4 Frame type/number of panes Fixed Triple 12.2 Typical frame U-values (W/m2K) for conventional windows (Aschehoug and Baker.2.3 2.5 2.8 3.36 1.2. low-e.19 0.3 2.low-e.1 Double Triple 10. Heat is transferred by radiation and also by convection of the surrounding air and ofthe infill gas within the window cavity. i. They have proposed Eq. It can be concluded that better spacers help increase the temperature on the bottom edge of the inside pane and hence reduce the risk of condensation.3 Typical centre-glass U-values for conventional high-performance glazing (Aschehoug and Baker. reinforcedvinyl Wood. argon filled Quadruple glazing. 1995) Frame material Aluminium Aluminium Spacer type Aluminium Aluminium Single 12.78 1.2 — 3. The thermal conductivity of edge-seals as determined by Svendsen and Fritzel (1995) do not consider all modes of heat transfer involved within the window. = UACg + UfAf + Acg + Af (3.6 2.62 . air filled Double glazing.2) Table 3.2) to obtain U0. low-e.4 5.4 — Operable Double 12. (3. vinyl Fibreglass insulated Aluminium insulated Aluminium 4.0 1.4 5. Thus.3 2. low-e.7 — 2.

glazing and frame.01 DGW >2.03 0.04 0.06 0.50 WINDOWS IN BUILDINGS where = length of inner perimeter of window frame.04 0. 1995) Wood/PVC frame 0. 1995) Wood/PVC frame 0.6 DGW 1.05 0.6 DGW 2.00 0.05 0. take account of infihl gas convection and as suchcannot be usedto determine theminimum indoorpanetemperature required to assessthe condensation occurrenceon a window.2 DGW 1.1 0.06 0. can be measured or calculated separately.04 0. Local variations in surface temperature Table 3.05 Metal frame with thermal break 0.04 0.4 TGW 0.4(b) Glazing U-value W/m2K g valuesfor stainlesssteel spacer in W/mK (Frankand Wakili.02 * Double-glazed window ** Triple-glazed window .01 DGW* >2.00 0.0 TGW 1.0 TGW >2.00 0.09 0.06 0. m = due to the combined thermal effects of the çL/g linearthermaltransmittance spacer. therefore.2.07 0. Table 3.00 0.06 0.2.03 0.01 0.02 Table 3.05 Metal frame without thermal break 0.04 0.05 Metal frame without thermal break 0.4(a) Glazing U-value W/m2K values for aluminium spacer in W/mK (Frank and Wakili.7 DGW2.06 0.03 0.08 0.02 0.4 TGW** >2.6 TGW 1.02 0.01 0. W/mK The frame and the centre-glass U-value are independent of each other and.0 * Double-glazed window ** Triple-glazed window 0.07 0.7 0.07 TGW 1.04 0.05 0. however.0 DGW 1. Introducing new spacer materials suchas silicon foam or thin stainless steel profiles can reduce l/Jg by up to 50% and therefore cause a rise of the surface temperatures of the glass edge which will reduce the risk of condensation (Frank and Wakili.08 Metal frame with thermal break 0.4(a)—(c)shows the linear thermal transmittance I'g obtained by Frank and Wakili (1995) for the three different spacer bars. 1995).06 0. None of the above numerical methods for analysing frame and edge-seal heat loss.2.

03 Metal frame without thermal break 0.THERMAL PROPERTIES OF WINDOWS Table 3. 1995.04 0. All multiple-glazedwindowshave one or more cavities. Curcija andGoss.01 0. Any model attempting to quantify local heat transfer rates in the bottom region of a double-glazed window or any model intended to determine the temperature distribution across the face of the glazing must account for both the natural convection of the infill gas and conduction within the edge-seal. 1995) Wood/PVC frame 0. it starts to circulate within the cavity.2.03 0.03 Metal frame withthermal break 0. glass and frame components. (1997) have developed numerical models for the heat transfer at the bottom edge of a double-glazed window. krypton.03 0. When the infill gas is subjected to a temperature differenceacross its boundaries. multiple-glazedwindows.8 *Doubleglazed window **Tripleglazed window 0. L. This is in contrast to previous work that solely addressed the edge-seal conduction effects (Frank and Wakili.00 0.7 TGW 0. 3. Each cavity contains a gas such as air.2 DGW 1. 1994) but this approachis very laborious.01 DGW* >2. 1995. Wright and Sulivan (1995)provide a numerical. In heat transferscience this is known as the conjugate problem. 1995).3. e. Analytical. Svendsen and Fritzel. The infill gas ascends along the hotter pane of glass and descends along the colder pane. Figure 3. These models will be explained later in this chapter. Their method requires a computational fluid dynamics (CFD) procedure to determine the infill gas flow pattern. xenon or sulphur hexafluoride.3 Free convection analysis for an enclosure The free convectionflow inside an enclosedspace is a complexphenomenon. argon. empirical andnumerical solutions are usually required in dealing with suchproblems.04 can be found using very detailed numerical methods (Wright and Sullivan.two-dimensionalconduction analysis for the frame and edge-seal whichcan be extended to account for infill gas motion. Muneer et al.3 TGW** >1.02 0.g. The two panes are maintained at different .1 shows a sectional view of a double-glazedwindow where a gas is contained betweentwo verticalpanes of glass separated by a distance.02 0. natural convection due to circulating gas and conduction through the edge-seal and frame. These models combine the effects of radiative energy exchange between the glass panes.4(c) Glazing U-value W/m2K 51 'g values for foam spacer in W/mK (Frank and Wakili.00 0.

1) where NUL is the Nusselt number based on the window cavity width (L) and k is the thermal conductivity of the infill gas.2) where c and n are constants and RaL is the Rayleigh number (based on the window cavity width.3. 3. L) which is defined as .52 WINDOWS IN BUILDINGS Figure 3.1 The boundaryconditionsfor the analysisof naturalconvection in an enclosedcavity temperatures (Th> Ta). For free convection within an enclosure.1.1. while the remaining walls are shown to be insulated from the surroundings.1 Empirical correlations The average free-convectionheat transfercoefficientwithin the window cavity. Note that at low Nusselt numbers heat transport takes place via effects of gas conduction and convection. is normally represented by the following formula: — NULk L (3. the Nusselt number is generally obtained as follows: NUL = cRa1t1 (3.3. h.3.3.

l.3f) v= p where g gravitational rn/s2 GrL = Grashofnumber(based on the characteristic length) Pr = Prandtlnumber T = cold pane temperature.3. K Tb =hot pane temperature. Practically speaking this limit is reached when the cavity width is 2—4 mm.3b) (3.3.THERMAL PROPERTIES OF WINDOWS 53 RaL = GrLPr= — gfl(Th T)L3 (3. With increasing Rayleigh number.3.l.3.3.3a) GrL TC)L3 (3. The Nusselt numberin this case is NUL = 1.l.l.3. the flow intensifiesand becomes concentrated in thin boundary layers adjoining the two panes.3e) (3. Jrrinfihl gas dynamic viscosity.m2/s = infill gas thermal diffusivity. infihl gas volumetric thermal expansion coefficient.4) For small Rayleigh numbers. Thecore becomes nearly stagnant. Tf T ± Tb (3. m2/s c = infihl gas specific heat at constantpressure. although additional cellscan develop .1. J/kgK acceleration.1. RaL1O3.3. Ns/m2 p=infill gas density. kg/rn3 K' All thermophysical properties of the infihl gas are evaluated at the film temperature.3c) (3.1. K v =infill gas kinematic viscosity.3d) PCp (3. the buoyancy-driven flow is weak and heat transfer is primarily by conduction across the infihl gas.

2 x 106. and 5<Ar<83 • Wright (1990) NUL = Max[Nui. 0.e. A few examples are given below.8) subject to Ar E 110 • Inaba (1984) NUL = Max[l.54 WINDOWS IN BUILDINGS in thecorners and side wall boundary layers eventually undergoing transition to turbulence (Bejan.5Arl10 and where C1 0. 1990).10) where . • El-Sherbiny.1. 0.2S (3.50/(l + (0. 0. I <Pr<2 x 10 and 104<RaL< i0 where Ar is the window aspect ratio.15.3. 0. 0. ratio of window height (H) to cavity width (L) • Raithby.1.3.6) subject to l0<Ar<40.7) subject to F>70° and 3.062Ra/3] (3.046Ra'3 (3.5) subject to 1<A<40.1.1.3.288(RaL/Ar)114.2have been developedto predict the heat transfer within a cavity.Hollands and Raithby (1982) NUL = Max1.Nu2] (3. Many correlations in the form of Eq.1.49/Pr)916)419.3.3.75C1{(sin FRaL)/Ar}'14. 3.9) subject to 5 x 103<RaL sin F<1.3. C = Min[0.271A°'21(RaLsin O.1.l4Pr°°84] F is the angle of the enclosure cavity from the horizontal in degrees.3. i. • MacGregor and Emery (1971) NUL = 0. Incropera and DeWitt.1. 1<Pr<20 and 106<RaL<109 NUL = 0. 0. 1984. Hollandsand Unny (1977) NUL = Max[l.42Ra25Pr0012A03 (3.29C1(sin FRaL)'131 (3.

turbulent or in the transition regime between laminar and turbulent. l996a). computer modelling has.2 CFD work Convection in a sealed glazing unit is complicated as it is affected by many parameters. Figure 3. especially for two-dimensional modelling. Mostcurrent models (Curcija and Goss. temperature differential.242[RaL/Ar]°272 subject to Ar • Muneer and Han (1996a) NUL = Max[1. This assumption reduces the number ofvariables that need to be considered and. Convectiveheat transfer within a window cavity may be laminar.3. 1995a). Muneer and Han.gives all the features of the physicalprocesses.1. Furthermore. h 3.THERMAL PROPERTIES OF WINDOWS 55 Nui = 0. at least qualitatively. however.11) and shows the relationship between h. Thus the control of this parameter will yield beneficial savings in the energy consumption due to space heating load. 1995) of heat transfer for a fenestration system are limited to two dimensions — along the height and depth of the enclosure cavity. the properties of the infill gases.Methods for improving the design of . (3. 1995. 1994. such as the width of the cavity.3. most importantly. the height of the cavity. now gained favour. cavity width and the window height. 1994.1.11) Thecritical path ofthermalnetwork in a multiple-glazedwindow with low-e coated panes is (Muneer and Han. the temperature differential and. the temperatures of the hot and cold panes. Wright and Sullivan.36(GrLPr)O245A28] (3.2 provides design charts based on Eq.76 x l0'°Ra298 subject to RaL 110 Nu2 = 0.3. 0. vective Heat transfer in a multiple-glazingsystem consists of fully coupled conand radiationheat transfer in the cavity of the insulated glazing unit and conduction heat transfer in the solid parts. Due to economicand efficiency advantages. Wright and Sullivan. The usefulness of CFD computer modelling in window research has been proven world-wideby several researchers(Curcija and Goss.3. Traditionally. the exterior window surfaces are exposed to forced (outdoor) and natural (indoor) convective heat transfer.03Ra4134 Nu1 = 1 subject to i04 <RaL x iO iO' + 1. heat transfer characteristics ofmultiple-glazedwindowswere researched using experimental methods.07Ra3 to subject 5 x io <RaL < 10 5 Nu1 = 0.

5m —0.51- f I T I I I 12 I I I T I --*- Air.0 - ' v v I ' v ' v I ' ' ' ' x T I -4-.5m K Argon.—--®---®I Oi3 10 12 14 16 18 I I I 20 22 24 26 28 50 100 Gap width(mm) 3.3.5 2.0 1.5m 0.56 3.0 (c) t3=20°C.5 1..Kryptor 5m —. 2 m Argon.0 1.5 2.5 1. 0._ 1 I L T T I T I T I T I 1 0. t2=—10°C 12:2=:: I W—. t2 = 0°C 2.5 3.51I 8 10 I I I I I 12 14 16 20 22 24 26 28 Gap width(mm) 18 50 100 Figure 3. 0.0 WINDOWS IN BUILDINGS (a) t3 =20°C.0 - (b) t3 = 20°C.5 3.2m 01 8 I I 10 14 16 20 22 24 Gap width(mm) 18 26 28 50 100 -I'.0 2.2 Cavityconvection heat transfercoefficient (W/m2K)..Xenon.Xenon.2m 3. 1 m ___________________________________ T I -4. t2 = 10°C .5 3. Theboxed text indicates the in gas and the cavity height fill . 1 m T I Krypton.Xenon.

Kakac et al.2.3.2.2. 1992.. The partial differential equations which describe the conservation of energy. Zero heat transfer has been assumed through the edge-seal areas.3.3) 0)) (3.3.+ Ov =k1—+I02u (32T 9T'I (3. the warm and cold sides are set as isothermal. T=Th atx=0 T=T atx=L OT —=0 aty=0 andy=H (zeroheatflux) Oy u=v=0 aty=0 y—H x=0 andx=L where u=velocityin x (horizontal) direction v=velocityin y (vertical) direction Table 3. The physical balancesfor laminartwo-dimensionalgas flow can be described mathematically by assuming that the fluid is Newtonian with compressibility and viscous dissipation effects neglected and the fluid properties treated as constantexceptin theformulation ofthebuoyancy term(Shaw.3.3.1 (seecolour plate section) shows . Incropera and DeWitt.1 shows the boundary condition of Muneer and Han's (l996a) model.1) vOv) 0u1 = = + 02u Op (3.3.1 givesthe details ofthe CFD test conditions and the results for the double-glazingheattransferanalysis. Plate 3.momentum and mass at any given point in the flow within a vertical cavity are: lOT OT 0T1 pCp1-5-+u---+v-—1 IOu Ou + u.THERMAL PROPERTIES OF WINDOWS 57 the windowsystem be developed by looking at the distribution of the local can heat flux on the indoor and outdoor window surfaces and the motion of the flow to identify areas with high heat transfer rates. 1985.2. 1990).2) p+ Ox IOv 102v 02v) u—+ v—j — Op —+ pgf3(T—Tf) (3.4) Figure 3.

58

WINDOWS IN BUILDINGS

Table 3.3.1 CFD results for enclosure convection coefficient

(refer to Figure 3.3.6 for notation)
Enclosure gap, mm

h, W/m2K (Han,
20 24
1.99 1.38 1.04

1996)
50

4

8

12

16

28

1.9 1.9 6.25 3.16 2.2 Air,0.6 m height, t3=20°C, i2=0°C 4.05 2.05 1.47 1.31 1.34 Argon, 0.6 m height, t3=20°C, t20°C Krypton, 0.6 m height, t3=20°C, t2=0°C 2.15 1.14 0.96 1.00 1.03 1.21 0.74 0.73 0.76 0.77 Xenon, 0.6 m height, t3=20°C, t2=0°C 6.25 3.14 2.16 1.76 1.64 Air, 1 m height, t3=20°C, t2=0°C 6.25 3.14 2.14 1.71 1.53 Air, 1 m height, t3=20°C, t2=5°C 6.15 3.1 2.15 1.79 1.73 Air, 1 m height, t3=20°C, 2= —5°C 6.3 3.17 2.17 1.76 1.63 Air, 1 m height, t3=25°C, t2=5°C —5°C 6.15 3.11 2.12 1.74 1.64 Air, 1 m height, t3 = 15°C, — — 1.68 1.49 Air, 2.4 m height, t3=20°C, t2=0°C

2.07 2.13 1.4 1.38
1.03 1.01 1.8

0.76
1.67 1.52 1.79 1.65 1.68
— — — —

0.75 0.75
1.75 1.57 1.84 1.7 1.73
— — —

t

1.66 1.85 1.77 1.75
— —

Air, 2.4 m height, r3=20°C, t2=—lO°C Air, 2.4 m height, t3=30°C, t2=—10°C Air, 2.4 m height, = 30°C, 2= —20°C

— —

— — —

— —

— — —

t

1.57 1.19 0.94

temperature and velocity raster plots ofthe convectiveflow within two air-filled cavities. These plots were obtained by Han (1996)as part ofa contractual work undertaken for Living Design, a Glasgow-baseddouble-glazingmanufacturer. The plots demonstrate the transition from a conduction-dominant to convection-dominant regime as the cavity increases from 8 mmto 20 mm width. Plate 3.2 (see colour plate section)which includes velocity raster plots for a range of cavity widths of reinforces this point. The flow structure seems to get more complicated, with a thickening of the inactive core, as the width increases. In effect at widths of 50 mm or over the flow within the enclosure becomes channel flow. Hence further increase in the cavity width will have negligible effect on h. Using the data of Table 3.3.1, the ratio ofthe convective thermal resistance for eachwindow cavity to the convectivethermal resistance for a 4 mmcavity has been plotted against window cavitywidth as shown in Figure 3.3.3. Several interesting points may be noted here. The change from a conductive to a convective flow of heat through the cavity is noticeable as the contour of the line changes from linear to curvilinear. Following this logic it may be noted that for heavier gases the onset of convection occurs at narrower cavities. This pointwill be explained more fully later in this chapter. The attenuation effects of the cavity height may also be noted for air-filled cavities. Once again a theoretical discussionon this pointis provided later. Figure 3.3.4 compares the results of Muneer and Han's work with measured results for eight window
samples.

3.3.2.1 The effect of infill gases on convection heat transfer The useof inert gases can dramatically reduce the convective heat transfer coefficient. The optimum width (corresponding to minimum h) decreaseswith

4.50

4.00

3.50

3.00

2.50

2.00

1.50

1.00

0.50

0.00
15

0

5

10

20

25 Cavity width,mm

30

35

40

45

50

Figure 333 Variation of the ratio of convective thermal resistance of window cavity to the resistance of a 4 mm cavity
(see Figure 3.3.8)

60

WINDOWS IN BUILDINGS
10

-

• Inaba(1984) • El Sherbiny et al. (1982)
A Raithby et al. (1977)

x

*

Wright(1990)

/

1 1

10

100

1000 RaL

10000

100000 1000000

Figure 3.3.4 Comparison of cavity convection models
increasing molecular mass, M. For a given window height of 1 m the optimum width occurs at 8, 12, 16 and 20 mm for xenon, krypton, argon and air respectively (Muneer and Han, 1996a). This behaviour can be explained as follows. The buoyancy force, responsible for the ensuing convection, is

proportional to the density differential between the cold and hot gas (PcPh), in the immediate neighbourhood of the cold and hot sides of the enclosure. Using the ideal gas relationship Muneer and Han (1996b)have shown that

PCPh[_]M
where

(3.3.2.5)

p = gas pressure, kPa R = universal gas constant (= 8.3144 kJ/kmole K) M = infill gas molecular mass, kg/kmole
For fixed p and T values, the above relationship shows that the driving buoyancy force is directly proportional to the molecular mass of the gas. With a heavier gas (larger buoyancy force) the onset of convection, as well as the point where is minimum, occurs at reduced cavity width.

h

3.3.2.2 The effect of cavity width on convection heat transfer
Significant improvements in energy efficiency may be achieved by increasing the cavity width between two glazings. For small enclosure widths of say

4 mm, where almost all the energy is transported via conduction, the con-

THERMAL PROPERTIES OF WINDOWS ductive—convective

61

coefficientdrops sharply with increasing gap width. This is due to conduction heat transfer resistance being directly proportional to the thickness of the conducting material. Further increase in the gap width results in a less rapid decrease in and eventually becomes almost constant.

h

h

3.3.2.3 The effect of window height on convection heat transfer

on h. The decreasein with height has also been addressed by El-Sherbinyet al. (1982). Experimental tests carried out by Rayment et a!. (1992) shed further light on this subject. Muneer andHan (1995b)have shown that for an air-filled window of20 mmcavity widthunder temperature conditions of20°C on the warm side and 0°C on the cold side, for a 1 m high window cavity is 1.64 W/m2K and for a 0.6 m high window = 1.90 W/m2K. The decrease in with height is due to attenuation effects. The effect of window height, however, only applies when convection dominates. In the conduction regime, there is practically no
Figure 3.3.5 shows the effect of the increase in enclosure height

h

h

h

h

effect.

3.3.2.4 The effect of temperature difference on convection heat transfer increases. Muneer and Han By increasingthe temperature difference, (1996a) have also illustrated this point graphically (Figure 3.3.2). The decrease in the optimum gap occurs because when the temperature difference is in-

h

10

20

30

40

50

60

Cavity width, mm

Figure 3.3.5 Reduction of convection coefficientdue to attenuation with increased cavityheight (see Figure 3.3.6)

62

WINDOWS IN BUILDINGS

the flow results in the motion of two increasingly independent boundary layers and this enables convection to dominate conduction effects at
creased,

a smaller cavity width. This phenomenon has also been explained by Elder

(1965) and Gill (1966). The temperature levels have no measurable impact on as long as the temperature differential remains constant(Muneer and Han,

h 1

996a).

3.3.2.5 The effect oflow-emissivity coatingson glazingU-value

A low-emissivity (low-e)

coating on the glass has the ability to reflect the longwaveradiation. In air-filledcavities with uncoated surfaces, the longwave radiation exchange between glass surfaces is high, amounting to about 60% of the total heat exchange across the cavity. With one of theglass surfaceshaving a coating with emissivity less than 0.2 (compared with 0.88 for the uncoated glass surface), the radiation exchange is reduced by approximately 75% and consequently the U-value is reduced (Button and Pye, 1993).

3.3.3 Computer tools for obtaining windowU-value

and Calc3-02 have been developed to respectively deal with analysis for double- and triple-glazing. Calc3-Ola and Calc3-Olb respectively provide approximate and precise solutions for double-glazed windows. The inputs are the internal and external ambient temperatures and the window specifications, e.g. window height, cavity width, infill gas type, pane thickness and emissivityof the coated surface. If more accuracy is needed (see Calc3-Olb.xls,one for each gas), the calculation is initiated by assuming a U-value for the glazing unit under consideration. An approximate U-value may be obtained from Calc3-Ola. Next, the centre glazing temperature of the panes, t1, t2, t3 and t4 (see Figure 3.3.6), are calculated from the followingequations:
Excel workbooks Calc3-01

= T + (1/h0)U[T1— T0] = To + [Rgi + (1/h0)]U[T1— T0]
13

(3.3.3.1) (3.3.3.2) (3.3.3.3) (3.3.3.4)

= T1 — [Rg2+ (1/h)]U[T1 — T0]

= T1 — (1/h1)U[T1 — T0]

The gas properties are determined at tf using an average of 12 and t3. The NUL and GrL numbers are then calculated using Eqs (3.3.1.11) and (3.3.1.3b)respectively.As mentioned previously, if the NUL number is less than unity, NUL= 1. Then hr, and are calculated using the relevant equations in Section 3.2.1. h1 and h0 are given the standard values of 8.6 and 16.7 W/m2K

h

h

THERMAL PROPERTIES

OF WINDOWS

63

'II

ill
Indoor glass —F1 I'll!
Surface#4, ti, C4 Centreline

Outdoorglass Surface#1, ti, €i

liii!
Surface#3,t3,C3

i
III

1:1

ill
Surface#2, t2,E2 T0

T

!

T

1/h
q

Rg2


t3

________
L.....1

(


t2

Rgi


tI

i//Iso

• T0

i/hr__f.....J

Figure 3.3.6 Centre/me thermalcircuitfor double-glazed windows

respectively. Finally, the U-valueis calculated from Eq. (3.2.1.2). The processis continued iteratively until convergenceis achieved in the U-value.

Example 3.3.1

An argon-filled double-glazedwindow with a 12 mm cavity and 0.4 m height has one low-emissivitycoating, C2 = 0.12 (see Figure 3.3.6). The internal and external ambient temperatures are 20°C and 0°C respectively. The thickness of each glass pane is 4 mm and the glass thermal conductivity is 1 W/mK. Compute the centre-glazing U-value using the approximate and precise
methods. The respective solutions are provided by workbooks Calc3-Ola and Calc3OlbAr.xls. Launch workbook Calc3-Olaandactivate the sheet 'Ucg-Ar'. Insert the given data in cells C4:Cl4.The required centre-glazing U-value is 1.56 as shown in Cell E4. Note that Calc3-Ola provides an approximate solution. The Calc3-0Ia.xls workbook also enables calculation of the edge, frame and overall U-value. Activate the 'Ueg,Uf&Uo' sheet by clicking its tab. Provide the required input data in the cell range B3:B7. The requisite outputs are produced in the cell range B8:B14. Launch the workbook Calc3-OlbAr.xls.This workbook provides a precise solution. Activate the sheet 'Argon'. Insert the given data in cells C4:C18 as shown. The computed U-value is 1.54 (Cell E4). Note that the results obtained by both methods areapproximately the same. Thus, even an approximate solution would be adequate, at least at the preliminary design stage. The Calc3-Olb family of workbooks also enables the

64

WINDOWS IN BUILDINGS

of the effect of wind speed on U-value. Within Calc3-Ola the external wind resistance is set to the normal exposureconditions recommended by CIBSE (1986).
investigation

Example 3.3.2

Plot thetemperature contours across the cavity for the followingdouble-glazed windows. See Figure 3.3.6to followthe notation. Note that a further visual aid is provided within Calc3-Olb.xls. Air filled, float glass =0.88, 2 =0.88, =0.88, 0.88) Air filled, one low-c (c1 =0.88, c2=0.12, 83—0.88, 84—0.88)

(i

Argon filled, one low-c (c=0.12, 82=0.88, 83=0.88, 84=0.88) Argon filled, one low-c =0.88, 82=0.12,83=0.88,84=0.88) Argon filled, one low-e =0.88, 62=0.88, 83=0.12, 64—0.88) Argon filled, one low-e (6=0.88, 62=0.88, 63=0.88, 64=0.12) Kryptonfilled, one low-e (2 = 0.88, 62 = 0.05, 23 = 0.88, = 0.88) Xenon filled, one low-e =0.88, 62=0.05, 83=0.88, 64 ocr 0.88)

( (i

(i

6

Note that manufacturers claim respective emissivities of 0.12 and 0.05 for Sn02 and Ag coatings. The solution has been obtained by using Calc3-Olbfor each of the above cases. Figure 3.3.7 shows the temperature contours for all cases. It may be

oir,flo41 glrr.c Air.low-c )rrO 2) Xc,low-c (e=O05) Ar. low-c )rrwo.12)
Kr. low-c )O=O 05)

Ar, low-r(r.wO 12) Ar. low-c (e0.l2)

N 4c
0)

ttc

I
0

I
41)

Figure 3.3.7 Temperature variation along window depth (see Example 3.3.2 and Figure 3.3.6). Note that the bottom two cases in the box are presented by a commontemperature profile

THERMAL PROPERTIES OF WINDOWS 65 noted that a significant improvement in the temperature of the inner glazing surface results with the use of low-e.8°C. In cold climates.9.88.83=0.88. Thevalues oft4for theabove cases arerespectively 9. If the walls are assumed to be at the temperature of cases: the internal air (ta) find the dry resultant temperature.88) Case 5: Air filled.88. one low-e (6=0. so the energy absorbed by the low-e coating is conducted to the outer surface and removed away by the external air. tres.1. 830. The wind speed is assumed to be 3 m/s. 11. Next. one low-e (c=O.xls.84=0. 63=0.1.0 m height with a 12 mm cavity. + tr= At At4 where A is the wall area.9.8.88. 62=0.6 for notation). (3. 11. 64=0.88.12.8 and 9. 9.88) Case 4: Air filled.88. is to be calculated using Eq.3. The calculated value of 4-es for the third and fourth cases is 14.3. Activate the sheet 'Air'andinsert the given data for case 1 as shown in Example 3.88.12) The workbook Calc3-OlbAir. It is not surprisingthat tres has notchanged significantly due to the fact that thewindow area is much smaller A t . The bedroom is 2.88. float glass (8i=O.4 m. one low-e (6i=O. 83=0.3 m and a length of 3. the low-e coating should be placed on surface number two of the double-glazed window (see Figure 3. the mean radiant temperature (tr) is to be calculated for each case by using the formula. UK has a height of 2.3 A singlebedroom in a detached house in Edinburgh. Example 3.84=0.3).9°C. Finally.l2. the window area. Refer to the schematicdrawing within sheet 'Air' of the workbook Calc3OlbAir.3.88. 82=0. 84=0.0 m width and 1. The following data are provided: Lg12'Lg344mm and = kgl2 kg34 1.9°C while for the remaining casesit is 14.88) Case 3: Air filled. In places where solar energy is not wanted. 82=0. the dry resultant temperature. 63=0. and is the wall surface temperature. Assume that the internal and external ambient temperatures are 15°C and 0°C respectively. 840. 82=0.xlsis required to obtain the centre-glazing temperature (t4). Then activate the sheet 'CALC' and record the value oft4(cell 08).12.88) Case 2: Air filled.88.88. the low-e coating should not be placed on surface number three as it is easier for the energy absorbed by the coating to transfer into the room ratherthan across the cavity to the outside.7 m wide. 82=0.0 W/mK. for the following Case 1: Air filled. Along the width of the bedroom is an air-filleddouble-glazedwindow of2.88. one low-e (8=0. inert gas-filled windows. tres.

5.0.25 Wind speed.75 1. 1993).66 WINDOWS IN BUILDINGS of the room walls. In the above section it was demonstrated that two factors. i.0- op Middle 0. convection within cavity and frame conduction as well as edge effects.4 Temperature stratificationin windows It is customary to assume that the interior glass pane of a window is at a uniform temperature.2. ignores the effect of infill gas convection within the window cavity and its impact on the temperature distribution ofthe glass pane. contribute to a nonuniform temperature distribution.25 I 1.5. This assumption. This stratification dictates that the lowest temperature of the interior pane would be experiencedat its lower edge where condensation initiates.5 1. there would be a noticeable difference in tres. Muneer et al.e. if the window area was much larger. however. It is therefore logical to expect a non-uniform heat flux through any double-glazing.8 Flux distribution over an air-filled double-glazing .8.0 - 1.75 2.2. compared to the area 3.A sample of theflux measurements undertaken by Han (1996) is presented in Figure 3.3. Condensation prediction charts (Buttonand Pye.3. The plot shows increased flux at the bottom end of the glazing which inevitably comes into contactwith the cold air descending along the outer pane.25 I Bottom 0. The interior pane temperature is normally assumed to be equal to the centre-glazingtemperature. are produced under the simplified 3. rn/s Figure 3. that are generally used to predict condensation occurrence. (1996)and Abodahab and Muneer (1998) have shown the invalidity of this assumption and demonstrated the temperature stratification effect along the height of the inner pane. However.I 0.

68 + [1 O. T.1) where y is the critical height at which transition occurs. The Nusselt number and Rayleigh numbers are based on the height of the vertical plate as the characteristic dimension. The temperature difference at any point y may then be obtained by the following equation (Incropera and DeWitt. The density of the air close to the plate is less than that which is further removed. The charts use the following parameters to predict the condensation occurrence: inside air temperature. therefore.2) In the case of a vertical plate with a uniform flux (qp — constant) the temperaturedifference T = T9 —T) will vary with y. 3. Then. Churchill and Chu (1975) have shown that without loss of any serious accuracy the correlations obtained for the isothermal plate maybe used for the case of constantheat flux. Bejan (1984) has shownthat transition from laminar to turbulent flowwithin the boundary layer occurs when a T.the temperature at the midpoint ofthe plate T(H/2) may be obtained in an iterative manner. This will be demonstrated via the followinganalysis. ifNu and Ra are defined in terms of the temperature difference at the midpoint of the plate AT(H/2) and using the relationship i — qp AT(H/2) along with Eq.4.1.4. (3.4. indoor relative humidity. Buoyancy forces.THERMAL PROPERTIES OF WINDOWS 67 assumption of isothermal plate temperature. As pointed out above the assumption of uniform pane temperature is invalid. induce a free convection boundary layer in which theheatedair rises verticallyandis replaced by air from the quiescent region. Assume the plate surface temperature.4. and U-value of the glazing.1.2).492/Pr)9116]419 (3. H For laminarflow.1. outside air temperature. increasingfrom a value of zero at the leading edge.1 Free convection along a verticalplate with a uniformheat flux Consider the case of vertical plate immersed in a quiescent fluid such as air. Nu 0. Ra gfl(T —T)y3 i09 (3. 1990): .670Ra'4 + (0. to be greater than air temperature.

1998) Design details Window number Trade name g* 2.83 1.4.Table 3.1 Window configurations (Abodahab and Muneer. 200 and T0 = 0°C (W/m2 °C) .14 0.53 1.72 1 2 3 4-Airl2-4 4-Arl2-4 4-Airl2-E4 4 4-Ar12-E4 4 mm float glass + 4 mm float glass + 4mm float glass + 4mm float glass + 5 4E-Krl2-E4 1. calculated at the glazing centre with T.91 6 7 4E-XelO-E4 4 + Air2O-4 12 mm air gap + 4 mm float glass 12 mm argon gap + 4 mm float glass 12 mm air gap + 4 mm low-emissivity coated glass 12 mm argon gap + 4mm low-emissivity coated glass 4 mm low-emissivity coated glass + 12 mm krypton gap + 4 mm low-emissivity coated glass 4 mm low-emissivity coated glass + 10 mm xenon gap + 4 mm low-emissivity coated glass 4 mm float glass + 20 mm air gap with 50 mm high baffle placed centrally on the bottom edge + 4 mm float glass 2.86 2.82 *Urer: The reference U-value.

1 Design schematicand nodalarrangement window design consideredby Muneeret al. Note: all dimensions are in mm and k values in W/mk .1. A detailed cross-section of double-glazed window edge-seal is shown in Figure 3.4. It also shows the corresponding temperature of the glazing Centreline - B61 for Figure 3.l5(_)ATH/2 (3. These may. (1997).2 shows the variation of the measured and computed temperatures against the window height for three window samples. They have also shown that there is a decrease in temperature at both the bottom and top edges of the window. be thermally brokento avoid excessive conduction heat transfer.4.4.1 shows the specifications of the double-glazedwindowsthat were tested by Abodahaband Muneer (1998) to investigate the temperature stratification along the height of the inner pane.4. Figure 3.4. The window edge spacers are usually made ofaluminium. however.3) Table 3.1.THERMAL PROPERTIES OF WINDOWS 69 AT = 1. Abodahab and Muneer (1998) have shown that the temperature along the vertical centreline of the window inner pane increases as the height increases.

1 0. 1998) .2 0.3 A .25 0.2 0. m 0. 4E-Kr12-E4 0 0.4 I 10 0.45 _______________-- Window height.35 .4.35 0.3 0..05 0.2 0.45 Window height.35 0.m 24 22 20 V 18 a 16 g I14 12 0 0. T-measured I IIIT-computed I IA T-Pilkjngton 0.05 0.25 Window height. m Figure 3.15 0. A A A A U 15 14 13 12 11 .1 0.3 0.4 I IAT-Pilkington 0.4 0.•::''' •• 0.15 0.15 0.45 12 20 19 18 17 16 AAAAAAA I .2 Measured and computedtemperature variations for three windowconfigurations (Abodahab.05 0.25 0.1 0.70 WINDOWS IN BUILDINGS 22 21 AAAAAAAAAAAAAAA64AA 20 19 U 18 . a 13 I 4-Ar12-4 0 •T-measured T-computed 0.

°C y = height of the arbitrary point from the window bottom edge. andcold sides.5) where = dimensionless temperature t.1. At the bottom of the cavity this cold gas turns its direction and comes intocontactwith the bottom edge ofthe indoorglazing. Plate 3.4.5 and 1°C it was shown via measurementsthat there is an average increase of 2.2 demonstrates the interplay of the two factors which determine the longitudinal temperature variation. For the respective indoor and outdoor conditions of 15. = temperature of the inner glass pane at an arbitrary point. The infihl gas in the sealed space flows upward along the indoor glazing and In order to compare the temperature variation in diverse window configurations with disparate geometry.3 (see colour plate section) compares those infrared images with a photograph of a double- .4) (3. Note that the latter work was carried out on behalf of Pilkington plc. Figure 3. internal temperature and external weather conditions.1.5°C between the highest and lowest longitudinal temperatures. For most double-glazeddesigns this is the dominant factor and it is responsible for the rise in the temperature along the height. two dimensionlessparameters T and * are introduced: (3. The cold bridging effect of the spacer is partly responsible for the temperature drop at the bottom and top edges of the glazing.. The descending gas becomes progressively colder.alternating betweenthehot downward along the outdoor glazing.4.4. m = dimensionlessheight Abodahab(1998) has shown that the temperature ofthe inner paneincreases with the window height. These are: • Thecirculation ofthe gas within the enclosure. This effect has also been demonstrated by Frank and Wakili (1995) and Wright and Sullivan (1995). • The conduction effect due to the metallic spacer which acts as a thermal bridge between the cold external environment and the warm internal environment. Infrared photographs of ordinary double-glazed and superinsulated windows were takenat the LawrenceBerkeleyLaboratory in California.THERMAL PROPERTIES OF WINDOWS 71 bottom as obtained by Button and Pye (1993). Later in this chapter. a two-dimensional temperature model that involves cold bridging effects of the edge-seal will be presented.

4)).2. Only a précis of that publication is given below.72 WINDOWS IN BUILDINGS glazed window in an experimental bedroom.[q/k]y[RaPr]'15 where the expression for uniform heat flux Rayleigh number is (3.2 Physicalmodel for longitudinal temperature profile ofmultiple-glazed windows This work has been presented in an article by Abodahab and Muneer (1998). AT= c{q/k]y[RaPr]m where m is a constant and c is defined by Eq. (3.1) where 5 is the thickness of the thermal boundary layer and y is defined above.8). It can be seen that condensation occurred also on the side edges of the window due to cold bridging by the edge spacer. A generalised form of Eq.4. The infrared and photographic images of the double-glazed window clearly demonstrate the above-mentioned phenomenon ofgas circulation and its effect of lowered temperature at the bottom edge of the cavity. The physical situation of a heated glass wall is neither isothermal nor constant heat flux.2. (3. For gaseous fluids such as those used in window enclosures.2) was given by Abodahab and Muneer (1998): .1.4) zT— ty — (Tf—273.2. Bejan (1984) has used scaling law analysis to show that ö y[RaPr]'5 (3.5) A new parameter is introduced: . Also iXT.2) Ra == g/3y4q/cvk (3.4.3) All physical properties Pr. The room was occupied with one adult person the night before the photograph was taken.4. The photograph shows water droplets forming at the bottom edge ofthe window. (3. v.4.15) (3.2. the latter condition is closer to reality for heat transferthrough a double-glazing system.2. (3.4. and k are evaluated at the enclosure film temperature Tf (see Eq.4.2.2.4. However.3.4. 3.

1°C.8) into Eq. Note that Calc3-03.4. a temperature stratification of2.4. Abodahab and Muneer (1998) have shown that the semi-analyticalmodel presented by Eq. e. Example 3.5) and (3. Na = L.2.xls. Becausethe above model is not applied for the bottomand top edges of the glazing.01°C and a root mean square error of 1.g. The bottom and top temperatures are respectively9. Assume that the internal andexternal ambient temperatures are 20°C and 0°C respectively. A definition for UrefS provided in Table 3.0 W/mK. an average height of 25 mm has to be discarded from the bottom and from the top as shown in cells D6 and D25 respectively.2.2. Substituting Eqs (3. (3. c0 and c1 are given numericalvalues. c1 = —0. (3.4.9 and 12.2.2.7) = C0 + CUref (3.2.xlsprovides a solution only for air-filled windows. Refer to the schematicdrawing in sheet 'Main' of the workbook Calc3-03. For other gases the reader may wish to make suitable modifications in the 'Caic' sheet of the above workbook.86 W/m2K.6) Na = c[RaPrIm where C (3.2°C. (3.4) then becomes (3. Given data: Lg12=Lg34=4mm and kgI2=°kg34= 1.0 m height has a Uref of2.4.4. These valuesare c0= 1.4.1563.e.4) and rearranging the terms yields = 4+ (qy/k)(c0 + c1Uf)([gf3y4qpr]/{vk])m (3.4.4.xls workbook.2°C. Compute the temperature stratification along the height of the inner pane.4. 20 locations as shown in the cell array D6:D25.1.2. The window height is divided into equal segments.3).2. Activate the sheet 'Main'.1 An air-filled double-glazedwindow of2.4.THERMAL PROPERTIES OF WINDOWS 73 Napier Number.2.4.9) estimates the longitudinal temperature distribution for any multiple-glazedwindow with a mean bias error of 0. Insert the given data in the cell array C5:C12. The resultant temperature stratification is shown in the cell array F6:F25.31 and m=—0.8) m.2. .Tk/(qy) Equation (3. i.2. The solution is provided by the Calc3-03.9) This equation represents the physical model which may be usedto compute the longitudinal temperature variation of the inner glass pane of any multiglazed window. The followingexample shows the relevant calculation for a particular case.4.

3. argon. As mentioned previously.1 consists of an aluminium spacer.1. then the insulation benefits of the glass can be significantly diminished.g. Double-glazedwindows containing high-performance glass require a compatiblehigh insulation in theiredge spacer and frame.2 shows a comparison of the temperature difference between the centre glazing and the bottom node (node 3) for the two scenarios. plastic. which consist of two panes of glass separated by an edge-seal. also decreases with a lowered U-value. 1993).3. While a conventional double-glazedwindow with low-e and argon infill gas achieves U-values between 1. desiccant. an aluminium and a perfectly insulated spacer. Reference is now made to column 5 of Table 3. i. and the frame. It was shown above that the thermal transmittance of a complete doubleglazed window depends on three components: the glazing unit. the descendinginfill gas would approach the temperature of the cold outer glazing. Once again an increasing differential between t3 and T3 is noticeable between cases 1 and 2. (1997) to study the cold bridge effects ofwindow edge-sealand frame on the temperature distribution along the height ofthe inner glazing.g. Column 4 shows an increasing trend for the above temperature differential. The above two comparisons demonstrate the effect of higher resistance to the heat flow due to convection. This conclusion can be verified by Eqs (3.e.2. krypton. The temperature differential (t3—T3) also increases from .8 to 7. The above referenceprovides further details of the numerical model andits experimental validation.4. Ifthis is not achieved.4. The edge-seal shown in Figure 3. poorly insulated frames ofwood. Thus T3.2) and (3. and polysulphide. A two-dimensionalnumerical technique was used by Muneer et al.The edgeseal isolates the space between the two glass panes and consequently creates an insulative cavity suitable for the use of low-emissivitycoatings and/or lowconductivity infihl gases.3 Multimode thermalmodel Modern double-glazed windows use hermeticallysealed. Note that the position of all nodes is shown in Figure 3.4. The indoor centre-glazingtemperature t3 increases and the outdoorpanetemperature t2 decreases with the decrease of U-value due to increased cavity thermal resistance. and between cases 3 and 4.4. the bottom temperature of the inner pane. normal. aluminiumor steel may give U-values ranging from 2. polyisobutylene. and severe. insulative glazing units. The thermaleffect of the edge-seal for different conditions of exposure. This increases the overall U-value of the window and lowers the local temperatures at the edge region.8 W/m2K.74 WINDOWS IN BUILDINGS 3.0 W/m2K (Button and Pye. e. The presence of the aluminium spacer creates a thermal short circuit at the window edge-seal. may be calculated from tables given in Iso 6946 (1983). or xenon. Therefore.3. the temperature differential (t3—T3) will increase quite rapidly due to the combined effect of the elevated t3 and lowered 1'3. Table 3.3).3 and 1. e.3. This has also been demonstrated by Frank and Wakili (1995). respectively shown in columns 4 and 5. sheltered. This phenomenon can be explained as follows.4. the edge-seal (spacer).

0°C respectively. (B) (t3—T3) is the temperaturedifferencebetweencentre-glazingand node 3.7 2 3 4 5 6 4+Ajrl2+4 4+Argl2+4 4+Ajrl2+E4 4+Argl2+E4 4E+Kr12+E4 4E+Xe10+E4 3.14 0. case 1 to 3 and from case 2 to 4 due to the increased radiative resistance between the respective cases. its relative weighting is only 17%. The relative weightingof the edge effect in this case is 60%. in absolute terms.8 2.83 1. Several counteracting influences may be identified.radiation° only 2.86 W/m2k. (C) Temperaturevalues calculated by the numericalmodel. (D) Temperature values calculated by the numericalmodel. a comparison between columns 4 and 5 illustrates the effect of the conductivity of the edge-seal. for super-insulated windows.53 1. thus tending to bring down T3. All heat transfer modes are involved.5 5. the descending colder gas gets cooler with decreasing U-value.91 Convection.0 5.4 5. Given L512=Lg34=°4 mm. Cold bridge effects are excludedby replacing the aluminium spacer by an infiniteresistance. Thirdly. e.2 Cold bridge effects demonstrated'via the difference between centreline and bottom edge temperature(Muneer et al. Column 5 provides the values of (t3—T3) for a hypothetical situation wherein the aluminium spacer resistance is replaced with an infinite resistance.5 2. radiationC 3. Example 3. The situation in cases 5 and 6 is very involved.4 m height and 1 m width uses an aluminium spacer.9 3. The centre-glazing U-valueis 2. .g.xls wherein a schematic of the window structure is shown.72 1. It is clear that for an air-filled window (case 1) the edge conductivity has a minor influence.THERMAL PROPERTIES OF WINDOWS 75 Table 3. T3 may not be lowered. Refer to the sheet 'Main' of the workbook Calc3-03. for the low-emissivity xenon windows the spacer conduction is the major contributorin reducing the edge temperature. However. Secondly. Assume that the internal and external ambient temperatures are 20°C and 0°C respectively.4. 0. Firstly.9 4. 1997) (t3_T3)B (t3—T3) UvalueA Case no. (t3—T3) decreases due to increased thermal contact between these nodes for heavier. This is due to the fact that in the latterdesign t3 will be higher by a wide margin.node 3 loses less heat by radiation and therefore tends to elevate T3. 1 Windowtype Conduction.86 2.2 A double-glazed window with a 12 mm air gap.3 (W/m2K) 2. It should be borneinmindthat eventhough(t3—T3) increases.. convection.4. For any given case.Cases 1 and 6 lie at the extreme ends of the comparative exercise.4 5. inert gasesas shown by Muneer and Han (1996b).3 2.4 (A) Internal and externalambient temperatures are assumed20°C.

1. The temperatures of all nodes. The assumption is valid as the thermal resistance of the external film of air is high compared with the internal thermal resistances within the building. Activate the sheet 'Air'and insert the given data in cell array B4:B16. This approach assumes that the complete building is a single unit that loses heat to the external environment during the coolingdown period. Super-windows may also experience this phenomenon due to the fact that the temperature of the outer pane drops as a result of the low U-value.The sheet 'CalcAir' containsthe thermal resistance network related to Figure 3. To quantify the condensation problem encountered on windows. a simplified model for assessing the nocturnal temperature drop inside a building subjected to a daytime heating schedule has been provided by Muneer and Abodahab (1998).3.6°C.1. and high relative humidity.xls. This temperature is 9. 1993) are generally used to predict condensation occurrence.4.The iteration has been checked by performing an energy balance for each finite element as shown in the cell array B64:N68.1 Model for assessment of the potentialof condensation occurrence Condensation prediction charts (Button and Pye. They are produced under the assumption of isothermal plate temperature as pointed out in Section 3.1) using sheet 'air' of workbook CaIc3-04.4. Frequency of condensation occurrence on double-glazed windows 3.xls.4. Condensation may occur on the outer pane of a window if its temperature goes below the outer dew-point temperature.76 WINDOWS IN BUILDINGS kg12kg34=l. This happens when thereis a clear sky. From Newton's law of cooling it may be shown that ln(T1_T0)=_(i'4)r (3. It initiates at the coldest part of the glazing at window's bottom edge and then propagates upwards as the external temperature and/or the internalhumidity increases.88.5 It is common experiencethat within dwellingscondensation occurs on glazing surfaces during the early hours ofthe morning.5.This phenomenon was demonstrated by Plate 3. find the temperature at the bottom edge of the glazing (node 3 of Figure 3.1) .4.5. shown in Figure 3. The answer to this example can be obtained via the workbook Calc3-04. have been iteratively calculated as shown in the cell range A33:U61. 3.1.O W/mk and 2=E3=O. The required temperature t3 is given in cell D5 of the sheet entitled 'Air'.

The work of Webb and Concannon (1996) has shown similar behaviour. These two properties enable the determination ofthe indoor dew-point temperature by using the principles of psychrometry.1) is called the time constant.5.2) To determine the potential frequency of condensation occurrence on a building window the internal air temperature. The temperature of the building structure.e. seconds T0 = external air temperature at a certain time. wind direction. the longwaveradiation loss to the clear sky. m2 c = building-specific heat. can be affected by the external air temperature.kg/rn3 at T1 = buildinglumpedtemperature a certain time. with the bottom edge inner glazing temperature. should be known. creases with The annual average of condensation occurrence for any given location dea reduction in the U-value. By comparing the internal air dew-point temperature. wind speed.1.1 gives a relationship between T. A large time constantmeans that the building requires a longer time to cool down or heat up between specific temperatures. °C = elapsedtime. infiltration and/or ventilation. Td. During the unheated period. (3. Relative humidities within a single-occupancybedroom were measured by Muneer and Abodahab (1998) for the above-mentioned experimental bedroom. This is due to the thermal capacity of the building structure that dampens the influence of the changing weather over a short period of time.5. and T0 for the bedroom during the nocturnal cooling-down period. and relative humidity. the building structure cools down due to the influenceof the external environment. J/kg K h = heat transfer coefficient from the W/m2K building surface to external air.5. Thus. By knowing the external ambient temperature T0. the internal ambient temperature T can be obtained using a regression of the form T1=aT0+b (3. the heating system is switched off and the house is cooled down by the external environment. °C Theterm (pVc/hAb) in Eq.1. Figure 3. when there are no solar effects. T0. in general.5. It was thus shownthat the relativehumidity changes only slightlyduring the cooling-down period. t3. p = building density.1 shows that the potential . an assessment of the occurrence of condensation may be made. i. Table 3.THERMAL PROPERTIES OF WINDOWS 77 where Ab = buildingsurfacearea. T. if t3<TdPcondensation will occur on the glazing edge. 4.

c) a) U- D 'L .S 2 E 0 0 •0 q) Q) en 1 It.

1 Averaged annual potential (Abodahaband Muneer. window No.3 a more realistic assessment of the potential of condensation . 1998) Year Window no.xlsand Calc3-05. Initiate the calculations by running the macro thus: press 'Alt+F8' then click the 'Run' button in the Macro dialogue box. kg/kg of dry air).g.8°C. launch Excel. The indoor relative humidity is 70%. 7 Edinburgh 117 79 of frequency of condensation occurrence Manchester 97 88 London 58 50 44 38 31 28 14 106 100 90 78 71 82 70 61 46 55 34 frequency of condensation occurrence will decrease by about 40—50% if less efficientglazing(e. however. m3/kg). shown that by using the work described in Section 3.xls.4. Example 3. The former workbook is usedto obtain the temperature at the bottom edge of the glazing (13).2. The following psychrometric properties: humidity ratio (Wa. kJ/kg). specific volume (v.xls within this environment. 3 Window no.then open Calc3-05.5. as shown in Example 3. It is worth mentioning that the above DLL file provides a useful facility to compute the psychrometricproperties for any numberof inputdata.g.2. Because the workbook 'Calc3-05. 6 Window no.Insertthe indoor air temperature (DBT = 12°C) and the relative humidity (RH=70%) in cells ClO andD10 respectively.1 Window no. 13 was found to be 5.The required value for DPT is foundto be equal to 6.. The solution is provided by workbooks Calc3-04.dll'. windowNo. °C)andwet-bulb temperature (WBT. °C) are calculated in the cell array F10:J10. dew-point temperature (DPT. Traditionally the potential for condensation occurrence on windows was based on the centre ofglazing temperature (Carmody et al.1 For the window of Example 3. 5 Windowno.4.7°C.4. Of course these results apply for the bedroom underdiscussion but may be taken as indicative of the domestic scene in Northern Europe.THERMAL PROPERTIES OF WINDOWS Table 3. 1996). 6).xls. specific enthalpy (h. Muneer and Abodahab (1998) have. Find the potential for condensation occurrence at the bottom edge of the window.xls'contains a macro entitled 'Prog72. Thusit maybe concluded that condensation will occur on the glazingedge as 13 is less than the dew-point temperature. The indoor and outdoor air temperatures are respectively 12°C and 0°C. consider the early-morningcondition of an occupied bedroom.5. as shown in Calc3-05. 4 Window no.as shown in cell 110. 2 Window no. 1) are replacedby the most efficientones (e.

0.45 0.1) uses a 50 mm high glass baffle.5.2 Innovative developments As mentioned the previously. The identificationof this phenomenon has led to an innovative development (Muneer et al.. placed mid-way in the cavity on the bottom spacer.75 0.4.2 shows the benefit achieved by the use of an in-cavity baffle.5. It therefore predicts the onset of condensation with a much higher accuracy.65 0.2 Temperature distribution along the height of indoor pane of windows with and withouta baffleplate (Han. This helps in keepingthe temperature at the bottompart ofthe inner pane at a higher value than would be otherwise. Provided the internal and external environments are maintained at 20°C and 0°C. an average increase of 2°C edge temperature is achieved by the above innovation.1 shows that the window under discussion is least susceptible to condensation occurrence. 1996) . 7 of Table 3. 1996). Figure 3. The baffle plate constricts the descending cold air hugging the outer glass pane from hitting the bottom part of the inner glass. Table 3.5.5.7 0.6 a55 0.80 WINDOWS IN BUILDINGS occurrencecan be made. 3. circulating infill gas within the window enclosure is responsiblefor thelowered temperature ofthe bottompart ofthe inner pane. The innovation (window no. Recall that the latter procedure takes into account the temperature stratification effects of the inner glass pane due to the circulation of infill gas and the cold bridging of the spacer.5 0.met Figure 3.4 100 200 300 400 500 600 Height.

Edinburgh. (1995) Linear transmittance of different spacer bars. ASHRAE (1993) ASHRAE Handbook ofFundamentals. Proceedingsof the Window Innovations '95 conference. (1996) Residential Windows. and Heschnog. Edinburgh. J. W.THERMAL PROPERTIES OF WINDOWS 81 References Abodahab. T. 2. Norton & Company. JournalofFluid Mechanics26. El-Sherbiny. UK. N. Energy Conversion and Management 39. J. (1996) Investigation of Thermal Characteristics of Multiple Glazed Windows. and Baker. (1995) Frame and edge-seal technology — an international view.American Societyfor Heating. G. PhD thesis. mt. Frank. Elder. 257. J. B. Proceedings of the Window Innovations '95 Conference. and Muneer. P. Atlanta. Trans ASMEJ. London. 0. and Raithby. JournalofFluid Mechanics 23. PhD thesis. 244. p. Button. S. Selkowitz. USA. (1994) Two dimensional finite element model of heat transferin complete fenestration systems. Wiley. E. B.Canada. Napier University. and Goss. Refrigeration and Air Conditioning Engineers. and Chu. McGraw-Hill. T. (1984) Convection Heat Transfer. T. K. Bejan.. D. London. Gill. (1998) Temperature Distribution Models for Double-glazed Windows and their Use in Assessing Condensation Occurrence. P. (1975) Correlating equations for laminar and turbulent free convection from a vertical plate. Heat & Mass Transfer 18. (1966) The boundary layer regime for convection in a rectangular cavity. Napier University.American Society for Heating. D. 1127. K. Int.H. H. W. Thermal properties of building structures. 515. J. Refrigeration and Air Conditioning Engineers. and Wakili. Curcija. New York. W. p. CIBSE (1989) CIBSE Guide A3. UK. W. Han. 253. (1965) Laminar free convection in a vertical slot. A. S. USA. . D. New York. A. Inaba. 99. Aschehoug. (1973) Thermal Comfort. Fanger. (1982) Heat transfer by natural convection across vertical and inclined layers. British Standards Institution (1989) BS 6993. J. M. Toronto. G. UK. Atlanta. ASHRAE Transactions 100. 1323. Part I. ASHRAE (1985) ASHRAE Handbook ofFundamentals. (1984)Experimental study of naturalconvectionin an inclined layer. London. Churchill. S. (1993) Glass in building: A Guide to Modern Architectural Glass Performance. Toronto. Abodahab. W. 0. Heat Mass Transfer 27. Heat Transfer 96. and Pye. Carmody. (1998) Free convection analysis of a window cavity and its longitudinal temperature profile. CIBSE (1986) CIBSE Guide to Current Practice Volume A.. N. L. New York. Pilkington Glass Ltd. Hollands.

eds: K. M. D. (1996) Gas flow in window enclosures and its effect on temperature distribution. T. Abodahab. Southampton. B. 1463. Brebbia. B. K. Muneer. W. New Orleans. Hollands.82 WINDOWS IN BUILDINGS Incropera. B.. (1993) Human Thermal Environments: The effects of hot. J. Building Services EngineeringResearch & Technology (BSER&T) 17. Muneer. 223. (1995b) Environmental impact evaluation of high performance windows for sustainable buildings. Glasgow. and Han. Energy Conversion & Management 37. New Orleans. eds G. T. 183. and Viskanta. 287. ProceedingsASHRAE.. Raithby. (1990) Fundamentalsof Heat and Mass Kakac. Muneer. T. DC. Italy. and Han.. A. In ComputationalMethods and Experimental Measurements VII. UK. B. . P. Ito. Porteous.Part 1 — Steady state thermal properties of building components and building elements. N. convection. and Han. (1971) Prandtlnumber effects on natural convection in an enclosed vertical layer. BuildingServices EngineeringResearch & Technology (BSER&T) 18. C. P. (1996a) Design charts for multiple glazed windows. D. N. Proceeding International Workshop on EnvironmentalImpact Evaluation of Buildings and Cities for Sustainability.fundamentals MacGregor. (1994) Heat transmission characteristics of multiple glazed windows-measuredand modelled results. F. ISO (1983)Thermal insulation calculation methods. and Oka. (1972) A field study on the convectiveheat transfercoefficient on exterior surface of a building. p. ProceedingsofAdvances in Fluid Mechanics (AFM '96) Conference. 233. Muneer. and Gilchrist. 11—13 June 1996. Energy Conversion and Management 39. T. 253. Computational Mechanics Publications. 717.. Heat Transfer May. MacGregor and C. G. T. (1998) Frequency of condensation occurrence on double glazings in the United Kingdom. N. A. (1995a) Use of CFD for thermal analysis of double glazings. UK. and Emery. andapplications. (1997) Combined conduction. and Abodahab. J. (1977) Analysis of heat transfer by natural convection across vertical fluid layers. Semi-annualmeeting. Muneer. K. N. T. B. R. T. Carlomagno and C.. B. Muneer.Florence. Wiley.S.Hemisphere. E. G. T. and Unny. Muneer. Transfer. 542. K. and Han. USA. Washington. and DeWitt. p. and Han. K. T. Aung. Computational Mechanics Publications. comfort andperformance. Proceedings Northsun '94 (Solar energy at high latitudes). and radiation heat transfer model for double glazed windows. A. (1996b) Simplified analysis for free convection in enclosures — application to an industrial problem. London. (1985)Naturalconvection. New York. Muneer. Taylorand Francis Ltd. Parsons. Kimura. UK. F. and Han. ISO 6946. T. moderate and cold environments on human health. Journal of Heat Transfer 99. Abodahab. R.

Rose. Shaw. T. P.University of Sheffield.THERMAL PROPERTIES OF WINDOWS 83 Rayment. B. (1982) Modelling the heat transfer through opaque building components for the detailed computation of thermal performance under varying climatic conditions. L. Webb. Canada. M. F. ASHRAE Transactions 101. (1992) A study of glazing heat losses and trickle ventilators. Building Research Establishment Occasional Paper. (1995) Spacers for highly insulated windows. M. 327. and Fritzel. Solar Energy 25. and Sullivan. L. London. H. J. and Concannon. Hemel Hempstead. J. P. (1992) Using Computational Fluid Dynamics. 90. PhD thesis. P. Department of Building Science.. Proceedings of the Window Innovations '95 Conference. S. 5 and 6 June. p. J. (1996) In the cool of the night. Wright. J. R. 38. Internal Report BS63. Svendsen. (1980) A simple procedure for assessing thermal comfort in passive heated buildings. 0.. (1995) A 2-D numerical model for glazing system thermal analysis. CIBSE. S. (1990) The Measurement and Computer Simulation of Heat Transfer in Glazing Systems. and Seymour. Prentice Hall International. Sharples. University of Waterloo. Wright. C. Building Services Journal. Wray. P. W. Fishwick. . Toronto.

Page blank in original .

The energy bill associated with the energy consumption due to heating of buildings is over 25% for the UK and 35% for the USA.4 WINDOWS AND SOLAR HEAT 4. In the northern hemisphere. and about 5% for cooling. The regulations make no allowance for the functions of the window. The result is that limiting window areas is recommendedwhile prescribing minimum allowable U-values. through passive solar design in buildings. Direct use of the sun's energy. Passive solar design uses a building's form and fabric to capture. is one of the most economically attractive ways of utilising renewable energy. e. passive solar design enhances the internal environment of buildings. Environmental conditioning in non-residential buildings accounts for about 15% of the UK energy consumption.1.g. its transparency to solar heat and daylight. 1984). Studies such as the one undertaken by ETSU (1985) suggest that a significantpart ofthe building heating energyrequirement could be met by solar energy technologies.Passive solar concepts offer an ideal opportunity to reduce energy bills for industrial.free from overshadowing.1 Windows as solar energy providers Within the United Kingdom. Owens (1982. store and distribute solar energy received. this form of solar energy technology is both proven and economical. CEN (1997) and Weir (1998) have presented the design requirements of any given window and have demonstrated the need for an . the European Unionor the USA. one third for lighting.is unreservedlyimportant. 4. Most building regulations discriminate against windows in favour of ordinary opaque insulating materials. commercialand domestic buildings. more energyis used to maintain comfortable internal environments in buildings than for any other single purpose. nearly two thirds of which is used for space heating.minimisingglazing on north-facing walls and incorporating complementary energy-efficiency features. Moreover. thereby reducing the demand for heat and artificial light. suchas adequate roof and wall insulation and automatic controls on heating systems.for both space heating and daylighting. it requires siting buildings so that large glazed areas can face south. Without an added capital expenditure penalty.1 Energy balance of a window Therole ofwindowsin the exploitation of passive solar energy.

A synopsis of the former two references will be presented below.2) Q 'critical = U(T1 — To)/(x) (4. to provide an expression for the critical ting the heat collection irradiation threshold ('critical). with referenceto Eq. The low levels of'critical show that the newer windowsare wellplaced for providing passive solar heating solutions. .1. Energy balance on any given window may be written in terms of its transmission characteristics. (4. and the internal and external temperatures.g. This is demonstrated by Example 4.1. for windows in passive solar heatedenclosures has been given by Duffie and Beckman (1991) as (). Table 4. provides data for a number of commercially available windows. the incident solar radiation. Numerical values of this transmittance—absorptance productfor an averaged-sizedroom with an average window size are included in Table4.4 W/m2K. 1991).1. the German manufacturer Interpane's. The critical threshold radiation is also included in this table.3) It was shown in Chapter 3 that the rapid development in superinsulated window technology has enabled the manufacture of glazing with an extremely low U-value. An expression for the glazing transmittance and room absorptance product. An important relationship which essentially represents the steady-state enis the ergy balance of any solar heat collector with an absorber area of Hottel—Whillier—Bliss model. Thus 'critical may be obtained as I — T0)] (4.This equation can be rearranged by setto nil.1.1) where TG and ID are respectivelythe transmittance of the window for global and diffuse solar irradiation. Eq.1. The equation enables calculation of useful energy gain (Q) as a function of the internal (T. (ti) = tGR/[R + (1 — cIR)TDAw/ARI (4.86 WINDOWS IN BUILDINGS objective energy balance.1.1. Muneer (1999) has presented the calculation procedure for (icc).1. ) and external (T0) ambient temperatures: R A Q = AWFR[IW(IL)— U(T FR and are respectively the heat removal factor and irradiance on the given window (Duffie and Beckman.3. 1991).2) (Duffie and Beckman.1. Aw and AR are the respective window and room surface areas. and the diffuse absorptance for the room surface.3) it may be shown that it is now feasible to collect solar energy for space heating even under overcast conditions. triple-glazing with xenon has a U-value of 0. Techniques for obtaining the above-mentioned transmittances for any multiglazed window will be presented in Section 4. e. (4. The latter may be defined as the minimum level of solar irradiance to sustain a net energy gain through any glazing system.1. based on the work of Muneer (1999). Thus.1.

Owens (1982. Note: It is possible to show that under overcast conditions the vertical slope irradiance for any window having an aspect of north-east to north-west including south is around half of the horizontal irradiance. Plots such as these are easy to produce within Excel by invoking the histogram function. there are added difficulties in establishinghow the building . Using the above assumption the threshold ofhorizontal global irradiance for any given window will be twice that of 'critical• Table 4.8 1.1.4°N).1 Design data for double-glazed windows (Muneer.WINDOWS AND SOLARHEAT Table 4. i.1 Figure 4.1.62 0.1 estimate the fraction of the time each of the given window design will provide a net contribution of solar energy towards space heating. Using the 'critical data presented in Table4. 1999) Solution for Example4.56 65 43 36 27 24 1o W/m2K 130 87 February October 45 57 f-value Float glass with air Low-emissivity Low-emissivity Low-emissivity Low-emissivity *SnO2 coating **Agcoating glasswithair* glasswithArgon* glass with Krypton** glass with Xenon** 2. Further. England (51.1 presents data for the frequency of occurrence of horizontal solar irradiation for the months of February and October for Bracknell. It is well known that solar radiation possesses temporal and spatial variability.1 includesthe essential answers to the problem under discussion.65 0.xlsworkbook may be used for obtaining the Uvalues.8 1.5 1.9 48 63 66 68 55 67 74 79 81 fractionaltime (per centofthe month) when irradianceexceedsthe criticalvalue thuscontributing towards space heating requirements Note that the Calc3-Ola. February to October.56 0.1.1. In this respect the reader is referred to Muneer (1997a).1. Since the newer windows operate with quite low levels of 'critical it is reasonable to assume that the corresponding irradiance climate will be experiencedunder an overcast sky.0 72 54 0.1. The ordinary U-value characterises only the thermal losses without taking into account the solar gains. will be in excess of thef-values provided in Table 4. Example 4. 1984) has argued for the inclusion of solar heattransmission of any given window to determine its energy balance.1.1.Itmust be bornein mind that for any given window design the actual contribution towards space heating during the nine-month duration.1 Window design U-value W/m2K 'critical 87 (ta) W/m2K 0.e.62 0.

Fd the dirt or overshadowingfactor and HDDis the degree-dayvalue for the heating season. Owen has produced an equation to obtain an effective U-value of a window. Using dynamic thermal analysis for a building of a given design. the window transmission. England its use responds to this variable energy source. The effective U-value may be obtained as and Ueff U where is the utilisation factor.6 and 0. Note that a negative value of Ueff indicates a net flow of heat inwards into the building envelope. and Fd may be assumed as having values of 0. it is a useful means of establishing comparisons between various window designs.1. W/m2 Figure 4. An extract from the above reference is provided in Table 4. provided that the following approach is not used outside the context of conventional buildings.1. f [l000fuzFdIw/(24HDD)] (4.! of the CEN (1997)document provides data for and HDDfor a number of locations world-wide. This is defined as the U-value modified by the quantity of solar heat entering the window. Table A.4) f I . typically. As pointed out by Owens (1984) all simplified approaches are susceptible to limitations but.88 WINDOWS IN BUILDINGS 100 90 80 70 60 ci) E 50 0 (a 40 30 cci 0 U- 20 10 0 0 100 200 300 400 500 600 Global irradiance.2. Note that the term degree-day is increasingly being called 'Kelvin-day'.1.1 Frequency of occurrence of solar irradiance for Bracknell.8. Owens (1984) has suggested that.

5 —0. In this respect information is available on the World Wide Web on the amountof transmitted solar radiation through multi-glazing for over 200 sites within the USA. Using Eq. Likewise. Ueff for the given sites were obtained and these are included in the given table.4).1 0. France Berlin.5 —0.3 0. Heating Degree kWh/rn2 (Jeff season days North East/west South North East/west South Sept/May Sept/May Sept/May Sept/May Nov/March Sept/May Sept/May Sept/June 2900 3465 2625 3335 1401 2700 2500 3757 202 198 230 203 110 200 146 175 350 370 410 358 239 347 306 443 505 547 590 518 442 510 511 1.WINDOWS AND SOLARHEAT Table 4. The European Union-funded European Solar Radiation Atlas provides these data for a large number of locations.1 —2. particularly for use within buildings. 1982) 89 Location Uccle.9 1.87 1 = 22. By linking slope irradiation data with commonly availabledata for heating degree-days it is possible to obtain Ueff for any given location. Other worldwide web-based sources are presently identified for the acquisition of solar radiation data.2 containssolar radiation and heating degree-daydata for a few world-widelocations.5) ERI = heating season energy rating (hundred Btu/hour-ft2) SC = shading coefficient defined as the quotient of solar transmittance and 0. under the auspices of ASHRAE a large programme of work is under way to address the availability of such data for North American sites.9 —0. —98 and —2376 (Pilkington.2 Effective U-value for selected sites (Owen. This type of data is downloadable directly . Italy London. The ERI uses the followingformula: ERI = l00{(0.1 0. In the USAan 'energy rating index' (ERI) method is used to account for the impact of solar transmission through windows.2 —0.5 Muneer (1989)has presented tables for monthly-mean slope irradiation for a number of UK locations.7 1.Denmark Trappes. England Edinburgh.3 0. (4. Table 4.5 —0.1.2 0.9 Btu/hour-ft2 (average value) Typical values ofERI for clear-single. modelling and reporting of the available solar energy. Germany Rome. A lower value of ERI indicates higher window efficiency.1 0.4 709 1. Japan Solar irradiation.9 0.1. Within the past decade there has been an explosion in the field of measurement. Scotland Sapporo.0 —0.clear-doubleandlow-ecoated doubleglazing are respectively quoted as 329.Belgium Copenhagen.1.7 —0.3 0. 1993).1 0.87SCI) where — (T1 — T0)U— (air leakage)] (4.0 1.1.3 —0.1 —0.

irradiation.entpe. New developments related to all aspects of window technology are available at: http://EETD.cie. The terms solar radiation.html Other sources of information are the interactive European Solar Radiation Atlas produced by the European Union (Palz and Grief. Irradiation availability of arbitrary sloped surfaces is a prerequisite for determining a given window's thermal balance. luminance and illuminanceare frequentlyencounteredinthe literature and anote on theiruse is perhaps appropriate at this stage. the energy exchange between the surface and its surroundings. Theinterception ofsolar radiation by arbitrarysurfacesis a function of their geometry and a determinant of their microclimatic interaction.e. .co. A number of associated Excel workbooks are also provided to ease the necessary computations.78 . Thus it is important to understand the physics of solar radiation and daylight.at/cie/home.org/staff/ceg/sunangle/index. available at: http://solstice. irradiance.org/renewables/solrad/ http://eosweb. the CIBSE Guide J for Weather and Solar Data. Other related websites are: http://www.crest. and the International Daylight Measurement Programme (IDMP) websites: http://www.39 to 0. 1996).90 WINDOWS IN BUILDINGS from the followingwebsite: http://rredc.gov/DATDOCS/Surface_Solar_Energy.LBL.Based on an algorithm developed by Dr Bernard Yallop.gov/solar/.html and http://idmp. Solar radiation (W/m2)or luminance (candela/rn2) refer to the energy emanating from the sun. a well-known astronomer at the Royal Greenwich Observatory in England. The terms irradiation (Wh/m2 or J/m2) and illumination (lumenhour/m2)refer to the cumulativeenergyincident on asurfacein agiven period of time.html. radiance. the energy balance ofthe window requires an assessment of the incident solar radiation.fr/. Luminance is the energy contained within the visible part of the solar radiation spectrum (0.html. A most user-friendly and interactive tool is the sun angle calculator. Irradiance (W/m2)and illuminance(lx) refer to the rate ofincident energy. solar radiation affects its thermal balance and thus has a significant influence on the building's internal en- vironment. i.gov/CBS/NEWSLETTER/ NL10/windows. 4.crest. the latter toolprovides an accurate routine for obtaining the sun's position at any moment in time during the period 1980--2050.nasa.tm).2 Solar radiation availability While daylight is a building's form giver.nrel.larc. The following section provides an insight into the developmentofthis science. This section has shown that during daylight hours.

DN= 1 for 1 Januaryin any given year.2. A more precise model for EOT developed by Yallop (1992) will be presented later on in this section. EOT The solarday defined above varies in length throughout the year due to: (a) the tilt of the earth's axis with respect to the plane of the ecliptic containing the respective centres of sun and the earth.2. This correction is needed in addition to the equation of Solar time time. The difference between the standard time and solar time is defined as the equation of time. and (b) the angle swept out by the earth—sun vector during any given period of time.l236sinx — 0. 1 Septemberand25 December.2.1538 sin2x + 0.1. AST is the time to be used in all solar geometry calculations. It is to apply the corrections due to the difference between the longitude necessary of the given locality (LONG) and the longitude of the standard time meridian (LSM).2. the standard time (as recorded by clocks running at a constantspeed) differs from the solar time. 4. non-leap year.2 Equation of time. DN is defined as the numberofdays elapsed since the start of the year up to a given date. which depends upon the earth's position in its orbit. In many solar energy applications one needs to calculate the day number (DN) corresponding to a given date. In any EOT is nil for 15 April.1 Solar day A solar day is defined to be the interval of time from the moment the sun crossesthe local meridian to the next time it crosses the same meridian. 4.1.6O8cos2x (4.the time required for one fullrotationof theearth is less than a solar day by about 4 minutes.1.1) where x= 360 (DN—l)/365. EOT.WINDOWS AND SOLARHEAT 91 4.2. 13 June. In a given year. EOT may be obtained as expressed by Woolf (1968): EOT = 0.242. 4.1 Solarradiationfundamentals In the followingparagraphs fundamentals of solar radiation are introduced. The object is to enable the user to perform computations related to solar radiationtransmission through any given window.OO43cosx + 0. Dueto the fact that the earth rotates in a diurnal cycle as well as moves forward in its orbit.3 Apparent solar time. . Thus. Thus.

A simple formulation for DEC is provided in Kreider and Kreith (1981): DEC = sin'{O. The algebraic sign preceding the longitudinal correction terms.013t The solar declination angle. month (m). if m>2 then y =y and m= m— 3. (4. (4. EOT is expressed in hours. day (D).460+36000.4) where UT is the universal time = h+(min/60)+(s/3600).2.915 sin G+0.466 sin 2L+0. The LSM and LONG themselves have no sign associated with them. GHA= 15UT—180—C+L— Ifnecessaryadd or subtract multiples of360 degreesto G.92 WINDOWS IN BUILDINGS AST = standardtime(localcivil time) + EOT ± [(LSM — LONG)/15] (4. hour (h). DEC=tan'(tan sin The equation of time.020 sin 2G L = 280.39795cos[O. Declination may also be defined as the angular position of the sun at noon (apparent solar time) with respect to the equatorial plane. e=23.5}/36525 (4.2.053 sin 4L The Greenwich Hour Angle.528+35999.4393—0. The present routine is valid for the period 1980—2050 and has an accuracy of 3 seconds for the EOT and 1 minute of arc for DEC. The obliquity of the ecliptic.1. 4. DEC The anglebetween earth—sun vector and the equatorial plane is called the the solar declination angle. For a given year (y).770t+C = L—2.5] + [365. The followingintermediate terms need to be obtained. Yallop's (1992) algorithm given below enables a high-precisioncomputation {(UT/24)+ D + [30.2). L and GHA to set them in the range 0 to 360°. otherwise y =y— 1 and m =m+9.4 Solar declination.05t C = 1.2.2. EOT=(L—C—)/15 ) . minute (mm) and second (s).98563(DN — l)]} (4.2.25(y — 1976)] — 8707.4).2. As an adopted convention DECis considered to be positive when the earth—sun vector lies northwards of the equatorial plane. DEC.In Eq.3) of EOT and DEC.2) In Eq. [x] denotes integer part ofx. the G = 357. contained in the squared brackets should be inserted as (+ )ve for longitudes which lie east of LSM and vice versa. In the expression above for t.6m + 0.

2.5 Solar geometry Refer to Figure 4.2. the latitude (LAT) and longitude (LONG) of the location. daily values of EOT and DEC computed for noon GMT. 4.6) cos(SOLAZM) = cos(SOLALT) Normal to sloped Figure 4. These co-ordinates which describe the sun's position are dependent on the apparent solar time (AST). These routines are. part of the various workbooks provided in the accompanying compact disc. 1997a) and the accuracy of the EOT and DEC values are far higher than those obtainable via Yallop's algorithm presented above.WINDOWS AND SOLARHEAT 93 In this text all Visual Basic for Application (VBA) routines for EOT and DEC are based on Yallop's algorithm.2.2. the elevation angle above the horizon. Where high computing speed is required this file may be used with the users' own routines or workbooks.1. and solar azimuth (SOLAZM). the azimuth from north of the sun's beam projection on the horizontal plane (clockwise = + ye). 1997a) .1. and DEC.The sun's position in the sky can be described in terms of two angles.2.5) (4. of course. The electronic data file Data4-O1.xls contains a complete four-year (leap) cycle. The solar geometry may be obtained as follows: sin(SOLALT) sinLATsin DEC+ cos LATcos DECcos SHA cosDEC(cos LAT tan DEC— sin LAT cos SHA) (4. solar altitude (SOLALT).1 Solargeometryfor a slopingsurface (Muneer. These data were provided by the Royal Greenwich Observatory for Muneer'searlier book (Muneer.

that the sun's beam strikes a sloped surface of any given tilt can then be calculated from the solar altitude. The sign convention adopted for WAZ is the same as that used for SOLAZM. Corrections have therefore to be made for the above refraction and altitude effects.1 for further details. (4.8) H in the above equation is the station height in metres above sea level.2. Eq. the sun will appear in the morning slightly earlier. These may be used as illustrations for the computational routines discussed above.8) for SOLALT (degree) refers to the instance of actual sunrise or Astronomers define sunrise and sunset as sunset: SOLALT = —0. i. air. Further. azimuth and the orientation of the surface as expressed by its wall azimuth angle (WAZ). The angle of incidence. This is due to the refraction of light by the terrestrial atmosphere.2. Using Eqs (4.1.INC.2.1.for locations which are higher than the sea level. Muneer (1997a) has presented an exhaustive account of models and electronic tools.1 shows the angles relevant to the determination of sun's position and the geometry for a tilted surface. Using the latter referencea number of Excel workbooks have been developed.0347I1° (4. Examples extracted from the above reference are given below.e.2.94 WINDOWS IN BUILDINGS The solarhour angle.2)(in that order). the sunrise/sunset instance may be computed by setting SOLALT =0. on a compact disc. See Section 4.do not occur at the time when the sun's elevation is zero.2.5) and (4. for preciselyobtaining solar geometry.6 Sunrise and sunset the moment at which the centre of the solar disk is along thehorizon ofthe earth. INC = cos1[cos SOLALTc0s(SOLAZM — WAZ) sinTLT + cos TLTsin SOLALT] (427) Figure 4. 4. clockwisefrom North is considered positive. SHA. . The actual sunrise and sunset. These workbooks enable solar radiation computations to be performed with much more ease within the spreadsheet environment. however.2. Hence actual sunrise appears slightly before astronomical sunrise and actual sunset occursafterastronomical sunset. A ray oflighttravelling in vacuum from the sun which is actually belowthe earth'shorizonis bent towards the earth by the heavier medium. is the angular displacement ofthe sun from the local meridian at 15° per hour.8333 — 0. An interactive sun position calculator based on Yallop's algorithm is available via the World Wide Web.

LONG=3.2.12 h (7 minutes and 10 seconds) DEC = 0. the sunrise and sunset events were respectively to occur at 0540 and 1820 hours AST. Note that global refers to total amount of irradiation and diffuse to the irradiation received from the sky and that reflected from the ground.40° (0° 21') AST = 11. solar altitude and sun's azimuth for Edinburgh.3° Note that a supplementary workbook. With possible to perform energy-system simulations using hourly or sub-hourly data.WINDOWS AND SOLARHEAT 95 Example 4. This workbook provides mid-hour (based on solar time) sun geometry for the 24-hour cycle. Calc4-04b.2 Hourlyhorizontal irradiation The frequency at which solar radiation data are required depends on the apthe advent of cheap yet powerful desktop computers it is now plication. AST. if any. WAZ. Find the angle of incidence ofthe sun's beam on a surface with a given tilt of45° and orientation 150 west of south.95°N. is 195° ( 180+ 15). For the hours neighbouring these events the geometry is computed in the followingmanner. is also available on the compact disc. Calc4-02 (for AST). The choice of AST in Calc4-04bis justified on the grounds that most meteorological office data are available in solar time. UK [LAT=55.1 Calculate the EOT. For example. taking into account the sunrise and sunset events. computed from the corresponding horizontal global and diffuse energy data. if for a given date and location. 4.67 h EOT Sunrise time = 0611 hours GMT Sunset time = 1828 hours GMT SOLALT = 3430 SOLAZM = 174° INC = 19. find the time for sunrise and sunset for the given date (assume H=0). Note that the required input value of the wall azimuth angle. DEC. and likewise. Calc4-03 (for sunrise andsunset times) and Calc4-04a (for solar geometry) the followingare obtained: = —0. For the sunrise event the geometry is computed at midpoint between the sunrise and the integer hour immediatelyfollowing sunrise.2. Under an overcast sky the diffuse and global irradiation become equal. . Such simulations. at mid-point between the integer hour preceding the sunset and sunset event itself. Also. however. require reliable estimates of slope irradiation and illuminance. the geometry would respectively be computed at 0550 and 1810 hours AST.20°W and LSM=0] at 12 noon LCT on 21 March 1997. Using Excel workbooks Calc4-01 (for EOT andDEC).

34 1.217°N.97 3.65 0.37 0.67 2.86 4.83 0.72 1.51 2.02 1.23 4.54 1.31 3.61 0. l.37 0.15°N.89 1.51 2.03 0.22 Cambridge52.19 1.16 2.99 2.57 0.81 2.28 4.21 0.2.61 2.75 3.6 0.63 2.18 0.417°N.25 Newcastle54.32 2. 1.34 0.31 1.35 Global 0.74 1.77 Diffuse 0.75 3.45 0.26 4.48 0.96 2.23 3.54 0.18 0.33 4.37 0.46 4.16 Manchester 53.86 1.71 1.48 4.79 0.4 4.93 0.49 0. 0.96 4.1 Edinburgh 55.43 0.06 2.78 1.17 Leicester52.18 1.63 3.67 1.38 2.28 1.27 0. 3.26 1.44 0.44 0.53 2.32 0.85 2.45 0.233°W Global 0.43 0.18 Diffuse 0.1°E 1.82 4.0°W Global 0.86 1.42 2.37 0.28 0.76 5 6 7 8 9 10 11 12 Lerwick60.06 2.35 1.l°W Global 0.33 0.35 0.73 3.56 2.52 1.3 4.93 3.42 0. 3.76 1.87 Diffuse 0.45 1.43 0.09 Dundee 56.55 1.34 2.22 0.42 1.37 3.49 2.83 Diffuse 0. 5.2.27 1.29 3.01 2.39 0.23 3.59 2.26 0.62 0.66 4.42 Global Diffuse 0.29 4.7 1.983°N.39 0.6 1. 3.37 3.75 1.33 0.26 0.0 1.95 1.44 1.39 0.15 1.38 0.36 0.07 1.51 3.38 4.51 4.1 Monthly-averaged horizontaldaily globaland diffuse irradiation (kWh/rn2)— UK sites(Muneer.67 2.3 0.94 2.9 Diffuse 0.94 4.26 2.133°N. l.99 3.35 4.41 0.55 0.69 0.15 Aldergrove54.567°W Global 0.06 2.29 0.06 1.19 1.217°W Global 0.46 0.42 2.200°W Global 0.5 4.32 4.7 0.29 0.32 2.48 3.51 1. 4.5 1.4 0.0 1.23 1.31 Belfast 54.82 Diffuse 0.133°N.88 2.42 0.94 3.32 2.99 2.53 0.36 0.71 2.62 1.31 1.06 2.65 2.383°N.46 3.28 1.61 1.5 2.12 0.06 1.68 1.54 2.29 4.67 0.07 1.72 4.37 2.72 4.35 2.48 1.11 3.82 Diffuse 0.49 1.28 1.31 2.83 3.25 1.59 .22 2.01 0.38 0.88 2.45 0.37 4.183°W 0.09 2.19 1.22 Birmingham52.36 0.58 Diffuse 0.97 Diffuse 0. 2.43 1.14 2. l.39 1.65 1.28 2.15 2.68 2.22 0.59 2.3 0.26 4.16 5.583°N.24 2.48 1.36 2. 4.34 3.93 2.98 3.73 1.08 2.52 1.66 1.44 2.87 1.38 Aberporth 52.29 0. 6.47 3.1 Eskdalemuir 55.16 2.89 2.92 5.35 1.35 0.81 2. 1.633°N.3 4.35 2.96 2.39 3.08 3.76 4.27 3.8 1.21 Liverpool 53.583°W Global 0.86 3.21 Aberdeen 57.5°N.01 Diffuse 0.45 2.59 3.96 WINDOWS IN BUILDINGS Table 4.51 2.34 0.63 2.38 1.28 0.0 1.06 1.81 1.49 5.45 4.05 3.917°W Global 0.29 0.28 0.71 2.79 2.27 4.76 2.55 2.59 2.63 1.69 1.23 0.19 4.65 0.29 1.5°N.12 2.9 2.92 Diffuse 0.083°W Global 0.65 0.15 4.99 3.76 1.14 0.51 1.45 0.63 1.03 1.26 1.82 1.33 0.41 0.65 1.57 1.36 0.833°W Global 0.81 4.92 3.69 1.16 3.34 0.25 0.38 1.34 0.34 3.07 1.86 1.47 1.37 0.22 2.25 4.3 0.83 1.85 1.08 2.25°W 3.47 0.34 4.67 1.45 0.32 2. 1999) Month Global Diffuse 1 2 3 4 3.75 2.5°W Global 0.26 3.08 Diffuse 0.61 Diffuse 0.65 1.32 0.27 0.41 1.2 1.47 0.83 4.36 2.12 2.483°N.28 2.39 0.11 2.76 3.317°N.35 0.37 0.69 1. 2.12 2.47 0.28 0.1 2.47 1.21 1.89 1.91 2.11 Glasgow 55.73 0.4 0.58 0.0 3.3 4.07 1.0 Diffuse 0.9l7°W Global 0.76 5.41 0.51 4.36 3.45 3.867°N.32 0.89 1.1 2.03 0.36 1.2°W Global 0.56 3.21 Sheffield 53.0 3.51 1.29 0.78 2.95°N.8 2.650°N.26 1.

2.65 Diffuse 0.07 4.12 1.8 1.6 2.l83°N.167°W Global 0.78 0.52 2.76 1.75 1.36 1.05 9 10 11 12 London 51.53 2.217°W Global 0.68 2.95 1.4 4.5 2.38 3.64 0. 0.WINDOWS AND SOLARHEAT 97 Table 4.27 Diffuse 0.89 Cardiff 5l.32 3.64 5.52 5.83 1.84 1.55 5.44 0.62 1.69 1.7 1.22 1. 4. 2.99 Diffuse 0.9 2.0 2.22 2.517°N.61 5.91 0. Therefore.17 2.12 Diffuse 0.73 1. the electronic tools presented herein will enable: (a) the decomposition of daily into hourly values.3 1.05 3. pressure. Table 4.7 1.53 2.24 2.07 4. There is extensiveactivity to produce an atlas of the solar energy resource.32 The UK radiation measurement network is one of the best in Europe.92 1.483°N.65 1.64 0.52 5.81 0.32 3.43 5.75 1. 4.37 0.28 1.383°N.1 0. 0.41 2. 2.1 (Continued) Month 1 2 3 4 5 6 7 8 3.33 0. Examplesin this respect are the work of Mani (1981)for India.42 4.76 5. In this section models are presented which enable computation of horizontal irradiance and illuminance.183°W Global 0.4 0.48 0.52 0.51 1. methods are required for estimations to be carried out from long-term records of daily irradiation or other meteorologicalparameters such as humidity.98 0.69 3.96 2.44 3.32 Diffuse 0.43 0.58 2.36 0.93 0.9 Jersey 49.28 2.79 1.4 2.36 2.2.04 2.68 2.64 0.26 4.06 3.88 Bristol 51.59 0. Palz and Grief(1996)for Europeand Duffie and Beckman (1991) for the USA.017°W Global 0. yet long-term hourly data for the latter quantities are available for only a limited numberofsites.49 0.14 2.56 1.62 1.15 4.33 4.32 4.03 2.53 5.01 Plymouth 50.89 2.94 1.42 4. and (b) the horizontal.43°N. It will be shown in this section that provided monthly-averaged horizontal dailyirradiation data for the global anddiffuse components are available.58 0.02 1.95 1.84 2.01 2.53 0.59 0.35 1.55 0.2.45 0.3 2.75 0.42 0.56 0.46 2.53 Diffuse 0.48 3.17 3.56 2.1 Hourly horizontal globalirradiation It will be shown in the following section that Liu and Jordan (1960) models may be used to obtain hourly irradiation data from long-term records of .86 4.87 Easthampstead 51.58 0.01 1.64 4.58°W Global 0.1 2.58 0.383°N.7 2.1 4.33 5.2.55 2.8 1.53 0. 3.32 2.5 2.53 0.57 0.783°W Global 0.1 presents monthly averaged horizontal daily irradiation data for a number ofUK locations.46 2.59 2. sunshine and cloud-cover.55 2.68 4.92 1. hourly irradiation can then be used to obtain slope irradiation.21 2.0 2.82 0.

based on variables routinely recorded by meteorological networks. 0.5°W)/Heathrow (51. 0. When evaluated.O must be monthly-averaged daily values. the irradiance under a cloudless sky 'G. Kasten and Czeplak (1979) have shown that for any given cloud amount. A new approachdeveloped by Page (1997). with a slightly poorer performance compared with the SRM and CRM under intermediate and clear sky conditions. nitrogen and carbon dioxide). Research teams at Napier University in Scotland. Models based on sunshine are considered to be more reliable as the input variable is properly registered by a sunshine recorder in contrast to a single spot reading of the cloud amount based on a visual estimate. Aughton (53. for the generation of world-wide data tables.3°N. 0. reliable computational methods based on other meteorological data may be used.98 WINDOWS IN BUILDINGS detailed hour-by-hour data.2 The cloud radiation model (CRM) Cloud data are extensivelyused for estimating global and diffuse irradiance. In order to determineglobal irradiance 'G under a cloud amount of N. water vapour. the ratio ofglobal irradiance to global irradiance under a corresponding cloudless sky is independent ofsolar elevation.4°W) will be discussed later in this chapter. The SRM. 2. The amount of cloud is a measure of the proportionof the sky-coverwhile the durationof bright sunshine determines the frequency of cloud-cover.5°W)/Ringway (53. Bracknell (51.3°W). called the 'PRM' for (Page radiation model) is based on a combination of sunshine and cloudcover data to calculate solar radiation.2°W). these models have been found to confirm the findings of Bennett (1969).0°N. Sheffield University in England and Waterloo University in Canada have been engaged in separate projects to develop solar radiation models. ozone and aerosols. 6. Data4-02. Readers will be able to perform their own computations using a number of relevant workbooks and sample data files provided on the accompanying compact disc.3°N. however. Building simulation programs. need .2.2°N.l°W)and Stornoway (58. Page's model has been found to perform effectively under overcast conditions.3°N.2°N. The so-called meteorological radiation model (MRM) is based on sunshine fraction andtemperature data. 4. In the absence of measurements.xls provides simultaneous meteorological and irradiation (horizontal and slope) data and these may be used for undertaking independent evaluations of the above procedures. Finningley (53. 1 . Such models estimate the beam transmission through the terrestrial atmosphere and its attenuation due to the presence of clouds and mixed gases (such as oxygen.0°W).2. 2. the CRM and Page's combined approach using data from UK sites: London (51. In the present text the MRM has been renamed SRM (sunshine radiation model). The cloud-cover radiation model (CRM) uses information regarding the amount of sky covered with clouds. According to Bennett (1969) 70—85% of solar radiation variance is explained by sunshine durationin contrast to the figure of 50% quoted against cloud amount.6°N.

3 A B 30 45 C 0.2. that part of the electromagneticspectrum whichstarts with ultraviolet (UV) radiation and ends in the near-infrared (NIR) region.3 + 0.73 0.2.9a) thus: 'G.2. The summation of all energy received at individual wavelengthsin the solar spectrum equals the solar constant. 1998) Latitude. The Calc4-05.4 .3 Solar radiation transmission through a terrestrial atmosphere Most broadbandsolar radiation sensors work in the 300 nm to 3 tm band since this region covers98% ofthe energyradiatedby the sun.2 54.75 0.O depends on solar altitude.0 51.2 Coefficients for CRM (Eq.WINDOWS AND SOLARHEAT 99 known first.1 910 979 956 902 948 58.9c) The values ofthe A.2.2.e. (4.2.71 0. i.71 D 3.2.9b) The diffuse component is calculated as follows: 1D/1G = 0.7 3.O = (A sin(SOLALT)— B) N is usually expressedin oktas ofthe sky obscured by clouds where an okta is a one-eighthportionof the hemisphere. SOLALT. This is the rate of the energy received on a surface normal to the sun's rays at the top of the earth's atmosphere and at a sun—earth distance equal to 1 AU (1 Astronomical Unit.4 3.2. N Hamburg Stornoway Aldergrove Finningley London 58.o has been shown to be independentofthe solar altitude: IG/1G. AU= 1.70 0.2. and may be obtained (4.7(N/8)2 (4.4 3. However. C and D coefficientsare provided in Table 4. B. even with the coefficientsoriginally presented by Kasten and Czeplak (1979) the CRM performs quite satisfactorily.9a—c)refers) (Gui et al.496 x 10" m). 'Sc (= 1367 W/m2).6 53. An improvement in performance results from the use of locally fitted coefficients.. 4.The ratio ofglobal irradiance 'G for a given cloud amountN to 'G.5 34 36 49 3.xlsworkbook provides an easy means of generating solar irradiance data based on the CRM procedure. The distribution ofthe solar spectral irradiance is notuniform in the range of 250 nm to 25 tm. 'G.O = 1 — C(N/8)D (4. The energy of the solar Table 4. occurring at the vernal and autumnal equinoxes.

2. the attenuation of light through a medium is proportional to the distance traversed in the medium and the local flux of radiation.2.10) k and m are respectively known as the total attenuation coefficient and air mass. 02.13) .0005 mbar. Eq.12) Sunshine radiation model estimates the horizontal beam and diffuse components from only ground-based meteorologicaldata air temperature.10) becomes t t. beam and global irradiance.285k'°°648 (4. Thus.10)may be written as 'B = 'E where 'B and 'E are respectively the beam and extraterrestrial irradiance on a horizontal surface. part ofthe incident energyis scattered in all directions.ozone and water vapourtransmittances. play a significant role in the absorption of solar radiation.2. Thus. (4. However. Definingthe transmission coefficientas r=exp(km). Therefore if Tr and are the respective transmittances for the Rayleighand Mie scatteringandrg.the phenomenon iscalled Rayleighscattering. visible-band(46%) and NIR (46%).100 WINDOWS IN BUILDINGS spectrumis approximately distributed as follows: UV (8%). The SRM is therefore a useful tool and can provide reliable estimates of horizontal diffuse. 'E may be computed from = 1367[1 +0. H20 and aerosols. while that part which arrives at the earth's surface directly from the sun is called direct or beam radiation. Ifthe particle size is smaller than the wavelength.4 The sunshine radiationmodel (SRM) (4. According to the Bouguer—Lambert law. 03. Eq. (4. DBR = 0. 'B = IEtrtgtotw 4.2. When solar radiation enters the earth'satmospherea part of the incident energy is lost by themechanisms ofscattering and absorption. The SRM for non-overcast conditions consists of regression between the hourly diffuse to beam ratio (DBR = ID/IB) and the beam clearness index (kB IB/'E).11) When an electromagneticwave strikes a particle. and the mixed-gases. The naturally occurring gases.O33cos(0. The atmosphere extends to an altitude of 100 km or a pressure of 0. relative humidity (or wet-bulb temperature) and sunshine duration.0172024DN)]sin(SOLALT) (4.2. N2 and CO2(mixed gases).2.2. 'B = IEexp(—km) (4. The scattered radiation is called diffuse radiation.2. Usually air molecules cause Rayleigh and aerosols Mie scattering. Such data are readily available world-wide.iftheparticle is ofthe order of thewavelengththeprocess is known as Mie scattering.

2. The procedure is quite involved. but it nevertheless produces good estimates of irradiation once the hourly sunshine and cloud cover are provided. (1998) have provided a detailed account of the development of SRM.2. t t Li + rn'°2+ 1 — m + m102j tr) 0.6 Page clear sky radiationmodel This clear sky irradiation modelwas originally developed by Page as a key component needed in the development of the European Solar Radiation Atlas . Global irradiance for a non-overcast sky is then obtained as the sum of diffuse andbeamirradiance. however. In particular.2.14) and (4. diffuse and global irradiance. The workbooks were built to generate all-sky and clear-sky irradiance. 4. ifthey so wish. Gul et al. the SRM respectively computes 'B and 'D using Eqs (4.2. In the case of overcast skies the direct radiation reaching the earth is nil. The model enables hour-by-hour computation of beam. A condensed version of the procedure is presented here.2. attenuatedwith the hourly sunshine fraction (SF) for non-overcast conditions. Further details are available in Muneer (1997a).5 Page radiationmodel(PRM) Theradiationmodel developedby John Page of Sheffield University was based on the work undertaken for the developmentof the EuropeanSolar Radiation Atlas. The Calc4-06. the durationofbright sunshine within the hour. Its accuracy is most precise for clear sky conditions and worst during overcast periods.84(1 —t)) (4. is defined as 18 = (SF)Inzrttgtozw (4. remain 'transparent'to the ratherelaborate procedure by the use ofworkbooks presently available on the accompanying compact disc. the PRM procedure returns more precise values of overcast irradiation which will be of use in daylight design. Readers may.WINDOWS AND SOLARHEAT 101 The beam irradiance. The cloudless sky model estimatesthe global irradiation from the sum of theestimated beam and diffuse irradiation fallingon a horizontal surface.2.xlsworkbook provides an easy means of generating solar irradiance data based on the SRM procedure. hence there is no beam component. In this case the diffuse irradiance is modelled as = ID = IE'ttgotw 1 0.13). i.15) 4.2.14) For a given valueof SF.5(1 —m where and are functions of m and ta..2.e. The overcast model determines the hourly values of overcast day irradiances from thedailymodels as a function of solar altitude using a sine-based second-order polynomial.

a second beam irradiance calculation is performed. The cloudless sky diffuse irradiance is then obtained as the product of Kd.102 WINDOWS IN BUILDINGS (Page. In this respect.8662mTLK5r(m)) (4. is obtained by m = 1/{sinSOLALT + O. First the value of diffuse transmittance is established. is then calculated as the sum of beam and diffuse components.8 Page overcast sky radiation model The key initial input into the daily overcast sky model is the daily transmittance for the overcast conditions. This is the theoretical diffuse irradiance on a horizontal surface with the sun vertically overhead for the air mass 2 Linke turbidity factor.2. m. TLK. The model provides control values of the maximum acceptable values of hourly global irradiance using the cloudless sky model described above. 4.the reader is referred to the work of Page (1997). 4. 1997).2.17) the Rayleigh optical depth at air mass m and is calculated using the work of Louche et al.2. afterincorporating the standard corrections to the mean solar distance.2. This is the ratio of the overcast day daily .2. and air mass adjustment for station height. 'Gc. selected. Further details may be obtained from Kinghorn and Muneer (1998).2. Note that a definition of some of the parameters presently under discussionis beyond the scope ofthis book. (1986). The diffuse modelling is carried out in two stages. Total global irradiation underclear sky conditions. Global irradiation is estimated from the sum of estimated beam and diffuse irradiation falling on a horizontal surface. In this model the diffuseirradiation ofthe cloudlesssky depends on the solar altitude and air mass 2 Linke turbidity factor. As the sky becomes more turbid the diffuse irradiance increases while the beam irradiance decreases. The air mass. The Linke turbidity factoris set to its lowest possible value. The cloudlesssky model predicts the irradiance on the horizontal surfaces as a function of solar altitude and air mass 2 Linke turbidity factor.16) Kd is the correction factor to allow for the varying solar distance and TLK is the air mass 2 Linke turbidity factor.7 Page extreme clarity clearsky model In order to establish credible upper limits over the wide range of air mass 2 Linke turbidity factors that may occur in practice.07995)106364} t5r(m) is (4.50572(SOLALT + 6. Thesolar beam irradiation normal to the beam depends on thesolar altitude and the Linke turbidity factoris obtained first: 'Bn = I 367Kdexp(—O. the diffuse transmittance and a solar altitude dependent function.

2 Table 4.18) 'Do is the overcast day irradiation and Trd the corresponding transmittance for a solar elevation of90° and A(0). Using the CRM. 4. SRM and PRM models. This associated hourly overcast radiation model is based on the estimation of the overcast sky transmittance with the sun directly overhead combined with the application of an overcast sky solar elevation function to estimate the overcast day global irradiance value at any solar elevation. The formula used to calculate the overcast day irradiation is of the form 'Do = KdTrd[A(0) + A(l) sinSOLALT + A(2) sin2 SOLALT] (4.WINDOWS AND SOLARHEAT 103 irradiation. on a horizontal plane to the extraterrestrial daily irradiation.4.2. the PRM produces better irradiation estimates under overcast conditions as has been shown by Kinghorn and Muneer (1998).3. Note that Calc4-07a. provides a more detailed evaluation of the above three procedures for obtaining irradiation on a horizontal plane.xls workbooks may be used to generate the output shown in Table 4. It may be seen that for the non-overcast conditions.19) The DLL-based workbooks Calc4-07a. Ifthe hourly transmittance were constant.30 for all months. Example 4.xlsand Calc4-07b respectively provide easy means of generating solar irradiance data for all-sky and clear-sky conditions based on the PRM procedure.2. the daily transmittance for overcast conditions in theUKis estimated using the work ofCowley(1978). On the otherhand.xls and Calc4-07a. Global irradiance is then obtained as = (IGc — ID0)SF+ 'Do (4. a second-order sine polynomial is used to calculate the overcast day solar elevation function. Calc4-05.2. For hours with cloud = 7 oktas. the CRM and SRM provide global irradiation estimates which are betterthan the corresponding PRM values.2.2.xlsis a DLLbased workbook.2. the daily transmittance for overcast conditions is set as being equal to 0.3 provides measured hourly weather data for Bracknell. Calc4-06. the correction would be equal to the sine ofthe solar altitude.xls. A(1) andA(2)are constants dependent on sky type. For hours with cloud =8 oktas.9 Conversion from daily global irradiation to hourlydiffuse irradiance The daily globalradiationinformationhas next to be linked with the overcast sky diffuse irradiance model.2. As this is not so. based on the material presented in this • reference.2. Table 4. west London. . The text box provided below should be referred to for the execution of all such files. obtain the respective hourly horizontal global and diffuse irradiation values.

England(August 1990)* of Hour Date 19 ending 5 6 7 8 9 10 11 12 13 Measured 0 12 Solar irradiation.siy press ALT+F8 keys • Select the relevant macro from the dialog box • Either click the Run button.W/m2 CRM SRM 6 31 PRM 6 31 124 6 31 19 19 19 19 29 44 55 107 167 198 50 85 116 142 160 169 169 45 76 106 132 151 161 161 225 321 19 19 19 19 19 19 19 19 19 404 463 491 294 340 253 105 65 14 15 16 17 18 334 296 243 177 489 375 399 319 223 123 443 282 227 162 66 17 1 1 19 19 20 20 20 20 20 20 20 20 20 20 20 20 19 20 5 6 7 8 9 10 11 104 30 6 6 45 161 29 103 30 6 6 29 163 313 96 30 6 6 29 161 329 441 548 176 242 295 456 577 661 616 591 456 571 20 20 20 20 12 13 14 15 16 17 18 19 428 643 523 303 253 65 37 1 483 456 404 390 175 102 705 704 658 538 381 335 491 625 721 770 770 720 589 20 28 6 286 113 28 6 411 299 110 28 6 *Data4O2xls refers For all DLL-based workbooks the followingprocedure MUST be used: • Launch Excel software • Open the DLL-based workbook • Enterdata in the relevant worksheet(s) • Simultanec.2.104 WINDOWS IN BUILDINGS Table 4.3 Comparison radiation models for Bracknell. or simultaneouslypress ALT + R keys .

3 58.6 51.4 53.3 58.3 In many instances it is important to obtain estimatesofclear-sky irradiation so that the effect of large glazed facades on the cooling load of a buildingmay be investigated.2 Bracknell 51.0 53.2 —56 —86 —55 —101 —40 —22 —75 —77 Aughton Stornoway 88 93 96 95 81 95 129 91 142 178 213 —95 122 108 149 K MBE* RMSE** index Mean Bias Error Root Mean Square Error Hourly clearness 4. The PCSRM has been validated by Page (1997) using clear-sky data from Indian locations and shows consistency with an optimum value of 3.WINDOWS AND SOLARHEAT Table 4.5 51.2.2 51.5 51.4 53.5 for the Linke turbidity .2.10 Page clear-sky radiation model (PCSRM) It will be demonstrated here that the Page clear-sky radiationmodel provides an accurate means of obtaining global irradiation for worldwide locations.2.4 Comparisonof hourly globalirradiation models Latitude Location London 105 N 51.5 CRM MBE* (W/m2) MRM PRM 6 —1 RMSE** (W/m 2) MRM PRM CRM 54 63 57 83 57 69 81 71 (a) Overcastskies.5 for the air mass 2 Linke turbidity was found to produce reasonably good estimates for world-wide locations. The only parameter required is the turbidity factor. The PCSRM originally developed as part of the new European Solar RadiationAtlas is a powerful and useful facility for suchan exercise.4 53. Example 4.2 4 —11 98 122 153 K > 0.2 <K<0.3 33 38 33 53 53 51 55 51 Finningley Aughton Stornoway London 2 5 —2 55 37 2 —6 —1 65 67 63 74 34 43 47 131 64 66 80 68 92 82 38 71 49 122 66 64 99 113 130 (b) Intermediate skies. K < 0.0 53. The PCSRM was also used in the production of clear-day tables for world-wide locations within the CIBSE Guide J for Weather and Solar data.2. As a matter of fact this procedure was adopted in theproduction of CIBSE Guide J.6 18 21 Bracknell 5 20 22 Finningley Aughton Stornoway (c) Clear skies. 0. An average value of 3. London Bracknell Finningley 0 —3 —2 1 17 13 87 72 58.0 53. The following example demonstrates the use of the PCSRM.

1 19.5 19.5°E Measured data* 'G 'D Hour 1 Computed data 'G W/m2 0 0 0 0 0 0 20 192 A0 DBT 24.2 23.1 16.9 17. W/m2 .2 16.7 34.0 22.9 19.0 16.5 33.9 32. SRM and PRM are good tools when missing records of measured irradiance are to be filled-in to create complete (and unbroken) time series.6 16.9 24. The CRM.8 34.8 16.5 23.2 16. Longitude Date: 26 October 1971 = 73°E.6 shows the performance of the PCSRM on a monthly-averaged basis.2.4 30.7 17.3 28.xlsworkbook to evaluate the PCSRM against clearsky data for Jodhpur(India) data provided in Table 4.1 25.2.3 19.4 22.xls is a DLL-based workbook. Table 4.6 17. Use the Calc4-07b.0 33.2 18. Table 4.7 W/m2 0 0 0 0 0 0 38 243 465 W/m2 0 0 0 0 0 0 10 51 W/m2 0 0 0 0 0 0 20 68 101 120 129 132 132 129 120 101 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 21.2 25.4 16.0 34.1 28.8 WBT 17.8 26.4 17. India Latitude = 26.8 27.5.4 17. W/m2 Hourly diffuseirradiance.5 Evaluation of Page clear sky radiation model(PCSRM) for Jodhpur.9 20.0 19.2 16.5. °C Hourly global irradiance.2.4 71 80 81 88 87 88 78 66 45 6 415 613 756 830 830 755 612 414 191 20 17.7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 68 20 0 0 0 0 0 0 Mani (1981). Calc4-07b. Such time-series are required for serious simulation work on building energy performance.2 31. The text box provided above should be referred to for the execution of all such files.6 26.106 WINDOWS IN BUILDINGS factor.4 21.9 18.3°N. Dry-bulb temperature. °C Wet-bulb temperature. The computed values of 'G and 'D are included in Table 4.8 640 752 826 820 740 600 405 185 13 20 21 22 23 24 DBT WBT ID 34.2. LSM = 82.6 18.

6 Evaluation of PCSRM for Jodhpur. If. (1984) have respectivelypresented models forAustralian and Indian locations. however. models for specific locations are available they may indeed be used instead for achieving higher accuracy: .2. 4. (4. (1982) followed their procedure to develop a regression model for the USA.2.WINDOWS AND SOLARHEAT 107 Table 4.2. fitted for the mean global data and this maybe usedto estimate the horizontal diffuse irradiance in the absence of a regression model for any given location. India (data averaged over a 10-year period) Latitude = 26. Longitude = 73°E. Muneer and Saluja (1986) have provided Eq. LSM = 82.2.11 Hourlyhorizontal diffuse irradiation Followingthe approachof Liu andJordan (1960). Erbs et al. Spencer (1982) and Muneer et al.5°E Month February Hour 10 11 12 13 SF 1 1 1 1 1 1 1 1 1 1 1 Measured* Computed % error Jo 607 740 802 803 Measured Computed % error 1o I April 14 9 10 11 12 13 14 May 9 10 11 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 729 577 754 883 944 942 865 603 771 604 747 822 822 748 625 830 976 1051 1051 —1 1 2 2 3 8 10 11 11 143 159 166 168 164 194 227 243 120 129 132 102 96 121 254 256 260 231 132 135 135 102 —16 —19 —20 —40 —42 —37 —42 —44 —47 —60 —63 —46 —48 —51 —51 —52 —51 —50 —1 —5 12 13 976 685 882 1022 1094 1094 1022 882 807 882 882 343 532 670 742 742 669 531 96 125 134 135 135 135 135 12 13 14 15 892 953 951 884 751 758 824 823 October November 11 12 13 14 14 15 15 15 16 17 6 7 7 —5 —2 1 259 274 278 281 276 269 132 141 147 134 131 133 102 88 —31 3 9 10 11 12 13 363 541 85 100 108 115 14 15 663 727 723 654 526 2 3 2 1 116 116 102 110 122 126 126 121 110 10 13 10 9 5 8 *Mani (1981). Orgill and Hollands (1977) have presented a case to correlate hourly diffuse ratio ('DuG) and hourly clearness index (Kr).20).3°N.

2. 4. 4.108 WINDOWS IN BUILDINGS ID/IG 1. on the ratio of hourly to daily global irradiation (rG): It / . For this purpose one needs to obtain hourly data from records ofmonthly-averageddaily data sets. Example 4.2.l24lK — l2.2.20) to obtain thecorresponding diffuse irradiance estimates. The original work in this field is attributed to Whillier (1956).7166K (4.2.xlsmay be used to obtain the above estimates for 'D The estimated and measured diffuseradiationvalues arerespectivelygiven in columnsJ and L of the workbook.12 Monthly-averaged hourlyhorizontal global irradiation averagedhourlyirradiation and temperature data. cosw—cosw rG =—(a +b cosw) 24 cosw—wcosw Often window designerneeds to perform abbreviated analysis usingmonthly- a (4.6609 — 0.13 Monthlyaveraged hourly horizontal diffuse irradiation Long-term averages of hourly diffuse irradiation can be computed from monthly-average values of daily diffuse irradiation if the ratio of hourly to daily diffuse irradiation. rD.xls contained in the compact disc.047) and b' = 0.5016 sin(w — 1. It is evident that using the given regression provides reasonable estimates of the desired quantity.2. (4. Calc4-08.4 Hourly global irradiance values are provided for Bracknell in the electronic workbook Data4-02.37lK + 3.047). Liu and Jordan (1960) and Collares-Pereira and Rabl (1979) extended that work to develop a relationship which takes into account the effect of the displacement ofthe hour from solar noon (w) and the day-length (w). Evaluate these estimates against the measured values reported in the above workbook.2. Use Eq.22) . is known.2.006 — 0.20) Muneer (1997a) has provided an extensive survey of the relevant diffuse radiation models for locations world-wide.4767 sin(w — 1. Liu and Jordan (1960) have presented such a model: cosw rj = 24cos w — w cosw — (4.21) a' = 0.409 + 0.2.76l6K + 9. The following section presents models for these tasks. which are readily available in the meteorological archives.2.

2.2°W) (Muneer. 1 1 J March April 1o 'G 0 0 3 3 19 17 0 0 14 11 8 8 61 42 153 56 197 117 67 47 164 100 36 28 117 72 208 128 8 8 58 42 147 0 0 14 11 381 281 1119 81 286 172 258 156 361 211 75 53 53 178 108 231 142 92 61 147 May June 'D Iç 1 jc 'D 94 186 114 0 97 269 156 297 178 361 281 156 Computed values Jan. March April May June '0 'G 0 0 0 11 10 0 0 12 10 73 0 0 66 47 35 29 12! 77 211 133 319 187 68 50 167 87 60 192 110 294 170 406 87 60 192 110 170 68 50 167 35 29 121 1 'D 88 'G 'D Iç 66 142 218 101 191 144 99 246 112 153 57 165 99 265 157 99 157 376 211 77 211 133 294 265 406 0 0 66 47 144 99 246 0 0 12 10 0 0 0 0 11 73 57 165 10 88 380 278 1116 686 1996 1252 3200 376 295 184 348 363 215 416 231 211 415 238 467 223 442 250 494 263 223 442 250 494 263 415 238 467 252 1 123 144 270 164 319 187 363 215 416 231 153 295 184 112 66 142 101 191 123 218 144 270 164 1904 3750 2264 4372 2468 348 201 201 252 . 7. 3.2.xlsenables such computations. Table 4. Using Eqs (4. Scotland.5 9.5 Wh/m2 0 0 0 0 19 17 'G 'D Feb. 1997a) Month Time.5 11.5 13.21)and (4. h 6.5 15.2.5 Totals 17.5 Table 4.22)or Calc4-09.2.1 mayeasilybe decomposedinto hourly values.5 12.1. The text box provided below Example 4.xlsthe monthly-averaged daily irradiationdata presented in Table 4.2 should be referred to for the execution of all such files.2.7 Monthly averaged hourly horizontal irradiation for Eskdalemuir (55.WINDOWS AND SOLARHEAT 109 Example 4. It is evident that except the sunrise/sunset period the procedure under discussion provides reasonable estimates of averaged-hourly irradiation.21) and (4. Calc4-09.2.2. Table 4.2.7 includes measured and computed hourly irradiation data.22) obtain hourly estimates of irradiation.5 Measured values Jan. For any given window design these data may then be used for undertaking the critical threshold analysis presented in Section 4.7 provides the monthly averaged daily irradiation for Eskdalemuir. Using Eqs (4.5 8.5 36 64 28 47 125 172 75 103 219 267 133 164 333 369 189 217 358 411 214 242 406 458 236 267 83 81 56 194 114 289 169 403 233 428 258 500 281 16.3°N.2.5 14. 'D l 'D 10 203 400 233 422 261 486 281 308 181 94 244 147 164 103 86 61 144 397 244 464 267 358 217 414 233 303 183 225 139 94 186 114 347 197 269 156 683 2014 1250 3247 1936 3900 2397 4669 2717 Feb.xlsis a DLLbased workbook. Note that Calc4-09.5 10.2.

FIR: FIR = (IG — ID)/1E (4.110 WINDOWS IN BUILDINGS 4. Precise estimation of this is therefore important for all solar energy-relatedwork. circumsolar and background sky-diffuse irradiance.2. More recent slope irradiance models therefore treat the sky-diffuse component as anisotropic.2. .2. Hourly beam irradiance on a surface of slope TLT (IB.TLT) is obtained by IB.e. Muneer's model (Muneer. very often the sky-diffuseirradiance is the dominant component. is not as simple. The most common of all the skyclarity indices is the 'clearness function' of solar irradiation. 1990a. Computation ofthe diffuse component on a surfaceofgiven orientation and tilt. Under non-overcast conditions the constituent components of sky-diffuse irradiance are circumsolar (sun's aureole) and background diffuse irradiance. i. Chicago and Washington DC to demonstrate the anisotropic nature of the luminance distribution of overcast skies. The simplestof all slope irradiance models assumes an isotropic sky. diffuse irradiation is not isotropic in nature and is an angular function ofthe solar altitude and azimuth.23a) rB = max[O. cos(INC)/sin(SOLALT)] (4.23b) where 'B is the difference between horizontal global (IG) and diffuse irradiance (ID). In this model the slope diffuse irradiation for surfaces in shade and sunlit surfaces under overcast sky is computed as . the development of sky-diffuse models began with the work of Moon and Spencer (1942)who used measured data from Kielin Germany. The sky clarity indices are used to 'mix' the above-mentioned components. Withthe exceptionof clear skies.2.b. however. Many anisotropic sky-diffuse models use one form or another of a sky clarity index to describe the prevailing condition. Historically. Most models under this category decompose non-overcast irradiance as the sum of two components. 5 However.TLT = 'D cos2 T 4.TLT = IBrB (4.14 Hourly slope irradiation The task ofcomputing beam (direct) energy is a matter purely related to solar geometry. 1987.24) A lower value of FIR indicates higher turbidity. 1995) and Saluja and Muneer (1987) treat the shaded and sunlit surfaces separately and further distinguish between overcast and non-overcast conditions of the sunlit surface.2. 'D.

Cb cI .0 j36 (0 0 .

N > > . E E q) > H :1 N N H .

E3c: : 00-Cl) 0. 0 Cb a L .CA) Q. 3 Cb 0 Cl) 0• -.

C) 0 a 0 1 .

) b E -J CN U) ci) .0 0 0 C.) — 0 0 0 :0 0 C.

1340 hours Sky luminancedistributionplots for Garston. UK — overcast sky .3 Luminance distribution plot for Garston. UK Plate 5. 1105hours Measured Computed Thin overcast sky 17 March 1992.Measured Computed Heavyovercast sky 15 January 1992.

6987 F2[forNorthern Europe] (4.73.2. Non-overcast skies.TLT = ID[TF(l — FIR) + FJRrB] + 2b{ir(3 + 2b)}' [sin TLT — TLTcosTLT — iv sin2 (4. UK.08000 — 1.00263 — 0.29a) 2b{iv(3 + 2b)}' = 0.2. exhibit a continuously decreasing behaviour of b and therefore the followingequations.2.27) TF = cos2(TLT/2) (TLT/2)] sky TF is obtained using a value of b.68 for sun-facing surfacesunder an overcast sky. on the other hand.29c) (4.2. based on Calc4-10.8400F2[forJapan] (4.2.2. are recommended: 2b{iv(3 + 2b)}' = 0. obtained via data from 14 world-widelocations (Muneer.2.1 andthe Appendix for further details. 1995).2.1.8 provides hourly horizontal and vertical irradiance for Edinburgh.xls.00333 — 0.TLT = ID[TF] 111 (4.820F— 2.26) where TF is the surface tilt factor.15 Frequency distribution of irradiation of solar energy system performance can be achieved via detailed simulations using hourly or sub-hourly weather data or by simpler computer Prediction .6883F2[forSouthern Europe] (4. 1990a. Estimates of slope irradiance using the isotropic and Muneer models.WINDOWS AND SOLARHEAT 'D. which corresponds to the appropriate and azimuthal condition. 4.2.050 F— 2.29d) 2b{ir(3 + 2b)}' = 0. A considerableamount ofdata on solar radiation and daylightis availablevia the World Wide Web. are included. while b= 1.712 F — 0.0260F2[fortheworld] Example 4. The analysis suggests that the accuracy of predictionsimproves significantly when the isotropic model is replaced with the anisotropic model.415F— 0. See Section 4. For the European climate a shaded surface is modelled with b=5.2.04000—0.6 Table 4.29b) 2b{m(3 + 2b)} = 0. A sunlit surfaceunder a non-overcast sky is modelled as ID.

Figures 4.3 0. In this respect the daily clearness index. 1997a) Date: 12 August 1993.7 0.95°N) (Muneer. is widely used as a categorisingparameter.8 Measured and computed vertical irradiation for Edinburgh (55.2.2. 1997a) — USA . KT.6 KT 0.112 WINDOWS IN BUILDINGS Table 4.5 0.8 0. USA 0.6 Fractional time.f Figure 4.2. respectively. for the USA.4 show the KT distributions.2 Generalised cuives for frequency of solar radiation (Muneer. Units: W/m2 Hour 6 7 8 9 10 11 'G 38 ID 31 South Isotropic Muneer North IsotropicMuneer East IsotropicMuneer 30 55 55 96 269 329 391 581 80 102 172 37 47 51 29 41 36 61 72 81 86 112 102 142 135 108 12 13 14 15 16 17 18 19 542 594 340 763 507 422 221 132 223 204 284 269 215 280 187 121 20 32 122 64 22 84 98 97 77 104 69 62 77 117 79 73 101 83 170 542 414 332 376 210 137 80 104 69 103 13 44 131 16 11 454 336 288 96 77 100 91 106 339 212 135 108 140 455 342 283 299 161 40 96 176 202 271 28 112 190 251 96 77 100 467 400 464 250 235 411 364 418 214 523 140 94 61 61 101 101 67 43 63 180 12 76 55 94 61 61 67 43 524 322 209 73 44 29 8 43 23 14 32 11 34 312 205 68 32 11 445 393 449 222 553 308 171 43 23 12 73 9 methods based on statistical analysis of long-term measurements.9 0.4 0.1 0.2.5 0.2 KT 0.7 0.4 0.2 — 4.

9 1. 1997a) .7 KT 113 0.4 0.1 KT +0.6 0.4 0.6 Fractional time.7 0. f Figure 4.8 0.6 Fractional time.2.5 0.2 0.5 0.3 Generalised curves for frequencyof solar radiation — India (Muneer.3 0.2.55 0.4 Generalised curves for frequencyofsolar radiation— UK (Muneer.0 0 0.3 0.f Figure 4.4 0.8 0.WINDOWS AND SOLARHEAT 0.1 0 0.1 0.5 0.2 0. 1997a) United Kingdom KT 0.

Theabove curves. Passive solar architecture involves the design 4.3. However.1—4. Often the transmission properties usually made available are those for normal incidence. some solar energy. as well as reject.4.3. Johnson (1991) has provided an interesting discussion on this item. present the insolation character for the respectiveregions. as an intermediate step. Glass may be specified by its normal incidence solar . transmission data one has to resort to fundamental physical principles for derivation of these. Almost all buildings collect. Selection of the above-mentioned 'optimum' glazing system requires a detailed hour-by-hour thermal simulation of the buildingand. England.t ofglass. By selectinga suitable glazing system the designermay optimise the admitted solar energy so that the maximum amount of fossil fuel is conserved and yet the building does not overheat during the summer months. incidence angle-dependent. Some of these data were provided by Pilkington plc of Merseyside. The work of Jones (1980) is summarised in Sections 4. KL is the productof the absorption coefficientKof glass and its thickness L. Hawas and Muneer (1984) and Lloyd (1982).114 WINDOWS IN BUILDINGS India and the UK. the transmission characteristicsofthe candidate glazing systems. absorptance(ce) and reflectance(p) of most glass-basedglazing systems are determined by the 'KL' value and the refractive index.1 provides transmission data for single. A small number of missing data were filled in using fitting routines. Jones (1980) has provided an excellentsummary of glazing transmission properties and the relevant physics.3 Solar radiation transmission of buildings to collect the apamount of energyto be used for the provision ofheating and lighting propriate needs. absorption and reflection characteristics ofglazing systems Theangularvariationsin transmittance (t). . glazing manufacturers are somewhat 'coy' about providing the transmission models or characteristics. Table 4.3. The transmission properties are most preciselyobtained via experimentation. 4.and double-glazing systems. In the absence of readily available.1 Transmission. These frequencycurves enable analysis of the availability of solar energy above a given threshold. 1986). also known as generalised KT-curves. This is due to the fact that such glazing systems may only approximately be modelled on the basis of Fresnel reflection properties. Jones' work includes derivation of solar heat gain factors and is the basis of the sections on 'Estimation of plant capacity' (Section A9) and 'Thermal response of buildings' (Section AS) within CIBSE Guide A (CIBSE. This is particularly the case for heat-absorbing glazing systemsas pointed out by Jones (1980).3. The respective plots are based on the works of Liu and Jordan (1963).

* % Solar radiant heat Direct 82 88 85 82 80 78 74 70 67 62 56 49 49 62 56 Glass (type/thickness) Single Sheet/3 2** 3 Light Direct Light Diffuse Shading Total 84 90 88 86 84 83 80 78 76 73 Coefficient 89 91 4 5 6 8 10 12 15 19 25 90 89 88 87 85 84 82 81 78 74 82 84 83 82 81 97 103 101 80 78 77 75 74 71 98 97 95 92 89 87 84 79 74 79 74 68 72 27 87 83 Clear coated/3 Clear coated/4 Clear coated/6 Antisun float/6 — bronze Suncool float/6 — bronze Double 80 79 77 50 10 67 73 72 70 46 9 70 69 65 8 46 6 67 60 54 5 56 45 69 64 64 64 59 62 24 75 72 66 16 Clear float + clear float Clear float + clear coated Clear coated + clear coated Clear float + suncool float/6 Triple 80 75 67 9 79 76 18 Triple clear float Two clear coated + one clear/i *Unless stated the thicknessof glass 72 66 60 67 59 77 71 = 4 mm **Clearfloat thickness transmittance.3)) components thus: .3.2)) and perpendicular (Eq.3. (4.3. T0 and its refractive index t.3.1) The procedure to obtain the angular dependence of the properties under discussion starts with obtaining the ratios of intensities of the reflected beam to the incident beam radiation for polarised parallel (Eq.2 Characteristics for single-glazing (4. sin Oi = /1 sin0r 4.1 Glazing daylightand solar radiation transmission(indicative values). (4.3. The respective angles ofrefraction (Or) and incidence (Of) are related by the well-known Snell's Law.WINDOWS AND SOLARI-lEA T 115 Table 4.

3. Figure 4. and then averaged values obtained for each one of the three properties. TB would be taken as the average of TB calculated for the perpendicular and parallel components.1 presents solar transmission characteristics for five glazing configurations. Thus.6) p = R + [R(1 — R0)2(l — F)2]/[l — — F)2] R(l The glazing properties T.3. For beam (direct) radiation. absorptance and reflectance may then be obtained as 1=1(1 -R)2(l -F)]/[1 -R(1 -F)2] — = [F(l — R)(l + R(l — F))}/[1 — F)2] R(l (4. at 50 intervals. was originally developed by Weir (1998).3. The transmittance.5 to 87.e. (4.7) ponents of polarised light.3.8) = pir/2 Jo I cB(O)sin2OdO (4.3. i.5) (4.2) R1 = sin2(01 — Or)/Sifl2(Oi + Or) For any given angle of incidence the fraction of energy absorbed is (4.3. Transmission functions for other glazing combinations may be developedfrom first principles if basic data related to refractive indices and normal incidence transmissivityare available.3.9). perpendicular and parallel components. The relevant physical analysis and computer tools .3) F = (1 R — R)[(1 — exp(—KL/cosOr)I (434) where represents any one of the above two components (parallel or perpendicular). highlighted in Eqs (4.l)—(4.116 R11 WINDOWS IN BUILDINGS = tan2(O1 — Or)/tan2(Oi + Or) (4.5°.3.3. In the case of diffuse radiation the transmission (Td) and absorption (cd) values are averaged over angles of incidence varying from 0 to 90°. An Excel spreadsheet.3. c and p are calculated for each of the two com- = I. Jo and TB(0)5i1120d0 (4.9) Jones (1980) has recommended that the above numerical integration be carried out between incidence angles of 2. which undertakes the computations and integrations.

(c) coatedheat-absorbing glass.7 ___I 0 102030405060708090 Angleofiocideoce indegrees 02 0102030405060708090 0102030405060708090 indegrees Angieofinddencoin degrees F represents the solar factor are introduced in the followingparagraphs. The following assumptions are made for the interaction of shortwave energy with glass. triple.WINDOWS ANDSOLARHEAT (a) (b) 117 1.1 Glazing solar transmission characteristics (Pllkington.1 boc. The analysis is presented below. Here the numerals 1.oj nbsorbodandM. The irradiation is assumed to be unpolarised and the parallel and perpendicular components of irradiation are assumed to be equal.cd OOS.7INU 0. builds upon the work of Weir and enables easy numericalmanipulation.rd.th 0. (e) float glass + floatglass.od . ._____ O. (b) coatedclear glass.3 Characteristics for double-glazing The characteristicsofdouble-. The transmissivity and absorptance are calculated for the above components and then an average value adopted.dIo. 2 and 3 refer to the system elements. 2I :! .4 Eo.oasmkaioo Eoeegy ndmlOodby (C) 0102O30405060708090 (d) Angloofirn. presently included in the compact disc.or indeed quadruple-glazingsystemsmay be obtained by the extensionoftheprocedure laid out above for single-glazing.bcd eIes.:.a.eolejccadby sllee000 Eoorny O.3.xls. WorkbookCalc4-11.1 Eo. 4.6 0.ideneein degrees 0102030405060708090 Angle of ioeideoee (C) Figure 4. ciri' respectivelydenote the proportion ofthe incident radiation transmitted by the entire system and the fraction of the energy absorbed in the nth element. t. (a) Float glass.3. 1969)... 0.OU \ Angleofincidence indegrees 8 0. (d) Solarshield glass. This workbook will be more appropriately introduced later in this section. 1 being for the outermost elementof the system and so on.

14) + [1BtIBrBp3B/{Z(1— P1BP2B)}] (4.2)—(4.13) = ttt3/Z 1B = 1B + [1BtIBp2B/(1 — P1BP2B)] (4.3. tB = X1BT2B/(l — P1BP2B) (4. 4.3.17)have been extracted from the work of Jones (1980).15) (4. Note that Eqs (4.3. which comprehensivelyproduces the solar transmission characteristics for single-.3. Z = (1 — P1BP2B)(1 Thus.10) iB = 1B + 1BT1Bp2B/(l — PIBP2B) z2B (4. — P2BP3B) BPIBP3B (4.17) 2B = 2Bu1B(l — P2BP3B + t2BP3B)/Z 3B = 3Bt1Bt2B/Z It has to be mentioned once again that the characteristicsfor sky-diffuseand ground-reflectedradiation are obtained by replacing the subscript B with that for the respectivecase.and triple-glazed windows. Calc4-11.3.3.3.3.16) (4.5 Computer-aided tools for obtaining window solar transmission characteristics An Excel workbook. double.3.12) = 2Bt1B/(l — PIBP2B) The characteristics for sky-diffuse and ground-reflected radiation are obtained likewise by replacing the subscript B with that for the respective case.3.118 WINDOWSIN BUILDINGS For direct radiation.xlsenables the determination of the angular dependence of direct solar and daylight transmission through multi-glazed windows. with any corn- .3. For ease of presentation a dummy variable Z is introduced which represents either one of the terms in any given expression or the entire denominator.11) (4.4 Characteristics for triple-glazing For triple-glazing the systems expressions for transmittance and absorptance become quite involved. 4.3.

'double' (row 8) and'triple' (row 13) worksheets. and4e-4-4e(enter 0. One such compendium. These data are thenrespectivelycollatedin the 'single' (row 5).e. is Pilkington (1998). The former type of worksheet provides the necessarytransmissivity for the direct. Microsoft Excel does provide a facility to fit polynomial functions if x (independent variable). transmission models in the form of polynomial equations relating the direct tranmissivity to the angle of incidence are often desired. Calc4-11. However.WINDOWS AND SOLARHEAT 119 bination of uncoated and low-emissivity coated glass. Using such normal incidencedata in conjunction with the Calc4-11. However. Data for other types ofglass ofvarying thicknessesand more innovative coatings suchas silver. row 8 for 'double' and row 13 for 'triple'. The coated glass transmission data refers to Pilkington-K type material with a low-e coating of tinoxide.3.86 in cell C4) for a double-. with a high degree of accuracy. The computational chain provided in the suite of Eqs (4. This table enables the selectionofthe glasscombination for themulti-glazedwindow bykeyinginthe type /position of each element of the glazing system in cells B3:D5. as used by the German glazingcompany Interpane. y (dependent variable)data are provided in acolumn-wisemanner. Note that data for the direct component of daylight transmission are also provided in a graphical format within each of the above-named sheets.4e-4 (enter0. row 5 for 'single'. This workbook. and the latter sheet the transmissivity for diffuse component of solar radiation.xlsworkbook.72 in cell B4) for a single-.3.xls. It is possible to use Excel's Solver tool to fit polynomial functions.e. or indeed non-linear models.3.Whereas direct radiation transmissivitiesare presented in each of the above-mentionedworksheets as a function of the incidence angle. Most users would be content with the numerical output of the angular transmission data available for each of the respective glazing worksheet. 4.2. the diffuse radiation transmissivityis provided as a single value (refer to (Eq.7.2)—(4. The object is to minimise the objective function.8)). A demonstration of the use of the Solver tool is provided in Example 2.72 in cell D4) for a triple-glazedwindow. the sum of squares of errors. contains the basic property data for the normal incidence solar transmission (T0 value) for the commonly used glass (Table KL2 within 'KL'). i . located on the extremeleft-hand end. Note that (cell A5) stands for refractive index. There is an ever-growing demand for window transmission data from architects and the building services engineeringcommunity. Ifany other glazingmaterial is used thentheusermaykey-in therelevant data as discussedabove.l7) is carried out within the worksheets TR1D through to T3d. the user may easily obtain the angular transmission characteristicsfor multi-glazed windows. The default glazing combinationis provided as 4e (enter 0. includes 10 worksheets and a description of each of these worksheets is provided below.the accuracy ofsucha fit is far from satisfactory. by varying a given set of 'seed' coefficients. The worksheet called 'KL'. i. In this respect glass manufacturers are continually publishing new material. which provides normal incidence solar transmission data for a large number of glass types. i. is included. may be inserted within Table KL1 by theuser.

(plots (a)—(d))and double-glazing(plot (e)).3. normal and severe conditions as being those which are respectively exposedunderwind regimesof 3. 1998).3.1(a)—(e) presents the angular transmission characteristics of single. (4. 1993.18) where and h0 are respectively the internal and external surface conductance. 8 and 24 km/h. It may be used with almost equal effectiveness for obtaining direct or total angular transmittance provided the normal incidence transmittance (ta) 15 known: = 1. The work undertaken by Pilkington (1969) covers an analysis of the total heat contributionof multi-glazedwindows (solar factor) under steady-state conditions.19) (4. (1993) have provided a very simple yet robust model for obtaining angle dependent transmittance (r) of multiple-glazing systems.3. The total solar transmission is also known as the solar factor.3.(Eq. For a singleglazing it is defined as F=x+fx glazing (Eq.4. The symbol 'F identifiesthe curve for total solar transmission underthe above-mentionednormal condition of exposure.120 WINDOWS IN BUILDINGS 4.3.3.2. CIBSE (1986) defines sheltered.018[cosO+ sin3 Ocos8] (4. F For any given set of glazing the radiation that is neither reflected nor transmitted is absorbed andtherefore raises the temperature of the glass. and triple(4. (4.20)) It is possible to extend this analysis to double.20) (4.3 and 0.3.3. For a double-clear glazingthe F factors corresponding to the above conditions are 0. Figure 4.22) .21) and F respectively refer to the solar radiant transmission and solar factor for the nth pane ofa multi-glazing. of the solar energy absorbed by single-glazingthat is released inwards has been shown to be J h1 + h0 (4. The fraction. These plots have been extracted from information contained in Pilkington (1969.n being the number of the innermost j For building surfaces.21)) as F—t1F2 F='r1'r2F3 where pane.3. Baker et al.6 Total solar transmission. 0.

6. It will be shown in this section that with the development of superinsulated windows. Shading coefficients are calculated for radiation at normal incidence.3.2 Heat capacitance effects in multi-glazed windows To date. shading coefficient and visible transmittance. however.and short-wavelength shading coefficients for a number of glazing combinations. The program was developed at the Lawrence BerkeleyLaboratory and its main features are its ability to compute U-value.1 Shading coefficient Solar radiantproperties are often compared by theirshading coefficient which is derived by comparing them against thetotal transmittance ofa clear glass of thickness between 3 and 4 mm. The shading coefficientsare therefore fairly constantfor almost all angles ofincidence.WINDOWS AND SOLARHEAT 121 4. Window 4. 1994). Strictly speaking under the varying conditions of solar radiation experienced under European and North American climates.3. The program contains optical spectral data files for glass manufactured by ten major manufacturers.1 is another useful tool for obtaining the radiation transmission characteristicsofmulti-glazing(LBL.html. The program with a companion manual is availablewithout cost from its producers as wellas being downloadable via the internet from the following website: http://windows. all glass manufacturers have reported values of F under the assumption of steady-state conditions.lbl. the total shading coefficient. mass x specific heat capacity (c) — ftc — total heat transfer coefficient (h) x surface area (A) The time constantrepresents the time taken by any given solid to change its temperature to 37% of its initial value. For other angles of incidence the glass is compared with clear glass in a likewise condition. Sometimesthe long. in particular for the newer designs of superinsulated windows.6. The latter has a total transmittance of 0. 4. For any glazing. The thermal time-constant. For further discussion on this item referenceis made to Suryanarayana (1995).and short-wavelengthcomponents are given separately so that their individual effects can be examined. the above assumption is not valid for proper load estimations. the above analysis becomes somewhat redundant.87. solar heat gain coefficients. tttc = pLgC .Pilkington (1998)is a useful source to obtain the long. The most commonly reported property is. of any body is z.gov/materials/materials.

3.2 Time constant for inner glazing hse or cavity conductance Glazing Single— 3 mm Single— 4 mm Double.8 h1 (W/m2K) 8.3 8. 4 mm) are respectively the density and thickness of the glass. c= 750 J/kgK. Passive solar design seeks to optimise that contribution (ETSU.The assumption ofsteady-state transfer of solar heatis thus no longer valid bearing in mind the large irradiance fluctuations under European and North American climates. However.3 h (W/m2K) 25. Eh is the sum ofthe inner and outer heat transfer coefficients. This implies that the heating and cooling-down periods ofthe modern glazing systems are considerable and this warrants a dynamic solar energy transmission analysis. with little or no me- chanical assistance.3 8.0 25.1 t (minutes) 3. passive solar design makes use of sunlight to heat and light buildings.6 n/a n/a air Krypton Xenon 1. 1985).3. Historically. A comparison of these makes it clear that the assumption of steady-state transferof thermal energy is not valid. 1 low-e — 4 mm Infihlgas (W/m2K) 16. Incidental heat gains and daylight already make a contribution to the energy needs of most buildings.and double-glazing while Figure 4.3 8. float 4 mm Double.The adoptionof passive solar techniques can substantially reduce the amount of energy needed to operate a building.122 WINDOWS IN BUILDINGS Table 4.0 14.16 0. This is due to the fact that the time-constant for a low-e coated super-windowis four times that ofa single glazing.2 shows the time-constants for single. Figure 4.3.1 summarises the performance of one such house.46 9.75 5. It was found that. 4.7 5.92 9.4.and low-tech doubleglazing were valid due to their short thermal memories. the steady-state conditions for single. with the advent of superinsulated windowsthe time-constant has quadrupled to nearly quarterofan hour. Table 4.0 8. I low-c — 4 mm Double.7 16. like with like.2 13. . The energy consumption ofthese houses was then compared with similar houses without passive solar features.22 where p (2500 kg/rn3) and t5 (nominal value. 1988).3 8.9 13. passive solar measures reduced heating bills by 40%.2 shows the irradiance plot for Edinburgh for a frequencyofinterval of one minute. In the early 1980s over 100 solar houses built at Milton Keynes in the UK were closelymonitored (ETSU.4 Passive solar buildings Combined with the appropriateenergy efficiency measures.

Figure 4.2 One-minute frequency. time-series plot for solar radiation (W/m2) measured at Edinburgh (28 America This type of variation in the intensity of irradiation is common in Europe and North 1600—1700 hours). .400 300 200 100 0 C\J C'J o c'j 0( C'J (0(00 (0(00 (0(00 (0(00 J (0(00 • (0(00 9090 9 (0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0(0 t- March 1993.3.

or with thermostatically controlled fans. Overheating can be avoided by providing thermal mass. sun-spaces provide pleasant extensions to the home for up to three quartersof the year: typically conditions within them are like outside conditions three months nearer summer. Such direct gain passive designs. When solar gains are enough to raise the sun-space above house temperature the heat collected can be let into the house by opening communicating doors and windows. As well as thermal benefits. vents at floor androoflevel and reflective blinds. which can be constructed at little or no extra cost.1 Energyflows for a solarhouse.124 WINDOWS IN BUILDINGS 2500 2000 1500 1000 500 Useful incidentalgains4300kWh J A' S' 0' N1 D' J 'F'M' A' M J Figure 4. and this reduces heat losses from the spaces house and warms any ventilation air which passes through the sun-space. . Design studies in the UK show that sun-spaces canbe equally effective in large and small houses and can make passive design possible even in high-density urban housing. July to June (Muneer. Another method ofcapturing solar energy is to add an extra highly glazed unheated room — a sun-space or side of the house. With good design and proper use. Solar gains always make sunconservatory — to thesouth warmer than the outside air. and many of these are highly glazed. Sun-spacesopen up a different range of design possibilitiesand unlike direct gain features. In the UK. around 50000 house extensions are built annually.4. their thermal performance can be similar to or even betterthan directgaindesigns. 1999) Worthwhile improvements in the energy performance of buildings can be achieved by arranging windows and skylights to allow solar energy into areas that need to be heated. typically have large windows on the south side andsmaller glazed areas facing north. they can readily be added to many existing houses.

but less well in more unpredictable climates.&2 Design schematicfor a passive solar design (Muneer. or replace the masonry wall with a water wall. Thewall itself is massive. Solarradiation Direct gain Solarradiation Attached sun-space Figure 4. Figure 4. Trombe walls work well in some climates where there is a regular alternative ofsunny days and cold nights. which collects heat during daylight hoursand releasesit to the building interior via conduction and convection. 1999) . and uninsulated. Other design approaches insulate the mass wall and use blown air to transfer heat into the house. A Trombe wall can be thoughtof as a shallow sun-space designedsolely for solar heat gain and cooling.2 shows the design schematic of the above-mentioned passive solar heating systems. so that solar heat collectedduring the day radiates into the living spaces overnight. User-operated vents at the top of the wall allow warm air to flow into thehouse. dark coloured externallyto maximise heat absorption. it conducts heat slowly from outside to inside over 6—12 hours.4.WINDOWS ANDSOLARHEAT 125 One of the more complex passive solar design is what is known as the 'Trombe wall' — a glazedheat collector.

The sites considered were Easthampstead (England).argon-filled 'Kappafloat' windows and a medium-weightconcrete floor. comfort and energy-exchange. daylight penetration. Eskdalemuir (Scotland) and Aldergrove (NorthernIreland). QMAX FQ F5Q5 -. wind-conductance.126 WINDOWS IN BUILDINGS 4. These require long-term.)(I (1-Fc)(1-cz)(J-Fs)Qs Fs)Qs4? C (1-aL)(1-FL)QL C (a) Conductance network (b) Energyinput fornodes Figure 4. The forthcoming CIBSE Guide J addresses the present lack of availability of such quality-controlleddata. hourly data sets of solar radiation and other weather parameters. 1990b) . The simulation procedure was based on precise models for slope irradiation. ambient temperature and wind speed for four sites.çFLQL R 5L( l-FL)QL cx(lFs)Qs F (lFp)Qp w Fc(I-a.1 Simulation of passive solar heated buildings CIBSE has identified the pros and cons of a number of sophisticated design tools. The building design involved a southern facade of double-glazed. Aberporth (Wales). The simulation study used data from three years which consisted of hourly horizontal global and diffuse irradiation.4. An hourly simulation design study was undertaken by Muneer (1990b) to determine the performance ofsolar-heated and daylitoffice buildings underthe climate of the United Kingdom.43 Thermal model fora direct-gain solar-heatedoffice building (Muneer. These windowsoffer a high daylight and a moderate solar radiant heat transmittance.

The other three nodes for wall. I F 28 26 0 x Indoorenvironmental temperature Outside ambient temperature oo p p 20 L) 3 Ninth decile 18 16 14 12 10 8 Fourth decile - N W N U N N N X 6 4. floor. Q..4. Likewise the convective proportions of the solar gains.3 shows the thermal model for the type of building considered by Muneer (1990b).floor and three inner nodes I. walls and inside glazing and infrared radiation exchangesbetweenthe ceiling. FLQL are added to node R. QE represents the convective heat input due to office equipment. The conductance network takes into account the convective heat exchangesbetween room air and ceiling. and 13. R. 1990b) . and lighting gains.WINDOWS AND SOLARHEAT 127 Figure 4.4.glazing. glass and room air were considered to be non-capacitative due to their small thermal mass and consequently small time-constants. and floor. hour Figure 4.A fraction of the total sensible heat emitted by the occupants. The energy inputs to the various nodes shown in Figure 4.4.walls.4 Free temperature swings for a solar-heatedoffice buildingin west London (Muneer. 2 0 0 *N N N U I N N I N 12 First decile U N N N N K I 4 8 16 20 24 Timeofday. isconvective and hence is added to the room convective node. Thedynamic thermal model has five capacitative nodes:ceiling.3(b) are as follows.

4.4.4. The aim is to demonstrate the possibility of the satisfactory functioning of such buildings under the influenceof local climate. Figure 4.5 shows the breakdown of the above-mentioned energy quantities.4 shows the fourth to ninth decile bandwidth for the indoor temperature and the first to ninthdecile bandwidth outdoortemperature.4. Thestudy showed that 50—60% savings might be achieved in the primary electrical consumption by exploiting daylight potential. 1990b) . In passive solar buildings the above is rarely the case.5 Contribution of solarenergyto heatingand lightingof an office buildingin the UK (Muneer. For an office environment Wray (1980) has shown that the optimum comfort temperature is about 22°C with a comfort range of about 4°C.4. 280 " 240 E 200 - Daylight contribution Electrical consumption Solarheat gains Auxliary heat 0 160 - 120 0 •0 0 E 80 - ': HR Easthampstead ilift Aberporth Aldergrove jJft __ Eskdalemuir Figure 4. An environment in which the mean radiant temperature (TR) and air temperatures (Ta) are equal is termed as thermally uniform. The solar gains in conjunction with the occupancy and equipment gains were found to be adequate for most of the space heating requirements.2 Examples of solar-heated buildings The objectofthis section is to report briefperformance results of a few solarheated buildings in the UK. Thus the comfort temperature lies in the interval of 18—26°C.128 WINDOWS IN BUILDINGS One way of evaluating the free (uncontrolled) temperature swings of solar heated buildings is to plot percentiles or deciles of hourly temperatures. Solar-heated buildings require special treatment of thermal comfort evaluation. Figure 4.

6 presents the monitored winter performance of the Berm House as reported by Muneer (1990b).to heavy-weightdesign the temperature would also remain within the comfort band for longer periods. In an observational study (Davies. The windows cover the entire southern facade of the cube-shaped test room.8 respectively show the performance of the test room on a long-term and short-term (an average winter day) basis. SouthWales. The main part of the building is covered with 1. The building faces 30° west of south and is built on the bank of the river Usk. About half of the outerwall area is covered with double-glazedwindows. only 40 days out of 494 were noted in which the classroom temperature was below 16°C. In a medium. Another example of effectively utilising solar gains for space heating is the 'Berm House' in Caer Llannear Monmouth. Figures 4. With a more realisticthermal mass it would be possible to stabilise the temperature and bring it down to fall within the comfort band. The respective measured U-values have been found to be 0. Thus.5°C.4. which act as thermal diodes. with the other sides and door being heavily insulated. Facing almost due souththe two blocks ofextension to theexistingbuildingcontainan assemblyhall. The unit has been successfulin producing sauna-like conditions. with the collaboration ofNordanwindow manufacturers. Thethird example is the Napier solar 'sauna' (Muneer. No conventional heating is used. At the heart of the 'sauna' design are high-performance triple-glazed windows. During the summer the mean daily temperature exceeded 23°C on 34 out of 494 days. It may be seen that both designs have been successful in producing extremely warm conditions indoors.direct gain system formaximisingthe temperature boost. The project's aim was to investigate the feasibility of using superinsulated windows in a lightweight. in 1995. Hourly records of temperature and incident solar radiation have been maintained.HEAT WINDOWS AND SOLAR 129 The first exampleis the extension of St George's School (now St Mary's College).a gymnasiumand classrooms. The heating season in Scotland typicallylasts up to 10 months in the year. Two window designs have been tested. exceptthose daysin which heavyovercast occurred. Itwasnoted that. which is followedby a solar corridor and another wall with single-glazed windows. 1976) carried out between 1 January 1969 and 28 July 1970.near Liverpool. Almost half ofthe southelevation is a directgain system with two skins ofglass with a 62 cm cavity in between. the building is heated by solar radiation and the heat from the lighting and the occupants.4. The building contains 10 main rooms. Argon.5 m of soil with a garden at the top. Opened in 1962. 1997b).73 and 0. The maximum temperature noted was 24.4.4 W/m2K. the temperature for at least 4 to 5 hours aroundmid-day remained in thewarmto unpleasant end ofthe comfort band. an experimental test unit was developed in Edinburgh.and xenon-filledglazing have been evaluated for their U-value performance and their contribution to solar energy gains. each for a period of one year.7 and 4. Wallasey. 8 of which are accommodation units. . Figure 4. Throughout the monitored period the temperatures were found to be higher than in the older building which was heated by conventional means.

and xenon-filledtriple-glazing. Figure 4. This measured performance for Edinburgh may now be used to investigate the potential of duplicating such an application for other locations.4. i.4.Indoor (minimum) — — Indoor (maximum) * (-) 14 12 E 10 8 6 4 2 0 Figure 4. 1990b) An abbreviated way of analysing the potentialfor the use of solar energy is to plot the annual receipt of vertical surface radiation (supply)against heating degree-days(demand). e.9 presents a supply and demand scenario for various British locations. Figure 4. Muneer's experiments have shown that even under the cold and dull climate of the east of Scotland the superinsulated windows have a good potential to provide comfort conditions within unheated conservatories. The performance for many other European and North American .4.130 WINDOWS IN BUILDINGS 24 22 20 18 16 —+— Outsideambient(minimum) —*-- Outsideambient(maximum) —0--. maintain the temperature of the conservatory above 20°C for over three-quarters of the time.6 Winter heating performance ofthe 'Berm House'. highest heating demand and lowest level of supply of the solar energy resource. Records have indicated that either of the above two windows. It may be noted that Edinburgh presents the severest test.South Wales (Muneer. argon.9 shows data for seven UK locations.g.e.

7 Thermal performance of the Napiersolar 'sauna' (Muneer. Solid line(sun-space temperature). Solid line (Sun-spacetemperature). . IIIIIIIIrIIIIririr -5.4. S . Maximum daily temperatures (Celsius) forthe solar sauna with xenon windows. dashed line (outside ambient temperature) 65 55 45 35 25 :1_ ]MrrFrrl IlIllIrlIl IFIIt!IIIII(IIJIIIIIIIpIIrIIIIIuIIIIII!!IpIII[pIrIIIIruIIIIIIlII!III Illitlil *1111+11+11+fl*tI. dashed line(outside ambient temperature) Figure 4. 1997b) locations may thus be extrapolated andindeed it would be reasonable to expect a better thermal performance. It was shown above that a larger supply of insolation and reduced demand for heat may provide an overall better thermal performance for passive solar .WINDOWS AND SOLARHEAT 131 Maximum dailytemperatures (Celsius) forthesolarsauna withargon windows.

9 Solarirradiation (resource) versusheatingdegree-days (demand) diagram for UK locations (Muneer.) (U • London • Manchester 3000 3100 3200 Birmingham Newcastle Belfast C C 2900 3300 3400 3500 Heating degree-days (°C) Figure 4.4. 1997b) C) 28 27 Cardiff c'J 26 25 24 23 22 2800 I I I Edinburgh I 0CG.4. 1997b) heating using the now commonlyavailable super-insulatedwindows.3 Economics ofsolar heating The primary function of a well-designed solar space heating system is to optimisethe reduction of the auxiliary energy consumption in such a way that the minimum overall cost solution is obtained. For maximum exploitation of solar energy for space heating applications there must be a sufficient heating demand.4. With an abundance of available insolation but little or no demand for . However. 00 D D (UI. 16 February 1996 (Muneer.8 Hourly temperatures and incidentradiation for the Napiersolar 'sauna'. the situation may alter when economicsis brought into the picture.132 WINDOWS IN BUILDINGS Figure 4. This will be discussed in the following section. 4.

1 provides a summary of the results for a given solar-heated building.4.47 45.67 680 760 710 1 000 4 720 4 560 5 100 1 550 2 450 7 230 2 640 2300 4 650 1 090 3 540 3 290 2660 2 410 Ajaccio 43.WINOQW5 AND OLAFf AT 133 heating.05 830 590 3160 1 090 1160 3 510 . the laws of economics would work against the use of solar energy.82 45.13 52.MacGregor analysed insolation and other weather data from 25 world-wide locations to obtain the energy quantities on the two sides of the equation — solar irradiation on a vertical surface (supply) and heating load on a candidate building (demand).67 55.20 35.80 890 910 810 3 790 2 530 2210 Rome New York Messina Albuquerque 40. The lack of uptake of deliverablesolar energy for De Bilt for almost a third ofthe year is noticeable (nil heating load during summer).heating demand and solar energy contributionto space heating (MacGregor.42 kWh/m2 Heating demand kWh 14 020 Solar contribution kWh 5 690 930 670 770 670 690 790 780 660 760 640 7 400 6 800 7 750 4 240 3 270 4 250 Copenhagen Eskdalemuir Hamburg Berlin Abepporth De Bilt Valentia 6 430 6 540 5 540 5 760 1 550 3 070 3 230 3 810 2 780 London Brussels 660 1 050 770 700 820 4 920 5 410 11 320 5 970 4 610 3 470 2 670 2 620 5 040 3 090 Winnipeg Wurzburg Paris Tours Bolzano Limoges 2 570 3 110 2 500 Turin Nice Madison 46.78 60.92 41. The last column of Table 4.18 43.15 55.10 51.10 which compares the solar fuel savings for Lerwick (60°N) against De Bilt (52°N).13 41.4. Table 4.63 52.38 52. Based on the above argument MacGregor (1981) has postulated the idea that the northern(or high-latitude) locations are the ones which aremost suited for an economicalapplication of solar energy.4.1 provides an indication of the annual solar contribution towards saving of conventional fuel.32 53.80 49. This point is further amplified by Figure 4.1 Annualirradiation on south-facingverticalsurface.47 50.4.82 47. 1981) Irradiation Location Bethel Lerwick Latitude °N 60.80 48.90 49.93 51. It is apparent that the maximum exploitation of solar energy concurs with higher-latitude locations.77 38. Table 4.

The shades comprise fins bolted at an angle of 45°.10 Solar energysupplyand heatingdemandfor two European locations.11.12.9 W/m2K has been quoted for the low-e coated doubleglazing. Simulations have indicated that the temperature may rise above 25°C for only 5% of the occupied period per year.4.4 Designing buildings against overheat During the 1990s a large number of buildings have been built which use large glass facades and yet effectively rely on naturalmeans ofventilating/cooling.4. Solar gains and daylight are fullyexploited with high levels of insulation and the right choice of thermal properties for the building structure. One example of avoiding overheating of such buildings is that of Oxford Instruments headquarters (Bunn. The design brieffor this buildingwas to maintain comfort conditions without introducing mechanical cooling.It is not possible to enumerate here the very many fine examples which clearly demolish the dictum that a large glass facade results in summer overheating. Most ambient lighting is switched off by timers. 1981) 4.4. January—December (MacGregor.134 WINDOWS IN BUILDINGS kWh 1000 500 J FMAMJ J ASOND J FMAMJ J ASOND Figure 4. yet usingpassive solar energy during winter. The 4800 m2 building which accommodates 110 people uses 50% glazing for the north and south facades. 1998). Occupants tend to open these when additional ventilation is required. Refer to Figure 4. Specific strategies for environmental control against overheat are shown in Figure 4.5°C and the external ambient is above 15°C. The Tokyo Electric Power Company's new R&D Centre is another good exampleofa buildingwhich is designedagainst overheat. The building is located in Abingdon Business Park in Oxfordshire.4. the latter protected by a fixed brise soleil. A nominal value of 1. By automatically controlling the . The windowsare all operable by hand levers. The atrium roof has glazed smoke vents which act as ventilators when the internal temperature rises above 21. althoughnight cooling is carried out using an air handling plant. The passive solar architecture sets out to thermally condition the space using the building fabric.

4. This actionprevents excessive buildingcooling loads. the slats are raised to admit beneficial energy from the sun.13 shows another innovative natural ventilation technique developed by Monodraught Limited. England (Monodraught. however. 1998). 1998) angle of the slats.12 a reduced the cooling load to a fifth that of interior blinds with the same solar shading characteristics.If there is a low level ofsunlight. UK (Bunn. .4.4.13.5 Windcatcher system Figure 4.WINDOWS AND SOLARHEAT 135 Summer day Fully openable windowsacross entireelevation underdirect occupantcontrol Summernight ** * * * Secureflight-time ventilationachieved bysupplyingair from raised roofvoid Iii IT ________ / Exposed concrete coffers Figure 4. Studies have indicated that the blinds system shown in Figure 4. the system ensures that irrespective of the direction of the wind there will always be one side that will be subject to wind pressure. 1998).4.11 Naturalventilation for OxfordInstruments buildingin Oxfordshire.4. This technique provides controlled ventilation within buildings by exploiting roof-level wind admittance. With reference shown to the 'wind chimney' shown in Figure 4. 4. Further information on this building project is provided in the CADDET Newsletter (CADDET. the blind prevents direct solar radiation from entering the room beyond certain distance.

12 Blind control for overheating avoidance (CADDET.4. 1998) .Sun at low angle ofincidence Angle of incidence Sun athigh angle of incidence a — Lineofsight To affordoutside views as much as possible. slats are controlledat minimumangles required for solae shading Sun hidden by cloud furlongperiods Th Th —Line of sight jiiiij Son becoming hidden by laad -Th -Th -Th Line of sight Th Th Line of sight Th Blind is raised to secure maximumoutside views Light reflected by blinds makeslighting more effective Flat Figure 4.

1998) In summer months doors and windowsare more likely to be opened and this aids cross-ventilation.13 Schematic the operation ofthe Monodraught 'Windcatcher' natural ventilation system (Monodraught.4 m/s Iclosed ksqTemperature.WINDOWS AND SOLARHEAT 137 SUMMERDAYTIMEOPERATION Cool air Warm air 2 mIs open NIGHT-TIMEOPERATION-' (and mid-season) Cool air Warmair — vailingwind. Warm air rises to ceiling level and is carried out through the passive stack element.4. with fresh air coming in through windows on the windward side of the building and being exhausted through the passive stack. In mid- . 5C say of Figure 4. The advantage of this type of ventilation system is that it is not dependent on openable windows or vents in the side of the building.

4. and Steemers. 391—393. December. CEN/TC 129/WG9 N. Both. 1998. Newsletter. Italy. DOC.138 WINDOWS IN BUILDINGS season or during daytimes or weekendswhen the building may not be in use the system continues to ventilate naturally. TheBuilding Research Establishment. Since the building may be otherwiseclosed at night. Secretariat UNI. The dampers are programmed to close again at set times in the morning and a temperature limit can be introduced to ensure that the building is not over-cooled. . the prevailing windpressure pushes air throughout the passive stack. the system achieved a ventilation rate of five air changes per hour corresponding to a wind speed of 4. 4. With the wind speed dropping to 1. with temperatures ranging up to 30°C. Fanchiotti. Centre for the Analysis and Dissemination of Demonstrated Energy Technologies. Any prevailing wind carries this air down into the building below. (1993) Daylighting in Architecture. No. In a study undertaken at the University of Hertfordshire two lecture theatres were monitored during the summer of 1998. 4. has undertaken a series oftests to evaluate the effectiveness of the above Windcatcher system. Solar Energy 12. This leavesthe building interior fresh without security being compromised. Overnight cooling resulted in a 2°C drop in temperature ofthe building. flow visualisation and the cooling effect of the naturally ventilated air were investigated. A. day. opaque sky cover and percentage of possible sunshine.and night-time cooling strategies were monitored. James and James Ltd. CEN (1997) Glass in building — Determination of energy balance value — Calculation method. November.UK. 19 Rev. References Baker. Comité Européen de Normalisation. January 1997. Ventilation rates.5 rn/s the ventilation rate dropped to 1. Bennett.. 2. Building ServicesJournal20—24.5 m/s. N.6 Night-time cooling Night-time coolingor 'free cooling' is considered to be one of the most important aspects of natural ventilation strategies. 20133 Milano. I. (1998) Precise services. Via Battistotti Sassi 11B.Itwasnoted that in August 1998. K. London. CADDET (1998) Energy efficient windows.5—2 air changes. (1969) Correlation of daily insolation with daily total sky cover. R. The volume control dampers can be programmed to open fully at night to encapsulate the cool night air. The volume control dampers at the base of the system at ceiling level will control the amount of the airflow through the system depending on the internal temperature levels and external weather conditions. Bunn.

LBL (1994) Window 4. 53. Kinghorn. J.A.B. B. (1986) An analysis of Linke turbidity factor. Butterworth-Heinemann. 333.P.WINDOWS AND SOLARHEAT 139 of Building Services London. (1979) Solar and terrestrial radiation dependent on the amount and type of cloud. F. McGraw-Hill. 6.A. July 1982. 393. Solar Energy 4. T. 177—189. Solar Energy Unit. 2.G. Solar Energy 64. Kasten. Chartered Institute 28(6). (1991) Solar Engineering of Thermal Processes. Jones.A. New York. London. (1981) Solar Energy Handbook. (1979) The average distribution of solar radiation — correlations between diffuse and hemispherical and between daily and hourly insolation values. A. (1980) Solar radiation throughwindows— theoryand equations. J. LBL-35298. (1984) Generalized monthly Kr-curves for India.. Lawrence Berkeley Laboratory. M. Solar Energy 7. J. Engineers. Energy Cony.. and Beckman. M. Erbs. B. (1982) Estimation of the diffuse fraction ofhourly. (1963) The long-term average performance of flat plate solar energy collectors. Solar Energy 22. Boston. 293—302 ETSU (1985) Passive solar design. Solar Energy CIBSE (1986) CIBSE Guide A. New York. S. London. 99—108 Hawas. T. (1998) Solar radiation models based on meteorological data.. (1976) The contribution of solar gain to space heating. and Kambezidis.A. Wiley. G. Solar Energy 18. MeteorologicalMagazine 107. Louche. Cardiff. Klein. Energy Technology Support Unit. Cowley. (1998) Daylight illuminance frequency distribution: Review of computational techniques and new data for UK locations.E. H. R. and Rabi. Pen. and Muneer. University College. Report No. Kreider. and Muneer. Liu. (1982) A study of some empirical relations described by Liu and Jordan. & Mgmt 24. Muneer.H. (1991) Low-e GlazingDesign Guide.S.C. Solar Energy 24.L. . Johnson. J. G. and Kreith. Solar Energy 37. Berkeley. CA. W. (1978) The distribution over Great Britain of global solar radiationon a horizontal surface. D.H. 361.F.Y.G. A. and Czeplak. USA (March 1994).1 — A PC Program. P. M. D. and Jordan.H. T.C. Gul. F. 155. M. Energy Technology Support Unit. R. DTI. Lloyd. and Duffie. daily andmonthly average global radiation. Davies. Duffie. BSER&T 1. 357. diffuse and total solar radiation. M. DTI. ETSU (1988) Using energyfrom the sun. (1960) The inter-relationship and characteristic distribution ofdirect. 185. 83—91. and Jordan. 139. and Iqbal. R. LR&T 30. Liu. T. Collares-Pereira.Y. 1.

BSER&T 3. St Helens. J.K. Green Enertopia Conf.E. Springer-Verlag. and Sahili.Pilkington Glass Ltd. Oxford. 357. Mani. Ilium.G. Pilkington Glass Ltd.T.F. P. Palz.G. diffuse solar radiation. M. T. and Spencer. T. LR&T 27. Owens. London 37. Owens. (1989) Monthly mean solar irradiation availability for the United Kingdom. (1999) Solar energy. (1984) Correlation between hourly diffuse and global radiation for New Delhi. 265. Monodraught Ltd. Pye. sunshine and cloud in the UK. (1996)European Solar RadiationAtlas. J. (1977) Correlation equation for hourly diffuse radiation on a horizontal surface. Muneer. New Delhi. M. Architectural Press.G. Kent. (1997b) A solar powered sauna.K. and Hollands.. St Helens. Solar World Forum. BSER&T 11. and Saluja. 4. Pilkington Glass Ltd. 141. Muneer. eds: D. Hawas. 209. D. Muneer. A. England. (1990b)Solar heated and daylitoffices: Design study. W. T. (1981) A comparison of climatic suitability of various locations in the European community for solar space heating. Muneer. 707. (1984) Heat reflective coatings on glass. T. (1986) Correlation between hourly diffuse and global solar radiation for the UK. Muneer. England. Tonbridge. P. Pilkington (1998) Glass. T. Pilkington (1993) Glass in Building: A guide to modern architectural glass performance.T. Berlin. J. London. Eng.140 WINDOWS IN BUILDINGS MacGregor.A.W. BSER&T 11. PhD thesis. October 1997. Brighton. andGrief. Report presented to CIBSE Solar Data Task Group. Page. St Helens. T.S. (1987) Solar Radiation Modellingfor the United Kingdom. I: Sloped surfaces.T. (1997) Proposed quality control procedures for the Meteorological office data tapes relating to global solar radiation. 1—3 September 1997. 153. MillerFreemanUK Ltd. Muneer. BSER&T 5. Orgill. Muneer. England. 37. Moon. P. (1982) Effective U-value. & Mgmt 24. (1981)HandbookofSolar RadiationDatafor India. Pilkington (1969) Windows and Environment. (1997a) Solar Radiation and Daylight Modelsfor Energy Efficient Design of Buildings. Button and B. T. England. Trans. Solar Energy 19. Energy Cony. Proc. High Wycombe. CNAA. Korea. Soc. T. Cheju. Allied Publishers. K. . BSER&T 10. T. 81—85. Proc. 189—192. 75. G. In Kempe's Engineers Year Book. 4. BSER&T 7. (1942) Illumination from a non-uniform sky.. K. Muneer. Muneer. T. Muneer. Monodraught (1998) Windcatcher natural ventilation system (brochure). (1995) Solar irradiance and illuminance models for Japan. 2. (1990a) Solar radiation model for Europe.

CA.WINDOWS AND SOLARHEAT 141 Saluja. Edinburgh. G. Royal Greenwich Observatory. Solar Energy 25. Arch.y. T. Napier University. 197. Geoph. G. Inst. (1968) Report NASA TM-X-1646. (1987) An anisotropic model for inclined surface solar irradiation. 6. 11. . Solar Energy 29. (1982) Correlation equation for hourly diffuse radiation on a horizontal surface. Woolf. Cl. Met. B. A. 327.O. (1998) L?fe Cycle Assessment ofMulti-glazed Windows.W. J. West Publishing Company. England. UK.D. W. PhD thesis. USA Wray. NASA. 19. (1980) A simple procedure for assessing thermal comfort in passive solar heated buildings. Yallop. Proc. H.M. (1995) Engineering Heat Transfer. Series B 7. Weir. Cambridge. (1956) The determination of hourly values of total radiation from daily summations. Suryanarayana. N. New York.S. (1992)A simplemodel for solar declination (unpublished report). Mech. Moffet Field. Whillier. Engrs 201. Spencer. Biokl. and Muneer.

Page blank in original .

Daylighting offersdramaticreductions in the energycostsin buildings with the high lighting and cooling loads typical of commercial premises. In the newer offices lighting accounts for some 50% ofthe electricalload. This is due to the fact that the luminous efficacy of daylight is much higher. and yield significant savings in electricityconsumption through the reduced need for artificial lighting and the consequent reduced need for cooling. Natural light gives the time of day andthemood of seasons to enter' (Ander. software design tools will be presented to enable computation of daylight penetration through multi-glazing. In this chapter a review of recent daylight research is undertaken. It contributes to occupiers' comfort and satisfaction.4. So good lighting design reduces cooling needs as wellas directelectricityconsumption. however. Careful building design can produce comfortable levels of interior lighting without glare.1 Daylighting fundamentals Thefundamental definitions in presented this sectionare primarily based on BS 4727. Computation of daylight factor will be introduced in Section 5.5 WINDOWS AND DAYLIGHT ThefamousarchitectLouisKahn has been very aptly quoted as having said 'A room is not a room without natural light. 5. The designofnon-residential. Daylight is. The BREEAM method of environmental assessment cites good daylight as a contribution to healthy building design. Additionally. a greatly under-utilised energy resource.Passive solar daylit spaces are popularwith building occupants and daylighting systemscan be fitted to existing buildings. BREEAM designguidelinesdemand a daylight factor of 2% for 80% of the occupied space of the building. Part 4: Glossary of terms particular to lighting and colour. Perimeter daylighting uses glass in various locations to allow light to entertheperimeter rooms of a building. Electric lights generate more heat than light but daylight does not. Daylight is a crucial and critical part ofa building's design. 1995). passive solar buildingsusually starts with consideration ofdaylighting. which in turn can have implications for absenteeismand productivity benefits. In the Netherlands health regulations forbid buildings where staff sit further than 6 m from a window. . For commercialoffices it has been reported that a saving of 30—70% by exploitation of daylight results in a payback period of 2 to 3 years.

It is measured in candela (cd) which is equal to one lumen per steradian.3 I/luminance (E) (5.2 Luminous intensity (I) Thisis definedas the power of a source or illuminated surface to emit light in a given direction. The SI unit of luminousflux is lumen (im).1. The relationship between the flux 0 (lumen) from a light source and the uniform intensity (cd) in a solid angle w (steradian) is given by I çb=I.1) This is the luminous flux incident per unit area.1. 5. The unit lumen-hour is usually associated with the science of illumination and it represents the quantity of light emitted by a one-lumenlampoperating for one hour.144 WINDOWS IN BUILDINGS 5. and proportional to thecosine of the angle made by the normal to the surface with the direction of the light The direct illuminance at a point which results from a large source of a uniform diffusing surface may be obtained for a small element on the source surface.1. Candela is also defined as the luminous intensity of a monochromatic light source that emits radiation at a frequency of 540 THz and an emissive power of 1/683 Watt per steradian. The result is then integrated for the whole source surface.4 Luminance (L) Luminance is measure of brightness of a surface.w 5. can be calculated from the followingexpression: rays. It may also be defined as the intensity oflight emitted in a given directionper projectedarea ofa luminous or . or received by a surface. It is measured in lux (lx) which is equal to one lumen per square metre. Illuminance is inversely proportional to the square ofthe distance ofthe light source. An alternative way of defining lumen is that it is the amount of light which falls on a unit area when the surface area is at unit distance from a source of one candela.1 Luminous flux (4) This is defined as the light emitted by a source. A lumen is the amount of light emitted in a unit solid angle by a source of one candela output.1.1.1. The direct illuminance E (lx) at a point located at distance H.2) 5. below the centre of a disc source of radius R. E= LR2 a R2 (5.

1998). A more detailed discussionon the sky luminance distribution will be presented later in this chapter. can also be used which expressesthe light emitted in terms of lumens per unit area. apostilb (asb). Therelationship between the two units is given by Luminance (asb) = it x luminance (cd/rn2) (5. With the advent of .using digital control technology provides a smooth logarithmic variation. Modern buildings frequently employ designs and equipment which enable maximum exploitation ofsolar heatand light with the added possibility of glare and overheat avoidance using weather-sensitive. can provide up to 92% protection against solar heat. The maximum possible dimming down results in an electricalpower consumption saving of 85% (TLC. One such building is the new European Court of Human Rights in Strasbourg which. The latter can be considered as the effective source of daylight.1. The unit ofluminance is cd/rn2 ofa projected area. controlled blinds. on demand.5 Source of daylight Thesun providesthe daylight in two ways. Under this assumption and for a sky luminance of L cd/rn2. however.3) 5. The most modernofdaylight-assistedlighting controls have the capability to provide dimmingup to 1% ofthe full range (TLC.WINDOWS AND DAYLIGHT 145 reflecting surface.1.2 Windows as daylight providers Recent developments in technology have shown that significant savings in electrical consumption within the building sector are possible through exploitation of daylight. 5.The latest equipment. 5. A non-SI unit.6 Luminance distribution of the sky Originallyit was assumed that the sky had a uniform luminance. shown that the sky has a non-uniform luminance. Research has shown that daylight has an important bearing on the human brain chemistry. The human eye is very sensitive in the 0—10% dimming range where any sudden change is uncomfortable. Light entering via eyes stimulates the nerve centres within the brain which controls daily rhythms and moods. the horizontal illuminance due to an unobstructed sky hemisphereis given by (5.1. Some of the other part is scattered by the atmosphere and produces the blue sky. 1998).4) Measurements have.1. Part of the sun's energyreaches the earth's surface as direct sunlight.

maintaining optimal mood conditions for longer (Cawthorne. however. however. • Buildingswith low daylight factorcreate environments with homogenous lighting. Useoflarge window areas avoids this. comfortable for the occupant. A typical office worker could spend 50% or more of their time in environments of 0. 1994). • One unwanted aspect to the presence of windows is the generation of disability glare and discomfort glare. These are summarised below. Lighting quality is characterised by many factors including quantity. It is recognisedthat a holistic approach to lighting design is required to provide environments which are pleasing to the eye. Theabove limit will. ceilings and floors. 1996). 1991). content and contrast. creating conditions likened to those found externally. Windows as energy providers are key elements in any solar energy building design and a more detailed discussion on their energetic impact is provided later in this chapter.5. • On a window/floorarea ratiobasis. . 1996). maintaining potential for larger overall energy saving due to reduced lighting loads. building aspect. This could be improved by rethinking the office environment. and ultimately must be suited to the tasks being carried out andthe comfort of buildingoccupants. and building construction. finishes applied to walls. Christoffersen(1995)foundthatratios much above 25% significantlyreduced net energy savings for buildings. work performance and general productivity varies. Collins. 1980. location and architecture.1—100 lx (Espiritu. when contrasting fields of brightness and darkness exist. 1992).146 WINDOWS IN BUILDINGS superinsulatedwindows it is possible to provide much larger glazed areas thus exploiting daylight and passive solar heat gains. It is dependent upon window size. A more detailed discussionon glare is provided in Section 5.increase with the use of superinsulated windows. construction and transmission properties. room characteristics. Research related to daylight quality and its effects on perceived comfort. • People prefer environments with windowsanddaylight conditions (Wyon • and Nilsson. avoiding window proportions of 40—55% (Boubekri and Boyer. but that window/floor arearatiosaround 25% allowedquality daylight to be transmitted. The average person receivingmore than 1000 lx from natural daylight for less than one hour per day is not receiving sufficient levels to maintain optimalmood. The impact of these can be minimised by appropriately sizing glazing areas in a facade. Weir(1998)has undertaken an exhaustive literature reviewofthe above factors and hencedrawn a number of conclusions. 1975). and may recover from operations and illness more quickly in environments which are daylit and afford an exterior view (Loe and Davidson. Discomfort glare occurs. whereas those with high daylight factor transmit more quality • daylight. having little contrast and holding limited interest for the occupant. and which do not limit work productivity (Loe and Davisdon.

Occupants who arecontented with their environment find it easier to channel theirattentionto work tasks.1995) Atlanta Chicago 300 240 183 Denver 300 Los Angeles 300 228 156 New York Seattle 300 243 192 300 246 198 No daylighting Sidelighting only Side toplighting 300 231 171 + 240 174 . distraction is reduced and work productivity is increased. anddiffuse compared to thelightfrom the sun. Using at least two windows on different walls or a skylight may improve the balance of the admittedlight. Admitting daylight into a building is an important function of windows.3 Design of daylighting systems The introductionof natural light is a powerful architectural tool as it defines and shapes the interior spaces. This factor is more significant in commercial buildings compared to houses.WINDOWS AND DAYLIGHT 147 • Improvements in daylight penetration to the indoor environment can significantly lessen energy consumption on artificial lighting systems (Zeguers and Jacobs. The more light admitted. Reflectiveground surfaces and walls may be used to increase daylight admitted through southand north-facing windows.2. Location of windows within a building shellcan be selected so that direct sunlight reflects offinterior walls and floors and provides more diffuse and even light. is that people prefer daylit environments and enjoy the benefitsassociated with windows. What is clear. It has also more visible light with a significantly smaller infrared Table 5. It is therefore difficult to produce confident relationships between daylight level andimprovements/detrimentsto work output and quality. However. the daylight quantity should be controlled and only recommended values should be used. 5. however. Admitting daylight via windows andskylightscan save up to halfof theenergy than could be consumed ifthere was no daylighting.2. theirlocation.1 Lighting energy costs (MWh) in an officebuilding (Ander. The daylight admitted into a building should be balanced to provide visual comfort and to enable occupants to perform their visual tasks. 1997).1 shows the importance of daylighting in reducing energy consumption due to using artificial lighting in an office building. Lightfrom the sky is cooler. Table 5. and the average reflectanceof the interior and exterior surfaces ofa room. Research on the provision of daylight and its impact on work performance and productivity is still in its infancy. The balance of light in a space depends on the overall number and size ofwindows. Daylight may save energy by reducing relianceon artificiallighting. the better people can see. gentler.

• Changeand varietyChange and variety are at theheart ofdaylighting. 1 Overcast sky: defined as being a sky in which at least 80% of the sky dome is obscured by clouds. They influencethe appearance of the interior spaces within a building. Sunlight is fundamentally good both therapeutically and visually. Small windows. In Section 5. for example. • Sky conditions Sky conditions vary the nature and quantity of the light entering a building. During the day theview in is difficult becausetheexternal light a window such as trees and other buildings may affect the amount of daylight entering a space.1 When Design issues the design process of daylighting system is employed for a building. north windows can provide the best quality of daylight of any orientation. 5.3. View out depends to an extent upon location. Thus. the colour of a white wall will vary with time ofday but it will always appear white due to human adaptation to the natural light. 2 Clear sky: defined as being a sky in which no morethan 30% of the sky dome is obscured by clouds. • Sunlight Optimisation of daylight increases the overall level of light and assists in providing change and variety.All otherforms of light change the perceivedcolour to a greater or lesserdegree. It usually includes widely varying luminance from one area of the sky to another and tends to change quite rapidly. certain issues should be considered. • Colour Daylightis thecolour reference. size. These changes of daylight conditions are due to weather. season and time of day. • View Windows provide an importantlink to the outside world. The illumination levels produced by the overcast sky may vary from a few hundred to several thousand lx. In buildings. Three types of sky conditions are utilised to estimate illumination levels within a space. break up a view. • External obstructionsand orientation External obstructions that surround with the outside world and the view out are important aspects of orientation.148 WINDOWS IN BUILDINGS component. contact . 3 Cloudy sky: defined as being a sky in which 30—80% of the sky dome is obscured by clouds. shape and detailing of the window. The awareness of the daylight pattern. Windows also allow theview in. depending on the density of the clouds. in situations where daylight is desired with minimal solar heat gain.6 all-sky luminance distribution models and relevant Excel workbooks will be presented.and the window is the medium through which it delivered.

WINDOWS AND DA YLIGHT

149

is greaterthan that inside. At night the reverse occurs, allowing view in. Using curtains andblinds may, however, control the problem of privacy. • Glazing The type of glazing can significantly affect daylighting. Clear glass, for instance, provides goodvisibility.Tinted or reflective glass helps reduce glare. Much of the solar gain is absorbed by the tinted glass and then released inside the building. Thus, tinted glass has less effect on reducing air conditioning. In addition, tinted glass produces conditions where the occupants switch on the lights, thus increasing energy use. Photochromic glass changes colour according to the ambient light conditions. This also helps reduce glare. Electrochromic glass becomes over opaquewhen a current is applied to it and thus can be used for privacy.

5.3.2 Design strategies
The strategies an effective daylighting system may be summarised as for
follows.

• Extending the perimeter form of a building may improve the building's
by increasing daylighting • performanceaperture high in a wall allows deeper penetration of daylight. Locating an

the total

area.

• Using the optimum glazing area to achieve daylight saturation, as addi-

• Reflectingdaylight within space results in increasing room brightness by and evening out brightness patterns. • spreading ceiling from the fenestration area will help increase the Sloping the
space. ceiling • brightnessdirect beamdaylight on critical visual tasks. Poor visibilityand Avoiding

tional glazing area will increase the cooling loads more than it will reduce the lighting loads.

of the

further into a

discomfort will result if excessive brightness differencesoccur in the vicinity of critical tasks. • When harshness ofdirectlight is a potential problem, vegetation, curtains, or louvres can help soften and distribute light more uniformly.

5.3.3 Design elements
The most significant design determinant when implementing daylighting strategies is the geometry of a building's walls, ceilings, floors, windows, and how each relates to the other. An understanding of the effects of the various building elements will provide the basis for manipulating form to achieve adequate lighting levels.

• Exterior elementsExterior elementscan be usefulcontrols for fenestration.
Examples are overhangs, light shelves, horizontal louvres, vertical louvres (fins), daylight tracking and reflecting systems.

150

WINDOWS IN BUILDINGS

• In-walland roofelementsTheamount of daylight within a building canbe
glazing technology have reduced this liability. The emerging window technology has been discussed in the first chapter of this book. • Interior elements These may be further categorised as room geometry, room surface reflectance and interior shading control. A brief discussion of these items follows. increased by increasing the total amount of glazing area. However, admittingmore light may bring in unwanted heat gain. Recent advances in

The depth that daylight will penetrate is dependent on the ceiling height relative to the top of the window. A high window will allow daylight to strike the ceiling plane and be reflected into the interior space. In a typical building with a window height of 2.5 m and a room width of 3.7 m, daylight can penetrate about 6 m from the window elevation. The depth of the room has a direct effect on the intensity of illumination. With deeper rooms, the same quantity of incoming light is distributed over a larger area. The ceiling is the most important surface in reflectingthe daylight coming intoa space onto the workplane. The next most importantelementis the back wall, followed by the sidewalls, and finally, the floor. Several types of manual interior control devices can be usedto eliminateexcessivebright spots and also get daylight where it is needed. Examples are venetian blinds, draperies, and roller shades.
5.3.4 Lighting controls The demandfor energy efficiency has led to an increasing application of intelligent lighting controls. These may be categorised as: Time-scheduling, Presence detection, Daylight linking, Manual switching, and Intelligent luminaires.

5.4 Daylight factor

In design studies it has become customary to specify interior daylighting in terms ofdaylight factor (DF). This is the ratio of the internal illuminanceto the
external illuminance, available simultaneously and is usually expressed as a percentage. The daylight factor is divided into three components: the direct skylight (sky component), SC, the externally reflected component, ERC, and the internally reflected component, IRC. Thus

DF = SC + ERC + IRC

(5.4.1)

The SCand ERC are both found by considering the geometry ofthe visible skyor the external reflected surfacesat a point on a horizontal plane within the

WINDOWS AND DAYLIGHT

151

room. They can be found from a Waldram diagram, BRE protractors, BRE tables, and Pilkington paper pot diagrams. TheIRC is basedon inter-reflectiontheory andcan be foundfrom formulae, BRE tables, and BRE nomograms (Building Research Establishment, 1986). 5.4.1 Calculation of sky component (SC) The sky component is the ratio of illuminance at any given point that is received from a sky of known luminance distribution to the horizontal illuminanceunder an unobstructed sky hemisphere.Likewise,the external and internal reflected components are, respectively, the ratios of the illuminance received after reflectionsfrom external and internal surfaces to the horizontal illuminanceunder an unobstructed sky hemisphere.An electroniclookuptable for the skycomponent, basedon the CIE standardovercast sky, is provided in CalcS-01.xls. This table was originally published by Hopkinson et al. (1966). The BRE protractor and the accompanying literature (Building Research Establishment, 1986) enable estimation of the above three components. It is worth mentioning that Calc5-Ol.xlsrequires the datafile 1n3-5.CSV which is included in the compact disc. TheBRE sky component tables maybe usedto obtainthedaylight level at a referencepointin a horizontal plane ifthe height and width of the window and the distance of the reference point from the window are known. Figure 5.4.1 shows these details. The electronic lookup table, Calc5-01.xls, gives the sky component at the intersections of the ratio H/D and W/D. The geometric construction has to be such that the horizontal and vertical planes drawn through the given point to meet the window wall perpendicularly form two bounding edges of the window. If the window sill is above the referenceplane theheight of the sill above the referenceplanemust also be takeninto account. Example 5.4.1 demonstrates the use of CalcS-0l.xls.

5.4.2 Calculation of externallyreflectedcomponent(ERC)

If direct entryof sunlight or skylight is restricted through thewindow then it is necessary to calculate the ERC. In this case the obstruction's luminance is taken as a fraction of the obscured sky's luminance. Usually the reflectanceof
the obstruction is taken as 0.1—0.2. If a significant proportion of the sky is obscured then the value of the ERC is so smallthatdaylighting is oflittle value. Further information on the computation of ERC is provided by Pritchard
(1995).

5.4.3 Calculation of internallyreflectedcomponent (IRC) on the work undertaken by the UK Building Research Establishment. The design chart is for a 6 m square room with a 3 m ceiling and a singlewindow.
Figure 5.4.1 provides minimum values of the IRC. This information is based

152

WINDOWS IN BUILDINGS

C)

2

5

7

10

15

20

25

30

35

40

45

50

Window areaas parCentoffloor a area

Figure 5.4.1 Minimum value of the internallyreflectedcomponent

In summary the sky component, SC, is by far the dominant part of the daylight factor.
Example 5.4.1 Consider the case ofa single-glazed rectangular window shown in Figure 5.4.2. The external obstruction runs along the entirelength of the room. Obtain the sky component using CalcS-01.xls. Also obtain the ERC and IRC given the reflectance of the external and internal surfaces are respectively 0.1 and 0.5. The floor reflection factor may be taken as 20%, the wall reflection factor is 50% and the window/floor area=0.2.
5.4.4 Calculation of SC

With referenceto Figure 5.4.2, compute the following:
(a) W/D = Tan20° = 0.364,H1/D = Tan45° = 1.0, andH2/D = Tan 35° = 0.7

Note that the vertical portion of the window between H1 and H2 receives daylight from the sky canopyand the portion between H2 and H3 receives the
wall-reflectedcomponent (ERC). (b) Open workbook CalcS-0l.xls. Notice that this includes a FORTRANbased DLL program [Fr305.dll].Therefore, Excel should be launched first and then Calc5-01.xlsopened within Excel(seebox below).Note the two relevant files1n3-5.csv andFr305.dll must be loaded in the same PC folder as Calc5-Ol.xls.

WINDOWS AND DAYLIGHT

153

Sky canopy

Figure 5.4.2 Window schematicfor Example 5.4.1 (Muneer, 1997)

(c) Insert the values for the H1/D, W/D pair in cells B4 and B5 respec-

(d) Press Alt+F8. The Macro dialog box appears. Click on Run button to run the FORTRAN-based DLL program [Fr305.dll].The sky component 2.02 (cell B6).

tively.

85%. Appreciation of brightness is governed not only by the actual luminance of the habitated environment but also by the brightness of the .6 Calculation of IRC Using the data provided for the wall and floor reflection factors and window/ floor area the IRC is read off from Figure 5.364. (f) The actual sky component is then obtained by subtraction of the above two values and then doubling the result (due to symmetry). sun-facing or shaded aspects.68%.4. more precise procedures are presented which take into account the window aspect and the real sky luminance distributions. It must be borne in mind that the above computation ofsky component does not take into account the effects of window orientation. it represents the effectiveness of the window as a lighting provider.85%.84=1. The ERC is now obtained as the product of its (subtended) SC and reflectance= 1. 2(1. Thus. SC=l. Secondly. 5.18 (cell B6) is obtained. The actual sky component of the external obstruction is therefore.1 as 0. Constancy is associated with the concept of adaptation.17% and IRC=0.1 = 0. It is therefore only an approximate method of obtaining internal illuminance. The ERC in this case is thus of negligible value.68%. sky component at point P=2 x 0. or simultaneously press ALT+ R keys (e) Repeat steps (b) and (c) for the H2/D. The value of 1. Thus.346)= 1.4. Firstly. Hopkinson et al.18—0.67 x 0. W/D pair.In later sections.g.154 WINDOWS IN BUILDINGS For all DLL-based workbooks the following procedure MUST be used: • Launch Excel software • Open the DLL-based workbook • Enter data in the relevant worksheet(s) • Simultaneouslypress ALT +F8 keys • Select the relevant macro from the dialog box • Either click the Run button. For this combination Calc5-01. due to its simplicity the above procedure for sky component estimation is widely used.17%.4. However. the daylight factor= 2.5 Calculation of ERC Note that W/D=0.xls provides a SC value of 0.346. the daylight factor remains constant even though the outdoorilluminancemay fluctuate.7%.67%.364 and H3/D=Tan 20°=0. 5. e. (1966) have enumerated the advantages of daylight factor as follows. We note that for the given situation. ERC=0.

visual of the interior of a room does not change radically even though appreciation the actual luminancewill be higher as a result ofthegreater amountofdaylight surroundings which govern the level penetration resulting from brighter skies. Discomfort glare. A generalised form ofglare quantification can be derived by studying the averageresponse of a large numberof people to the same glare situation.8. One solution is to design in such a way that the recommended level of internal illuminance is attained during a certain agreed proportion of the working period throughout the year. especially if the window faces east or west. The relevant material is presented in Section 5. As a result.e. or sand. 55 Glare Glare is the excessive brightness contrast within the field ofview. An example of this type of approach is to be found in the work of Hunt (1979) for Bracknell andKew in the UK. Glare can be classified into two types: discomfort and disability.4. Windows may give rise to glare. display screens.1 Discomfort glare This is defined as glare which causes visual discomfort without necessarily impairing the vision.WINDOWS AND DAYLIGHT 155 of visual adaptation. water. Eq. shows the illuminance frequency distributions for world-widelocations. 5.g.5. At any given point the daylight factor will result in wide variations in internalilluminance. as presented by Baker et al. The excessive contrast between foreground and background may disturb the eye's ability to distinguish objects from their background and to perceive detail.1)provides a very rough estimate ofDF and is therefore not recommended for serious design work. An average but 'loose' value of DF may also be calculated in a single step. The human eyecan function quitewell over a wide range of luminous environments. Daylight reflection from such surfaces intensifiesthe glare problem. extracted from Pilkington's (1993) design book and Kinghorn's work undertaken at Napier University. People orient windowstoward an interesting view such as snow.3.8. It is possible to undertake illuminance frequencyanalysis for other locations too using the procedure in the following section. The bright sky may be close to the line of view. (5. Glare is a subjective phenomenon and difficult to quantify. It may cause reflectionson work surfaces. Figure 5. from a light source can be expressed as: . (1993). Although easy to use. Light from the sun may shine directly or by reflection to create glare. but it cannot function well ifextreme levels of brightness are present in the field of view at the same time.Acceptanceof a given illuminancelevel as thecriterion of an appropriate visual environment poses a problem in relation to the variability of available daylight.

kix (b) C h -0 5 10 HongKong Lisbon. Portugal Athens.Portugal Athens.3 Cumulative frequencyof (a) globaland (b) diffuseilluminance for world-wide locations x 6 Discomfort glare = L6 Lb < x 0.1) where L. .France Usbon. France Nantea. NewYork Bracknell.France Perth.45 (5.UK F 15 20 25 30 35 40 45 Measured diffuse illuminance. klx Figure 5.4. cv is the angular size of the source and P is the position index which indicates the effect of the position of the source on glare. Greece Albany. France Lyon. NewYork Bracknell. NigerIa Nantee. Australia Darwin.156 (a) WINDOWS IN BUILDINGS I 10 20 30 Hong Kong Lagos. Australia Albany. Greece Lyon.UK 40 50 60 70 80 90 100 110 120 130 Measured globalillumirtance. is the luminance of the source.5. Lb is the average luminance of the background.

% wall area 50 60 Figure 5. and thendecreases as the window size increases beyond 50%.. Lateral positior (mean value) ..1 shows that perceived glare rises from 1.WINDOWS AND DAYLIGHT 157 5.5..66 ' . - — A0 _____ 20 30 40 10 Window size.1 Influence of windowsize on glare (Boubekri and Boyer. Boubekri and Boyer (1992) studied the effect of window size on sunlight presence and glare and noted that the discomfort glare of sunshine can compete with the positive psychologicaleffects of sunlight.1 Intolerance zoni — olerancezone — — /' Frontalpositioi (predicted) . Its effect can be expressed as a shift in • Small windows glaring source is small and perceived sensation is not • disturbing. Perceived glare is in the tolerable range.2 Disabilityglare the adaptationlevel of the eye.5. The reasons for this may be summarised as follows: This is defined as glare which impairs the ability to see detail without necessarilycausing visual discomfort. except when the window size is 40—55% of the wall area.5. Figure 5.7 as the window area increases from 20% to 50% of the wall area. Medium windows a highcontrastbetween glare sourceand surrounding — — adjacent wall leads to a higher perceivedglare level. 1992) .4 to 4. r ________________________________ ________________ I I Room reflection=O.. The occupant experiences discomfort when the perceived glare value rises above a value of 4. Boubekri and Boyer found that window size accounts for less than 30% ofvariation in perceivedglare. A survey carried out in nine UK schools showed that 47% of teachers complained of thermal and/or visual discomfort from the sun. S. and that they were more tolerant towards the sun than pupils..

6 Availabilityof daylight In theUnited Kingdom the latest update of the building regulations aims to improve the lighting efficiency (CIBSE. the designer is prompted to encourage the maximum use ofdaylight. KG = [680f V()IG(2) dA]/[f IG(A) d] (5. Studies such as these open up the 'market' for large glazing facades with the obvious consequence of exploiting daylight as well as passive solar energy. These in turn require values ofhorizontal global and diffuse illuminance. 5. To meet the deemed to satisfy approachset out in the newregulation. including the use of solar altitude. x1 and x2 determine the lower and upper limits of the visible bandwidth. Boubekri and Boyer's (1992) work has shown that adequate space layout and optimum placement of furniture in the office have a large potential to reduce discomfort glare. 5.6. KG.1 Horizontal global and diffuse illuminance The luminousefficacy of daylight depends upon the way in which the radiant energy is shared between the visible and invisible (infrared and ultraviolet) parts ofthe spectrum. Muneer (1997) has presented a survey of these models. water vapourcontent.This in turn depends upon a numberoffactors including V) is the CIE spectral sensitivityofthe human eye and IG(2) is the solar .158 WINDOWS IN BUILDINGS • Large windows— thoughtheglare source is large. With the growing use of daylight in office buildings. 1996).6.In the absence ofmeasurements it is necessary to resort to luminous efficacy models to obtain an estimate of global and diffuse illuminancefrom othermeasured or estimated atmospheric parameters. i.1) spectral irradiance. Several approaches have been adopted.e. Linke turbidity factor and many other atmospheric parameters. Efficient and more precise design of windowsin buildings requires the development of inclined surface illuminance models. Research has also indicated that a lateral viewing positionkeeps the levels ofdiscomfort glare well within the acceptablecomfort limits even for quite large window areas. is expressed as the ratio of illuminance (lx) to irradiance (W/m2) which can be found through the integration of the whole spectrum. raising the adaptation level of the eye and reducing the glare sensation and the level of discomfort. the issue of glare continues to receivewidespread attention. thecontrast betweenthe source and the surroundings is small. The global luminous efficacy of daylight.

Pleijel showed that clear and overcast skies vary little in luminous efficacy with solar altitude (and thus with time of year) but that there is a marked decreasein the efficacy ofthe sun's beam radiation at solar altitudes less than 30°. England. 5.6 — 74. 5.6. overcast or average) and the altitude of the sun.1 Luminous efficacy The human eyesare distinctly different from those of other animals. the human eye has a much larger proportion ofcones.5lK + 57. the sky alone and for the global radiation(sun plus sky).342lK (5. Thus.WINDOWS ANDDAYLIGHT 159 state of the sky(clear.3) . overcast and average skies.2) KD is the diffuse irradiance luminous efficacy and a. the global efficacy was found to vary between 105 and 128 lm/W. Determinationshave been made by workers in different parts ofthe world of the luminous efficacy of daylight from simultaneous measurements of the illuminance and irradiance.6.6. i. b1. It is desirable to obtain luminous efficacies under all-sky conditions. Blackwell's measurementswere related to global radiation with clear. the luminous efficacy is different for the sun alone. The meanefficacieswere found to be 119. 1990) for all-sky luminous efficacy is given as et KG orlCD = a1 + bl + c1 cos z + d1ln(i) (5. l. denotes the optical of the cloud cover and is the atmospheric precipitable water transparency content. This differenceis due to the proportion ofrods andcones in the nerve endings of the human eye's retina. KG and = 136. the 5. Dogniaux (1960) at Uccle.6. In particular. c1 and d are the coefficients which are functions of sky clarity.e.6. 120 and 116 lm/W respectively. With average skies. Blackwell (1954) at Kew. 1997). Belgiumand Drummond (1958)at Pretoria in South Africa.1. Such models are presented in the followingsections. Worthy of note among the earliest measurement efforts are Pleijel (1954) in Scandinavia.1.3 Muneer—Kinghornmodel The philosophybehind thesemodelsis that the luminous efficacy is most significantly influenced by the sky clearness index (Muneer and Kinghorn.2 Perezmodel The Perez model(Perez al. The cones respond over the 400—730nm wavelength band with a varying strength of the output signal.1..

the Muneer—Kinghorn model was usedto produce long-term illuminancedata sets for a number of UK sites.4.9797K (5. at least for the temperate belt of the world. When considering the design of an integrated natural and artificial lighting system.4) The above models have been rigorously evaluated against an extensive set of measured data gathered from a variety of locations across the UK as part of the International Daylight Measurement Programme. Data related to the site location andsurface geometry are to be provided within the 'Site' sheet. as shown presently.2 — 39.828K1+ 49.6. such as school gymnasia.6. Findings confirm that in the event of unavailability of long-term measured illuminance data.1 for getting help in the execution of this workbook. Average luminous efficacy models also show promise. Note that Calc5-02. Indeed. it is of primary concern to recognise what quantity of daylight is Buildings . Refer to the text box provided under Example 5. certain local county councils have adopted policies whereby commissioned new structures.xlsand Calc5-03. using measured irradiance values in conjunction with a luminous efficacy model is the most reliable method of synthetically generating an illuminancedatabase.6. Muneer and Angus (1993) have shown that for UK locations the respective average luminous efficacies of 110 lm/W and 120 lm/W for global and diffuse components are comparable with otherluminous efficacy models. Figures 5. namely 'Site'. 'Compute'and 'Graph'. must incorporate daylighting intotheir design before being considered for approval. If a validation of any of the three models is required the user may enter the relevant measured data under the 'Measured data' columns within the 'Compute' sheet. The similarity between the computed and measured traces is remarkable. it also enhances corporate image and public perception towards environmental concern. respectively depicting the calculated global and diffuse illuminance values against their corresponding measured counterparts. The sheet named 'Graph' is dynamically linked and therefore will provide a visual means of evaluation. Each of the above two workbooks contain three sheets. respectivelyusing the above mentioned Perez and Muneer—Kinghornmodels. During the preparationof the CIBSE Guide J. Not only is this advantageous from the energysaving perspective.xlsis a DLL-based file.160 WINDOWS IN BUILDINGS KD = 130.1(a) and (b) demonstrate this point.2 Frequency of occurrence of llluminance are increasingly being designed today to exploit natural daylight availability as effectively as possible.xlsenable luminous efficacy and zenith luminance computations. The workbooks Calc5-02. 5. Modern office complexes can now even incorporate the use of light-pipe technology to bring daylight to central cells within the structure that are remote from fenestration.

1 Variation in (a) global and (b) diffuse illuminance derived from measured irradiance (Edinburgh. Apr11 1993 data) .6.WINDOWS AND DAYLIGHT 161 x 0 Ct C E Ct 0 0 2nd 3rd 4th 5th 6th 7th 8th x 0) 0 C Ct C E 0) 0) 2nd 3rd 4th 5th 6th 7th 8th Figure 5. Napier University.

For example.68 for 20% of the time. such as prevailing cloud-cover and sunshine. It therefore provides an indication of the prevalent atmospheric clarity. respectively. network of synthesised solar irradiance datasets. Kinghorn and Muneer (1998)presented the data shown in Figures 5.162 WINDOWS IN BUILDINGS available. recorded over periods of up to 20 years. these data may be used to construct a widespread. The curvesenable querying the availability of solar energy above a given threshold. Due to the lack of long-term historical solar illuminance measurements. various researchers have developed techniques to mathematically model levels of illuminance at frequent intervals. 5.2(a)and (b) for the cumulativefrequency ofglobal and diffuse illuminancefor a number of locations in the UK. for Indian locations. Through creation of a syntheticallyderived solar illuminancedatabase the frequency of distribution. k . This type ofinformation is a useful aid for daylight designers for exploring the potentialfor energy savings via modernphotoelectric controls.2.for the USA. Strictly speaking the above insolation frequency distributions are for daily basedquantities. Figures 4.1 Liu and Jordan's method The background to this method was discussedearlier on in Chapter 4.2. KT <0. The generalisedKT-curves present the insolation character for the respective regions. 1998).4 showed the KT distributions. or below. Such data are routinely and reliably recorded at many more locations than solar irradiance.6. It is therefore evident that the distributions for the Indian locations are flatter.e. Thus the above curves may also be used to obtain the frequency occurrence of hourly global irradiation. Hence. A comprehensive review of procedures available that allow daylight frequency levels to be estimated from meteorological data is presented in the literature (Kinghorn and Muneer. hourly and subhourly. and 0. Using such measurement data.7. of daylight can be determined. Themost accurate method of enablingdaylight estimationsis by interlinking the luminous efficacy models with thecorresponding irradiance measurements.2—4. This means that for the Indian subcontinent the daily clearness index (KT) varies in a narrower range and thus indicates a stable solar climate. during a month for which = 0. 56% and 30%. In contrast the corresponding figures for KT the USA are. Previous work in the field ofirradiance modelling has focused on developing estimations based primarily on the availability of other meteorological data. respectively. yet accurate.2.73 for 70% fractional time. Access to such data enables the lighting designer to predict the frequency of occurrence of natural illuminance above. or cumulative frequency. However. Recall that the dimensionless quantity KT represents the ratio of global to extraterrestrial irradiation. a predetermined level. Whillier(1953) has shownthat the hourly and daily distribution curves are very similar to each other. i.6. India and the UK.

20 10 '0 10 20 30 40 50 Figure 5.3 Orientation factors It is reasonable to assume that. The .CQ) >u) ow P (b) Diffuse illuminance.6. klx g C a) 90 80 70 60 Ca0 I • • • Stornoway 1982—94 Manchester 1982—94 Aberporth 1975—94 London 1975—94 -c Oa. 1999) 5.. o 50 40 >. klx 163 0 0) C > C) a) 00 .6.WINDOWS AND DAYLIGHT (a) Global illuminance. even with overcastconditions.2 Cumulative frequencyof (a) globaland (b) diffuse illuminance for UK locations (Kinghorn. windows facing the direction of the sun will receive more light than those facing away.

00 Orientation factor(Littlefair. However. A model that enables the estimation of ILTLT is presented below.4) and dTLT is the ratio ofthe exterior illuminancein theplaneofthe window to theexterior horizontal illuminance under CIE overcast sky conditions. Orientation factors Direction North-facing window East-facingwindow South-facingwindow West-facingwindow Orientation factor(Hunt.77 1. Orientation weighted DF = Orientation factor x Overcast sky DF. 1990) 0.55 1.r is defined as the modifieddaylight factor. it being the ratioofthe internal to external illuminancein the plane of the window with a tilt angle of TLT. where p is the albedo (reflectivity) of the underlying land mass. i.20 1.15 1.164 WINDOWS IN BUILDINGS CIBSE Window Design Guide contains a set of orientation factors (see table below) which should be used when calculating the DF for energy-saving estimates.e. Hunt carried out at the BRE. dTLT= 0.R.4 Slope illuminance In the previous section it was shown that orientation factors may be used to takeaccount of a window's orientation with respect to the sun. Haves and Littlefair's (1988) procedure to obtain internal illuminance is summarised below: IL1 = ILTLT.6.r (5. For an unshaded vertical window.21 5. r = d/dTLT (5.Sp. England duringthe 1970s.396+O.6) where d is the standard CIE daylight factor (Section 5. .G. 1979) 0.97 1.6.5) IL1 is the desired horizontal internalilluminanceandILTLT the external slope illuminance. Since then improvementsin orientation factor have been proposed by Littlefair (1990) and these are included in the table below.6. orientation factors fail to take account of the dynamicvariations in daylight as pointed out by Haves and Littlefair (1988). In that respect orientation factors provide a slightly better assessmentof daylight admittance than calculations based on the simple daylight factor approach. The latter team have introduced a procedure to address the above-mentioned shortcomingsof the daylight factor. This information was based on the work of D.04 1.

Note that bç and bh describe the radiance or luminance distribution of the sky vault thus. The respective tilt factors are obtained from TFh = 0. For any sloped surface of angle /3.8) is the horizontal extraterrestrial illuminance.9). Thus.8) ILE in Eq. (5.6.WINDOWS AND DAYLIGHT 165 5.9) Kinghorn and Muneer have shown that for the respectivecases.6.7) refer to the global and diffuse components.2.4755K Once TF values are obtained.6.UK L .10) (5.the product of extraterrestrial irradiance and its luminous efficacy of 96.11) TFf = 0. values of bf and bh are routinely returned from Eq.6.4. Example 5. i. The subscripts G and = (ILG — ILD)/ILE (5. L0 (1 +bsinO) Lz (l+b) where L0 is the luminance ofa sky patch at an angle 0 from the horizon.4 + 0.6.rB] D in Eq. cos (INC)/ sin(SOLALT)] FIL that r (5.6.8 lmW'.23b)).7) was definedin Chapter 4 (Eq. (4.6.6. TF represents the ratio ofbackground-skyirradiance (or illuminance)on a slope to the horizontal diffuse irradiance (or illuminance): — — TF= ILTLT cos2+ {(32b) [sinfl f3cosfl sin2]} (5.rB+ ILD[(l — FJL)TF+FJL. (5.1 Muneer—Kinghornmodel Thismodel(Kinghornand Muneer.l99lK + 0.1 Five-minute averaged horizontal global and diffuse irradiance and horizontal global and diffuse illuminance measured values are provided for Garston. The ratio TF is the tilt factor.6. (5. the tilt factor varies with clearness index. 1999) is along the lines ofMuneer's (1990) work on solar irradiance modelling. rB = Max[0.0793K (5. ILILT Recall (ILG — ILD). CalcS-04. shaded (sh) and sun-facing (sO components of the sky vault.e.xls enables computation of slope ifluminance based on the Muneer—Kinghorn model. K.0014K — 0.4 — 0. the zenith luminance and b the radiance or luminance distribution index.6.

Insert the given Year and Month in cells B7:B8. Use the Kinghorn and Muneer (1999) model to obtain the diffuse slope illuminanceson east. converted to daylight factor values. Note that validation data and a corresponding linked graph are also included in the Calc5-04workbook.8 * To bediscussedin Section 5. the factors change with the sky clarity. against the orientation-weighted daylight factors. Using Eq.2 3. that the slope illuminance-based computations produce in effect 'dynamic' values of orientation factors.e. The followingtable enables comparison between the present set of results. On 1 April 1992at 0930hours the global and diffuse irradiation on a horizontal surface were recorded as 392 W/m2 and 245 W/m2 respectively. A similar procedure for the east-facing surface will yield the corresponding value as 19. (c) Refer to Eq.2 Hunt (1979) 1.and y-axes in the 'Graph' sheet are respectively the measured and computed global slope illuminance. LONG. and LSM in cells B4:B6.38°W) in Chapter 8.5) and (5. Insert the given site data LAT.7 3.166 WINDOWS IN BUILDINGS (51.6.2 klx and 27.3 klx.and north-facing windows. The diffuse sky illuminance on the north-facing surface is thus obtained as 10.6. Aspect=0° (north orientation).6.5) the internal illuminances are then obtained as 872 lx and 458 lx for the above respective cases. i.6).044. . The x.1 2.6. (5. (b) Activate sheet 'Compute'.2.6 The above comparison shows that even the simple technique oforientationweighteddaylight factor produces reasonable accuracy for the above estimates. Note that d=0. (5. Tilt=90° (vertical surface) and p=0.2.3 klx. (a) Open workbook Calc5-04.Sp.6. Note.6)obtainthe internal illuminancesfor the above two aspects and hence compare the daylight factors obtained with this procedure against those obtained using the orientation factor weighted daylight factor.496 and r=0. Using the procedure highlighted in Eqs (5.5 (north) Orientation-weighted Orientation-weighted (east) Slope illuminancebased (north) Slope illuminancebased (east) Luminance distribution based* 2. 0.71°N. however.022 and dTL-r==O.xls.7 Littlefair (1990) 2. dTLT=O. Comparisonof daylightfactor (DF) expressedas per cent Overcastsky = 2.Thus for p=O.3 1. The corresponding illuminances were recorded as 42.396+O. Insert the given surface data in cells J4:J6.7 klx.

0 52.4 33.7 17.0 11.6 6.7 19.6 13.4 28.8 11.2 34.7 14.0 41.4 13.4 22.3 1 8 30 428 309 515 12.5 13.5 Computed data (kix) ILE IL5 ILw 7.8 46.8 18.0 I Measured data (klx) ILE ILs 1Lw 11.7 10.9 1 10 58.8 7. 1 August 1991 ILN 5.7 11.0 9.4 26.2 63.1 27.2 8.0 14.9 27.8 37.9 66.8 14.0 46.0 .7 5.9 19.9 35.7 57.1 1 12 21.2 31.5 44.4 13.6.8 17.6 12.4 49.1 67.0 11.6 5.4 5.0 30.3 37.9 5.2 12.8 Year (W/m2) ILG (klx) 132 Month Day ILN 505 653 Hour Minute 'G (W/m2) ILD (kix) 58.1 for Garston.6 21.7 4.1 64.8 12.4 10.0 55.6 10.6 1 13 4.8 11.5 10.6 60.1 10.8 17.6 37.3 76.1 Comparison of measured and computed illuminance for Example 5.1 7.9 1 9 26.2 50.7 26.5 29.5 24.5 1 14 270 75.9 1991 1991 1991 1991 1991 1991 1991 1991 1991 98 810 229 443 8 8 8 8 8 8 8 8 8 1 15 16 30 30 30 30 30 30 30 30 15.8 13.3 14.2 10.6.4 11.1 1 11 73.2 12.3 35.Table 5.3 11.6 18.4 1 217 307 286 390 88 324 219 89.1 14.

(1983).13) This model claims to be site and season independent.. The latter team have also demonstrated the site-specific nature of many of the older models.xls may be used to obtain Perez zenith luminance estimates given horizontal diffuse illuminance. (1993). Such luminance values can be determined as a function ofthe corresponding zenith luminance. Calc5-02.5 Zenith luminance for Daylightdesign buildingsrequires knowledge not only ofthe total amount of external illuminance available but also of the distribution of luminance across the sky vault. With an overall bias error around 1% and RMS errors ranging between 0.7 kcd/m2 for clear skies to 1. a deep plan side-lit room which may receive daylight from only a small portion of the sky.6.6. . it represents a significant improvement over the above models. Notable among these are Kittler (1970). Some ofthe above models are not applicable in the tropical belt as they use a tangent function ofthe solar altitude (Rahim et al. Previous studies have shown that the sky vaultcan be considered as a series ofseparate skypatches. (1966) and more recently Littlefair (1994) have shown that under dark. Nagata (1970).5. 5.1 shows a comparison of measured and estimated slope illuminanceusing the Kinghorn and Muneer (1999) model. Under such conditions the zenith luminance (Lx) may be obtained as L = {9/(77t)}ILG 5. Liebelt (1975).6. 1993). A number of researchers have proposed zenith luminance models for clear skies. each with its own individual degree of luminance. The model is represented by L/ID = a1 +b1cosz+c1exp(—3z) +d (5. Once zenith luminance computations are available absolute luminance values of each sky patch can be easily determined. This is of particular interest when considering external illuminanceas a lighting provider in. overcast skies the Moon and Spencer (1942) proposed CIE overcast-sky model shows good agreement with measurements. Sky luminance distribution data are of great importance to the lighting designer considering daylighting as an option for the provision of internal illuminance.6.5 kcd/m2 for intermediate skies.12) Perez et al. Dogniaux (1979).168 WINDOWS IN BUILDINGS Table 5.1 All-skyzenith luminance models (5. Nakamuraet al. say.6. Krochmann (1970). Hopkinson et al. (1990)have developed an all-skyzenithluminancemodel which has been tested against extensive data from five locations in North America. Karayel et al. (1985) and Rahim et a!.

xls cannot be opened by double-clicking it from the Windows Explorer. (a) Open workbooks Calc5-02. ILD and sheet 'Compute'. Copy the horizontal global anddiffuse illuminance data (cells C75:D86)from Data5-Ola and pasteit in cells P6:Q17 ofsheet 'Compute'. (1).xls. L . UK (51.2 Five-minute averaged horizontal global and diffuse irradiance. or simultaneously press ALT+ R • • keys (b) Activate sheet 'Site' of Calc5-02 and insert the given LAT. Notice that Calc502. Use the Perez et al. Table 5. cells B6:B17 with month (8) and cells C6:C17with day (d) Choose an hour's data from Data5-Ola for the purpose of calculations. fill cells A6:A17 with year (1991). Therefore.WINDOWS AND DAYLIGHT 169 Example 5. LSM in cells A2:C2 as shown. Copy the zenith luminancedata (cellsQ75:Q86) from Data5Ola and paste it in cells R6:R17 of sheet 'Compute'. For all DLL-based workbooks the following procedure MUST be used: • Launch Excel software • Open the DLL-based workbook • Enter data in the relevant worksheet(s) • Simultaneously press ALT+ F8 keys Select the relevant macro from the dialog box Either click the Run button. manually.xlsand Data5-Ola.xls includes a FORTRAN-based DLL program [Fr309b. LONG.2 shows the results. (1990) procedure to compute horizontal global and diffuse illuminance and zenith luminance for 1 August 1991.38°W) in Data8-02.g.dll]. Calc5-02. for the hour beginning 1100. Excel should be launched first and then Calc5-02. horizontal global and diffuse illuminance and zenith luminance data are provided for Garston. Copy the Hour andMinutedata (cells A75:B86)from Data5-Ola and paste it in cells D6:E17 of sheet 'Compute' of Calc5-02.6.6. Copy the averaged horizontal global and diffuse irradiance data (cells 175:J86) from Data5-Ola and paste it in cells F6:G17 of sheet 'Compute'. Compare the estimates against the respective measurements.71°N. (c) Activate sheet 'Compute' of Calc5-02 and. e. are shown in cells L6:N17 of (e) The computed values of ILG. Sheet 'Graph' of Calc502 shows a plot of the computed parameters against the corresponding measured values.xls. 0.xls opened within Excel (see box below).

4 13.0 11.6 55.2 8.9 11.3 12.0 33.8 28.5 11.1 43.8 31.2—3 for Garston.5 26.9 36.1 27.7 77.1 40.0 36.4 34.Table 5.8 36.7 47.3 65.5 34.3 10.3 14.8 14.2 Input/output data for Examples 5.8 67.9 23.7 9.8 14.7 29.8 35.0 40.3 46.1 39.6 .6.4 35.5 16.5 38.5 39.3 67.6 43.2 11.3 37.9 13.7 31.8 10.8 12.0 37.8 24.1 24.8 33.8 37.6.3 11.4 25.6 57.4 9.3 28.9 34.5 26.2 8 8 8 8 8 8 8 8 8 1 II 40 8 1 11 8 1 11 45 50 500 455 364 277 205 228 309 703 592 295 350 219 286 373 283 258 299 336 16.8 8.0 46.4 8. 1 August 1991 Perez Year IL0 (klx) 16.1 34.2 53.0 25.7 12.7 30.8 41.3 44.9 Month Day IL0 (kix) Hour Minute 'G (W/m2) Ij (W/m) Measured data IL0 (klx) L (kcd/m2) L (kcd/m2) 13.7 10.1 13.1 47.9 29.5 48.0 37.6 9.1 41.8 1991 1991 1 1 Ii 355 319 307 249 193 0 11 5 1991 1991 1 1 1 11 11 10 15 11 1 1 11 13.9 11.5 43.1 42.4 14.1 11.5 10.6 44.6 1991 1991 1991 1991 1991 1991 1991 1991 8 1 11 55 400 59.7 42.5 13.3 37.8 51.5 33.8 27.6 35.6 79.8 33.6 76.0 43.6 9.9 23.3 35.1 Muneer—Kinghorn IL0 (klx) IL0 (kix) IL0 (kix) L (kcd/m2) 12.4 1 11 11 20 25 30 35 12.6 36.9 37.5 37.0 27.

xlsand Data5-Ola.xls shows a comparison between the computed values and their corresponding measured values. provides an indication of the relative brightness ofthe skyvault.xls. and Lz are shown in cells Q6:S17of (e) The computed values of ILG.6.2 using the Muneer—Kinghorn(1997) and Kinghorn— Muneer (1999) models. AZIMUTHand SLOPE in cells A2:E2 as shown.6.23 to —0. ILD and locations are provided in Data5-Olb and Data5-02 respectively.6. Example 5.62 when the sky condition changes from overcast to clear skies. may be representedby a unique value ofb.2 and 5.5.6. or luminance. L 5. distribution of each half of the sky-vault.2 presents the results. 'sun-facing' (by the sun) and 'shaded'(opposite sun).9)—(5. Zenith luminance. 14) Note that to enable zenith luminance computations. One such modelling and visualisation package is the . is then calculated in the followingmanner as a function of these indices and the corresponding diffuse illuminance of the sky vault: L l.6.2 Muneer—Kinghornmodel As stated above the luminous distribution index. (d) Repeat step (d) from the above example. e.6. LSM. L.l1)). In this case.3 Repeat Example 5. sheet 'Compute' of CalcS-03.xls. (a) Open workbooks Calc5-03. values of bh and bf have to be determined first (see Eqs (5.3 to obtain for Garston or Fukuoka. Notice that the horizontal global anddiffuse illuminancedata should be pasted in cells U6:V1 7 of sheet 'Compute'and the zenith luminancedata in cells W6:W17 of sheet 'Compute'. Japan. IL The reader may wish to try the above Examples 5.6.5ILD (1+bsh'\ + (1+bsf + + 2b 3 2bS) 3 (5. Table 5. (b) Activate sheet 'Site' of CalcS-03 and insert the given LAT.g. these indices are denoted by bf and bsh respectively.6 Sky luminance distribution On the daylighting front recent advances in computer graphics technology allow the realistic modelling of complex building interiors with a minimal training time. b.WINDOWS AND DAYLIGHT 171 5. LONG.6.6. (c) Repeat step (c) from the above example.6. Sheet 'Graph' of CalcS-03. Steven(1977)has shown b to vary from 1. Measured data for both IL0. It has been shown by Steven (1977) and Muneer (1987) that the radiance.

dand e. copyrighted UNIX based software produced at the Lawrence Berkeley Laboratory in California. Skylight is a non-uniform extended light source. c. The RADIANCE system is free.6. it is essential to be able to estimate skylight distribution from routine measurements such as irradiance. a building energy simulation expert at DeMontfort. In the following section luminance distributions for all-sky conditions are presented. defined as the ratio between the sky luminance at the considered point and the luminance of an arbitrary reference point is given by It is a physically based program which allows precise estimation of in- L l =f(O. sky-luminanceangular distribution is the necessaryinformation required for calculating daylight penetration into any properly described environment. Perez et al. DeMontfort University in Leicester. luminance distributions.6.16a) . (1993) have presented an all-sky model which is a generalisation of the CIE standardclear-sky formula (CIE. for (5. x = x1 + x2z + (x3 + x4z). 1973). (e) It can estimate daylight factors using real sky luminance distributions.172 WINDOWS IN BUILDINGS RADIANCE lighting simulationsystem(Ward. (b) It has the capability to model geometries with realistic. (c) It supports a wide variety of reflection and transmission models.15) For x=a. ) = [1 + aexp(b/cosO)][1 + cexp(d) + ecos bins (skyclearness) 2—8 (5. Mardaljevic (1995) has undertaken the first comparison of RADIANCE results to actual room illuminances under a real sky. illuminance predictions were found to agree quite favourably with the internal illuminances. Its intensity and spatial distribution vary as a function of prevailing sky conditions.undertaken by the Building Research Establishment at Watford. 1994). In addition to direct sunlight. This expression includes five coefficients that can be adjusted to account for luminance distributions ranging from totally overcast to clear skies.. England is shown in Plate 5. rather than assuming the worst case scenario used by the CIE overcast sky model. (d) It can import building scene descriptions from CAD systems. The main features of RADIANCE have been enumerated by Mardaljevic (1995) as follows: (a) terior illuminance. Using 700 scans of measured sky luminance measurements. The relative luminance li. Details of acquiring this software have been given by Mardaljevic (1995).1 (see colour plate section). Because actual sky-luminance distribution data are available only in a handful of locations.A sample RADIANCE output of a simulation of the Queen's Building. b. This plot has been provided by John Mardaljevic.

(1993). (5. Details ofthe sky elementgrid adopted for these computations are shown in Figure 5.xls.6. (b) Highlight and copy cell range D705:ER705.Japan.xls and Calc5-06.3. however. enable computation of the sky luminance distribution. In the relative co-ordinate scheme the azimuth of any given sky patch is its angular separation from the sun. (h) Copy cell range C2:C146from step 'f' into cell range B4:B148 of sheet 'Sky_Lum data' of the CalcS-05.xlsworkbook. (d) Paste the above data in the cell range C2:C146ofsheet 'Measured data'.16b) (5. For further details the reader is. in Calc5-05. Sort command. which respectivelyuse the absolute and relative co-ordinate scheme for the sky patch.WINDOWS AND DAYLIGHT 173 otherwise c + c2z1)} — I d = exp{t4di -i. Click OK to execute.e.6. the absolute framework. Use data for the skyscan for 6 June 1994 at 12 noon LCT to evaluate the Perez all-sky luminance distribution model (Eq. has been normalised against zenith luminance. are presented.xls the dimensionlessparameter 1.. The other co-ordinate scheme. Example 5. (c) Launch the Calc8-O1. (f) Copy cell range C2:C146 to the corresponding column of the 'Sorted data' worksheet. referred to the article by Perez et al.xlsto compute the above luminancedistribution and compare it against measured dataset. (e) Highlight cell range A2:C146 and then use the Data.6.6. . The dialog box 'Sort' appears.xls workbook. Presently.xls measured luminance distribution data for Fukuoka. i. Use Calc5-05. allowsthe user to obtain the luminancedistribution ofthe sky with the north direction being the reference.d2zfl ± d3 ± Ad4 exp{(Afci (5. b. A listing of these coefficientshas been provided by Muneer (1997).xls workbook.xls and Calc5-06.xls . Make sure that the 'Sort by' and 'Then by' boxes respectively contain 'Altitude' and 'Azimuth' and the corresponding 'Ascending' radio buttons are selected. Calc5-05. (g) Launch the Calc5-05.6. Thefollowingprocedure is to be used: (a) Launch the Calc8-07.xlsworkbook. c.4 In the electronic file Data8-07.16c) The coefficients a.15)refers). d and e are adjustable functions of irradiance conditions. Note: The measured data for Fukuokais transformed form due south to a due north referenceusingthe workbook Calc8-O1.

3 Sky scan recording map for (a) Garstonand Sheffield(PRC Krochmann scanner— relative co-ordinate system) and (b) Fukuoka (EKO scanner— absolute co-ordinate system) .174 WINDOWS IN BUILDINGS (b) Figure 5.6.

it is possible to plot several hundred thousand to a few million data points. patch and a close similarity betweenthe measured and computed values.9 and 0. time and solar data in cell range B5:Bl8 of the 'Main' worksheet.txt. (1999) and plots similar to those present in Figure 5. A comparison of measured and computed sky scan plots is demonstrated in Figure 5. Plots such as the one under discussion may be obtained simply by double-clicking the LD Vals.and triple-glazings the BRE Digest (Building Research Establishment. The data required by the above exe file are to be provided in the LD_Indat. double.exe produces a scheme of the UK sky scan system.4 included in the latter reference. The CIE sky component table.8 to . (1966) have presented daylight transmission curvesfor single-. This versatile tool enables quick graphical review of very largedata files.exe file which is also available in the Exe_files folder.WINDOWS AND DA YLIGHT 175 (i) Provide the site.7 Luminance transmission Hopkinson et al. Further validation workon luminancedistribution has been undertaken by Chain et al.csv file. A sample is provided in the 'Exe_files'folder.exe' for x-y plots is also included.csv.exe file which is available in the Exe_files folder. Note that the file LD_Numbr. presented in Chapter 3. Note that within CalcS-05 and Calc5-06.6.xls a doughnut diagram is also provided. 1986) recommendsrespectivecorrection factors of0. These plots enable a better visual comparison between the measured and computed datasets. 5. (j) The final computed data is now available in cell range E4:El48 of 'SkyLumdata' sheet. This file draws data from the file Data XY. such as those encountered in solar energy and daylight research. e.6.csv file. The plot shows numerical values of the luminance of each sky Plates 5. For double. incorporates transmission loss for a singleglazing.g. The file LD Vals.4.and triple-glazingsofclear float glass design.exe (seethe latter folder) may then be used (by double-clickingthe file name icon from Windows Explorer) to produce plots of the sky scan. This diagram.2 and 5. A directly executable file 'Plot_XY. Colour plots such as these may be obtained by using the LD_Fill. The data required are once again provided via the LD_Indat. The inner sectors represent the measured and the outer ones the computed luminance data. Once again the Perez model seems to produce reasonable estimates of the sky-luminance distribution. included within the sheet named 'Graph' enables a coarse method of validation. (k) The measured andcomputed columns of measured and computed data are then saved in LD_Indat.3 (see colour plate section) show colour luminance distribution plots for Garston (UK) data.

4 Comparison of(a) measured and (b) computed luminance sky distribution for Fukuoka. 12 hours LCT .176 WINDOWS IN BUILDINGS (a) (b) Figure 5. 6 June 1994.6.

7.89. This is demonstrated via Example 5. it is possible to investigate the fine detail ofdaylight penetration through multiple glazings without any loss of be multiplied accuracy by combining the above luminosity estimates with the glazing transmission characteristics.88) Using a similar procedure the incidence angle for SP2 is obtained as 40. Also obtain the respective angular transmittances for a single-glazingassuming normal incidence transmittance. This is certainly a simple but approximate approach. Compute the incidence angle between a normal to the window and the luminance emanating from these sky patches. the angular separation of the sky patch and a normal to the window glazing is required.4.xls enables such computations to be performed quite easily.xlsmust NOT be saved after the above execution since the formulae for the sun's position are deleted by the entry of the above numerical data. Example 5.This material is presented in the followingsections. Refer to Figure 5.7. The transmittance is provided in cell N13 ( 0.WINDOWS AND DAYLIGHT 177 to the total daylight factor. Caution: File Calc5-05.1. SF3 and SP4 are then obtained from consideration of symmetry.7. The angle between the above two points is then read off from cell J13 as 42. In this particular case the geometry for a normal to thewindow maybe usedinstead of sun's position to obtain the desired solution.3° and transmittance as = 0. A step-by-step procedure using CalcS-05.9. . 5.1 Incidence angle of luminance from a given sky patch and transmission through glazing In order to obtain the transmission of luminance from any given sky patch discussed above.1 Refer to Figure 5. SP1—SP4are shown.2 wherein four sky patches. = 0. 80. Note that the latter are the altitude and azimuth co-ordinates of the mid-point of SF1.xls is provided herein: Enter t = 0.9 in cell B25.3°. In the present context. CalcS-05.1 which providescomputational details for any given sky patch and its angular separation from the sun's position. Enter altitude and azimuth of the secondpoint The altitude and azimuth ofthe normal to the window is provided in cells D8 and E8 as 0. Enter altitude and azimuth of thefirst point The altitude and azimuth (measured from true north) of the centre of SP1 are to be provided in cells D13 and E13 as 40. 95.7.

7. UK were provided.15)) .1 wherein illuminance data for Garston.178 WINDOWS IN BUILDINGS North (b) N = north P = sky patch S = position ofsun in sky Z = zenithofsky vault zZOS = solarzenithangle LAOS = solar altitude angle LNOA = solarazimuth angle LNOB = sky patchazimuth angle LBOP = sky patch altitude angle ZPOS = elongation anglebetween P and S Figure 5. (5.2 Refer to Example 5.7.6. 1997) Example 5. Using Perez all-sky luminance distribution model (Eq.1 (a) Geometry of a given sky patch (SP) and (b) skypatch and solargeometry(Muneer.6.

90 3. The RADIANCE software package undertakes this type of analysis for representation of the illuminance environment within buildings. The total internalilluminance due to sky may thenbe obtained by summing up the contributions of all relevant sky patches. Using Calc5-02.2.7. A derivation for estimating the illurninance on a horizontal surface due to the given sky patch is now presented: obtain Area of sky patch (r dz)(r di) where dx and determine the size ofthe sky patchand r is the radialdistance of the patch.30°andan average of the computations.2) t of 0. Let E(IL) represent the internal horizontal illuminanceat the referencepoint P due to the given sky patch. Using the above fixed values of ir/l8 for dx and dii.8%. used in CalcS-05.1.xls 1. SP.computedis obtained as 5865 cd/rn2.1(a) shows a sketch of the sky hemisphere and the geometry for any given sky patch.72 L= L computed x (L/L) 5. It may easily be shown that L(IL) = zL(rdcx)(rdJi)sin/r2 (5.19 2.7.95 Total internalilluminance = Note that this example indicates a daylight factor of 3.xls. Figure 5.6.7. In CalcS-05.885 may be assumed for the above four sky patches.xls LZ. Of course the present procedure provides the most precise values of internal illuminances as it takes into account the detailed luminance distribution of the complete sky canopy. (5. As mentioned above.1) d where L is the luminance of the sky patch in cd/rn2.4.WINDOWS AND DAYLIGHT 179 the internalillurninancefor the reference point shown in Figure 5.4. . Sky patch (Figure The followingtable summarisesthe rest L/Lz Calc5-05.2) 10 090 12840 17 010 23170 A(IL) 179 228 302 411 1120 SP1 SP2 SP3 5P4 2. Lx(IL) = tL(ir/18)2sin (5. Use data obtained in the above example.7. Basedon Example 5.an average incidenceangle of41. A comparison with DF obtained from other simpler procedures was presented in Example 5.xls these values may be reset at 10° (ir/18 radians) intervals.2) cd/rn2 5865 5865 5865 5865 lx Eq.7.1.

Note that Eq. (5.9 Innovative developments A high densityofurban structures canoften leadto theloss ofdaylight amenity. Daylight is gathered via a polycarbonate dome at roof level and then transmitted . Equation (5. Pay is the average room reflectance. have patented a 'Sunpipe' system which can pipe daylight to internal spaces with the added advantage that excessive solar gains are avoided.9. A simple formulation for the average daylight factor has been presented in Pritchard (1995): DFav(%) =AwOT/{2AT(l Pav)I (5. an English company specialising in building services products. 0 is the angle subtended by the skyat thecentre of the window and is the glazing transmittance.8 Optimisation of glazing area Wilkinson (1992) has undertaken a study of the window area and the interaction of the costs of electric lighting and the penalty paid in increased heating costs. The mirrored light transmission system is crudely analogous to the fibre opticcable.8.1. the difference being that in the latter device the transmission efficiency is much higher because the signal is carried throughby means of a multiple refraction effect. One innovative solution that has emerged in recent years is to capture daylight via a mirrored lightpipe and then direct it to those areas of buildingwhich are starved ofdaylight. A schematic of the Monodraught system is shown in Figure 5. Using Eq.8. The c1 and c2 constants will respectively include the window daylight and thermal transmission characteristics.8.8. Improper designof large glazed facadesmay also cause the problem ofglare as has been pointed out above in this chapter. 5. (5.1) a relationship for the total energy expenditure on electric lighting and heating was presented by Wilkinson (1992): A t Total energy costs = [c1/A] + [c2A] (5.8. Monodraught.2)may be differentiated against the window area to provide an expression for an optimum value of A.1) is very approximate and should be handled with care. The constant c1 will additionally include information related to the illuminance requirements for any given building application and the design data for the top-up electric lighting.180 WINDOWS IN BUILDINGS 5.1) where and ATare respectivelythe window andtotal room surface area.2) The first term on the right-and side of the equation represents the costs associated with top-up electric lighting (inversely proportional to the window area) while the second term reflects the heating costs which linearly increase with glazing area.

some with four pipe bends. 1998). The internal surface of the sunpipe is coatedwith a highly reflective mirror finish material (typicallywith a reflectancein excess of 0. 0 the angle between the incident downwards . In another innovation the sunpipe has been combined with a natural ventilation system which draws air from the rooftop opening to provide cooling for the building fabric. Measurements on a sunpipe system undertaken at Liverpool and Nottingham Universitieshaveindicated that under hazy sunshine conditions a lighting level of 900 lx was recorded internally at approximately 2 m below the outlet lightdiffuser. The total number of buildings within the UK with mirrored lightpipes now exceeds 750. Shao et al. (1998) have presented simple design formulae for the design of mirror lightpipes. One such formula relates the transmittance ('r) of the lightpipe to p. Such systemsare described in further detail in Chapter 4. In typical installations sunpipes ofup to 6 m in length have been used. A further innovation in the development of the sunpipe system is its integration with electric lighting within the sunpipe so that the devicecontinues to function as a light source even during the hours of darkness (Payne.WINDOWS AND DAYLIGHT Hemispherical. lightpipes are a relatively new concept in Europe.95)which helps in achievinga reasonable illuminance indoors when daylight is introduced via a light diffuser. the pipe surface reflectance.1 A schematicof the Monodraught Iightpipe system to interior spaces within buildings.9. perspexdome 181 Figure 5. The building types have ranged from insurance companies to health clinics and from residential to university buildings. Monodraught sunpipe systems have successfully been installed in buildings across the UK. However. The light-reflecting tube is adaptable to incorporate any bends around building structural components. Although commonly used in the USA and Australia.

(1975) Windows and people: a literature survey — psychological reaction to environments with and without windows. New York. A. and Fontoynont. UK. 69—74. K. Van NostrandReinhold. B. Cent. R. London. Bátim. Boubekri. Cawthorne. provided an excessive number of bends or long pipe lengths are avoided. Watford. Chain. LR&T 24. Cah. Paris. Collins. SBI report 258. D. Fanchiotti. (1999)A comparative model of luminance. and Steemers. (1998). N. Solar Energy 65. Sci. M. NBS Building Sience Series 70. Thebroad conclusions that were drawn were that lightpipes can be effective devices for introducing daylight into buildings. Lightingandthe Biological Clock.L. International Conference on Illumination. (1954) Five years continuous recordings of total and diffuse solar radiation at Kew Observatory. 24. Res. the length (L) of the lightpipe. M. (1991)Buildings. 285—295. DuMortier. Dogniaux. correlated colour temperature and spectral distribution of skylight: Comparison with experimental data. Publication 895. (1995) Daylight Utilisation in Office Buildings. Building Research Establishment. The Martin Centre for Architectural and Urban Studies. CIE (Commission Internationale de l'Eclairage) (1973) Standardisation of luminous distribution on clear skies.D. References Ander. M. University of Cambridge. the effective diameter of the pipe (5. Chartered Institute of Building Services Engineers. (1992) Effect of window size and sunlight presence on glare.J. (1960) Données météorologiques concernant l'ensoleillementet l'éclairage naturel. London. (1995) Daylighting. PhD thesis. Metorological Office. Baker. (1993) Daylighting in Architecture. C. James and James Ltd. . Estimating daylight in buildings: Part 2. Building Research Establishment (1986) BRE Digest. CIBSE (1996) Building Services SupplementMay 1996. Danish Building Research Institute (1995). Christoffersen. G.L.9. and Boyer. Blackwell.. D.182 WINDOWS IN BUILDINGS light and the pipe axis. L..1) where x= L tan 0 deff Field monitoring of buildings installed with sunpipes was carried out by Shaoet al.C. and deff. Use of as large a pipe diameter as possible was also recommended. Met. London. 44.

P. D. (1988) Daylight in dynamic thermal modelling programs: Case study. (1979) Improved daylight data for predicting energy savings from photoelectric controls. Navvab. BSER&T 11. Biological Psychiatry 35. Architectural Press. Kinghorn. T. J. Illum. BSER&T9. R. Petherbridge. and Selkowitz. Gesundsheitsingenieur 96. Trans. Muneer. Met. J. (1975) Leuchtdichte. (1994) Low illumination experienced by San Diego adults: association with atypical depressive symptoms. LR&T 11. (1970) Some considerations concerning the zenith luminance ofthe cloudless sky. United Kingdom. 3. R. Solar Energy 53. T. Energy Management.Sept.. Muneer. Soc.J. Daylight and skylight at Pretoria: The luminous efficacy of daylight. Haves. (1987) Solar Radiation Modellingfor the United Kingdom. Mardaljevic. 127. R. P. 149. 551. 11. P. Espiritu. E. D. and Littlefair. T. M. and Davidson. Circular No. and Longmore. London.WINDOWS AND DAYLIGHT 183 Dogniaux.. Building and Environment 25. Hunt. 139—150. Series B 9. C. London. (1966) Daylighting. D. Arch. S. T. Muneer. LR&T (to be published).J. P. (1994) A comparison of sky luminance models with measured data from Garston. and Muneer. Energy and Buildings 6.J.G. (1990) Predicted annual lighting use in daylitbuildings. 16—18.2. 183. LichttechnikB22. P.E. P. (1990) Solar radiation model for Europe. Liebelt. (1970) Uber die horizontal beleuchtungs-starke und die zenitleuchtdichtedes klaren himmels. Drummond. 9—23. and Spencer./Oct. P. Littlefair. Karayel. LR&T 27. pp 43—54 Littlefair.und strahldichteverteilungdurch tageslicht. CIE E-3. (1998) Daylight illuminance frequency distribution: Review of computational techniques and new data for UK locations. Krochmann. Wien. (1979) Variations qualitatives et quantitatives de composante du rayonment solaire sur une surface horizontale par ciel serein en fonction du trouble atmosphérique. M. Moon. Heinemann. PhD thesis. Bruxelles. CNAA. Oxford. Kittler. (1997) Solar Radiation and Daylight Models. Hopkinson. D. Kinghorn. (1996) A holistic approach to lighting design. 181. Publication IRM. NY 37.J. Engng. (1983) Zenith luminance and sky luminance distribution for daylighting applications. (1999) All-skyzenithluminance models for the UK. (1995) Validation of a lighting simulation program under real sky conditions. R.R.C. Loe. 153. 315. and Muneer..G. 403—407. Neeman.J. . A. LR&T 30. 707. D. (1942) Illumination from a non-uniform sky. T. (1958) Notes on the measurement of naturalillumination II. Serie b(62).

Pleijel. Computer GraphicsProceedings. Ward. D.. A. Whillier. Shao.R. T. 103. A. Pritchard. and Angus. Koga. R. LR&T 29. PhD thesis. J. Eng.M. (1990) Modellingdaylight availability and irradiance components from direct and global irradiance.P. (1998) Mirror lightpipes: Daylighting performance in real buildings. M. Proceedingsof the 18th CIB conference.J. 457. Trans. Y. 185.D. Ilium. Napier University. 41.184 WINDOWS IN BUILDINGS Muneer. (1977) Standard distribution of clear sky radiance.. (1998) A brighter future for natural daylight. Steven. Nakamura. and Michaisky. Payne. P. (1953) Solar Energy Collection and its Utilisation for House Heating. (1993) Modelling skylight angular luminancedistribution from routine irradiance measurements. M. (1985) Luminance distribution of intermediate sky. 10. G. Proc. D. Oki.. L. M. Q. Winter. (1992)The effect of glazing uponenergy consumption within buildings. Wyon. 234—239. PhD thesis. MIT. Cambridge. R. ProceedingsofLux Europa. 6. Roy. T. and Hayashi. 271. LR&T 30.A. G. Addison Wesley Longman Limited. T. Bangkok. Solar Energy44. (1994)The RADIANCE lighting simulation and rendering system. (1993) The modified equation for the zenith luminance of the clear sky. (1997) Energy savingand the perception of visual comfort. (1995) Lighting.. and Seals R. Rahim. Soc..M. I. Building Resources Worldwide. shops and colleges in Sweden. G. H. Muneer. R. .A. Zeguers. Met. 37. Seals. and Matsuzawa. (1970) Luminance distribution of clear skies. Oxford. (1954) The computation of natural radiation in architecture and town planning. Architectural Press. J. (1980) Human experience of windowless environments in factories. LR&T 27. Wilkinson. offices. Meddelande Bull. Perez. Part 2 — Theoretical considerations. Japan 186. LR&T 24. and Jacobs. Edinburgh. 1.D. Nagata. Annual Conference Series. 30. Inst.. M. T. (1993) Daylight illuminancemodels for the United Kingdom. J. Tridonic Lighting Components. Basingstoke. TLC (1998) Luxcontrol. T. Y. Nakamura. Elmualim. Light & Visual Environment9.. and Yohannes. (1998) Life Cycle Assessment ofMulti-glazed Windows. Arch. Perez. Weir. 2nd Lux Pacjfica Conf.C. D. Stockholm 25. and Kinghorn. Statens Namnd for Byggnadsforskning. MA. Soc. R. (1997) Luminous efficacy of solar irradiance: Improved models. 99. and Nilsson.J. J. Pilkington (1993) Glass in Building. 10—13 November 1993.. Harlow. Ineichen. H.J. M&E Design November 1998. 223. M.

6 ACOUSTIC PROPERTIES OF WINDOWS Ifthe well-beingof building occupants is notat the forefront of designcriteria. including hearing damage. It may originate from transportation. If the definition of . 6. though well designed in terms of thermalperformance and daylight transmission as discussed throughout Chapters 3—5. This chapter will consider issues of aural comfort. either immediatelyor directly. • Receiver variables focus upon tasks and activitiesbeing performed by the occupants. Noise pollution manifests itself in many ways. For this reason it is important to have correct acoustic properties of windows. Thus it may be deduced that • Noise source variables concern the nature. the building. may be described as pollution. will still be deemed unsatisfactoryby its occupants: work performance is also heavily influencedby the acoustics of the working environment. and Window the type of glass used.1 Aural comfort Attempts to apply consistent comfort principles ofthermal comfort analysis to building acoustics have been explored in the recent past. the effect they have upon health and productivity of building occupants.Noise is unwanted sound. composition and origin of the noise. Sound is a form ofwave energy. Annoyance very often occurs because the interference experiencedby an individual is not the result of an action he or she benefitsfrom. This evokes a need for corrective action. industry. community life. and as such. Windows play an important role in attenuating exterior noise to an acceptable level for building occupancy. the receiver (the occupant) and the transmission medium (the window). • building'svariables include the number and thicknessof glazingpanes. exterior noise sources which impact upon the interior environment. and how windows can be designed to act as appropriatefilters to attenuate noise. communication disruption and the creation of unnecessary stress. Achieving this is a function of three variables: the noise source. and other sources.

1.1. 11% of the subjects were bothered by road traffic noise.1 Road trafficnoise Roadtraffic noise generallylow frequency.186 WINDOWS IN BUILDINGS thermal comfort is that condition where neither a warmer.and is a function ofvehicletype.1 illustrates sound pressure and level for typical locations. Recommended maximum LA0q Level dB(A) 30—40 40—45 40—45 45—50 Bedrooms Living rooms Offices Private/smallconference rooms Large offices Educational Classrooms(15—35 people) Classrooms(>35 people) Music/drama spaces Health and Welfare General wards Smallconsulting rooms Diagnosisrooms 40 35 30 55 50 45 .a survey carried out in England and Wales on 14 000 households between 1985 and 1987 showed that 14% of the adult population suffer from domestic noise annoyance. Complaints about noise in general have increased over the past few decades.2 Noise sources Most noise consists of a wide spectrum of frequencies. is road surface. 6. and 7% by aircraft noise. Some typical road traffic noise levels are Table 6.1. nor cooler environment is desired. then acoustic comfort can be paralleled as that condition where neither too quiet nor too noisy an environment exists. frequency levels and strengths must be identified.One of the dominant problems is traffic noise. The increase in complaints received is a combination of several factors.1 shows interior noise targets for some common building uses. Increased public awareness about entitlement to quieter environmentsand a wider knowledge ofthecomplaint procedure are partlyresponsible. 6. topography.1 Interior noise targets Location Dwellings a role. and speed.but increases in general environmental noise levels also play There are thresholds ofbackground noise level which should not be exceeded in order to establish conditions of comfort. For example. while Figure 6. Table 6. Before action can be taken to minimise the effect of noise.2.

Road traffic noise decreases by approximately 3 dB(A) with each doubling of the distance away from the road.1.1.1 — — 70 60 0.1 Soundpressureand level shown in Table 6.00002 0 Threshold ofhearing _______________________________________________ Figure 6.2.01 — 50 Average traffic(kerb) Conversational speech Typical business office 40 0.0001 — — 10 0. .ACOUSTIC PROPERTIES OF WINDOWS Sound 187 Pressure (N/rn2) Sound level (dB) 140 100_ — 130 Threshold ofpain Pneumatic drill 120 10 110 Loudcar horn (1 metre) Pop group(20metres) Insidetube train 90 100 Insidebus 80 0.001 — Livingroom(suburbs) Library Bedroom at night Broadcasting studio 30 20 0.

in accordance with International Standard ISO 1996—1 (ISO.2. averagespeed 62 mph.2 Rail trafficnoise a similar spectrum of frequency to road traffic electrictrains have more emphasis towards middle frequencies. and is generallyplotted in contours around runways. grassed terrain 20 m from busy main road. NNI has since been replaced with LAeq (equivalent continuous A-weighted sound pressure level in decibels). This is a composite ofthe number of aircraft movementsand the peak noises which they generate. The noise levels associated with high-speed. 1995) Distance from track Open grasslandterrain 25 50 100 200 Noise level dB(A) LAeq 67 64 59 54 .2. and landing with high-frequency reverse thrust noise. noise. In the past. paved 20 m from busy main road.3 Aircraft noise Aircraft noiseis a function of altitude. diesel passenger trains and freight trains. although with higher frequenciesdecaying more quickly. averaging 120 trains per day are shown in Table 6. load and type of aircraft. 1996a).2.1 Roadtraffic noise levels (Pilkington. with takeoff noise dominated by low frequencies. Table 6. Table 6.188 WINDOWS IN BUILDINGS Table 6. average speed 31 mph. Railway locomotive noise has 6.2 Rail traffic noise levels (Pilkington.3 shows the probable community annoyance due to aircraft noise at varying levels of NNI and LAeq. averagespeed 31 mph. because it is more predictable and noise disturbance is shorter in period than the constancy ofroad traffic noise. noise in areas surrounding airportshas been assessed using the Noise and Number Index (NNI).2. 1995) Noise 1evel dB(A) Situation 20 m from busymotorway. despite higher noise levels.2. The expressionofrail noise is normally in terms of a 24-hour period since most traffic is regulated and timetabled. Rail noise annoyance is generally more accepted than road traffic noise.2. weather conditions. screened by houses LAb LAq 77 67 57 80 70 60 6.2.

LAb (traffic noise level measured with standard 'A' scale weighting).1 Speech The range of frequencycovered by adult speech lies between 500 and 2000 Hz.. This difference has psycho-acoustic implications. LAeq. The problem ofmaking definite conclusionsfrom the surveys is that different types ofnoise have been assessed by different parameters.3.3 Theacousticenvironment One problem experiencedin assessingthe exposureof an office worker to noise is that internally generated noise levels can often exceedthose of external noise sources. and different measurement locations. compared with external noise where no control is available. A study of all office occupants revealed that 48% of workers regarded the internal noise level as unsatisfactory. and to what level. Seventy per cent of workers seated within 7 m of the noise-affectedwindow revealed the external noise disturbance unacceptable. Thebuilding occupant has a greater element of control over noise generated internally. especiallyif the noise source is ofa singletone or is impulsivein nature. external annoyance may remain the focus of complaint. 1997) recommends a value of 5 dB(A) although this is regarded by some as inaccurate. 6. particularly where the noise is impulsive or tonal (Wilson et al. The background noise level in the office during unoccupied periods averaged 45—50 dB. NNI. A particular noise can cause annoyance at a level as much as 10 dB below background levels.ACOUSTICPROPERTIES OF WINDOWS Table 6. The type of questionnaire used is also highly variable. as with an alarm or construction noise. 1995) 189 NNI 35 LAeq (12 hour) Probable community annoyance Low Moderate High Very high 45 55 60 57 66 75 80 +/—4 +/—4 +/—4 +/—4 Mostconclusionsconcerning the ways in which people are affected by noise have been drawn from the results of subjective questionnaire-based surveys which are used to assess the numberof people affected by traffic noise. as annoyance levels are so subjectivein nature. 1993).2. BS 4142 (British Standards Institution. The traditionalway ofaccounting for tonalor impulsivenoise was to add a penalty value to the measured dB(A).3 Aircraft noise levels(Pilldngton. Baker (1993) found that noise experiencedby workers sittingjust2 m from a window wall affected by traffic noise averaged 60 dB. Suppression of higher frequencies is crucial in maintaining privacy of con- . 6. Despite this.

In the 1960s this was altered to three bands with centres: 500. while Table6.3.2 Sound level/distance relationship for speech intelligibility (Adams and McManus.2 60 65 70 0.2 1. generallyusedfor machinery and services noise). Table 6. Prevailing values much in excess of those listed will render normal speech unintelligible. in terms of background noise level in dB(A).0 4.3. since aural intelligibility relies on these most.1 Suggestedmaximum PSIL values for effective communication Room type Smallprivate office Conference room Concert hall Bedroom Livingroom Classroom/lecturetheatre Max.1 are suggested for good speechcommunication. 1200—2400 Hz and 2400—4800 Hz.3 Table 6.7 Over 77 Over 70 0.2 lists sound level/distance relationships for speech intelligibility.3 Backgroundsound levelsfor telephoneconversations Quality of conversation Satisfactory Slightly difficult Difficult Sound level dB(A) 58 68 82 NR 50 60 75 Unsatisfactory Over 82 Over75 .3. with a fourth centred at 4000 Hz added later.4 0. and background noise rating (NR.0 2. acceptable PSIL (dB) 45 30—35 25 30 45 30 Table 6. In addition to developing noise criterion curves Beranek (1989) defined the speech-interference level (SIL).3. covering most frequencies of speech. The average sound pressure level is calculated for three octave bands: 600—1200 Hz. 1994) Background sound level dB(A) 48 53 58 63 68 73 77 Max. except at very close distances. This enabled the influenceof background noise levels on speech to be quantified. The average sound pressure level in the four bands is termed the preferred octave speech interference level (PSIL). The PSIL levels shown in Table 6. distance for NR 40 45 50 55 intelligibility (m) 7.190 WINDOWS IN BUILDINGS versation. 1000 and 2000 Hz.3.3.2 Too noisy for speech Table 6.

exposure to loud noises is generally not experienced. fatigue. The issue of personal susceptibility to noise remains a contentious one. and decision-making criteria. High noise levels render speech unintelligible. The effect of noise upon human health is variant and a function of the nature. making people less reasonable . more general physiological. From 35 dB(A) the time required to fall asleep increases and the duration of sleep falls. including headaches. In the office environment.4 Health and productivityeffects Numerous studies have been carried out to investigate the effect of loud noise on human health. psychological and social effects occur. Studies to date have failed to identify any reliable formulation of the problem. and the problem is associated with longer exposure to lower sound levels. In industry this can leadto inefficiency andpossiblyserious or even fatal accidents.Environmental stressorslike noise affect the operation of cognitive processes. which contradicts the common belief that certain individuals are more sensitive to noise than others and experience exacerbated effects due to noise. Interference with communication is one of the most common complaints of noise. Firstly.g. and can restrict the understanding of warning signals. however. shows the quality at various background sound levels for 6. Perception and memory are controlled by these higher processinglevels. activity disruption. nausea. The results of the study highlighted two points. Problems in measurement and quantification of such effects mean that a relationship between noise discomfort and productivity loss is very difficult to develop. which could affect emotional responses. errors of memory. Noise exposure can lead to a number of symptoms being experienced. but have led to a better understanding of its nature. no correlation between sensitivity to noise and noise exposure was found. Secondly. attention and action. or prolonged exposure to higher levels of noise on hearing ability. Psychologists suggest that information is handled on different processinglevels. nervousness.ACOUSTIC PROPERTIES OF WINDOWS 191 of conversation telephone conversations. increased accident risk. capacity for conversation and work efficiency. duration and location of the noise. The impact of more moderate-intensity noise levels upon both human health and work productivity is. e. less well defined than for loud noise impacts. Increased agitation or annoyance can also result. and feelingsof isolation. At lower sound levels. Smith and Stansfield (1986) studied the effects of aircraft noise on selfreported everyday errors. accentuated irritability and aggressiveness. reduced concentration. however. Sounds of only 20 dB(A) engender increased alertness during sleep. with higher levels controlling the actions oflower levels. a positive relationship was foundbetween aircraft noise exposure and the frequency in occurrence of minor errors.

Holloway (1969) suggests that if nonmasking noise can cause problems in speech intelligibility. impulsive tones are generally more annoying than constantnoises ofthe same sound energy level. • Noise. not only its intensity. One suggestion for differences in personal response is that expectationsdiffer between individuals. disorientation and general unpleasantness in response to infrasonic noise sources have been published. yet the impact of low-frequencynoise had received little or no attention (Broner. His findingssuggestthat very few people are annoyed at noise levels below 45 dB(A). At 65 dB(A) approximately 25% of the population are highly annoyed. where the unknown element oflow-frequencynoise effects had been neglected. especially in telephone communication. There is a need to interpret and understand noise signals. requiring more effort. There is an increase in the number of people annoyed as the outside noise level rises above 60 dB(A). raise level of arousal and improve concentration and performance in relation to simple tasks. • Loud noise is inherently annoying and distracting. Annoyance due to noise is often associated with psychiatric ill-health. Reports of nausea. Two noises may have the same sound energy level.192 WINDOWS IN BUILDINGS than they might otherwisebe. 1978). may reduce arousaland have a detrimental effect on performance. Annoyance was defined by the World Health Organisation (WHO) as afeeling ofdispleasureevokedby noise. particularly loud or monotonous ambient noise. This implies that the person listening acts as an information processor of limited capacity. Many studies have investigated the effects of infrasonic noise on people. Shultz (1978) has made an attempt to quantify annoyance levels by way of noise level. The focus of problems due to noise from these sources is annoyance. Leventhall (1973)noted adverse effects on performance at noise levels lower than 80 dB(A). especially speech. Man-made sources of noise at low frequency (20—100 Hz) include compressors. cars and ships. although responses to low-frequencysound have varied from sleep disturbance to threats of suicide in people who are otherwise disturbed. which produces annoyance. but one may be foundmore annoying than the other if it is expected to be quieter. Another reason may be the nature of the noise. Loudness of speech is also known to rise as difficulty in understanding increases. increase the difficulty of the listener.boilers. however. Adams and McManus (1994) summarised the effects on task performance and communication in four statements: • Noise may stimulate people. Broadbent (1980) notes that it is the meaning ofnoise.then it is possible that difficulties can arise in other tasks performed simultaneously with listening. Noise whichis insufficientto mask speechsignals can. • Loud noise interfereswith physiologicalmechanismswhich are essentialto complex task performance. but that improvements in performance sometimesoccurred at .

All of these effects.ACOUSTICPROPERTIES OF WINDOWS 193 that higherlevels. impulsive. whether directly or indirectly. except in conditions where this is required for short periods. irritability. Whatis certain. It cannot be measured using electronic equipment. • Annoyance due to disturbance from noise sources is one of the most frequently received complaints. This reduces the efficiency of work tasks performed. and prejudice. that attenuation of external noise sources to an acceptable level via sound-reducing glazing solutions is beneficial to work performance. • A large number of factors influence noise perception. similar to that experiencedin higher-frequencynoise. reduced efficiency.or containing errors. Complete silence for prolonged periods can be very disturbing due to the effects of sensory deprivation and feelings of isolation. negatively influencing work performance and productivity.This suggests individuals are affected by the arousal effect. Reasonable noise levels penetrating from the exterior can also be beneficial to work productivity. • . is that tonal. type of community. nausea. Noise within the workplace can be beneficialto efficientoperation of machinery and equipment. Interference with communication is one of the most common complaints associated with noise. nervousness. It is therefore difficult to draw firm conclusionsrelating to noise and perceivedaural comfort. recording studios. The complete elimination of noise is not encouraged either. and feelings of isolation. e. Noises which are tonal or impulsive in nature are found to be most annoying. reduced ability to converse. Findings may be summarised as follows: • Noise exposure can lead to a number of symptoms being experienced. reduced concentration. • A substantial majority of workers sitting close to a window wall find disturbance from external noise sources to be unacceptable. as it gives an indication of the state of working order. • Where the noise disturbance in an environmentis deemed unsatisfactory. and is only partially correlated with dB(A) measurements. • A direct relationship is confirmed between noise exposure and frequency ofminorerrors.g. increased numbers of accidents. disruption to tasks. and may eventually lead to serious errors in instances where workers experience fatigue. including headaches. aggressiveness. then. themaintasksbeing carried out may remain unaffected but at theexpense of secondary tasks becoming inefficient. including personal health. fatigue. loud and disturbing noises cause annoyance. social habits and class. It may also lead to the early onset of fatigue in workers. limit an individual's capacity to focus on a specific task. psychological well-being. personality. however. which is detrimental to work task performance. Disturbances in communications is both inefficient and dangerous. reducing productivity and performance. It is easily seen.

Adams and McManus (1994) conclude from a review of research techniques that the relationship between measured noise and noise level effects is extremely complex. • The example office is in a city-centre location. including personal health. wall construction. An environment which is neither too noisy nor too quiet is optimal to work performance. orientation and construction: • Each window configuration is installed within a typical open-plan office setup. community type.5 Noise transmission throughglazing This section concerned with the noise that penetrates from the outside of a is building (the source) to the working environment inside (the receiver). andwindow sash and frame construction remaining constantacross all permutations. There is a clear need to design windows such that annoyance and disturbances caused by unwanted sound sources are minimised. The section which follows provides a method for assessing the sound attenuation properties provided by windows.1. on an intermediate floor of a building and has one external wall. cavity width and infill gas for a given window configuration are evaluated.194 WINDOWS IN BUILDINGS • Complete elimination ofnoise sources can be equally detrimental to work performance. . while ensuring that thermal and visual qualities are not compromised in terms of occupant comfort and potentialenvironmental impact. psychologicalwell-being. location. and prejudice. with office location and orientation. Theaim is to evaluate noise-attenuation properties for any given window construction. 6. Frequency-dependent sound reduction is then converted into a single number characterising the acoustical performance for the window. There are a large number of influences on noise perception. due to the attenuation properties ofthewindow (thetransmitting medium). and varies from person to person. Artificiallysuppressed acoustic environments could reduce arousaland distract concentration. personality. The procedure for calculating aural performance of windows will be illustrated via Example 6. • Frequency-dependent values of sound insulation as a function of the • thickness of glazing. social class. social habits.5. An office building will be considered by way of demonstrating the relevant analysis. The followingare the salient features of the computational procedure. The following assumptions are made regarding the office environment.

based on acoustic velocity. Figure 6. SF6 (C0= 128 m/s). Values of C0 for air. Performing a proportional analysis. due to their similar C0 values. It is assumed therefore that if noise attenuation is an important criterion in building design. The method is not suitable for evaluating triple-glazed window constructions. Iso (1996b)and Weir (1998). and work performance is not compromised.11 No acoustic benefit is achieved through the use of argon in place of air in window cavities (Pilkington. there is no noticeable benefit below a cavity width of 50 mm. Xenon was foundto have a very similar C0 value to sulphur hexafluoride. The noise reduction achievable using a triple-glazed window with two 8 mm cavities is less than that for a doubleglazed window with a cavity width of 20 mm. There is a sharp increase in sound reduction whenthe cavity width rises above 50 mm as shown in Figure 6. Calculations can be performed for any window/wall proportions or single-/doubleglazed options.13 2. argon. double-glazed windows with a cavity greater than 50 mm would be preferred over tripleglazed options.2. Sound reduction due to the cavity gas Rgasis a logarithmic function of C0.1 illustrates the frequency-dependent sound reduction for gas infill within cavity widths less than 50 mm.ACOUSTIC PROPERTIES OF WINDOWS 195 The presentanalysis based on the work of Pilkington (1995).23 2. Example6. for which published data are available. the manufacturer should be approached for a detailed breakdown of sound-insulation properties. . this example can be adapted to calculate attenuation properties for other locations. krypton. By changing the prevailing noise spectrum and building construction details.5. the acoustic velocity within that medium. Xenon is therefore likened to SF6 for acoustic calculations. it is possible to interpolate between data values for air/argon and xenon/SF6 to produce data for krypton gas. For this reason argoncan be likened to air in acoustic calculations. 1997). whereby task operation is not detrimentally affected.5.5. The mid-pane in such windows causes acoustic coupling. Noise-reductionvalues for single-glazing panesand numerous double glazed windows are widely available through the literature and manufacturers' brochures. and the attenuation value of the overall window is reduced due to vibration.39 2. Generally speaking. Beranek is and Ver (1992). To obtain data on triple-glazed windows. varying noise sources and different building details.1 assesses the ability ofwindow constructions to attenuate noise to an acceptable working level. xenon and SF6 are presented below: Air Argon Krypton Xenon SF6 C0 (m/s) 340 44 169 135 128 Log C0 2.53 2.

196 80 - WINDOWS IN BUILDINGS 70 60 50 40 30 20 IS 125 250 500 Frequency (Hz) 1000 2000 4000 *4 mm.E 30 25 20 15 10 5 o U) 0) .1 For the data given below find the Sound Reduction Index.4 mm 14 mm+(<50(Xe*4 mm J Figure 6.5.1 Frequency-dependent sound-reduction values for doubleglazed windows incorporating two 4 mm glass panes and a cavityof less than 50 mm C • C .r(. R (dB) for the facade (compositewall) and also find the single-numberratingwhich takes into .1 0 0 100 200 300 400 Glass separation (mm) Figure 6.(c50)Kr.5.r50)uiru-4mm —I4 mrc.5.2 Relative soundinsulation with increased air space width Example 6.

f(Hz) Sound reduction. and has frequency-dependentvalues as follows (Pilkington. kryptonfilled window. These data were reported by Weir (1998). the frequency-dependent noise-reduction values are compared to the .is given by (Pilkington. Next. in this case a doubleglazed. (60% of facade) =21 m2 Wall area. 23 23 (dB) A R 500 38 1000 50 2000 59 The first requirement is to evaluate the frequency-dependentvalues of airborne sound insulation for the selected window design.ACOUSTICPROPERTIES OF WINDOWS 197 account the given sound spectrum. R. 1969): Frequency (Hz) Rwaii (dB) 125 40 250 45 500 53 1000 59 2000 64 Hence.5. R for a frequency of 250 Hz25 dB. The single-frequencysound reduction index.5. Given data Office location: city centre Window type: 4-12Kr-4 Window area. The window to be used is a 12 mmkrypton fill between two 4 mm panes of glass. m2 Rwaii = sound window area (m2) reduction index of masonry wall at 250 Hz. An example calculation is presented here for noise at 250 Hz frequency. Thefrequency-dependentairborne sound insulation valuesmaythenbe used to obtain single numbers characterising the acoustical performance for the given window configuration. (6. A summary of the sound-reduction values. Awaii (40% of facade) = 14 m2 Glazingproperties 125 250 Frequency.incorporating a glazed area.1) A1= total area of wall and window (facade area).1). This calculation is repeated for noise frequenciesin the range 125—2000 Hz. using Eq. m2 Awaii= area of the masonry wall.and double-glazed windows is shown in worksheet Calc6-01. dB = sound reduction index ofwindow at 250 Hz. R. 1969) R = 10 Log where A1 IORwaIl/ A + 1ORw/lO (6. R. for a composite wall. for a number of single. is calculated for the above frequencies. The single-frequencysound reduction Index. dB Rwaii is takenfor that ofa masonry wall ofminimum thickness300 mm.

xls. Note that the wall sound reduction data must be provided in thecell range D7:H7 of the latter sheet. Spectrum adaptation terms.3.00 500. given in dB. after shifting it in accordance with this procedure. Thereferencecurve is iteratively shifted downwardsin steps of 1 dB towards the calculated curve until the sum of unfavourable deviations isjust below the value of 10 dB.00 30.00 Frequency.xls automates the procedure of computing R. • Select thewindow type for which sound reduction properties are required. The sum ofunfavourable deviations is then obtained for the calculated and referencedata. the calculated value is above the reference value. • Copy and paste the sound reduction values for the given window con- struction.to sheet 'SRI calculation'. = Rweighted.5.3 Reference values forairborne soundreduction in octavebands (ISO.198 WINDOWS IN BUILDINGS 60. the difference is set to nil. The workbook Calc6-01.00 250.5.In this example Rweighted 41 dB.00 35. The procedure is as follows: • Open the workbook Calc6-0l.00 C a) 40. for any given frequency.00 45. of the referencecurve at 500 Hz. If. • Select the worksheet labelled 'Basic data'. 1996b) relevant reference curve. in decibels.00 2000. Rweighted. are added to the single-number rating to take account ofthe particular sound spectra. ISO (1996b)provides a detailed account of the procedure to calculate C0 and Ctr.00 50.00 r a) 55. For this reason a single adaptation term is calculated respectively for background outdoor(C0) and traffic noise (Ctr). The predominant noise source in a city centre is the background outdoor noise and traffic noise. The value. cell range D9:H9. .00 125. The term unfavourable deviation is defined as the difference between the reference and calculated value. gives the sound reduction index. for each frequency.00 Figure 6. The curve of reference values for airbornesound in octave bandsis illustrated in Figure 6. C0 and Ctr thus characterising the acoustical performance of any given window. Hz 1000.

ACOUSTICPROPERTIES OF WINDOWS 199 • Activate the 'Rweighted Calculation' sheet.Wiley. (1969) Noise and efficiency: the spoken word. C = —4 dB and Ctr = _9 dB. Thus. Acoustics: description and measurement of environmental noise.E. a building that provides excellent thermal and visual conditions is of limited use if the auralperformance is such that worktasksare compromisedand buildingusers are dissatisfied. (1989)Balanced noise-criterion(NCB) curves. As previously mentioned. (1994) Noise and Noise Law: A practical approach. ofthe AcousticalSoc. and mental health. F. C and Ctr values. Theaural analysis presented above is important to thedesigner and architect in providingbuildings whichare pleasant and comfortable places for work. References and McManus. 650. London. 483. Holloway.J. this means that the sound insulation value of the facade is 37dB (41—4=37) for outdoors and 32 dB (41—9=32) for traffic noise. L. on Architecture. (1978)The effects of low frequency noise on people: a review. ISO (1996a) 1996—1. ISO (1996b)717—1. D. (1992) Noise and VibrationControl: Principlesand applications. The above example provides an efficient and simple-to-use procedure for analysing nose transmission through glazing systems. (1980) Noise in relation to annoyance. ofthe 3rd European Conf. E&FN Spon. Broadbent. AcousticalSoc. (1993) Thermal comfort evaluation for passive cooling: a PASCOOL task.Y. N. J. C.L.p.L. British Standards Institution(1997) BS 4142: Rating industrial noise affecting mixed residential and industrial areas. Proc. L. The single-number quantity of the krypton cavity-filleddouble-glazedwindowis hence given as: Rweighted(C: Ctr) = 41(—4: —9) dB This implies that the facade has a frequency-dependentairborne sound insulation value of 41 dB with an outdoor noise correction factor of —4 and a traffic noise correction factorof—9. ofAmerica86. Baker. and Yer. .L.S. New York. Beranek. New Scientist. M. Beranek. 103. N. of America68. and which maximiseemployee productivity. performance. Broner. Adams. • Cell range B3:B5 provides the Rweighted. of Sound and Vibration 58. 15. Acoustics: rating of sound insulation in buildings and of building elements airbornesound insulation.In simplerterms. J. I.

.. Cu SfB (31). Napier University. Pilkington plc (1997) Glass and noise control. (1998) Life Cycle Assessment of Multi-glazed Windows. and everyday errors. St Helens. Pilkington Glass Products. Wilson. Pilkington plc (1995) Glass and Noise Control. H. PhD thesis. (1993) Measurementsof background noise levels in naturally ventilated buildings. T. Proc. (1973) Man made infrasound. Pilkington plc (1969) Windows and Environment. (1978) Synthesis of social surveys on noise annoyance.J. J. Smith. 129. Environment andBehaviour 18. Nicol. 377. G. Proc. Pilkington Glass Ltd. (1986) Aircraft noise exposure. Centre National de la Recherche Scientifique. UK. and Singh. Weir. A. 214. 1997. Pilkington Glass Products. S. and Stansfield. F. Technical Bulletin. on Infrasound. ofthe Colloq. R. M. July. of the Institute ofAcoustics 15. ofAmerica 64. of the AcousticalSoc. Paris. May. 1995. Shultz. noise sensitivity.G. England. its occurrence and some subjective effects.200 WINDOWS IN BUILDINGS Leventhall.

It is necessary to consider the impact which raw material extraction. accounting for over 50% of the UK's total CO2 emissions. Carbon dioxide (C02) has many influences on the environment. manufacturing processes. use patterns. This has prompted research into new material production. energy production. shows the carbon emissionsresulting from the production of some commonly used building materials. manifested in rising global temperatures. and has a very large part to play in global warming issues. A world-wide rise in temperature.1. product design. water vapour. processesand activities have on local. extracted from Buchanan and Honey (1994). CO2. have also increased in concentration since the commencement of industrialisation. regional or global environments to be assessed. and ozone in the atmosphere.scientists. which accounts for about 5 billion tonnes of CO2 pro- . Table 7.designers. methane (CH4). methane. animal habitats. transportation needs and waste disposal requirements have on both social and natural environments. manufacturing methods. The effects of rising concentrations of C02. Impacts on the environment include effects on the atmosphere and the world's natural resources. noise pollution. not found naturally in the atmosphere. such as chlorofluorocarbons (CFC5).managers and environmentalistsalike. and waste disposal management. recyclingtechnology. technologists. LCA enablesthe effectswhich products. nitrous oxides (NO). cause sea levels to rise and affect crop production. Warming of this magnitude would alter climates throughout the world. and the availability of raw materials and primary fuel for the future. Other gases. estimated at 2°—6°C over the next century could result from these changes. Greenhouse gases absorb infrared radiation from the sun's heat which would normally be reflected back into space. and is caused by important natural gases. The natural greenhouse effect maintains the surface of the earth at a temperature which enables inhabitation. in addition to human health factors. and ozone (03).0. fuel depletion. Much of the CO2 in the atmosphere is a direct result of fossil fuel burning.7 WINDOWS AND THEIR LIFE CYCLE Since the late 1960s life cycle assessment (LCA) has become an increasingly importanttool for engineers. This retained heat in the atmosphere is reported to be very slowly increasing the earth's temperature. In the building and construction trade CO2 is the largest environmental concern. was first noted at the beginning of the nineteenth century by Jean-Baptiste Fourier.

which use reduced quantities of energy and which contribute reduced amounts of greenhouse gases to the atmosphere in their use phase. The life cycle of a productis not simply its period of useful employment. (1994) state that LCA has become one of the most actively considered techniques for the study and analysis ofstrategies to meet environmental challenges. The other two major influences on CO2 production are cement manufacturing and changes in tropical land use. use and maintenance to disposal or recycling. process or activity with a view to identifyingthose stages of the life cycle which have the greatest adverse effect on the environment. A simplified illustration of the LCA process is shown in Figure 7. but includes each stage from raw material extraction. 1995). 1995). Field et al. specific to the LCA of multi-glazed windows. in addition to the underlying requirements of administration and workforce services. retailing. 7.0. The development of products with low embodied energy contents.andconclude that it is one of the most promising approaches to integrate environmental knowledge and data into a framework for action. All units in kg carbon per m of material (Buchananand Honey. has become of prime importance. .1. There are four main stages to an LCA: • Planning • Inventory analysis • Impact assessment • Improvement analysis EachLCAstage is discussedin detail throughout this chapter.1. manufacturing. accounting for about 1 and 2 billion tonnes respectively(Callander.1 Carbon emissions Carbon released Treated timber Glue laminated timber Structural steel Reinforcedconcrete Aluminium 22 82 8132 182 Carbon stored 250 250 15 Nett carbon emitted —228 —168 8117 182 6325 0 0 6325 duction annually. the transportation effects must be accounted for. This alone is responsible for a yearly CO2 concentration rise of 0.202 WINDOWS IN BUILDINGS from the productionof commonly used building resulting3 materials.1 Definition of life cycle assessment LCA takes a holistic view of theentirelifecycle of a product. 1994) Table 7. At eachstage.4% (Callander.

LCA is a relatively new science and its application to window design is even more recent (Weir.caremust be takennot to generalisethe resultstoo far. technical staff. LCA results for very many window designs are not readily available. 1998). Without close liaison with such personnel it is easy for the LCA to be misdirected.2 Planning The main purpose of the planning phase is to define the investigation boundaries. LCA is a very specific tool and whereas it can be applied to almost any product. categorisingofresearch and developmentneeds or aggregationoftotal product environmental burdens. transportation managers. new productdevelopment.WINDOWS AND THEIR LIFE CYCLE 203 LCAboundary Fresh water Energy Air emissions and watereffluent By-products Figure 7.1 Input/outputdata requirements for a comprehensive life cycle assessment (Weir. System limitations may include such considerations as data availability. LCA may be used for many different purposes. such that the most up-to-date and relevant information is included in the analysis. future environmental effects and technology development requirements. including the comparison of a range of products. The example LCA which follows is based . 1998) 7. process or activity. Thus. During the planning of an LCA the objectivemust be clearlyborne in mind while the system limitations are definedandthe methods for data collection and evaluation are set out. windows users. Throughoutall stages of the window life cycle it is necessaryto be in close communication with production managers. designengineers and building construction experts.1. identification of harmful stages in a product life cycle. allocation of process resources.

2 m and manufactured in Norway using Scandinavianpinefrom well-managedtree plantations. in alternative building types.1. krypton and xenoncavity gases. production processes. based in any location desired.2.5 m 6. The objective of this LCA is to outline the key sources of environmental impact throughout the life cycle of a multi-glazed window. in a defined location. The useful life of the window is considered to be 20 years. A plan of the office space is shown in Figure 7.204 WINDOWS IN BUILDINGS upon a specific window type.2 m by 1. The window considered is of timber frame construction.4 m Ceiling height 2. manufacturing techniques and life cycle activities to reduce this burden over the holistic life of a window. 1998) .8 Figure 7. A clear definition of investigationinclusionsand exclusionsandthe methods of data retrievalmust also be set outduring the planning phase. and to identify realistic improvements which can be made to the window design. measuring 1. manufactured in a specific location and used within a specific building type.1 Example office dimensions and layout (Weir. The LCA used in this chapter is subject to the following conditions. The principles and techniques used in this analysis can easily be generalised and used to perform subsequent LCA studies on different window types and constructions. It is transported by sea and road to an Edinburgh location where it is installed within a typical three-story office building.and triple-glazed windows.2. with and without the use of low-emissivity glazing coatings. and therefore form the investigation boundaries here: Inclusions • raw material extraction • manufacturing processes • packaging requirements • transportation requirements • in-use repairs and maintenance • waste disposal and recycling • fuel production and use • energy generation Soothwall Exclusions • manufacture of capital equipment • maintenance of capital equipment • manufacture of services production • maintenance of services • energy required to raise capital for 13. The setting of these have a significantinfluenceon the results generated. argon. double. A comparison of 18 multi-glazed window constructions is presented showing the difference between single-. and the use ofair. The followingset the boundaries for most general LCAs.

and simplifies the process of turning data into meaningful information. the energy content or 'embodied energy' of the raw materials must be quantified. The functional unit adopted here is Joules per unit output (J/window). This can be achieved by close examination and measurement of manufacturing operations. 7. each of these raw materials requires to be quantified in terms of mass or volume. or by averaging energy consumption over the number of window units manufactured annually. Fourthly. Embodied energy is a measure of the total energy required to extract. Thirdly. There are several sources of literature which provide this information. This is done by examining and listing the main components which comprise a finished window. It may be helpful to reduce the data acquired from many different sources to the same common unit. and is not somethingwhich is carried out as a 'oneoff' exercise. information and knowledgefor the second. TheLCAofmulti-glazedwindowsrequires severaldata-acquisition methods to be adopted. however. or a component product whichmay be used as part of a larger product. as LCA is an iterative process. process and manufacture a raw material intoa finished product. The overall objective of the LCA may be achieved through a fifth or sixth iteration of the process. technical. but points to the fact that LCA is a commitment to ongoing environmental improvement.1 Raw material extraction In order to quantify the environmental burden associated with raw material extraction it is first necessary to identify all major raw materials required for the production ofmulti-glazed windows. It would be very time intensive to assess each raw material prior to it arriving at the factory door. This can be done by dismantling a finished window and weighingor measuring the volume of each component. third.2 Manufacturing The total energyconsumed within the manufacturing plant is not limited to the consumption of energy for electricity and machinery power. It is acceptable to do this.WINDOWS AND THEIR LIFE CYCLE 205 It may be necessaryto change the objectivesof theLCAif no method can be found to obtain the data required. This may be a lengthy procedure. the production and manufacturing processes need to be observed such that raw material wastage can be assessed and quantified. and finally. Assessment is made by detailed measurement of individual processes. process or activity provides furtherdata. iterations of the LCA procedures. This is termed the 'functional unit'.2. Secondly. 7. . Care must be taken.2. etc. to ensure that the basis of this information is in keeping with the objectives and boundaries of the LCA. building services and support facilities. but also includes all administrative. The first LCA of a product. For simplicitythis is divided into the following sections.

2.4 Use The utilisation phaseof a number of products has little impact on energy consumption and environmental impact. It is necessary to calculate the energy consumption associated with space heating and electric lighting. This may employ several different data-acquisition methods. of finished windows to the construction site. analysis of historical data. As the number ofLCAs being performed increases.2. Additionally. as discussedin Chapters 3—5. Energy. This is discussedat length in Chapter 3. raw materials. from physical measurements to database searches. air emissions. however. as briefly outlined in the planning phase above. reduce the need for electric lighting and thus reduce the electricity requirement over a lifetime of use. This is a measure of all matter which crosses the boundary defined in the planning phase.5 Recycling or disposal life within the building envelope is over is very difficult to assess. windows. Consideration must be given to the transportation of raw materials to the factory site.available information on recycling and disposal will become more readily accessible.2. 7. surveys. 7. questionnaires. and of disposal/ recyclingof windows at the end of their useful life. which are designed to permit quality daylight into an occupied space. the life cycle is divided into a number of stages: • materials and manufacture • transportation • use • disposal and recycling . using the transmission properties of windows. Windowswhich aredesigned to high thermal performance specificationssignificantly reduce the requirement for space heating over a lifetime ofuse. to gain a holistic view of window performance. The design and construction of windowsdo.3 Transportation All stages and methods of transportation should be assessed and aggregated. 7. impact heavily on energy consumption.3 Inventory analysis The energy associated with recyclingor disposing ofwindows once theiruseful The inventory analysis is a methodical quantification of inputs and outputs. To simplifythe process.enabling engineers and environmentalists to make better and more informedjudgements on the recycling and/or disposal ofmulti-glazed windows.206 WINDOWS IN BUILDINGS 7. theoretical calculations and individual interviews.water-borne effluentand solid waste are examined and measured.

3.000114 0. The second issue concerns the consequences which must be addressed regarding the energy input and environmental burdens created as a result of production. heating.9 0. The first ofthese is the associated benefits. 7. Timber processing. Administration and services needs are accounted for in the factory as a separate but necessaryissue.069 . marketed and sold there must be administration personnel. a design team. In this analysis three infihl gases are investigated.1 Percentagecomposition Component Nitrogen Oxygen CO2 ofair by volume (Fernieand Muneer.1 Inert infill gas There are two issues to be dealt with when considering the environmental impact ofcavity infill gases. lighting. and management. heating. krypton and xenon are all present in the atmosphere. and compared to the use of air in cavities. the window requires detailed inventory analysis work due to the number of components involved and the complexprocessesrequired to manufacture a finished window.1 shows the percentage content of gases in air. services.3. factory workers. in terms of improved thermal properties and reduced heat loss from thewindow. Trace gases include Table 7. aluminium optimising. glazing unit production. 1996) Percentage 78 21 Argon Krypton Xenon Trace gases 0. and technical design needs are all analysed individually.3. material use and environmental impact.1 Materials and manufacture Material requirements and manufacturing processes are simpler for some products than for others. These issues are discussed in full in Chapter 4.03 0.WINDOWS AND THEIR LIFE CYCLE 207 7. Table 7. Each of these has implications for energy consumption. To sustain a workforce requires services.0000087 0. painting.3. • Argon • Krypton • Xenon Argon.with a final aggregation to sum the data collected. and production of the pure gases is a process of separation from other components present.1. Each main material or component part within the finished window is discussed below. assembly. administration. For the product to be produced. lighting and materials.

carbonmonoxide. Filling the glazing units with inertgas is a carefully monitored and controlled process. 1996). The equipment required for the separation ofgases from air was reportedby BOC in Middlesborough.208 WINDOWS IN BUILDINGS Waste N2 Pure N2 80000rnTh Losses 140000rn/h 4600rn/h 15/h Liquid 02 tlOflUfll 500 m3Th Gas02 20 000 rn/h Kr Xe purification CH4 / Gas 02 Kr91. ozone.3% Xe 6. krypton andxenon gases to fill this gap is 12 kJ. The process of separating inert gases from air is shown in Figure 7. Gas is detected using special .1 Gas productionschematic for BOG Middlesborough plant (Fernie and Muneer.2 and 8.requiring a gas volume of 17.and neon. liquid oxygen pump (2. The actual yield of gas from the column is between 31% and 54%. Krypton and xenon. 1996) watervapour.4 kW). such that there is no 'flushing' process taking place. helium. 13. Production rate data is basedon an average yield rate of42%. flash vaporiser (4 kW). 508 MJ and4.3.9% CH4 1.8% Figure 7. methane.8 kW). heater (10.3 kW). UK. krypton and xenon would be of 16. consisting of a reactor (4 kW).3. require the running of a crude krypton/xenon column.1 m2. however. For a window measuring 1. to have a power rating of 12 MW.5 GJ respectively (Fernie and Muneer.6. This is the only equipment required for extracting argon gas.2 m by 1. The energy requirement to produce argon. and the cavity gap for argon.2 m the area of glazing is 1. extracted from Fernie and Muneer (1996).hydrogen.1. The gas quantities and pressure regulation ensure that the mixing of gas with the displaced air is minimised. 12 and 8 mm respectively. and purification column (2.8 litres.The panes are filled via filling and control holes drilled in the spacer. and the amount of gas flow is regulated using flow valves.

whereby two trees are planted to replace the felling of one. once the correctconcentration is achieved in the pane. The loss of vacuum ensures that each timber pore is filled with preservative. Details of the energy analysis are as follows (Weir and Muneer. and is obtained from organised.3 Timber sash and frame The timber usedin the window manufacturing process is 100% softwood.1.1 kWh/rn2low-e coating 1. One very populartreatment applied to softwoods is preservative impregnation. but are minimal due to the use of efficient controls. and the gas flow is stopped by the closingof a solenoid valve.2 Low-emissivity coating Float glassis produced 24 hours a day.WINDOWS AND THEIR LIFE CYCLE 209 probes.1. Norway 30%.4 kWh/rn2low-e coating 2. and a new vacuum at a higher pressure is created over a longer period of time. The vacuum is maintained for a few minutes before a dischargepump is operated and the vacuum is lost. Timber has a high strength to weight ratio.About2 billion tonnes of CO2 enters the atmosphere as a direct result of forest destruction each year (Callander. for the production of low-e coating for MJ per pane.3 kWh per pane.The gross area of glass produced is 3097 m2/hour (Liggett. does not corrode. 1997): Natural gas load Electrical load Total load 3165 kW 783 kW 3948 kW Energy input Energy input Total input 1. the energy content is 2. which soaks into the timber pores. Sweden 40%).1 m2. 7. It possessesgoodthermaland sound-insulating properties. and given the right conditions and treatment will not rot. thus lengthening the window life cycle. Plantation forestry accounts for approximately 10% oftimber requirements world-wide. a window measuring 7.3. and is easily formed for many applications. Exact levels of gas wastage are not known. The chamber is then emptied of preservative. All timbercomponents are impregnated with a white spirit-based solution. and requires a weekly input of 5000 tonnes of materials for a typical float glass production plant.4 Therefore. . well-establishedtree management programmes in Scandinavia (Finland 30%. The preservative tank is then opened and the chamber is filled. The net yield of glass allowing for breakage and trimming of the glazing edges has been accounted for in the calculations. The glass ribbon produced is 3. To achieve impregnation all timber sections are enclosed in a designated chamber and a vacuum is created. 1995). The annealing lehr works at a speed of 870 rn/hour. and slows down the process of decay in the timber.56 m wide. A final discharge completes the impregnation process.7 kWh/rn2low-e coating 0.and can be produced in a variety of thicknesses. 1996). and the timber remains within the drying chamber for 1—2 days.3. or 8.

6 GJ/m3. Energy content is a measure of the total energy inputrequired to extract raw materials and manufacture them into useful materials or finished articles. (1994) was found to be most closelyrelated to the LCAboundaries under Section 7. but excludes use and recycling. processing and manufacture. Work by West et al. estimated that 0. and 4. requiring differing treatment processes and having varying energy contents. must be keptcentral to all decisionmaking in the analysis. sash 10. 1994) Timber type Rough timber Glulam timber Hardboard Softwood Energycontent (MJ/m3) 848 4500 20600 15500 12900 Particleboard . Allowing for the waste within the window production process and given the window dimensions and type. incorporating many differencesin procedure. according to the water content in the timber. of 192 MJ. The density of softwood varies greatly. a contribution for the glue involved in laminating must be added.5 GJ/m3. extracted from Buchanan and Honey (1994). waste 12. andhence an energy content of 2.51 GJ/m3. but an average of 500 kg/m3 is used. (1994) produced data for indigenous softwood which includesentries for There transportation.94 kg. This information must be used within the boundaries of the analysis.Varying boundary inclusions can cause energy content valuations to differ greatly. Table 7.2. Honey and Buchanan (1992).6 GJ/m3 of softwood timber is used in this analysis. exclusive of machining. timber must pass through a laminating process where two or three sections of timber are glued and compressed together. This gives an energy content for the timber component. The working boundaries.2. and normally includes the energy associated with packaging and transportation.77 GJ/m3. Further studies by Hollinger and Hunt (1990). in their analysis.03 GJ/m3 respectively. softwood having an energycontentof 15.01 kg.210 WINDOWS IN BUILDINGS are many different uses for timber. Analysis by West et al.2 Energy content for some popular forms of timber used in construction (Buchananand Honey. a total mass of 37 kg of timber is used (frame 14. machinery and available technology.3. Manufacturing methods maybe widelyvarying in nature.3. Therefore.2. Halliday (1991) and Dinesen andTrabergy-Borup (1994) arrived at figures of 0.3. material extraction. shows the energy content for some popular forms of timber used in construction. To provide material for the window sash. These final values are widely dispersed due to wide variance in the underlying analyses. as set out in the planning phase of an LCA. This indicates an energy content of 2. 1. Research has shown energy contents which are both higher and lower than thefigures shown in Table 7. to complete the analysis for the embodied energy of timber.05 kg).5 kg ofglue is Table 7.

Aluminium is used for many different purposes within the window construction. Therefore. and secondary aluminum is recycledmaterial. This is because the energy- .melting and reprocessing secondary aluminium uses a fraction ofthe energy to produce 1 kg of formed aluminium (50 MJ/kg compared to 225 MJ/kg for primary aluminium). is not discarded but sold to a metal merchant for recyclingas secondary aluminium.WINDOWS AND THEIR LIFECYCLE 211 required per square metre of laminate surface. The glue is assumed to have an energy intensity of 160 MJ/kg. Aluminium profiles are fitted along the bottom edge of the sash to protect the timber from water penetration. and the casting of ingot. is hence 195 MJ. an average of 0. inclusiveof timber processes and laminating glue. Each layer covers an area of 0. electrolysis of alumina. The total embodied energy for the timber sash and frame.3. while the remaining job requires two layers of glue and three sections of timber. Thetotal energy inputrequired to produce and process the aluminium for one window. The most energyintensive stages in the procedure include the crushing of raw bauxite.04 m2 of glue is applied to each sash (expressingthis in mass terms gives 0. The total waste at the window factory site. and the glazing unit spacer are also made of aluminium to ensure a light-weightdesign.2 m.02 kg of laminate glue). which equates to 3 MJ per sash. wherebylengths offormed aluminiumare cut to the desired size.306 kg per window. Further work by Young and Vanderburg (1995) produced data for both primary and secondary aluminium. as illustrated in Figure 7. Primary aluminium is produced from raw bauxite. various sections of the ventilation grills.3.2. assuming that 100% primary aluminium is used. the energy content of aluminium production is much debated. Twenty-five per cent of timber laminated involves one layer of glue and two sections of timber. As noted previously. taking into account the fact that only 30% of sashes are laminated. The window-openingmechanism. Aluminium smelting and forming is a highly energy-intensiveprocedure.1. is estimated to be 518 MJ. of 0. The final processingis performed in the window factory. and the total waste from each process was calculated as shown in Table 7. Again. while Saito and Shukuya (1995)quote a content of503 MJ/kg for pressed and finished aluminium window frames.3. The total mass of aluminium used in a standardwindow of 1.07 m2. and must be considered in the LCA of the window unit.3. All other processingis performed at the aluminiumplant. having gross energy requirements of225 MJ/kg and 50 MJ/kg respectively. to produce soluble aluminium. The main consideration for material consumption and energy use comes at the aluminium manufacturing stage.4 Aluminium The processes involved in cutting and assembling the aluminium components for the window construction involve minimal quantities of energy. 7. The process of manufacturing aluminium components from raw bauxite is a multi-stage procedure.2 m by 1. Buchanan and Honey (1994) quote a content of 130 MJ/kg for general aluminium.

24 0.30 .4kgbauxite 1 kgaluminium Figure 7.99 0.212 WINDOWS IN BUILDINGS 6.16 1.17 1. 1998) Aluminiummass Windowcomponent Glazing unit spacer Frame ventilation Outer protection Window mechanism Total (kg) Totalaluminiummass % Waste 3 17 17 17 (kg) 0. 1995) intensive processes of smelting and electrolysisare not required for secondary aluminium.42 0.3. UNIDO (1989) estimate that 27% of the world's total aluminium production comes from recycling aluminium.2 m by 1.25 0.2 Production ofaluminium from rawbauxite (Saito and Shukuya.2 m (Weir.66 0.20 2.3 Estimate of total aluminium mass incorporated into one window measuring 1.3. This percentage is likely to be lower for building materials due to design for long life spans in construction Table 7.19 1.

There is less discrepancy between these values than for aluminium and timber. 7. West et al. lighting and heating estimated. security mechanisms. 16.3. the work of West et al.4 lists the mass of materials used. An energy and material analysis for the glass sheet manufacturing was carried out. 7.3. Having discussed the major material inputs 7. administration. and an energycontentof 13 MJ/kg ofsheet glass is used. and small offcuts which are not able to be used. which must be accounted for in the LCA study. 7. Further milling is required to .1. Saito and Shukuya (1995).1 m2 was measured to be 21. han- dles. requiring minimal processing on the factory site.3.If the assumption is madethat about a quarterof the aluminium used is recycled.5 Sealed glazing unit The production of large glass panes supplied to the factory is also an energyintensive process. services. then the energy intensity for the aluminium used in the win- dow becomes 409 MJ. (1994). (1994) was found to be compatible with the aims ofLCA. the manitselfmust be considered and the energy consumption for ufacturing process production.6 Components A number of components are assembled into the window construction. Table 7.26 kg having an energy content of 289 MJ. Wastage is minimised due to the high levels of technology implemented and amounts to approximately 5.5% of the total glass utility.2 kg in the finished product.3. and associated embodied energy values. The mass of glass required for a double-glazed unit of 1. A small amount of material is wasted due to breakages.Againthis is sub-dividedfor simplicity. and 18. and fewer differencesbetween manufacturing methods exist.6 MJ/kg respectively. Again. and Dinesen and Trabergy-Borup (1994) reportthe energy content per kg of sheet glass manufactured to be 13 MJ/kg. seals and ventilation grills are fitted to the completed sash and frame.1. This is perhaps because the technology used is standardised.9 MJ/kg.WINDOWS AND THEIR LIFE CYCLE 213 components.3. Taking this into account means that the average glass consumption for one unit is 22.2.1 Timber sash and frame production Sections oftimberare first moulded using an energy-intensive sequence ofmills and saws which provide the necessary profiling. All components are bought as finished products.2 Manufacturing to the window system. The total embodied energy of materials used for components was calculated to be 144 MJ.

3. 1998) . An aluminium spacer is fitted to separate the glazings. programmed cutting tables. It is possible therefore to divide the annual energy consumption by the number of units manufactured. the edges aroundlow-e coated glass sheets are ground to ensure that the adhesive seals the unit adequately.2.04 — Rubber Total 67.214 WINDOWS IN BUILDINGS Table 7. Approximately 6 MJ of energy are consumed per finished window. Before the unit is sealed it is filledwith inertgas. and is ready for assembly into the window sash and frame. 1998). The glass sheets are precision cut on large. Glass panes are then washed to remove dirt and dust particles from the interior of the unit. Secondly.5 0.11 0.3.3 Sealedglazingunitproductionline process (Weir. and is kept in place by a butile compound. 7. Finally. the unit is sealed with a butile compound.9 MJ and 16. The first task in the glass production procedure is to plan the cutting schedule. The equipment used in these processes has a relatively constantload. Figure 7.4* Total embodied energy (MJ) 57.6 349* 160.17 Embodiedenergy (MJ/kg) 68. The average energyconsumption for the production of one sash and one frame were estimated to be 16.4 Materials and embodied energy of components Material Zinc Steel Plastic Mass incorporated Mass incorporated in frame (kg) in sash (kg) 0.3 MJ respectively(Weir. The final two stagesofproduction for the sash and frameinclude assemblyand impregnation.2 Sealed glass unit production Theproductionprocess the manufacturing of sealed glazing units is shown for in Figure 7.05 0.3.03 0.0 143.0* 148.74 0.3.3.8 * Extracted from Buchanan and Honey (1994) shape each sash and frame for the ironmongery and accessories added later.45 0.0* 77 11.

WINDOWS AND THEIR LIFE CYCLE

215

7.3.2.3 Aluminiumprocessing

The cutting machinery aluminium uses very little energy, and has a load for of approximately 4 kW. For a typical finished window, the factory power processing of aluminium consumes a very small quantity of energy, approximately 0.2 MJ per finished window. This is negligible, and it is clear that the energy of production to produce aluminium from raw bauxite is of a greater concern. 7.3.2.4 Lighting and factory services Lighting and factory services are not directly related to the manufacturing of the finished window, but are an essential element of the overall production process. Without these elements, production would be impossible. They are, however, not so easily subdivided into manufacturing stages, and must be accounted for in a unit overhead. Included in these are the factory services of administration, technicaldesign, heating, lighting, and weekendand night-time loads. The average power load, annual energy consumption, and energy consumed per window is shown in Table 7.3.5. Maintenance and repairs are kept outside the investigationboundaries for all stages of production, including machining, lighting, services and transportation. During the product use phase, maintenance and repairs become important, as the length of useful life for a window is a direct function of the treatment it receives while in-situ. 7.3.3 Transportation transportation requirements throughout a product life cycle demands detailed inventory analysis based on the source and destination of all materials, fuels and components. In addition, information on the final destinationsof finished products must be known, and the means by which theyare transported. Different modes of transport have varying energy consumption
Assessing

Table 7.3.5 Average electrical power load and energy consumption for non-machining requirementsper window (Weir, 1998)
Energyuse Administration Technicaldesign Heating
Lighting

Average power load (kW)
70 57 49 256
378 338

Annual energy consumption (GJ) 463
380 1558 3390 5006 3036

Energyconsumption per unit production (MJ)
2.1
1.7 11.3

24.5
36.2

Night-time use
Weekends

Total

21.9 97.7

216

WINDOWS IN BUILDINGS

Table 7.3.6 Energy consumption associated with various modes distancesand loads (Buchananand Honey, 1994) Mode oftransport

of transport for given

Energy consumption (MJ/Kg)
0.11

Road30km Road 50 km Rail 100 km
Rail 200 km Rail 500 km

0.19 0.23 0.15 0.37

implications. Table 7.3.6 lists the energy consumption associated with various modes of transport, for given distances and loads. Materials are sourced from a number oflocations and countries throughout the world. Raw bauxite for the production ofaluminium is shipped 12 000 km from Argentina, while timberis transported by rail from Norway, Swedenand Finland. Glass sheets are bought from a variety of locations, depending upon market availability, currency changes and manufacturing needs. Components are sourced from all over Europe, again according to market and currency variations. Accurate measurement of energy consumption used in transporting these materials to the factory would require thorough analysis and reliable data on market sourcing and transportation systems, and would warrant an independent LCA study of its own. Therefore, literature and past research is extensivelyused. An estimation oftransportation requirements associated with the main material sourcing for the production of one window unit is shown in Table 7.3.7. Energy of transportation of materials was estimated to be 65 MJ per window (Weir, 1998). Window deliveriesto the UK are made in a seriesof stages. Firstly, units are transported by road to Stavanger, where they are transferred to a shed, prior to
Table 7.3.7 Energyconsumption associatedwithtransportingmaterials to factory site per window (Weir, 1998) Total energy
Material
Bauxite Aluminium Timber Timber

Source location Argentina Norway Norway
Sweden

Average distance
(km)

Modeof
transport Ship Road Rail Rail Rail
Ship

Material mass per window
(kg)
14.7

consumption

per window
(Mi)
52.5 0.1

12 000
80
100

2.3
11.1 14.8 11.1

200

2.6 2.2
3.2 3.8

Timber Glass Components Total

Finland UK Norway

400 600
100

21.2
1.6

Road

0.4
64.8

WINDOWS AND THEIR LIFE CYCLE

217

shipping. Orders are shipped to Aberdeen once per week. Aberdeen is used as

the shippingpoint for thewhole ofthe UK. Unitsare again stored before being transported, by road, to site. The number of handling exercises involved with this transport system means that breakages are a more common occurrence.It is not cost effective to transportwindowsin individual units, but thefunctional unitof this LCA is measured in J/window, and to this end, on a mass basis, the transportation analysis is the same. For example, a tilt and turn window weighing37.5 kg andtransported from the factory to an Edinburgh city centre location consumes fuel energy as follows:
80 km by road, Moi—Stavanger 525 km by ship Stavanger—Aberdeen 200 km by road Aberdeen—Edinburgh Total 0.304 MJ/kg 0.03 MJ/kg 0.76 MJ/kg
11.4 MJ 1.1 MJ

28.5 MJ 41.0 MJ

The total energy consumption per window unit for transportation of raw materials and finishing is therefore estimated to be 106 MJ.
7.3.4 Use

In any LCA, certain phases of the life cycle will be highlighted as having a greater or lesser impact on the environment than others. Raw material sour-

cing, manufacturing, transportation and recycling/disposalof most products have the greatest impact on the environment, with little or no consideration being required for theirgeneral use. Windowsdiffer greatly in theirLCA. Their use phase impacts heavily on the environment. Window performance affects the indoor environment, building services operation, energy consumption of buildings and the well-being of occupants. Thermal insulating properties impact upon indoor air temperatures, draught sensation, radiant temperatures, and building user comfort, influencingenergy use for heating and ventilation. No analysis would be complete without designingto optimisethe well-beingof occupants, while simultaneously minimising environmental impact generated from the life cycle. If the well-being of building occupants is not at the forefront ofdesign criteria, the building, thoughwell designed in terms of thermal performance, aesthetics and light transmission, will be unsatisfactory to its occupants. Work performance is heavily influenced by the working environment. Energy consumption used for space heating over a lifetime of use is calculated for a number ofwindow constructions. To consider the energy analysis of a window in isolation of the building of which it is a part would be both misleading and of limited use. For this reason the example office, having dimensions and layout as shown in Figure 7.2.1, is used to compare the space heating energy requirement associated with each of 18 windows used in construction. The following assumptions are made regarding the office environment, location, orientation and construction:

218

WINDOWS IN BUILDINGS

• The office environment is maintained at an environmental temperature

which is high enough to meet thermal comfort needs, yet low enough to inhibit the early onset of fatigue in workers. It is therefore reasonable to assume that undernormal operating conditions, a setpoint temperature of 19°C is maintained, using building controls and/or energy management
systems. office location and orientation, wall construction, and window sash and

• Each window configuration is installed within a typical office setup, with
permutations. • The example office remaining is on an intermediate floor of a building and has one frame construction

constantacross all

external wall. It is surrounded above, below and on all internal sides by spaces at the same temperature.

It is assumed that the space is heated throughout the heating season (1 October to 30 April in the northernhemisphere),over the hours 8 a.m. to 6 p.m., the windows are openable during summer months to provide natural ventilation,and that no artificialcooling is provided. The analysis is performed over a period ofoneyear using a typical oneyear sample ofweather data. The analysis is based on a glazedarea equivalent to 60% ofthe external facade requiring 18 example-sizedwindows to cover this area. Annual space heating consumption for the office space, using each of 18 window constructions over the facade is illustrated in Figure 7.3.4. Refer to Table 7.3.8for adefinition ofwindow types.
25

20

0
C

0 0 0
w

0. E

1

() C
0) C

Ii ,,,,, ,/g =!=Iii.iiiiii J• 2'
. . .

4

=

b

b

Windowtype

Figure 7.3.4 Annual space heating energy consumption for example office for a glazed facade proportion of 60% (Weir, 1998). See Table 7.3.8 for windowtype definitions

WINDOWS AND THEIR LIFE CYCLE Table 7.3.8 Definitionof window types used in this chapter (Weir, 1998) Window type Single4 mm Single 6 mm Single 8 mm Single 10 mm
4-20air-4

219

Definition Singleglazed,4 mm thick glass Singleglazed,6 mm thick glass Singleglazed, 8 mm thick glass Singleglazed, 10 mm thick glass Double glazed,4 mm thick glass, 20 mm air cavity Double glazed,4 mm thick glass, 16 mm argon cavity Double glazed,4 mm thick glass, 12 mm krypton cavity Double glazed,4 mm thick glass, 8 mm xenon cavity Double glazed,4 mm thick glass, 20 mm air cavity,

4-l6Ar-4 4-l2Kr-4
4-8Xe-4 4e-20air-4 4e-20air-e4

one low-emissivity
4e- 16Ar-4 4e-l6Ar-e4 4e-l2Kr-e4
4e-8Xe-e4 4e-20air-4-20air-e4

Double glazed,4 mm thick glass, 20 mm air cavity, two low-emissivity coatings Double glazed,4 mm thick glass, 16 mm argon cavity,

coating

one low-emissivity coating Double glazed,4 mm thick glass, 16 mm argon cavity,

4e-l6Ar-4-l6Ar-e4 4e-l2Kr-4-l2Kr-e4
4e-8Xe-4-8Xe-e4

two low-emissivity coatings Double glazed,4 mm thick glass, 12 mm krypton cavity, two low-emissivity coatings Double glazed, 4 mm thick glass, 8 mm xenon cavity, two low-emissivity coatings Triple glazed, 4 mm thickglass, two 20 mm air cavities, two low-emissivity coatings Triple glazed, 4 mm thickglass, two 16 mm argon cavities, two low-emissivity coatings Triple glazed, 4 mm thickglass, two 12 mm krypton cavities, two low-emissivity coatings Triple glazed,4 mm thickglass, two 8 mm xenon cavities, two low-emissivity coatings

Like the thermal characteristics of windows, the visual performance is important to buildinguser satisfaction andinfluencestheelectricityconsumption used for artificial lighting. Good lighting in a building may be defined as the provision of 'adequate' light, in the right place, at the right time, enabling occupants to perform tasks in comfort, to a high degree of efficiency, without suffering eyestrain or fatigue. Guidelines have been in place for many years to aid designers in the provision of sufficient lighting. What is less well documented is the definition of 'adequate' lighting. Use of appropriate luminaires,

lighting control strategies,energy-savingschemes,exploitation ofdaylight, and the impact which environments have upon building occupants must be considered in the standard of service provided. Windows which permit light to enter an office space in a manner which is comfortable and satisfactory to its occupants, and which help create environments that permit maximum work productivity to be achieved, are sought. Energy consumption used for electric lighting, based on a top-up lighting control strategy, for each window construction is calculated. To do this re-

onto the working plane within a building structure.5 Annual artificial lightingenergyconsumption for a glazed facade proportion of60%and a top-uplightingcontrolstrategy(Weir. A number of assumptions are made with regard to lighting control strategies within the office environment: quires that • Occupants do not override the lighting controls. • Dimming control is linear in nature and does not employ step functions. 0 E C 0 0 >' 0) )1) w C I Ct. 1998). See Table 7. Window type Figure 7. Annual energy consumption on artificial lighting for each of the 18 window constructions is illustrated in Figure 7.5.8 for window type definitions . is assessed. as discussed throughout Chapter 5. • Building occupants do not object to gradation in daylight/artificial light throughout the passage of a day. based again on a glazed area covering 60% of the external facade. C 0 U.3.3. The analysis is performed over a period of one year.220 WINDOWS IN BUILDINGS the light transmitted through windows. • Both lamps and control units are suitably maintained to ensure good performance.The extentto which light is transmitted through windowsis a function ofthe optical properties of the material used.3.

paperless offices. in addition to environmentally sensitive disposal of waste. energy generation and environmental emissions. It was found that around 96% by volume of the waste generated from the demolition of an existing building was reusable or recyclable. recycling and use of waste to generate energy. After more thanfive decades ofprofligateuse of resources and badly managed waste disposal. newspaper reclaims and other initiatives make the public more aware of recyclingpossibilities. enabling consumers to make more informed decisions about disposal and recycling of household and commercial waste.3. However. the largest proportion of embodied energy for a given productis predominantly in material extraction and processing. .WINDOWS AND THEIR LIFE CYCLE 221 7. The British Building Research Establishment (BRE) carried out a recent research programme to demonstrate the re-use and recyclingof materials (Hobbs and Collins. However. A Best Practicable Environmental Option (BPEO) that imposes least damage on the environment as a whole is sought. commercial and industrial wastes end up in landfillsites (Jones.Only a small percentage (4%)was sent to landfill sites. re-use. Bottle banks. The need to plan for disposal and recycling in the product conception phase is emphasised. In the construction industry.5 Disposaland recycling Waste handling and management must also be considered in a thorough LCA.corporations and industries have a responsibility too. was found to be of the order of 1500 MJ (excluding choice of cavity gas and lowemissivitycoatings). and place each community member in a positionof responsibility. Re-use. During the past decades there has been considerable interest in recycling products and materials. and remains the only sustainable solution in the long term. Embodied energy ofmaterials for a double-glazedwindow. 1993).Re-using materials omits this embodied energy in new constructions. When a product reaches the end of its useful life it is tempting to focus attention on its successor.littleprogress has been made in the practical implementation of research findings and government recommendations. textile re-use. of window units would therefore show significant benefitsin terms of material sourcing. Re-use and recycling of demolition and construction waste is no exception. approximately 85% of total energy consumption used in extracting raw materials. form the framework of sustainable development.or to forget that the productrequires disposing ofand! or recycling.Beyond the issue ofpersonal responsibility. Ninety five per cent of household. Public-awareness campaigns have raised the issue. while recyclingof materials significantly reduces environmental impact. attentionhas been turned to reduction of waste at source. which is heavily focused on sustainable developmentsolutions (Scottish Office. 1996). The Scottish Office Development Department produced a planning policy guideline on waste management in March 1996. with the emphasis on reducing waste at source. Research is ongoing and coincides well with the aims of LCA. manufacturing and transportation. These four management solutions. in this LCA. or recycling. This equates to 8 tonnes of waste per person annually. 1997).

and all energy used for artificial lighting. an iterative process. manufacture of finished window units. however.3. in combination with the trend towards more open disclosure of environmental information by companies.6.4. and the desire by consumers to be guided towards the least harmful products. packing and transportation of both materials and finished products is illustrated in Figure 7.5. It is a question of defining and characterising the consequenceswhich result from the inputs and outputs quantified in theinventory analysis. 7. effluent output. and development work to find reliable. fossil fuels and nuclear power. presented in Figures 7.222 WINDOWS IN BUILDINGS no on-goingconstruction activity could be found to support this initiative to date.3.4 Impact assessment requires a qualitative and quantitative approach to analyse how raw material Impact assessment focuses on how the product affects the environment. relative to the average UK plant mix. The energycontent ofmaterial requirementsandmanufacturing processesis defined above in the inventory analysis.The total energy consumption throughout the life cycle (20 years) of each of 18 window constructions is illustrated in Figure 7. To begin analysing thegreenhouse gases emitted due to the burning of fossil fuels. this summary of inventory analysis data forms the basis of the impact assessmentwhich follows. Used in conjunction with the energy consumption calculated for space heating and artificial lighting.1 show the emissions created for each MJ of electrical energy generated. water production. air emissions and solid waste affect the environment. and as data quality improves. Each unit of energy consumed in developingmaterials. It is assumed that all energy used to produce raw materials and to manufacture them into finished components or window units. is electrical energy. and in turning them into finished window products and in providing adequate space heating and artificial lighting.7. generated using a mix of renewable energy. energy generation.3. LCA is.4 and 7. as outlined in the introductory paragraphs. This use. The quantitative analysis does not include the disposal and recycling of materials due to a lack of quality data relating to this. and those emissions generated from thecombustion ofnatural gas. The data shown in Table 7. TheWorld ResourceFoundation(1995) considers LCA to be a vital. A summary of all energy requirements involved in the extraction of materials. . on-going tool. All space heating energy requirements are assumed to be provided for by the combustion of natural gas in conventional central heating networks. economic markets for recycled materials is in its infancy. Transportation ofwaste to processingcentres remains a stumbling block to recycling initiatives.3.these elements may be included in the analysis with confidence. has its own implications for greenhouse gas production and the environment. an analysis of the methods of energy production must be made.

2 m by 1. 1998) The UK plant mix seems to be in a state of change.WINDOWS AND THEIR LIFE CYCLE 223 Figure 7. 21% naturalgas. and 32% nuclear (DTI. 1998). and 28. However. 1997. 2.2 m (Weir. during 1998 the above mix had changed to 39% coal. 27% natural gas.3. .5% oil. 2% oil. In 1997 the electricity supply mix was 42% coal.6 Summaryofenergyconsumption per finished window unit(raw materialextraction to site arrival) for a windowmeasuring 1.5% nuclear.

1998).4.1 Transportation of materials. Where transportation Table 7. See Table 7. and consideration should be given to the most appropriateroute. finished products. there are several environmental impacts to be considered. Reduced air pollution and emissions result directly from maximising vehicle capacity. basedon a top-up lightingcontrolstrategy(Weir.3.2 kg 0.3.058 kg SO2 2. waste or recycled resources is necessary.694 g CO.. E '0 C 4 . 0.02 mg 1.4.945 g 4. WINDOWS IN BUILDINGS — — — 40C — 0 C 35C 0 0.778 g — NO. 1998) USA emissions CO2 Electricity generation (UK Plant mix) 0.7 Total energyconsumption over windowlife for a glazed facade of 60%. and in back-hauling materials.224 50C 4b(.60g Natural gas combustion . eliminating empty journeys.8 for window type definitions 7.189 kg SO2 UK emissions NO. and from burning natural gas (Weir. 0. Ensuring that products areproperly contained reduces therisk ofwaste due to breakages on ajourney.1 Emissionsgeneratedas a result ofproducing 1 MJ ofelectricityin the USA and UK. 0.• 'i" I II U Materials and manufacture Windowtype Space heating /1/ 0 Electric lighting proportion Figure 7.

3 Greenhouse gas and hydrocarbon emissions generated as transportation of materials and finished window units (Weir.05 0.7 10. Research work by Buchanan and Honey (1994) was used to evaluate emissions generated for various modes of transport on a distance and load basis.40 a result of CO (kg) 28. Table 7.08 0.0 1.4. again related to our example case where 60% of the example office facade is glazed and a top-up lighting control strategy is adopted.03 3.07 45 2.02 4.4 0.4. Weir (1998) has presented a case for the importance of transportation in a range of LCA studies.1 5. drawing a number ofconclusions.13 0.28 0. 1998) Material Transported to factory Transported to factory Timber Transported to factory Glass Transported to factory Components Transported to factory Finishedwindow Transported to site units Total Bauxite Aluminium CO2 (kg) 472.3.6 2.carbon monoxide.03 2.2 3.30 Hydrocarbons NOx (kg) 45. rail and ship transport.5x103 2. the generation ofgreenhousegas emissions.6 5.03 4 . For the distances andloads listed above.1 3. These Table 7.28 0.9x103 0. the greenhouse gas emissions generated over a lifetime of use are illustrated in Figure 7.2 Transportationemissions per kg load.2 Table 7.05 0.6x104 0. The results are illustrated in Table 7. The major impact is easily identified as carbon dioxide.WINDOWS AND THEIR LIFE CYCLE 225 in addition to the locating ofmanufacturing and production plants in relation to market places.as a result of transportation of materials and finished window units.64 477 1.01 0.0 2.16 29 2.1 NOx (MG) 2.4 0.2 1. Adding the impact associated with electrical energy generation.00 (kg) 3.nitrousoxides.perkm travelled(Buchananand Honey.05 1.4. 1994) Mode oftransport CO2 (G) CO (MG) 16. the combustion of natural gas and transportation requirements throughout the life cycle of a multi-glazed window unit.4.6 0.1.03 2.1xl03 0.2 shows the quantities ofcarbon dioxide. but other greenhousegases are also emitted as pollution. and hydrocarbons which result from road.7 Hydrocarbons (MG) 2.8x104 0.6x103 0.20 0.500 3.7 16. was evaluated.7 1.4. whereas contributions with more than 10% occurred regularly. In none ofthe LCAs did transport contribute less than 5% of the energy-related interventions or impacts.0x104 0.34 0.2 Road 30 km Road 50 km Road 100km Rail 200 km Rail 500 km Ship 0.17 0.

If windows were selected according to their life cycle energy use or total environmental burden./I' 0 .8 for windowtype definitions other gasesinclude sulphur dioxide (SO2). Using the information presented in Figures 7. 1998).4.4. See Table 7. It is seen that an energy saving of more than 150 GJ can be made over a lifetime of use by selecting a triple-glazed . then this presents an improvement. carbon monoxide (CO) and hydrocarbons. The conditions fixed during the introductory stages of an assessment set the limits for possible improvements. This requires taking an objective view of a product life cycle and assessing the impact which changes would make on the environment. 7.1 Total greenhouse gas emissions over windowlife for a glazed facade proportionof60%. basedon a top-uplightingcontrol strategy(Weir.5 Improvement analysis Improvement analysis involves decision making to reduce environmental burdens. The conclusion of any analytical study is borne out of underlying scope and goal definitions. or thermal performance properties. rather than purely for aesthetics.226 70 WINDOWS IN BUILDINGS 60 50 40 30 20 10 //f/Y/// //1//I/1.3.7 and 7.1 as a guide towards selectingand designingwindows which use reduced quantities ofenergy throughout their life cycle provides one source of improvement. .The outcome ofan LCA study is no different. •th b Windowtype COproduction • Otherpollutants Figure 7. nitrous oxides (NO).3.

proactive and strategic response to necessary and sustainable environmental initiatives. the aims ofthe improvement analysis should focus on meeting therequirements ofthe present without compromisingthe ability of future generations to meet their own needs. Improvement analyses which aim merely to improve environmental accountability with approaches which are not sustainable do not display sufficient initiative.1 Products Extending the useful life of a productmeans that fewer units are required over 1987).pragmatic assertions. rather than providing definitive solutions. aims to answer all possiblequestions arising. To this end. expressedas 'an approach to progress which meets the needs of the present withoutcompromising the ability of future generations to meet their own needs' (WCED. The following recommendations. which actually possesses better thermal transmission properties. The energy savings calculated in the example presented are for a specific building construction. poses questions to which the answers demand further research activity. Product design changes. Much ofthe followingdiscussion. improved waste management facilities. questions whichmaybe addressed in a second. rather than a similar window having xenon gas cavities (4e8Xe-4-SXe-e4). and in some instances merely providesthe researcher with the 'right' questions to ask. or fourth iterationof research techniques. initsgeneric form. however. could be made as part of a long-term. LCAis a commitment to on-goingimprovement. be achieved. based solely on the consequences upon human interest in the short term. relating to the product.No LCA. or suggested consumer use changes. orientation and use. seek to outline future improvements which may be practically and realisticallyimplemented. third. The most common definition of sustainable developmentstems initially from the Brundtland Report. manufacturing process changes.WINDOWS AND THEIR LIFE CYCLE 227 window with two low-emissivity glazingcoatingsand argon gas cavities (4e16Ar-4-16Ar-e4). remain unresolved. Little room is allowed for defensive.5. materials and management issues. 7. a given period of time to satisfy the same consumer need. Single-glazed windows do not possess sufficient thermal insulation properties. An unrestrained improvement analysis leaves too much scope for individual interpretation and poor judgement values. This is because the energy associated with producing pure xenon gas is so great that the benefit of reduced space heating requirements can never be gained. Environmental sustainability provides a benchmark on which to base criteria for life cycle improvement analysis. The useful life of a .highlightingproblems which. while windows incorporating xenonfilled cavities are too energy intensive in their manufacture stage. but the same general principles would extend to many building types. as yet. a purposeful and challenging goal which is more rewarding than abstract and ambiguous attempts to 'lessen environmental burdens'. raw material substitutions. Further improvements to the LCA results presented could.

5.5. and will need to be replaced less frequently.1.228 WINDOWS IN BUILDINGS product can be extendedin a number of ways. stress and environmental degradation over long life cycles.5 Repairs Products which allow the replacement of dysfunctional parts with ease. 7.1. 7. Adaptable products 7.2 Adaptability make allowances for continual improvement and upbe usedfor more than one function. employs of resources but poses many questions about collection.5. Windows with high-quality seals.5.5. of a product or part. The quality of finish and wood treatments applied throughout the manufacture stages are critical in creating windows which are durable. but standardisation of components would ensure easy maintenance and longer life cycles. in an environment for which they were designed. This is less applicable grading and/or may to windows. limiting down-time. components and finishes will remain reliable for many decades. trans- . processing and waste.1. all ofwhich are dependent upon design criteria and quality of manufacture. once retired from a clearly defined duty.1. Durable items 7.1. Reliable products 7.3 Reliability are able to meet performance criteria for a defined period of time. Windows are one building component which lend themselves well to re-use.1 Durability can withstand wear.6 Re-use Re-use efficient use portation. and have better longevity. speed the return of systemsto normal operating conditions.1.5. This has been demonstrated by Hobbs and Collins (1997). without failing. Again. and having little impact on operating efficiency. standardisation of window components makes repairs quicker and simpler.4 Maintenance Windows which are easily and practically maintained reduce the time required for repairs and preventative work. 7.

7.2 Recycling Complete products or component parts may be regenerated in one of two ways.prior to manufacturing.7 Remanufacture Remanufacture of parts is an industrial process which restores worn units to as-new condition. is often the dominant source of environmental burdens in an LCA.5.5. There are several means by which the useful life of materials can be extended. Again this would increase the longevity of windows requiring fewer replacements over building life cycle. Additionally. Closed-loop recyclingrecoversmaterials that are suitable substitutes for virgin materials. but having capital equipment that may be adapted for a number of working materials enables a wider choice of markets to be sourced and more informed decisionsabout the environmental impact of materials to be made. 7.WINDOWS AND THEIR LIFE CYCLE 229 7.5. management issues are central to enprovement analysis vironmental improvements. Product design factors .2.2. either directly or indirectly via fuel extraction and combustion to generate energy. Open-loop recycling recovers materials a finite number of times before final disposal.5.3 Management issues and material selection issues are important to imwork.3 Reduced material intensiveness of material resources can result in less waste and reduced environmental impact. Energy embodied in material sourcing and processing. Conservation 7. Timber.2. and which are used to manufacture the same product or component again. and therefore requires less packaging and reduced transportation energy.5. and associated efficiency with which processes are carried out. glass and aluminium components within the window structurelend themselveswell to recyclinginitiatives.1. a 7. The planned use of materials and energy.1 Substitution Material substitutionmay be restricted by the need to use existing operating plant.5. impact heavily on LCA results. 7.2 Materials Extending the useful life of materials used to manufacture products and components can have a significant effect on the outcome of an LCA study. A material which is manufactured from smaller quantities ofmaterial is likely to be lighter in weight.

Application of the best available technology limits material waste and inefficiencies in product and resource handling.5 Facilities management Flexible manufacturing principles can extend the useful life of facilities and equipment.3.2 Process control Control processes which suppress influence of external disturbances and the which ensure goodperformance stability should be adopted where practicable.3. 7.3.3. 7. re-use. 7. waste or recycled resources is necessary.5. 7. Packaging and products should be designed to complement each other. Wherever possibleheat losses from one system should be used to perform work in other processes.4 Inventory handling and Appropriatelycontrolledresource material handling wherebyoverstocking eliminated limits waste due to spills and deterioration. 7.7 Transportation of materials. a number of considerations are offered.such that output is consistent and efficient. adoptingthe best available technology. the use of packaging should be eliminated.3. Consideration should also be given to the layout ofprocesses. productchanges and use ofreduced material intensivenessapply to packaging as well as finished products. and wherever practically possible. is 7. finished products.5.3.5. The provision of services should also be energy efficient.5. Processes which create major environmental burdens should be replaced by more benign ones. Reduced air pollution and emissions result directly from maximisingvehicle capacity and in Where transportation .6 Packaging The same principlesofmaterial substitution.5. and accident risks are reduced to a minimum.5.230 WINDOWS IN BUILDINGS 7.5. recycling.1 Process substitution Designers should be aware of the best available technology to perform a manufacturing requirement.3 Energy efficiency Use ofenergy throughout the production process should be viewed holistically.3.

Ensuring that products areproperly contained reduces the risk ofwaste due to breakages on ajourney. and consideration should be given as to the most appropriate route. eliminating empty journeys. A general trend is seen throughout the assessment of energy consumption and global environment impact.9 Product labelling Once a full LCA has been carriedout. activities and a selection of materials further up the product development and component manufacturing tree can have a larger detrimental effect upon local and global environments than the final product itself. double. product labellingprovides the consumer and retailer with information which allows them to make an informed choice.3. in Figure 7.3.5. and having air. the optimal window construction. 7.or triple-glazed construction.8 Selection ofsuppliers Selection of environmentally responsible suppliers can significantly influence the environmental impact of product's life cycle. These windows are generally of double.2. thepresent analysis is subject to uncertaintiesandtherefore it may only be used to indicate trends. The example LCA shown in this chapter is not detailed enough to make a definitiveselection.WINDOWS AND THEIR LIFE CYCLE 231 back-hauling materials. Provision of information on window U-values and the embodied energy of materials and manufacture would allow window specifiers and designers to make informed decisions regarding energy consumption and building performance. Furthermore. 7. however. relating to material selection and window . in terms of energy consumption and greenhouse gas emissions was found to be: 4e-l6Ar-4-l6Ar-e4 triple-glazedwindow with two 16 mm cavities. and for the example building shown in Figure 7. It can be seen.5. and two low-emissivity coatings. 7. filled with argon gas. and the window life is assumed to be 20 years.1.7 that a number of window constructions consume very similar quantities of energy throughout theirlifecycle. where the window area is 60% of the south-facingfacade.6 Selection of the 'optimum' window type For the range of single-. but does indicate that a number of window constructions consume smaller quantities of energy throughout their life cycle. argon or krypton gas-filled cavities. incorporating low-emissivity glazing coatings.3.and triple-glazed window options analysed. Processes.

Research Report 91-3. Callander. (1996) Monetary. J.R. 129. London. . as the impact ofusing materials with higher energycontents is offset by reduced requirements for space heating and artificial lighting energy consumption.232 design.Windows designed for longevity should possessgoodthermalinsulating and daylight transmission properties. The Stationery Office. (1995) Scientificaspects ofthe framework convention on climatic and Assessment 38.A. 12. University of Canterbury. WINDOWS IN BUILDINGS Windows which possess poor thermal insulating properties are characterised by a predominant energy requirement for space heating over window life. Danish Building Research Institute. T. Windows which provide superior thermal insulating properties by use of highly energy intensive materials in construction are characterised by a predominant energy requirement in material sourcing and manufacture. DTI (Department of Trade and Industry) (1998) Digest of United Kingdom Energy Statistics. D. Dinesen. (1994) An energy life cycle assessment model for building design. and Hunt. G. 43.A. B. M. (1994) Life-cycle analysis of automobiles:a critical review ofmethodologies. CIB DTI (Department of Trade and Industry) (1997) Digest of United Kingdom Energy Statistics. change and national greenhouse gas inventories. The Stationery Office. Department of Civil Engineering. and Muneer. Department of Civil Engineering. energyandenvironmental cost of infill gases for double glazings. Hollinger.. and Clark. 337. and Buchanan. EnvironmentalMonitoring Honey. University of Canterbury. F. (1994)Energy and carbon dioxideimplications of building construction. JournaloftheRoyal Society. 1P3/97. Building Services EngineeringResearch and Field III. (1992) Environmental impacts of the New Zealand building industry. Proceedings of the 1st International Conferenceon Buildings andthe Environment. Isaacs. 205. London. Energy and Buildings 20. Research Report 92-2. B. and Trabergy-Borup.P. (1991) Feasibility of using timber for medium rise office structures. J. J. and Collins. Technology 17. 16—20 May 1994.NewZealand20. Building Research Establishment. R. B. D. and Honey. Halliday. B. T. (1990) Anthropogenis emissions ofcarbon dioxide and methane in New Zealand. References Buchanan. Journal ofMinerals. (1997) Demonstration of reuse and recycling of materials: BRE energy efficient office of the future. Design life is also seen to be critical to window selection. Metals and Materials Society 46. J. Fernie. Hobbs.

Weir. Warmer Bulletin. and Muneer. monetary and environmental costs. Proceedings of the First International ConferenceofBuildings and the Environment. Vienna.WINDOWS AND THEIR LIFE CYCLE Jones. London. Liggett. UNIDO (1989) United Nations Industrial Development Organisation. (1997) Low-emissivitycoatings in high-performance double-glazed windows: Energy.CIB 16—20 May 1994. IMechE. W. Young.T. Our Common Future. P. SB. Metals and Materials 46. J. (1996)EnergyandEmissions Square Metre ofPilkingtonK Glass. G. and Howard. Weir. Zimbabwe. and Shukuya. World Resource Foundation (1995) Information sheet.H. Industry and Development — Global Report 1989/90. Scottish Office (1996) National planning policy guideline. M. Journal ofMinerals. P. Building Services EngineeringResearch and Technology 18. Napier University. (1994) Embodied energy and carbon dioxide emissions for building materials. M. Oxford Press. . PhD thesis. processing and recycling. Waste: Handling. T. J. ISES '95 Congress. per Pilkington Glass Products. G. Suffolk. 4. Harare. 125. Mechanical Engineering Publications Ltd. West. and Vanderburg. WCED (1987) World Commission on Environment and Development. Saito. March. (1993) Underlying 233 Trendsin the UK Waste Sector.. NPPG 10: Planning andwaste management. (1995) Energy and material use in the production of architectural windows for passive solar heating. Atkinson. (1995) Applying environmental life cycle analysis to materials. N. 1996. August 1995. (1998) L4fe Cycle Assessment of Multi-glazed Windows.

Page blank in original .

Professor P. horizontal and vertical illuminance and irradiance (solar radiation) measurements have been carried out for four sites. MostIDMP data are quality controlled in accordance with the rules set out by CIE (1994). As a result.html.satisfy a long-standing need. Garston. i.http://idmp. hopefully. Manchester. In Japan daylight data were recorded at 14 stations.entpe. The slope data arepresented herein Excelformat.fr/ and http://www.1. 1998). A selection of data is included in the CD-ROM accompanying this publication.cie. Tregenza as part of an EPSRCfunded project (Tregenza. the reader is referred to the followingwebsites.8 SOLAR RADIATION AND DAYLIGHT DATA 8. new activity commencedin the field ofdaylight measurement on a world-widescale. 1995). The measurements at Edinburgh were respectively undertaken at Napier and Heriot-Watt universities. other than that supplied be required. Sheffield and Edinburgh.e. Since the data gathered from the IDMP were reported by research parties from a variety of world-wide monitoring stations.co. This activity was later named the International Daylight Measurement Programme (IDMP). The purpose of collating the measured solar data is to provide a qualitycontrolled daylight and solar radiation database which will. In the UK as a result of the CIE call. Japanese data were checked for quality by the research team headed by Professor H.1 International daylight measurement programme The Commission Internationalde 1'Eclairage(CIE) declared the year 1991 as the International Daylight Measurement Year (IDMY). R. Nakamura at Kyushu University. This information is shown in Table 8.initially by the respective station managers and then by the overall co-ordinator.at/cie/home. 8. However. The UK quality-controlled data provided to the authors have been rigorously checked. should more extensive data. Thedata . the entire solar irradiance and illuminancedataset could not be included with this book. Measured solar irradiance and illuminance IDMP data for the UK and Japan have been compiled on a CD-ROM and duly reported (Muneer and Kinghorn. Muneer (1997) has presented details of IDMP stations throughout the world.2 Slope data Daylightis a highly variable resource and therefore warrants measurement at a shorttime interval.1.

20°N 40.52°N 51.82°E 139.50°E 141.OO°N G G R S SR R R R S R 36.92°E 7.40°N G R R *Station classificationas follows: R: researchclass SR: simplified researchclass G: general class 5: simplified general class .90°E 135.60°W Garston Albany Ann Arbor Cape Canaveral 42.67°N 67.30°W 2.78°E 136.05°N 34.27°N 55.83°N 35.33°N Fukuoka Garston Kiyose 35.38°W 73.53°E 135.38°E 139.27°N 28.15°E 9.50°W O.58°N Longitude 151.77°N 35.07°W I16.l5°N 48.75°E 4.57°N 45.70°N 59.97°E 136.48°E O.50°N 60.12°N Kyoto Nagoya (Daido) Nagoya (Meijo) Osaka Sapporo Suita 3507°N 34.95°N 55.57°E 30.33°E 135.67°N 35.55°N 37.85°N 58.15°N G G S R S Tokyo Toyota Tsukaba Uozu Eindhoven Lisbon Moscow Voeikovo R G R 14005°E 137.02°N 35.60°N Ukraine United Kingdom United Kingdom United Kingdom United Kingdom United Kingdom USA USA USA Kiruna Norrkoping Geneva Karadag Edinburgh (Napier) Edinburgh (Heriot-Watt) Manchester Sheffield 17.lO°E G S G S 46.87°S 51.50°E 139.63°N 33.80°N 29.92°E 5.85°W 83.15°E 49. 1997) Country Australia Canada China China France France France France France Germany Germany Greece Location Sydney Latitude 33.58°N 43.20°W 3.57°N 45.47°E 5.71°N G G G R R l.97°E 23.93°N 53.17°N 36.38°N 51.12°N 47.13°E 37.38°E 3678°N 51.70°E 103.72°E Type* R Calgary Beijing Chongquing G R G S S Chanbery Grenoble Nantes Strasbourg Vaulx en Velin Freiburg Hamburg Athens 45.80°N 53.70°N 42.27°E G G G G R R S 16 12°E 6.55°E 130.72°W 80.05°N 39.77°N l.90°E 34.236 WINDOWS IN BUILDINGS Table 8.78°E Singapore Sweden Sweden Sweden Switzerland Singapore Gavle l.38°W 139.50°E 106.32°W 7.82°E 137.97°N India Israel Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Japan Netherlands Portugal Russia Russia Roorkee Bet Dagan Ashikaga Chofu 29.1 World-wide IDMP stations (Muneer.42°N 38.50°N 53.85°N 32.l2°E 20.75°N 52.72°E 77.25°W 46.71°N 35.1.57°E 3.27°E 114.90°N G R G S 547°E 9.

Napier (55.50°W).2. Further. data for the nearest period to the mean day have been selected.48°E) data are presented as 10-minuteaveragesof 1-minutespot readings.1 Illuminancethresholds(lx) 237 Sun at zenith Clear sky. Table 8.20°W) Universities and Garston (51. Edinburgh. undersunlight (10 h solar time) Clear sky. This approach provides a simple.215 included varies from a frequency of 5 minutes for three UK stations. method for determination of monthly average radiation values required in a multitude of solar energy engineering applications.2.2.2. Since these data are comparatively more scarce than.3.3—8. Table 8.20°W) and Heriot-Watt (55. format and structure of the files provided on the CD-ROM. Fukuoka (33. 2.1 provides approximate threshold values of illuminance. Tables 8. Sky luminance and radiance distribution data from one Japanese and two UK . An attempt has been made to supply these data using the mean day of each month as a centre point for each weekly period. 3.95°N. 8.50°N. i. Approximately one week of data per month is included for each site considered. 3. Garston and Fukuoka. A list ofthe electronicdata files is provided in the front section of the book. A summary of the UK and Japanese slope data included in the accompanying CD-ROM is presented in Table 8. yet reliable.25°W) and Sheffield (53.2.SOLARRADIATION AND DAYLIGHT DATA Table 8.e. 130. undersunlight (15 h solar time) Sun at horizon Flashlight at 1 m distance Streetlightsat night Candlelightat 0. 1. In accordance with the work of Klein (1977).001 0.38°N.2 m distance End of civil twilight End of astronomical twilight Full moon at zenith 103 000 65 000 35 000 355 250 175 10 4. This information may be useful in acquiring a sense of the prevailing daylight illuminance levels for the locations under discussion.3 0.3 Sky scan data sites are provided on the accompanyingCD-ROM.1 presents a summary of radiance and luminance data recorded by sky scanners at Sheffield.71°N.2.4 respectivelyprovide details of the notation.93°N.38°W) to half-hourly time series for Manchester (53.52°N. the mean days of each month are reported as those for which daily horizontal extraterrestrial irradiance values are approximately equal to the mean monthly values. In instances where no measured data are available for the mean day of each month. 0.

1991*) Data8-02.xls Julian days 15—21 17(17) 15—21 13—19 16—22 11—15 3—7 No data available 1—5 No data available 13—19 13—19 16 (47) 16 (75) 15 (105) 1—5 No data available 13—19 9—15 15—21 19—23 14—20 January February March April May 9—15 14—20 13—19 13—19 15 (135) June 14—20 13—19 13—19 9—15 15—21 July 11(162) 17 (198) August 1—5 2—6 5—9 1—5 September 1—5 14—20 13—19 1—5 1—5 No data available No data available No data available 14_20* 13_19* 12_18* October November December 12_18* l4_20* 16 (228) 15 (258) 15 (288) 14 (318) 16 (344) 1—5 19—25 45—51 73—79 103—109 133—139 162—168 196—202 225—231 256—262 288—294 314—320 348—354 No data available 45—51 73—79 103—109 133—139 162—168 196—202 225—231 256—262 288—294 314—320 348—354 *Indicates data substituted from another year to produce a complete year of information .2.xls Sheffield (1994) Month Dates Mean day (Julian day) Dates Dates 19—26 Data8-06.xls Dates 15—21 14—20 Fukuoka Manchester (1993) Data8-04.xls Napier (1993. 1992*) Data8-05.Table 8. 1993*) Data8-03.2 Slope illuminancedatabase Garston Henot-Watt (1994.xls (1994) Data8-01.xls Julian days (1992.

e.2.2. skyluminance distribution data arerespectivelycontained in files Data8-08 and Data8-09 for Garston and Sheffield.3 Format of slopeilluminance Column Edinburgh sites Month Day Year Hour Minute Evg 239 data files Garston Month Day Year Hour Minute Evg EVd Fukuoka Date Time SOLALT Manchester Year Julian day Hour Eeg Eed Eegn Eege Sheffield A B C Year Julian day Hour D E F G H SOLAZM Evg EVd EVb Evgn Eg Eed Eegn Eege Ed Evgn Evge Evgs I J Evgn Evge Evgs Ecg Eegw Eeg Eegw Evg K L E5 Evgw Eeg Evgw M N Eg Ed Eegn Eege Eegs E5 Eeg Eed Evg Ed Evg Evge Evgw Evg EVd Evgn Evge Evgs 0 P Eed Eeb Eedn Eede E5 Eege Eegs Eegw Evg RH Tdb Evgw Q S R T U V W X Y Egw SOLALT SOLAZM CefCvf Ed. the information contained on the accompanying CD-ROM comes in approximately 5-day blocks around the mean day of each month. of the sky.SOLARRADIATION AND DAYLIGHT DATA Table 8. capable of concurrent luminance and radiance recording. E. purposely developed by the EKO Company of Tokyo. The Japanese data were recorded using a twin-headed sensor. To ensure concurrency between comparative data. Eedw Tdb Tdp RH Tdb SOLALT SOLAZM L. Both instruments measure their required parameters by considering individual elements.xls and Data8-1O.2. Sky luminance and radiance distribution data recorded at Fukuoka are respectively contained in the files Data8-07. slope data. Modelling procedures may be used to synthetically generate Table 8. In the case of the UK data. The UK data were measured using a PRC Krochmann sky scanner.3. i.1 data from that given in Table 8.xls. 1993). Each sky patchcan be recognised in terms of its solar geometry. the sky scanner data are presented (as far as practicable) in accordance with the slope data measurement period. A comparative evaluation ofthese instruments has previously been conducted against IDMP data from Geneva. the altitudinal and azimuthal co-ordinates individual to . or patches. Switzerland (Ineichen and Molineaux. CefCvf RH Tdb Ee say.

xls Days Fukuoka(1994) Data8. west-facing surface Direct (beam normal) illuminance Diffuse horizontal illuminance Global horizontal illuminance Verticalilluminance.west-facing surface Zenith luminance Relative humidity Solaraltitude angle Solar azimuth angle Dry bulb temperature Dewpoint temperature Units Dimensionless Dimensionless W/rn2 W/rn2 W/rn2 W/m2 W/m2 W/m2 W/rn2 lx lx lx lx lx lx lx Cd/rn2 % Ce C. north-facingsurface Verticalirradiance.240 WINDOWS IN BUILDINGS Table 8.3. E5.ç Eb Eed Eeg Eege Eegn Eegs Eegw Eb Ed E5.east-facingsurface Vertical illuminance. south-facingsurface Verticalilluminance.1 Sky scan database Fukuoka (1994) Data8-07.xls Julian days 15—21 44—50 76—81 103—109 133—138 162—168 191—197 230—236 256—262 288—294 314—320 348—354 Radiance Sheffield (1993) Data8-09.. south-facingsurface Verticalirradiance. north-facing surface Vertical illuminance.xls Luminance Garston (1992) Data8-08.10. Evgw Evgs RH SOLALT SOLAZM Tdb TdP °C Table 8.xls Days 11—15 3—7 1—5 1—5 1—5 Month January February March April May June July Days 11—15 3—7 1—5 1—5 1—5 No data available No data available 14—20 14—17 18—19 2—6 5—9 1—5 1—5 1—5 1—5 1—5 August September No data available No data available No data available 17—19 15—21 18—24 20—26 2—6 5—9 1—5 1--S 1—S 1—5 1—5 October November December . east-facing surface Verticalirradiance.4 Nomenclatureused in slope illuminance data files CIE abbreviation Parameter Shade ring correction factorfor horizontal diffuse irradiance Shadering correction factorfor horizontal diffuse illuminance Direct (beam normal) irradiance Diffuse horizontal irradiance Global horizontal irradiance Verticalirradiance.2.

2 Sky patch geometryfor UK locations(sun-relative co-ordinatesystem)* 241 Patchno.6. Altitude (°) 1 2 3 4 5 6 6 12 12 6 132 Azimuth (°) Patchno. Altitude (°) 54 54 Azimuth (°) 40 121 60 122 80 123 Patch no. Altitude(°) 18 Azimuth (°) Patch no. Altitude (°) Azimuth (°) 240 61 30 0 71 30 150 81 252 62 30 15 72 30 165 82 30 315 92 42 105 264 63 30 30 73 30 180 276 64 30 45 74 30 195 84 30 345 94 42 135 104 83 30 300 91 30 330 93 288 65 30 60 75 30 210 85 42 0 95 300 66 30 75 76 30 225 86 312 67 42 15 96 42 90 101 42 120 103 42 270 113 42 150 105 42 165 106 102 42 240 111 42 255 112 54 42 285 114 54 42 300 115 42 315 116 30 90 77 30 240 87 42 30 97 42 180 107 42 330 117 54 324 68 30 105 336 69 30 120 348 70 30 135 80 30 285 78 30 255 88 42 45 98 79 30 270 89 42 60 99 42 210 109 90 42 75 100 42 195 42 225 110 108 42 345 118 54 0 119 54 20 120 54 Patchno. Altitude (°) Azimuth (°) Patchno. Altitude (°) Azimuth (°) Altitude (°) Azimuth (°) Patch no. Altitude (°) Azimuth(°) Patch no. Altitude (°) Azimuth (°) Patchno. Patch no. 0 41 18 120 51 18 12 24 43 18 42 18 132 36 44 18 156 48 45 18 168 60 46 18 180 72 47 18 192 84 96 48 18 49 18 216 59 18 50 18 Altitude (°) Azimuth (°) Patch no.3. Altitude (°) Azimuth(°) Patch no.SOLARRADIA TION AND DAYLIGHT DATA Table 8. Altitude(°) 52 18 144 53 18 54 18 55 18 56 18 57 18 204 58 18 228 60 18 Azimuth (°) Patchno. Altitude (°) Azimuth (°) Patchno.3(a) . Altitude (°) 0 11 6 24 13 6 144 6 36 14 6 156 6 48 15 6 168 25 6 6 60 16 6 180 7 6 8 9 10 72 17 6 192 6 84 18 6 204 28 6 96 19 6 216 29 6 108 20 6 228 30 6 120 21 6 240 31 22 6 252 32 18 23 6 264 33 18 24 6 276 34 18 26 6 288 35 18 6 300 36 18 27 6 312 37 18 6 324 38 18 6 336 39 18 6 348 40 18 108 Azimuth (°) Patch no. Altitude (°) Azimuth (°) *See 54 240 131 54 260 132 54 280 133 66 120 141 66 150 142 66 180 143 100 124 54 300 134 66 210 144 54 120 125 54 140 126 54 320 135 66 240 145 54 340 136 66 270 146 160 127 66 0 137 54 180 128 54 200 129 66 30 138 66 60 139 78 0 149 220 130 66 90 140 66 300 147 66 330 148 78 60 150 78 120 78 180 78 240 78 300 90 — 90 — 90 — 90 — 90 — 90 — Figure 5.

can be divided into bands of solar altitude.xls workbook. the sensor is elevated to an angle of 6°.xls and Data8-09. Rotatingclockwise. Though only 145 sky patches are considered by both sky scanners.09. Data for patches 145—150 should be averaged to obtain zenith luminance in files Data8-08. The azimuthal increments of each band are shown in Table 8. The sensor then rotates counter-clockwise to the next elevation of 18° andrecords sky patchproperties at 12° until again returning to due south.xls and Data8-09. Commission International de 1'Eclairage. Also.and 30-minute intervals. the reported value may not be representative of the sky during the time-scale the scan was conducted. Therefore. This is due to the data collection techniques of the respective sky scanners.7.3. the sensor records the properties of each sky patch at 12° intervals until it has returned to its initial starting point. A hemisphere. Technical Report CIE 108-1994. while the centre point (number 145) relates to the zenith of the sky hemisphere.3(b). Figure 5. References CIE (1994) Guide to recommended practice of daylight measurement. Fukuoka sky luminance distribution data may be transposed into the same format as Data8-08. Starting at a southerly vector.the sensor is measuring the properties of the zenith of the sky canopy.2. . For comparative purposes. representing the sky canopy. when the sensor reaches a vertical position the zenith luminance may have changed since initialisation of that scan.xls.3 represents the individual sky patches as respectivelymeasured by the PRC Krochmann (German) and EKO (Japanese) sensors. zenith luminance is the last patch to be measured. Both PRC Krochmann and EKO instruments operate on the following principle. This clockwise-counter clockwise rotation continues through the further altitudinal and azimuthal divisions of the sky. The sky scans at Garston and Sheffield sites are respectively reported at 15. This is due to this scanner recording an additional five values of zenith luminance during each complete sky scan.6. solar position and sky zenith was shown in Figure 5.xls via Calc801. These bands lie parallel to the horizon and represent increments in solar altitude of 12°. TC 3.At this point. 1987). Thus. The EKO sensor used for Japanese data collection operates in the followingmanner.242 WINDOWS IN BUILDINGS Note that the geometric relationship between sky patches. Further.6. each band is azimuthally subdivided to produce an array of sky patches.until the sensor is finallyvertical. A short note on the merit of this techniqueofsky subdivisionis availablein the literature (Tregenza. each patch. There exists a slight difference between the numbering of each sky patch.1. The outer bandwidths are those at the horizon. it is seen that 150 patches are reported during each complete sky recorded by the PRC Krochmann instrument at both UK locations. as shown in Figure 5.

May 1995. S. Tregenza.R. Final report for EPSRCresearch grant GR/K07829. Muneer. T. T. D.R. Solar Energy 19.SOLARRADIATION AND DAYLIGHT DATA 243 Ineichen. and Molineaux. B. Klein. Tregenza. and Kinghorn. Architectural Press. (1998) Solar irradiance and daylight illuminance data for the United Kingdom and Japan. 13. (1987) Subdivision of the sky hemisphere for luminance measurements. (1997) Solar Radiation and Daylight models. Renewable Energy 15. P. (1977) Calculationofmonthly average insolation on tilted surfaces. Oxford. . P.A. 325. (1995) Standardisation of UK daylight data.Lighting Research and Technology 19. 318. (1993) Characterisation and comparison of two sky scanners: PRC Krochmann and EKO instruments lEA Task XVII expert meeting Geneva. Muneer. P.

Page blank in original .

org/software-central/html/forum. and http://www.html.gov/ The website for the US-based National Climatic Data Centre provides details of the protocolto be used for downloading solar andmeteorological data for many US and global sites.efficientwindows. and http://eosweb.BADC. An International Energy Agency (lEA) database which provides bibliographic information on energy publications. sunshine and wind data.larc.org International Energy Agency website for Energy Conservation in Buildings.html http://EETD. http://solstice.AC.shtml Websites which provide a calculator for solar geometry.RL.UK The Rutherford Appleton Laboratory based website provides the UK Meteorological Office data for the UK and the EU.gov/DATDOCS/Surface_Solar_Energy.LBL.org LBL window research-related websites.badc.de/stn/html Energy Technology Data Exchange (ETDE).APPENDIX Web-based and other software material or http://www.com/ http://www.ac.ncdc.nrel. The updates are made daily thus enabling access to recent solar radiation.html Websites for obtaining solar radiation data for US sites.fiz-karlsruhe.crest.org/staff/ceg/sunangle/index. temperature. http://www. www. cloud-cover.rl. and http://solstice.nasa.html. http://www.uk/data/ synoptic-new/radtlist. http://rredc.ecbcs. . Included in the listis a wide range ofCD products.dialog.crest.noaa.gov/CBS/NEWSLETTER/NL1O/windows.gov/solar/. The network is fairly exhaustive and a listing of the concerned stations may be obtained from: http://www.

Bracknell. The program comes on a suite of five floppy discs with a briefand easy-to-read manual. Autoevaluator Micro Surveys property Systems Ltd. London. Available from the UK Meteorological Office. Ontario. Solar radiation data Solar radiation data for the United Kingdom. www.246 WINDOWS IN BUILDINGS http://www. FRAME(1992) Developed by Enermodal Engineering Limited edge. . this program analyses heat transfer through frame. sash.susdesign. The data were collectedby the Meteosat satellite.uk/'jm Website that provides validation data for RADIANCE daylight simulation package using measured sky scan luminance distributions. ASHRAE Handbook CD ASHRAE Customer Services.iesd.fr/ Information on a world-widenetwork of 48 stations measuring daylight and solar radiation under the CIE-International Daylight Measurement Programme. Atlanta Tel: + 1-404-636 8400 Contains three volumes: HVAC Equipment and systems (1995). divider and window RESFEN 3.dmu. HVAC Applications(1996). http://idmp. http://SATELLIGHT. RESFEN 3. and Fundamentals (1997). England.entpe.entpe. of Waterloo.0 is a PC program for calculating the heating and cooling energy use of windowsin residential buildings.ac. London Road. Tel: + 44-181-395 3512 A PC software which allows enables cost evaluation of energy-efficiency measures for buildings.fr/ A database for daylight and solar radiation for a quarterof a million locations within Western and Central Europe.0 (1997) Available from the Lawrence Berkeley National Laboratory (LBL) in California.com/windowheatgain A shareware program that calculates the solar heat gain through vertical windows in temperate climates.

The program is capable of analysing thermal and daylight transmission characteristics of any combination of glazing layers.APPENDIX 247 for the United States and its territories. Solar radiation data THERM 2.lbl.this database provides information for 239 stations. A detailed manual is accompanied by software.0 (1998) Available from LBL.gov/ materials/materials. gas layers. Golden—Colorado. which may be loaded.0 is a PC program for analysing twodimensional heat transfer through building products including multi-glazed windows.html. 1617 Cole Boulevard. The software is downloadable via the website: http://windows. Prepared under the auspices of the University of Lowell Photovoltaic program. on PCs from a set of two floppy discs. VISION (1992) A computer program to be used as a design and rating tool for evaluating the heat transfer. this compendium provides information for 55 countries.1 (1994) Developed at the Lawrence BerkeleyLaboratory (LBL) in California this is a PC programfor analysing the thermal performance of fenestration products. THERM 2. WINDOW 4. . University of Waterloo in Canada. International Solar Irradiation Database.gov. Optical data for over a thousand glazing products is included in its library. Staff at LBL may be contacted at Plross@lbl. Available from National Renewable Energy Laboratory (NREL). solar gain and temperature profile throughglazing. Theprogram is available from the Advanced Glazing System Laboratory. frames and spacers.

Page blank in original .

134 Brundtland Report. 143 Brise soleil. 185 Adaptability. 188 Aluminium. 194—9 Autoevaluator software. 185—6 Aural performance. 115—17 triple glazing. product. energy analysis. 207—209 Artificial lighting: controls. 100 BREEAM design guidelines. 27 Candela (unit). 129 Best Practicable Environmental Option (BPEO). 9 Air change rate. 117—18 glazing systems. 118 Acoustic environment. 246 Aural comfort. 41 Air-conditioning systems. 215 Apparent solar time (AST). 5 Aircraft noise.INDEX Absorption/reflection characteristics: double glazing. 186 noise rating (NR). 246 Background noise levels. condensation prevention. 219—20 ASHRAE Handbook CD. 91—2 Argon. windows. windows. as inert infill gas. 190 Baffleplates. 227 C-based Dynamic Linked Libraries (DLL). 211—13. 115 single glazing. 221 Bouguer—Lambert law. 150 energy consumption. 228 Aerogel filled windows. 80 Berm House (example of solar heating). 144 . 189—90 Acoustic properties.

3—4.250 INDEX Carbon dioxide. 10 assessmentmodel. conversion to hourly diffuse irradiance. 98—9 Cold mirrors. 145—7 sources. 67—72 free flow in enclosed space. 232 Dew-point temperatures. 58—60 effect of temperature difference. 56 Cloud-cover radiation model (CRM). 11—12 Convection: along vertical plate with uniform heat flux. 76—9 avoidance. 55—8 effect of cavity width. 61 Daily global irradiation. 66. 143 measurement programme. 77 Disability glare. 76 innovative prevention developments. 37—40 Commission International de l'Eclairage (CIE). 48 Comfort criteria. energy analysis. 145 systems design. environmental concerns. 157—8 Discomfort glare. 13 double-glazed windows. indoor. 235 Component manufacturing. 17 free convection flow analysis. 55—8 Computers. 154—5 calculation ofcomponents. 8 Cold-bridging phenomenon. 155—6 Disposal see recycling/disposal . use in design. 14. 151—4 Daylighting: availability. 61—2 effect of window height. 17—18 Condensation. 103 Day number (DN). 80 prediction charts. 235 role of windows. 147—50 Design life. 201—202 Cavity convection heat transfer coefficient. 76 surface. 91 Daylight factor (DF). 213 Computational Fluid Dynamic (CFD) modelling. 150. 51—5 Convection flow analysis: CFD modelling. 180 advantages. 60—61 effect ofinfill gases. 14. 158 fundamentals.

228 Dynamic Linked Libraries (DLL). 47—51 Effective emissivity. 222—4 costs. energy analysis. 114 selection. 91. 117—18 aural performance. 180 in productuse. 154 Facilities management. 217—20 see also life cycle assessment Energy efficiency. 151. daylight. 215 Finite element analysis (FEA). 51—2 empirical correlations. 9 Glare. optimisation. aural performance. 43 Glazed areas. 224 Energy balance. 230 Energy rating index (ERI). 39 Durability. 27 Edge-seal thermal analysis. 150. 46 Externally reflected component (ERC). 52—5 Gas-filledwindows. 245 Equation of time (EOT). 145 External heat transfer coefficient. 103 . 213 Glass thermal resistance. 13—14 Energy consumption. 14 see also multi-glazing Dry resultant temperature. 195 for energy conservation. 89 Energy Technology Data Exchange (ETDE).INDEX 251 Double glazing: absorption/reflection characteristics. 92 European Court of Human Rights building. 2—3 legislation. 85—90 Energy conservation. 195 Geometric media. 230 Factoryservices. 17 FORTRAN-based Dynamic Linked Libraries (DLL). 155—8 Glass sheet manufacturing. product.46 Emissions. 114 Global irradiance. 27 FRAME (1992) program. USA. energy analysis.from electricitygeneration. 246 Free convection flow analysis. 180 Glazing systems: absorption/reflection characteristics.

multi-glazing. 97—8 monthly-averageddiffuse. 160. 236 International Daylight Measurement Year (IDMY). 108 monthly-averaged global. 160. 37—40 Illuminance. 121—2 Heat transfer: double-glazedunits. 191—4 Heat balance equation. 201—202. 4. 47 Internally reflected component (IRC). daylight. effect of noise. 8 Horizontal global/diffuse illuminance. 235 stations. 226—31 Inert gas fill. 38 Heat capacitance effects. 237 Impact assessment. 43 through glass. 58—60 Infiltration effects. 11 Interior noise targets. human requirements. 150. 38 thermal comfort. 103. 54—5 Holographic coatings. 186 Internal environment effects. 144 data availability. 201 Goal Seek facility. 43—45 from building. 222. 11 Internal heat transfer coefficient. data extraction for LCA. 107—108 global. 224—5 Improvement analysis. 2 heat balance equation. Microsoft Excel. 86 Hourly horizontal irradiation. 225 Health. 53 Greenhouse gases. energy analysis. 108 Hourly slope irradiation. 235 Inventory analysis. effect on convection heat transfer. 222—3 transportation. 207—209 Infill gases. 43 to internal glass surface. 95—7 diffuse. 162 thresholds. energy balance. 110—111 Human requirements. 24—6 Grashofnumber. 40—4 1 from outer glass surface. 158—9 Hottel—Whillier—Blissmodel. 206 . 42—3 within a cavity.252 INDEX Global warming. 154 International Daylight Measurement Programme (IDMP). 151.

226—31 inventory analysis. 230 Irradiance modelling. 177 transmission incidence angle. 228 Management issues. energy conservation. 204 multi-glazed windows. 144 Luminance. 202 impact assessment. as inert infill gas. 126 Krypton. 201 definition. 102. 162—3 Long wavelength radiation exchange. 207—209 Kyoto Protocol. 207—213 . 162 Irradiation: hourly horizontal. 213—15 Material requirements: data extraction for LCA. 105 Liu and Jordan. 22—3 Lotus. improvement analysis. daylight frequency. data extraction for LCA. 222—5 improvement analysis. product. 159 model. 13—14 Life cycle assessment (LCA). 39 Lookup tables. 95—7 hourly horizontal global. 225 Lighting see artificial lighting Lightpipes. 177—9 Luminous efficacy. 206—22 investigations. 229—31 Manufacturing. 18 Low-emissivity coatings. 144 Luminous intensity. 159 Luminous flux. 175. 3—4. 209 Lumen (unit). 180—82 Linke turbidity factor. 1-2-3 spreadsheet. 97—8 Kappafloat windows. 62 energy analysis.INDEX 253 Inventory handling. 205 planning. 13. 5 Legislation. 203—206 transportation. 159—60 Muneer—Kinghorn Perez model. 144 Maintenance. 205. 144—5 transmission. 6—8 effect on U-value.

129 National Climatic Data Centre. 72—3 thermal transmission. 18—19 as CAD tool. 54—5 Optimum window selection. 13 longitudinal temperature profile. 19 for dynamic graphs. design against overheating. 103 Overheating. 134—8 Night-time cooling. 134—5 Oxford Instruments building. 19 Goal Seek facility. 245 Natural ventilation. 31—4 Vlookup function. 229 useful life. US-based. 24—6 for lookup tables. 186 transmission through glazing. 22 functions. 181 Multi-glazing: advances. 191—4 low frequency. 164—6 Muneer—Kinghorn Muneer--Kinghorn model (zenith luminance). 231—2 Orientation factors. 194—9 Noise and Number Index (NNI). 37 heat capacitance effects. 171 Napier sauna (example of solar heating). 229 Meteorological radiation model (MRM) see sunshineradiation model Microsoft Excel. 138 Noise: annoyance levels. 186—9 targets for interior. 163—4 Overcast day irradiation. 192 health and productivity effects. avoidance. 134 . 22—3 for sequential computation. 74—6 model (luminous efficacy). 188 Nusselt numbers. 192 sources.254 INDEX reduction. 100 Monodraught 'Sunpipe' system. 19—22 intensiveness Solver facility. 23 Mie scattering. 121—2 life cycle assessment. 52. 159—60 Muneer—Kinghorn model (slope illuminance). 180. 40—42 Multimode thermal model.

230 Page radiation model (PRM). 172.0 (1997) program. 237. 190 Process control. 126—8 Perez model (luminous efficacy). 122—5 simulation. 239 Radiation heat transfercoefficient. 105—106 extreme clarity clear sky. 121 Single glazing. 98. 245 St George's School (example of solar heating). 206. 214 Seasonal affective disorder (SAD). 191—4 RADIANCElighting simulation system. energy conservation. 188 Raw material extraction. 227—8 Product use. 102—103 Passive solar design. daylight. 152—4 . 4 Shading coefficient. improvement analysis. 52 Rayleigh scattering. data for LCA. 230 Process substitution. 213. 229 Repairs. 101—102. product. 228 Recycling/disposal: data extraction for LCA. absorption/reflection characteristics.INDEX 255 Packaging. 179 Radiation distribution data. 221—2 product. 115—17 Sky clarity/clearnessindices. 151. 110. 246 Road traffic noise levels. 159 Preferred octave speech interference level (PSIL). product. 102 overcast sky. product. 228 RESFEN 3. 101 clear sky (PCSRM). 13—14 Relative humidities. 186—7 Room surface temperature effects. 217—20 Productivity. 205 Rayleigh numbers. 231 Product life. 230 Productlabelling. effect of noise. 150. energy analysis. 159 Sky component (SC). 228 Remanufacture. product. 206. 129 Sauna (example of solar heating). 47 Rail traffic noise level.229 Regulations. 77 Reliability. 100 Re-use. 12 Rutherford Appleton Laboratory. dataextraction for LCA. 129 Sealed glazing units.

90 Solar radiation: availability. 235 Solar day. 148 Sky luminance distribution. 5 collection with superinsulated windows. 93—4 Solar heating. 238 Snell's law. 94 Solar illuminance. 28 Solar transmission: . 242 Sky scan data. 239 database. 99—100 Solar time. 122—5 building examples. 240 recording map. 92 Solar design. 86—90 current technology. 87—90. 241. 111—14 Solar luminance. 93 Solar data measurement programme. 237 data. 171—5 Sky patch geometry. 164. 110 Slope illuminance. effect on daylighting. 239 database. 128—31 economics. 90 comparison of models. 240 data files format. 132—3 potential. 91 models. 91 Solar declination (DEC). 126—8 Solar hour angle. 90 Solar irradiation. 28. 85 Solar factorsee total solar transmission Solar geometry. 145. 115 Solar altitude. 174 Sky-diffuse models. 237 data file nomenclature. 237. 93 Solar azimuth. 98 transmission. 28. 130—32 simulation.256 /NDEX Sky conditions. 104 frequency distribution. 104 data sources. 90 data. passive see passivesolar design Solar energy. 247 fundamentals. 235. 114 transmission through terrestrial atmosphere.

124 Sunblinds. 66 Therapeutical aspects. 98.231 Sustainable development. 28 Substitution. 146 Superwindows. 120 aluminium. 227 Temperature variation: longitudinal. 121—2 for larger glazed areas. 37—40 Thermal properties. energy analysis. 51 in multi-glazing. 118—19. 50 Spectral splitting coatings. 71 physical model. 198 Speech-interference level (SIL). 90 Sun-spaces. windows. 217—18 Spacers: total. 8 Spectrum adaptation terms. 94—5 Sunshine radiation model (SRM). 100—101 Superinsulated windows: advantages. design against overheating. 213—14 Tokyo Electric Power Company building. 224—5 .0 (1998) program. 134—6 Sunpipes see lightpipes Sunrise/sunset. 9 Transportation: data extraction for LCA. 190 Standard time. 209—11. 5 Sound Reduction Index (R). 50 foam. 6 heat capacitance. 4 THERM 2. 215—17 impact assessment. energy consumption. 134—5 Total solar transmission (F). 72—3 stratification.INDEX 257 computer-aided tools. materials.8 Supplier selection. 196—7 Space heating.48 stainless-steel. 37 Thermal transmittance see U-value Timber sash and frame. 121 Sound insulation. 229 Sun angle calculator. 206. 120 Transparent insulation material (TIM). 247 Thermal comfort.

168 all-sky models. 6 Windcatcher (natural ventilation) system. 135. 207—209 Xerogel filled windows. as inert infill gas.1 (1994) program. 9 Zenith luminance. 45 components. 247 World Resource Foundation.258 INDEX improvement analysis. 42 computer tools for analysing. 168—9 Muneer—Kinghorn model. 12 VisiCalc spreadsheet. 247 Visual Basic for Applications (VBA) routines. 14 Trombe wall. 19—20 effect of low-emissivitycoatings. 118 aural performance. 171 137—8 . 18 VISION (1992) program. WINDOW 4. 9—10 Ventilation effects. 62 modified for solar gain. 45 Vacuum windows. 125 U-value: British Standard computation. 8 derivation. 23 Waste management see recycling/disposal Wavelength selective coatings. 93 language. 26 Vlookup function. 195 for energy conservation. 87—8 multi-glazed windows. 230—31 Triple glazing: absorption/reflection characteristics. 62—6 definition. 222 Xenon.

com [ I O\SC- g 1780750 642095 .architecturalpress. Muneer. The informationis presented in easy-to-usecharts. tables and software • Presents state-of-the-art information on glazingand in-fill gases • Exhaustive coverage of thermal. daylight and sound transmissionproperties Glass and glazing systems lie at theheart of window design. This information is the result of many years of research undertaken by the authors. The accompanying CD-ROM demonstrates the computation examples in the book. visual and acoustic environments. Windows in Buildings meets this need: it presents state- of-the-art information on high efficiency glazing systems and their energetic and environmental impact on modern buildings. Weir and J. glazing systems and other window materials. N. An understanding of their interaction and relation to built structures is crucial to the modern designer. to enable their suitable and appropriateincorporation into modern buildings.WINDOWS IN BUILDINGS THERMAL. Frontcover photo:Western MornngNevasand Pnntirry Works in Plymouth. Kubie • Highlights theenvironmental impact and efficiencyof glazing • Free CD-ROMto accompany the book • Easy-to-use charts. VISUAL AND SOLARPERFORMANCE 1. ACOUSTIC. Abodahab. oesrgnedby Grimshaus Partners. courtesyof SaintGobain Solaglass ISBN O•7506-4209-2 ___ ArchitecturalPress of An imprint Butterworth-Heinemann http://www. G.courtesyofPilkinyton Glass Ltd N. It contains over fifty Excel spreadsheets showing window design data and routines for analyzing a building's thermal. solar. tables and software for the design and selection of glass. and Sackcoverphoto: SamplesoF SGG Climalit.

Sign up to vote on this title
UsefulNot useful