Ventilation of Buildings - Hazim B. Awbi

Administrator200079a5coverv05b.jpg

Ventilation of Buildings

Essential reading

Building Services Design Methodology, a practical guideD. Bownass pbk: 0-419-25280-0

Spon Press

Building Services Engineering, 3rd editionD. Chadderton hbp: 0-419-25730-6

pbk: 0-419-25740-3Spon Press

Naturally Ventilated Buildings, building for the senses,the economy and societyEdited by D. Clements-Croome hbk: 0-419-21520-4

Spon Press

Indoor Air Quality IssuesD. L. Hansen hbk: 1-56032-866-5

Spon Press

Information and ordering detailsFor price availability and ordering visit our website www.sponpress.comAlternatively our books are available from all good bookshops.

Ventilation of BuildingsSecond edition

Hazim B. Awbi

First published 2003by Spon Press11 New Fetter Lane, London EC4P 4EE

Simultaneously published in the USA and Canadaby Spon Press29 West 35th Street, New York, NY 10001

Spon Press is an imprint of the Taylor & Francis Group

2003 Hazim B. Awbi

All rights reserved. No part of this book may be reprinted or reproducedor utilised in any form or by any electronic, mechanical, or other means,now known or hereafter invented, including photocopying and recording,or in any information storage or retrieval system, without permission inwriting from the publishers.

British Library Cataloguing in Publication DataA catalogue record for this book is availablefrom the British Library

Library of Congress Cataloging in Publication DataA catalog record for this book has been requested

ISBN 0415270561 (pbk)ISBN 0415270553 (hbk)

This edition published in the Taylor & Francis e-Library, 2005.

To purchase your own copy of this or any of Taylor & Francis or Routledgescollection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.

ISBN 0-203-63447-0 Master e-book ISBN

ISBN 0-203-63782-8 (Adobe eReader Format)

Contents

Preface to the second edition ixPreface to the rst edition xi

1 Human comfort and ventilation 1

1.1 Introduction 11.2 Heat balance equations 11.3 Environmental indices 91.4 Thermal comfort models 131.5 Thermal discomfort 221.6 Indoor air quality 37

References 45

2 Ventilation requirements 48

2.1 Introduction 482.2 Indoor contaminants 492.3 Ventilation rates 652.4 Air change effectiveness and age of air 782.5 Types of ventilation systems 822.6 Energy requirement for ventilation 89

References 93

3 Air inltration calculation and measurement 97

3.1 Introduction 973.2 Air leakage characteristics of buildings 983.3 Air inltration calculation and modelling 1063.4 Measurement of air inltration 121

References 135

4 Principles of air jets and plumes 137

4.1 Introduction 1374.2 Free air jet 1374.3 Wall jet 150

vi Contents

4.4 Effect of buoyancy 1564.5 Jet interference 1634.6 Plumes 173

References 183

5 Air diffusion devices 186

5.1 Introduction 1865.2 Air diffusion glossary 1865.3 Performance of air terminal devices 1885.4 Types of air terminal devices 1965.5 Selection of air terminal devices 213

References 220

6 Design of room air distribution systems 221

6.1 Introduction 2216.2 Model investigations 2216.3 Case studies 2636.4 Design procedures 275

References 301

7 Natural, hybrid and low energy ventilation 304

7.1 Introduction 3047.2 Wind characteristics 3057.3 Wind-induced ventilation 3087.4 Buoyancy-induced ventilation 3137.5 Combined wind and buoyancy ventilation 3177.6 Characteristics of natural ventilation openings 3187.7 Natural ventilation strategies 3247.8 Combined natural and mechanical (hybrid) ventilation 3387.9 Ventilation for re control 341

References 345

8 Computational uid dynamics in room air ow analysis 348

8.1 Introduction 3488.2 Transport equations 3488.3 Turbulence models 3538.4 Solution of transport equations 3648.5 CFD applications to room air movement 386

References 444

9 Measurement of indoor climate 450

9.1 Introduction 4509.2 Measurement of air temperature 4509.3 Measurement of radiant temperature 457

Contents vii

9.4 Measurement of humidity 4639.5 Measurement of pressure 4669.6 Measurement of air velocity 4699.7 Measurement of volume ow rate 4829.8 Measurement of thermal comfort 4929.9 Measurement of air pollutants 4949.10 Tracer gas measurements 4989.11 Air ow visualization 500

References 507

Appendix A Air inltration calculation software 509

A.1 Air inltration development algorithm (AIDA) 509A.2 Lawrence Berkeley Laboratory (LBL) model 512A.3 AIOLOS model 512A.4 COMIS model 512A.5 CONTAMW model 513

References 513

Appendix B CFD codes 514

Index 517

Preface to the second edition

During the last three decades ventilation philosophy has been experiencing majorchanges. In the rst decade of this period, considerable efforts were made towardsunderstanding the mechanisms of air inltration in buildings in order to control andoften reduce the fortuitous ventilation and conserve energy. In some cases, the reduc-tion in air inltration created problems associated with the air quality in buildings andthe generic term sick building syndrome came into being. The second decade of thesame period experienced concerted efforts to understand the causes of sick buildings,which resulted in the introduction of new ventilation concepts, such as the age of air,and new air quality units, and a consensus for increased outdoor air ow rates. Inthe third decade, the emphasis on reducing energy consumption and environmentalconsciousness has focused the minds of researchers and designers alike on the poten-tial of natural ventilation and user control of the local environment. As a result ofthese changes, new ventilation standards and guidelines have been written to reectthe importance of ventilation on the quality of indoor environment.

Almost 12 years have passed since the publication of the rst edition of Ventilationof Buildings. During this time ventilation has truly been established as a disciplinefor scientists to research and engineers to apply. Previously, ventilation was consid-ered to be on the fringe of scientic research; this is no more to be. The subject nowattracts respected researchers from different backgrounds all round the world, whoare engaged in developing new theories and applications into more effective and sus-tainable ventilation systems. This shift on emphasis did not happen accidentally butas a result of mans needs for better indoor environment.

In this second edition, almost every chapter from the rst edition has been thor-oughly updated and a new chapter on natural ventilation has been added. Recentdevelopments in ventilation concepts and room air distribution methods are includedwithin the updated chapters. It was the intention that this edition provides the readerwith recent developments in the subject but, at the same time, emphasizes the practicalaspects that are needed for modern ventilation system design.

With the recognition of ventilation as a science, much good quality research hasbeen done. I tried to capture some of this in the new edition, but in a text of this kindit is impractical to include all of the good quality research that has been done and hopethat I have not omitted the obvious!

Finally, my thanks and appreciations to those colleagueswho provided their researchdata and illustrations to include in this edition. My thanks also go toMichelaMarchettifor producing the CAD drawings for some of the new illustrations.

Hazim B. AwbiReading, UK, 2002

Preface to the rst edition

The term air-conditioning is generally understood to mean the heating, cooling andcontrol of moisture in buildings. Consequently, this involves heating and cooling loadcalculations in addition to the design of plant components, ductwork and control sys-tems. Ventilation, on the other hand, refers to the provision of sufcient quantities ofoutside air in the building for the occupants to breathe and to dilute the concentrationof pollution generated by people, equipment and materials inside the building. Thisis necessary for both air-conditioned and non-air-conditioned buildings. The dilutionof indoor contaminants is inuenced by the quantity and quality of the outside airsupplied to the building as well as the way this air is distributed around the space. Themethod of air distribution is also a vital part of any air-conditioning system. In thisbook, the emphasis is on the last two themes, namely quantifying the outdoor air owrate to a building and distributing this ventilation air around the space.

During the last 15 years ventilation philosophy has been experiencingmajor changes.In the rst part of this period, considerable efforts were made to understand the mech-anisms of air inltration in buildings in order to control, and often to reduce, thefortuitous ventilation and conserve energy. In some cases, the reduction in air inltra-tion created problems associated with the air quality in the building and the genericterm sick building syndrome came into being. The second half of the same periodexperienced concerted efforts to understand the causes of sick buildings, which resultedin the introduction of two new air quality units by Professor P. O. Fanger, namely theolf and the decipol, and a consensus for increased outdoor air ow rates. As a result ofthese changes, some ventilation standards, which initially recommended a reductionin outdoor air requirement for occupancy, had to increase these rates beyond thoserecommended prior to this period. Reecting the current attitude, the new ASHRAEStandard 62-1989 has increased the minimum outdoor air per person from 2.5 to7.5 l s1, i.e. a threefold increase.

This book has been designed to complement rather than replicate the HVAC hand-books such as those by ASHRAE and CIBSE. Where appropriate the theory of a designproblem is given to broaden the readers horizon of the subject. Recent developmentsin ventilation requirements, thermal comfort, indoor air quality and room air dis-tribution are also included. The text is intended for the practitioner in the buildingservices industry, the architect, the postgraduate student taking courses or research-ing in HVAC or general building services and the undergraduate studying buildingservices as a major subject. The book assumes that readers are familiar with thebasic principles of uid ow and heat transfer and knowledgeable with regard to the

xii Preface to the rst edition

thermal characteristics of building fabric. However, Chapter 7 requires more advancedknowledge of partial differential equations, which describe the turbulent ow and heattransfer processes of uids.

Chapter 1 presents the theory and practice of thermal comfort and indoor air qual-ity (AIQ) in some detail, including recent advances in these areas based largely on thework of Fanger and his associates. Chapters 2 and 3 describe the procedures used indetermining ventilation rate requirements for different occupancy and the methodsof determining air inltration rates and the design of passive and smoke ventilationtechniques. Chapters 4 and 5 outline the theory of different types of air jets, includ-ing the inuence of buoyancy, Coanda force and obstructions and the aerodynamiccharacteristics of various air terminal devices used in practice. Chapter 6 presents, insome detail, the theory of physical modelling as applied to room air movement anddiscusses the inuence of various uid and heat ow parameters on the modellingprocess. This chapter also contains case studies involving reduced-scale and prototypemeasurements. Design procedures for different room air distribution methods are alsopresented here. Chapter 7 is devoted to the theory of computational uid dynamics(CFD) as applied to ventilation and room air movement and presents the results ofrecent publications to illustrate the state of the art in this rapidly expanding eld ofventilation research. Finally, Chapter 8 deals with the principles of measuring air tem-perature, radiant temperature, humidity, pressure, air velocity, air ow rate, thermalcomfort, indoor air contamination and air ow visualization.

This book provides the reader with recent developments in the subject which arelargely missing from other titles currently available. The ultimate aim is, of course,better design of ventilation systems to reduce the frequency of public complaints aboutHVAC systems which have increased in recent years. Finally, with the subject ofintelligent buildings of the future frequently being mentioned, it is hoped that thisbook will provide the designer of ventilation systems for these buildings with some ofthe necessary information.

Hazim B. Awbi1991

1 Human comfort and ventilation

1.1 IntroductionThe purpose of a ventilation system is to provide acceptable microclimate in the spacebeing ventilated. In this context, microclimate refers to thermal environment as wellas air quality. These two factors must be considered in the design of a ventilationsystem for a room or a building, as they are fundamental to the comfort and well-being of the human occupants or the performance of industrial processes within thesespaces. In a modern technological society, people spend more than 90% of their timein an articial environment (a dwelling, a workplace or a transport vehicle). As aresult of the energy-saving measures which started in the early 1970s these articiallycreated internal or indoor environments have undergone radical changes, somebeing positive and others negative. On the positive side, increased levels of thermalcomfort have become fait accompli through improved thermal insulation and moreadvanced air-conditioning or heating system design. On the negative side, a deteriora-tion of the indoor air quality has been experienced particularly among air-conditionedbuildings [1]. Indeed, the term sick building syndrome is synonymouswith the energy-saving era. These indoor air quality problems have been associated with poor plantmaintenance, high concentrations of internally generated pollutants and low outdoorair supply rates.

The designers and operators of ventilation systems should be familiar with the com-fort requirements and the quality of air necessary to achieve acceptable indoor climate.These require knowledge of the heat balance between the human body and the inter-nal environment, the factors that inuence thermal comfort and discomfort as well asthe indoor pollution concentrations that can be tolerated by the occupants. In recentyears, there has been a ux of published information, notably from Scandinavia andthe USA, on these important issues of the built environment. The fundamental workthat has been done on thermal comfort and air quality will be discussed in the rstpart of this chapter and more recent developments will be presented in the second part.This chapter will then become the fulcrum for the remaining chapters of this book.

1.2 Heat balance equations

1.2.1 Body thermoregulation

The primary function of thermoregulation is tomaintain the body core, which containsthe vital organs, within the rather narrow range of temperature which is vital for theirproper functioning. The temperature control centre is the hypothalamus, a part of the


209 W

349 W

488 W

Figure 1.1 Rectal temperature as a function of ambient temperature for three activities.

brain that is linked to thermoreceptors in the brain, the skin and other parts of the bodysuch as the muscles. The hypothalamus receives nerve pulses from the temperaturesensors and coordinates information to different body organs to maintain a constantbody core temperature. The thermoreceptors are particularly sensitive to changes intemperature and temperature change rates as small as +0.001 and 0.004Ks1 canbe detected. Temperature regulation is carried out by controlling metabolic heat pro-duction rate, control of blood ow, sweating, muscle contraction and shivering inextreme cold situations. Under normal conditions the body core temperature, tc, isapproximately 37 C and this is maintained at a constant value despite changes in theambient temperature, as shown in Figure 1.1 for three levels of activity [2]. The gureshows the ability of the temperature-regulating mechanisms to maintain a constantcore temperature up to a certain ambient temperature beyond which core temperaturecannot be maintained because evaporative cooling of sweat becomes ineffective. It isalso shown in Figure 1.1 that the core temperature is not always constant but dependson activity, i.e. increases with increase in metabolism, and may be as high as 39.5 Cin extreme activities.

While the body core temperature remains almost constant over a wide range ofambient temperatures, the skin temperature changes in response to changes in theenvironment and is usually different for different parts of the body. However, thevariation in skin temperature over the body is reduced when the body is in a state ofthermal equilibrium and comfort. The variation of mean skin temperature, ts, for anude person with ambient temperature is shown in Figure 1.2, which also shows therectal temperature for the purpose of comparison [3].

1.2.2 Heat transfer between body and environment

The heat balance equation for the human body is obtained by equating the rate of heatproduction in the body by metabolism and performance of external work to the heatloss from the body to the environment by the processes of evaporation, respiration,radiation, convection and conduction from the surface of clothing. Thus:

S = M + W + R + C + K E RES (1.1)

Human comfort and ventilation 3

Figure 1.2 Variation of mean skin and rectal temperatures for a nude body with ambienttemperature.

where S = heat storage in body ( tb/ where is time), W; M = metabolic rate, W;W = mechanical work, W; R = heat exchange by radiation, W; C = heat exchangeby convection, W; K = heat exchange by conduction, W; E = evaporative heat loss,W and RES = heat loss by respiration, W. A positive value of S indicates a rising bodytemperature, tb, a negative value suggests a falling tb, and when S = 0 the body is inthermal equilibrium. The expressions that are used to calculate the terms on the right-hand side are given below, but for a more detailed assessment of each term readersshould consult the books by Fanger [4] and McIntyre [5].

Metabolism (M) This is the body heat production rate resulting from the oxidationof food. Its value for every person depends upon their diet and level of activity andmay be estimated using the equation below:

M = 2.06 104V(Foi Foe) W (1.2)where V = air breathing rate, l s1; and Foi and Foe are the fraction of oxygen in theinhaled and exhaled air, respectively. The value of Foi is normally 0.209 but Foe varieswith the composition of food used in the metabolism; for fat diet Foe 0.159 and forcarbohydrates Foe 0.163.

The rate of metabolism per square metre of body surface area can be obtained usingequation (1.2) and Du Bois area given as:

AD = 0.202m0.425H0.725 m2 (1.3)where m = body mass, kg and H = body height, m. For an average man of mass 70 kgand height 1.73m, AD = 1.83m2. Metabolism is often given the unit met whichcorresponds to the metabolism of a relaxed, seated person, i.e. 1met = 58.15Wm2.Typical activity values are given in Table 1.1.

Work (W) If work is produced by the body, the metabolism increases to provideextra energy to perform this work. Because the thermal efciency (W/M) of a humanis poor, i.e. less than 20%, for every watt of work power produced an increase in


Table 1.1 Metabolic rates of different activities [6]

Activity Metabolic rate

(Wm2) (met)

Reclining 46 0.8Seated, relaxed 58 1.0Standing, relaxed 70 1.2Sedentary activity (ofce, dwelling, school, laboratory) 70 1.2Standing activity (shopping, laboratory, light industry) 93 1.6Standing activity (shop assistant, domestic work, machine work) 116 2.0Medium activity (heavy machine work, garage work) 165 2.8

metabolism of 5W will be needed. Most of the work produced by a person is positive;however, in some cases negative work can occur such as a personwalking down a steephill, in which case some of the potential energy will be converted to heat in the musclesto keep a constant walking speed.

Radiation (R) The heat exchange by radiation occurs between the surface of thebody (clothing and skin) and the surrounding surfaces such as internal room surfacesand heat sources or sinks. It is calculated using the StefanBoltzmann equation:

R = fefffcl(t4cl t4r

)Wm2 (1.4)

where feff = factor of effective radiation area, i.e. ratio of the effective radiation area tothe total surface area of clothed body; fcl = factor of clothing area, i.e. ratio of surfacearea of clothed body to surface area of nude body; = emissivity of clothed body; =StefanBoltzmann constant, i.e. 5.67 108 Wm2 K4; tcl = surface temperatureof clothing, K and tr = mean radiant temperature, i.e. effective temperature of roomsurfaces, K.

The factor fcl is used in equation (1.4) because the heat transfer is based on the surfacearea of the nude body, i.e. Du Bois area, AD. Because some parts of the human bodyact as a shield to other parts, the factor feff has a value of 0.696 for a seated personand 0.725 for a standing person; a mean value of 0.71 is usually used. The combinedemissivity of skin and clothing, , is approximately 0.97 and, for long-wave radiation,this is independent of the skin or clothing colour. Substituting the values for feff , fcl, and , equation (1.4) reduces to:

R = 3.9 108fcl(t4cl t4r

)Wm2 (1.5)

The mean radiant temperature, tr, can be either calculated from areas, view factorsand temperatures of room surfaces (e.g. see Fanger [4]) or measured directly (seeChapter 9). The clothing temperature, tcl, depends on the metabolic rate, thermalresistance of clothing and air temperature (see later).

The range of temperatures in the indoor environment is usually small (typically1030 C) and in this case the fourth power law of equation (1.5) can be adequatelyreplaced by the linear equation:

R = fclhr(tcl tr

)(1.6)


where the radiant heat transfer coefcient, hr, can be approximated by:

hr = 4.6(1 + 0.01tr) (1.7)

For normal room conditions, hr 5.7Wm2 K1.Convection (C) The heat transfer between a body and the surrounding air is pri-

marily by convection which can be either free (natural) caused by buoyancy or forced(mechanical) caused by a relative movement between the body and air. The generalheat convection equation is:

C = fclhc(tcl ta) Wm2 (1.8)

where hc = convective heat transfer coefcient, Wm2 K1; ta = air temperature, C.The convective heat transfer coefcient, hc, depends on the mode of heat transfer. Forfree convection it is given by:

hc = 2.38(tcl ta)0.25 (1.9)

For forced convection it is:

hc = 12.1

(v) (1.10)

where v is the relative velocity between the body and air, m s1.Combined radiation and convection (R + C) The radiation and convection heat

transfer may be combined into a single equation to give the sensible heat transfer fromthe body to the surroundings, i.e. using equations (1.6) and (1.8):

R + C = fcl[hr(tcl tr) + hc(tcl ta)]

This equation may be simplied to:

R + C = fclh(tcl to) (1.11)

where h is a combined radiation and convection heat transfer coefcient, W m2 K1,and to is the operative temperature, C. These are given by:

h = hr + hc (1.12)to = (hr tr + hcta)/(hr + hc) (1.13)

The operative temperature can be dened as the average of the mean radiant andair temperatures weighted by their respective heat transfer coefcients. At air speedsv 0.4m s1 and tr < 50 C, to is approximately the simple average of the air andmean radiant temperatures. This is an environmental index that is described later.


Conduction through clothing (K) The heat conduction through clothing using thenormal heat conduction equation is:

K = hcl(ts tcl) Wm2 (1.14)where hcl = heat conductive coefcient of clothing, Wm2 K1; ts = average skintemperature, C.

The conductive coefcient is often replaced by the reciprocal of the thermal resis-tance of the clothing, Icl. The clothing thermal resistance is usually given the unit clowhich is given by:

1 clo = 0.155m2 KW1Hence:

hcl = 1/(0.1555Icl) = 6.45/Icl (1.15)where Icl has the unit clo.

Typical values of Icl for different clothing ensembles are given in Table 1.2. Theclothing area factor fcl is directly related to the thermal resistance of clothing and iscalculated from the following relations:

fcl ={1.00 + 0.2Icl for Icl < 0.5 clo1.05 + 0.1Icl for Icl < 0.5 clo

(1.16)

The mean skin temperature, ts, may be estimated from an empirical relation derivedby Fanger [4] which relates the mean skin temperature to activity as follows:

ts = 35.7 0.0275(M W) C (1.17)This formula has been derived for 4 > M > 1.

Table 1.2 Thermal resistance of clothing ensembles [6]

Clothing ensemble Icl

(m2 KW1) (clo)

Nude 0 0Shorts 0.015 0.1Typical tropical clothing ensemble: 0.045 0.3briefs, shorts, open-neck shirt with short sleeves, lightsocks and sandals

Light summer clothing: 0.08 0.5briefs, long lightweight trousers, open-neck shirt withshort sleeves, light socks and shoes

Light working ensemble: 0.11 0.7light underwear, cotton work shirt with long sleeves,work trousers, woollen socks and shoes

Typical indoor winter clothing ensemble: 0.16 1.0underwear, shirt with long sleeves, trousers, jacket orsweater with long sleeves, heavy socks and shoes

Heavy traditional European business suit: 0.23 1.5cotton underwear with long legs and sleeves, shirt, suitincluding trousers, jacket and waistcoat, woollensocks and heavy shoes


Evaporative heat loss (E) Heat loss by evaporation is partly due to diffusion ofwater vapour through the skin tissues, Ed, and partly due to evaporation of sweat fromthe skin surface, Esw. In both cases heat is absorbed from the skin and this processcontrols the rise in body temperature. The water diffusion is a continuous processthat occurs even in a cool environment but the sweat evaporation only occurs in a hotenvironment and when the body activity is higher than normal.

The diffusion heat loss depends on the difference between the saturated vapourpressure at skin temperature, pss, and the vapour pressure of the surrounding air, pa.Olesen [3] gives the following equation for Ed:

Ed = 3.05 103(pss pa) Wm2 (1.18)where pss and pa are in pascals (Pa).

The saturated vapour pressure may be obtained using an equation such as that givenin the CIBSE Guide [7] which is:

log10 pss = 30.59051 8.2 log10 ts + 2.4804 103ts 3.14231 103/ts(1.19)

Here, pss is in kPa and ts is absolute skin temperature, K.Over the skin temperature range 27 C < ts < 37 C a linear expression for pss can

be approximated to within 3% error using steam table data. Thus:

pss = 256ts 3373 Pa (1.20)where ts is in C.

Equations (1.18) and (1.20) may be combined to give:

Ed = 3.05 103(256ts 3373 pa) Wm2 (1.21)pa can be calculated from a knowledge of the air temperature and relative humidityas shown in Section 2.2.9. The value of Ed represents the minimum heat loss byevaporation.

McIntyre [5] gives a different expression for Ed as follows:

Ed = 4.0 + 1.2 103(pss pa) Wm2 (1.22)where pss and pa are in Pa.

The evaporation of sweat from skin is the most effective way of maintaininga constant body temperature in a hot environment and at a high metabolic rate.Consequently, the rate of heat loss by sweat is inuenced by the ambient temperatureand metabolism. The maximum heat loss by sweat occurs when the skin is completelywet and is given by:

(Esw)max = fpclhe(pss pa) Wm2 (1.23)where he is the evaporative heat transfer coefcient, Wm2 Pa1, and fpcl is a per-meation factor of clothing for water vapour which, for porous clothing, may berepresented by [8]:

fpcl = 1/(1 + 0.143hc/hcl) (1.24)


where hc is the convective transfer coefcient and hcl is the heat conductance ofclothing, both in Wm2 K1. The coefcient he in equation (1.23) may be representedby the convective coefcient hc through the Lewis relation, i.e.:

he = 16.7hc

Hence

(Esw)max = 16.7fpclhc(pss pa) Wm2 (1.25)

The effect of clothing absorbing sweat, which is then transmitted to the surface bythe capillary action of the clothing fabric, is not considered in the above equations.When this happens the evaporation of sweat takes place from within the clothing andnot from the skin. This reduces the efciency of sweat in removing excess body heatand more sweat will be required to produce the same heat loss from the skin surface.A sweating efciency may be dened by:

sw = 1/(1 + h/hcl) (1.26)

where h is the combined radiative and convective heat transfer coefcient. The actualheat removed by sweating then becomes:

Esw = sw(Esw)max (1.27)

Fanger [4] has produced a formula based on experimental measurements correlatingEsw to metabolism as follows:

Esw = 0.42(M W 58.15) Wm2 (1.28)

This equation applies to 4 > M > 1. The diffusion loss, Ed, is used to calculate evapo-rative losses when the skin is not wet with sweat. However, Ed is ignored when the skinis completely wet and in this case equation (1.25) or (1.28) is used to calculate Esw.

Respiration heat loss (RES) Inspired air is both warmed and humidied by itspassage through the respiratory system. The sensible and latent heat losses are pro-portional to the volume ow rate of air to the lungs that in turn is proportional to themetabolic rate. The sensible heat loss is given by:

Sres = 0.0014M(34 ta) Wm2 (1.29)

where ta is the ambient air temperature, C. Sres is a small quantity in comparison withthe latent heat loss which is given as:

Lres = 1.72 105 M(5867 pa) Wm2 (1.30)

where pa is the ambient water vapour pressure, Pa. Respiration heat loss is onlysignicant at high activity and under normal sedentary activity it is less than 6Wm2and can be neglected.


1.3 Environmental indicesIn the preceding section it was shown that there are four environmental parameters air temperature, ta, mean radiant temperature, tr, air velocity, v, and water vapourpressure in the air, pa and three personal parameters metabolism, M, work, Wand thermal insulation of clothing, Icl. The heat storage, S, is not included in thepersonal parameters when heat balance between the body and the environment ispresent. During the last 80 years a number of studies have been carried out to deneenvironmental indices in terms of these parameters.

Current environmental indices may be divided into three categories: direct; rationaland empirical. Direct indices are based on the measurement of a simple instrumentthat responds to the environmental variables in a manner similar to that of a human.An example is a globe thermometer which responds to changes in air temperature, airvelocity and radiant temperature. Rational indices are based upon models of humanresponses to the thermal environment taking into consideration thermoregulation andheat exchange between the body and the environment. These models can be used topredict human responses to certain environmental conditions. The operative temper-ature, to, is a rational index which combines the heat exchange between the bodyand the environment by radiation and convection which forms the basis of ASHRAEStandard 55-1992 [9].

An empirical environmental index is based on a model of the energy exchangebetween the human body and the environment which is validated by subjecting a largepopulation sample of known activity and clothing to a range of environmental condi-tions and recording their thermal sensation. The body is represented by a single node,two nodes or multi-node system for the purpose of evaluating the energy exchange.The data is then analysed to assess the effect of each variable and from this an envi-ronmental index is established. Some of the earliest works in this area using a singlenode system was that by Houghton and Yaglou [10], which introduced the effectivetemperature index that was later adopted by ASHRAE, and the well-known work car-ried out by Fanger and his collaborators in Denmark which introduced the predictedpercentage of dissatised, PPD [4]. These and other more recent models are describedin Section 1.4 but in the following sections environmental indices that are based on asingle equivalent temperature are rst introduced.

1.3.1 Effective temperature (ET)The effective temperature index (ET), derived by Houghton and Yaglou [10], com-bines the effect of dry-bulb and wet-bulb temperatures with air movement to yieldequal sensations of warmth or cold. This scale, which was adopted by ASHRAE,was used by HVAC designers for almost 50 years until it was replaced by the new(or revised) effective temperature scale (ET). Criticisms of the old ET scale were thatit overemphasized the effect of humidity in cooler and neutral conditions, underem-phasized its effect in warm conditions, and failed to account for air velocity under hotand humid conditions. This scale is no longer in use now.

Results of tests carried out by Gagge [11] in environmental chambers have shownthat skin wettedness is an excellent predictor of discomfort under transient environ-ments. He then developed the new index (ET) which is dened in terms of theoperative temperature, to, and combines the effects of air and radiant temperatures aswell as water vapour pressure. Gagge correlated his measurements with a two-node


1 h

50%

rh

80%

rh100%

rh

Isotherms forET*, w, discomfort

20% rh

Figure 1.3 The new effective temperature scale-lines of constant ET [12].

model that was used to evaluate skin temperature, skin wettedness and evaporationrate within the zone of operation. A single-node model, such as the PMV/PPD model,is only valid under steady-state thermal conditions and to study transient effects ornon-uniform thermal environments, two- or multi-node models are required. In theET scale, skin temperature, ts, skin wettedness, w, and clothing permeability indexare used to dene the thermal state of a person. The skin wettedness, w, is the ratio ofthe actual evaporative loss at the skin surface to the maximum loss that could occurin the same environment, i.e. when the skin is completely wet. This scale has beenadopted by ASHRAE [12] to replace the old ET scale.

The ASHRAE effective temperature (ET) is the dry-bulb temperature of uniformenclosure at 50% relative humidity (rh) in which people have the same net heatexchange by radiation, convection and evaporation as they do in varying humidities ofthe test environment. The ET scale is based on equal air and operative temperatures,a clothing resistance of 0.6 clo, an air speed of 0.2m s1, a sedentary activity ( 1met)and an exposure time of 1 h. As shown in Figure 1.3 the new effective temperaturefor thermal comfort at these conditions and an rh of 50% is 23.5 C for w = 0.06,i.e. no sweating. The gure also gives the values of ET for different skin wettednessratios. The principle used to develop ET has been extended to higher levels of activ-ities, representing higher levels of sweating. These are discussed in more detail in theASHRAE Handbook of Fundamentals [12].

1.3.2 Operative temperature (to)

TheASHRAE Standard 55-1992 [9] uses the operative temperature as the environmen-tal variable for evaluating thermal comfort at different activity and clothing insulation.


It is dened as the uniform temperature of a radiantly black enclosure in which anoccupant would exchange the same amount of heat by radiation plus convection as inthe actual non-uniform environment. It is expressed by equation (1.13), which denesthe average of the mean radiant and air temperatures, weighted by their respective heattransfer coefcients. The ASHRAE Standard species environmental conditions thatare acceptable to 80% or more of the occupants. It is mainly applicable to sedentaryactivity (1.2met) with normal winter or summer clothing ensembles, i.e. 0.81.2 clowinter clothing or 0.60.8 clo summer clothing. The acceptable ranges of operativetemperature and humidity for winter and summer are dened by the shaded areas inthe psychrometric chart of Figure 1.4 which is based on 10% dissatisfaction [9]. Thegure shows an overlap of the winter and summer zone in the range of to = 2324 Cbecause in this region people in summer clothing would tend to feel slightly cool whilethose in winter garments would be near the slightly warm sensation.

The maximum average air speed in the occupied zone is specied by the Standard as0.15m s1 in winter and 0.25m s1 in summer environments. However, in summer,the comfort zone (see Figure 1.4) may be extended beyond 26 Coperative temperature

15

10

5

0

5

10

Dew

poi

nt te

mpe

ratu

re (

C)

Effe

ctive

tem

pera

ture

ET*

15

20

20

20 ET* 26 ET*

25 300

5

10

Hum

idity

ra

tio (g

kg

1 )

15

70%

100%

rh 60%

50%

30%

Operative temperature (C)

Winter Su

mmer

Figure 1.4 Acceptable ranges of operative temperature and humidity for winter andsummer clothing and sedentary activity, 1.2met [9].


if the average air speed is increased by 0.27m s1 K1 of increased temperature toa maximum of 0.8m s1. Air speeds of 0.8m s1 are unacceptably high in normaloccupancy as loose paper, hair and other objects may be blown about at this speed.

The comfort zone temperature in Figure 1.4 should be decreased when the activitylevel is higher than sedentary, i.e. M > 1.2met. The operative temperature for activityrange 1.2 < M < 3 is obtained using [9]:

to active = to sedentary 3(1 + clo)(met 1.2)

However, the minimum allowable temperature is 15 C.

1.3.3 Dry resultant temperature (tres)

Missenard (seeMcIntyre [5]) was the rst to dene the dry resultant temperature whichwas the equilibrium temperature recorded by a 90-mm-diameter globe thermometer.He also dened the wet resultant temperature as the temperature recorded by a globethermometer that is partly kept moist by wet muslin. This temperature is rarely usednowadays. Missenard selected a 90-mm-diameter sphere so that the ratio of radiantto convective heat exchange with the environment is the same as a human body,i.e. 1 : 0.9. In still air the dry resultant temperature measured by a 90-mm-diameterglobe is:

tres = 0.47ta + 0.53tr (1.31)

However, Missenard gave no correction for air movement and tres was undened inmoving air.

CIBSE [13] has adopted the dry resultant temperature as a recommended environ-mental index for design but allowance has been made for air speed. It is dened as thetemperature recorded by a thermometer at the centre of a blackened globe of 100mmdiameter. The original globe thermometer of Missenard was increased from 90 to100mm so that, at low air speeds, tres becomes the mean of ta and tr, i.e.:

tres = 0.5(ta + tr) (1.32)

Taking the effect of air speed into account, CIBSE denes tres by:

tres =[tr + ta

(10 v)

]/[1 +(10 v)] (1.33)

where v is the air speed, in m s1. Equation (1.33) reduces to equation (1.32) whenv = 0.1m s1, i.e. still air.

The CIBSE Guide [13] gives a table of recommended tres values in still air envi-ronment and a relative humidity range of 4070%. These values range from 13 Cfor heavy activity in factories to 26 C in swimming pools. They implicitly take intoaccount the activity and clothing insulation. The recommended value for ofces is20 C. Where air speeds depart from the still air value (i.e. v > 0.1m s1), CIBSE rec-ommends an elevation of tres according to Figure 1.5, but speeds greater than 0.3m s1are not recommended, except in summer.


1

0

Req

uire

d el

eva

tion

of t r

es

(C)

0 0.2 0.4 0.6Air velocity (m s1)

0.8 1.0

2

3

Figure 1.5 Correction to dry resultant temperature to allow for air movement.

1.4 Thermal comfort models

1.4.1 Fangers model

The environmental indices discussed so far were based on statistical analyses ofextensive experimental data obtained in the laboratory or on site. Therefore, eachindex strictly applies only over the range of physical conditions covered by the exper-iments. Because of the large tasks involved in conducting such experiments, most ofthe measurements were restricted to lightly clothed sedentary subjects and it wouldbe imprudent to extend the range of application of these indices beyond that cov-ered experimentally. Fanger [4] has developed a more fundamental approach thatis based on physical assessment of the body heat exchange with the environmentand supported by extensive tests on human subjects placed under strictly controlledenvironments.

Thermal comfort equation Fanger approached the problem of arriving at a universalenvironmental index from the principle that a person senses his/her own temperatureand not those of the environment. He stipulated three requirements for achievingthermal comfort:

1 The body must be in thermal equilibrium with the environment, i.e. the rate ofheat loss to the environment balances the rate of heat production. This impliesa steady-state condition, i.e. a single-node model is used.

2 Thermal sensation is related to skin temperature and therefore the mean skintemperature, ts, should be at an appropriate level. Measurements have shownthat ts decreases with increasing metabolic rate (equation (1.17)).

3 There should be a preferred rate of sweating, e.g. at sedentary activity peopleprefer not to sweat. The rate of sweating increases with metabolic rate as givenby equation (1.28).

Applying the three thermal comfort requirements and using the heat exchangeexpressions presented in Section 1.2, Fanger reduced the heat balance equation


(equation (1.1)) to the functional relationship:

f (M,W , Icl, ta, tr, v,pa) = 0 (1.34)

As a result, the function f represents a thermal comfort expression which includes thethree personal parameters M, W and Icl and the four environmental parameters ta, tr,v and pa. Theoretically, any combination of the four environmental parameters maybe chosen to satisfy the expression, hence achieving thermal comfort, for a personwith a given clothing (Icl), metabolism (M) and work (W).

On substituting the appropriate formulae for heat exchange between the body andenvironment in the heat balance equation (1.1) as well as the expressions for skintemperature (equation (1.17)) and the evaporative losses (equations (1.21) and (1.28)),Fanger arrived at the now well-known thermal comfort equation given below:

(M W) 3.05 103[5733 6.99(M W) pa] 0.42[(M W) 58.15] 1.7 105M(5867 pa) 0.0014M(34 ta)

= 3.96 108fcl[(tcl + 273)4 (tr + 273)4

] fclhc(tcl ta) (1.35)where the clothing surface temperature is given as:

tcl = 35.7 0.028(M W) 0.155Icl{(M W) 3.05 103 [5733 6.99(M W) pa] 0.42[(M W) 58.15] 1.7 105M(5867 pa) 0.0014M(34 ta)} (1.36)

The convective heat transfer coefcient, hc, in equation (1.35) is calculated usingequations (1.9) and (1.10) and taking the greater of the two values. The clothing areafactor, fcl, is obtained from equation (1.16).

The thermal comfort equation (1.35), which includes the personal and environmen-tal parameters, is too complex to solve except by computer. For this reason, comfortcharts representing solutions to the comfort equation are often used by various combi-nations of two variables to represent a line of a third variable as a constant. Figure 1.6is a typical comfort chart with the mean radiant temperature as the ordinate, the airtemperature as the abscissa and a family of lines of constant air speed. Many other simi-lar charts may be drawn to provide a quick reference for designers of environmentalsystems, see [4, 12].

Many experiments were carried out in the 1970s, mainly on lightly clothed sedentarypeople in environmental chambers, to validate Fangers comfort equation. In parti-cular the inuence of such variables as age, race, seasonal variations, adaptation andbackground colour and noise on the preferred temperature were tested. The variationin the preferred temperature by the people tested was within 1.5K. These experimentsconrmed the predictions of the comfort equation under mainly sedentary activity butthere have been few studies at high activity levels. However, under these conditionspeople are less sensitive to the environmental temperature and any errors thatmay arisefrom applying the comfort equation are not expected to be signicant in assessing thecomfort at a high level of activity.


35

0.20.1

0.530

Sedentary 1 metMedium clothing Icl = 1.0 clo

25

20

15

10

5

00 5 10 15 20

Air temperature (C)30 3525

t mrt =t

Mea

n ra

dian

t tem

pera

ture

(C)

Relative velocity 1.5m

s1

Figure 1.6 Comfort chart for 1met activity, 1 clo and 50% rh (w = 0).

Predicted mean vote (PMV) The thermal comfort equation is only applicable toa person in thermal equilibrium with the environment. If not in equilibrium, the bodywill be under physiological strain to activate the effector mechanism that is necessaryto change the skin temperature to achieve a new heat balance. Fanger assumed thatthe thermal sensation at a given activity level is related to this strain. He used theheat balance equation (1.35) to predict a value for the degree of sensation using hisown experimental data and other published data for different activity levels. Thethermal sensation index that has been adopted by Fanger is based on the seven-pointpsychophysical scale:

cold cool slightly cool neutral slightly warm warm hot3 2 1 0 +1 +2 +3

This index is termed the predicted mean vote (PMV) which is the mean vote onewould expect to get by averaging the thermal sensation vote of a large group of peoplein a given environment. The PMV is a complex mathematical expression involvingactivity, clothing and the four environmental parameters. It is expressed by:

PMV = (0.303e0.036M + 0.028){(M W) 3.05 103[5733 6.99(M W) pa] 0.42[(M W) 58.15] 1.7 105M(5867 pa) 0.0014M(34 ta) 3.96 108fcl[(tcl + 273)4

(tr + 273)4] fclhc(tcl ta)} (1.37)


where

tcl = 35.7 0.028(M W) 0.155 Icl{3.96 108fcl[(tcl + 273)4

(tr + 273)4] + fclhc(tcl ta)} (1.38)The convective heat transfer coefcient, hc, again takes the greater of the values givenby equations (1.9) and (1.10). Equation (1.38) is a transcendental equation whichcan only be solved by an iterative process. Equations (1.37) and (1.38) are moreconveniently solved using a computer. A computer program written in FORTRAN isgiven in [6].

Predicted percentage of dissatised (PPD) From the experimental data available tohim, Fanger correlated the percentage ratio of the people whowere dissatisedwith thethermal environment with the predicted mean vote. This ratio was called the predictedpercentage dissatised and is shown plotted against PMV in Figure 1.7. The gureshows a symmetrical curve with a minimum value of 5% corresponding to the lowestPD subjects, i.e. in an ideal environment. ISO Standard 7730 [6] recommends a PPDlimit of 10% corresponding to 0.5 PMV +0.5. Figure 1.7 may be representedby the expression [6]:

PPD = 100 95 exp{0.03353(PMV)4 + 0.2179(PMV)2} (1.39)The simplied expression below [14]may be used giving good accuracy for |PMV | 2:

PPD = 5 + 20.97|PMV |1.79 (1.40)Although Fangers thermal comfort criterion is the most comprehensive one derived

to date it still has some deciencies. The criterion produces good results for the stan-dard conditions of sedentary activity and light clothing (representing the conditionsat which experimental data were obtained), but it is less satisfactory at more extremeconditions of activity and heavy clothing and in less well-controlled environments,such as free-running buildings. Further work will be required to test the criterion atthese extreme conditions. However, measurements at high activity and in free-runningbuildings where transient thermal conditions exist become less certain as it becomes

8060403020

108654

2.0 1.5 1.0 0.5 0Predicted mean vote

0.5 1.0 1.5 2.0

Pred

icted

per

cent

age

of

diss

atisf

ied

Figure 1.7 Predicted percentage of dissatised as a function of predicted mean vote.


difcult to maintain a constant activity or constant clothing ensembles for sufcientlylong periods.

Application of PMV/PPD For a room with a uniform thermal environment, a sin-gle value of PMV and PPD can be obtained using equations (1.37) and (1.38) andFigure 1.7 or equation (1.39) from the values of ta, tr, v and pa (or relative humidity)at a single point in the room and known values of M and Icl. If the calculated PMVand PPD are outside the recommended range, then a simple adjustment to roomtemperature can restore them to acceptable values.

In most rooms, however, complete thermal uniformity does not exist, owing to thepresence of hot or cold surfaces which produce temperature gradients and varying airvelocity in the room. Hence, to analyse the comfort level in a room, measurement ofthe environmental variables should be carried out at various points in the occupiedzone and these are then used to calculate the distribution of PMV throughout the zone.Normally, the occupied zone is divided into 1 or 0.5m square grids and measurementsof ta, tr and v are made at each grid point for three or four heights above the oor.Typical heights normally chosen are 0.15m (ankle level), 0.6m (back level, seated),1.2m (head level, seated) and 1.8m (head level, standing). The mean radiant temper-ature at each point can be calculated using equation (1.33) from a measurement of thedry resultant temperature, tres, using a 100mm globe thermometer. Thus:

tr =[1 +(10 v)] tres (10 v)ta (1.41)

The water vapour pressure usually has the same value throughout the room and onlyone measurement of the relative humidity in the room is needed to obtain pa. ThePMV values at each grid point are volume averaged to obtain the mean PMV for thewhole of the occupied zone, from which the PPD for the zone is calculated. TypicalPMV contours obtained for a test room with a side-wall air supply (heating) havingone window and two outside walls are shown in Figure 1.8 [14]. It can be seen that thedistribution of PMV for most of the occupied zone is below 0.5 which correspondsto PPD > 10%. The average PMV in the occupied zone is 0.69 which correspondsto a PPD of 15.7%. This is partly due to a low room temperature and partly due to

PMV < 0.75

0.75 < PMV < 0.50

Extract

Supply device

= 0.75= 0.50

Figure 1.8 PMV contours for a room heated by a side-wall air supply.


0.25 < PMV < 0

0 < PMV < 0.25

Extract

Supply device

= 0.00

Figure 1.9 Corrected PMV contours for the room in Figure 1.8.

thermal non-uniformity of the room (i.e. presence of a cold window and two walls).Therefore, this environment would not be suitable for long-term occupancy.

By altering the air temperature level in the room to give a zero average PMV value, itwould be possible to examine the thermal uniformity of the occupied zonemore closely.The effect of moderate changes in the average room air temperature on the velocityeld and on the temperature gradients in the occupied zone will be negligible. Thiscan be achieved by adjusting the supply temperature to produce the necessary changein the mean temperature of the occupied zone. The change in room temperature willproduce an average PMV for the occupied zone of zero. It can be calculated eitherfrom equations (1.37) and (1.38) by setting PMV = 0 or from PMV tables givenby Fanger [4]. To examine the thermal uniformity of the occupied zone the meanPMV value is subtracted from the PMV value at each grid point in the occupied zoneand new or corrected PMV contours are then plotted. From the corrected PMVs thenew PPDs are calculated to give what Fanger called the lowest possible percentage ofdissatised (LPPD). The average of LPPDs in the occupied zone should be less than10% for an acceptable environment. If LPPD > 10% the causes should be examinedclosely to suggest improvements to the air distribution system or thermal insulationof the building or both.

The PMV values in Figure 1.8 were increased by +0.69 to give an average valuefor the occupied zone of zero. The corrected PMV distribution is shown in Figure 1.9which is in the range0.25 < PMV < +0.25 and this is quite acceptable for long-termoccupancy. The supply air temperature in this case was increased by 2.45K and theaverage LPPD was 5.24%.

1.4.2 Multi-node thermal comfort models

The human body is not a simple thermal system that can be assumed homogenousunder different environmental conditions. Different parts of the body react to theenvironment in different ways and are normally at different temperatures. It has beenrecognized that treating the human body as a homogeneous system does only pro-duce accurate predictions under transient environmental conditions or in non-uniform


environments. Consequently, models were developed such that the human body is rep-resented by two or more nodes to allow for inhomogeneous effects in the environment.Such models of the human body are called multi-node models.

The earliest such model was that developed by Gagge [11] in the 1970s which repre-sents the human body by two nodes, an internal core and the skin (which also includeda layer of clothing) with heat transfer taking place between the two nodes by bloodow. Under transient conditions sweating from the skin was also considered in thetotal energy balance. This model was adequate for predicting the overall responseof the body at a moderate activity in a homogeneous environment that can also betransient. Since then a number of models involving, in some cases, a large number ofnodes have been developed for dealing with transient and inhomogeneous environ-mental conditions. Thellier et al. [15] have developed a 25-node model by dividing thebody into six segments (head, trunk, arms, hands, legs and feet) with each segmentmade up of four layers (core, muscle, fat and skin) and the 25th node is the bloodforming a thermal link with the other nodes. The model was validated using experi-mental measurements under transient and inhomogeneous thermal environments withgood correlations.

Conceio [16] has developed a model by dividing the human body into 35 cylindri-cal and spherical elements with each element made up of three layers (internal, centraland external), i.e. using a total of 105 nodes. The internal layer was represented bya vein and an artery except for an element in an extremity (e.g. head, ngers andfoot) when the artery was not considered. The results from this model were used tocompare the local skin temperature at four locations of nude people in an environ-mental chamber under transient thermal environments with good correlations. Thenumber of layers was later extended [17] to 12 with one layer representing the core,two layers for the muscle, two layers for the fat and seven layers for the skin with thepossibility of clothing protection for each element. The extendedmodel was again eval-uated by simulating people exposed to inhomogeneous environments and predictingthe percentage of dissatised in steady- and unsteady-state conditions.

1.4.3 Models-based on eld measurements

The temperature indices described in Section 1.3 have largely been derived fromdata obtained in environmental chambers and laboratory measurements where con-sistent control of the environmental parameters is maintained during the experiments.Although such tests provide accurate evaluation of the inuence of each parameter ora group of parameters on the thermal sensation of people, the subjects involved arenot in their familiar surroundings nor are they engaged in their usual work activitiesthroughout the test periods. The subjects in the laboratory tests may therefore lackthe opportunity to adjust to the thermal environment as people do in normal life.This anomaly can be overcome by conducting eld studies where peoples thermalresponse is investigated in their usual work environments. The results from these mea-surements could then be used to develop thermal comfort models for the occupantsof actual buildings. A number of eld investigations are reported in the literature,e.g. [1825].

Humphreys [18, 19] analysed the results of more than 30 eld studies of thermalcomfort conducted over a period of 40 years in about 12 countries of different climaticconditions. All the investigations were conducted indoors, the majority of which were


24 22 20 18 16 14 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34

14

Monthly mean outdoor temperature (C)

16

B

18

20

22

24Nat

ural t

empe

ratu

re (

C)

26

28

30

32

34 Free-running buildings, regression line AOther buildings, regression line B

A

Figure 1.10 Relationship between the comfort temperature and monthly mean outdoortemperature [18].

in ofces, i.e. light activity. The mean temperatures for these studies range from 17 Cfor English homes in winter to 37 C for ofces in Iraq in the summer. In the latterstudy, the temperaturewas adjusted to 34 C to allow for the abnormally high air veloc-ities present. The temperatures recorded in these investigations were air and/or globetemperatures which differed little in most cases. From the temperature measurementsand the survey results it was possible to establish a comfort temperature, tc, for thesubjects involved. This is the temperature at which the respondent to the survey is ther-mally neutral. Allowing for the high air movement in a hot environment, Humphreysobtained a range of comfort temperatures from 17 to 30 C (see Figure 1.10).

Other, but less wide-ranging, surveys in ofces were carried out by Auliciems [20]in Australia, Fishman and Pimbert [21] in England and Schiller et al. [22] in USA.These studies produced comfort temperatures of 20.523.1 C; 22.0; 22.0 C (winter)and 22.6 C (summer), respectively. In the study by Schiller et al. the winter comforttemperature compares well with the comfort temperature recommended by ASHRAEStandard 55-1992 [9], however the comfort temperature in the summer was lowerthan that recommended by the Standard (24.5 C). This indicates that people tend toprefer cooler conditions in the summer than those suggested by the Standard, whichis largely based on laboratory test data.

The results from these and other eld studies [2225] suggest that people tend toadjust to their environment. However, one must remember that eld data normallyhave lower degrees of correlation than laboratory-based measurements and furtherresearch is needed for an accurate assessment of the thermal acceptability of dailyenvironments in different types of buildings, particularly those that are naturallyventilated.

1.4.4 Adaptive models

Based on the believe that people naturally adapt to their thermal environment bymaking various adjustments to themselves and their surroundings in order to reduce


discomfort and physiological strain and the fact that a single-node model such as thePMV/PPD model does not always predict conditions in free-running buildings, newmodels called Adaptive Models have been developed during the last two decades. Thesingle-node model may be acceptable to buildings with environmental control systemsbut would not apply in the case of unconditioned, e.g. naturally ventilated buildings.In such cases, there will invariably be interaction between occupants behaviour andthe environment to inuence their thermal sensation. Thermal interaction with theenvironment can take different forms, e.g.

Clothing changes Changes of posture and activity Movement between different thermal zones in the building Adapting the available thermal control devices to change the environment.

These interactions are achieved under conscious control which will also affect thehuman physiological regulatory mechanisms. They are also time-dependent and adifferent type of adaptation may occur at different times, e.g.

Short-term adaptation: changing of clothing, posture, environmental control. Long-term adaptation: seasonal changes in clothing, seasonal changes in activity,

changes in furnishing and changes in building control devices.

The adaptive models do not actually predict comfort responses of people but ratherthe thermal conditions under which people are expected to be comfortable in build-ings. Based on eld tests, such as those mentioned earlier, it has been found that anacceptable degree of comfort in residential and ofce buildings could be accomplishedover a very wide temperature range. Another signicant parameter which is inher-ent in most adaptive models is the external weather condition and its inuence onboth peoples perception and on the climate inside. Mean monthly external tempera-ture is used as a parameter in most adaptive models developed thus far to predict anacceptable or a comfort temperature, tc, inside a building. Extensive studies in uncon-ditioned (free-running) buildings [1829] have found a correlation between the meanoutdoor temperature and the indoor comfort temperature. Most of these correla-tions are between the comfort (neutral) temperature tc and the monthly mean outdoortemperature tm. Some of these correlations are given below:

Humphreys relation for free-running buildings [18]:

tc = 12.1 + 0.53tm (1.42)

Humphreys relation for climate-controlled buildings [19]:

tc = 24.2 + 0.43(tm 22) exp{(tm 22)/24

2}

(1.43)


Auliciems relation [20]: tc = 17.6 + 0.31tm (1.44)Williamson et al.s relation [27]: tc = 10.5 + 0.58tm (1.45)Nicols relation [28]: tc = 17.0 + 0.38tm (1.46)De Dear and Bragers relation [29]: tc = 17.8 + 0.31tm (1.47)

A mean expression for free-running buildings based on the relations (1.42) to (1.47)(not including equation (1.43)) is:

tc = 14.0 + 0.45tm (1.48)Equation (1.48) applies to a range of 35 > tm > 10 with an accuracy of 2K in tc.

Adaptive models are useful to provide data that may be used in design calculations,specifying building temperature set points and energy assessments for a building.

1.5 Thermal discomfortAlthough the body may be thermally neutral in a particular environment, i.e. thethermal comfort equation is satised, there may not be thermal comfort if one partof the body is warm and another is cold. This local thermal discomfort can be causedby an asymmetric radiant eld (e.g. cold windows or warm heaters), by contact witha warm or cold oor, by vertical air temperature gradient or by a local convectivecooling of the body (draught). The rst three sources of local discomfort are brieydiscussed here but more emphasis will be directed towards the results from studies onthe effect of draught, since this is a common occurrence in naturally or mechanicallyventilated buildings in winter.

1.5.1 Asymmetrical thermal radiation

Non-uniform or asymmetrical thermal radiation in a space may be caused by cold/hotwindows, walls, ceilings, solar patches in the room or heating panels. In ofce and res-idential buildings asymmetric radiation is mainly due to cold windows and hot ceilingswhereas in factory and food storage buildings it can be due to infrared heaters, hotor cold equipment and so forth. The radiation asymmetry can be described by a para-meter called radiant temperature asymmetry, tpr, which is dened as the differencebetween the plane radiant temperatures of the two opposite sides of a small plane ele-ment. In this context, the plane radiant temperature, tpr, is the uniform temperature ofan enclosure in which the incident radiant ux on one side of a small plane element isthe same as in the existing environment. The plane radiant temperature is a parameterwhich describes the effect of thermal radiation in one direction as compared with themean radiant temperature, tr, which describes radiation from all surrounding surfaces.

For surfaces of high thermal emissivity ( 1), as is the case for most buildingmaterials, the plane radiant temperature, tpr, can be calculated from a knowledge ofthe temperatures of all the surrounding surfaces and their respective view factors withrespect to a small plane element at which tpr is calculated, i.e.:

t4pr = t41Fp1 + t42Fp2 + + t4nFpn (1.49)where tpr = plane radiant temperature (K); tn = temperature of surface n (K); Fpn =view factor between a small plane element and surface, n and

Fpn = 1.


Since the sum of all the view factors is unity, for small differences between thetemperatures of the enclosure surfaces, the plane radiant temperature becomes themean value of the surface temperatures weighted by the respective view factor foreach surface. Hence, equation (1.49) may be linearized as follows:

tpr = t1Fp1 + t2Fp2 + + tnFpn (1.50)

where all the temperatures are in C. The view factors may be obtained from heattransfer textbooks or CIBSE or ASHRAE Handbooks.

The effect of thermal radiation asymmetry on human discomfort has been studiedby several investigators, e.g. McIntyre [5] and Olesen [30]. Figures 1.11 and 1.12show results of a study carried out at the Technical University of Denmark by Fangeret al. on seated and lightly clothed (Icl = 0.6 clo) subjects exposed to a heated ceilingand a cold wall in a climatic chamber [30]. The subjects were otherwise in a state ofthermal equilibrium with the environment in the chamber, i.e. the people were onlydissatised with the radiant asymmetry. The effect of radiant temperature asymmetry,tpr, on the percentage of dissatised due to a general warm or cold sensation isshown in these two gures for the heated ceiling and cold window respectively. It maybe concluded from this study that people are more sensitive to asymmetric radiationcaused by a warm ceiling than by a cold vertical surface. The inuence of a cold ceilingand a warm wall was also studied but people were found to be less sensitive to theseforms of radiant asymmetry.

The work of Fanger et al. on radiant asymmetry resulted in establishing limitsfor plane radiant temperatures by ISO Standard 7730 and ASHRAE Standard 55-1992. The limit in the vertical direction is 5K and in the horizontal direction is 10K.Both of these limits refer to either a small horizontal plane (vertical asymmetry) or asmall vertical plane (horizontal asymmetry) 0.6m above the oor. From Figures 1.11and 1.12 these limits correspond to a percentage of dissatised of 7% and 5% for thevertical and horizontal asymmetry respectively.

8060

Dis

satis

fied

(%)

40

20

10864

2

10 4 8 12Radiant temperature asymmetry, tpr (C)

16 20 24

Figure 1.11 Effect of radiant temperature asymmetry due to a heated ceiling on thepercentage of dissatised.


1.5.2 Vertical air temperature gradient

The air temperature in enclosures is not usually constant but increases vertically fromoor to ceiling or varies horizontally from an external wall or window to an inter-nal wall. The vertical temperature gradient in particular is a source of discomfort toa seated or a standing person. If this gradient is sufciently large, local warm discom-fort can occur at the head, and/or cold discomfort can occur at the feet, although thebody as a whole may be thermally neutral. Figure 1.13 shows the effect of vertical airtemperature gradient on the percentage dissatised from an investigation in a climatechamber [31]. Seated subjects were individually exposed to various air temperaturedifferences between head (1.1m above the oor) and ankle (0.1m above the oor)whilst they were in thermal neutrality. ISO Standard 7730 recommends a maximumair temperature gradient of 3K between 1.1 and 0.1m above the oor, which corre-sponds to a percentage of dissatised of 5% according to Figure 1.13. However, theASHRAE Standard 55-1992 recommends the same temperature gradient between 1.7

8060

Dis

satis

fied

(%)

40

20

10864

2

10 4 8 12Radiant temperature asymmetry, tpr (C)

16 20 24

Figure 1.12 Effect of radiant temperature asymmetry due to a cold wall or window on thepercentage of dissatised.

806040

20

10864

2

1

Dis

satis

fied

(%)

0 2 4 6Air temperature difference between head and feet,

t1.1 t0.1 (K)

8 10 12

Figure 1.13 Effect of vertical air temperature difference between 1.1 and 0.1m above ooron the percentage of dissatised.


and 0.1m above the oor, which relates to a standing, not a seated, person as in thecase of the ISO limit.

1.5.3 Warm or cold oor

In winter, the sensation of cold feet is a very common cause of thermal discomfortamong sedentary people in residences and ofces. This can be due to a low oortemperature and/or the general thermal state of the person. The effect of cold or warmoors on the comfort of people with bare feet and with footwear has been extensivelyinvestigated [3133]. In studying the effect of cold/warm oors the subjects were keptin thermal neutrality, so dissatisfaction was only related to discomfort due to coldor warm feet. For bare feet, the ooring material was signicant in determining anacceptable oor temperature. The recommended temperature range for a carpetedoor is 2128 C giving an expected percentage of dissatised of 15%. The ooringmaterial was not found to be signicant for people with footwear. An optimum oortemperature of 25 C was obtained for a sedentary (seated) person and 23 C fora standing or walking person. When the results for seated and standing persons werecombined, Figure 1.14 was obtained which shows an optimum oor temperature of24 C corresponding to a percentage of dissatised of 6%. A value of 10% dissatisedgives a oor temperature range of 19.528 C.

ISO Standard 7730 recommends a oor temperature range of 1926 C for light,mainly sedentary activity in winter and oor heating system design temperature of29 C. The oor temperature range recommended in ASHRAE Standard 55-1992 is1829 C. These ranges are applicable to peoplewearing appropriate indoor footwear.

1.5.4 Draught

Draught is dened as an undesired local cooling of the human body caused by airmove-ment. It is one of the most frequent causes of complaint in heated or cooled buildingsand in transport vehicles. As mentioned earlier in this chapter the thermoreceptors inthe skin are very sensitive to changes in skin temperature. Draughts produce a coolingeffect of the skin by convection, which is dependent on the temperature difference

Figure 1.14 Effect of oor temperature on the percentage of dissatised with footwear.


between the air and skin, air speed and the magnitude of the uctuations in the airspeed, i.e. turbulence level.

Occupants who are subjected to draughts in winter tend to elevate the room temper-ature to counteract the cooling sensation thereby increasing the energy consumption.In extreme cases ventilation systems are shut off or air supply outlets are blocked offwith a consequent deterioration of the indoor air quality. Draughts are caused by a highmean air speed and/or high turbulence intensity (dened as the ratio of the standarddeviation of the velocity uctuation to the mean velocity). Serious draught complaintscan often occur at mean speeds lower than those recommended by standards (seeSection 1.3) when the turbulence intensity is high.

Effect of mean air speed The few early studies of the effect of draught were mainlyconcerned with the mean air speed and excluded the effect of turbulence. One ofthe earliest investigations was carried out by Houghton et al. [34] in 1938 in whichten male, medium-clothed subjects were each subjected to a jet of air from a ductpositioned a short distance from the back of the neck or the ankles. The fall in skintemperature of the neck and ankles produced by a combination of air speed and airtemperature was correlated with the percentage of subjects dissatised. A 1.8K fall inskin temperature corresponded to 10%dissatised and a 2.4K fall to 20%dissatised.Figure 1.15 shows the variation of air speed with air temperature for the neck regioncorresponding to 20% dissatised. It was also found that the fall in ankle temperaturerequired to produce discomfort was the same as the neck, but larger speeds wereneeded to produce the same effect. The results of this investigation were later adoptedto develop the ASHRAE effective draught temperature (see Chapter 6).

McIntyre [35] carried out a series of experiments on lightly clothed subjectsin a chamber of temperature 21 or 23 C with a jet of low-turbulence air froma 150-mm-diameter nozzle directed at the cheek from a distance of 300mm. Thesubjects were given a 2min exposure and the jet temperature was varied between 17and 23 C. Speeds of 0.35m s1 and higher were detected by all subjects whatever jettemperature was selected (i.e. 1723 C). However, at speeds below 0.35m s1 the

0.4

0.3

0.2

Air v

elo

city

(ms

1 )

0.1

018 19 20

Air temperature (C)21 22 23

Figure 1.15 Variation of air speed with temperature at the back of the neck to produce20% dissatisfaction.


percentage of people who were able to detect the draught increased as the jet temper-ature decreased. Jet speeds of 0.15m s1 or lower at 23 C were undetected by morethan 20% of the subjects.

The effect of low-turbulence draught on the heads of 50 lightly clothed subjects(0.8 clo) was investigated by Mayer and Schwab [36] in a climatic chamber. Theair was supplied to the head of each seated subject from four different directions:vertically upward, vertically downward, horizontally from the front and horizontallyfrom the back. The air temperature was maintained at 23 C and was also equal tothe mean radiant temperature. Five air speeds of 0.1, 0.2, 0.3, 0.4 and 0.45m s1were used in the experiments whilst the turbulence intensity was unchanged at 5%.Each person was exposed to each speed for 10min when he or she was thermallyneutral. The results of percentage of dissatised because of draught against mean airvelocity are shown in Figures 1.16 and 1.17. The percentage of dissatised due todraught at the face directed upward and downward is given in Figure 1.16 whichshows a higher dissatisfaction with an upward draught. Heat transfer measurementsfrom the heated head of a manikin subjected to upward and downward draughts atthe face produced higher heat transfer coefcients for the upward draught than forthe downward draught, which supports the subjects sensation to this type of draught.This may be explained by the presence of an upward natural convective current due tobuoyancy which enhances an upward-directed draught but diminishes a downward-directed draught at the face. The results for draught at the neck were found to liebetween those shown in Figure 1.16 but no difference was found between an upwardand a downward air motion. The dissatisfaction votes for draught towards the faceand draught towards the neck are shown in Figure 1.17. In common with the ndingsof other investigations these results show the neck to be more sensitive to draught thanthe face.

100

80

60

40

20

0 0.1 0.2 0.3Mean air velocity (m s1)(turbulance intensity 5%)

Perc

en

tage

of d

issat

isfie

d be

caus

e of

dra

ugh

t at f

ace

0.4 0.5

Figure 1.16 Effect of air speed on the percentage of dissatised by an upward anddownward draught at the face.


100

80

60

40

20

0 0.1 0.2 0.3Mean air velocity (m s1)(turbulance intensity 5%)

Perc

en

tage

of d

issat

isfie

d be

caus

e of

dra

ugh

t at n

eck/

face

0.4 0.5

Face

Neck

Figure 1.17 Effect of air speed on the percentage of dissatised by a horizontal draughttowards the face and neck.

The experiments on the effect of draught reported so far were conducted with a lowturbulence intensity air motion, i.e. approximately 5%. Measurements in numerousmechanically ventilated buildings have indicated turbulence intensities greater than30% at a height of 1.1m above the oor and in excess of 20% at a height of 0.1mabove the oor [37, 38]. These values refer to the highest mean speeds measured in thespaces investigated (0.3m s1) and much higher intensities were measured at lowermean speeds. Measured air speeds and turbulence intensities in a number of naturallyventilated buildings with different heating systems were found to be slightly lowerthan those in the mechanically ventilated buildings, but the turbulence intensity wasstill more than 25% at a height of 1.1m above the oor for the higher mean speedrange.

The results of these studies prompted Fanger and his associates [39] to repeat theearlier experiments on the effect of draught using turbulence intensities of similar val-ues as those obtained from the eld measurements. They exposed 100 seated subjects(50 male and 50 female) in a climatic chamber to a turbulent air ow typical of realventilated spaces. During the experiments the mean speed was varied from 0.05 to0.4m s1 at air temperatures of 20, 23 and 26 C. The subjects body as a wholewas kept thermally neutral by modifying clothing ensembles to give a range of insu-lation from 0.58 to 0.91 clo. The turbulence intensity of the draught was in the range3065%. The subjects were exposed to draught at the back of the neck but limitedtests were also carried out with the feet and elbow exposed to draught. Measurementsof air velocity and temperature were taken at a distance of 0.15m from the part ofthe body subjected to draught. Figure 1.18 shows the percentage of subjects who feltdraught at the head region as a function of the mean speed at the neck. The headregion comprises the head, neck, shoulders and upper part of the back. The scaleused in the abscissa represents the square root of the speed since the convective heattransfer rate is approximately proportional to the square root of the mean velocity.


Figure 1.18 Effect of mean draught speed on the percentage of dissatised due to the feelingof draught at the head region.

The effect of the temperature of draught on the percentage dissatised is signicant.The line for the percentage of dissatised for a temperature of 23 C is similar to thecurve in Figure 1.17 representing the results from Mayer and Schwab [36] at the sametemperature but much lower turbulence intensity. This is rather surprising since higherturbulence intensity produces a greater cooling sensation as will become more obviouslater in this section.

Based on the above measurements, a draught chart which identies the percentageof subjects dissatised due to draughts in ventilated spaces was produced, as shown inFigure 1.19. The chart excludes discomfort due to draughts of speed below 0.04m s1since it has been argued that at such low speeds the source of discomfort is somethingother than draught. The chart is represented by the expression:

PD = 13, 800{[(v 0.04)/(ta 13.7) + 0.0293]2 8.57 104} (1.51)where PD = percentage of dissatised; v = mean air speed (ms1) and ta = airtemperature (C).

The effect of draught on 50 subjects wearing clothing of 0.86 clo was studied byBerglund and Fobelets [40] in an environmental chamber. The draught was generatedby a variable-speed fan producing an air speed in the range of 0.050.5m s1 at adistance 0.3m upstream of the subjects. The turbulence intensity varied from about44% at the lowest speed down to 5% at the highest speed. The air temperature wasvaried between 19 and 26 C. A linear regression analysis of the test data producedthe following expression for the percentage experiencing draught (PED):

PED = 113(v 0.05) 2.15ta + 46 (1.52)Equations (1.51) and (1.52) are plotted in Figure 1.20 for air temperatures of 20 and24 C for comparison. Equation (1.51) due to Fanger and Christensen [39] produces


Figure 1.19 Draught chart for sedentary people wearing normal indoor clothing.

Equation 1.52

Equation 1.51

20C 24

C

20 C

24 C

[39]

[40]

[40]

Figure 1.20 Comparison of percentage of dissatised due to draught using equations (1.51)and (1.52).

a higher PD than equation (1.52) of Berglund and Fobelets. The difference may beattributed to the larger turbulence intensity used in Fanger and Christensens study(between 30% and 65%) and possibly also the difference in the questionnaires usedin the two investigations.

ASHRAE Standard 55-1992 and the ISO Standard 7730 specify maximum meanair speeds of 0.15m s1 in winter and 0.25m s1 in summer. Taking the higher limit,it can be seen from the draught chart in Figure 1.19 that more than 25% of the


occupants may be dissatised if this limit is used in design. This may explain whydraught complaints are so common in ventilated spaces.

Effect of turbulence The rate of change of skin temperature provides a stimu-lus for the thermoreceptors in the skin to initiate signals to the brain which sensesthis change [41]. Using an electrical analogue computer to simulate the human skin,Madsen [42]measured the heat ow through the receptorswhen the simulated skinwasexposed to sinusoidal air velocities of the same amplitude but varying frequency. Heobtained a maximum heat ow through the thermoreceptors at a frequency of 0.5Hz.Fanger and Pederson [43] conducted experiments on 16 sedentary subjects exposedat the back of the neck to uctuating air speeds with varying amplitudes and frequen-cies and different air temperatures. Each subject was exposed to mean speeds in therange 0.10.3m s1, frequencies of 00.83Hz and turbulence intensities of 6090%.Figure 1.21 shows the results for a mean air speed of 0.3m s1 and in a remarkableagreement with Madsens predictions maximum discomfort occurred at frequenciesbetween 0.3 and 0.5Hz. These studies have shown that a turbulent air velocity is lesscomfortable than a laminar air velocity of the same average value. Mayer [44] carriedout laboratory measurements of the convective heat transfer coefcient of a heatedarticial head under a range of air speeds and turbulence intensities. The heat transfercoefcient increased nearly fourfold when the standard deviation of the uctuatingvelocity increased from 0 (laminar ow) to 0.22m s1.

In occupied spaces the air movement is rather random and not well dened whichis characteristic of turbulent ow. Typical air velocity signals in three rooms withdifferent heating systems are shown in Figure 1.22 [40]. The small velocity uctuationsclose to the oor in the case of the oor heating system are due to the downdraughtover the oor from a window. The higher uctuations in the radiator-heated roomare caused by the interaction of convective currents produced by the radiator withdowndraught from a window, since the signal was measured close to the window.Except for the difference in the turbulence intensity (TI) between the two signals themean velocity is almost the same.

Measurements of air velocities in 12 mechanically ventilated buildings with airchange rates of 4 and 8 per h are reported in [37]. Similar measurements were per-formed in 22 buildings of various types including ofces, schools, lecture rooms,meeting rooms, auditoria, clean rooms, industrial halls and a swimming pool [38].

Frequency (Hz)Notuncomfortable

Uncomfortable

0.00625

Figure 1.21 Degree of discomfort as affected by the frequency of air velocity uctuation.

0.30.20.1

00 60

Warm air Head level v = 0.09 m s

1 sd = 0.06 TI = 0.65

120 180 240 s

0.30.20.1

00 60

RadiatorHead level v = 0.13 m s

1 sd = 0.07 TI = 0.51

120Time (s)

Air v

elo

city

(ms

1 )

180 240 s

0.30.20.1

00 60

Floor heating Ankle level

v = 0.14 sd = 0.02 TI = 0.17

120 180 240 s

Figure 1.22 Air velocity signals in three rooms with different heating systems.

(a)

(b)

Thorshauge [37]Hanzawa et al. [38]Melikov et al. [45]

Figure 1.23 Variation of standard deviation with mean velocity in mechanically andnaturally ventilated buildings: (a) 0.1m above oor and (b) 1.1m above oor.


(a)

(b)

Thorshauge [37]Hanzawa et al. [38]Melikov et al. [45]

Figure 1.24 Variation of turbulence intensity with mean velocity in mechanically andnaturally ventilated buildings: (a) 0.1m above oor and (b) 1.1m above oor.

The air change rates for these buildings were wide ranging: from 1.6 ach for aswimming pool to 25 ach for a clean room.

Velocity measurements were also carried out in eight naturally ventilated and heatedbuildings [45]. The results from these three investigations are shown in Figures 1.23and 1.24 for heights of 0.1 and 1.1m above the oor. Figure 1.23 shows that thestandard deviation of the velocity uctuations increases linearly with increasing meanspeed, with higher uctuations present in the mechanically ventilated buildings. Theturbulence intensity (see Figure 1.24) is higher at low mean speeds and graduallydecreases withmean speed until an asymptotic value is attained. These results representa range of turbulence intensity of about 1580%. The mean speeds in the naturallyventilated buildings were generally lower than in themechanically ventilated buildings.

The effect of turbulence intensity on the sensation of draught was investigatedin experiments carried out on 50 seated subjects (25 male and 25 female) in adraught chamber and each subject exposed to three levels of turbulence intensity(TI < 12%, 20% < TI < 35% and TI > 55%) from behind with a constantair temperature of 23 C [46]. These TI levels have the same ranges as those obtainedin eld measurements. Apart from local draught discomfort the subjects were dressedto achieve thermal neutrality with the environment. In each experiment the sedentarysubjects were exposed to six mean air speeds ranging from 0.05 to 0.4m s1. The airvelocity was measured at the neck level (1.1m above the oor) and at a distance of0.15m from the neck as in previous experiments. The effect of turbulence intensity on


[40]

Low turbulence, TI < 12% Medium turbulence, 20%


Dissatisfied 10%

15%

20%

Figure 1.26 A three-dimensional representation of the three parameters inuencing the percent-age of dissatised due to draught at the head region.

(a) (b)

Figure 1.27 Relationship between mean air speed and

Ventilation of Buildings - Hazim B. Awbi

Documents

Transcript of Ventilation of Buildings - Hazim B. Awbi