Advances in Building Energy

Advances i n Bui l di ng Energy Research
ABER_prelims 6/12/08 1:52 PM Page i

Editor-in-Chief
Mat Santamouris
University of Athens, Greece
Editorial Board
Professor O. Seppanen
Technical University of Helsinki, Finland
Dr M. Sherman
Lawrence Berkeley Laboratory, USA
N. Fintikakis
Associate Editor for UIA, Greece
Dr P. Wouters
Belgian Building Research Institute, Belgium
Professor F. Allard
University of La Rochelle, France
Dr H. Akbari
Lawrence Berkeley Laboratory, USA
Professor E. Maldonado
University of Porto, Portugal
Professor Lee S. E.
University of Singapore, Singapore
Professor A. Papadopoulos
Aristotle University of Thessaloniki, Greece
Professor F. Nicols
University of Strathclyde, UK
Dr E. Erell
Ben Gurion University, Israel
Professor H. Yoshino
Tohuku University, Japan
Professor F. Haghighat
Concordia University, Canada
Professor R. Lamberts
University of Santa Catharina, Brazil
Professor J. Clarke
University of Strathclyde, UK
Professor A. Athienitis
Concordia University, Canada
Professor J. Khedari
University of Bangkok, Thailand
Professor F. Butera
Technical University of Milan, Italy
Professor M. Wilson
Metropolitan University, London
Professor K. Voss
University of Wuppertal, Germany
ABER_prelims 6/12/08 1:52 PM Page ii
Advances i n Bui l di ng Energy Research
Volume 2
Editor-in-Chief Mat Santamouris
publishing for a sustainable future
London Sterling, VA
ABER_prelims 6/12/08 1:52 PM Page iii
First published by Earthscan in the UK and USA in 2008
Copyright Mat Santamouris, 2008
All rights reserved
ISSN 1751-2549 (Print)
ISSN 1756-2201 (Online)
ISBN 978-1-84407-517-1
Typeset by Domex e-Data, India
Printed and bound in the UK by Cromwell Press, Trowbridge
Cover design by Giles Smith
For a full list of publications please contact:
Earthscan
Dunstan House
14a St Cross Street
London EC1N 8XA, UK
Tel: +44 (0)20 7841 1930
Fax: +44 (0)20 7242 1474
Email: earthinfo@earthscan.co.uk
Web: www.earthscan.co.uk
22883 Quicksilver Drive, Sterling, VA 20166-2012, USA
Earthscan publishes in association with the International Institute
for Environment and Development
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-in-Publication Data
Advances in building energy research / editor-in-chief, Mat Santamouris.
v. cm.
Includes bibliographical references.
1. BuildingsEnergy conservation. I. Santamouris, M. (Matheos),
1956-TJ163.5.B84A285 2007
696dc22
2007004087
The paper used for this book is FSC-certified. FSC
(the Forest Stewardship Council) is an international
network to promote responsible management of the
worlds forests.
Advances in Building Energy Research volume 2, 2008.
Advances in Building Energy Research is published annually. Periodicals Postage Paid at
Rahway, NJ. US agent: Mercury International, 365 Blair Road, Avenel, NJ 07001.
POSTMASTER: Address changes to ADVANCES IN BUILDING ENERGY RESEARCH, 365
Blair Road, Avenel, NJ 07001.
ABER_prelims 6/12/08 1:52 PM Page iv
Contents
List of Figures and Tables vii
List of Acronyms and Abbreviations xiii
1 Evolution of Cool-Roof Standards in the US 1
Hashem Akbari and Ronnen Levinson
2 A Review of Innovative Daylighting Systems 33
A. Tsangrassoulis
3 Physically Based Modelling of the Material and
Gaseous Contaminant Interactions in Buildings:
Models, Experimental Data and Future Developments 57
P. Blondeau, A. L. Tiffonnet, F. Allard and F. Haghighat
4 The Application of Urban Climate Research in the
Design of Cities 95
Evyatar Erell
5 Solar Air Conditioning: A Review of Technological and
Market Perspectives 123
S. Oxizidis and A. M. Papadopoulos
6 Experimental Methods in Ventilation 159
M. Sandberg, H. Lundstrm, H. O. Nilsson and H. Stymne
7 A Review of Optical Properties of Shading Devices 211
Athanassios Tzempelikos
Index 241
ABER_prelims 6/12/08 1:52 PM Page v
ABER_prelims 6/12/08 1:52 PM Page vi
List of Figures and Tables
FIGURES
1.1 Locations of the eight ASHRAE-defined climate zones in the US 11
1.2 Locations of the 16 California climate zones 17
1.3 Total savings per unit roof area (cooling equipment savings plus
15-year NPV of energy savings with TDV; US$/m
2
) by California climate
zone, simulated for a prototypical non-residential Title 24 building with
a cool low-sloped roof 18
1.4 Adoption of commercial building energy codes by US states as of
May 2007 23
1.5 Adoption of residential building energy codes by US states as of
May 2007 24
2.1 Simulation of luminance distribution in a room with no shading
system (upper) shaded by fabric roller (middle) and shaded by
Venetian blinds (lower) 36
2.2 Schematic representation of two geometrical characteristics:
cut-off angle for solar rays and angle of acceptance for diffuse radiation 36
2.3 Retrolux blind 37
2.4 Okasolar-W system 38
2.5 Creation and operational principle of a compound parabolic
concentrator (CPC) using two parabolas and definition of
maximum acceptance angle () 38
2.6 Construction scheme of the Microlouver (by Siteco) 39
2.7 Prismatic glazing shading operation 40
2.8 Geometric configuration of a typical prismatic glazing section 40
2.9 Laser-cut light-deflecting panel (LCP) operational principle 41
2.10 Heliostat designed and installed at BartenBach LichtLabor,
Innsbruck, Austria 42
2.11 Inflatable heliostat prototype 43
2.12 Mechanism for alignment of a solar tracker 44
2.13 Definition of acceptance angle for fibre optics 45
2.14 Himawari system 45
2.15 HSL 3100 hybrid lighting system 46
2.16 Solux daylighting system installed in the University of Athens
during the Universal Fibre Optics research programme 46
2.17 Heliobus system installed at Graubunden, Switzerland 48
2.18 Horizontal light pipe 49
2.19 Horizontal light pipe proposed by Canziani et al (2004) 49
2.20 Prototype daylighting system developed by Whitehead (2006) 50
2.21 Anidolic ceiling 50
3.1 Elemental interaction phenomena between gaseous contaminants
and water vapour in poly-dispersed porous materials 85
ABER_prelims 6/12/08 1:52 PM Page vii
viii ADVANCES IN BUILDING ENERGY RESEARCH
5.1 Operational principles of heat pumps: (a) mechanically driven;
(b) thermally driven 127
5.2 Schematic representation of sorption refrigerator 128
5.3 Basic cycle for adsorption and absorption in a (lnP, -1/T) diagram:
(a) closed systems; (b) chemical reaction systems 129
5.4 Temperature profile of a TCHP cycle 132
5.5 Working media paths in a sorption system 133
5.6 S-T diagram for a discontinuous sorption system process 135
5.7 Entropytemperature (ST) diagram for a sorption heat pump 136
5.8 Equivalent Carnot cycles of a sorption heat pump 137
5.9 Cost of solar cooling systems v. vapour compression cooling 142
5.10 Primary energy ratio of solar-assisted absorption and electrically
driven vapour compression chillers 145
6.1 Percentage of people who find air quality just acceptable at a
given ventilation rate 160
6.2 Sketch of the dependence of concentration and velocity upon
ventilation flow rate 161
6.3 The sequence outdoor ambient to the human 161
6.4 Weekly average values of ventilation flow rate as a function of a
season in a house ventilated by mechanical extract ventilation 162
6.5 Distribution of air change rate in similar naturally ventilated dwellings
(unbroken line) and similar mechanically extract ventilated (dashed line) 162
6.6 (Left) A person simulator produces 100W of heat; (right) the
photographer Magnus Mattsson 163
6.7 A thermal manikin is exposed to two different environments,
one actual with non-uniform and one imaginary with uniform
climatic conditions 168
6.8 Supply air terminals: (top) High-velocity supply, medium-velocity
supply; (bottom) low-velocity supply intended for displacement
ventilation, local ventilation 170
6.9 Interpretation of thermal length 171
6.10 A flow dominated by buoyancy (Ar
s
= 0.97): Overhead view of
position of the front on the floor 173
6.11 Velocity at a point in the occupied zone where there is a supply
of isothermal heated air and cold air 174
6.12 Assessment of velocity in a room ventilated by displacement ventilation 174
6.13 Time constant 175
6.14 Concentration depends upon both how the contaminant and the
air are distributed 176
6.15 Kitchen exhaust hood used for laboratory test of capture efficiency:
Visualization with smoke and a laser light sheet; oil is sprayed into
a hot pan by means of a spray-nozzle 177
6.16 Concentration recorded in a room ventilated by displacement
ventilation 178
ABER_prelims 6/12/08 1:52 PM Page viii
LIST OF FIGURES AND TABLES ix
6.17 Bidirectional bulk density flow through a doorway studied in a
two-dimensional model: The cooler air moves towards the left as a
gravity current along the floor 179
6.18 Unidirectional boundary layer flow through an opening 180
6.19 Flow in a stairwell studied with the salt bath method: The density
difference is made visible with the shadowgraph method 180
6.20 A ventilation system, a ventilation system with infiltration, and purging
flow rate 181
6.21 Definition of displacement height (D) and roughness height (z
0
) 182
6.22 Wind tunnel test of a hybrid-ventilated school at a scale of 1:200:
In the background is shown the spires and the roughness elements 183
6.23 Visualization by the sand erosion method of wind-driven flow
through openings 184
6.24 The heated full-scale thermal manikin with 33 individually controlled zones:
This manikin was especially constructed for climate evaluation in 1991 184
6.25 Dummies simulating the heat generated by a person 185
6.26 Manikin with 20 individually controlled segments 186
6.27 Manikin with 120 independently controlled zones 187
6.28 Illumination system, flow markers and a digital camera 195
6.29 Particle image velocimetry (PIV) recording and interrogation process 196
6.30 Orientation of the measuring plane 197
6.31 Supply of warm air from a ceiling diffuser: Temperature distribution
recorded with an infrared camera 197
6.32 Components in a passive system 199
6.33 Comfort zone diagrams adapted for Comfortina-type manikins 203
7.1 Glazing system with N layers 214
7.2 Between-glazing and exterior venetian blinds 217
7.3 Discretization of venetian blinds 218
7.4 Direct-to-direct transmittance 220
7.5 Directly irradiated parts of venetian blind 220
7.6 View of a cell formed by adjacent slats showing how the cell is divided
into segments, s
i
, for the calculation of direct solar transmittance;
and side view of a cell showing case where some of the direct solar
passes between adjacent slats without touching either of them 222
7.7 Side view of horizontal blinds for calculating optical properties for sky
and ground diffuse radiation 223
7.8 Description of cut-off angles 226
7.9 Examples of interior roller shades 229
7.10 Spectral and directional transmission properties (different incidence
angles) of a common roller shade 230
7.11 Screen model rendering of intersecting orthogonal crossed cylinders 231
7.12 Categorization of fabrics for draperies/shades 233
7.13 Bidirectional reflection and transmission distribution functions 234
7.14 Detection of the light transmitted through a sample 235
ABER_prelims 6/12/08 1:52 PM Page ix
TABLES
1.1 Cool-roof energy savings measured in six California non-residential
buildings 3
1.2 Thermal resistance of insulation below a cool roof (solar reflectance
0.55) that yields the same annual energy expenditure (cost at
US$0.66/therm and US$0.08/kWh) as a low, medium or high level
of insulation (3,11 or 38ft
2
h F BTU
1
) below a conventional roof
(solar reflectance 0.20) 9
1.3 Roof thermal transmittance (U-factor) multipliers for cool roofs on
buildings other than low-rise residential buildings 10
1.4 Roof thermal transmittance (U-factor) multipliers for cool roofs on
buildings other than low-rise residential buildings 10
1.5 Multiplier by which the thermal transmittance (U-factor) of a residential
roof assembly can be increased without raising annual energy use when
the solar reflectance of the roofs surface is increased to 0.55 (cool)
from 0.10 (conventional) 13
1.6 Ceiling thermal transmittance (U-factor) multipliers for residential
cool roofs: It is possible that these multipliers should be replaced
by their reciprocal to yield values less than or equal to unity 14
1.7 Ceiling thermal resistances (ft
2
h F BTU
1
) prescribed by ASHRAE
Standard 90.2-2007 for ceilings under conventional (non-cool) and
cool residential roofs, derived from ASHRAE Standard 90.2-2007:
Reduced requirements for cool-roofed buildings are shaded 15
1.8 Life-cycle cool-roof savings per unit roof area (cooling equipment
savings plus 30-year NPV of energy savings with TDV; US$/m
2
)
by California climate zone for a non-residential building with a
steep-sloped roof 19
1.9 Life-cycle cool-roof savings per unit roof area (cooling equipment
2
)
by California climate zone for a residential building with a
low-sloped roof 20
1.10 Life-cycle cool-roof savings per unit roof footprint area (cooling
equipment savings plus 30-year NPV of energy savings with TDV;
US$/m
2
) by California climate zone for a residential building with
a steep-sloped roof 21
1.11 Solar reflectance and thermal emittance requirements of a 2007
residential cool-roof rebate programme administered by two
California utilities (Pacific Gas & Electric and Southern California Edison) 28
3.1 Representative adsorption isotherm models for building applications 66
3.2 Compilation of sorption and diffusion coefficients: Data fit with
one-phase models (K K
H
) 69
3.3 Compilation of sorption and diffusion coefficients: Data fit with
two-phase models (K K
p
) 72
5.1 Main features of possible solar sorption cooling technologies 138
5.2 Physical characteristics of several chillers of 400kW cooling capacity 143
x ADVANCES IN BUILDING ENERGY RESEARCH
ABER_prelims 6/12/08 1:52 PM Page x
LIST OF FIGURES AND TABLES xi
5.3 Strategic factors of the external environment 148
5.4 Strategic factors of the internal environment 149
5.5 Strategies for development (TOWS matrix) 151
6.1 Description of the connection between the measured quantities
and equivalent temperature 167
ABER_prelims 6/12/08 1:52 PM Page xi
ABER_prelims 6/12/08 1:52 PM Page xii
List of Acronyms and
Abbreviations
ABL atmospheric boundary layer
ACT acoustic ceiling tile
A/D analogue/digital
AS aspect ratio
ASHRAE American Society of Heating, Refrigerating and Air-Conditioning Engineers
BTDF bidirectional transmittance distribution function
BT(R)DF bidirectional transmission (or reflection) distribution function
C degrees Celsius
CAD computer-aided design
CASE Codes and Standards Enhancement
C
6
F
6
hexafluorobenzene
CCD charge-coupled device
CEC California Energy Commission
CET corrected effective temperature
CFC chlorofluorocarbon
CFD computational fluid dynamic(s)
CIE Commission Internationale de lEclairage
clo average clothing
cm/s centimetres per second
CO
2
carbon dioxide
COP coefficient of performance
CPC compound parabolic concentrator
CRP carpet
CSP computer-simulated person
CSUMM Colorado State University Mesoscale Model (a hydrostatic,
primitive-equation, three-dimensional Eulerian model of the Earths
boundary layer climate)
2D two dimensional
3D three dimensional
dB decibel
DDC direct digital control
DEM digital elevation model
DSM demand-side management
EPA US Environmental Protection Agency
EPF energy performance factor
ESCO energy service company
ET effective temperature; equivalent temperature
F degrees Fahrenheit
ft
2
square feet
g gram
GAX generator/absorber heat exchanger
ABER_prelims 6/12/08 1:52 PM Page xiii
GB gypsum board
GHG greenhouse gas
GW gigawatt
GWI global warming impact
GWP global warming potential
HCFC hydro-chlorofluorocarbon
HFC hydrofluorocarbon
HOE holographic optical element
HVAC heating, ventilation and air conditioning
H/W height-to-width ratio
Hz hertz
IAQ indoor air quality
IAS ideal adsorbed solution
IDC integration of directional coefficients
IECC International Energy Conservation Code
IR infrared
ISO International Organization for Standardization
J joule
K kelvin
kg kilogram
kHz kilohertz
kJ kilojoule
kJ/K kilojoule per kelvin
kW kilowatt
kWh kilowatt hours
kWh/m
2
kilowatt hours per square metre
lb ft
2
pounds per square foot
LCP laser-cut light-deflecting panel
LDA laser Doppler anemometer
LEED Leadership in Energy and Environmental Design
LES large eddy simulation
LiBr lithium bromide
m
2
square metre
MHF mean heat flux
MJ megajoule
MJ/m
2
megajoules per square metre
m/s metres per second
MTV mean thermal vote
MW megawatt
N
2
O nitrous oxide
NG natural gas
NH
3
ammonia
NPV net present value
O&M operation and maintenance
ODP ozone-depletion potential
OSB oriented strand board
xiv ADVANCES IN BUILDING ENERGY RESEARCH
ABER_prelims 6/12/08 1:52 PM Page xiv
Pa pascal
PB particle board
PC personal computer
PCI park cool island
PER primary energy ratio
PFC polyfluorocarbon
PG&E Pacific Gas & Electric
PHF perceived heat flux
PIER Public Interest Energy Research Program
PIV particle image velocimetry
PLY plywood
PMMA polymethyl methacrylate
PMV predicted mean vote
ppb parts per billion
PPD predicted percentage dissatisfied
PSV particle streak velocimetry
PV photovoltaics
R&D research and development
RES renewable energy source(s)
RH relative humidity
RT resultant temperature
s second
s
1
per second
SAC solar-assisted cooling
SCE Southern California Edison
SET standard effective temperature
SF
6
sulphur hexafluoride
SMUD Sacramento Municipal Utility District
SRI solar reflectance index
SWOT strengths, weaknesses, opportunities, threats
TCHP thermochemical heat pump
TDV time-dependent valuation
TES thermal energy storage
TOWS threats, opportunities, weaknesses and strengths
TVOC total amounts of VOC
micro
m micrometre
UHI urban heat island
US United States
UV ultraviolet
VCC vapour compression cooling
VOC volatile organic compound
W watt
W/m
2
watts per square metre
Wh/m
2
watt hours per square metre
WIS Advanced Window Information System
LIST OF ACRONYMS AND ABBREVIATIONS xv
ABER_prelims 6/12/08 1:52 PM Page xv
ABER_prelims 6/12/08 1:52 PM Page xvi
Keywords cool roofs; solar reflectance; thermal emittance; solar reflectance index; building energy
efficiency standards; ASHRAE 90.1; ASHRAE 90.2; California Title 24; International Energy Conservation
Code (IECC); Leadership in Energy and Environmental Design (LEED); Energy Star; Florida Building Code
(FBC); Revised Ordinances of Honololu (ROH); City of Chicago Energy Conservation Code
INTRODUCTION
Roofs that have high solar reflectance (high ability to reflect sunlight: spectrum 0.32.5m)
and high thermal emittance (high ability to emit thermal radiation: spectrum 480m) stay
ADVANCES IN BUILDING ENERGY RESEARCH 2008 VOLUME 2 PAGES 132
Abstract
Roofs that have high solar reflectance and high thermal emittance stay cool in the sun. A
roof with lower thermal emittance but exceptionally high solar reflectance can also stay
cool in the sun. Substituting a cool roof for a non-cool roof decreases cooling electricity
use, cooling power demand and cooling equipment capacity requirements, while slightly
increasing heating energy consumption. Cool roofs can also lower the citywide ambient
air temperature in summer, slowing ozone formation and increasing human comfort.
Provisions for cool roofs in energy efficiency standards can promote the building-
and climate-appropriate use of cool roofing technologies. Cool-roof requirements are
designed to reduce building energy use, while energy-neutral cool-roof credits permit
the use of less energy-efficient components (e.g. larger windows) in a building that has
energy-saving cool roofs. Both types of measures can reduce the life-cycle cost of a
building (initial cost plus lifetime energy cost).
Since 1999, several widely used building energy efficiency standards, including
ASHRAE 90.1, ASHRAE 90.2, the International Energy Conservation Code and Californias
Title 24 have adopted cool-roof credits or requirements. This chapter reviews the
technical development of cool-roof provisions in the ASHRAE 90.1, ASHRAE 90.2 and
California Title 24 Standards, and discusses the treatment of cool roofs in other
standards and energy efficiency programmes. The techniques used to develop the
ASHRAE and Title 24 cool-roof provisions can be used as models to address cool roofs
in building energy efficiency standards worldwide.
1
Evolution of Cool-Roof Standards
in the US
Hashem Akbari and Ronnen Levinson
doi:10.3763/aber.2008.0201 2008 Earthscan ISSN 1751-2549 (Print), 1756-2201 (Online) www.earthscanjournals.com
ABER_CH_01 7/11/08 10:19 AM Page 1
cool in the sun. The same is true of roofs with lower thermal emittance but exceptionally
high solar reflectance. Roofs that stay cool in the sun by minimizing solar absorption and
maximizing thermal emission are hereafter denoted cool roofs.
BENEFITS OF COOL ROOFS
Low roof temperatures lessen the flow of heat from the roof into the building, reducing
the need for space cooling electricity in conditioned buildings. Since building heat gain
through the roof peaks in mid to late afternoon, when summer electricity use is highest,
cool roofs can also reduce peak electricity demand. Prior research has indicated that
savings are greatest for buildings located in climates with long cooling seasons and short
heating seasons, particularly those buildings that have distribution ducts in the plenum
(Akbari, 1998; Konopacki and Akbari, 1998; Akbari et al, 1999).
Cool roofs transfer less heat to the outdoor environment than do warm roofs (Taha,
2001). The resulting decrease in outside air temperature can slow urban smog formation
and improve human health and outdoor comfort. Reduced thermal stress may also
increase the lifetime of cool roofs, lessening maintenance and waste (Akbari et al, 2001).
Earlier studies have measured daily air-conditioning energy savings and peak-power
demand reduction from the use of cool roofs on buildings in several warm weather
climates, including California, Florida and Texas. Cool roofs on non-residential buildings
typically yielded measured summertime daily cooling energy savings and peak-power
demand reductions of 10 to 30 per cent, though values have been as low as 2 per cent
and as high as 40 per cent (see Table 1.1) (Konopacki et al, 1998). For example:
Konopacki et al (1998) measured summer daily cooling energy savings per unit roof
area of 67, 39 and 4Wh/m
2
(18, 13 and 2 per cent, respectively) for three California
non-residential buildings two medical offices in Davis and Gilroy and a retail store
in San Jose. Assuming an aged solar reflectance of 0.55, estimated annualized
cooling energy savings (daily savings number of cooling days per year) were 6.4,
3.7 and 0.6kWh/m
2
(16, 11 and 2 per cent, respectively), while peak-power demand
reductions per unit roof area were 3.3, 2.4 and 1.6W/m
2
(12, 8 and 9 per cent,
respectively).
Hildebrandt et al (1998) measured summer daily cooling energy savings of 23, 44
and 25Wh/m
2
(17, 26 and 39 per cent, respectively) in an office, a museum and a
hospice in Sacramento, California. Estimated annualized cooling energy savings were
1.3, 2.6 and 2.2kWh/m
2
, assuming an aged solar reflectance of 0.55.
Konopacki and Akbari (2001) estimated summer daily cooling energy savings of
39Wh/m
2
(11 per cent) and a peak-power demand reduction of 3.8W/m
2
(14 per cent)
in a large retail store in Austin, Texas. Estimated annualized cooling energy savings
were 6.8kWh/m
2
Parker et al (1998a) measured summer daily cooling energy savings of 44Wh/m
2
(25 per cent) and a peak-power demand reduction of 6W/m
2
(30 per cent) for a
school building in Florida. Estimated annualized cooling energy savings were
4.7kWh/m
2
2 HASHEM AKBARI AND RONNEN LEVINSON
ABER_CH_01 7/11/08 10:19 AM Page 2
Evolution of Cool-Roof Standards in the US 3
T
A
B
L
E

1
.
1
C
o
o
l
-
r
o
o
f

e
n
e
r
g
y

s
a
v
i
n
g
s

m
e
a
s
u
r
e
d

i
n

s
i
x

C
a
l
i
f
o
r
n
i
a

n
o
n
-
r
e
s
i
d
e
n
t
i
a
l

b
u
i
l
d
i
n
g
s

(
A
)

D
A
V
I
S
(
B
)

G
I
L
R
O
Y
(
C
)

S
A
N

J
O
S
E

(
D
)

S
A
C
R
A
M
E
N
T
O

(
E
)

S
A
C
R
A
M
E
N
T
O

(
F
)

S
A
C
R
A
M
E
N
T
O

M
E
D
I
C
A
L

O
F
F
I
C
E
M
E
D
I
C
A
L

O
F
F
I
C
E
R
E
T
A
I
L

S
T
O
R
E
O
F
F
I
C
E
M
U
S
E
U
M
H
O
S
P
I
C
E
R
o
o
f
A
r
e
a

(
m
2
)
2
9
5
0
2
2
1
0
3
0
6
0
2
2
9
0
4
5
5
5
5
7
T
y
p
e
B
u
i
l
t
-
u
p
B
u
i
l
t
-
u
p
B
u
i
l
t
-
u
p
F
o
u
r
-
p
l
y

w
i
t
h

B
u
i
l
t
-
u
p

g
r
a
v
e
l
C
o
m
p
o
s
i
t
e
c
a
p
s
h
e
e
t
s
h
i
n
g
l
e
/
f
l
a
t

b
u
i
l
t

u
p
M
a
t
e
r
i
a
l
A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

A
s
p
h
a
l
t

c
a
p
s
h
e
e
t

w
i
t
h

l
i
g
h
t

g
r
e
y

w
i
t
h

l
i
g
h
t

g
r
e
y

w
i
t
h

t
a
n

g
r
a
n
u
l
e
s
w
i
t
h

l
i
g
h
t

g
r
e
y

w
i
t
h

l
i
g
h
t

g
r
e
y

w
i
t
h

t
a
n

g
r
a
n
u
l
e
s
g
r
a
n
u
l
e
s
g
r
a
n
u
l
e
s

g
r
a
n
u
l
e
s
g
r
a
n
u
l
e
s
I
n
s
u
l
a
t
i
o
n

t
h
e
r
m
a
l

1
.
4

(
R
-
8

r
i
g
i
d
)
3
.
4

(
R
-
1
9

R
a
d
i
a
n
t

b
a
r
r
i
e
r
3
.
4

(
R
-
1
9
)
N
o
n
e
1
.
9

(
R
-
1
1
)
r
e
s
i
s
t
a
n
c
e

(
m
2
K
/
W
)
f
i
b
r
e
g
l
a
s
s
)
S
t
r
u
c
t
u
r
e
M
e
t
a
l

d
e
c
k
W
o
o
d

d
e
c
k
W
o
o
d

d
e
c
k
M
e
t
a
l

d
e
c
k
W
o
o
d

d
e
c
k
W
o
o
d

d
e
c
k
P
l
e
n
u
m

t
y
p
e
R
e
t
u
r
n

p
l
e
n
u
m
V
e
n
t
i
l
a
t
e
d

p
l
e
n
u
m
V
e
n
t
i
l
a
t
e
d

p
l
e
n
u
m
R
e
t
u
r
n

p
l
e
n
u
m
V
e
n
t
i
l
a
t
e
d

p
l
e
n
u
m
V
e
n
t
i
l
a
t
e
d

p
l
e
n
u
m
C
e
i
l
i
n
g

t
y
p
e
T
i
l
e
s
T
i
l
e
s
T
i
l
e
s
T
i
l
e
s
T
i
l
e
s
T
i
l
e
s
P
r
e
-
c
o
a
t
i
n
g

2
5
%

g
r
a
n
u
l
e

l
o
s
s

2
5
%

g
r
a
n
u
l
e

l
o
s
s

2
5
%

g
r
a
n
u
l
e

l
o
s
s

2
5
%

g
r
a
n
u
l
e

l
o
s
s

2
5
%

g
r
a
n
u
l
e

l
o
s
s

2
5
%

g
r
a
n
u
l
e

l
o
s
s

c
o
n
d
i
t
i
o
n
a
n
d

b
u
b
b
l
i
n
g
a
n
d

c
r
a
c
k
i
n
g
a
n
d

c
r
a
c
k
i
n
g
a
n
d

b
u
b
b
l
i
n
g
a
n
d

c
r
a
c
k
i
n
g
a
n
d

c
r
a
c
k
i
n
g
P
r
e
-
c
o
a
t
i
n
g

0
.
2
4
0
.
2
5
0
.
1
6
0
.
2
4
0
.
2
5
0
.
1
6
s
o
l
a
r

r
e
f
l
e
c
t
a
n
c
e
P
o
s
t
-
c
o
a
t
i
n
g

0
.
6
0
0
.
6
0
0
.
6
0
0
.
6
0
0
.
6
0
0
.
6
0
s
o
l
a
r

r
e
f
l
e
c
t
a
n
c
e

a
f
t
e
r

o
n
e

y
e
a
r
D
e
g
r
a
d
e
d

0
.
5
5
0
.
5
5
0
.
5
5
0
.
5
5
0
.
5
5
0
.
5
5
(
w
e
a
t
h
e
r
e
d
)

s
o
l
a
r

r
e
f
l
e
c
t
a
n
c
e
ABER_CH_01 7/11/08 10:19 AM Page 3
T
A
B
L
E

1
.
1
C
o
o
l
-
r
o
o
f

e
n
e
r
g
y

s
a
v
i
n
g
s

m
e
a
s
u
r
e
d

i
n

s
i
x

C
a
l
i
f
o
r
n
i
a

n
o
n
-
r
e
s
i
d
e
n
t
i
a
l

b
u
i
l
d
i
n
g
s

(
C
o
n
t
d
)
(
A
)

D
A
V
I
S
(
B
)

G
I
L
R
O
Y
(
C
)

S
A
N

J
O
S
E

(
D
)

S
A
C
R
A
M
E
N
T
O

(
E
)

S
A
C
R
A
M
E
N
T
O

(
F
)

S
A
C
R
A
M
E
N
T
O

M
E
D
I
C
A
L

O
F
F
I
C
E
M
E
D
I
C
A
L

O
F
F
I
C
E
R
E
T
A
I
L

S
T
O
R
E
O
F
F
I
C
E
M
U
S
E
U
M
H
O
S
P
I
C
E
S
u
p
p
l
y

d
u
c
t
I
n
s
u
l
a
t
i
o
n

t
h
e
r
m
a
l

U
n
i
n
s
u
l
a
t
e
d
0
.
8
1

(
R
-
4
.
6
)
0
.
3
5

(
R
-
2
)
U
n
i
n
s
u
l
a
t
e
d
0
.
8
1

(
R
-
4
.
6
)
0
.
3
5

(
R
-
2
)
r
e
s
i
s
t
a
n
c
e

(
m
2
K
/
W
)
L
o
c
a
t
i
o
n
C
o
n
d
i
t
i
o
n
e
d

s
p
a
c
e
P
l
e
n
u
m
P
l
e
n
u
m
C
o
n
d
i
t
i
o
n
e
d

s
p
a
c
e
P
l
e
n
u
m
P
l
e
n
u
m
S
a
v
i
n
g
s
M
e
a
s
u
r
e
d

d
a
i
l
y

6
7

(
1
8
%
)
3
9

(
1
3
%
)
4

(
2
%
)
2
3

(
1
7
%
)
4
4

(
2
6
%
)
2
5

(
3
9
%
)
c
o
o
l
i
n
g

e
n
e
r
g
y

s
a
v
i
n
g
s

(
W
h
/
m
2
)
C
o
o
l
i
n
g

d
a
y
s
/
y
e
a
r
1
1
0
1
1
0
1
6
5
1
6
5
1
6
5
1
6
5
D
e
g
r
a
d
e
d

a
n
n
u
a
l

6
.
4
3
.
7
0
.
6
1
.
3
2
.
6
2
.
2
c
o
o
l
i
n
g

e
n
e
r
g
y

s
a
v
i
n
g
s

(
k
W
h
/
m
2
)
D
e
g
r
a
d
e
d

p
e
a
k
-

3
.
3
2
.
4
1
.
6
n
/
a
n
/
a
n
/
a
p
o
w
e
r

d
e
m
a
n
d

r
e
d
u
c
t
i
o
n

(
W
/
m
2
)
N
o
t
e
:
n
/
a

=

n
o
t

a
v
a
i
l
a
b
l
e
S
o
u
r
c
e
:
(
a
)
(
c
)

K
o
n
o
p
a
c
k
i

e
t

a
l

(
1
9
9
8
)
;

(
d
)
(
f
)

H
i
l
d
e
b
r
a
n
d
t

e
t

a
l

(
1
9
9
8
)
ABER_CH_01 7/11/08 10:19 AM Page 4
Cool roofs on residential buildings yielded measured summertime cooling energy savings
and peak-power demand reductions that ranged from negligible to 80 per cent. For
example:
In a study of 11 Florida homes, Parker et al (1998b) measured average summer daily
cooling energy savings of 7.7kWh (19 per cent) per house and an average peak-
power reduction of 0.55kW (22 per cent) per house. The daily electricity savings in
individual houses ranged from 0.9kWh (0.2 per cent) to 15.4kWh (45 per cent) and
the peak-power reduction ranged from 0.2kW (12 per cent) to 0.99kW (23 per cent).
These initial savings resulted from increasing the solar reflectance of the shingle
roofs to 0.70 from 0.08.
Akbari et al (1997) measured summer daily energy savings of 14Wh/m
2
(80 per cent)
and peak demand savings of 3.8W/m
2
(30 per cent) in a single-story, flat-roofed
house in Sacramento. The savings resulted from increasing the solar reflectance of
the roof to 0.70 from 0.18.
NEED FOR COOL-ROOF STANDARDS
It is difficult for a building owner to assess the influence of roof properties on the lifetime
cost of heating and cooling energy, which depends upon:
climate- and building-specific hourly uses of heating and cooling energy;
hourly valuations of energy;
the time value (discounting) of money; and
the service life of the roof.
Building owners may also be unaware of the societal benefits of cool roofs, such as lower
peak-power demand (reducing the likelihood of power failures on hot days) and lower
outdoor air temperatures (improving comfort and slowing the formation of smog). Hence,
without cool-roof standards, owners will tend to choose roofs that minimize initial
construction cost, rather than the aggregate cost of construction and lifetime energy
consumption.
Provisions for cool roofs in energy efficiency standards promote their building- and
climate-appropriate use, and also stimulate the development of energy-saving cool-roof
technologies. For example, several manufacturers have introduced novel cool non-white
roofing materials, including fibreglass asphalt shingles, clay and concrete tiles, and metal
products (Akbari and Desjarlais, 2005). The development and long-term performance of
cool-roof technologies are described by Akbari et al (2005a, 2005b), Levinson et al (2005b,
2005c, 2005d, 2007) and Berdahl et al (2008).
TYPES OF REQUIREMENTS IN STANDARDS
Building energy efficiency standards typically specify both mandatory and prescriptive
requirements. Mandatory requirements, such as practices for the proper installation of
insulation, must be implemented in all buildings subject to the standard. A prescriptive
requirement typically specifies the characteristics or performance of a single component
ABER_CH_01 7/11/08 10:19 AM Page 5
of the building (e.g. the thermal resistance of duct insulation) or of a group of components
(e.g. the thermal transmittance of a roof assembly).
All buildings regulated by a particular standard must achieve either prescriptive or
performance compliance. A proposed building that meets all applicable mandatory and
prescriptive requirements will be in prescriptive compliance with the standard. Alternatively,
a proposed building can achieve performance compliance with the standard if:
it satisfies all applicable mandatory requirements; and
its annual energy use does not exceed that of a comparable design (also known as a
standard or reference) building that achieves prescriptive compliance.
Prescribing the use of cool roofs in building energy efficiency standards promotes the
cost-effective use of cool roofs to save energy, reduce peak-power demand and improve
air quality. Another option is to credit, rather than prescribe, the use of cool roofs. This can
allow more flexibility in building design, permitting the use of less energy-efficient
components (e.g. larger windows) in a building that has energy-saving cool roofs. Such
credits are energy neutral, but may still decrease peak-power demand and improve air
quality. They may also reduce the initial cost of the building.
This chapter reviews the technical development of cool-roof provisions in the ASHRAE
90.1, ASHRAE 90.2 and California Title 24 building energy efficiency standards, and
discusses the treatment of cool roofs in several other standards and energy efficiency
programmes. The techniques used to develop the ASHRAE and Title 24 cool-roof
provisions can be used as models to address cool roofs in building energy efficiency
standards worldwide.
DEVELOPMENT OF STANDARDS
In 1999, the American Society of Heating, Refrigerating and Air-Conditioning Engineers
(ASHRAE) first credited cool roofs on non-residential and high-rise residential buildings in
ASHRAE Standard 90.1-1999: Energy Standards for Buildings Except Low-Rise Residential
Buildings (ASHRAE, 1999). In 2001, ASHRAE amended its standards for low-rise
residential buildings to credit cool roofs, implementing the revisions three years later in
ASHRAE Standard 90.2-2004: Energy-Efficient Design of Low-Rise Residential Buildings
(ASHRAE, 2004b).
In January 2001, the state of California followed the ASHRAE approach by crediting in
its Title 24 Energy Efficiency Standards for Residential and Non-Residential Buildings the
use of cool roofing products on non-residential buildings with low-sloped roofs (CEC,
2001). In 2005, California upgraded Title 24 to prescribe minimum values of solar
reflectance and thermal emittance for low-sloped roofs (i.e. roofs with a ratio of rise to run
not exceeding 2:12) on non-residential buildings (CEC, 2006). As of June 2007, California
is evaluating proposals to prescribe in the 2008 Title 24 standards minimum values of
solar reflectance and thermal emittance for low-sloped roofs on non-residential buildings,
and for both low-sloped and steep-sloped roofs on residential buildings. Other localities,
such as Florida and Chicago, have adopted custom cool-roof requirements in their
energy codes.
ABER_CH_01 7/11/08 10:19 AM Page 6
Note that the building envelope requirements of the ASHRAE and California Title 24
standards apply only to envelope components (e.g. roofs) that enclose conditioned
spaces.
ASHRAE STANDARD 90.1
Recognizing the potential for solar-reflective roofs to reduce the conditioning energy use
of commercial buildings, the ASHRAE Standard 90.1 committee organized a task force in
1997 to analyse the energy-saving benefits of cool roofs in different climates, and to
propose modifications to the standard to account for the effect of roof solar reflectance.
This section summarizes the cool-roof analysis performed for ASHRAE Standard 90.1
(Akbari et al, 1998).
Cool roofs versus roof insulation
Solar-reflective roofs with high thermal emittance stay cool in the sun, reducing the flow
of heat from the roof to the buildings conditioned space. This can decrease the need for
cooling energy in summer and increase heating energy use in winter. The winter heating
energy penalty is usually smaller than the summer cooling energy savings because in
winter the sun is low, the days are short, the skies are often cloudy, and heating occurs
mainly in early morning and early evening.
Roof insulation impedes the flow of heat between the roof and the conditioned space,
slowing both heating of the building when the roof is warmer than the inside air and
cooling of the building when the roof is cooler than the inside air. One can develop an
energy-neutral trade-off between the solar reflectance of the roofs surface and the
thermal transmittance of the roof assembly.
Survey of the radiative properties of roofing products
The task force surveyed the solar reflectance and thermal emittance of various roofing
products, including fibreglass asphalt shingles, elastomeric coatings, membranes, metal
panels, clay tiles and concrete tiles. The solar reflectance of shingles ranged from 0.03 to
0.26, with most between 0.10 and 0.15. Roofing membranes, such as black single-ply
roofing, smooth bitumen, grey single-ply roofing, and nominally white (actually grey)
granule-surfaced bitumen exhibited solar reflectances of 0.06, 0.06, 0.23 and 0.26,
respectively. Gravel roofs had solar reflectances of about 0.12 to 0.34, depending upon
gravel colour. The thermal emittances of these non-metallic surfaces were about 0.8 to
0.9. Bare, shiny metal roofs have higher solar reflectance (about 0.60), but their low
thermal emittances (about 0.10) make them as hot as a dark roof when wind speed is low.
These data suggest that a conventional dark low-sloped roof could be conservatively
assumed to have a solar reflectance of about 0.20.
An asphalt-aluminium coating has a solar reflectance in the range of 0.30 to 0.61. A
freshly applied white elastomeric coating typically has a solar reflectance of 0.60 to 0.85,
while that of a new white single-ply roofing membrane usually exceeds 0.70. Soiling and
weathering typically reduce the solar reflectances of elastomeric and membrane white
roofs by about 0.10 to 0.15 within the first year, with little change in solar reflectance
thereafter. It was therefore assumed that a cool low-sloped roof should have an initial
ABER_CH_01 7/11/08 10:19 AM Page 7
solar reflectance not less than 0.70, an aged solar reflectance not less than 0.55, and a
thermal emittance not less than 0.80.
Building energy simulations
The DOE-2.1E building energy simulation programme (DOE-2, 2007) was used to estimate
the influences of the solar reflectance of the roofs surface and the thermal resistance
of the roofs insulation on the conditioning energy uses of residential and non-residential
buildings with low-sloped roofs. The residential model applies to guest rooms in hotels,
patient rooms in hospitals and high-rise residential apartments. The buildings were
simulated with electric cooling; gas heating; low, medium and high levels of roof
insulation (insulation thermal resistances of R = 3, 11 or 38ft
2
h F BTU
1
); a roof thermal
emittance of 0.80; and roof solar reflectances of = 0.05, 0.15, 0.45 and 0.75. The 19
simulation climates ranged from very hot to very cold.
The thermal transmittance, or U-factor, of a roof assembly is the reciprocal of the
sum of the thermal resistances of the roof assembly (including insulation) and its
surrounding air films. In each climate, simulated values of annual cooling energy use
(kWh), annual heating energy use (therms) and annual conditioning energy expenditure
(US dollars at US$0.08/kWh and US$0.66/therm) were each regressed to the solar
absorptance, = 1 , of the roofs surface, and to the thermal transmittance, U, of the
roof assembly.
Each climate-specific energy use or energy expenditure E was well fit by an
expression of the form:
. [1]
This result was used to find combinations of roof solar absorptance and roof assembly
thermal transmittance U that yield equal annual energy expenditure E. It was also used to
determine the extent to which increasing the solar reflectance of a roof from 0.20
(conventional roof) to 0.55 (aged cool roof) can decrease the need for roof insulation
without increasing annual energy use. Table 1.2 shows for various cities the thermal
resistance of insulation required under a cool roof to achieve the same annual energy use
as low, medium and high levels of insulation below a conventional roof. The use of a cool
roof reduced the need for insulation in all cases, though more so in hot climates than in
cold climates.
Cool-roof credits
ASHRAE Standard 90.1 permits both prescriptive and performance (energy cost budget)
compliance. ASHRAE Standard 90.1-1999 includes two forms of credits for a cool roof,
defined as one with a minimum initial solar reflectance of 0.70 and a minimum thermal
emittance of 0.75. For performance compliance, a cool roof on a proposed building is
assigned a solar absorptance of 0.55 (solar reflectance of 0.45), while a non-cool roof on
a proposed building and the roof on the design building are each assigned a solar
absorptance of 0.70 (solar reflectance of 0.30). We note that the solar reflectance of 0.45
assigned to a cool proposed roof is less than that assumed in the preceding analysis; this
E C C C U C U = + + +
0 1 2 3

ABER_CH_01 7/11/08 10:19 AM Page 8
may be a typographical error. The standard should be corrected to assign a solar
reflectance of 0.55 (solar absorptance of 0.45) to a cool proposed roof.
For prescriptive compliance, ASHRAE Standard 90.1-1999 (section 5.3.1.1)
approximates the benefits of a cool-roof surface by adjusting the thermal transmittance
of the proposed roof assembly. The standard includes the following adjustment to the
thermal transmittance of the roof assembly with a cool surface:
[2]
where U
roof adj
is the adjusted roof thermal transmittance for use in demonstrating compliance;
U
roof proposed
is the thermal transmittance of the proposed roof, as designed; and F is the
roof thermal transmittance multiplier from Table 1.3. Since F 1, this has the effect of
decreasing the assumed thermal transmittance (increasing the assumed thermal
resistance) of a proposed roofing assembly with a cool surface.
U U F
roof adj roof proposed
=
TABLE 1.2 Thermal resistance of insulation below a cool roof (solar reflectance 0.55) that yields the
same annual energy expenditure (cost at US$0.08kWh and US$0.66/therm) as a low, medium or high level
of insulation (3, 11 or 38ft
2
h F BTU
1
) below a conventional roof (solar reflectance 0.20)
LOCATION CDD50
a
RESIDENTIAL BUILDING NON-RESIDENTIAL BUILDING
R = 3 R = 11 R = 38 R = 3 R = 11 R = 38
Honolulu, HI 9804 0.0 3.5 19.4 0.0 3.9 16.5
Miami, FL 9261 0.1 4.3 21.0 0.3 4.5 18.2
Tampa, FL 8022 0.4 4.9 21.4 0.5 5.0 19.2
Phoenix, AZ 7858 0.9 6.2 26.2 0.9 5.9 22.0
Lake Charles, LA 6860 0.7 5.6 24.2 0.7 5.5 21.3
San Diego, CA 5170 0.1 4.2 19.6 0.2 4.2 16.5
Fort Worth, TX 6200 1.1 6.7 27.5 1.1 6.4 23.5
San Bernardino, CA 4854 0.9 6.0 23.9 0.9 5.7 21.1
Atlanta, GA 4922 1.0 6.4 25.9 1.0 6.1 22.4
San Francisco, CA 2486 1.7 8.1 31.2 1.3 6.9 24.8
Amarillo, TX 4262 1.4 7.4 29.5 1.4 7.1 26.1
Portland, OR 2320 2.0 8.6 31.4 1.8 8.0 27.4
Seattle, WA 1716 2.2 9.2 33.9 1.9 8.3 28.6
Boise, ID 2748 1.9 8.4 31.7 1.8 7.9 27.7
Vancouver, Canada 1468 2.2 9.1 32.0 1.9 8.3 28.5
Minneapolis, MN 2701 2.4 9.7 34.5 2.1 8.9 31.2
Halifax, Canada 1447 2.4 9.7 35.1 2.2 9.2 32.2
Bismarck, ND 2222 2.3 9.4 33.5 2.2 9.0 31.5
Anchorage, AK 684 3.0 10.9 36.8 2.6 10.0 34.4
Edmonton, Canada 880 2.8 10.4 36.0 2.5 9.7 33.3
Note: a = cooling degree days based on 50F
Source: Akbari et al (1998)
ABER_CH_01 7/11/08 10:19 AM Page 9
Revisions
ASHRAE Standard 90.1-2001 (ASHRAE, 2001) retains the same provisions for cool-roof
credits. The current version of this standard, ASHRAE Standard 90.1-2004 (ASHRAE,
2004a), tabulates thermal transmittance multipliers by US climate zone (see Figure 1.1),
rather than by heating degree days (see Table 1.4).
ASHRAE STANDARD 90.2
The procedure for incorporating the effect of roof solar reflectance in the ASHRAE
Standard 90.2 residential standards was similar to that followed for ASHRAE Standard 90.1.
This section summarizes the cool-roof analysis performed for ASHRAE Standard 90.2
(Akbari et al, 2000).
Building-energy simulations
The Standard 90.2 task group used DOE-2.1E to simulate, in 29 climates, the influence of
a solar-reflective roof on the heating and cooling energy uses of a residential building
prototype used in previous 90.2 analyses and described by Akbari et al (2000). Parameters
varied in the prototype buildings included presence or absence of an attic; duct location
(attic or conditioned space);
1
thermal resistance of duct insulation (2, 4 or 6ft
2
h F BTU
1
);
roof solar reflectance (0.10, 0.25, 0.50 or 0.75); and thermal resistance of ceiling insulation
TABLE 1.3 Roof thermal transmittance (U-factor) multipliers for cool roofs on
buildings other than low-rise residential buildings
HDD65
a
(HDD18)
b
ROOF U-FACTOR MULTIPLIER
0900 (0500) 0.77
9011800 (5011000) 0.83
18012700 (10011500) 0.85
27993600 (15012000) 0.86
> 3600 (> 2000) 1.00
Notes: a = heating degree days based on 65F
b = heating degree days based on 18C
Source: ASHRAE 90.1-1999 (ASHRAE, 1999, Table 5.3.3.1B)
TABLE 1.4 Roof thermal transmittance (U-factor) multipliers
for cool roofs on buildings other than low-rise residential
buildings
CLIMATE ZONE ROOF U-FACTOR MULTIPLIER
1 0.77
2 0.83
3 0.85
48 1
Source: ASHRAE 90.1-2004 (ASHRAE, 2004a, Table 5.5.3.1)
ABER_CH_01 7/11/08 10:19 AM Page 10
Z
o
n
e

1

i
n
c
l
u
d
e
s
H
a
w
a
i
i
,

G
u
a
m
,
P
u
e
r
t
o

R
i
c
o
,
a
n
d

t
h
e

V
i
r
g
i
n

I
s
l
a
n
d
s
A
l
l

o
f

A
l
a
s
k
a

i
n

Z
o
n
e

7
e
x
c
e
p
t

f
o
r

t
h
e

f
o
l
l
o
w
i
n
g
B
o
r
o
u
g
h
s

i
n

Z
o
n
e

8
:
B
e
t
h
e
l
D
e
l
l
i
n
g
h
a
m
F
a
i
r
b
a
n
k
s

N
.

S
t
a
r
N
o
m
e
N
o
r
t
h

S
l
o
p
e
N
o
r
t
h
w
e
s
t

A
r
c
t
i
c
S
o
u
t
h
e
a
s
t

F
a
i
r
b
a
n
k
s
W
a
d
e

H
a
m
p
t
o
n
Y
u
k
o
n
-
K
o
y
u
k
u
k
S
o
u
r
c
e
:
A
S
H
R
A
E

(
2
0
0
7
)
F
I
G
U
R
E

1
.
1
L
o
c
a
t
i
o
n
s

o
f

t
h
e

e
i
g
h
t

A
S
H
R
A
E
-
d
e
f
i
n
e
d

c
l
i
m
a
t
e

z
o
n
e
s

i
n

t
h
e

U
S
ABER_CH_01 7/11/08 10:19 AM Page 11
(1, 11, 19, 30 or 49ft
2
h F BTU
1
). Buildings were cooled electrically, and heated with an
electric heat pump, electric resistance or natural gas. All roofs were assigned a thermal
emittance of 0.80.
In each climate, simulated values of annual cooling energy use (kWh), annual heating
energy use (therms) and annual conditioning energy expenditure (US dollars at
US$0.08/kWh and US$0.69/therm) were each regressed to the solar absorptance of the
roofs surface, and to the thermal transmittance U of the roof assembly.
Each climate-specific energy use or energy cost E was well fit by an expression of the
form:
[3]
This result was used to find combinations of roof solar absorptance and roof-assembly
thermal transmittance that yield equal annual energy cost. It was also used to determine
the multiplier by which the thermal transmittance of a roof assembly can be increased
without raising annual energy use when the solar reflectance of the roofs surface is
increased to 0.55 (cool white steep-sloped roof) from 0.10 (conventional dark steep-
sloped roof). Table 1.5 shows this multiplier for various prototype configurations in the
29 US cities simulated. Multipliers exceeded unity in all but four cities, and were at least
0.94 in all cities that is, all but four cities exhibited positive savings, and the penalties in
cold climates were small.
Cool-roof credits
ASHRAE Standard 90.2-2004 permits both prescriptive and performance (energy cost
budget) compliance. The standard includes two forms of credit for a cool roof, defined as
a roof with a minimum initial solar reflectance of 0.65 and a minimum thermal emittance
of 0.75; and/or a solar reflectance index (SRI) of at least 75 calculated under the medium
wind-speed conditions specified by ASTM Standard E1980: Standard Practice for
Calculating Solar Reflectance Index of Horizontal and Low-Sloped Opaque Surfaces
(ASTM, 1998). SRI is a relative index of the steady-state temperature of a roofs surface
on a typical summer afternoon. SRI is defined to be zero for a clean black roof (solar
reflectance 0.05, thermal emittance 0.90) and 100 for a clean white roof (solar reflectance
0.80, thermal emittance 0.90). Thus, warm surfaces have low SRI, and cool surfaces have
high SRI.
For performance compliance (section 8.8.3.1), a cool roof on a proposed building
is assigned its actual solar absorptance, or possibly a solar absorptance of 0.35; the
standards language is ambiguous. A non-cool roof on a proposed building and the
roof on the design building are each assigned a solar absorptance of 0.20 (solar
reflectance of 0.80). However, the authors believe the latter to be a typographical
error, since the logical value would be a solar absorptance of 0.80 (solar reflectance
of 0.20).
For prescriptive compliance (section 5.5), ASHRAE Standard 90.2-2004 approximates
the benefits of a cool-roof surface by adjusting the thermal transmittance of the proposed
ceiling (the authors believe that ceiling may actually mean roof assembly). The standard
E C C U C U C U = + + +
0 1 2
2
3
.
ABER_CH_01 7/11/08 10:19 AM Page 12
TABLE 1.5 Multiplier by which the thermal transmittance (U-factor) of a residential roof assembly can be
increased without raising annual energy use when the solar reflectance of the roofs surface is increased
to 0.55 (cool) from 0.10 (conventional)
LOCATION MULTIPLIER
CDD65
a
HDD65
b
DUCTS DUCTS IN ROOFS WITH
IN ATTICS CONDITIONED NO ATTICS
SPACE
Honolulu, HI 4329 0 2.62 2.62 1.69
Miami, FL 4127 141 2.19 2.19 1.67
Brownsville, TX 3563 659 1.60 1.60 1.61
Phoenix, AZ 3815 1154 1.53 1.53 2.39
Jacksonville, FL 2657 1437 1.52 1.52 1.59
Tucson, AZ 2763 1554 1.49 1.49 2.76
Lake Charles, LA 2624 1683 1.46 1.46 2.02
El Paso, TX 2046 2597 1.42 1.42 1.68
Los Angeles, CA 943 1309 1.38 1.38 1.64
San Diego, CA 766 1076 1.37 1.37 1.69
Las Vegas, NV 3067 2293 1.37 1.37 1.65
Fresno, CA 1884 2602 1.34 1.34 1.56
Charleston, SC 2010 2209 1.33 1.33 1.58
Fort Worth, TX 2415 2304 1.31 1.31 1.64
Fort Smith, AZ 1895 3351 1.24 1.24 1.63
Sacramento, CA 1144 2794 1.22 1.22 1.61
Albuquerque, NM 1211 4361 1.19 1.19 1.40
Los Angeles, CA 470 1291 1.16 1.16 1.55
St Louis, MO 1437 5021 1.11 1.11 1.50
Washington, DC 1044 5233 1.09 1.09 1.23
Dodge, KS 1371 5353 1.09 1.09 1.34
North Omaha, NE 1051 6047 1.08 1.08 1.27
Denver, CO 623 6007 1.06 1.06 1.33
Winnemucca, NV 604 6444 1.06 1.06 1.28
New York, NY 1002 5090 1.05 1.05 1.28
Bismarck, ND 408 8666 1.02 1.02 1.25
Redmond, OR 194 6732 1.01 1.01 1.26
Madison, WI 521 7495 1.01 1.01 1.18
Seattle, WA 127 4867 0.97 0.97 1.12
Fairbanks, AK 29 14095 0.97 0.97 1.09
San Francisco, CA 69 3239 0.94 0.94 0.98
Notes: Ducts in the attics have R-4 insulation (4 ft
2
h F BTU
1
); ducts in the conditioned space are uninsulated
a = cooling degree days based on 65F
b = heating degree days based on 65F
Source: Akbari et al (2000)
ABER_CH_01 7/11/08 10:19 AM Page 13
includes the following adjustment to the thermal transmittance of the ceiling under a cool
roof:
[4]
where U
ceiling adj
is the adjusted ceiling thermal transmittance for use in demonstrating
compliance; U
ceiling proposed
is the thermal transmittance of the proposed ceiling, as designed;
and Multiplier is the ceiling thermal transmittance multiplier from Table 1.6. Since
Multiplier 1, this has the effect of increasing the assumed thermal transmittance
(decreasing the assumed thermal resistance) of a proposed roofing assembly with a cool
surface. Hence, we believe the multiplier values to be in error. It is possible that each value
in Table 1.6 should be replaced by its reciprocal to yield multipliers that do not exceed unity.
Revisions
The current version of this standard, ASHRAE Standard 90.2-2007 (ASHRAE, 2007), retains
the same cool-roof credits for performance compliance. However, the cool-roof credits
for prescriptive compliance have been modified. Rather than specify ceiling thermal
transmittance multipliers, the new standard prescribes reduced thermal resistances for
ceilings under cool roofs in climate zones 1 to 3 (see Table 1.7).
CALIFORNIA TITLE 24 STANDARDS
In 2001, cool-roof credits were added to Californias Title 24 Standards. The standards
were upgraded in 2005 to prescriptively require cool roofs on non-residential buildings
with low-sloped roofs. The California Energy Commission is currently (2007)
considering adding prescriptive cool-roof requirements for all other buildings to the
2008 standards.
Cool-roof credits (2001)
A Codes and Standards Enhancement (CASE) study prepared in 2000 by the Pacific Gas
& Electric company indicated that cool roofs would cost-effectively save energy and
U U Multiplier
ceiling adj ceiling proposed
=
TABLE 1.6 Ceiling thermal transmittance (U-factor) multipliers for residential
cool roofs: It is possible that these multipliers should be replaced by their
reciprocal to yield values less than or equal to unity
ZONE CEILINGS WITH ATTICS CEILINGS WITHOUT ATTICS
1 1.50 1.30
2 1.25 1.30
3 1.20 1.20
4 1.15 1.20
5 1.10 1.10
6,7,8 1.00 1.00
Source: ASHRAE 90.2-2004 (ASHRAE, 2004b, Table 5.5)
ABER_CH_01 7/11/08 10:19 AM Page 14
reduce peak-power demand in California (Eilert, 2000). In January 2001, the state of
California followed the approach of ASHRAE Standards 90.1 and 90.2 by adding a
cool-roof credits to Title 24 (CEC, 2001). Roofs are considered cool if they have an initial
solar reflectance not less than 0.70 and a thermal emittance not less than 0.75. An
exception lowers this minimum initial solar reflectance requirement to 0.40 for tile roofs.
Cool roofs were not made a prescriptive requirement. For performance compliance, a
cool roof on a proposed building was assigned a solar absorptance of 0.45 (solar
reflectance of 0.55). The roof of a standard (design) building was assigned a solar
absorptance of 0.70 (solar reflectance of 0.30), as was a non-cool roof on a proposed
building.
Prescriptive requirements for low-sloped roofs on non-residential
buildings (2005)
In 2002, the Berkeley Lab Heat Island Group began to investigate the prescriptive
requirement in Title 24 of cool roofs for non-residential buildings with low-sloped roofs.
The analysis approach was similar to that used to develop ASHRAE Standards 90.1 and
90.2. Steps included reviewing the physics of cool roofs; reviewing measurements of
cool-roof energy savings reported in the literature; investigating the market availability of
cool roofs; surveying cost premiums (if any) for cool roofs; reviewing roofing material
durability; investigating the environmental consequences of cool roofs; and performing
hourly simulations of building energy use to estimate the energy and peak-power demand
savings potentials of cool roofs (Levinson et al, 2005a).
A review of low-sloped roofing technologies indicated that cool options (solar-
reflective products or coatings) were available for nearly all types of low-sloped roofs,
TABLE 1.7 Ceiling thermal resistances (ft
2
h F BTU
1
) prescribed by ASHRAE Standard 90.2-2007 for
ceilings under conventional (non-cool) and cool residential roofs, derived from ASHRAE Standard
90.2-2007: Reduced requirements for cool-roofed buildings are shaded
CLIMATE CEILINGS WITH ATTICS CEILINGS WITHOUT ATTICS
ZONE WOOD FRAME STEEL FRAME WOOD FRAME STEEL FRAME
CONVEN- COOL CONVEN- COOL CONVEN- COOL CONVEN- COOL
TIONAL ROOF TIONAL ROOF TIONAL ROOF TIONAL ROOF
ROOF ROOF ROOF ROOF
1 30 20 30 20 13 10 19 10
2 30 24 30 24 22 17 19 17
3 30 27 30 27 22 18 22 18
4 38 38 38 38 22 22 22 22
5 43 43 43 43 26 26 26 26
6 49 49 49 49 38 38 38 38
7 49 49 49 49 38 38 38 38
8 52 52 52 52 38 38 38 38
Source: ASHRAE Standard 90.2-2007 (ASRAE, 2007, Tables 5.2 and 5.6.1)
ABER_CH_01 7/11/08 10:19 AM Page 15
including the three dominant products: built-up roofing, modified bitumen and single-ply
membrane. A cool roof was defined as a roof with:
an initial thermal emittance not less than 0.75 and an initial solar reflectance not less
than 0.70; and/or
an initial thermal emittance (
initial
) less than 0.75 and an initial solar reflectance not
less than 0.70 + 0.34 (0.75 (
initial
).
The second term in this expression is the solar-reflectance premium required to ensure
that under ASTM E1980 medium wind-speed conditions, the aged (weathered)
temperature of a roof with low thermal emittance will not exceed that of an aged
(weathered) cool roof with high thermal emittance.
DOE-2.1E building energy simulations performed in Californias 16 climate zones (see
Figure 1.2) indicate that the use of a cool roof on a prototypical California Title 24 non-
residential building with a low-sloped roof yields average annual cooling energy savings
of 3.2kWh/m
2
, average annual natural gas deficits of 5.6MJ/m
2
, average source energy
savings of 30MJ/m
2
, and average peak-power demand savings of 2.1W/m
2
. Total savings
initial cost savings from downsizing cooling equipment plus the 15-year net present value
(NPV) of energy savings with time-dependent valuation (TDV) ranged from US$1.90/m
2
to US$8.30/m
2
(see Figure 1.3).
The typical cost premium for a cool low-sloped roof is US$0.0/m
2
to US$2.2/m
2
. Cool
roofs with premiums up to US$2.2/m
2
are expected to be cost-effective in climate
zones 2 to 16; those with premiums not exceeding US$1.9/m
2
are expected also to be
cost-effective in climate zone 1. Therefore, the 2005 California Title 24 Standards
adopted a cool-roof prescriptive requirement in all California climate zones for non-
residential buildings with low-sloped roofs. Non-residential buildings with low-sloped
roofs that do not meet this new prescriptive requirement may still achieve performance
compliance.
Proposed prescriptive requirements for steep-sloped non-residential
roofs, steep-sloped residential roofs and low-sloped residential
roofs (2008)
In 2005, the Berkeley Lab Heat Island Group began to investigate the merits of adding
to the 2008 Title 24 Standards prescriptive requirements for the use of cool roofs on
all other types of buildings, including non-residential buildings with steep-sloped
roofs, residential buildings with steep-sloped roofs and residential buildings with
low-sloped roofs. The methodology was similar to that used to consider the
prescriptive requirement in the 2005 Title 24 Standards of cool low-sloped roofs for
non-residential buildings. In these 2008 cycle analyses, the MICROPAS building
energy simulation tool (MICROPAS, 2007) was used to simulate the hourly energy use
of prototypical residential and small non-residential buildings (Akbari et al, 2006; Wray
et al, 2006).
ABER_CH_01 7/11/08 10:19 AM Page 16
Steep-sloped roofs on non-residential buildings
Berkeley Lab developed a non-residential prototype building that prescriptively complies
with the 2005 Title 24 Standards. The energy use of this building was simulated with
conventional and cool versions of three different steep-sloped (5:12) roofing products:
fibreglass asphalt shingle, concrete tile and polymer-coated metal. Each conventional
product had a solar reflectance of 0.10. The cool shingle had a solar reflectance of 0.25,
Source: Eley Associates
FIGURE 1.2 Locations of the 16 California climate zones
ABER_CH_01 7/11/08 10:19 AM Page 17
while the cool tile and cool metal products each had a solar reflectance of 0.40. All
products were assigned a thermal emittance of 0.90.
Total savings per unit roof area defined as initial cost savings from downsizing
cooling equipment, plus the 30-year NPV of TDV energy savings ranged from US$2.8/m
2
to US$24.4/m
2
across Californias 16 climate zones (see Table 1.8). The typical cost
premium for a cool steep-sloped roofing product is US$0.0/m
2
to US$2.2/m
2
. Cool roofs
with premiums of up to US$2.2/m
2
are expected to be cost-effective in all 16 climate
zones. At the time of writing, California is considering including in its 2008 Title 24
Standards a prescriptive cool-roof requirement in all climate zones for non-residential
buildings with steep-sloped roofs.
Low-sloped roofs on residential buildings
Berkeley Lab developed a residential prototype building that prescriptively complies with the
2005 Title 24 Standards. The energy use of this building was simulated with conventional
( = 0.20) and cool ( = 0.55) versions of a low-sloped (horizontal) built-up roof.
While the 2005 Title 24 Standards prescriptively require a sub-roof radiant barrier for
residential buildings in some climate zones (2, 4 and 8 to 15), radiant barriers are not
1.9
4.0
4.1
4.3
5.1
5.2
5.7
5.8
6.6
6.6
6.9 6.9
7.2
7.6
7.7
8.3
0
1
2
3
4
5
6
7
8
9
10
1 3 5 16 4 11 12 2 10 7 141 31 5 6 9 8
California climate zone
T
o
t
a
l

s
a
v
i
n
g
s

[
e
q
u
i
p
m
e
n
t

+

1
5
-
y
e
a
r

N
P
V

e
n
e
r
g
y
]

(
U
S
$
/
m
2
)
C
o
s
t

p
r
e
m
i
u
m
(
0

-

2
.
2

U
S
$
/
m
2
)
Source: Levinson et al (2005a)
FIGURE 1.3 Total savings per unit roof area (cooling equipment savings plus 15-year NPV of energy
savings with TDV; US$/m
2
) by California climate zone, simulated for a prototypical non-residential Title
24 building with a cool low-sloped roof
ABER_CH_01 7/11/08 10:19 AM Page 18
usually installed in houses with low-sloped roofs (in climates zones where radiant
barriers are prescriptively required, a house without a radiant barrier would have to
demonstrate performance compliance). Without a radiant barrier (typical of low-sloped
roofs, in general, and pre-2000 construction, in particular), total savings defined as
initial cost savings from downsizing cooling equipment, plus the 30-year NPV of TDV
energy savings ranged from US$2.4/m
2
to US$8.2/m
2
across Californias 16 climate
zones (see Table 1.9). With a radiant barrier, the savings ranged from US$2.5/m
2
to
US$4.7/m
2
. The negative savings occurred in coastal California climate zones with
minimal summertime cooling requirements. The presence of a sub-roof radiant barrier
reduces cool-roof energy savings, just as the presence of a cool roof reduces radiant-
barrier energy savings.
The typical cost premium for a cool roof is US$0.0/m
2
to US$2.2/m
2
. Cool roofs with
premiums of up to US$2.2/m
2
are expected to be cost-effective in some climates zones.
At the time of writing, California is considering including in its 2008 Title 24 Standards a
prescriptive cool-roof requirement in hot Central Valley climates for residential buildings
with low-sloped roofs.
TABLE 1.8 Total savings per unit roof area (cooling equipment
2
) by
California climate zone for a non-residential building with a steep-
sloped roof
CLIMATE ZONE SHINGLE TILE METAL
1 2.8 4.8 5.8
2 5.9 10.4 12.0
3 4.7 8.2 9.6
4 6.3 11.3 12.8
5 5.1 8.9 10.4
6 10.1 17.8 20.7
7 9.1 15.8 18.7
8 11.6 20.4 23.9
9 11.9 20.8 24.4
10 8.6 15.6 17.5
11 7.4 13.5 15.1
12 6.9 12.5 14.0
13 8.3 15.3 17.0
14 9.1 16.6 18.5
15 11.4 21.2 23.3
16 5.2 9.2 10.6
Note: Cool fibreglass asphalt shingles were assigned a solar reflectance of 0.25; cool
concrete tiles and metal products, 0.40; and all conventional products, 0.10
Source: original material for this chapter
ABER_CH_01 7/11/08 10:19 AM Page 19
Steep-sloped roofs on residential buildings
Berkeley Lab developed a residential prototype building that prescriptively complies with
the 2005 Title 24 Standards. The energy use of this building was simulated with
conventional and cool versions of three different steep-sloped (5:12) roofing products:
fibreglass asphalt shingle, concrete tile and polymer-coated metal. Each conventional
product had a solar reflectance of 0.10. The cool shingle had a solar reflectance of 0.25,
while the cool tile and cool metal products each had a solar reflectance of 0.40. All
products were assigned a thermal emittance of 0.90.
The 2005 Title 24 Standards prescriptively require a sub-roof radiant barrier for
residential buildings in some climate zones (2, 4 and 8 to 15); but they are not present on
most existing houses. Without a radiant barrier (typical of pre-2000 construction), total
savings initial cost savings from downsizing cooling equipment, plus the 30-year NPV of
TDV energy savings per unit roof area ranged from US$1.7/m
2
to US$18.6/m
2
across
Californias 16 climate zones (see Table 1.10). With a radiant barrier, the savings ranged
from US$1.3/m
2
to US$12.1/m
2
. Cool shingles incurred smaller saving (and penalties)
than did cool tiles and cool metal products because the solar reflectance of the cool
TABLE 1.9 Total savings per unit roof area (cooling equipment savings plus 30-year NPV
of energy savings with TDV; US$/m
2
) by California climate zone for a residential building with
a low-sloped roof
CALIFORNIA CLIMATE ZONE WITHOUT RADIANT BARRIER WITH RADIANT BARRIER
1 2.4 2.5
2 1.9 0.1
3 0.4 1.1
4 0.9 0.4
5 0.6 1.4
6 0.7 0.2
7 1.2 0.1
8 2.9 0.9
9 4.2 1.8
10 5.9 2.5
11 5.3 2.5
12 4.2 1.6
13 5.9 3.0
14 4.9 2.1
15 8.2 4.7
16 2.3 0.1
Note: The cool roof was assigned a solar reflectance of 0.55; the conventional roof, 0.10. We note that while
Californias Title 24 Standards prescribe the installation of sub-roof radiant barriers for residential buildings in
California climate zones 2, 4 and 8 to 15, it is not a common building practice for homes with low-sloped roofs.
The shaded values are appropriate to each climate zones radiant-barrier requirement
ABER_CH_01 7/11/08 10:19 AM Page 20
shingle exceeded that of the conventional shingle by 0.15, rather than by 0.30. The
negative savings occurred in coastal California climate zones with minimal summertime
cooling requirements. The presence of a sub-roof radiant barrier reduces cool-roof energy
savings, just as the presence of a cool roof reduces radiant-barrier energy savings.
The typical cost premium for a cool roof is US$0.0/m
2
to US$2.2/m
2
. Cool roofs with
premiums of up to US$2.2/m
2
are expected to be cost-effective in some climates zones.
At the time of writing, California is considering including in its 2008 Title 24 Standards a
prescriptive cool-roof requirement in hot Central Valley climates for residential buildings
with steep-sloped roofs.
COOL-ROOF PROVISIONS IN OTHER STANDARDS AND PROGRAMMES
Many US states have adopted building energy efficiency codes from ASHRAE Standard
90.1 or the International Energy Conservation Code (IECC). Aside from California, these
include the cities of Atlanta (Georgia) and Chicago (Illinois); the states of Florida, Georgia,
and Hawaii; and the territory of Guam. Cool-roof requirements have also been developed
TABLE 1.10 Total savings per unit roof area (cooling equipment savings plus 30-year NPV of energy
savings with TDV; US$/m
2
) by California climate zone for a residential building with a steep-sloped roof
CALIFORNIA CLIMATE ZONE WITHOUT RADIANT BARRIER WITH RADIANT BARRIER
SHINGLE TILE METAL SHINGLE TILE METAL
1 0.9 1.7 1.7 0.6 1.2 1.3
2 2.2 3.0 5.0 1.3 2.2 2.9
3 0.2 0.2 0.5 0.1 0.1 0.3
4 0.9 1.1 2.2 0.5 0.9 1.2
5 0.4 0.2 0.9 0.2 0.1 0.4
6 0.7 1.0 1.7 0.5 0.6 1.0
7 1.2 1.7 2.7 0.7 1.0 1.4
8 3.1 4.5 6.8 1.8 2.9 3.9
9 3.5 5.4 7.7 2.2 3.7 4.8
10 5.7 8.9 12.4 3.5 5.9 7.3
11 5.5 8.8 12.2 3.5 6.1 7.6
12 3.7 5.8 8.2 2.4 4.1 5.2
13 6.8 11.1 14.8 4.1 7.4 8.8
14 5.3 8.3 11.5 3.3 5.7 7.0
15 8.5 13.8 18.6 5.6 9.8 12.1
16 2.8 4.5 5.9 1.6 2.6 3.3
Note: Cool fibreglass asphalt shingles were assigned a solar reflectance of 0.25; cool concrete tiles and metal products, 0.40;
and all conventional products, 0.10. While Californias Title 24 Standards prescribe the installation of sub-roof radiant barriers
for residential buildings in climate zones 2, 4 and 8 to 15, it was not common practice in pre-2000 construction. The shaded
values are appropriate to each climate zones radiant-barrier requirement.
ABER_CH_01 7/11/08 10:19 AM Page 21
by several voluntary energy-efficiency programmes, including the US Environmental
Protection Agency (EPA) Energy Star label, the Leadership in Energy and Environmental
Design (LEED) Green Building Rating System, and the cool-roof rebate programmes
offered by the state of California and its utilities.
An earlier report by Eley Associates (2003a) summarizes the history of cool-roof
credits and requirements in these policies through 2003, with particular attention paid to
the political process of their development. Here we review the treatment of cool roofs in
several of these standards and programmes through 2007, focusing on technical details.
STANDARDS OTHER THAN ASHRAE 90.1, ASHRAE 90.2 AND TITLE 24
International Energy Conservation Code
The 2003 International Energy Conservation Code (IECC) does not explicitly address the
use of cool roofs. However, section 801.2 allows commercial buildings to comply with the
2003 IECC by satisfying the requirements of ASHRAE Standard 90.1, which, in turn, offers
cool-roof credits. The 2003 IECC provides neither direct nor indirect cool-roof credits for
residential buildings (ICC, 2003).
The 2006 IECC retains the link to ASHRAE Standard 90.1 for commercial buildings, and
explicitly offers cool-roof credits for residential buildings through performance
compliance. Table 404.5.2(1) assigns to the roof on the reference residential building a
solar absorptance of 0.75 (solar reflectance of 0.25) and a thermal emittance of 0.90, while
the roof on the proposed building is assigned its proposed values of solar absorptance
and thermal emittance (ICC, 2006).
The adoption as of May 2007 of IECC and/or ASHRAE standards by individual US
states is detailed in Figures 1.4 (commercial building codes) and 1.5 (residential building
codes).
Chicago, Illinois
To mitigate urban heat islands, the city of Chicago, Illinois, added a provision to section
18-13-303 of its 2001 Energy Conservation Code requiring that low-sloped roofs (those
with a ratio of rise to run not greater than 2:12) exhibit an initial solar reflectance not less
than 0.65, and a solar reflectance of at least 0.50 three years after installation. Medium-
sloped roofs (those with a ratio of rise to run greater than 2:12 and less than or equal to
5:12) were required to have initial and three-year solar reflectances of at least 0.15. Both
low- and medium-sloped roofs were required to have a minimum thermal emittance of
0.90. Roofs or portions of roofs that use photovoltaic, solar-thermal or roof garden
systems were exempt from these requirements (Chicago, 2001).
The cool-roof provisions of this code have been amended several times since 2001.
The current (2007) code requires that low-sloped roofs installed by 31 December 2008
have an initial solar reflectance not less 0.25, while those installed after that day must use
products that qualify for the US EPA Energy Star label (initial and aged solar reflectances
not less than 0.65 and 0.50, respectively). Medium-sloped roofs must have an initial solar
reflectance of at least 0.15.
Chicagos cool-roof standard has been weakened by the elimination of its thermal
emittance requirement and the establishment of a very low minimum solar reflectance
ABER_CH_01 7/11/08 10:19 AM Page 22
requirement for medium-sloped roofs. The effects of weakened standards on roof surface
temperature are quantified below in the discussion of the Energy Star programme.
Florida
The state of Florida first offered cool-roof credits for residential buildings in the 2001
edition of the Florida Building Code. The codes whole-building performance method for
compliance (Form 600A) multiplies the area of each envelope component by a summer
point multiplier and a winter point multiplier to estimate its contributions to the summer
cooling load and winter heating load. A cool-roof credit introduced in 2001 allows a
proposed home with a white roof (solar reflectance 0.65, thermal emittance 0.80) to
multiply its summer point multiplier by a credit factor of 0.55 and its ceiling winter point
multiplier by a credit factor of 1.044 (FBC, 2001, sections 607.1.A.5 and 607.2.A.3.6). This
reduces the estimated summer ceiling heat load of a white-roofed proposed home by
45 per cent and increases its estimated winter ceiling heat load by 4.4 per cent. The
current (2004) version of the code retains this white roof credit (FBC, 2004).
Source: Building Codes Assistance Project (BCAP, 2007)
FIGURE 1.4 Adoption of commercial building energy codes by US states as of May 2007
ABER_CH_01 7/11/08 10:19 AM Page 23
The 2007 code will use EnergyGauge USA FlaRes2007 (EnergyGauge, 2007), rather than
point multipliers, to estimate the annual energy use of residential buildings (FBC, 2007).
EnergyGauge USA FlaRes2007 is a building energy model based on DOE-2.1E that
incorporates an improved attic model (Parker, 2005). The 2007 code will require that the
annual energy budget of a proposed (as-built) home not exceed that of a reference
(baseline) home whose roof has a solar reflectance of 0.25 and a thermal emittance of
0.90. Radiative properties that make the proposed roof cooler than the reference roof (e.g.
a solar reflectance above 0.25) will permit the consumption of more energy elsewhere in
the building. Conversely, radiative properties that make the proposed roof warmer than
the reference roof (e.g. a solar reflectance below 0.25, or a thermal emittance less than
0.90) will require increased energy efficiency in other parts of the home. The solar
reflectance of 0.25 to be assigned to the roof of the reference home in the 2007 code is
greater than the reference roof solar reflectance of 0.15 used to generate the cool-roof
credit factors present in the 2004 code (Parker, 2007).
If the initial solar reflectance of the proposed homes roofing product has not been
measured by the manufacturer, it will be set to 0.04. If the initial thermal emittance is
Source: Building Codes Assistance Project (BCAP, 2007)
FIGURE 1.5 Adoption of residential building energy codes by US states as of May 2007
ABER_CH_01 7/11/08 10:19 AM Page 24
unmeasured, it will be assigned a value of 0.90 (FBC, 2007). We note that the latter
provision can significantly overestimate the true thermal emittance of a bare-metal roofing
product, which is typically less than 0.20. We address the influence of low thermal
emittance on roof surface temperature in our discussion of the Energy Star programme.
In 2004, the Florida Building Code adopted prescriptive and performance cool-roof
credits for commercial buildings that are essentially the same as those in ASHRAE
Standard 90.1-2004. The only difference is that the thermal transmittance multipliers used
for prescriptive compliance in section 13.404.1.C.1 are mapped to Florida climate zones,
rather than to US climate zones (FBC, 2004; Swami, 2007). The proposed 2007 Florida
Building Code retains the same cool-roof provisions.
Hawaii
Building energy codes in Hawaii are set by county ordinance, rather than by state law.
Over 80 per cent of the states population live in the counties of Honolulu and Maui
(Census, 2007). In 2001, the county of Honolulu amended its ordinances to prescriptively
require that roof assemblies on low-rise residential buildings include at least one of the
following:
insulation with a thermal resistance of 19ft
2
h F BTU
1
;
2 inches of continuous foam-board insulation;
a radiant barrier and attic ventilation; or
a cool roof with a solar reflectance not less than 0.70 and a thermal emittance not
less than 0.75 (ROH, 2001, section 32-14.2).
The county of Maui followed suit in 2005 (MCC, 2004). The current (2007) ordinances of
Honolulu and Maui retain this cool-roof requirement. The counties of Kauai and Hawaii
neither credit nor prescribe the use of cool roofs on residential buildings. In 2001, 2002
and 2005, respectively, the counties of Honolulu, Kauai and Maui adopted cool-roof
credits for commercial and high-rise residential buildings based on ASHRAE Standard
90.1-1999 (Wiig, 2007). The building envelope prescriptions for these buildings (e.g.
section 32-8 of ROH, 2004) use a modified cool-roof credit adopted from the Hawaii
Model Energy Code (Eley Associates, 2003b). By requiring that the product of the roof
thermal transmittance (BTU ft
2
h
1
F
1
), roof solar absorptance and a radiant barrier credit
(0.33 if present, 1 if absent) be less than 0.05BTU ft
2
h
1
F
1
, this provision allows for the
use of reduced insulation under a roof with high solar reflectance. The envelope
prescription does not set a minimum requirement for the thermal emittance of the roof.
Hence, it may credit the use of a solar-reflective but hot bare-metal roof, as addressed in
the discussion of the Energy Star programme.
The county of Hawaii follows ASHRAE Standard 90.1-1989 for commercial buildings,
which neither prescribes nor credits cool roofs.
Guam
Guams code for non-residential and high-rise residential buildings (adopted in 1995) and
its code for low-rise residential buildings (adopted in 2000) establish identical prescriptive
ABER_CH_01 7/11/08 10:19 AM Page 25
requirements for roofs on air-conditioned buildings (Eley Associates, 2007). For these
buildings, mass roof assemblies that is, roofs made of concrete 4 inches or greater in
thickness; having heat capacity per unit area greater than 7.0BTU ft
2
F
1
; and/or weighing
more than 35lb ft
2
must have:
a cool (high albedo) surface of solar reflectance not less than 0.70 and thermal
emittance not less than 0.75;
R-11 insulation (11ft
2
h F BTU
1
) in the interior furring space;
2 inches of continuous insulation; or
thermal transmittance not exceeding 0.12BTU h
1
ft
2
F
1
.
Air-conditioned buildings with other types of roofs are required to have more insulation
and/or a lower thermal transmittance than mass roofs, but cannot apply a cool-roof
surface towards prescriptive compliance.
Air-conditioned buildings that meet the codes mandatory requirements but not its
prescriptive requirements can achieve performance compliance via either ASHRAE
Standard 90.1-1989 or via a building envelope trade-off option. The latter requires that the
energy performance factor (EPF) of a proposed building not exceed that of a reference
(budget) building. The EPF of a building includes the EPF of its roof, which for mass roofs
and roofs on metal buildings (buildings with metal sheathing and metal framing) is defined
as the product of the roofs area, thermal transmittance and solar absorptance. In EPF
calculations, the solar absorptance of a proposed or reference roof is set to 0.30 if the roof
is cool (solar reflectance 0.70, thermal emittance 0.75), or 0.70 otherwise. The roof of
the reference building may or may not be cool, depending upon whether the roof
assembly chosen for the reference building uses a cool surface to comply with the code
(Eley, 2007).
VOLUNTARY ENERGY EFFICIENCY PROGRAMMES
US EPA Energy Star label
To qualify for its Energy Star label, the US EPA currently requires that low-sloped roofing
products (those installed on roofs with a ratio of rise to run not exceeding 2:12) have initial
and three-year aged solar reflectances not less than 0.65 and 0.50, respectively. Steep-
sloped roofing products (those installed on roofs with a ratio of rise to run greater than
2:12) must have initial and three-year aged solar reflectances not less than 0.25 and 0.15,
respectively (EPA, 2007).
We note that the Energy Star cool-roof requirements have two weaknesses. First, by
specifying neither a minimum thermal emittance nor a minimum solar reflectance index,
they permit the use of bare-metal roofs with high solar reflectance but low thermal
emittance. Under ASTM E1980 medium wind-speed conditions, the surface of an aged
bare-metal roof with a solar reflectance of 0.50 and a thermal emittance of 0.15 would be
about 12K (22F) hotter than that of an aged white roof with a solar reflectance of 0.50
and thermal emittance of 0.80. Second, the minimum three-year aged solar reflectance
required for a steep-sloped roof (0.15) excludes only the hottest of roofing materials, such
as granule-surfaced fibreglass asphalt shingles coloured with conventional dark pigments.
ABER_CH_01 7/11/08 10:19 AM Page 26
Many cool roofing products for steep-sloped roofs attain an aged solar reflectance of at
least 0.30. Under these conditions, the surface of a roof with a solar reflectance of 0.15
and a thermal emittance of 0.80 will be 10K (18F) hotter than that of a roof with a solar
reflectance of 0.30 and a thermal emittance of 0.80.
LEED Green Building Rating System
The Leadership in Energy and Environmental Design (LEED) Green Building Rating System
assigns one rating point for the use of a cool roof in its Sustainable Sites Credit 7.2 (Heat
Island Effect, Roof). LEED Version 2.0 (2001) requires a cool roof to either cover at least
75 per cent of its surface with materials that have initial and three-year aged solar
reflectances of at least 0.65 and 0.50, respectively, and a thermal emittance of at least
0.90; or cover no less than 50 per cent of its surface with vegetation (GBC, 2001).
LEED Version 2.1 (2002) requires a cool roof to either:
cover at least 75 per cent of its surface with Energy-Star compliant products that also
have a thermal emittance of at least 0.90;
cover no less than 50 per cent of its surface covered by vegetation; or
cover at least 75 per cent of its surface with a combination of these two materials
(GBC, 2002).
Compared to version 2.0, version 2.1 reduces the minimum initial solar reflectance
required for steep-sloped roofs (ratio of rise to run greater than 2:12) to 0.25 from 0.65,
and the minimum aged solar reflectance to 0.15 from 0.50.
We note that the minimum thermal emittance requirement of 0.90 in versions 2.0 and
2.1 is unnecessarily high, as most high-emittance materials have thermal emittances in
the range of 0.80 to 0.95. The LEED requirement of 0.90 tends to exclude many cool
materials, such as white roofs, whose thermal emittances may lie slightly below 0.90. This
issue is compounded by the high uncertainty (up to 0.05) in measuring the thermal
emittance of thermally massive materials. The less stringent minimum thermal-emittance
requirement of 0.75 used in the ASHRAE and Title 24 Standards definitions of a cool roof
is designed to include all high-emittance materials, most of which are expected (though
not required) to exhibit thermal emittances of at least 0.80.
LEED Version 2.2 (2005) uses SRI, rather than solar reflectance, thermal emittance or
Energy-Star compliance, to qualify a non-vegetated cool roof (GBC, 2005). LEED Version 2.2
requires a cool roof to either:
cover at least 75 per cent of its surface with products that have a minimum SRI of 78
(low-sloped roofs) or 29 (steep-sloped roofs);
have at least 50 per cent of its surface covered by vegetation; or
use a combination of vegetation and high-SRI materials that satisfy a particular
formula (GBC, 2005).
We note that the SRI requirements in the current version of LEED (version 2.2) are about
those achieved by a low-sloped roof with a solar reflectance of 0.65 and a thermal
ABER_CH_01 7/11/08 10:19 AM Page 27
emittance of 0.90, and by a steep-sloped roof with a solar reflectance of 0.28 and a
thermal emittance of 0.90 (since the SRI of this cool low-sloped surface is actually 78.9,
we recommend that its required SRI be increased to 79 from 78). We welcome both the
simplicity of the SRI requirement and the ability to use truly cool materials whose thermal
emittances are less than 0.90.
California cool-roof rebate programmes
From 2001 to 2005, the state of California and several of its utilities offered rebates of
US$0.10/ft
2
to US$0.20/ft
2
for the installation of cool roofs with initial solar reflectance not
less than 0.70 and initial thermal emittance not less than 0.75. Since the current (2005)
Title 24 Standards now prescriptively require cool roofs on non-residential buildings with
low-sloped roofs, recent rebate programmes have focused on residential buildings.
In January 2006, the Sacramento Municipal Utility District (SMUD) began offering a
rebate of US$0.20/ft
2
for the installation of a residential flat roof with a solar reflectance
greater than 0.75 and a thermal emittance greater than 0.75. In May 2007, the SMUD
programme was expanded to offer a rebate of US$0.10/ft
2
for the installation of a steep-
sloped residential roof with a solar reflectance greater than 0.40 and a thermal emittance
greater than 0.75 (SMUD, 2007).
In January 2007, two California utilities Pacific Gas & Electric (PG&E) and Southern
California Edison (SCE) began a new programme offering rebates of US$0.10/ft
2
to
US$0.20/ft
2
for retrofitting existing homes in certain California climates with cool roofs
(PG&E, 2007; SCE, 2007). The solar reflectance and thermal emittance requirements of
this programme are detailed in Table 1.11. Note that all qualifying products must have a
thermal emittance of at least 0.75.
The low-sloped roof requirements of the PG&E/SCE programme are designed to
promote the use of white roofs. The two levels of rebates for steep-sloped roofs (Tier 1:
US$0.10/ft
2
for solar reflectance between 0.25 and 0.39; Tier 2: US$0.20/ft
2
for solar
reflectance not less than 0.40) are designed to encourage the use of existing cool-
coloured products (most of which lie in Tier 1) and to spur the development and sale of
improved cool-coloured products (Tier 2).
TABLE 1.11 Solar reflectance and thermal emittance requirements of a 2007 residential cool-roof rebate
programme administered by two California utilities (Pacific Gas & Electric and Southern California Edison)
ROOF REBATE INITIAL INITIAL REBATE
SLOPE TIER SOLAR THERMAL (US$/FT
2
)
REFLECTANCE EMITTANCE
Low
a
n/a 0.70 0.75 0.20
Steep
b
Tier 1 0.250.39 0.75 0.10
Steep
b
Tier 2 0.40 0.75 0.20
Notes: a = ratio of rise to run less than or equal to 2:12
b = ratio of rise to run greater than 2:12
Source: Pacific Gas & Electric (PG&E, 2007); Southern California Edison (SCE, 2007)
ABER_CH_01 7/11/08 10:19 AM Page 28
CONCLUSIONS
Since the late 1990s, the quantification of energy savings offered by the use of cool roofs
has led both ASHRAE and the state of California to add cool-roof credits and/or
requirements to their energy efficiency standards for both residential and non-residential
buildings. Many US states have adopted cool-roof credits from ASHRAE Standard 90.1
(1999 or later), IECC 2003 or IECC 2006. Several US cities and states other than California
have developed custom cool-roof provisions to their energy standards. Voluntary energy
efficiency programmes, such as the US EPA Energy Star label, the LEED Green Building
Rating System of the US Green Building Council and rebate programmes offered by
California and its utilities, have established qualifications for cool roofs.
While cool-roof requirements have occasionally been too strict for example,
excluding many cool materials by setting a minimum thermal emittance of 0.90, rather
than one of 0.75 they are more often too lax. Particularly problematic are those
definitions that (a) allow the use of hot bare-metal products on low-sloped roofs by
specifying neither a minimum thermal emittance nor a minimum SRI; and/or (b) allow the
use of all but the hottest materials on steep-sloped roofs by qualifying products with an
aged solar reflectance as low as 0.15. We have also found ambiguities and outright errors
in several cool-roof standards. These issues suggest that more care needs to be taken to
ensure that cool-roof standards are both accurate and effective.
The standards described in this chapter were developed primarily by workers at
several US research laboratories. We expect that cool-roof standards will be further
refined to incorporate improvements in building energy analysis and cool-roof technology.
However, the methods used to develop the cool-roof provisions in the ASHRAE and
California Title 24 standards can be used as models to address cool roofs in building
energy efficiency standards worldwide.
AUTHOR CONTACT DETAILS
Hashem Akbari: (corresponding author) Heat Island Group, Lawrence Berkeley National Laboratory, Berkeley,
CA, 94720 US; H_Akbari@LBL.gov
Ronnen Levinson: Heat Island Group, Lawrence Berkeley National Laboratory, Berkeley, CA, 94720 US
ACKNOWLEDGMENTS
This chapter was supported by the California Energy Commission (CEC) through its Public
Interest Energy Research Program (PIER) and by the Assistant Secretary for Renewable
Energy under Contract No DE-AC02-05CH11231.
We wish to thank Javier Ceballos, City of Chicago; Danny Parker, Jeff Sonne and
Muthusasmy Swami, Florida Solar Energy Center; Howard Wiig, State of Hawaii; and
Charles Eley, Architectural Energy Corporation for clarifying building codes.
NOTE
1 The location of distribution ducts can strongly influence the energy performance of cooling
systems. Leaky and/or poorly insulated ducts in attics have exhibited delivery efficacies as low as
ABER_CH_01 7/11/08 10:19 AM Page 29
50 per cent (Jump and Modera, 1994). Delivery efficacy falls as the attic temperature rises. Parker
et al (1998a) have developed a model to account for the influence of attic temperature upon the
delivery efficacy of the distribution system.
REFERENCES
Akbari, H. (1998) Cool roofs save energy, ASHRAE Transactions, vol 104, no 1B, pp783788
Akbari, H. and A. Desjarlais (2005) Cooling down the house, Professional Roofing, March,
www.professionalroofing.net/article.aspx?A_ID=609
Akbari, H., S. Bretz, H. Taha, D. Kurn and J. Hanford (1997) Peak power and cooling energy savings of high-albedo roofs,
Energy and Buildings, vol 25, no 2, pp117126
Akbari, H., S. Konopacki, D. Parker, B. Wilcox, C. Eley and M. Van Geem (1998) Calculations in support of SSP90.1 for
reflective roofs, ASHRAE Transactions, vol 104, no 1B, pp984995
Akbari, H., S. Konopacki and M. Pomerantz (1999) Cooling energy savings potential of reflective roofs for residential and
commercial buildings in the United States, Energy, vol 24, pp391407
Akbari, H., S. Konopacki and D. Parker (2000) Updates on revision to ASHRAE Standard 90.2: Including roof reflectivity for
residential buildings, in ACEEE 2000 Summer Study on Energy Efficiency in Buildings, vol 1, Pacific Grove, CA, August,
American Council for an Energy Efficient Economy, Washington, DC, pp111
Akbari, H., M. Pomerantz and H. Taha (2001) Cool surfaces and shade trees to reduce energy use and improve air quality in
urban areas, Solar Energy, vol 70, no 3, pp295310
Akbari, H., R. Levinson and P. Berdahl (2005a) Review of residential roofing materials, Part I: A review of methods for
the manufacture of residential roofing materials, Western Roofing Insulation and Siding, January/February,
pp5457
Akbari, H., R. Levinson and P. Berdahl (2005b) Review of residential roofing materials, Part II: A review of methods for the
manufacture of residential roofing materials, Western Roofing Insulation and Siding, March/April, pp5258
Akbari, H., C. Wray, T. T. Xu and R. Levinson (2006) Inclusion of solar reflectance and thermal emittance prescriptive
requirements for steep-sloped nonresidential roofs in Title 24, http://energy.ca.gov/title24/2008standards/prerulemaking
documents/2006-05-18_workshop/2006-05-19_NONRESDNTL_STEEP-SLOPED_COOL_ROOFS.PDF
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) (1999) ASHRAE Standard 90.1-1999:
Energy Standard for Buildings Except Low-Rise Residential Buildings, SI Edition, American Society of Heating,
Refrigerating and Air-Conditioning Engineers, Atlanta, GA
ASHRAE (2001) ASHRAE Standard 90.1-2001: Energy Standard for Buildings Except Low-Rise Residential Buildings, SI Edition,
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA
ASHRAE (2004a) ASHRAE Standard 90.1-2004: Energy Standard for Buildings Except Low-Rise Residential Buildings, SI
Edition, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA
ASHRAE (2004b) ASHRAE Standard 90.2-2004: Energy-Efficient Design of Low-Rise Residential Buildings, American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA
ASHRAE (2007) ASHRAE Standard 90.2-2007: Energy-Efficient Design of Low-Rise Residential Buildings, American Society of
Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA
ASTM (American Society for Testing and Materials) (1998) ASTM E 1980-98: Standard Practice for Calculating Solar
Reflectance Index of Horizontal and Low-Sloped Opaque Surfaces, in Annual Book of ASTM Standards, vol 04.12,
American Society for Testing and Materials, Philadelphia, PA
BCAP (2007) Status of residential and commercial building state energy codes, www.bcap-energy.org
Berdahl, P., H. Akbari, R. Levinson and W. A. Miller (2008) Weathering of roofing materials an overview, Construction and
Building Materials, vol 22, no 4, April, pp423433
ABER_CH_01 7/11/08 10:19 AM Page 30
CEC (California Energy Commission) (2001) 2001 Energy Efficiency Standards for Residential and Nonresidential Buildings,
P400-01-024, California Energy Commission, Sacramento, CA
CEC (2006) 2005 Building Energy Efficiency Standards for Residential and Nonresidential Buildings, CEC-400-2006-015,
California Energy Commission, Sacramento, CA
Census (2007) US Census Bureau State and County Quickfacts, http://quickfacts.census.gov
Chicago (2001) Amendment of Title 18 of Municipal Code of Chicago Concerning Energy Efficiency Requirements, Journal of
the City Council of Chicago, 6 June, p60939
Chicago (2007) City of Chicago Energy Conservation Code, Index Publishing Corporation, Chicago, IL
DOE-2 (2007) Lawrence Berkeley National Lab DOE-2 website, http://simulationresearch.lbl.gov/dirsoft/d2whatis.html
Eley (2007) Pers comm with Charles Eley, Architectural Energy Corporation, 17 August
Eley Associates (2003a) Assessment of Public Policies Affecting Cool Metal Roofs, Final report prepared for the Cool Metal
Roofing Coalition, www.coolmetalroofing.org/elements/uploads/casestudies/TMI_CaseStudy_28.pdf
Eley Associates (2003b) Hawaii Commercial Building Guidelines for Energy Efficiency,
www.archenergy.com/library/general/hawaiigl
Eley Associates (2007) Guam Building Energy Code, http://eley.com/guam
Eilert, P. (2000) High Albedo (Cool) Roofs: Codes and Standards Enhancement (CASE) Study, Pacific Gas & Electric report,
www.energy.ca.gov/title24/2001standards/associated_documents/2000-11-17_PGE_CASE.PDF
EnergyGauge (2007) EnergyGauge USA FlaRes2007 Energy and Economic Analysis Software, www.energygauge.com
EPA (US Environmental Protection Agency) (2007) Roof Products Criteria for US EPA Energy Star Program,
www.energystar.gov/index.cfm?c=roof_prods.pr_crit_roof_products
FBC (Florida Building Commission) (2001) 2001 Florida Building Code, Florida Building Commission, Tallahassee, FL,
www.floridabuilding.org
FBC (2004) 2004 Florida Building Code, Florida Building Commission, Tallahassee, FL, www.floridabuilding.org/
FBC (2007) Proposed Modification to the Florida Building Code: Chapter 11, Energy Efficiency,
www.dca.state.fl.us/FBC/thecode/Res_Chapter_11.rtf
GBC (US Green Building Council) (2001) Leadership in Energy and Environmental Design Green Building Rating
System for New Construction and Major Renovations (LEED-NC), Version 2.0, US Green Building Council,
www.usgbc.org
GBC (2002) Leadership in Energy and Environmental Design Green Building Rating System for New Construction and Major
Renovations (LEED-NC), Version 2.1, US Green Building Council, www.usgbc.org
GBC (2005) Leadership in Energy and Environmental Design Green Building Rating System for New Construction and Major
Renovations (LEED-NC), Version 2.2, US Green Building Council, www.usgbc.org
Hildebrandt, E., W. Bos and R. Moore (1998) Assessing the impacts of white roofs on building energy loads, ASHRAE
Technical Data Bulletin, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, vol 14,
no 2, pp2836
ICC (International Code Council) (2003) 2003 International Energy Conservation Code, www.iccsafe.org
ICC (2006) 2006 International Energy Conservation Code, www.iccsafe.org
Jump, D. and M. Modera (1994) Energy Impacts of Attic Duct Retrofits in Sacramento Houses, LBL-35375, Lawrence Berkeley
National Laboratory, Berkeley, CA
Konopacki, S. and H. Akbari (1998) Simulated Impact of Roof Surface Solar Absorptance, Attic, and Duct Insulation on Cooling
and Heating Energy Use in Single-Family New Residential Buildings, LBNL-41834, Lawrence Berkeley National Laboratory,
Berkeley, CA
Konopacki, S. and H. Akbari (2001) Measured Energy Savings and Demand Reduction from a Reflective Roof Membrane on a
Large Retail Store in Austin, LBNL-47149, Lawrence Berkeley National Laboratory, Berkeley, CA
ABER_CH_01 7/11/08 10:19 AM Page 31
Konopacki, S., L. Gartland, H. Akbari and L. Rainer (1998) Demonstration of Energy Savings of Cool Roofs, LBNL-40673,
Lawrence Berkeley National Laboratory, Berkeley, CA
Levinson, R., H. Akbari, S. Konopacki and S. Bretz (2005a) Inclusion of cool roofs in nonresidential Title 24 prescriptive
requirements, Energy Policy, vol 33, no 2, pp151170
Levinson, R., P. Berdahl and H. Akbari (2005b) Solar spectral optical properties of pigments, part I: Model for deriving
scattering and absorption coefficients from transmittance and reflectance measurements, Solar Energy Materials & Solar
Cells, vol 89, pp319349
Levinson, R., P. Berdahl and H. Akbari (2005c) Solar spectral optical properties of pigments, part II: Survey of common
colorants, Solar Energy Materials & Solar Cells, vol 89, pp351389
Levinson, R., P. Berdahl, A. A. Berhe and H. Akbari (2005d) Effects of soiling and cleaning on the reflectance and solar heat
gain of a light-colored roofing membrane, Atmospheric Environment, vol 39, pp78077824
Levinson, R., P. Berdahl, H. Abkari, W. Miller, I. Joedicke, J. Reilly, Y. Suzuki and M. Vondran (2007) Methods of creating solar-
reflective nonwhite surfaces and their application to residential roofing materials, Solar Energy Materials & Solar Cells,
vol 91, pp304314
MCC (Maui County Code) (2004) A Bill for an Ordinance Amending Title 16, Maui County Code, Pertaining to Energy Efficiency
Standards for Buildings, www.co.maui.hi.us/files/ordinance/LF-Ord3240_etkoujogl.pdf
MICROPAS (2007) MICROPAS product website, http://micropas.com
Parker, D. (2005) Technical Support for Development of an Attic Simulation Model for the California Energy Commission, Florida
Solar Energy Center report FSEC-CR-1526-05, www.fsec.ucf.edu/en/publications/pdf/FSEC-CR-1526-05.pdf
Parker, D. (2007) Pers comm from Danny Parker, Florida Solar Energy Center, 13 August
Parker, D., J. Huang, S. Konopacki, L. Gartland, J. Sherwin, and L. Gu (1998a) Measured and simulated performance of
reflective roofing systems in residential buildings, ASHRAE Transactions, vol 104, no 1, Atlanta, GA
Parker, D., J. Sherwin and J. Sonne (1998b) Measured performance of a reflective roofing system in a Florida commercial
building, ASHRAE Technical Data Bulletin, American Society of Heating, Refrigerating and Air-Conditioning Engineers,
Atlanta, GA, vol 14, no 2, pp712
PG&E (Pacific Gas & Electric) (2007) Pacific Gas & Electric Cool-Roof Rebate Program, www.pge.com/myhome/saveenergymoney/
rebates/remodeling/coolroof
ROH (Revised Ordinances of Honolulu) (2001) Revised Ordinances of Honolulu, City and County of Honolulu
ROH (2004) Revised Ordinances of Honolulu, City and County of Honolulu
ROH (2007) Revised Ordinances of Honolulu, www.honolulu.gov/refs/roh
SCE (Southern California Edison) (2007) Southern California Edison Cool-Roof Rebate Program, www.sce.com/Rebatesand
Savings/Residential/_Heating+and+Cooling/CoolRoof
SMUD (Sacramento Municipal Utility District) (2007) Sacramento Municipal Utility District Residential Cool-Roof Program,
www.smud.org/rebates/cool%20roofs
Swami, M. (2007) Pers comm from Muthusamy Swami, Florida Solar Energy Commission, Developer of FLA/COM
performance compliance software, 14 June
Taha, H. (2001) Potential Impacts of Climate Change on Tropospheric Ozone in California: A Preliminary Episodic Modeling
Assessment of the Los Angeles Basin and the Sacramento Valley, LBNL-46695, Lawrence Berkeley National Laboratory,
Berkeley, CA
Wiig, H. (2007) Pers comm from Howard Wiig, Institutional Energy Analyst, Department of Business, Economic Development
and Tourism, Hawaii, 14 June
Wray, C., H. Akbari, T. T. Xu and R. Levinson (2006) Inclusion of Solar Reflectance and Thermal Emittance Prescriptive
Requirements for Residential Roofs in Title 24, www.energy.ca.gov/title24/2008standards/prerulemaking/
documents/2006-05-18_workshop/2006-05-17_RESIDENTIAL_ROOFS.PDF
ABER_CH_01 7/11/08 10:19 AM Page 32
Keywords daylighting systems; shading; simulation
INTRODUCTION
After the 1970s energy crisis, interest in daylighting among other solar applications
began to be re-examined in an effort to reduce energy consumption. Today, with the
advancing problem of global warming, daylighting is considered a very important strategy
to replace the electric energy demanded by artificial lighting, reducing not only lighting
(and cooling) consumption, but efficiently decreasing peak electrical loads.
The literature presents a variety of results in relation to energy savings through
daylighting, which are difficult to compare since the data refer to a particular place,
building and system. Based on simulation or field measurements results, energy savings
are somewhere between 20 and 50 per cent of a buildings electricity consumption
(Bodart and De Herde, 2002). It is evident that maximization of daylight use can be
increased through the development of innovative systems. Although many systems have
been tested during major research programmes, such as IEA TASK 21 (2002), market
Abstract
One of the challenges of architectural design is to bring daylight into the core of a building
in an effort to improve visual environment and, at the same time, to increase energy
savings. This can be performed through the use of daylighting systems, the operational
principles of which are presented in this chapter. Using sunlight/skylight redirection or
exclusion these systems can affect a buildings energy behaviour since their adoption is
subject to a compromise between shading needs and daylight provision. The choice of the
proper daylighting system should be based on a solid rating system, which is quite difficult
to develop since a large number of parameters are involved. These include the site, climate,
openings, operational and integration constraints, and the objectives that the design team
has initially set. Although a large number of daylighting systems are available to the building
industry today, their use is limited for reasons such as high initial cost, maintenance and
variability in their performance parameters.
2
A Review of Innovative
Daylighting Systems
A. Tsangrassoulis
ABER_CH_02 6/12/08 3:31 PM Page 33
penetration is very limited for the following reasons:
high cost due to the precision required in the manufacturing process;
difficulty in quantifying energy savings due to the lack of relevant data (such as
bidirectional transmittance distribution function (BDTF) or solar heat gain coefficients);
risk of glare or over-illuminance due to their control.
Given that these drawbacks are corrected, it is quite possible that the systems efficiency
will increase considerably. Furthermore, during the summer months, reducing solar gains
and attaining adequate daylighting levels can be difficult, leading to an optimized design
procedure in order to satisfy both needs. Traditionally, this role is left to shading systems
(e.g. blinds, fabric rollers, etc.), which minimize daylight supply. Daylighting can contribute
to an improvement in lighting quality (e.g. sufficient lighting levels and colour rendition)
and to better occupant health and comfort since it is a mood motivator and affects
productivity due to reduced stress and fatigue.
During the past few decades, new issues in daylighting design have been raised in
conjunction with new demands by the building industry. One of the challenges is to bring
usable daylight deep into the core of a building or into non-daylit areas, which may result
in significant savings in energy consumption while creating an attractive visual
environment. Conventional fenestration systems can provide adequate daylighting in
areas near building openings (perimeter area). The increase of this area can be realized
either by using a design strategy in relation to a different building form (i.e. with an atrium)
or a system that offers specialized spectral and/or angular selectivity. The latter can be
called a daylighting system in the sense that it can extend performance beyond that of
conventional solutions. An overview of the main principles involved, providing sketches
and short descriptions of the systems elements, has been presented in Kischkoweit-
Lopin (2002). This chapter, instead, deals with the physical processes that define the
operation of these systems.
TAXONOMY
Depending upon the position of the daylighting systems within the building envelope, two
main categories exist (McCluney, 1998):
1 fenestration systems that are built in between glazing or can be attached to the
exterior/interior of the window frame;
2 core systems that are used to introduce daylight from a distance from the building
envelope.
Another classification can be performed that examines the use of diffuse skylight or direct
sunlight. According to Matusiak (2004):
Daylighting systems with shading are systems that:
rely primarily on diffuse skylight and reject sunlight;
generally use direct sunlight, reflecting it onto ceilings or to locations above eye
height.
34 A. TSANGRASSOULIS
ABER_CH_02 6/12/08 3:31 PM Page 34
Daylighting systems without shading are:
diffuse light-guiding systems;
direct light-guiding systems;
light-scattering or diffusing systems;
light transport systems.
Of course, in an effort to achieve seasonal adjustments, daylighting systems can be fixed
or moveable (e.g. sun-tracking devices).
The introduction of daylighting systems to a building is based on the fulfilment of one
or more of the following goals:
increase in lighting levels in the areas away from the building perimeter;
increase in uniformity;
improvement of visual comfort;
provision of sufficient shading.
FENESTRATION SYSTEMS
Fenestration systems can be separated according to their reflective or refractive operating
principles. Many daylighting systems restrict the view outside; thus, they have to be
placed on clerestory windows away from the horizontal line of sight. The most common
shading systems that can be considered as daylighting systems are Venetian blinds.
These can be used either internally or externally, causing redistribution of daylight
(sunlight or skylight) to ceilings. The efficiency of Venetian blinds in terms of light
redistribution is rather low and depends upon the blind section and its material (and the
associated reflectance). Only the optimum orientation of these blinds can allow maximum
reflection without preventing outward vision. Visual connection to the exterior
environment is vital for human beings, providing access to quality daylight.
Reflecting blinds are used as an internal multifunctional system that is able to fulfil all
complex demands of fenestration systems. The blinds are equipped with a high specular,
high reflective upside surface with reflectance above 90 per cent, and with a grey diffuse
downside surface. A further improvement is the perforation of the blinds surface, a
technique able to provide a visual connection with the exterior environment even when
the blind system is closed.
The optimum use of a blind system can be performed by automatic control depending
upon the external climatic conditions and the internal demands. DiBartolomeo et al (1996)
have tested such a system with the following control options:
direct sun block;
adjustment of blind angle to maintain the design workplaces illumination; and
maintenance of the total illuminance from lights and daylight within the designs
illuminance range.
Ordinary blinds have a curved section in order to achieve mechanical stiffness and
adjustment precision.
A Review of Innovative Daylighting Systems 35
ABER_CH_02 6/12/08 3:31 PM Page 35
FIGURE 2.1 Simulation of luminance distribution in a room with no shading system (upper) shaded by
fabric roller (middle) and shaded by Venetian blinds (lower)
Source: Department of Architecture, University of Thessaly
2
2
7
5
Acceptance angle for diffuse radiation

Solar rays cut-off angle
FIGURE 2.2 Schematic representation of two geometrical characteristics: cut-off angle for solar rays and
angle of acceptance for diffuse radiation
ABER_CH_02 6/12/08 3:31 PM Page 36
Since it is an expensive solution to continuously and automatically adjust blind positions,
a variety of new designs have been produced. The main idea behind these designs is to
maintain the blinds, if possible, in a horizontal position, offering an unobstructed view, and
to regulate solar gains during heating/cooling periods. Such a system, developed by
Koester (2004), is presented in Figure 2.3.
Retrolux is a toothed blind that consists of two parts, with the outer used for the suns
retro-reflection, while the inner part reflects the suns radiation by using a very limited
number of reflections. As a result, the reflection of the retro-reflection in the window
panes does not generate disturbing glare effects in the interior space. It is evident that in
maintaining a constant blind position, performance depends upon the position of the sun.
Thus, in order to maintain performance, designers should use quite complex optical
systems where direct sunlight can be partly reflected depending upon the incident angle.
During the summer, the transmittance is normally low, while for low solar altitudes during
winter, the transmittance is high. Due to their accurate shape and highly reflective
material, these systems are enclosed between a double-glazing insulating unit. These
systems reflect direct sunlight backwards and mainly transmit diffuse radiation. An
important advantage is that these systems do not affect the colour of daylight and do not
require maintenance. Numerous optical designs with various optical properties exist. A
typical example of this system is seen in Figure 2.4.
Through special surface treatment of three-dimensionally designed metal or plastic
optical structures, compound parabolic concentrators (CPCs) can be constructed for the
directional selection of radiation. These structures have a strong directional selectivity in
their transmittance/reflectance (Kuckelkorn, 2002). CPCs can be created by using two
parabolas, as presented in Figure 2.5; its two-dimensional geometry was proposed by
Winston (1974).
Any ray entering the CPCs entry aperture with an angle of incidence smaller than the
angle () will emerge from the exit aperture. In all other cases, rays are reflected back
H
i
g
h

s
u
n
R
E
T
R
O
F
l
a
t

s
u
n
R
E
T
R
O
IN
T
R
O
Source: Koester LichtPlannung (www.koester-lichtplanung.com/pages_gb/product_01.html)
FIGURE 2.3 Retrolux blind
ABER_CH_02 6/12/08 3:31 PM Page 37
(depending, of course, upon the surface roughness and local curvatures). For symmetrical
CPCs, the concentration ratio (C) is given by:
C = 1/sin(). [1]
Increase of C means that has to be decreased, which, in practice, signifies that the
collector accepts light from a very small part of the sky. Taking into account the azimuth
Source: OKALUX (www.okalux.com)
FIGURE 2.4 Okasolar-W system

FIGURE 2.5 Creation and operational principle of a compound parabolic concentrator (CPC) using two
parabolas and definition of maximum acceptance angle ()
ABER_CH_02 6/12/08 3:31 PM Page 38
and slope of the window where the system is to be installed, asymmetric profiles can be
designed in order to accept light from a predefined part of the sky (e.g. excluding sunlight
during summer). A similar system used for skylights is the Microlouver by Siteco. This
system consists of specially shaped and aluminium vapour-deposited louvres that are
designed to retro-reflect radiation coming from the southern hemisphere while
transmitting light from the northern hemisphere.
The solar heat gain coefficient is 14 per cent, with direct transmittance varying from 5 to
55 per cent, depending upon the suns position, and diffuse transmittance of 38 per cent.
Refraction is also used as a design principle for daylighting systems. The most
representative system in this category is the prismatic glazing, which is manufactured
from ultraviolet (UV) light- and weather-resistant polymethyl methacrylate (PMMA). The
geometry of the prisms is designed to reflect direct sunlight by means of internal total
reflection, and to transmit diffuse light. Due to the narrow angle of this operation, the
prismatic glazing has to be tilted according to the height of the sun. Typical prism
geometry is presented in Figure 2.7.
Following the analysis presented by Lorenz (2001), the two prism angles are critical for
performance.
Angle is determined in such way that the following equation applies:
[2]
where is the prism angle, n is the refractive index and n
G
is the angle that defines if
transmission or rejection of solar radiation will occur. If the suns elevation (on the cross-
sectional plane of the prism) is larger that n
G
, then the suns rays are to be rejected;
otherwise, they are transmitted as shown in Figure 2.8.
Prismatic systems with various angles, such as and , have been designed, and in
some of them one facet of the prism is coated with a reflective layer. On certain occasions,
sin( ) sin(arcsin ) n n
n
G
=

*
1
Source: Siteco (www.siteco.de/de/produkte/tageslichtsysteme.html)
FIGURE 2.6 Construction scheme of the Microlouver (by Siteco)
ABER_CH_02 6/12/08 3:31 PM Page 39
prismatic glazing is enclosed in a double-glazing unit for protection. Prismatic glazing can
cause visual distortion; thus, in many cases they are placed in clerestory windows.
Laser-cut light-deflecting panels (LCPs) (Edmonds, 1993) consist of PMMA panes in
which narrow parallel grooves have been cut out using a laser beam. A typical cross-
section is presented in Figure 2.9.
FIGURE 2.7 Prismatic glazing shading operation
FIGURE 2.8 Geometric configuration of a typical prismatic glazing section
n
G
ABER_CH_02 6/12/08 3:31 PM Page 40
Depending upon the geometry of LCPs and the angle of incidence, light can be deflected
upwards (Edmonds et al, 1995). Thus, in the case of a vertical window equipped with this
system, light flux is redirected to the ceiling and improves light levels in deep rooms,
while at the same time permitting a clear view to the exterior environment.
Holographic optical elements (HOEs) use diffraction effects to redirect sunlight and
are manufactured by exposing a photographic film to a coherent and monochromatic
laser beam. There are two parameters that affect their operation:
1 the period of the photolithographic structure; and
2 the wavelength of the radiation.
If << anti-reflective surfaces are created, >> light redistribution is possible. Thus,
their use in laminated glazing can offer either shading or redirection of sunlight to the
ceiling in an effort to increase light levels in areas away from windows. The angle of
diffraction is dependent upon the wavelength and this can be achieved only for a certain
angle of incidence. For all other angles of incidence, diffraction does not occur and the
HOE is transparent. Although their efficiency is relatively poor (depending upon the
wavelength), the technology is sophisticated (IEA, 2000).
CORE SYSTEMS
Core systems introduce daylight into the core of a building away from its perimeter zone.
These systems usually consist of three functions:
collecting;
transmitting; and
emitting.
A heliostat is, essentially, a mirror that tracks the sun and redirects sunlight onto a specific
receiving area. If this area is, for example, a skylight, then direct sunlight can be guided
FIGURE 2.9 Laser-cut light-deflecting panel (LCP) operational principle
ABER_CH_02 6/12/08 3:31 PM Page 41
through the core of the building. In common configurations there are two mirrors
employed: one of them is actually the heliostat, while the second one is static and is used
to direct sunlight to a predefined direction.
Due to the divergence of the suns rays (~0.5), there is a difference between the light
flux emitted from the heliostat and the light flux that impinges upon the receiving area.
This difference is due to:
mirror surface roughness;
local curvature of the mirror surface; and
tracking error.
The above parameters define the efficiency of the heliostat in transporting solar light flux.
It is evident that if the distance between two plane mirrors increases, then the light
transportation efficiency decreases. In an effort to overcome this problem, various shapes
of concave heliostats have been introduced that are capable of focusing the light beam
on the receiving area. A lens can be used to collimate the beam as well, with a further
increase in efficiency.
Glass and metal are the most common materials used for constructing the mirror
surfaces of heliostats, although the potential of stretched reflective films (which can be
metal, polymeric or composite construction) on a frame has been exploited by Murphy
Source: BartenBach LichtLabor (www.bartenbach.com)
FIGURE 2.10 Heliostat designed and installed at BartenBach LichtLabor, Innsbruck, Austria
ABER_CH_02 6/12/08 3:31 PM Page 42
(1983). Until recently, heliostat designs based on this principle were extremely limited. The
main problem associated with this type of heliostat is the expected lifetime of the material
and the rigidity of the construction.
During 2002, Sankrithi presented a prototype inflatable heliostat in which a mirrored
film is located between two equally sized inflated transparent domes. By adjusting the
pressure in these domes, various focal distances can be achieved. The whole system is
moved by using two motors and, due to its minimal weight, there are low torque
requirements.
A crucial element of heliostat efficiency is the proper tracking mechanism. The
required accuracy of this mechanism depends upon the receiver acceptance angle. Since
the application of heliostats, tracking has been performed with azimuth-elevation axes
using two motors. The motors can be controlled directly through a proper controller and
through use of irradiance sensors. These sensors help the motors to align the heliostat in
the correct direction and can be placed, for example, on the receiver. When clouds block
solar radiation, the tracking mechanism should be capable of following the suns path
using a time-based algorithm; after the suns appearance, the sensors take over control
again. Because of the level of accuracy required, rotation speed in both axes is extremely
low. Such a control mechanism is presented in Figure 2.12.
This mechanism consists of two pairs of photo-sensors separated by small fins. The
controller is trying to equalize signals from four sensors, in this way achieving perfect
alignment with the sun.
Transferring daylight with fibres is quite an old concept; Robieux (undated) was the
first to suggest the use of concentrating parabolic mirrors associated with solid light
guides. There are different types of fibre optics according to the material that has been
used for their construction, as presented in a review by Jacobs (2002). Plastic fibres made
FIGURE 2.11 Schematic drawing of the inflatable heliostat prototype
ABER_CH_02 6/12/08 3:31 PM Page 43
of PMMA have a clear PVC sheath and losses of around 3 per cent per metre. Strands
of fibres (0.25 to 3mm in diameter) can be bundled together for the increase of
cross-section. However, the section includes not only the core but the cladding and the
coating. Even with the coating removed, the cladding and the space between the fibres
leaves an unused part of the bundle section. Plastic fibres have a maximum operational
temperature of 80C.
Glass fibres are extruded from borosilicate glass with a cladding of lead silicate having
losses of > 6 per cent per metre and are produced in small diameters (~0.05mm). A
special type of glass fibre is the silica fibre, which comes with a polymer cladding and
typical diameters of 1mm. Packing losses are approximately 20 per cent, while maximum
operational temperature can reach 350C.
The last category is the liquid fibre, which normally has a large diameter (>13mm) and
consists of a pressurized tube with a Teflon cladding, filled with a liquid and plugged on
both ends with PMMA solid taps. The liquid that is especially produced to have low
infrared (IR) transmission makes this type of fibre suitable for daylighting applications. A
critical parameter for the coupling of a collection system with the fibre is the latters
acceptance angle. Core material has a refracting index n
1
, while the cladding has a lower
refracting index n
2
. If certain incidence conditions are fulfilled, a ray undergoes total
internal reflection on the surface separating the core and its cladding. These conditions
occur when the rays incidence angle is smaller than
A
:
[3]
There are some systems that have been designed to carry daylight through fibres.
A
n n =
sin ( ).
1
1
2
2
2
Source: BartenBach LichtLabor, system Solux (2004, available from www.bomin-solar.de)
FIGURE 2.12 Mechanism for alignment of a solar tracker
ABER_CH_02 6/12/08 3:31 PM Page 44
During 1979, the Himawari system (Himawari, 2004, available from www.himawari-
net.co.jp) was developed.
This system is actually a heliostat with small hexagonal Fresnel lenses the number
of lenses being 12, 36 or 90, depending upon the size capable of concentrating sunlight
to glass fibres with a diameter of 1mm. The other end of the fibres forms a luminescent
spotlight in the interior of the building.
Hybrid solar lighting systems (Muhs, 2000) consist of five major elements:
1 light sources (both sunlight and electric lamps);
2 sunlight collection and tracking systems;
3 light distribution systems;
4 hybrid lighting control systems; and
5 hybrid luminaires.
A system of two parabolic mirrors is used to concentrate solar radiation. The primary
reflective dish reflects solar radiation on a secondary mirror located at the focal point.
Since the secondary mirror is a cold mirror, only visible radiation is reflected to a bundle
of optical fibres placed in the centre of the dish. At the other end of the fibres an acrylic
rod is attached, with its top hemisphere sandblasted in order to produce a uniform
scattering surface, and both the top hemisphere and concave end mirror coated with
aluminium. This system is used as a luminaire and its similarity with a linear fluorescent
lamp is profound (Earl and Muhs, 2001). The systems total efficiency is approximately
50 per cent if sunlight is transferred to one floor.
n1 n2
FIGURE 2.13 Definition of acceptance angle for fibre optics
Source: La Foret Engineering, Himawari (2004, available from www.himawari-net.co.jp/e_page-index01.html)
FIGURE 2.14 Himawari system
ABER_CH_02 6/12/08 3:31 PM Page 45
The Solux system was developed by Bomin Solar Research (Solux, 2004, available from
www.bomin-solar.de) and is a heliostat system in which sunlight is captured by a Fresnel
lens (concentration 10,000) and then transmitted to the building interior through the use
of a liquid fibre optic. The Fresnel lens has a diameter of 1m with a focal length of 80cm
and visible transmittance of 94 per cent. To optimize the coupling efficiency into the liquid
fibre, this cone angle (which is defined by lens diameter and focal point) must be equal to
or smaller than the acceptance angle of the light fibre. To avoid thermal damage of the
PMMA end of the liquid fibre, a closed loop water system with a heat exchanger is used
to remove infrared radiation from the concentrating sunlight. The system was tested
during the 2002 Universal Fibre Optics Project.
The end of the fibre in the building interior was coupled with a flat emitter. The emitter
is a sheet of Perspex material with a white dot pattern screen printed on it (Prismex).
These dots allow the light that is trapped inside the panel through total internal reflection
to be released. By varying the size and distance of the dots, any arbitrary distribution of
luminance across the surface can be created. The design of the luminaire is mostly
Plastic Primary Mirror
Secondary Mirror
Receiver Module
Tracking Mechanism
Source: Sunlight Direct (www.sunlight-direct.com)
FIGURE 2.15 HSL 3100 hybrid lighting system
Source: Universal Fibre Optics Project (2002, www.learn.londonmet.ac.uk/portfolio/1999-2001/ufo.shtml)
FIGURE 2.16 Solux daylighting system installed in the University of Athens during the Universal Fibre
Optics research programme
ABER_CH_02 6/12/08 3:31 PM Page 46
influenced by the overall geometry of the Prismex with its printed area, which acts like a
large, even emitter. The top emitting surface can be covered with a high specular sheet
( = 93 per cent) in an effort to increase luminous intensity.
In an attempt to use larger cross-sections and to increase the ability to transfer light
flux, Callow and Shao (2003) used a set of light rods that are sufficiently small to be fitted
into most existing buildings without structural modification. During testing, rods were
found to have a transmittance equal to that of a light pipe which had an aspect ratio six
times smaller. Rods would collect both direct and diffuse light throughout the year, with
high collection efficiency.
Light pipes are quite simple daylight devices, which through the use of a highly specular
(aluminium-anodized plastic films or polymeric multilayer optical stacks) or refractive (Optical
Lighting Film) material can guide sunlight into the core of a building. Light pipes transmit
less solar radiation into the building interior, and thus have an advantage over ordinary
skylights. A simple formula for the transmittance of a reflective light pipe is:
[4]
where R, and d are the reflectance, the suns elevation and the diameter of the light pipe,
respectively. It is evident that the light pipes performance depends upon the suns
elevation; thus, during the winter months, transmittance is extremely low. Some new
designs incorporate light redirection systems (such as laser-cut panels) on top in an effort
to modify the angle of incidence of light rays.
Increase of light pipe performance can be achieved by using sun-tracking devices,
which redirect sunlight at a proper angle through the pipe; but their complexity and
intense maintenance requirements are really a disadvantage. Following other applications
in the field of daylighting and solar concentration (Welford and Winston, 1978), Molteni et al
(2001) used the concept of anidolic optics to design an external collector that is
characterized by an 80 input acceptance angle.
Oakley et al (2000) presented light-level measurements from four buildings in the UK,
concluding that light pipes can be effective devices for introducing daylight into buildings
provided that they are designed in such a way as to avoid large aspect ratios and bends.
A light pipe variant is the prism light guide that is actually a hollow pipe made of
transparent acrylic material in which one of its surfaces has longitudinal prisms. Normally,
the smooth surface of the transparent material is in the interior of the pipe. This
configuration causes the light that enters the pipes inner space to be reflected by total
reflection at the outside surface, which has a 90 prismatic ridge. Of course, total
reflection occurs providing that it is less than 27.5.
During the European research project Arthelio (Kasse et al, 19982000), a similar
system was examined with the aim of providing daylighting in deeper building interiors.
Two prototypes have been installed at the 3M distribution centre in Carpiano, Italy, and at
the Semperlux Building in Berlin, Germany. Both prototypes use the hollow light prism
guide concept with heliostats as a light collection system on the roof.
Rosemann and Kaase (2005) report a high user acceptance of both systems, while the
sunshine probability affects the energy balance of the system. Hollow horizontal light
T R
d
=
41tan

ABER_CH_02 6/12/08 3:31 PM Page 47
pipes with polygonal sections can be used to transport daylight from the faade to the
building core. Such a system has been examined by Beltrn et al (1997). Their paper
mentions some additional parameters that have to be considered:
The light pipe size should be quite small in order to fit with other building
subsystems within the ceiling plenum.
The cross-section of the light pipe must be varied in order to study the changes
in illuminance efficiency and distribution.
The reflector system needs to partially collimate incoming sunlight in order to
minimize inter-reflections within the transport section of the light pipe, and to
maximize the efficiency of the system.
The shape of the light pipe transport cross-section may have to be altered and
various reflector options may have to be investigated in order to redirect daylight
to the work plane.
Such a configuration is depicted in Figure 2.18.
Beltrn et als (1997) conclusions were that the light pipe can achieve work-plane
illuminance levels consistently above 200lx for about seven hours per day throughout the
year. Lower, but still useful, levels of daylight (>100lx) are provided for a greater range of
sun angles. This is one of the most critical parameters that directly affects light pipe
efficiency. Normally, efficiency is quite high as long as the suns rays impinge upon the
system with a small angle of incidence.
Recently, Canziani et al (2004) presented a new type of horizontal light pipe that
includes a planar closing element and a sunlight collecting and deflecting device that
optimizes the direction of the incoming solar rays as the solar position varies.
Source: Heliobus AG (www.heliobus.com)
FIGURE 2.17 Heliobus system installed at Graubunden, Switzerland
ABER_CH_02 6/12/08 3:31 PM Page 48
In a recent research programme initiated and led by Lorne Whitehead at the University of
British Columbia, Canada (Whitehead, 2006), a daylighting system was developed that
demonstrates both the feasibility and cost-effectiveness of a solar canopy system that
guides direct sunlight deep into the interior of a building. The design of the canopy system
is compatible with standard building construction techniques, which is critically important
for widespread adoption of the technology. This system consists of a number of small
moveable mirrors (mirror array) that redirect sunlight into a light pipe.
FIGURE 2.18 Horizontal light pipe
FIGURE 2.19 Horizontal light pipe proposed by Canziani et al (2004)
ABER_CH_02 6/12/08 3:31 PM Page 49
A similar system is the Aszen (Kinney and Burnett, 2002) collection system, which
consists of two systems of highly reflective blinds placed horizontally and vertically and
controlled by two motors.
During 2002, Scartezzini and Courret presented a horizontal daylight transportation
system where anidolic elements had been used in both ends.
Source: Structured Surface Physics Laboratory, University of British Columbia (www.physics.ubc.ca/ssp/solar_canopy.htm);
Whitehead (2006)
FIGURE 2.20 Prototype daylighting system developed by Whitehead (2006)
FIGURE 2.21 Anidolic ceiling
ABER_CH_02 6/12/08 3:31 PM Page 50
Anidolic systems attempt to solve two primary radiation transfer design problems:
1 maximization efficiency;
2 creation of controlled illuminance distribution.
A system that is commonly applied to solar collection is the compound parabolic
concentrator (CPC), although some designs do not use parabolic profiles and are referred
to as CPC type.
Anidolic systems have, in principle, the following properties:
The number of reflections can be minimized.
Angular selectivity can be achieved though appropriate design.
Consequently, such a system can achieve high efficiency in terms of transmitting light flux.
The main advantage of the system is the efficient collection and redistribution of
diffuse skylight, achieving an increase in daylight levels in the deeper part of the room
equal to 32 per cent, with no external obstructions, while the percentage increases to
45 per cent when the system is installed in an urban environment. Depending upon the
climate, some protection against solar gains is needed during the summer months.
PERFORMANCE METRICS
Innovative daylighting systems are complex optical devices, and it is important to know
their performance parameters through simulation or measurements. Performance can be
estimated either for the system itself by measuring/simulating its luminous and radiant
properties or for the system combined with a test cell. The latter has the advantage that
systems with arbitrary complexity can be measured, while the main disadvantage of this
method is the time needed to investigate all possible combinations of solar position and
sky luminance distribution.
Aizlewood (1993) has used criteria to define the performance of innovative daylighting
systems, percentage increase in illuminance and the glare index when the system is
compared with clear glass. Moecks (1998) work was a continuation of Aizlewoods paper,
proposing not clear glass as a base case, but an ordinary Venetian blind system, which is
more likely to be present in an office environment. According to his paper, quantitative and
qualitative criteria must be used. Among a number of criteria mentioned the following
have been used during his experiments:
luminous intensity distribution of the system;
illuminance on the working plane;
uniformity;
total light flux entering the space;
light flux on the upper back part of the space;
luminance of the window;
total area of the space that is over-illuminated;
spherical illuminance values.
ABER_CH_02 6/12/08 3:31 PM Page 51
Although the above list sufficiently describes the impact of the system to the visual
environment, additional criteria should be added, such as:
estimation of daylight savings;
estimation of glare indices;
visual contact with the exterior environment;
estimation of radiant properties (such as g-value).
By using a facility that reproduces sky luminance distributions over the hemisphere, a
daylighting system in an artificial sky can be examined. Errors associated with sky coverage
(due to the limited number of lamps used to cover the entire area of the sky) and with
parallax have to be taken into account, while the presence of a solar simulator is essential.
A method that uses artificial sky in order to estimate the performance of daylighting
systems is the integration of directional coefficients (IDC) method, which has been
presented by Papamichael and Beltrn (1993). The method uses a scale model of the
system and a space in an artificial sky. Using a collimated beam of light, the directional
illuminance coefficients are estimated. These coefficients are defined as follows:
C(azimuth, elevation) = Ei(azimuth, elevation)/Ee(azimuth, elevation) [5]
where Ei(azimuth, elevation) is the interior illuminance due to the collimated beam in the
direction specified by the pair of (azimuth, elevation), and Ee(azimuth, elevation) is the
exterior illuminance due to, and normal to, the collimated beam. A similar procedure is
applied for the calculation of skylighting as well. It is evident that if the above coefficients
have been calculated at reference points of interest, then they can be used in an algorithm
to determine daylight factors or actual daylight illuminance values under any sun, sky and
ground conditions. This method is similar to that proposed by Tregenza and Waters (1983).
Daylight coefficients for sunlight have to be calculated separately using a collimated
beam of light for various positions. Depending upon the system examined, the number of
solar positions should be increased or decreased accordingly.
Direct measurement of the performance of a daylighting system is not an easy task
since the majority of todays facilities have a limitation on the sample size. In general,
characteristics with hemispherical or conical incidence or collection of light can be
measured either using gonio-photometers (measurement of intensity distributions or
bidirectional transmittance distribution function: BTDF) or integrating sphere
photometers. With the gonio-photometric measurements, the radiation transmitted by the
sample into different directions is measured in these directions, while in sphere
photometers light flux transmitted by the sample in all directions is collected and spatially
integrated by the sphere (Aydinli and Kaase, 2002). As mentioned above, sample size is
of critical importance so that the gonio-photometer can measure a system and not a
material sample. The bidirectional gonio-photometer of Ecole Polytechnique Fdrale de
Lausane (EPFL) (Scartezzini et al, 1997) allows the use of samples with a maximum size
of 40 x 40cm, while the same size is used in a similar facility at the Fraunhofer Institute
for Solar Energy Systems.
ABER_CH_02 6/12/08 3:31 PM Page 52
A measurement device that can calibrate the segment luminous intensity distribution
of hollow prism light pipes has been developed and built at the Technical University of
Berlin (Kloss et al, 2000; Kloss, 2001). It essentially consists of an arc that can move
around a segment of the hollow light guide and a rail system that can linearly drive the
hollow light guide through the arc. As a result, all segments of the hollow light guide
system can be measured. The maximum length of light pipes that can be measured is
30m.
Simulation of BTDF for complicated daylighting systems can be performed using
specialized ray-tracing programmes such as ASAP (see www.breault.com) and TracePro
(see www.lambdares.com); but the estimated data should be treated with caution since
construction inaccuracies can affect the systems performance considerably. Another
issue is that these programmes are not linked at least directly to a lighting simulation
software in order for luminance distribution along interior surfaces to be estimated on an
hourly basis.
To examine the influence of a daylighting system in a building, thermal simulation
models should be used. In these models, the energy transport through the individual
building components is calculated on the basis of hourly values for the external climatic
conditions and for the assumed behaviour of the occupants of the building. The problem
with these models in terms of radiation is the difficulty of estimating the dynamic
performance of complex shading/daylighting systems. For todays models, evaluating the
performance of such a system can be performed by using empirical or flux-transfer
methods, radiosity-based methods or ray-tracing techniques. Energy+ (available from
www.energyplus.gov) supports light pipes and shelves, while complex fenestration
systems can be used as long as precalculated BTDFs are given.
A different approach is used by Esp-r (available from www.esru.strath.ac.uk/
Programs/ESP-r.htm). In this case the thermal algorithm is linked with Radiance (available
from www.radiance-online.org) for daylight calculation. If daylight coefficients are known,
they can be used as well.
DAYSIM (daylighting analysis software that predicts the annual daylight availability, as
well as electric lighting use in a building; see http://irc.nrc-cnrc.gc.ca/ie/light/daysim.html)
can be used in conjunction with Radiance to calculate daylight coefficients and then
estimate illuminance levels in the interior according to a predefined time step. Although it
is possible to simulate daylighting systems, only a limited number of systems (with simple
geometric properties) can be simulated with Radiance.
According to Jan de Boer (2006) a database with BTDFs of daylighting systems
should be created using gonio-photometric measurements and a calculation method
that superimposes the external luminance distribution and the BTDFs of the respective
system (TALISYS is available from www.talisys.de). This method allows one to
determine and evaluate the effects of different daylighting systems on the illuminance
distribution.
CONCLUSIONS
Innovative daylighting systems on top of their shading properties provide higher-
quality work environments that are more comfortable for occupants and also provide
ABER_CH_02 6/12/08 3:31 PM Page 53
owners with greater space value. Their main functions are to bring daylight into the building
core, to increase uniformity and to reduce glare. Nevertheless, these characteristics are
often associated by high illuminance and possible obstruction of views.
It is true that adoption of daylighting systems is rather limited because of equipment
and installation costs, high maintenance of control systems, reliance on weather and
degradation of optical properties due to air pollution. Another issue is that, in many cases,
the choice of a daylighting system is tailor made and requires detailed design.
An increase in the use of innovative daylighting products can be promoted if there is
an associated technology push to justify their benefits. These systems are normally used
in high-profile projects where cost is irrelevant and they are seen as more or less a
marketing tool. The outcomes of various European research programmes have identified
the barriers that prevent the widespread adoption of these systems and have highlighted
the necessity of improving the following issues:
data concerning the characterization of daylighting systems;
design tools, especially software tools;
demonstration and advertisement of existing case studies; and
cost reduction through mass production (if possible).
A. Tsangrassoulis: Department of Architecture, University of Thessaly, Pedion Areos, 383 34 Volos, Greece;
atsagras@uth.gr
REFERENCES
Aizlewood, M. E. (1993) Innovative daylighting systems: An experimental evaluation, Lighting Research Technology, vol 25,
no 4, pp141152
Aydinli, S. and H. Kaase (2002) Measurement of Luminous Characteristics of Daylighting Materials, Subgroup A4 Report IEA
Task 21, ECBCS Annex 29, www.iea-shc.org/task21
Beltrn L., E. Lee and S. Selkowitz (1997) Advanced optical daylighting systems: Light shelves and light pipes, Journal of the
Illuminating Engineering Society, Berkley, CA, pp91106
Bodart, M. and A. De Herde (2002) Global energy savings in offices buildings by the use of daylighting, Energy and Buildings,
vol 34, pp421429
Callow, J. and L. Shao (2003) Air-clad optical rod daylighting system, Lighting Research and Technology, vol 35, no 1, pp3138
Canziani, R., F. Peron and G. Rossi (2004) Daylight and energy performances of a new type of light pipe, Energy and
Buildings, vol 36, pp11631176
de Boer, J. (2006) Modelling indoor illuminance by complex fenestration systems based on bidirectional photometric data,
Energy and Buildings, vol 38, pp849868
DiBartolomeo, D., E. Lee and F. Rubinstein (1996) Developing a Dynamic Envelope/Lighting Control System with Field
Measurements, IESNA 4-7/8/1996, Cleveland, US
Earl, D. and J. Muhs (2001) Preliminary results on luminaire designs for hybrid solar lighting systems, in Proceedings of
Forum 2001: Solar Energy The Power to Choose, 2125 April 2001, Washington, DC
ABER_CH_02 6/12/08 3:31 PM Page 54
Edmonds, I. (1993) Performance of laser cut panels in daylighting applications, Solar Energy Materials and Solar Cells, vol 29,
no 1, pp126
Edmonds, I., G. Moore, G. Smith and P. Swift (1995) Daylighting enhancement with light pipes coupled to laser-cut light-
deflecting panels, Lighting Research and Technology, vol 27, no 1, pp2835
IEA (2000) IEA Report Task 21: A Source Book on Daylighting Systems and Components, Report LBNL-47493, July
IEA (2002) Daylight in Buildings: A Source Book on Daylighting Systems and Components, IEA SHC Task 21 Report, IEA SHC
Task 21/ECBCS Annex 29
Jacobs, A. (2002) State of the Art Review of Fibre-Optic Cable Technology, Final report of Universal Fibre Optics project,
ERK6-CT1999-00011, www.learn.londonmet.ac.uk/packages/helioptics/results.html
Kasse, H., T. Muller, A. Mingozzi, S. Bottiglioni, J. Ejhed, G. Rehm and O. Pelin (19982000) Final Report: Project Arthelio
Intelligent and Energy Optimized Lighting Systems Based on the Combination of Daylight and the Artificial Light of Sulfur
Lamps, C.n. JOR3-CT97-0177
Kinney, L. and T. Burnett (2002) Innovations in direct beam daylighting systems, Paper presented at 5th International
Conference on Energy-Efficient Lighting, 2931 May 2002, Nice, France
Kischkoweit-Lopin, M. (2002) An overview of daylighting systems, Solar Energy, vol 73, no 2, pp7782
Kloss S.-H. (2001) Ein Goniophotometer zur Messung des Lichtstromes und der Lichtstarkeverteilung von hohlen Lichtleitern,
PhD thesis, TU Berlin, Germany
Kloss, S.-H., T. Muller, A. Rosemann and H. Kaase (2000) Goniophotometrie an Hohllichtleitern: Licht 2000, Tagungsband der
14, Gemeinschaftstagung der Lichttechnischen Gesellschaften in Goslar, Germany
Koester, H. (2004) Dynamic Daylighting Architecture: Basics, Systems, Projects, Birkhauser Publishers
Kuckelkorn, T. (2002) Nichtabbildende Konzentratoren fr den Sonnenschutz in Gebuden, PhD thesis, Technische Universitat
Munchen, Germany
Lorenz, W. (2001) A glazing unit for solar control, daylighting and energy conservation, Solar Energy, vol 70, no 2,
pp109130
Matusiak, B. (2004) Lighting Systems in Smart Energy Efficient Buildings: A State of the Art, Report STF22A04504, Norwegian
Research Council, Norway
McCluney, R. (1998) Advanced fenestration and daylighting systems, Paper presented to Daylighting 1998 International
Conference on Daylighting Technologies for Energy Efficiency in Buildings, 1013 May 1998, Ottawa, Canada
Moeck, M. (1998) On daylight quantity and quality and its application to advanced daylight systems, Journal IESNA, vol 27,
no 1, pp319
Molteni, S. C., G. Courret, B. Paule, L. Michel and J. L. Scartezzini (2001) Design of anidolic zenithal: Lightguides for
daylighting of underground spaces, Solar Energy, vol 69, pp117129
Muhs, J. (2000) Design and Analysis of Hybrid Solar Lighting and Full-Spectrum Solar Energy Systems, SOLAR2000
Conference, Madison, Wisconsin
Murphy L. M. (1983) Technical and Cost Benefits of Lightweight, Starched Membrane Heliostats, SERI/TR-253-1818, Solar
Energy Research Institute, Golden Colorado, May
Oakley, G., S. B. Riffat and L. Shao (2000) Daylight performance of lightpipes, Solar Energy, vol 69, no 2, pp8998
Papamichael, K. and Beltrn, L. (1993) Simulating the daylight performance of fenestration systems and spaces of arbitrary
complexity: The IDC method, Paper presented to 3rd International Conference of the International Building Performance
Simulation Association, BS93, 1618 August, Adelaide, Australia
Robieux, J. (undated) Patent No 75-14582 (INPI) Intern Class, F03G7/02; H02KX/G02B5/10
Rosemann, A. and H. Kaase (2005) Lightpipe applications for daylighting systems, Solar Energy, vol 78, pp772780
Sankrithi, M. (2002) Low-Cost Inflatable-Support Heliostats to Enable Cost-Effective Large-Scale Solar Thermal Power, Final
technical report, US Department of Energy Inventions and Innovation Grant No DE-FG36-01GO11023, A000, 24 November
ABER_CH_02 6/12/08 3:31 PM Page 55
Scartezzini, J.-L., R. Compagnon, C. Roecker, and L. Michel (1997) Bidirectional photogoniometer for advanced glazing
material based on digital imaging techniques, Lighting Research and Technology, vol 29, no 4, pp197205
Scartezzini, J. L. and G. Courret (2002) Anidolic daylighting systems, Solar Energy, vol 73, no 2, pp123135
Tregenza, P. and I. M. Waters (1983) Daylight coefficients, Lighting, Research and Technology, vol 15, no 2, pp6571.
Universal Fibre Optics Project (2002) Final Report, ERK6-CT-1999-00011, www.learn.londonmet.ac.uk/packages/
helioptics/results.html
Welford W. and R. Winston (1978) The Optics of Nonimaging Concentrators: Light and Solar Energy, Academic Press
Publications, San Francisco, CA
Whitehead, L. (2006) Solar Canopy Illuminance System, www.physics.ubc.ca/ssp/solar_canopy.htm
Winston, R. (1974) Principles of solar concentrators of a novel design, Solar Energy, vol 16, pp8995
ABER_CH_02 6/12/08 3:31 PM Page 56
Keywords Indoor air quality; modelling; adsorption; diffusion; surface reactions; building materials;
mass transfer
Abstract
Although potentially having a significant influence on indoor air quality (IAQ), interactions
between building materials and gaseous contaminants have often been neglected or
crudely modelled in IAQ simulation tools. During the past 20 years, empirical source and
sink models have progressively given way to physically based models; but confusion still
remains on their applicability, as well as on the adequate experimental data to input for the
model parameters. Thus, demonstration is first made that models relating macroscopically
the room air phase and material concentrations through adsorption and desorption
constants are not scientifically sound. Instead, elemental models combining diffusion
equations and local sorption equilibria should be used. The compilation of sorption and
diffusion data presented in the second part of this chapter underlines the fact that such data
cannot be considered independently from the mass transport equations used to fit the
measurements. As a result, a thorough analysis of diffusion processes in polymers and
porous media is presented in order to define and relate the diffusion coefficients. Finally,
the last part of the chapter discusses the way in which existing models could be extended
to account for the contributions of temperature, multi-component mixtures, humidity and
chemical transformations within materials. Still based on fundamental considerations,
the proposed methodology consists of implementing new functionalities to describe the
temperature dependence of the model parameters, elemental models representing the
interactions between gaseous contaminants and water, as well as kinetic models coming
from the fields of atmospheric and surface sciences.
3
Physically Based Modelling of
the Material and Gaseous
Contaminant Interactions in
Buildings: Models, Experimental
Data and Future Developments
P. Blondeau, A. L. Tiffonnet, F. Allard and F. Haghighat
ABER_CH_03 6/12/08 3:31 PM Page 57
INTRODUCTION
Interactions between gaseous contaminants and materials in buildings involve a variety of
phenomena, including volatile organic compound (VOC) emissions from new or newly
applied covering materials, deposition of reactive species (ozone and sulphur dioxide, as
well as some nitrogen oxides) from the room air to the material surfaces, and what is
commonly called the reversible sink effect of the walls and furnishings (or
adsorption/desorption processes; but this term proves to be confusing and somewhat
inappropriate, as will be shown later on). These phenomena have been extensively
studied since the mid 1980s. First, the source/sink behaviour of building materials has
been characterized from environmental chamber studies. Most of them lead to the
proposal of empirical or semi-empirical models obtained by fitting an appropriate
mathematical expression to a set of experimental data, usually concentration at the air
exhaust of the chamber as a function of time (Dunn and Tichenor, 1988; Chang and Guo,
1992; Colombo and De Bortoli, 1992; Haghighat and Zhang, 1999). Although such
experiments are still relevant today, especially for standardization on material emissions,
researches have now also clearly been directed toward the development of physically
based models to be implemented in indoor air quality simulation tools.
This chapter reviews these models and provides a compilation of adequate
correlations and experimental data for the model parameters. Indeed, as noticed by
Meininghaus and Uhde (2002), there are still remarkably few sets of relevant diffusion,
sorption and reactivity data to be found in the literature compared with the vast
amount of existing emission and sink data. Therefore, synthesizing the existing
information and searching for correlations to generate more data is another key point
to strive for, for simulations of realistic configurations of building operation. Finally, the
chapter gives some tracks for future developments by presenting the way in which the
influence of temperature, humidity or, more generally, mixtures of contaminants could
be implemented in the equations describing the contaminant transports within
building materials. Since chemical transformations and, especially, secondary
emissions from building materials have recently become of great concern in health-
related indoor air quality studies (Wilkins et al, 2006), the way in which homogeneous
and heterogeneous reaction models could be implemented in these equations is also
discussed.
On no account is the purpose of the models description or compilation of data to be
exhaustive. Instead, the aim is to show the way in which existing models can be classified
based on different approaches, and subsequently to distinguish between what is
scientifically sound and what is not. The emphasis is also put on demonstrating that a
deeper look at the fundamentals of diffusion and sorption may be helpful in understanding
what has been observed experimentally, extending the model abilities, but also
simplifying the problems.
DESCRIPTION OF EXISTING MODELS
Physically based models aimed at representing the interactions between gaseous
contaminants and materials can first be classified as macroscopic and elemental models.
Then, in the case of elemental models, distinction is to be made between what will be
58 P. BLONDEAU, A. L. TIFFONNET, F. ALLARD AND F. HAGHIGHAT
ABER_CH_03 6/12/08 3:31 PM Page 58
called one-phase and two-phase models. The physical principles and related equations
defining these classes are presented hereafter.
MACROSCOPIC MODELS
The most widely used model for predicting either the VOC emission from materials or the
adsorption and desorption cycles of airborne contaminants is the one-sink model initially
proposed by Tichenor and co-workers (Tichenor et al, 1991, 1993) and subsequently developed
by many authors (Jrgensen et al, 1999; Won et al, 2000, 2001; Elkilani et al, 2003). It
considers that at any time, the mass flux at the material surface is the difference between
the adsorption and the desorption rates of the material. Analogous to Langmuirs theory,
the former is assumed to be proportional to the contaminant concentration in the bulk air
of the room, while the latter is assumed to depend upon the amount of contaminant
adsorbed at the material surface. Therefore, the governing equations for indoor and
surface concentrations of a non-reactive compound, C
(mg/m
3
) and C
s
(mg/m
2
),
respectively, are:
[1]
[2]
where C
inl
(mg/m
3
) is the pollutant concentration in the inlet air, V (m
3
) the room volume,
Q (m
3
.s
1
) the airflow rate through the room, k
a
(m.s
1
) the adsorption rate coefficient, k
d
(s
1
) the desorption rate constant, and n a constant that may account for potential
non-linearities in the desorption process but is typically assumed to equal unity. A (m
2
) is
the surface area of the material that is exposed to the room air, which means that
increased surface areas due to surface roughness and internal pores, as well as the
effects of bulk mass transport and diffusive transport processes, are implicitly lumped
into the adsorption and desorption rate coefficients.
ELEMENTAL MODELS
Unlike the one-sink model described above, and similar models that relate
macroscopically the room concentration to a bulk concentration within the material or at
the material surface, elemental models can account for the concentration gradient that
is, non-uniformity of the concentrations within the material. They usually describe the
contaminant transport as a serial assembly of two equations, one representing the
diffusion transport in the material and the other the contaminant transport at the material
surface. Depending upon whether the boundary layer resistance is considered or not, the
latter is the boundary layer equation, a simple partitioning equation of the contaminant
between the room air and the material, or a combination of the two. Although showing
similarities, elemental models can be classified into two broad groups namely, one-
phase and two-phase models, depending upon whether the material is seen as a
homogeneous solid or porous medium.
dC
dt
k C k C
s
a d s
n
=
,
V
dC
dt
Q C C Ak C Ak C
inl a d s
n

= ( ) ,
Modelling of Contaminant Interactions in Buildings 59
ABER_CH_03 6/12/08 3:31 PM Page 59
One-phase models
One phase models consider that materials are bulk homogeneous solids where Ficks law
of diffusion applies. As a result, the equation describing the transient diffusion of gases
through the material is (Little et al, 1994; Zhang and Xu, 2003):
[3]
where C
mat
(mg.m
3
) is the material-phase concentration, t (s) is time and D
mat
(m
2
.s
1
) the
diffusion coefficient of the species in the material. When coupling Equation 3 in series
with the boundary layer equation:
[4]
where C (mg.m
3
) is the air phase concentration and h (m.s
1
) the convective mass transfer
coefficient, continuity is obtained by assuming that the solid phase and near-surface air
phase concentrations constantly remain in equilibrium at the material surface. This
equilibrium relation is most often referred to as the linear adsorption isotherm of the
pollutant/material system, and is defined as:
[5]
with x (m) being the spatial dimension perpendicular to the surface and K being the partition
coefficient of the compound (note that K is used here as a generic term characterizing the
contaminant partitioning between the air phase and the solid phase; distinction between
what should be K
H
or K
p
will be made later on). It is of interest to point out here that the
combination of Equations 3 and 4 is the more general form describing the pollutant transport
between the bulk air phase of the room and the material surface. In many situations (Dunn,
1987; Little et al, 1994), one-phase models have been developed and applied based on the
assumption that C
mat
= KC
at abscissa x = 0, which implicitly means that h is infinite and

amounts to considering that the boundary layer resistance to mass transfer is negligible.
Two-phase models
Two-phase models apply for porous media and thus consider that the diffusing material is
made up of a solid phase and an air phase. In most cases diffusion through solid grains can
be considered to be negligible compared to porous diffusion. Moreover, unlike simple
diffusion, diffusion in porous materials is constrained by adsorption/desorption phenomena
all along the diffusion path. Under these conditions, the contaminant transport within the
material can be represented by the equation (Axley, 1991; Lee et al, 2005; Li and Niu, 2007):
[6]
where C (kg/kg
air
) and C
s
(kg/kg
mat
) are the local air phase and sorbed phase concentrations
in the pores, respectively. The first term of Equation 6 accounts for the gas diffusion along
D C
C
t
C
t
p
mat
air
s
=

C KC x
mat
= = at 0
=
C
t
h C
=
C
t
D C
mat
mat mat
2
ABER_CH_03 6/12/08 3:31 PM Page 60
the spatial dimension x (m), D
p
(m
2
.s
1
) representing the diffusion coefficient of the
contaminant; the latter has different expressions depending upon the assumptions made
on the diffusion process, as will be discussed further on. The second term of Equation 6
accounts for the accumulation of the diffusing species in the air phase of the pores, with
(m
3
air
/m
3
mat
) being the material porosity. Finally, the last term accounts for the
accumulation of the species at the pore surfaces, with
mat
(kg.m
3
) and
air
(kg.m
3
) being
the material and air densities, respectively. Since C
s
is related to C through the adsorption
isotherm of the gas and material system:
[7]
substituting Equation 7 into Equation 6 yields:
[8]
In the end, Equation 8 has the form of a Fickian diffusion equation expressed here in
terms of the air phase concentration in the pores. The latter can directly be coupled in
series with the boundary layer equation, Equation 4, to describe the contaminant
transports from the bulk air phase of the room to the depths of materials, and reversibly.
It is important to note that the porosity sometimes appears in the left-hand side of
Equations 6 and 8 to account for the fact that the actual cross-sectional area offered to
diffusion is not the total surface area exposed to the room air, but times this surface
area. Here, is considered to be included in the definition of D
p
, as will be discussed
later on.
From a more general point of view, it is also interesting to note that when dedicated
empirical models have been developed to represent the sink effect and the VOC emission,
physically based models such as the one-phase and two-phase models described above
can either serve as sink or source models. The latter will actually only differ from the initial
conditions considered. Taking the VOC content of the material as the initial condition of
the model will result in a source model; but this model will automatically turn to a sink
model as soon as the concentration gradient between the room air and the material is
reversed. Details on the way to determine the initial VOC concentrations in building
materials are provided further on.
Finally, it is important to point out that one-phase and two-phase models were
described considering a single slab of material; but they can easily be extended to
multilayer walls, as shown by Kumar and Little (2003), Zhang and Niu (2004), Li and Niu,
(2007) and others.
SYNTHESIS
Macroscopic or elemental models?
As previously described, macroscopic models aim to characterize the dynamics of gas
and material interactions through the sorption and desorption rate constants k
a
and k
d
.
However, still following the principles of Langmuirs theory of adsorption, some authors
D C
f
C
C
t
p
mat
air
=

\
)
C f C
C
t
f
C
C
t
s
s
=

( ) , or
ABER_CH_03 6/12/08 3:31 PM Page 61
have also proposed to define material and air equilibrium partition coefficients as the ratio
of the two rate constants:
[9]
To some extent, this illustrates the ambiguity that arose out of the definition of sorption
processes. Obviously, there is great confusion between what is the sorption dynamics
(i.e. the sink effect) and the sorption equilibrium (i.e. the sorption isotherm) of a given gas
and material system. When developing models, it is of fundamental importance to
consider that at the surface of non-porous materials, and at any point within porous
materials, the local sorption kinetics are much faster than the diffusion process in the
material. Practically, this means that the sorption transports at the solid surfaces can be
considered to be instantaneous: the molecules in the bulk air or in the pore space, on
the one hand, and the adsorbed molecules, on the other, are in equilibrium with each
other; their concentrations are related by the sorption isotherm of the contaminant and
material system. However, as diffusion will contribute to establish a concentration
gradient between the bulk air phase of the room and the depths of the material, the
sorption equilibrium must be considered locally. Especially in the case of porous
materials, space-varying pore-phase concentrations will define different equilibrium
points within the material.
Finally, there is evidence that the sorption parameters to consider in models are the
adsorption isotherms (or pseudo-absorption isotherms) of the contaminants. The sorption
dynamics of a wall or furnishings do not relate to the sorption transports themselves, but
to the kinetics of boundary layer and solid diffusion. Hence, there is no physical sense in
modelling the sink effect by macroscopically relating the room concentrations to the
corresponding material bulk concentrations through adsorption and desorption constants,
or the ratio of the two. Besides, these constants would depend upon the material
thickness, instead of being intrinsic properties of the gas and material systems
investigated. Such models, as well as the related experimental data, are not scientifically
sound and should be disregarded.
One-phase or two-phase models?
One-phase models consider the materials as homogeneous solid media and,
consequently, hold that adsorption/desorption processes occur at the material surface
exposed to the room air. On the other hand, two-phase models consider diffusion within
the pores and adsorption/desorption transports occurring locally at the pore surfaces all
along the diffusion path.
Theoretically, one-phase models should only hold for materials such as polymers that
do not have any measurable porosity, but allow for gas diffusion. Indeed, there is
fundamentally no good reason to consider that sorption will occur only at the material
surface when the pore surface area available is usually much larger; for instance,
Blondeau et al (2003) report effective surface areas as high as 500, 1750 and 1600m
2
.kg
1
for brick, particle board and gypsum, respectively. Moreover, one must keep in mind that
K
k
k
a
d
=
ABER_CH_03 6/12/08 3:31 PM Page 62
adsorption/desorption is a surface phenomenon. Therefore, if the material has no
porosity, or is modelled that way, the sorbed phase concentrations should not be defined
in units of mass of gas per mass of material, but in units of mass of gas per unit surface
area exposed to the bulk air (in the case of homogeneous porous materials, the specific
surface area of the material is correlated to its volume; consequently, the sorbed phase
concentration can equally be expressed in units of kg/m
2
or kg/kg
mat
). However, one would
then have to face the problem of solving the diffusion equation with boundary conditions
defined in terms of surface concentrations, which is tricky to say the least.
In the end, the physical consistency of one-phase models applied to non-porous
polymeric-like building materials (e.g. paints and vinyl flooring) may come from a different
interpretation of the contaminant partitioning between the air phase and the solid phase
at the material surface. As suggested in many references dealing with the transport of
gases in solid polymers (Treybal, 1981; Sun et al, 1997; Rutherford et al, 2005), the mass
transports at the material surface should be seen as absorption/desorption phenomena
rather than adsorption/desorption processes, and modelled analogously with Henrys law
for dilute liquid solutions. For gas/solid systems, the latter will be expressed as:
[10]
where C (g/m
3
air
) and C
mat
(g/m
3
mat
) are the equilibrium air phase and material
concentrations, respectively, and K
H
is a temperature-dependent constant that represents
the solubility of the species in the material. Similarly to adsorption processes, the kinetics
of absorption are likely to be so rapid compared to diffusion transports that they can be
treated as instantaneous. On this assumption, the near-surface air phase and material
concentrations will constantly remain in equilibrium, as described by Equation 10, but also
by Equation 5. Therefore, one-phase models as described by Equations 3 to 5 are most
certainly relevant for the modelling of gaseous species transports in polymeric materials;
but K should be referred to as the pseudo contaminant solubility in the polymeric material
(K
H
) rather than the partition coefficient of the adsorption isotherm (K
p
).
Nodal, zonal or computational fluid dynamics (CFD) models
In addition to the criteria used above, source and sink models can be classified as nodal,
zonal and computational fluid dynamics (CFD) models based on the level of accuracy used
to describe the contaminant transport in the room air. Nodal models consider that the
contaminant is instantaneously and perfectly mixed within the room air. They do not require
any information about the airflow distribution; but accordingly do not provide any
information on the contaminant distribution within the room. Most of the source and sink
models were implemented based on this assumption. In the CFD approach, the room and,
in some cases, the material is divided into a large number of small control volumes or
elements (Yang and Chen, 2001; Murakami et al, 2003; Ataka et al, 2004). Detailed
information about the airflow in the bulk air phase of the room, as well as at the material
surface, is obtained by solving the equations expressing the conservation of momentum
and energy. This may prove to be useful in accurately evaluating the convection mass
transfer coefficient, which is of central importance when boundary layer diffusion is the
C K C
mat H
=
ABER_CH_03 6/12/08 3:31 PM Page 63
rate limiting process of the gas and material interaction. Finally, zonal models are
intermediate between CFD and nodal models. These models divide the room volume into
a small number of cells where the air is assumed to be perfectly mixed. The airflow rate
between adjacent cells is given either by the Bernouilli equation or specific empirical
equations representing specific driving flows, such as thermal plumes, jets or boundary
layer flows. Solving the coupled heat and mass conservation equations within each cell
provides a rough overview of the airflow and resulting contaminant concentration
distributions within the room. Damian et al (2003) have presented a representative example
of how these room models can be coupled with Equations 4 and 8 to account for the
simultaneous influence of ventilation and interactions with materials on IAQ.
EXPERIMENTAL DATA
As well as knowing which models are relevant or which models are the most suitable for a
given application, the problem of modelling the material and contaminant interactions is also
to input adequate experimental data for the model parameters namely, the convection
coefficients at the material surface, the adsorption isotherm or the so-called solubilities for
polymeric materials, the contaminant diffusivities and the initial concentrations in the
materials. Somewhat surprisingly, this proves to be extremely tricky, especially for sorption
and diffusion properties. The reasons are explained hereafter, together with a compilation of
existing data and correlations for gas and building material systems.
CONVECTIVE MASS TRANSFER COEFFICIENTS
The convection average mass transfer coefficient of Equation 4 strongly depends upon
the characteristics of the airflow at the material surface. It can be measured directly
(Matthews et al, 1987; Haghighat and Zhang, 1999; Morrison et al, 2003) or indirectly
using the naphthalene technique, as described by White (1991). Assuming isothermal
conditions (i.e. the boundary layer only originates from a concentration gradient) and low
enough indoor air velocities to consider that the flow is laminar, h (m.s
1
), can be obtained
from correlations derived from the heat and mass transfer analogy. These correlations
relate the Sherwood number (Sh) to the Rayleigh (Ra) and Schmidt (Sc) numbers in the
case of natural convection, and to the Reynolds (Re) and Schmidt (Sc) numbers in the case
of forced convection. For a characteristic length L (m), these are defined by:
[11]
[12]
[13]
[14]
Sc =

D
m
Re =

U L
Ra
g CL
D
c
m
=

3
Sh
hL
D
m
=
ABER_CH_03 6/12/08 3:31 PM Page 64
where D
m
(m
2
/s) is the molecular diffusivity of the contaminant in air, U
(m/s) the air

velocity outside the boundary layer, g the gravitational constant (m.s
2
), (m
2
.s
1
) the
kinematic viscosity of air and C (kg/m
3
) the concentration gradient within the boundary
layer. On the assumption that the ideal gas law applies,
c
(m
3
.kg
1
), the volumetric
expansion coefficient of the species, is given by:
[15]
where V
M
(m
3
/mol) is molar volume, and Mand M
air
(kg/mol) are the molecular weights of
the contaminant and air, respectively.
Based on the assumption that the boundary layer airflow is similar to unidirectional
flow over a flat plate with a characteristic length scale taken as the length of the wall in
the direction of the flow, Bejan (1995) has suggested the following correlations for natural
and forced convection with laminar flow.
For natural convection:
[16]
[17]
For forced convection:
[18]
[19]
In the case of mixed convection, the dominating convection mechanism is decided by the
smaller of natural and forced film thicknesses. Bejan (1995) has shown that the problem
consists merely in studying the ratio:
[20]
If less than unity, the dominating mechanism is natural convection, and Equations 16 and 17
apply. Otherwise, forced convection prevails and h can be derived from Equations 18 and 22.
Still assuming isothermal conditions but turbulent flow at the material surface, White
(1991) suggests that h be calculated from the equation:
[21]
which is also derived from the heat and mass analogy for forced convection over a flat
plate.
Sh Sc
L
= 0 037
1
3
4
5
. Re ,
Re Sc
Ra
Sc
Sc
Ra
Sc
L
L
L
L
1
2
1
4
1
4
1
2
1
3
1
4
1 1 if or if <
Re
.
Sh Re Sc Sc
L
= < 1 128 1
1
2
1
2
. if
Sh Re Sc Sc
L
= 0 664 1
1
2
1
3
. if
Sh Ra Sc Sc
L
= < 0 8 1
1
4
1
4
. if
Sh Ra Sc
L
= 0 671 1
1
4
. if
c
M air
M air air
V M M
V C M M MM
=

( )
( )
ABER_CH_03 6/12/08 3:31 PM Page 65
TABLE 3.1 Representative adsorption isotherm models for building applications
MODEL C
S
= f (C) MODEL PARAMETERS
Linear K
P
: Partition coefficient
Langmuir C
s
0
: Monolayer sorbent concentration
K
L
: Langmuir constant
Freundlich K
F
: Freundlich constant
n
F
: Freundlich exponent
Temkin a, b: Temkin constants
Brunauer Emmett C
s
0
: Monolayer sorbent concentration
and Teller (BET) K
BET
: Brunauer, Emmett and Teller
(BET) constant
C
sat
: Saturation concentration
Source: Axley (1993)
SORPTION PARAMETERS
The sorption properties of interest for the modelling of both the sink effect and VOC
emissions is the contaminant adsorption isotherm for porous materials, and the pseudo
absorption constant K
H
of Henrys absorption law for non-porous materials such as polymers.
When various adsorption isotherm models exist (see Table 3.1), the adsorption equilibrium of
gaseous contaminants in buildings has most often been described as a linear correlation
between the sorbed phase and air phase concentrations (e.g. see Equation 5). To some
extent, this is most certainly a reasonably good assumption in many circumstances since
contaminant concentrations in indoor settings typically do not exceed few parts per billion
(ppb), on the one hand, and, for physical consistency, all sorption isotherm models should
reduce to linearity at low concentration, on the other. However, as noticed by Elkilani et al
(2003), the upper limit of what is called low concentrations is not clearly defined. In particular,
the linear isotherm assumption may fail in representing the first stages of the VOC emission
process. Indeed, newly manufactured materials often exhibit pore-phase VOC concentrations
that are several orders of magnitude higher than those commonly found in indoor settings.
Within this concentration range, the adsorption isotherm may show non-linearities, as shown
by Tiffonnet et al (2002) when dealing with the emission of acetone from particle board.
Practically, sorption parameters have been determined in the context of characterizing
either the sink or source behaviour of building materials. Many experiments have been
carried out and partition or absorption coefficients for many gas and material systems can
be found in the literature. Much of this data is, nevertheless, inadequate, uncertain or
proves to be tricky for modellers and practitioners. The reasons are fourfold:
1 Many authors have determined K values (K
p
or K
H
) by fitting Equation 5 or other
macroscopic models to measured concentration profiles (non-equilibrium
C K C
s p
=
C
C K C
K C
s
s L
L
=
0
1
C K C
s F
n
F
=
C aln bC
s
= ( )
C
C K
C
C
C
C
C
C
K
C
C
s
s BET
sat
sat sat
BET
sat
=
\
)
\
0
1 1
))

ABER_CH_03 6/12/08 3:31 PM Page 66
concentrations). K was either defined by the ratio of adsorption and desorption
constants, k
a
/k
d
, or the ratio of rate coefficients describing the rates of mass transfer
among the sinks (Colombo et al, 1993; Jrgensen et al, 1999, 2000; Won et al, 2000,
2001; Singer et al, 2004). However, as discussed in section Synthesis, such data
characterizes the sorption dynamics (which includes the contribution of diffusion)
rather than the sorption equilibrium, and is therefore not relevant. The most
compelling proof is that the proposed values would prove to depend upon the
material thickness when the partition coefficient is a physical property of the gas and
material system.
2 In other cases, the partition coefficients have been determined simultaneously with
the contaminant diffusivities by fitting one-phase or two-phase models to transient
sorption or desorption data (He et al, 2005; Li and Niu, 2005). Such models consider
local sorption equilibrium and thus do provide equilibrium sorption data even if drawn
from dynamic tests. The problem here is that two parameters are to be determined
simultaneously; the best mathematical fitting does not provide any guarantee on the
physical consistency of each parameter.
3 The units used for the air phase concentrations, sorbed phase concentrations and
partition coefficients add one step of complexity to the problem. As discussed earlier,
in the case of porous materials, the surface area available for adsorption is the
effective surface area of the material (sometimes called BET surface area) and the
sorbed phase concentrations should be expressed accordingly in units of mass of
contaminant adsorbed per square metre of pore surface area. As a result, partition
coefficients determined from sorbed phase concentrations expressed in units of
mass of contaminant per nominal surface area of the material (surface area exposed
to the bulk air of the test chamber) have no sense; the proposed values would
actually prove to vary with the size of the material sample tested, which is not
consistent. On the other hand, since the effective surface area of a porous media
depends linearly upon its volume, the sorbed phase concentrations can also be
expressed in units of mg/kg
mat
or mg/m
3
mat
, which is undoubtedly more convenient
for modelling purposes, but raises another critical problem regarding the data to
input in the model: as defined by Equation 10, or the linear model presented in
Table 3.1, the partition coefficient describing the adsorption or absorption equilibrium
is defined by the ratio of the sorbed phase and air phase concentrations. Thus,
whether these two concentrations are in units of mg/kg or mg/m
3
, K (K
p
or K
H
) will be
dimensionless and, in practice, most often presented that way in the literature.
However, it is essential to consider that K will actually have units of kg
air
/kg
mat
or
m
3
air
/m
3
mat
, so we have:
[22]
where
mat
and
air
(kg/m
3
) are the material and air densities, respectively. Given that
mat
is from two to three orders of magnitude higher than
air
, partition coefficients
K
kg
kg
K
m
m
air
mat
air
mat
air
mat
l
l
l
l
=
l
l
l
l
l
3
3
ABER_CH_03 6/12/08 3:31 PM Page 67
determined from volume concentrations will also be from two to three orders of
magnitude higher than those determined from mass concentrations. Although
being of fundamental importance when proposing or reporting K values, this
information is most often omitted or hard to find. Systematically detailing the
units of partition coefficient data would definitely be an easy way to avoid
confusions and mistakes.
4. Although some experiments have been carried out in dry conditions, most of the
existing sorption data were determined at typical relative humidity levels indoors
(from about 40 to 60 per cent; see Table 3.2). The problem is that humidity can either
decrease or promote the sorption of contaminants due to the competition for the
active sites or other complex interaction processes within the material. This will be
discussed in section Influence of humidity. We simply note here that partition
coefficients determined in humid conditions may not correctly represent the intrinsic
sorption properties of the contaminant species.
Tables 3.2 and 3.3 present a non-exhaustive compilation of relevant sorption data based
on the previous remarks. A distinction has been made between the data that were fit
from one-phase and two-phase models or, similarly, what should actually be considered
to be absorption coefficients of Henrys law (K
H
) or partition coefficients of the linear
adsorption isotherm model (K
p
). As one will notice from the first list, some data have
been determined from one-phase models when the materials investigated are likely to
be porous. Such data can, nevertheless, be input in two-phase models, provided that
adequate units are used.
DIFFUSION COEFFICIENTS
Diffusion models and definition of diffusivities
Diffusion of gases in non-porous building materials such as synthetic floorings, paints and
other kinds of polymeric materials can be characterized by a bulk diffusivity that depends
both upon the diffusing species and the chain structure of the polymer. Diffusion through
polymers requires that a hole or passage of sufficient size be available and, thus, in turn,
depends upon the thermal motion of the polymer chains (Treybal, 1981). Practically, this
means that small molecules are expected to diffuse faster than larger ones, and that the
diffusion rate of any species will increase with temperature, as will be further discussed
in section Influence of temperature. A more detailed description of the mechanisms
involved in gas diffusion in polymers and their practical implications on VOC emissions
can be found in Cox et al (2001a).
Unlike diffusion in polymers, diffusion of gases in porous materials involves a complex
variety of processes, including air phase diffusion and surface diffusion within the porous
interstices (Treybal, 1981; Ruthven, 1984; Masel, 1996).
Air phase diffusion puts into play two distinct diffusion processes depending upon the
pore size considered: molecular diffusion occurs in the largest pores where collisions
between gaseous molecules are predominant, while Knudsen diffusion occurs in the
smallest pores where the mean free path of the diffusing species is limited by the pore
dimensions. The mean free path, (m), is the average distance travelled by molecules
ABER_CH_03 6/12/08 3:31 PM Page 68
TABLE 3.2 Compilation of sorption and diffusion coefficients: Data fit with one-phase models (K K
H
)
REFERENCE MATERIAL/CONTAMINANT K D (M
2
/S) COMMENTS
Little et al Carpet 1a/styrene 4200 4.1 10
12
K in units
(1994) Carpet 1b/styrene 6500 3.6 10
12
of m
3
air
/m
3
mat
Carpet 1a/4- 1400 5.2 10
12
ethenylcyclohexene
Carpet 1a/ethylbenzene 1500 10.2 10
12
The numbers
and xylenes following carpet
identify different
Carpet 1b/ethylbenzene 2400 4.3 10
12
samples
and xylenes
Carpet 1a/4-phenylcyclohexene 81,000 0.59 10
12
K and D determined
Carpet 1b/4-phenylcyclohexene 67,000 0.50 10
12
as the best fit of
Carpet 3/formaldehyde 11,000 3.2 10
12
published
Carpet 3/acetaldehyde 1 6.4 10
12
emission data
Carpet 3/2,2,4-trimethylpentane 59,000 0.060 10
12
Carpet 3/1,2-propanediol 180,000 0.065 10
12
Carpet 3/2-ethyl-1-hexanol 450,000 0.088 10
12
Carpet 4/styrene 5700 3.1 10
12
Carpet 4/4-ethenylcyclohexene 1700 2.1 10
12
Carpet 4/ethylbenzene 5300 1.5 10
12
and xylenes
Carpet 4/4-phenylcyclohexene 170,000 1.2 10
12
Bodalal et al Carpet backing material/toluene 6171 43.10 10
12
K in units
(2000) of m
3
air
/m
3
mat
Carpet backing material/nonane 6216 28.30 10
12
T = 23C 1C
Carpet backing material/decane 14,617 5.42 10
12
RH = 50% 5%
Carpet backing 24,255 2.79 10
12
material/undecane
Plywood/cyclohexane 348 155 10
12
Plywood/ethylbenzene 1636 40.4 10
12
Plywood/decane 6948 12.8 10
12
Floor tile/ethylbenzene 1920 16 10
12
Floor tile/nonane 2142 14.8 10
12
Floor tile/decane 13,045 2.09 10
12
Floor tile/undecane 26,647 0.85 10
12
Cox et al Vinyl flooring/n-butanol 810 77 T = 25.6 0.3C
(2001b) Vinyl flooring/toluene 980 34 Dry conditions
Vinyl flooring/phenol 120,000 3000
Vinyl flooring/n-decane 3000 420
Vinyl flooring/n-dodecane 17,000 260
Vinyl flooring/n-teradecane 120,000 1300
Vinyl flooring/n-pentadecane 420,000 38,000
ABER_CH_03 6/12/08 3:31 PM Page 69
H
)
(Contd)
2
/S) COMMENTS
Yang et al Gypsum board/ethylbenzene 1550 K in m
3
air
/m
3
mat
(2001) Gypsum board/ethylbenzene 10.05 T = 23C 1C
Gypsum board/ethylbenzene 34.90 RH = 50% 2%
Tiffonnet et al Acrylic paint/acetone 2.80 at 20C K in g
air
/g
mat
(2002) 2.56 at 25C (original data in
2.53 at 30C units of g
air
/m
2
mat
)
He et al ACT2/tetradecane 400 500 10
12
ACT: acoustic
(2005) ceiling tile
ACT3/hexanal 5524 109 10
12
CRP: carpet
CRP1/4-phenylcyclohexene 66,833 235 10
12
GB: gypsum board
CRP1/heptane 300 609 10
12
OSB: oriented
strand board
CRP2/dodecane 986 192 10
12
PB : particle board
CRP2/limonene 1441 163 10
12
PLY : plywood
CRP2/styrene 268 364 10
12
K in units
of m
3
air
/m
3
mat
CRP3/4-phenylcyclohexene 10,512 117 10
12
Partition coefficients
CRP3/dodecane 3840 174 10
12
determined
CRP3/limonene 1400 325 10
12
simultaneously
CRP3/styrene 275 455 10
12
with diffusivities
CRP3/tetradecane 22,476 12 10
12
and initial
CRP3/tridecane 7336 81 10
12
concentrations
CRP4/dodecane 42,549 75 10
12
by fitting one-phase
CRP4/iso-octane 50,000 9 10
12
model to various
CRP4/tridecane 69,942 72 10
12
emission data
CRP5/iso-octane 84,151 1 10
12
measured in
CRP6/tridecane 71,560 136 10
12
small chamber tests
GB1/-pinene 80,329 10 10
12
Proposed values
GB2/-pinene 80,200 5 10
12
are optimal values
GB3/-pinene 52,744 16 10
12
from the best fit
OSB1/-pinene 194,080 0.2 10
12
of emission
OSB2/-pinene 21503 38 10
12
profiles
OSB2/furan 60,256 6 10
12
PB0/-pinene 849 99 10
12
The numbers
PB0/camphene 1090 106 10
12
following the
PB0/hexanal 550 74 10
-12
material name
PB0/limonene 334 126 10
12
identify different
PB4/-pinene 86 800 10
12
samples
12
ABER_CH_03 6/12/08 3:31 PM Page 70
between collisions with other particles. Assuming that the ideal gas law applies, it can be
determined from the equation (Mulder, 1991):
[23]
where T (K) is temperature, N
A
the Avogadros number, R (J.mol
1
.K
1
) the universal gas
constant, d (m) the diameter of the gas molecules, and P (Pa) the pressure of the diffusing
system. In the absence of data for the molecule diameter, the mean free path can also be
estimated from the relation (Treybal, 1981):
[24]
=

(
\
)
3 2
2
0 5
.
.
P
RT
M
=
RT
d N P
A
2
2
H
)
(Contd)
2
/S) COMMENTS
PB5/-pinene 492 655 10
12
12
PB5/hexanal 70,927 16 10
12
PB5/limonene 1898 900 10
12
12
PB6/hexanal 19,650 66 10
12
PLY1/3-carene 37,562 388 10
12
PLY2/3-carene 125,632 0.4 10
12
PLY2/-pinene 80,560 0.4 10
12
PLY2/p-cymene 29,736 0.4 10
12
PLY3/3-carene 19,676 22 10
12
PLY3/-pinene 6749 0.1 10
12
PLY3/limonene 11,454 36 10
12
PLY3/p-cymene 15,974 44 10
12
Elkilani et al Carpet fibres/toluene 0.86 at 25C Dry conditions
(2003) 0.73 at 30C
0.50 at 35C Original data in
0.34 at 45C units of m
Carpet fibres/1,2-dichlorobenzene 1.60 at 25C converted to
0.83 at 30C g
air
/g
mat
using
0.74 at 35C the specific
0.36 at 45C surface area
Carpet fibres/1,1,1-trichloroethane 0.18 at 25C of the carpet
0.16 at 30C fibres provided
0.13 at 35C by the authors
0.11 at 45C
ABER_CH_03 6/12/08 3:31 PM Page 71
TABLE 3.3 Compilation of sorption and diffusion coefficients: Data fit with two-phase models (K K
p
)
REFERENCE MATERIAL/ K D (M
2
/S) COMMENTS
CONTAMINANT
Silberstein (1989) Gypsum/ 5 In units g
air
/g
mat
formaldehyde
Radulescu- Gypsum/ 0.094 Dry air
Bouilly et al (2006) tetrachloroethene T = 20C
Tiffonnet et al (2002) Gypsum/acetone 0.69 at 20C 1.60 x 10
6
In units g
air
/g
mat
Dry air
Particle board/ 0.70 at 20C 1.41 x 10
6
K
p
values for acetone
acetone 0.60 at 25C 1.46 x 10
6
in particle board
0.46 at 30C 1.51 x 10
6
were determined
Concrete/ 0.25 at 20C 4.92 x 10
7
from the low
acetone concentration
asymptote of non-
linear isotherms
Effective diffusivities
in the materials
were determined
independently
from mercury
intrusion
porosimetry
tests (see
Blondeau et al,
2003)
Meininghaus Carpet with styrene 0.381 3.5 10
7
K in units of g
air
/g
mat
et al (2000) budadiene rubber T = 24 0.5C
(SBR) backing/ RH = 45 3%
n-octane
Aerated concrete/ 0.12 7.6 10
7
n-octane
Solid concrete/n-octane 0.031 1.0 10
7
Brick/n-octane 0.016 3.2 10
7
Gypsum/n-octane 0.105 8.4 10
7
Carpet with SBR 0.170 4.5 10
7
backing/ethyl-acetate
Aerated concrete/ > 2.614 > 5.7 10
7
ethyl-acetate
Solid concrete/ > 0.590 > 0.5 10
7
ethyl-acetate
Brick/ethyl-acetate 0.129 4.7 10
7
Gypsum/ethyl-acetate 0.133 11.2 10
7
ABER_CH_03 6/12/08 3:31 PM Page 72
where (Pa.s) and M (kg.mol
1
) are the dynamic viscosity and molecular weight of the
diffusing system, respectively.
For a given pore radius, r (m), the dominant diffusion mechanism is determined by the
ratio of the pore diameter and the mean free path of the gas molecules. In the range 2r/ from
roughly 0.2 to 20, both molecular and Knudsen diffusion have influence. When 2r/ is much
less than unity, diffusion is induced by the bouncing of the molecules to and from the pore
walls, and is thus represented by the Knudsen diffusivity, D
k
(m
2
/s), which is defined as:
[25]
On the other hand, molecular diffusion prevails when 2r/ is much larger than unity. In
such conditions, the pore diffusion rate depends upon the contaminant molecular
diffusion coefficient in air, D
m
(m
2
/s), which can either be found in tables or evaluated from
online tools such as the one developed by the US Environment Protection Agency (EPA,
2006). This molecular diffusion coefficient in air is in the order of 10
5
to 10
4
m
2
/s for most
gases at normal temperature (Do, 1998).
Since building materials usually exhibit a wide range of pore sizes (see Blondeau et al,
2003), both molecular and Knudsen diffusions can occur simultaneously within a unit
volume of material. Considering, furthermore, that the porous structure is not made up of
straight and cylindrical pores, but a series of interconnecting tortuous paths of varying
cross-sectional area, the contaminant flux in the air phase of the pores, J
p
(kg.m
2
.s
1
), can
be described using Ficks law by defining an effective diffusion coefficient, D
e
(m
2
.s
1
):
[26]
with:
[27]
In Equation 27, , the connected-pore volume fraction, accounts for the real cross-sectional
area offered to diffusion, while , the tortuosity factor, accounts for the geometric constraints
that give a longer diffusion path in the flow direction.
p
, the constriction factor, stands for the
effects of varying cross-sectional areas within a pore. It is often omitted, either because it is
close to unity or because these effects are included in the tortuosity factor. Finally, D (m
2
/s)
is the mean diffusion coefficient in the pores of the material. The latter can be calculated from
the pore-volume distribution and an extension of the well-known Bosanquet equation:
[28]
where f(r)dr is the volume fraction for the pores with radii between r and r + dr.
Surface diffusion consists in the jumping of adsorbed gas molecules from one
adsorption site to another. It occurs even at very low sorbed phase concentrations and
can be modelled using Fickian-like diffusion laws expressed in terms of the sorbed phase
D
D D r
f r dr
m k
=
\
)
1 1
0
( )
( ) ,
D
D
e
p
=

J D C
p air e
=
D
r RT
M
k
=
2
3
8

ABER_CH_03 6/12/08 3:31 PM Page 73
concentration, C
s
(kg
species
/kg
mat
). This way, the contaminant flux in the flow direction x, J
s
(kg.m
2
.s
1
) is given by:
[29]
or: [30]
where
mat
(kg.m
3
) is the bulk density of the porous material, and D
s
(m
2
.s
1
) the surface
diffusivity of the contaminant.
As surface diffusion occurs in parallel to molecular or Knudsen diffusion processes,
the total diffusion flux is the sum of the contributions of pore and surface diffusions:
[31]
Therefore, substituting Equations 26 and 30 into equation 31 gives:
[32]
Finally, if further assuming a linear adsorption isotherm:
[33]
Equation 32 turns to the classic form of Ficks second law of diffusion, written in terms of
only the pore gas-phase concentration:
[34]
with:
[35]
As defined by Equation 35, D
p
(m
2
.s
1
) represents the more general form of the porous
diffusion coefficient used in Equation 8. However, care must be taken that if the basic
form of Ficks first law for diffusion:
[36]
is taken as the reference mass transport model in the porous material, the diffusion
coefficient of interest is the so-called apparent diffusivity of the contaminant and material
system, D
app
(m
2
.s
1
). The latter embeds both the air phase and surface diffusion
coefficients and the slope of the adsorption isotherm, K
p
:
[37]
D
D D K
K
app
e
mat
air
s p
mat
air
p
=
.
D C
C
t
app
=

2
D D D K
p e
mat
air
s p
=

.
J D C
tot air p
= ,
=
C
C
K
s
p
,
J D D
C
C
C
tot air e mat s
s
=

\
)
.
J J J
tot p s
= .
J D
C
C
C
s mat s
s
=

,
J D C
s mat s s
= ,
ABER_CH_03 6/12/08 3:31 PM Page 74
In the field of indoor air-quality modelling, it is common to ignore surface diffusion, either
due to the lack on knowledge on the actual surface diffusivities of contaminants in
building materials, or because it is thought to be negligible compared to air phase
diffusion. The relevance of this assumption can be assessed by comparing the two terms
of the right-hand side of Equation 35.
After Blondeau et al (2003), the effective diffusivities of VOCs in building materials
range from 10
6
to 10
7
m
2
.s
1
. On the other hand, Treybal (1981) indicates that surface
diffusivities are typically of the order of 10
7
to 10
9
m
2
/s at ordinary temperatures for
physically adsorbed gases. However, considering that surface diffusivities significantly
decrease with decreasing sorbed phase concentrations, and that the latter are usually low
compared to industrial applications, D
s
values of the order of 10
9
m
2
/s are more likely.
Partition coefficients also lie in a very wide range depending upon the material and
contaminant systems considered. K
p
values are generally of the order of 10
2
to 10
1
; but
Silberstein (1989) reported a value of K
p
as high as 5.5kg/kg
air
for formaldehyde in gypsum
board (see Table 3.2). As the density of building materials typically ranges from 200kg.m
3
to 3000kg.m
3
, one can finally note that
mat
D
s
K
p
may be of the same order of magnitude
as
air
D
e
, and will most of the time be from one to three orders of magnitude smaller.
Therefore, contrary to what one might think, surface diffusion may be significant and even
the dominant diffusion process in some practical situations. This indicates that diffusion
in porous materials can be controlled by sorption, which will prove to be of central
importance when discussing the order of magnitude of diffusivities, as well as the
influence of environmental parameters on the sorption and diffusion properties of gases
in building materials.
Another noteworthy point coming from the above discussion, and which is of great
importance in regard to the following sub-section, is that diffusion coefficient data
without any reference to the related diffusion model are worthless and unusable for
modelling issues.
Experimental data
Most of the contaminant diffusivities in building materials have been determined either
from diffusion-cell methods or by fitting one-phase or two-phase models to emission or
sink data.
The general principles as well as the drawbacks of diffusion-cell methods have been
described by Park et al (1996) and, later on, by Haghighat et al (2002). In the frame of
IAQ-related studies, two kinds of diffusion-cell methods have been used: the cup method
(Kirchner et al, 1999; Hansson and Stymne, 2000; Meininghaus and Uhde, 2002) and twin
chamber methods (Bodalal et al, 2000; Meininghaus et al, 2000). The cup method
consists of measuring the diffusion capacity of the tested system by weighting the
diffusion loss of VOC through a material sample covering a cup that contains the VOC at
saturation concentration (temperature and humidity are controlled). In twin chamber
methods, the diffusion capacity is assessed by measuring the VOC concentration in the
supply and exhaust air of two small-scale chambers placed on each side of the material
sample to be tested. In both cases, the introduction of the measured steady-state
concentrations and VOC flow rate in Ficks first law of diffusion provides the intended
ABER_CH_03 6/12/08 3:31 PM Page 75
diffusion coefficients. These may be solid-phase, effective or apparent diffusivities,
depending upon whether the assumption of porous material is considered or not. In all
cases, one problem is that the concentrations at the material sample surfaces can hardly
be measured. As a result, the concentration gradient in the material is most often taken
as the bulk air concentration difference between the primary and secondary side of the
cell; it includes the resistance to diffusion in the boundary layers, which leads to an
underestimation of the actual diffusivity of the diffusing system (Meininghaus et al, 2000;
Blondeau et al, 2003). In the specific case of the cup method, another drawback lies in the
material being subjected to VOC saturation concentration, which is unrealistically higher than
typical VOC concentrations in indoor settings. As diffusion coefficients have been shown to
be concentration dependent at high concentrations, the data may not be representative of
actual diffusivities in materials as installed in buildings (Blondeau et al, 2003).
Haghighat et al (2002) proposed an extensive compilation of published VOC
diffusivities in building materials. As for sorption properties, the data derived from
concentration profiles measured in environmental chambers may be classified into two
groups: those where the diffusivities were determined simultaneously with other
parameters (partition coefficient and, in some cases, convective mass transfer coefficient
and initial concentrations) from measured transient concentrations (Bodalal et al 2000; He
et al, 2005; Li and Niu, 2005), and those where the sorption properties were determined
first by considering the steady state, diffusivity then being defined as the best fit of the
mass transport equation to the sorption or desorption concentration profiles
(Meininghaus et al, 2000; Cox et al, 2001b). For the reasons explained previously, the
former are much more uncertain than the latter. This was confirmed by Li and Niu (2005),
as well as He et al (2005), who report potentially large uncertainties on diffusivities due to
cross-uncertainties between each parameter.
From a practical point of view, one common limitation of all the methods discussed
above is that the experiments are costly in time and, perhaps, money. In this context,
Blondeau et al (2003) suggested another method based on the analysis of the material
porosity first, and then the application of Carniglias empirical model to determine the
tortuosity factor . The effective diffusivity of any gas in the tested material can finally be
computed from Equation 27, provided that its molecular diffusivity in air is known. A
comparison between diffusivities determined that way, and data coming from traditional
diffusion-cell methods, revealed a good agreement and emphasized the potential of the
methodology to generate a large number of data with limited experimental effort.
Table 3.2 presents diffusion coefficients that were determined simultaneously with
partition coefficients and, in some cases, initial concentrations. It shows that measured
D
e
(or D
app
) values are of the order of 10
6
to 10
7
m
2
.s
1
, while measured solid-phase
diffusivities vary in a wider range from 10
10
to 10
13
m
2
.s
1
. In the latter case, however,
distinction has to be made between two kinds of data. True solid-phase diffusivities
(i.e. diffusivities in materials such as vinyl flooring or carpet backing) lie in the range from
10
10
to 10
14
m
2
.s
1
, which is in agreement with typical diffusion coefficients in polymers
at ordinary temperatures for buildings (Treybal, 1981; Sun et al, 1997). On the other hand,
many data have been determined based on the assumption of solid diffusion when the
tested materials are likely to be porous, and would thus be better characterized by
ABER_CH_03 6/12/08 3:31 PM Page 76
pore-phase diffusivities. If such data are to be input in two-phase models, comparing
Equations 3 and 8 and further assuming a linear adsorption isotherm (with K
p
K
H
K
p
in
units of g
air
/g
mat
being the measured partition coefficient) show that measured D
mat
values
can be converted to D
p
values from the relation:
[38]
Given that
mat
is expected to be two to three orders of magnitude higher than
air
, D
mat
measured for porous materials is expected to be one order of magnitude lower than
D
p
if K
p
< 10
3
, and several orders of magnitude lower otherwise. Most of the data where
K
p
and D
mat
values were determined simultaneously are consistent with this (see Table 3.2).
INITIAL CONCENTRATIONS
Initial concentrations are a key parameter for the modelling of VOC emissions from
materials. However, knowing the initial VOC content of building materials faces the
problem of industrial secrecy, and no information is to be expected from manufacturers.
Besides, this information is usually hard to get experimentally. Methods such as the
solvent extraction or high thermal desorption (Tichenor, 1996) have been widely used but
carry the risk that the chemical materials are transformed during operation. For dry
materials, an interesting alternative has recently been proposed by Cox et al (2001a), who
developed a method based on cryogenic grinding and fluidized bed desorption to
measure the concentrations of volatile organic compounds in vinyl flooring. Finally, a third
way to access the initial concentrations is first to integrate, over time, the emission
profiles measured in environmental chambers, and then to divide the estimated total
mass emitted by the material mass or volume (Little et al, 1994; Elkilani et al, 2003). This
method has the advantage that many emission data exist and thus can provide a
substantial amount of information within a reasonable time period. However, it may lead
to a great underestimation of the initial concentrations if significant amounts of VOC
remain in the material at the end of the experiment.
Regarding the modelling issue, it is important to consider that the measured VOC
contents, C
0
tot
(kg
species
/m
3
mat
), can directly be used as the boundary condition of one-phase
models. On the other hand, the parameter to input in two-phase emission models is the
initial concentration in the air phase of the pores, C
0
(kg/kg
air
). The latter can be determined
by considering that the total mass of VOC contained in the material is the lump sum of
the VOC contained in the gas phase and in the sorbed phase that is to say:
[39]
where C
0
s
(kg
species
/kg) is the initial sorbed phase concentration, m
mat
(kg) the mass of the
material sample, and m
air
(kg) the mass of air in the material sample. Since, the last two
parameters are correlated through the relation:
[40]
m
m
air air
mat
mat
=
,
m C m C m C
mat tot air mat s
0 0 0
= ,
D
K
D
p
mat
air
mat
=
\
)
.
ABER_CH_03 6/12/08 3:31 PM Page 77
substituting Equation 39 into equation 40 and rearranging gives:
[41]
Finally, if assuming a linear isotherm, we have:
[42]
CORRELATION WITH PHYSICAL PROPERTIES
In spite of research efforts, there is evidence that there is insufficient data on the sorption
and diffusion properties of contaminants in building materials to model the complexity of
real indoor settings. The proposal of standards, or at least validated methods, to get such
data would definitely be a first step in progressing towards informative databases.
However, experimental methods find their limitations in the huge number of gases and
materials encountered in buildings, which results in a prohibitive number of systems to be
tested. Identical materials, furthermore, may exhibit different properties depending upon
the manufacturer and manufacturing process. For instance, Blondeau et al (2003) noted
that even slight differences in the porous structure of gypsum board samples would result
in differences in the contaminant diffusivities from one sample to another. Recently, Seo
et al (2005) carried out porosity tests on a commercial gypsum board and reported a
porosity as high as 68 per cent that is to say, more than twice the porosity measured by
Blondeau et al (2003)!
In this context, several authors have searched for correlations between the
experimental parameters of interest and the gas properties, such as van Der Waals molar
volume, mean diameter of the molecule, saturation vapour pressure/boiling point, or
molecular weight (Berens and Hopfenberg, 1982; An et al, 1999; Bouhamra and Elkilani,
1999; Bodalal et al, 2000; Cox et al, 2001b; Meininghaus and Uhde, 2002; Elkilani et al,
2003, He et al, 2005). The general trends are that VOC diffusion coefficients often
decrease as the molecular weight of the compound increases, and sorption constants
generally increase as the saturation vapour pressure of the compound decreases (Little
and Hodgson, 1996).
These conclusions are somewhat consistent with the principles of diffusion in
polymers: species having a high molecular weight are generally large molecules; they will
diffuse slowly through the polymer chains. On the other hand, the relationship between
diffusivity and molecular weight is harder to establish analytically for porous materials.
Three cases are to be distinguished: if Knudsen diffusion is the controlling mechanism
within the pores of the material, Equation 25 indicates that the diffusion coefficient is
expected to vary as a function of M
1/2
. If molecular diffusion prevails, the Chapman-Enskog
equation (Treybal, 1981), which is often used to assess the molecular diffusivity of gases
in air, relates D
m
to M in a non-explicit way. It would, nevertheless, show that molecular
diffusivities also decrease as the molecular weight of the species increase. Finally, if pore
phase diffusion is negligible compared to surface diffusion, which has been
C
C
K
tot
air
mat
p
0
0
=

C C C
tot
air
mat
s
0 0 0
=

ABER_CH_03 6/12/08 3:31 PM Page 78
demonstrated to be the case in some circumstances, Equation 35 indicates that
diffusivities are defined by the partition coefficient of the diffusing system. Therefore, D
p
will be expected to vary the same way as K
p
that is to say, to increase with molecular
weight.
Regarding sorption, some authors found a linear increase in K
H
with the reciprocal of
saturation vapour pressure (An et al, 1999; Elkilani et al, 2003), while, in a similar way,
others concluded that the logarithm of K
H
linearly correlates with the logarithm of
saturation vapour pressure (Cox et al, 2001b; Won et al, 2001; Weschler, 2002). Practically,
these correlations are of great interest in the sense that they provide a convenient means
to extrapolate the existing data to other compounds. Since vapour pressure depends
upon temperature, they may also serve to estimate K
H
at temperatures other than the one
set experimentally (see section Influence of temperature). Care should, nevertheless, be
taken when extrapolating the data since the correlations do not account for molecular
structure or functionality, and thus only apply for compounds belonging to the same
chemical group; in other words, when drawing K
H
versus vapour pressure for different
VOCs
,
the plots will not lie on the same unique straight line (Elkilani et al, 2003).
Furthermore, the linear relationships may prove to depend upon temperature. As an
example, Elkilani et al (2003) noted that K
H
in carpet fibres steadily increases with the
boiling point of the species (and decreases with their vapour pressure) at 25C, but levels
off in the boiling point range of 100C to 200C at temperatures up to 30C.
MODELLING THE INFLUENCE OF ENVIRONMENTAL PARAMETERS
The influence of indoor temperature, relative humidity and, to a lesser extent, airflow and
gas mixtures on both the VOC emissions and sink effect has been widely investigated by
experimental means during the past 20 years (Haghighat and De Bellis, 1998; Wolkoff,
1998; Fang et al, 1999). Although useful in understanding and interpreting the phenomena,
these experiments, nevertheless, cannot fit real configurations of building operation where
hundreds of contaminants are present in the air, on the one hand, and environmental
parameters are constantly varying over time, on the other. This section discusses the way
in which physically based models could be extended to address this issue.
INFLUENCE OF TEMPERATURE
Whether they have been carried out in real buildings or in small-scale chambers, most of
the experiments dealing with the influence of temperature on gas and material
interactions conclude on a lower adsorption of building materials or an increased
emission at higher temperatures. For instance, Seifert et al (1989) found that the emission
of 4-phenyl cyclohexene from carpets increased fivefold when increasing the temperature
from 25C to 50C. More recently, when studying the influence of temperature on the
emission rate of VOC from two carpets, a PVC flooring and an acrylic paint, Van Der Wal
et al (1997) concluded that the general trend was a faster emission with increasing
temperature. They also noted that, except for paint, the total emitted mass was higher at
50C than at 23C.
Numerically, the influence of temperature on the sink effect and VOC emission can be
modelled by considering the way in which diffusivities and sorption properties vary with
ABER_CH_03 6/12/08 3:31 PM Page 79
temperature. In both cases, a distinction has once more to be made between polymeric
and porous materials.
Influence of temperature on diffusivities
Diffusion coefficients in polymers at different temperatures are often found to be well
predicted by:
[43]
where D
A0
(m
2
.s
1
) is a constant and E
a
(J.mol
1
) is the energy of activation for diffusion.
The influence of temperature on pore diffusivity is somewhat more difficult to assess
since the latter may depend upon the sorption properties of the diffusing system, which
are themselves strongly influenced by temperature (see below). If sorbed phase
concentrations are low enough so that surface diffusion can be neglected (D
p
D
e
), pore
diffusivities will diminish to effective diffusivities. The latter have the following
temperature-dependent form (Do, 1998):
[44]
where D
e0
(m
2
.s
1
) is the effective diffusivity at some reference temperature T
0
(K). By
reference to Equation 25 for Knudsen diffusivity, and the Chapman-Enskog equation for
molecular diffusivity, will actually have the value of 0.5 if Knudsen diffusion is the
controlling mechanism within the pores of the material, and about 1.75 if molecular
diffusion prevails. will range from 0.5 to 1.75 in the intermediate region where both
processes control the pore diffusion.
If surface diffusion in the porous medium is to be considered, the influence of
temperature will be stronger as surface diffusivities also follow an Arrhenius-type law for
temperature dependency:
[45]
with D
s
(m
2
.s
1
) being the surface diffusivity at infinite temperature and E
s
(J.mol
-1
) the
activation energy for surface diffusion.
Practically, Equations 43 to 45 indicate that within typical temperature ranges for
buildings, the influence of temperature on gas diffusivities will be significant or almost
negligible depending upon the materials and the concentration levels that are put into
play. For instance, based on the results of Kumar and Siddaramaiah (2005) and other
references, the activation energy for VOC diffusion in polymers at ordinary temperatures
may be thought to range from 20 to 30 kJ/mol. Taking these values for E
a
in Equation 43
gives that D
A
will be increased about twofold when the temperature is increased from
20C to 35C. On the other hand, Equation 44 indicates that the corresponding increase
in D
e
would not exceed 10 per cent in the case of a porous material operating at moderate
sorbed phase concentrations.
D D e
s s
E
s
RT
=

\
)
,
D D
T
T
e e
=

(
\
)
0
0
,
D D e
A A
E
a
RT
=

0
,
ABER_CH_03 6/12/08 3:31 PM Page 80
After Do (1998), the temperature dependency of the partition coefficient K
p
follows an
Arrhenius-type law:
[46]
with K
p
being the partition coefficient at infinite temperature, H
ads
(J.mol
1
) the heat of
adsorption, R (J.mol
1
.K
1
) the universal gas constant and T (K) the absolute temperature.
On the other hand, the temperature dependency of the absorption constant in polymers,
K
H
, follows a Vant Hoff equation, which by integration yields:
[47]
Here the pre-exponential factor K
H
is the absorption constant at infinite temperature and
H
abs
(J.mol
-1
), the activation energy for absorption (or heat of absorption), is the
difference between the internal energies of the adsorbate in the absorbed and gaseous
states.
In the end, one can note that Equations 46 and 47 have similar shapes. Furthermore,
since H
ads
and H
abs
have negative values, they both indicate that K
H
and K
p
decrease with
increasing temperatures, and the higher the heat of sorption, the greater the temperature
sensitivity of the sorption parameters. By way of illustration, Elkilani et al (2003) found H
abs
absolute values of 36.6, 56.9 and 20.3kJ.mol
1
for absorption of toluene, 1-2-dichlorobenzene
and 1,1,1-trichloroethane in carpet fibres, respectively. As shown in Table 3.2, it results in
a 41, 53 and 28 per cent decrease in the absorption coefficient when the temperature is
increased from 25C to 35C. Tiffonnet et al (2002) found a lower heat of absorption
of 7.5 kJ/mol for acetone in acrylic paint, reflecting slighter changes in K
H
with
temperature (see Table 3.2). Finally, still based on the results of Tiffonnet et al (2002), the
acetone heat of adsorption in particle board would have a value of about 30 kJ/mol that
is to say, the same order of magnitude as typical heats of absorption in polymers.
However, there is too little data to draw definitive conclusions on the compared
temperature sensitivity of K
p
and K
H
.
If contaminant and material interaction models were to be developed in conjunction
with heat-transfer models, the influence of temperature on mass transports within porous
materials could be automatically accounted for by considering that C
s
is a function of both
the species air-phase concentration and temperature. Therefore, Equation 8 would
translate to:
[48]
Then, assuming a linear isotherm and substituting Equation 46 into Equation 48 would
yield:
[49] D C K e
p
mat
air
p
H
ads
RT
=
\
)
\
)
\
)
C
t
K
H
RT
e
T
t
mat
air
p
ads
H
ads
RT

2
D C
C
t
C
C
C
t
C
T
T
t
p
mat
air
s s
=

\
)
K K e
H H
H
abs
RT
=

\
)
K K e
p p
H
ads
RT
=

\
)
,
ABER_CH_03 6/12/08 3:31 PM Page 81
INFLUENCE OF CONTAMINANT MIXTURES
The few experiments that have addressed the problem of multi-component sorption in
building materials lead to different conclusions. Cox et al (2001b) examined the effect of
simultaneous sorption of two VOCs on vinyl flooring by exposing a vinyl flooring sample
to a gas stream containing phenol at 28,000g.m
3
and n-dodecane at 33,000g.m
3
.
Although the tested concentrations were much higher than those typically found
indoors, the sorption of one VOC was found to be unaffected by the simultaneous
sorption of the other compound. On the other hand, Huang et al (2006) recently showed
from a statistical analysis that the uptake of toluene by a ceiling tile as a single
compound was higher than as part of a mixture. Finally, Jrgensen and Bjrseth (1999)
concluded from dynamic adsorption tests that the sorption of -pinene and toluene on
wool and nylon carpets was enhanced when both compounds were present in the
chamber. It seems that the nature of the VOC interactions in materials depends upon the
type of material, the type of compound and probably the concentrations that are put into
play. Moreover, while the results of the experiments have been analysed solely in terms
of potential changes in the sorption properties of the gas and material systems
investigated, and will be considered that way here, the observed discrepancies might
actually result from complex reactivity phenomena at the material surfaces and should
be modelled accordingly (see the section on Modelling chemical transformations within
porous materials).
Regarding polymeric materials, Schwope et al (1989) indicated that VOCs in pure
polymers generally behave ideally that is, follow Henrys law if the concentration of
VOCs in the material phase is lower than 1 per cent by weight. Practically, this means that
except in the case of heavily polluted environments or emissive materials containing large
amounts of residual VOCs, linear models would still apply for multi-component absorption
systems. Moreover, the absorption constants measured for single compounds would be
unaffected by the presence of other gases in the air.
In the case of porous media, a variety of theories and related multi-component
adsorption isotherm models have been proposed. To date, their efficacy has only been
investigated for industrial separation applications and for air cleaning applications in
industrial applications (Wood, 2002). In general, multi-component adsorption isotherm
models state that under isothermal conditions at atmospheric pressure, the equilibrium
sorbed phase concentration of a given component i, C
si
, will depend upon not only the air
phase concentration of that component, but the air phase concentrations of all
compounds in the mixture. Whatever the model, the computed sorbed-phase
concentrations will always be lower than those predicted from the individual adsorption
isotherms of the compounds, thus reflecting the competition between species for a
limited number of active sites at the material surface. As a result, experiments showing
increased adsorption capacities for contaminant mixtures would tend to suggest that
phenomena other than only competition for sorption occur within the material
(e.g. absorption in condensed gases and surface reactions).
Still based on the premise that contaminant concentrations are usually low indoors,
the extended Langmuir equation for component molecules that do not interact at the
material surface is likely to be the simplest and most suitable multi-component adsorption
ABER_CH_03 6/12/08 3:31 PM Page 82
model for building applications. The latter relates the sorbed phase concentration of a
given species i to the air phase concentration of all other compounds j in the mixture as:
[50]
where the model parameters are defined for each component as outlined in Table 3.1.
Consequently, one practical constraint of using the extended Langmuir equation is that
both C
s0
and K
L
must be determined for all compounds when only K
p
values are available
in many situations.
It is important to note here that for very low air-phase concentrations and/or a small
number of compounds in the mixture, will be much smaller than 1 and Equation
50 simplifies to a series of n uncoupled single-component linear isotherms. Therefore, in
this case, multi-component competition can be ignored. On the other hand, if
does not prove to be negligible compared to 1, Equation 50 will contribute to couple the
equations describing the contaminant transports within porous materials. For instance, if
considering a mixture of two gases in air under constant temperature, Equation 8 for the
first component becomes:
[51]
Then, substituting the partial derivatives of Equation 50 into Equation 51 gives:
[52]
The equation for the second component is obtained by simply interchanging the
subscripts 1 and 2 in Equation 52.
Another multi-component adsorption isotherm model that may be interesting for
indoor air applications is the ideal adsorbed solution (IAS) theory of Myers and Prausnitz
(1965). Although it has been developed for applications where porous adsorption occurs
in the form of capillary condensation, which is unlikely for most building and pollutant
systems, applying this theory with Langmuir equations as individual adsorption isotherm
models yields the Langmuir extended equation. Besides, if considering linear adsorption
isotherm models for individual compounds, the method will return the same models for
the mixture of compounds, which reflects the fact that concentrations are low enough to
ignore the competition for sorption. All of these results are consistent and tend to prove
D C
C K
K C K C
C K C
K C
p
mat
air
s L
L L
s L
L
1
2
1
1
0
1
1 1 2 2
1
0
1
2
1
1 1
1
1
=

( KK C
C
t
C K K C
L
mat
air
s L L
2 2
2
1
1
0
1 2 1
)
\
)
l
l
l
l
l
(( ) 1
1 1 2 2
2
2

K C K C
C
t
L L
D C
C
t
C
C
C
t
C
C
C
t
p
mat
air
s s
1
2
1
1 1
1
1 1
2
2
=

\
)
K C
Lj j
j
n
=
1
K C
Lj j
j
n
=
1
C
C K C
K C
si
si Li i
Lj j
j
n
=
0
1
1
,
ABER_CH_03 6/12/08 3:31 PM Page 83
that the IAS theory is probably more general than one usually thinks. Compared to the
extended Langmuir equation, it has the advantage that different individual adsorption
isotherm models can be used as inputs of the method; this may prove to be useful when
dealing with compounds, such as water vapour, with isotherms that exhibit such strong
non-linearities that they cannot be approximated by linear or even Langmuir models. On
the other hand, the IAS is numerically much more difficult to implement than the extended
Langmuir equation.
To a lesser extent, the presence of contaminant mixtures in the air may also influence
the diffusion of each individual species within materials. Although little is known on this
topic, the fundamentals of diffusion presented in the section on Diffusion coefficients
point out the fact that the question of multi-component diffusion in porous materials
cannot be treated independently from the question of the dominant diffusion mechanism:
the effect of simultaneous diffusion of dilute compounds in the air can probably be
ignored if only pore air-phase diffusion is to be considered. On the other hand, it may
really become significant through the above-mentioned competition for adsorption if
surface diffusion prevails. In such a case, the diffusion coefficients are expected to be
lower than those for single species.
INFLUENCE OF HUMIDITY
The experiments dealing with the influence of humidity on the sink (Jrgensen et al, 1999;
Won et al, 2000, 2001; Zhang and Niu, 2004) and source (Haghighat and De Bellis, 1998;
Wolkoff, 1998; Fang et al, 1999) behaviour of materials have led to different and sometimes
divergent conclusions regarding the influence of humidity. A meaningful example is
provided by Haghighat and De Bellis (1998): when measuring the VOC emissions from a
water-based acrylic paint and a polyurethane plastic finish varnish, the authors came to
the conclusion that, during the first stage of the emission process, the total amounts of
VOC (TVOC) emitted from the paint were higher at 32 per cent relative humidity (RH) than
at 62 per cent RH. On the other hand, the TVOC emissions from the varnish proved to
increase with increasing humidity. Nevertheless, in this latter case, the authors also
noticed that unlike the TVOC, individual compounds such as toluene, m,p-xylene
and 1,3,5-trimethylbenzene showed higher emission rates at 32 per cent RH than at
62 per cent RH.
Of all the studies quoted above, few have addressed the question of how humidity
physically interacts with contaminants in materials. Recently, Radulescu-Bouilly et al
(2006) presented a theory that may serve as a first step in understanding and modelling
these complex interactions for porous materials. First, this theory is based on the principle
that the transport of water vapour in building materials can be modelled analogously to
the transport of any gas, except that a distinction is made between monolayer and
multilayer adsorption, on the one hand, and capillary condensation, on the other hand.
Next, three elemental interaction mechanisms are considered (see Figure 3.1). First is the
competition for adsorption at the surfaces of the largest pores, the latter defined as those
where capillary condensation does not occur; the authors suggest that it should be
modelled using the extended Langmuir equation: Equation 50. The second and third
interaction processes come from the possible capillary condensation of water vapour in
ABER_CH_03 6/12/08 3:31 PM Page 84
the thinnest pores of the material. As gases are soluble in liquids to a greater or lesser
degree (Sander, 1999), the pores that are filled with liquid water will give rise to
absorption/desorption mass transports inside the material, with the result of increasing its
mass capacity. In addition, the pores that are filled with liquid water no longer contribute,
or contribute differently, to the porous diffusion of gases in the material. Therefore, the
diffusivity of VOCs may be decreased significantly when humidity increases. Using, first,
published data for the pore-size distribution of some materials, on the one hand, and,
second, the Laplace equation to determine the critical pore radius below which the pores
will be filled with liquid water, and the resulting water content of the materials, and, third,
Henrys law to represent the absorption/desorption of gases in water, Radulescu-Bouilly
et al (2006) show that increasing the air humidity can contribute to change from one
dominant interaction phenomenon to another. This may explain the non-linearities in the
observed sink or source behaviour at various humidity levels, as well as the differences in
the influence of humidity depending upon the contaminant and material systems
investigated.
With regard to polymeric materials, Cox et al (2001b) suggested various possible
mechanisms to explain the observed changes in the adsorption capacity of vinyl flooring
in humid conditions. Some relate to direct interactions between sorbed VOC and water
vapour within the material, while others suggest that sorbed water molecules change the
Absorption/desorption
in condensed water
Competition for
sorption at the pore
surfaces
Capillary
condensation
of water in the
thinnest pores
Source: adapted from Radulescu-Bouilly et al (2006)
FIGURE 3.1 Elemental interaction phenomena between gaseous contaminants and water vapour in
poly-dispersed porous materials
ABER_CH_03 6/12/08 3:31 PM Page 85
structural properties of the material and, subsequently, the sorption properties of the VOC
and vinyl flooring system. In all cases, these processes nevertheless remain very difficult
to represent analytically.
MODELLING CHEMICAL TRANSFORMATIONS
WITHIN POROUS MATERIALS
During the past 20 years, many studies have emphasized the potentially strong influence
of homogeneous chemistry on the concentration levels of some hazardous contaminants
(Nazaroff and Cass, 1986; Weschler and Shields, 1997; Blondeau et al, 1998, Sarwar et al,
2002). All have focused on chemical reactions occurring in the bulk air phase of the
rooms. However, the kinetic coefficients reported for those reactions, which are most of
the time directly transposed from atmospheric chemistry mechanisms, show that
chemical reactions are likely to proceed much faster than diffusion in materials.
Therefore, there is evidence that organic or non-organic species may also be
transformed through homogeneous processes when transported in the pores of porous
building materials.
Similarly, surface chemical transformations (also called heterogeneous processes or
chemisorption, in contrast to the reversible physical adsorption considered previously) are
likely to be important for a large number of species, including nitrogen oxides and their
radicals; oxygen; ozone and oxygen radicals; sulphur dioxide and sulphate ions; organic
and inorganic acids; and a large number of volatile organic compounds (Axley, 1991). They
may be considered even more important for porous materials where connected pores
offer huge surface areas for surface reactions.
In the field of indoor air analysis, surface chemistry has most often been modelled
using the highly simplified deposition velocity approach (Nazaroff and Cass, 1989). The
latter is based on the assumption that the removal rates of contaminants, r (kg.s
1
), are
related directly to the bulk air-phase species concentration, C:
[53]
where V
d
(m.s
1
) is the deposition velocity of the compound and A
s
the nominal surface
area of the material (i.e. surface area exposed to the room air). From the computational
point of view, the deposition velocity model has the great advantage that r can be included
in the equations describing the contaminant mass conservation in the bulk air phase of
the room: no equations and unknowns need to be added in the problem formulation to
account for the contribution of surface reactions. Furthermore, the model is linear and
thus does not call for advanced solving methods. Physically, the use of the deposition
velocity model is much more questionable. First, this model does not account for
boundary layer resistance: the removal rate should be defined as a function of the
contaminant near-surface concentration, rather than its concentration in the bulk air.
Practically, this means that the deposition velocity of a given compound and material
system will also depend upon the characteristics of the airflow at the material surface,
which is not fully satisfactory from a physical point of view. Another drawback of the
model is that it can only handle unimolecular decomposition reactions although other
r V A C
air d s
= ,
ABER_CH_03 6/12/08 3:31 PM Page 86
reactivity processes may exist (see below); moreover, it does not consider the yield of
product species that may be important from an indoor air quality point of view. Finally, the
deposition velocity model is not suitable for porous materials where, as mentioned above,
the surface area available for surface chemistry is much higher than the material surface
area that is exposed to the room air.
Considering that homogeneous and heterogeneous reactions will occur in parallel in
porous materials, and will subsequently be constrained by the diffusion and sorption
processes that determine the amounts of reagents locally available, Axley (1991) suggested
that the two phenomena may be modelled analogously by extending Equation 8 to:
[54]
where r, here in units of kg.m
3
.s
1
, is the local removal or yield rate of the compound
due to any kind of transformation within the material. To be more precise, r will actually
be the sum of the contributions of each homogeneous or surface reaction j involving the
compound as a reagent or as a product:
[55]
Based on the knowledge and theories developed in the field of atmospheric chemistry, r
air,j
(kg.m
3
.s
1
) will be defined as (Seinfeld, 1986):
[56]
where x is the stoichiometric coefficient of the compound in reaction j, k
j
(s
1
) the kinetic
coefficient of the reaction, n
j
the number of reagents involved in reaction j (i.e. 1, 2 or 3
for unimolecular, bimolecular or termolecular reactions, respectively) and C
k
(kg/kg
air
)
represents the local pore air-phase concentration of these reagents. Finally, r
air,j
is negative
if the compound is involved as a reagent in the reaction, and positive if acting as a reaction
product.
In the fields of catalysis and surface science, surface reactions are frequently
described as unimolecular decomposition or bimolecular combination reactions, although
more complex interactions may exist (Masel, 1996). By analogy with the modelling of
homogeneous reactions, the local rate of removal of reagents or yield of product species
resulting from reaction j is in both cases given by the kinetic coefficient of the reaction, k
j
(in units of s
1
for a unimolecular reaction and s
1
.kg
1
for a bimolecular reaction), multiplied
by the stoichiometric coefficient of the compound and the mass(es) of reagent(s)
adsorbed that is:
.
[57]
As for homogeneous reactions, r
surf,j
will be negative if the compound is involved as a
reagent in reaction j, and positive otherwise. For a bimolecular reaction involving species
r xk C
surf j mat j sk
k
n
j
,
=
=
1
r xk C
air j air j k
k
n
j
,
, =
=
1
r r r
air j surf j
j j
=
, ,
.
D C r
f
C
C
t
p
air
mat
air
=

\
)
2
1
,
ABER_CH_03 6/12/08 3:31 PM Page 87
A and B as reagents, substituting Equation 50 into Equation 57 then leads to the so-called
Langmuir-Hinshelwood model:
[58]
To date, the implementation of surface chemistry models is of limited interest since not
one relevant surface reaction on building materials has been elucidated to the point that
the mechanism is understood and the kinetic constant is known. Only Spicer et al (1986)
suggested a catalytic conversion of nitrogen dioxide to nitric oxide on some surfaces and
Ryan and Koutrakis (1993) indicated that the decomposition of ozone on latex paints may
be governed by a bimolecular reaction with sorbed water. Experiments are needed in
order to gain more knowledge and to support research in this area. Then, considering that
most reagent species will follow a Langmuir adsorption isotherm model in porous
building materials, and that Langmuir-Hinshelwood type rate equations often work under
conditions where the adsorption of the reagents does not follow a Langmuir adsorption
isotherm (Masel, 1996), Equation 58 definitely comes in as a good candidate for numerical
developments in this field.
In the end, it is important to note that from a numerical point of view, the
implementation of homogeneous and/or heterogeneous chemistry models in the equations
describing the contaminant transport in building materials will lead to systems of non-linear
and coupled equations. Moreover, not only the contaminant of interest for indoor
air-quality analysis, but all the species involved as reagents in the transformations, must
be considered in the model. This may result in large systems to be solved and, possibly,
prohibitive calculation times.
CONCLUSION
Although potentially having a great influence on predicted contaminant concentrations,
modelling homogeneous and heterogeneous processes that occur indoors still remains
the drawback of most indoor air-quality simulation tools. The review of existing
gas/material interaction models presented here shows that research in this field has
suffered confusion regarding which physical phenomena to consider and the way in
which to model data. Frequently, misinterpretation occurs between sorption dynamics
and sorption equilibrium. From the knowledge developed in the related fields of surface
science and catalysis, it seems obvious that the former results from the dynamics of
diffusion within the materials. Therefore, elemental models that couple local equilibrium
sorption models with diffusion equations are most probably a more realistic fit than what
has been referenced as macroscopic models.
Another confusing problem comes from the nature of the sorption phenomena
involved depending upon the materials considered. In most studies, the contaminant
transfers from the bulk air to the material (and vice versa) are referred to as adsorption/
desorption, and are modelled accordingly, when they actually consist of absorption/
desorption processes in the case of non-porous polymeric materials. Practically, this is of
no consequence since Henrys law for ideal absorption in solids or liquids is analogous to
a linear adsorption isotherm model. However, from a general viewpoint, it underlines the
r
xk C K C C K C
K C K C
surf j
mat j sA LA A sB LB B
LA A LB B
,
( )
. =

0 0
2
1
ABER_CH_03 6/12/08 3:31 PM Page 88
central importance of the porous features of building materials for modelling purposes:
knowing whether the material is porous or not determines the kind of mass-transport
model to be used. Furthermore, having information on the pore-size distribution of porous
materials may help one to assess the contaminant diffusivities in the air phase of the
pores and, subsequently, to determine whether surface diffusion is to be considered or
not in the model.
Without going into philosophical considerations, physically based models somewhat
imply universality or, at least, generality. As a result, this chapter placed particular
emphasis on demonstrating that, depending upon the initial conditions and the
parameters set, elemental models can either represent the emission or the sink effect of
any material, predict the transition from one phenomenon to another, and possibly
account for the decomposition of reactive species (what is referred to as deposition) by
defining surface reactions. To date, very little is known about the latter point. Developing
knowledge on this topic and, especially, combination reactions (e.g. oxidation, reduction
or hydrolysis of contaminants at the material surfaces) comprise a real challenge for the
next years. Implementing multi-component facilities and coupling the heat and mass
transfer equations are undoubtedly other steps to overcome before accurately modelling
the variety and great complexity of all indoor gas and material interactions.
P. Blondeau: (corresponding author) LEPTAB, University of La Rochelle, France; patrice.blondeau@univ-lr.fr
A. L. Tiffonnet: LUSAC, University of Caen, France
F. Allard: LEPTAB, University of La Rochelle, France
F. Haghighat: Centre for Building Studies, Concordia University, Montreal, Canada
REFERENCES
An, Y., J. S. Zhang and C. Y Shaw (1999) Measurements of VOC adsorption/desorption characteristics of typical interior
building materials, HVAC Research, vol 5, pp297316
Ataka, Y., S. Kato, S. Murakami, Q. Zhu, K. Ito and T. Yokota (2004) Study of effect of adsorptive building material on
formaldehyde concentrations: Development of measuring methods and modeling of adsorption phenomena, Indoor Air,
vol 14, pp5164
Axley, J. W. (1991) Adsorption modelling for building contaminant dispersal analysis, Indoor Air, vol 2, pp147171
Axley, J. W. (1993) Modeling sorption transport in rooms and sorption filtration systems for building air quality analysis,
Indoor Air, vol 3, pp298309
Bejan, A. (1995) Convection Heat Transfer, 2nd edition, Wiley and Sons, New York
Berens, A. R. and H. B. Hopfenberg (1982) Diffusion of organic vapors at low concentrations in glassy PVC, polystyrene and
PMMA, Journal of Membrane Science, vol 10, pp283303
Blondeau, P., F. Allard, M. Sprandio and F. Haghighat (1998) Detailed modeling of gas-phase chemistry mechanisms in IAQ
simulation, ASHRAE Transactions, vol 104, no 2, pp13091317
Blondeau, P., A. L. Tiffonnet, A. Damian, O. Amiri and J. L. Molina (2003) Assessment of contaminant diffusivities in building
materials from porosimetry tests, Indoor Air, vol 13, pp310318
ABER_CH_03 6/12/08 3:31 PM Page 89
Bodalal, A., J. S. Zhang and E. G. Plett (2000) A method for measuring internal diffusion and equilibrium partition coefficients
of volatile organic compounds for building materials, Building and Environment, vol 35, pp101110
Bouhamra, W. S. and A. S. Elkilani (1999) Investigation and modeling of surface sorption/desorption behaviour of volatile
organic compounds for indoor air quality analysis, Environmental Science and Technology, vol 20, pp531545
Chang, C. S. and Z. Guo (1992) Characterization of organic emissions from a wood finishing product wood stain, Indoor Air,
vol 2, pp146153
Colombo, A. and M. De Bortoli (1992) Comparison of models used to estimate parameters of organic emissions from
materials tested in small environmental chambers, Indoor Air, vol 2, pp4957
Colombo, A., M. De Bortoli, H. Knppel, E. Pecchio and H. Vissers (1993) Adsorption of selected volatile organic compounds
on a carpet, a wall coating, and a gypsum board in a test chamber, Indoor Air, vol 3, pp276282
Cox, S. S., A. T. Hodgson and J. C. Little (2001a) Measuring concentrations of volatile organic compounds in vinyl flooring,
Journal of the Air and Waste Management Association, vol 51, no 8, pp174185
Cox, S. S., D. Zhao and J. C. Little (2001b) Measuring partition and diffusion coefficients for volatile organic compounds in
vinyl flooring, Atmospheric Environment, vol 35, pp38233830
Damian, A., P. Blondeau and F. Allard (2003) Distribution of room air contaminant concentrations as a function of ventilation
and air cooling, Proceedings of the 7th International Conference Healthy Building 2003, Singapore, vol 2, pp410415
Do, D. D. (1998) Adsorption Analysis: Equilibria and Kinetics, Imperial College Press, London
Dunn, J. E. (1987) Models and statistical methods for gaseous emission testing of finite source in well-mixed chambers,
Atmospheric Environment, vol 21, pp425430
Dunn, J. E. and B. Tichenor (1988) Compensating for sink effects in emissions test chambers by mathematical modeling,
Elkilani, A. S., C. G. J. Baker, Q. H. Al-Shammari and W. S. Bouhamra (2003) Sorption of volatile organic compounds on
typical carpet fibers, Environment International, vol 29, pp575585
EPA (US Environmental Protection Agency) (2006) EPA On-line Tools for Site Assessment Calculation Diffusion Coefficient
Estimation, www.epa.gov/athens/learn2model/part-two/onsite/estdiffusion.htm
Fang L., G. Clausen and O. Fanger (1999) Impact of temperature and humidity on chemical and sensory emission from
buildings materials, Indoor Air, vol 9, pp193301
Haghighat, F. and L. De Bellis (1998) Material emission rates: Literature review, and the impact of indoor air temperature and
relative humidity, Building and Environment, vol 33, pp261277
Haghighat, F. and Y. Zhang (1999) Modelling of emission of volatile organic compounds from building materials estimation
of gas-phase mass transfer coefficient, Building and Environment, vol 34, pp377389
Haghighat, F., C.-S. Lee and W. S. Ghaly (2002) Measurement of diffusion coefficients of VOCs for building materials: Review
and development of a calculation procedure, Indoor Air, vol 12, pp8191
Hansson, P. and H. Stymne (2000) VOC diffusion and absorption properties of indoor materials consequences for indoor air
quality, Proceedings of Healthy Building 2000, Espoo, Finland, vol 4, pp151156
He, G., X. Yang and C. Y. Shaw (2005) Material emission parameters obtained through regression, Indoor and Built
Environment, vol 14, pp5968
Huang, H., F. Haghighat and P. Blondeau (2006) Volatile organic compound (VOC) adsorption on material: influence of gas
phase concentration, relative humidity and VOC type, Indoor Air, vol 16, pp236247
Jrgensen, R. B. and O. Bjrseth (1999) Sorption behaviour of volatile organic compounds on material surfaces the
influence of combinations of compounds and materials compared to sorption of single compounds on single materials,
Environment International, vol 25, pp1727
Jrgensen, R. B., O. Bjrseth and B. Malvik (1999) Chamber testing of adsorption of volatile organic compounds (VOCs) on
material surfaces, Indoor Air, vol 9, pp29
ABER_CH_03 6/12/08 3:31 PM Page 90
Jrgensen, R. B., T. H. Dokka and O. Bjrseth (2000) Introduction of a sink-diffusion model to describe the interaction
between volatile organic compounds (VOCs) and material surfaces, Indoor Air, vol 10, pp2738
Kephalopoulos, S., H. Knoeppel and M. De Bortoli (1996) Testing of sorption and desorption of selected VOCs on/from indoor
materials, Proceedings of the 7th International Conference on Indoor Air Quality and Climate, Indoor Air 1996, Nagoya,
pp6166
Kirchner, S., J. R. Badey, H. N. Knudsen, R. Meininghaus, D. Quenard, K. Saarela, H. Sallee and A. Saarinen (1999) Sorption
capacities and diffusion coefficients of indoor surface materials exposed to VOCs: Proposal of new test procedures, in
Proceedings of the 8th International Conference on Indoor Air Quality and Climate, Indoor Air 1999, Edinburgh, vol 1,
pp430435
Kumar, H. and J. C. Little (2003) Characterizing the source/sink behavior of double-layer building materials, Atmospheric
Kumar, H. and Siddaramaiah (2005) A study of sorption/desorption and diffusion of substituted aromatic probe
molecules into semi interpenetrating polymer network of polyurethane/polymethyl methacrylate, Polymer, vol 46,
pp71407155
Lee, C. S., F. Haghighat and W. S. Ghaly (2005) A study on VOC source and sink behaviour in porous building materials
analytical model development and assessment, Indoor Air, vol 15, pp183196
Li, F. and J. Niu (2005) Simultaneous estimation of VOCs diffusion and partition coefficients in building materials via inverse
analysis, Building and Environment, vol 40, pp13661374
Li, F. and J. Niu (2007) Control of volatile organic compounds indoors development of an integrated mass-transfer-based
model and its application, Atmospheric Environment, vol 41, pp23442354
Little, J. C. and A. T. Hodgson (1996) A Strategy for Characterizing Homogeneous, Diffusion-Controlled, Indoor Sources and
Sinks, Standard Technical Publication 1287, American Society for Testing and Materials, pp294304
Little, J. C., A. T. Hodgson and A. J. Gadgil (1994) Modeling emissions of volatile organic compounds from new carpets,
Masel, R. I. (1996) Principles of Adsorption and Reaction on Solid Surfaces, Wiley and Sons, New York
Matthews, T. G., A. R. Hawthorne and C. V. Thompson (1987) Formaldehyde sorption and desorption characteristics of
gypsum wallboard, Environmental Science and Technology, vol 21, pp629634
Meininghaus, R. and E. Uhde (2002) Diffusion studies of VOC mixtures in a building material, Indoor Air, vol 12,
pp215222
Meininghaus, R., L. Gunnarsen and H. N. Knudsen (2000) Diffusion and sorption of volatile organic compounds in building
materials impact on indoor air quality, Environmental Science and Technology, vol 34, pp31013108
Morrison, G. C., Z. Ping, D. J. Wiseman, M. Ongwandee, H. Chang, J. Portman and S. Regmi (2003) Rapid measurement of
indoor mass-transfer coefficients, Atmospheric Environment, vol 37, pp56115619
Mulder, M. (1991) Basic Principles of Membrane Technology, Kluwer Academic Publishers, Dordrecht, The Netherlands
Murakami, S., S. Kato, K. Ito and Q. Zhu (2003) Modeling and CFD prediction for diffusion and adsorption within room with
various adsorption isotherms, Indoor Air, vol 13, pp2027
Myers, A. L. and J. M. Prausnitz (1965) Thermodynamics of mixed-gas adsorption, American Institute of Chemical Engineers
Journal, vol 11, pp121127
Nazaroff, W. W. and G. R. Cass (1986) Mathematical modeling of chemically reactive pollutants in indoor air, Environmental
Science and Technology, vol 20, pp924934
Nazaroff, W. W. and G. R. Cass (1989) Mass-transport aspects of pollutant removal at indoor surfaces, Environment
International, vol 15, pp567584
Park, I., D. D. Do and A. E. Rodrigues (1996) Measurement of the effective diffusivity in porous media by the diffusion cell
method, Catalysis Reviews: Science and Engineering, vol 38, pp189247
ABER_CH_03 6/12/08 3:31 PM Page 91
Radulescu-Bouilly, C., F. Allard, P. Blondeau, B. Collignan, R. Popescu and A. Sjberg (2006) A physically-based analysis of the
interactions between humidity and VOCs in building materials, Proceedings of the Healthy Building 2006 International
Conference, Lisboa, vol IV, pp8994
Rutherford, S. W., R. E. Kurtz, M. G. Smith, K. G. Honnell and J. E. Coons (2005) Measurement and correlation of
sorption and transport properties in polystyrene POSS membranes, Journal of Membrane Science, vol 263,
pp5765
Ruthven, D. M. (1984) Principles of Adsorption and Adsorption Processes, Wiley and Sons, New York
Ryan, P. B. and P. Koutrakis (1993) The reactive chemistry of ozone in indoor environments, CIAR Currents: The Newsletter
of the Center for Indoor Air Research, Linthicum, MD, pp13
Sander, R. (1999) Compilation of Henrys Law Constants for Inorganic and Organic Species of Potential Importance in
Environmental Chemistry (Version 3), www.henrys-law.org
Sarwar, G., R. Corsi, Y. Kimura, D. Allen and C. J. Weschler (2002) Hydroxyl radicals in indoor environments, Atmospheric
Schwope, A. D., W. J. Lyman and R. C. Reid (1989) Methods for Assessing Exposure to Chemical Substances, US EPA Report
(EPA 560/585015), vol 11
Seo, J., S. Kato, Y. Ataka and Q. Zhu (2005) Evaluation of effective diffusion coefficient in various building materials and
absorbents by mercury intrusion porosimetry, Proceedings of the 10th International Conference on Indoor Air Quality
and Climate, Indoor Air 2005, Beijing
Seifert, B., D. Ullrich and R. Nagel (1989) Volatile organic compounds from carpeting, in Proceedings of the 8th World Clean
Air Congress, The Hague, vol 1, pp253258
Seinfeld, J. H. (1986) Atmospheric Chemistry and Physics of Air Pollution, Wiley and Sons, New York
Silberstein, S. (1989) A Gypsum Wallboard Formaldehyde Sorption Model, National Institute for Standards and Technology
(NIST), Document no NISTDIR 89-4028, Gaithersbourg, US
Singer, B. C., K. L. Revzan, T. Hotchi, A. T. Hodgson and N. J. Brown (2004) Sorption of organic gases in a furnished room,
Spicer, C. W., R. W. Coutant and G. F. Ward (1986) Investigation of Nitrogen Dioxide Removal from Indoor Air, Gas Research
Institute, 1986, Document no GRI86/0303, Chicago, US
Sun, Y. M., C. L. Lin, Y. K. Chan, and C. H. Wu (1997) Sorption and diffusion of organic vapors in poly[bis(trifluoroethoxy)
phosphazene] and poly[bis(phenoxy)phosphazene] membranes, Journal of Membranes Sciences, vol 134,
pp117126
Tichenor, B. A., Z. Guo, J. E. Dunn, L. E. Sparks and M. A. Mason (1991) The interaction of vapour phase organic compounds
with indoor sinks, Indoor Air, vol 1, pp2335
Tichenor, B. A., Z. Guo and L. E. Sparks (1993) Fundamental mass transfer model for indoor air emissions from surface
coatings, Indoor Air, vol 3, pp263268
Tichenor, B. A. (1996) Overview of source/sink characterization methods, ASTM STP, vol 1287, pp919
Tiffonnet, A. L., P. Blondeau, F. Allard and F. Haghighat (2002) Sorption isotherms of acetone on various building materials,
Indoor and Built Environment, vol 11, pp95104
Topp, C., P. V. Nielsen and P. Heiselberg (2001) Influence of local airflow on the pollutant emission from indoor building
surfaces, Indoor Air, vol 11, pp162170
Treybal, R. E. (1981) Mass Transfer Operations, McGraw Hill International, Boston, US
Van Der Val, J. F., A. W. Hoogeveen and P. Wouda (1997) The influence of temperature on the emission of volatile organic
compounds from PVC flooring, carpet and paint, Indoor Air, vol 7, pp215221
Weschler, C. J. (2002) Connections: particles, sensory offending filters, the sink effect and nasal pungency thresholds,
Proceedings of Indoor Air 2002, Monterey, US
ABER_CH_03 6/12/08 3:31 PM Page 92
Weschler, C. J. and H. C. Shields (1997) Potential reactions among indoor pollutants, Atmospheric Environment, vol 31,
pp34873495
White, F. M. (1991) Heat and Mass Transfer, Addison-Wesley, New York
Wilkins, C. K., P. Wolkoff, P. A. Clausen, M. Hammer, S. T. Larsen and G. D. Nielsen (2006) Effects of repeated exposure to
limonene oxidation products in a mouse bioassay, Proceedings of the Healthy Building 2006 International Conference,
Lisboa, Portugal
Wolkoff, P. (1998) Impact of air velocity, temperature, humidity and air on long-term VOC emission from building products,
Won, D., R. L. Corsi and M. Rynes (2000) New indoor carpet as an adsorptive reservoir for volatile organic compounds,
Environmental Science and Technology, vol 34, pp41934198
Won, D., R. Corsi and M. Rynes (2001) Sorptive interactions between VOCs and indoor materials, Indoor Air, vol 11,
pp246256
Wood, G. O. (2002) Review and comparisons of D/R models of equilibrium adsorption of binary mixtures of organic vapors on
activated carbons, Carbon, vol 40, pp231239
Yang, X. and Q. Chen (2001) A coupled airflow and source/sink model for simulating indoor VOC exposures, Indoor Air,
vol 11, pp257269
Yang, X., Q. Chen, J. S. Zhang, Y. An, J. Zeng and C. Y. Shaw (2001) A mass transfer model for simulating VOC sorption on
building materials, Atmospheric Environment, vol 35, pp12911299
Zhang, L. Z. and J. L. Niu (2004) Modeling VOCs emissions in a room with a single-zone multi-component multi-layer
technique, Building and Environment, vol 39, pp523531
Zhang, Y. and Y. Xu (2003) Characteristics and correlations of VOC emissions from building materials, International Journal of
Heat and Mass Transfer, vol 46, pp48774883
ABER_CH_03 6/12/08 3:31 PM Page 93
ABER_CH_03 6/12/08 3:31 PM Page 94
Keywords urban microclimate; urban planning; architectural design
INTRODUCTION
The quality of life of millions of people living in cities can be improved if the factors
that affect the urban microclimate are understood and the form of the city responds to
them in a manner that is appropriate to its environment. Underlying this approach is
the idea that climatically responsive urban design is vital to any future notion of
sustainability: it enables individual buildings to make use of natural energy, it
enhances pedestrian comfort and activity in outdoor spaces, and it may encourage city
dwellers to moderate their dependence upon air conditioning in buildings and upon
private vehicles.
It has been known for well over a century that cities generate their own climate
(Howard, 1833). However, urban climate research has evolved as a specialist discipline
within meteorology and climatology only in the past 50 years or so. During this period, the
focus of research has shifted from descriptive studies of the properties of the urban wind
field, temperature and humidity, to an experimental approach designed to investigate the
physical processes responsible for the unique meteorological conditions found in cities.
Breaking down complex urban forms into basic components, notably the urban canyon,
has allowed researchers to isolate the effects of individual factors. Comprehensive
Abstract
In spite of the growing body of research on urban climatology and the increasing demand
for architects and urban planners to practise climate-conscious design, there is too little
evidence of the application of urban climatology in practice. This chapter explores the
relationship between climatologists and urban planners, seeking to establish some of
the reasons for this state of affairs. It then sets out a methodological framework for the
application of climatology in the planning process, outlining possible goals for such
intervention, as well as its limitations. The chapter then attempts to establish the effects
on the urban microclimate of a broad range of decisions taken routinely by architects
and planners, based on an extensive survey of applied research in the field.
4
The Application of Urban Climate
Research in the Design of Cities
Evyatar Erell
ABER_CH_04 6/12/08 3:32 PM Page 95
reviews of progress in research on urban climatology, in general, may be found in Arnfield
(2003) and Grimmond (2006).
Architecture (the design of buildings) and urban planning (the design of the urban
context within which individual buildings are constructed) have traditionally responded to
climatic conditions, as, indeed, they have to other aspects of the location of towns. The
importance of environmental factors in the design of settlements was recognized at least
as early as Roman times, as is recorded in the writings of Vitruvius (1999), who warned
planners, for example, that if the streets run straight in the direction of the winds then
their constant blasts rush in and sweep the streets with great violence. More recently,
there have been numerous manuals describing modern methods of bio-climatic
architecture, many of which owe much to the pioneering work of Olgyay (1963).
Two modern inventions have diverted the attention of architects and allowed them,
for a time, to ignore the effects of climate: first, the automobile, which replaced human
beings as the focus of attention in the design of outdoor space; and, second,
advanced space heating and cooling systems, which provided comfortable and stable
interiors regardless of the vagaries of climate, and allowed architects to avoid some of
the consequences of designing buildings that showed no regard to local conditions.
Climatic considerations have nonetheless had a major effect on the plans of a number
of new neighbourhoods or towns during recent years (Gotz, 1982; Etzion, 1990; Evans and
de Schiller, 19901991; El-Shakhs, 1994). In all of these cases, planning decisions were
made primarily on the basis of an intuitive understanding of local conditions, and less on
the basis of scientific analysis of the meteorological conditions and likely urban effects. The
only variable treated in a quantitative manner in some modern neighbourhood plans is solar
access rights, which are based on geometric rules derived from apparent solar position.
Unfortunately, most publications on the application of urban climatology to the
planning process fall into one of two categories: they are either cases studies of conditions
in existing settlements (Potchter, 1988); or they provide only general recommendations, but
not detailed design tools (Landsberg, 1968; Golany, 1996). Borve (1982), Westerberg and
Glaumann (1990/1991) and Pressman (1996) offered general guidelines for so-called winter
cities; Aynsley and Gulson (1999) proposed design strategies for humid tropical cities, in
general; while Emmanuel (1995) suggested strategies specific to Colombo, Sri Lanka. Herz
(1988) drew up recommendations for the Sahel region.
The awareness of the importance of climatological inputs in the process of urban
planning is growing in recent years. However, even where environmental concerns are the
subject of much public debate and where urban planners are interested in climatic
aspects of design, the use of climatic information is not systematic and climatology has
little impact on the planning process (Eliasson, 2000). The reasons for this are not only
conceptual and knowledge based, but are also related to technical matters, policy issues,
organizational aspects and the market.
ON THE DIALOGUE BETWEEN URBAN PLANNERS
AND URBAN CLIMATOLOGISTS
Mills (1999) has suggested that the difficulty in applying research in urban climatology to
architectural design problems may be explained by the fact that despite the common
96 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 96
interest in the urban climate, these fields pursue different research interests, employ
contrasting methodologies and present results differently. Each of these contrasts
deserves further elaboration.
Urban climatology now encompasses issues as diverse as the effect of street design
on pollution dispersion and the effect of cities on the hydrological balance; but a large
proportion of the studies has been devoted to the phenomenon of the urban heat island.
Although it has been demonstrated that the urban heat island occurs mostly on clear
nights with no wind conditions that are not necessarily frequent in many cities it has
been the focus of considerable research since it was proposed that its maximum intensity
was related to the population size of cities (Oke, 1973). The success of models devoted
to the urban street canyon (Nunez and Oke, 1977; Oke, 1981; Nakamura and Oke, 1988)
has led to its widespread use as the preferred unit of study irrespective of the fact that
most urban spaces do not conform to this simplified generic form. Architectural research,
meanwhile, has tended to focus mostly on daytime phenomena and on issues of human
thermal comfort.
Research by architects on urban climate is typically concerned with observing the
climatic behaviour of urban spaces, with the underlying assumption that successful
examples may then be examined to elucidate the fundamental physical characteristics that
are most responsible for creating the desirable conditions. It often relies on studies of
vernacular architecture, seeking to apply the distilled experience of previous generations to
modern-day situations (see, for example, Krishan, 1996). The emphasis on physical form
and material is in marked contrast to the interest of climatologists in studying the
processes and the fundamental principles that drive them, which may require isolation of
a process and its presentation in abstract terms. Meteorological models necessarily involve
simplification of the real world, and applying the insights gained by such methods to
planning in the complex reality of a city may therefore be quite difficult.
Climatic research by architects generally has a strong focus on practical application.
As Mills (1999) noted, the results of architectural research are often formulated in terms
of guidelines or methodologies for other designers. These are frequently demonstrated
with examples from the real world, which are presented as proof that the underlying
principles discovered through the research may be applied in practice.
A PARADIGM FOR CLIMATE-CONSCIOUS URBAN PLANNING
DESIGN GOALS
Oke (1988a) with mid-latitude cities in mind, proposed the following objectives for urban
planners considering an appropriate response to climate:
maximize shelter;
maximize dispersal of pollutants;
maximize urban warmth;
maximize solar access.
Considering the aspect ratio of streets as the only parameter to be modified in order to
achieve these objectives, Oke (1988a) suggested that a height-to-width ratio (H/W) of 0.4
The Application of Urban Climate Research in the Design of Cities 97
ABER_CH_04 6/12/08 3:32 PM Page 97
was a compromise that would lead to satisfactory performance with respect to all of
them. This implies a very low density compared to what is found in many existing urban
centres. However, several studies (Manins et al, 1998; Mills, 2006) have shown that a
compact city performs best on a number of measures, especially reduction of energy use
in transport implying a much higher density is desirable.
Okes (1998a) objectives were framed in a very narrow and specific framework. Page
(1972) discussed the effects of microclimate climate on a very broad range of issues
encompassed in the field of urban planning and design. These include optimization of
land-use patterns in relation to different activities to be carried out in the town;
identification and development of suitable microclimates for various activities, such as
parks or recreation; identification of adverse microclimatic factors likely to affect the
detailed design of urban systems, such as high local winds; optimization of building form
in relation to external climatic inputs, such as solar radiation and wind; optimization of
building form in relation to microclimatic modification of the immediate exterior domain
of the building, such as the high winds induced near ground level by tall buildings;
constructional safety, especially with respect to high winds; selection of appropriate
building materials; planning of the construction process itself in view of climatic
constraints; control of water runoff; assessment of building running costs (heating,
ventilation and air conditioning (HVAC), lighting, etc.) in advance of construction;
optimization of the operating environment of transport systems (e.g. avoidance of ice
hazards); and control of the environmental impact of a transport system on its adjacent
urban systems (e.g. with respect to air pollution by vehicles).
As Page (1968) demonstrated, urban microclimate may affect our lives in diverse
ways. To summarize this section, it may be useful to organize these effects into two main
categories:
1 The effect of microclimate on human activity, especially pedestrian, in the spaces
between buildings. The urban fabric consists of buildings and open space,
which may, in turn, be classified according to intended patterns of use. Where
pedestrian access is considered valuable, design of the outdoor spaces intended
for humans should provide optimal conditions, as appropriate in the local climatic
conditions.
2 The effect of microclimate on the performance of buildings, especially with respect
to energy conservation. The magnitude of modifications to the microclimate
resulting from the effects of the urban fabric has drawn the attention of researchers
to the need for tools to predict them and to devise design strategies to respond to
them (Taha, 1978; El-Sioufi, 1987; El Nahas, 1996). The impetus for some of the
research has been the proliferation of computer software for building thermal
analysis, which relies on meteorological data to predict interior conditions. While
many building simulation codes are now considered to be quite accurate, significant
errors may be introduced to the simulations as a result of weather data, which is
based on regional averages but which may not be representative of site-specific
conditions (Even and Williamson, 2006).
98 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 98
SUBSIDIARITY
Architects and planners must deal with a multitude of factors. Often, the demands of
different consultant experts introduce conflicting requirements upon the architect, so the
design of urban space frequently involves a process of optimization. It is thus of great value
to be able to establish the benefits of a particular approach in general terms without resorting
to a unique policy required to achieve the desired goal. Furthermore, if there is more than one
solution that may yield the required result, the preferred solution is one that may be applied
as late as possible in the planning process, and which thus has the least impact upon other
aspects of the design. This approach, of seeking the solution for a particular issue at the
lowest possible level of the planning process, may be termed subsidiarity.
The following example may serve to illustrate this principle. Solar access, especially
with respect to direct radiation, has generated much research interest (Gupta, 1984;
Littlefair, 1998; Bourbia and Awbi, 2004; Robinson and Stone, 2004; Bozonnet et al, 2005;
Ali-Toudert and Mayer, 2006). The reasons for the profusion of studies into this particular
aspect of the urban microclimate are self-evident: solar radiation is the driving force of all
climatic systems; it may be studied through the application of simple rules of geometry;
and its recommendations are likewise formulated in terms of geometrical restrictions on
building volumes or the proportions of streets. Guptas (1984) work is instructive: using
the rather basic computing tools of the time, he produced an excellent analysis of the
effects of solar radiation on urban geometry in hot climates. Yet, while acknowledging the
fact that external shading devices such as pergolas, awnings, etc. may be used to limit
the solar exposure of building openings thus undermining much of the rationale for
the analysis the study then proceeds to study the effects of structural shading (i.e. the
shadows cast by the basic form of the building) as the primary criterion for assessing the
relative merits of the different options. These include the choice of building form, building
height, street orientation and street height-to-width ratios all of which could be decided
upon prior to the study of window shading.
COMPLEXITY
The previous example illustrates another common shortcoming of some attempts to
apply the scientific method to urban planning. In order to analyse a particular question,
researchers often simplify the issue, studying its effects in isolation from other factors
that may be involved. This has clear advantages as far as analysis of the physical
processes is concerned; yet, great care must be used in the synthesis of research results
into an overall planning strategy that may be applied to a specific urban location. Thus,
deriving an optimal urban form on the basis of exposure to solar radiation risks
overlooking the effects of other factors, such as energy emitted by long-wave radiation or
heat exchange by convection. The formulation of design recommendations on the basis
of such research must also take care to very carefully define the goals to be achieved:
design for pedestrian comfort and for building energy savings may lead to contradictory
requirements.
The role of climatologists in a real-life design team also includes responsibility for
analysing climatic conditions in the urban area in question in order to identify the critical
ABER_CH_04 6/12/08 3:32 PM Page 99
issues to be addressed by the proposed plan. Some cities may have only one clearly
defined period in which meteorological conditions introduce significant stress that
deserves the planners attention. Yet, many cities, such as New York, for instance,
experience cold winters and hot-humid summers, both conditions requiring careful
analysis with respect to possible planning intervention. Givoni (1989) presents such an
analysis for a number of different climates. The report deals with the definitions of human
thermal comfort, provides a general description of the relevant characteristics of the
urban climate, and discusses the effects of various planning features, housing types and
vegetation on the urban microclimate. Its broad scope is an indication of the complex
nature of the issue and the multifaceted response required in order to be of value in the
context of city planning.
ECONOMIC VIABILITY
The recommendations of urban climatologists with respect to city planning may often
have significant economic implications. Urban development is driven to a great degree by
economic considerations, and zoning regulations often reflect the desire of city planners
to attract investment by real-estate developers. Street width, for instance, is generally
determined by the requirements of vehicular access, while building height reflects the
desire to maximize the value of land. Thus, any recommendation concerning the height-
to-width ratio of streets, which has a major influence on the canopy layer climate of cities,
may also have considerable economic implications. Any explanation for the relative lack
of success in implementing climate-related strategies in urban planning must therefore
consider the lack of a practical framework to assess their economic effects too.
CLEAR AND IMMEDIATE BENEFITS
In order for a particular recommendation based on urban climate considerations to be
adopted by planners, the proposed strategy should have clear and immediate benefits.
For example, the effect of urban climate modification on building energy consumption
could be estimated, taking into account the urban heat island, as has been demonstrated
from measured temperature data for Athens (Santamouris et al, 2001) and London
(Kolokotroni et al, 2006). To realize this aim, urban climatology is required to develop
sufficiently accurate and reliable predictive tools. In the absence of quantitative studies on
the effect of proposed designs upon climate, and on the basis of well-documented
evidence from other planning professions, decision-makers, in general, tend to
downgrade the importance of climatic considerations in urban planning.
COMPREHENSIVE APPROACH TO PROBLEM-SOLVING
In order to apply urban climatology effectively in the process of town planning, a
comprehensive approach must be adopted. Recommendations must be based on
analysis of all factors influencing the urban microclimate. Conclusions based upon the
study of one or even several factors in conjunction may be misleading if they fail to take
into account the effect of other significant processes. With the increase in computing
power, computerized modelling may be capable of providing accurate and comprehensive
analysis of the urban microclimate. Once such models become reliable and accurate
100 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 100
enough, they should be applied wherever possible to inform the decision-makers of the
microclimatic implications of urban planning strategies under consideration.
For the models to be useful, they must allow the study of the particular issues that are
foremost in the architects mind. In other words, they must be formulated so that the inputs
include parameters related directly to the architects decision-making process. The
following section illustrates the effect of planning decisions on the resulting microclimate.
THE EFFECT OF DESIGN DECISIONS ON THE URBAN MICROCLIMATE
Modern urban planning is a complex process, generally carried out by teams of
professionals from a number of fields. While the architect or town planner may have
overall responsibility for producing the plan and for coordinating the project, crucial inputs
are provided by consultants from other disciplines. The process is generally driven
primarily by economic forces in response to market demand for housing, retail space, etc.
The input of the urban climatologist is therefore but one of many that compete not only
for the architects attention, but also for appropriate consideration in the process of
optimization that takes place in urban planning.
Mills (2006) noted that while the meteorologically ideal settlement serves a useful
pedagogical purpose, it does not recognize planning realities where climate issues are
rarely a dominant concern. So, while a statement of desirable outcomes and the means
of achieving them may be a logical means of applying urban climatology, climatologists
seeking to inform the decision-making process are rarely in the position of generating a
plan. Rather, they may be required to respond to proposals made by other members of
the design team. In order to do so effectively, urban climatologists must be aware of the
effect that climatic considerations may have on the issues that the planners must resolve
when preparing a town plan. This section is therefore structured to deal with the effects
of various features of the urban form on microclimate from the perspective of an urban
planner.
URBAN DENSITY
The density of a city is generally determined by economic considerations, reflected in the
price and availability of land. It is also influenced, and in turn has an effect upon, the
overall form of the town. Architects and planners typically measure urban density by
means of the number of dwellings per unit area of the site (Knowles, 2003) or by the ratio
of the total built floor area to the area of the site, an index more suited to non-residential
development. Urban climatologists, on the other hand, refer to density by different
measures: the plan area density, which is the ratio of the buildings footprint to the total
area of the site; or the frontal area density, which is the ratio of its (windward) elevation
to the site area (Grimmond and Oke, 1999).
Density has a direct effect on the exposure of urban surfaces to direct solar radiation.
Such exposure may be considered beneficial in cold climates or detrimental in hot
conditions, where shade is considered desirable. However, whereas the latter objective is
fairly easy to achieve for example, by the addition of specialized shading elements such
as pergolas and blinds the former imposes stringent limitations on the overall built
volume that can be constructed in a given site, as well as upon its geometry. The so-called
ABER_CH_04 6/12/08 3:32 PM Page 101
solar envelope has been the subject of numerous studies by architects, with the aim of
ensuring appropriate solar exposure of buildings (Knowles, 1981) or of open space
(Yezioro et al, 2006).
Solar exposure of buildings has been a primary concern of many planners in climates
with cold winters to allow for passive solar heating. It has sometimes been investigated
in isolation from other aspects of urban climatology because it only requires knowledge
of geometry. Littlefair (1998) surveyed a variety of graphic methods to establish solar
exposure, though a substantial number of computer-aided design (CAD) programmes
now perform the task automatically, given a geographical location and time of year.
However, the study raises the intriguing question of how to establish the criteria for solar
exposure. It proposes that rather than mandate solar access on the winter solstice, when
the sun is lowest, it may be preferable to define a heating season and aim to maximize
gains over the whole period. Meeting the first requirement in high-latitude locations
requires exceedingly large distances between adjacent buildings in order to provide what
may be marginal benefits because sunshine hours are short and insolation is limited.
Guaranteeing exposure when insolation levels are higher may fill a proportion of heating
requirements that is only slightly smaller, yet requires far less stringent geometric
limitations on building height.
In tropical climates, exposure to solar radiation is generally undesirable. Planners in
such locations are therefore concerned with creating urban geometries that maximize
shade. Narrow streets and a dense urban matrix sometimes are often recommended for
desert locations (Mazouz and Zerouala, 1998), although they restrict ventilation; but
designs for warm-humid locations must maximize airflow while providing shade in public
spaces too (Emmanuel, 1993), so tropical cities may be less dense.
In temperate climates, design for solar access reflects a desire to accommodate
sometimes contradicting requirements in response to seasonal weather and solar
exposure. Swaid (1992) proposed that operable screens be installed on building rooftops
to restrict solar access into street canyons when radiation is excessive, yet which are
capable of being folded away when full exposure is desirable. He simulated the effect of
adjustable screens on canyon air temperature, and reported significant differences among
the various configurations tested. Knowles (2003), too, suggested that the solar envelope
of a building should be adjustable, terming the zone that lies between the summer and
winter extremes the interstitium.
The effect of urban thermal mass on air temperature may be seen as being analogous
to the effect of thermal mass on interior temperatures of buildings (Ratti et al, 2003). Erell
and Williamson (2007) suggest that since the three-dimensional geometry of the city
results in much larger surface area compared with a flat rural site with the same plan area,
effective thermal mass is greater in the former, which tends to reduce the diurnal
temperature swing. Field experiments carried out in Adelaide, Australia, have shown that
the diurnal amplitude of air temperature in a deep street canyon was 60 to 70 per cent of
the rural one, and was the result not only of elevated night-time minima, but also of
reduced daytime maxima. This suggests that the dense urban structure often found in
traditional quarters of many desert cities reduces the diurnal temperature extremes
characteristic of such climates.
102 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 102
STREET ORIENTATION
The planning of streets is often the primary input in the master plan of a city. While it is
true that it is the architecture of the buildings that give a city character, it is the street
layout that determines its structure. In modern town planning, streets are generally
planned in response to the requirements of transport systems. Transport planners draw
up the appropriate road links in response to the density and type of expected land use in
accordance with national standards or accepted best practice.
The orientation of streets with respect to the path of the sun or to the prevailing winds
is now rarely considered during the design process, although its effect on microclimate
was recognized even in historical times: Vitruvius (1999), discussing orientation in
approximately 20 BC, wrote:
. . . the orientation of streets and lanes according to the regions of the heavens. This
process will be properly accomplished if, with foresight, the lanes are kept from facing
into the path of the prevailing winds. For if the winds are cold, they injure; if hot, they
corrupt; if moist, they are noxious.
Kenworthy (1985), discussing the regular street pattern of some ancient cities such as
Miletus (in Asia Minor), proposed that the opposite was true: promoting exposure to
regional winds on an urban scale has been an aim of city planners from ancient times.
What, then, is the orientation (relative to wind direction) that best promotes
ventilation? Kenworthy (1985) found that while streets oriented parallel to the prevailing
winds would appear to offer the least aerodynamic resistance, scale-model tests of
orthogonal block grids showed that the maximum wind speed at street level was, in fact,
measured when the wind blew at a small angle to the main axis. Bottema (1999) also
found that parallel flow resulted in lower roughness (z
0
) than normal flow, but noted that
long buildings aligned with the wind created flow channelling that actually impaired
ventilation and removal of pollutants because of reduced vertical mixing. If the street
width is less than twice building height, shelter is enhanced, but ventilation is reduced.
Several studies on the orientation of streets, usually in a grid scheme, have
recommended various orientations on the basis of exposure of buildings or of street
surfaces to direct solar radiation. Gupta (1984) found that in composite climates, an
eastwest street with continuous wall surfaces was the optimum configuration (from
building energy considerations), but in low-latitude locations, a northsouth street axis
gave buildings equal solar protection to an eastwest oriented street. A similar study
(Mills, 1997) compared the effect of building group configuration on the thermal stresses
affecting individual buildings on the basis of two measures: solar exposure (controlling
heating) and sky view factor (controlling cooling by long-wave radiation). The study
provides a useful insight into the effects of urban geometry; but recommendations for
different climate zones (defined by latitude) are, of necessity, too simplistic, not least
since they ignore the effects of convection on building energy needs.
Considering solar exposure of streets, rather than buildings, Knowles (1981) arrived at
different conclusions: a Spanish grid (in which streets are oriented at 45 to the cardinal
points of the compass) was found to be preferable to the so-called Jeffersonian grid
ABER_CH_04 6/12/08 3:32 PM Page 103
(in which streets are oriented eastwest and northsouth). This is because in the Spanish
grid, part of the street is shaded while the opposite sidewalk is exposed to the sun all year
round, allowing pedestrians to choose between different conditions. The Jeffersonian
grid, in contrast, maximizes midday exposure on the northsouth streets in both summer
and winter, while eastwest streets will be in shadow most of the time in winter, unless
they are very wide, and exposed to the sun during summer. The effects of street
orientation on solar access in low-latitude desert cities were also modelled by Bourbia and
Awbi (2004) and Ali-Toudert and Mayer (2006), the latter emphasizing that the effects of
street orientation on thermal comfort should be considered with respect to the daily
patterns of use by pedestrians since solar access varies with time of day. Shashua-Bar and
Hoffman (2003) pointed out that extensive planting of trees minimizes the effect of street
orientation with respect to solar access. However, the question of street orientation
should be considered not only in relation to solar access (or protection), but also with
respect to wind direction, especially where lack of ventilation is a major problem (Ahmed,
2000).
STREET ASPECT RATIO
Practising architects typically refer to the height of building faades facing a street or to
the width of the street; but rarely, if at all, do they refer to the ratio between them, except
in a qualitative sense. Climatologists, on the other hand, have found it useful to study the
city by means of a simplified form referred to as an urban canyon: a semi-infinite street
with a rectangular cross-section bounded on both sides by continuous buildings of
equal height. This prototypical urban space is characterized by means of its aspect ratio,
or the ratio of building height (H) to the width of the street (W). The aspect ratio of a street
is one of the most important controls on microclimatic conditions in the street, affecting
the transfer of energy, mass and momentum occurring in the space between buildings.
Radiant exchange
The canyon aspect ratio affects radiant exchange in different ways during the daytime and
at night. Deep streets have a small sky view factor and therefore lose less heat at night
by long-wave radiation than shallower ones. Oke (1981) demonstrated how canyon
geometry could explain the formation of the nocturnal urban heat island (UHI), previously
thought to be linked to the population size of the city (Oke, 1973). Okes (1981) experiment
is supported by numerous field studies (Barring et al, 1985; Yamashita et al, 1986;
Runnalls and Oke, 1998; Goh and Chang, 1999; Livada et al, 2002; Chow and Roth, 2006;
Erell and Williamson, 2006) and has formed the basis for much subsequent research on
the UHI.
The effect of the canyon aspect ratio on air temperature at night is well understood:
the city cools down more slowly than rural areas and has higher minimum temperatures.
However, the processes that occur during the daytime are more complex. Deep street
canyons restrict access to direct solar radiation and provide shade at street level, but they
also act as radiation traps: solar radiation that might otherwise have been reflected back
to the atmosphere from high-albedo surfaces experiences multiple-reflections among
canyon surfaces. The overall balance of these opposing tendencies depends inter alia on
104 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 104
the exact proportions of the street and on the thermal properties of the surfaces. Several
studies of air temperature in urban streets show that they are warmer than rural areas in
the daytime as well as at night (Santamouris, 1998; Giridharan et al, 2004), while others
demonstrate that air temperature in city streets may, in fact, be slightly cooler during the
day (Steinecke, 1999; Runnalls and Oke, 2000; Erell and Williamson, 2007).
The canyon aspect ratio also has an effect on the overall albedo of a city with a regular
geometry. Aida and Gotoh (1982), using a numerical model, found that the maximum solar
absorption in a mid-latitude city such as Tokyo occurs when the canyon width is
approximately twice the block width. Soler and Ruiz (1994) confirmed this finding in an
empirical study comparing the intensity of reflected radiation in satellite images of urban
and rural areas near Barcelona with terrestrial measurements. It should be noted that
existing high-albedo cities are characterized by either high density (where roof albedo
dominates) or very low density (where the albedo of the ground is most important). In
medium-density cities, multiple reflections among canyon facets emphasize the
contribution of walls in addition to the horizontal surfaces.
The effect of street canyon geometry on radiant exchange may have a great effect on
human thermal comfort, an issue often overlooked in modern street design. In hot
climates with high radiant loads, net radiant balance may be more important than
convective exchange, and the benefit to pedestrians of shade outweighs minor
modifications to air temperature that might be measured in the street (Pearlmutter et al,
1999, 2006).
Airflow
Although airflow in cities takes place in open spaces of all types and sizes for example,
backyards, parks and plazas, in addition to streets it is streets that have received the
most attention from both climatologists and designers addressing this issue (although for
different reasons).
Climatologists have found the two-dimensional street canyon a useful generic urban
form to model airflow, using the aspect ratio (H/W) as one of its primary geometric
descriptors. Oke (1987) classified urban airflow into three basic regimes isolated
roughness flow, wake interference flow and skimming flow identifying flow patterns
associated with certain ranges of street aspect ratios. Airflow in street canyons with
uniform building heights and an aspect ratio of approximately unity is characterized by a
lee vortex generated by the above-roof wind, and may display a corkscrew pattern as the
cross-canyon circular motion is combined with an along-canyon component (Hotchkiss
and Harlow, 1973; Yamartino and Wiegand, 1986). The effect of the canyon aspect ratio
on airflow patterns in street canyons that are either much shallower or much deeper, or
which have buildings of unequal height on either side (step up or step down with respect
to the wind), has been the subject of several recent studies (Baik and Kim, 1999; Baik et al,
2000). In deep canyons, two major counter-rotating vortices may occur in some conditions,
the lower one driven by the upper (So et al, 2005). The thermal flows generated by
unequal heating of canyon surfaces further complicate the picture (Kim and Baik, 2001;
Xie et al, 2005, 2006). Assimakopoulos et al (2006) report on field studies that show the
existence of intermittent vortices that cannot be modelled by quasi-steady state models.
ABER_CH_04 6/12/08 3:32 PM Page 105
They note that existing computational fluid dynamic (CFD) models are still incapable of
fully reproducing the turbulent nature of airflow in reality, and cannot be used to predict
the evolution of airflow at a particular urban location for any length of time. The
agreement between predicted and actual airflow is especially poor when wind speed is
low less than 2m/s when ventilation might be difficult to provide (Dixon et al, 2006).
Large eddy simulation (LES), requiring very substantial computing resources, though
much more powerful than CFD, is also still not sufficiently developed to be applied as a
tool for generating design strategies for even simplified generic forms of the urban
structure.
The geometry of street canyons is of primary importance in the study of pollutant
dispersion, especially automobile exhaust fumes. Studies have been carried out using
scale models in wind tunnels (Cermak, 1995; Kastner-Klein and Plate, 1999; Al-Sallal et al,
2001), as well as numerical simulation, once computing power had developed sufficiently
to allow numerical simulation of airflow in the presence of arrays of bluff objects
representing buildings (Kim and Baik, 1998, 2004; Chan et al, 2001, 2003). The flow
regime in a canyon (and, hence, the effectiveness of flushing out pollutants) depends not
only upon the wind speed but also, to a great degree, upon the canyon aspect ratio
(So et al, 2005). As might be expected, deeper canyons are more likely to suffer from high
concentrations of pollutants than shallower ones, while irregular building height promotes
turbulence and therefore improves pollutant removal. The effect of roof geometry is also
important (Rafailidis, 1997; Kastner-Klein and Plate, 1999). Much of the pollutant removal
occurs at intersections, which are characterized by unique flow regimes (Chan et al, 2003;
Dobre et al, 2005).
NEIGHBOURHOOD AND BUILDING TYPOLOGY
A feature of urban development clearly visible to all city dwellers is building typology.
Many streets conform to the urban canyon model that has proved such a useful tool for
urban climatologists. Yet, many more do not. Gupta (1984) identified three basic building
forms, which may translate into quite different urban morphologies: pavilion, court and
street. Steemers et al (1997) found six generic urban forms: pavilions; slabs; terraces;
terrace courts; pavilion courts; and courts different combinations of which produced
distinct wind patterns in wind tunnel experiments.
The complexity of real cities, as opposed to simplified conceptual models composed
of generic building types and regular street patterns, makes analysis of the effect of
urban form on micro-scale environmental behaviour very difficult. Computerized image-
processing techniques offer a method of dealing with such conditions by analysing urban
texture using a digital elevation model (DEM) of the site in question (Ratti and Richens,
1999, 2004). The DEM is a compact way of storing urban 3D information using a 2D
matrix of elevation values: each pixel represents the height of a building (or part of it) and
can be displayed in shades of grey as a digital image. The computer model may then be
analysed for any geometric characteristic desired, such as the sky view factor from the
ground at any point, the extent of shadowing or sunlight availability. The technique was
used in an analysis of building energy consumption and urban texture (Ratti et al, 2005)
using the LT model
1
to calculate energy requirements and the DEM to differentiate
106 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 106
between core sections of buildings, which require more energy to service, and perimeter
areas, which are more amenable to daylighting, ventilation and passive heating. A further
study using this technique suggested that in a hot-arid climate, courtyard buildings (of
certain proportions) were a more appropriate building type than the pavilion type on the
basis of their surface-to-volume ratio, shadow density, daylight distribution and sky view
factor (Ratti et al, 2003). The same study also suggests that courtyard buildings would
be less suitable for tropical climates, where a narrow temperature variation would not
benefit from the high thermal mass and reduced ventilation characteristic of this
configuration. Texture analysis with image processing can also be used in the study of
winds and dispersal of pollutants in the city (Ratti et al, 2006). The variation of wind
speed above the ground is affected by the morphology of the surface, which can be
described by means of the roughness length (z
0
) and the displacement height (z
d
). The
roughness length is, in turn, a function of the average height of the buildings and a
measure of their density, such as the frontal area density (for a given wind direction) or
the plan area density (Grimmond and Oke, 1999). These can be obtained easily from a
DEM and the aerodynamic roughness derived for the site being studied. Using this tool,
analysis of hypothetical arrangements of long buildings arranged in parallel rows showed
that counter to design intuition and common sense, aligning the streets parallel to the
wind resulted in poorer ventilation and slower pollutant removal than aligning them
perpendicular to the wind.
SIZE, TYPE AND LOCATION OF URBAN PARKS
Vegetation affects conditions in the city in a variety of ways. For instance, several studies
found that vegetation dry-precipitates and adsorbs pollutants and, by doing so, decreases
the mass of airborne gases and particulate matter (Raza et al, 19901991; Taha et al,
1997). However, the following section deals only with the effects of vegetation on the
energy balance and upon air temperature.
The effect of local parks in a non-homogeneous urban area has been the subject of
intense study, especially once it became clear that the microclimate of built-up areas
differed substantially from that of rural areas. The so-called park cool island (PCI), a
manifestation of the more general oasis effect, is the converse of the urban heat island
(UHI): empirical findings show that air temperature in moderate- to large-sized parks may
be substantially lower than temperature in surrounding built-up areas (Jauregi,
19901991; Kanda and Moriwaki, 1998), although there are significant variations among
different types of parks.
Landscaping, specifically the incorporation of planted areas within the urban fabric,
may reduce differences between natural terrain and the urban surface. Bonan (2000)
demonstrated in an empirical study of a town that the availability of water and the
resulting increase in evaporation was the main factor responsible for lower surface
temperature and air temperature above the lawns, rather than the mere presence of
plants. This helps to explain the finding that increasing the proportion of planted areas in
a city tends to reduce daytime maximum temperatures, but may often have little or no
effect on night-time minima (Urano et al, 1995). Emmanuel (1997) noted that the main
effect of trees is to reduce radiant exchange at the ground surface: this may decrease
ABER_CH_04 6/12/08 3:32 PM Page 107
daytime maximum temperature, but would also restrict nocturnal cooling, actually leading
to higher minima.
Classification of urban parks
Spronken-Smith and Oke (1998) distinguished between park cool islands that are defined
by surface temperature, which may be quite large, and cool islands defined by air
temperature, where the effects of surface temperature variation are diluted by near-
surface turbulent mixing and advection by wind. During daytime, surface temperatures
are affected by the presence or absence of shade, by surface albedo, by water availability
and by the thermal properties of the soil. These properties govern the receipt of solar
radiation, its absorption and the role of evaporative cooling. At night the thermal
properties of surfaces and the radiative geometry are the major controls on cooling.
Urban parks vary substantially with respect to the above factors, and may be classified
according to the arrangement of vegetation (Spronken-Smith and Oke, 1998): grass;
grass with tree border; savannah (grass with isolated trees); garden; forest; and multi-use.
Park cool islands may develop either during the daytime or at night. However, a given
urban park will display a regular diurnal pattern, indicating that the formation of PCIs may
be the result of a number of mutually exclusive factors. Spronken-Smith and Oke (1999)
found that daytime PCIs formed as a result of the combined effects of soil moisture and
shading: trees shade the surface, while grass is typically cooler than most solid surfaces
during the daytime if it is well irrigated. The relative coolness of irrigated parks therefore
peaks in the afternoon (forest type) or early evening (garden, savannah and multi-use
types). However, trees also inhibit nocturnal long-wave radiative cooling by blocking off
part of the sky, while excess moisture increases the thermal capacity of the soil and slows
down surface cooling. Night-time PCIs therefore typically form in relatively dry urban parks
with a sparse tree cover. They are driven by long-wave radiative cooling (since the sky view
factor is close to unity), and since evaporative fluxes are generally weak at night,
evaporation does not play a significant role in the formation of this type of PCI. In such
parks, daytime temperatures may sometimes be higher than in neighbouring urban areas.
However, an edge effect exists that applies within distances of about 2.2 to 3.5 times the
height of the park border, and which results in weaker radiative cooling where the sky view
factor is reduced by the obstructing features, such as perimeter trees or buildings.
Effects of size
The extent of vegetated area required to produce measurable effects on air temperature
is of great interest to urban planners. Upmanis et al (1998) found that the magnitude of
the intra-urban temperature difference between parks and their urban surroundings at
night increased with increasing park size, although large differences were also found
within the parks and in the built-up areas. These were attributed to the degree of exposure
to the sky and, hence, to the intensity of radiant cooling. Saito et al (19901991), in
contrast, reported cool islands in clumps of vegetation less than 200m across, although
the effect of vegetation was limited to the planted area itself and was not felt at distances
as little as 20m away from the park edge. Ca et al (1998) reported on the basis of
measured data that an urban park 0.6km
2
in area can reduce air temperature in a commercial
108 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 108
area 1km downwind by as much as 1.5C; but Shashua-Bar and Hoffman (2000), who
reported differences of up to 3C between air temperature in tree-shaded urban avenues
and nearby non-shaded reference points when wind speed was very low, noted that the
cooling effect declined at an exponential rate with increasing distance from the border of
the planted area, and vanished less than 100m away from the sites studied. Numerical
modelling (Bruse and Fleer, 1998) indicates that small parks of only tens of metres across
may create temperature differentials of 2C or more. However, the horizontal gradients of
air temperature represented by the models may be quite large, and their spatial patterns
shift constantly with wind speed and direction, as demonstrated by Upmanis and Chen
(1999) in Goteborg. A pronounced difference in air temperature is almost always observed
very close to the edge of the park, a phenomenon also noted by Santamouris (2001) in a
study carried out in Athens.
The effect of trees on urban air temperature
The presence of trees in the urban matrix may affect air temperature at a variety of spatial
scales, from individual streets (Shashua-Bar and Hoffman, 2000) to city-scale
modifications (Huang et al, 1987). However, the magnitude of this effect may depend
upon a variety of factors because the interaction between trees and other constituents of
the urban environment is so complex (Oke, 1989). Trees intercept not only solar radiation,
but also long-wave radiation from the ground, building surfaces and the sky. The
dissipation of this heat load depends upon the water balance and wind climate of the tree.
In the presence of unrestricted water, transpiration will cause substantial cooling.
However, water supply to the root system may be restricted; stomata may be physically
blocked by particulates; or the heat load may be excessively high, leading to closure of
the stomata. Furthermore, at night (in the absence of sunlight) there is no photosynthesis,
so the stomata are closed and the tree is not cooled by transpiration. The response of
trees to increased energy loading, which may occur when individual trees are planted in
extensive paved areas such as parking lots, for example, will vary with species, humidity
of the atmosphere and how much of the crown is exposed (Kjelgren and Montague,
1998). Species from hot or arid habitats may be tolerant of high temperatures or able to
dissipate heat with small leaves but the evapotranspiration rates from such trees may
accordingly be lower than those of broadleaf trees, and thus they have a smaller effect on
air temperature in their surroundings.
Grimmond et al (1996) found that in a neighbourhood with a relatively dense tree cover,
the latent heat flux increased as a fraction of available energy compared to an otherwise
similar neighbourhood with a sparse tree cover, as did the storage flux, whereas the
sensible heat flux decreased. However, in absolute terms, all fluxes, including the sensible
heat flux, were enhanced at this neighbourhood. Trees and shrubs lowered the albedo and
surface temperatures, thereby reducing the loss of solar and long-wave radiation,
respectively resulting in an increase in the overall amount of energy to be dissipated
compared to the sparsely planted neighbourhood. As a result of the difference in fluxes,
the maximum air temperature above the canopy was about 1C higher in the densely
planted neighbourhood, while temperatures below the trees at noon and in the early
afternoon were similar or up to 0.5C higher than in the sparsely planted neighbourhood.
ABER_CH_04 6/12/08 3:32 PM Page 109
In contrast to Grimmond et al (1996), most researchers report that trees reduce urban
air temperature. This is usually attributed to evapotranspiration; but Shashua-Bar and
Hoffman (2003) suggested that the cooling is due almost entirely to shading, which
more than offsets the exchange of sensible heat between the tree canopy and the air.
The actual reduction in air temperature might be overstated in many cases due to the
difficulty of measuring air temperature accurately in the presence of strong radiant loads
(Erell et al, 2005).
A model specifically designed to predict the general microclimatic effect on energy
consumption of augmenting urban vegetation was proposed by Sailor (1998). The
meso-scale model (2 x 2km grid) predicts that by increasing the vegetated fraction of the
core of a hypothetical city by 0.065, annual cooling loads could be reduced by 3 to 5 per
cent simply by lowering the background air temperature. Shashua-Bar and Hoffman (2002)
proposed a quantitative model for predicting the cooling effect of trees in an urban
setting; but their method incorporates site-specific empirical factors to account for
convective exchange, the magnitude of which is not systematic and for which there is no
method of calculation. The method is, nonetheless, demonstrated in an empirical study
of the effect of trees on air temperature in the streets of Tel Aviv, Israel, which found that
the overall cooling effect of trees could be as high as 3K, depending, in addition to the
shade coverage, upon the depth of the urban canyon, the albedo of canyon walls and
street orientation (Shashua-Bar and Hoffman, 2004).
The effects of vegetation on building energy consumption
Landscaping and careful planting of vegetation near buildings have been credited with
energy savings of anything up to 80 per cent in hot, dry climates (Meier, 19901991).
Several studies have been carried out to quantify the effect of vegetation, especially trees,
on the energy consumption of buildings (Simpson and McPherson, 1998; Simpson,
2002); but much of the evidence remains anecdotal.
The mechanisms by which vegetation affects the energy exchange between buildings
and the environment may be summarized as follows:
Vegetation can reduce energy consumption in buildings in hot climates if air
temperature is reduced near the planted area. However, it should be noted in this
context that heat transfer through building walls is driven by differences in surface
temperature, rather than by air temperature. Furthermore, the reduction in air
temperature resulting from evapotranspiration is accompanied by an increase in the
vapour content of the air. Therefore, the air-conditioning system must deal with an
increased latent heat load, offsetting, to some extent, any gains from a lower
sensible heat load.
Plants may shade building surfaces, reducing the radiant load on the envelope. This
may be beneficial in hot climates, but detrimental in cold ones. In temperate climates
with distinct heating and cooling seasons, deciduous trees are often planted, and
vine-covered trellises are common in many Mediterranean areas. However, the timing
of defoliation and the permeability of the bare trees vary widely from species to
species (Canton et al, 1994) and may not match the desired pattern of exposure to
110 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 110
the sun. McPherson et al (1988) found that in the middle latitudes, cooling loads
were most sensitive to shading on the roof and on the west wall, while heating loads
were affected most by exposure of the south (equator-facing) and east walls. The
direct effect of shading building surfaces by plants was studied by Papadakis et al
(2001). Thick foliage producing a full shade effect resulted in a reduction of the
surface temperature of a light-coloured concrete wall by up to 8C, with concomitant
reductions in heat flux through the surface. When wind speed was negligible, air
moisture content in the planted canopy was up to 2g/m
3
greater than in the
surrounding air. Likewise, a dense growth of ivy can block radiant exchange at the
wall surface almost entirely (Hoyano, 1988). However, several studies show that
where the shaded area has a limited spatial dimension for instance, beneath a
pergola (Hoyano, 1988) or in the shade of a liman (small clump of trees in an artificial
floodplain in the desert) (Schiller and Karschon, 1974) the effect on air temperature
at a height of 1m above the ground is negligible.
Plants may reduce wind speed near buildings, limiting unwanted infiltration, but also
restricting ventilation and reducing convective exchange at building surfaces. The
first two mechanisms are self-explanatory; but the third has less well-known
consequences. For instance, in hot climates, wind is an asset for unshaded houses
because it helps to remove radiant heat at the external building surfaces (McPherson
et al, 1988), reducing temperature differentials between the interior and exterior.
However, in poorly insulated houses, especially in cold climates, increased
convective exchange at the building envelope results in increased loads on building
heating or cooling systems.
Plants in warm climates may reduce temperatures of ground surfaces through
evapotranspiration (although planted surfaces may be warmer than bare soil in cold
climates), with two effects: cooler surfaces emit less infrared radiation, thus reducing
the radiant load on building surfaces; and they release less sensible heat to the
adjacent air, so that buildings are exposed to cooler ambient air.
Roof gardens (or planted roofs) are perhaps the most obvious example of the use of
plants to control building energy performance, and are sometimes credited with
improving the urban microclimate as well (Wong et al, 2003). The shading and
evapotranspiration of the plants contribute to lower surface temperatures and thus to
lower heat gains through the roof. Experiments show that the surface temperature of
an exposed roof can be reduced substantially by the addition of an irrigated lawn on
a fabric matrix (Onmura et al, 2001). A complete thermal model of a planted roof
(Palomo Del Barrio, 1998) showed that the contribution of the planting to the thermal
performance of the roof depends mainly upon the density of the foliage, the
composition, density and thickness of the substrate, and its water content. However,
unless the thermal conductivity of the soil is particularly low or the thickness of the
substrate is considerable, the thermal resistance provided by the planting and
substrate is usually insufficient during the cold season, even in mid-latitude countries
with relatively mild winters. However, much of the work in this field remains
empirical and lacks a comprehensive theoretical framework at the urban scale
(Niachou et al, 2001).
ABER_CH_04 6/12/08 3:32 PM Page 111
BUILDING AND PAVING MATERIALS
Roofing materials, wall finishes and paving blocks are typically specified by architects for
a variety of reasons, such as cost, durability and appearance. Their thermal properties,
especially insulation, are usually assessed in the context of the energy budget of the
building. However, the properties of the materials that form the surfaces that make up the
urban canopy also have an effect on urban microclimate, which may be modified through
the selection of suitable finishes.
The role of materials in the formation of the urban heat island has been the subject of
some conjecture. Oke (1981) demonstrated that the nocturnal urban heat island could be
explained not only by the effects of street canyon geometry on long-wave radiant
exchange, but also by differences in the thermal properties of urban and rural surfaces.
However, he also noted that, in reality, differences in thermal admittance between urban
and rural materials were too small and were, in any case, affected to a great degree by
rural soil moisture. A later study (Oke, 1982) suggested that differences in thermal inertia
between rural and urban sites could be accounted for by considering the role of increased
surface area in the city, as well as differences in moisture availability. However, the
difficulty of measuring the storage flux directly in situ on a neighbourhood scale has thus
far limited efforts to explain this mechanism satisfactorily.
The effect of the thermal properties of the substrate on the surface temperature and
on the magnitude of the energy fluxes was investigated by Asaeda et al (1996), who
demonstrated a substantial difference in the behaviour of asphalt and concrete (painted
black to create similar albedo). The former has a lower thermal conductivity, and the
radiant energy it absorbs at the surface during the daytime results in a higher temperature
and therefore greater emission of long-wave radiation and higher rate of sensible heat
loss. Conversely, an asphalt surface also cools down more quickly than concrete at night.
Buildings in the warmer parts of the Mediterranean are often whitewashed to reduce
the external radiant load on the buildings and, hence, their surface temperature during the
daytime. The same strategy has been proposed to reduce convective heat transfer from
pavements and buildings to the air, although the temperature of air near the surface is also
affected by several additional factors. Paving materials with a variety of colours and
textures have been investigated by several researchers to evaluate the effect of exposure
to intense sunlight on surface temperature (Bretz et al, 1998; Doulos et al, 2004; Synnefa
et al, 2006). The studies found that during the daytime, all paving materials had a mean
temperature higher than that of the ambient air, while at night all surfaces were cooler, in
some cases by almost 6C (Synnefa et al, 2006). However, differences among the samples
are illuminating: compared to an exposed concrete tile serving as the reference, Synnefa
et al (2006) found that some of the coatings tested resulted in mean daytime
temperatures lower by approximately 4C to 5C, with a maximum difference in excess of
7C, while Doulos et al (2004) found that the difference between the coolest surface
(white marble) and the warmest one (asphalt) was over 22C.
The albedo of roof surfaces may have less effect on air temperature in the canopy
layer than the properties of the pavement or the ground because it affects heat transfer
above roof level. However, since it affects the energy performance of a house directly, it
has been the subject of much research. The obvious strategy with respect to roofing
112 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 112
materials in hot climates is to employ light-coloured materials. However, progress in the
production of wavelength-selective paints now allows other colours to be specified,
which although visually fairly dark, nevertheless have relatively high albedo (Levinson et al,
2007; Synnefa et al, 2007). The so-called cool coatings can reduce the temperature of
roof tiles having the same appearance (i.e. colour) by between 1.5C and 10C (for green
and black tiles, respectively). Some of the coatings have been tested to evaluate the
long-term effects of exposure to ultraviolet (UV) radiation and mechanical damage, and
show relatively little deterioration over extended periods (Bretz and Akbari, 1997; Synnefa
et al, 2007). Some of the reduction in reflectance is attributed to soiling and can therefore
be reversed by rinsing.
High surface temperatures contribute directly to pedestrian discomfort by imposing a
long-wave radiant load (Pearlmutter et al, 2006, 2007), and may also affect air temperature
in close proximity. However, the thermal properties of urban surfaces may also have an
effect on the climate of a city as a whole: if roof albedo is modified on a large proportion
of city roofs, the cumulative effect on the climate of a city may be significant. Using a one-
point model of Sacramento, California, Taha et al (1988) suggested that maximum air
temperature (in the early afternoon on a summer day) could be reduced by about 4C if
urban albedo was increased from 0.25 to 0.4, while reducing urban albedo to only 0.1
would result in an increase of more than 4C relative to the base case. Taha et al (1997)
found, using the Colorado State University Mesoscale Model (CSUMM), that feasible
albedo changes could lower daytime maximum temperatures in central Los Angeles by
up to 2C, with smaller reductions predicted for the suburbs. The actual albedo of most
urban areas varies from about 0.1 to about 0.2 (Taha, 1997; Sailor and Fan, 2002), with a
mean value of about 0.14 for urban centres (Oke, 1988b), so there appears to be
substantial potential for cooling if the large-scale application of high-albedo surfaces can,
in fact, be attained (Bretz et al, 1998). The scale of modification possible depends upon
the area of solid surfaces, such as rooftops, streets and other paved areas as a proportion
of the total urban area. However, the micro-scale intra-urban variation of air temperature
in the urban canopy, observed in several field studies (Eliasson, 1996; Erell and
Williamson, 2007), means that the benefits of such a strategy may not be felt uniformly
over the entire city.
Extensive glazed areas also affect radiant exchange in a street canyon. While the use
of mirror glass results in extremely high albedo, sunlight transmitted to building interiors
through transparent windows creates urban surfaces that have extremely low albedo.
Tsangrassoulis and Santamouris (2003) showed that increasing the window-to-wall ratio
resulted in a lower canyon albedo because sunlight is transmitted by the glass to building
interiors. In very deep canyons (H/W > 2), the predicted albedo for all sun angles as well
as for uniform diffuse radiation was between 0.01 to 0.03 meaning that practically all of
the incident solar radiation is absorbed. Deep streets with extensive glazed areas in
adjacent buildings may therefore be viewed as a very efficient trap for solar radiation.
DISCUSSION AND CONCLUSIONS
As the above survey has demonstrated, there is a substantial body of knowledge on the
effects upon microclimate of a wide variety of planning decisions. However, it is equally
ABER_CH_04 6/12/08 3:32 PM Page 113
evident that very little of this knowledge is being applied in practice, except in a
haphazard, piecemeal way.
Page (1968) identified three reasons why scientific information available to
researchers might, nonetheless, be rejected by practitioners:
1 The potential user considers the information irrelevant.
2 The user considers the information provided inapplicable in the form presented.
3 The user considers the information provided incomprehensible.
Page (1968) also noted that in order to be of use to urban planners, urban climatology
must first be predictive descriptive science is not in itself sufficient. Furthermore, the
problem is not to produce an idealized climatological plan, but to produce a workable
evolutionary plan that is economically viable and accepts that the planner must consider
other factors, such as the requirements of transportation systems.
Bitan (1988) noted that climatology affected urban design at a variety of spatial scales,
and that failure to address the effects at each of them, in turn, would necessarily restrict
the effectiveness of any intervention. However, in reality, there are very few projects
where climatologists take part in ongoing design processes over many years, beginning
with environmental impact assessments at the stage of site selection for large urban
developments and continuing throughout the entire planning process. The number and
hierarchy of authorities that are involved in the planning and regulatory approval of such
projects require a commitment that is rarely found in practice.
Oke (1984) suggested that what climatologists require is an ability to demonstrate the
importance of climatic information in the design of a settlement, and the predictive power
to foretell the climatic impact of alternative design strategies. According to Oke (1984),
several shortcomings in microclimate research at that time prevented it from becoming
an applied science, including a lack of quantitative techniques and relationships; lack of
standardization, generality and transferability; and the absence of clear guidelines for
those wishing to learn and use climatological principles in settlement planning.
Advances in addressing any of the following issues would represent real progress in
applying urban climatology in practical urban design:
What are the urban forms that will provide the best microclimate for any given
density, taking into consideration meso-scale climate, latitude, etc.?
Can street orientation be used as a policy guideline to achieve desirable microclimatic
conditions in a city?
Can the aspect ratio of streets be used as a policy guideline to achieve desirable
microclimatic conditions in a city? What are the quantitative limits that may be
usefully applied in realistic planning situations in high-density cities with regard to
specific goals of solar access and airflow?
What are the most appropriate neighbourhood layouts for every climate type? Are
some building typologies better at providing good microclimates than others?
What are the real effects of vegetation in various urban contexts? There seems to be
a consensus that plants improve the microclimate of cities; but numerous
114 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 114
quantitative studies have produced a wide range of estimates regarding the
magnitude and spatial extent of the effect of urban parks.
What are the city-scale modifications to microclimate that can be achieved by
implementing particular types of building and paving materials? The benefits to
individual buildings may be clear; but although models suggest that the compound
effect of enforcing their use all across an urban area may be substantial, the
magnitude of even local-scale effects is still open to some question.
Can urban climatology provide detailed site-specific weather data modified from
regional weather stations to allow more accurate simulation of building energy
performance, including not only air temperature, but also humidity and wind
characteristics?
The answers to these questions will probably only come about through research
collaboration between planners and urban climatologists.
Associate Professor Evyatar Erell: The Desert Architecture and Urban Planning Group, The Jacob Blaustein
Institutes for Desert Research, Ben-Gurion University of the Negev, Israel; tel +972-8-6596875; www.bgu.ac.il/
CDAUP; E-mail: erell@bgu.ac.il
NOTE
1 LT is the name Baker and Steemers (2000) gave to a computer model dealing with lighting and
thermal properties of buildings.
REFERENCES
Ahmed, K. S. (2000) Problems and prospects of bioclimatic urban design: An analytical study of shading and ventilation
needs of the urban spaces in the humid tropics, PLEA 2000: Architecture City Environment, Cambridge, UK
Aida, M. and K. Gotoh (1982) Urban albedo as a function of the urban structure a two-dimensional numerical simulation,
Boundary Layer Meteorology, vol 23, pp415424
Al-Sallal, K. A., M. M. AboulNaga, and A. M. Alteraifi (2001) Impact of urban spaces and building height on airflow
distribution: Wind tunnel testing of an urban setting prototype in Abu Dhabi City, Architectural Science Review vol 44,
no 3, pp227232
Ali-Toudert, F. and H. Mayer (2006) Numerical study on the effects of aspect ratio and orientation of an urban street canyon
on outdoor thermal comfort in hot and dry climate, Building and Environment, vol 41, pp94108
Arnfield, J. (2003) Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the
urban heat island, International Journal of Climatology, vol 23, pp126
Asaeda, T., V. T. Ca and A. Wake (1996) Heat storage of pavement and its effect on the lower atmosphere, Atmospheric
Environment, vol 30, no 3, pp413427
Assimakopoulos, D. N., C. Georgakis and M. Santamouris (2006) Experimental validation of a computational fluid dynamics
code to predict the wind speed in street canyons for passive cooling purposes, Solar Energy, vol 80, pp423434
ABER_CH_04 6/12/08 3:32 PM Page 115
Aynsley, R. and L. Gulson (1999) Microclimate and urban planning in the humid tropics, Planning in the Hothouse 27th
National Congress of the Royal Australian Planning Institute, Darwin, Australia
Baik, J. and J. Kim (1999) A numerical study of flow and pollutant dispersion characteristics in urban street canyons, Journal
of Applied Meteorology, vol 38, no 11, pp15761589
Baik, J., R, Park, H. Chun and J. Kim, J. (2000) A laboratory model of urban street-canyon flows, Journal of Applied
Meteorology, vol 39, no 9, pp15921600
Baker, N. and Steemers, K. (2000) Energy and Environment in Architecture: A Technical Design Guide, E. and F. Spon, London
Barring, L., J. O. Mattsson and S. Lindqvist (1985) Canyon geometry, street temperatures and urban heat island in Malmo,
Sweden, Journal of Climatology, vol 5, pp433444
Bitan, A. (1988) The methodology of applied climatology in planning and building, Energy and Buildings, vol 11, pp111
Bonan, G. B. (2000) The microclimates of a suburban Colorado (USA) landscape and implications for planning and design,
Landscape and Urban Planning, vol 49, pp97114
Borve, A. B. (1982) Settlement planning under Arctic conditions in northern Norway, Energy and Buildings, vol 4, pp6770
Bottema, M. (1999) Towards rules of thumb for wind comfort and air quality, Atmospheric Environment, vol 33,
pp40094017
Bourbia, F. and H. B. Awbi (2004) Building cluster and shading in urban canyon for hot dry climate, Part 2: Shading
simulations, Renewable Energy, vol 29, pp291301
Bozonnet, E., R. Belarbi and F. Allard (2005) Modelling solar effects on the heat and mass transfer in a street canyon, a
simplified approach, Solar Energy, vol 79, pp1024
Bretz, S. and H. Akbari (1997) Long-term performance of high-albedo roof coatings, Energy and Buildings, vol 25, pp159167
Bretz, S., H. Akbari and A. Rosenfeld (1998) Practical issues for using solar-reflective materials to mitigate urban heat
islands, Atmospheric Environment, vol 32, no 1, pp95101
Bruse, M. and H. Fleer (1998) Simulating surfaceplantair interactions inside urban environments with a three dimensional
numerical model, Environmental Modelling and Software, vol 13, pp373384
Ca, V. T., T. Asaeda and E. M. Abu (1998) Reductions in air conditioning energy caused by a nearby park, Energy and
Canton, M. A., J. L. Cortegoso and C. de Rosa (1994) Solar permeability of urban trees in cities of western Argentina,
Cermak, J. E. (1995) Physical modelling of flow and dispersion over urban areas, in Cermak, J. E., Davenport, A. G., Plate, E. J.
and Viegas, D. X. (eds) Wind Climate in Cities, Kluwer Academic Publishers, Dordrecht/Boston/London, pp383403
Chan, A., E. So and S. Samad (2001) Strategic guidelines for street canyon geometry to achieve sustainable street air
quality, Atmospheric Environment, vol 35, pp56815691
Chan, A., W. Au and E. So (2003) Strategic guidelines for street canyon geometry to achieve sustainable street air
quality part II: Multiple canopies and canyons, Atmospheric Environment, vol 37, pp27612772
Chow, W. and M. Roth (2006) Temporal dynamics of the urban heat island of Singapore, International Journal of Climatology,
vol 26, pp22432260
Dixon, N. S., J. W. D. Boddy, R. J. Smalley and A. S. Tomlin (2006) Evaluation of a turbulent flow and dispersion model in a
typical street canyon in York, UK, Atmospheric Environment, vol 40, pp958972
Dobre, A., S. Arnold, R. J. Smalley, J. W. D. Boddy, J. Barlow, A. S. Tomlin and S. Belcher (2005) Flow field measurements in
the proximity of an urban intersection in London, UK, Atmospheric Environment, vol 39, pp46474657
Doulos, L., M. Santamouris and I. Livada (2004) Passive cooling of outdoor spaces: The role of materials, Solar Energy,
vol 77, pp231249
El Nahas, M. M. (1996) Energy Saving Through Urban Design A Microclimatic Approach, PhD thesis, Department of
Architecture, University of Adelaide, Adelaide
116 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 116
El-Shakhs, S. (1994) Sadat City, Egypt, and the role of new town planning in the developing world, Journal of Architectural
and Planning Research, vol 11, no 3, pp239259
El-Sioufi, M. (1987) Urban Patterns for Improved Thermal Performance, PhD thesis, Department of Architecture, University of
Michigan, MI
Eliasson, I. (1996) Intra-urban nocturnal temperature differences: A multivariate approach, Climate Research, vol 7, pp2130
Eliasson, I. (2000) The use of climate knowledge in urban planning, Landscape and Urban Planning, vol 48, pp3144
Emmanuel, R. (1993) A hypothetical shadow umbrella for thermal comfort enhancement in the equatorial urban outdoors,
Architectural Science Review, vol 36, pp173184
Emmanuel, R. (1995) Energy efficient urban design guidelines for warm-humid cities: Strategies for Colombo, Sri Lanka,
Journal of Architectural and Planning Research, vol 12, no 1, pp5879
Emmanuel, R. (1997) Summertime urban heat island mitigation: Propositions based on an investigation of intra-urban air
temperature variations, Architectural Science Review, vol 40, no 4, pp155164
Erell, E. and T. Williamson (2006) Simulating air temperature in an urban street canyon in all weather conditions using
measured data from a reference meteorological station, International Journal of Climatology, vol 26, no 12,
pp16711694
Erell, E. and Williamson, T. (2007) Intra-urban differences in canopy layer air temperature at a mid-latitude city, International
Journal of Climatology, vol 27, pp12431255
Erell, E., V. Leal and E. Maldonado (2005) Measurement of air temperature in the presence of a large radiant flux: An
assessment of passively ventilated thermometer screens, Boundary Layer Meteorology, vol 114, pp205231
Etzion, Y. (1990) A desert solar neighbourhood in Sde-Boqer, Israel, Architectural Science Review, vol 33, pp105111
Evans, J. M. and S. de Schiller (19901991) Climate and urban planning: The example of the planning code for Vicente
Lopez, Buenos Aires, Energy and Buildings, vol 1516, pp3541
Giridharan, R., S. Ganesan and S. Lau (2004) Daytime urban heat island effect in high-rise and high-density residential
developments in Hong Kong, Energy and Buildings, vol 36, pp525534
Givoni, B. (1989) Urban Design in Different Climates, WMO/TD, no 346
Goh, K. and C. Chang (1999) The relationship between height to width ratios and the heat island intensity at 22:00 for
Singapore, International Journal of Climatology, vol 19, pp10111023
Golany, G. S. (1996) Urban design morphology and thermal performance, Atmospheric Environment, vol 30, no 3, pp455465
Gotz, L. (1982) Integration of climate in planning and building illustrated in a case of extreme climatic conditions, Energy and
Grimmond, C. S. B. (2006) Progress in measuring and observing the urban atmosphere, Theoretical and Applied Climatology,
vol 84, pp322
Grimmond, C. S. B. and T. R. Oke (1999) Aerodynamic properties of urban areas derived from analysis of surface form,
Journal of Applied Meteorology, vol 38, pp12621292
Grimmond, C. S. B., C. Souch and M. D. Hubble (1996) Influence of tree cover on summertime surface energy balance fluxes,
San Gabriel Valley, Los Angeles, Climate Research, vol 6, pp4557
Gupta, V. K. (1984) Solar radiation and urban design for hot climates, Environment and Planning B: Planning and Design,
vol 11, pp435454
Herz, R. K. (1988) Considering climatic factors for urban land-use planning in the Sahelian zone, Energy and Buildings, vol 11,
pp91102
Hotchkiss, R. S. and F. H. Harlow (1973) Air Pollution Transport in Street Canyons, US EPA Report (EPAR473029)
Howard, L. (1833) The Climate of London, Harvey and Darton, London
Hoyano, A. (1988) Climatological uses of plants for solar control and the effects on the thermal environment of a building,
ABER_CH_04 6/12/08 3:32 PM Page 117
Huang, Y. J., H. Akbari, H. Taha and A. H. Rosenfeld (1987) The potential of vegetation in reducing summer cooling loads in
residential buildings, Journal of Climate and Applied Meteorology, vol 26, pp11031116
Jauregi, E. (19901991) Influence of a large urban park on temperature and convective precipitation in a tropical city, Energy
and Buildings, vol 1516, pp457463
Kanda, M. and R. Moriwaki (1998) Environmental effect of Meiji-shrine forest as a sink of energy and pollutants, Second
Urban Climate Symposium, Albuquerque, New Mexico
Kastner-Klein, P. and E. J. Plate (1999) Wind-tunnel study of concentration fields in street canyons, Atmospheric
Kenworthy, A. T. (1985) Wind as an influential factor in the orientation of the orthogonal street grid, Building and
Environment, vol 20, no 1, pp3338
Kim, J. and J. Baik (1998) A numerical study of thermal effects on flow and pollutant dispersion in urban street canyons,
Second Urban Environment Symposium, Albuquerque, New Mexico
Kim, J. and J. Baik (2001) Urban street canyon flows with bottom heating, Atmospheric Environment, vol 35, pp33953404
Kim, J. and J. Baik (2004) A numerical study of the effects of ambient wind direction on flow and dispersion in urban street
canyons using the RNG k- turbulence model, Atmospheric Environment, vol 38, pp30393048
Kjelgren, R. and T. Montague (1998) Urban tree transpiration over turf and asphalt surfaces, Atmospheric Environment,
vol 32, no 1, pp3541
Knowles, R. (1981) Sun, Rhythm, Form, The MIT Press, Cambridge, MA
Knowles, R. (2003) The solar envelope: Its meaning for energy and buildings, Energy and Buildings, vol 35, pp1525
Kolokotroni, M., I. Giannitsaris and R. Watkins (2006) The effect of the London urban heat island on building summer cooling
demand and night ventilation strategies, Solar Energy, vol 80, pp383392
Krishan, A. (1996) The habitat of two deserts in India: Hot-dry desert of Jaisalmer (Rajasthan) and the cold-dry high altitude
mountainous desert of Leh (Ladakh), Energy and Buildings, vol 23, no 3, pp217229
Landsberg, H. E. (1968) Climates and urban planning, Symposium on Urban Climates and Building Climatology, Brussels,
Belgium
Levinson, R., H. Akbari and J. Reilly (2007) Cooler tile-roofed buildings with near-infrared reflective non-white coatings,
Building and Environment, vol 42, pp25912605
Littlefair, P. (1998) Passive solar urban design: Ensuring the penetration of solar energy into the city, Renewable and
Sustainable Energy Reviews, vol 2, pp303326
Livada, I., M. Santamouris, K. Niachou, N. Papanikolaou and G. Mihalakakou (2002) Determination of places in the great
Athens area where the heat island effect is observed, Theoretical and Applied Climatology, vol 71, pp219230
Manins, P. C., M. E. Cope, P. J. Hurley, P. W. Newton, N. C. Smith and M. Lo (1998) The impact of urban development on air
quality and energy use, 14th International Clean Air & Environment Conference, Melbourne, Australia, 1822 October 1998
Mazouz, S. and M. S. Zerouala (1998) Shading as a modulator for the design of urban layouts based on vernacular
experiences, Energy and Buildings, vol 29, pp1115
McPherson, E. G., L. P. Herrington and G. M. Heisler (1988) Impacts of vegetation on residential heating and cooling, Energy
and Buildings, vol 12, pp4151
Meier, A. K. (19901991) Strategic landscaping and air-conditioning savings: A literature review, Energy and Buildings,
vol 1516, pp479486
Mills, G. (1997) The radiative effects of building groups on single structures, Energy and Buildings, vol 25, pp5161
Mills, G. (1999) Urban climatology and urban design, 15th International Congress of Biometeorology and International
Conference on Urban Climatology, Sydney, Australia
Mills, G. (2006) Progress toward sustainable settlements: A role for urban climatology, Theoretical and Applied Climatology,
vol 84, pp6976
118 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 118
Nakamura, Y. and T. R. Oke (1988) Wind, temperature and stability conditions in an eastwest oriented urban canyon,
Atmospheric Environment, vol 22, no 12, pp26912700
Niachou, A., K. Papakonstantinou, M. Santamouris, A. Tsangrassoulis and G. Mihalakakou (2001) Analysis of the green roof
thermal properties and investigation of its energy performance, Energy and Buildings, vol 33, pp719729
Nunez, M. and T. R. Oke (1977) The energy balance of an urban canyon, Journal of Applied Meteorology, vol 16, pp1119
Oke, T. R. (1973) City size and the urban heat island, Atmospheric Environment, vol 7, no 8, pp769779
Oke, T. R. (1981) Canyon geometry and the nocturnal urban heat island: Comparison of scale model and field observations,
Journal of Climatology, vol 1, no 3, pp237254
Oke, T. R. (1982) The energetic basis of the urban heat island, Quarterly Journal of the Royal Meteorological Society, vol 108,
no 455, pp124
Oke, T. R. (1984) Towards a prescription for the greater use of climatic principles in settlement planning, Energy and
Oke, T. R. (1987) Boundary Layer Climates, Methuen, London and New York
Oke, T. R. (1988a) Street design and urban canopy layer climate, Energy and Buildings, vol 11, pp103113
Oke, T. R. (1988b) The urban energy balance, Progress in Physical Geography, vol 12, no 4, pp471508
Oke, T. R. (1989) The micrometeorology of the urban forest, Philosophical Transactions of the Royal Society London B,
vol 324, pp335349
Olgyay, V. (1963) Design with Climate, Princeton University Press, Princeton, NJ
Onmura, S., M. Matsumoto and S. Hokoi (2001) Study on the evaporative cooling effect of roof lawn gardens, Energy and
Page, J. K. (1968) The fundamental problems of building climatology considered from the point of view of decision-making by
the architect and urban designer, Symposium on Urban Climates and Building Climatology, Brussels, Belgium
Page, J. K. (1972) The problem of forecasting the properties of the built environment from the climatological properties of
the green-field site, in Taylor, J. A. (ed) Weather Forecasting for Agriculture and Industry, David & Charles, London
Palomo Del Barrio, E. (1998) Analysis of the green roofs cooling potential in buildings, Energy and Buildings, vol 27,
pp179193
Papadakis, G., P. Tsamis and S. Kyritsis (2001) An experimental investigation of the effect of shading with plants for solar
control of buildings, Energy and Buildings, vol 33, pp831836
Pearlmutter, D., A. Bitan and P. Berliner (1999) Microclimatic analysis of compact urban canyons in an arid zone,
Pearlmutter, D., P. Berliner and E. Shaviv (2006) Physical modeling of the pedestrian energy exchange within the urban
canopy, Building and Environment, vol 41, pp783795
Pearlmutter, D., P. Berliner and E. Shaviv (2007) Integrated modeling of pedestrian energy exchange and thermal comfort in
urban street canyons, Building and Environment, vol 42, pp23962409
Potchter, O. (1988) Climatic aspects in rural settlement development in hot, arid zones: A case study of the Central Jordan
Valley, Energy and Buildings, vol 11, pp7390
Pressman, N. (1996) Sustainable winter cities: Future directions for planning, policy and design, Atmospheric Environment,
vol 30, no 3, pp521529
Rafailidis, S. (1997) Influence of building areal density and roof shape on the wind characteristics above a town, Boundary
Layer Meteorology, vol 85, pp225271
Ratti, C. and P. Richens (1999) Urban texture analysis with image processing techniques, Computers in Building: Proceedings
of the CAADfutures 1999 Conference, Atlanta, Georgia, 78 June 1999
Ratti, C. and P. Richens (2004) Raster analysis of urban form, Environment and Planning B: Planning and Design, vol 31,
pp297309
ABER_CH_04 6/12/08 3:32 PM Page 119
Ratti, C., D. Raydan and K. Steemers (2003) Building form and environmental performance: Archetypes, analysis and an arid
climate, Energy and Buildings, vol 35, pp4959
Ratti, C., N. Baker and K. Steemers (2005) Energy consumption and urban texture, Energy and Buildings, vol 37,
pp762776
Ratti, C., S. Di Sabatino and R. Britter (2006) Urban texture analysis with image processing techniques: Winds and
dispersion, Theoretical and Applied Climatology, vol 84, pp7790
Raza, S. H., M. S. R. Murthy, O. B. Lakshmi and G. Shylaja (19901991) Effect of vegetation on urban climate and healthy
urban colonies, Energy and Buildings, vol 1516, pp487491
Robinson, D. and A. Stone (2004) Solar radiation modelling in the urban context, Solar Energy, vol 77, pp295309
Runnalls, K. and T. R. Oke (1998) The urban heat island of Vancouver, BC, Second Urban Environment Symposium,
Albuquerque, New Mexico, 26 November 1998
Runnalls, K. and T. R. Oke (2000) Dynamics and controls of the near-surface heat island of Vancouver, British Columbia,
Physical Geography, vol 21, no 4, pp283304
Sailor, D. J. (1998) Simulations of annual degree day impacts of urban vegetative augmentation, Atmospheric Environment,
vol 32, no 1, pp4352
Sailor, D. and H. Fan (2002) Modeling the diurnal variability of effective albedo for cities, Atmospheric Environment, vol 36,
pp713725
Saito, I., O. Ishihara and T. Katayama (19901991) Study of the effect of green areas on the thermal environment in an urban
area, Energy and Buildings, vol 1516, pp493498
Santamouris, M. (1998) The Athens urban climate experiment, PLEA 1998: Environmentally Friendly Cities, Lisbon, Portugal,
14 June 1998
Santamouris, M. (ed) (2001) Energy and Climate in the Urban Built Environment, James & James, London
Santamouris, M., N. Papanikolaou, I. Livada, I. Koronakis, C. Georgakis, A. Argiriou and D. N. Assimakopoulos (2001) On the
impact of urban climate on the energy consumption of buildings, Solar Energy, vol 70, no 3, pp201216
Schiller, G. and R. Karschon (1974) Microclimate and recreational value of tree plantings in deserts, Landscape Planning,
vol 1, pp329337
Shashua-Bar, L. and M. Hoffman (2000) Vegetation as a climatic component in the design of an urban street: An empirical
model for predicting the cooling effect of urban green areas with trees, Energy and Buildings, vol 31, no 3, pp221235
Shashua-Bar, L. and M. Hoffman (2002) The Green CTTC model for predicting the air temperature in small urban wooded
sites, Building and Environment, vol 37, pp12791288
Shashua-Bar, L. and M. Hoffman (2003) Geometry and orientation aspects in passive cooling of canyon streets with trees,
Shashua-Bar, L. and M. Hoffman (2004) Quantitative evaluation of passive cooling of the UCL microclimate in hot regions in
summer, case study: Urban streets and courtyards with trees, Building and Environment, vol 39, pp10871099
Simpson, J. R. (2002) Improved estimates of tree-shade effects on residential energy use, Energy and Buildings, vol 34,
no 10, pp10671076
Simpson, J. R. and E. G. McPherson (1998) Simulation of tree shade impacts on residential energy use for space conditioning
in Sacramento, Atmospheric Environment, vol 32, no 1, pp6974
So, E., A. Chan and A. Wong (2005) Large-eddy simulations of wind flow and pollutant dispersion in a street canyon,
Soler, M. R. and C. Ruiz (1994) Urban albedo derived from direct measurements and LANDSAT 4 TM satellite data,
International Journal of Climatology, vol 14, pp925931
Spronken-Smith, R. A. and T. R. Oke (1998) The thermal regime of urban parks in two cities with different summer climates,
International Journal of Remote Sensing, vol 19, no 11, pp20852104
120 EVYATAR ERELL
ABER_CH_04 6/12/08 3:32 PM Page 120
Spronken-Smith, R. and T. R. Oke (1999) Scale modelling of nocturnal cooling in urban parks, Boundary Layer Meteorology,
vol 93, pp287312
Steemers, K., N. Baker, D. Crowther, J. Dubiel, M.-H. Nikolopoulou and C. Ratti (1997) City texture and microclimate, Urban
Design Studies, vol 3, pp2550
Steinecke, K. (1999) Urban climatological studies in the Reykjavik subarctic environment, Iceland, Atmospheric Environment,
vol 33, pp41574162
Swaid, H. (1992) Intelligent urban forms (IUF): A new climate-concerned urban planning strategy, Theoretical and Applied
Climatology, vol 46, pp179191
Synnefa, A., M. Santamouris and I. Livada (2006) A study of the thermal performance of reflective coatings for the urban
environment, Solar Energy, vol 80, pp968981
Synnefa, A., M. Santamouris and K. Apostolakis (2007) On the development, optical properties and thermal performance of
cool colored coatings for the urban environment, Solar Energy, vol 81, no 4, pp488497
Taha, H. G. (1978) An Urban Micro-Climate Model for Site-Specific Building Energy Simulation: Boundary Layers, Urban
Canyon and Building Conditions, PhD thesis, Department of Architecture, University of California, Berkeley, CA
Taha, H. (1997) Urban climates and heat islands: Albedo, evapotranspiration and anthropogenic heat, Energy and Buildings,
vol 25, pp99103
Taha, H., H. Akbari, A. Rosenfeld and J. Huang (1988) Residential cooling loads and the urban heat island the effects of
albedo, Building and Environment, vol 23, no 4, pp271283
Taha, H., S. Douglas and J. Haney (1997) Mesoscale meteorological and air quality impacts of increased urban albedo and
vegetation, Energy and Buildings, vol 25, pp169177
Tsangrassoulis, A. and M. Santamouris (2003) Numerical estimation of street canyon albedo consisting of vertical coated
glazed facades, Energy and Buildings, vol 35, pp527531
Upmanis, H. and D. Chen (1999) Influence of geographical factors and meteorological influences on nocturnal urban-park
temperature differences a case study of summer 1995 in Goteborg, Sweden, Climate Research, vol 13, pp125139
Upmanis, H., I. Eliasson and S. Lindqvist (1998) The influence of green areas on nocturnal temperatures in a high latitude city
(Goteborg, Sweden), International Journal of Climatology, vol 18, pp681700
Urano, A., Y. Morikawa and M. Nishimura (1995) Urban environment analysis of the Kanto Plain with 1-dimensional and
3-dimensional numerical simulations, Pan Pacific Symposium on Building and Urban Environmental Conditioning in Asia,
Nagoya, Japan, March 1995
Vitruvius (1999) Ten Books on Architecture, Cambridge University Press, Cambridge, UK
Westerberg, U. and M. Glaumann (1990/1991) Design criteria for solar access and wind shelter in the outdoor environment,
Wong, N. H., Y. Chen, C. L. Ong and A. Sia (2003) Investigation of the thermal benefits of rooftop garden in the tropical
environment, Building and Environment, vol 38, pp261270
Xie, X., Z. Huang, J. Wang and Z. Xie (2005) Thermal effects on vehicle emission dispersion in an urban street,
Transportation Research Part D, vol 10, pp197212
Xie, X., C. Liu, D. Leung and M. Leung (2006) Characteristics of air exchange in a street canyon with ground heating,
Yamartino, R. J. and G. Wiegand (1986) Development and evaluation of simple models for the flow, turbulence and pollutant
concentration fields within an urban street canyon, Atmospheric Environment, vol 20, pp21372156
Yamashita, S., K. Sekine, M. Shoda, K. Yamashita and Y. Hara (1986) On relationships between heat island and sky view
factor in the cities of Tama River basin, Japan, Atmospheric Environment, vol 20, no 4, pp681686
Yezioro, A., I. G. Capeluto and E. Shaviv (2006) Design guidelines for appropriate insolation of urban squares, Renewable
Energy, vol 31, pp10111023
ABER_CH_04 6/12/08 3:32 PM Page 121
ABER_CH_04 6/12/08 3:32 PM Page 122
Keywords cooling; solar systems; urban environment; SWOT analysis
INTRODUCTION
Solar-assisted cooling appears to be a promising alternative to conventional electrically
driven air-conditioning units from an environmental point of view since it results in
decreased CO
2
emissions and, in the case of prevailing solar-cooling technologies, the
elimination of chlorofluorocarbons (CFCs) and hydro-chlorofluorocarbons (HCFCs). In
addition, SAC can contribute significantly to dealing with continuously increasing peak
electricity loads during summer, which burden both the generation and transfer capacities
of electricity utility companies, as well as consumers. The heat-driven, solar-assisted
chillers seem, in this sense, likely to have an excellent potential in the space
Abstract
The close relation between maximum insolation, and a buildings cooling loads and peak
electricity demand indicates that solar-assisted cooling (SAC) is an interesting method of
successfully handling this issue. Furthermore, from an environmental viewpoint, SAC
arises as a promising alternative to conventional, electrically driven air conditioning since
it results in decreased carbon dioxide (CO
2
) emissions and does not use of ozone-
depleting or greenhouse gas (GHG) refrigerants. The utilization of solar energy in air
conditioning may in this sense be the challenge for the next major growth of the solar
thermal market. Nevertheless, there is a long way to go, and SAC technologies must
prove that they are competitive. This chapter discusses the principles and the state of the
art of SAC. Technical and non-technical advantages and barriers are also presented, as
well as the possibility of using active SAC technologies, and technical issues raised by
the use of electrically driven conventional air-conditioning systems are identified. Finally,
by utilizing strengths, weaknesses, opportunities, threats (SWOT) analysis, conclusions
are drawn with respect to measures and strategies for enhancing SAC technologies.
5
Solar Air Conditioning:
A Review of Technological
and Market Perspectives
S. Oxizidis and A. M. Papadopoulos
ABER_CH_05 6/12/08 3:32 PM Page 123
air-conditioning businesses, although so far they are not really a competitive alternative,
in market terms, to conventional vapour compression cooling (VCC) systems.
Increased environmental concern and sensitivity during recent years, combined with
technological development in solar energy use and continuous urbanization, have given rise to
new priorities for the building sector (Santamouris, 2001). The rapid industrialization and
urbanization of the second half of the 20th century led to a dramatic increase in the size and
density of cities worldwide, leading to a corresponding upsurge in energy consumption. Thus,
energy efficiency within buildings must take into account the particular requirements and
constraints of urban buildings themselves, as well as the urban microclimate and environment.
In this sense, there are two major concerns that occupy the interest of researchers
and decision-makers: first, the urban environment that alters the microclimate, putting
124 S. OXIZIDIS AND A. M. PAPADOPOULOS
NOMENCLATURE
Q heat (kJ)
T temperature (C)
H enthalpy (kJ)
S entropy (kJ/K)
W work (kJ)
COP coefficient of performance
c
p
specific heat (kJ/kgK)
GREEK
change in quantity
efficiency
SUBSCRIPTS
des desorption
sor sorption
ev evaporation
con condenser
dec decomposition
syn synthesis
M medium
L low
H high
g gas
c complex
l liquid
s solid
HE heat engine
HP heat pump
R refrigerator
ABER_CH_05 6/12/08 3:32 PM Page 124
a heavy burden on cooling loads and increasing demand for air conditioning in buildings,
which is reflected in the sale of air conditioners; and, second, the fact that this burden can
only marginally be covered by existing generation plants or the capacity of
transport/distribution electricity grids. There are several reasons for the boost in the
cooling demand of urban buildings:
Climatic aspects:
Urban heat island effect. Higher air temperatures experienced in cities due to the
urban heat island effect are a significant factor of increased demand for cooling
(Papadopoulos, 2001; Santamouris et al, 2001).
Urban microclimate, which reduces the potential of natural cooling. In densely
built cities with buildings that create street canyons, wind velocities are very low
and vortices lead to the entrapment of hot air masses, thus reducing natural
cooling options (Alvarez and Molina, 2003; Geros et al, 2005).
Global warming. The major feature of climate change is higher temperatures due
to global warming. In addition, the frequency and intensity of heatwaves are
expected to increase (Cartalis et al, 2001).
Architectural aspects:
Buildings oriented to maximize solar gains. After the first oil crisis of the
1970s, efforts focused on the construction of buildings with low energy
consumption for heating, disregarding the effect this would have on cooling
loads.
Current trend in modern architecture. The vast majority of office buildings
have been, and are being, constructed by making extended use of transparent
surfaces, with the consequence that interior solar loads are enormous
(Cheung et al, 2005).
Occupancy and regulatory trends:
Continuously increasing capacities of electric equipment used in buildings. The
excessive use of electric and electronic office and house equipment has a
proportional effect to internal energy gains (Katsakos, 2007).
Rise of living standards increase in thermal comfort demands. Todays
buildings occupants are placing higher demands on thermal comfort, a fact that,
combined with stricter regulations for indoor air quality (IAQ), increases the
demand for ventilation and, hence, for cooling fresh air (Chatzidimoula, 2004;
Avgelis et al, 2005).
Change of work schedules. Traditionally, in areas with hot summers such as the
Mediterranean belt, there used to be a break during noon and work continued in
the afternoon. Lately, even in these countries the working hours follow the
Western and Northern European profile of a continuous schedule.
Availability of HVAC systems:
Diminishing prices for air conditioners. The price of room air conditioners has
fallen dramatically, while cheap air conditioners introduced to the market by
lesser-known manufactures present low performance coefficients (BSRIA, 2000;
Dethman, 2004).
Solar Air Conditioning: A Review 125
ABER_CH_05 6/12/08 3:32 PM Page 125
This increased demand for cooling results in a series of problems, which can be divided
into problems of global and regional/local scale:
Global scale:
Conventional refrigerants present a high ozone-depletion potential (ODP) and
global warming impact (GWI). The hydrofluorocarbon (HFC) and polyfluorocarbon
(PFC) refrigerants that replaced HCFCs and CFCs in new chillers may be less
drastic, but are still GHGs and are therefore considered responsible for climate
change.
Increased primary energy consumption leads to higher emissions of pollutants
from power plants.
Regional/local scale:
Summer peak electricity demand cannot be covered by installed plant capacity
due to the high cost of new plants and transmission and distribution networks,
which private power utilities in the liberalized electricity market are less and less
willing to undertake. In addition, environmental restrictions (e.g. the Kyoto Protocol)
do not allow for the installation of large fossil fuel-driven generation plants.
Any active cooling technology electrically or thermally driven rejects high
quantities of heat enforcing the urban heat island (UHI), therefore adding to the
vicious cycle.
From the above, one can deduce that refrigeration is one of the major technological
challenges that the developed world has to face. In order for this to happen, a reduction
in demand is almost mandatory since on the production side (electricity generation) the
options are limited. This will lead to a replacement of a vast quantity of vapour
compression by alternative cooling technologies. There are several solutions that can be
considered as technically viable alternatives, the most important of which are:
Methods that deal with the issue of increased primary energy consumption:
energy-conscious design of buildings, featuring sun protection, enhanced thermal
insulation and utilizing the thermal storage properties of the buildings envelope;
applying low-energy cooling techniques, such as evaporative, radiative and
ground cooling and ventilation;
active cooling techniques (mainly solar technologies).
Alternative, but still conventional, technologies and economies of scale:
cool storage in the building;
district cooling systems;
distributed energy generation systems, preferably combined heat and power
systems;
natural gas-fired cooling systems.
All of these technologies are suitable to address the problem of reducing high peak
electricity loads; but the only ones with a true positive environmental impact are the active
solar and the low-energy cooling technologies, which result in a lower primary energy
ABER_CH_05 6/12/08 3:32 PM Page 126
ratio (PER) for the cooling energy to be delivered to the building. However, in an urban
context, it is very difficult to utilize low-energy cooling technologies due to the lack of
appropriate heat sinks and the constraints imposed by the built environment (e.g. heavily
built sites and canyon geometry).
THERMALLY DRIVEN ACTIVE SOLAR COOLING TECHNOLOGIES
A heat pump (see Figure 5.1) can operate either by consuming electricity (with an electric
compressor) or fossil fuel (with a gas engine) during the production and use of mechanical
work (through compression) either through vapour compression heat pumps or by
consuming thermal energy through thermally driven heat pumps.
Thermally driven pumps differ fundamentally from the vapour compression heat
pumps because they convert heat of a given temperature to heat of another temperature
without any intermediate use of mechanical work through an electric compressor or gas
engine. The term thermal compression is often used to point out the feature of
compression by mechanical means.
Thermally driven heat pumps utilize sorption phenomenon. Sorption heat pumps can be
operated with different working pairs (e.g. lithium bromide/water, zeolite/water, salt/
ammonia or metal/hydrogen) and employ different operating principles, which may be
absorption, adsorption or chemisorption (chemical reaction). Although the physical
principles of these systems may vary, all sorption systems can be evaluated in a similar way.
Sorption cooling technologies are the prevailing option when using solar energy for
air-conditioning purposes.
Figure 5.2 provides a schematic representation of a simple ideal sorption heat pump.
The system is driven by heat input Q
des
(heat of desorption or heat of generating vapour)
at temperature T
des
(high temperature), releases heat Q
sor
(heat of sorption) at T
sor
(medium
Heat Sink
T
High
Temperature
Heat Sink
T
Medium
Temperature
Heat Sink
T
Mechanical
Energy
W
Heat Source
T
Heat Source
T
Q
M
Q
H
Q
C
Q
Heat Pump
Thermal
Energy
Q
H
Q
C
Heat Pump
(a) (b)
FIGURE 5.1 Operational principles of heat pumps: (a) mechanically driven; (b) thermally driven
ABER_CH_05 6/12/08 3:32 PM Page 127
temperature) and Q
con
at T
con
(medium temperature), and produces cold Q
ev
at T
ev
(low
temperature). As a result, the temperature profile of the process is T
ev
< T
con
< T
sor
< T
des
,
though often when the heat pump works on a cooling mode the extraction of heat during
sorption and condensation is often assumed to take place under the same temperature
(T
con
= T
sor
= T
m
) because for both processes, the heat is transferred to the same
environmental heat sink.
CLASSIFICATION OF SORPTION HEAT PUMPS
In their capacity of driving heat pumps, sorption phenomena can be categorized as gas
absorption by a liquid, gas adsorption on a micro-porous solid, or chemical reaction
between a gas and a solid. The most important difference between these three processes
concerns the nature of the process itself.
Adsorption involves the use of solids for removing substances from gaseous
solutions. The process of adsorption concerns separation of a substance from one phase,
accompanied by its accumulation or concentration at the surface of another. On the other
hand, absorption is the process in which material transferred from one phase to another
(e.g. liquid) infiltrates the second phase to form a solution.
In the case of adsorption, physical adsorption on the surface occurs without modifying
the solid skeleton itself. At the surface of solids, unsaturated and relatively weak
intermolecular forces of the van der Waals type, as well as electrostatic forces, are in play.
Hence, when a solid is exposed to a gas, the gas molecules will form bonds with it and
become attached though not chemically bonded. The inverse process the release of
gas molecules from the solid surface is termed desorption. When the gas molecules
form a chemical bond with the surface of the solid, the phenomenon is called
chemisorption (Schawe, 2001). In this case, chemical changes occur that induce
modification of the solid itself. This difference is a result of the variance of the equilibrium
of each process (see Figure 5.3):
HIGH TEMPERATURE HEAT SOURCE
(SOLAR COLLECTORS)
MEDIUM
TEMPERATURE
HEAT SINK
(AMBIENT)
Q
con
T
con
T
des
T
sor
T
ev
Q
ev
Q
des
Q
sor
Q
sor
T
con
T
des
T
sor
T
ev
Temperature of condensation
Desorption temperature
Sorption temperature
Evaporation temperature Q
ev
Q
des
Q
con
Heat of sorption
Heat of evaporation
Heat of desorption
Heat of condensation
CONDENSER
DESORBER SORBER
EVAPORATOR
COOLING
WATER
MEDIUM
TEMPERATURE
HEAT SINK
(AMBIENT)
THERMAL COMPRESSOR
CHILLED
WATER
LOW TEMPERATURE HEAT SOURCE
(BUILDING COOLING SYSTEM)
FIGURE 5.2 Schematic representation of sorption refrigerator
ABER_CH_05 6/12/08 3:32 PM Page 128
Absorption is bi-variant and induces no volume modification.
Adsorption is also bi-variant and induces no volume modification of the solid.
Chemisorption is mono-variant and induces volume modification of the solid.
When using the sorption phenomena (heating, cooling and heat storage), a whole range
of refrigerant fluids is available as working pairs (sorbate/sorbent) that can be useful for a
wide range of applications, from cryogenics to high-temperature heat storage. The pairs
of interest in the field of sorption for building applications are:
adsorption: water/zeolite, water/silica gel, methanol/activated carbon and
ammonia/activated carbon;
absorption: water/lithium bromide and ammonia/water;
chemical reaction: ammonia/inorganic salts.

Sorbate Concentration
refrigerant vapour (absorbate or
absorbate) / liquid (absorbent) or
solid (absorbent) desiccant
isosteric network
refrigerant
(sorbate) liquid /
vapor
equilibrium line
refrigerant
(sorbate) liquid /
vapor
equilibrium line
Q
dec Q
con
Q
ev
Q
syn
T
dec
T
syn
T
con
T
ev
T
gen
T
T
(a)
(b)
P
P
T
con
T
ev
T
abs
/T
abs
P
con
P
ev
P
con
P
ev
reaction between sorbent sorbate
equilibrium line
Q
gen
Q
abs
Q
con
Q
ev
FIGURE 5.3 Basic cycle for adsorption and absorption in a (lnP, 1/T) diagram: (a) closed systems; (b)
chemical reaction systems
ABER_CH_05 6/12/08 3:32 PM Page 129
All three processes can be utilized in open systems (desiccant cooling), where the
elements comprise the sorbent, the sorbate (usually water) and air for heat and mass
transport, while in closed systems, there are the sorbent and the sorbate (usually water)
in an evacuated system. In open systems, air is the heat and water vapour the
transporting fluid, while in closed evacuated systems no fluid supports the heat transfer
and either special heat exchangers have to be installed or the heat exchangers have to be
coated by a sorption material.
Absorption heat pumps
The two main types of absorption systems with respect to the working fluid are lithium
bromide (LiBr) (absorber)/water (refrigerant) and water (absorber)/ammonia (NH
3
)
(refrigerant).
Since both systems use water, their application range is limited to temperatures above
4C, which is not a problem with regard to air-conditioning applications. On the other hand,
concerning the advantages of LiBr/water over water/ammonia systems, the LiBr/water
solution is a non-volatile absorbent of LiBr, which eliminates the need for a rectifier, such
as in the water/NH
3
system. In addition, water, as a refrigerant, has a high vaporization heat.
Furthermore, LiBr/water systems present a lower generator operating temperature, which
makes them more suitable for solar utilization and, in principle, achieves a higher
coefficient of performance (COP), compared to the water/ammonia systems.
Two rather minor problems for the LiBr/water solution arise as a result of its proclivity
for crystallization this problem has substantially been reduced through the use of
microprocessor and direct digital control (DDC) (Dorgan et al, 1995) and its
corrosiveness to some metals. Several additives can be added to the solution as corrosion
inhibitors or in order to improve the heat mass-transfer performance (Srikhirim et al, 2001).
LiBr/water systems are categorized by the number of times that the solution is heated
to produce refrigerant vapours. This is referred to as the number of effects. A single-effect
system uses the heat input once, while the double-effect system uses the heat input for
one desorption effect and uses the warm refrigerant vapours as the heat source for the
second effect (Dorgan et al, 1995). Triple-effect systems are under research; but no triple-
effect machine is yet available commercially.
The water/ammonia solution has been widely used since the invention of the
absorption cycle systems for both cooling and heating purposes, although after the
introduction of the LiBr/water solution, its use became limited to industrial applications for
refrigeration purposes. Both NH
3
(refrigerant) and water (absorbent) are highly stable over
a wide range of operating temperatures and pressures. Ammonia has a high latent heat
of vaporization, which is necessary for the efficient performance of the system. It can be
used for low-temperature applications (as low as 50C) as the freezing point of NH
3
is
77C. Since both NH
3
and water are volatile, the cycle requires a rectifier to strip away
water that normally evaporates with NH
3
. Without a rectifier, the water will accumulate in
the evaporator and reduce the systems performance, since it will freeze in the low
operating temperatures. There are other disadvantages such as its high pressure, toxicity
and corrosive action to copper and copper alloy. However, the water/NH
3
system is
environmentally friendly and inexpensive (Srikhirim et al, 2001).
ABER_CH_05 6/12/08 3:32 PM Page 130
Water/ammonia systems have several disadvantages compared to LiBr/water
systems: their COP, for instance, is lower. Since they require a higher generator inlet
temperature, they operate at a 10 to 15 per cent lower solar fraction than LiBr/water
systems. Furthermore, they require higher pressures, resulting in higher pumping power.
A more complex system requiring a rectifier to separate ammonia and water vapour at the
generator outlet is necessary, otherwise captured water will evaporate within the
absorption cycle, affecting temperature and pressure. Finally, there are restrictions on
in-building applications of water/ammonia cooling units because of the toxicity hazards
associated with the use of ammonia (Li and Sumathy, 2000).
A common classification for the water/ammonia systems refers to the number of
evaporator stages involved. The number of stages describes the number of basic absorption
cycles that are combined to create a particular configuration. The number of effects,
furthermore, describes how often a major portion of the driving heat is reused within the
system in order to provide cooling. As a rule, increasing the number of effects increases
efficiency, and increasing the number of stages requires more hardware. However, for a
given number of stages, there are only limited numbers of effects possible. For a cycle with
n stages, the maximum number of effects is 2
n
1 (Cheung et al, 1996). The most common
systems are the single-stage ones, which feature one stage of evaporation/absorption, and
the double-stage ones with two stages of evaporation/ absorption.
Adsorption heat pumps
The main adsorbents used for air-conditioning purposes are silica gel and zeolite, which
use water as a refrigerant, while the pair-activated carbon/methanol, which has been used
since 1930, can be employed for refrigeration purposes as well since methanol can be
cooled below 0C, although methanols latent heat of vaporization is smaller than waters.
On the other hand, zeolite/water systems demand generation temperatures of 170
o
C,
while the silica gel/water and, to a lesser degree, the active carbon/methanol systems can
utilize heat in temperatures below 100C. Therefore, the most common pair for air-
conditioning applications is the silica gel/water system.
When fixed adsorbent beds are employed, which is the common practice, these
cycles can be operational without moving parts other than magnetic valves. This results
in low vibrations, mechanical simplicity, high reliability and availability and, consequently,
a longer lifetime. The use of fixed beds also results in an intermittent cycle operation, with
adsorbent beds changing between adsorption and desorption stages. Hence, if a
constant flow of vapour from the evaporator is required in order to provide continuous
cooling, two or more adsorbent beds must be operated out of phase (Critoph, 1999).
Compared to absorption, adsorption chillers are more expensive and their commercial
availability is still limited. In principle, their COP is lower than those of absorption chillers;
but they can utilize lower heat temperatures and thus can be driven by flat-plate solar
collectors more efficiently.
Chemical reaction (thermochemical) heat pumps (TCHPs)
In general, the operational principle of a chemical reaction heat pump is quite similar to
that of an adsorption heat pump. The temperature profile of a thermochemical heat
ABER_CH_05 6/12/08 3:32 PM Page 131
pumps complete cycle is depicted in Figure 5.4, along with the heat fluxes at every step
of the cycle:
1 At high temperature T
dec
, the decomposition (endothermic reaction) of the solid
material S.nG takes place by utilizing high-temperature heat. The reaction leads to
the products S.(n-m)G (solid or liquid) and MG (gas).
2 After the products are separated, the solid loses sensible heat to the environment.
3 Concurrently with step 2, the gas loses sensible heat to the environment.
4 The gas condenses at temperature T
con
, releasing heat either to the environmental
heat sink or to useful heating applications.
5 The forming liquid loses more sensible heat to the environment.
6 The liquid extracts heat from the environment at temperature T
ev
to evaporate and
increase its enthalpy.
7 The formed gas increases its enthalpy through the addition of more sensible heat.
8 The products of the initial reaction are brought together so that the reverse
exothermic reaction can take place, delivering the useful heat (stored as chemical
energy) at medium temperature T
syn
.
9 The addition of heat to the solid brings it back to initial conditions, ready to
decompose and repeat the cycle.
FIGURE 5.4 Temperature profile of a TCHP cycle
ABER_CH_05 6/12/08 3:32 PM Page 132
THERMODYNAMICS OF SORPTION HEAT PUMPS
As already mentioned, the temperature of sorption and condensation can be the same;
thus, T
con
= T
sor
= T
m
. Adopting this assumption and in relation to the phase diagram of
Figure 5.3, an enthalpytemperature (HT) relation is depicted in Figure 5.5.
In Figure 5.5, the dotted line corresponds to the working cycle of the refrigerant, along
with the following processes:
(a) Desorption of the complex: the heat of desorption H
H(c)
is supplied by the high-
temperature heat source (solar collectors).
Source: adjusted from Benach (1984)
FIGURE 5.5 Working media paths in a sorption system
ABER_CH_05 6/12/08 3:32 PM Page 133
(b) Cooling of sorbent and gaseous working fluid: sensible heat [c
p(s)
(T
H
T
M
) +
c
p(g)
(T
H
T
M
)] is released to ambient air.
(c) Sorption of the complex: the heat of sorption H
M(c)
is disposed to ambient air.
(d) Heating of the complex: sensible heat [c
p(c)
(T
H
T
M
)] is supplied from a high-
temperature heat source (solar collectors).
(e) Vaporization of working fluid: latent heat H
L(l)
is supplied from a low-temperature
environmental heat source (building cooling system).
(f) Heating of gaseous working fluid: sensible heat [c
p(g)
(T
M
T
L
)] is supplied.
(g) Condensation of working fluid: the heat of condensation H
M(g)
is disposed to
ambient air.
(h) Cooling of liquid working fluid: sensible heat [c
p(l)
(T
M
T
L
)] is disposed.
Thus, the calculation of the minimum entropy production for the system is:
[1]
The entropytemperature (ST) diagrams of Figure 5.6, for a discontinuous process, and
Figure 5.7, for a continuous one, depict the work related to the low- and high-temperature
cycles of the process presented in Figure 5.5. The Carnot diagram equivalent to the overall
cycle of Figure 5.6 is given in Figure 5.8. However, two different modes are distinguished:
they are coupled together and assumed to run simultaneously. These two modes can be
described as follows:
1 a high-temperature Carnot heat engine, operating between desorption and
condensation temperature;
2 a low-temperature heat pump, operating between evaporation and sorption
temperature.
Thus, relationships for the Carnot efficiencies can be established as follows.
For the heat engine cycle:
[2]
For the heat pump cycle:
[3]
For the refrigeration cycle:
[4]
And the overall efficiencies (coefficients of performance) derived from the Carnot
efficiencies and Figure 5.8 are as follows.
R
ev
HP
ev
sor ev
Q
W
T
T T
=
HP
sor
HP
sor
sor ev
Q
W
T
T T
=
HE
HE
des
des con
des
W
Q
T T
T
=

.
=
( ) +
+
S
c
T
c
T
T T
c c
T
c
T
p l
L
p g
M
M L
p p s
M
p c
H
( ) ( ) (g) ( ) ( )
( ) T T
H M
.
ABER_CH_05 6/12/08 3:32 PM Page 134
COP for heating mode, assuming that the heat of condensation is not used in heating
applications but is disposed to an environmental heat sink:
[5]
COP for heating mode, assuming that the heat of condensation is used for heating
applications:
[6]
COP for cooling mode:
[7] COP
Q
Q
T T T
T T T
COOLING
ev
des
ev des con
des sor ev
= =
( )
( )
.
COP
Q Q
Q
T T T T
T T T
HEATING
sor con
des
sor des con ev
des sor ev
=
+
=

( )
.
COP
Q
Q
T T T
T T T
HEATING
sor
des
sor des con
des sor ev
= =
( )
( )
.
FIGURE 5.6 ST diagram for a discontinuous sorption system process
ABER_CH_05 6/12/08 3:32 PM Page 135
It should be noted again that the above relationships assume temperature-independent
heats of sorption and desorption and do not take into account system-related sensible
heat.
Specifically, the basic factors that contribute to less-than-ideal system performance
are as follows (Benach, 1984):
heat of formation in the sorption process;
heat capacities of the working substances;
heat capacities of system construction materials.
The most common thermodynamic requirements related to sorption heat pumps can
therefore be summarized as follows:
The heat output per mass of the working medium should be as high as possible.
The working medium should remain isochoric during the entire cycle.
The working medium should preserve its characteristic properties during many
recycling periods.
Simple regeneration of the working medium should be possible.
FIGURE 5.7 Entropytemperature (ST) diagram for a sorption heat pump
ABER_CH_05 6/12/08 3:32 PM Page 136
TECHNICAL AND MARKET CHARACTERISTICS
OF SOLAR-ASSISTED COOLING (SAC) SYSTEMS
In this section, commercial technologies are considered, even those which have only had
a significant number of demonstration projects or have limited applications in niche
markets. These technologies are absorption, adsorption and desiccant cooling. A review of
applied systems based on these technologies can be found in Sayigh and McVeigh (1992);
Papadopoulos et al (2003); Henning (2004); Balaras et al (2007) and Henning (2007b).
The main features of these technologies, with respect to their coefficient of
performance, application, input required and stage of availability are presented in Table 5.1.
Combining the current trends of the available solar collectors and those of refrigeration
technologies, a correlation between the two areas can be produced, which is also depicted
in Table 5.1. The most important features of SAC technologies with respect to their market
applicability are discussed in the following sub-sections.
Q
des
T
dec
T
con
T
sor
T
M
Q
ev
Q
sor
= Q
ev
- W
HP
(Q
sor
= Q
ev
+ W
HE
)
W
H
E

=

-

W
H
P
T
ev
DESORBER
CONDENSER
SORBER
Q
con
= Q
des
- W
HE
High temperature cycle
A Carnot Engine (CE)
that supplies the work
to drive the low
temperature cycle
Low temperature cycle
A Carnot Heat Pump
(CHP) or
A Carnot Refrigeration
(CR) cycle
EVAPORATOR
FIGURE 5.8 Equivalent Carnot cycles of a sorption heat pump
ABER_CH_05 6/12/08 3:32 PM Page 137
T
A
B
L
E

5
.
1
M
a
i
n

f
e
a
t
u
r
e
s

o
f

p
o
s
s
i
b
l
e

s
o
l
a
r

s
o
r
p
t
i
o
n

c
o
o
l
i
n
g

t
e
c
h
n
o
l
o
g
i
e
s
T
E
C
H
N
O
L
O
G
Y
C
O
P
A
P
P
L
I
C
A
T
I
O
N
I
N
P
U
T

(
O
C
)
A
V
A
I
L
A
B
I
L
I
T
Y
C
O
L
L
E
C
T
O
R
S
C
A
P
A
C
I
T
Y

(
K
W
)
L
i
t
h
i
u
m

S
i
n
g
l
e

e
f
f
e
c
t
0
.
5
0
.
7
5
A
C
8
0
1
0
0
C
F
P
b
r
o
m
i
d
e
/
w
a
t
e
r
2
0
5
0
0
0
D
o
u
b
l
e

e
f
f
e
c
t
0
.
8
1
.
2
A
C
1
0
0
1
6
0
C
V
T
T
r
i
p
l
e

e
f
f
e
c
t
1
.
3
1
.
7
A
C
1
6
0
2
3
0
L
V
T
,

C
C
G
A
X
A
C
L
W
a
t
e
r
/
S
i
n
g
l
e

s
t
a
g
e
0
.
5
A
C
/
R
1
2
0
1
5
0
C
F
P
,

V
T
a
m
m
o
n
i
a
T
w
o

s
t
a
g
e
1
.
2
1
.
3
A
C
/
R
C
G
A
X
1
A
C
/
R
L
H
a
l
f

e
f
f
e
c
t
0
.
2
0
.
3
A
C
/
R
L
D
i
f
f
u
s
i
o
n
0
.
0
5
0
.
2
R
1
4
0
2
0
0
C
O
n
e

s
t
a
g
e

(
s
i
l
i
c
a

g
e
l
/
w
a
t
e
r
)
0
.
3
0
.
7
A
C
6
0
9
0
C
F
P
5
0
1
0
5
0
T
w
o
s
t
a
g
e

(
s
i
l
i
c
a
g
e
l
/
w
a
t
e
r
)
0
.
3
5
A
C
5
0
7
5
L
T
w
o
-
b
e
d

s
y
s
t
e
m

0
.
4
0
.
6
A
C
L
(
A
C
/
m
e
t
h
a
n
o
l
)
A
d
s
o
r
b
e
r
/
c
o
l
l
e
c
t
o
r
0
.
1
0
.
3
5
R
C
C
h
e
m
i
c
a
l

r
e
a
c
t
i
o
n

s
y
s
t
e
m
s
0
.
3
0
.
7
A
C
/
R
6
0
9
0
L
F
P
L
i
q
u
i
d
>
1
D
4
5
7
0
L
F
P
,

A
F
C
S
o
l
i
d
R
o
t
a
r
y
0
.
5
1
D
4
5
9
5
C
F
P
,

A
F
C
2
0
3
5
0
F
I
X

B
E
D
D
6
0
C
F
P
N
o
t
e
s
:
A
C

=

a
i
r

c
o
n
d
i
t
i
o
n
i
n
g
;

A
F
C

=

a
i
r

f
l
a
t

c
o
l
l
e
c
t
o
r
s
;

C

=

c
o
m
m
e
r
c
i
a
l
;

C
C

=

c
o
n
c
e
n
t
r
a
t
i
n
g

c
o
l
l
e
c
t
o
r
s
;

D

=

d
e
h
u
m
i
d
i
f
i
c
a
t
i
o
n
;

F
P

=

f
l
a
t

p
l
a
t
e
;

L

=

l
a
b
o
r
a
t
o
r
y
;

R

=

r
e
f
r
i
g
e
r
a
t
i
o
n
;

V
T

=

v
a
c
u
u
m

t
u
b
e
.
S
o
u
r
c
e
:
o
r
i
g
i
n
a
l

m
a
t
e
r
i
a
l

f
o
r

t
h
i
s

c
h
a
p
t
e
r
A b s o r p t i o nA d s o r p t i o nD e s i c c a n t
ABER_CH_05 6/12/08 3:32 PM Page 138
SIZE OF SYSTEMS
Currently, the size of most systems is quite large, resulting in limited suitability for
domestic applications. Nevertheless, apart from the existence of this smallest capacity
available limit, there is no respective upper capacity limit. Solar-assisted cooling (SAC)
systems can therefore be installed in the majority of large public buildings (e.g. offices,
complexes, big hotels, etc; see Table 5.1 for the range of capacity for cooling systems).
Hence, in order for solar-assisted chillers to take advantage of the already wide and
mature market of solar collectors, domestic-sized units should be developed. This is the
current trend, where a number of manufactures are introducing or are about to introduce
low-power systems of less than 20kW capacity (Mugnier, 2006; Henning, 2007a).
TYPES OF WORKING MEDIUM PAIRS USED (SORBENT/REFRIGERANT)
There are many parameters that influence the selection of the appropriate working
medium pair: the most important are recognized to be temperature and pressure range;
thermodynamic efficiency; the history of applications; the complexity in handling the
system; safety; and environmental concerns. Despite the plethora of different
possibilities, two pairs have until now dominated the market and can be considered as
state of the art: H
2
O/LiBr for air conditioning and NH
3
/H
2
O for refrigeration. Both pairs are
utilized in absorption technology.
With regard to further options, research has focused on improving the characteristics
of these well-known pairs in order to overcome their associated problems, and on the use
of new working-medium pairs. According to Ziegler (2002) within the first line of approach,
several salt mixtures, based on the LiBr system have been proposed. It seems that until
now the problem of corrosiveness and poor heat transfer has not yet been solved. The
same holds for the change from salt solutions to hydroxides or acids. Within the second
line of approach, there are two alternatives: first, organic pairs are often suggested a
good choice if, for any reason, the standard pairs cannot be applied. Nevertheless, they
still do not play a significant role. A more radical solution is the switch to solid sorbents
(adsorption), where zeolite attracts much attention, although silica-gel chillers are already
commercially available. Moreover, the basic research seems to be more active in the field
of silica gel, as well as in the field of ammonia salts. Another alternative considered is the
use of chemical heat pumps (chemisorption instead of physisorption for adsorption), a
perspective that gains more attention in the effort to use thermal energy for cooling
provision. An extended review of the whole range of working pairs used in chemical heat
pumps and their applications can be found in Wongsuwan et al (2001).
PROSPECTIVE MARKETS
The most attractive potential markets for SAC are areas with hot, sunny climatic
conditions. However, taking into consideration the trend monitored during the last few
years all over Europe namely, the rapid increase in summer air-conditioning loads one
should not overlook Northern European areas as prospective markets. In addition,
increasing environmental sensitivity in these countries, along with the growth of the solar
collectors market and the comparatively mild cooling load, has encouraged the installation
and use of SAC.
ABER_CH_05 6/12/08 3:32 PM Page 139
There are various markets for solar air conditioning in developed and developing
countries. Concerning developed countries where air-conditioning loads are steadily
increasing, SAC has a good potential since it can successfully meet peak electricity loads
during the summer, while also fulfilling environmental regulations for example,
concerning ozone-depletion potential (ODP) and global warming potential (GWP). On the
other hand, in developing or underdeveloped countries, the lack of an extended electricity
grid gives SAC an opportunity to cover current and future air-conditioning loads in a
sustainable way.
Another market for SAC could be areas with specific problems in electricity
transmission systems, such as islands or isolated rural areas. One can also easily deduce
that countries with a high number of installed solar collectors, such as Austria, Cyprus,
Germany, Greece and Israel, are very likely to embrace SAC since solar energy is already a
seasoned alternative to the use of fossil fuels. The same holds for countries such as Japan
where sorption refrigeration controls an essential portion of the air-conditioning market.
COMMERCIAL REQUIREMENTS AND CRITERIA
As for most heating, ventilation and air conditioning (HVAC) systems, the most important
commercial requirements and criteria for SAC systems are considered to be:
efficiency and partial load characteristics;
useful lifetime;
turnkey costs;
operation and maintenance costs;
payback period of the investment;
technical availability and reliability;
footprint, weight and volume;
noise;
technical support;
safety;
environmental performance.
The following sub-sections analyse these requirements for SAC systems, while (wherever
possible) a short comparison is provided with vapour compression cooling (VCC) systems.
Efficiency and part-load characteristics
As noted in Table 5.1, sorption cooling technologies essentially present lower efficiencies
than vapour compression systems. If the solar collectors efficiency is also considered,
the overall result is rather poor. However, taking into account the efficiency of power
plants in the latter systems and the solar fraction in the former, a relative balance is
achieved, and in cases where solar fraction exceeds 50 per cent, there are actually
savings in terms of primary energy. Nevertheless, this assumption does not consider the
parasitic energy consumption (pump energy).
Concerning partial load characteristics, SAC systems perform adequately, producing
partial load efficiencies that are as good, or even better, than those for full-load conditions
ABER_CH_05 6/12/08 3:32 PM Page 140
since their effective capacity can be varied by simply changing the rate at which the
solutions are pumped through the systems. In addition, since solar radiation is the prime
factor in generating cooling loads, the cooling output of SAC systems follows the air-
conditioning demand of buildings.
Another emerging issue concerns the fact that VCC systems consume energy of a
much higher quality namely, electricity compared to sorption chillers, which use
thermal energy, whether it is solar, waste heat or heat converted from chemical, but non-
fossil, energy sources.
Lifetime
The lifetime of SAC technologies is of the same magnitude as that of conventional
systems or other low-energy cooling technologies. In fact, the relative lack of large
moving parts and the corresponding absence of vibrations are leading to longer lifetimes
compared to a conventional compressor system. The same hypothesis stands for solar
collectors as well, since they have already proved to have a lifetime equal to that of
sorption chillers.
The American Society of Heating, Refrigerating and Air-Conditioning Engineers
(ASHRAE, 1999) suggests 23 years of service life for absorption chillers, the same for
centrifugal electric chillers and 20 years for reciprocating electric chillers.
Turnkey costs
The initial investment cost for SAC systems is significantly higher than that for vapour
compression systems since SAC systems consist of the sorption heat pump, the solar
collectors array, the auxiliary heater and the storage system. Moreover, the costs of the
chiller units alone are higher because of the larger number of heat exchangers required.
In this sense, the cost of cooling towers is also higher because they have to be of a larger
capacity in order to dispose of significantly more and higher temperature heat. In addition,
it should be emphasized that today, most VCC systems that are installed are air cooled.
The installation costs of SAC systems are higher than of those of VCC systems since the
former are larger, heavier and more technically sophisticated, although the lack of rotating
parts lowers the foundation costs.
Figure 5.9 depicts the capital costs of several solar cooling systems (prices refer to
the lower-capacity machines). The cheapest system (single-effect absorption chiller
with flat-plate collectors) costs almost five times the price of a vapour compression
cooling system. It is obvious that the cost of collectors dominates the capital cost of
SAC systems; but it should be emphasized that in such systems, solar collectors are
usually all-year-round heat sources, providing space heating and, probably, domestic hot
water as well.
In a thorough economic study by Syed et al (2002), several SAC configurations
(including photovoltaic-driven electric chillers) were compared for a 1MW cooling
installation. Syed et al (2002) estimated that SAC systems cost between 650 Euros/kW to
730 Euros/kW for single-effect absorption chillers with flat-plate collectors. The latter
constitute the larger portion of the total capital cost, amounting to 50 to 60 per cent. The
costs of electric chillers were estimated at approximately 215 Euros/kW, 195 Euros/kW
ABER_CH_05 6/12/08 3:32 PM Page 141
and 125 Euros/kW for a centrifugal, water-cooled screw and packaged air-cooled screw
vapour compression chiller, respectively.
Operation and maintenance costs
The operation and maintenance (O&M) costs of SAC sorption systems are lower than
those of VCC. Nevertheless, one cannot fail to notice that systems with water as a
refrigerant operate at very low pressure levels, close to technical vacuum conditions,
which results in some technical complexity that leads to increased maintenance costs. In
particular, vacuum maintenance is critical for the smooth function and lifetime of the
system, so special care is needed. In addition, SAC systems require an advanced system
of monitoring and control.
In general, maintenance costs for absorption chillers range from 0.6 to 1.25 times the
maintenance costs of vapour compression chillers (Elsafty and Al-Daini, 2002). Syed et al
(2002) estimated maintenance costs at 1 and 4 per cent of the combined capital costs of
chiller and cooling tower for the absorption and vapour compression chillers,
respectively.
Concerning fuels costs, SAC systems (electricity for pumps and fans, and natural gas
or oil for auxiliary system) are likely to be cheaper than VCC systems (electricity for pumps,
fans and compressors). These points, though, are based on the assumption of a solar
collector array large enough for primary energy savings to be achieved and of fairly flat
rates for electricity. In the case of escalating load tariffs imposed by electric utilities, SAC
performs even more economically (Papadopoulos et al, 2006).
Where the cost or availability of water is of importance, the issue arises of using water
in the cooling towers, which benefits the economic performance of air-cooled VCC systems.
Several studies have showed that annual economic performance of SAC systems is poor
compared to that of VCC systems. Balaras et al (2007) calculated that for an office building in
Madrid, the best SAC system (single-effect absorption chiller with flat-plate collectors) has
15 per cent higher annual costs than the electric compression cooling system.
Source: adjusted from Kim (2007)
FIGURE 5.9 Cost of solar cooling systems v. vapour compression cooling
ABER_CH_05 6/12/08 3:32 PM Page 142
Payback period
According to Henning et al (1998) who studied the efficiency of solar sorption cooling
systems in three areas in Germany, Denmark and Italy, the lowest payback periods are
achieved with desiccant cooling systems. The payback period does not significantly increase
for solar fraction values up to some 70 per cent, so SAC can reach high solar fractions without
leading to oversized and, hence, disproportionately expensive collector arrays.
With respect to SAC, Tsoustos et al (2003) concluded that even in regions with
increased air-conditioning demand and solar energy availability, such as Greece, there is a
strong need for investment incentives and for conventional-energy taxation, which would
help to internalize the full environmental costs of using conventional fuels for cooling.
Technical availability and reliability
Once more, the lack of moving parts and the concept of a thermally driven system result in
both exceptional technical availability and reliability. Nevertheless, the absence of an
extended sales and technical support network is a major disadvantage that has to be tackled.
Footprint, weight and volume
Both the footprint and weight of SAC systems constitute a significant drawback. Sorption
chillers are heavier and larger than VCC chillers, so the addition of the needed solar
collector array really puts a big constraint on the applicability of SAC to buildings.
Generally, absorption chillers are almost two times heavier than water-cooled vapour
compression chillers. When considering air-cooled chillers, the difference in weight is
even larger. Vapour compression chillers also have half the footprint and volume of
absorption chillers.
Table 5.2 compares the weight, footprint and volume of three 400kW chillers from two
major manufactures.
In addition, beyond the need for a quite large area/room for the installation of the
sorption chiller (in terms of surface and volume), an extended roof area is also needed for
the installation of a large collector array if significant energy savings are to be achieved.
Noise
Again, the lack of reciprocating and rotating parts results in a vibration-free and noiseless
operation that can be important in specific applications. Typical sound levels for sorption
TABLE 5.2 Physical characteristics of several chillers of 400kW cooling capacity
CHILLER TYPE SINGLE-EFFECT WATER-COOLED AIR-COOLED VAPOUR
ABSORPTION (WATER VAPOUR COMPRESSION
FIRED) COMPRESSION
Weight 12kg/kW 7kg/kW 5.5kg/kW
Footprint 5m
2
2.5m
2
2.5m
2
Volume 11m
3
4.5m
3
4.5m
3
Source: Carrier Corporation (2007); Yazaki Energy Systems Inc (2007)
ABER_CH_05 6/12/08 3:32 PM Page 143
chillers are 80dB(A), while for vapour compression chillers, levels of 95dB(A) (depending
upon the type of compressor: centrifugal, reciprocating or rotary).
Technical support
Currently, there is still a lack of practical knowledge of integrated systems covering their
whole life cycle of design, engineering, establishment, control, operation, maintenance
and, eventually, disposal. This also reflects the absence of skilled and competent
technicians.
Safety
In principle, SAC systems are considered to be safer than any conventional systems
because of the natural working fluids used and the lack of specialized electricity
installation.
Environmental performance
A clear and sound assessment of the environmental performance of SAC should definitely
take into account the use of solar collectors throughout the year, and not just on a
seasonal base. Thus, solar thermal energy should be considered to provide hot water,
space heating and space cooling. According to this perspective, there are three
parameters to be used for assessing environmental performance:
1 primary energy saving, or primary energy ratio;
2 global warming potential;
3 ozone-depletion potential.
Considering these three parameters, SAC appears to have a favourable environmental
performance when compared to conventional systems. In fact, most SAC systems use
water as a refrigerant, in no case are ozone-depleting substances employed and, when
combined with an adequate area of solar collectors, the system can result in essential
energy savings for both heating and cooling, performing better both from a primary
energy ratio and global warming potential point of view. Lamp and Ziegler (1998)
determined the solar fraction necessary for a SAC (absorption) system to consume less
primary energy compared to a VCC system, given the electricity generation efficiency
depicted in Figure 5.10. Balaras et al (2007) estimated that the primary energy saving
for several configurations of SAC systems and for a collector array of 3m
2
/kW
cooling
varies from 36 to 53 per cent compared to a VCC system for an office building in
Madrid.
An assessment of environmental performance should include water consumption and
thermal pollution caused by the chillers. In terms of this criterion, sorption chillers with a
COP of 0.7 reject 2.43 units of heat for every unit of cooling served, while electric chillers
with a COP of 3 reject only 1.33 units of heat for every unit of cooling served. The heat
that is rejected in sorption chillers, therefore, results in much higher water consumption
in these cooling towers.
ABER_CH_05 6/12/08 3:32 PM Page 144
TECHNICAL AND NON-TECHNICAL BARRIERS TO COMMERCIALIZATION
There are additional barriers that discourage the commercialization of SAC systems.
Technical barriers
Beyond the technical issues reported earlier, additionally technical constraints exist in
solar-assisted air conditioning.
First, SAC technologies have not yet reached an acceptable level of technological
sophistication and are still considered to be under development. There is a need, for
example, for cost-efficient and operationally friendly heat storage, as well as a backup
system.
SAC systems demand advanced electronic control systems, a field that is developed
enough to allow for the implementation of SAC systems in buildings, though not
necessarily to the point of achieving optimized, unobstructed and reliable daily operation
over many years in frequently harsh environments. Over the past decade, progress in the
field of building management systems and remote control is expected to provide much
needed assistance in this sector (Canbay et al, 2004; Hipkiss, 2005).
1.5
1
0.5
0
0 0.2
Solar Fraction f / Efficiency of Power Generation
P
r
i
m
a
r
y

E
n
e
r
g
y

R
a
t
i
o

P
E
R
0.4
COP = 3
COP = 5
Double
effect
Single
effect
0.6 0.8 1
Electrical driven compression chillers
Solar assisted sorption chillers
Source: Lamp and Ziegler (1998)
FIGURE 5.10 Primary energy ratio of solar-assisted absorption and electrically driven vapour compression chillers
ABER_CH_05 6/12/08 3:32 PM Page 145
Non-technical barriers
The main non-technical barrier to adopting SAC systems is the fact that solar refrigeration
is still not competitive in terms of the energy services provided by the utilities. Specifically,
retail electricity prices were fairly low throughout the period of 1995 to 2005, fuelled by
competition that followed liberalization in most European countries, where steeply rising
time-dependent tariffs where not common. However, this trend has begun to change and
prices are not expected to decrease, due to the demand for new investments in power
generation capacities and to CO
2
emission charges (Richmond and Kaufmann, 2006).
Another significant barrier is the relative lack of public awareness of available systems,
together with a lack of environmental sensitivity. In addition, the absence of an appropriate
sound legislation, encouraging environmentally friendly energy systems (investment incentive,
energy tax and cost of externalities), does not allow SAC systems to take advantage of all
aspects of their environmental performance when compared to conventional systems.
Limited access to funding also presents a barrier to SAC, which is characterized by a high
initial investment. The complexity and cost of different national testing and certification
procedures are barriers to trade in solar systems. Indeed, the lack of sound certification
and of standardization of SAC systems constitutes another barrier.
Finally, architects, developers and construction companies are not interested in
adopting active or sophisticated systems such as SAC, but rather adopt conventional
solutions such as VCC systems.
EXTERNAL FACTORS THAT WILL ACCELERATE COMMERCIALIZATION
The most important external factor that can influence the commercialization of SAC is
energy costs. An increase in the cost of electricity will accelerate not only technological
development, but the acceptability of such systems. More specifically, if more electricity
utility companies adopt time-dependent tariffs or demand-charging schemes, this will
definitely lead to an increase of interest for SAC systems. An increase in energy prices can
also be initiated by environmental policies that include the costs of environmental
externalities within conventional fuel prices. In fact, environmental issues are already
putting pressure on electric utilities, placing a heavy burden on the construction and
operation of new generation plants.
Environmental policies are highly influential in determining the success of SAC
systems within the air-conditioning market. In the European Union, in particular, a series
of actions involving the promotion and propagation of renewable energy sources (RES)
not only refer to research and development (R&D), and dissemination and integration of
low-energy technologies, but also highlight regulatory measures as well. Of course, the
direction of additional funding towards R&D is a crucial element that may prove to be
significant since SAC technologies must continue to evolve if they are to be competitive.
In addition, regarding private-sector strategies, SAC systems could be the subject of
integrated demand-side management programmes by both electricity and gas utilities
since, during the summer, the electricity grid is frequently overloaded, in contrast to gas.
Covering peak electricity loads appears to be a particularly significant problem in
liberalized electricity markets since private utility companies are not willing to invest the
huge capital needed to construct power plants that will be working for only days or weeks
ABER_CH_05 6/12/08 3:32 PM Page 146
per year. Another factor that may contribute to the penetration of SAC within the
commercial market is the expansion of energy service companies (ESCOs), as these can
more easily meet the high initial capital investment required by SAC systems,
compensating for it by the lower annual energy cost.
A SWOT ANALYSIS APPROACH
SWOT is a strategic analysis tool, used widely in industrial management since the 1980s,
and has two elements: an external analysis of opportunities (O) and threats (T) and an
internal analysis of strengths (S) and weaknesses (W). The external analysis focuses on
the external environment and the opportunities and threats that this may represent for
technology, a product or investment. Opportunities or threats are defined according to
individual perceptions, assumptions about what the future holds or which aspect of the
environment is considered. The internal analysis examines the strengths and weaknesses
of technology in relation to its users and competitive systems. The aim of the SWOT
analysis is to produce a TOWS matrix, which contains the strategies that have to be
applied in order to face threats and opportunities by utilizing the strengths and improving
weaknesses.
When applied to a typical industrial product, the SWOT analysis focuses on six main
points:
1 product to be sold;
2 process of selling the product;
3 customer;
4 distribution;
5 financial aspect (prices, costs and investments needed); and
6 administration of the whole scheme (Humphrey, 2004).
In a slightly varying line of approach, SWOT analysis is considered to incorporate Porters
Five Forces model as an outside-in business unit strategy tool, which is applied by means
of identifying the five fundamental competitive forces (Wu and Wu, 2005):
1 entry of competitors (how easy or difficult is it for new entrants to compete; which
barriers exist?);
2 threat of substitutes (how easy can our product or service be substituted with
cheaper options?);
3 bargaining power of buyers (how strong is the position of buyers; can they work
together to order large volumes?);
4 bargaining power of suppliers (how strong is the position of sellers; are there many
or only a few potential suppliers; is there a monopoly?);
5 rivalry among existing players (is there strong competition between existing players;
is one player dominant or are all equal in strength/size?).
Various published sources also consider the government to be an additional sixth
competitive force.
ABER_CH_05 6/12/08 3:32 PM Page 147
TABLE 5.3 Strategic factors of the external environment
INCLINES LIST PROBABILITY POSSIBLE EFFECTS PRIORITY MATRIX
OF SUCCESS ON TECHNOLOGY
OPPORTUNITIES
Ozone depletion Large Mean Mean
Climate change Mean Mean Mean
(global warming)
Climate change Mean Small Small
(Higher frequency
of heatwaves)
Stricter legislations Large Mean Large
on energy performance
of buildings
Existing legislation on Large Large Large
the promotion of renewable
energy sources (RES)
Expanding markets in Large Mean Mean
developing countries
New approach of energy Mean Small Small
systems in the liberated
market
Integration of externalities Mean Large Large
within energy costs
Increasing environmental Mean Small Small
awareness
Inability of electric Large Mean Large
utilities to deliver
peak load
Environmental and Large Mean Mean
financial constraints
for new power plants
Possibility of introducing Large Mean Large
time-dependent tariffs
Increasing demand for Large Mean Mean
building air conditioning
Development of energy Large Mean Mean
service companies
(ESCOs)
High penetration of Large Large Large
solar thermal collectors in
many countries, increasing
rates in many others
Lack of need for night-time air Large Small Small
conditioning in offices and
most public buildings
ABER_CH_05 6/12/08 3:32 PM Page 148
TABLE 5.3 Strategic factors of the external environment (Contd)
INCLINES LIST PROBABILITY POSSIBLE EFFECTS PRIORITY MATRIX
OF SUCCESS ON TECHNOLOGY
THREATS
Liberalization and Mean Mean Mean
deregulation of the
electricity market can
keep electricity costs low
Power companies and Large Mean Mean
customers prefer
conventional methods
(i.e. TES) in order to
smooth out peak load
demand
Domination of the Large Mean Large
air-conditioning market
by vapour compression
cooling (VCC) systems
Architects and construction Mean Small Small
companies are not interested
in active solar systems
TABLE 5.4 Strategic factors of the internal environment
INCLINES LIST PROBABILITY OF POSSIBLE EFFECTS PRIORITY MATRIX
SUCCESS ON TECHNOLOGY
STRENGTHS
Lower energy Large Large Large
consumption (favours
primary energy ratio)
Hybrid systems Large Mean Mean
(desiccant vapour
compression cooling
(VCC), absorption and
adsorption; natural gas
cooling)
Provide free solar Large Large Large
heating during
winter as well
Modular and easily Large Mean Mean
adapted to conventional
heating and cooling systems
ABER_CH_05 6/12/08 3:32 PM Page 149
TABLE 5.4 Strategic factors of the internal environment (Contd)
INCLINES LIST PROBABILITY OF POSSIBLE EFFECTS PRIORITY MATRIX
SUCCESS ON TECHNOLOGY
Do not use ozone-depletion Mean Mean Mean
potential (ODP) refrigerants
Do not contribute to global Large Mean Mean
warming impact (GWI)
Thermally driven cooling Mean Mean Mean
systems already have a
relatively essential share
in the air-conditioning market
Solar thermal collectors Mean Mean Mean
are already proven to be
commercially sophisticated
and sound technology
WEAKNESSES
Many applications are not Large Large Large
adequately sophisticated
Highly sophisticated systems Large Large Large
require advanced control and
monitoring
Low overall efficiency Large Large Large
Large footprint; sizeable systems Large Mean Large
require substantial free space
Need for solar energy storage Mean Mean Mean
High initial cost Large Large Large
Falling prices for VCC systems Large Mean Mean
Very limited commercial Mean Large Mean
representation
Lack of sound certification Large Mean Mean
and standardization of
systems
Not autonomous systems; Large Mean Mean
require auxiliary (backup)
system
Lack of imaginary strategies Mean Mean Mean
and marketing techniques for
renewable energy source
(RES) technologies
Complete absence of Large Small Mean
skilled and educated technicians
for maintenance and repair
Source: Original material for this chapter
ABER_CH_05 6/12/08 3:32 PM Page 150
T
A
B
L
E

5
.
5
S
t
r
a
t
e
g
i
e
s

f
o
r

d
e
v
e
l
o
p
m
e
n
t

(
T
O
W
S

m
a
t
r
i
x
)
I
N
T
E
R
N
A
L

F
A
C
T
O
R
S
S
T
R
E
N
G
T
H
S
W
E
A
K
N
E
S
S
E
S
E
X
T
E
R
N
A
L

F
A
C
T
O
R
S
L
o
w
e
r

H
y
b
r
i
d

s
y
s
t
e
m
s
P
r
o
v
i
d
e
M
a
n
y
H
i
g
h
l
y
L
o
w
L
a
r
g
e
H
i
g
h

e
n
e
r
g
y

(
d
e
s
i
c
c
a
n
t
/
f
r
e
e

s
o
l
a
r
a
p
p
l
i
c
a
t
i
o
n
s
s
o
p
h
i
s
t
i
c
a
t
e
d
o
v
e
r
a
l
l
f
o
o
t
p
r
i
n
t
;
i
n
i
t
i
a
l
c
o
n
s
u
m
p
t
i
o
n
v
a
p
o
u
r
h
e
a
t
i
n
g
a
r
e

n
o
t
s
y
s
t
e
m
s
e
f
f
i
c
i
e
n
c
y
s
i
z
e
a
b
l
e
c
o
s
t
s
(
f
a
v
o
u
r
a
b
l
e

c
o
m
p
r
e
s
s
i
o
n
d
u
r
i
n
g
a
d
e
q
u
a
t
e
l
y
r
e
q
u
i
r
e

s
y
s
t
e
m
s
p
r
i
m
a
r
y
c
o
o
l
i
n
g
,

w
i
n
t
e
r
s
o
p
h
i
s
t
i
c
a
t
e
d
a
d
v
a
n
c
e
d

r
e
q
u
i
r
e
e
n
e
r
g
y

a
b
s
o
r
p
t
i
o
n

a
n
d

c
o
n
t
r
o
l

a
n
d
s
u
b
s
t
a
n
t
i
a
l
r
a
t
i
o
)
a
d
s
o
r
p
t
i
o
n
/
N
G

m
o
n
i
t
o
r
i
n
g
f
r
e
e

s
p
a
c
e
c
o
o
l
i
n
g
)
S
T
R
E
N
G
T
H
S

A
N
D

W
E
A
K
N
E
S
S
E
S

A
N
D

O
P
P
O
R
T
U
N
I
T
I
E
S

(
S
O
)

O
P
P
O
R
T
U
N
I
T
I
E
S

(
W
O
)

S
T
R
A
T
E
G
I
E
S
S
T
R
A
T
E
G
I
E
S
O
P
P
O
R
T
U
N
I
T
I
E
S
S
t
r
i
c
t
e
r

T
a
x

(
o
r

o
t
h
e
r
)

p
e
n
a
l
t
i
e
s
I
m
p
o
s
i
t
i
o
n

o
f
l
e
g
i
s
l
a
t
i
o
n
s

o
n

h
i
g
h

(
p
r
i
m
a
r
y
)
r
e
s
t
r
i
c
t
i
o
n
s

o
n
,
o
n

e
n
e
r
g
y

e
n
e
r
g
y
-
c
o
n
s
u
m
i
n
g
a
n
d

p
r
o
m
o
t
i
o
n

o
f
p
e
r
f
o
r
m
a
n
c
e

b
u
i
l
d
i
n
g
s
s
o
l
a
r
-
a
s
s
i
s
t
e
d
o
f

b
u
i
l
d
i
n
g
c
o
o
l
i
n
g

(
S
A
C
)
s
y
s
t
e
m
s

t
o
p
r
o
f
e
s
s
i
o
n
a
l

a
c
t
o
r
s
(
a
r
c
h
i
t
e
c
t
s

a
n
d
c
o
n
s
t
r
u
c
t
i
o
n

c
o
m
p
a
n
i
e
s
)
E
x
i
s
t
i
n
g

l
e
g
i
s
l
a
t
i
o
n

o
n

t
h
e

p
r
o
m
o
t
i
o
n

o
f

r
e
n
e
w
a
b
l
e
e
n
e
r
g
y

s
o
u
r
c
e
s

(
R
E
S
)
I
n
t
e
n
s
i
f
y

R
&
D

p
r
o
g
r
a
m
m
e
s

o
n

S
A
C

s
y
s
t
e
m
s
S
u
b
s
i
d
i
z
e

R
E
S

s
y
s
t
e
m
s
ABER_CH_05 6/12/08 3:32 PM Page 151
T
A
B
L
E

5
.
5
S
t
r
a
t
e
g
i
e
s

f
o
r

d
e
v
e
l
o
p
m
e
n
t

(
T
O
W
S

m
a
t
r
i
x
)

(
C
o
n
t
d
)
I
N
T
E
R
N
A
L

F
A
C
T
O
R
S
S
T
R
E
N
G
T
H
S
W
E
A
K
N
E
S
S
E
S
E
X
T
E
R
N
A
L

F
A
C
T
O
R
S
I
n
t
e
g
r
a
t
i
o
n

P
r
o
m
o
t
i
o
n
I
n
t
r
o
d
u
c
t
i
o
n

o
f
I
n
v
o
l
v
e
m
e
n
t

o
f

p
r
i
v
a
t
e
I
n
t
r
o
d
u
c
t
i
o
n

o
f
o
f

e
x
t
e
r
n
a
l
i
t
i
e
s

a
n
d

p
r
o
v
i
s
i
o
n
e
n
v
i
r
o
n
m
e
n
t
a
l
c
o
m
p
a
n
i
e
s

(
u
t
i
l
i
t
i
e
s

a
n
d
e
n
v
i
r
o
n
m
e
n
t
a
l
w
i
t
h
i
n

e
n
e
r
g
y

o
f

i
n
c
e
n
t
i
v
e
s
t
a
x
e
s

(
g
r
e
e
n
m
a
n
u
f
a
c
t
u
r
e
s

o
f

s
o
l
a
r
t
a
x
e
s

(
g
r
e
e
n
c
o
s
t
s
f
o
r

d
e
m
a
n
d
-
s
i
d
e
t
a
x
e
s
)

o
n

t
h
e
c
o
l
l
e
c
t
o
r
s

a
n
d

r
e
f
r
i
g
e
r
a
t
i
o
n
t
a
x
e
s
)

o
n

t
h
e
m
a
n
a
g
e
m
e
n
t
p
r
i
c
e

o
f

e
n
e
r
g
y
e
q
u
i
p
m
e
n
t
)

i
n

R
&
D

a
n
d
p
r
i
c
e

o
f

e
n
e
r
g
y
I
n
a
b
i
l
i
t
y

o
f
(
D
S
M
)

p
r
o
g
r
a
m
m
e
s
p
r
o
m
o
t
i
o
n
a
l

p
r
o
g
r
a
m
m
e
s
e
l
e
c
t
r
i
c

u
t
i
l
i
t
i
e
s

f
r
o
m

b
o
t
h
t
o

d
e
l
i
v
e
r

e
l
e
c
t
r
i
c

a
n
d
p
e
a
k

l
o
a
d
g
a
s

u
t
i
l
i
t
i
e
s
P
o
s
s
i
b
i
l
i
t
y

A
l
l
o
w
a
n
c
e
s

a
n
d

i
n
c
e
n
t
i
v
e
s

f
o
r

e
n
e
r
g
y
o
f

i
n
t
r
o
d
u
c
i
n
g

s
e
r
v
i
c
e

c
o
m
p
a
n
i
e
s

(
E
S
C
O
s
)

i
n

d
e
l
i
v
e
r
i
n
g
t
i
m
e
-
d
e
p
e
n
d
e
n
t

e
n
e
r
g
y

m
a
n
a
g
e
m
e
n
t

w
i
t
h
i
n

b
u
i
l
d
i
n
g
s
t
a
r
i
f
f
s

H
i
g
h

A
d
o
p
t
i
o
n

o
f

m
a
r
k
e
t
i
n
g
I
n
c
e
n
t
i
v
e
s

t
o

r
e
p
l
a
c
e

o
l
d
F
i
n
a
n
c
i
n
g

R
&
D
p
e
n
e
t
r
a
t
i
o
n

t
e
c
h
n
i
q
u
e
s

b
y

n
e
w
c
o
l
l
e
c
t
o
r
s

w
i
t
h

h
i
g
h
e
r
p
r
o
j
e
c
t
s

f
o
r
o
f

s
o
l
a
r

s
e
r
v
i
c
e
s

t
h
a
t

c
a
n

b
e
e
f
f
i
c
i
e
n
c
y
c
h
e
a
p
e
r
t
h
e
r
m
a
l

p
r
o
v
i
d
e
d

b
y

s
o
l
a
r
c
o
l
l
e
c
t
o
r
s
;
c
o
l
l
e
c
t
o
r
s

t
h
e
r
m
a
l

c
o
l
l
e
c
t
o
r
s
p
r
o
p
a
g
a
t
i
o
n

o
f
i
n

m
a
n
y

s
o
l
a
r

c
o
l
l
e
c
t
o
r
s
c
o
u
n
t
r
i
e
s
,

i
n

o
r
d
e
r

t
o

r
e
d
u
c
e

c
o
s
t
s
i
n
c
r
e
a
s
i
n
g

r
a
t
e
s

i
n

m
a
n
y

o
t
h
e
r
s
ABER_CH_05 6/12/08 3:32 PM Page 152
S
T
R
E
N
G
T
H
S

A
N
D

W
E
A
K
N
E
S
S
E
S

A
N
D

T
H
R
E
A
T
S

(
W
T
)

S
T
R
A
T
E
G
I
E
S
T
H
R
E
A
T
S

(
S
T
)

S
T
R
A
T
E
G
I
E
S
T
H
R
E
A
T
S
L
i
b
e
r
a
l
i
z
a
t
i
o
n

P
r
i
c
e

p
e
n
a
l
t
i
e
s

o
n

l
a
r
g
e
D
e
t
e
r
m
i
n
a
t
i
o
n

o
f
a
n
d

d
e
r
e
g
u
l
a
t
i
o
n

e
l
e
c
t
r
i
c
i
t
y

c
o
n
s
u
m
e
r
s
h
i
g
h

c
o
s
t
o
f

e
l
e
c
t
r
i
c
i
t
y

(
d
u
e

t
o

c
o
o
l
i
n
g

f
o
r

b
l
a
c
k
o
u
t
s

o
r
m
a
r
k
e
t

c
a
n

o
r

o
t
h
e
r

u
s
e
s
)
i
n
a
b
i
l
i
t
y

t
o
k
e
e
p

e
l
e
c
t
r
i
c
i
t
y

p
r
o
v
i
d
e

e
l
e
c
t
r
i
c
i
t
y
c
o
s
t
s

l
o
w
D
o
m
i
n
a
t
i
o
n

T
a
x

p
e
n
a
l
t
i
e
s

o
n

t
h
e
T
a
x

p
e
n
a
l
t
i
e
s
o
f

t
h
e

p
r
i
c
e

o
f

l
o
w
-
e
f
f
i
c
i
e
n
c
y
o
n

t
h
e

p
r
i
c
e
a
i
r
-
c
o
n
d
i
t
i
o
n
i
n
g

c
o
n
v
e
n
t
i
o
n
a
l
o
f

l
o
w
-
e
f
f
i
c
i
e
n
c
y
m
a
r
k
e
t

b
y

a
i
r
-
c
o
n
d
i
t
i
o
n
i
n
g
c
o
n
v
e
n
t
i
o
n
a
l
V
C
C

s
y
s
t
e
m
s
s
y
s
t
e
m
s
a
i
r
-
c
o
n
d
i
t
i
o
n
i
n
g
s
y
s
t
e
m
s
S
o
u
r
c
e
:
o
r
i
g
i
n
a
l

m
a
t
e
r
i
a
l

f
o
r

t
h
i
s

c
h
a
p
t
e
r
ABER_CH_05 6/12/08 3:32 PM Page 153
Nevertheless, when the SWOT methodological approach is being expanded from a
product to a technology and a market branch, one has to consider factors such as the
customers awareness, the technologys image, the importance of its attributes to the
potential customers, and the importance of its attributes to decision- and policy-
makers who determine the boundary conditions of the building sector as a market
(Terrados et al, 2007). Furthermore, one has to keep in mind that the complexity of the
building markets response to technological changes makes the assessment of
impacts difficult (Tsoutsos et al, 2003). In this sense, SWOT analysis can be useful
since it can provide a framework for medium- and long-term planning, especially if
economies of scope are to be achieved. This framework takes the shape of a TOWS
matrix, providing the results of analysis of the solar cooling market (Papadopoulos
et al, 2004)
Tables 5.3 to 5.5 present the results derived from the SWOT analysis of SAC
technologies, which takes into consideration the features described in this section. The
strategic factors of the external and internal environment are presented, as well as a
resulting TOWS matrix of the proposed development strategies.
CONCLUSIONS
Solar cooling, as an exclusive solution or assisted by a conventional cooling system, cannot
be considered a competitive financial and technological alternative to conventional air-
conditioning systems, at least under the prevailing conditions in the construction and
energy markets of most countries. As a result, one can understand its difficulties in
entering competitive, conservative and demanding building services markets. On the other
hand, solar air conditioning potentially offers an elegant model for a clean, sustainable
technology, which is consistent with the international commitment to sustainable
development. The increasing effort in applied research during the last decade, combined
with the collateral advantages arising from the use of solar cooling technologies, together
attest to the high potential for commercial applicability. The main advantages of solar air-
conditioning technologies can by summarized in the:
reduction of summer peak loads in electricity demand;
close seasonal and hourly coincidence of solar radiation and a buildings cooling
demand profile;
absence of any ozone-depleting refrigerants;
decreased primary energy consumption;
decreased global warming impact; and
maximization of the exploitation of solar energy for heating, hot-water production and
cooling by the same solar system.
According to this line of thought, the option of solar cooling becomes an increasingly
attractive one. Nevertheless, there are significant barriers towards its commercialization.
Targeted actions should be adopted concerning R&D, dissemination and promotion of
technology, providing effective incentives to users. Another set of actions may refer to the
dominant technology of electrically driven VCC in terms of introducing environmental
ABER_CH_05 6/12/08 3:32 PM Page 154
taxes or penalties for low overall energy efficiency in order to offset the hidden costs of
these conventional systems.
Nevertheless, the real momentum that solar-assisted cooling can take advantage of is
the substantial increase in demand for air conditioning in urban areas over the last decade.
It is not logical to expect that this increasing demand can be covered by electricity utilities
without incurring a significant economic and environmental penalty. It is therefore
reasonable to expect that a significant part of this demand will redefine the market,
providing access for new technologies and approaches, including SAC systems. This
prospect can offset the current financial unattractiveness of SAC since electric utilities will
be forced to shift their marginal costs that are due to increased peak loads to consumers.
SAC is a favourable tool for demand-side management in relation to both electric and
natural gas utilities. Furthermore, the activities of energy service provision schemes,
especially in relation to the management of large buildings, is another tool for propagating
SAC technologies in the market, especially since these companies can ensure the
financing of the high initial capital investment required by SAC systems.
Reforms in the energy market may lead to a new framework for the entire sector:
investors, energy producers, brokers and service companies become active in an
expanding, more competitive market, utilizing new technologies and taking advantage of
developing primary resource potential. But even in this liberalized environment, some
form of regulation is necessary. The success story of wind power and the growing
popularity of photovoltaics are good examples for such a balancing act: the market has to
find its way; but mature technologies have to be available and this can only be done with
support during their infancy period. Nevertheless, climate change policies cannot be
effective when solely based on economic instruments and technological advances. A
broader social consensus is needed for any regulatory measure to succeed.
S. Oxizidis: (corresponding author) Laboratory of Heat Transfer and Environmental Engineering,
Department of Mechanical Engineering, Aristotle University Thessaloniki, Box 483, 54124 Thessaloniki,
Greece; tel (30) 2310 996048; fax (30) 2310 996012; oxis@aix.meng.auth.gr
A. M. Papadopoulos: Laboratory of Heat Transfer and Environmental Engineering, Department of Mechanical
Engineering, Aristotle University Thessaloniki, Box 483, 54124 Thessaloniki, Greece
ACKNOWLEDGEMENTS
The authors wish to thank the European Commission of Training and Mobility of
Researchers Programme (DG Research) for funding the ATREUS project under Research
Training Networks, HPRN-CT-2002-00207.
REFERENCES
Alvarez, S. and J. L. Molina (2003) Cooling by natural sinks, in Santamouris, M. (ed) Solar Thermal Technologies for
Buildings: The State of the Art, James & James, London, pp140162
ABER_CH_05 6/12/08 3:32 PM Page 155
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) (1999) HVAC applications, in ASHRAE
Handbook, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA
Avgelis, A., S. Oxizidis, A. Stylianou and A. M. Papadopoulos (2005) Energy consumption and indoor quality relationship in a
naturally ventilated building, in Proceedings of the Urban Air Quality Conference, Valencia, Spain, 2931 March
Balaras, C.-A., G. Grossman, H.-M. Henning and C.-A.-I. Ferreira (2007) Solar air conditioning in Europe an overview,
Renewable and Sustainable Energy Reviews, vol 11, pp299304
Benach, R. (1984) Chemical Heat Pump/Thermochemical Energy Storage Systems I: A Reconnaissance, TNO, The Hague
BSRIA (2000) World Market for Air Conditioning Update 2001, www.bsria.co.uk/, accessed June 2006
Canbay, C. S., A. Hepbasli and G. Gokcen (2004) Evaluating performance indices of a shopping centre and implementing
HVAC control principles to minimize energy usage, Energy and Buildings, vol 36, no 6, pp587598
Carrier Corporation (2007) Performance data at www.carrier.com, accessed 31 October 2007
Cartalis, C., A. Synodinou, M. Proedrou, A. Tsangrassoulis and M. Santamouris (2001) Modifications of energy demand in
urban areas as a result of climate changes: An assessment for the southeast Mediterranean region, Energy Conversion &
Management, vol 42, pp16471656
Chatzidimoula, E.L. (2004) Contribution to the Natural Ventilation of Internal Spaces via Controlled Use of Ventilation
Openings, PhD thesis, Aristotle University of Thessaloniki, Department of Civil Engineering, Thessaloniki, Greece
Cheung, K., Y. Hwang, J. Judge, K. Kolos, A. Singh and R. Radermacher (1996) Performance assessment of multistage
absorption cycles, International Journal of Refrigeration, vol 19, pp473481
Cheung, C. K., R. J. Fuller and M. B. Luther (2005) Energy-efficient envelope design for high-rise apartments, Energy and
Critoph, R. E. (1999) Rapid cycling solar/biomass powered adsorption refrigeration system, Renewable Energy, vol 16,
pp673678
Dethman, L. (ed) (2004) Energy Star Home Products Program: Market Evaluation Report No 2, Seattle, WA
Dorgan, C. B., S. P. Leight and C. E. Dorgan (1995) Application Guide for Absorption Cooling/Refrigeration Using Recovered
Heat, ASHRAE, Atlanta, GA
Elsafty, A. and A. J. Al-Daini (2002) Economical comparison between a solar-powered vapour absorption air-conditioning
system and a vapour compression system in the Middle East, Renewable Energy, vol 25, pp569583
Geros, V., M. Santamouris, S. Karatasou, A. Tsangrassoulis and N. Papanikolaou (2005) On the cooling potential of night
ventilation techniques in the urban environment, Energy and Buildings, vol 37, pp243257
Henning, H. M. (ed) (2004) Solar-Assisted Air-Conditioning in Buildings: A Handbook for Planners, Springer Wien, New York, NY
Henning, H.-M. (2007a) Solar cooling, in Proceedings of ISES Solar World Congress 2007, Beijing, China
Henning, H.-M. (2007b) Solar assisted air conditioning of buildings an overview, Applied Thermal Engineering, vol 27,
pp17341749
Henning, H. M., T. Erpenbeck, C. Hindenburg and S. Paulussen (1998) Solar cooling of buildings possible techniques,
potential and international development, in Proceedings of Eurosun 1998, Ljubljana, vol III, pp2.20-12.20-6
Hipkiss, R. (2005) Energy management needs intelligence, Energy World, vol 328, pp2021
Humphrey, W. S. (2004) The Personal Software Process Technical Report,
www.sei.cmu.edu/pub/documents/00.reports/pdf/00tr022.pdf/, accessed June 2004
Katsakos, K. (2007) The Effect of Electric and Electronic Equipment on the Cooling Load of Buildings, Diploma thesis, Aristotle
University of Thessaloniki, Department of Mechanical Engineering, Laboratory of Heat Transfer and Environmental
Engineering, Thessaloniki, Greece
Kim, D.-S. (2007) Solar Absorption Cooling, PhD thesis, Technische Universiteit Delft, The Netherlands
Lamp, P. and Z. Ziegler (1998) European research on solar assisted air conditioning, International Journal of Refrigeration, vol
21, no 2, pp8999
ABER_CH_05 6/12/08 3:32 PM Page 156
Li, Z. F. and K. Sumathy (2000) Technology development in the solar absorption air conditioning systems, Renewable and
Sustainable Energy Reviews, vol 4, pp267293
Mugnier, D. (2006) Thermally Driven Water Chillers: Market Status and New Developments, IEA SHC Programme Task 38
Meeting, Aix les Bains, France
Papadopoulos, A. M. (2001) The influence of street canyons on the cooling loads of buildings and the performance of air
conditioning systems, Energy and Buildings, vol 33, pp601607
Papadopoulos, A. M., S. Oxizidis and N. Kyriakis (2003) Perspectives of solar cooling in view of the developments in the
air-conditioning sector, Renewable and Sustainable Energy Reviews, vol 7, pp419438
Papadopoulos, A. M., D. H. Oxizidis and I. Samlidis (2004) Constrains and potential for the propagation of solar refrigeration in
the buildings sector, Ktirio, Scientific Edition, vol AB, pp4150
Papadopoulos, A. M., A. Stylianou and S. Oxizidis (2006) Impact of energy pricing on buildings energy design, Management
of Environmental Quality, vol 17, no 6, pp753762
Richmond, A. K. and R. K. Kaufmann (2006) Energy prices and turning points: The relationship between income and energy
use/carbon emissions, Energy Journal, vol. 27, no 4, pp157180
Santamouris, M. (2001) Solar and natural resources for a better efficiency in the built environment, in Gordon, J. M. (ed)
Solar Energy: The State of the Art ISES Position Papers, James & James, London
Santamouris, M., N. Papanikolaou, I. Livada, I. Koronakis, C. Georgakis, A. Argiriou and D. N. Assimakopoulos (2001) On the
impact of urban climate on the energy consumption of buildings, Solar Energy, vol 70, pp201216
Sayigh, A. A. M. and J. C. McVeigh (1992) Solar Air Conditioning and Refrigeration, Pergamon Press, New York
Schawe, D. (2001) Theoretical and Experimental Investigations of an Adsorption Heat Pump with Heat Transfer between Two
Adsorbers, PhD thesis, Institut fr Kernenergetik und Energiesysteme der Universitt Stuttgart, Stuttgart, Germany
Srikhirim, P., S. Aphornratana and S. Chungpaibulpatana (2001) A review of absorption refrigeration technologies, Renewable
and Sustainable Energy Reviews, vol 5, pp343372
Syed, A., G. G. Maidment, R. M. Tozer and J. F. Missenden (2002) A Study of the Economic Perspectives of Solar Cooling
Schemes, CIBSE National Conference, Part 2, London, October
Terrados, J., G. Almonacid and L. Hontoria (2007) Regional energy planning through SWOT analysis and strategic planning
tools.: Impact on renewables development, Renewable and Sustainable Energy Reviews, vol 11, pp12751287
Tsoutsos, T., J. Anagnostou, C. Pritchard, M. Karagiorgas and D. Agoris (2003) Solar cooling technologies in Greece: An
economic viability analysis, Applied Thermal Engineering, vol 23, pp14271439
Wongsuwan, W., S. Kumar, P. Neveu and F. Meunier (2001) A review of chemical heat pump technology and applications,
Applied Thermal Engineering, vol 21, pp14891519
Wu, X. and Z. Wu (2005) Research on strategic environmental analysis of small hydropower enterprises in China: A case
study of ZhpngDaSanChuan Hydropower Ltd, in IEEE International Engineering Management Conference Proceedings, vol
I, Article number 1559115, pp184189
Yazaki Energy Systems Inc (2007) Performance data at www.yazakienergy.com, accessed 31 October 2007
Ziegler, F. (2002) State of the art in sorption heat pumping and cooling technologies, International Journal of Refrigeration, vol
25, pp450459
ABER_CH_05 6/12/08 3:32 PM Page 157
ABER_CH_05 6/12/08 3:32 PM Page 158
Keywords ventilation, thermal comfort, measurement techniques, temperature, velocity, tracer gas
VENTILATION
People are the customers and consumers of ventilation. Therefore, it is of great
importance to plan the indoor environment with people in mind, and not just building
surfaces and air. The focus must be on people as the final control point, not just the
building itself. This position is necessary in order to develop healthy, realistic and
sustainable indoor environments. In many companies today, the costs involving personnel
are about 100 times higher than energy costs. It is therefore very uneconomical to save
energy at the expense of peoples well-being.
Ventilation refers to the process of introducing outdoor air, distributing the air to
different rooms and finally distributing the air within the room. The purpose of supplying
Abstract
This chapter describes different experimental methods related to the function of
ventilation. Methods of thermal comfort are also treated. A classical, comprehensive
review paper on this subject is found in Hitchin and Wilson (1967). Much information
provided by this paper is still valid, although today the standard technology is digital, not
analogue. In this chapter we examine a range of developments in the field since Hitchin
and Wilsons paper. With arrays of sensors in digital cameras and heat cameras, etc.,
we can record velocities, temperatures and concentrations over whole fields in space
instead of point-wise. Adding tracers that follow air movement is a qualitative
visualization method that has now become a quantitative method for recording velocities
over large fields. Optical techniques such as laser Doppler anemometry are now a
standard technique for recording the velocity vector at a specific point. Passive tracer
gas techniques enable one to take measurements in the field without tying up expensive
and bulky equipment. There has been a shift from pure technical issues to a focus on
people. Concern about air quality has led to development of manikins that can breathe.
Instruments for assessing thermal comfort have been developed that combine the effect
of radiation, velocity and temperature in an index.
6
Experimental Methods
in Ventilation
M. Sandberg, H. Lundstrm, H. O. Nilsson and H. Stymne
ABER_CH_06 6/12/08 3:33 PM Page 159
air is to dilute contaminants generated within the room that may be harmful or annoying
and, finally, to remove them from the room. The first contaminant considered was bio-
effluents, followed by emissions from building materials; today the focus is on small
particles. A prerequisite for air to maintain good hygiene is that the supply of air must be
cleaner than indoor air. It is now acknowledged that heating, ventilation and air-
conditioning (HVAC) systems may be a pollutant source. The contaminant exposure in a
space is a balance between pollutant generation rates and sinks, where the main sink is
ventilation. The amount of air required to reduce the concentration of contaminants is the
hygienic ventilation flow rate and is related to the issue of indoor air quality (IAQ).
Ventilation requirements have been explored in several studies. Figure 6.1 displays the
result from a study carried out at the Danish Technical University.
Figure 6.1 shows that the majority of people have very modest requirements, although
others may be very sensitive. Frequently, ventilation systems have a dual role to play: in
addition to being a means of controlling indoor air quality, ventilation systems are used for
cooling or heating. When used as a cooling system, ventilation flow rates often exceed
hygienic flow rates.
Ventilation systems provide a means of transporting outdoor air into and from a
building. The two main types are mechanical (fan powered) and natural ventilation. With
a fan-powered system, the ventilation flow rate can, at least in theory, to a great extent be
controlled. Natural ventilation relies on outdoor air being introduced into a building
through purpose-provided openings or cracks in the buildings fabric. The engine for this
flow is provided by the kinetic energy in the wind and the difference in potential energy
between columns of air with indoor air density and outdoor air density, respectively. The
ventilation flow rates in a naturally ventilated house are inherently disparate due to varying
driving forces. A combination of mechanical and natural ventilation is known as hybrid
ventilation (Heiselberg, 2006).
Building ventilation rates are important and need to be measured, including the whole
building ventilation flow rate and the flow rate to individual rooms. The flow into a building
160 M. SANDBERG, H. LUNDSTRM, H. O. NILSSON AND H. STYMNE
0
10
20
30
40
50
Ventilation rate (I/s person)
0-1 1-2 2-3 3-4 4-5 5-6 6-7 7-8 8-9 9-10 A
i
r

q
u
a
l
i
t
y

j
u
s
t

a
c
c
e
p
t
a
b
l
e

(
%
)
Source: Clausen (2007)
FIGURE 6.1 Percentage of people who find air quality just acceptable at a given ventilation rate
ABER_CH_06 6/12/08 3:33 PM Page 160
is through the ventilation system and through leakages. An additional path of inflow to a
room is through inflow of unused air from other rooms (i.e. air that has not been
contaminated).
A necessary side condition for a proper supply of outdoor air is that it will not cause
thermal discomfort by creating substantial heat loss from a part of the human body. This
trade-off between the positive effect of diluting the contaminant and increasing velocity
is shown in Figure 6.2. Velocities of the order of 15cm/s can be considered a draught.
Recording such low velocities is a challenge.
Figure 6.2 points out that the first remedy is source control.
THE SYSTEM VIEW
Before ventilation air, taken from outdoors, arrives inside, it passes through a number of
processes that will affect its quality (concentration) and temperature. Figure 6.3 shows the
different parts in the system that must be dealt with when exploring ventilation and
thermal comfort.
People are the customers and consumers of ventilation, and in terms of their
preferences and requirements, are not alike. The room is the enclosure that is closest to
the individual. Ventilation air is supplied to the room, and the shape of the room affects
the motion of air generated by the ventilation system. Through radiation exchange, people
sense surface temperature. The temperature difference between room surfaces and
Experimental Methods in Ventilation 161
Annoyance
Total
annoyance
draught
noise
dry air
particles
smell
gaseous pollutants
Air flow rate
FIGURE 6.2 Sketch of the dependence of concentration and velocity upon ventilation flow rate
FIGURE 6.3 The sequence outdoor ambient to the human
Building
Building
Envelope
Outdoor
Ambient
Room Human
ABER_CH_06 6/12/08 3:33 PM Page 161
room air generates the movement of air that penetrates the occupied zone. Within the
occupied zone, air movement may be caught by upward flowing air, which envelopes
people at a higher temperature than is the case with ambient air.
A building is subdivided into apartments or offices, which, in turn, are subdivided into
rooms. Large airflows are generated through doorways, which act as conduit between
rooms. Flow contact between floors is established by the flow through stairwells. Air then
enters the building through the building envelope through purpose-provided openings or
cracks. In the case of open windows and open doors, large quantities of air flow in and
out of a building. Even if a building is ventilated by a mechanical ventilation system,
ventilation flow rate varies with time (see Figure 6.4).
In a sufficiently tight house, an extract ventilation system creates an under-pressure, and,
in theory, ventilation is not influenced by minor disturbances, with a resulting nearly constant
total flow rate. Figure 6.4 depicts a peak in flow rate during the summer. The cause is
probably the behaviour of the occupant, who leaves windows partly open during shorter or
longer time periods. As a result, the influence of occupant behaviour is an important factor.
With respect to ventilation, occupancy behaviour transforms technically identical houses
into different houses. Figure 6.5 shows the measurements of ventilation airflow rate during
Source: adapted from Stymne et al (2006)
FIGURE 6.4 Weekly average values of ventilation flow rate as a function of a season in a house ventilated
by mechanical extract ventilation
300
250
200
150
100
50
0
Mar
F
l
o
w

r
a
t
e

[
m
3
/
h
]
May Jul Sep Nov Jan Mar
1
0,8
0,6
0,4
0,2
0
0 0,25 0,5 0,75 1
Air change rate [
1
h
]
Source: adapted from Stymne et al (2006)
FIGURE 6.5 Distribution of air change rate in similar naturally ventilated dwellings (unbroken line) and
similar mechanically extract ventilated (dashed line)
ABER_CH_06 6/12/08 3:33 PM Page 162
identical weather conditions in terraced houses, of which 29 were naturally ventilated and
22 extract ventilated. The distribution shown is the cumulative distribution depicting the
fraction of houses with an air change rate less than the value given on the vertical axis.
The fact that technically identical houses are not alike is an important consideration when
making a survey of the ventilation conditions in a building stock. A sufficiently large number
of technically identical buildings must be included in the sample. For example, the near
ambient, located in an urban area with a high building density or a in a rural area, is particularly
important when the building is ventilated through natural ventilation. Planning experiments
and selecting suitable instrumentation will necessitate that one knows beforehand the
ambient that will be recorded. Therefore we will, in order of priority, go through the
characteristics of the individual parts in the sequence depicted in Figure 6.3. We only deal
with spaces that people normally inhabit and therefore exclude areas such as crawl spaces.
PEOPLE
MICROCLIMATE AND TEMPERATURE REGULATION
The human bodys temperature regulation system permits physiological adjustment to
thermal stress and provides thermal comfort under a variety of conditions. Heat is
primarily produced through metabolism, which emanates from digestion and muscle
exercise. In normal conditions, this results in an average deep body temperature of about
37C. Within a certain interval, the bodys temperature control system strives to maintain
this temperature when internal or external thermal disturbances arise. The effects
resulting from a change in climatic conditions produce a reaction within the body
concerning physiological autonomic responses, but also activate behavioural regulation.
Autonomic regulation is controlled by the hypothalamus, which regulates the different
avenues of heat loss.
The human bodys temperature-regulating centre works in a similar fashion to a
thermostat. Temperature set point may change during different physiological conditions.
Source: Hans Lundstrm University of Gavle
FIGURE 6.6 (Left) A person simulator produces 100W of heat; (right) the photographer Magnus Mattsson
ABER_CH_06 6/12/08 3:33 PM Page 163
Thermo-sensitive receptor nerve endings send signals to the regulation centre. In a cool
environment, stimulation of the cold receptors may lower the set point, and heat loss can
decrease by means of vasoconstriction and shivering. In a warm environment, on the
other hand, the set point becomes elevated and heat loss can increase by means of
vasodilatation in the skin and sweating. The temperature regulatory centre is connected
to receptors in the skin, as well as in the deep body core and the brain itself. These
receptors consist of a net of fine nerve endings that are specifically activated by heat or
cold stimuli. These temperature receptors are especially sensitive to rapid changes in
temperature and are highly susceptible to adaptation. The number of active receptors
determines, to some extent, the sensation of temperature (strand and Rodahl, 1986).
A person at rest produces approximately 100W of heat. If clothing and ambient
conditions are adequate, the same amount of heat is lost to the environment. The whole
body heat balance is therefore in equilibrium, and the individual feels thermally neutral.
With increasing ambient temperature, convection and radiation diminish. Sweat
evaporation has to compensate for this in order to balance the heat production of 100W.
Sweating is associated with a sensation of warmth and, eventually, discomfort. In a cooler
environment, convection and radiation increase, producing total heat losses greater than
100W. The physiological response is to reduce skin and extremity blood flow in order to
lower the external temperature gradient. In this case, the individual feels cool or cold and
uncomfortable. A normal response is to add or remove clothing.
In other words, for comfortable climatic conditions, dry heat loss can only vary within
a certain narrow range. In a similar way, the heat balance of a skin segment can be
analysed. Optimal local heat balance and a comfortable skin temperature can only be
maintained within a specific span of convective, radiative and conductive heat loss.
Higher local heat losses will be felt as cool or cold, and lower local heat losses as warm.
The human bodys response to the thermal environment has been found to depend
primarily upon six factors (Parsons, 1993):
1 air temperature;
2 mean radiant temperature;
3 air velocity;
4 relative humidity;
5 physical activity; and
6 thermal resistance of clothing.
These factors are critical in steady state conditions. If the exposure is short or intermittent,
the length of the exposure can also have an impact. The final decision on whether a
combination of these factors represents a comfortable situation or not depends upon the
sensory reception of the climate. In order to produce correlations between this human
thermo-sensation and different climatic disturbances, it must be assumed that the human
body cannot differentiate between the heat loss sensations of draughts, radiation or
conduction.
Body build, sex, age, geographical differences, food and beverages, as well as several
other environmental factors, also influence the perception of climatic comfort. The effect
ABER_CH_06 6/12/08 3:33 PM Page 164
of these components is not taken into consideration in any of the above indices. They may
have an impact upon the perception of thermal comfort and eventually contribute to the
occasionally significant individual variation in response.
TWO DIFFERENT WAYS OF LOOKING AT THERMAL COMFORT
Two different types of models are widely known for predicting steady-state thermal
comfort. One model is based on the heat balance of the human body. Another approach
assumes, to a certain degree, an adaptation to the thermal environment:
1 Predicted mean vote (PMV) model: a method for calculating a steady-state thermal
comfort index derived from heat balance calculations and climate chamber studies.
The model assumes a relation between optimal thermal conditions, using the steady-
state heat balance equation for the human body and thermal comfort ratings from
panels of subjects.
2 Adaptive principle: a relation for steady-state thermal comfort obtained from large
field studies. The model assumes that people will adapt to thermal conditions, using
only indoor and outdoor temperatures as dependent variables.
THE PREDICTED MEAN VOTE (PMV)
PMV (Fanger, 1970) describes the conditions for climatic comfort, and methods and
principles for evaluating and analysing different environments from a climate comfort
point of view, with the indices of predicted mean vote (PMV) and predicted percentage
dissatisfied (PPD). The concept defines conditions that must be fulfilled in order for a
person to be in whole body climate comfort. An additional condition is the absence of
any local climate discomfort. The intention was to develop a comfort equation where just
the above-mentioned six factors are needed in order to calculate in which thermal state a
normal person is found. These calculation and estimation methods now comprise an
international standard (ISO 7730, 2006) and are used all over the world.
The model is, in essence, a regression equation that relates PMV on the seven-point
thermal sensation scale of a group of people exposed to a certain environment to the
calculated result of the basic heat balance equation. This equation uses the heat balance
for the human body and assumes a connection between the deviation from optimal
thermal balance and thermal comfort vote. The greater the deviation, the more the
comfort vote departs from zero. The PMV equation is primarily a steady-state model. It is
a semi-empirical equation for predicting the mean rating on an ordinal rating scale of
thermal comfort for a group of people. PMV is currently the most widely used thermal
comfort index. ISO Standard 7730 uses limits of PMV and PPD, as well as local
recommendations, as definitions of a comfort zone.
THE ADAPTIVE PRINCIPLE
The premise of the adaptive approach is expressed by the adaptive principle:
If a change occurs that produces discomfort, people react in ways that tend to restore
their comfort.
ABER_CH_06 6/12/08 3:33 PM Page 165
The adaptive principle consequently suggests that people will adapt to certain climatic
conditions. For instance, in warmer climate, when environmental thermal comfort
parameters point at a higher PMV, people will become adapted to the higher temperatures
and still feel comfortable. As an analogy, a corresponding PMV principle could be
written as:
No change should occur that produces discomfort. In a properly designed environment,
will people maintain their comfort?
The adaptive principle (Nicol and Humphreys, 2001) is supported by field comfort studies
in many different environments. From these global field studies of thermal comfort ratings
and temperatures, Humphreys (1976) found that comfort temperature differed between
groups of people. This means that comfort conditions calculated with the heat balance
models did not fully agree with the comfort conditions found in the field. By linking the
comfort rating to human behaviour, the adaptive principal links the comfort temperature
to the climatic situations in which individuals find them. Comfort temperature is a result
of the interaction between the subjects and the thermal environment. Nicol and
Humphreys (2001) also conclude that people with more opportunities to adapt
themselves to the environment will be less likely to suffer discomfort.
The adaptive model is essentially a regression equation that relates the desired
temperature indoors to the monthly average temperature outdoors. The only input
variable used is the average outdoor temperature, which has an indirect impact upon the
human heat balance. Consequently, the adaptive model does not include the six classical
thermal parameters that have an impact upon the human heat balance and, therefore,
upon thermal sensation.
CLIMATE INDICES WITH SIMILAR PRINCIPLES
A number of indices have been developed, all according to similar methods, by
transforming the real environment to a standard environment, which gives the same
thermal climate experience:
effective temperature (ET);
resultant temperature (RT);
corrected effective temperature (CET);
new effective temperature (ET*, or ET star);
standard effective temperature (SET);
PMV* (PMV star).
EQUIVALENT EXPERIENCED TEMPERATURE
Equivalent temperature (ET) is a recognized measure of the effects of non-evaporative
heat loss from the human body (Dufton, 1936, Madsen et al, 1984, Nilsson, 1999). It is
particularly useful whenever complex interactions of various heat fluxes are present.
Equivalent temperature is derived from the operative temperature by including the effect
of air velocity on a heated body.
ABER_CH_06 6/12/08 3:33 PM Page 166
The well-known operative temperature (t
op
) only considers air temperature and mean
radiant temperature, and is defined for the actual air velocity, whereas equivalent
temperature (t
eq
) is defined for a standard low air velocity.
One advantage of t
eq
is that it expresses the effects of combined thermal influences
in a single figure, and is easy to interpret and explain. It is particularly useful for differential
assessment of climatic conditions. However, the underlying hypothesis is that the t
eq
value
always represents the same subjective response irrespective of the kind of combinations
of heat losses.
As mentioned earlier, it is often difficult to explain the combined effects of different
heat losses to people. It is therefore very useful to convert these values into something
easier to understand, such as experienced temperature, which is a more straightforward
concept. This equivalent experienced temperature (t
eq
) is then calculated with Equation 1
according to (Nilsson, 2004a):
[1]
where:
q
T
= measured manikin heat loss during the actual conditions (W/m
2
);
R
T
= total insulation, seated, no clothing (m
2
K/W);
t
s
= manikin surface temperature (C);
t
eq
= equivalent temperature of the uniform, homogenous environment (C).
The equivalent temperature in Table 6.1 shows low values, as could be expected. The
results should be compared to the climate of an unclothed person, sitting in the air
stream during the same conditions as the manikin. The definition today (Nilsson and
Holmr, 2003) reads: the equivalent temperature (t
eq
) is the temperature of an imaginary
enclosure with the mean radiant temperature equal to air temperature and still air in which
a person has the same heat exchange by convection and radiation as in the actual
conditions.
t t R q
eq s T T
=
"
TABLE 6.1 Description of the connection between the measured
quantities and equivalent temperature
ACTION INFLUENCE
Increased air speed Lower t
eq
Decreased air temperature Lower t
eq
Decreased mean radiant temperature Lower t
eq
Decreased air speed Higher t
eq
Increased air temperature Higher t
eq
Increased mean radiant temperature Higher t
eq
ABER_CH_06 6/12/08 3:33 PM Page 167
THE ROOM
The flow in a room can be subdivided into two regimes: primary airflow generated by the
air from the supply device and secondary flow generated by entrainment of air into the
primary airflow. The primary air stream is a jet when isothermal air is supplied, and a
buoyant jet when ventilation air is heated or cooled. If supplied along a horizontal surface
with a low velocity, it becomes a gravity current if warm air is supplied at the ceiling or
cool air is supplied at floor level. The primary air stream can further be subdivided into a
near field region and a far field region. Within the near field region, the supply device
leaves a footprint on the flow, whereas in the far field region, details of the effect of the
supply devices have disappeared. A room is an enclosure that constrains airflow, and a
characteristic feature of room airflow is the occurrence of recirculation zones. Outdoor air
is limited to the supply of outdoor air through the ventilation systems; but internal airflow
rates can be much larger. A characteristic feature of room airflow is the strong interaction
between airflows generated by different sources.
Heat transfer occurs at surfaces with a temperature that varies from room air
temperature; there is also an exchange of heat by radiation between surfaces of different
temperatures. Walls are lined by boundary layer flows, which are seldom fully developed,
but exist in a transition between laminar and turbulent flow. An upward flowing boundary
layer flow surrounds individual people, starting at the bodys ankles.
In general, airflow in a room is not fully developed, which makes it difficult to predict
flow according to computational fluid dynamics. Most of the frequency content of
turbulent kinetic energy is below 2Hz.
actual non-uniform
environment
uniform enclosure
t
eq
= 24
C
t
r
= 26
C
v
a
= 0.2 m/s
t
a
= 22
C
t
r
= t
a
=t
eq
= 24
C
v
a
0 m/s
R + C R + C
_
=
Note: It is assumed that the posture, the activity level and the clothing are the same in both environments. The dry heat
exchange, as well as the equivalent temperature, becomes the same in both environments. The heat exchange can, of course,
be negative in warm environments and positive in cold environments.
Source: Nilsson (2004a)
FIGURE 6.7 A thermal manikin is exposed to two different environments, one actual with non-uniform
and one imaginary with uniform climatic conditions
ABER_CH_06 6/12/08 3:33 PM Page 168
SUPPLY AIR TERMINALS
Supply of air to a room can occur through a ducted system with supply air terminals at the
end of the duct. The other alternative is ventilation through vents mounted in the walls,
which is used in connection with mechanical extract ventilation and natural ventilation.
Figure 6.8 shows supply air terminals for mechanical ventilation. With respect to supply
velocity B
inlet
one makes a distinction between high-velocity supply devices, medium-velocity
devices and low-velocity devices. A high-velocity supply can amount to 10m/s. A medium-
velocity supply device or low-velocity supply device is used when supplying air directly to the
occupied zone. Local ventilation occurs where air is supplied directly towards peoples faces
from a ventilation supply device located, for example, in a computer (Melikov, 2004). A new
type of local ventilation is where air is blown from a perforated cushion around the neck of
a person and upwards (Nielsen et al, 2007; see Figure 6.8(d)).
PARAMETERS CHARACTERIZING SUPPLY DEVICES AND AIR STREAMS
FROM SUPPLY DEVICES
The parameter characterizing an isothermal supply is the flow force (momentum flux)
(Newton). The specific momentum flux is the flow force per unit mass:
[2]
where q is the volumetric flow rate and A
inlet
is a characteristic opening area of the supply
terminal. The momentum flux of supply air terminals with a flat front can be determined
by blowing against a balance and recording the reaction force (Nordtest, 2000). The
Reynolds number is defined as:
[3]
where [m
2
/s] is the kinematic viscosity.
The parameter characterizing a non-isothermal supply with temperature difference T
(K) between supply air and room air is the specific buoyancy flux B:
[4]
where g (m/s
2
) is the acceleration of gravity.
Alternatively the specific buoyancy flux can be express in the load E
C
(W), removed by
the ventilation system:
[5]
where C
P
is the specific heat at constant temperature and is the density.
When the ventilation air is used for cooling or heating and supplied with a low velocity,
we obtain a buoyant jet. In the beginning, the momentum flux dominates; but the
buoyancy flux gradually becomes more important. When supplied vertically, the transition
B g
E
C T
C
p
=
B qg
T
T
m
s
=

l
l
l
l
l
4
3
Re
Inlet
= =
U A
m

m q UB A UB = = ( )

l
l
l
l
l
Inlet Inlet Inlet
2
4
2
m
s
ABER_CH_06 6/12/08 3:33 PM Page 169
(a)
(b)
(c) (d)
FIGURE 6.8 Supply air terminals: (a) High-velocity supply; (b) medium-velocity supply; (c) low-velocity
supply intended for displacement ventilation; (d) local ventilation
ABER_CH_06 6/12/08 3:33 PM Page 170
from momentum driven to buoyancy driven is given by the thermal length, sometimes
called jet length, l
T
:
Round buoyant jet Plane buoyant jet (in metres). [6]
The thermal length is only dependent upon the flow force of the supply air and the heat
to be removed by the ventilation system.
With no buoyancy, B = 0 and the thermal length is infinite.
The physical meaning of the thermal length is illustrated in Figure 6.9. Cold air is supplied
upwards. Initially, the momentum dominates; but the buoyancy directed downwards
gradually becomes relatively stronger, and when it takes over, the flow turns downwards.
Pure buoyancy sources are heat sources and sinks. Pure buoyancy sources give rise to
plumes. A human body is a heat source; an example of a heat sink is a passive chilled beam.
VELOCITY IN THE OCCUPIED ZONE
The highest velocity occurs in the primary air stream, where predicting velocity is of the
utmost importance with respect to the risk of draught. In a room, airflow is constrained
by room surfaces. Therefore, the room length L in the direction of the flow and the room
width W are significant parameters. In a long room, the jet from a ceiling-mounted supply
device may separate from the ceiling before it reaches the opposite wall. If it reaches the
opposite wall, it follows the wall downwards and continues along the floor and a backflow
is generated. Between air streams, there is an exchange of momentum (for an example,
see Carrilho da Graca and Linden, 2003). The ratio between the velocity in the room and
l
T
m
B
=
2 3 /
l
T
m
B
=
3 4
1 2
/
/
.
FIGURE 6.9 Interpretation of thermal length
g
q
1,85 l
T
1,85 l
T
g
q
ABER_CH_06 6/12/08 3:33 PM Page 171
the inlet velocity can be written as
[7a]
For isothermal supply the number of parameters is reduced:
[7b]
For small Reynolds numbers, there is a dependence on the Reynolds number (T. Valentino,
personal communication). In particular, the length of the potential core (region of the jet
with supply velocity) of the jet is a function of the Reynolds number. With increasing
Reynolds number, the length decreases to finally attain the asymptotic value of about six
times the characteristic length of the supply terminal. This may be important in
relation to personal ventilation.
The second parameter reflects the constraint by the room length L on the
development of the flow. As mentioned earlier, the length of the room must be larger than
six times the characteristic length of the supply air terminal in order for a jet to be
developed. Some modern supply devices are so large (Elvsn and Sandberg, 2008) that
the flow discharged from the supply device will not reach an asymptotic state.
The third parameter provides information about the forces dominating from the
beginning, the momentum of the supply air or the buoyancy of the supply air. It can be
expressed in terms of supply velocity, characteristic length of the supply device and the
temperature difference:
[8]
where Ar
s
is what is called the densimetric Archimedes number of the supply within
ventilation engineering. It is used in the design of ventilation (Nielsen, 2007). In fluid
mechanics, this parameter is called the densimetric Froude number. If this parameter is of
the order of 1, the flow is dominated by buoyancy from the very beginning. One example
is displacement, where cold air is supplied with a low-velocity diffuser.
The fourth parameter is the ratio of the width b of the supply to the width of the room.
When the width of the flow begins to approach the width of the room W, the further
development of the flow will be constrained by the room.
For isothermal flow, the increase in width is proportional to the entrainment coefficient
. For an axisymmetric jet with an entrainment coefficient of 0.0535, the growth in width is:
[9]
For a flow dominated by buoyancy, the spread is controlled by the density difference.
Figure 6.10 shows how a buoyant flow (cold air) discharged from a low-velocity air
b x x x ( ) = = 2 0.107 .
l
T
Inlet
Inlet
Inlet
s
A
U
g
T
T
A
Ar
=
=
1
A
Inlet
U x
U
f
L
A
b x
W
R
Inlet
Inlet
( )
=
( )
\
)
Re, , .
U x
U
f
L
A A
b x
W L
R
Inlet
Inlet
T
Inlet
T
( )
=
( )
\
)
Re , , , .
l l
ABER_CH_06 6/12/08 3:33 PM Page 172
terminal is spread. The tests were conducted in a model with water as the operating fluid
and salt was added to the supply to obtain a density difference. The point X
D
indicates the
position where the air from the supply air terminal hits the floor. From the point of impact
the air is spread as a gravity current on the floor.
It is reasonable to assume that the front is spread with a velocity proportional to
where h is the local thickness of the air stream:
[10]
where k is a parameter to be determined experimentally.
The last parameter is the ratio between the thermal length and the length of the room
in the direction of the supply. There is always a risk that in the case of a supply of cold air
at ceiling level, the jet may separate from the ceiling and descend into the occupied zone
and cause a draught. It is difficult to predict where it will separate. However, in order to
avoid separation from the ceiling, a rule of thumb is that the thermal length should be
much larger than the room length.
Figure 6.11 shows the velocity at a point in the occupied zone as a function of the
supply flow rate. The supply air terminal is a high-velocity supply air terminal located at
ceiling level. With a supply of air that features the same temperature as room air, velocity
increases linearly with the flow rate (supply velocity). If we need to heat the room with
warm air, the supply air remains under the ceiling at low flow rates, which gives rise to
b x t k g
T x t
T
h x t dt
t
( ) ( )
=
( ) ( )
( ) ( )
0
g
T
T
h

Source: Sandberg and Mattsson (1993)
FIGURE 6.10 A flow dominated by buoyancy (Ar
s
= 0.97): Overhead view of position of the front on the floor
ABER_CH_06 6/12/08 3:33 PM Page 173
low velocities. With the supply of isothermal air or warm air, the velocity in the occupied
zone is less than the supply velocity. If we supply cold air with a low velocity, the velocities
in the occupied zone will be high due to the tendency for the colder and heavier air to
descend into the occupied zone.
In displacement ventilation, buoyancy dominates and the maximum velocity in the
occupied zone close to the supply device may be greater than the supply velocity.
A CHARACTERISTIC ROOM VELOCITY SCALE FOR BUOYANCY-DRIVEN FLOW
Horizontal flow driven by buoyancy is comprised of gravity currents (see Figure 6.12).
A characteristic room velocity scale based on the specific buoyancy flux and the room
width is:
[11]
This velocity scale can be used to assess the velocity in the room generated by displacement
ventilation (Sandberg and Etheridge, 1996, p449; Cehlin and Sandberg, 2007).
U
B
W
R
=

(
\
)
1/3
(in m/s).
Isothermal
Heating
Supply velocity
V
e
l
o
c
i
t
y

i
n

t
h
e

o
c
c
u
p
i
e
d

z
o
n
e
Cooling
FIGURE 6.11 Velocity at a point in the occupied zone where there is a supply of isothermal heated air and
cold air
Radial
U
R
[m/s]
Two-dimensional
W
FIGURE 6.12 Assessment of velocity in a room ventilated by displacement ventilation
ABER_CH_06 6/12/08 3:33 PM Page 174
This velocity scale is only dependent upon the width of the room and the load to be
removed by the ventilation system. It can therefore be calculated in the early design
stage. If the velocity is larger than the comfort criteria required, a displacement ventilation
system cannot be used.
TIME CONSTANT OF THE VELOCITY FIELD IN THE ROOM
Knowledge of the velocity fields time constant is necessary for determining sampling rate
and integration time (number of samples) required to record the mean velocity and
standard deviation of the velocity fluctuations within a certain accuracy. The time constant
is a measure of the memory of the velocity field. Samples should be independent to
provide new information. If the time separation between two samples is twice the time
constant, they are regarded as independent. Sampling more frequently does not yield
additional benefits. The time constant can be determined from the autocorrelation function
of the velocity fluctuations. Figure 6.13 displays the recorded time constant in different
parts of a room. Within the occupied zone, the time constant can be as low as 1s.
The inset in Figure 6.13 shows an example of two energy spectra derived from velocity
fluctuations. The mean velocity is the same, but the frequency content is different. To be able
to retrieve information about the frequency content of velocity fluctuations without distortion,
the velocity signal must first pass a low-pass filter. Through this process, the frequency
content is limited to a maximum frequency f
max
. To avoid aliasing the sampling interval, the
filtered signal should, according to the Whittaker-Shannon theorem, be less than 1/(2 f
max
).
THE VENTILATION PROCESS
Under ideal conditions in an airtight room, outdoor air is supplied to the room from a well-
defined location and extracted from the room at a well-defined point. Furthermore, the
Source: Sandberg (1994)
FIGURE 6.13 Time constant
ABER_CH_06 6/12/08 3:33 PM Page 175
assumption is made that the contaminants released within the room do not react with each
other. The resulting concentration due to a release of contaminant in the room depends
upon both how the air and the contaminant are distributed within the room (see Figure 6.14).
The distribution of room air is quantified by its age at each point in the room. The local
mean age of air
p
at an arbitrary point P describes the length of time that the flow of air
molecules passing by P has, on average, spent within the building. The local mean age
of air within and leaving the room is independent of the airflow pattern and is equal to:
[12]
where V is the volume of the room. The volume divided by the flow rate is also called the
nominal time constant of the ventilation system. Since there is a local mean age of air at each
point, it is meaningful to define the mean age () of the whole volume of air within the room.
The time
t
, that it takes, on average, to replace the air within the room is:
[13]
The flow rate, q (kg/s), of outdoor air provides the dilution capacity in the sense that the
lowest concentration attained due to a local release with a rate m
*
in the room is:
[14]
This relation is only based on mass conservation (what goes in, goes out) and holds true
irrespective of airflow pattern within the room. This concentration is attained in the extract
point where all contaminants released within the room are mixed with ventilation air.
C
m
q
=
*
.

r
= 2 .
e
=
V
q
q
q
m
C
o
n
t
a
m
i
n
a
n
t
A
ir
C=
m
q
e
=
q
V
FIGURE 6.14 Concentration depends upon both how the contaminant and the air are distributed
ABER_CH_06 6/12/08 3:33 PM Page 176
An ideal state is uniform mixing where the concentration everywhere is equal to the
concentration in the extract.
At steady state, the total amount M (kg) of contaminants in the room (according to
chemical engineering parlance) is given by the equation:
[15]
where
C
is the hold-up time, which is the same as the mean age of contaminant leaving
the room. Equation 15 provides a recipe for air-quality control: to minimize the amount of
contaminant in the room, the contaminant should be removed as fast as possible and not
be spread throughout the room. This is the basis for using exhaust hoods (see Figure 6.15).
However, if a spread of the contaminant cannot be avoided, there are two strategies.
One is to mix it with the ambient air in order to attain the minimum concentration
(Equation 14): this is the idea behind mixing ventilation. The other principle is
displacement ventilation, which is based on the principle of stratified flow caused by the
density difference between layers of air with different temperatures. This air distribution
principle can be used when there is a need for cooling. There are two layers in the room:
one lower layer of air with supply air quality and an upper layer with polluted air. The height
of the interface depends upon the heat load and the flow rate. Figure 6.16 displays a room
ventilated by displacement ventilation. Two people sitting down were simulated. A tracer
M m
C
=
i
/
FIGURE 6.15 Kitchen exhaust hood used for laboratory test of capture efficiency: Visualization with smoke
and a laser light sheet; oil is sprayed into a hot pan by means of a spray-nozzle
ABER_CH_06 6/12/08 3:33 PM Page 177
gas source was located on the body of the person to the right. The concentration
distribution is shown. The reference concentration is the concentration in the extract,
which has been put equal to 1.
In Figure 6.16, the two layers appear clearly and there is no contaminant in the lower
layer. If the room had been ventilated by a mixing ventilation system, the concentration
should (apart from a region close to the source with enhanced concentration) have been
equal to 1 everywhere.
THE RELATION BETWEEN THE LOCAL MEAN AGE OF AIR
AND THE CONTAMINANT CONCENTRATION
For one and only one source there is a direct correspondence between the local mean age of
air. If a passive contaminant is generated in each point of the room with the same rate
everywhere then the concentration at point P is proportional to the local mean age at that point
[16]
where M
is the total release rate within the room. This relation is the theoretical
background to the homogeneous emission method for recording the local mean age.
THE BUILDING
FLOWS BETWEEN ROOMS
Rooms stay in contact with each other through flows generated by the density difference
of air masses or by pressure differences introduced as a result of mechanical ventilation
or wind loading. Flow generated by a density difference can be subdivided into two cases.
In bulk density flow, the greatest difference in density (temperature) is between the two
C
V
P p
=

Source: Stymne et al (1991)
FIGURE 6.16 Concentration recorded in a room ventilated by displacement ventilation
ABER_CH_06 6/12/08 3:33 PM Page 178
rooms. The other case is boundary layer flow, where the main difference in density is
between room air and room surface. Boundary layer flow occurs when one has a room-
dominating heat source or a heat sink in the room. In the bulk density case, the only
source of buoyancy is due to different room air temperatures. The room air surfaces are
assumed to be in thermal equilibrium with room air and have attained the same
temperature. These conditions can be achieved by carrying out experiments with water
as an operating fluid and adding salt to the water to obtain a density difference
corresponding to the density difference in air. This experimental method is called the salt
bath method and is a simple method to study flows generated by density difference. An
advantage of using water is that the flow can easily be visualized by adding various
coloured dyes. Different parts can be tinted simultaneously with different colours.
Fluorescent dye lit up with a laser light gives a clear picture of the flow. In the experiment
depicted in Figure 6.17, salt has been added to the water in the compartment on the right-
hand side. The door between the adjacent rooms has suddenly been opened. A
bidirectional flow with the heavier fluid (the cooler air) moves towards the left as a gravity
current, whereas the lighter, heavier fluid moves towards the right.
Figure 6.18 shows a boundary layer flow generated by a heat source in a buildings
hall. The boundary layer flow passes into an adjacent room as a unidirectional flow where
it is spread as a three-dimensional gravity current.
Density differences between floors in a house give rise to a bidirectional flow. Figure 6.19
shows the flow in a stairwell studied with the salt bath method. The flow is visualized with
the shadowgraph method (Settles, 2001), which makes the density difference visible. On
the reverse of the model there is a transparent paper so that a distant light source, such
as a projector, can render the density difference visible.
Flows between floors are dealt with in Heiselberg and Li (2007).
Source: Claes Blomqvist, University of Gavle
FIGURE 6.17 Bidirectional bulk density flow through a doorway studied in a two-dimensional model: The
cooler air moves towards the left as a gravity current along the floor
ABER_CH_06 6/12/08 3:33 PM Page 179
DIFFERENT ROUTES FOR THE SUPPLY OF OUTDOOR AIR TO A ROOM
A building consists of many rooms and, therefore, many individual zones. In terms of
airflow within a building, a zone is a space that is sharply marked off from its surroundings
with respect to the transport of air. Normally, a zone is physically marked off from its
surroundings by partition walls. Zones can also be established by pressure control
achieved by supplying air.
Figure 6.20 displays the different routes for flow of outdoor air to a room. The top
image depicts direct supply of outdoor air through a ventilation system. In the middle is
shown air supply through the ventilation system and infiltration through external walls.
However, outdoor air can also arrive from other rooms. What matters is whether or not
the air has a dilution capacity when arriving inside. It has a dilution capacity if it is
unused air that is, it has not been used for dilution of a contaminant released in the
actual room. The bottom image shows the purging flow rate U, which is the flow rate of
Source: Blomqvist (2008)
FIGURE 6.18 Unidirectional boundary layer flow through an opening
Source: Blomqvist and Sandberg (2004)
FIGURE 6.19 Flow in a stairwell studied with the salt bath method: The density difference is made visible
with the shadowgraph method
ABER_CH_06 6/12/08 3:33 PM Page 180
outdoor air to the ventilated space that is available for dilution of contaminants in the
actual room.
According to the multi-zone theory, the purging flow rate can be expressed as follows
(Sandberg, 1984):
[17]
where:
P
ij
= the transfer probability from zone j into zone i;
q
i
= flow rate of outdoor air directly to zone i.
Within a room there is strong recirculation; nevertheless, there is a purging flow of
particles that never return. The maximum purging flow rate is equal to the total airflow
rate to the building.
The supply of outdoor air through a ventilation system is recorded in the duct. The
supply of air through a ventilation system, as well as infiltration, can be recorded with
the tracer gas technique, otherwise known as the constant concentration method. The
purging flow rate can be obtained by a constant release of tracer gas and recording the
equilibrium concentration.
U q P q
i i ij j
j i
=

q
q
q
FIGURE 6.20 A ventilation system, a ventilation system with infiltration, and purging flow rate
ABER_CH_06 6/12/08 3:33 PM Page 181
OUTDOOR AMBIENT
The ambient surrounding a building is important with respect to wind and pollutants.
When a house is provided with a natural ventilation system, wind environment is
important. One engine for natural ventilation is the kinetic energy of wind. The outdoor
ambient climate affects both wind speed and wind profile. If the building is located in
open terrain or in an urban area, this is an important consideration. Pollutants in the
ambient (e.g. small particles), pollen and exhaust from vehicles may penetrate the building
and affect indoor air quality. Pollutant distribution around buildings should be considered
when choosing the location of ventilation intakes.
WIND ENVIRONMENT
The logarithmic mean wind profile over terrain is as follows (Stull 1988):
and [18]
friction velocity [19]
where:
z
0
= roughness height;
D = zero-plane displacement height;

w
= surface stress;
= von Karman constant.
Values of between 0.41 and 0.37 have been reported in published papers.
The roughness height z
0
is defined as the height where the wind speed becomes zero.
Protrusion of objects (e.g. buildings, forests, etc.) above a surface displaces the entire
flow upwards. To account for this, a displacement height D is introduced. Historically, the
zero-plane displacement height was introduced in order to be able to retain the
logarithmic profile over a rough surface (see Figure 6.21).
U
w
U z
U
z D
z
( ) =

\
)
e
0
z
0
z
0
+D
0
D
h
Logarithmic profile
U(z)
Transition layer
Mean building
height
Source: adapted from Meystayer in Fenger et al (1998)
FIGURE 6.21 Definition of displacement height (D) and roughness height (z
0
)
ABER_CH_06 6/12/08 3:33 PM Page 182
In a wind tunnel, the atmospheric boundary layer (ABL) flow can be reproduced. As a
result, a wind tunnel can serve as an experimental tool for exploring the effect of the
ambient on the pressure on buildings or the flow through openings. Models of buildings
and the ambient are typically reproduced on a scale of between 100th to1000th of the
original. Wind tunnel technique is an established experimental technique. The ABL over
different types of terrain is generated by roughness elements and spires (see Figure 6.22)
(Blomsterberg et al, 2007).
The properties of the boundary layer are quantified in terms of the shape of the vertical
velocity profile and roughness height and friction velocity. The roughness height is
determined from the vertical velocity profile, and the wall shear stress can be recorded with
a Preston tube (Goldstein, 1996). There are standards for doing wind tunnel tests (VDI, 2000).
FLOW THROUGH THE ENVELOPE
A flow (infiltration) occurs through the building envelope as a result of leakages consisting
of small cracks or small holes. To the list of parameters in Equations 7a and 7b that deal
with mechanical supply, we must now add another parameter, t/A, where t is the length
(wall thickness) and A is the linear size of the opening area A. Figure 6.23 shows a wind
tunnel study of wind-driven flow through openings. The arrow indicates wind direction.
For small t/A, the opening is a sharp-edged opening; but for large t/A, the opening
is a duct and the wind is ducted into the room. The flow depicted in Figure 6.23 has been
visualized by spraying semolina powder on the floor. The sand erosion method is
thoroughly dealt with in Dezi (2006).
Source: Leif Claesson, University of Gavle
FIGURE 6.22 Wind-tunnel test of a hybrid-ventilated school at a scale of 1:200: In the background is
shown the spires and the roughness elements
ABER_CH_06 6/12/08 3:33 PM Page 183
MEASUREMENT OF HEAT FLUX FROM THE HUMAN BODY
MEASUREMENTS WITH THERMAL MANIKINS
A heated thermal manikin represents the ultimate heat flux transducer. The whole manikin
surface is heated to, and controlled at, the same skin temperature as the human body
t/A=0.16
t/A=1.29
Source: Leif Claesson, University of Gavle
FIGURE 6.23 Visualization by the sand erosion method of wind-driven flow through openings
Source: Nilsson (2004a)
FIGURE 6.24 The heated full-scale thermal manikin with 33 individually controlled zones: This manikin
was especially constructed for climate evaluation in 1991
ABER_CH_06 6/12/08 3:33 PM Page 184
surface. The early manikins comprised only one zone. Todays manikins provide 16, 20, 35
or even more individually heated and controlled zones. To obtain a single figure for t
eq
,
local heat losses are added and the total value divided by the total body surface area gives
the dry heat loss to be used for the t
eq
computation.
Manikin with constant surface temperature
In this case, the surface temperature of the manikin is uniform and constant
(approximately 34C) over all zones.
Manikin with distributed surface temperature
Here, the manikin has a distributed temperature over all zones. Usually, the temperature
declines from the torso to the extremities.
Manikin with constant heat loss
This manikin type operates with constant heat loss across the different zones. Small radial
as well as inward axial heat losses may influence the measurements by this technique.
Manikin with adaptable surface temperature
Here, the surface temperature of the manikin is allowed to change as a function of dry
heat loss using an expression derived from the basis of comfort criteria.
Manikin with heated sensors
This method measures the surface temperature of the sensors at constant heat loss with 18
small sensors on a manikin body, each containing two heated surfaces with two different
power levels. The surface temperature of the two surfaces with known heat loss is measured
and a linear model is used to calculate the surface temperature for an unheated surface.
SIMPLIFIED OR ADVANCED MANIKINS
Researchers around the world have developed many different geometrical configurations
in order to simulate a human body through a manikin. These manikins are different with
Source: Magnus Mattsson, University of Gavle
FIGURE 6.25 Dummies simulating the heat generated by a person
ABER_CH_06 6/12/08 3:33 PM Page 185
respect to size, form, heat generation, number of zones, regulation principles, etc. The
evolution of these manikins is shown in Figures 6.25 to 6.27.
Simple
Figure 6.25 shows a simple but still efficient dummy for simulating the heat generated by
human occupancy. In this case, there is constant heating with no regulation or measurement
of heat loss (Mattsson, 1999; see also Figure 6.6, where the temperature of a dummy and
a real person is compared). These dummies are relatively cheap, and a large number can be
manufactured to simulate a group of individuals for example, as a class of students.
Standard
Figure 6.26 shows Comfortina, a thermal manikin of Nille type belonging to the University
of Aalborg, Denmark. Comfortina has individually controlled segments enabling
assessment of local thermal comfort or discomfort. It measures heat loss or equivalent
temperature, with up to 20 independently controlled zones (Nilsson et al, 2007).
Advanced
Figure 6.27 displays an advanced automotive manikin (ADAM), which measures heat loss
at 120 independently controlled zones. Manikin data is transferred to a computer model
that simulates human thermoregulatory responses in real time. Still another computer
model predicts human thermal comfort (Fan, 2006; see also www.nrel.gov/vehicles
andfuels/ancillary_loads/adam.htmlUT).
FIGURE 6.26 Manikin with 20 individually controlled segments
Source: Nilsson et al (2007)
ABER_CH_06 6/12/08 3:33 PM Page 186
The variations do, of course, reflect different subjects of interest, such as effects on
airflow, thermal comfort, as well as pollutant production and exposure.
OTHER HEAT FLUX MEASUREMENT TECHNIQUES
Today, heat flow and equivalent temperature are measured with various instruments and
devices. The most developed method is a man-sized heated manikin, which realistically
simulates the three-dimensional heat exchange of a human body.
MEASUREMENT OF CONVECTIVE AND RADIATIVE HEAT LOSSES
WITH HEATED SENSORS
Heated sensors can be used to determine body surface heat losses through convection
and radiation.
ELLIPSOID SENSORS
A heated ellipsoid sensor may serve as a representative physical model of the human
body (Brel and Kjr, 1982; Madsen et al, 1984). The heat exchange of the sensor is
assumed to correspond to the total (and uniform) convective and radiative heat exchanges
of the human body surface. The instrument directly calculates equivalent temperature.
FLAT SURFACE SENSOR
Heated flat surface sensors give reliable and relevant estimates of dry heat losses. The
three-dimensional nature of human heat transfer requires special arrangements of several
flat sensors or measurements on several locations for example, on the surface of a
body-shaped dummy. In order to provide a single figure for the overall t
eq
values,
Source: NREL (2006)
FIGURE 6.27 Manikin with 120 independently controlled zones
ABER_CH_06 6/12/08 3:33 PM Page 187
individual values are added and area is weighted in a representative way (Mayer and
Schwab, 1993).
LOCAL DISCOMFORT METER
Here, a double-sided heated skin element is used (Madsen et al, 1992). The difference in
mean heat flux (MHF) from the two opposite elements can directly be transformed to a
thermal asymmetry. The electrical signals received from the skin element are transformed
by a microprocessor into a value called perceived heat flux (PHF), which is equivalent to
the sensation of local thermal discomfort. Calibration of the instrument is done in the
same way as for the manikins.
MEASUREMENT OF BASIC CLIMATIC PARAMETERS
With measurements of air temperature, mean radiant temperature and air velocity, radiant
and convective heat losses can be calculated and t
eq
determined using appropriate
equations.
POINT MEASUREMENT TECHNIQUES FOR TEMPERATURE AND VELOCITY
Classical techniques are point-measuring techniques. Measurements of temperature, air
velocity, etc. are performed at discrete positions. Measuring at fixed positions
corresponds to the Eulerian approach to fluid mechanics. Observations are made at fixed
positions and the flow passes by. During the last two decades, there has been substantial
advances in whole field measuring techniques, which makes it possible to take
continuous measurements over parts of a room. Here, measurements of velocities are
based on the Lagrangian approach to fluid mechanics, where the displacement of tracer
particles ideally follows air motion.
Many of the measurement techniques are indirect in that the methods rely on the
physical interpretation of the quantity measured. Hot-wire techniques, for instance, are
based on relating heat transfer to velocity. A Pitot tube senses pressure, and velocity is
inferred from the measured pressure by using the Bernoulli equation. Flow rate is
recorded by relating the rate of change of tracer gas concentration to flow rate.
TEMPERATURE MEASUREMENTS
CONTACT MEASUREMENTS
Years ago, the mercury thermometer was the ultimate choice for measuring the indoor
climate; today, however, electrical thermometers such as resistance thermometers and
thermocouples are usually used. Electrical thermometers can be made small enough to
have a fast time response and may also be connected to data loggers and computers. But
the accuracy is not necessarily higher; a mercury thermometer can be calibrated to read
0.001K.
Electrical thermometers for laboratory use are usually resistive temperature sensors
or thermocouples. Resistive temperature sensors rely on the fact that the electrical
resistance of a material changes as its temperature alters. Two key types are the metallic
devices and thermistors. Metallic sensors are usually made of a platinum coil or film,
while thermistors consist of a piece of semiconductor material made of metal oxides
ABER_CH_06 6/12/08 3:33 PM Page 188
pressed into a small bead and covered with glass or epoxy. The thermocouple, which is
the most commonly used temperature sensor, has a quite different working principle. A
thermocouple is created when two dissimilar metal conductors are joined. The contact
point produces a small voltage as a function of temperature. This thermoelectric voltage
is known as the Seebeck effect and can be used to determine temperature.
Significant features of platinum resistive sensors are high reliability and excellent long-
term stability. They are well standardized and are very common in industrial applications
and whenever high accuracy is important, such as with reference sensors for calibration.
However, they are not the first choice for air temperature measurements in indoor climate
research.
Thermistors have low cost and high resolution, and they can be made very small. The
smallest bead thermistors available are approximately 0.1mm in diameter. Thermistors are
used in many low-cost temperature instruments and are the standard sensor employed in
the great variety of digital household thermometers available on the market. This does not
mean that thermistors are inaccurate. They can be calibrated to read one thousandth of a
degree, and standardized types now available guarantee an accuracy of 0.2K in the
room-temperature range.
Thermocouples are easy to use. In many applications, the thermocouple wire itself
can be employed as a sensor simply by twisting the two wires together. Thermocouple
wire is available in different sizes and with different insulations. For indoor climate use,
0.2mm to 0.5mm diameter wire is common; but wires as small as a few micrometres in
diameter are available. There are, in fact, two junctions in a thermocouple circuit. One is
at the measurement point and is called the hot junction; the second one is known as the
cold junction and appears at the connection inside the instrument. Since a thermocouple
only measures the temperature difference between two junctions, it is necessary to know
the temperature at the cold junction in order to calculate the temperature at the
measuring point. The cold junction temperature is measured by means of a temperature
sensor inside the instrument. Consequently, any errors in the measurement of the cold
junction temperature will be added to the total measurement error. In a low-cost
instrument, this sensor may have poor accuracy and is sometimes not even specified.
This source of errors is often overlooked when using thermocouples. A typical
thermocouple holds an uncalibrated accuracy of approximately 0.3K in the room-
temperature range; but the total error is highly dependent upon the instrumentation used.
The maximum resolution attained is in the order of 0.05K without using specialized
instruments.
Measuring temperatures in still air invites some pitfalls due to the poor heat conduction
of air. In order to measure the resistance of a resistive temperature sensor, a small current
must be passed through the sensor, causing self-heating of the sensor. It is thus important
to design the measurement electronic circuit so that self-heating is low enough to allow for
temperature measurements of the desired accuracy. For this reason, in more expensive
data loggers, the excitation current may be controlled. The thermocouple does not
generate self-heating or suffer from this problem. With thermocouples, however, errors
may occur if the thermocouple is mounted in a temperature gradient. Since the metal wires
in the thermocouple are good thermal conductors, heat may be conducted through the
ABER_CH_06 6/12/08 3:33 PM Page 189
wires out to, or from, the measuring junction, thus causing a different temperature in the
junction than in the surrounding air.
Errors may also be introduced due to radiation exchange between the sensor and the
surrounding surfaces. This is usually not a problem in the occupied zone under normal
conditions where the temperature difference between the air and surrounding surfaces
are low. But when measuring temperatures in an environment with surfaces at a
temperature that is quite different from air temperature, such as large windows, chilled
ceilings or heating radiators, this effect should always be considered. Here, the
thermocouple may have an advantage. If the insulation on the thermocouple wire is
stripped off on a piece close to the measuring junction, we will have a sensor with small
emissivity that is less sensitive to radiation effects. Bare metal wires have a much smaller
emissivity than insulated wires.
NON-CONTACT MEASUREMENTS
The above-mentioned temperature sensors are contact sensors, which means that they
must be in thermal contact with (and ideally in thermal equilibrium with) the media in
which the temperature is measured. Strictly speaking, a contact temperature sensor is
only able to measure its own temperature and it is important to mount the sensor so that
it attains the temperature of the media that we want to assess. There are also non-contact
temperature meters, which make use of the fact that all objects emit radiation. The
amount of radiation can be measured and related to temperature using the Planck law of
radiation. Using this technique, surface temperature can be measured remotely. Infrared
radiation thermometers have long been employed in industry for measuring high
temperatures, where contact sensors are difficult to apply. Today, the availability of low-
cost infrared semiconductor sensors has opened a new broadened market for infrared
temperature measurements. Handheld infrared thermometers are now available at
reasonable prices and are excellent for making a quick check of surface temperatures
in rooms. However, it cannot be used for measuring air temperature.
The Planck law only holds for a perfectly black surface. A real surface, usually called a
grey surface, emits less radiation than a black surface. The radiation from a grey surface is
related to the black surface radiation by a correction factor called emissivity. The emissivity
has a value in the range of 0 to 1, where 1 applies to a black surface. Thus, in order to use
the thermal radiation as a measure of temperature, the emissivity must be known.
High emissivity surfaces, such as building materials, painted surfaces or human skin,
hold an emissivity close to 0.9. Low-cost instruments have the emissivity fixed at that
value, and with this assumption we can measure temperatures at this kind of surface with
an accuracy in the order of 1K, well enough for many indoor climate applications. More
expensive instruments have provision for the emissivity to be set manually, which allows
for measurements with higher accuracy, provided that the emissivity is known. The more
sophisticated instruments also are provided with optics that allow for a small measuring
spot. Typically, a measuring spot of approximately 20mm diameter can be achieved at a
distance of a couple of metres from the instrument. A fundamental characteristic of
infrared thermometers is, however, that measuring temperature on low emissivity
surfaces, such as metals, is very difficult and prone to large errors.
ABER_CH_06 6/12/08 3:33 PM Page 190
AIR VELOCITY MEASUREMENTS
In the occupied zone, velocity measurements are performed in combination with air
temperature measurements in order to investigate the risk of draught. But air movements
in other zones, such as close to air terminals, may also be of interest. In the occupied
zone, velocities are low, typically in the range of 0.1m/s to 0.5m/s, while the velocity from
an air inlet terminal may be up to a few metres per second. Air velocity measurements
may also be used in ducts as a means of estimating the volume flow.
There are basically two types of instruments in use for measuring air velocities in
indoor climate research: thermal anemometers and laser Doppler anemometers. Thermal
anemometry has been around for almost 100 years and is a well-established technique.
Laser Doppler techniques have evolved rapidly during the last decades; but laser Doppler
anemometers (LDAs) are still very expensive. The two techniques are really very different,
each one having its pros and cons. A thermal anemometer relies on the fact that
convective heat loss from a heated body varies with velocity, while the LDA measures the
movement by small seeding particles following the air stream.
THERMAL ANEMOMETERS
There are essentially two types of thermal anemometers that are used for indoor climate
measurements: the hot-wire anemometer and the hot-sphere anemometer.
The hot-wire anemometer has a sensor consisting of a thin metal wire of platinum or
tungsten, typically 5m in diameter and a few millimetres long. The sensor is heated by
means of electrical joule heating. It is usually operated in constant temperature mode,
which means that electrical current through the sensor is controlled by means of
electronic controlling circuitry, so that the temperature of the sensor wire is kept at a
constant value. The electrical power delivered to the sensor to maintain its temperature
constant is a measure of the heat transfer from the sensor to the air and can be related
to velocity. This relationship must be achieved by means of calibration to known
velocities. The main characteristics of the constant temperature hot-wire anemometer are
a fast response and high spatial resolution, making it useful for unsteady and turbulent
flows. Sampling rates of up to 100kHz are possible. Hot wires are prone to drift with time
and need to be frequently recalibrated. Accuracy depends upon many parameters (and,
of course, upon the quality of the calibration equipment used). A typical value is 5 per
cent and up to 1 per cent in extreme cases.
The original application area for hot-wire anemometers is high velocity measurements in
wind tunnels. Here, the wire is operated at a high temperature. However, for low velocity
flows, as in indoor climate applications, the wire temperature must be kept low, typically less
than 100C, in order to minimize errors due to natural convection from the heated sensor
wire. A decrease in wire temperature results in an increase in sensitivity to air temperature.
As the sensor wire temperature is constant, any change in air temperature results in a change
in temperature difference between the wire and the air and will thus affect the output from
the anemometer and appear as an error in the velocity measurements. This means that hot-
wire anemometers usually must be temperature compensated when measuring low
velocities. Temperature compensation is accomplished by measuring the air temperature
with a separate temperature sensor and using a compensation scheme in the software.
ABER_CH_06 6/12/08 3:33 PM Page 191
The hot-wire anemometer is most sensitive to the velocity component perpendicular
to the wire and must be oriented properly in order to get correct readings. Room air
movements may not have a well-defined direction and in these situations the hot sphere
anemometer is often preferred. The hot-sphere anemometer has a spherical sensor,
usually consisting of a glass sphere typically 2mm to 3mm in diameter, covered with a
nickel layer. This type of anemometer is essentially omni-directional and may be regarded
as an air speed meter or draught meter, mostly used in a free air stream such as in the
occupied zone. It is not suitable for measurements close to surfaces, such as in boundary
layer measurements. The time response is limited to about 20Hz. In order to minimize
disturbances due to natural convection from the sensor itself, the sensor temperature is
low, typically 20K to 30K above ambient air temperature, and this is accomplished by
operating the anemometer in the constant temperature difference mode. This means that
the hot sphere temperature is controlled by electronic controlling circuitry such that the
temperature difference between the sensor and the ambient air is always kept at a fixed
value. For this reason, the hot sphere anemometer is always equipped with a temperature
sensor close to the velocity sensor. It is important to bear in mind that in order to give
correct measurement results, the two sensors must be immersed in air with equal
temperature. This must always be considered when temperature gradients are expected
in the flow, as may be the case, for instance, in natural convection flows.
In addition to the above-mentioned types of thermal flow sensors, a variety of sensors
of different configurations exist for use in different applications. Hot-wire anemometers
may be equipped with more than one sensor wire in order to make it possible to sense
more than one component of the flow field. The most common is the X-probe with two
perpendicular wires, which can detect the velocity vector within a cone of 10.
Sensors made of a glass fibre rod covered with a nickel layer (fibre film sensors) are
sometimes used instead of the solid wire sensors. Fibre film sensors are not as fragile as
the solid wire sensors and are often used in high-velocity measurements in wind tunnels
or for measurements in water. Hot-sphere anemometers are sometimes equipped with
bead thermistors as sensors instead of the nickel-covered glass sphere. Thermistors can
make reasonably good omni-directional anemometers at low cost, and it is not uncommon
that indoor climate laboratories assembly their own thermistor anemometers for
measuring velocities at many points simultaneously (McNair, 1972; Lundstrm et al, 1990).
A fundamental drawback with thermal sensors is that they cannot sense flow
reversals, so measurements in circulating and highly turbulent flows should be handled
with caution. Strictly speaking, we may say that when measuring velocities by means of
thermal anemometers, we must have some a priori information about the character of
the flow field. Visualization with smoke may accomplish this. By applying a puff of smoke
into the flow and by watching where it is going, we can estimate the velocity and
direction, but also get rough information about turbulence intensity. This is a simple but
very useful technique that should not be overlooked when investigating low velocity
flows in air.
For more detailed information about thermal anemometry and the practical problems
involved, the reader is referred to the available literature in the field (e.g. Christman and
Podzimek, 1981; Fingerson, 1994; Bruun, 1995; Melikov et al, 1998a). An extensive investigation
ABER_CH_06 6/12/08 3:33 PM Page 192
of uncertainties in low-velocity measurements when using hot-sphere anemometers is
presented by Popiolek et al (1998) and Melikov et al (1998b). Lundstrm et al (2007) have
investigated the use of hot-wire anemometers at low velocities and varying air temperatures.
LASER DOPPLER TECHNIQUES
In the laser Doppler anemometer, the scattered light from small seeding particles passing
through the intersection of two laser beams is measured. The laser beams interfere in the
intersection (the measuring volume of the LDA), and when a particle passes, the scattered
light has a frequency that is proportional to the velocity of the particle. The frequency
depends upon particle velocity, as well as laser wavelength and the angle between the
two laser beams. The laser wavelength is very stable, and once set up the laser Doppler
anemometer needs no recalibration. Light from the measuring volume is scattered in all
directions, although the light scattered in the forward direction from the laser beams has
the highest intensity. Today, many laser Doppler anemometers use backscatter and have
the feature that both laser optics and receiving optics, as well as the detector, may be
assembled in one single unit. The measuring volume is slightly ellipsoid shaped and the
size depends upon the diameter of the laser beams, the angle between the beams and
the focal length of the optics used. Typical size of the measuring volume is a few
millimetres, down to about 0.05mm. However, it is important to remember that the
sampling rate cannot be exactly controlled because a measurement occurs when a
seeding particle passes the measuring volume.
Measurements are intrusion free and the accuracy is generally higher than for thermal
anemometers figures as high as 0.1 per cent have been reported. In contrast to thermal
anemometers, the laser Doppler anemometer measures the true velocity component. For
measurements of more than one velocity component, more laser beams of different
wavelengths are added, and laser Doppler anemometers that simultaneously measures
one, two or all three velocity components are available.
A drawback, however, is the need for seeding. In air, seeding is usually accomplished by
adding smoke. Supplying smoke to the measuring volume without disturbing low-velocity
flow is sometimes tricky and deserves some exercise. In addition, a confined space will
soon be filled with smoke, and measurements must be interrupted for ventilation.
The comparably large measuring volume and the fact that use of a laser Doppler
anemometer close to a surface may introduce errors from reflections make the laser
Doppler anemometer less useful for boundary layer measurements. Here, the hot-wire
anemometer usually performs better. The hot-wire anemometer is also superior when
making time-series analysis due to the difficulty in controlling the sampling rate in the
laser Doppler anemometer.
ULTRASONIC ANEMOMETERS
Although uncommon for measuring indoor velocities, the ultrasonic anemometer is
another type of anemometer worth mentioning here. Ultrasonic anemometers are based
on the measurement of the transit time of transmitted sound waves between two points.
This can be accomplished in different ways; but commercial anemometers available
today usually utilize the pulse method. Here, the anemometer measures the time taken
ABER_CH_06 6/12/08 3:33 PM Page 193
for an ultrasonic pulse to travel from one transducer to the other, and then compares it
with the time taken for another pulse to travel in the opposite direction. By using three
pairs of transducers, the true velocity vector can be calculated.
The ultrasonic anemometer was presented for the first time during the 1940s; but
commercial anemometers have been on the market only for a couple of decades. The
main application is in meteorology for outdoor wind measurements. These types of
ultrasonic anemometers are rather large, but are sometimes used for indoor climate
measurements. However, the future trend seems to be smaller-sized ultrasonic sensors,
and we will certainly see more of these anemometers in the future. The sampling rate is
typically 20Hz for ultrasonic anemometers and accuracy in the order of a few per cent may
be achieved.
MEASURING VELOCITY AND TEMPERATURE
WITH WHOLE-FIELD MEASURING TECHNIQUES
Whole-field techniques are relatively new and under development. An overview of whole-
field measuring techniques, including absorption tomography for recording concentrations,
is given in Sandberg (2007).
All techniques are optical in the sense that pictures are taken by an array of sensors
sensible to different wavelengths of light depending upon the application. The velocity
components can be recorded with particle image velocimetry (PIV) or particle streak
velocimetry (PSV). Both methods are based on adding tracer particles, which ideally follow
air motions; a digital camera records their displacement during a specified time interval.
This corresponds to the Lagrangian formulation of fluid mechanics. With PIV, the
displacement of groups of particles is recorded in contrast to the PSV method, where the
displacement of a single particle is recorded. PIV can provide high-resolution information
over small regions, whereas the PSV method can cope with large areas.
Temperature distribution can be recorded and visualized with an infrared camera and
a measuring screen that ideally attains room air temperature.
PARTICLE DISPLACEMENT
In both PIV and PSV, velocity is determined by measuring the displacement X in room
coordinates: X = X(t + t) X(t) = (X(t + t) X(t), Y(t + t) Y(t), Z(t + t) Z(t)) of seeded
particles in a time interval t.
For a known time step t, the components of the velocity vector U(X, t) are derived
from the fundamental definition of velocity:
[20]
Equation 20 is the definition of velocity, and both PSV and PIV are therefore direct
methods. A planar light sheet is used and the velocity components lying in the light sheet
are the in-plane components. Traditionally, the notation assumes that the X and Y
components are the in-plane components, while the Z component is the out-of-plane
component. A well-defined time step is generated by a pulsed light source or by
interrupting the light beam by a chopper or a shutter.
U X t
X
t
X t t X t
t
Y t t Y t
t
Z t t Z t
t
( , )
( ) ( )
,
( ) ( )
,
( ) ( )
= =

\
)
.
ABER_CH_06 6/12/08 3:33 PM Page 194
BASIC EQUIPMENT FOR MEASURING VELOCITY
Here, the objects are markers following air motion. The markers are illuminated by light of
a suitable wavelength (see Figure 6.28). Through an optical system, the light is turned into
a light sheet, usually with a cylindrical lens. The light sheet constitutes the measuring
volume. With a system measuring two velocity components, proper measurements
require that the flow field is essentially two dimensional and the light sheet is oriented
accordingly. To determine the out-of-plane component, stereo-photogrammetry can be
used. At least two imaging systems are then required and the thickness of the light sheet
is set to approximately 10cm.
TRACER SEEDING SYSTEM
Here, the air must be seeded by particles that follow air motion. As a result, there is a
compromise between the need for particles to be small enough to follow air motion, but
large enough to reflect sufficient light in a suitable illumination.
In the PIV mode, particles with a diameter of approximately 1m are used. Smoke
generators for example, by boiling paraffin oil can generate such particles. In the PSV
mode, individual particles must be identified; therefore, the particles must be larger
because light sources are not strong enough to produce a good exposure of small
particles. In the PSV mode, the particles used in air are normally neutral buoyant helium-
filled soap bubbles with a diameter of about 1mm to 2mm. According to simulations by
Muller et al (2001), a soap bubble follows variations in air velocity reasonably well, up to
a frequency of 100Hz. The requirements on tracer particles for PIV are dealt with in
Melling (1997).
d
i
d
0
Lens
Image
Light source
Light sheet
CCD Array
X
Y
Z
Interrogation
volume
x
y
Tracers
Z
Image
coordinate
system
Room coordinate
system
FIGURE 6.28 Illumination system, flow markers and a digital camera
ABER_CH_06 6/12/08 3:33 PM Page 195
PARTICLE IMAGE VELOCIMETRY (PIV)
An overview of PIV is given in Adrian (1991, 2005). The PIV recording and interrogation
process are illustrated in Figure 6.29.
As a result of using small-diameter particles (droplets), a pulsed high-power laser is
employed as a light source. Seeding density is high and the displacement of whole
groups of particles is recorded. Through correlation analysis, the displacement of the
group of particles is identified. A drawback with PIV is the small area that the system can
manage: approximately 0.2 x 0.2m; nevertheless, detailed information can be obtained.
PIV systems are sold commercially.
PARTICLE STREAK VELOCIMETRY (PSV)
In PSV, seeding density is low and the movement of individual tracers can be recorded (a
synonym for particle streak velocimetry is particle tracking system). As a result, PSV can
only record where there are tracers. In order to measure areas between tracers, data must
be interpolated. An advantage of PSV is that the system can cope with a large area and
standard lighting systems can be used. A drawback is that within the interpolated regions,
accuracy is lower at about 20 to 25 per cent.
MEASURING AIR TEMPERATURE WITH INFRARED THERMOGRAPHY
Air is transparent to infrared radiation and therefore a measuring screen that adopts air
temperature must be used (see Figure 6.30), making this technique intrusive. A modern
x
x
y
y
z
z
d
0
d
1
Flow field
Object plane
Image plane
x
c
x
c
Interrogation
spot
Interrogation
analysis of the
images
Displacement
Resulting
vector
Interrogation volume
Source: adapted from Soloff et al (1997)
FIGURE 6.29 Particle image velocimetry (PIV) recording and interrogation process
ABER_CH_06 6/12/08 3:33 PM Page 196
infrared camera typically has a resolution of 320 x 240 pixels, which is much lower than
commercial charge-coupled device (CCD) cameras and is sensitive to infrared radiation in
the range 7.5 to 13m. The absolute accuracy is about 2C, while the relative accuracy
is about 0.1C. A measuring screen is affected by radiation from ambient room
surfaces, which gives rise to a systematic error. A remedy for minimizing error is to cover
the two sides of the screen with different materials (Cehlin, 2006). One side can be of
paper, while the other may be aluminium foil.
Figure 6.30 shows how the measuring screen can be placed when recording the radial
supply temperature from a low-velocity diffuser for displacement ventilation placed at
floor level. Disturbance is minimized when the screen is placed in the symmetry plane.
Measurement screen
Supply device
FIGURE 6.30 Orientation of the measuring plane
FIGURE 6.31 Supply of warm air from a ceiling diffuser: Temperature distribution recorded
with an infrared camera
ABER_CH_06 6/12/08 3:33 PM Page 197
By taking repeated measurements along the other positions indicated in the figure, a
three-dimensional picture of the temperature distribution can be obtained. This
information may be used in computer-aided design (CAD) programmes showing how a
floor area is affected by the direct supply of the chilled air to the occupied zone.
A solid screen restricts the applicability of the method. Porous screens (10mm thick
with holes of 9mm in diameter) oriented horizontally have been used to record the
temperature in a plume above a heat source (Streblow et al, 2007).
Figure 6.31 shows an application of this method. Here, the recorded temperature
distribution of the supply of warm air from a ceiling diffuser is depicted.
TRACER GAS METHODS
Ventilated rooms feature strong recirculation; therefore, the velocity field does not reflect
the rate by which a contaminant is removed. Tracer gas methods are the ventilation
measuring method because the tracer gas is a substitute for a passive contaminant.
Through ventilation, the tracer gas is diluted and the rate of dilution provides information
about the airflow rate to a room and parameters related to the ventilation process. A tracer
gas can also be seen as an indicator of air molecules. We can therefore record the time
that it takes to replace the markers in the room with non-marked air molecules coming
from outside. Tracer gas methods are the most general technique for measuring
ventilation airflow rates in rooms, including infiltration. They are the only methods
available for recording the total flow rate, including infiltration, to a room, and are the only
methods for recording the age of air and the time that it takes to replace air.
TRACER GASES
Ideally, a tracer gas should be a gas not normally present in the indoor environment. In the
review by Hitchin and Wilson (1967), tests are mentioned that feature radioactive tracers.
Today, there is a great concern about health effects and the impact of gas on the
environment. Common tracer gases are sulphur hexafluoride (SF
6
) and nitrous oxide
(N
2
O). Occasionally, carbon dioxide (CO
2
) is used as a tracer gas despite the fact that it is
present in peoples ambient and exhalation air. A gas used in connection with passive
tracer gas techniques is hexafluorobenzene (C
6
F
6
). This gas can be detected at low
concentrations.
METHODS OF DETECTING TRACER GASES
A gas analyser consists of a detector and a gas sampling unit. A gas chromatograph with
an electron capture detector or mass spectrometer can detect sulphur hexafluoride and
hexaflurorobenzene at low concentrations. A gas chromatograph is primarily a laboratory
instrument. With an infrared gas analyser or a photoacoustic detector, one can detect
nitrous oxide, carbon dioxide and sulphur hexafluoride. These two types of analysers can
be used in field trials.
TECHNIQUES FOR INJECTING TRACER GAS AND SAMPLING
Here, the two main techniques are active or passive. In an active technique, the pressure
in the gas cylinder is used for injecting the gas. The gas is mixed with room air using
ABER_CH_06 6/12/08 3:33 PM Page 198
mixing fans or other means that promote amalgamation. Air samples are taken by using
a pump, and when sampling at many points, electro-mechanical valves are used for
switching between sampling points.
Passive systems are based on molecular diffusion both for injecting and sampling gas.
As a gas source is used, a small container is provided with a capillary or a permeation
membrane in order to control the emission rate. Figure 6.32 depicts a source with a
capillary. The emission rate can be controlled with a metal wire. Molecular diffusion is
temperature dependent and careful calibration is required for the actual gas used. For
passive sampling, an adsorption tube is used (see Figure 6.32).
METHODS FOR MEASURING FLOW RATE
Various methods for measuring the flow rates to a house or air change rates by tracer gas
are dealt with in ISO 12569 (2000). Depending upon injection strategy, we can classify
them as the decay method, the constant injection method, the constant concentration
method and the arbitrary injection method.
The decay rate method is the simplest method and is used to obtain discrete flow
rates over short periods of time. A small quantity of tracer gas is mixed into the room air.
Metal
wire
FIGURE 6.32 Components in a passive system
ABER_CH_06 6/12/08 3:33 PM Page 199
The source is then removed and the decay in the concentration of tracer gas is measured
over a period of time. Mixing fans may be used to ensure that the tracer gas concentration
is the same in all points of the room. The method is based on the fact that when there is
complete mixing, the local mean age in its point is equal to
p
. The coefficient of the
exponential decay is equal to the inverse of the local mean age and thus:
Q = V
e
. [21]
To obtain the flow rate, an additional parameter, the volume V of the ventilated zone, must
be known. The decay method is therefore an indirect method.
With the constant injection method, a tracer gas is released with constant and known
mass (volume) rate. According to Equation 10, the attained equilibrium concentration
gives the flow rate as:
[22]
This is therefore a direct method that is used, both with active and passive methods,
for estimating long-term average flow rates and the airflow through long ventilation ducts.
With a time-varying flow rate, bias is associated with the method. In order to achieve
the average flow rate, we need to know the average of the inverse concentration; but we
only know the inverse of the average concentration: 1/C. These two quantities differ from
each other.
The constant concentration method is used for continuous airflow rate measurements
in one or more zones. Here, the concentration of tracer gas in a zone is measured and a
computer controls the tracer gas injection rate so that the concentration in each zone is
kept at a constant target value. By keeping the concentration constant, the effect of flow
through doorways is neutralized and the inflow of outdoor air to each individual room
(zone) can be recorded (see Figure 6.20).
By using system identification methods, the flow rate can be retrieved from the time
history of the concentration generated by a time-varying injection of tracer gas. This is a
so-called inverse problem that is very sensitive to the accuracy of the data. One example
of the application of the method is found in Okuyama (1990).
METHODS FOR MEASURING THE LOCAL MEAN AGE OF AIR
For field measurements, the most common method is the homogeneous emission
method (see Equation 16), based on use of passive tracer gas techniques. The room is
subdivided into a number of sub-volumes: V
i
. The injection rate m
.
i
into sub-volume V
i
is
proportional to the volume of the sub-volume:
[23]
where M
.
is the total injection rate into the room. The decay technique is also relatively
simple. The local mean age at the point P is obtained from the area of the graph of the
decay curve and the initial concentration C(0):
m
V
V
M
i
i

=
q
m
C
=
.
ABER_CH_06 6/12/08 3:33 PM Page 200
[24]
Procedures for doing the measurements can be found in ISO 16000-8 (2007).
MEASURING DUCT FLOW
Perhaps the most widely used flow-metering principle is to record the pressure
difference over some obstruction in the duct. In a laboratory set-up, orifice flow meters
are the most common and rely on measuring the pressure difference over a sharp-
edged orifice placed in the duct. This is a standardized and very accurate method.
Orifice flow meters are seldom used for field measurements; but modern ventilation
installations may be provided with similar devices that feature connections for a
differential pressure meter for monitoring the flow. Sometimes a bend in the ducts is
used for this purpose.
The volume flow can also be calculated from velocity measurements in the duct. Since
the flow profile in the duct is unknown, velocities in a number of points in the duct must
be measured, and the flow is calculated from the velocity values according to a
standardized scheme. The velocity is usually measured by means of a hot-wire
anemometer or a Pitot tube. The instrument is introduced through a hole in the duct wall
and traversed across the duct. The Pitot tube (named after Henri Pitot in 1732) measures
the fluid velocity by converting the kinetic energy of the flow into pressure using the
Bernoulli equation.
There are instruments on the market that can directly measure the flow through
an air inlet or exhaust terminal. These instruments consist of a hood, which is placed
over the air terminal and through which the airflow must pass. The hood is equipped
with a flow-measuring device based on either differential pressure measurements
or a hot-wire type of instrument. The accuracy is about 5 per cent for these
instruments.
The bag method can be used to measure the true volumetric flow through an air
exhaust terminal. The method relies on measuring the inflation time of a polyethylene bag,
which is suddenly connected to the air terminal. Flow in ducts may be measured using
the constant injection method and monitoring the concentration. Equation 22 gives the
flow rate. Here, it is essential that the duct is sufficiently long that injected tracer gas will
be fully mixed with the air.
DATA COLLECTION AND DATA HANDLING
Today, measuring instruments are often delivered as complete systems that connect
directly to a computer and come with a dedicated software package. This applies to, for
instance, laser Doppler anemometers, modern hot-wire anemometers and gas monitors.
For measuring devices that deliver an analogue output signal, such as temperature
sensors, a separate data logger must be used. A data logger can be a multi-channel
analogue/digital (A/D) card inserted in the computer or a separate instrument connected
P
C t dt
C
P
0
0
( )
( )
ABER_CH_06 6/12/08 3:33 PM Page 201
to the computers serial bus. During the last few years, many manufacturers have
presented a large variety of small low-cost data loggers using USB connection. But there
are great differences between different data loggers, and it is important to ensure that the
data logger has the resolution and accuracy needed. Thermostats have a high output and
are easily connected to most data loggers, while thermocouples are more demanding.
Thermocouples, which are the most commonly used temperature sensor for laboratory
use, have a low output signal, and a good-quality instrument is recommended.
For field measurements, where the measuring points are often spread over an
apartment or a whole building, several small data loggers may be beneficial. During the
last decade, small battery-operated data loggers have become available from several
vendors. They are equipped with typically 1 to 5 inputs to which sensors for temperature,
humidity, etc. can be connected. These type of loggers store data in their internal memory
for weeks or months and can then be connected to a PC for saving data onto the
computer. More sophisticated models feature connection to a host computer via wireless
radio communication or via a telephone line.
The availability of user-friendly software has considerably reduced frequently time-
consuming computer work. Today, LabVIEW (a trademark of National Instruments) is
probably the most commonly used software dedicated to measurement and data
handling. Many instrument manufacturers supply software drivers for operating their
instruments with LabVIEW.
ASSESSMENT AND EVALUATION METHODS
COMFORT ZONE DIAGRAM EVALUATION
In order to make comfort evaluation independent of clothing, the construction of new
comfort zone diagrams can be made by inserting any seated total insulation available
(Nilsson, 2004b). Equation 25 shows how a relationship between the equivalent
temperature level and mean thermal vote (MTV) can be established for each manikin body
part. The heat loss corresponding to a certain level of comfort, or discomfort, in the
diagram is consequently considered to be the same. The shape of the zones is, however,
changed according to the clothing used:
[25]
where:
t
eq,zone
= equivalent temperature in the zone (C);
t
s
= manikin surface temperature (here 34C) (C);
R
T
= total insulation, seated (m
2
K/W);
a, b = linear regression constants (W/m
2
);
MTV
zone
= mean thermal vote in the zone (here no data).
The equation is valid for an interval of seated whole-body total insulation (R
T
) between
0.9clo and 1.9clo. By using Equation 25 and the information in Table 6.1, it is, for the first
time, possible to make a comfort zone diagram (see Figure 6.33) that applies to a specific
t t R a b MTV
eq zone s T zone ,
. . = ( )
ABER_CH_06 6/12/08 3:33 PM Page 202
clothing combination used in a given situation. One must insert in Equation 25
corresponding local values of the surface temperature t
s
(usually 34C), a and b, and local
total insulation value R
T
of the clothing and air layer, together with a MTV
zone
for the four
borders of the three shaded comfort zones in the legend of Figure 6.33. Now t
eq,zone
can be
calculated for those four borders, for all zones and the whole body. The plotted result forms
the evaluation background in the clothing-independent comfort zone diagram. This is only
done once for each clothing combination and reflects the insulation distribution of the
clothing used. As could be expected, diagrams for clothing with less insulation show an
increased sensitivity in most zones, except the normally unclothed head and hands. The
opposite, decreased sensitivity for heat loss variations, can be observed for diagrams with
increased clothing insulation.
These new methods have also recently become an international standard: ISO 14505:
Ergonomics of the Thermal Environment Thermal Environment in Vehicles. The comfort
zone diagram shown in Figure 6.33 corresponds to the figures D1 and D2 in the new
standard (ISO 14505-2, 2006).
A new benchmark study is providing requirements for the design and development of
computer manikins and computational fluid dynamic (CFD) benchmark tests for comfort
evaluation. The main idea is to focus on people. That the comfort requirements of
occupants should be the decisive factor in thermal climate will prevail. It is therefore
important to use comfort simulation methods that originate from people, not just the
temperature of surfaces and air.
In order to obtain accurate results from the numerical simulation, knowledge or
correct simulation of the manikin surface heat transfer is of utmost importance. The near
Note: This figure shows the results with no clothing (IT = 0.9clo) for 17 segments and all of the manikin body. Abbreviations
refer to L = left; R = right; U = upper. A spreadsheet can be downloaded from the Thermal Manikin Network (Nilsson, 1999)
as well as the CFD Benchmark sites (see www.cfd-benchmarks.com). Note that the comfort zones of acceptance are much
narrower for summer clothing, except for the less sensitive head and hands.
Source: www.cfd-benchmarks.com
FIGURE 6.33 Comfort zone diagrams adapted for Comfortina-type manikins
35
30
25
20
15
10
5
0
E
q
u
i
v
a
l
e
n
t

t
e
m
p
e
r
a
t
u
r
e
,

t
e
q

(
C
)
L
.

F
o
o
t
R
.

F
o
o
t
L
.

L
o
w

l
e
g
R
.

L
o
w

l
e
g
L
.

T
h
i
g
h
R
.

T
h
i
g
h
P
e
l
v
i
s
H
e
a
d
T
o
p

o
f

h
e
a
d
L
.

H
a
n
d
R
.

H
a
n
d
L
.

F
o
r
e
a
r
m
R
.

F
o
r
e
a
r
m
L
.

U
p
p
e
r

a
r
m
R
.

U
p
p
e
r

a
r
m
C
h
e
s
t
B
a
c
k
A
l
l
too hot
warm but
comfortable
neutral
cold but
comfortable
too cold
1.5
0.5
0.5
1.5
ABER_CH_06 6/12/08 3:33 PM Page 203
surface flow field in a room or around a heated body is characterized by a combination of
natural (free) and forced convection, developing boundary layers. The restricted validity of
the heat-transfer models often used originates from the assumptions that have been
made to solve special boundary layer flows. These assumptions are consequently not
always valid for flows that can be commonly found in the indoor environment.
Nielsen et al (2003) introduced two benchmark tests focusing on the airflow around
virtual thermal manikins or computer-simulated persons (CSPs). Now a new benchmark
test for a CFD manikin or a CSP is available. One purpose of this test is to create a series
of very detailed and accurate full-scale measurements to serve as a comparison with CFD
predictions. The ideas behind a CFD manikin benchmark test, which define the boundary
conditions around a real, as well as a CFD, manikin, are as follows:
It is of great importance to be able to verify that simulated heat losses equal
measured heat losses converted into equivalent experienced temperature in order
to support comparisons with human experiences presented in clothing-independent
comfort zone diagrams.
If different versions of virtual CFD manikins can be tested with the same boundary
conditions, it is possible to make comparisons, and perhaps make some new decisions,
on a geometrical level of the design, turbulence model used, type of grid, etc.
It is very useful to have the results presented not only as a whole-body influence, but
also with local information on how the thermal climate varies over the human body.
The development of these virtual models gives us a more efficient and complete
complement to traditional evaluation of the thermal environment.
Different avenues of heat loss from a buildings occupants are best recorded directly
with a thermal manikin or some other heat loss-detecting instrument. Today, it is
frequently more time-consuming to measure indirect discrete values of air and
radiant temperatures, as well as air velocity. This is, however, foreseen to change in
the near future when it becomes more important to have better control of the indoor
environment due to increased demands and new legislation and regulations.
TEST ROOMS AND SCALE MODELS
A model of a building or a room can be a full-scale replica of a building or room, or a building
or room on a reduced scale (scale models). The advantage of a model is that it can easily be
altered and sophisticated instrumentation can be used. Topics, for example, that can be
studied in a model are airflow pattern, the risk of draught and the risk of short-circuiting of air.
FULL-SCALE REPLICA
In the design process, full-scale rooms are used when there are special requirements
concerning ventilation and cooling or heating. These special requirements can either be a
non-standard room shape or, for example, a high cooling load in terms of W/m
2
floor area,
which introduces a risk of draught. Special conditions will make it uncertain whether
standard solutions will work; as a result, a full-scale room is built to test that the thermal
comfort requirements can be met. An advantage of a full-scale room is that people can be
exposed to the climate and therefore can sense it. Special full-scale test rooms (climate
ABER_CH_06 6/12/08 3:33 PM Page 204
chambers) are built with the purpose of exploring peoples sensation of climate. Test
people are used in such experiments and two important issues related to the use of such
people are the question of internal and external validity. In such experiments, one variable
is frequently varied (the independent variable), often in such a way that the test people are
unaware of the change. The internal validity of the experiment concerns itself with
whether the observed effects were caused only by the independent variable. The external
validity tells us to what degree the findings can be generalized to other people and other
situations (Liebert and Langenbach-Liebert, 1995).
As shown by Oseland (1995), a context effect on thermal assessment may be
obtained in the sense that the judgement of the thermal conditions can be influenced by
the surrounding environment. Therefore, it is important that the appearance of the test
room is similar to a real room. This means that the test room must be decorated and
furnished. If an installation does not perform well in a laboratory test, it will in all
probability not work in practice due to future disturbances and difficulties.
SCALE MODELS
In general, a scale model is used when it is not possible to do a full-scale test in the
laboratory or in the field. The aim of using a scale model may be as follows:
During the design process, predictions can be made by using scaling laws to transform
results from scale models to those at full scale (Etheridge and Sandberg, 1996,
Chapter 14). These scaling laws are based on the premise that there will be similarity
between the model and the full scale, which implies that the non-dimensional numbers
relevant to the actual physical process involved, such as a Reynolds number, a Froude
number and a Peclet number, will be equal. The principle problem with a model in
reduced scale is that the different physical phenomena are affected to varying degrees
by the change in scale. Before the introduction of computational fluid dynamics, scale
models were used, but are now less frequent due to the problems with scaling of
non-isothermal phenomena.
To isolate a physical process for example, the effect of a source of buoyancy can
be studied by using a solution of salt and water to generate a density difference. This
method is sometimes called the salt bath method. With the use of water as an
operating fluid, there is no radiation exchange between room surfaces. As a result,
we cannot expect that the recorded density distribution exactly represents the
temperature distribution in air. The scaling is based on the fact that the reduced
gravity is the same in the model and at full scale. An example of this
approach is found in Hunt and Linden (1999), who studied the combined effect of
wind and buoyancy in natural ventilation. Examples of the use of the salt bath
method can be seen in Hunt and Linden (1999, Figures 10, 17 and 18).
To calibrate a CFD model with the aid of data from the scale model. By doing so, the
scale model becomes the real model. The rationale behind this approach is that
despite the fact that the model is not real, it has all the complexity of reality and all
the physical processes present in reality.
=

g g


ABER_CH_06 6/12/08 3:33 PM Page 205
Scale models of buildings are used in wind tunnels in order to study wind effects in the
built environment, as well as the dispersion of contaminants.
DISCUSSION AND CONCLUSIONS
Buildings are large objects and a room is a large space. This large scale in itself introduces
a problem: one cannot measure everywhere. A strong recirculating flow exists inside
rooms, and the velocity in the occupied space is very low (in the order of 10 to 20cm/s).
It is crucial to perform measurements on several occasions; in real buildings,
measurements are carried out to quantify the performance of the system. Mock-up rooms
are built in laboratories when designing a building and in climate chambers when carrying
out research. The application of computational fluid dynamics has renewed the need for
measurements with the purpose of calibrating (validating) CFD models. Development of
virtual manikins to be used in computers is an ongoing process. The basic idea behind
these virtual manikins is to connect the results from measurements using thermal
manikins to human sensation, and then to implement virtual manikins with the same
properties as the thermal manikins. The crucial thing is to predict the correct heat transfer
rate (see Brohus and Nielsen 1996; Loomans, 1998; and Nilsson, 2004a). This has been
shown to be difficult, leading to renewed interest in measuring the heat flux from different
parts of the human body with segmented thermal manikins.
Among thermal climate variables, velocity is the most difficult variable to record. In
order to assess draught, the speed (magnitude) of the velocity field can be recorded point-
wise with omni-directional probes. The velocity vector field can then be recorded with
whole-field measuring techniques. With digital cameras, pictures of tracers following air
motion are recorded. Particle image velocimetry (PIV) is a commercial technique for
measuring two-dimensional velocity fields at a high accuracy. However, PIV can only cope
with fields in the order of 0.2 x 0.2m. Due to the limited size of the region that can be
measured, PIV is not a general-purpose method for studies of room air. Another
disadvantage is that it requires a high-powered laser light source. One non-commercial
technique is particle streak velocimetry (PSV), which can cope with areas up to 10m
2
. An
advantage here is that standard light sources can be used. With PSV, velocity is recorded
only at locations where a tracer happens to be located. Within regions where there is no
tracer, data interpolation is required and the interpolation may introduce a relative error of
up to 20 per cent.
Due to the problem of scaling non-isothermal flow, CFD and multi-zone models have,
to a great extent, replaced scale models. However, there is a renewed interest in the use
of scale models for calibrating CFD models. In this approach, the scale model is the real
model. The rationale behind this is that despite the fact that the model is not, in fact, real,
it has all the complexity of reality and all the physical processes present in reality. Its small
size, furthermore, makes it easy to instrument.
A characteristic feature of internal flows in buildings is the occurrence of recirculating
air. As a result, the velocity field does not adequately provide information about the
transport of contaminants. The tracer gas technique is the ventilation measuring
technique because it is possible to record the dilution of a tracer gas and the purpose
of ventilation systems is to dilute and remove contaminants in air. This is a versatile
ABER_CH_06 6/12/08 3:33 PM Page 206
method: the rate of outdoor air entering a building can be recorded and the distribution
of air within a room can be quantified in terms of the airs age. Malfunctioning of systems
through short-circuiting can be easily detected. In a building with many rooms, the rooms
are connected to each other in a complex way. Outdoor air, for example, can reach a room
through different passages. The purging flow rate is the amount of air available for dilution
and can easily be recorded through the tracer gas technique.
Full-scale test rooms are an excellent tool where people can sense the indoor climate.
As shown by Oseland (1995), a context effect on thermal assessment may be obtained in
the sense that peoples perception of thermal conditions can be influenced by the
surrounding environment. As a result, it is important that the test room appears similar to
real life. In general, laboratory tests can be seen as a screening procedure; if a system
does not work in the laboratory, it will not work in reality.
This chapter is an attempt to briefly explain some of the more well-known
measurement techniques used in indoor climate research. Due to continuous development
in the field, a review of this kind may, of course, never be complete. However, we have
tried, as closely as possible, to cover current state-of-the-art technology. We have tried to
apply a practical approach, avoiding unnecessary formulae (which, in any case, may be
found in textbooks). The intention is, rather, to inform the reader of the different techniques
available, their pros, and cons, and the benefits expected.
For more detailed information, the reader is referred to existing literature in the field
(e.g. Doeblin, 1990) for a general approach on measurement techniques; for fluid
mechanics measurements, see Goldstein (1996) and Tavoluaris (2005).
M. Sandberg: (corresponding author) Department of Technology and Built Environment, Laboratory of
Ventilation and Air Quality, University of Gavle, Sweden; msg@HiG.se
H. Lundstrm: Department of Technology and Built Environment, Laboratory of Ventilation and Air Quality,
University of Gavle, Sweden
H. O. Nilsson: Department of Technology and Built Environment, Laboratory of Ventilation and Air Quality,
H. Stymne: Department of Technology and Built Environment, Laboratory of Ventilation and Air Quality,
ACKNOWLEDGEMENTS
We gratefully acknowledge the help from our colleagues Claes Blomqvist, Leif Claesson,
Per-ke Elvsn, Magnus Mattsson and Hans Wig.
REFERENCES
Adrian, R. J. (1991) Particle imaging techniques for experimental fluid mechanics, Annual Review of Fluid Mechanics, vol 23,
pp261304
Adrian, R. J. (2005) Twenty years of particle particle image velocimetry, Experiments in Fluids, vol 39, pp159169
strand, P. and K. Rodahl (1986) Textbook of Work Physiology, 3rd edition, McGraw-Hill, New York
ABER_CH_06 6/12/08 3:33 PM Page 207
Blomqvist, C. (2008) Conversion of electric heating in buildings: An unconventional alternative, Energy and Buildings, under review
Blomqvist, C. and M. Sandberg (2004) Air movements through horizontal openings in buildings: A model study, The
International Journal of Ventilation, vol 3, pp110
Blomsterberg, ., M. Sandberg and . Wahlstrom (2007) Behovsstyrd hybridventilation-mer n frstrkt sjlvdrag, in
Broschyr 2007, Formas, Stockholm
Brohus, H. and P. Nielsen (1996) CFD models of persons evaluated by full-scale wind channel experiments, in Proceedings of
the 5th International Conference on Air Distribution in Rooms (Roomvent 96), Technical University of Denmark and Danvac,
pp137144
Brel and Kjr (1982) Thermal Comfort Meter Type 1212: Instruction Manual, No 033-0699, Brel and Kjr Sound and
Vibration Measurement A/S, Nrum, Denmark
Bruun, H. H. (1995) Hot-Wire Anemometry, Principles and Signal Analysis, Oxford University Press, Oxford
Carrilho da Graca, G. and P. F. Linden (2003) Simplified modeling of cross-ventilation airflow, ASHRAE Transactions, vol 109,
pp114
Cehlin, M. (2006) Visualization of Airflow, Temperature and Concentration Indoors, KTH Research School, University of Gvle,
Gvle
Cehlin, M. and M. Sandberg (2007) Time Evolution of Gravity Currents Discharged from Low Velocity Diffusers, Roomvent
2007, Helsinki, Finland
Clausen, G. (2007) Human interaction with the indoor environment scientific background for design criteria: Human
comfort, health and productivity, in Seppnen, O. and Steri, J. (eds) Roomvent 2007, FINVAC ry, Helsinki
Christman, P. J. and J. Podzimek (1981) Hot-wire anemometer behaviour in low velocity air flow, Journal of Physics E:
Scientific Instruments, vol 14, pp4651
Dezi, G. (2006) An Assessment of Wind Comfort by Sand Erosion, University of Eindhoven, Eindhoven
Doeblin, E. O. (1990) Measurement and Systems: Application and Design, 4th edition, McGraw-Hill, Eindhoven, Holland
Dufton, A. (1936) The equivalent temperature of a warmed room, JIHVE (now the Journal of CIBSE), vol 4, pp227229
Elvsn, P.-. and M. Sandberg, (2008) Buoyant jet in a ventilated room: Velocity field, temperature field and air-flow patterns
analyzed with three different whole-field methods, Building and Environment, accepted for publication
Etheridge, D. and M. Sandberg (1996) Building Ventilation: Theory and Measurement, John Wiley, Chichester
Fan, J. (ed) (2006) Proceedings of the 6th International Thermal Manikin and Modeling Meeting, Hong Kong, 1618 October 2006
Fanger, P. (1970) Thermal Comfort, Danish Technical Press, Copenhagen
Fenger, J., O. Hertel and F. Palmgren (eds) (1998) Urban Air Pollution: European Aspects, Kluwer Academic Publisher,
Dordrecht, The Netherlands
Fingerson, L. M. (1994) Thermal anemometry, current state, and future directions, Review of Scientific Instruments, vol 65,
pp285300
Goldstein, R. J. (1996) Fluid Mechanics Measurements, Taylor and Francis, Washington, DC
Heiselberg, P. (2006) Hybrid ventilation, in Santamouris, M. and Wouters, P. (eds) Building Ventilation: State of the Art, James
& James, London
Heiselberg, P. and Z. Li (2007) Experimental study of buoyancy driven natural ventilation through horizontal openings, in
Proceedings of Roomvent, vol 2, Helsinki, Finland, pp141150
Humphreys, M. (1976) Field studies of thermal comfort compared and applied, Journal of the Institution of Heating and
Ventilating Engineers, no 44, pp 527
Hunt, G. R. and P. F. Linden (1999) The fluid mechanics of natural ventilation displacement ventilation by buoyancy-driven
flow assisted by wind, Building and Environment, vol 34, pp707720
Hitchin, E. R. and C. B. Wilson (1967) A review of experimental techniques for the investigation of natural ventilation in
buildings, Building Science, vol 2, pp5982
ABER_CH_06 6/12/08 3:33 PM Page 208
ISO 12569 (2000) Thermal Performance of Buildings: Determination of Air Change in Buildings Tracer Gas Dilution Method,
International Organization for Standardization, www.iso.org
ISO 7730 (2006) Moderate Thermal Environments: Determination of the PMV and PPD Indices and Specification of the
Conditions for Thermal Comfort, International Organization for Standardization, www.iso.org
ISO 14505-2 (2006) Ergonomics of the Thermal Environment Thermal Environment in Vehicles, International Organization for
Standardization, www.iso.org
ISO 16000-8 (2007) Determination of Local Mean Ages of Air in Buildings for Characterizing Ventilation Conditions,
International Organization for Standardization, www.iso.org
Liebert, R. M. and L. Langenbach-Liebert (1995) Science and Behavior: An Introduction to Methods of Psychological Research,
Prentice Hall International, New York
Loomans, M. (1998) The Measurement and Simulation of Indoor Airflow, PhD thesis, Eindhoven University of Technology, Eindhoven
Lundstrm, H., C. Blomquist, P. O. Jonsson and I. Pettersson (1990) A microprocessor-based anemometer for low air
velocities, in Proceedings of the 2nd International Conference of Engineering Aero- and Thermodynamics of Ventilated
Rooms, Roomvent 90, Oslo, Norway, 1315 June, ppB127
Lundstrm, H., M. Sandberg and B. Moshfegh (2007) Temperature dependence of convective heat transfer from fine wires in
air: A comprehensive experimental investigation with application to temperature compensation in hot-wire anemometry,
Experimental Thermal and Fluid Science, vol 32, pp649657
Madsen, T., B. Olesen and N. Kristensen (1984) Comparison between operative and equivalent temperature under typical
indoor conditions, ASHRAE Transactions, vol 90, part 1, pp10771090
Madsen, T., E. Soehrich and Z. Popiolek (1992) Measurement of draft sensation by a new skin element, Thermal Insulation
Laboratory, Technical University of Denmark, vol 8
Mattsson, M. (1999) On the efficiency of displacement ventilation, with particular reference to the influence of human
physical activity, Centre for Built Environment, Royal Institute of Technology, Gvle, Sweden
Mayer, E. and R. Schwab (1993) Evaluation of heat stress by an artificial skin, Proceedings of Indoor Air 93: The 6th
International Conference on Indoor Air Quality and Climate, vol 6, pp7378
McNair, H. P. (1972) A combined thermistors thermometer and anemometer, in Cockrell, D. J. (ed) Fluid Dynamic
Measurements, Leicester University Press, Leicester, pp250255
Melikov, A. K. (2004) Personalized ventilation, Indoor Air, vol 14, pp157167
Melikov, A. K., T. Sefker, Z. Popiolek, I. Care and W. Finkelstein (1998a) Requirements and guidelines for low-velocity
measurements, ASHRAE Transactions B, vol 104, no 1, pp15291542
Melikov, A. K., T. Madsen, and G. Langkilde (1998b) Impact of natural convection on the accuracy of low-velocity measurements
by thermal anemometers with omnidirectional sensor, ASHRAE Transactions, vol 104, no 1, part B, pp15191528
Melling, A. (1997) Tracer particles and seeding for particle image velocimetry, Measurement Science and Technology, vol 8,
pp4061416
Muller, D., B. Muller and U. Renz (2001) Three-dimensional particle streak tracking (PST): Velocity measurement of a heat
exchanger inlet flow, Measurement Science and Technology, vol 30, pp645656
Nicol, J. and M. Humphreys (2001) Adaptive thermal comfort and sustainable thermal standards for buildings, in Moving
Thermal Comfort Standards into the 21st Century, CD ROM
Nielsen, P. V. (2007) Analysis and design of room air distribution systems, HVAC and Research, vol 13, pp987997
Nielsen, P. V., S. Murakami, S. Kato, C. Topp and J. H. Yang (2003) Benchmark Tests for a Computer Simulated Person, Version
of 7 November 2003, Aalborg University, Aalborg, Denmark
Nielsen, P. V., N. M. Bartholomaeussen, E. Jakubowska, H. Jiang, O. T. Jonsson, K. Krawiecka, A. Mierzwjewski, S. J.
Thomas, K. Trampezynska, M. Polak and M. Soennichsen (2007) Chair with integrated personalized ventilation for
minimizing cross infection, in Roomvent 2007, FINVAC ry, Helsinki
ABER_CH_06 6/12/08 3:33 PM Page 209
Nilsson, H. O. (1999) The International Thermal Manikin Network, http://hem.passagen.se/nilssons/mer_manikin_network.htm
Nilsson, H. O. (2004a) Comfort Climate Evaluation with Thermal Manikin Methods and Computer Simulation Models, Royal
Institute of Technology, University of Gvle and the Swedish National Institute for Working Life, National Library of
Sweden, Sweden, http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-3726UT
Nilsson, H. (2004b) Evaluation and visualisation of perceived thermal conditions, European Journal of Applied Physiology,
Special Issue 5I3M, pp714716
Nilsson, H. and I. Holmr (2003) Comfort climate evaluation with thermal manikin methods and computer simulation models,
International Journal of Indoor Air Quality and Climate, vol 13, pp2837
Nilsson, H. O., H. Brohus and P. V. Nielsen (2007) CFD modeling of thermal manikin heat loss in a comfort evaluation
benchmark test, in Proceedings of the Roomvent 2007 Conference, Helsinki, Finland, 1315 June,
www.roomvent2007.org
Nordtest (2000) NT VVS 122-2000 Jet Momentum Flux: Determination Based on Weighing of Air Issuing from a Supply
Device, Nordtest, Espoo, Finland
NREL (National Renewable Energy Laboratory) (2006) ADAM: Thermal Manikin, www.nrel.gov/vehiclesandfuels/
ancillary_loads/adam.html
Okuyama, H. (1990) System identification theory of the thermal network model and an application for multi-chamber air flow
measurement, Building and Environment, vol 25, pp349363
Oseland, N. A. (1995) Predicted and reported thermal sensation in climate chambers, offices and homes, Energy and
Parson, K. (1993) Human Thermal Environments, Taylor and Francis, London
Popiolek, Z., A. K. Melikov, F. E. Jorgensen, W. Finkelstein and T. Sefker (1998) Impact of velocity and temperature
fluctuations on the accuracy of low-velocity measurements indoors by thermal anemometers, ASHRAE Transactions,
vol 104, no 1, part B, pp15071518
Sandberg, M. (1984) The multi-chamber theory reconsidered from the viewpoint of AOR quality studies, Building and
Sandberg, M. (1994) Measurement techniques in room air flow, in Roomvent 94, Krakow, Poland
Sandberg, M. (2007) Whole-field measuring methods in ventilated rooms, HVAC and Research Journal, vol 13, pp951970
Sandberg, M. and Etheridge, D. (1996) Building Ventilation, John Wiley, Chichester
Sandberg, M. and M. Mattsson (1993) Density Currents Created by Supply from Low Velocity Devices, Research report,
National Swedish Institute for Building Research, Gvle
Settles, G. S. (2001) Schlieren and Shadowgraph Techniques, Springer Verlag, Berlin
Soloff, S. M., R. J. Adrian and Z.-C. Liu (1997) Distortion compensation for generalized stereoscopic particle image
velocimetry, Measurements Science Technology, vol 8, pp14411454
Streblow, R., R. Rank and D. Muller (2007) An analysis of thermal plume above a cylindrical heat source in a room
with a vertical temperature gradient, in Seppnen, O. and Steri, J. (eds) Proceedings of Roomvent 2007, FINVAC ry,
Helsinki
Stull, R. B. (1988) An Introduction to Boundary Layer Meteorology, Kluwer Academic Publishers, Dordrecht, The Netherlands
Stymne, H., M. Sandberg and M. Mattsson (1991) Dispersion pattern of contaminant in a displacement ventilated room
implications for demand control, in Proceedings of the 12th AIVC Conference, Ottawa, Canada
Stymne, H., G. Emenius and C. A. Boman (2006) Passive tracer gas measurements of long term variation of ventilation in
three Swedish dwellings, International Journal of Ventilation, vol 5, pp333343
Tavoluaris, S. (2005) Measurements in Fluid Mechanics, Cambridge University Press, New York
VDI (2000) Umweltmeteorologie Physikalische Modellierung von Strmungs- und Ausbreitungsvorgngen in der
atmosphrischen Grenzschicht Windkanalanwendungen, Verein Deutscher Ingenieure e.V., Dsseldorf Germany
ABER_CH_06 6/12/08 3:33 PM Page 210
Keywords shading; optical properties; venetian blinds; roller shades
NOMENCLATURE
V standard photopic luminous efficiency function
E illuminance (lux)
E
s
normalized spectral solar power distribution
F
12
view/shape factor from 1 to 2
J irradiance (W/m
2
)
s length/slat width (m)
h distance between slats (m)
I irradiance (W/m
2
)
f view factor
nr number of inter-reflections
Po portion for each specular inter-reflection (percentage)
g total solar energy transmittance (percentage)
AS aspect ratio (percentage)
D Diameter (m)
Abstract
The optical properties of shading devices have an impact upon the daylighting and
thermal performance of a buildings perimeter spaces. This chapter presents an
overview of past and current approaches in computing the optical properties of shading
devices. Optical properties are determined by the different components of visible and
solar transmittance, reflectance and absorptance. Emphasis is given to the three major
types of shading used today: venetian blinds, roller shades/screens and draperies.
Comparisons between experimental procedures and analytical and numerical modelling
techniques are presented, and limitations in existing models and standards are
discussed. The final section describes the latest generalized methods and models for
characterizing complex fenestration/shading systems.
7
A Review of Optical Properties
of Shading Devices
Athanassios Tzempelikos
ABER_CH_07 6/12/08 3:34 PM Page 211
S material spacing (shades) (m)
L luminance (cd/m
2
)
dw solid angle (sr)
GREEK LETTERS
absorptance/solar altitude (percentage/degree)
s
relative surface solar altitude (degree)
p
profile angle (degree)
blind tilt angle/surface inclination (degree)
surface orientation (degree)
s
relative surface solar azimuth (degree)
solar incidence angle/polar coordinate (degree)
wavelength (m)
reflectance (percentage)
shining factor (percentage)
transmittance (percentage)
solar azimuth/polar coordinate (degree)
equivalent conductance for inward-flowing fraction (W/m
2
K)
cut-off angle (degree)
SUBSCRIPTS/SUPERSCRIPTS
f front
b back
dir direct
diff diffuse
sol solar
vis visible
s spectral/solar/surface
ex exterior
p profile
sky sky radiation
sys system property
h hemispherical
gl glazing
bl venetian blind
total total solar energy property
in inward flow/interior
sc screen
sh roller shade
x horizontal property
y vertical property
i, j, k surface indices
212 ATHANASSIOS TZEMPELIKOS
ABER_CH_07 6/12/08 3:34 PM Page 212
INTRODUCTION
The optical properties of shading devices have a decisive impact upon the thermal and
daylighting performance of perimeter spaces. It is therefore surprising that, currently,
there is no widely recognized procedure for measuring or calculating shading optical
properties (ISO 15099, 2003). Consequently, many shading manufacturers provide limited
or incomplete technical specifications of the optical and thermal properties of their
products. The data provided tends to be integral values that cannot satisfactorily resolve
the often strong angular dependency of shading devices. As a result, the products cannot
be reliably modelled in energy simulation tools since the software engines usually import
sets of properties from existing databases (WINDOW 5/THERM, 2003).
Optical properties of window shading refer to transmittance (), absorptance () and
reflectance () of window shadings with or without the window itself (for both the visible
and the solar spectrum). This chapter presents an overview of modelling and experimental
approaches previously used to predict the properties of shading devices, and reviews the
assumptions and limitations of these methodologies.
Despite the large number of shading devices available, the chapter focuses on the
three most common types of moveable shading devices used today in buildings: venetian
blinds, roller shades and draperies. Exterior fixed window obstructions, such as light
shelves, awnings, overhangs and fins, are excluded from the review since the properties
of these systems are generally known. Moreover, occupants usually have no control over
these fixed systems; therefore, they are generally easy to model and use in building
energy simulation software.
The location of the shading layer (interior, intermediate or exterior) is an important
factor; in this review, all three possible locations for the shading layer are considered:
interior, exterior and between the glazing panes (integrated).
Research on shading devices is divided into three categories: analytical models,
numerical models (e.g. ray-tracing, Monte Carlo) and experimental methods. Hybrid
models also exist (e.g. combining some experimental results with computer simulation).
The chapter summarizes and extends the findings of an earlier review report
(Tzempelikos, 2005).
OPTICAL PROPERTIES OF WINDOW SHADING DEVICES
The optical properties of windows equipped with shading devices depend upon:
1 the type of glazing(s) used;
2 the type of shading device used; and
3 the location of the shading device with regard to window glazing(s).
Venetian blinds, roller shades and draperies can be treated as shading layers located
parallel to the glazing plane, with intimate thermal and optical contact. The case of
venetian blinds is, of course, more complicated since they are a discrete shading device,
in contrast with shades and draperies, which are considered continuous or
homogeneous systems. The shading layer is usually modelled as a one-dimensional
layer similar to a pane or film, so the optical properties of the shading system are
A Review of Optical Properties of Shading Devices 213
ABER_CH_07 6/12/08 3:34 PM Page 213
a function of the device geometry and the position in the assembly. In general, the
equations remain similar to those for window glazing.
The solar optical properties of windows are usually calculated from the optical
properties of the individual glass layers. The three basic quantities transmittance ()
(assumed the same in front and back), front reflectance (
f
) and back reflectance (
b
) are
used (usually given for normal incidence) and they are wavelength dependent. These
properties can be estimated according to ISO 9050 (2003), or using any other standard
tool or method with acceptable results. Figure 7.1 shows a glazing system consisting of
N glass layers separated by non-absorbing gas layers.
Solving the recursion equations between layers (i) and (j), including inter-reflections,
yields:
[1]
[2]
[3]
A detailed explanation of the above may be found in Rubin et al (1998); Hollands et al
(2001); ISO 9050 (2003) or ASHRAE (2005). Because the above properties are a function
of wavelength, spectral-averaged values are obtained by integration over wavelength.
Moreover, angular properties for coated and uncoated glazings are calculated using optics

j i
b
j j
b
j j j i
b
j i
b
j j
f
, ,
, ,
, ,
=
2
1
1
1

i j
f
i j
f i j j j
f
j j
f
j i
b
, ,
, ,
, ,
=
1
1
2
1
1

i j
i j j j
j j
f
j i
b
,
, ,
, ,
=
1
1
1
FIGURE 7.1 Glazing system with N layers
ABER_CH_07 6/12/08 3:34 PM Page 214
theory. A description of glazing and window system properties is available in ASHRAE
(2005).
The WINDOW software (WINDOW 5/THERM, 2003) uses a large database of thermal
and optical properties of different types of glazings and windows, and can perform a
detailed optical and thermal analysis of all kinds of window systems (excluding shading
devices). Therefore, it is used by most of the major advanced building simulation software
to accurately simulate thermal and optical properties of windows (Zmeureanu, 1998).
DOE-2 software imports fenestration analysis results from WINDOW 4.1 (Reilly et al,
1995). TRNSYS software has a simplified option where the user inputs the window
transmittance (type 19) or window U-value (type 35), but also offers a more detailed option
(type 56), where the results of WINDOW software could be used as input (TRNSYS, 2003).
In ESP-r software, reflection, absorption and transmission of short-wave radiation are
evaluated separately for the visible and solar spectral portions, employing a large optical
properties database supported by an automated link to WINDOW software. In ESP-r,
advanced glazings and shading device systems can be supported. The thermal/optical
properties can vary with time to support the modelling of blinds and shading devices
(which can be controlled as a function of time, irradiance or temperature). ESP-rs special
materials facility can be used to impart behaviour to individual glazing layers (i.e.
transmittance switching in the case of an electrochromic layer). However, modelling a
venetian blind system in ESP-r is quite complex: building the geometry is not simple; the
optical properties are given for five incidence angles only; there are no correlations for
convective heat transfer between the blind and the air, etc. Therefore, researchers and
designers cannot use the software to model blinds.
The IMAGE project (Clarke et al, 1998) was a major European project aimed at
assessing the overall performance of advanced glazing systems using computer
simulation. In contrast to shading devices, a substantial amount of work has been done
in optical and thermal characterizations of complex glazings/prisms (Lorenz, 1998; Sullivan
et al, 1998; Beck et al, 1999; Karlsson and Roos, 2000, Lorenz, 2001), producing analytical
models based on detailed geometry and optics theory for the considered glazing system.
A particular characteristic of a shading device compared to normal glazings or films
is that the incident solar radiation may change direction while being transmitted or
reflected at the layer. For a complete evaluation of the spatial distribution of daylight,
calculations using the full matrix of transmission, forward and backward reflection and
absorption at each component for every angle of incidence are required. Consequently, it
is important to know the following solar properties of the solar shading device for each
angle of incidence:
Transmittance:
direct-to-direct transmittance:
dir,dir
;
direct-to-diffuse transmittance:
dir,diff
;
diffuse-to-diffuse transmittance:
diff,diff
.
The same respective parameters have to be calculated for reflectance:
dir,dir
,
dir,diff
and
diff,diff
.
ABER_CH_07 6/12/08 3:34 PM Page 215
Absorption:
[4]
[5]
In general, the above properties are a function of wavelength. Usually, a distinction is
made between visible or luminous/daylight transmittance,
vis
, and solar transmittance,
sol
. For both cases, the transmittance function has to be evaluated as a function of
incidence angle . The visible transmittance at incidence angle can be defined as:
[6]
where D
65
() is the spectral relative power distribution of the Commission Internationale
de lEclairage (CIE) standard illuminant and V() is the standard photopic luminous
efficiency function.
The solar transmittance at incidence angle is defined as:
[7]
where E
s
() is the normalized spectral power distribution of solar radiation. For a window
equipped with a layer-type shading device, the equations used for glazings remain the
same in concept. However, transmittance and reflectance are split into a backward and a
forward value. These two values are not necessarily equal because of redirection of
radiation. The sum of (
dir,dir
+
dir,diff
) is equal to the direct-to-hemispherical transmittance
dir,h
(and similarly for the reflectance). Note that according to ISO 15099 (2003), there is
no existing international standard for measuring these optical properties.
Once a beam transmitting through or reflecting from a solar shading device is split into
a direct and a diffuse part, the diffuse part continues its route through the system. This
implies that even for normal incidence solar radiation, for all other panes, films and
shading layers in the window, the
diff,diff
and
diff,diff
values are required. Thus, the values for
normal incidence provide insufficient information.
The following sections describe the methodologies and concepts used to extract
information about optical and solar properties of venetian blinds, roller shades and
draperies. In most cases, the shading device itself (without the window) is examined.
Starting from standards and analytical models, a review is provided of computer-based
numerical methods and experimental methods, and/or combinations than can be used for
simulating the optical properties of shading devices. Finally, the latest methods and
models of complex fenestration/shading systems are presented.
VENETIAN BLINDS
Venetian blinds are possibly the most complex type of shading in terms of optical and
thermal properties. They consist of separate, equally spaced horizontal louvres;
therefore, their properties change with louvre characteristics, tilt angle and angle of

sol s
E d ( ) ( ) ( , ) =

vis
D V d ( ) ( ) ( ) ( , ) =
65
Diffuse:
diff diff diff diff diff
= 1
, ,
.
Direct:
dir dir dir dir diff dir dir dir diff
= 1
, , , ,
;
ABER_CH_07 6/12/08 3:34 PM Page 216
incidence (see Figure 7.2). Most publications refer to interior daylighting conditions and
control of blinds without discussing their optical/thermal properties (Lee and Selkowitz,
1995; Lee et al, 1998; Dubois, 2003). Overviews of daylighting systems include several
types of venetian blinds and describe their function and characteristics, but provide very
limited (Laar and Grimme, 2002) or no information (Kischkoweit-Lopen, 2002) about their
optical (or thermal) properties. Because of the complexity of these systems and
limitations in existing models, many researchers work experimentally, trying to measure
the optical/thermal properties of different types of venetian blinds. The experimental
findings can be well used for simulation and perhaps this is the most accurate way of
modelling venetian blinds. The disadvantage of experiments is that they have to be
repeated for all kinds of blinds, which is very time-consuming. Analytical and numerical
models could handle all types of venetian blinds by changing the corresponding
parameters. The disadvantage of analytical models for venetian blinds is that they are
usually complex. However, there are some basic standards and analytical methods that
can be used to model their properties.
ASHRAE METHOD
The American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE,
2005) uses the following approach: first, it categorizes venetian blind systems (as well as
roller shades and draperies, as discussed later) in the exterior, between glass and interior.
For all cases, solar attenuation coefficients are used. These quantities serve the purpose
of the so-called shading coefficient for glazings and they describe the interaction of the
incident radiation with the shading system. In the case of exterior venetian blinds, blind
systems are separated into four categories, depending upon the colour and the width over
spacing ratio. For average values of transmittance and particular values of profile angles,
unshaded fractions and exterior solar attenuation coefficients are given (discussed in the
following section). Sample light transmittance, reflectance and absorptance data for
interior medium-coloured venetian blinds (slat angle 45 and width-to-slat spacing 1.2) are
provided for normal incidence only.
ASHRAE provides more data on specific venetian blind products mainly on thermal
characteristics. For cases that are not covered by the tables given in ASHRAE (2005), a
standard procedure for calculating optical properties based on Equation 1 is described in
FIGURE 7.2 Between-glazing and exterior venetian blinds
ABER_CH_07 6/12/08 3:34 PM Page 217
detail. In general, the methods and information described in ASHRAE (2005) are not
considered a systematic, detailed analysis of venetian blind systems, but provide some
data for estimating optical and thermal properties of particular venetian blind products
(the data, however, is not generalized). Another disadvantage is that venetian blinds are
assumed to be perfect diffusers.
ISO 15099
ISO 15099 (2003) also presents a solution technique for multilayer solar-optical models.
For a shading device consisting of parallel slats, optical properties can be determined as
a function of slat properties, geometry and position. The calculation is based on three
assumptions:
1 The slats are non-specular reflecting.
2 Any effects on window edges may be ignored.
3 Due to the assumption of non-specular reflection, a slight curving of the slats may be
ignored.
Based on these assumptions, the procedure considers two adjacent slats and subdivides
each of the slats into five equal parts, front (f) and back (b) (see Figure 7.3).
Source: ISO 15099 (2003)
FIGURE 7.3 Discretization of venetian blinds
ABER_CH_07 6/12/08 3:34 PM Page 218
Every slat is divided into five elements (the benefit of considering more elements is
supposed to be negligible). Notice that different properties can be assigned to every
element, particularly to every side of the slat. The process described below has to be
solved for every wavelength band required by the properties of the elements or by the rest
of the transparent system where the shading device is installed. The irradiance/illuminance
on each subsurface (f,i) and (b,i), with i from 0 to n (n = 6 here, including the vertical side
as n= 0) and normally for each spectral interval is calculated from:
[8]
[9]
where F
pq
is the view (shape) factor from surface p to surface q. The radiosity from the
external environment (J
ex
) is equal to the incident solar radiation, and it is assumed that
there is no internal radiation source (reflections from room interior, etc.).
Diffuse-to-diffuse transmission and reflection
Because specular reflection is excluded, view factors between sub-surfaces and between
the external environment and layers can be calculated using known diffuse radiation
exchange theory (Siegel and Howell, 1972). Diffuse-to-diffuse transmission is the ratio of
the radiation reaching the internal environment, divided by the incident solar radiation,
after solving the following equations (this procedure is not explained in ISO 15099, 2003):
[10]
for each wavelength. Similarly, for diffuse-to-diffuse reflection:
[11]
Direct-to-direct transmission and reflection
The beam radiation passing through the slats without touching can be calculated from the
tilt angle and aspect ratio of the slats for any incidence angle (see Figure 7.4). The direct-
to-direct transmittance is equal to:
[12]
for each wavelength and is a function of the incidence angle. No information is given on
how to actually calculate the direct-to-direct transmission. Obviously, the direct-to-direct
reflectance (
dir,dir
) is equal to zero since the slats are assumed to be perfect diffusers.
Direct-to-diffuse transmission and reflection
First, the parts of the slat (k) directly irradiated are computed for a certain incidence angle,
(see Figure 7.5). The view factors between the external environment and directly irradiated
dir dir
dir dir
ex
E
J
,
,
=
diff diff
b ex
ex
E
J
,
,
=
diff diff
b n
ex
E
J
,
,
=
E E F E F
b i b k f k b k b k b i f k b k f k f k b i
k
, , , , , , , , , , ,
[( ) ( ) ] =

i i
E E F E F
f i f k b k f k f k f i b k f k b k b k f i
k
, , , , , , , , , , ,
[( ) ( ) ] =

i i
ABER_CH_07 6/12/08 3:34 PM Page 219
parts are equal to unity, whereas view factors between the inside and outside are set to
zero to exclude the direct-to-direct component. The obvious equations are given for each
incidence angle:
[13]
dir diff
dir diff
f n
ex
b n
ex
E
J
E
J
,
,
,
,
'
!
1
1
1
+
1
1
1
'
!
1
1
1
+
1
1
1
=
'
!
1
1
1
1
1
11
+
1
1
1
1
1
1
'
!
1
1
1
1
1
1
+
1
1
1
1
1
1
Jex
Edir,dir
FIGURE 7.4 Direct-to-direct transmittance
Source: ISO 15099 (2003)
FIGURE 7.5 Directly irradiated parts of venetian blind
ABER_CH_07 6/12/08 3:34 PM Page 220
Absorption
Radiation that is neither transmitted nor reflected is absorbed by the blind, according to
Equations 4 and 5 for the direct and the diffuse components. This is the procedure
recommended by ISO 15099 (2003). All of the calculated quantities are essentially a
function of wavelength. Therefore, spectral averaging over the visible and the solar
spectrum should give the average visible and solar transmittance, reflectance and
absorptance of venetian blind systems. Following the above procedure, ISO 15099 (2003)
has an appendix containing tables of the optical properties of venetian blinds for four
different types of slats, two different slat angles (45 and 80) and incident solar radiation
values, including a translucent white slat system.
ENERGYPLUS METHODOLOGY
EnergyPlus (2007) models venetian blinds as a series of equally spaced slats that are
oriented horizontally or vertically (Crawley et al, 2002; EnergyPlus, 2007). The procedure
is similar to the one described by ISO 15099 (2003); but the geometry and equations
are all clearly stated and analysed. All of the slats are assumed to have the same optical
properties. The overall optical properties of the blind are determined by the slat geometry
(width, separation and angle) and the slat optical properties (front-side and back-side
transmittance and reflectance). Blind properties for direct radiation are also sensitive to
the profile angle, which is the angle of incidence in a plane that is perpendicular to the
window plane and to the direction of the slats. The following assumptions are made in
calculating the blind optical properties:
The spectral dependence of inter-reflections between slats and glazing is ignored;
spectral-average slat optical properties are used.
The slats are perfect diffusers. They have a perfectly matte finish so that reflection
from a slat is isotropic (hemispherically uniform) and independent of angle of
incidence (i.e. the reflection has no specular component). This also means that
absorption by the slats is hemispherically uniform with no incidence angle
dependence. If the transmittance of a slat is non-zero, the transmitted radiation is
isotropic and the transmittance is independent of the angle of incidence.
Inter-reflection between the blind and wall elements near the periphery of the blind
is ignored.
If the slats have holes through which support strings pass, the holes and strings are
ignored. Any other structures that support or move the slats are ignored.
The direct-to-direct and direct-to-diffuse transmittance of a blind is calculated using the
slat geometry depicted in Figure 7.6, which shows the side view of one of the cells of
the blind. For the case highlighted, each slat is divided into two segments so that the cell
is bounded by a total of six segments, denoted by s
1
to s
6
. The lengths of s
1
and s
2
are
equal to the slat separation, h, which is the distance between adjacent slat faces.
Lengths s
3
and s
4
are the segments illuminated by direct radiation. In the case shown in
Figure 7.6, the cell receives radiation by reflection of the direct radiation incident on s
4
and (if the slats have non-zero transmittance) by transmission through s
3
, which is
ABER_CH_07 6/12/08 3:34 PM Page 221
illuminated from above. The goal of the blind direct transmission calculation is to
determine the direct and diffuse radiation leaving the cell through s
2
for unit direct
radiation entering the cell through s
1
.
Direct-to-direct blind transmittance
As shown in Figure 7.6, some of the direct radiation passes through without hitting the
slats. Using the geometry, the equation is as follows:
[14]
where w can be expressed as a function of slat angle and profile angle:
[15]
where s is the slat width,
p
is the profile angle and is the slat angle.
Direct-to-diffuse transmittance
This is performed by using the radiosity method and view factors, similarly to the method
described in ISO 15099 (2003). However, all of the equations are explicitly presented and
the view factors between the six segments (two on each slat and two on the openings)
are computed using the cross-string method.
Diffuse-to-diffuse transmittance
Assuming that all surfaces are perfect diffusers, each slat is divided in two new sections of
equal length and the radiosity theory is again applied. The diffuse-to-diffuse calculations
are performed separately for solar, visible and long-wave slat properties in order to obtain
w s
p
p
=

i
cos( )
cos( )

dir dir
w
h
,
= 1
Note: In the figure, the profile angle is shown as
s
(
p
in the text) and slat angle is shown as
b
(
p
in the text)
Source: EnergyPlus (2007)
FIGURE 7.6 View of a cell formed by adjacent slats showing how the cell is divided into segments, s
i
, for
the calculation of direct solar transmittance; and side view of a cell showing case where some of the direct
solar passes between adjacent slats without touching either of them
ABER_CH_07 6/12/08 3:34 PM Page 222
the corresponding solar, visible and long-wave blind properties. Moreover, the diffuse-to-
diffuse solar properties of blinds are separated based on radiation coming from the ground
and radiation coming from the sky. The properties are obtained by integrating sky and
ground elements, as described in Equation 16 and Figure 7.7, treating each element as a
source of direct radiation of intensity I(
p
) incident on the blind at profile angle
p
:
[16]
If the sky radiation is uniform, then I
sky
is independent of
p
and Equation 16 becomes:
[17]
For the ground-diffuse reflectance,
p
is integrated from /2 to 0. A correction is also
made to the blind solar/optical properties in order to account for the amount of radiation

diff diff
sky
dir dir dir diff p p
d
, , ,
( ) cos( ) . =
i
0
2

diff diff
sky
dir dir p dir diff p sky p
I
,
, ,
( ) ( ) ( ) co
=
l
l
l
i i ss( )
( ) cos( )
.

p p
sky p p p
d
I d
0
2
0
2
i
Note:
s
denotes the profile angle,
p
in the text
FIGURE 7.7 Side view of horizontal blinds for calculating optical properties for sky and ground diffuse radiation
ABER_CH_07 6/12/08 3:34 PM Page 223
incident on the blind that is reflected or absorbed by slat edges (thickness). This
correction is not available in any other method.
Blind-glazing system
EnergyPlus (2007) also provides all of the required equations for calculating transmittance
and absorptance values for a combination glazing/blind system for both interior and
exterior blinds. These analytical expressions, presented below, can be directly used for
building simulation models.
Interior blind
The direct-to-hemispherical transmittance (sum of
dir,dir
+
dir,diff
) for a glazing/blind system
with an interior blind is given by:
[18]
where is the incidence angle,
p
the profile angle, gl stands for glazing and bl stands
for blind.
The diffuse system transmittance is equal to:
[19]
where the blind diffuse transmittance is described in Equation 17 for sky or ground
radiation.
Exterior blind
The direct-to-hemispherical transmittance for a glazing/blind system with an exterior blind
is given by:
[20]
The diffuse system transmittance is equal to:
[21]
Overall, the ISO 15099 (2003) method is similar, in principle, to the method explained
thoroughly in EnergyPlus (2007), except that the slats are divided into two parts instead
of five. The ISO 15099 (2003) and EnergyPlus (2007) results agree to within 12 per cent,
except for very small transmittance values. A comparison table given in the EnergyPlus
guide (2007, Table 28) provides more detailed data for particular products.

diff diff
sys diff diff
bl f
diff
gl
diff
gl f
diff
bl b
,
,
,
, ,
. =
i
i 1

dir h
sys
p dir dir
bl f
p dir
gl diff
gl
dir
gl
, ,
,
,
( , ) ( ) ( ) = i
i
ff
dir diff
bl b
diff
gl f
diff
bl b
di
i
i
,
,
, ,
1
\
)
rr diff
bl
p diff
gl
diff
gl f
diff
bl b
,
, ,
( )
.

i
i 1

diff diff
sys diff
gl
diff diff
bl f
diff
bl f
diff
gl b
,
,
,
, ,
=
i
i 1

dir h
sys
p dir
gl
dir dir
bl f
p dir diff
bl
p , ,
,
,
( , ) ( ) ( ) ( ) = i

diff
bl f
dir diff
bl f
p diff
gl b
diff
bl f
diff
gl
,
,
, ,
,
( ) i i
i 1
,,b
\
)

ABER_CH_07 6/12/08 3:34 PM Page 224
PREVIOUS RESEARCH ON THE SOLAR OPTICAL PROPERTIES
OF VENETIAN BLINDS
The first (and pioneer) extensive study on the effect of slat-type shading on heat and light
transmission was published by Parmelee and Aubele in 1952. For the first time, a
distinction between specular and diffuse slat surfaces was made. Transmittance and
absorptance of specular and diffuse slats were then plotted for different spacing
(opening) ratios, blind tilt angles and profile angles based on analytical expressions
presented in the appendices of the paper. Opening ratios were expressed as a function of
blind geometry and profile angle. The main assumptions in this study were:
the slats have negligible thickness;
slats are flat and opaque;
edge effects are negligible; and
graphs provided are strictly applicable to monochromatic radiation.
One of the most significant efforts in analytically determining blind solar optical properties
was completed by Pfrommer et al (1996). An analytical model derived from the fundamental
geometrical and physical characteristics of blind systems was developed for solar radiation
transport through slat-type blinds so that it could be directly used in a thermal simulation
engine. Light transmission through blinds was separated according to four paths:
1 direct-to-direct transmittance;
2 direct-reflected transmittance from a slat surface, then processed as
specular/diffuse/combinations;
3 diffuse-to-diffuse directly transmitted; and
4 diffuse-reflected transmittance on a slat surface and then transmitted.
Direct-to-direct transmittance is calculated from geometrical properties of the blind
(opening ratio described by Parmelee and Aubele, 1952). For direct-reflected
transmittance, a shining factor, , is used to separate direct-to-diffuse and specular-
reflected components. Direct-to-diffuse transmittance is calculated by recursive
calculations for two inter-reflections:
[22]
where is the slat reflectance (assumed constant and independent of wavelength), f
1
is
a view factor between the illuminated slat area and the inside, f
2
is between the
illuminated slat area and the upper slat, and f
3
is between the upper slat and the inside.
These view factors were approximated based on a method for infinitely small illuminated
areas (i.e. configuration factors). The term (1
diff
) accounts for the absorbed and directly
reflected component. Transmittance of the specular-reflected component is calculated
using the possible number of inter-reflections from the system geometry, nr
i
, and the
portion for each specular inter-reflection, Po
i
:
[23]

dir
spec
nr
i
i
Po =
1
i .

dir diff diff
f f f
,
( ) ( ) = i i i i
1
2
2 3
1
ABER_CH_07 6/12/08 3:34 PM Page 225
Diffuse-to-diffuse transmittance is computed by integrating incoming diffuse radiation
between the cut-off angles (
1
,
2
) (see Figure 7.8) and is given by:
[24]
where I
sky
is the sky radiation and () is calculated from:
[25]
s being the slat width, the slat angle and h the distance between slats. An excellent
comparison of Parmelee and Aubele (1952) with the Pfrommer et al (1996) model and with
experimental measurements identifying limitations and future work was published by
Chantrasrisalai and Fisher (2004).
The Advanced Window Information System (WIS) is a European software tool for
calculating optical and thermal properties of commercial and innovative window
systems, a big part of which are venetian blinds. One of the unique elements in this
software is the combination of glazings and shading devices. In fact, the way in which
WIS treats the solar optical properties of a layer-type shading device has been the basis

( )
sin( ) cos( ) tan( )
,

=

1
s s
h
i i i
diff diff
sky
sky
I d
I d
,
cos( ) ( )
cos( )
=
i i
i i

1
2
90
90
Cut-off angle
1
Cut-off angle
2
Source: Pfrommer et al (1996)
FIGURE 7.8 Description of cut-off angles
ABER_CH_07 6/12/08 3:34 PM Page 226
of ISO 15099 (2003). A comparison between WIS, Monte Carlo simulation and radiosity
results is available by Van Dijk and Oversloot (2003). In the WIS optical sub-model for
blinds, the spectral direct-hemispherical reflectance of the slats measured at normal
incidence is used throughout, although light strikes them at different angles when
passing between slats. The main assumptions are, first, that the slats are considered
Lambertian surfaces and, second, that the curvature of the slats is neglected. Using the
same procedure with ISO 15099 (2003), the slats are separated into five sections, as
depicted in Figure 7.3. For diffuse-to-diffuse properties, view factors are calculated
based on a three-dimensional model (LAMAS), developed under the European PASCOOL
project (Molina et al, 2000).
Parasol version 3 (Hellstrom et al, 2007) is another computer programme for
calculating the solar optical and thermal properties of windows with shades and blinds.
Users can input direct and diffuse transmittance and reflectance. For shades, the
transmittance is independent of the angle of incidence. For blinds, the ISO 15099 method
is used, in general, and specular and diffuse reflection are taken into account.
Effective properties of absorptance, reflectance and transmittance of venetian blinds
were calculated by Yahoda and Wright (2004), based on a radiosity theory. Results are
presented as a function of slat aspect ratio and blind tilt angle in categorized graphs. In
addition, the effect of curved slats was analysed by modifying view factors to account for
self-viewing.
Kuhn (2006) developed a generalized analytical model for the solar optical properties
of interior/exterior blinds, shades and other shading/glazing systems. The total solar
energy transmittance for a glazing/shading system with interior venetian blinds is
calculated by the following expression as a function of solar incidence angle, , profile
angle,
p
, blind tilt angle, , and wavelength, :
[26]
where
total
(
p
, , , ) is the radiation (at wavelength ) entering the room, equal to:
[27]
and q
in
total
(
p
, , , ) is the inward-flowing fraction of radiation (at wavelength ), equal to:
[28]
q g
total
in
p
gl gl
Absorbed in glazing
( , , , ) ( , ) ( , ) =

i
i i
i

( )
( , ) ( , , ) ( )
( , , )
1
1

gl bl
p diff
gl
bl
p
ddiff
gl
Reflected by blind and thenabsorbed in glazing
( )

i
i
i

gl
bl
p
diff
bl
diff
gl
in
( , )
( , )
( ) ( ) 1
1

tternal
Absorbed by theblind
\
)

total p
gl bl
p
Directly transmitted
( , , , ) ( , ) ( , , ) = i

i i i
gl bl
p diff
gl
diff
bl
( , ) ( , , ) ( ) ( , )
1
bbl
p
diff
gl
Reflected back inside from glass
( , , ) ( )
( )
i

g q
total total p total
in
p
( , , ) ( , , , ) ( , , , ) =
ABER_CH_07 6/12/08 3:34 PM Page 227
where:
[0,1] is an adjustable parameter for the outward-flowing fraction of energy that is
reflected by the interior blind and then absorbed by the glazing;

gl
is the glazing absorptance;

2
= 18W/m
2
K;
, where U
gl
is the glazing thermal conductance.
The model assumes that the solar optical properties are known (previously measured) for
each wavelength (i.e.
bl
(
p
, )) and identifies the parameters that determine each
property. It can be used for exterior blinds (appropriately modified) and for other shading
systems with known properties.
Wright and Kotey (2006) developed another general method for predicting beam and
diffuse radiation through multilayered glazing/shading systems. The method allows for
unequal front and back layer properties, and detailed solutions were obtained using matrix
manipulation. Kotey and Wright (2006) also published the algorithm for the case of
venetian blinds.
A disadvantage of all of these methods is that the direction of transmitted light cannot
be predicted. Therefore, the models are good for predicting solar radiation transmitted
through systems of blinds; but they are not accurate enough for detailed interior work,
plane illuminance prediction and daylighting simulation.
Another model for the optical properties of venetian blinds was inspired by the fact
that, for many cases, the slat reflectance is highly anisotropic, strongly peaked about the
specular direction (Breitenbach et al, 2001). The spectral specular and diffuse reflectances
at normal incidence are required as inputs to the model. In this study, Breitenbach et al
(2001) measured the visible and total solar energy transmittance of an integrated venetian
blind system using a gonio-spectrometer for different slat angles and incidence angles.
They then developed a model as follows: for the light that is diffusely reflected and that
undergoes two or more inter-reflections, a fraction F of it is assumed to proceed as a
specular component at each reflection. The fraction 1-F is reflected backwards. It retraces
its path, with a fraction absorbed at each reflection on its way back. This model therefore
contains one adjustable parameter, F, whose magnitude depends upon the degree to
which the diffusely reflected light is peaked about the specular direction. The fraction of
incident light undergoing n specular reflections was also calculated and a good
agreement was obtained between measured and predicted values.
The ALTSET project was part of the European Commissions Standards Measurement
and Testing programme from 1996 to 1999. Its main objective was to develop procedures
to measure the angular dependence of luminous and total solar energy transmittance of
complex glazings. Fixed and variable venetian blinds were included in the project. An
overview of recent developments in modelling optical and thermal properties of complex
windows (Rosenfeld et al, 2000) explains and compares available models, such as the
Breitenbach et al (2001) model with measurements and WIS results.
internal
gl
U
=
1
1 1
2
ABER_CH_07 6/12/08 3:34 PM Page 228
Due to the complexity of the problem and inaccuracies in the existing models, many
researchers prefer to work experimentally, obtaining the optical properties of venetian
blinds and then using them to create simulation models. An example is the work
published by Papamichael and Beltrn (1993), named the integrated directional
coefficients method. This hybrid method is based on the combination of scale-model
photometry and computer simulation. Physical scale models are first used to
experimentally determine a set of directional illuminance coefficients. These are defined
as the ratio of the interior illuminance on a point due to light coming from a set of
spherical coordinates (,), divided by the exterior illuminance due to light incident from a
direction normal to (,). The general state-of-the-art method is described in the section
on solar optical properties of complex fenestration and shading systems (page 233).
An experimental procedure for measuring the average solar transmittance of non-
homogeneous shading devices, such as venetian blinds, is presented by Aleo et al (1994),
to be used together with the LAMAS project. Two pyranometers were used (one exterior
and one moving interior), and the average direct and diffuse solar transmittance was
measured for three different types of venetian blinds at several tilt angles.
In another study, the total and the diffuse-to-diffuse solar transmittance of an
integrated venetian blind system was measured and modelled as a function of blind tilt
angle and profile angle (Athienitis and Tzempelikos, 2002; Tzempelikos and Athienitis,
2003). In the same study, the optimum blind tilt angle, which functions as a means of
blocking direct sunlight while maximizing view, was calculated as a function of blind and
solar geometry.
ROLLER SHADES, SCREENS AND DRAPERIES
ROLLER SHADES
Roller shades are probably the most commonly used shading devices due to their easy
operation and relatively low price. They are usually manually controlled to satisfy
occupants needs concerning privacy and glare issues (see Figure 7.9). The optical
Source: Chapter author
FIGURE 7.9 Examples of interior roller shades
ABER_CH_07 6/12/08 3:34 PM Page 229
properties of roller shades are generally assumed to be constant and independent of
incidence angle: they are considered perfect diffusers.
EnergyPlus (2007) model shades are perfect diffusers; however, a distinction
between shades and screens is made, as explained below. This means that direct
radiation incident on the shade is reflected and transmitted as hemispherically uniform
diffuse radiation: there is no direct component of transmitted radiation. It is also assumed
that the transmittance, reflectance, and absorptance are the same for the front and back
of the shade and are independent of angle of incidence. In this case, the glazing/shade
system transmittance is equal to:
[29]
where
sh
,
sh
,
diff
gl
are constant diffuse properties of the shade and the glass and
gl
()
is the glass transmittance, which depends upon the incidence angle. In published work
about modelling transmitted daylight through roller shades (Tzempelikos and
Athienitis, 2007), this same approach was followed. Nevertheless, recent studies have
shown that roller shades do not always perform like perfect diffusers, although they
could have a specular component and directional properties, depending upon their
construction and material (M. Collins, pers comm, 2007). Moreover, there is a strong
dependence upon wavelength visible and the solar optical properties are not equal
(see Figure 7.10).

diff
sys gl sh
diff
gl
sh
( )
( )
=
i
i 1
0 500 1000 1500 2000 2500
Wavelength (nm)
S
h
a
d
e

t
r
a
n
s
m
i
t
t
a
n
c
e

(
%
)
Transmittance (30 deg)
0
2
4
6
8
10
12
14
16
18
20
Source: Measurements by Dr M Collins, University of Waterloo
FIGURE 7.10 Spectral and directional transmission properties (different incidence angles) of a common
roller shade
ABER_CH_07 6/12/08 3:34 PM Page 230
SCREENS
EnergyPlus (2007) differentiates between shades and screens, defining screens as
exterior insect protection devices made up of metallic or non-metallic materials. The model
assumes that the screen is composed of intersecting orthogonally crossed cylinders, with
the surface of the cylinders assumed to be reflecting diffusely. The cylinder diameter and
spacing are supposed to be known and equal in both dimensions (see Figure 7.11).
The exterior screen is modelled as a planar semi-transparent sheet having specular
transmittance that is dependent upon the angle of incidence of beam solar radiation. The
screen transmittance algorithm includes two components. The first accounts for the
blockage of the suns rays by the screen material (i.e. the beam solar radiation passing
through the screen openings without hitting the screen material). The aspect ratio (AS) of
the screen must initially be determined. This is equal to the screen material (cylinder)
diameter, D, divided by the material spacing, S:
[30]
Then, the direct beam transmittance is computed as a function of the aspect ratio (AS), the
surface solar azimuth,
s
(equal to the difference between the surface orientation, , from true
north, and the solar azimuth, ), and the relative surface solar altitude,
s
(equal to the
difference between the surface inclination,
s
(from vertical), and the solar azimuth, ). For a
vertical window/shade,
s
would be equal to the solar zenith angle (or 90). Given the screen
diffuse reflectance,
sc
, and the screen aspect ratio, AS, the model takes the direction of solar
incidence,
s
and
s
, and calculates the direct beam transmittance, , as follows:
[31]
where
x
dir
and
y
dir
are the horizontal and vertical components of direct beam
transmittance of the shade, respectively, calculated by the solar and screen geometry:
[32]

dir
x
s s
AS ( , ) cos( ) sin( ) tan( ) cot [ cos (
cos
=

1 1
2
2 1
i i i
(( ) cos( )
cos( )
)]

s s
i
l
l
l
l
l

dir dir
sc
s s dir
x
s s dir
y
s s ,
( , ) ( , ) ( , ) = i

dir dir
sc
s s ,
( , )
AS
D
S
=
FIGURE 7.11 Screen model rendering of intersecting orthogonal crossed cylinders
ABER_CH_07 6/12/08 3:34 PM Page 231
and:
[33]
where and are given by:
[34]
[35]
The second part,
sc
dir,diff
(
s
,
s
), accounts for the additional flux of transmitted beam solar
radiation by diffuse reflectance (scattering) from the screen material (direct-to-diffuse
transmittance) and is calculated with an empirical algorithm (EnergyPlus, 2007).
It should be noted that although this method has some limitations (standard screen
geometry, equal spacing, empirical diffuse inward fraction), it is the only analytical method
available today for transmission through screens. Moreover, although EnergyPlus (2007)
refers to screens, the method can be well applied to roller shades using the appropriate
aspect ratio.
DRAPERIES
ASHRAE (2005) categorizes draperies based on appearance, yarn colour and weave
according to the chart shown in Figure 7.12. Classes may also be approximated by eye.
All of the existing models assume that draperies are perfect diffusers, with constant
properties independent of the incidence angle.
The only detailed study on the solar optical properties of draperies was completed
recently by Kotey et al (2007). The drapery was modelled as a series of uniformly arranged
rectangular pleats with optical properties that are dependent upon geometry, the optical
properties of the fabric and the profile angle of incident radiation. The fabric was assumed
to transmit and diffusely reflect any incident beam radiation. A radiosity-based method
was then applied to compute the transmittance and reflectance for different folding ratios
and profile angles.
SUMMARY
The experimental techniques discussed in this chapter, although giving an estimation of
transmittance values for venetian blind systems, can be used in simulation only if the
blind/glazing system is assumed to be a perfect Lambertian surface. However, this is not
usually the case since most venetian blinds have a specular component and the spatial
distribution of transmitted daylight cannot be modelled accurately. Numerical methods
provide a solution to this problem. In some cases, ray-tracing techniques using RADIANCE
(Reinhart and Walkenhorst, 2001) or Monte Carlo-based techniques (Tsangrassoulis et al,
1996; Campbell and Whittle, 1997) were used to predict daylight in rooms equipped with
exterior venetian blinds. The following section presents the latest developments and

l
l
l
l

tan tan( ) sec(cos (
cos( ) cos( )
cos( )
)) .
1 1
i
i
s s
=
cos cos ( ) cos ( ) sin ( )

1 2 2 2
s s s
i

dir
y
s s
s s
AS ( , )
cos[tan (tan( ) sec( ))] sin[tan (ta
=

1
1 1
i nn( ) sec( ))] tan( ) cot ( )

s s s s
i i i 1
2
2

l
l
l
l
l
ABER_CH_07 6/12/08 3:34 PM Page 232
state-of-the-art methods for measuring and computing optical properties of complex
fenestration/shading systems.
SOLAR OPTICAL PROPERTIES OF COMPLEX FENESTRATION AND SHADING
SYSTEMS: LATEST DEVELOPMENTS AND STATE OF THE ART
An accurate evaluation of daylight distribution through complex fenestration systems
(such as venetian blinds) requires knowledge of solar optical effective properties as a
function of both incident and emerging directions of light. This is achieved using the
concept of bidirectional transmission (or reflection) distribution functions (BT(R)DFs). The
idea was first proposed by Klems and Warner (1995) and was then developed by
Andersen (2002). BTDFs are defined as follows (see Figure 7.12):
[36]
where:
(
1
,
1
) and (
2,
2
) are the polar coordinates of the incoming and transmitted light flux,
respectively;
BT R DF
L
L d
( ) ( , , , )
( , , , )
( , ) cos( )

1 1 2 2
2 1 1 2 2
1 1 1 1 1
= =
i i
LL
E
cd
m lx
2 1 1 2 2
1 1
2
( , , , )
( )
( )

l
l
l
l
l
Source: ASHRAE (2005)
FIGURE 7.12 Categorization of fabrics for draperies/shades
ABER_CH_07 6/12/08 3:34 PM Page 233
L
1
(
1
,
1
) and L
2
(
1
,
1
,
2
,
2
) are the luminances of an element of incoming and
transmitted light flux;
d
1
is the solid angle subtended by incoming light flux; and
E
1
(
1
) is the illuminance on the sample plane due to the incident light flux.
BT(R)DFs are measured with bidirectional gonio-photometers. Most of these techniques
are based on a scanning process, which is usually very time-consuming. However, there
are other methods that are based on a video approach, detecting the light through digital
video capture after being collected on a device-specific projection surface. As explained
by Andersen and Scartezzini (2005) and Andersen and de Boer (2006):
. . . light emerging from the sample is reflected by a diffusing triangular panel towards
a charge-coupled device (CCD) camera, used as a multiple-points luminance meter. The
diffusing coating of the screen is necessary both because the camera must be able
to capture light reflected by any area of the screen independently of the location of this
area and to avoid any correlation between the cameras position and the
Source: Andersen and Scartezzini (2005)
FIGURE 7.13 Bidirectional reflection (a) and transmission (b) distribution functions
ABER_CH_07 6/12/08 3:34 PM Page 234
measurements. After six 60 rotations of the screen-camera system, the emerging light
distribution is fully determined. For reflection measurements, some additional
constraints appear due to the conflict of incident and emerging light flux. The incoming
beam needs to penetrate the measurement space and reach exactly the sample
surface, therefore requiring a special opening through the measurement space
envelope and the removal of screen covers when the latter is obstructive.
Andersen et al (2003b) also presented the first extensive comparison of BTDF data with
ray-tracing simulation. Particular applications for venetian blinds can be found in Andersen
et al (2003a, 2003b, 2005). The inclusion of the specular component in the assessment
of BTDF (Andersen and Scartezzini, 2005) set the framework for future work on computing
solar optical properties of venetian blind products.
Using measured or simulated BTDFs, researchers can model the optical properties of
venetian blinds and then use one of the available methods for processing transmitted flux
with known directional properties. Reinhart and Andersen (2006) used gonio-photometer
and integrated sphere measured data and then used RADIANCE with the Perez et al (1993)
sky model and a daylight coefficients approach (Tregenza, 1983) in order to model the
performance of a translucent panel. De Boer (2006) presented a method for including
BTDF raw data sets into an appropriate format for daylighting simulation. The new version
of EnergyPlus incorporates BTDFs in its daylighting simulation engine, according to
Crawley et al (2002).
CONCLUSION
An overview of past and current approaches for computing the optical properties of
shading devices was presented in this chapter; emphasis was placed on venetian blinds,
roller shades/screens and draperies, the most commonly used systems. Although the
international standard ISO 15099 (2003) describes a general procedure for calculating
(a) (b)
Source: Andersen and Scartezzini (2005)
FIGURE 7.14 Detection of the light transmitted through a sample: (a) specular component against diffuse
transmission; (b) light transmission and detection with the digital imaging-based photogoniometer
ABER_CH_07 6/12/08 3:34 PM Page 235
these properties, there is still no existing standard for experimental measurement
procedures.
Several analytical and numerical methods have been developed in order to model the
optical behaviour of fenestration systems and shading attachments. In most of the
studies, objectives are to calculate the effective solar optical properties: direct-to-direct,
direct-to-diffuse and diffuse-to-diffuse transmittance and reflectance.
For direct-to-direct properties, the quantities are computed based on the shading
geometry (e.g. blind tilt angle, openness ratio and screen construction) and the solar
geometry. For direct-to-diffuse properties, empirical calculations are often used for estimation
of the specular component, based on material properties and geometry. Otherwise, the
shading surface is assumed to be a perfect diffuser and radiosity-based methods are utilized.
Finally, for diffuse-to-diffuse properties, a radiosity analysis is performed, based on the
system geometry and incoming radiation from the sky and the ground.
The main assumptions in most of the models are that:
The blinds are considered perfect diffusers.
Proper inter-reflections are ignored.
Slat curvature/thickness/edge effects are not considered.
Some of the analytical methods which can be used directly in building simulation
software overcome the above limitations by modifying transmittance modes. However,
the models assume that some of the basic properties are already known (from
measurements). Moreover, the direction of reflected/transmitted light is usually unknown
and therefore the models are not accurate enough for detailed daylighting simulation.
Venetian blinds are probably the most complex of all shading systems; therefore the
majority of research has focused on this type of shading. Roller shades are modelled as
perfect diffusers that are independent of shade transmittance, and their properties are
considered constant on both shade surfaces. An advanced model for simulating optical
properties of screens was only found in EnergyPlus (2007). More research is needed in
this direction since, depending upon shade construction and material properties, the
specular component could have an effect and, as a result, the direct-to-diffuse
transmittance may not be modelled accurately.
Ray-tracing techniques provide another viable solution and include the specular
component. Recently, development of bidirectional transmission (or reflection)
distribution functions (BT(R)DFs) permitted a more accurate evaluation of radiation
transmission/reflection through complex fenestration/shading systems. This method set
the framework for modelling the solar optical properties of any fenestration/shading
system. Transmitted/reflected components are computed as a function of both incident
and emerging directions of radiation. These functions are measured with bidirectional
gonio-photometers that scan a hemisphere around the measured sample and can include
the specular components. Comparison of BTDF measured data with ray-tracing simulation
has shown good agreement. The bidirectional transmittance distribution functions
(measured or simulated) can be used to model the optical properties of complex shading
systems; researchers may then use one of the available methods for processing
ABER_CH_07 6/12/08 3:34 PM Page 236
transmitted flux with known directional properties (i.e. ray-tracing) in order to evaluate
detailed illuminance distributions inside buildings.
Athanassios Tzempelikos: Solar Buildings Research Network, Department of Building, Civil and Environmental
Engineering, Concordia University, 1455 de Maisonneuve West, EV. 6-139, Montreal, Quebec, Canada H3G
1M8; www.solarbuildings.ca; tel 514-848-2424/7080; agtzemb@alcor.concordia.ca
ACKNOWLEDGEMENTS
The author would like to thank the National Research Council of Canada for financial
support, and especially Aziz Laouadi and Christoph Reinhart from the Institute of Research
in Construction in Ottawa for providing resources and useful technical comments.
REFERENCES
Aleo, F., S. Sciuto and R. Viadana (1994) Solar transmission measurements in outdoor conditions of non-homogeneous
shading devices, in Proceedings of the European Conference on Energy Performance and Indoor Climate in Buildings,
Lyon, France, pp10741079
Andersen, M. (2002) Light distribution through advanced fenestration systems, Building Research and Information, vol 30,
no 4, pp264281
Andersen, M. and J. de Boer (2006) Goniophotometry and assessment of bidirectional photometric properties of complex
fenestration systems, Energy and Buildings, vol 38, pp836848
Andersen, M. and J. L. Scartezzini (2005) Inclusion of the specular component in the assessment of bi-directional distribution
functions based on digital imaging, Solar Energy, vol 79, no 2, pp159167
Andersen, M., M. Rubin and J.-L. Scartezzini (2003a) Comparison between ray-tracing simulation and bi-directional
transmission measurements on prismatic glazing, Solar Energy, vol 74, no 3, pp157173
Andersen, M., J. L. Scartezzini, M. D. Rubin and R. C. Powles (2003b) Bi-directional light transmission properties assessment
for venetian blinds: Computer simulations compared to photogoniometer measurements, in Proceedings of ISES Solar
World Congress, Gotenborg, Sweden, pp187198
Andersen, M., M. D. Rubin, R. C. Powles and J. L. Scartezzini (2005) Bi-directional transmission properties of venetian blinds:
Experimental assessment compared to ray-tracing calculations, Solar Energy, vol 78, no 2, pp187198
ASHRAE (American Society of Heating, Refrigerating and Air-Conditioning Engineers) (2005) ASHRAE Handbook:
Fundamentals, American Society of Heating, Refrigerating and Air-Conditioning Engineers, Atlanta, GA, Chapter 30
Athienitis, A. K. and A. Tzempelikos (2002) A methodology for simulation of daylight room illuminance distribution and light
dimming for a room with a controlled shading device, Solar Energy, vol 72, no 4, pp271281
Beck, A., W. Krner, O. Gross and J. Fricke (1999) Making better use of natural light with a light-redirecting double-glazing
system, Solar Energy, vol 66, no 3, pp215227
Breitenbach, J., S. Lart, I. Langle and J. L. J. Rosenfeld (2001) Optical and thermal performance of glazing with integral
venetian blinds, Energy and Buildings, vol 33, pp433442
Campbell, N. S. and J. K. Whittle (1997) Analyzing radiation transport through complex fenestration systems, in Proceedings
of 5th IBPSA Conference, Prague, Czech Republic, pp173180
Chantrasrisalai, C. and D. E. Fisher (2004) Comparative analysis of one-dimensional slat-type blind models, in Proceedings of
SimBuild 2004, Boulder, CO, pp110
ABER_CH_07 6/12/08 3:34 PM Page 237
Clarke, J. A., M. Janak and P. Ruyssevelt (1998) Assessing the overall performance of advanced glazing systems, Solar
Energy, vol 63, no 4, pp231241
Crawley, D. B., L. K. Lawrie, C. O. Pedersen, R. K. Strand, F. C. Winkelmann, W. F. Buhl, M. J. Witte, R. J. Henninger, D. E.
Fisher and D. Shirey (2002) EnergyPlus: New, capable and linked, in Proceedings of 2nd eSim Conference, Montreal,
Quebec, Canada, pp244251
De Boer, J. (2006) Modelling indoor illumination by complex fenestration systems based on bidirectional photometric data,
Dubois, M.-C. (2003) Shading devices and daylight quality: An evaluation based on simple performance indicators, Lighting
Research and Technology, vol 35, no 1, pp6176
EnergyPlus (2007) Engineering Reference: The Reference to EnergyPlus Calculations, US Department of Energy,
Washington, DC
Hellstrom, B., H. Kvist, H. Hakansson and H. Bullow-Hube (2007) Description of Parasol 3.0 and comparison with
measurements, Energy and Buildings, vol 39, pp279283
Hollands, K. G. T., J. L. Wright and C. G. Granqvist (2001) Solar Energy: The State of the Art, ISES Position Papers, James &
James, London, Chapter 2
ISO 15099 (2003) Thermal Performance of Windows, Doors and Shading Devices: Detailed Calculations, 1st edition,
International Organization for Standardization, Geneva
ISO 9050 (2003) Glass in Building: Determination of Light Transmittance, Solar Direct Transmittance, Total Solar Energy
Transmittance, Ultraviolet Transmittance and Related Glazing Factors, International Organization for Standardization,
Geneva
Karlsson, J. and A. Roos (2000) Modeling the angular behavior of the total solar energy transmittance of windows, Solar
Energy, vol 69, no 4, pp321329
Kischkoweit-Lopin, M. (2002) An overview of daylighting systems, Solar Energy, vol 73, no 2, pp7782
Klems, J. H. and J. L. Warner (1995) Measurement of bi-directional optical properties of complex shading devices, ASHRAE
Transactions, vol 101, no 1, pp791801
Kotey, N. A. and J. L. Wright (2006) Simplified solar optical calculations for windows with venetian blinds, in Proceedings of
the 31st Conference of the Solar Energy Society of Canada Inc (SESCI) and 1st Solar Buildings Conference (SBRN),
Montreal, Quebec, Canada, 2024 August 2006
Kotey, N. A., J. L. Wright and M. R. Collins (2007) A simplified method for calculating the effective solar optical properties of
a drapery, in Proceedings of the 32nd Conference of the Solar Energy Society of Canada Inc (SESCI) and 2nd Buildings
Conference (SBRN), Calgary, Alberta, Canada, 1014 June 2007
Kuhn, E. (2006) Solar control: A general evaluation method for faades with venetian blinds or other solar control systems,
Laar, M. and F. W. Grimme (2002) German developments in daylight guidance systems: An overview, Building Research and
Information, vol 30, no 4, pp282301
Lee, E. S. and S. E. Selkowitz (1995) The design and evaluation of integrated envelope and lighting control strategies for
commercial buildings, ASHRAE Transactions, vol 101, no 1, pp326342
Lee, E. S., D. L. Di Bartolomeo and S. E. Selkowitz (1998) Thermal and daylighting performance of an automated venetian
blind and lighting system in a full-scale private office, Energy and Buildings, vol 29, pp4763
Lorenz, W. (1998) Design guidelines for a glazing with a seasonally dependent solar transmittance, Solar Energy, vol 63,
no 2, pp7996
Lorenz, W. (2001) A glazing unit for solar control, daylighting and energy conservation, Solar Energy, vol 70, no 2, pp109130
Molina, L. J., R. I. Maestre and E. Lindawer (2000) A model for predicting the angular dependence of the optical properties of
complex windows including shading devices, in Proceedings of PLEA, Cambridge, UK, pp765770
ABER_CH_07 6/12/08 3:34 PM Page 238
Papamichael, K. and L. Beltrn (1993) Simulating the daylight performance of fenestration systems and spaces of arbitrary
complexity: The IDC method, in Proceedings of 3rd IBPSA Conference, Adelaide, Australia, pp509515
Parmelee, G. V. and W. W. Aubele (1952) The shading of sunlit glass, ASHRAE Transactions, vol 58, pp377398
Perez, R., R. Seals and J. Michalsky (1993) All-weather model for sky luminance distribution preliminary configuration and
validation, Solar Energy, vol 50, no 3, pp235245
Pfrommer, P., K. J. Lomas and C. Kupke (1996) Solar radiation transport through slat-type blinds: A new model and its
application for thermal simulation of buildings, Solar Energy, vol 57, no 2, pp7791
Reilly, M. S., F. C. Winkelmann, D. K. Arasteh and W. L. Caroll (1995) Modeling windows in DOE2.1E, Energy and Buildings,
vol 22, pp5966
Reinhart, C. F. and M. Andersen (2006) Development of a radiance model for a translucent panel, Energy and Buildings,
vol 38, pp890904
Reinhart, C. F. and O. Walkenhorst (2001) Validation of dynamic: Radiance-based simulations for a test office with external
blinds, Energy and Buildings, vol 33, pp683697
Rosenfeld, J. L. J., W. J. Platzer, H. Van Dijk and A. Maccari (2000) Modeling the optical and thermal properties of complex
glazing: Overview of recent developments, Solar Energy, vol 69, no 6, Supplement, pp113
Rubin, M., K. Von Rottkay and R. Powles (1998) Window optics, Solar Energy, vol 62, no 3, pp149161
Siegel, R. and J. Howell (1972) Thermal Radiation Heat Transfer, McGraw-Hill Inc, Columbus, OH
Sullivan, R., L. O. Beltrn, E. S. Lee, M. D. Rubin and S. E. Selkowitz (1998) Energy and daylight performance of angular
selective glazings, in Proceedings of Thermal Performance of Exterior Envelopes VII, Clear Water, FL, pp319328
Tregenza, P. R. (1983) The Monte Carlo method in lighting calculations, Lighting Research and Technology, vol 15, no 4,
pp163170
TRNSYS 15 (2003) Reference Manual, TRANSSOLAR/TESS
Tsangrassoulis, A., M. Santamouris and D. Assimakopoulos (1996) Theoretical and experimental analysis of daylight
performance for various shading systems, Energy and Buildings, vol 24, pp223230
Tzempelikos, A. (2005) Literature Review on Convection Heat Transfer Correlations in Skylight Cavities, and on Thermal and
Optical Prediction Models of Window Shadings, Final report submitted to IRCNRCC, May 2005
Tzempelikos, A. and A. K. Athienitis (2003) Modeling and evaluation of a window with integrated motorized venetian blinds,
in Proceedings of 3rd ISES World Congress, Gotenburg, Sweden
Tzempelikos, A. and A. K. Athienitis (2007) The impact of shading design and control on building cooling and lighting
demand, Solar Energy, vol 81, no 3, pp369382
Van Dijk, D. and H. Oversloot (2003) WIS, the European tool to calculate thermal and solar properties of windows and
window components, Proceedings of the 8th IBPSA Conference, Eindhoven, The Netherlands, pp259266
WINDOW 5/THERM (2003) NFRC Simulation Manual, Lawrence Berkeley National Laboratory, CA
Wright, J. L. and N. A. Kotey (2006) Solar absorption by each element in a glazing/shading layer array, ASHRAE
Transactions, vol 112, no 2, pp312
Yahoda, D. S. and J. L. Wright (2004) Methods for calculating the effective long-wave radiative properties of a venetian blind
layer, ASHRAE Transactions, vol 110, no 1, pp463474
Zmeureanu, R. (1998) Defining the Methodology for the Next Generation HOT2000 Simulator, Internal report submitted to
CANMET
ABER_CH_07 6/12/08 3:34 PM Page 239
ABER_CH_07 6/12/08 3:34 PM Page 240
ABL see atmospheric boundary layer
absorption
contaminant interactions 59, 623, 66,
81, 88
shading devices 21516, 221
solar cooling 12831, 138
see also sorption
acceptance angles 36, 45
access, solar 99
active tracer gas methods 1989
ADAM see advanced automotive manikin
adaptive principle 1656
adsorption
contaminant interactions 5960, 623,
66, 79, 88
solar cooling 1289, 131, 138
see also sorption
adsorption isotherm models 66
advanced automotive manikin (ADAM)
1867
Advanced Window Information System (WIS)
2267
air
contaminant modeling 634
mean age 176, 178, 2001
temperature 102, 1968
urban design 102, 1056
velocity 1913
see also ventilation
air conditioning 26, 12357
air phase diffusion 6873, 75
albedo 105, 11213
alternative cooling technologies 126
ALTSET project 228
American Society of Heating, Refrigerating
and Air-Conditioning Engineers (ASHRAE)
614, 29, 21718, 232
anemometers 1914, 201
anidolic systems 501
architectural design 95121, 125
Arrhenius-type law for temperature
dependency 801
artificial sky 52
ASHRAE see American Society of Heating,
Refrigerating and Air-Conditioning
Engineers
aspect ratios 97, 1046, 114, 231
atmospheric boundary layer (ABL) 183
attics 1315
axisymmetric jets 172
bag method 201
bead thermistors 189, 192
behaviour 162, 166
benchmark comfort zone tests 2034
bidirectional flow 179
bidirectional transmission (or reflection)
distribution functions (BT(R)DFs) 2334,
2367
bidirectional transmittance distribution
function (BTDF) 523, 233, 2356
bimolecular reactions 878
blind systems 359, 213, 21629,
232, 236
boundary layers 59, 625, 17980
BTDF see bidirectional transmittance
distribution function
BT(R)DFs see bidirectional transmission
(or reflection) distribution functions
building materials
sorption and diffusion 6973
urban design 11213, 115
see also non-porous materials; porous
materials
bulk density flow 1789
buoyancy 16970, 1725
California 34, 6, 1421, 289
canopy systems 49
carbon/methanol adsorption heat
pumps 131
Carnot diagram 134, 137
CCDs see charge-coupled devices
ceilings 1415, 50, 197
CFD see computational fluid dynamics
models
charge-coupled devices (CCDs) 234
chemical reaction heat pumps 1312, 138
I ndex
ABER_Index 6/12/08 1:52 PM Page 241
chemical transformations 868
chemisorption cooling technologies 1289
Chicago, Illinois 223
city design 95121
see also urban areas
climate
urban 95121, 125
zones 11, 17
climatic parameter measurement 188
closed systems 129
clothing 2023
coefficient of performance (COP) 1301,
135, 138, 144
comfort 125, 1646, 2024
Comfortina 186, 203
commercial level
cool roofs 223
SAC 1407
competitive aspects
contaminant interactions 825
SAC systems 1234, 1467, 154
compound parabolic concentrators (CPCs)
379, 51
computational fluid dynamics (CFD) 634,
2036
computer-simulated persons (CSPs) 204
constant concentration method 181, 200
see also tracer gas technique
constant injection method 200
contact temperature measurements
1889
contaminants
interactions 573
tracer gas methods 1989, 2067
ventilation 160, 1768, 2067
continuous sorption processes 136
control systems 35, 43, 145
convective loss measurements 1878
convective mass-transfer coefficients
635
cool coatings 113
cooling
solar-assisted 12357
vegetation 10710
cool-roof standards 132
COP see coefficient of performance
core systems 34, 4151
costs
cool roofs 16, 1819, 21
daylighting systems 54
SAC systems 1412, 146
CPCs see compound parabolic concentrators
credits, cool roof 6, 89, 1215, 23, 25
CSPs see computer-simulated persons
cup method 756
cut-off angles 226
data loggers 2012
daylighting systems 3356, 217
DAYSIM 53
decay rate method 2001
DEM see digital elevation model
densimetric Archimedes number 172
density
building ventilation 1789
urban 978, 1012, 105
dependency, temperature 801
deposition velocity approach 867
desiccant cooling systems 129,
1378, 143
design, city 95121
desorption 59, 63, 88, 128
see also absorption; adsorption; sorption
developed countries 140
diffuse reflectance 228
diffuse-to-diffuse transmittance and
reflectance 219, 2226, 236
diffuse transmittance 235
diffusion 602, 878, 221, 230, 236
diffusion-cell methods 75
diffusion coefficients 6877, 801, 84
diffusivity 6875, 78, 805
digital cameras 195
digital elevation model (DEM) 1067
digital image-based photogoniometers 235
dilute liquid solutions 63
dilution capacity 176, 180
dimensions of rooms 1714
direct beam transmittance 231
directional illuminance coefficients 229
directional transmission 230
direct-reflected transmittance 225
direct-to-diffuse transmission and reflectance
21922, 236
direct-to-direct transmittance and reflectance
21922, 225, 236
direct-to-hemispherical transmittance 224
discontinuous sorption processes 135
displacement height 182
displacement ventilation 1745, 1778
DOE-2 software 215
doorways 179
242 ADVANCES IN BUILDING ENERGY RESEARCH
draperies 2323
ducts 13, 181, 201
economic level
SAC systems 1412
urban planning 100
effects, number of 1301
efficiency see energy efficiency
electrical thermometers 188
electricity
cool roofs 25
elemental models 5861
ellipsoid sensors 187
emission profiles 77
emissivity 190
energy
conservation 223, 98
consumption 1101
efficiency 132, 1401
energy performance factor (EPF) 26
EnergyPlus 2214, 2301
energy savings
cool roofs 25, 16, 1821
SAC systems 1445
vegetation 110
Energy Star Label, EPA 267
entropytemperature (ST) diagrams
1356
envelopes 25, 1834
environmental level
parameter modeling 7981
SAC systems 124, 1434, 146,
14850
urban climates 96
Environmental Protection Agency (EPA), US
267
EPA see Environmental Protection Agency
EPF see energy performance factor
equivalent temperature (ET) 1667
ESP-r 53, 215
ET see equivalent temperature
European glazing systems performance
215
exhaust hoods 177
experienced temperature 167
exposure, solar 99, 1014
exterior fixed window constructions 213,
217, 224
extract ventilation 162
fan-powered ventilation see mechanical
ventilation
FBC see Florida Building Code
fenestration systems 3441, 2335
see also windows
fibre optics 437
Ficks law of diffusion 601, 74
field comfort studies 166
flat surface sensors 1878
Florida Building Code (FBC) 235
flow markers 195
flow rate measurement 199200
flows between rooms 17880
forced convection 65
Fresnel lenses 456
full-scale replicas 2045, 207
gaseous contamination interactions 5793
glass fibres 44
glazing systems 113, 214
see also fenestration systems; windows
global urban cooling demand 126
global warming 125
gonio-photometers 52, 2346
gonio-spectrometers 228
Green Building Rating System, LEED 278
ground-diffuse reflectance 2234
ground surface temperature 111
Guam 256
Hawaii 25
heat balance, human 164
heat flux 109, 1847
heating, ventilation and air conditioning
(HVAC) systems 125, 140, 160
heat island effect 97, 107, 112, 125
heat pumps 12738
Heliobus system 48
heliostats 413, 456
Henrys law for dilute liquid solutions 63
Henrys law for ideal absorption 66, 68, 82,
85, 88
heterogeneous processes see chemical
transformations
Himawari system 45
HOEs see holographic optical elements
holographic optical elements (HOEs) 41
homogeneous emission method 200
homogeneous reactions 87
horizontal light pipes 489
horizontal venetian blinds 223
INDEX 243
hot-sphere anemometers 192
hot-wire anemometer 1912, 201
humans
behaviour 162, 166
microclimates 98
temperature regulation 1635
humidity 68, 846
HVAC see heating, ventilation and air
conditioning systems
hybrid solar lighting systems 456
hygienic ventilation flow rates 160
IAQ see indoor air quality
IAS see ideal adsorption theory
ideal adsorption theory (IAS) 834
IECC see International Energy
Conservation Code
illumination systems 195
IMAGE project 215
indoor air quality (IAQ) 57, 64, 75, 160
infiltration 111, 1801, 183, 198
inflatable heliostats 43
infrared thermography 1968
infrared thermometers 190
initial concentrations 778
injection strategies 198200
inlet velocity 1712
insulation 78
integrated directional coefficients method
229
internal environment 14950
International Energy Conservation Code (IECC)
22
ISO 14505 203
ISO 15099 21821, 224, 2356
isothermal supply 169, 172, 174
Jeffersonian grids 1034
jet length 171
Knudsen diffusion 68, 734, 78
Kuhns model 2278
LabView 202
Lambertian surfaces 227, 232
landscaping 1078
Langmuir equation 824
Langmuir-Hinshelwood model 88
Langmuirs theory of adsorption 59, 612
large eddy simulation (LES) 106
laser-cut light-deflecting panels (LCPs) 401
laser Doppler anemometers (LDAs) 191, 193
LCPs see laser-cut light-deflecting panels
LDAs see laser Doppler anemometers
Leadership in Energy and Environmental
Design (LEED) 278
LEED see Leadership in Energy and
Environmental Design
LES see large eddy simulation
light pipes 479
linear adsorption isotherm models 60, 68, 74,
778, 81, 83, 88
liquid fibres 44
lithium bromide/water absorption heat
pumps 130, 138
living standards 125
local discomfort meters 188
local equilibrium sorption models 88
local urban cooling demand 126
low-sloped roofs 1516, 1819, 22, 289
luminaires 467
macroscopic models 59
maintenance costs 142
mandatory energy efficiency standards 56
manikins 1847, 2034, 206
market level
mass roof assemblies 26
material contamination interactions 5793
mean age of air 176, 178, 2001
mechanical ventilation 160, 162, 16970
medium pairs 139
mercury thermometers 188
metallic sensors 1889
meteorological models 97
microclimates
human temperature regulation 1635
urban 95121, 125
Microlouvers 39
mixed convection 65
mixing ventilation 177
mixtures of contaminants 824
modelling
material and gaseous contamination 5793
thermal simulation 53
see also simulation
molecular diffusion 68, 734, 78, 80
momentum flux 16970
Monte Carlo-based techniques 232
multi-component sorption 823
natural convection 65
natural cooling 125
natural ventilation 160, 1623, 182
neighbourhood typology 1067, 114
nodal models 634
noise 1434
non-contact temperature measurements 190
non-isothermal supply 169
non-porous materials 68, 88
non-residential buildings 24, 89, 1519
non-white roofing materials 5
occupancy trends 125
Okasolar-W system 38
one-phase models 60, 623, 6971
one-sink model 59
open systems 129
operation costs 142
optical properties of shading devices
21139
optimization 989
orientation of streets 1034, 114
orifice flow meters 201
outdoor environment 2, 1804
parabolas 379
Parasol version 3 227
parks, urban 10711
particle image velocimetry (PIV)
1946, 206
particle streak velocimetry (PSV)
1946, 206
partition coefficients 678, 75, 81
partition diffusion 79
part-load characteristics 1401
passive tracer gas methods 199
paving materials 11213, 115
payback period 143
peak-power demand 2, 126
perfect diffusers 221, 230, 236
performance
buildings and microclimates 98
cool roofs 6, 8, 12, 26
COP 1301, 135, 138, 144
light pipes 47
SAC systems 144
sorption heat pumps 1356
photogoniometers 52, 2346
photometric measurement 52
physically based modeling 5793
Pitot tube 201
PIV see particle image velocimetry
Planck law 190
planning, urban 96115
plastic fibres 434
platinum resistive sensors 189
PMV see predicted mean vote
point measurement techniques 188
policy level 146
pollution 106, 182
polymeric materials 78, 802, 856
porous materials
contaminant interactions 623, 78, 802,
84, 869
diffusion 601, 6877, 80
Porters Five Forces model 147
predicted mean vote (PMV) 165
predictive approaches 114
prescriptive aspects 56, 89, 1218
prices
air conditioners 126
SAC systems 146
primary airflow 168, 171
primary energy 1267, 145
prismatic systems 3940
Prismex 467
prism light 47, 53
profile angles 221, 229
prototype daylighting systems 50
prototype inflatable heliostats 43
pseudo absorption constant 62, 66
PSV see particle streak velocimetry
pulse method 1934
purging flow rate 181
Ra see Rayleigh number
RADIANCE 232, 235
radiant barriers 1821
radiant exchange 1045
radiation 190
see also solar radiation
radiative loss measurement 1878
radiative properties 79
radiosity theory 227
Rayleigh number (Ra) 64
ray-tracing techniques 53, 232, 236
Re see Reynolds number
rebate programmes 28
recursive equations 214, 225
reflectance
bidirectional 234
diffuse 228
direct-to-diffuse 21922, 236
INDEX 245
direct-to-direct 21922, 225, 236
ground-diffuse 2234
roller shades 230
screens 231
shading devices 21516
solar 12, 79, 12, 1516, 249
venetian blinds 21920
reflecting blinds 35
refraction 3941
refrigerants 126, 12831
see also cooling
regional urban cooling demand 126
regulation of human temperature 1635
regulations 125
relative humidity 68
reliability 143
replicas 2045
residential buildings 5, 89, 1316,
1822, 24
resistive temperature sensors 1889
Retrolux blind 37
Reynolds number (Re) 64, 169, 172
roller shades 22930, 236
roof gardens 111
roofs 132, 11213
room velocity scale 1745
room ventilation 16878
roughness height 182
SAC see solar-assisted cooling
safety 144
salt bath method 179, 205
Sc see Schmidt number
scale models 2046
Schmidt number (Sc) 64
screens 1978, 2312
secondary airflow 168
seeding 193, 195
sensors 43, 18792
Sh see Sherwood number
shades 233
shading 349, 99, 11011, 21139
Sherwood number (Sh) 64
silica gel/water adsorption heat
pumps 131, 138
simulation
cool roofs 8, 1012
daylighting 513
manikins 1847
see also modelling
sink effect 58, 612, 66, 7980
site-specific weather data 115
size of cooling system 139
size of room 1714
skylights 412
slat properties 21829
solar-assisted cooling (SAC) 12357
solar lighting systems 456
solar radiation
access 99
cool roofs reflectance 12, 79, 12, 1516,
249
exposure 99, 1014
transmittance 216
solar reflectance index (SRI) 12, 279
Solux System 46
sorbent/refrigerants 1301, 1389
sorption
chemical transformations 87
coefficients 6972
cooling technologies 12737
correlations 79
diffusion coefficients 76
elemental models 5960
local equilibrium 88
multi-component 82
parameters 668
processes 612
see also absorption; adsorption
source/sink behaviour 58
Spanish grids 1034
spectral aspects 2301
specular aspects 225, 2278, 235
SRI see solar reflectance index
ST see entropytemperature diagrams
stairwells 17980
standards
comfort zones 203
cool-roof 132
steep-sloped roofs 1621, 289
stoichiometric coefficient 87
streets 97, 1036, 114
strengths, weaknesses, opportunities,
threats (SWOT) analysis 14754
structural shading 99
subsidiarity 99
summer point multipliers 23
supply velocity 16970, 1724
surface chemistry models 878
surface diffusion 735, 80
SWOT see strengths, weaknesses,
opportunities, threats analysis
system identification methods 200
system view 1613
TALISYS 53
technology
alternative cooling 126
environmental effects 14850
temperate climates 102
temperature
anemometers 1912
building and paving materials 11213
contaminant modeling 7981
diffusivities 814
ET 1667
heat pumps 1306
measurement 18890, 1949
microclimates and human regulation
1635
street aspect ratio 1045
thermal manikins 185
vegetation 10711
test people 205
test rooms 2046
texture analysis 107
thermal anemometers 1913
thermal aspects
building and paving materials 11213
comfort 125, 1646
cool roofs 12, 79, 1516, 249
thermal-driven active solar cooling
technologies 12737
thermal manikins 1847, 206
thermal transmittance 910, 14
see also transmittance; U factor
thermistors 1889, 192
thermochemical heat pumps 1312
thermocouples 18990, 202
thermometers 188
threats, opportunities, weaknesses,
strengths (TOWS) matrix 147, 1514
time constants 175
time varying flow rate 200
Title 24 standards, California 6, 1421
TOWS see threats, opportunities,
weaknesses, strengths matrix
tracer gas methods 181, 1989, 2067
see also constant concentration method
tracer seeding system 195
tracking mechanisms 434, 47
transmittance
bidirectional 234
diffuse 235
direct-to-diffuse 21922, 236
direct-to-direct 21922, 225, 236
direct-to-hemispherical 224
roller shades 230
screens 231
venetian blinds 21929
see also thermal transmittance
trees 10910
TRNSYS software 215
tropical climates 102
turnkey costs 1412
twin chamber methods 75
two-phase models 603, 72
typology of buildings 1067, 114
U-factor 810, 14
see also thermal transmittance
ultrasonic anemometers 1934
United States of America (US) 132
urban areas
climate research 95121
cooling demand 1246
density 978, 1012, 105
SAC systems 155
US see United States of America
Vant Hoff equation 81
vapour compression cooling (VCC) 124, 126,
1415
VCC see vapour compression cooling
vegetation 10711, 11415
velocity
deposition 867
measuring 188, 1919, 206
ventilation 1715
venetian blinds 356, 213, 21629, 232, 236
ventilation 103, 159210
video approaches 2345
virtual manikins 204, 206
visible transmittance 216
visualization 184, 192
VOCs see volatile organic compounds
volatile organic compounds (VOCs) 61, 66,
77, 7980
volume flow 201
volumetric expansion coefficient 65
voluntary approaches 268
INDEX 247
water/ammonia absorption heat
pumps 130, 138
weather data 97, 115
whitewashing 112
whole-field measuring techniques 1949
wind 111, 1823
windows 21415
see also fenestration systems
WINDOW software 215
wire sensors 189, 1912
WIS see Advanced Window Information
System
zeolite/water adsorption heat
pumps 131
zonal models 634

Advances in Building Energy

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Advances in Building Energy

Uploaded by

Copyright:

Available Formats

Advances i n Bui l di ng Energy Research

ABER_prelims 6/12/08 1:52 PM Page i

Acceptance angle for diffuse radiation

Source: original material for this chapter

A Review of Innovative Daylighting Systems 47

at abscissa x = 0, which implicitly means that h is infinite and

(m/s) the air

66 P. BLONDEAU, A. L. TIFFONNET, F. ALLARD AND F. HAGHIGHAT

Modelling of Contaminant Interactions in Buildings 73

78 P. BLONDEAU, A. L. TIFFONNET, F. ALLARD AND F. HAGHIGHAT

Experimental Methods in Ventilation 173

178 M. SANDBERG, H. LUNDSTRM, H. O. NILSSON AND H. STYMNE

Experimental Methods in Ventilation 181

Experimental Methods in Ventilation 205

224 ATHANASSIOS TZEMPELIKOS

cos cos ( ) cos ( ) sin ( )

You might also like