Building and Environment 40 (2005) 15051511

Performance evaluation of green roof and shading for thermal

protection of buildings
Rakesh Kumar

, S.C. Kaushik
Centre for Energy Studies, Indian Institute of Technology, New Delhi-110016, India
Received 29 July 2003; received in revised form 5 July 2004; accepted 22 November 2004
Abstract
The present paper describes a mathematical model for evaluating cooling potential of green roof and solar thermal shading in
buildings. A control volume approach based on nite difference methods is used to analyze the components of green roof, viz. green
canopy, soil and support layer. Further, these individual decoupled models are integrated using Newtons iterative algorithm until
the convergence for continuity of interface state variables is achieved. The green roof model is incorporated in the building
simulation code using fast Fourier transform (FFT) techniques in Matlab. The model is validated against the experimental data
from a similar green roof-top garden in Yamuna Nagar (India), and is then used to predict variations in canopy air temperature,
entering heat ux through roof and indoor air temperature. The model is found to be very accurate in predicting green canopy-air
temperature and indoor-air temperature variations (error range 73.3%, 76.1%, respectively). These results are further used to
study thermal performance of green roof combined with solar shading. Cooling potential of green roof is found adequate (3.02 kWh
per day for LAI of 4.5) to maintain an average room air temperature of 25.7 1C. The present model can be easily coupled to different
greenhouse and building simulation codes.
r 2005 Published by Elsevier Ltd.
Keywords: Foliage; Canopy; Green rooftop; Leaf area index; Shading
1. Introduction
Enormous use of ground for various purposes has lead
to disappearance of green planted surfaces. In order to
prevent dangerous and uncomfortable urban heat island
effects the indispensable need of planted surfaces is quiet
inevitable as is conrmed by many researchers viz. [15].
Space constraints have further reduced the applicability of
green surfaces in various areas surrounding the building
envelope. Consequently, planted roofs become the only
promising and stabilizing choice in the present scenario.
Good thermal protection can greatly reduce the high
thermal loads that badly affect the comfort conditioning
of building during summers. Eumorfopoulou and
Aravantinos [2] reported that planted roofs contribute
not only in reducing the thermal loads on the buildings
shell but also in reducing urban heat island effects in
densely built areas having a little natural environment.
Akbari et al. [6] have described the cooling energy
potential of shade trees by reduction of the local
ambient temperature. For their biological functions
such as photosynthesis, respiration, transpiration and
evaporation, the foliage materials absorb a signicant
proportion of the solar radiation. Thermal protection
techniques of green roof can provide a great degree of
reduction in the local air temperature near canopy, thus
reducing the incoming heat ux into the building. A
study done by Onmura [4,7] revealed that in closed
spaces with the planted roofs, the air temperature
beneath the plants is lower than that of the air above,
by nearly 451C. Thermal performance of green roof
ARTICLE IN PRESS
www.elsevier.com/locate/buildenv
0360-1323/$ - see front matter r 2005 Published by Elsevier Ltd.
doi:10.1016/j.buildenv.2004.11.015

Corresponding author. Tel.: +91 11 2659 1253;

fax: +91 11 2659 1266.
E-mail addresses: krakesh1999@hotmail.com (R. Kumar),
sckaushik@ces.ernet.iitd.in (S.C. Kaushik).
have been studied in great detail recently by several
researchers. It is worth mentioning the works on
prediction of thermal performance of green roofs by
Del Barrio [3] and Good [8]; and on implementation of
green roof in the buildings by Dominguez and Lozano
[9]; Eumorfopoulou and Aravantinos [2] and Takakura
et al. [10]. Capelli et al. [11] predicted thermal behavior
and effectiveness of vegetation covers with different
average absorptance for solar radiation and diffusive
properties, which shields roof-covering structures of
different masses. In a study conducted by Niachou et al.
[12], investigation of green roof is done in two phases; in
the rst phase, extensive data measurements for
temperature, both indoor and outdoor are considered,
and in the second phase, thermal properties of green
roof are studied using a mathematical approach.
Hoyano [13,14] conducted an experimental study on
effect of rooftop lawn planting on thermal environment
and also described for climatological uses of plants
for solar control and the efforts on the thermal
environment.
Many of the studies predict thermal performance of
green roof localized to experimental site or employ
several numerical techniques to evaluate thermal per-
formance. This restricts the applicability of the green
roof to particular buildings and hence thermal space
conditioning of different building cannot be predicted,
as green roof model is to be coupled to the building
simulation code. Effect of parametric variations in
thermal components of green roof on cooling potential
is also not described. This model improves upon these
aspects by incorporating thermal modeling of green roof
components, parametric variations in the green roof
components and coupling the model to the building
simulation code. The process of heat transfer into the
planted roof is very different from bare roof, both
qualitatively and quantitatively. The analysis of green
roof can be classied into three sub-regions, viz. green
canopy, soil, and the roof support. Dynamic perfor-
mance of each of the sub-region can be evaluated and
further coupled with each other using the boundary
conditions.
The objective of studying green roof is multifold: to
determine the effect of variations in foliage character-
istics, viz. leaf area index (LAI) and foliage height
thickness on thermal performance of green canopy,
estimation of thermal load reduction in the building and
evaluation with thermal shading on building space
conditioning. Results obtained from the simulation of
green roof model are validated with experimental data
from an existing green roof with similar foliage
characteristics and building parameters in Yamuna
Nagar, Haryana state of India.
2. Mathematical formulation
The green roof top modeling described is based on
Del Barrio [3]. Fig. 1 shows the model of green roof
ARTICLE IN PRESS
Nomenclature
d average leaf thickness
D
VT
vapor diffusivity
h
aN
heat transfer coefcient to air
h
g
convective coefcient of vapor transport
H
c
height of green canopy
H
s
height of soil layer
H
R
thickness of roof support
i nodal points
L canopy layer thickness
p
l
leaf vapor pressure
p
a
air vapor pressure
p
g
vapor pressure at soil surface
r
e
mean canopy resistance to sensible heat
transfer
T
c
temperature of canopy
T
a
temperature of air
T
sky
sky temperature
T
g
ground temperature
T
so
soil Temperature
T
N
air temperature
Greek letters
t
s
shortwave transmittance
t
t
transmittance of the leaf tissue
y
a
air specic humidity
c
s
solar radiation
r density
l
su
thermal conductivity of support material
Fig. 1. Schematic of green roof model.
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1506
constituting three elements of green canopy, soil and the
roof support. A very important parameter for green
canopy is denoted by LAI. It is the ratio of area of
leaves to the area of the base occupied. Fig. 2 shows the
average LAI for soyabean crop.
The governing equation of energy balance for the
green canopy model is
rC
p
d LAI
dT
c
dt
1 t
s
1 t
s
r
1
1 t
s
r
g
c
s
1 t
l
sT
4
sky
sT
4
g
2sT
4
c

2LAI
rC
p
r
e
T
c
T
a

2LAI
rC
p
gr
e
r
i

p
c
p
a
; 1
where p
a
is the vapor pressure of the canopy air, p
c
is the
vapor pressure at the soil surface and p
N
is the outdoor
air vapor pressure. Here, the rst term denotes net solar
radiation absorbed by the canopy. t
s
r
g
c
s
represents net
solar radiation reected by the ground. The second term
denotes net thermal radiation in long wave as absorbed
by the canopy. T
g
is the ground temperature, T
sky
is the
sky temperature, r
e
indicates the mean canopy
resistance. The third term signies net convective
heat transfer from green canopy to air. Finally, the
fourth term denotes net transpiration losses from plant
to air.
Canopy air temperature can be determined from the
following equation:
rc
a
L
dT
a
dt
2LAI
rC
p
r
e
T
c
T
a

h
g
T
g
T
a
h
a1
T
1
T
a
; 2
where h
g
is the convective coefcient of vapor transport
and h
aN
is the heat transfer coefcient between canopy
air and outdoor air, r
e
is the mean canopy resistance to
the sensible heat transfer, and T
N
is the outdoor air
temperature.
Energy balance for air specic humidity is given by
raL
qy
a
qt
h
g
p
g
p
a
h
a1
p
a
p
1
: (3)
Soil model is governed by the following:
rC
p
x; t
qT
so
y; t
qt

q
qy
l LD
vT
o; T
so

qT
so
y; t
qy

; 4
where l is the soil thermal conductivity, x is the local
volumetric moisture content in the porous medium
domain, L is the latent heat of vaporization and D
nT
is
the non-isothermal vapor diffusivity.
This equation is reduced into algebraic form using
nite difference method. Backward differencing in space
and time has been used. The roof support is considered
as a homogenous material. The overall one-dimensional
heat conduction equation is taken as
rC
p
qT
su
y; t
qt
l
su
q
2
T
su
y; t
qt
2
; (5)
where T
su
y; trepresents the temperature eld, r is the
density, and l
su
is the thermal conductivity of the
support material.
Finite volume method [15] is used to discretize the
governing equation in to algebraic equation. Backward
differencing and explicit time-marching scheme is used.
Mixed or Robin boundary condition is applied on the
bottom of the roof support layer, whereas, specied
temperature or Dirichlet boundary condition is applied
at the top of the roof support layer to preserve
continuity.
The three decoupled models are then coupled by the
interfaces where the relations between the connected
models are described. Newton iteration technique is
employed to determine the interface state variables that
satisfy the constraints of ux continuity, together with
continuity of state variables.
The room temperature determined from building
simulation code is governed by one-dimensional heat
conduction equation. The resultant heat ux from green
roof model is incorporated in the building simulation
code [16] in Matlab using FFT. Solar thermal shading is
incorporated in the building simulation code by
ascertaining the fraction of solar radiation received by
various building components based on building loca-
tion, orientation and shaded portions.
A room size of 6 5 4 with occupancy of 4 people is
considered for modeling. An air change rate of 5 h
1
is
taken for a window area of 2 m
2
. The predicted results of
green roof model coupled to building simulation code
are obtained for parametric values shown in Table 1.
However, the model is adaptive to include any value
of the parameter specied by the user as per building
specication and selected green roof parameters.
ARTICLE IN PRESS
Fig. 2. Average calculation of LAI for foliage.
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1507
3. Results and discussions
The parametric values used in this research for
various constants are:
Soil thermal diffusivity 2.2e(4) m
2
/s
Vapor diffusivity 2.e( 6) m
2
/s
Latent heat of evaporation 2257 kJ/kg
The accuracy of the presented model is veried on the
basis of a comprehensive set of experimental data. Five
temperature and humidity sensors (8160TFF10) sup-
plied by Luftopus-200/200I were placed in each room at
the corners and the center of green canopy layer exactly
at the central height of the foliage (0.20, 0.40, 0.60 and
1.0 m above the soil surface). Five platinum resistance
thermometers (PT-500) were used to measure the soil
temperature at a height of 0.10, and 0.20 m above the
roof surface. Three temperature humidity sensors were
placed at 0.50, 1.5 and 2.5 m above the oor in side the
room. Fig. 3 shows the photo of the experimental site
with marked circles to indicate the testing points.
The performed simulations extend over validation
and parametric study of green roof with cooling energy
potential. Fig. 4 shows the roof top garden to which the
modeling simulation results are validated. Validation of
simulated results with experimental observations is
shown in Fig. 5 for canopy air temperature. And it is
found that the results matched experimental data within
an accuracy of 73.3%. Further, validation of indoor air
temperature obtained from building simulation code
(after incorporating green roof model) is shown in Fig.
6. Results matched within an accuracy of 76.1% with
experimental data obtained from 8 check points to
measure the room temperature.
Effect on indoor air temperature is also shown in Fig.
6. It is found that the indoor air temperature is reduced
by an average of 7.2 1C. However, the reduction of
temperature follows a pattern with a maximum reduc-
tion observed in peak heating period (12:0015:00 h) and
a minimum in off sunshine period.
LAI is one of the important parameter affecting the
micro-climate of the green canopy and hence the
interiors of the building. Fig. 7 shows the effect of
parametric variations in LAI on canopy air temperature.
A peak reduction of 9.3 1C in canopy air temperature is
observed for a cycle of 8 days from June 18. Peak
canopy air temperature and temperature width are both
reduced with increased in LAI as summarized in Fig. 7.
Hence, it is observed from the gure that uctuations of
the order of 11.6 1C are reduced to only 3.6 1C with an
increase in LAI from 0.5 to 3.5.
ARTICLE IN PRESS
Table 1
Planted roof parameters and properties
Element Effective conductivity (l
w
) Thickness Specic heat (J/kg K)
Green canopy Varying (on foliage characteristics) 0.4 m Variable
Soil 1.21 W/mK 840
Sieve layer 0.4 W/mK 0.002 1340
Wash off layer 0.035 W/mK 0.03 Varying
Reinforced Concrete 1.58 W/mK 0.006 880
Roof plaster 0.72 W/m-K 0.018 820
Fig. 3. Schematic of experimental site.
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1508
The long-term dynamic performance of green roof for
shielding heat radiations is presented with parametric
variations in foliage thickness and LAI in Figs. 8 and 9.
Heating ux entering green roof with a foliage height of
0.6 m is 1.94 Wh m
2
(6984 J) and this increased
further by nearly 4 times (14,400 J) Whm
2
for the bare
roof as seen from Fig. 8. Also, the uctuations in the
entering heat ux are very less for green roof as
compared to bare roof.
It is important to determine the heat ux penetrating
the building affected by green roof for various leaf area
indices denoting the extent of coverage of the roof.
Internal gains for the room as shown in Fig. 9 are an
average 3.02 kWh (10872 kJ) (5.18 kWh for bare roof &
2.16 kWh for LAI of 4.5). The results show an increase
in the entering heat ux with decreased in LAI. Increase
in LAI in general leads to a reduction in coverage area
and hence the insulating effect. However, it also relates
to lessening of biological activities by foliage materials,
affecting the micro-climate of the building.
Finally, building space conditioning is shown in Fig.
10. Average indoor air temperature of a (6 5 4)
room is 28.47 1C without green roof and this is reduced
by 3.3 1C, leading to an average indoor air temperature
of 25.7 1C. For shading combined with green roof,
temperature reduces by 2.1 1C leading to an average
indoor air temperature of 23.6 1C.
Diurnal uctuations in indoor air temperature
show reduced temperature width with green roof.
Hence, it is seen from the gure that the temperature
width of 10.2 1C in indoor temperature without
green roof is reduced to only 5.1 1C with green roof
ARTICLE IN PRESS
24
26
27
28
29
30
31
48
Time (Hrs)
C
a
n
o
p
y

a
i
r

t
e
m
p
e
r
a
t
u
r
e

(

C
)
72 96 120
Predicted Canopy air temperature data
10 - 14 June, 2001
LAI: 1.5
Experimental measurements data
Fig. 5. Validation for green canopy air temperature.
Fig. 4. Green roof top garden (picture view from side).
100 200 300 400 500 600 700
20
25
30
35
40
Ambient air temperature
Station : Yamuna Nagar, Haryana
Time : 1May to 30 May, 2001
Indoor air temperature (Theoretical)
Measurements
I
n
d
o
o
r

a
i
r

t
e
m
p
e
r
a
t
u
r
e


(

C
)
Time (Hrs)
Fig. 6. Prediction and validation of indoor air temperatures for
monthly cycle.
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1509
which has further dropped to only 2.1 1C while shading
is included.
4. Conclusions
1. Thermal performance of green roof model is pre-
sented with extensive set of validation with experi-
mental data for canopy air temperature and indoor
air temperature.
2. Parametric variations in LAI and foliage height
thickness are carried out to determine the modulation
of canopy air temperature, and reduction in tem-
perature width and to estimate the penetrating heat
ux. Results determine that larger LAI reduces the
canopy air temperature, stabilized the uctuating
values and reduced the penetrating ux by nearly
4 W/m
2
.
3. Indoor air conditioning of the building is predicted.
Green roof combined with solar thermal
shading reduced averaged indoor air temperature by
5.1 1C, from the average indoor air temperature for
the bare roof. Green roof provided a cooling
potential of 3.02 kWh per day which is found to be
adequate.
Earth covering methods for thermal protection like
green roofs and solar thermal shading provide a very
effective solution in building habituation for present
state of affairs being very energy efcient and eco
responsive. Results of this investigation are applicable to
the entire building geometrical realm.
ARTICLE IN PRESS
Time (Hrs)
C
a
n
o
p
y
a
i
r
t
e
m
p
e
r
a
t
u
r
e
(
0
C
)
24 48 72 96 120 144 168 192
20
25
30
35
1-8 June, 2001
40
Ambient air temperature
LAI =1.5
LAI = 2.5
LAI = 3.5
Fig. 7. Effect of various LAI on canopy air temperature.
50 100 150
7
6.5
6
5.5
5
4.5
4
3.5
3
2.5
2
Time (Hrs)
H
e
a
t

f
l
u
x

(
W

/

m
2
)
bare roof
Time : 1 to 8 June, 2001
Roof Area : 30 Sq.m.
fht = 0.6
fht = 0.5
fht = 0.4
fht = 0.3
fht = 0.2
Fig. 8. Heat ux through roof for various foliage height thickness.
Time (Hrs)
H
e
a
t
f
l
u
x
(
W
/
m
2
)
0 24 48 72 96 120 144 168 192
-2
0
2
4
6
Bare roof
LAI = 0.5
LAI = 2.5
LAI = 3.5
17 - 25 May, 2001
Yamuna Nagar
Fig. 9. Total heat ux through planted roof for various leaf area
indices.
50 100 150
Time (C)
I
n
d
o
o
r

a
i
r

t
e
m
p
e
r
a
t
u
r
e

(

C
)
20
22
24
26
28
30
32
34
36
38
40
Ambient air temperature
Station : Yamuna Nagar, Haryana
Time : 1 to 8 June, 2001
Indoor air temperature With bare roof
Green roof
Green roof + Shading
Fig. 10. Indoor air temperature variations (from building simulation
code).
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1510
Acknowledgements
The authors are thankful to Ministry of Non-
Conventional Energy Sources (MNES) New Delhi, for
its nancial help in carrying out this research work. We
are also thankful to Indian Meteorological Department
(IMD) New Delhi, for providing the Weather Data.
References
[1] Wong NH, Chen Y, Ong CL, Sia A. Investigation of thermal
benets of rooftop garden in the tropical environment. Building
and Environment 2003;38:26170.
[2] Eumorfopoulou E, Aravantinos D. The contribution of a planted
roof to the thermal protection of buildings in Greece. Energy and
Buildings 2003;27:2936.
[3] Del Barrio EP. Analysis of the green roofs cooling potential in
buildings. Energy and Buildings 1998;27:17993.
[4] Onmura S, Matsumoto M, Hokoi S. A study on evaporative
cooling effect by roof lawn garden. Proceedings of the European
conference on energy performance and indoor climate in buildings
1994:6349.
[5] Haefeli P, Lachal B, Weber W. Experiences with green roofs in
Switzerland. Proceedings of the PLEA98, Portugal 1998:3658.
[6] Akbari H, Kurn DM, Bretz SE, Hanford JW. Peak power and
cooling energy savings of shade trees. Energy and Buildings
1997;25:13948.
[7] Onmura S, Matsumoto M, Hokoi S. Study on evaporative
cooling effect of roof lawn gardens. Energy and Buildings
2001;33:65366.
[8] Good W. Factors in planted roof design. Constructions Specica-
tion 1990;43(11):132.
[9] Dominguez J, Lozano A. Green roof systems. Comput. Mech.
Publication 1998;1:61524.
[10] Takakura T, Kitade S, Goto E. Cooling effect of greenery cover
over a building. Energy and Buildings 2000;31:16.
[11] Cappelli M, Cianfrini C, Corcicone M. Effects of vegetation
roof on indoor temperatures. Heat Environment 1998;16(2):
8590.
[12] Niachou A. Analysis of the green roof thermal properties and
investigation of its energy performance. Energy and Buildings
2001;33:71929.
[13] Hoyano A, He J, Horiguchi T, Wang G. Experimental study on
heat budget for foliage layer on lawn planting. Part 1. Effect of
rooftop lawn planting on thermal environment. Journal of
architecture Planning and Environmental Engineering, Transac-
tions of AIJ 1994;462:319.
[14] Hoyano A. Climatological uses of plants for solar control and the
efforts on the thermal environmental of a building. Energy and
Buildings 1988;11:2936.
[15] Patankar SV. Numerical heat transfer and uid ow. New York:
Hemisphere Publishing Corporation; 1980.
[16] Kaushik SC, Sodha MS, Bansal PK, Bhardwaj SC. Solar thermal
modeling of a non-air conditioned building: evaluation of overall
heat ux. Energy Research 1982;6:14360.
ARTICLE IN PRESS
R. Kumar, S.C. Kaushik / Building and Environment 40 (2005) 15051511 1511