Building and Environment: Irina Susorova, Melissa Angulo, Payam Bahrami, Brent Stephens
Building and Environment: Irina Susorova, Melissa Angulo, Payam Bahrami, Brent Stephens
Building and Environment: Irina Susorova, Melissa Angulo, Payam Bahrami, Brent Stephens
a r t i c l e i n f o a b s t r a c t
Article history: A mathematical model of an exterior wall covered with climbing vegetation has been developed to
Received 27 February 2013 evaluate the thermal effects of plants on heat transfer through building facades. This model allows for
Received in revised form analysis of how various plant physiological parameters such as leaf area index, average leaf dimension,
15 April 2013
and leaf absorptivity can improve facade thermal performance by reducing the exterior wall surface
Accepted 23 April 2013
temperatures and heat flux through the façade. The model has been verified with a set of experiments
that measured both bare and vegetated facade thermal performance of an educational building in Chi-
Keywords:
cago, IL, during the summer. A sensitivity analysis was also conducted to elucidate the relative impacts of
Green facades
Facade vegetation
plant characteristics, weather conditions, climate zones, wall assembly types, and facade orientation on
Facade thermal performance vegetated facade thermal performance. Overall, results herein show that a plant layer added to the facade
Vegetation models can improve its effective thermal resistance by 0.0e0.7 m2 K/W, depending on a range of inputs for wall
Energy efficient parameters, climate zones, and plant characteristics (particularly leaf area index). These improvements
Buildings are especially pronounced in predominantly warm climates with high solar radiation and, to a lesser
extent, low wind speeds. The model developed herein can ultimately be used both to access facade
thermal improvements in existing buildings retrofitted with green walls and to design green walls for
optimal energy efficiency in new construction.
Ó 2013 Elsevier Ltd. All rights reserved.
0360-1323/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.buildenv.2013.04.027
2 I. Susorova et al. / Building and Environment 67 (2013) 1e13
Nomenclature Rplant effective thermal resistance of the plant layer (m2 K/W)
RH relative humidity
cpair specific heat of air at constant pressure (29.3 J/mol C) Sbw, Svw heat stored in the wall material of the bare and
cpwall specific heat of wall material (J/kg K) vegetated facades (W/m2)
Cbw, Cvw convection to/from the bare facade and vegetated SRbw, SRvw shortwave radiation to/from the bare facade and
facade (W/m2) vegetated facade (W/m2)
D leaf characteristic dimension (m) t time (s)
ea partial water vapor pressure of air (kPa) Tair air temperature (K)
es(T) water vapor pressure of air at saturation (kPa) Tdewpoint dewpoint temperature (K)
Fgr, Fsky view factors to the ground and to the sky Tgr ground temperature (K)
gasll, gasul actual stomatal conductance of the lower and upper Tin air temperature of the room interior (K)
leaf surface (mol/m2 s) Tinbw, Tinvw measured interior surface temperature of the bare
gbh boundary layer conductance for heat transfer and vegetated facade (K)
(mol/m2 s) Tleaf temperature of the leaf surface (K)
gbv boundary layer conductance for vapor (mol/m2 s) Tsky clear sky temperature (K)
gc convective conductance (mol/m2 s) Tsbw, Tsvw exterior surface temperature of the bare and vegetated
gr radiative conductance (mol/m2 s) facade (K)
gsll, gsul typical stomatal conductance of the lower and upper Vair wind speed (m/s)
leaf surface (mol/m2 s) XR radiative exchange between the bare facade and the
gv vapor conductance (mol/m2 s) plant layer (W/m2)
hbw, hvw convection heat transfer coefficient of the bare and awall facade absorptivity for solar radiation
vegetated facade surface (W/m2 K) aleaf leaf absorptivity for solar radiation
Imax maximum total solar radiation incident on a surface g thermodynamic phsycometer constant
(W/m2) (0.000666 1/ C)
It total solar radiation incident on a surface (W/m2) g0 apparent phsycometer constant (1/ C)
k material conductivity (W/m K) D slope of the saturation vapor pressure function
L wall thickness (m) (kPa/ C)
LAI leaf area index εleaf leaf emissivity of longwave radiation
LRbw, LRvw longwave radiation to/from the bare facade and εwall facade surface emissivity of longwave radiation
vegetated facade (W/m2) εsky sky emissivity (approximately 1)
Pair atmospheric pressure at sea level (100 kPa) hroot minimum value of soil moisture in the root zone of the
Qbw, Qvw total heat flux through the wall with the bare facade plant
and vegetated facade (W/m2) hwilt level of soil moisture below which permanent wilting
Qleaf the total radiation absorbed by a vertical layer of plants of the plant occurs
(W/m2) k attenuation coefficient
rasll, rasul actual stomatal resistance of the lower and upper leaf r wall material density (kg/m3)
surface (m2 s/mol) l latent heat of vaporization of water (kJ/kg) (2501 kJ/kg
rsll, rsul typical stomatal resistance of the lower and upper leaf at 0 C)
surface (m2 s/mol) s StefaneBoltzmann constant (0.00000005673)
Rbw, Rvw effective thermal resistance of the bare facade and s transmissivity of total radiation through a plant layer
vegetated facade (m2 K/W) q tilt angle of the facade surface
time period. The purpose of this study is to develop a compre- the thermal resistance analogy to evaluate the effect of facade
hensive mathematical model of a vegetated wall (without soil) that orientations, leaf density, and insulation layer placement on facade
can be used to evaluate the effects of climbing plants on the ther- thermal performance and cooling load reductions. These previous
mal performance of a building facade for various input parameters, studies have relied primarily on plant characteristics measured
including weather conditions, climate zones, facade orientation, during individual experiments as model inputs as opposed to
wall assembly types, and, importantly, a variety of important plant computing them directly, which is a focus of this work.
characteristics. Mathematical models simulating the effect of plants in green
roofs were developed by Alexanri [9] and Sailor [10], whose model
2. Model development was also implemented in the EnergyPlus energy analysis software
[11]. Both dynamic mathematical models simulated one-
Several previous studies have constructed heat transfer models dimensional vertical heat flux through a green roof assembly,
for thermal analysis of vegetated facades. For example, Holm [7] which included a layer of soil. The heat flux through the horizontal
developed a model to predict thermal improvements in a facade plant layer was simulated using the calculations of vegetation
covered with plants for different orientations, climates, and wall canopy models commonly used in climatology. However, green roof
construction types. However, this model was based on a particular models cannot be directly used to evaluate thermal properties and
set of plant characteristics measured in laboratory tests and did not heat flux through vegetated walls because of the absence of the soil
account for plant evapotranspiration. Di and Wang [2] constructed layer and the difference in heat exchange processes.
a mathematical model of heat flux through vegetated facades based At present, despite multiple studies on vegetation thermal
on experimental investigations to calculate conductive heat trans- effects, there are no published models of vegetated walls capable
fer and energy use reductions. Kontoleon and Evmorfopoulou [8] of simulating the effect of plants on the facade thermal perfor-
built a simplified mathematical model of a vegetated wall using mance for variable plant characteristics, facade properties, building
I. Susorova et al. / Building and Environment 67 (2013) 1e13 3
orientations, and weather conditions. Therefore, the model plant layer affects the facade thermal performance only during
described in this work offers an improvement over previous models relatively warm weather when the indoor room temperature is
by computing plant physiological processes, including evapo- lower than the facade exterior surface temperature and heat flux is
transpiration and radiative and convective heat exchange between in the direction from outdoors to indoors. In cold weather, when
the plant layer, the facade, the surrounding environment, and the the direction of the heat flux is reversed from the indoors to out-
ground and using individual plant characteristics inputs (e.g., leaf doors the plant layer has little effect; the model assumes the value
absorptivity, typical leaf dimension, leaf area index, radiation of the plant thermal resistance to be zero in cold weather.
attenuation coefficient, and leaf stomatal conductance) and The vegetated facade model calculates the effects of plants on
weather data to simulate the impacts of vegetated walls on façade the reduction in the facade surface temperature and heat flux
thermal performance. The next section describes the model through the exterior wall. The model does not account for wind
development in more detail. reduction effects because information is lacking on convective heat
transfer coefficients at vertical vegetated facade surfaces.
2.1. Vegetation model description
2.2. Facade energy balance and heat flux calculations
The proposed vegetated facade model simulates the one-
dimensional heat flux through the depth of a vertical layer of Fig. 1 describes the 1-D surface energy balance used in the
vegetation. The model accounts for short-wave radiative trans- vegetated wall model.
mission through the plant layer, long-wave radiative exchange A building facade receives shortwave radiation SR from the sun
between the plant layer and the environment, convective heat (direct, scattered from the sky, and reflected from the ground and
transfer to and from vegetation, and evapotranspiration from atmosphere) and also exchanges longwave thermal radiation LR
leaves. The model represents a layer of vertical plants rooted in the between the ground, sky, and surrounding surfaces [11,19,20]. Ra-
ground, such as vines or flat vertical shrubs growing directly diation absorbed by the facade is either emitted back into the
against a building exterior wall without the use of soil or substrate. environment, transferred to or away from the wall by convection to
Several model assumptions summarized below were based on the outdoor air C, or transferred to the building interior by con-
common assumptions made in other studies modeling vegetation duction Q, where energy can also be stored in the wall material S. At
[9,10,13e17]: any moment in time, the energy balance of a bare wall (bw) is thus:
The model considers plants only during the growing season. SRbw þ LRbw þ Cbw ¼ Qbw þ Sbw (1)
Leaves are uniformly distributed and oriented in the vegeta-
tion layer. Individual leaf angles are not considered. The energy balance of the vegetated wall (vw) includes an
Plant parameters including leaf absorptivity, leaf dimension, additional term (XR) for the radiative exchange between the leaves
leaf area index, radiation attenuation coefficient, and plant of the plant layer and the facade surface, as shown in Eq. (2).
stomatal conductance are constant and do not change with
season. SRvw þ LRvw þ XR þ Cvw ¼ Qvw þ Svw (2)
For calculating emitted radiation from the plant surface, the
The developed model solves the facade energy balance equa-
temperature of a leaf is assumed to be the same as air
tions with respect to heat flux Q by calculating its components. The
temperature.
facade incident shortwave radiation SR, longwave radiation LR,
Heat flow through a layer of vegetation occurs only hori-
convective heat flux C, and heat flux to and from the bare and
zontally; vertical heat flux is not considered.
vegetated facades Q are shown in Eqs. (3)e(11).
External factors that may vary with height, such as wind
speed, are assumed to be constant for the low-rise facades SRbw ¼ It awall (3)
considered.
The level of soil moisture at plant roots is constant; no pre-
SRvw ¼ It awall s (4)
cipitation is considered.
Air beneath stomatal pores is saturated with water.
LRbw ¼ εwall εsky sFsky Tsky
4 4
Tsbw þ εwall εgr sFgr Tgr
4 4
Tsbw
Most plants can typically survive only within a narrow tem-
perature range between 0 C and 40 C [14,15,18]. Therefore, the (5)
where L is the wall thickness (m), cpwall is the specific heat of the 2.4. Convective heat transfer coefficient
wall material (J/kg K), r is the wall material density (kg/m3), and t is
the time (s). The convective heat transfer coefficient for the bare facade is
The energy balance equations of the bare and vegetated facades calculated with an equation suggested by the EnergyPlus Engi-
are solved for the exterior surface temperature at one-minute time neering Ref. [11]
steps in this work. The model advances forward in time computing
the current facade surface temperature based on the temperature h ¼ a þ bV þ cV 2 (17)
at the previous time step using a numerical bisection method. The
energy balance is calculated using the R statistical software package where h is the convective surface heat transfer coefficient (W/
[22]. The energy balances for the bare and vegetated walls use m2 K), V is the local wind speed (m/s), and a, b, and c are the ma-
multiple model input parameters that are either known or terial roughness coefficients. The model assumes that the facade is
computed. Weather parameters are taken from weather data files made of a material with a medium-rough surface for which the
for different geographic locations available from the U.S. Depart- material coefficients are 10.79, 4.192, and 0. Thus, the convective
ment of Energy [23]. Plant and facade parameters are taken from heat coefficient hbw for the bare facade is estimated as:
I. Susorova et al. / Building and Environment 67 (2013) 1e13 5
Fig. 2. Heat flux through the bare facade (left) and vegetated facade (right) expressed as an electrical circuit diagram.
where Vair is the air speed at the bare facade (m/s). Although some where gr is the radiative conductance obtained from tables by
studies showed that vegetation on walls can decrease wind speed Campbell [14] and gbh is the boundary layer conductance for heat
at vegetated wall surfaces [6,14], we are not aware of robust studies transfer through air calculated as
on decreased values of convective heat transfer coefficients for
wide ranges of vegetated facade types. In the absence of better data, rffiffiffiffiffiffiffiffi!
Vair
the present model assumes hbw ¼ hvw. However, this assumption gbh ¼ 1:4 0:135 (22)
D
should be explored in future studies.
A typical broad leaf is often represented as a flat thin plate that 0:5gasul gbv 0:5gasll gbv
has radiative and convective heat exchange with the surrounding gv ¼ þ (23)
gasul þ gbv gasll þ gbv
atmosphere and evaporates water from its surface (conductive heat
transfer is insignificant) [15]. Leaf surface temperature is a complex where gasll is the actual stomatal conductance of the lower leaf
function of multiple factors, such as solar radiation, air humidity, surface; gasul is the actual stomatal conductance of the upper leaf
wind speed, and internal leaf carbon dioxide concentration. The surface, and gbv is the boundary layer conductance for vapor
leaf surface temperature in the vegetation model is found using the calculated as
calculations by Campbell [14]
rffiffiffiffiffiffiffiffi!
" # Vair
g0 Qleaf es ðTÞ ea gbv ¼ 1:4 0:147 (24)
Tleaf ¼ Tair þ (19) D
D gc cpair Pair g0
þ g0
Pair The slope of the saturation vapor pressure function is given by
where Tair is the air temperature ( C); g0 is the apparent phsyc- 4217es ðTÞ
ometer constant (1/ C); D is the slope of the saturation vapor D ¼ (25)
ð240:97 þ Tair Þ2
pressure function (kPa/ C); Pair is the atmospheric pressure at sea
level (kPa); Qleaf is the total radiation absorbed by a vertical layer of The total radiation absorbed by a vertical layer of plants Qleaf is
plants (W/m2); gc is the heat conductance (mol/m2s); cpair is the calculated as
specific heat of air at constant pressure (J/mol C), es(T) is the water
vapor pressure of the air at saturation (kPa), and ea is the partial Qleaf ¼ It aleaf þ εleaf sF Tsky
4 4
þ Tgr εleaf sðTair Þ4 (26)
water vapor pressure of the air (kPa). The apparent phsycometer
constant is
where Tgr is the temperature of the ground which is assumed to be
ggc the same as the outside air temperature Tair. It should be noted that
g0 ¼ (20)
to calculate the radiation emitted by a vertical layer of plants, the air
gv
temperature Tair was used instead of the leaf temperature Tleaf, not
where g is the thermodynamic phsycometer constant and gv is the yet calculated.
vapor conductance through air. The heat conductance through air The water vapor pressure of the air at saturation and the partial
gc is water vapor pressure of the air are given by
6 I. Susorova et al. / Building and Environment 67 (2013) 1e13
Fig. 3. South facade of Siegel Hall on the campus of the Illinois Institute of Technology (left) and the instruments measuring two facade areas (right).
Fig. 4. Bare and vegetated facade temperatures measured on a sunny day (August 31, 2012) and a cloudy day (September 1, 2012).
At the same time, the vegetated facade exterior surface tempera- during times of direct sun exposure. At night and during the cloudy
ture typically closely matched the ambient air temperature and was day, the temperature gradient of the vegetated facade was actually
never higher than 36.8 C. The highest difference between the bare higher (mean 1.4 C for the bare and 1.8 C for the vegetated fa-
and vegetated exterior surface temperatures occurred around cades), presumably because its surface could not reject heat as
14:00 (5.7 C) and coincided with the highest values of solar radi- effectively as the bare surface. However, the absolute difference
ation measured. This confirms that the most important plant effect was often low (i.e., <1 C), meaning that conduction is not as
is shading walls from solar radiation. important during these periods.
The average bare facade exterior surface temperature was
29.6 C on the sunny and 24.2 C on the cloudy day and the average 3.3. Model validation
vegetated facade exterior surface temperature was 28.6 C on the
sunny and 24 C on the cloudy day. On the sunny day, the vegetated These experimental results were used to validate the wall
facade exterior surface temperature was consistently lower than model by comparing the predicted and measured facade tem-
that of the bare facade (mean difference of 1.1 C), but it was peratures. Knowledge of the surface temperature reductions be-
approximately the same at night (Fig. 4). On the cloudy day, these tween the bare and vegetated facades allowed for estimation of
two surface temperatures were closer throughout the day, but the other facade properties, such as a decrease in heat flux through
exterior surface temperature behind plants was lower on average the wall (assuming a constant R-value for the wall itself) and thus
(mean difference of 0.4 C) likely because plants absorbed diffuse the instantaneous effective R-value of the plant layer. Several
solar radiation passing through the clouds. assumed plant and wall parameters were used in conjunction
The average interior surface temperature of the bare facade was with the measured weather data (air temperature, solar radia-
24.9 C on the sunny day and 22.9 C on the cloudy day, while that tion, relative humidity, and wind speed at the facade) as inputs in
of the vegetated facade was 24.2 C on the sunny day and 22.1 C on the model. Due to equipment limitations, the experiment
the cloudy day. The interior surface temperature of the vegetated measured only the global horizontal solar radiation, which was
facade was always lower than that of the bare facade (mean dif- adjusted by modeling solar radiation on both a horizontal and a
ference of 0.9 C). The peak interior surface temperature difference vertical surface using solar geometry for Chicago [29]. Correla-
between the bare and vegetated facades occurred around 18:00, tions between the two surfaces were calculated and applied to
which did not correspond to the peak difference in the exterior the measured data to adjust the direct solar radiation data to a
surface temperatures. This can be explained by the thermal lag of vertical surface. The validation was performed with the data for a
the facade, which appears to be approximately 4 h for the 0.2 m sunny day (August 31, 2012) and a cloudy day (September 1,
thick masonry wall. Although not a focus of this study, it should be 2012).
noted that this thermal lag effect can be beneficial for reducing The vegetation on the south facade consisted of a layer of
cooling loads during peak hours in some climates even when peak mature Boston ivy (Parthenocissus tricuspidata). Several important
solar radiation occurs several hours prior. plant characteristics were either assumed based on information
from biophysics literature or directly measured, including:
3.2.2. Temperature gradient (difference between the exterior and
interior wall surfaces) Leaf absorptivity coefficient: 0.5 (assumed).
The measured temperature gradient from the exterior to inte- Average leaf dimension: 0.12 m (measured).
rior surfaces served as a surrogate for heat conduction through the Average leaf area index: 1.8 (measured).
wall because the R-value of the wall itself is the same in both Radiation attenuation coefficient: 0.4 (assumed).
vegetated and bare cases. This gradient depended most strongly on Typical stomatal conductance: 0.2 mol/m2 s (assumed).
changing exterior surface temperatures because they varied much
more than interior surface temperatures did (indoors was inside a The leaf absorptivity coefficient and the stomatal conductance
conditioned space). During sun exposure, the temperature gradient values were adopted from the literature on biophysics and are
of the vegetated façade (Tsvw Tin) was consistently lower than that considered appropriate for Boston ivy [14,15,25]. The average leaf
of the bare facade (Tsbw Tin), with a mean of 4.4 C and maximum dimension was found directly by measuring the width of several
of 11.4 C compared to a mean 4.8 C and a maximum of 16.1 C. typical leaves in the plant layer. The radiation attenuation coeffi-
This difference, which depended on the time of day, suggests the cient was found by visually evaluating a leaf angle of the plant layer.
vegetated layer can have significant energy benefits particularly The leaf area index was estimated by measuring the area of a typical
I. Susorova et al. / Building and Environment 67 (2013) 1e13 9
leaf and counting the area of ivy leaves in an image of the vegetated The effective thermal resistance of the plant layer measured
facade under study. during the experiment varied throughout the day and reached a
The portion of the wall used in the experiment consisted of the maximum of 0.14 m2 K/W based on these measured values, which
two layers of cored brick with a small air gap in the middle and suggests that the plant layer reduced conductive heat transfer by as
metal reinforcement elements, according to architectural drawings. much as another w10 cm layer of brick or w0.5 cm of expanded
Several assumptions for key facade parameters were made based polystyrene insulation in this particular climate and during these
on information on building materials in the literature [19,20]. conditions [19,20]. Measured versus modeled correlations were
generally lower for the facade temperature gradients than the
Wall thickness: 0.2 m (measured) surface temperatures alone, particularly for the cloudy day. Co-
Wall absorptivity: 0.7 for buff brick (assumed) efficients of determination (R2 values) between the measured and
Wall emissivity: 0.9 (assumed) modeled bare and vegetated facade temperature gradients were
Wall thermal resistance: 0.25 m2 K/W (assumed) 0.96 and 0.92 on the sunny day and only 0.67 and 0.51 on the
Wall density: 672 kg/m3 (assumed) cloudy day, respectively. However, the variations in measured
Specific heat: 468 J/kg K (assumed) versus modeled facade surface temperatures and temperature
gradients were small on the cloudy day for both facades and were
The thermal resistance value of 0.25 m2 K/W used for the vali- generally within the uncertainty of both the facade surface tem-
dation was assumed to be lower than brick typical thermal resis- perature measurements (0.75 C) and the facade temperature
tance of 0.40 m2 K/W [19] because of a combination of old facade gradient measurements (estimated as 1.06 C when added in
construction, the likely abundance of metal elements in the wall quadrature), which suggests that the model successfully repro-
composition, the large amount of poorly insulated windows and duced the measured results in all cases.
steel beams near our measured location, and the presence of large
air cavities between the brick layers. The measured wall was
composed of cored bricks where the total volume of solid brick 4. Discussion: sensitivity to weather and plant parameters
material is estimated to be 60% based on details typical of the
construction period (1940s). The remaining 40% of the wall volume Using the validated model, the sensitivity of conductive heat
is assumed to be air. Thus, with the assumed solid brick density of transfer differences between the bare and vegetated facades
1120 kg/m3, the actual wall density was estimated as approximately (Qbw Qvw) and the maximum plant instantaneous effective R-
60% of this number, or 672 kg/m3 [19]. Similarly, the actual wall value (Rplant) to different weather parameters was analyzed. Heat
specific heat was estimated as 60% of the assumed solid brick flux was estimated using the interior-exterior surface temperature
specific heat of 790 J/kg K, or 468 J/kg K. differences and constant thermal resistances for each facade. The
Using the experimental weather conditions, the exterior surface sensitivity of multiple model parameters was explored by changing
temperatures for the bare (Tsbw) and vegetated facades (Tsvw) were the values of one parameter at a time while keeping other pa-
estimated at each one-minute time step using the model. Then, the rameters constant. Effects of the following inputs were evaluated:
differences between the exterior and interior wall surface tem- incident solar radiation, It (0 W/m2, 400 W/m2, and 800 W/m2);
peratures of the bare (Tsbw Tinbw) and vegetated (Tsvw Tinvw) outdoor air temperature, Tair (15 C, 20 C, 25 C, 30 C, and 35 C);
facades (i.e., temperature gradients through the brick layer) were relative humidity, RH (20%, 40%, 60%, 80%, and 100%); and wind
calculated using the measured interior surface temperature (which speed, Vair (0.5 m/s, 1.5 m/s, 2.5 m/s, 3.5 m/s, and 4.5 m/s). The
remained near 24 C in the air-conditioned space). The time- indoor temperature was assumed to be 24 C in both cases. These
varying modeled values for exterior surface temperatures on the parameter ranges represent an air-conditioned environment in
bare and vegetated wall for the sunny and cloudy day used in this warm weather conditions typical of plant growing seasons. The
work were first compared with the raw experimental data as impact of each parameter is explored individually in the subse-
shown in Fig. 5. Overall, the measured and modeled values for quent section. The model parameters that were held constant
exterior surface temperatures on both the sunny and cloudy days included a leaf absorptivity coefficient of 0.5; an average leaf
were closely correlated. Coefficients of determination (R2 values) dimension of 0.15 m; an average leaf area index of 2; a radiation
between the measured and modeled facade surface temperatures attenuation coefficient of 0.5; a typical stomatal conductance of
for the bare and vegetated facades were 0.97 and 0.96 on the sunny 0.2 mol/m2 s; a wall thickness of 0.2 m; a wall absorptivity of 0.7; a
day and 0.87 and 0.86 on the cloudy day, respectively. wall emissivity of 0.9; a wall thermal resistance of 0.4 m2 K/W; and
Fig. 5. Comparison of the measured and modeled exterior surface temperatures for the bare and vegetated facades on a sunny day (August 31, 2012) and a cloudy day (September 1,
2012).
10 I. Susorova et al. / Building and Environment 67 (2013) 1e13
Fig. 6. Sensitivity of heat flux differences between the bare and vegetated facades (left) and plant effective thermal resistance (right) to incident solar radiation for relative humidity
50% and wind speed 1 m/s.
a heat capacity of 0 J/kg K. Figs. 6e8 summarize the results of the also less effective in cooling the facade exterior surface temperature
sensitivity analysis. and decreasing the heat flux through the facade. Concurrently, the
effective thermal resistance of a plant layer gradually decreases
4.1. Sensitivity to incident solar radiation when the air temperature rises. For the tested modeled range in air
temperatures values between 15 C and 35 C (and with It ¼ 800 W/
Increasing solar radiation incident on the facade had the largest m2, RH ¼ 50%, and Vair ¼ 1 m/s), the reduction in facade surface
impact on the reduction in exterior surface temperature and heat temperatures due to the plant layer varied from 12.3 C to 13.8 C.
flux through the bare and vegetated facades. Therefore, the positive Similarly, the heat flux reduction through the vegetated facade
effect of the vegetation cover and its effective thermal resistance varied from 31 W/m2 to 34 W/m2, and effective plant R-values
increases significantly with incident solar radiation; that is, when varied from 0 m2 K/W to 0.22 m2 K/W.
the solar radiation level is the highest, the plant layer resistance is
highest due to blocked transmission of radiation to the exterior wall 4.3. Sensitivity to relative humidity
surface. Conversely, when solar radiation is zero at night, the
effective thermal resistance of the plant layer is negligible. For the For the modeled relative humidity values between 20% and
range of solar radiation values explored (0e800 W/m2 at RH ¼ 50% 100% (with It ¼ 800 W/m2 and Vair ¼ 1 m/s), the reduction in
and Vair ¼ 1 m/s), the reduction in facade surface temperatures due facade surface temperatures due to the plant layer ranged from
to the plant layer ranged from 0 C and 13.9 C, corresponding to a 11.9 C to 14.2 C, heat flux reductions through the vegetated
heat flux reduction through the facade from 0 W/m2 to 35 W/m2. facade ranged from 30 W/m2 to 36 W/m2, and effective plant R-
Similarly, the modeled effective plant R-value ranged from 0 m2 K/ values ranged from 0.21 m2 K/W to 0.67 m2 K/W. When the rela-
W with no solar radiation to 0.67 m2 K/W with the highest level of tive humidity of air is low, plants significantly decrease the rate of
solar radiation. Therefore, the plant layer is shown to be particularly evaporation as a way to protect themselves from drying out [15].
beneficial in climates and locations with high level of solar inso- When the relative humidity is high, the rate of evaporation from
lation. The plant layer is also particularly effective in cooling east plants greatly increases. This process is controlled by stomatal
and west facades that are exposed to the highest levels of solar pores on the plant surface, which open and close depending on
radiation. surrounding humidity. Because of this, a facade covered with the
plant layer cools better at higher humidity levels. With increasing
4.2. Sensitivity to outside air temperature relative humidity, the exterior surface temperature reduction be-
tween the bare and vegetated facades increases and the plant
The modeled exterior facade surface temperatures and effective thermal resistance increases as well, although the
heat fluxes generally increased in response to increasing air tem- magnitude of effects is small relative to solar radiation and outdoor
peratures. At higher air temperatures, the facade plant layer was temperature.
Fig. 7. Sensitivity of heat flux differences between the bare and vegetated facades (left) and plant effective thermal resistance (right) to relative humidity for solar radiation 800 W/
m2 and wind speed 1 m/s.
I. Susorova et al. / Building and Environment 67 (2013) 1e13 11
Fig. 8. Sensitivity of heat flux reductions between the bare and vegetated facades (left) and plant effective thermal resistance (right) to wind speed for solar radiation 800 W/m2 and
relative humidity 50%.
4.4. Sensitivity to wind speed [30] (0.34 m2 K/W for a 0.16 m thick ivy layer), Wong [12]
(0.36 m2 K/W for a turf grass layer, 0.57 m2 K/W for trees, and
With increasing wind speed, the facade exterior surface tem- 1.61 m2 K/W for shrubs), and Kontoleon and Evmorfopoulou [8]
perature decreases as it cools more effectively due to convection. As (0.5 m2 K/W for a 0.25 thick ivy layer).
a result, when the wind speed increases, the reduction in heat flux
between the bare and vegetated facades and the effective plant R- 4.5. Sensitivity to plant parameters
value also decreases, although the modeled differences are small
relative to the impacts of solar radiation. For the tested wind speed The sensitivity analysis also evaluated the impacts of multiple
values between 0.5 m/s and 4.5 m/s (and with It ¼ 800 W/m2 and plant parameters, including leaf area index (LAI), radiation atten-
RH ¼ 50%), the reduction in facade surface temperatures due to the uation coefficient (k), leaf absorptivity (aleaf), leaf typical dimension
plant layer ranged from 8.5 C to 15.4 C, heat flux reduction (D), and plant stomatal conductance (gs). Among the evaluated
through the facade ranged from 21 W/m2 to 39 W/m2, and effective parameters, only LAI and gs were shown to have a significant effect
plant R-value ranged from 0.19 m2 K/W to 0.61 m2 K/W. on the plant layer thermal performance; therefore, this analysis is
Overall, the heat flux reduction between the bare and vegetated limited to calculating the vegetated facade properties with varying
facades and the plant effective thermal resistance are most sensi- LAI and attenuation coefficient. Additionally, because solar radia-
tive to changes in solar radiation, followed by wind speed, relative tion was shown to be the most important weather parameter in the
humidity, and outdoor air temperature. The sensitivity analysis previous sections, an extreme condition in a hot climate was cho-
showed that adding a plant layer to a facade works best as a passive sen to compare the impacts of LAI and k on the performance of the
cooling strategy in climates with high solar radiation and low wind. vegetated facade. A day in Phoenix, AZ with the highest solar ra-
The plant instantaneous effective R-value for the conditions used in diation (June 10th according to TMY3 weather data) was chosen to
the sensitivity analysis typically varied between 0.27 and explore the largest impact these plant parameters could likely have
0.40 m2 K/W, with a maximum value of 0.71 m2 K/W, which sug- on reducing facade temperatures and thus heat flux through walls.
gests that a plant layer with the assumed characteristics could The calculations were also performed for north, east, south, and
provide as much additional thermal resistance as w2.5 cm of west facades.
expanded polystyrene insulation. The effective plant R-values As we have shown, a layer of vegetation absorbs a fraction of
modeled in this study generally match plant layer thermal resis- incident solar radiation and transmits the rest through its foliage.
tance values previously found in other studies, including Minke The fraction of total absorbed radiation increases with higher
Fig. 11. Sensitivity of heat flux through the bare and vegetated facades (left) and plant effective thermal resistance (right) to leaf area index for a west facade in Phoenix on the
brightest day (k ¼ 0.5).
I. Susorova et al. / Building and Environment 67 (2013) 1e13 13
include solar radiation, wind speed, relative humidity, and outdoor [10] Sailor DJ. A green roof model for building energy simulation programs. Energy
and Buildings 2008;40:1466e78.
air temperature, in order of importance. The analysis also showed
[11] Energy Plus engineering reference. United States Department of Energy;
that plant layers with dense foliage (high leaf area indices) and with October 10, 2010.
leaves parallel to the wall (high attenuation coefficients) are likely [12] Wong NH, Cheong DKW, Yan H, Soh J, Ong CL, Sia A. The effect of rooftop
the most successful in reducing facade surface temperatures and garden on energy consumption of a commercial building in Singapore. Energy
and Buildings 2003;35:353e64.
heat flux through facades. On hot sunny days, a plant layer on a [13] Jones HG. Plants and microclimate. A quantitative approach to environmental
brick facade was estimated to reduce its exterior surface temper- plant physiology. Cambridge: Cambridge University Press; 1992.
ature by 0.7e13.1 C, reduce the heat flux through the exterior [14] Campbell GS, Norman JM. An introduction to environmental biophysics. New
York: Springer; 1998.
wall by 2e33 W/m2, and provide an effective R-value of 0.0e [15] Gates DM. Biophysical ecology. New York: Dover Publicaitons, Inc; 2003.
0.71 m2 K/W, depending primarily on wall orientation, leaf area [16] Deardorff JW. Efficient prediction of ground surface temperature and mois-
index, and radiation attenuation coefficient. ture with inclusion of a layer of vegetation. Journal of Geophysical Research
1978;83:1889e903.
[17] Zhang JQ, Fang XP, Zhang HX, Yang W, Zhu CC. A heat balance model
Acknowledgements for partially vegetated surfaces. Infrared Physics and Technology 1997;38:
287e94.
[18] Levitt J. Responses of plants to environmental stresses: chilling, freezing, and
This research was conducted at the Illinois Institute of Tech- high temperature stresses. San Diego: Academic Press; 1980.
nology and was funded by the Wanger Institute for Sustainable [19] Larcher W. In: Physiological plant ecology. 4th ed. New York: Springer; 2003.
Research (WISER) Interdisciplinary Seed Funding Grant, 2011. [20] Straube J, Burnett E. Building science for building enclosures. Westford:
Building Science Press Inc; 2005.
[21] ASHRAE handbook fundamentals. Atlanta: American Society of Heating,
References Refrigerating and Air-Conditioning Engineers; 2009.
[22] Asan H. Numerical computation of time lags and decrement factors for
[1] Hoyano A. Climatological uses of plants for solar control and the effects on the diferent building materials. Building and Environment 2006;41:615e20.
thermal environment of a building. Energy and Buildings 1988;11:181e99. [23] The R Project for Statistical Computing; March 13, 2013. Available from:
[2] Di HF, Wang DN. Cooling effect of ivy on a wall. Experimental Heat Transfer http://www.r-project.org/.
1999:235e45. [24] Weather Data. United States Department of Energy; March 13, 2013. Available
[3] Evmorfopoulou EA, Kontoleon KJ. Experimental approach to the contribution from: http://apps1.eere.energy.gov/buildings/energyplus/weatherdata_about.
of plant covered walls to the thermal behaviour of building envelopes. cfm?CFID¼229916&CFTOKEN¼b582a9b31f7c0ad9-CE02C17D-E418-423A-
Building and Environment 2009;44:1024e38. 5F5546DCD5860F2C.
[4] Wong NH, Kwang Tan AY, Chen Y, Sekar K, Tan PY, Chan D, et al. Thermal [25] Monteith J, Unsworth M. In: Principles of environmental physics. 3rd ed. San
evaluation of vertical greenery systems for building walls. Building and Diego: Academic Press; 2007.
Environment 2010;45:663e72. [26] Nobel PS. Physicochemical and environmental plant physiology. Burlington:
[5] Peréz G, Rincón L, Vila A, González JM, Cabeza LF. Green vertical systems for Academic Press; 2009.
buildings as passive systems for energy savings. Applied Energy 2011;88: [27] Berdahl P, Martin M. Emissivity of clear skies. Solar Energy 1984;32:663e4.
4854e9. [28] Swinbank WC. Long wave radiation from clear skies. Quarterly Journal of the
[6] Perini K, Ottele M, Fraaij ALA, Haas EM, Raiteri R. Vertical gardening systems Royal Meteorological Society 1963;89:339e48.
and the effect on air flow and temperature on building envelope. Building and [29] Yu C. The intervention of plants in the conflicts between buildings and climate
Environment 2011;46:2287e94. e a case study in Singapore. PhD thesis, National University of Singapore;
[7] Holm D. Thermal improvement by means of leaf cover on external walls e 2006.
a simulation model. Energy and Buildings 1989;14:19e30. [30] Pielke RA. Mesoscale meteorological modeling. San Diego: Academic Press;
[8] Kontoleon KJ, Evmorfopoulou EA. The effect of the orientation and proportion 2002.
of a plant-covered wall layer on the thermal performance of a building zone. [31] Minke G, Witter G. Häuser Mit Grünem Pelz, Ein Handbuch zur Hausbe-
Building and Environment 2010;45:1287e303. grünung. Köln: Edition Fricke; 1985.
[9] Alexandri E, Jones P. Developing a one-dimensional heat and mass transfer [32] Hirose T. Development of the MonsieSaeki theory on canopy structure and
algorithm for describing the effect of green roofs on the built environment: function. Annals of Botany 2005;95:483e94.
comparison with experimental results. Building and Environment 2007;42: [33] Price J. Green facade energetics. Master of Science thesis, University of
2835e49. Maryland; 2010.