Sustainable Cities and Society 76 (2022) 103463

Contents lists available at ScienceDirect

Sustainable Cities and Society

journal homepage: www.elsevier.com/locate/scs

Investigation of indoor environment and thermal comfort of building

installed with bifacial PV modules
Oufan Zhao a, Wei Zhang a, *, Lingzhi Xie b, Wei Wang a, Mo Chen a, Zihao Li a, Jianhui Li a,
Xin Wu a, Xiding Zeng a, Shaomin Du a
a
College of Architecture and Environment, Sichuan University, Chengdu, China, 610065
b
Institute of New Energy and Low-carbon Technology, Sichuan University, Chengdu, China, 610065

A R T I C L E I N F O A B S T R A C T

Keywords: In comparison with common PV module, bifacial PV module can increase energy output by installing cells on
Bifacial PV module both sides of the module. Due to the thermoelectric effect in the rear, the impact of bifacial PV module on indoor
BIPV environment is more complex. To study indoor environment of bifacial PV module as building envelope, a
Indoor environment
building integrated bifacial PV module was established. Through the experiments and simulation of the PV
Thermal comfort
building, it is found that thermal stratification is obvious at noon in summer, which indicates the bifacial PV
Optimization design
module causes a large influence on the indoor temperature. PMV is between -1 and +1 in summer nights and
winter days but it becomes poor in sunny days due to the large solar radiation intensity. Additionally, the indoor
light environment is studied. The indoor lighting is ideal in general but overlarge when the solar radiation is too
strong. For the purpose of improving the indoor comfort, the coverage of PV module was changed and the heat
insulation reflective layer was installed behind the PV modules. With the optimization design, the indoor
environment has been significantly improved, and the period of annual thermal comfort has increased by 8%.

1. Introduction improves the overall power generation by absorbing the indoor reflected
radiation and the surrounding scattered radiation from both the front
PV applied on building as envelop such like sunshade (Kang, Hwang, and the rear (Lopez-Garcia, Casado, & Sample, 2019). The structural
& Kim, 2012; Akbari Paydar, 2020), roof (Kubota, Chyee, & Ahmad, model of bifacial PV module is shown in Fig. 1. Katsaounis et al. (2019)
2009; Thebault, Clivillé, Berrah, & Desthieux, 2020), skylight (Karthick, predicted the daily power generation of single-sided and bifacial PV
Kalidasa Murugavel, & Kalaivani, 2018), window (Guo et al., 2020) and modules through experiments and simulation, and found that the energy
other building components could provide power and better living output of bifacial PV modules could increase 15% at 45 ◦ tilt angles
environment for buildings (Yoo, 2019). Common PV modules are crys­ compared with single-sided PV modules, in that bifacial PV modules
talline silicon (C-Si) and thin film. Efficiency of C-Si is higher than that of could absorb more reflected and scattered radiation. Yin et al. (2021)
thin film, with the highest efficiency of 22% (Skandalos & Karamanis, found that in comparison with mono-facial solar cells, bifacial solar cells
2015), but the light transmittance of C-Si is poor. A thin film has good provide over 6% more energy yield. And as a result of enhanced effects
light transmittance and low cost, but its efficiency is low from 10% to of a reflective coating on the back, the output of power could be
14.5% (Thopil, Sachse, Lalk, & Thopil, 2020). As a result of receiving increased by 10%.
solar radiation from both sides, the efficiency of bifacial PV module is Although bifacial PV module has a lot of advantages, there is a few
higher than traditional PV modules such like C-Si and thin film. Ac­ researches of its application on buildings because it is a novel product.
cording to statistics, the production capacity of bifacial PV module has Myong and Jeon (2016) combined bifacial PV module with windows,
exceeded 45GW. And it is predicted that the share of bifacial PV module and found that the loss of lower direct plane-of-array irradiation at a
in the world market will steadily increase to more than 15% by 2024 higher tilt angle can be compensated by the power generation of bifacial
(Guerrero-Lemus, Vega, Kim, Kimm, & Shephard, 2016). PV module. Baumann et al. (2019) studied the combination of bifacial
In comparison with common PV module, bifacial PV module PV module and green roof, and found that the annual energy output of

* Corresponding author.
E-mail address: zhangwei821@scu.edu.cn (W. Zhang).

https://doi.org/10.1016/j.scs.2021.103463
Received 30 April 2021; Received in revised form 26 August 2021; Accepted 16 October 2021
Available online 20 October 2021
2210-6707/© 2021 Elsevier Ltd. All rights reserved.
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

bifacial PV module on indoor environment were analyzed. To further

improve the indoor comfort, the building was optimized. The coverage
of PV cells was changed and the heat insulation reflective layer was
installed behind the PV modules. Finally, the indoor environment of the
reconstructed building is analyzed and evaluated. As an innovative
research on indoor comfort of the building mounted with bifacial PV
modules, this study provides theoretical basis and technical support for
the future researches of BIPV indoor environment.

2. Methodology
Fig. 1. Structural model of bifacial PV module.
In this research, the performance of the bifacial PV module applied
bifacial PV systems mounted vertically with east-west orientation is on building was studied. In Comparison with traditional PV module, the
equal to that of the single-sided module installed horizontally. Accord­ energy output of bifacial PV module has increased by 5%-19% as a result
ing to the previous researches, most of them focus on the impact of the of indoor lighting and bifaciality. In the purpose of studying the indoor
performance of bifacial PV module on building energy consumption. environment, the surface temperature and room temperature of bifacial
Bifacial PV module can absorb solar radiation on both sides, which PV windows were tested, and the indoor thermal comfort is tested based
leads to a more complex impact on indoor environment, especially on the temperature result. Additionally, indoor illuminance of a week
thermal comfort. Currently, there are a few researches on indoor com­ was tested. Through the establishment of the corresponding simulation
fort of PV buildings, and majority of the PV modules used in these re­ model, the indoor temperature distribution at specific time, thermal
searches are common C-Si and thin film. Schmid and Uehara (2017) comfort and indoor lighting of the whole year were simulated.
found that the buildings with PV windows works on a 54% to 57 % lower
illuminance level compared with direct gain windows. Liu, Sun, Wilson, 2.1. The bifacial PV module and the PV building
and Wu (2020) found that the percentage of time with ‘Useful’ illumi­
nance (UDI100–2000lux) could be significantly increased by using CdTe As shown in Fig. 2, the PV building was built in a two-story pre­
PV windows as building envelope. Wang et al. (2018) compared the fabricated building in Guanghan City, Deyang, Southwest China (east
solar thermal gain coefficient (SHGC) between the traditional window longitude 104-105 ◦ and north latitude 30-31 ◦ ). The bifacial PV module
and the semi-transparent PV window, and found that the SHGC of was used in the room (length * width * height: 4.68m * 3.3m * 3.2m).
traditional windows was much higher than that of the semi-transparent The bifacial PV building is composed of three sides (southwest, south­
PV module, which can reach 0.15. Yoon, Shim, An, and Lee (2013) east and roof) of the external protection structure with the bifacial PV
analyzed the temperature of PV modules and traditional windows in module and three sides (northwest, northeast and floor plane) of the
different directions. The results indicated the maximum temperature internal protection structure. There are three bifacial PV modules
difference between PV module and the conventional window can reach mounted on the southwest, six bifacial PV modules mounted on the
10 ◦ C. Han, Lu, Peng, and Yang (2013) compared the indoor tempera­ southeast and the roof plane. At the same time, 6mm thick conventional
ture of a room which was mounted with a conventional window and windows are used in the other southwest, southeast positions and the
semi-transparent PV module. The result indicated that the maximum roof plane to provide sufficient natural lighting for the indoor
indoor temperature of the experimental room with the conventional environment.
window was close to 34 ◦ C in summer, while the temperature of the The bifacial PV module in this research was designed and produced
room with the semi-transparent PV module was only 29◦ C. Salehi et al. by our cooperator Suzhou talison PV Co., Ltd, and the specific plane size
(2019) simulated thermal comfort of the building installed with solar of it is shown in Fig. 3. The bifacial PV module consists of a total
chimney and found that the optimal time periods of operation of Solar thickness of 1 mm PV cell and two layers of glass with thickness of 2.5
Chimney mainly depend on climate. (Zhu, Wang, & Guo, 2019) verified mm, and the parameters of it are listed in Table 1.
utilization of solar energy with OVW (Optic-Variable Wall) for thermal The bifacial PV module not only absorbs direct solar radiation, but
comfort, and found that the comfort hour is significantly increased with also absorbs indoor reflected radiation and ambient scattered radiation
the use of OVW. From the above research, PV module as building en­ for power generation (Shoukry, Libal, Kopecek, Wefringhaus, & Werner,
velope can improve indoor light environment, reduce indoor tempera­ 2016). As shown in Fig. 4, the radiation part of the PV module returns to
ture in summer, and have a great impact on indoor thermal comfort. the indoor and outdoor environment through reflection, and the rest is
Though the energy output of bifacial PV module is higher than absorbed by the PV module and converted into electric energy for indoor
common PV module, there are few studies which connected it with in­ power generation. The converted heat enters the environment through
door environment, especially indoor thermal comfort . Both the front convection and radiation heat exchange.
side and back side of bifacial PV module produce heat, making the
thermal environment of PV building more complex (Zhang et al., 2020). 2.2. Test rig set up
And indoor thermal environment directly affects the sensible heat ex­
change of human body through convection and radiation, which is the For the purpose of testing the lighting of room, four light meters were
main factor influencing the comfort of human body (Giannopoulou arranged on the indoor 1.3m high horizontal plane, as shown in Fig. 5a.
et al., 2014). So there is a necessity to study the indoor environment, To study temperature distribution, thermocouples were placed at 0.1m,
especially thermal comfort of PV building. 0.7m and 1.1m of the indoor vertical plane, as shown in Fig. 5b. The
In the purpose of exploring the impact of bifacial PV module on in­ comfort meter could directly test the indoor comfort, as shown in Fig 5a.
door environment, a prefabricated building was established in Guan­ The comfort meter recorded data every 1min, and the experimental data
ghan, Sichuan Province, with the roof and two facades installed with of light meters and thermocouples were collected by multi-channel re­
bifacial PV modules. The thermal environment parameters were tested, corders every 1min. All the equipment used in this experiment is shown
including surface temperature of PV modules, indoor temperature dis­ in Table 2.
tribution and thermal comfort. Besides, indoor light environment also
was measured simply. Simulation models of thermal simulation and
light simulation were established to analyze indoor environment typi­
cally and annually. According to the test and simulation, the impact of

2
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig 2. The building integrated bifacial PV module.

Fig. 4. The energy conversion of bifacial PV module.

Fig. 3. The specific plane size of the bifacial PV module.

independent. Both of them are intended to analyze the indoor thermal
environment. Radiance was used to simulate and analyze the indoor
Table 1 light environment. Besides, the relationship between the experimental
The parameters of bifacial PV module. results and simulation results could be summarized in two aspects:
a The Structure of bifacial PV module Experimental data could be input into the simulation models as input
parameters and the simulation models could be validated with the test
Layer Thickness (mm)
Front glass 2.5 results.
PV cell layer 1 To study the temperature distribution, it is necessary to analyze the
Back glass 2.5 heat conduction of the whole PV building (Naghavi et al., 2021). Ac­
b. The performance parameters of PV module cording to the analysis of heat transfer process of bifacial PV module, the
equations of PV heat absorption were established in STAR CCM+. Due to
Parameters Value
Active area, A(m2) 1.474 the absorption of solar radiation and indoor emission radiation by the
Short circuit current, Isc(A) 9.89 PV module, the energy transfer process of the PV module which is shown
Open circuit voltage, Voc(V) 40.5 in Fig. 6 was considered as a 3-dimensional steady-state heat transfer
Current at the maximum power point, Imp(A) 9.39 problem with an internal heat source.
Voltage at the maximum power point, Vmp(V) 34.1
Absorptivity (%) 90
The differential equation of heat conduction is shown below (Yang
Reflectivity (%) 10 et al., 2019):
Heat transfer coefficient (W/(m*K)) 0.22
∂2 T ∂2 T ∂2 T
λ +λ 2 +λ 2 +S=0 (1)
∂x 2 ∂y ∂z
2.3. Simulation of indoor environment
where, λ is the thermal conductivity of PV module, W/(m•K); T is the
2.3.1. Simulation method temperature of PV module, ◦ C; S is the square source term, taking into
To study the indoor environment of the PV building, the experiments account the solar radiation absorbed by PV module and power output,
of temperature distribution, thermal comfort and light environment W.
were carried out. Based on the test results, one kind of CFD software: The internal and externa surfaces of the PV module are the bound­
STAR CCM+ was used to simulate and analyze the temperature distri­ aries of combined convection and radiation heat transfer, while the
bution; EnergyPlus V8-9-0 was used to simulate and analyze the thermal contact surface with ordinary windows is the boundary of the thermal
comfort. The simulations of STAR CCM+ and EnergyPlus are insulation. The specific boundary conditions show as follows (Peng, Lu,

3
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig. 5. Thermal and light environment test platform.

⃒ ⃒ ⃒ ⃒
Table 2 ∂T ⃒⃒ ∂T ⃒ ∂T ⃒ ∂T ⃒
− λ = − λ ⃒⃒ = − λ ⃒⃒ = − λ ⃒⃒ = 0 (4)
The specifications of the instrument. ∂x ⃒x=D ∂x x=0 ∂y y=L ∂y y=0
Equipment Manufacture Function Accuracy/Sensitivity
where, hci is the convection heat transfer coefficient of internal surface
Multi-channel JianTongtech Data collector The minimum
data recorder resolutions are 1 μV and of PV module, hco is the convection heat transfer coefficient of external
0.1 ◦ C surface of PV module, W/(m2 • K); ξ 1, ξ 2 are radiation heat transfer
Light meter JianTongtech Illuminance 0–100,000 lux; 1 lux; factors of inner and outer surfaces of PV module; hri, hro are the radiation
recorder ±4% heat transfer coefficients of the internal and external surfaces of PV
T-type JianTongtech Temperature − 20 ◦ C–100 ◦ C; 0.1 ◦ C;
module, W/(m2•K); D, L are the thickness and height of the PV module,
Thermocouples test ±0.5 ◦ C
Thermal imager Jiantongtech Thermal 0.1◦ C m; Ti, To, T1 and T2 are indoor air temperature, outdoor air temperature,
imaging PV module outdoor surface and internal surface temperature,◦ C.
Indoor thermal Jiantongtech Thermal − 20 ◦ C − 85 ◦ C; +0.5◦ C According to Peng et al. (2013), hri and hro are respectively:
comfort tester comfort 0~x223C 100%RH;±3%
( )
recorder RH 0.05–5 m/s, 0.3 m/s hri = σ T2 2 + Ti 2 T2 + Ti (5)
+ 2%
( )
hro = σ T1 2 + To 2 T1 + To (6)
Yang, & Han, 2013):
⃒ where, σ is the Stefan Boltzmann constant, 5.67*10− 8 W/(m2•K).
∂T ⃒
− λ ⃒⃒ = hco T1 − To + ξ1 hro T1 − To (2) The radiative heat transfer factors ξ1 and ξ2 on the outside and inside
∂x z=0 of the PV module are as follows (Peng et al., 2013):
⃒ 1 1 1
∂T ⃒⃒ = + − 1 (7)
− λ = hci T2 − Ti + ξ2 hri T2 − Ti (3) ξ1 εo ε1
∂x ⃒z=H

Fig. 6. The 3-D model of the PV module.

4
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig. 7. The specific size of the PV building (Left: Southwest wall. Middle: Roof top. Right: Southeast wall).

1 1 1 body surface area without clothes. ta is the ambient temperature (◦ C). tcl
= + − 1 (8)
ξ2 εi ε2 is the clothing surface temperature (◦ C). tr is the average radiation
temperature (◦ C). And the relationship between PPD and PMV (Tian
where, εo ,εi ,ε1 and ε2 the emissivity of outdoor environment, indoor et al., 2019) is shown as equation 15:
environment, external surface and internal surface of bifacial PV mod­
ule. PPD = 100 − 95exp − 0.03353PMV4 − 0.2197PMV2 (15)
The outdoor surface heat transfer coefficient of PV module is (Smolec In order to study the impact of bifacial PV modules on light envi­
& Thomas, 1993): ronment of the PV building, UDI was calculated with Radiance. UDI
hco = 5.7 + 3.8v (9) refers to the annual effective daylighting intensity, which is used to
evaluate the time proportion of the annual indoor illumination in the
where, v is the local outdoor wind speed, m/s. comfort zone (Mandalaki, Tsoutsos, & Papamanolis, 2014). In accor­
dance with UDI, the indoor lighting can be divided into three ranges:
Nuy λa
hci = (10) 300lx, 300lx-2000lx and 2000lx. It only fully evaluates the illuminance
Y level between 300 lux and 2000 lux, which means that the illuminance
where, λa is the thermal conductivity of indoor air, W/(m2•K); Nuy is the beyond the range is invalid and need to be regulated artificially. The
Nusselt number at y. calculation formula of UDI is shown in equation 16 (Nabil & Mardal­
At the same time, the heat source S in the PV module can be jevic, 2006):
simplified as the heat directly absorbed by solar radiation and generated tUD
UDI = (16)
by power generation (Yaman & Arslan, 2021): tT
S = G⋅θ + E (11)
Where, tUD is effective daylighting time, s; tT is total daylighting time, s.
2
Where, G is the radiation amount irradiated on the PV module, W/m ; θ
2.3.2. Simulation geometry and physical performance parameters
is the solar radiation heat conversion coefficient; E is the heat generated
The length (L), width (B) and height (H) of the PV building are
due to power generation.
4.68m, 3.3m and 3.2m respectively as shown in Fig .7. The positive
Based on the test results of indoor temperature and thermal comfort,
direction is 45 ◦ north by west, the altitude is 518m, and the positive
the indoor thermal comfort was simulated by using EnergyPlus V-8-9-0.
wall (southwest wall) is mainly composed of three PV modules and
PMV represents thermal comfort, which is the evaluation of people’s
single-layer glass module, where the wall in the right and upper parts of
satisfaction with the surrounding thermal environment. The equations
PV module are composed of six PV modules and single-layer glass
(Tian et al., 2019) are:
module respectively. It is connected with the southeast wall surface and
M − W = (C + R + Esk ) + ( Cres + Eres ) + ( Ssk + Scr ) (12) the roof surface, while the bottom and the left are 1355mm and 208mm
respectively according to the floor surface and the northwest wall sur­
Load = (M − W)− 3.05 ×10− 3 ×[5733− 6.99(M − W)− Pa]− 0.42 ×[M − W] face. At the same time, the side wall (southeast facing wall) and the roof
− 58.15]− 1.7 × 10−[ 5M × (5867− (
Pa)− 0.0014M(34−
)4 ]
ta ) plane are composed of six PV modules and single-layer glass modules of
− 3.96 ×10− 8 fcl × (tcl +273)4 − tr +273 − fcl hc ×[tcl − ta ] the above size.
Then, according to the actual size of the PV building and the per­
(13)
formance parameters of bifacial PV module, the numerical simulation
MV = 3.155 × 0.3 × e− 0.114M
+ 0.28L (14) models were established. The thermal model of the BIPV is illustrated in
Fig. 8. Considering that the thermal parameters of all PV modules are the
where M is metabolic heating rate (W/m2), W is mechanical work rate same, and the PV modules on each side are connected together, we make
(W/m2), C is the rate of heat convection loss. (W/m2), R is radiant heat a simplified treatment for the PV modules and glass on different sur­
loss of the skin (W/m2), Cres is convective heat loss rate of respiration faces. The southwest wall is composed of a PV module and conventional
(W/m2), Eres is evaporative heat loss rate of respiration (W/m2), Ssk is window with a size of 1669mm * 3024mm. The southeast wall is
the rate of skin heat storage (W/m2), Scr is the rate of body heat storage composed of a PV module and conventional window of a size 3018mm *
(W/m2). Load is the heat load of the body. Icl is the clothing heat resis­ 3411mm. Finally, the roof top is composed of a whole PV module of size
tance (◦ C•m2/W). fcl is the ratio of body surface area with clothes to of 3213mm * 3411mm and a conventional window. Because the room is

5
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig. 8. The thermal model of the PV building.

in a completely closed state during the test process, the model used was
closed window and door.
The light environment simulation model of the PV building is illus­
trated in Fig. 9, and the size of the model is consistent with that of the
original PV building.

2.4. Optimization disign of the prefabricated building

To further improve the quality of the indoor environment, the

coverage of PV was optimized and reflective insulation layer was
installed behind the PV board. It is measured that when the coverage
rate of bifacial PV is 74%, the PV building may have the best power
generation efficiency and indoor lighting. Besides, A translucent
reflective layer (20-100nm silver film) is proposed to be installed on the
back of the bifacial PV modules, as shown in Fig 10. The reflective layer
can reflect the near-infrared and infrared parts of the spectrum, improve
the reflectivity and radiation uniformity, reduce the temperature of the
module, improve the efficiency of the module, and reduce the indoor Fig. 10. Bifacial PV module with installation of reflective film.
temperature. Besides, it can let the visible light pass through without
affecting the indoor lighting. Then, EnergyPlus V-8-9-0 and Radiance more than 800W/m2, and reached 1100 W/m2 on July 5th. Because the
could be used to analyze the results with the optimization design. effect of outdoor solar radiation, on sunny days, the outdoor air was
heated by solar radiation, and the outdoor environment temperature
3. Results analysis of indoor thermal environment increased or decreased to a large extent in the morning and afternoon
respectively. On cloudy days, due to the low level of outdoor solar ra­
3.1. Test results diation, the heating effect on the outdoor air was limited, and the change
ratio of outdoor climate temperature was much smaller than that on
3.1.1. Outdoor climatic condition results sunny days.
The outdoor environment test results of a representative week
(2019/07/05-2019/07/12) are shown in Fig 11. It could be seen that 3.1.2. Test results of surface temperature of PV modules
four days were sunny, and the maximum solar radiation value reached To further analyze the temperature changes on different surfaces, the
temperature difference between different PV modules and the

Fig. 9. The light environment simulation model. Fig 11. Test results of outdoor environment.

6
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

conventional windows was tested with a thermal imager, and the test window was almost below 31◦ C. For the conventional window on the
results are illustrated in Fig. 12. From the test results, it can be seen that roof with strong solar radiation, the maximum temperature was only
the temperature of PV modules was higher than that of conventional 31.3◦ C. The temperature of the PV module could reach above 30◦ C, even
windows. The average temperature of the inner surface of conventional the highest temperature would be 33.3◦ C. The maximum temperature

Fig. 12. Thermal imaging test results of surface temperature.

7
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

difference between PV modules and conventional windows was in the thermal environment, the outdoor surface temperature of different PV
southwest, which could reach 4~x223C 5◦ C. And the minimum tem­ modules was tested.
perature difference was in the southeast, which was about 2~ x223C 3 The temperature change curve of the outdoor surface of PV module
◦
C. on the surface of the bifacial PV building is shown in Fig. 14. Because of
The test was not carried out when the outdoor temperature was the the effect of solar radiation on sunny days, the overall temperature of the
highest. And if it was done when the outdoor temperature was the PV module on the roof was significantly higher than that on the wall,
highest, the temperature difference would be higher. The temperature and the maximum temperature could reach about 64◦ C. The tempera­
difference between PV module and conventional window lead to a huge ture difference between the wall and the roof PV module reaches the
impact of PV module on the indoor thermal environment compare to maximum at about 1 pm, and the difference between the average out­
that of the conventional window. Therefore, it is necessary to analyze door surface temperature of the of the horizontal roof PV module and the
the impact of the PV module on the indoor thermal environment.
To analyze the thermal environment of bifacial PV building, the in­
door surface temperature of different PV modules was tested with the
thermal imager, and the test results are illustrated in Fig. 13. There were
obvious differences in the temperature of the PV modules in all di­
rections, including the indoor surface temperature of the top plane PV
module which could reach between 31.2◦ Cand 34.2◦ C, and the average
indoor surface temperature of the whole panel was 32◦ C. However, the
indoor surface temperature of PV module on the southeast wall was
29.5◦ C, and the maximum indoor surface temperature of PV panel was
only 30.6◦ C. The central temperature of the PV installed on the south­
west wall between the two was only slightly higher than that on the
southeast wall, and the highest temperature reached only about 32.4◦ C.
At the same time, the temperature difference between the lower and
upper part of the PV module on the southeast wall was large, which
indicates there is interaction between the PV panel temperatures on the
different planes. The temperature difference of PV modules on different
planes and the interaction between the temperature of the PV module
leads to a serious stratification of indoor ambient temperature. There­
Fig. 14. Surface temperature test results of PV module.
fore, in order to further analyze the impact of PV module on indoor

Fig. 13. Thermal imaging test results of PV module surface temperature.

8
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

southeast wall PV module could reach about 11◦ C on July 10th. All of (average thermal sensation index) of the space under different climates.
this indicates that the indoor thermal environment is huge impact by the As shown in the box chart, PMV is generally distributed between 0.25
PV module on the sunny days. and 2 in sunny days, with the minimum value of 0.003 and the
On cloudy days, the difference of the outdoor surface temperature maximum value of 2.747. The median line is at 0.3, and the difference
between the PV module in each plane is relatively low. According to between the lower and upper limits is large. On cloudy days, PMV
outdoor surface temperature change of PV module on July 11th, it was generally fluctuates at PMV =1, the minimum value is 0.340, the
found that the temperature of the horizontal roof PV module on cloudy maximum value is 1.513, the median line is at 1, and the difference
days could only reach 28◦ C, while the maximum the outdoor surface between the upper and lower limits is reduced. The results show that this
temperature of the PV module on the two vertical walls could also reach map is not satisfactory for different climate. As shown in the Fig. 16b,
about 27.5◦ C. It was mainly caused by the weak solar radiation on PDD mainly distributes between 10% and 80% in sunny days, with the
cloudy days, and the heat produced by solar radiation was far lower than minimum value of 5.01% and the maximum value of 94.19%. The me­
that on sunny days. At the same time, there is mainly diffuse radiation on dian value is 55%. It can be seen that the space comfort is not satisfac­
cloudy days, which leads to the small difference in the solar radiation tory in sunny days. On cloudy days, PDD mainly distributed between
received by the PV module on each plane. On sunny days, the outdoor 20% and 30%, the minimum value was 7.46%, the maximum value was
solar radiation mainly consisted of direct radiation, the solar radiation 51.62%, and the median line was at 20%. Compared with sunny days,
received by each plane was greatly affected by the solar height angle. the comfort of the space was more satisfactory on cloudy days.
Therefore, the temperature difference between the horizontal plane and Based on the above analysis, it can be seen that the comfort of the
the vertical plane was large on sunny days. space is always more satisfactory on cloudy days than on sunny days.
The reason is that the transparent glass as the enclosure structure,
3.1.3. Indoor temperature test result and analysis although it complements the indoor lighting, its poor heat insulation
According to the analysis of the outdoor surface temperature of the performance will lead to a greater impact on the indoor temperature in
PV module, the outdoor surface temperature of the PV module on the top sunny days. Besides, due to the heat absorption and power generation
plane is much higher than the temperature on the vertical surface on heating effect of bifacial PV modules, the temperature of modules will
sunny days, which lead to heating of the air at high levels of the bifacial rise, which will undoubtedly affect the indoor thermal environment.
PV building. Hence the temperature of the indoor ambient at top of the
room must be higher than that at the bottom, and the air density of the
high-temperature air is lower than that at low temperature, which lead
to significant vertical air temperature stratification in the PV building.
To analyze the thermal environment of the PV building, the vertical
temperature field in the area with intense human activity was tested
(under 1.1m high), and the test results are shown in Fig. 15. On cloudy
days (July 7th), due to the low temperature of the PV module on the
horizontal roof, the maximum indoor ambient temperature was only
about 30◦ C. The vertical stratification of air temperature in the bifacial
PV building was slight, and the difference between the temperature at
the highest position (1.1m) and the temperature at the lowest position
(0.1m) was only 0.5◦ C at noon. In the case of sunny days (July 9th), the
maximum temperature of the horizontal roof PV module reached 55 ◦ C
which led to the indoor ambient temperature in the highest position of
indoor air would reaching 32.1◦ C at noon, while the temperature in the
lowest position was only about 28.5◦ C. The difference in the indoor
ambient temperature between the highest and the lowest position
reached 3.6◦ C, greatly affecting the internal thermal comfort on sunny
days.

3.1.4. Indoor thermal comfort test result and analysis

Based on the test of temperature distribution of the PV building, the
indoor thermal comfort is further tested. Fig. 16a shows the PMV

Fig. 15. Indoor air temperature test results. Fig. 16. Indoor thermal comfort test results on sunny and cloudy days.

9
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

reflects the deviation between the numerical simulation results and the
measured results. It is suggested from ASHRAE Guideline 14 that if the
MBE and Cv(RMSE) of a building simulation model fall within10% and
30% respectively, the reality of the simulation could be proved (Coak­
ley, Raftery, & Keane, 2014). The calculation formulas are shown as
follows:
It could be seen from the figures above that the trend of the simu­
lation results was consistent with that of experimental test results. The
measured PV panel temperature and indoor temperature were slightly
lower than these in the numerical simulation. The main reason for the
difference is that the test results were affected by the occlusion of the
surroundings.

3.2.2. Simulation of surface temperature of PV modules and conventional

windows
Under the influence of solar radiation, the temperature change of PV
modules was more complex, and the subsequent impact on the indoor
thermal environment was more complex. STAR CCM+ was used to
analyze the temperature change of bifacial PV modules and conven­
tional windows. As shown in Fig. 19, except at 9 am when the sun had
just risen, the surface temperature of the conventional window was
higher than that of the PV module under the influence of the direct
sunlight. The temperature of PV module was higher than that of the
conventional window at the other times. Especially at 1 pm, the
maximum temperature difference between the PV module and conven­
tional window reached 6-7◦ C. Under the influence of the solar radiation,
the temperature of PV module on the roof plane is much higher than that
on the vertical plane, and reached 47.6◦ C. At the same time, the tem­
perature of the PV module in the southwest was the lowest, and the
maximum value was only 31.5◦ C. At 8 pm, under the influence of low
solar radiation angle in the afternoon, the average temperature of PV
module on the southwest was slightly higher than that on the roof.

3.2.3. Simulation results of indoor temperature field

Based on the simulation of surface temperature, the indoor temper­
ature field was simulated with STAR CCM+. The vertical temperature
Fig. 17. Comparison of Surface temperature of PV module
distribution at typical moments (9 am, 1pm and 8 pm) was simulated to
analyze the vertical temperature distribution of the PV building, and the
simulation results are shown in Fig 20.
Because the radiation absorbed by each surface was small, as the sun
rises in the morning, the surface temperature of the PV module on the
roof plane and the southwest wall was relatively low. The heating effect
of PV module on the indoor air was small at this time. From the Fig 20a,
it is found that the indoor temperature distribution is uniform, especially
in the middle part, and the air is not affected by the heating of the PV
module, which results in a relatively uniform temperature distribution
on the vertical surface at this time.
The temperature of the middle part increased to about 23◦ C at 1:00
p.m., because the indoor air has received sufficient radiation. Mean­
while, the solar radiation on the roof plane is greater than that on the
other two surfaces, the indoor temperature of the PV module on the roof
plane reaching about 50◦ C, and the temperature of the air at the top
which was directly affected by the PV module also reached 46.5◦ C. Due
to the negative heat transfer of the air, the stratification of the indoor
Fig. 18. Comparison of indoor temperature.
temperature was increased. When the solar radiation was strong. the
heating of the top PV module effectively affected the indoor vertical
3.2. Simulation results and validation plane temperature field distribution.
At 8 pm, as the outdoor solar radiation dropped below 100 W/m2,
3.2.1. Validation of temperature analysis and thermal comfort the surface temperature of PV module on each surface was decreased
The numerical simulation results and the experimental results were significantly, among which the temperature of PV module on the roof
used to analyze the reliability of the numerical model. During the vali­ plane was only 30◦ C. As a result, the maximum temperature of the in­
dation process, two statistical values (MBE and Cv(RMSE)) were chosen door air obviously decreased compared with that at 1 pm, which was
to evaluate the deviation between the experimental and simulation re­ close to the PV module move to 0.5m from the PV module, with a value
sults (Myong, 2016). The Mean Bias Error (MBE) that mainly represents of only 36◦ C. Due to the heating of the PV during daytime, the indoor
the average difference between the numerical simulation results and the temperature field in the vertical direction at this time also had obvious
measured results was used, while the root mean square error (Cv(RMSE)) stratification. The air temperature at the bottom was much lower than

10
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig. 19. The surface temperature of PV module at the typical moments.

Fig. 20. The vertical indoor temperature field at the typical moments.

the temperature at the top, only about 29.5◦ C. Similar to the results of the indoor ambient temperature in the ver­
Due to the large indoor depth, up to 4.68m, the indoor horizontal tical direction, the radiation received on each surface was small, espe­
temperature field was analyzed and is shown in Fig. 21 below. cially for the PV module in the southwest, where temperature was less

11
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

than 23◦ C at 9 am. As shown in Fig. 21a, the air temperature in the reduce thermal conductivity of bifacial PV module and transparent glass
indoor horizontal plane is almost unaffected by the PV module, and the and enhance the heat insulation performance, and it can reduce the solar
temperature distribution is relatively uniform, especially in the middle transmittance of transparent glass and the heat gain through the
part. windows.
Because the solar radiation is mainly eastward from the morning to
noon in summer, the solar radiation absorbed by the southeast wall PV 3.2.5. Simulation results of thermal comfort with optimization design
module was greater than that absorbed by the southwest wall PV module After the optimization of the building model, the simulation results
before 1 pm, which caused the surface temperature of the southeast wall of annual thermal comfort are shown in Fig. 23. The comfort degree of
PV module to be significantly higher than that of the southwest PV the whole year after the transformation has been improved, the comfort
module. As shown in Fig. 21b, the maximum temperature on the indoor time ratio of the whole year is 28%, which is 8% higher than the 20%
environment level at this time appear close to the southeast wall. before the transformation, and the time under the extreme conditions of
Finally, as the solar radiation in the afternoon is mainly concentrated + 3 and - 3 is significantly reduced. PMV is mostly between 0 and + 1 in
on the southwest wall, the surface temperature of the PV module on the summer and - 1 to 0 in winter.
southwest wall is gradually higher than that on the southeast wall in the
afternoon, which results in the highest temperature point on the indoor 4. Results analysis of indoor light environment
horizontal surface near the southwest wall at 8 pm. At the same time,
due to the temperature reduction of the PV module at night, the highest 4.1. Test results
temperature appears near from the PV module on the southwest wall.
The daylighting performance of the building in sunny days was
3.2.4. simulation results of indoor thermal comfort tested for one week, and the experimental results are shown in Fig 24. It
The annual indoor comfort of the building is simulated with Ener­ could be seen that using bifacial PV modules and transparent glass as the
gyPlus V8-9-0, and the results are shown as Fig. 22. building envelope can provide sufficient daylighting for the indoor
It can be seen from the above figure that the comfort time ratio of environment. However, in extreme cases, such as 12:00-14:00 noon, the
indoor environment of the building installed with PV module is 20% and indoor illumination will be far greater than the maximum illumination
the comfort at noon in summer is poor. Due to the strong solar radiation, required by indoor people. On July 8th, the noon illumination will even
the PMV value is + 2 ~x223C + 3, which is in the overheated state. The reach 4500lux.
simulation results at this time are similar to the experimental results.
Because the solar radiation is not too strong in summer morning and
evening, the indoor temperature is proper and the indoor environment is 4.2. Simulation results and validation
comfortable. In winter, the indoor environment is comfortable most of
the time during daytime because of the heating effect of the bifacial PV 4.2.1. Validation of indoor light environment
module. And in winter night, the indoor temperature is low and the The comparison of illuminance results on July 8th is shown as Fig. 25.
value of PMV is - 2~ x223C -1. The main reason for the poor thermal It could be seen from the figure that the trends of indoor illuminance
comfort is the poor thermal insulation performance of bifacial PV were similar between experiment and simulation. It could be seen from
module and transparent glass. To further improve the thermal comfort the figure that the trend of the simulation results was almost consistent
environment of the PV building, the way of installing reflective layer can with that of experimental results. The MBE and Cv(RMSE) of the indoor
illuminance of the PV building were -5.25% and 13.38%. Therefore, the

Fig. 21. The horizontal indoor temperature field on the typical moments.

12
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Fig. 22. Simulation of indoor annual comfort.

Fig. 23. Indoor PMV simulation with the optimization design.

Fig. 24. Average illumination test results. Fig. 25. Comparison of illuminance performance.

light environment model is reliable. 4.2.3. Simulation results of indoor light environment with optimization
design
4.2.2. Simulation results of indoor light environment The Simulation results of indoor light environment with optimiza­
Based on the test results and the validation, the simulation results of tion design is shown in the Fig. 27. The UDI has been increased by 20% -
light environment with Radiance are shown as Fig. 26. As shown in the 30% with the optimization design. The UDI reached 75 ~x223C 90% in
figure, UDI in the central area could reach 50% ~x223C 60%, while that the central area and 40%-50% near the conventional glass. These results
in the window area is only 20%. The illumination near the conventional show that the indoor light and thermal environment has been effectively
window area is too large because of the large area of conventional glass improved after changing the coverage of PV and installing the reflective
module and the high transmissivity of conventional glass. The trans­ layer.
missivity of bifacial PV module is almost zero, it can absorb the scattered
light and improve the lighting quality of the area near the PV module 5. Conclusions
and the central area, so that the illumination in these areas is better.
In this research, a building integrated bifacial PV module was built.
Through the test and simulation of the PV building, the impact of

13
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

which is in the overheated state. It is because that when the solar

radiation is strong, the transmission of transparent glass and heating
effect of bifacial PV module both will lead to indoor overheating
3 Additionally, the indoor lighting environment of the PV building was
studied. It is found that the indoor lighting is ideal in general.
Especially in the central area, the UDI can reach about 60%. But at
noon in summer, when the solar radiation is too strong, the indoor
illumination will excess, which is caused by the high transmissivity
of the transparent envelopes. The transparent envelopes could be
optimized to improve the quality of the indoor light environment.
4 The building was optimized to improve indoor environment and the
result was evaluated. The coverage of PV was adjusted to 0.74 and
reflective insulation layer was installed behind the PV module. With
the optimization design, the indoor annual thermal comfort time
ratio increased by 8%, and the UDI of the indoor environment has
increased by 25%-40% The results showed that optimization design
adopted has a good affect on improving the indoor environment of
building integrated bifacial PV module.
5 If bifacial PV module is installed on the sloping roof or flat roof, the
surrounding structure is recommended to be coated with high
reflectivity material: white paint. The solar radiation reflected by the
Fig. 26. Indoor light environment simulation.
high reflectivity material could be absorbed by the PV module, which
leads to improvement of efficiency of the PV module. If bifacial PV
module is installed on the wall with reflective film installed behind
it, active devices such as fans are recommended to use to enhance the
air circulation between the air gap (gap between PV module and
reflective film) and outdoor environment. As a result of the air cir­
culation, the surface temperature of the bifacial PV module could be
reduced and the efficiency of it could be increased.

Declaration of Competing Interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence
the work reported in this paper.

Acknowledgments

It is supported by National Key Research and Development Program

of China (No. 2016YFE0124500), Science & Technology Department
Foundation of Chengdu City, China (2017-GH02-00006-HZ), Joint
Laboratory of Integrated Energy and Environmental Systems for Green
Fig. 27. Indoor light environment simulation with the optimization design. Buildings with Science and Technology Department of Yibin City, China
(2019CDYB-17), and the Science and Technology Department Project of
bifacial PV module on indoor environment was explored. Finally, the PV Sichuan Province (No. 2020YJ0061).
building is optimized and the indoor environment with the optimization
design is analyzed. References

1 As shown in the test results, there is a great impact of the PV modules Akbari Paydar, M. (2020). Optimum design of building integrated PV module as a
movable shading device. Sustainable Cities and Society, 62.
on the indoor thermal environment because the surface temperature Baumann, T., Nussbaumer, H., Klenk, M., Dreisiebner, A., Carigiet, F., & Baumgartner, F.
of the PV module is far higher than that of conventional window on (2019). Photovoltaic systems with vertically mounted bifacial PV modules in
sunny days. In the terms of the indoor thermal environment influ­ combination with green roofs. Solar Energy, 190, 139–146.
Coakley, D., Raftery, P., & Keane, M. (2014). A review of methods to match building
ence analysis. Because of the huge temperature difference between
energy simulation models to measured data. Renewable and Sustainable Energy
the roof and the vertical wall, temperature stratification will occur in Reviews, 37, 123–141.
both the indoor vertical and horizontal temperature fields. The Giannopoulou, K., Livada, I., Santamouris, M., Saliari, M., Assimakopoulos, M., &
Caouris, Y. (2014). The influence of air temperature and humidity on human thermal
maximum temperature difference in the horizontal direction and in
comfort over the greater Athens area. Sustainable Cities and Society, 10, 184–194.
the vertical direction are 10◦ Cand 4.5◦ Crespectively. Guerrero-Lemus, R., Vega, R., Kim, T., Kimm, A., & Shephard, L. E. (2016). Bifacial solar
2 The indoor thermal comfort in the PV building was tested. It is found photovoltaics – A technology review. Renewable and Sustainable Energy Reviews, 60,
that the comfort of the space is always more satisfactory on cloudy 1533–1549.
Guo, W., Kong, L., Chow, T., Li, C., Zhu, Q., Qiu, Z., et al. (2020). Energy performance of
days than on sunny days in different climates. According to the test photovoltaic (PV) windows under typical climates of China in terms of transmittance
results, the indoor thermal comfort of the whole year was simulated, and orientation. Energy, 213.
it is found that in sunny evenings or cloudy days, indoor thermal Han, J., Lu, L., Peng, J., & Yang, H. (2013). Performance of ventilated double-sided PV
façade compared with conventional clear glass façade. Energy and Buildings, 56,
comfort is satisfied, and PMV is between - 1 and + 1. And the comfort 204–209.
at noon in summer is poor, and the PMV value is + 2 ~x223C + 3, Kang, S., Hwang, T., & Kim, J. T. (2012). Theoretical analysis of the blinds integrated
photovoltaic modules. Energy and Buildings, 46, 86–91.

14
O. Zhao et al. Sustainable Cities and Society 76 (2022) 103463

Karthick, A., Kalidasa Murugavel, K., & Kalaivani, L. (2018). Performance analysis of Skandalos, N., & Karamanis, D. (2015). PV glazing technologies. Renewable and
semitransparent photovoltaic module for skylights. Energy, 162, 798–812. Sustainable Energy Reviews, 49, 306–322.
Katsaounis, T., Kotsovos, K., Gereige, I., Basaheeh, A., Abdullah, M., Khayat, A., et al. Smolec, W., & Thomas, A. (1993). Theoretical and experimental investigations of heat
(2019). Performance assessment of bifacial c-Si PV modules through device transfer in a Trombe wall. Energy Conversion & Management, 34(5), 385–400.
simulations and outdoor measurements. Renewable Energy, 143, 1285–1298. Thebault, M., Clivillé, V., Berrah, L., & Desthieux, G. (2020). Multicriteria roof sorting for
Kubota, T., Chyee, D. T. H., & Ahmad, S. (2009). The effects of night ventilation the integration of photovoltaic systems in urban environments. Sustainable Cities and
technique on indoor thermal environment for residential buildings in hot-humid Society, 60.
climate of Malaysia. Energy and Buildings, 41(8), 829–839. Thopil, G. A., Sachse, C. E., Lalk, J., & Thopil, M. S. (2020). Techno-economic
Liu, D., Sun, Y., Wilson, R., & Wu, Y. (2020). Comprehensive evaluation of window- performance comparison of crystalline and thin film PV panels under varying
integrated semi-transparent PV for building daylight performance. Renewable Energy, meteorological conditions: A high solar resource southern hemisphere case. Applied
145, 1399–1411. Energy, 275.
Lopez-Garcia, J., Casado, A., & Sample, T. (2019). Electrical performance of bifacial Tian, H., Zhang, W., Xie, L., Ni, Z., Wei, Q., Wu, X., et al. (2019). Thermal Comfort
silicon PV modules under different indoor mounting configurations affecting the rear Evaluation of Rooms Installed with STPV Windows. Energies, 12(5).
reflected irradiance. Solar Energy, 177, 471–482. Wang, W., Zhang, W., Xie, L., Wu, Y., Tian, H., & Zheng, L. (2018). Experimental
Mandalaki, M., Tsoutsos, T., & Papamanolis, N. (2014). Integrated PV in shading systems Assessment of the Energy Performance of a Double-Skin Semi-Transparent PV
for Mediterranean countries: Balance between energy production and visual comfort. Window in the Hot-Summer and Cold-Winter Zone of China. Energies, 11(7).
Energy and Buildings, 77, 445–456. Yaman, K., & Arslan, G. (2021). A detailed mathematical model and experimental
Myong, S. Y., & Jeon, S. W. (2016). Efficient outdoor performance of esthetic bifacial a- validation for coupled thermal and electrical performance of a photovoltaic (PV)
Si:H semi-transparent PV modules. Applied Energy, 164, 312–320. module. Applied Thermal Engineering, 195.
Nabil, A., & Mardaljevic, J. (2006). Useful daylight illuminances: A replacement for Yang, J., Cheng, Y., Jia, J., Du, Z., Shi, Z., & Han, J. (2019). The impact of indoor air
daylight factors. Energy and Buildings, 38(7), 905–913. distributions on the thermal performance of a single layer semi-transparent
Naghavi, M. S., Esmaeilzadeh, A., Singh, B., Ang, B. C., Yoon, T. M., & Ong, K. S. (2021). photovoltaic facade. Building Simulation, 12(1), 69–77.
Experimental and numerical assessments of underlying natural air movement on PV Yin, H. P., Zhou, Y. F., Sun, S. L., Tang, W. S., Shan, W., & Huang, X. M. (2021). Optical
modules temperature. Solar Energy, 216, 610–622. enhanced effects on the electrical performance and energy yield of bifacial PV
Peng, J., Lu, L., Yang, H., & Han, J. (2013). Investigation on the annual thermal modules. Solar Energy, 217, 245–252.
performance of a photovoltaic wall mounted on a multi-layer façade. Applied Energy, Yoo, S.-. H. (2019). Optimization of a BIPV system to mitigate greenhouse gas and indoor
112, 646–656. environment. Solar Energy, 188, 875–882.
Salehi, A., Fayaz, R., Bozorgi, M., Asadi, S., Costanzo, V., Imani, N., et al. (2019). Yoon, J.-. H., Shim, S.-. R., An, Y. S., & Lee, K. H. (2013). An experimental study on the
Investigation of thermal comfort efficacy of solar chimneys under different climates annual surface temperature characteristics of amorphous silicon BIPV window.
and operation time periods. Energy and Buildings, 205. Energy and Buildings, 62, 166–175.
Schmid, A. L., & Uehara, L. K. S. (2017). Lighting performance of multifunctional PV Zhang, Z., Wu, M., Lu, Y., Xu, C., Wang, L., Hu, Y., et al. (2020). The mathematical and
windows. Energy and Buildings, 154, 590–605. experimental analysis on the steady-state operating temperature of bifacial
Shoukry, I., Libal, J., Kopecek, R., Wefringhaus, E., & Werner, J. (2016). Modelling of photovoltaic modules. Renewable Energy, 155, 658–668.
Bifacial Gain for Stand-alone and in-field Installed Bifacial PV Modules. Energy Zhu, Y., Wang, C., & Guo, X. (2019). Smart utilization of solar energy with Optic-Variable
Procedia, 92, 600–608. Wall (OVW) for thermal comfort. Energy and Buildings, 202.

15