Energy & Buildings 298 (2023) 113540

Contents lists available at ScienceDirect

Energy & Buildings

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

Assessing heat transfer characteristics of building envelope deployed BIPV

and resultant building energy consumption in a tropical climate
Hamza Jhumka a, Siliang Yang a, *, Christopher Gorse b, Sara Wilkinson c, Rebecca Yang d,
Bao-Jie He e, f, g, h, Deo Prasad i, Francesco Fiorito i, j
a
School of Built Environment, Engineering and Computing, Leeds Beckett University, Leeds, United Kingdom
b
School of Architecture, Building and Civil Engineering, Loughborough University, Loughborough, United Kingdom
c
School of Built Environment, University of Technology Sydney, Sydney, Australia
d
School of Property, Construction and Project Management, RMIT University, Melbourne, Australia
e
Centre for Climate-Resilient and Low-Carbon Cities, School of Architecture and Urban Planning, Chongqing University, Chongqing, China
f
Network for Education and Research on Peace and Sustainability (NERPS), Hiroshima University, Hiroshima, Japan
g
Key Laboratory of New Technology for Construction of Cities in Mountain Area, Ministry of Education, Chongqing University, Chongqing, China
h
State Key Laboratory of Subtropical Building Science, South China University of Technology, Guangzhou, China
i
School of Built Environment, University of New South Wales, Sydney, Australia
j
Department of Civil, Environmental, Land, Building Engineering and Chemistry, Polytechnic University of Bari, Bari, Italy

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

Keywords: Building-Integrated Photovoltaic (BIPV) is a viable technology towards increasing renewable energy production
BIPV glazing and achieving low carbon footprints for buildings. Mauritius, with a daily average of 5.6 kWh/m2 of solar ra­
Roof BIPV membrane diation over 2350 h annually, has been targeting at achieving its low carbon goals by focusing on photovoltaic
Double-skin façade
technology including the uptake of BIPV. However, BIPV has not been well researched in terms of its overall
Heat transmission
thermal impact especially overheating on the building envelope and the resultant energy performance for
Thermal impact
Cooling load buildings for the tropical climatic condition in Mauritius. This research, by means of validated simulation
modelling, adopted a novel approach of coupling thermal finite element analysis (FEA) with whole building
dynamic simulations to assess the heat transfer characteristics of BIPV either on facades or roof and the resultant
energy consumptions of a typical office building in Mauritius. The façade scenario had two options, namely BIPV
curtain wall and BIPV double-skin façade (BIPV-DSF), while the roof scenario also had two options, namely
uninsulated and insulated roof BIPV membranes. Results show that roof BIPV membrane options had a better
thermal performance in reducing overheating for the building compared to the BIPV façade options, with a
reduction in cooling load of 8% and 15% for the uninsulated and insulated BIPV membranes, respectively. In
terms of energy performance, both BIPV façade options were not capable of reducing the energy consumption of
the building, as the BIPV curtain wall resulted in 1.66% more net energy consumption on a yearly basis. This
shows an ineffectiveness of using vertical BIPV glazing for typical office buildings in Mauritius. Although the
BIPV-DSF achieved an annual net energy saving of 5.16% benefited from the BIPV energy production, it was not
as good as the net savings of 160% and 172% from the respective uninsulated and insulated roof BIPV membrane
options.

sector with energy generation through fuel and coal power stations, is
being a major catalyst of greenhouse gases in our atmosphere [2–4]. As
1. Introduction such, it can be deduced that global warming is a by-product of energy
consumption. Buildings are amongst the biggest energy consumers
Global warming is one of the major issues that needs to be tackled by worldwide, and reducing the energy demand of buildings can lower
countries worldwide, with a target of keeping the warming effect below their carbon emission footprint and help reduce the effect of global
1.5 ◦ C over the next two decades as a generally accepted consensus [1]. warming [5,6].
This warming effect, led by one of the main culprits that the energy

* Corresponding author.
E-mail address: [email protected] (S. Yang).

https://doi.org/10.1016/j.enbuild.2023.113540
Received 26 May 2023; Received in revised form 17 August 2023; Accepted 11 September 2023
Available online 14 September 2023
0378-7788/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Nomenclature Text external air temperature [K]

Tint internal air temperature [K]
Symbols TPV photovoltaic module temperature [K]
Ai inside surface area [m2] Tsi internal surface temperature [K]
Cp specific heat capacity [J/kgK] Tsup supply air temperature [K]
h convective heat transfer coefficient [W/m2K] Tsurf surface temperature [K]
hext external surface convective heat transfer coefficient [W/ Tz zone mean air temperature [K]
m2K] Tzi interzone air temperature [K]
hi internal surface convective heat transfer coefficient [W/ T∞ outside air temperature [K]
m2K] t time taken [s]
hint− h internal surface (horizontal) convective heat transfer U-value heat transfer coefficient or thermal transmittance [W/
coefficient [W/m2K] m2K]
hint− v internal surface (vertical) convective heat transfer V voltage in one-diode equivalent circuit [V]
coefficient [W/m2K] Vmpp voltage at maximum output [V]
I current in one-diode equivalent circuit [A] Voc open circuit voltage of photovoltaic module [V]
IL photovoltaic module photocurrent [A/W] xi calculated value [-]
Impp current at maximum output [A] yi simulated value [-]
Io diode reverse saturation current [A] Greek symbols
Is incident solar radiation on solar cell surface [W/m2] γ empirical photovoltaic curve-fitting parameter [-]
Isc short circuit current of photovoltaic module [A] ε emissivity of surface being subject to radiant energy [-]
i instance of study [-] εext emissivity of external surface [-]
k thermal conductivity [W/mK] εint emissivity of internal surface [-]
k Boltzmann’s constant [J/K] ρ density [kg/m3]
Mi measured data [-] σ Stefan-Boltzmann constant [W/m2K4]
MP average of measured data [-] Abbreviations
ṁi interzone air mass flow rate [kg/h] ACH air changes per hour
ṁinf outside air mass flow rate due to infiltration [kg/h] a-Si amorphous silicon
ṁsys supply air mass flow rate [kg/h] BAPV building-attached photovoltaic
n total number of data samples [-] BIPV building-integrated photovoltaic
NP number of values for a given time duration [-] BIPV-DSF building-integrated photovoltaic in double-skin facade
P calculation time interval [-] CdS cadmium sulphide
Pmpp maximum power of photovoltaic module [Wp] CdTe cadmium telluride
Q heat flux [W/m2] CIGS copper indium gallium selenide
QD− Tr directly transmitted heat flux [W/m2] CVRMSE cumulative variation of root mean squared error
Qelec converted electric power [W/m2] DSF double-skin facade
QETFE heat flux through ethylene tetrafluoroethylene layer [W/ DSSC dye-sensitised solar cell
m2] DX direct expansion
QEVA heat flux through ethylene–vinyl acetate layer [W/m2] ETFE ethylene tetrafluoroethylene
Qfloat heat flux through float glass layer [W/m2] EVA ethylene–vinyl acetate
Qglass heat flux through low iron glass layer [W/m2] FEA finite element analysis
QPV heat flux through photovoltaic module layer [W/m2] HVAC heating, ventilation, and air conditioning
QPVB heat flux through polyvinyl butyral layer [W/m2] MBE mean bias error
Qreflect reflected heat flux [W/m2] NOCT nominal operating cell temperature
Qsolar irradiance on solar cell surface [W/m2] OPV organic photovoltaic
PIR polyisocyanurate
Q̇i internal load [kW]
PSC perovskite solar cell
Q̇load net zone thermal load [kW] PV photovoltaic
Q̇sys air system thermal load [kW] PVB polyvinyl butyral
q elementary charge [C] PVC-P plasticised polyvinyl chloride
Rs photovoltaic module series resistance [Ω] RRMSE relative root mean square error
Si simulated data [-] SHGC solar heat gain coefficient
T local temperature [K] TPT Tedlar Polyester Tedlar
Ta ambient air temperature [K] VLT visible light transmittance
Tair external or internal air temperature [K]
Tc photovoltaic cell temperature [K] [K]

Mauritius recently embarked on several ambitious journeys, first lead towards more construction of mid- to high-rise buildings that
with the “Smart City Scheme” to promote the construction of smart cities require more electrical energy [9], the climate change mitigation mea­
around the island, and at the same time adopt a low carbon strategy to sure of Net Zero Carbon will force the building and construction sector
keep up with the United Nations climate change policies agreed at the towards a more sustainable and low energy approach during design,
COP26 [7,8]. The target is to achieve carbon neutrality by 2070 along construction, and operation [10]. Khoodaruth et al. [11] conclude that
with an intermediate goal of generating 40% of its energy through re­ the participation of the Mauritian policy makers and regulators are
newables by 2030 [8]. Whilst adoption of the Smart City Scheme would fundamental along with public sensitisation and financial investments

2
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

towards decarbonising the energy systems through 100% renewable unframed, roof tiles or a building component itself [24]. There are
energy by 2050 for the country. modules which are either opaque, or semi-transparent, that are mounted
Solar energy is one of the most widely adopted renewable energy on structural frames, or between glass panels to be fixed as glazing
sources to increase sustainability in most sectors including building and components on a building [25]. When they are integrated into the
construction [12]. With Mauritius receiving a daily average of 5.6 kWh/ building structure they are known as BIPV, whereas if they are fixed
m2 of solar radiation over an average of 2350 h annually, solar energy is onto a building component or structure then they are known as Building
one of the most viable options in terms of renewable source [11]. The Attached Photovoltaic (BAPV) [26]. Without knowledge of the
United Nations Development Programme (UNDP) with the support of mounting system used, it is difficult to differentiate between the two
the Green Climate Fund has been assisting Mauritius in achieving its since PV arrays constitute most of the visible portion in both cases. The
2030 Renewable Energy Roadmap by focusing on solar photovoltaic BAPVs have no significant effect on the functionality of the envelope of a
(PV) technology [13]. Ramgolam and Soyjaudah [14], while studying building, the BIPVs however, will have a direct impact on the envelope,
the potential of PV systems, found that Mauritius could yield an average as it is replacing a specific building component and; is a fundamental
of 1428 kWh/m2 of solar energy per year. While the PV system has an variable parameter for the energy balance of the building [18].
encouraging future as a sustainable and renewable energy technology, The CIBSE (Chartered Institution of Building Services Engineers)
the main barrier to its implementation is physical space [15]. On-site Knowledge Series KS15 [25] indicates that there are two types of BIPV
solar farms for individual buildings can take a considerable amount of systems, namely roof-integrated and façade-integrated, by providing a
space, while being aesthetically unpleasant and being an unwanted list of sub-classifications for both. The façade-integrated type is cat­
source of glare [16]. Since the 2000 s, an innovative form of PV systems egorised as vertical curtain walling, inclined wall glazing, rain-screen
has been developed and implemented globally, namely building- cladding and sun shading, while the roof-integrated type is cat­
integrated photovoltaic (BIPV). This system which uses PV arrays inte­ egorised as inclined roof, curved roof, skylights, and atrium. It should be
grated into the building envelope can form a cohesive design, con­ noted that the roof-mounted systems are usually placed on a structural
struction, and energy solution for buildings [17]. Several studies such as roof and considered as BAPV. It is stated that the roof-integrated system
Biyik et al. [18], Yu et al. [19] and You and Yang [20] reviewed and provides for the better energy performance, while the façade-integrated
showed the encouraging potential of BIPV as energy generators, while has more opportunities to be used as a secondary function such as rain
Shukla et al. [21] gave an indication of the lifecycle assessment showing screen or sun shading while providing for an architectural statement
a reasonable economic payback and carbon footprint reduction of BIPV [25]. Being transparent, the façade installations also provide for more
in general. daylighting capabilities [27,28]. Yu et al. [19] further categorised the
On the other hand, solar radiation that reaches the BIPV arrays is glazing-mounted BIPV into four distinct types, namely single layer,
partly absorbed by the building envelope itself [22]. The absorbed heat double-layer with closed air gap, double-layer with a ventilated air gap
travels through convection, if there is a gap between the arrays and the and vacuum type, while the last one had the lowest heat transfer rate as
building envelope, then it travels by conduction and radiation to the low as 0.6 W/m2K under the influence of solar irradiation. The double-
internal surface of the building to be transmitted to the indoor space layer PV also had low heat transmission at almost the same magnitude as
thermal load [23]. However, very little research has assessed the heat the vacuum type, but the single layer PV had the highest heat trans­
transfer characteristics of BIPV on building envelopes and its impact on mission with a rate as high as 5.5 W/m2K for certain solar cell materials.
the cooling load of buildings [19]. Most existing research is limited to They indicate that there is lack of research on the double-layer and
analysing PV module surface temperatures and the impact of solar vacuum types compared to the single layer type to provide for a more
transmission from glazed BIPV options even though BIPV glazing makes impartial assessment of their heat transfer capabilities and energy
up only 20% of the current market [21]. Therefore, this paper presents performance.
research aimed at investigating the comparative thermal impact of both
roof and glazed type BIPVs on the indoor air temperature of a typical 2.2. BIPV material and its applications in buildings
office building situated in the tropical climate of Mauritius. The BIPV
heat transmission will then be integrated into energy models to assess Three main types of BIPV material are monocrystalline silicon,
the overall effect of this heat gain on the energy consumption of the polycrystalline silicon, and thin-film coating [29], while some lesser
building and the resultant energy savings due to the energy production used materials are the non-silicon versions such as dye-sensitised solar
of the BIPV. cell (DSSC), perovskite solar cell (PSC) and organic photovoltaic (OPV)
In summary, this paper is subdivided into 6 sections, namely intro­ [30]. Monocrystalline silicon is the most efficient type, but also the most
duction, literature review, materials and methods, model calibration, expensive one, due to being made from a slice of a single crystalline
results and discussion, and conclusion. The introduction section gives silicon, while polycrystalline, which has larger cells, is made from multi-
insight on the background information leading to the research motiva­ crystalline silicon cast in a mould [31]. Thin-film is a micron sized
tion. Literature review covers the current studies with respect to energy coating made of cadmium sulphide (CdS), cadmium telluride (CdTe) and
and thermal impact of BIPV on buildings, where a problem statement amorphous silicon (a-Si), which is used on a variety of substrates ranging
concludes the review with clear guidelines on the significance of this from glass to steel [30]; although it is the least efficient, layering two or
research. The section of materials and methods provides an overview of more thin-film PV junctions on top of each other can improve the per­
the methodology including a description of the case study with the input formance both in electricity conversion efficiency and reduction of
parameters. Model calibration shows the calibration procedure of vali­ mechanical degradation over time [32,33]. Research has shown the
dating the simulation model. Results and discuss section presents the potentialities of using thin-film based BIPV on building envelopes in
numerical results obtained and the findings of this study. Finally, the terms of its thermal and energy performance [34,35].
section of conclusions restates the research conducted and summarises Some studies revealed the availability and feasibility of BIPV appli­
the main findings with a critical insight on the key metrics from the cation in different parts of the building envelope. Shukla et al. [21]
analysis. conclude the recent technologies classifying BIPV installations into
tiling products for roofs, glazing products for both roofs and façades, foil
2. Literature review products for curved roofs, and other products for shading and cladding
materials. They found that 80% of the current market is dominated by
2.1. Overview of BIPV the roof-mounted BIPV, using both glazed semi-transparent materials
and opaque tiling materials, while the façade-mounted installations only
PV modules come in several configurations namely framed, account for 20% of the market. Shukla et al. [21] also indicate that the

3
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

payback period for both installations with monocrystalline requiring (heating, ventilation, and air conditioning) in the assessment.
around 2.7 to 7.3 years, while thin-film costs could be recovered within By using a high reflectivity heat insulation film laminated with a
the first 8 months to 4 years at most. Previously, by surveying the Eu­ varied combination of thin-film semi-transparent PV modules of a-Si, Liu
ropean market, Pester [36] found that monocrystalline was the most et al. [50] discovered an average of 34.2% energy savings on the HVAC
used PV cell for the opaque application such as roof tiles, slates and in sub-tropical climate of Taiwan due to the low SHGC and low U-value
frameless laminates, while the thin-film was used mostly for the semi- achieved from the insulation. They also found a visible light reflectance
transparent glazing for windows, skylights and curtain walls. of only 5% using the insulation film, hence the impact on the visual
Furthermore, the thin coating consisting of copper indium gallium performance, or light pollution, of the glazed PV was negligible. In terms
selenide (CIGS) photovoltaic cells are used on lightweight metal roofing of a DSF variation to a semi-transparent a-Si thin-film glazing, Han et al.
also as a BIPV [37]. The foil type, which uses a-Si combined with other [51] determined that a ventilated gap in the PV module would provide a
cushioning structural material such as ethylene tetrafluoroethylene reduction of cooling load in subtropical climates. However, they only
(ETFE), comes in the form of rolls and membranes that are used on roof analysed temperature profiles with respect to energy generating ca­
surfaces by simple adhesion without the need for frames or mounting pacities of the BIPV without an actual analysis of the effects on cooling
elements [38]. Those flexible BIPV modules can take the shape of the loads.
roof structure including the curved roof [30].
2.4. Problem statement
2.3. Bipv-induced thermal impact on buildings
Even though literature has shown that thermal and energy perfor­
BIPVs are normally analysed through their electricity generation
mance evaluation for BIPV is a well-documented research topic, there
efficiency noted by energy production and their thermal performance
are considerable ambiguities about the thermal gains and subsequent
through the heat transfer coefficient; commonly known as the U-value
impact on energy savings, which makes the assessment of BIPV for the
[39]. When BIPV is transparent, or semi-transparent (glazed type), its
Mauritian climate a difficult task. Mauritius has a cooling only climate
thermal performance includes the solar heat gain coefficient (SHGC)
[52], while the energy savings stated in the literature are often a result of
which informs of the solar irradiation transmitted through the PV media
the impact of BIPV on space heating. Several factors, such as local codes,
into the building [40]. Another performance factor normally involved
construction practices and socio-economic contexts may also affect the
when dealing with glazed BIPV is the optical performance measured
impact of BIPV on local buildings [53]. Although thermal aspects are
through the visible light transmittance (VLT) and the glare probability
present in current literature, the impact of BIPV installation on the
value to assess the indoor visual effects [19,40,41], while an acceptable
cooling load of a building has not been researched.
daylighting can be maintained when the VLT is within the range of 25%
This paper proposes a novel approach of assessing the thermal per­
− 38% [42,43].
formance of a semi-transparent BIPV glazing and a roof BIPV membrane
A comprehensive literature review by Maghrabie et al. [30]
on buildings in Mauritius, using a case study of a typical office building
concluded that semi-transparent thin-film PV modules have superior
to analyse cooling load variations in the tropical climate of the country
performance levels with their capabilities to produce electricity while
and correlating the thermal performance to energy harvesting capabil­
reducing the cooling load through the partial absorption of solar radi­
ities of the BIPV for an overall energy assessment. The significance of the
ation incident on the envelope. However, most of the research was based
proposed research will provide a basis for the building professionals to
on the PV modules with a façade integration, and few of them were
assess the thermal impact and energy performance of BIPV on buildings
compared to roof-mounted systems [44]. Earlier, Ban-Weiss et al. [22]
in a broader tropical climate context such as the hidden problem of
provided an analysis of a roof-integrated BIPV consisting of thin-film a-
overheating.
Si noting a 5 ◦ C decrease on the upper surfaces, as a result of the
increased solar absorption and a 9.6 kWh/m2 reduction on the cooling
3. Materials and methods
load; however, the BIPV used for the research is no longer available on
the PV market and further research is needed to update the present roof-
This research aimed to investigate two types of BIPV materials on the
mounted thin-film BIPV.
fabric of a case study building in Mauritius. Since the case study building
In comparison, it was found that a crystalline PV module can convert
is existing and operational, a real retrofit is not economically feasible.
around 15–20% of incident irradiation into electrical energy, while
Thus, the research was essentially based on calibrated simulations for a
another 5–10% are reflected, or converted, into other energy sources
comparative study. The study comprised two areas – thermal perfor­
[30,45]. This could allow the BIPV modules to yield a surface temper­
mance and energy performance analyses. The thermal analysis consisted
ature of 60 ◦ C on hot and sunny days, which warrants the need for
of a detailed finite element heat transfer analysis and a dynamic building
ventilated gaps between the BIPV and building envelope to provide a
performance analysis. Specifically, the finite element heat transfer
thermal break in the energy transfer [30].
analysis was used to predict thermal performance of the selected BIPV
In recent years, Yang et al. [43,46–48] numerically analysed the
materials, while the dynamic building performance simulation was used
effect of a-Si, DSSC and PSC based BIPVs combined with a ventilated
to analyse the thermal impacts of the BIPV on the building cooling load.
double-skin façade (DSF) in Australian climate, which found that the
On the other hand, the energy performance analysis consisted of an
ventilated air gap produced energy savings from 34% up to 106%
energy modelling of the building usage and PV energy generation, which
depending on the region due to the heat harvesting capabilities of the air
assesses the whole building energy performance through parametric
gap. They also demonstrated that the heat recovery combined with the
scenarios.
electricity generated would potentially offset the energy demand due to
additional heat gains on the envelope [46], while the ventilated DSF
with BIPV could be beneficial to improvement of indoor thermal comfort 3.1. BIPV typologies
in summertime [43,48].
A numerical study through EnergyPlus simulation in Cameroon’s Literature showed the typical typologies of BIPV in the built envi­
tropical climates earlier showed an increase in the indoor temperature ronment. This research focused on analysing three main typologies as
by 4 ◦ C using BIPV, while the type of BIPV adopted was not specified follows:
[49]. The authors found that roof-integrated BIPV had a considerable
impact on the indoor hygrometric conditions than façade applications. 1) A flat roof based waterproofing BIPV membrane.
However, the study did not incorporate the indoor heat gains and HVAC 2) A façade based glazed BIPV curtain wall.

4
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 1. Typologies of BIPV models for analysis: roof BIPV membrane (left), BIPV curtain wall (middle), and naturally ventilated DSF with semi-transparent BIPV
glazing (right). The blue colour of the schematics represents the BIPV structure. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)

Fig. 2. The case study building – a view of south-west façade.

3) A naturally ventilated DSF comprising of a front layer of semi- Mechanical is a new approach towards BIPV analysis, but it has been
transparent BIPV glazing (also known as building-integrated photo­ commonly used to analyse regular PV modules [55–57].
voltaic in double-skin façade, shortened to BIPV-DSF).

Fig. 1 illustrates the typologies and building surfaces of the case 3.3. DesignBuilder modelling – Energy performance of BIPV
study that were substituted with the selected BIPV material for the nu­
merical simulations in assessing the thermal load variations of the DesignBuilder is a whole building dynamic simulation software that
building. provides a graphical user interface to the EnergyPlus simulation engine.
The typical workflow consists of the selection of a location and corre­
sponding weather data, the creation of the thermal building model ge­
3.2. Ansys modelling – Thermal performance of BIPV ometry and assigning occupancy and equipment operation schedules;
the results range from thermal load profiles to lighting illumination
Ansys is a multi-physics finite element analysis (FEA) programme profiles and other energy simulation data [58]. DesignBuilder is used
being widely used in both commercial and academic settings, which has mainly to evaluate façade options, daylighting analysis, visualisation of
different packages that can numerically model and solve static and dy­ site layouts and solar shading, thermal conditions, and the sizing of
namic heat transfer and fluid problems [54]. The package used in this HVAC equipment and systems [59]. A number of studies have success­
research was Ansys Mechanical, which performed steady state finite fully validated the thermal and energy performance simulation capa­
element heat transfer analysis for the individual BIPV materials. Ansys bilities of DesignBuilder [60–62].

5
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Table 1 Where ε is the emissivity of the surface being subject to radiant en­
Case study building features. ergy, σ is the Stefan-Boltzmann constant at 5.669 × 10-8 W/m2K4, Tsurf is
Type of features Parameters Details the temperature of the external surface being subject to radiant energy.
Climate features Location Pointe aux Canonniers,
Mauritius 3.5.2. Thermophysical modelling of BIPV materials
Climate classification Tropical monsoon climate The BIPV components used for this research were selected from
Architectural Total building floor area 2003 m2 commercially available products in the form of a semi-transparent triple
features Façade area 354 m2
laminated glass with 10% transparency from Onyx Solar (the manu­
Façade glazing area 172 m2
Mechanical Cooling system only Direct Expansion (DX) coil split
facturer) and a polymer-based roof waterproofing membrane from Axter
features system (the manufacturer). The simulated BIPV glazing consisted of three layers
Energy for building Electricity of float glasses and Polyvinyl Butyral (PVB) interlayers, while the central
operation float glass was etched with a-Si PV which had an efficiency of 4.74%. In
Conditioned area 1712 m2
this case, all the layers were joined together without any air gaps. The
Unconditioned area 291 m2
roof BIPV membrane consisted of an ETFE polymer with flexible CIGS
modules (with an efficiency of 16.6%) encapsulated between an ethyl­
3.4. Case study building overview ene–vinyl acetate (EVA) sheet on the photoactive side and a Tedlar
Polyester Tedlar (TPT) sheet on the backside. The membrane was used
The case study building is a typical office building located in Pointe along with adhesives and fasteners on top of Plasticised Polyvinyl
aux Canonniers, the north-west of Mauritius (as shown in Fig. 2). It has Chloride (PVC-P) waterproofing sheets and vapour barriers from the
around 471 m2 space area per floor over three floors, of which both the same manufacturer for a full waterproofing solution. The simulation
ground floor and first floor have a mezzanine having roughly 60% floor model included a thermally insulated and uninsulated version of the
space and 40% open space. The building height is 14.8 m, and it has a composite waterproofing membrane. The physical and thermal proper­
glazed façade oriented towards the south-west. Due to the building ties for each component of both the BIPV glazing and roof membrane
being in the Southern Hemisphere, the glazed façade is oriented towards were taken from the manufacturer data. Table 2 and Table 3 summarise
the south to limit heat gains in the building. the respective thermal properties used to model the BIPV glazing and
The building features are given in Table 1, while Fig. 3 shows the membrane in Ansys Mechanical.
south-west elevation of the façade with its curtain wall at ground floor With both BIPVs having transparent and semi-transparent materials
level and large glazed windows on the remaining floors, as well as a on the photoactive sides, the radiative heat fluxes were highly depen­
typical floor layout with its mezzanine. The building is used as an office dent on the transmissivity, reflectivity and absorptivity of each indi­
with common circulation areas and customary wet areas. The building vidual material. Table 4 presents the radiation properties of the
envelope consists primarily of lightweight 200 mm concrete blocks with transparent materials on the photoactive side for both BIPVs. Fig. 4 il­
cast-in-place concrete columns, beams and slabs for the structural ele­ lustrates the thermal model used for both the BIPV glazing and roof BIPV
ments. The curtain wall and façade windows are double glazed 6 mm membrane materials and the energy balance between incident radiative
glass with a 12 mm air gap. beam, power conversion and heat transfer to surrounding environment.
Furthermore, the thermal radiation properties were used to calculate
the boundary conditions for each layer (as per Fig. 4) in Ansys Me­
3.5. Development of numerical modelling
chanical. Onyx Solar (the manufacturer) had already been providing the
total solar reflection and direct solar transmission factors for the triple
3.5.1. Finite element heat transfer modelling
laminated glass at 51.3% and 7.4% respectively, which were used to
Heat transfer is energy transfer between material bodies due to
calculate the boundary conditions. Table 5 shows the boundary condi­
temperature differences; the governing equations used for finite element
tions of each layer based on calculations and assumptions, while the
steady state heat transfer analysis is derived from the first law of ther­
Appendix (attached in this paper) illustrates the meshing and mesh
modynamics for the conservation of energy [63] as follows:
refinement for the FEA simulation in Ansys Mechanical.
dT
Qconduction + Qconvection + Qradiation = ρCp (1)
dt 3.5.3. Dynamic building performance simulation
Where Q is the heat flux (in conduction, convection, and radiation), ρ The primary output of the dynamic building performance simulation
is the density of a material, Cp is the specific heat capacity of the ma­ is to analyse how the BIPV materials would impact the thermal loads of
the case study building by retrofitting the BIPV glazing and the roof BIPV
terial, dT is the temperature change with time taken as 1 s for a steady
dt
membrane to the existing building. The DesignBuilder software was
state calculation.
used to perform the proposed dynamic building simulation, which
Specifically, heat transfer through conduction is given by Fourier’s
adopts the EnergyPlus simulation engine to calculate the cooling load of
law as:
the building using the following equation [65].
dT
Qconduction = − k (2) Nsl
∑ Nsurfaces
∑ N
∑ zones
dΔ Q̇load = Q̇i + hi Ai (Tsi − Tz ) + ṁi Cp (Tzi − Tz ) + ṁinf Cp (T∞ − Tz )
Where k is the thermal conductivity of the material, dΔ
dT
is the tem­ i=1 i=1 i=1

perature gradient occurring in ×, y, and z directions. (5)

The convective heat transfer is given by Newton’s law of cooling as: Where Q̇i is the internal load, hi Ai (Tsi − Tz ) is the convective heat
Qconvection = h(Tsurf − Tair ) (3) transfer from the zone surfaces, ṁi Cp (Tzi − Tz ) is the heat transfer due to
interzone air mixing, ṁinf Cp (T∞ − Tz ) is the heat transfer due to infil­
Where h is the convective heat transfer coefficient in relation to air tration of outside air.
movement, Tsurf is the surface temperature in contact with air, Tair is the The input parameters for the baseline dynamic thermal model of the
external or internal air temperature. case study building are given in Table 6. The building operation loads in
Heat transfer through radiation is given by Stefan-Boltzmann law as: terms of lighting power, occupancy and equipment were calculated
based on the actual building operations.
Qradiation = εσ (Tsurf )4 (4)
The building operating schedules were based on standard office

6
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 3. Architectural drawings of the case study building: (a) south-west façade, (b) typical floor layout, (c) mezzanine layout.

occupancy settings, however, adjusted to match real life scenarios in the were incorporated and compared to the baseline simulation:
course of model calibration. By means of observation and interviews
with the building occupants, the schedules were changed to match a low ▪ Option 1 – The south-west façade of the building was
occupancy and remote working scenario due to the aftereffects of the substituted with a single BIPV laminate glazing (BIPV curtain
COVID-19 lockdowns in Mauritius. Thus, the schedules for occupancy, wall)
cooling, lighting and equipment uses are illustrated in Fig. 5. ▪ Option 2 – The roof of the building was substituted with an
The physical modelling with the envelope properties of the baseline uninsulated roof BIPV membrane
building was then performed in DesignBuilder with proper orientations ▪ Option 3 – The roof of the building was substituted with an
of each wall; while the cooling load simulation of the building was insulated roof BIPV membrane
performed for the baseline building first, and then the following retrofits

7
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Table 2 Fig. 6 and Fig. 7, respectively.

Thermal properties of triple laminated BIPV glazing layers.
Material Thickness Thermal Specific heat Density 3.5.4. Building energy modelling
(mm) conductivity (W/ capacity (J/ (kg/m3) The building energy modelling, in terms of energy usage and PV
mK) kgK) electricity generation of each of the proposed BIPV typologies, was
Low iron 6 1.05 750 2500 performed in DesignBuilder to assess the impact of the BIPVs on the
glass building’s energy end-use. For the whole building energy assessment,
PVB foil 1.52 0.236 1360 1100
Float glass 3.2 1.05 750 2500
a-Si PV 0.2 1.8 882 2200
module Table 5
Thermal boundary conditions for FEA simulation.
Boundary condition parameter Value Source
Table 3 Radiative heat fluxes
Thermal properties of roof BIPV membrane layers. Triple laminated BIPV glazing
(a-Si)
Material Thickness Thermal Specific Density
Qsolar 1027.948 W/ Based on Section 3.5.1 and
(mm) conductivity heat (kg/m3)
m2 Table 4
(W/mK) capacity (J/
Qreflect 527.337 W/m2 Based on Section 3.5.1 and
kgK)
Table 4
ETFE 0.3 0.16 2303 1740 Qglass 82.236 W/m2 Based on Section 3.5.1 and
EVA 0.5 0.311 2090 960 Table 4
2
CIGS PV module 0.33 3.7 300 5770 QPVB 94.982 W/m Based on Section 3.5.1 and
TPT 0.3 0.15 1250 1200 Table 4
2
Butyl adhesive 1 0.24 1400 1200 Qfloat 28.322 W/m Based on Section 3.5.1 and
Nylon fastener 2.35 0.25 1600 1150 Table 4
PVC-P membrane 1.5 0.17 900 1390 Qelec 48.725 W/m2 Based on Section 3.5.1 and
Polyisocyanurate 25 0.025 1500 33.6 Table 4
(PIR) insulation QD-Tr 76.068 W/m2 Based on Section 3.5.1 and
Vapour barrier 1.2 0.38 1700 1460 Table 4
2
QPV 170.278 W/m Based on Section 3.5.1 and
Table 4
Roof BIPV membrane (CIGS)
Table 4 Qsolar 959.333 W/m2 Based on Section 3.5.1 and
Radiation properties of BIPV transparent layers. Table 4
Qreflect 80.124 W/m2 Based on Section 3.5.1 and
Material Transmissivity Reflectivity Absorptivity Emissivity
Table 4
2
ETFE 0.926 0.065 0.009 0.94 QETFE 8.634 W/m Based on Section 3.5.1 and
EVA 0.900 0.020 0.08 – Table 4
Low iron glass 0.84 0.08 0.08 – QEVA 71.067 W/m2 Based on Section 3.5.1 and
PVB 0.82 0.07 0.11 – Table 4
Float glass 0.88 0.08 0.04 – Qelec 159.249 W/m2 Based on Section 3.5.1 and
Table 4
2
QPV 640.259 W/m Based on Section 3.5.1 and
▪ Option 4 – The south-west façade of the building was partially Table 4
Convective heat transfer coefficient
substituted with a naturally ventilated double-skin façade
hext 13 W/m2K BS EN ISO 6946:2017 [64]
incorporating a single BIPV laminate glazing in the front layer hint-v 2.5 W/m2K BS EN ISO 6946:2017 [64]
for the first and second floors, while the ground floor façade hint-h 0.7 W/m2K BS EN ISO 6946:2017 [64]
was kept as the BIPV curtain wall only Air temperatures
Text 32.9 ◦ C Simulation condition
Tint 25 ◦ C Simulation condition
The baseline and retrofits modelled in DesignBuilder are shown in

Fig. 4. Thermal energy balance schematics: BIPV glazing (left), roof BIPV membrane (right).

8
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Table 6 circuit, Rs is the PV module series resistance.

Input parameters for dynamic simulation of the building. Fundamentally, the generated PV electrical power is given by “P =
Input parameter Value IV”, where it is required to obtain the current I and voltage V through
three known I-V points. The three points are the short circuit current
U-value of the external wall 1.916 W/m2K
U-value of the internal wall 1.501 W/m2K (Isc), the open-circuit voltage (Voc) and the maximum power output
U-value of the roof 2.204 W/m2K (Pmpp), while the corresponding temperature coefficients are also
U-value of the intermediate floor 2.036 W/m2K crucial. Table 7 shows the inputs for the BIPV module obtained from
U-value of the external window 1.772 W/m2K manufacturer documentations, which were used as the BIPV module
Air permeability 0.7 ACH
Cooling set point 24 ◦ C
inputs in the DesignBuilder model. In this study, 508 m2 of roof space
Occupancy density (ground floor) 0.082 person/m2 and 342 m2 of façade space were used for the roof and façade BIPV
Occupancy density (first floor) 0.043 person/m2 retrofits, respectively. Specifically, the roof retrofit consisted of 146 ×
Occupancy density (second floor) 0.013 person/m2 500 W and 17 × 200 W CIGS roof membrane modules, while the façade
Heat gain from occupant 150 W/person
retrofit consisted of 209 × 62 W and 19 × 32 W semi-transparent a-Si
Lighting power density (ground floor) 3.9 W/m2
Lighting power density (ground floor mezzanine) 5.1 W/m2 modules. The PV electricity generation was allowed to run for the whole
Lighting power density (first floor) 4.7 W/m2 year including weekends to assess its complete capability.
Lighting power density (first floor mezzanine) 6.1 W/m2
Lighting power density (second floor) 5.1 W/m2 4. Model calibration
Equipment power density (offices) 2.3 W/m2
Equipment power density (common areas) 2.7 W/m2
4.1. Energy model calibration

the services to each zone were simulated to imitate a full operation Building energy consumption, as the core research output reflecting
scenario of the case study building to investigate the energy impact. The thermal and energy performance of the case study building, was selected
inputs for lighting and equipment uses for the building are shown in for calibration to match real data of the building to gauge the model’s
Table 6. To meet the cooling load calculated in Eq. (5), the HVAC sys­ suitability as an evaluation tool for the BIPV retrofits. The calibrated
tems consisting of a DX split system were modelled based on the Eq. (6) model had several office spaces switched off to match real life operations
used in association with the DesignBuilder simulation [65]. during the year 2021 for which annual energy usages were available.
( ) Specifically, a baseline model of building energy consumption was first
Q̇sys = ṁsys Cp Tsup − Tz (6)
calibrated using actual metered energy consumption from the utility
Where ṁsys is the mass flow rate provided by the DX split system, Cp is provider. Since it was the second year of operation of the building,
the zone air specific heat, Tsup is the supply air temperature, Tz is the several offices were still unoccupied on the second floor, east wing of
zone mean air temperature. first floor and west wing of ground floor. Thus, the building services
The BIPV construction material feature of DesignBuilder was used to systems to those offices were switched off for the baseline calibration.
simulate the PV electricity generation following the equivalent one- The ASHRAE 14 [66] monthly acceptance indices – Mean Bias Error
diode model, which was calculated by using the following equation (MBE) and Cumulative Variation of Root Mean Squared Error (CVRMSE)
[65]: – were used to evaluate the energy model calibration. The acceptable
[ ( ) ] criteria for the MBE and CVRMSE are within ± 5% and less than 15%,
I = IL − Io exp
q
(V + IRs ) − 1 (7) respectively, and both are calculated as follows:
γkTPV ∑ NP
(Mi − Si )
Where I is the current in the one-diode equivalent circuit, IL is the PV MBE = i=1 ∑NP (8)
i=1 Mi
module photocurrent, Io is the diode reverse saturation current, q is the
elementary charge at 1.602 × 10-19C, γ is the empirical PV curve-fitting ∑NP
i=1 Mi
parameter, k is the Boltzmann’s constant at 1.381 × 10-23 J/K, TPV is the MP = (9)
NP
PV module temperature, V is the voltage in the one-diode equivalent

Fig. 5. Operating schedules for the case study building.

9
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 6. 3D view of the baseline model.

Fig. 7. 3D view of the BIPV retrofit options.

10
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Table 7 The model calibration process, as illustrated in Fig. 8, was used to get
I-V characteristics of the BIPV modules for the retrofits. a calibrated energy model that would meet the ASHRAE 14 acceptable
Parameter Semi-transparent a-Si CIGS roof membrane criteria [66]. In summary, major parameters affecting the accuracy of
BIPV module BIPV module the simulation results including weather data, physical properties of the
62 W 32 W 500 W 200 W building envelope and BIPV modules, internal gains and other opera­
Area (m2) 1.39 0.71 2.33 0.94 tional settings were fine-tuned to match the measured electrical utility
Rated power (W) 72.32 32 500 200 data with an acceptable level of accuracy. Since the HVAC system was a
Short circuit current, Isc (A) 1.15 1.15 9.07 8.82 DX split system with the indoor temperatures defined by the occupants,
Current at max output, Impp 1.04 0.93 8.03 8.27
(A)
it was assumed that the setpoints varied considerably over the year. To
Temperature coefficient of 0.000104 0.000104 0.00073 0.00071 tackle this issue two setpoints were used for the baseline energy model
Isc (A/◦ C) calibration, namely 25 ◦ C during summer months when temperatures
Open circuit voltage, Voc (V) 94 50 77.2 29.2 were already high and 22 ◦ C during winter months when outdoor
Voltage at max output, Vmpp 70.5 34 62.4 24.2
temperatures were relatively low.
(V)
Temperature coefficient of − 0.26 − 0.14 − 0.216 − 0.082 Fig. 9 shows the final iteration of the simulated energy consumption
Voc (A/◦ C) and the utility energy measurements on a monthly basis for a whole year
NOCT ambient temperature 20 20 20 20 along with the percentage deviation (based on the CVRMSE) between
(◦ C)
NOCT cell temperature (◦ C) 45 45 48 48

Table 8
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
̅ Results of acceptance indices for energy model calibration.
∑NP ((Mi − Si )2 )
i=1 NP
Indices Summer Winter Annual Acceptance Met acceptance
CVRMSE = (10) (Oct – (Apr – (Jan – criteria criteria?
MP
Mar) Sep) Dec)
Where Mi and Si are the respective measured and simulated data at
MBE − 3% 2% − 0.3% Within ± 5% Yes
instance “i”, P is the calculation time interval; NP is the number of values CVRMSE 6% 5% 1.1% ≤ 15% Yes
at interval P, MP is the average of the measured data.

Fig. 8. Process of model calibration method [67].

Fig. 9. Monthly simulated energy consumption versus measured energy consumption.

11
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Table 9
Results of finite element heat transfer model calibration.
Scenario RRMSE Excellent accuracy Good accuracy Fair accuracy Poor accuracy

BIPV glazing 0.014% RRMSE < 10% 10% ≤ RRMSE ≤ 20% 20% ≤ RRMSE ≤ 30% RRMSE > 30%
Roof BIPV membrane 0.023%

Fig. 10. Isometric view of temperature distribution on the laminated BIPV glazing (left), side view of maximum and minimum temperatures on the laminated BIPV
glazing (right).

Fig. 11. Close-up cross section views of temperature and heat distributions on the laminated BIPV glazing.

both energy results. It can be seen that most of the corresponding 4.2. Finite element heat transfer model calibration
simulated and measured results matched well to each other; the highest
deviation occurred in April, while the remaining months stayed within In addition to the building energy modelling, the finite element heat
the 15% threshold of the CVRMSE. transfer modelling was another crucial part of the study, which was also
Since two temperature setpoints being used in different seasons of calibrated (as per the calibration process shown in Fig. 8) while using a
the year, the calibration was further divided into a summer period, a different evaluation method. To calibrate the results from the FEA, the
winter period and an annual period. As can be seen in Table 8, all the PV cell temperature (Tc ) was selected as the calibration parameter.
monthly MBE and CVRMSE indices were well within the set limits for all Fundamentally, PV cell temperature is determined through the Nominal
the periods, while the annual indices showing the closest fit to measured Operating Cell Temperature (NOCT) model [68] as per Eq. (11):
data. Thus, the model was deemed calibrated.
Is
Tc = Ta + (NOCT − 20) • (11)
800

12
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 12. Temperature profile along cross section of laminated BIPV glazing.

Fig. 13. Cumulative frequency heat metrics for: (a) inner glazing surface, (b) a-Si PV module surface.

13
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 14. Temperature and heat distributions on uninsulated roof BIPV membrane: (a) isometric view of temperature distribution, (b) side view of maximum and
minimum temperatures, (c) close-up cross section view of temperature distribution, (d) close-up cross section view of heat distribution.

Where Ta is the ambient air temperature, Is is the incident solar ra­ emitting the heat to its surroundings to provoke the high temperatures
diation on the cell surface. on the surface. Fig. 11 presents a close-up of the cross section, which
Using the data given in Table 5 and Table 7, the PV cell temperatures shows heat concentration occurring on the photoactive side of the a-Si
for both the a-Si BIPV glazing and the CIGS roof BIPV membrane were PV module. In this case, the resultant heat flux was predominantly
calculated, where only the uninsulated membrane scenario was occurring from the PV module to the inner surface of the glazing with
considered for the roof BIPV modelling calibration due to the avail­ the highest heat fluxes happening through the gaps between the strips on
ability of the data. The PV cell temperatures obtained through the the edges of the PV module.
ANSYS simulation were then compared to the calculated PV cell tem­ To further verify that whether the resultant heat was dominated by
peratures using the measure of Relative Root Mean Square Error the PV module, a temperature profile along the cross section of the
(RRMSE), which is calculated as follows [69]: laminated BIPV glazing is presented in Fig. 12. It is found that the
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
∑n highest temperature occurred on the a-Si PV module itself, while if a gap
2
substitutes the PV module layer the highest temperature occurred at the
1
i=1 (xi − yi )
RRMSE = n ∑ (12)
n
i=1 (yi )
2
outer PVB layer. This indicates that the PV module was absorbing a
considerable amount of radiative heat, which was then being emitted to
Where yi is the simulated value, xi is the calculated value, n is the
its surrounding. However, both trends being dropped similarly after the
total number of data samples (that is, the total number of the PV cell
entire glazing to reach the average inner glazing temperature of 41.4 ◦ C.
temperature samples).
Furthermore, temperature variations on the laminated BIPV glazing
In summary, the RRMSE values for both scenarios of BIPV glazing
surfaces were evaluated for a better assessment of the impact of the BIPV
and roof BIPV membrane along with the accuracy bands [69] are
glazing on building cooling loads. Fig. 13 shows the cumulative fre­
tabulated in Table 9. As can be seen, both scenarios of the FEA simu­
quency distribution of the temperature and resultant heat flux on both
lation were within the excellent accuracy, thus the models were deemed
inner glass surface and a-Si module of the laminated glazing. It can be
calibrated.
observed that the heat flux had a close correlation with the temperature
of the nodes happening on the inner glass surface. The curves show that
5. Results and discussion
the 15th percentile marker amounting to 41.2 ◦ C (temperature) and 136
W/m2 (heat flux), which means that for 85% of the inner glass surface
5.1. Finite element heat transfer analysis
area the temperature varied between 41.2 ◦ C and 41.4 ◦ C, while the heat
flux remained between 136 W/m2 and 137.4 W/m2. According to local
This section evaluates the capability of the finite element heat
thermal discomfort definition given by ASHRAE 55 [70], these resultant
transfer analysis to predict the heat transfer characteristics of the triple
temperatures varying on the inner glass surface (as a “Warm Wall”) can
laminated BIPV glazing and both insulated and uninsulated versions of
significantly lead to local thermal discomfort. Thus, a considerable
the roof BIPV membranes.
cooling energy certainly will be consumed to maintain thermal comfort
in the building, which is discussed in the Section 5.2 of this paper. For
5.1.1. Heat transfer through triple laminated BIPV glazing
the a-Si PV module surface, both curves had opposing profiles below the
The results from the finite element heat transfer analysis for the BIPV
15th percentile and above the 98th percentile, however, the discrepancy
glazing are illustrated in Fig. 10, where the temperature distribution
accounted for an overall 17% of the surface only, while for 83% of the
over the whole test piece is shown. The results show that the a-Si PV
nodes both heat metrics stayed relatively constant between 42.8 ◦ C and
module in the interlayer had the highest temperatures at 43.3 ◦ C, while
43.3 ◦ C for the temperature and 170 W/m2 and 300 W/m2 for the heat
the lowest temperature occurred at the edge of the glazing which was
flux. Apparently, the higher resultant temperatures of PV module sur­
enclosed in a frame at 40.489 ◦ C. The distribution clearly shows that the
face, to a large extent, led to the “Warm Wall” affecting thermal per­
a-Si PV module was heating up from the effect of incident radiation and
formance of the entire laminated BIPV glazing.

14
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 15. Close-up cross section view of temperature and heat distributions on insulated roof BIPV membrane.

5.1.2. Heat transfer through roof BIPV membrane maximum heat occurred under the CIGS PV module but, the heat flux
Fig. 14 provides the results of the finite element heat transfer anal­ directed towards the edges of the component rather than through the
ysis for the uninsulated roof BIPV membrane, where the temperature other membrane layers. Fundamentally, the PV modules were process­
distributions of the test piece are shown. Similar to the laminated BIPV ing maximum heat by absorbing and emitting it to the surrounding
glazing, the uninsulated roof BIPV membrane had the highest temper­ layers; however, in comparison with the BIPV glazing which is trans­
ature on the CIGS PV module layer at 56.745 ◦ C and the lowest tem­ parent and transmits incident radiation, the heat transfer of the roof
perature along the edges of the membrane which were the furthest from BIPV membrane was reduced considerably due to the opacity of the TPT
the heated PV modules at 48.303 ◦ C, while average inner surface tem­ back layer. In this case, heat transfer occurred mostly through conduc­
perature at the vapour barrier was about 53.78 ◦ C. As shown by the tion underneath the TPT layer and onto the waterproofing layers.
close-up cross sections of the uninsulated membrane in Fig. 14, the Fig. 15 shows the cross section of the insulated roof BIPV membrane.

15
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 16. Temperature profile along cross section of: (a) uninsulated roof BIPV membrane, (b) insulated roof BIPV membrane.

The temperature of the membrane was seen to reduce considerably from surface temperature was at an average of 55.54 ◦ C, which was higher
57.185 ◦ C to 39.518 ◦ C due to the inclusion of the PIR insulation panel, than the initial photoactive side of the uninsulated roof BIPV membrane;
and the average inner surface temperature at the vapour barrier was this means thermal resistances of the under layers were too low to resist
42.3 ◦ C. Basically, most of the heat occurred and transferred on the the amount of heat transmitted from the PV modules (GIGS), and
topmost layers of the membrane (above the insulation panel). Similar to consequently heat transfer to the adjoining construction materials dur­
the uninsulated version, the heat flux was diffusing sideways to the ing a normal installation would be high. For the insulated scenario, it
edges above the waterproofing layers. can be seen that both profiles had very similar trends peaking at the PV
Fig. 16 shows both temperature profiles along the cross section of the modules (CIGS) and decreasing towards the inner surface temperature.
uninsulated and insulated roof BIPV membranes, through both the CIGS However, unlike the uninsulated membrane, the insulated one with PIR
PV module and the gap (without the CIGS PV module) within the panel had a significantly lower inner surface temperature at an average
membranes. For the uninsulated scenario, even though the temperatures of 43.45 ◦ C with an effective temperature drop around 12 ◦ C noted over
through the CIGS module was slightly higher, both profiles followed the cross section. This demonstrates that the advantage of the PIR
similar trends peaking at the PV module/gap layer and then gradually insulation panel in controlling heat transmission through the roof BIPV
decreased to inner surface temperature. The average difference between membrane.
the two trends amounted to only about 0.253 ◦ C. However, the inner Fig. 17 presents the cumulative temperature and heat flux

16
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 17. Cumulative thermal metrics for roof BIPV membrane inner surfaces: (a) uninsulated, (b) insulated.

Fig. 18. Peak thermal loads/gains for different design options.

distributions of the inner surfaces for both the insulated and uninsulated internal gain (which was not presented as this was not a design vari­
roof BIPV (CIGS) membranes. Both distributions show a relatively able). With the baseline peaking at 133 kW, Option 1 with the BIPV
steady heat flux, of which 22 W/m2 for the uninsulated scenario and 13 curtain wall had a slight increase of about 1 % to reach 135 kW. In
W/m2 for the insulated scenario; while the temperature distributions comparison, Options 2 and 3 with the roof BIPV membranes had a most
over the simulated nodes of the inner surfaces were within the range of effective reduction with 8 % at 119.85 kW for the uninsulated roof
49–55 ◦ C (6 ◦ C in difference) and 39.5–43.5 ◦ C (4 ◦ C in difference), membrane and 15 % at 112.02 kW for the insulated roof membrane,
respectively, for the uninsulated and insulated scenarios. This has while Option 4 with the BIPV-DSF had a slight reduction of 2 % to reach
further demonstrated that the PIR insulation panel was active in cooling 130 kW. This clearly shows that the roof BIPV membranes had the most
the roof BIPV membrane surface down, as well as restraining the sig­ impact in terms of thermal transmission and overall heat balance; while
nificant heat transmission along the surface. the poorer thermal performance of the BIPV façades was likely due to
the nature of the BIPV glazing, as higher thermal mass in the form of
solid walls were largely substituted by BIPV glazing with a relatively
5.2. Dynamic building performance analysis
higher heat transfer coefficient and SHGC.
The hourly variations of building heat balance for each BIPV retrofit
The building’s peak cooling load results for each of the BIPV retrofit
option are presented in Fig. 19. The heat gains are shown in the positive
options from the dynamic building simulations are shown in Fig. 18. The
domain of the X-axis, while the heat losses in the form of cooling energy
design options’ cooling loads were compared to that of the baseline, and
or heat transferred out of the zones are shown in the negative domain.
the cooling loads were referred to the corresponding heat gains from the
The hourly data was plotted against the outdoor dry bulb temperature
façade elements, the roof, and the direct solar radiations, as well as the

17
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 19. Hourly heat balance for different design options: (a) baseline model, (b) Option 1 – BIPV curtain wall, (c) Option 2 – uninsulated roof BIPV membrane, (d)
Option 3 – insulated roof BIPV membrane, (e) Option 4 – BIPV-DSF.

and indoor average air temperature resulting from the heat balance and heat from the internal zones was leaking outside, which was similar to
cooling process. In the baseline case, all the heat gains remained in the Option 2 that creating a lower heat accumulation and hence a lower
positive domain with solar heat transmissions from glazing and cooling load during occupancy.
conductive heat gains from roof had the biggest impacts on the total heat
gains, while the BIPV retrofitting cases had diverging heat balances.
The BIPV curtain wall of Option 1 had a greatly reduced solar gain 5.3. Building energy performance analysis
compared to the baseline model. Before 8 am and after 6 pm, the internal
zones of the building were losing heat to the outside environment The monthly energy consumption for each of the BIPV retrofits and
through the low thermal resistance of the BIPV glazing as shown by the the baseline building by considering all the office spaces were opera­
façade gain being in the negative domain at those specific hours. The tional are shown in Fig. 20. Basically, all the scenarios had the similar
BIPV-DSF of Option 4 balanced the effect of low thermal resistance of monthly variation trend in energy across the year. On a yearly basis, the
the BIPV glazing with its insulating air gap but did little to reduce on the baseline building had an energy consumption of 73,349 kWh, while the
total cooling load of the baseline building. retrofits of BIPV curtain wall (Option 1) and BIPV-DSF (Option 4) had
Even though Option 2 and Option 3 had roofs retrofitted in, there energy consumptions of 79,649 kWh and 74,627 kWh respectively,
was the direct solar heat gains from existing glazing that had the biggest which accounted for an increase of 9% and 2% over the baseline. On the
impact on the thermal loads. The roof heat gain was also higher than other hand, the roof BIPV retrofits show less annual energy consumption
baseline for the uninsulated roof BIPV membrane (Option 2), which was in comparison with the baseline, saw 69,583 kWh and 5% less for the un-
losing heat to the surroundings before 9.30 am accounting for a lower insulated roof BIPV membrane (Option 2) and 60,762 kWh and 17% less
heat accumulation in the building when occupancy start to increase, and for the insulated version (Option 3). Even though the energy savings and
HVAC air cooling requirements were decreased and thus resulted in a energy excess of the different retrofit options were relatively linear
lower cooling demand. Option 3 (insulated roof BIPV membrane) had a compared to the baseline, it can be noticed that the histograms of BIPV-
net improvement on baseline and between 5:30 am and 10:30 am the DSF (Option 4) show matching closer to the baseline especially from
June to December.

18
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 20. Monthly energy consumption for different design options.

By extracting the cooling energy consumption from all the design 5.16% relative to the baseline. As a result, the heat transmissions
options, the daily consumption distributions are shown in Fig. 21. The through the semi-transparent BIPV modules were apparent, but the
daily cooling energy trendlines shows that the BIPV facades required cooling effect of BIPV-DSF would offer a degree of mitigation on the
more cooling energy due to the higher heat gains, with an average in­ transmitted heat gains from the BIPV itself. By comparison, both roof
crease of 12% cooling energy for the BIPV curtain wall and an average of BIPV membrane options (Option 2 and Option 3) achieved significant
2% more cooling energy for the BIPV-DSF relative to the baseline. It also net energy savings of 160% and 172%, respectively; the higher energy
can be seen that the roof BIPV membrane options in overall had a better saving from Option 3 (insulated roof BIPV membrane) demonstrated the
energy performance by reducing the cooling energy consumption to effectiveness of the thermal insulation applied.
below the baseline case by an average of 7% and 18% for the uninsu­
lated and insulated roof BIPV membranes, respectively. 6. Conclusions
As per the results obtained from both the dynamic building perfor­
mance and the building energy performance simulations, it was found In this paper, a novel approach to numerically assessing the heat
that having a lower thermal transmittance on thermal masses of the transfer and energy performance of BIPV was evaluated. Different BIPV
building envelope had a better effect than having a lower solar radiation retrofits in terms of thermal and energy performance for a typical office
transmission. Limiting the conductive heat gains with better thermal building in Mauritius were studied through finite element heat transfer
insulating materials provided by the roof BIPV membranes offers a analysis, dynamic building performance analysis, as well as energy
better energy consumption factor than limiting the direct radiative heat performance analysis. Specifically, the BIPV retrofit options presented in
gains provided by the BIPV facades. this study include two BIPV façade typologies (BIPV curtain wall and
By taking into account the energy generated through the BIPVs, a net BIPV-DSF) and two roof BIPV membrane typologies (uninsulated and
resultant energy consumption for each BIPV retrofit option in compar­ insulated roof BIPV membranes). The results show that both BIPV facade
ison with the baseline case is shown in Fig. 22. It can be seen that both and roof BIPV membrane options had high thermal transmission factor
BIPV façade options (Option 1 and Option 4) had the monthly net energy with their inner surfaces heated up to above 40 ◦ C and hence created
consumption close to the baseline case, especially for the Option 1 (BIPV more convective heat transfer to their adjacent air (that is, the indoor
curtain wall), of which the monthly net energy consumption was higher air). However, the intermediate insulation panel (for the roof BIPV
than the baseline from January to September. On a yearly basis, the membrane) and the ventilated air gap (for the BIPV-DSF) showed a net
Option 1 had about 1.66% more net energy consumption than the reduction in the heat transfer.
baseline, while the Option 4 (BIPV-DSF) received a net energy saving of The roof BIPV membranes had a better thermal performance effect

19
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 21. Daily cooling energy consumption: (a) BIPV facades, (b) roof BIPV membrane.

on the building by reducing its cooling load by 8% for the uninsulated footprint and provide a benchmark for future and existing buildings to
version and 15% for the insulated version; by contrast, both BIPV façade adopt the roof BIPV membrane technology in light of their low carbon
options could not reduce the cooling load effectively, while the BIPV and energy efficiency performance for Mauritius and the similar climate
curtain wall even led to an increase of cooling load for 1%. The results zones.
indicate that BIPV facades have had a negative outcome with respect to
traditional construction materials for the region and climate of 7. Limitations and future research
Mauritius, but the roof BIPV membrane whether insulated or uninsu­
lated could help reduce the cooling load and mitigate overheating issue Since the proposed research methodology was based solely on
for the building. Furthermore, the energy analysis shows a similar simulation, there was always an uncertainty factor to the outcomes of
outcome to the cooling load analysis with both BIPV façade options such numerical calculations although model calibration has been used to
adding more energy consumption to the building (as there was 1.66% mitigate as much as possible the uncertainty and major deviations in the
more annual energy consumption from BIPV curtain wall option), which results. Future research should look at the actual performance of both
could not be compensated by the PV electric power generated. Although BIPV façade and roof options through on-site experiments. Even though
the BIPV-DSF shows more promising results by reducing the net energy the roof scenario of this research was based on a roof BIPV membrane of
consumption (received a net energy saving of 5.16%), it generates too 500 W, it should be noted that in real life applications, smaller panels
little electric power to acknowledge its viability as a low energy solu­ would have been chosen to accommodate the actual energy consump­
tion. On the other hand, both uninsulated and insulated roof BIPV tion of the building instead of choosing the highest energy generating
membrane options achieved the annual energy saving for the building capacity system. Moreover, this research optimised the available space
by 160% and 172%, respectively. This net positive energy balance has a without considering actual usage of the roof space. Thus, future research
strong potential towards making the building reach a net zero carbon should take into account the effective roof space following the building

20
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. 22. Net resultant energy consumption for baseline and BIPV retrofits.

Fig. A1. Convergence between 4 mm and 10 mm mesh sizing factors for FEA.

CRediT authorship contribution statement

Table A1
Mesh sizing quality metrics. Hamza Jhumka: Methodology, Data curation, Formal analysis,
Finite element Sizing Nodes Elements Aspect Skewness Validation, Visualization, Writing – original draft. Siliang Yang:
model ratio Conceptualization, Methodology, Supervision, Validation, Visualiza­
Triple laminated 4 891,165 348,785 8.15 ± 0.769 ± tion, Investigation, Formal analysis, Writing – review & editing. Chris­
glazing 6.38 0.295 topher Gorse: Formal analysis, Writing – review & editing. Sara
Roof membrane 5 914,498 313,355 8.34 ± 0.66 ± Wilkinson: Formal analysis, Writing – review & editing. Rebecca
6.57 0.314
Yang: Writing – review & editing. Bao-Jie He: Writing – review &
editing. Deo Prasad: Writing – review & editing. Francesco Fiorito:
management plan in reality. On the other hand, this research focused on Validation, Writing – review & editing.
an existing office building with pre-set orientations with respect to the
zenith and azimuth angles of the sun. In general, a north orientation
provides better PV efficiencies in the Southern Hemisphere, while the Declaration of Competing Interest
office building in this study has a south-west orientation. This may
impact on any assessment of future constructions based on this research. 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.

21
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. A2. Mesh refinement – triple laminated BIPV glazing.

Fig. A3. Mesh refinement – roof BIPV membrane.

Data availability materials and their individual components being flat and having regular
shapes, a hexahedral dominant meshing type was used for the FEA of
Data will be made available on request. heat transfer study. To limit the processing time, smaller sections of the
BIPV modules were extracted for analysis with the triple laminated BIPV
Appendix glazing being reduced to a 300 mm by 320 mm section, while the roof
BIPV membrane being reduced to a 425 mm by 470 mm section.
Appendix – Mesh refinement A mesh refinement was performed by reducing the mesh sizes to
produce finer grids. The resultant heat emission from the inner surfaces
Meshing is a process of sub-dividing the 3D geometry of the analysed of both FEA (BIPV glazing and roof membrane) was used as the metric to
object into a finite number of elements and nodes. With both BIPV determine the optimum mesh sizing and proper results convergence.

22
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

Fig. A1 shows the convergence of both the BIPV glazing and roof BIPV [27] A. Ghosh, N. Sarmah, S. Sundaram, T.K. Mallick, Numerical studies of thermal
comfort for semi-transparent building integrated photovoltaic (BIPV)-vacuum
membrane FEA between a 4 mm and 10 mm mesh sizing factor for the
glazing system, Solar Energy 190 (2019) 608–616.
finite elements. A 4 mm mesh sizing factor was used for the triple [28] F. Roberts, S. Yang, H.u. Du, R. Yang, Effect of semi-transparent a-Si PV glazing
laminated glazing and a 5 mm mesh sizing factor was used for the roof within double-skin façades on visual and energy performances under the UK
membrane with their corresponding mesh statistics and quality metrics climate condition, Renewable Energy 207 (2023) 601–610.
[29] P. Pal, et al., Pre-feasibility analysis and performance assessment of solar
summarised in Table A1. The PV layer was thinner than the rest of the photovoltaic (PV) modules for the application of renewable power generation,
materials encapsulating them, which was further refined with a mesh Materials Today: Proceedings 39 (2021) 1813–1819.
sizing factor of 2 to properly analyse the heat transitions occurring [30] H.M. Maghrabie, M.A. Abdelkareem, A.H. Al-Alami, M. Ramadan, E. Mushtaha,
T. Wilberforce, A.G. Olabi, State-of-the-Art technologies for Building-Integrated
around the BIPV modules. Fig. A2 and Fig. A3 show the mesh generated photovoltaic systems, Buildings 11 (9) (2021) 383.
for both BIPV materials using Ansys Mechanical, respectively. [31] F. Hidayanti, The effect of monocrystalline and polycrystalline material structure
on solar cell performance, International Journal of Emerging Trends in Engineering
Research 8 (7) (2020) 3420–3427.
References [32] K. Xiao, X. Wu, X. Song, J. Yuan, W. Bai, C. Wu, C. Huang, Study on performance
degradation and damage modes of thin-film photovoltaic cell subjected to particle
[1] IPCC. Climate change widespread, rapid, and intensifying. 2022 [cited 2022 27 June]; impact, Scientific Reports 11 (1) (2021).
Available from: https://www.ipcc.ch/2021/08/09/ar6-wg1-20210809-pr/. [33] S. Moghadamzadeh, I.M. Hossain, T. Duong, S. Gharibzadeh, T. Abzieher, H. Pham,
[2] Tsang, D.C. and L. Wang, Low Carbon Stabilization and Solidification of Hazardous H. Hu, P. Fassl, U. Lemmer, B.A. Nejand, U.W. Paetzold, Triple-cation low-bandgap
Wastes. 2021: Elsevier. perovskite thin-films for high-efficiency four-terminal all-perovskite tandem solar
[3] M. Diesendorf, H. Saddler, Australia’s polluting power: Coal-fired electricity and its cells, Journal of Materials Chemistry A 8 (46) (2020) 24608–24619.
impact on global warming, World Wide Fund, 2003. [34] S. Attia, S. Bertrand, M. Cuchet, S. Yang, A. Tabadkani, Comparison of thermal
[4] R. Hepple, H.u. Du, H. Feng, S. Shan, S. Yang, Sustainability and carbon neutrality in energy saving potential and overheating risk of four adaptive façade technologies
UK’s district heating: A review and analysis. e-Prime - Advances in Electrical in office buildings, Sustainability 14 (10) (2022) 6106.
Engineering, Electronics and Energy 4 (2023) 100133. [35] S. Yang, F. Fiorito, A. Sproul, D. Prasad, Optimising design parameters of a
[5] K. Ahmed Ali, M.I. Ahmad, Y. Yusup, Issues, impacts, and mitigations of carbon Building-Integrated photovoltaic Double-Skin facade in different climate zones in
dioxide emissions in the building sector, Sustainability 12 (18) (2020) 7427. australia, Buildings 13 (4) (2023) 1096.
[6] R. Hepple, S. Yang, S. Khattak, Z.i. Qian, D. Prasad, Comparative analysis of CIBSE [36] Pester, S., Building-integrated photovoltaic systems: Challenges and opportunities for
admittance and ASHRAE radiant time series cooling load models, CivilEng 3 (2) manufacturers and specifiers. Information paper. 2012, Bracknell: BRE Press.
(2022) 468–479. [37] J. Ramanujam, D.M. Bishop, T.K. Todorov, O. Gunawan, J. Rath, R. Nekovei,
[7] Economic Development Board Mauritius, Smart City Scheme Guidelines. 2020, E. Artegiani, A. Romeo, Flexible CIGS, CdTe and a-Si: H based thin film solar cells:
Economic Development Board. A review, Progress in Materials Science 110 (2020) 100619.
[8] United Nations Framework Convention on Climate Change (UNFCCC). Mauritius - [38] Y. Yin, W. Chen, J. Hu, B. Zhao, X. Huang, Photothermal-structural-fluid behaviors
High-level Segment Statement COP 26. 2021 [cited 2022 22 March]; Available from: of PV-ETFE cushion roof in summer: Numerical analysis using three-dimensional
https://unfccc.int/sites/default/files/resource/MAURITIUS_cop26cmp16cma3_ multiphysics model, Energy and Buildings 228 (2020) 110448.
HLS_EN.pdf. [39] M.G.S. Conde, K. Shanks, Evaluation of available building integrated photovoltaic
[9] Board of Investment Mauritius, Smart Mauritius - Live, Invest, Work, Play. 2016, (BIPV) systems and their impact when used in commercial buildings in the united
Board of Investment. arab emirates, international Journal of Sustainable Energy Development (IJSED) 7
[10] L.-E.-E.-D.-N.-D. Moka certification to recognise Moka Smart City’s commitment to (1) (2019) 344–356.
sustainability, cited 2022 22 May https://www.moka.mu/en/blog/smart- [40] S. Yang, F. Fiorito, D. Prasad, A. Sproul, A. Cannavale, A sensitivity analysis of design
solutions/leed-nd-certification-moka-sustainability/ 2022 Available from. parameters of BIPV/T-DSF in relation to building energy and thermal comfort
[11] A. Khoodaruth, V. Oree, M.K. Elahee, W.W. Clark, Exploring options for a 100% performances. Journal of Building, Engineering 41 (2021) 102426.
renewable energy system in mauritius by 2050, Utilities Policy 44 (2017) 38–49. [41] X. Li, J. Peng, Y. Tan, Y. He, B. Li, X. Ju, J. Ji, S. Zhang, N. Li, Y. Chen, Optimal
[12] A. Chel, G. Kaushik, Renewable energy technologies for sustainable development design of inhomogeneous semi-transparent photovoltaic windows based on
of energy efficient building, Alexandria Engineering Journal 57 (2) (2018) daylight performance and visual characters, Energy and Buildings 283 (2023)
655–669. 112808.
[13] United Nations Environment Programme (UNEP), Accelerating the [42] P. Boyce, N. Eklund, S. Mangum, C. Saalfield, L. Tang, Minimum acceptable
transformational shift to a low-carbon economy in the republic of mauritius, transmittance of glazing, International Journal of Lighting Research and
United Nations Environment Programme, 2020. Technology 27 (3) (1995) 145–152.
[14] Y.K. Ramgolam, K.M.S. Soyjaudah, Unveiling the solar resource potential for [43] S. Yang, A. Cannavale, D. Prasad, A. Sproul, F. Fiorito, Numerical simulation study
photovoltaic applications in mauritius, Renewable Energy 77 (2015) 94–100. of BIPV/T double-skin facade for various climate zones in australia: Effects on
[15] B. Ghaleb, M. Asif, Application of solar PV in commercial buildings: Utilizability of indoor thermal comfort, Building Simulation 12 (1) (2019) 51–67.
rooftops, Energy and Buildings 257 (2022) 111774. [44] H. Sun, C.K. Heng, S.E.R. Tay, T. Chen, T. Reindl, Comprehensive feasibility
[16] R. Trattnig, G. Cattaneo, Y. Voronko, G.C. Eder, D. Moor, F. Jamschek, assessment of building integrated photovoltaics (BIPV) on building surfaces in
T. Buchsteiner, Smart glass coatings for innovative BIPV solutions, Sustainability high-density urban environments, Solar Energy 225 (2021) 734–746.
13 (22) (2021) 12775. [45] N. Rathore, N.L. Panwar, F. Yettou, A. Gama, A comprehensive review of different
[17] S. Yang, et al., Study of building integrated photovoltaic/thermal double-skin facade for types of solar photovoltaic cells and their applications, International Journal of
commercial buildings in sydney, australia, in Final conference of COST TU1403 Ambient Energy 42 (10) (2021) 1200–1217.
“Adaptive facades network”, Lucerne, Switzerland, 2018. [46] S. Yang, A. Cannavale, A. Di Carlo, D. Prasad, A. Sproul, F. Fiorito, Performance
[18] E. Biyik, M. Araz, A. Hepbasli, M. Shahrestani, R. Yao, L.i. Shao, E. Essah, A. assessment of BIPV/T double-skin façade for various climate zones in australia:
C. Oliveira, T. del Caño, E. Rico, J.L. Lechón, L. Andrade, A. Mendes, Y.B. Atlı, Effects on energy consumption, Solar Energy 199 (2020) 377–399.
A key review of building integrated photovoltaic (BIPV) systems, Engineering [47] Yang, S., et al., Studies on Optimal Application of Building-Integrated Photovoltaic/
Science and Technology, an International Journal 20 (3) (2017) 833–858. Thermal Facade for Commercial Buildings in Australia, in Proceedings of SWC2017/
[19] G. Yu, H. Yang, D. Luo, X.u. Cheng, M.K. Ansah, A review on developments and SHC2017. 2017. p. 1-10.
researches of building integrated photovoltaic (BIPV) windows and shading blinds, [48] S. Yang, et al., Numerical simulation modelling of Building-Integrated photovoltaic
Renewable and Sustainable Energy Reviews 149 (2021) 111355. Double-Skin facades, in: F. Bulnes, J.P. Hessling (Eds.), Recent Advances in
[20] S. You, H. Yang, The potential electricity generating capacity of BIPV in hong kong. Numerical simulations, IntechOpen, London, United Kingdom, 2021, pp. 61–75.
Conference Record of the Twenty Sixth IEEE Photovoltaic Specialists Conference- [49] M.A. Ekoe a Akata, D. Njomo, B. Mempouo, the effect of building integrated
1997, IEEE, 1997. photovoltaic system (Bipvs) on indoor air temperatures and humidity (Iath) in the
[21] A.K. Shukla, K. Sudhakar, P. Baredar, Recent advancement in BIPV product tropical region of cameroon, Future Cities and Environment 1 (0) (2015) 1.
technologies: A review, Energy and Buildings 140 (2017) 188–195. [50] H.-M. Liu, C.-H. Young, D.-J. Horng, Y.-C. Shiue, S.-K. Lee, Improving the
[22] G. Ban-Weiss, C. Wray, W. Delp, P. Ly, H. Akbari, R. Levinson, Electricity performance of a semitransparent BIPV by using High-Reflectivity heat insulation
production and cooling energy savings from installation of a building-integrated film, International Journal of Photoenergy 2016 (2016) 1–15.
photovoltaic roof on an office building, Energy and Buildings 56 (2013) 210–220. [51] J. Han, L. Lu, J. Peng, H. Yang, Performance of ventilated double-sided PV façade
[23] R.A. Agathokleous, S.A. Kalogirou, Double skin facades (DSF) and building compared with conventional clear glass façade, Energy and Buildings 56 (2013)
integrated photovoltaics (BIPV): A review of configurations and heat transfer 204–209.
characteristics, Renewable Energy 89 (2016) 743–756. [52] D. Sooben, N. Purohit, R. Mohee, F. Meunier, M.S. Dasgupta, R744 refrigeration as
[24] de Wild-Scholten, M., et al. A cost and environmental impact comparison of grid- an alternative for the supermarket sector in small tropical island developing states:
connected rooftop and ground-based PV systems. in 21st European Photovoltaic Solar The case of mauritius, International Journal of Refrigeration 103 (2019) 264–273.
Energy Conference. 2006. Dresden Germany. [53] K. Benis, I. Turan, C. Reinhart, P. Ferrão, Putting rooftops to use – A Cost-Benefit
[25] R. Rawlings, Capturing solar energy. CIBSE Knowledge Series: KS15, Chartered analysis of food production vs. energy generation under mediterranean climates,
Institution of Building Services Engineers, 2010. Cities 78 (2018) 166–179.
[26] A. Ghosh, Potential of building integrated and attached/applied photovoltaic [54] Stolarski, T., Y. Nakasone, and S. Yoshimoto, Engineering analysis with ANSYS
(BIPV/BAPV) for adaptive less energy-hungry building’s skin: A comprehensive software. 2018: Butterworth-Heinemann.
review, Journal of Cleaner Production 276 (2020) 123343.

23
H. Jhumka et al. Energy & Buildings 298 (2023) 113540

[55] S. Kumar Laha, P. Kumar Sadhu, A. Ganguly, A. Kumar Naskar, A comparative [62] M. Baharvand, M. Hamdan, M. Abdul, DesignBuilder verification and validation for
study on thermal performance of a 3-D model based solar photovoltaic panel indoor natural ventilation, Journal of Basic and Applied Scientific Research
through finite element analysis, Ain Shams Engineering Journal 13 (2) (2022) (JBASR) 3 (4) (2013) 8.
101533. [63] R.W. Lewis, P. Nithiarasu, K.N. Seetharamu (Eds.), Fundamentals of the Finite
[56] Panchenko, V., S. Chirskiy, and V.V. Kharchenko, Application of the software system Element Method for Heat and Fluid Flow, Wiley, 2004.
of finite element analysis for the simulation and design optimization of solar photovoltaic [64] British Standards Institution, BS EN ISO 6946: Building components and building
thermal modules, in Handbook of Research on Smart Computing for Renewable Energy elements - thermal resistance and thermal transmittance - calculation method,
and Agro-Engineering. 2020, IGI Global. p. 106-131. British Standards Institution, London, 2017.
[57] Y. Lee, A.A. Tay, Finite element thermal analysis of a solar photovoltaic module, [65] U.S. Department of Energy, EnergyPlus Version 9.6.0 Documentation in Engineering
Energy Procedia 15 (2012) 413–420. Reference. 2021.
[58] Garg, V., J. Mathur, and A. Bhatia, Building Energy Simulation: A Workbook Using [66] ASHRAE, Guideline 14, in Measurement of Energy and Demand Savings. 2002,
Designbuilder™. 2020: CRC Press. American Society of Heating, Refrigerating and Air-Conditioning Engineers:
[59] Manke, P., Y.K. Garg, and V.M. Das. Energy simulation tools and CAD interoperability: Atlanta, Georgia.
A critical review. in 2013 International Conference on Energy Efficient Technologies for [67] Yang, S., Studies on the Performances of Building Integrated Photovoltaic/Thermal
Sustainability. 2013. IEEE. Double-Skin Facade for Commercial Buildings in Australia, in School of Built
[60] G. Elshafei, A. Negm, M. Bady, M. Suzuki, M.G. Ibrahim, Numerical and Environment. 2020, University of New South Wales: Sydney. p. 299.
experimental investigations of the impacts of window parameters on indoor [68] V. Sun, A. Asanakham, T. Deethayat, T. Kiatsiriroat, A new method for evaluating
natural ventilation in a residential building, Energy and Buildings 141 (2017) nominal operating cell temperature (NOCT) of unglazed photovoltaic thermal
321–332. module, Energy Reports 6 (2020) 1029–1042.
[61] A. Fathalian, H. Kargarsharifabad, Actual validation of energy simulation and [69] Jadon, A., A. Patil, and S. Jadon, A Comprehensive Survey of Regression Based Loss
investigation of energy management strategies (Case study: An office building in Functions for Time Series Forecasting. arXiv preprint arXiv:2211.02989, 2022.
semnan, iran), Case Studies in Thermal Engineering 12 (2018) 510–516. [70] ASHRAE, Standard 55, in Thermal Environmental Conditions for Human Occupancy.
2017, American Society of Heating, Refrigerating and Air-Conditioning Engineers:
Atlanta, Georgia.

24