Solar Energy 140 (2016) 151–161

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Heat transfer studies of photovoltaic panel coupled with phase change

material
Karunesh Kant a,⇑, A. Shukla a, Atul Sharma a, Pascal Henry Biwole b,c
a
Non-Conventional Energy Laboratory, Rajiv Gandhi Institute of Petroleum Technology, Raebareli, UP, India
b
University of the Côte d’Azur, CNRS, J.A. Dieudonné Laboratory, France
c
Mines Paris Tech, PSL Research University, Centre for Processes, Renewable Energies and Energy Systems, Sophia Antipolis, France

a r t i c l e i n f o a b s t r a c t

Article history: The higher operating temperature of photovoltaic panels (above the standard operating temperature,
Received 28 August 2016 usually 25 °C) adversely affects the panel’s efficiency. PV panel coupled with phase change materials
Received in revised form 25 October 2016 (PCM) could be a feasible solution due to the higher energy storage density of such materials.
Accepted 2 November 2016
However, heat transfer studies of PCM combined with PV panel are quite intricate due to variable ambi-
ent conditions and changing thermo-physical property of the material with phase change. In the present
study, detailed heat transfer study of the PV panel coupled with PCM has been performed. The compu-
Keywords:
tational fluid dynamic study of the PV module coupled with PCM is carried out in which essential heat
Photovoltaic
Thermal regulation
transfer mechanisms between PV module to PCM and environment have been accounted for.
PV/PCV Additionally, its effect on power output has been investigated to see the variation of PV module operating
Phase change materials temperature. This study clearly suggests that for realistic simulation of heat and mass transfer studies of
PV panel attached with PCM, it is very important to consider following effects into account: convection
effect within melted PCM, the velocity of wind, and angle of inclination of PV panel. The results from the
study demonstrate the ability and importance of such model to reasonably simulate the thermal charac-
teristics of PV panels coupled with PCM. The maximum panel operating temperature with conduction
and convection effect is found to be 54.90 °C and 58.5 °C when convection mode in melted PCM (only
conduction mode) is not considered. It has also been shown that the higher wind velocity and tilt angle
leads to lower operating temperature of PV panels.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction studies have proven that the higher operational temperature of PV

module significantly decreases the power output (Armstrong and
The efficiency of photovoltaic (PV) panel depends on certain Hurley, 2010). Therefore appropriate cooling technologies must
physical parameters, such as solar radiation intensity falling on be used for enhancing the electrical efficiency of panels. The cool-
PV surface, PV panel operating temperature, heat loss from panel ing methodologies could be of two types i.e. active (methodologies
surface, and its material technology. For a typical PV panel, 5– which directly reduce the temperature of PV panel) and passive
25% radiated solar energy on PV panel front surface is transformed (which work as a heat sink and absorb the extra heat from PV
into electricity (Kant et al., 2016), and remaining is transformed panel). Due to higher operating and maintenance charge of active
into heat (Atkin and Farid, 2015; Malvi et al., 2011; Tonui and cooling technologies, the passive method of cooling of the photo-
Tripanagnostopoulos, 2007) that leads to enhance the module tem- voltaic panel is a preferred technique to maintain the PV module
perature. The electrical productivity of the solar photovoltaic cell is temperature at a level consistent with higher efficiency. The PCM
reduced at a higher temperature (Skoplaki and Palyvos, 2009), technique is a passive temperature regulation technique of PV
owing to increased internal charge carrier recombination rates modules which require less operational cost and has a higher
(Zhao et al., 2015; Sohel et al., 2014), which consequently results energy density (Hasan et al., 2010).
in a drop in the magnitude of maximum power production. Several In recent past, researchers have attempted to develop thermal
models to estimate the PV system’s operating temperature consid-
ering finer details of heat transfer. Jones and Underwood (2001)
⇑ Corresponding author. modeled heat transfer characteristics of the PV considering energy
E-mail addresses: [email protected] (K. Kant), amritanshu.shukla@gmail. conservation of PV module and its components. The methods of
com (A. Shukla).

http://dx.doi.org/10.1016/j.solener.2016.11.006
0038-092X/Ó 2016 Elsevier Ltd. All rights reserved.
152 K. Kant et al. / Solar Energy 140 (2016) 151–161

Nomenclature

TPV operating temperature of PV panel (°C) Pr Prandtl number

qrad radiative heat loss to environment (W/m2) Re Reynolds number
qconv convective heat loss to the environment (W/m2) Pm maximum Power output (W)
Pout power output of PV panel (W) Isc short circuit current (A)
qs radiation absorbed by PV panels (W/m2) Im current at maximum power (A)
a absorptivity of glass Vm voltage at maximum power (V)
G global solar radiation (W/m2) Voc open circuit voltage (V)
qr radiative heat loss to ground (W/m2) g efficiency of PV panel
e emissivity of glass surface Tref temperature at STC (°C)
F view factor bref temperature coefficient of PV panel
T temperature (°C) h angle of Elevation (°)
qsky radiative heat loss to the sky (W/m2) d angle of Declination angle (°)
Tamb ambient temperature (°C) w latitude (°)
U angle of inclination of PV panel from vertical (°) n no of days in year
qn natural convection heat loss (W/m2) Ghorizontal solar radiation on horizontal surface (W/m2)
hfree heat transfer coefficient due to natural convection Gmodule radiation incident on a tilted PV panel (W/m2)
(W/(m2 K)) Gincident solar radiation perpendicular to the sun (W/m2)
k thermal conductivity of air (W/(mK)) k thermal conductivity (W m
1 K
1)
L height of PV panels (m) Lf latent heat (J kg
1)
Ra Raleigh number P pressure (Pa)
l dynamic viscosity of air (Ns/m2) t time (s)
Cp specific heat (J/(kg K)) u velocity (m s
1)
ap thermal expansion coefficient of air (1/°C) v velocity in y direction (m s
1)
g acceleration due to gravity (m/s2) DT transition Temperature (K)
q density of air (kg/m3) b thermal expansion coefficient of PCM
qf heat loss due to forced convection (W/m2) r Stefan–Boltzmann constant: r = 5.67E
8 W/(m2 K4)
hforced heat transfer coefficient due to forced convection
(W/(m2K))

energy transfer from PV module to surroundings and power pro- of materials before applying with PV panels. Later, microencapsu-
ductivity were theoretically modeled which involved: long and lated PCMs were tested to regulate the temperature of BIPV by Ho
short wave radiation, heat loss due to convection over the panel et al. (2012). Maiti et al. (2011) performed a study on V trough PV
front surface and solar energy transformed into electricity. An panels (using solar reflectors to increase the intensity of radiation)
exergy study was carried out to forecast the thermal behaviour to prevent the temperature rising using PCM. Malvi et al. (2011)
of PV module by Duran Sahin et al. (2007). The analysis was presented generic joint photovoltaic (PV) thermal model incorpo-
applied to PV panels and its constituents for the estimation of rating PCM. Kibria et al. (2016) forecasted the thermal response
exergy losses, flows, and thus the effectiveness of PV module was of BIPV coupled with PCM with the developed thermal model.
estimated. Armstrong and Hurley (2010) established a mathemat- However, in their study, the outcome of wind velocity, convection
ical model for the study of PV module operating temperature, effect in melted PCM, and the consequence on power output was
incorporating weather conditions, material properties of compo- not considered. Huang et al. (2006) and Huang et al. (2004) per-
nents and the mounting methods. The model’s results were found formed a study of PCM filled in an aluminum container having a
in fair agreement with the experimental measurement of PV coating of solar radiation absorbing material on its front surface
panel’s temperature having adjustable wind velocities. to mimic a PV panel. Numerical simulation was performed using
Fontenault and Gutierrez-Miravete (2012) carried out a study of finite volume analysis to evaluate the performance of BIPV. Similar
numerical simulation of a hybrid PV thermal system by the consid- to the experiment described above, Hasan et al. (2010) got 10 °C
eration of FEM techniques using commercial package of Comsol temperature drop for 6 h at a heat flux of 415 W/m2, using a
Multiphysics by taking constant heat flux (1000 W/m2) and water eutectic combination of Capric and Lauric acid in an aluminum
as coolant and it was investigated that higher competence of the container. Lin and Ma (2016) developed an innovative Taguchi-
combined PV thermal system was achieved when using greater Fibonacci technique to optimize buildings with PCM and
flow rates of the coolant behind the panel. Ma et al. (2015), Islam air-based PVT collector’s design. Lin et al. (2016) investigated the
et al. (2016) and Browne et al. (2015) carried out an extensive performance and optimized buildings integrated PCM and PVT.
review on the thermal management of photovoltaic for enhancing Thermal models developed so far for PV/PCM systems, have
electrical efficiency. These reviews pointed that the use of PCM received wide interest, yet these require further modification,
could be a suitable option for passive thermal control of the photo- which could include convection effect within the melted PCM,
voltaic system. Stropnik and Stritih (2016) simulated the PV mod- fusion and solidification of the PCM, heat loss to the environment
ule coupled with PCM in TRNSYS software and it was investigated from PV panel’s front and back surface (by natural as well as forced
that the power output increases up to 7.3% annually. convection), heat loss due to radiation, variable climatic conditions
Hasan et al. (2014) studied the possibility of PCM for enhancing and the effect of the PV panel’s inclination angle on the heat loss
the electrical efficiency of the photovoltaic module in depth. Five mechanisms. In the present thermal model, these improvements
different PCMs (including paraffin waxes, salt hydrates, and combi- have been considered to study the thermal response of a PV mod-
nations of fatty acids) were characterised using DSC (Differential ule coupled with PCM under varying atmospheric conditions. This
scanning Calorimeter), to measure the phase transition properties is accomplished by considering the influence of thermal properties
 K. Kant et al. / Solar Energy 140 (2016) 151–161 153

of PCM, PV module’s materials layers and further investigating the and additional heat releases from the PV modules also as thermal
heat losses from the PV module’s surface to the environment along energy by means of different mode heat transfers techniques i.e.
with the heat, mass and momentum transfer in the PCM. This ther- convection and radiation, which in turn enhances the panel tem-
mal model thus incorporates some important factors which were perature. To analyse the PV panels, following assumptions have
not incorporated hitherto. been prepared concerning the conceptual PV panel structure,
The thermodynamic study of PV panel coupled with PCM has atmospheric conditions, and other factors.
been performed using finite element analysis approach with Com-
sol Multiphysics 5.0 version software (https://www.comsol.co.in/).  Properties of every layer in PV modules are isotropic and
The thermal model with and without PCM is validated with previ- homogeneous.
ous experimental studies Park et al. (2014) and Huang et al.(2006)  The radiation falling on panel’s front surface is equally
respectively. distributed.
 No dust or another agent is deposited on the panels surface that
affects the absorptivity of the panel
2. Simulation model and boundary conditions
 The melted PCM is Newtonian and incompressible.
 The flow due to the melting of PCM is laminar, and radiation
Typically, a PV module is composed of five layers as shown in
and three-dimensional convection effects of melted PCM are
Fig. 1 (Kant et al., 2016). The properties of layers used in the cur-
negligible.
rent study are given in Table 1. The PV system is assumed to be
 Two-dimensional convection and conduction modes of heat
fixed in a metal casing and the effects of the metal casing are not
transfer have been considered in the melted PCM.
incorporated in the model, since its lower surface area pertaining
to the panel surface area has an insignificant influence on the PV
The variation in PV module temperature is calculated by consid-
module operating temperature variation (Jones and Underwood,
ering the heat transfer taking place from module to the environ-
2001; Kant et al., 2016). The height of PV/PCM system is taken
ment and energy absorbed by PCM. The heat transfer paths to
for present study is 10 cm. The thickness of PCM layer is 2 cm filled
and from the PV/PCM system is shown in Fig. 1.
in 2 mm thick aluminum container. The commercial PCM taken in
this work is RT-35 and its properties are given in Table 2
3.1. Within solid parts
(Rubitherm GmbH, 2016).

In solid parts, i.e. PV panel’s layers, and aluminum container,

3. Mathematical formulations for the simulation work the heat transfer is governed by conduction mode only, therefore,
the heat transfer diffusion equation applied to the solid portions of
The processes which govern the PV module’s operating temper- the system. The diffusion equation is given by Eq. (1).
ature are quite intricate. This includes interior processes occurring
DT
in the semiconductor material during the bombardment of pho- qC p þ r:ð
krTÞ ¼ E þ tw ð1Þ
tons onto the solar cell, which cause the production of electricity Dt

Fig. 1. Physical model and path of heat transfer.

154 K. Kant et al. / Solar Energy 140 (2016) 151–161

Table 1
PV panel material properties (Kant et al., 2016).

PV layers PV panel layer properties

Density (q) (kg/m3) Specific heat (Cp) (J/kg K) Thickness (dm) (m) Thermal conductivity (k) (W/mK)
Glass face 3000 500 00.003 1.80
EVA 960 2090 0.0005 0.35
Silicon cells 2330 677 0.0003 148
Polyester/Tedlar Trilaminate 1200 1250 0.0005 00.2

Table 2 3.2.1. Radiation heat transfer

Thermophysical properties of PCM (RT-35) (Rubitherm GmbH, 2016). The total radiation absorbed by PV coupled with PCM (PV/PCM)
Properties Value depends upon solar radiation incident on from PV panel surface
Melting Temperature (°C) 35
and total radiation produced by the PV/PCM system. The total radi-
Latent Heat of Fusion (kJ/kg) 240 ated energy obtained by PV/PCM structure can be given as:
Specific heat (kJ/(kg K)) Solid 2
Liquid 2
qrad ¼ qs
ðqr þ qsky Þ ð5Þ
Density (kg/m3) Solid 880
Liquid 770 where qs, qr and qsky stand for the radiation absorbed by PV panels,
Thermal conductivity (W/(m K) Solid 0.2 radiative heat loss to the ground and sky correspondingly. The PV
Liquid 0.2 module efficiency depends on absorptivity of covering glass, and
Thermal expansion coefficient (1/K) 0.00091 it can be given as follows:
Kinematic Viscosity (m2/s) 5  10
6
qs ¼ aGmodule A ð6Þ
The equations relating Gmodule, Ghorizontal and Gincident represented
In Eq. (1) E represents the interior heat generation within solid as follows (Kant et al., 2016):
parts, q, Cp and k are respectively the density, specific heat and
thermal conductivity of each metallic layer; D/Dt is the Lagrangian Ghorizontal ¼ Gincident sin h
derivative and tw is the heat added from viscous frictions. Within Gmodule ¼ Gincident sinðh þ /Þ
solids, E, w is zero and velocity involved in the Lagrangian deriva- h ¼ 90
w þ d ð7Þ
tive u is zero. Applying the above-said conditions, the Eq. (1) can be
360 
d ¼ 23:45 sin 365 ð284 þ nÞ
rewritten in the form of:
Gmodule ¼ Ghorizontal
Sinh
sinðhþ/Þ

@T
qC p ¼ kDT ð2Þ where h, /, w and d are the elevation, inclination, latitude and decli-
@t
nation angle of the PV/PCM system, and Ghorizontal and Gincident the
radiation incident on surface at right angles to the sun’s direction
3.2. On the PV panel surface respectively.
The effective radiation leaving from the PV panel’s front surface
Above the front PV panel surface, we have considered solar radi- and aluminum container back surface mainly depend on the sur-
ation, longwave radiation from glass surface and convection terms: face emissivity, the surrounding temperature and the temperature
of the panel front surfaces and aluminum container back surface
@T

kglass ¼ ðhfree þ hforced ÞðT amb
T glass Þ þ eglass F rðT 4sky
T 4glass Þ and tilt angle. The heat radiated to the ground, qr is given as
@y (Kant et al., 2016):
þ aglass Gmodule ðtÞ ð3Þ
qr ¼ ealu min um F rAðT 4alu min um
T 4ground Þ ð8Þ
where hfree and hforced respectively signify the heat transfer coeffi-
It is supposed that the ground temperature is same as the ambi-
cient for natural and forced convection as defined in Eqs. (13) and
ent temperature. The radiated heat given to the sky is as follows
(17), Tamb, Tglass and Tsky are respectively the ambient, the covering
(Biwole et al., 2008):
glass surface, and the sky temperature as defined in Eq. (9), F is
)
the view factor as defined in Eq. (10) and Gmodule is the radiation inci- qsky ¼ eglass F rAðT 4glass
T 4sky Þ
dent on the PV module as properly defined in Eq. (7). The glass ð9Þ
Tsky ¼ 0:037536  T 1:5
amb þ 0:32  T amb
cover emissivity eglass is taken as 0.91 (Notton et al., 2005). On the
aluminum plate at back side of system, long wave radiation from The view factor F on PV/PCM structure can be expressed as
aluminum surface and convection terms are accounted with follow- follows:
ing equation: 8 9
> F front
sky ¼ 12 ð1 þ cos /Þ >
>
> >
>
@T <F =
front
ground ¼ 2 ð1
cos /Þ
1

kalu min um ¼ ðhfree þ hforced ÞðT amb
T alu min um Þ ð10Þ
@y > F rear
sky ¼ 2 ð1 þ cosðp
/Þ >
1
>
> >
>
þ ealu min um F r ðT 4ground
T 4alu min um Þ ð4Þ : ;
F rear
ground ¼ 12 ð1
cosðp
/Þ

where Tground and Taluminum correspondingly stand for ground tem-

perature and the aluminum back surface temperature. The alu- 3.2.2. Convective heat losses
minum sheet external surface emissivity ealu min um was taken as Convective heat losses are also considered for panel front sur-
0.85 (Bazilian et al., 2002). As reflected in Eqs. (3) and (4), longwave face and aluminum container back surface. The convective heat
radiation of the panel front surface (respectively rear surface) with loss occurs due to natural and forced convection. Therefore, the
the surrounding buildings and the ground (respectively the sky) heat loss by convection at the surface of the PV/PVM system is
was neglected. expressed as
 K. Kant et al. / Solar Energy 140 (2016) 151–161 155

qconv ¼ qn þ qf ¼ ðhfree þ hforced ÞAðT PV
T amb Þ ð11Þ qPCM ðTÞ ¼ qsolid þ ðqliquid
qsolid ÞBðTÞ ð19Þ

where qn and qf are heat losses as a result of natural and forced con- where qPCM is temperature dependent density of PCM. The densities
vection respectively, and TPV the panel’s surface temperature. of PCM in solid and liquid phase are qsolid and qliquid respectively.
The density of PCM varies linearly in transition zone according to
3.2.2.1. Heat loss due to natural convection. The mathematical rela- the function B(T) which can be given as:
tion for heat losses by natural convection is specified by consider- 8
>
< 0; T < ðT m
DTÞ
ing Newton’s law of cooling and is given by
BðTÞ ¼ ðT
T m þ DTÞ=ð2DTÞ; ðT m
DTÞ 6 ThðT m þ DTÞ ð20Þ
qn ¼ hfree AðT PV
T amb Þ ð12Þ >
:
1 T > ðT m þ DTÞ
The heat transfer coefficient because of natural convection hfree
Eq. (20) represents that B is zero when the PCM is in solid state
for PV panel front surface and aluminum container back surface
and 1 when it is melted and it linearly grows from 0 to 1 between
can be calculated by following expression (Incropera et al., 2011):
the two states (Biwol et al., 2014). Where Tm is melting tempera-
00 1 9
> ture and DT is the transition temperature of PCM. The specific heat
>
>
B B C >
> and latent heat of fusion for the PCM can be modeled using Eq.
B kB 0:67ðcos /ÞRaL
1=4
C 9 >
>
B L B0:68 þ   9=16  C

if


Ra 6 10 >
> (21):
B @ 4=9
A
L
>
>
B >
>
B >
1þ 0:492k

B
lCp = C pPCM ðTÞ ¼ C psolid þ ðC pliquid
C psolid ÞBðTÞ þ Lf DðTÞ ð21Þ
hfree B
¼B 0 1 >
B >
> where CP PCM is temperature dependent specific heat of PCM and
B B >
>
B B C > specific heat in solid and liquid phase are Cp solid and Cp liquid respec-
B k B0:825 þ  0:387RaL  C 9>
>
1=6

BL@   C

if


Ra > 10 >

>
@
8=27
A
L >
> tively. The Lf is the latent heat of fusion and D(T) is depends on tem-
>
9=16
1þ 0:492k
lCp
>
; perature T, DT and Tm of the PCM, which is modeled as Eq. (22):
 .pffiffiffiffiffiffiffiffiffi

ðT
T m Þ2
ð13Þ p:DT 2
DðTÞ ¼ e DT 2
ð22Þ
where L and RaL are the elevations of PV panel and Raleigh number
Function D is Delta function which is zero everywhere except in
respectively and the value of RaL is given by
interval [Tm
DT, Tm + DT]. It pinpoints on Tm and the value of its
g ap q2 C p jT PV
T amb jL3 integral is one. The main role of function D is to distribute the
RaL ¼ ð14Þ latent heat of fusion similarly nearby the melting point of PCM.
kl
The thermal conductivity of the PCM depending on its phase can
In above equation, the ap is thermal expansion factor of air and be modeled by Eq. (23):
given by
  kPCM ðTÞ ¼ ksolid þ ðkliquid
ksolid ÞBðTÞ ð23Þ
@q
1
ap ¼ ð15Þ
q @T PV p
3.3.2. Mass transfer
It is supposed that the melt PCM is a Newtonian fluid. The two-
3.2.2.2. Heat loss due to forced convection. Wind-induced forced dimensional energy conservation, mass and momentum transfer
convection heat losses puts a substantial effect on calculating the diffusion equation with the effect of bouncy driven force, solved
temperature of the panel and widespread values for the forced concurrently. The mass and momentum transfer equation modified
convection heat transfer coefficient (hforced) is existing in the liter- as follows to model the phase transition process:
ature (Skoplaki and Palyvos, 2009). The heat loss occurred due to !
wind induced forced convection is given by @u ! ! ! ! !
q þ q ðu rÞ u
lcdot r2 u ¼
rP þ F b þ F a ð24Þ
@t
qf ¼ hforced AðT PV
T amb Þ ð16Þ
Fb is a buoyancy force given by the Boussinesq approximation:
The value of hforced can be given by using the fundamental the- ! !
ory of heat transfer (Incropera et al., 2011), wind tunnel and field F b ¼
qliquid ð1
bðT
T m ÞÞ g ð25Þ
measurements (Kant et al., 2016). The value average hforced over
the PV panel is specified as follows (Incropera et al., 2011): and
!
0 !
1=2
0:3387Pr 1=3 ReL 5
F a ¼
AðTÞ  u ð26Þ
k
B 2 L ð1þð0:0468=PrÞ2=3 Þ1=4


if

ReL 6 5  10
hforced ¼ @ where the value of A(T) which is encouraged from the Carman–
2 kL Pr 1=3 ð0:037ReL4=5
871Þ


if

ReL > 5  105 Koseny relation in a porous medium while the magnitude of »(P)
ð17Þ from Darcy’s law and can be given as (Voller and Prakash, 1987)
and (Brent et al., 1988):

3.3. Heat and mass transfer in the PCM Cð1
BðTÞÞ2

AðTÞ ¼ ð27Þ
ðB3 ðTÞ þ qÞ
3.3.1. Heat transfer
If we assume that, the flow is laminar:
The heat transfer diffusion equation has been applied to the PV
panel, PCM, and the aluminum container that can be given as:
Cð1
BðTÞÞ2 !
rP ¼ :u ð28Þ
@T ! B3 ðTÞ
qC p þ r:ð
krTÞ þ qC p u :rT ¼ 0 ð18Þ
@t The value of C depends on the morphology of the PCM. Raising
The velocity field u in Eq. (18) is specified by Navier-Stokes its value leads to a reduction of the flow of liquid matter in the
equations for incompressible fluids. The density of the PCM can mushy region, and therefore to a flattering shape of the liquidus
be modeled using Eq. (19): front (mushy zone boundary closer to the liquid side of the PCM)
156 K. Kant et al. / Solar Energy 140 (2016) 151–161

while reducing its value will have the opposite effect (Brent et al.,
1988). Similar studies on the subject have considered the value of C
varying between 104 and 107. In this study, C is given the constant
value 106 due to the high viscosity of solid PCM. Constant q is cho-
sen very low so as to make Eq. (19) effective, even when B(T) is
zero. The significant value of q is fixed at 10
3. With the upsurge
in the temperature of the PCM and gets higher than Tm + DT, the
PCM becomes in the liquid phase, hence, B is 1 and consequently,
A and Fa are zero. In this situation, the common momentum con-
servation equation applies. During the transition state to crystal-
lization, 0 < B(T) < 1. A(T) upsurges accompanied by the melting
process until the added force Fa reaches greater than the convec-
tion and diffusion terms in Eq. (24) and the momentum equation
becomes analogous to the Darcy law for fluid flow in a porous
medium: Fig. 2. Mesh at a mid-height cross section of the numerical model.

! K
u¼
rP ð29Þ
l of elements (elements ref Fig. 2) selected for the numerical simu-
lation. The simulations were performed on a workstation having
where the permeability K is a function of B(T). When B(T) reduces,
processor Intel(R) Xenon(R), 64 GB of RAM. The model was two-
the velocity field also reduces until it turns into zero when the
dimensional and the simulation was performed for the atmo-
PCM converts completely to the solid phase. After solidification of
spheric conditions for May 2011, for the City of Allahabad, Uttar
the PCM, the value of B becomes zero as temperature reaches below
Pradesh, India (25.4500°N, 81.8500°E). Fig. 3 represents the Data
the Tm + DT. Accordingly, the added force controls all the terms in
of wind velocity, ambient temperature and solar radiation taken
the momentum conservation equation.
to carry out numerical investigation reported in the present study.1
The transient numerical investigations were carried out by adopting
3.4. Power output
constant (Newtonian) iteration procedures with Backward Euler
time stepping method. The maximum six numbers of iterations were
The power productivity of PV cells can be given by the following
assigned for each time step, and the damping factor was 0.9 and
relation (Dubey et al., 2013):
maximum time step was 5 s.
Pout ¼ Im V m ¼ ðFFÞIsc V oc ¼ gAG ð30Þ The present thermodynamic study with and without PCM con-
sidering the convective mode of heat transfer effect has been vali-
where Im is current and Vm is voltage output at maximum power, FF dated by previous experimental study Park et al. (2014) and Huang
is fill factor, and Isc and Voc are open circuit current and voltage et al.(2006) respectively. The validation of temperature variation of
respectively. The Voc and the FF decrease considerably with an without PCM with the study of Huang et al.(2006) is shown in Fig
increase in the temperature, while short-circuiting current 4. Table 3 gives the properties of materials that have been used as
increases slightly (Zondag, 2008). Thus, the efficiency of module PCM for the validation of our thermal model. The temperature vari-
can be represented by a relation as following (Evans and ation of the day is calculated from the model and is compared with
Florschuetz, 1977): the study of Park et al. (2014), as shown in Fig. 5. This comparative
g ¼ gref ½1
bref ðT PV
T ref Þ ð31Þ study clearly shows that results obtained with our thermal model
are in good match with the experimental results as measured by
where gref, bref, and Tref are respectively the panel’s electrical effi- Park et al. (2014). Such comparative analysis puts a stringent test
ciency, temperature coefficient, and temperature at STC. The value of the reliability and accuracy of the presently developed model
of bref is 0.004 1/K for crystalline PV panels (Notton et al., 2005). before using it further.
The total power output is given by
Pout ¼ gref ½1
bref ðT PV
T ref ÞAGmodule ð32Þ 5. Results and discussion

The present thermal model for PV/PCM has been validated first
4. Computational procedure, mesh dependency test, and model and then applied to examine the temperature variation of PV pan-
validation els coupled with PCM. The initial temperature of the module and
the PCM is equal to the ambient temperature. The results and dis-
The concurrent momentum, heat and mass transfer prevailing cussion of the current study are presented in following sub-
differential equations subjected to the boundary conditions are sections (i) melting and solidification of PCM, (ii) effect of natural
computed numerically by using the heat transfer module of the convection in PCM on PV panel temperature (iii) Effect of PV
COMSOL Multiphysics 5.0 software, based on the finite element panel’s tilt angle on its operating temperature (iv) Effect of wind
method. PV/PCM geometry, PV panel layers, and PCM properties speed on PV panel’s operating temperature (v) and PV panel power
were defined appropriately in COMSOL Multiphysics and triangu- output.
lar meshing was created for the overall mesh (ref. Fig. 2). The mesh
dependence study was done in order to improve the accuracy and 5.1. Melting and solidification of PCM
calculation time of the model i.e. to offer an accurate mesh-
independent solution and to reduce the overall calculation time. At the initial state (From 00:00 h to the 5:00 AM) the solar radi-
This helped in defining the appropriate size of elements for further ation is not available therefore the temperature of Polycrystalline
study. The three mesh sizes initially considered had 25,357, 38,579
and 45,136 finite elements and it is observed that the results 1
Data obtained from Indian Society of Heating, Refrigerating and Air-Conditioning
obtained were quite similar for all three mesh size. For the sake Engineers (ISHRAE) http://apps1.eere.energy.gov/buildings/energyplus/cfm/weath-
of exactness in the results, finer mesh size having 45,136 numbers er_data3.cfm/region=2_asia_wmo_region_2/country=IND/cname=India#instructions.
 K. Kant et al. / Solar Energy 140 (2016) 151–161 157

Fig. 3. Data of wind speed, ambient temperature and solar radiation.

Fig. 5. Model validation with Park et al. (2014).

day are shown in Fig. 6. Because of the lower density of melted

PCM, it moves upward near the PV panel and cooler PCM parts hav-
Fig. 4. Model validation with Huang et al. (2006).
ing higher density fill that space, as represented by arrows in Fig. 6.
By this way, the natural convection takes place during the melting
Table 3 of PCM. The smaller arrow represents melted PCM with lower
Thermal properties of the PCM (Park et al., velocity and larger arrow size represents melted PCM with higher
2014). velocity in the fig. The absence of arrow means PCM with zero
Thermophysical properties Value velocity. Fig. 7 displays the velocity of the melted PCM at mid-
Thermal conductivity (W/m K) 0.2
height of the PV panel. The melted PCM velocity at the initial stage
Density (solid) (kg/m3) 880 of melting (at 7:30 AM) is lower near the front aluminum plate and
Density (fluid) (kg/m3) 760 it increases with time.
Specific heat capacity (kJ/kg K) 2.1 As time increases, the process of melting gets enhanced and
Melting temperature (K) 298
after some time it is accelerated when the PCM in liquid phase
Heat of fusion (kJ/kg) 184
comes in contact with the back of the aluminum plate as it gets
warmer than the melt temperature of the PCM. The initiation of
melting starts at about 7:00 AM and completes at around
silicon cells does not increase. At daytime, with the solar radiation 9:00 AM. After the completing the fusion process, the melted
falling on the PV panel front surface, the temperature of PV panel PCM absorbs energy from PV panels in the form of sensible heat
starts growing and heat transfer starts taking place through PV as its temperature keeps increasing. During the sensible heating
panel to PCM. The PCM first absorbs and stores energy in the form of liquid PCM, the heated liquid PCM near the front aluminum
of sensible heat in its solid phase as its temperature increases. plate moves upward and circulates inside the container. The max-
When the temperature of the solid PCM reaches the situation of imum mean velocity in the PCM domain is seen at around
transition temperature, the energy starts getting stored in the form 11:00 AM with the value of 4.28  10
4 m/s.
of latent heat. The simulated results of temperature and velocity When the solar radiation starts decreasing at around 12:00 PM,
field inside the PCM domain at a different time interval of the the PCM temperature also starts decreasing, along with the veloc-
158 K. Kant et al. / Solar Energy 140 (2016) 151–161

Fig. 6. Temperature (°C) and velocity field of PV coupled with phase change materials at different time interval.

Fig. 7. Variation of mean velocity at mid height of PCM with Arc length (m).

ity field in the melted PCM (refer to Fig. 7 at 17 h, 17.5 h, and 18 h). et al., 2016; Park et al., 2014). However, it has been pointed out
The PCM starts solidifying at about 05:30 PM and is completely that the heat transfer due to convection plays a vital role in the
solidified at 08:00 PM. It is important to note that during solidifica- melting of PCM when container volume is large enough to allow
tion, the velocity induced due to natural convection of the PCM is for macroscopic fluid particle displacement (Buddhi and Bansal,
around zero, showing that conduction heat transfer is overriding. 1988; Costa et al., 1998; Liu et al., 2014). The convective mode of
Fig. 8 represents the variation of PCM melt fraction over the day heat transfer speeds up melting process and gives better heat
time. It shows that the process of fusion and solidification are car- transfer rate which causes a reduced temperature of PV panels pre-
ried out in a short period of time of about 2 h. sented in Fig. 9. More importantly, during the melting process, nat-
ural convection is dominant over conduction. This is because of the
5.2. Effect of convective heat transfer in PCM on PV panel temperature mass transfer inside the PCM cavity which leads to an enhanced
thermal conductivity of PCM (Farid and Kanzawa, 1989). Due to
The melting of PCM occurs due to both conductive and convec- the enhancement of PCM’s effective thermal conductivity, the
tive mode of heat transfer. Earlier studies on thermal modelling of melting process is accelerated and results in more heat transfer
PV panels with PCMs considered only the conductive mode of heat from the PV panel in comparison to the heat transfer without con-
transfer inside the PCM during its fusion and solidification (Kibria sidering the convection effect, which can be seen from Fig. 9, where
 K. Kant et al. / Solar Energy 140 (2016) 151–161 159

Fig. 10. Effect of tilt angle on PV panel operating temperature.

Fig. 8. Variation of melt fraction over the daytime of PCM.

panel is horizontal and convection is maximum for a vertical panel.

Overall the PV panel operating temperature decreases with incre-
the maximum panel temperature is 54.90 °C with conduction and ment in PV panel tilt angle.
convection effect, 58.5 °C with conduction only and 60 °C for PV
panel without having PCM at all. Thus, proper consideration of con-
vective mode of heat transfer is highly required for accurate ther-
5.4. Effect of wind speed on PV panel operating temperature
mal analysis of PV/PCM systems.

The operating temperature variation of PV panel is also depen-

5.3. Effect of PV panel’s tilt angle on PV panel operating temperature dent on the wind speed. Fig. 11 represents the effect of wind speed
on PV panel operating temperature. The developed thermal model
The PV module tilt angle affects the melting of the PCM is solved for three different wind speed i.e. 2 m/s, 3 m/s and 4 m/s
attached at its back by due to the convective heat transfer inside at 45° angle of inclination. From figure it is clearly shown that
the PCM. The developed thermal model is solved for three different when the wind velocity is higher, the PV panel operating temper-
angle of inclination from vertical i.e. 30°, 45° and 60°. Fig. 10 rep- ature tends to close to the ambient temperature, due to the higher
resents the effect of tilt angle on variation of PV panel operating convective heat loss and therefore increased heat transfer to the
temperature. As the tilt angle of PV/PCM system increases the surroundings from PV/PCM. The lower ambient temperatures lead
velocity of melted PCM decreases, which reduces the convective to enhanced heat losses from PV panel front and PCM container
heat transfer inside the PCM cavity. The natural convection heat back surface by natural and forced convection and also enhanced
transfer losses of air to the environment over the panel decreases radiative heat losses due to the temperature gradient between
along with the PV tilt angle, and there is no convection when the PV/PCM and ambient.

Fig. 9. Temperature variation of PV panel over the daytime.

160 K. Kant et al. / Solar Energy 140 (2016) 151–161

PV panel. In comparison to existing studies, the present thermal

model offers a comprehensive analysis as it includes convective
mode of heat transfer inside the PCM, fusion and solidification of
the PCM, influence of wind speed on natural and forced convection
from PV panel’s front and a back surface, radiative heat transfer to
ambient and ground, variable weather situations and the effect of
the panel inclination angle on the heat and mass transfer mecha-
nism in PCM along with heat loss from PV panels front and PCM
container back surface. All the heat transfer mechanisms are being
considered to develop a thoughtful understanding of heat and
mass transfer phenomenon in PV/PCM system. Thus the developed
thermal model can ably forecast the temperature response of solar
panel, which has been successfully tested by the experimental
results available in the literature. These results clearly indicate that
PCM offers an effective way of controlling the increase in the tem-
perature of the panel. The outcome of the study is summarized
below:
Fig. 11. Eeffect of wind Speed on PV panel operating temperature.
1. The simulation studies are performed for the month of May and
for the Indian climatic conditions of the City of Allahabad (Uttar
Pradesh) (25.4500°N, 81.8500°E) which have very good PV
panel installation potential, the operating temperature of the
panel with PCM attached its back surface is lower by a maxi-
mum of 6 °C when equated to the conventional PV module,
when considering convective mode of heat transfer effect in
the PCM, and reduced by 3 °C when considering only conduc-
tion mode of heat transfer, confirming the affect of PCM to
maintain the panel temperature due to adding PCM. Ultimately,
the productivity of the solar panel module is increased by about
5%.
2. The convective mode of heat transfer during the melting of PCM
plays a substantial role in carrying out the thermal analysis and
power output analysis of PV panel.
3. The maximum operating temperature reaches at 55.31 °C,
53.53 °C, and 52.32 °C with wind speed of 2 m/s, 3 m/s and
4 m/s respectively.
4. The maximum operating temperature reaches at 55.81 °C,
54.88 °C, and 54.32 °C with tilt angle of 60°, 45° and 30° respec-
Fig. 12. Variation of power loss with and without PCM.
tively from vertical.

Overall it can be summarized that the angle of inclination, wind

5.5. PV panel power output speed, convective heat transfer in PCM has substantial effect on PV
panel operating temperature.
The effect of operating temperature on PV panel power loss is
shown in Fig. 12. From this figure, it can be seen that the power
loss increases sharply, if PCM is not attached at the back side of
PV panel. Using PCM to regulate the PV module temperature, one References
can see the significant decrease in power loss, which otherwise
Armstrong, S., Hurley, W.G., 2010. A thermal model for photovoltaic panels under
would have occurred due to high PV module temperature. This varying atmospheric conditions. Appl. Therm. Eng. 30, 1488–1495. http://dx.
happens because of the isothermal phase change nature of PCM doi.org/10.1016/j.applthermaleng.2010.03.012.
Atkin, P., Farid, M.M., 2015. Improving the efficiency of photovoltaic cells using PCM
while changing its phase; the PCM delays the rise of panel temper-
infused graphite and aluminium fins. Sol. Energy 114, 217–228. http://dx.doi.
ature by absorbing the extra thermal energy of the system. The org/10.1016/j.solener.2015.01.037.
natural convection for the duration of the PCM melting also plays Bazilian, M.D., Kamalanathan, H., Prasad, D.K., 2002. Thermographic analysis of a
a substantial role in the reduction of PV module power loss. This building integrated photovoltaic system. Renew. Energy 26, 449–461. http://dx.
doi.org/10.1016/S0960-1481(01)00142-2.
is because of greater heat transfer rate from the panel to the Biwol, P., Eclachec, P., Biwol, P., Eclachec, P., 2014. Phase-change materials to
PCM and environment. The extreme power loss occurs with the improve solar panel’ s performance. Energy Build. 62, 59–67.
PV panel system with no PCM and the PV panel having PCM with Biwole, P.H., Woloszyn, M., Pompeo, C., 2008. Heat transfers in a double-skin roof
ventilated by natural convection in summer time. Energy Build. 40, 1487–1497.
natural convection effect yields the lowest power loss. The maxi- http://dx.doi.org/10.1016/j.enbuild.2008.02.004.
mum power loss is 14% for the panel with no PCM whereas the Brent, A.D., Voller, V.R., Reid, K.J., 1988. Enthalpy-Porosity technique for modeling
minimum power loss is 11% for the panel with PCM at its back. convection-diffusion phase change: application to the melting of a pure metal.
Numer. Heat Transfer Part A Appl. 13, 297–318.
Browne, M.C., Norton, B., McCormack, S.J., 2015. Phase change materials for
photovoltaic thermal management. Renew. Sustain. Energy Rev. 47, 762–782.
6. Conclusion http://dx.doi.org/10.1016/j.rser.2015.03.050.
Buddhi, D., Bansal, N.K., 1988. Solar thermal storage systems using phase change
A thermodynamic model of PV panel coupled with PCM at its materials. Int. J. Energy Res. 12 (12), 547–555.
Costa, M., Buddhi, D., Oliva, A., 1998. Numerical simulation of a latent heat thermal
back has been developed to study and recognize the heat, mass energy storage system with enhanced heat conduction. Energy Convers.
and momentum transfer features of a PCM integrated standalone Manage. 39, 319–330. http://dx.doi.org/10.1016/S0196-8904(96)00193-8.
 K. Kant et al. / Solar Energy 140 (2016) 151–161 161

Dubey, S., Sarvaiya, J.N., Seshadri, B., 2013. Temperature Dependent Photovoltaic Lin, W., Ma, Z., 2016. Using Taguchi-Fibonacci search method to optimize phase
(PV) efficiency and its effect on PV production in the world – a review. Energy change materials enhanced buildings with integrated solar photovoltaic
Procedia 33, 311–321. http://dx.doi.org/10.1016/j.egypro.2013.05.072. thermal collectors. Energy 106, 23–37. http://dx.doi.org/10.1016/j.
Duran Sahin, A., Dincer, I., Rosen, M.A., 2007. Thermodynamic analysis of solar energy.2016.03.013.
photovoltaic cell systems. Sol. Energy Mater. Sol. Cells 91, 153–159. http://dx. Lin, W., Ma, Z., Cooper, P., Sohel, M.I., Yang, L., 2016. Thermal performance
doi.org/10.1016/j.solmat.2006.07.015. investigation and optimization of buildings with integrated phase change
Evans, D.L., Florschuetz, L.W., 1977. Cost studies on terrestrial photovoltaic power materials and solar photovoltaic thermal collectors. Energy Build. 116, 562–
systems with sunlight concentration. Sol. Energy 19, 255–262. 573. http://dx.doi.org/10.1016/j.enbuild.2016.01.041.
Farid, M.M., Kanzawa, A., 1989. Thermal performance of a heat storage module Liu, S., Li, Y., Zhang, Y., 2014. Mathematical solutions and numerical models
using PCM’s with different melting temperatures: mathematical modeling. J. employed for the investigations of PCMs’ phase transformations. Renew.
Sol. Energy Eng. 111, 152. http://dx.doi.org/10.1115/1.3268301. Sustain. Energy Rev. 33, 659–674. http://dx.doi.org/10.1016/j.rser.2014.02.032.
Fontenault, B., Gutierrez-Miravete, E., 2012. Modeling a combined photovoltaic- Ma, T., Yang, H., Zhang, Y., Lu, L., Wang, X., 2015. Using phase change materials in
thermal solar panel. In: Proceedings of the 2012 COMSOL conference, Boston, photovoltaic systems for thermal regulation and electrical efficiency
USA. improvement: a review and outlook. Renew. Sustain. Energy Rev. 43, 1273–
Hasan, A., McCormack, S.J.J., Huang, M.J.J., Norton, B., 2014. Characterization of 1284. http://dx.doi.org/10.1016/j.rser.2014.12.003.
phase change materials for thermal control of photovoltaics using differential Maiti, S., Banerjee, S., Vyas, K., Patel, P., Ghosh, P.K., 2011. Self regulation of
scanning calorimetry and temperature history method. Energy Convers. photovoltaic module temperature in V-trough using a metal–wax composite
Manage. 81, 322–329. http://dx.doi.org/10.1016/j.enconman.2014.02.042. phase change matrix. Sol. Energy 85, 1805–1816. http://dx.doi.org/10.1016/
Hasan, A., McCormack, S.J.J., Huang, M.J.J., Norton, B., 2010. Evaluation of phase j.solener.2011.04.021.
change materials for thermal regulation enhancement of building integrated Malvi, C.S., Dixon-Hardy, D.W., Crook, R., 2011. Energy balance model of combined
photovoltaics. Sol. Energy 84, 1601–1612. http://dx.doi.org/10.1016/ photovoltaic solar-thermal system incorporating phase change material. Sol.
j.solener.2010.06.010. Energy 85, 1440–1446. http://dx.doi.org/10.1016/j.solener.2011.03.027.
Ho, C.J., Tanuwijava, A.O., Lai, C.-M., 2012. Thermal and electrical performance of a Notton, G., Cristofari, C., Mattei, M., Poggi, P., 2005. Modelling of a double-glass
BIPV integrated with a microencapsulated phase change material layer. Energy photovoltaic module using finite differences. Appl. Therm. Eng. 25, 2854–2877.
Build. 50, 331–338. http://dx.doi.org/10.1016/j.enbuild.2012.04.003. http://dx.doi.org/10.1016/j.applthermaleng.2005.02.008.
Huang, M.J., Eames, P.C., Norton, B., 2004. Thermal regulation of building-integrated Park, J., Kim, T., Leigh, S.-B.B., 2014. Application of a phase-change material to
photovoltaics using phase change materials. Int. J. Heat Mass Transf. 47, 2715– improve the electrical performance of vertical-building-added photovoltaics
2733. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2003.11.015. considering the annual weather conditions. Sol. Energy 105, 561–574. http://dx.
Huang, M.J.J., Eames, P.C.C., Norton, B., 2006. Phase change materials for limiting doi.org/10.1016/j.solener.2014.04.020.
temperature rise in building integrated photovoltaics. Sol. Energy 80, 1121– Rubitherm GmbH, 2016. Available at <http://www.rubitherm.de>.
1130. http://dx.doi.org/10.1016/j.solener.2005.10.006. Skoplaki, E., Palyvos, J.A., 2009. Operating temperature of photovoltaic modules: a
Incropera, F.P., DeWitt, D.P., Bergman, T.L., Lavine, A.S., Incropera, F.P., Lavine, A.S., survey of pertinent correlations. Renew. Energy 34, 23–29. http://dx.doi.org/
2011. Fundamentals of Heat and Mass Transfer, Dekker Mechanical 10.1016/j.renene.2008.04.009.
Engineering. John Wiley & Sons. Sohel, M.I., Ma, Z., Cooper, P., Adams, J., Scott, R., 2014. A dynamic model for air-
Islam, M.M., Pandey, A.K., Hasanuzzaman, M., Rahim, N.A., 2016. Recent progresses based photovoltaic thermal systems working under real operating conditions.
and achievements in photovoltaic-phase change material technology: a review Appl. Energy 132, 216–225. http://dx.doi.org/10.1016/j.apenergy.2014.07.010.
with special treatment on photovoltaic thermal-phase change material Stropnik, R., Stritih, U., 2016. Increasing the efficiency of PV panel with the use of
systems. Energy Convers. Manage. 126, 177–204. http://dx.doi.org/10.1016/j. PCM. Renew. Energy 97, 671–679. http://dx.doi.org/10.1016/j.
enconman.2016.07.075. renene.2016.06.011.
Jones, A.D., Underwood, C.P., 2001. A thermal model for photovoltaic systems. Sol. Tonui, J.K., Tripanagnostopoulos, Y., 2007. Air-cooled PV/T solar collectors with low
Energy 70, 349–359. http://dx.doi.org/10.1016/S0038-092X(00)00149-3. cost performance improvements. Sol. Energy 81, 498–511.
Kant, K., Shukla, A., Sharma, A., Biwole, P.H., 2016. Thermal response of poly- Voller, V.R., Prakash, C., 1987. A fixed grid numerical modelling methodology for
crystalline Silicon photovoltaic panels: numerical simulation and experimental convection-diffusion mushy region phase-change problems. Int. J. Heat Mass
study. Sol. Energy 134, 147–155. http://dx.doi.org/10.1016/ Transf. 30, 1709–1719.
j.solener.2016.05.002. Zhao, B., Chen, W., Hu, J., Qiu, Z., Qu, Y., Ge, B., 2015. A thermal model for amorphous
Kibria, M.A., Saidur, R., Al-Sulaiman, F.A., Aziz, M.M.A., 2016. Development of a silicon photovoltaic integrated in ETFE cushion roofs. Energy Convers. Manage.
thermal model for a hybrid photovoltaic module and phase change materials 100, 440–448. http://dx.doi.org/10.1016/j.enconman.2015.04.062.
storage integrated in buildings. Sol. Energy 124, 114–123. http://dx.doi.org/ Zondag, H., 2008. Flat-plate PV-thermal collectors and systems: a review. Renew.
10.1016/j.solener.2015.11.027. Sustain. Energy Rev. 12, 891–959. http://dx.doi.org/10.1016/j.rser.2005.12.012.