Solar Energy 80 (2006) 751–759

www.elsevier.com/locate/solener

Performance evaluation of solar PV/T system:

An experimental validation
Arvind Tiwari *, M.S. Sodha
Centre for Energy Studies, Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India

Received 2 September 2004; received in revised form 30 May 2005; accepted 6 July 2005
Available online 22 August 2005

Communicated by: Associate Editor Jean Rosenfeld

Abstract

In this communication, an attempt has been made to develop a thermal model of an integrated photovoltaic and
thermal solar (IPVTS) system developed by previous researchers. Based on energy balance of each component of
IPVTS system, an analytical expression for the temperature of PV module and the water have been derived. Numerical
computations have been carried out for climatic data and design parameters of an experimental IPVTS system. The
simulations predict a daily thermal efficiency of around 58%, which is very close to the experimental value (61.3%)
obtained by Huang et al.
 2005 Elsevier Ltd. All rights reserved.

Keywords: PV thermal; Solar energy; Water heater

1. Introduction
The energy payback time (EPBT) of a photovoltaic
(PV) system lies between 10 and 15 years depending
on insulation and the efficiency of the PV module. If
the efficiency can be increased then the energy payback
time can be reduced. In order to increase the efficiency
of the PV module, the temperature of the PV module
should be decreased (Zondag et al., 2003; Chow,
2003). Jones and Underwood (2001) have developed a
non-steady-state thermal model for a photovoltaic sys-
tem (BP 585) by considering the effect of heat capacity
of the PV module. Infield et al. (2004) have suggested
reducing the temperature of the PV module by flowing
*
Corresponding author. Tel.: +91 11 26591255.
E-mail address: arvindtiwari02@yahoo.com (A. Tiwari).
air between the PV module and the double glass wall
for space heating. They have developed a steady-state
model to evaluate an overall heat loss coefficient (U)
and thermal gain factor (g). Similar studies were carried
out by Tripanagnostopoulos et al. (2002), Zondag et al.
(2002), Prakash (1994) and Chow (2003) by flowing air
and water below the PV module to increase the electrical
efficiency of the PV module. Such a system is referred as
a photovoltaic–thermal (PV/T) or hybrid or combi-
panel.
Zondag et al. (2002) have developed 1D, 2D, and 3D
dynamical models of a combi-panel. They concluded
that the simple 1D steady-state model for computing
daily yield from combi-panel performs almost as well
as more time consuming 2D and 3D dynamic models.
Two types of combi-panel (PV/T hybrid) for water heat-
ing have been considered i.e. (i) tube-in-plate configura-
tion, Zondag et al. (2002), Chow (2003), Huang et al.

0038-092X/$ - see front matter  2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.solener.2005.07.006
752 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759

Nomenclature

A area Utw an overall heat transfer coefficient from

b breadth glass to water through solar cell and the ted-
C specific heat lar
dx elemental length X theoretical value
FR flow rate factor Xi theoretical value of ith term
h heat transfer coefficient Y experimental value
hp1 penalty factor due to tedlar through glass, Yi experimental value of ith term
solar cell and EVA
hp2 penalty factor due to the interface between Subscripts
tedlar and the working fluid 0 glass to ambient
I(t) incident solar irradiation a ambient
K thermal conductivity bs back surface of tedlar
L length c solar cell
M mass eff effective
m_ rate of flow of water mass f fluid
n number of observations fi inlet fluid
Q_ u rate of useful energy transfer fout outgoing fluid
t time G glass
T temperature i insulation
Ub an overall heat transfer coefficient from r reference
water to ambient w water
UL an overall heat transfer coefficient from so- T tedlar
lar cell to ambient through the back insula- th thermal
tion
Ut an overall heat transfer coefficient from so- Greek letters
lar cell to ambient through the glass cover a absorptivity
UT conductive heat transfer coefficient from so- b packing factor
lar cell to water through tedlar g efficiency
UtT an overall heat transfer coefficient from s transmissivity
glass to tedlar through solar cell

(2001) and Kalogirou (2001) and (ii) parallel plate con-
figuration, Prakash (1994) and Tiwari et al. (2005).
Chow (2003) has developed a detailed dynamic model
for photovoltaic–thermal collectors (tube-in-plate con-
figuration) by considering the heat capacity of the glass,
PV plate, absorber plate, tube bonding and insulation
etc. Prakash (1994) has analyzed the transient behavior
of a photovoltaic–thermal solar collector for co-genera-
tion of electricity and hot air/water for a parallel plate
configuration. He concluded that the overall efficiency
of the PV/T is significantly increased which can further
reduce the energy payback time.
Huang et al. (2001) have conducted exhaustive exper-
imental studies of an integrated PV/T system (IPVTS) of
45 l capacity under forced mode of operation. They have
used both tube-in-plate and corrugated polycarbonate
panel configurations and found that the latter gives a
good thermal efficiency due to proper thermal contact
between fluid and the PV module.
In this paper, a thermal model of an IPVTS system as
proposed by Huang et al. (2001) has been developed and
validated by their experimental results. The design
parameters and climatic data of IPVTS system have
been used for numerical computations (Fig. 4, Huang
et al., 2001).
2. Experimental integrated photovoltaic–thermal
system (IPVTS)
Fig. 1 shows a schematic diagram of an integrated
photovoltaic–thermal system (IPVTS) set-up considered
by Huang et al. (2001). It consists of a PV/T system
(0.516 m2) and an insulated cylindrical storage water
heater of capacity 45 kg. The PV/T system comprises a
PV module separated by a tedlar film from a channel
through which the cooling water flows. The bottom of
the channel is made of insulation material. The storage
tank is connected to PV/T system through insulated
pipes. A water pump (3 W, DC) with controller has been
used to circulate the water. A cross-sectional view of the
components of the integrated thermal system is shown in
 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759 753

PV/T Collector

PV Module
Heat Collecting
Plate

Controller

Two

AC110V + Hot Water

60Hz
Supply
-
Insulated Storage
Tank

Tw
- + Pump Cold Water
Supply
T fi = Tw

+ -

Fig. 1. Schematic diagram of integrated PV/T system (IPVTS).

Fig. 2a and the corresponding thermal resistance circuit
diagram is shown in Fig. 2b. Fig. 2c shows the flow of
water along the length of IPVTS module with elemental
length ‘dx’.
An expression for the temperature dependent electri-
cal efficiency of a PV module (Schott, 1985; Evans, 1981)
is given by
g ¼ gr ½1  bðT c  T r Þ
where b = 0.0045 C1 is the temperature coefficient
(Eq. (3), Zondag et al., 2002), Tc is the temperature of
the cells, Tr is the reference temperature (taken as
25 C) and gr is the efficiency of the module at the refer-
ence temperature. Eq. (1) can be used to evaluate gr in
the present case for a given PV module efficiency
g = 0.09 and cell temperature (Tc) with water flow.
(iv) One-dimensional (1D) heat conduction has been
considered for the present study (Zondag et al.,
2002).
(v) The transmissivity of EVA is approximately
100%.
(vi) A mean temperature is assumed across each layer.
(vii) Water flow between the tedlar and the insulation
material is uniform for forced convection.
ð1Þ (viii) The system is in quasi-steady state.
Following Fig. 2, the energy balances for different
components of the integrated photovoltaic–thermal sys-
tem are as follows:
3.1. Solar cells of module (Fig. 2a)
sG ½ac bc I ðtÞ þ ð1  bc ÞaT I ðtÞW dx

¼ ½U t ðT c  T a Þ þ U T ðT c  T bs ÞW dx þ gc I ðtÞbc W dx
3. Thermal model
ð2aÞ
The following assumptions have been made From the above equation, the rate of thermal energy
transferred from the solar cell to the back surface of
(i) The heat capacity of the PV/T system (2918 J/K,
the tedlar can be obtained, given by
Jones and Underwood (2001)) has been neglected
in comparison to the heat capacity of the water U T ðT c  T bs Þ ¼ hp1 ðasÞI ðtÞ  U tT ðT bs  T a Þ ð2bÞ
(188,550 J/K) in the storage tank.
(ii) No stratification in the storage tank due to forced where hp1 ¼ U tUþU
T
T
, is the penalty factor (not present in
mode of operation (Tfi ffi Tw). the corresponding equation for a conventional flat plate
(iii) The heat capacities of solar cell material, tedlar collector) due to presence of solar cell material, tedlar
and insulation have been neglected. and EVA.
754 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759

Ta
I(t)
Y

Glass
qu Δx
X
Solar cell and EVA

Water in Tedlar C f Tf C fTf

Water out x + Δx
x
Insulating Material
x Δx
Ta

a c

Ts
Ta Conductive Resistance

Convective Resistance

Radiative Resistance
Glass
Tg

Solar Cell and

EVA
Tc

Tedlar
Tbs

Water out
Water in

Insulating Structure

Ta
TSO
L

Fig. 2. (a) Cross-sectional view of an integrated photovoltaic/thermal system, (b) thermal resistance circuit diagram for photovoltaic/
thermal system and (c) heat flow balance over an elemental length Dx below tedlar.

3.2. Back surface the tedlar hT hT U tT

hT ðT bs  T f Þ ¼ hp1 ðasÞI ðtÞ  ðT f  T a Þ
U tT þ hT hT þ U tT
U T ðT c  T bs ÞW dx ¼ hT ðT bs  T f ÞW dx ð3aÞ ¼ hp1 hp2 ðasÞI ðtÞ  U tw ðT f  T a Þ ð3bÞ
With the help of Eqs. (2b) and (3a), one gets the rate of
thermal energy transferred from the back surface of ted- where hp2 ¼ U tThþh
T
T
, is the penalty factor due to presence
lar to the working fluid in a steady-state condition, given of the interface between tedlar and the working fluid
by through the channel.
 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759 755

3.3. Flowing water below the tedlar (Fig. 2c) Table 1

Design parameters of PV/T collector and storage tank
The net thermal energy available in the fluid between Parameters Values
the tedlar and the bottom insulation is carried away by Ac 0.516 m2
the flowing fluid in forced mode of operation i.e. Average mass flow rate 0.016 kg/s
b 0.467
F 0 ½hp1 hp2 ðasÞI ðtÞ  U tw ðT f  T a Þ  U b ðT f  T a ÞW dx
Cw 4190 J/kg K
m_ dT f hi 5.8 W/m2 K
¼ cf W dx ð4aÞ
n dx h0 5.7 + 3.8 V
hp1 0.8772
or hp2 0.9841
hT 500 W/m2 K
m_ dT f
F 0 ½hp1 hp2 ðasÞI ðtÞ  U L ðT f  T a ÞW dx ¼ cf W dx Kc 0.039 W/m K
n dx KG 1.0 W/m K
ð4bÞ Ki 0.035 W/m K
KT 0.033 W/m K
The above equation is similar to the expression for con- L 1.105 m
ventional flat plate collector except the terms hp1 and hp2 Lc 0.0003 m
as explained earlier. The presence of hp1 and hp2 reduces LG 0.003 m
the efficiency of an IPVT system in comparison to that Li 0.05 m
of a conventional flat plate collector. LT 0.0005 m
From Eq. (4b) one gets Mw 45 kg
  Ub 0.62 W/m2 K
Q_ u ¼ F R hp1 hp2 ðasÞeff I ðtÞ  U L ðT w0  T a Þ ð5Þ Ut 9.24 W/m2 K
UT 66 W/m2 K
where UtT 8.1028 W/m2 K
ðasÞeff ¼ sG fac bc þ aT ð1  bc Þ  gc bc g ¼ 0:66 (UA)T 0.44 W/m2 K
V 1 m/s
for data of Table 1, is close to the value for conventional ac 0.70–0.85
flat plate collectors. aT 0.50
In addition to the above equations, we use the follow- bc 0.90
ing relations between the parameters defined in the table gc 0.09
of nomenclature: sG 0.95
 1
LG 1
U L ¼ U tw þ U b ; U t ¼ þ ;
K G h0 3.4. Storage tank: (Fig. 1)
KT hT U tT
UT ¼ ; U tw ¼ ; The rate of thermal energy available from the IPVT
LT U tT þ hT
collector, Eq. (5), is transferred and stored in an insu-
 1 lated storage tank. The energy balance for the storage
Li 1
Ub ¼ þ and tank is given by
K i hi
  dT w
_ w
mC U F f Ac
ð LmC
Q_ u ¼ M w C w þ ðUAÞT ðT w  T a Þ ð7Þ
FR ¼ 1  e _ w Þ with Ac ¼ bL dt
Ac U L F 0
Further, it is important to note that there is no strat-
0
The calculation of F is given in Appendix A. ification of temperature in the storage tank due to forced
Eq. (5) is almost similar to the equation for the rate mode of operation. Hence one can assume that Tfi = Tw
of useful energy for a flat plate collector. Further it as shown in Fig. 1. Now Eqs. (5) and (7) can be com-
can be used to get an expression for an instantaneous bined and rewritten in the form of a first order differen-
thermal efficiency of the PV/T system as tial equation as
  dT w
Q_ u T fi  T a þ aT w ¼ f ðtÞ ð8Þ
gi ¼ ¼ F R hp1 hp2 ðasÞeff  U L ð6Þ dt
A c I ðt Þ I ðt Þ
where
which is almost same as the well-known Hottel–Whiller–
Bliss characteristic equation for flat plate collector (Eqs. ðUAÞT þ Ac F R U L
a¼ and
(3.57) and (3.58) of Tiwari, 2004) the difference being the M w Cw
two penalty factor, hp1hp2, which are not present in the 1
f ðtÞ ¼ ½Ac F R hp1 hp2 ðasÞeff IðtÞ þ fðUAÞT þ Ac F R U L gT a 
Hottel–Whiller–Bliss equation. M w Cw
756 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759

An analytical solution of Eq. (8) can be obtained the wind speed observed over the course of their tests.
with the following assumptions: This affects the convective heat transfer coefficient
between the glass surface and the ambient air, and hence
• ‘a’ is considered as constant over the time interval 0–t the heat flow from the PV cells to ambient through the
and glass. In Table 1 we have assumed a wind speed of
• f(t) is considered as f ðtÞ over the time interval 0–t. 1 m/s for comparison with the experimental data, but
additionally performed calculations for different wind
The solution of Eq. (8) with initial condition i.e. speeds, also reported below.
Twjt=0 = Tw0 becomes By using the data of Table 1, the gain factor and loss
terms of Eq. (6) for FR = 0.99 (see Appendix A) for the
f ðtÞ instantaneous efficiency can be calculated as follows:
Tw ¼ ð1  eat Þ þ T w0 eat ð9Þ
a
Gain factor ðgÞ ¼ F R hp1 hp2 ðasÞeff
After obtaining Tw from Eq. (9), the temperatures of
the PV module and base of PV module can be obtained ¼ 0:99  0:8774  0:984  0:66 ¼ 0:56
from Eqs. (2a) and (2b) for given climatic data and
design parameters of the unglazed IPVTS. Expressions The overall heat loss coefficient for IPVTS is given by
for Tbs and Tc are given below: UL = Utw + Ub where

hp1 ðasÞeff IðtÞ þ U tT T a þ hT T w hT U tT

T bs ¼ ð10Þ U tw ¼ ¼ 7:98 W=m2 K and
U tT þ hT hT þ U tT
U b ¼ 0:62 W=m2 K ðTable 1Þ
and
Hence UL = 7.98 + 0.62 = 8.6 W/m2 K.
sG ½ac bc þ aT ð1  bc ÞI ðtÞ  gc I ðtÞbc þ U t T a þ U T T bs The numerical value of UL is almost comparable to
Tc ¼
Ut þ UT the value obtained for single glazed flat plate collectors
ð11Þ (Tiwari, 2004).
However, the value of the term describing heat loss is
The cell temperature of module without water flow given by
can be written as    
T fi  T a T fi  T a
sG ½ac bc þ aT ð1  bc ÞI ðtÞ þ ðU t þ U T a ÞT a F RU L ¼ 8:51 where IðtÞ is in W=m2
T 0c ¼ ð12Þ IðtÞ IðtÞ
Ut þ UTa
where U T a ¼ ½KLTT þ h1i 1 .
To compare the results of the calculations with the 4.2. Numerical results
experimental results, the coefficient of correlation (r)
and root mean square percent deviation (e) have been The hourly data of solar radiation [I(t)] and ambient
evaluated by using the following expressions: air temperature obtained from Fig. 4 of Huang et al.
P P P (2001) are shown in Table 2. The design and climatic
nð X  Y Þ  ð X Þ  ð Y Þ data of Tables 1 and 2 have been used to evaluate water
r ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2  P qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
P 2  P ffi ð13Þ
n X  ð X Þ2  n Y  ð Y Þ2 temperature and cell temperature from Eqs. (9)–(11).
The results are reported in Figs. 3–5. The correlation
and coefficient and root mean square percent deviation have
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi also been evaluated and given in the same figures
P
ðe i Þ2 obtained by using Eqs. (13) and (14) respectively.
e¼ ð14Þ
N

where ei ¼ ½X iXYi i   100.

Table 2
Intensity and ambient temperature on 21/05/99
4. Results and discussion Time (h) Intensity (W/m2) Ambient temperature (C)
8 580 25
4.1. Analytical results 9 560 29
10 784 32
The design parameters and various heat transfer 11 852 34
coefficients of the integrated photovoltaic–thermal sys- 12 812 35
tem (IPVTS) used in the present study are given in Table 13 660 35
1. They correspond to the experimental system described 14 604 36
15 486 34
by Huang et al. (2001), except that they did not report
 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759 757

55.00 Tc increase in water temperature are in accordance with

Tw
the results reported by Huang et al. (2001).
50.00
Ta
Further, it can be observed from Fig. 4 that there is
45.00
fair agreement between experimental value of cell tem-
Temperature,°C

perature, Tc(exp) and the theoretical value, Tc(th). The

40.00 correlation coefficient and root mean square deviation
are found to be 0.98% and 7.22% respectively.
35.00 Fig. 5 gives the hourly variations of theoretical and
experimental water temperatures in the storage tank.
30.00
Although the theoretical value is higher than the exper-
25.00
imental value, the correlation coefficient and root mean
8 9 10 11 12 13 14 15 square percent deviation are 0.99% and 5.87% respec-
Time, hr tively, which supports the validity of the present model
developed in the present study.
Fig. 3. Hourly variation of temperature with time: solar cell
(Tc), ambient (Ta) and water (Tw).
The good agreement shown between experiment and
theory suggests that the assumption of a wind speed of
1 m/s made in the calculations is reasonable.
55.00
r = 0.98
The effect of mass flow rate on the hourly variation of
Tc(EXP)
e = 7.22% water temperature is shown in Fig. 6. It can be seen that
50.00
Tc(TH) the flow rate has only a small effect on the hourly vari-
ation of water temperature over the range of flow rates
Temperature,°C

45.00 considered. Hence one can conclude that the optimum

mass flow rate lies between 0.005 and 0.075 kg/s.
40.00 Fig. 7 shows the hourly variation of water tempera-
ture for different air speeds. It is evident from the figure
35.00 that as the air speed over the panel increases, the water
temperature decreases because of the increase in heat
30.00 loss from PV module through the glass.
8 9 10 11 12 13 14 15 The effect of the length of the PV module on the
Time, hr
hourly variation of water temperature is shown in
Fig. 4. Hourly variation of solar cell temperature—experimen- Fig. 8. The water temperature increases with increasing
tal and theoretical. length as expected. This is due to fact that the net ther-
mal energy available to the system increases on account
of the increased area exposed to the incident solar radi-
53.00 ation. It is also important to note that there is only mar-
Tw(EXP) r = 0.99
e = 5.87% ginal increase in water temperature for lengths greater
Tw(TH)
48.00 than 4 m. Hence one can conclude that the optimum
length of the PV module is 4 m for the present set of
Temperature,°C

43.00 design and climatic parameters.

38.00
50.00

48.00 0.005
0.016
33.00 46.00
0.028
44.00 0.075
Temperature,°C

28.00 42.00
8 9 10 11 12 13 14 15
40.00
Time, hr
38.00
Fig. 5. Hourly variation of water temperature—experimental 36.00
and theoretical.
34.00

32.00

From Fig. 3, it is evident that the solar cell tempera- 30.00

8 9 10 11 12 13 14 15
ture is higher than the water temperature as expected.
Time, hr
There is an increase in water temperature of about
17 C with respect to an ambient air temperature in Fig. 6. Hourly variation of water temperature with different
the evening. The trend in temperature variation and an mass flow rate in kg/s.
758 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759

53.00 50.00
1 300
2 48.00 500
3 700
48.00 4 46.00 1000
5 44.00

Temperature,°C
Temperature,°C

43.00 42.00

40.00

38.00 38.00

36.00

33.00 34.00

32.00

28.00 30.00
8 9 10 11 12 13 14 15 8 9 10 11 12 13 14 15
Time, hr Time, hr

Fig. 7. Hourly variation of water temperature with different Fig. 10. Water temperature with different values of the film
velocity of air in m/s. heat transfer coefficient in W/m2 K.

storage temperature are small. This indicates that the

75.00 1
2
optimum value of water mass is about 60 kg for present
70.00
3 set of design and climatic parameters, depending upon
65.00 4
5
requirement of water temperature.
Temperature,°C

60.00
The hourly variation of water temperature with dif-
55.00 ferent film heat transfer coefficients is shown in
50.00 Fig. 10. It shows that there is little variation in water
45.00 temperature for the various values of film heat transfer
40.00 coefficient considered in the present study.
35.00

30.00 4.3. An overall thermal efficiency

8 9 10 11 12 13 14 15
Time, hr
An overall thermal efficiency of the PV/T system can
Fig. 8. Hourly variation of water temperature with different be evaluated as follows:
module length in meters.
M w C w ðT f  T fi Þ
gth ¼ P
Ac IðtÞ  3600
Fig. 9 shows the hourly variation of water tempera- By using the results of Fig. 2 for DT = 48.57 
ture with different mass of water in the system. As the 30.39 = 18.18 C and climatic data of Table 2, the over-
mass increases (and hence the thermal capacity of all thermal efficiency of PV/T system is calculated to be
the water in the system increases), the temperature of 34.55%.
the water in the storage tank decreases. For a mass The overall thermal efficiency of the IPVTS system is
of water greater than 60 kg, further reductions in the given by
g ¼ gc =gp þ gth ¼ ðge =0:38Þ þ gth
63.00
20
¼ ð0:09=0:38Þ þ 0:3455 ¼ 0:58
58.00 40
60 This is slightly less than the experimentally evaluated
80
53.00 overall efficiency of about 61.3% (Table 1, Huang
Temperature,°C

48.00
et al., 2001). This is due to lower water temperatures pre-
dicted by the thermal model as shown in Fig. 5.
43.00

38.00
5. Conclusions
33.00

28.00 The thermal model of IPVTS presented in this study

8 9 10 11 12 13 14 15
is in good agreement with the experimental result
Time, hr
obtained earlier (Huang et al., 2001). The study also
Fig. 9. Hourly variation of water temperature with different indicates that there is a significant increase in the overall
mass of water in the system in kg. thermal efficiency of IPVTS from 24% (0.09/0.38) to
 A. Tiwari, M.S. Sodha / Solar Energy 80 (2006) 751–759 759

W Here, the mass flow rate (m) _ has been evaluated for
D
known experimental values of average intensity
. . (670 W/m2), temperature difference (20 C) and an effi-
Qu′ Qu″ K δ ciency of collector of IPVTS (50%).
The above values have been taken to evaluate the
water temperature and solar cell temperature of the
PV module for validation.

Fig. 11. Schematic view of the finned-tube configuration of the

flat plate collector, considered by Huang et al. (2001). References
58% because of additional thermal energy due to the Baker, L.H., 1967. Film heat transfer coefficient in solar
water flow. collector tubes at low Reynolds number. Solar Energy 11,
78–85.
Chow, T.T., 2003. Performance analysis of photovoltaic–
Appendix A
thermal collector by explicit dynamic model. Solar Energy
75, 143–152.
A schematic view of the finned-tube configuration of Duffie, J.A., Beckman, W.A., 1991. Solar Engineering of
the flat plate collector, considered by Huang et al. Thermal Processes, second ed. Wiley Interscience, New
(2001), is shown in Fig. 11. York.
Referring to Fig. 11 and following Duffie and Beck- Evans, D.L., 1981. Simplified method for predicting PV array
man (1991) and Tiwari (2004), one has output. Solar Energy 27, 555–560.
Huang, B.J., Lin, T.H., Hung, W.C., Sun, F.S., 2001. Perfor-
UL 8:1
m2 ¼ ¼ ¼ 66:18 mance evaluation of solar photovoltaic/thermal system.
Kd 204  0:0006 Solar Energy 70, 443–448.
or Infield, D., Mei, L., Eicker, U., 2004. Thermal performance
estimation of ventilated PV facades. Solar Energy 76, 93–98.
m ¼ 8:134 Jones, A.D., Underwood, C.P., 2001. A thermal model for
The fin efficiency factor (F) is given by photovoltaic systems. Solar Energy 70, 349–359.
Kalogirou, S.A., 2001. Use of TRYNSYS for modeling and
tanh m W D
2 simulation of a hybrid PV-thermal solar system for Cyprus.
F ¼
m W D
2
Renewable energy 23, 247–260.
Prakash, J., 1994. Transient analysis of a photovoltaic–thermal
For the values of W = 0.04 m and D = 0.006 m solar collector for co-generation of electricity and hot air/
F ¼ 0:99 water. Energy Conversion and Management 35, 967–972.
Schott, T., 1985. Operational temperatures of PV modules. In:
The flat plate collector efficiency (F 0 ) is given by Proceedings of 6th PV Solar Energy Conference, pp. 392–
396.
1
F 0 ¼ WU L with hT ¼ 500 W=m2  C Tiwari, G.N., 2004. Solar Energy: Fundamentals, Design
Dh
þ WUK L þ DþðWU L
W DÞF Modeling and Applications. CRC Press, New York and
d
Narosa Publishing House, New Delhi.
The value of hT has been evaluated by using the expres- Tiwari, A., Sodha, M.S., Chandra, A., Joshi, J.C., 2005.
sion given by Baker (1967). Performance evaluation of photovoltaic thermal solar air
By using the values of various parameters in the collector for composite climate of India. Solar Energy
above equation, one gets Materials and Cells, in press.
Tripanagnostopoulos, Y., Nousia, T.H., Souliotis, M., Yia-
F 0 ¼ 0:895 noulis, P., 2002. Hybrid photovoltaic/thermal solar system.
Now, the flow rate factor (FR) given by Solar Energy 72, 217–234.

 Zondag, H.A., de Vries, D.W., van Helden, W.G.J., van
mC_ f Ac U L F 0 Zolengen, R.J.C., van Steenhoven, A.A., 2002. The thermal
FR ¼ 0 1  exp  ¼ 0:9887
Ac U L F mC_ f and electrical yield of a PV–thermal collector. Solar Energy
ffi 0:99 for data of Table 1 72 (2), 113–128.