HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004

PHOTOVOLTAIC THERMAL (PV/T) COLLECTORS: A REVIEW

P.G. Charalambousa, S.A. Kalogiroub , G. Maidmenta and T.G. Karayiannisa

a
Department of Engineering Systems, London South Bank University,
103 Borough Road, London, U.K. SE1 0AA
b
Higher Technical Institute, P.O.Box 20423, Nicosia 2152, Cyprus

ABSTRACT

An extensive amount of research on PV-

thermal collectors has been carried out over The purpose of this extraction device is
the last 25 years. This paper aims at presenting twofold. Firstly, to cool the PV module and
a review of the most available literature on thus improve its electrical performance and
PV/T collectors. The review is presented in a secondly to collect the thermal energy
thematic way, in order to enable an easier produced, which would have otherwise been
comparison of the findings obtained by various lost as heat to the environment. This collected
researchers, especially on parameters affecting heat could be used, for example, for space
PV/T performance (electrical and thermal). heating or for domestic uses (showers and
The review covers analytical and numerical washing).
models, simulation and experimental work and
qualitative evaluation of thermal/electrical A considerable amount of work has been
output. carried out during the last 25 years and a
review of the most important available
literature on liquid and air PV/T collectors is
1. INTRODUCTION presented.

Photovoltaic thermal (PV/T) collectors are

devices that simultaneously convert solar 2. PERFORMANCE OF PV/T
radiation into electricity and heat. A PV/T COLLECTORS
collector shown in figure 1, typically consists
of a PV module on the back of which an 2.1 Analytical Models
absorber plate, a heat extraction device, is
attached. Several analytical and numerical models have
been developed, by various researchers,
Glass cover
PV module predicting the performance of PV/T collectors.
Adhessive The steady state thermal efficiency (ηth) of a
conventional flat plate solar collector is
calculated by (Duffie and Beckman, 1991):

Qu
ηth = (1)
G
Absorber
Water tubes and the useful collected heat (Qu) is given by:
Insulation
Qu = mC
& p (To − Ti ) (2)

Fig. 1 Cross section of a PV/T collector

1
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
Or, it is simply the difference between the efficiency factor, F’ and the heat removal
absorbed solar radiation and the heat losses: factor, FR) still apply. It was concluded that for
practical purposes F’ and FR of the PV/T
Qu = Ac [ S − U L (Tp ,m − Ta )] (3) collector may be considered identical to F’ and
As in the previous equation the mean absorber FR of the thermal collector.
plate temperature (Tp,m) is difficult to calculate Bergene and Lovvik (1995), proposed a
or measure since it is a function of collector detailed model predicting the performance of
design, the incident solar radiation and the PV/T collectors that was based on energy
entering fluid conditions, Hottel and Whillier transfer analysis and to some extent on the
(as explained in Duffie and Beckman, 1991) models for flat plate solar collectors presented
modified equation (3) as follows: by Duffie and Beckman (1991). The model
predicts PV/T efficiency (thermal + electrical)
to be about 60 to 80%.
Qu = Ac FR [ S − U L (Ti − Ta )] (4)
where, Sandnes and Rekstad (2002) have also
& p ⎡ − c L ⎤
AU F'
mC mC
& p developed an analytical model for the PV/T
FR = ⎢1 − e ⎥ (5) collector by modifying the well known models
AcU L ⎢ ⎥
⎣ ⎦ for flat plate collectors to include the effects of
1 the additional solar cells. Good agreement
UL between the simulation and the experimental
F'= (6) results was found.
⎡ 1 1 1 ⎤
W⎢ + + ⎥
⎢⎣ U L [ Do + (W − Do ) F ] Cb π Di h f i ⎥⎦ Sopian et al. (1996) analyzed with steady state
models the performance of single and double
tanh( x) pass PV/T air collectors. The results showed
F= (7) that double-pass PV/T collectors have superior
x
and efficiencies. Typically thermal, electrical and
combined (thermal + electrical) efficiencies for
U L ⎛ W − Do ⎞
x= (8) single-pass collectors were 24-28, 6-7, 30-
kδ ⎝⎜ 2 ⎠⎟ 35%, respectively and for double-pass
collectors were higher at 32-34, 8-9, 40-45%,
respectively. In addition, thermal and
The electrical efficiency (ηel) of a PV/T combined efficiencies increased as the packing
collector is calculated by factor (defined as the fraction of the absorber
(Tripanagnostopoulos et al. 2002): area covered by photovoltaic cells) decreased.
However, electrical efficiency decreased
I m Vm slightly. It was emphasized that the improved
ηel = (9) performance of the double-pass PV/T collector
GAc
(compared to the single-pass) was achieved at
very little increase in collector capital cost.
and typically, the dependence of the PV
module electrical efficiency on the module
2.2 Numerical models
temperature is given by (Zondag et al. 2003):
Zondag et al. (2002) prepared and run four
(
ηel = ηo 1 − 0.0045 ⎡⎣T − 25o C ⎤⎦ ) (10) numerical models for predicting PV/T collector
yield. One 3D dynamic and three steady state
Florschuetz (1979), extended the well known (3D, 2D and 1D) models. The simple 1D
Hottel-Whillier analytical model for flat plate steady state model performed almost as good
collectors so that with simple modifications it as the much more time consuming 3D dynamic
applies to combined PV/T collectors and all model. However, the 2D and 3D models
existing relations and supporting information provide more detailed information required for
available in the literature (such as the collector further collector optimization.
2
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
electrical efficiency as a result of reduced glass
Since the 3D dynamic model developed by transmissivity. None of the other features
Zondag et al (2002) was an extensive one significantly affected electrical performance.
(typically uses 2.5 hr simulation time for 1 hr The efficiencies with and without a selective
real life equipment operation), Chow (2003) absorber were virtually the same. The
developed an explicit dynamic model, based on dominant control of the electrical efficiency
control volume finite difference approach for a was the glass transmissivity. The optimum
single glazed PV/T collector. The model can combination of an air PV/T collector was
generate results for hourly performance found to consist of gridded-back PV cells, a
analysis, including instantaneous thermal / nonselective secondary absorber and a high
electrical gains and efficiencies. It was found transmissivity / low emissivity glass above the
that the maximum combined efficiency of a photovoltaic cells.
perfect collector can be over 70% and can
decrease to less than 60% for a low quality 2.4 Experimental work
collector.
Tripanagnostopoulos et al. (2002) built and
2.3 Modeling and simulation tested various PV/T collector models with both
water and air as working fluids. The
Kalogirou (2001) modeled and simulated a performance of these models was boosted by
PV/T system using the well known TRNSYS diffused reflectors made of flat aluminum
simulation program and a typical sheets. It was found that the water heat
meteorological year for Nicosia, Cyprus. The extraction results to a lower PV temperature
annual electrical efficiency of the standard PV than that of air heat extraction because water
system increased from 2.8% to 7.7% for the temperature from mains (20oC) was lower than
PV/T system operating at the obtained that of ambient air (29oC). The authors believe
optimum flow rate (25 l/hr). that the much higher heat transfer coefficient
of water, compared to that of air, is a second
Garg and Adhikari (1997) simulated the possible reason. Therefore, there is an
performance of single and double glass advantage of using water instead of air as heat
configurations PV/T air heating collectors. removal fluid especially during the warm
They found that increasing cell density results periods. The electrical efficiency of the basic
in very large values of electrical efficiency, PV/Twater model was 3.2% higher than that of
although thermal efficiency drops. The the simple PV module and an additional
combined efficiency increased with increase in advantage of using PV/T instead of plain PV
collector length, mass flow rate and cell modules comes from the higher thermal output.
density and decreased with increases in duct Moreover, a PV/T system with booster diffuse
depth. reflectors of concentration, C=1.35, can
achieve a percentage increase of electrical
Cox and Raghuraman (1985) performed efficiency up to 19.2% (compared to that of the
computer simulations towards improving the simple PV module), at an additional PV/T cost
solar absorptance and reducing the IR for the reflectors of only about 4%.
(infrared) emittance of flat plate air PV/T
collectors. They found that air PV/T collectors Lalovic et al. (1986) built a PV/T collector
are generally less efficient than liquid ones due using amorphous silicon photovoltaic cell and
to low PV cell packing factor, low solar its performance was tested. The diameter of
absorptance, high IR emittance and poor the copper tubes was 12 mm and tube spacing
absorber to air heat transfer coefficient. A low was 12 cm. The transmittance-absorptance
emissivity layer was added to the PV cells so product (τα) for the hybrid collector was rather
that the resulting combination was effectively a low, i.e. 0.53 and the slope of the performance
selective absorber. Moreover, low iron glass characteristic was also low 3.25 W/mK and
resulted in high thermal efficiency due to showed that the collector behaved as a double
reduced top loss coefficient and lower glazed collector with a selective surface. It
3
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
was noted that in order to increase the Zondag et al (2003) reported that the
efficiency of PV/T collectors, the aluminum uncovered sheet and tube collectors perform
back electrode of PV panels could be replaced poorest due to large heat losses. Moreover,
by indium tin oxide (ITO) which is transparent electrical efficiency of sheet and tube
to solar spectrum above 0.5 µm. In this way a collectors with two covers strongly deteriorates
part of the solar radiation not absorbed by the due to the second cover.
amorphous layers would pass onto the
absorber plate, which now has to be black. 2.5.2 Mass flow rate
Bergene and Lovvik (1995) found that the
Sopian et al. (2000) developed and tested a thermal efficiency may increase only by a
double pass PV/T collector suitable for solar factor of 0.10 if flow rate increases from 0.001
drying applications. Comparisons were made to 0.075 kg/s (the PV/T collector area was not
between the experimental and the theoretical given). Consequently, they suggest that when
results and close agreement between the two the flow rate is around 0.001 kg/s, there is not
values were obtained. much to gain on increasing it further. It was
also pointed out that with respect to electrical
2.5 Parameters (factors) affecting PV/T efficiency, flow rate is one of the most
performance important parameters (the other being inlet
fluid temperature). Moreover, at low flow rates
A number of parameters have been identified the electrical efficiency increases when W/D
to affect PV/T performance. These include increases, whereas at high flow rates the
mass flow rate, inlet temperature of working opposite occurs. In his study, Chow (2003)
fluid, number of covers, absorber to fluid showed that as mass flow rate in the tube
thermal conductance and absorber plate design increases from 0.002 to 0.016 kg/s, for a 2 m2
parameters such as tube spacing, tube diameter PV/T collector area (i.e. 0.001 to 0.008
and fin thickness. An analysis of these kg/sm2), the thermal and electrical efficiencies
parameters follows: also increase.

2.5.1 Covered vs uncovered PV/T Garg and Agarwal (1995), carried out
collector simulations for different solar cell areas, mass
Sandnes and Rekstad (2002) explained that the flow rates and different water masses by
effect of adding a glass cover to the PV/T solving the governing equations using an
collector, is to reduce the heat losses to the iterative finite difference method. The system
surroundings. However, the energy was composed of a PV/T collector, storage
absorptance is also reduced by reflection tank, pump and differential control. The
(around 10%) from the glass. They found that optimum flow rate was found to be 0.03 kg/s,
the simulated total electrical energy output for a 2 m2 PV/T collector area (i.e. 0.015
over a day for the plain PV module was 306.9 kg/sm2), for maximum thermal collector
Wh, for the PV/T without glass cover was efficiency. However, electrical efficiency was
339.3 Wh and for the PV/T with glass cover found to decrease at 0.03 kg/s and was
was 296.2 Wh. minimum when solar insolation was maximum
(which is expected as at this time absorber
Fujisawa and Tani (1997) found that the temperature is maximum); the average
thermal performance of the single covered electrical efficiency and the average daily
PV/T was as high as that of the flat plate combined efficiency increased as the total
collector (FPC) and that of the coverless PV/T water mass increased (by increasing number of
collector was inferior owing to the lack of heat- tubes, i.e. decreasing tube spacing) and the
insulating layer of air. On the other hand, the higher the mass flow rate was, the higher the
coverless PV/T collector produced the highest solar cell efficiency.
electrical energy.
Morita, Fujisawa and Tani (2000) determined
that maximum exergetic efficiencies for single
4
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
cover (of 13.36%) and coverless (of 11.92%) or increase costs for what may only be
PV/T collectors occur at optimum flow rates marginal improvements in performance.
of 0.0014 and 0.0049 kg/s, respectively, for a
PV/T collector area of 0.61 m2 (i.e. 0.002 and Chow(2003), refers to the two manufacturing
0.008 kg/sm2, respectively). defects found in PV/T collectors (imperfect
adhesion between PV plate and absorber plate,
Furthermore, Kalogirou (2001) in his imperfect bonding between absorber plate and
simulations using TRNSYS, found the tubes) and for a range of thermal conductances
optimum flow rate to be 25 l/hr (0.007 kg/s), 10000 W/mK to 25 W/mK (perfect to
for a 2.54 m2 PV/T collector area (i.e. 0.003 defective), found that the maximum combined
kg/sm2). efficiency of a perfect collector can be over
70% and for a low quality collector, may
Garg and Adhikari (1997) & (1999) and decrease to less than 60%.
Sopian et al (2000), in testing the performance
of PV/T air heating collectors noticed that 2.5.5 Design types
increasing air mass flow rate reduces cell and Zontag et al. (2003) compared the efficiency of
outlet temperatures and consequently seven different design types of PV/T
increases thermal and electrical efficiency. collectors. They observed that:
a) All channel concepts have a substantially
2.5.3 Absorber plate parameters higher efficiency than sheet and tube due to
Bergene and Lovvik (1995), elaborated on the the better heat transfer characteristics of
effect of tube spacing to tube diameter ratio channels.
(W/D). It was found that: b) In the case of free flow panel, evaporation
a) The thermal efficiency is approximately strongly reduces the thermal efficiency and
halved when W/D increases from 1 to 10, condensate on top of the glass causes
by keeping W constant. It was also additional reflection.
emphasized that different results are c) Since the sheet and tube design is the easiest
expected when increasing W whilst to manufacture (and is only 2% less in
keeping D constant. efficiency), it is the most promising of the
b) The fact that the speed of cooling liquid different design concepts examined.
increases when tube diameter is decreased
does not compensate for losses from the
fin. 3. QUALITATIVE EVALUATION OF
c) Increasing W/D from 1 to 10, decreases THERMAL / ELECTRICAL OUTPUT
outlet fluid temperature.
d) Even though electrical efficiency is not Electrical and thermal energies are not
heavily affected by fin size, combined qualitatively the same as explained by
efficiency is largely dependent on fin size. Fujisawa and Tani (1997); thermal energy
e) If thermal efficiency is of any importance, cannot produce work until a temperature
its dependence on the relative tube diameter difference exists between a high temperature
should be weighted against the cost of the source and a low temperature source, but
tubes electrical energy can completely transform
into work irrespective of the ambient
2.5.4 Absorber to fluid thermal conductance conditions. Use of exergy, defined as the
As Florschuetz (1979) pointed out, exceedingly maximum theoretical useful work obtainable
large values of absorber to fluid thermal from a system as it returns to equilibrium with
conductance, Uf, are not required for the heat the environment, enables qualitative
removal factor, FR, to be within a reasonable evaluation by comparing electrical and
value, especially with collectors of at least one thermal energy based on the same standard.
glazing. Design approaches must be The results showed that the coverless PV/T
undertaken carefully since they may increase collector produced 8% more electrical energy
the pressure drop to unacceptably high values than a standard PV module did and produced
5
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
41% of the thermal exergy of the FPC. 40% efficient and it is thus equivalent to
Moreover, the coverless PV/T collector assuming an electrical/thermal value ratio of
produced the largest available total (electrical 2.5. However, low temperature hot water from
+ thermal) exergy of 80.8 kWh, whereas the a PV/T is not as thermodynamically useful as
PV module and the FPC produced 72.6 kWh high temperature steam from a coal-fired
and 6.0 kWh, respectively. It is clear that the boiler. For the renewable energy market
total exergy of the coverless PV/T collector methodology, the grid connected photovoltaics
was 11% higher than the PV module and levelised energy cost (defined as the unit price
1287% higher than the FPC. The output of energy output that would result in the
density (exergy gain divided by installation system having a zero net present value ‘NPV’
area) of the coverless PV/T was the highest, over its lifetime) was found to be US $
76% higher than with use of separate PV 0.367/kWh and the solar hot water levelised
module and FPC (the output density of the energy was found to be US $ 0.087/kWh.
single cover PV/T was 57% higher). Thus, the ratio of electrical to thermal value is
4.24. This means that electricity is 4.24 times
Morita, Fujisawa and Tani (2000) performed a more valuable than hot water. In the
numerical analysis and determined the environmental valuation approach, based on
optimum operating conditions by using avoided emissions from the use of PV/T, the
exergetic evaluation for single and no cover ratio of electrical to thermal avoided CO2
PV/T collectors. They developed a steady state emissions was found to be 7.58 (including life
energy equilibrium equation based on the cycle emissions). They suggest that the most
analytical model on PV/T collectors of Cox realistic energy ratio for a PV/T collector is the
and Raghuraman (1985). As mentioned earlier, renewable energy market (i.e. 4.24), although
the maximum exergetic efficiencies were for they admit that there is no simple answer for
single cover 13.36% and for coverless 11.92%. determining what energy value ratio should be
The corresponding optimum flow rates were used, rather the ratio should be a parameter
0.0014 and 0.0049 kg/s, respectively and the selected for the circumstance applicable to a
corresponding optimum fluid temperatures particular installation.
were 83.6oC and 38.3oC, respectively. It is
clear that the cover glass has the function of
raising the maximum exergetic efficiency and 4. CONCLUSIONS
the optimum outlet temperature. It is also clear
that in order to obtain the most thermal exergy, A review of the available literature on liquid
the electrical exergy has to be sacrificed. On and air PV/T collectors which covers the work
the other hand, when one desires to obtain of the last 25 years was presented. The
positive electrical exergy, coverless PV/T is following conclusions have been reached:
more useful than PV/T with cover as the total
exergetic efficiency of coverless PV/T and It was found from analytical and numerical
medium fluid temperature (25oC), surpasses models that PV/T efficiencies could range
that of PV/T with cover (11.28% and 10.31%). from over 70% for a perfect collector and to
less than 60% for a low quality collector.
The significance of an exergy comparison is
not clear, according to Coventry and Air PV/T collectors are generally less efficient
Lovegrove (2003), if electrical or mechanical than liquid PV/T collectors. Moreover, since
work is not the only desired output from the the sheet and tube design is the easiest to
system, such as when the thermal output is hot manufacture and is only 2% less efficient, it is
water used directly for showers and washing. the most promising of design concepts.
They used three different methodologies
(thermodynamic, market based and For practical purposes the collector efficiency
environmental) for determining an electrical- factor (F’) and the heat removal factor (FR) of
to-thermal energy value ratio. For the a PV/T collector could be considered identical
thermodynamic methodology, a power plant is
6
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
to the corresponding ones of a thermal Cp specific heat of fluid [J/kg K]
collector. Di inside tube diameter [m]
Do outside tube diameter [m]
The thermal performance of a coverless PV/T F fin efficiency
collector is reduced especially at high F’ collector efficiency factor
temperatures due to the lack of heat-insulating FR heat removal factor
layer of air. However, coverless PV/T G solar irradiance [W/m2]
collectors have a better electrical performance. hfi heat transfer coefficient of
fluid [W/m2K]
The optimum flow rate was found to be in the Im PV current at maximum power
range of 0.001 to 0.008 kg/sm2, whereas a point [A]
value of 0.015 kg/sm2 was also reported and k thermal conductivity of the
thus, optimum flow rate studies could be fin [W/mK]
investigated further. m& fluid mass flow rate per unit
collector area [kg/s m2]
The thermal efficiency was approximately Qu useful collected heat by
halved when the tube spacing to tube diameter collector [W/ m2]
ratio (W/D) was increased from 1 to 10 (by S absorbed solar energy [W/ m2]
keeping D constant). It was emphasized that T temperature of PV module [K]
different results could be expected when Ta temperature of the ambient [K]
increasing W whilst keeping D constant and Ti fluid inlet temperature [K]
further investigation on this issue should be To fluid outlet temperature [K]
carried out. Tp,m average plate temperature [K]
UL overall collector heat loss
Exceedingly large values of absorber to fluid coefficient [W/m2K]
thermal conductance are not required for the Vm PV voltage at maximum
heat removal factor FR to be within a power point [V]
reasonable value, especially with collectors of W tube spacing [m]
at least one glazing. Greek Symbols
δ fin thickness [m]
It was suggested that the most realistic energy ηel PV/T electrical efficiency
value ratio is the renewable energy market (i.e. ηo PV electrical efficiency at
4.24), although there is no simple answer in standard conditions
determining what energy value ratio should be ηth PV/T thermal efficiency
used in comparing electrical and thermal τα transmittance absorptance product
energy output, of PV/T collectors, based on the
same standard.
REFERENCES
Finally, further work should be carried out
aiming at improving the efficiency of PV/T Bergene T. and Lovvik O.M. (1995) ‘Model
collectors and reducing their cost, making them calculations on a flat plate solar heat collector
more competitive and thus aid towards global with integrated solar cells’ Solar Energy 55,
expansion and utilization of this pp. 453-462.
environmentally friendly renewable energy
device. Chow T.T. (2003) ‘Performance analysis of
photovoltaic-thermal collector by explicit
dynamic model’ Solar Energy 75, pp. 143-
NOMENCLATURE 152.

Ac PV/T collector area [m2] Coventry J.S. and Lovegrove K. (2003)

Cb conductance of the bond between the ‘Development of an approach to compare the
fin and tube [W/mK] ‘value’ of electrical and thermal output from a
7
 HPC 2004 – 3rd International Conference on Heat Powered Cycles, Cyprus, October 2004
domestic PV/thermal system’ Solar Energy and analytical model’ Solar Energy 72, pp. 63-
75, pp. 63-72. 73.

Cox C.H. and Raghuraman P. (1985) ‘Design Sopian K.S., Yigit H.T., Liu H.T., Kakac
considerations for flat-plate S.and Veziroglu T.N. (1996) ‘Performance
photovoltaic/thermal collectors’ Solar Energy analysis of photovoltaic/thermal air heaters’
35(3), pp. 227-241. Energy Convers. Mgmt 37(11), pp. 1657-
1670.
Duffie J.A. and Beckman W.A. (1991) Solar
engineering of Thermal Processes, 2nd edn. Sopian K.S., Liu H.T., Kakac S.and Veziroglu
New York: John Wiley and Sons, Inc. T.N. (2000) ‘Performance of a double pass
photovoltaic thermal solar collector suitable
Florschuetz L.W.(1979) ‘Extension of the for solar drying systems’ Energy Convers.
Hottel-Whillier model to the analysis of Mgmt 41, pp. 353-365.
combined photovoltaic/thermal flat plate
collectors’ Solar Energy 72, pp. 361-366. Tripanagnostopoulos Y., Nousia Th., Souliotis
M. and Yianoulis P. (2002) ‘Hybrid
Fujisawa T. and Tani T. (1997) ‘Annual exergy photovoltaic/thermal solar systems’ Solar
evaluation on photovoltaic-thermal hybrid Energy 72(3), pp. 217-234.
collector’ Solar Energy Materials and Solar
Cells 47, pp. 135-148. Wysocki J.J. and Rappaport P. (1960) ‘Effect
of temperature on photovoltaic solar energy
Garg H.P. and Agarwal R.K. (1995) ‘Some conversion’ Journal of Applied Physics, 31(3),
aspects of a PV/T collector/forced circulation pp. 571-578.
flat plate solar water heater with solar cells’
Energy Convers. Mgmt 36(2), pp. 87-99. Zondag H.A., Vries D.W., Van Hendel W. G.
J., Van Zolingen R.J.C. and Van Steenhoven
Garg H.P. and Adhikari R.S. (1997) A.A. (2002) ‘The thermal and electrical yield
‘Conventional hybrid photovoltaic / thermal of a PV-thermal collector’ Solar Energy, 72,
(PV/T) air heating collectors: steady-state pp. 113-128.
simulation’ Renewable Energy 11(3), pp. 363-
385. Zondag H.A., Vries D.W., Van Hendel W. G.
J., Van Zolingen R.J.C. and Van Steenhoven
Kalogirou S.A. (2001) ‘Use of TRNSYS for A.A. (2003) ‘The yield of different combined
modelling and simulation of a hybrid PV- PV-thermal collector designs’ Solar Energy
thermal solar system for Cyprus’ Renewable 74, pp. 253-269.
Energy 23, pp. 247-260.

Lalovic B. (1986) ‘A hybrid amorphous silicon

photovoltaic and thermal solar collector’ Solar
Cells 19, pp. 131-138.

Morita Y., Fujisawa T. and Tani T. (2000)

‘Moment performance of photovoltaic/thermal
hybrid panel (Numerical analysis and exergetic
evaluation)’ Electrical Engineering in Japan
133(2), pp. 81-87.

Sandnes B. and Rekstad J. (2002) ‘A

photovoltaic/thermal (PV/T) collector with a
polymer absorber plate. Experimental study