PII:
Communicated by HANSJORG
GABLER
AbstractWe present test results on hybrid solar systems, consisting of photovoltaic modules and thermal
collectors (hybrid PV/ T systems). The solar radiation increases the temperature of PV modules, resulting in a
drop of their electrical efficiency. By proper circulation of a fluid with low inlet temperature, heat is extracted
from the PV modules keeping the electrical efficiency at satisfactory values. The extracted thermal energy can
be used in several ways, increasing the total energy output of the system. Hybrid PV/ T systems can be applied
mainly in buildings for the production of electricity and heat and are suitable for PV applications under high
values of solar radiation and ambient temperature. Hybrid PV/ T experimental models based on commercial PV
modules of typical size are described and outdoor test results of the systems are presented and discussed. The
results showed that PV cooling can increase the electrical efficiency of PV modules, increasing the total
efficiency of the systems. Improvement of the system performance can be achieved by the use of an additional
glazing to increase thermal output, a booster diffuse reflector to increase electrical and thermal output, or both,
giving flexibility in system design. 2002 Elsevier Science Ltd. All rights reserved.
1. INTRODUCTION
218
Y. Tripanagnostopoulos et al.
because the variation of the reflected solar radiation results to a non-uniform density of illumination on the PV surface, reducing the electrical
efficiency of the system.
In our laboratory we study hybrid systems with
commercial silicon PV modules of typical size
and thermal units using water or air as working
fluid. A study on the application of solar energy
systems in hotels of Cycladic islands in Greece
was our first step on hybrid PV/ T systems. We
used pc-Si PV modules of dark blue colour, for
aesthetic compatibility to the Cycladic architecture (white buildings of low height with blue
doors and windows) and water for the heat
extraction. The devices were similar to flat plate
thermal collectors (Tripanagnostopoulos et al.,
1996a,b) and to ICS solar water heating systems
(Tripanagnostopoulos et al., 1998). In addition,
we studied air type PV/ T systems, testing several
modes of PV cooling and heat extraction improvement (Tripanagnostopoulos et al., 2000). In the
present paper we include design considerations
and experimental results from constructed and
tested outdoors hybrid PV/ T systems. We determine several aspects for the efficient operation
of the hybrid PV/ T systems and we present
alternative designs with the use of an additional
glazing, a booster diffuse reflector or both, aiming
to the improvement of the total energy output
from the system. The combination of PV/ T
system with booster diffuse reflector is most
important and therefore the main part of test
results is referred to this concept.
2. HYBRID SYSTEM DESIGN
CONSIDERATIONS
219
Fig. 1. PV/ T systems with booster diffuse reflectors: (a) horizontal building roof system installation; (b) PV/ T1REF
experimental system with indication of diffuse reflected solar rays.
220
Y. Tripanagnostopoulos et al.
Heat extraction from PV modules by air circulation is used in hybrid PV/ T systems and several
projects aim to cost effective devices with increased total energy output. These improvements
together with projected lower costs of the PV
component of hybrids will aid the market penetration of these systems. Hybrid systems using liquid
as heat extraction fluid are less developed and
more steps are necessary for their improvement.
Our research on hybrid PV/ T systems is mainly
based on using commercial pc-Si PV modules and
the working fluid for heat extraction is the water
(model pc-PV/ WATER). Water is more suitable
for the weather conditions and the building needs
in lower latitude countries. We have also used
hybrid systems of a-Si PV module type (model
a-PV/ WATER), to compare the performance of
systems of lower electrical efficiency. In addition,
hybrid systems using air to remove heat from PV
modules of pc-Si and a-Si type (models pc-PV/
AIR and a-PV/AIR, respectively) were constructed and tested. With the aim to increase the
total energy output, we studied the basic hybrid
systems of pc-Si module (PV/ WATER and PV/
AIR) with an additional glazing (PV/ WATER 1
GL, PV/AIR 1 GL), a booster diffuse reflector
(PV/ WATER 1 REF, PV/AIR 1 REF) and both
(PV/ WATER 1 GL 1 REF,
PV/AIR 1 GL 1
REF). A plain pc-Si PV module with both its
Fig. 2. Cross section of the PV/ T experimental models: (a) PV/ WATER type; (b) PV/ WATER1GL type; (c) PV/AIR type; (d)
PV/AIR1GL type.
contact. This type of heat exchanger was handmade in our laboratory and as a result the thermal
contact was not very satisfactory. We expect that a
much better thermal contact can be achieved in
hybrid PV/ T systems that will be manufactured
industrially by a proper procedure.
We used thin copper sheet of thickness 5 3
10 24 m and copper pipes of 0.01 m inside and
0.012 m outside diameter. Aiming at sufficient
heat exchanger efficiency with acceptable cost,
the distance between the pipes was 0.08 m at their
arrangement on the copper sheet surface. The
input and output of the heat exchanger were
connected to the forced water circulation system
and thermocouples (TCs) of CuCuNi type were
installed for the measurement of the water input
and output temperature. We have also placed TCs
on the copper sheet surface, on the PV rear
surface and on the protective glazing of PV
module to record the temperature distribution on
several parts of the hybrid system and to calculate
the mean temperature of each part.
In hybrid systems with circulating air the
thermal unit is simpler, as we formed an air duct
between the PV rear surface and the thermal
insulation. The air circulation channel width was
0.1 m and plastic tubes were installed in opposite
positions at the lower and at the higher part of the
air duct for the entrance and the exit of the
circulating air, respectively. The width of the air
channel affects the heat extraction, with higher
values for smaller width and lower for larger
width. Considering that smaller width increases
pressure drop, it has been suggested (Bhargava et
al., 1991) to use air channel with minimum width
w50.1 m to balance the thermal output of air
heating system with the needed electrical energy
input for the fan. We have put TCs at the same
positions as in the hybrid systems with heat
extraction by water circulation and almost all tests
were performed with forced air circulation.
In all constructed hybrid systems we used
thermal insulation of polyurethane with thermal
conductivity K50.05 W K 21 m 21 and thickness
d50.05 m at the backside of the systems and
d50.02 m at the system edges. This insulation is
estimated to be sufficient considering the low
operating temperatures of the hybrid systems. The
aperture area A a of the pc-Si PV module was
A a 50.40 m 2 and of a-Si PV A a 50.27 m 2 . In the
hybrid systems with the additional glazing we
used a glass sheet of thickness 2310 23 m to
avoid as much as possible the increase of optical
losses by the additional absorption of solar radiation. This glazing simulates in transmittance in
221
222
Y. Tripanagnostopoulos et al.
~ p (T o 2 T i ) /A a G
hth 5 mc
The variation of thermal efficiency hth relative to
the fluid input temperature T i , the ambient temperature Ta and the incoming solar radiation G is
determined experimentally as a function of the
ratio DT /G with DT 5 T i 2 Ta . The function
hth 5 f(DT /G) is used for the performance determination of thermal collectors. It can be also
used for the hybrid PV/ T systems, as the thermal
part of them corresponds approximately to a
thermal collector. In stagnation operation (hth 5
0) of the PV/ T systems we used the water heat
exchanger temperature (T whe ) or air channel temperature (T air ), as the systems were operating
without fluid flow. These temperatures were used
to determine the corresponding ratio DT /G, with
DT 5 T whe 2 Ta or DT 5 T air 2 Ta .
During the tests the PV electrical output was
connected to a load, simulating real system operation. With the tested PV/ T systems at thermal
equilibrium under ambient conditions, we determine the values of current Im (in A) and voltage
hel 5 ImVm /A a G
In hybrid PV/ T systems we can consider the total
efficiency htot , which corresponds to the sum of
the electrical efficiency hel and the thermal efficiency hth of the system, for certain operating
conditions.
Studying the effect of booster diffuse reflectors
on the hybrid experimental models we calculated
thermal and electrical efficiency from the net
incoming solar radiation G on the PV module
surface (without the additional solar input from
the diffuse reflector). In concentrating solar devices the concentration ratio C is determined by
the ratio of the system aperture area to the
absorber surface area. In our work the calculation
of thermal (hth ) and electrical (hel ) efficiencies is
based on G and not on the amount C ? G, in order
to get the effective values of the corresponding
efficiencies, considering that the additional solar
input from the reflector affects hth and hel rising
their values by the increase of fluid output temperature T o (or T whe , T air ) and Im , Vm , respectively. The increase of system electrical and thermal
output (included in the effective values of hth
and hel ), provides a clearer comparison of the
results from the tested systems with and without
diffuse reflectors.
The study of the experimental PV/ T model
(PV/ WATER) with diffuse reflector includes tests
with variable percentage of the additional solar
radiation from the diffuse reflector, with respect to
its electrical efficiency as a function of the pc-Si
PV module temperature. We consider a concentration factor C that corresponds to a homogeneous increase of the incoming solar radiation on
PV module surface, with value C51 for net solar
input (without additional solar radiation from the
diffuse reflector) and C51.1 for an effective 10%
additional solar input, etc., up to C51.5 for 50%
additional solar input on PV module. Higher
values of factor C are not usually achieved in
practice, because of the use of diffuse reflector in
the collectorreflector geometry of Fig. 1.
Thermal performance of hybrid PV/ T systems
with booster diffuse reflector depends on the fluid
input temperature, ambient temperature, incident
net solar radiation and the concentration factor C.
223
Fig. 3. Measurement of Ct by using photosensors PS1 (Itot ) and PS2 (Inet ), for variable angle q of the insident solar radiation on
the diffuse reflector, which forms angle w with PV/ T system surface.
224
Y. Tripanagnostopoulos et al.
Fig. 4. Thermal efficiency hth results of systems pc-PV/ WATER, a-PV/ WATER, pc-PV/AIR and a-PV/AIR, as function of DT /G
operating values.
Thermal efficiency
pc-PV/ WATER
a-PV/ WATER
pc-PV/AIR
a-PV/AIR
PV/ WATER
PV/ WATER1GL
PV/ WATER1REF
PV/ WATER1GL 1REF
PV/AIR
PV/AIR 1GL
PV/AIR1REF
PV/AIR 1GL 1REF
Fig. 5. Electrical efficiency hel results of systems pc-PV/ WATER, a-PV/ WATER, pc-PV/AIR and a-PV/AIR, for the
corresponding DT /G values of Fig. 4 experiments.
225
and we consider that Ct 5 1. In practical applications the effective angle w is w 5908 (or slightly
greater) with w 51208 possible upper limit and the
results of Ct for the ratio L /R correspond to
variable values of PV module width (L) relative to
reflector width (R), considering L /R51 as maximum.
From Fig. 6 (angle w 5908) we observe that the
distribution of the total solar radiation on PV
plane (simulated by the use of the [OL]) is
uniform for angles 158 # q # 458 with a mean
concentration ratio Ct 1.4. For q , 158 and
ratio L /R $ 0.20 we obtain a mean value Ct 1.2
and for q $ 608 there is a less uniform distribution of solar radiation mainly in the range of
L /R # 0.40 with a mean value Ct 1.30. These
results show that the uniform increase of solar
radiation on PV module plane is more important
during Summer (higher sun altitude), when the
values of angle q are smaller (158 # q # 608)
than during Winter (lower sun altitude), when
there are larger angles q (q .608). Mean values
of Ct 1.35 and Ct 1.15, respectively can be
considered for these two operation periods of the
combined system.
The results of Fig. 7 are less effective regarding
the increase of the solar radiation on the PV plane
by the used diffuse reflector for angle w 51208
between them. For angles q # 308 there is a very
small effect in concentration factor Ct (Ct # 1.05
for all values of the ratio L /R). For angles 308 ,
q # 608 the obtained result is a smooth variation
of concentration factor Ct around the value of
Ct 1.2 and for angles q . 608 a higher value of
Fig. 6. Variation of the concentration factor Ct for various L /R ratio values, regarding the angles q 508 to q 5758 of the incident
solar radiation on the diffuse reflector, which forms an angle w 5908 with PV/ T system surface.
226
Y. Tripanagnostopoulos et al.
Fig. 7. Variation of the concentration factor Ct for various L /R ratio values, regarding the angles q 508 to q 5758 of the incident
solar radiation on the diffuse reflector, which forms an angle w 51208 with PV/ T system surface.
227
Fig. 8. Results of PV/ T system electrical efficiency hel for the diffuse reflector concentration factors C 5 1 to C 5 1.5 (step 0.1)
and for variable pc-PV module operating temperature.
Fig. 10 for electrical efficiency hel for the corresponding values of the ratio DT /G of Fig. 9. The
results of the system PV/ WATER are the same as
the results of Figs. 4 and 5 (of pc-PV/ WATER
system). They are included in Figs. 9 and 10 for
comparison with the results of the other PV/ T
systems. The experimental points of these PV/ T
systems for thermal efficiency are fitted linearly
and are included in Table 1, but the results of
electrical efficiency are presented only by points
in Fig. 10, as hel is function only of T pv . In the
diagrams of Figs. 9 and 10 experimental points in
the negative range of DT /G axis are included,
Fig. 9. Thermal efficiency hth results of systems PV/ WATER, PV/ WATER1GL, PV/ WATER1REF and PV/ WATER1GL1
REF, as function of DT /G operating values.
228
Y. Tripanagnostopoulos et al.
Fig. 10. Electrical efficiency hel results of systems PV/ WATER, PV/ WATER1GL, PV/ WATER1REF and PV/ WATER1GL1
REF, for the corresponding DT /G values of Fig. 9 experiments.
Fig. 11. Thermal efficiency hth results of systems PV/AIR, PV/AIR1GL, PV/AIR1REF and PV/AIR1GL1REF, as function
of DT /G operating values.
229
Fig. 12. Electrical efficiency hel results of systems PV/AIR, PV/AIR1GL, PV/AIR1REF and PV/AIR1GL1REF, for the
corresponding DT /G values of Fig. 11 experiments.
230
Y. Tripanagnostopoulos et al.
Fig. 13. Test results of the compared systems PV/ WATER, PV/AIR, PV/ FREE and PV/ INSUL from 9:30 h to 15:30 h,
regarding their PV operating temperature for the corresponding incident solar radiation G, ambient temperature Ta and wind
speed Vw .
air heat extraction does not reduces PV temperature significantly regarding the PV module with
both surfaces free to ambient. In addition, model
PV/ FREE is of lower temperature compared to
model PV/ INSUL, due to the radiation and
convection thermal losses from its two surfaces.
Therefore, the thermally insulated PV module
(PV/ INSUL) operates always at higher temperature than that of the other systems, because of the
lower thermal losses from its rear surface.
In Fig. 14 we can see the variation of thermal
efficiency hth of both PV/ WATER and PV/AIR
Fig. 14. Test results of the compared systems PV/ WATER, PV/AIR, PV/ FREE and PV/ INSUL regarding their electrical
efficiency hel , as well as thermal efficiency hth of systems PV/ WATER and PV/AIR, for the corresponding operating conditions
of Fig. 13.
231
Table 2. Test results of the compared systems PV/ WATER, PV/AIR, PV/ FREE and PV/ INSUL, regarding PV module
temperature T pv and electrical efficiency hel under different weather conditions
Ta
(8C)
G
(W m 22 )
PV/ WATER
T pv (8C)
hel
PV/AIR
T pv (8C)
hel
PV/ FREE
T pv (8C)
hel
PV/ INSUL
T pv (8C)
hel
15.0
29.0
35.0
850
910
898
32.0
38.1
47.2
31.9
41.2
48.4
0.134
0.126
0.121
33.9
42.7
54.1
0.132
0.124
0.114
50.2
55.2
68.7
0.118
0.113
0.098
0.134
0.128
0.123
232
Y. Tripanagnostopoulos et al.
Iref
Itot
K
L
~
m
R
PS
PV
T
Ta
T air
Ti
To
T pv
T whe
Vm
Vw
w
DT
b
q
hel
hth
htot
w
233
REFERENCES
Al-Baali A. A. (1986) Improving the power of a solar panel by
cooling and light concentrating. Solar and Wind Technology
3, 241245.
Bergene T. and Lovvik O. M. (1995) Model calculations on a
flat-plate solar heat collector with integrated solar cells.
Solar Energy 55, 453462.
Bhargava A. K., Garg H. P. and Agarwal R. K. (1991) Study
of a hybrid solar system solar air heater combined with
solar cells. Energy Convers. Mgmt. 31, 471479.
Brinkworth B. J., Cross B. M., Marshall R. H. and Yang H.
(1997) Thermal regulation of photovoltaic cladding. Solar
Energy 61, 169178.
Brogren M., Ronnelid M. and Karlson B. (2000) PVThermal
hybrid low concentrating CPC module. In Proc. 16 th
European PV Solar Energy Conf., Glasgow, UK. Vol. III,
21212124.
Cox C. H. and Raghuraman P. (1985) Design considerations
for flat-plate-photovoltaic / thermal collectors. Solar Energy
35, 227245.
Elazari A. (1998) Multi solar system. Solar multimodule for
electrical and hot water supply for residentially building. In
Proc. of 2 nd World Conference and Exhibition on Photovoltaic Solar Energy Conversion, Vienna, Austria, 2430
2433.
Florschuetz L. W. (1979) Extension of the HottelWhillier
model to the analysis of combined photovoltaic / thermal flat
plate collectors. Solar Energy 22, 361366.
Garg H. P., Agarwal R. K. and Bhargava A. K. (1991) The
effect of plane booster reflectors on the performance of a
solar air heater with solar cells suitable for a solar dryer.
Energy Convers. Mgmt. 32, 543554.
Garg H. P. and Adhikari R. S. (1997) Conventional hybrid
photovoltaic / thermal (PV/ T) air heating collectors: steadystate simulation. Renewable Energy 11, 363385.
Garg H. P. and Adhikari R. S. (1999) Performance analysis of
a hybrid photovoltaic / thermal (PV/ T) collector with integrated CPC troughs. Int. J. Energy Res. 23, 12951304.
234
Y. Tripanagnostopoulos et al.