Energy Convers. Mgmt Vol. 35, No. 7, pp.

621-633, 1994
~ Pergamon 0196-8904(93)E0024-F
Copyright © 1994 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
0196-8904/94 $7.00+ 0.00

EXPERIMENTAL STUDY ON A HYBRID PHOTOVOLTAIC-

T H E R M A L S O L A R W A T E R H E A T E R A N D ITS
PERFORMANCE PREDICTIONS

H. P. GARG, R. K. AGARWAL and J. C. JOSHI

Centre of Energy Studies, Indian Institute of Technology, Hauz Khas, New Delhi-110 016, India

(Received 4 Februao' 1993; received for publication 20 December 1993)

AbstractThe combined photovoltaic thermal system is a new kind of solar system which, so far, has
not seen intense research and developmental activities as other systems have seen. The most attractive
feature of such a system is that it could produce both electrical and thermal energies from the same panel.
These systems could play an important role in popularizing solar energy. In order to understand and
evaluate solar hybrid systems, an experimental study was undertaken. The system chosen is a ther-
mosyphon water heater with solar cells. The cells are pasted directly over the absorber plate. The solar
insolation falling on the aperture area of the collector is concentrated on the solar cells using a simple
parabolic reflector. The temperatures at several points in the system (or the absorber at several points,
of the solar cells, of the water in the tank at several positions, of the inlet and outlet water from the
collector) are measured by using thermocouples. The solar cells are arranged in three groups. A m o n g a
group, the solar cells are connected in series. The three groups can be connected either in parallel or series.
The electrical characteristics of the cells, i.e., the I V characteristics of various combinations with several
different load resistances are measured. The experiment was performed for several days with ther-
m o s y p h o n flow and without it. The heat transfer coefficients, heat loss coefficient of the tank and overall
heat loss coefficient of the collector are determined from experimental data. An attempt is made to fit the
various experimental observations, like water temperature in the tank, inlet and outlet water temperatures
of the collector, with the well known model of thermosyphon solar water heaters. The model equations
are modified to include the presence of the solar cells and the reflectors. The agreement between the
experimental and theoretical predictions, within the experimental errors, is excellent.

PV/T system Solar water heater Hybrid solar system

NOMENCLATURE

A C = Aperture area of collector (m 2)

A s = Surface area of storage tank (m 2)
Acell = Solar cells area (m 2)
Acc = Absorber plate area (m 2 )
C w = Specific heat of water (J/kg ~C)
M w = Mass of water in system (kg)
m= Water mass flow rate (kg/s)
S= Solar insolation (W/m 2)
S2 = Solar radiation absorbed by absorber plate (W/m 2)
T,n= Collector inlet water temperature CC)
Toot = Collector outlet water temperature ('C)
L= Ambient temperature CC)
rp= Temperature of absorber plate (°C)
TR= Temperature of metallic reflector ( C )
Mean water tank temperature (°C)
I= Time interval (s)
H L = Heat loss coefficient for collector (W/m 2 ~C)
Hs= Heat loss coefficient for storage tank (W/m 2 C )
Ur = Total heat loss coefficient of system (W/m 2 'C)
CLp = Plate heat loss coefficient for absorber (W/m 2 ~C)
ULR = Heat loss coefficient for metallic reflector (W/m 2 °C)
a= Absorptance of absorber plate
col = Collector efficiency
C ~ Efficiency of solar cell

INTRODUCTION

During the last 3 years, a considerable amount of work has been done on utilization of solar energy,
and several systems have been developed. Present solar systems can be divided broadly into two

621
622 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER

categories--thermal and photovoltaic. Both kinds of systems are widely used to harness solar energy
for mankind. The efficiency of thermal systems is high, as much as 50% or above, but the energy
generated is of low quality and has limited applications. Flat plate thermal systems are easy to
construct. On the other hand, in photovoltaic systems, solar energy is converted directly into a high
quality energy, namely electrical energy. The systems are more sophisticated than thermal systems.
A good infrastructure is needed for manufacturing such systems. The efficiency of photovoltaic
systems is rather low, at the most a few percent (bare cell efficiency is around 10%, while that of
panels is even less). A third kind of system which combines both photovoltaic and thermal systems,
known as photovoltaic-thermal (PV/T) or hybrid systems, has also been proposed. It combines both
systems into one unit. The output from the system is in the form of thermal and electrical energies.
Such systems should offer several advantages over separate thermal and photovoltaic systems. Hybrid
systems should be easy to construct as no new technology is involved.
There would be many design approaches to build a hybrid collector. The simplest one is to
construct a flat plate hybrid collector. The system has a black absorber plate over which silicon
solar cells are placed. The solar cells can either be directly pasted over the absorber plate or first
placed on a black thin insulating sheet which is then pasted over the metallic absorber. The second
alternative should be better because it is easier to remove the cells from a thin sheet rather than
from the metal plate. Since a large part of the absorber is covered by solar cells, a selective paint
could be used to make up the loss of absorptivity of the solar cell. Although no detailed study is
available on the absorptivity of the solar cell surface, it has been reported that the absorptivity of
the solar cell is slightly less than that of the black absorber surface.
The literature survey indicates that hybrid systems have not received much attention. Only a few
papers have appeared on this topic [5-12]. Hybrid systems, based on the mode of heat removal,
can be classified mainly into two categories--water systems and air systems. Two kinds of solar
water heater systems are normally used, forced flow and thermosyphon type. The latter are more
popular in the domestic sector of developing countries. In view of the widespread use of
thermosyphon systems, an experimental study of hybrid thermosyphon water heater is undertaken.
The forced flow solar water heater has been discussed, both theoretically and experimentally. The
situation for air systems seems to be worse. Recently, authors have discussed some aspects of air
systems theoretically, without concentrators and with plane mirror boosters. It has been shown that
these generate sufficient electrical energy to run the p u m p which circulates the air through the
system. Since a hybrid collector is self sufficient in electrical energy, these should be particularly
useful in rural areas of developing countries. Solar dryers are most needed in places where little
or no electrical power is available.
Thermosyphon systems do not need any electrical energy for operation. All of the electrical
energy generated by the solar cells would be available for use. It could be stored and later utilized
during the night-time for running a fan or television or some other utility. A hybrid solar water
heater of aperture area 2.97 m 2 was installed. The solar cells used are normal single crystal silicon
cells. The solar radiation is concentrated over the cells by using a simple aluminium foil parabolic
reflector. The electrical and thermal characteristics are measured. The experiment was performed
over a period of several days with thermosyphon flow and without it. The experimental data is
fitted with the known models of thermosyphon flows [1-4]. The equations of the model are modified
to include the presence of the solar cells. The two heat loss coefficients of the model, that of the
tank and the collector, are determined from experimental data.

EXPERIMENTAL SET-UP
The experimental set-up which has been studied is shown schematically in Fig. 2. It has two units,
a flat plate collector and a storage tank unit. The two units are connected by pipes. Glass wool
is used to provide insulation around the pipes. The thickness of the insulation is sufficient to assume
that no heat losses occur through the pipes. The tank unit has two concentric cylindrical tanks made
from galvanized iron sheets. The inner one contains the water, while the space between the inner
and the outer one is filled with glass wool. The water level in the inner tank is maintained constant
by a float valve. When filled to the maximum level, the tank contain 150 1. of water. The photograph
of the hybrid PV/T solar water heater is shown in Fig. 1.
 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER 623

Fig. 1. A hybrid photovoltaic thermal solar water heater installed at I.I.T., Delhi.

The collector unit has an aperture area 2.97 m 2 (2.48 m , i.20m). The absorber plate is
made of ~luminium. It is divided into four channels, 2.48m long and 0.14m wide. The solar
insolation is concentrated on the absorber area by means of parabolic reflectors placed at the
side of each channel. The reflectors are made from thin aluminium foils. The concentration
ratio is 2.1. The solar cells are not pasted directly over the aluminium absorber. They are first
fixed over a cloth-like material. A silicon resin is used to obtain a good contact with the cloth.
The cloth prevents short circuiting of adjacent cells. The silicon resin has high thermal conductivity
and high electrical resistance. The collector consists of 102 single crystal silicon cells, 26 each
in the two outer channels and 25 each in the two middle channels. The total surface area of the
cells is 0.6 m 2. The cells are grouped in three strings, each string having 34 cells. In a string,
the cells are connected in series. The three strings, from outside, can be connected either in
parallel or in series. The electrical characteristics can, thus, be studied under m a n y different load
conditions.
About 43% of the absorber area is covered by solar cells. The absorptivity of the solar cell
surface is not equal to that of the black paint. If a normal black paint is used, the
thermal performance of the system would not be high. In order to improve the thermal
performance of the system, a selective paint is used on the area not covered by the solar cells.
The cloth with the solar cells is then pasted over the aluminium absorber. The water carrying
copper tubes of 1.9 cm outer diameter are brazed behind the aluminium sheets. The system has
a single glazing of tempered glass of thickness 4 mm. The collector is placed at an angle of 45 ° facing
south.
624 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER

Ho--I.TIrIM //'If0 c m / / / A
50 cm T
Cold water
- - 7 0 cm ~!
~ermometer

T2,T3,T4,T 5 Plate temperature

Solar cell temperature

T8,T 9 Storage water temperature

= TI0 Collector inlet water temperature
~t= TI l Collector outlet water temperature

Ther

12lm --

Fig. 2. Schematic layout of the hybrid solar water heater.

INSTRUMENTATION
Solar radiation data are measured by an Eppley pyranometer placed in the plane of the collector.
The data are recorded automatically by a Kipp and Zonen recorder. The ambient temperature
was recorded by an automatic potentiometric recorder along with a copper-constantan
thermocouple wire. Both of the measuring instruments were calibrated before the start of the
experiment.
In order to evaluate system performance, two kinds of quantities are to be measured--thermal
and electrical. The thermal quantities are measured by two kinds of thermo-sensors, thermometers
and thermocouples. The locations of the various sensors are shown in Fig. 2. The collector inlet
and outlet water temperatures are measured by mercury thermometers placed in the thermowell.
The absorber, cell and tank water temperatures are measured by precalibrated copper-constantan
thermocouples. The thermocouples /'1-/'5 are placed below the cloth and in the contact with the
metal absorber. T6, which measures the cell temperature, is pasted over the cell. Care is taken so
that radiation does not increase its temperature. From the electric point of view, the system has
been divided into three strings, i.e. A, B, C. Each string has 34 silicon solar cells connected in series.
1-V curves of the solar cells were plotted by the 2000 Omnigraphic recorder (x-y plotter). The short
circuit current (I~:), open circuit voltage (Voc), maximum power voltage (Vmp), maximum power
current (Imp) and maximum power (PM) have been measured under clear sky condition by using
the M P P T (maximum power point tracker) at intervals of 1 h.
 G A R G et al.: H Y B R I D P H O T O V O L T A I C - T H E R M A L S O L A R W A T E R H E A T E R 625

2 2 2 2 2 2

4 4 4 4 4 4 4 ~3
13 ~3 13

Reflectors

6 5 7

Metallic plate

Solar cell

I ! I
Fig. 3. Successive stages o f solar radiation absorbed by absorber plate.

MATHEMATICAL MODEL

Several authors have earlier discussed thermal modelling of thermosyphon water heaters [1-4].
We have studied a thermosyphonic solar water heater where solar cells are pasted on fiat plate
collectors[13]. Here, we are not going to repeat all the energy balance equations of the
thermosyphonic solar water heater. In the present experimental set-up under study, there are two
new components, solar cells and reflectors. The modified energy balance equation of the system
is

FRS2 = UT(Tm - T a ) + (M,,,Cw + M m f p m ) A T m

At ' (1)
where
UT = UL Ac + UsA~. (2)

.... Solar radiation (exp.)

Ambient temperature (exp.)
1000 -
{\
r\ / \ I \
I
~: 800 ~ ~1 / ~ I ~ I I I
I ~ I I I i ~ 40
i i
i i
•~ 600 t i ~ I ~ J L %
I ~ -- 30

t ~ 20
m 200 p ~- J'
t Feb. 27, 1989 Feb. 28, 1989 March 1, 1989 ~" , March 3, 1989 -- 10 E-
ol I [ Iiii I I Iiii l I Iii1 I I Ii/I I I
8 16 24 8 8 16 24 8 8 16 24 8 8 16 24 8 8 16 24
Time (h)
Fig. 4. An hourly variation of solar intensity and ambient temperature.

ECM 35,7 F
626 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER

The Mm Cpm term should be written MiCpiwhere M~ is the mass and Cp~ the specific heat of the
material, i.e. the total heat capacity of the system is defined as

C = MwC. + ~ Micpi. (3)

t'

The mechanism of absorption of the direct and reflected parts of solar radiation by the solar
cells and absorber plate is shown in Fig. 3.
The efficiency of the collector is given by (Ac = Acol)

mCw(To~t- T~n) (4)

q = SAgo1
The heat balance equation for the absorber plate is given by
$2 - m C w ( r o m - T i . ) = UL,,A=(Tp- Ta) + ULR(A¢ -- A c . ) ( T R - Ta). (5)
The reflector is made of metallic material. It is in contact with the metallic absorber plate. The
temperature of the reflector, TR, would be very close to the plate temperature Tp. No error will
creep into the analysis if we assume ULR = Uep = Ur~ and TR = Tp. Using these and putting
equation (4) in equation (5), the heat balance equation for the absorber plate becomes
$2 - nSAco, = U ~ ( T p - Ta)A~o,. (6)
Equation (6) is used to calculate the plate temperature which is given by
$2 - - nSAcol
Zp = Up aAcol ~ Ta• (7)

The successive stages of passage of solar radiation through the collector, illustrated in Fig. 3,
are
(1) solar radiation falling on the collector aperture area;
(2) solar radiation transmitted through the glazing;
(3) solar radiation reaching the absorber plate directly;
(4) solar radiation reaching the absorber plate after reflection from the reflectors;
(5) total radiation on the absorber plate [3, 4];
(6) solar radiation falling on the solar cell;
(7) solar radiation falling on the black absorber area;

160 --
Feb. 27, 1989

140 -- ,~'~\T2
II \\
120
//W,', .... Experimental
-- II XX

o II / \\%~
~ 100 -- III ~X
Ut ~X Feb. 28. 1989 March 1, 1989 M a r c h 2. 1989 March 3, 1989
T~ T~ ~..7_T ~ T,
& 80 - JH ~
lU W i t h o u t %X T~ ,I'T< F[_ / ~ - . NNX
Ii~ water ~
[-
JI flow ~.~
l/
(jl T 4 \ "\ : x'{X
60

40-- -,i III

l//
II
II
111 T4 \'~
\
"~ ibm/--
i
.',
X
i;ll`
l,
'~,
"
xX
With water f l o w ( t h e r m o s y p h o n )

20 I I 118/II I I 11811 I I l ilitA I I 18//~ I I

8 12 noon 16 8 12 noon 16 8 12 noon 16 8 12 noon 16 I 12 noon 16 18
Time (h)
Fig. 5. The variation of absorber plate temperature with time, with and without water flow.
 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER 627

(8) converted into electrical energy by the solar cells;

(9) transmitted by the solar cells to the metallic plate below [6-8];
($2) solar radiation absorbed by the absorber plate [a (7, 9)].
In the equations, there appears the solar cell efficiency r/c. The efficiency of a single solar cell is
a function of its temperature. The following linear relationship is used to evaluate qc at a given
mean absorber temperature Tp,

r/c = r/r[l - (Tp - Tr)], (8)

where r/r ( = 10%) is the cell efficiency at a reference temperature Tr (=25°C). An expression for
the temperature coefficient fl is obtained on the assumption that at some temperature T, the cell
efficiency must drop to zero. fl is then given by I / ( T - Tr). If T is taken to be 275°C, then fl turns
out to be 0.0041, a value in close agreement with the one taken by several authors [5-12]. In this
work, we have taken fl = 0.0041.
In solar hybrid systems, there are two kinds of efficiencies--thermal and electrical. The thermal
efficiency, as usual, is defined by

MwCw(Tmf,,, -- Tmi) (9)

Y/thermal
Ac I S dt
J
The electrical efficiency is defined as

~/ele :
f
'
(Maximum power)systemdt
(lO)
Ace u f s dt
The total system efficiency is

Yltot = tlthermal + Ylele" (11)

R E S U L T S AND D I S C U S S I O N S
In this paper, the authors have experimentally studied a hybrid photovoltaic-thermal solar water
heater provided with reflectors, the design of which has already been explained. In Ref. [13], the
theoretical model of an almost similar system, but without reflectors, has been theoretically studied

70

Feb 28, 1989

60

S
o
~ 50 NX

\
Cu
J
E 40 I
P
[- /
/
.... Experimental

3O

2O I I I I I I I I
9 10 11 12 noon 13 14 15 16 17
T i m e (h)

Fig. 6. The v a r i a t i o n o f a b s o r b e r plate t e m p e r a t u r e with time, w i t h w a t e r flow.

628 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER

in depth. Thus, for theoretically calculating the various parameters, such as inlet and outlet water
temperatures and the storage water temperature, the energy balance equations given in Ref. [13]
are used here also by incorporating the required modifications in the values of solar radiation
absorbed by the system. This difference is obviously due to the reflectors incorporated with the
system.

EXPERIMENTAL OBSERVATIONS
The experimental observations of the system were carried out from mid-January 1986 to
mid-March 1989 in order to study the thermal and electrical behavior of the system. Of these
observations, the data collected on February 27, 28, March 1, 2 and 3 are presented in this paper.
Meteorological data of solar radiation and ambient temperature in Delhi for the above mentioned
dates (February 27-March 3, 1989) are given in Fig. 4.
Observation of the thermal behaviour of the system was carried out on February 27, 1989
without water flow. This is achieved by keeping the water level in the tank below the return water
level. This stops the water flow. However, since the plate temperature is above the boiling point
of water (between 10 a.m. and 2 p.m.), the temperature of the water in the channels increased
considerably. This leads to temperature stratification of the water. The difference between the water
temperature at the top and bottom of the channel would be quite large. Since the water temperature
at the bottom of the channel is very low as compared to that at the top, heat transfer from the
plate to the water at the bottom of the collector would be large. Thus, the plate temperature at
the bottom /'4 is lower than that at the top T:, where the heat transfer is comparatively less due
to the higher temperature of the water at that level. As is clear from Fig. 5, which depicts the
time-dependent variation of T2, 7"3 and T4, which are the plate temperatures measured at three
points along the length of the same channel on successive days February 28-March 3, 1989, the
experiment was continued with flow of water due to the thermosyphonic effect by keeping the water
level in the tank above the return water pipe level. During all these days, the water was not
withdrawn from the system at all.
The flow of water in the channels leads to higher heat transfer between the plate and the water
and, thus, the plate temperature at all three points (T2, T3 and T4) reduces considerably at any
instant of time, as compared to that without water flow in the system.

160 - -

Feb 2 7 , 1 9 8 9

140 -- 90-
// \ \
// \ X
II \\ \ March 1, 1989
Ii T 6 \ \\

120
iiii / \\ \\
80-- ~__~ T6
ii \\\ // ~-\
r,.) ii \\
/ \
iI \\
ii \\
.o iI \\ / // T 3 xxXx
100 ,/ 70-
//
II
/
/ \ \
\ \
\\
3 \\
I /

,ii;; \\,\\\
k \
II \\
E II \\
[-o II \\
II \\
80 /t g 60- II \\
l/ \X
II k\
!~ g II \\
II . . . . . Experimental ~x II .... Experimental \\
ii \\
\\ I/ X\
60 XX
\X
\%
Xx
\%

40 I t I I I 4o I I I
10 12 n o o n 14 16 17 10 12 n o o n 14 16 17
Time (h) Time (h)

Fig. 7. The variation of solar cell temperature and absorber plate temperature with time.
 G A n G et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER 629

80--
o T7 T7

7(I T7 \ ~ \_

-- T7 / // /\r - ~ . ~ . ~ ~"-~.~ // /k..T9 "~"~-.~..~.//f////\T9

K 60
E
0~ \I~ ~ - . - ~ I/ I! T9 ~"~-/ // M a r c h 2, 1989 M a r c h 3. 1989

5(I ~ / / M a r c h 1, 1989
--- 1/111\ T 9
..~ 40 - 7 / / Feb 28. 1989 T 7 - T o p water temperature
e.4 1/ --- Experimental
/// T 9 - B o t t o m water temperature
3(I -
P.
o
2(1 I I I I I I I I I I I i I I i t I
9 12 18 24 6 12 18 24 6 12 18 24 6 12 18 24 6 12
noon noon noon noon noon

Time (h)
Fig. 8. The variation of storage tank water temperature with time for 4 consecutive days.

The plate temperatures (T~, T3 and Ts) at the same height from the bottom of the collector (which
is kept tilted) is noted continuously throughout the four days of the experiment. T~, 7"3 and T5 are
the temperatures of the plate in the first, second and third channels of the collector, respectively.
The time-dependent variations of these temperatures for 1 day are plotted in Fig. 6. From the
figure, it is seen that T~, T3 and T5 are almost the same at any instant of time. However, a slight
difference is observed during the hours when the thermosyphon mode is working.
In Fig. 7, the variations of the cell temperature T6 on the middle point of the second channel
and the plate temperature T3 at a point very near to this cell are plotted against time for 2 days
of the experimental observations. On the first day (February 27, 1989), the observations were
carried out without a flow in the collector, whereas on March 1, 1989, there was flow in the collector
due to the thermosyphon effect. In both cases, it is seen that, at any instant of the day, the difference
between T3 and T6 is negligible. Thus, the assumption that the average cell temperature can be taken
the same as the average plate temperature is validated.
The storage water temperatures TT, /'8 and T9 at the three different depths of the storage tank
were measured continuously for 4 days. Of these, the temperatures at the top T7 and the bottom
T9 are plotted against time in Fig. 8. From the figure, it is seen that the maximum temperatures
(both at the top and the bottom) on consecutive days keeps on increasing during the first 3 days
of the experiment. On the fourth day, the maximum temperature at the top T7 is almost the same
as that on the third day. As expected, T7 is higher than T9 during the initial 3 days, but on the

80 --

March 2, 1980 M a r c h 3, 1989

70 -- March 1, 1989
.,/~ ",TII ~ \\TII
// ~ ~ I1 / \ / \
60 -- Feb 28, 1989 TI 0
C
o /
/ / / \\
\ /
// / T I 0/ / - \
/
,, \\
\ \
// TI0//~
/
\\
\
"\
\
/ / \ /" \
,/ / x / / \ /
/ / \
/
50 - - / // TIO - - /
/
& /
/
/
/ /
E
40 ~ II T I 0 - Hybrid collector inlet water temperature
/
/
// TI 1 - Hybrid collector outlet water temperaturc

30 f " .... Experimental

2O I I ill I I 11/ I I i [/! I I I J

10 12 n o o n 14 16 10 12 n o o n 14 16 10 12 n o o n 14 16 10 12 noon 14 16

Time (h)

Fig. 9. The variation of collector inlet and outlet water temperature with time.
630 GARG et al.: HYBRID PHOTOVOLTAIC-THERMALSOLAR WATER HEATER

2 0.74 kW/m 2

10.6/2 kW/m2 ~ ~ Parallel combination

10 20 30 40 50
V o l t a g e (V)

Fig. 10. I - V characteristics o f the solar hybrid system.

fourth day, these two temperatures become almost equal. This shows that there is no stratification
of the water temperature in the storage tank on the fourth day. This may be due to the fact that
the storage water temperature at this stage becomes truly periodic in nature.
The collector inlet water temperature T,0 and outlet water temperature Tit are noted for all 4
days. The variations of these temperatures with time are shown in Fig. 9. The difference between
Tl~ and T,0 at any instant of the first day (February 28, 1989) is large. As the days progress, this
difference is observed to be decreasing. This is due to the fact that the temperature stratification
in the storage tank reduces considerably towards the fourth day (March 3, 1989) because no water
is withdrawn from the system during these days.
In Fig. 10, the I - V characteristics of the system are plotted for the series and parallel combination
of the strings A, B, and C for different values of incident solar radiation (0.62 kW/m 2, 0.74 kW/m 2
and 0.92 kW/m2). From the figure, it is seen that the current increases with incident solar radiation
for any value of voltage. The open circuit voltage Vo~ is almost independent of incident solar
radiation. Vo~ for the series combination is larger than that of the parallel combination. For any

18-- With water flow (thermosyphon)

/-xC fN C C
.--.x
/ \ C \\ I
/ \ 11 \\ /
Without / /
14-- I \
water flow I \
/
l \ I
C /
/ 1 A
/ \ I A X
IO-- / \ I /AN x
I /--\ III if/B\\
iI A \ I / -- \
o Ii B \\ /III:IIB "",, ~ iiii!///~\~\%
I/ I X ~ B ~ \ \ \\ I' II1/ \ \ \\ ~
//fll \\ I
__[llt,~--~'\ \\ III \X" 1 \\\ I /ll ! \\\ I Iiii ll[
11 \ x XX III XXA I II \~\II
I II ~\ I
t
Feb 27, 1989 Feb 28, 1989 %~ 1, 1989 March 2. 989 .. March 3, 1989 ~\~.~,--

12 noon
I 1161/ 8I 12 noon
I 116II I
8
I
12 noon 16
I II I
8 12 noon
I
16
I//I
8
I
12 noon
I
16
T i m e (h)

Fig. 11. T h e variation o f p o w e r with time o f the individual string.

 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER 631

With water flow (thermosyphon)

40--
I
Without / -I \ ,...I I
/
water flow / 1I \ / x / II \
/ /. x / 1I \ //1I\\ / /'~\ \
30-- Ii \\ I fx \ 1/f\\
I I \ \ \\ / / \\
1 I i \ \ I i \\\\ / / \\
/--'x II \\ II II / I \\
// \\ jl \ ~ I \\\ I / \ \
II II \ ~ Ii II II \ \
20-- / / ~\
iI / / \ \\ \ \\ II \~ Ii II \~\ i] I \~
O
//I \ \ \l II ~ \~ 0

10--

Feb 27, 1989 \, Feb 28, 1989 ~ March 1, 1989 ~ March 2, 1989 ~ March 3, 1989
x

I Iii I I I III I I II I I 1161/

12 noon 16 8 12 noon 16 8 12 noon 16 8 12 noon 8 12 noon 16
T i m e (h)
Fig. 12. T h e variation o f p o w e r with time o f A, B and C connected in series (I) and parallel
(II) c o m b i n a t i o n s .

given value of voltage and incident solar radiation, the current is larger in the parallel combination
than that in the series combination.
In Fig. 11, the individual variations of the electric power of the strings A, B and C with time
are shown for 5 days. It is seen from the figure that, on the first day (February 27, 1989), when
there is no water flow, the power output is less for all three strings of the cells as compared to the

6°V

N/m2°C)
55
G
o
1.5

2.0
E

50
2.5

o
~q
3.0

3.5
4r

42.5 I I I
16 p.m. 20 24 8a.m.
T i m e (h)

Fig. 13. The t a n k water temperature as a function o f time (from 4 p.m. to 6 a.m.) for several values o f
tank heat loss coefficient, U s .
632 GARG et al.: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER
Experimental mean
60 T7
T8 U s = 2 . 5 W/m 2°C
T9~ / ~

Computed mean
55 II I
/ I - U L = 5 W/m 2 °C
/ I
I I - U L = 8 W/m 2 °C
/ I
/ I
50 / /
o - U L = 12 W/m 2 °C
iI iI

45 I
I
I
I
I

,,//Y
I
I
I Ii
II I f
40 I //////
I
Experimental

/ Theoretical
/
O I
35 /
/ I# I'
I
/
30 ~ /I ii 1
/ t

251 I I I I I
10 12 noon 14 16 18
T i m e (h)

Fig. 14. T h e tank w a t e r t e m p e r a t u r e as function o f day time ( f r o m 7 a.m. to 5 p.m.) for several values
o f c o l l e c t o r heat loss coefficient, U s = 2.5 W / m 2 °C.

output on the following days when there is water flow. The increase in power when there is water
flow is due to the decrease o f the cell temperature and consequent increase in cell efficiency.
In Fig. 12, the variations o f power out with time are presented for the series and parallel
combinations o f the strings A, B and C for 5 days. Here also, it is seen that the power output is
higher when there is water flow in the system.
Using the mathematical model for the thermosyphonic system developed in Ref. [13], with the
modification due to the effect o f reflectors, the variation o f the mean water tank temperature during

60
TII
Ti0

50

C
o

40

[..
30

20 I I I I I I I
10 11 12 noon 13 14 15 16
Time (h)
Fig. 15. The collector inlet and outlet water temperature as a function of time.
 GARG et aL: HYBRID PHOTOVOLTAIC-THERMAL SOLAR WATER HEATER 633

the night h o u r s with time is c a l c u l a t e d for v a r i o u s values o f the overall heat transfer coefficient
o f the tank, Us. These v a r i a t i o n s are p l o t t e d in Fig. 13. T h e e x p e r i m e n t a l l y o b s e r v e d v a r i a t i o n o f
the storage w a t e r t e m p e r a t u r e is also s h o w n in the same figure. It is clear from the figure that the
best-fitting value o f Us is 2.5 W / m 2 °C,
The value o f Us = 2.5 W / m 2 °C has a l r e a d y been o p t i m i z e d by a curve-fitting technique. By
m a k i n g use o f this value o f Us, the s t o r a g e w a t e r t e m p e r a t u r e with t h e r m o s y p h o n i c w a t e r flow is
c a l c u l a t e d for different values o f the overall collector heat transfer coefficient UL. These curves are
p l o t t e d against time in Fig. 14 for F e b r u a r y 28, 1989. The e x p e r i m e n t a l l y observed values o f storage
water t e m p e r a t u r e (7"7, T8 a n d Tg) at different p o i n t s o f the storage t a n k for the same d a y are also
p l o t t e d in the same figure. T h e m e a n o f TT, T8 a n d T9 is calculated, a n d it is c o m p a r e d with
theoretically calculated values o f the m e a n s t o r a g e w a t e r t e m p e r a t u r e . It is f o u n d that there is g o o d
a g r e e m e n t between the e x p e r i m e n t a l a n d theoretical curves when the m e a n t e m p e r a t u r e is
calculated by m a k i n g use o f UL = 5.0 W / m 2 °C. In Fig. 15, the v a r i a t i o n o f e x p e r i m e n t a l l y o b s e r v e d
a n d theoretically c o m p u t e d values o f the inlet a n d outlet water t e m p e r a t u r e s o f the collector T,0
a n d T~j with time are shown. The g o o d a g r e e m e n t between the e x p e r i m e n t a l a n d theoretical curves
o f b o t h Tl0 a n d T,~ proves the validity o f the theoretical model.

CONCLUSIONS

(i) The t h e r m a l a n d electrical efficiencies o f the h y b r i d solar water heater are 33.5 a n d 3.35%,
respectively.
(ii) The a s s u m p t i o n that the average cell t e m p e r a t u r e can be taken the same as the average plate
t e m p e r a t u r e is validated.
(iii) There is no stratification o f the w a t e r t e m p e r a t u r e in the storage t a n k on the fourth day.
(iv) There is m o r e electric p o w e r in the parallel c o m b i n a t i o n t h a n in the series c o m b i n a t i o n .

REFERENCES
1. D. J. Close, The performance of solar water heaters with natural circulation, Sol. Energy 6, 23 (1962).
2. C. L. Gupta and H. P. Garg, System design in solar water heater with natural circulation, Sol. Energy 12, 163 (1968).
3. K. S. Ong, A finite difference method to evaluate the thermal performance of a solar water heater, Sol. Energy 16, 137
(1974).
4. G. L. Morrison and J. E. Braun, System modelling and operation characteristics of thermosyphon solar water heater,
Sol. Energy 34, 389 (1985).
5. L. W. Florschuetz, On heat rejection from terrestrial solar array with sunlight concentration, IEEE Photo~,oltaic
Specialists Conf., p. 318 (1975).
6. Susan D. Hendrie, Evaluation of Combined Photovoltaic/Thermal Collectors. ISES, Atlanta, Ga, U.S.A. (1979).
7. L. W. Florschuetz, Extension of the Hottel-Whillier model to the analysis of combined photovoltaic/thermal fiat plate
collectors, Sol. Energy 22, 361 (1979).
8. Akio Suzuki and Susumu Kitamura, Combined photovoltaic and thermal hybrid collector, J. Appl. Phys. 19, 79 (1980).
9. Martin Wolf, Performance analysis of combined heating and photovoltaic power systems for residences, Energy
Com,ers. 16, 79 (1976).
10. D. I. Evans and I. W. Florschuetz, Terrestrial concentrating photovoltaic power system studies, Sol. Energy 20, 37
(1978).
1I. S. N. Sharma, S. S. Mathur and T. C. Kandpal, Analytical performance evaluation of combined photovoltaic -thermal
concentrator receiver systems with linear absorbers. Energy Convers. Mgmt 27, 361 (1987).
12. S. N. Sharma and T. C. Kandpal, Optimum concentration ratio for a combined photovoltaic thermal concentra-
tor receiver system, Energy Convers. Mgmt 26, 355 (1987).
13. A. K. Bhargava, R. K. Agarwal and H. P. Garg, Study of a photovoltaic thermal system: thermosyphonic solar water
heater combined with solar cells, Sol. Energy. In press (1994).