2011

IEEE First Conference on Clean Energy and Technology CET

Heat Extraction from Gradient Layer using External

Heat Exchangers to Enhance the Overall Efficiency
of Solar Ponds
Yusli Yaakob\ Abhijit Date2, Aliakbar Akbarzadeh3
Energy Conservation and Renewable Energy Group
School of Aerospace, Mechanical and Manufacturing Engineering
RMIT University
P.O Box 71, Bundoora, Victoria, 3083, Australia
lyusli.yaakob@rmit.edu.au
2abhijit.date@rmit.edu.au
3aliakbar.akbarzadeh@rmit.edu.au
Abstract- A salinity gradient solar pond is a combined

generation of electricity specifically at remote area locations.

It holds the promise of providing an alternative to Diesel
generation of electricity specifically at remote area locations.

solar collector and thermal energy storage system, and in

past heat has been successfully extracted from the lower
convective zone (LCZ) in working ponds.
discusses possibility

of

This paper

heat extraction from the

non

Figure 1 and 2 shows a typical experimental solar pond

which consists of three regions (surface zone, gradient and
lower zone). The surface or upper convective zone (UCZ) is a
homogeneous layer of low salinity brine or fresh water. The
gradient layer or non-convective zone (NCZ) constitutes a
thermally insulating layer that contains a salinity gradient to
suppress convection currents. The lower zone (LCZ) is a
thermally convective zone with a homogeneous salt
concentration and the brine in this region is much hotter than
that in the surface zone.

convective zone (NCZ) using an external heat exchanger.

Here, two methods of heat extraction from different levels
within

the

NCZ

using

external

heat

exchanger

are

presented. The first method uses thermosyphon effect to

transfer the heat from different levels in NCZ to the
binary fluid.

The second method uses pumps (forced

convection)

for

theoretical

modeling

thermosyphon

heat
based

extraction.

This

paper

and

experimental

heat

extraction

presents

results

method.

for

Later

theoretical predictions and experimental results have been

compared. This investigation shows good prospects for

...
'C.c

application of this system for heat extraction from NCZ of

5-10% salinity

cBl
0<1)
Q.I

large solar ponds. By extracting heat from the NCZ the

efficiency of a solar pond could be increased up to 30%.

Dai average ambient temperature

UCZ

Keywords- Salinity-gradient solar pond; heat extraction;

thermosyphon; heat exchangers.

I.

NCZ

00

INTRODUCTION

A solar pond is a low-cost solar collector with long term

thermal storage capability. It utilizes a large body of salted
water with a density gradient from to bottom that absorb solar
radiation and stores the thermal energy [1]. The solar pond
technology has been used in Israel [2] and USA [3,4] for
power generation and low-temperature process heat
applications respectively. Solar pond is a simple and low cost
solar energy system for a relatively longer period of time. It
holds the promise of providing an alternative to Diesel

978-1-4577-1354-5/11/$ 26.002011 IEEE

LCZ

00

50

1 00

15

30 Salintly ( % by wt.)

Temperature (DC)

Fig. I The schematic of salinity gradient solar pond

23

2011

IEEE First Conference on Clean Energy and Technology CET

experimental study were found to be localised, further

experimental studies using an external heat exchanger (which
would be more practical for medium- to large-scale solar
ponds) needed to be conducted. The similar study has been
conducted by[15] both numerically and experimentally for the
0.64 m2 mini solar pond. The result showed the improvement
of the solar pond performance.
The current study proposes an alternative method of heat
extraction from the NCZ, by drawing the brine through
external heat exchangers using small pumps and re-injection
at difference levels (Figure 3). This study is also looking at the
possibility of circulating the water in the heat exchanger
without using pumps. This paper presents theoretical
modeling and experimental results for thermosyphon based
heat extraction system. The experimental results will then
used for validation of the theoretical modeling.

Fig. 2 A 50m2 experimental salinity gradient solar pond at RMIT University

Conventionally heat has been extracted from solar ponds

solely from the LCZ by two common methods. The first
method is by withdrawing the hot brine through an external
heat exchanger as experimented in EI Paso, Texas [4], Kutch
in India [5], Beith Ha'arava in Israel [2] and Singapore [6].
The second method is by using an in-pond heat exchanger,
where the cold water is used as heat transfer fluid as
demonstrated at Pyramid Hill, Victoria [7] and Marshad in
Iran [8] . Tabor [9] has discussed both of these methods in his
review of solar pond technology, and stressed that both
method have convenient practical merit.

Heat

Temperature
sensors

-t;;;:;:;;-',--,

Hot

pump

NCZ
______

The main concerns with solar ponds is the thermal

performance to store and deliver the heat which is in average
of 15-20% [10]. Many researchers have studied ways of
improving the thermal performance of solar ponds. This could
be done by improving the water clarity to allow better
penetration of sunlight to the storage zone [11], putting a
reflector to reflect additional solar radiation into solar pond
[12] and by improving the method of heat extraction from
LCZ [13].

Temperature
! d'e!.1 !..

}; a

'.

-----

__________________

;.
j.
;..

Pump

-C

...... ____

;.
;.
!

;.
Fig. 3 Proposed heat extraction system from NCZ

II.

THEORETICAL MODELLING OF THERMOSIPHON HEAT

EXCHANGER

An alternative method of extracting heat from solar pond

was investigated theoretically by Andrews and Akbarzadeh
[14] with the aim of increasing the overall energy efficiency
of collecting solar radiation, storing and delivering heat. In
this method, heat is extracted from the NCZ as well as, or
instead of, from the LCZ. The theoretical analysis showed that,
heat extraction from the gradient layer has the potential to
significantly reduce heat losses to the surface and
consequently increase the overall energy efficiency of the
solar pond by up to 50%. Leblanc et al. [7] have carried out an
experimental study of the heat extraction process proposed by
[14] at RMIT University, Bundoora East, Victoria, Australia
for a period of two months by using an in-pond heat
exchanger made of polyethylene pipe. Results indicate that
heat extraction from the gradient layer increases the overall
efficiency of the solar pond by up to 55% compared with the
conventional method of extraction from the LCZ. Leblanc et
al. [7] concluded that although convection currents in the

978-1-4577-1354-5/11/$ 26.00 2011 CO IEEE

c::::::>
c::::::>

24

Thermosiphon refers to a passive heat exchanger based on a

temperature differential with gravity causing the liquids to
circulate by density variations without the need for a
mechanical pump. Tundee et al.[16] and Singh et al. [17] have
applied the similar concept of gravity assisted heat exchanger
in their experiments to remove the heat from solar pond and
potentially produce the electricity by combining it with
thermoelectric modules.
The modelling consists of 15 equations with 15 of the
following unknowns were solved using a Visual Basic-Excel
computer program.

(U,

V, h i' h o,Tco,Tho,Ts, M,;; ,/2,ReI ,Re2,Red

Nu ,Nu
i
o

'

2011

IEEE First Conference on Clean Energy and Technology CET

[p.g.p(L -IXThi -Tho)]

+ [PogoPoLc(Th' _ ( Tho; Th, ))]

fz Ts
Re2 Red

Copper

M'

coil

' ]+[/,( pv' ]

pu
H
H
+ [ pu2 ] + [ pu2 ]

(5)

[r.( 2L;Lc
ki

ko

(6)

Where, the friction factors in fully developed flow in the

thermosiphon and coil are given by equations (7) and (8).

J;

12

Fig. 4 Thermosiphon heat exchanger system schematic diagram

The following mathematical model of the system was

derived to determine the performance of the heat exchanger,
based on the thermosiphon heat exchanger system as shown in
figure 4.

Cp(Tco -TcJ u4 2 Cp(Thi -ThJ

,,;, Cp(]Co - ) mllho [( Th, 1ho ) -T, ]

ff

1c,

(0.79lnRel-1.64(
3000 < ReI < 5xl06

(7)

(0.79lnRez -1.64tZ
3000 < Rez < 5xl06

(8)

The Reynolds numbers for the flow in thermosiphon and

through the circular coil are given by equations (9) and (10).

The energy balance in the system is defined through equation

(1), considering the heat flow into the cooling coil and the
heat taken away by the cooling water by equation (2) and the
heat transfer rate due to forced convection between the hot
water outside the coil wall and inside the coil by equation (3).

me

puD

ReI

=--

Rez

(9)

Jl

pVDc

(10)
Jl
The Nusselt number due to convection heat transfer from the
coil waH and inside the copper tube are given by equations (11)
and (12)

(1)

Nuo

0.913Re618Pr1/3

(2)

(11)
(12)

where, the Reynolds number inside the copper tube is;

(13)

The difference of velocity due to the restriction by cooling

coil in the system is given by;

The convection coefficient heat transfer inside and outside of

the copper tube are expressed by equation (14) and (15)
respectively.

(4)

Nu hid
Nu hoLe
I

The available pressure due to buoyancy and the pressure loss

using Darcy's pressure drop correlation in the system are
expressed by equations (5) and (6).

978-1-4577-1354-5/11/$ 26.00 2011 IEEE

25

(14)
(15)

2011

IEEE First Conference on Clean Energy and Technology CET

III. EXPERIMENTAL STUDY

An experimental study was conducted to measure the actual
performance of a thermosiphon heat exchanger. The
arrangement is shown in Figure 5. The water will rise to fill
the U-shape siphon as the vacuum pump is turned on. The
vacuum pump is shut off when the water reached the desired
level, where the pressure is approximately -10 kPa. The data
acquisition system is switched on for five minutes to check the
stability of all the temperatures before the heater is turned on.
After five minutes, the cold water valve is opened to allow tap
water to flow through the heat exchanger coil and the flow
rate in is measured. As the tank water gets hotter, it starts to
rise due to the natural convection process, and as it flows
through the heat exchanger, it will lose some heat making it
colder. Therefore it will fall back to the tank, and the cycle is
repeated. The outcome from this process is that the
temperature of the cold water outlet will rise to create a
temperature difference between inlet and outlet.

(b)
Fig. 5 (al Schematic diagram of thermosiphon heat exchanger (bl
Thermosiphon H=heat exchanger experimental set-up

A monitoring system using a DataTaker (DT800) data

acquisition system was used to allow continuous temperature
measurements for six different locations. Thermocouples were
placed at the inlet and outlet of the cold water; inlet and outlet
of the hot water on the siphon; in the tank; and next to the rig
to measure the ambient temperature.

A flow visualisation technique, adopted from [18], was

used to measure the velocity of the hot water in the U-shaped
thermosiphon by injecting dye into the system. The dye is
injected, while the water is circulating in the system, to the
siphon pipe through a small tube. A Canon digital camera was
used to capture the movement of the dye in the video mode. A
stopwatch was used to record the time as the dye travel passed
by the marked scale with 0.1 m interval on the clear pipe
section. By analysing the series of video obtained, the distance
travelled by the dye during a particular time was calculated in
order to determine the flow velocity in thermosiphon.

Heated water
outlet
Visualisation

IV. RESULTS AND DISCUSSION

Copper coil

Theoretically the system is initiated with input values of the

physical geometry, the water properties with relevant
coefficient of loss factor and the temperature of hot water and
cooling water. The system has been modelled with the effect
of cooling water mass flow rate through the heat exchanger
coil of 0.033 Lis, 0.04 Lis, 0.05 Lis, 0.067 Lis and 0.1 Lis.
The experiments have been carried out accordingly applying
the same mass flow rates. As a result, the input temperatures
of hot and cold water were obtained as the system is reaching
the steady states as shown in table 1. These temperatures
becomes an input values to the theoretical modelling.
Immersion Heater

TABLE!
EXPERIMETALS TEMPERATURES OUTCOMES

(a)

me

(Lis)

0.033
0.04
0.05
0.067
0.1

978-1-4577-1354-5/11/$ 26.00 2011 CO IEEE

26

Thi, oc

40.11
38.28
36.66
34.05
32.34

Tci, oC

17.06
16.6
16.89
16.72
16.55

2011

IEEE First Conference on Clean Energy and Technology CET

The obtained results from theoretical and experimental

were compared and are shown in figure 6 to 8.

Theoretically the velocity of hot water shows a slight

decreased of velocity between minimum and maximum of the
cooling water flow rates as shown by the values in figure 7.
The similar phenomenon occurs in experimental, where the
velocity is decreased with the increased of the cooling water
mass flow rates from 0.0581 m/s at 0.033 Lis to 0.0538 mls at
0.1 Lis. The effect of buoyancy dominantly has driven the
velocity of the hot water due to the temperature difference,
where the larger temperature difference will cause more heat
loss from hot water and make the water less dense and flow
with high velocity.

45.00
40.00
35.00
30.00
f!
.'l 25.00

.,
0- 20.00

()

15.00

10.00
5.00

2600
-fr-Thi

-o-Tha Thea

....... Tci

...... Tea Thea

....... Teo Exp

-o-Tha Exp

0.00 +----------j
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11
Cooling water mass flow rates (Lis)

...
"
l:

Fig. 6 Temperature difference of cooling water and hot water

Figure 6 shows the temperature differences of cold and hot

water for theoretical and experimental respectively.
Temperature difference of cold water are getting decreased
with the increased of the cooling water mass flow rates,
approximately 16 aC for flow rate of 0.033L1s and reduced to
around 5 DC at flow rate of 0.1 Lis theoretically. This
phenomenon clearly due to the effect of the mass flow rates,
where the lower the flow rates will caused more heat to be
absorbed by the cooling water inside the coil and vice versa.
The results from experiments shows slightly higher
temperature on the output, but it curve proved the similar
phenomenon is achieved. The hot water temperature
difference is remains constant approximately of 1.2 DC with
the variations of cooling water mass flow rates, for both
experimental and theoretical. This condition is due to the
effect of the constant hot water temperature in the tank, where
the return hot water will gradually gain the heat again before it
re-enter the thermosiphon heat exchanger.

....... Qe Thea

...... Qe Exp

-0- Qh Exp

-+- Qh Thea

The amount of heat extracted by the cooling water and the

heat removed from the hot water theoretically and from
experiment is shown in figure 8. Apparently, the theoretical
results show the balance of the heat transfer phenomenon with
heat which has been extracted by the sooling water and heat
that had been removed from the hot water. This is to proved
the validity of the modelled equations for this thermosiphon
system. But, the experimental result shows a larger difference
between heat extracted by cooling water and the heat removed
from the hot water, approximately 8%-12% compared to
theoretical result. The results are proportional to the obtained
velocities as discussed previously. This is understandable, as
velocity is one of the parameters to determine the quantity of
the energy produced. Theoretically, 1644 w of energy can be
produced and 2283 w of energy has been extracted from this
thermosiphon system.

0.0550

Collectively, all results in figure 6, 7 and 8 show a good

agreement between theoretical and experiments. The
sensitivity of velocity of hot water, u though with small
changes may lead to large discrepancy on the heat extracted
by cooling water and heat removed from the hot water. As an
example, with the difference of 0.0003 mls of hot water
velocity between mass flow rate of 0.033 Lis and 0.04 Lis has
resulted a difference of 304.46 W of heat removed from the
hot water as shown in figure 8.

0.0500
Exp

0.0450 +---------j
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11
Cooling water mass flow rate (Lis)

Fig. 7 Variation of hot water velocity, u inside the thermosiphon

978-1-4577-1354-5/11/$ 26.00 2011 IEEE

1600

Fig. 8 Energy Extracted from variation of cooling water mass flow rate

...... u hot

1800

Cooling water mass flow rate (Lis)

-0- u hot Thea

2000

1000 +------,---------"
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 0.11

2116

1200

iii 0.0600

Qi
>

2283

2200

1400

0.0650 ,------,

2400

27

2011

V.

IEEE First Conference on Clean Energy and Technology CET

CONCLUSIONS

h
s

new method of extracting heat from the NCZ of a solar

pond using thermosiphon heat exchanger has been studied
theoretically and experimentally. Both theoretical and
experiment work has been carried out to realised the concept
of this heat extraction system. The theoretical modelling show
valid results and encouraged a further study. A prototype
thermosiphon system with a coiled heat exchanger has been
manufactured and tested in laboratory. The preliminary test
results obtained from it showed a good agreement at the
variations of cooling water mass flow rates from 0.033 Lis to
0.1 Lis. The 2283 W of heat has been extracted from this
experiments has provide a positive motivation of
implementing this system to the improved the overall
efficiency of heat extraction from solar ponds. Further study
to observe the thermosiphon heat exchanger performance in
solar pond is necessary in order to prove the concept of this
initial result.
A

d
C
T

REFERENCES
[I]
[2]

[3]
[4]

Proceeding of the3rd International Conference on Progress in

[7]

The authors would like to express their appreciation to the

School of Graduate Research, RMIT University for providing
financial support for attending 8th SEE conference.

[8]

NOMENCLATURE

[11]

!1

Dynamic viscosity (N.s/m2)

g
LJP

f
Re
Nu
A

D
d

L
I

u
Th
Tc
h

Exp
Theo

[12]

Density (kg/m3)
Thermal expansion (K'I)
Prandtl number

Pr

[9]
[10]

Mass flow rate of cooling water (kg S'I)

Specific heat (J/kg.C)

fJ

[13]

[14]

Thermal conductivity (W/m. C)I Losses

gravity (m2/s)
Pressure difference caC)

[15]

Losses due to friction

Reynolds number
Nusselt number
Area (m2)
Diameter of thermosiphon pipe (m)
Diameter of copper tube (m)
Height (m)
Length of copper tube (m)
Velocity surrounding heat exchanger (m/s)
Velocity in Siphon (m/s)
Temperature of hot water (0C)

[16]

[17]

[18]

Temperature of cold water (0C)

Convection heat transfer coefficient (W/m2 0C)
Experiment
Theoretical

Subscripts
I
2
a

thermosiphon
circular heat exchanger coil
inlet
outlet

978-1-4577-1354-5/11/$ 26.00 2011 CO IEEE

1993.
A. Kumar and V. V. N. Kishore, "Construction and Operational
Experience of A 6000 m2 Solar Pond at Kutch, India, " Solar
Energy, vol. 65, pp. 237-249, 1999.
T. H. Kho, et al., "Design and Performance Evaluation of a Solar
Pond for Industrial Process Heating, " Int. J Sustainable Energy,
vol. 10, pp. 83-101, 1991.
1. Leblanc, et al., "Heat Extraction methods from salinity gradient
solar ponds and introduction system of heat extraction for
improved efficiency, " Solar Energy, in Press, 2010.
M. R. Jaefarzadeh, "Heat extraction from a salinity-gradient solar
pond using in pond heat exchanger, " Applied Thermal Engineering,
vol. 26, pp. 1858-1865, 2006.
H. Tabor, "Solar Ponds, " Solar Energy, vol. 27, pp. 181-194, 1981.
Y. F. Wang and A. Akbarzadeh, "A study on the transient
behaviour of solar ponds, " Energy, vol. 7, pp. 1005-1017, 1982.
Y. F. Wang and Yagoobi, "Effect of water turbidity on thermal
performance of salt gradient solar pond, " Solar Energy, vol. 54, pp.
301-308, 1995.
S. Aboul-Enein, et al., "Parametric study of a shallow solar-pond
under the batch mode of heat extraction, " Applied Energy, vol. 78,
pp. 159-177, 2004.
A. Akbarzadeh, et al., "Solar Pond Technologies: A review and
Future Directions, " in Advances in Solar Energy. vol. 16, ed:
Earthscan, UK, 2005, pp. 233-294.
1. Andrews and A. Akbarzadeh, "Enhancing the thermal efficiency
of solar ponds by extracting heat from the gradient layer, " Solar
Energy, vol. 78, pp. 704-716, 2005.
M. M. Ould Dah, et aI., "The influence of the heat extraction mode
on the performance and stability of a mini solar pond, " Applied
Energy, vol. 87, pp. 3005-3010, 2010.
S. Tundee, et aI., "Heat extraction from salinity-gradient solar
ponds using heat pipe heat exchangers, " Solar Energy, vol. 84, pp.
1706-1716, 2010.
R. Singh, et aI., "Electric power generation from solar ponds using
combined thermosyphon and thermoelectric modules, " Solar
Energy, vol. 85, pp. 371-378, 2011.
T. T. Lim, "Flow Visualisation : Techniques and examples, " pp.
43-72, 2000.
Solar Ponds,

[5]

ACKNOWLEDGMENT

me

N. D. Kaushika, "Solar ponds: A review, " Energy Conversion and

vol. 24, pp. 353-376, 1984.
H. Z. Tabor and B. Doron, "The Beith Ha'arava 5MW(e) Solar
Pond Power Plant (SPPP) - Progress Report, " Solar Energy, vol.
45, pp. 247-253, 1990.
A. H. P. Swift, et aI., "Operational Result for 3355 m2 Solar Pond
in EI Paso, Texas, " presented at the Solar Engineering, 1987.
H. Xu, et al., "operating experience with the EI Paso Solar Pond, "

Management,

[6]

Cp

cold
hot
heat exchanger coil surface
copper tube
heat exchanger coil
thermosiphon pipe

28