Energy Conversion and Management 101 (2015) 239254

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Experimental investigation of a building integrated photovoltaic/thermal

roof collector combined with a liquid desiccant enhanced indirect
evaporative cooling system
Mahmut Sami Buker , Blaise Mempouo, Saffa B. Riffat
Institute of Sustainable Energy Technology, University of Nottingham, Nottingham NG7 2RD, UK

a r t i c l e

i n f o

Article history:
Received 11 February 2015
Accepted 8 May 2015
Available online 7 June 2015
Keywords:
Solar thermal roof collector
Liquid desiccant dehumidication
Indirect evaporative cooling
Tri-generation
Numerical model
Energy analysis

a b s t r a c t
Large consumption of limited conventional fossil fuel resources, economic and environmental problems
associated with the global warming and climate change have emphasized the immediate need to transition to renewable energy resources. Solar thermal applications along with renewable energy based cooling practices have attracted considerable interest towards sustainable solutions promising various
technical, economic and environmental benets. This study introduces a new concept on solar thermal
energy driven liquid desiccant based dew point cooling system that integrates several green technologies; including photovoltaic modules, polyethylene heat exchanger loop and a combined liquid desiccant
dehumidication-indirect evaporative air conditioning unit. A pilot scale experimental set-up was developed and tested to investigate the performance of the proposed system and inuence of the various
parameters such as weather condition, air ow and regeneration temperature. A cost effective,
easy-to-make polyethylene heat exchanger loop was employed underneath PV panels for heat generation. In addition, a liquid desiccant enhanced dew point cooling unit was utilized to provide air conditioning through dehumidication of humid air and indirect evaporative cooling. The experimental results
show that the proposed tri-generation system is capable of providing about 3 kW of heating, 5.2 kW of
cooling power and 10.3 MW h/year power generation, respectively. The ndings conrm the potential
of the examined technology, and elucidate the specic conclusions for the practice of such systems.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Utilization of sunlight for cooling is a long-sought goal. As the
demand for cooling is proportional to the solar intensity, thus the
time of peak cooling need coincides with the time of maximum
resource occurs. Given this relation, it is no doubt that there has
been a considerable interest to produce economical solar cooling
technologies. Heat-activated systems that mainly driven by heat
input from solar thermal energy have been introduced allowing
simultaneous production of heat and cooling/refrigeration [1].
Compared to the conventional vapour compression systems, thermally driven air-conditioning technique would be an effective
alternative in terms of increasing primary energy savings with less
power consumption and therefore less greenhouse gas emissions
and hazardous materials and pollutants depleted to the environment [2].
Corresponding author. Tel.: +44 7874364852.
E-mail addresses: Mahmut.Buker@nottingham.ac.uk, msbuker@gmail.com
(M.S. Buker).
http://dx.doi.org/10.1016/j.enconman.2015.05.026
0196-8904/ 2015 Elsevier Ltd. All rights reserved.

A number of experimental investigations have been introduced

in the literature for solar cooling systems. Huang et al. [3] conducted an experimental study of solar heat driven ejector cooling
system for cooling load of 3.5 kW and cooling time of 10 h within
two locations; Taipei and Tainan, respectively. Bermejo et al. [4]
investigated a hybrid solar/gas cooling plant composed of a double
effect LiBr + water absorption chiller with cooling capacity of
174 kW, a linear concentrating Fresnel collector and direct-red
natural gas burner. Another experimental study was performed
by Ge et al. [5] employing a two-stage solar driven rotary desiccant
cooling system with newly developed silica gel-haloid composite
desiccant to achieve reducing regeneration temperature and high
energy performance. A solar electric-vapour compression refrigeration system was tested by Bilgili [6]. Eicker et al. [7] developed
and tested a new photovoltaic-thermal (PVT) system to produce
both power and cooling energy for night radiative cooling of
buildings.
In addition, several studies have investigated the solar assisted
liquid desiccant cooling systems. To name a few, Katejanekarn
et al. [8] presented an experimental study of a solar-regenerated

240

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

Nomenclature
Aeff
C
Di
Do
e
F
F0
Gr
g
h
I
K
L
m
M
_
m
Nu
Q
T
(TPV)eff
U
W
V

effective area, m2
specic heat
inner diameter of heat exchanger tube, m
outer diameter of heat exchanger tube, m
percent deviation
n efciency
at plate collector efciency factor
Grashoff number
gravitational acceleration, m/s2
heat transfer coefcient, W/m2K
incident solar radiation, W/m2
thermal conductivity, W/mK
cubic root of dwelling volume, m3
variable dened to solve differential equations
mass, kg
mass ow rate, m/s
Nusselt number
energy, W
temperature, C
effective PV module temperature, C
thermal transfer coefcient, W/m2K
distance between tubes, m
wind speed, m/s

Greek letters
a
absorption
bp
temperature coefcient for PV efciency, Cbs
tilt-angle of PV panels,
d
thickness, m
e
emissivity
g
efciency, %
,
Stefan-Boltzmann constant, W/m2 K4
s
transmittance
v
kinematic viscosity, m2/s
q
density, kg/m3
n
packing factor
k
thermal conductivity, W/mK

liquid desiccant ventilation pre-conditioning system in hot and

humid climate of Thailand. They reported 1.2 C temperature drop
and 11.1% relative humidity reduction of the delivered air. Li et al.
[9] experimentally studied an open cycle solar desiccant dehumidication air-conditioning system in Hong-Kong and a payback period around 7 years was reported. In another study [10], the authors
experimentally investigated a desiccant air handling unit powered
by vacuum tube solar collectors in a moderately humid climate
with regeneration solely by solar energy and the electrical consumption based performance indicator was found to be 4.5. Few
researchers have studied the solar enhanced liquid desiccant air
conditioning; Qi et al. [11] for commercial buildings in cities of four
main climate regions including Singapore, Houston, Boulder and
Los Angeles; Fong et al. [12] for high tech ofces in subtropical climate; Aly et al. [13] for modelling a solar-powered open absorption cycle with two desiccant solutions and Zeidan et al. [14] for
modelling and simulation of solar liquid desiccant regenerator in
open absorption cooling cycle; and they all indicate that solar
assisted liquid desiccant cooling systems have a large potential
especially in hot and humid areas but additional research and
development is still needed to prove their potential.
Combining solar power and a liquid desiccant cooling system
can provide cooling demands for residential and building applications in addition to thermal comfort and indoor air quality.

Subscripts
a
air
air
air layer
abs
absorber layer (PV cell)
c
cover layer
cool
cooling
cv
convective
db
dry bulb
dp
dew point
e
electrical
eq
equilibrium
eV,cool evaporative cooler
f
uid
hein
inner wall of heat exchanger
heo
outer wall of heat exchanger
heo, hein outer wall to inner wall of heat exchanger
hein, w inner wall of heat exchanger to water
in
inlet
l
loss
o
overall
out
outlet
p
pipe
pl
EVA plastic back of a PV module
pv, heo PV cell layer to outer wall of heat
rd
radiative
rs
reference situation
s
sky
sa
supply air
t
thermal
tl
tile
tl, heo
tile to outer wall of heat exchanger
v
vapour
w
water
wb
wet bulb

Although substantial studies have been performed on the investigation of the regenerators of liquid desiccant cooling systems from
solar thermal applications, regeneration heat capture by such a
unique heat exchanger loop with PV modules has yet to be considered. In this study, the performance of the liquid desiccant cooling
system with polyethylene heat exchanger loop unit driven regenerator unit is experimentally investigated. The proposed system
comprises polyethylene roof loop, a desiccant cooling system with
a liquid desiccant-based dehumidication unit and an indirect
evaporative cooling unit to dehumidify and cool the supply air.
An experimental unit was set up and the system elements were
examined experimentally. The system arrangement is described
and experimental data is reported.
2. System conguration
The overall conguration of the combined system is presented
in Fig. 1. The operation of the system is divided into three main
processes; roof, dehumidication and cooling. A polyethylene heat
exchanger is employed to capture heat from solar PV system to
supply regeneration heat to regenerate the dilute desiccant solution. Regeneration temperature is at around 6070 C to the desiccant regenerator as shown Fig. 1. The desiccant cooling unit
utilized in this system can provide thermal comfort and good

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

241

Fig. 1. The system conguration of the combined system.

indoor air quality through the desiccant dehumidication. An indirect evaporative cooling unit is also integrated with the liquid desiccant dehumidication unit to supply supplementary cooling
effect to full the cooling demands. The proposed system can operate in both dry and humid regions as employing an integrated liquid desiccant-dew point evaporative cooling unit. In humid
environmental conditions, ambient air is dehumidied through
the liquid desiccant dehumidication process and then introduced
to the dew-point cooling system for cooling effect. However, ambient air can be plainly supplied to the indirect evaporative cooler
without any need of a dehumidication process. Next section is
about built-up of the experimental system and description of the
sub-units.

3. Experimental setup
3.1. Roof unit
Building integrated photovoltaic thermal (BIPVT) systems,
inherently, have potential to appear as a major source of renewable
energy on the roof top of a building to generate electrical energy
and produce thermal energy required for thermal applications
[15]. This hybrid technology combines PV and solar thermal components into a single module to boost the solar conversion efciency and achieve economical use of space. However, as PV/T is
a newly emerged technology, various technical challenges relevant
to PV/T found in practice e.g., current technical status, difculties
and problems that prevented wide-scale applications of these systems [16]. Moreover, inability of converting already-built photovoltaic (PV) systems into PV/T systems could be also mentioned
as a practical drawback. Therefore, polyethylene heat exchanger
as a concealed heat extraction component without damaging original conguration of PV modules represent an effective solution to
full aforementioned pitfalls and practical limitations of the existing systems. Moreover, polyethylene heat exchanger propitiously
blends into its surroundings free from any add-on appearances
and has a dual function as absorbing heat and passively cooling
of photovoltaic modules via extracting waste heat from solar panels. Fig. 2 shows the schematic diagram of the roof unit.
The method and mathematical equations presented in this
study, are not necessarily the most accurate available; but they
are widely applied, easy to use and adequate for most of the design
computation. A modelling program, incorporating location dependent radiation data, PV collector data, liquid desiccant

dehumidication and dew point cooler data was developed and

applied. A ow chart for the modelling is given in Figs. 3 and 4
illustrates the cross-sectional view of the roof units heat transfer
prole.
Employing 84 Sharp NU-R245 (J5) PV modules in the system,
incoming solar energy initially absorbed by the PV cells can be
expressed as:

Q abs sc  aabs  Aeff  I

The part of incoming solar energy is used to generate power

energy by silicon cells, partly carried through the heat exchanger
and the rest is dissipated to the ambient due to temperature difference between solar module and atmosphere. A view factor (proportion of the radiation which leaves one surface that strikes
another) of 1 is assumed as the heat loss between the back plate
and the tubes would be minimal since the corrugated roof surface
also transfers heat to the water owing through heat exchanger.
So, the loss can be expressed:

Q l hcv  T c  T a ec  r  T 4c  T 4s hair  T pl  T heo

epl  r  T 4pl  T 4heo

Combination above equations can provide a revised expression

of the useful heat energy (Qt):

Q t Q abs  Q l

Following empirical relation can estimate the PV cover temperature (Tc) [34]:

T c 30 0:0175  I  300 1:14  T a  25

This relation is not only applied for standard pc-Si PV modules

but in PV/T systems, the relation also accounts for the system operating conditions, such as the effect of heat extraction uid.
Therefore, the parameter (TPV)eff is considered to correspond to
the PV module temperature in terms of operating conditions of
PV/T systems. Thus, for the proposed system, (TPV)eff is given by
[34]:

T PV eff T c T PV=T  T a

McAdams widely used convective heat transfer coefcient of

the ambient air in solar collector theory:

hcv 5:7 3:8  V

The sky temperature as a function of the ambient temperature

can be obtained by Swinbanks equation:

242

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

Fig. 2. Schematic diagram of the roof unit.

Ta
T s 0:037536  T 1:5
a 0:32

where hrd1 and hrd2 are the radiation heat transfer through front and
rear side of the collectors, respectively and given by:

hrd1 ec  r  T 4c  T 4s

hrd2 epl  r  T 4pl  T 4heo

The associated convective heat transfer coefcient, hair, assuming a natural convective stagnant air layer between the PV module
and heat exchanger can be expressed as [37]:

hair

Nu  K air
dair

10

where thermal conductivity of air is Kair and thickness of the air

layer between PV and heat exchanger is dair, respectively. Then
Nusselt number can be given by [35]:


 
bs
 Gr1=3
Nu 0:06  0:017 
90

11

Then, Grashoff number is given by [35]:

Gr

g  T pl  T heo  d3air
V 2air  T air

12

In addition, instantaneous thermal efciency of the roof unit is

given by:

gt

Fig. 3. Flow chart of thermal analysis computational algorithm.

Qt
Aeff  I

13

Fig. 4. Heat transfer of the PV/T system.

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

Conductive heat ow takes place across the PV cover, cell layer,

and EVA plastic layer on the panel rear. Under natural convective
air layer between PV module and polyethylene heat exchanger
which is also measured, conductive effort, again, takes place
between aluminium layer and heat exchanger wall and then, heat
is eventually introduced to the water. The heat gain of water is as
equal as the useful heat energy itself and is given by:

Q t Aeff  U t  T abs  T w

14

Overall heat transfer coefcient between PV cell and water can

be expressed as:
1

1
1
1
U t U 1
c;abs U abs;pl U heo;hein U hein;w

Heat transfer coefcient (Uc,

(absorber) can be expressed as:

U c;abs

15

abs) from cover to PV cell layer

kc
dc

16

Then, average temperature at the absorber (cell layer) becomes:

T abs T c 

Qt
Aeff  U c;abs

kabs
dabs

18

Qt
Aeff  U abs;pl

19

Heat transfer coefcient (Uheo, r) from outer wall to interior wall

of heat exchanger is given by [35]:

U heo;hein

2  p  khe
h i
ln DDo

20

The average temperature at the inner wall of polyethylene heat

exchanger is given by [35]:

T hein T heo 

Qt
Aeff  U heo;hein

21

Heat transfer coefcient (Uhein,hein) from interior wall of heat

exchanger to the water is given by:

U hein;w

Nur  kr
Di

22

where Nusselt number, for such a smooth conduit case, is given by

[36]:
0:4
Nur 0:023  Re0:8
r  Pr r

23

The root mean square percent deviation (e) and the coefcient
of correlation (r) is computed in order to match the theoretical
results with the experimental outcomes by using the following
expressions:

s


Rei 2
Xi  Y i
 100
e
where ei
N
Xi
nRX  Y  RX  RY
r q q
nRX 2  RX2  nRY 2  RY2
Stating the expressions by:

26

Then, the n efciency factor (F) can be given as:

o
tanh m WD
2

o
m WD
2

27

The at plate collector efciency (F0 ) then becomes:

F0

WU t
Do U abs;pl

1
WU t
t
WU
Do WD
K
o F

28

The ow rate factor (FR) is given as:

FR




_ f
Aeff U t F 0
mC
1

exp

_ f
mC
Aeff U t F 0

29

_ of water is calculated for the measured

The mass ow rate (m)
experimental values of average solar intensity (761.54 W/m2), and
outside temperature (37 C) for the given day. Given values are
used to evaluate the water temperature and solar cell temperature
of the PV modules for validation.
3.2. Energy efciency

The temperature at the outer wall of the polyethylene heat

exchanger is given by:

T pl T abs 

Ut
Kd

17

Heat transfer coefcient (Uabs,pl) from PV cell (absorber) to the

EVA plastic layer is given by:

U abs;pl

m2

243

24

25

In order to determine the instantaneous power energy efciency and power output, some of the widely used energy equations are employed as below;
Temperature dependent power energy efciency of a PV module (ge) can be given as:

ge grc 1  bp T c  T rc

30

The rate of instantaneous solar energy available on solar cell

and power energy output is given by:

Q e as  n  It  Aeff

31

Packing factor for a PV module is stated as:

Acell
Acoll

32

The overall conguration of the solar assisted liquid desiccant

cooling system is presented in Fig. 1 with various system components. The system has three main loops; polyethylene heat exchanger loop, dehumidication loop and cooling loop. The proposed
roof system consists of polyethylene heat exchanger loop underneath PV modules to form a PV/Thermal roof collector. The complete roof structure has several layers, an outer cover; a layer of
photovoltaic cells beneath the cover; EVA plastic layer at the back
of PV adjacent to the PV cells layer, polyethylene heat exchanger
and roof support. There are eighty-four (84) Sharp NU-R245 (J5)
photovoltaic solar panels mounted on the roof which are entirely
for power generation without any thermal function. The solar panels will convert part of the incoming solar radiation into power
energy due to the photogalvanic effect of the silicon cells and
remaining heat energy will be conveyed through the circulating
water across the heat exchanger. The hot water can be used for
heating & cooling, domestic hot water supply, food drying, natural
ventilation inside the building and more. Part of the heat required
to drive regeneration unit of the desiccant dehumidication system is expected to be obtained from waste heat dissipated by solar
panels. Fig. 5 illustrates the experimental setup built for the polyethylene heat exchanger system. As the heat exchanger is loosely
contacted with the corrugated roof, any radiative exchange would
be miniscule and thus disregarded.
The PV solar collectors absorb a greater proportion of incoming
solar radiation on the PV cell surface, while dissipating remainder
to ambient as waste heat. From top to bottom, a transparent and

244

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254
Table 2
Technical parameters of the roof unit.
Component

Parameter

Value

PV modules

No. of module
Module dimensions
Cell type:

84
1652  994  46
Monocrystalline
0.91
14.9%
245 W
K = 80%

Packing factor
Conversion efciency
Module peak power
Estimated power/performance
efciency
Maximum voltage, Vm
Maximum current, Im
Open circuit voltage, Voc
Short circuit current, Isc
Fig. 5. Integration of the polyethylene loop underneath PV modules.

thermally resistant outer cover enables to reduce dissipation and

secures that maximum solar radiation reaches the photovoltaic
cells. Meanwhile, the temperature of the cell layer needs to be lowered so as to facilitate the solar power conversion efciency of PV
panels. Heat ows from ambient to the photovoltaic cells in case of
lower temperature of cell surface. Thus, adjusting circulating water
temperature at a lower range helps to reduce the cell surface temperature to some extent. Although rst-of-its-kind polyethylene
heat exchanger, due to its physical structure, is loosely adhered
to the rear surface because of the aluminium frame of the PV modules, less exible pipes and risers, relatively low temperature difference between the cell surface and circulating water is
expected. Furthermore, the physical structure of the poly heat
exchanger prevents the use of thermally conductive adhesives
due to uneven contact to the rear of PV modules. Consequently,
any decrease in the cell temperature will collaterally enhance both
power and thermal efciencies. Solar optical parameters of the PV
modules are presented in Table 1.
The piping system is placed below the roof truss and joint by
exible coupled connectors with valves. In order to provide a leak
free connection, the pipes and ttings are engaged tightly via welding with soldering iron. The supply pipe feeds cold water into the
polyethylene heat exchanger whilst the return pipe transports
the hot water from the heat exchanger. The PV modules and steel
corrugated roof support are well clamped together, and polyethylene heat exchanger is located in between. Technical parameters of
the PV solar collectors and polyethylene heat exchanger are shown
in Table 2.
Circulating pumps should cope with head pressures every time
it turns on, in close loop solar roof systems. Also, the pumps must
be compatible with solar applications in point of size, hydraulics,
speed stages, efciency and power consumption. The details of
the circulation pumps are presented in Table 2.
K type thermocouples with a 0.25% of accuracy, 250 C of maximum temperature measurement and 1.6 s of response time were
inserted for various temperature measurements. A Kipp & Zonen
pyranometer with the sensitivity of 4.47  104 V/Wm2 and
response time of 1.66 s were integrated on a vertical surface
(10 C South) to measure the global solar radiation. Wind speed

30.5
8.04
37.5
8.73

Exposed roof area (Aeff)

and Tilt angle (bs)

Active total area

South facing active area of PV
arrays with heat exchanger
Tilt angle

140 m2
40 m2

Polyethylene heat
exchanger

The internal diameters of tubes

The external diameters of tubes
Distance between tubes
Length
Height
Max temperature allowed in
poly HE
Max pressure allowed in poly
HE

Di = 0.0027 m
Do = 0.0043 m
W = 0.001 m
L = 10 m
H=1m
60 C

Model:
Max delivery head
Max operating pressure
Permissible temperature range
Mains connection

WILO
6m
10 bar
10 C to
+110 C
1230, 50 Hz

Dimensions
Capacity

0.9 m 0.45 m
120 l

Circulating pump

Hot water storage

10

1 MPa

meter was also utilized to nd out the heat ow rate through natural convection between the PV cover and ambient and connected
along with all other measuring devices to a Datataker DT500 data
logger.
3.3. Liquid desiccant dehumidication unit
Recently, substantial amount of research has mainly focused on
desiccant substances and their practical applications particularly
on dehumidication and cooling practices as a solar assisted zero
carryover liquid desiccant air conditioner [17], experimental performance test of a liquid desiccant dehumidication system under
tropical climates [18], an aqueous lithium chloride desiccant air
dehumidication and regeneration [19], and a liquid desiccant system with a structured packing dehumidier/regenerator [20]. The
desiccants are either natural or synthetic materials having the ability of absorbing or adsorbing moisture through water vapour pressure difference between the ambient air and the desiccant surface.
They could be solid, either polymer sorbent or porous material as
silica gel and zeolite, liquid solution as lithium bromide solution
and lithium chloride with water, etc. . ., [21]. In terms of energy
consumption, desiccant air conditioning systems consume less

Table 1
Solar optical parameters of PV cells and glass cover.
PV module type

Absorptance

a
Mono-crystalline-Si
Single glazing

0.9

Emissivity
e

Thermal conductivity,
(w/m C) K

Thickness
d

Reference
efciency %

Temperature
coefcient, (/C) bp

0.96
0.9

149
1

0.02

14.9

0.03

Transmittance

s
0.91

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

245

Fig. 6. Liquid desiccant based dehumidication with indirect evaporative cooling unit.

than a quarter of energy comparing to its counterparts; conventional vapour-compression systems [22]. Liquid desiccant dehumidication systems require lower regeneration and have
advanced utilization exibility and low pressure drop on air side
[23]. However, as a major drawback of conventional liquid desiccant systems, liquid desiccant particles on the absorbers could be
carried along by the process air ow over an extended surface or
packing material. This corrosive solution could eventually have a
disturbing effect on the indoor air quality and comfort of the residents. Corrosion and carry-over of liquid desiccant can be diminished by employing an effective direct contact core with a
semi-permeable micro-porous membrane for direct-liquid contact
[24]. In this study, a new, basic and low cost heat and mass exchanger core is utilized for liquid desiccant dehumidication. This
selectively permeable multiple porous membranes are employed
to reduce the carry-over of liquid desiccant particles and increase
heat and mass exchange efciency. Fig. 6 shows the overall liquid
desiccant air conditioning system consisting of a dehumidier,
regenerator and indirect evaporative cooling units.
The liquid desiccant, in the dehumidier unit, is sprinkled
evenly along the porous membrane of the desiccant core over the
humid air owing upwards whilst the heat and mass exchange
between the humid air stream and desiccant solution takes place
within the membrane. Moisture from the process air condenses,
permeates into the desiccant side and is fully absorbed into the
strong desiccant solution. As a result of water absorption, the

moisture content of the desiccant solution increases and the strong

solution becomes diluted. Also, there is a water core with porous
foam membrane in the dehumidication unit. Water from the
sump at the bottom of the unit, as shown in Fig. 6, is sprayed down
over the fresh air streaming upwards that causes water evaporation in the air channel and humidies the process air. A plate heat
exchanger is assembled between the desiccant and water core to
allow heat transfer between desiccant and water ows and
pre-cool the desiccant solution in the dehumidication process.
In addition, a regeneration core with porous foam sheets, similar
to the desiccant core, is integrated to the regenerator unit wherein
the impure desiccant solution is heated by retrieved solar heat and
air ow removes moisture to form a strong desiccant solution.
Then, the regenerated desiccant solution is sent back to the desiccant core for better dehumidication performance. In this experimental study, environmentally friendly, relatively less corrosive,
lower cost, lower density and viscosity potassium formate
(HCOOK) solution was utilized as a liquid desiccant [25].
The performance evaluation of the liquid desiccant
air-conditioning system is based on two main parameters, the
enthalpy and humidity effectiveness. The humidity, also called
moisture, effectiveness is the ratio of actual change in air humidity
to the maximum change capacity and can be computed both for
the regenerator and dehumidier units by:

ew

wa;in  wa;out
wa;in  weq

33

246

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

conditions and can be calculated from sensible effectiveness, latent

effectiveness, and the ratio of latent to sensible energy differences
across the unit (e.g., dehumidier and regenerator). It is given by:

eh

Fig. 7. The dehumidier and regeneration unit.

Weq represents the condition for when air humidity ratio at the
interfacial area in equilibrium with the liquid desiccant solution,
in kgH2O/kgair. This moisture content is the minimum level for ideal
air dehumidication and is calculated based on partial vapour pressure of the desiccant solution, PV by [2]:

weq 0:62197

PV
1:013  105  P V

34

Similarly, the enthalpy effectiveness (or combined heat and

mass transfer effectiveness) is dened as the ratio of actual air
enthalpy change to the maximum change capacity in ideal

ha;in  ha;out
ha;in  heq

35

where heq, J/kg, refers to equilibrium state of air enthalpy with the
liquid desiccant solution and can be calculated in terms of air
humidity ratio with air temperature equal to the liquid desiccant
solution.
The liquid desiccant dehumidication units utilized in the
experimental setup is shown in Fig. 7. The dehumidier unit with
the outer dimensions of 90  50  130 cm has a desiccant core
inside the dehumidication part with the dimensions of
50  50  50 cm and water core having with dimension of
35  50  50 cm inside the adjacent water cooling part. In addition
to the dehumidier, a regenerator unit is employed with the given
dimensions of 30  30  85 cm along with a counter ow regeneration core with 30x30x30 cm dimensions. There are three pipe
connections between the dehumidier and regeneration units.
For those connections, one is for strong desiccant ow to the dehumidier; another is for weak desiccant ow to the regenerator and
the last one for balance of desiccant solution level between two
units. Pre-heated hot water from the roof part is rst collected in
hot water storage and the temperature is detected there if the temperature level is satisfactory for regenerating the weak desiccant
solution which is usually at around 6070 C. It should be noted
that the heat source for the desiccant regenerator is of decisive
among other units to run the complete desiccant system. If the
temperature level is not in desired level, then, an auxiliary three
phase water heater submerged in the water tank connected to
the solar PV panels heats the water until the desired range of
regeneration temperature and stops automatically until the temperature level drops. Also, a fan coil for heat dumping could be considered in case of excessive water temperature (>70 C). For the
heat input to regenerate the weak potassium formate (HCOOK)
as desiccant solution, a plate heat exchanger is employed in the
regeneration unit. 25 l of desiccant solution having the mass concentration of 75% were stored in the liquid desiccant system. In
order to circulate the potassium formate between the dehumidier
and regenerator units, two centrifugal magnetically coupled circulation pumps were integrated enabling secure and leakage free
operation. Overall control over the performance of the system covering pumps and fans is handled by a digital controller. HMT130
humidity and temperature transmitters were used for data recording regarding air temperature and humidity at the both inlet and

Fig. 8. Schematic diagram of the indirect evaporative cooler.

247

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

Fig. 9. Experimental indirect evaporative cooling unit.

Table 3
Performance data of the indirect evaporative cooling unit.
Fan setting (m3/h)

300

600

900

1200

1500

Volumetric ow rate of supply air (l/s)

Inlet air temperature (C)
Inlet air enthalpy (kJ/kg)
Supply air Temperature (C)
Supply air relative humidity (%)
Temperature difference across the cooler (C)
Exhaust air temperature (C)
Cooling capacity (W)
COPcooler
Wet bulb effectiveness
Dew point effectiveness

63.5
30.03
64.76
15.75
78.24
14.28
16.31
1032
12.1
0.96
0.61

108.3
30.44
66.02
17.45
76.55
13
12.6
1117.5
10.5
0.91
0.56

143.4
29.57
63.37
19.99
75.69
9.58
11.9
1173.3
9.6
0.73
0.44

171.6
29.88
64.30
22.69
72.99
7.19
8.7
1211.84
7.4
0.68
0.42

202.3
30.24
65.40
22.17
72.5
7.07
9.3
1229.82
6.3
0.67
0.40

3.4. Indirect evaporative cooling unit

900

Temperature,C

60

800
50
700
40
600
30

500
Ta,C
Tw,C
(Tpv)eff,C
I, W/m

20

400

Incident solar radiation,W/m

1000

70

300

10

10

12

14

16

Time,hr
Fig. 10. Hourly variation of PV, ambient, water temperatures and incident solar
radiation.

outlet of the dehumidier and regenerator units. The HMT130

probes used for this experimental study have operation temperature range of 40 to 80 C with 0.2 C accuracy and relative
humidity range of 0100% with 1.5% accuracy and output signal
range of 010 V. Also, liquid desiccant and water temperatures in
the system were measured by using K-type thermocouples.

Evaporative cooling can be an efcient and environmentally

attractive alternative to mechanical vapour compression for air
conditioning to maintain thermal comfort especially in hot/dry
regions. These systems generally consume only a quarter of the
electric power that conventional vapour compression systems
require. Hence, such systems will save substantial amount of electric power and therefore contribute to minimizing greenhouse gas
emissions [26]. In addition to the greenhouse gas emission savings,
thermally driven desiccant enhanced evaporative cooling systems
can potentially lower the peak electricity demand and associated
electricity infrastructure expenses [27]. Indirect evaporative cooling systems, also called dew point evaporative cooling, use the
latent heat of water evaporation as a natural driving resource
and usually show high performance as their coefcient of performance (COP) is reported between the ranges of 820 [28].
Evaporative cooling systems are generally separated into two main
techniques as direct and indirect systems. Indirect evaporative systems separate process air from water so does not allow direct contact between air and water, dissimilar to the direct evaporative
cooling systems. So, the supply air is cooled sensibly without a signicant increase in the moisture content [29]. This feature makes
IEC systems even more attractive to meet comforting indoor air
quality demands. Thus, reducing air temperature below the wet
bulb temperature, or also called sub-wet bulb temperature cooling,
has been the main topic of many studies to develop a new and

248

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

The entire cooling capacity of the indirect evaporative cooling

unit can be obtained by:

70

Temperature,C

60

r = 0.994

_ sa hin  hsa
Q cool m

e = 4.03%

50
40
30
20
08:00

(TPV)eff EXP
(TPV)eff THEO

r = 0.98

As widely known, the indirect evaporative cooler coefcient of

performance is the ratio of the overall cooling capacity to the electric power consumed during the operation. In dew point cooler
case, there are two equipment that consume the overall electric
power, water circulation pump to run the water from water tank
to the distribution nozzles and air supply fan. Then, COP of the system is given by:

Tw EXP

e = 7.63%

10:00

36

Tw THEO

12:00

14:00

16:00

Time, hr
Fig. 11. Hourly variation of PV cover temperature- experimental and theoretical.

0.15

COPeV:cooler

Q cool
_ elec;cooler
W

37

The indirect evaporative cooler unit effectiveness can be evaluated by two important parameters, dew point effectiveness and
wet bulb effectiveness and are given by:

edp

T in;db  T sa;db
T in;db  T in;dp

38

ewb

T in;db  T sa;db
T in;db  T in;wb

39

nPV w/o PH

0.14

0.13

0.12

0.11

0.10
08:00

10:00

12:00

14:00

16:00

Time
Fig. 12. Effect of poly heat exchanger on power conversion efciency.

innovative evaporative cooling techniques as a comparative study

on the performance assessments of novel M cycle based air coolers
[30], a desiccant-enhanced evaporative air conditioner [31], and a
review on the recent developments in solar assisted liquid desiccant evaporative cooling technology [32]. In this regard, indirect
evaporative cooling systems, or dew point coolers, aim to reduce
air temperature under the wet bulb temperature versus inlet air
dew point temperature by using modied heat and mass exchanger congurations, techniques and designs.
The cross-section view of the indirect evaporative cooling system proposed is depicted in Fig. 8. The unit consists of cross-ow
core with bre sheets stacked together and separated by channel
guides located on one side of the each sheet to form adjacent
wet and dry channels. Each sheet is coated with polyethylene
material to prevent water penetration from the wet channels.
Also, the wet channel surfaces are plated with brous coating
material with superior water absorption ability and large capillary
forces to draw off water from the water lm surface to allow saturation. The perforated heat exchanger sheets have orderly
deployed numerous holes alongside the ow paths of dry air at
the process air ow area. As shown in Fig. 8, the shape of the heat
and mass exchanger disperses air from the dry channels to the wet
channels through perforations to create air ows perpendicular to
the primary air ow. As a result of heat and moisture exchange, the
warmer and highly saturated air is evacuated to the atmosphere.
The pre-cooling process of the process air can create a major temperature difference between the wet and dry channels so that heat
absorption capacity and cooling effectiveness of the system
increase substantially.

The indirect evaporative cooler unit, used in this study, with the
dimensions of 125  85  75 cm is shown in Fig. 9. The core of the
cooler is a cross ow heat and mass exchanger with dry and wet
channels. A submerged water circulation pump is integrated in
the water sump at the bottom of the unit for water circulation
between the collection tank and nozzles at the top of the heat
and mass exchanger sprinkling water equally all over the wet
channels of the core, shown in Fig. 6. The sprayed water is distributed alongside the vertical wet channel surfaces and collected
in the water tank then again pumped through the distribution
pipes to the nozzles. Two HMT130 humidity and temperature
probes which have direct link to the Datataker DT500 datalogger
were connected to measure air temperature both at the inlet and
outlet of the indirect evaporative cooling unit.
Overall coefcient of performance of the combination of liquid
desiccant dehumidication unit and indirect evaporative cooler is
determined by the ratio of whole cooling capacity supplied over
the thermal heat input at the regenerator to regenerate the dilute
desiccant solution. In addition, electrical coefcient of performance
of the entire unit is determined as the ratio of overall cooling
capacity delivered by the dehumidier and dew point cooler over
the total electric power consumption of the system including fans,
pumps, electrical heater etc. Both thermal and electrical COP of the
system is given by the following equations:

4500
Useful heat gain, W

4000

Heat gain, W

PV Energy Efficiency (%)

nPV with PH

3500

3000

2500

2000
08:00

10:00

12:00

14:00

Time, hr
Fig. 13. Useful heat gain.

16:00

249

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

_ stand for
_ w; m
_ des ; I and W
where subscripts Tw, Tdes, Tabs, Ta, Tsa m
temperature of outlet water and desiccant solution, temperature
of collector surface, ambient and supply air temperatures, mass
ow rate of water and desiccant solution, solar radiation and power
consumed by fans and pumps, respectively.
Total uncertainty for overall system efciency can be expressed
as;

0.20
0.18
0.16
0.14

2

0.12

@g
_ des
@m

0.15

0.20

0.25

0.30

0.35

0.40

4. Results and discussion

90

30

80
25
70
20
60
T_dehum _in (C)
T_dehum _out (C)
RH_dehum_in (%)
RH_dehum_out (%)

15

40
200

10

50

100

50

Relative Humidity, %

100

35

150

Time, mins
Fig. 15. Air temperature and relative humidity variation across the dehumidier.

COP Therm;DehumCooler

COP Elec;DehumCooler

Q cool;ov erall
_ elec;cooler
W

40

Q cool;ov erall
_ elec;total
W

41

3.5. Uncertainty analysis

The experimental uncertainties were determined by applying
Gauss propagation law. The result R is calculated as a function of
the independent variables x1, x2, x3, . . ., xn and w1, w2, w3, . . ., wn
represents the uncertainties in the independent variables. Then,
uncertainty R is expressed as [37]:

"

@R
w1
@x1

2

@R
w2
@x2

2




@R
wn
@xn

2 #1=2
42

The independent parameters measured in the experiments are:

solar radiation, temperature of solar thermal roof, temperature of
outlet water, ambient temperature, ow rates and power consumption of the cooling unit.
It is obtained from the above equations that the gt is the function of several variables, each subject to uncertainty:

gt f T w ; T des ; T abs ; T a ; T sa m_ w ; m_ des ; I; W_ Rp;f

4.1. Roof unit

Various heat transfer coefcients and solar optical parameters
regarding the roof part of the proposed system are given in
Tables 1 and 2. The wind speed was observed over the course of
test to detect the heat ow rate through natural convection
between the PV panels and ambient air, and was measured as
1.2 m/s in average. Also, average solar radiation data during the
day and outside air temperature were found to be
761.5 W/m2and 37.2 C, respectively. The climate data along with
the design parameters provided in Tables 13 were evaluated to
estimate the water and PV cell temperatures in Figs. 10 and 11.
The root mean square percent deviations and correlation coefcients were also calculated and shown in Fig. 11 in order to compare the theoretical results with the experimental outcomes.
During the series of test sessions, no working uid leakage was
observed both in the roof and cooling units.
Fig. 10 presents the hourly variations of effective PV module
temperature (TPV)eff, and water temperature Tw with the ambient
temperature, Ta and incident solar radiation, I. It is shown that
the PV module temperature throughout the operation remains
higher than the water temperature as expected. The increase in
water temperature circulating through the heat exchanger reaches
up to 16 C throughout the testing. Considering the changes in
temperature tendencies and rise in water temperature, there is a
fair agreement with the results obtained by [33].
Fig. 11 compares the theoretical an experimental values of
effective PV module temperature (TPV)eff, attained as a result of
simulation and experimental testing. The theoretical values considering the effects of polyethylene heat exchanger indicate that
(TPV)eff (THE) are in accordance with the experimental values,
(TPV)eff (EXP). The correlation coefcient and root mean square

80

40

70
30
60
20
50
10

T_reg _in (C)

T_reg _out (C)
RH_reg _in (%)
RH_reg _out (%)

0
0

50

100

150

40

Relative Humidity, %

0.10

Fig. 14. Effect of uid temperature on the overall thermal efciency of the PV/T
system.

Temperature, C

RP;f

44

(Tw-Ta)/I, C.m/W

2 
2 
2 
2 
2 
2 31=2
@g
@T@ g wT des @T@ g wT abs @T
wT a @T@ gsa wT sa @@m_gw wm_ w
a
des
abs
7
7

2
2
5

2
wmdes @@Ig wI @W_@ g wW_ P;f

wT w

0.10
0.08
0.05

wR

@g

6 @T w
wR 6
4

Temperature, C

System Thermal Efficiency, %

0.22

30
200

Time, mins

43
Fig. 16. Air temperature and relative humidity variation across the regenerator.

250

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

could have various impacts over the thermal efciency of the system as follows: lower ambient temperature may lead to higher
heat dissipation to the environment, thus causing to lower thermal
efciencies. Higher solar insolation leads to enhanced heat transfer
from PVs to the heat exchanger and this will consequently tend to
greater efciencies. Moreover, higher inlet temperatures of the
water results in lower efciencies due to reduced heat transfer
across the poly heat exchanger. Conned frame of PV modules also
enables to prevent higher heat dissipation to the ambient, so that
PV modules could be mentioned as a good heat source to the water
owing through polyethylene tubes and risers. The emissivity is
also lower and relatively higher thermal conductivity should also
be noted.

70

Temperature, C

60
50
T_reg _water (C)
T_cool _water (C)
T_desiccant_cold (C)

40
30
20
10
0
0

50

100

150

200

4.2. Liquid desiccant dehumidication unit

Time, mins
Fig. 17. Liquid desiccant and water temperature variation.

values are obtained as 99% and 4.03%, respectively. Fig. 11 also

compares the hourly variations of experimental and theoretical
values of the water temperature in the system. Despite higher theoretical water temperature values, the correlation coefcient and
root mean square percent deviation rates are attained as 98% and
7.63% respectively. These values clearly support the validity of
the model utilized in the present study.
Fig. 12 illustrates the degree of polyethylene heat exchanger
inuence over the electric power conversion efciency of the PV
modules, gpv. For with Polyethylene HE case, the increase of cell
efciency as a result of passive cooling off via water circulation
would lead to an increase on power conversion efciency, gpv.
On the other hand, higher cell temperature would cause a substantial decrease on the cell efciency, gpv. Nevertheless, it was found
that for both cases examined in this study, the cell efciency, gpv
is always better off with Poly HE case than without Poly HE case.
This implies that from the viewpoint of the rst law of thermodynamics, the with Poly HE case would be a better choice for PV
systems to enhance the overall energy output of PV panels.
Fig. 13 shows the variation of useful heat through circulating
water with a mass ow rate of 0.0493 kg/s. The useful heat generated by the polyethylene heat exchanger roof unit ranges between
2.23 kW and 4.33 kW for the given test day.
Fig. 14 illustrates the collector efciency as a function of the
ratio of the difference between the working uid inlet and ambient
air temperatures to the solar insolation incident on the collector
surface. The thermal efciency varies adversely with increasing
(Tw  Ta)/I ratio. The momentarily changing climate conditions

Wet-bulb effectiveness
Wet-bulb temperature, C

Wet-bulb effectiveness

0.54

24

0.53

23

0.52

22

0.51
21
0.50
20

0.49

19

0.48
0.47

24

25

26

27

28

29

30

18
31

Inlet air temperature,C

Fig. 18. Variation in wet-bulb effectiveness of the dehumidier.

Wet-bulb temperature,C

25

0.55

The liquid desiccant dehumidication system, as shown in

Fig. 7, was experimentally analysed in a series of tests to examine
the system performance under the given climate conditions. A
temperature control room was constructed to control inlet air temperature and humidity. 20 l potassium formate HCOOK solution
with a concentration of 75% and ow rate of 3 l/min was employed
and the inlet air mass ow rate was about 0.06 kg/s at around 30 C
where the relative humidity of the air was between 70 and 85%.
The reduction in the relative humidity (RH) was directly proportional to the increase in relative humidity of the inlet air. The average relative humidity at the outlet of the dehumidier was about
61.8%.
Fig. 15 shows the performance of the dehumidier unit under
the climate conditions provided. The reduction in the relative
humidity (RH) was directly proportional to the increase in relative
humidity of the inlet air. The average relative humidity at the outlet of the dehumidier was about 60.5%. Also, Fig. 16 demonstrates
the performance of the regenerator unit and the regeneration
water inlet ow rate was 0.074 kg/s. As can be seen in Fig. 16,
the average increase in the air temperature blowing across the
regenerator is around 6 C while the maximum change in the relative humidity is noted as 21%. Furthermore, Fig. 17 displays the
temperature variations of regeneration water, cooling water and
liquid desiccant solution during the operation. The average temperature of the water supplied to regenerate the dilute potassium
formate (HCOOK) desiccant solution was around 57.7 C where the
temperature of regenerated and cooled desiccant solution driven
to the dehumidier core and cooling water were 21.73 C and
12.41 C, respectively. It should be stressed that the wet-bulb
effectiveness is the main parameter to evaluate the performance
of the dehumidication core. Fig. 18 presents the dehumidier
wet-bulb effectiveness and change in wet-bulb temperature
accordingly. Based on the experimental results, the maximum
wet-bulb effectiveness was found to be at around 55% and the variation of wet-bulb temperature was found to be in the range of 18
24 C.
4.3. Indirect evaporative cooler
The indirect evaporative cooler unit employed in the experiment was experimentally tested in a series of tests under the given
environmental conditions (air temperature: 30 C, RH: 80%).
Fig. 18(a-b-c-d-e) shows the cooler performance at ve different
blower settings 300, 600, 900, 1200, and 1500 m3/h, respectively.
A FX-060 digital blower controller is set to maintain inlet air speed
with the given ow range. A 2 kW fan heater and 10 l capacity
humidier were employed to keep the control room temperature
at around 30 C and RH of 80%, respectively. The water supplied
to the cooler was about 15 C. Fig. 19(a) shows the variation of
air temperature across the cooler at the lowest fan speed of

251

35

35

30

30

Temperature, C

Temperature, C

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

25
20
15
10
5

Inlet Air Temperature

Supply Air Temperature
Exhaust Air Temperature

(a)

25
20
15
10
5

Inlet Air Temperature

Supply Air Temperature
Exhaust Air Temperature

(b)

0
0

10

15

20

25

30

10

15

35

35

30

30

25
20
15
10
5

Inlet Air Temperature

Supply Air Temperature
Exhaust Air Temperature

(c)

20

25

30

Time, mins

Temperature, C

Temperature, C

Time, mins

25
20
15
10
5

Inlet Air Temperature

Supply Air Temperature
Exhaust Air Temperature

(d)

0
0

10

15

20

25

30

10

15

Time, mins

20

25

30

Time, mins

35

Temperature, C

30
25
20
15
10
Inlet Air Temperature
Supply Air Temperature
Exhaust Air Temperature

(e)
0

10

15

20

25

30

Time, mins
Fig. 19. (a-b-c-d-e) Indirect evaporative cooler performance data: (a) warm air with 300 m3/h, (b) warm air with 600 m3/h, (c) warm air with 900 m3/h, (d) warm air with
1200 m3/h (e) warm air with 1500 m3/h.

300 m3/h. The inlet air temperature varies between 29.0 C and
31.23 C and the supply air temperature is 14.317.6 C with the
temperature difference of 14 C is recorded across the cooler. At
the fan speed of 600m3/h, the maximum exhaust air temperature
is recorded as 13.9 C as presented in Fig. 19(b). The minimum
and maximum supply air temperature attained across the cooler
is 15.5 C and 22.69 C, respectively. Also, Fig. 19(e) shows the temperature variations across the cooler at the maximum blower setting of 1500m3/h. the maximum supply air reaches to 23.2 C with
a maximum temperature decrease of 8.9 C during the cooling
operation.
Table 3 shows the average performance data obtained during
the test session. The measured supply air volumetric rate at ve
different blowers setting extend between 63.5 l/s and 202.3 l/s as
the minimum and maximum ow rates, respectively. One can

deduct from the summary data that the air temperature decrease
across the cooler is inversely proportional to the blower speed.
However, the cooling capacity shows an increasing trend from
1032 W at 300 m3/h to 1229 W at 1500 m3/h as the air ow rate
increases. Moreover, the wet-bulb effectiveness and dew point
effectiveness of the cooler at ve different fan speeds are presented
as wet-bulb effectiveness ranges between 0.96 and 0.67 and dew
point effectiveness is from 0.61 to 0.40, respectively. So the effectiveness of the cooling system decreases as the fan speed increases.
In addition, Fig. 20 presents the cooler COP calculated using the
experimental data recorded in accordance with each fan speed setting. The COP is calculated as the ratio of cooling capacity delivered
over the electrical power consumption of each fan and pump
employed in the unit. It is illustrated that COP of the unit and inlet
air ow rate vary in inverse proportion to one another. The

252

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

90

14
COP

Relative Humidity (%)

80

COP

12

10

70
60
50
40
RH_Dehumidifier Inlet (%)
RH_Cooler Inlet (%)
RH_Cooler Outlet (%)

30
20

4
300

600

900

1200

1500

100

Fig. 20. Indirect evaporative cooler COP at different fan settings.

Overall dehumidication and cooling performance of the system was investigated in a series of tests by combining the liquid
desiccant dehumidication unit with the indirect evaporative
cooler. Hot and humid air introduced from the temperature and
humidity control room is rst dehumidied by the liquid desiccant
based dehumidier, and the processed air is, then, delivered to the
indirect evaporative cooling unit, also called dew point cooler, to
further cooling and adjusting the air quality. HMT130 humidity
and temperature probes were used for data recording regarding
air temperature and humidity at the both inlet and outlet of the
dehumidier and evaporative cooling units. Fig. 21 introduces
the air temperature variation across the dehumidier and indirect
evaporative cooler. The temperature of the air inlet to the dehumidier is ranged from 27.5 C to 34.5 C and the average temperature drop across the dehumidier and cooler units is around
4.5 C and 7.5 C, respectively. Also, average air temperature at
the outlet of the cooler is about 19.45 C. One can conclude that
substantial ratio of temperature drop (nearly 2:3) takes place
across the cooler. Moreover, Fig. 22 presents the change in relative
humidity across the dehumidier and cooler units. The process air
has the mass ow rate of 0.099 kg/s and liquid desiccant ow rate
is around 3 l/min. According to the experimental data, the range of

Temperature (C)

40

30

20

10
T_Dehumidifier Inlet (C)
T_Cooler Inlet (C)
T_Cooler Outlet (C)

200

300

400

500

400

500

Time (min)
Fig. 21. Variation of air temperature across the combined dehumidication-cooling
unit.

7000
6000

Cooling Capacity (W)

4.4. Combined dehumidication and indirect evaporative cooling

system

100

300

Fig. 22. Variation of relative humidity across the combined dehumidicationcooling unit.

maximum COP of 12.1 is noted at the minimum blowing rate of

300 m3/h and the minimum COP of 6.3 at 1500 m3/h.

200

Time (min)

Fan Speed (m/h)

5000
4000
Indirect Evaporative Cooler cooling capacity

3000

Overall cooling capacity

2000
1000
0
0

100

200

300

400

500

Time, mins
Fig. 23. Cooling capacity of indirect evaporative cooler and combined unit.

air relative humidity at the dehumidier inlet was between 78.1%

and 83.5% and average drop in relative humidity across the unit
was found to be 25.7%. At inlet and outlet of the cooler, the average
relative humidity ratios were noted as 55.78% and 60.44%,
respectively.
Fig. 23 illustrates the comparison and variation of cooling
capacities
both
for
only
cooler
and
combined
dehumidication-cooler unit. The maximum cooling capacities
delivered by only cooler and the combined unit were 881 W and
6673 W, respectively. The experimental data proves that the main
contribution to the overall cooling capacity belongs to the liquid
desiccant based dehumidication unit during the operation of
the combined dehumidication-cooling system.
Table 4 summarizes the overall performance data of the building integrated PV/T roof collector combined with a liquid desiccant
enhanced indirect evaporative cooling system. Utilizing the renewable energy based-solar thermal source, the roof unit with an average water temperature of 30.25 C and mass ow rate of 6 l/min
from the roof unit, supplies the part of the regeneration heat input
around 3 kW required to adjust the concentration of the dilute liquid desiccant solution. The remaining heat input required for the
regeneration is calculated to be 4 kW where the total heat input
needed in the regenerator is estimated to be 7 kW for the average
regeneration temperature of 6065 C. In addition, the air temperature decreases from 30.99 C to 26.45 C at the dehumidier unit
and further to 19.45 C at the indirect evaporative cooler. As for the
relative humidity variation, it is reduced from 80.39% to 55.78%

253

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254
Table 4
Experimental performance data of the overall system.
Parameter

Value

Parameter

Value

Poly HE roof unit water supply temperature (C)

Poly HE roof unit water mass ow rate (kg/s)
Poly HE useful heat gain (W)
PV efciency with Poly HE (%)
PV efciency w/o Poly HE (%)
Regenerating water temperature (C)
Regenerating heat input (W)
Regenerating water mass ow rate (kg/s)
Liquid desiccant ow rate (kg/s)
Dehumidier inlet air mass ow rate (kg/s)

30.25
0.074
2982.7
12.4
11.2
57
5845
0.074
0.037
0.099

Dehumidier inlet air temperature (C)

Dehumidier inlet air relative humidity (%)
Ind. Evp. Cooler inlet air temperature (C)
Ind. Evp. Cooler inlet air relative humidity (%)
Ind. Evp. Cooler outlet air temperature (C)
Ind. Evp. Cooler outlet air relative humidity (%)
Ind. Evp. Cooler cooling capacity (W)
Overall cooling capacity (W)
Combined dehumidier-cooler thermal COP
Combined dehumidier-cooler electrical COP

30.99
80.39
26.45
55.78
19.45
60.44
780
5135
0.73
6.75

across the dehumidier where supply air relative humidity drops

to 60.44% through dew point cooling. The cooling capacity of the
cooler is about 881 W where the overall cooling capacity is found
to be 6673 W for the liquid desiccant enhanced indirect evaporative cooling system. Also, thermal and electrical COPs of the combined cooling system are estimated to be 0.73 and 6.75,
respectively. A Gauss propagation based uncertainty analysis was
performed to estimate the uncertainty in calculation of the parameters. The maximum uncertainties inuencing the efciency of the
proposed system were estimated to be 2.47% for the roof unit,
3.75% for the cooling part, respectively.

that theoretical model makes a fair agreement with the experimental results obtained. In addition, the proposed tri-generation
system is capable of providing about 3 kW heating power, cooling
capacity of 5.2 kW where average power generation of the PV system was found to be 10.3 MW h/year. The electrical and thermal
COP of the combined dehumidication-cooling system was calculated as 0.73 and 6.75, respectively. With the satisfactory experimental results obtained, one can conclude that the proposed
system could meet the cooling needs and improve indoor air quality for occupants.
Acknowledgements

5. Conclusion
In this study, the integration of innovative building integrated
solar thermal collector with a pilot scale combined cooling system
was introduced and experimentally investigated. A low cost polyethylene heat exchanger loop unit underneath the PV modules
was built and tested utilizing solar energy as a heat source and
employing water as an environmentally friendly working uid.
Also, a liquid desiccant enhanced indirect evaporative cooling system was developed and tested experimentally including a liquid
desiccant based dehumidier and a dew point evaporative cooler
with a cross-ow core. The environmentally friendly, less corrosive, lower cost, lower density and viscosity potassium formate
(HCOOK) solution was used as a liquid desiccant also considering
its thermodynamic and physical benets comparing to its conventional counter-parts. Heat input to regenerate the dilute desiccant
solution was partly harnessed from the roof loop and the remaining heat demand was met by employing a water immersion heater
with rod thermostat. The combined experimental prototype was
operated under the given operational conditions to estimate the
ability of such a combined small scale system for building applications. The concluding remarks are outlined as follows.
Based on the experimental data recorded, water temperature
owing through the tubes and risers of polyethylene heat exchanger loop could reach up to 35.5 C subject to outside parameters,
mostly affected by air temperature and global solar radiation.
Thus, experimental ndings prove that polyethylene heat exchanger loop can function efciently as a heat extraction component for
solar assisted heating and cooling systems. Overall power efciency data show that PV modules can improve power energy performance by 10.7 % due to achieved collector cooling as the cold
water ow forms passive cooling effect and partially removes the
waste heat from PV modules. Hence improved performance of
the photovoltaic systems help to mitigate the energy supply issues
and contribute to the CO2 reduction towards a sustainable environment. Else, waste heat removal away from the PV panels is likely to
enhance life cycle of the solar collectors since high operating temperatures may shorten the life-span. The validation of the thermal
model for the roof part was performed by comparing the experimental results with the estimated data. The comparison shows

The research described in this paper was conducted at the

University Of Nottingham, and supported nancially and in-kind
by Geo Green Power Ltd and the Sustainable Construction iNet,
Project Number: CR&D 106154.
References
[1] Otanicar Todd, Taylor Robert A, Phelan Patrick E. Prospects for solar coolingan
economic and environmental assessment. Sol Energy 2012;86(5):128799.
[2] Jradi M, Riffat S. Experimental investigation of a biomass-fuelled micro-scale
tri-generation system with an organic Rankine cycle and liquid desiccant
cooling unit. Energy 2014;71:8093.
[3] Huang BJ et al. System performance and economic analysis of solar-assisted
cooling/heating system. Sol Energy 2011;85(11):280210.
[4] Bermejo Pablo, Pino Francisco Javier, Rosa Felipe. Solar absorption cooling
plant in Seville. Sol Energy 2010;84(8):150312.
[5] Ge TS et al. Performance investigation on a novel two-stage solar driven rotary
desiccant cooling system using composite desiccant materials. Sol Energy
2010;84(2):1579.
[6] Bilgili Mehmet. Hourly simulation and performance of solar electric-vapor
compression refrigeration system. Sol Energy 2011;85(11):272031.
[7] Eicker Ursula, Dalibard Antoine. Photovoltaicthermal collectors for night
radiative cooling of buildings. Sol Energy 2011;85(7):132235.
[8] Katejanekarn Thosapon, Chirarattananon Surapong, Kumar S. An experimental
study of a solar-regenerated liquid desiccant ventilation pre-conditioning
system. Sol Energy 2009;83(6):92033.
[9] Li Yutong, Lin Lu, Yang Hongxing. Energy and economic performance analysis
of an open cycle solar desiccant dehumidication air-conditioning system for
application in Hong Kong. Sol Energy 2010;84(12):208595.
[10] Bourdoukan P, Wurtz E, Joubert Patrice. Experimental investigation of a solar
desiccant cooling installation. Sol Energy 2009;83(11):205973.
[11] Qi Ronghui, Lin Lu, Huang Yu. Energy performance of solar-assisted liquid
desiccant air-conditioning system for commercial building in main climate
zones. Energy Convers Manage 2014;88:74957.
[12] Fong KF et al. Solar hybrid cooling system for high-tech ofces in subtropical
climateRadiant cooling by absorption refrigeration and desiccant
dehumidication. Energy Convers Manage 2011;52(8):288394.
[13] Aly A Ayman, Zeidan B El-Shafei, Hamed M Ahmed. Solar-powered open
absorption cycle modeling with two desiccant solutions. Energy Convers
Manage 2011;52(7):276876.
[14] Zeidan El-Shafei B, Aly Ayman A, Hamed Ahmed M. Modeling and simulation
of solar-powered liquid desiccant regenerator for open absorption cooling
cycle. Sol Energy 2011;85(11):297786.
[15] Agrawal Basant, Tiwari GN. Optimizing the energy and exergy of building
integrated photovoltaic thermal (BIPVT) systems under cold climatic
conditions. Appl Energy 2010;87(2):41726.
[16] Zhang Xingxing et al. Review of R&D progress and practical application of the
solar photovoltaic/thermal (PV/T) technologies. Renew Sustain Energy Rev
2012;16(1):599617.

254

M.S. Buker et al. / Energy Conversion and Management 101 (2015) 239254

[17] Lowenstein, Andrew, Steve Slayzak, Eric Kozubal. A zero carryover liquiddesiccant air conditioner for solar applications. In: ASME 2006 International
Solar Energy Conference. American Society of Mechanical Engineers, 2006.
[18] Jain Sanjeev, Tripathi Sagun, Das Rajat Subhra. Experimental performance of a
liquid desiccant dehumidication system under tropical climates. Energy
Convers Manage 2011;52(6):24616.
[19] Fumo Nelson, Goswami DY. Study of an aqueous lithium chloride desiccant
system: air dehumidication and desiccant regeneration. Sol Energy
2002;72(4):35161.
[20] Bassuoni MM. An experimental study of structured packing dehumidier/
regenerator operating with liquid desiccant. Energy 2011;36(5):262838.
[21] Daou Kadoma, Wang RZ, Xia ZZ. Desiccant cooling air conditioning: a review.
Renew Sustain Energy Rev 2006;10(2):5577.
[22] Lowenstein Andrew. Review of liquid desiccant technology for HVAC
applications. HVAC&R Res 2008;14(6):81939.
[23] Oliveira Armando C et al. Thermal performance of a novel air conditioning
system using a liquid desiccant. Appl Therm Eng 2000;20 13:121323.
[24] Isetti Carlo, Nannei Enrico, Magrini Anna. On the application of a membrane
airliquid
contactor
for
air
dehumidication.
Energy
Build
1997;25(3):18593.
[25] Riffat SB, James SE, Wong CW. Experimental analysis of the absorption and
desorption rates of HCOOK/H2O and LiBr/H2O. Int J Energy Res
1998;22(12):1099103.
[26] Riangvilaikul B, Kumar S. An experimental study of a novel dew point
evaporative cooling system. Energy Build 2010;42(5):63744.
[27] Goldsworthy Mark, White Stephen. Optimisation of a desiccant cooling system
design with indirect evaporative cooler. Int J Refrig 2011;34(1):14858.

[28] Zhan Changhong et al. Comparative study of the performance of the M-cycle
counter-ow and cross-ow heat exchangers for indirect evaporative cooling
Paving the path toward sustainable cooling of buildings. Energy
2011;36(12):6790805.
[29] Qiu Guoquan, Liu Hao, Riffat Saffa B. Experimental investigation of a liquid
desiccant cooling system driven by ue gas waste heat of a biomass boiler. Int J
Low-Carbon Technol 2013;8(3):16572.
[30] Caliskan Hakan, Dincer Ibrahim, Hepbasli Arif. A comparative study on
energetic, exergetic and environmental performance assessments of novel
M-Cycle based air coolers for buildings. Energy Convers Manage 2012;56:6979.
[31] Woods Jason, Kozubal Eric. A desiccant-enhanced evaporative air conditioner:
numerical model and experiments. Energy Convers Manage 2013;65:20820.
[32] Buker Mahmut Sami, Riffat Saffa B. Recent developments in solar assisted
liquid desiccant evaporative cooling technologya review. Energy Build
2015;96:95108.
[33] Tiwari Arvind, Sodha MS. Performance evaluation of solar PV/T system: an
experimental validation. Sol Energy 2006;80(7):7519.
[34] Chong WT et al. Techno-economic analysis of a windsolar hybrid renewable
energy system with rainwater collection feature for urban high-rise
application. Appl Energy 2011;88(11):406777.
[35] Zhao Xudong et al. Theoretical study of the performance of a novel PV/e roof
module for heat pump operation. Energy Convers Manage 2011;52(1):60314.
[36] Dufe John A, Beckman William A. Solar engineering of thermal processes, vol.
3. New York etc.: Wiley; 1980.
[37] Buker Mahmut Sami, Mempouo Blaise, Riffat Saffa B. Performance evaluation
and techno-economic analysis of a novel building integrated PV/T roof
collector: an experimental validation. Energy Build 2014;76:16475.