Post on 26-Dec-2015
nline at www.sciencedirect.com
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9
Available o
www. i ifi i r .org
journal homepage: www.elsevier .com/locate/ i j refr ig
Preliminary experimental characterization of a three-phaseabsorption heat pump
A. Rosato*, S. Sibilio
Seconda Universita degli Studi di Napoli, Dipartimento di Architettura, via San Lorenzo, 81031 Aversa, Italy
a r t i c l e i n f o
Article history:
Received 5 August 2012
Received in revised form
23 October 2012
Accepted 14 November 2012
Available online 23 November 2012
Keywords:
Absorption cycle
Thermally driven chiller
Chemical heat pump
Lithium chloride
Solar cooling
Trigeneration
* Corresponding author. Tel./fax: þ39 081 81E-mail address: antonio.rosato@unina2.it
0140-7007/$ e see front matter ª 2012 Elsevhttp://dx.doi.org/10.1016/j.ijrefrig.2012.11.015
a b s t r a c t
In this paper a recently commercialized three-phase absorption heat pump that is capable
of storing energy internally in the form of crystallized salt (LiCl) with water as refrigerant
has been experimentally investigated during summer period. The tests have been per-
formed with the aim to investigate the operation logic of the machine and to highlight both
the reliability and the efficiency of the system over an operating conditions range of great
practical interest.
The measured performance have been compared with those of a conventional elec-
trically driven vapor compression refrigerating system from an energy, environmental and
economic point of view in order to assess the suitability of the absorption heat pump: this
comparison showed that the absorption system is potentially able to guarantee an energy
saving, a reduction of carbon dioxide emissions and a lower operating cost only in case of
the most part (at least 70%) of required thermal energy is supplied by solar collectors.
ª 2012 Elsevier Ltd and IIR. All rights reserved.
Caracterisation experimentale preliminaire d’une pompe achaleur a trois phases
Mots cles : cycle a absorption ; refroidisseur a entraınement thermique ; pompe a chaleur chimique ; chlorure de lithium ; refroidissement
solaire ; trigeneration
1. Introduction
The worldwide cooling demand has drastically increased over
the last few years. This has led to the installation of a large
number of electrically driven air conditioning systems
(Balaras et al., 2007; Henning, 2007) with a dramatic rise in
electricity consumption, which is nowadaysmostly generated
from fossil fuels. This trend has caused important
22530.(A. Rosato).ier Ltd and IIR. All rights
environmental problems such as ozone layer depletion and
global warming.
In this context, there is a clear need to develop more
sustainable technologies in order to minimize the environ-
mental impact of cooling applications. Absorption heat
pumps have emerged as a promising alternative to conven-
tional vapor compression cycles (Fiskum et al., 1996; Florides
et al., 2002; McMullan, 2002; Wang et al., 2011), since they
reserved.
Nomenclature
Latin letters
B natural gas-fired boiler
c specific heat (kJ kg�1 K�1)
C operating cost (V)
CO2 carbon dioxide equivalent emission (kg CO2)
COP coefficient of performance
CW10 ClimateWell10
CWIC2 CW10 internal software
CUng natural gas Unit Cost (V Nm�3)
CUel electric energy Unit Cost (V kWh�1)
E energy (kJ)
EDC electrically driven chiller system
EFB fraction of Eth,TDC produced by natural gas-fired
boiler
FC fan-coil
IHE internal heat exchanger
HD heat dissipator
HWS hot water storage
LHV lower heating value (kWh Nm�3)
M water mass flow meter
MCHP micro combined heat and power generation
MG natural gas volumetric flow meter
P power (kW)/pump
PER primary energy ratio (%)
PES primary energy saving (%)
PHE plate heat exchanger
R electric resistance
SUN Second University of Naples
T temperature/resistance thermometer
TC0 temperature of water going towards heat
dissipator before by-pass valve (�C)TDC thermally driven chiller_V volumetric flow rate (m3 s�1)
Greeks
a CO2 emission factor for electric energy
(kgCO2 kWh�1)
b CO2 emission factor for primary energy
(kgCO2 kWh�1)
D difference (%)
h efficiency
r density (kg m�3)
Subscripts
B boiler
cool cooling
el electric
FC fan-coil
HD heat dissipator
in inlet
IHE internal heat exchanger
MCHP micro combined heat and power generation
ng natural gas
out outlet
th thermal
TDC thermally driven chiller
w water
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9718
can use low grade energy sources that are environmentally
friendlier instead of electricity. Several scientific papers
studied the integration of different types of commercially
available absorption systemswith cogeneration units by using
the surplus of heat coming from the cogeneration device
during the warm season for activating the absorption cycle
and providing a combination of electric, heat and cooling
energy (Angrisani et al., 2012; Chicco and Mancarella, 2009;
Hernandez-Santoyo and Sanchez-Cifuentes, 2003; Serra et al.,
2009). In comparison to the traditional units based on separate
energy production, these plants (called trigeneration systems)
showed a significant potential in terms of energy savings and
reduction of CO2 emissions (Huicochea et al., 2011; Kavvadias
et al., 2010; Li et al., 2006; Lin et al., 2007).
There are several technologies of thermally activated
chillers commercially available today, e.g. standard absorp-
tion system using LiBr/water or NH3/water and salt-water
absorption chiller (Srikhirin et al., 2001) and/or chemical
heat pump (Wongsuan et al., 2001). Chemical heat pump is
a new and promising technologywhich is capable of operating
with low temperature heat sources: salt-water solutions such
as lithium chloride (LiCl)/water, sodium sulphite (Na2S)/water,
and calcium chlorides (CaCl2)/water, etc. have been used (Boer
et al., 2002; Conde, 2004; Ogura et al., 2003). Absorption chillers
aremore common atmediumor larger scale, while small scale
units are in process of becoming commercial.
In this paper a recently commercialized chemical heat
pump using LiCl/water as a working fluid pair has been
experimentally investigated. It is a three-phase absorption
system that is capable of storing energy internally in the form
of crystallized salt (LiCl) with water as refrigerant; the triple-
state process, so called because it uses solid, liquid and
vapor at the same time, makes this thermally driven chiller
(TDC) particularly different from other chemical heat pumps
or standard absorption processes (which use liquid and vapor
phases).
The unit was patented in 2000 (Olsson et al., 2000) and it
has been developed by the Swedish company ClimateWell�
via five generations of prototypes. The 4th generation of
machines, that was the first to be sold commercially as from
2007 under the name CW10, is installed at the laboratory of
Second University of Naples (Fig. 1). It consists of two identical
units, so called barrels, that work together. Each barrel
consists of four different vessels: the reactor (absorber/
generator), the condenser/evaporator, the solution vessel and
the refrigerant vessel. The reactor and condenser/evaporator
are the active parts of the unit with a vapor channel between
them, while the two other vessels are stores for salt solution
and the refrigerant; the unit is operated as a closed system
under vacuum conditions and there are heat exchangers in
the reactor and condenser/evaporator; solution and refrig-
erant are pumped from the storage vessels over these heat
exchangers and then flow under gravity back to the storage
vessels (Bales and Ayadi, 2009).
The machine is connected to three external circuits: the
thermal supply, the heat sink and the cooling supply. The
Fig. 1 e Schematic of complete CW10machine on the left (Udomsri and Bales, 2011) and single barrel on the right (Bales and
Ayadi, 2009).
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 719
process occurring in each barrel works in batch mode, with
a separate desorption (charge) phase followed by absorption
(discharge) phase:
U during the charging phase the reactor is connected to the
thermal supply, while the condenser/evaporator is con-
nected to the heat sink; the solution is heated by the
thermal source via the heat exchanger in the reactor
becoming steadily more concentrated, and when it reaches
saturation point further desorption can result in the
formation of solid crystals that fall under gravity into the
vessel. These then get transferred to the storage vessel.
Here they are prevented from following the solution into
the pump by a sieve, thus forming a form of slurry in the
bottom of the vessel; at the same time water is evaporated
and steam is released to the condenser/evaporator;
U during the discharging phase the reactor is connected to
the heat sink, while the condenser/evaporator is connected
to the cooling supply circuit; the saturated solution is
pumped over the heat exchanger in the reactor where it
absorbs the refrigerant evaporated in the condenser/evap-
orator. The solution becomes unsaturated in the reactor,
but when it goes to the solution store it has to pass through
the slurry of crystals, where some of the crystals are dis-
solved to make the solution fully saturated again. In this
way the solution is kept saturated as long as there are
crystals available and the net result is a dissolving of the
crystals into saturated solution.
Since the energy is stored in a chemical form, no energy
shouldbe lost to the surroundings;whena barrel is charged, the
energy stays stored in the barrel until there is a coolingdemand.
A by-pass valve is installed in the machine for regulating
thewater temperature going towards the cooling supply to the
set value: by-pass valve position can vary between 100% (by-
pass valve fully open) and 0% (by-pass valve fully closed).
A plumbing unit switches the flows between the external
circuits and the relevant heat exchangers in the two barrels.
The machine has its own control system that makes all the
“swaps” of the machine which changes the state from
charging to discharging and vice versa. The control system
also sends signals to the plumbing unit to control all the
valves in order to change the circuit connections and it guar-
antees that the machine works automatically and
independently.
The unit can be operated so that one barrel is charged
while the other one is discharged: this gives quasi-continuous
operation, but when the units are swapped at the end of
charge/discharge, there is a period without cooling supply.
More generally, the CW10 unit can be operated in seven
different modes: “manual”, “normal”, “full cycles”, “double”,
“timer”, “turbo” and “test”. In this paper the performance of
the system have been experimentally investigated during
both “normal mode” and “double mode” operation. “Normal
mode” is the default and the fully automatic mode, where the
barrels alternate in charging and discharging: during this
operation mode the machine is always able to both provide
cooling energy and use the supplied thermal energy. In
“doublemode”, both barrels are charged and discharged at the
same time. This should result in higher cooling/heating power
when discharging and higher charging power when charging;
however, running in this mode the discharging delivery and
the charging power is not continuous.
The machine control system recognizes when a swap
should take place. It then sends signals to the plumbing unit
which automatically makes all the necessary connections. A
swap is performed when one of the following conditions is
verified:
1) charging barrel: level reaches 100% will trigger a swap
independent of discharging barrel status;
2) charging barrel: level reaches above 80% in combination
with condition 3 or condition 4;
3) discharging barrel: level has reached below 40% and TC0/
Tw,TDC,in is higher than 0.67 and TC0 is higher than 15 �C in
combination with condition 2;
4) discharging barrel: level has been 3% or less for 15 min in
combination with condition 2;
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9720
where TC0 is the temperature of internal water going towards
the heat dissipator before the by-pass valve and Tw,TDC,in is the
temperature of water coming from the heat source before
entering the machine. The level of each barrel is determined
by measuring the weight of the water in the barrels.
The same TDC model installed at the laboratory of Second
University of Naples (SUN) has been already investigated by
Udomsri et al. (2011, 2012), Bales and Nordlander (2005), Bales
and Ayadi (2009). Udomsri et al. (2011) presented the moni-
toring results of CW10 driven by district heat from a network
supplied by a centralised combined heat and power fired with
municipal waste; they investigated the systemduring “normal
mode” operation and found amaximum thermal coefficient of
performance during the hottest period of around 0.50;
however, the figure was only 0.41 for the completemonitoring
period during the summer of 2008. According to themonitored
results obtained from the demonstration, a system simulation
model for the TRNSYS environment has been calibrated by
Udomsri et al. (2012) and used to find improved system design
and control. Bales and Nordlander (2005) carried out just few
of the planned experiments on CW10model during “full cycle”
operation due to lack of time before the machines were
shipped. Of these,most hadmissing data due to an error in the
logger program that limited the duration of saved data,
resulting in an even smaller amount of recorded results; due
to these problems, no direct calculations of the thermal
coefficient of performance was possible. They tested also the
TDC model DB220 produced by ClimateWell�, a TDC model
less recent than CW10. According to the available measure-
ments obtained for CW10 model, Bales and Ayadi (2009)
developed a grey box simulation model for the TRNSYS envi-
ronment; the TDC unit model was verified against the
measured data and showed reasonable agreement, but the
authors stated that more data would be needed be needed to
make sure the parameters are correct and to verify them
properly. The model was also used for parametric studies in
order to determine the effect of boundary conditions on the
thermal coefficient of performance.
Even if some data have been already available in literature,
the CW10 unit has not been yet investigated during “double
Fig. 2 e Experim
mode” operation, and the experimental results regarding
“normal mode” operation are still quite limited. For these
reasons in this paper the performances of CW10 model have
been experimentally investigated during both “normal mode”
and “double mode” operation in order to better highlight the
system operation and performance. In the following the
experimental set-up and the results gathered during the
experiments (thermalpower supplied, coolingpowerdelivered,
coefficient of performance, temperature levels, etc.) will be
presentedandanalyzed indetail. Inaddition themeasureddata
have been used to compare the performance of the experi-
mentally investigated thermally driven chiller with those of
a conventional vapor compression refrigerating system from
an energy, environmental and economic point of view in order
to verify the suitability of CW10 model. The measurements
reported in the following canbe alsoused to verify theaccuracy
of the recently developed TRNSYS simulation model (Udomsri
et al., 2011, 2012; Bales and Nordlander, 2005; Bales and Ayadi,
2009) in order to carry out a techno-economic analysis for
studying and evaluating the viability of trigeneration plants
using the TDC model investigated in this paper.
2. Experimental set-up
A schematic view of the test apparatus of the Built Environ-
ment Control Laboratory of Second University of Naples
(SUN), detailing instrumentation components, is shown in
Fig. 2. The experimental set-up is located in Frignano,
a municipality in the Province of Caserta (around 20 km far
from Naples).
As stated above, the TDC unit experimentally investigated
in this paper is the 4th generation of a chemical heat pump
(model CW10), patented in 2000 (Olsson et al., 2000) and sold
by the Swedish company ClimateWell�. The machine has
been described in detail in the previous section.
As can be derived from Fig. 2, the unit is supplied by the
thermal power recovered from a micro-cogenerator based on
a natural gas fuelled reciprocating internal combustion engine
(commercialized by AISIN-SEIKI company) and stored in 1000 L
ental set-up.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 721
tank; taking into consideration that the TDC needs a charging
temperature at least 50 �C larger than theheat sink temperature
to start the absorption process and that the temperature of hot
waterflowing fromtheMCHP system into the storage cannot be
higher than around 70 �C, a natural gas-fired boiler (B) with
a rated thermal output of 32 kW has been installed before the
inlet of the TDC system. The hot water storage (HWS) is insu-
latedwith50mmflexiblepolyurethane layerand is furthermore
equipped with an auxiliary 4.0 kW electric resistance (R) fed by
the micro-cogenerator. A water to air heat exchanger with
a ratedpower equal to 30 kW is installed as heat sink. Thewater
cooled by the TDC is pumped towards a fan-coil with a rated
total cooling capacity equal to 10.95 kW devoted to satisfy the
cooling load of a part of the entire laboratory.
Variable speed wet rotor pumps (P1, P2, P3 and P4) have
been installed in order to circulate the water within the
experimental plant; three different pump revolution steps can
be manually set for each pump with a maximum mass flow
rate equal to 22.8 l min�1 for pump P3 and to 14.4 l min�1 for
the other pumps.
The experimental plant is well instrumented (Fig. 2) in
order to measure the following parameters:
� water temperature in the key-points of the plant (at the inlet
and outlet of TDC, FC, HD, B, HWS, MCHP);
� ambient temperature;
� water volumetric flow rate in the key-points of the plant
(flow rate entering TDC, FC, HD, B, HWS, MCHP);
� natural gas volumetric flow rate entering both micro-
cogenerator and natural gas-fired boiler;
� electricpower suppliedbymicro-cogenerator to the end-user.
Water and ambient temperatures are measured by using
resistance thermometers Pt100; water mass flow rate is ob-
tained by using an ultrasonic mass flow sensor, while
a thermal mass flow meter is installed to evaluate the natural
gas volumetric flow rate; three wattmeters measure the
electric flows entering and exiting the unit. Two resistance
thermometers are used for measuring the hot water temper-
ature within the tank. Table 1 summarizes the main charac-
teristics of the plant instrumentation.
Table 1 e Main characteristics of the plantinstrumentation.
Parameter Instrument Operatingrange
Accuracy
T Resistance
thermometer
Pt100
�50 O 100 �C �0.2 �C
_Vw Ultrasonic
volumetric
flow meter
0 O 50 l min�1 �2.5% of
full scale
_Vng Thermal
volumetric
flow meter
0 O 5.0 Nm3 h�1 �0.8% of
reading
�0.2% of
full scale
Pel,MCHP Wattmeter 0 O 6 kW 0.2% of
full scale
0 O 10 kW
The TDC installed at SUN lab is equipped with an internal
software (named CWIC2) by means of which several operation
systemparameters canbemonitoredand recorded: inparticular,
the machine internal software provides the values of some
parameters that cannot be directly derived by using our instru-
mentation, i.e. thewater temperatureTC0, the levelofeachbarrel
during system operation, the by-pass valve position, etc.
Based on the direct measurements, the parameters listed
beloware calculated inorder to evaluate theplant performances:
Pth;TDC ¼ _Vw;TDC$rw$cw$�Tw;TDC;in � Tw;TDC;out
�(1)
Pth;HD ¼ _Vw;HD$rw$cw$�Tw;HD;in � Tw;HD;out
�(2)
Pcool;FC ¼ _Vw;FC$rw$cw$�Tw;FC;out � Tw;FC;in
�(3)
where the water specific heat and the water density, respec-
tively, have been assumed equal to cw ¼ 4.187 kJ (kg K)�1 and
rw ¼ 990 kg m�3.
The signals coming from the resistance thermometers
Pt100 are acquired by three cFP-RTD-124 analog inputmodules
(producedbyNational Instruments�),while the signals coming
from the other sensors are managed by two cFP-AI-110 analog
input modules (produced by National Instruments�). Each
acquisition device is a 16-bit resolution system with eight
current outputs (4O 20 mA). The digital data coming from the
modules are sent to a personal computer. The software Lab-
View 8.2 is used to define the acquisition frequency and to
monitor and/or record all the directlymeasured and calculated
parameters. The experimental data presented in the following
sections have been recorded every 10 s.
Additional details regarding the above-presented experi-
mental plant can be found in Rosato and Sibilio (2012) and
Angrisani et al. (2012).
3. Experimental results
In the following the operating conditions and themain results
gathered during the experiments are highlighted and deeply
analyzed. The data are presented separately for “normal” and
“double” mode operation.
Given the constraints of the experimental set-up, the
experiments have not been conducted over the entire range of
machine operation; however the achieved results allows to get
useful information on the system performance in relation to
a range of operating conditions of great interest in the practice
not yet fully exploited experimentally.
During both the tests in “normalmode” and “doublemode”
the set value of water temperature going towards the cooling
supply was 13 �C.In the last section the measured data are used to compare
the performance of CW10 unit with those of a conventional
electrically driven vapor compression refrigerating system
from an energy, environmental and economic point of view.
3.1. Normal mode operation
The test in “normal mode” has been performed the 19th
October 2011 from 11:01 until 17:44. In Figs. 3 and 4 the
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
11:01 11:30 11:59 12:28 12:57 13:25 13:54 14:23 14:52 15:21 15:49 16:18 16:47 17:16
Tem
pera
ture
(°C
)
Time (hh:mm)
Text Tw,TDC,in Tw,TDC,out Tw,FC,in Tw,FC,out Tw,HD,in Tw,HD,out
Fig. 3 e Water temperature values measured during “normal mode” operation.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9722
operating conditions related to the experiment performed in
“normal mode” are reported as a function of the time: in Fig. 3
the water temperature values measured in the key-points of
the plant are depicted, while Fig. 4 shows the level of both
barrels and thewater volumetric flow rates flowing through the
thermally driven chiller, the heat dissipator and the fan-coil.
Fig. 5 depicts the thermal power supplied to the TDC system
(Pth,TDC), thecoolingpowerproducedbytheTDCsystem(Pcool,FC)
and the thermal Coefficient Of Performance (COPth,TDC) as
a functionof the timeduring“normalmode”operation.Pth,TDC is
calculated by using Eq. (1), while Eq. (3) provides Pcool,FC; the
instantaneous values of COPth,TDC are defined as follows:
COPth;TDC ¼ Pcool;FC=Pth;TDC (4)
As can be derived from Figs. 3e5, TDC operation in “normal
mode” starts around 11:01 with Barrel B charging and Barrel A
discharging; at around 13:00 the barrels are “swapped”, so that
Barrel A charging starts and cooling is provided thanks to the
Barrel B discharging; two additional “swaps” are performed
around 14:30 and 16:00, respectively: as a consequence, both
Barrel A and Barrel B have been charged and discharged two
times through the experiment. Each swap between barrels is
due to the fact that charging barrel level reaches 80% and
discharging barrel level has reached 40% with both the ratio
TC0/Tw,TDC,in higher than 0.67 and TC0 values larger than 15 �C.
Fig. 3 shows that the water temperature coming from the
boiler towards the TDC (Tw,B,out) is around 81.5 �C (quite lower
than that one suggested by the manufacturer for the TDC, i.e.
85e120 �C) and the temperature drop across the machine is
about 5e10 �C; the temperature of the water coming from the
HD towards the machine during charging/discharging periods
(Tw,HD,out) oscillates between around 26 and 30 �C. Water
temperature coming from the TDC towards the fan-coil
(Tw,FC,in) is around 15 �C, with a minimum value of 12.6 �Cachieved around 11:10.
Except during the “swap” between the barrels, the volu-
metric flow rate through both the thermally driven cooling
system ( _Vw;TDC) and fan-coil ( _Vw;FC) is 14.4 l min�1 (15.0 l min�1
is suggested as minimum water flow rate by ClimateWell�),
while 22.8 l min�1 is the water flow ( _Vw;HD) pumped towards
the HD (Fig. 4).
As can be derived from Fig. 5, during the “swap” between
the two barrels the TDC cannot deliver cooling; during the
charging/discharging periods, cooling capacity increases till
reaching a maximum and then slightly reduces: maximum
value of cooling power gathered during the test is about
3.5 kW. The measured data agree well with those reported by
the manufacturer that suggests about 3.0 kW as cooling
capacity in case of Tw,TDC,in ¼ 80 �C, Tw,HD,out ¼ 30 �C and
Tw,FC,out ¼ 20 �C. During Barrel A discharging the values of
Pcool,FC are slightly higher than the those achieved during
Barrel B discharge. So the plot shows that the two units
worked differently, with Barrel B performing poorer than the
other: also Bales and Nordlander (2005) found a different
performance between two barrels by experimenting the
model DB220.
Fig. 5 shows that the COPth,TDC (defined by Eq. (4)) is not
constant: it increases during discharging phase till reaching
a maximum value and then becomes zero when “swap”
period starts; the maximum value of COPth,TDC measured
during Barrel A discharging is around 0.6, quite higher than
the greater value of COPth,TDC achieved during Barrel B
discharging.
The cumulative cooling energy supplied by the TDC system
(Ecool,FC) and the cumulative thermal energy supplied to the
TDC system (Eth,TDC) throughout the experiment are equal to
57818.1 kJ and 1836501.1 kJ, respectively; as a consequence,
a value of 0.31 can be calculated for the thermal Coefficient of
Performance by considering the energies associated to the
charge/discharge cycles as follows:
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
11:01 11:30 11:59 12:28 12:57 13:25 13:54 14:23 14:52 15:21 15:49 16:18 16:47 17:16
Bar
rel l
evel
(%)
Time (hh:mm)
Barrel A level
Barrel B level
Vw,TDC = Vw,FC=14.4 l min-1
Vw,HD = 22.8 l min-1
Fig. 4 e Volumetric water flow rate and barrel level during “normal mode” operation.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 723
COPth;TDC ¼ Ecool;FC=Eth;TDC (5)
The values of COPth,TDC found in this work agree quite well
with those measured by Udomsri et al. (2011).
In Table 2 the duration of both charging/discharging pha-
ses and “swap” periods are reported: as can be derived from
this table, the three “swaps” between barrels have a duration
of around 5 min; regarding the charging/discharging phases,
the first one shows a duration quite higher than that the other
ones.
The experiment described in Figs. 3e5 has been repeated in
order to verify its repeatability and a good agreement between
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
11:01 11:30 11:59 12:28 12:57 13:25 13:54 14:23
Pow
er (
kW)
Time (hh:m
Pth,TDC Pcool,FC
BARREL A is dischargingBARREL B is charging
BARREL B is dischargingBARREL A is charging
SWAP SW
Fig. 5 e Pth,TDC, Pcool,FC and COPth values mea
the results reported above and those achieved during the
repeated test has been found. The presented data agrees well
also with the values recorded by the CW10 internal software
(named CWIC2).
3.2. Double mode operation
The data related to the experiment carried out in “double
mode” have been gathered the 20th October 2011 from 11:11
until 15:41.
The water temperatures and volumetric flow rates
measured during the test carried out in “double mode” are
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
0.55
0.60
0.65
0.70
14:52 15:21 15:49 16:18 16:47 17:16
CO
Pth
(-)
m)
COPth,TDC
BARREL A is dischargingBARREL B is charging
BARREL B is dischargingBARREL A is charging
AP SWAP
sured during “normal mode” operation.
Table 2 e Duration of both charging/discharging phases and swap periods.
1st chargingphase
2nd chargingphase
3rd chargingphase
4th chargingphase
1stswap
2ndswap
3rdswap
Duration
(min)
116.5 87.3 93.3 89.1 5.3 5.3 5.0
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9724
reported in Figs. 6 and 7; Fig. 8 depicts the thermal power
supplied to the TDC system (Pth,TDC) and the cooling power
produced by the TDC system (Pcool,FC) as a function of the time
during “double mode” operation. Pth,TDC is calculated by using
Eq. (1), while Eq. (3) provides Pcool,FC.
As can be derived from Figs. 6e8, TDC operation in “double
mode” starts around 11:11 with both barrels discharging
(cooling power is provided); at around 11:20 the discharging
phase stops and both barrels start charging. A “swap” is per-
formed at the end of each charging period due to the fact that
Barrel A becomes completely charged (level ¼ 100%). Barrel A
and Barrel B have been charged and discharged six times
through the experiment.
Fig. 6 shows that the water temperature coming from the
boiler towards the TDC (Tw,B,out) is around 80.5 �C during
charging phase and theminimum temperature drop across the
machine is around 10 �C; the temperature from the HD to the
machine (Tw,HD,out) oscillates between around 21 and 27 �Cduring charging phase and between around 28 and 39 �C during
discharging periods. Minimum water temperature coming
from the TDC towards fan-coil (Tw,FC,in) is around 15 �C.Except during the “swap” between the barrels, the volu-
metric flow rate through both the thermally driven cooling
machine ( _Vw;TDC) and fan-coil ( _Vw;FC) is 14.4 l min�1, while
22.8 l min�1 is the water flow pumped towards the HD (Fig. 7).
As can be derived from Fig. 8, during both charging and
“swap” phases the TDC system cannot provide cooling power;
during discharging periods, cooling capacity increases till
reaching a maximum and then becomes zero: maximum
12
17
22
27
32
37
42
47
52
57
62
67
72
77
82
87
11:11 11:27 11:43 11:59 12:15 12:30 12:46 13:02 13:18
Tem
pera
ture
(°C
)
Time
Text Tw,B,out Tw,TDC,out Tw
Fig. 6 e Water temperature values measu
value of cooling power gathered during the test is about
3.0 kW. The measured values of Pcool,FC are significantly
(around 50%) lower than the expected ones: in fact, thanks to
the concurrent discharge of both barrels, “double mode”
operation should result in higher cooling power in comparison
to the “normal mode” operation. This could be due to the low
water flow rate entering the absorption system.
Compared to the test carried out in “normal mode”,
a higher charging power has been measured during “double
mode” operation (as expected). However the manufacturer
does not provide any information regarding the operation in
“double mode” and, therefore, it is not possible a comparison
with the measured values.
The cumulative cooling energy provided by the TDC
system (Ecool,FC) and the cumulative thermal energy supplied
to the TDC system (Eth,TDC) throughout the experiment are
equal to 8383.3 kJ and 167266.6 kJ, respectively; as a conse-
quence, a very low value (0.05) is obtained for the thermal
Coefficient of Performance by using Eq. (5).
In Table 3 the duration of both charging/discharging pha-
ses and “swap” periods is reported: as can be derived from this
table, the five “swaps” have a duration around 6.5 min; the
discharging phase has a duration of about 13.5 min; regarding
the charging phase, the duration oscillates between 23.7 and
29.8 min.
The experiment described in Figs. 6e8 has been repeated in
order to verify its repeatability and a good agreement between
the results mentioned above and those achieved during the
repeated test has been found. The data reported above agrees
13:34 13:50 14:06 14:21 14:37 14:53 15:09 15:25 15:41
(hh:mm)
,FC,in Tw,FC,out Tw,HD,in Tw,HD,out
red during “double mode” operation.
60
64
68
72
76
80
84
88
92
96
100
11:11 11:27 11:43 11:59 12:15 12:30 12:46 13:02 13:18 13:34 13:50 14:06 14:21 14:37 14:53 15:09 15:25 15:41
Bar
rel l
evel
(%
)
Time (hh:mm)
Barrel A level
Barrel B levelVw,TDC = Vw,FC = 14.4 l min-1
Vw,HD = 22.8 l min-1
Fig. 7 e Volumetric water flow rates and barrel level during “double mode” operation.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 725
well also with the values recorded by the CW10 internal
software (named CWIC2).
4. Energy, economic and environmentalanalysis
In order to assess the suitability of the thermally driven chiller
experimentally investigated in this paper, in the following its
measured performances are compared with those of an
0
2
4
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
11:11 11:27 11:43 11:59 12:15 12:30 12:46 13:02 13:18 13
Pth
,TD
C(k
W)
Time (hh
Pth,TDC
Fig. 8 e Pth,TDC and Pcool,FC values measur
electrically driven vapor compression chiller (EDC) from an
energy, economic and environmental point of view. The
comparison is performed by assuming that:
U TDC operates with the same water temperature and mass
flow rates measured during the experiments;
U thermal energy required by TDC is supplied by solar
collectors with the auxiliary thermal energy, required in
case of scarce solar irradiation, provided by a natural gas-
fired boiler.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
2.2
2.4
2.6
2.8
3.0
3.2
:34 13:50 14:06 14:21 14:37 14:53 15:09 15:25 15:41
Pco
ol,F
C(k
W)
:mm)
Pcool,FC
ed during “double mode” operation.
Table 3 e Duration of both charging/discharging phases and swap periods.
Charging phases Discharging phases Swap periods
1st 2nd 3rd 4th 5th 6th 1st 2nd 3rd 4th 5th 6th 1st 2nd 3rd 4th 5th
Duration
(min)
29.8 25.7 23.7 27.2 25.5 24.3 13.2 13.5 13.7 13.7 13.5 13.3 6.5 6.3 6.7 6.5 6.7
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9726
Due to the poor performance of CW10 during “double
mode” operation, the following analysis will be limited to the
experimental data gathered during “normal mode” operation.
In order to compare TDC with EDC from an energy point of
view, the Primary Energy Ratio (PER) has been evaluated. This
parameter is defined as the ratio between the useful energy
output supplied to the end-user (Ecool,FC) and the primary
energy consumption; as a consequence, the values of Primary
Energy Ratio (PER) for both TDC and EDC can be calculated as
follows:
PERTDC ¼ Ecool;FC=ðEFB$Eth;TDC=hBÞ$100 (6)
PEREDC ¼ Ecool;FC=ðEel;EDC=hPPÞ$100 ¼ COPel;EDC$hPP$100 (7)
where Ecool,FC is the cumulative cooling energy provided by the
TDC during “normal mode” operation (equal to 57818.1 kJ),
Eth,TDC is the cumulative thermal energy supplied to the TDC
during “normal mode” operation (equal to186501.1 kJ), hB is
the efficiency of the natural gas-fired boiler, hPP is the effi-
ciency of Power Plant (PP) producing electric energy, Eel,EDC is
the electric energy required by EDC for providing the same
cooling energy Ecool,FC of TDC, COPel,EDC is the electric Coeffi-
cient of Performance of EDC (defined as the ratio between the
cooling power supplied by EDC and the electric power
consumed by EDC), EFB is the fraction of Eth,TDC provided by
the natural gas-fired boiler (so that the difference (1 � EFB) is
the fraction of Eth,TDC recovered from solar collectors).
25
45
65
85
105
125
145
165
185
205
225
245
265
285
0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
PE
R (
%)
EF
PER_TDC PE
Fig. 9 e Primary energy ratio and primar
Fig. 9 shows the values of both PERTDC and PEREDC at
varying EFB from 0.1 (natural gas-fired boiler produces 10% of
Eth,TDC) to 0.9 (solar collectors field provides 10% of Eth,TDC).
The data depicted in this figure have been obtained by
assuming the following values:
U hB ¼ 0.9;
U hPP ¼ 0.46 (Rosato and Sibilio, 2012);
U COPel,EDC ¼ 2.
The value of hPP includes transmission and distribution
losses.
In the same figure the values of Primary Energy Saving
(PES ) are also reported. The parameter PES allows to evaluate
the potential of primary energy saving; so it is defined as re-
ported below:
PES ¼ ½1� ðPEREDC=PERTDCÞ�$100 (8)
Positive values of PES mean that TDC allows for an energy
saving in comparison to EDC.
Fig. 9 denotes that PES increases at decreasing the value of
EFB till reaching its maximum value (around 70%) when
EFB ¼ 0.1. From this figure it can be derived that the thermally
drive chiller investigated in this work is suitable from an
energy point of view (PES > 0) if compared to a conventional
electrically driven refrigerating systemwith COPel,EDC ¼ 2 only
in case of EFB < 0.3, i.e. only when the most part (at least 70%)
-230
-210
-190
-170
-150
-130
-110
-90
-70
-50
-30
-10
10
30
50
70
0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
PE
S (%
)
B (-)
R_EDC PES
y energy saving as a function of EFB.
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
60
80
1.01.52.02.53.03.54.04.55.05.56.06.57.07.58.08.59.09.5
10.010.511.011.512.0
0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
CO
2(%
)
OCgk(
snoissimetnelaviuqe
edixoidnobra
C2)
EFB (-)
CO2_TDC CO2_EDC DeltaCO2
C
Fig. 10 e Carbon dioxide equivalent emissions as a function of EFB.
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 727
of thermal energy required by the TDC is recovered from solar
collectors.
However the choice of the energy conversion technology
cannot be based only on the energy performances, but it
should be also affected by the assessment of the environ-
mental impact. In the following the carbon dioxide equivalent
emissions of both TDC and EDC have been assessed by using
the following formulas:
CO2;TDC ¼ ½b$ðEFB$Eth;TDC=hBÞ��3600 (9)
CO2;EDC ¼ a$Eel;EDC=3600 (10)
where a represents the equivalent CO2 emissions in the
power plant for 1 kWh of electric energy consumed and
0.30.50.70.91.11.31.51.71.92.12.32.52.72.93.13.33.53.73.94.14.34.54.74.9
0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Ope
rati
ng c
ost
(€)
EFB
C_TDC C_E
Fig. 11 e Operating cost as a function of E
b represents the equivalent CO2 emissions for 1 kWh of
primary energy consumed. The following values have been
assumed:
� a ¼ 0.523 kgCO2 kWh�1 (Rosato and Sibilio, 2012)
� b ¼ 0.2 kgCO2 kWh�1 (Rosato and Sibilio, 2012).
The equivalent CO2 emissions due to electricity production
are typical of the mix of technologies adopted in the Italian
geographic area.
Fig. 10 shows the values of CO2,TDC and CO2,EDC as function
of EFB. The percentage difference DCO2 between CO2,TDC and
CO2,EDC is also reported in Fig. 10:
DCO2 ¼ ½1� ðCO2;TDC=CO2;EDCÞ�$100 (11)
-240
-220
-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
60
80
0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
C (%
)
(-)
DC DeltaC
C
FB during “normal mode” operation.
-400-380-360-340-320-300-280-260-240-220-200-180-160-140-120-100
-80-60-40-20
020406080
0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
PE
S (%
),
CO
2 (%
),
C (
%)
EFB (-)
PES for COP_EDC=1.5
PES for COP_EDC=3
DeltaCO2 for COP_EDC=1.5
DeltaCO2 for COP_EDC=3
DeltaC for COP_EDC=1.5
DeltaC for COP_EDC=3
CC
Fig. 12 e Comparison between absorption heat pump and electric driven chiller.
i n t e r n a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9728
Data reported in Fig. 10 show that, in comparison to the
EDC, thermally driven chiller investigated in this paper allows
for a reduction of CO2 emissions only in case the fraction of
Eth,TDC provided by the natural gas-fired boiler is lower than
around 36%; as a consequence, the thermal energy supplied to
the TDC coming from solar collectors has to be larger than 64%
in order to guarantee the suitability of the TDC in comparison
to the EDC from an environmental point of view.
However the evaluation of economic performance
indices is also necessary to complete the analysis of the TDC
suitability. As known, the assessment of investment profit-
ability depends on country conditions, as feed-in tariffs,
bonus payment, market mechanism and even tax rebates.
As a consequence, estimating economic benefits is made
difficult by the large number of parameters involved and by
the fact that incentives are often assigned according to
complex schemes. In the following only the operating cost of
the TDC has been evaluated and compared to that one of
EDC in order to give a general indication. Natural gas and
electricity prices in the domestic sector vary largely across
Europe: TDC system financial viability in the Italian market is
investigatedbyassuminganelectricenergypriceCUelequalto0.18
V kWh�1 (Rosato and Sibilio, 2012) and a natural gas price CUng
equal to0.80VNm�3 (RosatoandSibilio, 2012).Theoperatingcost
of both TDC and EDC has been estimated by using the following
equations:
CTDC ¼ EFB$Eth;TDC=�3600$hB$LHVng
�$CUng (12)
CEDC ¼ Ecool;TDC=ð3600$COPEDCÞ$CUel (13)
where LHVng is the Lower Heating Value of natural gas
(assumed equal to 9.593 kWh Nm�3).
The percentage difference between CTDC and CEDC is
calculated as follows and reported in Fig. 11:
DC ¼ ½1� ðCEDC=CTDCÞ�$100 (14)
Fig. 11 shows that, if compared with the EDC, the TDC
allows for an operating cost reduction when the parameter
EFB becomes lower than around 0.28: this means that the TDC
allows to reduce the operating cost only in case the percentage
of Eth,TDC recovered from solar collectors is higher than 68%.
Taking into consideration that the performance of electric
driven chiller is affectedby the externalweather conditionsand
loads, the comparison between the absorption chiller and the
electric driven chiller has been performed by considering two
additional values (1.5 and 3.0) of COPel,EDC. The comparison has
been performed from an energy, environmental and economic
point of view. Fig. 12 shows the results of this comparison.
Fromthisfigureitcanbederivedthat,comparedtotheelectric
driven chiller with COPel,EDC ¼ 1.5, the thermally drive chiller
investigated in thiswork is suitable frombothanenergypoint of
view(PES> 0)andaneconomicpointof view (DC> 0) only incase
the thermal energyprovided by solar collectors isnot lower than
60%ofthermalenergyrequiredbytheTDC; theabsorptionchiller
allowsfor reducing thecarbondioxideemissions ifpercentageof
Eth,TDC recovered from solar collectors is higher than 52%.
In comparison to the electric driven chiller with COPel,EDC ¼3.0, the thermally drive chiller investigated in this work allows
for saving both energy and money only in case the thermal
energy provided by solar collectors is not lower than 80% of
thermal energy required by the TDC; the absorption chiller is
suitable from an environmental point of view if percentage of
Eth,TDC recovered from solar collectors is higher than 75%.
5. Conclusions
The 4th generation of a three-phase absorption chiller/heat
pump that is capable of storing energy internally in the form
of crystallized salt (LiCl) with water as refrigerant, patented in
2000 and sold by the Swedish company ClimateWell�, has
been experimentally investigated. Data have been gathered
i n t e rn a t i o n a l j o u r n a l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7e7 2 9 729
during two different systemmodes operation: “normal mode”
and “double mode”. The performed tests showed a maximum
coefficient of performance COPth,TDC equal to about 0.6 while
the machine was operating in “normal mode”; the measured
system performance during “double mode” was significantly
worse than that measured during “normal mode” operation.
The measured data have been used to compare the
performance of the thermally driven cooling systemwith that
one of a conventional electrically driven refrigerating
machine. The comparison has been performed from an
energy, economic and environmental point by assuming that
the thermal energy required by the TDC is supplied by both
a solar collectors field and a natural gas-fired boiler. The
comparison pointed out that, in comparisonwith the EDC, the
TDC allows for a reduction of both primary energy
consumption, carbon dioxide emissions and operating cost in
case of at least 70% of thermal energy required by the TDC is
recovered from solar collectors (instead of provided by
a conventional natural gas-fired boiler). Comparison between
electric driven chiller and absorption heat pumphas been also
performed by considering two different values of COPel,EDC.
However additional tests should be carried out in order to
highlight the system performance over a wider range of
operating conditions; in addition a comparison of the experi-
mental data against the simulation model developed by
Udomsri et al. (2011, 2012) has to be performed in order to
verify the accuracy of the model, and the suitability of the
model itself for both determining the effect of boundary
conditions on the machine efficiency and for evaluating the
viability of the thermally driven chiller CW10 in comparison to
traditional systems via a techno-economic analysis.
r e f e r e n c e s
AISIN-SEIKI. http://www.aisin.com/ (accessed 26.07.12.).Angrisani, G., Rosato, A., Roselli, C., Sasso, M., Sibilio, S., 2012.
Experimental results of a micro-trigeneration installation.Appl. Therm. Eng. 38, 78e90.
Balaras, C.A., Gaglia, A.G., Georgopoulou, E., Mirasgedis, S.,Sarafidis, Y., Lalas, D.P., 2007. European residential buildingsand empirical assessment of the Hellenic building stock,energy consumption, emissions and potential energy savings.Build. Environ. 42, 1298e1314.
Bales, C., Ayadi, O., 2009. Modelling of commercial absorptionheat pump with integral storage. In: Proceedings Effstock2009. Swedvac, Stockholm, Sweden.
Bales, C., Nordlander, S., 2005. TCA Evaluation, LabMeasurements, Modelling and System Simulations. No. 3-9809656-4-3. Hogskolan Dalarna, Borlange, Sweden.
Boer, R., Haije, W.G., Veldhuis, J.B.J., 2002. Determination ofstructural, thermodynamic and phase properties in theNa2SeH2O system for application in a chemical heat pump.Thermochim. Acta 395, 3e19.
Chicco, G., Mancarella, P., 2009. Distributed multi-generation:a comprehensive view. Renew. Sust. Energ. Rev. 13, 535e551.
ClimateWell�. http://www.climatewell.com/index.html#/innovation/Documentation/TechSpecs (accessed 26.07.12.).
Conde, M.R., 2004. Properties of aqueous solution of lithium andcalcium chlorides: formulations for use in air conditioningequipment design. Int. J. Therm. Sci. 43, 367e382.
Fiskum, R.J., Adcock, P.W., DeVault, R.C., 1996. United StatesDepartment of Energy thermally activated heat pumpprogram. In: Energy Conversion Engineering Conference.Proceedings of the 31st Intersociety, vol. 732, pp. 736e741.
Florides, G.A., Kalogirou, S.A., Tassou, S.A., Wrobel, L.C., 2002.Modelling, simulation and warming impact assessment ofa domestic-size absorption solar cooling system. Appl. Therm.Eng. 22, 1313e1325.
Henning, H.M., 2007. Solar assisted air conditioning of buildings ean overview. Appl. Therm. Eng. 27, 1734e1749.
Hernandez-Santoyo, J., Sanchez-Cifuentes, A., 2003. Trigeneration:an alternative for energy savings. Appl. Energy 76, 219e227.
Huicochea, A., Rivera, W., Gutierrez-Urueta, G., Bruno, J.C.,Coronas, A., 2011. Thermodynamic analysis of a trigenerationsystem consisting of a micro gas turbine and a double effectabsorption chiller. Appl. Therm. Eng. 31, 3347e3353.
Kavvadias, K.C., Tosios, A.P., Maroulis, Z.B., 2010. Design ofa combined heating, cooling and power system: sizing,operation strategy selection and parametric analysis. Energ.Convers. Manag. 51, 833e845.
Li, H., Fu, L., Geng, K., Jiang, Y., 2006. Energy utilization evaluationof CCHP systems. Energy Buildings 38, 253e257.
Lin, L., Wang, Y., Al-Shemmeri, T., Ruxton, T., Turner, S., Zeng, S.,Huang, J., He, Y., Huang, X., 2007. An experimentalinvestigation of a household size trigeneration. Appl. Therm.Eng. 27, 576e585.
McMullan, J.T., 2002. Refrigeration and the environment e issuesand strategies for the future. Int. J. Refrigeration 25, 89e99.
National instruments�. http://www.ni.com/ (accessed 26.07.12.).Ogura, H., Yamamoto, T., Kage, H., 2003. Efficiencies of CaO/H2O/
Ca(OH)2 chemical heat pump for heat storing and heating/cooling. Energy 28, 1479e1493.
Olsson, R., Kaarebring-Olsson, M., Jonsson, S., 2000. A ChemicalHeat Pump. Patent No: WO0037864, Sweden.
Rosato, A., Sibilio, S., 2012. Calibration and validation of a modelfor simulating thermal and electric performance of an internalcombustion engine-based micro-cogeneration device. Appl.Therm. Eng. 45, 79e98.
Serra, L.M., Lozano, M.A., Ramos, J., Ensinas, A.V., Nebra, S.A.,2009. Polygeneration and efficient use of natural resources.Energy 34, 575e586.
Srikhirin, P., Aphornratana, S., Chungpaibulpatana, S., 2001. Areview of absorption refrigeration technologies. Renew. Sust.Energ. Rev. 5, 343e372.
TRNSYS. TRNSYS Coordinator, Solar Energy Laboratory,University of Wisconsin-Madison, 1500 Engineering Drive,1303 Engineering Research Building Madison, WI 53706, U.S.A.
Udomsri, S., Bales, C., Martin, A.R., Martin, V., 2011. Decentralisedcooling in district heating network: monitoring results andcalibration of simulation model. Energy Buildings 43,3311e3321.
Udomsri, S., Bales, C., Martin, A.R., Martin, V., 2012. Decentralizedcooling in district heating network: system simulation andparametric study. Appl. Energy 92, 175e184.
Wang, K., Abdelaziz, O., Kisari, P., Vineyard, E.A., 2011. State-of-the-art review on crystallization control technologies forwater/LiBr absorption heat pumps. Int. J. Refrigeration 34,1325e1337.
Wongsuan, W., Kumar, S., Neveu, P., Meunier, F., 2001. A reviewof chemical heat pump technology and applications. Appl.Therm. Eng. 21, 1489e1519.