Preliminary experimental characterization of a three-phase

absorption 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
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 efciency 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.
2012 Elsevier Ltd and IIR. All rights reserved.
Caracte risation expe rimentale pre liminaire dune pompe a`
chaleur a` trois phases
Mots cles : cycle a` absorption ; refroidisseur a` entranement thermique ; pompe a` chaleur chimique ; chlorure de lithium ; refroidissement
solaire ; trige ne ration
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 nowadays mostly generated
from fossil fuels. This trend has caused important
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
* Corresponding author. Tel./fax: 39 081 8122530.
E-mail address: antonio.rosato@unina2.it (A. Rosato).
www. i i i r . or g
Available online at www.sciencedirect.com
j ournal homepage: www. el sevi er. com/ l ocat e/ i j ref ri g
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9
0140-7007/$ e see front matter 2012 Elsevier Ltd and IIR. All rights reserved.
http://dx.doi.org/10.1016/j.ijrefrig.2012.11.015
can use low grade energy sources that are environmentally
friendlier instead of electricity. Several scientic papers
studied the integration of different types of commercially
available absorption systems withcogeneration 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 signicant potential in terms of energy savings and
reduction of CO
2
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 NH
3
/water and salt-water
absorption chiller (Srikhirin et al., 2001) and/or chemical
heat pump (Wongsuan et al., 2001). Chemical heat pump is
a newand promising technology whichis capable of operating
with low temperature heat sources: salt-water solutions such
as lithiumchloride (LiCl)/water, sodiumsulphite (Na
2
S)/water,
and calciumchlorides (CaCl
2
)/water, etc. have been used (Boer
et al., 2002; Conde, 2004; Ogura et al., 2003). Absorption chillers
are more commonat mediumor 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 uid 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 ve generations of prototypes. The 4th generation of

machines, that was the rst 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 ow 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
Nomenclature
Latin letters
B natural gas-red boiler
c specic heat (kJ kg
1
K
1
)
C operating cost (V)
CO
2
carbon dioxide equivalent emission (kg CO
2
)
COP coefcient of performance
CW10 ClimateWell10
CWIC2 CW10 internal software
CU
ng
natural gas Unit Cost (V Nm
3
)
CU
el
electric energy Unit Cost (V kWh
1
)
E energy (kJ)
EDC electrically driven chiller system
EF
B
fraction of E
th,TDC
produced by natural gas-red
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 ow meter
MCHP micro combined heat and power generation
MG natural gas volumetric ow 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 ow rate (m
3
s
1
)
Greeks
a CO
2
emission factor for electric energy
(kgCO
2
kWh
1
)
b CO
2
emission factor for primary energy
(kgCO
2
kWh
1
)
D difference (%)
h efciency
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 nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 718
process occurring in each barrel works in batch mode, with
a separate desorption (charge) phase followed by absorption
(discharge) phase:
Uduring 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;
Uduring 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
shouldbelost tothe surroundings; whena barrel is charged, the
energystays storedinthe barrel until there is a coolingdemand.
A by-pass valve is installed in the machine for regulating
the water 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 ows 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
double mode, 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
veried:
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/
T
w,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;
Fig. 1 e Schematic of complete CW10 machine on the left (Udomsri and Bales, 2011) and single barrel on the right (Bales and
Ayadi, 2009).
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 719
where TC0 is the temperature of internal water going towards
the heat dissipator before the by-pass valve and T
w,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 red with
municipal waste; they investigated the systemduring normal
mode operation and found a maximumthermal coefcient of
performance during the hottest period of around 0.50;
however, the gure was only 0.41 for the complete monitoring
period during the summer of 2008. According to the monitored
results obtained fromthe demonstration, a systemsimulation
model for the TRNSYS environment has been calibrated by
Udomsri et al. (2012) and used to nd improved system design
and control. Bales and Nordlander (2005) carried out just few
of the planned experiments on CW10 model during full cycle
operation due to lack of time before the machines were
shipped. Of these, most had missing data due to anerror inthe
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
coefcient 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 veried 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 coefcient of performance.
Even if some data have been already available in literature,
the CW10 unit has not been yet investigated during double
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 (thermal power supplied, coolingpower delivered,
coefcient of performance, temperature levels, etc.) will be
presentedandanalyzedindetail. Inadditionthemeasureddata
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
reportedinthe following canbe alsousedtoverify the accuracy
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 combustionengine
(commercialized by AISIN-SEIKI company) and stored in 1000 L
Fig. 2 e Experimental set-up.
i nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 720
tank; taking into consideration that the TDC needs a charging
temperature at least 50

Clarger thanthe heat sinktemperature

to start the absorption process and that the temperature of hot
water owing fromthe MCHP systeminto the storage cannot be
higher than around 70

C, a natural gas-red 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-
latedwith50mmexiblepolyurethanelayer andisfurthermore
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 kWis installedas heat sink. The water
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 ow
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 ow rate in the key-points of the plant
(ow rate entering TDC, FC, HD, B, HWS, MCHP);
natural gas volumetric ow rate entering both micro-
cogenerator and natural gas-red boiler;
electric power suppliedbymicro-cogenerator tothe end-user.
Water and ambient temperatures are measured by using
resistance thermometers Pt100; water mass ow rate is ob-
tained by using an ultrasonic mass ow sensor, while
a thermal mass ow meter is installed to evaluate the natural
gas volumetric ow rate; three wattmeters measure the
electric ows 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.
The TDC installed at SUN lab is equipped with an internal
software (named CWIC2) by means of which several operation
systemparameters canbemonitoredandrecorded: 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, thelevel of eachbarrel
during system operation, the by-pass valve position, etc.
Based on the direct measurements, the parameters listed
beloware calculatedinorder toevaluate theplant performances:
P
th;TDC

_
V
w;TDC
$r
w
$c
w
$

T
w;TDC;in
T
w;TDC;out

(1)
P
th;HD

_
V
w;HD
$r
w
$c
w
$

T
w;HD;in
T
w;HD;out

(2)
P
cool;FC

_
V
w;FC
$r
w
$c
w
$

T
w;FC;out
T
w;FC;in

(3)
where the water specic heat and the water density, respec-
tively, have been assumed equal to c
w
4.187 kJ (kg K)
1
and
r
w
990 kg m
3
.
The signals coming from the resistance thermometers
Pt100 are acquired by three cFP-RTD-124 analog input modules
(producedbyNational Instruments

), whilethesignals 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 (4 O20 mA). The digital data coming from the
modules are sent to a personal computer. The software Lab-
View 8.2 is used to dene the acquisition frequency and to
monitor and/or recordall the directlymeasuredandcalculated
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 the main 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 normal mode and double mode
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
Table 1 e Main characteristics of the plant
instrumentation.
Parameter Instrument Operating
range
Accuracy
T Resistance
thermometer
Pt100
50 O 100

C 0.2

C
_
V
w
Ultrasonic
volumetric
ow meter
0 O 50 l min
1
2.5% of
full scale
_
V
ng
Thermal
volumetric
ow meter
0 O 5.0 Nm
3
h
1
0.8% of
reading
0.2% of
full scale
P
el,MCHP
Wattmeter 0 O 6 kW 0.2% of
full scale
0 O 10 kW
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 721
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 the water volumetric owrates owing throughthe
thermally driven chiller, the heat dissipator and the fan-coil.
Fig. 5 depicts the thermal power supplied to the TDC system
(P
th,TDC
), thecoolingpower producedbytheTDCsystem(P
cool,FC
)
and the thermal Coefcient Of Performance (COP
th,TDC
) as
afunctionof thetimeduringnormal mode operation. P
th,TDC
is
calculated by using Eq. (1), while Eq. (3) provides P
cool,FC
; the
instantaneous values of COP
th,TDC
are dened as follows:
COP
th;TDC
P
cool;FC
=P
th;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/T
w,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 (T
w,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
(T
w,HD,out
) oscillates between around 26 and 30

C. Water
temperature coming from the TDC towards the fan-coil
(T
w,FC,in
) is around 15

C, with a minimum value of 12.6

C
achieved around 11:10.
Except during the swap between the barrels, the volu-
metric ow rate through both the thermally driven cooling
system (
_
V
w;TDC
) and fan-coil (
_
V
w;FC
) is 14.4 l min
1
(15.0 l min
1
is suggested as minimum water ow rate by ClimateWell

),
while 22.8 l min
1
is the water ow (
_
V
w;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 T
w,TDC,in
80

C, T
w,HD,out
30

C and
T
w,FC,out
20

C. During Barrel A discharging the values of
P
cool,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 COP
th,TDC
(dened 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 COP
th,TDC
measured
during Barrel A discharging is around 0.6, quite higher than
the greater value of COP
th,TDC
achieved during Barrel B
discharging.
The cumulative cooling energy supplied by the TDCsystem
(E
cool,FC
) and the cumulative thermal energy supplied to the
TDC system (E
th,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 Coefcient of
Performance by considering the energies associated to the
charge/discharge cycles as follows:
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
T
e
m
p
e
r
a
t
u
r
e

(

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 nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 722
COP
th;TDC
E
cool;FC
=E
th;TDC
(5)
The values of COP
th,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 rst 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
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 ow 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
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 14:52 15:21 15:49 16:18 16:47 17:16
C
O
P
t
h
(
-
)
P
o
w
e
r

(
k
W
)

Time (hh:mm)
Pth,TDC Pcool,FC COPth,TDC
BARREL A is discharging
BARREL B is charging
BARREL A is discharging
BARREL B is charging
BARREL B is discharging
BARREL A is charging
BARREL B is discharging
BARREL A is charging
SWAP SWAP SWAP
Fig. 5 e P
th,TDC
, P
cool,FC
and COP
th
values measured during normal mode operation.
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
B
a
r
r
e
l

l
e
v
e
l

(
%
)
Time (hh:mm)
Barrel A level
Barrel B level
V
w,TDC
= V
w,FC
=14.4 l min
-1
V
w,HD
= 22.8 l min
-1
Fig. 4 e Volumetric water ow rate and barrel level during normal mode operation.
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 723
reported in Figs. 6 and 7; Fig. 8 depicts the thermal power
supplied to the TDC system (P
th,TDC
) and the cooling power
produced by the TDC system (P
cool,FC
) as a function of the time
during double mode operation. P
th,TDC
is calculated by using
Eq. (1), while Eq. (3) provides P
cool,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 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 (T
w,B,out
) is around 80.5

C during
charging phase and the minimum temperature drop across the
machine is around 10

C; the temperature from the HD to the
machine (T
w,HD,out
) oscillates between around 21 and 27

C
during charging phase and between around 28 and 39

C during
discharging periods. Minimum water temperature coming
from the TDC towards fan-coil (T
w,FC,in
) is around 15

C.
Except during the swap between the barrels, the volu-
metric ow rate through both the thermally driven cooling
machine (
_
V
w;TDC
) and fan-coil (
_
V
w;FC
) is 14.4 l min
1
, while
22.8 l min
1
is the water ow 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
value of cooling power gathered during the test is about
3.0 kW. The measured values of P
cool,FC
are signicantly
(around 50%) lower than the expected ones: in fact, thanks to
the concurrent discharge of both barrels, double mode
operationshould result in higher cooling power incomparison
to the normal mode operation. This could be due to the low
water ow 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 (E
cool,FC
) and the cumulative thermal energy supplied
to the TDC system (E
th,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
Coefcient 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 fromthis
table, the ve 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
Table 2 e Duration of both charging/discharging phases and swap periods.
1st charging
phase
2nd charging
phase
3rd charging
phase
4th charging
phase
1st
swap
2nd
swap
3rd
swap
Duration
(min)
116.5 87.3 93.3 89.1 5.3 5.3 5.0
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 13:34 13:50 14:06 14:21 14:37 14:53 15:09 15:25 15:41
T
e
m
p
e
r
a
t
u
r
e

(

C
)

Time (hh:mm)
Text Tw,B,out Tw,TDC,out Tw,FC,in Tw,FC,out Tw,HD,in Tw,HD,out
Fig. 6 e Water temperature values measured during double mode operation.
i nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 724
well also with the values recorded by the CW10 internal
software (named CWIC2).
4. Energy, economic and environmental
analysis
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
electrically driven vapor compression chiller (EDC) from an
energy, economic and environmental point of view. The
comparison is performed by assuming that:
UTDC operates with the same water temperature and mass
ow rates measured during the experiments;
Uthermal 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-
red boiler.
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
B
a
r
r
e
l

l
e
v
e
l

(
%
)

Time (hh:mm)
Barrel A level
Barrel B level
V
w,TDC
= V
w,FC
= 14.4 l min
-1
V
w,HD
= 22.8 l min
-1
Fig. 7 e Volumetric water ow rates and barrel level during double mode operation.
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
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 13:34 13:50 14:06 14:21 14:37 14:53 15:09 15:25 15:41
P
c
o
o
l
,
F
C
(
k
W
)
P
t
h
,
T
D
C
(
k
W
)
Time (hh:mm)
Pth,TDC Pcool,FC
Fig. 8 e P
th,TDC
and P
cool,FC
values measured during double mode operation.
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 725
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 dened as the ratio between the useful energy
output supplied to the end-user (E
cool,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:
PER
TDC
E
cool;FC
=EF
B
$E
th;TDC
=h
B
$100 (6)
PER
EDC
E
cool;FC
=E
el;EDC
=h
PP
$100 COP
el;EDC
$h
PP
$100 (7)
where E
cool,FC
is the cumulative cooling energy provided by the
TDC during normal mode operation (equal to 57818.1 kJ),
E
th,TDC
is the cumulative thermal energy supplied to the TDC
during normal mode operation (equal to186501.1 kJ), h
B
is
the efciency of the natural gas-red boiler, h
PP
is the ef-
ciency of Power Plant (PP) producing electric energy, E
el,EDC
is
the electric energy required by EDC for providing the same
cooling energy E
cool,FC
of TDC, COP
el,EDC
is the electric Coef-
cient of Performance of EDC (dened as the ratio between the
cooling power supplied by EDC and the electric power
consumed by EDC), EF
B
is the fraction of E
th,TDC
provided by
the natural gas-red boiler (so that the difference (1 EF
B
) is
the fraction of E
th,TDC
recovered from solar collectors).
Fig. 9 shows the values of both PER
TDC
and PER
EDC
at
varying EF
B
from 0.1 (natural gas-red boiler produces 10% of
E
th,TDC
) to 0.9 (solar collectors eld provides 10% of E
th,TDC
).
The data depicted in this gure have been obtained by
assuming the following values:
Uh
B
0.9;
Uh
PP
0.46 (Rosato and Sibilio, 2012);
UCOP
el,EDC
2.
The value of h
PP
includes transmission and distribution
losses.
In the same gure 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 dened as re-
ported below:
PES 1 PER
EDC
=PER
TDC
$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
EF
B
till reaching its maximum value (around 70%) when
EF
B
0.1. From this gure 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 system with COP
el,EDC
2 only
in case of EF
B
< 0.3, i.e. only when the most part (at least 70%)
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
-230
-210
-190
-170
-150
-130
-110
-90
-70
-50
-30
-10
10
30
50
70
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 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
P
E
S

(
%
)
P
E
R

(
%
)

EF
B
(-)
PER_TDC PER_EDC PES
Fig. 9 e Primary energy ratio and primary energy saving as a function of EF
B
.
i nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 726
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:
CO
2;TDC
b$EF
B
$E
th;TDC
=h
B

3600 (9)
CO
2;EDC
a$E
el;EDC
=3600 (10)
where a represents the equivalent CO
2
emissions in the
power plant for 1 kWh of electric energy consumed and
b represents the equivalent CO
2
emissions for 1 kWh of
primary energy consumed. The following values have been
assumed:
a 0.523 kgCO
2
kWh
1
(Rosato and Sibilio, 2012)
b 0.2 kgCO
2
kWh
1
(Rosato and Sibilio, 2012).
The equivalent CO
2
emissions due to electricity production
are typical of the mix of technologies adopted in the Italian
geographic area.
Fig. 10 shows the values of CO
2,TDC
and CO
2,EDC
as function
of EF
B
. The percentage difference DCO
2
between CO
2,TDC
and
CO
2,EDC
is also reported in Fig. 10:
DCO
2
1 CO
2;TDC
=CO
2;EDC
$100 (11)
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
60
80
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
9.0
9.5
10.0
10.5
11.0
11.5
12.0
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
C
O
2
(
%
)
O
C
g
k
(
s
n
o
i
s
s
i
m
e
t
n
e
l
a
v
i
u
q
e
e
d
i
x
o
i
d
n
o
b
r
a
C
2
)

EF
B
(-)
CO2_TDC CO2_EDC DeltaCO2
C
Fig. 10 e Carbon dioxide equivalent emissions as a function of EF
B
.
-240
-220
-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
60
80
0.3
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
2.3
2.5
2.7
2.9
3.1
3.3
3.5
3.7
3.9
4.1
4.3
4.5
4.7
4.9
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
C

(
%
)
O
p
e
r
a
t
i
n
g

c
o
s
t


(

)

EF
B
(-)
C_TDC C_EDC DeltaC
C
Fig. 11 e Operating cost as a function of EF
B
during normal mode operation.
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 727
Data reported in Fig. 10 show that, in comparison to the
EDC, thermally driven chiller investigated in this paper allows
for a reduction of CO
2
emissions only in case the fraction of
E
th,TDC
provided by the natural gas-red boiler is lower than
around 36%; as a consequence, the thermal energy supplied to
the TDCcoming fromsolar 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 prot-
ability depends on country conditions, as feed-in tariffs,
bonus payment, market mechanism and even tax rebates.
As a consequence, estimating economic benets is made
difcult 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 nancial viability in the Italian market is
investigatedbyassuminganelectricenergypriceCU
el
equalto0.18
V kWh
1
(Rosato and Sibilio, 2012) and a natural gas price CU
ng
equal to0.80VNm
3
(RosatoandSibilio, 2012). Theoperatingcost
of both TDC and EDC has been estimated by using the following
equations:
C
TDC
EF
B
$E
th;TDC
=

3600$h
B
$LHV
ng

$CU
ng
(12)
C
EDC
E
cool;TDC
=3600$COP
EDC
$CU
el
(13)
where LHV
ng
is the Lower Heating Value of natural gas
(assumed equal to 9.593 kWh Nm
3
).
The percentage difference between C
TDC
and C
EDC
is
calculated as follows and reported in Fig. 11:
DC 1 C
EDC
=C
TDC
$100 (14)
Fig. 11 shows that, if compared with the EDC, the TDC
allows for an operating cost reduction when the parameter
EF
B
becomes lower than around 0.28: this means that the TDC
allows to reduce the operating cost only in case the percentage
of E
th,TDC
recovered from solar collectors is higher than 68%.
Taking into consideration that the performance of electric
drivenchiller is affectedby the external weather conditions and
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 COP
el,EDC
. The comparison has
been performed from an energy, environmental and economic
point of view. Fig. 12 shows the results of this comparison.
Fromthisgureitcanbederivedthat, comparedtotheelectric
driven chiller with COP
el,EDC
1.5, the thermally drive chiller
investigatedinthis workis suitable frombothanenergy point of
view(PES>0) andaneconomicpoint of view(DC>0) onlyincase
thethermal energy providedby solar collectors is not lower than
60%of thermal energyrequiredbytheTDC; theabsorptionchiller
allowsfor reducingthecarbondioxideemissionsif percentageof
E
th,TDC
recovered fromsolar collectors is higher than 52%.
In comparison to the electric driven chiller with COP
el,EDC

3.0, 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
E
th,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
-400
-380
-360
-340
-320
-300
-280
-260
-240
-220
-200
-180
-160
-140
-120
-100
-80
-60
-40
-20
0
20
40
60
80
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
P
E
S

(
%
)
,

C
O
2

(
%
)
,

C

(
%
)
EF
B
(-)
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
C
C
Fig. 12 e Comparison between absorption heat pump and electric driven chiller.
i nt e r na t i o na l j o ur na l of r e f r i ge r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 728
during two different system modes operation: normal mode
and double mode. The performed tests showed a maximum
coefcient of performance COP
th,TDC
equal to about 0.6 while
the machine was operating in normal mode; the measured
system performance during double mode was signicantly
worse than that measured during normal mode operation.
The measured data have been used to compare the
performance of the thermally driven cooling system with 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 eld and a natural gas-red boiler. The
comparison pointed out that, in comparison with 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-red boiler). Comparison between
electric driven chiller and absorption heat pump has been also
performed by considering two different values of COP
el,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 efciency and for evaluating the
viability of the thermally drivenchiller CW10 in comparisonto
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.,
Saradis, Y., Lalas, D.P., 2007. European residential buildings
and 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 absorption
heat pump with integral storage. In: Proceedings Effstock
2009. Swedvac, Stockholm, Sweden.
Bales, C., Nordlander, S., 2005. TCA Evaluation, Lab
Measurements, Modelling and System Simulations. No. 3-
9809656-4-3. Hogskolan Dalarna, Borlange, Sweden.
Boer, R., Haije, W.G., Veldhuis, J.B.J., 2002. Determination of
structural, thermodynamic and phase properties in the
Na
2
SeH
2
O 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 and
calcium chlorides: formulations for use in air conditioning
equipment design. Int. J. Therm. Sci. 43, 367e382.
Fiskum, R.J., Adcock, P.W., DeVault, R.C., 1996. United States
Department of Energy thermally activated heat pump
program. 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 of
a domestic-size absorption solar cooling system. Appl. Therm.
Eng. 22, 1313e1325.
Henning, H.M., 2007. Solar assisted air conditioning of buildings e
an 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., Gutie rrez-Urueta, G., Bruno, J.C.,
Coronas, A., 2011. Thermodynamic analysis of a trigeneration
system consisting of a micro gas turbine and a double effect
absorption chiller. Appl. Therm. Eng. 31, 3347e3353.
Kavvadias, K.C., Tosios, A.P., Maroulis, Z.B., 2010. Design of
a 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 evaluation
of 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 experimental
investigation of a household size trigeneration. Appl. Therm.
Eng. 27, 576e585.
McMullan, J.T., 2002. Refrigeration and the environment e issues
and 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. Efciencies of CaO/H
2
O/
Ca(OH)
2
chemical heat pump for heat storing and heating/
cooling. Energy 28, 1479e1493.
Olsson, R., Kaarebring-Olsson, M., Jonsson, S., 2000. A Chemical
Heat Pump. Patent No: WO0037864, Sweden.
Rosato, A., Sibilio, S., 2012. Calibration and validation of a model
for simulating thermal and electric performance of an internal
combustion 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 efcient use of natural resources.
Energy 34, 575e586.
Srikhirin, P., Aphornratana, S., Chungpaibulpatana, S., 2001. A
review 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. Decentralised
cooling in district heating network: monitoring results and
calibration of simulation model. Energy Buildings 43,
3311e3321.
Udomsri, S., Bales, C., Martin, A.R., Martin, V., 2012. Decentralized
cooling in district heating network: system simulation and
parametric study. Appl. Energy 92, 175e184.
Wang, K., Abdelaziz, O., Kisari, P., Vineyard, E.A., 2011. State-of-
the-art review on crystallization control technologies for
water/LiBr absorption heat pumps. Int. J. Refrigeration 34,
1325e1337.
Wongsuan, W., Kumar, S., Neveu, P., Meunier, F., 2001. A review
of chemical heat pump technology and applications. Appl.
Therm. Eng. 21, 1489e1519.
i nt e r na t i ona l j o ur na l o f r e f r i g e r a t i o n 3 6 ( 2 0 1 3 ) 7 1 7 e7 2 9 729