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.