Experimental investigation on the performanceof a solar powered lithium bromide–waterabsorption cooling system

Ming Li a,*, Chengmu Xu a, Reda Hassanien Emam Hassanien a,b,Yongfeng Xu a,c, Binwei Zhuang a

a Solar Energy Research Institute, Yunnan Normal University, Kunming 650500, Chinab Agricultural Engineering Department, Faculty of Agriculture, Cairo University, Cairo 12613, Egyptc Zhejiang Solar Energy Product Quality Inspection Center, Haining, Zhejiang 314416, China

A R T I C L E I N F O

Article history:

Received 7 January 2015

Received in revised form 11 July

2016

Accepted 30 July 2016

Available online 26 August 2016

A B S T R A C T

The performance of solar cooling absorption system needs further research, due to its poor

coefficient of performance (COP). Therefore, this study investigated the performance of a

23 kW solar powered single-effect lithium bromide–water (LiBr–H2O) absorption cooling system.

Furthermore, the space heating mode was also investigated and the improvement methods

were analyzed and discussed. The cooling system was driven by a parabolic trough collec-

tor of 56 m2 aperture area and used for cooling a 102 m2 meeting room. Research results

revealed that the chiller’s maximum instantaneous refrigeration coefficient (chiller effi-

ciency) could be up to 0.6. Most of the time, in sunny and clear sky days the daily solar heat

fraction ranged from 0.33 to 0.41 and the collectors field efficiency ranged from 0.35 to 0.45.

At the same time, chiller efficiency was varied from 0.25 to 0.7 and the daily cooling COP

was varied from 0.11 to 0.27, respectively.

© 2016 Elsevier Ltd and IIR. All rights reserved.

Keywords:

Solar cooling

Single-effect absorption chiller

Lithium bromide–water

Parabolic trough solar collector

(PTC)

Cooling performance

Étude expérimentale de la performance d’un système derefroidissement solaire à absorption au bromure delithium-eau

Mots clés : Froid solaire ; Refroidisseur à absorption simple effet ; Bromure de lithium-eau ; Collecteur solaire cylindro-parabolique

(PTC) ; Performance de refroidissement

* Corresponding author. Solar Energy Research Institute, Yunnan Normal University, Kunming 650500, China. Fax: +86 871 65517266.E-mail address: lmllldy@126.com (M. Li).

http://dx.doi.org/10.1016/j.ijrefrig.2016.07.0230140-7007/© 2016 Elsevier Ltd and IIR. All rights reserved.

i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/ locate / i j re f r ig

ScienceDirect

1. Introduction

The global climate changes resulted in an increasing of energydemand for traditional air conditioning and heating systems

in buildings which consumes more fossil fuel, meanwhile, in-creasing the carbon emissions. Therefore, using solar energyfor air conditioning becomes one of the promising approachesto reduce energy consumptions and negative environmentalimpacts from buildings. There are two main types of solarcooling namely solar-thermal-driven and solar-PV-driven airconditioning technologies. Solar-PV-driven system uses a con-ventional vapor compression air conditioning cycle in whichthe electrical input is provided by solar PV panels. The solar-thermal-driven utilizes solar thermal energy to power thegenerator of an absorption refrigeration systems and it has threecooling technologies such as solar absorption refrigeration, solaradsorption refrigeration and solar ejector refrigeration (Otanicaret al., 2012; Siddiqui and Said, 2015). However, solar absorp-tion refrigeration is the most matured technologies and it hasbeen studied more extensively than other systems. The mainadvantage of the solar absorption cooling technology is thatthe coefficient of performance (COP) is higher than that of otherthermally operated cycles (Hassan and Mohamad, 2012). Fur-thermore, in a comparison between the different solar electric,solar thermal from the point of view of energy efficiency andeconomic feasibility, solar electric and thermo-mechanicalsystems appear to be more expensive than thermal sorptionsystems. Absorption and adsorption are comparable in termsthat adsorption chillers are more expensive and bulkier thanabsorption chillers. The total cost of a single-effect LiBr–H2Oabsorption system is estimated to be the lowest (Kim andInfante Ferreira, 2008). So due to relatively simple configura-tion and low requirements for heat sources, numerous studiesof solar absorption refrigeration system are using single-effect LiBr–H2O absorption chillers recently (Agyenim et al., 2010;González-Gil et al., 2011; Iranmanesh and Mehrabian, 2013;Izquierdo et al., 2014; Lamine and Said, 2014; Lizarte et al., 2013).Furthermore, double-effect LiBr–H2O absorption chiller systemcoefficients of performance are almost twice higher than thoseobtained with single-effect systems.With double-effect systemsit is possible to obtain coefficients of performance as high as1.12 at condenser temperatures of 30 °C but they need gen-erator temperatures higher than 140 °C to reach evaporatortemperatures as low as −5 °C (Domínguez-Inzunza et al., 2014;Iranmanesh and Mehrabian, 2014; Li et al., 2014a, 2014b;López-Villada et al., 2014). But there are some high require-ments for heat sources maybe more than 120 °C. So double-effect LiBr–H2O system is sophisticated to construct, highinvestment and operation costly. Nevertheless, Avanessian andAmeri reported that the CO2 emission of the single-effect systemwas respectively about 1.9 and 1.7 times higher than direct-fired and hot-water double-effect (Avanessian and Ameri, 2014).

The efficiency enhancement of the solar absorption coolingsystem components is essential to increase the COP of thewhole system (Fong et al., 2011; Said et al., 2015). The perfor-mance evaluation of a 35 kW LiBr–H2O absorption machinedriven by 72 m2 heat pipe evacuated tube collector with a gasbackup system showed that the actual average COP of thesystem was 0.33 while the maximum and minimum valueswere 0.50 and 0.17 respectively. These results were obtainedbased on the average collector efficiency and the solar frac-tion of 0.55 and at an ambient temperature of 32 °C (Ketjoy et al.,2013). Lu et al. (2013). investigated one two-phase thermo-syphon silica gel–water solar adsorption chiller and LiBr–H2O

Nomenclature

SymbolsAc collection area of PTC [m2]cp specific heat [J kg−1 K−1]COPs,av average COP of the whole system.Ei input solar radiation into PTC [W]Ei,tot total input solar radiation energy into

PTC [MJ]Ib solar beam radiation [W m−2]mF mass flow rate of water which flow

passing by PTC array [kg s−1]mF,g mass flow rate of water flow passing by

the chiller’s generator [kg]mF,c mass flow rates chilled water [kg s−1]mF,e mass flow rates of cooling water [kg s−1]mw,t water mass in the tank [kg]Pte,ins instantaneous thermal power of PTC [W]Pte,ins,s instantaneous power of heat collection

system [W]Pr the gained refrigeration quantity from

the absorption chiller system [W]Pg consumption thermal power of

generator [W]Pe absorption power from chilled water of

water evaporate [W]Pc output power into cooling water by

absorber and condenser [W]Qtc,ins the received useful solar radiation

energy [W]Qp,loss pipeline heat losses [W]Qt,loss heat losses of the hot storage tank [W]Qr total refrigeration quantity [MJ]Qh the gained heat of the system [MJ]T1 – T15 the temperatures of probes in Fig. 1 [°C]Ta corridor temperature (T15 in Fig. 1) [°C]Tr indoor (meeting room) temperatures [°C]

Greek symbolsΔT temperature difference [°C]ηte,ins the instantaneous thermal efficiencyηte,ins,s the instantaneous thermal efficiency of

heat collection systemηr thermodynamics coefficientηoel the optical end loss ratio of PTCηr,av the average refrigeration efficiencyθ incidence angle [°]

AbbreviationsCOP coefficient of performanceLiBr–H2O lithium bromide–waterPTC parabolic trough solar collector(s)

47i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

absorption chiller with new medium compound parabolic con-centrator (CPC) solar collectors. Results revealed that theefficiency of the medium temperature evacuated-tube CPC solarcollector can reach 0.5 when the hot water temperature is125 °C. The average solar COP of absorption system was 0.19(Lu et al., 2013). On the other hand, Ali et al. (2008) investi-gated the performance assessment of an integrated free coolingand solar powered single-effect LiBr–H2O absorption chiller inOberhausen (Germany). The plant includes 35.17 kW coolingabsorption chiller, 108 m2 evacuated tube collectors, a 6.8 m3

hot water tank, a 1.5 m3 cold water tank and a 134 kW coolingtower.The results illustrated that the chiller efficiency in somecooling months could be up to 70% while it was about 25%during the 5 years period of the plant operation; the monthlyaverage efficiency value of collectors varies from 34.1% up to41.8% and the five-year average value is about 28.3%. Hang et al.(2014) investigated the energy performance analysis of solarabsorption cooling system, a 23 kW double-effect absorptionchiller driven by a 54 m2 external compound parabolic con-centrator (XCPC) solar collectors. The results showed that thedaily average collector efficiency changed between 36% and 39%.The average coefficient of performance (COP) of the LiBr ab-sorption chiller was between 0.91 and 1.02 with an average valueof 1.0, and the daily solar COP was approximately at 0.374. Onother study, the performance analysis of a mini-type solar-powered absorption cooling system with a cooling capacity of8 kW, solar collector’s area of 96 m2 and a water storage tankof 3 m3 was investigated.The experimental results revealed thatthe average values of predicted mean vote and predicted per-centage of dissatisfied of the test room were 0.22 and 5.89,respectively, and the power consumption reduced by 43.5%.Meanwhile, the theoretical model predicted that the solar ra-diation intensity has a great impact on the performance of thesolar powered absorption cooling system compared to theambient temperature (Yin et al., 2013). Lazzarin et al. (1993) re-ported their field test results for a solar cooling plant and gavethe performance of a typical day in terms of energy collectedand temperature of the solar collector.They indicated that bothenergy collected and temperature of the solar collector weresignificantly varied during the day because the solar input con-stantly varies. Subsequently, the performance of the systemmay deteriorate and the solar powered refrigeration system willnot be able to work consistently (Xu et al., 2011). Venegas et al.(2011) investigated the influence of operational variables on theperformance of a solar absorption cooling system. Resultsshowed that the most important variables which effected onthe daily cooling energy produced and the daily averaged solarCOP are the amount of the collected solar energy, the wind ve-locity and wind direction. On the other hand, it was reportedthat the optimum mass flow rates of hot water passing throughthe generator on a double-effect absorption chiller with 100 tof cooling capacity and evacuated tube collector have an im-portant role on reducing the auxiliary energy (Iranmanesh andMehrabian, 2014). Subsequently, three alterative designs (heatstorage, cold storage, and refrigerant storage) for 24-hour-operating solar-powered LiBr–H2O absorption air-conditioningsystems were analyzed. The results indicate that continu-ously operating solar-powered LiBr–H2O absorption withrefrigerant storage was the most suitable alternative design fora 24-hour cooling effect (Al-Ugla et al., 2015). It was demon-

strated that, to obtain the high coefficient of performance fora small size absorption chiller which was developed from anold out-of-order commercial chiller, this chiller should be op-erated at 85 °C of hot water temperature supplied to thegenerator (Prasartkaew, 2014).Therefore, Calise et al. (2010) simu-lated different solar cooling systems to find out the appropriateoperation parameters for maximizing the COP of solar coolingsystems. However, there are many challenges remaining to befurther studied to enhance the COP of the absorption refrig-eration technology. This study investigates the experimentalperformance of a single-effect lithium bromide absorption re-frigeration system driven by PTC for air conditioning.Furthermore, it discusses and analyzes appropriate methodsfor improving the cooling performance.

2. System description

This experiment has been conducted in Kunming, China. Thecollector was a parabolic trough solar collector (PTC). Fig. 1shows the schematic diagram for the main parts of the refrig-eration experimental system which includes PTC array, hotwater storage tank (with a supplemental water tank), single-effect lithium bromide absorption chiller (TX-23), cooling towerand blower coils installed in the meeting room. Fig. 1 also il-lustrates the positions of temperature test points. Table 1 showsthe sensors of temperatures and its instruction. Fig. 2 showsthe main parts of the refrigeration experimental system. Themain measurements and instrument accuracy of the moni-toring system are shown in Table 2.

The parameters and heat efficiency values at different tem-peratures of PTC array were shown in Table 3.The heat receiver,glass–metal evacuated absorber tube, was the core compo-nent of PTC and the specifications were shown in Table 2 too.The PTC array is made up with two PTCs in series and the arrayspace between two PTCs was 5.6 m. Table 4 shows some pa-rameters of refrigeration chiller. The capacity of hot waterstorage tank was 1 m3. Table 5 shows the cooling tower pa-rameters, the same type of cooling water has been used in the

Table 1 – Instructions for the temperature probes inFig. 1.

Temperatureprobe

Measurement explain

T1 Inlet temperature of PTC1T2 Outlet temperature of PTC1T3 Inlet temperature of PTC2T4 Outlet temperature of PTC2T5 Hot water inlet temperature of chillerT6 Hot water outlet temperature of chillerT7 Chilled water inlet temperature of chillerT8 Chilled water outlet temperature of chillerT9 Cooling water inlet temperature of chillerT10 Cooling water outlet temperature of chillerT11 Outlet temperature of hot water tankT12 Inlet temperature of hot water tankT13, T14 The indoor temperature of meeting roomT15 The outdoor temperature of meeting room

48 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

cooling tower.The area of end air-conditioning room was 102 m2

and 3 Brower coils were installed on the top of the meetingroom (relevant parameters are shown in Table 6). The room isused to conference room which can accommodate more than20 persons. The height of the room is 3.5 m. And the lengthand the width are 13.6 m and 7.5 m, respectively.There are twowindows with 10 m2 transparent fiberglass in the north wall.The south wall nearby the corridor has two glass doors andthe area of each door is 3 m2. And then, the last two walls aremade up with reinforced concrete with 10 cm thickness.

3. Energy conversion analysis

Fig. 3 shows the heat flow diagram of the absorptionchiller, its main energy conversion processes are asfollows:

(1) The input solar radiation into PTC:

E I Ai b c= cosθ (1)

Fig. 1 – Scheme diagram of solar powered absorption cooling system.

49i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

where Ib, Ac and θ are the solar beam radiation, the collec-tion area and the incident angle of sun ray, respectively.(2) The useful energy transformed from the received solar

radiation Qtc,ins. According to Fig. 1,

Q m c T T T Ttc ins F p, = −( ) + −( )[ ]2 1 4 3 (2)

where mF is the mass flow rate of water flowing in the re-ceiver of PTC array, and cp is the specific heat capacity atconstant pressure.

(3) The pipeline heat loss between PTC and hot water storagetank Qp,loss,1:

Q m c T T T T T Tp loss F p, ,1 11 1 2 3 4 12= −( ) + −( ) + −( )[ ] (3)

(4) The calculation of the heat loss of hot water storage tankQt,loss is relatively complicated.The heat loss of hot water

Fig. 2 – Main parts of the solar absorption cooling system.

Table 2 – Main measurements and instruments accuracy of the monitoring system.

Instrument Model Range Accuracy Applicationscope

Maximumrelativeerror

Maximumabsolute

error

Uncertainty(B class)

Pyranometer Kipp & Zonen CMP-6 0–2000 (W/m2) ±5% 0–1000 (W/m2) ±10% ±100 W/m2 57.7348 W/m2

Thermocouples T −200 to 350 (°C) ±0.4% 0–150 (°C) ±0.93% ±1.4 °C 0.8083 °CWind speed transducer EC-9S 0–70 (m/s) ±0.4% 0–10 (m/s) ±2.8% ±0.28m/s 0.1617 m/sElectromagnetic flow

meterKROHNE OPTIFLUX5300

DN 25; 0–12(m/s)

±0.15% 0–5 (m/s) ±0.36% ±0.018m/s 0.0104 m/s

Pressure transducer YOKOGAWA EJA430E 0.14–16 (MPa) ±0.055% 0–2 (MPa) ±0.44% ±0.0088 MPa 0.0051 MPa

Table 3 – PTC array parameters.

Parameter Value

Aperture area 56 m2

Aperture width 2.5 mOrientation North–South (ψ = 0°)Length of PTC 26 mFocal distance of parabolic trough 1.1 mThe width of focal spot 5 cmPTC efficiency values at different

temperature77.5% (30 °C), 68.4%(50 °C), 57.3% (70 °C),38.4% (90 °C)

Inner diameter of metal pipe of receiver 4 cmOuter diameter of glass pipe of receiver 11 cm

Table 4 – Parameters of absorption chiller.

Parameter Value

Model TX-23Working pair Lithium bromide–

waterAmbient temperature 28 °C–36.0 °Chot water temperature at chiller inlet 50 °C–90.0 °Chot water temperature at chiller outlet 49 °C–80.0 °COutlet temperature of chilled water 10.0 °C–14.0 °CInlet temperature of chilled water 15.0 °C–22.0 °CCooling water temperature at chiller inlet 20.7 °C–27.8 °CCooling water temperature at chiller outlet 23.5 °C–29.3 °CPower dissipation 2.3 kWRefrigeration capacity 23 kWChilled water rate 4.0 m3/hHot water rate 5.7 m3/h

50 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

tank was tested in accordance with ISO 9459 andEN12976. Before the experiment, the water in the tankwas heated by PTC and the temperature was not low than60 °C. In order to ensure the temperature consistency inthe tank and avoid tank thermal stratification, an ex-ternal water loop was adopted. The water was pumpedby hot water pump out of the tank in the bottom andback into the tank in the top and then the hot water wascirculated through the external water pump. The pumpand the circulation pipes were wrapped with insula-tion material, the thickness of which was no less than3 cm. There temperature sensors was arranged in equalintervals in the vertical direction inside of the hot watertank. When the values of the sensors were close, there

are no thermal stratification in the tank. And then thecirculation pump must be shut down.The hot water tem-perature was 63.6 °C. The water was placed for 15 hourswithout any interference. And then, the circulation pumpwas turn on once again and the hot water was circu-lated until the temperatures of sensors were the samenearly. The temperature was 49.8 °C. In real experimen-tal test we can use the simple formula to calculate it asfollows:

Pm c T

tt loss avw t p

, ,,=

ΔΔ

(4)

where mw,t is the water mass in the tank, Δt is the con-tinue time of experiment, and ΔT is the temperaturedifference of water with no thermal stratification in the tankwithin the time of Δt. When, Δt→0, it can be gained the in-stantaneous heat loss power.(5) The pipeline heat loss between hot water storage tank

and chiller:

Q m c T Tp loss F g p, , ,2 11 52≈ −( ) (5)

where mF,g is the mass flow rate of water flow passing bythe chiller’s generator.(6) The heat absorbed from chilled water in the evapora-

tor of refrigeration chiller, Qp,loss,2 (this issue will beanalyzed in the next section).

(7) The output heat from absorber and condenser to coolingwater, this part of heat can be equivalent to the outputheat Qc from cooling tower.

(8) Qp,a,3 is the heat absorbed of pipeline between chiller andmeeting room which comes from the out surface ofpipeline.

(9) Qa,room is the heat absorbed by blower coils which comesfrom the air in the meeting room.

Ei

PTC array PipelineHot water Storage tank

Absorption coolingchiller

Coolingtower

The end of air-condition

Pipeline

pipeline

Qa,room

Qtc,ins

Qp,loss,1

Qt,loss

Qp,loss,2

Qp,a,3 QcQe

Sun

Fig. 3 – Heat flows diagram of the solar absorption cooling system.

Table 5 – Parameters of cooling tower, tank and pumps.

Parameter Value

Model BLT-10Wind rate 10.5 m3/hCooling water rate 10 m3/hDynamo power 0.75 kWCapacity of tank 1 m3

Insulation thickness of tank 3 cmPump of PTC 0.33 kWHot water pump 0.78 kWChilled water pump 0.78 kWCooling water pump 0.98 kW

Table 6 – Parameters of blower coil.

Parameter Value

Model EKCW800KTWind rate 1360 m3/hCooling power 7200 WInput power 130 W

51i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

4. Thermal efficiency of heat collectionsystem

4.1. Thermal efficiency of PTC array

The instantaneous thermal power and instantaneous thermalefficiency are respectively as follows:

P m c T T T Tte ins F p, = −( ) + −( )[ ]2 1 4 3 (6)

ηθte ins

te ins

b c

PI A,

,

cos= (7)

Because of the roof orientation factor the PTC array couldbe shaded by adjacent buildings after 4:30. Therefore, the ex-periments could only be conducted before 4:30.The experimentshave been also conducted under the windy weather condi-tions, the tested peak wind velocity reached 8 m s−1, which waslargely influential to the thermal efficiency of PTC array undersuch weather condition, because the focal line was frequentlymoved out of the absorber tube under such high wind velocity.

Fig. 4 gives the test results of thermal efficiency of PTC array.Obviously, when the irradiance changed from 0.80 to 0.90 kWm−2 the instantaneous thermal efficiency ranged from 0.50 to0.65.At the same time, the system instantaneous thermal powerwas changed between 24 and 27 kW. In the calculations, themass flow rate of water mF was tested as 0.602 kg s−1.

4.2. Energy losses from the pipelines

Here, the test of energy losses of the pipeline between PTCand hot water storage tank, energy losses of the pipelinebetween hot water storage tank and chiller, and energy lossesof the pipeline between chiller and meeting room wereconducted.

The heat conduction coefficient of thermal insulation layerof pipeline was between 0.037 and 0.040 W m−1 K−1 and the outsurface of the thermal insulation was wrapped with alumi-num alloy layer.The three types of pipelines heat loss are shownin Table 7. It is worth pointing out that because the tempera-ture of chilled water inside the tube was lower than that ofthe outside, the heat was absorbed from outside to inside oftubes. Therefore, the energy losses of the pipeline betweenchiller and meeting room were cooling losses, rather than heatlosses.

4.3. Heat losses of hot water storage tank

The heat loss of water storage tank was tested. At first, thetank was filled with 688 kg hot water. And 15 hours later, thetemperature of water has been decreased from 63.6 °C to49.8 °t. It can be estimated that the average heat lossespower was about 640 W. Obviously, the average powerof heat losses increased with the increase of hot watertemperature.

Fig. 4 – Test results of instantaneous efficiency and instantaneous power of PTC array.

Table 7 – Parameter of pipes and the energy losses of the pipelines.

Pipeline type Inner tubediameter

Outer tubediameter

Insulationthickness

Pipelength

Total energylosses (kW)

Note

Pipeline between PTC and hot water storage tank 4 cm 12.5 cm 4 cm 78 m 3.90–6.20 60–90 °C (heat loss)Pipeline between hot water storage tank and chiller 4 cm 12.5 cm 4 cm 18 m 0.90–1.40 60–90 °C (heat loss)Pipeline between chiller and blower coil 4 cm 10 cm 3 cm 55 m 0.27–0.55 7–16 °C (cooling loss)

52 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

4.4. Thermal efficiency of heat collection system

When the pipeline heat losses and heat losses of hot watertank are considered, the instantaneous power of heat collec-tion system can be expressed as:

P P P Ptc ins s tc ins p loss loss, , , , ,= − − t (8)

While the instantaneous thermal efficiency of heat collec-tion system can be expressed as:

ηθte ins s

tc ins p loss loss

b c

P P PI A, ,

, , ,

cos=

− − t(9)

where Pp,loss and Pt,loss are pipeline heat losses and hot water tankheat losses, respectively. Thus when the solar beam radiationchanged between 0.80 and 0.90 kW m−2, the instantaneousthermal efficiency and the instantaneous power ranged from0.32 to 0.42 and from 17 to 21 kW, respectively (Fig. 5).

5. Performance of the refrigeration system

This experiment was conducted mainly to test and investi-gate the performance of the refrigeration system. During theexperiment many parameters such as solar beam radiation,wind velocity, ambient temperature, inlet and outlet tempera-ture of PTC, hot water storage tank, heating, chilled water, andcooling water were tested.

5.1. Energy analysis of refrigeration process for coolingchiller

The thermodynamics coefficient (ηr) was considered as theeconomy evaluation which was defined as:

ηr e gP P= (10)

where Pe is the absorption power from chilled water of waterevaporates, and Pg is the consumption of heat energy.

According to Fig. 1, the consumption thermal power of gen-erator is as follows:

P m c T Tg F g p= −( ), 6 5 (11)

The absorption power from chilled water of water evapo-rates is:

P m c T Te F e p= −( ), 7 8 (12)

The output power into cooling water by absorber and con-denser is as the following:

P m c T Tc F c p= −( ), 10 9 (13)

In the three equations above, mF,g, mF,c and mF,e are mass flowrates of hot water, chilled water and cooling water, respec-tively. T5–T10 are the inlet and outlet temperatures of hotwater, chilled water and cooling water (see Fig. 1 andTable 1).

According to the analysis above, if the power consumed bycirculating water pump is ignored, the cycle refrigeration co-efficient of the chiller (chiller efficiency or COP of the chiller)can be expressed as:

ηre

g

F e p

F g p

F e

F g

PP

m c T Tm c T T

m T Tm T T

= =−( )−( )

= −( )−( )

,

,

,

,

7 8

6 5

7 8

6 5(14)

5.2. Experimental results

The cooling performance of the absorption chiller for ten daysof experiment was conducted. The experiment was con-ducted from April to May, 2014 at Yunnan Yi Tong Solar Scienceand Technology Co. Ltd (in Kunming, China).

At first the water in tank must be heated by PTC. When thetemperature of hot water in the tank reached more than 65 °C,

Fig. 5 – Thermal efficiency and power of heat collection system.

53i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

the refrigeration chiller was turned on manually. In order tounderstand the chiller performance under different operat-ing temperatures, the different water temperatures is adoptedto driven chiller and the refrigeration performance can be ob-tained and recorded. But in order to protect refrigeration chiller,the test experiment must be shut down when the hot watertemperature in the tank descended to around 40 °C. All of thecomponents are controlled manually. During experiment, thetest value of mF, mF,g and mF,c are 0.602 kg s−1, 1.36 kg s−1 and0.90 kg s−1 respectively.

The experimental results are shown Table 8 and Figs. 6–9show the changing curves of system performance param-eters along with time.The figures show the experimental resultsin two days and Table 8 gives the experimental results for re-frigeration performance during 14 days.

According to Fig. 6, during the experiment, the wind speedis high in most sunny and cloudy days and the maximum windvelocity vw reached 8 m s−1 (generally it ranged between 2 and5 m s−1), which had a huge impact on the thermal efficiencyof PTC system.

Fig. 7 shows that when the operating temperature of re-frigeration chiller (the temperature of hot water) increased from

39 °C to 80 °C, the lowest temperature of chilled water couldreach 9 °C. At the same time, the temperature of cooling waterchanged between 20 °C and 35 °C.The cooling process was dis-connected. At the beginning of the experiment, thetemperatures of hot water and chilled water decreased rapidly.After chiller operating steadily, the system performance pa-rameters changed stably along with time. The higher the inlettemperature of hot water, the lower the outlet temperature ofthe chilled water is.

Fig. 8 shows the changing curves of refrigeration coeffi-cient and refrigeration power of chiller along with time. It showsthat the refrigeration coefficient ηr increased along with therefrigeration quantity Pr increase of chiller.

Fig. 9 shows that in the first thirty minutes after refrigera-tion chiller operating, the indoor temperature of meeting roomdecreased sharply. After reaching a certain value, the indoortemperature began to gradually stabilize. Sometimes, the tem-perature fluctuated slightly.The indoor temperature increasedalong with the ambient temperature.

Table 8 shows the average value of performance coeffi-cient of the refrigeration system during each experimental day,where Ei,tot is the total incident solar radiation energy (the total

Table 8 – The average coefficient of performance for the refrigeration system.

Date (2014) Weather condition Ei,tot (MJ) Qr (MJ) ηr,av COPs,av

Apr. 4 Cloudy and windy 448 54 0.18 0.12May 10 Cloudy and windy 643 84 0.21 0.13May 11 Cloudy and windy 549 58 0.17 0.11May 13 Sunny and windy 1300 220 0.51 0.17May 14 Sunny and windy 1200 170 0.45 0.14May 15 Sunny and windy 1122 234 0.57 0.21May 16 Sunny and windy 1120 267 0.58 0.23May 17 Sunny and windy 1280 284 0.51 0.22May 18 Sunny and windy 1262 313 0.55 0.25May 19 Sunny and windy 877 288 0.59 0.23May 20 Cloudy and windy 690 158 0.51 0.21May 21 Sunny and windy 967 270 0.57 0.27May 22 Sunny and windy 1110 262 0.47 0.24May 24 Sunny and windy 1084 234 0.50 0.22

Fig. 6 – Solar beam radiation and wind velocity vs. time.

54 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

input solar radiation energy into PTC), Qr is the total refrig-eration quantity, ηr,av is the average refrigeration efficiency andCOPs,av is the average COP of the refrigeration system.The totalincident solar radiation energy Ei,tot is given as:

E A I dti tot c b oel St

t

s bt

s st

, cos,

,= −( )∫ θ η1 (15)

where ts,bt and ts,st are the time of experiment starting andfinishing, respectively, ηoel is the optical end loss ratio ofPTC (Xu et al., 2014). The total refrigeration quantity is givenas:

Q m c T T dtr F e pt

t

r bt

r st= −( )∫ ,

,

,

7 8 (16)

where tr,bt and tr,st are the times of refrigeration chiller operat-ing and stopping, respectively. The average refrigerationefficiency expressed as:

ηr avr

F g pt

t

Q

m c T T dtr bt

r st,

,,

,=

−( )∫ 5 6(17)

COPs,av is shown as:

COPQE

s avr

i tot,

,

= (18)

Table 8 illustrates that the average refrigeration efficiency(ηr,av) and the average COP (COPs,av) of the whole system were

Fig. 7 – The inlet and outlet temperatures of heating, chilled and cooling of chiller vs. time.

Fig. 8 – The refrigeration coefficient and refrigeration power of the absorption cooling chiller.

55i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

relatively low under cloudy weather conditions. In such weathercondition, the solar radiation was relatively small. So the hotwater could not be heated to high temperature. In addition,when the sun is covered by cloud, the pump was running withfixed frequency all the time which resulted in more pipelineheat losses.Therefore, the average refrigeration efficiency andthe average COP of the system were relatively low. In con-trast, under clear weather conditions, the refrigeration efficiencyand system COP were high. The experiment results revealedthat the average refrigeration efficiency of chiller changedbetween 0.17 and 0.60, and the refrigeration coefficient of thewhole system ranged from 0.11 to 0.27.

5.3. Space heating performance

The solar refrigeration system located at the tropical areasallows the refrigeration applications to be undertaken all theyear around. However, in high latitudes with four obviousseasons of the year, we need to maximize the utilization of thesystem by conducting the cooling mode in summer and heatingmode in winter. Thereby the solar energy utilization effi-ciency will be improved.

The energy conversion process of this system was rela-tively simple in the heating mode compared to the coolingmode, and the gained heat of the system can be expressed asEq. (8).

In the heating mode, the refrigeration unit (cooling chiller)acted as the heat exchanger. The output heat power from thecooling chiller can be expressed as Eq. (12).

The system heating performance are tested with experi-ment and the results are shown in Table 9 and the systemcharacteristic parameter variation curves are given in Figs. 10and 11. Fig. 10 gives the variations of the running tempera-ture of the absorption chiller, where T7 and T8 are the inletand outlet temperatures of the heat exchanger, respectivelyand the measurement positions of T5 to T8 are shown inFig. 1.

Fig. 11 gives the changes of indoor temperatures, where, Tr1,Tr2 and Tr3 are the indoor temperatures and Ta (T15 in Fig. 1) isthe corridor temperature. It shows that the indoor tempera-tures increased very quickly at the first 10 minutes of heatsupply process.

Table 9 gives the heat collected by PTC under differentweather conditions. It shows that the average gained heat rate

Fig. 9 – The variation of indoor temperatures during refrigerating.

Table 9 – The gained heat of system in the heatingmode.

Date(2014)

Weathercondition

Ei,tot

(MJ)Qh

(MJ)Average gained

heat rate

Mar. 2 Sunny and windy 1032 412 0.399Mar. 29 Cloudy and windy 423 164 0.387May 12 Cloudy and windy 197 81 0.411

Fig. 10 – The running temperatures of absorption chillerwhen it was under the heating mode (Mar. 29).

56 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

was about 0.4. In the practical applications, the total gainedheat of the system can be estimated as:

Q Eh te ins s i tot= η , , , (19)

5.4. Improvement analysis and discussion

The experimental results revealed that the operating tem-perature and cooling temperature had a huge influence on thecooling performance of cooling system.

Fig. 12 shows the changes of refrigeration coefficient at dif-ferent running temperatures. The refrigeration quantity wasincreasing along with the increase of the hot water tempera-ture. At the same time, the refrigeration coefficient was alsosubstantially increasing with the increase of the hot water tem-

perature, where T5 is the hot water temperature at inlet asshown in Fig. 1 and Table 1.

In absorption refrigeration system, the function of coolingwater is to exclude the absorption heat in the absorber and con-densation heat in the condenser.Therefore, the intensity degreeof absorption and condensation is related to temperature dif-ference between inlet and outlet of cooling water in condenserand absorber. Fig. 13 shows the cooling performance influenceby cooling water at different water temperatures, where, T9 andT10 are the inlet and outlet temperatures of cooling water, re-spectively (see Fig. 1 and Table 1). Because the cooling pump wasworking in intermittent, the temperature of cooling water wasdisplayed in the shape of fluctuating with time. It can be alsoseen that when the cooling water temperatures reached themaximum value, the temperature difference between inlet andoutlet of cooling water was about 1.4 °C, while when the coolingwater temperature was down to the minimum value, the tem-perature difference between inlet and outlet of cooling waterwas about 3 °C.Therefore, when the inlet temperature of coolingwater was 21 °C, the chiller refrigeration coefficient was about0.68, when the inlet temperature of cooling water was 27.6 °Cthe refrigeration coefficient dropped to about 0.39. The refrig-eration coefficient of the system is very close to that of thereference (Ali et al., 2008). And the operation model was similarto that mentioned in Prasartkaew (2014). In other words, in orderto increase the refrigeration efficiency and refrigeration quan-tity, the temperature of cooling water should be as low aspossible.

In order to increase the comprehensive efficiency of thewhole system (including refrigeration efficiency and heatingefficiency), a certain operation conditions should be con-ducted as follows.

First, the area of PTC should be increased. Subsequently, tem-perature of the hot water and the quantity of heat storage willbe improved.Thus, the cooling chiller can be started in a shortperiod of time and operated at a capacity of 23 kW for a longrunning time.

Fig. 11 – The variations of room temperatures during theheating mode (Mar. 29).

Fig. 12 – Instantaneous refrigeration coefficient vs. running temperatures.

57i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

Second, the energy losses of each part of energy conver-sion materials should be reduced as far as possible, such aspipeline heat loss, hot water tank heat loss and so on. Accord-ing to Table 6 the heat losses of system pipeline were relativelyhigh; the main reason is that the length of pipelines amongPTC, hot water tank and cooling chiller was too large. There-fore, to reduce the pipelines heat losses, the length of thesepipelines should be reduced as possible. In order to tackle theheat losses problem, every part of the heat transfer pipelinesshould be shortened. According to the experimental measure-ments, the length of pipeline between PTC and hot water tankcan be reduced to 50 m; the pipeline between hot water tankand cooling chiller can be shortened to 5 m. Thus, the heatlosses between PTC and hot water tank will be approxi-mately decreased from 3.90–6.2 kW to 1.40–2.30 kW. In addition,the heat losses between hot water tank and the cooling chillerwill be also reduced from 0.90–1.40 kW to 0.25–0.40 kW. Sub-sequently, the gained heat of the whole system will be increasedby 3.15–4.90 kW and the average COP of the whole system willbe more than 0.3. Because the system was located in a windyarea, in order to improve system performance, system designoptimization work must be carried out. One of the ways is thatboth of hot water storage tank and the cooling chiller shouldbe installed in a small room so as to minimize the convec-tion heat loss of them.

Third, to increase the chiller refrigeration quantity Qr andthe COP of the whole cycle, the temperature of cooling watershould be as low as possible. So the way of continuous coolingshould be used. According to Fig. 13, when continuous coolingis adopted, the refrigeration efficiency immediately raised to0.7. On the contrary, in intermittent cooling model, the averagerefrigeration efficiency was only about 0.55.

6. Conclusion

The cooling performance of the single-effect LiBr–H2O absorp-tion chiller driven by PTC has been investigated. The results

revealed that the chiller’s average refrigeration coefficient ηr,av

was between 0.18 and 0.60, and the average COP of the wholerefrigeration cycle COPs,av was from 0.11 to 0.27 under differ-ent weather conditions. Moreover, the refrigeration quantityof the refrigeration chiller has been increased with the in-creasing of heat temperature. Meanwhile the refrigerationcoefficient also substantially increased with the rising of hotwater temperature.The cooling water temperature could largelyinfluence on the performance of the refrigeration chiller.

The values of ηte,ins,s and Pte,ins,s were relatively low, due to thepipeline was too long and its heat loss was too high. In addi-tion, the values of ηr,av and COPs,av were also relatively lowbecause the hot water temperature was too low. Therefore, inorder to improve the cooling performance of the existing systemthe length of heat transfer pipelines should be shorten to reducethe heat losses. Furthermore, area of PTC and the operatingtemperature of the aperture should be increased and the watertemperature for cooling system should be as low as possible.

Acknowledgements

The present study was supported by National Natural ScienceFoundation, China (Grant No.: U1137605).

R E F E R E N C E S

Agyenim, F., Knight, I., Rhodes, M., 2010. Design and experimentaltesting of the performance of an outdoor LiBr-H2O solarthermal absorption cooling system with a cold store. Sol.Energy 84, 735–744.

Al-Ugla, A.A., El-Shaarawi, M.A.I., Said, S.A.M., 2015. Alternativedesigns for a 24-hours operating solar-powered LiBr-waterabsorption air-conditioning technology. Int. J. Refrigeration 53,90–100.

Ali, A.H.H., Noeres, P., Pollerberg, C., 2008. Performanceassessment of an integrated free cooling and solar poweredsingle-effect lithium bromide-water absorption chiller. Sol.Energy 82, 1021–1030.

Fig. 13 – Cooling performance influent by cooling water temperature.

58 i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9

Avanessian, T., Ameri, M., 2014. Energy, exergy, and economicanalysis of single and double effect LiBr-H2O absorptionchillers. Energy Build. 73, 26–36.

Calise, F., Palombo, A., Vanoli, L., 2010. Maximization of primaryenergy savings of solar heating and cooling systems bytransient simulations and computer design of experiments.Appl. Energy 87, 524–540.

Domínguez-Inzunza, L.A., Hernández-Magallanes, J.A., Sandoval-Reyes, M., Rivera, W., 2014. Comparison of the performance ofsingle-effect, half-effect, double-effect in series and inverseand triple-effect absorption cooling systems operating withthe NH3–LiNO3 mixture. Appl. Therm. Eng. 66, 612–620.

Fong, K.F., Chow, T.T., Lee, C.K., Lin, Z., Chan, L.S., 2011. Solarhybrid cooling system for high-tech offices in subtropicalclimate—Radiant cooling by absorption refrigeration anddesiccant dehumidification. Energy Convers. Manag. 52, 2883–2894.

González-Gil, A., Izquierdo, M., Marcos, J.D., Palacios, E., 2011.Experimental evaluation of a direct air-cooled lithiumbromide–water absorption prototype for solar airconditioning. Appl. Therm. Eng. 31, 3358–3368.

Hang, Y., Qu, M., Winston, R., Jiang, L., Widyolar, B., Poiry, H., 2014.Experimental based energy performance analysis and lifecycle assessment for solar absorption cooling system atUniversity of Californian, Merced. Energy Build. 82, 746–757.

Hassan, H.Z., Mohamad, A.A., 2012. A review on solar coldproduction through absorption technology. Renew. Sustain.Energy Rev. 16, 5331–5348.

Iranmanesh, A., Mehrabian, M.A., 2013. Dynamic simulation of asingle-effect LiBr-H2O absorption refrigeration cycleconsidering the effects of thermal masses. Energy Build. 60,47–59.

Iranmanesh, A., Mehrabian, M.A., 2014. Optimization of a lithiumbromide–water solar absorption cooling system withevacuated tube collectors using the genetic algorithm. EnergyBuild. 85, 427–435.

Izquierdo, M., González-Gil, A., Palacios, E., 2014. Solar-poweredsingle-and double-effect directly air-cooled LiBr-H2Oabsorption prototype built as a single unit. Appl. Energy 130,7–19.

Ketjoy, N., Yongphayoon, R., Mansiri, K., 2013. Performanceevaluation of 35 kW LiBr-H2O solar absorption cooling systemin Thailand. Energy Procedia 34, 198–210.

Kim, D.S., Infante Ferreira, C.A., 2008. Solar refrigeration options-a state-of-the-art review. Int. J. Refrigeration 31, 3–15.

Lamine, C.M., Said, Z., 2014. Energy analysis of single effectabsorption chiller (LiBr-H2O) in an industrial manufacturing ofdetergent. Energy Procedia 50, 105–112.

Lazzarin, R.M., Romagnoni, P., Casasola, L., 1993. Two years ofoperation of a large solar cooling plant. Int. J. Refrigeration 16,185–190.

Li, Z., Ye, X., Liu, J., 2014a. Performance analysis of solar air cooleddouble effect LiBr-H2O absorption cooling system insubtropical city. Energy Convers. Manag. 85, 302–312.

Li, Z., Ye, X., Liu, J., 2014b. Optimal temperature of collector forsolar double effect LiBr/H2O absorption cooling system insubtropical city based on a year round meteorological data.Appl. Therm. Eng. 69, 19–28.

Lizarte, R., Izquierdo, M., Marcos, J.D., Palacios, E., 2013.Experimental comparison of two solar-driven air-cooled LiBr-H2O absorption chillers: indirect versus direct air-cooledsystem. Energy Build. 62, 323–334.

López-Villada, J., Ayou, D.S., Bruno, J.C., Coronas, A., 2014.Modelling, simulation and analysis of solar absorption power-cooling systems. Int. J. Refrigeration 39, 125–136.

Lu, Z.S., Wang, R.Z., Xia, Z.Z., Lu, X.R., Yang, C.B., Ma, Y.C., et al.,2013. Study of a novel solar adsorption cooling system and asolar absorption cooling system with new CPC collectors.Renew Energy 50, 299–306.

Otanicar, T., Taylor, R.A., Phelan, P.E., 2012. Prospects for solarcooling – an economic and environmental assessment. Sol.Energy 86, 1287–1299.

Prasartkaew, B., 2014. Performance test of a small size LiBr-H2Oabsorption chiller. Energy Procedia 56, 487–497.

Said, S.A.M., El-Shaarawi, M.A.I., Siddiqui, M.U., 2015. Analysis ofa solar powered absorption system. Energy Convers. Manag.97, 243–252.

Siddiqui, M.U., Said, S.A.M., 2015. A review of solar poweredabsorption systems. Renew. Sustain. Energy Rev. 42, 93–115.

Venegas, M., Rodríguez-Hidalgo, M.C., Salgado, R., Lecuona, A.,Rodríguez, P., Gutiérrez, G., 2011. Experimental diagnosis ofthe influence of operational variables on the performanceof a solar absorption cooling system. Appl. Energy 88, 1447–1454.

Xu, S.M., Huang, X.D., Du, R., 2011. An investigation of the solarpowered absorption refrigeration system with advancedenergy storage technology. Sol. Energy 85, 1794–1804.

Yin, Y.L., Zhai, X.Q., Wang, R.Z., 2013. Experimental investigationand performance analysis of a mini-type solar absorptioncooling system. Appl. Therm. Eng. 59, 267–277.

59i n t e rna t i ona l j o u rna l o f r e f r i g e r a t i on 7 1 ( 2 0 1 6 ) 4 6 – 5 9