new test methodologies to analyse direct expansion solar assisted heat pumps for domestic hot water

10
New test methodologies to analyse direct expansion solar assisted heat pumps for domestic hot water Jorge Faca ˜o , Maria Joa ˜o Carvalho National Energy and Geology Laboratory (LNEG), Estrada do Pac ßo do Lumiar 22, 1649-038 Lisbon, Portugal Received 21 January 2013; received in revised form 1 November 2013; accepted 24 November 2013 Available online 20 December 2013 Communicated by: Associate Editor C. Estrada-Gasca Abstract Since there are not specific standards for testing direct expansion solar assisted heat pumps for domestic hot water, new testing meth- odologies are proposed supported by laboratory experiments. Two methodologies were developed for performance measurement: mod- ified BIN method and long term performance prediction with a TRNSYS model validated with specific experimental conditions. The long term performance prediction is a methodology similar to the already obtained for solar thermal systems. A system was tested in Lisbon during one year, covering almost all possible local weather conditions. The hot water tapping test cycle used was in agreement with recent standards EN16147:2011 or EN15316-3-1:2007. The influence of average daily air temperature, dew point temperature and solar irradiation was analysed. The seasonal performance factor was calculated for two cities in Portugal (Lisbon and Porto) and for additional four cities in Europe (Davos, Athens, Helsinki and Strasburg). The establishment of a procedure to calculate the seasonal performance of this kind of systems is very important according to the directive 2009/28/EC of the European Parliament and of the Council. Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. Keywords: Direct expansion solar assisted heat pump; Annual seasonal performance factor; Long term performance prediction 1. Introduction The energy in the actual economical context presents more and more expensive and rare. The performance of a heat pump increases with increasing the low temperature medium in the evaporator. Assisting the heat pump with solar energy promotes the global efficiency of the system. Then the principal objective in the interaction is to reduce the primary energy consumption for the same level of com- fort and to reduce the initial investment and maintenance costs. The principle of the heat pump has been known since the middle of the nineteenth century; however there was a little incentive to develop them in a time of cheap and abundant energy coming from fossil fuel (Dinc ßer and Kanoglu, 2010). The heat pumps are thermal machines which move the heat from cold source to hot sink, using a relatively small amount of high-quality drive energy. They could be classified according to different points of view. According to the thermodynamic cycle there are two main groups: vapour compression heat pumps and absorption heat pumps, although there are other cycles more specific. Electric motor or combustion engine are nor- mally used to drive the compressor in vapour compression cycle. In absorption heat pumps the mechanical work is only to drive a small hydraulic pump. Heat pumps are gen- erally classified by their respective heat sources and sinks. Depending on cooling requirements, various heat source 0038-092X/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.solener.2013.11.025 Corresponding author. Address: Solar Energy Laboratory, National Energy and Geology Laboratory (LNEG), Estrada do Pac ßo do Lumiar 22, 1649-038 Lisbon, Portugal. Tel.: +351 210924600; fax: +351 217163688. E-mail address: [email protected] (J. Faca ˜o). www.elsevier.com/locate/solener Available online at www.sciencedirect.com ScienceDirect Solar Energy 100 (2014) 66–75

Upload: maria-joao

Post on 21-Dec-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Available online at www.sciencedirect.com

www.elsevier.com/locate/solener

ScienceDirect

Solar Energy 100 (2014) 66–75

New test methodologies to analyse direct expansion solar assistedheat pumps for domestic hot water

Jorge Facao ⇑, Maria Joao Carvalho

National Energy and Geology Laboratory (LNEG), Estrada do Pac�o do Lumiar 22, 1649-038 Lisbon, Portugal

Received 21 January 2013; received in revised form 1 November 2013; accepted 24 November 2013Available online 20 December 2013

Communicated by: Associate Editor C. Estrada-Gasca

Abstract

Since there are not specific standards for testing direct expansion solar assisted heat pumps for domestic hot water, new testing meth-odologies are proposed supported by laboratory experiments. Two methodologies were developed for performance measurement: mod-ified BIN method and long term performance prediction with a TRNSYS model validated with specific experimental conditions. Thelong term performance prediction is a methodology similar to the already obtained for solar thermal systems. A system was tested inLisbon during one year, covering almost all possible local weather conditions. The hot water tapping test cycle used was in agreementwith recent standards EN16147:2011 or EN15316-3-1:2007. The influence of average daily air temperature, dew point temperature andsolar irradiation was analysed. The seasonal performance factor was calculated for two cities in Portugal (Lisbon and Porto) and foradditional four cities in Europe (Davos, Athens, Helsinki and Strasburg). The establishment of a procedure to calculate the seasonalperformance of this kind of systems is very important according to the directive 2009/28/EC of the European Parliament and of theCouncil.Crown Copyright � 2013 Published by Elsevier Ltd. All rights reserved.

Keywords: Direct expansion solar assisted heat pump; Annual seasonal performance factor; Long term performance prediction

1. Introduction

The energy in the actual economical context presentsmore and more expensive and rare. The performance of aheat pump increases with increasing the low temperaturemedium in the evaporator. Assisting the heat pump withsolar energy promotes the global efficiency of the system.Then the principal objective in the interaction is to reducethe primary energy consumption for the same level of com-fort and to reduce the initial investment and maintenancecosts.

0038-092X/$ - see front matter Crown Copyright � 2013 Published by Elsevie

http://dx.doi.org/10.1016/j.solener.2013.11.025

⇑ Corresponding author. Address: Solar Energy Laboratory, NationalEnergy and Geology Laboratory (LNEG), Estrada do Pac�o do Lumiar 22,1649-038 Lisbon, Portugal. Tel.: +351 210924600; fax: +351 217163688.

E-mail address: [email protected] (J. Facao).

The principle of the heat pump has been known sincethe middle of the nineteenth century; however there wasa little incentive to develop them in a time of cheap andabundant energy coming from fossil fuel (Dinc�er andKanoglu, 2010). The heat pumps are thermal machineswhich move the heat from cold source to hot sink, usinga relatively small amount of high-quality drive energy.They could be classified according to different points ofview. According to the thermodynamic cycle there aretwo main groups: vapour compression heat pumps andabsorption heat pumps, although there are other cyclesmore specific. Electric motor or combustion engine are nor-mally used to drive the compressor in vapour compressioncycle. In absorption heat pumps the mechanical work isonly to drive a small hydraulic pump. Heat pumps are gen-erally classified by their respective heat sources and sinks.Depending on cooling requirements, various heat source

r Ltd. All rights reserved.

Nomenclature

A aperture area of collector, m2

bu collector efficient coefficient (wind dependence),s/m

b2 heat loss coefficient at (Te � Ta)=0, W/(m2 K)c2 collector efficiency coefficient, Ws/(m3 K)E total tapping daily energy, kWhEL long wave irradiance, W/m2

G00 net irradiance, W/m2

H solar irradiation, MJ/(m2 day)hj FrequencyKh incident angle modifierPs saturation water vapour pressure, PaPER primary energy ratio_Qevap useful power extracted from evaporator collec-

tor, W_Qcondensation condensation power, W_Qcond useful heat power in condenser, WSPF seasonal performance factorT temperature, �CTa ambient air temperature, �CTair average ambient air temperature, �CTair,int Indoor ambient temperature, �C

Tc refrigerant condensing temperature, �CTdp atmospheric dew point temperature, �CTe refrigerant evaporating temperature, �CTwater tank storage water temperature, �CUAcomp compressor overall heat transfer coefficient-heat

transfer area product, W/�CUAcond condenser overall heat transfer coefficient-heat

transfer area product, W/�Cuwind wind velocity, m/s_W compressor power consumption, W

Greek symbols

b tilted angle of the collector with respect to hor-izontal, degrees

r Stefan–Boltzmann constant, W/(m2 K4)es sky emissivityg ratio between total gross production of electric-

ity and primary energy consumption for electric-ity production

go collector optical efficiencyhL longitudinal angle of incidence, �hT transversal angle of incidence, �

J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75 67

and heat sink arrangements are possible in practical appli-cations. The six basic source and sink pairs of heat pumpare as follows:

� air to air,� air to water,� water to water,� water to air,� ground to air, and� ground to water.

A direct expansion solar assisted heat pump DX-SAHPfor domestic hot water preparation is composed of anunglazed absorber, domestic hot water store, a thermo-block, which comprises the electrically powered compres-sor, the thermostatic expansion valve, the auxiliaryheating element and the controller. Inside the absorber cir-culates the heat pump refrigerant as heat transfer fluid. Inthe system the solar absorber absorbs energy from solarirradiation to evaporate the refrigerant. While the refriger-ant is passing through the absorber it collects energy fromthe surroundings and turns into vapour. The evaporatedrefrigerant enters into the compressor that raises the pres-sure. In the condenser, which is integrated as an immersedheat exchanger in the lower part of the store, the refriger-ant condenses and transfers its latent heat to the domesticwater in the store. Before the refrigerant returns to theabsorber, a thermostatic expansion valve reduces its pres-sure. As emergency heater an electrical heating element is

located in the lower part of the store at the height of theheat exchanger. The compressor is controlled by a thermo-stat inserted inside the store.

There are in literature several analytical and experimen-tal studies about DX-SAHP for preparation of domestichot water (Morgan, 1920; Chaturvedi and Abazeri, 1987;Chaturvedi and Shen, 1984; Anderson et al., 2002; Ander-son and Morrison, 2007; Axaopoulos et al., 1998; Ayeet al., 2002; Guoying et al., 2006; Scarpa et al., 2011; Itoet al., 1999; Kuang and Wang, 2006; Chow et al., 2010;Li et al., 2007; Huang and Lee, 2003; Chyng et al., 2003;Huang and Lee, 2007; Mette et al., 2009).

Chaturvedi and Shen (1984) have concluded that thedirect expansion solar assisted heat pump has a better per-formance when compared with a classical air source heatpump. They found heat pump COP’s in the range 2–3and solar collector efficiency according to enthalpy differ-ence from 40 to 70%.

Anderson and Morrison (2007) and Anderson et al.(2002), have published several studies related to DX-SAHPand Morrison (1994) has developed a specific Type (com-ponent library) for TRNSYS. The numerical model pre-sented here is based on the Morrison’s model, but withhot water tapping test cycle and COP definition accordingthe new standards (EN 16147:2011; EN 15316-3-1:2007).

Axaopoulos et al. (1998) have compared the perfor-mance of a DX-SAHP with a thermosyphon solar systemand the heat pump has presented a better behaviour. Theheat pump COP was above 3. However the analysis is

Fig. 1. Direct expansion solar evaporator (left) and storage tank withcompressor (right).

68 J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75

not a comparison of energy consumption but of tempera-ture reached in storage tank.

Aye et al. (2002) have compared for the main Austra-lians cities three technologies to produce DHW: solar ther-mal system, air source heat pump and DX-SAHP. Thecomparison was in terms of electrical energy consumption,life cycle cost and reduction of greenhouse gas emissions.The right choice to minimize the life cycle cost dependsheavily in climatic conditions and local price of electricity.

Guoying et al. (2006) have presented a numerical simu-lation of DX-SAHP. They have compared the energy con-sumption of an electrical boiler, solar thermosyphonsystem with electrical backup and DX-SAHP operatingin Nanjing, China. The DX-SAHP has presented the bestperformance.

Recently Scarpa et al. (2011) have compared the perfor-mance of solar thermal systems and DX-SAHP, both sys-tems with gas boiler as a backup. It was a theoreticalanalysis and they have used a variable capacity compressorVCC. The heat pump presents higher energy savings whencompared with the solar thermal system.

There are several papers where the analysis ofDX-SAHP is without taping cycle in thermal storage(Chow et al., 2010; Li et al., 2007; Kuang et al., 2003;Huang and Lee, 2003; Chyng et al., 2003; Huang andLee, 2007). In these cases it is necessary to measure themass flow rate of refrigerant for COP calculation.

In the existing standards to test heat pumps for domestichot water, there is the European Standard EN16147: 2011.However the standard does not allow solar assistance inevaporator and the test is made for only one ambienttemperature.

The standard ISO 9459-4:2012 based on CTSS method(Component Testing – System Simulation) could beapplicable to test the system. However the methodologyconsumes a significant time to test each component and itrequires several laboratory instrumentation andrequirements.

We propose here the development of an “open source”

test approach for the determination of the thermal perfor-mance of these systems, without any intrusive measure-ments, similar to the Dynamic System Test (DST)method applied to solar thermal systems (ISO 9459-5:2007). Mette et al. (2009) have presented an analysis basedin DST method, however the methodology considers thesystem as a “black box” and the model for the system isnot well set up.

The taping cycle in thermal storage and SPF (seasonalperformance factor) calculation is according to EN16147:2011. The outside air temperature was variablecontrarily to the standard and there is an influence of solarirradiance in evaporator, which is not the case in EN16147:2011. With test facility used it was impossible to con-trol the inlet water temperature always closer to 10 �C, arequirement of EN 16147:2011. Nevertheless in tappingcycles the energy calculation takes the inlet water tempera-ture variation into account. It was also impossible to

control rigorously the ambient temperature where thethermal storage is located around 20 �C as cited by thestandard.

The analysis presented here is for a system with a heatstorage tank of 300 L, evaporator of 1.6 m2 and set-pointof 50 �C. According to EN 16147:2011 the tapping cyclefor the test is selected by manufacturer. Tapping test cycleXL was tested but the system was unable to achieve anaverage outlet temperature of 40 �C at 21:30 (tapping testtype bath). The system satisfied always the requirementsof the hot water tapping test cycle L, and then it was thetapping cycle adopted in the analysis. The heat pump ther-mal performance with hot water tapping test cycle M isworst then with cycle L.

In Section 2 the system in analysis is classified accordingTask 44, Section 3 presents the experimental analysis donefor climatic conditions in Lisbon during one year and Sec-tions 4 and 5 present the two methodologies to calculatethe seasonal performance factor: BIN method and long-term prediction term. Section 6 presents the principal con-clusions of the paper.

2. System classification

The system in analysis is presented in Fig. 1. Table 1 pre-sents the main dimensions of the system.

The association of solar energy with heat pumps coulduse several heat sources for the heat pump and/or morethan one heat sink for the solar collector. Within the Task44 “Solar and Heat Pump Systems” of IEA’s Solar Heatingand cooling Programme that has been started in the begin-ning of 2010, combine solar thermal and heat pump sys-tems have been analyzed in detail. Elimar et al. (2010)have proposed a systematic classification of such systemsin order to gain an overview of the possible alternativeson the one hand and to provide a basis for considerationof the respective pros and cons from technical point of viewon the other hand. Fig. 2 presents the classification of DX-SAHP according to Elimar et al. publication. Basically, thevisualisation presents the energy flow in the system. Energyflows, that have to be purchased, like electricity or naturalgas, are shown at the system boundary to the left, useful

Table 1Principal characteristics of the system.

Rotary compressor nominal power 455 WFluid R134aTank nominal volume 300 LTank effective volume 274 LAero-solar evaporator surface 0.8 � 2 = 1.6 m2Set-point 50 �CDT differential 5 �C

Fig. 3. Daily SPF as a function of daily average outside air temperature.

J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75 69

energy flows like domestic hot water to the right. Environ-mental energy sources such ambient air, are shown at theupper part. The components, solar collector and heatpump are introduced in centre of the chart. The energyflows connecting components are represented by lines. Onthe right side of Fig. 2 is presented the notation schemeaccording to same publication. The sources and sinks ofthe solar collector (S) and heat pump (HP) are indicatedwith superscripts (sources) and subscripts (sinks) aroundthe abbreviation SHP.

Fig. 4. Daily SPF as a function of daily average dew point temperature.

3. Experimental analysis

The system was tested in Lisbon during one year cover-ing almost all local possible weather conditions with theevaporator outdoor. During the whole year it was impossi-ble to respect always the inlet water temperature near 10 �Cand indoor temperature of 20 �C as recommend by stan-dard EN 16147: 2011. The uncertainties of measurementare in agreement with same standard. The performanceof the system is presented in Figs. 3–5 as a function of out-door conditions: daily average ambient temperature, dailyaverage dew point temperature and daily solar irradiation.Fig. 6 includes daily average air temperature and irradia-tion dependence. The thermal performance of the heatpump presents a correlated dependence with daily averageambient temperature and with daily average dew pointtemperature. For daily solar irradiation is difficult to deter-mine a dependency of SPF with this parameter.

Fig. 7 presents the thermal energy extracted in storagetank according to tapping cycle L. The big extractionsoccur in the morning – 7:00–8:00 – and in the evening –

Fig. 2. DX-SAHP classification with energy flo

20:00–21:00. This fact forces the compressor to work inthese periods. Fig. 8 presents the ON–OFF-cycles of thecompressor and solar irradiance in summer and winter.The compressor is working mainly in periods with low irra-diance. Therefore the influence of solar irradiation in sys-tem performance was smaller than as expected and this isdue to the profile of taping cycle recommended.

w chart (left) and notation scheme (right).

Fig. 5. Daily SPF as a function of daily solar irradiation.

Fig. 6. Daily SPF as a function of daily solar irradiation and average dailyair temperature.

Fig. 7. Thermal energy extracted in tapping cycle L.

70 J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75

4. BIN methodology

The seasonal performance factor for heat pumps fordomestic hot water preparation is calculated for fixed con-ditions of ambient air temperature. The European standardEN14825, 2012 presents a methodology to calculate theSPF – Seasonal Performance Factor for heat pumps tospace heating for several part loads, i.e. for several coldsource temperatures. The method used was the BIN

method and it uses the hourly average evaporator ambientair temperature as an input.

A new methodology based in BIN method is proposedhere, which uses as input the daily average evaporatorambient air temperature and daily average SPFs fromexperimental analysis. As presented in last section, the heatpump has presented a good relationship with daily averageair temperature.

The annual performance factor was calculated from Eq.(1).

SPF ¼Xn

j¼1hj� E

. Xn

j¼1hj� E=SPFj

� �ð1Þ

where E is the total tapping thermal daily energy and h thedaily average temperature frequency.

The results for Lisbon are presented in Table 2. The SPFused was according to the fitted curve obtained in experi-mental analysis (see Fig. 3). Since the system was testedfor daily average ambient temperature in the range of 6–30 �C, some, points were extrapolated by the fitting, forothers cities. Below �3 �C the fitting gives a SPF less than1. In this cases SPF was considered equal to 1, like in aclassical electrical heater.

5. Long-term performance prediction

The calculation of long term performance predictionbased on a modified BIN method and on experimentalmeasurements of SPF correlated with ambient air temper-ature, as shown above, may be influenced by the fact thatthe experimental SPF was measured using an inlet watertemperature that was not constant and equal to 10 �Cand an indoor temperature (storage tank ambient temper-ature) that was not always 20 �C, as recommended by thestandard EN 16147: 2011.

An alternative way to the determination of long termperformance prediction is the development of a numericalmodel to characterize the system and specific tests to vali-date the model, i.e., determine its characteristic parameters.This approach is identical to the one used for solar systemcharacterisation according to ISO 9459-5: 2007.

With a model the system has not to be tested for all pos-sible climatic conditions. It has only to be tested in specificconditions to validate the model. A numerical model wasdeveloped in TRNSYS (2005) for the system and validatedfor specific test sequences. With this model it was possibleto calculate the long term performance prediction and esti-mate the thermal behaviour for climatic conditions differ-ent from the location where tests were performed.

5.1. The model

The model was developed in TRNSYS environment andit is based on Morrison’s work (Morrison, 1994). The typeused to simulate the storage tank was type 4 instead of type38 used by Morrison. The equation for heat transfer due tocondensation was also different. Since there is not an

Fig. 8. On–off-cycles of the compressor and solar irradiance in winter and summer.

Table 2BIN table for Lisbon with the daily average outside air temperature as aninput.

Bin (T) Frequency (hj) Cumulative (%) SPF = 0.051Tair + 1.1091

7 2 0.55 1.58 3 1.37 1.59 7 3.29 1.6

10 10 6.03 1.611 18 10.96 1.712 22 16.99 1.713 28 24.66 1.814 31 33.15 1.815 28 40.82 1.916 21 46.58 1.917 20 52.05 2.018 22 58.08 2.019 23 64.38 2.120 28 72.05 2.121 26 79.18 2.222 24 85.75 2.223 16 90.14 2.324 11 93.15 2.325 11 96.16 2.426 4 97.26 2.427 6 98.90 2.528 2 99.45 2.529 1 99.73 2.630 1 100.00 2.6

Annual SPF 1.96

Table 3Incidence angle modifier.

hT / hL 50� 10� 20� 30� 40� 60� 70�

Kh (hT) 1.00 1.00 1.00 1.00 1.00 0.95 0.91Kh (hL) 1.00 1.00 1.00 1.00 1.00 0.95 0.91

J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75 71

available type for the system, a new type was implementedin FORTRAN, compiled and linked to work in TRNSYS.

The heat pump evaporator was simulated as an uncov-ered solar collector but with an influence of condensationwhen its temperature is below the dew point temperature.

The heat flux received by the evaporator was calculatedby the Eq. (2).

_Qevap ¼ _Qcondensation þ G00Agoð1� buuwindÞKh

� A ðb2 þ c2uwindÞðT e � T aÞ½ � ð2Þ

The incidence angle modifier (Kh) was considered from so-lar keymark certificate and presented in Table 3. The samesolar collector was tested with water as a working fluid in adifferent laboratory and certified by DIN CERTCO.

The condensation effect was introduced when the evap-orator temperature was below dew point temperature(Morrison, 1994).

_Qcondensation¼ 0:0163Aðb2þC2uwindÞðP sðT eÞ�P sðT dpÞÞif T e� T dp

_Qcondensation¼ 0 if T e > T dp

ð3Þ

The temperature dependent saturated vapour pressure ofwater can be calculated by (Vajen et al., 2003):

P sðT Þ ¼ 102exp 19:016� 4064:95

T þ 263:25

� �ð4Þ

The net irradiance takes into account the relative long waveirradiance.

G00 ¼ Gþ 0:85ðEL � rðT a þ 273:15Þ4Þ ð5Þ

EL ¼ esrðT a þ 273:15Þ4ð1þ cosbÞ=2 ð6Þ

Fig. 9. Procedure flowchart adopted.

72 J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75

The sky emissivity can be quantified in terms of atmo-spheric dew point temperature

es ¼ 0:711þ 0:56T dp=100þ 0:73ðT dp=100Þ2 ð7Þ

The work power absorbed by the compressor and the use-ful heat liberated in the condenser heat pump is a functionof evaporator and condenser temperature, and availablefrom compressor manufacturer data. The curves from themanufacturer were fitted by second order polynomialregarding evaporator temperature and first order regardingcondenser temperature.

_Qcond ¼ aþ bT s þ cT 2s þ dT c þ eT sT c þ fT 2

s T c ð8Þ

_W comp ¼ a1 þ b1T s þ c1T 2s þ d1T c þ e1T sT c þ f1T 2

s T c ð9Þ

The heat in condenser inside the storage tank is also a func-tion of the global heat transfer coefficient and the temper-ature differential between the refrigerant temperature andwater temperature in the storage tank.

_Qcond ¼ UAcondðT c � T waterÞ ð10Þ

The first law of thermodynamics applied of the systemgives the following energy balance.

_Qcond ¼ _W comp þ _Qevap � UAcompðT c � T air;intÞ ð11Þ

For system simulation and considering the last equationwe have to solve a system of non-linear equations. Theunknown variables are: _Qcond , _W comp, _Qevap, Tc and Ts.The Type developed in FORTRAN for TRNSYS environ-ment solves the system of non-linear equations by Newton–Raphson method with sub-relaxation factors. Fig. 9 pre-sents the flowchart of the new Type.

Since the system does not work for exterior air temper-ature below �5 �C, the domestic hot water preparation isassured in these cases by an electrical resistance with apower of 1000 W. To simulate the storage tank a modifiedType 4 was adopted with 11 nodes with the thermostat innode 10 and the heat exchanger in node 11. The heat losscoefficient was considered a function of storage tank ambi-ent temperature. The tapping load was implemented inTRNSYS in agreement with the European standard EN16147, 2011.

5.2. Model experimental validation

The model implemented presents several parametersthat must to be set. The dimensional parameters are knownand imposed in the model. To validate the model four spe-cific days have been chosen: cold day with high irradiation,cold day with low irradiation, warm day with high irradia-tion and warm day with low irradiation. An optimizationprocess was developed varying parameter values and per-forming an expressive number of simulations for the spe-cific days that allowed the determination of global heattransfer coefficients:

� condenser heat transfer coefficient immersed insidethe storage tank: 70 W/�C;

� heat loss coefficient of the compressor: 12 W/�C;� storage tank heat loss coefficient as a function of ambi-

ent temperature: 24.564–0.386Tair,int(kJ/hr m2 K);� optical efficiency go of 80%;� b2 of 32 W/(m2K);� bu of 0 (s/m);� c2 of 6.4 Ws/(m3K).

It must be noted that some parameter values led todivergence solution or solutions without physical sense.The values presented here gave the best solution and theyhave been predicted with a vast number of simulationsand also with the help of GenOpt (2004) subroutines.

Table 4 presents experimental validation of the modelfor four specific days. The maximum relative error in objec-tive function, electrical compressor energy consumption,was about 5.0%. It was a reasonable value and similar tothe error achieved by Mette et al. (2009) for an analysisof similar system.

5.3. Long term performance prediction

Using the model and characteristic parameters deter-mined by simulation and experimental validation it is

Table 4Model experimental validation.

Day Tair (�C) G (MJ/m2) Tair,int (�C) Exp. comp.energy (kWh)

Model. comp.energy (kWh)

Relativeerror (%)

Cold/high irradiation 3 February 2012 6.2 22.9 10.2 8.49 8.91 5.0Cold/low irradiation 25 January 2012 14.7 7.4 12.8 7.58 7.42 2.1Warm/high irradiation 24 June 2012 30.3 22.2 23.2 4.69 4.73 0.9Warm/low irradiation 14 June 2012 21.3 18.0 21.3 5.58 5.55 0.6

Fig. 10. Average monthly SPF variation evaluated with LTPP.

J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75 73

possible to determine the long term performance predictionbased in a simulation for several climatic conditions. Atime step of 5 s was used for the annual simulation. Thedaily tapping load is the same for every day of the year.

Fig. 10 presents the average monthly SPF variation eval-uated by LTPP for two cities in Portugal (Lisbon andPorto) and four cities in Europe (Davos, Athens, Helsinkiand Strasburg). The monthly SPF variation during the yearin warm cities was small. Cold cities present maximum rel-ative monthly SPF variation of 64% in Helsinki and 62% inDavos.

Table 5 summarizes the results and compares the LTPPwith BIN method. In general LTPP presents higher annualperformance factor. It must be noted that the SPF fittingused in modified BIN method was obtained with experi-mental tests without the required conditions, i.e. indoortemperature of 20 �C and inlet water temperature of10 �C. In LTPP based on the numerical model developedthe system was simulated with the required conditions.The primary energy ratio PER could be easily calculatedas a function of g the ratio between total gross productionof electricity and the primary energy consumption for

Table 5Annual performance factor evaluated by LTPP and BIN method.

City Annual averageambient temperature (�C)

Annual SPF

LTPPAnnual SPFBIN method

Lisbon 16.8 2.09 1.96Porto 14.5 2.06 1.84Athens 18.4 2.10 2.00Davos 2.8 1.66 1.27Helsinki 5.6 1.70 1.36Strasburg 11.0 1.95 1.61

74 J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75

electricity production, PER = gSPF. According the Euro-pean directive, 2009/28/EC only heat pumps with PER of1.15 should be considered as renewable energy equipment.

6. Conclusions

The experimental analysis done for one year in Lisbongave a SPF with a good fitting with daily average air tem-perature where the evaporator is located. The irradiationinfluence was difficult to be correlated. With tapping cycleadopted (L) the compressor works mostly in periods withlow irradiance. This fact justifies the irradiation non-influ-ence in the heat pump performance. A new methodologybased in BIN method was implemented to characterizedirect expansion solar assisted heat pumps for domestichot water. The input in BIN method was the daily averageair temperature instead of hourly air temperature as used inspace heating applications. A simulation model was imple-mented in TRNSYS and validated against experimentalresults. The model is important to estimate the long termperformance prediction and to simulate the thermal behav-iour for conditions that were not tested in laboratory.Comparing BIN method with LTPP method the last onehas presented higher annual seasonal performance factor.With LTPP method the inlet water temperature and indoorair temperature were always respected the standard recom-mendations. The results of BIN method for low average airtemperature were extrapolated and were not validated inreal operation. The direct influence of solar irradiancewas taken into account in LTPP, but not in the usedBIN method. In BIN method the experimental analysiscould present more time consuming and could be moreexpensive than LTPP.

Acknowledgements

The authors wish to thank Fundac�ao para a Ciencia e aTecnologia under FCT PTDC/ENR/70844/2006 researchproject for the test facility used and the IEA-SHC Task44 participants for their discussion inputs.

References

Anderson, T.N., Morrison, G.L., 2007. Effect of load pattern on solar-boosted heat pump water heater performance. Solar Energy 81, 1386–1395.

Anderson, T.N., Morrison, G.L., M. Behnia, M., 2002. Experimentalanalysis of a solar-boosted Heat pump Water heater with integralcondenser. In: Proceedings of Solar 2002. Australian New ZealandSolar Energy Society.

Axaopoulos, P., Panagakis, P., Kyritsis, S., 1998. Experimental compar-ison of a solar-assisted heat pump vs. a conventional thermosyphonsolar system. Intl. J. Energy Res. 22, 1107–1120.

Aye, Lu, Charters, W.W.S, Chaichana, C., 2002. Solar heat pump systemsfor domestic hot water. Solar Energy 73 (3), 169–175.

Chaturvedi, S.K., Abazeri, M., 1987. Transient simulation of a capacity-modulated. Direct-expansion, solar-assisted heat pump, solar energy39 (5), 421–428.

Chaturvedi, Sushil K., Shen, James.Y., 1984. Thermal performance of adirect expansion solar-assisted heat pump. Solar Energy 33 (2), 155–162.

Chow, T.T., Pei, G., Fong, K.F., Lin, Z., Chan, A.L.S., He, M., 2010.Modeling and application of direct-expansion solar-assisted heat pumpfor water heating in subtropical Hong Kong. Appl. Energy 87, 643–649.

Chyng, J.P., Lee, C.P., Huang, B.J., 2003. Performance analysis of a solar-assisted heat pump water heater. Solar Energy 74, 33–44.

Dinc�er, Ibrahum, Kanoglu, Mehmet, 2010. Refrigeration Systems andApplications. second ed. Wiley.

Directive 2009/28/EC of the European Parliament and of the Council,Official Journal of the European Union, 2009.

Frank Elimar, Michel Haller, Sebastian Herkel, Jorn Ruschenburg, 2010.Systematic Classification of Combined Solar Thermal and Heat PumpSystems, EuroSun 2010 International Conference on Solar Heating,Cooling and Buildings, 28 September–1 October 2010, Graz, Austria.

EN 15316–3-1, 2007. Heating systems in buildings – method forcalculation of system energy requirements and system efficiencies –Part 3. 1 Domestic hot water systems, characterisation of needs(tapping requirements), CEN.

EN 16147:2011. Heat pumps with electrically driven compressors – testingand requirements for marking of domestic hot water units.

EN14825: 2012. Air conditioners, liquid chilling, packages and heatpumps, with electrically driven compressor, for space heating andcooling – testing and rating at part load conditions and calculation ofseasonal performance.

GenOpt, 2004. A Generic Optimization Program, User Manual Version2.0.0, Ernest Orlando Lawrence Berkeley National Laboratory.

Guoying, Xu, Xiaosong, Zhang, Shiming, Deng, 2006. A simulation studyon the operating performance of a solar-air source heat pump waterheater. Appl. Thermal Eng. 26, 1257–1265.

Huang, B.J., Lee, C.P., 2003. Long-term performance of solar-assistedheat pump water heater. Renew. Energy 29, 633–639.

Huang, B.J., Lee, C.P., 2007. Performance evaluation method of solar-assisted heat pump water heater. Appl. Thermal Eng. 27, 568–575.

ISO 9459–4, 2012. Solar heating – domestic water heating systems – part 4:System performance characterization by means of component testsand computer simulation.

ISO 9459–5, 2007. Solar heating — Domestic water heating systems —Part 5: System performance characterization by means of whole-system tests and computer simulation, first edition.

Ito, S., Miura, N., Wang, K., 1999. Performance of a heat pump usingdirect expansion solar collectors. Solar Energy 65 (3), 189–196.

Kuang, Y.H., Wang, R.Z., 2006. Performance of a multi-functional direct-expansion solar assisted heat pump system. Solar Energy 80, 795–803.

Kuang, Y.H., Sumathy, K., Wang, R.Z., 2003. Study on a direct-expansion solar-assisted heat pump water heating system. Intl. J.Energy Res. 27, 531–548.

Li, Y.W., Wang, R.Z., Wu, J.Y., Xu, Y.X., 2007. Experimental perfor-mance analysis on a direct-expansion solar-assisted heat pump waterheater. Appl. Thermal Eng. 27, 2858–2868.

Mette, B., Druck, H., Bachmann, S., Muller-Steinhagen, H., 2009.Performance Testing of Solar Thermal Systems Combined With HeatPumps. ISES Solar World Congress 2009, Johannesburg South Africa(October 11–14).

J. Facao, M.J. Carvalho / Solar Energy 100 (2014) 66–75 75

Morgan, R.G., 1920. Solar assisted heat pump. Solar Energy 28 (2), 129–135.

Morrison, G.L., 1994. Simulation of packaged solar heat-pump waterheaters. Solar Energy 53 (3), 249–257.

Scarpa, F., Tagliafico, L.A., Tagliafico, G., 2011. Integrated solar-assistedheat pumps for water heating coupled to gas burners; control criteriafor dynamic operation. Appl. Thermal Eng. 31, 59–68.

TRNSYS, 2005. A Transient Simulation Program. User’s Manual,Version 16, Solar Energy Laboratory, University of Wisconsin-Madison, 2005.

Klaus Vajen, Marcel Kramer, Elimar Frank, 2003. Extended ModelEquation for uncovered collectors, ISES Solar World Congress, 14–19June 2003, Gothenburg.