Energy performance comparison of three innovative HVAC systems for renovation through dynamic simulation

http://www.diva-portal.org

This is the published version of a paper published in Energy and Buildings.

Citation for the original published paper (version of record):

Gustafsson, M., Dermentzis, G., Myhren, J A., Bales, C., Ochs, F. et al. (2014)

Energy performance comparison of three innovative HVAC systems for renovation through dynamic simulation.

Energy and Buildings, 82: 512-519

http://dx.doi.org/10.1016/j.enbuild.2014.07.059

Access to the published version may require subscription.

N.B. When citing this work, cite the original published paper.

Permanent link to this version:

http://urn.kb.se/resolve?urn=urn:nbn:se:du-15163

ContentslistsavailableatScienceDirect

Energy and Buildings

jo u r n al h om ep age :w w w . e l s e v i e r . c o m / l o c a t e / e n b u i l d

Energy performance comparison of three innovative HVAC systems for renovation through dynamic simulation

Marcus Gustafsson

, Georgios Dermentzis

, Jonn Are Myhren

, Chris Bales

, Fabian Ochs

, Sture Holmberg

, Wolfgang Feist

aEnergyandEnvironmentalTechnology,Buildingtechnology,HögskolanDalarna,79188Falun,Sweden

bUnitforEnergyEfficientBuildings,UniversityofInnsbruck,Technikerstraße13,A-6020Innsbruck,Austria

cBuildingtechnology,HögskolanDalarna,79188Falun,Sweden

dFluidandClimateTechnology,DepartmentofCivilandArchitecturalEngineering,KTHRoyalInstituteofTechnology,SchoolofArchitectureandtheBuilt Environment,Brinellvägen23,10044Stockholm,Sweden

ePassiveHouseInstituteRheinstr.44/46,D-64283Darmstadt,Germany

a r t i c l e i n f o

Articlehistory:

Received27March2014

Receivedinrevisedform23July2014 Accepted26July2014

Availableonline4August2014

Keywords:

Energyperformance Dynamicsimulation HVAC

Microheatpump Heatrecovery Ventilationradiator TRNSYS

MATLABSimulink Renovation

a b s t r a c t

Inthispaper,dynamicsimulationwasusedtocomparetheenergyperformanceofthreeinnovative HVACsystems:(A)mechanicalventilationwithheatrecovery(MVHR)andmicroheatpump,(B)exhaust ventilationwithexhaustair-to-waterheatpumpandventilationradiators,and(C)exhaustventilation withair-to-waterheatpumpandventilationradiators,toareferencesystem:(D)exhaustventilationwith air-to-waterheatpumpandpanelradiators.SystemAwasmodelledinMATLABSimulinkandsystemsB andCinTRNSYS17.Thereferencesystemwasmodelledinbothtools,forcomparisonbetweenthetwo.

AllsystemsweretestedwithamodelofarenovatedsinglefamilyhouseforvaryingU-values,climates, infiltrationandventilationrates.

ItwasfoundthatAwasthebestsystemforlowerheatingdemand,whileforhigherheatingdemand systemBwouldbepreferable.SystemCwasbetterthanthereferencesystem,butnotasgoodasAorB.

Thedifferenceinenergyconsumptionofthereferencesystemwaslessthan2kWh/(m²a)between SimulinkandTRNSYS.Thiscouldbeexplainedbythedifferentwaysofhandlingsolargains,butalsoby thefactthattheTRNSYSsystemssuppliedslightlymorethantheidealheatingdemand.

©2014ElsevierB.V.Allrightsreserved.

1. Introduction

About40%ofthetotalenergyuseintheEU-27isaccountedfor bythebuildingsector.Thus,thebuildingstockplaysanimportant partintheworktowardstheinternationalgoalsoflowerenergyuse [1].Two-thirdoftheenergyusedinhouseholdsintheEU-15goesto spaceheating[2],andthelargestpotentialforsavingenergyinthis sectorliesinrenovationandupgradingofoldbuildingstomodern energystandards[3].Suchrenovationmeasuresincludechanging windows,insulatingroofsandexternalwallsandchangingHVAC systems.Thelatterofthesewastheobjectiveofthisstudy.

Manystudieshavepreviouslybeenconductedwithinthefield ofHVACsystemsandenergyuseofbuildings,bothresidentialand commercialbuildings.Boji ´cetal.[4]comparedthreeHVACsys- temsfor heating and coolingof an office building. Wanget al.

[5]madeacomparisonofthreeHVACsystemsforahypothetical

∗ Correspondingauthor.Tel.:+4623778563.

E-mailaddress:mgu@du.se(M.Gustafsson).

apartment building. The study included 17 climate zones with various temperatureand humidity conditions, and thesystems comparedwereadirectexpansionsplitsystem,asplitair-source heatpumpsystemandaclosed-loopwater-sourceheatpumpwith boileranevaporativefluidcooler.

AstudycarriedoutbyGustafssonetal.[6],treatingtwoHVAC systemssimilartosystemsAandBofthisstudy,indicatedthat thesesystemsmayindeedbepotentialalternativestoother,more established,systems.TheresultsofthisandthestudybyWangetal.

[5]alsoconfirmthenecessitytovarytheclimaticconditionswhen comparingHVACsystems.

A complete building retrofit includes many more steps and aspectsthanthosecoveredinthisstudy.Maetal.[7]proposesasys- temicapproach,goingallthewayfromplanningtopost-evaluation.

ThepresentstudyfocusesonthechoiceofHVACsystemsaftera renovationofthebuildingenvelope.

Thereisawiderangeoftoolsforsimulationofbuildingenergy performance.In a studybyOchset al.[8], thesimulationtools MATLABSimulink[9]andTRNSYS17[10]areusedtomodelaren- ovatedmulti-familyhouse.Accordingtothisstudy,therearemajor http://dx.doi.org/10.1016/j.enbuild.2014.07.059

0378-7788/©2014ElsevierB.V.Allrightsreserved.

Nomenclature

COP coefficientofperformance

HD annualheatingdemandofbuilding(kWh/(m²a)) HVAC heating,ventilationandairconditioning MVHR mechanicalventilationwithheatrecovery U heat transfer coefficient for building parts

(W/(m²K))

Greekletters

␩ airchangerate(h⁻¹) Subscripts

inf infiltration vent ventilation

differencesbetweenthetoolsregarding themodellingofwalls, zonenodes,windowsandshading,andthetimestepofthesolver (fixed inTRNSYS, adaptive in MATLABSimulink).However,the studyalsoshowsgoodagreementinresultsbetweenthetwotools.

Inthisstudy,theenergyperformanceofthreeinnovativeheat- ing and ventilation systems was investigated through dynamic simulationandsetinrelationtoa referencesystem,basedona commonair-to-waterheatpump.ThechoiceofsystemsA,BandC wasbasedontheirpotentialsuitabilityforbuildingrenovationand ontheneedtofillagapintheresearch,whileDisanestablished typeofsystem,thussuitableasreference.Allofthetestedsystems wereimplementedinamodelofagenericsinglefamilyhouseand testedfortworenovationlevelsinsevendifferentclimates.

Thesecondobjectiveofthisstudywastocontributetothecom- parisonof differentsimulationtools. Thereference systemwas modelled in both MATLAB Simulink and TRNSYS17, toenable detectionofsystematicdifferences.

2. Methodology

2.1. Buildingmodelandboundaryconditions

Thebuildingmodelledinthisstudyisasemi-detachedsingle familyhouse,withatemperedfloorareaof78m²andavolumeof thetemperedzoneof187m³.ItwasdefinedwithintheFP7project iNSPiRe[11]asatypicalEuropeansinglefamilyhouseconstruction.

TheactualbuildingislocatedinLondon,UK,andconsistsoftwo floorsandanunheatedattic,withaninsulatedceilingbetweenthe topfloorandtheattic.Inthemodel,theatticwasexcluded,andthe ceilingofthetopfloorwastakentobetheupperlimitofthebuilding envelope.Solargainsoftheroofwerethusdisregardedandthe ceilingwasassumedtoexchangeheatdirectlytotheambientair.

Thewesternwall,adjacenttotheneighbouringhouse,wastakento beadiabatic.Thewholetemperedareawasmodelledasonezone, withstairsandintermediatefloorasinternalwalls.Simulationsin TRNSYS17comparingthesinglezonemodeltoamodelwithone zoneperfloorandonezonefortheatticshowedadifferencein heatingdemandandheatloadoflessthan3%fortheclimateof London.

Foropenwindowventilationandshading,theboundarycondi- tionsusedinthisstudywerethesameasthoseusedwithin[11], andtoalargeextentalsowithinIEASHCTask44[12].Internalgains fromoccupantsandelectricalequipmentwerebasedonthesame scheduleasin[11]and[12],butsincethelivingareaforthebuild- ingusedinthisstudywasonly78m²,comparedto140m²in[12], thenumberofoccupantswasreducedfromfourtotwoandthe gainsfromelectricalequipmentandlightingwerescaleddownby 50%.Theventilationratewastakentobe0.4h⁻¹andtheinfiltration

ratewascalculatedfromasimplifiedmodelofthebuildingenve- lopetobe0.1h⁻¹.Fortwoofthestudiedclimates,theinfluenceof airchangeratesontheHVACsystemswastested.Theinfiltration ratewasincreasedbystepsof0.1to0.2h⁻¹and0.3h⁻¹.Theventi- lationratewasbothdecreasedandincreasedbythesameamount to0.3h⁻¹to0.5h⁻¹.Whilevaryingonetheseparameters,theother onewaskeptatitsdefaultvalue.

Thedesiredindoortemperature,whichwasusedtocontrolthe heatingsystems,wassetto20^◦C.Transmissionlossestotheground weremodelledby settingthedisturbed groundtemperatureas boundarytemperatureforthegroundfloor.Thedisturbedground temperaturewasapproximatedasasine, whichwascalculated accordingtostandardISO13370[13].

Besideventilationandinfiltrationrates,thesensitivityanaly- siscomprised climaticconditions and heatingdemand.Climate dataforsevendifferentEuropeanlocationswereused,aslisted inTable1.Thechosenlocations,thesameasusediniNSPiRe[11], representcontinentalandcoastalclimatesaswellasarangeof averageambienttemperatureandrelativehumidity.Datafilesfrom Meteonorm[14],basedonlong-termmeasurements,wereusedto generateweatherdataforthesimulations.

Foreachclimate,tworenovationlevelsweredefined.EnerPHit standard(HD25)[15]andPassiveHousestandard(HD15)[16]were usedtodefinehouseswithheatingdemandsof25kWh/(m²a)and 15kWh/(m²a),respectively,assumingaheatrecoveryefficiency of85%anddisregardingcoolingdemand.Forthetestedsystems whichdidnotincludeheatrecovery,theactualheatingdemand washigher.Thedifferenceinheatingdemandwithorwithoutheat recoverywaslargerforthecolderclimates,wheretheheatrecov- eryhasalargerimpact.InsulationthicknessesandrelatedU-values werecalculatedusingthepassivehousecalculationtoolPHPP[17].

The appliedU-values foreach climateand renovation levelare listedinTable1.

2.2. Investigatedsystems

Allofthetestedsystemsweresettoprovidespaceheatingand ventilation,whiledomestichotwaterusewasleftoutofthestudy.

Thecoolingdemandwasevaluatedbymeasuringthenumberof hourswithindoortemperature above26^◦C. The comparisonof thesystemsdidnotincludeaneconomicanalysis,andpractical detailsoninstallationwerenotconsidered.Totalenergyconsump- tionincludedheatpumpcompressor,auxiliaryheater,pumpfor thespaceheatingcircuitandventilationfans.Allenergyconsumed wasthuselectricity.

ThelayoutsoftheinvestigatedHVACsystemsaredescribedin Fig.1.

2.2.1. SystemA

SystemAisbasedonamicroheatpump,incombinationwith mechanicalventilationwithheatrecovery(MVHR),withelectric radiatorsasbackupforpeakheatloads.Theheatpumpusesthe exhaustairoftheheatrecoveryunitassourceandprovidesheatto thesupplyairoftheventilationsystem.Thus,onecompactunitcan beusedforcombinedventilationandheatingorcooling(reverse operationforcooling).Freshoutdoorairflowsintotheheatrecov- eryunit,whereitisheatedwitharecoveryefficiencyofupto95%.

Itisthenfurtherheatedbythemicroheatpumpuptomaximum 52^◦C,ashighertemperaturesmaycauseodourproblems,tosupply spaceheating.

Incomparisontoanairsourceheatpump,theevaporatoruses thebenefitofslightlyhighersourcesidetemperatureandoflatent heat.Theevaporatorextractsheatfromtheairusingthelatentheat ofcondensation,andthehighersourcesidetemperatureimproves thecoefficientofperformance(COP)oftheheatpump.However, theairvolumeflowrateintheevaporator,whichisequaltotheflow

Table1

LocationsforclimaticdataandcorrespondingU-valuesusedinsimulations.

Location U-values[W/m²K]

HD25 HD15 Both

Walls Floor Roof Walls Floor Roof Windows Doors

Stockholm 0.126 0.128 0.126 0.057 0.057 0.057 0.90 0.80

Gdansk 0.150 0.153 0.150 0.074 0.075 0.075 0.90 0.80

Stuttgart 0.235 0.244 0.237 0.143 0.146 0.144 0.90 0.80

London 0.277 0.290 0.279 0.175 0.180 0.176 0.90 0.80

Lyon 0.320 0.337 0.323 0.198 0.204 0.199 0.90 0.80

Madrid 0.500 0.544 0.509 0.361 0.383 0.365 0.90 0.80

Rome 0.621 0.689 0.634 0.456 0.492 0.463 0.90 0.80

Fig.1. SchematiclayoutofstudiedHVACsystems:(A)mechanicalventilationwithheatrecovery(MVHR),microheatpumpandelectricradiators,(B)exhaustventilation withexhaustair-to-waterheatpumpandventilationradiators,(C)exhaustventilationwithair-to-waterheatpumpandventilationradiatorsand(D)exhaustventilation withair-to-waterheatpumpandpanelradiators.

rateinthecondenser,islimitedtothehygienicflowrate(toohigh flowrateleadstodryindoorair).Thus,thesourcepowerislimited.

Theheatingcapacityofthemicroheatpumpisintherangeof1kW.

Therefore,thissystemcanonlybeimplementedinflatsorsmall houseswithverylowenergydemandsuchasPassiveHouses.The advantagesofthemicroheatpumparethecompactness,givingthe possibilityofintegrationintothefac¸ade,andcostreduction[18].

InordertomodelsystemAin MATLABSimulink,theEFKOS model[19]wasused.TheEFKOSmodelwasoriginallydeveloped forair-to-waterheatpumps,butthemicroheatpumpisanexhaust air-to-airheatpump.Therefore,theinputdataforthemodelwere chosensothattheoutputdatawouldfittotheheatingtestpoints ofPassiveHouseComponentCertificateofthecompactunitAeros- martmDrexel&Weiss[20].Ameanvolumeflowrateof160m³/h

wastakenforthemeasuredtest pointsofthecertificate.Inthe studiedbuildingwithadefaultventilationrateof0.4h⁻¹andavol- umeof187m³,thevolumeflowrateis74.8m³/h.Toadapttothis, theheatingcapacitywasscaleddownwhilekeepingtheratioof heatingcapacityandvolumeflowrateconstant,assumingtheCOP tobeindependentofthevolumeflowrate.Table2showsnominal performancedataforthemicroheatpumpforinletairtempera- turestotheheatrecoveryunit,calculatedfor78%heatrecovery efficiency.

2.2.2. SystemB

InsystemB,themechanicalexhaustventilationprovidesanair- to-waterheatpumpwithairfromthelivingzone,whileatthesame timecreatingthelowpressureneededtodrivetheairflowthrough

Table2

RatedperformanceofmicroheatpumpforinletairtemperaturestotheHRCunit.

Airtemperature[^◦C]

−2 2 7

Heatingoutput[kW] 1.03 1.18 1.34

COP[dimensionless] 2.22 2.73 3.07

theventilationradiatorsintothebuilding.Theheatpumpextracts heatfromtheairanddeliversheatedwatertotheradiators.The numberofventilationradiatorswaschosenbasedonthedesired ventilationrateandtheidealairflowandpressuredropforoneven- tilationradiator[21].Theradiatorswerethensizedtocovertheheat loadofthebuildingatadistributionandreturntemperaturesof 35/30^◦C.Theheatpumpmodelwasbasedonperformancedatafor anexistingair-to-waterheatpumpforexhaustair,aspresentedin Table3[22].Thesameheatingcapacity,plusanauxiliaryheaterof 1.5kW,wasusedforalllocationsandrenovationlevels.Thewater flowratewasheldconstantatthenominallevelaccordingtotest standards[23].

Intheventilationradiators,outdoorairflowsinthroughaduct inthewallandisheatedbytheradiatorpanelsbeforeenteringthe room.Theheatoutputofaradiator,eitheroftraditionalorven- tilationtypeisproportionaltothemeantemperaturedifference betweentheradiatorsurfaceandtheairincontactwiththeheated radiatorsurfaces.Becauseofthelowersurroundingairtempera- tureofaventilationradiatorcomparedtoatraditionalradiator,it canworkwithalowersupplywatertemperature.Thedirectcon- tactwithoutdoorairgivesthesystemthequalityoffastthermal response,astheheatoutputisautomaticallyadjustedwithany changeofambientairtemperature.Ventilationradiatorshavealso beenproventoperformwellintermsofthermalcomfort,givinga stableanduniformindoorclimate[24],andthelowwatertemper- atureisbeneficialfortheperformanceoftheheatpump.Froma renovationperspective,ventilationradiatorsincombinationwith mechanicalexhaustventilationcanbeacompetitivesolution,given thatthereisalreadyawaterheatingsysteminthehouse.

2.2.3. SystemC

SystemChasthesameconfigurationassystemB,butwithareg- ularair-to-waterheatpumpwithoutheatrecoveryfromexhaust air.Theheatpumpmodelwasbasedonmanufacturerperformance dataforanexistingair-to-waterheatpump,aspresentedinTable4,

Table5

Nominalcapacityandflowrateofscaleddownair-to-waterheatpump.

HPcapacity[kW] HPwaterflowrate[kg/s]

Location HD25 HD15 HD25 HD15

Stockholm 1.90 1.70 0.121 0.108

Gdansk 1.70 1.50 0.108 0.095

Stuttgart 1.90 1.60 0.121 0.102

London 1.60 1.40 0.102 0.089

Lyon 1.70 1.40 0.108 0.089

Madrid 1.90 1.60 0.121 0.102

Rome 1.70 1.40 0.108 0.089

witha nominalcapacityof 3kWand anominalCOP of 3.27at A2/W35[25].

AssumingtheCOPtobeindependentofsize,thenominalheat- ingcapacitywasscaleddowntocovertheaverageheatloadover 24hduringthewholeyear.Thenominalwaterflowrate,asgiven byteststandards[23],wasscaledaccordingly,toallowusingthe sameperformancedata.Anauxiliaryheaterof1kWwasemployed whennecessary.InTable5,thenominalcapacityandflowratefor eachlocationandrenovationlevelarelisted.

Forthesensitivityanalysisonventilationandinfiltration,the heatpumpwassizedtofitthenewloads,asshowninTable6.

2.2.4. SystemD

TheheatpumpofthereferencesystemDisthesameasthe oneusedinsystemC,andwassizedthesameway.Thetraditional panelradiatorswereassumedtobeinplacebeforetherenova- tionandsizedtocovertheheatloadofthebuildingwithoutthe extrainsulationappliedforHD25orHD15.Designdistributionand returntemperaturesweretakentobe90/70^◦C,butintherenovated housesstudied,theactualradiatorwatertemperatureswouldbe lower,duetothelowerheatload.

2.3. Simulationtools

SystemAwasmodelledinMATLABSimulinkandsystemsBand CweremodelledinTRNSYS17,whilethereferencesystemDwas modelledinbothtools,toenablecomparisonbetweenthisandthe othersystemswhilereducingtheriskofsystematicerrors. This alsoallowedforacomparisonbetweenthetwosimulationtools.

AtimestepoffiveminuteswasusedintheTRNSYSsimulations, whileMATLABSimulinkusesanadaptivetimestep.

Table3

Ratedperformanceoftheexhaustair-to-waterheatpumpofsystemB.

Watertemperature[^◦C] Airflowrate[l/s]

30 40 50 60 70

Heatingoutput [kW]

35 1.14 1.30 1.42 1.46 1.50

45 1.15 1.24 1.30 1.35 1.37

COP[dimensionless] 35 4.46 4.76 5.12 5.24 5.43

45 3.34 3.49 3.72 3.86 3.91

Table4

Ratedperformanceoftheair-to-waterheatpumpofsystemsCandD.

Watertemperature[^◦C] Airtemperature[^◦C]

−15 −7 2 7 10 12 20 30

Heatingoutput [kW]

35 1.70 2.60 3.00^* 4.50 4.60 4.80 5.00 5.50

45 1.50 2.30 2.80 4.00 4.20 4.40 4.60 5.00

55 1.30 2.00 2.70 3.70 3.80 4.00 4.30 4.70

COP

[dimensionless]

35 1.90 2.85 3.27^** 4.64 4.83 5.00 5.27 5.70

45 1.60 2.20 2.65 3.55 3.67 3.75 4.04 4.41

55 1.16 1.70 2.12 2.74 2.87 3.10 3.30 3.45

*Nominalheatingoutput.

**NominalCOP.

Table6

Nominalcapacityandwaterflowrateofheatpumpforvaryingventilationandinfiltrationrates.

Location Parametricvariation HPcapacity[kW] HPwaterflowrate[kg/s]

HD25 HD15 HD25 HD15

Stockholm vent=0.3 1.70 1.40 0.108 0.089

vent=0.5 2.20 1.90 0.140 0.121

inf=0.2 2.20 1.90 0.140 0.121

inf=0.3 2.40 2.10 0.153 0.134

Rome vent=0.3 1.60 1.30 0.102 0.083

vent=0.5 1.80 1.50 0.115 0.095

inf=0.2 1.80 1.50 0.115 0.095

inf=0.3 1.90 1.60 0.121 0.102

InMATLABSimulinkthecomplexbuildingmodeloftheCarnot Blocksetwasused.TheheatpumpmodelinsystemDwasbased onperformancemapdata.Theheatingcapacitywasapproximated linearlydependingonthesourceinletair temperatureand the sinkoutletwatertemperature.TheCOPwasbasedonCarnotCOP and the Carnot performance factor. In the initialization of the model(pre-processing)theCarnotperformancefactorandthelin- earcoefficientsfortheheatingcapacitywerecalculatedinorderto achievethebestpossibleagreement.

InTRNSYS,theheatpumpwasmodelledusingaperformance mapwithdataonheatingcapacityandcompressorpowerfor a rangeoftestingpoints.TheheatoutputandCOPoftheheatpump werecalculatedinthemodelthroughinterpolationbetweenthese points.

TheventilationradiatormodelofsystemsBandCwasbased onanExcelmodelprovidedbyaradiatormanufacturer,whichin turnwasbasedonmeasurementsontheirownproducts[26].A linkembeddedinTRNSYSwasusedtoconnecttheExcelmodelto therestofthesystem.

2.4. Controls

Allheatingsystemsandauxiliary heaterswerecontrolledby on/offdifferentialcontrollerswithhysteresis.Thegoverningtem- peraturewas theindoorair temperature. Thesetpoint for the primaryheatingsystemswas20^◦C,withupperandlowerdead bandsof0.25K.Similarly,theauxiliaryheatershadasetpointof 19.75^◦Candallowedthetemperaturetovarybetween19.5^◦Cand 20^◦C.Thesetpointfortheauxiliaryheaterwassettoalowervalue toavoidoperationduringhourswhentheprimarysystemcould managetheheating.Theventilationwasrunningindependentlyof theheatingcontrolsignals,butinsystemsAandBtheheatpumps werebypassedwhennoheatingwasneeded.

3. Results

Fig.2showstheannualheatingsupplybyheatpumpandcom- pressorof thereference systemand Fig.3theelectrical energy consumptionofthereferencesystem,comparingMATLABSimulink andTRNSYSforalllocationsandrenovationlevels.InFig.2,the solidanddashedlinesmarktheheatingdemandwithheatrecov- eryfortheHD25andtheHD15houses,respectively.Thedifference betweenthetwotoolsexceeded5%onlyforMadridandRome, wherethedifferencewas7%and6%respectivelyforsuppliedheat- ingand11%and8%respectivelyforelectricalenergyconsumption.

Thetrendsweresimilarforbothenergystandardsofthehouse.In TRNSYS,allsystemsovershottheidealheatingdemand,whichwas definedastheheatingrequiredtokeeptheindoorairtemperature atorabove20^◦C atalltimes,by1kWh/(m²a)to2kWh/(m²a).

InSimulink,systemDfollowedtheidealheatingdemandmore closely.TheidealheatingdemandsimulatedinSimulinkwashigher thaninTRNSYSforallclimatesexceptforMadridandRome,where

Fig.2.AnnualheatingsuppliedbysystemDinMATLABSimulinkandinTRNSYS.

SolidlinemarksheatingdemandofHD25withheatrecovery;dashedlinemarks heatingdemandofHD15withheatrecovery.

Fig.3.ElectricalenergyconsumptionofsystemDinMATLABSimulinkandinTRN- SYS.

itwaslower.Theseasonalperformancefactoroftheheatpump wasaround0.1higherinSimulinkthaninTRNSYS.Intermsofgains andlossesofthehouse,somedifferenceswerenotedinabsorbed solarenergy.The solargainsin MATLABSimulinkwerearound 3kWh/(m²a)higherthaninTRNSYSfortheclimatesofMadridand Rome,whileforotherclimatesthesolargainswere2.5kWh/(m²a) to5kWh/(m²a)lowerinMATLABSimulinkthaninTRNSYS.

TherelativeenergyconsumptionofsystemsA,BandCcompared tothereferencesystemDisshowninFig.4.SystemAissetin relationtotheperformanceofsystemDinMATLABSimulink,while BandCaresetinrelationtotheTRNSYSmodelofsystemD.

SystemAhadthelowestenergyconsumptionforbothrenova- tionlevelsinallclimates.Thelargestsavingscomparedtosystem DwereseenfortheHD15incoldclimates,withamaximumof36%

fortheclimatesofStockholmandGdansk.

System B showed a similar trend, but with less difference betweenthecoldestandthewarmestclimates,andalsolessdif- ferencebetweenthetworenovationlevels.FortheHD25houseit wasclosetosystemAinenergyconsumptioninallclimates.The maximumenergysavingcomparedtosystemDwas23%forthe HD15houseinStockholm.

Fig.4.Energyconsumptionoftestedsystemscomparedtothereferencesystemfor varyingclimateandheatingdemand.

Fig.5. Energyconsumptionoftestedsystems(HD25)comparedtothereference systemforvaryingventilationrate.

Fig.6. Energyconsumptionoftestedsystems(HD25)comparedtothereference systemforvaryinginfiltrationrate.

ForsystemC,theenergyusewasconsistentlylowerthanthe reference,withonlysmallvariationswiththeclimates.Itwasthe bestsystem,togetherwithAandB,fortheHD25houseinRome.The energysavingscomparedtosystemDrangedfrom3%fortheHD25 houseinMadridandRometo8%fortheHD15houseinStockholm.

Theinfluenceofventilationrateontheenergyperformanceis showninFig.5andtheinfluenceofinfiltrationrateinFig.6,both fortheHD25house.SystemAwasaffectedinapositivedirection relativetosystemDwhentheventilationratewasincreasedand inanegativewaywhentheinfiltrationratewasincreased.Witha ventilationrateof0.5h⁻¹,theenergyperformanceofsystemAwas betterthanthereferencefortheclimateofRome.ForsystemB,both waysofincreasingtheairchangeratewerefavourablecompared tosystemD.Withaninfiltrationrateof0.2h⁻¹,systemBhadthe bestenergyperformanceforStockholm,andwithaninfiltration rateof0.3h⁻¹itwasthebestsystemalsoforRome.SystemCwas notsignificantlyinfluencedinanywayinrelationtothereference systembyeitheroftheseparameters.

Fig.7showsthenumberofhourswithroomtemperatureabove 26^◦Cforthereferencesystem.InLyon,MadridandRome,theroom

Fig.7.Numberofhourswithroomtemperatureabove26^◦Cforthereferencesys- teminMATLABSimulinkandinTRNSYS.

temperaturereachedabove26^◦C for1000hto2000hper year, withslightlyhigherfiguresfortheHD15house.Allsystemsmet thecriteriatokeepthetemperatureabove19.5^◦Catalltimes.

4. Discussion

In the comparison betweenthe two simulation tools, some deviationswereobservedfortheperformanceofsystemD.Therel- ativelyhighdifferencesinpercentageforRomeandMadridcould partlybeexplainedbythelowheatingdemandfortheselocations.

Also,inTRNSYSallthetestedsystemsprovided1kWh/(m²a)to 2kWh/(m²a)morethantheidealheatingdemand,notablywitha mutualdifferenceoflessthan1kWh/(m²a)betweenthem.There alsoseemedtobeadifferenceinhowthetwotoolshandlesolar gainsofahouse.Thisshowedinthenumberofhourswithover- heating,wherethesimulationsinMATLABSimulinkgaveahigher numberinmostcases.Itcouldalsohaveinfluencedthedifference inannualheatingdemand.

Thehousewasmodelledasasinglezone, disregardingsolar gainstotheroof.Fortheheatingdemand,particularlyforLondon andcolderclimates,thisapproachmakeslittleornodifferenceto theresultscomparedtoamodelincludingroofandattic,sincethe solargainsarerelativelysmallduringtheheatingseason.Inwarmer climatesthedifferencecouldbemoresignificant,andifthecooling demandistobedeterminedallsolargainsshouldbetakeninto account.

Heatingsystemsthatarebasedonheatrecoveryofventilation airarealwayslimitedbytheventilationrate.ForbothsystemsAand Binthisstudy,therelativelylowventilationratelimitedtheheat- ingcapacityoftherespectiveheatpumpsandincreasedtheneed forauxiliaryheating.Varyingtheventilationrate,it wasshown thattheperformanceofthesesystemssystemrelativetotheref- erencesystemimprovedwithahigherventilationrate,andvice versa.Whenitcomestovaryingtheinfiltrationrate,AandBare affectedinoppositeways.InsystemB,theexhaustfanenablesthe exhaustairheatpumptoutilizeboththeventilationandinfiltra- tionairtodeliverenergytotheventilationradiators.InsystemA, theheatrecoveryunitcanonlymakeuseoftheventilationpart, whiletheinfiltratedairjustaddstotheheatlosses.

Allsystemswerecomparedforthesameinfiltrationrate.How- ever,theinfiltrationrateofahouseisdependentonthepressure differencebetweenindoorandoutdoor,whichinturndependson thetypeofventilationsysteminstalled[27].SystemA,usingbal- ancedventilation,would havea lowerinfiltrationratethan the othersystemsforthesamehouse.

In the present study,theuse of mechanical ventilation was accountedforonlyduringthemonthswhenthehouserequired heating.InSweden,buildingregulations[28]donotallowreplac- ingmechanicalventilationwithopeningwindows.Extendingthe ventilationperiodwouldstrikethehardestonsystemA,sincethe MVHRconsumesmoreenergythanasimpleexhaustfan.Forthe

HD15houseinStockholm,applyingmechanicalventilationallyear wouldincreasethetotalenergyconsumptionofsystemAby8%, whereastheincreaseforsystemsB,CandDwouldbe2–3%forthe samecase.However,bypassingtheheatrecoveryunitduringsum- merwouldreducetheimpactonenergyconsumptionforsystem A.

ForsystemC,somesavingswereseenduetothelowerwater temperatureenabledbytheuseofventilationradiators.However, thewatertemperatureinthereferencesystemwasalreadylow, sincetheexistingradiatorsweresizedforahigherheatload.The largestreductioninenergyconsumptionwasseenforsystemA, wheretheairheatrecoverycutdowntheheatingdemandsignifi- cantlycomparedtothatofothersystems.SystemBconsumedless energythansystemCduetothehighersourcesidetemperatureof theexhaustairheatpump.

SystemAwasthesystemthatbenefittedthemostfromahigher renovationstandard.Ithadthelowestenergyconsumptionforboth renovationlevelandallclimates,butfortheHD25houseitcon- sumedalmostasmuchenergyassystemB,despitetheadvantage ofMVHR.Thisconfirmsthepremisethatthemicroheatpumpis bestappliedinverylowenergybuildingsuchasPassiveHousesand suggeststhatasystemlikeBwouldbepreferableinhouseswith higherheatingdemand.

Inacompletebuildingretrofit,itmaynotalwaysbefeasible fromtheeconomicpointofviewtoachievePassiveHousestandard.

FortheclimatesofGdanskandStockholm200-300mmofextra insulationisrequiredtogofromHD25toHD15level.Thiswillof courseincreasetheinvestmentcostssignificantly,eventhoughthe totalinsulationthicknesscouldbereducedbychoosingbetterinsu- latingwindowsandimprovetheairtightnessinsuchcoldclimates.

Ontheotherhand,insulationofthefloormaynotalwaysbefea- sible,thusincreasingtheneedforimprovementsonotherpartsof thebuildingenvelope.

Thelevelofinsulationcanalsobeimportantforthechoiceof heatingsystemintermsofthermalcomfort.Ifventilationradiators areusedinheavilyinsulatedhouses,asinsystemsBandC,there couldbeproblemswithcolddraftwhentheoutdoortemperature isatornearthebalancetemperatureofthehouse.Astheheating systemwillnotbeactiveabovethebalancetemperature,alower balancetemperaturewillallowcolderairtobesuppliedthrough theradiators.

Inwarmerclimates,athickerinsulationleadstoslightlyhigher indoortemperaturesduringsummer,thusoccasionallyincreasing thecoolingdemand.TheobservedindoortemperaturesforLyon, MadridandRomeinthisstudyindicatethatthetestedhousewould needacoolingdeviceintheseclimates;aservicewhichcouldbe providedbyreversingtheoperationoftheheatpump.

Acompleteenergysystemforahouseneedalsoincludedomes- tic hot water. Air-to-water heat pumps, like the ones used in systemsB,CandD,arenormallydesignedtohandlebothspace heatingandhotwater.SystemA,ontheotherhand,wouldrequirea complementtotheair-to-airmicroheatpumptobeabletoprovide thisservice.Inheavilyinsulatedhouses,whereheatlossesaremin- imized,the relative importanceof hot water usewillnaturally becomelarger.

TheheatpumpusedinsystemsCandDwasscaleddownfrom 3.0kWtoheatingcapacitiesrangingfrom1.2kWto2.4kW,assum- ingthattheCOPremainedthesame.Inreality,theCarnotefficiency, andthustheCOP,mightnotbeindependentofthecapacityofthe heatpump.Forascalingdownofthisrelativelysmallmagnitude,it maynothaveagreatimpact,butitshouldbetakenintoconsider- ationthatitcouldaffecttheresultofsystemsCandD,especiallyfor thewarmerclimateswheretheheatpumphasbeenscaleddown more.

TheTRNSYSheatpumpmodelusedwasnotdesignedforvari- ablespeed.TheexhaustairheatpumpofsystemBwould,inreality,

beabletovarythecompressorspeedtocopewithhigherloads,and wouldthereforeneedtouselessauxiliaryheatingthanthemodel did.ThemicroheatpumpofsystemAalsohasthepotentialto performslightly betterwithaspeedcontrolledcompressor.The influenceofcontrolstrategycouldbeasubjectforfuturestudies.

5. Conclusions

In dynamic simulation of building energy performance, the resultsaretosomeextentdependentonthechoiceofsimulation tool.ThedifferencesbetweenMATLABSimulinkandTRNSYS17 wereinthisstudyfoundtobelargerforwarmerclimates,possi- blybecauseofdifferencesinhowsolargainsaretreatedinthetwo tools.Also,theTRNSYSsystemssuppliedslightly morethanthe idealheatingdemand.Still,themagnitudesofthedeviationswere acceptable.

BothsystemsAandBweremorefavourableincolderclimates;

systemAduetotheheatrecoveryandsystemBduetothehigher sourcesidetemperatureoftheheatpump.Accordingtotheresults ofthisstudy,systemAisthebestoptioninwell-insulatedhouses withlowinfiltrationandhighventilationrate.Foralessinsulated housewithhigherinfiltrationrate,locatedinthesameclimate,sys- temBwouldhavethebestenergyperformance.Theperformance ofsystemCshowsthatsomeenergycanbesavedbyapplyingven- tilationradiatorsinsteadoftraditionalpanelradiators,althoughin thiscasethepanelradiatorsweresizedforahigherheatloadand thusalsoenabledalowwatertemperature.SystemCwasbetter thanthereferencesystem,butnotasgoodasAorB.

In future studies of retrofitted buildings, it is suggested to include the useof domestic hot water, as this will make up a largerpartofthetotalenergyconsumptionwhenthespaceheating demandisloweredthroughrenovationofthebuildingenvelope.In warmclimates,coolingdemandshouldalsobeconsidered.

Acknowledgement

Theresearchleadingtotheseresultshasreceivedfundingfrom theEuropeanUnion’sSeventhProgrammeforresearch,technolog- icaldevelopmentanddemonstrationundergrant agreementNo 314461.TheEuropeanUnionisnot liableforanyusethatmay bemadeoftheinformationcontainedinthisdocumentwhichis merelyrepresentingtheauthorsview.

References

[1]TheEuropeanParliamentandtheCounciloftheEuropeanUnion.Directive 2010/31/EUontheenergyperformanceofbuildings.OfficialJournalofthe EuropeanUnion,L153(2010)13–35.

[2]EuropeanEnvironmentAgency.EnergyandEnvironmentReport,2008.ISBN:

978-92-9167-980-5.

[3]EuropeanCommissionEnvironmentalImprovementPotentialsofResidential Buildings(IMPRO-Building).OfficeforOfficialPublicationsoftheEuropean Communities,2008.ISBN:978-92-79-09767-6.

[4]M.Boji ´c,N.Nikoli ´c,D.Nikoli ´c,J.Skerli ´c,I.Mileti ´c,Asimulationappraisalofper- formanceofdifferentHVACsystemsinanofficebuilding,EnergyandBuildings 43(2011)1207–1215.

[5]W.Wang,J.Zhang,W.Jiang,B.Liu,Energyperformancecomparisonofheating andair-conditioningsystemsformulti-familyresidentialbuildings,HVAC&R Research17(3)(2011)209–322.

[6]M.Gustafsson,J.A.Myhren,C.Bales,ComparisonoftwoHVACsolutions:acase study,in:CLIMA2013,Prague,CzechRepublic,2013.

[7]Z.Ma,P.Cooper,D.Daly,L.Ledo,Existingbuildingretrofits:methodologyand state-of-the-art,EnergyandBuildings55(2012)889–902.

[8]F.Ochs,G.Dermentzis,D.Siegele,A.Konz,W.Feist,Useofbuildingsimulation toolsforrenovationstrategies—arenovationcasestudy.PartofEuropean7th FrameworkProgrammeprojectiNSPiRe,2013.

[9]MathWorks. Simulink–Simulation and Model-Based Design.

http://www.mathworks.se/products/simulink/(accessed25.02.14).

[10]S.A.Klein,A.Beckman,W.Mitchell,A.Duffie,TRNSYS17—atransientsys- temssimulationprogram,in:SolarEnergyLaboratory,UniversityofWisconsin, Madison,2011.

[11]iNSPiRe,EuropeanCommission7thFrameworkProgrammeproject.Proposal number:314461;Title:DevelopmentofSystematicPackagesforDeepEnergy RenovationofResidentialandTertiaryBuildingsincludingEnvelopeandSys- tems;Duration:2012-10-01–2016-09-30.

[12]TheReferenceFrameworkforSystemSimulationsoftheIEASHCTask44/HPP Annex38–PartB:BuildingsandSpaceHeatLoad,2012.

[13]ISO 13370:2007–Thermalperformanceofbuildings—Heattransfer viathe ground—Calculationmethods.

[14]Meteonorm.http://meteonorm.com/(accessed25.02.14).

[15]PassiveHouseInstitute.EnerPHit:Certificationcriteriaforretrofits,2012.

[16]Passive House Institute—Passive House requirements.

http://www.passiv.de/en/02 informations/02 passive-house- requirements/02passivehouse-requirements.htm(accessed25.02.14).

[17]PHPP—The Energy Balance and Passive House Planning Tool.

http://passiv.de/en/04phpp/04phpp.htm(accessed25.02.14).

[18]F. Ochs, G. Dermentzis, W. Feist, Fac¸ade integrated active components intimber-constructions forrenovation—acasestudy,NSB,Lund, Sweden, 2014.

[19]TheReferenceFrameworkforSystemSimulationsoftheIEASHCTask44/HPP Annex38–PartC:HeatPumpModels—A5.HeatpumpmodelEFKOS,2013.

[20]PassiveHouseInstitute,PassiveHouseInstitute—PassiveHouseSuitableCom- ponentCertificate.CompactHeatPumpSystem,Drexel&WeissAerosmart m.

[21]Acticon(2008).Easy-Vent—Friskluftibostäder,26–27.(inSwedish).

[22]N—Indata,NIBE—IndatatillTMF:sprogramver2.1förNIBEF750.(inSwedish).

[23]EuropeanCommitteeforStandardization.EuropeanStandardEN14511-2—Air conditioners,liquidchillingpackagesandheatpumpswithelectricallydriven compressorsforspaceheatingandcooling—Part2:Testconditions,2007.

[24]J.A.Myhren,S.Holmberg,Designconsiderationswithventilation-radiators:

comparisonstotwo-panelradiators,EnergyandBuildings41(2009)92–100.

[25]ViessmannVitocalTechnicalGuide.Vitocal200-STypeAWB201.B,2011.

[26]M.Ivonen,2007.PurmoAirSimulatorVers.05.11.2007.

[27]SverigesCentrumförNollenergihus.FEBY12–Kravspecifikationförnollenergi- hus,passivhusochminienergihus(inSwedish),2012.

[28]Boverket.SwedishBuildingRegulations,BBR19,BFS2011:26(inSwedish).

ISBN978-91-86827-41-0,2011.