ContentslistsavailableatSciVerseScienceDirect
EnergyandBuildings
journalhomepage:www.elsevier.com/locate/enbuild
Review
Quantitativeenergyperformanceassessmentmethodsforexistingbuildings
ShengweiWang∗,ChengchuYan,FuXiao
DepartmentofBuildingServicesEngineering,ThePolytechnicUniversity,Kowloon,
article
info
abstract
Articlehistory:
Received14May2012
Receivedinrevisedform21August2012Accepted22August2012
Keywords:
Energyquantification
EnergyperformanceassessmentExistingbuildings
Buildingenergyperformanceassessmentiscrucialtoascertaintheefficiencyofenergyuseinbuildingsandisthebasistomakeanydecisionforenhancingenergyefficiency.Inordertoassesstheenergyperformanceofexistingbuildingsquantitatively,theenergyuseoftheassessedbuildingsshouldbequantifiedfirst.Thequantifiedenergyusewillbethenusedtocomparewiththeassessmentcriteriatodeterminetheenergyperformancequantitatively.Thispaperpresentsanoverallreviewonthestateoftheartoftheresearchandapplicationsofquantitativeenergyperformanceassessment.Aframeworkisproposedforcategorizingtheenergyquantificationmethodsandperformancebenchmarkingmethodsforenergyperformanceassessmentforexistingbuildings.Energyquantificationmethodsareclassifiedintothreecategories,i.e.thecalculation-based,measurement-basedandhybridmethods,accordingtotheenergydataacquisitionapproaches.Energyperformanceassessmentmethodsareclassifiedaccordingtotheassessmentscopeanddepthofassessment,i.e.whole-buildingbenchmarkingmethodatbuildinglevelandmulti-levelassessmentmethod.
©2012ElsevierB.V.Allrightsreserved.
Contents1.2.
3.
4.
Introduction.........................................................................................................................................Anoverviewofenergyperformanceassessment...................................................................................................2.1.Backgroundofenergyperformanceassessment............................................................................................2.2.Theobjectivesofbuildingenergyperformanceassessment................................................................................
2.2.1.Energyperformanceassessmentforclassification................................................................................2.2.2.Energyperformanceassessmentfordiagnosis....................................................................................
2.3.Energyperformanceassessmentfornewbuildingsandexistingbuildings................................................................2.4.Theapplicationsofenergyperformanceassessment.......................................................................................
2.4.1.Buildingenvironmentassessmentschemes.......................................................................................2.4.2.Energycertification................................................................................................................2.4.3.Whole-buildingbenchmarking....................................................................................................2.4.4.Hierarchicalassessmentanddiagnosistools......................................................................................
Energyquantificationmethods.....................................................................................................................3.1.Calculation-basedquantification............................................................................................................
3.1.1.Dynamicsimulationforenergycalculation.......................................................................................3.1.2.Steady-statemethodsforenergycalculation.....................................................................................
3.2.Measurement-basedquantification.........................................................................................................
3.2.1.Energybill-basedmethods........................................................................................................3.2.2.Monitoring-basedmethods........................................................................................................
3.3.Hybridquantificationmethod...............................................................................................................
3.3.1.Calibratedsimulation..............................................................................................................3.3.2.Dynamicinversemodels...........................................................................................................
Energyperformanceassessmentapproaches.......................................................................................................4.1.Whole-buildingbenchmarking..............................................................................................................
873
873873873873874874874874876876876876876877878880880881881881882883883
∗Correspondingauthor.Tel.:+85227665858;fax:+85227746146.E-mailaddress:beswwang@polyu.edu.hk(S.Wang).
0378-7788/$–seefrontmatter©2012ElsevierB.V.Allrightsreserved.http://dx.doi.org/10.1016/j.enbuild.2012.08.037
874
S.Wangetal./EnergyandBuildings55(2012)873–888
5.
4.1.1.Statisticalbenchmark..............................................................................................................4.1.2.Calculatedbenchmark.............................................................................................................
4.2.Hierarchicalassessmentanddiagnosis.....................................................................................................Discussionandconclusions.........................................................................................................................References...........................................................................................................................................
883884884885886
1.Introduction
Thebuildingsectoristhelargestenergyconsumerexceedingindustryandtransportationsectors.Currently,theglobalenergyusecontributedbybuildingsisabout40%[1],ofwhichasignificantproportionmightbewastedduetovariousfaultsinbuildingdesign,constructionandparticularlyinoperationstages[2,3].Eliminatingbuildingenergyproblemsandenhancingbuildingenergyefficiencyisapromisingapproachtorelievetheincreasinglyaggravatedenergyproblems.
Buildingenergyperformanceassessmentiscrucialtoascer-taintheefficiencyofenergyuseandisthebasistomakeanydecisionforenhancingenergyefficiency.Itcanprovidebuildingownersand/ortenantswithunderstandableinformationregardinghowmuchenergybeingconsumedandhowtheperformancebeingappraisedcomparingwithbenchmarks,whichconsequentlyshouldbeamotivationforthebuildingownerstoimprovetheenergyperformancewhentheperformanceisdeficientandcertaincost-effectiveenergysavingopportunitiesbeidentified.Thisini-tiativeprovidesbenefitsofreducingoperationsandmaintenancecostsfromthereductionofenergyconsumption.ItalsofacilitatestheCO2emissionfromthebuildingsector.
Whentheenergysavingandenergyefficiencyissuesarecon-cerned,essentialquestionsconfrontedbymostprofessionalsandresearchersarehowtodefineandhowtoassesstheenergyper-formanceofbuildings?Thetermofenergyefficiencyisdefinedas“usinglessenergywithoutcompromisingtheperformanceofthebuilding[4]”andwhichcanbeachievedbyreducingunnec-essaryenergyuseinbuildings.Thequestionishowmuchenergyshouldbeconsideredastheminimalenergyinputinordertoprovidetherequiredservicestoabuilding?Inotherwords,howtoassessthebuildingperformancereferringtoenergyaspectandjudgewhetheranyunnecessaryenergyuseoccursinabuildingisanessentialquestiontobeanswered.However,thereisnounanimousagreementonthisfundamentalquestionsofar.Therefore,thispapersummarizestheresearcheffortsandapplicationsconcerningbuildingenergyperformanceassessmentparticularlyforexistingbuildingsandfocusesontwocriticalissues:(1)themethodologiestoquantifyenergyuseinexistingbuildings,and(2)themethodstoassessenergyperformanceforexistingbuildings.
2.Anoverviewofenergyperformanceassessment
2.1.Backgroundofenergyperformanceassessment
Energyperformanceisatermtoindicatethequalityofabuild-inginenergyuse[5,6].Energyperformanceindicators(EPI)arequantifiablemeasurestoassessenergyperformance.ThemostcommonlyusedEPIformanybuildingtypesisenergyuseinten-sities(EUI),i.e.kWh/m2.Buildingenergyperformanceismainlydeterminedbysixfactors:(1)climate,(2)buildingenvelop,(3)buildingservicesandenergysystems,(4)buildingoperationandmaintenance,(5)occupants’activitiesandbehaviorand(6)indoorenvironmentalqualityprovided,assummarizedinIEAAnnex53project[7].
Theenergyperformanceassessmentapproachesinbuild-ingsectorcanbeclassifiedintotwomajorcategories,namelyperformance-basedandfeature-specificapproaches.Using
performance-basedapproach,assessmentresultsareobtainedbycomparingtheperformanceindicators(e.g.EUIorCO2emission)againstestablishedbenchmarks.Whileusingfeature-specificapproaches,creditsareawardedwhencriteriaofspecifiedfeaturesaremet.Thefinalscorewillbegradedaccordingtothetotalawardedcreditsofallitemsassessed[8].
Useoftheperformance-basedapproachtoassesstheenergyperformanceofabuildingismuchmorepreciseandoftenprefer-ableasitisbasedonthequantifiableperformanceindicators.However,thedevelopmentoftheassessmentmethodsandtheassessmentproceduresismoredifficult,whichinvolvestheestab-lishmentofanappropriatemethodforquantificationoftheenergyperformanceandtheappropriatecriteriatojudgetheperformanceoftheassessedbuildings.Mostcurrentresearchandapplicationsofenergyperformanceassessmentmethodsbelongtosuchquan-titativeperformance-basedapproach.
2.2.Theobjectivesofbuildingenergyperformanceassessment
Energyperformanceassessmentschemesandmethodsareestablishedmainlyfortwopurposes:energyclassificationandenergyperformancediagnosis.Energyclassificationprovidesuni-formorauthorizedmeanstocommunicateabuilding’srelativeenergyefficiencyandcarbonemissionstoboththeownersandthepublictoencourageongoingefficiencyandconservationgains.Energyperformancediagnosisaimsatdetectingfaultsanddiagnos-ingthecausesofpoorperformanceinbuildings,andaccordinglyprovidingspecificenergyefficientmeasurestoimproveenergyperformance.
2.2.1.Energyperformanceassessmentforclassification
Energyclassificationisaninformationinstrument,whichpro-videsbuildingownersorpublicswithinformation,regardingtheenergyperformanceoftheassessedbuildings.Suchinformationisusuallyexpressedinaverypracticalandunderstandableforms(1–100,orA–M,orpoor–excellent),whichencouragesthebet-terperformancewithhigheracknowledgementandmotivatesbuildingownerstoimprovetheenergyperformance.Severaltyp-icalenergyclassificationinstrumentshaveemergedinpractice,includingenergybenchmarking,energyrating,energylabelingandenergycertification.Eachofthemhasitsuniquenessinclassify-ingthequalityanddisplayingthelevelofenergyperformancewhilesometimestheyhaveoverlappingmeaningsandevencanbereplacedbyeachother[9].
2.2.1.1.Energybenchmarking.Itisoftenreferredaswholebuildingbenchmarking.Itusesasimplemethodtoinformdecisionmakerswitharelativeenergyperformancelevelbycomparingthewhole-buildingenergyperformanceindexoftheassessedbuildingwithpre-setbenchmarks.
2.2.1.2.Energyrating.Comparedwithotherclassificationinstru-ments,theuniquefeatureoftheenergyratingisitsstringentaddressinginenergyquantificationmethods.EuropeanStandardEN15603:2008[10]proposestwoprincipaltypesofenergyrating,i.e.the“calculatedenergyrating”andthe“measuredenergyrating”(or“operationalrating”).Calculatedratingsaresubdividedinto“standardrating”(orassetrating)and“tailoredrating”,according
S.Wangetal./EnergyandBuildings55(2012)873–888
875
towhetherthecalculationconditions(includingoccupancy,operationalschedules,andindoor–outdoorclimateconditions)usestandarddatasettingoractualdata.
2.2.1.3.Energylabeling.Energylabeling,assigninganenergyper-formanceclassorlabeltothebuilding,requiresthedevelopmentofascalerelatedtoalabelingindex(LI)[9].Thelabelingscaledeterminesthepercentileintervals(bands)toenergyclasses,forinstance,top10%forA,orawardingcreditsaccordingtoenergyreductionpercentages.Itaddresseshowtodisplaytheassessmentresultswithdistinctivelevels,comparingwithreferencedperfor-mance.Itisworthnoticingthatsettingalabelingscaleisaquitesubjectiveprocesswhichismorelikelytobeapolicydecisionratherthanatechnicalanalysis.
2.2.1.4.Energycertification.Energycertificationisaproceduretoassessenergyperformanceandtoproduceanenergycertificatebyanauthorizedinstituteorperson.Acertificationprogramgen-erallyincludesan“energyrating”processtoquantifyenergyuse,anauthorized“energylabeling”scaletoclassifythecorrespond-ingperformance,aswellasaminimumrequirementtoeliminatetheunacceptableperformance.InEurope,buildingenergyperfor-mancecertificationhasbecomecompulsory.Fornewbuildingsaswellasexistingbuildingssubjecttomajorrenovation,energyper-formancecertificateshavetobemadeavailabletotheownersortothetenantswhenbuildingsareconstructed,soldorrentedout.
2.2.2.Energyperformanceassessmentfordiagnosis
Anenergyclassificationprogramonlyprovidestheoverallenergyperformanceresultsatbuildinglevelwhichmayhelptoidentifybuildingswithapoorenergyperformance.Butinterpretingthedataisoftendifficult.Thesuggestionsforenergyperformanceimprovement(ifavailable)arealsoverygeneralandnotfacilityspecific.Incontrast,buildingenergydiagnosisrequiresamorecomprehensiveapproachprovidingmorespecificinformationtoidentifytherootcausesofenergyproblemsassociatedwithsys-temsandfacilities.Ingeneral,energyperformancediagnosisisusedtodetectwhetherenergyinefficienciesoccurandtoidentifywhatarecauses.Correspondingmeasuresarerecommendedtofixtheproblemsandenhancetheenergyefficiency.
Theenergyperformanceassessmentatdifferentlevelsmayprovidedifferentdetailsforfaultdiagnosis.Atbuildinglevel,theenergyconsumersinabuildingconsistofofficeequipment,lighting,HVACandotherservicesystems.Energyperformanceatbuildinglevelcangiveadefinitiveoverallassessmentwhencom-paredwithsimilarbuildings.However,itcannotprovideeffectiveinformationforfaultdiagnosis.Theassessmentatsystemlevel,HVACsystemforinstance,thecooling/heatingloadintensityisausefulindicatortoassessthedemandsideperformance.Furtherassessmentatcomponentlevel,e.g.theCOPofacertainchiller,canexactlyassesstheperformanceofthischilleranddecidewhetheranyimprovingmeasureisneeded.
2.3.Energyperformanceassessmentfornewbuildingsandexistingbuildings
Fornewbuildings(i.e.buildingsunderdesign),thereisnooperationalperformancedataavailable.Theenergyperformanceassessmentthereforeshouldbebasedoncalculatedratings.Forexistingbuildings,bothcalculatedandmeasuredratingmethodsareapplicable,whilethelatterisapreferredinpractice.
Inprinciple,energyassessmentofexistingbuildingscanalsousethecalculation-basedapproach.However,itissubjecttomuchgreaterconstraintsinexistingbuildings.Firstly,itisoftenprob-lematicinobtainingthedetaileddesigndata,includingthedata
ofbuildingenvelope,air-conditioningsystemandoccupationcon-ditions,formodelingabuilding,particularlyanoldbuilding.Siteinspectionstoconfirmthoseinputdetailsarealsodifficultandtimeconsuming.Moreover,thediscrepanciesbetweenthepredictedandactualbuildingconsumptioncanbedetectedeasily,whichmayweakenthecredibilityofthisassessmentmethod.
Ontheotherhand,themeasuredrating,whichusesthemetereddataintheassessment,experienceslessconstrainsandhastheben-efitofrepresentingtheactualuseofbuildings.Inexistingbuildings,measuringenergyuseisusuallymucheasierthancollectingandfeedingvariousinputparametersrequiredincalculationmodels.Themainadvantagesofmeasurement-basedapproachesarealsoresultedfromthefactthattheyreflecttherealbehaviorofbuildingsandallowdetectingtheinfluenceofallthefactors(buildingdesign,operation,occupantbehavior,comfortconditionsandclimate)onthebuildingenergyperformance.Forexample,theIEAAnnex53projectproposedanewmethodologyforanalysisandevaluationofbuildingenergyuse[7].Theultimateoutcomeofthisprojectistostrengthentherobustpredictionofenergyusageinbuildingsbyinvestigatingtherelationshipbetweentheactualmonitoredenergydataandallthemaininfluencingfactors,particularlytheeffectsofactualhumanbehavior.
Duetotheconsiderationsoninputavailabilityandoutputcredi-bility,therearesomegeneralrulestochooseanappropriateenergyperformanceassessmentapproachfornewbuildingsandexist-ingbuildings.Energyperformanceofnewbuildingscanonlybeassessedusingcalculatedapproacheswhilebothcalculatedandmeasuredapproachesareapplicableforperformanceassessmentofexistingbuildings.Whenusingthecalculatedratings,standardenergyratingusuallyintendstodrawupanenergyperformancecertificatewhiletailoredenergyratingoftenisusedforoptimiza-tion,validationanddiagnosis.
2.4.Theapplicationsofenergyperformanceassessment
Buildingenergyperformanceassessmentmethods/toolsdif-fersignificantlyregardingissuesconcerned(e.g.environmental,energy),theirobjectives(e.g.certification,decisionmaking,perfor-mancediagnosis),andthedetailsconcerned(e.g.whole-buildinglevelonly,multi-level).Accordingtoabovedifferences,theappli-cationsofenergyperformanceassessmentcanbeclassifiedintofourcategoriesincluding:(1)buildingenvironmentassessmentschemes,(2)energycertification,(3)whole-buildingbench-markingtools,(4)hierarchicalassessmentanddiagnosistools.Table1presentsasummaryonthetypicalexamplesofapplica-tiontoolsandapplicablebuildings,energyquantificationmethodsandenergyperformanceassessmentmethodsandthedeliverablesassociatedtothesefourcategories.
2.4.1.Buildingenvironmentassessmentschemes
Theassessmentoftheeffectivenessofenergyuseisoneofthemajortasksofmostcommonlyusedbuildingenvironmentassess-mentschemeswhiletheymightalsoincludewater,waste,materialandsite.TypicaltoolsincludeLEED(LeadershipinEnergyandEnvironmentalDesign)inUSA[11],BREEAM(BuildingResearchEstablishmentEnvironmentalAssessmentMethod)inUK[12],CASBEE(ComprehensiveAssessmentSystemforBuildingEnviron-mentalEfficiency)inJapan,GreenStarinAustralianandHK-BEAM(BuildingEnvironmentalAssessmentMethod)in[13].Mostenvironmentalassessmentschemesaremarket-drivenandvoluntarywhichmainlyaimatthecommercialbuildings,particu-larlyatpublicandofficesbuildings.
Mostearlyvisionsaddressonlytheassetassessmentofnewbuildings.Supplementaryversionsfortheoperationalassessmentofexistingbuildingsarealsodeveloped.Suchas,the“LEED-EB:O&M”(ExistingBuildingOperationandMaintenance)isaspe-Table1
Asummaryoftypicalapplicationsofenergyperformanceassessment.
Targeted
TypicalExamples
ApplicableBuildings
EnergyQuantificationMethods
EnergyPerformanceAssessmentMethods
DeliverablesApplicationsEnvironmentalLEED(US),BREEAM(UK),Newandexistingbuildings,CalculatedbydynamicComparethe‘assessedbuilding’witha
4–7gradelevels
AssessmentHK-BEAM(HK)
majorityforcommercialsimulationorapprovedself-referenced‘baseline’buildingandscoringSchemes
buildings
simplifiedmethods,ortheperformanceaccordingtothereductionMeasuredenergydata(forpercentageofenergy/costorCO2emissions
existingbuildings)
GreenStar(Australian)
NewandexistingofficeGreenStarEnergyCalculator,Noreferencebuildingforcomparison,directly1Starto6Stars
buildings
Approvedsimulationprograms
scoringtheperformancebythepredictedgreenhousegasemissions
BuildingEnergyASHRAEbEQ(US)
Newandexistingbuildings
AssetratingandoperationalCompareEUI’sforReferenceBuildingswithAnenergylabelandcertificate
Certification
rating
CBECSData
DOEenergyassetrating(AR)
AllcommercialbuildingsAssetratingmethod
Compareas-builtenergyperformanceamongAnenergycertificate
(US)
similarbuildings
HELP(houseenergylabeling
Existingsingle-familyhousesCalculatedbyahybridmethodComparetheNHACvalueofassessedbuildingEPBDenergycertification
procedure)(EU)
(identificationmethod)withaminimalperformancerequirementELOforlargebuildings(>1500m2),EMforsmallbuildingsMainlyuseassetratingUseEnergyPerformanceIndicator(EPI)(Denmark)
(calculation-based)
asabasisforEPBDenergycertification,EPA-Wforexistingdwellings,EPA-Uforexistingnon-residential,thebuildingwithbetterperformanceEPCfornewbuildings(Netherlands)
thantheminimalrequirementwillget“Energiebedarfsausweis”fornewbuildingsandrenovatedanenergycertificate
buildings(Germany)
EnergyAdviceProcedure,EnergyCharter,PassiveHousePlatform(Belgium)
EnergyPerformanceAssessmentforExistingDwellingsSimplifiedBuildingEnergyCompareenergyuseandCO2emissionwith(EPA-ED),EnergyPerformanceAssessmentforModel,SimpleHourly
minimumenergyperformancerequirementsNon-ResidentialBuildings(EPA-NR)(EU)
Calculation,DetailedDynamicforacertificate
Simulation
Whole-BuildingENERGYSTARExistingresidential&MeasuredenergyusewithComparewholebuildingenergyusewiththe1–100ranking(75+canapplyBenchmarking(US)
commercialbuildings
use-adjustment
distributionalbenchmarktableswhich
forEnergyStarlabel)
Tools
establishedbyregressionmodelsandresultingstandarderrors
Cal-Arch(US)
Existingbuildings
Measuredenergyuse
CompareEUIswiththedistributionaltableofTheposition(percentage)inasimilarbuildingsfromCEUSdatabase
EUIsdistribution
EnergySmartOfficeLabelExistingofficebuildings
Measuredenergyuse
ComparewholebuildingenergyusewiththeGradedSystemefficiencyand(Singapore)
distributionalbenchmarktableswhich
IEQ
establishedbyregressionmodelsandresulting(Top25%canapplyforLabel)
standarderrors
HierarchicalEARM-OAM(Energy
Existingofficebuildings
DisaggregatingthemeasuredProgressivemulti-levelassessmentprocedure:Multi-levelassessmentresultsAssessmentandAssessmentandReportingenergyusefromenergysurveyassesstheperformancebycomparingtheandadvisesonenhancingDiagnosisTools
Methodology-Office
intoEnd-uses
performanceindicatorswithtypicalorbestenergyefficiency
AssessmentMethod)(UK)
referencevalues(atbuilding,systemandequipmentlevels)
AMethodtoAssesstheEnergyExistingcomplexescomprisingDisaggregatingthemeasuredAssesstheenergyperformanceinlandlordsideAnumericalscore(between1PerformanceofExisting
ofamixpremises
energyusefromenergybillandtenantsideseparately,comparetheEUIand4)
CommercialComplexes(HK)
intoend-usesfordifferentwithsimilartypesofpremisesinotherpremises
buildings.
EnergyefficiencydiagnosisforAirconditioningsystemsinBasedonsub-meteredenergyAssesstheperformanceofHVACsystemsandEnergysavingpotential
airconditioningsystemsexistingbuildings
usedata
componentsbycomparingtheenergyassociatedwithretrofitoptions
(China)
efficiencyindicatorswithrecommendedvaluesinregulations
876S.Wangetal./EnergyandBuildings55(2012)873–888S.Wangetal./EnergyandBuildings55(2012)873–888
877
cializedversionofLEEDfortheassessmentofexistingbuildingsduringtheoperationandmaintenancestage,“BEAMPlusforExist-ingBuildings”isaversionofHK-BEAMfortheassessmentofexistingbuildings.Theenergyperformanceassessmentmethodsusedbythetoolsinthiscategorytypicallyincludebothfeature-specific(e.g.,prerequisites)andperformance-basedapproaches.Foranenvironmentalassessmentscheme,theassessmentframe-workisusuallyapplicabletobothnewandexistingbuildings,buttheemphasisandthemethodsofassessmentaredifferent.Theuseofsimulationsoftwareremainsthegenericmethodforassessingnewbuildingswhilemeasuredenergydataareoftenusedforassessingexistingbuildings.Besidesdevelopingtheirownmethodstoassesstheenergyperformance,someenvironmentalassessmentschemesmaydirectlyemploythespecializedenergyratingsystemsforenergyperformanceassessment.Forexample,asanoption,the“LEED-EB:O&M”usetheoperationalenergyratingresultsfromEnergyStartoallocatethepointsofenergyperformanceforexistingbuildings[11].
2.4.2.Energycertification
ManyenergyperformancecertificationmethodshavebeendevelopedbyvariousEPBD(EnergyPerformanceofBuildingsDirective)participants.EuropeanUnioniscurrentlythemajorcatalystwhodrivesenergyefficiencypoliciesinbuildingsec-tors.ThissituationisfurtherhighlightedwiththeimplementationoftheEPBD,whichrequiresthebuildingenergyperformanceassessmentascompulsoryinEurope[14].SuchasELO(energymanagementschemeforlargebuildings)andEM(energycerti-ficationschemeforsmallbuildings)aretwomandatoryschemesaddressingnewandexistingbuildingsinDenmark.TheEnergyPerformanceAdvice(EPA-W)forexistingdwellings,EPA-Ufornon-residentialbuildingsandEnergyPerformanceCoefficient(EPC)fornewbuildingsarewellestablishedintheNetherlands.EnergyPer-formanceAssessmentforExistingDwellings(EPA-ED)andEnergyPerformanceAssessmentforNon-ResidentialBuildings(EPA-NR)aretwoprojectssupportedbytheEuropeanCommissionandwithparticipatingofEuropeanmemberstates.SummarizingthoseEPBD-relatedassessmentmethods,energyperformancecertifica-tioncovers:(1)Theestablishmentofageneralcalculationmethodforbuildingenergyperformance;(2)Thesettingofthemini-mumenergyperformancerequirementsfornewbuildingsandformajorrenovationoflargeexistingbuildings,and(3)Acertifica-tionschemeforratinganddisplayingtheenergyperformanceinbuildings.Althoughoperationalratinghavebeenrecommendedforexistingbuildingassessment,actuallymostoftheenergyquantifi-cationmethods(bothfornewandexistingbuildings)areusingtheassetratingbasedonstandardbehavior.
InUS,similarratingsystemsincludetheASHRAE’sBuildingEnergyQuotient(bEQ)programandtheDOEenergyassetrat-ing(AR)program[15,16].bEQisanenergylabelingprogramslatedtoincludebothanassetlabelandanoperationallabel,therebybeingabletoencompassnewconstructionandexistingbuildings.Theassetlabelprovidesamodeledassessmentofthebuildingperformance,takingintoaccountdesigncomponentssuchasmechanicalsystems,envelope,orientation,andday-lighting.Afterayear’soperation,andsubsequentlyayearofdata,thebuild-ingcanthenobtainanoperationalrating.Theoperationallabelprovidesinformationonmeasuredenergyusebasedonbuildingtypeandoperations.TheARprogramisastandard,voluntaryandentry-levelenergyperformanceassessmentmethodforallcom-mercialbuildingsatthenationallevel.Asanassetratingmethod,ittakesintoaccountonlythephysicalcharacteristicsofthebuild-ing,suchasthebuildingenvelope,themechanicalandelectricalsystems,andothermajorenergy-usingequipment.Thisallowsthecomparisonofbuildingenergyperformanceonanequalfootingbyeliminatingthewidevariationduetodifferencesinoperation
andmaintenance,plugloads,andoccupantbehavior.Theassetratingsystemwillresultinanenergycertificateandhelpcommer-cialbuildingownersandoperatorsidentifyandimplementspecificactionablestrategiestoimproveefficiencyintheirbuildings.Theprogramisstillunderdevelopmentandisrecruitingcommercialbuildingsforpilotssincethespringof2012.
2.4.3.Whole-buildingbenchmarking
Whole-buildingbenchmarkingparticularlyreferstoanoveralloperationalassessmentofexistingbuildings,whichusesthemea-suredwhole-buildingenergyperformanceindexcomparingthestatisticalbenchmarks.“EnergyStar”[17]and“Cal-Arch”[18]aretwowell-establishedwhole-buildingbenchmarkingtoolsinUSA.“EnergySmartOfficeLabel”isasimilarbenchmarkingtooldevel-opedinSingaporealthoughitisnamedasa“Label”program.
2.4.4.Hierarchicalassessmentanddiagnosistools
Inordertoobtainanddiagnosetheperformanceofspecificsystemsorfacilities,theassessmentmightbeundertakenusingahierarchicalapproach.Suchspecificassessmentsareusuallycom-paniedwithadetailedenergyauditinexistingbuildings.Forinstance,Leeetal.proposedamethodtoassesstheenergyperfor-manceofacommercialcomplexcomprisingofpremises,includingoffices,restaurantsandretailshopsin.Inthismethod,theenergyperformancesofeachpremise,inbothtenantsideandlandlordside,areseparatelyauditedandassessed[8].EARM-OAM(EnergyAssessmentandReportingMethodology-OfficeAssess-mentMethod)[19,20]isatypicalprogressivelydetailedmulti-levelassessmentproceduretoassessandinterprettheenergyperfor-manceatsystemlevel.InChina,theenergyefficiencydiagnosisforair-conditioningsystemsisamorespecifictooltoexclusivelyassessanddiagnosetheperformanceofair-conditioningsystems.
3.Energyquantificationmethods
Energyquantificationisthebasisofanyquantitativeenergyperformanceassessmentmethod.Itistheprocessofdetermin-ingtheamountofenergyuseorenergyperformanceindicatorsofagivenbuildingbasedonrelevantinformationcollected.Util-itybills,buildingauditdata,end-usesub-meteringsystemorBMSmonitoringsystem,andcomputersimulationsarecommonsourcestoquantifybuildingenergyuses[21].Accordingtothosedatasources,theenergyuseofanexistingbuildingcanbequan-tifiedthroughthreedifferentapproaches,e.g.Calculation-basedApproach,Measurement-basedapproachandHybridApproach,asshowninFig.1.Usingcalculation-basedapproach,modelscanbebuilttobeverycomplicatedasfordynamicsimulationorsim-pleusingsteady-statemethods.Tobuildsteady-statesimplifiedmodels,bothforwardmodelingandinversemodelingmethodsareapplicable.Usingmeasurement-basedapproach,datacanbecollectedusingdifferentmeans,fromthesimpleway(e.g.energybill)todetailedway(e.g.sub-meteringsystem,BMSmonitoring).Inpractice,inordertocalibratethecalculatedresultsortoiden-tifydynamicmodelingparameters,longorshorttermmonitoringisusuallyconductedtosupportortosupplementcalculations.Insuchcases,thequantificationofenergyuseisachievedusingHybridApproaches.
3.1.Calculation-basedquantification
Acompletemethodforbuildingenergycalculationnormallyconsistsofthreeparts,i.e.inputs(influentialfactors),calculationmodel,andoutputs(energyperformanceindicators).Selectionofinputsandoutputsaremainlydeterminedbyhowthecalcula-tionmodelsarebuilt.Tomodelabuilding,howtoconsiderthe
878
S.Wangetal./EnergyandBuildings55(2012)873–888
Ene rgy Quantific ation Metho dCalcula tio n-b ase dHybr id met hodMeasureme nt-basedDynamic nSimulatioSteady-state MethodCalib rate d nSimulatioDyn amic Inv erseModelingMonitor ing-base dMetho dBill- base dMethodForwardModelin gInverseModelingSub-metering/NILM/ BMSEnerg yDisaggreg ationFig.1.Energyquantificationmethodsforexistingbuildings.
dynamiceffectsofbuildingsandsystemsarequitedifferentinvari-ouscalculationmethods.Consequently,calculation-basedmethods
canbeclassifiedasdynamicmethodsandsteady-statemethods.Dynamicmethods,oftenadoptedbydetailedsimulations,arecapa-bleofcapturingbuildingdynamicssuchasthermaldynamicsofenvelopeandsystemdynamicssuchasthatofonlinecontrolstrate-gies.Whileinthesteady-statemethods,thedynamiceffectsareignoredorsimplifiedbycorrelationfactors.Suchsimplificationgreatlydecreasesthecalculationcomplexity.
Forbothdynamicmethodsandsteadystatemethods,thecal-culationprocedurecanbeestablishedintwodifferentmodelingapproachesbasedontheinputsandoutputsselected,i.e.forwardmodelingandinversemodeling.Modelsusingforwardmodelingmaybenamedasdeterministicmodels.Modelsusinginversemod-elingmaybenamedasregressionmodels.Aclassicalmethodofforwardmodelingistopredictenergyusebasedonfirstprinci-pleswithgivenconditionsasinputsandbuildingcharacteristics.Thismodelingmethod,alsoreferredasphysicalmodeling,beginswithadescriptionofthebuildingsystemorcomponentanddefinesthebuildingaccordingtoitsphysicaldescription.Mostdynamicmethods,suchasvariousbuildingsimulationtools,adoptforwardmodelingtocreateathermodynamicbuildingmodelusingfunda-mentalengineeringprinciples.
Usinginversemodeling,theinputsandoutputsshouldbefirstlyavailable,andthemodelparametersorstructurewillbetrainedandidentified.Therefore,thismodelingmethodisalsoreferredtoasparameteridentificationorsystemidentification.Twotypesof
inversemodelshavebeenreportedintheliterature,i.e.steadystateinversemodelsanddynamicinversemodels.
Thissectiononlyintroducestheenergyquantificationmeth-odspurelybasedoncalculation.Whenmeasurementsneedtobeinvolved,suchasusingmeasuredenergybilltocalibratethepredictedenergyconsumptionsfromsimulations(i.e.calibratedsimulation),orusingshort-termmonitoreddatatotraintheparam-etersofdynamicinversemodels,themethodsareclassifiedashybridapproach.
3.1.1.Dynamicsimulationforenergycalculation
3.1.1.1.Basicprinciplesofenergysimulation.Dynamicsimulationprograms,alsocalleddetailedsimulationtools,havebeenacceptedaspowerfultoolsforanalyzingbuildingenergyperformance[22,23].Theproceduretobuildadynamicsimulationusuallyusesaforwardmodelingapproach.ThegeneraldataflowandmaintaskstocompleteatypicalsimulationareshowninFig.2.
Usingdynamicsimulation,requiredinputsshouldbefirstlycol-lectedandthenbefedintoaso-calledsimulationenginetodescribedetailedmathematicmodels.Typicalinputsforadynamicsim-ulationtoolmayincludefourgroupsofparameters,i.e.weatherconditions,buildingdescription,systemdescriptionandcompo-nentdescription.Theweatherconditionsgenerallyincludethedryandwetbulbtemperatureofoutdoorair,solarradiationintensity,windspeed,etc.Buildingdescriptiondatamainlyincludeloca-tion,designandconstructiondata,thermalzones,internalheatgain,infiltrationandusageprofiles,etc.Forsystemdescription,
sInputWeathe rConditio ns onSimulati eEnginBuildin g d LoaOutput sBuildingDescriptio nFeedbacksSystem nDescriptioSystem sAnalysiFeedbacks yEnerg anc ePerform orsIndicat entCompon nDescriptioPlantAnalysisFig.2.Generaldata-flowandmainstepsofdetailedsimulation.
S.Wangetal./EnergyandBuildings55(2012)873–888
879
systemtypesandsizes,controlschedules,aswellasoutdoorairrequirementsarerequired.TheinputsforcomponentdescriptionaremainlyaddressedHVACcomponents,includingtheequipmenttypesandsizes,performancecharacteristic,loadassignmentsandauxiliaryequipment[24].
Thecorepartofmostsimulationprogramsisthesimulationenginewhichdescribesthedetailsofmathematicalsimulationalgorithms.Asimulationenginegenerallyinvolvesthreemajorsteps,i.e.thermalloadscalculation,systemsimulationandcen-tralplantanalysis.Thebuildingloadsubprogramsareemployedbymostprograms,whichusethebuildingdescriptioninformationandweatherdatatocalculatethehourlyheatingandcoolingloadsofabuilding.Heatbalancemethodandweighting-factormethodaretwobasicmethodstocalculatethecooling/heatingloadreleasedfrombuildingenvelopestoairconditionedspace.Thesystemsub-programsareusedtosimulatetheperformanceofairhandlinganddistributionsystems,andrelatedcontrolschedulesinmain-tainingtherequiredspacetemperatureandhumidity.Theplantsubprogramstaketheoutputsofthesystemsubprogramanduseinputsofspecificequipmenttocalculatethefuelandelectricityrequirementsofthebuildinganditsenergysystems[24].Simpli-fiedandsteady-statemethodsaregenerallyacceptabletosimulatetheperformanceofsystemsandequipment.
Whenthesubprogramsofbuildingload,systemandplantareestablished,therearetwosequencestosolvethesesubprogrammodels,i.e.sequencemodelingandsimultaneousmodeling.Thecalculationprocedureinsequencingmodelingiscompletedstepbystep.Thereisnodatafeedbackfromlattersubprogramstoformersubprogramswhichmaycausetheassignedloadmismatch-ingwiththeactualhandlingcapacity,andreducetheaccuracyofsimulations.Incontrast,usingsimultaneousmodelingmethod,thesub-modelsofbuilding,systemandplantaresimultaneouslycalcu-latedineachtime-step.ItguaranteestheactualairstatesafterthehandlingofHVACsystemscanbetimelyfedbacktothecalculationofthermalload.Asaresult,thesimulationaccuracyisenhancedwhilethecomputationcomplexityisincreasedgreatly.
Theenergyperformanceinformationisthefinaloutputofthesimulations.BesidestheannualenergyuseofabuildingwhichisgenerallyexpressedasEUIororCO2emission,moredetailedenergyperformanceindicatorsincludingtheheating/coolingloadofbuild-ingandeachthermalzone,equipmentefficiencyofsystemandcomponent,areusuallyprovided.
3.1.1.2.Representativesimulationtools.TherearevariousdynamicsimulationstoolsavailableamongwhichDOE-2,EnergyPlusandTRNSYSarewell-knownandcommonlyused,whicharebrieflyintroducedasfollows.
DOE-2:DOE-2[25]usestheweightingfactormethodtocalculateheattransferbetweenspacesandusessequencemodelingmethodtosolvesimulationsubprograms.DOE-2ispowerfultosimulatealmostalltypesofbuildingsenvelopes.ForHAVCsystemsmod-eling,thecommonlyusedtypesofsystemsarepredefinedwithinthetemplateofDOE-2.ItalsoofferspossibilitiesforuserstoaddcasespecificfunctionsfornewsystemswhilethisprocessinvolvesasignificanteffortandrequiresagoodunderstandingofDOE-2.TheoriginalinterfaceofDOE-2isnotfriendlyenoughbutmorethanonehundredsoftwaretoolswithbetterinterfaces,SuchasVisualDOE,eQUESTandPowerDOE,aredevelopedonthebasisofDOE-2.
EnergyPlus:EnergyPlus[26]combinesthebestfeaturesofDOE-2andBLAST[27]anditisoftenreferredasa“newgeneration”simulationengine.ItusesheatbalancemethodofBLASTtocalcu-latethethermalloadswhileadoptsandimprovesthestructureofDOE-2toorganizesubprogramsbyusingthesimultaneousmodelingmethod.Forbuildingthermalperformancesimulation,
variousfactorsincludingshading,verdurization,windandsnowaredetailedconsideredinbuildingmodelingwhichenablesEner-gyPlustosimulatealmostalltypesofcomplicatedbuildings.Forsystemmodeling,user-configurableheatingandcoolingequip-mentcomponentsfortheeachofthemostpopularsystemtypesareincluded.Thiscomponent-basedHVACmodelingmethodgivesusersmuchmoreflexibilityinmatchingtheirsimulationtotheactualsystemconfigurations.EnergyPlusalsocontainsinter-zonalairflow,moistureabsorptionanddesorption,definitionsofmorerealisticHVACsystemcontrolsandradiantheatingandcoolingsystemstoincludemorenewHVACtechnologiesandconcepts.Asasimulationengine,EnergyPlusdoesnotofferauser-friendlygraphicalinterface.Severaluserinterfaces,suchasDesignBuilder,areofferedbythird-parties.
TRNSYS:TRNSYS[28]adoptacompletelydifferentsimulationstructureasEnergyPlusandDOE-2.Itusestheconceptof“compo-nent”toassembleasimulationmodel.Eachcomponentrepresentsamathematicaldescriptionofasubsystem,equipmentorather-malandmasstransferprocess.TRNSYSprovidesacalculationplatform,bywhichuserscancall,modifyandevendefinetheirowncomponents,andthenassembletogetherforsimulation.Allconnectedcomponentsareintegratedasawholeandthenbesolvedsimultaneously.ComparingTRNSYSwithEnergyPlusandDOE-2,thethermalloadcomponentismuchsimplerwhiletheabilitiestosimulateHVACperformanceandcontrolsystemsaremorepowerful.Inaddition,thecomponent-connectionanduser-definingfeaturesenableTRNSYSwithexcellentflexibilitytosimulatevariousbuildingsandsystems,particularlysuitablefornovelsystems.
3.1.2.Steady-statemethodsforenergycalculation
Steady-statemethodsaredevelopedmainlyforsimplifiedbuildingenergycalculation.Steady-statemethodshavetheadvan-tagesofhighcomputationspeedandsimplificationinmodelingduetoignoringofdynamiccharacteristics.Therearetwooppo-siteapproachestobuildsteady-statemodelsforenergycalculationinpractice,i.e.forwardmodelingandinversemodeling.Simpli-fiedBuildingEnergyModel(SBEM)isatypicalsteady-statemodeldevelopedfortheimplementofEPBDusingforwardmodelingapproach.Modelingexamplesusinginversemodelingapproachincludewhole-buildingregressionmethods,degree-daymethod,variablebasedegree-daymethod,BINandmodifiedBINmethod,equivalentfullloadhourmethodaredevelopedusinginversemod-elingtechniques.
3.1.2.1.Forwardmodelingapproach.SimplifiedBuildingEnergyModel(SBEM)wasoriginallybasedontheDutchmethodologyNEN2916:1998(EnergyPerformanceofNon-ResidentialBuildings).IthasbeenmodifiedtocomplywiththerecentCENStandardsastheNationalCalculationMethod(NCM)ofenergyperformancecal-culationforEPBD.SBEMadoptsa(quasi)steadystatemethodtocalculatetheheatingandcoolingdemandonamonthbasis.Inthismethod,dynamiceffectsofabuildingaretakenintoaccountbyintroducingcorrelationfactors,alsocalledutilizationfactors,forbothcoolingandheatingcalculation.Thefactorsdeterminetowhatextentheatgainsareusefulfortheheatingdemand(duringheat-ingmode)andtowhatextentheatlossesareusefulforthecoolingdemand(duringcoolingmode).Thebasicformulationsofthemodel[29]are:
QNH=QL,H−ÁG,H·QG,H
(1)QNC=QG,C−ÁL,C·QL,C
(2)
where,QNHandQNCarethemonthlyenergydemandforheatingandcoolingrespectively.QL,HandQL,Careheatlossesforheatingandcoolingmoderespectively.QG,HandQG,Careheatgainsforheating
880
S.Wangetal./EnergyandBuildings55(2012)873–888
andcoolingmoderespectively.ÁG,HandÁL,Carethegainutilizationfactorforheatingandcoolingmoderespectively.
Theutilizationfactorsrepresenttheportionofgains(duringtheheatingseason)oroflosses(duringthecoolingseason)thatcon-tributetothereductionintheheatingdemandorinthecoolingdemand.Thedeterminationofthevalueofutilizationfactorshasbeenintroducedinmanyrelatedresearchpapers[14,30].
Thismonthlysteady-statesimplifiedmethodcalculatesenergyuseinaforwardapproach.Havingcalculatedtheenergydemandforheating/cooling,theenergyuseforheatingandcoolingcouldbeeasilydeterminedinEqs.(3)and(4),onamonthlybasis:
EH=
QNH
SCoP(3)EC=
QNC
SEER(4)
whereSCoPandSEERareforshortof“SystemCoefficientofPer-formance”and“SystemEnergyEfficiencyRatio”ofHVACsystems,whichrepresenttheoverallefficiencyofHVACsystemduringheat-ingandcoolingperiod,respectively.
Asthenationalenergyperformancecalculationmethodadvo-catedbyEPBD,SBEMmethodhasbeenwidelyadoptedanddeeplyinvestigatedinEurope(e.g.UNI10344inItaly,EN832inEurope)[31,32].TherearealsomanyexamplesoftheapplicationsofSBEMforenergyperformanceassessment,suchasbuildingregulationscompliancefornewbuildingdesigns,assetratingforbothnewandexistingbuildings.TheapplicationofSBEMin“Assetrat-ing”ofexistingbuildingisillustratedinaninternationalstandard(ISO13790:2008),Energyperformanceofbuildings—Calculationofenergyuseforspaceheatingandcooling[33].TwoEuro-peanprojectsforenergyperformanceassessment,namely,“EnergyPerformanceAssessmentforExistingDwellings”(EPA-ED)and“EnergyPerformanceAssessmentforexistingNon-Residentialbuildings”(EPA-NR)alsoadoptedthismethodtoquantifybuildingenergyperformance.
3.1.2.2.Inversemodelingapproach.Inversemodelingapproachrelatesenergyperformanceindicators(energyloadorconsump-tion)tooneormoreimportantenergyinfluentialfactors(e.g.weatherconditions,floorarea).Usingthisapproach,astructureorphysicalconfigurationofthebuildingorsystemisassumedfirstandmodelcoefficientsarethenidentifiedwithgivendatabystatisticalanalysisorregressionanalysis.Forsteady-stateinversemodeling,themodelcoefficientsand/orevenmodelstructurearefixedoncebeingidentified.Steady-stateinversemodelingmethodsareusu-allyusedattwolevels,i.e.atwholebuildinglevelandatHVACsystemlevel.
Inversemodelingforwholebuildinglevelconsumptionisthesimplestformtocalculatebuildingenergyperformance.Usingthismethod,thewholebuildingenergyuse(orEUI)canberegressedagainstseveralimportantinfluencingparameters.Linear,change-pointlinearandmultipleregressionsarewidelyacceptedtechniquestocorrelateenergyusewithweatherdataand/orothervariables.Ingeneral,theprocedureofusingwholebuildingenergymodelinvolvescreatingawholebuildingmodel,anumberofdifferentregressionmodelsfortheparticularbuilding,andthencomparingtheresultsandselectingthebestmodelusingR2andCV(RMSE).Inmostcasesthewhole-buildinginversemodelhastheform:
E=C+B1V1+B2V2+···+BnVn
(5)
whereEistheenergyuseestimatedbytheequation.Cisaconstantterminenergyunits.BnrepresentstheregressioncoefficientofanindependentvariableVn.
Therearemanyexamplesthatusingsuchmethodstoanalyzethewholebuildingenergyperformance[34].Forinstance,alinear
regressionwasusedtostudyEUIforofficebuildingsregisteredinthe1992USCommercialBuildingEnergyConsumptionSurvey(CBECS).Thestrongestdeterminantparameterswerefoundtobefloorarea,numberofworkers,personalcomputers,owneroccupancy,operationhoursandpresenceofeconomizersandchillers[35].AmultipleregressionmodelwasalsousedtosearchforexplanatoryfactorsofEUIforofficebuildings.Inthisstudy,themostsignificantfactorswerefoundtobethebuildingage,operatinghours,floorarea,numberofconsumersandasubjectivequalitativedescriptionofuserbehaviorandmaintenance[36].Moretypicalsteady-stateinversemodels,includingconstantormeanmodels,modelsadjustedforthedaysinthebillingperiod,twoparametermodels,threeparametermodelsorvariable-baseddegree-daymodels,fourparametermodels,fiveparametermod-els,andmultivariatemodelsareintroducedinASHRAE’sIMT(InverseModelingToolkit)[37].
Besidestheactualdatafromstatisticalsurveyofexistingbuildings,thedatafromdetailedsimulationscanalsousedforiden-tificationofinversemodels.Josephetal.[38]usedtheenergydatafromaDOE-2simulationtoestablishanenergyregressionmodelforanofficebuilding.Throughinversemodelinganalysis,atotaloftwelvebuildingdesignparameters(sixfromthebuildingload,fourfromtheheatingventilationandair-conditioning(HVAC)sys-temandtwofromtheHVACrefrigerationplant)wereidentifiedandconsideredasinputsintheregressionmodels.Thedifferencesbetweenregression-predictedandDOE-simulatedannualbuildingenergyusearelargelywithin10%.Leeetal.[39]proposedamultiplelinearregressionmodeltoassesstheenergyperformanceofexist-ingbuilding.Thismodelwasestablishedbasedondetailsimulationandhasbeenadoptedasanalternativemethodforcertainbuild-ingsinthelatestvisionofHK-BEAM.Furthermore,shesummarizedthegeneralstepstoestablishaninversemodelingbysimulation:(1)Generationofpredictionsofenergyperformancedataforasuf-ficientnumberofbuildingsbyawellrecognizedsimulationtool.(2)Identificationofkeyinfluentialparametersandsuitablemath-ematicalformsfortheregressionmodels.(3)Determinationofthecoefficientsassociatedwitheachindependentvariableinthemod-elsbymultipleregressionanalysisbasedonthesimulationresults.(4)Verificationoftheaccuracyoftheresultantregressionmodels,andtheadequacyfortheirintendedpurposes.
Therearealsosometypicalsteady-stateinversemodelingmethodsforbuildingloadcalculation,suchas“Degree-day”and“Variablebasedegree-day”method,“BinandmodifiedBin”method,and“Equivalentfull-loadhour”(EFLH)method.
3.1.2.2.1.Degree-daymethodandvariablebasedegree-daymethod.Thedegree-daymethodassumesthatsolarandinternalheatgainoffsetsheatlosseswhenthemeanoutdoortempera-tureis65◦F(basetemperature),andthattherateofheatgainorlossisproportionaltothedifferencebetweenthemeandailytemperatureandbasetemperature.Degreedaysareessentiallythesummationoftemperaturedifferencesovertime,andhencetheycapturebothextremityanddurationofoutdoortemperatures.Thismethodisgenerallyusedinresidentialbuildingswhereinfiltrationandenvelopeconductiontransmissiondominatethebuildingloadwhilenotsuitableforcommercialbuildingswithlargeinternalheatgains.Thevariablebasedegree-daymethodisamodifiedversionofdegree-daymethodwithvariablebasetemperatures.Thebasetemperatureisnotfixedtobe65◦Fbutdeterminedbytheactualbalancepointofacertainperiod.Thismodificationenablesthedegreedaymethodtobebetterfortheapplicationsinpracticebytakingnon-constantindoortemperatureandvariationofinternalandexternalheatgainsintoaccount.
3.1.2.2.2.BinandmodifiedBinmethod.TheBinmethodcalcu-latesthebuildingloadbydeterminingthenumberofhoursperyearthattheaverageoutdoortemperatureofthelocationunderstudywascontainedinatemperaturebandor“Bin”.Addingthe
S.Wangetal./EnergyandBuildings55(2012)873–888
881
loadforeachofthesetemperaturebinsdeterminestheyearlyenergyrequirement.RefinementstothismethodincludetheuseofcoincidentwetbulbtemperaturestocalculatelatentloadsineachtemperatureBin.TheBinmethoddoesnotaccountforsolarheatgains.InthemodifiedBinmethod,solarheatgainsareconsideredbyusingasimplifiedtemperaturedependentrepresentation[40].3.1.2.2.3.Equivalentfull-loadhour(EFLH)method.Thismethodneglectsoperationalaspectsandisstrictlyempiricallybasedonexperiences.Equivalentfull-loadhouristheratioofannualcooling/heatingenergyrequirementstoratedcapacityofthecooling/heatingequipment.TheEFLHforacertainlocationisrelatedtobuildingfunction,typeandtheoperationalstrategiesofHVACsystem.Itisusuallyestimatedaccordingtoexpert’sexperiencesorbasedonthestatisticsurvey.Thismethodhasverylimitedapplicabilityandisgenerallyusedforroughestimationsofannualcoolingandheatingrequirementsforbuildingsusingself-containedairconditionersratherthanusingcentralizedair-conditioningsystems.
3.2.Measurement-basedquantification
Foranycalculation,thereareinevitablediscrepanciesbetweenthepredictionsandactualconsumptions,whichmayweakenthecredibilityofresults.Fornewbuildings,calculation-basedmethodistheonlychoice.Forexistingbuildings,however,energyquantifi-cationbasedonmeasureddatasoundstobeabetterchoice[41].Therearemanyoptionstomeasuretheactualbuildingenergycon-sumption,fromthesimplestenergybilltothedetailedend-usesub-meteringandmonitoring.
3.2.1.Energybill-basedmethods
Energybillisatypeofhighqualitymeasurementdata.Itwouldbepreferredasthemostcost-effectivemethodtoquantifytheenergyusesinceenergybillsarereadilyavailableinmostexistingbuildings.However,theoriginaldatafrommonthlybillsprovideinsufficientinformationforenergyperformanceassessment,par-ticularlyformulti-levelassessmentanddiagnosis,becausethistypeofdataisinherentlyaggregatedacrossend-uses.Disaggre-gatingenergybillsintoend-useprovidesabetterunderstandingonenergyuseandresultsinabetterassessmentforsystemsandequipment.
Energybilldisaggregatingisaprocesstoapportionthetotalenergyconsumptionsfromenergybillsintoenduseofmainsys-temsandequipmentwithanacceptablelevelofaccuracy.Energydisaggregationmethodsgenerallyadopttwodifferentalgorithms.Oneisthe“estimationalgorithm”whichreliesontheknowledgeoftheusagecharacteristicsofvariousfacilitiesatbottomlevel.Theotheroneisthe“disaggregationalgorithm”whichisbasedontheunderstandingontheinteractiverelationshipsbetweentheend-usesandtheirhigh-levelparents.Therearethreetyp-icalenergybilldisaggregationmethodswithmoredescriptionprovidedbelow,whichmaybeusedforenergyperformanceassessment.
3.2.1.1.Bottom-upcalculationmethod.Abottom-upcalculationtodisaggregateenergybillintomaintypesofend-usewaspresentedbyFieldetal.[20].Usingthismethod,theenergyuseofeachitemiscalculatedindividuallyandthenbesummedforreconciliationwithavailablemeteredinformation,i.e.energybills.Forthecalcu-lationofeachtypeofend-use,theinformationsuchastheratedelectricalload,electricalloadfactor,theusagepatternandtheusagefactorarerequired.Alargediscrepancybetweenthecalcu-latedandmeteredconsumptionindicatesthecollectedcalculationparametershavelargeruncertaintyandthedisaggregatedresultsareunacceptable.Asaresult,thecalculationinformation,particu-larlytheusagepattern(e.g.usagefactor,timeschedules)needsto
betunedinajustifiedrangetoreconcilethesummedenergyusewiththemeasuredconsumption.ThismethodwasemployedinaprototypeoftheEnergyAssessmentandReportingMethodology(EARM)project,namelyOfficeAssessmentMethod(OAM)fortheanalysisofenergyuse[19].
Leeetal.alsodevelopedasimilarmethodtobreakdownthetotalenergyuseintofourmajorend-uses,i.e.air-conditioning,lighting,officeequipmentandmiscellaneousequipment[8].Theindividualenergyuseofeachtypeofequipmentwascalculatedbymultiplyingthenumber,ratedconsumption,utilizationfactor(UF),andtheestimatedoperationhoursthroughouttheyear,inaccordancewithEq.(6):
AECEquip,i=(
Ni,j·Wi,j)·UFi·AOHr
(6)
i
Thebreakdownmayneedtobecarriedoutinaniterativemanner,involvingfine-tuningofthevalueoftheutilizationfactor,inorderthatthesumofenergyuseofallcontainedequipmentwillmatchthetotalelectricityconsumptionasdeterminedfromtheelectricitybills.
AEC
Bill=
AECEquip,i
(7)
i
Breaking-downtheenergybillusingthisbottom-upmethod,however,isalabor-intensive,experience-dependedprocessduetotheextensiveinspectionofcalculationparametersfromvari-ousfacilities.Inaddition,thereconciliationbetweenthesummedconsumptionandthetotalamountatwholebuildinglevelcannotguaranteetheaccuracyofdisaggregatedconsumptionatbottomlevel.
3.2.1.2.Bottom-upshort-termmeasurementmethod.Usingabovecalculationmethod,theactualpowerinputandefficiencyofequip-mentarealwaysdifferentwiththeratedvalueandtheactualoperationhoursarealsodifferentwiththatfromoperationalschedules,whichmaycausesignificantdiscrepancyinenergyestimation.Incontrast,insitumeasurementmayprovidemoreaccuratevaluesandmoreactualschedule.Davidetal.devel-opedamethodtodisaggregatetheenergybillintoprincipalendusesusingashorttermmeasurements(STM)method[42].Inthismethod,buildingconsumersaredividedintoseasonalvary-ingend-users(i.e.HVAC)andnon-seasonalvaryingend-users(i.e.lighting,hotwaterandothers),accordingtovariabilityandsea-sonaldependence.Thenon-seasonalend-usesarealsoreferredastosteady-stateconsumersbecausetheirconsumptionsarecon-sideredtoberelativelyconstantthroughouttheyear.Foreachtypeofnon-seasonalend-users,theconsumptionshapes(e.g.theconsumptionofaweek)intypicaldays(e.g.weekdaysandweek-enddays)canbeconstructedbyinsitumeasurement.Theannualconsumptioncanbethencalculatedsimplybymultiplyingtheweeklyenergyusebythenumberofworkingweeksthroughoutayear.TheenergyuseofHVACisthenobtainedbysubtractingtheconsumptionsallnon-seasonalend-usesfromthetotalenergyconsumption.
Parkeretal.[3]suggestedaprotocolformeasuringthesteady-stateconsumptionforindividualappliancesinahouseusingameteringsystemcalledTED(TheEnergyDetective).Itinvolvesmanuallyswitchingtheappliancesinthehouseoffandonwhileanotherpersonrecordstheassociatedchangeinpowerconsump-tion.Thistechniquewasreportedtotake2–4hfor2people,andcanidentifythesteady-stateconsumptionofmostelectricloadsover10W.Thismanuallyswitchingmeasurementprotocolmaybecosteffectiveandlaboraffordableforsimpleresidentialcus-tomers.Whileforcommercialbuildings,thismethodmightnotbeappropriatesincetoomanyfacilitiesareinvolved.
882
S.Wangetal./EnergyandBuildings55(2012)873–888
3.2.1.3.Top-downdisaggregationalgorithm.Incontrasttothebottom-upapproach,LawrenceBerkeleyLaboratorydevelopedatop-downdisaggregationalgorithm,namedEnd-useDisaggrega-tionAlgorithm(EDA),tobreakdownwhole-buildingelectricalloadintothreetypesofenduses(HVAC,lightingandmiscellaneous).IntheEDA,thesumoftheendusesisconstrained,tobeequaltothemeasuredwhole-buildingload.Inaddition,theload/temperaturerelationshiphelpstocharacterizetheair-conditioningenduse,providinganadditionalconstraintontheremainingendusesandpreventingsomeoftheerrorspossiblewithsimpleproration.Thedisaggregationisperformedintwomajorsteps.First,apreliminaryend-useprofileofbuildingisdrawnbyusingtheon-sitesurveydataandtheDOE-2program.Meanwhile,theload/temperaturerela-tionshipisalsoestablishedwiththeresultsfromsimulations.Inthesecondstep,theinitialend-usedistributionswillbeadjustedaccordingtothetwoconstraintsabove.Theadjustmentprocessshouldbecarriedoutiterativelyifanyconstraintisnotsatisfied[21].Itisworthnoticingthatthisenergydisaggregationmethodisdevelopedforhourlyloadbreaking-downbyauthorswhilethealgorithmalsomakessenseorevenbebetterformonthlyorannualenergybilldisaggregationsinceabetterdependencybetweenHVACloadandweatherconditionscanbeobtainedinalongerperiod.
3.2.2.Monitoring-basedmethods
Energybillscanonlyprovidethetotalconsumptionofawholebuilding.Disaggregatedenergyconsumptionsfromenergybillsmayprovideend-userconsumptiondatawhiletheaccuracyanddetailsarestillverylimited.Moreaccurateanddetailedenergyuseinformationshouldbemonitoredbymoresophisticatedmeteringsystemsorplatforms.
3.2.2.1.End-usesub-metering.Onesolutiontoobtainenergyuseofindividualloadsistoplaceseparatemeteringhardwareoneachrelevantcircuitbranch,whichisreferredasend-useseparatemeteringorsub-meteringtechnology.Theresearchonsub-meteringtechnologyhasbeencarriedoutsince1980s.Forinstance,inordertoestablishanend-useempiricaldatabaseforcommercialbuildings,theU.S.DepartmentofEnergy(DOE)consignedPacificNorthwestLaboratory(PNL)tocompleteareportentitled“Com-mercialBuildingEnd-UseEnergyMeteringInventory”[43].Themajorfindingsfromtheinventoryarethatwell-documentedandwidely-availabledatabaseofsub-meteredend-usedataisbothdesirableandvaluableforthebuildingresearchcommunitywhiletherearefewwell-documentedsourcesofmeteredend-usedatainthepublicdomain.Therefore,PNLrecommendedthatDOEestab-lishagovernment-sponsoreddatabaseofend-usemetereddata.Thedatashouldbereviewedandprocessedsothatitispubliclyavailableinaneasy-for-useandcommonformat.
Sub-meteringsystemismainlyusedtoprovidedetailedenergydataforresearchorvalidationpurposeswhileisusuallyconsideredtobetooexpensiveforpracticalapplicationincommonbuildings[21,44,45].Sub-meteringtechnologyisabletoclearlyseparatedif-ferentusesofelectricitydependingontheconfigurationofthehardware.Iftheexistingelectricconfigurationiswelldesignedandorganized,e.g.thesametypeofend-useisconnectedinthesameelectricalcircuit,onlyfewmetersneedtobeinstalledinthemainelectricentrances.However,formostexistingbuildings,particu-larlyoldbuildings,thecircuitsareusuallymixedtoservicedifferentend-users,duetoyearsofretrofit,maintenanceandrelocation.Insuchcases,circuitreorganizationoranumberofadditionalmetersarerequiredtoensurethatnootherelectricloadsaremixedwiththetargetedmeasurement.Consequently,theimplementationofsub-meteringinsuchbuildingsmightbeverydifficultandexpen-sive.Ingeneral,togetafullsignatureofbuildingenergyuse,metersmustbeleftinplacelongenoughtorecordusageduringseveral
differentweatherandoccupancyconditions,evencontinuouslyonlinemonitoring.
3.2.2.2.Non-intrusiveloadmonitoringmethod.Non-intrusiveloadmonitoring(NILM)methodisapatternrecognition-basedmethodwhichiscapableofgatheringdetailedenergy-usedatawithoutsub-metering.Workperformedbyresearchershasshownthatindi-vidualloadscanbedetectedandseparatedbyrapidsamplingofpoweratasinglepointservinganumberofappliances,forexampletheelectricalserviceentranceforawholebuilding.TheapplicationofNILMgenerallyconsistsoftwomodes,i.e.samplingmodeanddisaggregationmode.Inthesamplingmode,theoperatingchar-acteristicandusagepatternofeachend-usearedeterminedfromthedatacollectedoveraperiodofseveraldaysusingatleastonecurrentsensorperappliance.Thisprocessisalsoreferredasmodeltrainingwhereend-usecharacteristicscanidentifiedbyshort-termsub-metering.Inthedisaggregationmode,onlythemainelectricentranceismonitored.Theelectricsignalisgenerallyanalyzedusingpatternrecognitionmethodstodisaggregatethemonitoredenergyuseintoend-uses.
Non-intrusiveloadmonitoringmethodhasbeenreportedtoworkwellforresidentialbuildings[46,47].Forinstance,MarceauandZmeureanupresentedacomputerprogramtodisaggregatethetotalelectricityconsumptionofahouseusingnon-intrusiveloadmonitoringmethod.Theresultsshowthattheerrorsinestimatingtheenergysharesofthreemajorendusers(waterheater,base-boardheaterandrefrigerator)arelessthan10%formostevaluationscenarios[48].
Non-intrusiveloadmonitoringmethodcanalsobeusedincom-mercialbuildingsbuttheapplicationwillbemoredifficultduetothelargercomplexityanddiversityoffacilities.Leslieetal.devel-opedaNILMmethodandapplieditinanoffice/laboratorybuilding.Usingthismethod,besidesthemonitoredconsumption,equipmentstart-upandshut-downeventsarealsocentrallyobserved.Indi-vidualloadsaredistinguishedbymatchingthesetransienteventstoknownpatterns.TheprototypeofthisNILMmethodperformedwelltoinlaboratorytests[49].
3.2.2.3.BMS-basedmethods.Actually,theavailabledatafromproperlydesignedandmaintainedBMS(BuildingManagementSystem)aregenerallysufficienttoobtainaclearpictureoftheenergyuseoftypicalHVACsystems.Ifnotsufficient,fewdedicatedmeasurementdevices,suchaselectricitymetersandtemperatureloggers,maybeemployedasasupplementtoBMSmonitoring.Inthisway,theenergyperformanceofHVACsystemandenergycon-servationopportunitiesmaybereadilyidentifiedandimplementedatveryloworevennegligiblecosts.CasestudiesdevelopedbytheHARMONACprojecthaveshownthatBMScanbeapowerfulplatformforenergyperformancemonitoring[50].
3.3.Hybridquantificationmethod
Usinghybridquantificationmethods,themajorityeffortsarestillpaidonthecalculationanalysiswhilemeasurementsaregener-allysupplementedtoreducecalculationdiscrepanciesortoidentifymodelparameters.Calibratedsimulationanddynamicinversemodelingaretwotypicalhybridmethodstoquantifyenergycon-sumption.
3.3.1.Calibratedsimulation
Thesimulationonlyaftercarefulcalibrationcanprovidecred-ibleresults.Calibratedsimulationisaprocesswhichusesanacknowledgedbuildingsimulationprogramtotunetheinitialguess-valuesofinputstotheprograminaheuristicmannersothatthepredictedenergyusematchescloselytothemeasureddata[51].
S.Wangetal./EnergyandBuildings55(2012)873–888
883
Claridgeetal.developedamethodologyfortherapidcalibrationofcoolingandheatingenergyusesimulationsforcommer-cialbuildingsbasedontheuseof“calibrationsignatures”thatcharacterizethedifferencebetweenmeasuredandsimulatedper-formance[52,53].Themethodisdescribedandthenitsuseisdemonstratedintwoillustrativeexamplesandtwocasestudies.Thereportcontainscharacteristiccalibrationsignaturessuitableforuseincalibratingenergysimulationsoflargebuildingswithfourdifferentsystemtypes:single-ductvariablevolume,single-ductconstant-volume,dual-ductvariable-volumeanddual-ductconstant-volume.
Reddyetal.developedageneralproceduretocalibratethedetailedbuildingenergysimulationprograms(e.g.DOE-2),withmeasureddatathroughASHRAEfundedproject1051-RP.Thispro-cedureconsistsoffivesteps:(1)identifyacapablebuildingenergyprogramandsetupthesimulationinputfiletobeasrealisticanderror-freeaspossible.(2)Reducethedimensionalityoftheparam-eterspacebyperformingwalk-throughauditsandheuristics.(3)Performacoarsegridsearchwhereinthedimensionality-reducedinfluentialparameterstoidentifythestrongandweakparameters,andtonarrowtheboundsofvariabilityofthestrongparameters.(4)Performaguidedgridsearchtofurtherrefinethepromisingparam-etervectorsolutions.(5)Usethissmallsetofsolutionstomakepredictionsaboutintendedchangestothebuildinganditssystems,anddeterminethepredictionuncertaintyoftheentirecalibrationprocess.Theresultsofapplyingthiscalibrationmethodologytotestbuildingsaswellaspracticalguidelinesweresummarizedinafollowedcompanionpaper[,55].
Calibratedsimulationisusuallyaveryusefulapproachformea-suringthesavingsfromenergyconservationretrofitstoexistingbuildings.Inordertofindouthowmucheffortandresourcesarenecessarytoproduceasatisfactorymodelforretrofittingstud-ies,acomprehensivecalibrationmethodologywasdevelopedbyPedrinietal.[56].Thecalibrationisdividedintothreesteps,wherethreesimulationmodelswithsuccessivelyincreasinginputinfor-mationdetailsandaccuracyarecreated.Theinputsofeachsteparecomefrom(1)buildingdesignplansanddocumentation;(2)walk-throughandaudit;(3)end-useenergymeasurements,respectively.Bycomparingthesimulatedresultsofeachstepwiththeactualenergyuseofanexistingbuilding,thegainsinaccuracyandtheinfluenceofinputsareclearlycaptured.Atthefirststage,allinputsarefrombuildingplansanddocumentationwithoutanyvisittothesite,thesimulatedannualelectricityconsumptionis114%higherthantheactualconsumption.Thesecondsimulationmodel,withabetterdescriptionofthebuildingzonesandinternalloadparam-etersobtainedatauditstage,decreasedtheannualconsumptiondifferenceto0.1%.Atthethirdstage,thescheduleandinternalpowerdensityarewelltunedbytheresultsfromend-usemonitor-ing.Itshowsasimilardiscrepancyintheannualconsumption(0.2%)asthatatstagetwo.However,asshowninFig.3,theend-usedataprovidedbymodelsatstagetwoandstagethreeareverydifferent.Comparedwiththeend-useresultsatstagethreewhichwasvali-datedbymonitoredend-useprofiles,theenergyuseoflightingwassignificantlyoverestimatedatthefirsttwostagesanddecreaseswiththerefinementsofthemodelinputs.Thecalibrationresultsatstagetwoalsoindicatethatalthoughthesimulatedannualenergyuseofwholebuildingfitwellwiththeutilitybilldata,thereliabilityandaccuracyatend-uselevelstillcannotbeguaranteed.
Toreducetheiterativeeffortsandthedependenceontheexpe-riencesofcalibratorsinmanualcalibrations,SunandReddy[57]proposedageneralframeworkforanalyticcalibrationthathasafirmmathematicalandstatisticalbasis.Itisbasedontherecogni-tionthatacalibrationcanbetreatedasanoptimizationproblem.Theproposedmethodologyconsistsoffourdistinctprocesses:(1)Sensitivityanalysistoidentifyasubsetofstronginfluen-tialvariables.(2)Identifiabilityanalysistodeterminehowmany
0% 20% 40% 60% 80% 100%Step179%Plan3%18%Step270%Audit16%15%Step336%End-use31%33%Ligh ting
Equipment
CoolingFig.3.End-useresultsfromthreemodelingwithdifferentinputdetails[56].
parametersofthissubsetcanbetunedmathematicallyandwhichspecificonesarethebestcandidates.(3)Numericaloptimizationtodeterminethenumericalvaluesofthebestsubsetofparameters.(4)Uncertaintyanalysistodeducetherangeofvariationoftheseparameters.A“synthetic”exampleinvolvinganofficebuildingisusedtoillustratethattheproposedanalyticalprocedureisabletocorrectlytunetheparametersbacktotheirreferencevalues.
3.3.2.Dynamicinversemodels
Toidentifytheparametersoreventheconfigurationofinversemodels,thetrainingdataforsteady-statemodelsmaycomefromthedetailedsimulationorstatisticalsurveyofotherexistingbuild-ings.Whilefordynamicmodels,thetrainingdatashouldbebasedonmeasurementsofthebuildingconcerned.Therefore,calculat-ingtheenergyuseofabuildingusingdynamicinversemodelsmustinvolveinsitumeasurementsformodelidentification.Unlikesteady-stateinversemodels,dynamicmodelsarecapableofcap-turingdynamiceffectssuchasthermalmasswhichtraditionallyrequiresthesolutionofasetofdifferentialequations.Thedisad-vantagesofdynamicinversemodelsarethattheyareincreasinglycomplexandneedmoredetailedmeasurementsto“tune”themodel[58].
Typicalexamplesofdynamicinversemodelsincludeautore-gressivemovingaverage(ARMA)models,Fourierseriesmodelsandartificialneuralnetworkmodels[2].ARMAmodelsassumelinearrelationshipsbetweenthepresentandlaggedvaluesoftheresponsevariableandthoseofthedrivingterms.Thesemodelsarehelpfulforonlyshorttermpredictionsbasedonlongtermper-formancedata.Forinstance,Reddy[59]usedmultivariateARMAmodelstopredicttheindoorairtemperatureofthreeresidencesrelatedtotheoutdoortemperature,solarradiation,andwhole-houseelectricityconsumption.HourlyenergyuseincommercialbuildingsshowsperiodicvariationsindailyandannualcyclessuchthattheFourierseriesfunctionalformsaresuitableinmodelingthisbehavior.Abushakra[60]developedamethodtopredictandevaluatetheenergyperformanceoflargecommercialandinsti-tutionalbuildings.Usingthismethod,thelong-termpredictions(wholecoolingseason,forinstance)ofinternalloadsaremodeledbyFourierseriesmodels.Artificialneuralnetworks(ANN),afamilyofnon-linearregressiontechniques,areusedformodelingbuildingenergyuse.Abrahametal.[61]presentedamodelidentificationmethodtogenerateannualheatingandcoolingelectricitycon-sumption,throughtheuseofanANNmodel.
Anothertypeofdynamicinversemodels,whichcanrepresentthephysicalpropertiesofthebuildingsystem,isgraymodels.Graymodelscanusuallydescribethebehaviorsoftheconcernedsys-temandexplainthesystemphysically.Theparametersofmodels
884
S.Wangetal./EnergyandBuildings55(2012)873–888
areidentifiedusingmeasureddata.Richaleetal.proposedagraymodelmethodnamedHELP(HouseEnergyLabelingProcedure)tocertificateenergyuseforspaceheatinginsingle-familyhouses.HELPestablishesasimplephysicalmodeltocalculateanormalizedheatingannualconsumptionunderstandardclimatewithstandardoperations.Themodelparameters(e.g.UAandgA)arederivedfromacontinuousrecordingofinternaltemperatureinresponsetooutdoorclimateandinternalloadsbyidentificationanalysis[62].Dynamicinversemodelingtechniquesarealsousuallyemploydeterministicorsimplifiedphysicalmodelswithparametersiden-tifiedusingbuildingorsystemperformancedata.Usingsuchmodels,themodelisofphysicalnaturewhilethemodelparam-eterswhichrepresentcertainphysicalpropertiesareidentified.Comparingwithpuredatadrivingmodels(e.g.ANNmodels),thismethodrequireslessperformancedata.Inaddition,theidentifiedmodelsmayhaveacceptableaccuracyevenifwhenthemodelsareusedbeyondtherangecoveredbytheperformancedatausedforparameteridentificationorfitting.Wangetal.[63]developedadeterministicchillermodelforHVACsystemperformancesim-ulation.Usingthismodel,theparameterscanbeidentifiedwithverylimitedchillerperformancetestdataorusingcataloguedataatfullloadandpartialload.WangandXu[–66]proposedasimplifiedequivalentthermalnetwork(RCmodel),forlong-termenergyperformancepredictionofexistingbuildingsusingshort-termoperationdata.Itisaparameteridentification-basedmethod,whichconsistsoftwoparts.Onepartisthesimplifiedenergy3R2Cmodelsofbuildingenvelopes.Theotherpartisthesimplified2R2Cmodelforbuildinginternalmass.Parameteridentificationtech-niquesareemployedtodeterminethekeyparameters(e.g.thevaluesofRandC)ofbothmodels.Theparametersof3R2Cmodelaredeterminedbycomparingthefrequencycharacteristicsofthosesimplifiedmodelsandtheirtheoreticalfrequencycharacteristicsbasedontheeasilyavailabledetailedphysicalpropertiesofexte-riorwallsandroof.Theparametersofthebuildinginternalmassmodel(2R2C)areoptimizedusingshort-termmonitoredopera-tiondata.Ageneticalgorithm(GA)estimatorisusedtooptimallyidentifytheseparameters.
Insummary,theenergyuseinformationofanexistingbuildingcanbeobtainedviaacalculation-basedmethod,ameasurement-basedmethodorahybridmethod,withdifferentdetailsandcosts.Theselectionofanappropriatemethodiseventuallydeterminedbyhowtousethequantifiedenergyusedatatoassessenergyper-formanceofthebuildingconcerned[9].
4.Energyperformanceassessmentapproaches
Energyperformanceassessmentreferstoschemestojudgetheenergyperformanceofbuildingbycomparingwithrelevantbench-marks.Withquantifiedenergyuse,appropriatebenchmarkingcriteriaandcomparisonscaleareessentialforafairanduse-fulassessment.Accordingtothecomparativescopeanddepth,energyassessmentmethodcanbeclassifiedintotwomajorcat-egories,namelywhole-buildingbenchmarkingandhierarchicalassessment/diagnosis.
4.1.Whole-buildingbenchmarking
Whole-buildingbenchmarkingmethodisasimpleandeffec-tivemethodtoassesstheoverallbuildingenergyperformance,bycomparingthewhole-buildingenergyperformanceindexoftheassessedbuildingwithreferencebenchmarks.Toselectordevelopanappropriatereferencebenchmark,thekeyquestioniswhethertheenergyperformancedataofawidenumberofsamplebuildingsareavailable.Fortheaffirmativeanswer,thebenchmarkscanbeestablishedbystatisticalanalysisofsimilarbuildingstocks,i.e.statisticalbenchmark.Alternatively,thebenchmarkswillbe
establishedbycalculationfromahypotheticalreferencebuilding,i.e.calculatedbenchmark.
4.1.1.Statisticalbenchmark
Therearetwobasictypesofmodelstousetheenergydatabasetoestablishthestatisticalbenchmarks,i.e.thesimplenormal-izedmodelandtheregression-basedmodel.Usingthesimplenormalizedmodel,theenergyperformanceindicator(e.g.floor-area-normalizedEUI)isusuallyobtainedbynormalizingtheenergyusewithfloorareawhichaccountsforonlyonebuildingfeaturethataffectsenergyconsumption.Toaccountfortheeffectofotherenergyinfluentialfactors,benchmarkshavebeenconstructedbyusingregression-basedmodel.
Cal-Archisarepresentativewhole-buildingbenchmarkingtoolusingasimplenormalizedmodel.Itassessestheoverallenergyperformanceofanexistingbuildingbycomparingthefloor-area-normalizedEUIwiththebenchmarktableofsimilarbuildingsintheenergydatabaseofCalifornia’s1992CommercialEndUseSurvey(CEUS)[67].Thebenchmarktableisthepercentagedistribution(afrequencyhistogram)ofEUIsforasubsetofbuildings(thesametypebuildings)intheCal-Archdatabase.TheEUIoftheassessedbuildingisnotedwithanarrowpointingtothecorrespondingposi-tioninthehistogram,whichgraphicallyshowshowabuilding’sEUIfitswithinthestatisticaldistribution.Theassessmentresultisnotascorebutarelativerankingwhichinformsthepercentageofbuildingsinthedatabasethathavehigher/lowerEUIsthantheassessedbuilding.InChina,asimilarbenchmarkingsystemwasproposedtoassesstheenergyperformanceofexistinglargepub-licbuildings[68].Thissystemrankstheenergyperformanceofagivenbuildingbythecomparisonofarea-normalizedEUIwithsimilarbuildingsinthesameclimatezone.AbuildingwithlowerEUIvaluewillbeawardedwithhigherscoresandbeconsideredtobemoreenergyefficientthanothers,viceversa.Forabovetwosimplenormalizedbenchmarkingsystems,thecomparisondataforbothassessedbuildingsandreferencebuildingsaretheactualEUIswithoutanyadjustmentforweatheroranyotherfactors.
Withoutaccountingforsufficientenergy-relatedfactors,therearelimitationsofusingthesimplenormalizedEUIforenergyper-formancebenchmarking,suchasusersshouldonlycomparetheirbuilding’sEUItothatofsametypebuildingsinthesameclimatezone,otherwisethecomparisonismeaningless.Inaddition,theusagecharacteristicsandoccupancyfactorshaveagreatimpacttodeterminetheenergyuseofabuilding.Ignoringofthesefactorswillweakenthefairnessandcredibilityofanenergyperformancebenchmarkingtool.Therefore,adjustmentsofclimateandotherfactorsintobenchmarkingprocessarerecommendedbysomeresearchers.Forinstance,SharpdevelopamethodtocorrelatesomeimportantcharacteristicsofbuildingusewithEUIusingamultivariatelinear-regressionapproach.ThismethodwasadoptedinAsia-PacificEconomicCooperationEnergyBenchmarkSystemwhichhasbeenslightlymodifiedtobethebasisoftheEnergyStarbenchmark[35,36].
EnergyStarisaregression-basedbenchmarkingtool,basedonbuildingcharacteristicandenergyusedatafromtheDOE/EIACom-mercialBuildingEnergyConsumption(CBECs)survey.UsingtheobserveddataofCBECs,abuilding-specificregressionmodelisdevelopedforeachapplicabletype(e.g.office,grocery,schools,hotel,andhospitalbuildings)ofbuildings.Usingthismodel,theannualenergyconsumptioncanbepredictedwithgivenbuildingcharacteristicsandoccupancyfactors(e.g.grossfloorarea,occu-pantnumber,thenumberofcomputersandoperationhours).Theprocesstoassesstheenergyperformanceofagivenbuilding(e.g.BuildingA)inEnergyStarprogramconsistsoftwosteps.ThefirststepistomaptheEUIson1–100ratingscaleforacertaintypeofbuildings.TheresultingregressionmodelisusedtopredictEUIsforeachobservationinthereferencedatasetandtheseEUIsare
S.Wangetal./EnergyandBuildings55(2012)873–888
885
mappedontoa1–100scale(i.e.themappedratingscale).ThesecondstepistousethemappedEUIsratingscaletoassesstheper-formanceofBuildingA.TheEUIofBuildingAisalsopredictedusingthesamebuilding-specificregressionmodel.ThepredictedEUIisusedtoadjustthemappedEUIsratingscaleintoa“customizedratingscale”usingacorrectionfactor.Ontheotherhand,themea-suredEUIofBuildingAisweathernormalizedbaseda30-yearaverageweatheryear.Theenergyperformance(i.e.theEnergyStarscore)ofBuildingAisfinallydeterminedbycomparingtheweather-normalizedEUIwiththe“customizedratingscale”.Buildingsthatscore75(orabove)andmeettheindoorenvironmentcriteriarequirementsareeligibletoobtaintheEnergyStarLabel[67].
4.1.2.Calculatedbenchmark
Adifferentapproachforgeneratingdatabaseistoapplybuildingenergysimulationtoavarietyofbuildingtypesforarangeofenergyparameters.However,sincethebenchmarkforacertainbuildinghastobeestablishedbysimulation,creatingonlyonecustomizedself-referencebuildingisdefinitelymucheasierthangeneratingavarietyofdatasets.Asaresult,usingaself-referencebuildingasthecomparisoncriterionispreferabletoestablishingabenchmarkfromasimulateddatabase.Aself-referencebuildingisalsocalledasa“notionalbuilding”,whoseshape,geometricaldimensionsandfunctionallayoutaretotallyidenticaltotheassessedbuilding[8].
Thisself-referencebenchmarkingmethodhasbeenwidelyadoptedinvariouscountriesforcompliancejudgmentofnewbuildingsandhasbeguntobeusedtoassesstheenergyper-formanceofbothnewandexistingbuildingsinmanyenergycertificationandenvironmentalassessmentschemes.Itdoesnotrequireadatabaseforthecomparison,norastatisticalanaly-sisofthebuildingstockcomparisonscenario[9].Generally,theperformance(gradesorcredits)ofanassessedbuildingisdeter-minedbythepercentagereductionofannualenergyuseinrelationtotheself-referencebuilding.Fig.4showsfivedifferentassess-mentscalesforclassifyingtheenergyperformancebasedontheenergy(orequivalentindicators)savingpercentagesofacertainself-referencedbuilding.
4.2.Hierarchicalassessmentanddiagnosis
Hierarchicalassessment(i.e.multi-levelassessment)referstotheassessmentofenergyperformanceextendingfrombuildingleveltosystemand/orcomponentlevel.Thismethodismoreeffec-tivetoidentifyanddiagnoseproblematicareasandtoprovidemorespecificenhancementrecommendations,bytakingmorefeaturesaffectingenergyuseintoaccount.
Usingstatisticalwhole-buildingbenchmarkingmethod,theenergyperformanceisjudgedagainstsimilarbuildings.However,sometimes,itisdifficulttodefine“similarbuildings”foracomplexbuildingwhichcomprisesamixpremises,includingoffices,restau-rantsandretailshops.Thesepremisessharethesamecentralizedbuildingservicesbutdiffersignificantlyinenergyuse.Inthiscase,theenergyperformanceshouldbeassessedhierarchicallybytakingaccounttheenergyusecharacteristicsofdifferenttypesofpremises.Leeetal.proposedahierarchicalmethodtoassesstheenergyperformanceofsuchexistingcomplexbuildingsin[8].
Inthismethod,theenergyperformanceofwholebuildingisseparatedintotenantsideandlandlordside.Thetenantsideper-formanceisfurtherdividedintoindividualpremisesfordetailedassessment.Foracertainpremises(e.g.anofficepremises),theenergy-billedconsumptionisdisaggregatedandnormalizedintoEUIsofprincipalend-uses(lighting,officeequipmentandmiscel-laneousequipment).ThecorrespondingEUIsarethencomparedagainstperformancescaleswithsimilarpremisesindatabase.Eachend-useobtainsanumericscoreaccordingtothecompar-isonanalysis.Asetofweightingsaregiventothescoresintheassessmentofvariousend-usesforthecomputationofaweightedsumofscores,whichreflecttherelativeenergyperformanceoftheassessedpremises.Atlandlordside,themethodforassessingthelandlord’senergyperformancecanbedividedintotwolev-els.Level-1assessmentisforbuildingsequippedwithsufficientsub-meterstoallowthetotalelectricityconsumptiontobebro-kendownintoindividualcentralservices.Theperformancesofthesecentralservicesareassessedseparatelybycomparingtheirperformanceindicatorswithcorrespondingbenchmarks.Level-2assessmentisforbuildingswithoutsub-meterswheretheenergyperformanceofallcentralservicesystemswouldbeassessedtogether.
Thishierarchicalassessmentmethodisnotonlymorefairthanthemethodbenchmarkingthewholecomplexasawholebutalsomoreeffectivetoidentifywhetherthepoorperformanceofthisbuildingismainlycausedbytenantsideorlandlordside.Ifcausedbytenantside(landlordsideissimilar),thenitcanbefurtherlocatedtoacertainpremisesoreventoaspecificend-user.Suchinformationissignificantlyvaluablesothatmoreeffortscanbepaidonthepartswithlargestenergysavingpotential.
Inasingle-typebuilding,hierarchicalassessmentisalsoveryusefulfordetailedassessmentandspecificdiagnosis.Ahierarchi-calassessmenttreecomprisingofenergyperformanceindicesofdifferenttypesofend-use,wasasdevelopedbyFieldetal.toassesstheperformanceofbuildingsystems[20].AsshowninFig.5,the
AMHEBE DLEEBREE AM0123456710012345671024610 11 12 13 14 15CALENERCBACEN0%
10%
20%
B30% 40% 50% 60%
Energy Sav ing Percent ageA70%
80%
90%
100%Fig.4.Energyperformanceassessmentscalesusingself-referencebenchmarks[9,41].
886
S.Wangetal./EnergyandBuildings55(2012)873–888
Whole- bui lding EUI (Pri mary or CO2 equivalen t)Buildin glevelLighting kWh /m2Ventilat ion kWh/ m2Other systemsSystemlevelLig hting W/ m2Hours/y rVentilati on W/m2Hours/yrCompon entlevelLightlevelLuxEfficiency(W/m2)/1 00 lxVen rat e(l/s) /m2EfficiencyW/(l/s)Fig.5.Treediagramofenergyperformanceindicesformulti-levelassessment[20].
energyperformanceisassessedatmulti-levels.Theoverallenergyperformanceatbuildinglevelisindicatedandassessedbywhole-buildingEUIvalue.Atsystemlevel,thetotalenergyuseofwholebuildingisbrokendownintotheconsumptionsofindividualsys-tems.Theperformanceofeachsystemcanbeeasilydeterminedbycomparingtheconsumptionvaluewithreferencevalue.Toobtainmoredetailsandinterprettheperformanceofacertainsystem,forexamplethelightingsystem,theassessmentshouldbeextendedtocomponentlevel.Atthislevel,theinfluentialfactorsoftheper-formanceofalightingsystemareclassifiedastechnicalaspect(lightingintensity)andmanagementaspect(operationalhours).Intechnicalaspect,theindividualcontributionsfromthelightlevelandlightingefficiencycanbefurtheridentified.Suchinterpreta-tioncanusedtodetectanddiagnosetheexactlyenergyproblemsforaspecificfacility.
Usingmulti-levelassessment,top-downapproachisusuallyusedtoassessanddiagnosetheenergyperformancesprogressively.EARM-OAM(EnergyAssessmentandReportingMethodology-OfficeAssessmentMethod)isatypicalprogressivelydetailedmulti-levelassessmentmethod,whichconsistsofthreestages,i.e.initialstage,intermediatestageandadvancestage.Atinitialstage,abroad-brushassessmentusingsimplefossilfuelandelectricitycon-sumptionisusedtojudgetheperformanceatbuildinglevel.Theassessmentresultalsodetermineswhethermoredetailedinvesti-gationisnecessaryandworthy.Iftheanswerisyes,theassessmententerstheintermediatestagewheresomekeyinfluentialfac-tors,suchasextendedoccupancy,extremeweatherorunusuallyenergyend-usewillbeconsideredtointerprettheassessperfor-manceresult.Theassessmentprocesscanbeterminatedifabovementionedfactorscanclearlyinterprettheresultsatthisstage.Otherwise,moredetailedend-useanalysisneedstobeundertakenatadvancedstage.Theexactcausesofpoorperformancecanbediagnosedandspecificsuggestionsonenhancingenergyefficiencycanberecommended[19].
TheInternationalPerformanceMeasurementandVerificationProtocol(IPMVP)alsodevelopedageneralprocedureforselect-inganappropriateapproachtoquantifytheenergysavingsinaretrofittingproject.Inthisprocedure,theenergyperformancequantificationandassessmentwillbeincreasinglydetailedstepbystep:(a)ingeneral,onecantrytoperform“MonthlyUtilityBillBefore/AfterAnalysis”,(b)Ifthisdoesnotwork,heperforms“DailyorHourlyBefore/AfterAnalysis”,(c)Ifthisdoesnotworkagain,heperforms“ComponentIsolationAnalysis”,(d)Ifthisdoesnotworkneither,heperforms“CalibratedSimulationAnalysis”[58].Inthisway,themostcosteffectiveofmeasurementandverifica-tionmethodcanbeguaranteedsincethecomplicatedandhigh-cost
approachesarenotinvolveduntilthelesscostlyapproachescannotbeused.
5.Discussionandconclusions
Thequantitativeenergyperformanceassessmentforexistingbuildingsconsistsofaquantificationprocessoftheenergyuseandacomparison(judgment)processofperformanceindicators.Theeffectivenessandcredibilityofanassessmentaremainlydeter-minedbythesetwoprocesses.
Forenergyquantification,usingsimulationtoolsmightbe,inprinciple,themostpowerfulmethodsbyprovidingabundantanddetailedoutputs.However,theuseofsimulationtoolstocalcu-latetheenergyperformanceinexistingbuildingisusuallynotverycost-effectiveinpractice.Thebiggestchallengeiscollectionofper-formancedataandsystemparameters.Theavailabilityofdataisoftenproblematicinexistingbuildings.
Usingmeasurement-basedmethods,energybillsgiveaneasyaccesstotheoverallperformanceatbuildinglevel.Foramoredetailedassessment,itinvolvesadisaggregationprocesstoestablishasplitoftotalenergyintoenduses.Hardware-basedsub-meteringsystemcanofferaccurateandabundantend-usecon-sumptioninformationwithrelativehighcost.Non-intrusiveloadmonitoringmethodiscapableofgatheringdetailedenergy-usedatawithlesscostwhileitstillfacesmanychallengeswhenusedforcomplexbuildings.
Hybridmethodscombinethefeaturesofcalculation-basedandmeasurement-basedmethods.Theymayprovidemoreflexibilityinquantifyingtheenergyuseofexistingbuildings.However,themanyavailablehybridmethodsgenerallyusethecalculationandmeasurementina“parallel”ratherthanan“integrated”approach.Measurementsareoftennotinvolvedinthesimulationprocesseveninthecasesofcalibratedsimulationsexceptthecasesofdevel-opingidentifiedinversemodels.
Therearevariousenergyclassificationschemesavailabletoassesstheoverallenergyperformanceofexistingbuildingsusingwhole-buildingbenchmarkingmethod.However,thesystematicmulti-levelenergyperformanceassessment/diagnosismethodsareunfortunatelyverylimited.Inpractice,themulti-levelenergyperformanceassessmentanddiagnosisarenormallyconductedmanually.Lackofgeneric,effectiveanduser-friendlytoolsforpracticalenergyperformanceassessmentanddiagnosisisaseri-ouslimitationforimplementingenergyenhancementmeasuresinexistingbuildings.Futureresearchonenergyperformanceassess-mentofexistingbuildingsshouldpaymoreeffortsindevelopingthesystematicandeffectiveassessmentanddiagnosismethods.
S.Wangetal./EnergyandBuildings55(2012)873–888
887
References
[1]DOE,2010Buildingsenergydatabook,Officeofenergyefficiencyandrenew-ableenergy,USDepartmentofEnergy,2011.
[2]XuXinhua,Model-basedbuildingperformanceevaluationanddiagnosis,Ph.D.
Dissertation,DepartmentofBuildingServicesEngineering,ThePolytechnicUniversity,April,2005.
[3]D.Claridge,C.Culp,M.Liu,etal.,Campus-widecontinuouscommissioning
ofuniversitybuildings,in:ProceedingsofACEEESummerStudyonEnergyEfficiencyinBuildings,PacificGrove,CA,August20–25,2000,2000,pp.3.101–3.112.
[4]UnitedNationsEnvironmentalProgram(UNEP)andUnitedStatesEnvironmen-talProtectionAgency(EPA),Reducinggreenhousegasemissions:Theroleofvoluntaryprograms,UnitedNationsEnvironmentalProgramPublication,1997.[5]B.Poel,G.vanCruchten,C.A.Balaras,Energyperformanceassessmentofexist-ingdwellings,EnergyandBuildings39(4)(2007)393–403.
[6]EN15217,Energyperformanceofbuildings—Methodsforexpressingenergy
performanceandforenergycertificationofbuildings,2007.
[7]IEAECBCSAnnex53,Annex53TotalEnergyUseinBuildings:Analysis&Eval-uationMethods,www.ecbcsa53.org
[8]W.L.Lee,F.W.H.Yik,J.Burnett,Amethodtoassesstheenergyperformanceof
existingcommercialcomplexes,IndoorandBuiltEnvironment12(5)(2003)311–327.
[9]L.Pérez-Lombard,J.Ortiz,R.González,I.R.Maestre,Areviewofbenchmarking,
ratingandlabellingconceptswithintheframeworkofbuildingenergycertifi-cationschemes,EnergyandBuildings41(2009)272–278.
[10]CEN,Energyperformanceofbuildings—overallenergyuseanddefinitionof
energyratings,EN15603:2008,EuropeanCommitteeforStandardization,2008.
[11]U.S.GreenBuildingCouncil(USGBC),LEED—GreenBuildingRatingSystemfor
ExistingBuildings:Operations&Maintenance(LEED-EBTM),January2008.[12]BREEAM,BREEnvironmentalAssessmentMethod,BuildingResearchEstablish-mentUK.http://www.breeam.org/
[13]HK-BEAMSocietyHK-BEAM—anenvironmentassessmentforairconditioned
officepremises.Version4/04:newbuildingsandversion5/04:existingbuild-ings.BuildingEnvironmentCouncil,,2004.
[14]V.Corrado,H.E.Mechri,E.Fabrizio,Buildingenergyperformanceassessment
throughsimplifiedmodels:applicationoftheISO13790quasi-steadystatemethod,Proceedingsofthe10thConferenceoftheInternationalBuildingPer-formanceSimulationAssociation,3-6September(2007),TsinghuaUniversity,Beijing,China,79-86.
[15]M.J.McCabe,N.Wang,CommercialBuildingEnergyAssetRatingProgram—
MarketResearch,PNNL-21310,PacificNorthwestNationalLaboratory,Rich-land,WA,2012.
[16]R.E.Jarnagin,ASHRAEBuildingEQ:Programwillhelpowners,operatorsassess
buildingsandguidegooddecisions,ASHRAEJournal51(12)(2009)18–19.[17]EnvironmentalProtectionAgencyandtheU.S.DepartmentofEnergy,ENERGY
STARprogram,1992.
[18]CaliforniaEnergyCommission,CAL-ARCH(CaliforniaEnergybenchmarking
tool),2003.
[19]RobertCohen,MarkStandeven,BillBordass,AdrianLeaman,Assessingbuilding
performanceinuse1:theProbeprocess,BuildingResearch&Information29(2)(2001)85–102.
[20]A.J.Field,J.Soper,P.Jones,W.Bordass,P.Grigg,Energyperformanceofoccupied
non-domesticbuildings:Assessmentbyanalysingend-useenergyconsump-tion,BuildingServicesEngineeringResearchandTechnology18(1)(1997)39–46.
[21]H.Akbari,Validationofanalgorithmtodisaggregatewhole-buildinghourly
electricalloadintoenduses,Energy20(December(12))(1995)1291–1301.[22]J.P.Waltz,ComputerizedBuildingEnergySimulationHandbook,TheFairmont
Press,Lilburn,GA,2000.
[23]M.S.Al-Homoud,Computer-aidedbuildingenergyanalysistechniques,Build-ingandEnvironment36(2001)421–433.
[24]J.M.Ayres,E.Stamper,Historicaldevelopmentofbuildingenergycalculation,
ASHRAETransactions:Symposia101(1995)841–848.
[25]B.Birdsall,W.F.Buhl,K.L.Ellington,A.E.Erdem,F.C.Winkelmann,Overviewof
theDOE-2buildingenergyanalysisprogram,Version2.1D,LawrenceBerkeleyLaboratoryLBL19735-Rev.1,1990,Berkeley,CA.53pp.
[26]D.B.Crawley,L.K.Lawrie,F.C.Winkelmann,W.F.Buhl,Y.J.Huang,C.O.Peder-sen,etal.,EnergyPlus:creatinganew-generationbuildingenergysimulationprogram,EnergyandBuildings33(4)(2001)319–331.
[27]BuildingSystemsLaboratory,BLAST3.0UsersManual,DepartmentofMechan-icalandIndustrialEngineering,BuildingSystemsLaboratory,UniversityofIllinois,Urbana-Champaign,IL,1999.
[28]SolarEnergyLaboratory,TRNSYSVersion15UserManualandDocumentation,
SolarEnergyLaboratory,MechanicalEngineeringDepartment,UniversityofWisconsin,Madison,WI,February2000.
[29]D.VanDijk,M.E.Spiekman,P.DeWilde,Amonthlymethodforcalculating
energyperformanceinthecontextofEuropeanbuildingregulations,in:Build-ingSimulation2005ProceedingsoftheNinthInternationalIBPSAConference,Montreal,15–18August,2005.
[30]V.Corrado,E.Fabrizio,Assessmentofbuildingcoolingenergyneedthrougha
quasi-steadystatemodel:simplifiedcorrelationforgain-lossmismatch,EnergyandBuildings39(5)(2007)569–579.
[31]J.Sjösten,T.Olofsson,M.Golriz,Heatingenergyusesimulationfor
residentialbuildings,in:ProceedingsoftheEighthInternationalIBPSA
Conference,11–14August,2003,Eindhoven,Netherlands,2003,pp.1221–1226.
[32]
J.Jokisalo,J.Kurnitski,PerformanceofENISO13790utilisationfactorheatdemandcalculationmethodinacoldclimate,EnergyandBuildings39(2)(2007)236–247.
[33]
ENISO13790:Energyperformanceofbuildings—Calculationofenergyuseforspaceheatingandcooling(ENISO13790:2008),EuropeanCommitteeforStandardization(CEN),Brussels,2008.
[34]
T.Olofsson,S.Andersson,J.-U.Sjögren,Buildingenergyparameterinvestiga-tionsbasedonmultivariateanalysis,EnergyandBuildings41(1)(2009)71–80.[35]
T.Sharp,Energybenchmarkingincommercialofficebuildings,in:Proceedingsofthe1996ACEEESummerStudyofEnergyEfficiencyinBuildings,vol.4,AmericanCouncilforAnEnergy-EfficientEconomy,Washington,DC,1996,pp.321–329.
[36]
W.Chung,Y.V.Hui,Y.M.Lam,Benchmarkingtheenergyefficiencyofcommer-cialbuildings,AppliedEnergy83(2006)1–14.
[37]
K.Kissock,J.Haberl,D.Claridge,DevelopmentofaToolkitforCalculatingLin-ear,Change-pointLinearandMultiple-LinearInverseBuildingEnergyAnalysisModels:FinalReport,submittedtoASHRAEunderResearchProject1050-RP,UniversityofDaytonandEnergySystemsLaboratory(December),2001.
[38]
J.C.Lam,K.K.W.Wan,D.Liu,C.L.Tsang,Multipleregressionmodelsforenergyuseinair-conditionedofficebuildingsindifferentclimates,EnergyConversionandManagement51(2010)2692–2697.
[39]
W.L.Lee,W.H.Yik,J.Burnett,Simplifyingenergyperformanceassessmentinenvironmentalassessmentmethod,BuildingServicesEngineeringResearchandTechnology2(2)(2001)113–132.
[40]
D.E.Knebel,Simplifiedenergyanalysisusingthemodifiedbinmethod,ASHRAE(1983)1-1–1-6.
[41]
W.L.Lee,F.W.H.Yik,J.Burnett,Assessingenergyperformanceinthelatestver-sionsofbuildingenvironmentalassessmentmethod(HK-BEAM),EnergyandBuildings39(3)(2007)343–3.
[42]
DavidRobison,PacificPower,Theuseofshort-termmeasurementstodecom-posecommercialbillingdataintoprimaryenduses,ACEEE1992SummerStudyonEnergyEfficiencyinBuildings,vol.3,1992,ACEEE,Washington,DC.
[43]
J.A.Heidell,R.P.Mazzucchi,R.W.Reilly,Commercialbuildingend-useenergymeteringinventory,PacificNorthwestLaboratoryReportPNL-5027.
[44]
C.H.Philip,W.K.Yu,Chow,Adiscussiononpotentialsofsavingenergyuseforcommercialbuildingsin,Energy32(February(2))(2007)83–94.[45]
MaryAnnPiette,SatKartarKinney,PhilipHaves,Analysisofaninformationmonitoringanddiagnosticsystemtoimprovebuildingoperations,EnergyandBuildings33(October(8))(2001)783–791.
[46]
G.W.Hart,Nonintrusiveapplianceloadmonitoring,ProceedingsofIEEE80(1992)1870–11.
[47]H.Pihala,Non-intrusiveApplianceLoadMonitoringSystemBasedonaModernkWh-meter,VTTPublications,1998.
[48]
M.L.Marceau,R.Zmeureanu,Nonintrusiveloaddisaggregationcomputerpro-gramtoestimatetheenergyconsumptionofmajorendusesinresidentialbuildings,EnergyConversionandManagement41(September(13))(2000)13–1403.
[49]
K.Leslie,Norford,B.Steven,Leeb,Non-intrusiveelectricalloadmonitoringincommercialbuildingsbasedonsteady-stateandtransientload-detectionalgorithms,EnergyandBuildings24(1)(1996)51–.
[50]
M.Masoero,C.Silvi,J.Toniolo,EnergyperformanceassessmentofHVACsys-temsbyinspectionandmonitoring,in:Proceedingsof10thREHVAWorldCongressClimate2010,Antalya,9–12May2010,2010.
[51]
T.Reddy,Literaturereviewoncalibrationofbuildingenergysimulationprograms:uses,problems,procedures,uncertainty,andtools,ASHRAETrans-actions112(2)(2006)226–240.
[52]
D.Claridgeetal.,MethodologyforCalibratingSimulations,submittedtoAnnex40,April,2004.
[53]
D.E.Claridge,Usingsimulationmodelsforbuildingcommissioning,in:Interna-tionalConferenceonEnhancedBuildingOperation,EnergySystemsLaboratory,TexasA&MUniversity,2004.
[]
T.A.Reddy,I.Maor,C.Panjapornpon,Calibratingdetailedbuildingenergysim-ulationprogramswithmeasureddata-partI:generalmethodology(RP-1051),HVAC&RResearch13(2)(2007)221e41.
[55]
T.A.Reddy,I.Maor,C.Panjapornpon,Calibratingdetailedbuildingenergysim-ulationprogramswithmeasureddata-PartII:applicationtothreecasestudyofficebuildings(RP-1051),HVAC&RResearch13(2)(2007)243e65.
[56]
A.Pedrini,F.S.Westphal,R.Lamberts,Amethodologyforbuildingenergymod-elingandcalibrationinwarmclimates,BuildingandEnvironment37(2002)903–912.
[57]
T.JianSun,AgamiReddy,Calibrationofbuildingenergysimulationprogramsusingtheanalyticoptimizationapproach(RP-1051),HVAC&RResearch12(1)(2006)177–196.
[58]
J.S.Haberl,C.H.Culp,2005,Reviewofmethodsformeasuringandverifyingsav-ingsfromenergyconservationretrofitstoexistingbuildings,EnergySystemsLaboratoryReportESL-TR-03/09-01,EnergySystemsLaboratory,TexasA&MUniversity,CollegeStation,TX.
[59]
T.Reddy,Applicationofdynamicbuildinginversemodelstothreeoccu-piedresidencesmonitorednon-intrusively,in:ProceedingsoftheASHRAEDOE/BTECC/CIBSEconference,ThermalperformanceoftheexteriorenvelopesofbuildingIV,December4–7,Orlando,FL,19,pp.6–671.
[60]
B.Abushakra,AnInverseModeltoPredictandEvaluatetheEnergyPerformanceofLargeCommercialandInstitutionalBuildings,ConcordiaUniversity,Quebec,1999.
888
S.Wangetal./EnergyandBuildings55(2012)873–888
[61]A.Yezioro,B.Dong,F.Leite,Anappliedartificialintelligenceapproachtowards
assessingbuildingperformancesimulationtools,EnergyandBuildings40(4)(2008)612–620.
[62]V.Richalet,etal.,HELP(houseenergylabelingprocedure):methodologyand
presentresults,EnergyandBuildings33(3)(2001)229–233.
[63]S.W.Wang,J.Wang,J.Burnett,Mechanisticmodelofcentrifugalchillersfor
HVACsystemdynamicssimulation,BuildingServicesEngineeringResearchandTechnology21(2)(2000)73–83.
[]S.W.Wang,X.H.Xu,Parameterestimationofinternalthermalmassofbuilding
dynamicmodelsusinggeneticalgorithm,EnergyConversionandManagement47(13–14)(2006)1927–1941.
[65]S.W.Wang,X.H.Xu,Simplifiedbuildingmodelfortransientthermalper-formanceestimationusingGA-basedparameteridentification,InternationalJournalofThermalSciences45(4)(2006)419–432.
[66]X.H.Xu,S.W.Wang,Amixed-modebuildingenergymodelforperformance
evaluationanddiagnosisofexistingbuildings,BuildingServicesEngineeringResearchandTechnology29(1)(2008)73–83.
[67]N.Matson,M.A.Piette,ReviewofCaliforniaandnationalmethodsforenergy
performancebenchmarkingofcommercialbuildings,LawrenceBerkeleyNationalLaboratory,LBNLNo.573,2005.
[68]XueZhifeng,Studiesonenergyefficiencyforlarge-scalecommercialbuildings,
Ph.D.Dissertation,Beijing,TsinghuaUniversity,October2005.
因篇幅问题不能全部显示，请点此查看更多更全内容
Copyright © 2019- awee.cn 版权所有 湘ICP备2023022495号-5
违法及侵权请联系：TEL:199 1889 7713 E-MAIL:2724546146@qq.com
本站由北京市万商天勤律师事务所王兴未律师提供法律服务