天津科技大学外文翻译

天津科技大学电气工程及其自动化毕业设计的外文翻译样板

RenewableEnergy57(2013)587e593

ContentslistsavailableatSciVerseScienceDirect

RenewableEnergy

journalhomepage:w

http://wendang.chazidian.com/locate/renene

Electricgriddependenceonthecon gurationofasmall-scalewindandsolarpowerhybridsystem

M.Mikatia,M.Santosb,*,C.Armentac

ÅFTechnology,Lindholmspiren9,41756Gothenburg,Sweden

DepartamentodeArquitecturadeComputadoresyAutomática,FacultaddeInformática,UniversidadComplutensedeMadrid,C/ProfesorGarcíaSantesmases,s/n,28040Madrid,Spainc

DepartamentodeFísicaAtómica,Molecular,yNuclear,FacultaddeCienciasFísicas,UniversidadComplutensedeMadrid,CiudadUniversitaria,s/n,28040Madrid,Spain

ba

articleinfo

Articlehistory:

Received16October2012Accepted18February2013

Availableonline25March2013Keywords:

RenewableenergyHybridsystemWindpowerSolarpowerMicrogrid

Energydistributionef ciency

abstract

Resultsarepresentedfromthemodelingofasmall-scaledistributedpowersystemcontainingpowerdemands,photovoltaicarrays,small-scalewindturbinesandanelectricgridconnection.Detailedmodelsofthephotovoltaicarrayandthewindturbinearebrie yexplainedinadditiontothesolarandwindrecoursemodels.Asubunitisde nedtoconsistofapowerdemandtogetherwithpowercontributors.Itisshownhowthegriddependency(orrenewableenergycontribution)isaffectedbytheconnectionofsubunitsandaccordingtotherelationshipbetweenthepowerdemandpatternsandrenewableresourcepatterns.Theoutcomefromseveralcasestudiesispresentedusingsimulatedpowerdemandsandrenewableresources.Inascenariowithsubunitpowerdemandscharacteristicforalargehouseholdandasmallfactory,itisshownthatthecouplingofsubunitsreducesannualgridpowertransfersbymorethan10%andincreasestherenewablepowercontributiontothedemandbyalmost7%.

Ó2013ElsevierLtd.Allrightsreserved.

1.Introduction

Windandsolartechnologyhaveseenarapidgrowthduringthelastdecades.Theinstalledworldcapacityisincreasingexponen-tiallyandproductioncostsarecontinuouslyreduced[1,2].Spain,forexample,hadbytheendof2011awindandsolarpowercontributiontoelectricdemandsofrespectively16%and3%[3],andwindpowerinDenmarkrepresentedalmost26%ofelectricde-mands[4].Moreover,countrieswithlargehydropowerresourceshaverenewableenergycontributionsfarbeyond20%.

Amajorportionofcurrentrenewablepowerproductionisachievedbycentralizedlarge-scalegeneration.However,therearemanyfactorsindicatingthatafuturepowersystemshouldconsistofamixtureofnumeroussmall-scaledistributedpowergeneratorsconnectedintoamicrogrid(orsmartgrid)togetherwiththecentralizedgeneration.Thecentralizedgeneratorsandthemicro-gridclustersareconnectedthroughthemainelectricgrid.

Thebene tsofmicrogridsincludeimprovedoverallsystemreliability,reducedcentralgenerationreserverequirements,localvoltagesupportandreducednetworklosses[5e7].Inaddition,theinvestmentcostofsmall-scalepowergeneratorsmakesitrealistic

*Correspondingauthor.

E-mailaddresses:monir.mikati@http://wendang.chazidian.com(M.Mikati),msantos@dacya.ucm.es,msantospenas@http://wendang.chazidian.com(M.Santos),cardeu@ s.ucm.es(C.

Armenta).0960-1481/$eseefrontmatterÓ2013ElsevierLtd.Allrightsreserved.http://wendang.chazidian.com/10.1016/j.renene.2013.02.018

forhomeownerstobecomepowerproducerswhichmayincreaseconsciousnessforenergyuse.

ForageneralreviewofmicrogridsseeRefs.[5,8],seealsoRef.[6]regardingthecontrolofmicrogrids.InRef.[7]numerouscasestudiescanbefoundandinRefs.[9,10]microgridsandsmartgridsareexaminedinaBritishandaGermanscenario,respectively.

Thisarticleputsfocusonthegainsofconnectingsmall-scaledistributedpowersystemsincomparisontoindependentcells.Theanalysisbecomesvaluablewhenconsideringthedesignofdistributedpowersystemsonacommunitylevel.Italsoworksasanincentiveforthecouplingofdistributedpowercells.

Withthegrowthrenewabletechnologiessuchasgrid-connectedphotovoltaic(PV)arrays,anincreasedamountofsmall-scaledistributedgenerationisbeingintroducedintothepowersystem.SomecountrieshaveevenacceptedbuildingcodeswhichdemandthatPVarraysandsolarwaterheatingsystemsmustbeinstalledwhenconstructingnewbuildings[2,11].

Thereareseveralreasonswhydistributedpowergeneratorsshouldavoidthemainelectricgrid.Thekeyargumentsinthecontextofthisworkarethelongdistancedistributionlossesofthemaingriditselfandthevoltagetransformationlossesbetweenthedistributedpowercellsandthemaingrid.Onlythelongdistancedistributionlossesare5e20%[1].

Inthisstudy,thepowersystemshowninFig.1isinvestigatedintwocon gurations.Eachcon gurationconstitutesoftwosubsystems:S1andS2.Everysubsystemconsistsofapowerdemand,acontrol

天津科技大学电气工程及其自动化毕业设计的外文翻译样板

588M.Mikatietal./RenewableEnergy57(2013)587e593

I

ndependentconfiguration

Coupledconfiguration

S2

S2

Fig.1.Schematicofthetwosystemcon gurations:left,independent;right,coupled.Thearrowsmarkthedirectionofthe owofelectricpower.

system,asmall-scalewindturbineandaPVarray.Aconnectiontotheelectricgridisassumedpossible,aswellasthepossibilitytodeliverexcesspowertothegrid.Theonlydifferencebetweenthecon gu-rationsisthatthesubsystemscanexchangepowerinthecoupledcon guration(Fig.1,right).Thisisnotpossibleintheindependentcon guration(Fig.1,left),whosesubsystemshavetoexchangealltheirpowerdirectlywiththegrid.ThecoupledpowersysteminFig.1canbeseenasthesimplesthybridpoweredmicrogridpossible.

Thepurposeoftheanalysisistocomparethepowertransfersoftheindependentandthecoupledsystemwiththemaingridandtoinvestigatehowtheconnectionofsubsystemsaffectsthegridde-pendencyandtherenewablecontributiontothepowerdemands.

Thisstudyhasbeencarriedoutbysimulation,usingpowerdemandstypicalforlargehouseholdsandasmallfactoryorstore.

Theorganizationoftherestofthepaperisasfollows.Section2describesallpartsofmodelaswellasthemodelstructure.InSection3,theresultsfromthesimulationsarepresentedanddis-cussed.Conclusionsendthepaper.2.Developmentofthemodel

Modelingisausefultoolinordertostudycomplexsystemsindifferent elds[12].Inthispaper,allmodelsdevelopedtorepresentthescenarioinFig.1wereimplementedinMatlab/Simulinkwithasamplingperiodof10min.Simulationsweremadeonlyconsid-eringelectricalpower ow;hencenoconsiderationsweremaderegardingaspectssuchaspowerquality.

Onedayofeverymonthwasusedandtakenascharacteristicfortherespectivemonthregardingwindfeaturesandsolarradiationdurationandintensity.Stochasticelementswereincludedinthesolarandwindmodelstoimitatetheunpredictablenatureoftheserenewableresources.Themodeledresourcesmaybetunedtomatchsitespeci cconditionsusingwindspeedandsolarradiationmeasurements.Hence,themodelcanbeusedtoevaluatecertainenergysupplysystemsusingresourceuncertaintiesthatmimicparticularlocalconditions.

Thefollowingsectionsbrie ydescribethemodelsandmethodsusedtosimulatethescenarioshowninFig.1.DetaileddescriptionsandexplanationsofthemodelscanbefoundinRef.[13].2.1.Thesolarresource

Solarradiationonearthcanbemodeledasthesumofadirectanddiffusecontribution,anditisreferredtoasglobalortotalradiation.Thedirectradiation(orbeamradiation)representsthesolarradiationreceivedbyaplanewithouthavingbeenscatteredbytheatmosphere.Thediffuseradiationrepresentsthescatteredsolarradiationandisheremodeledasbeingisotropic,henceuni-formlydistributedfromallovertheskydome.Thisisaverysimplewayofmodelingdiffuseradiationandmoreaccuratemethodsexist.Thissimpli cation,however,hasaminorin uenceontheendresultsofthestudy.

Thesolarresourcemodelwasbasedonaclearskyradiationmodel,givenbyRef.[14].Theclearskymodelestimatesthebeamanddiffuseradiationtransmittedthroughclearatmosphericcon-ditionsonahorizontalplane,givenazenithangle,analtitudeandcorrectionfactorswhichaccountfortheclimatetype.Fourdifferentclimatetypesareused;tropical,mid-latitudesummer,sub-arcticsummerandmid-latitudewinter.Radiationonatitledplaneiscalculatedatanyinstantbyusingthelatitude,theorientationofthecollectorplane,thetimeofyearandthehouroftheday.

Themajorityofthemodel’sparametersareeasilydeterminedsincetheyaregivenbythegeometricalrelationshipbetweenthesunandtheearth.Ontheotherhand,theeffectoftheatmosphericconditionsontheincomingsolarradiationmustbecarefullystudiedbeforerealizingsitespeci cmodeling.Thein uenceoftheatmosphericconditionsis,intheclearskymodel,determinedbythealtitudeandtheclimatetype.

Partiallyrandomattenuationoftheradiationwasaddedtotheclearskymodelwiththeaimtoapproachtheunpredictabilityofrealworldconditions.Thelevelofalloweddistortionwasde nedseparatelyforeverymonth,duetochangesintheatmosphericconditionsthroughouttheyear.Thedistortionoftheclearskyra-diationmaybetunedtomatchspeci clocations.However,inthisstudythemainpurposeofthedistortionistointroducesimplebutrealisticuncontrollablereductionsofPVpoweroutput.Thisisimportantsincethesimulationsamplingperiodwas10min.

Thedistortionwasassumedtoreducethedirectradiationmomentarily,andforanumberoflargerperiodsthroughouttheday.Thesmaller uctuationsofthedirectradiationareprovoked rstlybyselectinganallowedreductionlevel,andthentheactualreductionisgeneratedusingarandomselectionfunction.Thelargerperiodswithoutdirectradiation(duringdaytime)http://wendang.chazidian.comtitudehasbeensetto40 N,heightto600mandtheclimatetypecorrespondstomid-latitudesummer.Thecollectorplaneisfacingsouthanditis

天津科技大学电气工程及其自动化毕业设计的外文翻译样板

M.Mikatietal./RenewableEnergy57(2013)587e593

589

April

)

2m/W( TGSolar timeMay

)

2

m/W( TGSolar time

Fig.2.ExamplesofmodeleddailyirradianceoncharacteristicdaysforAprilandMay.

tiltedtoanangleequaltothelatitude.AscanbeseeninFig.2,therandomdistortionismodeledtoaffectonlythedirectradiation.Thediffuseradiationisassumednottobeaffectedbypassingcloudssinceitismodeledasisotropic.Again,thisisasimpli cationsincethecloudswillrendertheincomingdiffuseradiationfromtheskydomesomewhatheterogenous.2.2.Thewindresource

Thewindresourceonaparticularsitedependsonamultitudeofaspects,suchas:theearth’sthermalmass,topographyandlatitude[15e17].Anincreasedelevationabovegroundgenerallyleadstosmootherwindssincesurfaceroughnessprovokesunstablebehavior.Inthisstudythefollowingsimplemethodwasusedtogeneratedailywindspeedpro les.Thewindwasconsideredtoconsistofanaveragewindspeedpro le,withtwotypesofrandomdistortionsuperposed:

Turbulence,thatdisturbsalldatapointsofthedaywiththesameprobability.

Stronger uctuationsthataffectonlypartsoftheday.

Theaveragewindspeedpro lewascreatedbydividingthedayintosixequallylonglinearsegmentsof4h.Turbulencewasthengeneratedbylettingeverydatapoint uctuateaccordingtoanormaldistributionfunction.Thestronger uctuationswereaddedanddesignedtolastfor20mineachtime.Finally,acertainamountofrandomizationwasincludedintothemeanwindpro leinordertogeneratenewwinddayseverytimethemodelwasrun.

Itispossibletoadjustthedistortionoftheaveragewindpro letoincreaseorsmooth-outuncertainties.TwocomparisonsareshowninFig.3betweenrecordedwindmeasurementsandwindpro lesgeneratedwiththemodel.Windmeasurementswererecordedin(above)Celestún,Méxicoat20m,the29thofOctober2006andin(below)Ambewela,SriLankaat20m,the20thofMay2001.

2.3.Thephotovoltaicarray

PlentyofliteratureexistsonthemodelingofPVsolarcellsandmodules[18e22].ThisstudystartsfromtheequivalentcircuitshowninFig.4andmainlyfollowsthemethoddescribedbyRef.[19].TheresultofthedevelopmentproposedinRef.[19]isaSimulinkmodelofaPVmodulewherethenumberofcellsinpar-allelandseriescanbede ned,aswellasparameterscommonlyknownsuchastheopencircuitvoltage,theshortcircuitcurrentandtheidealityfactor.Inthisway,commercialPVmodulescanbemodeledusingthemanufacturer’sdatasheet.

)

s/

m (eedps

nd WiTime

)

s/m(d eepsd niWTime

http://wendang.chazidian.comparisonsbetweenmodeleddailywindspeedpro lesandwindmeasure-mentsrecordedwitha10-minsamplingperiod.Themodeledwindpro leshavedeliberatelybeenelevatedtofacilitatethecomparison.

Inadditiontothedevelopmentof[19],thefollowingexpressionwasusedtoaccountforthevariationsintheopencircuitvoltagewithtemperature[20]:

V OC¼VOC;0LITCÀT

ref

(1)

whereVoc,0istheopencircuitvoltageatthereferencetemperatureTref,LIisatemperaturecoef cient,andTCisthecelltemperature.

InFig.5,acomparisonismadebetweenPVmoduleperfor-mancecurvesgeneratedwithmodelandthespeci cationsfromthemanufacturerofaPVmodule.Theclosenessbetweenthemodel’sandthemanufacture’scurrentevoltageandpowerevoltagerelationshipsgivesreliabilitytothemethodused,atleastata xedcelltemperature.2.4.Thesmall-scalewindturbine

Greatdifferencesarefoundinwindturbinetechnologydependingonsizeandcapacity.Windturbineswithanominalcapacitylessthan10kWareusuallysimplerthanlargeturbinesandexcludefeaturessuchasgearboxandbladeregulation.Theyareoftenequippedwithapermanentmagnetsynchronousgenerator[1,17,24]andneednoexternalpowersupply.Mostcommonly,theangularvelocityatveryhighwindspeedsislimitedusingafurlingmechanism[25,26],whichturnstherotordiscawayfromthewind

Fig.4.Simpli edequivalentcircuitofasolarcell.

天津科技大学电气工程及其自动化毕业设计的外文翻译样板

590M.Mikatietal./RenewableEnergy57(2013)587e593

http://wendang.chazidian.comparisonbetweenIeVandPeVcurvesofthemodeledphotovoltaicpanel(left)andthespeci cationsgivenbythemanufacturer(right)ofthephotovoltaicpanelNE-165U1[23].

directionandstronglyreducesthepowerextraction.Consequently,thepowercurveofafurlingequippedwindturbinehasadrasticreductioninoutputcapacityafteracertainwindspeed,incontrasttolargerwindturbinesthatcommonlymaintainthepoweroutput.

Inthisstudy,generalwindturbineequationsfrome.gRefs.[15,25,27e30]wereusedtosimulatethepowerextractionfromthewind,theaerodynamicsoftherotor,andtheelectricaloutput.Adirectcurrentmachinewasusedtorepresentthepermanentmagnetsynchronousgeneratorwitharecti edoutput,amethodalsousedbyRef.[27].Additionally,simplemodi cationsweredevelopedtomodelthefurlingeffectonthewindturbinepoweroutput.ThepowercurveofthewindturbinemodelispresentedinFig.6.

2.5.Modelstructure

ThewindandsolarresourcemodelswereconnectedwiththemodelsofthePVmodules,thewindturbineandthepowerde-mandsaccordingtoFig.1.Threepowerdemandswereusedintheanalysis;theyareshowninFig.7andrepresentthepowerdemandoftwohouseholds(house1and2)andonesmallfactoryorstore.ItcanbeseeninFig.7thatthetwohouseholdshavesimilarelectricity

Power (kW)

consumptionpatterns,characterizedbypowerdemandpeaksduringtheearlymorninghoursandintheevening.Incontrast,theelectricityconsumptionofthesmallfactoryismorecontinuousandinphasewiththesolarresource.TakingthedayinFig.7asrepre-sentativefortheentireyear,annualenergydemandsareabout9.0,7.5and15.8MWhforhousehold1,household2andthesmallfactory,respectively.

Someadditionalelementswereincludedinthemodeltoac-countforlossesinthepowersystem.Theireffectonthe owofpowerwascalculatedusingtheiref ciencies.Thefollowingele-mentswereincluded:

PowerinvertertoconvertPVoutputfromDCtoAC,theef -ciencyoftheinverterisshowninFig.8.

BoostconvertertoadjustthePVoutputvoltage,h¼80%[31]. Buck-Boostconvertertoadjustwindpoweroutputvoltage,h¼90%[32].

Filtertosmooththeinverteroutput,h¼97%.

Transformationbetweentheinverterandtheelectricgrid,h¼95%[33]

.

Power (W)

Wind speed

Hour

Fig.6.Powercurveofthemodeledsmall-scalewindturbine.

Fig.7.Thethreepowerdemandsusedduringsimulations.

天津科技大学电气工程及其自动化毕业设计的外文翻译样板

M.Mikatietal./RenewableEnergy57(2013)587e593591

)

(% ynceiciffEPercent of full load

Fig.8.Ef ciencyofthepowerinverterusedduringsimulations.AdaptedfromRef.[21]

.

Itwasassumedthatthetransformationlossesbetweentheinverterandthepowerconsumerswerenegligible.2.6.Conditionsoftotalsystemsimulations

20PVmoduleswereconnectedinparallelineachsubsystemtogetherwitha7kWwindturbine.EachPVmodulehadanominalcapacityof165W,andthecut-inandcut-outspeedsofthewindturbinewere3and25m/srespectively.ThePVcellef ciencywasconsideredasconstant.Bothsubsystemswereassumedtobecloselylocated,hencetheirwindandsolarresourcesweresaidtobethesame.

Thegridpowertransfersofeachsubsystemintheindependentcon gurationwerecalculatedbysubtracting,ateveryinstant,thepowerdemandfromthesumofconvertedPVandwindpoweroutput.Eachsubsystemwilldeliverpowertothegridifthatequationturnsoutpositiveanditwillbeextractingpowerfromthegridifitisnegative.Sincethecoupledsystemcanexchangepowerbetweensubsystems,thepowertransferswiththegridinthiscon gurationwerecalculatedasthesumofthepowertransfersofsubsystem1and2ateveryinstant.

ThewindpoweroutputwasgivenahighercontributionprioritythanthePVarraysifthecollectivewindpowerandPVoutputexceededthepowerdemandsofthesubsystems.Thiswasdeter-minedduetothecostperinstallednominalcapacity,whichgenerallyislowerforwindturbinesthanforPVarrays[2,21].

Thesolarresourcewassettorepresentsasitelocatedat40 NandatmosphericconditionsapproachingthoseofMadrid,Spain.Thewindresourcewasestablishedascontinuousbutmoderate.Yet,itmustbeemphasizedthatnofurthereffortwasmadetomatchthesimulationconditionstothisparticularlocation.Thepowersystemmodelwasstudiedusing ctionalrenewableresources.3.Results

3.1.Resultsfromaone-daysimulationexample

InFig.9,aone-daysimulationoftheentiresystemispresentedforthemonthofMarch.ThebottompartofFig.9showsthewindandsolarresourcesonthatparticularday.ThediagramsaboveshowtheACwindandPVpowergenerationtogetherwiththepowerdemandsofthetwosubsystems.Thepowerdemandsofsubsystem1and2correspondtohousehold1andthesmallfactory.ThetoppartofFig.9showsthepowertransfersbetweenthegridandsubsystems1and2respectively.Italsoshowsthepowertransfersfortheentirecoupledsystem.

Asexpected,thewindresourceisfairlystablethroughouttheday,alteredbycontinuoussmall uctuationsandafewlargerones.

Fig.9.Exampleofaone-daysimulationforthemonthofMarch.ThepowerdemandofS1andS2correspondstohousehold1andthesmallfactoryrespectively.

Theincomingirradiancereachesabout900W/m2aroundmiddaybutiscontinuouslydistortedbytheatmosphere.Thedirectradia-tionisabruptlyeliminatedduringperiodsbypassingclouds.Thisleadstoamoderatebutuninterruptedwindpowergenerationandastrong,butdiscontinuousPVpowergeneration.Itcanbenotedthat,forbothsubsystems,thereareperiodswhenthedemandexceedsthesupplyandviceversa.Thisisalsore ectedinthepo-wertransferplotatthetopofFig.9.Inbothcon gurationsthereareperiodsofpositivepowertransferwhenelectricityisdeliveredtothegridandnegativepowertransferwhenelectricityistakenfromthegrid.Theidealcaseimpliesceropowerexchangewiththegrid.Thiswouldmeanthatnopower owstoorfromthegridandrenewablepowersupplyequalsdemand.

ItcanbeseeninthepowertransferplotatthetopofFig.9thatthecoupledsystemsometimestransfersmorepowerthanthetwosubsystemsindividually.Thishappens,forexample,around19:00h.However,thepowertransfersofthecoupledsystemcanneverexceedthecollectivepowertransferofbothsubsystems,becauseitisthesumofthetwo.Onotheroccasionsthepowertransfersofthecoupledsystemapproacheszero.Thishappens,forexample,at09:00handbefore16:00h,anditoccurswhenthesignsofthesubsystem’spowertransferareopposite.Thatis,ac-cordingtothemodel,thecouplingofsubsystemswillreducepowerexchangewiththegridif,atanyinstant,onesubsystemisgener-atingexcesspowerwhiletheotherismomentarilyinapowerde cit.Ontheotherhand,theconnectionofsubsystemsissuper- uousduringperiodswhenbothsubsystemssimultaneouslyhaveanexcessorshortageintheirpowergeneration.3.2.Longtermeffectsofcouplingthesubsystems

Thesubsequentresultswereobtainedfromone-yearsimulationrunsusing rstly:thepowerdemandsofhousehold1andthesmallfactory,andsecondly,thepowerdemandsofhousehold1and2.Thewindandsolarresourcesweresimilarinbothsimulationrunswhenmeasuredinaverageenergyperdayandtotalannualresource.Averagedailysolarirradiancewasbetween2.5and4.5kWh/m2andtheaveragedailywindspeedwasbetween4.4and6.8m/s.Theannualef ciencyofthePVarraywas10.8%andthecapacityfactorofthewindturbinewas20.0%.

Thedailysolarradiationgeneratedwiththemodelisnotablylowerthantheannualaverageofdailysolarradiationmeasuredby