Page - 84 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

Energies2019,12, 57 generation. Themodelparameterswere tunedbasedonexperimentaldata toallowtheprediction ofsystemresponsesunderdifferentoperationconditions.Historical irradiationandwinddatawere applied to estimate thepower suppliedby thePVandWT,while three typical loadprofileswere consideredtounderstandtheir impactsonsystemoptimization. Section3defines threeperformance indexes forevaluatinghybridpowersystemsequippedwithdifferentcomponentsandmanagement strategies.Weappliedthree typical loads tooptimizesystemdesignbytuningthecomponentsizes andpowermanagement. Theresults showedthat theoptimizationprocessescaneffectivelyreduce theenergycostsby38.9%,40.0%,and28.6%andgreatly improvesystemreliabilityby4.89%,26.42%, and5.08% for household, laboratory, and office loads, respectively. The guaranteed sustainable operationperiodsunderextremeweatherconditionswerealsoestimated. Theresults revealedthat systemsustainabilitycanbeimprovedbytheuseofasub-optimaldesignorchemicalhydrides.Wealso discuss thecriticalpricesof implementingachemicalhydrogengenerationsystem.Conclusionsare thendrawninSection4. 2. SystemDescriptionandModelling Thegreenbuilding, as showninFigure1 [34], is located inMiao-LiCounty inTaiwan. Itwas constructedbyChinaEngineeringConsultants Inc. (CECI)andwasequippedwithahybridpower systemthatconsistedof10kWPVarrays,6kWWTs,800Ahlead-acidbatteries, a3kWPEMFC,anda 2.5kWelectrolyzerwithahydrogenproductionrateof500L/h.Thebuildingwasautonomousand didnotconnect to themaingrid, i.e., itselectricitywassuppliedcompletelybygreenenergy, suchas solarandwind. Theenergycanbestoredforusewhenthegreenenergy is less thanthe loaddemands. Thesecomponentswereoriginallyselectedtoprovideadailyenergysupplyofabout20kWhbased on theNationalAeronautics andSpaceAdministration (NASA)data [34], as illustrated inTable 1. Solarenergywasabundant in thesummerbutpoor in thewinter, sowindenergywasexpected to compensate forsolarenergy inthewinter.However,ChenandWang[32]appliedtheVantagePro2 Plus Stations [35] tomeasure the realweather data on the building site and found that thewind energywasnotsufficient tocompensate for thereducedsolarenergy in thewinter. Furtheranalysesof theenergycostsalsorevealedthat thewindenergywasnoteconomicallyefficient for thisbuilding, as illustrated inTable2. Therefore, the followingcomponentselectionprinciplesweresuggestedto improvesystemperformance [32]: (1) Energy sources: the use of PV and PEMFC in the green buildingwas suggested, because solarenergywas themosteconomical energysourceandthePEMFCcouldguaranteeenergy sustainability. ThePEMFCcanbe regardedas anenergy source thatprovides steadyenergy andasanenergystoragesystemwhencoupledwithahydrogenelectrolyzer.Consideringthe transportation, storage,andefficiencyofenergyconversion, thePEMFCwithchemicalhydrogen generationbyNaBH4 [36]wassuggestedfor thesystem. (2) Energystorage: thelead-acidbatterywassuggestedbecauseof itsgreaterthan90%efficiency[37]. ThoughthePEMFCwithahydrogenelectrolyzercanalsostoreenergy, theconversionefficiency fromelectricity into hydrogenwas only about 60% [33]. Therefore, the total energy storage efficiencywas about 36%, because the PEMFC converted hydrogen into electricitywith an efficiencyofabout60%[38].Note that theLiFebatteryhasahigherefficiency(more than95%) but ismuchmore expensive than a lead-acid battery. Therefore, the lead-acid batterywas preferredfor thegreenbuilding. That is, theselectionofmultipleenergysourcesandstoragesdependedonthe localconditions andloadrequirements. 84