Page - 82 - in Short-Term Load Forecasting by Artificial Intelligent Technologies

energies Article ImpactsofLoadProfilesontheOptimizationof PowerManagementofaGreenBuildingEmploying FuelCells Fu-ChengWang* andKuang-MingLin DepartmentofMechanicalEngineering,NationalTaiwanUniversity,Taipei10617,Taiwan; r07522809@ntu.edu.tw * Correspondence: fcw@ntu.edu.tw;Tel.:+886-2-3366-2680 Received: 25October2018;Accepted: 24December2018;Published: 25December2018 Abstract:Thispaperdiscusses theperformance improvementofagreenbuildingbyoptimization proceduresandthe influencesof loadcharacteristicsonoptimization. Thegreenbuildingisequipped withaself-sustainedhybridpowersystemconsistingofsolarcells,windturbines,batteries,proton exchangemembrane fuel cell (PEMFC), electrolyzer, andpowerelectronicdevices.Wedevelopa simulationmodelusingtheMatlab/SimPowerSystemTMandtunethemodelparametersbasedon experimental responses, so thatwecanpredictandanalyzesystemresponseswithoutconducting extensiveexperiments. Threeperformance indexesare thendefinedtooptimize thedesignof the hybrid system for three typical loadprofiles: thehousehold, the laboratory, and theoffice loads. Theresults indicatethatthetotalsystemcostwasreducedby38.9%,40%and28.6%forthehousehold, laboratoryandoffice loads, respectively,while thesystemreliabilitywas improvedby4.89%,24.42% and5.08%.That is, thecomponentsizesandpowermanagementstrategiescouldgreatly improve systemcost andreliability,while theperformance improvement canbegreatly influencedby the characteristicsoftheloadprofiles.Asafetyindexisappliedtoevaluatethesustainabilityofthehybrid powersystemunderextremeweatherconditions.Wefurtherdiscuss twomethods for improvingthe systemsafety: theuseofsub-optimalsettingsor theadditionalchemicalhydride.Adding20kgof NaBH4 canprovide63kWhandincreasesystemsafetyby3.33,2.10,and2.90days for thehousehold, laboratoryandoffice loads, respectively. In future, theproposedmethodcanbeappliedtoexplore thepotentialbenefitswhenconstructingcustomizedhybridpowersystems. Keywords:hybridpowersystem; fuelcell; solar;wind; fuel cell;optimization; cost; reliability 1. Introduction Today’s energy crises andpollutionproblemshave increased the current interest in fuel cell research. Oneof themostpopular fuel cells is theproton exchangemembrane fuel cell (PEMFC), whichcan transformchemical energy intoelectrical energywithhighenergyconversionefficiency byelectrochemical reactions. At theanode, thehydrogenmolecule ionizes, releasingelectronsand H+protons.At thecathode,oxygenreactswithelectronsandH+protons throughthemembraneto formwater. Theelectronspass throughanelectrical circuit to create currentoutputof thePEMFC. ThePEMFChasseveraladvantageousproperties, includinga lowoperatingtemperatureandhigh efficiency.However, italsohasverycomplexelectrochemicalreactions,soattemptstodevelopdynamic models for PEMFCsystemshavebecomeanactive research focus. For example, Ceraoloetal. [1] developedaPEMFCmodel thatcontainedtheNernstequation, thecathodickineticsequation,andthe cathodic gas diffusion equation. Similarly, Gorgun [2] presented a dynamic PEMFCmodel that includedwaterphenomena,electro-osmoticdraganddiffusion,andavoltageancillary. Thesemodels haveservedas thebasisofmanyadvancedcontrol techniquesaimedat improvingtheperformance Energies2019,12, 57;doi:10.3390/en12010057 www.mdpi.com/journal/energies82