Performanceofaglucosefedperiodicanaerobicba

Performanceofaglucosefedperiodicanaerobicbaffledreactorunderincreasingorganicloadingconditions:2.ModelpredictionK.Stamatelatoua,L.Lokshinab,V.Vavilinb,G.Lyberatosa,*aDepartmentofChemicalEngineering,UniversityofPatras,1Karatheodorist.,26500Patras,GreecebWaterProblemsInstituteofRussianAcademyofSciences,Gubkinastr.bd.3,119991Moscow,RussiaReceived6June2002;receivedinrevisedform7October2002;accepted23October2002AbstractAmodelwasdevelopedfortheanaerobicdigestionofaglucose-basedmediuminaninnovativehigh-ratereactor,theperiodicanaerobicbaffledreactor(PABR).ThemodelconsiderseachPABRcompartmentastwovariablevolumeinteractingsections,ofconstanttotalvolume,onewithhighsolidsandonewithlowsolidsconcentration,withthegasandliquidflowsinfluencingthematerialflowsbetweenthetwosections.Forthesimulationofglucosedegradation,thebiomasswasdividedintoacidogenic,acetogenicandmethanogenicgroupsofmicroorganisms.Thekineticpartofthemodelaccountedforpossibleinhibitionofaci-dogenesis,acetogenesisandmethanogenesisbyvolatilefattyacids.Themodelsucceededinpredictingthereactorperformanceuponstepincreasesintheorganicloadingrate.
2002ElsevierScienceLtd.Allrightsreserved.Keywords:Periodicanaerobicbaffledreactor;Dynamicresponse;Glucose;Model;Inhibition1.IntroductionTheperiodicanaerobicbaffledreactor(PABR)isaninnovativereactor,initiallydevelopedbySkiadasandLyberatos(1998).ThePABRinteriorspaceisseparatedintocompartments,eachconsistingofonedownflowandoneupflowpart(Stamatelatouetal.,inpress).Thecompartmentsarearrangedinacircularmannerintheannularregionbetweentwoconcentriccylinders.InanABR,thewastewaterentersthedigesterthroughaninfluentportandpassesbelowandaboveaseriesofbafflescomingintocontactwiththeaccumulatedbio-mass.Thedegradedwastewatercomesoutofthereactorthroughaneffluentport.InthePABR,theseinfluentandeffluentportsarenotfixed,buttheycanbeswitchedperiodically,sothatthefeedentersthereactorthroughallitscompartmentssuccessivelywithinapredefinedperiod.Thus,thePABRexhibitsaperiodicdynamicbehaviour,whichseemstobebeneficialinthecaseofsevereoverloading(SkiadasandLyberatos,1998;Skiadasetal.,2000).TherehavebeensomeattemptstomodelPABRinthepast,takingintoconsiderationthatbiomassaccu-mulatesinthereactor.Inordertosimulatethisfact,Skiadasetal.(2000)allowedthesolidretentiontimetobelongerthanthehydraulicretentiontimebutconstant.Thiswasdonebyintroducingaconstantempiricalbiomassretentionfactor.Althoughthisapproachpro-videsaparticularlyusefulpracticalmeansformodelinghigh-ratedigestionsystems,thephysicalfactors,suchasthesolidssettlingvelocityorupwardfluidvelocity,whichdeterminebiomassaccumulation,arenotexplic-itlyaccountedfor.Thisfact,then,limitsthepredictiveabilityofthemodel,astheappropriatevalueoftheretentionfactormaybesignificantlydifferentunderdifferentoperatingconditions.AnimprovementonthePABRmodelingframework,whichrelaxedtheas-sumptionofaconstantbiomassretentionfactor,wasmadebyStamatelatouandLyberatos(2002),whode-velopedamodelfortheanaerobicdigestionofgelatininthePABR.ThemodelfollowedtheapproachofBuijsetal.(1982),whoinmodelinganotherhigh-ratesystem,theupflowanaerobicsludgeblanketreactor(UASBR),consideredthemainpartofthereactorasconsistingoftwoseparateregions.ThemodelofStamatelatouandLyberatos(2002),usedkineticsforBioresourceTechnology88(2003)137–142*Correspondingauthor.Fax:[phone].E-mailaddress:[e-mail](G.Lyberatos).0960-8524/03/$-seefrontmatter
2002ElsevierScienceLtd.Allrightsreserved.PII:S0960-8524(02)00277-8gelatindegradationthatwereobtainedthroughbatchexperiments.Inthepresentstudy,thehMETHANEisimulationmodel(Vasilievetal.,1993;Vavilinetal.,1994),aver-satilekineticmodeldescribinganaerobicdigestionki-neticsinhomogeneoussystemswasproperlyadaptedfordescribingthekineticsofglucosedigestioninaPABRreactor.ThedynamicresponseofthePABRduringsuccessiveincreasesintheorganicloadingratewassimulatedsatisfactorilybythisPABR-adoptedversionofthehMETHANEimodel,supportingourunder-standingofthestateofthereactorundertheseupsets.2.MethodsThePABRconfigurationandoperationmodeaswellastheseriesofexperimentsconductedinPABRaredescribedinapreviouspaper(Stamatelatouetal.,inpress).Thefeedsyntheticmediumusedduringtheex-perimentscontainedglucose,caseinhydrolysate,yeastextract,NaHCO3,andtraceelements.Anexperimentconsistingofthreephaseswasconductedataconstanthydraulicretentiontime(4d)andswitchingperiod(2d).Ineachphase,thefeedconcentrationwasincreasedin-stantlytoanewvalue(Table1).Thedurationofeachphasewasdeterminedbythetimerequiredforasteadyperiodicstatetobereached.2.1.PABRmodelingInthesequelthemodelisdescribedintwoparts,thefirstpertainingtotheflow-frameworkofthePABR,andthesecondtothekineticmodelassumptions.2.1.1.ThegeneralPABRconfigurationmodelframeworkThemodelwasbasedonthepreviousworkofStamatelatouandLyberatos(2002).TheschemeofthePABRmodelframeworkisrepresentedinFig.1.Al-thougheachcompartmentconsistedofonedownflowandoneupflowsection,themodelconcernedonlytheupflowsection,assuminginstantaneoustransferinthedownflowsection.Thissimplificationisjustifiedbythenegligiblysmallvolumeofthedownflowsection.So,theterm‘‘