WaterPractice™•Vol.1•No.5©2007WaterEnvironmentFederation•doi:10.2175/193317707X256991TheImportanceofAerobicMixing,BiofilmThicknessControlandModelingontheSuccessorFailureofIFASSystemsforBiologicalNutrientRemovalDipankarSen1*,CliffordW.Randall2,RhodesR.Copithorn3,MarkkuHuhtamäki4,GregFarren5,WayneFlournoy61SantaClaraValleyWaterDistrict,5750AlmadenExpressway,SanJose,CA95118,USA.2DepartmentofCivilEngineering,VirginiaTech,Blacksburg,VA,USA.3Stearns&Wheler,Cazenovia,NY,USA.4JuuroconOy,Naantali,Finland.5AnneArundelCountyDPW,MD,USA.6EntexInc,NC,USA.*Towhomcorrespondenceshouldbeaddressed.Email:DSen@valleywater.org.ABSTRACTResearchonfullscaleIntegratedFixedFilmActivatedSludge(IFAS)andMobileBedBiofilmReactor(MBBR)systemsshowedthatproblemsdevelopedwithexcessivebiofilmthicknesswhenthemixingandrollpatternsdecreased.Evaluatedmediaincludedfixedbedcordmedia,movingbedspongemediaandmovingbedplasticmedia.Threestrategiesarerecommended.(1)ProcessmodelingshouldbeusedtodeterminethesolublebiodegradableCOD(SCODbio)andNprofilesalongtheaerobiccellsandadjustthevolume,quantityofmediaorMLSSMCRTifSCODbioconcentrationinanycellisexcessive.(2)WhereprocesschangescannotlowertheSCODbioandcontrolthickness,themixingandaerationsystemsshouldbeimprovedbyswitchingtoamorevigorousrollpatternwithcoarsebubblediffusersandjetmixing.(3)Thetypeofmediaselectedshouldbesuchthatthebiofilmspecificsurfaceareadoesnotdecreasesignificantlyasthethicknessincreasesfrom0.1to1mm.KEYWORDS:IFAS,MBBR,biofilm,thickness,mixing,modeling,Aquifas,dynamicsimulation.INTRODUCTIONTheoperatingdatafromfullscaleplantstreatingmunicipalwastewaterandoperatedasIntegratedFixedFilmActivatedSludge(IFAS)andMobile(orMoving)BedBiofilmReactors(MBBRs)systemsshowedthatproblemsdevelopedduetoexcessivebiofilmthicknesswhentheintensityofmixingandaerationrollpatternfellbelowcertainthresholds.ItwasalsoobservedthatexcessivebiofilmthicknessoccurredwhenthesolublebiodegradableCOD(SCODbio)inthemixedliquoroutsidethebiofilmincreasedwhileoperatingtheplantwithafixedlevelofmixingandrollthatwasinadequate.Inseveralinstances,thechangesinSCODbioandthicknesscouldbeattributedtoanincreaseinloading,decreaseintheMCRTofmixedliquor(MLSSMCRT),reductionintheefficiencyoftheaerationsystemandlossofsomediffusers.Theplantsstudiedincorporatedbiofilmsupportmediawithintheaerobicand,insomeinstances,theanoxiczonesoftheactivatedsludgesystem(Figure1).Thetypesofmediaappliedincludefixedbedcordmedia(RinglaceandBioweb),movingbedspongemedia(LinporandCaptor);andmovingbedplasticmedia(Kaldnes,Hydroxyl,Bioportzandother).Whatisunfortunateisthatwiththeexceptionofplantswithspongemedia(LinporandCaptor),theincreaseinthicknessalsoledtoareductioninbiofilmspecificsurfacearea(SSAb-m2ofbiofilmsurfaceperm3oftankvolume),thusreducingtheorganicremovalsandthenitrificationanddenitrification,whichfurtherexacerbatestheproblems.Theseobservationspointtotheneedforbetterprocessmodelsforthedesignandinstallationofthemedia.ThemodelingiscomplicatedbecausethethicknessofbiofilmthatSenetal.WaterPractice™•Vol.1•No.5©2007WaterEnvironmentFederation•doi:10.2175/193317707X2569912developsatacertainmixingintensityandthespecificsurfaceareaofbiofilmonacarrierparticlearedependentonthenatureandstructureofthemediaandontheCOD,NH4N,NOxNandDOconcentrationsineachcellofamulti-cellreactor.IB.IFAS–fixedbedEg:Ringlace,Bioweb(cord)ReactorInfluentRAS1A.ActivatedSludge1AScreenIC.IFAS–MovingBedEg:Linpor,Captor(sponge)Kaldnes,Hydroxyl,Bioportz(plastic)ScreenID.MBBRSamemediaas1CClarifierisnotusedinsometertiaryandsidestreamapplications1C1DEffluentIB.IFAS–fixedbedEg:Ringlace,Bioweb(cord)ReactorInfluentRAS1A.ActivatedSludge1AScreenIC.IFAS–MovingBedEg:Linpor,Captor(sponge)Kaldnes,Hydroxyl,Bioportz(plastic)ScreenID.MBBRSamemediaas1CClarifierisnotusedinsometertiaryandsidestreamapplications1C1DEffluentFigure1:PlantConfigurations.Aswithactivatedsludgesystems,theprocessmodelhastocomputetheaerationrequirementsforeachaerobiccellfortheDOset-pointspecified.However,inIFASandMBBRsystems,theaerationrequirementscomputedmustalsobeevaluatedagainstthethresholdsformixing(tokeepthemediasuspended)andshear(tocontrolthebiofilmthickness)andthephysicalcharacteristicsofthemedia.Dynamicsimulationisimportantinordertodetermineiftheaerationsystemdesignmighthavetobemodifiedtoinsurethatsufficientairisprovidedformixingthemediaduringthelowloadhours.TheexcessDOsuppliedduringthelowloadhoursmustbeaccountedforinthedesignofthepostanoxiccells.Inselectinganaerationsystem,thedesignerneedstoevaluatewhetherthesystemcanoperateeffectivelyandprovidethemixingwhensomediffusersbreakandtheairdistributionisdisturbedinthevicinityofthebrokendiffuser.Also,thesizeandnumberofaerobiccellsshouldbeevaluatedtodetermineiftheSCODbioconcentrationissoelevatedinthefirstaerobiccellthatproblemswiththenatureandthicknessofthebiofilmarecreated.Themodelshouldhelpthedesignerresizeandoptimizethelayoutofthefacility,educatethedesigneraboutthephysicalparametersthatdifferentiatevarioustypesofmedia,andallowthedesignertoincorporatethesedifferencesintotheprocessdesign.Dataregardingthe