<<

OutlineOutline •• IntroductionIntroduction •• IssuesMembraneIssues •• OtherIssuesOtherIssues OutlineOutline •• IntroductionIntroduction •• MembraneIssuesMembraneIssues ––GeneralfoulingGeneral ––ImpactofSRTorF/MImpactofSRTorF/M ––ImpactofMLSSImpactofMLSS ––ImpactofwetweatherImpactofwetweather •• OtherIssuesOtherIssues WhatWhat isis membranemembrane fouling?fouling? •• MembranefoulingisthelossofpermeabilitywithMembranefoulingisthelossofpermeabilitywith timetime •• Inpractice,thisisobservedasanincreaseintheInpractice,thisisobservedasanincreaseinthe TMPrequiredtomaintainflowthroughtheMBRTMPrequiredtomaintainflowthroughtheMBR

•• Forengineers,theincreaseinTMPForengineers,theincreaseinTMP needstoberelatedtothefluxrateneedstoberelatedtothefluxrate andnormalizedfortemperatureandnormalizedfortemperature thisiscalledatemperaturethisiscalledatemperature correctedcorrected ““PermeabilityPermeability ”” oror ““SpecificFluxSpecificFlux ”” DefinitionDefinition ofof termsterms

Po

J=Q/A=membraneflux(m/s)

Membrane permeability = J/TMP

Pe TMP=P oPe (Pa) A Membrane fouling: TMP

time PermeabilityPermeability J L = P TMP

Typical units in USA: gal/(ft 2. d.lb/in 2) or gfd/psi

Europe and Asia: L/(m 2. h.bar) or LMH/bar

Strict SI Units: m2. s/kg TemperatureTemperature correctedcorrected permeabilitypermeability

(-0.0239( T -20)) o J ⋅ e L20 C = P TMP

The above equation corrects for temperature effects on the viscosity of water. This equation is accurate within 5% for a temperature range of 5 to 40 oC.

WHY DO WE DO THIS?? Because changes in the viscosity of water directly impact TMP TemperatureTemperature correctioncorrection

2.0

1.8 As water temperature decreases - viscosity of water increases 1.6 s •

1.4

1.2

Actual 1.0 Calculated

0.8

0.6 Absoluteviscosityofwater,mPa 0.4 All other conditions equal -

0.2 this increases the TMP Need to Need equation different a use 0.0 0 5 10 15 20 25 30 35 40 45 Temperature,ºC KeyMBR ResearchProjects ListList ofof KeyKey ResearchResearch ProjectsProjects

• 1999 2000WERFstudyinSanDiego • 1999 2000BureauofReclamationII StudyinSanDiego • 2000 2004STOWAprojectinthe Netherlands • 2002 2005WERFstudyinSan Francisco • 2001 presentKingCountyinSeattle • 2003 2004BureauofReclamationIII StudyinSanDiego • 2002 presentOn goingresearchby Anjou Recherche • 2006 presentEUfundedAmadeus initiative Understanding MBRFouling MBRMBR foulingfouling theorytheory

• Basicfundamentalsofmembranefoulingin MBRs arethesame regardlessofthemanufacturerorconfiguration(Pressureor Vacuum) • Membranefoulingresultsfromtheinteractionbetweenthemixed liquorandmembranematerial – Complexmixtureoforganics – Metabolicbyproductsandpossiblyinfluentsubstrateorpartiall y degradedinfluentsubstrate – Cellsandmicrobes – Cellularandmicrobialdebris – Inertsuspendedsolids – Dissolved inorganics (possibleprecipitants) ResistanceResistance inin --seriesseries modelmodel •• SimplisticmodelSimplisticmodel •• WidelyusedwithlowWidelyusedwithlow pressurepressuremembranes (MF/UF/MBR)(MF/UF/MBR) •• CanbeusedtoprovidepowerfulinsightstoCanbeusedtoprovidepowerfulinsightsto MBRfoulingMBRfouling

J = membrane flux, m/s TMP TMP = trans-membrane pressure, Pa J = . w = absolute viscosity of water, kg/m s -1 w ⋅ R T RT = total resistance to , m ResistanceResistance inin --seriesseries modelmodel

•• RRT=R=R M+R+R F+R+R C

•• RRT =Totalresistance=Totalresistance

•• RRM =Membrane=Membrane

•• RRC =CakeLayer=CakeLayer

•• RRF == FoulantsFoulants – OrganicAdsorption – InorganicPrecipitation – Poreblocking MembraneMembrane resistance,resistance, RR MM

• RM isthehydraulicresistance duetothemembranealone DeterminingR M

• RM canbedeterminedby 90000 performingacleanwaterflux 80000 profileonacleanmembrane 70000 60000 • RecordTMPandtemperature 50000 12 -1 40000 RM = 3x10 m for3differentfluxrates TMP(Pa) 30000

20000 • PlotTMPvs.m*J,slopeisR M y = 3E+12x + 21823 10000 R2 = 0.9996 0 5.E-09 6.E-09 7.E-09 8.E-09 9.E-09 1.E-08 Viscosity*Flux(kg/s 2) OtherOther resistanceresistance termsterms TMP • RT isobtainedduringnormalMBRoperation R T = – Increaseswithtimeortotalvolumefiltered w ⋅ J

– Influencedbyresistanceofthefiltrationcake,R C

– Influencedbythedegreeof foulant presentonthemembrane,R F

• RF canberoughlyestimatedatanypointinanoperation cycle – Drainthemixedliquorfromthemembranetank(airoff) – Fillthemembranetankwithmembranepermeateandperform fluxprofile thisprovidesR M+R F (possiblysomeresidualR C that ’swhythisisan estimate )

• SubtractR M (this was obtained before run began )andyou canapproximatetheamountof foulant ,R F • RemainderofR T isattributedtoR C HydrodynamicHydrodynamic forceforce balancebalance

• Membranefluxcontrolstherateofmaterialtransported

tothemembranesurface,J SS • Theliftforcecontrolstherateatwhichrejectedmaterial

isre suspendedtothebulk,V L

•• NormalMBRoperationNormalMBRoperation

–– JJss ≤≤ VVL –– i.e.Operatingatsubi.e.Operatingatsub criticalfluxcriticalflux Critical Flux CriticalCritical fluxflux

• Conventionallydenotesfluxbelowwhichfoulingdoes nottakeplace – Membranepermeabilityremainsasitwasinpurewater • StrictcriticalfluxdefinitiondoesnotapplytoMBR • Fieldetal.,1995firstadaptedthisconcepttolow pressuremembranes • Le Clech etal.,2003 furtherdevelopedthecriticalflux conceptfor MBRs IllustrationIllustration ofof criticalcritical fluxflux

20 gfd 8.7 MLSS = 8 g/L 18 gfd 10.9 SCFM = 30 scfm gfd 13.1 16 gfd 15.3 gfd 17.5 14

12

10

8

VacuumPressure,inHg 6

4

2

0 0 2 4 6 8 10 Time,minutes FactorsFactors affectingaffecting criticalcritical fluxflux

•• SpecificMBRhydrodynamicsSpecificMBRhydrodynamics – Hollow fiber versusflatsheet – Coarseaerationdistribution – Pressurevs.VacuumMBRsystems •• MixedliquorpropertiesMixedliquorproperties – Degreeof » Moredisperse flocs withhighercolloidalmaterialisdifferentthana well flocculatedsludge – Viscosity

» ThemixedliquorviscosityimpactstheefficiencyofV L » Higherviscosity lowerscouringefficiency ImportanceImportance ofof coarsecoarse bubblebubble airair Cross-flow AdaptedfromBérubé etal.,2005 AWWA MTC velocity

Constantfluxexperiments Single Phase = Water alone Dual Phase = Air/Water

Conclusion: Maintaining clean, well-functioning, and well-distributed coarse bubble air is critical SludgeSludge PropertiesProperties

Filamentous Colloidal Microorganisms Material

ParticleSize

Extracellular Polymeric Substances(EPS) CriticalCritical FluxFlux IllustrationIllustration

MLSS=1012g/L Air=30scfm

AdaptedfromFan etal.,2006 Water Research V40 RM RC RF ––JssJss =tomembrane=tomembrane

––VVL =awayfrommembrane=awayfrommembrane

––JssJss ≥≥ VVL (rapidfouling)(rapidfouling) “Typical” MBR FoulingMechanisms

PhotosadaptedfromMiuraetal.,2007 ““TypicalTypical ”” MBRMBR foulingfouling mechanismsmechanisms

• Organicsarethemostcommon foulant undernormal operatingconditionsin MBRs – Conservativeflux – Wellfunctioning/distributedcoarseaeration – ControlledMLSS • Organicfoulingis primarily attributedtothesolubleor colloidalorganicspresentinthemixedliquor – Particles ≤ 6 m – Notincorporateintolargerfloc – Notyetclearwhethercolloidalorsolubleisculprit(likelybo th) » Researchhashighlightedtheimportanceofsolublecarbohydrate or polysaccharides,butthereisalsoliteraturetothecontrary • Increasedsoluble/colloidalorganiccontentresultsin increasedmembranefoulingrates ExtracellularExtracellular PolymericPolymeric SubstancesSubstances (EPS)(EPS) andand SolubleSoluble MicrobialMicrobial ProductsProducts (SMP)(SMP)

Hydrolysis EPS

Active Cell SMP Substrate

Adsorption and Diffusion/Shear flocculation OrganicOrganic foulingfouling

AdaptedfromLesjean etal.,2005 Water Science and Technology OrganicOrganic foulingfouling

70

Total SMP = 7.0x + 36.8 2 60 R = 0.77

50

40

30

SMPconcentration,mg/L 20 SMP = soluble microbial products 10 (soluble protein + soluble carbohydrate)

0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Steadystatefoulingrate,LMH/bar •d

AdaptedfromTrusselletal.,2006 InorganicInorganic foulantsfoulants

• LessseverethanorganicfoulingformostmunicipalMBR applications • Certainwaters(e.g.hardwaters)canslowlydevelopan inorganicfoulinglayer – LowpHclean(mostcommoniscitricacid)willcontrolthis – Thiscleancanbedoneasinfrequentlyasannuallyatmany facilities • Coagulantsaretypicallyusedinmunicipalwastewater treatmentfacilities – Highcoagulantdosescreatehydroxideprecipitants(e.g. Fe(OH) 3),orcoagulantcarryover(e.g.notboundupin mixedliquor)thatwillresultininorganicfouling – Itappearsthatanoccasionallowdoseofcoagulantcanhelp reducesolubleandcolloidalorganicfouling CoagulantCoagulant AdditionAddition

AdaptedfromHolbrooketal.,2004 Water Environment Research PolymerPolymer AdditionAddition

•• BenefitsofspecializedpolymeradditionorBenefitsofspecializedpolymeradditionor ““fluxflux enhancersenhancers ”” arecurrentlybeingresearchedarecurrentlybeingresearched – Reducesmixedliquororganiccontent(SMP) – Allowsforincreaseinmembranefluxbyreducingcolloidal organics •• BenefitshavenotbeendemonstratedonlongBenefitshavenotbeendemonstratedonlong termterm basisbasis – Short termincreaseinmixedliquorfilterabilityoccurs – Highdosesrequiredforlongerruntimes – Long termimpactsonsludgeproperties(e.g.post polymer addition)havenotbeendemonstrated PolymerPolymer AdditionAddition

AdaptedfromYoonetal.,2005 Water Science & Technology OtherImportant FoulingMechanisms ChangesChanges inin MLSSMLSS concentrationconcentration • IncreasesintheMLSSconcentrationareimportant

– IncreasestheJ SS tothemembranesurface – Increasesthemixedliquorviscosity – Combinationcanresultinoperationabovethecriticalflux withoutchangingthemembraneflux • Differentresearchershavereacheddifferentconclusions onthe “maximum ” MLSSconcentrationformembrane fouling • Thisisbecausethe “maximum ” MLSSdependson – Membranehydrodynamics(e.g.flatsheet,hollowfiber,pressure vs.vacuum,etc.) – Membranefluxrate – Re suspendingefficiency(e.g.airrate,noair? crossflow velocity, “jet ”,mixedliquorviscosity) ChangesChanges inin mixedmixed liquorliquor propertiesproperties

• Mixedliquorviscositycanchangedramaticallywithout theMLSSconcentrationchanging! – Mixedliquorviscosityhasbeen>2timesgreaterdependingon properties(e.g.200vs.400 mPa .s at18g/L) – Mixedliquorviscositydependsuponthedegreeofflocculation, extracellular polymericsubstance(EPS)concentration,and filamentconcentration • Mixedliquorfilterabilitycanchangewithoutchanging MLSSconcentration

– Ifde flocculationoccurs,adramaticincreaseintheR C willoccur » Increaseincolloidalcontent » Disperse flocs andsinglecells » Dramaticchangescanbequantifiedbytimetofilter(TTF) MixedMixed liquorliquor viscosityviscosity

AdaptedfromCuietal.,2003 OtherOther importantimportant mixedmixed liquorliquor propertiesproperties forfor MBRMBR foulingfouling • Key foulants arisefrombiomass,termed extracellular polymericsubstances(EPS) – unboundfractionoftenreferredtoassolublemicrobialproduct (SMP) – boundfraction(EPS) • Thesecanbefurtherfractionatedintochemicaltypes, namely: – polysaccharide(orcarbohydrate) – protein ChemicalChemical foulantfoulant studiesstudies

• Difficulttoubiquitouslyidentifykey foulant • Generally,highconcentrationsofSMPareasignificantconcern – Membranefoulingwillincrease – Newresearchisshowingimportanceofmolecularweightofsolubl eorganic (e.g.>10 kDa and<100 kDa ) • HighconcentrationsofEPSdonotalwaysresultinincreasedfou ling rates – HighEPScanbeasignofgoodflocculation(e.g.lowcolloidal andsoluble organiccontent)

– “Sticky ” EPScanresultatlowEPSconcentrationsandproducehighR C IsPoreSizeImportant? MFMF vsvs UFUF •• AmuchdebatedtopicAmuchdebatedtopic •• SomebelievethatMFhasahigherfoulingSomebelievethatMFhasahigherfouling tendenacytendenacy thanUFmembranesthanUFmembranes •• SomebelievetheMFandUFmembranesinSomebelievetheMFandUFmembranesin MBRsMBRs willproducesignificantlydifferenteffluentwaterwillproducesignificantlydifferenteffluentwater qualities,possiblyimpactreactordesignbythequalities,possiblyimpactreactordesignbythe retentionofadditionalorganicsretentionofadditionalorganics •• HermanowiczHermanowicz et.al(2006)clarifiedaNovaket.al(2006)clarifiedaNovak publicationthatsuggestedwhetheranMBRisMForpublicationthatsuggestedwhetheranMBRisMFor UFwouldimpactthebiologicaldesignUFwouldimpactthebiologicaldesign – HavingeitheranMForUFproducedsimilarCODatthe sameconditions DynamicDynamic CakeCake LayerLayer (Lee(Lee etet al.al. 2001)2001) • Solids(microbial floc )protectthe membranefromdirect exposuretoorganics • Actsasa “secondary ” membrane • Membranefouling ratewillincrease withaless effectivedynamic cakelayer – Poorflocculation USUS BureauBureau ofof Rec.Rec. ReportReport (2000)(2000)

RapidfoulingRapidfouling attributedtoMFattributedtoMF modulemodule ImpactofSRTorF/Mon MembraneFouling RationaleRationale F S = o M θH ⋅ XMLVSS •• TheSMBRprocessiscurrentlyTheSMBRprocessiscurrently limitedtoanMLSSconcentrationoflimitedtoanMLSSconcentrationof 10g/L10g/L •• TheF/MratioisakeyparametertoTheF/Mratioisakeyparameterto optimizereactortankdesignoptimizereactortankdesign –– Smalltank(lowHRT)Smalltank(lowHRT) –– Smalltank(highF:M)Smalltank(highF:M) RationaleRationale

Capital O&M Present Worth, $ Worth, Present

θθθH, time EquipmentEquipment andand ApparatusApparatus

•• PilotPilot scalescale SMBRSMBR •• TreatingTreating primaryprimary effluentfromeffluentfrom theCityofSantheCityofSan FranciscoFrancisco ’’sSEPsSEP – COD=325mg/L – TSS=98mg/L MembraneMembrane OperationOperation andand CharacteristicsCharacteristics •• Zenon500CModuleZenon500CModule •• Nominal=0.035Nominal=0.035 mm •• Flux=30L/mFlux=30L/m 2. hh •• Air=14L/sAir=14L/s •• IntermittentIntermittent aerationaeration •• 9minoperating9minoperating cyclefollowedbycyclefollowedby 30secrelax30secrelax ExperimentalExperimental MethodsMethods

•• Initialoperatingconditions:Initialoperatingconditions: MCRT=10d(F/M=0.34gCOD/gVSSMCRT=10d(F/M=0.34gCOD/gVSS .d)d) •• Dissolvedoxygen>2mg/LDissolvedoxygen>2mg/L •• ConstantMLSS=8g/LConstantMLSS=8g/L •• SteadySteady statedatacollectionbeganstatedatacollectionbegan after3MCRTsafter3MCRTs •• 2weeksteady2weeksteady statedatacollectionstatedatacollection periodperiod •• MCRTwassteadilydecreased(5,4,3,2MCRTwassteadilydecreased(5,4,3,2 d)d) – F/M(0.53,0.73,0.84,1.4gCOD/gVSS.d) MembraneMembrane PerformancePerformance atat 1010 --dd MCRTMCRT (F/M=0.34(F/M=0.34 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux

40 300 Start-up Chemical Clean Large Foam Event

35 250

30

200 25 C, LMH/bar C, o

20 150

15 100

10 Specific Flux @ 20 @ Flux Specific

50 5

0 0 50 70 90 110 130 150 170 190 Days of Operation MembraneMembrane PerformancePerformance atat 55 --dd MCRTMCRT (F/M=0.53(F/M=0.53 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux 40 300

35 250

30

200 25 C, LMH/bar o

20 150

15 100

10 @ Specific Flux 20

50 5

0 0 180 190 200 210 220 230 240 250 260 270 280 Days of Operation MembraneMembrane PerformancePerformance atat 44 --dd MCRTMCRT (F/M=0.73(F/M=0.73 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux 40 300 Intermittent Coarse Air Failure Foam Event

35 250

30

200 25 C, LMH/bar C, o

20 150

15 100

10 Specific Flux @ 20 @ SpecificFlux

50 5

0 0 270 280 290 300 310 320 330 Days of Operation MembraneMembrane PerformancePerformance atat 33 --dd MCRTMCRT (F/M=0.84(F/M=0.84 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux

40 300 Routine Feed Routine Feed Intermittent Coarse Line Cleaning Line Cleaning Air Failure

35 250

30

200 25 C, LMH/bar C, o

20 150

15 100

10 Specific Flux @ 20 Flux@ Specific

50 5

0 0 355 360 365 370 375 380 385 390 395 Days of Operation MembraneMembrane PerformancePerformance atat 22 --dd MCRTMCRT (F/M=1.4(F/M=1.4 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux

40 300 Foam Event

35 250

30

200 25 C, LMH/bar o

20 150

15 100

10 Specific Flux @ 20

50 5

0 0 390 395 400 405 410 415 Days of Operation EffectEffect ofof F/MF/M onon SteadySteady --StateState FoulingFouling RateRate

MCRT, d 10 5 4 3 2 4.0

3.5

3.0 y = 1.661x2.1977 R2 = 0.9517

2.5

2.0

1.5

1.0

0.5

0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

F/M, g COD/g VSS.d SteadySteady --statestate FoulingFouling RateRate vsvs sCODsCOD

Soluble COD COD Rejection

90 100

80 sCOD = 3.8x + 61.3 90 R2 = 0.30 80 70

70 60 60 50 COD Rejection = -1.5x + 63.0 R2 = 0.24 50 40 40 30 30

20 COD Membrane Rejection, % 20

10 10

0 0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Steady-state fouling rate, LMH/bar.d SteadySteady --statestate FoulingFouling RateRate vsvs SMPSMP

Protein Carbohydrate Total

70

Total SMP = 7.0x + 36.8 60 R2 = 0.77

50

40

30 SMPp = 2.8x + 20.5 R2 = 0.36

20 SMPc = 4.2x + 16.2 R2 = 0.72 10

0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Steady-state fouling rate, LMH/bar.d ConclusionsConclusions •• Highorganicloadingrates(F/M)Highorganicloadingrates(F/M) increasedmembranefoulingratesincreasedmembranefoulingrates •• BiologicalfoamingwascontrolledBiologicalfoamingwascontrolled mechanicallymechanically •• IncreasedsteadyIncreasedsteady statemembranefoulingstatemembranefouling ratescorrelatedwithSMP,notsCODratescorrelatedwithSMP,notsCOD •• UnderstandingmembranefoulingathighUnderstandingmembranefoulingathigh organicloadingratesallowsengineersorganicloadingratesallowsengineers todesignacompactSMBRwithout:todesignacompactSMBRwithout: – excessivemaintenancecostsor – failingtomeetthedesigncapacity WhydoeshighF/Mcause membranefouling EffectEffect ofof F/MF/M onon SteadySteady --StateState FoulingFouling RateRate

MCRT, d 10 5 4 3 2 4.0

3.5

3.0 y = 1.661x2.1977 R2 = 0.9517

2.5

2.0

1.5

1.0

0.5

0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

F/M, g COD/g VSS.d EquipmentEquipment andand ApparatusApparatus

•• BenchBench scalescale SMBRSMBR •• TreatingTreating primaryprimary effluentfromeffluentfrom theCityofSantheCityofSan FranciscoFrancisco ’’sSEPsSEP – COD=325mg/L – TSS=98mg/L MembraneMembrane OperationOperation andand CharacteristicsCharacteristics •• MitsubishiMitsubishi SteraporeSterapore ® •• NominalporesizeNominalporesize =0.4=0.4 mm •• Membraneflux=18Membraneflux=18 L/mL/m 2. hh •• CoarsebubbleairCoarsebubbleair =0.4L/s=0.4L/s •• 9minoperating9minoperating cyclefollowedbycyclefollowedby 30secrelax30secrelax ExperimentalExperimental MethodsMethods

•• Operatingconditions:Operatingconditions: MCRT=10d(F/M=0.50gCOD/gVSSMCRT=10d(F/M=0.50gCOD/gVSS .d)d) MCRT=2d(F/M=2.34MCRT=2d(F/M=2.34 gCOD/gVSSgCOD/gVSS .dd)) •• Dissolvedoxygen>2mg/LDissolvedoxygen>2mg/L •• ConstantMLSS=1.4g/LConstantMLSS=1.4g/L •• SteadySteady statedatacollectionbeganstatedatacollectionbegan after3MCRTsafter3MCRTs •• 2weeksteady2weeksteady statedatacollectionstatedatacollection periodperiod ToolsTools UsedUsed toto UnderstandUnderstand MembraneMembrane FoulingFouling • Steady statemembranefoulingrateduring operation • Molecularweightdistributionofinfluent, SMPandeffluent • FTIRofcleanandfouledmembranes • Batchfiltrationexperimentsexpressedas ModifiedFoulingIndex(MFI) – Stircellfiltrationofsteadystatemixedliquor withUF(NMWCO=300 kDa ,PES) – DatapresentedasMFIat20 oCand210 kPa • Fouledmembraneresistances FouledFouled MembraneMembrane ResistanceResistance TermsTerms

•• R=RR=R M+R+R F+R+R C •• RR =Total=Total resistanceresistance

•• RRM =Membrane=Membrane

•• RRC =CakeLayer=CakeLayer

•• RRF =Foulants=Foulants – OrganicsAdsorption – Inorganic Precipitation MembraneMembrane PerformancePerformance atat 1010 --dd MCRTMCRT (F/M=0.50(F/M=0.50 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux

40 600 Start up 66 Days at 10-d MCRT (F/M = 0.50 gCOD/gVSS.d) 35 500

30

400 25 C, LMH/bar C, o

20 300

15 200

10 Steady-state fouling rate Specific Flux @Flux 20 Specific

100 Chemical Cleaning 5

0 0 0 10 20 30 40 50 60 70 80 Days of Operation MembraneMembrane PerformancePerformance atat 22 --dd MCRTMCRT (F/M=2.34(F/M=2.34 gCOD/gVSSgCOD/gVSS .d)d)

Flux Specific Flux

40 600 25 Days at 2-d MCRT (F/M = 2.34 gCOD/gVSS.d) 35 500

30 Chemical Chemical Cleaning Chemical Cleaning Cleaning 400 25 C, LMH/bar C, o

20 300

15 200 Steady- 10 state Specific Flux Specific @ 20

100 5 Improper Wasting Volumes 0 0 75 80 85 90 95 100 105 Days of Operation SteadySteady --StateState MembraneMembrane FoulingFouling RatesRates

Steady-state Fouling F/M MCRT SMP SMP Total SMP Rate @ 20oC c p . . gCOD/gVSS d d LMH/bar d mg/L mg/L mg/L 0.50 10 2.60 24 14 38 2.34 2 59.0 10 49 59 •• MembranefoulingratesincreasedMembranefoulingratesincreased withF/MwithF/M •• TotalSMPconcentrationincreasedTotalSMPconcentrationincreased withF/MwithF/M •• UnlikepilotUnlikepilot scalework,scalework, SMPcSMPc diddid notincreasewithincreasingF/MnotincreasewithincreasingF/M CarbohydrateCarbohydrate MolecularMolecular WeightWeight IncreasedIncreased atat LowLow MCRTMCRT (High(High F/M)F/M)

> 10 kDa 10 kDa - 1 kDa < 1 kDa

25

20

15

10

5 Carbohydrate concentration, mg/L

0 Influent SMP - 10 d SMP - 2 d Effluent - 10 d Effluent - 2 d Sample ProteinProtein MolecularMolecular WeightWeight IncreasedIncreased atat LowLow MCRTMCRT (High(High F/M)F/M)

> 10 kDa 10 kDa - 1 kDa < 1 kDa

70

60

50

40

30

20 Protein Concentration, mg/L 10

0 Influent SMP - 10 d SMP - 2 d EFF - 10 d EFF - 2 d Sample FouledFouled MembraneMembrane FTIRFTIR ResultsResults Virgin (Blue) 1 0 0 10-d MCRT (Green)

8 0

6 0 2-d MCRT %T

4 0

2 0 1 0 4 0 0 0 3 0 0 0 2 0 0 0 1 0 0 0 6 5 0 Wavenumber[cm-1]

3380 - indicates OH stretching 1660 and 1540 - indicates NH and COO - (protein) 1060 - indicates CO stretching of polysaccharides FouledFouled MembraneMembrane ResistanceResistance 1010 --dd MCRTMCRT Fouling Resistance During 10 days MCRT Operation

25

Before cleaning y = 4.2319x 20 R2 = 0.9979 y = 0.4014x 2 Physical R = 0.5914 15 y = 4.0956x Chemical R2 = 0.9757

10

5

0 0 5 10 15 20 25 30 35 40 Viscosity * Flux (g/s2) After 66 d of operation without a chemical clean FouledFouled MembraneMembrane ResistanceResistance 22--dd MCRTMCRT

Fouling Resistance During 2 days MCRT Operation

25

Before cleaning y = 2.0627x R2 = 0.9919 20

15 y = 1.7468x Physical R2 = 0.9895 y = 0.4303x 10 R2 = 0.9315 Chemical

5

0 0 5 10 15 20 25 30 Viscosity * Flux (g/s2) After 5 d of operation without a chemical clean FouledFouled MembraneMembrane ResistanceResistance TermsTerms

A B

RMembrane 9% RMembrane 21% RCake 3%

RCake 15% RFoulant 64%

RFoulant 88%

R = 4.23x1012 m-1 R = 2.07x1012 m-1 Fouled membrane R distribution for SMBR: A) 10-d MCRT (0.5 gCOD/gVSS .d) B) 2-d MCRT (2.34 gCOD/gVSS .d) BatchBatch FiltrationFiltration ResultsResults • Operatingmembranepermeability wassimilarwhenanalyzed63and 71LMH/barforthe2 dand10 d MCRTs • Factorof2intotalfouled resistance • Usedabatchfiltrationtestto betterunderstandthese differencesandimportanceof variouscomponentstofouling • Stircellfiltrationofsteady statemixedliquorwithUF (NMWCO=300 kDa ,PES) • DatapresentedasMFIat20 oC and210 kPa BatchBatch FiltrationFiltration ResultsResults

ModifiedFoulingIndex,10 3 s/L 2 SRT, d Mixed Liquor Soluble SS Mixture Effect 10 17 11 2 4 2 47 27 12 8

• Highersludgeresistanceobservedat2 dMCRT • Reductioninsludgefilterabilitywas observedasmembranefouling – Fouledresistances4.23(10 d)and2.07(2 d)withmeasurementwasmadeonmembrane permeate – Fouledpermeability71(10 d)and63(2 d) withmeasurementwasmadeduringoperation BatchBatch FiltrationFiltration ResultsResults

ModifiedFoulingIndex,10 3 s/L 2 SRT, d Mixed Liquor Soluble SS Mixture Effect 10 17 11 2 4 2 47 27 12 8

• MFIwashigherforallfractionsatMCRT=2d • Sludgewascentrifugesat12,000gfor15 minutes – Solublefractionwassupernatant – Suspendedsolids(SS)fractionwaspellet • SSwasmeasuredby resuspending pelletwith batchstircellpermeate • SolubleMFIwasalmost3timeshigheratlow MCRT • SSMFIincreased6timesatlowMCRT(sticky cake) • Mixtureeffectwasobservedatbothconditions SMBR Sludge- Low EPS/High Colloidal Material

Activated Sludge- High EPS/Low Colloidal Material EPSEPS DataData

Mean Concentration, mg/gVSS MCRT, d < 1 kDa 10 kDa - 1 kDa > 10 kDa Total Carbohydrate 10 4.3 17.6 7.8 29.7 2 5.5 6.8 18.3 30.6 Protein 10 30.6 46.2 14.4 91.2 2 48.5 11.3 60.9 120.7

Nodifferenceintotalcarbohydrateconcentration mostcommonlycited foulant MoretotalproteinatlowMCRT MorehighmolecularweightorganicsatlowMCRT CarbohydrateCarbohydrate EPSEPS

A B

< 1 kDa < 1 kDa 14% > 10 kDa 18% 26%

10 kDa - 1 > 10 kDa kDa 60% 22% 10 kDa - 1 kDa 60% Carbohydrate Carbohydrate Total Concentration: 29.7±1.7 mg/gVSS Total Concentration: 30.6±1.5 mg/gVSS Dramatic shift between the >10kDa and 10-1 kDa range A) 10-d MCRT (0.5 gCOD/gVSS .d) B) 2-d MCRT (2.34 gCOD/gVSS .d) ProteinProtein EPSEPS

A B

> 10 kDa 16% < 1 kDa 34% < 1 kDa 40% > 10 kDa 51%

10 kDa - 1 10 kDa - 1 kDa kDa 50% 9% Protein Protein Total Concentration: 91.2±6.6 mg/gVSS Total Concentration: 120.7±20.3 mg/gVSS AGAIN - a dramatic shift between the >10kDa and 10-1 kDa range A) 10-d MCRT (0.5 gCOD/gVSS .d) B) 2-d MCRT (2.34 gCOD/gVSS .d) ConclusionsConclusions

•• Highorganicloadingrates(F/M)Highorganicloadingrates(F/M) increasedmembranefoulingratesincreasedmembranefoulingrates •• IncreasedsteadyIncreasedsteady statemembranefoulingstatemembranefouling ratescorrelatedwithtotalSMPratescorrelatedwithtotalSMP •• MWofcarbohydrateandproteinSMPMWofcarbohydrateandproteinSMP increasedwithF/MincreasedwithF/M •• MembranerejectedhigherMWSMPMembranerejectedhigherMWSMP •• FTIRindicatedproteinandcarbohydrateFTIRindicatedproteinandcarbohydrate presenceonfouledmembraneswithpresenceonfouledmembraneswith strongeradsorptionsresultingfromthestrongeradsorptionsresultingfromthe 22dMCRTconditiondMCRTcondition ConclusionsConclusions •• MembranefoulingwasprimarilyduetoMembranefoulingwasprimarilydueto theadsorptionoforganicsandRtheadsorptionoforganicsandR F waswas dominateresistancetermoffouleddominateresistancetermoffouled membranesmembranes

•• RRC increasedwithF/MandthiswasincreasedwithF/Mandthiswas attributedtochangesinattributedtochangesin flocfloc propertiesproperties thatresultinathatresultina ““stickysticky ”” cakecake •• Sludgefiltrationresistance(MFI)Sludgefiltrationresistance(MFI) increasedwithF/MincreasedwithF/M •• MFIofsuspendedsolidsincreased6MFIofsuspendedsolidsincreased6 times,supportingtheincreasingtimes,supportingtheincreasing importanceofthecakelayerwithimportanceofthecakelayerwith increasingF/MincreasingF/M