Engineering4(2018)597–609Contents lists available at ScienceDirectEngineeringResearch
WatershedEcology—Article
AFloatingIslandTreatmentSystemfortheRemovalofPhosphorusfromSurfaceWaters
MarkT.Browna,?,TreavorBoyera,R.J.Sindelara,SamArdena,AmarPersauda,SherryBrandt-WilliamsbabDepartmentofEnvironmentalEngineeringSciences,UniversityofFlorida,Gainesville,FL32611,USASt.JohnsRiverWaterManagementDistrict,Palatka,FL32177,USAarticleinfoabstract
Thegoalofthisprojectwastodesign,build,andtestapilot-scale?oatingmodulartreatmentsystemfortotalphosphorus(TP)removalfromnutrient-impairedlakesincentralFlorida,USA.Thetreatmentsys-temconsistedofbiologicalandphysical–chemicaltreatmentmodules.First,investigationsofprospectivebiologicalandphysical–chemicaltreatmentprocessesinmesocosmsandinbench-scaleexperimentswereconducted.Thirteendifferentmesocosmswereconstructedwithavarietyofsubstratesandcombi-ànationsofmacrophytesandtestedforTPandorthophosphate(PO34)removalef?cienciesandpotentialarealremovalrates.Bench-scalejartestsandcolumntestsofseventypesofabsorptivemediainadditiontothreecommercialresinswereconductedinordertotestabsorptivecapacity.Onceisolatedprocesstestingwascomplete,a?oatingislandtreatmentsystem(FITS)wasdesignedanddeployedforeightmonthsinalakeincentralFlorida.Phosphorusremovalef?cienciesofthemesocosmsystemsaveragedabout40%–50%,providinganaverageuptakeof5.0gámà2áaà1acrossallmesocosms.Thebest-performingmesocosmswereasubmergedaquaticvegetation(SAV)mesocosmandanalgaescrubber(AGS),whichremoved20and50mgámà2ádà1,respectively,foranaverageremovalof5.5and12.0gámà2áaà1fortheSAVandAGSsystems,Oftheabsorptivemedia,thebestperformancewasalumresidual(AR),whichàreducedPO34concentrationsbyabout75ˉter5minofcontacttime.Ofthecommercialresinstested,thePhosXresinwassuperiortotheothers,removingabout40%ofphosphorusafter30minand60ˉter60min.Underbaselineoperationconditionsduringdeployment,theFITSexhibitedmeanà3àPO34removalef?cienciesof53%;usingthe50thand90thpercentileofPO4removalduringdeployment,and the footprint of the FITS system, yielded ef?ciencies for the combined FITS system of 56% and 86%, respectively, and areal phosphorus removal rates between 8.9 and 16.5 gámà2áaà1.
ó 2018 THE AUTHORS. Published by Elsevier LTD on behalf of Chinese Academy of Engineering and Higher Education Press Limited Company. This is an open access article under the CC BY-NC-ND license
Articlehistory:Received2February2018Revised10April2018Accepted26April2018Availableonline14August2018Keywords:PhosphorusremovalLakeecosystemFloatingislandtreatment1.IntroductionInFlorida,USA,asinmanyregionsoftheworld,lakeecosys-temsareundergoingsigni?cantchangescausedbyhuman-inducedalterationofwatersheds,whichhavedegradedwaterquality.Thereisincreasingevidencethatlakeecosystems,moresothanothertypesofsystems,mayhaveecologicalthresholds[1],whichifexceededmakerecoveryexceptionallydif?cultandpotentiallyanonexistentalternative—duetoalternativestablestatesandasymmetricalrecoverytrajectories(hysteresis)[2].Theecologicalengineeringoflacustrinerecoveryeffortsmustrec-ognizethesecriticalissuesandadaptproactiveapproachesin?Correspondingauthor.E-mailaddress:mtb@u?.edu(M.T.Brown).ordertoavoidecosystemthresholdsaswellastoattemptrestora-tionofthoselakesthathavealreadypassedthem.Atpresent,157lakesinFlorida(totalingabout1.42?105hm2)arenutrientimpairedand116lakes(totaling1.59?105hm2)thatarecurrentlynotimpairedaredegrading,leadingtotheirquestion-ablestatusinthefuture[3].Therecognitionofimpairedwatersultimatelyrequirestheadoptionoftotalmaximumdailyloads(TMDLs)forthepollutantsthathavecausedtheimpairment.IfthelakeispartofaSurfaceWaterImprovementandManagement(SWIM)Program,thedevelopmentofpollutionloadreductiongoals(PLRGs)isrequiredbythewatermanagementdistrictsthatareresponsibleforthelake,andoftenprovidesthescienti?cbasisfortheTMDLs[4].AnintegralpartoftheTMDLprocessisthedevelopmentoftheBasinManagementActionPlan(BMAP)thatlistseachstakeholder’srestorationstrategiesformeetingtherequiredpollutantreductions.598M.T.Brownetal./Engineering4(2018)597–609Somelakesrepresentnutrientsinksbecausetheyreceivesur-facerunofffromtheirwatershedsandfromthewatershedsofalltheirtributaries.Becausetheyhaveareasoflow?owwithhighresidencetimes,theytendtoconcentratenutrients,thusrepre-sentinginterestingchallengesforrestoration.Wholesaleremovalofnutrientsthroughlakebottomdredgingisonealternative,albeitacostlyone[5–7].Furthermore,itlacksreductionoftheexternalanthropogeniccomponentthatisde?nedbytheTMDLprocessintheCleanWaterActandisalmostuniversallyrequiredinordertomaintainlowlevelsofnutrientandsedimentbuildupwithinthelake.Restorationcanalsoincludeconstructionoflarge-scalewetlandtreatmentcells,wherelakewatersarepumpedthroughthesystemtosequesternutrientsandslowlyreducein-lakenutri-entlevels[8,9].Apopularalternativeistheuseofalumtreatmentsystems,inwhichachemicalreactionprecipitatesthephosphorus(P)andsequestersthenutrientatthebottomofthewaterbodyorinanoff-line?occulentholdingpond.Inallthesecases,restorationrequiresasigni?cantquantityoflandincloseproximitytothelakeinquestion,eitherfortheconstructedwetlandorforthedisposalofsediments.Thelandsurroundingmanylakesiscompletelydevelopedandthereforedoesnotaffordtheopportunityforland-basedrestorationefforts.Theuseof?oatingislandtechnol-ogymaybeasolutiontothelackofopportunityforland-basedapplications.However,simplebiologicaltreatmentsystemsrelyingsolelyonplantuptakerequiretoomuchsurfaceareawithinthelakeand/oritstributaries.Innovativealternativescombiningplantcomponentsandadsorptivetechnologywithminimalbyproductgenerationandsmallfootprintsmayovercomemanyoftheobsta-clestonutrientreductioninurbanwatershedswithimpairedlakes.Inthisproject,wedesigneda?oatingislandwatertreatmenttechnologythatincorporatedbothabiologicalcomponent(?owthroughwetlandcells)andaphysical–chemicalcomponent(?uidizedbedsofabsorptivemedia)fortheremovaloftotalphosphorus(TP)andorthophosphate(PO3à4)fromlakewater.We?rsttestedvariousgrowingmediaandmacrophytecombina-tionstodeterminethecombinationsunderhighvolume?owconditionsandseveraldifferentabsorptivemedia.Tominimizethefootprintofthetreatmentislandandtomaximizeremovalef?ciency,weusedsolarphotovoltaic(PV)panelstoprovidepowerforwaterpumps,sincetheislandhadtobeportableandself-contained.Theoverallobjectivesoftheprojectwerethreefold.The?rstobjectivewastodemonstrateatapilotscalethatoneofseveralnutrient-removalprocessescanachieveef?cientandsustainableTPreductionoflakewater.ThesecondobjectivewastoassessthetotalyearlyTPreductioninkgáaà1andinkgámà2áaà1fromeachprocesstested,bothatbenchscaleduringoptimizationevaluationandthroughthe?oatingislandtreatmentsystem(FITS)thatwaseventuallydeployed.Thethirdobjectivewastoassessthecostofcapital,operation,andmaintenancethatwasrequiredtoupscaletheprocesstomeetFloridaStateef?uentcriteriabasedontheexperimentalpilotstudy.Inaddition,theFITSmustbeamooredormobile?oatingplatformwithouttheneedforanupland,teth-eredpowersource,andmustnotbeasigni?cantobstructionofnavigation.TheFITScannotcauseecologicallysigni?cantchangesinwaterchemistrybeyondreductionsinnutrientconcentrationsbetweenthein?uentandef?uent,andtheremustbenochemicalsintheef?uentthatwerenotinthein?uent;?nally,theremustbelittlelossofesthetics.2.MethodsSixmonthsofdatawerecollectedfrombiologicalmesocosmsandbench-scaletestsofabsorptivemediainpreparationforthedesignanddeploymentoftheFITS.Methodsforeachcomponentoftheprojectaregivennext.2.1.Biologicalmesocosmexperiments2.1.1.MesocosmdesignThemesocosm-scaletestsofbiologicaltreatmentsystemswereconductedonthecampusoftheUniversityofFlorida(UF),inGainesville,FL,USA.Fig.1showsanoverviewofthecompletebio-logicalmesocosmexperiment.AstoragetankwasnecessarytocontrolinputconcentrationsofPtothemesocosms.Inordertoensurealargequantityofwater,whichwasnecessary,theexper-imentsusedGainesville’spublicwatersupply.Anactivatedcarbon?lterwasusedtostripchlorinefromthepotablewaterpriortointroductiontothemesocosms.Figs.2and3provideconstructiondetailsforfourtypesofmeso-cosms:macrophytes,bio?lm,algaescrubber(AGS),andverticalbio?lter(VTF).Themesocosmswere0.75m3(0.5mwide?3.0mlong?0.5mdeep),constructedfromplywood,andlinedwith40mil(1mil=0.0254mm)polyvinylchloride(PVC)sheeting.Onceconstructed,eachmesocosmwasinoculatedwithwaterfromtheUFsewagetreatmentplantaswellaswaterfromLakeAliceontheUFcampustoinsuremulti-organismseeding.Mesocosmexperimentsweredesignedtoquicklyprovidebasicinformationregardingtheselectionofappropriatemacrophytesandgrowingmedia.Becauseofspaceandequipmentlimitations,themesocosmexperimentswereconductedintwophases.Inall,12mesocosmswereusedintwophases,asfollows:FirstPhase:??EMC:Emergentmacrophytemesocosmwithrecycledconcrete(RC)asthesubstrate;??EML:Emergentmacrophytemesocosmwithlavarockasthesubstrate;??VET:Vetivergrass(Chrysopogonzizanioides)mesocosmplantedhydroponicallywithinthemesocosm;??SAV:MesocosmcontainingthesubmergedaquaticplantCerato-phyllumdemersum;??BFJ:Bio?lmmesocosmcontainingthreePVCracks(0.5m?0.5m)holdingabout0.25kgofjute?bernettingper-pendiculartothedirectionof?ow;??BFB:Bio?lmmesocosmcontainingthreebalesofcoconut?ber(0.5m?0.5m?0.5m)weighingapproximate0.5kg,eachinstalledperpendiculartothedirectionof?ow;??AGS:Algaescrubbermesocosmconsistingofashallowchannelsaturatedwithlocallyobtainedalgae.SecondPhase:??PLM:Mesocosmcontainingemergentmacrophytes(samesixspeciesusedpreviously)plantedinaplasticregrindsubstrate;??PLV:Mesocosmcontainingvetivergrassplantedinaplasticregrindsubstrate;??BIB:MesocosmcontainingBioballsonly;Bioballsareacommer-ciallyavailablehighsurfaceareaplasticsubstratedesignedtopromotebio?lmgrowthtotreathighnutrientlevels;??PLO:Mesocosmcontainingplasticregrindwithoutplants;??VTF:Mesocosmcontainingaverticaltrickling?ltercomprisedofaplasticsubstrateinthemesocosmwithasolarpumptopro-videwatertohangingjutecurtains(0.5m?0.5m)holdingabout0.25kgofjuteabovethemesocosm.2.1.2.MesocosmoperationAllmesocosmswereoperatedusingadaylighton/nighttimeoffcyclethatwasmodi?edduringthelengthoftheexperimenttomatchdaylighthours.Eachevening,thestoragetank(Fig.1)was?lledwithwaterandspikedwithmonopotassiumphosphate(KH2PO4)toraisethemesocosmin?uentPO3à4tomatchtheM.T.Brownetal./Engineering4(2018)597–609599Fig.1.Biologicaltreatmentmesocosmexperimentaldesignlayout.Theactivatedcarboncolumnwasusedtoremovechlorinefromin?uentwater.Fig.2.Detailsoftheconstructionofthemesocosms.(a)Macrophytemesocosms(EMC,EML,VET,PLM,PLV,BIB,andPLO);(b)bio?lmmesocosms(BFJandBFB).1in=2.54cm.expectedlevelincentralFloridalakes(220lgáLà1ofP).Inaddition,à1nitrate(NOàwithsodiumnitrate3)levelswerespikedto800lgáL(NaNO3)toensurethatthesystemwasnotnitrogenlimited.SincewemaintainedaconstantconcentrationofPinthein?uent,massloadingofPwascontrolledineachmesocosmbyhydraulicloadingrates(HLRs).WevariedtheHLRsoverawiderange(from2.5to60cmádà1)totesttheresponseofvegetationandgrowingmedia;thus,themesocosmsexperiencedawiderangeofmassloadingandsubsequentPuptake.Mesocosmexperimentsweredesignedtoquicklyprovidebasicinformationregardingtheselectionofappropriatemacrophytesandgrowingmedia.Sinceamajorconcerninthelongrunwastheestheticsofthe?oatingislands,weselectedmacrophytesbasedonacombinationof?oweringandPuptake.Mesocosmscontainingemergentmacrophyteswereplantedwithanequaldis-tribution(tenplants)ofthefollowingspecies:Canna?accida,Sag-gitarialancifolia,Pontederiacordata,Peltandravirginica,Orontiumaquaticum,andHymenocallisspp.Inaddition,wewereinterestedin?ndinggrowingmediawithsuf?cientporespacethatwouldaccommodaterelativelyhighHLRsofupto60cmádà1.The?rstsetofmesocosmswasoperatedfrom24Aprilto15July2009andthesecondsetwasoperatedfrom20Julyto15December600M.T.Brownetal./Engineering4(2018)597–609Fig.3.Detailsoftheconstructionofthemesocosm.(a)AGS;(b)VTF.2009.Duringtheoperationofbothsets,individualmesocosmswerediscontinuedorwerecontinuedforlongerthantheopera-tionalphasesgivenabove,dependingontheirperformance.Duringtheoperationalphaseofthe?rstsetofmesocosms,thelengthofoperationvariedforindividualmesocosms.Themesocosmcon-tainingRCconsistentlyexhibitedaveryhighpH,theresultoftheconcretedust,andwediscontinueditsoperationafterfourweeks.ThehighestPremovalef?ciencywasfoundintheBFJ;however,asthejutedegraded,thedemandforoxygenincreased,resultinginverylowdissolvedoxygen(DO)inef?uentwaters,aconditionthatwouldbeunacceptableforalakeapplication.TheBFBalsosufferedfromlowDO,anditsoperationwashaltedaftereightweeks.ThevegetationintheSAVmesocosmsuccumbedtotem-peratureduringtheearlysummer,andwehalteditsoperationaftersixweeks.Thesecondsetofmesocosmswasconstructedbasedontheknowledgegainedinthe?rsttrials,whichwereprimarilyintendedtotestdifferentgrowingmedia.Thenewmediumwasaplasticregrindythatweobtainedfromaplasticsmanufacturer,whichwasdestinedforland?llingbecauseofitspoorqualityforrecycling.Thechipsizeoftheregrindwasabout0.5cmandaffordedexcellenthydrauliccharacteristics,withacross-sectional?owthroughthemediumthatwashighenoughtomeetourrequirementsofHLRsofupto60cmádà1.Inadditiontotheregrind,wetestedacommer-ciallyavailableplasticsubstratewithahighsurfaceareatopromotebio?lmgrowth(Bioballs,BIB),andaVTF.TheVTFwasdesignedasadripbio?ltercomposedofjutestripshungverticallywithwaterdis-chargedovertheminordertoovercometheanoxiaofcontinuouslyinundatedjuteintheBFJofthe?rstsetofmesocosms.
TheHLRsofthe?owthroughthemesocosmsinthesecondsetwerechangedapproximatelyevery30dduringthe148dofoper-ation.Theywerestartedat30cmádà1,thenincreasedto45cmádà1,increasedagainto70cmádà1,decreasedto45cmádà1,andwere?nallydecreasedto20cmádà1.TheVTFmesocosmwasoperatedat55cmádà1withvariationintheHLRresultingfromthefactthatwaterwaspumpedovertheverticaljuteviaasolarpump.2.1.3.SamplingprocedureformesocosmexperimentsWatersampleswerecollectedeveryotherdayduringthe?rsttwomonthsofoperationandeveryfourthdaythereafter.Allwatersampleswerecollectedandcompositedasdailysamplesinbrown(opaque)Nalgenebottlesthatwerespeci?ctoeachmesocosm.Adrip-samplingori?ceattheout?owofeachsystemwasadjustedtoprovidea500mLsampleevery12h.Samplebottleswerewashedthreetimeswithsamplewaterbeforebeingplacedonthedripline.Attheendofeachsampleday,sampleswererefrig-àeratedat4°CforPO3analysis.Phosphorusanalyseswereper-4formedwithinthreedaysofsamplecollection,inaccordancewithEPAMethod365.1[10].Flowratesthroughthemesocosmswereadjustedeverymorn-ingandmeasuredtwiceadayinordertoensureconstantratesusingastopwatchanda1000mLgraduatedplasticcylinder.DO,temperature,andpHweremeasureddailyineachmesocosm.2.2.Bench-scalejartestsandcolumnexperiments2.2.1.MaterialsTwoclassesofmaterialswereevaluatedinthiswork:adsor-bentsandion-exchangeresins.Theadsorbentsevaluatedinthisworkconsistedofwastebyproductsandnaturalmaterials,asdescribedinTable1.Theadsorbents,drinkingwatertreatmentalumresidual(AR),ironslag(IS),andsteelslag(SS)werecrushedwithamortarandpestleandsievedthroughUSStandardsieves30and40,toyieldaparticlesizerangeof420–595lm.Drinkingwatertreatmentferricresidual(FR),RC,andlimestone(LS)weredriedunderambientlaboratoryconditionsbeforebeingcrushedasdescribedabove.ClassF?yash(FA)wasreceivedinpowderedformandwasusedinitsoriginalstate.Theion-exchangeresinsPlasticregrindisgroundorchopped?ash,runners,sprues,andnon-contaminatedrejectedpartsfromplasticmanufacturingthatareproducedbyamolderininitialmoldingprocesses.Thesematerialsarecrushedtosmallersizeandrecycledwithvirginmaterials.Regrindappliestopost-industrial(pre-consumer)waste.
yM.T.Brownetal./Engineering4(2018)597–609Table1
Adsorbentsusedinjartestsandcolumnexperiments.AdsorbentDrinkingwatertreatmentARDrinkingwatertreatmentFRGranulatedblastfurnaceISBasicoxygenfurnaceSSFARCLSSourcePeaceRiverManasotaRegionalWaterSupplyAuthority,Arcadia,FLDavidL.TippinWaterTreatmentFacility,Tampa,FLCivil&MarineInc.,CapeCanaveral,FLLevyEnterprises,Valparaiso,INBoralMaterialsTechnologies,Tampa,FLFloridaConcreteRecyclingInc,Gainesville,FLFloridaRockIndustriesInc.,Gainesville,FLDescription601SurfacewatertreatmentplantthatusesaluminumsulfatetotreatwaterfromthePeaceRiverSurfacewatertreatmentplantthatusesferricsulfatetotreatwaterfromtheHillsboroughRiverNon-metallicbyproductfromironproductionByproductofmanufacturingsteelfrompigironCombustionbyproductofcoalConcreteaggregatecollectedfromdemolitionsitesNaturalrockminedfromvariouslocationsusedinthisworkaredescribedinTable2,andwereusedasreceived.2.2.2.TestwatersSincethequantityoftestwaterforjarandcolumntestswasrel-ativelysmall,weusedwatercollectedfromtwocentralFloridalocations.The?rstlocationwasSanfordAvenueCanal,atributarytoLakeJesup,nearSanfordFlorida;thesecondsourcewasLakeAliceonthecampusoftheUFinGainesville.Table3showsthewaterqualityforSanfordAvenueCanalandLakeAlicetakenfromJanuarythroughAugust2009.2.2.3.JartestsJartestswereconductedtoinvestigatetheeffectivenessoftheàadsorbentsandion-exchangeresinsinremovingTPandPO34fromthetestwaters.Theadsorbentsweredosedgravimetrically(0.5–8gáLà1)andtheion-exchangeresinsweredosedvolumetri-cally(0.5–4mLáLà1).APhippsandBirdPB-750JarTesterequippedwitheither2Lor500mLjarswereused.Thebaselineexperimen-talprotocolwasasfollows:Addadsorbentorion-exchangeresintotestwater,rapidlymixat100ráminà1for60min,andallowtosettlefor30min.TheFR,IS,SS,RC,andLSwererapidlymixedat200ráminà1becausethesematerialsweredenserthantheAR,FA,andion-exchangeresins.Samplesweretakenafter5and30minofmixing,andafter60minofmixingand30minofset-àtling.SampleswereanalyzedforTP,PO34,pH,turbidity,ultravioletabsorbanceat254nm(UV254),totalorganiccarbon(TOC),andtotalnitrogen(TN).UV254andTOCarecommonlymeasuredsurro-gatesfororganicmatter.2.2.4.ColumnexperimentsThecolumnexperimentswereconductedusingacolumnwithaninnerdiameterof0.7854cmandaheightof2cm,andwith25lmporesizepolyethylenefritsoneachend.Thecolumnwasthen?lledwith1mLofwetadsorbentorion-exchangeresin.Byde?nition,1bedvolume(BV)wasequalto1ml.Tubingwascon-nectedtoallowforanup-?owdirection.The?owrateusedwas2mLáminà1or2BVáminà1.Beforeacolumnexperimentwasstarted,thesystemwas?ushedbypumping120BVofdeionized(DI)waterthroughthecolumn.Watertobetreatedwas?lteredthroughaWhatmanGF/A?lter(1.6lmporesize)tocontrolthecloggingoffrits.Twoscenariosweretested:continuous?owandintermittent?ow(12-h-on/12-h-off),where1hlongsamplesweretakenevery3handleftto?owwithoutsamplingfor12horlefttorestfor12hduringthenight,respectively.Apredeterminedbreak-àthroughPO3concentrationwassetat50%removal,afterwhich4oneextrasamplewastakenatthenextsamplepointtocon?rmtheendofanexperiment.Watersampleswereanalyzedforthesameparametersasthejartests.Table2
Ion-exchangeresinsusedinjartestsandcolumnexperiments.Ion-exchangeresinPhosXMIEXDowex22ManufacturerStructure2.3.Floatingislandtreatmentsystem2.3.1.MaterialsFig.4isadrawingofthe‘‘as-built”FITS.ThedesignutilizedaHobieCatòhaulandframeworkwithadditional?oatationaddedtosupporttheweightofthetreatmentsystems(estimatedatabout1100–1200lb(1lb=0.4536kg)whenfullofwater).Theadditional?oatingstructureutilizedStyrofoamdock‘‘billets”thatwereplacedbetweenthetwohaulsoftheHobieCatò.SolmeteXOricaWatercareDowChemicalMacroporouspolymerresinimpregnatedwithironoxideparticlesMacroporous,polyacrylicanion-exchangeresinwithstrong-base,typeIIfunctionalgroupsMacroporous,polystyreneanion-exchangeresinwithstrong-base,typeIIfunctionalgroupsTable3
Waterqualityoftestwatersusedinjartestsandcolumnexperiments.LocationSampleMonthof2009SanfordAvenueCanalRawFilteredaRawFilteredaRawFilteredaRawFilteredaAprilAprilJuneJuneJulyJulyAugustAugust7.67.76.97.67.77.77.67.5pHTurbidity(NTU)4———————àPO34(gáLà1ofP)TP(gáLà1Chloride)(mgáL1851855555141416—à1)Sulfateà1(mgáL)333366191821—TOCà1(mgáL)141534317.27.18.38.1UV254à1(cm)0.6410.5931.631.640.2280.2190.2250.228222215107107421387507272329250195143482423529327LakeAliceNTU:nephelometricturbidityunit;TOC:totalorganiccarbon;UV254:ultravioletabsorbanceat254nm.aWhatmanGF/A?lter.
一种去除地表水中的磷的浮岛处理系统 - 图文
