modeling pcb loadings reduction scenarios to the lake ......modeling pcb loadings reduction to the...

Modeling PCB Loadings Reduction Scenarios to the

Lake Washington Watershed: Final Report

March 2014

Alternate Formats Available

Modeling PCB Loadings Reduction Scenarios to the Lake Washington Watershed: Final Report

Prepared for: US EPA Region 10

Submitted by: Richard Jack, Jenée Colton, Curtis DeGasperi, and Carly Greyell King County Water and Land Resources Division Department of Natural Resources and Parks

Funded by the United States Environmental Protection Agency Grant PC-J28501-1

Disclaimer: This project has been funded wholly or in part by the United States Environmental Protection Agency under assistance agreement PC-00J285-01 to King County. The contents of this document do not necessarily reflect the views and policies of the Environmental Protection Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

ModelingPCBLoadingsReductiontotheLakeWashingtonWatershed:FinalReport

KingCounty i March2014

Acknowledgements Afantasticgroupofpeoplehavemadethisprojectpossibleandcontributedtoitssuccess.ColinElliottoftheKingCountyEnvironmentalLaboratoryprovidedlaboratoryprojectmanagementservicesthroughoutthefieldstudy.ManythanksgotoBenBudka,MarcPatten,DavidRobinson,BobKruger,JimDevereaux,andStephanieHess,wholedtheyear‐longfieldsamplingprogram.TheknowledgeablestaffatAXYSAnalyticalServicesprovidedhighqualityadviceandanalyticalservices.ArchieAllenandKirkTullaroftheWashingtonDepartmentofTransportationprovidedcriticalassistanceduringsitingandsamplingofthehighwaybridgerunofflocation.JohnWilliamsonoftheWashingtonDepartmentofEcologygraciouslyallowedusaccessandsharedspaceattheBeaconHillairstation.LauraReedandvariousstaffatSeattlePublicUtilitiessupportedusinsiteselectionandsetupofmultipleSeattlesamplinglocations.SpecialthanksgotoJonathanFrodge,JennyGaus,RonStraka,andPatrickYamashitafortakingextratimetomeetwithusandprovideinformationduringsiteselection.GinnaGrepo‐GroveofEPARegion10assistedwithdatavalidationruleinterpretation.

ThanksalsogototheTechnicalAdvisoryPanelmemberswhoprovidedfeedbackatourinitialresultsandloadingestimatepresentation.SpecialthankstopanelmembersJonathanFrodgeandHeatherTrimwhocontributedsubstantiallytothetributaryflowandcontaminantloadingextrapolationapproachpresentedintheloadingsreport.

WethankalloftheLakeWashingtonLoadingsprojectadvisorypanelmembersforreviewingthedraftprojectdocumentsandtheirconstructivecriticismsthroughout.

TechnicalAdvisoryPanelMembers(inalphabeticalorder)were:

FredBergdolt,WADept.ofTransportation BetsyCooper,KingCountyWastewaterTreatmentDivision JonathanFrodge,SeattlePublicUtilities JennyGaus,CityofKirkland JoanHardy,WADept.ofHealth RachelMcCrea,WADept.ofEcology DougNavetski,KingCountyWaterandLandResourcesDivision AndyRheaume,CityofRedmond RonaldStraka,CityofRenton BruceTiffany,KingCountyWastewaterTreatmentDivision HeatherTrim,PeopleforPugetSound/FutureWise PatrickYamashita,CityofMercerIsland

ValuableadviceandfeedbackontechnicalaspectsofbothfateandbioaccumulationmodelswereprovidedbyGregPelletieroftheWashingtonStateDepartmentofEcology(Ecology).WeacknowledgeDebLesterandJimSimmondswhoprovidedmanysuggestionsthroughoutthatresultedinimprovementstotheprojectoverallandtothecomponentreports.


KingCounty ii March2014

Citation KingCounty.2014.ModelingPCBLoadingReductionstotheLakeWashingtonWatershed:

FinalReport.PreparedbyRichardJack,JenéeColton,CurtisDeGasperiandCarly

Greyell,WaterandLandResourcesDivision.Seattle,Washington.


KingCounty iii March2014

Table of Contents ExecutiveSummary.................................................................................................................................................v

1.0. Introduction.................................................................................................................................................1

2.0. ProjectFindings..........................................................................................................................................3

2.1 FieldStudy...............................................................................................................................................3

2.2 tPCBLoadingsEstimates...................................................................................................................7

2.3 tPBDELoadingsEstimates.............................................................................................................10

2.4 LakeWashingtonPCBFatemodel..............................................................................................11

2.5 LakeWashingtonPCBBioaccumulationmodel....................................................................12

2.6 LoadReductionTissueRecoveryScenarios...........................................................................13

3.0. NextStepsandRecommendedActions.........................................................................................18

3.1 UpdatetheConceptualModelofPCBsintheLocalEnvironment................................19

3.2 SourceControl‐DevelopaPCBInventory.............................................................................19

3.2.1 PerformRegionalSourceTracingUsingCongenerData.............................................20

3.3 ConductOutreachandEducationtoDecision‐makersandPublic...............................20

3.4 EvaluateEffectivenessofStormwaterTreatmentandLIDMethods..........................21

3.5 DevelopAirshed&WashoffModels..........................................................................................22

4.0. Conclusions................................................................................................................................................24

5.0. References..................................................................................................................................................26

Figures Figure1. Fieldstudysamplingstations......................................................................................................4

Figure2. tPCBaverageandstandarddeviationconcentrationsinmajorwaterpathways.6

Figure3. tPBDEaverageandstandarddeviationconcentrationsinmajorwaterpathways..............................................................................................................................................7

Figure4. PredictedwholefishtissueconcentrationswithPCBloadreductions..................14

Figure5. PredictedresponseofwatercolumntPCBconcentrationsto10,25,50,and85percenttPCBloadreductions...................................................................................................15

Figure6. PredictedresponseoftPCBsedimentconcentrationsto10,25,50,and85percenttPCBloadreductions...................................................................................................15


KingCounty iv March2014

Figure7. PredictedfilletconcentrationsoftPCBsunder10,25,50,80,and90percentloadingreductionscenarioscomparedtoWADOHhumanhealthfilletscreeningconcentrations................................................................................................................................17

Tables Table1. Average,25thand75thpercentiletPCBloadingratesforpathwaystoLake

Washington,LakeUnion/ShipCanal,andtoPugetSound(g/yr)............................10

Table2. Average,25thand75thpercentiletPBDEloadingratesforpathwaystoLakeWashington,LakeUnion/ShipCanal,andtoPugetSound(g/yr)............................11

Table3. PredictedsteadystatetPCBconcentrationsinLakeWashingtonwaterandsedimentunderselectedloadreductionscenarios........................................................14

Table4. WADOHfilletscreeningconcentrationsfortPCBs..........................................................16


KingCounty v March2014

EXECUTIVE SUMMARY WashingtonDepartmentofHealth(WADOH)issuedafishconsumptionadvisoryonLakeWashingtonnorthernpikeminnow,cutthroattroutandyellowperchin2006becauseofhumanhealthrisksfromPCBcontamination(WADOH2004).Givenconcernsabouthowtoaddresstheadvisory,in2010,KingCountywasawardedaU.S.EnvironmentalProtectionAgency(USEPA)PugetSoundScientificStudiesandTechnicalInvestigationsAssistanceGranttoestimateloadingoftotalpolychlorinatedbiphenyls(tPCBs)andtotalpolybrominateddiphenylethers(tPBDEs)toLakeWashington,LakeUnion,andPugetSound.ThegrantalsosupportedmodelingthepotentialdecreaseinLakeWashingtonresidentfishtissueconcentrationsassociatedwithselectPCBloadingreductionscenarios.Thisreportisthefinalproductofthegrantproject,whichsynthesizes:

a) Thefieldstudycomponents(KingCounty2013a),

b) EstimatesoftPCBloadingstoLakesWashingtonandUnion,andthesubsequentloadexportedtoPugetSoundbytheShipCanal(KingCounty2013b),

c) ModelingeffortstodescribefateoftPCBloadsinthelakeandtPCBbioaccumulationinfish(KingCounty2013c),and

d) Loadreductionscenariooutcomes.

Theprojectelementsweredevelopedinconjunctionwiththeproject’sadvisorypanel.Thefindingsincludetheircollectiverecommendationsforfurthertechnicalandmanagementactions.ThisreportalsopresentsrecommendationsfornextstepstobetterunderstandPCBsources,pathways,andbioaccumulationinLakesWashingtonandUnion,andsubsequentinputstoPugetSound.

Theproject’sfieldstudycomponentcollectedandanalyzed146samplesforPCBsandPBDEsfrom:

1) AmbientLakeWashingtonwaters

2) AmbientLakeUnionandShipCanalwaters

Andalsofromthefollowingsixinputpathways:

3) Threecreeksduringbothbaseflowandstormconditions,

4) TheCedarandSammamishRivers,

5) Threecombinedseweroverflow(CSO)dischargelocations,

6) Sixmunicipalstormwaterdischargelocations,

7) Twocombinedwetanddryatmosphericdepositionlocations,and

8) Onehighwaybridgestormwaterdischarge.

Themajorfindingsareasfollows.tPCBloadingstoLakeWashingtonwereestimatedat672g/yrwhichismorethantheestimated140g/yrloadexportedfromthelakeoutletattheMontlakeCut.Therefore,LakeWashingtonservesasapartialsink(repository)forPCBs,primarilyduetosedimentaccumulationandburial.LakeWashingtonPCBsareexportedtoLakeUnionandtheShipCanalandtheyalsovolatilizefromthelakesurface.Of


KingCounty vi March2014

theestimated2,023g/yrtPBDEloadtothelake,only800g/yrexitsattheMontlakeCuttoLakeUnionandtheShipCanal,thusLakeWashingtonactsasasinkforPBDEsaswell.

LakeUnionandShipCanaltPCBandtPBDEloadsincreasefromtheoutletofLakeWashingtonattheMontlakeCuttoPugetSound,partiallyduetoloadsfromthecombinedsewer/stormwatersysteminSeattle.CSOdischargesareestimatedtocontributemoretPCBsandtPBDEstoLakeUnionandtheShipCanalthantoLakeWashington.AveragetPCBloadingfromtheShipCanal’sHiramM.ChittendenLockstoPugetSoundwasestimatedat360g/yr.Normalizedtowatershedarea,360g/yrequals6.27x10‐4g/km2/daywhichiscomparabletootherpublishedurbanwatershedrates,buthigherthanruralPugetSoundwatershedrates.AveragetPBDEloadingfromtheShipCanal’sHiramM.ChittendenLockstoPugetSoundwasestimatedat990g/yrtPBDEs.BasedonfatemodelingofinstantaneoustPCBloadreductions,thelargestchangesinLakeWashingtonsedimentandwaterconcentrationswouldoccurwithin20years;upto40yearsisrequiredtoreachequilibrium.While20to40yearsisalongtime,intheabsenceofsubstantialwatershed‐wideeffortstoreducetPCBloads,theexistingfishconsumptionadvisoryisprojectedtoremainindefinitely.

DespiteabanontheproductionandmanyusesofPCBsinthelate1970s,foodwebbioaccumulationmodelingpredictsthatapproximatelyanadditional85percentreductionintPCBloadingisrequiredtoreduceLakeWashingtonfishtissueconcentrationsandremovetheexistingWADOHfishconsumptionadvisory.Toprogresstowardreductionsofthismagnitudewerecommend:

TraceandidentifyongoingPCBsourcesincurrentandhistoricallyusedmaterials.

Developingastate‐wideand/orregionalPCBinventorytoenabletargetedsourcecontrolactions,

Conductoutreachandengagedecision‐makersandthecommunityindiscussionaboutthewidespreadsourcesofPCBs,andthefinancialandregulatorychallengesinherentincontrollingsuchsources,

EvaluatetheeffectivenessoftreatmenttechnologiesandbestmanagementpracticesforPCBremoval,and

Developbulkdeposition/washoffmodelstodeterminecontributionofatmosphericPCBdepositiontostormwaterrunoff.


KingCounty 1 March2014

1.0. INTRODUCTION WashingtonDepartmentofHealthissuedafishconsumptionadvisoryonnorthernpikeminnow,cutthroattroutandyellowperchin2006becauseofhumanhealthrisksfromPCBcontamination(WADOH2004).Inaddition,asimilarlypersistentandmoremodernchemical,polybrominateddiphenylethers(PBDEs)isalsobioaccumulatinginLakeWashingtonfish(KingCountyunpublisheddata).ElevatedconcentrationsofPCBsandPBDEsarenotuniquetoLakeWashington,asevidencedbychemicalconcentrationsmeasuredinmarinemammalandfishtissuefromPugetSound(Rossetal.2000,Rossetal.2004,Krahnetal.2007,Westetal.2008,Sloanetal.2010,Cullonetal.2005).However,todatetherehavenotbeenanystudiesfocusedonhowthesechemicalsaregettingintoLakeWashingtonfishorthequantityofthesechemicalsenteringLakeWashington,LakeUnionandtheShipCanal(LakeUnion),andPugetSoundfromthiswatershed.Therefore,primarygoalsofthisprojectaretohelpfillPCBandPBDEdatagapsfortheLakeWashingtonwatershedandPugetSoundbasinandprovideinformationandtoolsneededtodirectmanagementofPCBsandreducehealthrisksassociatedwithLakeWashingtonfishconsumption.

In2010,KingCountywasawardedaU.S.EnvironmentalProtectionAgency(USEPA)PugetSoundScientificStudiesandTechnicalInvestigationsAssistanceGrantto(1)estimateloadingoftotalpolychlorinatedbiphenyls(tPCBs)andtotalpolybrominateddiphenylethers(tPBDEs)toLakeWashington,LakeUnionandPugetSound,and(2)modelthepotentialdecreaseinLakeWashingtonfishtissueconcentrationsassociatedwithselectPCBloadingreductionscenarios.tPCBswerethefocusofthemodelingbecauseLakeWashington’scarpandnorthernpikeminnowarecurrentlyunsafetoeatinanyamountandcutthroattroutorlargeyellowperchareonlysafeinlimitedamounts(WADOH2004).WhilesmallmouthbassarenotpartoftheWashingtonDepartmentofHealth(WADOH)advisory(2004),theyhavesincebeenfound(KingCounty2013e)tohavesimilartPCBconcentrationstothoseinnorthernpikeminnow.

UnderstandingtherootcausesofthehighPCBlevelsinfishiscriticaltoattainingsafetissueconcentrations.tPCBswerenotmodeledinLakeUnionduetoalackofresidentfishtissuedataforLakeUnion.tPBDEswerealsonotmodeledbecausetheyarenotpartoftheexistingLakeWashingtonfishconsumptionadvisoryandinformationabouttheirconcentrationsininvertebratesandpreyfishwaslacking.

Tofocusthefieldandmodelingefforts,threemanagementquestionsweredeveloped:

1) WhichtypesofimportpathwaysarethehighestprioritiesfortPCB/tPBDEloadingreduction?

2) HowwillpotentialloadingreductionsfromthesepathwaysreducethemagnitudeofresidentfishtissuePCBconcentrationsandtheneedforafishconsumptionadvisoryonLakeWashington?

3) HowlongmightthesystemtaketorespondtothesehypotheticaltPCBloadingreductions?



Thisreportisthefinalproductofthegrantprojectwhichsynthesizes:

a) Thefieldstudycomponents(KingCounty2013a),

b) LoadingestimatestoLakesWashingtonandUnion,andtheloadexportedtoPugetSoundbytheShipCanal(KingCounty2013b),

c) ModelingeffortsdescribingthefateofPCBloadsinthelakeandPCBbioaccumulationinresidentLakeWashingtonfish(KingCounty2013c),and

d) Loadreductionscenariooutcomes,Section2.5ofthisreport.

Theprojectelementsweredevelopedinconjunctionwiththeproject’sadvisorypanel.Thefindingsincludetheircollectiverecommendationsforfurthertechnicalandmanagementactions.

ThisprojectcollectedthefirstextensivemeasurementsoflowleveltPCBandtPBDEconcentrationsinwholewaterfromLakeWashingtonandtheShipCanal,aswellasinavarietyofotherwaterinputpathways(KingCounty2013a).Thepathwaysincludedwetanddrybulkatmosphericdeposition(dustandrainfall)becausePCBsandPBDEsaresemi‐volatilecompoundswhichmayseasonallycyclebetweensoils,water,andtheatmosphere(Hornbuckleetal.1994,Kayaetal.2012).ThetPCBandtPBDEwateranddepositiondatawerethenusedtoestimatecontaminantloadingstoLakesWashingtonandUnionandsubsequentlyfromthemajorityofthe(WRIA8)watershedtoPugetSoundviatheHiramM.ChittendenLocks.Loadingsestimatesarepresentedinaseparatereport(KingCounty2013b)thataddressedthefirstmanagementquestionabove.

HistoricalandprojectPCBdata(KingCounty2013a)werethenusedtodevelopfateandbioaccumulationmodelsforLakeWashington(KingCounty2013c).Incombinationwiththeloadingsinformation,thesemodelswereusedtoinformpriorityloadingreductionpathwaysandestimatethelengthoftimetheLakeWashingtonecosystemmighttaketorespondtochangesintPCBloadings.

Thisfinalreportconcludeswithadescriptionofthemagnitudeofloadingreductionsneededtoachievesafefishtissuelevels.SuggestionsandrecommendationsarealsoprovidedfornextstepstobetterunderstandPCBsources,pathways,andbioaccumulationinLakesWashingtonandUnion,andsubsequentinputstoPugetSound.TheseincludeactionsthatmaybeusedtoreducePCBloadstoLakeWashingtonfromthewatershed.



2.0. PROJECT FINDINGS Theprojectconsistedoffourcomponents:afieldstudy,loadingsanalysis,fateandbioaccumulationmodeling,andloadreductiontissuerecoveryscenarios.Detailsofthefirstthreecomponentsareprovidedinpriorreports(KingCounty2013a,b,c)andaresummarizedbelow.Adetaileddescriptionofthescenariosexaminedtoreducefishtissueconcentrationsisalsopresentedinthissection.

2.1 Field Study ThisprojectcollectedandanalyzedsamplesfromlakereceivingwatersandsixdifferentlakeinputpathwaytypestosupportdevelopmentofmassloadingsestimatesoftPCBsandtPBDEstoLakesWashingtonandUnion,andsubsequentlytoPugetSoundfromthemajorityofthegreaterCedar‐SammamishLakeWashingtonWaterResourceInventoryArea(WRIA8).Thereceivingwatersevaluatedincluded:(1)LakeWashingtonand(2)LakeUnionandtheShipCanal.Theinputpathwaysincluded:(1)threecreeksinthewatershed;sampleswerecollectedduringbothbaseflowandstormconditions;(2)theCedarandSammamishRivers;(3)threecombinedseweroverflow(CSO)locations;(4)sixstormwaterdischargelocations;(5)highwaybridgerunoff;and(6)combinedwetanddryatmosphericdeposition.Atotalof146sampleswerecollectedandanalyzedforPCBandPBDEcongenersandselectconventionalparameters.Wholewatersamples(whichincludedbulkairdeposition)weresuccessfullycollectedaccordingtothestudydesignspecifiedintheQualityAssuranceProjectPlan(QAPP)(KingCounty2011a)withafewmodifications(KingCounty2013a).AnoverviewofthesamplinglocationsbytypeisprovidedinFigure1.

SomeoftheanticipateddataqualityobjectivesofprecisionforPCBandPBDEcongeners1werenotmetasoutlinedintheQAPP.Particularlylowreproducibilityoccurredinambientlakeandriverwaters:LakeWashington,theShipCanal,andtheCedarandSammamishrivers.ThisstudywasunabletoreliablymeasuretheselowtomoderatetPBDEconcentrationsinregionalambientlakeandriversamplesbecauseofbackgroundlaboratorycontaminationandtheoverallubiquityofPBDEs.

1Thereare209uniquePCBand209uniquePBDEcongeners.Eachhasslightlydifferentchemicalpropertiesbasedontheirspecificformulaandshape.ThisprojectanalyzedallPCBcongeners,butresultsareexpressedasasum(total).TheprojectonlyanalyzedthemostcommonninePBDEcongenersandreportedtheirtotals.



Figure 1. Field study sampling stations.



tPCBandtPBDEconcentrationsdetectedinLakeWashingtonandriversampleswerethelowestofallsamplesmeasured.Thiswasexpectedgiventhatparticle‐associatedcontaminantslikePCBsandPBDEsareprovidedalongsettlingtimeinthelakeandthelargesurfaceareaofthelakeprovidesanadditionallosspathwayforthesecontaminantsthroughvolatilization.Thus,detectedtPCBconcentrationsinLakeWashingtonrangedfrom36to415pg/Landfrom32to1,572pg/LfortPBDEs.WhendatacollectedfromLakeWashingtonandthetwoShipCanalstationsarecompared,thehighestdetectedtPCBconcentrationswereattheBallardLocksstation(upto583pg/L)indicatinginputsintoLakeUnion/ShipCanalrepresentasubstantialcontributionoftPCBs.tPBDEconcentrationsinLakesWashingtonandUnionandtheShipCanalwereoftentooclosetomethodblankconcentrationstoreliablymeasure(i.e.,theywerenotdetected);however,thethreehighestmeasurementswerealsodetectedattheBallardLocksandMontlakeCutlocations(upto2,148pg/L).Comparedtootherinputpathwaysandthelakeitself,tPCBandtPBDEconcentrationsintheCedarandSammamishRiverswerebothrelativelylow.ThetPCBsintheCedarandSammamishriversrangedfrom10to267pg/L,whiletPBDEsrangedfrom3to3,150pg/L.TheselevelsaresimilartotPCBandtPBDEconcentrationsmeasuredbyHerreraEnvironmentalConsultants(2011a)inforested,agricultural,andresidentialtributariestotheSnohomishandPuyallupRivers,whicharealsoinWesternWashington.Comparedtothecurrentstudy,tPCBswerehigherduringstormeventsinthecommercial/industrialtributariessampledbyHerrera(2011a);however,thiswasnotthecasefortPBDEs.

tPCBandtPBDEconcentrationsmeasuredinmajorcreeksdifferedbylocation,withThorntonCreekexhibitingthehighestconcentrations,upto10,527pg/LtPCBsand20,910pg/LtPBDEsduringstorms.Asexpected,tPCBandtPBDEconcentrationswerelowerincreeksduringbaseflowconditions(minimaof105and59pg/L,respectively).tPCBandtPBDEconcentrationsvariedwidelyinstormwaterandhighwayrunoff.ThehighesttPCBconcentrationwasobservedintheFremontstormwater(165,685pg/L),whilethesecondhighestconcentrationwasdetectedinthehighwaybridgerunoff(16,133pg/L).tPBDEconcentrationsinstormwaterrunoffweresimilarbetweenSeattledischarges(Madrona,SewardParkandFremontDrainages),whilesmallercitydischarges(Kirkland,MercerIslandandRenton)exhibitedmorevariability,butweregenerallylowerthanthoseinSeattle.

ThehighesttPCBandtPBDEconcentrationsdetectedatCSOlocationswerealwaysobservedattheDexterCSO(upto565,108and212,174pg/L,respectively).ForbothtPCBsandtPBDEs,theSewardParkCSOhadthelowestreportedconcentrations(minimaof2,301and6,703pg/LfortPCBsandtPBDEs,respectively),whileconcentrationsattheBallard150CSOwereintermediate.TheaveragetPBDEconcentrationforallCSOswas82,898pg/L(N=8),whiletheaverageCSOtPCBconcentration(101,426pg/L,N=8)wasanorderofmagnitudehigherthantheaveragetPCBconcentrationofanyotheraqueousinputpathway.ThehighvariabilityofCSOtPCBconcentrationsgeneratedsignificantuncertaintyaroundthisarithmeticaverageconcentration.BecauseanarithmeticaverageCSOtPCBconcentrationwasrequiredformodelingandloadingsestimates,additionaldata(45samples)collectedfromCSOsintheLowerDuwamishRiverBasinWaterwaywereincludedtoenhancethereliabilityoftheestimatedconcentration.The53samplecombined



datasethadamorereliablearithmeticaverageof70,543pg/L,lowerthanthisproject’sCSOaverage,butstillmuchhigherthananyothermeasuredinputpathway’saverageconcentration.Figures2and3belowsummarizetheaveragetPCBandtPBDEwaterconcentrationsfoundinthefieldcomponentoftheproject.

PCBs pg/L

0

2.0x

104

4.0x

104

6.0x

104

8.0x

104

105

3.0x

105

AmbientStormwaterCSOs

Streams Baseflow

Rivers

Streams Storms

CSO

Stormwater Bridges

Stormwater All

Stormwater Seattle

Stormwater Small Cities

n=8n=4

n=25

n=10

n=11

n=3

n=16

n=9

Figure 2. tPCB average and standard deviation concentrations in major water pathways. Stormwater All is the average and standard deviation of Seattle, small city, and bridge runoff measurements.



Figure 3. tPBDE average and standard deviation concentrations in major water pathways. Stormwater All is the average and standard deviation of Seattle, small city and bridge runoff measurements.

Ofthetwoairdepositionsamplinglocations,airdepositionratesforBeaconHillwereconsistentlyhigherthanSandPoint.However,thedifferenceintPCBdepositionratesbetweensitesisgenerallywithinafactoroftwo.tPBDEdepositionrateswereaboutfourtimeshigheratBeaconHillthanSandPoint.AveragetPCBdepositionrateswere3.4ng/m2/d,whileaveragetPBDEdepositionrateswere18.1ng/m2/d.

Thisproject’sdataprovidethefirstextensivemeasurementsoflow‐leveltPCBandtPBDEconcentrationsinwholewaterfromLakeWashingtonandtheLakeUnion/ShipCanal,aswellasfromtheriver,creek,stormwater,bulkdeposition,CSO,andhighwayrunoffpathwaystoLakeWashington.Thesedatawereusedintheloadingsanalysisandmodelingphasesoftheproject(KingCounty2013b,c),whicharesummarizedbelow.

2.2 tPCB Loadings Estimates Followingcompletionofthefieldstudy,contaminantloadingpathwaystoLakesWashingtonandUnionweredevelopedbycombiningcontaminantconcentrationdatawithlong‐termflowestimates.tPCBandtPBDEmassloadingestimatestoLakeWashingtonandLakeUnion/ShipCanalaswellasthecontaminantexporttoPugetSoundweredeveloped.DetailsofthetPCBandtPBDEloadingcalculationapproachandresultingestimatesarepresentedinapreviousreport(KingCounty2013b).tPCBfindingsarehighlightedbelowastheprojectusedtheseloadingestimatestodevelopfateandbioaccumulationmodels.



Approximately70percentofthetPCBloadtoLakeWashingtoncomesfromlocaltributarywatershedsaroundthelakeandtheirstormwaterrunoff.Thethreesampledcreeks(Thornton,JuanitaandMay)representarangeinthetypeandintensityofdevelopmentandlanduse.Amongthesecreeks,theThorntonCreek’sbasinhadthehighestamountofcommercial/industrialdevelopmentthatoccurredpriortothebanonPCBmanufactureanduselimitationsinthelate1970s.ThisbasinhadthehighestestimatedtPCBmeanloadingof56g/yr(2.0g/km2/yr).TheMayCreekbasinhadthelowestamountofpre‐1979commercial/industrialdevelopmentandalsothelowestestimatedtPCBmeanloadingof23g/yr(0.66g/km2/yr).TheJuanitaCreekbasinhadanintermediateamountofpre‐1979commercial/industrialdevelopmentandanestimatedtPCBmeanloadingof17g/yr(0.93g/km2/yr),whichfallsbetweenthearealestimatesforThorntonandMaycreeks.Theseloadingestimatesandadditionalcorrelationanalyses(KingCounty2013b)suggestthatthepredominantsourceoftPCBstoLakeWashingtonisstormwaterrunofffromdevelopedareas–possiblylinkedtopre‐1979commercial/industrialdevelopment.TheseresultsareconsistentwiththeconceptualmodeloftPCBsourcesandpathwaysemergingfromotherstudies(Beltonetal.2007,Diamondetal.2010,RodenburgandMeng2013)thatsuggestPCBsourcesareconcentratedinurbancenterscontainingoldercommercialandindustrialbuildingswheretransformers,lightballasts,paints,caulks,andsealantswereused.Therefore,extrapolationofthesampleddrainagestotributarieswithoutconcentrationorflowmonitoringwasdonebasedontheirpercentageofpre‐1979industrial/commerciallanduseandrainfall.Thisproducedanestimatedtotallocaldrainageloadof450g/yr(1.2g/km2/yr)(KingCounty2013b).

LoadingestimatesforthetwomajorriverstoLakeWashington(CedarandSammamish)suggesttPCBloadsof56and41g/yr,respectively(97g/yrcombined).Thesearesomewhatlowerthanthetributaryload,althoughthereisahighdegreeofuncertaintyintheseestimatesduetothelowreportedtPCBconcentrationsinrivers;concentrationswereclosetothosedetectedinmethodblanks.ThecontributionofatmosphericdepositiontothesurfaceofLakeWashingtonwasalsoestimatedtoberelativelysignificantfortPCBs(110g/yr)representing14percentofthetotalloadtothelake.

Despitehavingthehighestmeasuredconcentrations,CSOsandhighwaybridgerunoffcontributedrelativelylittletotheoveralltPCBloadstoLakeWashington,12and2.9g/yrrespectively.Togetherthesepathwayscontributeapproximately2.2percentofthetPCBannualloadtoLakeWashington(14.9g/yr).Theseloadingswererelativelysmall,despitetheirhighconcentrations,duetotheirsmallcontributingvolumes.

EstimatedaveragetPCBloadingof672g/yrtoLakeWashingtonisgreaterthantheestimatedloadexportedfromthelakeoutlet(140g/yr).ThisisbecausethelakeactsasapartialsinkforPCBs,primarilyastheresultofsedimentaccumulationandburial.PCBsarealsolostfromthelakesurfacebyvolatilizationbackintotheatmosphere.

tPCBconcentrations,andhenceloading,increasefromtheoutletofLakeWashingtontotheoutletofLakeUnion/ShipCanaltoPugetSound.ThisispartiallyduetoCSOswhichdischargealargermassofPCBstoLakeUnionandtheShipCanal(58g/yr)thantoLakeWashington(12g/yr).TheaveragetPCBloadingestimatefromLakeUnionandtheShipCanaltoPugetSoundwas360g/yr.



The360g/yrloadingratefromtheShipCanal’sHiramM.ChittendenLockstoPugetSoundrepresentsabout4.5percentofthemediantotalloadofPCBstotheentirePugetSoundassummarizedbyNortonetal.(2011).GriesandSloan(2009)estimatedtPCBloadstotheDuwamishRiverfromtheGreenRiverusingsevensuspendedparticulatesamples.Theyestimatedthat0.6to1.21g/daytPCBsentertheDuwamishRiverfromtheGreenRiver(attheconfluencewiththeBlackRiverinTukwila).Theirflowratinganalysissuggeststhiswouldresultinthedischargeof153.3to2,263g/yrtPCBsfromtheGreenRiverintotheDuwamishWaterway.Theseloadingsarewithinasimilarrangeandboundthoseestimatedinthisstudy.However,GriesandSloan’s(2009)studymayunderestimateloadingsbyexcludingthedissolvedfractionofPCBs.Incomparison,theLakeWashingtonprojectbasedloadingsestimatesonwholewaterresultsattheHiramM.ChittendenLocks.Aswithmanyfieldefforts,loadingestimatesfrombothstudiesarenotdirectlycomparable(e.g.,differentmethodsofestimatingannualflow)andhavesubstantialassociateduncertaintiessuchaslimitedsamplesizes.

Onanareabasis,theGreaterLakeWashington(Cedar‐Sammamish)watershedis1,572km2;therefore,the360g/yrloadingratetoPugetSoundequalsanarea‐normalizedloadingrateof6.27x10‐7kg/km2/day.Thisarealoadingrateisnearlyanorderofmagnitudehigherthanthe7.39x10‐8kg/km2/daymaximumratecalculatedbyGriesandOsterberg(2011)fromfiveotherPugetSoundwatersheds:theSkagit,Snohomish,Nooksack,Stillaguamish,andPuyallup.However,thesewatershedshaverelativelylesspre‐1979development,havelessurbanareaoverall,andhavemoreagriculturalandforestlandthantheGreaterLakeWashingtonwatershed.Approximately600,000peopleliveintheGreaterLakeWashingtonwatershed(USCensus2010),andPCBsourcesarepartoftheoverallolderinfrastructuresupportingthesepeople‐nearly10percentofthestatewidepopulation.Therefore,whiletheestimatedloadingratefromtheLakeWashingtonwatershedisthehighestofthesixwatersheds,thisisplausible.

ThetPCBloadingestimatesdocumentedinKingCounty’sloadingsreport(2013b)werethefirstestimatesofPCBloadingsinthismajormetropolitanwatershed.Asummaryofthemeanloadsbypathway,alongwiththeir25thand75thpercentilevalues,isprovidedinTable1.TheseestimateswereusedinthefateandbioaccumulationmodelsdiscussedbelowtosimulatetheresponseofLakeWashingtonfishtissueconcentrationstoreductionsintPCBloading.



Table 1. Average, 25th and 75th percentile tPCB loading rates for pathways to Lake Washington, Lake Union/Ship Canal, and to Puget Sound (g/yr).

Pathway 25th

Percentile Mean

75th Percentile

Local drainages and rivers to Lake Washington 190 450 620

Highway runoff to Lake Washington 1.7 2.9 4.1

CSOs to Lake Washington 4.9 12 12

Direct atmospheric deposition to Lake Washington 84a 110 140a

Totals to Lake Washington 333 672 889

Lake Washington to Lake Union/Ship Canal 73 140 140

Local drainages to Lake Union/Ship Canal 22 40 53

Highway runoff to Lake Union/Ship Canal 0.52 0.87 1.2

CSOs to Lake Union/Ship Canal 23 58 59

Direct atmospheric deposition to Lake Union/Ship Canal 3.6a 4.8 6.0a

Lake Union/Ship Canal to Puget Sound 190 360 540

aMinimumandmaximumreportedbecauseonly2locationsweresampled.

2.3 tPBDE Loadings Estimates SimilartothetPCBloadsdescribedabove,tPBDEmassloadingestimatestoLakeWashingtonandLakeUnion/ShipCanalweredevelopedaswellastheloadexportedtoPugetSound.tPBDEloadingswere,ingeneral,moreuncertainthantPCBloadingestimates.ThiswasduetotheanalyticalchallengesinmeasuringlowlevelPBDEsinriversandLakeWashington.Theestimatesarelikelybiasedlow,becauseonlynineofthemostcommonPBDEswerequantifiedbythechosenmethod.Despitetheanalyticalchallengesandthemorelimitedanalytelist,overallPBDEloadingstoLakeWashingtonweremorethandoubletPCBloadings.Table2illustratesthatthe25thto75thpercentilerangesofPBDEloadingestimatesspananorderofmagnitudeformostpathways,whereasfortPCBs(Table1)the25thand75thpercentileloadingsspannedlessanorderofmagnitudeforallpathways.

AswithtPCBs,localdrainagescomprisedthemajorityofthetPBDEloadtoLakeWashington(820g/yr).Atmosphericdepositiontothelakesurfacealsocontributedarelativelylargemass.AsisthecasefortPCBs,aportionofthetPBDEloadingsaresequesteredinsedimentsorvolatilizefromthesurfaceofLakeWashington.Ofthe2,023g/yrtPBDEloadtothelake,only800g/yrexitsattheMontlakeCuttoLakeUnionandtheShipCanal.ThisisaugmentedbyCSOs,localdrainagesandatmosphericdepositionto990g/yrtPBDEsthatareexportedtoPugetSoundviatheHiramH.ChittendenLocks.



Table 2. Average, 25th and 75th percentile tPBDE loading rates for pathways to Lake Washington, Lake Union/Ship Canal, and to Puget Sound (g/yr).

Pathway 25th

Percentile Mean

75th Percentile

Local drainages and rivers to Lake Washington 200 820 1,200

Highway runoff to Lake Washington 1.3 19 19

CSOs to Lake Washington 3.3 14 26

Direct atmospheric deposition to Lake Washington 190a 590 980a

Totals to Lake Washington 416 2,023 2,755

Lake Washington to Lake Union/Ship Canal 330 800 940

Local drainages to Lake Union/Ship Canal 28 69 95

Highway runoff to Lake Union/Ship Canal 0.39 5.6 5.8

CSOs to Lake Union/Ship Canal 16 68 120

Direct atmospheric deposition to Lake Union/Ship Canal 8.3a 25 42a

Lake Union/Ship Canal to Puget Sound 280 990 1,400

aMinimumandmaximumreportedbecauseonly2locationsweresampled.

2.4 Lake Washington PCB Fate model ThefatemodelwasoriginallydevelopedbyGobasetal.(1995)andhasbeenappliedtoLakeOntario,SanFranciscoBayandPugetSound(Gobasetal.1995,Davis2004,PelletierandMohamedali2009).Itincorporatesmultipleprocessesincludingpartitioningtoparticulatesandorganiccarbon,LakeWashington‐specificsedimentationrates,andvolatilizationfromthelakesurface.Boththefateandbioaccumulationmodelsusethephysical/chemicalparametersofPCB‐118asasurrogatefortherangeofPCBchemicalpropertiesacrossallcongeners.ThechemicalcharacteristicsofothercongenerswithintherangeofthePCBsmostcommonlyfoundinLakeWashingtonfishtissuewereexaminedinthesensitivitysectionsofthefatemodel(KingCounty2013c).

ThefateandbioaccumulationmodelsforLakeWashingtonweredevelopedtoreliablyforecasttherecoveryoffishfromPCBcontaminationunderhypotheticalmanagementscenarios.Thefirstmodelingobjectivewastodevelopaquantitativeunderstandingofthelong‐termfateofPCBsinLakeWashington.Thesecondobjectivewastoprovidequantitativeestimatesofthetimeandmagnitudeoftheresponseoflakewater,sedimentandfishtissuetoreductionsinPCBloading.Thesetwomodelingobjectivesaddressoverallprojectstudyquestions2and3describedinSection1.0above.

Althoughthefatemodelisasimple,two‐compartmentboxmodel,itperformedwellwhentestedagainstobservedwaterandsedimentconcentrations.Thepredictedequilibrium



watertPCBconcentration(95pg/L)closelymatchesthe92pg/Lvolume‐weightedbestestimateofcurrentconcentrationsinLakeWashington.ThepredictedequilibriumsedimenttPCBconcentration(18µg/kgdry)wasapproximatelyone‐thirdthemeanobservedconcentrationbasedonobserveddata(55µg/kgdry,N=69).However,themeanobservedsedimenttPCBconcentrationissuspectedtobebiasedhighduetonon‐randomstudydesigns(KingCounty2013c).Manysedimentsampleswerecollectedfromgreaterthan3cmdeep(e.g.,0‐10cmgrabs),whichreflectolder,morecontaminatedsedimentasdescribedinUSGSandEcology(Ecology2010)sedimentcores.Thus,themodel‐predictedsedimentconcentrationmaybeclosertothetruemeansedimentconcentrationthanindicatedbycomparisonstoavailabledata.Thefatemodelestimatesabout98percentofthetotaltPCBmassresidesinthesurfacesedimentcompartment;examiningallcompartments,biotaareasmall(~4percent)reservoirforPCBs(KingCounty2013c).

TheLakeWashingtonimportandexportloadingsestimatesindicatethatLakeWashingtonisapartialsinkforPCBs;theexportloadissmallerthanthetotalimportload.Thefatemodeldemonstratedthatsedimentburial(44%)isthedominantlosspathwayofPCBsfromLakeWashington,followedbyvolatilization(24%).Lossbyoutflow(16%)anddegradation(9%)arelessimportantpathways.TheremainingsevenpercentoftPCBmassremainsinLakeWashington’swater.

Accordingtomodeltesting,theresponsetimeforLakeWashingtonsedimentandwaterconcentrationstoreachequilibriumunderaconstantloadingisapproximately40years;however,themostrapidchangesoccurforwaterandwithinthefirst20years.Sensitivityanalysisofthefatemodelfoundittobeverysensitivetothelogarithmoftheoctanol‐waterpartitioncoefficient(logKow)2andPCBloads.Whentheseinputswerevariedintheuncertaintyanalysisindependentlyandtogether,itwasdeterminedthattPCBloadingestimatescontributedmuchmoreuncertaintythanlogKow.However,fatemodeluncertaintyremainedwithintherangeofestimateduncertaintyinobservedwaterandsedimentconcentrations.Therefore,uncertaintyinthefatemodel‐predictedwaterandsedimentconcentrationsislessthanuncertaintyassociatedwithobservedwaterandsedimentdata.

2.5 Lake Washington PCB Bioaccumulation model Thebioaccumulationmodel(KingCounty2013c)wasoriginallydevelopedbyGobasetal.(1995),ArnotandGobas(2004),andadaptedforuseinSanFranciscoBay(GobasandArnot2005),andtheStraitofGeorgia(Condon2007).FurtheradaptationsallowedmodelingcontaminantbioaccumulationinselectPugetSoundbiota(PelletierandMohamedali2009).BoththefateandbioaccumulationmodelsselectedforLakeWashingtonhavedemonstratedhistoriesofutilityandsuccess.

BioaccumulationmodeltestingusedbestestimatesofwaterandsedimenttPCBconcentrationsfromobserveddataaswellasthewaterandsedimentconcentrationswhichweregeneratedbythefatemodelwith672g/yrtPCBloadtoLakeWashington.

2Theoctanol‐waterpartitioningcoefficientisaunitlessmeasureofachemical’shydrophobicity.



Bioaccumulationmodeloutputtissueconcentrationswerethencomparedtoobservedtissueconcentrations.Thebioaccumulationmodel‐predictedtissueconcentrationsmatchedobservedconcentrationsbetterwhenthefatemodeloutputsedimentconcentrationwasusedasaninput(18µg/Kgdry)ratherthanobservedsedimentdata(55µg/Kgdry).

TestingindicatedthemodelperformedadequatelyandsimilarlytoapplicationsofthismodelinPugetSound,GeorgiaBasinandSanFranciscoBay(GobasandArnot2005,Condon2007,PelletierandMohamedali2009)withabias3of1.2,i.e.,themodelover‐predictsLakeWashingtontissueconcentrationsby20percent(KingCounty2013c).Similartothefatemodel,thebioaccumulationmodelishighlysensitivetologKow.ThismeansthatthefishtissueconcentrationspredictedbythebioaccumulationmodelcanchangesignificantlywiththevalueselectedforthelogKow.However,uncertaintyanalysisindicatedthatvariabilityintheobservedtissuedatawasgreaterthanmodeluncertaintyforseveralfishspeciesincludingsmallmouthbass,cutthroattroutandnorthernpikeminnow.Theseuncertaintyresultsindicatethereismoreuncertaintyintheobservedfishtissuedatathanthemodel‐predictedtissueconcentrationsforthesespecies.Forlargeyellowperch,uncertaintyinthemodel‐predictedconcentrationswasgreaterthantheobservedtissuedata.

Althoughsedimentconcentrationsareordersofmagnitudehigherthanwaterconcentrations,lowertrophiclevelbiotaassociatedwithwater(e.g.,daphnia,copepods)playanimportantroleinthebioaccumulationofPCBsinuppertrophiclevelfishofLakeWashington.Themodeledspeciesbestmatchingobservedtissuedatawerenorthernpikeminnowandsmallmouthbass,twospeciesofhighconcernintheWADOH(2004)consumptionadvisory.Modeledlargeyellowperchconcentrationstendedtooverestimatetheobservedtissuedata.Thecauseofthisdifferenceisunknown;however,yellowperchdietsshift,becomingmorepiscivorouswithage.ToestimateyellowperchfishbodyburdenmoreaccuratelywouldrequireadditionaldataonLakeWashingtonyellowperchdietsthroughouttheirdevelopmentandseasonally.

2.6 Load Reduction Tissue Recovery Scenarios Inthefinalprojectphase,thefateandbioaccumulationmodelswerecoupledtotesttPCBloadingreductionscenariosandinformwaterqualitymanagers,stakeholdersandothersofthemagnitudeofchangeneededtoreachPCBfishfilletconcentrationsthataresafeforhumanconsumption(definedastheWADOHunrestrictedconsumptionscreeninglevelforPCBs).Theloadreductionscenariosdidnotincludeactivecleanupsofexistingcontaminatedsediments(e.g.,dredging,capping)ortheexclusivereductionofanyparticularpathway/source.

Themethodsandresultsofthesescenariomodelrunsarepresentedbelow.TohelpunderstandhowpotentialloadingreductionswillreducefishtissuetPCBconcentrationsin

3Modelbiasisthegeometricmeanofthedifferencesbetweenobservedandpredictedtissueconcentrationsforeachmodeledspeciesortaxon.Modelbiasabove1.0indicatesthemodelover‐predictstissueconcentrations,whilebiasbelow1.0indicatesunder‐prediction.



LakeWashingtonresidentfish,thecurrentestimatedtPCBloadof672g/yrwasreducedby10,25and50percentandeachreducedtotalloadwasusedasinputtorunthefatemodel.Theresultingpredictedsteady‐statesedimentandwaterconcentrations(Table1)werethenusedasinputstothebioaccumulationmodel.ThepredictedwholefishtissueconcentrationswereapproximatelyproportionaltothecorrespondingpercentreductionintotaltPCBload(Figure2).

Table 3. Predicted steady state tPCB concentrations in Lake Washington water and sediment under selected load reduction scenarios.

tPCB Loading Reduction Scenario

Water Sediment

(pg/L) (µg/kg dw)

Base case (load = 0.672 kg yr-1/logKow = 6.65) 95 18

10% tPCB load reduction 85 17



Observed 25th and 75th-percentile concentrations 51 - 118 11 - 53

Figure 4. Predicted whole fish tissue concentrations with PCB load reductions.

Thefatemodel‐predictedresponsetimetochangesinthetotaltPCBloadscenarios,suggeststhattheresponseofwaterconcentrationsisrelativelyquick,perhapsayearortwo(Figure5).However,theoverallsystemresponseislimitedbythesediments,which



requireabout20yearsforthemajorityofchangetooccurandabout40yearstoreachsteady‐stateequilibriumwiththereducedtPCBload(Figure6).Relativetosediment,thetypicalresponsetimeoforganismstochangesinPCBloadsisveryshort(days)(GobasandArnot2010).

Figure 5. Predicted response of water column tPCB concentrations to 10, 25, 50, and 85

percent tPCB load reductions.

Figure 6. Predicted response of tPCB sediment concentrations to 10, 25, 50, and 85 percent

tPCB load reductions.

ToestimatethemagnitudeofchangeneededtoreachsafetPCBlevelsinfish,thewaterandsedimentconcentrationsusedinthebioaccumulationmodelweredecreasedby10,25,50,80and90percent,andtheresultingpredictedwholebodytissueconcentrationswere

0

20

40

60

80

100

120

140

0 20 40 60 80 100

tPCB (pg/L)

Time (years)

10% reduction

25% reduction

50% reduction

85% reduction

0

10

20

30

40

50

0 20 40 60 80 100

tPCB µg/kg dw)

Time (years)

10% reduction

25% reduction

50% reduction

85% reduction



convertedtofilletconcentrationsforspeciesincludedinthecurrentLakeWashingtonfishconsumptionadvisory:cutthroattrout,largeyellowperch,northernpikeminnowaswellassmallmouthbass.FilletconcentrationswerethencomparedtoWADOHhumanhealthscreeningconcentrations(McBrideunpublishedmemorandum).WADOHdevelopedfillettPCBscreeningconcentrationstoevaluatefishtissuedataforpossiblerestrictionsonhumanconsumption.ThesescreeningconcentrationsaccountforlossofPCBsbycookingandarebasedonmeanconcentrationsinskin‐onfillets(Table3).Thepredictedtissueconcentrationsfromthebioaccumulationmodelarebasedonwholefish.Thus,species‐specificwhole‐to‐filletratiosbasedonDOH(2004)wereusedtoconvertthepredictedwholetissuetofillettissueconcentrations.4

Table 4. WADOH fillet screening concentrations for tPCBs.

Consumption Restriction Screening PCB Concentrations in

Fillet (ug/kg wet)

No Restriction (“safe”) <46

Possible Restriction 46-380

Do Not Eat >380

ThecomparisonofpredictedfillettissueconcentrationswithWADOHscreeninglevelsindicatesthatsubstantial(approximately85%)reductionsinsedimentandwaterconcentrationsarerequiredtobringtheaveragefishfilletconcentrationsofallfourspeciestothe“NoRestriction”level(<46ug/kgww,Figure7).ThepredictedtissueconcentrationinFigure7isbasedonthefatemodel‐predictedsedimentandwaterconcentrationsunderazeroreductionassumption.Errorbarswereaddedontheobservedandpredictedmeanforcomparisonandshowvariabilitycanexistforsomespeciesinbothactualtissuedataandthatpredictedbythebioaccumulationmodel.Forexample,thestandarddeviationonthemeannorthernpikeminnowfilletconcentrationisordersofmagnitudelargerthantheerroronthepredictedmeanconcentration(tissueconcentrationsbasedonlowor25thpercentileandhighor75thpercentileestimatesoftPCBloads,seenoteinFigure5).

Itshouldbenotedthatthisanalysisisonlyusefulasageneralindicationoftherequiredmagnitudeofsedimentandwaterconcentrationsreductionsnecessary,becauseinadditiontouncertaintiesassociatedwiththebioaccumulationmodelandobservedtissuedata,therearealsouncertaintiesinthewhole‐to‐filletratiosandWADOHfishtissuescreeninglevels.Thereisadditionaluncertaintyassociatedwithpredictedconcentrationsinlargeyellowperchduetotheiroverallpoorermodelfit,althoughtheircurrenttPCBconcentrationsarenotashighasotherspecies.Nevertheless,thisanalysisdemonstratesthatverylargereductionsintPCBloadsareneededtobringLakeWashingtoncutthroattrout,northernpikeminnowandsmallmouthbasstissueconcentrationsdowntolevelsthataresafeforhumanconsumption.

4Whole‐to‐filletratiosof0.5wereusedforsmallmouthbass,yellowperchandnorthernpikeminnowand0.68forcutthroattrout.



Figure 7. Predicted fillet concentrations of tPCBs under 10, 25, 50, 80, and 90 percent loading

reduction scenarios compared to WADOH human health fillet screening concentrations.



3.0. NEXT STEPS AND RECOMMENDED ACTIONS

Attheoutsetofthisproject,anadvisorypanelconsistingofrepresentativesfromseveraljurisdictionsandagencieswasestablishedtoprovideinputonprojectdesignandexecution,aswellasbrainstormfutureactionsandrecommendations.Panelmemberswere:

FredBergdolt,WADepartmentofTransportation

BetsyCooper,KingCountyWastewaterTreatmentDivision

JonathanFrodge,SeattlePublicUtilities

JennyGaus,CityofKirkland

JoanHardy,WADepartmentofHealth

RachelMcCrea,WADepartmentofEcology

DougNavetski,KingCountyWaterandLandResources

AndyRheaume,CityofRedmond

RonaldStraka,CityofRenton

HeatherTrim,PeopleforPugetSound/FutureWise

BruceTiffany,KingCountyWastewaterTreatmentDivision

PatrickYamashita,CityofMercerIsland

Throughoutthethree‐yearproject,advisorypanelmembersreviewedandcritiquedthefieldstudydesign,modelingframeworkandassumptions,andallprojectdeliverables.Projectteamandadvisorypanelmembersalsodiscussedthescenarioresultsdescribingthe~85percentreductionintPCBloadingsrequiredtoachievefishtissueconcentrationsthatarebelowtheWADOH“No‐Restriction”screeninglevel.Aspartofthisdiscussion,theadvisorypanelprovidedrecommendationsfornextstepsandactions,focusingontPCBs.Thefollowingsectionsdescribethecollectiverecommendationsofthepanel.ManyoftheseideasaddresssomeformofPCBsourcecontrol,althoughtheymaynotmeetthedefinitionsofsourcecontrolusedinpotentiallyapplicablelawsandregulationssuchastheModelToxicsControlAct(Chapter70.105DRCW).SourcecontrolisseenastheultimatesolutiontoreducingPCBloadingstoLakeWashington.However,sourcecontrolisaslowprocess;therefore,additionalactionsarerecommendedthatcanalsoreducePCBloadingstotheLake.Advisorypanelmembers,otherjurisdictions,businesses,andgovernmentalandnon‐governmentalorganizationsareencouragedtoconsiderimplementingoneormoreofthesenextstepsthroughouttheregion.



3.1 Update the Conceptual Model of PCBs in the Local Environment

PerhapsoneofthemostfundamentalfindingsofthisstudyisthepervasiveandactivenatureofPCBsourcesthroughoutthewatershed.Localtributariestothelakeareresponsibleforapproximately70percentoftheestimated672gannualloadingtothelake(333to889g/yrestimatedrange).InurbanareaslikeSeattleandthesurroundingmetropolitanarea,PCBsarenotjustlegacycontaminantspresentinsediment.TheemergingconceptualmodelforLakeWashingtonissimilartothosedevelopinginotherurbanareassuchasSanFranciscoBay(Tsaietal.2002),Camden,NJ(Beltonetal.2007),Chicago(RodenburgandMeng2013)andToronto,Canada(Robsonetal.2010).Thesemodelspostulate,investigate,anddescribeanurbanplumeordomeofPCBsinandaroundcities;conceptuallysimilartothepollutionplumesatmosphericscientistshavedescribedforheat,carbondioxide,particulates,andotherairpollutants(HuddartandStott2010).Urbanareasarelongknowntotrapheat,CO2andparticulatesbeneathaninversionlayerwhichmayelongatedownwinddependingonprevailingwinds(Munn1976).PCBsalsoformanurbanplumeofairpollutionastheyvolatilizefromcaulks(Robson,etal.2010),paints,andolderlightballasts,particularlythoseballastswhichfail(Guoetal.2011).ThesegaseousPCBscanthenprecipitateontoparticulatesorsurfacefilmsandgetpickedupinstormwaterrunoffordissolvedirectlyintorainorwaterbodies.PCBscanalsowashdirectlyintowaterbodiesfromflakingpaintorfromabradedjointcaulksusedinconcretesurfaces(Ecology2011,CityofTacoma2013).PCBs,whileubiquitousintheurbanenvironmentappeartobeconcentratedinolder(e.g.,pre‐1979)buildingsandinfrastructure.Previousstudiessuggestthatresidentialbuildingsandstructuresolderthan1945andnewerthan1980arelesscommonlyassociatedwithPCB‐containingbuildingmaterials(Robsonetal.2010,Ecology2011).

AlloftheseprocessesarepartoftheemergingunderstandingthatPCBsarenotsimplyhistoricalcontaminants,butinsteadhavenumerouson‐going,diffuseurbansources(Diamondetal.2010).ThesizeofthegreaterLakeWashingtonPCBplumeisunknown,butBrandenbergeretal.(2010)foundthaturbanPugetSoundPCBdepositionwasmorethanthreetimeshigherthaninruralareas.

Thisconceptualmodelofdiffusesourcesthroughoutthewatershed,asopposedtodistinctreleasesandpointemissions,isconsistentwithfindingsofpreviousattemptstoidentifyhotspotsofPCBcontaminationinLakeWashingtonsurfacesediments(Era‐Milleretal.2010).Todate,noPCBsurfacesedimenthotspotshavebeenfound.ThisprojectfocusedonPCBtransportpathways;futureinvestigationswillneedtoinvestigatethesepathwaystotraceandidentifyPCBsourcesincurrentandhistoricallyusedmaterials.

3.2 Source Control - Develop a PCB Inventory Thefirststepinsourcecontrolisidentifyingwheresourcesarelocated.Therefore,werecommendconductinganurbaninventoryofPCBstocks.Developingalistofcurrentlyusedproductsandin‐situsourcessuchaspaints,caulks,ballasts,andcontaminatedsiteswithhistoricalreleasestosoil,sedimentsand/orgroundwaterwouldemphasizeand



reinforcethecross‐basincollaborationnecessarybyregionalpartnerstoidentifythematerialscontributingPCBstotheregion’senvironmentalmedia.

Aspartofaninventory,arankingoftherelativemassandcontributionsfromdifferentsourcesisrecommended.Forinstance,Robsonetal.(2010)estimatedthat13metrictonsofPCBsremaininbuildingsealantsinToronto,acitywithapopulationof2.6million.TheGreaterSeattlemetropolitanareahas3.5millionresidents.WhilethisprojectisunabletoaccountfordifferencesinbuildingagebetweenTorontoandSeattle,scalingforpopulationsuggeststhattheSeattleregioncouldhave17+metrictonsofPCBsincurrentuseassealantsandcaulksalone.Ecology(2011)estimated850g(0.00085metrictons)ofPCBswerecurrentlyinpaintaloneon2,286buildingsconstructedbetween1950and1977intheDiagonalAvenueSouthsub‐basinoftheDuwamishWaterway–anareaofabout9,000acreswithatotalof27,000buildings.

WhileconductingsourcetracingactivitiesintheLowerDuwamishWaterwayarea,Ecology(2011)developedapartialinventoryofPCB‐containingbuildingmaterials.However,gettingpropertyaccessandpermissiontosamplewaschallenging;only32ofthe92contactedpropertyownersagreedtocompositesamplecollection(materialscombinedfrommultiplebuildings).Ecology’s(2011)inventoryusedcompositesamplingtoallowindividualpropertyownerstoremainanonymous.ItislikelythatliabilityconcernsandthegenerallackofawarenessregardingthewidespreadnatureofPCBsourcesaresignificantbarrierstocreatingalocal,actionableinventory.Nevertheless,

3.2.1 Perform Regional Source Tracing Using Congener Data Narrowingthelistofbuildings,sites,andproductscontainingPCBstoasub‐basinlevelwouldalsoallowfordevelopmentandtestingofpilotprojectsinsomeofthemostcontaminatedareasand/oronthemorecontaminatedrunoff.Thisproject’sdatasetpresentsarichopportunityforresearcherstoanalyzePCBcongenerpatternsandpotentiallymatchthemtourbansources.EcologyrecentlyfundedaSpokaneRiverToxicsTaskForceprojecttoanalyzeforPCBsincommonlyusedmunicipalproducts,suchaspaints,caulksandsealants,andresultsofthiseffortmaybecombinedwitheffortstolinksourcestoloadsintheGreaterLakeWashingtonwaterandairsheds.EffortstobetterunderstandandidentifyPCBsourcesintheLakeWashingtonwaterandairshedswouldformthebasisforfuturesourcecontrolactions.

3.3 Conduct Outreach and Education InadditiontoupdatingtheconceptualmodelanddevelopinganactionableinventoryofPCBsources,anoutreachandeducationstrategyisneededtoengagedecision‐makersandthecommunityatlargeinadiscussionofthefinancialandregulatorychallengesfacingeffortstoreducePCBloadingstoLakeWashingtonandotherwaterbodies.ThesediscussionshavebegunintheLowerDuwamishandSpokaneRiverwatersheds,buttodatehaveyettooccurwithintheLakeWashingtonwatershed.

PCBspresentinbuildingmaterials,whetherassociatedwithcaulk,paints,orfromspillsoftransformersorcapacitors,areregulatedbyEPAandEcology(EPA2005,WAC173‐303‐9904).Thefederalregulatorythresholdformostproductsandwastesis50mg/kg;



transformerandcapacitorwaste(evenwhendrainedofPCBoils)isregulatedabove2mg/kgbyEcology(ModelToxicsControlActdangerouswastecodeWPCB,WAC173‐303‐9904).TheselevelsareordersofmagnitudehigherthanPCBlevelsofconcernintheenvironment.LakeWashingtonjurisdictionsareencouragedtosupporteffortstoreformfederalregulationsforPCB‐containingmaterialsremaininginuse.

Thedegreetowhichsuspectpaintorcaulksaretestedpriortobuildingdemolitionorrenovationisunknown.Paints,caulksorspilledPCB‐containingoilscanalsocontaminateunderlying/adjacentplaster,wood,masonry,andconcrete.EPA(undated)hasissuedguidanceforcontractors;however,thisisanemergingareaofwastemanagement.Knowledgeoftheserequirementsforbothprotectionofwaterqualityandworkerhealthisunknown.EnhancedscreeningofbuildingspotentiallycontainingPCB‐ladencaulksorpaintsisonestrategytofurtherlimitreleasesofPCBsfromthesematerials.GiventhereluctanceofLowerDuwamishareapropertyownerstoconsenttosamplingtheirbuildings’paintsandcaulks(Ecology2011),financialorregulatoryincentivestofacilitatetestingandsafePCBremovalduringdemolitionorrenovationsof1950‐1980eraindustrial,commercialandinstitutionalbuildingsandinfrastructuremayencouragewidertestingandsourcecontrol.Forinstance,buildingownersthatvoluntarilytesttheirbuildingmaterialsforinventorypurposescouldreceiveanin‐usegraceperiodtoallowthemtograduallyremovethepaint,caulk,etc.whilesuchmaterialsdiscoveredthroughotherinvestigations(e.g.,cityorcountystormwatersourcecontrolstudies)wouldrequiremoreurgentremoval.Thesetypesofregulatorychangesandapproachesrequirebroadercommunityandlegislativeeducationanddiscussions.

WhilethetimeframenecessaryforLakeWashingtonfishtissueconcentrationstosubstantivelydeclineundera“noaction”scenarioiscurrentlyunknown,underanyloadreductionscenarios,atleast20to40yearsarerequiredtoreachnewequilibrium.LakeWashingtonsurfacesedimentconcentrationshavedeclinedsubstantiallysincetheirpeakinthe1970s(Era‐Milleretal.2010)buttherateofdeclineasevidencedbysedimentcores,hasslowedoverthepast2decades.Basedontherelativelyconstantsurfacesedimentconcentrationsoverthepasttwodecades,currentsedimentandwaterconcentrationsarelikelyinroughequilibriumwithcurrentloadings.ThisplateauofsedimentconcentrationsindicatesthatthecurrentrateofPCBsourceremovalsuchasbuildingrenovationandtransformerreplacementresultsinaslowtonodeclineinoverallPCBloading.Continuingthestatusquowillresultinthecurrentfishadvisorypotentiallyremainingindefinitely.

Ecology’s“ChemicalActionPlan”(Ecology2013)forPCBsiscurrentlyinpreparationandpresentsanidealopportunitytodevelopaneducationandoutreachstrategyaimedatovercomingthefinancialandregulatorychallengesfacingeffortstoreducePCBloadings.

3.4 Evaluate Effectiveness of Stormwater Treatment and LID Methods

Todate,literaturereviewshavenotfoundanystudiestestingtheeffectivenessofvarioustechnologiesfortreatingPCB‐contaminatedstormwaterrunoff(Herrera2011b).SeveralpromisingtechnologieshavebeenandcontinuetobetestedintheLowerDuwamisharea(Kalmar2010,Schmoyer2012).Electro‐coagulationandchitosan‐enhancedsandfiltration



aretwoofthesemoreintensivestormwatertreatmenttechnologies.Theseexpensivetreatmenttechnologiesaretypicallyusedindiscretewastewaterandconstructionstormwatersettingstoaddressspecific,contaminatedsourcesordischarges.FundingsuchtechnologiesonthescalenecessarytoreduceLakeWashingtonPCBloadings~85percentwouldbeextremelychallengingandtodateitisunknowniftheyareeffectiveatreducingmodest(ng/Lorless)PCBconcentrationsinwater.ThecostofstormwaterrunofftreatmentneedstobeevaluatedatthebasinscaletoestimatethecostpermassofPCBsremoved.ThisinformationshouldthenbecomparedtotheeffectivenessofotherPCBremovaloptions(e.g.,paintremoval,conveyancepipecleaning,etc.)toidentifythebestoptionstomaximizetheenvironmentalbenefitoflimitedpublicfunds.

Highefficiency‐streetsweepingtocapturesmallerparticulatesisabasicsourcecontroltechniquethathasthepotentialtoreduceloadingsfromimpervioussurfaces.Similarly,catchbasinandstormwaterconveyancepipecleaningareotherpossibletools,whichmayreducestormwaterrunoffloadings,particularlyfromhistoricsources.BothrelyonbulksolidsremovaltopotentiallyremovePCBsthathaveadheredtosolidspresentonthestreetsurfaceandwithintheconveyancesystem.Theefficacyandcost‐benefitofthesestrategieshasalsoyettobedemonstratedforwidespreaduseinwesternWashingtonurbanwatershedsorforPCBs.Moreextensiveuseofrapidandrelativelylowcost(<$100/sample)immunoassayanalyticalmethods(EPA1996)toscreenPCB‐contaminatedmaterialsincludingstreetdirt/solidsisencouraged.

EPA,Ecology,andotherregulatoryagenciesaremovingtowardlowimpactdevelopment(LID)asapreferredstormwatermanagementtool(EPAetal.2007,EPA2013).However,recentreviewsofLIDtechnologies(Herrera2011b,TaylorandCardno2013)havebeenunabletodocumenttheeffectivenessofLIDinreducingPCBloads.Forinstance,eventhoughtechnologieslikepermeablepavementsmayreducestormwaterflowsthroughinfiltration,theirPCBcontaminantloadsmayremaintrappedonsurfaceparticulateswithsomeportionre‐volatilizinglaterintotheurbanatmosphereanddepositingelsewhere.AdditionalLIDdesignconsiderationssuchasincorporationofspecificsoilamendmentslikebiochar(charcoal)mayenhancePCBretentioninswales,bioretentionfacilitieslikeraingardens,orgreenroofs.Unfortunately,theefficacyofanyspecificamendmentordesignelementiscurrentlyunknown,asLIDstormwatermanagementpracticeshavenotbeenevaluatedforimpactsonurbanPCBcycling.GiventhesubstantialinterestinLIDtechnologyforstormwaterflowcontrol,volumereduction,andothercontaminantremoval(e.g.,dissolvedcopper),understandingtheefficacyanddesignrequirementsforPCBremovalisnecessaryandtimely.

3.5 Develop Airshed & Washoff Models Thecurrentdatasuggestthatbulk(wetanddryforms)aerialdepositionisasignificantcontributortostormwaterloads;however,thedegreetowhichbulkaerialdepositioncontributestoPCBstormwaterloadsfromdifferentlandsurfacessuchasgrass,pavement,orroofingisunknown.Furthermore,therelativecontributiontoLakeWashingtonairdepositionfromtheLowerDuwamishValley,anidentifiedregionalPCBhotspot,isunknown.WindpatternsarelikelytocarryPCBsvolatilizedfromtheLowerDuwamishwater,sediments,anduplandstowardsLakeWashington’swatershedandmaycontribute



totheregionalairdepositionthatinfluenceswatershedloadings.TheregionaleffectsoffuturecleanupsintheLowerDuwamishorelsewherehavenotbeenestimatednorquantified.

MostPCBcontrolstudieshavefocusedontracingprominentPCBsources(Beltonetal.2007,Bottsetal.2007,Cargill2008,KCIWandSPU2005,Schmoyer2007,KingCounty,2011).Thesetypesofinvestigationshavefocusedonin‐pipestormwatersystemsamplingandanalysistotraceandidentifyPCBcontributionsfromcommercialorindustrialsources.Ifonecanbedeveloped,aregionalrunofforwashoffmodelcouldbeusedtoprioritizelandusesorsub‐basinsformorerigoroussourcetracingsamplingandanalysiseffortssuchasthese.

Awashoffmodelwouldsimulatetherelationshipsbetweenbulkaerialdepositionandrunofffromdifferentsurfacesandlanduses.Awashoffmodelisnecessarytodefinetheroleofaerialdepositioninstormwatercontaminationandextendthepredictivecapabilityofthebulkdepositiondatatodifferentlandsurfacesordevelopmenttypes,whichwouldhelpfocussourcecontrolstudies.Forexample,stormwatersamplesexceedingexpectedconcentrationsfrombulkdepositionalonecouldbethefocusofadditionalsourcetracingefforts.Suchanair‐land‐watermodel,whetherstatisticalormechanistic,wouldbemultivariateasmultiplefactorssuchasparticulates,temperature,rainfallandwindspeedallcorrelatewithbulkdepositiondata(KingCounty2013d).FurtherunderstandingofPCBvolatilizationandcyclingfromurbansurfacesinWesternWashingtonorasimilarclimatewouldbenecessarytodevelopauseful,predictivewashoffmodel.Thedevelopmentofsimplerelationshipsbetweenbulkdepositionandstormwaterconcentrationssimilartotherelationshipstestedintheregionaldrainageextrapolations(KingCounty2013b)wouldbeagoodfirststeptowarddevelopingawashoffmodel.



4.0. CONCLUSIONS ThisprojectevaluatedsixmajorPCBandPBDEcontaminantpathwaystoLakesWashington,UnionandtheShipCanalinSeattle,Washingtontoestimateloadings.TheloadingswerelinkedwithafatemodelandabioaccumulationmodeltofurtherunderstandhowchangesinloadingsaffectfishtissueconcentrationsofPCBs.

ThemostsignificantloadtoLakeWashingtonwasfromlocaldrainagestormwaters(stormwaterdischargesandcreekstormflows),whichcontributeapproximately67percentoftheannualtPCBload.Directatmosphericdepositionandriverscontributeapproximately14percentoftheannualtPCBload.Despitehavingthehighestaveragemeasuredconcentrationsbyatleastanorderofmagnitude,theoverallloadsfromCSOsandhighwaybridgerunoffcombinedwerelessthanthreepercentoftheestimatedtotaltPCBloading.Combined,thesixpathwaysareestimatedtocontributeapproximately672g/yrofPCBsonaveragetoLakeWashington(25thpercentile‐333g/yr,75thpercentile‐889g/yr).LakeWashingtonout‐flowaccountsfor140g/yrofPCBsenteringLakeUnionwhereitisaugmentedto360g/yrbyCSOs,localstormwaterdrainages,andatmosphericdepositionbeforedischargingtoPugetSound.MuchoftheremainingloadisburiedinLakeWashingtonsedimentsorvolatilizedthroughthelakesurface.Thus,LakeWashingtonactsasbothasourceandasinkforPCBs.EstimatesoftPBDEloadingstoLakesWashington,UnionandPugetSoundwerecompleted,butconsideredmoreuncertainduetothechallengesposedbymeasuringPBDEsinambientwatersatlevelsclosetothoseinmethodblanks.TheaveragetotaltPBDEloadfromallassessedpathwaystoLakeWashingtonisestimatedtobe2,023g/yr(25thpercentile–416g/yr,75thpercentile–2,755g/yr).

LakeWashingtonbiotabioaccumulatePCBsresultinginthecurrentfishtissueconcentrationswhicharesomeofthemostcontaminatedinthestate(WADOH2004,JackandColton2011).Thebioaccumulationmodelindicatesthatasubstantialreduction(~85%)incurrentPCBloadsisnecessarytoreducePCBbodyburdensinresidentfishtosafelevels(belowpublichealthfishconsumptionadvisorylimits)andeliminatethecurrentfishconsumptionadvisory(WADOH2004).Thefateandbioaccumulationmodelsdevelopedforthisprojectpredictalarge~80‐85percentreductioninPCBloadingsisnecessarytoremovethecurrentfishconsumptionadvisoryfornorthernpikeminnow,cutthroattrout,andyellowperch.Althoughnotcurrentlysubjecttotheconsumptionadvisory,an85percentreductionshouldalsoreducePCBconcentrationsinsmallmouthbasstobelowWADOHscreeninglevels.Carpwerenotincludedinthemodelingeffort;additionaldatawouldbenecessarytodetermineifan85percentPCBloadreductionwouldresultintissueconcentrationssufficienttoremovethecarpconsumptionadvisory.Lakesedimentandwaterconcentrationswoulddeclinesubstantiallywithin20yearsofreducedloadings.However,steady‐stateconditionsbetweenloadings,water,andsedimentconcentrationswilltheoreticallytakeapproximately40yearstoachieve.While20to40yearsisalongtime,intheabsenceofsubstantialwatershed‐wideeffortstoreducetPCBloads,theexistingfishconsumptionadvisoryisprojectedtoremainindefinitely.

Theproject’sadvisorypaneldiscussedanumberofpotentialstrategiestohelptargetandeventuallyreduceoverallPCBloadings.SupportingWashingtonStateandregionalpartnersindevelopinganurbaninventoryofpotentialPCBsourcesisoneoftheproject’s



toprecommendations.Asecondrecommendationistodevelopanoutreachstrategywhichengagesdecisionmakerswhileactivelyeducatingthecommunity‐at‐largeaboutthefinancialandregulatorychallengesfacingeffortstoreducePCBloadingstoLakeWashingtonandotherwaterbodies.Also,anumberofjurisdictionsandstormwatermanagersarealreadyactivelypursuingLIDtechnologiestotreatormanagestormwatervolumes,nutrients,metals,andsuspendedsediment.Sincestormwater,asdirectdischargeorstormflowsincreeks,isthemostsignificantloadingpathway(estimatedtobeabout70percentofthetotalloadingtoLakeWashington),studyingandunderstandingtheefficacyofLIDtechnologiesandensuringtheyareeffectiveatPCBremovalisanimportantnextstep.Lastly,furtherinvestigationofPCBairdeposition,volatilization,cycling,andwashoffinWesternWashingtonurbanareasisrecommended,sincewetanddrydepositionarebothsuspectedtobeimportantcontributorstooverallstormwaterloads.

Inconclusion,acombinationofaggressivesourceidentification,removal,andstormwatertreatmentarerecommendedtoworktowardsachievingtheestimated~85percentloadreductionnecessarytolowerLakeWashingtonfishtissueconcentrationstosafelevels.



5.0. REFERENCES

Arnot,J.A.andF.A.P.C.Gobas.2004.Afoodwebbioaccumulationmodelfororganic

chemicalsinaquaticecosystems.Environ.Toxicol.Chem.23:2343‐2355.

Belton,T.,J.Botts,L.Lippincott,andE.Stevenson.2007.AnIndustrialEcologyApproachto

PCBSourceTrackdowninCamdenCity,NewJersey.InOptimizingContaminantTrackdown:FocusingonWastewaterTreatmentPlantsandRelatedSystems.NY

AcademyofSciences.

Botts,J.A.,J.Spadone,B.McKenna,A.Kricun,T.Belton,R.Hindt,andD.Dutton.2007.PCB

MinimizationPlansandSourceTrackdown.InOptimizingContaminantTrackdown:FocusingonWastewaterTreatmentPlantsandRelatedSystems.NYAcademyof

Sciences.

Brandenberger,J.M.,P.Louchouarn,L‐JKuo,E.A.Crecelius,V.Cullinan,G.A.Gill,C.Garland,J.Williamson,andR.Dhammapala.2010.ControlofToxicChemicalsinPugetSound,Phase3:StudyofAtmosphericDepositionofAirToxicstotheSurfaceofPugetSound.WashingtonDepartmentofEcology,OlympiaWA.

http://www.ecy.wa.gov/pubs/1002012.pdf

Cargill,D.2008.LowerDuwamishWaterwaySourceControlStatusReport‐April2008throughAugust2008.WashingtonStateDepartmentofEcology,Lacey,WA.Publicationnumber08‐09‐068.AccessedDecember19,2013at:https://fortress.wa.gov/ecy/publications/summarypages/0809068.html

CityofTacoma.2013.TheaFossSourceControlInvestigations.AccessedDecember52013at:http://www.cityoftacoma.org/cms/One.aspx?portalId=169&pageId=7283

Condon,C.D.2007.Development,evaluation,andapplicationofafoodweb

bioaccumulationmodelforPCBsintheStraitofGeorgia,BritishColumbia.Thesis:MasterofResourceManagement.SchoolofResourceandEnvironmental

Management.SimonFraserUniversity,Barnaby,B.C,Canada.

Cullon,D.L.,S.J.Jeffries,P.S.Ross.2005.Persistentorganicpollutantsinthedietofharbor

seals(Phocavitulina)inhabitingPugetSound,Washington(USA),andthestraitofGeorgia,BritishColumbia(Canada):AfoodbasketapproachEnvironmental

ToxicologyandChemistry.24(10):2562‐2572.



Davis,J.A.2004.Thelong‐termfateofpolychlorinatedbiphenylsinSanFranciscoBay

(USA).Environ.Toxicol.Chem.23:2396‐2409.

Diamond,M.L.,L.Melymuk,S.A.Csiszar,andM.Robson.2010.EstimationofPCBStocks,

EmissionsandUrbanFate:WillourPoliciesReduceConcentrationsandExposure?

Env.Sci.Tech.44:2777‐2783.

Ecology.2010.GeneralCharacterizationofPCBsinSouthLakeWashingtonSediments.

PreparedbyEra‐Miller,B.,R.Jack(KingCounty),andJ.Colton(KingCounty),

Olympia,Washington.AccessedJune24,2013https://test‐fortress.wa.gov/ecy/testpublications/SummaryPages/1003014.html

Ecology.2011.SurveyofPotentialPCB‐ContainingBuildingMaterialSourcesSummary

Report.PreparedbySAIC,Bothell,WA.AccessedDecember20,2013at:https://fortress.wa.gov/ecy/gsp/DocViewer.ashx?did=6496

Ecology.2013.http://www.ecy.wa.gov/programs/swfa/pbt/pcb.html

EPAundated.ContractorsHandlingPCBinCaulkDuringRenovation.EPAdocumentnumberEPA‐747‐F‐09‐004.AccessedDecember18,2013at:http://www.epa.gov/waste///hazard/tsd/pcbs/pubs/caulk/pdf/contrctrs.pdf

EPA,1995.EPARegion10SOPFortheValidationofMethod1668Toxic,Dioxin‐like,PCB

Data.AccessedJune12,2013http://yosemite.epa.gov/r10/OMP.NSF/b380f72ce20f16c188256f0100719be8/77c

4f4ee8e0e02a98825731a0069a5fd/$FILE/qar01668.pdf

EPA,1996.TestMethodsforEvaluatingSolidWaste,Physical/ChemicalMethods,SW‐846.

Method4020:ScreeningforPolychlorinatedBiphenylsbyImmunoassay.AccessedDecember20,2013at:http://www.epa.gov/osw/hazard/testmethods/sw846/pdfs/4020.pdf

EPA2005.ManagingYourEnvironmentalResponsibilities:APlanningGuidefor

ConstructionandDevelopment.PartI,SectionVIII.EPAdocumentnumber

EPA/305‐B‐04‐003.AccessedDecember18,2013at:

http://www.epa.gov/compliance/resources/publications/assistance/sectors/const

ructmyer/myerguide.pdf

EPA,NationalAssociateofCleanWaterAgencies,NaturalResourcesDefenseCouncil,Low

ImpactDevelopmentCenter,AssociationofStateandInterstateWaterPollution

ControlAdministrators.2007.GreenInfastructureStatementofIntent.Memorandum.4pgs.AccessedJanuary15,2014at:



http://water.epa.gov/infrastructure/greeninfrastructure/upload/gi_intentstatement.pdf

EPA2013.“MythBustersFactsheet”series.EPAdocumentnumber:EPA841‐N‐12‐003A

AccessedDecember2,2013at:http://water.epa.gov/polwaste/green/bbfs.cfm

Era‐Miller,B.,R.Jack,J.Colton.2010.GeneralCharacterizationofPCBsinSouthLake

WashingtonSediments.WashingtonDepartmentofEcology,Lacey,WA.Publication

No.10‐03‐014.AccessedDecember5,2013at:

https://fortress.wa.gov/ecy/publications/summarypages/1003014.html

Gobas,F.A.P.C,M.N.Z’GraggenandX.Zhang.1995.TimeresponseoftheLakeOntario

EcosystemtovirtualeliminationofPCBs.Environ.Sci.Technol.29:2038‐2046.

Gobas,F.A.P.CandJ.A.Arnot.2005.SanFranciscoBayPCBFoodWebBioaccumulationModel:FinalTechnicalReport.PreparedfortheCleanEstuaryPartnershipbyFrankGobas,PhDandJonArnot,MET,SimonFraserUniversity.

Gobas,F.A.P.CandJ.A.Arnot.2010.FoodwebbioaccumulationmodelforpolychlorinatedbiphenylsinSanFranciscoBay,California,USA.Environ.Toxicol.Chem.29:1385‐1395.

Gries,T.andD.Osterberg.2011.ControlofToxicChemicalsinPugetSound:

CharacterizationofToxicChemicalsinPugetSoundandMajorTributaries,2009‐10.WashingtonDepartmentofEcology,Lacey,WA.EcologyPublicationnumber11‐03‐008.AccessedJanuary13,2014at:https://fortress.wa.gov/ecy/publications/publications/1103008.pdf

Gries,T.andJ.Sloan.2009.ContaminantLoadingtotheLowerDuwamishWaterwayfrom

SuspendedSedimentintheGreenRiver.WashingtonDepartmentofEcology,Lacey,WA.EcologyPublicationnumber09‐03‐028.AccessedDecember20,2013at:

https://fortress.wa.gov/ecy/publications/publications/0903028.pdf

Guo,Z.,X.Liu,K.A.Krebs,R.A.Stinson,J.A.Nardin,R.H.Pope,andN.F.Roache.2011.

LaboratoryStudyofPolychlorinatedBiphenylContaminationandMitigationin

Buildings.Part1EmissionsfromSelectedPrimarySources.USEPAResearch

TrianglePark,NC.AccessedDecember5,2013at:

http://nepis.epa.gov/Adobe/PDF/P100F9XG.pdf

HerreraEnvironmentalConsultants.2011a.ControlofToxicChemicalsinPugetSound:Phase3DataandLoadEstimates.WashingtonDepartmentofEcologyPublication



11‐03‐010.AccessedDecember24th,2012.http://www.ecy.wa.gov/biblio/1103010.html

Herrera.2011b.LiteratureReviewofExistingTreatmentTechnologiesforIndustrial

Stormwater.HerreraEnvironmentalConsultants,Seattle,WA.AccessedDecember

19,2013at:http://www.wastormwatercenter.org/files/library/literature‐review‐

existing‐treatment‐technologies‐for.pdf

Hornbuckle,K.C.,J.D.Jeremaison,C.W.Sweet,andS.J.Eisenreich.1994.SeasonalVariations

inAir‐WaterExchangeofPolychlorinatedBiphenylsinLakeSuperior.Environ.Sci.

Tech.28(8):1491‐1501.AccessedJanuary14,2014at:

http://dx.doi.org/10.1021/es00057a018

Huddart,D.andT.Stott.2010.EarthEnvironments:Past,PresentandFuture.Wiley,John&

Sons,NY.

Jack,R.andColton,J.2011.LakeWashingtonPCBandPBDEloadings.PresentationatSalishSeaEcosystemConference.October25‐27,2011.VancouverBritishColumbia.Abstractavailableat:http://www.verney.ca/assets/Abstract%20book‐

Final_Nov14.pdfPages38and39.

Kalmar,J.2010.TechicalMemorandum:Short‐termtreatmentfacilitycompliance

assessment(Task1C)NorthBoeingField,Seattle,Washington.LandauAssociates,Seattle,WA.DatedNovember19,2010.AccessedDecember20,2013at:http://www.epa.gov/region10/pdf/sites/ldw/slip4/Boeing_NBF_Landau_111910_S

TTS_Compliance_Assess_TM.pdf

Kaya,E.,Y.Dumanogulu,M.Kara,H.Altoik,A.Bayram,T.Elbir,andM.Odabasi.2012.Spatialandtemporalvariationandair‐soilexchangeofatmosphericPAHsandPCBsinanindustrialregion.AtmosphericPollutionResearch3(2012)435‐449.Accessed

January14,2014at:http://www.atmospolres.com/articles/Volume3/issue4/APR‐12‐051.pdf

KCIWandSPU.2005.KingCountyandSeattlePublicUtilitiesSourceControlProgramfor

theLowerDuwamishWaterway:June2005ProgressReport.AccessedDecember

19,2013at:

http://www.kingcounty.gov/~/media/environment/wastewater/IndustrialWasteP

rogram/docs/Duwamish/ArchivedProgressRpts/June05/progress_report_06_05_n

ofigs.ashx

KingCounty.2011.LowerDuwamishWaterwaySourceTracinginKingCountyCombinedSewerSystem:SamplingandAnalysisPlan.PreparedbyD.WillistonandR.Jack,



KingCountyWaterandLandResourcesDivision,Seattle,WA.AccessedDecember

19,2013at:

http://your.kingcounty.gov/dnrp/library/wastewater/iw/SourceControl/Studies/LDW_Source_Tracing_SAP_Final_Aug2011.pdf

KingCounty.2013a.LakeWashingtonPCB/PBDELoadingStudyDataReport.Preparedby

JenéeColtonandRichardJack.WaterandLandResourcesDivision,Seattle,

Washington.

KingCounty.2013b.PCB/PBDELoadingEstimatesfortheGreaterLakeWashington

Watershed.PreparedbyCurtisDeGasperi.WaterandLandResourcesDivision,

Seattle,Washington.

KingCounty.2013c.DevelopmentofLakeWashingtonPCBFateandBioaccumulationModels.PreparedbyCurtisDeGasperi,JenéeColtonandCarlyGreyell.WaterandLandResourcesDivision,Seattle,Washington.

KingCounty.2013d.LowerDuwamishWaterwaySourceControl:BulkAtmosphericDepositionStudy:DataReport.PreparedbyJenéeColton,CarlyGreyellandRichardJack,KingCountyDeptofNaturalResourcesandParks,WaterandLandResourcesDivision,ScienceSection.Seattle,Washington.

KingCounty.2013e.LakeWashingtonTissueDataAddendumtoEstimatingPCBandPBDELoadingstotheLakeWashingtonWatershed:DataReport.TechnicalMemorandumPreparedbyCarlyGreyell,RichardJack,andJenéeColton.KingCountyDeptofNaturalResourcesandParks,WaterandLandResourcesDivision,ScienceSection.Seattle,Washington.

Krahn,M.M.,Hanson,M.B.,Baird,R.W.,Boyer,R.H.,Burrows,D.G.,Emmons,C.K.,Ford,J.K.B.,Jones,L.L.,Noren,D.P.,Ross,P.S.,Schorr,G.S.,Collier,T.K.,2007.Persistent

organicpollutantsandstableisotopesinbiopsysamples(2004/2006)fromSouthernResidentkillerwhales.MarinePollutionBulletin54:1903–1911.

McBride,D.unpublishedmemorandum“TotalPCB(Provisional)ScreeningValuesEnglish

sole–PugetSound(MarineWaters)”.EmailfromJoanHardy,WADepartmentof

HealthtoJenéeColtonofKingCounty,receivedAugust28,2013.

Munn,R.E.1976.AirPollutionMeteorologyinManualonUrbanAirQualityManagement.

Suess,M.J.andCraxford,S.R.eds.WorldHealthOrganizationEuropeanSeries

Publication1.,Copenhagen,Denmark.



Norton,D.,D.Serdar,J.Colton,R.Jack,andDLester.2011.ControlofToxicChemicalsin

PugetSound:AssessmentofSelectedToxicChemicalsinthePugetSoundBasin,

2007‐2011.WashingtonDepartmentofEcology.Olympia,WA.Pub.No.11‐03‐055

AccessedJanuary14,2014at:

https://fortress.wa.gov/ecy/publications/summarypages/1103055.html

Pelletier,G.andT.Mohamedali.2009.ControlofToxicChemicalsinPugetSound:Phase2,

Developmentofsimplenumericalmodels:Thelong‐termfateandbioaccumulation

ofpolychlorinatedbiphenylsinPugetSound.WashingtonDepartmentofEcology.

Olympia,WA.Pub.No.09‐03‐015.http://www.ecy.wa.gov/biblio/0903015.html

Robson,M.,L.Melymuk,S.A.Csiszar,A.Giang,M.L.Diamond,andP.A.Helm.2010.

ContinuingSourcesofPCBs:Thesignificanceofbuildingsealants.EnvironmentInternational36:506‐513.

Rodenburg,L.A.andMeng,Q.2013.SourceApportionmentofPolychlorinatedBiphenylsinChicagoAirfrom1996to2007.Env.Sci.Tech.47:3774‐3780.

Taylor,W.J.andCardnoTechInc.2013.FinalWhitePaperforStormwaterManagementProgramEffectivenessLiteratureReview:LowImpactDevelopmentTechniques.AccessedDecember2,2013at:

http://www.ecy.wa.gov/programs/wq/psmonitoring/ps_monitoring_docs/SWwor

kgroupDOCS/LIDWhitePaperFinalApril2013.pdf

Schmoyer,B.2007.SourceTracingintheLowerDuwamishWaterway,Seattle,WA.InOptimizingContaminantTrackdown:FocusingonWastewaterTreatmentPlants

andRelatedSystems.NYAcademyofSciences.AccessedDecember18,2013at:http://www.nyas.org/asset.axd?id=0f0f0d12‐2d2a‐4449‐93a0‐383d9d4e61ec

Schmoyer,B.2012.PCBSourceTracingintheDuwamishWaterway.MSPowerpoint

presentation.AccessedDecember20,2013at:http://www.cforjustice.org/wp‐content/uploads/2012/05/060512_Spokane_SPU_LDWSC1.pdf

Sloan,C.A.,B.F.Anulacion,J.L.Bolton,D.Boyd,O.P.Olson,S.Y.Sol,G.M.Ylitalo,L.L.Johnson.

2010.PolybrominatedDiphenylEthersInOutmigrantJuvenileChinookSalmon

FromTheLowerColumbiaRiverAndEstuaryAndPugetSound,WA.Archivesof

EnvironmentalContaminationandToxicology,58(2):403‐414.

WAC173‐303‐9904.AccessedDecember18,2013at:

http://apps.leg.wa.gov/wac/default.aspx?cite=173‐303‐9904



WADOH.2004.FinalReport:EvaluationofContaminantsinFishfromLakeWashington,

KingCounty,Washington.WashingtonStateDepartmentofHealth,Olympia,WA.

AccessedDecember4,2013at:

http://www.doh.wa.gov/Portals/1/Documents/Pubs/333‐061.pdf

West,J.E,S.M.O'Neill,andG.M.Ylitalo.2008.Spatialextent,magnitude,andpatternsof

persistentorganochlorinepollutantsinPacificherring(Clupeapallasi)populations

inthePugetScienceofTheTotalEnvironmentVolume.394(Issues2‐3):369‐378.