Professional Documents
Culture Documents
PROCEEDINGS
Formof
Topic Lastname Firstname Authors Title
presentation
MaleneToernqvistFront,CharlotteRiis,AndersG. DelineationofcontaminantplumesusingLowLevelMIHPT(LLMIHPT)
Christensen,NancyHamburger,PederJohansen,Lone
1A.6 ToernqvistFont Malene TolstrupKarlby Lecture
3DModellingofthesalt/freshwaterinterfaceincoastalaquifersofLowerSaxony(Germany)based
1A.4 Deus Nico NicoDeus,JrgElbracht,BernhardSiemon onairborneelectromagneticmeasurements(HEM) Lecture
JacquelineAnneFalkenberg,MetteMarieMygind,Anne ScreeningforFluorinatedCompounds(PFAS)aroundPotentialSourcesofPollutionatDanishDefence
MetteBrunerLindof,JetteKjgeOlsen,JensDengs Establishments
1A.5 Falkenberg JacquelineAnne Jensen,AndersG.Christensen Lecture
"Multiplelinesofevidence"approachappliedtoAssessmentofResidualDenseNonAqueousPhase
1A Simone Laura LauraSimone,ThomasHeld Liquid(DNAPL) Poster
EzioCrestaz,AntonioAncilletta,LeonardoPatata,Michele GroundwaterflowandtransportmodellingaimedatanexploratoryspatialdataanalysisofUst
1A Ancilletta Antonio Pellegrini,FeliceTatangelo,StefanoIsidori Kamenogorskaquifersystem,EastKazakhstan Poster
Groundwaterflowandtransportmodellingasatoolfordevelopingandevaluatingremedialactions
1A Coetsiers Marleen MarleenCoetsiers,PieterjanCallewaert,MicheleRemonti casestudyofacomplexgroundwatercontamination Poster
AndresaOliva,MarcusBaessa,EliasIsler,MarcoPede, OptimizingofTheCharacterizationofNAPLContaminatedSiteUsingUltraVioletOpticalScreening
1A Oliva Andresa HungKiangChang ToolandGeophysicalSurvey Poster
TracinganextensiveMtBEplumeinasandyaquifertryingtoescapemonitoringwellsusingthorough
1A Larsen LarsChristian LarsChristianLarsen,PederJohansen hydrogeologicalsiteunderstanding Poster
1A Lobo M.Carmen M.MarGilDiaz,AnaM.Guerrero,M.CarmenLobo Monitoringofthedoramectindiffusionintosoilamendedwithpigmanure Poster
ChristopherTeasdale,JeanHall,JohnMartin,David GroundGasMonitoring:ExaminingSpatialTrendsonShortandLongTemporalScales
1A Teasdale Christopher Manning Poster
CharlotteRiis,GroLilbaek,AndersG.Christensen,Mikael DetailedHydrogeologicalCharacterizationofaSoftChalkAquiferthroughHPTLoggingandCoring
1A Riis Charlotte Joergensen,MartinStrmose Poster
NiloofarShoari,JeanSebastienDub,Shoja'eddin StatisticaldistributionofleftcensoredconcentrationdataExamplefromasoilcharacterizationstudy
1A Shoari Niloofar Chenouri inMontreal Poster
1A Kaman Harun HarunKaman,Mahmutetin,CevatKrda Determinationofsaltbudgetforlargeirrigatedareas Poster
1A Kaman Harun HarunKaman,AbdullahSayc,Assist.Selukzmen Determinationofirrigationwatersalinity:acasestudy Poster
InvestigationofirrigationwatersqualitiesinAntalyaKumlucaregion,Turkey
1A Kaman Harun HarunKaman,Selukzmen,Hasanetin,AbdullahSayc Poster
1A Kozubek Petr PetrKozubek VariabilityinGroundwater:NaturalAttenuationRatevs.ShortTermFluctuations Poster
MariadeLurdesDinis,JoaquimGis,AntnioFiza,Jos FateofHazardousElementsinSoilsSurroundingaCoalFiredPowerPlantComplex
1A Fiza Antnio SoeiroCarvalho,AnaCristinaMeiraCastro Poster
BiologicaldegradationofUVfiltersinmarinesediments:alaboratorymicrocosmstudy.
AngelaVolpe,MichelePagano,PaolaGrenni,AnnaBarra
1A Volpe Angela Caracciolo,GiuseppeMascolo,SimonaRossetti Poster
BestAvailableTechniqueforSurfaceWaterDischargeMonitoringatFinnishOutdoorShootingRanges
1A Takala Mikael MikaelTakala,JenniTakala Poster
1A Zakariadze Nino NinoZakariadze PreparationofMedicalGeologyAtlas Poster
HubertdeJonge,LoneTolstrupKarlby,NannaMuchitsch, Anovelpassivesamplingapproachforinsitumeasurementofgroundwaterandorganicpollutant
1A deJonge Hubert MadsTerkelsen fluxes:conceptualapproachandcasesfromtheCapitalRegionofDenmark Poster
TageVikjrBote,PerLoll,ThomasLarsen,Bjarke Vaporintrusionstateoftheart FreeSession
1B.5S Mller MadsGeorg Hoffmark,ArneRokkjr,MadsGeorgMller Proposal
1B.6 Wagelmans Marlea MarleaWagelmans Protocolsforecologicalriskassessment Lecture
Residentiallocationcontaminatedwithcumene:buildingteamconstructionresultsinasuccessful(in
1B.7 Ramakers Peter PeterRamakers,JoostVanSchijndel,GerardBorggreve situ)remediation Lecture
PaulFrognerKockum,MrtaLndell,GunnelGransson, Quantificationofcontaminanttransportfromsediment
1B.9 FrognerKockum Paul YvonneOhlsson Lecture
Assessmentofriskscausedbythepermeationoforganiccontaminantsingroundwaterthrough
1B.9 Otte Piet PietOtte,MartinSchans,MartinMeerkerk,FrankSwartjes polyethylenedrinkingwaterpipes Lecture
1B.10 Leroi Steve SteveLeroi,AdrienKahn anapproachtoRiskassessmentandmanagementofcontaminatedlandinAsia Lecture
1B.8 Provoost Jeroen JeroenProvoost,JanBronders,IlseVanKeer Probabilisticriskassessmentforsixvapourintrusionalgorithms Lecture
1B.8 Provoost Jeroen JeroenProvoost,KarenVictor Newconceptsinvapourintrusion Lecture
JooSungAhn,SangWooJi,YongChanCho,SeungJun Predictingtheoccurrenceofacidrockdrainagethroughageochemicalstreamsedimentsurvey
1B Ahn JooSung Youm,GilJaeYim Poster
ChristaCornelis,MirjaVanHolderbeke,AnnColles,Griet Contaminantintakefromeggsofbackyardchicken:whataresafesoilconcentrations?
1B Cornelis Christa VanGestel Poster
JesperAlroeSteen,MarenHostrup,MetteMarieMygind, Occurenceofantimonyinsoilandgroundwateratformershootingranges
BoergeHansen,HenrikSoegaardLarsen,JacquelineAnne
1B Steen JesperAlroe Falkenberg Poster
CamillaPease,JenniferGill,DavidSchofield,MattPannett, RecentAdvancesinHumanHealthToxicologyofTrichloroethylene(TCE)anditsApplicationinSoil
1B Pease Camilla AndrewGoddard ScreeningCriteria Poster
PaulBardos,JuergenBraun,Miroslavernk,DanielElliott, Nanoremediationallyouwantedtoknow(apracticalguidetonanoremediation) FreeSession
1C.23S Bardos Paul ElsaLimasset,HansPeterKoschitzky Proposal
PaulBardos,StephanBartke,NicolaHarries,HansPeter Nanoremediationyourfuturebusinessopportunities(strategicandmarketintelligence) FreeSession
1C.24S Bartke Stephan Koschitzky Proposal
MattiasVerbeeck,RichardLookman,JohanGemoets, InsituZincBioprecipitationthroughOrganicSubstrateInjectioninaHighflowAquifer:From
1C.12 Verbeeck Mattias BeatrijsLambi LaboratorytoFullscale Lecture
RichardBewley,RickParkman,PaulBardos,Marcusvan TheRegulatoryBasisforSustainableRemediationPracticeintheEuropeanUnionandUnitedKingdom
1C.2 Bewley Richard Zutphen,JonathanSmith Lecture
1C.3 Patek Regine ReginePatek PublicFundingschemeforRemediationProjectsinAustria Lecture
AcceleratingTrichloroethyleneRemediationinSaproliteandFracturedCrystallineBedrockbyInsitu
ChemicalOxidationandInsituChemicalReductionaSuccessfulCaseStudyofCombinedRemedies
1C.15 Maalouf GeorgeY. GeorgeY.Maalouf at a Challenging Site Lecture
1C.19 Pensaert Stany StanyPensaert EnvironmentaldredgingofachromiumcontaminatedfjordinValdemarsvik,Sweden. Lecture
1C.7 Birnstingl Jeremy JeremyBirnstingl CombinedRemedySynergiesExamplesandConceptualRoadMap Lecture
1C.9 Schmidt KathrinRachel KathrinRachelSchmidt,SarahGaza,AndreasTiehm Aerobicbiodegradationoftrichloroethenewithoutauxiliarysubstrates Lecture
ChallengesandhopesforscalingupanelectrodialyticremediationmethodfortreatingCCA
KrzysztofKowalski,SanneSkovNielsen,PernilleErland contaminatedsoil
1C.20 Kowalski Krzysztof Jensen,ThomasLarsen,MadsTerkelsen,LisbethOttosen Lecture
CharlotteRiis,MartinBymose,DortePade,EvanCox, FullScaleElectrokineticsEnhancedBioremediation(EKBIO)ofPCEDNAPLSourceAreainClayTill
1C.12 Riis Charlotte JamesWang,DavidGent,MadsTerkelsen Lecture
CombinedFentonlikeoxidationandCO2spargingforthetreatmentofgroundwatercontaminatedby
1C.17 Baciocchi Renato DanielaZingaretti,IasonVerginelli,RenatoBaciocchi organiccompounds Lecture
GerardBorggreve,AlbertSmits,DennisScheper,Adri FullyAutomatedEnhancedBiodegradationofChlorinatedEtheneswithanOnSiteAnaerobic
1C.10 Borggreve Gerard Nipshagen,ReneTjassens,MichielPluim Bioreactor Lecture
AlbertSmits,GerardBorggreve,DennisScheper,Mart TheAdvantageofBenchScaleTreatablityStudiesasaDecisionMakingToolforaFullScaleIsco
1C.17 Smits Albert Jansen,MichaelMueller ApproachinanInnovativeTenderProcedure Lecture
Batchtestsandfieldapplicationofinsituremediationofgroundwatercontaminatedwithchlorinated
Gammeltoft AnneGammeltoftHindrichsen,JohnUlrikBastrup,Jan solventsbydirectinjectionofnanoscaleZeroValentIrononthreelocationsinDenmark.
1C.22 Hindrichsen Anne Slunsk Lecture
SILPHESInvestigationofchemicaltreatmentsfortheremediationofrecalcitrantchlorinatedsolvents
RomainRodrigues,StphanieBetelu,FrdricGarnier,
StfanColombano,AntoineJoubert,DavidCazaux,
1C.10 Rodrigues Romain GuillaumeMasselot,TheodoreTzedakis,IoannisIgnatiadis Lecture
PollutionofsoilandgroundwaterbyindustrialoilsdumpinginJaramaRiverBasin(Madrid,Spain)
1C.7 Villarroya Fermn FermnVillarroya,EsperanzaMontero,JuanPedroMartn Lecture
EdwardvandeVen,ArtLobs,TimDeBouw,Richard ImplementingInsituChemicalOxidationonanindustrialEXratedsite
1C.16 Lookman Richard Lookman Lecture
1C.7 Beaulieu Michel MichelBeaulieu LacMegantic:Therehabilitationofatownfollowingapetroleumloadedtrainexplosion Lecture
Meetingthechallengesforbioremediationofchlorinatedsolventsposedatoperationalsites:a
1C.9 Bewley Richard RichardBewley,PaulaHick,AntheaRawcliffe comparisonofcasestudies Lecture
Formof
Topic Lastname Firstname Authors Title
presentation
GroLilbaek,JacquelineAnneFalkenberg,AndersG. EstimationofRemediationRatesforChlorinatedSolventsinConfinedUnsaturatedMedia
1C.8 Lilbaek Gro Christensen,HelleOvergaard Lecture
1C.3 Nathanail C.Paul C.PaulNathanail ProgresstowardsanISODocumentonSustainableRemediation Lecture
BuildingaNetworkbasedExpertStakeholderFrameworkforSustainableRegeneration
1C.2 Rizzo Erika FilipAlexandrescu,ErikaRizzo,LisaPizzol,AndreaCritto Lecture
PetrKozubek,JanNemecek,VladislavKnytl,Eliska DNAPLtreatedbyapplicationofsurfactantsfollowedbyISCO
1C.14 Kozubek Petr Kosinova Lecture
1C.21 TaylorKing Christopher ChristopherTaylorKing SurfactantEnhancedAquiferRestorationatFormerChemicalWorks Lecture
1C.7 Bastrup JohnUlrik JohnUlrikBastrup,JieCheng,DanielChiang RemediationinChina Lecture
ChristosTsakiroglou,MScKaterinaTerzi,Alexandra OptimizingthepropertiesofnanofluidsfortheefficientNAPLremediationinporousmedia
1C.22 Tsakiroglou Christos SikiniotiLock,KataHajdu,ChristosAggelopoulos Lecture
OliverTrtschler,HansPeterKoschitzky,BerndLidola, SteamAirInjectioninFracturedBedrock:ResultsandLessonsLearnedofaCHCRemediationAtThe
1C.27 Trtschler Oliver IsabellKleeberg,StefanSchulze SiteBiswurm(VillingenSchwenningen,Germany) Lecture
Complexboundaryconditionsforinsituthermaltreatments(ISTT)conductedduringlandrecycling
1C.27 Hiester Uwe UweHiester,MartinaMller andremediationbeneathbuildings Lecture
1C Wagelmans Marlea MarleaWagelmans,NielsvanRas Usesedimentsforeconomicallyviablegroundwaterpollutionmanagement Poster
AlessandradeFollyd'Auris,MarcoTagliabue,Roberto Thermaldesorptiontechnology:twocasestudies
1C deFollyd'Auris Alessandra Bagatin Poster
SuccessfullystimulatedbiologicalinsituremediationVOCcontaminatedsourcezones
1C Henssen Maurice MarleaWagelmans,AnnekeRoosma,MauriceHenssen Poster
YasuhideFurukawa,TetsuoYasutaka,TakashiKobayashi, Acasestudyoffiveoptionsforsustainableinsituremediationofcontaminatedsoil
1C Furukawa Yasuhide TakaakiShimizu Poster
MaritsaKurashvili,TamarVarazi,MarinaPruidze,Gia Modeltestingofnewapproachforcleaningsoilspollutedwithorganochlorinepesticides
1C Kurashvili Maritsa Khatisashvili Poster
EnhancedReductiveDehalogenationofChloroetenesbyApplicationofCheeseWheyinSitu:aCase
1C Stavelova Monika MonikaStavelova,IvaDolinova,MariaBrennerova StudywithField,ChemicalandMolecularBiologyAnalysis Poster
SusanaDelReinoQuerencia,OlgaEscolano,Manuel ISCOapplicationinasaltmarshareacontaminatedbyfuelinsouthernSpain(BIOXISOILproject)
1C Escolano Olga RodriguezRastrero,JorgeBueno,RociMilln Poster
TamarVarazi,MaritsaKurashvili,MarinaPruidze,Gia DifferentToolstoAvoidDeepContaminationandtoImprovePhytoremediationTechnology
Khatisashvili,GeorgeAdamia,NinoGagelidze,Marlen
1C Varazi Tamar Gordeziani,GeorgeZaalishvili,MarkSutton Poster
VicenMart,NuriaRoyo,OriolGibert,CesarAlberto DenitrificationcapacityofleachatesfromwetlandsoilsandvegetalbiomassintheLlobregatriver
Valderrama,MariaRosarioMartnez,DavidRibas,Irene basin
1C Mart Vicen Jubany Poster
DrM.MarGilDiaz,DrAguedaGonzlez,JuanAlonso, ImmobilizationofCrinsoilusingstabilizednanoscalezerovalentiron.Impactonbarleyplants.
1C Lobo M.Carmen PalomaPinilla,M.CarmenLobo Poster
AndreaAmantia,MScGiacomoDonini,MscEnvironmental PhytoGroundwaterContainmentinaLargePollutedSiteinSouthernItaly
1C Canova Fabio CarlosHerrarte,FabioCanova Poster
1C Jez Erika DomenLestan,ErikaJez,NezaFinzgar RemediationoftoxicmetalscontaminatedsoilusingEDTAsoilwashing Poster
NereaOtaegi,EkainCagigal,Miroslavernk,JanSlunsk, TestingonemergingNanoparticlesforArsenicremovalunderrealconditionsonapilotfieldsite,in
1C Otaegi Nerea JulianBosch Asturias,Spain Poster
YiKuoChang,ChengDiDong,ChiuWenChen,ChihFeng Applicationofhydrocycloneseparationsystemforcontaminatedsoiltreatmentandflowsimulation
1C Chang YiKuo Chen,ZhenWeiHong,YiLiangShu,ChiaYingLi Poster
FabienLION,NicolasAubert,GeoffreyBoissard,Stfan Frenchnationalapproachfordelineationofcontaminantsourcezones:combiningmassbalance,mass
1C LION Fabien Colombano fluxandmassdischargemethods Poster
IsolationandIdentificationofCultivableAnthraceneDegradingFungifromPAHContaminatedSoilsfor
PatriciaGodoy,MScAndreaCaldern,StudentAngela theirApplicationinPhytoremediation
LpezEugenio,StudentRocoReina,ResearchScientist
1C Aranda Elisabet InmaculadaGarcaRomera,ElisabetAranda Poster
PetraNajmanov,OndejLhotsk,VladislavKnytl,Jan RemovalofhexavalentchromiumbyapplicationofnZVINANOFERSTAR
1C Najmanov Petra Slunsk,JanFilip Poster
SaraDastoli,ElenaRomano,AlessandraPolettini,Raffaella Coordinatedapproachfortreatmentofdredgedsedimentfromsmallharbours
Pomi,GiorgiadeGioannis,MaurizioMorelli,AldoMuntoni,
RobertoPeretti,AntonelloSerci,BarbaraVillani,Antonello
1C Dastoli Sara Zucca Poster
AnettGeorgi,GlennGillies,KatrinMackenzie,FrankDieter ColloidalFezeolitesanovelmaterialforsorptionsupportedinsituchemicaloxidation(ISCO)
1C Georgi Anett Kopinke Poster
PlantresponseofchromiumtolerantandsensitiveclonesofSilenevulgarisaffectsrhizosphere
1C Lobo M.C. A.E.PradasdelReal,M.C.Lobo,A.PrezSanz bacterialcommunities Poster
SusanaMachado,JooPacheco,JosTomsAlbergaria, RemediationofwaterandsoilcontaminatedwithAmoxicillinusinggreenzerovalentiron
1C Machado Susana CristinaDelerueMatos nanoparticles Poster
BurcuOzkaraovaGungor,AyseKuleyin,FeryalAkbal,M. FentonProcessforThePurificationofPollutedWaters
1C Gungor BurcuOzkaraova OyaOrkun,DilekGm Poster
Adsorptionbasedtechnologiesforinnovative,sustainable&onsitehydrocarbonremediation
1C Agut Adeline AdelineAgut Poster
1D.2 Mars JanFrank JanFrankMars,PeterRamakers,ReinderSlager Flowers4Brabant Lecture
1D.3 Wagelmans Marlea JannekeWittebol,MarleaWagelmans Biologicaltreatmentofmicropollutantsindrinkingwaterresources Lecture
CorinneKoot,AnnemiekWiegman,ReinderSlager,BSc RegionallyapproachedgroundwatermanagementinZwolle:preventingrisksandutilizing
1D.2 Koot Corinne MartijnvanHouten opportunities Lecture
1D.2 Vasin Sandra SandraVasin,HermannJosefKirchholtes AGroundwaterManagementPlanforStuttgart Lecture
Howtogetacameltogothroughtheeyeofaneedle:Successfulsiteremediationofaformer
explosivesproductionsite:safehousing,workinganddrinkingwaterproductiononalongtermbasis
1D.3 Weingran Christian ChristianWeingran,H.GeorgMeiners Lecture
Successandfailurefactorsareawidegroundwatermanagement
1D.2 Alphenaar Arne ArneAlphenaar,FrankSwartjes,PietOtte,ReinderSlager Lecture
TheFinalCountdownSuccessfulremediationpoliciesleadstotheendoftheDutchSoilProtection
2.3S Gadella Michiel MichielGadella,CoMolenaar Act Lecture
2.5 Wortelboer Rob RobWortelboer ContaminatedSludgeBeneficiallyUsedinFoundations Lecture
DennisScheper,GerardBorggreve,AlbertSmits,Adri WhatCanYouDoforOneandaHalfMillionUrbanRedevelopmentThroughImplementationofNew
2.4 Scheper Dennis Nipshagen,DickSpecht,LuukWallinga Technology Lecture
Usinginnovativegeotextileconstructionsasaninsitubioremediationtechniquetoremediate
2.6 Lauwerijssen Chiel ChielLauwerijssen contaminatedsedimentsandtoimprovewaterqualityofshallowlakes Lecture
2.5 Dedecker Dirk DirkDedecker,IrFilipDeNaeyer,Lic.EddyVanDyck Flemishpolicyontheuseofexcavatedsoil Lecture
2.6 Dickinson Claire ClaireDickinson,HilaryAllen TheCircularEconomyMaximisingtheReuseofSoilsMakingithappen Lecture
EffectsofPlantGrowthPromotingMicroorganismsonWheatPlantGrowthandAminoAcidContent
AdemGnes,NurgulKtr,MetinTuran,FikrettinSahin,
2 Gnes Adem MedineGulluce,Prof.GlerayAar,Haticet Poster
ViktoriaFeigl,MnikaMolnr,vaUjaczki,Orsolya Ecotoxicityofbiocharsfromorganicwastesfocusingontheiruseassoilameliorant
Klebercz,IldikFeketeKertsz,MriaTolner,Emese
2 Feigl Viktoria Vaszita,KatalinGruiz Poster
LetsmakegroundwaterSTRONGerawatersystembasedapproachtowards3Dspatial
3.4 Romijn Reinier ReinierRomijn,AlmerBolman development. Lecture
3.4 Knudsen Theis TheisKnudsen,ChristianHelweg,LarsJacobsen AquiferThermalEnergySystemsinareasofdrinkingwaterandgroundwaterpollutions Lecture
Ecosystemservicesofthegroundwaterandthesubsurface;fillingtheknowledgegap
JohannesP.A.Lijzen,drs.SophieVermooten,HansPeter
3.5 Lijzen JohannesP.A. Broers,SuzannevanderMeulen,MichielRutgers Lecture
3 Mars JanFrank JanFrankMars,LindaMaring,AlbertdeVries Towardsasustainableandresilientcity Poster
3 Nathanail C.Paul C.PaulNathanail,MatthewAshmore TheHOMBREBR2:BrownfieldREMIT/RESPONSEformodellingtheurbansubsurfacesystem Poster
EdinoValcovich,CarloAntonioStival,RaulBerto,Giovanni Greenroofstechnologicalanalysisaccordingtoaperformancebasedapproach
4 Berto Raul Cechet Poster
Open ApplicationofTrap&TreatTechnologyforAchievingSustainableRemediationofContrasting
Topic Wilson James JamesWilson,Clauslund ContaminantPlumes
DELINEATION OF CONTAMINANT PLUMES USING LOW-LEVEL MIHPT (LL-MIHPT)
1
NIRAS, Sortemosevej 19, DK-3450 Alleroed, Denmark, +45 4810 4200, maf@niras.dk, cer@niras.dk and agc@niras.dk
2
Capital Region of Denmark, Department of Regional Development, Kongens Vaenge 2, DK-3400 Hilleroed, Denmark, +45
3866 5000, nancy.hamburger@regionh.dk and peder.johansen@regionh.dk
3
COWI, Parallelvej 2, DK-2800 Kgs. Lyngby, +45 5640 0000, ltka@cowi.dk
Plumes of dissolved contaminants may be widely distributed in the saturated zone, i.e. to great depths
and over large areas. Also the concentration levels of contaminants in the plume generally are much
lower than the concentrations found in the hotspot area. Characterization and delineation of
contaminant plumes therefore typically require many investigation points at great depths. Often the
delineation of a contaminant plume is conducted using traditional drilling techniques and installation of
screened wells. However, this method is both time-consuming and resource-intensive and only a
limited number of discrete depths can be screened in each well. Furhermore, screen depths are often
chosen based on expected flow and contaminant distribution patterns in the saturated zone, rather
than on detailed hydrogeological data and vertical contaminant distribution.
On behalf of The Capital Region of Denmark, Department of Regional Development, and in
cooperation with experts from Geoprobe Systems (US), NIRAS A/S (DK) has tested a novel tool, Low-
Level MIHPT (LL-MIHPT), for delineation and characterization of contaminant plumes in groundwater.
LL-MIHPT is based on the existing high resolution direct push investigation tools MIP and HPT
developed by Geoprobe Systems. The HPT probe (Hydraulic Profiling Tool) is used to continuously
map the hydrogeological conditions (permeability), while the MIP probe is used to continuously map
VOC contamination in soil and groundwater. The MIHPT system is a combination of the MIP and HPT
systems and has proved to be very efficient for field investigations in hotspots and areas with high
contaminant levels. However, the detection limits of the standard MIHPT system are too high for
delineation of contaminant plumes where the concentration levels are significantly lower than in the
source zones. The Low-Level MIHPT system is developed with the objective to detect contamination
at low concentrations and thus provides a means for conducting more time efficient and cost-effective
delineation of contaminant plumes in unconsolidated saturated formations.
The LL-MIHPT systems has been tested at two sites located in the towns of Farum and Slangerup in
Denmark as part of ongoing field investigations at the two sites.
The purpose of this project was to test the LL-MIHPT technique for delineation of contaminant plumes
in groundwater at two sites with different geological formations; sandy and clayey, respectively. The
objectives have thus been to determine at which concentration level the LL-MIHPT system could
detect the site specific contaminants and to investigate the correlation between observed LL-MIHPT
responses and results from analysed water samples from targeted depths.
9 LL-MIHPT logs to 20-25 meters below surface have been carried out. At each log water samples
were collected at specific depths with the GeoProbe for verification of the observed responses from
the LL-MIHPT and for correlation of contamination levels. For further correlation of the LL-MIHPT data
core samples were collected at three locations.
The results from the field tests show that it is possible with the LL-MIHPT to track relatively low
concentrations of chlorinated solvents and BTEXs in the saturated zone. Hence, for chlorinated
solvents a detection limit in the order of 10 g/L can be expected. For comparison the detection limit
for chlorinated solvents with the standard MIP system is in the order of 1-10 mg/L.
Based on the results and experiences obtained from the field tests the new LL-MIHPT system shows
good promise for delineation of contaminant plumes in the saturated zone with simultaneous retrieval
of hydrostratigraphic data from the saturated zone. Thus, LL-MIHPT logs followed by depth specific
groundwater sampling with the GeoProbe system is considered to be an optimal set-up for delineation
and characterization of contaminant plumes in saturated zones in unconsolidated geological
formations.
The field tests were conducted and evaluated in the fall of 2013 and spring of 2014. Since then NIRAS
A/S has obtained good results with the LL-MIHPT system at field investigations at several other sites.
Session: 1A.4 - Novel monitoring approaches II
NICO DEUS, State Authority for Mining, Energy and Geology (LBEG), Germany
DR. JRG ELBRACHT, State Authority for Mining, Energy and Geology (LBEG), Germany
DR. BERNHARD SIEMON, Federal Institute for Geosciences and Natural Resources (BGR), Germany
Abstract
Airborne electromagnetic methods are able to distinguish high conductive saltwater from resistive
freshwater. Coastal areas such as the German Bight are at risk to large storms and rising sea level
that may affect the salt-/freshwater interface (SFWI). The understanding of the processes and the
current status of the dynamic SFWI is very important for the water management in these areas.
Therefore the German Federal Institute for Geosciences and Natural Resources (BGR) has focused
on research projects applying their geophysical airborne system to investigate the SFWI on the
German North Sea coast. The helicopter electromagnetic data (HEM) were collected with a Resolve
system operating at six-frequencies (SIEMON et al., 2015 in press). The spatial conductivity distribution
derived from the HEM data provides information on water salinity and lithology.
Based on the information from the conductivity distribution the State Authority for Mining, Energy and
Geology planed to create a 3D model of the SFWI for Lower Saxonian coastline.
Introduction
The public water management of Lower Saxony has to provide about 8 million people with fresh water.
With a portion of 71% of the public water supply in Germany (BGR, 2010), the fresh groundwater is the
most important resource for urban, agricultural and industrial activities. Some of the aquifers used for
groundwater extraction are affected by intruding saltwater from different sources (GRUBE et al., 2000).
Especially the coastal aquifers may be vulnerable for sea water intrusion, which is the landward
encroachment of seawater into freshwater aquifers (IVKOVIC et al., 2012). This process could be
enhanced or initiated by anthropogenic activities. Because of the importance of the saltwater intrusion
problems, the State Authority for Mining, Energy and Geology (LBEG) planed, based on a pilot project
in the area of Esens (DEUS, 2012), to generate a statewide saltwater map for Lower Saxony with a
focus on the coastal aquifers influenced by sea water intrusion.
In Germany the use of fresh water as drinking water is limited through the thresholds for different
parameters in the Trinkwasserverordnung (German drinking water regulation) (BMJ, 2013). Those
are 250 mg/l chloride, 250 mg/l sulfate and 200 mg/l sodium (BMJ, 2013). The threshold for the
electrical conductivity at 20C is 2790 S/cm (BMJ, 2013) which correlates with an electric resistivity of
4 m. These parameters were used to characterize the groundwater and divide it into salt-/freshwater
areas.
Geophysical methods are widely used in groundwater exploration. Particularly airborne geophysical
methods using helicopters enable economic and ecological mapping of near surface sediments and
resources. Electromagnetic and geoelectrical methods are able to map hydro geological units and
distinguish conductive clayey material from resistive sand and gravel as well as conductive saltwater
from resistive freshwater (KIRSCH, 2006).
Methods
The resistivity data from the electromagnetic measurements were put into the geologic modeling
software GOCAD in form of vertical resistivity profiles and horizontal resistivity maps for different
depths (see fig. 2a & b). Combining this resistivity information with groundwater analyses in the study
area and the distribution of
the low resistivity layers
(Lauenburg Clay & Holocene
brackish to marine
sediments) provide
reasonable information
about the SFWI progression.
The maximum landward
extend of the salt-/freshwater
interface has been picked on
each horizontal depth map.
In the next step, the vertical
resistivity profile has been
analyzed on the appearance
of saltwater and saltwater up
coning zones. The SFWI has
been digitized in these
profiles as curves. The Elbe
River and the sea level of the
North Sea have been used
as boundary and seaward
saltwater head for the NE
part of the study area.
Figure 2: a) HEM results: Horizontal resistivity map atr a depth of 5 m below sea level through connecting the data
(b.s.l.). The resistivity is shown in the color scale bar on the right site. Low resistivities
are represented through purple to red and yellow colors. Green and blue colors points gained in the steps
represent higher resistivity above 30 m (STEUER et al., 2013). b) West-East orientated
before. Finally the depth of
vertical resistivity profiles in the study area. The colors are equal to 2a (STEUER et al.,
2013). the SFWI surface model has
been calculated in meters below the Digital Elevation Model (DEM, 5m raster).
Results
The SFWI dips from sea level at the Elbe River towards the inland. The maximum depth for the SFWI
investigation is 60 m below sea level (b.s.l.) due to the limited depth of investigation achievable with
HEM. In figure 3 the distribution of the SFWI is shown in meters below the DEM.
Figure 3: Elevation map showing the SFWI in the study area. The depth of the SFWI below ground level is illustrated through
the different colors (see color bar). In the southern part of the study area the Lauenburg clay that potentially may have
resistivities similar to those of brackish groundwater is deposited (see red surface).
The depth of the SFWI dips landwards which is a well known phenomenon for seawater intrusion
(W ERNER et al., 2013). It dips rapidly within a few kilometers from sea level down to about 60 m b.s.l.
The groundwater table in the whole northeastern part of the study area is about 0 to 1 m a.s.l., which
approximately corresponds to the ground level in this area.
Combining the areas of salty groundwater with the distribution of Lauenburg clay we can see that
there is not much overlapping between these two phases. In almost the whole area a discrimination of
the Lauenburg clay based on the geologic model from saltwater indications in the HEM data was
possible. An additional help have been the groundwater analyses data with chloride concentration for
a specific filter pipe depth. These data has been integrated into the interpretation workflow and has
been used as hard data.
Discussion
Airborne electromagnetic is a valuable tool to derive a 3D-surface of the salty groundwater in coastal
aquifers. The identification of the SFWI based on the HEM data is a qualitative approach. We are not
able yet to get a quantitative chloride concentration of the groundwater. Therefore further combined
measurements and analyses must be executed.
A next step relating to this study will be the evaluation of the SFWI with several Direct Push drillings
through the interface. Electric resistivity as well as groundwater samples will be taken every few
meters while drilling to get a comparison to the identified SFWI in the HEM data. An important aspect
will be the calibration between electric resistivity and chloride concentration in the fluid phase to get a
better sense for the correlation of resistivity and concentration that could be adopted for following
studies.
Due to the limited HEM investigation depth (which is depending on the lithology) we have to use
further data, such as a combination of geoelectric data and chloride concentration derived from
groundwater samples, to identify the SFWI at greater depth.
References
DEUS, N. (2012): Kartierung der Kstenversalzung mit Hilfe geophysikalischer Daten und 3D-
Modellierung im Raum Esens (Ostfriesland). Hannover (unverff. Masterarbeit Univ. Hannover), 92pp.
GRUBE, A, W ICHMANN, K., HAHN, L. & NACHTIGALL, K.H. (2000): Geogene Grundwasserversalzung in
den Poren-Grundwasserleitern Norddeutschlands und ihre Bedeutung fr die Wasserwirtschaft. TZW-
Schriftenreihe, 9, Karlsruhe, 203 pp.
IVKOVIC, K.M., MARSHALL, S.K., MORGAN, L.K., W ERNER, A.D., CAREY, H., COOK, S., SUNDARAM, B.,
NORMAN, R., W ALLACE, L., CARUANA, L., DIXON-JAIN, P. & SIMON, D. (2012): National-scale vulnerability
assessment of seawater intrusion: summary report. Waterlines Report Series No 85, Australia, 185 pp.
KIRSCH, R., 2006, Hydrogeophysical properties of permeable and low permeable rocks, in Kirsch, R.,
ed., Groundwater Geophysics A Tool for Hydrogeology: Springer, 122.
SIEMON, B., COSTABEL, S., VO, W., MEYER, U., DEUS, N., ELBRACHT, J., GNTHER, T. & W IEDERHOLD,
H. (2015): Airborne and ground geophysical mapping of coastal clays in Eastern Friesland, Germany.
Geophysics, Special Issue: Near Surface Geophysics (In press).
STEUER, A., SIEMON, B., IBS-VON SEHT, M., PIELAWA, J. & VO, W., (2013): Technischer Bericht
Hubschraubergeophysik Befliegung Glckstadt, Sommer 2008 Frhjahr 2009. BGR-Bericht,
Archiv-Nr. 0131097, Hannover.
W ERNER, A.D., BAKKER, M., POST, V.E.A, VANDENBOHEDE, A., Lu, C., ATAIE-ASHTIANI, B., SIMMONS,
C.T. & BARRY, D.A. (2013): Seawater intrusion processes, investigations and management: Recent
advances and future challenges.- Advances in Water Resources, 51 (2013), 3-26.
Session 1A.5 - Persistence of historical and emerging subsurface contaminants.
Screening for Fluorinated Compounds (PFAS) around Potential Sources of Pollution at Danish
Defence Establishments
1 2 2 1
Jacqueline Falkenberg , Anne Mette Lindof , Mette Marie Mygind , Jette Kjge Olsen , Jens Dengs
1 1
Jensen and Anders G. Christensen
1
NIRAS, Denmark (jaf@niras.dk)
2
Ministry of Defence, Estates and Infrastructure Organisation (EIO), Environmental and Nature Section,
Denmark
Background
Perfluoroalkyl and polyfluoroalkyl substances (PFASs) have been in focus the last 10 15 years as
persistent and undesirable contaminants. Concern about health and environmental risks led to the EC
regulation in 2006 of a specific compound, PFOS (perfluorooctane sulfonic acid) and its derivatives (salts,
amides, halides etc.). These are only allowed in very low concentrations or in certain closed loop production
systems. This meant that the import or production of PFOS in fire-fighting foams was banned in the EU in
2006, but stockpiles could be used until 27 June 2011. PFOS and its derivatives have been listed since 2009
as a POP (Persistent Organic Pollutant) under the Stockholm Convention. Certain other specific PFAS
substances and their derivatives including PFOA (perfluorooctanoic acid) are included on the EU Candidate
List as Substances of Very High Concern (SVHC).
Because of this concern, many reports (Danish, Swedish, and other EU lands) have been produced with
surveys identifying products or waste which might contain PFAS as well as environmental reports mapping
the occurrence of PFAS in soil, groundwater and biota.
Objective
Due to general awareness of the risk of soil and groundwater contamination with PFAS especially PFOS and
PFOA, the Estates and Infrastructure Organisation (EIO) have implemented screening for PFAS while
monitoring groundwater quality at some active and former air bases. The initial focus has been at air bases
where fire-fighting exercises have been carried out. The results of the initial screening have been
disseminated along with results from other point sources in a publication by the Danish Environmental
Protection Agency.
Apart from the further investigations around fire-fighting areas and former waste dumps, the EIO have now
implemented an extended project plan to identify and document other activities with products containing
PFAS used by the Danish Defence.
The main focus for these investigations is to identify the threat to groundwater reservoirs.
Method
The first phase of the project involves interviews with personnel working for the Danish Defence and Danish
Emergency Management Agency. The objective is to attempt to identify products that might have contained
PFAS such as cleaning agents, impregnation agents (water repellent agents), paints, hydraulic oil as well as
fire-fighting foams. It is the content of PFAS in the final product, the potential flux to soil and groundwater
and the risk for migration in the environment which is of interest.
Furthermore, activities whereby these products or activities might cause soil or groundwater contamination
i.e. spill, waste water, storage, general handling, decanting etc., are identified and located at Defence
Establishments.
Based on this information activities and products which might have caused pollution with PFAS have been
identified and investigations initiated to screen for possible pollution of groundwater. Both water and soil
samples are analysed for a selection of at least 15 PFAS components.
The second phase of the project has been implemented and involves supplementary investigations to
delineate soil and groundwater contamination with PFAS.
Results
The initial groundwater screening at Danish military air bases has demonstrated that PFAS substances are
present in the groundwater at some, but not all sites, where firefighting activities have occurred. The level of
pollution ranges from <10 ng/l to 7500 ng/l for sum of 9 PFAS. In contrast to investigations at some civil air
ports or air bases in other countries, PFOS and PFOA have not been the dominating components indicating
that the fire-fighting foams have been produced by a manufacturer with a different formulation. The C-6 and
C-7 perfluorinated carboxylic acids (perfluorohexanoic acid - PFHxA and perfluoroheptanoic acid - PFHpA)
have dominated in the initial screening and for this reason EIO has decided to analyse samples for a broader
range of PFAS.
The newest results indicate that additionally a C-4 and C-5 perfluorinated carboxylic acids as well as 6:2 FTS
(6:2 fluorotelomer sulfonate) are present in groundwater samples at Danish Defence Air bases.
More results from the on-going investigations are presented in the lecture at AquaConSoil.
At present, Denmark has no quality criterion for PFAS in groundwater, drinking water or soil. Results are
often compared to the German criterion for drinking water of 100 ng/l for the sum of PFOS and PFOA.
The results provide an estimation of groundwater loads around Danish Ministry of Defence properties and an
indication of the PFAS composition in groundwater and soil samples. Determination of the PFAS
composition is an important contribution with respect to the future definition of a comprehensive Danish
quality criterion which should include both individual and sum values for relevant PFAS.
Furthermore, it would be reasonable to expect that the PFAS profile (composition) is dependent on the
source and therefore of great interest for future identification of point sources of pollution.
AquaConSoil Copenhagen 2015
INTRODUCTION
Enhanced reductive dechlorination (ERD) is frequently used to remediate sites contaminated with
chlorinated hydrocarbons (CHCs). Many experiences have shown that the success probability of the
remediation is strongly dependent on the distribution of the CHCs within the aquifer. Dense Non1Aqueous
Phase Liquids (DNAPL) can serve as a long term reservoir for the continued dissolution of CHC to
groundwater.
Furthermore, the injection of electron donors into groundwater in order to stimulate reductive
dechlorination can further contribute to the dissolution of the DNAPL droplets, ganglia, and pools (e.g.
beneficial effects of formed microbial surfactants acting as cosolvents [7]). Since these processes are
comparatively slow, the success of ERD strongly depends on the occurrence and mass of such DNAPL
bodies. Different investigation methods exist to assess residual DNAPL in an aquifer, but none of them
allows the determination of all of the characteristics (e.g. location, quantity, pore saturation) at the same
time in a reliable way and all of them have pitfalls [4]. A site investigated by ARCADIS in northern
Germany is a practical example of how a so called multiple1lines of evidence approach can lead to a
better understanding of the site and allows the preparation of a reliable conceptual site model. Different
investigation methods were used at this site, including a combination of Membrane Interface Probe (MIP)
measurements and Partitioning Tracer Tests (PTT) in the laboratory and in the field.
AquaConSoil Copenhagen 2015
The contamination at the site was caused by high volumes of CHCs, released by a dry cleaning facility on
the northern side of an inner1city water channel. The CHCs are present both in the shallow unconfined
aquifer and in a second deeper confined aquifer, separated from the first one by several meters of glacial
till. Concentrations of CHCs up to 150 mg/L are present in the deeper groundwater, and the plume
extends several hundred meters down1gradient of the source area. To explain the observed distribution of
CHCs, the DNAPL must have penetrated through the glacial till to the deeper aquifer, presumably via
fissures in the clayey till, and subsequently have been transported (partly as free product phase) down1
gradient by the groundwater flow.
The dry cleaner facility is currently not operational and the site is being redeveloped. The building formerly
overlying the source zone has been demolished. The first phase of planned remediation was to
hydraulically isolate the first aquifer (via a concrete pile wall) and then remove the soil and shallow
groundwater within the source, and replace it with clean material, preparing the site for redevelopment. In
the weeks before the beginning of the construction works, a second remediation phase was planned
comprising the injection of a vegetable oil emulsion into the deeper aquifer in order to initiate ERD. The
time available for the investigation and the injection was very limited, posing a further challenge in the
remediation of the site.
Membrane Interface Probing (MIP) - To assess the feasibility of ERD, the deeper aquifer was investigated
using a MIP direct push tool to log the relative concentration of CHCs. In order to investigate the second
aquifer, sixteen MIP probings were conducted (3 up1gradient, 9 within the source area, and 4 down1
gradient on the opposite side of the channel). The MIP probing was performed in two steps. First the
upper aquifer was investigated. Then, before penetrating the aquitard to reach the deeper aquifer, a
protection casing was installed to isolate the shallow aquifer and avoid cross contamination and the
investigation of the second aquifer was performed through this protection casing. The MIP was calibrated
before each boring by means of flash tests. Since the CHC1sensitive sensor DELCD is desensitized by
high contaminant concentrations, the detector had to be changed several times during the investigation.
Figure 1: Location of the MIP borings (DP 1-16) and of the source area (highlighted in red).
AquaConSoil Copenhagen 2015
Sealed-Liner and Groundwater Sampling - In order to calibrate the MIP signals, soil and groundwater
samples were taken at discreet locations applying also direct push drilling equipment (Macro1Core).
Within the most contaminated area of the deeper aquifer, 2 liner samples were taken; in addition to these,
a sample was taken in the aquitard (glacial till), where the MIP results suggested the presence of a highly
contaminated layer.
The undisturbed soil cores were taken at first with a polytetrafluoroethylene liner in order to minimize the
contaminant sorption to the sampling material. These liners were unfortunately too soft for the site
sampling conditions and got severely damaged. PVC liners were therefore used for further sampling.
A Geoprobe Systems SP116 was used to perform the groundwater sampling close to the location of an
MIP point. The screen point was pushed to the maximum sampling depth (27,5 m bgl) and then
successively pulled to the next two sampling points. The sampling depths were defined based on the MIP1
results in order to obtain samples with a low, medium and high CHC concentration. At each sampling
depth a low1flow (0,511,5 l/min) sampling was performed. The field parameters (redox potential, dissolved
oxygen, pH, temperature, electrical conductivity) were measured every two minutes and the sample was
taken after 14 minutes, as soon as the sampling piping had been purged.
In-situ Tests A push1pull1partitioning tracer test (PTT) was conducted in the zone of highest
contamination (based on the MIP investigation findings). A new well was built to perform this test,
screened only for a length of 1 m at the depth where the MIP1results indicated a high contamination.
The PPT was carried out as push1pull1PPT. In order to increase the solubility of the tracers, these were
dissolved in methanol first by a 1:1 ratio before being diluted in the injection water. The tracers
concentrations were chosen in order to obtain measurable tracer concentrations at the end of the test.
The following tracers were diluted in 500 l drinking water at the concentrations indicated below:
Conservative tracers:
o 21Propanol 75 mg/l
o n1Butanol 100 mg/l
Partitioning tracers:
o n1Hexanol 175 mg/l
o n1Heptanol 100 mg/l
o 21Ethyl111hexanol 100 mg/l
To assure an optimal dissolution of the tracers in water, optical and analytical checks were made and the
tracers solution was mixed for over one hour before injection. Then it was injected and extracted again at
a rate of 2,5 l/min with a 2 hours pause (so called shut1in period) between the push (injection) and the
pull (extraction) phases. The tracers and CHC's concentrations were measured in the mixing tank before
the beginning of the test and in the water extracted in the pull phase every 15 minutes. The analytical
test results were interpreted according to [5] and [6].
Laboratory Tests The partitioning tracer tests (PTTs) were also repeated in laboratory. After shortening
the liner cores at the extremities to a length of 40 cm to eliminate the most disturbed sample portions, they
were used in laboratory as soil columns for the PTTs. The same tracers used for the in1situ push1pull1PPT,
with the addition of two further reactive tracers (2,31dimethyl121butanol (DMB) and 2,41dimethyl131pentanol
(DMP) ), were applied in the laboratory column test in the following concentrations:
Conservative tracers:
o 21Propanol 800 mg/l
o n1Butanol 800 mg/l
AquaConSoil Copenhagen 2015
Partitioning tracers:
o Dimethyl121butanol (DMB) 100 mg/l
o 2,41dimethyl131pentanol (DMP) 100 mg/l
o n1Hexanol 800 mg/l
o n1Heptanol 350 mg/l
o 21Ethyl111hexanol 100 mg/l
Respectively 19,3 ml and 20,2 ml (corresponding approximately to the estimated pore volume) of tracers
solution were infiltrated in the columns and leached for 100 h with 886 ml and 924 ml water to a achieve
minimum 4,51fold pore volume exchange. The leachate was collected and samples were taken on a
regular basis.
The same soil used for the PPT was used in two small columns to determine the maximum oil retention
capacity by infiltrating a rapeseed oil emulsion (12 % in water) (Newmann Standard Formulation 1901
6722) for three hours and afterwards tap water again for three hours at the same rate to flush non1
retarded oil. The TOC of the soil was determined before the beginning and after the end of the test on
homogenized samples. The results were used to calculate the oil retention capacity of the two samples.
Moreover 18 soil samples were taken during drilling of the test well used for the in situ tests and stored in
headspace vials and overlaid with methanol. The samples were analyzed for CHC concentration. Four soil
samples were also taken from the discarded parts of the liner samples and analyzed to determine the
concentrations of CHC, iron, manganese (through sequential extraction), TOC, TIC and acid neutralization
capacity.
Furthermore a diffusion test was carried out on the soil sample taken from the contaminated glacial till and
Oil Red O shake tests were conducted on two glacial till and one deep aquifer samples. For the diffusion
test a contaminated sample from the glacial till (soil rod) was installed in a column, surrounded by clean
sand with the same grain size characteristics of the deeper aquifer. The column was leached for 10 days
in a stop1go mode (24 h leaching, 24 h stop) to achieve a 101fold exchange of the pore volume with
uncontaminated groundwater. The leachate was collected on a regular basis and analyzed.
In shake tests, 2 5 g of soil, a very small amount of the hydrophobic dye Oil Red O and 15 ml deionized
water were shaken in a 50 ml centrifuge vial. If CHC phase is present, the sample turns red, while no color
changes are recognizable in a sample if no phase is present (the dye remains undissolved) [3].
RESULTS
The values detected by MIP1DELCD (Dry Electrolytic Conductivity Detector), PID (Photo Ionization
Detector), FID (Flame Ionization Detector), EC (Electric Conductivity), as well as the probe temperature
and velocity were graphically recorded as a function of the depth and automatically interpreted. While the
DELCD only reacts to CHC, the other two detectors are sensible to other compounds, but to a different
extent. The MIP locations relative to the source area are shown in Figure 1 and the MIP1DELCD results
are summarized in Figure 2.
AquaConSoil Copenhagen 2015
10
-5
Height [m AMSL)
-10
-15
DP -20
DP 05
07 DP
DP DP
01 DP DP
02 11
03 04
DP DP -25
08 10
DP DP DP DP
13 09 DP DP DP 15 16
14 06 12
-30
5940186,0 5940166,0 5940146,0 5940126,0 5940106,0 5940086,0 5940066,0
Northing (Gauss-Krger coordinates)
<1 V 1 V-2 V 2V-3V 3 V-4 V > 4V Top deeper aquifer Ground level
Figure 2 DELCD signal (green: DELCD-Signal < 1 V, yellow: DELCD Signal 1 - 2 V, orange: DELCD-Signal 2 - 3 V, red:
DELCD-Signal 3 - 4 V, purple: DELCD-Signal > 4 V; the blue dash shows the location of the aquitard bottom,
representing the top of the deep aquifer) in the direction North to South, approximately corresponding to
the groundwater flow direction.
The comparison of the DELCD signal with the results of the CHC analyses conducted on the soil samples
shows a very good correlation (Figure 2), especially in cohesive soil like the glacial till. The geological
profiling received from the EC1probe data showed that the contaminants accumulated above less
permeable layers in the aquifer (e.g. slightly clayey layer 10,1 1 12,6 m bgl) or in layers with a high organic
material content and therefore a high adsorption capacity (e.g. 13,5 1 15,5 m bgl), which have delayed the
DNAPL migration. Only the peaks identified by the MIP at approx. 17 m and 19 m bgl were not detected
through the traditional soil sampling. This discrepancy is probably due to the lower resolution and
precision of the soil sampling with the drilling technology used on site within the sandy layer (mud auger).
AquaConSoil Copenhagen 2015
DELCD-Signal [V]
Concentration [mg/kg TM]
1 10 100 1.000 10.000 100.000
0
6
Depth [m bgl]
10
12
14
16
CHC
18 PCE
TCE
20
Figure 3: Boring log, graphic representation of the vertical contaminant distribution in the test well and MIP-log in
the closest direct-push investigation point.
Oppositely, the groundwater concentration in the samples taken with the direct push technology did not
correlate well to the results of the MIP1detection. This is probably due to sampling method used: while in
the penetration phase the screen point is sealed by the screen sheath, forming a mechanical annular seal
above the screen interval. When the screen point is pulled up, the borehole can become a preferential
pathway and allow a short circuit with the lower portion of the aquifer.
The breakthrough curves and the recovery curve for the in situ push1pull PTT are shown in Figure 4. The
tracers concentrations are expressed as ratio between the measured concentration (C) at the sampling
time and the start concentration (C0) as a function of the ratio between the extracted groundwater volume
(Vext) at the sampling time and the total injected volume (Vinj).
AquaConSoil Copenhagen 2015
140%
Tracer breakthrough curves
120%
100%
80%
C/C0
60%
Cumulative tracer recovery
40%
20%
0%
0,0 0,5 1,0 1,5 2,0
V extr /V inj
Propanol12 Butanol11
Hexanol11 Heptanol11
21Ethyl111hexanol Cum. recovery Propanol12
Cum. recovery Butanol11 Cum. recovery Hexanol11
Cum. recovery Heptanol11 Cum. recovery 21Ethyl111hexanol
The strong oscillation of the concentrations measured in the first test phase is due to the low analytic
precision (error up to 20 % caused by the high dilution grade of the highly concentrated samples). The
water volume measurement system (water meter) constitutes a further imprecision source in the test (error
up to 10 %). In ideal conditions the inflection point of the breakthrough curve should correspond to a
Vext/Vinj = 1. If there is a retardation, the curves of the reactive tracers are flatter. In real ambient
conditions (e.g. non1linear or non1instantaneous adsorption, strong diffusion processes) the inflection point
can be reached by Vext/Vinj values lower than 1, as in this case.
Moreover the recovery curves show that part of the injected solution (approx. 20 %) could not be
extracted. This loss is probably due to hydraulic effects (losses especially at the upper and lower
extremities of the screen, due to the relatively short screen in thick highly conductive aquifer) can be
reduced by injecting higher volumes for the test. By correcting the breakthrough curves taking into
account the estimated volume loss, the inflection point shifts to approximately 1.
In order to calculate the retardation (R) and the average DNAPL saturation in the interrogated portion of
the aquifer (S), the test data have to be statistically elaborated. Schroth et al. [6] developed a method to
estimate the retardation factor using the comparison between the breakthrough curve of a conservative
tracer and that of a reactive tracer. The fundamental assumption is that the longitudinal dispersivity ( ) L
under ideal transport conditions is a property of the aquifer and therefore it should be identical for a
AquaConSoil Copenhagen 2015
reactive and a conservative tracer. Thus, differences in the shape of their breakthrough curves may be
attributed to retardation of the reactive tracer in comparison to the conservative one.
By applying the simplified method for estimating retardation factors from push1pull tests as described in [6]
to compare the breakthrough curves of butanol to those of the partitioning tracers and assuming a porosity
of 20 %, the retardation factors for the tested partitioning tracers could be calculated. The residual
saturation could then be calculated as follows [5]:
(1)
where R is the retardation factor and K is the tracers partitioning coefficient (NAPL/water). For this
project the weighted mean of the partition coefficient PCE/water TCE/water (based on PCE/TCE ratio on
site) was considered. The results of the statistical elaboration to compare the breakthrough curves of
butanol with those of the partitioning tracers are summarized in Table 1.
The statistical elaboration with propanol as a conservative tracer delivered implausible results, probably
because of measurement errors. With butanol as a conservative tracer, an average NAPL saturation of
0,3 3,2 % of the pore volume (preliminary values) was estimated. In order to evaluate the significance
and reliability of these values, the breakthrough curves of the two conservative tracers were reciprocally
compared according to the same method used for the partitioning tracers and a retardation factor of
1,6 % could be calculated. This indicates that any calculated retardation below this value does not
certainly imply the existence of residual phase. The results of the test cannot therefore demonstrate nor
quantify the NAPL saturation in this case with certainty.
The same assumptions and residual saturation calculation methods applied for the in1situ PTT were used
to interpret the results of the partitioning tracer column tests in the laboratory and led to the calculation of
average NAPL saturations of maximum 0,6 %. Considering the measurement inaccuracy, this value
cannot be considered as a final proof of the existence of free phase in the investigated aquifer portion. As
expected, the tests showed that the results of the tests conducted with strongly partitioning tracers have a
lower variance. This confirmed that by low saturation values the accuracy of the PTT can be increased by
using reactive tracers with a high partitioning coefficient.
The oil retention capacity calculated on a medium to fine sand sample was 5,4 g oil/kg soil and in the fine
sand sample 17,3 g oil/kg soil. These values are in the range of usual retention capacities determined in
laboratory tests for these particle sizes [1].
In the frame of the diffusion test relevant quantities of PCE and TCE (respectively 87 % and 43 % of the
initial concentrations) were leached. Although no quantitative information about the duration and extent of
the diffusion can be inferred from the test, it can be assumed that the CHC emissions will last for a long
time and cause high contaminant concentrations in the aquifer. If only CHCs dissolved in water reached
the deeper aquifer through the aquitard, they would be expected to contaminate a limited, upper aquifer
portion. However, since the Oil Red O shake tests showed that free phase was present in one of the
samples taken in the glacial till, the CHCs could have migrated as DNAPL to the deeper aquifer.
AquaConSoil Copenhagen 2015
CONCLUSIONS
The combination of the results of the different technologies applied allowed achieving a high definition
picture of the residual DNAPL in the subsurface:
The first aquifer and the aquitard separating it from the deeper aquifer are very contaminated. In
particular, the geometry of the contaminated area and the presence of DNAPL in some aquitard
portions (shown by dye shake tests) suggest the presence of fractures of similar preferential
pathways allowing the migration of the DNAPL.
As shown by the diffusion test, CHCs are expected to emit into the deeper aquifer for a long time
in relevant concentrations.
The contamination in the deeper aquifer at the moment involves only the southern part of the
source zone, therefore the remediation effort can be focused on this area.
The contamination in the deeper aquifer follows an anisotropic, meandering pattern, so that the
contamination migrates preferentially through highly contaminated ganglia, sinking along the
groundwater flow direction to deeper portions of the aquifer.
REFERENCES
[1] Borden, C., Clayton, M., Weispfenning, A. M., Simpkin, T., Liebermann, M.T.(2008) USERS
GUIDE Development of a Design Tool for Planning Aqueous Amendment Injection Systems,
ESTCP Project ER10626.
[2] Divine, C.E., McCray, J.E., Wolf Martin, L.M., Blanford, W.J., Blitzer, D.J., Brusseau, M.L., Boving,
T.B. (2004) Partitioning tracer tests as a remediation metric: Case study at Naval Amphibious
Base Little Creek, Virginia Beach, Virginia. Remediation, Spring 2004, 7131.
[3] EPA (2004) Site Characterization Technologies for DNAPL Investigations, Hazardous Waste
Cleanup Information (CLUIN) system web site at http:/cluin.org.
[4] ITRC (2000) Dense Non1Aqueous Phase Liquids (DNAPLs): Review of Emerging
Characterization and remediation Technologies (Technology overview, June 2000).
[5] Istok, J.D., Field, J.A., Schroth, M.H., Davis, B.M., Dwarakanath V. (2002) Single1well Push1Pull
partitioning tracer test for NAPL detection in the subsurface, Environmental Science Technology
36: 270812716.
[6] Schroth, M. H., Istok, J. D., Haggerty, R. (2001) In situ evaluation of solute retardation using
single1well push1pull tests, Advances in Water Resources 24, 1051117.
[7] Suthersan S. S., Payne F.C. (2004) In situ Remediation Engineering, CRC Press
Groundwater flow and transport modelling aimed at an exploratory spatial data
analysis of Ust-Kamenogorsk aquifer system, East Kazakhstan
#
Crestaz Ezio, Ancilletta Antonio , Isidori Stefano, Patata Leonardo, Pellegrini Michele, Tatangelo Felice
Saipem SpA (ENI Group), Via Toniolo, 1 61032 Fano (PU), Italy
#
Corresponding author; e-mail antonio.ancilletta@saipem.com
ABSTRACT
Current paper focuses on groundwater flow and transport modelling as a tool for conducting an Exploratory
Spatial Data Analysis (ESDA) of the highly permeable alluvial aquifer of the huge urban and industrial
settlement of the City of Ust-Kamenogorsk, located in eastern Kazakhstan near the confluence between two
main rivers. With a population of about 300,000 inhabitants, spanning through an area of approximately 35
2 2
km , and a 6 km industrial complex located immediately upstream of the city, the area has been impacted,
since WWII, by diffused metals and radioactive pollution.
As part of a larger remediation project on soils, surface water and groundwater in the area, national Kazakh
Institutions and World Bank co-funded specific actions towards assessment of groundwater quality status
and flow conditions, i.e. estimate of discharges and flow directions, as well as assessment of potential risks
at more sensitive areas as main city Oktyabrski water intake. Piezometric and quality monitoring networks
have been designed, new wells and piezometers drilled, pumping tests performed to assess hydrogeological
parameters, and periodic monitoring campaigns conducted over the period 2012-2013. Resulting data and
emerging interpretation evidences have been used as the basis for ESDA, based on a 3D groundwater flow
and transport finite element model. Such a model, implemented using DHI-WASY Feflow 6.1 numerical
code, supported uncertainty analysis, leading to the establishment of a Spatial Decision Support System
(SDSS) to assist ongoing and future remediation projects and policies.
Different alternative interpretative scenarios have been numerically investigated, leading to relevant
conclusions concerning: 1) confirmation of complex patterns in groundwater flow paths, mainly governed by
geometry of the underlying low permeable formations; 2) high groundwater flow velocity, particularly in the
main plain below eastern Ust-Kamenogorsk; and 3) strong interconnection between rivers and aquifer
system, upper Ulba river acting as a major recharge body and Irtysh river being the main drainage reference.
Given a complex pattern of ongoing contamination, with more soluble contaminants as Selenium migrating
towards the banks of Irtysh river and less mobile ones, as Cadmium, remaining trapped in the proximity of
main industrial complex, one relevant outcome has been that Oktyabrski water intake looks like relatively
protected. Being close to Irtysh river and entirely bounded by other surface water bodies, most of the
pumped water is leaking from such surface water system.
INTRODUCTION
The City of Ust-Kamenogorsk experienced a rapid expansion of mining and heavy industry during the Soviet
Era. Industry includes facilities related to metallurgical, high technology, machine building and energy
production activities, resulting in large quantities of hazardous waste and relevant local and regional impacts
on environmental status (Fig. 1).
It has long been suspected that contaminated groundwater in industrial areas is due to leakage of process
water and untreated water from old sewage networks and treatment plants, generally observed to be in poor
condition. Contamination within the Ust-Kamenogorsk area has been found to originate mainly from
leachates from metallurgical waste dumps and ash ponds, effluents from metallurgical plants and polluted
run-off from industrial plant areas.
A clear relationship has been documented between observed plumes of contaminated groundwater, mainly
dominated by metals, and the location of waste deposits (UHPP, UMP ash and sludge ponds), industrial
plants (Central Industrial Complex, CIC) and unsewered urban areas.
There is a strong hydraulic connection between the Irtysh and Ulba rivers and the underlying alluvial aquifer,
which, face to the multiple sources of contamination, leads to a challenge groundwater
protection/remediation framework.
Fig. 1 Hydrogeologic and hydrologic key features, monitoring network and reference
piezometry.
Most of the documentation from former studies, which have been used in the framework of current analysis,
is not explicitly referred to in text nor in references, due to confidentiality issues and permission constraints.
Current paper mainly refer to contents of unpublished Saipem reports (2011, 2012a, 2012b, 2012c).
HYDROGEOLOGY
Paleozoic basement rocks, mainly schists and coarse-grained intrusive rocks, outcrop W and NW of CIC
(Fig. 1). The limited outcrop, close to CIC about 20 m higher than neighboring areas, looks like spatially
correlated to the other outcrops about 2 km far away to NW. Basement continuity in depth reasonably acts
as, at least partial, groundwater divide, affecting both groundwater flow and contaminants migration.
Actually the Paleozoic basement quickly deepens to E, towards Ulba river, where it is overlain by coarse
sand and gravel quaternary deposits with maximum thickness of about 100 m. Neogenic clay overlays
quaternary deposits, with generally limited thickness in the order of few meters in the main plain and close to
rivers, up to few tens meters in the bounding hills, where UMP ash ponds, about 40 m higher than main
industrial complex, have been built. Apart from some minor silty-clay lens in upper part of coarse alluvial
deposits, finer sediments tend to increment in depth, being locally prevalent at aquifer bottom.
Main aquifer is characterized by high to very high hydraulic conductivities. The sedimentological framework
and the many evidences of (partially) abandoned meanders, as is the case for the Komendantka channel,
the pond in the central city metallurgical park and the water emergence in narrow elongated alluvial plain
located at W of the isolated crystalline outcrop just near the CIC, strengthen the hypothesis of relevant
variability in hydrogeological properties spatial patterns (Fig. 1).
Piezometric interpretation (Fig. 1) highlights main groundwater flow directions, being controlled by surface
waters of both main rivers and partially abandoned meanders. As Ulba river has much higher gradients than
Irtysh in the study area, roughly flowing in a SW direction, it is also responsible of most groundwater
recharge. Groundwater flows towards SW and WSW within the eastern part of the main plain, aquifer
emerging at Komendantka and at the many abandoned meanders in the area. In the western sector,
groundwater flow directions sharply modify, being prevalent a W-WNW direction, parallel to end section of
Brajinskij channel. The presence of abandoned meanders in the area suggest at least groundwater flow
components occurring parallel to Irtysh river. Particular attention must be devoted to higher sector of the
study area, immediately N of CIC, where local recharge is mainly related to UMP sludge ponds and
groundwater flow directions are affected, as previously noted, by crystalline outcrops.
-4 -3
Hydraulic conductivities, varying in the range of 2.3 10 m/s up to 2.1 10 m/s, with a log-normal
-4
average of 9.0 10 m/s, have been reported in former studies. The existing groundwater monitoring
network has been integrated by construction of 65 new piezometers, further to 6 wells, conceived for
execution of pumping and infiltration tests. Resulting hydraulic conductivities have been in the order of 2.5
-3
10 m/s in the main alluvial plain, higher than former estimates although still of the same order of magnitude.
-4
Hydraulic conductivity in the order of 4 10 m/s was estimated for the sector NW of CIC. Hydraulic
conductivities spatial distribution has been estimated, based upon new hydraulic tests and stratigraphic data.
Estimate of vertical recharge include effective infiltration due to rainfall and snow melting at spring time, and
leakage from human made infrastructures. The former component has been tentatively estimated in previous
studies, ranging from dominating evapotranspiration and null effective infiltration, up to values of about 200
mm/yr infiltration and 300 mm/yr evapotranspiration. Low effective infiltration estimates would find
justification in limited local rainfall rates, further to climate conditions, dominated by warm/dry summers and
very cold winters, causing soils freezing and consequent cut of direct recharge. The latter components
include sewage network and industrial channels, most of them being buried but few clearly visible from aerial
photos. A previous study focused on large scale leakage estimate resulting in an overall value of 7,050,000
m3/year over approximately 56.5 km2, equivalent to a recharge rate of approximately 125 mm/yr. It further
provided spatially disaggregated estimates of local leakage from within the CIC, the largest estimated
component being related to the heating thermal plant in western sector.
Specific considerations on recharge estimate have been demanded to calibration stage, including
comparison between direct infiltration and estimated in/out flows along boundary conditions.
Hydrochemical data collected during four quarterly monitoring campaigns confirmed the general groundwater
quality framework discussed in previous studies, yet supporting a more detailed conceptualization of
contaminants distribution and/or plumes, S and W of CIC.
Among Metals, most of them such as Cadmium, remain trapped close to the CIC area (
Fig. 3), while few others, as Selenium, are widespread in the area of investigation (
Fig.4). General inorganic parameters such as Nitrates and Sulphates represent the most persistent problem
for groundwater. Radioactivity analyses highlight large contamination by Alpha-activity, with concentrations
apparently increasing with depth.
Based on inferred plumes of most relevant and mobile contaminants, and a ranking on reliability of such
plumes, weighted probability raster analysis lead to the computation of an hydraulic conductivity index (Fig.
1). Such an index has been used in the framework of estimation of hydraulic conductivity spatial distribution
as a guidance, properly combined with other stratigraphic evidences and sedimentological considerations.
Few groundwater pathways extending from SE to NW, parallel to Irtysh river, have then found further
confirmation on the identification of abandoned meanders and minor streams (Fig. 2; rif. to Fig. 1 for plane
location).
Fig. 2 Snapshots from Google Earth aereal photos of upper part of Komendantka channel and
Metallurgical park
Modelconceptualization
Former 2D and 3D flow and fate and transport (F&T) models have been refined, in the framework of a post-
auditing activity based on evidences from new drillings, piezometric and quality monitoring networks,
interpretation of pumping tests and identification of paleo-rivers. All models have been implemented using
the state-of-the-art DHI-WASY finite element code/environment Feflow (Diersch, 2006).
Formerly interpreted groundwater system geometry has been confirmed by the new stratigraphic data. The
groundwater divide, extending towards N from the local crystalline basement outcrop W of CIC, clearly
disconnects main flood plain, below CIC and Ust-Kamenogorsk downtown, from minor aquifers located NW
of the divide. Pumping tests interpretation outcomes confirmed a stratigraphic transition, within main plane,
from highly permeable deposits S of CIC to less permeable finer sediments in the western sector of CIC
(calibrated values in the range 1.5 10-3 5 10-4 m/s), further to a transition to finer deep deposits in main
plane.
Lower hydraulic conductivity has been assumed in the S, close to Irtysh river, consistently with information
provided by Committee of Water Resources of the Ministry of Environment and Water Resources of the
Republic of Kazakhstan (CWR). Extended ponds, possibly causing finer sedimentation, observed
drawdowns at Oktyabrski well field higher than expected, and no impact of more soluble contaminants
whose preferential directions are reasonably controlled by upstream paleo-rivers, provide supporting
evidences to such an assumption.
discretization mesh, with 2D refinements below CIC, along main rivers and Komendantka channel,
and a 3D detail of vertical discretization and groundwater divide location W of CIC (Fig. 5);
hydrogeological properties distribution showing spatial variations in both 2D and 3D (
Fig. 6);
boundary conditions (BCs,
Fig. 7), of I type (Dirichlet) along upper inflow and low outflow limits, and of III type (Cauchy) along
Ulba and Irtysh rivers, and at internal abandoned, yet wet, meanders as Komendantka channel. High
riverbed vertical hydraulic conductivity of the latter BCs, consistent with the coarse deposits,
definitely converted, from a practical standpoint, III Type BCs to Dirichlet BCs. Surface water
elevation control groundwater flow, upper part of Ulba mainly recharging the groundwater system,
and lower Ulba and Irtysh draining it. Locally Irtysh and associated ponds leak to groundwater city
supply well system of Oktyabrski.
Fig. 5 - 3D mesh discretization and detail section showing vertical discretization and groundwater divide located S of
CIC in a view from S30W (vertical exageration 20:1)
Fig. 6 - 3D mesh discretization and hydrogeological properties distribution in a view from S30W
(vertical exageration 20:1)
Fig. 7 - Boundary conditions in a view from S30W and 2D infiltration (vertical exageration 20:1)
Overall recharge has been estimated taking into account both effective infiltration and leakage from human
made infrastructures over Ust-Kamenogorsk urban area and CIC. Based on bibliographic estimates,
although affected by a large variability (Hrkal et al., 2001), and by a former study on leakage from urban
infrastructures, an overall mean recharge value of 183 mm/yr has been used. Recharge has been spatially
disaggregated based on prevalent lithology close to ground surface, ranging from a minimum of 50 mm/yr,
close to Irtysh ponds and NW of CIC, up to 200 mm/yr in the coarse sediments main plain. Local higher
values of 383 mm/yr has been finally set in the SE sector of CIC, based on the evidence of presence of
surface water channels and pools probably used for cooling purposes. Provided that above recharge
estimates are cautelative, in the upper limit of the variability range, no further effort has been spent at this
stage on vertical recharge analysis, in consideration of the key contributions of rivers to groundwater flow, as
clearly highlighted by model fluid mass balance (Cauchy vs. Dirichlet BCs in Tab. 1).
Model calibration
Piezometry is strongly controlled by surface water bodies. So it is a relatively poor calibration indicator on its
own, particularly for validation of interpreted hydraulic conductivities and true velocities spatial distribution.
Contaminants plumes provide some guidance towards assessment of groundwater flow directions, but still
poor information about true groundwater flow velocities. Highly soluble contaminants, as Se (
Fig.4), moved far away since early stages of the contamination history in 1940ies, and still their partition
coefficients Kd can be highly variable depending upon local conditions (i.e. relevance of finer sediments on
adsorption phenomena, Wang and Chen, 1997; Appelo and Postma, 2007). On the other hand low mobility
metals, as Cd (
Fig. 3), remained sorbed in the close proximity of CIC.
K spatial distribution has been inferred based on hydraulic tests interpretation, stratigraphic data and
evidences of preferential pathways resulting from expert judgement of contaminants distribution (hydraulic
conductivity index in Fig. 1). Observed vs. computed piezometric data comparison and statistical
convergence parameters (Fig. 8; Tab. 1) document steady-state model calibration at current date, 70 years
after simulation start in the early 1940ies, at the time CIC was built.
Computed contaminants spatial distribution at calibration is consistent with observed patterns, clearly
highlighting the different behavior of high and low solubility compounds. Se and Cd computed spatial
distributions at current date are definitely not so different from simulated patterns after further 30 years in
absence of environmental remediation and groundwater containment measures. Refer to simulation in next
paragraph for further details.
Fig. 8 - Computed vs observed piezometric data at calibration for current date (70 years after CIC
construction)
Statistical convergence
parameters
Comp-Obs -0.04
|Comp-Obs| 0.36
(Comp-Obs)2 0.25
Tab. 1 - Fluid mass balance and statistical convergence parameters at calibration for current date
(70 years after CIC construction)
Simulationscenarios
Based upon calibrated model, various long term scenarios have been designed and simulated, focusing on
30 and 130 years from current date. Few of them are briefly discussed here below.
A first scenario refers to the absence of any containment/remediation action, without removal of
contamination source at CIC. Groundwater steady-state flow field remains unmodified (Fig. 9). Actually the
Oktyabrskij well field, in the proximity of Ulba river, is impacted by groundwater flowing in the eastern part of
CIC. Local scale water balance analysis highlights that most of pumped water is drained from surface water
bodies, including both Ulba and ponds connecting the river to the abandoned meander of Komendantka
channel (Fig. 1); so groundwater contamination N of the well field does not affect pumped water quality.
Simulated Cd and Se concentrations (Fig. 10) at 30 years is similar to current conditions, highlighting the
different behavior of the two contaminants and the impact of persistent sources of contamination in CIC area.
Cd plume remains confined SW of CIC, while Se strongly extends westward, resulting in a widespread
contamination over most of the domain downstream of CIC.
Fig. 9 - Simulation scenario in absence of any remediation/containment action: steady state
piezometry and computed pathlines at model and local (between CIC and Octybriski) scales
A second scenario refers to the activation of a pumping system located S of CIC, aimed at full containment of
3
groundwater flow. Overall estimated pumping rate is 1260 m /h, unevenly distributed in space, with higher
rates in the eastern sector to account for the leakage to groundwater system from neighboring Ulba river.
Piezometry and streamtraces (Fig. 11) also highlight the relatively slow transient behavior of groundwater
system towards steady state conditions and full groundwater containment.
30 days 1 year
Fig. 11 - Simulation scenario with 14 wells hydraulic barrier S of CIC at an overall pumping regime of 1260 m3/h (72 m3/h
3
on western 7 wells and 108 m /h on eastern 7 wells): piezometry and groundwater pathlines at different times since
barrier activation (in red undisturbed piezometry, in blu transient piezometry affected by pumping
ACKNOWLEDGEMENTS
We acknowledge here the many valuable contributions from reports and modeling activities implemented by
former consultancies, which, for confidentiality issues, we could not explicitly reference. We particularly
acknowledge the many colleagues, who provided helpful and valuable support, among them the Project
Director Greganti Savio, the Project Manager, Scipioni Piero, and the former Project Technical Manager
Tabani Luciano. We would also like to thank CWR and World Bank representatives for the many valuable
discussions and helpful guidelines.
CONCLUSIONS
A Feflow groundwater flow and transport model of the aquifer system of Ust-Kamenogorsk urban area, in far
east Kazakhstan downstream of a huge industrial site (CIC) and related sludge ponds, has been presented.
Model scopes included the exploratory spatial data analysis of quantitative and qualitative framework, as
delineated from legacy data and outcomes of new bore holing activities and monitoring campaigns, and the
assessment of best policies towards contaminated groundwater containment.
The identification of a groundwater divide west of CIC, related to crystalline basement outcrops elongated in
a NS direction, permitted to discriminate between two major hydrogeological domains, at least partially
disconnected. The eastern part, hugely impacted by contamination from CIC, has been conceptualized,
thickness and hydraulic conductivity sensibly increasing towards Ulba river. Paleorivers (Komendantka
channel, Metallurgical park and lower section of Brajinskij channel) clearly identified from aereal images in
the lower plain, indirect evidences from more soluble contaminants plumes spatial patterns, stratigraphic
information and sedimentological interpretation all supported the working hypothesis that preferential higher
hydraulic conductivity would occur parallel to Irtysh river, and that finer sediments would be present in the
close proximity to the river.
Focusing on main objectives of protecting Irtysh river and the Oktyabrski city water intake, S of CIC and Ust-
Kamenogorsk downtown, potentially impacted by contamination originating from the industrial complex,
following key model outcomes have emerged:
full groundwater flow containment at a wells barrier S of CIC would demand for a very high overall
pumping rate, estimated in 1260 m3/h;
pumping should be unevenly distributed, with higher rates in eastern part of the wells barrier, due to
the higher hydraulic conductivities in the main plane and the proximity to Ulba river. Pumping would
result in a strong impact on groundwater flow directions and increasing inflow from Ulba river. A
relatively long transient period would be required before hydraulic barrier effectiveness would be
guaranteed;
high pumping rates at Oktyabrski city intake induce large leaking inflow from bounding surface water,
namely Irtysh river and ponds, connecting to Komendantka channel, eastward. This is probably the
main reason why no significant contamination is reported at the wells field, according to information
provided by local Institutions, although in current conditions the well field is potentially affected by
groundwater pathlines flowing below eastern part of the CIC.
Simulation scenarios focused on the hypothesis that no contamination removal would occur at CIC and no
containment measures would be put in place immediately downstream, highlight that overall contamination
spatial patterns would not sensibly modify, even in the medium-long term, up to 30 years from now.
Assessment of technical and economic implications of groundwater containment policies, as well as of
alternative approaches, as partial groundwater containment aimed anyway at impacting on groundwater flow
directions in the lower alluvial plane, is currently ongoing.
REFERENCES
Anderson, M.P. and Woessner W.W., 1992. Applied Groundwater Modeling: Simulation of Flow and
Advective Transport. Academic Press Inc., San Diego CA, USA
Appelo C.A.J. and Postma D., 2007. Geochemistry, groundwater and pollution. 2nd Ed., A.A. Balkema
Publishers/Taylor & Francis Group
Diersch H.J.G., 2006. FEFLOW, Finite Element Subsurface Flow & Transport Simulation System, White
Papers Vol. I. Wasy, Institute for Water Resources Planning and Systems Research Ltd., Berlin, Germany
Hrkal Z., Gadalia A. and Jucker C., 2001. Contamination of groundwaters by heavy metals in the City of Ust
Kamenogorsk, north-Eastern Kazakhstan. Environmental Geology (2001) 41: 174-182
SAIPEM S.p.A., 2011. Design of Monitoring Well System: Deep and Multilevel Monitoring Wells, SPC-00-
BH-E-94004 Rev.0 Contract No. UKER-DD2 Loan 4848, October 2011, unpublished report
SAIPEM S.p.A., 2012a. Groundwater Monitoring and Design Contract, Inception Report, SPC-00-BE-E-
94005 Rev.1 Contract No. UKER-DD2 Loan 4848, February 2012, unpublished report
SAIPEM S.p.A., 2012b. Technology Screening for Groundwater Treatment System, SPC-00-BE-E-96001
Rev.0 Contract No. UKER-DD2 Loan 4848, September 2012, unpublished report
SAIPEM S.p.A., 2012c. Groundwater Monitoring and Design Contract, Report about installation of the
underground water monitoring system, SPC-00-BE-E-94030 Rev.2 Contract No. UKER-DD2 Loan 4848,
November 2012, unpublished report
Wang F. and Chen J., 1997. Modeling Sorption of Trace Metals on Natural Sediments by Surface
Complexation Model. Environ. Sci. Technol., 31, pp. 448-453
ABSTRACT
A three-dimensional groundwater flow and transport model was developed for a contaminated
industrial site in Belgium in order to evaluate the currently ongoing remedial activities and to develop
a hydrogeological barrier to prevent further of site spreading of polluted groundwater. The
groundwater model was used to predict the likely future contaminant distribution under different
groundwater abstraction scenarios.
The site is contaminated with among others CVOCs and trichloroethylene (TCE) was characterized
as the parent product and cis-1,2-dichloroethelyne (DCE) and Vinylchloride (VC) as the degradation
products. The groundwater contamination with chlorinated solvents is present until a depth of 11 m
below surface where a poorly permeable clay layer is present. To stop the migration of contamination
from the industrial site to neighboring parcels a reactive wall (funnel & gate principle) with three gates
was installed in 2001. Additionally two source areas, suspected to contain pure product DNAPL, are
pumped with vacuum enhanced recovery (VER) technique since 2001. In 2006 a third VER extraction
was installed below the adjacent street to tackle pollution that had in the past already spread beyond
the location of the reactive wall before its installation.
Recent analytical groundwater results showed the presence of additional source zones of CVOCs
which result in migration of impacted groundwater in a northwesterly direction. No reactive barrier is
currently present to mitigate this migration risk and it is not technically feasible to extent the existing
funnel & gate barrier into this area. To prevent further off-site spreading a hydrogeological barrier with
extraction wells was proposed.
The groundwater flow model was developed to evaluate the feasibility of a hydrogeological barrier
and to design the barrier taking into account stability constraints, the influence on the groundwater
flow near the reactive barrier and a desirable low extraction rate to minimize the treatment costs of
extracted impacted groundwater. The hydrogeological barrier needs to be designed in such a way
that no negative impact is expected on the surrounding remediation infrastructure and on other
contaminations present at the site. Additionally, the model should be able to evaluate the existing
remedial infrastructure at the site and to simulate future remediation options.
The 3D groundwater flow and solute transport model was developed using the finite-difference model
codes MODFLOW 2000, MODPATH and MT3D, using the Groundwater Vistas (version 6) interface.
A regional groundwater flow model was develop and calibrated using historical groundwater level data
from the period before remedial infrastructure was present. The regional model has an extension of
2
ca 19 km . The model represents the phreatic aquifer until a depth of 11 m below surface.
Telescopic Mesh Refinement was used to develop a local, more detailed, groundwater flow model for
the area of the contaminated industrial site from the calibrated regional groundwater flow model. The
local model contains 3 layers in accordance with observations at the site: an upper sandy layer, a
middle sandy clay layer and a lower weakly sandy clay layer. All the remediation infrastructure
present at the site was incorporated in the local groundwater flow model. Recent groundwater level
data was used to calibrate the local model, which was further used for dimensioning the
hydrogeological barrier and which was extended into a solute transport model to predict the further
spreading and evolution of the groundwater contamination in the future.
The calibrated groundwater flow model was then used to optimize the properties of the
hydrogeological barrier taking into account the site specific constraints, i.e. minimal lowering of water
table and minimal volume of extracted water for stability reasons, minimal treatment costs and not
influencing the groundwater flow pattern near the funnel & gate infrastructure. This resulted in a
proposed pumping well configuration of 11 extraction wells extracting groundwater from different
3
depths with a total pumping rate of 5,4 m /day.
The contaminant transport model was also used to obtain a better estimate of the duration of the
currently installed remediation and to visualize the impact of the installed infrastructure and extraction
systems on the current and future evolution of the contaminant plume. The groundwater will also be
used for strategic planning of remediation options in the future. The model has also proven to be a
strong visualization tool for current and future scenarios, that is used in communication with
stakeholders like neighbors, contractors, the sites management, etc.
Theme C: Assessment and monitoring
1 2 1 1
Andresa Oliva ; Marcus Paulus Martins Baessa ; Elias Isler ; Marco Aurlio Zequim Pede ; Chang Hung
1
Kiang
1. Universidade Estadual Paulista, Unesp, Campus de Rio Claro, Av. 24 A, 1515, Rio Claro Sao Paulo
Brazil, CEP: 13506-900, Phone/Fax +55 19 3532-5119; (aoliva@rc.unesp.br; mpede@yahoo.com;
chang@rc.unesp.br)
2. PETROBRAS Petrleo Brasileiro S.A. CENPES, Av. Horcio Macedo, 950, Rio de Janeiro Brazil,
CEP: 21941-915, Phone/Fax +55 21 2162-6662; (marcus.baessa@petrobras.com.br)
ABSTRACT
In order to establish an adequate remediation plan for contaminated sites located in Espirito Santo and
Sergipe, Brazil due to 13,000 liters of diesel spill and due to disposal of oily sludge and crude oil,
respectively, it is fundamental to know the contaminant distribution in subsurface, normally achieved by
pertinent chemical analyses in soil samples. However, the number of samples is normally limited, making
the accurate delimitation of the contamination plume difficult and costly. In face of that in this study were
executed field tests to improve investigation of residual volatile organic solvents identification, using
electroresistivity method with 2D and 3D electrical imaging and ultra-violet optical screening tool
(UVOST). In both studied sites, it was considered that high conductivities mean possibly contaminated
portions, assertive confirmed by UVOST that it uses laser induced fluorescence (LIF) to detect the
presence and absence of fluorescing compounds, such petroleum fuel and coal tar wastes. UVOST was
designed to provide rapid sampling and real time relatively low cost analysis of the chemical
characteristics of subsurface soil and groundwater to quickly distinguish contaminated areas from
noncontaminated areas. In sites were executed geophysical and UVOST surveys where occurred the
diesel spill and disposed oily sludge and crude oil, in a location close to the leak and disposed area and in
an area there is no evidence of contamination, for definition of the background. The resistivity values in
the first area vary from 3,000 ohm to 1 Ohm.m and in the second vary from 2,500 ohm to 2 Ohm.m. These
values are correlated to local geology, constituted by sandy and clay sediments. The resistivity values
lower than 5 Ohm.m in both areas were associated with the presence of organic contamination. These
geoelectrical anomalies (resistivity lower than 5 Ohm.m) were checked using UVOST tool, that which
confirmed the presence of contamination of crude oil (signal RE % 7,8% - 29,4%) in the first area and
diesel (signal RE%186,2 222) in the second area. The UVOST data provided analytic identification and
quantification, although the quantification may be relatively imprecise.The test performed in these sites
showed the potential of the electrical imaging techniques and UVOST tool in identification and delimitation
of contaminated sites. The use of these methods is recommended, in addition to analytical methods.
In order to establish an adequate remediation plan for contaminated sites it is fundamental to know the
contaminant distribution in subsurface, normally achieved by pertinent chemical analyses in soil samples.
However, the number of samples is normally limited, making the accurate delimitation of the
contamination plume difficult and costly
In face of that, indirect methods of investigation have been applied more and more, especially geophysical
methods using techniques of Electrical Imaging and Grounding Penetrating Radar (GPR) and ultra-violet
optical screening tool (UVOST).
The results of the geophysical data interpretation, associated with ultra-violet optical screening tool
(UVOST) that it uses laser induced fluorescence (LIF) to detect the presence and absence of fluorescing
compounds, such petroleum fuel and coal tar wastes and associated with the preliminary conceptual
model and historical data of the area, can drive the installation of monitoring wells and soil sampling,
minimizing the investigation efforts and, consequently, reducing the investigation costs.
Thus, electroresistivity method with 2D and 3D electrical imaging and ultra-violet optical screening tool
(UVOST) was used to establish an adequate site management and related remediation strategies about
the distribution pattern of a 13,000 liters of diesel spill in Linhares, Espirito Santo state (Brazil) (Figure 1)
and due to disposal of oily sludge and crude oil in Carmopolis, Sergipe state (Brazil) (Figure 2).
The surveys allowed the identification of resistivity vertical and horizontal variation, the generation of a
geoelectrical model for the area and the analysis of contaminant geoelectrical response.
Finally, the respective electric resistivity values were correlated to the local geological characteristics and
the conductive anomalies detected were correlated with local in-situ measurements by using laser
induced fluorescence (LIF) to detect the presence and absence of fluorescing compounds, such
petroleum fuel and coal tar wastes. UVOST was designed to provide rapid sampling and real time
relatively low cost analysis of the chemical characteristics of subsurface soil and groundwater to quickly
distinguish contaminated areas from noncontaminated areas.
In order to analyze the variability of resistivity in subsurface in the horizontal and vertical directions, and
mainly, the geoelectrical response of impact of diesel, oily sludge and crude oil in subsurface, the
electroresistivity method was used, applying 2D and pseudo 3D electrical imaging, using the resistivimeter
AGI Super Sting R8/IP (Figure 3).
In the first area were applied twenty seven (27) 2D electrical imaging using 28 electrodes and two (2)
pseudo 3D electrical imaging by merged of the 14 lines (3D1) and 9 lines (3D2).
For the 2D electrical imaging the electrodes spacing was 5 m, reaching depths of approximately 32.7 m.
For the pseudo 3D1 and 3D2 electrical imaging, the electrodes spacing was also 5 m, covering a volume
of 385,882 m and 188,287 m , reaching depth of approximately 21 m and 23.4 m, respectively.
3 3
In the second area were applied eighteen (18) 2D electrical imaging using 28 electrodes (Figure 4) and
five (5) pseudo 3D electrical imaging by merged of the 3 lines (3D1), 5 lines (3D2), 3 lines (3D3), 4 lines
(3D4) and 3 lines (3D5).
For the 2D electrical imaging the electrodes spacing was 1.5 m, reaching depths of approximately 7 m.
For the pseudo 3D1, 3D2, 3D3, 3D4 and 3D5 electrical imaging, the electrodes spacing was also 1.5 m,
covering a volume of 1,082 m , 2,165 m , 1,086 m , 2,082 m and 1,082 m , reaching depth of
3 3 3 3 3
The Dipole-dipole array was used for both areas. This array was chosen taking into account two aspects:
the good horizontal resolution and good coverage of the measures. The data processing was performed
using software EarthImager 3D and EarthImager 2D (AGI, USA), and was used the smoothness
constrained inversion method (GROOT-HEDLIN e CONTABLE, 1990).
Figure 3 Resistivimeter AGI Super Sting R8/IP, steel stake and passive electrical resistivity imaging
cable.
UVOST delineate nearly any petroleum NAPL including gasoline, diesel, crude oil, kerosene, and many
others. It can be deployed by any type of direct push platform such as Geoprobe (Figure 4).
The UVOST system uses a sapphire window in the side of the direct push probe to measure front-face
fluorescence of the petroleum NAPL as the probe is advanced into the soil with nearly any DPT platform.
PAH fluorescence of fuels/oils is directed back to the surface where it is analyzed. Responses are
indicated in real-time on a graph of UVOST signal vs. depth (Figure 5).
In the first and second area were applied ten (10) UVOST.
UVOST were performed outside the contaminated areas to identify the background in both areas.
Figure 4 UVOST deployed in the Geoprobe.
Figure 6 and 7 - Disposal of the electrical imaging and UVOST surveys in the first and second area.
In the first area the resistivity ranges values vary from 2,500 ohmeters to 2 ohmeter. Respective electric
resistivity values were correlated to the local geological characteristics. By verification, resistivity values
lower than 5 ohmmeters were associated with the presence of diesel in the subsurface (Figure 8).
In the second area the resistivity ranges values vary from 3,000 ohmeters to 1 ohmeter. Respective
electric resistivity values were correlated to the local geological characteristics. By verification, resistivity
values lower than 5 ohmmeters were associated with the presence of oily sludge and crude oil in the
subsurface (Figures 9).
Among the 10 points analyzed by the LIF technique in the first area and second area, three of them
showed significant anomalies in first area (Figure 10) and six in the second area (answers above 2.5% RE
standard) (Figure 11).
In the first area the fluorescence logs showed callouts with similar standards to each other for the 4 peaks
of the waveforms, expected for diesel fuel. However, with loss of the light ranges polycyclic (lower blue
peak expected) due to biodegradation (Figure 10).
In the second area the fluorescence logs showed callouts with similar standards to each other for the 4
peaks of the waveforms, expected for crude oil. The LIF 07 showed waveforms slightly different from the
others, indicating contribution of other oil (Figure 11).
Figure 8 Electrical Imaging 3D illustrating the anomalies tested by the LIF technique in the first area.
Figure 9 Electrical Imaging 3D illustrating the anomalies tested by the LIF technique in the second area.
Figure 10 LIF technique results in the first area.
Figure 12 Results about the correlations between geophysical and LIF results in the first area.
Figure 13 Results about the correlations between geophysical and LIF results in the second area.
CONCLUSIONS
The UVOST data provided analytic identification and quantification, although the quantification may be
relatively imprecise.The test performed in these areas showed the potential of the electrical imaging
techniques and UVOST tool in identification and delimitation of contaminated sites. The use of these
methods is recommended, in addition to analytical methods.
REFERENCES
TRACING AN EXTENSIVE MTBE PLUME IN A SANDY AQUIFER TRYING TO ESCAPE MONITORING WELLS USING
THOROUGH HYDROGEOLOGICAL SITE UNDERSTANDING
1 2
Lars Chr. Larsen , Peder Johansen
1
Orbicon A/S, Ringstedvej 20, DK4000 Roskilde, Denmark,
2
Capital Region of Denmark, Kongens Vnge 2, DK3400 Hillerd, Denmark
lcla@orbicon.dk, (+45) 23 42 47 26
Abstract:
A gasoline station, has caused a serious contamination of gasoline components including MtBE in the
underlying sandy aquifer. BTEX-components and MtBE have been documented in the plume about
150 meters downstream the source area, however MtBE has recently been observed in water supply
wells at two well fields at a distance of 1.8 km and 2.1 km down gradient the source zone area. The
investigations have been through several phases. Initially a conceptual model for the sources of
pollution and dispersal routes for the site was established. Based on the initial conceptual model an
investigation program was set up, including a flux transect in the sandy aquifer immediately down
gradient of the source area. Subsequently it turned out that the pollution plume was located
significantly deeper and had spread further than expected and the investigations were extended
through several phases. In the latest phase, after establishing 47 screens in 17 wells with depths up to
65 meters below ground surface (bgs) the extent of MtBE contamination was investigated. MtBE-
concentrations in the plume range from 8000 g/l 80 meters down gradient the source at 32-35 meters
bgs area to 540 g/l at 150 meters down gradient at a depth of 45 - 48 meters bgs.
In addition hydrogeological tests were performed in existing monitoring and water supply wells. A total
of 45 slugtests and a long term pumping test was carried out. However, delineation of the
contamination was still not achieved.
At this point the strategy was changed into assessing the risk of the existing water abstraction by
studying the degradation of MtBE and BTEX in the deeper parts of the aquifer through isotopic
fractionation analyses. As it turned out there was no degradation of MtBE and BTEX down gradient
the source area.
The investigations so far strongly indicate, that MtBE observed in the water abstraction wells 1.8 kms
down gradient originate from this gasoline station, although other sources are also possible. The flow
path from the source area to the wells fields is not yet fully understood.
Introduction
A gasoline station in Birkerd located in the Capital Region of Denmark, has caused a serious
contamination of gasoline components including MtBE in the underlying sandy aquifer. BTEX-
components and MtBE have been documented in the plume about 150 meters downstream the source
area, however MtBE has recently been observed in water supply wells at two well fields at a distance
of 1.8 km and 2.1 km downstream the source area, see Figure 1.
Figure 1: Map showing location, well fields and source area.
In the period 1996 to 2007 remedial measures in the form of air sparging and pump and treat in the
aquifer underlying the source area was performed. However, the effect of the remedial effort was not
fully evaluated, The goal was to remove NAPL phase gasoline from the aquifer, which was succeeded.
The aspect of MtBE was not considered at the time.
The aim of this investigation was to delineate the residual contamination in the groundwater,
determine the flux of contamination from the source area, to assess the impact to the water supply
wells at nearby well fields and finally to assess the expected duration of this impact.
I order to calculate flow velocity and contamination flux, slugtests were performed in 47 screens at 5
depth levels. The resulting calculated hydraulic conductivities can be seen in Figure 3. It is surprising,
that there are differences in the order of two magnitudes in the values of a material, which is
geologically described as fairly homogeneous medium to fine grained alluvial sand. Overall, the
hydraulic conductivity tends to increase with depth. Mean values rise from about 710-5 m/s just
below the water table to about 410-4 m/s 40 meters below water table, at the bottom of the aquifer.
Considering the formation of the sand aquifer as an alluvial plain in front of a retreating ice front, this
is expected, as the energy in the braided rivers decreases as the distance to the ice front increases.
The implication for the contamination plume is that it will tend to dive deeper, because of a
downward hydraulic gradient resulting from the fact that water is abstracted more rapidly at the
bottom part of aquifer.
Figure 3: Hydraulic conductivity vs. depth. Red crosses show the individual measurements; Blue circles show mean values from
each of the five depth intervals. The blue line is a trend line for hydraulic conductivities vs. depth.
Another interesting feature that can be attributed to the geological formations history is that at a
local scale the groundwater gradient apparently can have different directions at different depths.
Figure 5 shows a conceptual model for understanding the flow in a braided river system. Figure 5
illustrate the hydraulic conductivity in the primary aquifer varies by more than two orders of
magnitude. In flow channels there will typically be high hydraulic conductivity and in the
intermediate sand, the hydraulic conductivities will be lower.
Figure 5: Direction and dip of the water Figure 4: Conceptual model for understanding the flow in a braided river
table as logged in 3 wells; OB10; OB12 and system. The overall direction of flow is rom the front of the box to the back of
OB13 over time. The circumference shows the box. For convenience the pressure level on the front of the box is set to be
the direction and the radius shows the P1 and at the back of the box P2, where P1 is greater than P2. K1 is the
gradient. hydraulic conductivity of stream channels k1 and k2 the hydraulic conductivity
in the surrounding sand k2 where k1 >> k2. The wells B1 and B2 are screened
across the through-flow channel, while B3 is screened across one cut off flow
channel. The pressure level in well B3 will be close to P1 because of the lower
hydraulic conductivity in the intermediate sand, while the pressure falls evenly
between B1 and B2 because of the relatively higher conductivity in the stream
channel.
Flux calculations
Based on the measured hydraulic conductivities and concentrations and the groundwater gradient, the
MtBE flux in the sand aquifer is calculated just down gradient the source area, where the
contamination plume is elevated.
Figure 6 shows the flux as grams pr. square meter per year.
Figure 6: MtBE flux in the sandy aquifer, just downstream the source area as grams per square meter per year. The location of
the transect is shown in Figure 2.
By integrating the flux profile in figure 6, the resulting MtBE flux is estimated to approx. 12 kg
MtBE/year through the transect.
Isotopic fractionation analyses
Because of the very high cost of deploying investigation wells in a steeply diving plume, it was
abandoned to delineate the contamination down gradient by performing additional wells. Instead the
strategy shifted into assessing the risk to the existing water supply by studying the degradation of
MtBE and BTEX in the deeper parts of the aquifer through isotopic fractionation analyses in order to
assess the risk to the existing drinking water wells. The objective was to find out whether the plume
biologically degraded before it reaches the water supply wells.
In order to document any biological degradation, isotope fractionation analysis was performed on
seven water samples along a transect from the source area to the to the farthest well, where MtBE
and benzene were detected (180 ms down gradient).
Biodegradation is considered to be the only significant mechanism that gives rise to the isotopic
12 13
C/ C fractionation of MTBE and BTEX components. Detection of enrichment of the heavier isotopes
down gradient source area must therefore be considered to be a certain indication of degradation.
Interpretation of isotopic fractionation of MtBE and BTEX is particularly easy as the original isotopic
composition lies in a narrow range, so that fractionation thus appears clearer.
Figure 7: 13C enrichment as a function of MtBE concentrations. Inset shows the normal path for degradation versus dispersion-
dilution.
The conclusions of the performed isotopic fractionation analyses is that MtBE down gradient from
the source area, i.e. down gradient well P0is are not biodegraded, since there is no enrichment in 13C.
Immediately below the source area, at P0, there is weak evidence of biodegradation of MtBE.
For benzene (not shown), there are clear indications that biodegradation are occurring the first part
of the plume, after which there are no further indications of biodegradation. This probably reflects
that the redox conditions in the source area are methanogenic, while the redox conditions in the first
25 meters of the plume are aerobic. After the first 25 meters, the conditions change again to
anaerobic.
Geological re-interpretation
The results of isotopic fractionation tests coincided with the reports of low level finds of MtBE in
three water supply wells down gradient the source area. MtBE was detected in the two well fields:
Ellebk and Sandholm, about 2 km downstream from the source area. MtBE (or benzene) was
however not detected in the Birkerd Well Field, which is closer to the source area, see Figure 8. All
well fields abstract drinking water from an underlying limestone aquifer. No other sources of MtBE in
the area are known.
Figure 8: Map showing well fields and findings of MtBE in water abstraction wells. Blue contours shows the potentiometric
surface in the sand aquifer, the red line shows the location of the geological profile in Figure 9.
The odd distribution of finds in water supply wells called for a re-interpretation of the geology in the
area. The result is show in Figure 9.
Figure 9: Geological cross section from source area to Sandholm Well Field, location is shown in Figure 8.
The re-interpretation explains why Sandholm and not Birkerd Well Field is affected by the MtBE
plume. At Birkerd Well Field the sandy aquifer is separated from the limestone aquifer by a clayey
deposition of either clayey till or alluvial clay. This is confirmed by pumping tests in one of the wells in
Birkerd Well Field.
Conclusion
The investigations have shown that it is essential to have a thorough knowledge of the geological and
hydrogeological conditions in order to understand the pathways of plumes both horizontally and
vertically in an aquifer.
Hydrogeological tests may to some extent reduce the need for the establishment of monitoring wells.
In glacial meltwater deposits plumes tend to dive steeply due to the increase of hydraulic conductivity
with depth. In this case study the he vertical transport was not originally expected, since the plume
was in a fairly homogenous sandy aquifer.
Plumes with BTEX and MtBE can extend very far (> 2 km) and deep under unfavourable degradation
conditions. The degradation potential can be documented by performing isotope fractionation analysis.
1b. Risk assessment and management
MONITORING OF THE DORAMECTIN DIFFUSION INTO A SOIL AMENDED WITH PIG MANURE
Introduction
Human and veterinary drugs are continuously being released in the environment mainly as a result of
the manufacturing processes, the disposal of unused or expired products, and the excreta (Pretovic et
al., 2005). Many drugs are stable and can therefore accumulate in eco-systems. Although the amount
of pharmaceuticals and their bioactive metabolites introduced into the environment is likely low, their
continuous environmental input may lead to a high long-term concentration (Pretovic et al., 2005).
Thus, it is need to study how pharmaceutical production and consumption affect the environment. The
analysis of drugs in the environment constitutes a difficult task, first, because of the complexity of the
matrices, and second, because of the normally very low concentrations of the target compounds. In
the last decade, the impressive progress of both the extraction techniques and liquid chromatographic
methods has led to quantify low levels of pharmaceuticals in environmental matrices.
Veterinary medicines are widely used in treating cattle and can be released into the environment,
either directly in feces or urine or indirectly after the application of the manure as an organic fertilizer.
Numerous veterinary medicines, such as hormones, antibiotics, and antiparasitics, have been
detected in soil and water (Boxall et al., 2006; Martnez-Carballo et al., 2007; Petrovic et al., 2005). As
emerging contaminants, these feces-borne drugs are suspected of causing adverse effects in both
humans and wildlife. In the last few years, several studies have focused on the behavior of veterinary
medicines on the environment (Accinelli et al., 2007; Blackwell et al., 2007; Kolar et al., 2006; Kolar
and Kouh Eren 2006).
Avermectins are macrolide endectocides that are now widely used in veterinary medicine because of
their broad spectrum activity against ecto- and endoparasites, their high efficiency, and their high
safety margin (Kolar and Kouh Eren 2006). Doramectin is one of the avermectins most used
(Goudie et al., 1993). Doramectin is excreted mainly through feces as non-metabolized drug, and its
excretion profile depends strongly on the drug formulation, dosage, animal species, and sex of the
animal (Gil-Diaz et al., 2011; Kolar et al., 2006; Taylor, 1999). The use of cattle residues contaminated
with avermectins as an organic amendment to soil could potentially be a source of diffuse pollution
with these veterinary medicines in various environmental media, especially in the soil, surface water,
and groundwater. Furthermore, avermectins have a low solubility in water (25 g/L for doramectin)
and a strong affinity for lipids and organic matter. Therefore, avermectins could persist in the soil for
long periods of time, thus affecting soil biodiversity (Kolar et al., 2008; Rmbe et al., 2010; Suarez,
2002). The chemical structure of doramectin is shown in Figure 1.
This study presents a field experiment which evaluates the potential diffusion and persistence of
doramectin in soils amended with manure from pigs treated with doramectin. A 213-day monitoring
study was performed.
Experimental design
The study was conducted at an experimental pig farm located in Aranjuez (Madrid, Spain), where anti-
parasite treatment of the pigs was usually performed using doramectin. Manure from 40 female pigs
treated with doramectin (Dectomax, Pfizer, Injectable Solution 1%, 0.3 mg/kg) were collected for 12
days, mixed with waste straw, and then, applied to an olive grove soil. This area has not been
previously treated with animal manure, at least in the last 10 years. The soil and manure properties
were determined according to the Spanish official methodology (MAPA 1994) and are shown in Table
1. In brief, electrical conductivity (EC) and pH were measured in a 1:2.5 soil-to-water ratio; the organic
matter and total nitrogen contents were determined using the WalkleyBlack and Kjeldahl methods,
respectively; the percentage of carbonates was measured using a calcimeter; the available
phosphorus was evaluated using sodium bicarbonate at pH 8.5 as an extractant; and available
nutrients were extracted with 0.1 N NH4Ac and assessed using flame atomic absorption spectrometry
(FAAS) (AA240FS, Varian, Victoria, Australia). The soil texture was analyzed using a Bouyoucos
densimeter.
Two plots, A and B (14 x 5 m) were treated with 4.6 kg/m 2, of manure, according to the traditional
practices for nitrogen application. Analogous control plots were also included. The mean concentration
of doramectin in the manure was 101.4 18.1 ng/g (dry feces). Soil samples (0-30 cm) were collected
to monitor the doramectin levels on days 0, 1, 6, 29, 63, 85, 115 and 213. To evaluate drug mobility,
deeper samples (30-60 and 60-90 cm) were collected on days 115 and 213. Two soil samples per plot
were collected with a soil probe (each plot was divided into two subplots in which two composite
samples were taken). Soil samples were air dried, passed through a 2 mm screen, and mixed prior to
analysis to ensure homogeneity.
An analytical method was developed to determine the concentration of doramectin in the soil and pig
manure samples and was described in detail by Gil-Daz et al (2011). The method was based on
extraction with acetonitrile followed by a clean-up using solid-phase extraction (SPE); the extract was
evaporated to dryness under a stream of nitrogen at 45C and a derivatization reaction was performed
using 1-methylimidazole, trifluoroacetic anhydride, triethylamine and trifluoroacetic acid to obtain a
fluorescent derivative (Berendsen et al., 2007). Three replicates were performed for each sample. The
quantitative analysis of doramectin was carried out by high performance liquid chromatography
(HPLC) using fluorescence detection (Gil-Daz et al., 2011). A Waters HPLC system (Milford, MA,
USA) equipped with the following components was used: a gradient pump (600 Controller) equipped
with a degasser, a 717 plus autosampler, a column heater, and a fluorescence detector model 2475
Multi . An aliquot of 20 L of derivatized extract was injected on a Phenomenex Luna C18 (2) column
(150 x 4.6 mm i.d.; 5 m particle size) with a Phenomenex C18 pre-column (4.0 x 3.0 mm i.d.; 5 mm
particle size). The column temperature was maintained at 35C. The mobile phase consisted of
methanol, acetonitrile, and water (48:47:5, v/v/v) and was pumped at a rate of 1.2 mL/min. The
fluorescence detector was set at an excitation wavelength of 365 nm and an emission wavelength of
470 nm.
Statistical analyses
Differences in doramectin concentration between the different sampling times were evaluated by the
one-way ANOVA at a significance level of p<0.05 using the SPSS statistical package for Windows,
release 16.0 (SPSS Inc, IBM Company), followed by a Tukey post-hoc test to determine which means
are in fact different.
Table 1 shows the mean values of the main physico-chemical properties of the soil used in the present
study. It is a sandy loam soil, with an alkaline pH (pH 8.1) and moderate nutrients content. Thus, the
application of some organic fertilizer would be required to improve soil fertility and the nitrogen
requirements of the crop.
After the application of the manure to the field, the doramectin levels were monitored over a period of
213 days (Table 2). In control soil samples doramectin was not found. In general, both treated plots, A
and B, showed a similar behaviour. The maximum concentration of doramectin was detected six days
after treatment (4.04 1.63 ng/g dry feces). On day 29, the doramectin concentration in soil
decreased significantly, and mean levels were similar to those found on day 63 in plot A and on day
85 in the two plots. On day 115, the doramectin level was higher at 30-60 cm than the superficial
zone. In samples from day 213, traces of doramectin were found from the surface to a 90 cm depth.
This result demonstrates some mobility of the doramectin molecules after seven months, probably
associated with organic matter because doramectin presents a high organic carbon-normalized
sorption coefficient (approximately 7500 for a silty loam soil, according to Kolar and Kouh Eren
(2006)). There are limited data about field experiments that evaluate doramectin diffusion into soil. Our
data are consistent with Taylors estimates in which 1.8 ng/g of doramectin could be found in soil
where sheep treated with doramectin had grazed (Taylor, 1999). Published data on doramectin
concentrations related to ecotoxicity effects in soil fauna are higher than the values found in the
present study (Kolar et al., 2008; Surez et al., 2009). In general, the mobility of pharmaceutical
compounds in soil depends on drug concentration and source, soil type and environmental factors
such as rain intensity (Drillia et al., 2005; Wu et al., 2010). During the experiment, the total
2
precipitation was 387 L/m . Studies about sorption and mobility of pharmaceutical compounds in soil
have shown high mobility in soil organic matter-poor soils, which emphasizes the potential transport of
these drugs to groundwater (Drillia et al., 2005; Chefetz et al., 2008). Thus, the application of irrigation
water and organic amendments such as biosolids or manure containing pharmaceutical compounds
could suppose an important risk introducing drugs into soil and groundwater. In the case of
doramectin, its sorption to soil not likely to be pH dependent because it is an un-ionized molecule in
the normal soil pH range (between 5 and 9). In a monitoring study on the presence of veterinary
medicines in the United Kingdom, Boxall et al. (2006) did not find doramectin in soil samples, although
it was found in sediment samples.
The obtained results evidence that under the present experimental conditions, doramectin diffusion
from manure to soil was moderate, since a low concentration of doramectin in soil was found even
after a recent application. However, successive applications of manure from pigs treated with
doramectin in a specific area would produce an accumulation of this drug in the soil, and it could reach
toxic levels for soil fauna. It should also be taken into consideration that the successive application of
manure in a same agricultural area is a common practice in the current farming productive system,
intensive type. Moreover, it should be taken into account the potential interaction of this drug with
other contaminants that could be found in soil samples.
Conclusions
The application of pig manure containing doramectin under the specified conditions led to the
presence of low levels (below 5 ng/g) of the drug in the soil. Some mobility of doramectin through the
soil profile was observed and seven months after the manure application, traces of doramectin were
still detected from the surface of the soil to a 90 cm depth. Thus, successive applications of manure
from pigs treated with doramectin in a specific area would produce an accumulation of this drug in the
soil. Further research about the effects of successive land applications of manure containing
doramectin should be performed.
Acknowledgements.
The authors thank IMIDRA (Community of Madrid) for supporting Project FP 13 CONT, and Ministerio
de Educacin y Ciencia for supporting Project CTM2007-66306-C02-02/TECNO.
References
Accinelli, C.; Koskinen, W.C.; Becker, J.M.; Sadwsky, M.J. Environmental Fate of Two Sulfonamide
Antimicrobial Agents in Soil. J. Agric. Food Chem. 2007, 55, 26772682.
Berendsen, B.J.A., Mulder, P.P.J., Hans (J.)A. van Rhijn. The derivatisation of avermectins and
milbemycins in milk: New insights and improvement of the procedure. Anal. Chim. Acta 2007, 585,
126-133.
Blackwell, P.A.; Kay, P.; Boxall, A.B.A. The dissipation and transport of veterinary antibiotics in a
sandy loam soil. Chemosphere 2007, 67, 292299.
Boxall, A.B.A.; Fogg, L.A.; Baird, D.J.; Lewis, C.; Telfer, T.C.; Kolpin, D.; Gravell, A.; Pemberton, E.;
Boucard, T. Targeted monitoring study for veterinary medicines in the environment. 2006. Science
Report: SC030183/SR.
Chefetz, B., Mualem, T., Ben-Ari, J. Sorption and mobility of pharmaceuticals compounds in soil
irrigated with reclaimed wastewater. Chemosphere 2008, 73, 1335-1343.
Drillia, P., Stamatelatou, K., Lyberatos, G. Fate and mobility of pharmaceuticals in solid matrices.
Chemosphere 2005, 60, 10341044.
Gil-Daz, M.; Prez-Sanz, A.; Martn, M.; Lobo, M.C. Potential Diffusion of Doramectin into a Soil
Amended with Female Pig Manure. A Field Experiment. J. Agric Food Chem. 2011, 59, 10635-10640.
Goudie, A.C.; Evans, N.A.; Gration, K.A.F.; Bishop, B.F.; Gibson, S.P.; Holdom, K.S.; Kaye, B.; Wicks,
S.R.; Lewis, D.; Weatherley, A.J.; Bruce, C.I.; Herbert, A.; Seymour, D.J. Doramectin - a potent novel
endectocide. Vet. Parasitol. 1993, 49, 5-15.
Kolar, L.; Flajs, V.C.; Kuner, J.; Marc, I.; Pogacnik, M.; Bidovec, A.; Van Gestel, C.A.M.; Kouh
Eren, N. Time profile of abamectin and doramectin excretion and degradation in sheep faeces.
Environ. Pollution 2006, 144, 197-202.
Kolar, L.; Kouh Eren, N. Veterinary parasiticides are they posing an environmental risk? Slov. Vet.
Res. 2006, 43(2), 85-96.
Kolar, L.; Kouh Eren, N.; Hogerwerf, L.; Van Gestel, C.A.M. Toxicity of abamectin and doramectin to
soil invertebrates. Environ. Pollution 2008, 151, 182-189.
MAPA, Ministerio de Agricultura. Mtodos Oficiales de Anlisis, 1994; vol. III, Spain.
Martnez-Carballo, E.; Gonzlez-Barreiro, C.; Scharf, S.; Gans, O. Environmental monitoring study of
selected veterinary antibiotics in animal manure and soils in Austria. Environ. Pollution 2007, 148,
570-579.
Petrovic, M.; Hernando, M.D.; Daz-Cruz, M.S.; Barcel, D. Liquid chromatographytandem mass
spectrometry for the analysis of pharmaceutical residues in environmental samples: a review. J.
Chromatogr. A 2005, 1067, 1-14.
Rmbke, J.; Coors, A.; Fernndez, A.A.; Frster, B.; Fernndez, C.; Jensen, J.; Lumaret, J.P.; Porcel-
Cots, M.A.; Liebig, M. Effects of the parasiticide ivermectin on the structure and function of dung and
soil invertebrate communities in the eld (Madrid, Spain). App. Soil Ecol. 2010, 45, 284-292.
Surez, V.H. Helminthic control on grazing ruminants and environmental risks in South America. Vet.
Res. 2002, 33, 563-573.
Surez, V.H., Lifschitz, A.L., Sallovitz, J.M., Lanusse, C.E. Effect of faecal residues of moxidectin and
doramectin on the activity of arthropods in cattle. Ecotoxicol. Environ. Saf. 2009, 72, 15511558.
Taylor, S.M. Sheep scab-environmental considerations of treatment with doramectin. Vet. Parasitol.
1999, 83, 309-317.
Wu, C., Spongberg, A.L., Witter, J.D., Fang, M., Czajkowski, K.P., Ames, A. Dissipation and leaching
potential of selected pharmaceutically active compounds in soils amended with biosolids. Arch.
Environ. Contam. Toxicol. 2010, 59, 343-351.
soil manure
pH 8.1 8.4
Carbonates (%) 27 -
P (mg/kg) 35 450
Clay (%) 5 -
Texture class,
sandy loam -
USDA 1994
Table 2. Mean doramectin concentration (ng/g) in soil samples treated with pig manure (nd: not
detected; LOQ: limit of quantification). Means with different letters in the same column indicate
significant differences (p < 0.05; Tukey).
0 0-30 nd nd
a
1 0-30 0,780,07 <LOQ
60-90 nd nd
1. School of Civil Engineering & Geosciences, Newcastle University, Newcastle upon Tyne, UK, NE1
7RU
2. TerraConsult Ltd, Bold Business Centre, Bold Lane, Sutton, St Helens, Merseyside, UK, WA9 4TX
KEYWORDS
Carbon dioxide, methane, ground gas, monitoring, landfill, Coal Measures, high temporal frequency
ABSTRACT
Understanding the exchange of carbon dioxide (CO 2) and methane (CH4) between the geosphere and
atmosphere is essential for the management of anthropogenic emissions. In this paper an innovative
approach of expressing ground gas compositions is presented, using data derived from regulatory
monitoring of boreholes in the unsaturated zone at infrequent intervals (typically 3 months) with data
from a high frequency monitoring instrument measuring half-hourly over periods of weeks. Similar highly
variable trends are observed for timescales ranging from decades to hourly for boreholes located close
to sanitary landfill sites. Additionally, high frequency monitoring data confirm the effect of meteorological
controls on ground gas emissions. For a two week monitoring period, only air was present in a borehole
during 70% of the data capture. There is a clear weakness in current point monitoring strategies that
may miss emission events and this needs to be considered along with obtaining baseline data prior to
starting any engineering activity.
INTRODUCTION
The exchange of carbon dioxide (CO2) and methane (CH4) between the geosphere and
atmosphere is a key component of the global carbon cycle. At the advent of industrialisation,
atmospheric CO2 concentration was 280 ppm and has subsequently increased so that it now exceeds
400 ppm (IPCC, 2007). This figure is projected to reach 600-800 ppm by the close of the century
(IPCC, 2007). It is believed that the rise in CO2 concentration in the Earths atmosphere is linked to
global climate change (Hansen and Sato, 2004, Hansen et al., 1981). With a global warming potential
(GWP) 21 times greater than CO2 (Nosalewicz et al., 2011), CH4 emission to the atmosphere is
considered to be one of the greatest environmental challenges of the 21 st Century. In addition to
emission of CO2 from the combustion of fossil fuels, which account for the major part of the post-
industrial increase (Hofmann et al., 2009), CO2 exchange between the coupled plant-soil system and
the atmosphere is a major control of atmospheric CO2.
Emissions of gases from soils vary in their origin. Near surface fluxes of CO2 and CH4 from
biologically-active soils occur in response to microbial respiration, and occur rapidly after the
formation of the gas (Maier and Schack-Kirchner, 2014). Concentrations of CO2 and CH4 in soils
dominated by plant roots and associated microbial systems are typically very low, below 1% v/v
(Hirano et al., 2003). Below the biologically-active soil, sedimentary geological processes produce
CO2 and CH4. Proportions vary according to specific geological circumstances. Natural fluxes of CO2
and CH4 to the atmosphere include contributions from both geological and biological sources. Artificial
sources of gases within soil include sanitary landfill, where anaerobic microbial processes
characteristically produce gas containing up to 70% CH4 and 30% CO2 (Bergmaschi and Harris, 1995,
Yavitt et al., 1995). Geological gases migrate through permeable and fractured formations until
trapped or released to the atmosphere. Similarly, gases produced in landfills migrate laterally as well
as vertically, if containment systems fail. Monitoring systems in the vicinity of landfills are designed to
detect migration, so that appropriate action can be taken.
For well-informed management of deep engineering processes that have the potential to affect
greenhouse gas emissions, there is a need to understand the background levels of CH 4 and CO2 in
soils, superficial unconsolidated sediment deposits and groundwater systems and, importantly, the
extent of their temporal variability. In northern England, the compositions of gases within shallow
unconsolidated sands capped by glacial till are specifically considered as unsaturated sands are a
potential gas migration pathway. Landfills have been constructed in these formations, and monitoring
regimes are in place to assess the integrity of containment, with boreholes outside the landfill to
permit determination of gas compositions and so detect any migration. Thus, there is access to a
resource of historical monitoring data that extends back several decades. CO2 and CH4 in the sands
may be derived from sources other than landfill, including natural migration from underlying strata, in
this case, coal-bearing. In this paper, historic periodic and recent high frequency monitoring data
obtained from outside landfill sites are compared, with reference to data from a site with a similar
geological setting but distant from landfill. The use of different temporal scales of observation has
major implications for the planning and interpretation of ground gas monitoring procedures, with a
wide range of applications. The work presented here extends material previously published by
Teasdale et al. (2014).
DATA COMPILATION
Data Sources
Monitoring data were obtained for two landfill sites in Cheshire, UK. The sites differ geologically
and produce contrasting gas signatures. Site 1, a closed landfill, is situated in thick Quaternary
deposits with no known potential external (geological) source of CH4 or CO2. Conversely, Site 2 is
situated in Quaternary deposits overlying thin Permo-Triassic bedrock, with underlying Coal Measures
(Carboniferous; Westphalian) which provide a potential source of geogenic gas. Additionally, Site 2 is
adjacent to a second older landfill, constructed to a lower technical specification. The Control Site was
chosen at a location with similar geology to Site 1, but with no associated landfill and so is used as a
control. Details of the ground conditions at each site can be found in Teasdale et al. (2014).
Historically, gas composition has been measured in accordance with regulatory requirements
using hand-held gas monitors, as frequently as daily (Site 2) to as infrequently as quarterly (Site 1).
Measurements were made using a Geotech UK GA2000 Landfill Gas Analyser until 2012, after which
a Gas Data Ltd GFM435 Landfill Gas Analyser was used. One of the limitations of changes of
instrumentation over a period of several years is that data sets may not be directly comparable. Both
gas monitoring instruments employ a dual beam infrared absorption method to quantify the
concentration of CH4, CO2 and O2 in a flowing gas. The balance is assumed to be nitrogen (N2).
However, the instruments differ in their measurement of borehole flow. The GA2000 pressure
transducer is equipped with a set resistor in order to minimise flow. It is argued that this will more
closely reflect true borehole conditions as the act of opening the borehole valve will disrupt equilibrium
conditions. Conversely, the GFM435 pressure transducer does not have this specification.
High temporal resolution data capture was achieved using a GasClam instrument, by which gas
compositions can be recorded at intervals as short as three minutes (IonScience, 2012). The
GasClam also records atmospheric pressure and temperature as well as gas concentration, avoiding
an additional limitation of hand-held gas monitors (Morris et al., 2008). The instrument has, according
to the manufacturer, a range 0-100% for CH4 and CO2, with a stated detection limit of 0.1% for both
gases. In practice, observed non-zero minimum CH4 and CO2 concentrations were 0.4% and 0.1%
respectively, and observed maximum values were 57.1% and 10.8% respectively.
Two graphical approaches have been used to present the gas monitoring data. In the first,
compositional data are compared in terms of the measured CH4, CO2 and O2 content normalised to
N2 (assuming N2 to be the balance, i.e. 100 ( CH4, + CO2 + O2)), in a ternary plot. This allows the
relative proportions of CH4 and CO2 to be compared irrespective of any dilution by air, the O 2/N2 ratio
indicating whether this has occurred and the extent to which oxygen has been removed, as N2 can be
regarded as non-reactive (Bergmaschi and Harris, 1995). In a plot of this type (which is analogous to
the Piper diagram in hydrogeology), end member compositions can be identified, so that an array of
observed data points can be explained as mixtures of gases from different sources, and the
characteristics of one borehole can be compared with another. The second graphical approach is to
show variation with time in absolute gas concentration (% v/v) measured at high frequency and how
this is affected by atmospheric pressure conditions.
Monitoring Wells near a Landfill where there is no Known Input from Coal-Bearing Rocks (Site 1)
Figure 1 shows the gas composition recorded over a period of 10 years in boreholes at Site 1.
BH 1 is located on the southern boundary of the landfill and produced a CH4-rich gas signature (70:30
CH4/CO2) which is attributed to landfill gas given the close proximity of the borehole to the edge of the
landfill. Values recorded at 100% O2/N2 correspond to air. There is some evidence of mixing between
landfill gas and air at this location, but this was observed on few occasions (points between the two
end member compositions).
CO2 / N2 (Relative %)
0
100
BH 1 10
90
20
80
30
70
40
60
50
50
60
40
CO2 / N2 (Relative %)
70
30 LANDFILL GAS
80 0
100 20
AIR 90 10
90 10
100 20
80 0
O2 / N2 CH4 / N2
0 10 20 30 40 50 60 70 80 90 100
(Relative %) 30 (Relative %)
70
40
60
50
BH 2 60
50
40
70
CO2 / N2 (Relative %)
30
80
0 20
100
90
10 10
90
100
20 0
O2 / N2 80 CH4 / N2
0 10 20 30 40 50 60 70 80 90 100
(Relative %) 30
70
(Relative %)
40
60
50
BH 3 60
50
40
70
30
80
20
90
10
100
0
O2 / N2 CH4 / N2
0 10 20 30 40 50 60 70 80 90 100
(Relative %) (Relative %)
Figure 1 Gas compositions for monitoring wells on Site 1 (2004-2014), indicating (Figure 2a)
expected plotting positions of landfill gas and air.
BH 2 is approximately 25 metres south of the landfill perimeter, and shows a more diffuse scatter
of data. There is still a strong indication of the presence of landfill gas but the proportion of CH4 is
lower (Figure 2), suggesting mixing with more CO2-rich gas, or removal of CH4 by biological
processes. BH 3, 100 m away from the landfill, records no landfill gas (Figure 2).
High frequency Monitoring where there is no Known Input from Coal-Bearing Rocks (Site 1)
50
CO2
O2 1020
40 Atmospheric Pressure
30
1000
20
10 980
0
960
0 13 0 13 0 13 13 13 13 13 13 13 13 13 13 13 13
0/ 2 0/ 2 1/ 2 /20 /20 /20 /20 / 20 /20 /20 / 20 /20 /20 /20
/1 /1 /1 /11 /11 /11 / 11 /11 /11 /11 /11 / 11 / 11 / 11
30 31 01 02 03 04 05 06 07 08 09 10 11 12
50 CO2 10
O2
40 5
Differential Pressure
30
0
20
-5
10
-10
0
-15
3 3 3 3 3 3 3 3 3 3 3 3 3 3
01 01 01 01 01 01 01 01 01 01 01 01 01 01
/ 10/ 2 / 1 0/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2 / 1 1/ 2
30 31 01 02 03 04 05 06 07 08 09 10 11 12
C Gas monitoring time-series data against atmospheric pressure from Control Site
60 1040
50 CH4
CO2 1020
40 O2
Atmospheric Pressure
30
1000
20
10 980
0
960
4 4 4 4 4 4 4 14 14 14 14 14 14 14
/2 01 /201 /201 /2 01 /201 /2 01 /201 / 20 /20 / 20 / 20 / 20 / 20 /20
/02 /02 / 02 /02 /02 / 02 / 02 /02 / 02 /02 /02 /02 /02 / 02
04 05 06 07 08 09 10 1 1 1 2 1 3 1 4 1 5 1 6 1 7
Figure 2 Gas monitoring time-series data collected from GasClam against atmospheric pressure and
differential pressure (BH 1 Site 1), and gas monitoring time-series data against atmospheric pressure
(Control Site)
High frequency data for the two-week period from BH 1 at Site 1 are plotted (Figure 3) for
comparison with monitoring data for 10 years for the same borehole (as in Figure 1). In two weeks, a
pattern of variation in gas composition is observed that covers the range of data observed during ten
years of periodic monitoring. Additionally, mixing of air and landfill gas is more fully resolved, as
demonstrated by the greater number of data points between the two end member compositions.
Crucially, this demonstrates that occurrences of elevated emissions of CO2 and CH4 over a ten year
period are much more frequent than suggested by measurement under the requirements of a normal
regulatory regime.
CO2 / N2
(Relative (%)
0
100
10
90
20
80
30
70
40
60
50
50
60
40
70
30
80
20
90
10
100
0
O2 / N2 0 10 20 30 40 50 60 70 80 90 100
CH4 / N2
(Relative %) (Relative %)
Figure 3 Ternary Plot of high frequency monitoring data obtained from BH 1 Site 1 for the Period
29/10/2013 12/11/2013
Monitoring Wells near a Landfill that Show Inputs to Ground Gases from Coal-Bearing Rocks (Site 2)
At Site 2, which overlies Coal Measures, monitoring wells are located around the landfill
perimeter. From hand-held monitor generated data, the CH4/N2 ratio frequently exceeds 80%,
suggesting a geogenic gas influence arising in two boreholes (BH W and BH Y). In BH W, the array of
data is consistent with mixing of air with predominantly geogenic CH4, with some observations
extending towards CO2/N2 that may reflect a landfill gas component.
Boreholes BH Y and BH Z clearly show more variable gas compositions, with CH4/N2 and CO2/N2
ratios approaching 100% (Figure 4). Compared with Site 1, where gas composition was almost
exclusively air or landfill gas (with some mixing in between), boreholes at Site 2 show a more complex
pattern. The complication and greater mixing of gases is most likely to be due to additional sources
associated with the geological characteristics of the site. The CH4-rich compositions observed are
consistent with the presence of geogenic gas derived from underlying Coal Measures.
High frequency Monitoring near a Landfill that Show Inputs to Ground Gases from Coal-Bearing
Rocks (Site 2)
50
1020
40
30 CH4
CO2 1000
20 O2
Atmospheric Pressure
10 980
0
960
4 4 4 4 4 4 4 14 14 14 14 14 14
/2 01 /201 /2 01 /201 /201 /201 /201 /20 /20 /20 /20 /20 /20
/07 /07 / 07 / 07 / 07 / 07 /07 /07 /07 /07 /07 /07 /07
09 10 11 12 13 14 15 16 17 18 19 20 21
50 CO2 10
O2
40 Differential Pressure 5
30
0
20
-5
10
-10
0
-15
4 4 4 4 4 4 4 4 4 4 4 4 4
/2 01 /2 01 /2 01 /2 01 /2 01 /2 01 /2 01 /201 /2 01 /2 01 /2 01 /201 /2 01
/ 07 / 07 /07 / 07 /07 /07 /07 /07 /07 /07 /07 / 07 /07
09 10 11 12 13 14 15 16 17 18 19 20 21
50 CO2 25
O2
Temperature (C)
40 20
Temperature
30
15
20
10
10
5
0
0
4 4 4 4 4 4 4 4 4 4 4 4 4
/2 01 /2 01 /2 01 /2 01 /2 01 /2 01 01 01 01 01 01 01 01
/07 /07 /07 /07 /07 /07 7/ 2 /07
/2
/07
/2
/07
/2
/07
/2
/07
/2
/07
/2
09 10 11 12 13 14 1 5/ 0 16 17 18 19 20 21
Figure 5 Gas monitoring time-series data collected from GasClam against atmospheric pressure,
differential pressure and temperature at BH W, Site 2 (08/07/2014 21/07/2014)
An immediately obvious limitation of the GasClam device is that multiple instruments are required
to sample multiple boreholes simultaneously at high frequency. Data sets are not directly comparable
as recordings were made during different conditions. The intended purpose is to illustrate the
variability in the data, different timescales and comparisons with other sites.
Shown in Figure 5 is the variation in ground gas composition for BH W at Site 2 measured using
the same protocols as at Site 1. Temperature appears to be the key influence on emissions of CH 4
and CO2 for this monitoring well. CH4 and CO2 are negatively correlated with temperature while O2
concentration is positively correlated with temperature and variability in gas composition can be
ascribed to a diurnal timescale. However, the data presented here is more complex than diurnal
cycling, as shown in Figure 6.
30 30 1040
25
Temperature
20 1020
Temperature (C)
Atmospheric Pressure
20
15
10 15 1000
5 10
0 980
5
-5
0 960
14 14 14 14 14 14 14 14 14 14 14 14 14
7 /20 / 20 /20 /20 /20 /20 /20 /20 /20 / 20 /20 /20 / 20
9/ 0 /07 1/ 0
7
2/ 0
7
3/ 0
7
4/ 0
7
5/ 0
7
6/ 0
7
7/ 0
7
8/ 0
7
9/ 0
7
0/ 0
7
1/ 0
7
0 10 1 1 1 1 1 1 1 1 1 2 2
Figure 6 O2 concentration time-series data against atmospheric pressure and temperature from
BH W, Site 2 (08/07/2014 21/07/2014)
During periods of static high pressure (08/07/2014 12/07/2014), ambient temperature is the main
control on gas composition and a close relationship exists that varies on a diurnal time scale.
However, when atmospheric pressure decayed by 12 mbar (12/07/2014-13/07/2014), the correlation
weakened, but stabilised once more when atmospheric pressure increased and stabilised at 1013
mbar on 14/07/2014. While the relationship with atmospheric pressure is less explicit for BH W, Site
2, it still has an influencing role.
As for BH 1, Site 1, that was located on the landfill perimeter, the high frequency resolution
data from BH W, Site 2, may be normalised to N2 and transformed into a ternary plot as shown in
Figure 7.
CO2 / N2 (Relative %)
0
100
10
90
20
80
30
70
40
60
50
50
60
40
70
30
80
20
90
10
100
0
O2 / N2 CH4 / N2
0 10 20 30 40 50 60 70 80 90 100
(Relative %) (Relative %)
Figure 7 Ternary Plot of High Temporal Frequency Data obtained from BH W Site 2 for the Period
08/07/2014 21/07/2014
Unlike the point measurement data obtained over a 16 year period that shows gas composition
approaching CH4/N2 end member concentration and a greater spread (mixing) of gases, the selected
two week period consistently shows a strong CH4 concentration (65-85%) with very little CO2
contribution (typically less than 4%). This supports the case for a geogenic gas source.
DISCUSSION
The historical data measured using portable gas monitoring devices reveal patterns in gas
composition that relate to the geology of the site and subsequent landfill activity, where appropriate.
The high frequency monitoring of data allows temporal variability to be constrained, and indicates
links with weather conditions. Taken together, these observations have significant implications for the
understanding of ground gas emissions to the atmosphere with particular relevance for deep
engineering activities that might perturb ground emissions, as well as for the design of monitoring
programmes
Gas may migrate from geological and non-geological sources through different pathways to the
atmosphere at locations where glacial deposits overly coal-bearing rocks. This is a two-way process,
as monitoring data clearly indicate that air enters boreholes (and other voids) in periods of high
atmospheric pressure; it will also enter permeable formations wherever they are exposed naturally
(e.g. valley sides). Dilution of gases in a mixture by equivalent proportions occurs with the addition of
one non-reactive gas component (Romanak et al., 2012). Thus, any CO2 and CH4 present are diluted
by air (N2 + O2) during periods of increasing atmospheric pressure.
The sources of ground gas can be distinguished from their compositions; taking a ratio to N2
enables the relative proportions of CH4 and CO2 to be determined, allowing for dilution by air. In this
study, gases derived from sands overlying coal bearing rocks are shown to be richer in CH4 than
those gases derived from landfill. Historical monitoring data shows evidence of mixing of ground gas
and air, to extents that vary from occasion to occasion. Observations of gas compositions at a site
with neither landfill nor coal-bearing sources show that the natural ground gas is air with a small
proportion of CO2, and CH4 below detection
Faults and permeable strata in the bedrock act as natural conduits for gas flow in the subsurface.
Similarly, mine shafts and other voids provide man-made channels for gas to flow to the surface. In
the instance of abandoned mine shafts, the height of the water table becomes an important factor in
gas movement. As the groundwater recharges and fills the mine void, a piston effect is achieved that
drives the gas towards the surface. However, as CO2 is 58 times more soluble in water at standard
temperature and pressure (STP) than CH4 (Hooker and Bannon, 1993) a proportion of this gas may
be dissolved and so removed from the system.
Examining the high temporal resolution data, it becomes clear that in addition to the potential
pathways that are available to CH4 and CO2 in the unsaturated zone, atmospheric pressure is a key
controller of gas movement. As air pressure falls, gas is released from the unsaturated zone. With
increasing atmospheric pressure, air is forced into the ground, thereby producing a diluting effect on
the concentration of CH4 and CO2. As was seen from the monitoring data, under a negative air
pressure gradient, pressure in a borehole builds up and creates a positive borehole flow. In other
words, the pressure in the borehole is greater than the atmosphere, giving a focused induced flow of
gas from the ground to the atmosphere.
The data collected from continuous monitoring show considerable variation with time, and are
consistent with the CH4 and CO2 data collected by hand-held meters. Using these, measurements are
made at specific times/dates. If the periodicity of the sampling points is superimposed on the high
frequency data series, gas composition is measured as a snapshot view of a highly variable system,
and measured concentrations will vary considerably (as shown in Figure 1 andFigure 2 in particular).
From a regulatory perspective, high temporal resolution allows a clearer understanding of the
processes that are occurring in the near-surface ground gas regime. As has been noted, air pressure
and depth of water table are key factors. Current UK practice is to take point measurements from all
monitoring wells on a site for a minimum investigation period. It is required that at least one of the
point measurements needs to be taken during falling atmospheric pressure.
The data presented here show high temporal variability. The periodicity of the cycling of gases
can be as short as a few hours to as long as a few days. Under current regulatory practice, two point
measurements would have been made during the selected two week period shown in Figure 2. For
70% of that monitoring period, only air (N2 + O2) was present in the test borehole. Thus there is a high
likelihood of missing an emission event. To be certain of the ground gas regime for a site where CH4
and CO2 are likely to pose a hazard, a high temporal resolution data set may be required.
Furthermore, a longer statutory monitoring period could be necessary to clarify short-term diurnal
trends and identify any longer-term seasonal/annual variations in the ground gas regime.
ACKNOWLEDGEMENTS
The authors would like to thank EPSRC for financial support in the form of a DTA studentship to CJT.
REFERENCES
Methods
The hydrogeological characterization of the chalk aquifer has comprised: Coring and installa-
tion of 7-channel CMT screens using Sonic Drill to 40 m bgs, geological logging and sub-
sampling of cores, synchronous water level measurements at different seasons, continuous
logging of water level in 11 screens at specific depths across the contaminated aquifer vol-
ume, pumping test, 11 HPT (Hydraulic Profiling Tool) logs to 47.8 m bgs with uup to 19 dissi-
pation test per log, installation of 6 nested piezometers using the GeoProbe system at 12
and 22 m bgs, respectively.
Results
The geology at the site consist of chalk overlain by quarternary deposits. The upper part of
the chalk exhibits evidence of deep glaciotectonic deformations. Chalk is a fractured media,
but by core sampling the chalk at the site appears loose, soft and unhardened. This means
that the fractures cannot be identified visually. Both upwards and downwards vertical pres-
sure gradients are observed in the chalk and the horizontal pressure gradients are generally
very small in the area.
The hydrogeological investigations show that the permeability is higher in the upper part of
the chalk (8-25 m bgs) than in the deeper part (25- >48 m bgs). There a significant horizontal
flow in the upper part of the chalk. This flow mainly occurs in the horizons of high permeabil-
ity that extend consistently over larger distances. Water level data from the pumping test
confirm the presence of coherent flow horizons, since the change in pumping rate was regis-
tered instantly (<30 minutes).
Consistent with these findings, the HPT logs show a significant shift in the vertical pressure
gradient at 25 m bgs throughout the area. Below this depth, the absolute pressure decreases
with depth indicating a downward vertical pressure gradient. This is confirmed by the in-
creased HPT pressure, indicating larger layers of low permeable material. The flow in the
section will primarily be downwards, which corresponds with the observed contaminant dis-
tribution. The downward vertical pressure gradient in the deeper part of the chalk indicate the
presence of a high permeable horizon at even larger depth, that drains the water from the
upper part of the chalk.
The results illustrate the difficulty connected with correlating the high permeable horizons,
observed in the HPT logs, and thus the difficulty in correlating corresponding depth for water
level measurements.
The chalk aquifer is under significant pressure and can thus be characterized as a confined
aquifer. Pressure changes in the aquifer propagates very fast even over large distances.
As expected, the investigations have not provided direct evidence of the presence or fre-
quency of fractures in the chalk, but the aquifer is nonetheless considered a dual porous,
fractured media based on indirect indications on hydraulic responses from HPT logs and dis-
sipation test as well as from background knowledge about chalk.
The large transient variability in the pressure levels caused by the extraction strategy in the
water supply wells together with the small horizontal gradients in the area have stressed the
importance of performing synchronous water level measurements within a very narrow time
frame. Continuous logging of the water level at specific depths across the chalk aquifer has
shed light on the vertical variability in the pressure propagation in the specific flow horizons.
The detailed investigations comprising HPT-logs, core sampling and depth specific screen
installation have illustrated the possibility that the water divide can be placed differently at
different flow horizons. This understanding challenges the common conceptual perception of
water divides and overall flow conditions and may have a significant importance at the site in
question.
Statistical distribution of left-censored concentration data Example from a soil
characterization study in Montreal
Email: niloofar.shoari.1@ens.etsmtl.ca
Abstract
Concentration data sets often contain left-censored observations, i.e. those concentration measurements
that fall below the limit of detection of analytical instruments. This fact complicates any statistical analysis
of data including the goodness of fit procedures. For this reason, environmental data are commonly
assumed to follow the lognormal distribution without performing any appropriate analysis to validate such
assumption. In this paper, we apply two goodness of fit procedures that can accommodate the problem of
left-censoring. These procedures employ the Akaike Information Criterion, and the Anderson-Darling test
statistic. Among three candidate distributions (lognormal, Weibull, and gamma), we identify the distribution
that fits the best to concentration data obtained from an existing site characterization study. Results show
no evidence that supports the common assumption about the lognormality of data. Moreover, we show the
inadequacy of substituting the left-censored observations with a constant using the Anderson-Darling test
statistic.
1. Introduction
In environmental studies, many statistical inference procedures rely on the model that is best supported by
the data. In numerous studies, environmental data are assumed to follow a lognormal distribution, merely
because the previous studies were based on this assumption; see El-Shaarawi, (1989); Johnson (1995),
Huybrecht (2002); Lu et al. (2003); Baccarelli (2005); Caudill (2007), and Leith (2010) among many others.
Nevertheless, some studies showed that other right-skewed distributions are more appropriate for
representing the frequency distributions of contaminants. For example, Singh et al. (2002) suggested
calculating the 95th upper confidence limit based on assuming a gamma distribution for environmental data.
Thompson (2000) assumed a Weibull distribution to estimate the distributional parameters of PAH in
sediment and oyster samples collected at a South Carolina estuary. Some other references are Tsukatani
and Shigemitsu (1980), Kan and Chen (2004), and EFSA (2010).
Its worth mentioning that previous studies were based on complete data sets; however, to our knowledge,
no study has explored the underlying distribution of contaminant concentration data due to complications
frequently caused by left-censored observations. These are concentrations that are below the limit of
detection (LOD) of measuring instruments and thus cannot be quantified. In practice, left-censored
observations are replaced by an arbitrary constant such as LOD/2, and the usual goodness of fit tests are
employed. Alternatively, environmental data are assumed to be lognormal by default, without further
supporting the appropriateness of this assumption. Either way, an inappropriate distribution may be used,
substantially altering the inferences and decisions.
Since the shape of the distribution of environmental data are right-skewed, lognormal, Weibull, and gamma
distributions are plausible candidates to model contaminants concentration data sets obtained from a soil
characterization study in Montreal, Canada. The objective of this study is to explore which of these
candidate distributions adequately represents left-censored contaminant data. We use two procedures to
judge the adequacy of fit of lognormal, Weibull, and gamma distributions: a) goodness of fit test based on
the Anderson-Darling (AD) test statistic, and b) Akaike information criterion (AIC). Finally, we demonstrate
the inappropriateness of substituting censored observations before data analysis through the AD test
statistic.
2. Methodology
2
The concentration data is said to follow a lognormal distribution, ~( , ), with mean and
2
variance in logarithmic scale, and the pdf is given by
2)
1 1 2
(| , = exp[ ( ) ]
2
2 2
2)
(| , = ( )
2
The mean and variance 2 in linear scale are
2
= exp ( + 2),
2
2 = 2 (exp( ) 1)
When concentration data follow a Weibulll distribution, ~(, ), the pdf is given as
1
(|, ) =
exp [ ( ) ] > 0, > 0
where parameters and are the shape and scale parameters, respectively. The cdf of the Weibull
distribution is defined as
(|, ) = 1 ()
1
= (1 + )
2 1
2 = 2 { (1 + ) [(1 + )]2 }
() = 1 , > 0
0
A random variable is said to follow a gamma with shape, , and rate, , shown as ~(, ), if it pdf
is given by
1
(|, ) = > 0, > 0
()
2 = 2
2.2 Estimation methods for left-censored data
2.2.1 Maximum Likelihood Estimation (MLE)
This method utilizes a likelihood function to estimate distributional parameters, , of a population. Let
1 , 2 , , be contaminant concentrations and let = (1 , 2 , , ) denote the vector of detection
limits below which measuring instruments cannot detect concentration values. The obtained concentration
data set consists of pairs ( , ), = 1,2, , , where indicates whether an observation is censored ( =0)
or uncensored ( =1). For left-censored data, the likelihood function is given by
( ; ) ( ; )1
=1
where ( ; ) is the density function (pdf) of an assumed parametric family of distributions and the ( ; )
is the cumulative density function (cdf) of censored observations associated with the assumed distribution.
Maximization of the likelihood functions provides estimates of the distributional parameters.
This method assumes that data follow a lognormal distribution. The logarithm of observations, including the
censored ones, are plotted against their normal scores resulting in the model
ln( ) = + 1 ( )
where 1 ( ) refers to the inverse cumulative normal distribution for plotting positions . The estimates
of the mean and standard deviation ( and ) are employed to predict censored observation. The
predicted observations are combined with uncensored ones resulting in a complete data set.
This method relies on estimating the cdf, (x) = Pr(X x) , in a non-parametric way, meaning that it does
not rely on a parametric distributional assumption. The F(x) determines the probability that a
concentration is less than or equal to a certain value x.
The parametric statistical methods to handle left-censored data (i.e., MLE and rROS) require assuming a
probability density function for the data. The most commonly assumed distribution in environmental
literature is the lognormal distribution. However, the right-skewed shape of concentration data suggests
that they can potentially be represented by a Weibull or gamma distribution. We evaluate the goodness of
fit of lognormal, Weibull, and gamma distributions based on the Akaike information criterion (AIC), and the
Anderson-Darling (AD) test statistic.
2.3.1 Akaike Information Criterion (AIC)
The AIC criterion assesses how well a certain model fits the data, penalizing the model with more
parameters. In this study, we calculate the AIC for the ith models as
= 2 2
where defines the likelihood of fitting the ith model, and is the number of parameters in that model.
Among the three values, the smallest indicates the preferred model, . The AIC difference for
each model is given by = . To select the best model among the three candidates, we
calculate the Akaike weights ( ) as
i
exp( 2)
=
3=1( 2)
The values indicate the probability that a given model is the best among all models evaluated herein.
In goodness of fit test based on the AD test statistic, the null hypothesis assesses whether a random sample
of 1 , 2 , , comes from a specified distribution,(|) with distributional parameter , i.e., 0 :
(|). The (|) in our study refers to cdf of lognormal, Weibull, and gamma distributions described
earlier in section 2.1. The AD test statistic measures the squared distance between the hypothesized and
the empirical (based on the Kaplan-Meier estimate) cumulative distribution functions.
Let (1) < (2) < < () < (+1) < (+2) < < be an ordered data set in which the
smallest observations are left-censored and the remaining r observations are uncensored. For
convenience, assuming that (.) is the cdf of an assumed distribution evaluated at (.) , we have (1) < (2) <
< () < (+1) < (+2) < < () . DAgostino and Stephens (1986) define the computing formula
for the AD statistic for single censored data as
+1 +1
1
= (2 1){ln[1 (+1) ] ln (+1) } 2 ln (+1)
=1 =1
1
[( )2 ln (+1) 2 ln (+1) + 2 (1 (+1) ]
We have applied the afore-mentioned goodness of fit criteria to concentration data sets obtained from soil
samples collected for a site characterization study conducted in Montreal, Canada. Data sets consist of
concentration measurements of 22 contaminant; the sample size varies between 13 and 62 observations,
and the censoring percentage ranges between 4% and 72%. The results of the application the AIC and AD
goodness of fit are reported in Table 1. The model selection based on Akaike weights (4th column in table
1) indicates that the Weibull distribution generally has a higher chance of being the best one among those
considered in the set of candidate models. On the other hand, the results based on the AD test statistic (5th
column in table 1) clearly show that lognormal, Weibull, and gamma often provide good fit to the majority
of contaminants (p-values > 0.05). Therefore, there is no evidence that concentration data exclusively follow
a lognormal (or any other) distribution. Moreover, larger p-values are often obtained when a Weibull
distribution is fitted to data, suggesting that Weibull is the most appropriate model among the other
candidates. This result is consistent with the model selected previously based on Akaike weights.
Since substitution of left-censored observations with a constant is widely practiced, we also investigated
the effect of substitution on the goodness of fit procedure. For all contaminant data, the censored
observations are substituted with a LOD/2 and the goodness of fit test based on AD statistics for complete
data is performed. The resulting p-values are very small (always <0.05), suggesting the inappropriateness
of all of the models under study. Therefore, censored observations have to be taken into account in order
to correctly identify the underlying distribution of data. Although previous research has shown that
substitution method is not appropriate for environmental censored data (e.g., El-Shaarawi and Esterby
(1992), Helsel (2006)), we confirm this finding using a model selection criterion; that is the Anderson-Darling
test statistic. Note that previous research have mainly examined the performance of substitution using data
simulation studies.
Tabel 1. Comparison of models used to evaluate the contaminant concentration distributional form using
Akaike Information Criterion (AIC) and Anderson-Darling (AD) test statistic a
Sample size
Contaminant model Akaike weight P-values of AD
(censoring %)
Lognormal 0.33 0.22
36
Cadmium Weibull 0.34 0.12
(72%)
Gamma 0.34 0.09
Lognormal 0.46 0.23
13
Mercury Weibull 0.32 0.07
(4%)
Gamma 0.23 0.07
Lognormal 0.27 0.01
36
Molybdenum Weibull 0.39 0.02
(42%)
Gamma 0.35 0.02
Lognormal 0.34 0.36
36
Tin Weibull 0.36 0.19
(28%)
Gamma 0.30 0.13
Lognormal 0.06 0.06
46
Benzo fluoranthene Weibull 0.48 0.56
(22%)
Gamma 0.46 0.70
Lognormal 0.07 0.03
62
Indeno.pyrene Weibull 0.68 0.54
(37%)
Gamma 0.25 0.17
Lognormal 0.10 0.05
62
Benzo perylene Weibull 0.70 0.52
(37%)
Gamma 0.19 0.18
Lognormal 0.23 0.11
62
Benzo pyrene Weibull 0.74 0.76
(27%)
Gamma 0.03 0.05
Lognormal 0.13 0.12
62
Fluoranthene Weibull 0.84 0.80
(21%)
Gamma 0.03 0.07
Lognormal 0.52 0.30
62
Fluorene Weibull 0.47 0.52
(37%)
Gamma 0.01 0.01
62 Lognormal 0.56 0.39
Methyl1naphtalene (47%) Weibull 0.39 0.33
Gamma 0.05 0.03
Lognormal 0.62 0.58
62
Methyl2naphtalene Weibull 0.35 0.31
(47%)
Gamma 0.03 0.01
Lognormal 0.27 0.72
60
Naphtalene Weibull 0.53 0.59
(55%)
Gamma 0.21 0.05
Lognormal 0.24 0.08
Dibenzo (ah) 61
Weibull 0.69 0.22
anthracene (46%)
Gamma 0.08 0.03
Continued
Sample size
Contaminant model Akaike w P-values of AD
(censoring %)
Lognormal 0.44 0.54
62
Phenanthrene Weibull 0.56 0.96
(16%)
Gamma 0.0 0.01
Lognormal 0.15 0.14
62
Pyrene Weibull 0.81 0.85
(21%)
Gamma 0.04 0.09
Lognormal 0.23 0.39
61
Acenaphtene Weibull 0.65 0.72
(51%)
Gamma 0.12 0.05
Lognormal 0.55 0.44
62
Acenaphtylene Weibull 0.41 0.37
(53%)
Gamma 0.04 0.03
Lognormal 0.12 0.12
62
Anthracene Weibull 0.76 0.60
(40%)
Gamma 0.12 0.09
Lognormal 0.02 0.04
Benzo (c) 61
Weibull 0.17 0.28
Phenanthrene (49%)
Gamma 0.81 0.72
Lognormal 0.17 0.09
62
Benzo (a) anthracene Weibull 0.79 0.79
(27%)
Gamma 0.04 0.07
Lognormal 0.33 0.16
62
Chrysene Weibull 0.65 0.80
(23%)
Gamma 0.01 0.03
a The best model for each compound is italicized and in bold font.
4. Conclusions
Parametric inference relies on the knowledge of the form of the population distribution. While environmental
data sets are commonly assumed to be lognormally distributed, they may have arisen from other right-
skewed distributions. Assessing the distributional form for uncensored data is relatively easy; however,
complications occur when left-censored observations are present. This study evaluated the distribution of
environmental data containing left-censored values through two procedures of goodness of fit assessment:
a) the Akaike information criterion, and b) the Anderson-Darling test statistic. The data sets used herein are
concentration measurements obtained from a site characterization study conducted in Montreal, Canada.
The models under study are Weibull, gamma, and lognormal distributions. The following conclusions were
derived:
The lognormal distribution is not always appropriate to model the concentration data. In fact, for
the majority of contaminants in this study, the Weibull distribution better fits the data compared to
the lognormal and gamma distributions.
Substitution of censored values with a constant should be avoided since this procedure modifies
the true underlying distribution of concentration data.
Reference
Baccarelli, A.; Pfeiffer, R.; Consonni, D.; Pesatori, A. C.; Bonzini, M.; Patterson, D. G.; Landi, M. T. Handling
of dioxin measurement data in the presence of non-detectable values: overview of available methods and
their application in the Seveso chloracne study. Chemosphere, (2005) 60(7): 898-906.
Caudill, S. P.; Wong, L. Y.; Turner, W. E.; Lee, R.; Henderson, A.; Patterson, D. G. Percentile estimation
using variable censored data. Chemosphere, (2007) 68(1): 169-180.
D'Agostino, R.B.; Stephens M.A. Goodness-of-fit Techniques Marcel Dekker, New York (1986)
ElShaarawi, A. H. Inferences about the mean from censored water quality data. Water Resources
Research, (1989) 25(4): 685-690.
European Food Safety Authority, Management of left-censored data in dietary exposure assessment of
chemical substances. EFSA Journal (2010), 8: 1-96.
Helsel, D. R; Fabricating data: How substituting values for nondetects can ruin results, and what can be
done about it, Chemosphere (2006): 65 (11), 2434-2439, ISSN 0045-6535
Huybrechts, T.; Thas, O.; Dewulf, J.; Van Langenhove, H. How to estimate moments and quantiles of
environmental data sets with non-detected observations? A case study on volatile organic compounds in
marine water samples. Journal of Chromatography A, (2002) 975(1): 123-133
Johnson, R.A.; Gan, D.R.; Berthouex, P.M. Goodnessoffit using very small but related samples with
application to censored data estimation of PCB contamination. Environmetrics (1995): 341-348.
Kan, H. D.; Bing-Heng Chen. Statistical distributions of ambient air pollutants in Shanghai,
China. Biomedical and Environmental Sciences (2004): 366-372.
Leith, K. F.; Bowerman, W. W.; Wierda, M. R.; Best, D. A.; Grubb, T. G.; Sikarske, J. G. A comparison of
techniques for assessing central tendency in left-censored data using PCB and p, p DDE contaminant
concentrations from Michigans Bald Eagle Biosentinel Program. Chemosphere, 2010 80(1): 7-12.
Lu, H-C; Chuen-Jinn T; I-Fu Hung. Atmospheric lead concentration distribution in Northern Taiwan.
Chemosphere (2003): 1079-1088.
Singh, A.; Singh, A. K.; Iaci, R. J. Estimation of the exposure point concentration term using a gamma
distribution. Technology Support Center Issue, EPA/600/R-02/084, EPA Technology Support Center for
Monitoring and Site Characterization, Las Vegas, Nevada (2002)
Thompson, R. E., Voit, E. O.,Scott, G. I. Statistical modeling of sediment and oyster PAH contamination
data collected at a South Carolina estuary (complete and leftcensored samples). Environmetrics, (2000),
11(1): 99-119.
Tsukatani, T.; Shigemitsu, K. Simplified Pearson distributions applied to air pollutant concentration.
Atmospheric Environment (1980), 14(2): 245-253.
DETERMINATION OF SALT BUDGET FOR LARGE IRRIGATED AREAS
1,* 2 2
Harun Kaman , Mahmut etin , Cevat Krda
1
Akdeniz University, Faculty of Agriculture, Department of Agricultural Structures and Irrigation, 07058
Campus, Antalya, Turkey
2
Cukurova University, Faculty of Agriculture, Department of Agricultural Structures and Irrigation,
01330 Balcali, Adana, Turkey
*
Corresponding author: hkaman@akdeniz.edu.tr
Abstract
In this study, quantification of salt budget elements in large-scale irrigated areas was investigated. The
study was carried out in the Akarsu Irrigation District, covering 9495 ha command area, of Lower
Seyhan Plain in the Mediterranean coastal plain located in Southern Turkey. The farmers in the District
commonly use surface irrigation methods. The field studies were carried out in 2008 hydrological year.
In order to determine the soil salinity (ECe) change in crop rooting zone, soil saturation, porosity and
bulk density maps for 0-0.9 m soil profile were produced in GIS media. Then, ECe and total dissolved
-1
salts (TDS in mg l ) of soil samples taken from 0-30, 30-60 and 60-90 depths were determined in the
laboratory. Regression analysis was done, and a meaningful relation was found between TDS and
ECe. Finally, TDS in crop rooting zone was estimated by using the relationship between ECe and
TDS. The map calculations and zonal statistics of GIS showed that the net change in soil salinity
-1
(STDS) over 0-0.90 m soil depth was STDS=-0.026 Mg ha in 2008, indicating a small decrease in
soil salinity. Salt balance equation was resulted in a closure error in the order of the magnitude of S=-
-1
2.84 Mg ha . This indicates that the amounts of salts exported are greater than the salts imported. In
this study, the salt contribution of fertilizer applications to the soil and GW system was overlooked;
furthermore, the salinity change in the soil layers between GW table and root zone was not considered
in the mass balance calculations. These two potential factors may be the source of this closure error.
Eighty-one percent of S occurred in NIS, indicating mostly GW contribution, leaching and washing
out of the salts by precipitation in winter.
Keywords: Irrigation, crop rooting zone, salt balance, hydrologic year.
Introduction
Lack of rainfall or insufficient rainfall in the months when it is time for plants to grow, necessitates
the use of irrigation methods in agricultural production. Therefore, irrigation is a method of
guaranteeing an on-going agricultural production in high amount. As in some other agricultural
applications, various problems may emerge in due course of irrigation practices. These problems
result from the unbalance among soil-plant-water relations. The fact that farmers lack of enough
knowledge and experience about when to grow, how much to produce and which method to use using
irrigation method, leads to improper irrigation practices. In addition to that, the use of primitive poor
performer irrigation methods by the farmers is also very common. This not only hinders sustainable
irrigated farming but also leads to salinity and alkalinity problems, and production capacity of soil
becomes unusable. Over irrigation leads to the rise of ground water and this then increases the
salinity rate of the soil in time (etin and Krda, 2003). In dry periods, the ground water level increases
up to the crop rooting-zone and becomes critically important for glycophytes which are the large
majority of plant species not salt-tolerant and damageable fairly easily by high salinity. On the other
hand ground water level in the irrigation districts is closely related to irrigation method (DS, 1982;
etin and Diker, 2003) which accelerates the extent and severity of salinisation.
Monitoring and evaluation of ground water level and ground water salinity for the purpose of
sustainable agricultural production and water management is a necessity in agriculture. With the
precautions taken considering the findings obtained as a consequence of this monitoring and
evaluation can help solve the irrigation related problems (Bykcangaz and Deirmenci, 2002). The
level of ground water and its salinity rate and soil salinity are considered to be among the most
significant factors affecting agricultural production. If, in dry spells, the ground water levels increases
up to the root level and reaches to a critical value, it can be expected that ground water level is closely
related to the irrigation; therefore, that irrigation rate is low (DS, 1982; etin and Diker, 2003).
Lower Seyhan Plain is one of the largest areas of irrigation in Turkey. Irrigation has been
gradually introduced to the plain since the construction of Seyhan Regulator in 1942, and with the
construction of Seyhan Dam in 1956 (etin and zcan, 1999). Since then, important gains have been
achieved regarding irrigation agriculture. However, with the launch of irrigated farming in the region,
the problems such as drainage, salinity and alkalinity in the irrigated and non-irrigated fields started to
emerge (etin and zcan, 1999).
This research was carried out in Akarsu Irrigation District. The farmers in the Lower Seyhan Plain
generally employ surface irrigation method which is a poor performing method. Drainage and ground
water level related problems have emerged as a result of the deficiency in the irrigation methods and
poor irrigation performance, the soil structure and lack of enough drainage systems and etc. These
problems negatively affect high potential of agricultural production. This study aims to investigate the
quantification of mass balance elements of salinity in Akarsu Irrigation District.
Following the land survey, soil samples are collected from the region and observation points were
determined for ground water level in the region (Figure 2). Water samples were biweekly collected
from the irrigation waters diverted to the District, and daily samples from precipitation events and
drainage waters at the catchment outlet. Additionally, 108 drainage observation wells were taken
under review. Groundwater samples from 108 GW observation wells were collected and analyzed for
salinity. The chemical analysis including electrical conductivity (EC) and total dissolved salts (TDS) of
water samples were done in the lab. Depths to groundwater table in total of 108 GW observation wells
were also measured, concurrently with the collection of GW samples.
In order to determine the salinity change in crop rooting zone, first of all, soil saturation, porosity
and bulk density maps for 0-0.9 m soil profile were produced using GIS media. Then, the ECe and
-1
total dissolved solids (TDS in mg l ) of soil samples (0-30, 30-60 and 60-90 cm depth) taken nearby
GW observation wells were determined in the laboratory. Regression analysis was done, and a
meaningful relation was found between TDS and ECe.
Figure 2. Spatial distribution of soil and groundwater sampling sites in Akarsu Irrigation District.
Table 1. The lands covered by the plants cultivated in the research region (%).
Crops % of total irrigated land
Wheat 17.7
Citrus 29.2
1. crop corn 39.6
2. crop corn 1.3
Cotton 8.5
Melons 0.5
Vegetables 0.3
Others 4.2
The research area has had a drainage and irrigation system for about 50 years. In the regions
where irrigation applications are densely performed, the problems related to the increase in the level
of ground water and salinity may emerge. In this regard, some information related to the salt budget
can be seen in Table 2 prepared to cover the whole hydrological year, irrigation season (IS) and non-
irrigation season (NIS) of 2008 so that one can make comparisons between the periods.
-1
Table 2. Salt budget elements in Mg ha unit determined for the whole observation period plus
irrigation and non-irrigation season of the 2008 hydrological year.
-1
Mass of salts (Mg ha ) Hydrologic Irrigation
Non-IS
Year Season (IS)
Precipitation (SP) 0.989 0.242 0.747
Irrigation (SI) 3.874 3.506 0.368
Drainage (SQ) 7.714 4.293 3.420
Change in 0.9 m soil depth (STDS) -0.026 - -
Change in groundwater (GWTDS) 3.112 - -
Subterranean lateral flows (SLFTDS) +0.010 +0.008 +0.002
Salt balance error (Closure Error)
-2.841 -0.538 -2.304
[S=SI+SP+SLSF-SQ]
In the hydrological year 2008, TDS maps for 0-0.9 m soil depth produced for early October/07
-1
and late September/08 were converted to the Mg ha unit by utilizing soil saturation, porosity and bulk
density maps in GIS media. The map calculations and zonal statistics showed that the net change in
-1
soil salinity (STDS) over 0-0.90 m soil depth was STDS=-0.026 Mg ha , indicating a small
decrease in soil salinity which is desirable in irrigated agriculture.
On the other hand, the net GW salinity change over the 3 m soil layer, corresponding the
-1
designed depth of open drainage canals, was GWTDS=+3.112 Mg ha , an indication of an increase
-1
in GW salinity. Imported salt by subsurface lateral flows was nearly nil (0.01 Mg ha ). However, the
-1
salt load in irrigation water and precipitation were calculated as 3.87 and 0.99 Mg ha , respectively,
-1
indicating that yearly total salt imported was about 4.86 Mg ha by the two events, precipitation and
irrigation. The percentages of salts imported by irrigation and precipitation were 10 and 76,
respectively, in NIS.
-1.
Salt balance equation was resulted in a closure error of S=-2.84 Mg ha This figure
indicates that the amounts of salts exported from the area by drainage are greater than the salts
imported to the District.
Conclusion
In this study, the salt contribution of fertilizer applications to the soil and GW system was
overlooked; furthermore, the salinity change in the soil layers between GW table and root zone was
not considered in the mass balance calculations. These two potential factors may be the source of this
closure error. Eighty-one percent of S occurred in NIS, indicating mostly GW contribution, leaching
and washing out of the salts by precipitation in winter.
Acknowledgement
Authors gratefully acknowledge that this work was funded by the European Union QUALIWATER
project (INCO-CT-2005-015031) and partial support received from Cukurova University Research
Fund (Project #: ZF2006KAP1 & ZF2009KAP1). The authors also would like to thank the Research
Fund of Akdeniz University for its partial support.
References
Bykcangaz, H., Deirmenci, H. 2002. Drenaj sularnn sulamada yeniden kullanlmas. (Turkish) Su
Havzalarnda Toprak ve Su Kaynaklarnn Korunmas, Gelitirilmesi ve Ynetimi Sempozyumu,
1820 Eyll 2002, Antakya, s. 614617.
Din, U., Sar, M., enol, S., Kapur, S., Sayn, M., Derici, M., avugil, V., Gk, M., Aydn, M., Ekinci,
H., Aca, N., Schlichting E. 1995. ukurova Blgesi Topraklar. (Turkish) ukurova niversitesi
Ziraat Fakltesi yardmc ders kitab, No:26, Adana.
DS, 1982. ASO IV. Merhale Projesi Planlama Drenaj Raporu. (Turkish) Ankara.
etin, M., Diker, K. 2003. Assessing drainage problem areas by GIS: A case study in the Eastern
Mediterranean Region of Turkey . Irrig. and Drainage 52:343353.
etin, M., Krda, C. 2003. Spatial and temporal changes of soil salinity in a cotton field irrigated with
low-quality water. Journal of Hydrology 272: 238249.
etin, M., zcan, H. 1999. Problems encountered in the irrigated and non-irrigated areas of the Lower
Seyhan Plain and recommendations for solution: a case study. (Turkish with English Abstract)
Tr. J. of Agr. and For. 23 (1):207217.
DETERMINATION OF IRRIGATION WATER SALINITY: A CASE STUDY
1,* 1 2
Harun Kaman , Abdullah Sayc , Seluk zmen
1
Akdeniz University, Faculty of Agriculture, Department of Agricultural Structures and Irrigation, 07058
Campus, Antalya/Turkey
2
Dzce University, Faculty of Agriculture, Department of Biosystem Engineering, 81620, Dzce/Turkey
*
Corresponding author: hkaman@akdeniz.edu.tr
Abstract
Salinity has been the most important problem of irrigated farming, and it is basically a consequence of
irrigation practices. Salinity decelerates plant growth and decreases the yield. The improper practices
in irrigation, the insufficient or over use of water decrease the efficiency of water and decrease the
yield. On the other hand, climate changes have caused increase of temperatures and decrease of
water resources in spite of increase demand of water by all sectors. Increased population and
industrial development cause additional strain on availability of water. Therefore it is important to
determine irrigation water salinity in irrigated agriculture which is the largest user of available water
resources. The work undertaken here investigates irrigation water salinity under an existing irrigation
management in a large scale irrigation district. The study was carried out in Karaman-Ermenek District
in Turkey. The farmers in the District commonly use conventional irrigation methods. For irrigation
water salinity, water samples were collected from 19 wells randomly selected sites, representing the
whole area. Analyses of irrigation waters EC (S/cm) were done. Results showed that mean EC was
261 S/cm. EC of water samples exhibited range of variability from 112 to 580 S/cm. As a result,
irrigation water salinity has quite low a level.
Keywords: Water quality, groundwater well, salinity.
Introduction
Soil and water resources are considered to be among the most important natural richness. In
the socio-economic developments of the communities, it is very important to improve the resources
and benefit from them wisely. Water is a natural resource without which all living creatures cannot live.
When we lack of it, the amount of plant production significantly decreases. Agriculture, which is the
most water consuming sector, covers all tasks related to the production of cultivated plants. Agriculture
is responsible for 19% of total national income, for 9% of the total export. Agriculture provides jobs for
51% of the community (Klner et al., 2002).
The predictions that suggest world population will be 8 billion in 2025, makes it clear that food
security will be the most serious problem of the world in the future. To meet the needs of the
increasing world population, there is a need for a two-time increase in production in 50 years (Howell
et al., 2001).
For people to meet their food needs in a safe way is closely related to the increase in
agricultural production and in the size of agricultural lands. There was a 1% increase in the amount of
land where irrigated farming is done in 2000s, but it is suggested that this increase needs to be at
least 2.25% (FAO, 1988). According to the recent predictions, it has been reported that there will be a
two times more need for food, clothing, housing and fresh water in 2050 (Postel et al., 1996).
The predictions suggest that the number of people who are under the risk of hunger and
malnutrition is 800 million in developed countries. The solution to the food safety problems of the
people living in these areas is dependent on the solution of water management in countryside
(Rockstrm, 2003).
As known, the most effective factor limiting the growth of plants in dry lands and semi-arid lands
is to lack of enough water needed for growing in the root zone (Falkenmark and Rockstrm, 1993; Lal,
1991). Therefore, it is a compulsory to carry out irrigated farming in dry and semi-arid areas. It is
expected that the sizes of lands will extend and irrigated farming will increase accordingly, and
effective use of water will increase food production (Yudelman, 1994) and as a consequence of these,
the need for water will increase in parallel to the increase in the world population. However, the
increase in water consumption causes very significant problems. For example, underground water
resources are used up, other water ecosystems are polluted; besides, irrigated farming causes many
other environmental problems. As can be understood from the reasons suggested above, the share of
irrigations is very big in agricultural production. On the other hand, many other problems also emerge
in irrigated farming, such as salinity and etc. With this regard, the purpose of this study is to
investigate the electrical conductivity of the irrigation water in the land where this study was carried
out.
600
Irrigation water EC (S/cm)
500
400
300
200
100
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Irrigation water wells no
Figure 1. EC analysis of water samples.
With this regard, the electrical conductivity of the irrigation water that we have investigated in
this study, was found to be C1 and C2. As reported by ktren Asri et al (2010), the irrigation water in
C2 class does not constitute any problem with regards to salinity.
In a study carried out by Arslan et al (2007) in Bafra Ovas to investigate the quality of
underground water in the region, climate parameters and soil and water resources were investigated.
In another study which was carried out by rtem and Sar (2011) to examine the irrigation system in
Balkesir Plain, the following issues were investigated, such as the plantation rates by year together
with weather data without any irrigation and soil parameters, planned and earned crop pattern
irrigation yield and irrigation planning. There are a lot of similar studies in the literature investigating
the irrigation practices in one region. Some of them were planned as narrow-scoped ones and some
of them were planned as in-depth studies considering the facilities available to use. This study has
been planned as a preliminary study and planned as a narrow scoped one.
Conclusion
In the study that we carried out as researchers, the findings obtained with the analysis of the
irrigation water, did not come up with any factor which may negatively affect the yield. However, as
can be understood from the scope of the study, we made an analysis of irrigation water for one
season. As known, the quality of the irrigation water with regards to EC is not the only factor affecting
plant production. There is a need for the anion and cation analysis of the soil. In brief, it is very
beneficial to monitor both the soil and irrigation water periodically.
Acknowledgement
The authors would like to thank the Research Fund of Akdeniz University for its partial support.
References
Arslan H., Gler M., Cemek B., Demir, Y. 2007. Assessment of groundwater quality in Bafra Plain for
irrigation. (Turkish with English Abstract) Journal of Tekirdag Agricultural Faculty, 4(2): 219-226.
Falkenmark M., Rockstrom J. 1993. Curbing rural exodus from tropical drylands. AMBIO-0122 no
71993
FAO. 1988. World Agriculture Toward 2000: A FAO Study N. Alexandratos (ed.) Bellhaven Press
London 338 s.
Howell T.A., Evett S.R., Tolk J.A. 2001. Irrigation Systems ans Management to Meet Future Food
Fiber Needs and to Enhance Water Use Efficiency. USDA-ARS Water Management User Unit
Bushland Texas USA.
rtem E., Sar T. 2011. Investigation of Balkesir Plain irrigation. (Turkish with English Abstract) J. Fac.
Eng. Arch. Gazi Univ., 26(2): 461-469.
Klner N., akmak ., Eri A., Kanber R., Knac E., Yurdakul O. 2002. TBTAKn Tarm Sektrne
Ynelik Yaklam Ve Politikalarn Belirlemesine likin Yaplan Deerlendirme almas.
(Turkish) TBTAK-TOGTAG, ittage Raporu. Baslmam. Ankara, 146 s
Lal R. 1991. Current Research on Crop Water Balance and Implications for the Future. In: Soil Water
Balance in the Soudano Sahelian Zone. Eds.
ktren Asri F., Demirta E.I., Ar N., Arpacolu A.E., zkan C.F. 2010. Determination of irrigation
water qualities in greenhouses of Antalya-Serik region. (Turkish with English Abstract) Akdeniz
niversitesi Ziraat Fakltesi Dergisi, 23(2): 145-150.
ktren Asri F., Demirta E.I., Ar N., zkan C.F. 2013. Determination of irrigation water qualities of
Bilecik-Osmaneli district. (Turkish with English Abstract) Akdeniz niversitesi Ziraat Fakltesi
Dergisi, 26(1): 49-55.
Postel S.L., Daily C.D., Erlich P.R. 1996. Human Appropriation of Renewable Fresh Water Science.
Vol. 271. No. 5250. Issue 9. American Association for the Advencement of Science. s. 785-799.
Rockstrm J. 2003. Rewsilence Bilding and Water Diment Management for Drought Mitigation.
Physics and Chemistry of the Earth. 28: 869-877
Yudelman, M. 1994. Feeding the World. Int. Irrig. Manage. Institute Rev. 8 (1) 415.
INVESTIGATION OF IRRIGATION WATERS QUALITIES IN ANTALYA-KUMLUCA REGION,
TURKEY
1,* 2 1 1
Harun Kaman , Seluk zmen , Hasan etin , Abdullah Sayc
1
Akdeniz University, Faculty of Agriculture, Department of Agricultural Structures and Irrigation, 07058
Campus, Antalya/Turkey
2
Dzce University, Faculty of Agriculture, Department of Biosystem Engineering, 81620, Dzce/Turkey
*
Corresponding author: hkaman@akdeniz.edu.tr
Abstract
Both water and population distribution is not parallel in the World, water resources in many different
parts of the world are not fairly available. Therefore, water shortage in some parts of the world and
some countries has been an up to date problem. On the other hand, salinity has been the most
important problem of irrigated farming. Salinity decelerates plant growth and decreases the yield. The
main objective of this paper was to investigate potential use of irrigation waters in large-scale irrigation
schemes. The study was carried out in Antalya-Kumluca district in Turkey. Antalya, located in the
South West of Turkey by the Mediterranean coastline, has significant agricultural contribution to the
economy. Almost 31% of the Province population settled in the countryside and they are all engaged in
farming. Agriculture in Antalya is very strong in this region with regards to quality, yield and agro
biodiversity. For irrigation waters qualities of a large area, water samples were collected from 40
irrigation cannel and wells randomly selected sites, representing the whole area. Analyses of irrigation
waters including pH and EC (S/cm) were done in the laboratory. Results showed that mean pH and
EC were respectively 7.7 and 802 S/cm. EC of water samples exhibited range of variability from 190
to 1736 S/cm. In the same way, pH of water samples exhibited range of variability from 7.3 to 8.2.
Keywords: Irrigation cannel, irrigation wells, EC, pH.
Introduction
Water, which is considered to be a renewable resource, has been under the risk of losing this
characteristics in some fields. As a consequence of that, the provision of new water resources and
developing them for use has been very costly, and even impossible. What is worse is that, majority of
the society shows no interest in the fact that water will surprisingly be an inhibiting factor in food
production in the future (IFPRI, 2004). On the other hand, salinity appears the biggest environmental
problem in irrigation practices. Salinity happens when inappropriate irrigation method is used and
when the drainage system of the land is not good enough (Ghassemi et al., 1995). According to FAO
predictions, almost half of the irrigated lands is under the risk of salinity and alkalinity which are
defined as silent enemy. In many studies carried out in the field, it is pointed that about 20-30 million
hectare irrigated land causes significant decrease in crop yield as a consequence of salinity (El-Ashry,
1991; Rhoades, 1987; Kayasseh and Schenk, 1989). Besides, according to recent predictions, the
amount of irrigated land size where there was a decrease in crop yield in 1980s was 30-46 million
hectare. Similar case is true for Turkey too. There has been a significant amount of salinity and
sodification problem in the irrigated lands.
As a consequence of rapid population increase and industrial development, the increase in the
demand for water, the increase in the demand for water as a consequence of rapid increase in
population and industrial developments, there is a need for a national plan to sustain the proper use of
water resources and match it with the term of sustainable development to be able to stand against
qualitative and quantitative deterioration of water, environmental pollution and possible climatic
changes. All type of limitations set for development of water resources and their proper use, can be
removed with the help of a national plan. An appropriate observation network can be developed, and
improper technology can be improved.
Production is made in field crop cultivation, and garden cultivation is made in some parts of
Turkey and cultivation is made in greenhouse. On the other hand, as suggested by Snmez and
Kaplan (2004), the use of hotbed should be limited in the regions where intensive agriculture is
commonly practiced. In other words, a hard work should be done on the soil surface to be able to
harvest the best crops. It is stated that this case leads to deterioration in the quality of irrigation water.
Similarly, in a study carried out by zmen (2013), it was found that a lot of problems emerged in the
irrigation of Dzce Plain depending on the characteristics of the soil cultivated, the quality of the water
and the technology employed in cultivation. The research and findings related to irrigated farming are
quite common in literature. Therefore, the relationship between soil-plant-water need to be carefully
examined to increase agricultural production.
As explained above, conservation of soil and water resources and their development and
making them sustainable are extremely important. With this regard, in this study, it was the aim to
determine the quality of irrigation waters with regards to pH and EC in the region where plant
production is mostly done in greenhouses.
8.5
pH of irrigation water
8.0
7.5
7.0
6.5
EC analysis of the irrigation water was given in Figure 2. EC values of the irrigation water
samplings changes between 190 S/cm and 1736 S/cm. Average EC value is 802 S/cm. As
suggested by ktren Asri et al (2010; 2013), EC value of an irrigation water is 250>, it is class C1, if
EC value is between 250-750, it is class C2. If the EC value is between 750 and 2250, it is class C3.
As ktren Asri et al (2010) claimed, the irrigation water resources which are in class C1 and C2
cause no salinity problem. It is suggested that if the irrigation water is in class C3 in a region, drainage
system should be appropriate and the plants to be cultivated need to be salt resistant at average level
(ktren Asri et al., 2010).
In many studies in which agricultural irrigation is investigated, samplings are taken from the soil
where cultivation will be done as much as possible. For example, in a study carried out by Snmez
and Kaplan (2004), irrigation waters and the changes in salinity were investigated in three different
periods. In the study carried out by Arslan et al (2007) in Bafra Plain to evaluate underground water in
the region with regards to irrigation quality, soil and water resources were investigated as well as
climate parameters. In the study carried out by rtem and Sar (2011) in Balkesir Plain to investigate
the irrigation system in the region , plant vegetation rates, planned and earned plant pattern, irrigation
yield and irrigation plan were investigated. There are many studies investigating irrigation systems of
some regions with similar findings to this study in the literature. Some of the researches are narrow-
scoped, and some of them are in-depth studies considering the facilities provided. This study is a
preliminary one, and it is narrow-scoped.
2000
EC of irrigation water (S/cm)
1600
1200
800
400
Conclusion
According to the findings of the study, it is concluded that the pH values of irrigation waters will
cause no problem for plant production. However, if C3 class irrigation waters are used with regards to
electrical conductivity, it is suggested that the plants to be grown need to be salt-resistant at more than
average level.
On the other hand, pH and electrical conductivity values of irrigation waters were determined
and evaluated for one season. Soil samples need to be taken as well as irrigation water samples. In
these samplings, all anion and cation analysis should be done as well as pH and electrical
conductivity. In conclusion, all analysis mentioned here need to be investigated together with plant
species and types.
Acknowledgement
The authors would like to thank the Research Fund of Akdeniz University for its partial support.
References
Anonim, 2000. Antalya ili uzun yllk iklim verileri. (Turkish) Antalya Meteoroloji Blge Mdrl,
Antalya.
Arslan H., Gler M., Cemek B., Demir, Y. 2007. Assessment of groundwater quality in Bafra Plain for
irrigation. (Turkish with English Abstract) Journal of Tekirdag Agricultural Faculty, 4(2): 219-226.
El-Ashry M.T. 1991. Policies for Water Resource Management in Semi-Arid Regions. - International
Journal of Water Resources Development 7 (4) 230- 236.
Ghassemi F., Jakeman A.J., Nix H.A. 1995. Salinization of Land and Water Resources Human,
Causes, Extent, Management and Case Studies. Centre for Resource and Environmental
Studies. The Australian National University. Canberra/Australian. 576 s.
IFPRI. 2004. Water and Food to 2025. Policy Responses to Threat of Scarcity.
rtem E., Sar T. 2011. Investigation of Balkesir Plain irrigation. (Turkish with English Abstract) J. Fac.
Eng. Arch. Gazi Univ., 26(2): 461-469.
Kayasseh M., Schenk C. 1989. Reclamation of Saline Soils Using Calcium Sulphate from the Titanium
in Industry-Ambio 18 (2) 124 127.
ktren Asri F., Demirta E.I., Ar N., Arpacolu A.E., zkan C.F. 2010. Determination of irrigation
water qualities in greenhouses of Antalya-Serik region. (Turkish with English Abstract) Akdeniz
niversitesi Ziraat Fakltesi Dergisi, 23(2): 145-150.
ktren Asri F., Demirta E.I., Ar N., zkan C.F. 2013. Determination of irrigation water qualities of
Bilecik-Osmaneli district. (Turkish with English Abstract) Akdeniz niversitesi Ziraat Fakltesi
Dergisi, 26(1): 49-55.
zmen S. 2013. Irrigation water use and water table in Dzce area. (Turkish with English Abstract)
Dzce niversitesi Bilim ve Teknoloji Dergisi, 1: 4757.
Snmez ., Kaplan M. 2004. Determination of salinity of soil and irrigation waters of greenhouses in
Demre Region. (Turkish with English Abstract) Akdeniz niversitesi Ziraat Fakltesi Dergisi,
17(2): 155-160.
Rhoades J.D. 1987. The Problem of Salt in Agriculture. Yearbook of Science and the Future.
Encyclopedia Britannia Chicago.
PAPER:
1a. Assessment and monitoring
VARIABILITY IN GROUNDWATER: NATURAL ATTENUATION RATE VS. SHORT-TERM
FLUCTUATIONS
Petr Kozubek
ENACON s.r.o., Krcska 16, 140 00 Prague, Czech Republic +420 241 404 911,
kozubek@enacon.cz,
Keywords: monitored natural attenuation, attenuation rate, chlorinated hydrocarbons
Introduction
Natural attenuation processes include a variety of physical, chemical, and/or biological processes that
act without human intervention to reduce the mass or concentration of contaminants in soil and ground
water. These in-situ processes include biodegradation, dispersion, dilution, sorption, volatilization;
radioactive decay; and chemical or biological stabilization, transformation, or destruction of
contaminants. Method of monitored natural attenuation (MNA) refers to the reliance on natural
attenuation processes to achieve site-specific remediation objectives within a reasonable time frame.
MNA should not be viewed as a no-action approach but rather as a directed control over the course
of natural attenuation processes to run according to the previous estimates and assumptions and to
go towards the remediation objectives.
The site is located at the edge of agricultural land, meadow and forest on the small hill. From the
geological point of view, the site is situated within the granite massive represented by double mica
medium to fine grained granite. Quaternary deluvial sediments represented by sandy to stony loams
thick from 0.3 to 2 m overlie weathered granite. The thickness of Quaternary sediments increases at
the foot of the hill.
Groundwater aquifer is bound in fractured granite with partly confined groundwater level at about 6 to
7 m below surface level at the top hill area declining up to 15 to 20 m at the slopes of a hill. The
groundwater flow regime is driven by the intensity of the bedrock fracturing, i.e. the level of fractures
opening and character of their filling. Well-logging performed within the site remediation at the top hill
area showed vertical downward groundwater flow to the lower situated fractures and joints.
Considering the site location on the small hill, the groundwater generally flows to two different
directions, to the southwest and to the southeast, respectively.
Public and individual groundwater resources are located approximately 600 m from the site. Public
groundwater resources are formed by eight dug wells supplying the near village of 1,300 inhabitants.
The nearest dug well is situated approximately 500 m southeast from the former waste dump. In
addition, several domestic wells are located in two near settlements located approximately 400 to 600
m from the site.
The six-year project of monitored natural attenuation was initiated in June 2009 and completed in April
2015. The project has been divided into two three-year phases. The initial phase comprised 12
quarterly monitoring rounds of 18 selected monitoring wells representing the wide-ranging site
surroundings (top hill area, slopes of the hill, and areas of public and individual groundwater
resources). The frequency of monitoring was reduced to semiannual for the latter three-year phase.
Main contaminants chlorinated aliphatic hydrocarbons, organochlorine pesticides, chlorides, nitrates,
and nitrites are analyzed in the collected groundwater samples. Natural attenuation parameters
involving ethen as the final product of natural reductive dechlorination of chlorinated ethenes, further
geochemical indicators as dissolved oxygen, sulphates, iron, manganese, total organic carbon), and
microbial abundance and phosphorus as macronutrient are monitored next to the main contaminants.
During the entire project the uniform approach of samples collection is being applied (constant type of
sampling pumps and in-well equipment, constant sample collection depth in the individual wells,
constant duration of wells purging prior their sampling, constant way of samples handling after their
collection etc.) in order to eliminate errors potentially induced by the variations of sample collection
methods.
The low-flow sampling with a constant volume of purging specified for the individual wells was applied
for groundwater samples collection during the project. Gas chromatography/mass spectrometry
according to US EPA 8260 method and gas chromatography according to US EPA 8081 method are
used for laboratory analyses of organic contaminants, chlorinated aliphatic hydrocarbons and
organochlorine pesticides, respectively.
Results
The remedial action in 2008 led to the decrease of the groundwater contamination however its level
remained high in both the source zone and contamination plumes formed along the tectonic fractures
and discontinuities. The source zone contained CHCs up to 28 mg/l, OCP up to 0.7 mg/l, and
inorganic contaminant in range of 1,000 to 2,000 mg/l at the start of the monitored natural attenuation
project. Elevated concentrations of CHCs above 1 mg/l were observed in monitoring wells located to
the distance of 200 m from the source zone.
The effect of the removal of the disposed industrial waste from the quarry became fully evident in the
last two years, i.e. after four years after waste removal and only in the direct area of a former quarry.
During the initial four years after the waste removal, the development of groundwater contamination in
time was influenced by the outstanding fluctuations of concentration of individual subject contaminants
(see graph 1 below). The point attenuation rate for the area of a former quarry was estimated for 0.36
year-1 (corresponding to half-life of approx. 1.92 year).
The decrease of level of contamination has not been observed in the close vicinity of a former quarry
in wells located in radius of 20 m due to persisting inflow of contamination from the bedrock underlying
a former quarry and also due to release of contamination introduced to the quarry vicinity during its
lifetime (desorption of contaminants bound to the filling of the fractures, diffusion from fractures with
limited circulation etc.). Just the last sampling campaigns have indicated the certain improvement
however not statistically evaluable due to fluctuations of concentrations (Mann-Kendall test used for
evaluation of trends in data; e.g. Wiedemeier et al., 1999).
The stabilized development without detectable decrease of concentrations has been observed also in
the area farther from the source zone. The exception is well HV-26 located on the preferential pathway
on the fracture in granite massive with unambiguous decreasing trend of CHCs and other monitored
contaminants.
However, the data plot of development of concentration in distance from the source zone comparing
individual years (annual averages) shows certain improvement even in the area where no trend was
detected due to seasonal fluctuations (see graph 2).
Graph 1: Development of concentration of sum of CHC during the project in the area of a former
quarry
Graph 2: Development of concentration of sum of CHC in distance from the source zone comparing
start of the project (2009 annual average), mid of the project (2012 annual average) and end of the
project in 2014 (annual average)
Conclusion
The project of Monitored Natural Attenuation applied at the site of a former industrial waste dump for
period of six years showed that the attenuation occurs with significantly lower intensity than was
expected at the start of the project. The observed development of the contaminants concentrations is
significantly fluctuating when short-term seasonal effects lead to fluctuations with higher amplitude
than changes caused by the natural attenuation processes.
These aspects is necessary to keep in mind when planning projects of monitored natural attenuation
or projects based on the long-term monitoring in order to eliminate effects of other aspects which
influence the concentrations of contaminants (sampling methods, extreme natural conditions).
Furthermore, the frequency of monitoring should be set to correspond with the observed/expected
attenuation rate.
The performed was granted by Operational Programme Environment and Region Vysocina.
References
U.S. EPA (1999): Use of Monitored Natural Attenuation at Superfund, RCRA Corrective Action, and
Underground Storage Tank Sites, Office of Solid Waste and Emergency Response (OSWER),
Directive 9200.4-17P, Final, Washington, D.C., April 21, 1999, p. 41.T.H. Wiedemeier, H.S. Rifai, C.J.
Newell and J.T. Wilson (1999): Natural Attenuation of Fuels and Chlorinated Solvents in the
Subsurface, John Wiley & Sons, Inc., 1999, p. 617.
T. McHugh, C.J. Newell, D. Adamson, K. Hamel, L. Molofsky and L. Beckley (2013): Improved
Understanding of Sources of Variability in Groundwater Sampling for Long-Term Monitoring Programs,
SERDP Project ER-1705. SERDP, 2013, pp. 1-54.
U.S. EPA (2009): Statistical Analysis of Groundwater Monitoring Data at RCRA Facilities, Unified
Guidance, Office of Resource Conservation and Recovery. EPA 530/R-09-007, March 2009, p. 888.
Theme 1. Dealing with contamination of soil, groundwater and sediment
Abstract. A study was carried out in the vicinity of a Portuguese coal-fired power plant in order to
determine the extent and degree of soils contamination that may be related to the dispersion of the
atmospheric emissions from the this coal-fired power plant. Metals and soluble ions concentrations have
been determined both in-situ and in soil samples collected at sites potentially affected by the stacks
emissions. The analysis of trace elements was performed by means of X-ray fluorescence. The
concentrations of twenty trace elements were determined: As, Ba, Ca, Co, Cu, Fe, Hg, K, Mn, Mo, Ni,
Pb, Rb, S, Sb, Se, Sr, Ti, Zn, Zr, even though very few measurements were obtained for some of these
elements in some specific areas around the power plant. The spatial distribution obtained by kriging in
combination with the analysis of the data by Principal Component Analysis (PCA) showed three
important situations in what concerns to metal concentration in soils and according to soil uses and
geographic localization: the power plant (peak concentrations in the immediate area around the stacks),
in the prevailing wind direction (from N-NW to E-SE and SE-S) and the area comprised between 6 and
20 km from the power plant (mostly agricultural and pasture areas). Considering that soils concentration
for the studied elements may had its origin in the past atmospheric emissions from the coal-fired power
plant and were accumulated in surface soils over the past decades, the height of the stacks and the
dispersion by seasonal winds, could explain the observed enrichment of some elements in the area
between the 6 and 20 km. However, some of the typical heavy metals associated with coal combustion
emissions such as Cd, Cr, Hg and Ni were not significantly registered in the area under study and this
may suggest the contribution from other industrial plants present in the region.
1. Introduction
Many hazardous air pollutants are associated with emissions from coal-fired power plants. Coal is
composed mainly by organic material with inorganic material as minor constituent. Organically coal
consists primarily of carbon, hydrogen and oxygen and with lesser quantities of sulphur and nitrogen.
The inorganic constituents include minerals and trace elements such as As, B, Be, Cd, Cl, Cr, Co, Cu,
F, Hg, Mn, Mo, Ni, Pb, Se, V, and various acid gases.
Following coal combustion, some trace elements become concentrated in ash and may be released
to the environment either through leachate from ash disposal dams or in the gas or particulate phase in
flue gas. Flue-gas emissions from coal burning are characteristically enriched in nanoparticles, toxic and
greenhouse gases, volatile organic compounds and volatile inorganic species, including As, B, Cd, Hg,
Se, organometallic compounds, and other gas components (Hower et al., 2013; Ribeiro et al., 2010).
The solid products of coal burning include fly ash, bottom ash and boiler slag, which may be enriched
in several trace elements. In what concerns to metals and metalloids, atmospheric emissions and
leaching of solid combustion by-products during their disposal or sedimentation, are the main pathways
to soils of the surrounding area.
Soils around a coal power plant are enriched in Ag, Cd, Co, Cr, Cu, Fe, Hg, Ni, Ti and Zn (Klein
and Russell, 1973) and in Mo, As, Mn, Pb, Be, and V (Mandal and Segupta, 2006). Bityukova et al.,
(2000) observed that the levels of As, Cr, Mn and V were more than three times higher and the levels
of Pb and Zn were more than five times higher than the background levels for industrial polluted soils
1
resulting from fossil fuel combustion. Dragovi et al., (2008) observed an enrichment of Cd, Co, Cr, Cu,
Hg, Mn, Ni, Pb and Zn in soils surrounding a coal-fired power plant in Serbia and Keegan et al., (2006)
observed an increase of As in soils around a coal-burning power station in Slovakia. Thus, soil at sites
in the surrounding area of a coal-fired power plants potentially present higher concentrations in various
pollutants depending upon the power plant technology (filtration systems), coal characteristics, and
climatic variations, including humidity and rainfall. These factors change over time mostly due to the
technology development and due to the different types of coal used as fuel.
A study was carried out in the vicinity of a Portuguese coal-fired power plant in order to determine
the extent and degree of soils contamination that may be related to the dispersion of the atmospheric
emissions from the plants stacks. Metals and soluble ions have been determined both in situ and in soil
samples collected at sites potentially affected by these atmospheric emissions. At the same time, dust
collectors were set in these sites to study the flux deposition of the particle matter and establish a relation
between the deposition flux and soil contamination patterns. The variation of the concentration pattern
with the distance to the emitting source was also analyzed as well as the influence of local
meteorological parameters in the concentration pattern. To achieve this, spatial distribution obtained by
kriging in combination with the analysis of the data by Principal Component Analysis (PCA) were used.
The present study was conducted in the vicinity of a coal-fired power plant located in the southwest
coastline of Portugal, 6 km south-east from the town of Sines. This power plant is located in an important
industrial center with others heavy industrial plants, including a petrochemical complex, a refinery and
a resins plant. The power plant has been operational since 1986 with two stacks of 225 m height and it
is fuelled by bituminous coal. It has a coal storage capacity over than 1 million tons in 4 active piles with
150 000 tons each and in one passive pile with 700 000 tons (Dinis et al., 2012; 2013; 2014).
Data on this emission source is reported in the European database European Pollutant Release
and Transfer Register for the period between 2007 and 2012 (Table 1) (E-PRTR, 2015).
Table 1: Pollutant releases to air from Sines coal-fired power plant (extracted from E-PRTR, 2015)
Releases to air 2007 2008 2009 2010 2011 2012
Arsenic and compounds 40.0 kg 31.0 kg 33.0 kg - 27.2 kg 34 kg
Chromium and compounds 127 kg - - - - -
Mercury and compounds 107 kg 209 kg 95.0 kg 54.7 kg 76.6 kg 98.2 kg
Nickel and compounds 126 kg 271 kg 341 kg 254 kg 305 kg 336 kg
Zinc and compounds - 265 kg 349 kg 220 kg 290 kg 354 kg
Particulate matter (PM10) 587 ton 394 ton 130 ton 99.7 ton 286 ton 178 kg
Despite the efficiency of the stacks filtration system has been improved over the years and reduced
significantly the amount of particulate matter released, the atmospheric emissions from this power plant
still present a high fluctuation in what concerns to the amount of trace metals released into the
environment. Therefore, the potential increase in soil metals concentration remains due to both past and
present fugitive emissions.
The historical (10 years) meteorological data referring to the study area indicates that the prevailing
wind direction is from the sector N-NW. The sampling periods were performed during March/April and
September/October 2011 and the prevailing wind direction during these periods were from N-NW and
from N-NE, respectively (Dinis et al., 2012).
The design of the sampling area was based on the influence of the stacks emissions considering
the potential effect on soils contamination up to a distance of 20 km from the coal plant. The sampling
area was defined by two concentric circles with a radius of 6 and 20 km centered in the plant stacks.
The spatial heterogeneity and the deposition flux of dust fallout were assessed in both areas.
A total of 50 samplings sites, represented in Figure 1, were defined in the area under study: 13
sampling sites within the 6 km circle, with urban and sub-urban characteristics, and 37 sampling sites
within the 6-20 km circle, mainly rural. The sampling sites were selected according to the influence from
2
the height of the stacks, the distance to residential areas, the use of soil and accessibility to the sites.
Sampling procedure was carried out in two different periods, representing dry or rainy atmospheric
conditions (Dinis et al., 2013; 2014).
Figure 1: Study area and sampling sites around the coal-fired power plant (source: Dinis et al., 2013, 2014).
The analysis of trace elements concentration in soil was performed by means of X-ray fluorescence
(XRF) with a field portable equipment. Two measurements were done in-situ and repeated in the
laboratory with soil samples collected in the upper 5 cm of soil at each sampling site.
The element concentration detected in surface soils were compared with soil quality guidelines
according to soil use: agricultural, residential, commercial and industrial (Dinis et al., 2012; CCME,
2011).
The elemental analysis of the soil samples collected in the study area measured the concentration
of the following trace elements: As, Ba, Ca, Co, Cu, Fe, Hg, K, Mn, Mo, Ni, Pb, Rb, S, Sb, Se, Sr, Ti,
Zn, Zr, even though very few measurements were obtained for some of these elements in some specific
areas around the power plant. Mercury was identified only in 16 measurements, nickel in 11
measurements, selenium in 8 measurements and antimony in 6 measurements. The results presented
in this study refers to two sampling campaigns, N 2 (SC2) and N 3 (SC3), carried out between
March/April and between September/October 2011. The average values and the statistical description
of the obtained results from both sampling campaigns are presented in Table 2 and Table 3. The number
of cases for each detected element is also presented (N) showing the variation in abundance of several
elements in the study area.
3
Table 2: Statistical summary of trace element concentrations measured in the SC2 (mg/kg)
Element Average Median S.D. Skewness Max. Min. N
As 71.22 70 13.28 0.31 101 52 41
Ba 1679 1679 195.16 - 1817 1541 2
Co 330.75 252 183.52 1.87 603 216 4
Cu 41.33 39 12.42 0.75 61 29 6
Fe 17839.85 10066 17631.61 1.49 74805 829 41
Hg 12 12 - - 12 12 1
Mn 517 280 658.05 2.82 3392 59 36
Mo 11 9 3.46 1.73 15 9 3
Ni - - - - - - -
Pb 23.61 18 25.84 5.08 162 8 33
Rb 55.39 52 31.06 0.69 138 10 41
Sb - - - - - - -
Se - - - - - - -
Sr 71.49 51 78.23 2.68 386 10 41
Ti 3143.06 2738 1728.07 0.25 6468 647 35
Zn 52.51 38 61.34 3.48 351 9 37
Zr 112.1 107 71.13 0.45 259 16 41
Table 3: Statistical summary of trace element concentrations measured in the SC3 (mg/kg)
Element Average Median S.D. Skewness Max. Min. N
As 69.51 63 19.79 0.72 122 35 42
Ba 1001 1001 209.3 - 1149 853 3
Co 320.89 276 116.26 0.68 495 187 5
Cu 72.75 46 74.24 2.57 252 31 5
Fe 17606.27 9440 19296.29 1.53 89231 782 42
Hg 14 - - - - - 1
Mn 558.45 389 585.95 2.09 3091 49 34
Mo 18.42 15 9.26 2.25 51 9 14
Ni - - - - - - -
Pb 29.58 19.5 41.27 5.81 300 10 31
Rb 49.72 42 35.31 0.87 152 3 42
Sb 82 82 - - 82 82 1
Se 4 4 - - 4 4 1
Sr 78.69 48.50 85.01 2.59 506 5 41
Ti 4141.67 3214 4105.6 3.78 27900 625 35
Zn 93 41. 310.21 8.24 2676 10 37
Zr 99.85 75 77.72 1.04 304 10 41
Some of the selected elements were detected in very few sampling points: Ba, Co, Cu, Hg and Mo
in addition to Ni for sampling campaign N 3. Elements content in soils were compared with reference
levels according to soil use: agricultural, residential commercial or industrial. The Canadian Soil Quality
Guidelines for the Protection of Environmental and Human Health were used for comparison with the
selected elements (CCME, 2011).
The concentration values obtained for As and Co are above the respective guideline level for all
types of soil use while Cu concentration is below the guideline level only for commercial and industrial
soil use and Pb and Sn concentrations are below the respective guideline level for all types of soil use.
Barium, mercury and molybdenum concentrations exceed the respective guideline levels for agriculture
and residential soils. Nevertheless, it should be pointed out that some of these elements were detected
in very few sampling sites and are not statiscally representative.
The same variation pattern was observed in the second sampling campaign: Pb, Zn and Cu
concentrations are below the guidelines limits and As and Co concentrations are above the guidelines
levels for all types of soil use. Barium, Hg and Mo concentration values are above the respective
guidelines levels for agricultural and residential soil use.
On average, the concentration values are very similar for both sampling campaigns with the
exceptions of Hg, Sb and Se, which were detected only in one of the two presented sampling campaigns
and of Ba, Cu and Zn, which concentrations present a fluctuation from SC2 to SC3. The concentrations
of the selected elements within the 0-6 km area and within the 6-20 km area are presented in Table 4.
4
Table 4: Average values and statistical description of measurements for selected elements from SC2 and SC3.
For both sampling campaigns, the obtained concentrations are higher in the outer sampling area
(6-20 km) with the exception of Ba, Rb and Sr for SC2 and Co, Sr and Ti for SC3.
Some of the typical heavy metals associated with coal combustion emissions such as Cd, Cr, Hg
and Ni were not detected in the sampling area or at least in a significant number of cases. This may
result from the limitations of the field XRF, in particular high soil moisture, metals interference (Fe tends
to absorb Cu X-rays, whereas Cr levels will be enhanced in the presence of Fe), lack of sensitivity due
to peak overlapping (high detection limits demanding more field measurement time, e.g., Cd and Hg are
not detected easily by XRF because of their detection limit) (Hunter, 2011).
The dispersion of the atmospheric emissions from industrial processes and consequently the
deposition on soils is mainly influenced by wind parameters, rainfall, resuspension of dust from earths
surface, intensified agricultural activities, road traffic and other sources of emissions that may be
present. In cases where none of these factors is present, the intensity of particles deposition and
consequently the concentration on soil should decrease with the distance to the main source of pollution.
The element concentrations were interpolated to unsampled sites by ordinary kriging. Iso-
concentration maps have been made for some elements with the interpolated data from both sampling
campaigns. The distribution profiles are presented in Figure 3 and Figure 4.
5
Ti (mg/kg) Mn (mg/kg)
120000 120000
S-N
S-N
105000 3600 105000 900
3200 800
2800 700
600
2400
100000 100000 500
2000
400
1600
300
1200 200
95000 800 95000 100
400 0
90000 90000
Zr (mg/kg) Pb (mg/kg)
120000
120000
260 115000
250 36
115000 240
230 34
220 32
210
200 110000 30
190
110000 180 28
170
160 26
150
140 24
S-N
105000
S-N
105000 130 22
120
110 20
100
90 18
80 100000
100000 70 16
60
50 14
40 12
30
20 95000 10
95000 10
0 8
90000 90000
Figure 3: Iso-concentrations maps for Ti, Mn, Zr and Pb (data from SC2).
As (mg/kg) Fe (mg/kg)
120000 120000
120 46000
44000
115000 115 115000 42000
110 40000
38000
105 36000
100 34000
110000 110000 32000
95
30000
90 28000
85 26000
S-N
S-N
24000
105000 80 105000 22000
75 20000
18000
70
16000
65 14000
100000 100000 12000
60
10000
55 8000
50 6000
4000
95000 45 95000 2000
40 0
90000 90000
Zn (mg/kg) Rb (mg/kg)
120000 120000
230 150
220 140
115000 115000
210
200 130
190 120
180
170 110
110000
110000 160 100
150
140 90
130 80
S-N
S-N
120 105000
105000 110 70
100 60
90
50
80
70 100000 40
100000 60
30
50
40 20
30
95000 10
20
95000 10 0
90000
90000
140000 145000 150000 155000 140000 145000 150000 155000
W-E W-E
Figure 4: Iso-concentrations maps for As, Fe, Zn and Rb (data from SC3).
6
The iso-concentration maps for the selected trace elements showed a pattern concentration
described by either marked element enrichment for distances between the 6 and 20 km in the NE-E and
E-SE sectors and peak concentrations in the immediate area around the stacks. The stacks are locate
at approximately 141000 m (x) and 107600 (y). The distribution of the selected elements as well as
other elements in surface soil seems to follow a similar pattern observable in both sampling campaigns.
Nevertheless, no regular trends could be observed for elemental concentration and distance from the
power plant.
During both sampling campaigns the prevailing wind directions were from N-NW and N-NE and this
trend was confirmed with data from a 10 years period of meteorological observations. This could explain
at some extent the metals concentration pattern in soils. However, peak concentrations in other wind
directions rather than the prevalent one, suggest the contribution from others sources. An important
feature in atmospheric emissions is the influence of stack height in the dispersion and deposition
patterns: about 25% of emitted particles from a 40 m stack fell within 5 km and 34% within 10 km
(Keegan et al., 2006). It should be pointed out the presence of the coal storage in the immediate area
around the stacks and the peak concentrations in the immediate vicinity of the power plant is likely to
be predominately from local fugitive sources in the form of windblown ash and dust.
The results from the application of the principal component analysis (PCA) showed that there four
significant principal components: 1) As, Ti, Zr, Rb and Mn; 2) Pb; 3) As and 4) Pb and Zn. In general,
for these sets of significant association, the highest concentration values were observed near the power
plant (S8, S9, S12, S14), in the SE direction (S16, S17, S23, S24) and in the NE direction (S7, S32,
S33).
The spatial distribution obtained by kriging in combination with the analysis of the data by principal
component analysis showed three important situations in the study area according to soil uses, wind
direction and geographic localization: the power plant (peak concentrations in the immediate area
around the stacks), in the prevailing wind direction (from N-NW to E-SE and SE-S) and the area
comprised between 6 and 20 km from the power plant (mostly agricultural and pasture areas). These
findings overlap with the pattern of the particulate matter flux deposition registered in the study area.
Also, soils concentration pattern was observed for most of the detected elements in some repetitive
areas in all sampling campaigns, which explains the correlation found between some elements. Soil
concentration and deposition flux correlation made also possible the identification of common sources
of contamination to the area under study (Dinis et al., 2011; Dinis et al., 2012).
4. Conclusions
A study was carried out in the vicinity of a Portuguese coal-fired power plant in order to determine
the extent and degree of soils contamination related to the atmospheric emissions from this industrial
facility. The concentration of trace elements have been determined both in-situ and in top soil samples,
collected at sites potentially affected by power plant operations. The obtained results were above of the
international soil quality guidelines for some of the identified elements. On average, the concentrations
of several elements measured in soil, such as As, Ba, Co, Hg, Mo, Ni and Se, are above the soil quality
guidelines for agricultural and residential soils. Nevertheless, with the exception of As and Mo very few
measurements were obtained for these elements.
The concentrations were used to evaluate the spatial distribution obtained by kriging in combination
with the analysis of the data by principal component analysis (PCA). A similar spatial pattern was
observed for both sampling campaigns: on average the concentration is higher within the outer area (6
- 20 km). High concentrations were measured in the NE, E and SE directions, influenced by the wind
parameters and the height of the stacks, which are expected to have transported the atmospheric
emissions for relatively long distances from the source. The peak concentrations measured in soils
around the stacks could not have fallen from the plume vertically and may be directly related with the
coal storage near the stacks, in addition to fugitive emissions from ground-level areas enriched in these
elements by other means.
Considering that soils concentration for the measured trace elements had its origin in the past
emissions from the power plant s stacks and were accumulated in surface soils over the past decades,
the height of the stacks and the dispersion by seasonal winds, could explain the observed enrichment
of some elements within the 6 - 20 km area. However, some of the typical heavy metals associated with
coal combustion emissions such as Cd, Cr, Hg and Ni were not significantly registered in the area under
study and this may suggest the contribution from other sources related to coal combustion or to the
power plant, in addition to others industrial plants present in the region (Dinis et al., 2011).
7
ACKNOWLEDGMENTS
This work was supported by FEDER grants through the COMPETE (Programa Operacional Factores
de Competitividade) and by national grants through the FCT (Fundao para a Cincia e Tecnologia)
namely via the project FCOMP01-0124-FEDER-009745 (FCT reference: PTDC/ECM/100735/2008).
Strategic project FCT- UID/ECI/04028/2013.
REFERENCES
Baba, A., 2000. Assessment of radioactive contaminants in by-products from Yatagan (Mugla, Turkey)
coal-fired power plant. Environmental Geology Journal, 41: 916-921.
Bitykova, L., Shogenova, A., Brike, M., 2000. Urban Geochemistry: a study of element distributions in
the soils of Tallinn (Estonia). Environmental Geochemistry Health, 22: 172-193.
CCME 2011. Canadian Environmental Quality Guidelines, http://ceqg-rcqe.ccme.ca/.
Dinis, M.L., Fiza, A., Gis, J., Carvalho, J.S., Castro, A.C.M., 2011. Assessment of direct radiological
risk and indirect associated toxic risks originated by Coal-Fired Power Plants. Radioprotection
Journal, 46, n 6: S137-S143.
Dinis, M.L., Fiza, A., Gis, J., Carvalho, J.S., Castro, A.C.M., 2012. Assessment of heavy metals in
soils in the surroundings of a coal-fired power plant, Proceedings of the International Conference
on Environmental Science and Technology, June 25-29, 2012 in Houston, Texas, USA.
Dinis, M.L., Fiza, A., Gis, J., Carvalho, J.S., Castro, A.C.M., 2013. Radiological Impact Associated to
Technologically Enhance Naturally Occurring Radioactive Materials (TENORM) from Coal-Fired
Power Plants Emissions, Proceedings of the Symposia on Waste Management, Phoenix, Arizona,
February 2428.
Dinis, M.L., Fiza, A., Gis, J., Castro, A.C.M., Carvalho, J.S., 2014. Radiation Dose rates and Exposure
Associated to Technologically Enhanced Naturally Occurring Radioactive Materials (TENORM) at
Locations of a Portuguese Coal-fired, Proceedings of the Symposia on Waste Management,
Phoenix, Arizona, March 0206.
Dragovic, S., Mihailovic, N., Gaji, B., 2008. Heavy metals in soils: distribution, relationship with soil
characteristics and radionuclides and multivariate assessment of contamination sources.
Chemosphere, 72(3), 491495. doi:10.1016/j. chemosphere.2008.02.063.
E-PRTR, 2015. European Pollutant Release and Transfer Register database,
http://prtr.ec.europa.eu/Home.aspx.
Hower, J.C., O'Keefe, J.M.K., Henke, K.R., Wagner, N.J., Copley, G., Blake, D.R., Garrison, T.M.,
Oliveira, M.L.S., Kautzmann, R.M., Silva, L.F.O., 2013. Gaseous emissions from the Truman
Shepherd coal fire, Floyd County, Kentucky: a re-investigation following attempted mitigation of the
fire. Int. J. Coal Geol. 116117, 6374.
Hunter, D., 2011. Field X-Ray Fluorescence Measurement. USEPA, SESDPROC-107-R2.
Keegan, T.J., Farago, M.E., Thornton, I., Hong, B., Colvile, R.N., Pesch, B., Jakubis, P.,
Nieuwenhuijsen, M.J., 2006. Dispersion of As and selected heavy metals around a coal-burning
power station in central Slovakia. Sci Total Environ. 358(1-3):61-71.
Klein, D., Russell, P., 1973. Heavy Metals: Fallout around a Power Plant. Env Sci and Tech, 7, n 4:
357-358.
Mangal, A., Sengupta, D., 2006. An assessment of soil contamination due to heavy metals around a
coal-fired thermal power plant in India. Environmental Geology Journal, 51: 409-42
Ribeiro, J., Flores, D., Ward, C., Silva, L.F.O., 2010. Identification of nanominerals and nanoparticles in
burning coal waste piles from Portugal. Sci. Total Environ. 408, 60326041.
8
Topic 1a Assessment and monitoring
POSTER - 4976
c
angela.volpe@ba.irsa.cnr.it; dmichele.pagano@ba.irsa.cnr.it; egrenni@irsa.cnr.it; fbarracaracciolo@irsa.cnr.it;
g
giuseppe.mascolo@ba.irsa.cnr.it; hrossetti@irsa.cnr.it
Abstract
The biodegradation potential of marine sediments towards two selected UV-filters was tested through
a laboratory investigation. Sediments collected from two locations in Italy were spiked with 3-(4-
methylbenzylidene)camphor (4-MBC) and 2-ethylhexyl 4-(dimethylamino)benzoate (EH-DPAB), and
used to set up eight microcosms that were operated under either aerobic or anaerobic conditions. UV-
filters concentration in the sediment was monitored over 360 days. Results showed that native
microbial communities of both sediments were able to perform the complete removal of EH-DPAB,
whereas 4-MBC was scarcely affected by biodegradation. Degradation kinetics of EH-DPAB was first-
order, with kinetic rate constants around 0.02 d-1 for both sediments under both aerobic and anaerobic
conditions. Microbiological characterisation of microcosms showed that the composition and structure
of the sediment microbial communities changed after the different treatments with respect to the native
microbial population.
1. Introduction
Organic UV-filters are emerging environmental contaminants whose persistence has recently
raised concern, owing to their potential endocrine disrupting effects. The spreading of these
compounds into the aquatic environment occurs as a consequence of their wide use in the formulation
of personal care products, where they are needed to protect human skin from solar UV-radiation.
Moreover, several categories of goods and materials (e.g., polymers, varnishes, adhesives, etc.) also
contain UV-filters to reduce photodegradation caused by the exposure to sunlight (Gago-Ferrero et al.
2012; Giokas et al. 2007). The release of UV-filters into the environment occurs mainly through direct
input, due to recreational activities, as well as through the discharge of effluents from wastewater
treatment plants. In fact, UV-filters removal may not be effective during conventional wastewater
treatments, and this issue is currently a matter of study (Li et al. 2007; Tsui et al. 2014; Weiss et al.
2006).
Most UV-filters are hydrophobic compounds, characterised by low water solubilities and high
octanol-water partition coefficients (Diaz-Cruz et al. 2008). Therefore, in aquatic environments they
accumulate in sediments and in biota, causing adverse effects to aquatic organisms (Coronado et al.
2008; Kaiser et al. 2012; Kunz and Fent 2006). The consequent biomagnification through the trophic
chain may eventually pose potential risks to human health.
Up to date, there are very scarce data regarding the fate of UV-filters in the environment.
Biological degradation has been investigated with reference to wastewater treatment either by
activated sludge (Kupper et al. 2006; Liu et al. 2012) or through bioslurry reactors inoculated with
selected fungal strains (Rodriguez-Rodriguez et al. 2012). Information regarding the potential
biodegradation of UV-filters in sediments is even more exiguous (Liu et al. 2013). No data are
currently available with regard to biodegradation of UV-filters in marine sediments.
Our study was focused on two chemicals belonging to two different classes of widespread UV-
filters, i.e., 3-(4-methylbenzylidene) camphor (4-MBC) and 2-ethylhexyl 4-(dimethylamino) benzoate
(EH-DPAB). They are characterised by very low water solubilities, and, therefore, may be frequently
detected in sediments at high concentrations. Two sets of laboratory-scale batch reactors were set up
using two sediments collected, respectively, in the area of La Spezia harbour (Northern Italy) and at
the estuary of the Sarno river (Southern Italy). The sediments were spiked with the target compounds,
and reactors were operated under different conditions (i.e., either aerobic or anaerobic; either
containing living microorganisms or sterile). The UV-filters removal was monitored over time,
determining the kinetics of biodegradation. The development of microbial consortia into microcosms
was simultaneously investigated, determining the change of composition and structure of microbial
population associated to contaminants degradation.
2.1. Sampling and characterisation of the sediments. The sediments collected at the two
sampling sites will be referred to in the following as S1 (La Spezia harbour) and S2 (Sarno river). After
sampling, they were characterised by determining: a) water content; b) pH; c) concentration of UV-
filters; d) total microbial abundance; and c) phylogenetic composition of the native microbial
population. The samples were stored at 4C and were then used as received.
2.2. Preparation of batch microcosms. Aerobic microcosms were prepared in Pyrex bottles,
which were closed by teflon-faced butyl rubber stoppers, and were periodically opened for aeration.
Anaerobic microcosms were prepared in cylindric glass vessels with flat bottom and glass heads,
equipped with two screw-cap necks and a glass stopcock, to assure a continuous nitrogen flow during
preparation and every time that opening was required for sampling. A magnetic stirrer bar was
introduced into each reactor.
4-MBC and EH-DPAB were not detected in the sampled sediments. Therefore, during the setup of
each microcosm, a weighed amount of wet sediment was suspended into synthetic salt water and
-1
spiked with a methanol solution containing the two UV-filters (1500 mg L each). The final
concentration obtained for each compound was 50 mg per kg of dry sediment (mg/kgdry) in
microcosms prepared with S1, and 25 mg/kgdry in those prepared with S2. Sodium sulphide (2 mM)
was added to anaerobic microcosms as a reducing agent. Resazurin (2 mg L-1) was also added as a
redox indicator, and anaerobic reactors were fluxed with N2 until the disappearance of the purple
colour from the liquid phase.
Four active reactors (i.e., containing living microorganisms) were obtained using the unaltered
sediments. A series of four sterile microcosms were prepared, as above described, using previously
autoclaved sediments (two consecutive cycles at 120 C for 20 min), and adding 5 g L-1 of sodium
azide to maintain sterility. Experimental conditions for all microcosms are summarised in Table 1.
During reactors operation, each time the complete decay of EH-DPAB was detected, the sediment
was spiked again with the UV-filter by adding a methanol solution of EH-DPAB (1500 mg L-1) to the
suspension. All reactors were kept in the dark at 22 C, in a refrigerated thermostat incubator, and
were periodically shaken by hand. Samples of 10 mL of sediment suspension were periodically
withdrawn for chemical analysis. Further samples were withdrawn for microbiological analysis at times
corresponding to the complete decay of EH-DPAB.
2.3. Analytical methods. Suspended sediment samples were centrifuged, and the collected solid
was freeze-dried. Each sample was split into three subsamples, which were extracted by pressurized
liquid extraction with in-cell purification, using a Dionex ASE 300 system. Cells with a volume of 66
mL, equipped with a cellulose filter on the bottom, were filled by introducing 3 g of silica gel (activated
by heating at 105C for 24 h), then 0.5 g of sediment mixed with 3 g of silica gel and 3 g of Cu powder,
and diatomaceous earth to complete the filling. Extraction was performed with a mixture of ethyl
acetate and hexane (80:20 v:v), by running three static cycles of 5 min each at 160C. Extracts were
evaporated first by a rotary evaporator, then concentrated to the volume of 0.2 mL by a nitrogen
stream, and finally reconstituted in 1 mL of acetonitrile for HPLC analysis. Blanks were obtained by
applying the above procedure to sediments not previously spiked with the UV-filters.
UV-filters in the sediment extracts were analysed by the HPLC-UV technique. A Dionex Ultimate
3000 System, equipped with a Kinetex C18 column (2.6 , 100A, 150 x 2.1 mm) and a C18
SecurityGuard cartridge (Phenomenex), was used. The detection wavelength was 300 nm. The eluant
was a mixture of water (A) and acetonitrile (B), fed in gradient mode for 10 min (from 80/20 A/B to
10/90), then in isocratic mode for 5 min, and return to initial conditions in 1 min. External standard
calibration was carried out in the range 0.5 - 30 mg L-1 for each compound.
2.4. CARD-FISH analysis. Slurry samples were immediately fixed in formaldehyde (2% v/v final
concentration) for 3 h at 4C, and further processed to extract cells from the sediment particles as
described previously (Barra Caracciolo et al., 2005). Biomass extracted from the slurry was harvested
on polycarbonate membrane filters (pore size 0.2 mm, diameter 47 mm, Millipore) by gentle vacuum
(<0.2 bar), and filters were stored at -20C until performing Catalysed Reporter Deposition FISH
(CARD-FISH) according to previously published procedures (Matturro et al., 2013). Oligonucleotide
probes labelled with horseradish peroxidase (HRP) (Biomers, Germany) were used to identify and
quantify members of Bacteria (Eub338-I, -II, -III probes) and Archaea (Arch917 probe). Total microbial
abundance (N. cells/g dry sediment) was determined by direct count, using 4-6-diamidino-2-
phenylindole (DAPI) as a DNA fluorescence agent (Barra Caracciolo et al., 2005). Hybridized cells
were analysed and quantified by epifluorescence microscopy (Olympus, BX51) by counting
fluorescent cells (at least 100 cells per grid) on random grids on filters sections.
3.1. Biodegradation of UV-filters. During the monitoring period of the microcosms, the
concentration of both UV-filters did not vary in all the sterile control reactors. This result confirmed that
no abiotic degradation processes occurred, and any change of the UV-filters concentration in the
active microcosms was ascribable to biologically mediated reactions.
The concentration of 4-MBC decreased only by 30% after 360 d in the anaerobic reactor containing
S1, whereas it did not appreciably vary in S2 after 300 d (data not shown). These results suggested
that biodegradation did not affect the compound 4-MBC. Conversely, for both sediments S1 and S2,
the concentration of adsorbed EH-DPAB sharply decreased since the beginning of the microcosms
operation. The decay of EH-DPAB concentration occurred under both aerobic and anaerobic
conditions. After the decrease of the UV-filter concentration below the detection limit, the second spike
of the sediment was performed, and the decay was again observed. This result confirmed the
biological nature of the EH-DPAB loss The curves reported in Figure 1 illustrate the concentration
trends of EH-DPAB in the four active reactors after the second spike of the contaminant (i.e., in the
time range 180-360 d from the beginning of the experiment). Correlation of concentration data led to
assess that degradation kinetics was first-order. The calculated kinetic rate constants k (Table 2) were
comparable for the two sediments under aerobic and anaerobic conditions. These findings suggested
60
50
10
0
0 30 60 90 120 150 180
60
50
EH-DPAB (mg Kg-1)
b
40
S1-Anaerobic
S1-Anaerobic Control
30 S2-Anaerobic
S2-Anaerobic Control
20
10
0
0 30 60 90 120 150 180
Time (d)
k (d-1)
a
Microcosm t1/2
S1-Aerobic 0.020 35
S2-Aerobic 0.019 36
S1-Anaerobic 0.019 36
S2-Anaerobic 0.016 43
a
calculated as ln 2/k
that, in all the active microcosms investigated, mixed aerobic and anaerobic microbial consortia might
actually be present, i.e., local micro-environments might exist in the complex structure of the
sediments, where microbial species able to degrade EH-DPAB may grow under different reaction
conditions. Further investigations are needed to assess the mechanism of biodegradation. In fact, the
input of methanol might have induced potential co-metabolic pathways associated to a source of
readily available organic carbon in the investigated system.
2.0E+06
1.5E+06
dry
N. cells/g
1.0E+06
5.0E+05
0.0E+00
S1-Aerobic S2-Aerobic S1-Anaerobic S2-Anaerobic
Figure 2. Abundance of bacteria, archaea and total cells in the different treatments
estimated at the end of the microcosms operation.
4. Conclusions
The results of this study suggested that native microorganisms in sediments may have developed
the ability to degrade a few UV-filters, although the massive spreading of these compounds in the
environment has occurred only in recent times. Nevertheless, the potential risks associated to the
accumulation of metabolites deserve further study.
Acknowledgements
The activities described in this publication were funded by the Project Bandiera RITMARE - La
Ricerca Italiana per il Mare coordinated by the National Research Council and funded by the Ministry
for Education, University and Research within the National Research Program 2011-2013.
References
Barra Caracciolo A, Grenni P, Cupo C, Rossetti S (2005). In situ analysis of native microbial
communities in complex samples with high particulate loads, FEMS Microbiol Lett 253: 55-58.
Coronado M, De Haro H, Deng X, Rempel MA, Lavado R, Schlenk D (2008). Estrogenic activity and
reproductive effects of the UV-filter oxybenzone (2-hydroxy-4-methoxyphenyl-methanone) in fish,
Aquat Toxicol 90:182-187. doi:10.1016/j.aquatox.2008.08.018
Diaz-Cruz MS, Llorca M, Barcelo D (2008). Organic UV filters and their photodegradates, metabolites
and disinfection by-products in the aquatic environment, TRAC-Trend Anal Chem 27:873-887.
doi:10.1016/j.trac.2008.08.012
Gago-Ferrero P, Silvia Diaz-Cruz M, Barcelo D (2012). An overview of UV-absorbing compounds
(organic UV filters) in aquatic biota, Anal Bioanal Chem 404:2597-2610. doi:10.1007/s00216-012-
6067-7
Giokas DL, Salvador A, Chisvert A (2007). UV filters: From sunscreens to human body and the
environment, TRAC-Trend Anal Chem 26:360-374. doi:http://dx.doi.org/10.1016/j.trac.2007.02.012
Kaiser D, Sieratowicz A, Zielke H, Oetken M, Hollert H, Oehlmann J (2012). Ecotoxicological effect
characterisation of widely used organic UV filters. Environ Pollut 163:84-90.
doi:10.1016/j.envpol.2011.12.014
Kunz PY, Fent K (2006). Estrogenic activity of UV filter mixtures, Toxicol Appl Pharm 217:86-99.
doi:http://dx.doi.org/10.1016/j.taap.2006.07.014
Kupper T, Plagellat C, Braendli RC, de Alencastro LF, Grandjean D, Tarradellas J (2006). Fate and
removal of polycyclic musks, UV filters and biocides during wastewater treatment, Water Res
40:2603-2612. doi:10.1016/j.watres.2006.04.012
Li W, Ma Y, Guo C, Hu W, Liu K, Wang Y, Zhu T (2007). Occurrence and behavior of four of the most
used sunscreen UV filters in a wastewater reclamation plant, Water Res 41:3506-3512.
doi:10.1016/j.watres.2007.05.039
Liu YS, Ying GG, Shareef A, Kookana RS (2012). Biodegradation of the ultraviolet filter
benzophenone-3 under different redox conditions, Environ Toxicol Chem 31:289-295.
doi:10.1002/etc.749
Liu YS, Ying GG, Shareef A, Kookana RS (2013). Degradation of six selected ultraviolet filters in
aquifer materials under various redox conditions, Ground Water Monit R 33:79-88.
doi:10.1111/gwmr.12027
Matturro B., Heavner G.L., Richardson R.E., Rossetti S., 2013. Quantitative estimation of
Dehalococcoides mccartyi at laboratory and field scale: comparative study between CARD-FISH
and Real Time PCR. J Microbiol Meth 93, 127-133.
Rodriguez-Rodriguez CE, Baron E, Gago-Ferrero P, Jelic A, Llorca M, Farre M, Diaz-Cruz MS, Eljarrat
E, Petrovic M, Caminal G, Barcelo D, Vicent T (2012). Removal of pharmaceuticals,
polybrominated flame retardants and UV-filters from sludge by the fungus Trametes versicolor in
bioslurry reactor, J Hazard Mater 233:235-243. doi:10.1016/j.jhazmat.2012.07.024
Tsui MMP, Leung HW, Lam PKS, Murphy MB (2014). Seasonal occurrence, removal efficiencies and
preliminary risk assessment of multiple classes of organic UV filters in wastewater treatment
plants, Water Res 53:58-67. doi:http://dx.doi.org/10.1016/j.watres.2014.01.014
Weiss S, Jakobs J, Reemtsma T (2006). Discharge of three benzotriazole corrosion inhibitors with
municipal wastewater and improvements by membrane bioreactor treatment and ozonation,
Environ Sci Technol 40:7193-7199. doi:10.1021/es061434i
ENVIRONMENTAL COMPLIANCE MONITORING AT FINNISH OUTDOOR SHOOTING RANGES
-BEST AVAILABLE TECHNIQUES AND PRACTICES FOR SURFACE WATER
In Finland, there are almost one thousand active outdoor shooting ranges, which vary in size and
environmental conditions. The environmental impacts of the ranges depend mostly on their location.
The ranges located on a ridge/esker near to the water extraction well have very different
environmental and health risks to those located on peatland or clay soil. The EU directives and Finnish
Environmental Protection Act require operators, including active shooting ranges, to use best available
techniques (BAT) and best environmental practices (BEP) to prevent or reduce environmental risks
and impacts. The most recent national BREF-document is Environmental Impact Management at
Shooting Ranges Best Available Techniques (BAT), which was released September 2014. In this
document the compliance monitoring for surface water discharge is described. Some examples of this
new perspective approach are covered in this presentation.
The surface water discharge criteria is based on the annual average concentration of the constituents
of concern in receiving water body. To determine the annual average concentration of the discharge
the national legislation requires monthly sampling for selected monitoring points. In the BREF
document an alternative to monthly sampling is the use of passive sampling methods. Also if the
operator can reliably determine the range discharge profile i.e. the relationship of metal concentration
vs. flow rate, the sampling frequency can be reduced. Typically the highest discharge of metals from a
range occurs in the spring time during the snow melt down. In these cases, shooting range metals are
usually bound to colloids or humic particles. In such case it is possible to carry out the monitoring only
in the springtime. The monitoring results in this case give the worst case discharge and hence if the
worst case is below the allowable discharge then it can be deduced that the annual discharge is at
allowable levels also. However for a more reliable result the use of a passive sampler is advisable.
There are passive samplers that measure the flow weighted average concentration. This type of
sampling device is optimal when determining the annual metal load from a range.
An example of the concentration fluctuation due to flow rate and monitoring result from passive
sampler is presented in figure 1.
Figure 1. Flow rate vs. Pb concentration in range discharge stream.
The acceptable discharge from the range depends on the distance from the range to the receiving
water system and the amount of water flowing to the stream from and outside of the range and
therefore mixing the range water to the external water. The acceptable discharge can be determined
by the water balance of the catchment areas. The compliance point for catchment area is usually the
end point of a stream that enters a larger surface water body. This point can receive water from other
sites and operators, so it is impossible to specify what site or operator is responsible for the discharge.
Therefore sites and operation located on the same catchment area need to have their own monitoring
points.
The range specific target concentration for surface water discharge from the range is based on the
size of the range and surrounding catchment area. The operators can determine the catchment area
from topography maps that are freely available at the national land surveys web site or from other
maps, making the method easily available to all range operators. Dilution factor is the range
catchment area divided by the total catchment area of the compliance monitoring point.
For example if the catchment area of a stream that flows from a range to the water system (e.g. sea,
lake, river) is 10 ha (100 000 m2) and the catchment area of the shooting range is 20 000 m 2, the
dilution factor of range water to the external water is 100 000 m 2 / 20 000 m2 = 0,2. The environmental
quality standard for lead in surface water is 7,2 g/l, so the acceptable concentration for lead in the
water from the range to the ditch is 7,2 g/l / 0,2 = 36 g/l.
Figure 2. The range monitoring point at a catchment area.
If the annual average concentration of the water from the range monitoring point is below this
acceptable concentration, the concentration of the water system will not exceed the acceptable level
(environmental quality standard). If the concentration exceeds the criteria, actions to reduce the metal
load to the stream are required.
This method does not take into account sedimentation due to chemical changes in the discharge nor
filtration. Therefore the method gives a conservative estimate of the allowable discharge.
Theme 1. Dealing with contamination of soil, groundwater and sediment
Developments in technologies, policies, concepts, regulation, management
1a. Assessment and monitoring
Nino Zakariadze1,2, Shota Adamia1, Giga Kvartskhava1,2, Nino Gagelidze1,2, Shorena Jadugishvili1
1.GEO-ECO TAO - 27,Griboedov St., 0108 Tbilisi. Georgia; Tel: + (995 5) 99 26 20 06
2.Georgian Technical University, Agricultural Science and Biosystem Engineering Department -
77 KostavaSt, 0175 Tbilisi, Georgia; Tel: + (995 5) 99 22 67 33
nzakariadze@psi.ge; nzakariadze@geoecotao.ge; sh_adamia@hotmail.com;
giga.kvartskhava@gmail.com; n.gagelidze@agruni.edu.ge; Shorena.gia@caucasus.net
Initiative
GEO-ECO TAO group takes the initiative to make a first step for establishing Georgian Medical
Geology Association in 2015 at Georgian Technical University.
The background of GEO-ECO TAO - multi approach studies on Assessment of Natural and
Anthropogenic Impacts in parallel with related Human Health Problems carried on since 2001 enables
the team members to attract high professionals and start with creation of valid data base on Natural
and Man Made Threats in Georgia and Health and Social State of the Communities exposed to those
threats.
The territory of Georgia is divided into 9 regions and 2 autonomous republics - Abkhazia (occupied)
and Adzharia.
Despite the small area the country has varied topography with various natural resources (water,
mineral, flora and fauna) and at the same time is characterized by various natural hazards - high
seismic activity and earthquake related phenomena - floods, destruction of relief stability, avalanches,
landslips (creeps), mudflows etc. The severe exhaustion of natural resources, in result of
anthropogenic impacts - waste accumulation and heavy pollution of air, water and soil in result of
anthropogenic activities must be mentioned as well.
The Man Made and Natural impacts in most cases are interdependent and are reflected in two
interdependent domains - Health and Social state of the population, the damage of which is often
grave, continuous and even irreversible.
Two cases are presented for the poster on Environmental Health policy analyses and discussion:
I. Hazardous Industrial Discharges to Surface Water Bodies and Morbidity in Kvemo Kartli
Region
The contaminated water is permanently used by local population for irrigation of arables, vegetable
gardens, vineyards and orchards.
At the same time the fruits and crops are often washed in the loaded water of Mashavera River before
being brought for sale.
The alluvial plains of both sides of the Mashavera River are natural grassland used for grazing of
sheep and cattle, which several times per year are flooded after heavy rainfall.
Among risk groups (employees and inhabitants consuming local products) the incidence of leukemia
and anemia is significantly increased. The limited screening programs indicate the rising
procancerogenisc trends.
Though, the latency of industrial disease and diversity of clinical signs hinder health state assessment
of vulnerable communities.
Very often the alarming news on increased morbidity is conveyed not from health care, social or
environmental organizations, but private insurers, incurring looses with the threat to go under.
The single health programs are carried separately and do not coincide with studies on water and soil
chemical contamination.
Asymmetry in Morbidity Evaluation Strategies
But between these organizations no agreement is reached upon standardization and unification of the
assessment methodologies (selection of assessment parameters and calibration of evaluation scale)
for disease and disabilities prevalence in separate regions:
Territorial
1. Regional 1 - - -
2. Separate 2 - 2 2
Classification of Disease
1. Contagious 1 1
2. Noncontagious - 2 2
3. Oncological - 3
4. Accidental (disaster/labor related) - 4
Hazardous Factors
1. Probable 1 - 1 1
2. Verified 2
Study of Cofactors +
Inadequate studies of risk factors and disharmonious use of various parameters lead to information
asymmetry and become the source for uncertainties in health statistics, especially in non-contagious
morbidity.
II. Illegal Tree Felling, Landslides and Ecomigration from Adchara Region
The quality of education was strongly and irreversibly affected by brain drain;
The variety of evaluating strategies and methodologies looks rather more problematic in organizations
focused on environmental impact and risk assessment. The big players - Georgian Ministry of
Environment and Natural Resources and international foundations working in this field are using their
own reporting systems, mostly based on requirements of donor organizations.
Overwhelming majority of works falls to the share of hundreds of NGOs and consulting companies
(local and international).
NGO Environmental domain is saturated with huge number of sporadic, single value studies, based on
inappropriate methodologies, invalid data and false models elaborated on the basis of wrong
approaches and ill posed tasks. Research Task Design always remains as the weakest point of every
study.
However, very useful research is often buried in this grey literature and loses significance with the laps
of time.
NGOcracy appeared to be the reason for duplication of financing and money laundering. Sometimes,
one single issue under study requires unlimited number of Projects and Workshops, but is it is
impossible to find among them even two, similar Risk Assessment Matrices.
Continuous and useless atomization of the main tasks of the issues under study and misuse of
targeted resources prepare the ground for imperfect decision making what hinders problem resolution
(if any defined) and finally leads to severe damages of environment, irreversible looses including
grave impacts on human health.
MG GIS Necessity
The decision making in the fields of Health Care and Environment will be improved significantly when
Unified Methodologies are elaborated and Harmonized Research is conducted for assessment of
environmental hazards in parallel with health risks.
Goal
Medical Geology Atlas is one of the very important tools of Decision Support Systems (DSS) in the
process of fundamental investigation and monitoring of Environmental Health problems.
Preparation of MG Atlas considers collecting evidence based, statistically valid data on prevalent
disease and disabilities in separate regions of Georgia and linking these data to the digitalizing
information on existing natural and man-made hazardous factors of each, corresponding region.
Data Sources in Georgia: National Centre for Disease Control; Ministry of Labor, Health and Social
Affairs; Departments of Health of the Local Governments; Ministry of Environmental Protection and
Natural Resources; Academia Institutions; Tenures and Individual Experts; Insurance Companies and
Health Care Service Providers; Local Industries; Local Communities; Case Reporting Teams (CRP).
Standardized and user friendly MG data base will be available to all interested parties:
At State level: Health and Environmental Policy Makers; Scientific and Education Institutions;
Legislative bodies.
At Regional level: Health Care Service Providers; Insurers; Industrial Companies; Communities.
For present the main task for the Project Team is to become familiar with the Best Practices of Data
Collection, Data Entry and Data Storage for Medical Geology Atlas:
International Partnership
We seek for advices from international partners and are interested in hands-on experience sharing: to
start with this complicated task, develop the strategy for Georgian Medical Geology Atlas and upgrade
permanently the quality of results.
ThS 1A.1 Passive sampling
Keywords: passive sampling, groundwater discharge, contaminant mass discharge, groundwater risk
assessment, chlorinated solvents.
ABSTRACT
Groundwater contaminant mass flux (discharge) receives increased focus as one of the key
parameters in relation to risk assessments at sites with contaminated groundwater. The mass flux is
relevant for e.g., the ranking and prioritizing of sites for remediation purposes, evaluating ongoing
remediation, and monitoring or natural attenuation. Traditionally, flux concentrations are derived from
pumping methods or indirectly calculated from solute concentrations, hydraulic gradients and hydraulic
conductivities. This contribution presents a novel passive sampling approach for direct assessments of
the Darcy flux and solute fluxes in groundwater. The Fluxsampler is a hydraulically active unit filled
with sorbents that retain specific groups of chemical solutes, and an environmentally neutral tracer
salt. The sorbent and salt are contained in a permeable housing designed to fit in 2-inch groundwater
wells. Upon installation of the sampler in a groundwater filter casing, three sorbent sections intercept
the horizontal component of the mass flux. The groundwater also passes compartment with salt, that
dissolves proportionally with the groundwater volume passing the sampler housing. The installation
procedure is simple and robust, within approximately 20 minutes per well on average. The sampler is
after end installation, typically 1-3 months, brought to the laboratory, and analysed for the accumulated
mass of solutes in the sorbent sections and the loss of tracer salt. The Darcy flux and solute fluxes are
then derived from loss of salt and accumulated mass of solutes, respectively. Optionally, the horizontal
flow direction can be derived from a triangular vector calculation on the content of the three
subsections of the sorbent phase. This contribution will present the basic principles underlying the
passive sampling approach, and present results from two studies that were performed at VOCl
contaminated sites in the Capital Region of Denmark. Flux assessments and groundwater flow
directions were derived with both traditional methods and with the Fluxsampler. The results document
this approach as a valuable supplement for traditional measurements both for site investigation and for
on-going monitoring. The prospects of the method, advantages and limitations are discussed.
INTRODUCTION
The contaminant mass flux or mass discharge is increasingly recognized as the one the key
descriptors for groundwater risk assessments. An assessment of the mass discharge is highly relevant
for a.o. the preliminary ranking and prioritizing of sites for remediation purposes, evaluating ongoing
remediation or natural attenuation, and general monitoring purposes.
Traditionally, transect methods are used for contaminant mass based on groundwater gradients and
hydraulic conductivities. More recently, different methods are proposed including advanced pumping
methods1 and passive sampling methods2 . In the current paper we introduce a novel passive
sampling technique, Fluxsampler, for assessments of the groundwater Darcy flux and solute mass
discharge in groundwater. The sampling technique builds on combining sorbents retaining specific
groups of chemical solutes with an environmentally neutral tracer salt, dissolving proportionally with
sample volume 3. These features combined in a single passive sampler, a small Sorbicell cartridge,
can be used in recipient water4 and groundwater5 for monitoring of time-weighed solute
concentrations. The FluxSampler is based on the same generic sampling principles, however evolves
a special mounting that is suitable for direct mass discharge in groundwater. The purpose of the
present paper is to present the conceptual approach, and interpret the results with reference to a set
of conventional data for three cases from the Capital Region of Denmark.
Figure 1 shows the Fluxsampler, a cylinder-shaped unit with three sorbent sections (a, b, and c)
encompassing a central triangular shaped tracer salt section. Upon installation of the sampler in a
borehole, the sorbent sections intercept the mass flux, Jc. The groundwater further passes the centre
compartment and tracer salt dissolves proportionally to water flux q passing the sampler. The sampler
is after end installation brought to the laboratory, analysed for content of solutes in the sorbent
sections, and loss of tracer salt. The flux parameters Jc , q and cw are derived from loss of tracer salt
and accumulated mass of solutes. The horizontal flow direction can be derived from a triangular vector
calculation on the content of solutes in section a, b, and c.
Fig 1. Picture of Fluxsampler and schematic of the installation in groundwater wells for the vertical and
horizontal delineation of groundwater- and mass discharge.
Flux samplers are directly installed in HDPE or PVC filter casings with internal diameter of 50 mm on a
stainless steel foot for positioning the sampler at the center of the filter screen. Aluminum poles with
click-couplings are used for compass orientation of sampler and pushing the sampler to depth (Figure
2). In all cases, the samplers were placed such that the section a was pointed towards north. At one of
1
Wheeldon, J.G. 2008. An evaluation and implementation guide for current groundwater mass flux
measurement practices. U.S. Air Force Institute of Technology.
2
For review see Verreydt, G. et al. 2010. Ground Water Monitoring & Remediation 30, 2: 114126.
3
De Jonge, H. and Rothenberg, G. 2005. Environ. Sci. Technol., 39, 274-282.
4
Rozemeijer, J. et al. 2010. Environ. Sci. Technol., 44, 13531359.
5
Heinicke, G. and Andersson. M.T. 2009. DHI Nordic Water Technology Verification Centers, 80144.
the three sites (Holten), a range of samplers were installed in meshed stainless steel filter filters, to
allow for more direct capillary contact with the sediment formation.
Fig. 2. Picture of the orientation of the sampler within the well and mounting with aluminium click-
poles.
Installation periods were from 8-10 weeks (see below). After end sampling, the samplers were
retrieved, transported to the laboratory in transportation tube, and processed in the laboratory of
Sorbisense. Upon arrival in the laboratory, the sampler mesh fabric for each section was cut, and the
sorbent from sections a, b, and c (40 gram each) were collected separately. A subset of 10 g of
sorbent material was sent to Eurofins laboratories (Vejen, Danmark) for extraction and quantification of
VOC compounds by GC-MS. The remainder of tracer salt is determined by gravimetric weighing.
The volume of groundwater passing the sampler during exposure to groundwater, was derived from
the loss of tracer salt Mtracersalt (g) according to:
V
Q=
t
With Q is discharge (m3 t-1); t = length of sampling period. We made the simple assumption that
groundwater divergence or convergence and vertical flow can be ignored, hence the horizontal water
flux Jw was calculated by dividing the sample volume by the vertical cross section area A of the
sampler (m2):
V
Jw =
A t
With Jw = water flux (m t-1). Again assuming that groundwater flow divergence or convergence around
the sampler, and the effect of diffusion is negligible, the solute mass flux, Jc (g m-2 t-1) was derived
from:
Ms
Jc =
A t
The solute concentration, Cs (g m-3) was finally calculated from dividing the total mass of solute by the
volume of water exposed to the sampler:
Cs = M s / V
The sample volume range is determined by the starting amount of tracer salt (110 gram) and the
solubility product. The optimal sampling range of 7 - 65 L of water corresponds to 5-50% of the initial
tracer salt amount leached. If we assume that 7L 65 L is the accurately determined volume range,
this should be divided by the cross sectional area of the sampler (0,013 m2), corresponding to a 0,5 - 5
m of horizontal groundwater flux as the optimal measuring range.
The lower end of solute measuring range, the detection limit, is determined by the smallest detectable
mass of solute that can be extracted on the resin; for VOCs, this is approximately 0,2 ug/g, or
approximately 8 ug per compartment. Detection limits of the solute flux, Js, are then calculated by
dividing the detection limit of one compartment (ug) by cross sectional area and sampling time. For
example, the lowest solute flux trichloroethene (TCE) that can be measured over a period of 4 weeks
is calculated as:
0,2E 6 40
J TCE,low = = 22E 6 gramTCE m 2 day 1
0,013 28
Note that the calculation of the concentration assumes that the sorbent behaves like an infinite sink,
meaning that the sorption capacity is not exceeded. The higher end of the measuring range is
therefore determined by the sorption capacity. Although solute specific, a guideline for total VOC load
capacity is 2,5 mg/gram, or 300 mg of total VOC load for the sampler (sum of section a, b, c). Hence,
the highest VOC flux that can be measured over a period of 4 weeks is:
2,5E 3 120
J VOC,high = = 0,82gramVOC m 2 day1
0,013 28
The horizontal flux direction is derived from a triangular vector calculation from mass content of the
three sorbent sections, ma, mb, and mc.
Figure 3. Schematic top-view of sampler. Solute mass accumulation, left: flow from north to south,
section ma > mb mc ; right: flow from East to West, mb > ma > mc.
Section a, b and c each represent a 120 degrees segment of the circumference of the sampler. When
the section A is oriented towards the north (angle =0), section a intercepts the segment = 240o-60o;
section b intercepts segment 180o-240o; and section c intercepts the segment 60o-180o. The vectors
, have the direction =0, 240, and 120 degrees respectively, in the x, y coordinates
, and
(counterclockwise rotation, see figure below):
Fig. 4. Schematic of the derivation of the direction of mass movement from the mass accumulated on
the three adjacent adsorbent compartments. The vectors , , and each intercept a 120 degrees
segment of the circumference of the sampler.
Using the atan2 function accounts for the signs of both vector components, and places the angle in
the correct quadrant. Finally, the angle should be corrected for if the section a is not pointing north,
and the degrees are corrected as to follow the clockwise compass angles.
Field deployments are presented here from three contaminated sites from the Capital Region of
Denmark. All of the three sites had a longer data record so that a valid reference dataset was available
for an interpretation of the results obtained with the Fluxsampler.
The first site, located in Ballerup, is contaminated with chlorinated solvents and remediation - through
drainage of primary groundwater - and monitoring has occurred since 1994. The geology of the site is
a limestone chalk overlain by morainic sand deposits. During the 2014 measuring campaign,
groundwater level were recorded from 14 wells, purged groundwater samples were taken from 13
wells, SorbiCells were mounted in 8 wells, and subsequently Fluxsamplers were installed in 5
selected wells 6 in the period of 8-5 to 24-7-2014 (77 days).
The groundwater contamination in Helsinge, site two, originates from a former furniture factory. The
geology of the site is quarternary deposits, mixed sandy silty, and clayey glacial meltwater deposits,
overlying the Damian chalk. In 2013 and 2014, an intensive site risk assessment was carried out
including soil, pore-gas and groundwater investigations 7. A total of 10 Flux samplers were installed in
the depth range of 11,40 to 28,30 m.u.t.in the period of 18-6 to 18-8-2014 (61 days).
Site three is located at a former gas-work in Holten. The geology of the site is quarternary mixed
sandy and silty glacial meltwater deposits, overlying the chalk at 35 m.u.y. Soil and groundwater at the
site is heavily contaminated with chlorophenols, hydrocarbons, and BTEX, and has been managed by
pump-and treat to protect a nearby waterworks, located NE from the source the plume. A site-
conceptual model was provided by 3D mapping of geological structure, groundwater samples, and
monitoring of the hydraulic gradients at the site with subsequent modelling of hydraulic equipotential
lines. A total of 12 Fluxsamplers were deployed and retrieved in the period of 23-5 to 25-7-2014 (62
days) within a depth interval of 9,50 to 38,0 m.u.t.
6
COWI, December 2014, ref. A020578
7
Dansk Miljrdgivning, November 2014, ref. 08008086
RESULTS AND DISCUSSION
The first site in Ballerup provides for a direct head-to-head comparison of regular groundwater
samples, Sorbicell, and results from Fluxsampler. This results are shown in Table 1 for sum of PCE,
TCE and sum of degradation products. No groundwater samples were taken from borehole B104,
because there was no sufficient recharge after pumping. The other boreholes were prone to slow
recharge after pumping, and hence the three subsequent well purges were performed over a period of
6 hr; table 1 shows concentration levels after the third purging. Groundwater samples and results from
Sorbicells detected VOCl at comparable levels, with somewhat lower concentrations for GW samples.
It should be noted here, that GW measured VOCl levels in borehole B103 decreased subsequently
after purge 1, 2, and 3 by almost a factor 2. The distribution of VOCl is clearly dominated by
degradation products (up to 100%).
The mass discharge and flux concentrations levels are ranked similarly as reference concentrations,
however the flux concentrations in the source area well are considerably lower reference
concentrations, whereas comparable levels are detected in the plume. Somewhat speculative, we
subscribe the difference in the source area to non-equilibrium conditions, most likely due to ongoing
degradation of chloroethenes in the source area.
DCE, 1,1-DCA; 4 Stagnant well;,no groundwater sample could be obtained; 5 No Sorbicell installed in this well.
Darcy fluxes measured in the five wells were very different, in the range 0,01-0,05 m day-1. Most of the
mass is discharged in southeast direction. Mass discharge in the source area is approximately 10 mg
m-2 day-1. Some of the directions are statistically uncertain, either due to low levels of VOCl, or in the
case of B104 probably due to a low Darcy Flux. The direction of the mass transport in the source and
plume area is consistent with the delineation of the plume, pointing in a south- to southeast direction
(Figure 5). No consistent flow pattern and equipotential lines could be derived from the groundwater
levels, hence this site is adjacent to a hydraulic groundwater divide. This corresponds well with the
very low Darcy fluxes measured in well B104 (Table 1, Figure 4).
Fig. 5. Data from Ballerup site (COWI). Delineation of VOCl plume derived from groundwater
concentrations in 13 wells, together with groundwater- and VOCl mass discharge and direction derived
with Fluxsampler in 5 wells. Arrows indicate direction and magnitude of mass discharge as measured
with the Fluxsampler.
Equipotential lines at the Helsinge site showed clear gradients pointing east-southeast for the shallow
wells adjacent the source area, and south-southwest for the deeper wells (Figure 6). This complex 3D
pattern most probably relates to the mixed deposits and inclusion of tilted sand layers. The mass
discharge measured at the shallow depth, in the source zone, is southeast, consistent with
equipotential lines. Mass discharge in the source zone (well B3-1) is 300 times higher than the
Ballerup site, and the flux concentrations in this zone were derived at a level of 5.300 ug/L, which is
matching the contour levels from the groundwater sampler (Figure 6, top right). The direction of mass
discharges in the deeper wells almost exclusively point in south bound direction which corresponds
well with the hydraulic equipotential lines (Figure 6, bottom).
In some of the wells calculation of flow directions were either not significant, and in fact two wells are
pointing towards west (these are marked as question marks in Figure 6, bottom). It may very well be
reflecting local deflections of the low, but the measurements were not yet repeated. It should always
be taken into account, that the fluxsampler represents a smaller volume, and as such should be
considered almost a point measurement. Darcy fluxes measured at this site were remarkably
constant, with a mean value of 0,04 m day-1 (15 m year-1) and RSD of 17% of the mean. These values
seem certainly plausible yet somewhat higher than the estimates made from the hydraulic gradients
and k-values, that were in the range 3-6 m year-1.
The presented data show great promise for the potential of implementing the Fluxsampler in various
stages of contaminated site management. The limitations of the method representing a small volume
should always be kept in mind, however the delineation of the mass discharge and - direction is
certainly crucial and difficult to deduct indirectly with standard transect methods. It is outside the scope
of this paper, however we plan to use the presented data as shown in Figure 5 and 6, and other sites
such as the former gaswork site in Holten (data not shown here), to calculate mass discharges along
transects perpendicular to the general flow direction in the plume. The aim is to accomplish the
ultimate goal to derive estimates of total mass of VOCls leaving the source areas and various planes
of compliance. Also we intend to use the data to derive a statistical estimate on the number of
Fluxsampler measurements needed in order to derive the general flow pattern in the primary aquifers.
Finally, we would continue the Fluxsampler approach as to include anionic markers for groundwater to
elucidate more accurately the 2D and 3D flow directions of groundwater in source areas and areas
with very low contents of organics.
Fig. 6. Data from Helsinge site (DMR). Delineation of VOCl plume derived from groundwater
concentrations, together with groundwater- and VOCl mass discharge and direction derived with
Fluxsampler; top left: hydraulic equipotential lines in upper groundwater; top, right: groundwater
concentration contour plots and results from Fluxsampler in upper groundwater; Bottom left: hydraulic
equipotential lines in deeper groundwater; bottom, right: groundwater concentration contour plots and
results from Fluxsampler in deeper groundwater.
CONCLUSIONS
At three sites within the Capital Region of Denmark, reference data were collected with regard to the
delineation of VOCl plumes, hydraulic gradients and estimation of groundwater discharge and flow
direction. Data from two sites are presented here and were compared to results obtained from a novel
passive sampler yielding directly measured Darcy flux, VOCl mass discharge, and mass flow
directions. The first site, Ballerup, was shown to be close to a groundwater divide and the hydraulic
gradients could not be used to elucidate the direction of groundwater flow. The horizontal distribution
and shape of the plume, as obtained from regular groundwater samples, reflected in retrospect the
mass flux away from the source area in SW direction. The results obtained with the passive sampler
confirmed this pattern, viz. a non-uniform distribution of Darcy fluxes, primarily S to SW-bound flow
directions, and consistently correlated solute (flux) concentrations and total mass flux estimates. At the
second site in Helsinge, the hydraulic gradients were much more pronounced, and plume delineation
revealed a complex 3D distributed pattern with horizontal gradients SW directions in the source area
close to the GW-water table, then a curved pattern with primarily SE and S-bound mass transport
away from the source area at larger depth. The results of the passive sampling technique were
consistently related to this, viz. a more uniform distribution of relatively high Darcy fluxes, SE-bound
mass transport in the source area, and mainly S-bound VOCl-fluxes deeper in the primary
groundwater. The results from these studies illustrate the large potential of the passive sampling
method for derivation of groundwater- and mass discharge for the purpose of groundwater risk
assessments.
Protocols for ecological risk assessment
M.H.A.B. Wagelmans MSc.
Bioclear BV, PO Box 2262, 9704 CG, Groningen
E-mail contact: wagelmans@bioclear.nl
1. Introduction
In the Netherlands, in case of expected ecological risks caused by soil pollution, the Triad approach can be
used to assess site specific ecological risks. This approach combines and integrates three lines of evidence
from a site to give a complete and realistic overview of the site specific ecological risks. The three fieldw of
expertise that are combined are environmental chemistry, toxicology (bioassays) and ecological field
observations. Although this method is fully validated and has been used in the Netherlandw ever since its
development (Chapman et al. 1987, Nobis 98-1-28, Jensen and Mesman 2006) a standard and legal
protocol for this type of research was missing.
From an evaluation of all Triad projects in the Netherlandw (project management by Bioclear) it was
concluded that authorities find it hard to judge on specialized work which is not formalised in legal protocols
(Wagelmans et al., 2009) which hampers decision making. In 2009 the Dutch Council of State (Raad van
State) has rejected the results of a specific Triad research. Because no protocol or guidelines for sampling
and sampling strategy for Triad research were present, the research protocol followed was compared to the
protocol for standard chemical soil research. However, the goals of Triad research significantly differ from the
goals of a standard soil research. Therefore, the protocol for standard soil research is not applicable to Triad
research and alignment of protocols for standard chemical soil research and Triad research was needed.
4. Example site
During development of both the technical protocol and the Dutch National Standard weve tested the draft
versions on practical applicability on the topsoil of a former landfill polluted with heavy metals (figure 1). The
site is being used for agriculture, grazing of sheep and cows. The site is part of a bigger redevelopment plan
in the area. Future use of the site will remain agriculture as part of a natural green zone. Based on model
calculations remediation of the site is needed due to possible ecological risks.
The first step was a workshop with all stakeholders (farmer, owner of the site, water regulatory authority, two
different local authorities, the provincial authority and scientists). Stakeholders were asked about future
development and use of the site. Scientists presented the research proposal containing chemical analyses
on soil and earthworms from the site, bioavailability analyses, bioassays with bacteria and ecological fiend
observations of nematodes, and the nitrification potential. Based on input from the stakeholders the research
proposal was extended with additional measurements on heavy metals in plants and bioassays with plants
and calculations of the agricultural risk based on grazing by cows and sheep.
In this workshop it was agreed to use the cleanest sample from the site as site specific reference sample, to
give the same weight to all of the results and to use a numerical method of integration of the results (RIVM
method [5]). Also agreements were made about the site specific risk boundary.
4.1. Sampling
After this workshop the ecological risk assessment has been carried out. Sampling has been carried out
according to protocol SIKB BRL 2301. Samples were taken in a square of 5 * 5 m per sampling point. For
chemical analyses, bioassays and nitrification potential composite samples were taken in this square using a
soil drill. Samples were taken from the first 30 cm because the highest density of soil life is found in this
layer. For nematode analyses samples were taken from the top 10 cm by means of a small gouge.
Earthworms were collected from two sodw of 25*25*25 cm (figure 2). Per sample 100 g of grass was
collected for chemical analyses of heavy metals.
Analyse/ sample Unit BC1 BC2 BC3 BC4 BC5 BC6 BC7 BC8 BC9 BC10
Chemical parameters
Dry weight (Dw) % 55.1 56 55.2 58.1 68.7 70.7 63.7 55.0 72.6 69.9
Metals
Arsenic-total mg/kg dw 9.3 15 9.6 11 8.4 10 7.1 8.6 <4 6.8
Barium-total mg/kg dw 680 450 200 590 450 150 330 650 98 51
Cadmium-total mg/kg dw 1.1 2 0.80 5.4 2.1 0.34 0.54 0.97 <0.17 0.24
Mercury-total mg/kg dw 0.67 1.4 0.83 2.4 0.87 0.26 0.77 0.28 0.13 0.16
Lead-total mg/kg dw 310 460 270 380 390 110 230 720 61 36
Molybdenum-total mg/kg dw <1.5 <1.5 <1.5 4.2 2.3 <1.5 <1.5 <1.5 <1.5 <1.5
Nickel-total mg/kg dw 22 20 16 34 17 20 15 20 12 12
Zinc-total mg/kg dw 490 960 450 750 470 190 400 900 170 91
PAH
Anthracene mg/kg dw 0.62 0.63 0.11 0.34 0.58 0.52 1.3 0.08 0.17 0.04
Benzo(a)anthracene mg/kg dw 4.7 2.6 0.83 1.4 1.9 1.6 5.0 0.44 0.59 0.19
Benzo(a)pyrene mg/kg dw 4.4 2.5 0.92 1.3 1.9 1.4 3.9 0.46 0.62 0.21
Benzo(ghi)perylene mg/kg dw 3.1 2.0 0.82 0.91 1.5 0.98 2.3 0.35 0.47 0.17
Benzo(k)fluoranthene mg/kg dw 2.4 1.4 0.51 0.73 1.0 0.70 2.0 0.24 0.30 0.10
Chrysene mg/kg dw 5.1 2.6 0.96 1.5 2.0 1.5 4.5 0.48 0.61 0.21
Fenanthrene mg/kg dw 3.1 3.0 0.58 1.8 3.3 2.2 5.4 0.44 0.72 0.17
Fluoranthene mg/kg dw 8.7 5.3 1.6 3.3 4 3.6 10 0.99 1.5 0.44
Indeno-(1,2,3- mg/kg dw 3.6 2.2 0.92 1.1 1.7 1.2 2.6 0.39 0.48 0.19
c,d)pyrene
Naftalene mg/kg dw <0.2 <0.5 0.043 0.38 <0.2 0.22 <0.5 0.04 0.08 0.01
Toxische druk (%) 33.1- 62.2 14.8- 68.7 17.9- 3.6 37.5 50.8 0-11.7 0
(msPAF) 74.2 43.6 59.9
*. For the calculation of the toxic pressure the lowest and highest organic matter and fraction < 2 m were used.
d= limit of detection
red=Higher than Dutch Intervention value or msPAF > 25 % (above which ecological risks can be expected)
Based on chemical analyses ecological risks can be expected at a big part of the location. Six of the ten
samples were chosen for the rest of the ecological risk assessment. First the availability of heavy metals was
measured. In none of the samples the availability was above the limit of detection.
Bioaccumulation was measured in earthworms and grass from the location. The concentration of heavy
metals in earthworms was comparable to that in worms from clean locations (personal communication
RIVM). Also no correlation was found between the concentrations of metals in soil and that in earthworms.
This was also found for the grass samples. Based on these results, bioaccumulation in earthworms and
grass is not present and risks of accumulation in the food chain are low.
4.3. Toxicology
Results of Microtox analyses are shown in figure 3. This figure shows there is no correlation between the
observed effects and the concentration of metals in soil (shown as toxic pressure). Probably the effects are
not caused by the metal pollution but by confounding factors.
1
0,9
0,8
0,7
0,6
effect
4.4. Ecology
In the ecological part of the Triad at this site nitrification potential and nematode community structure were
determined. Figure 4 shows the results of the nitrification potential. No correlation was found between the
nitrification potential and the pollution present (as toxic pressure). The values for nitrification are similar to
that in reference soils found by Bioclear (results not published).
0,07
0,06
0,04
0,03
nitrificatie
0,02
0,01
0
0 20 40 60 80
Toxische druk
Table 2 shows the results of the nematode analyses. In this table the MI is Maturity Index which is a
measure for the health of the nematode population. A low number means a disturbed population, a number
between 2,5 and 3,5 means a stable population.
Tabel 2. Results nematode analyses
The amount of nematodes is high, higher than the average amount of nematodes at comparable agricultural
sites (Rutgers et al, 2007). However, the amount of species is lower than described by Rutgers et al, also in
the clean (reference) samples. There is no correlation between the MI and the toxic pressure nor between
the amount of species and the toxic pressure. That means the pollution has no negative effect on
nematodes.
4.5. Summarized
Although high concentrations of metals were found leading to a high toxic pressure at the site, the available
fractions of metals were low. Based on this low availability, negative effects were not expected. The results
of the measurements conform this expectation. No significant correlations were found between the toxic
pressure and the organisms. Also no bioaccumulation of metals occurred in earthworms and grass from the
location.
Based on the ecological risk assessment was concluded that the pollution in the topsoil does not cause a
negative ecological effect nor a negative effects on grazers. Remediation of the site is not necessary and
redevelopment of the area can take place as planned. This result was communicated with the stakeholders.
Stakeholders were pleased with the process and felt taken seriously during the entire process. The
redevelopment of the entire area will take place as planned.
5. Conclusions
The Dutch National Standard on ecological risk assessment and the technical guideline on sampling and
sampling strategy are valuable tools in risk communication. It increases acceptance for the research plan,
the results and the final decision because stakeholders have influence on the process. In the process their
questions and objections will be answered and solved in an early stage. Because of this acceptance by
stakeholders, it is easier for decision makers to make a final decision based on site specific research,
especially now it has been formalized in standard protocols. By using the standard and technical guideline
risk assessment is not a project for only scientists anymore but it becomes a process for both decision
makers and scientists together. Also legal authorities can determine whether the research was performed
correctly following the technical guidelines which is positive for future acceptance of Triad projects.
References
[1] Chapman P.M. (1986). Sediment quality criteria from the Sediment Quality Triad an example.
Environmental Toxicology and Chemistry 5, 957964
[2] Nobis 98-1-28 (2000). TRIADE benadering voor beoordeling van bodemkwaliteit.
[3] Jensen J., Mesman M. (2006). Ecological Risk Assessment of Contaminated Land. Decision support for
site specific investigations. ISBN 978-90-6960-138-0 RIVM Bilthoven.
[4] Wagelmans M., M. Mesman, S. Kools, D. Lud, J. Lahr, Evaluatie van de toepassing van TRIADE bij het
beoordelen van ecologische risicos, SKB, Gouda, 2009.
[5] Mesman M., Schouten A., Rutgers M., Dirven-van Breemen E. Handreiking TRIADE: Locatiespecifiek
ecologisch onderzoek in stap 3 van het Saneringscriterium. RIVM 711701068, 2007
[6] NEN 5737:2009 Soil- Process of site specific ecological risk assessment of soil pollution, 2010
[7] SIKB BRL protocol 2301, Preparation and implementation of fieldwork for Triad Research of land soils,
2013
[8] Wagelmans M., Lieten S., Rutgers M., Mesman M. Triade-onderzoek voormalige stort Knoopslaan te
Wassenaar, Eindrapport SKB PT 8447, 2010.
[9] Rutgers M., C. Mulder, A.J. Schouten, J. Bloem , J.J. Bogte, A.M. Breure, L. Brussaard, R.G.M. de
Goede, J.H. Faber, G.A.J.M. Jagers op Akkerhuis, H. Keidel, G.W. Korthals, F.W. Smeding, C. ten Berg C,
N. van Eekeren, Typeringen van bodemecosystemen in Nederland met tien referenties voor biologische
bodemkwaliteit, RIVM 607604008, 2007
Title:
Residential location contaminated with cumene: building team construction results in a successful (in-
situ) remediation
Oral presentation
Abstract:
Cumene or isopropylbenzene - is a volatile aromatic hydrocarbon with an odor threshold below
the chemical-analytical detection limit. The soil contamination with cumene was caused by a road
accident in which a truck loaded with Komol fell over and 10,000 l of Komol with cumene as the
principal compound - poured out all over the soil surface. In 2012, soil investigation showed:
a soil contamination source location of about 750 m2 until 7 m below ground level
a ground water contamination up to 350 m from the source location and to a depth of 25 m below
ground level.
The Dutch province of North Brabant the problem owner - was confronted with many
questions. How to assess the risks of this compound for which no general risk assessment for soil
contamination exists? How to translate risk levels into remediation goals? What remediation
approach will abolish all odor nuisance for the residents of the location? How to prevent nuisance
during the remediation activities (as much as possible) and how to protect the health of the
residents and contractors laborers? The province of North Brabant had to address these and
many other questions in such way that the final aim of the remediation would be achieved in the
most effective way and with respect to the complex situation with many stakeholders involved.
The presentation on Consoil will show the results of the investigation (conceptual model), risk
assessment, in-situ remediation measurements and the applied process of a building team
construction, which have resulted altogether in a successful remediation.
The soil and ground water investigation has based on a conceptual model, which gave insight in
the information gabs, research questions to be answered and a clear visual insight in the
contamination. The risk assessment consists of two levels: technical risk assessment concerning
the human and environmental risks of the contamination and risk management (using the
RISMAN method) to clarify and to address properly all project related risks. One of the most
important issues of the technical assessment is the impact of the extremely low odor threshold
level on the remediation goals to be achieved and the remediation activities needed to secure
these goals. The conceptual model mentioned above and risk management have been the
starting points of the project approach and project risks are determined and prioritized by the
stakeholders involved, e.g. extreme odor nuisance, conventional explosives (second world war)
and groundwater extraction. In 2012, an evaluation of the most appropriate contract form and
market analyses followed the risk management session hold and as a result the building team
construction has been chosen as the best suiting contract form and a short list of contractors was
generated who were invited to compete for the contract. Furthermore, the advantages of the
building team construction will be shown:
- the province of North Brabant is able to influence the process in such a way that the
province can take its responsibility in this complex situation in which residents are
involved
- members of the team (province, consultant Tauw and contractor NTP) depend on each
other for the realization of the project and project risks are addressed and solved in the
most effective way possible
Finally, the remediation consists of both excavation activities and in-situ measurements. The
possible in-situ remediation measurements biological, fysiological and/ or chemical - have been
evaluated and it has been decided to implement a bioscreen to remediate the ground water
contamination and to imply a bio blanked with a slow release compound on the floor of the
excavation.
Author and presenter: Peter Ramakers
Co-authors: Joost van Schijndel and Gerard Borggreve
ThS 1B.9 Risk modelling
1 1 1 1 2
Paul Frogner-Kockum , Mrta Lndell Gunnel Gransson , Yvonne Ohlsson , Henrik slund
1
Swedish Geotechnical Institute, SE-583 91 Linkping, Sweden
2
University of Gothenburg, SE-40530 Gteborg, Sweden.
1
paul.frogner-kockum@swedgeo.se
Abstract
The Swedish Geotechnical Institute (SGI) recently got a government mandate to develop approaches
to increase the remediation rate of contaminated areas in Sweden. More effective risk assessments
can contribute to this. In Swedish risk-assessments of sediment, estimates of contaminant transport at
the sediment/water interface have so far been done using diffusive flux models. However, difficulties
exists using models for sediments: diffusive fluxes are subject to considerable uncertainty in model
coefficients, and the overall fluxes (diffusive and advective) in sediment might be much greater than
only diffusive fluxes. The aim of this project is to test different methods to quantify contaminant fluxes
from sediments. A comparison of contaminant fluxes between in situ measurements with the benthic
flux chamber method (BFC), diffusion calculations from sediment samples/water samples and surface
leaching tests from sediment samples show that fluxes from BFC measurements and surface leaching
tests are similar in magnitude while fluxes estimated by diffusion calculations deviate significantly from
the others. Calculated fluxes deviate at least by a factor of 2 from measured values by the BFC and
surface leaching tests, which means that modeled fluxes seems to underestimate spreading of
pollutants. Thus, the study indicates that risk assessment based on calculated fluxes may give
misleading estimation of contaminant spreading from sediment where this would be underestimated.
By sequential leaching, it also evident that it should not either be recommended to base a risk
assessment on total concentration in sediments (although it may be useful as a first screening of
contaminants to search for hotspots). This study shows that copper mainly is tightly bound in the
sediment, which explains why the measured flux rate of copper was so low.
1. Introduction
In the Swedish method for risk assessment of sediment, estimates of contaminant transport at the
sediment/water interface have so far been done using diffusive flux models. Diffusive flux models are
based on concentration gradients between sediment pore water and overlying water or between pore
water concentrations at different sediment depths (Eek et. al., 2010). However, difficulties exists using
models for sediments: diffusive fluxes are subject to considerable uncertainty in model coefficients,
and the overall fluxes (diffusive and advective) in sediment might be much greater than only diffusive
fluxes (Hampton and Chadwick 2000). In this study, a method for in situ measurement of contaminant
transport from sediments to the overlying water column is developed and analyzed.
The study was initiated as one part of a government mandate that recently was given to the Swedish
Geotechnical Institute with the aim to develop approaches to increase the remediation rate of
contaminated areas in Sweden. An effective risk assessment can contribute to this.
The project aims to improve the methodology to quantify the contaminant fluxes from sediment to the
overlying water mass and thus contribute to an improved risk assessment of contaminated sediments.
Quantify contaminant transport in situ with the help of the BFC methodology compared to
diffusion calculations and surface leaching tests. The comparison is made in order to evaluate
the usefulness of these methods in a risk assessment context and when one method is
preferred over the other.
Make recommendations about when different methods are suitable in a risk assessment
context.
2. Method
The project has, as shown in figure 1, been implemented in the following steps:
(1) Sediment sampling and characterization by total metal concentrations, pore water analysis and
sequential leaching.
(2) Quantification of the fluxes using three different methods: a diffusion calculation method, an in situ
method (the benthic flux chamber method) and a surface leaching test.
(3) The results are evaluated and the fluxes compared.
Figure 1. Work schedule of the study divided in three different steps: characterization, quantification of
fluxes by three different methods and comparison of fluxes.
The characterization was done by analyzing the total metal concentration in the sediment and different
metals leachability by using a sequential leaching test (UNEP / IOC / IAEA, 1995) and information
from the metal concentrations in the pore water. Total concentration in sediments gives a first idea of
pollutant sources present in the sediment, but not how lose or hard contaminants are bounded in or to
particles and organic substances. The sequential leaching thus provides complementary information
on how different pollutants are bounded in the sediment and to which fractions (in minerals, at mineral
surfaces or to an organic fraction). These two methods together thus made it possible to find out what
is hard and loosely bound and how large fraction of the total content that can be spread and become
bioavailable.
Sediment pore water was analyzed to gives information on the actual amount of metals that are
available in the pore water and that can be transported out of the sediment. In step (2), three different
methods are used in order to quantify the metal transport out of the sediment. A diffusion model (Li
and Gregory, 1974; Awakura et. al., 1988), a surface leaching test in the laboratory (NEN 7347) that
tries to simulate what happens in reality and an in situ method (Benthic flux chamber) that measures
what happens in reality (Thorbergsdttir and Gslason, 2004; Thorbergsdttir et. al., 2004). In step (3)
the metal fluxes (concentration per surface unit and unit of time) measured and calculated from the
three different methods are compared.
2.1 Sediment sampling and characterization
Helsingborg city has an environmental monitoring program of coastal sediments since 1995, and
elevated concentrations with respect to Zn, Co and Cu in a sampling point labeled KED has been
detected. The sampling point KED is located in a small marina called Knhaken harbor adjacent to the
larger Bulk harbor. Elevated metal concentrations in the sampling point KED most likely originate from
a pipe that drains surface water from larger parts of the southern Helsingborg. The high total metal
content in the sediment samples was one of the reason why Knhaken harbor was chosen to test if
contaminant fluxes measured with the benthic flux chamber method (Hampton and Chadwick 2000)
would differ compared to diffusion calculations based on pore water samples (Li and Gregory, 1974;
Awakura et. al., 1988) and surface leaching tests based on sediment samples (NEN 7347).
Figur 2. Sampling points in the Knhaken harbor close by the larger bulk harbor (harbor basin to the
left in the figure). The figure shows approximate location of the sampling points in the Knhaken
harbor with the sampling site KED furthest into the harbor closest to the surface water pipe outlet
(storm water), and sampling points K1, K2 and K3.
Sediment cores were taken in the sampling points (Figure 2) by using a haps-corer sampler. The
Haps-corer sampler is a tube sampler that routinely has been used in the Helsingborg coastal
monitoring program for about 20 years (Gransson, 2013). Hence, we used a haps-corer sampler in
order to facilitate the comparison of the results in this study (FLUXSED project) with that of the
previous sampling and analyzes made in the Helsingborg coastal monitoring program. When studying
the redox profile of the sediment in the sampling location closest to KED in the Helsingborg coastal
monitoring program it is seen that only the uppermost 2-5 centimeters are oxygenated and that the
rest of the sediment is reduced (anoxic). Thus sediment samples were taken from the sediment core in
sections of 5 cm each at the following depths: 0-5 cm, 5-10 cm and 10-15 cm. However, the choice of
5 cm sections was also important to be sure to get enough pore water in order to make the pore water
analyzes that the diffusion calculations are based on (Li and Gregory, 1974; Awakura et. al., 1988).
The sequential leaching was performed according to the method UNEP / IOC / IAEA (1995). The
method includes leaching with a 25% acetic acid (fraction 1), followed by a residual leaching (fraction
2) on the same sediment sample. Fraction 1 consists mainly of exchangeable metals but also of the
less readily exchangeable metals that are organically bound and/or carbonate-bound (Rao et. al.,
2008). Acetic acid used in this step does not affect silicates or iron and manganese minerals (UNEP /
IOC / IAEA, 1995). The residual-fraction, consists of the more strongly bound metals released with the
"Four Acids" method (ALS, MS ME61A). The four acids method is a mixture of several strong acids
(ALS, MS ME61A) which releases all metals in the sample because the majority of silicates and iron
and manganese minerals are dissolved (ALS Global, 2014).
In order to estimate the transport of contaminants in the dissolved phase out from the sediment have
the diffusive flux of cobalt and zinc been calculated by applying Fick's first law:
J = - D C/x
where J is the diffusive flux, D is the effective diffusion coefficient and C/x is the concentration
gradient.
In the literature diffusion coefficients for specific ions, e.g. cobalt (II) is available (Ullman and Aller,
1982). These coefficients are temperature and pressure adjusted in aqueous solutions in the range
-5 -6 2 -1
between 10 - 10 cm s .
In natural systems, it is rare with free ions and metals are transported to a varying degree bound to
organic or inorganic ions or complexes. Coefficients for these complexes are unknown and are
therefore difficult to calculate or to propose. This means that the diffusion coefficients used in these
calculations, which have been determined in laboratory experiments (Ribeiro et al., 2002) and in pure
solutions (often at higher concentrations than in the natural environment) will not be entirely relevant in
a natural environment. But this modeling is as close as we can get in a simulation of natural
conditions.
In the diffusion calculations a correction also must be introduced due to that transport distance is not a
straight line which is because of the sediment porosity. A turtuositet of 1.3 and a porosity of 0.7 have
therefore been adopted.
The method NEN 7347 or alternatively CEN / TC351 / WG1 1N 179 is used for surface leaching test of
inorganic substances from materials with fine particles such as silt and sediment where diffusion
controls the transport of leached substances. The principle is that the material is packed in a cylinder.
This cylinder is then placed in a larger cylinder of water. A pre-determined amount of water is added
and this volume is replaced with fixed time intervals. During surface leaching the sample is leached in
a cylinder with water. The water flows not through the sample as the leaching of substances takes
place through the sample surface by diffusion. The leachate is then analyzed for the ions of interest.
In this study a deviation from the standard-method NEN 7347 was made as the sediment is sampled
in the same PE pipes as later was used as test cylinders. The tube was cut off to two cylinders on
which a base plate was glued to. This deviation from the standard method was made in order to get
undisturbed samples and thus not receive re-worked sediment material in the samples. As shown in
the pore water analyzes from different sections of the sediment there is a clear concentration gradient
in sediment profile which is why the sediment must not be re-worked. After this deviation and changes
in sample preparation from the original sample preparation procedure in NEN 7347, the interest
increased significantly of using NEN 7347 as a comparison method to the BFC method in this project.
This is because the new sampling methodology preserves the sediment core in an unaffected
condition. After the changes in sample preparation the NEN 7347 method mimics natural conditions in
that a column of water (which is replaced at regular intervals to avoid that equilibrium occurs) is in
contact with the undisturbed sediment top layer.
The sampling is done with a so-called Benthic flux chamber (BFC) deployed on the sea floor (Figure
3). It is a closed system at the sediment-water interface due to that the chamber is open in the bottom.
At given times after the flux chamber is deployed on the seafloor, the contaminant transport into the
flux chamber and thus the accumulation of contaminants in the chamber is measured with time.
The sampling container of the BFC has the form of a straight block having a square open towards the
sediment. The equipment has a base that is 30 x 30 cm and a height of 50 cm. The chamber also has
a stirrer for mixing the water-mass (homogenizing) and a source of power to the stirrer. One end of
tubing is attached to the chamber and the other end goes up to the water surface. This tubing is
attached to a pontoon at the water surface and at each sampling occasion water is pumped to the
surface by means of a peristaltic pump.
Sampling was carried out so that the samples all together were less than 1% of the total water volume
in the box. This meant that corrections for the total sample volume were not needed as <1% was
considered to be negligible.
The samples were analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS) and Gas
Chromatography-Mass Spectrometry (GC-MS) to determine the flux for different types of contaminants
from the sediment per unit area and time. An entrance to the chamber was also made for a probe
containing electrodes for measuring pH, redox, oxygen and conductivity. A frame with lead weights
was coupled to the chamber so that it would stand firmly on the sediment to ensure there is no
leakage in submerged conditions at the bottom.
It is important to keep the water in the BFC chamber at 10-20% oxygen-saturation so that the redox
conditions in the sediment surface do not change (Tengberg et al., 2005).
Figure 3. A schematic drawing of a Benthic flux chamber with a stirrer, lead weights and a probe to
measure pH, redox, oxygen and conductivity. The chamber has the dimensions 30x30x50 cm and is
open in the bottom to allow transport of pollutants into it.
According to NEN 7347 the sediment should be re-worked in a sample cylinder before surface
leaching tests are done. This procedure made the NEN 7347 method irrelevant for comparison with
other methods (See chapter 2 where the sampling preparation procedure was changed to make NEN
7347 more suitable for comparison). It may be seen from sediment analysis. When studying the redox
profile (Figure 4) of the sediment in the sampling location KED it is seen that only the uppermost 2-5
centimeters are oxygenated and that the rest of the sediment is reduced (anoxic).
Figure 4. Redox profiles from the report, "Helsingborg coastal monitoring program 2012". R0.5 Nov
respectively R1 Nov are two sample points that is close to the Knhaken harbor. Thus the sediment
(between years 1987-2012) in the coastal area outside the Knhaken harbor has been oxygenated
with a redox transition at about 2-5 cm depth.
From the results, it is also seen that the organic content decreases with depth (Figure 5) and that
metal content increases in a gradient between the surface sample and further down into the sediment
(Figure 6).
120
100
80
mg/l
60 DOC
40 TOC
20
0
KED 0-5 cm KED 5-10 cm KED 10-15 cm
Figure 5. Organic carbon in sediments pore water with depths at the sample location KED. The
organic carbon content decreases with depth in KED as in the other analyzed sediment core K1 in
Knhake harbor.
60
50
40
Cu
ug/l
30 Zn
20 Co
10
0
KED 0-5 cm KED 5-10 cm KED 10-15 cm
Figure 6. Metal content in sediments pore water with depts at the sample location KED. The metals
Cu, Zn and Co increases with depth in KED as in the other analyzed sediment core K1 in Knhaken
harbor.
The sequential leaching of the sediment regarding the metals cobalt, zinc and copper is described in
figure 7, 8 and 9. It is clear from the histograms that a relatively lower amount of copper, compared
with cobalt and zinc, is released in the sequential leaching step with acetic acid (HOAc). It was not
possible to detect any general trend in terms of fractionation with depth in the sediment between
different stations. As shown in Figure 7, 8 and 9 mainly copper is strongly bound in the sediment,
while zinc and copper both persists as strongly and loosely bound.
cobolt
8,00E+02
7,00E+02
6,00E+02
5,00E+02
ppm
4,00E+02
3,00E+02 Co - MS61
2,00E+02 Co - HOAc
1,00E+02
0,00E+00
K1 K2 K3 KED K1 K2 KED K1 KED
0 2 7
2 5 15
Station number Depth of subsample's top/bottom ends [cm]
Figure 7. Sequential leaching of cobalt for each station, grouped by sediment samples and range of
depths in the sediment. HOAc is acetic acid leaching and MS61 (read: MS ME61A) is the residual
concentration leached with several strong acids. The combined bar height represents the total content.
zink
7,00E+04
6,00E+04
5,00E+04
4,00E+04
ppm
3,00E+04
Zn - MS61
2,00E+04
Zn - HOAc
1,00E+04
0,00E+00
K1 K2 K3 KED K1 K2 KED K1 KED
0 2 7
2 5 15
Station number Depth of subsample's top/bottom ends [cm]
Figure 8. Sequential leaching of zink for each station, grouped by sediment samples and range of
depths in the sediment. HOAc is acetic acid leaching and MS61 (read: MS ME61A) is the residual
concentration leached with several strong acids. The combined bar height represents the total content.
copper
2,50E+03
2,00E+03
1,50E+03
ppm
1,00E+03
Cu - MS61
Cu - HOAc
5,00E+02
0,00E+00
K1 K2 K3 KED K1 K2 KED K1 KED
0 2 7
2 5 15
Station number Depth of subsample top/bottom ends [cm]
Figure 9. Sequential leaching of copper for each station, grouped by sediment samples range of
depths in the sediment. HOAc is acetic acid leaching and MS61 (read: MS ME61A) is the residual
concentration leached with several strong acids. The combined bar height represents the total content.
The results highlight the inappropriateness of using only total concentrations for risk assessment of
sediments. A classification based on total concentration is a first step (Trin 1) in the Norwegian
national guidelines for risk assessment of contaminated sediments (SFT TA-2230/2007) where it is
used for screening that will give a first indication of hotspots of contaminats in sediments. The
inappropriateness of basing a risk assessment on total concentrations is shown by the comparison
between total concentrations, pore water analysis, surface leaching tests and the BFC measurements
of this study. The study shows that a risk assessment based only on total concentrations had
assessed the risk of spreading copper larger than the dispersion of cobalt because its total
concentration is greater in the sediment. A risk assessment based on total concentrations had proved
misleading, as can be seen when we study the spread of both substances with the BFC chamber and
with surface leaching tests. In the use of these methods shows that cobalt fluxes dominates over
copper and zinc fluxes. Even pore water analyzes show the inadequacy of using total concentrations
as a basis for a risk assessment as the cobalt content dominates over both zinc and the copper
content in the pore water, contradicting the trend that emerges by studying the total concentrations.
However, total analysis may be used for screening of contaminants in sediments to get a first opinion
where hotspots are as it is used in the Trinn 1 of the Norwegian national guidelines for risk
assessment of sediments.
A comparison of contaminant fluxes between Benthic Flux Chamber (BFC) measurements, diffusion
calculations and surface leaching tests shows that fluxes from BFC measurements and surface
leaching tests are similar in magnitude while fluxes estimated by diffusion calculations deviate
significantly from the others. Calculated fluxes deviate at least by a factor of 2 from measured values
by the BFC and surface leaching tests, which means that modeled fluxes seems to underestimate
spreading of pollutants. Differences between calculated and measured values have also been
observed in previous studies. In a study of Pakhomova et. al. (2007) measurements by the BFC
method resulted in significant increased fluxes of iron and manganese compared to calculated fluxes
(Pakhomova et. al., 2007).
Figure 10. Comparison of contaminant fluxes between Benthic Flux Chamber (BFC) measurements,
diffusion calculations and surface leaching tests
4. Conclusions
When comparing measured contaminant fluxes with different methods, this study shows that fluxes
which are measured with the benthic flux chamber and surface leaching tests are similar in magnitude
while modelled fluxes differ from the others. Modelled fluxes deviate at least by a factor of 2 from
benthic flux chamber measurements. On the other hand, gives surface leaching tests similar results as
benthic flux chamber measurements, which means that modelled fluxes under estimate contaminant
transport from sediments. Based on the results from this study, modelled fluxes are not recommended
in a risk assessment perspective as it is shown to underestimate the contaminant spreading from the
sediment. The total metal analysis compared to the sequential leaching results also indicates that it
should not either be recommended to base a risk assessment on total concentrations in sediments.
Although the total Cu content was high in the sediment, the sequential leaching shows that Cu is
mainly strongly bound in the sediment, which also explains why fluxes of Cu from the sediment
measured were low. However, total analysis may be used for screening of contaminants in sediments
to get a first opinion where hotspots are.
From the results of this study we recommend that the Benthic flux chamber is used together with a
surface leaching test as the surface leaching test only will give the diffusive transport component while
the Benthic flux chamber may give contribution from both diffusion and advection.
5. References
Almroth, E., Tengberg, A., Andersson, J.H., Pakhomova, S., Hall, P.O.J., 2009 Effects of re-
suspension on bentic fluxes of oxygen, nutrients, dissolved inorganic carbon, iron and
manganese in the Gulf of finland, Baltic Sea.
ALS Global., 2014., Schedule of services and fees 2014 EUR. 2014. Hmtad den 6 oktober 2014, frn
http://www.alsglobal.com/en/Our-Services/Minerals/Geochemistry/Service-Schedule
Awakura, Y., Toshiya., D., and Hiroshi, M., 1988, Determination of the Diffusion Coefficients of
CuSO4, ZnSO4., and NiSO4 in Aqueous Solution. Metallurgical Transaction B, Volume 19b,
February 1988.
Eek, E., Cornelissen, G. and Breedveld, G.D., 2010, Field Measurement of Diffusional Mass Transfer
of HOCs at the Sediment-Water Interface. Environmental Science and Technol 44:6752-6759.
Gransson P, Karlfelt, J., and Bertilsson-Vuksan, K.S., 2013. Kustkontrollprogram fr Helsingborg.
Miljnmnden i Helsingborg., rsrapport 2011 & 2012.
Hamton, T., W., Chadwik, D., B., 2000, Quantifying In situ Metal Contaminant Mobility in Marin
sediments, Technical report 1826, Spawar system Center (SSC) San Diego, CA 92152-5001.
Li, Y.H. and Gregory, S., 1974, Diffusion of ions in sea water and in deep-sea sediments. Geochim.
et. Cosmochim. acta, 38:703-714.
Pakhomova, S.V., Hall, P.O.J., Kononets, A.G.R., Tengberg, A., Vershinin, A.V., 2007, fluxes of iron
and manganese across the sediment-water interface under various redox conditions, Mar. Chem.
107:3, 319-331.
Rao, C.R.M., Sahuquillo, A., & Lopez Sanchez, J.F. (2008). A Review of the Different Methods Applied
in Environmental Geochemistry For Single and Sequential Extraction of Trace Elements in Soils
and Related Materials. Water, Air, and Soil Pollution, 189(1-4), 291333.
Ribeiro, A.C.F., Lobo, V.M.M., and Natividade, J.J.S. (2002). Diffusion Coefficients in Aqueous
Solutions of Cobalt Chloride at 298.15 KJ. Chem Eng Data 47, 539-541.
SFT, 2007, Veiledare for risikovurdering av forurenset sediment (TA-2230/2007)
Tengberg, A., Hall, P.O.J., Anderson, U., Linden, B., Styrenius, O., [et al.], 2005, Intercalibration of
benthic flux chambers II, Hydrodynamic characterization and flux comparision of 14 different
designs. Mar. Chem. 94:147173.
Thorbergsdttir I. M. and Gslason S.R., 2004, Internal loading of nutrients and certain metals in the
shallow eutrophic Lake Mvatn, Iceland. Aquatic Ecology 38:191-207.
Thorbergsdttir I. M., Gslason S.R., Ingvarson, H. R. and Einarsson ., 2004, Benthic oxygen fluxes
in highly productive sub-arctic Lake Mvatn, Iceland: In situ benthic chamber study. Aquatic
Ecology 38:177-189.
Ullman, W.J. & Aller, R.C., 1982, Diffusion coefficients in nearshore marine sediments. Limnol.
Oceanogr. 27:552-556.
UNEP/IOC/IAEA, 1995., Reference Methods for Marine Pollution Studies No 63. Manual for the
geochemical analyses of marine sediments and suspended particulate matter. UNEP, 1995.
Assessment of risks caused by the permeation of organic contaminants in
groundwater through polyethylene drinking water pipes.
P.F. Otte*1, M.L. van der Schans*2, M.A. Meerkerk*2, F.A. Swartjes*1
*1
National Institute for Public Health and the Environment, the Netherlands
*2
KWR Watercycle Research Institute
Abstract
National Institute for Public Health and the Environment and KWR Water Cycle Research Institute
have developed a new method for assessing the risks caused by the permeation of contaminants
through drinking water pipes. The core of this method is a step-by-step approach that should lead
to sound decisions on adequate measures to be taken. The approach incorporates the use of risk-
based trigger values for groundwater, a state-of-the-art model for the calculation of permeation
and, if there is any doubt, the execution of verification measurements in drinking water.
It is suggested that this method replaces the current methods used by the competent authority for
soil and groundwater quality and by the drinking water companies.
The current approach used to calculate the permeation is based on research and interpretations
from the period 1985-1995. The calculation of the permeation is part of the model used for the
derivation of the so-called Intervention Values and for the assessment of the urgency of soil and
groundwater remediation. The water companies use signal values to assess the permeation risk.
These signal values are at a much lower level than the Intervention Values and these differences
cause problems regarding decisions about measures to be taken.
In this study, a new model is being developed for the assessment of the permeation. Subsequently,
the practicality and the related uncertainties will be evaluated. The new permeation model will be
used to derive risk limits for frequently occurring organic contaminants in groundwater. It is noted
that the resulting risk limits are undergoing evaluation and require further improvement before
implementation.
Based on the derived risk limits it is proposed that the Intervention Value for groundwater is used
as a trigger value for potential risks caused by permeation of contaminants through drinking water
pipes.
It is also proposed that the web-based risk assessment package Sanscrit, used by competent
authorities in the Netherlands to decide on measures concerning contaminated soil and
groundwater, be adapted to suit the new insights. The approach developed enables regulators,
water companies and advisors to take sound decisions concerning the risk of permeation.
Introduction
The supply of safe drinking water is governed by the Dutch Drinking Water Act. The quality of the
drinking water has to comply with the drinking water standards. If drinking water pipelines are in
contact with contaminants in the soil and groundwater, there is a risk that these contaminants will
permeate through the wall of the water pipe and end up in the drinking water. The greatest risk
concerns polyethylene (PE) connection pipes. In order to manage these risks, drinking water
companies test the quality of the soil and groundwater using signal values 1. Signal values for soil
and groundwater are included in the Water Work Sheets (KIWA, 2004). If signal values are
exceeded, it is assumed that there is a potential risk and the drinking water companies take
measures. Measures consist of control measurements or the replacement of connecting pipes.
The current signal values for soil and groundwater for polyethylene (PE) drinking water pipes have
been set at a low level for organic contaminants. However, verification measurements often
1
Signal values were published in 1989 in an interim inspection guideline and were implemented in the so-called Water Work Sheet. The
term signal values should not be confused with the term signaling values from the third Water Management Memorandum of 1989.
These are used to assess the water bed quality.
indicate that there is no permeation, even in contamination situations where organic contaminants
occur in concentrations that are from ten to a hundred times higher than the signal values.
The approach adopted to tackle soil and groundwater contamination is set out in the Soil Protection
Act and the Circular on Soil Remediation. The essence of the approach is that first of all the
seriousness of a case of groundwater contamination is established on the basis of whether the
Intervention Values for groundwater have been exceeded. The Intervention Value acts as a trigger
value. If these values have been exceeded, a more in-depth investigation is carried out, in order to
assess the actual health risk at the location. For the assessment of the health risk the total
exposure to contaminants is tested against the Maximum Permissible Risk Level (MTRhuman). One of
the seven exposure routes that is assessed is exposure as a result of the consumption of drinking
water following permeation of contaminants through the water pipe. The total exposure is
calculated using the CSOIL exposure model. The exposure resulting from permeation of organic
contaminations through polyethylene drinking water pipes is calculated on the basis of
experimental research from the 1980s.
Because of the difference between the approach based on the Drinking Water Act and that based
on the Soil Protection Act it can occur that drinking water companies replace existing drinking
water pipes because of concerns about the drinking water quality, even though there is no reason
for measures to be implemented on the basis of the Soil Protection Act. This results in additional
costs and it sends out a contradictory message about health risks, drinking water quality and soil
quality to the parties and residents involved.
The purpose of this study is to improve the current methods relating to the assessment of
permeation risks and, on the basis thereof, to make a proposal for an assessment framework which
is suitable for both policy domains (drinking water and soil contamination).
The drinking water companies assess the permeation risk on the basis of the signal values that are
included in Water Work Sheet 2.2 C (KIWA, 2004). Separate values have been derived for PE and
PVC pipes. Permeation through PE is greater than through PVC. As far as the establishment of the
signal values is concerned, no distinction was made with regard to the dimensions of the drinking
water pipes, neither for type of PE. It is likely that the method devised by Vonk (1985) was used as
the basis for the establishment of the signal values, combined with expert judgement.
Table 1 gives the various standards and signal values for several common contaminants. If the
signal values are exceeded, the water company carries out a more in-depth investigation or the PE
pipes are replaced by impermeable pipes.
Table 1: Drinking water standard and signal values for groundwater for PE drinking water pipes
(KIWA, 2004), Target and Intervention Values for groundwater (Ministry of Infrastructure and the
Environment, Circular on Soil Remediation 2013). Concentrations in g/L.
Name of contaminant Drinking water Signal value for Target Value Intervention
standard possible risk of groundwater Value
permeation groundwater
Benzene 1 10 0,2 30
Ethyl benzene 1 - 4 150
Toluene 1 10 7 1000
Naphthalene 0.1 0.5 0,01 70
Vinyl chloride 0.1 - 0,01 5
1,2-dichloroethane 3 - 7 400
1,1-dichloroethene 1 - 0.01 10
**
1,2-dichloroethene (cis) 1 - 0.01 20**
1,2-dichloroethene (trans) 1 - 0.01** 20**
1,1,1-trichloroethane 1 10 0.01 300
Trichloroethene (tri) 10 10 24 500
(sum of tri and
Tetrachloroethene (per) 25 0.01 40
per)
Pyridine 1 - 0.5 30
Tetrahydrophurane 1 - 0.5 300
MTBE (Methyl-tert-butyl 1 306/ 4600* - 9400 ***
ether)
ETBE (Ethyl tert-butyl 1 110/ 1600* - -
ether)
Dimethylphthalate 1 - - -
*
Suggested values for LDPE (PE40) based on a permissible drinking water concentration of 1 g/l
or 15 g/l respectively (Meerkerk, 2008)
**
The sum of cis and trans 1,2-dichloroethene
***
Concerns an indicative level for serious contamination.
- No standard available.
The Circular on Soil Remediation, part of the regulations that are included in the Soil Protection Act
(Ministry of Infrastructure and the Environment, 2013), includes Target and Intervention Values for
groundwater for over 100 contaminants. Intervention Values are used to assess the seriousness of
a case of soil or groundwater contamination. If these values have been exceeded, a decision should
be taken about the urgency of the remediation. Intervention Values indicate the threshold above
which unacceptable human and/or ecotoxicological risks cannot be categorically ruled out. Target
and Intervention Values for groundwater are applied generically and have no relation to the local
function or the local use of soil and groundwater. If a soil and groundwater contaminant is lower
than the Intervention Value, it is assumed that unacceptable health risks are absent. Exposure
resulting from the consumption of drinking water following permeation through the water pipe is
one of the ten exposure routes that is taken into consideration.
The life-long average total exposure is tested against the Maximum Permissible Risk for people
(MTRhuman). Baars (2001) defines the Maximum Permissible Riskhuman) as follows:
The MPRhuman is defined as the amount of a substance (usually a chemical substance) that any
human individual can be exposed to daily during a full lifetime without any significant health risk.
For a detailed description of the human-toxicological risk assessment for soil contaminants refer to
Swartjes and Cornelis (2011) and Brand (2007).
The CSOIL exposure model calculates, via three routes, the exposure to contaminants that enter
the drinking water pipes via permeation: the consumption of drinking water, the inhalation of
vapours during showering, and skin contact with water during showering.
The method for determining the permeation was applied for the first version of CSOIL (Van den
Berg, 1995 and Van den Berg, 1997). Underlying the method and the quantification of permeation
coefficients is an interpretation of Vonks (1985) research. This research also forms the basis of the
current CSOIL (Brand et al., 2007). The parameters and permeation constants were reported in
Otte et al. (2001).
A difference with the derivation of signal values is that the human health risk is assessed on the
basis of an average drinking water concentration during a 24-hour period and a comparison of the
exposure with the Maximum Permissible Risk (Brand, 2007). The signal values (Ministry of VROM,
1989 and KIWA, 2004) were based on an assessment of the calculated maximum concentration in
drinking water after 8 hours stagnation and compared with the drinking water standards.
For the human health-related risk limits that underlie the Intervention Values, the exposure route
via the consumption of drinking water is small compared with that of other exposure routes. For 15
of the 134 contaminants the calculated exposure via drinking water is greater than 5% of the total
exposure: these are cresols, several chlorophenols and phthalates, d-HCH, carbaryl, pyridine,
tetrahydrophurane and tetrahydrothiophene (Lijzen et al., 2001).
A practical step-by-step plan was drawn up for the assessment of permeation risks (see figure 1).
The step-by-step plan is a combination of testing against a trigger value in groundwater, modelling
of the extent of permeation and, if there is any doubt, verification measurements in drinking water.
The aim is to use the step-by-step plan and the proposed test criteria therein to achieve cohesion
between the assessment related to concern for the quality of soil and groundwater (Soil Protection
Act) and that related to concern for the quality of the drinking water (Drinking Water Act).
As PE connecting pipes are the most sensitive to permeation, the step-by-step plan starts with the
question as to whether there are PE drinking water pipes present (1) and whether these are in
contact with contaminants (2). Then an initial test is carried out on the basis of a trigger value (3)
and the charting of any taste or smell complaints. Then a model calculation can be carried out to
estimate the permeation. If necessary, verification measurements in drinking water are carried out
(5), then a decision about the measures is taken.
Provisionally, the Intervention Values for groundwater are proposed (in step 3) as trigger values.
Permeation calculations for 17 selected contaminants and verification measurements, carried out
on the drinking water in contaminant situations, show that at the level of the Intervention Value it
is unlikely that the drinking water exceeds the drinking water standard. The charting of smell and
taste complaints may be a useful addition to the assessment. The assessor is advised, if there is
any doubt, to contact the drinking water company to check whether there have been any
complaints. The verification measurement (step 5) is preferably carried out after a period (eight
hours) of not using tap water (stagnation period). If this shows a concentration in between the
detection limit and the drinking water standard, a permeation risk cannot be ruled out and the
advice is to carry out a repeat measurement. If the assessment results in the conclusion that
action needs to be taken (6), a choice can be made between soil and groundwater remediation and
the replacement of the drinking water pipes.
The step-by-step plan will enable risks resulting from permeation to be identified in good time and
measures to be taken to guarantee the drinking water quality. This approach also means that
unnecessary measurements, and additional costs, are avoided as much as possible.
Figure 1: Step-by-step plan for assessment of the permeation risk caused by organic contaminants
in PE drinking water pipes.
Although drinking water pipes are impermeable, organic contaminants present in the groundwater
can permeate through the free spaces between the polymer chains of the drinking water pipes and
then enter the drinking water in the drinking water pipes. These free spaces in the pipeline material
are of a molecular size and are therefore not visible to the naked eye. If the pipe material is a
polymeric plastic, this process is called permeation (Comyn, 1985).
There have been various known cases of contaminants from the soil and groundwater being found
in drinking water pipes. Goodfellow et al. (2002), for example, refers to eleven cases of
contaminated drinking water in drinking water pipes in the UK in the period 1998 - 1999. In the
USA over 100 incidents were reported in the period up to 1992 (Glaza and Park, 1992). Bromhead
(1997) provides an overview of incidents resulting from permeation in drinking water pipes in
Belgium, Germany, The Netherlands, Norway, the UK and the USA.
Large-scale permeation incidents were reported in The Netherlands in the 1980s, including in
Rotterdam (Van der Heijden, 1985). Currently permeation occurs in isolated cases. For example,
there were three cases in the province of North Holland of contaminants being found in the drinking
water (personal report by PWN, 2013) and there were six cases in Brabant (personal report by
Brabant Water, 2014).
The assessment of the permeation is an important part of the step-by-step plan. The level of the
trigger value (3) is determined in part on the basis of a permeation calculation. In addition, a
permeation calculation is important for location-specific assessment using the Sanscrit assessment
package (4). Sanscrit (www.sanscrit.nl) is a system that supports decision-making for remediation
operations in cases of serious soil contamination. The CSOIL exposure model (Brand, 2007) for the
assessment of human risks is included in Sanscrit.
Permeation of contaminants from soil and groundwater through PE drinking water pipes includes
partition and diffusion processes (see figure 2). This figure shows the following processes:
1. flow of contaminants to the drinking water pipe via diffusion;
2. partition over the solid soil phase, the pore water and the gas soil phase;
3. partition from the pore water, and/or the soil air in the polymer;
4. diffusion through the polymer;
5. partition from the polymer to the drinking water in the drinking water pipe;
6. diffusion in the drinking water;
7. travel from the drinking water to the outlet.
Figure 2: Diagram of a PE drinking water pipe and the processes that are important for the extent
of permeation. The following are identified: the partition of the contaminants over the groundwater
and the PE pipe, the diffusion of the contaminants through the PE wall and then the partition of the
contaminants between the PE pipe and the drinking water.
The quality of the drinking water has to meet the drinking water standards of the Drinking Water
Decree (Government Gazette, 2011a). The drinking water standard is thus the statutory
assessment criterion for the assessment of permeation risks.
In periods during which no water is used, the concentration of contaminants may gradually
increase as a result of permeation, until water is used and the concentration instantly drops to
zero. This is followed by a new period of charging of the drinking water. At the end of the
relatively long stagnation period during the night the concentration will reach its maximum (peak
concentration). The assessment of the drinking water quality and the calculation of risk threshold
values can be carried out in various ways. Two options are:
Test the average concentration in the drinking water over 24 hours.
Tests the peak concentration in the drinking water after 8 hours stagnation in the drinking
water pipe.
Testing of a calculated peak concentration is a cautious test in which mixing in the pipes (for
example as a result of dilution) is not taken into account and the drinking water meets the
standard everywhere at all times. This is a more cautious test than a test of the average
concentration and therefore results in stricter (lower) risk threshold values in groundwater. This
starting point underlies the signal values currently in force. The average concentration (over 24
hours) is a guide to the estimation of health risks in the case of life-long exposure and has been
used for over 20 years in risk models such as CSOIL and Sanscrit.
The legislation does not provide any clarity on assessment procedures. The prescribed sampling
method for quality control of the drinking water is based on a sample that is used to measure an
average concentration. The drinking water standards are in general derived from a Negligible Risk
Level (Maximum Permissible Risk/100); average concentrations are possibly the closest to these.
However, this study takes both options into account.
Modelling of permeation
Modelling permeation can be carried out using the model included in the current CSOIL (Brand,
2007). This model is based on permeation coefficients that have been determined empirically and
on the basis of expert judgement (expressed in m2.day-1). The concentration in the drinking water
is calculated on the basis of the groundwater concentration and the dimensions of the drinking
water pipe, the period of charging of the drinking water during stagnation and the average
drinking water consumption during a 24-hour period (refer to Brand, 2007).
As part of this study, research is being carried out into whether an improved model can be drawn
up to calculate permeation. This new model is based on the same diffusion experiments as the
approach of Vonk / Van den Berg (Van den Berg, 1995). In this model, the permeation from
groundwater to drinking water via the pipe wall of the PE water pipe is dependent on the
concentration gradient and the diffusion coefficient. The determination of this permeation
coefficient (P) is central to this study. Vonk / Van den Berg ascertain the permeation coefficient on
the basis of the interpretation (expert judgement) of empirical data. The permeation concept
consists of two physico-chemical processes (diffusion and partition).
For the calculation of the diffusion coefficient two different approaches will be evaluated: First there
is the approach developed for the determination of the diffusion coefficient in plastic packaging
materials used for food products (Piringer, 2000) and the parameterisation of Brandsch et al.
(2000) (called the Piringer/Brandsch model in this study).
Secondly a regression analysis of experimental determined diffusion coefficients reported in other
studies will be performed. The analysis is aimed at predicting the coefficient based on molecular
characteristics such as mass and the octanol-water partition coefficient (Kow).
The model approach according Piringers equation is as follows:
where:
Dp : diffusion coefficient [m2.s-1]
Mr : molecular mass of a contaminant [D]
T : temperature [K]
AP : polymer-specific diffusion parameter [m2/s]
The factor 0.135 in the equation is regarded as a good estimate for most hydrocarbons and other
contaminants with a low polarity (Piringer, 2000a).
AP = AP /T (Eq. 2)
where:
AP : the upper limit of the polymer-specific diffusion parameter [-]
: a polymer-specific activation energy [-]
According to this approach the diffusion coefficient (Dp) depends on the molecular mass of the
contaminant and the properties of the polymer. The molecular mass is representative of the size of
the molecule and therefore of the resistance that affects the molecule during diffusion. The
properties of the polymer are included by the so-called polymer-specific diffusion parameter, which
depends on the polymer-specific activation energy. This can be plastically presented as the power
that is required to force the polymer chains apart in order to make diffusion of a molecule possible.
The determination of the partition coefficient PE/groundwater is based on the analogy with the
octanol-water partition coefficient (Kow). After all, a contaminant will, depending on the
hydrophobia or lipophilia of the contaminant, spread in a similar manner between a polar medium
(water) and an apolar medium (PE material or octanol). So in this study, an empirical relationship
has been derived between the Kw PE/water and the Kow on the basis of measurements and
regression analysis.
The permeation coefficient can be calculated as follows:
P = Dp * Kw (Eq. 3)
where:
P : permeation coefficient [m2.s-1]
Dp : diffusion coefficient for PE [m2.s-1]
Kw : partition coefficient PE/groundwater [-]
For this study the permissible concentration for a group of seventeen organic contaminants in
groundwater was calculated that corresponds to a concentration that is regarded as permissible in
drinking water (risk threshold value). The risk threshold value is the concentration in the
groundwater at which, following permeation through a PE pipe, the drinking water concentration is
equal to the drinking water standard.
The permissible concentration in groundwater, CgwT [g.L-1], can be calculated as follows:
where
CdwT : permissible concentration in drinking water [g/L]
d : wall thickness of water pipe [m]
di : diameter of water pipe [m]
Dp : diffusion coefficient [m2/s]
Kw : partition coefficient of PE/water in accordance with OW model [-]
t : duration of stagnant drinking water [s]
The risk threshold values resulting from the model calculations were compared with the
Intervention Value for groundwater and the current signal values. This shows that, both on the
basis of the Vonk / Van den Berg approach and on the basis of the model that was developed as
part of this study, the calculated permissible concentration in groundwater ( CgwT) is higher than
the Intervention Value for groundwater in an assessment against the drinking water standard and
on the basis of average concentrations in the drinking water. Even according to the strictest
scenario (peak concentration), for most contaminants a permissible concentration in groundwater
(CgwT) is calculated that is higher than the Intervention Value.
Moreover, practical data relating to groundwater contaminants and measurements in the drinking
water were used to investigate whether the Intervention Value is a safe trigger value for potential
risks caused by permeation. These measurement data were issued by the drinking water company
PWN, which carried out a study to identify the permeation risk. A list was drawn up of 50 locations
that are most at risk from permeation. At these locations the (maximum or average)
concentrations measured in groundwater were significantly higher than the signal values and
Intervention Values. The drinking water at these locations was sampled after 8 hours stagnation in
the water pipe. The residents were asked not to use any water that night.
On the basis of this data set it can be concluded that there are permeation risks only in the case of
very high concentrations in the groundwater, much higher than the Intervention Values.
A comparison of the series of measurements and initial model calculations shows that both
permeation models greatly overestimate the actual (measured) permeation (at least by a factor of
ten).
The PWN data set is unfortunately small as regards both the range of investigated contaminants
and the total number of measurements. The data set is too small to make proposals for signal
values on the basis of these measurement data, but it does provide an indication of the size of the
risk threshold values.
It can be concluded that the Intervention Value is a solid and practical trigger for identifying a
permeation risk. Model calculations are valuable for excluding permeation risks. It is likely that the
model overestimates the actual risk for most situations. An additional safety net for situations
involving PE pipes is provided by the assessment in accordance with the step-by-step plan, in that
it takes into account taste or smell complaints and the possibility of verification measurements.
It is concluded that assessment of permeation risks can take place on the basis of the step-by-step
plan. In this step-by-step plan the Intervention Value should be used as a trigger value for either
excluding impermissible risks or further investigation..
A new permeation model has been developed. The uncertainty of the model is estimated to be
relatively large. A comparison of the model results with field measurements shows that this model
will overestimate the actual permeation in most cases. This is in accordance with the valid risk
concept, whereby in the case of great uncertainty a cautious approach is followed and fairly strict
(low) risk threshold values are thus calculated. For the assessment of the risk, other factors are
also taken into account in accordance with the step-by-step plan.
The following steps are recommended with a view to improve the permeation model:
The extent to which dissolved contaminants in groundwater are available for permeation should
be investigated. This is connected with the adopted access and analysis method for
groundwater samples, as is known from the literature relating to bioavailability.
Permeation experiments should be carried out for those contaminants (often contaminants with
a high lipophilicity) for which there is no reliable estimation of the permeation coefficients on
the basis of the literature. This is difficult from an experimental point of view and, partly for
this reason, published research gives different results. Maybe passive samples in and outside
drinking water pipes could provide an answer.
Research should be carried out into the possibility of deriving a kinetic model, including
parameterisation.
References
- Brand E., P.F. Otte, j.P.A. Lijzen, 2007. CSOIL 2000 an exposure model for human risk
assessment of soil contamination. A model description. Report 711701054. Bilthoven, The
Netherlands
- Comyn, J., 1985. Introduction to polymer permeability and mathematics of diffusion. In: J.
Comyn (Ed.). Polymer permeability, Chapman and Hall, London.
- Glaza E.C., J.K. Park. 1992. Permeation of organic contaminants through gasketed pipe
joints. Journal AWWA, 84(7): 92-100.
- KIWA Nederland B.V., 2004. Waterwerkblad WB 2.2 C Leidingmaterialen; Buizen van PE,
http://www.infodwi.nl/uploadedFiles/Website/VEWIN/WB%202.2%20C.pdf (Mai 2014).
- Ministry of VROM, 1989. Memo Dealing with risks, appendix no. 5 of the National Policy
Plan, Ministry of Housing, Spatial Planning and the Environment (in Dutch). Tweede Kamer
der Staten-Generaal, vergaderjaar 1988-1989, stuk 21.137 nr. 5; SDU uitgeverij, Den
Haag.
- Otte, P.F., J.P.A. Lijzen, J.G. Otte, F.A. Swartjes, C.W. Versluijs, 2001.
- Evaluation and revision of the CSOIL parameter set. RIVM report 711701021/2001, RIVM,
Bilthoven, the Netherlands.
- Swartjes F.A., C. Cornelis, 2011. Human health risk assessment. Chapter 5. In: Swartjes
FA, editor. Dealing with contaminated sites. From theory towards practical application-
Dordrecht: Springer publishers; 2011.
- Van den Berg, R., 1995. Human exposure to soil contamination. A quantitative and
qualitative analysis resulting in proposals for human-toxicological C-quality standards (in
Dutch). RIVM report 725201011, RIVM, Bilthoven, the Netherlands.
- Van der Berg, R., 1997. Account of data and procedures for the first series of Intervention
Values: from RIVM reports to the Memo Intervention Values soil remediation (in Dutch).
RIVM Rapport 715810012, Bilthoven, the Netherlands.
- Van der Heijden, B.G., 1985. Some experience with the permeation of organic compounds
through plastic water pipes (in Dutch). H20 18, nr 5: 88-95.
- Vonk, M.W., 1985. Permeation van permeation of organic compounds through pipe
materials, Mededeling no. 85, Nieuwegein (KIWA N.V.)
AN APPROACH TO RISK ASSESSMENT AND MANAGEMENT OF CONTAMINATED LAND
IN ASIA
Disclaimer
The aerial pictures used throughout this article are aerial pictures of an industrial area somewhere in Asia. Throughout this
article we will use these pictures of the same location to illustrate the approach which was developed. The pictures used are not
from the actual site, but will serve to set out the methodology used and illustrate the results obtained. By using these pictures
there is no implication meant whatsoever as to the activities, land use and quality of the land, subsurface soils and groundwater
on the location. The aerial pictures were available through Google Earth.
Introduction
The area under investigation is about 2 km by 2 km and is situated in one of the larger cities in Asia.
The area underwent an important expanding phase and development in the 50 and 60 of the
previous century. A report by an international consultancy was made available, compiling the results of
3 sampling campaigns over a period of about 3 weeks. SITA Remediation was requested to review the
report and provide (1) an estimation of the overall remediation cost and (2) an average remediation
cost per m2.
The report contained information in relation to 38 boreholes and 16 groundwater wells. The report
provided field observations : borelogs, PID measurements during drilling, water level registrations, field
measurements like pH and EC of groundwater samples, .
Illustration the spread of the sampling points on the industrial area under redevelopment (yellow line 2 km).
In order to get to know the sites up close and meet the client, a site visit was organized. Although the
site visit did not reveal any additional data, the site visit provided a feel of the surroundings and the
conceptual build-up to the industrial plants in the area. This became also clear that not the full area
was under investigation and up for redevelopment.
The size of the individual site would not work as a measure or evaluation criterion to determine the
environment impacts on the site. Some sites were large sites already redeveloped, some of the large
sites had only been green land in the past. The complete opposite case was a small site that had
clearly a CS2-plume in the groundwater.
The Challenge
The limited but analytically good site investigation report posed a double challenge. The sampling
points were not close enough to the actual activities and too limited to make any practical assessment.
At the same time going through the analytical reports some low concentrations were registered and
some regional impact appear in the groundwater analytical results.
What was clearly missing was the footprint of the contamination. With some of the sites being
completely demolished and derelict, no clear information was available to provide quantitative
information as a measure for comparison of any scale of contamination.
By an arbitrary check on Google Earth, it became clear that a lot of the sites had been redeveloped
recently and often several times over a period of about 12 years (2002-2012). The aerial pictures
available provided a massive set of information, at intervals of sometimes 3 months, with relevant
changes on the sites.
In order to compile the information, a CAD file was built and all information was registered in different
layers by time- and geo-referencing.
Although the data set now looked promising and provided an insight in what was happening more
clearly, the challenge was still to make sense of all the available information.
If something would be provided to the client as an output of the assessment, it should have the
following criteria :
The solution to processing all the data into a system lies in using the Source-Path-Receptor-concept.
A multi-criteria matrix can be built by scoring Source/Path/Receptor criteria. The summation of the
scores in the matrix provides a measure to assess environmental impacts in the subsurface. Not only
a score for the impacts on the sites are obtained, but also the impact to the direct surrounding
environment.
Source Score
The source score would have two components or be a composite of two scores. First there would
need to be a measure for the scale of the possible impact. Second the signature of the contamination
in terms of human toxicology and exposure in the current and possible future land use would need to
be evaluated. This last measure would be strictly limited to the situation on the site.
Path Score
The path score represents the possible distribution and migration of impacts beyond the site boundary.
Possible ways of migration can be dust that travels to adjacent sites or groundwater that flows in the
direction of next door sites.
Receptor Score
The receptor score is a translation of the sensitivity of the surrounding area in terms of land use. This
means that residence time, type of activity, sensitivity of the subjects, of the surrounding sites are
taken into account.
Scoring each separate part of the multi-criteria matrix
Source Score.
Scale of operations
In order to determine the scale of the operations and activities on site, an analysis was made of the
sets of aerial pictures available. Together with the observations during the site visits, a core or footprint
of the activities could be established.
By measuring the surface area of several relevant sections of the operations, an overall activities &
2
operations footprint was calculated (m ).
Impact to Health
For the impact to health score the measured analytical chemical compounds in either soil or
groundwater were listed independent of their concentration and special (distance) relation to any
potential source on the site. The human health score would then be a measure of the carcinogenic
characteristics of the chemicals measured and also would take into account compound accumulated
exposure if several chemicals are involved. In the table below additional information is stated like
Activity on the site & Guiding parameters, this information is stated to allow appreciation or re-
evaluation of the score if more information was gathered over time.
Example of source scoring in the multi-criteria matrix
The source score is represented by two scores surface and impact to health, deliberately the sum of
both scores is not given in the table. The sum of the two scores is compounded in the final score, but
the above representation allows the reader to see the weight and contribution given by each score on a
scale of 5.
Path score.
Receptor score
The table below illustrates the score card results. Sites with a higher score are sites that are under
suspicion of having higher impacts, with more impact to the surrounding environment.
Overall
Source Path Receptor
Score
score score
SITE 1 3 5 4 4 16
SITE 2 3 5 4 4 16
SITE 3 4 4 2 1 11
The score card presentation on a map
In line with the score card results, a legend coloration was established to provide fast interpretation on
a map. But not only a color is given, also the individual scores and the overall score is given in an
informational window. This again facilitates the review of maps by checking the score and allowing the
map to become a life document, that can change over time. The use of both the score matrix and the
map allows consultation of relevant data prior to consulting the investigative detailed information.
The overall coloration gives a ranking or priority of the site and its environmental presumed possible
impact, including the possible impact to the surrounding environment.
Discussion and conclusion :
Taking into account the initial very limited data set, it seemed impossible to provide an assessment of
2
initial remediation costs for an industrial area of 4 km .
By looking at all the data in the investigative report and acquiring additional information from aerial
pictures, a large data set was obtained. The full integration of this dataset in the concept of the
Source-Path- Receptor-linkages by providing a score, has allowed to provide a ranking of the sites.
The ranking can be plotted on a map to visually represent the sites more under suspicion of
contamination than others. Additionally, the impact towards the surroundings are taken into account.
Indirectly the score would represent a presumed remediation cost ranking, whereby the surface area
of the footprint analysis can be a measure for the impacted subsurface.
The methodology presented in this article actually provides a tool which can be used by urban
planners when redeveloping larger industrial areas. It can be a ranking for site investigation priority or
can be a tool for the follow-up of the site investigation results themselves. Additionally, more add-on
information can be integrated to provide for more scores and information, which can be used as a tool
to measure and represent the actual status of an on-going redevelopment process.
Probabilisticassessmentfor sixvapourintrusionalgorithms
JeroenProvoost1,JanBronders2,IlseVanKeer2
1
IndependentResearcher,Finland,ResearchGate:http://www.researchgate.net/profile/Jeroen_Provoost2/
2
FlemishInstituteforTechnologicalResearch(VITO),Boeretang200,2400Mol,Belgium
1 Introduction
Manycountrieshavedevelopedcontaminatedlandmanagement(CLM)policiestoreduceriskstohumansand
ecosystems originating from soil pollution (Provoost et al. 2008a). These policies often make use of model
algorithmstocalculateasoilscreeningvalues(SSV)totriggerfurtheractions,suchashighertierssitespecific
human health risk assessment, or remediation actions to reduce the risks (Provoost et al. 2013). One of the
major pathways of exposure for humans is inhalation of indoor air as a result of subsurface contamination
with volatile organic chemicals (VOC), called vapour intrusion (VI) (DOH 2006). VI is a process not yet well
understood. Swartjes (2007) looked into the variation of exposure by applying seven EU models for human
health risk assessment, and concluded that the variation for exposure due to indoor air inhalation is large.
Provoostetal.(2008a)investigatedthereasonsforthevariationinSSVbetweendifferent(EU)countriesand
theeffectofharmonisingalgorithmsandparametersontheSSV.TheconclusionwasthatSSVforVOCvary1to
4ordersofmagnitude(OoM)between(EU)countries.Thisisduetotheapplicationofdifferentalgorithmand
itsparametervalues(scientificelements),whilepoliticalelements,liketoxicologicalcriteria,haveamoderate
to high influence on the variation in SSV. Although geographical elements differ throughout Europe, they
seemed to have a limited impact on the variation. The Common Forum recently released a note, named
TowardsanEuropeanresearchagendaforcontaminatedlandmanagement,whichidentifiesgapsandneeds
forthedifferentstepsofCLM.Thenotestates,forthesiteinvestigationstep,thatphasepartitioninginthesoil
and the resulting VI in buildings needs further investigation. Furthermore it recognizes, for the step risk
assessment,thataonesizefitallmodelalgorithmisnotrealistic.Adeeperunderstandingisrequiredrelated
tothecontentandthelimitsofthemodelsandpredictedvalues(CommonForum2013).
Ifalgorithmsareappliedpredictingthesoilairandindoorairconcentrationasaresultofsoilpollution,andthe
related human exposure, is complex and is affected by numerous factors. These factors are translated into
screening algorithms parameters. Parameters can be divided generally in three categories: environmental,
buildingandphysicochemical(McAlaryetal.2011,Provoostetal.2010),witheachcategorycharacterisedby
adegreeofuncertaintyorvariability.Thealgorithmsoutputisthereforesubjecttotwosourcesofparameter
variation: uncertainty and variability. Variability regards variation that can be naturally expected, while
uncertaintyregardsprecisionbywhichaquantityismeasured(VanBelle2008).Mostofthepresentalgorithms
forVIcalculate(deterministic)pointestimatesbasedonasetofdefaultparametervaluesandthereforegive
noindicationofthevariationandconservatismofthepredictedairconcentrations.
2 Methodology
Measureddatafromtwowelldocumentedsiteswereusedtocomparedpredictionsandobservations,andare
describedinmoredetailinProvoostetal.(2014).Thefirstsite,AstralinVilvoorde(Belgium),iscontaminated
with the aromatic hydrocarbons benzene (BE) and ethylbenzene (EB), while the second site, CDOT site in
Denver(USA),containsamongotherchlorinatedcontaminantsmainlytrichloroethene(TCE).Forbothsitesthe
majorityofthecontaminantsourceforVIwaslocatedinthevadosezone.
The algorithms selected were the Swedish dilution factor algorithm from 1996 (SEPA 1996), dilution factor
algorithm from Norway (SFT 1999), Johnson and Ettinger model (United States) (Johnson et al. 1999), CSoil
(Netherlands)(Brandetal.2007),VolaSoil(Netherlands)(Waitzetal.1996)andVlierHumaan(regionFlanders
inBelgium)(OVAM,2004).
Foreachalgorithmaprobabilisticassessmentwithsensitivityanalysiswasapplied,aswellasthecalculationof
thedeterministicpredictedconcentrations.Probabilitydistributionfunctions(PDF)werederivedfromthedata
gathered for three groups of parameters (soil, building and physicochemical properties for the 3
contaminants) and fed into the algorithms. Data were obtained from site measurements and literature. The
PDF in algorithms input were propagated into an output distribution of the predicted (soil and indoor) air
concentration (probabilistic analysis), and allows an analysis of the contribution of each parameter to the
2/7
variation(sensitivityanalysis).Theresultsaredisplayedasboxandwhiskerplotsdisplayingthedistributionof
air concentrations, and sensitivity stacked bar charts (ranking of parameters according to the correlation
between parameter and algorithm output). Algorithm parameters were adapted to site specific conditions
whereneeded.
Thepredictedconcentrationswereevaluatedagainstobservedairconcentrationstodeterminetheaccuracyof
each algorithm, while the level of conservatism was obtained by comparing the 95 percentile predicted
concentrationfromtheprobabilisticrangewiththedeterministicpredictedconcentration.Theanalysisresults
in a ranking of algorithms for accuracy and conservatism in predicting soil air and indoor air concentrations,
andthusdeterminestheirsuitabilityforregulatorypurposes.
TheaccuracyandconservatismofscreeninglevelalgorithmsisobjectifiedbycalculatingtheMaximumrelative
Error(ME),RootMeanSquaredError(RMSE)andCoefficientofResidualMass(CRM),asdescribedbyLoague&
Green(1991),forthepairedpredictedandobservedairconcentrations.ThesecriteriawereappliedinProvoost
etal.(2008b,2009,2014),andalsointhisstudy,forinteralgorithmcomparison,andprovidedarankingofthe
algorithmsastotheiraccuracy.
The boxandwhiskers plots include the tolerable concentration in air as a reference to inhalation risks of
indoorair.TolerableconcentrationswereobtainedtheWorldHealthOrganization(WHO1996,2010).
Tobeusefulforaregulatorypurposescreeninglevelalgorithmsshouldbesufficientlyconservativeandmay
resultinfewfalsenegativepredictions.However,theyshouldnotbetooconservativebecausethentheymight
haveinsufficientdiscriminatorypower(Provoostetal.2009).
3 Results
The boxandwhiskers plot (Figure 1ab) display the minimum, 25 percentile, median, 75 percentile and
maximumpredictedsoilorindoorairconcentration,aswellasthepredicteddeterministicconcentration()
and, in the case of indoor air concentrations, the tolerable concentration in air () for the pertaining
contaminant.Alsoforeachofthecontaminantstheobserved(measured)soilairandindoorairconcentrations
are displayed to contrast with the predictions. The boxandwhiskers plots provide an insight in the spread
(range of the values from the highest to the lowest value), and the midspread (range of middle 50% of the
values)oralsocalledinterquartilerange.Thelocationofthemedianlinerelativetothe25and75quartiles
indicatestheamountofskewnessorasymmetryinthedata.
DF SE: Dilution Factor algorithm from Sweden, DF NO: Dilution Factor algorithm from Norway, JEM: Johnson and Ettinger model, Vl-H:
Vlier-Humaan, Obs: observed concentrations, Box plot: minimum, median or maximum concentration, box is 25 or 75 percentile
concentration, deterministic concentration
Figure 1a. Box-and-whiskers plot for predicted and observed soil air concentrations by algorithm and
contaminant
3/7
DF SE: Dilution Factor algorithm from Sweden, DF NO: Dilution Factor algorithm from Norway, JEM: Johnson and Ettinger model, Vl-H:
Vlier-Humaan, Obs: observed concentrations, Box plot: minimum, median or maximum concentration, box is 25 or 75 percentile
concentration, deterministic concentration, tolerable concentration in air
Figure 1b. Box-and-whiskers plot for predicted and observed indoor air concentrations by algorithm and
contaminant
The sensitivity analysis allows for the ranking of dominant parameters to the variation in predicted air
concentrations. Parameters were grouped by physicochemical, soil or building parameters (Table 1ab)
resultinginanoverallcontributionofthegrouptothetotalvariation(Figure2ab).
Table1a:Summaryofparameterscontributing>20%tothevariationinsoilair
DFSE1996
VolaSoil
DFNO
Group
Type
CSoil
VlH
JEM
Parameter
Octanolwaterpartitioncoefficient PC Var X X
Henryscoefficient PC Var X
Solubility PC Var X X X
Organiccarbonwaterpartitioningcoefficient PC Var X X
Initialconcentration S Var X X X X X X
Organiccarbonfractionsoil S Unc X X X X X
Var:variability,Unc:uncertainty,PC:physicochemicalproperty,S:soilproperty,DFSE1996:dilutionfactoralgorithmfromSweden,DF
NO:dilutionfactoralgorithmfromNorway,JEM:JohnsonandEttingermodel,CSoil:CSoilalgorithm,VolaSoil:VolaSoilalgorithm,VlH:
VlierHumaanalgorithm
4/7
Table1b:Summaryofparameterscontributing>20%tothevariationinindoorair
DFSE1996
VolaSoil
DFNO
Group
Type
CSoil
VlH
JEM
Parameter
Octanolwaterpartitioncoefficient Var PC X X
Henryscoefficient Var PC X
Solubility Var PC
Organiccarbonwaterpartitioningcoefficient Var PC X X
Initialconcentration Var S X X X X
Waterfilledporosity Unc S X
Airfilledporosity Unc S X
Airpermeability Unc S X
Organiccarbonfractionsoil Unc S X X X X
Soilbuildingpressuredifferential Unc S X
Fractionairinconcrete Unc B X
Var:variability,Unc:uncertainty,PC:physicochemicalproperty,S:soilproperty,B:buildingproperty,DFSE1996:dilutionfactoralgorithm
fromSweden,DFNO:dilutionfactoralgorithmfromNorway,JEM:JohnsonandEttingermodel,CSoil:CSoilalgorithm,VolaSoil:VolaSoil
algorithm,VlH:VlierHumaanalgorithm
DF SE: Dilution Factor algorithm from Sweden, DF NO: Dilution Factor algorithm from Norway, JEM:
JohnsonandEttingermodel,VlH:VlierHumaan
Figure 2a: Stack bars of the percentage that physico-chemical and soil parameter
5/7
DF SE: Dilution Factor algorithm from Sweden, DF NO: Dilution Factor algorithm from Norway, JEM:
Johnson and Ettinger model, Vl-H: Vlier-Humaan
Figure 2b: Stack bars of the percentage that physico-chemical, soil and building
parameter values contribute to the variation in indoor air concentrations by algorithm
and contaminant
4 Discussion
4.1 Soil air
Someoftheboxandwhiskersplotsshowanegativeskewnesstowardsthehigherconcentrations.Withsome
exceptionthedeterministicallypredictedconcentrationsareinthemidspreadoftheboxplot(figure1a).
Thedeterministicpredictedsoilairconcentrationsareoverallhigherthanthemedianobservedconcentrations
with the exception of VolaSoil and CSoil for benzene. The algorithms that most frequently overpredict (less
accurate)theobservedsoilairconcentrationsaretheDFNO,VolaSoilandtheDFSEalgorithms,whiletheJEM,
CSoilandVlHhaveahigheraccuracy.
6/7
5 Conclusions
According to this study, the screeninglevel algorithms that have a higher degree of conservatism for their
defaultparametersetaretheJohnsonandEttingermodel(JEM),DilutionFactoralgorithmfromSweden(DFSE
1996),VlierHumaanandVolaSoil.Fromthese4algorithmstheJEMandVolaSoilhavearelativehighaccuracy
(discriminative power). For the latter two algorithms different parameters, that are variable and uncertain,
contributetothevariationinindoorairconcentration.Differencesbetweenparametersthatdrivethevariation
wereobservedbetweenthearomaticandchlorinatedhydrocarbons.
For TCE, the default parameter set of VlierHumaan, CSoil and DF SE 1996 should be adapted to arrive at a
higher deterministically predicted indoor air concentration when a more conservative approach is required.
ThedeterministicallypredictedairconcentrationsforBEandEBseemtobesufficientlyconservative.
It is shown that the probabilistic approach allows for an improved insight into the relative importance of
parameters in the risk estimates. A probabilistic approach should be applied to more sites to confirm the
findingsofthisstudy.
Some algorithms are more accurate, but might produce (frequently) false negative predictions. A balance
between accuracy and conservatism need to be sought and this balance determines the suitability for site
specificassessmentand/orderivingofSSV.
Giventheabovetwoalgorithmsforsitespecificriskassessmentcouldbeapplied.Someofthealgorithmsare
more accurate, depending on contaminant and source (groundwater/soil) and a toolbox with algorithms
could be developed. For the algorithms in the toolbox fixed parameters could be defined that can be
standardized (e.g. physicochemical parameters) and flexible parameters for which a range can be provided
(geographicalparametersrelatedtothesoilandbuilding).
ThispapercontainscontentfromProvoostetal.2014.Thefullpaperanddataset(parametervaluesandPDF)
canbedownloadedfromResearchGate.net.
7/7
6 Reference
Brand E, Otte PF, Lijzen JPA (2007). CSOIL 2000: an exposure model for human risk assessment of soil contamination, RIVM report
711701054,http://www.rivm.nl/bibliotheek/rapporten/711701054.html
Common Forum (2013). Towards an European research agenda for contaminated land management, Common Forum, note 2013.046,
http://www.commonforum.eu
DOH(2006).GuidanceforevaluatingsoilvaporintrusioninthestateofNewYork,NewYorkStateDepartmentofHealthDOH,Centerfor
Environmental Health, Bureau of Environmental Exposure Investigation, New York.
http://www.health.ny.gov/environmental/investigations/soil_gas/svi_guidance/docs/svi_main.pdf
JohnsonPC,KemblowskiMW,JohnsonRL(1999).Assessingthesignificanceofsubsurfacecontaminantvapormigrationtoenclosedspaces:
sitespecific alternatives to generic estimates, Journal of Soil Contamination, vol. 8, issue 3, pages 389421,
http://openagricola.nal.usda.gov/Record/IND22015155
LoagueKM,GreenRE(1991).Statisticalandgraphicalmethodsforevaluatingsolutetransportmodels:overviewandapplication,Journalof
ContaminatedHydrology,vol.7,pages5173,http://dx.doi.org/10.1016/01697722(91)900383
McAlary T, Provoost J, Dawson H, Swartjes F (editor) (2011). Book on Dealing with Contaminated Sites From Theory towards Practical
Application, chapter 10 Vapour intrusion, January 2011, pages 409453, Springer Publishers, Netherlands, ISBN 9789048197569,
http://dx.doi.org/10.1007/9789048197576_10
OVAM(2004).Basisinformatievoorrisicoevaluatie,Deel3H,FormulariumVlierHumaan,report,http://www.ovam.be
Provoost J, Reijnders L, Swartjes F, Bronders J, Carlon C, DAlessandro M, Cornelis C (2008a). Parameterscausing variation between soil
screening values and the effect of harmonization, Journal of Soils and Sediments, vol. 8, issue 5, pages 298311,
http://dx.doi.org/10.1007/s1136800800260
Provoost J, Reijnders L, Swartjes F, Bronders J, Seuntjens P, Lijzen J (2008b). Accuracy of seven vapour intrusion algorithms for VOC in
groundwater, Journal of Soils and Sediments Protection, Risk Assessment, and Remediation, vol. 9, number 1, pages 6273, DOI:
http://dx.doi.org/10.1007/s113680080036y
ProvoostJ,BosmanA,ReijndersL,BrondersJ,TouchantK,SwartjesF(2009).Vapourintrusionfromthevadosezonesevenalgorithms
compared,JournalofSoilsandSedimentsProtection,RiskAssessment,andRemediation,http://dx.doi.org/10.1007/s1136800901274
ProvoostJ,ReijndersL,BrondersJ(2013).Accuracyandconservatismofvapourintrusionalgorithmsforcontaminatedlandmanagement,
JournalofEnvironmentandPollution,vol.2,no.2,CanadianCenterofScienceandEducation,Toronto,Canada,ISSN19270909,EISSN
19270917,http://dx.doi.org/10.5539/ep.v2n2p71
ProvoostJ,ReijndersL,BrondersJ,VanKeerI,GovaertsS(2014).ProbabilisticRiskAssessmentforSixVapourIntrusionAlgorithms,Journal
ofEnvironmentandPollution,vol.3,No.2,ISSN19270909(print),ISSN19270917(online),CanadianCenterofScienceandEducation,
Toronto,Canada,http://dx.doi.org/10.5539/ep.v3n2p1
ProvoostJ,TillmanFD,WeaverJW,ReijndersL,BrondersJ,VanKeerI,SwartjesF(2010).BookonAdvancesinEnvironmentalResearch,
volume5,chapter2VapourIntrusionintoBuildingsALiteratureReview,DanielsJ.A.(Ed.),NovaSciencePublishersInc.,NewYork,ISBN
9781616687441,https://www.novapublishers.com/catalog/product_info.php?products_id=14005
SEPA (1996). Development of Generic guidelines values, Model and data used for generic guideline values for contaminated soils in
Sweden,SwedishEnvironmentalProtectionAgency,Report4639
SFT (1999). Guidelines for the risk assessment of contaminated sites, Norway, report 99:06,
http://www.miljodirektoratet.no/old/klif/publikasjoner/andre/1691/ta1691.pdf
SwartjesFA(2007).InsightintothevariationincalculatedhumanexposuretosoilcontaminantsusingsevendifferentEuropeanmodels,
IntegratedEnvironmentalAssessmentandManagement,vol.3,issue3,pages322332,http://dx.doi.org/10.1002/ieam.5630030303
Van Belle G (2008). Statistical rules of thumb, Wiley Series in Probability and Statistics Wiley, pp. 99100, ISBN13: 9780470144480,
http://dx.doi.org/10.1002/9780470377963
Waitz MFW, Freijer JI, Kreule P, Swartjes FA (1996). The VOLASOIL risk assessment model based on CSOIL for soils contaminated with
volatilecompounds,RIVMreport715810014,http://www.rivm.nl/bibliotheek/rapporten/715810014.pdf
WHO (1996). Environmental Health Criteria 186, Ethylbenzene, ISBN 9241571861, retrieved from
http://apps.who.int/bookorders/anglais/detart1.jsp?codlan=1&codcol=16&codcch=186
WHO (2010). WHO Guidelines for indoor air quality, ISBN 9789289002134, retrieved from
http://www.who.int/indoorair/publications/9789289002134/en/
Acomputationaltextanalysisofthevapourintrusioncorpus
JeroenProvoosta,KarenVictorb
a
Independentresearcher,Finland,jprovoostresearch@yahoo.co.uk
b
BirminghamCityUniversity,SchoolofEnglish,Birmingham,UnitedKingdom
Keywords
computational,text,analytics,sentimentanalysis,Knime,datamining,vapour,intrusion
Introduction
Humanscanbeexposedtochemicalsoriginatingfromcontaminatedlandandthereforetherelated
healthrisksareassessedtoreduceittoacceptablelevelsifneeded.Oneoftheexposureroutesfor
humans living on or near a contaminated site is vapour intrusion (VI). This is the process where
volatile organic contaminants migrate to the indoor air of a building so humans can inhale the
contaminatedair.Overthelast25yearanincreasingnumberofpublicationshavebecomeavailable
on the topic of VI. The purpose of this study was to provide a content analysis of relevant
publicationsanduncoveremergingareasofresearch.
Methods
The corpus was compiled by searching the Web of Science for articles containing the terms
vapour or vapour and intrusion in the title or abstract. In addition relevant reports were
collectedfromothersources.Atotalof348relevantpublicationswereselectedresultinginacorpus
of around 3 million words. The corpus was analysed by using the computational textmining tool
Leximancer. The tool first divided the corpus in 330.000 text blocks of each 2 sentences long and
identifiestermsthatfrequentlyoccurinthecorpus,andusedtheseasconceptseedwords.Thenext
stepbuildsathesaurusoftermsforeachconceptbasedonitscooccurrencewiththeseedwords.
Additionalconceptswerediscoveredduringthisanalysisresultinginatotalofaround300concepts.
Leximancerthencalculatedthecooccurrenceoftheconceptsandcreatesconceptmapsindicating
therelationshipandconnectionsbetweentheconcepts,aswellasgroupingofconceptsintological
themes.ThefreeopensourceKnimedataminingtoolwasusedinadditionforfurtheraggregation
andvisualisationoftherawdatafromLeximancer.Severalfiguresandgraphs,likeheatmaps,were
producestoinvestigatetherelationshipbetweenconceptsandpublications.Leximancerallowsfora
sentiment analysis of the identified concepts by using an inbuild thesaurus of sentiment terms
(negativeversuspositive).Inthisstudythesentimentanalysiswasusedasanadjunctanalysisonly,
indicatingtherelativefrequencyofsentimentassociationsforeachterm.
Results
Theresultsdemonstratethatthemajorthemesandassociatedconcepts(betweenbrackets)are:[1]
soil (air concentration), [2] model (parameters), [3] diffusion (coefficient), [4] site, [5] (volatile)
compounds,[6](ground)water,[7]risk(assessment),[8]house(construction),[9]uncertainty,[10]
monitoring and [11] (hazardous) waste. Emerging areas of research are [1] probabilistic risk
assessment of software model parameters with Monte Carlo analysis (uncertainty), [2] how
hazardouswasteisrelatedtothesiteandhowsampling,monitoringandtheremediationofground
waterisorganised,[3] (ad)sorptionandequilibrium phasedistributionin the waterphase andthe
needtoclarifythemasstransferandtransportofcontaminants,includingthediffusionthroughthe
boundaryfluxlayer.
Conclusions
More fundamentally, this study raises questions whether the way in which VI is modelled, for
exampletheuseoftheHenryconceptforcalculatingthesoilairconcentration,isrepresentingthe
reality. The sentiment analysis shows that a set of publications and their year of publication are
associated with negative terms, although the overall sentiment of the corpus is positive. The
sentimentanalysisshouldbetreatedcautiously,giventhecomplexityandshortcomingofcomputer
drivensentimentanalysis.
ThisstudycontributestotheinsightinthedirectionofVIbyexaminingthechangesintheliterature.
The results from this study suggest that VI research is continually changing and will continue to
evolve.Itisthuspossibletotracktheevolutionofsciencebylookingatsemanticrelationshipsand
clustersofwords.
Paper submitted for AquaConSoil 2015 proceedings POSTER - Risk Assessment Theme
Keywords: Human health risk assessment, trichloroethylene, trichloroethene, C4SL, LLTC, PBPK modelling.
ABSTRACT
Trichloroethylene (TCE; trichloroethene; CAS No. 79-01-6) is a soil contaminant found commonly at a
range of sites across Europe. Soil screening values are based upon a human health value as derived
from the most up to date toxicology information on a substance. The most recent toxicology review for
TCE, as performed by the USEPA in 2011, has been met with some controversy in the approaches taken
and choices made in the evaluation. In particular, the derivation of a Reference Dose (RfD) for oral
exposure and Reference Concentration (RfC) for inhalation exposure have relied heavily on
physiologically-based pharmacokinetic (PBPK) modelling for route-to-route and species-to-species
extrapolation and specific choices of toxicological benchmark with associated uncertainty factors that may
or may not be acceptable by all international regulatory bodies.
Historically, the key human health effect of concern for TCE has been kidney cancer. In the EPA 2011
review, the most sensitive effects are three-fold, identified as fetal heart malformations, developmental
immunotoxicity and effects on the thyroid in adults. Kidney cancer remains a sensitive effect, but not the
most sensitive in the new evaluation. All of these effects are considered together, as they all occur at
similar intake doses, making the evaluation relatively complex in comparison to other contaminants. This
has led to a significant reduction in the RfD and RfC values since before 2011 and it is important to
understand the basis of these changes before using the USEPA values in EU contaminated land risk
assessment. Although the reference values are lower than previous, the new RfD (0.5 g/kg/day) and
RfC (2 g/m3; equivalent to an intake of 0.57 g/kg/day) as derived by the USEPA, would be considered
to be 15-fold higher than a minimal risk position in a UK Human Health assessment context.
In March 2014, the UK Department for the Environment, Food and Rural Affairs (Defra) published an
approach for the derivation of Category 4 Screening Levels (C4SLs) for use within the contaminated land
regime (Part 2A of the Environmental Protection Act 1990) in England and Wales. C4SLs are designed
to represent low risk and to provide a simple test for deciding when land is suitable for use and definitely
not contaminated land. The change in UK contaminated land policy has increased the onus for detailed
review of toxicology information by a suitably competent person and in a new framework introduced a
new term, low level of toxicological concern (LLTC) that represents defined risk levels. The presentation
explains the choices that have been made in the US EPA review, and how they have been used in a UK
evaluation using the framework. In this presentation, we demonstrate how the Defra C4SL Project
approach has been applied to derive new oral and inhalation LLTC values for TCE, in order that a set of
C4SLs for different land use scenarios can be derived using an appropriately modified version of the
exposure model CLEA Version 1.06.
A suggested value for both the oral and inhalation LLTC for trichloroethylene is 0.3 g/kg/day.
Category 4 Screening Level (C4SL) Criteria (England & Wales) a C4SL approach for TCE
In 2014, Defra in the UK published a report from its Research Project SP1010 Development of Category
4 Screening Levels for Assessment of Land Affected by Contamination (accessed at
http://randd.defra.gov.uk/Default.aspx?Module=More&Location=None&ProjectID=18341 on 27 April
2015). It was accompanied by a Defra Policy Companion Document dated December 2014, with the
intent that the resulting C4SLs should be used in the context of Defras April 2012 revised statutory
guidance for Part 2A under the Environmental Protection Act 1990 (England & Wales), to help with
determining that land is low risk in Category 4 and therefore is definitely not contaminated in the legal
sense. The research report and policy companion document included the derivation of six C4SLs for
arsenic, cadmium, chromium VI, benzene, benzo[a]pyrene and lead; trichloroethylene (TCE) was not
included.
Here, we follow the harmonized risk assessment principles established in the toxicological framework of
the Defra SP1010 project (Figure 1) and apply them to TCE. The ultimate aim of the framework is to
define a health-based guidance value in the form of a low level of toxicological concern (LLTC), which is a
health-based guidance value that can be input into the Contaminated Land Exposure Assessment
(CLEA) model version 1.06 as an intake of low concern, and coupled with exposure can be shown to
lead to a scenario of low risk. It is important to state that the principles of a minimal risk Health Criteria
Value (HCV) defined within Environment Agency (England & Wales) SR2 and SR3 guidance from 2009,
remain consistent with this new framework illustrated in Figure 1. The framework approach starts with an
overview of the authoritative toxicology data reviews for TCE.
The Environment Agency (England & Wales) last produced a soil guideline value (SGV) for TCE in 2004,
and this was based on a minimal risk HCV of 5.2 g/kg/day (oral and inhalation intakes). Since this time,
new toxicology reviews have been published (which calls for a significant reduction in health-based
guidance values for TCE) that change the conclusions of the (now withdrawn) SGV.
In 2011, the US Environmental Protection Agency (EPA) performed a lengthy and comprehensive review
of TCE toxicology, and proposed candidate reference doses (RfDs) based around three pivotal studies
(as outlined in Table 1). It is critical to understand the underpinning basis of these RfDs and what health
effects can be caused by TCE before choosing to use these RfDs in EU contaminated land risk
assessments, in order to evaluate the level of current concern and know the acceptance of risk these
numbers from US EPA represent.
In 2013, the US Agency for Toxicity and Disease Substance Registry (ATSDR) published an addendum
to its toxicity profile for TCE published in 1997. This was largely a reiteration of the US EPA preferred RfD
value of 0.51 g/kg/day as a Minimal Risk Level (MRL).
In 2014, both the US National Toxicology Program (NTP) and IPCS reviewed new evidence on
carcinogenicity and NTP concluded that the current listing in the Report on Carcinogens for TCE as
reasonably anticipated to be a human carcinogen should be changed to known to be a human
carcinogen based on sufficient evidence of carcinogenicity from studies in humans regarding kidney, liver
and non-Hodgkin lymphoma. TCE is also genotoxic in vitro and in vivo, which means it should be
considered as a non-thresholded carcinogen, even though its metabolites are likely to be causative of
carcinogenicity.
There is at present no review or critique of the new US toxicology evaluation in the UK. The US EPA
values are beginning to be used in contaminated land evaluations (e.g. the Suitable 4 Use levels, S4ULs).
However, the US EPA values would not be considered as minimal risk values in UK contaminated land
practice, nor may they be considered as low risk as per Defra guidance.
Figure 1 A Toxicological Framework to define a Low Level of Toxicological Concern (LLTC) for land contaminants as used
in the context of implementing the revised statutory Part 2A guidance of the Environmental Protection Act 1990 (England
and Wales) for developing C4SLs. (Reproduced from the Defra Project SP1010 report).
Table 1 Pivotal non-cancer toxicology data as a basis for candidate oral reference doses (RfDs) in
the US EPA review.
POD
Ref Dose
Health Effect value POD type UF Pivotal study
mg/kg/day
mg/kg/day
Human
Equivalent Dose 100
(99th percentile) (10 for use of a LOAEL;
based on a 3.16 as a PBPK model Keil et al (2009)
Thymus
LOAEL (0.35 was used for interspecies 0.00048 30 week drinking
decreased 0.048 mg/kg/day) extrapolation; mg/kg/day water study in
weight expressed as an 3.16 as PBPK model was B6C3F1 mice
internal dose used for interindividual
(0.139 variability)
mg/kg/day)
Peden-Adams et al
Immunotoxicity 1000 (2006) from
delayed type (10 for use of a LOAEL; 0.00037 placental,
0.37 LOAEL
hypersensitivity 10 for interspecies and 10 mg/kg/day lactational and
in pups for intrahuman variability) drinking water
exposure.
Human
Equivalent Dose
(99th percentile) 10
based on a (3.16 as a PBPK model Johnson et al
BMD1% (0.065 was used for interspecies (2003) Rat
Fetal heart 0.00051
0.0051 mg/kg/day; extrapolation; developmental
malformations converted to 3.16 as PBPK model was
mg/kg/day
study drinking
0.0142 mg TCE used for interindividual water
oxidized /kg/day variability)
as expressed as
an internal dose)
All of the most sensitive and serious non-cancer (development) endpoints are based upon animal studies;
cancer endpoints are kidney, liver and non-Hodgkins lymphoma, for which there are human data. The
preferred oral RfD in the US EPA evaluation, as also adopted by the ATSDR in 2013, is that derived in
relation to fetal heart malformations as the most sensitive endpoint in rats (Johnson et al (2003). PBPK
modelling has been used to predict internal human relevant doses from the rat drinking water study.
Cancer risk was also reviewed in the US EPA evaluation using the human inhalation data in Charbotel et
al (2006) on incidence of kidney cancer with TCE exposure; for inhalation the unit risk is 4 x 10-6 per
g/m3 and the oral slope factor, resulting from PBPK modelling of route-to-route extrapolation from the
same inhalation study, is 5 x 10-2 per mg/kg/day. The total unit risk of 2 x10-6 from all cancers (kidney,
NHL and liver, with consideration of early life susceptibility) from drinking 1gTCE/L water is calculated in
US EPA 2011 page 5-162; similarly, it is 4.8 x 10-6 per g/m3 air in US EPA 2011 page 5-159.
Below, we interpret this new information in the context of the toxicological framework in Figure 1 and
apply the guidance from Defra Project SP1010, particularly recommendations from the Defra policy
companion document around the choices of margins and cancer risk (not considered in the three US EPA
RfDs directly) that can be considered in deriving an LLTC using these data.
Tables 2 and 3 capture the options as to how an oral and inhalation LLTC respectively, can be derived for
TCE for use in contaminated land risk assessment. The key endpoints chosen are the most sensitive
(and potentially severe/fatal) non-cancer endpoint of fetal heart malformations (FHM)(Johnson et al,
2003) and human total cancer unit risk values of 2 x 10-6 per g/L drinking water and 5 x 10-6 per g/m3
air, both accounting for early life susceptibilities and lifetime exposure (US EPA 2011 evaluation).
Table 2 Results of using the Toxicological Framework to interpret the data and provide evidence
such that an appropriate Oral LLTC can be defined for a trichloroethylene C4SL
Potential Basis for LLTC Value Unit Comments
2004 Minimal Risk Health Criteria Value 5.2 g/kg Index dose mouse inhalation study dose-
(Environment Agency) now withdrawn as BW/day independent cancer endpoints. ELCR 1 in
no longer defensible 100,000. No longer valid.
Using Johnson et al 2003 - based on 0.65 g/kg Applied dose with no PBPK modelling Based
BMD01 (also similar to provisional WHO BW/day upon a BMD01 for Fetal heart malformations in
EU 2011 4th Ed drinking water guideline 20 rat.
g/L) with margin of 100 Incidence FHM = 1 in 10000.
Human cancer risk = 1 in 22,000
US EPA RfD 0.5 g/kg BMD01 but translated to HED99 using PBPK
Using BMD01 Johnson et al (2003) as BW/day modelling.
most sensitive endpoint - fetal heart N.B. This intake is 130-fold lower than the
malformations with internal human BMD01 for FHM not considering PBPK
equivalent dose using PBPK modelling modelling.
with a margin of 10. Human cancer risk = 1 in 28,600
Current European drinking water standard 10 g/L Based on unit risk of 2 x 10-6 per gTCE/L
http://dwi.defra.gov.uk/consumers/advice- (USEPA 2011 total risk of all cancer, including
leaflets/standards.pdf early life susceptibility & lifetime exposure)
Oral Cancer Risk at 10 g/L is
1 in 50000
0.3 g/kg (10 g x 2L per day)/70 kg
BW/day
Oral LLTC Based upon a Cancer Risk of 0.3 g/kg Protective of a sensitive developmental
1 in 50000 BW/day immunotox endpoint (RfD 0.37 g/kg/day)
(also equates to the current UK drinking Protective of fetal heart malformations: the
water standard which avoids LLTC is 216-fold lower than the BMD01 of
disproportionately targeting soil) and is 0.065 mg/kg/day. Approx notional incidence
suitably protective of non-cancer rate of 1 in 22000.
endpoints.
Oral LLTC driven by FHM, based on 0.14 g/kg i.e. BMD10 = 0.7 mg/kg/day divided by a margin
BMD10 and a margin of 5000 BW/day of 5000
(i.e. as per policy companion for non- Incidence FHM = 1 in 50000
thresholded carcinogens, use similar for Cancer risk = 1 in 102000 (minimal risk)
this severe developmental endpoint)
Considerations of minimal risk for FHM
Using Johnson et al 2003 - based on 0.07 g/kg Applied dose BMD10 = 0.7 mg/kg/day divided by
BMD10 (UF 10000) BW/day a margin of 10000 - with no PBPK modelling
Using Johnson et al 2003 - based on 0.02 g/kg Applied dose BMDL01 = 0.02mg/kg/day
BMDL01 (95th %ile) and margin of 1000 BW/day divided by a margin of 1000 with no PBPK
modelling
Minimal Risk HCV using Johnson et al 0.02 g/kg Applied dose BMDL10 = 0.23 mg/kg/day
2003 for the most sensitive endpoint. BW/day divided by an uncertainty factor of 10000
BMDL10 (UF 10000) as per EA SR2
guidance
From the above evidence in Table 2, it is suggested to set an Oral LLTC at 0.3 g/kg BW/day for use in
developing a C4SL for TCE. This value is:
a) Equivalent to an excess lifetime cancer risk via the oral route of 1 in 50000, as based upon an
evaluation of human study data (which is in accordance with Defras policy companion relating to
a low risk ELCR).
b) Is lower than the RfD (0.37 g/kg BW/day) for developmental immunotoxicology endpoints.
c) Represents low risk of fetal heart malformations, a developmental endpoint, with this intake being
216-fold lower than a highly sensitive point of departure (BMD01).
d) Equates to the intake of the current drinking water standard
Table 3 Results of using the Toxicological Framework to interpret the data and provide evidence
such that an appropriate Inhalation LLTC can be defined for a trichloroethylene C4SL
Potential Basis for LLTC Value Unit Comments
2004 Minimal Risk Health Criteria Value 5.2 g/kg Index dose mouse inhalation study dose-
(Environment Agency) now withdrawn as BW/day independent cancer endpoints. ELCR 1 in
no longer defensible 100,000. No longer valid.
Based on Cancer Risk of 1 in 50000 1.1 g/kg Assuming a 70kg person inhales 20m3 air per
[Using 2011 US EPA evaluation] BW/day day, containing 4 gTCE/m3
4 g/m3 Cancer risk = 1 in 50000, based on a unit risk of
5 x 10-6 (human study Charbotel et al 2006).
US EPA RfC 0.57 g/kg Assuming a 70kg person inhales 20m3 air per
BW/day day, containing 2 gTCE/m3
The basis is cited as the non-cancer effects
(as per Table 1 above): fetal heart 2 g/m3 The dose in g/kg BW/day is based upon route
malformations, adult and developmental to route PBPK extrapolation from oral non
immunotoxicity. PBPK modelling has been cancer effects (drinking water studies)
used to perform route-to-route Cancer effects: Kidney tumours; NHL, and liver
extrapolation for these three most sensitive cancers are based on human inhalation
endpoints originally from drinking water exposure data.
study data in rodents. Cancer risk = 1 in 100000, based on a unit risk
of 5 x 10-6 (human study Charbotel et al 2006).
Recommended Inhalation LLTC (at the 0.3 g/kg Non-cancer toxicity drives the risk: assumes the
current time) equate with the oral LLTC, BW/day toxicological consequences of absorption
as driven by the most sensitive non-cancer through the gut and lungs is the same, as per
endpoints. Remains suitably previous EA 2004 guidance. Factors in no
precautionary as per Defra guidance. PBPK extrapolation or differential route-specific
metabolism at the moment these assumptions
carry high uncertainty and methods have not
been fully considered and reviewed at National
level in the UK.
From the above evidence in Table 3, it is suggested to set an Inhalation LLTC at 0.3 g/kg BW/day for
use in developing a C4SL for TCE. This value is:
a) Equivalent to the oral LLTC from Table 2 above; making a current assumption that the inhalation
route kinetics are the same as the oral route kinetics leading to the same non-cancer effects. This
is a suitably precautionary position for current regulatory use in the UK, as PBPK modelling is
used in multiple ways in the US EPA assessment to perform route-to-route extrapolations,
internal human dose estimates and incorporating oxidative metabolism assumptions.
b) Protective of all cancer effects by the inhalation route; ELCR is calculated at 1 in 200000 at this
intake dose.
From the above analysis, it is proposed that both the Oral and Inhalation LLTC value for trichloroethylene
is set at a suitably precautionary 0.3 g/kg BW/day. These LLTC values would be lower than the RfD
(0.5 g/kg BW/day) and RfC (2g/m3; equivalent to 0.57 g/kg BW/day) proposed by the US EPA in 2011
following their comprehensive review of new toxicological data, but protective of all non-cancer and
cancer effects.
It should also be noted that the most sensitive effects (fetal heart malformations and immunotoxicity) are
developmental in nature and therefore sensitive time windows of exposure (i.e. during gestation) requires
careful consideration in the exposure modelling side of a risk-based evaluation, where the C4SL should
be derived for pregnant women and the developing fetus, as the most sensitive human receptors
potentially exposed to TCE.
PBPK modelling techniques and metabolism studies are useful in refining risk assessment across species
and performing route-to-route extrapolations. PBPK modelling and assumptions around oxidative
metabolism have been used heavily in the US EPA assessment. Measured data to parameterize the
models is not comprehensive, and it is possible that there is a high degree of uncertainty in the modelling
underpinning the US EPA values. It is important before using the EPA values directly in contaminated
land risk assessment, that there is a review and discussion in UK/EU about the parameterisation of the
PK models used, a critique of the underlying assumptions made and data on metabolism reviewed in
order that there is full transparency about the uncertainties in the US EPA evaluation.
Given these specific details for TCE modelling, and indeed the general use of PBPK modelling and
metabolism data, are yet to be debated and reviewed at national level in the UK for contaminated land
risk assessment purposes, it remains suitably precautionary to set the LLTCs for TCE at values slightly
lower than the US EPA reference doses until such a review is performed.
In deriving the C4SLs using these LLTCs for the various exposure scenarios and considering the risks of
developmental effects, the most sensitive human receptors should be the pregnant woman and
developing foetus during the sensitive window of gestational exposure; for cancer effects the receptor is
the standard residential child and commercial adult worker.
References
ATSDR. 1997. Toxicological Profile for Trichloroethylene. Atlanta, GA: Agency for Toxic Substances and
Disease Registry. 335 pp.
ATSDR. 2013. Addendum to the Toxicological Profile for Trichloroethylene. Atlanta, GA: Agency for Toxic
Substances and Disease Registry. 120 pp.
Charbotel, B; Fevotte, J; Hours, M; Martin, J-L; Bergeret, A. (2006). Case-control study on renal cell
cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects. Ann Occup Hyg
50: 777-787.
Environment Agency (England & Wales) 2004 Contaminants in Soil: Collation of Toxicological Data and
Intake Values for Humans. Trichloroethene. Science Report Tox 24.
EPA. 2011a. Toxicological Review of Trichloroethylene (CAS No. 79-01-6) In Support of Summary
Information on the Integrated Risk Information System (IRIS). EPA/635/R-09/011F. U.S. Environmental
Protection Agency. 1200 pp.
EPA. 2011b. Toxicological Review of Trichloroethylene Appendices (CAS No. 79-01-6) In Support of
Summary Information on the Integrated Risk Information System (IRIS). EPA/635/R-09/011F. U.S.
Environmental Protection Agency. 1269 pp.
IARC. 2014. Trichloroethylene, Tetrachloroethylene and Some Other Chlorinated Agents, IARC
Monographs on the Evaluation of Carcinogenic Risks to Humans. vol. 106, Lyon, France: International
Agency for Research on Cancer.
Johnson PD, Goldberg SY, Mays MZ, et al. 2003. Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat. Environ Health
Perspect 111(3):289-292.
NTP. 2011. Trichloroethylene. In Report on Carcinogens. 12th ed. Research Triangle Park, NC: National
Toxicology Program. pp. 420-423.
Key words: soil quality values, transfer to eggs, dioxins, PCBs, pesticides, metals, backyard chickens
Introduction
Growing your own food has received renewed interest during recent years. Allotments, community
gardens, urban gardens, are initiated or enlivened. These initiatives are often planned on
redeveloped land, where concern for soil contamination is justified. Attention generally focusses on
growing own vegetables, but the practice of keeping backyard chicken for fresh eggs remains
somewhat outside the picture. Incidences of e.g. dioxin contamination in commercial eggs due to
contaminated feed are well-known. But what about eggs from private chicken runs in backyards?
Chickens in backyards generally have plenty access to (bare) soil and pick soil organisms, which
could result in uptake and transfer to the egg. Recent studies have indeed confirmed that eggs from
backyard chickens contain higher levels of dioxins and PCBs than commercial eggs and that soil is a
potential exposure source, be it directly or indirectly. But is this problem limited to dioxins and PCBs or
are more chemicals of concern? And what are safe levels in soil? To answer these questions, we
used the S-Risk model to calculated safe levels in soil for residential and allotment scenarios,
including consumption of backyard eggs. The S-Risk module allows calculation of transfer to chicken
eggs through feed and soil intake. However, where transfer factors from feed to chicken eggs can be
found for chemicals such as dioxins, PCBs, persistent pesticides and a range of metals, transfer
factors from soil to egg are generally lacking. We therefore developed an approach to estimated soil-
to-egg transfer factors using feed-to-egg transfer factors, bioavailability estimates and earthworm
bioaccumulation factors. We verified this approach with experimental data on dioxins and DDT. We
also evaluated whether feed-to-egg transfer factors could be derived from QSARs that are used to
predict fat biotransfer for organic chemicals without experimental transfer factors. This approach is
suggested by some authors. Taking all information together, we calculated safe soil concentrations for
the residential and allotment scenarios, the results of which also revealed what chemicals could be of
concern when keeping backyard chickens for egg consumption. As expected, dioxins and PCBs
require priority, also in view of soil levels found. In addition, DDT (despite of its ban since the 1970s).
S-Risk model
The S-Risk model is an online software tool (www.s-risk.be), that allows the calculation of human
health based screening values for soil quality and of human health risks for contaminants in soil. It
consists of 4 modules: a) within soil distribution module, b) fate and transfer module from soil to other
environmental compartments, c) exposure module to quantify human exposure to the relevant
environmental media, and c) risk characterization module comparing exposure with toxicological
criteria and concentrations with limit values. The model is a stead-state conservation of mass model. It
contains a database with chemical properties consisting of metals, aromatics, PAHs, chlorinated
aliphatics and aromatics, . Default exposure scenarios are provided, establishing exposure
pathways and exposure factor values. The S-Risk model was used to calculate the soil quality levels
with modifications detailed below to represent specific scenarios of local gardening. For this,
application type I with automated iteration procedure was run.
Scenarios
Based on a review and discussion with experts, a set of scenarios considered relevant for local
gardening was constructed. A distinction was made between a) the situation of a vegetable garden
and/or chicken run near the house, where exposure due to the residential function is combined with
exposure from local food production, and b) an allotment and/or chicken run in a dedicated area at
distance from the gardeners home, where only exposure due to gardening and consumption of local
food is considered. A selection of scenarios used in the calculations is given in Table 1. Further
scenario details can be found in the English summary report (Cornelis et al., 2014).
An extended list of chemicals was agreed, taking into account chemicals that are part of the Flemish
legislation on soil contamination, chemicals considered relevant following Flemish biomonitoring,
chemicals known to accumulate in chicken eggs, chemicals included in Dutch lists with land use
specific limit values for soil.
Chemical properties, biological transfer factors and toxicological reference values were taken from the
S-Risk substance data sheets (available for download from www.s-risk.be). For additional chemicals
selected from Dutch lists, Dutch background documents were taken as a starting point. For the
remaining chemicals and for those chemicals were updating of the information was considered
necessary, data were searched for in the literature. Specific attention was given to: a) actualisation of
the toxicological reference values, and b) information to estimate transfer from soil and feed to eggs
for persistent organic pollutants.
For inorganic chemicals, bioconcentration factors (BCF) related to total soil concentration are required
input values of the S-Risk model. When available, bioconcentration relationships were used,
incorporating the influence of soil properties on plant uptake.
For organic chemicals, concentrations in plants due to root uptake can be calculated using chemical
properties (log Kow, log Koa, ), soil and plant properties (QSARs); or BCF values related to pore
water concentrations can be entered in the S-Risk model. For all organic chemicals considered, the
QSAR option was used.
In addition to concentration in the plant following root uptake, particle deposition (due to soil
resuspension) and soil splash is modelled as well.
Table 1: Properties of selected scenarios used to calculated soil quality criteria for local gardening
scenarios with residential function scenarios without residential function
egg consumption X X X X
local vegetable fraction high moderate zero high moderate high high moderate
Exposure pathways
ingestion of and dermal uptake from soil X X X X X X X X
ingestion of and dermal uptake from indoor dust X X X X X
local egg consumption X X X X
local vegetable consumption X X X X X X X
ingestion of and dermal uptake from drinking-water X X X X X
inhalation outdoor air X X X X X X X X
inhalation indoor air X X X X X
inhalation showering X X X X X
Fraction of local consumption
potatoes 0.5 0.39 0 0.5 0.39 0.5 0.5 0.39
root vegetables 1 0.36 0 1 0.36 1 1 0.36
bulb vegetables 1 0.52 0 1 0.52 1 1 0.52
fruit vegetables 1 0.39 0 1 0.39 1 1 0.39
cabbages 1 0.21 0 1 0.21 1 1 0.21
leafy vegetables 1 0.36 0 1 0.36 1 1 0.36
pulses 1 0.42 0 1 0.42 1 1 0.42
stem vegetables 1 0.10 0 1 0.10 1 1 0.10
eggs 1 0.6 0.6 0 0 1 0 0
Exposure frequency
days per week 7 7 7 7 7 1 1 1
weeks per year 52 52 52 52 52 40 40 40
Transfer factors from feed and soil to chicken eggs
Transfer to chicken eggs requires two transfer factors (expressed as concentration in egg relative to
intake): a biotransfer factor (BTF) from feed to egg and a biotransfer factor (BTF) from soil to egg.
The primary source for feed-to-egg BTFs for metals was Sheppard et al. (2010). If data were lacking in
this information source, a literature search was performed. The BTF from soil to egg was set equal to
the BTF from feed to egg, except for lead and arsenic. In the case of lead, we used information from
Waegeneers et al. (2009) to estimate the soil-to-egg BTF. In the case of arsenic, we used the same
approach as applied for persistent organic chemicals (see hereafter). Seen the low transfer of metals
to chicken eggs, more accurate estimation of soil-to-egg BTFs for the other metals than lead and
arsenic was not considered a priority.
For organic chemicals feed-to-egg BTFs were searched for in the literature. As persistent pesticides,
dioxins and PCBs tend to accumulate in eggs, it was considered important to derive soil-to-egg BTFs
for these groups of chemicals. Transfer from soil to eggs results from two pathways: 1) the direct
ingestion of soil, and 2) the indirect intake through accumulation in soil invertebrates (earthworm) and
ingestion of these invertebrates. The soil-to-egg BTF was therefore calculated as follows:
= ( + )
With
The daily intake of earthworms was set at 20 g fw/d (Waegeneers et al., 2009) with a dry matter
content of 16 % (BBL, 2006). Daily soil ingestion was set at 30 g/d (default value S-Risk model). We
validated the model to the extent possible by using experimental data for DDT, dioxins and PCBs
(Colles et al., 2014, Cornelis et al., 2011, Van Overmeire et al., 2009). These compounds are the most
relevant based on occurrence in backyard chicken eggs and human toxicity. The final transfer factors
are listed in Table 2 and discussed below.
Table 2: Biotransfer factors for chicken eggs (backyard chicken) for persistent organic compounds
chemical BTFfeed-egg (d/kg) RBAsoil (-) BCFworm (dm BTFsoil-egg calculated
basis) (d/kg)
DDT 9.8 0.25 11.2 14
dieldrin 10.6 0.12 14.7 18
endrin 4.9 0.12 14.7 8
hexachlorobenzene 15.5 0.4 6.5 17
pentachlorophenol 0.28 1 14.6 0.7
-HCH 1.3 0.6 5 1.5
-HCH 19 0.6 5 21
-HCH 2.4 0.6 5 2.8
chlordane 20 0.5 0.24 11
dioxins 8.3 n.a. n.a. 0.011
dioxin-like PCBs 10 n.a. n.a. 30
marker PCBs 7.8 (PCB-28) n.a. n.a. 26
10 (PCB-153)
n.a.:not applicable
For DDT, dieldrin, endrin and hexachlorobenzene we derived the BTFfeed-egg from the BCFs in Kan
(1978). To convert between BCF and BTF we used a feed consumption of 123 g/d (default S-Risk
value). There are no experimental BTFfeed-egg values for pentachlorophenol. Brambilla et al. (2009)
found a 30 times lower transfer factor of pentachlorophenol from contaminated wood shavings
compared to dioxins. We tentatively used this ratio to set the BTF from feed to egg for
pentachlorophenol. The BTFfeed-egg for chlordane was taken from Foster et al. (1972). The BCFs from
soil to worm were taken from US-EPA (2007). For hexachlorobenzene, the BCF is taken from Vampre
et al. (2010). The BCF for lindane (-HCH) was derived from Vlkov and Hofman and used for all
HCHs (we did not search for more differentiated values). The BCF for chlordane was taken from
Menzie et al. (1992). The RBAsoil for DDT was taken from a bioavailability study with mice (Smith et al.,
2012). The RBAsoil for dieldrin was taken from Wang et al. (2013) who measured the bioaccessibility of
organochlorine pesticides in dust. In that study the bioaccessibility of DDT was similar to the
bioavailability in Smith et al. (2012). BCF soil to worm and RBAsoil for endrin were taken the same as
for dieldrin by lack of data. The RBAsoil value for hexachlorobenzene is very tentative and derived from
Ma et al. (2008) and comparing the DDT data with Smith et al. (2012). The bioaccessiblity and
bioavailability of pentachlorophenol seems to be high (Pu et al., 2003, Puglisi et al., 2009, Spagnuolo
et al., 2010). We therefore set it at 1. RBAsoil for lindane was taken from the bioaccessibility study of
Tao et al. (2009). The RBAsoil for chlordane was set at 0.5, being the value for other organochlorine
pesticides with similar log Kow.
For endosulfan we estimated a value from the study of Bargar et al. (2001a,b) with transfer to eggs
measured after injection of the dose. As transfer will not be significant we set the BTF from feed to
eggs and soil to eggs at the same value.
For PAHs the BTF to eggs was set at 0. The CONTEGG study showed that PAHs could not be
detected in eggs of backyard chicken (Van Overmeire et al., 2009).
For lipophilic compounds, it seems that the transfer from soil to egg (including earthworms) is much
higher than the transfer from feed in backyard chickens. We therefore used the BTF feed-egg from
published studies assessed in earlier work (Cornelis et al., 2007) for dioxins, dioxin-like PCBs and
marker PCBs, and derived their BTFsoil-egg from the studies in backyard chickens conducted in
1
Flanders and Belgium (Colles et al., 2014, Cornelis et al., 2011, Van Overmeire et al., 2009) .
Most other organic contaminants in the list did not have experimentally determined biotransfer factors
for chicken eggs. We therefore searched for regression equations that could predict the biotransfer.
We found several equations.
McKone (1993) used the regression equations of Travis and Arms (1988) and assumed the feed-fat
transfer factor being the same for cattle and chickens. The transfer factor accounts for the fat content
in chicken eggs and the feed intake of a chicken.
= 1.6. 106
US-EPA (2005) also assumed no difference between cattle and chickens for the fat-based transfer
factor. The fat-based transfer factor (RTI, 2005) is given as
The transfer factor feed to egg is obtained through multiplication with the fat content of an egg. US-
2
EPA (2005) takes a fat content of 8 % . As the BTFfat was derived for chemicals with a log Kow
between -0.67 and 8.2, substances with a log Kow falling outside that range will have a BTF equal to
the value at either log Kow 0.67 or 8.2 (whathever applies).
S-Risk uses the equations of Travis and Arms (1988) for meat and milk, limiting them to a value
corresponding to a log Kow < -1 for meat and < -1.8 for milk. We assumed the BTF for milk could be
converted on a fat basis to a BTF for eggs, using 4 % fat in milk and 9.5 % fat in eggs.
Leeman et al. (2007) analyzed the data for a large set of substances to derive bioconcentration factors
to eggs that can be used for risk-based screening. Bioconcentration factors (as P95 values per log
1
The same was done for DDT, but for DDT the developed model gave better correspondance with the
experimental data than when it was used for dioxins and PCBs (for dioxins and PCBs we had no information on
the RBAsoil).
2
In our calculations for this project we used a fat content of 9.5 %, based on the NEVO tables which report a
range of 9 9.5 % (http://nevo-online.rivm.nl)
Kow class) are given. We converted them to a BTF using a feed intake of 123 g/d. For the lower log
Kow classes (where we have no experimental data) the chemicals in the database are mostly
pesticides and veterinary medicines, while our chemicals are aromatic substances, chlorinated
chemicals,
We compared the different approaches and also listed available experimental values from the
publications we collated. Results are given in Figure 1.
1E+03
1E+02
beta-HCH
dieldrin
DDT
1E+01 PCBs and dioxins
endrin heptachloor
1E+00
alfa-HCH
gamma-HCH
1E-01
BTF egg (d/kg)
McKone (1993)
1E-02
HHRAP (2005) 9,5%
chloorpyrifos S-Risk
1E-03
experimental values
Leeman (2007)
1E-04
1E-05
1E-06
1E-07
0 1 2 3 4 5 6 7 8 9
log Kow
Figure 1: Comparison of approaches to estimate a feed to egg transfer factor with some experimental
transfer factors
Conclusion of this exercise is that it is not possible to make an estimate of the transfer of organic
chemicals from feed (or even soil) to eggs. It can not be excluded that the proposal of Leeman (2007)
could work, but seen the differing chemical types in their database and the ones we consider at lower
log Kow, we can not verify this.
We therefore decided not to calculate scenarios with transfer to eggs for those chemicals without
experimental data.
Results
Results of calculated soil quality criteria for selected contaminants, for which either vegetable
consumption or egg consumption influences the soil quality criteria are shown in Table 3. The
extended table can be found in Cornelis et al. (2014).
Table 3: Calculated soil quality criteria for local gardening (mg/kg dm)
scenarios with residential function scenarios without residential function
egg consumption X X X X
local vegetable fraction high moderate zero high moderate high high moderate
metals
arsenic 18 24 27 34 40 26 96 152
cadmium 8.4 21 91 8.7 22 9.2 9.5 29
mercury (inorganic) 0.85 1.35 6.8 0.85 1.35 0.95 0.95 1.6
lead 200 225 232 274 282 396 885 1222
nickel 99 125 157 125 152 168 265 447
Polycyclic aromatic hydrocarbons
benzo(a)anthracene 26 32 38 26 32 67 67 143
benzo(a)pyrene 3.6 4.9 6.3 3.6 4.9 7.3 7.3 16
benzo(b)fluoranthene 29 37 44 29 37 73 73 162
benzo(k)fluoranthene 29 37 44 29 37 74 74 163
dibenzo(a,h)anthracene 3.3 4.4 5.5 3.3 4.4 7 7 16
indeno(1,2,3-cd)pyrene 29 38 44 29 38 75 75 165
Pesticides
Aldrin + dieldrin 0.03 0.05 0.06 0.33 0.2 0.03 0.38 0.23
DDT+DDD+DDE 0.61 1 1.0 27 25 0.62 62 55
hexachlorobenzene 0.09 0.14 0.14 1.6 2 0.09 3.4 5.5
-Hexachlorocyclohexanea) 31 66 153 77 313 32 226 1300
-Hexachlorocyclohexanea) 0.52 0.88 0.92 4 4 0.52 7.4 16
-Hexachlorocyclohexanea) 0.02 0.03 0.04 0.03 0.04 0.02 0.03 0.05
Pentachlorophenol 2.2 3.2 6.0 3.4 4.7 2.5 4.3 6.6
Dioxins and PCBs
PCDD/Fb) (TEQ) 7.2E-6 1.2E-5 1.2E-5 3.2E-4 5E-4 7.3E-6 4.8E-4 1E-3
Dioxin-like PCBsb) (TEQ) 2.3E-6 3.8E-6 3.8E-6 2.4E-4 3.4E-4 2.3E-6 4.6E-6 1E-3
Marker-PCBs 0.02 0.036 0.036 1.72 2.34 0.02 3.1 6.2
a): the risks of the three HCH isomers should be combined. At the level of the soil quality criteria this means that if
+ + > 1, there is a risk.
/
b): the risks for PCDD/F en dl-PCB should be combined. At the level of the soil quality criteria this means that if + > 1, there is a risk.
/
Discussion
At present, soil quality criteria considering the consumption of locally grown vegetables are available
for some countries. However, safe levels in soil, considering consumption of eggs from backyard-
chickens, are lacking. Studies have shown that these eggs contain higher levels of persistent organic
pollutants than eggs from commercial chicken farms (including conventional and organic type eggs).
The source of this increased levels is considered to be the soil compartment, where backyard
chickens generally have plenty of access to. Ingestion of soil and soil invertebrates can lead to
increased intake and transfer to eggs. Calculations of safe soil concentrations considering the
consumption of eggs from backyard chickens shows that these soil concentrations are substantially
lower than commonly used quality criteria for the persistent compounds like dioxins and PCBs and the
older pesticides. Although the use of the latter class of pesticides is generally prohibited in Europe,
studies in Flanders show occurrence of some of these in the environment, chicken eggs and humans.
Calculation of safe levels requires the use of biotransfer factors. Although feed-to-egg biotransfer
factors were available for most of the persistent compounds, soil-to-egg biotransfer factors are often
not available. We therefore developed an approach to estimate these from the feed-to-egg biotransfer
factors, relative bioavaibility and accumulation in earthwoms. Existing QSAR models derived from
meat or milk transfer factors are not suitable to estimate feed-to-egg biotransfer factors. The
classification of Leeman et al. (2007) could be useful as a first approach.
Acknowledgements
This study was financed by the Public Waste Agency of Flanders (OVAM). We also thank Frank
Swartjes (RIVM, NL), Paul Rmkens (Alterra, NL), Chris Collins (University of Reading, UK), and
Preben Bruun (Ministry of the Environment, Denmark) to answer our questions about soil quality
criteria for local gardening in their country.
References
Bargar T.A., Scott G.I. & Cobb G.P. (2001a). Uptake and distribution of three PCB Congeners and
endosulfan by developing white leghorn chicken embryos (Gallus domesticus),
Arch.Environ.Contam.Toxicol. 41(4): 508-514.
Bargar T.A., Scott G.I. & Cobb G.P. (2001b). Maternal transfer of contaminants: Case study of the
excretion of three polychlorinated biphenyl congeners and technical-grade endosulfan into eggs by
white leghorn chickens (Gallus domesticus), Environmental Toxicology and Chemistry 20(1): 61-
67.
Brambilla G., Fochi I., De Filippis S.P., Iacovella N. & di Domenico A. (2009). Pentachlorophenol,
polychlorodibenzodioxin and polychlorodibenzofuran in eggs from hens exposed to contaminated
wood shavings, Food Additives and Contaminants Part A-Chemistry Analysis Control Exposure &
Risk Assessment 26(2): 258-264
BBL (2006). Resource Conservation and Recovery Act Facility Investigation Phase II Report - NAO
Flint Operations Site - appendix H: ecological risk evaluation supporting information,
http://epa.gov/region05/cleanup/rcra/buickcity/index.html
Colles A., Cornelis C., Van de Mieroop E. & Paulussen M. (2013). Dioxins, PCBs and DDT in soil and
egg samples from Menen, Wevelgem and Wervik (report in Dutch). VITO, MOL, Belgium.
Cornelis C., Servaes K., Van den Bosch B., Vloemans P., Hufkens N. & Touchant K. (2011b). Dioxins
and PCBs in eggs and vegetables of private gardens interpretation of results (report in Dutch).
VITO, MOL, Belgium
Cornelis, C., Van Holderbeke, M., Bierkens, J. (2014). Soil quality criteria for gardens summary
report, VITO, Mol, Belgium (available on request).
Foster T.S., Morley H.V., Purkayastha R., Greenhalgh R. & Hunt J.R. (1972). Residues in eggs and
tissues of hens fed a ration containing low levels of pesticides with and without charcoal, Journal of
Economic Entomolology 65: 982-988.
Leeman W.R., Van den Berg K.J. & Houben G.F. (2007). Transfer of chemicals from feed to animal
products: the use of transfer factors in risk assessment, Food Additives & Contaminants 24(1): 1-
13.
Ma R., Entwistle J. & Dean J. (2008). In-Vitro bioaccessibility studies of POPs - presentation.
http://www.reading.ac.uk/web/FILES/shes/MA.pdf.
McKone T.E. (1993). The precision of QSAR methods for estimating Intermedia transfer factors in
exposure assessments, SAR and QSAR in Environmental Research 1(1): 41-51.
Menzie, Charles A, Burmaster D.E., Freshman J.S. & Callahan C.A. (1992). Assessment of methods
for estimating ecological risk in the terrestrial component: a case study at the Baird & McGuire
Superfund Site in Holbrook, Massachusetts, Environmental Toxicology and Chemistry 11: 245-260.
Pu X., Carlson G.P. & Lee L.S. (2003). Oral bioavailability of pentachlorophenol from soils of varying
characteristics using a rat model, Journal of Toxicology and Environmental Health, Part A 66(21):
2001-2013.
Puglisi E., Vernile P., Bari G., Spagnuolo M., Trevisan M., Lillo E.d. & Ruggiero P. (2009).
Bioaccessibility, bioavailability and ecotoxicity of pentachlorophenol in compost amended soils,
Chemosphere 77(1): 80-86.
RTI (2005). Methodology for predicting cattle biotransfer factors.
Sheppard S.C., Long J.M. & Sanipelli B. (2010a). Verification of radionuclide transfer factors to
domestic-animal food products, using indigenous elements and with emphasis on iodine, Journal of
Environmental Radioactivity 101: 895-901.
Smith E., Weber J., Rofe A., Gancarz D., Naidu R. & Juhasz A.L. (2012). Assessment of DDT Relative
Bioavailability and Bioaccessibility in Historically Contaminated Soils Using an in Vivo Mouse
Model and Fed and Unfed Batch in Vitro Assays, Environmental Science & Technology 46(5):
2928-2934.
Spagnuolo M., Puglisi E., Vernile P., Bari G., de Lillo E., Trevisan M. & Ruggiero P. (2010). Soil
monitoring of pentachlorophenol by bioavailability and ecotoxicity measurements, Journal of
Environmental Monitoring 12(8): 1575-1581.
Tao S., Lu Y., Zhang D., Yang Y., Yang Y., Lu X. & Sai D. (2009b). Assessment of Oral
Bioaccessibility of Organochlorine Pesticides in Soil Using an In Vitro Gastrointestinal Model,
Environmental Science & Technology 43(12): 4524-4529.
Travis C.C. & Arms A.D. (1988). Bioconcentration of organics in beef, milk and vegetation,
Environmental Science & Technology 22: 271-274.
US-EPA (2005a). Human health risk assessment protocol for hazardous waste combustion facilities,
uitgegeven door United States Environmental Protection Agency.
US-EPA (2007). Guidelines for developing Ecological Soil Screening Levels (Eco-SSLs) - Attachment
4-1: Exposure factors and bioaccumulation models for derivation of wildlife Eco-SSLs.
Vampre T.M., Fuccillo R. & de Andrea M.M. (2010). Eisenia Andrei (Oligochaeta) As Bioindicator of
Soil Contamination by Hexachlorobenzene (Hcb), Revista Brasileira de Ciencia do Solo 34(1): 59-
66.
Van Overmeire I., Pussemier L., Waegeneers N., Hanot V., Windal I., Boxus L., Covaci A., Eppe G.,
Scippo M.L., Sioen I., Bilaur M., Gellynck X., De Steur H., Tangni E.K. & Goeyens L. (2009).
Assessment of the chemical contamination in home-produced eggs in Belgium: general overview of
the CONTEGG study, Science of the Total Environment 407: 4403-4410.
Vlkov K. & Hofman J. (2012). A comparison of POPs bioaccumulation in Eisenia fetida in natural
and artificial soils and the effects of aging, Environmental Pollution 160(0): 49-56.
Waegeneers N., De Steur H., De Temmerman L., Van Steenwinkel S. & Viaene J. (2009c). Transfer of
soil contaminants to home-produced eggs from Belgium: levels, contamination sources and health
risks, Science of the Total Environment 407(15): 4438-4446.
Wang W., Huang M.J., Wu F.Y., Kang Y., Wang H.S., Cheung K.C. & Wong M.H. (2013). Risk
assessment of bioaccessible organochlorine pesticides exposure via indoor and outdoor dust,
Atmospheric Environment 77(0): 525-533.
OCCURENCE OF ANTIMONY IN SOIL AND GROUNDWATER AT FORMER SHOOTING
RANGES
Jesper Alroe Steen1, Maren Kann Hostrup2, Mette Marie Mygind2, Boerge Hansen2, Henrik
Soegaard Larsen1, Jacqueline Falkenberg1
jas@niras.dk
Background: Up until 2014, site investigations at military shooting range locations in Denmark
have focused on lead as the primary contaminant of concern. Studies of sampling and analytical
practice in other NATO countries have identified antimony as a contaminant of possible
concern. Currently, Denmark has no quality criterion for antimony in soil or groundwater, which
presents problems with respect to the drivers for site investigation and risk assessment.
Prior to site remediation at three shooting range locations, the Danish Defense have analyzed
soil and groundwater samples in order to assess the occurrence of antimony. Subsequently, site
remediation has been performed based on findings from the site investigations including the
screening for antimony and the final risk assessment.
The site investigations have provided information on the presence of both lead and antimony as
critical contaminants at former shooting range locations.
Aim: The aim of this presentation is to present recent studies on the occurrence of antimony in
soils and groundwater at former military shooting ranges in Denmark. The presentation will
include information concerning projectiles and activities that form the basis for the occurrence of
antimony at shooting ranges, a correlation between the presence of both lead and antimony,
and a comparison with contaminant levels at similar sites in other countries.
A secondary aim of the presentation is to highlight the process (focus on communication with
environmental authorities) involved for site remediation if no national quality criteria are derived.
The process requires the derivation of clean-up goals and approval by the environmental
authorities based on documentation of the risk assessment process.
Project: The project has included a literature study on the occurrence of antimony at shooting
range soils, including a correlation for the presence in the soil of both lead and antimony (no
national data existed).
Pre-investigations, including discrete soil and groundwater sampling, prior to site remediation, at
three Danish shooting range locations have been performed in order to assess the occurrence
of antimony. At the three locations, antimony was found in top soil in concentrations up to 515
mg/kg DW, 265 mg/kg DW and 12 mg/kg DW, respectively, and lead was present in
concentrations up to 18,600 mg/kg DW.
As a significant supplement to the discrete soil sampling, Multi Incremental Sampling (MIS) has
been used to estimate average contaminant concentrations in target areas.
Groundwater samples have been collected from screened intervals in wells in the target area.
Antimony concentrations in groundwater (ranging from non-detects to 2.2 g/l) have been
compared to the average antimony concentrations in soil in target areas illustrating the potential
for leaching of antimony to the groundwater.
Based on these investigations, risk assessment has been performed in order to evaluate the
potential risks to human health due to soil contact and on groundwater quality.
In relation to site remediation at the three specific shooting range locations, a proposal for a soil
criteria for antimony has been prepared. The Danish EPA has undertaken to publish by early
2015, a national Danish soil and groundwater criterion for antimony which will be included in the
presentation.
The presentation will also include data from the three site remediations, including the procedure
for clean-up documentation.
Conclusion: Antimony and lead do co-exist in soil at shooting range locations. As the bullet
core consist of 92-95% lead, 2-5% antimony and traces of other metals, contaminant
concentrations at shooting range locations will be dominated by lead, but antimony can be a
significant contaminant.
The mobility of antimony due to leaching from contaminated soil is greater than the mobility of
lead, however both contaminants are strongly adsorbed to soil particles and transport in the
unsaturated zone is thus greatly retarded.
For future investigations at shooting range locations, it will be relevant to sample for antimony,
especially in relation to the pending national Danish soil and groundwater criteria for antimony.
Paper submitted for AquaConSoil 2015 proceedings POSTER - Risk Assessment Theme
Keywords: Human health risk assessment, trichloroethylene, trichloroethene, C4SL, LLTC, PBPK modelling.
ABSTRACT
Trichloroethylene (TCE; trichloroethene; CAS No. 79-01-6) is a soil contaminant found commonly at a
range of sites across Europe. Soil screening values are based upon a human health value as derived
from the most up to date toxicology information on a substance. The most recent toxicology review for
TCE, as performed by the USEPA in 2011, has been met with some controversy in the approaches taken
and choices made in the evaluation. In particular, the derivation of a Reference Dose (RfD) for oral
exposure and Reference Concentration (RfC) for inhalation exposure have relied heavily on
physiologically-based pharmacokinetic (PBPK) modelling for route-to-route and species-to-species
extrapolation and specific choices of toxicological benchmark with associated uncertainty factors that may
or may not be acceptable by all international regulatory bodies.
Historically, the key human health effect of concern for TCE has been kidney cancer. In the EPA 2011
review, the most sensitive effects are three-fold, identified as fetal heart malformations, developmental
immunotoxicity and effects on the thyroid in adults. Kidney cancer remains a sensitive effect, but not the
most sensitive in the new evaluation. All of these effects are considered together, as they all occur at
similar intake doses, making the evaluation relatively complex in comparison to other contaminants. This
has led to a significant reduction in the RfD and RfC values since before 2011 and it is important to
understand the basis of these changes before using the USEPA values in EU contaminated land risk
assessment. Although the reference values are lower than previous, the new RfD (0.5 g/kg/day) and
RfC (2 g/m3; equivalent to an intake of 0.57 g/kg/day) as derived by the USEPA, would be considered
to be 15-fold higher than a minimal risk position in a UK Human Health assessment context.
In March 2014, the UK Department for the Environment, Food and Rural Affairs (Defra) published an
approach for the derivation of Category 4 Screening Levels (C4SLs) for use within the contaminated land
regime (Part 2A of the Environmental Protection Act 1990) in England and Wales. C4SLs are designed
to represent low risk and to provide a simple test for deciding when land is suitable for use and definitely
not contaminated land. The change in UK contaminated land policy has increased the onus for detailed
review of toxicology information by a suitably competent person and in a new framework introduced a
new term, low level of toxicological concern (LLTC) that represents defined risk levels. The presentation
explains the choices that have been made in the US EPA review, and how they have been used in a UK
evaluation using the framework. In this presentation, we demonstrate how the Defra C4SL Project
approach has been applied to derive new oral and inhalation LLTC values for TCE, in order that a set of
C4SLs for different land use scenarios can be derived using an appropriately modified version of the
exposure model CLEA Version 1.06.
A suggested value for both the oral and inhalation LLTC for trichloroethylene is 0.3 g/kg/day.
Category 4 Screening Level (C4SL) Criteria (England & Wales) a C4SL approach for TCE
In 2014, Defra in the UK published a report from its Research Project SP1010 Development of Category
4 Screening Levels for Assessment of Land Affected by Contamination (accessed at
http://randd.defra.gov.uk/Default.aspx?Module=More&Location=None&ProjectID=18341 on 27 April
2015). It was accompanied by a Defra Policy Companion Document dated December 2014, with the
intent that the resulting C4SLs should be used in the context of Defras April 2012 revised statutory
guidance for Part 2A under the Environmental Protection Act 1990 (England & Wales), to help with
determining that land is low risk in Category 4 and therefore is definitely not contaminated in the legal
sense. The research report and policy companion document included the derivation of six C4SLs for
arsenic, cadmium, chromium VI, benzene, benzo[a]pyrene and lead; trichloroethylene (TCE) was not
included.
Here, we follow the harmonized risk assessment principles established in the toxicological framework of
the Defra SP1010 project (Figure 1) and apply them to TCE. The ultimate aim of the framework is to
define a health-based guidance value in the form of a low level of toxicological concern (LLTC), which is a
health-based guidance value that can be input into the Contaminated Land Exposure Assessment
(CLEA) model version 1.06 as an intake of low concern, and coupled with exposure can be shown to
lead to a scenario of low risk. It is important to state that the principles of a minimal risk Health Criteria
Value (HCV) defined within Environment Agency (England & Wales) SR2 and SR3 guidance from 2009,
remain consistent with this new framework illustrated in Figure 1. The framework approach starts with an
overview of the authoritative toxicology data reviews for TCE.
The Environment Agency (England & Wales) last produced a soil guideline value (SGV) for TCE in 2004,
and this was based on a minimal risk HCV of 5.2 g/kg/day (oral and inhalation intakes). Since this time,
new toxicology reviews have been published (which calls for a significant reduction in health-based
guidance values for TCE) that change the conclusions of the (now withdrawn) SGV.
In 2011, the US Environmental Protection Agency (EPA) performed a lengthy and comprehensive review
of TCE toxicology, and proposed candidate reference doses (RfDs) based around three pivotal studies
(as outlined in Table 1). It is critical to understand the underpinning basis of these RfDs and what health
effects can be caused by TCE before choosing to use these RfDs in EU contaminated land risk
assessments, in order to evaluate the level of current concern and know the acceptance of risk these
numbers from US EPA represent.
In 2013, the US Agency for Toxicity and Disease Substance Registry (ATSDR) published an addendum
to its toxicity profile for TCE published in 1997. This was largely a reiteration of the US EPA preferred RfD
value of 0.51 g/kg/day as a Minimal Risk Level (MRL).
In 2014, both the US National Toxicology Program (NTP) and IPCS reviewed new evidence on
carcinogenicity and NTP concluded that the current listing in the Report on Carcinogens for TCE as
reasonably anticipated to be a human carcinogen should be changed to known to be a human
carcinogen based on sufficient evidence of carcinogenicity from studies in humans regarding kidney, liver
and non-Hodgkin lymphoma. TCE is also genotoxic in vitro and in vivo, which means it should be
considered as a non-thresholded carcinogen, even though its metabolites are likely to be causative of
carcinogenicity.
There is at present no review or critique of the new US toxicology evaluation in the UK. The US EPA
values are beginning to be used in contaminated land evaluations (e.g. the Suitable 4 Use levels, S4ULs).
However, the US EPA values would not be considered as minimal risk values in UK contaminated land
practice, nor may they be considered as low risk as per Defra guidance.
Figure 1 A Toxicological Framework to define a Low Level of Toxicological Concern (LLTC) for land contaminants as used
in the context of implementing the revised statutory Part 2A guidance of the Environmental Protection Act 1990 (England
and Wales) for developing C4SLs. (Reproduced from the Defra Project SP1010 report).
Table 1 Pivotal non-cancer toxicology data as a basis for candidate oral reference doses (RfDs) in
the US EPA review.
POD
Ref Dose
Health Effect value POD type UF Pivotal study
mg/kg/day
mg/kg/day
Human
Equivalent Dose 100
(99th percentile) (10 for use of a LOAEL;
based on a 3.16 as a PBPK model Keil et al (2009)
Thymus
LOAEL (0.35 was used for interspecies 0.00048 30 week drinking
decreased 0.048 mg/kg/day) extrapolation; mg/kg/day water study in
weight expressed as an 3.16 as PBPK model was B6C3F1 mice
internal dose used for interindividual
(0.139 variability)
mg/kg/day)
Peden-Adams et al
Immunotoxicity 1000 (2006) from
delayed type (10 for use of a LOAEL; 0.00037 placental,
0.37 LOAEL
hypersensitivity 10 for interspecies and 10 mg/kg/day lactational and
in pups for intrahuman variability) drinking water
exposure.
Human
Equivalent Dose
(99th percentile) 10
based on a (3.16 as a PBPK model Johnson et al
BMD1% (0.065 was used for interspecies (2003) Rat
Fetal heart 0.00051
0.0051 mg/kg/day; extrapolation; developmental
malformations converted to 3.16 as PBPK model was
mg/kg/day
study drinking
0.0142 mg TCE used for interindividual water
oxidized /kg/day variability)
as expressed as
an internal dose)
All of the most sensitive and serious non-cancer (development) endpoints are based upon animal studies;
cancer endpoints are kidney, liver and non-Hodgkins lymphoma, for which there are human data. The
preferred oral RfD in the US EPA evaluation, as also adopted by the ATSDR in 2013, is that derived in
relation to fetal heart malformations as the most sensitive endpoint in rats (Johnson et al (2003). PBPK
modelling has been used to predict internal human relevant doses from the rat drinking water study.
Cancer risk was also reviewed in the US EPA evaluation using the human inhalation data in Charbotel et
al (2006) on incidence of kidney cancer with TCE exposure; for inhalation the unit risk is 4 x 10-6 per
g/m3 and the oral slope factor, resulting from PBPK modelling of route-to-route extrapolation from the
same inhalation study, is 5 x 10-2 per mg/kg/day. The total unit risk of 2 x10-6 from all cancers (kidney,
NHL and liver, with consideration of early life susceptibility) from drinking 1gTCE/L water is calculated in
US EPA 2011 page 5-162; similarly, it is 4.8 x 10-6 per g/m3 air in US EPA 2011 page 5-159.
Below, we interpret this new information in the context of the toxicological framework in Figure 1 and
apply the guidance from Defra Project SP1010, particularly recommendations from the Defra policy
companion document around the choices of margins and cancer risk (not considered in the three US EPA
RfDs directly) that can be considered in deriving an LLTC using these data.
Tables 2 and 3 capture the options as to how an oral and inhalation LLTC respectively, can be derived for
TCE for use in contaminated land risk assessment. The key endpoints chosen are the most sensitive
(and potentially severe/fatal) non-cancer endpoint of fetal heart malformations (FHM)(Johnson et al,
2003) and human total cancer unit risk values of 2 x 10-6 per g/L drinking water and 5 x 10-6 per g/m3
air, both accounting for early life susceptibilities and lifetime exposure (US EPA 2011 evaluation).
Table 2 Results of using the Toxicological Framework to interpret the data and provide evidence
such that an appropriate Oral LLTC can be defined for a trichloroethylene C4SL
Potential Basis for LLTC Value Unit Comments
2004 Minimal Risk Health Criteria Value 5.2 g/kg Index dose mouse inhalation study dose-
(Environment Agency) now withdrawn as BW/day independent cancer endpoints. ELCR 1 in
no longer defensible 100,000. No longer valid.
Using Johnson et al 2003 - based on 0.65 g/kg Applied dose with no PBPK modelling Based
BMD01 (also similar to provisional WHO BW/day upon a BMD01 for Fetal heart malformations in
EU 2011 4th Ed drinking water guideline 20 rat.
g/L) with margin of 100 Incidence FHM = 1 in 10000.
Human cancer risk = 1 in 22,000
US EPA RfD 0.5 g/kg BMD01 but translated to HED99 using PBPK
Using BMD01 Johnson et al (2003) as BW/day modelling.
most sensitive endpoint - fetal heart N.B. This intake is 130-fold lower than the
malformations with internal human BMD01 for FHM not considering PBPK
equivalent dose using PBPK modelling modelling.
with a margin of 10. Human cancer risk = 1 in 28,600
Current European drinking water standard 10 g/L Based on unit risk of 2 x 10-6 per gTCE/L
http://dwi.defra.gov.uk/consumers/advice- (USEPA 2011 total risk of all cancer, including
leaflets/standards.pdf early life susceptibility & lifetime exposure)
Oral Cancer Risk at 10 g/L is
1 in 50000
0.3 g/kg (10 g x 2L per day)/70 kg
BW/day
Oral LLTC Based upon a Cancer Risk of 0.3 g/kg Protective of a sensitive developmental
1 in 50000 BW/day immunotox endpoint (RfD 0.37 g/kg/day)
(also equates to the current UK drinking Protective of fetal heart malformations: the
water standard which avoids LLTC is 216-fold lower than the BMD01 of
disproportionately targeting soil) and is 0.065 mg/kg/day. Approx notional incidence
suitably protective of non-cancer rate of 1 in 22000.
endpoints.
Oral LLTC driven by FHM, based on 0.14 g/kg i.e. BMD10 = 0.7 mg/kg/day divided by a margin
BMD10 and a margin of 5000 BW/day of 5000
(i.e. as per policy companion for non- Incidence FHM = 1 in 50000
thresholded carcinogens, use similar for Cancer risk = 1 in 102000 (minimal risk)
this severe developmental endpoint)
Considerations of minimal risk for FHM
Using Johnson et al 2003 - based on 0.07 g/kg Applied dose BMD10 = 0.7 mg/kg/day divided by
BMD10 (UF 10000) BW/day a margin of 10000 - with no PBPK modelling
Using Johnson et al 2003 - based on 0.02 g/kg Applied dose BMDL01 = 0.02mg/kg/day
BMDL01 (95th %ile) and margin of 1000 BW/day divided by a margin of 1000 with no PBPK
modelling
Minimal Risk HCV using Johnson et al 0.02 g/kg Applied dose BMDL10 = 0.23 mg/kg/day
2003 for the most sensitive endpoint. BW/day divided by an uncertainty factor of 10000
BMDL10 (UF 10000) as per EA SR2
guidance
From the above evidence in Table 2, it is suggested to set an Oral LLTC at 0.3 g/kg BW/day for use in
developing a C4SL for TCE. This value is:
a) Equivalent to an excess lifetime cancer risk via the oral route of 1 in 50000, as based upon an
evaluation of human study data (which is in accordance with Defras policy companion relating to
a low risk ELCR).
b) Is lower than the RfD (0.37 g/kg BW/day) for developmental immunotoxicology endpoints.
c) Represents low risk of fetal heart malformations, a developmental endpoint, with this intake being
216-fold lower than a highly sensitive point of departure (BMD01).
d) Equates to the intake of the current drinking water standard
Table 3 Results of using the Toxicological Framework to interpret the data and provide evidence
such that an appropriate Inhalation LLTC can be defined for a trichloroethylene C4SL
Potential Basis for LLTC Value Unit Comments
2004 Minimal Risk Health Criteria Value 5.2 g/kg Index dose mouse inhalation study dose-
(Environment Agency) now withdrawn as BW/day independent cancer endpoints. ELCR 1 in
no longer defensible 100,000. No longer valid.
Based on Cancer Risk of 1 in 50000 1.1 g/kg Assuming a 70kg person inhales 20m3 air per
[Using 2011 US EPA evaluation] BW/day day, containing 4 gTCE/m3
4 g/m3 Cancer risk = 1 in 50000, based on a unit risk of
5 x 10-6 (human study Charbotel et al 2006).
US EPA RfC 0.57 g/kg Assuming a 70kg person inhales 20m3 air per
BW/day day, containing 2 gTCE/m3
The basis is cited as the non-cancer effects
(as per Table 1 above): fetal heart 2 g/m3 The dose in g/kg BW/day is based upon route
malformations, adult and developmental to route PBPK extrapolation from oral non
immunotoxicity. PBPK modelling has been cancer effects (drinking water studies)
used to perform route-to-route Cancer effects: Kidney tumours; NHL, and liver
extrapolation for these three most sensitive cancers are based on human inhalation
endpoints originally from drinking water exposure data.
study data in rodents. Cancer risk = 1 in 100000, based on a unit risk
of 5 x 10-6 (human study Charbotel et al 2006).
Recommended Inhalation LLTC (at the 0.3 g/kg Non-cancer toxicity drives the risk: assumes the
current time) equate with the oral LLTC, BW/day toxicological consequences of absorption
as driven by the most sensitive non-cancer through the gut and lungs is the same, as per
endpoints. Remains suitably previous EA 2004 guidance. Factors in no
precautionary as per Defra guidance. PBPK extrapolation or differential route-specific
metabolism at the moment these assumptions
carry high uncertainty and methods have not
been fully considered and reviewed at National
level in the UK.
From the above evidence in Table 3, it is suggested to set an Inhalation LLTC at 0.3 g/kg BW/day for
use in developing a C4SL for TCE. This value is:
a) Equivalent to the oral LLTC from Table 2 above; making a current assumption that the inhalation
route kinetics are the same as the oral route kinetics leading to the same non-cancer effects. This
is a suitably precautionary position for current regulatory use in the UK, as PBPK modelling is
used in multiple ways in the US EPA assessment to perform route-to-route extrapolations,
internal human dose estimates and incorporating oxidative metabolism assumptions.
b) Protective of all cancer effects by the inhalation route; ELCR is calculated at 1 in 200000 at this
intake dose.
From the above analysis, it is proposed that both the Oral and Inhalation LLTC value for trichloroethylene
is set at a suitably precautionary 0.3 g/kg BW/day. These LLTC values would be lower than the RfD
(0.5 g/kg BW/day) and RfC (2g/m3; equivalent to 0.57 g/kg BW/day) proposed by the US EPA in 2011
following their comprehensive review of new toxicological data, but protective of all non-cancer and
cancer effects.
It should also be noted that the most sensitive effects (fetal heart malformations and immunotoxicity) are
developmental in nature and therefore sensitive time windows of exposure (i.e. during gestation) requires
careful consideration in the exposure modelling side of a risk-based evaluation, where the C4SL should
be derived for pregnant women and the developing fetus, as the most sensitive human receptors
potentially exposed to TCE.
PBPK modelling techniques and metabolism studies are useful in refining risk assessment across species
and performing route-to-route extrapolations. PBPK modelling and assumptions around oxidative
metabolism have been used heavily in the US EPA assessment. Measured data to parameterize the
models is not comprehensive, and it is possible that there is a high degree of uncertainty in the modelling
underpinning the US EPA values. It is important before using the EPA values directly in contaminated
land risk assessment, that there is a review and discussion in UK/EU about the parameterisation of the
PK models used, a critique of the underlying assumptions made and data on metabolism reviewed in
order that there is full transparency about the uncertainties in the US EPA evaluation.
Given these specific details for TCE modelling, and indeed the general use of PBPK modelling and
metabolism data, are yet to be debated and reviewed at national level in the UK for contaminated land
risk assessment purposes, it remains suitably precautionary to set the LLTCs for TCE at values slightly
lower than the US EPA reference doses until such a review is performed.
In deriving the C4SLs using these LLTCs for the various exposure scenarios and considering the risks of
developmental effects, the most sensitive human receptors should be the pregnant woman and
developing foetus during the sensitive window of gestational exposure; for cancer effects the receptor is
the standard residential child and commercial adult worker.
References
ATSDR. 1997. Toxicological Profile for Trichloroethylene. Atlanta, GA: Agency for Toxic Substances and
Disease Registry. 335 pp.
ATSDR. 2013. Addendum to the Toxicological Profile for Trichloroethylene. Atlanta, GA: Agency for Toxic
Substances and Disease Registry. 120 pp.
Charbotel, B; Fevotte, J; Hours, M; Martin, J-L; Bergeret, A. (2006). Case-control study on renal cell
cancer and occupational exposure to trichloroethylene. Part II: Epidemiological aspects. Ann Occup Hyg
50: 777-787.
Environment Agency (England & Wales) 2004 Contaminants in Soil: Collation of Toxicological Data and
Intake Values for Humans. Trichloroethene. Science Report Tox 24.
EPA. 2011a. Toxicological Review of Trichloroethylene (CAS No. 79-01-6) In Support of Summary
Information on the Integrated Risk Information System (IRIS). EPA/635/R-09/011F. U.S. Environmental
Protection Agency. 1200 pp.
EPA. 2011b. Toxicological Review of Trichloroethylene Appendices (CAS No. 79-01-6) In Support of
Summary Information on the Integrated Risk Information System (IRIS). EPA/635/R-09/011F. U.S.
Environmental Protection Agency. 1269 pp.
IARC. 2014. Trichloroethylene, Tetrachloroethylene and Some Other Chlorinated Agents, IARC
Monographs on the Evaluation of Carcinogenic Risks to Humans. vol. 106, Lyon, France: International
Agency for Research on Cancer.
Johnson PD, Goldberg SY, Mays MZ, et al. 2003. Threshold of trichloroethylene
contamination in maternal drinking waters affecting fetal heart development in the rat. Environ Health
Perspect 111(3):289-292.
NTP. 2011. Trichloroethylene. In Report on Carcinogens. 12th ed. Research Triangle Park, NC: National
Toxicology Program. pp. 420-423.
Contact person
Paul Bardos, Information Manager, NanoRem Project (www.nanorem.eu)
r3 environmental technology ltd, H9, TOB1, Whiteknights, Reading, RG6 6AT, UK
(www.r3environmental.com; paul@r3environmental.co.uk), paul@r3environmental.co.uk
Introduction
This is the second of two free sessions intended to provide delegates with sufficient infor-
mation to decide if nanoremediation and new techniques in nanoremediation is a viable ac-
tivity for their organisation. It is particularly targeted at practitioners such as site own-
ers/managers service providers (consultants contractors), and regulators.
Since early 2014, the EU FP7 NanoRem project (www.nanorem.eu) has been carrying out
an intensive development and optimisation programme for different nanoparticles (NPs),
along with analysis and testing methods, investigations of fate and transport of the NPs and
their environmental impact. NanoRem is a 14 million international collaborative project with
28 Partners from 12 EU countries, and an international Project Advisory Group (PAG)
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
providing linkages to the USA and Asia. It is a major initiative, which will support the effec-
tive deployment of nanoremediation technologies in Europe. As part of its work it offers
these two sessions to provide key information for organisations considering diversifying into
nanoremediation, or already beginning that process.
The first free session focuses on providing a practical grounding in nanoremediation theory
and practice with particular reference to applied examples in the field.
The second free session focuses on providing business and strategic intelligence for dele-
gates with interests in using nanoremediation at their sites or developing nanoremediation
activities at their organisations.
Stephan Bartke, UFZ, Germany: What will drive the EU nanoremediation market till
2025 assessment of factors determining opportunities and challenges for the take-
up of nanoremediation (15 minutes).
This presentation will review stakeholder meeting and focus group information collected by
NanoRem over 2014/2015 on factors determing market developments. The intention is to
provide an informative but not prescriptive talk. Part of the strategic and business intelli-
gence provided will be from self-learning as the session delegates interact with each other in
a structured way in small groups.
Various external determinants from economy, technology development, politics and society
affect the property market in general and the industry for contaminated land remediation in
particular. In order to develop an exploitation strategy that considers the medium to longer
term potential market development for nZVI, any analysis has to deal with an uncertain fu-
ture. The factors (i.e. drivers and uncertainties related to driver development) that foster or
inhibit the evolution of the market need to be better understood. It is unclear how the factors
likely to influence the nanoremediation market development are linked, and how they are
likely to develop in the future. It is challenging, therefore, to make any straightforward predic-
tions regarding the emerging nanoremediation market. As a result, traditional supply and
demand modelling is unsuitable. A simple use of neo-classic economics theory will fail to
reveal and quantify the role that nano-remediation options could unfold under different future
market conditions, because there are severe caveats regarding standard assumptions made
in such traditional market assessments. A scenario approach will therefore be used to help
forecast potential market developments. The outcomes are utilisable for: real-world busi-
ness development, deducing strategies for market activities; informing policy development,
identifying governance options for market expansion; and/or informing regulatory authorities,
highlighting the potential for nanoremediation.
(i) What the factors (drivers and uncertainties) are in the nanoremediation market-
system.
(ii) What the extent of the factors impacts is.
(iii) How the factors interdepend.
Scenario design and analysis differ, but usually a stepwise approach is taken. In NanoRem,
the following procedure was selected:
1) Conducting a present situation analysis to establish the baseline for scenario devel-
opment and a framework for factor identification.
2) Filtering and systematising factors that drive or inhibit market development. Estab-
lishing key determinants.
3) Projection of key factors developments and producing consistent stories about ways
the system might evolve in the future. Identification of multiple alternative develop-
ment trajectories is possible.
4) Deciding on planning-oriented exploitation strategies, which may be responsive or
proactive. Development of governance recommendations.
Figure 1 below, gives an overview of the the work conducted so far as part of a scenario
approach, and ongoing work. The steps are discussed further in the following sections. The
overall approach is discussed in more detail in Bardos et al. 2015 and the workshop is re-
ported in further detail in Tomkiv et al. 2015
-
-
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
To fulfil step one of the scenario development approach, a baseline understanding of the
nanoremediation market and the set of factors with the potential to influence the future de-
velopment of the nanoremediation market was established. This was achieved via key-
informant interviews and literature analysis, taking into account the market status quo and
risk-benefit appraisal (cf. Bardos et al. 2014). This preliminary research helped establish a
variety of external determinants from economy, technology development, politics and society
that may affect:
The property market in general;
The industry for contaminated land remediation broadly, and;
The potential evolution of nanoremediation in particular.
Expert engagement (key informant interviews and expert discussion) was utilised to establish
the most worthwhile timeframe for the scenario approach. A consensus was reached that
evolution of the market up to 2025 was the most appropriate scope. It was felt that a very
long-term assessment would be impossible due to the significance of unknown and uncertain
factors. After several iterations with expert involvement, a condensed list of 22 potentially
influencing factors was established.
To aid step two of the scenario design process, a Market Opportunities session was includ-
ed in the Sustainability and Markets workshop (cf. Tomkiv et al. 2015). The Oslo event was a
participatory approach to conclude on the systematic interdependencies and importance of
the factors in order to capture through engagement with the participants with their different
backgrounds the high diversity of drivers and inhibitors affecting the market system.
The 22 factors determined in the preliminary research stage were grouped into different cat-
egories (policy, economy, society, communication, technology and megatrends). The use of
categories helped to align the factors with appropriate expertise for later discussions. In order
to further condense the list of factors and remove less important factors, the list was sent to
the workshop participants in advance of the workshop. Participants were asked to provide
feedback on how important they perceived each factor to be for the development of the EU
nanoremediation market from present to 2025. Participants scored each factor according to
the following scale:
(0) = Negligible relevance the factor is not an important driver or inhibitor;
(1) = Minor relevance the factor might have a limited but not so important effect;
(2) = Considerable relevance the factor is likely to have a notable (indirect) effect;
(3) = Key relevance this factor is most certainly among those of utmost importance to
push or pull the nanoremediation market development.
The responses (20 respondents) were collated and an average score (the arithmetic mean
as the sum of the scores collected from all the respondents, divided by the number of the
respondents) was calculated for each factor. The results are shown in Table 1
below, in descending order of obtained scores.
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
Table 1 Preliminary factors and their perceived importance with regards to influenc-
ing nanoremediation market development in the EU up to 2025
The scorings indicate that several factors influence the markets development. Some of the
scorings, e.g. the ability to treat emerging contaminants with NPs, are surprising and may
indicate either bias or epistemic issues in the mind of the responders. As no factor had a
scoring > 2.50, it was concluded that no factor is likely to singlehandedly push or pull
nanoremediation market development.
In order to create scenarios, the interdependencies of the factors determined to be important
needed to be better understood. Stakeholders were provided with the factors in Table 1
including short descriptions of each factor. During the workshop, stakeholders were asked to
provide opinions, comments and suggestions about the factors and were also asked to iden-
tify and discuss the interrelations of the factors. In order to do this, stakeholders were divided
into smaller groups based on their field of expertise. The groups formed were Regulators /
Policy makers, Technology, Communication, Economy and Society. Participants in the re-
spective groups were asked to discuss the influence of three or four factors of their respec-
tive expert domain on the full list of factors identified to be of importance.
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
For each group a poster with an empty influence matrix was provided showing a short list of
factors from the respective field of a groups expertise in the rows on the vertical axis and the
full list of factors in the columns on the horizontal axis. Figure 2 illustrates the influence ma-
trixs outline.
The participants in a first phase were asked to review and provide opinions, comments and
suggestions about the collected factors in the rows. Next, they were asked to identify and
discuss the interrelations of the development of each of their factors in the lines on the full list
of factors, i.e. to discuss pairwise the influence of development of a factor from the vertical
axis on the development of a factor from the horizontal axis. For the assessment, again a
scoring was re-quested (as indicated in Figure 2, too): Considering the European Union in
2025, the impact of the development of the factor in the row on the development of the factor
in each column was gauged using the scale:
(0) = No impact;
(1) = Weak / delayed impact;
(2) = Medium impact;
(3) = Strong / direct impact.
At the end of this phase, each group had filled in their part of the influence matrix. After the
work-shop it was possible to add up the rows and to obtain a full matrix containing all factors
in the rows and columns however, this was not done at the Oslo event, as the key objective
here was to best substantiate the assessments within the matrix. Therefore, in a subsequent
phase of the event, taking up the World CafTM format, the participants were invited to dis-
cuss one by one the results of the other groups. Facilitators guided these discussions from
the identification to the review of the linkages of factors. At the end of the phase, the partici-
pant returned to their home table and revisited their assessments based on the feedbacks
collected from the other groups. At the end of the Session, the annotated posters and notes
of facilitators were collected and interpreted.
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
After the workshop, the information collected from the group sessions was analysed and the
factors that are more active in influencing other factors were identified, as well as those that
are more driven by the active ones. These relationships are expressed by the active sum
and passive sum as indicated in Figure 2 above and in Table 2, below. Table 2 lists the
five most active factors from those recorded in order of their activity (i.e. how influential a
factor is relative to other factors).
These dimensions form the conceptual scheme for the scenario states of the nanoremedia-
tion market. These scenario states show four potential future states for the market, see Fig-
ure 3 below, (going clock-wise in each quadrant of the matrix):
I. Validated information is broadly available AND there is comprehensive dialogue be-
tween stakeholders, in particular those from science, policy and regulation.
II. Validated information is lacking and uncertainty is still significant BUT there is com-
prehensive dialogue between stakeholders, in particular those from science, policy
and regulation.
III. Validated information is lacking and uncertainty is still significant AND there is no or
only minimum dialogue between stakeholders, in particular those from science, policy
and regulation.
NanoRem Project Nr.: 309517 AquaConSoil 2015, Copenhagen, Special session 1C.24S
IV. Validated information is broadly available BUT there is no or only minimum dialogue
between stakeholders, in particular those from science, policy and regulation.
The final steps of the scenario analysis will be the discussion of the key factors develop-
ments within the different scenario states, and based on these, the deduction of scenario
storylines. These discussions will finally inform the conclusion of recommendations for the
exploitation strategy.
The Scenario approach will continue to be expanded, with scenario development continuing
over Year 3 of the NanoRem project. Further expert engagement in focus groups and work-
shops has been planned in order to discuss the scenario scheme and future market states
that reflect the interactions identified in the Oslo workshop and specify directions of factor
development under the potential market states. The scenario storylines concluded from
these activities will be used to derive conclusions on the medium to longer term exploitation
opportunities and recommendations for entrepreneurs and policymakers.
Nicola Harries, CL:AIRE, UK; Elsa Limasset, BRGM, France; Stephan Bartke, UFZ,
Germany; Paul Bardos, r3 environmental technology ltd, UK. Discussion of possible
market trends and opportunities (60 minutes)
Discussion in groups, each with a NanoRem facilitator to provide a chance for active discus-
sion and exchange of ideas about market prospects, drivers and recommendations we can
conclude on the day. A short report of findings will be provided on
http://www.nanorem.eu/news.aspx by the end of August 2015.
Paul Bardos, r3 environmental technology ltd, UK. Wrap Up and Clinic Offer (10
minutes)
Note: Delegates will be provided with a NanoRem web link for take home materials from
special session presentations that they can use within their own organisations to support
further decision-making
References:
ABSTRACT: Present study investigates whether remediation through in-situ metal bioprecipitation is
attainable in a high flow zinc-contaminated aquifer at a galvanizing company in Maasmechelen,
Belgium. Both laboratory feasibility tests and a double long-term field pilot test were conducted. The
laboratory microcosm tests showed that more than 99% of the dissolved zinc was removed from the
water phase after the addition of an organic substrate (i.e. sodium lactate, glycerol or a commercial
emulsified vegetable oil) and sulfate. The three different organic substrates proved equally effective. A
4-week anaerobic leaching test indicated that the formed zinc precipitates are stable. In addition,
substrate addition seemed to result in a higher leachability of arsenic and manganese. During the 232
day testing period, the field pilot tests showed a concentration decrease for dissolved zinc of 2 to 3
orders of magnitude. During the field test, there was no indication of arsenic mobilisation whereas
concentrations of dissolved manganese increased significantly. Only a limited rebound of zinc
concentrations could be detected 2.5 years after the last substrate addition.
INTRODUCTION
Throughout the years, experience showed that remediation techniques like Pump & Treat (P&T)
seemed less suitable for metal contaminated aquifers due to high sorption coefficients. In order to
obtain a significant mass removal through P&T, a considerable volume of groundwater should be
extracted resulting in high energy demands over a long period of time. One of the more modern
approaches for the remediation of metal contaminated aquifers is in-situ metal bioprecipitation (ISBP)
(Sierra-Alvarez et al., 2006; Vanbroekhoven et al., 2011). ISBP relies on the widespread natural
presence of sulfate reducing bacteria (SRB) in temperate soil and aquifer environments (Muyzer and
Stams, 2008). SRB reduce sulfate to sulfide by consuming an organic substrate and/or H2 as electron
donors. The formed sulfide reacts with dissolved divalent metal ions (i.e. zinc) resulting in poorly
soluble metal sulfide precipitates (i.e. zinc sulfide or sphalerite).
The sulfate reduction rate is subject to different parameters like type of organic substrate, redox
potential, pH, etc. At low sulfate concentrations and/or high availability of organic substrate,
competition between methanogenic bacteria and SRB may be problematic (Annachatre and
Suktrakoolvait, 2001). The long-term stability of the zinc sulfide is still one of the uncertain factors in
applying ISBP. If subsurface conditions remain at neutral pH and moderately reduced, ISBP can offer
a long-term sustainable solution.
Site Description. The site is located in Maasmechelen, Belgium. Throughout the years, the metal-
processing activities caused a relatively large groundwater pollution plume with zinc (more than 200 m
long and 70 m wide).
Due to the proximity of the River Meuse, the subsoil consists mainly of heterogeneously layered
Pleistocenic coarse sand and gravel, resulting in a high hydraulic conductivity (app. 45 m/day) and a
high groundwater velocity (0.2 - 1 m/day). The groundwater level at the site is between 3 and 4 m bgl
(below ground level). Zinc concentrations in the plume vary from 1.000 to 100.000 g/L where the
legal Flemish limit concentration is 500 g/L. The local groundwater is relatively oxidized (1 - 3 mg/L),
naturally low in DOC (< 1 mg/L) and relatively low in sulfate (40 50 mg/L).
MATERIALS AND METHODS
Laboratory Tests. Serum bottles (160 mL) with Teflon-coated rubber septa and crimp caps were
used for test batches. These bottles were filled in an anaerobic chamber with 30 g of aquifer material
and 100 mL of local groundwater. A quadruple microcosm test was set up, consisting of an abiotic
control (adding 0.5% formaldehyde) and 3 different types of organic substrate, being:
At the beginning of the microcosm test, 0.02 %C organic substrate (0.04 %C for EOS) was added to
the groundwater. After 45 days, another dose of substrate (0.02 %C), sulfate (300 mg/L) and zinc (9
mg/L) was added to the microcosms. All batches were incubated at ambient temperature. After 45
days, water samples were taken and analysed for pH, redox potential and zinc. Three months after the
start of the incubation, the same analyses were done, elaborated with analyses for iron, sulfate and
volatile fatty acids.
The stability of the formed zinc precipitates was investigated through batch extraction tests under
anaerobic conditions. After separation of the aquifer material through decantation and filtration, all
samples were equilibrated to non-polluted local groundwater in three successive steps:
After each equilibration period, the water phase was filtered (0.45 m Millipore membrane filtration)
before adding nitric acid and analyzing for zinc.
Field Tests. Within the contamination plume, two pilot zones were installed, each one consisting of 3
injection wells (HDPE, 50 mm, screened from 4 7 m bgl) and 8 monitoring wells (HDPE, 50 mm,
screened from 4 6 m bgl). Figure 1 shows the general set-up of both pilot zones.
Monitoring well
Injection well
Groundwater flow direction
In September 2011, the field test started with a reference monitoring. The first substrate injection
th
event was executed in both pilot zones on October 17 2011. In the first testing zone, all three
injection wells were gravitationally injected with 2.0 m of 5 vol% glycerol solution containing 45 g/L
potassium sulfate. Each injection well was then flushed with 1.0 m locally collected non-polluted
groundwater. In the second testing zone, all three injection wells were gravitationally injected with 53
kg of EOS diluted in water to a total volume of 265 L. Alternately, all three injection wells in this zone
were injected with 1.0 m of a 45 g/L potassium sulfate solution, followed by flushing with 1.0 m
locally collected non-polluted groundwater. A second glycerol injection in zone 1 took place on
nd
February 2 2012. All injection wells were injected with a 1.0 m 10 vol% glycerol solution containing
70 g/L potassium sulfate. Each injection well was then flushed with 1.0 m locally collected non-
polluted groundwater. In the second testing zone, no additional substrate but only additional sulfate
was added on this date (1.0 m locally collected non-polluted groundwater with added potassium
sulfate (70 g/L) was injected in each injection well, followed by flushing with 0.5 m locally collected
th
non-polluted groundwater). On April 11 2012, a third and final substrate injection event took place. In
the first testing zone, all three injection wells were injected with 2.35 m of a 6 vol% glycerol solution
containing 71 g/L potassium sulfate, followed by flushing with 1.0 m locally collected non-polluted
groundwater. In the second testing zone, all three injection wells were injected with 53 kg of EOS
diluted in water to a total volume of 265 L. Alternately, all three injection wells in this zone were
injected with 2.0 m of a 83.5 g/L potassium sulfate solution. After each injection, wells were flushed
with 1.0 m locally collected non-polluted groundwater.
Within the total pilot test duration (232 days), 9 groundwater monitoring rounds were done to evaluate
the effects of the injections. To examine potential rebound of zinc concentrations, one additional
monitoring round was performed 6 months after the last substrate addition, bringing the total pilot
duration to 393 days. Groundwater samples were collected from the monitoring wells, after purging,
using a peristaltic pump (Eijkelkamp Agrisearch Equipment). Temperature, electrical conductivity (EC),
pH and redox measurements (Eh) were performed in the field. Other measured parameters such as
heavy metals, nitrate, sulfate, methane, dissolved organic carbon (DOC), iron and manganese were
performed by commercial certified laboratories. In each pilot zone, 2 monitoring wells were installed
hosting mesocosm socks (2 socks per well). These mesocosms were made by filling nylon filtercloth
bags with approx. 100 g of sieved (< 2 mm) local aquifer material. All mesocosms were installed at
app. 5 m bgl. Per well, one mesocosm sock was collected 100 days after the first injection event and
the other mesocosm 200 days after the first injection event. The aquifer material from the socks was
analyzed for zinc, iron and manganese. Batch anaerobic leaching tests were performed similar to
those done during the laboratory test (see above) to investigate the stability of the formed zinc
precipitates.
Laboratory Tests. Microcosm tests showed decreasing redox potentials and the formation of zinc
precipitates, removing more than 99% of zinc initially present in the water phase. The three tested
organic substrates showed equal effectiveness. After three months of incubation, the added sulfate in
the biotic batches was reduced to low concentrations. The abiotic controls showed no evidence of
bioprecipitation, showing elevated zinc concentrations in the water phase after three months of
incubation.
The results of the anaerobic leaching tests showed that the formed precipitates are stable under
tested conditions, especially in case of the tested substrates glycerol and EOS. The same tests
showed increasing solubility (leachability) of arsenic and manganese due to substrate addition and
following microbial reduction of ferric and manganese oxides.
th
Field Tests. During the first injection-event on October 17 2011, the groundwater EC monitoring
showed a fast and good distribution of the injected products (organic substrate and potassium sulfate)
in several monitoring wells. Figure 2 shows the increase of EC during the time of injection, indicated
by the red frame. These observations indicate a high response rate, due to the aquifer characteristics
mentioned above and give a first idea of the injections zone of influence, estimated at approx. 7.5 m.
During 232 days of field testing, zinc concentrations were reduced by 2 to 3 orders of magnitude in
both testing zones as a result of bioprecipitation following the organic substrate and sulfate injections.
In both pilot zones, an overall zinc concentration reduction of 96 97 % was observed. A general view
on zinc concentrations for individual monitoring wells over time is showed in figure 3.
Mesocosm socks soil analyses showed no significant effects of the injections for the analyzed
parameters iron, manganese and zinc. These observations led to the assumption that a substantial
decrease in water phase zinc concentration through bioprecipitation only yields a small increase of
solid phase zinc concentration, making the process hard to detect by mesocosm monitoring or aquifer
sampling. Leaching test results demonstrated that zinc mass leached from mesocosm soil was higher
than leaching from laboratory microcosm samples, but lower than the control soil.
Zone 1 / Injection event 1
Injection IF1, 2 en 3
12000
10000
,
P2
8000
M1
EC (S/cm)
M2
6000
M3
M4
4000
M5
2000
0
8:10 9:30 10:30 11:45 12:45 13:45
Time
20000
k
M6
PB302
EC (S/cm)
15000
M7
M8
10000
M9
M10
5000
0
10:30 13:30 14:30 15:30
Time
Neither one of the test zones showed arsenic mobilization. By formation of iron and zinc sulfides,
arsenic may have been coprecipitated (Labrenz et al., 2000). Field observations for manganese were
consistent with the laboratory microcosm results. As a result of the organic substrate injection,
manganese groundwater concentrations increased from 0.01 0.06 mg/L to 0.4 6.5 mg/L.
In order to examine the potential rebound of zinc concentrations (redissolution of ZnS precipitates),
additional monitoring rounds were performed in October 2012, July 2013 and November 2014,
respectively 6, 15 and 31 months after the last substrate addition. Based on these monitoring results,
aerobic conditions seem to be reinstalled approximately 1 year after the last substrate injection,
resulting in a limited rebound of soluble zinc concentrations. The rebound can be partially explained by
fresh polluted water entering both test zones.
To obtain detailed information about the potential rebound, groundwater concentrations will be
regularly monitored until November 2018, being 6,5 years after the latest substrate injection. At the
time of the conference, monitoring results for a period of 3 years after the last substrate addition will
be presented.
[Zn]
[Zn]
0
5000
10000
15000
20000
25000
30000
35000
40000
0
10000
20000
30000
40000
50000
60000
70000
80000
90000
100000
27/09/2011 27/09/2011
27/11/2011
27/11/2011
27/01/2012
27/01/2012
27/03/2012
27/03/2012
27/05/2012
27/05/2012
27/07/2012
27/07/2012
27/09/2012
27/09/2012
27/11/2012
27/11/2012
27/01/2013
27/01/2013
27/03/2013
27/03/2013
date
27/05/2013
date
27/05/2013
27/07/2013
Zone 2 (EOS)
27/07/2013
Zone 1 (glycerol)
27/09/2013
27/09/2013
27/11/2013
27/11/2013
27/01/2014
27/01/2014
27/03/2014
27/03/2014
27/05/2014
27/05/2014
27/07/2014
27/07/2014
27/09/2014
27/09/2014
M10
P2
PB302
M5
M4
M3
M2
M1
CONCLUSION
The results of present study show that bioprecipitation of zinc is a valuable groundwater remediation
technology, even in highly permeable, poorly reduced aquifers. Both highly soluble organic substrates
(glycerol) as less soluble substrates (EOS) seem equally applicable for this technique. Rebound
effects seem to be limited 3 years after the latest substrate addition. Nonetheless, a continued long-
term monitoring campaign will be necessary to further investigate the impact of recovering aquifer
conditions on the measured zinc concentrations in the water phase.
ACKNOWLEDGMENTS
Funding for this study was partially provided by the SME subsidy granted to VITO by the Flemish
Government. Our gratitude goes out to the concerned VITO and Santerra employees and Antea
Groups field work team.
REFERENCES
Annachhatre, A.P. and Suktrakoolvait, S. 2001. Biological sulfate reduction using molasses as a
carbon source. Water Environ. Res., 73: 118-126.
Labrenz, M., Druschel, G.K., Thomsen-Ebert, T., Gilbert, B., Welch, S.A., Kemmer, K.M., Logan, G.A.,
Summons, R.E., De Stasio, G., Bond, P.L., Lai, B., Kelly, S.D. and Banfield, J.F. 2000. Formation
of sphalerite (ZnS) deposits in natural biofilms of sulfate-reducing bacteria. Science, 290: 1744-
1747.
Muyzer, G. and Stams, A.J.M.. 2008. The ecology and biotechnology of sulfate-reducing bacteria.
Nature Rev. Microbiol., 6: 441-454.
Sierra-Alvarez, R., Karri, S., Freeman, S. and Field, J.A. 2006. Biological treatment of heavy metals in
acid mine drainage using sulfate reducing bioreactors. Wat. Sci. & Technol., 54: 179-185.
Vanbroekhoven, K., Van Roy, S., Diels, L., Gemoets, J, Verkaeren, P., Zeuwts, L., Feyaerts, K. and
Van den Broeck, F. 2008. Sustainable approach for the immobilization of metals in the saturated
zone: in-situ bioprecipitation. Hydrometallurgy, 94: 110-115.
1C.2 - Integrating sustainable remediation into other policies
Bewley, Richard.J.F.
AECOM Infrastructure & Environment UK Limited
4th Floor, Bridgewater House, Whitworth Street, Manchester, M1 6LT, United Kingdom
Tel: +44 (0)161 907 3500
richard.bewley@aecom.com
Parkman, Rick
AECOM Infrastructure & Environment UK Limited
4th Floor, Bridgewater House, Whitworth Street, Manchester, M1 6LT, United Kingdom
Tel: +44 (0)161 907 3500
rick.parkman@urs.com
Bardos, Paul
Environmental Technology Ltd
Room 120, Soil Research Centre, Russell Building, Whiteknights, PO Box 233, Reading, RG6 6DW,
United Kingdom
Tel: +44 (0)118 378 8164
paul@r3environmental.co.uk
Smith, Jonathan.W.N
Shell Global Solutions International B.V.
Lange Kleiweg 40, 2288 GK Rijswijk, The Netherlands
Tel: +31 70 447 3851
jonathan.w.smith@shell.com
Paragraph 1 of the preamble to the Priority Substances Directive 2013, which states:
As a matter of priority, causes of pollution should be identified and emissions of pollutants
should be dealt with at source, in the most economically and environmentally effective
manner.
During the implementation of remedial works engagement with stakeholders and sustainable waste
management practices, including deviation from the waste hierarchy on sustainability grounds, all
have a basis in various UK legislative documentation.
Finally, in relation to verification and monitoring, CIS Guidance Note 17 requires consideration of the
need to reduce the monitoring effort through use of indicator parameters and the cost benefit of
the number of wells versus the level of information that will be obtained.
It is therefore evident that, both at EU level and within the UK framework, there are key legislative
hooks which support, underpin or require sustainability to be embedded throughout the process of
investigating, assessing and remediating contaminated land.
CL:AIRE & NICOLE, 2015. A review of the legal and regulatory basis for sustainable remediation in
the European Union and the United Kingdom. Joint SuRF-UK/NICOLE report. CL:AIRE, London.
Prepared by AECOM.
EU Legislation
Consolidated Version of The Treaty on The Functioning of The European Union, 1957.
Directive 2000/60/EC Establishing a Framework for Community Action in the Field of Water Policy
(The Water Framework Directive)
Directive 2004/35/EC on Environmental Liability with regard to the Prevention and Remedying of
Environmental Damage
Directive 2006/118/EC on the Protection of Groundwater Against Pollution and Deterioration (The
Groundwater Directive)
Directive 2008/98/EC on Waste and Repealing Certain Directives (Waste Framework Directive)
Directive 2010/75/EU on Industrial Emissions (Integrated Pollution Prevention and Control) (Industrial
Emissions Directive)
European Commission, Guidance Note 17 on preventing or limiting direct and indirect inputs in the
context of the Groundwater Directive 2006/118/EC (CIS Guidance Note No. 17)
EA, 2004. Model Procedures for the Management of Land Contamination (CLR 11).
In addition, the ALSAG [1] provides the financial basis of the subsidies - the contaminated waste
contribution. This is levied as a tax on certain contributory activities (eg landfill, incineration). The total
revenue from contaminated waste contributions from 1990 to the end of 2014 amounted to 1.24 billion
Euro (earmarked money for remediation). In addition to the federal funding for site remediation these
financial means are used to finance the compilation and assessment of contaminated sites, as well as
those decontaminations, which the federal government has to carry out itself due to legal regulations.
The UFG [2] lays down the objectives, principles and conduct of environmental subsidies of the federal
government. These include in addition to remediation, urban water management, corporate
environmental support and the purchase of emission reduction units to decrease greenhouse gas
emissions (JI/CDM-Programm). KCP was entrusted with the handling of all these environmental
subsidies as a private partner by the Minister of the Environment.
Other major legal bases that apply at EU level are the General Block Exemption Regulation (GBER) -
Commission Regulation (EU) N651/2014 of 17 June 2014 declaring certain categories of aid
compatible with the internal market in application of Articles 107 and 108 of the Treaty
(GBER, 2014 [5] as well as at a national level, the subsidy guidelines 2015 for remediation of
contaminated sites (FRL 2015) [6].
The guidelines on state aid for environmental protection of the EU ensure conformity of national
funding with the rules of the common market. The current FRL 2015 [6] issued by the Minister of the
Environment on the basis of the UFG [2] regulate the details of the subsidy system, in particular the
volume of subsidies, and are therefore the most important legal basis for use in practice. The FRL
2015 [6] have been notified to the European Commission, thus the subsidies given to companies is
compatible with the principles of the common market.
4) Subsidy requirements
The following subsidy requirements are considered essential in accordance with UFG [2] and FRL
2015 [6]:
- The affected area is designated in the register of potentially hazardous sites regulation act including
priority classification.
- The contamination originated before 01.07.1989.
- The applicant is either the regional authority, the land owner, a party authorized to dispose or legally
obligated to remediate the contaminated site
- The applications for subsidies must be submitted before remediation is started
- An existing legally binding administrative decision (approval or contract) for the remediation
measures.
For all subsidized actions the subsidy recipient must perform procurement procedures in accordance
with the Federal Procurement Act.
5) Procedure of subsidies
After designation of an area as a contaminated site including priority classification a subsidy
application for rehabilitation or remediation can be submitted to KPC. An essential part of the
application is the variant analysis Thus the solution is determined in regard to the remediation
objective with the greatest possible environmental benefits at acceptable and justifiable
macroeconomic expenses. For the best option submitted costs are presented on the basis of an
estimate. The subsidy applications will be reviewed by KPC, the respective volume of subsidies will be
determined in accordance with FRL 2015 and a corresponding subsidy proposal is finally established.
The subsidy proposal will then be submitted to the "Commission on matters concerning the
remediation of contaminated sites" for assessment and evaluation. This Commission is constituted on
the basis of UFG [2] and acts as an advisory body of the responsible Minister of the Environment.
After positive evaluation of the subsidy proposal by the commission the Minister of the Environment
gives his approval. The actual subsidy agreement is usually concluded between the subsidy recipient
and the BMLFUW (represented by the KPC) after the completion of the planning and regulatory
approval process prior to construction. The disbursement and control of proper use of subsidies is
carried out by or via KPC. KPC in turn is subject to the control of the BMLFUW and the Court of
Auditors.
6) Volume of subsidies
The potential volume of subsidies is determined in accordance with FRL 2015 [6] by three criteria:
- Priority class of the contaminated site.
- Detectability or legal obligation of a person "responsible for the pollution".
- Subsidy applicant is a competitor (company) or not.
Depending on these criteria maximum subsidy rates of 55 to 95% of the eligible costs are possible.
The perpetrator of contamination after 1959 is considered to be responsible for pollution, with the
exception that the relevant environmental permits (eg water law or commercial law) were submitted to
and complied by in view of the contamination-related measures. It follows that for contamination
before the end of 1959, the highest subsidy category (65 to 95%) may always be granted. If there is a
legally obliged competitor, "responsible for the pollution", no subsidies are granted for this brownfield,
even if third parties (eg a local authority) - not responsible for the contamination - apply for a subsidy.
Should the legally obliged competitor, responsible for the pollution apply, then a "De-minimis" financial
support (currently a maximum of EUR 200,000, -) can be granted.
If by the subsidized measures increase the value of the affected property, the subsidy will be reduced
accordingly.
7) Research subsidies
As part of the subsidies granted for the remediation of contaminated sites [2] financial means can be
made available - in accordance with UFG - for research and development as well as for studies. The
aim of these subsidies is the application of advanced technologies that minimize both the resulting
emissions and the remaining contamination in the previously contaminated area. In the period
between 1993 and 2014 - according to the Environmental Support report 20144 [7] - the minister of
the environment approved of and authorised 36 research projects and studies with a total investment
of 18.7 million Euro that in turn were subsidized with 14.8 million. This corresponds to an average
subsidy rate of 79.1%.
The subsidy rates are based on the Research and Technology Promotion Act and in accordance with
the guidelines for the promotion of economic and technological research, technological development
and innovation. Depending on the organizational assignment to the categories "pre-industrial
technology development", "applied research" and "basic research", maximum subsidy rates of up to
25%, 50% and 100% are possible [8].
A special thematic focus of research funding is the further development of in-situ remediation
technologies and their combined use.
Literature
[1] Altlastensanierungsgesetz (ALSAG [1]; BGBl. Nr. 299/1989 idF BGBl. I Nr. 103/2013):
Bundesgesetz vom 7. Juni 1989 zur Finanzierung und Durchfhrung der Altlastensanierung, mit
dem das Umwelt- und Wasserwirtschaftsfondsgesetz, BGBl. Nr. 79/1987, das
Wasserbautenfrderungsgesetz, BGBl. Nr. 148/1985, das Umweltfondsgesetz, BGBl. Nr.
567/1983, und das Bundesgesetz vom 20. Mrz 1985 ber die Umweltkontrolle, BGBl. Nr.
127/1985, gendert werden.
[2] Umweltfrderungsgesetz (UFG [2]; BGBl. Nr. 185/1993 i. d. F. BGBl. I Nr. 51/2015): Bundesgesetz
ber die Frderung von Manahmen in den Bereichen der Wasserwirtschaft, der Umwelt, der
Altlastensanierung und zum Schutz der Umwelt im Ausland.
[3] http://www.publicconsulting.at/kpc/de/home/umweltfrderung/fr_betriebe/altlasten/
[4] http://www.umweltbundesamt.at/umweltinformation/altlasten/statistik/
[5] Verordnung (EU) Nr. 651/2014 der Kommission vom 17. Juni 2014 zur Feststellung der
Vereinbarkeit bestimmter Gruppen von Beihilfen mit dem Binnenmarkt in Anwendung der Artikel
107 und 108 des Vertrags ber die Arbeitsweise der Europischen Union (AGVO 2014)
[6] Frderungsrichtlinien 2015 fr die Altlastensanierung oder sicherung: Richtlinie des
Bundesministers fr Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft auf Grund der 13
und 29 ff des Umweltfrderungsgesetzes (UFG [2]), BGBl. Nr. 185/1993, idF BGBl. I Nr. 51/2015.
[7] Bundesministerium fr Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft (Hg.):
Umweltfrderungen des Bundes 2014, Wien.
[8] http://www.umweltfoerderung.at/uploads/infoblatt_antragstellung_altlastenforschung.pdf
[9] Bundesministerium fr Land- und Forstwirtschaft, Umwelt und Wasserwirtschaft (Hg.): Leitbild
Altlastenmanagement Sechs Leitstze zur Neuausrichtung der Beurteilung und Sanierung von
kontaminierten Standorten. Abteilung VI/3, Abfallbehandlung und Altlastensanierung, Wien 2009.
Session 1C.15 Combined Treatment Technologies 2
George Y. Maalouf, Rogers & Callcott Environmental , 426 Fairforest Way, Greenville, SC, USA 29607,
Phone: 864-232-1556, Email: george.maalouf@rogersandcallcott.com
Abstract
A recent release of approximately 7,600 Kilograms of trichloroethylene (TCE) that occurred in the mid-90s
resulted in a 6-hectare dissolved groundwater plume. Initial TCE concentrations in the source area
measured as high as 81,000 mg/Kg in soil and 1,200 mg/L in groundwater, near its water solubility limit.
The owner has set an ambitious clean-up objective to reach the allowable TCE concentration in drinking
water of 0.005 mg/L throughout the entire plume by the year 2025.
Meeting the site clean-up goal required an aggressive approach of intensive interim remedial measures
while completing the Remedial Investigation, Risk Assessment, Feasibility Study, and preparing for the
final Remedial Action. A remediation strategy was developed to meet the various technical and schedule
requirements of this particularly challenging site characterized by relatively high source area
concentrations, low permeability saprolite overlying fractured bedrock, a low natural attenuation rate, and
a large plume area with limited accessibility. Two remediation technologies were applied that are not
usually considered in consort for the same site. This paper presents remediation performance following a
full scale injection of In-Situ Chemical Oxidation (ISCO) and In-Situ Chemical Reduction (ISCR) reagents.
Approximately 83% of the contaminant mass was removed by the interim remedies, which consisted of
excavation, soil vapor extraction (SVE) and In-Situ Thermal Desorption (ISTD) in the source area, and a
Pump & Treat system along the property boundary. In the final remedy, these were replaced by SVE in
the vadose zone, aggressive ISCO injection for rapid and complete contaminant mass removal in the
source area aquifer, and a series of passive, long-lasting permeable reactive barriers (PRBs) using ISCR
to address long-term contaminant advection and diffusion in the plume area. The latter two technologies
are rarely applied at the same site because of their antagonistic redox environment and must be given
careful consideration during design and implementation. However, their application as parts of the same
combined remedy can be greatly beneficial at sites with similar challenges. The efficacy of combining the
antagonistic ISCO/ISCR remedial approaches was pilot tested at the same site prior to the full scale
implementation. Modeling and monitoring were conducted during design and implementation as a basis
for reagent requirements, separation between reagents, injection point spacing, and scale-up for future
expansion of the treatment. The injection wells were installed using sonic drilling methods to allow for a
close examination of the overburden and bedrock core in the target injection intervals. Due to the low-
permeability saprolite and random fracturing patterns in bedrock, traditional injection methods have
proven unsuccessful at this site. Hence, both reagents were emplaced as slurry in fractures induced
using a hydraulic injection technique at 1.2 meter vertical spacing.
For the ISCO source area treatment, 75 metric tons of potassium permanganate (KMnO4) and sand blend
(50% each) were injected in three stages between 2011 and 2014, including the pilot study. The ISCR
treatment consisted of injecting 658 metric tons of zero-valent iron (ZVI) in three PRBs that effectively
divide the plume into four segments, thereby drastically shortening its lifespan. Performance monitoring
has been conducted at least quarterly. As of the first quarter of 2015, 13 of the 15 monitoring wells in the
source area have experienced >99.9% reduction in TCE concentrations with the remaining two wells at
99% and 89% reduction from the ISCO baseline levels. Permanganate was observed in one discrete
fracture past the first ZVI PRB, but reducing conditions have been sustained in the immediately-adjacent
ZVI injection and monitoring wells; therefore, the permanganate has not adversely affected the PRBs.
The ZVI performance monitoring wells have exhibited significant reductions in TCE concentrations
ranging between 83.2% and 99.8% at the most upgradient PRB, between 73.7% to >99.9% at the mid-
PRB, but at lower rates in the most downgradient PRB (71.4% - 74.8%), where groundwater flow rates
are relatively lower and bedrock is less fractured.
Introduction
Remediation of a chlorinated solvent plume in groundwater presented several challenges due to: 1) high
contaminant concentrations in the source area, 2) low permeability saprolite overlying fractured bedrock,
3) low natural attenuation rate, 4) large plume area with limited accessibility, and 5) a very aggressive
remediation timeframe selected by the owner. After completing a Remedial Investigation / Risk
Assessment / Feasibility Study (RI/RA/FS), Rogers & Callcott Environmental was tasked with identifying
an aggressive remedial approach to meet these challenges. Working with Geo-Cleanse International, a
novel remedy was developed and pilot-tested in 2011. Based on the success of the pilot study and the
FS, a Record of Decision (ROD) was issued for full scale implementation. Installation of the final remedy
was completed in mid-2014. Monitoring results through the first quarter 2015 are presented in this paper.
Background
The site is a former manufacturing facility located in the Inner Piedmont geologic block of Upstate South
Carolina (Figure 1), in an area characterized by the metamorphic and igneous rocks of the Six-Mile and
Paris Mountain Thrust Sheets and the Table Rock Plutonic Suite. Hydrogeology of the site is
SiteLocation
Figure1.GeologicMapofSCshowingsitelocation(Referenceasshown)
characterized by a dual groundwater system comprised of a saprolite aquifer and a fractured crystalline
bedrock aquifer. The saprolite/partially weathered rock zone is 1 and 10 meters below ground surface
(m. bgs) and is primarily comprised of silty sand and sandy silt derived from the weathering of the
underlying bedrock. The upper bedrock exhibits varying degrees of fracturing and weathered zones in a
matrix of mica schist and gneiss, feldspar gneiss and granite. The depth to rock ranges from 2 to 30 m.
bgs. Groundwater flow at the site is partially radial in an overall southwestern direction towards a surface
water discharge area (Figure 2), and occurs approximately 1.5 to 20 m. bgs. The average groundwater
seepage velocity is approximately 15 m/year.
It is estimated that approximately
5,200 liters of trichloroethylene
(TCE) were released between the
time the facility was constructed in
1991 and the discovery of the
N release in 1996. Site
investigations revealed a
groundwater plume of TCE and
the breakdown product cis-1,2-
dichloroethene (cDCE) that covers
approximately 6 hectares, with
concentrations historically greater
than 250 mg/L in the source area
and up to 96 mg/L in the mid-
plume area. Approximately 130
metric tons of shallow impacted
soil were excavated from the
source area during the removal of
the TCE storage tank and
associated sumps and piping.
Several interim remedial
measures were applied to address
the remaining impacted soils in the
source area and mitigate off-site
plume migration. These measures
effectively met their goals,
recovering approximately 83% of
the total estimated quantity of
Scale = 50 m discharged TCE. The recovered
Figure2.TCEPlumeshowingsourceareaandgroundwaterflowdirection amounts were attained by soil
removal and shallow soil vapor
extraction (SVE) in the excavation
(8%), in-situ thermal desorption (ISTD) in
the vadose zone and saprolite and
bedrock aquifer within the limits of the
source area (69%), and groundwater
pump and treat along the downgradient
property boundary (6%). Despite the
success of the interim remedies,
groundwater modeling predicted a
remediation timeframe of seven decades,
and additional removal would be required
to meet the aggressive remediation goals
of reaching drinking water Maximum
Contaminant Levels (MCLs) of 0.005
mg/L within twelve years as directed by
the owner.
Remedial Approach
A remediation strategy was developed to
meet the various technical and schedule
requirements for this particularly
Figure3.Lowpermeabilitysaprolite (top)andtransmissivebedrock
challenging site by in-situ treatment within the saprolite, bedrock and the in-between transition zone
(Figure 3). The strategy included a novel combination of no-maintenance technologies coupling an
aggressive source-area remedy for quick source mass removal with a long-lasting series of plume area
barriers to address long-term advection and diffusion of TCE from inaccessible areas, thereby shortening
the lifespan of the plume. In-situ chemical oxidation (ISCO) using potassium permanganate (KMnO4) in
the source area was integrated with in-situ chemical reduction (ISCR) using zero valent iron (ZVI) barriers
in the downgradient plume area. A large scale pilot study of the remedial approach was implemented in
mid-2011 by hydraulically injecting reagents as high-solids slurries to distribute large reagent volumes
effectively within specific and focused target zones in the saprolite and fractured bedrock. Modeling and
monitoring were conducted as part of the design and implementation as a basis for reagent requirements,
injection point horizontal and vertical spacing, scale-up for future expansion of the treatment, and to
monitor any interaction between the antagonistic reagents. Following two years of monitoring, a scale-up
of the pilot study was implemented in 2013 and 2014.
Injection Mechanics
Traditional injection methods of liquid reagents had been tested at this site but proved to be unsuccessful
due to the nature of the low permeability soils and random fracturing. Hydraulic slurry emplacement, the
selected method for the pilot study and final remedy construction, is a process in which a fluid is
introduced at a pressure sufficient to overcome the in-situ confining stress and the material strength of the
formation, in order to mechanically induce or enhance fluid flow into the formation. Slurry can be injected
at multiple discrete depth intervals within a boring. Injections are conducted at discrete depth intervals as
opposed to pressurizing the entire borehole to focus the nucleation stresses in a horizontal plane. The
injection is begun by first cutting a kerf through the casing and into the exposed borehole wall. The kerf
is cut with a water jet or water blaster at a pressure of up to 10,000 psi (which in weathered saprolite has
been shown to extend 15-20 cm into the soil) transforming the borehole from a pressurized cylinder to a
pressurized disk. A slurry is prepared using tanks and mixing systems on a support trailer and then
injected into the formation. Stress induced by the injection pressure accumulates in the hoop formed by
the kerf in the horizontal plane, nucleating a horizontal emplacement. Once the emplacement is
nucleated, the initial horizontal form influences the course of propagation briefly, but soon natural forces
become significant and the final form of the emplacement can be quite irregular. This process is repeated
for each vertical treatment interval.
Pilot Test
The permanganate component of the pilot test targeted a portion of the source area, and the ZVI portion
targeted a mid-plume location near the property boundary (Figure 4). The treatments each targeted the
vertical interval from the water table in the saprolite to approximately 3 m into bedrock, covering the
zones in which impacted groundwater is found.
A total of five soil borings were utilized for
the ZVI slurry injection. One of the
locations was constructed with a 10-cm
N diameter, Schedule 40 PVC casing that
was grouted into place. Injection in the
remaining four was conducted through
open boreholes that were completed as
screened wells for potential future
monitoring within the barrier. The locations
were oriented in a line perpendicular to the
direction of groundwater flow. The boring
locations were installed with a horizontal
spacing of approximately 9 m. The ZVI was
injected into four to six discrete depth
intervals within each boring. The vertical
spacing was selected based upon
hydrologic modeling of groundwater
capture into the permeable slurry
Figure4.Pilotstudyplotsshowingplumepriortotreatment emplacements. This approach was
believed to be a very conservative design
intended to ensure both horizontal and
vertical overlap between emplacements. A granular form of ZVI was utilized rather than microscale or
nanoscale form, due to the much longer lifetime in the subsurface. The amount of ZVI required was
based upon three main factors: initial concentrations of
VOCs entering and exiting the barrier, the travel
velocity of the VOCs, and the length and vertical
thickness of the treatment zone. The ZVI slurry was
prepared by blending the ZVI with guar gum in an
aqueous matrix, which thickens the slurry to suspend
the ZVI in the fluid (Figure 5). An enzyme is included in
the slurry to slowly break down the guar gum over a
period of hours following injection (as verified by bench
testing), in order to expose the ZVI to react with the
groundwater. Approximately 66.4 metric tons of ZVI
Figure5.ZVISlurry
were emplaced into the five boring locations, forming a
45-m barrier line.
Two locations were utilized for the KMnO4 slurry emplacement application. Based upon the results of the
ZVI injections, it was determined that constructing the borehole with the PVC casing was the optimal
design. Therefore, the two locations were constructed with 10-cm diameter, Schedule 40 PVC casing that
was grouted into place. The locations were approximately 7.5 m apart, which was considered a
conservative spacing. The KMnO4 product that was emplaced was RemOx-S, manufactured by Carus
Corporation. Sand was blended into the permanganate in a 50/50
ratio by weight (Figure 6) in order to reduce crusting due to
manganese dioxide precipitation; the sand selected has a grain size
distribution nearly identical to that of the KMnO4 in order to minimize
grain sorting and associated permeability reduction during injection.
The permanganate slurry was prepared using bentonite to suspend
the solid permanganate in the slurry, rather than guar gum as used
with the ZVI, because permanganate reacts with the guar gum.
Tests and calculations demonstrated that the very small amount of
bentonite utilized was insufficient to measurably reduce the
permeability of the emplaced slurry. The amount of KMnO4 required
for the pilot test was based upon residual contaminant mass
estimates and permanganate oxidant demand bench tests. The
amount of permanganate utilized reflected the combined total of the
soil oxidant demand plus the VOC oxidant demands, which was Figure6.KMnO /sandblend
4
12.8 metric tons of KMnO4 and sand blend. Permanganate was
emplaced at five vertical intervals in each boring location.
The pilot study scale-up in the source area consisted of two injection events one year apart of an
additional 62.8 metric tons of KMnO4/Sand blend in 77 injection intervals. Seventeen injection wells were
used covering an area of approximately 390 square meters and extending 24 meters into the saprolite
overburden and 4 m into the underlying fractured granitic gneiss. Figure 11 shows the number of injection
intervals in each zone and injected quantities.
The full scale barriers consisted of injecting an additional 591.5 metric tons of granular ZVI in 57 injection
wells and 345 intervals in three barriers. The PRB lengths range from 73 m to 161 m with 57 injections
wells installed at 6 m of separation. Vertically, the injections were completed in the overburden
soil/saprolite, the transition zone of highly weathered and fractured rock and in existing discrete fractures
of the competent bedrock, up to 13 m into the underlying granitic schist and gneiss (Figure 12). The
location of the source treatment area and three PRBs is shown on Figure 13.
Figure12.CrosssectionofPRBAimmediatelydowngradientofthesourcetreatmentareashowingtheinjectionintervals(x)and
totalZVI(USTons)injectedineachboring.Thescaleisinfeetwithnoverticalexaggeration.Yellowishgreenshadingrepresentsthe
highlyweatheredtransitionzone.Thebluelinerepresentsthewatertable.
Injection borings were drilled with sonic methods and completed with solid PVC casing. This allowed for
the entire drilled column to be retrieved and examined for optimal placement of the injection intervals,
which proved invaluable in identifying the weathered transition zone and discrete bedrock fractures
targeted for injection.
Full Scale Results
Performance monitoring has been conducted on a
quarterly basis. So far the data demonstrate that the
ISCO treatment has been successful in almost
complete destruction of the contaminant mass in the
source area. Despite a rebound in some of the wells
following an initial drop immediately after the 2011 pilot
injection, TCE concentrations remained significantly
below baseline levels after 21 months in eight of the N
monitoring wells (Figure 9, and as shown on slides (a) KMnO4
and (b) of Figure 14). The first main injection event
(2013) affected further destruction of the source mass ZVIPRBA
but resulted in a localized but significant TCE
concentration increase beneath the footprint of the
former TCE handling area (slide (c) of Figure 14).
Following three quarterly sampling events after the
2014 injection, 13 of the 15 monitoring wells in the ZVIPRBB
source area have experienced >99.9% reduction in
TCE concentrations, with the remaining two wells at
99% and 89% reduction from the ISCO baseline levels,
and most of the source area concentrations remain
below the target level of 0.005 mg/L (slide (d) in Figure
14). The monitoring well with the least reduction is 50 m
located at the downgradient edge of the source area ZVIPRBC
and is expected to improve as the source is depleted.
Approximately six months following the last injection,
permanganate was observed in one discrete fracture
Figure13.PlumemapasofJanuary2015showingthe
past the first ZVI PRB, but reducing conditions have KMnO4sourcetreatmentareaandthethreeZVIPRBs
been sustained in the immediately-adjacent ZVI
injection and monitoring wells. Further sampling and a camera survey in the affected wells show the
permanganate presence is limited to a single fracture indicating no adverse effect on the PRBs.
The ZVI performance monitoring wells have exhibited significant reductions in TCE concentrations but at
a lower rate than the ISCO treatment area as expected. At the most upgradient barrier immediately
beyond the ISCO treatment area (PRB A), TCE concentrations have dropped by more than 99% below
baseline levels in four of the five monitoring wells and 83.2% in one well. A similar trend is observed at
the mid-plume barrier (PRB B) with reductions between 73.7% and >99.9%. The lowest concentration
reduction rates are observed at the most downgradient barrier (PRB C) with decreases of 71.4% and
74.8% from baseline. This area of the site is characterized by extremely small water level gradients and a
non-existent transition zone of weathered rock between the saprolite and competent bedrock. A faster
than expected reduction in concentration is being observed in a bedrock monitoring well located
approximately 70 m downgradient of PRB C, likely a result of the discrete fracture flow regime. The
measured contaminant concentrations in most wells meet or exceed those obtained from modeling
results. Performance monitoring continues with room for future adjustments if necessary.
Acknowledgments
The writer would like to thank the owner for the unwavering commitment and support of this project, and
Mr. Dan Bryant, Ph.D. of GeoCleanse International, and Mr. Bill Slack, Ph.D. of FRx for their work on this
project and contribution to this paper.
N
Scale = 12 m
(a) (b)
(c) (d)
Figure14:TCEplumeinsourceareawithpreandpostISCOtreatmentprogression.(a)June2011,
priortopilotstudy;(b)May2013,21monthspostpilotinjectionshowing2injectionwells;(c)May
2014,9monthspostfirstmaininjectionshowing7injectionwells;(d)January2015,6monthspost
supplementalinjectionshowing10injectionwells. Indicateswellsusedineachinjectionevent.
ThS. 1C.19
1 2 3 4
S. Pensaert , H. Nollet , E. Rombaut and X. Lepre
Abstract: The sediments in Valdemarsviken have been contaminated with chromium from earlier
discharge of untreated wastewater from tannery operations. The remediation of the Valdemarsvik fjord
is very important for the water quality in the Baltic Sea. Contaminated sediments in the inner part of
Valdemarsviken were dredged, dewatered, stabilized and placed as fill on land. As dredging of the
sediments in these areas would affect the stability conditions along parts of the shores and therefore
required the stabilization of quays, shores and the landfill area by lime-cement columns. The dredging
of contaminated sediments covered about 200,000 m spread over a 350,000 m large area. During a
detailed design phase, prior to the start of the works, the site conditions and sediments were
monitored and investigated thoroughly. Lab scale tests were performed to verify that the chosen
process for the dredging, dewatering and stabilization meet the set requirements as regards to
strength of the backfilled stabilized sediments specified by the client. The lab work also formed the
basis of process design for the on-site solidification plant. The dredging operations met the stringent
environmental demands towards turbidity and the generation of spill. The sediments were dredged by
using a hydraulic excavators equipped with a level-cut clamshell bucket or a visor bucket. The
dredged sediments were pumped to a continuous mixing unit where they were stabilized with a
cementitious binder. Any water released from the activities was collected and treated by a specially
designed on-site water treatment plant.
1
Head of R&D Department, DEME Environmental Contractors (DEC NV), Haven 1025, Scheldedijk 30, B-2070
Zwijndrecht, Belgium) pensaert.stany@deme.be
2
Project Engineer, DEME Environmental Contractors (DEC NV), Haven 1025, Scheldedijk 30, B-2070
Zwijndrecht, Belgium) Nollet.hendrik@deme.be
3
Project Manager, DEME Environmental Contractors (DEC NV), Haven 1025, Scheldedijk 30, B-2070
Zwijndrecht, Belgium) rombaut.elke@deme.be,
4
Project Engineer R&D, DEME Environmental Contractors (DEC NV), Haven 1025, Scheldedijk 30, B-2070
Zwijndrecht, Belgium) lepere.xavier@deme.be
ThS. 1C.19
INTRODUCTION
The Valdemarsviken fjord is located in Valdemarsviks municipality in the southeastern part of the
stergtland country on the eastern coast. Valdemarsviken represents the only threshold fjord on the
East Coast and stretches about 11 km from the town of Valdemarsvik out to the mouth at the Gryt
archipelago in the Baltic Sea.
The sediments in Valdemarsviken have been heavily contaminated with chromium from the earlier
discharge of untreated wastewater from tannery operations at the former Lundberg Lder which
opened in 1873 and closed in 1960. Chromium III salts were used extensively in the tanning process
as it enables faster and cheaper production of highly resistant and durable leathers. Today, more than
80% of finished goods are still tanned using basic chromium salts (Thorstensen, 1993).
The remediation of the Valdemarsviken fjord is very important for the water quality in the Baltic Sea, as
it has has long produced the biggest single release of chromium to the Baltic Sea, i.e. about 250 kg
per year.
Valdemarsviken is primarily used as a recreation area and for boating, fishing and swimming. The
existing plans for continued development of the area are linked to the development of marine
operations and to some extent housing. There are also plans for new areas for outdoor recreation and
sports activities.
In light of this, Valdemarsvik municipality decided to remediate the Valdemarsviken fjord with respect
to chromium contaminated sediments. The remediation consisted of the removal of the majority of the
contaminated bottom in the inner part of the fjord. The dredged and dewatered bottom sediments had
to be used as filling of a valley area about two kilometer south-east of the village. The dredging works
covered an area of about 350,000 m and will remove as much as 400 tons of chromium which equals
a reduction in the emission of chromium by 90%. The project helps to reduce the load to the Baltic
Sea and to achieve the environmental objective of a non-toxic Environment.
The remediation measures included dredging, treatment of dredged sediment, water treatment, reuse
of dredged material, shoreline reinforcements, construction of quay, clearing of obstacles to dredging
etc.
SITE INVESTIGATION
The sediments in Valdemarsviken were mainly contaminated with chromium. The average chromium
concentration observed during site investigation varied from about 1000 mg/kg DM at the deepest
levels up to about 3000 mg/kg DM in the near-surface sediments. The maximum measured chromium
concentration was approximately 7000 mg/kg DM. Chromium occured primarily as trivalent chromium
in the sediments, hexavalent chromium was initially not analyzed at levels greater than the detection
limit, although during the works it was detected several times. In addition the sediment samples
contained low to moderate levels of copper, lead, zinc, arsenic and mercury.
According to the main site investigation report, the total amount of chromium was estimated at 700 -
800 tons.
In order to achieve the remediation objective, a dredging design level was established based on a
remediation target of 500 mg/kg DM.
Based on the site investigation report about 200,000 m of sediment needed to be dredged.
In Valdemarsviken there are soil layers of clay with low shear strength (the so-called quickclays) and
high risk of landslides. Dredging of the sediments in these areas would affect the stability conditions
along parts of the shores affected by the dredging. The topographical and geotechnical conditions in
the bay, with steep banks and loose soil layers of clay mean that even relatively small load changes
can result in landslides. The near-shore dredging in some areas had to be combined with soil
reinforcement measures.
ThS. 1C.19
Soil reinforcement measures were executed in three areas along the coast line (areas are marked as
1, 2 and 3 on Figure 2). The proposed technique by the client was to perform the ground
reinforcement with vertical lime cement columns (KC-columns). This dry mixing method was used to
stabilize the very week soils along the shore line in Valdemarsvik.
The dosing and blend of lime and cement was defined by an extensive testing program. This testing
program covered the testing of binding agents KC (K=lime; C=cement) 50/50 and KC 25/75. Based on
the results from the performed tests, eventually the area was treated by using the binding agent KC
25/75, 28 kg/m binder and a rate of ascent of 25 mm/rotation (in combination with expanded
production monitoring).
The testing results with KC 25/75 and 28 kg/m showed that the columns reached a compressive
strength of about 50 kPa after 2 two days and about 150 kPa after 14 days. The surface will be
subjected to loads about 7 days after the installation of the columns. A conservative assessment of the
strength in the columns of 50 kPa after 7 days means that the strengthened clay has a compressive
strength of > 20 kPa.
Purpose designed drilling rigs were used for the installation of the KC-columns, equipped with a
special mixing tool at the end of the mixing rod. The drilling rigs installed the columns from a working
platform or reinforced clay surface (Figure 1). Columns with a diameter of 60 cm were installed up to a
depth of 22 m. Dosing and mixing of the dry binder blend with the soil took place while the rod was
withdrawn. The KC-mixture was delivered by bulk trailers from the supplier and stored into a shuttle on
site. The binder was transported from the shuttle to the rig through connecting hoses using
compressed air. By changing the rotation speed of the feeding wheel, the binder quantity was
adjusted. The columns were installed in a fixed pattern (Figure 1).
The mixing device moved in a corkscrew motion till it reached the fast bottom or the designed length.
From that moment the operator opened the valve and lifts the mixing device with a predefined lifting
and rotation speed. The operator controlled the whole process while producing the column. When
arrived near the surface the mixture addition was stopped and the length of the stabilized column was
recorded. About 400,000 m of lime-cement columns were installed.
ThS. 1C.19
After installation of the KC-columns the columns, validation tests based on in-situ vane tests were
executed in the installed columns. Those tests were performed after 12-16 days and 26-32 days. If the
strength was not enough or if weak spots were noticed in the columns, the amount of binder was
increased (e.g. 32 or 34 kg/m).
ENVIRONMENTAL DREDGING
The volume of contaminated sediments that required dredging was estimated to be about 200,000 m,
to be dredged over a surface of about 350,000 m. This implied the dredging of rather thin layers of
sediment (Figure 2). The depth of dredging ranged up till 14m.
The choice of the dredging technique was considered in the context of the entire project:
- The dredging method had to be compatible and fully integrated with the materials handling,
transport, treatment, and ultimate disposal of the dredged material.
- The dredging execution had to meet the stringent environmental demands towards turbidity
and the generation of residual spill.
An evaluation was performed of several environmental dredging techniques taking into consideration
the site conditions, sediment characteristics, environmental dredging performance standards,
equipment capabilities, operating methods and strategies.
It was believed that dredging using a hydraulic excavator equipped with a level-cut clamshell bucket or
a visor bucket could provide optimum dredging production, debris management, and dredging
accuracy for the site specific conditions (Figure 3 and 4). In order to obtain extreme accuracy in
dredging, the dredge was equipped with the latest positioning system. High accuracy dredging is
essential for the successful execution of clean-up projects. It leads to less volume to be transported,
stored and treated. Furthermore, it leads to less turbidity and spillage.
1
2
The level-cut clamshell bucket was designed for the dredging works in Valdemarsviken. The bucket
size was adjusted to the depth of the cut to maintain a higher fill factor and limit the amount of water
ThS. 1C.19
captured in the bucket. This was done to prevent overfilled buckets which can have a significant
negative impact on the water quality and residuals.
When thick banks of sediment needed to be removed, multiple dredging lifts of 30 cm each helped in
limiting bank instability and sloughing that can contribute to residuals.
The enclosed level-cut clamshell bucket has been developed to reduce spillage and leakage from the
clamshell while being raised through the water column. The total content of chromium in the re-
sedimented materials within the dredging area of Valdemarsviken may after dredging not exceed 6
tons of chromium (which equals an average of less than 20 g Cr/m).
Figure 3. The Ijburg in combination with a level-cut clamshell bucket will be used to perform the
dredging works in Valdemarsvik
In addition to the operational controls, a mobile silt screen was installed at the dredge around the
reach area of the excavator (see figure 3). Fixed silt screens were installed further downstream the
fjord. The maximum allowed increase in turbidity in Valdemarsviken immediately outside the area
where turbidity-causing works were in progress (50 m from the dredge) was 9 NTU in relation to a
reference point upstream.
The dredged sediments were loaded into sealed barges (400 - 600 m) for transport to the stabilization
plant onshore.
ThS. 1C.19
The area to be dredged was extensively sampled by using a manual core sampler, allowing
undisturbed sampling at the desired depths. Disturbed and undisturbed samples were collected.
Disturbed samples were used to perform the labscale stabilization tests. The in-situ parameters of the
sediments were determined on the undisturbed samples. For samples taken at the top of the
sediments attention was paid to pump off the water layer as shown at Figure 5 to ensure an in-situ
quality of the sediments.
The frequency of the sampling was adjusted in correlation of the volumes to dredge in each part of the
area. Each sampling location was individually characterized to have a good overview of the variations
of the physical and chemical properties of the sediments. Finally the individual samples were
composited into three composite groups, based on the location, density, granulometry, visual
observation Feasibility trials on lab scale were performed on these three composite samples.
The basic requirement of the client was that the sediment should be backfilled at the Grnns area,
and should have a shear strength of at least 25 kPa, 1 month after placement.
In order to achieve this objective, starting from the dredged sediments, all technically possible options
were evaluated. Although dewatering of the sediments by means of a membrane filter press was the
most interesting technical solution as it produced filter cakes with the necessary shear strength, the
investment and space required for this technique was too important and the executional time frame too
short.
Solidification of sediments, as originally suggested by the client, turned out to be the most cost-
efficient method within the time and space constraints. Therefore mechanical dredging was obviously
preferred above hydraulic dredging as a minimal dilution with water is achieved which is in favour of
the binder consumption.
ThS. 1C.19
Different binders available in the Valdemarsvik area were tested with different concentrations and
curing times. Cement-based binders gave the best results meeting sufficient strength with the lowest
dosage. An overview of these results is given in figure 6. Note we tested the shear strength indirectly
by means of determining the Unconfined Compressive Strength (UCS), divided by a factor 2.
140
120
100
UCS (kPa)
80
60
40
20
0
14 28 56
Curing time (days)
The binder dosage depends on the water content of the dredged sediments. This one varies during
the whole dredging activity; a lot of parameters are involved in these variations (dredged zone,
variation in sediment granulometry, depth of the layer to dredge, slope of the bottom ). Thanks to the
lab trials, it was possible to establish a relationship between the water content of the sediments and
the amount of binder needed to meet the criteria. This relationship is shown in figure 7.
ThS. 1C.19
Figure 7. Relation between water content and cement consumption to achieve 50 or 80 kPa UCS (=
25 or 40 kPa shear strength)
As the consumption of the binder during stabilization of the sediments is increasing exponentially with
increasing water content of the sediment (see figure 7) the appropriate dredging technique was
selected. Mechanical dredging by means of a level-cut clamshell or a visor bucket was used, and
after transportation of the sediments by barges to the unloading quay supernatant water was carefully
removed and treated in a tailor made water treatment plant on site, consisting of a flocculation unit,
lamella separator, sand and activated carbon filters. The sediments were unloaded by a clamshell
excavator from the barges, screened, and pumped into the continuous mixing facility. The mixing
facility had a throughput between 200 and 300 m/h.
Just before the mixing plant the water content of the sediments was measured by means of an in-line
density measurement. In combination with the flow measurement, this density value was used to
calculate the dosing of the binder. A continuous measurement and process control ensured the right
dosage of binder at any time. During start-up of the plant the binder dosage was frequently verified by
means of portable XRF measurements.
Immediately after mixing the sediments were transported by dumper trucks onto the Grnns area
where they were spread open by long reach excavators.
During the project, lots of samples were taken for verification of the stabilization process. The
samples were moulded, stored in a so-called library, and used to determine the shear strength after 1,
3, 7, 14, 28, 56 days. The strength after 28 days was seen as the contractual target. As it was
possible to establish a good correlation between the early strength development (1 and 3 days) and
the 28 days, early predictions could be made.
ThS. 1C.19
Finally the backfilled Grnns area of about 75000 m was covered with a synthetic geomembrane,
bentonite mats and moraine materials.
Figure 8. From right to left: unloading of barges, mixing plant, spreading of the stabilised sediments
into the backfilling area. In the front the water buffers from supernatant and run-off water can be seen.
REFERENCES
I. INTRODUCTION
Today, there are many in situ remediation technologies used regularly around the world in countries
with established clean-up legislation, ranging from extractive (pump-and-treat, thermal), chemical (in
situ chemical oxidation / reduction) to biological (enhanced / monitored natural attenuation). That so
many options are available is testimony in itself that no one technology is ideal in all circumstances
if that were so, it would follow that that technology alone would be employed. The variety of
technologies available is instead a direct reflection of the variety of strengths and weaknesses they
each present, be it in the balance of achievable cost vs. time, the degree of intrusion necessary for
their furtherance, their optimum concentration range for maximum efficiency, or their suitability to a
given geological, hydrological or geochemical setting.
Given the natural heterogeneity of most contaminated systems, it is perhaps surprising therefore that
in many cases, single technologies are applied alone the very process of clean-up will itself
progressively change the setting over time, whether this simply be in terms of contaminant
concentration, or other factors too, such as geochemistry or contaminant availability, relative
distribution and composition. It would follow that the performance and efficiency of the selected
technology will vary across the site or over time, and its selection must therefore, at the very least, be
a compromise.
Treatment Trains
The recognition of the above is leading progressively to the increased uptake of the treatment train
approach, in which technologies are being used in combination, whether this be spatial: one
technology for a source area, another technology for the plume for example or temporal: one
technology used for gross mass reduction, followed by another for polishing to mandated clean-up
targets. Such combined use may either be by design or by default switching to another technology
on failure of the initial technology to secure the desired targets within a given timeframe. Both
approaches have relevance, but clearly, proactive technology integration by design would be
preferential to integration as a remedy for real or perceived failure of the initial technology.
History one-size-fits-all
Notwithstanding the above, technology integration is still not as widespread as the benefits it confers
would suggest. This may be down to a number of possible factors. Technology selection may
arguably still reflect vestiges of the early days of environmental remediation, where the one-size-fits-
all model was a more accurate reflection of the limited remedial technology choices available
excavation was for the most part the approach for soils; extraction was the approach for groundwater.
As new technologies arose, these were typically reviewed within an options matrix, and frequently,
one technology was selected as the alternative to excavation or pump-and-treat.
Theme 1C: Strategies for remediation and brownfield regeneration
Lecture Session (ThS): 1C.7
It may or may not be fair to attribute the use of single technologies to a one-size-fits-all habit or the
inertia of familiarity, but the approach would undoubtedly have offered the convenience of simple
communication and familiar contractual models. But in fairness, it has only been with time that the
high frequency of diminishing returns and progressively asymptotic performance of in situ
technologies has become apparent. Indeed, this remains a focus of study today, for example, in the
form of elucidation and control of matrix back-diffusion (Seyedabbasi et al. 2012) and optimization of
contaminant availability for extraction / destruction.
A clear oversight of each of the technology sectors is required in order to select and balance them
optimally, and to manage an integrated treatment project through to successful conclusion. If this is
achieved, combining remedies has the potential not only to reduce overall remediation cost and time,
but also to increase net efficiency and flexibility in relation to addressing new information and
challenges, and adjustment to an evolving conceptual site model that should naturally be refined
through the greater insight afforded by the intense, invasive and dynamic nature of any remediation.
The intention of this paper is therefore to provide a simple conceptual overview that places each
predominant technology sector in context, and points to key considerations necessary for successful
technology integration.
At the top end of the concentration scale, physical systems can offer maximum mass removal per unit
cost, with little or no top-end limitation. Removal of free-product for example by physical systems
clearly represents a peak of mass-reduction efficiency. As concentrations are progressively reduced
however, physical systems can begin to struggle, with diminishing returns, asymptotic performance
and rebound having long been recognized as characteristic observations (Mackay and Cherry, 1989).
Similarly, where the starting concentrations are themselves low, the physical cost of the system
compared to the contaminant mass may itself represent an inefficiency as defined above.
In contrast, at the bottom end of the concentration scale, biological systems can present benefits of
simplicity, low unit costs, broad matrix applicability and elimination of rebound. Further up the
concentration scale however, reagent requirements and/or timescale can become excessive,
especially as the domain of smear and residual free product (for aerobically degradable
Theme 1C: Strategies for remediation and brownfield regeneration
Lecture Session (ThS): 1C.7
contaminants) is reached. This is generally the point at which physical systems, from the other
extreme, also begin to exhibit an efficiency tailing.
Chemical oxidation technologies, in contrast, conveniently exhibit their greatest efficiency broadly at
the point in the concentration spectrum where both biological and physical approaches struggle.
Reagent demand stoichiometry and the requirement for aqueous-phase reaction presents for
chemical technologies an analogous high-end efficiency tailing to that of bio, albeit at a higher
concentration. As contaminant concentrations reduce, efficiency also decreases as molecular
collision frequency between oxidant and contaminant reduces, and mass-transfer limitations begin to
dominate, typically requiring multiple applications of oxidant for decreasing incremental benefit.
Indeed, based on the evaluation of field performance of multiple projects, the success rate of ISCO in
meeting maximum contaminant levels (MCLs) has been low (15%, n = 39) (Krembs et al. 2011).
Greater success has been observed by the same authors for ISCO performance based on higher-
concentration targets (24%, n = 74), with the majority of full-scale sites that attained closure (89%)
having coupled ISCO with another technology (ibid.)
The schematic in Figure 5 thus not only serves to illustrate the efficiency optima of different
technology classes in relation to each other within a concentration spectrum where in the range one
approach may be more suitable than another for example but also illustrates the principle of
maintaining optimal performance efficiency through segueing from one technology class to the next
as remediation proceeds, especially if the required concentration reduction is great (i.e. several orders
of magnitude). Moreover, given that the performance tailings of each class are gradual, the figure
illustrates that the exact switch points may be relatively forgiving if each technology class is
employed.
It should be noted however that all technology classes might not be necessary on all projects. Where
concentrations are modest, (enhanced) bio alone may suffice, on all or part of a site. Similarly, chem-
bio integration is common, whether by design or by default. Phys-bio may also be employed, not just
with respect to facilitating the closure of pump-and-treat systems, but also following excavation or
thermal treatment. Rarely however is the sequence reversed. Similarly, rarely is bio not a component
of a combined remedy program although this may simply involve MNA rather than a form of
reagent-enhanced bio.
Compatibility considerations include foresight in physical system design, for example, design of
extraction installations to transition into injection arrays, and selection of (ISCO) reagents that are
compatible with the infrastructure and are not corrosive to pipes, wells or fittings or that are likely to
block the same. Similarly, integration efficiency is increased by selection of ISCO reagents that
promote rather than inhibit follow-on bio. For aerobically degradable contaminants, ISCO reagents
are generally synergistic to bio given that the ambient aquifer redox is invariably raised and chemical /
geochemical electron acceptors are consequently relatively abundant following treatment at least in
the short to medium term. The challenge is greater for anaerobic bio, however by the same token
the raised aquifer redox will need to be dropped to a status conducive to the biodegradation of the
target species, which may require significant reagent additions to overcome competing electron
acceptor demands. This challenge is exacerbated if the ISCO reagent itself either decomposes
incompletely (non-productive longevity) or leaves a residue that itself presents a competing electron
2-
sink (e.g. MnO2, SO4 ).
Technology switch points are important, but may or many not need to be objective. In many
cases, and under certain contractual arrangements, an engineers judgment based on professional
experience may alone suffice, albeit supported by appropriate testing. Similarly, spatial technology
integration generically does not require switch points; these are germane to regions where
technologies are used sequentially.
Theme 1C: Strategies for remediation and brownfield regeneration
Lecture Session (ThS): 1C.7
Clear identification of optimal switch points is important where transitions are abrupt, but becomes
less important as additional technology links are added and the integration becomes increasingly
seamless. For example, a physical treatment such as excavation or product recovery followed by
biotreatment would be an example of an abrupt transition, where issues of residual contamination and
smear neither of which is easily quantified can significantly impact the cost and performance of
the follow-on technology. In contrast, the intervening use between physical and biotreatment of an
intermediate recovery enhancement technology such as the proprietary Regenesis PetroCleanze
product (www.regenesis.com), or mass reduction technology such as a suitable ISCO reagent, will
soften the transition. This inclusion will provide degrees of freedom in the timing of the transition that
accommodates varied rates of performance across a site, for example, related to contaminant
distribution and natural heterogeneity, thus greatly facilitating the project management of the
integration transition.
Alternatively, switch points may be based on the achievement of transitional targets. This may be a
predetermined mass reduction or an intermediate concentration target that would ideally be based on
the tipping point of technology cost, as discussed earlier (concentration-based integration). In all
cases, the determination of the switch points should be established such that they result directly in
project cost reductions or project time-savings, or both. The following case studies are offered as
examples.
The reagent supply and application was cost-neutral: reagent and injection cost were equal to
operational cost of the DPE system over the treatment period, and were therefore offset by the
necessary extraction shutdown during the reaction phase prior to extracting the reagent-promoted
recoverable mass. The reagent-enhanced recovery promoted an 80% concentration reduction in four
weeks from the preceding six-month asymptote (Figure 6). There was no significant rebound this
was attributed to the removal of sorbed and smeared contamination and the establishment of a
Theme 1C: Strategies for remediation and brownfield regeneration
Lecture Session (ThS): 1C.7
favorable and abundant electron acceptor status for subsequent bio. Performance targets and
schedule were secured without introducing additional cost. Cost over-run and penalty charges were
avoided. The cost-benefit is therefore hard to enumerate but would be proportionally significant on
the basis of the previous asymptotic performance, should this have continued.
Case Two
A 10,000 100,000 g/L TCE plume originating from activities at a metalworking site in Wisconsin,
USA, extended through an area of 21 hectares (51 acres). A pump-and-treat (P&T) system was
installed in 2000 to secure the objectives of volatile organic compound (VOC) mass reduction and
hydraulic control. The plant had a capacity of 265 L/min (70 gpm), passing extracted water through
an air stripper followed by activated carbon. Post-installation operation and maintenance (O&M)
costs were approximately $200,000 per year, including energy costs of $10,000 per year. The plant
was successful in securing plume control, and through the first seven years of operation, removed a
total of 1,049 kg (2,312 lb) VOC, running at an average 93% capacity. A performance review at that
time projected that the strategy would not secure remedial targets within the next 30 years.
Evaluation of available remedial options appropriate for the site identified a strategy to transition from
the existing physical system through a program of in situ chemical oxidation (ISCO), enhanced natural
attenuation (ENA) and eventual monitored natural attenuation (MNA) as the optimum way forward.
The strategy built upon the containment and mass reduction secured by the initial P&T, and utilized
technology advances in ISCO and ENA that had become available since the initial system had been
commissioned. ISCO was undertaken through an injection campaign using a proprietary catalyzed
sodium percarbonate formulation (RegenOx , supplied by Regenesis, San Clemente CA) via 43 wells
in the remaining core plume areas. Thereafter, following a six-month aquifer re-equilibration period,
the controlled-release electron donors HRC-X and 3-D Micro-Emulsion (Regenesis, CA) were
applied through the same 43 core area wells plus an additional 100 direct push points in the
surrounding plume.
The ISCO campaign resulted in a source area VOC concentration reduction of 78 100% from 2007
levels which ranged from 10,000 200,000 g/L, securing the switch point target of 10,000 g/L in
all wells. Thereafter, the one-time electron donor application in the same area secured a further
reduction in target VOCs of 93% (average) over its first two years. On the basis of the results to
2011, the remedial completion projection was revised to 2016 from the previous P&T-based projection
Theme 1C: Strategies for remediation and brownfield regeneration
Lecture Session (ThS): 1C.7
which extended beyond 2041. The active groundwater extraction and treatment system was
deactivated in May 2011.
Overall, a projected cost saving in excess of $6M (at 2011 rates) was secured as a direct result of
transitioning technologies, with program completion brought forward by an estimated 25 years.
V. CONCLUSIONS
The use of technologies in appropriate spatial and/or sequential combination offers the potential of
securing improved remedial performance for reduced time, cost and resources. A single technology,
equally applicable and efficient at all stages of a project, may be considered the exception rather than
the rule.
The benefit of transition may arise from the natural synergy of one technology priming the way for
another, or may follow from the refinements in a conceptual site model (CSM) naturally secured
through monitoring the response of the aquifer to remedial intervention. The benefit of transition may
also simply occur through new and improved technologies becoming available during the course of
longer projects.
To regard transitions as an implicit mark of failure of the initial technology may be to disregard the
progression and facilitation secured within a complex dynamic system, and can indeed represent a
hindrance to securing the timely and effective transitions necessary for maintaining continued optimal
remedial efficiency on a site.
Prior identification of objective switch points can be helpful, but it is perhaps more important to
recognize the benefit of, and to anticipate, technology transitions as a natural counterpart of remedial
progression.
ACKNOWLEDGMENTS
2
Regenesis is grateful to GEO Ltd of Ferrybridge, UK, and to Symbiont, WI, USA for the case study
data presented in this paper.
REFERENCES
Krembs, F.J, W.S. Clayton, and M.C. Marley. (2011). Evaluation of ISCO Field Applications and
Performance. In R.L. Siegrist, M. Crimi and T.J. Sipkin (Eds.), In Situ Chemical Oxidation for
Groundwater Remediation. Pp.319-353. Springer Science+Business Media, LLC. New York,
Heidelberg, Dordrecht, London.
Mackay, D.M., and J.A. Cherry. (1989). Groundwater Contamination: Pump-and-Treat Remediation.
Environ. Sci. Technol. 23(6): 630-636.
Seyedabbasi, M.A., Newell, C.J., Adamson, D.T., & Sale, T.C. (2012). Relative contribution of DNAPL
dissolution and matrix diffusion to the long-term persistence of chlorinated solvent source zones.
Journal of Contaminant Hydrology, 134, 69-81.
-oO0Oo-
NOTES
Regenesis (www.regenesis.com) designs, produces and sells remediation products optimally designed for
combined use and seamless integration within the in situ remediation spectrum, from enhanced physical
recovery (PetroCleanze), through chemical oxidation (RegenOx, PersulfOx), to enhanced bioremediation
(aerobic ORC-Advanced and the ORC family; anaerobic 3D-MicroEmulsion and the HRC family), and
rapid risk-reduction / bio-enhancement / matrix back-diffusion control (PlumeStop).
Session 1C.9: Bioremediation of chlorinated solvents in groundwater 1
Water Technology Center, Department Environmental Biotechnology, Karlsruher Str. 84, 76139 Karlsruhe,
Germany; phone: +49 721 9678-223/-137; email: kathrin.schmidt@tzw.de, andreas.tiehm@tzw.de
Introduction
Trichloroethene (TCE) is a priority pollutant (EU, 2008; ATSDR, 2012; US-EPA, 2012) and among the
most frequently detected contaminants in groundwater (Conrad et al., 2010; Mattes et al., 2010). The
currently available bioremediation measures have certain drawbacks like e.g. the need for auxiliary
substrates (Figure 1). Oxidation of chloroethenes under aerobic conditions represents a promising way to
deal with the shortcomings of anaerobic reductive dechlorination (Tiehm and Schmidt, 2011).
TCE
Figure 2 Consumption of oxygen and auxiliary substrates (ammonia, methane and acetate)
during cometabolic and metabolic degradation of TCE.
Aliquots of the two groundwater microcosms with cDCE were transferred into chloride- and ammonium-
free mineral salts medium (Schmidt et al., 2014). This enriched mixed bacterial culture was maintained
active without any auxiliary substrate for five years, with ten transfers into fresh mineral salts medium
(10% inoculum). The only carbon substrate was TCE, which was added repeatedly after its degradation.
Results
In our study, the aerobic biodegradation of TCE as the sole growth substrate was demonstrated. This new
process of metabolic TCE degradation was first detected in laboratory groundwater microcosms (Figure
3). Stable carbon isotope fractionation was observed (Figure 4) providing a reliable analytical tool to
assess this new biodegradation process at field sites.
Figure 4 Compound specific stable carbon isotope fractionation (Rayleigh plot) during
aerobic trichloroethene (TCE) degradation in mineral salts medium (three replicate assays denoted
by different symbols). The error bars for the isotope analysis are smaller than the plotted symbols
(total analytical error: 0.5 ). Figure taken from Schmidt et al., 2014.
Further experiments with the enriched mixed bacterial culture in mineral salts medium showed sustained
long-term TCE biodegradation down to concentrations below the detection limit. Aerobic TCE degradation
resulted in stoichiometric chloride formation (Figure 5) and bacterial growth (Schmidt et al., 2014).
Figure 5 Aerobic trichloroethene (TCE) degradation and chloride formation in mineral salts
medium. (A) TCE concentration with repeated spiking, (B) cumulative TCE degraded and
cumulative chloride formed. The error bars represent 6 % and 1 % reproducibility on standard
analyses for TCE and chloride, respectively. Figure taken from Schmidt et al., 2014.
Aerobic metabolic TCE degradation might represent a new and promising concept for monitored natural
attenuation (MNA) approaches or engineered bioremediation of contaminated sites (enhanced natural
attenuation, ENA). Those remediation approaches can have advantages compared to reductive
dechlorination or cometabolic degradation and can be considered cost-efficient and environmental safe.
Based on these results the assessment of aerobic metabolic TCE degradation at field sites is highly
recommended. The observed stable carbon isotope fractionation provides a reliable analytical tool to
monitor and quantify aerobic oxidative degradation pathways in the field.
References
ATSDR - Agency for Toxic Substances & Disease Registry, 2012. Priority list of hazardous substances.
http://www.atsdr.cdc.gov/SPL/index.html. 21-6-2012.
Conrad, M.E., Brodie, E.L., Radtke, C.W., Bill, M., Delwiche, M.E., Lee, M.H., Swift, D.L., Colwell, F.S.,
2010. Field evidence for co-metabolism of trichloroethene stimulated by addition of electron donor to
groundwater. Environmental Science & Technology 44 (12), 4697-4704.
EU - European Union, 2008. Directive 2008/105/EC of the European Parliament and of the Council of
16 December 2008 on environmental quality standards in the field of water policy, amending and
subsequently repealing Council Directives 82/176/EEC, 83/513/EEC, 84/156/EEC, 84/491/EEC,
86/280/EEC and amending Directive 2000/60/EC of the European Parliament and of the Council. OJ L
348, 24.12.2008, p. 8497.
Mattes, T.E., Alexander, A.K., Coleman, N.V., 2010. Aerobic biodegradation of the chloroethenes:
pathways, enzymes, ecology, and evolution. FEMS Microbiology Reviews 34, 445-475.
Schmidt, K.R., Gaza, S., Voropaev, A., Ertl, S., Tiehm, A., 2014. Aerobic biodegradation of trichloroethene
without auxiliary substrates. Water Res. 59, 112-118.
Schmidt, K.R., Augenstein, T., Heidinger, M., Ertl, S., Tiehm, A., 2010. Aerobic degradation of cis-1,2-
dichloroethene as sole carbon source: Stable carbon isotope fractionation and growth characteristics.
Chemosphere 78 (5), 527-532.
Tiehm, A., Schmidt, K.R., 2011. Sequential anaerobic/aerobic biodegradation of chloroethenes - aspects
of field application. Current Opinion in Biotechnology 22 (3), 415-421.
Tiehm, A., Schmidt, K.R., Pfeifer, B., Heidinger, M., Ertl, S., 2008. Growth kinetics and stable carbon
isotope fractioning during aerobic degradation of cis-1,2-dichloroethene and vinyl chloride. Water
Research 42 (10-11), 2431-2438.
Acknowledgment
Financial support provided by the German Federal Ministry of Economics and Technology (grant number:
16224 N) is acknowledged. The authors thank Michael Deusch, Siegmund Ertl, Markus Friedrich, Holger
Hansel, Michael Heidinger, and Andrey Voropaev for their support.
CHALLENGES AND HOPES FOR SCALING UP AN ELECTRODIALYTIC
REMEDIATION METHOD FOR TREATING CCA CONTAMINATED SOIL
1 2 1
Krzysztof Piotr Kowalski *, Sanne Skov Nielsen , Pernille Erland Jensen , Thomas Hauerberg
2 3 3 1
Larsen , Mads Terkelsen , Carsten Bagge , Lisbeth Ottosen
1 Technical University of Denmark, Department of Civil Engineering, Kgs. Lyngby, Denmark
2 Orbicon A/S, Roskilde, Denmark
3 Center for Regional Development, Capital Region of Denmark
*main author, presenter, krkow@byg.dtu.dk
Keywords: Electrodialytic treatment, pilot plant, arsenic, copper, suspension, soil fine fraction, soil washing
INTRODUCTION
Soils polluted with heavy metals and arsenic compounds are a challenge for contaminated soil
handling procedures. Traditionally, a remediation of such material involves excavation of the
contaminated soil, followed by an immobilization of metal contaminants by solidification/stabilization
technology prior to disposal of the material in a permitted landfill site or on-site. However those
remedial actions based on solidification/stabilization technology are no longer considered a permanent
solution because the pollutants are not removed from the contaminated media resulting in a need for
future monitoring of a pollutants faith on site and questionable longevity of the solidified/stabilized
materials (Dermont et al., 2008).
Therefore, there is a great need for a novel treatment technology that can efficiently separate
heavy metals and arsenic from the soil. Among the technologies that have been in focus recently are:
1) flotation; 2) soil washing; 3) phytoremediation and 4) electrokinetic methods. All processes are
affected by soil properties and speciation of the polluting compounds and process efficiency may have
to be enhanced by chemicals addition, like surfactants, which were reported to improve washing and
flotation (Reynier et al., 2013; Torres et al., 2012).
Another approach could be a combination of two technologies, like presented in this paper
uniting soil washing and electrodialytic remediation. Application of the electrodialytic (ED) remediation
developed at the Danish Technical University has been well described in last decades, but was mostly
applied as laboratory and bench scale experiments with very limited mass of treated material (Ottosen
& Hansen, 1992). The ED method combines the technique of electrodialysis with the electromigration
of ions in the polluted soil. During the process it is possible to separate ions from soil with help of
applied electric field to electrodes isolated from soil by ion-exchange membranes. Anions, like arsenite
and arsenate, are removed to compartment with anode through a anion-exchange membrane and
cations, of copper and chromium, are moved into a compartment with cathode through a cation-
exchange membrane. Laboratory experiments proved that the ED remediation process applied on
saturated soil was very effective for the removal of copper and to a certain extent chromium (Hansen
et al., 1997). However As was not removed significantly in this system. This is mainly because the soil
is acidified during the remediation, and in acidified soil As will mainly be present as uncharged species
(H3AsO3 in case of As(III) and H3AsO4 in case of As(V)). Such uncharged species are not transported
in the applied electric field. Faster remediation was showed to take place in a stirred suspension of soil
than in the water saturated soil (Jensen et al., 2007; Ottosen et al., 2012). Moreover, increasing pH to
pH=3-4 significantly increased arsenic species removal reaching a removal efficiency of 60% (Sun et
al., 2012). It was also shown that remediation of the fine fraction of soil was faster and more energy
efficient than remediation of the whole soil. Therefore if the contamination is primarily bound to the fine
fraction of the soil, which is often the case, it may well be feasible to first implement soil washing to
separate the less contaminated coarse fractions from the highly contaminated sludge, and then
subject the fine fraction (sludge) to ED remediation, which would then be more efficient both due to the
smaller mass to be treated and due to the higher efficiency of the treatment (Jensen et al., 2006). The
concept was introduced earlier (Jensen et al., 2012). Another issue for application of the ED
remediation might be membrane fouling, that is a major factor to consider when applying
electrodialysis, especially in a full scale. Fouling is the accumulation of undesired deposits on the
membrane surface which affects its performance. Among the causes of these deposits scaling (salt
precipitation) and clogging (deposit of macromolecules, colloids) are anticipated to influence the ED
process the most (Araya-Farias and Bazinet, 2006; Mondor et al., 2009). Therefore the use of any
chemicals agents shall be limited, especially surfactants during the soil washing procedure as the
anion exchange membranes were reported to be sensitive to anionic organic compounds (Lee et al.,
2008)
This work presents some results from laboratory and pilot plant investigations of the technology
duet for decontamination of one of several former wood preservation plant left polluted by Cu, Cr and
As (CCA) in Denmark. CCA was originally used in the impregnation process, and hot spot
concentrations of copper, chromium and arsenic (1000-2000 mg Cu/kg of soil, 300-600 mg Cr/kg and
200-1200 mg As/kg) are clearly above the Danish clean soil criteria of 500 mg Cu/kg, 500 mg Cr/kg
and 20 mg As/kg. The Collstrop site only poses minimal risk to the nearby recipients and the
groundwater resource in the area, but it is a standard site for wood preservation. The site is therefore
used by The Capital Region of Denmark for testing remediation methods. The objective is that these
methods in the future can be applied on similar sites where remediation is required due to risks
towards groundwater and recipients.
The aim of this study is to discuss the feasibility of the soil washing and ED process
combination and its implementation for CCA contaminated soil treatment. The pilot scale
investigations gave an opportunity to evaluate especially the ED process, to define challenges that it
may face in the full scale and foremost outline directions for improvements to develop a cost-efficient
soil remediation technology.
METHODOLOGY
Soil sampling
Soil samples were obtained from two locations at the Collstrop contaminated site in Hillerd,
Denmark. The samples differed in contamination level. Soil samples from location A were mainly used
for lab scale investigations, and soil fines were obtained by wet-sieving the original soil through a
63m sieve with water and soil from location B was excavated and washed in a soil washing facility
Umweltschutz Ost GmbH in Germany. The initial concentrations are shown in Table 1. The soil
samples were air dried at room temperature prior laboratory experiments, while for pilot investigations
it was used as obtained from the soil washing facility.
During the soil washing in the German facility a high molecular weight flocculant, Magnafloc 120
L, was used to treat recycled water. Therefore the soil from location B was also sieved in laboratory by
wet-sieving and obtained fractions were mixed to get a fraction that mimics the soil washing without
undesirable compound.
Table 1 Heavy metals and arsenic content in Soil samples from the two different locations
Location A (the hot spot location) Location B excavated and washed in Germany
for lab scale investigations
Soil-A Fine-A Soil-B Fine-B Fine-M
Soil sieving results [% m/m]
X > 4mm 2.5 0.6
4mm > X >1mm 1.9 1.5
1mm > X >250 m 26.7 12.9 1.9
250 m > X> 125 m 39.2 26.0 5.3 7.6
125 m > X > 63 m 16.7 23.4 17.3 9.7
63 m > X 13.0 100 35.6 75.5 82.7
Heavy metals and arsenic content [mg/kg TS]
Cu 1130 90 8500 1600 183 80 260 10 480 4
Cr 300 20 1900 200 10 5 200 10 50 0.5
As 214 5 2080 400 220 80 1040 20 740 40
Analytical procedures
Solid samples of 1.0 g of dry soil and 20.0 mL (1:1) HNO3 were heated at200 kPa (120 C) for 30 min
according to Danish Standard 259. The liquid was separated from the solid particles by vacuum
through a 0.45 m filter and diluted to 100 mL. The concentrations of metals and As in prepared liquid
samples were measured with ICP (Varian 720-ES, software version: 1.1.0). Examination of the pH-
dependence of metals and arsenic mobilization from soil used the following procedure: 5.0 g of dry soil
and 25 mL of acid, HNO3 in various concentrations, or basic, NaOH in various concentrations,
solutions were used to make soil suspensions that were mixed for 7 days to reach equilibrium. After
this time the suspensions were filtered (0.45 m) and the filtrates were analyzed.
Electrodialytic remediation lab and pilot scale
The electrodialytic experiments in lab scale were conducted in cylindrical cells made from polymethyl
methacrylate and were built from 3 compartments; two compartments with electrolytes on sides
separated by ion exchange membranes from the middle compartment, where the soil slurry was
2
placed as shown on Figure 1. An anion exchange membrane (50 cm , AR204 SZRA MK III from
Ionics, GE Water & Process Technologies) was placed between anode (+) and middle compartment
2
and cation exchange membrane (50 cm , CR67 MK III from Ionics, GE Water & Process
Technologies) was placed between catode and middle compartment. Platinated rods from
Permascand were used as electrodes. Electrolytes (500 mL of 0.1 M NaNO3, pH= 2 0.2 adjusted
with HNO3) were circulated through the electrode compartments and the soil slurry in the middle
compartment remain suspended. Application of a DC electric field to electrodes immersed in the
electrolytes (compartment I and III) or directly in the treated slurry (compartment II) induces water
splitting reactions that depends on polarity of electrode:
+ -
- reaction on anode (+): 2 H2O O2(g) + 4H + 4e
-
- reaction on cathode (-): 2H2O + 2e- H2(g)+ 2OH
The ED cell can be used in two different setups, i.e. 3C and 2C that corresponds to use of all 3
compartments or only 2 compartments, when aiming to remove only one type of ions. Each concept
has its advantages and disadvantages, as for examples 2C setup was reported to have lower energy
consumption and shorter time for acidification when removing heavy metals from suspension of harbor
sediments and ashes (Kirkelund et al., 2015; Pedersen et al., 2014). Therefore a placing of electrode
in the slurries compartment enabled to apply current between the slurry and one electrolyte at a time,
depending what kind of ions were aimed to remove. To remove cations compartments II and III were
connected, where the electrode M was working as anode and the compartment I was inert in the
process. Similarly connection of compartment I and II, where the electrode M was working as catode
was used to remove anions from the slurry.
Figure 1 Principle of electrodialytic remediation of suspended soil slurry (AN anion exchange
membrane, CAT cation exchange membrane) (Sun et al., 2012)
The laboratory experiments were performed on 35 0.2 g of soil material that was suspended in 350
5 mL of water, resulting in liquid to solid (L/S) ratio 10. A current of 20 0.1 mA was used resulting in
2
a current density of 0.4 mA/cm . After end of the experiments removal efficiency for each element was
calculated as the mass of the actual heavy metal and arsenic in the electrode components
(membranes, solutions in electrode compartments and on electrodes) divided by the total mass of the
element found in all parts of the cell. Therefore the final results were divided into 4 groups that are
defined as following:
- Anode Comp. (+) sum of the element masses from electrolyte, anode and anion exchange
membrane;
- Catode Comp. (-) sum of the element masses from electrolyte, catode and cation exchange
membrane;
- Mobilized from soil mass of the element found in filtrate after passing 0.45 m filter;
- Left in soil mass of the element in dried soil after filtration with 0.45 m filter.
The upscaled electrodialytic remediation was following the same principles as the laboratory cell, but
enabled to treat 1000L of soil slurry. The compartments with ion exchange membranes where
submerged in the slurry that were mixed with help of compressed air applied from the bottom of the
2
tank, as shown on Figure 2. The upscaled plant used bigger membrane area, i.e. 2200 cm each, and
25L of electrolytes were circulated. There were used different initial volumes of slurry, but the L/S ratio
was the same as in the lab experiments.
The energy consumption of the ED process can be calculated by equation E x = VIdt/W, where
E is the power consumption per miligram of removed element (Wh/mgremoved x); V - voltage between
working electrodes (V); I - current (A); t - time (h); W - the mass of the element found in the
electrolytes compartment (mg). For the upscaled treatment plant power meters (SparOmeter kWh
meter from Elma Instruments) were used to measure a factual energy consumption of all electrical
equipment, also for electrolytes pumps and stirring systems.
Figure 2 Left: The prototype bench scale electrodialytic treatment with aeration system (in a bottom),
electrode cassettes and an electrode liquid recycling system (green with cation exchange membranes
and catholyte and yellow with anion-exchange membranes and anolyte). Right: Conceptual model of
electrochemical batch. Only the setup for removal of cations, Cr and Cu, is shown. For removal of As, the
electrodes and membrane are shifted.
RESULTS
Soil characteristic and implications on the treatment procedure
The particle size distribution obtained by wet-sieving of soil A indicates that the soil consisted
mostly of size fraction between <1-0.125 mm (Figure 3). However these fractions are the least
contaminated by As, Cr and Cu. The highest contents of Cu, exceeding the Danish clean soil criteria
of 500 mg/kg, were in sand (> 1mm) and finest fractions (<0.063 mm). All fractions surpass As the
criteria that is 20 mg/kg, but the highest As content was found in the finest fractions. Chromium
exceeding the clean soil criteria was found only in the finest fraction. The presence of the
contaminants in coarse fractions can be a consequence of organic material occurrence, especially
preserved wood particles. Though accumulation of contaminants in the finest fraction of soil brings two
advantages, enable to reduce mass of treated material and increase initial contaminants content.
Therefore washing of soil B was performed on 14 tonnes of material and was aiming to obtain the
finest fraction that would have been used in the pilot scale investigations. The outcome of soil washing
was production of 4.4 tonnes of fine fraction (fine-B), which properties can be found in Table 1. The
fine-B material was only exceeding the clean soil criteria for arsenic content, as the soil-B did not have
high Cr and Cu concentration when compared to soil-A.
Figure 3 Particle size distribution (a) and concentration of contaminants in each fraction size for soil
sample A (As, Cr and Cu soil limits refer to Danish clean soil criteria)
The pH of slurry is a crucial parameter for electrodialytic remediation as it affects heavy metals
and arsenic release. The As mobilization dependences on pH for fine fractions, obtained from the soil
washing facility and from wet sieving to mimic the one from the washing facility, are presented on
Figure 4. The differences in the As mobilization are significant and are most probably caused by the
fact that the fine fraction from the washing facility was mixed with settled material from treating
recirculated water used for washing. The recirculating water treatment was based on flocculation with
commercially available agent, Magnafloc 120. Unfortunately the use of the flocculation could have not
been avoided. However results obtained from fine-M material show more efficient As mobilization from
the fraction depending on the pH. To achieve more than 80% of As mobilization it is necessary to use
either very low pH<1 or high pH, i.e. above 12. The mobilization of arsenic compounds at low pH may
be due to dissolution of contaminant bearing minerals and aluminum and iron hydroxides that are
known for their adsorption capacities (co-precipitation). However the low pH conditions are limiting
occurrence of arsenic anions, as its species starts to dissociate at pH above 2 (pKa1 of arsenate-
H3AsO4 is 2.2 and pKa1 of arsenite-H3AsO3 is 9.3). Applying the basic conditions, especially pH above
12, reveals a different pattern for contaminants mobilization. At high pH the arsenic distinguish oneself
as the most mobilized. The phenomena may be explained with desorption of As anions from
amorphous iron oxide that starts at pH above 9 (Dixit and Hering, 2003).
Figure 4 Figure 4 Dependency of pH on As mobilization from fine fractions B (from soil washing facility)
and M (from wet sieving in lab)
Left in soil Mobilized from soil Catode Comp. (-) Anode Comp. (+)
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
As Cr Cu Fe
Figure 5 Performance of ED laboratory experiment, where cations were removed at low pH, in a first
phase lasting for 2 days, and anions at high pH, in a second phase lasting for 15 days
60%
50%
40%
30%
20%
10%
0%
Fine-B Fine-M
Figure 6 Comparison of ED laboratory experiment performed on fine-B, fine fraction from soil washing
facility, and fine-M, fine fraction from laboratory sieving. The experiments were performed in 2C cell
setup, where slurry were kept at high pH 10-11
Figure 7 Variations of main parameters during ED pilot remediation applied for fine-B suspensions (90kg
of soil fraction in 880 L of water) with and without NaOH initial addition: applied current (a) to electric
field; pH (b) and conductivity (c) of suspensions; concentration of As mobilized to suspensions water
phase (d) and total of As removed to electrolytes (e)
Figure 8 Comparison of energy of the ED processes used for removal of Cu (a) and As (b) from soil
material suspensions in laboratory cell (Lab fine-A and Lab fine-B) and in pilot runs with soil A (Pilot
soil-A - refers to energy used since suspension was acidified to pH around 2) and fine B with NaOH
addition (Pilot Fine-B+NaOH) and without any chemicals addition to the suspension (Pilot Fine-B),
where only As was aimed for removal
Conclusions
Laboratory studies proved that it is possible to apply electrodialytic remediation for separation of
arsenic and copper from soil materials. Chromium compounds were removed in a minor degree. The
amount of treated material can be reduced with help of soil washing that enables to separate the most
contaminated soil fraction, which was found to be the finest fraction (<0.063 mm).
However different approaches have to be applied to remove cations and anions, especially when
scaling up the process. To remove cations it is necessary to mobilize them at low pH, but for anions
removal, especially arsenic the high pH of the suspension has to be used to have sufficient anions
mobilization. The factors influencing the ED pilot scale, beside pH of treated suspension, are the
stirring routine to maintain material suspension, solution used for electrolyte and agents added during
soil washing, which can be eliminated when performing the washing on-site. Basing on experiences
from this and previous ED-based remediation processes an upscaling of the process has been
proposed to treat contaminated soil according to following on-site procedure:
1. Soil excavation.
2. Separation and soil fractionation in a soil washing facility on-site.
3. Performing on-site EDI remediation on the fine fraction slurry
4. Soil regeneration.
Acknowledgement
The project is a collaboration between DTU-Byg at the Technical University of Denmark and Orbicon
A/S with project owner Center for Regional Development, Capital Region of Denmark. We would like
to acknowledge Center for Regional Development, Capital Region of Denmark for funding the project
and their support.
Appendix
Picture 1 Pilot plant with 5 1000L tanks for soil fraction slurry
References
Araya-Farias, M., Bazinet, L., 2006. Effect of calcium and carbonate concentrations on anionic membrane fouling
during electrodialysis. J. Colloid Interface Sci. 296, 242247. doi:10.1016/j.jcis.2005.08.040
Dermont, G., Bergeron, M., Mercier, G., Richer-Laflche, M., 2008. Soil washing for metal removal: A review of
physical/chemical technologies and field applications. J. Hazard. Mater. 152, 131.
doi:10.1016/j.jhazmat.2007.10.043
Dixit, S., Hering, J., 2003. Comparison of arsenic (V) and arsenic (III) sorption onto iron oxide minerals:
Implications for arsenic mobility. Environ. Sci. Technol. 37, 41824189.
Hansen, H.K., Ottosen, L.M., Kliem, B.K., Villumsen, A., 1997. Electrodialytic remediation of soils polluted with
Cu, Cr, Hg, Pb and Zn. J. Chem. Technol. Biotechnol. 70, 6773. doi:10.1002/(SICI)1097-
4660(199709)70:1<67::AID-JCTB662>3.0.CO;2-V
Jensen, P.E., Ahring, B.K., Ottosen, L.M., 2007. Organic acid enhanced electrodialytic extraction of lead from
contaminated soil fines in suspension. J. Chem. Technol. Biotechnol. 82, 920928. doi:10.1002/jctb.1762
Jensen, P.E., Ottosen, L.M., Allard, B., 2012. Electrodialytic versus acid extraction of heavy metals from soil
washing residue. Electrochim. Acta 86, 115123. doi:10.1016/j.electacta.2012.07.002
Jensen, P.E., Ottosen, L.M., Pedersen, A.J., 2006. Speciation of Pb in industrially polluted soils. Water. Air. Soil
Pollut. 170, 359382. doi:10.1007/s11270-005-9008-7
Kirkelund, G.M., Magro, C., Guedes, P., Jensen, P.E., Ribeiro, A.B., Ottosen, L.M., 2015. Electrodialytic removal
of heavy metals and chloride from municipal solid waste incineration fly ash and air pollution control residue
in suspension test of a new two compartment experimental cell. Electrochim. Acta.
doi:10.1016/j.electacta.2015.03.192
Lee, H.J., Hong, M.K., Han, S.D., Shim, J., Moon, S.H., 2008. Analysis of fouling potential in the electrodialysis
process in the presence of an anionic surfactant foulant. J. Memb. Sci. 325, 719726.
doi:10.1016/j.memsci.2008.08.045
Mondor, M., Ippersiel, D., Lamarche, F., Masse, L., 2009. Fouling characterization of electrodialysis membranes
used for the recovery and concentration of ammonia from swine manure. Bioresour. Technol. 100, 566
571. doi:10.1016/j.biortech.2008.06.072
Ottosen, L.M., Jensen, P.E., Kirkelund, G.M., Dias-, C., Hansen, H.K., 2012. Electrodialytic Remediation of heavy
Metal polluted Soil treatment of water saturated or suspended soil 28, 103108.
Pedersen, K.B., Ottosen, L.M., Jensen, P.E., Lejon, T., 2014. Comparison of 2-compartment, 3-compartment and
stack designs for electrodialytic removal of heavy metals from harbour sediments. Electrochim. Acta.
doi:10.1016/j.electacta.2014.12.003
Reynier, N., Blais, J.-F., Mercier, G., Besner, S., 2013. Treatment of Arsenic-, Chromium-, Copper- and
Pentachlorophenol-Polluted Soil Using Flotation. Water, Air, Soil Pollut. 224, 1514. doi:10.1007/s11270-
013-1514-4
Sun, T.R., Ottosen, L.M., Jensen, P.E., Kirkelund, G.M., 2012. Electrodialytic remediation of suspended soil--
Comparison of two different soil fractions. J. Hazard. Mater. 203-204, 22935.
doi:10.1016/j.jhazmat.2011.12.006
Torres, L.G., Lopez, R.B., Beltran, M., 2012. Removal of As, Cd, Cu, Ni, Pb, and Zn from a highly contaminated
industrial soil using surfactant enhanced soil washing. Phys. Chem. Earth, Parts A/B/C 37-39, 3036.
doi:10.1016/j.pce.2011.02.003
Full-Scale Electrokinetics-Enhanced Bioremediation (EK-BIO) of PCE
DNAPL Source Area in Clay Till
Charlotte Riis (cer@niras.dk), Martin Bymose, and Dorte Pade (NIRAS A/S, Alleroed,
Denmark)
Evan Cox (Geosyntec Consultants, Waterloo, Ontario, Canada)
James Wang (Geosyntec Consultants, Columbia, Maryland, USA)
David Gent (US Army Engineer R&D Center, Vicksburg, Mississippi, USA)
Mads Terkelsen (Capital Region, Hilleroed, Denmark)
In 2011, a field pilot test was carried out at the Skuldelev site, Denmark to assess the ability of the
novel EK-BIO technology to treat PCE DNAPL source material in clay till with interbedded deposits
of sand. The EK-BIO pilot test demonstrated the transport and distribution of amendments (lactate
and microbial culture KB-1) through clay soils in the target area. Results from groundwater and
clay soil sampling showed significant reductive dechlorination of PCE to cisDCE, VC, and ethene
coupled with significant levels of dechlorinating microorganisms (Dhc with vcrA), indicating that PCE
dechlorination in clay soil was achieved by EK-BIO with KB-1 bioaugmentation. Based on the
successful pilot test, a full-scale remediation was implemented.
Approach/Activities. The full-scale EK-BIO implementation targeting a PCE source area at the
Skuldelev site was initiated as a Worlds First in December 2012. The treatment zone, addressed by
a network of 15 electrode wells, covers an area of 100 m2 to a maximum depth of 10 m bgl. The
overall area is divided into two sub-areas, which are treated in alternating stages, each for a period
of three months. The polarity of the electrodes is changed to alter the current directions and electric
field orientations in alternating stages in order to optimize the EK transport efficiency and to achieve
treatment of the entire target zone for remediation. Performance monitoring comprises monthly
water sampling for TOC and field measurements, as well as quarterly water sampling for complete
characterization of contaminant composition and degradation processes. In addition, soil sampling is
also performed at the end of select stages of operation to assess treatment in clay materials.
In this work we investigate an innovative cleanup approach based on the combination of chemical
oxidation by H2O2 (Fenton-like process) with carbon dioxide sparging for the treatment of groundwater
contaminated by organic compounds. The performance of this process was evaluated through slurry-
phase lab-scale tests performed on a soil-water system spiked with MtBE, using a given H2O2
concentration (6%) while fluxing CO2 for a fraction of the whole reaction time, but without any other
amendment. The obtained results were compared with those achieved applying H2O2 (6% concentration)
alone or amended with either EDTA used as chelating agent, or a combination of EDTA and KH2PO4 , the
latter used as H2O2 stabilizing agent. Furthermore, carbon dioxide sparging without H2O2 was also tested
and used as reference to evaluate the extent of MtBE stripping by CO2. The obtained results in terms of
H2O2 stability, showed that the hydrogen peroxide lifetime of the combined H2O2-CO2 process was longer
than the one observed in the tests performed with H2O2 or H2O2-EDTA, but shorter than the one observed
using KH2PO4 and EDTA. The removal efficiency achieved by the proposed new Fenton-like formulation
was significantly higher than the one obtained using KH2PO4 and EDTA and rather close to the one
observed using EDTA alone as chelating agent. The obtained results showed that fluxing CO2 into the
oxidizing solution has basically two effects which may explain the good performance achieved: on the one
hand, it leads to carbonic acid ionization with a consequent decrease of pH to acidic values, that
increases the H2O2 stability and makes the Fentons process more effective.; on the other hand, the CO2
sparging enhances the contaminant removal by stripping. Overall, these preliminary results showed that
the combination of hydrogen peroxide with CO2 sparging could represent a feasible alternative to
traditional Fenton-like formulations, allowing to avoid the use of persistent amendments such as EDTA,
while leading to comparable removal efficiencies.
KEYWORDS: Groundwater contamination, In situ chemical oxidation, CO2 sparging, MtBE, Hydrogen
Peroxide.
1 Introduction
In situ chemical oxidation (ISCO) is one of the main clean-up options for the treatment of groundwater
contaminated by a wide range of organic contaminants (Krembs et al., 2010). This remediation technique
involves injecting oxidants (such as hydrogen peroxide, potassium permanganate or sodium persulfate)
into the subsurface to destroy the compounds of concern (ITRC, 2005). Many studies so far have
demonstrated the effectiveness of the ISCO processes in contaminated site remediation (Siegrist et al.
2011). Among the available oxidation processes, the Fentons one gained an increasing interest due to its
ability of treating a wide range of contaminants (Huling and Pivetz, 2006).
In the classic Fentons reaction, hydrogen peroxide reacts with Fe2+ ions to form the hydroxyl radical OH:
(1)
H 2O2 Fe2 OH OH Fe3
In this process, an external source of iron is needed and the reaction environment needs to be acidified in
order to keep the iron in solution, thus avoiding its precipitation. Fenton-like process are typically
performed at neutral pH, without any iron addition, as this process relies on the iron and manganese
minerals naturally occurring in the soil; chelating agents, such as Ethylenediaminetetraacetic acid (EDTA)
are used in this case to extract and keep in solution the transition metals of the soil used as catalyst in the
Fenton-like process (Vicente et al. 2011), whereas stabilizing agents are added in order to increase the
lifetime of hydrogen peroxide which would be otherwise quite short at neutral pH conditions (Watts and
Teel, 2005). Actually, the hydrogen peroxide lifetime in the subsurface is a key factor to be considered in
the design of ISCO based on this oxidant, as it affects the contact time of the oxidant with the contaminant
and ultimately the delivery of H2O2 in the subsurface (Baciocchi et al. 2014). With respect to other
oxidants, such as potassium permanganate or persulfate, hydrogen peroxide persists in soil and aquifer
for relatively short times (from minutes to hours) thus potentially limiting the radius of influence (ROI) of
the treatment (Innocenti et al. 2014). To overcome this limitation, various reagents that enhance the
lifetime of H2O2 in the aquifer can be used (Baciocchi et al. 2003). Namely, the most common H2O2
stabilizers involves various forms of phosphate that reduce the availability of inorganic reactants
responsible of H2O2 disproportionation via complexation or precipitation reactions (Petri et al. 2011).
The use of these amendments is often questioned by regulatory agencies for their secondary impacts on
the environment. For instance EDTA is blamed for its excessive persistence in the environment due to its
low biodegradability (Tandy et al. 2004). To overcome this limitation, the use of other organic acids, such
as phytate, citrate and malonate, was recently proposed (Watts et al. 2007, Schmidt et al. 2011).
In this work, we investigate an innovative formulation of the Fenton-like process, based on the
combination of chemical oxidation by H2O2 with carbon dioxide sparging, that avoids the use of persistent
amendments.
The performance of this process was evaluated through slurry-phase lab-scale tests performed on a soil-
water system spiked with MtBE, using a given H2O2 concentration (6%) while fluxing CO2 for a fraction of
the whole reaction time, but without any other amendment. The obtained results were compared with
those achieved applying H2O2 (6% concentration) alone or amended with either EDTA used as chelating
agent, or a combination of EDTA and KH2PO4, the latter used as H2O2 stabilizing agent. Furthermore,
carbon dioxide sparging without H2O2 was also tested and used as reference to evaluate the extent of
MtBE stripping by CO2.
2.1 Reagents
Hydrogen peroxide (H2O2, 30% w/w solution) from Sigma -Aldrich was used in the different batch tests.
Titanium (IV) oxysulphate and Sulphuric acid (96%) from Sigma-Aldrich was used in the determination of
hydrogen peroxide. Ethylenediaminetetraacetic acid (EDTA), Methyl Tertiary Butyl Ether (MtBE 99,8%),
fluorobenzene and Sodium chloride were purchased from SigmaAldrich. All solutions were prepared with
deionized water.
The soil used in this study is a top soil collected in an area located in the centre of Italy. The soil was
sieved to separate the coarser fraction (d > 2 mm), corresponding to approximately 32% by weight of the
natural soil. The fraction below 2 mm used in the different ISCO tests was mixed and homogenised using
a quartering procedure and its particle size distribution was evaluated by applying the ASTM D422
standard procedure. This fraction was characterized measuring the total organic carbon (TOC), the soil
oxidant demand (SOD), the iron, manganese, chlorides and sulphates content. Table 1 summarizes the
main chemical and physical properties of the soil fraction below 2 mm.
Two sets of slurry-phase batch experiments were carried out. A first set (Table 2a) was performed to
evaluate the stability of hydrogen peroxide in a soil-water system using EDTA alone, EDTA with KH2PO4
or CO2 as amendments. In addition, a test using hydrogen peroxide without adding any amendment was
also performed and used as reference. Then, to evaluate the contaminant removal achievable with the
different combinations of amendments, a second set of slurry-phase tests was carried out on a soil-water
system artificially contaminated by MtBE (Table 2b). As control, a further test applying only carbon dioxide
sparging without H2O2 was performed in order to evaluate the extent of MtBE stripping achievable by CO2
fluxing.
All tests were carried out in triplicate at 20 2 C in 250 mL glass bottle sealed with a rubber cap using 50
g of air dried soil and applying a 5 l/kg liquid to solid ratio. The soil was mixed with the solution (i.e. water
or a MtBE solution depending on the type of test) for 5 minutes and then hydrogen peroxide was added
determining the beginning of the experiment. In the tests with EDTA and KH2PO4 both amendments were
added 24 h before adding H2O2. In the tests performed with carbon dioxide, instead, before adding
hydrogen peroxide the solution was saturated with CO2 by fluxing 1000 ml/min for 20 minutes through a
ceramic septum. In each test, the initial hydrogen peroxide concentration was equal to 6 % (w/w). All other
conditions of the different tests are reported in Table 2. The soil-water system was mixed through the
entire experiment duration using a magnetic stirrer set at 350 rpm. Before sampling, the mixing was
stopped for a few minutes to allow settling of the soil particles and the formation of a clear supernatant
layer. In the H2O2 stability tests a liquid sample was collected at 2, 4, 6, 8, 24 and 48 h and then analysed
for H2O2 concentration. In the oxidation experiments a liquid sample was collected after 24 h and added to
a 20 ml glass vial for GC-MS analysis.
The hydrogen peroxide concentration was measured using an UV-VIS spectrophotometer TU-18805
Double Beam after colour development with titanium sulfate technique. H2O2 concentration was measured
in duplicate to obtain an average value. MtBE in the water samples collected in the different batch tests
was measured by static headspace gas chromatography-mass spectrometry detection (GC-MS) with
fluorobenzene as internal standard. For headspace analysis, 10 ml samples were stirred for 40 min at 70
C in 20 ml headspace vials containing 5 g NaCl to increase vapour phase concentration by the salting-
out effect. The headspace gas (250 l) was injected (split 1:10) with an auto-sampler (AOC-5000 plus)
into a Shimadzu GC/MS (GC_QP2010SE) equipped with a TR-5MS column (30 m x 0.25 mm ID x 0.25
um film thickness: 5% Phenyl (equiv) Polysilphenylene siloxane). Helium (0.9 ml/min) was used as
carrier gas. Initial oven temperature was 35 C, which was held for 3 min, and then ramped to 180 C (at a
30 C/min rate) and maintained for other 2 min. The operating MS settings were as follows: Ionization
voltage, 0.87kV; interface temperature, 230C; detector temperature, 200C. The selected ions monitoring
(SIM) acquisition mode was used and the ions fragments selected were: 73 (MtBE) and 96
(fluorobenzene). All GC-MS analysis were performed in triplicate to obtain an average value.
3 Results
Figure 1 reports the decomposition kinetics of hydrogen peroxide observed in the tests carried out using
different amendments for H2O2 stabilization. Furthermore, the results obtained using hydrogen peroxide
without adding any amendments are also reported as reference. From this figure it can be noticed that
using the oxidant without any amendment led to a rapid decomposition of H2O2 after few hours from the
beginning of the test. The addition of KH2PO4, EDTA, CO2 or a combination of them significantly
enhanced the hydrogen peroxide lifetime. This is particularly evident when both KH2PO4 and EDTA were
added to the oxidant solution. Indeed, the results reported in Figure 1 clearly show that for this
formulation, 40% of the initial hydrogen peroxide concentration (i.e. C/C0 = 0.4) was still detected in
solution after a contact time of 48 hours. On the contrary, the addition of EDTA or CO2 exerted a
significant effect on H2O2 stabilization only in the early stage of the test (<10 h) where it can be observed
that the decomposition kinetic was significantly slowed down with respect to the test with H2O2 only.
Differently, for extended contact times (e.g. 48 hours) the addition of EDTA or CO2 was not effective
anymore and the oxidant concentration approached the values obtained using hydrogen peroxide without
adding any amendment.
1.0
H2O2
H2O2
0.9
H2O2+EDTA
H2O2+EDTA
0.8 H2O2+EDTA+KH2PO4
H2O2+EDTA+KH2PO4
0.7 H2O2+CO2
H2O2+CO2
0.6
C/C0
0.5
0.4
0.3
0.2
0.1
0.0
0 10 20 30 40 50
Time (h)
Figure 1: Kinetics of H2O2 decomposition observed in the tests performed using different amendments. The results
are expressed in terms of normalized hydrogen peroxide concentration (C/C0).
As shown in Figure 2, the different amendments used had a different effect on the solution pH, which was
around 7 in the test where H2O2 only was used, whereas it was reduced below 5 when KH2PO4 or CO2
were used as amendments.
8
6
pH
H2O2
H2O2
3
H2O2+KH2PO4
H2O2+KH2PO4
H2O2+CO2
H2O2+CO2
2
0 5 10 15 20 25
Time (h)
Figure 3 reports the MtBE removal efficiency observed at the end of the tests carried out using the
different amendments. Besides, the results obtained using hydrogen peroxide without adding any
stabilizer are also reported as reference. Finally, in order to evaluate the stripping effect of CO2, the MtBE
removal observed applying only carbon dioxide sparging without H2O2 are also shown in the same figure.
Making reference to the obtained results, it can be noticed that the higher removal efficiency (i.e. up to
95%) was achieved in the test carried out using hydrogen peroxide and EDTA. The test performed with
the use of H2O2 alone showed a removal efficiency of 70% and no significant improvement was observed
in the test carried out using both KH2PO4 and EDTA. This in principle could suggest that the best oxidant
formulation to treat groundwater contaminated by MtBE is the one based on the use of EDTA alone.
However, for a full-scale application, where the oxidant delivery in the subsurface is commonly
accomplished through permeation by vertical groundwater wells, it should be considered that with the
formulation based on EDTA alone, the hydrogen peroxide lifetime (see Figure 1) and consequently the
radius of influence (ROI) of each injection well would be very limited, i.e. requiring a layout of the injection
system characterized by a high number of injection wells. In this view, although characterized by a lower
removal efficiency, the use of both KH2PO4 and EDTA could be preferable especially in the case of large
contaminated areas to be treated, as it significantly enhances the H2O2 lifetime and consequently would
allow for a relevant increase of the ROI of an injection well.
As to the new process presented in this work, the obtained results show that the combination of the
Fenton-like process with CO2 sparging was very effective leading to a reduction of MtBE up to 90%.
Furthermore, the test performed only with CO2 sparging showed that in that case the MtBE removal was in
the order of 20%, thus suggesting that the effect of stripping was rather limited and consequently that the
removal observed in the combined process (H2O2 + CO2) can be mainly ascribed to oxidation by hydrogen
peroxide. These preliminary findings are very interesting and encouraging for further developments of this
process. Indeed, in a field-scale application the combination of hydrogen peroxide with CO2 sparging could
in principle avoid the use of persistent amendments such as EDTA while leading to comparable removal
efficiencies.
100%
90%
80%
70%
MtBE Removal
60%
50%
40%
30%
20%
10%
0%
H2O2
H2O2 H2O2 + EDTA
EDTA+H2O2
H2O2 + EDTA+KH2PO4
EDTA+H2O2+KH2PO4 H2O2+CO2
H2O2+CO2(20min)
CO2
CO2(20min)
Figure 3: Comparison of MtBE removal observed at the end of the tests performed using different
amendments.
4 Conclusions
The results obtained in this work suggest that a Fenton-like oxidation system combined with carbon
dioxide sparging can be a promising remediation option for the treatment of organic compounds in
groundwater. The results obtained using the new combined process were comparable or even better than
the one obtained with traditional formulations, based on the use of amendments. Namely, the hydrogen
peroxide lifetime measured in slurry-phase tests was higher than the one observed for EDTA alone but
lower than the one observed using KH2PO4 and EDTA, although this result was affected by the short time
chosen for CO2 fluxing, i.e. 20 minutes compared to 24 hours of test duration. The MtBE removal
efficiency achieved by the new formulation (90%) was significantly higher than the one obtained using
KH2PO4 and EDTA (70%) and comparable, although slightly lower, with the one observed using EDTA
alone as chelating agent (95%). CO2 fluxing into the oxidizing solution was shown to have basically two
effects which may explain the good performance achieved: on the one hand, it leads to carbonic acid
ionization with a consequent decrease of pH to acidic values, that increases the H2O2 stability and makes
the Fentons process more effective. On the other hand, the CO2 sparging enhances the contaminant
removal by stripping.
Overall, these preliminary results showed that the combination of hydrogen peroxide with CO2 sparging
could represent a feasible alternative to traditional Fenton-like formulation, allowing to avoid the use of
persistent amendments, such as EDTA, while leading to comparable removal efficiencies.
References
1. Baciocchi, R., Boni, M. R., DAprile, L. (2003). Hydrogen peroxide lifetime as an indicator of the
efficiency of 3-chlorophenol Fentons and Fenton-like oxidation in soils. J. Hazard. Mater. 96(2),
305-329.
2. Baciocchi, R., DAprile, L., Innocenti, I., Massetti, F., Verginelli, I. (2014). Development of technical
guidelines for the application of In-Situ Chemical Oxidation to Groundwater remediation, J. Clean.
Prod. 77, 47-55.
3. Huling, S.G., Pivetz, B.E. (2006). In-situ chemical oxidation. USEPA/600/R-06/072.
4. Innocenti, I., Verginelli, I., Massetti, F., Piscitelli, D., Gavasci, R., Baciocchi, R.(2014). Pilot-scale
ISCO treatment of a MtBE contaminated site using a Fenton-like process, Sci. Total Environ. 485-
486, 726-738.
5. ITRC (2005). Technical and Regulatory Guidance for In Situ Chemical Oxidation of Contaminated
Soil and Groundwater - Second Edition. I.T.R.C.
6. Krembs F.J., Siegrist R.L., Crimi M.L., Furrer R.F., Petri B.G. (2010). ISCO for Groundwater
Remediation: Analysis of Field Applications and Performance. Ground Water Monit R 30: 42-53.
7. Petri, B.G., Watts, R.J., Teel, A.L., Huling, S.G., Brown, R.A. (2011). Fundamentals of ISCO using
hydrogen peroxide, in: In Situ Chemical Oxidation for Groundwater Remediation, Springer New
York, pp. 33-88.
8. Schmidt, J. T., Ahmad, M., Teel, A. L., Watts, R. J. (2011). Hydrogen peroxide stabilization in one-
dimensional flow columns. Journal of contaminant hydrology, 126(1), 1-7.
9. Siegrist R.L., Crimi M., Simpkin T.J. (2011). In situ chemical oxidation for groundwater
remediation. Springer New York (USA).
10. Tandy, S., Bossart, K., Mueller, R., Ritschel, J., Hauser, L., Schulin, R., Nowack, B. (2004).
Extraction of heavy metals from soils using biodegradable chelating agents. Environmental
Science & Technology, 38(3), 937-944.
11. Vicente, F., Rosas, J. M., Santos, A., Romero, A. (2011). Improvement soil remediation by using
stabilizers and chelating agents in a Fenton-like process. Chem. Eng. J. 172(2), 689-697.
12. Watts R.J., Teel A.L. (2005). Chemistry of Catalyzed H2O2 Propagations (Modified Fenton's
Reagent) for In Situ Soil and Groundwater Remediation. J. Environ. Eng. 131, 612-622.
13. Watts, R. J., Finn, D. D., Cutler, L. M., Schmidt, J. T., Teel, A. L. (2007). Enhanced stability of
hydrogen peroxide in the presence of subsurface solids. Journal of contaminant hydrology, 91(3),
312-326.
Theme 1. Dealing with contamination of soil, groundwater and sediment
1c. remediation technologies and approaches
ABSTRACT:
Overvieuw
A special category of soil remediation covers sites that have been contaminated by the use of
solvents, especially chlorinated ethenes. These contaminants can be removed by enhanced natural
attenuation using the so-called TCE concept. Artificially cultivated bacteria (Dehalococcoides
ethenogenes, DHC) with the required carbon source and nutrients are introduced into the soil in order
to stimulate anaerobic degradation. The process can be controlled by constantly monitoring the
influent, infiltrate and bioreactor streams for pH, oxygen and ORP. The strength of this concept lies in
its relatively short active degradation phase with complete degradation of the chlorinated
hydrocarbons to harmless end products.
The general remedial approach involves extraction, amendment and re-circulation of groundwater
in the targeted aquifer zone. Extracted groundwater is passed through an anaerobic bioreactor to
enrich the water with DHC. After filtration, the feed is infiltrated through injection screens or wells
located up gradient of the treated zone. This anaerobic bioremediation approach can be applied in
urbanized areas with limited site access to prevent site disruption and reduce impacts to the
environment. This semi-passive method can expedite the time of remediation, compared to passive
approaches. Over the last ten years, NTP Enviro has optimized the use of the bioreactor system in
several bio-remediation projects, including the development of an automated application of the carbon
source. The fully automated remotely controlled bioreactor system provides clients with a complete
biodegradation of chlorinated ethenes.
Specifications
The system consists of a control unit, a bioreactor and a filtration and charging system. Once the DHC
inoculum is added to the bioreactor, their growth is enhanced by feeding the system, until a cell count
5
of 10 cells/ml is achieved. Bioreactor operating conditions include ORP levels of less than -300 mV,
while oxygen levels are reduced to zero at a temperature of about 20 C. The bioreactor operates at
3
an extraction/infiltration flow rate of about 10 m /hr. Depending on the size of a pore volume for the
targeted zone, the system may be able to treat a site within a period of several months. Based on
previous experiences, amendment of one pore volume of the targeted aquifer zone is sufficient for
complete chlorinated ethenes treatment at most sites.
Process and performance monitoring
Monitoring our process and performance is crucial to the success of the enhanced bioremediation
projects. Therefore, defined parameters such as oxygen concentration, pH, water temperature (C)
and oxidation-reduction potential (ORP) are monitored continuously. When the parameters exceed
threshold levels, the bioreactor is automatically switched off. Monitoring wells in the field are used to
assess the efficacy of enhanced bioremediation. The design and operation of this in-situ system of
extraction and infiltration can be adjusted depending on the volumes of groundwater to be treated, site
specific circumstances and remedial targets to be met.
Technology application
Leakages that had occurred because of a former dry cleaner in the center of The Hague in The
Netherlands left the shallow and deeper groundwater up to 15 m gl contaminated with chlorinated
solvents (PCE and TCE).
Soil properties, in the area of The Hague, are characterized by sediments of medium fine sand and
heterogeneous intermediate layers of clay and/or peat. Groundwater contamination plumes had
3
spread to a volume of 400,000 m underneath private properties and form a thread for the deeper
aquifers and the abstraction of drinking water.
We opted for a phased approach, with the infiltration and extraction filters all in the public road. The
plume was divided into six successive steps which were provided with a carbon source and micro-
organisms. After completion of the first phase, the extraction wells of phase 1 were used as infiltration
wells for phase 2. This function shift is also applied in subsequent phases, for which a total of 55 wells
are placed. The wells are placed in so-called pulse borings. The filter section of the wells was
3
designed at 6-11 and 12-16 m -gl. On average, each filter with a capacity of 1.6 m /hr extracted and
infiltrated a total of 10 to 12 m/h.
The adjustments that have been made during the active phase are the reason for this;
the difference in the length of time between the various phases is the result of the differences in
distance between the extraction- and infiltration wells.
After the bacteria and nutrients were introduced into the soil, the conditions for biological remediation
improved, leading to the next results:
cell copies DHC in the infiltration water 5.1 * 10 / ml in a flow of 10 m / hour;
4 3
The contaminants are biologically degraded within the course of a few months to a few years at most.
After this time the risks will have been removed and we are left with clean groundwater. The location in
The Hague is situated in an urban area; even so, the entire remediation system could be placed
without problems. All required wells and pipes were placed underground. The remediation unit was on-
site for approximately sixteen months. This is very short when compared to more traditional
remediation techniques. This also meant that the disturbance to the community was kept to a
minimum. Furthermore, this technique is sustainable and environmentally friendly, as it requires very
little energy and only natural nutrients are used. In a period of sixteen months, the bacteria and
nutrients were introduced into the soil. A large portion of the contamination had already been entirely
degraded in that time. The remediation project was therefore a great success.
Conclusions
The TCE treatment results in a complete degradation to ethane/ethane.
Using the fully automated telemetric system, it is possible to continuously control the pH, redox and
oxygen in the groundwater.
Logging data indicated the working of the TCE unit with minimum human supervision and
maximum efficiency.
TCE is a sustainable solution for large plume remediations in comparison to conventional
techniques.
The project was completed in time, with no rebound observed and with no cost overruns.
Theme 1. Dealing with contamination of soil, groundwater and sediment
1c. remediation technologies and approaches
ABSTRACT:
Overview
A marshalling yard for trains has been contaminated as a result of leaks, spills and filling losses of
hydrocarbon fuels. An in situ remediation by Biosparging, Bioventing and Nutrient dosing has been
carried out in the period from 2004 to 2010. After remediation for an amount of half a million Euro, a
large residual TPH contamination remained in the soil.
Laboratory tests
By linking the results of an aliphatic aromatic TPH split group with substance group properties, a
mathematical insight is obtained into possible remediation techniques. It was found that 43% of the oil
is soluble in water and that more than 84% of the contamination is moderately aerobically biological or
chemically degradable.
A desk study indicated that chemical oxidation would give the best results with activated sodium
persulfate. On a laboratory scale, tests were carried out with alkaline activated persulfate and
hydrogen peroxide activated persulfate. The hydrogen peroxide activated persulfate showed a
decrease of 38-54% in the ground. However, the TPH was almost completely mobilized to the aqeous
phase. Treatment with alkaline activated persulfate showed a destruction of 49-54% of TPH in the
ground, with 10 times less mobilization of TPH to the water phase.
Full-Scale Application
The Full-scale remediation was carried out using a fully automated remotely controlled ISCO unit. The
unit ensures complete telemetric monitoring, operating 24 / 7. Essential parameters, such as the injection
pressure, the amount and flow rate of injection as well as the soil temperature were monitored
continuously.
During the first injection period, an amount of 6.260 kg alkaline activated Klozur persulfate, was
injected over 15 ISCO injection wells in the most polluted areas.
Application of chemical oxidation with activated Klozur persulfate leads to an increase in the pH, the
redox potential and dissolved oxygen levels.
Comparison of the pollutant load after completion of the chemical oxidation with the load prior to the
chemical oxidation, is showing a decrease of 49% of TPH in the ground.
A strong mobilization of product into the aqueous phase was not observed.
A second activation of the remaining Klozur persulfate will be performed later this year, followed by a
phase of enhanced aerobic bioremediation.
Conclusions
The innovative tender procedure has led to a smart solution, according on a ISCO approach with
activated Klozur persulfate based on the results of a bench scale treatability study.
It has been shown that bench scale testing is an indispensable tool in designing ISCO projects.
The bench scale treatability test and full scale field results correspond to each other.
After the first injection period of ISCO treatment was applied the TPH-concentration levels were
successfully reduced with 49%.
Session 1C.22
Jan Slunsk
NANO IRON, s.r.o. Czech Republic, +420 608 366 310, slunsky@nanoiron.cz
Keywords: nZVI, Chlorinated solvents, Batch tests, full-scale remediation, sandy aquifers
Background
Chlorinated solvents has been widely used on electrical substations in Denmark as a degreasing
agent to clean the transformers. The use has in many cases led to soil and groundwater
contamination. For the last 10 years, Geo have been doing environmental investigations and
remediation on sub stations for DONG Energy A/S. The remediations was carried out as part of a
large-scale initiative by DONG Energy A/S, regarding environmental investigations and remediation of
contamination on electrical substations. In 2012 and 2013 full scale remediation with stimulated
reductive dechlorination (SRD) have been carried out on 3 electrical substations contaminated with
chlorinated solvents PCE and TCE /1, 2, 3/. During the SRD molasses was used as substrate, and the
bacteria culture KB1 was added in order to increase the number of specific degraders on the sites.
Degradation of the mother compounds PCE and TCE was observed but also accumulation of
degradation products cis-DCE and VC. Because of the accumulation of degradation products it was
decided to investigate the possibility of using nanoscale zero valent iron (type NANOFER 25S) to
achieve full degradation of the chlorinated solvents and their degradation products. The purpose of
this paper is to present the results from the treatability study and from the following full scale
application of nZVI to the contaminated groundwater on the three sites.
The contaminated sites (site 1, Farum, site 2, Helsingr and site 3, Borup) are located in the northern
part of Zealand. All of the sites are electrical substations contaminated with Perchloroethylene (PCE)
and trichloroethylene (TCE) and their degradation products sum of dichlorethylene (DCE) and vinyl
chloride (VC). The overall geology for the three sites, is claytill with various contents of sand lenses
underlain by a sandy secondary aquifer. The pollutants have spread through the claytill down to the
secondary aquifer and are considered to impose a risk to the primary aquifer from where ground water
are extracted for drinking water purpose. On Figure 1 the extension of the ground water contamination
on the three sites are presented.
Figure 1: The extend of the groundwater contamination on the three sites site 1, Farum, site 2, Helsingr and site
3, Borup.
nZVI
The injected product used in the three cases described in this paper is nZVI NANOFER 25S from
NANO IRON, s.r.o. The product is an aqueous dispersion of Fe(0) nanoparticles with a surface
modification which is based on a combination of biodegradable organic and inorganic stabilizer. The
nano particles exhibit an average particles size of 50 nm, an average surface area of 20-25 m2/g, a
narrow particle size distribution of 20-100 nm and an iron content in the range of 80-90 wt. %. The
product is designed for the application in the reduction technology of groundwater remediation. /6/
The nZVI is a strong reducing agent and react with contaminants like PCE, TCE, DCE and VC as well
as natural occurring oxidized compounds. The redox conditions in the area to be treated is therefore of
great importance in order to avoid the nZVI reducing natural occurring oxidized compounds instead of
the contaminants.
Treatability study
In order to investigate if nZVI could enhance the degradation of contaminants on field scale level, a
treatability study was conducted in the laboratory /2, 3, 4/. The tests were carried out with groundwater
extracted from the contaminated sites. On site 1 and 2 groundwater samples was taken from the area
treated with SRD. On site 3 groundwater samples was taken both from an area treated with SRD and
from an area which had not been treated with SRD. In each batch a test series with a blind test, a test
with 1 g/l added nZVI and a test with 5 g/l added nZVI was carried out. An overview of tests and days
of sampling are presented in Table 1.
Table 1: Overview of tests and days of sampling.
The samples were analysed for chlorinated solvents, degradation products, ethane and ethylene. On
Figure 2 results from the tests are presented. In all three tests the concentration of DCE decreased
more when adding nZVI to the groundwater than without added nZVI. In the samples from site 1,
Farum, a significant increase in the concentration of ethylene was observed when the concentration of
nZVI was increased to 5 g/l.
Site 1 (Farum)
Site 2 (Helsingr)
Site 3 (Borup)
The degree of dechlorination (DOD%) which are a parameter that cab describe the development of
the reductive dechlorination process is calculated for each batch and presented in Figure 3. The
groundwater samples from all three sites show that the batch where no nZVI has been added shows
no increase in DOD% whereas the batches with added nZVI shows an increase in DOD% over time.
Figure 3: DOD calculated for the batch tests from the three sites, Farum, Helsingr and Borup.
The results from the treatability study conducted in the laboratory showed in general increased DOD%
when adding nZVI to the ground water. The tests showed positive results, and based on the results, It
was decided to do field scale remediation with direct injection of nZVI slurry into the ground water on
the three locations.
Arkil A/S have carried out the injections as direct injections under supervision of Geo and NANO
IRON, s.r.o,. The nZVI was the type NANOFER 25S and was delivered from NANO IRON, s.r.o.
On site 1, Farum, nZVI was applied in areas formerly treated with SRD. The treatment zone (TZ) had
approximately a volume of 960 m3. nZVI was injected in 15 points with maximum 5 m between each
point. In each point nZVI slurry was injected in 4 to 7 depths from 14 to 16 m bgs. There was injected
around 13 kg nZVI mixed in 1.3 m3 of water per point. In total the amount of nZVI injected was around
195 kg. The placement of injection points are presented on Figure 4.
At site 3, Borup, the injections were also carried out in three treatment zones as presented on Figure
6. nZVI was applied in the source area treated with SRD (treatment zone 1 and 2), and in the plume
area not treated with SRD (treatment zone 3). In treatment zone 1 the injections were carried out in 9
incline boreholes drilled under the transformer in order to prevent contamination left under the
transformer to leak to the underlaying secondary aquifer. In each boring 500 l of nZVI slurry was
injected, corresponding to 60 kg of nZVI in total. In treatment zone 2 and 3 the injections were carried
out as direct injection in 24 points. The volume of treatment zone 2 and 3 was approximately 2,650 m3.
In each point 2,500 l slurry was injected in 5-6 depths. In total 65,000 l nZVI slurry was injected at site
3 during the first injection round. Approximately 5 months after the first injection round, nZVI was
reinjected in the plume area. There were injected app. 1,950 l nZVI slurry in 13 points. In each point
there were injected 300-350 l of nZVI slurry in 6 depths.
Figure 6: Injection points and treatment zones at site 3, Borup.
Results of full scale remediation at two sites with prior treatment with SRD
Results from selected monitoring wells at site 1, Farum and site 2, Helsingr are presented in Figure
7. These sites were treated with SRD prior to the injections with nZVI. The red dottet line show the
stop criteria for the remediation and the black line show the injection time for nZVI.
Figure 7: Results from monitoring wells on site 1, farum (to the left) and site 2, Helsingr (to the right) presented
as time series. On these sites SRD were carried out in the treatment zones prior to injection with nZVI. The black
lines show the time of injections with nZVI.
Both Site 1, Farum and site 2, Helsingr show a significant decrease in the concentrations of PCE and
TCE as well as the degradation products DCE and VC after injection of nZVI. Together with a
significant increase in ethane and ethylene, this indicates that full degradation of chlorinated solvents
are taking place at the sites. The molar fractions for all monitoring wells in the treatment zones at site
1, Farum and site 2, Helsingr, are presented in Figure 8. The overall picture show an increase of the
fraction of ethane and ethylene at both remediation sites corresponding to the results shown in Figure
7.
Figure 8: Molar fractions for all monitoring wells in the treatment zones at site 1, Farum (left) and site 2, Helsingr
(right) before injection with nZVI and after injection with nZVI.
The DOD% is plotted in Figure 9 for the two sites. The DOD% is increasing significantly at site 1,
Farum after the injection of nZVI. Before injections with nZVI the DOD% was in the interval between
app. 20% and 70% showing that the PCE and TCE were degrading to DCE and VC but not fully
degraded to ethane and ethylene. After injection of nZVI at the site, the DOD% has increased to
between 40% and 90% indicating that the chlorinated solvents are fully degraded to ethane and
ethylene. At site 2, Helsingr there is also a significant increase in DOD% after injection of nZVI. 9
months after injection of nZVI the DOD% is between 60% and 100% where prior to the injection with
nZVI the DOD was between 20% and 70% indicating that the injection with nZVI has increased the
degradation potential and full degradation of chlorinated solvents and their degradation products are
occurring.
Figure 9: Degree of dechlorination (DOD%) for all monitoring wells in the treatment zones at site 1, Farum (left)
and site 2, Helsingr (right). The black lines show the time of injections with nZVI.
Comparison of results of nZVI done in an area without prior treatment with SRD and nZVI done
in an area with prior treatment with SRD.
At site 3, Borup, nZVI was injected in 3 treatment zones. Treatment zone 1 and 2 have been treated
with SRD by injection of molasses and bacteria, but in treatment zone 3 nZVI was injected directly to
the groundwater without prior treatment with SRD, since it was evaluated that the redox conditions of
the groundwater was suitable for dechlorination by nZVI. The measured concentrations in the
groundwater from selected monitoring wells are presented in Figure 10.
Figure 10: Results from monitoring wells on site 3, Borup. The results to the left is from the area pre-treated with
SRD (TZ1 and TZ2) and the results to the right is from the plume (TZ3). The black lines show the time of
injections with nZVI.
In treatment zone 1 and 2 where SRD have been carried out, a significant decrease in the
concentrations of sum DCE and VC is observed after injection of nZVI to the groundwater. Together
with an observed increase in the concentrations of ethane and ethylene, this indicates that full
degradation of the chlorinated solvents are taking place after the injection of nZVI in treatment zones 1
and 2. In treatment zone 3 (the plume) a decrease is observed in the concentration of sum DCE, but
the concentration of VC is constant. The molar fractions for all monitoring wells in the treatment zones
are presented in Figure 11.
Figure 11: Molar fractions for all monitoring wells in the treatment zones at site 3, Borup. The results to the left is
from the area pre-treated with SRD (TZ1 and TZ2) and the results to the right is from the plume (TZ3).
The plot of molar fractions show that the dominating fraction in treatment zone 1 and 2 after injection
of nZVI is ethane and ethylene whereas the dominating fraction in treatment zone 3 is VC and DCE.
That corresponds to the results shown in Figure 10.
The DOD for monitoring wells in the treatment zones at site 3, Borup is presented in Figure 12.
Figure 12: Degree of dechlorination (DOD%) for all monitoring in the treatment zones at site 3, Borup. The results
to the left is from the area pretreated with SRD (TZ1 and TZ2) and the results to the right is from the plume (TZ3).
The black lines show the time of injections with nZVI.
The DOD% increases significant in treatment zone 1 and 2 after injection with nZVI. Before injection
with nZVI the DOD% is between 20% and 60% indicating that the dechlorination process stops at
DCE and VC. After injection with nZVI the DOD% increases to between 60% and 100% showing that
full degradation of the chlorinated solvents are occurring. In treatment zone 3 there is not observed
any change in the DOD% before and after injection with nZVI.
The results from site 1, Farum and site 2, Helsingr, where SRD have been conducted before
injection with nZVI, show increasing concentrations of VC, ethylene and ethane and increasing
DOD%, after injection of nZVI at the locations, indicating that the chlorinated solvents are fully
degrading and the degradation products DCE and VC is no longer accumulating. The concentrations
of chlorinated solvents and degradation products in the groundwater is close to the stop criteria for the
remediation and it is evaluated, that the contamination can be removed to below the stop criteria with
a supplementing injection of nZVI at the two sites.
At site 3, Borup the injections with nZVI were carried out both in the source area where SRD have
been conducted and in the plume area where SRD was not implemented. The results from the source
area show that the concentration of ethane and ethylene is increasing, and the DOD% calculated for
the monitoring wells are also increasing significant after injection of nZVI. In the plume area there is
not observed a decrease in the concentrations of chlorinated solvents or degradation products and the
DOD% is constant before and after injection of nZVI.
The results of the three remediations with injection of nZVI strongly indicates that pre-treatment with
SRD using molasses as donor will increase the degradation of chlorinated solvents and their
degradation products at contaminated sites, and that the problems that were observed with
accumulation of degradation products DCE and VC was solved using nZVI in a treatment train with
SRD.
References
/1/ Hillerd. Borupvej 49. Borup Transformerstation T12. Reinjektion med substrat og bakterier
2013. GEO projekt nr. 33089/27333/35970. Rapport 30, 2013-10-07
/2/ Helsingr. Esrumvej 77-79. Helsingr Transformerstation. Statusrapport for fuld skala
oprensning med SRD med anbefalinger om restoprensning. GEO projekt nr. 30389. Rapport 16,
2014-02-10
/3/ Farum, Rugmarken 42. Farum Gydegrd Transformerstation. Status for SRD oprensningen.
Moniteringsrunde 4-7 (monitering 1-4 efter reinjektion med melasse og bakterier) samt
beskrivelse af batchforsg med nZVI. GEO projekt nr. 31832. Rapport 9, 2014-09-03
/6/ http://www.nanoiron.cz/en
SILPHES INVESTIGATION OF CHEMICAL TREATMENTS FOR THE REMEDIATION OF
RECALCITRANT CHLORINATED SOLVENTS
1 1 1 1
Romain RODRIGUES , Stphanie BETELU , Frdric GARNIER , Stfan COLOMBANO , Antoine
2 3 4 5 1
JOUBERT , David CAZAUX , Guillaume MASSELOT , Thodore TZEDAKIS , Ioannis IGNATIADIS
1
BRGM (French Geological Survey), 3, avenue Claude Guillemin, BP 36009, 45060 Orlans Cedex 2,
France.
2
SERPOL, 2, chemin du gnie, BP 80, 69633 Vnissieux Cedex, France.
3
SOLVAY, 1, avenue de la Rpublique, 39501 Tavaux Cedex, France.
4
ADEME (French Environment and Energy Management Agency), 20, avenue du Grsill, BP 90406,
49004 Angers Cedex 1, France.
5
Laboratoire de Gnie Chimique, Universit Paul Sabatier, 118, route de Narbonne, 31062 Toulouse
Cedex 9, France.
Contact:
Since the 1990s, in situ chemical remediation is frequently considered because of good treatment
efficiency without carrying out an excavation or without additional processing step. Among the in situ
remediation technologies, those for the treatment of chlorinated solvents are highly widespread,
essentially for the treatment of PCE and TCE by using strong oxidizing or reducing agents. This study
is accomplished within the framework of SILPHES financed by ADEME, the French Environment and
Energy Management Agency (AMI 2013 program). SILPHES is a "technology demonstrator" project
which aims at developing innovative solutions for in situ remediation of a mixture of recalcitrant
chlorinated solvent, mainly composed of hexachlorobutadiene, hexachloroethane, PCE, TCE and
hexachlorobenzne, directly extracted in the groundwater of a polluted site in Tavaux, France.
SILPHES is organized around two fundamental and complementary tasks:
- The remediation of chlorinated solvents point source pollution. This part is devoted to the
optimization of the treatment of the point source, which includes physical, chemical and
thermal treatments;
- The remediation of chlorinated solvents plume. This part is devoted to the improvement of
environmental diagnosis and the design and monitoring of natural attenuation and
bioremediation.
Here, only the remediation of the residual phase of chlorinated solvents, remaining after a pump and
treat technology, with chemical treatments has been studied. Remediation of chlorinated solvents can
be performed with two chemical processes: oxidation and reduction. Oxidative processes were firstly
used to in situ remediation of chlorinated compounds, with the used of Fentons reagent (Leung et al.,
1992) and permanganate (Schnarr et al., 1998). Reductive processes have been then developed to
find a solution to the main disadvantage to in situ oxidation, i.e. the degradation of subsurface organic
matter (Siegrist et al., 2006), and to take advantage of the reducing conditions generally prevailing in
the lower part of the soil (Vizier, 1971).
In this study, aliphatic and aromatic chlorinated compounds can be encountered. Molecules have
different chemical affinities with reactants: chlorinated compounds with double bonds are generally
oxidizable, and chlorinated compounds with multiple chlorines are generally reducible. Both oxidation
and reduction can be considered and have been studied in laboratory.
2) Methods
a) Chemicals
The mixture of chlorinated compounds is extracted from groundwater polluted site located at Tavaux
(France). It is composed of hexachlorobutadiene (52 58 % w/w), hexachloroethane (12 16 % w/w),
perchloroethylene (7 9 % w/w), pentachlorobenzene (4 % w/w), carbon tetrachloride (3 5 % w/w),
trichloroethylene (2 % w/w) and hexachlorobenzene (1 % w/w) (see Fig. 1). The remaining percentage
includes tetra- and trichlorobenzenes, trichloropropane, tetrachloroethanes, pentachlorobutane and
pentachloroethane.
1% 5 - 19%
2%
Hexachlorobutadiene
3 - 5%
Hexachloroethane
4%
Perchloroethylene
7 - 9%
Pentachlorobenzene
52 - 58% Carbon tetrachloride
12 -16% Trichloroethylene
Hexachlorobenzene
Other chlorinated compounds
The chemical degradation of the mixture was investigated with three oxidizing agents potassium
permanganate KMnO4, Fenton's reagent H2O2 and sodium persulfate Na2S2O8 (ESCTP, 1999; Brown,
0
2003; Osgerby, 2006) and three reducing agents zero-valent iron Fe , with or without surfactant,
and sodium dithionite Na2S2O4. The theoretical stoichiometry for each reactant has been established
by considering the complete dechlorination of the mixture, and different concentrations of reactants
have been studied.
Tab. 1. Mass theoretical stoichiometry for all reactants used against the mixture
of chlorinated compounds.
Hydrogen peroxide has been tested in laboratory and for field application since the beginning of the
1990s (Leung et al., 1992; Ravikmur and Gurol, 1994; Gates and Siegtist, 1995). He is generally used
with iron salt to form Fentons reagent which is more reactive: many active oxygen species like
hydroxyl radical, superoxide, hydroperoxide anion or singlet oxygen are produced. Hydrogen peroxide
was purchased from VWR International (30 % solution) and ferrous iron from Acros Organics (99 %
purity).
Permanganate is a strong oxidant used since the late 1990s for in situ degradation of chlorinated
compounds (Schnarr et al., 1998; Siegrist et al., 2001). Permanganate can cleave the double bonds of
other compounds (Yan and Schwartz, 2000; Ogino et al., 1990), so it is effective for the degradation of
chlorinated organic compounds such as chloroethenes. The reactivity of permanganate is a function of
the pH; the most common reaction in groundwater environment involves three electrons.
Permanganate potassium was purchased from VWR International.
The used of persulfate for ISCO applications is more recent (Pugh, 1999), and a few fields
applications have been developed. Persulfate can be activated to form more active species (Hoag et
al., 2000, Block et al., 2004). Like hydrogen peroxide, reactivity of persulfate results from the formation
of sulfate radical, which has similar redox potential than hydroxyl radical, with the advantage of being
more stable in the subsurface. Persulfate sodium was purchased from Prolabo (98 % purity).
Zero-valent iron was first use for in situ remediation at the beginning of the 1990s in permeable
reactive barriers (Gillham, 1993), and has been extensively developed for source and plumes
treatment. ZVI particles react directly on the pollutant, but the reactivity is more important as ferrous
ions and hydrogen produced can also react (Matheson and Tratnyek, 1994; Johnson et al., 1996).
However, at long term, an oxide/hydroxide passivation layer of ferric ions is formed on the iron surface
(Farrell et al., 2000), reducing significantly the remediation efficiency (Szecsody et al., 2000).
Nanoscale zero-valent iron (ZVI) was purchased from Nanoiron, Czech Republic (NF25, stabilized
water dispersion of nanoscale zero-valent iron (14-18 %), and NF25S, stabilized water dispersion of
nanoscale zero-valent iron (14-18 %) with 3% of surfactant). Iron micro-particles enhanced by noble
metal hot spots (palladium, nickel and/or silver) were purchased from Biorem, Belgium (Biocat, slurry
of micro-scale iron in water soluble or water insoluble matrixes). Three different formulations have
been tested, with different ratio of each metallic spot.
Dithionite is known to be a strong reducing agent and it is efficient for in situ remediation for
chlorinated ethenes (Lee and Batchelor, 2004). It can also be used with zero-valent iron to dissolve
and reduce the iron oxide (III) to regenerate the iron (II) species at the iron surface (Xie and Cwiertny,
2010), and improve the lifetime of iron in subsurface. Sodium dithionite was purchased from Silox,
Belgium.
b) Experimental method
Experiments were performed in 100 mL penicillin vials filled with 50 mL of deionized water and a
volume 1 or 0.1 mL of the mixture of the chlorinated compounds mainly composed of
hexachlorobutadiene and hexachloroethane. Two or three different concentrations of reactants
calculated from theoretical stoichiometry were added to the batch system after sparging with nitrogen
to remove dissolved oxygen. Replicates of each experimental condition were performed to monitor
degradation with time. Vials were sealed with PTFE septa, and were stirred at 100 rpm in the dark in a
thermostatic chamber maintained at 12C, the average groundwater temperature.
A gas chromatograph was used to analyze all of the chlorinated compounds in samples, in aqueous
phase and headspace. Chloride was analyzed using an ion chromatograph (Dionex ICS-3000).
First, the most widespread oxidative and reducing agents have been tested to compare the relative
efficiency of two opposite processes. 1 mL of the mixture (1.73 g) was introduced in the vials. The
three oxidants were used at theoretical stoichiometric ratio, at 2 times the stoichiometric ratio and at 3
times the stoichiometric ratio. Suspensions of nanoscale iron were used at 0.038 times the
stoichiometric ratio, 0.192 the stoichiometric ratio and 0.384 the stoichiometric ratio. Sodium dithionite
was used at 0.357 times the stoichiometric ratio, 0.714 times the stoichiometric ratio and 1.07 times
the stoichiometric ratio.
Chloride ions are a good indicator for the quantification of the remediation of chlorinated solvents, as it
is one of total degradation products in both oxidation and reduction. Degradation percentages based
on chloride ions concentrations obtained with the best concentrations of reactants studied are shown
in Fig. 2.
14% 14%
After 15 days After 23 days
12% 12%
After 29 days After 80 days
Yield of degradation (%)
10% 10%
After 49 days
8% 8%
6% 6%
4% 4%
2% 2%
0% 0%
Na2S2O8 Fenton's reagent KMnO4 Na2S2O4 NF25 NF25S
(stoichiometric ratio) (stoichiometric ratio) (2 (1,07 (0,37 (0,37
stoichiometric ratio) stoichiometric ratio) stoichiometric ratio) stoichiometric ratio)
The low values can be explained by worse operating conditions. Amount of chlorinated solvents
introduced represents a concentration beyond what it is expected to found for the in situ residual
remediation, but it permits to test the limits of the reactivity of oxidizing and reducing agents.
After 49 days of incubation with oxidant, Fentons reagent and sodium persulfate has shown very low
chloride concentrations in all case. Potassium permanganate has shown the best results for oxidative
process, with a yield of degradation at 4 % at 2 times the stoichiometric ratio. After 80 days of
incubation with reductant, the yield of degradation has reached 2.3 % with sodium dithionite, 7.00 %
with NF25 and 9.3 % with NF25S. Reduction is more efficient than oxidation for the remediation of this
particular mixture of chlorinated compounds. NF25 and NF25S have shown the best efficiency and
were used as the limiting reactant. Increasing the amount of iron should enhance the degradation of
chlorinated solvents.
Results obtained with gas chromatography have shown that more than thirty different compounds are
quantified. Some of them are preferentially dechlorinated. With nZVI, the evolution of concentrations
and relative accumulation of chloroethanes has allowed to established three dechlorination pathways,
as proposed by Song and Carraway (2005): hydrogenolysis the break of a C-Cl bond to form a C-H
bond , -elimination the removal of two chlorines and the formation of a C=C double bond instead
of a C-C bond and dehydrohalogenation the elimination of HCl and the formation of a C=C double
bond (see Fig. 3). A complete dechlorination is observed as there is production of ethane and
ethylene.
Fig. 3. Degradation pathways for hexachloroethane with nZVI.
With dithionite, few variations in concentrations were observed over the time of incubation, and no
total dechlorination products have been observed. These results are in agreement with ion
chromatography results, and confirm that dithionite is not very efficient. The use of potassium
permanganate has shown a better efficiency against molecules with double bond, such as
chloroethenes and, to a lesser extent, hexachlorobutadiene. Finally, Fentons reagent and sodium
persulfate have not shown significant change over the time.
This experiment investigated the degradation of the mixture with amounts of reagents widely in excess
to define the limits of treatment efficiency, so the limiting reactant was the mixture of chlorinated
compounds. Only 0.1 mL of the mixture (0.173 g) was introduced in the vials, to be assimilated to an in
situ residual concentration.
Suspensions of nanoscale iron were used at 3.85 times the stoichiometric ratio (0.96 g). All Biocat
slurries were used at 6 times the stoichiometric ratio (1.5 g of bimetallic iron particles in 8.5 g of
matrix).
100%
90%
Yield of degradation (%)
80%
70%
60%
50%
40%
30%
20%
10%
0%
Control NF25 NF25S Biocat A Biocat A Biocat B Biocat B Biocat C Biocat C
sample soluble insoluble soluble insoluble soluble insoluble
The percentage of remediation of the mixture with all Biocat slurries ranges from 48 % with Biocat B
water soluble slurry to 85 % with Biocat B water insoluble slurry. The two types of matrixes used with
Biocat have shown different reactivity. With the water soluble matrix, the contact surface with iron is
more important for the more soluble compounds of the mixture such as chloroethanes and
chloroethenes, as iron particles can be encountered in water-soluble phase. Thus, in relative
abundance, ethane and ethylene are the two most compounds in presence in headspace gas
chromatography. The opposite has been observed with the water insoluble matrix: the less soluble
compounds such as hexachlorobenzene and hexachlorobutadiene are the most dechlorinated, as iron
particles can be encountered in a water-insoluble phase. Thus, in relative abundance, butane is the
most compound in presence in headspace gas chromatography (see Fig. 5). Benzene represents only
a few percentage as hexachlorobenzene represents only 1 w/w % of the mixture.
4) Conclusion
This study is a first approach to find the best reactant to the remediation of a mixture of recalcitrant
chlorinated compounds mainly composed of hexachlorobutadiene, hexachloroethane and PCE. Two
chemical processes oxidation and reduction were investigated. Experimental results have
demonstrated that the reductive process is more efficient than the oxidative process to this particular
mixture of chlorinated solvents. However, the most widespread reactants have shown limited reactivity
on this mixture, with low yield of dechlorination.
Innovative reactants, iron micro-particles enhanced by noble metal hot spots, have demonstrated
broad reactivity on chloroethenes, chloroethanes, and more recalcitrant compounds such as
hexachlorobenzene and hexachlorobutadiene. The production of butane, propane, ethane, methane,
and benzene with iron micro-particles enhanced by noble metal hot spots have shown that a complete
dechlorination of a wide range of compounds in the mixture has been reached.
Further studies on chlorinated solvents taken individually have to be done to clearly identify the
chemical kinetics and the degradation pathways of each molecule, to quantify precisely intermediate
products of each initial chlorinated solvent.
References
Block P.A., Brown R.A., Robinson D. (2004). Novel activation technologies for sodium persulfate In
Situ Chemical Oxidation. Proceedings, Fourth International Conference on the Remediation of
Chlorinated and Recalcitrant Compounds, Monterey, CA, USA.
Brown R.A. (2003). In Situ Chemical Oxidation: performance, practice, and pitfalls. AFCEE
Technology Transfer Workshop, February 24-27, 2003, San Antonio, Texas.
ESCTP (1999). Technology status review: in situ oxidation. ThermoRetec Consulting Corporation,
HydroGeoLogic, Inc. and Coleman Research Corporation Energy & Environmental Group.
Farrell J., Kason M., Melitas N., Li T. (2000). Investigation of the Long-Term Performance of Zero-
Valent Iron for Reductive Dechlorination of Trichloroethylene. Environmental Science &
Technology 34:514-521.
Gates D.D., Siegrist R.L. (1995). In Situ Chemical Oxidation of trichloroethylene using hydrogen
peroxide. Journal of Environmental Engineering 121:639-644.
Gillham R.W. (1993). Cleaning Halogenated Contaminants from Groundwater. U.S Patent n
5,266,213.
Hoag G.E., Chhedda P.V., Woody B.A., Dobbs G.M. (2000). Chemical oxidation of volatile organic
compounds. U.S. Patent n 6,019,548.
Johnson T.L., Scherer M.M., Tratnyek P.G. (1996). Kinetics of halogenated organic compound
degradation by iron metal. Environmental Science & Technology 30:2634-2640.
Lee W., Batchelor B. (2004). Abiotic reductive dechlorination of chlorinated ethylenes by soil.
Chemosphere 55:705-713.
Leung S.W., Watts R.J., Miller G.C. (1992). Degradation of perchloroethylene by Fentons reagent:
Speciation and pathway. Journal of Environmental Quality 21:377-381.
Matheson L.J., Tratnyek R.G. (1994). Reductive dehalogenation of chlorinated methanes by iron
metal. Environmental Science & Technology 28:2045-2053.
Ogino T., Hasegawa K., Hoshino E. (1990). Kinetics of the permanganate oxidation of endo-
dicyclopentadiene in non-aqueous organic solution: Formation and detection of the manganate (V)
ester intermediate. Journal of Organic Chemistry 55:2653-2656.
Osgerby I.T. (2006). ISCO Technology overview: do you really understand the chemistry?
Contaminated Soils, Sediments and Water. In: Calabrese E.J., Kostecki P.T., Dragun J. (Eds.).
Springer US, 287-308.
Pugh J.R. (1999). In Situ Remediation of soils containing organic contaminants using the
electromigration of peroxysulfate ions. U.S. Patent n 5,976,348.
Ravikmur J.X., Gurol M. (1994). Chemical oxidation of chlorinated organics by hydrogen peroxide in
the presence of sand. Environmental Science and Technology 28:394-400.
Schnarr M., Truax C., Farquhar G., Hood E., Gonullu T., Stickney B. (1998). Laboratory and controlled
field experiments using potassium permanganate to remediate trichloroethylene and
perchlorothylne DNAPLs in porous media. Journal of Contaminant Hydrology 29:205-224.
Siegrist R.L., Urynowicz M.A, West O.R., Crimi M.L., Lowe K. (2001). Principles and Practices of In
Situ Chemical Oxidation Using Permanganate. Battelle Press, Columbus, OH, USA, 336p.
Siegrist R.L., Crimi M.L., Munakata-Marr J., Illangasekare T., Dugan P., Heidersheidt J., Jackson S.,
Petri B., Sahl J., Seitz S., Lowe K., Van Cuyk S. (2006). Reaction and transport processes
controlling in situ chemical oxidation of DNAPLs. Strategic Environmental Research and
Development Program (SERDP), Project CU-1290, 235.
Song H., Carraway E.R. (2005). Reduction of Chlorinated Ethanes by Nanosized Zero-Valent Iron:
Kinetics, Pathways, and Effects of Reaction Conditions, Environmental Science & Technology
39:6237-6245.
Szecsody J., Williams M., Fruchter J., Vermeul V., Evans J. (2000). Influence of sediment reduction on
TCE degradation, remediation of chlorinated and recalcitrant compounds. In: Chemical Oxidation
and Reactive Barriers, Wickramanayake G. (Ed). Battelle Press: Columbus, OH, Vol. C2-6, 369-
376.
Vizier J.F. (1971). Etude de ltat doxydo-rduction du sol et de ses consquences sur la dynamique
du fer dans les sols hydromorphe. Cah. O.R.S.T.O.M., srie Pdologique, vol. IX, n4.
Xie Y., Cwiertny D.M. (2010). Use of dithionite to extend the reactive lifetime of nanoscale zero-valent
iron treatment systems. Environmental Science & Technology 44:8649-8655.
Yan Y.E., Schwartz F.W. (2000). Simulating the in situ oxidative treatment of chlorinated ethylenes by
potassium permanganate. Water Resources Research 36:3031-3042.
ThS 1C.16 In Situ Chemical Oxidation (ISCO)
Abstract
At the VOPAK Terminal ACS site in Antwerp an in-situ chemical oxidation (ISCO) remediation project
is being executed to restore groundwater quality. The site is locally polluted with organic solvents
(chlorinated aliphatic hydrocarbons, BTEX and volatile petroleum hydrocarbons).
Part of the accessible heavily polluted soil was removed by excavation. The remaining pollution mass
must be treated in situ, due to infrastructure present preventing further excavation. ISCO is being
applied, injecting both air/ozone (gas) and hydrogen peroxide (liquid) into the soil. The hazardous
properties of these oxidants and the EX-rating of the site imply that strict health and safety measures
are necessary, the chemical site remaining fully operational during the remediation. To limit
uncontrolled migration of contaminated vapours and ozone, the ISCO system is combined with soil
vapour extraction (SVE). We demonstrate that this combined remedial approach is applicable to
remediate complex soil and groundwater pollution at an EX-rated industrial site. The project is co-
funded by the EU Life+ project VOPAK-EXPER03.
Introduction
In Situ Chemical Oxidation (ISCO) is a well-tried approach to remediate aquifers polluted with
(oxidizable) organic pollutants such as chlorinated ethenes, gasoline, etc. It can be used to treat both
source and plume zones. In most cases liquid oxidants are used, such as permanganate, persulfate or
Fentons reagent [1]. One of the issues in using ISCO in heterogeneous soils, is the difficulty to
properly bring oxidant and contaminants into close contact. ISCO using a combination of ozone/air
injection and liquid hydrogen peroxide offers possibilities in this respect [2]. This technology is being
applied at full scale to remediate the VOPAK ACS site in Antwerp, Belgium (figure 1). The
groundwater level at the site normally is around 1 m below ground level (bgl). The soil is sandy to a
depth of approximately 3 m bgl, then becoming finer, with locally peat pieces being present to a depth
of 5 m bgl. Deeper, poorly permeable layers (clay) are present. The soil and groundwater are locally
polluted with organic solvents. Groundwater concentrations of chlorinated ethenes (PCE and partial
dechlorination products TCE, DCE, VC) were measured in the range 1 - 80 mg/L. Chlorinated ethanes
1,1,1-TCA and 1,1-DCA are present in concentrations up to 342 and 143 mg/L respectively. Non-
chlorinated solvents are also present at high concentration levels in the groundwater (BTEX and
volatile petroleum hydrocarbons >100 mg/L). The soil remediation project approved by the Flemish
Environmental Agency (OVAM) defined the following target values to be reached for the groundwater
(Table 1).
ThS 1C.16 In Situ Chemical Oxidation (ISCO)
Fig.1. Aerial view of the site. The treatment area is outlined in red.
To meet these objectives special attention was devoted to the design, automation & safety controls,
phased implementation (starting farthest away from the tank farm, and with the source zone
excavation), monitoring and field quality control of injection wells (pressure test, distribution of raising
gas bubbles) and the extraction system (leaks, water locks,).
was discontinuous, 10 min on and 50 min off. In total (all active injection points) 100-300 g of ozone
was injected per hour. Peroxide 35% was diluted to 7% with tap water and injected at a rate of 240
mL/min. To capture air/ozone surplus of the ISCO process, SVE was operational (via the horizontal
drains; 0.2-0.3 bar underpressure at a rate of appr. 150 Nm per drain). Ozone monitoring (sewer,
monitoring wells, SVE influent, soil gas measuring points, ambient air) was applied to verify that any
ozone excess was effectively captured by the venting system.
During the second ISCO application phase (started in November 2013), in and around the former
excavation area, ozone and VOC emissions were detected. This was solved by restoring liquid-tight
conditions of the concrete floor (finished in May 2014). Apart from that, some leaks in the piping and
connection between piping were repaired and three extra SVE drains 4A-C were installed (prior to
concrete floor repair) as indicated in figure 2. Finally the air/ozone injection regime (per zone) was
adjusted.
Fig.3. Concentration trends in example monitoring wells (positively influenced by the remediation)
ThS 1C.16 In Situ Chemical Oxidation (ISCO)
Fig.4. Concentration trends in monitoring well P451 (ISCO not yet operational in that zone) and P481
and P482 that are affected by the ISCO but still show high concentrations
Phase 3: tiered implementation in the tank farm and specific safety measures
The third operational phase of the remediation efforts was started in March 2015, additionally applying
ISCO/SVE in the tank farm (around P451, near tanks 68 and 70). The lessons learned during the first
2 remediation phases were taken into account to avoid ozone emissions that might lead to corrosion of
the tank bottoms. The remediation therefore was started with discontinuous air injection in three
injection filters only (2 Nm3/h/filter; 2 min on followed by 16 min off), and SVE running via vertical
extraction wells. Measurements (PID, LEL, O2 and ozone) in the SVE influent air and groundwater
monitoring wells P451, P471, P472 were performed during a period of several days after start-up. Also
the underpressure was recorded and PID measurements performed in measuring points installed at
the tank bases. In the field the absence of leaks in above-ground conduits/pipes was checked visually
and by using PID. SVE and air injection being operational without any problems, ozone and peroxide
were added to the process (2 Nm3/h/filter of air; 50 g/h of ozone and hydrogen peroxide 7% at a rate
of 120 ml/min). Previously mentioned monitoring measurements were continued. The SVE is
continuously operated and subsidence measurements of solvent tanks 68 and 70 are performed by
VOPAK. So far no subsidence has been observed.
After proven safe operation the ISCO will be further scaled up to the complete number of injection
points in the tank farm.
ThS 1C.16 In Situ Chemical Oxidation (ISCO)
From March 2013 to December 2014 a total of 6970 L (35%) H2O2 was consumed. The dosed
peroxide is approximately twice that of the injected ozone mass. Stoichiometrically, however, ozone
can oxidize more pollutant mass than peroxide. Since the main pollutant mass at the site is non-
chlorinated (> 90% of the total mass), most of the oxidant is being consumed to destroy the non-
chlorinated solvents, rather than the chlorinated solvents. Part of the total oxidant mass injected is also
lost to other reactions (e.g. oxidation of soil organic matter, reduced minerals, auto-oxidation of
peroxide: 2H2O2 O2 + 2 H2O and ozone decomposition: 2O3 3O2).
Roughly estimated, based on consumed amounts (see Fig. 5) so far, max. about 1000 kg of pollutant
mass was directly destroyed by the ISCO. About the same quantity was extracted by the MPE/SVE
and captured on the activated carbon, and about 5000 kg removed by the excavation. It is believed
that a substantial part of the pollutant mass is also biologically degraded in situ by aerobic bacteria,
owing to the high amounts of oxygen delivered to the subsoil.
ThS 1C.16 In Situ Chemical Oxidation (ISCO)
Conclusions
The results thus far available show that the applied ISCO is in combination with hot-spot excavation
and MPE/SVE an efficient treatment technology to treat the complex cocktail of pollution (both
chlorinated and non-chlorinated solvents) present in the low to mid concentration ranges. Based on
the results thus far available, it seems that ozone and peroxide injection can be safely implemented
even underneath solvent storage tanks, but requires careful scaling up, continuous SVE, a
discontinuous oxidant injection regime that needs to be adjusted for each (sub)location and meticulous
monitoring of process and safety parameters. At locations with high remaining contaminant load
(NAPL), ISCO may be in the long run also effective, but will require a very long remediation time
and large amounts of oxidants.
Acknowledgements
This project is partly financed by the EU Life+ project VOPAK-EXPER03 [3]. Patrick Versele and
Wendy Hendrickx are closely involved in the project for Vopak and Paul Baert (Badeco) as safety
expert.
References
[1] S.G. Huling, B.E. Pivetz, 2006. In-Situ Chemical Oxidation. U.S. EPA Engineering Issue, 60pp
[3] http://www.vopak-expero3.be/
ThS 1C.7 Strategies for remediation and brownfields regeneration
Michel Beaulieu
The town of Lac Mgantic, 6 000 inhabitants, is located in the Canadian Province of Quebec south east
corner, only 35 kilometers away from the American state of Maine. Build along the lake which gave its
name to the city, Lac Mgantic is a major northern transit route for trains circulating from the earth of the
continent to the Atlantic seaboard. Petroleum is one of the goods carried over that railway line,
th
increasingly so over the last years. Late on July 5 2013, a convoy consisting of 5 engines and 72 tankers
filled with crude petroleum extracted from the Bakken formation in North Dakota reached the town of
Nantes, 13 kilometres uphill from Lac Mgantic. On its way to the St-Johns Irving refinery in the
neighbouring eastern Canadian province of New Brunswick, it stopped for the night at the top of a slope.
Few minutes before 1 oclock at night, the breaks of the unattended train let go. The convoy started to roll
down the 1.2% slope. At 1:16 in the morning, entering Lac Mgantic downtown full speed, the train
derailed. Massive explosions and a huge fire followed, killing 47 people and destroying most of the town
centre buildings. Around 30 buildings, including the municipal library and the city archives, were burned
down. 37 others buildings, still standing, were more or less impacted by the fire and the pollution.
Involved almost immediately, the Quebec Sustainable Development, Environment and Climate Change
Ministry had to take charge of the following environmental interventions, after that the responsible
company failed to do so. This paper describes the works done by the ministry up to May 2015.
THE EMERGENCY INTERVENTION
Informed of the event by the firemen already fighting the blaze, the ministry emergency unit reached the
th
site at 5h30 in the morning of the 6 of July and was soon joined by the TAGA, a laboratory on wheels
able to sample and analyse in real time contaminants present in the air. Together with the other units
involved (firemen, public security department, etc) the urgency team identified four distinct problems that
had to be quickly managed:
The smoke
In the end, it will be established that the convoy carried 7 679 000 litres of light crude petroleum. From
this, 5 978 000 litres burned down or were spilled in the environment. But that information was not
th
available on the morning of July 6 . The exact nature of the petroleum transported was also unknown.
At 7.30, the TAGA moved next to the fire and took position at strategic locations, generating data on the
th
air quality. It kept doing so for days. The fire was finally put under control on July 7 . The next day, the
TAGA analyses showed a notable air contamination decrease. The data generated all along those critical
th
days contributed at first to establish the high risks area and, from the 10 of July, to allow the return of
more than 2 000 evacuated persons into their homes.
The petroleum accumulation in soils
The train derailed in a curve located on the former town rail yard, a flat area lying next to the city centre.
From there, a gentle slope leads to the shores of Lac Mgantic, less than 300 meters south-west. It was
given that part of the spilled hydrocarbons which had infiltrated the soils would migrate towards the lake.
1 th
As soon as the security perimeter was relaxed , on July 19 , 24 trenches were dug to stop this migration
and 32 pumping wells were put in place to recover the hydrocarbons.
Following the derailment, a flow of hydrocarbons had infiltrated the pluvial and sewer networks, following
this preferential path towards the lake. The pressure created by the petroleum wave flowing through the
pipes displaced all the manholes on its way. Confinement measures had to be set up (blockages,
retention pounds and stoppers) at the exit of the storm waterl drains for hindering the lake contamination.
At the same time, the sewer system was diverted and the polluted water brought toward an improvised
oily water recuperation pool. The town waste water treatment plant could then be cleanup up before being
put back in operation.
The navigation area next to the impacted zone was shut down, as well as the marina. The pumping and
treatment of the oily water recovered in the booms were undertaken. The lakes sediments and surface
water were sampled and analysed. A drinking and surface water follow up program was also established.
The results indicated that the contamination found in the lake was not a threat to the ecosystem. The rock
fill bordering the lake was cleaned up using high-pressure hoses.
The head of the Chaudire River lies next to where the hydrocarbons reached the lake. The river flows in
a northwest direction over a distance of 175 kilometres to meet the St-Lawrence River. On its way, it
crosses cities using its water as drinking water source. Despite all interventions, 100 000 litres of
petroleum reached the Chaudire River. Even if it did not have such accurate information at this time, the
ministry was aware of the risks incurred by the population deserved by the river. Aerial inspections
allowed the following of the hydrocarbons contamination front on the river. It led to the setting up of 14
booms and absorbing coils. At the same time, the flow of the Mgantic dam was lowered down to its
3 3
minimum, from 27 m /s to 3,5 m /s, a way of slowing down the spreading of the contamination. Three
municipalities lying downriver, St-George, Ste-Marie and Lvis, as well as two food processing plants,
were ordered to stop pumping water from the river. Alternative water sources were put in place and had to
be used.
1
A security perimeter has been established at first as there was still a risk of explosion. Later on, the remains of
the train were submitted to an extensive investigation to understand exactly what happened. Finally, human
remains had to be searched for.
THE CONTAMINATION ASSESSMENT
The impacted area was made off a fill more or less 2 meters thick, lying on a waterproof clay layer
more than 50 meters thick;
The natural imperviousness of the clay layer explained why no contamination of the groundwater had
been measured;
The spilled hydrocarbons had migrated from the impacted area towards the lake and the river by
infiltrating the underground sewer pipes or the granular fill which surrounded them;
Most of the contaminated soils were found close by the impacted zone;
Soils were contaminated by C10-C50, PAHs, OVCs, sulphur, and, more slightly, iron;
The heavy metals found in the soils and sediments were related to the former historical activities or
natural background of this area;
The volume of soils contaminated over the B criteria was estimated to be between 103 000 m3 and
126 000m3.
The 37 residential, commercial and institutional buildings which did not burned down and still stood in the
impacted area were assessed.
The banks and beds sediments appeared to be contaminated at different levels between the Mgantic
dam and the Sartigan dam in St-George town, a total length of about 85 kilometres. More than 150
workers cleaned manually the banks and, as much as possible, the bed of the river during summer and
fall 2013.
th
On September 18 , the analysis results showed that water and sediments quality near the drinking water
intakes were safe. The towns of St-George, Ste-Marie and Lvis were thus allowed to get again their
drinking water from the river. A water quality monitoring program was put in place and alternative sources
of drinking water were made ready to be used, in case hydrocarbons would later be detected.
The Chaudire River experts committee put in place by the ministry had the task to analyse all the data
gathered through the 2013 assessment campaign as well as additional water quality data generated
during the spring 2014. They also had to elaborate a management plan for the residual contamination.
th
The committee established in its May 12 2014 report that:
After the 2014 spring flood, the results of the survey program showed that there were no contaminants
in the brute water of the river;
At fall 2013, the extent and the level of the sediments contamination on the upper Chaudire was
important. The higher concentrations were found in the first 30 kilometres but, even there, were
variable;
Next to the St-George, Ste-Marie and Lvis drinking water intakes, sediments were only slightly or not
at all contaminated by petroleum hydrocarbons.
The perturbations related to the spill had major consequences on the upper river benthic organisms
communities,
Nothing showed that the spill had, in 2013, major consequences on the adult fish population.
Meanwhile, on the long term, the presence of petroleum at the bottom of the river could harm certain
species, for example by diminishing the food availability (benthic organisms) or by affecting the
reproductive system.
The report finally recommended updating the state of the sediments contamination, identifying the river
sections where the contamination level could still pose a risk for the ecosystem as well as the
environmental issues associated with the different methods foreseen for cleaning up the sediments.
In September 2013, 4 tenders were prepared to select the firms which could qualified to compete for
excavation, transport, confinement, demolition, environmental control and soil treatment contracts As for
the soil treatment, two major decisions had to be taken before letting the tender out, first, at which level
should the town polluted soils be cleanup up and, second, what should be done with the treated
excavated soils? The answers to those questions would have a direct impact on the quantity of soils to be
excavated and the choice of the cleanup technology, all technologies not allowing the same level of
cleanup.
The A, B and C generic criteria appearing in the Quebec Soil Protection and Contaminated Land
Rehabilitation Policy associates levels of contamination and possible land use. Roughly, to the A value, a
soil is not contaminated and would allow anything. Beyond B, it cannot be used for residential purposes.
Beyond C, all uses are restricted, including industrial and commercial. Since 1998, article 9 of the Quebec
Regulation respecting Hazardous Waste states that every person who accidentally releases a hazardous
material into the environment shall immediately recover the hazardous material and remove all
contaminated material that is not cleaned or treated on site, meaning bringing back the site to the quality it
had before the spill.
Consequently, the cleanup target for the Lac Mgantic residential and mixed zoning sectors had to be the
A criteria. As for the former rail yard, where some historic contamination pre-existed the spill, B would be
the target. For their part, the excavated soils should be treated in a way that they could be put to beneficial
use afterwards, ideally in Lac Mgantic itself. Once those objectives were selected, the ministry could
issue the tender aimed at selecting the soil treatment technologies. It was published on November 25
2013.
Lac Mgantic is located in the Appalachian mountain range. Snow may start falling mid-November and the
weather drop below zero by the beginning of December. All works would then stop till the thawing and the
drying of the soils, which may last till mid-may. Most of the works, especially the excavation works, must
consequently be performed between May and November.
The 2013 works
A lot had already been done at fall 2013. The aim was then to give back, as much and quickly as possible,
a normal life to the town and its citizens. To do so the following actions had to be taken :
Building a bridge over the Chaudire River to re-establish circulation between the two sides of the
town, the former bridge having been contaminated and damaged;
Repairing the railway and reopening it to traffic to allow for the movement of goods in, out and through
the town. To allow the reconstruction, contaminated soils found under the rails embankments had to
be excavated;
Cleaning part of the former unused rail yard to allow the construction on this free space of a new
commercial street and commercial buildings, where destroyed shops, including the food supermarket,
could be relocated;
Cleaning up the soils around the water and sewer pipes network before rebuilding them.
It was also feared that the spring thaw may contribute to the leaching of the soils, mobilising the
hydrocarbons that were still contained and polluting the Mgantic Lake and the Chaudire River. It was
thus important to increase the retention and treatment capacities for oily water. A new and larger
treatment centre was build.
3
Finally, three contaminated temporary storage platforms having a 75 000 m capacity allowing the
beginning of the cleanup work after the thaw were constructed. Those platforms were located in an
industrial sector three kilometres away from downtown.
The soils
At the beginning of June, once the soils had thawed and were dry enough to support the weight of the
machinery, the first phase of the contaminated soils excavation works and their replacement by clean soils
started. This work took place in the burned down area and next to the Agns bridge. It is also at this time
that the treatment of the excavated soils started. It was initially planned to use three technologies
(physico-chemical, biodegradation and thermal), the use of one or another depending of the level of
contamination and soil types met.
Ultimately, the physicochemical and thermal technologies were put aside for diverse reasons.
Contaminated soils were thus treated by biodegradation. To speed up the treatment tempo at first, some
soils were sent to two existing authorised biotreatment centres located outside of the region. This allowed
getting quickly the necessary space to store excavated soils and hinder potential delays in the work.
Next to 90% of the environmental work has by now been completed in downtown Mgantic. The only
contaminated soils still in place are the one located under the, by December 2014, still standing 37
buildings and infrastructures damaged by the fire as well as on a portion of the railway tracks next to the
city centre.
To this day, 240 000 m3 of soils have been excavated and carried to the treatment platforms or in
authorised treatment centres. It has been first envisioned to reuse those soils in Lac Mgantic, namely to
fill up the excavation. Taking into account the level of cleanup finally reached, the B value, the soils were
finally reused 116 kilometres westward of Lac Mgantic, in a town called Asbestos, on the site of a the
former Jeffrey asbestos mine. There, they are mixed with fertilizing solid waste and spread on the
asbestos tailings, concurring to their greening.
The pumping and the recovery of the contaminated water never ceased since the ministrys first
interventions. All together, more than 100 millions of oily water has been treated since July 2013. The
pumping went on the whole winter, namely under the buildings, to make sure that the containment
measures were in effect till the end of the cleanup work.
An intensive assessment study of the Lac Mgantic and Chaudire River sediments has been conducted
during the summer 2014 by the companies targeted by the orders and their final report is awaited.
Meanwhile, after the sediments assessment done by the ministry following the 2014 spring flood, the
committee concluded that it was necessary to act only at kilometres 4,5 where the concentrations were
still high, sometimes higher than in 2013. Work was done in this sector, by the same companies.
2
The work aimed at removing a first 15 centimetres thick layer of contaminated sediments on a 2 675 m
area, followed by the removal of a second 15 cm layer if the level of residual contamination still justified it.
The drying of the excavated sediments, the treatment of the effluent and the damaged banks stabilisation
were also done. A report is awaited later on this year.
To provide the environmental follow up of the Chaudire River, the Ministry has put in place 4 projects in
August 2014.
1. The follow up of the rivers water quality in flood time (till fall ends 2014);
2. The sampling of fish to estimate the state of their communities;
3. The fish flesh analysis to evaluate their state of contamination;
4. The benthos sampling which will allow to estimated the sediments toxicity.
Finally, the ministry still performs biotests on different types of organisms to document the sediments
toxicity and evaluate the toxicological risks.
With part of its centre burned down and many of the standing building damaged, the citizens and
authorities of Lac Mgantic had to decide what should be done. Save the damaged buildings? Rebuild as
it was? Demolish and start from scratch? Gatherings involving thousands of citizens where held to dream
of the new city. Finally, it was decided to start anew and aim for a greener, more sustainable city. Once
this decision was taken, it was possible to proceed to the demolition of the standing 37 buildings as well
as plan the final cleanup of the contaminated soils found underneath. The buildings demolition took place
during the winter 2015, freeing the ground for the contaminated soils excavation. After the thaw, the soils
will be excavated and carried to the treatment platforms. It is thus foreseen that by the end of 2015, all the
impacted contaminated lands will have been cleaned up, either to the A or to the B values, following the
agreements between the actors involved (government, the town and the new company owning the
railways). The soil treatment should be completed for its part at the latest in March 2017.
CONCLUSIONS
The cleanup work in Lac Mgantic should be completed at the end of 2015. The reconstruction of
downtown will then start, the cleanup level reached in the soils allowing all uses. It is foreseen that the
contaminated soil treatment will be completed by March 2017. The soil, having been treated to the B level,
would then have been all carried at Asbestos and used to green the Jeffrey mine asbestos tailings,
contributing to the environmental quality improvement of this area.
The results of the sediments assessment done in 2014, as well as the results of the diverse toxicological
evaluation projects actually pursued by the ministry will indicate if it is necessary to plan more
interventions on the Chaudire River.
It is estimated that the assessment and cleanup works done in Lac Mgantic as well as on the Chaudire
River will cost approximately 200 millions$. Legal prosecutions are currently ongoing to get the
responsible parties pay for the cleanup.
More than 200 members of the ministry participated in one another interventions which happened on Lac
Mgantic or on the Chaudire River. Along the way, the ministry had to adapt its way of doing things, for
example setting up, as soon as July 2013, a team of managers, professionals, and technicians specially
dedicated to support the various governmental actions. Besides facing complex scientific and technical
issues, those persons had to work on a tight schedule, under the always present scrutinising eyes of the
media, without ever losing sight of the drama lived by the population, victim of the greatest environmental
catastrophe which ever happened in Quebec.
Notes
(1) All the data gathered by the Ministry as well as the maps and reports were made public as soon
as they were validated or confirmed. Those documents can be consulted at the following address
http.((www.mddelcc.gouv.qc.ca(lac-megantic)index.htm )
Session 1C.9 - Bioremediation of chlorinated solvents in groundwater 1
Bewley, Richard.J.F.
AECOM Infrastructure & Environment UK Limited
4th Floor, Bridgewater House, Whitworth Street, Manchester, M1 6LT, United Kingdom
Tel: +44 (0)161 907 3500
richard.bewley@aecom.com
Hick, Paula
AECOM Infrastructure & Environment UK Limited
2 City Walk, Leeds, LS11 9AR, United Kingdom
Tel +44 (0)113 204 5119
paula.hick@aecom.com
Rawcliffe, Anthea
AECOM Infrastructure & Environment UK Limited
4th Floor, Bridgewater House, Whitworth Street, Manchester, M1 6LT, United Kingdom
Tel: +44 (0)161 907 3500
anthea.rawcliffe@aecom.com
ACKNOWLEDGEMENTS
The authors wish to thank the respective clients for agreeing to allow publication of information from
these projects, Regenesis for participation in the remedial designs and review of this document, Tor
Drilling for undertaking the injection of reagents at both sites, Telluric Remediation Management
Limited as contractor at Site 1 and ALcontrol and Jones Environmental Forensics laboratories for
undertaking the contract analyses.
At the first site, TCE concentrations averaging 2900g/l showed little evidence of natural degradation,
with degradation products being less than two orders of magnitude to that of the parent compound (cis
-1,2-dichloroethene (DCE)) <10g/l in the source zone and vinyl chloride (VC) <1g/l. Groundwater
was mostly aerobic, so conditions were unsuitable for reductive dechlorination, and elevated
concentrations of competing electron acceptors such as sulphate were also present. The remedial
strategy consisted of an injection of a lactate-based, hydrogen release compound (HRC). This was
accompanied by HRC primer to acclimatise the groundwater for reductive dechlorination, which was
duly demonstrated by successive increases, followed by decreases in the degradation products, DCE,
VC and ethene. An order of magnitude reduction in total chlorinated ethenes was achieved within the
source area in two years and, despite the initial absence of any significant reductive dechlorination, no
inoculation with Dehalococcoides was necessary. A soil vapour extraction scheme, undertaken in
parallel with the groundwater injection regime, successfully reduced vadose zone concentrations of
TCE in soil above the impacted area. The aeration had no effect on redox conditions and no inhibitory
effects on reductive dechlorination proceeding within the underlying groundwater.
The second site by contrast had been subject to a much greater degree of impact (primarily by PCE),
including the presence of localised Dense Non Aqueous Phase Liquid (DNAPL) in a more cohesive
geological formation consisting of a gravelly clay. Intrinsic reductive dechlorination was already well
advanced but access to significant areas of the source of the contamination was restricted. Therefore,
the overall aim was to mitigate the potential for off-site migration as a first priority, whilst achieving a
reasonable degree of mass removal within the source and plume subject to the constraints of on-site
operations.
A three-stage strategy was implemented involving (i) periodic injection of HRC as a migratory barrier
hydraulically down gradient of the source at the site boundary to protect off-site receptors, (ii)
application of a percarbonate-based chemical oxidation reagent, Regenox followed by HRC in the
source area and (iii) HRC alone in the downgradient plume, the HRC applications comprising both a
primer and an extended release formulation HRC-X. Notwithstanding some continuing dissolution into
the aqueous phase within the source area and immediately down hydraulic gradient from it,
degradation had been proceeding at a steady rate in localities towards the boundary. At the periphery
of the site, total chlorinated hydrocarbons rose from tens of thousands to over 100,000g/l following
initial mobilisation before being reduced by two-orders of magnitude to concentrations of less than
1,000g/l, five years after treatment commenced.
The paper will discuss the variation in Chlorinated Hydrocarbon Compound (CHC) behaviour in
response to the various active remedial intervention according to its nature, severity and distribution
and how remedial progression can be assessed using the Chloride Index, in conjunction with mass
removal.
INTRODUCTION
At Site 1, investigations that had taken place during a change in site ownership identified the presence
of trichloroethene (TCE) at several locations across the site, relating to a series of historical losses.
Most of the plumes associated with these losses were attenuating, apart from one that originated from
the site of a former storage tank, where up to 20mg/l of TCE had previously been detected (Figure1).
Figure 1: Plan of Site 1 illustrating location of The presence of TCE (but none of its
groundwater source area degradation products) was also
identified in samples taken from a
culverted brook within the site boundary
and hydraulically downgradient from the
source area, although concentrations
had not exceeded the Environmental
Quality Standard (EQS) of 10g/l. The
geological sequence beneath the site
comprised between 1m and 6m of made
ground, typically consisting of clayey
sand with varying proportions of gravel,
underlain by alluvial deposits. Shallow
groundwater was encountered towards
the base of the made ground or top of
the alluvial deposits typically at a depth
of between 1.8m and 2.4m in the
proximity of the source area. The
inferred groundwater flow regime was
convergent upon the line of the culverted
brook.
A review of the chemical and geochemical parameters demonstrated that conditions within the
groundwater were sub-optimal for reductive dechlorination (Table 1), particularly the presence of
elevated nitrate, sulphate and non-detectable iron (II). This was supported by the relatively low
concentrations of cis -1,2-dichloroethene (DCE) in relation to TCE at the site and the absence of vinyl
chloride (VC) above detection limits: Whilst a maximum concentration of 20,200g/l TCE was
identified, with DCE mostly non-detectable or below 50g/l. No evidence of DNAPL had been provided
or inferred from any of the previous site investigations.
Due to the operational nature of the site, an in situ remedial approach was essential and
bioremediation through Enhanced Reductive Dehalogenation (ERD) was selected as the best
practicable technique. The remedial design was facilitated using in-house software provided by
Regenesis, according to the groundwater characteristics set out in Table 1, and incorporated both a
microbial demand factor and a safety factor of respectively three times and two times greater than the
theoretical requirement.
The primary source of carbonaceous substrate required to promote reductive dehalogenation through
prolonged release of lactate was HRC supplied by Regenesis. This additionally incorporated HRC
primer, a modified version of HRC, to achieve acclimation of the saturated zone through a more rapid
reduction in redox potential and to overcome competing electron acceptor sinks, primarily sulphate.
The remedial design had identified a requirement for 52 injection points in total, 10 within the source
area and the remainder within the plume at approximately 3m spacings. Due to access constraints, it
was not possible to achieve a uniform grid spacing, however a total of 11 points were located within
Building A, encompassing the source area with an additional 42 hydraulically downgradient (Figure 2).
Figure 2: Location of injection points for HRC and soil The majority of the latter (34 in total)
vapour extraction wells at Site 1 were installed in a 10m width
corridor running between Buildings
A, and B, the distance between each
point ranging from approximately
2.5m to 5m. A cluster of 8 injection
points were also installed around
monitoring well BH406, located
towards the southern edge of the
plume within Building B.
with N being the number of chlorine atoms of the individual cVOC compound and concentration being
expressed in moles/l.
As degradation proceeds the Chloride Index falls: theoretical plumes consisting solely of PCE, TCE,
DCE, VC or ethene would have respective chloride indices of four, three, two, one, or zero. As seen in
Figure 3, the reduction in total mass of organochloride was accompanied by a substantial shift in
Chloride Index in most wells, the exception being well MW301, where less than 10g/l TCE was
detected.
The remedial target was reached or surpassed in all wells after 12 months and concentrations of total
CHCs continued to fall over the following year. After 24 months, concentrations of TCE fell from 2,913
830g/l prior to treatment to 49 46g/l. Whilst VC (non-detectable prior to treatment) was now
present, its concentration (38 26g/) was less than an order of magnitude to that of ethene (540
344g/) after 24 months (mean standard error).
Soil concentrations of TCE in the vadose zone were also reduced to below the remedial target of
144 mg/kg after eight months of operation of the SVE system and, on this basis, it was agreed with the
regulators that no further remedial action was required.
SITE 2
At the second site, which had experienced a number of historical losses of primarily PCE, the geology
consisted of made ground to a maximum depth of approximately 2.3m underlain by natural soft to stiff
brown sandy gravely clay encountered at depths of between 1.75m and 5.0m. Shallow groundwater
was encountered (typically at less than 1m depth) in the made ground and the drift deposits although it
was discontinuous across the site with low recharge rates. Several separate sources of contamination
were present (illustrated by the pink areas in Figure 5) and there was already evidence of significant
reductive dechlorination taking place with elevated DCE, VC and ethene detectable, in contrast to
Site 1 (Table 2). Also the concentrations of PCE present, typically in the tens and hundreds of mg/l in
source areas were indicative of DNAPL.
The primary objective of the remedial strategy at this site was to protect neighbouring properties
hydraulically downgradient, with the secondary objective being to effect reductions in contaminant
mass within the source and plume areas such that natural attenuation could address residual
concentrations. As with Site 1, operational constraints required an in situ remedial strategy, with the
additional challenge of DNAPL in areas of limited access. A three-stage approach was implemented:
Figure 5: Plan of Site 2 illustrating areas for treatment: The barrier, located in the green area in
source, (pink), plume (blue) and barrier (green) Figure 5, comprised two parallel rows of
10 off-set injection points at
approximately 2m to 3m spacings
(i.e. 20 injection points in total). During
January 2009, approximately 54.45kg of
HRC was injected at each location via
the direct push methodology, between
2m and 6m below ground level. This was
repeated some 14, 31, and 50 months
later (March 2010, August/September
2011 and March 2013), the latter
including a further five points extending
back into the site along the assumed
flow path.
At the same time HRC was injected within the plume area (blue zones in Figure 5) at 95 locations in a
herringbone pattern at approximately 2m 3m spacings, predominantly in new direct push locations,
as well as a number of existing fixed wells. A total of 4,658kg of HRC were injected in these areas.
Localised pumping and manual removal of DNAPL from a couple of locations (i.e. where access
permitted) was also undertaken. Since 2011 approximately 88 litres of DNAPL have been removed.
Groundwater and vapour monitoring have been undertaken regularly since the commencement of the
remedial treatment. Monitoring of selected dedicated indicator wells across the source, plume and
peripheral areas has allowed the observation of trends and assessment of the remedial treatment
program.
Most source areas of Site 2 tended to demonstrate an initial increase in TCE and/or DCE following
application of RegenOx, representing desorption of the parent compound from the solid phase into
aqueous solution with rapid conversion to DCE. This was also observed in plume area wells subjected
to only HRC, which is considered to represent the release of sorbed contamination or dissolution of
NAPL through biosurfactants production resulting from stimulation in microbial activity.
The changes in concentration of the key determinands arising as a result of degradation at most
locations have been compounded by the migration of degradation products from treatment areas
hydraulically up gradient and also the application of repeat injections of HRC. Whilst this makes the
tracking of progress difficult, three illustrations are provided of the substantial progress made in the
overall treatment.
These findings are presented in Figure 8 as the total CHCs in wells from within the barrier zone. Refer
to Figure 7 for well locations.
CONCLUSIONS
Conditions at Site 1 were unfavourable for natural attenuation of TCE, with no evidence of
Dehalococcoides activity through minimal DCE and the absence of VC and ethene. Despite these
circumstances, a steady mass reduction accompanied by a sequential increase and subsequent
decrease of daughter compounds were clearly evident following a single injection round of HRC,
without any additional requirement for bioaugmentation by Dehalococcoides.
Reductive dehalogenation was already well established at Site 2, but it was apparent that a significant
mass of contamination (assumed to be present as DNAPL) was present within areas of limited
accessibility. Its ongoing diffusion into the aqueous phase presented significant challenges not only in
terms of providing the requisite amount of hydrogen to achieve degradation but also understanding the
overall extent of degradation achieved. Nonetheless, a comparison of current concentrations with
peak concentrations encountered has demonstrated a substantial reduction in overall mass across the
site as a whole, following a sequential application of chemical oxidation and ERD. The effects of the
latter were also observed in changes to the Chloride Index, and the success of the bio barrier in
attenuating the residual contamination migrating to the site boundary is also clearly evident.
In situ ERD can therefore be an effective approach at active sites for protecting off-site receptors,
despite heterogeneous conditions, DNAPL and restricted access.
Background
The consequences of vapour transport in the unsaturated zone above groundwater aquifers is often
overlooked and the advantages of the investigation and mapping of contaminant transport in confined and
unsaturated media is underestimated.
At a site in Zealand, Denmark, a long standing contamination of the overlying moraine sands and clays with
trichloroethylene (TCE) has been investigated and delineated to a relatively limited area of about 100 m. At
a depth of 12 25 meters under the moraine clay, an unsaturated sandy horizon of about 10 - 15 meters
overlies the regional groundwater reservoir.
Investigation of the soil air content in the unsaturated sand horizon using Geoprobe soundings indicated that
the volatisation of trichloroethylene (TCE) from the TCE hotspot area in the moraine clay has led to diffusion
to the underlying unsaturated zone and contamination of an area of about 5.000 m with soil air
concentrations of 500 1.000 g TCE/m. This contamination in the confined unsaturated zone under
moraine clays constitutes a TCE contaminant mass of about 20 30 g which can continue to burden the
groundwater reservoir for decades.
The contaminant situation with TCE in the unsaturated zone was further complicated by the presence of a
severe contamination with perchloroethylene (PCE-hotspot) at an adjacent site which causes an addition
source of PCE and TCE in the unsaturated and confined sandy horizon providing a one directional and
increasing contribution with both PCE and TCE during the vacuum tests.
Objective
To estimate the amount of contaminant that can be removed by remedial activities involving ventilation of the
unsaturated zone, vacuum ventilation and tracer tests have been performed.
Method
The investigation involved 5 vacuum ventilation tests from two wells each with two 5 m screens in both the
upper and lower part of the unsaturated zone, i.e. just above the aquifer, and one well with a 5 m screen in
just the top of the unsaturated zone. The three wells were situated at distances of 16, 60 and 62 m
respectively, se figure 1. Vacuum was applied individually to the screens while measuring the differential
pressure in the other screens.
The vacuum (max. capacity 103 m/h at a vacuum of 150 mbar) was applied for a period of between 4 and
94 hours during the 5 ventilation tests. The lengthiest test was for the upper screen in the well closest to the
TCE hotspot. Before and after each vacuum test, a soil air sample from the vacuum well was sampled and
analysed.
At regular intervals during the vacuum tests, the content of volatile gases was screened with a PID
(PhotoIonisation Detector), and the content of oxygen, carbon dioxide and methane was measured with a
field instrument.
Furthermore, a tracer test with injection of 20 litres carbon monoxide gas at a well 16 m from the well closest
to the TCE hotspot was carried out to evaluate gas transport time during vacuum extraction. The carbon
monoxide content was measured continually at the vacuum well with an INNOVA gas meter.
Results
The ventilation tests demonstrate and compare the vacuum that can be achieved in the respective screens
in the three wells as illustrated by calculation of the radius of influence according to the vacuum applied as
shown in figure 2.
As expected, the influence due to the pumping from the vacuum well decreases with distance, so the well at
a distance of 62 m to the vacuum well is less affected than the well at a distance of 16 m. The differential
pressure measurements indicate a predominantly homogeneous unsaturated sand horizon in all directions.
-4
The tracer test with carbon monoxide indicated a gas transport speed of 4 x 10 m/s with a vacuum of 54
mbar at the well closest to the TCE hotspot.
Data treatment of the results allowed calculation of the radius of influence at different depths and positions
with different vacuum (pumping) strategies. If the unsaturated sand is sufficiently permeable, the vacuum
test reveals a low differential pressure between wells which provides a large radius of influence, whereas a
large differential pressure indicates a less permeable horizon in which case the radius of influence during
remediation is less, indicating the need for additional vacuum wells.
Furthermore, the change in soil gas concentrations from start to stop of the vacuum ventilation test indicates
the soil gas concentrations that can be achieved under steady state conditions. For the well close to the TCE
hotspot, an increase in TCE soil gas concentrations by a factor of 5 was observed during the ventilation test.
Due to the soil gas contamination with PCE and the degradation product TCE at the PCE hotspot at the
adjacent site, the development in soil gas concentration at the well nearest to the PCE hotspot demonstrated
an increase in soil gas concentrations with PCE and TCE by a factor of 17 and 7 respectively.
Conclusion and relevance
The vacuum ventilation and tracer tests provided valuable information about the extent of pollution in the
confined unsaturated zone and allowed estimation of the potential rates of removal by full scale vacuum
extraction as illustrated in figure 3.
Based on the tests, a theoretical extraction rate of about 2.6 kg /year was calculated for the well closest to
the TCE hotspot, which indicates that remediation by ventilation should provide an adequate solution to the
clean-up of TCE in the unsaturated and confined sandy horizon.
PROGRESS TOWARDS AN ISO DOCUMENT ON SUSTAINABLE REMEDIATION
C Paul Nathanail
School of Geography, University of Nottingham, Nottingham NG7 2RD, UNITED KINGDOM and Land
Quality Management Ltd UNITED KINGDOM. paul.nathanail(at)nottingham.ac.uk +44 115 951 5151
Sustainable remediation is the elimination and/or control of unacceptable risks in a safe and timely
manner whilst optimising the environmental, social and economic value of the work involved.
Since remediation often links into the beneficial redevelopment and reuse of brownfield sites,
sustainable redevelopment is inherently connected with sustainable remediation. Considering the
reuse of a site from the beginning of a remediation project is a fundamental component of sustainable
remediation and therefore sustainable remediation can act either as a natural precursor to or as a
subset of sustainable redevelopment. Much value can be achieved through successfully integrating
remediation into the redevelopment process to exploit synergies, while minimizing costs and
environmental impacts associated with bringing brownfield sites back to beneficial use.
Redevelopment is not the only driver for remediation. Risk assessment, under specific legal contexts,
can demonstrate the need for remediation. Such risk based land management precedes
considerations of sustainable remediation. Sustainable remediation is about how to manage risks that
require intervention; it is not an excuse for doing nothing when you have such risks.
In recent years sustainable remediation has been the focus of special sessions at conferences,
including AquaConsoil, of entire conferences (the third conference on Sustainable Remediation was
held in September 2014 in Ferrara, Italy) and of many peer reviewed papers. Following the setting up
of the Sustainable Remediation Forum (SuRF) in the USA many other countries have established
SuRF networks (e.g. Netherlands, Italy, Australia & New Zealand) or projects (United Kingdom).
Sustainable remediation has also been the subject of a NICOLE working group. A joint statement was
issued by NICOLE and the Common Forum.
Following a proposal submitted by the British Standards Institute (BSI), a working group has been set
up under the auspices of ISO Technical Committee TC190 Soil Quality to develop an informative
standard on sustainable remediation of soil and groundwater. The working group developed material
prepared by a group of volunteer authors into a Committee Draft that was voted on in 2014 and
approved to be developed into a Draft International Standard. The DIS is all but complete and will be
discussed at the next meeting of TC190 in October 2015 and will be voted on by members of TC190
with a view to it being accepted as an ISO document.
Sustainable remediation assessment involves comparing the social, environmental and economic
performance of short listed remediation strategies any of which could achieve the project objectives.
As such it is a relative assessment that seeks to identify the most sustainable of those short listed
options. It seeks to answer the questions: Given that we need to remediate to achieve [these
objectives], what is the most sustainable of our candidate remediation options?
This means that a tiered approach to assessment can be adopted. The principle of parsimony, well
known to computer modellers, can be applied. Simple qualitative approaches may suffice and more
complex, and costly, fully quantitative techniques are only rarely needed.
The choice of relevant indicators however is an important one, whatever method of sustainable
remediation assessment is adopted. A relevant indicator is a single characteristic that represents a
sustainability effect, whether beneficial or adverse, that can be compared across alternative
remediation strategies to evaluate their relative performance. Characteristics that do not vary among
the alternatives are not useful indicators. As such the choice of indicators will be a site specific one
and no generic list is possible.
Indicators need to reflect the economic, environmental and social aspects of remediation strategies.
The economic aspects of remediation have traditionally been the focus of both voluntary and enforced
action. Polluters, other responsible parties or developers of sites have sought the lowest cost
remediation that fits with their project objectives and framing. Remediation, whether of small-scale
neighbourhood sites to large urban or peri-urban sites, can have important impacts on individuals or
communities, including nuisance, loss of amenity and improving health. Site remediation activities
often have direct environmental impacts including air emissions, water consumption, non-renewable
energy utilization, landfilling of waste and resource usage. Of course the remediation often delivers
considerable environmental benefits, including protecting biodiversity, restoring water quality and
reversing damage to soil functionality.
Governance is the use of institutions, structures of authority and even collaboration to allocate
resources and coordinate or control activity in society or the economy. Corporate and regulatory
governance are central to remediation strategy selection. As such, governance can help or hinder the
adoption of sustainable remediation. The site specific legislative, policy and institutional structures will
have a major influence on the potential for achieving sustainable remediation.
Sustainable remediation benefits alignment motives among those who decide how remediation should
be carried out. The concept of sustainable remediation can be required by law or corporate policies.
Other influences (beyond legislative, policy and institutional controls) have a role to play in bringing
about a situation where decision makers and their stakeholders pursue remediation that is
sustainable.
This paper is based on the collective efforts of active members of the ISO working group and of those
volunteers who prepared text before the working group was set up. Their role is acknowledged by the
author who accepts responsibility for any errors or omissions in the above text.
1C.2 - Integrating sustainable remediation into other policies.
Lisa Pizzol, Ca Foscari University of Venice, Department of Environmental Sciences, Informatics and
Statistics, Calle larga S.Marta 2137; phone: +39 (0)41 2348652; e-mail: lisa.pizzol@unive.it
Keywords: social networks, stakeholder participation, social science methods, sustainable regeneration,
Nexsus research project, Vega Science and Technology Park
Introduction
The involvement of stakeholders in regeneration processes has recently received increasing attention
in the literature on the sustainable management of land and of regenerated areas. Progress is quite
recent in this area, going back one decade or so (RESCUE 2005, REVIT 2007). Stakeholders and
stakeholder involvement has first been conceptualized in the business and management literatures, for
example by Freeman (1984) and Freeman and Reed (1983). Approaches to stakeholder participation
have also emerged in other disciplinary contexts, such as social activism, adult education, applied
anthropology, complex systems, natural resource management and ecology (Reed 2008). All these
have been successfully applied in the environmental management literature (Reed et al. 2009) and an
emerging trend can be discerned in the literature on land management and regeneration (Cundy et al.
2013). Especially the new drive towards sustainable requalification includes the participation of
stakeholders as one of its key principles (Wehrmeyer et al. 2007, Cundy et al. 2013). In this paper, we
argue that further progress in understanding and managing stakeholder involvement in regeneration will
be shaped by the further incorporation of concepts and approaches from the social sciences into
regeneration theory and practice. We illustrate this by outlining the knowledge gains offered by a social
network approach to stakeholder involvement. This approach is being developed as part of a Marie
Skodowska-Curie individual fellowship whose incumbent (identified as the first author above) is housed
and coordinated in the Department of Environmental Science, Informatics and Statistics at CaFoscari
University of Venice, Italy. The name of the research project is Nexsus, that is, Network-based Expert-
Stakeholder Framework for Sustainable Remediation)
Involving stakeholders in brownfield regeneration has not always been seen as a priority. It was only
after the publication of the Brundtland report in 1987, that participation became a norm in sustainable
development (Gross and Bleicher 2010, Reed 2008). From that moment on, there was a growing interest
in who is in and why, (Reed et al. 2009), in environmental management and the management of
contaminated land. The concept of stakeholders, nurtured in the business management literature,
became widely used in environmental management. Stakeholders can be conceptualized both in a
narrow and in a broad sense. In a narrow definition, the focus is on the economic survival of the firm
(Mitchell et al. 1997). Stakeholders are individuals without whose support the organisation would cease
to exist (Bowie, 1988 as cited in Mitchell et a. 1997: 858). Freeman and Reed (1983: 91) also provide
a narrow definition of the stakeholder concept as an identifiable group or individual on which the
organization is dependent for its continued survival. A broader definition of stakeholders is offered by
Freeman and Reed (1983) alongside their narrow definition. The wide sense of stakeholder is that of
any identifiable group or individual who can affect the achievement of an organizations objectives.
(1983: 91). For practical purposes, it is important to have a way to identify the stakeholders who play a
role in a specific case of brownfield regeneration. This process is called stakeholder analysis and
consist of: i) dening aspects of social or environmental processes affected by decisions or actions; ii)
identifying social actors, (individuals, groups or organisations) who are affected by or can affect some
aspects of the phenomenon; and iii) prioritising these actors for involvement in decision-making (Reed
et al. 2009).
Reed et al. (2009) also distinguish between normative and instrumental approaches to stakeholder
analysis. Whereas the former is motivated by some principled argument (e.g. stakeholders should be
involved as it is their right to do so), instrumental approaches have pragmatic ends in view (e.g. in
order to get stakeholders to accept a given technology, they need to be involved). Regardless of the
approach pursued, it is important to recognize that the three steps defining the social and
environmental systems affected by decisions, identifying groups and organisations affected by or
affecting those systems and prioritising stakeholders for inclusion in decision-making require a social
scientific methodology.
Since the earliest attempts to consider stakeholder participation as part of contaminated land
management (e.g. Clarinet 2002), questionnaires were used to collect data, often in comparative
perspective. The focus was both on Europe (Clarinet 2002) and on European countries in comparison
to the US and China (Hou and Tabbaa 2014).
Other social science methods, such as semi-structured interviews or focus groups, were less frequently
used, but the review of stakeholder analysis methods by Reed et al. (2009) reveals considerable
potential for successful application in sustainable regeneration. For example, these authors discuss, in
addition to the two methods mentioned above, the following eight methods for stakeholder analysis:
This list of methods suggests the broad scope of social science methods applicable to the analysis of
stakeholder involvement in sustainable regeneration. These methods are of varying usefulness
depending on the case discussed and on the research focus of the investigators. In most case, social
science researchers opt for a combination of methods. In what follows we will introduce the research
design of the Nexsus research project and illustrate the benefits and challenges of applying a particular
combination of social science methods to investigate a case of sustainable brownfield regeneration.
This section introduces the design of the Nexsus research project. The aim of Nexsus (Network-based
Expert-Stakeholder Framework for Sustainable Remediation) is to provide a novel approach to
conceptualize stakeholder involvement in brownfield regeneration by means of a case study. The
approach is new in that it proposes: (1) a new conceptualization of stakeholder involvement, applicable
to sustainable regeneration; (2) a research design that includes several methods linked to each other
as part of a research process; (3) a case study to serve as an empirical basis for developing a framework
for stakeholder involvement in sustainable regeneration.
The theoretical argument is that, in any regeneration project, stakeholders are not isolated individuals
or organizations, but are social actors involved in structured relationships (i.e. relationships that are
stable and enduring) with each other. Describing and understanding these structures is important when
one recognizes that the management of contaminated land is a multi-actor and multi-purpose process
or, in conceptual terms, a governance process. Governance refers to the structures and processes by
which people in societies make decisions and share power (Folke et al. 2005: 444). Governance
assumes that managing contaminated land is not the task of only one designated actor (e.g. clean-up
company) and it also not the exclusive precincts of experts (from a biochemical and physical perspective
only) (cf. Bodin and Prell 2011). Instead, different actors make their influence felt in brownfield
requalification projects. This has already been recognized in the sustainable regeneration literature, for
example in the rainbow diagram that classifies stakeholders according to whether they are affecting,
affecting and affected or only affected by a regeneration project (Chevalier and Buckles 2008; Cundy et
al. 2013). What has been less recognized so far is how actors involved in the same requalification project
might influence each other, via communication ties. Hou and Tabbaa (2014), for example, approache
the growing interested for sustainable land management in terms of stakeholder demand and
institutional processes. Regulators, site owners, consultants and broader social groups can all make
demands on regeneration project managers or on other actors for achieving sustainability in
requalification. We argue that the flow of demands and also of other forms of communication can best
be investigated by means of social network analysis, which enables the study of how resources, goods
and information flow through particular configurations of social ties (Bodin and Prell 2011: 10). Social
network requires, however, a number of preparatory steps that are integrated in a unitary research
design.
The research design for network analysis draws on the work of Prell and her colleagues (2009). The
steps in identifying the network and measuring the ties among actors are described in the table below
The strength of this succession of steps is that it starts with an empirically grounded understanding of
sustainable regeneration in terms that the stakeholders themselves consider relevant. The
understanding of the sustainable requalification issues and of who is involved in this process is done in
terms of a dialectic relationship (Prell et al. 2009). Clarifying the set of sustainable regeneration issues
over which actors interact, via the semi-structured scoping interviews, also allows the identification of
the key actors involved, by means of the ego-centred maps, in which respondents are asked to
nominate, at the end of the interview, those they communicate with on the topics they have specified in
the interview. At the end of the first two steps, the researchers have a list of sustainable regeneration
issues and also a list of potential network members. The third step involves the synthesis of the collected
information into one tool a questionnaire which allows the identification of social networks and the
measurement of the nature and intensity of sustainable regeneration concerns. The fourth step allows
the testing of the overall governance concept characterizing the diversity of actors involved in a
network and of more specific propositions such as those formulated by Hou and Tabbaa (2014).
Finally, the fifth step addresses the need to generalize the findings and make them applicable to different
cases in which the configuration of stakeholder involvement is likely to play a role, as expected in most
of the complex environmental management processes (Liu et al. 2007). Network indicators such as the
centrality of individual actors, the density of the network as a whole or the existence of cliques (or
cohesive subgroups within a generally loose configuration) within the network (Prell 2012) can signal
for other projects to what extent the involvement of all stakeholders or of specific categories of
stakeholders is effective or not.
Both the identification of networks organized around specific issues and of sustainable regeneration
concerns require the use of a a strategic case study. By the latter, we mean a case that displays all the
conceptual constructs of interest, namely: active networks of actors who communicate on ongoing
matters of concern related to the regeneration of brownfield sites and a concern with the sustainability
of regeneration at different time scales. The case study chosen for the Nexsus project, briefly described
below, meets both these criteria.
The Vega Science and Technology Park consists of four ex brownfield sites (identified as Vega 1, Vega
2, Vega 3 and Vega 4), located in the first industrial area of Porto Marghera, Venice, Italy. Vega 1 has
undergone a regeneration process between 1993 and 2006, which has resulted in eight new or
refurbished structures for scientific research, technology development and service provision (Vega
2006). Between 2007 and 2011, no regeneration projects were carried out. Since 2012, the Vega 2 area
is undergoing a new regeneration phase aimed at establishing a new exposition, residential and
commercial space. There are a number of projects and strategies that are currently considered and
discussed for the future of the Vega 2 area, both over the immediate term and long term.
The Vega 2 case fits the conceptual requirements of the Nexsus research project. It enables the
simultaneous identification of active networks and of sustainability issues and of the relationships
between them. First, it allows the comparison of social networks linking actors who communicate about
projects with immediate implications and of those in which actors discuss strategies with longer term
implications. Sustainability networks are thus captured in relation to immediate and to more distant
goals. Second, the Vega 2 case allows the identification of sustainability goals (via open-ended
questions) and also the assessment of the relevance of sustainability indicators (derived from the
sustainable regeneration literature and from the interviews with key respondents). Third, and most
importantly, the Nexsus research can reveal in the case of Vega 2 how stakeholders with different
sustainability concerns are related to each other as part of networks.
Conclusion
This paper reports on an ongoing research project designed to use social network analysis as an
operational definition of stakeholder involvement in sustainable regeneration projects. The collection of
data via the research design outlined in Table 1 is currently at step 3. However, the steps carried out so
far have demonstrated the utility of a social science approach in conceptualizing and measuring
stakeholder involvement in regeneration projects. Social network analysis is not meant to replace
previous stakeholder analysis approaches, such as those based on questionnaire research, but to shed
light on a previously underexplored facet of stakeholder involvement in sustainable regeneration
projects. Our focus on communication ties among actors can inform both researchers and practitioners
on the following topics: how is information likely to flow among stakeholders; who are the centrally
located and the peripherally positioned actors; how are actors with similar kinds or levels of sustainability
concerns positioned in relation to each other etc. These and similar questions are important for
understanding how decisions are or can be communicated and how power relationships for example
between those who make decisions and those who have to comply with such decisions are played out
among stakeholders.
References
Bleicher, Alena, and Matthias Gross. 2010. Sustainability Assessment and the Revitalization of
Contaminated Sites: Operationalizing Sustainable Development for Local Problems. International
Journal of Sustainable Development & World Ecology 17 (1): 5766. doi:10.1080/13504500903488263.
Bodin, rjan, Ramirez-Sanchez, Saudiel, Ernstson, Henrik, and Prell, Christina. n.d. A Social Relational
Approach to Natural Resource Governance. In Social Networks and Natural Resource Management:
Uncovering the Social Fabric of Environmental Governance, 328. Cambridge, New York: Cambridge
University Press.
Chevalier, Jacques M., and Buckles, Daniel J. 2008. SAS2: A Guide to Collaborative Inquiry and Social
Engagement. Los Angeles, London: Sage.
Folke, Carl, Thomas Hahn, Per Olsson, and Jon Norberg. 2005. Adaptive Governance of Social-
Ecological Systems. Annual Review of Environment and Resources 30 (1): 44173.
doi:10.1146/annurev.energy.30.050504.144511.
Freeman, Edward R., and Reed, David L. 1983. Stockholders and Stakeholders: A New Perspective
on Corporate Governance. California Management Review XXV (3): 88106.
Hou, Deyi, and Abir Al-Tabbaa. 2014. Sustainability: A New Imperative in Contaminated Land
Remediation. Environmental Science & Policy 39 (May): 2534. doi:10.1016/j.envsci.2014.02.003.
Liu, Jianguo, Thomas Dietz, Stephen R. Carpenter, Marina Alberti, Carl Folke, Emilio Moran, Alice N.
Pell, et al. 2007. Complexity of Coupled Human and Natural Systems. Science 317 (5844): 151316.
doi:10.1126/science.1144004.
Mitchell, Ronald K., Bradley R. Agle, and Donna J. Wood. 1997. Toward a Theory of Stakeholder
Identification and Salience: Defining the Principle of Who and What Really Counts. The Academy of
Management Review 22 (4): 85386. doi:10.2307/259247.
Paul BARDOS, Anita LEWIS (r3 Environmental Technology Limited, United Kingdom), Stephen
NORTCLIFF (University of Reading, United Kingdom), MATIOTTI (Aquater, Italy), Franck MAROT
(ADEME, France), Terry SULLIVAN, Terry SULLIVAN (BNL, USA), and Franck MAROT (ADEME,
France). n.d. Review of Decision Support Tools for Contaminated Site Management, and Their Use in
Europe. Umweltbundesamt GmbH (Federal Environment Agency Ltd).
http://www.commonforum.eu/Documents/DOC/Clarinet/brownfields.pdf.
Prell, Christina. 2011. Social Network Analysis: History, Theory, Methodology. Los Angeles, London:
Sage.
Prell, Christina, Klaus Hubacek, and Mark Reed. 2009. Stakeholder Analysis and Social Network
Analysis in Natural Resource Management. Society & Natural Resources 22 (6): 50118.
doi:10.1080/08941920802199202.
Reed, Mark S. 2008. Stakeholder Participation for Environmental Management: A Literature Review.
Biological Conservation 141 (10): 241731. doi:10.1016/j.biocon.2008.07.014.
Reed, Mark S., Anil Graves, Norman Dandy, Helena Posthumus, Klaus Hubacek, Joe Morris, Christina
Prell, Claire H. Quinn, and Lindsay C. Stringer. 2009. Whos in and Why? A Typology of Stakeholder
Analysis Methods for Natural Resource Management. Journal of Environmental Management 90 (5):
193349. doi:10.1016/j.jenvman.2009.01.001.
RESCUE Project. 2005. The RESCUE Manual: Best Practice Guidance for Sustainable Browneld
Regeneration. Land Quality Press, a Division of Land Quality Management Ltd.
REVIT Project. 2007. Working towards More Effective and Sustainable Browneld Revitalisation
Policies, Stakeholder Engagement a Toolkit. Interreg IIIB Project. http://www.revit-
nweurope.org/selfguidingtrail/27_Stakeholder_engagement_ a_toolkit-2.pdf.
Vega - Parco Scientifico Tecnologico di Venezia. 2006. Bilancio Di Responsabilit Sociale 2005.
Wehrmeyer, Walter, Kalliope Pediaditi and Kate Burningham. 2007. Measuring Sustainability: Whats in
a Number?. http://www.claire.co.uk/index.php?option=com_phocadownload&view=file&id=31:case-
study-bulletins&Itemid=230.
PAPER:
1C.14 - Combined treatment technologies 1
DNAPL TREATED BY APPLICATION OF SURFACTANTS FOLLOWED BY ISCO
Petr Kozubek 1, Jan Nmeek 1, Elika Kosinov 2, Vladislav Knytl 2
1 ENACON s.r.o., Krcska 16, 140 00 Prague, Czech Republic +420 241 404 911,
kozubek@enacon.cz, nemecek@enacon.cz
2 Dekonta a.s., Steheleves, Czech Republic kosinova@dekonta.cz, knytl@dekonta.cz
Introduction
The presence of DNAPL is often limiting factor for the removal of chlorinated aliphatic hydrocarbons
(CHCs) from the subsurface. Residual DNAPL usually results in the rebound effect causing repeated
increase of contaminants concentrations. Surfactants (surface active agents) are chemical compounds
with specific molecular structures typically composed of a strongly hydrophilic head and a hydrophobic
tail. In a DNAPL-contaminated aquifer, this specific property of surfactants can increase DNAPL
solubility and lower DNAPL-water interfacial tension to the point that physical mobilization takes place.
Application of surfactants thus can enhance the remedial action. Surfactants flushing is a mature
technology in the petroleum-engineering field. The technology has been shown to be effective also for
DNAPL sites recently. A surfactant flushing system typically consists of a network of injection and
extraction wells. A mixture of injected fluid and mobilized contaminant is captured through extraction
wells that requires further treatment. This treatment could be technically demanding (e.g. due to the
tendency of surfactants to form foam). Considering that coupling of surfactants application with
another remedial technology seems to be suitable. In-situ chemical oxidation (ISCO) is promising
technology for the combination with surfactants. In addition, the main weakness of ISCO application to
treat DNAPL thus could be solved.
Test site
The mechanical engineering site situated in the western part of the Czech Republic has been selected
for the pilot test performance. The hotspot area is located in the north-eastern portion of the main
production hall where degreasing by trichloroethylene TCE and tetrachloroethylene PCE was
performed in the past.
From the geological point of view, the site belongs to Upper Proterozoic Unit of Central Bohemia,
called Tepelsky crystalline complex formation. The complex is formed by two-mica schist. In the area
of the pilot test, the upper most layer is formed by fill represented by sandy-clayey loams with
fragments of stones thick approximately 0.3 to 1 m. Fill materials are underlain with mica-schist
elluvium formed by sandy to sandy-clayey loams with fragments of bedrock stones up to 12 cm. The
bedrock was encountered in a depth of 1.2 to 2.5 m.
The shallow aquifer is bound to the zone of weathered bedrock and overlying sandy to clayey loams.
The aquifer permeability is relatively low represented by hydraulic conductivity of 10-6 to 10-7 m/s.
The groundwater contamination at the area of pilot test before its performance indicated presence of
DNAPL (PCE 31.3 mg/l and TCE 23.1 mg/l).
The separate application of surfactant was based on injection of solution of surfactant into two wells
IN-1 and IN-2. Another two wells MV-1 and MV-2 were used of groundwater extraction. The extracted
water was treated via striping and trap of chlorinated hydrocarbons on activated carbon. The treated
water was then injected again to wells IN-1 and IN-2.
This phase of pilot test was performed in three in three sequences, each in duration of 30 days. The
volume of surfactant was the same in all three sequences of 100 kg. The initial concentration of
surfactant was 50 g/l, the latter two ones of 100 g/l. Surfactant was distributed equally 1:1 to both
injection wells IN-1 and IN-2. The groundwater extraction was initiated approximately 16 hours after
surfactant injection from wells MV-1 and MV-2. The extraction was conducted for period of 21 days.
Between the second and third sequences the extraction was prolonged for another 21 days in order to
compare the efficiencies of steps with and without surfactant injection.
Approximately one year after the initial phase of pilot test his second phase has been performed
comprised combined application of surfactant and chemical oxidant. Surfactant in concentration of 50
g/l has been injected to the groundwater via 4 injection wells MV-1, MV-2, IN-1 and IN-2 in total
volume of 2 cubic meters. Subsequently, after 16 hours, 10 cubic meters of activated sodium
persulfate in concentration of 100 g/l was injected to the same four wells. The groundwater monitoring
followed the application. Once certain increase of CHCs has been detected the injection of persulfate
was repeated after 70 days from the start of test.
Results
The application of surfactant resulted in the multiply increase of concentrations of chlorinated
hydrocarbons (up to 13 times) confirming the efficiency of surfactant to mobilize the residual DNAPL;
see following graph for well MV-1.
Graph 1: Development of CHCs and surfactant concentrations during surfactant injection, the arrows
show individual sequences of surfactant injection (red ones injection of surfactant and subsequent
groundwater extraction, black one only extraction without preceding application of surfactant)
The application of surfactant was also evaluated from the remedial efficiency point of view. The
pump&treat method applied at the site prior the pilot test removed 2.2 kg of chlorinated hydrocarbons
during 4 months. In comparison, 10.6 kg of CHCs, i.e. approximately 5 times more, was removed by
the pilot test in duration of 85 days (incomplete 3 months).
In case of combined application the similar effect of surfactant has been observed, i.e. immediate
increase of CHCs concentrations up to factor 6. Application of oxidant resulted in reduction of
chlorinated hydrocarbons mobilized by surfactants. The low level of CHCs remained for 65 days where
certain increase of concentrations was observed and another injection of oxidant was performed
(without previous preceding application of surfactant). More intensive drop of CHCs concentrations
was observed after the second injection (see following graph). The similar development has been
observed in all four injection wells.
Graph 2: Development of CHCs during combined application of surfactant and oxidant (the vertical
lines show injection of surfactant (green) and oxidant (red))
Conclusion
The results of pilot test indicate the applicability of combined use of surfactants and in-situ chemical
oxidation at the DNAPL-contaminated sites. Surfactant has very positive influence of dissolution and
desorption of DNAPL that become available for the subsequent chemical oxidation.
Surfactant does not remain in high concentrations for long time, it is being degraded and chemically
oxidized. The results of the pilot test and previous laboratory experiments do not conclude that
oxidation of surfactant is preferential to oxidation of chlorinated hydrocarbons.
The performed work was granted by Czech Ministry of Industry and Trade.
References
Keuper B.H., Stroo H.F., Vogel C.M. a Ward C.H. eds.: Chlorinated Solvent Source Zone
Remediation. Springer. New York. 2014
Kram M.L., Keller A.A., Rossabi J. a Everett L.G. (2001): DNAPL Characterization methods and
approaches, Part I: Performance comparisons. Ground Water Monitoring & Remediation, 21 (4), 109-
123.
Liang C., Wang Z.S. a Bruell C.J. (2007): Influence of pH on persulfate oxidation of TCE at ambient
temperatures. Chemosphere 2007, 66 (1), 106-113.
M
iraglio M.A. (2009): Base-Activated Persulfate Treatment of Contaminated Soils with pH Drift from
Alkaline to Circumneutral. Diplomov prce, Washington State University, 2009.
Pankow J.F. a Cherry J.A. (1996): Dense Chlorinated Solvents and other DNAPLs in Groundwater:
History, Behaviour, and Remediation. Waterloo Press., 522 pp.
Session 1C.21 - New remediation technologies 2
Summary
The site (Plate 1) was previously one of the most contaminated in south-west England, in terms of the range
of organic, inorganic and other contaminants present and the magnitude of these. It has a long industrial
history as a munitions and chemical warfare plant during World War One, followed by zinc and lead smelting
with sulphuric acid production until 1972, and then production of pharmaceuticals, agrochemicals and
refrigerants until 2008, when the site was closed. Preliminary investigations undertaken to support a
planning application for redevelopment of the site identified multiple contaminant source zones with impact
from metals, petroleum hydrocarbons and chlorinated solvents (mainly chloroform and
trichlorofluoromethane at concentrations likely to represent Dense Non-Aqueous Phase Liquid (DNAPL)).
The potential risk of Unexploded Ordnance (UXO) and mustard gas also posed unusual Health and Safety
management challenges.
To address data gaps within the existing site conceptual model, generated by others, ERM completed a
programme of High Resolution Site Characterisation (HRSC), using a combination of a CPT/Membrane
Interface Probe (MIP) and, for the first time in the UK, a BATTM sampler to allow discrete groundwater
samples to be collected and hence provide accurate assessment of chlorinated solvent mass distribution.
The HRSC works identified that a majority of the contaminant mass was within the underlying confined
Fluvial Glacial Sand (FGS) aquifer, present to a depth of up to 14.5m below ground level (bgl). The results
are shown on Figure 1.
ERM also completed an innovative sequence stratigraphy assessment of the macro-scale traditional and
micro-scale HRSC geological data obtained. The results identified discrete channel structures representing
higher permeability pathways, interfluve swamp deposits comprising laminae and thin beds of fine sands,
which could be preferential pathways, but with relatively low pore volume owing to their thinness and
siltiness. It was determined that organic contaminant plumes in the interfluve swamps would be more
difficult to remediate as contaminants will be trapped in the matrix, sorbed onto the peats and in low
permeability strata (Figure 2). This exercise combined with the HRSC was instrumental in allowing the FGS
aquifer to be remediated in a cost effective and sustainable manner, by refining the treatment zone to reflect
locations of greatest mass in areas where in-situ remediation could be realistically targeted.
Figure 1: HRSC data used to define treatment zone Figure 2: Sequence stratigraphy
extent interpretation (Plan View)
Key:
It was identified that a combination of remediation techniques would be most appropriate to address
chlorinated solvent source zones in shallow soils (soil excavation and ex-situ soil vapour extraction),
perched water (dewatering) and deeper groundwater (hydraulic recovery and Surfactant Enhanced Aquifer
Restoration (SEAR)):
1. Shallow Soils (Source Treatment): A key aim of this work was to reduce soil volumes for off-site
disposal to a minimum and a method to allow the material to be treated and re-used on-site was
therefore designed. During this evaluation four key criteria were considered; achieve the site specific
remediation targets, minimise off-site disposal, maintain the current geotechnical properties of the soil
and minimise contaminant release to atmosphere.
2. Perched Water (Pathway Interception): A methodology was required to inhibit migration of perched
water from the chlorinated solvent source zones towards the Rhines surrounding the site. A number of
techniques were considered, including the installation of a permeable reactive barrier and lining the
surrounding channels with an impermeable membrane. Celtic designed a low cost and pragmatic
solution, comprising installation of a trench to capture water along the site boundary followed by water
treatment.
3. Perched Water (Source Treatment): Celtic had to develop a treatment methodology that not only
allowed perched water within the shallow soils to be removed in a controlled manner but could also be
undertaken in conjunction with the soil excavation work. Celtic designed a series of interconnecting
trenches, filtration sumps and settlement lagoons to remove a large proportion of the suspended solid
content prior to the water treatment system therefore maintaining a high treatment efficiency.
4. Deep Saturated Zone (Source Treatment): The remediation of the DNAPL located within the deep
saturated zone presented the most technically challenging problem, with several potential techniques
initially considered and sequential application of conventional hydraulic recovery (Stage 1) and SEAR
(Stage 2) eventually implemented. The hydraulic recovery was aimed at not only removing contaminant
mass but also further delineating the contaminant plume and assessing which wells yielded the greatest
mass recovery, with the aim of increasing efficiency of the SEAR and sustainably by precisely targeting
wells with the greatest contaminant mass. Celtic designed laboratory bench and column tests to be
undertaken in their UK and Canadian Laboratories to determine the most suitable surfactant mix,
amount and concentration.
Remedial Implementation
Shallow Soils: Approximately 22,000 tonnes of contaminated material was excavated from five locations
across the site. By applying a stringent excavation, segregation, assessment, classification and stockpile
management system, 100% of the excavated material remained onsite. Approximately 50% of the material
was immediately re-used onsite (without treatment being required), with 48% requiring treatment. Treatment
was via ex-situ soil vacuum extraction utilising the same principles as an aerated biopile. In order to achieve
a sufficient pore volume exchange rate to achieve the treatment programme over 4km of aeration pipework
was installed underneath the contaminated soils and connected to a high flow vacuum system. Recovered
vapours were treated via activated carbon for removal of VOCs prior to atmospheric discharge. Water
extracted from the soils was collected and separated from the extracted vapour and then was treated via
the on-site water treatment system. Over an 11 week period contaminant concentrations in the soils were
reduced to a level where they could be safely re-used on-site.
Deep Saturated Zone (Source Treatment) - Stage 1: A total of 59 injection/ extraction wells were installed
to a maximum depth of 13.5m bgl and connected via a network of above ground pipework (Plate 4). Detailed
logging and core recovery during well installation enabled more detailed delineation of the plume. The
recovered cores were also used in laboratory trials of the design of the surfactant injection works.
The first phase of pumping recovered over 1,000 kg of contaminant mass from over 5,000 m3 of groundwater
in 6 weeks of operation. All groundwater was re-injected back into the aquifer following on-site treatment.
Deep Saturated Zone (Source Treatment) - Stage 2, Laboratory and Field Trials: Prior to full-scale
implementation of the SEAR, comprehensive in-house laboratory trials involving vial tests and soil column
tests were carried out, with the aim of selecting the most suitable surfactant for both the contaminants of
concern and the aquifer geology. Initial criteria for selection of appropriate surfactants for use in the trial
included:
Environmental acceptability i.e. the surfactant must be biodegradable and have low toxicity;
Ability to cause significant reduction in the interfacial tension between water and non-aqueous
phase liquids (NAPL); and
Low critical micelle concentration (CMC). The surfactant concentration at which micelles form must
be low in order for the surfactant to be effective.
Chloroform and trichlorofluoromethane (freon-11) had been identified as the principal contaminants of
concern and chloroform was selected as the substance to be used in the trials. Freon-11 could not be used
due to its low boiling point (24 C) and high vapour pressure (89 kPa at 20 C) which would result in
significant losses hence significant margins of error in all stages of handling during the trial. However, due
to similar viscosities and densities of chloroform and freon-11, it is likely that a surfactant chosen on the
basis of successful solubilisation of chloroform should also be effective with freon-11.
The vial testing (Plate 5) assessed 8 different surfactants and 27 mixes comprising different concentrations
and co-solvent combinations. The vial testing identified that a 2% solution of polysorbate 20 was the most
effective solution at solubilising chloroform and produced a corrected eluent concentration of chloroform of
54,000 mg/l, which was complete solubilisation of the 1 ml of chloroform in a 30 ml vial (rest filled with
deionised water). This solution was therefore progressed to the soil column testing stage (Plate 6).
Subsequent soil column testing produced effluent results of 7,000 mg/l using Tween 20 at 2%, compared
to 3,500 mg/l in the control (only twice as high as the control but as expected due to very high solubility of
chloroform of 8,000 mg/l) and indicated the likely success of the selected surfactant in simulated aquifer
conditions at residual saturation.
Prior to surfactant injection, site specific effective porosity and zone of influence were determined by tracer
testing and injection testing. Sodium bromide was used as the conservative tracer, detectable by field testing
using a conductivity probe. 0.5% Potassium permanganate was used as a visual marker. The tracer testing
concluded that the effective porosity of the deep aquifer was 7% and a zone of influence of at least 7 m was
confirmed.
Deep Saturated Zone (Source Treatment) - Stage 2, DNAPL Delineation: In order to maximise mass
recovery, and effectively target the surfactant injection to areas of residual NAPL mass, a detailed
delineation was undertaken. Given the lack of free phase DNAPL identified in boreholes the delineation
focussed on dissolved concentrations, from three rounds of groundwater monitoring, which were considered
indicative of residual NAPL. The EA R&D publication states that an up-gradient DNAPL source is likely to
be present where down-gradient monitoring wells record concentrations in excess of between 1% and 10%
of contaminant solubility. An intermediate value of 5% solubility was assumed for the purpose of this
estimate. Literature values for the solubility of the key contaminants, chloroform and Freon-11, are 8,950
mg/l and 1,100 mg/l respectively. However, Celtic commissioned Jones Environmental Forensics to conduct
a study of the combined solubility of chloroform and Freon-11 in the presence of groundwater sampled from
the site, which indicated that their solubility, when in combination, converge at approximately 3,000 mg/l.
As such, the area where concentrations exceed 150 mg/l was assumed to be indicative of the presence of
DNAPL for both substances. These areas are indicated in Figure 4.
0m 50 m
Deep Saturated Zone (Source Treatment) - Stage 2, SEAR: Based on laboratory trials the SEAR works
were designed to deliver a single effective pore volume within the target area of 1,361 m2 and average
aquifer thickness of 3.5 m. To effectively distribute the 343,000 litres of surfactant within the identified area,
eight boreholes were selected for surfactant injection at the locations shown on Figure 5, below.
CBH37
CBH38
179530 CBH39
CBH48 CBH40
CBH41
CBH42
179520
179510
0m 50 m
179500
352510 352520 352530 352540 352550 352560 352570 352580 352590 352600
The surfactant solution was prepared to the required 2% concentration by dilution with treated site water
within a mixing tank using paddle mixers. The solution was passed through an additional in-line static mixer
prior to passage through a custom injection manifold which allowed the flow rates to individual boreholes to
be controlled and measured. Simultaneous injection into all injection boreholes was carried out and the
injected solution was allowed to equilibrate within the aquifer for a period of 7 days prior to commencement
of hydraulic recovery. Additional boreholes and additional pumps were added during the second phase of
hydraulic recovery to target abstraction of the injected surfactant solution from the areas with the highest
contaminant concentrations. As such, the recovery effort increased during the course of the works and the
resultant overall increase in recovery volumes led to a proportional increase in mass recovery.
During the post injection hydraulic recovery daily chemical analysis of both groundwater influent/effluent
and in-borehole concentrations, as well as recording abstracted water volumes, was carried out to carefully
monitor treatment efficiency (Figure 6) and to ensure that the recovery targeted the correct areas of the
aquifer. The recovered contaminant mixture comprised two principal components; chloroform (37% of total
mass) and trichlorofluoromethane (59% of total mass). No NAPL was observed within any of the settlement
units within the water treatment system during system operation and monitoring.
Surfactant injection was carried out with the objective of enhancing mass recovery from the FGS by enabling
the recovery of residual mass which may not have been recoverable through hydraulic recovery alone.
Injection of surfactant was successful in this objective, recovering a conservative estimate of 550 kg of
contaminant mass over a two day period. Mass recovery gradually decreased in the 6 weeks of pumping
following surfactant injection, following which there was sufficient confidence that the additional mass
liberated by surfactant injection had been recovered, and the programme of treatment was completed, in
line with the asymptotic end point agreed with the regulatory authorities.
At the end of the remediation programme, over 11,690 m3 of groundwater had been treated, removing over
2,500 kg of VOCs in just over 3 months of active recovery.
Post works groundwater monitoring indicated significantly reduced concentrations in the aquifer (Figure 7
below), particularly in areas where pre-works concentrations were indicative of the presence of free phase
contaminant mass.
Figure 7: VOC Concentrations in CBH12, CBH17 and CBH24 Post Surfactant Injection
Note: Dashed lines denote pre-injection concentrations within CBH12 (Blue), CBH17 (Red), CBH24 (Green) and the
Sum of three boreholes (Purple).
The remedial strategy was presented to the client, regulators and other stakeholders (former site owner and
future site owner) and a risk reduction approach of mass recovery to the extent technically and practically
feasible with the identified techniques was agreed with the regulatory authorities. Validation works showed
the following:
Site Specific Concentrations were achieved for all soils post above ground SVE treatment, and all soils
were reused onsite;
The deep groundwater treatment works recovered over 2,500 kg of contaminant mass from perched
water and groundwater. Monitoring showed that VOC concentrations reduced from a pre-remediation
maximum of 9,510 mg/l to a maximum post-remediation concentration of 240 mg/l; and
Post remediation, on-going monitoring of the Rhines showed sustained decline in VOC concentrations
to below the selected quality standards, in all locations.
ThS 1C.7 Strategies for remediation and brownfield regeneration
Contaminated site management in China
Jie Cheng, John Ulrik Bastrup, Daniel Chang
Geo, Maglebjergvej 1, DK-2800 Kgs. Lyngby. jub@geo.dk, jich@geo.dk. +45 4588 4444
Key words: Market, China, Contaminated site
China is experiencing one of the worlds largest urbanization nowadays. Official data from Na-
tional Statistics Bureau shows that in 2013, the ration of urbanization growth in China is 53.7%,
which is 7.7% higher than the year before. One of the key consequences is the fast speed of
land development from original city center towards neighborhood. Industries located in these
1
areas are therefore moved out. According to the research report of the Brown Field Remedia-
1
tion and Re-development in China from World Bank in 2010 , the national trends of re-locating
industries have led to a large amount of brown field leftover in the major cities, waiting for re-
development.
1
Table 1. Brown fields in Chinese major cities due to industry re-location
City / Province Description
Beijing city More than 200 industries (polluters) have moved out from
the inner area of the fourth ring road, leaving more than 8
million m2 brown field for re-development
Shenyang city In 2008, more than 56 industries moved out; in 2009, the re-
location activities covered all industries with heavy pollution
in the main city.
Jiangsu province Over 400 chemical industries moved out of main city, over
6000 small industries closed down, leaving more than 200
million m2 brown field for re-development4.
Shanghai city More than 10 large and heavy industries moved out from the
main city
Zhejiang province More than 100 large industries moved out or closed down
since 2005.
Guangzhou city 147 large industries moved out or closed down since 2007.
Chongqing city 112 industries moved out from the main city in 2010.
1
Brown field: the field with contamination risk which might cause a complexity in the land re-development, refer-
ring to http://www.epa.gov/brownfields/about.htm. In this report, brown field shall be understood as the site with
contamination potentials, while the contaminated site refer to those sites with confirmed contamination.
Left behind by either aged large-scale state-owned factory or rare-monitored private production
facilities, the historical environmental performance on these brown fields is generally considered
poor. Contamination is often discovered on these sites, containing a wide range of categories
from heavy metal, POPs like pesticides and PCBs to organic compounds like hydrocarbons and
chlorinated solvents.
These contaminated brown fields have become an obstacle for Chinas rapid growth of urbani-
zation. Due to the potential short-term and long-term risk towards the site worker as well as fu-
ture land users, the consequential liabilities have caused a significant hesitation within land-
owners, land developers as well as relevant authorities, in which case, the development of a city
is delayed.
From 2005 to 2013, China has conducted the first national mapping of soil contamination. The
2
actual activity covered around 6.3 million km , which is approx. 2/3 of the entire area within the
country board. The result shows that, 16.1% of total sampling spots are confirmed as contami-
nated. More specifically, the result of confirmed contamination has been analyzed by the cate-
gory of historical land use (Table 2).
2
Table 2. First national contamination mapping result of China
Category of historical land use The contamination percentage*
Farmland 19.4%
Heavy-pollute industries 36.3%
Abandoned industrial sites 34.9%
Industrial zones 29.4%
Oil fields 23.6%
Mining sites 33.4%
*: The contamination percentage is calculated out of a certain category of land use respectively.
Although the mapping has not covered the entire country, it is already possible to conclude that
the contamination in industrial sites (or caused by industrial activities) is the major contamina-
tion source for the nation. Based on the total number of heavy industries and the common his-
torical behaviors, some estimation indicates that a conservative number of contaminated sites in
China sum up to half a million or more. These heavy industries comprise chemical industry, iron
& steel, non-ferrous metal processing, coking plants, electronic components production and
5
machinery, hazardous waste storage / land fill, etc.
2/9
2
Table 3. Contaminated site category in China
Historical
Contamination profile Protection target Relevant authorities
land use
High density of hot spots, high con-
Industrial Human health in
centration of contamination, deep EPB*, CHURD**
sites future land use
distribution.
Distribution in large area with shallow Food safety and
EPB, agriculture depart-
Farm land depth, low concentration of contami- ecological system
ment.
nation. in soil
Distribution in large area with shallow Water resource and
Oil fields EPB, Oil authorities
depth, large variety of concentration. ecological system
Distribution in large area with high
concentration some areas, the con- Water resource and EPB, Land resource
Mining sites
tamination normally follows the waste ecological system bureau.
discharge routes.
*: EPB, Environmental Protection Bureau.
**: CHURD, Committee of Housing, Urban and Rural Development.
As the land is a state-owned property in China, the administration of land becomes one of the
key responsibilities of the government. This responsibility is spread over a large and complicat-
ed administrative system from central government to local municipalities. Although it is too chal-
lenging to describe this system in detail, an outline of key stakeholders is provided here to give
the overall picture.
3/9
and distribution when it comes to utilization of the public resource, and coordinating the cross-
organization activities among different public sectors such as Land Resource, Environmental
Protection, Finance, Planning and Municipal Department, etc.
In 2004 and 2006, two individual accident caused by soil contamination in China happened in
Beijing and Wuhan respectively, which are nowadays commonly recognized as the start of pub-
lic awareness arising on soil contamination in China. Meanwhile, two vital international events
Olympic Games in Beijing (2008) and the World Exhibition in Shanghai (2010) have also trig-
gered a great effort in contaminated site management during their years preparation phase.
The year of 2004 is normally recognized as the starting year for initiating Chinese legislation re-
lated to contaminated site management.
A record is provided below to present the development history of the key Chinese legislations.
2004, Chinese EPA, The notice about consolidating the pollution prevention activities dur-
ing re-location process of industries
2008, Chinese EPA, The Comments on enhancing the prevention of soil contamination
2009, Chinese EPA, a series of drafts about contaminated site management such as Risk
assessment guideline on contaminated site
2010, Discussion and approved in principle, The Tentative Measures for contaminated
site management
2012, Chinese EPA, The notice about ensuring the environmental safety during the re-
development of industrial sites
2013, General Office of the State Council, Short-term work plan for soil environment
protection and comprehensive treatment
2014, Chinese EPA, four national guidelines:
National Environmental Protection Standard of P.R.C. Technical Guidelines for
Environmental Site Investigation (HJ25.1-2014)
4/9
National Environmental Protection Standard of P.R.C. Technical Guidelines for
Site Soil Remediation (HJ25.4-2014)
Although it has been 10 years since China started to develop relevant legislations, there is still
no special law for contaminated site management, unlike the other environmental aspects as
water, air emission, solid waste and noise. In the latest National Peoples Congress in 2015,
which represents the highest national authority in China, it is announced that such a law is go-
ing to be initiated in recent years. This is the first time that soil contamination is addressed in
such a strong voice in China.
Due to the lack of a solid law, the enforcement of contaminated site handling has been poor in
China. In general, it has been difficult for individual person or NGO organization to claim for
their environmental legal rights. On top of that, the awareness of soil contamination has not
been occurred among most of the municipal environmental authorities, nor among the potential
polluters. Currently most of the industries who pays attention to such an issue are the ones with
foreign mother group or background, in which case it might be mandatory for them to follow a
more strict policy for environmental, health and safety. However, when the legislation and the
corresponding enforcement is not in place, it is also quite flexible for these industries to imple-
menting these policies.
A gratifying news is that, due to the arising public awareness on soil and groundwater contami-
nation, more and more efforts are conducted in handling such an issue, started from the cases
involving long-history industries and widely-affected population. One good example is the Lan-
zhou oil chemistry group case happened in 2014. The case started from a serious benzene con-
tamination in drinking water in the major area of Lanzhou city, the capital city of Gansu province
with a population of 3.6 million, Northwest of China. After an urgent stoppage of water supply,
the investigation tracked the source of contamination back to the surrounding soil of the water
supply ditch, which is contaminated by the Lanzhou oil chemistry group, one of the biggest oil
industry in China. After one year, this group finally decided to invest more than 100 million RMB
to remediate the contaminated soil around the water supply network. The primary design has
been finished by the end of 2014. During the entire investigation and decision making process,
the voice from the polluter has continued to keep the public updated and involved, which has
6
been seen as a great improvement of Chinese industries handling such a responsibility .
Chinese EPA has also put constant efforts to enhance the enforcement of legislation. From the
1st of January, 2015, a new version of National Environmental Protection Law has been imple-
mented. With a clearer definition of liabilities and more serious legal consequence, this new
version is metaphorically called born with teeth, quoted from the Premier of P.R.C., Mr Li
Keqiang. Following this new law, a specific legislation for water recourse protection and water
pollution prevention is published in April, 2015, to give a strengthened force for protecting all
kinds of water resources, including groundwater. According to the announced work plan, it could
be expected that such a specific legislation for soil will be published in 2017. Consequentially, a
strengthened enforcement could be expected after the launch of this item.
5/9
5 Challenges and opportunities
Challenges within legislation set-up, administration, technical practice and finance have oc-
curred in Chinese market. It is obtained that, the pilot market in contaminated site management
start from those left from the re-location of old industries. Those sites are planned to be devel-
oped as new city area due to its potential real-estate value, hence theres an economic interest
to cover the remediation cost. Lessons learned from the conducted cases have drawn an over-
all picture of the existing challenges in following aspects, where the opportunities lies in as well:
6/9
However, these methods are debatable when it comes to a large site due to its capacity limits.
In some cases, methods like soil washing or soil vapor extraction are performed without know-
ing the local geology / hydrology sufficiently, which means not sufficient remediation and often
caused a cross-contamination problem with an extension of the source and/or plum and a sig-
nificant extra cost. Without a systematic approach for selecting the methods, some technique /
equipment suppliers intent to over illustrate the performance of their products, which also cre-
ates a certain confusion towards the project owner.
Although there are still challenges, more and more enterprises are still attracted by the tempting
market in China. The statistics shows that in 2011, the number of newly-registrated companies
with soil remediation as its main business was over 20, in 2012 this number became over 80. In
5
2013 this number became over 150 companies . However, it is also indicated that a major por-
tion of these companies are small-medium scale. The market demands for competent and ex-
perienced enterprises is still strong.
6 A pilot project
A pilot project which involved Danish expertise has been started. It was initiated from a diplo-
matic bond between the city of Wuxi, China and three Danish municipalities (Bycirklen), Jiangsu
province and the Capital region of Denmark. The target site is a former coking plant since
2
1950s associated with a steel plant next-door, and its area covers approx. 230,000 m . The in-
tention of this project is to introduce Danish knowledge and experience within contaminated site
management to the Wuxi municipality, in order to contribute to a sustainable approach to handle
such a problem, and eventually to encourage a sustainable urbanization.
Some pre-conditions have been set up to overcome the challenges mentioned in chapter 5, it
could be summarized as:
Both the local environmental authorities and the project owner have experienced unpro-
fessional performance previously, so a demand for a good practice is clearly expressed
from the beginning;
The cost of the project has been approved to be considered into the overall land develop-
ment cost, meaning it will be covered eventually by the increased land value;
Both the local environmental authority and its superior agree with the risk-based approach
in principal, making it possible to engage some advanced concept;
7/9
Technology exchange and transfer has been planned as one of the key purpose in this
project, including capacity building seminars, on-site demonstrations and day-to-day
communication.
Comparing with the mature practice in Denmark, some similarities in overall project process are
recognized: the entire process is divided in three phases namely preliminary investigation, de-
tailed investigation and risk assessment, and remediation. The decision making process and
practice described in these four draft guidelines (chapter 4) are considerably close to the Danish
concept, while some differences in detail are obtained during the practice.
One of the major differences is the requirement for knowledge regarding the site profile. The
importance of knowing the sub-surface is not valued enough yet, instead, most of the questions
from the authorities and the project owner focus on the contamination concentration and depth.
This also lead to a challenge when the consultant propose a systematic approach with a com-
prehensive conceptual model and risk assessment, as the Chinese side always jump to the
question for remediation methods and corresponding costs. Another challenge comes from the
different level of knowledge and experience. As the Chinese team have not experienced a
mega site like this, it is difficult to explain the benefits of applying some advanced concept e.g.
dynamic approach, etc.
The pilot project is still ongoing hence it is too early to conclude the outcome. Despite the chal-
lenges, the willingness of cooperation from all stakeholders are solid, and positive progress is
constantly obtained. As the result for the first phase, Danish practice has been well-accepted by
both the local authority and the project owner. This proved that with the right attitude and set-up,
it is possible to export and implement advanced knowhow to such an immature market as China.
It is expected from both sides, that this is just a beginning for a mutually beneficial collaboration
where a bridge between a knowledge pool and a great market potential could be well estab-
lished.
7 Conclusions
Facing the worlds largest urbanization nowadays, China is surely one of the markets with the
greatest potential in contaminated site management. The latest national mapping showed that
over 30% of former industrial sites have been confirmed as contaminated, while there are also
contaminated farm land, oil fields and mining sites. Relevant institutional set-up includes various
stakeholders and authorities, with Environmental Protection authorities acting as the dominating
party. Although China has been developing specific legislations for over 10 years, it is still chal-
lenging to initiate and implement a national law regarding contaminate site management. It is
also challenging regarding project financing, criteria definition, technique selection, and supervi-
sion & monitoring. A pilot project has been launched based on the diplomatic collaboration be-
tween Denmark and China, where the mentioned challenges were explored, and the possibility
of a positive result has been proven. China, as immature as it is now, should still be considered
as one of the most interesting market for the players in this industry.
8/9
References
2
1. Xie Jian, Li Fasheng, The Brown Field Remediation and Re-development in China,
World bank, Sep. 2010.
2. Gu Qingbao, Challenges and suggestions regarding contaminated site market in our
country, China ecological remediation net, March, 2015.
3. Hu Qing, Soil remediation faces a number of challenges, an urgent demand for national
standards. China ecological remediation net, March, 2014.
4. Contaminated site remediation: a market worth for exploring. Xinhua News, April 2015.
5. Opportunities and challenges in contaminated site remediation in China, China ecologi-
cal remediation net, April, 2014.
6. Lanzhou Oil Chemistry Group invest over 100 million for remedying the oil contamination,
Peoples Daily, March, 2015.
2
Brown field: the field with contamination risk which might cause a complexity in the land re-development, refer-
ring to http://www.epa.gov/brownfields/about.htm. In this report, brown field shall be understood as the site with
contamination potentials, while the contaminated site refer to those sites with confirmed contamination.
9/9
OPTIMIZING THE PROPERTIES OF NANOFLUIDS FOR THE EFFICIENT NAPL
REMEDIATION IN POROUS MEDIA
*Corresponding author: E-mail: ctsakir@iceht.forth.gr, Tel +30 2610 965212, Fax: +30 2610 965223
Abstract
A variety of aqueous suspensions (nanofluids) of zero-valent nano-particles (nZVI) are prepared by wet
chemistry techniques, their stability and longevity is evaluated by physic-chemical methods of
characterization, and their reactivity toward the dechlorination of per-chloro-ethylene (PCE) is examined
with tests in batch reactors. For assessing the mobility, longevity and reactivity of nZVI suspensions
(nanofluids), under flow-through conditions, visualization multiphase flow and transport experiments are
performed on glass-etched pore networks. The transient nZVI injection in porous media is simulated as
an advection- dispersion process by accounting for the nanoparticles attachment / detachment due to
the interactions of several mechanisms governing the deep bed filtration (e.g. Brownian diffusion,
interception, gravitational sedimentation, inertial impaction). The kinetic parameters of
attachment/detachment mechanisms are estimated with inverse modeling.
Keywords: porous media; zero valent iron; nanoparticles; numerical modeling; in situ remediation; dechlorination.
1. Introduction
Nanotechnology has special relevance to the in situ soil and groundwater remediation, due to the
combination of nanoparticles properties with fluid characteristics and thus the potential for injecting
nano-sized (reactive or adsorptive) particles into contaminated porous media (e.g. soils, sediments, and
aquifers). Among the various nano-materials explored for remediation (e.g. zeolites, metal oxides,
carbon nanotubes, etc) nanoscale zero-valent iron (nZVI) is currently the most widely used for the in
situ remediation of soils from a variety of toxic pollutants (e.g. chlorinated solvents) [1] due to its high
reactivity and its availability.
In spite of the great deal of work done for the in situ remediation of dissolved pollutants with nZVI
[2,3], little progress has been made toward delivery of nZVI to NAPL source zones. The dechlorination
rate is typically first order with respect to pollutant concentration, and hence the rate of DNAPL
dechlorination can also be limited by the availability of dissolved contaminant. This potentially limits the
prospects of rapid and substantial DNAPL mass depletion [4]. Nonetheless, the regions of maximum
dissolved contaminants that are most amenable to nZVI targeting are found in the DNAPL source zone
(trapped ganglia) as well as directly downgradient of the capture zone. One of the first ideas proposed
on NAPL targeting was the use of nZVI modified by an amphiphilic triblock copolymer [5]. Later on,
water-in-oil emulsions were synthesized with the nZVI being suspended in the water droplets [6]. More
recently, multifunctional nZVI-silica nanocomposites were found capable to adsorb to a DNAPL/water
interface [7]. Limited attention was paid on NAPL targeting by nZVI under flow-through conditions in
real porous media [8,9]. In macroscopic numerical models of nanoparticle transport in porous media, a
variety of solid-liquid mass-transfer mechanisms (e.g. particle attachment/ detachment/ blocking/
ripening, mechanical filtration, straining, etc) are commonly described as exchange terms [10,11].
The objective of the present study is to assess the longevity and mobility of nZVI-based nanofluids
under flow-through conditions that simulate the in situ injection of nZVI in polluted groundwater. nZVI
suspensions are prepared with wet chemistry techniques. For the stabilization of nanoparticles and
prevention of undesirable reactions (e.g. fast oxidation of nZVI), the nanoparticles (NPs) are coated by
a qualified selection of polymers and the coatings ensuring the longest stability are selected for reactivity
tests in batch reactors. Visualization studies of the transport of nZVI suspensions in a glass-etched pore
network are combined with a dynamic macroscopic model to quantify the mobility of nanoparticles.
2. Synthesis and characterization of nZVI suspensions (nanofluids)
Aqueous solutions of two polymers were prepared and tested as nZVI stabilizers: carboxyl-methyl-
cellulose (CMC) and CMC grafted to poly-dimethyl-acryl-amide (CMC-g-PDMAM). Aqueous
suspensions of nZVI were prepared by adding 4 solution under anoxic conditions in an aqueous
solution of 4 . 72 pre-grafted with the above-mentioned polymers acting as stabilizing coatings:
(1) CMC (MW CMC=90000 g/mol); (2) CMC-g-PDMAM (MW CMC-g-PDMAM=140000 g/mol). The protocol of
He and Zhao [12,13] was followed by keeping the w/w ratios / = 0.003, and of 4 / =
2. A 250 three-neck flask (Fig.1) was placed inside a cold bath at ~4 , and two sensors measuring
the pH, and redox potential, were inserted in the two necks. First, the flask was filled with 100 of the
polymer solution of concentration 10g/L, and 40 of deionized and degased water. In order to keep
anoxic conditions, a steady flow of nitrogen was injected in the solution during the entire procedure.
The solution was stirred steadily at ~800 rpm, and 10 of deionized and degased 4 . 72
solution of concentration 100 g/L were added in it. After 15 min, a yellowish or green solution evolved
depending on whether CMC or CMC-g-PDMAM was used. Afterwards, 50 of a fresh 4 solution
of concentration 5.4 g/L was added in the flask at flow rate 5 /. Gradually, the solution became
black indicating the generation of nZVI suspension. The stirring continued for 10 min to allow all bubbles
of evolved hydrogen to release. Then, the suspension was undergone vacuum filtration, degassed and
finally stored in amber glass vials which were filled with N 2 on top and closed tightly. The variation of
pH and Eh during the synthesis process is shown in Table 1. A series of tests confirmed the results of
[12] that the average size of nanoparticles decreases with the temperature decreasing, and this was
attributed to the very slow reaction rate at the lowest temperature tested (~4 ).
Table 1. pH and redox potential measurements during the production of polymer coated nZVI
The hydrodynamic diameter distribution and -potential of NPs were measured with Dynamic Light
Scattering (DLS) in a Zetasizer Nano-Instrument (Nano ZS, MALVERN) with dilution of the nZVI
concentration by several orders of magnitude. The characteristics of CMC-coated nZVI suspensions
prepared at concentrations 0.1 g/L, and 1.0 g/L are shown in Table 2 and Fig.2. Evidently, the CMC-
coated nZVI suspensions are characterized by a negative -potential, ensuring their stability, over a
broad range of pH values (Fig.2). It is worth mentioning that the actual Fe concentration of synthesized
suspensions, was measured by dissolving the iron in dense HNO3 and analysing the solution with flame
atomic absorption spectroscopy. It was found ~5 % less than its target value (0.1 and 1 g/L). The
transient evolution of the average size of CMC-coated nZVI of concentration 1.0 g/L is shown in Fig.3.
Evidently, the suspensions are stable for a couple of days, and progressively their agglomeration leads
to micro-particles (Fig.3). However, when a tightly sealed bottle of CMC-coated nZVI was opened after
18 days, the average size was almost identical to that measured on the day of production (Fig.3),
indicating the significance of storage conditions for nanoparticle stability.
0 10
0
-10
-10
-20
-potential (mV)
-potential (mV)
-20
-30
-30
-40
-40
-50 -50
(a) 0.1g/L CMC-coated nZVI
(b) 1.0 g/L CMC-coated nZVI
-60 -60
0 2 4 6 8 10 12 0 2 4 6 8 10 12
pH pH
Figure 2. -potential as a function of pH (a) 0.1 g/L nZVI, (b) 1.0 g/L nZVI.
500
400
300
Size (nm)
200
100
CMC-coated nZVI
0
0 5 10 15 20
Time (Days)
Figure 3. Transient variation of the average size of CMC-coated nZVI
A parametric study of nZVI synthesis conditions was done and the results are shown in Table 3.
The average particle size increases as the polymer to iron ratio goes to high or low values
indicating that there is an optimum ratio generating the smallest nanoparticles.
The average particle size increases profoundly as the sodium borohydride to iron ratio increases.
The molecular weight of polymers had no visible effect on particle sizes or -potential.
Increasing the temperature to ~18 , the average size of the polymer coated nZVI increases.
If a powder of CMC is used by skipping the lyophilization step, instantaneous sedimentation takes
place.
No sound improvement on the nanofluid stability was evident (Table 3) when using the CMC-g-PDMAM
coating. So, the CMC was selcetd as the nZVI coating in subsequent reactivity and mobility studies.
Its worth mentioning that the reported hydrodynamic particle size distributions (Fig.3, Tables 2, 3) are
intensity-based, and given that the scattering intensity is proportional to the 6 th power of particle size
whereas the particle volume is proportional to the 3rd power of particle size, the size distribution shifts
to lower sizes as one goes from intensity-based to volume-based and to number-based frequency
distribution functions. Subsequently, the volume-based and number-based average particle diameters
are expected to be lower than the intensity-based average particle which are used to quantify the
nanofluid characteristics.
-1
-1
10 10 -1
10
Cp/Cp0
Cp/Cp0
Cp/Cp0
1.0
(a) (b)
100
0.8
60
0.4
[Fe]inj,0=0.85 g/L 40
0.2
20
0.0 0
0 5 10 15 20 25 30 35 40 45 50 0 2 4 6 8 10 12 14 16 18
Number of pore volumes injected, PV Number of pore volumes injected, PV
Figure 7. (a) Dimensionless nZVI concentration in the effluent of the glass-etched pore network (breakthrough
curve). (b) Pressure drop measured across a central area of the pore network.
Any changes caused on the properties of CMC-coated nZVI nanofluid collected at the outlet of the
porous medium are shown in Table 4. The nZVI nanofluid injection for 3 hrs was followed by the flushing
of distrilled water for 2 hrs. In Table 4, sample 0 denotes the initial feed solution, sample 7 denotes
the residual solution in the syringes after 3 hrs, sample 6 denotes the residual solution in the glass
micro-model at the end of experiment, and all other numbers correspond to the effluent sampled every
1hr. The nZVI concentration refers to the time-averaged value over the sampling interval (Table 4).
5. Numerical simulation
We assume 1-dimensional fully developed flow of nanofluid (suspension of nZVI) injected at a constant
volumetric flow Q through a porous medium of length L and cross sectional area At. The transport of
nZVI is modelled as an advection - dispersion process by accounting for the deposition of particles on
pore-walls and their interactions with the suspension, namely
( m c) ( b s) (ud c) c
m DL (1)
t t x x x
where m is the porosity, c is the nanoparticle concentration in the nanofluid, s is the concentration of
nanoparticles deposited on the porous medium (defined as the mass of attached particles per unit mass
of the porous medium), b the initial bulk density of the porous medium, ud (=Q/At) is the Darcy flow
velocity, and DL is the longitudinal hydrodynamic dispersion coefficient. The nanoparticles deposit in
the soil grains can be regarded as the superposition of two deposits caused by (1) physicochemical
(ripening, blocking) and (2) physical (mechanical filtration, straining) mechanisms [17]. In this context,
the concentration of deposited particles, s, may be considered as the sum of two concentrations
associated with the aforementioned mechanisms 1 and 2 ( s s1 s2 ) [17], namely
( b s1 )
m ka ,1 1 1sb1 c b kd ,1s1 (2)
t
b2
( b s2 ) x
m ka ,2 1 c b kd ,2 s2 (3)
t d50
In Eq. (2) and (3), ka,i and kd,i are the kinetic constants of the particle attachment and detachment
(release) rates, respectively, 1 is an attachment amplification coefficient, and b1,b2 are exponential
parameters quantifying the non-linearity of deposition processes.
Due to the transient deposition of nanoparticles in the pore space, the properties of the porous
medium vary with the time according to the transient changes of porosity and permeability [10]. The
permeability is given by a Kozeny type [18] equation of the form
3
a
2
k k0 m 0 (4)
0 a
where 0 is the clean bed porosity, m is the time-dependent porosity, 0 is the clean bed specific area,
is the total specific area including the porous medium and deposited particles, k is the time- dependent
permeability, and k0 is the clean bed permeability. The pressure gradient across the porous medium is
given by Darcy law which is written as
P
ud (8)
x k
where P is the pressure and is the aqueous suspension viscosity.
By using the dimensionless variables: P*=P/P0, c*=c/c0, s*=s/smax, si*=s/smax (i=1,2), =x/L, =t/tcr,
k*=k/k0, and the characteristic parameters, tcr=L/ud, smax=s*0/b we get the following dimensionless
equations
c* 1 * s s1* s2* 1 c* 1 2c* 1 s* c*
c (9)
1 s* c0 0 1 s* Pe 2 1 s*
s1* c0 Daa ,1
s
1
1 s* 1 1 s* smax c* Dak ,1s1* (10)
2
s2* c0 Daa ,2 L *
s
1 s* 1 c Dak ,2 s2
*
(11)
d 50
P* L ud
(12)
P0 k
where the following dimensionless parameters appear: Pe, Dai,j, . The Peclet number, Pe, is defined
by the relationship
Lud
Pe (13)
DL
and represents the ratio of the characteristic time for NP dispersive transport of nanoparticles to the
corresponding one for convective transport. The Damkhler numbers, Dai,j (i=a,d j=1,2), are defined
through the general relationship
ki , j L
Dai a ,k , j 1,2 (14)
ud
and represent the ratio of the characteristic time of NP convective transport to the corresponding time
of NP attachment or detachment by mechanism 1 or 2. The dimensionless velocity defined by
L ud
(15)
P0 k0
depends on the initial permeability of the porous medium, and is handled as a constant parameter in
sensitivity analysis The aforementioned dimensionless equations coupled with the appropriate initial
and boundary conditions were solved by using ATHENA Visual Studio [19]. Sample results of sensitivity
analysis are shown in Figs.8-10, where the parameter values were obtained from the literature [20].
1.0
=0 0.016
=0.0
=0.216 0.015
=0.216
=0.431 0.014
0.8 =0.431
0.013
=0.647 =0.647
0.012
=0.863 =0.863
0.011
=1.0 =1.0
0.6 0.010
0.009
S/Smax
c/c0
0.008
0.007 (b)
0.4
(a) 0.006
0.005
0.004
0.2
0.003
0.002
0.001
0.0 0.000
0 20 40 60 80 100 120 140 160 0 20 40 60 80 100 120 140 160
Figure 8. Breakthrough curves of the concentration of (a) suspended and (b) deposited nanoparticles during the
injection (0-70 PV) and flushing (70-170 PV) of nZVI at various distances from the injection point.
The transient variation of the axial distribution of the nZVI concentration in suspension and deposit are
shown in Fig.8a,b. Gradually, the nZVI concentration in nanofluid increases during injection and
decreases during water flushing (Fig.8a). The concentration of deposit has the tendency to increase at
long distances from the injection point when the mechanical filtration, mechanism 2, becomes dominant
(Fig.8b).
1.0
Pe=0.225
Pe=0.5625
Pe=2.25
0.8
Pe=4.5
Pe=9
Pe=22.5
0.6
C/C0
0.4
0.2
0.0
0 20 40 60 80 100 120 140 160
Pore Volume
Figure 9. Transient response of the nZVI concentration as a function of Peclet number at the axial position
=0.863.
The effect of Pe on the nZVI transport is shown in Fig.9. At low Pe values, the porous medium is fully
saturated by nanoparticles very fast due to Brownian diffusion which is the dominant NP transport
mechanism. In contrast, for high Pe values, the NP convective transport prevails and eventually the
maximum of deposited NP concentration may occur at early times of water flushing (Fig.9).
0.5
0.25Daatt, 1 0.5
0.25Dad, 1
1Daatt, 1
Dad 1
0.4 2Daatt, 1
0.4 2Dad, 1
0.3
0.3
(a)
s/smax
s/smax
0.2
0.2
(b)
0.1
0.1
0.0
0 20 40 60 80 100 120 140 160 0.0
0 20 40 60 80 100 120 140 160
Figure 10. Effect of the mechanism 1, (a) attachment and (b) detachment Damkhler number on the transient
response of deposited nanoparticles at distance =0.647 from the inlet port.
As the NP attachment Damkhler number, Daa,1, increases, the deposition of nZVI is enhanced, the
percentage of nZVI reaching at the outlet of the porous medium decreases, and the total concentration
of deposited nZVI increases (Fig.10a). Respectively, the concentration of deposited nZVI increases
profoundly, as the NP detachment weakens and the corresponding Damkhler number decreases, and
approaches asymptotically a constant value for high values of this parameter (Fig.10b).
By introducing into the model the pore space properties of the glass-etched pore network used in
experiments (Table 5), the kinetic parameters of particle attachment and detachment rates were
estimated with inverse modeling of the experimental data of Fig.7, by using the ATHENA visual studio,
and the results are shown in Table 5 and Fig.11.
0.9
0.8
0.7
[Fe]/[Fe]inj,0
0.6
0.5
0.4
0.3 Experimental
Modeling
0.2
0.1
0.0
0 5 10 15 20 25 30 35 40 45 50 55
6. Conclusions
CMC-coated nZVI suspensions were prepared from 4 and CMC pre-grafted 4 . 72 solutions
by controlling carefully the conditions of the conventional wet-chemistry technique, and selecting the
protocol that ensures the highest nanoparticle stability, longevity, and reactivity. The stability and
longevity of suspensions was assessed by measuring physicochemical parameters of nanofluids, and
the most efficient coating was chosen. The reactivity of nanoparticles was measured with batch tests in
solutions of per-chloro-ethylene (PCE) in distilled water and found excellent. The mobility of CMC-
coated nZVI suspensions (nanofluids) was tested with flow-through experiments in a glass-etched pore
network. A relevant dynamic numerical model was developed, and used to estimate the kinetic
parameters quantifying the nZVI attachment / detachment rates on pore-walls. In the next step, the
nanofluid will be tested as an agent for the dechlorination of per-chloro-ethylene (PCE) source zone
under flow-through conditions.
Acknowledgements
The research is co-financed by the European Union (European Social Fund-ESF) and Greek national
funds in the context of the action EXCELLENCE II of the Operational Program Education and Lifelong
Learning (project: Optimizing the properties of nanofluids for the efficient in situ soil remediation-
SOILREM).
References
1. Cundy, A. B., Hopkinson, L., Whitby, R. L. D., 2008. Use of iron-based technologies in
contaminated land and groundwater remediation: A review, Sci. Total Environ., 400, 4251.
2. Tratnyek, P. G., Johnson, R. L., 2006. Nanotechnologies for environmental cleanup, Nano Today,
1, 4448.
3. Zhang, W. X., 2003. Nanoscale iron particles for environmental remediation: An overview, Journal
of Nanoparticle Research, 5, 323332.
4. Taghavy, A., Costanza, J., Pennell, K. D., Abriola, L. M., 2010. Effectiveness of nanoscale zero-
valent iron for treatment of a PCE-DNAPL source zone, J. Contam. Hydrol., 118, 128142.
5. Saleh, N., Sirk, K., Liu, Y., Phenrat, T., Dufour, B., Matyjaszewski, K., Tilton, R. D., Lowry, G. V.,
2007. Surface Modifications Enhance Nanoiron Transport and NAPL Targeting in Saturated
Porous Media, Environmental Engineering Science, 24, 4557.
6. Quinn, J., Geiger, C., Clausen, C., Brooks, K., Coon, C., OHara, S., Krug T., Major D., Yoon, W.
S., Gavaskar, A., Holdsworth, T., 2005. Field demonstration of DNAPL dehalogenation using
emulsified zero-valent iron, Environ. Sci. Technol., 39, 13091318.
7. Zhang, J., Sunkara, B., Le, L., John, V. T., He, J., Mcpherson, G. L., Piringer G., Lu Y., 2009.
Multifunctional colloidal particles for in situ remediation of chlorinated hydrocarbons, Environ. Sci.
Technol., 43, 86168621.
8. Phenrat, R., Fagerlund, T., Illangasekare, F., Lowry, T., Tilton, G.V, 2011. Polymer-Modified Fe0
Nanoparticles Target Entrapped NAPL in Two Dimensional Porous Media: Effect of Particle
Concentration, NAPL Saturation, and Injection Strategy, Environ. Sci. Technol., 45, 61026019.
9. Bishop, E. J., Fowler, D. E., Skluzacek, J. M., Seibel, E., Mallouk, T. E, 2010. Anionic
homopolymers efficiently target zerovalent iron particles to hydrophobic contaminants in sand
columns, Environ. Sci. Technol., 44, 90699074.
10. Tosco, T., Sethi, R., 2010. Transport of non-newtonian suspensions of highly concentrated micro-
and nanoscale iron particles in porous media: A modeling approach, Environ. Sci. Technol., 44,
90629068.
11. Grolimund, D., Borkovec, M., 2006. Release of colloidal particles in natural porous media by
monovalent and divalent cations, J. Contam. Hydrol., 87, 155175.
12. He, F., Zhao, D., 2007. Manipulating the size and dispersibility of zerovalent iron nanoparticles by
use of carboxymethyl cellulose stabilizers. Environmental Science and Technology, 41, 6216-6221.
13. He, F., Zhao, D., Paul, C., 2010. Field assessment of carboxymethyl cellulose stabilized iron
nanoparticles for in situ destruction of chlorinated solvents in source zones. Water Research, 44,
23602370.
14. Theodoropoulou, M. A., Karoutsos, V., Kaspiris, C., Tsakiroglou, C. D., 2003. A new visualization
technique for the study of solute dispersion in model porous media, J. Hydrol., 274, 176197.
15. Theodoropoulou, M. A., Sygouni, V., Karoutsos, V., Tsakiroglou, C. D., 2005. Relative permeability
and capillary pressure functions of porous media as related to the displacement growth pattern,
International Journal of Multiphase Flow, 31, 11551180.
16. Tosco, T., Gastone, F., Sethi, R., 2014. Guar gum solutions for improved delivery of iron particles
in porous media (Part 2): Iron transport tests and modeling in radial geometry, J. Contam. Hydrol.,
166, 3451.
17. Tosco, T., Petrangeli Papini M., Cruz Viggi, C., Sethi, R., 2014. Nanoscale zerovalent iron particles
for groundwater remediation: A review, J. Clean. Prod., 77, 1021.
18. Kozeny, J., 1927, ber kapillare Leitung des Wassers im Boden, Sitzungsber. Akad. Wiss. Wien,
136, 271-306.
19. M. C. Stewart, W. E., Caracotsios, M., 2008. Computer-Aided Modeling of Reactive Systems John
Wiley & Sons, Inc., Hoboken,, New Jersey.
20. He, F., Zhang, M., Qian, T., Zhao, D., 2009. Transport of carboxymethyl cellulose stabilized iron
nanoparticles in porous media: Column experiments and modeling, J. Colloid Interface Sci., 334,
96102.
ThS 1C.27 Thermal remediation 2
Introduction
In situ thermally enhanced remediation methods (steam-air injection and thermal wells), developed
and verified in several field applications by the Research Facility for Subsurface Remediation, VEGAS,
are proposed for source zone remediation of NAPL in the saturated and the unsaturated zone.
Contaminants can be effectively removed within several months if hydrogeological conditions are
suitable.
A novel application of steam-air injection was the removal of chlorinated hydrocarbons from a
fractured sandstone aquifer. In 2009 VEGAS successfully completed a pilot trial of steam-air injection
for the problem owner, the city of Villingen-Schwenningen in south-west Germany [6]. The test field
extended over 2,000 m of fractured rock including the upper of two different aquifers. The thermal
radius of steam propagation was 5 m in the target zone between 3 15 m bgs. The heating period by
steam-air injection lasted for 19 weeks. Contaminants that had penetrated into the sandstone matrix
were thermally desorbed during the conductive heating of the bedrock due to steam-air was flowing
through the fractures. More than 91 % of the total extracted mass (560 kg chlorinated hydrocarbons,
CHC) was removed from the groundwater fluctuation and unsaturated zones via the soil vapour
extraction system, less than 6% thereof (34 kg of CHC) via the groundwater containment. The CHC
values in the soil vapour and the groundwater were decreased by 95% and 85%, respectively.
Based on the results of the pilot, the steam-air enhanced remediation of the groundwater fluctuation
zone and of the unsaturated zone (2,900 m, 15 m thickness) for the entire site was designed. The site
was divided into nine treatment sections in size of 400 600 m, meaning 4,500 6,000 m of
fractured bedrock each. The duration of the steam-air injection phase (steam injection power of 400
kW) was calculated to last 33 months requiring 31 two-level injection wells and 34 soil vapour
extraction wells (SVE). The costs were estimated to 2.6 million EUR to treat ca. 43,000 m of
sandstone during four years of operation.
A data acquisition system and ADSL modem for the online monitoring of essential process parameters
(pressure, temperature, flow rates of steam, air, soil vapour and water etc.) is installed. Contaminant
concentrations in the extracted soil vapour and purified off-gas are analysed hourly using a process
gas-chromatograph.
To monitor the remediation process SVE samples from all the wells are analysed monthly.
Groundwater samples are analysed every 2nd month.
Implementation
The thermally enhanced remediation started in July 2012. It is currently in full operation and the steam
injection phase will last until January 2016. In the beginning of the remediation procedure (section 1
and section 2) the steam-air mixture was injected by means of four to five dual-screened wells and
300 500 kW to heat up the individual treatment sections. After passing the target temperatures and
obtaining a typical peak-shaped decrease in the contaminant mass removal (Figure 6) the next
remediation section was to be selected for steam-air injection. The steam-air injection in the first
section is stopped and the SVE will be maintained for at least one week. Then the next section will be
treated by steam-air injection and SVE.
During the treatment of section 2 an increase of contaminants in the soil vapour in the SVE wells of
section 1 which were during that time not in operation was observed. The observation of overpressure
indicated steam driven mobilization of CHC from section 2 in direction of section 1. Since heat storage
in the sandstone layers causes a slow cooling process ( T of 40 - 50 K takes 10 12 months) the
evaporation and desorption process of CHC present in the matrix might also continue.
In addition the steam and heat propagation was found to be wider by a factor of 1.5 2 as predicted
from the pilot study. In consequence the number of operating SVE wells was doubled to capture the
mobilized contaminants. During the following two months of operation the contaminant mass removal
increased steadily indicating a long desorption time of the impregnated CHC from the bedrock. The
process of desorption lasted for 5 months (see Figure 6, April - September 2013, section 2). In
consequence the remedial operation had to be adopted.
The wide heat propagation of 5 10 m in radius allows the simultaneous treatment and steaming of
two sections. Since the design steam and heat flux (less than 500 kW) was not changed this results in
an energy-effective doubling of the treatment time for each section (now 7 months) and thus sufficient
time for desorption of the contaminants. Overall this means that the time demand will be increased by
35%, the energy consumption by 25%.
In addition the SVE was revised and the extraction rate was increased by 50% to cover 4 5
treatment sections for pneumatic control. The heat propagation was found to extend over 3 sections
(see Figure 5) due to heat storage and heat conductivity.
In the beginning of April 2015 the steam-air mixture (600 kg/h) was injected via 16 injection wells and
approximately 600 m/h (800 900 kg/h) of hot soil vapour was extracted by means of 45 wells. In
total 13 m/h of groundwater was extracted and treated.
Table 1: Comparison of design and effective remediation strategy
Design of remediation process (2010) Applied remediation process (2013 2015)
9 sections 9 sections transient steaming of 2 sections
stepwise downgradient simultaneously
duration 3 months each section stepwise downgradient
2 injection levels, 5 wells duration 7 months each section
Heat propagation
The steam propagation in each section was observed and monitored during the first 6 8 weeks after
starting the injection by a continuously increase of the temperature. The steam front arrived at the
extraction wells four weeks after start-up.
The heating phase lasted 8 10 weeks (see Figure 4, section 1 and section 2) to reach constant
temperatures above the desired azeotropic temperature of 88C in the saturated zone. During the
treatment of section 1 and 2 the average temperature in the unsaturated zone above 11 m bgs was
below 65C. Hence, the azeotropic temperature was not met. Therefore the evaporation of the
contaminants was slower as indicated by the long lasting increase of mass removal during the
treatment of section 2 (see Figure 6).
Energy consumption
During the last 30 months of steam-air injection the power intake was 400 kW meaning an average
steam-air rate of 540 kg/h. In average 670 kg/h of steam-containing hot soil vapour with an average
temperature of 56C was extracted, meaning 540 kg/h of hot air.
The steam-air mixture provided the energy to heat the subsurface (Figure 5). In total 9,300 MWh of
energy were consumed. Energy losses in a dimension of 100 kW occurred due to the discharge of
warm groundwater (GW10 13, maximal 22C) and the extracted hot soil vapour after the steam
breakthrough. Thus, in total 5,200 MWh of energy, or more than 55% of the energy input, were direct
losses. The energy stored in the subsurface is estimated as 620 MWh for approx. 15,000 m of heated
soil (3 sections, area equipped with temperature sensors) with an average temperature of 80C in the
beginning of April 2015.
Contaminant removal
The remediation process starts by the steam propagation in the fractures and the direct heating of the
highly contaminated surface of the sandstone close to the fractures and its fast thermal desorption.
The first steam breakthrough at the extraction wells is typically accompanied by a high mass removal
rate of contaminants (Figure 6). The effect is mainly and mostly visible after 1 - 2 weeks of steam-air
injection in a new treatment section. During the first 8 10 weeks the bedrock is heated up while the
contaminant mass removal increases (Figure 7, section 2, redish area). Passing peak of contaminant
removal the desorption phase continues. Once the contaminant extraction went below 2 kg/d of CHC
removal, meaning 100 150 mg CHC per m of soil vapour the next treatment section was added to
the steamed area.
Figure 6: CHC removal rate, concentration in soil vapour and mass balance, mass flux of
steam, steam-air of single lines and soil vapour extraction, total and single lines
Until the end of March 2015 (simultaneous operation in sections 5-6-7) in total approximately 3,650 kg
of CHC were removed (Figure 6). Approximately 3,500 kg CHC were removed by SVE and 150 kg
CHC by the groundwater containment. Including the pilot trial in 2009, in total more than 4,000 kg
CHC were removed by the soil vapour extraction during the thermally enhanced remediation. The
mean mass removal rate of the current remediation is approximately 3.5 kg of CHC per day. During
the treatment of section 2 and section 6 the daily mass removal even exceeded 20 kg of CHC.
The concentration in the downstream plume range between 100 300 g/L captured by the pumping
wells GW10 13. This indicates still a CHC mass flux of 40 - 80 g/d. The goal of the remediation is to
decrease the maximum emission to less than 20 g CHC per day.
Literature
[1] GEOsens Ingenieurpartnerschaft (2007): 03.052a ehem. Stdt. Verbrennungsplatz Biswurm
Sanierungsuntersuchung Juni 2007, Bericht vom 10.10.2007
[2] Koschitzky, H.-P., Trtschler, O. (2009): Markierungsversuche zur Prfung der hydraulischen
Eignung hinsichtlich ISCO am Standort Biswurm in Villingen, Schwarzwald-Baar-Kreis,
Technischer Bericht Nr. 2009/07 (VEG 37)
[3] Koschitzky, H.-P., Trtschler, O., Lidola, B., Epp, M. und S. Schulze (2009): CKW-Schaden im
Kluftgestein. Neues Anwendungsfeld der thermischen In-situ-Sanierung mit Dampf-Luft-
Injektion. In: wlb, Wasser, Luft und Boden, Supplement TerraTech, 12/TT10, Vereinigte
Fachverlage GmbH, Mainz
[4] Koschitzky, H.-P., Trtschler, O. (2008): Thermische In-situ-Sanierungsverfahren:
Einsatzbereich, Dimensionierung und erfolgreiche Anwendung, Handbuch Altlastensanierung
und Flchenmanagement, 55. Aktualisierung, 3. Aufl., Mai 2008, Franzius, Altenbockum,
Gerhold (Hrsg.), C. F. Mller Verlagsgruppe, Mnchen, 5716, pp. 1 47, Grundwerk ISBN
978-3-8114-9700-9
[5] Ochs, S.O. (2006): Steam injection into saturated porous media process analysis including
experimental and numerical investigations, Mitteilungen Heft 159, Institut fr Wasserbau,
Universitt Stuttgart, ISBN 3-933761-63-8
[6] Trtschler, O., Koschitzky, H.-P., Wei, H., Lidola, B., Epp, M., Schulze, S. (2010). Removal
of CHC from fractured bedrock: results of a pilot application using steam-air injection at the
site Biswurm (Villingen-Schwenningen, Germany). In: F&U Confirm Publisher, Leipzig
Germany: Proceedings of ConSoil 2010, LeS ThS_A23, Salzburg, 22-24 September 2010,
ISBN 978-3-00-024598-5
[7] Hiester, U., Mller, M., Koschitzky, H.-P., Trtschler, O., Roland, U., Holzer, F., Edel, H.-G.
(2012). In situ thermal treatment (ISTT) for source zone remediation of soil and groundwater,
TASK Centre of Competence, Leipzig, Germany , www.ufz.de/task,
www.ufz.de/export/data/38/50022_ISTT_Guidelines_FINAL_PRINT.pdf
Session: 1c. Remediation technologies and approaches
1C.27 - Thermal remediation 2
COMPLEX BOUNDARY CONDITIONS FOR IN-SITU THERMAL TREATMENTS (ISTT) CONDUCTED DURING LAND
RECYCLING AND REMEDIATION BENEATH BUILDINGS
Keywords: Steam injection, thermal wells, SEE, TCH, CHC, DNAPL, SVE, pump and treat
INTRODUCTION
In-situ thermal treatments (ISTT) are applied successfully at several sites worldwide. The fields of
applications vary from low temperature enhanced microbiological degradation to high temperature
degradation of low volatile compounds (Figure 1). For volatile organic compounds (VOC) like
chlorinated hydrocarbons (CHC), benzene, toluene or xylene (BTEX), the co-boiling of water and non
aqueous liquid phase (NAPL), named steam distillation, is a process, dominating the progress during
source zone remediation. This process occurs mainly between 60 to 100C.
Figure 1: Application ranges of sub-surface heating for remediation and fields of application of
these guidelines [HIESTER ET AL. 2013 [1]]
In the recent past, more complex projects are conducted as well with treatments close to the
surrounding infrastructure like in inner city sites or next to or beneath buildings. Often, usage of
infrastructure and buildings must be enabled as well during the remediation works, e.g. for housing or
industrial production processes [HIESTER & SCHRENK 2008 [2]]. Furthermore, ISTT has been
successfully applied during land recycling projects. Combining in-situ remediation projects with
buildings under construction demands a professional project management to solve the complexity
caused by a high amount of individual processes, often interacting with each other.
For ISTT of CHC, the selection of the heating method is mainly dominated by the geological and
hydro-geological conditions in the contaminated layers. Steam injection and steam enhanced
extraction (SEE) is applied usually in sandy layers with a hydraulic permeability > 10-6 m/s, whereas
thermal conductive heating (TCH) with thermal wells enables an efficient heating in silty, loamy or
clayey formations, as well as in case layer thickness of several meters (Figure 2).
Figure 2: Scheme of thermal conductive heating (TCH) with thermal wells (THERIS method)
This article presents ISTT projects with several design restrictions caused by neighbours, structural
engineering or industrial production in the treatment areas during remediation. The remediation
management team (RMT) coordinated the interactions between remediation activities and third party
purposes. Difficult processes are named and hints for further projects are summarised. Furthermore,
the impact of more complex boundary conditions on the ISTT design, operation procedure and
measurement devices during the remediation are described.
Figure 3: Soil vapour concentration before remediation (left), after 2 months TCH (middle) and
after 3 months TCH (right)
Due to changes of site owners facilities, this existing building was demolished and replaced by a new
factory building (Figure 4). Requirements for the CHC source zone removal were a rapid, seriously
and efficient remediation concept. In-situ strategies were compared with conventional source zone
excavation and disposal as well as with drilling and replacement. Evaluating costs, time frame and
interface with the timeframe for superstructure implementation, in-situ thermal treatment (ISTT) was
evaluated as the most economical and efficient technology. To heat the low permeable loam, TCH by
using thermal wells was applied to vaporize the CHC (THERIS method).
Rainfall infiltration can cause cooling effects in an ISTT field. To reduce the amount of rainfall
infiltration into the treatment target zone (TTZ), the TCH field was covered by a construction foil
(Figure 4, right).
ISTT works were not allowed to interrupt foundation works or superstructure installation for the new
building. Thus, a close interface management between involved specialists for the installation of the
new building like pile foundation team, superstructure construction team and remediation specialists
was established to enable a failure-free operation of each lot.
Figure 4: Beginning of the CHC source zone remediation: Demolishing of former chemical
storage building (left), TCH remediation field (right)
Figure 5: Schematic intersection of the remediation field: TCH remediation (THERIS method)
during simultaneous foundation and superstructure works for a new factory building
Design revision for ISTT as well as for foundation and superstructure was undertaken by an
interdisciplinary team of experts. Parts of the design did not match with the location of installations and
the scheduled time frame of the remediation works. Thus, parts of the foundation and superstructure
including location of gates in the outer shell had to be redesigned. Moreover and because building
activities were conducted close to the remediation area, it was found, that foundation and super-
structure works may change boundary conditions for the ISTT field almost weekly (compare Figure 4,
Figure 6, Figure 7). To enable a better work flow the superstructure implementation, parts of the
remediation field were allowed to be used temporarily for equipment storage for superstructure works.
Figure 6: Midterm of the ISTT: Concrete works for building superstructure crossing the TCH
field (left), excavation for the basement close by TCH remediation
Two months after starting ISTT, about 3 m deep excavation for the basement close by the operated
ISTT was conducted (Figure 6). The top of the heater zone of some thermal wells in the ISTT field was
almost in a similar height (Figure 5). The SVE was continuously operated, and no CHC concentrations
were found in the vapour in the excavation pit. Pressure monitoring in the ISTT field detected a
negative pressure of approx. 150 mbar even at 8 m distance from recovery wells under operation.
To enable a detailed and rapid remediation management, remote controlled automated monitoring
systems for remediation process relevant parameter like subsurface temperature, pressure and
concentration and discharge in the SVE were established. For different reporting dates, contamination
in soil vapour based on single well measurements is illustrated in Figure 2. To enable a maximum of
transparency in the remediation process and progress, co-ordination measures about ongoing works
with other specialists as well as with authorities was conducted on a weekly base. ISTT was
completed within 4 months with no disturbance of the new factory building installation. The roof of the
new building was closed before removing the last SVE-container from the TCH field and the outer
shelter was closed immediately after removing the ISTT equipment. The former remediation area is
superstruct now by an aluminium die casting machine with a weight of approx. 100 tons.
Figure 7: End of ISTT: TCH field inside and outside the new building. Two weeks later, TCH
field was removed and outer shelter incl. roof of the new hall were closed.
To enable successful ISTT projects under complex conditions, the remediation management team
(RMT) has to involve specialists for non remediation related subsections to coordinate interactions.
However, ISTT during a land recycling project is only successful, if RMT and structural engineering
(like foundation and superstructure works) cooperate in time to avoid time delay in each subsection.
During the first drilling campaign, two additional SVE wells were drilled and installed (Figure 9,
middle). Steam injection and SVE pipelines were installed in a very sensitive environment, because
industrial production had to continue without disturbance during remediation works. Partwise, steam
injection wells and SVE wells were located close to machines operation areas and equipment storage
areas for the production (Figure 12, blue lines). Steam boiler and compressor for steam-air-injection
as well as SVE equipment were installed in containments outside of the building. Pipelines for steam
injection and SVE were suspended under the ceiling to avoid damages by the ongoing production.
Steam boiler (capacity 100 kg/h) and SVE were connected by programmable logic controller (PLC)
communication to avoid steam injection in case of a shutdown of the SVE. Moreover, pressure in the
injection line was continuously monitored by pressure sensors. In case of pressure losses in the
injection line, steam injection was automatically stopped.
25
temperature monitoring
6 steam extraction (SVE)
20 combined wells
steam injection
Number of wells
4 8 groundwater well
15
4
10 7
5 3
2 3
2
5
5 5 5
0
tender design (TD) TD modification due to additional extension of
contaminant the remediaton field
evaluation during (2. drilling campaign)
drilling works
Figure 9: Number of wells during tender design, modified tender and extended remediation
To monitor processes continuously and to enable a rapid remediation management, temperature, flow
and concentrations were measured automated by remote controlled monitoring and data sampling
systems. Data were evaluated on a daily base and reported to project partners on a weekly base.
Installation procedure, monitoring concept and safety devices were confirmed before start-up by
client`s health and safety officer, production manager, quality survey team as well as local authorities.
Figure 10: Drilling inside a building during continued industrial production (left), soil sample
(middle) and display of steam injection filter horizon (middle and right)
Steam and air were injected at first in the unsaturated zone in the ordinary detected TTZ, afterwards
even in the saturated zone. For the deeper steam injection wells, the filter screen was installed in a
loamy layer at the base on the saturated zone to inject steam at the base of the aquifer (Figure 10).
CHC were vaporised and recovered continuously by the SVE. During this first three months, 2.5 kg
CHC were recovered in average per week.
Former soil vapour evaluation wells were still present beneath the building. To observe the impacts of
steam injection to CHC indoor intrusion, additional indoor air samples were analysed continuously. Air
exchange between indoor and outdoor air could be described as high. However, CHC could be
detected permanently in the indoor air (Figure 12, first week). Based on literature values
[Grenzwertverordnung [3], TRGS 900 [4]], these concentrations were very low and can be evaluated
as uncritical for human health. The permanent presence of CHC in indoor air was a surprising issue,
because impacts from the remaining CHC from the former conventional SVE, starting in a depth of 5
m bgs, to indoor air were not expected before the steam injection. It could be demonstrated, that
steam-air-injection into the remaining source zone reduced CHC indoor air pollution immediately. The
measurement of some detected PCE concentrations in indoor air after 90 days steam-air-injection
could not be correlated to the remediation processes or activities from industrial production process.
Even CHC remediation was completed after four months in the tender designed TTZ (Figure 8), CHC
were still detected at the boundary of the remediation field. Thus, four additional wells were drilled. To
reduce costs, two of these wells were used either for steam injection or SVE (combination well,
Figure 9). Two temperature monitoring profiles were installed at the filter screen of two wells.
Figure 11: Cross section of the steam enhanced extraction (SEE) beneath a building
Indoor air pollution [mg/m]
PCE
TCE
10 1,1,1-TCA
CIS-DCE
0,1
0,01
0 30 60 90 120 150
Time [d]
Figure 12: Insulated steam injection pipe beneath machinery and storage area (left, blue line),
reduction of CHC indoor air pollution during steam injection
50 55 Temperatures [C]
45 50
40 45
Temperature in 9 m bgs
Temperature in 5 m bgs
35 40
180,00 180,25 180,50 180,75 181,00
Time [d]
Figure 13: Effect of intermitting air co-injection into a continued operated steam injection
SVE was operated to recover CHC in SA11 in SA12 and SA15. BL8 was operated to avoid
uncontrolled steam migration. Because the area around BL8 has been treated before, BL8 effected a
dilution of the contaminants from SA11, SA12 and SA15. The concentrations, presented in Figure 13
are the mixed concentrations of all four operated wells.
Well known from lab experiments and former field applications, air-co-injection prevents contaminant
condensation and accumulation at the steam front. Furthermore, the radius of influence of the injected
heat is increased, because less steam condenses close by the injection well. Even approx. 11 m
distanced from the injection wells, in three different depths, the rapid, repeatedly and pulsate cooling
of the recovered soil vapour as an effect of an interrupted air-co-injection could be observed. Due to
the dilution by the air-co-injection, simultaneous to increasing temperature at the SVE recovery well,
CHC concentrations in the SVE system drop down usually. For the overall mass balance, this effect is
minor but repeatable and dominated by PCE as the main contaminating CHC at this site.
Nevertheless, both effects cause layers of different permeability to be affected by this intermitting air-
co-injection, which is beneficial for the remediation process in stratified layers.
Figure 14: Temperatures during intermitting steam injection at the injection well and a
corresponding SVE well
At the injection well SA11, the interruption of the air-co-injection effects the decreasing temperature in
the injection depth at 5 and 7 m bgs. However, in the groundwater affected filter screen, the missing
air-flux caused a collapse of the steam front and a flooding of the bottom of the filter screen, illustrated
by temperatures below 100C in a depth of 9 m bgs. The cooling effect is reduced due to higher
groundwater temperature in the environment of the well after permanent steaming, but still present
after several hours. The cooling effect is tremendous by reaching temporarily less than 40C while
interrupting the injection for one hour.
At the recovery well SA12, a continuous cooling in a depth of 4 m and 6 m bgs could be observed.
This is caused by cooling of the silt by SVE. Steam had been injected into SA12 weeks before heating
the subsurface in that area actively. But, more interesting is the periodic increase of temperature in
SA12 in 8 m bgs, when injection in SA11 is interrupted. This indicates that during interruption of steam
injection, other preferential flow paths are addressed by the SVE. The rising groundwater level during
interrupted steam injection might have an impact as well. A detailed interpretation of this field data is
difficult, due to missing detailed groundwater level monitoring in the area of interest.
Figure 15: Average mass flow [kg/a] for different remediation methods
LITERATURE:
[1] HIESTER, MLLER, KOSCHIZKY, TRTSCHLER, ROLAND, HOLZER: Guidelines:
In situ thermal treatment (ISTT) for source zone remediation of soil and groundwater.
http://www.reconsite.com/fileadmin/dateien/Publikationen/ISTT_Guidelines_FINAL_PRINT.pdf
FUNDED BY FEDERAL MINISTRY OF EDUCATION AND RESEARCH (bmbf) and Helmholtz Centre for
Environmental Research (UFZ). May 2013.
[2] HIESTER, UWE & SCHRENK, VOLKER: Thermally enhanced in-situ remediation beneath buildings
during their continued useage - new source removal options for urban sites. ConSoil 2008.
Proceedings of the 10th International UFZ-Deltares/TNO-Conference on Soil-Water Systems in
cooperation with Provincia die Milano, 3.-6. June 2008, pp. 143-150, Stella Polare Congress
Centre - Fiera Milano, Italy
[3] Grenzwertverordnung 2011 (BGBl. II Nr. 429/2011)
[4] TRGS900, Begrndung zu Tetrachlorethen (Per), Ausschuss fr Gefahrstoffe, AGS-
Geschftsfhrung, BAuA, Ausgabe Januar 2012
Natural Attenuation in sediments as a tool for groundwater pollution control
Abstract
Contaminated groundwater plumes are prevented to further migrate through soil by natural seepage of
contaminated groundwater to surface waters. During this natural process, the contamination can be degraded
in the sediment and sludge at the bottom of the surface water. In The Netherlands this process (called Natural
Attenuation in Sediments, or NAS) is more and more accepted as a management option for contaminated
groundwater provided that biological degradation occurs and risks can be controlled. Research has been
performed in three cities in the Netherlands at four different locations contaminated with chlorinated ethenes
in order to gain insight in the possibilities and limitations of using NAS as a management option. Samples have
been taken at these locations from different sediment layers for chemical and molecular analyses. Biological
degradation processes seemed to prevail in mainly in the top layer of the sediment. Both reductive
dechlorination (anaerobic process) and micro-aerophilic degradation (aerobic process) capacity are present in
sediment and can contribute to removal of chlorinated ethenes from the environment. The thickness of the
layer in which biological degradation capacity is found is site specific and has implications for managing the
specific water bodies. Dredging will remove the active layer and will cause migration of the contamination to
surface water. Surface water (and its sediment) used to be regarded as a protected object (pro forma).
However, the interface from groundwater to surface water can have a valuable contribution to sustainable
remediation and control of contaminated groundwater.
1. Introduction
At various sites in the Netherlands further migration of contaminated groundwater plumes is prevented
by seepage of contaminated groundwater to surface waters. During this discharge in surface waters, the
redox conditions in which the contaminations are present undergoes a change from anaerobic to aerobic,
and the contamination passes a sediment layer (especially sludge) at the bottom of the surface water.
These processes are favourable for natural degradation of several contaminants like chlorinated ethenes
and BTEX.
This natural process is more and more accepted as a groundwater contamination management option due
to a change in remediation procedures in the Netherlands from site specific remediation to area oriented
remediation. We expect that Natural Attenuation in Sediments (NAS, see figure 1) will more often be
considered as a viable remediation option. Provided that the risks are manageable, NAS can be a valuable
contribution to area specific groundwater management and remediation plans.
In the cities of Amersfoort, The Hague and Haarlem (all in The Netherlands) the use of NAS to manage
groundwater contaminated with chlorinated ethenes at different sites is being considered. The main questions
that have arisen with these cities are 1) what are the possibilities and limitations of NAS at the site, 2) how can
NAS processes be monitored, 3) which parameters play a key role in NAS and 4) how can we manage our water
bodies taking NAS into account?
Research has been performed by Bioclear in 2011 and 2012 in order to close these knowledge gaps and to
answer the main questions posed by these cities with regard to NAS. In this article the results of this research
are presented and discussed.
The results of this research can be used within the scope of monitoring of remediation progress, to support the
development of area oriented groundwater management plans and in discussions with (local) authorities and
other stakeholders like Water boards.
What is NAS?
At a number of sites in the Netherlands contaminated groundwater already seeps through sediments to
surface waters. Provided that biodegradation of pollutants occurs (or can be expected) in the sediment and
that the surface water is not negatively influenced by seepage of contaminated groundwater this process will
be more and more accepted as an option to prevent further migration of groundwater plumes.
The most widely studied mechanism for biological degradation of chlorinated ethenes is reductive
dechlorination (figure 2). This is an anaerobic degradation process in which chlorinated ethenes are used
as electron acceptor and organic matter is used as electron donor (Bradley, 2003, Bradley and Chapelle,
2010). Perchloroethene (PCE) can be reductively degraded under iron reducing to methanogenic
conditions. Successively Trichloroethene (TCE), Dichloroethene (DCE) and Vinyl chloride (VC) are formed
as degradation products in this process. If complete degradation of PCE occurs, even tually the harmless
products ethene and ethane are formed. For a complete anaerobic degradation of chlorinated ethenes (to
ethene and ethane), the following preconditions need to be met:
Presence of strongly reduced, methanogenic, conditions characterized by the absence of
oxygen, nitrate and sulphate.
Presence of sufficient amounts of organic carbon (TOC).
Presence of dechlorinating bacteria Dehalococcoides spp.
If the conditions are not favourable, for instance because the redox conditions are not methanogenic but
iron- or sulphate reducing, partial degradation of PCE and TCE to DCE and VC occurs, resulting in the
accumulation of these daughter products.
Besides reductive dechlorination also another biological degradation process can occur in the presence of
(trace amounts of) oxygen. Surface water and probably also the upper layer of the sediment in water
bodies are aerobic. Under these conditions further degradation of DCE and VC can occur. Even under
anaerobic conditions, trace amounts of oxygen can be present and micro-aerophilic degradation of DCE
and VC to carbon dioxide and water can occur (figure 3). Up till now, one bacterium is shown to be able to
use DCE as a carbon source, Polaromonas sp JS666. The genes involved in VC degradation under micro-
aerophilic conditions are etnE and etnC (Schmidt and Thiem, 2008, Mattes et al., 2010).
Due to high organic matter contents in sludge, mostly methanogenic conditions are present, which are
favourable for biodegradation of chlorinated ethenes. As a result, chlorinated ethenes can be degraded by
reductive dechlorination in the sludge layer before they reach the surface water. In the upper sediment layer
(sludge layer) also aerobic to micro-aerophilic conditions may prevail. Under these conditions, degradation of
DCE and VC can occur.
Research conducted by Hamonts et al. has shown that dechlorinating micro -organisms are present in
contaminated sediments and biological dechlorination in sediments can occur due to the presence of
specific bacteria (Hamonts et al., 2009). Besides anaerobic degradation, aerobic degradation of DCE and VC
could theoretically take place in the upper sediment layers, the sludge. This means that sediment and sludge
layers from ditches, creeks, canals and rivers can play an active role in removing chlorinated ethenes from the
environment. Combined with an active pollution source removal, this can lead to an effective and efficient
approach of the widespread chlorinated ethene pollution.
This research has been performed on four different sites in the cities of Amersfoort, The Hague and Haarlem.
At all sites the groundwater is contaminated with chlorinated ethenes. In these cities contaminated
groundwater seeps to surface water. It is expected that natural attenuation processes occur in the sediment. In
order to gain insight in which layer the capacity for biodegradation is present, different sediment layers have
been sampled at the sites. Samples have been taken from within the flow path of the contaminated plume and
on the edges of the plume. In the sediment samples both chemical analyses and molecular analyses have been
performed. The first in order to retrieve information about contamination of the sediment layers, the latter in
order to retrieve information about biological processes in the sediment layers. Therefore analysis on micro-
organisms and enzymes known to be involved in biological degradation of chlorinated ethenes (see
introduction: What is NAS) have been performed
Site A
The site has been contaminated with chlorinated ethenes due to industrial activities in the past such as
graphic production, production of touch screens and membrane swit ches. The sandy soil at the site is
contaminated (mainly with TCE and its degradation products DCE and VC, in concentrations of maximum
4,000 g/l, 13,000 g/l and 65 g/l respectively) up to a depth of 5 m-bgl (meters below ground level).
The site is situated at a river which has a draining effect on the shallow groundwater. Reductive
dechlorination has already occurred in the groundwater at this site, showed by the presence of
degradation products DCE and VC.
Site B
East of a town canal various industries were located in the past which have caused soil and groundwater
pollution with chlorinated ethenes (PCE, TCE, DCE and VC in maximum concentrations of 27,000, 4,800,
3
6,800 and 14 g/l respectively in the source zone). In the source zone a total of 56,000 m of soil volume
is contaminated. The plume is larger but the total volume of the plume is not known. The pollution is
present in the first aquifer up to 11.5 m-bgl. The soil in this part of the city is sandy with local thin peat
layers. The conditions in the groundwater are sulphate reducing to methanogenic. The pollution migrates
towards the town canal. West of this canal no pollution has been found. The pollution probabl y seeps
through the sediment of the canal into the surface water.
2.2.3. Haarlem
South of a canal in Haarlem, two contaminated sites are located of which the contaminated plumes have
already merged. The pollution is present between 4 and 13 m-bgl. The soil is sandy and the conditions in
the groundwater are iron- to sulphate reducing. The groundwater flow direction is towards the canal
which has a draining effect. At the site mainly DCE is found in concentrations of maximum 24,000 g/l in
3
the source zone. A total of 200,000 m of soil volume (source and plume) is contaminated at this site.
2.3. Sampling
At all sites undisturbed sediment cores have been taken halfway the canal by means of transparent tubes
(1 m, diameter 3.6 cm) in a Beeker sampler by MUG Ingenieursbureau. Because of the transparent tubes
the sediment layers were visible. Immediately after sampling photos were taken of the tubes and the
thickness of the layers was determined. Samples were taken from each individual sediment layer by
cutting the sediment cores directly in the field and applying the appropriate conservation method for
each type of analysis. Table 1 shows the amount of tubes per site.
At the site Amersfoort B samples were also taken of the surface water just above the sediment layer.
2.4. Analyses
All sediment layers have been individually processed. At the sites in Amersfoort and Haarlem the top
centimetres were sampled separately for molecular analyses, even if it was not an individual layer. This
was done because it was assumed that oxygen would only penetrate within the first centimetres of the
sediment and this might change the biological processes in this part of the sediment compared to dee per
layers.
All individual samples, except for those from the first centimetres of the sediment, have been analysed for
the following parameters:
Dry matter (protocol NEN 12880) and organic matter content (protocol NEN 5754)
Concentration PCE, TCE, CIS, VC, ethene, ethane and methane (GC-MS, own protocol)
Concentration nitrate (protocol ISO/DIS 14256-1) and sulphate (own protocol)
Based on the results of the chemical analyses samples were selected for DNA extractions (own protocol).
The DNA extracts were subsequently analysed by means of q-PRC analyses (own protocol). The following
parameters have been analysed:
Total Eubacteria
Total Archaea
Dehalococcoides spp. (DHC), VC reductase (vcrA) as measure for reductive dechlorination
etnC (Alkene monooxygenase), etnE (Epoxyalkane coenzyme m transferase) and Polaromonas
JS666 as measure for micro-aerophilic degradation
McrA (measure for methane forming activity)
In this project large amounts of data were generated because many samples were collected and many
analyses per sample were performed. In order to find statistical significant relations per site and between
sites the statistical program JMP was used.
3.1. Amersfoort A
No contamination was found above the detection limits. At the site Amersfoort A no correlation was
found between the sediment depth and the organic matter content.
Statistical analysis showed that the organic matter content is correlated to the dry matter content but not
to sediment depth. DHC and vcrA (vinyl chloride reductase) are positively correlated. The micro-aerophilic
degradation is negatively correlated to sediment depth. Figure 4 shows the number of dechlorinating
bacteria and the number of micro-aerophilic degraders in the sediment layers. The number of micro-
aerophilic degraders slightly decreases with increasing depth. The number of dechlorinating bacteria
slightly increases with increasing depth. The reductive dechlorination capacity is higher in the deeper
sediment layers, in contrast with the findings in The Hague.
It is shown that the micro-aerophilic degradation capacity is most abundant in the top layer of the
sediment whereas the reduction dechlorination capacity is present in the entire sediment zone and is
higher in the deeper sediment layers. The dechlorinating capacity is highest in the sample taken in the
flow path of the contaminated groundwater plume.
3.2. Amersfoort B
The pollution is only found in the top layer of the sediment, mainly at the north side of the location, in the
flow path of the highly contaminated groundwater plume.
In sediment at site Amersfoort B the organic matter content decreases with incre asing depth. The organic
matter content and the dry matter content are negatively correlated. With increasing depth, the dry
matter content increases and the organic matter content decreases. Again DHC and vcrA (vinyl chloride
reductase) are positively correlated, as seen at the site Amersfoort A. With an increasing dry matter
content (so at increasing depth) the micro-aerophilic degradation capacity decreases. With increasing
depth the reductive dechlorination capacity decreases in the northern (more contaminated) sample. That
means that the reductive dechlorination capacity in the northern sediment is highest in the top layer. In
the southern (less contaminated) sample, we did not find a correlation between degradation capacity and
depth. At increasing concentrations of pollutant, the number of DHC and therefore the degradation rate
also increases.
In the groundwater near the canal the concentration of cis-DCE is 140 g/l and PCE is 730 g/l. In the
surface water near the northern sample, we have found low concentrations of pollution, cis-DCE (0.63
g/l) and PCE (1.6 g/l). The concentration in surface water is below the ecological guideline values from
the model OMEGA. No ecological risks are expected from this concentration. The concentration in the
surface water will further decrease by means of dispersion and aerobic degradation of cis-DCE.
3.4. Haarlem
Like at the sites in The Hague and Amersfoort B, a significant relation was found between depth, dry
matter and organic matter. With increasing depth the dry matter contents increases and the organic
matter contents decreases. Also we have found a significant correlation between methanogenic activity
and depth; with increasing depth the methanogenic activity decreases. And with increasing organic matter
contents the micro-aerophilic degradation capacity as well as the reductive dechlorination capacity
increase as well. High organic matter contents are found in the top layer of the sediment. The reductive
dechlorination capacity is found in the thin upper layer of the sediment (0-0.01 m). In this layer also
micro-aerophilic degradation can occur.
The maximum mass entering the canal has been calculated as follows:
Min = ((P*(1/2W))*v*p) * Cmax
Min (cis-DCE) = 0.44 kg/year cis-DCE;
Min (PER) = 0.005 kg/year VC.
The average mass that daily enters the canal is 1,208 mg cis-DCE and 13.8 mg VC. Based on the k-value
and this concentration the degradation rate has been calculated as k*concentration. Based on the highest k-
value that implies a degradation rate of 21.5 g/L/d for cis-DCE and 0.24 g/L/d for VC.
The contaminants are degraded in the aqueous phase of the sediment. The total amount of water in the
sediment has been calculated as P*1/2W*t*p. That means the volume is 4,500 L.
Based on the Total volume of the aqueous phase and the degradation rate the daily degraded mass can be
determined as degradation rate * volume, being 94.5 mg/d cis-DCE and 1.08 mg/d VC.
That means that the active sediment layer of 1.5 cm thick can degrade 8 % of the incoming mass. With a
factor 10 lower k-value the degradation efficiency decreases to 0.8 %. The efficiency will increase with
increasing thickness of the active sediment layer. A 10 cm increase in active sediment layer will increase the
efficiency to 52 %. Important to realize is that the site has been dredged prior to our research and the
sediment layer can typically reach a thickness of 70 to 80 cm.
The surface water just above the sediment layer contained a maximum of 5.6 g/l cis-DCE and 0.15 g/l VC.
That would mean that 98.7 % of the incoming cis-DCE pollution is degraded and 96.9 % of the VC pollution.
That means that the calculated degradation efficiency is a worst case scenario and the k-values might be
higher than used in the calculations.
3.5. Overall
At three out of four sites the organic matter content is negatively correlated with sediment depth. In
general, the biodegradation capacity is correlated to the organic matter content meaning that the
biodegradation capacity, both reductive dechlorination and micro-aerophilic degradation, is mainly
present in the upper layer of the sediment. The thickness of the biological active layer varies per location.
Some of the sites, like Haarlem, only have a very thin biological active layer. Dredging of the canal would
result in a removal of the biological degradation capacity, at least tempora ry. At the location Amersfoort
A dredging would not have such a drastic effect as the active layer is much thicker.
At most sites, the concentration of the contamination in the sediment is low. This has implications for the
calculated degradation rate, which is low at most sites except for Amersfoort B, which contains higher
concentrations of contamination in the sediment. As the degradation rate is concentration dependent the
degradation rate will increase with increasing concentrations of contaminants, if degradation capacity is
present.
4. Conclusions
This research has shown that the capacity for biological degradation is present in the upper, organic rich layers
of sediments from surface waters. Degradation rates within these layers seem to be sufficient to prevent
(ecotoxicological) risks due to seepage of contaminated groundwater into surface waters. Molecular tools are
indispensable for the determination of these processes and their degradation capacity.
Instead of regarding surface water as a protected object (pro forma) the interface from groundwater to surface
water can actually be used in sustainable remediation and control of contaminated groundwater. Provided that
the risks of this approach can be controlled, natural attenuation processes in sediments can provide a valuable
contribution to area specific groundwater management and remediation plans.
The next step in our research is to be able to actually measure the amount of contaminants that is
removed by biological processes in the sediment based on the actual degradation rate, the flux of
contamination by seeping groundwater and the thickness of the active layer. This will enable us to
calculate the amount of contaminants that may end up in the surface water and assess the risks
associated with this. Based on this it is possible to elaborate and guide a remediation option based on
NAS, resulting in cheaper remediation of the polluted groundwater.
Acknowledgements
The authors would like to thank the municipalities of Amersfoort, Haarlem and The Hague for enabling us to
perform this research.
Literature
Abe Y, Aravena R, Zopfi J, Parker B, Hunkeler D., Evaluating the fate of chlorinated ethenes in streambed
sediments by combining stable isotope, geochemical and microbial methods, Journal of Contaminant
Hydrology, 26;107(1-2), 2009.
Bradley, P.M., and F.H. Chapelle, Biodegradation of Chlorinated Ethenes. Chapter 3 in H.F. Stroo and C.H. Ward,
Eds. In Situ Remediation of Chlorinated Solvent Plumes. Springer. New York. p. 39-67, 2010.
Bradley P.M., History and Ecology of Chloroethene Biodegradation: A Review, Bioremediation Journal,
7(2):81109, 2003.
Ellis P.A., Rivett M.O., Assessing the impact of VOC-contaminated groundwater on surface water at the city
scale, Journal of Contaminant Hydrology, 91(1-2):107-27, 2007.
Hamonts K, Kuhn T, Maesen M, Bronders J, Lookman R, Kalka H, Diels L, Meckenstock RU, Springael D,
Dejonghe W., Factors determining the attenuation of chlorinated aliphatic hydrocarbons in eutrohic river
sediment impacted by discharging contaminated groundwater. Environmental Science & Technology,
43(14):5270-5, 2009.
Hamonts K, Kuhn T, Vos J, Maesen M, Kalka H, Smidt H, Springael D, Meckenstock RU, Dejonghe W., Temporal
variations in natural attenuation of chlorinated aliphatic hydrocarbons in eutrophic river sediments impacted
by a contaminated groundwater plume. Water Research 46(6):1873-88, 2012.
Imfeld G, Nijenhuis I, Nikolausz M, Zeiger S, Paschke H, Drangmeister J, Grossmann J, Richnow HH, Weber S.,
Assessment of in situ degradation of chlorinated ethenes and bacterial community structure in a complex
contaminated groundwater system, Water Research, 42(4-5):871-82, 2008.
Mattes, T.E., Alexander, A.K and Coleman, N.V. Aerobic biodegradation of the chloroethenes: pathways,
enzymes, ecology and evolution, FEMS Microbiology reviews 34 445-475, 2010
Zaan van der B, de Weert J, Rijnaarts H, de Vos WM, Smidt H, Gerritse J.,Degradation of 1,2-dichloroethane by
microbial communities from river sediment at various redox conditions, Water Research, 43(13):3207-16, 2009
Omega version 6.0, Optimal modeling in ecotoxicological assessment, RIZA, Lelystad, The Netherlands,
2006
THERMAL DESORPTION TECHNOLOGY: TWO CASE STUDIES
Abstract
Thermal desorption is a physical separation process consisting in heating polluted soils to remove
contaminants. Process effectiveness is strictly related to applied temperature and soil characteristics.
Laboratory studies on thermal desorption of soils coming from two different industrial sites located in the
North of Italy are described: i) the former one contaminated by DDX, i.e. DDT and its derivatives (DDE,
DDD); ii) and the second one contaminated by hydrocarbons and mercury.
For each soil, the experimental approach was: i) thermogravimetric tests in order to identify optimal
temperature range; ii) desorption treatments in static kiln in air flow for the optimization of operating
conditions.
In case of soil contaminated by DDT, two temperature ranges corresponding to two consecutive weigh
losses were identified: 200-300C and 480-600C. According to that, thermal desorption was performed
in the middle of the two ranges (250 and 550C) and slightly above (350 and 650C) for 1 and 2 hours.
An appreciable reduction in concentration of DDX after a treatment of 350C for one hour was noticed
(from 1600 to 1.9 mg/Kg). The same results were noticed by enhancing treatment temperature and time.
For soil contaminated by hydrocarbons and mercury, two temperatures were identified: 300 and 450C.
Considering a sample with starting values of 8 mg/kg of mercury and 10000 mg/kg of total hydrocarbons,
300C for 1 hour was enough for reducing the concentration under the limits set by Italian law.
According to the obtained results, thermal desorption treatment can be an effective remediation
technology for soils contaminated by different kind of contaminants.
1. Introduction
Pollutants such as oil and other chemical substances may contaminate large quantities of soil via spills,
leakage or other incidents.
For decontamination of these polluted soils, thermal treatments are commonly considered effective
because of their advantages of reliability and efficiency on a high variety of polluting compounds.
Thermal treatment has shown to be effective for removing contaminants with a high vapor pressure
such as organochlorides and mercury [1,2]. In some cases, depending on the contaminants, it can
promote also the transformation of the contaminants via chemical reactions, triggered by temperature,
with the decomposition of heavy compounds into their volatile fragments [3].
Generally, desorption temperature of organic contaminants is closely related to their boiling
temperatures, i.e. the increase of the first corresponds to the increase of the second; a complete
decontamination of the soil is usually reached at temperatures higher than the boiling point due to
interactions between contaminant and soil [4].
One of the contaminants considered in this study is mercury, a heavy metal defined as "toxic" because
its unfavorable biological effect on humans and livestock. Different kinds of treatment can be used for
the remediation of mercury contaminated soils such as soil-washing, stabilization/solidification, in situ
vitrification [5], but thermal desorption is considered to be the best intervention technology compared to
the former ones [1,6].
Thermal desorption of mercury contaminated soils is often reported [2] as an effective technology. Of
fundamental importance is the knowledge of mercury speciation present in soil: Hg (0) and Hg (I)
vaporize between 150 and 200C, while Hg(II) desorption temperature must reach 300C [7]. Huang
and coworkers [8] have shown that the presence of other metals do not affect thermodesorption of
mercury.
Another class of contaminants discussed in this study is DDX. In a not recent work, Geoffrey D. Silcox
and collaborators [9] use a rotary kiln, with excess of oxygen, for treatment of about 20000 ppm DDX
contaminated soil, by setting a temperature of 700C and withdrawing aliquots of the sample at different
times. It has been observed that after 10 minutes there has been a DDX reduction due to the well-known
quick conversion of DDT and DDD at high temperature, the only identified byproduct being DDE. The
optimum conditions under which the concentration of DDE reaches zero have been found to be 700C
for 30 min. Also in the work of Zheng Peng temperatures for the removal of DDT from a contaminated
soil are very high (950-1050C), and once again, the residue is composed only by DDE [10].
Bordas and associates in 1953 [11] correlated the inactivation of DDT with the amount of iron oxides in
the soil. Zhengping Hao and Tian have used mesoporous silicas functionalized with iron oxides for the
removal of DDT from aqueous solutions; the heat treatment of the adsorbent containing DDT was 450C
in air: temperature well below the temperature commonly used in the traditional technique of incineration
[12]. With a quantity of iron oxide of 4% by weight with respect to the solid support, at the end of the
heat treatment conducted at 350C for 2 hours, DDT resulted totally decomposed [13].
In the present paper two laboratory studies on thermal desorption of soil samples coming from two
different sites in the North of Italy are described: i) soil contaminated by Hg and DDX (DDT, DDE, DDD);
ii) soil contaminated by hydrocarbons and mercury. Our attention was focused on the research of the
experimental conditions for reaching the reclamation of soils according to the limits imposed by the
Italian regulations [14].
2. Experimental
The experimental approach was the same for all the soil samples of this work.
The first step was the collection of soils from the site of interest and their homogenization. Then
thermogravimetric measurements (TG) by means of Netzsch STA 449 F1 Jupiter apparatus were carried
out for the identification of temperatures for successive testing in static kiln. Finally, thermodesorption
tests were performed with the purpose of evaluating the effect of different temperatures and different
treatment times on the volatilization/degradation of contaminants. Tests were performed using the static
tube furnace LFT-TUBE-FORNACES-1200C by Lenton Thermal Designs Limited (Figure 1). A
resistance wire wound onto the quartz tube heats the oven, whose maximum temperature is 1200C.
The regulation of temperature is controlled by a thermocouple protected by a robust coating. For each
test, about 50 grams of soil were placed in a quartz pan then inserted inside the oven. Generally,
treatment temperatures were centered on the optimal ones identified in the TG tests.
Soil samples were analyzed before and after thermodesorption tests, measuring concentrations of the
contaminants of interest.
Contaminated soil used in this work were collected in two different areas of the north of Italy.
The first one is contaminated by Hg, DDT and its derivatives (DDE, DDD). Samples were collected
excavating with an excavator and homogenized in situ, then stored into polyethylene bags and
transferred in laboratory where they were stored at 4C. For laboratory studies samples were air-dried
and sieved in three fractions (< 2mm 0.063 mm < d < 2 mm, <0.063 mm). Fraction with diameter under
2mm was the 97.26% of the total, as reported in table 1. Consequently, tests were run on the two
fractions of main interest: <2mm and <0.063mm, since in most cases the finest particles retain
contaminants more strongly than the coarsest ones. Soil sample selected for test was named BH56.
For this soil, TG measurements were performed in a stream of air (50 ml/min) to simulate the real
situation, using a temperature ramp of 40C to 900C at 3C/min and 1 hour at 900C.
Figure 1: Tubular static kiln
The second kind of soil samples comes from a site contaminated by Hg, BTEX, C>12 and C<12.
Samples were collected and managed as previously described. In this case, soil was in wet form, without
air-dried. Moisture was measured following the procedure indicated in the Italian Ministerial Decree of
13/09/1999 [15].
TG measurements were made according to a ramp from 40 to 900C at 10C/min, using 50 ml/min of
air as carrier gas. Sample selected for test was named D43.
Figure 2 shows the comparison of two repeated measurements performed on sample BH56: a good
reproducibility of the samples treated in air was observed, with two main weight loss identifiable between
200-300C and 480-600C. According to this behavior, temperatures for the desorption in the tubular
static furnace were selected: 2 temperatures in the interval corresponding to the main weight loss,
respectively 250 and 350C, and 2 temperatures just above, respectively 550 and 650C.
Figure 2: TG measurements on sample BH56
For each temperature previously identified different heating times were tested, Table 2 shows all the
performed tests and the corresponding weight losses. As expected, the weight loss increases with
increasing temperature and the heating time.
Test % lost in
Sample Diameter
conditions weight
BH 56 <2 mm 250C, 1 h 0.61
BH 56 <2 mm 250C, 2 h 1.04
BH 56 <2 mm 250C, 3 h 1.14
BH 56 <2 mm 250C, 4 h 1.11
BH 56 <2 mm 350C, 1 h 1.33
BH 56 <2 mm 350C, 2 h 1.47
BH 56 <2 mm 350C, 3 h 1.54
BH 56 <2 mm 550C, 1 h 1.78
BH 56 <2 mm 650C, 1 h 2.81
BH 56 <0.063 mm 250C, 1 h 1.76
BH 56 <0.063 mm 250C, 2 h 2.13
BH 56 <0.063 mm 350C, 1 h 2.15
BH 56 <0.063 mm 350C, 2 h 2.15
BH 56 <0.063 mm 550C, 1 h 2.53
BH 56 <0.063 mm 650C, 1 h 3.09
Trends of DDT and its derivatives after thermal treatment are shown in figure 5: data are reported as
summation of DDX where DDE represents at least 98% of the total, in our cases. Starting values
(respectively 2128 mg/kg for the d<0.063mm and 1472 mg/kg for the d<2mm) are not shown for the
sake of clarity. Since the standard concentration limit reported in the Italian Legislative Decree of
03/04/2006 NO.152 [15] corresponds to 0.1 mg/kg, a treatment at 350C for 2 hours is required to reach
this limit. Such temperature is quite low if compared with those cited in literature, this may be attributed
to the presence of iron in an amount equal to 3% by weight; value obtained from XRF measurements
performed on the sample as received. This hypothesis is in agreement with the literature discussed in
the introduction relatively to the catalytic activity of iron in the DDT destruction [13].
From thermogravimetric tests (figure 6) two temperature ranges were identified, corresponding to the
observed greatest losses in weight, respectively between 100 and 300C, where water and light
hydrocarbons are desorbed, and between 300 and 450C where loss in weight is due to heavier
hydrocarbons. The total weight loss is 75%, which subtracted the 41% of relative humidity gives a value
of 34% of which 12% at 300C and 24% at 450C. Consequently, tests in the tubular oven were
performed at 300 and 450C as specified in table 3.
Figure 6. TG measurement on sample D43
An appreciable difference in weight loss between the two temperatures is observed in Table
3, while there are no significant variations increasing the heating time. For this reason, the
analytical measurements were performed only on soils treated for 1 and 2 hours.
Mercury concentration decreases from 5.8 mg/kg to 0.4 mg/kg after 1 hour treatment at
300C, then gradually goes to values close to zero for longer times and/or higher
temperatures (figure 7). Data obtained for BTEX and hydrocarbons are reported in table 4.
An evident decrease for all contaminants is observed after 1 hour of treatment at 300C,
and a complete removal at 450C for all the investigated treatment time.
Table 4: BTEX and hydrocarbon concentration in D43 first and after thermal treatment
4. Conclusions
This study presents the results of thermal desorption process to clean up soil contaminated from
different compounds such as mercury, DDX and a variety of hydrocarbons. Temperature and time were
considered the most important parameter for defining conditions of contaminants removal.
In case of soil contaminated from DDX and mercury, reduction in concentration was noticed after a
treatment of 350C for one hour (from 1600 to 1.9 mg/Kg).
For soil contaminated by hydrocarbons and mercury, two temperatures were identified: 300 and 450C.
Considering a sample with starting values of 6 mg/kg of mercury and 10000 mg/kg of total hydrocarbons,
300C for 1 hour was enough for reducing the concentration under the limits set by Italian law. At 450C
a complete removal of both contaminants was obtained.
In conclusion, thermodesorption under air flow conditions can be an effective remediation technology
for soils contaminated by a different kind of contaminants and may be convenient for in situ application.
5. References
M.J.C. Henssen, M.H.A.B. Wagelmans, A.S. Roosma, H.M.T. de Vries (Bioclear b.v.,
Rozenburglaan 13. PO Box 2262, 9704 CG Groningen, The Netherlands)
Summary
At several sites in the Netherlands Bioclear stimulated biological degradation of volatile organic
chlorinated compounds (VOC) in source zones is applied successfully. The technique often used
is based on direct injection of carbon source and, if needed additional specific dechlorinating
bacteria Dehalococcoides spp. (DHC). This technique is, instead of most in situ remediation
techniques, very well applicable in poorly permeable soils like clay and peat and at those sites
that are more difficult te reach (like urban areas). Direct push injections are typically used to
remediate a source zone. Research and field experience have shown that chlorinate d ethylenes
concentrations close to the maximum solubility can be biologically degraded.
Direct injections are performed with a small injection machine. Protamylasse can be used as
carbon source. Protamylasse does not only contain organic carbon but cont ains also enough
nutrients for the growth of dechlorinating bacteria. Protamylasse is an organic waste stream from
the potato starch industry. Which is an environmentally friendly (and cheaper) alternative
compared to chemically produced carbon sources like lactate, acetate or hydrogen release
compound. If injection of dechlorinating bacteria is desired, a culture with a high concentration of
dechlorinating bacteria is obtained from a lab.
The biological treatment of source zones in urban areas has proven to be a very interesting
method. Although remediation time may take various years, the cost of such a remediation is quite
low and hindrance during remediation is minimal. Especially for those source zone that pose no
human health risks at the moment but should be remediation in order to prevent flux of
contaminants into the groundwater, this approach is very suitable.
Introduction
Full scale applications have shown that chlorinated ethylene concentrations close to the maximum
solubility are biodegradable. The reason that biological DNAPL degradation is possible is the
presence of micro niches in the soil matrix. The conditions in these micro niches differ from the
overall soil conditions and can be favorable for anaerobic dechlorination. Besides dechlorination,
biological activity also accelerated the dissolution of free phase product up to a factor 16 which
has a positive effect on the total degradation proces. The accelerated dissolution is a result of the
lowered chlorinated ethylene concentration in the groundwater due to biological degradation.
When applied properly biological source zone treatment has several advantages:
- Capital and operational costs are usually lower than those of other DNAPL source zone
treatment technologies;
- The overall hindrance due to remediation activities is quite lower than with other
techniques (like ISCO, thermal treatment), also since no additional equipment or dosing
systems are needed at the site;
- Other contaminants present with the chlorinated ethylenes, specifically other chlorinated
organic compounds, may simultaneously be degraded.
- The technique can be used in combination with a number of other treatment technologies
as part of a larger overall site treatment strategy.
- By removing contamination mass, the overall treatment time frame at a DNAPL-
contaminated site is shortened and natural attenuation of the plume contaminants can be
used more effectively (lowered flux from source).
To verify the many promises of the biological source zone treatment technique, five full scale
biological treatment projects conducted in The Netherlands have been reviewed.
The first full scale source zone treatment application in The Netherlands was initiated in 2001 in
Appingedam. Several other source zone treatment projects have followed.
Biological source zone treatment accelerates natural degradation processes by adding carbon
source and, if needed, chlorinated ethylene degrading bacteria to the soil (Bio-augmentation). The
added carbon will result in redox conditions and organic carbon concentration favorable for
reductive dechlorination. Active bacteria will degrade solvents like tetrachloroethene (PCE) and
trichloroethene (TCE) to harmless ethene and ethane without the accumulation of toxic
degradation products like dichloroethene (DCE) and vinylchloride (VC). The degradation pathway
is visualized in figure 1.
Figure 1. Anaerobic biological dechlorination of tetrachloroethene (PCE)
The monitoring of the remediation processes of the five full scale locations is focused on the
applied technique, contaminants, redox parameters, total organic carbon (TOC), microbial
populations and formation of methane and sulfide (important safety issues). A typical monitoring
consisted of 4 rounds. The results of the monitoring are discussed in the next paragraphs. Table 1
shows an overview of the evaluated treatment projects.
In most cases the source zone remediation is executed by injection of carbon sources (and in
some cases additional bacteria) using direct injection technology. This technique is quite well
applicable for these situations, especially in densely build-on urban areas. The technique may
also be applied for injection underneath buildings.
Due to the low permeability of the soils treated (quite often the case in the Netherlands in the
upper soil layers) a grid with a distance of 1 to 2 meter between every injection point is needed
and used. In various cases Nutrolase (protamylasses), molasses and/or lactate was used for
carbon source. Carbon source is overdosed by a factor ten to ensure a successful treatment. In
3
four of the five location 0.23 to 0.85 kg TOC/m soil volume is added (TOC is a measure for the
total amount of carbon). At the location IJlst a carbon source shock load was applied, resulting in
an addition of 1.5 kg TOC/m3 soil volume.
Treatment results
Direct injections create a biologically active zone in which the contamination is degraded to
harmless degradation products. An important side effect of the biological active zone is that it
prevents migration of chlorinated ethylenes from the source zone to the plume. In other
words: the mass flux of contaminants leaking into the plume will decrease in course of time due to
the stimulated degradation in the source zone.
If natural attenuation occurs in de plume, source zone removal will result in a retreating plume and
immediate groundwater quality improvement. Figure 3 depicts a schematic overview of how direct
injection can treat a small heavy contaminated area resulting in a retreating plume. The
combination of Natural Attenuation in the plume and bio-source zone treatment therefore results
in a cost-effective solution, whereby remediation actions only take place in the relatively small
area of the source zone.
The last column of table 1 shows the remediation results of the six sites. On the location
Appingedam remains one contaminated well, the locations The Hague and IJlst are completely
remediated. Location Haren and a location in the west of the Netherlands show a very active
degradation process. The site in Groningen was closed with positive result after 8 years, with only
2 injection rounds. Retreating or stable plumes have been observed in five of the six evaluated
locations.
The duration of a biological source zone treatment depends on the type and amount of
contaminant present. If the source zone contains high concentrations or free phase product PCE
and/or TCE the treatment duration varies between 5 and 10 years. Sites contaminated with DCE
and VC can be treated very fast. Location The Hague showed that it was possible to treat the
DCE and VC contamination in one year.
Table 1. overview of the evaluated biological source zone treatment projects.
Site Size of Maximum conc Used carbon Longevity Injection Duration Treatment result
source (g/l) at t=0 source carbon grid (m) treatment
zone source (years)
(years)
Appingedam 65 m 2 PCE 69,000 Molasses/ lactate 9.0 1,5 13 Size of source
520 m 3 TCE 69,000 zone is strongly
DCE 150,000 reduced.
Complete plume
VC 14,000 removal
The Hague 90 m 2 PCE 5,900 Nutrolase 2.8 1,5 1 Source zone
1080 m 3 TCE 3,100 (protamylasses) treated in one
DCE 19,000 year. Max.
concentration
VC 700 after treatment:
0,005 mg/l VC
Haren 680 m 2 PCE 100 Nutrolase 2.6 2 2 Within 2 year all
6800 m 3 TCE 53,000 (protamylasses) TCE degraded to
DCE and VC.
DCE 32,000
Plume size is
VC 300 decreasing
West of The 156 m 2 PCE 40 Nutrolase 1.9 1 2 Source zone has
Netherlands 936 m 3 TCE 40 (protamylasses) decreased in
size. Active
DCE 140,000
degredation
VC 80,000 process is
ongoing
IJlst 250 m 2 PCE 57,000 Nutrolase Unknown 2 2 Size source zone
3750 m 3 TCE 42,000 (protamylasses) strongly reduced.
DCE 34,000 Plume is
decreasing in
VC 5,700 size.
Groningen 4,300 m 3 c-DCE 33,000 Protamylasses 3-4 years 2 8 Within 2 injection
VC 22,000 rounds (2006 and
2010) 90%
reduction of I-
value contour
was reached and
leakage of
contaminants to
lower gw-layer
was stopped
Carbon source
The feasibility and cost-effectiveness of the presented technique with direct injection of carbons
sources is very much dependent on the longevity of the carbons sources. If consumption of the
carbon source is taking place too quickly, re-injection may be needed and cost (and hindrance at
the site) will increase. Therefore longevity of the carbon source is important.
Based on the monitoring results from the various sites it can be concluded that longevity of the
substrates like protamylasse is on average 3 to 4 years and therefore comparable with slow
release carbon sources like HRC and emulsified oils. Re-injections are therefore normally
needed within a timeframe of approximately 4 to 5 years. Figure 4 shows a typical TOC
concentration development in time as we monitored it at various sites.
Figure 4, TOC concentration in time of well 3.2 at location Appingedam
The injection of large quantities of carbon source could cause soil acidification. Measurements at
the evaluated sites showed a decrease in pH in the first months after the injections. The lowest
measured pH was 5.5. Since dechlorinating bacteria can cope with pH 5.5, acidification causes no
inhibition of biological treatment processes. Within 1 or 2 years the pH had rebounded on all
locations to the original level. A typical pH development after carbon source injections is depicted
in figure 5.
Figure 5. pH development after carbon source injections in well 302 at location The Hague.
The injection of large quantities of carbon source might result in unwanted methane and sulfide
gas production. Two sites were investigated but no accumulation of methane and sulfide above
critical concentrations were measured. This shows that when designed properly, direct injection of
carbon source does not cause any human health or explosion risks.
Costs
The costs of biological source zone treatments is 20 to 30 euro/m3 soil volume, including
environmental supervision and monitoring (in source areas with a soil volume > 2,000 m3). The
costs are low compared with conventional treatment techniques such as chemical oxidation
(average 60 euro/m3) and electroreclamation or thermal treatment systems (170 and > euro/m3).
In the case Haren and 'Groningen' biological source zone treatment resulted in a cost reduction
of 70% compared to the alternative treatment techniques suggested for this site. The reduced
costs made it for the local government possible to raise enough funds to start the treatment
process. The specific remediation results for the location Haren are given below.
On site, carbon source as well as dechlorinating bacteria have been injected. Direct injections were
chosen is stead of infiltration en extraction because of the low permeability of the soil. The injections
have been performed within the area containing 99% of the contamination. A total 200 injections to 9
meter below ground level have been performed. The treatment progress is monitored from 2007.
The total costs of the treatment are 175,000 including extra injections after five years , quality checks
and monitoring of the treatment progress for ten years. This is 26.50 per cubic meter soil. The choice
for biological source zone treatment resulted in a cost reduction of 70% compared to alternative
treatment techniques. The reduced costs made it for the local government possible to raise enough funds
to start the treatment process.
Current status
Complete degradation of chlorinated ethylenes occurs. Derived from the amount of degradation products
(VC 130,000 g/l en ethene 130,000 g/l) it is concluded that more contamination was present than
expected. Within one year the size of the contamination plume has decreased with 30%. Gas sampling in
the unsaturated zone showed no vaporization of chlorinated ethylenes like VC. Human health ris ks were
thereby ruled out.
The treatment progress is according plan. The conditions for biological degradation are favorable and an
active bacterial population is present. Due to the large amount of chlorinated ethylenes present a re -
injection of carbon source is needed in time.
Biological source zone treatment, case Groningen
Ground and groundwater at a site in Groningen is contaminated with volatile chlorinated ethylenes
(VOCl) as a result of usage of degreasing agent for the polishing of glasses ( spectacles). The
contamination was mainly present in the poorly permeable clayey soil layer till 6 meter below ground
level (m-gl). In 1994 the groundwater at the site was remediated by extraction of groundwater and
aboveground treatment before discharging. Because of insufficient remediation results, was in 2002
Bioclear asked to formulate an alternative remediation approach based on in situ sti mulated biological
degradation. The site was mainly contaminated with the components cis-1,2-dichloroethylene (cis-DCE)
and vinyl chloride (VC). The maximum concentrations measured concerned respectively 33.000 g/l en
3
22.000 g/l. The extent of the contamination amounted approximately 4.300 m strongly (> I-value
according to the Dutch soil policy) contaminated soil volume.
The in situ biological remediation approach exists of injection of carbon source in the upper clayey soil
layer till 6 m-gl and ten-yearly monitoring of the groundwater in de upper soil layer and the underlying
aquifer. In 2006 the injections of carbon source were performed and in 2010 local additional carbon
source injections were performed.
Since the performance of the carbon source injections, the contamination completely degraded to
harmless end products (ethylene and ethane). Within the stimulated area the concentrations of
contamination are decreased below the T-value (Dutch soil policy remediation value). The total strongly
contaminated (> I-value) soil volume is from the start of the biological in situ remediation decreased with
approximately 90%. Furthermore the source of the contamination is removed as a result of the stimulated
biological degradation. Therefore leakage of contamination from the source zone in the upper clay layer
3
into the underlying aquifer has stopped. The remediation resulted in a limited volume (< 1.000 m ) of
remaining contamination with concentrations above the I-value. Moreover there are no risks as a result of
dispersion of contamination and no environmental protection areas are at risk.
Within eight years a successful remediation of the site was exectuted by in situ stimulated biological
degradation using only two carbon source injection rounds.
Sustainability
A biological source zone treatment can be subdivided in an active and passive phase. The active
phase is typically a few weeks of direct injections. The passive phase follows the active phase
and is the time needed for the bacteria to degrade the chlorinated ethylenes , this phase consist of
just monitoring activities. Because the active phase is relatively short the amount of energy
needed, compared to techniques like pump and treat, is small. Related to the energy demand is
the carbon dioxide emission, which is therefore smaller for biological source zone treatment.
Direct injections are performed with a small vehicle that can be employed in ha rd to reach areas.
If a proper design is applied no dangerous gasses like sulfide and methane are formed. And
because no corrosive chemicals or high temperatures are used no specific safety measurements
have to be taken. This makes direct injection a very safe technique with minimal hindrance for the
surroundings.
The resources needed for the direct injection consist mainly of carbon source. In four out of five
evaluated sites nutrolase (protamylasse) is used. This is a rest product from the potato process ing
industry. There is no need for chemical oxidizing agents, heat generation or the installation of
transport networks for chemical agents or groundwater, which are needed for techniques like
chemical oxidation and pump&treat. This makes biological source zone treatment a very
sustainable and resource efficient technique.
In the Netherlands focus has shifted in the last decade from cleaning up all contaminated sites as
much as possible (both source zone and plume areas) towards more risk-based approaches. This
in combination with the urgency to develop solutions in which management of groundwater
contamination (mainly plume areas) is combined with societal issues and needs like sustainable
energy and climate adaptation. In these situations still source zone removal stays important.
Source treatment can be very effective in the context of an area oriented approach. Lowering the total
mass in the source will decrease the mass flux of contaminants entering the groundwater. Due to this
lowered mass flux natural attenuation processes in groundwater and sediments may cope more easily
with the existing contaminants and the feed from sources. In situ biotreatment of source zones can
offer an efficient and sustainable option for this.
Conclusions
Based on the evaluation of various full scale biological source zone treatments it can be
concluded that:
- It is possible to remediate chlorinated ethylene concentration close to the maximum
solubility in one to ten years.
- Biological source zone treatment prevents migration of chlorinated ethylenes from the
source zone to the plume. If natural attenuation occurs in de plume, source zone removal
will result in a retreating plume;
- also in an area oriented approach - focusing on removal of the source and decline mass
flux entering the groundwater - bioremediation of source zones can be effective;
- Direct injection of carbon source and dechlorinating bacteria can be performed in difficult
to reach urban areas, causes little hindrance for the surroundings, are sustainable and
resource efficient.
- Biological source zone treatment does not have a negative effect on the soil properties
like pH and methane of sulfide gas production, if properly designed.
- Low permeable peat or clay layers in the soil do not have an adverse effect on the
treatment process.
- It is cheaper (20-50% of normal cost) compared to alternative source zone treatment
techniques and more sustainable.
Form of presentation Poster
Topic area Theme 1. Dealing with contamination of soil, groundwater and sediment
1c. Remediation technologies and approaches
[Keywords]
Sustainable remediation, In situ remediation, Anaerobic bioremediation, CVOC, Surf-UK framework,
Pump and treat
[Abstract]
Sustainable remediation has received much attention in recent years due to its importance for
redevelopment in urban areas. A few research groups in Japan have begun using quantitative tools to
determine the best approaches for sustainable remediation. In situ remediation for sites that have been
contaminated by chlorinated volatile organic compounds appears to be a promising technique.
However, case studies that can estimate the effectiveness of different technical options for in situ
remediation are required. We conducted a sustainability assessment of five options using the
environment, economy and social subcategories to conduct a sustainability rating of all the options.
The in situ remediation systems assessed were as follows: pump and treat; pump and treat with
solar-powered pumps; anaerobic bioremediation; anaerobic bioremediation with automatic injections
and pumping; and thermally enhanced bioremediation. The pump and treat system had the highest
levels of emissions of CO 2 and air pollution. The anaerobic system with automatic injections and
pumping had the lowest accident rate. Thermally enhanced bioremediation had the lowest cost. By
considering these subcategories of the sustainability assessment, it is possible to contribute to the
debate with stakeholders surrounding the site.
1. Introduction
In the last decade, a number of countries have developed legislation focused on soil. There are over
62,000 hectares of brownfield sites in England (Department for Communities and Local Government,
2007) and over 28,000 hectares in Japan (Expert Studying Group for Countermeasures against
Brownfields, Ministry of the Environment, Government of Japan, 2007). Useful approaches for
remediation have been developed over the past four decades. Cost-based approaches were the
principal methods of such decision-making in the 1970s. Since the 1990s, risk assessment approaches
have been used, such as that detailed in the contaminated land regime in the UK (Environment Agency,
2007). A sustainable approach is now becoming a new driver in the remediation of contaminated land.
Various new techniques which consider the need to reduce the environmental or social impacts are
now available for use with chlorinated volatile organic compounds (CVOC) In situ remediation can be a
sustainable technique; however, it still requires modification or implementation. While the Surf-UK
sustainability framework offers a set of categories which would widely be acceptable as a policy starting
point, this framework lacks options for more sustainable use of land. In addition, note that some
research groups in Japan have just begun to assess sustainable approaches by using various tools.
In this study, we provide a case study of a Japanese site and evaluate the relevant parameters
within sustainability subcategories to compare technical options. We assessed a small in situ
remediation technique with three options: a solar-powered groundwater pumping system, a thermally
enhanced system and an automatic operation system. We compared these options with pump and treat
and anaerobic bioremediation. The objective of this study was to conduct a case study of these options
within sustainability subcategories.
2. Methodology
2.1. Approach and output parameters
A quantitative approach in a sustainable framework was applied to consider remediation technology
TM
used. This study was conducted using the concept of SRT , which was devised by the US Air Force
(U.S. Air Force, AFCEE, 2010). The advantage of using this tool was the flexibility in selecting
remediation technology by using a Microsoft Excel spreadsheet in the early phase. By adopting this tool
with the Japanese inventory for a model site, the current methodology was made applicable to
Japanese sites. The inventory of remediation options used for this study was selected from MilLCA
(Japan Environmental Management Association for Industry, 2014).
The situation of this case study was site-specific; therefore, only selected parameters were used.
As shown in Table 1, only important sub-categories in the Surf-UK framework were considered in this
case study; therefore, some parameters such as impacts on ecology or indirect economic costs were
not significant for this case study. Moreover, most of the social category parameters were not
considered in this study because it was unlikely that there were significant differences between the
techniques used in this case study. After calculation of the environmental impacts, social parameters
and economic parameters resulting from the use of the selected methods, specific parameters were
compared with each other.
Anaerobic bioremediation
(with automatic injections and pumping) Thermally enhanced bioremediation
Fig. 2 System boundaries for each technique
2.3.1 Pump and treat
The pumping process was set up to ensure non-proliferation to the outside. Three wells and three
pumps (SQE5-35, 200V, 1.65kW, Grundfos) were set at the downstream side of the area. Pumped
groundwater was treated by aeration processing. The treatment system mainly consisted of an aeration
blower, a disposal blower, a vapour liquid separation tank and an active carbon tower. We assumed a
processing rate of 3,000 L/min for the aeration blower with an air-to-water ratio of 100:1. The total
electricity required for the treatment system was assumed to be 15.3 kW. The duration of this technique
was set as 25 years because several sites in Japan have used a similar system. In addition, the
operation time was calculated by 90% of the 25 years in this case study.
Fig. 5 CO2 emissions from the different systems Fig. 6 Air pollution emissions
A similar result was expected for CO2 emissions for thermally enhanced remediation, but the emissions
from this technology were somewhat lower than the other options. We can infer that the short duration
for this type of remediation can reduce its impact. It can be also assumed that long-term monitoring
over 15 years produced high environmental impacts. The results for the solar-powered pumping
system show that continuous use for 25 years can remarkably reduce the environmental impact
compared with pump and treat.
3.4 Discussion
The aim of this study was to assess the relative footprints and differences between several options for
remediation of contaminated soil. The difficulty of selecting options can be reduced by utilizing the
Surf-UK framework. Although numerical tools can provide useful information on sustainability, these
tools require the integration of enormous numbers of parameters. In earlier studies attempts were
made to use multi-criteria analysis (Harbottle et al. 2008), or costbenefit analysis or net present value.
In this study, we did not attempt to integrate sustainability subcategories. This was because the options
considered in this study were used very similar techniques and the data for assessment would be
enormous. In addition, we considered that it was most important to judge the possible techniques with
options at the early stage. The trial we conducted in this study was simple but establishes an easy
methodology for sustainable remediation that is particularly relevant to Japan. Future studies should
focus on the parameter sensitivity within sustainable appraisal and the other parameters in the
economic category.
As mentioned previously, one aim of this case study was to demonstrate the effectiveness of
sustainable remediation. International standards for sustainable remediation have been developed by a
group of experts as ISO/TC190. In addition, a local government in Japan has provided a calculation
tool that can estimate the environmental footprints of 20 remediation techniques. In the future, we
suggest that these two tools will be a good introduction and contribution to the debate on use of
sustainable remediation in Japan. The approach applied in this study has the potential to assist the
debate.
4. Acknowledgement
A part of this project was supported by the New Energy and Industrial Technology Development
Organization (NEDO).
References
Sustainable Remediation Forum-United Kingdom (SuRF-UK), 2009, A review of sustainability indicator
sets: How applicable are they to contaminated land remediation indicator-set development? CL:AIRE,
London, UK, ISBN 978-1-905046-18-8, www.claire.co.uk/surfuk
Department for communities and local government, 2007, Previously-developed land that may be
available for Development: England 2007 Results from the National Land Use Database of
Previously-Developed Land, London,
http://www.communities.gov.uk/documents/corporate/pdf/945914.pdf
Expert Studying Group for Countermeasures against brownfields, Ministry of the Environment,
Government of Japan, 2007, Current status of the Brownfields Issue in Japan Interim Report,
http://www.env.go.jp/en/water/soil/brownfields/interin-rep0703.pdf
Environmental agency, UK, 2009, Dealing with contaminated land in England and Wales
A review of progress from 2000-2007 with Part 2A of the Environmental Protection Act
U.S. Air force , AFCEE, 2010, Sustainable remediation tool user guide,
http://www.afcee.af.mil/resources/technologytransfer/programsandinitiatives/sustainableremediation
(currently unable to access)
Japan environmental management association for industry, 2014, MiLCA guide book,
http://milca-milca.net/download-files/MiLCAguidebook_En.pdf
SuRF-UK: A framework for assessing the sustainability of soil and groundwater remediation,
Anne K. Friisa, Axel C. Heimanna, Rasmus Jakobsena, Hans-Jrgen Albrechtsena, Evan Coxb,
Poul L. Bjerga, 2007, Temperature dependence of anaerobic TCE-dechlorination in a highly enriched
Dehalococcoides-containing culture, Water Research, Volume 41, Issue 2, January 2007, Pages 355
364
Harbottle, M. J., Al-Tabbaa, A., Evans, C. W., 2008, Sustainability of land remediation. Part1:overall
analysis, Proceedings of the ICE - Geotechnical Engineering, Vol.161, Issue 2, Pages 75-92, DOI
10.1680/geng.2008.161.2.75
MODEL TESTING OF NEW APPROACH FOR CLEANING SOILS POLLUTED WITH
ORGANOCHLORINE PESTICIDES
Kurashvili Maritsa, Adamia George, Amiranashvili Lia, Ananiashvili Tamar, Pruidze Marina, Varazi
Tamar, Gordeziani Marlen, Khatisashvili Gia
Abstract. Organochlorine pesticides belong to Persistent Organic Pollutants (POPs) as they have
high chemical stability and difficultly undergo biotic and abiotic transformations. They easily
accumulate in plants and animal tissues and incorporate into the food chain that causing a great
danger for health.
The goal of presented work is the development phytoremediation method targeted to cleaning
environment polluted with organochlorine pesticides (Lindane - 1,2,3,4,5,6-hexachlorocyclohexane,
DDT - dichlorodiphenyltrichloroethane and PCP - 1,2,3,4,5-pentachlorophenol), based on joint
application of plants and microorganisms. Created technology is based on following concept: initial
degradation of pesticides carry out by specially selected microorganisms; the forming dehalogenated
products easily uptake by the plants and undergo oxidative degradation via plant detoxification
enzymes. This approach can complete degradation of toxicants and their mineralization into nontoxic
compounds.
For this aim the plants with high activities of enzymes, directly participating in oxidative degradation of
organic pollutants and in conjugation of toxic metabolites, have been chosen. The strains of
rhizospheric microorganisms degrading organochlorine pesticides and stimulating plant growth have
been selected by screening. According to results of gas chromatographic analyses of residual
pesticides in incubation medium after cultivation, the best strains with high pesticide assimilating
capacities have been revealed. These strains can degrade organochlorine pesticides from cultivation
area and are usable for application in developed phytoremediation technology.
For testing the development technology, the laboratory model experiments for remediation of
artificially contaminated soils with pesticides (Lindane, DDT and PCP) by using consortia of
Pseudomonas strains and selected plants, have been carried out. After 1 month the soil samples
analyses show that content of toxicants in contaminated soils decreased by 60-80%. According to
obtained results the created technological approach will be effectively used for cleaning the soils
polluted with organochlorine pesticides.
Introduction
In the new ecological situation typified by massive emissions of technogenic pollutants in environment,
plants and microorganisms exhibited new capabilities for the absorption and metabolic degradation of
toxicants. This became the basis for development of phytoremediation technologies, based on unique
ability of plants and microorganisms to uptake and degrade of wide spectra of chemical pollutants
(among these organochlorine pesticides) via enzymatic transformation [1].
Concerning to biological methods of remediation that are generally most effective, economically
profitable and ecologically friendly technologies for rehabilitation of chemically polluted environment,
but in case of organochlorine pesticides, such biotechnologies have low productivity as only a few
bacterial strains are able to perform the full mineralization of these pollutants as a result of consequent
aerobic and anaerobic conversions [2]. For detoxification of organochlorine pollutants initial
dehalogenation of their molecules is necessary. As a result, the residual carbon skeletons become
accessible for the oxidizing enzymes that degrade toxic compounds up to cell regular metabolites. It
should be noted that data about plants ability to degrade organochlorine pesticides are very exiguous.
The using of plant-microbial tools for intensification of phytoremediation process may be effective in
case of pollution with high toxic and persistent organic pollutants such as organochlorine pesticides.
According to data, obtained as a results of our previous investigations of the molecular mechanisms of
organochlorine toxicants (DDT, Lindane and PCP) degradation in plants, it has been shown that in
vitro experiments detoxification enzymes of some plants have shown high ability of transformation of
organochlorine pesticides [3]. In spite of this, in model experiments the using of these plants for clean-
up of artificially contaminated soil was not enough effective. We supposed that it is caused by low
solubility and consequently low bioavailability of tested pesticides. For solution of this problem, we set
a goal to develop a new biotechnological approach based on the targeted use of plants and
microorganisms with high detoxification potential and corresponding capabilities for completely
removing these pollutants from the environment.
The goal of presented work is the development phytoremediation method targeted to cleaning
environment polluted with organochlorine pesticides, based on joint application of plants and
microorganisms.
120
Residual content of pesticide, %
100
80
60
40
20
0
Ps. PS63
Ps. PS8D
Ps. PS66
Ps. 8JL63
Ps. 4JL50
Ps. 4G14
Ps. 8JL63
Ps. TP335
Ps. TBM5
Ps. TBM6
Control
Ps. 751-6D
Ps. GN28
Model Experiments
The model experiments for elaboration of new approach of phytoremediation technology for cleaning
soils polluted by organochlorine pesticides have been carried out. In model experiments have been
used selected Pseudomonas strains and plants soybean, maize and alfalfa.
The obtained results show that for phytoremediation of soils polluted by DDT is most effective using of
concortia (Pseudomonas sp. TBM6 and Pseudomonas sp. 4JL50) with soybean (Table 1). In this case
DDT content in soils decreased by 80%. For phytoremediation of soils polluted by PCP is most
effective using of concortia (Pseudomonas sp. 8JL63 and Pseudomonas sp. TP335) with maize. In
this case PCP content in soils decreased by 67%. In experiments with Lindane, the content of
organochlorine pollutant in contaminated soils decreased by action of Pseudomonas sp. 4G14 and
maize by 71%.
Table 1. The results of model experiments using selected Pseudomonas strains and plants for
cleaning soils polluted with organochlorine pesticides. Duration 1 month; level of soil contamination
100 ppm; temperature - 25C; plants sowed after 7 days from inoculation of bacteria suspension.
Residual content of
Pesticide Bioremediation agent Phytoremediation agent pesticide, % from initial
concentration
Consortia of Pseudomonas
DDT strains (sp.TBM6 + sp. Soybean 19.7 1.0
4JL50)
Pseudomonas sp. 4G14 Alfalfa 37.8 1.9
Lindane
Pseudomonas sp. 4G14 Maize 29.2 1.5
Consortia of Pseudomonas
strains (sp. 8JL63 + sp. Alfalfa 44.4 2.0
TP335
PCP
Consortia of Pseudomonas
strains (sp. 8JL63 + sp. Maize 33.2 1.8
TP335
Thus, results of model experiments show that the developed new technological approach is usable for
successful phytoremediation of soils polluted by organochlorine pesticides.
Conclusion
Thus, as a result of carried out work, the new effective phytoremediation technology for cleaning of
soils contaminated with organochlorine pesticides has been developed. This technology is based on
join application of selected plants and microorganisms with high detoxification abilities.
References:
[1] Tsao D. (2013). Phytoremediation. Advances in Biochemical Engineering and Biotechnology.
Springer, Berlin Heidelberg New York, pages 207.
[2] Kvesitadze G., Khatisashvili G., Sadunishvili T., Ramsden, J. (2006). Biochemical Mechanisms of
Detoxification: Basis of Phytoremediation. Springer, Berlin Heidelberg New York, pages 262.
[3] Kurasvili M., Adamia G., Ananiasvili T., Varazi T., Pruidze M., Gordeziani M., Khatisashvili G.
(2014). Plants as tools for control and remediation of environment polluted by organochlorine
toxicants. Ann. Agrar. Sci., Vol. 12, 3, p. 84-87.
[4] Fatoki O., Awofolu R. (2003). Methods for selective determination of persistent organochlorine
pesticide residues in water and sediments by capillary gas chromatography and electron-capture
detection. Journal of Chromatography A. Vol. 983, 12, p. 225236.
POSTERPRESENTATION:1c. Remediation technologies and approaches
ISCOAPPLICATIONINASALTMARSHAREACONTAMINATEDBYFUELINSOUTHERNSPAIN(BIOXISOILPROJECT)
S.delReino1,M.RodrguezRastrero2,O.Escolano2,F.J.Daz2,J.Bueno1,JL.Fernndez1,G.Carretero1,R.Milln2
1AsociacinparalaInvestigacinyDesarrolloIndustrialdelosRecursosNaturales(AITEMIN).CalleMargaritaSalas14,Legans,Madrid
2CentrodeInvestigacionesEnergticas,MedioambientalesyTecnolgicas(CIEMAT)AvenidaComplutense22,28040Madrid(Espaa)(Espaa)
Email:susana.del.reino@aitemin.es
Keywords:
InsituChemicalOxidation(ISCO),
INTRODUCTION
Insituchemicaloxidation(ISCO)isaneffectivetechnologyforcleanupsitecontaminatedbyorganiccompounds.This
remediation involves the introduction of a chemical oxidant into the subsurface for the purpose of transforming
groundwaterorsoilcontaminantsintolessharmfulchemicalspecies.Commonlyappliedoxidantsarepermanganate,
hydrogenperoxide(withorwithoutiron),ozoneandpersulfate.ForISCOtobeeffective,theoxidantmustcontactthe
contaminant. This can be difficult in many soils and aquifers where natural heterogeneities can result in flow
bypassing around lower permeability zones or where the presence of natural compounds can generate non
productivereactionsthatconsumeoxidantcompromisingtheadequateoxidanttransportanddistribution.
ThiscasestudydescribestheremediationofhydrocarboninacomplexsiteinasouthernofSpain.Sitesoilsarethe
result of an accumulation of earthy anthropogenic materials on marsh deposits, showing important constraints for
ISCO development such as a temporary shallow water table and an irregular vertical and horizontal distribution of
contaminationaswellasahighcalciumcarbonatecontentandsalinity.
OBJECTIVES
The overall objective of the project is to evaluate, in real conditions, an ISCO application in a saltmarsh area
contaminatedbyfuelinsouthernSpain,hydraulicsandhydrogeochemicalyheterogeneous,byISCOtechniqueusing
asoxidantcatalyzedhydrogenperoxide(Fenton'sreagent).
Thisobjectivewillbeaddressedbythepartialachievementofthefollowingobjectives:
Conductathoroughsitecharacterizationinordertoevaluatesoilandwaterpollutionaswellaspedological
andhydrologicalproperties.
Study the impact of the physical and chemical subsurface heterogeneities in the success of the ISCO
technique.
Evaluate and validate the effectiveness of the modified Fenton reaction in terms of TPH reduction and
oxidationproductformation.
METHODOLOGY
Theinitialdirectpushtechnologieswereusedtomakeacompletesitecharacterization.Atotalof25soilpitswere
realized with pedological criteria including corresponding soil sampling and the soil profile description. Hydraulic
parameters like flow pattern or hydraulic conductivity were obtained by topographical survey of the site and by
pumpingtestsinsoilpitsandpreexistingboreholes.
In order to study the impact of the physical and chemical subsurface heterogeneities and to design the injection
programamoredetailedcharacterizationofsoilandgroundwater,consistingonwelldrillinginaselectedareaof900
m2(figure1)andthecontinuousmonitoringoftheundergroundwaterlevelevolutionbytheinstallationofautomatic
Divertyperegistrationsystemswererealized.
Monitoring wells were pumped, to evacuate of a minimum of three volumes of water, prior to collecting a
groundwatersamplewitha12Vtotakearepresentativesample.Groundwatersamplecollectiontookplace24hours
followingwellpurgingwithbailersmadeofTeflon.Samplingoccurredinaprogressionfromtheleastto themost
contaminatedwell.Likewise,measurementsfortemperature,pH,ORP,conductivityandO2weretakenatthetimeof
thesampling.
Figure1.Sitelayoutandthemonitoringnetwork.
RESULTS
Based on one of the test pits performed, the characteristics of the materials constituting the zone of action are
summarizedintheFigure2,
Anthropogenicfilllevel.Coarsegrainsize,basicpH;
abundant CaCO3; moderate organic matter
content;
variable salinity generally low; well rooted; well
oxygenated
(reddishcolours).
Shells level. level shells. Finegrained (clayey
sandy)trend
laminarstructure.Abundantpresenceoftracesof
astreidos, pectinides and others, occasionally
cobble and
Figure2.ConceptualSiteSoilModel.
Thisconceptualmodelindicatethat,fromahydraulicallypointofview,socalled"shelllevel"locatedwithinadepth,
between0.20and1.20m.,islimitedinhighbyaanthropogenicfillmorepermeableandindepthbyasludgelevel,
rich in clay, compacted and very impermeable. This physical heterogeneity can difficult oxidant distribution and
transportcapacityinthesubsurfaceandconsequentlytheefficiencyofthetreatment.Chromatographicsoilanalysis
showedpresenceofTPHintheshellslevel.
Water analysis in monitoring wells showed high content of natural scavengers (inorganic carbon, chlorides and
sulphates)thatcouldaffectthenaturaloxidantdemand,thatisanexpressionofhowmuchoxidantisconsumed,due
tothefactthattheycompetewiththecontaminantsofconcernforhydroxylradicals.
Thesefindings(subsurfaceheterogeneities,preferentialflowpathsandpresenceofscavengersthatconsumeoxidant)
hadimplicationsfortheprogramdesigninjection:
A. Tocushiontheimpactofphysicalheterogeneitiesandpresenceofpreferentialflowpaths:
Injections wells were constructed of polyvinylchloride (PVC) pipe with the screen interval placed in the
verticalsectionintendedfortreatment.Injectionswellareconstructedwithslotwellscreensandcoarsesand
pack located to the same depth or slightly deeper than, the bottom of shell level (the most contaminated
level).
Affectedlevelwasisolatebypackers(figure3).
Injectionhead
Piezometricpipe
Injectionreagents
routes
Packer
Figure3.Headinjectionandpackerusedtoisolateshelllevel
B. Tocushiontheimpactofpresenceofnaturalscavengers:
Hydrogenperoxidestabilizerswereused.
Hydrogenperoxideandcatalystsolutionwithironandhydrogenperoxidestabilizerwereinjectedseparately
inordertoincreasetheradiusofinfluence.
Assumingthesedesigncharacteristic,thereagentsinjectioninaselectedareastartedinJanuary2013.Afterinjection
of 9000L, by the help of an Automated injection device by Aitemin, that allowed to work in four injection wells
simultaneously,andwiththehelpofinjectionheadsattachedtoaplugarearrangedinsidewells(Figure4and5),a
new groundwater sampling were realized in order to evaluate the efficiency of oxidation (TPH removed). Injection
volumes, the concentration of TPH removed (g / L) and oxidation performance survey for each injection wells are
showninTable1.
Reagent injection
lines
a) b)
Figure4.Detailsoftheautomatedinjectiondevice
Figure5.Headinjectiondetail
Table1.Injectionvolumes,theconcentrationofTPHremoved(g/L)andoxidationperformancesurveyforeachwells.
Volumeof [TPH]removed Performance(%)
Wells oxidant(L) (ug/L) Legend
ZA1.1 0 7440 96% Inyectedand[TPH]
ZA1.2 148 1188 80% reduction
ZA1.3 0 1329 76%
Inyectedand[TPH]
ZA1.4 126,33 23520 99%
increase
ZA1.5 0 1329 63%
ZA2.1 0 1869 76% NoInyectedand
ZA2.2 148 54264 98% [TPH]reduction
ZA2.3 0 45155 99% NoInyectedand
ZA2.4 150 43210 92% [TPH]increase
ZA2.5 1442 12006 92%
ZA2.6 606 1074 72%
ZA2.7 0 20808 97%
ZA3.1 80 57 4%
ZA3.2 0 689 34%
ZA3.3 140 1280 62%
ZA3.4 392 1778 64%
ZA3.5 0 652 34%
ZA4.1 88 712 32%
ZA4.2 577 444 42%
ZA4.3 689 10 1%
ZA4.4 166 174 11%
ZA4.5 0 403 21%
ZA5.1 170 287 21%
ZA5.2 0 530 18%
ZA5.3 0 606 28%
ZA5.4 73 933 40%
ZA5.5 138 274 28%
I1 228 224 32%
S21 0 1265 32%
S22 294 9939 86%
S23 0 1380 63%
S24 10 286 47%
S15 0 532 93%
S26 605 20924 91%
S27 227 1250 45%
S28 110 1320 183%
Basedonperformanceresults,fourtypesofcasesweredistinguished:
WellswhereFentonreagentwasinjectedandTPHdissolvedconcentrationwasreduced(ZA1.2,ZA1.4,ZA2.2,ZA2.4,
ZA2.5,ZA2.6,ZA3.1,ZA3.3,ZA3.4,ZA4.1,ZA4.4,ZA5.4,S22,S26andS27).Thegreatestreductionoccuredinthose
wellswithahigherpreoperationalTPHconcentration,ZA2.4ZA1.4.Thepresenceofbyproducts(alcohols,aldehydes
andketone),absentspriortothestartofinjection,indicatedthatthedecreasewasduetohydrocarbonsoxidation
andnottoadilutioneffectresultingfromtheinjectionofreagentsolutions.
WellswhereFentonreagentwereinjectedbutTPHconcentrationincreased(ZA4.2,ZA4.3,ZA5.1,ZA5.5,I1,S24and
S28),phenomenonknownas"rebound"(inducedbyTPHdesorptionfromsoilorsolubilizationofFLSN).Presenceof
byproductsconfirmedthehydrocarbonsoxidation.However,TPHincreasewouldindicatesoilareasuntreated.
WellswhereFenton'sreagentwasntinjectedbutTPHconcentrationdecreased(ZA1.1,ZA1.3,ZA1.5,2.1,ZA2.3,
ZA2.7,ZA3.2,ZA3.5,ZA4.5,ZA5.2,ZA5.3,S21,S23,PC4).Thisfinding,coupledwiththebyproductpresenceinadjacent
wells,demonstratedthecorrectreagentdistributionthroughthesubsurface.
WellswhereFentonreagentwasntinjectedandtheconcentrationofTPHincreased(S15).Thissituationis
attributedtothesamephenomenonofsolubilizationordesorptionfromuntreatedareas.
Conclusions
Aneffectivesitecharacterizationthatallowsthedevelopmentofavalidconceptualsitemodel(CSM)isessentialfor
thedesignofISCOtechnologyandtodeterminetreatmentgoalsandoperatingconditions.Aconceptualsitemodel
shouldincorporatethelocationandmassofcontaminantofconcern,anunderstandingofthepedology,geologyand
hydrogeology,aquifergeochemistry, major migration pathways for thecontaminants of concern,groundwater flow
direction/gradient,andtheidentificationofsurfaceandsubsurfacestructuresandundergroundutilities.
In this study area, a deep knowledge of site conditions (pedologic, geochemical and lithologic characteristics of the
site, flow and mass transport) and its resulting conceptual site model (CSM) were fundamental tools to ensure the
successofISCO.
The injection program has been based on two strategies; the use of hydrogen peroxide stabilizers and the special
oxidantdeliverystrategy(targetlevelisolationandpressureinjection).Bothstrategiesweredesignedinordertosolve
somelimitationsthatcouldhaveendangeredthecontactbetweentheoxidantandTPH.Theselimitationsare,onthe
onehand,hydrogeochemicalconditionsthatlimitthepersistenceoftheoxidantinthesubsoilbyquicklyreactingto
thisviacatalysis(highpHandhighcontentofanionscarbonates,sulphatesandchlorides),andontheotherhand,
hydraulicconditionsthathinderthedistributionofoxidantinthebasement(highheterogeneityofsoilhorizons,low
permeabilityofthetargethorizonandshallow,spatialandtemporalvariabilityofthewatertable).
1 1, 1 1 1
Tamar Varazi , Maritsa Kurashvili , Marine Pruidze , Gia Khatisashvili , George Adamia , Nino
1 1 1 2
Gagelidze , Marlen Gordeziani , George Zaalishvili , Mark Sutton
1
Georgian Agricultural University, Durmishidze Institute of Biochemistry and Biotechnology
th
David Agmasheneblis kheivani, 240 (13 km), Tbilisi, 0159, Georgia. t.varazi@agruni.edu.ge,
Phone: +995 595 790 300
2
Chemical Sciences Division, Lawrence Livermore National Laboratory, California 94550 USA.
sutton18@llnl.gov, Phone: +001 925 424-2137
INTRODUCTION
Nowadays global chemical pollution of the ecosystems biosphere, waters, and soil tops the hazardous
level. Succesfylly and ecologically friendly method of cleansing of environment is a phytoremediation
selection and purposeful planting of those plants, which roots provide to uptake and enzymes to transform
the molecules of toxicants from environment.
Phytoremediation in comparison with other nonbiological and biological technologies has great
advantages. This advantage is manifested in the following aspects:
Is unique to chemical pollutants
Is unique to polluted objects
Ecologilally friendly
Cost-effective
Phytoremediation provides the control of erosion of soils
Phytoremediation provides design of landscape.
Here should be pointed shortcomings of phytoremediation Very high concentrations of pollutants e.g. at
oil accidental pouring, may significantly reduce greenery growth. Therefore, toxic compounds are more
rapidly disseminated from the pollution hotbeds into the environment (e. g. groundwater), than plants
provide their assimilation.
Nowadays is very actually to use such complex approaches, which provides effectiveness and successful
of remediation technologies.
The goal of the presented job is to elaborate the new technological approach of phytoremediation, based
on application different tools for improving of this technology:
detoxification abilities of microorganisms,
plants with high assimilation ability,
Natural biosurfactant for enhancing long chain hydrocarbons assimilation by plants.
Chelating agents for enhancing process of heavy metal uptake by plants.
Natural sorbents for avoid deep contamination of oil.
Attention have been paid to enhancement of soil fertility, which improves microorganism and plant growth
conditions and is important for bioremediation processes providing soil total rehabilitation According to
presented work, was used contaminants: crude oil, wax and heavy metal Pb.
MATERIALS AND METHODS
Microorganisms
Microorganisms of different taxonomic groups have been screened on solid and liquid areas contained
crude oil and wax. Microorganism collection cultures of Durmishidze Institute of Biochemistry and
Biotechnology as well as strains freshly isolated from polluted soils have been tested. After incubation in
cultivation area the concentration of residual pollutants have been estimated due to the gravimetric,
chromatographic and spectrophotometric methods. The criteria of strains selection at screening have
been:
Colony production ability.
Biomass accumulation potential.
Pollutants degradation ability.
Ability to use carbon atoms of toxic compounds as the only carbon source.
Plants
The following plant species were tested as vegetation to be used for purposeful planting during our
investigation: alfalfa (Medicago sativa) and soybean (Glycine max).
th
The plants were cultivated by the hydroponic method, on the running water. On the 5 day the seedlings
2+
were placed on lead (II) acetate solution of (concentration of Pb 400 mg/l) and after 5 days the lead
content was measured in different plant organs. Lead contents was measured in air-dried overground
parts and roots of tested plants, and the bioaccumulation coefficient (ratio between concentration of lead
in the plant tissue and the initial lead-containing nutrient solution) was determined by atomic-absorption
method at 283 nm.
Investigation of sorbents
Standard methods were applied to define physico-chemical features (volumetric weight, porosity,
displacement, filtration characteristics, oil absorption potential, absorption dependence on temperature,
pH, etc) of rocks [8].
Lead
Analysis of lead (II) content in plant and water samples was determined by atomic-absorption
spectroscopy [13].
In test objects pollutants concentration will be assayed periodically and soil purification level have been
stated. Tests have been carried out under different conditions (temperature, pH, salt concentration,
introduction of mineral and organic fertilizers etc).
During the test duration periodically, will be determined pollutant concentration in soil, deviations of
pollution rates and then the efficiency of phytoremediation will be estimated.
Degree of hydrocarbons utilization determined by formula:
= 100% / ()
degree of utilization, %;
contamination degree after wax introducing, ppm;
degree of contamination by the end of phytoremediation, ppm.
Fig. 2. Model experiment #2: Phytoremediation of soil artificially contaminated with crude oil and wax by
using of soybean, bacterial consortium (Pseudomonas sp. 6R67 + Bacillus spp.) and preparation of BB2
(complex of biosurfactants produced by strain of Baccilus spp., concentration 0.01%). Initial
contamination of soil by TPH - 117 000 ppm, by wax 58 500 ppm. Duration of experiment 3 months.
The effect of soil cleaning by presented technology equal 78-92% by model experiment # 1 and 85-92%
by model experiment #2 It has been established that soybean and alfalfa together with selected bacterial
consortium and natural biosurfactant are the best tools for phytoremediation of soils, polluted with oil
hydrocarbons. Bacterial consortium consisting Rhodococcus sp. MO227 and Pseudomonas sp. TP 335 is
best remediation agents among tested microorganisms.
The chromatographic analysis of soil samples (Fig. 3) show that in case of 43 000 ppm initial
contamination in the beginning of remediation the ratio between light and heavy petroleum
hydrocarbons (C8-15 : C16-35) decreased from 87 : 13 (Fig. 3A) to 6 : 94 (Fig. 3B). These results indicate
that the assimilation of C8-15 hydrocarbons take place mostly during bioremediation, than
phytoremediation.
Fig. 3. Chromatograms of hexane extracts of soil samples at the beginning (A) and after (B) remediation
process. Contaminated soil (initial contamination 43 000 ppm of TPH) is treated by bacterial consortium
(Rhodococcus sp. MO227 and Pseudomonas sp. TP 335). Duration of remediation: 2 months;
temperature: 20-25.
Absorbed lead was basically accumulated in roots. At presence of complexon EDTA in nutrient solution
the heavy metal content in over ground parts of plants was increased. This effect is especially important in
phytoextraction technologies.
CONCLUSION
Bacterial consortium consisted with Rhodococcus sp. MO227 and Pseudomonas sp. TP335 appeared to
be the best remediation agent among tested microorganisms and their consortia.In soils contaminated as
a result of oil pipeline spills (levels of contamination: from 43 000 to 117 000 ppm) after treatment by this
consortium the contamination is decreased by 78-92 %.
The assimilation of C8-15 hydrocarbons takes place mostly during bioremediation.
It has been established that soybean and alfalfa together with selected bacterial consortium and natural
biosurfactant are the best tools for phytoremediation of soils, polluted with oil hydrocarbons.
Lead basically accumulated in plant roots. At presence of complexon EDTA in nutrient solution the lead
content in over ground parts of plants was increased. This effect is especially important in phytoextraction
technologies.
ACKNOWLEDGMENTS
This work was supported by STCU Partner Project # P499. In addition, Lawrence Livermore National
Laboratory is operated by Lawrence Livermore National Security, LLC, for the U.S. Department of Energy,
National Nuclear Security Administration under Contract DE-AC52-07NA27344.
REFERENCES
1. Korte, F., Behadir, M., Klein, W., Lay, J.P., Parlar,H., Scheunert, I. Lehrbuch der okologischen
chemie. Grundlagen and Konzepte fur die Okologische Beureilung von Chemikalien. Suttgart, Georg
Thieme Verlag, 1992.
2. Arthur, E.L., Coats,J.R. Phytoremediation. In: Pesticide Remediation in Soils and Water. P. Kearney
and T. Roberts. Eds. New York, Wiley, 1998.
3. Salt, D.E., Blaylock, M., Nanda Kumar, P.B.A., Dushenkov, V.P., Ensley, B.D., Chet, I., Raskin, I.
Phytoremediation: a novel strategy for the environment using plants, Biotechnology, vol. 13, 1998,
pp. 468-474.
4. Tsao, D.T. Phytoremediation. Advances in Biochemical Engineering and Biotechnology. Berlin
Heidelberg New York, Springer, 2003.
5. Kvesitadze, G., Khatisashvili, G., Sadunishvili, T., Ramsden, J.J. Biochemical Mechanisms of
Detoxification in Higher Plants. Basis of Phytoremediation. Berlin Heidelberg New York, Springer,
2006.
6. AS 482.2 Guide to the sampling and investigation of potentially contaminated soils,1999.
7. Schnoor, Dee, 1997. Phytoremediation. Technology Evaluation Report TE-98-01. Ground-Water
Remediation Technologies Analysis Center. Ser E. Iowa City
8. Fomin, G.S., Fomin, A.G. Soil. Control of quantity and ecological safety according to international
standards. Moscow. BNII Standard, 2001.
9. NWTPH-HCID Hydrocarbon Identification Method for Soil and Water. 2001.
10. US EPA Method 3540C Soxhlet Extraction; Method 8572A. Semivolatile organic compounds in soils
and solid wastes using thermal extraction/gas chromatography/mass spectrometry,1996.
11. US EPA Method 418.1. Total petroleum hydrocarbons, 1997.
12. TNRCC method 1006. haracterization of C6-C35 petroleum hydrocarbons in environmental
samples, 2000.
13. Polyanskyi N.G. Analytical Chemistry of Elements. Lead. Moscow, Nauka, 1986.
14. Geyer G. Ultrahistochemie, 2-nd edn. Jena: Gustav Fisxher 1973.
Topic: 1C. Remediation technologies and approaches
Department of Chemical Engineering, Technical University of Catalonia (UPC), ETSEIB, Av. Diagonal,
1
1. INTRODUCTION
Biomass from wetlands is a natural source of organic material able to decrease nitrate present in
groundwater before the discharge to rivers [1]. The desired predominant mechanism of nitrate removal is
heterotrophic denitrification by supplying organic carbon that, additionally, enhances anoxic conditions in
subsurface environment. The stoichiometric reaction of a complete heterotrophic denitrification could be
described as eq. 1 [2].
- -
5CH2O + 4NO3 2N2(g)+ 4HCO3 + 3H2O + CO2(g) Eq. 1
Nevertheless, an excess of organic matter (CH2O) vs nitrate could activate other mechanisms as
Dissimilatory Nitrate Reduction to Ammonia (DNRA), that could be formulated as
- + -
2CH2O + NO3 + H2O NH4 + 2HCO3 Eq.2.
DNRA can be significant or even a dominant process in some ecosystems [3]. DNRA is a respiratory or
fermentative pathway where nitrate is slowly reduced to ammonium competing with denitrification for the
nitrate into riparian zones [3], [4].
Conditions favouring DNRA are less understood than denitrification although It is believed that
heterotrophic denitrification supplies more free energy than DNRA. Under nitrate limiting conditions or
high organic matter concentrations. However, DNRA could be favoured because more electrons can be
transferred per mole of nitrate, [6]. Nitrite could appear an intermediate in both, denitrification and DNRA.
The leaching of organic matter rich soils (peat or Horizon O layer) could be a mechanism to promote
denitrification in deeper subsurface layers as provides organic matter avoiding high concentrations of
organic matter that could promote DNRA pathways.
In the present work, the study at laboratory scale of the denitrification capacity of leachates of biomass
and wetland soils linking with the growing of Phragmites sp and Arundo donax reeds have been studied.
A close understanding of nitrate removal by using these materials ?has to improve knowledge of strengths
and weaknesses of Mediterranean wetlands to remediate nitrate contamination.
Samples of soils and degraded reed biomass were collected in two sites in the Llobregat river basin
(Barcelona, Spain) in February and May 2014 (figure 1 a) and b)).
c) Sample C2 d) Sample A2
In the first site three types of samples were taken. These samples consisted in degraded reed on the soil
in the middle of the reed field (February 2014) (C1) and degraded reed close to the riparian zone (C2)
(figure 1 c), both sampled in February. In May a sample of green leaves of Arundo donax was taken (CV).
In the second site a sample of degraded leaves of Phragmites sp. on the soil (A2, figure 1d) and degraded
stem of Phragmites sp.(B1) were taken.
2.2. Leachates
Leachates were obtained by using different amounts of the described materials, deionised water and
different leaching methods that are summarized in table 1. The three leaching procedures were:
percolation in a column filled with material, shaking with water and lixiviation by using a vertical stirrer
leaching test were tested to extract the maximum of Dissolved Organic Carbon (DOC) at room
temperature from the biomass samples.
2.3. Inoculum
Samples of soil in C1 and A2 were also taken in both sites and were used for inoculum extraction. The
procedure of extraction consisted in using an amber glass bottle sterilized by using acetone and left 24
hours at 100 C. After cooling the bottle at ambient temperature, 250 grams of wet soil were mixed with
500 ml of deionised water and were mixed in the bottle for 24 hours by using a vertical stirrer. The mix
was left settle down overnight and 250 ml of supernatant solution were separated by decantation and the
use of a plastic pipette. This volume was centrifuged (Centronic-BL II, Selecta) at 3500 rpm for 10
minutes. The Supernatant of the centrifuged tubes was discarded and the samples were centrifuged two
or three times to get pellets. These pellets were mixed in a tube that was centrifuged.
1 ml of pellet was mixed with 29 ml of culture medium. This culture medium consisted in a dissolution of
-3 - -3 -3 -3
100 mgdm NO3 , 2,1 gdm K2HPO4, 1 mgdm FeCl36 H20 and 200 mg/ dm of glucose. This mixture
was stirred 48 hours at room temperature (20-25C) and the obtained suspension was used for batch
experiments.
Batch tests were performed at room temperature (20-25C) in amber bottles previously sterilized with
acetone and heating at 100 C. 10 ml of inoculum were mixed to a final volume of 125 ml using the
- -3
leachate and a nitrate solution to reach a concentration of 25 mg NO3 dm . Some experiments were
performed using a dissolution of 200 mg/l glucose (DOC 80 mg/l) instead the leachates to test the correct
setup.
Blanks of inoculum (10 ml inoculum diluted to 125 ml with deionised water) and controls of TOC and TN
-3 -3
(mixtures of 200 mgdm glucose and 25 mgdm nitrate without inoculum) were also used. In all these
mixtures nitrogen gas was used to purge dissolved oxygen. The bottles were filled completely to avoid
head space ad were closed to avoid the air to enter in the experiment. After some days of experiment all
the samples were filtered at 0.45 m before analysis.
Nitrate, nitrite and ammonium were analysed by ion chromatography (Dionex ICS-1000/ICS-1100) with
ionic suppressor conductivity. The system was equipped with an Ionpac AS19 column (Dionex) and was
operated with 10-45 mM KOH mobile phase. Ammonium was analysed by means of the same ion
chromatography system equipped with an Ionpac CS16 column (Dionex) and using methanosulphonic
-3 + - -3
acid (30mM) as mobile phase. The detection limit was 0,01 mgdm for NH4 and NO2 and 0,02 mgdm
-
for NO3 .
DOC content (as Non-Purgeable Organic Carbon, NPOC) was determined by means of a Shimadzu TOC-
V CPH equipped with a carrousel ASI-V. Total N was determined by Shimadzu a TNM-1 module analyser
connected to TOC equipment. Samples were filtered through 0.45 m nylon membrane filters and
acidified with HCl solution to a pH lower than 2. Before analysis samples were purged during 20 minutes
-
in order to volatilise all the inorganic carbon present. The detection limit of the method was 1 mg DOCdm
3
.
HP 8453 UV-visible Spectrophotometer (Hewlett-Packard) was used to measure the absorbance signal at
254 nm. A Crison pH-meter was used to measure the pH of the dissolution.
3. RESULTS AND DISCUSSION
-3 -3 -3
Blanks of inoculum showed average values of 2,7 mgdm of nitrate, 0,3 mgdm nitrite, 0,2 mgdm
-3
ammonium and 5 mg DOCdm . Table 2 summarizes the characterization of leachates
-3
DOC in the materials of experiments 1 and 2 ranged from 10 to 31 mgdm . The ratio of the UV signals at
254 nm divided by DOC was also used to assess quantitatively the presence of aromatic compounds in
the leachates, which could be important in samples of soils. This ratios showed values from 0,03 to 0,06
3 3
A.Udm /mg that are in the range of fulvic acid (0.05 A.U.dm /mg). [7]
-3
Nitrate in the leachates from all the biomass showed values above 20 mgdm , with the exception of
-3
degraded Arundo donax in soil very close to the river that exhibit values below 2 mgdm .
A third leaching experiment was prepared by shredding green Arundo donax leaves sampled in May
-3 -3
2014. DOC of these filtered leachates reached 523 mgdm and total nitrogen 49.8 mgdm . These
values of nitrogen could correspond to organic forms from chlorophyll.
Comparing C2 samples, it was observed that the samples obtained by vertical stirrer obtained more
concentration of dissolved organic matter in the lixiviates.
a) Glucose Experiment 1 b) B1 Experiment 1
c) C2 Experiment 1 d) A2 Experiment 2
e) C2 Experiment 2 f) C1 Experiment 2
g) C2 Experiment 3 h) CV Experiment 3
All these characterized leachates were used as matrix in batch experiments 1 to 3. The reaction was
performed for a period of 6 to 11 days and results were compared with initial N species at T=0 (figure 2).
-3
Parallel experiments, where DOC was replaced by 100 mgdm glucose, showed complete denitrification
for all experiments after a period of 6 days or more (see figure 2a).
In the case of samples C2 for different experiments the results (figure 2, left samples) showed no nitrate
reduction when t=6 days and very slight reduction at 7 and 11 days. Sample B1 in experiment 1 showed
no denitrification after 6 days and sample C1 in experiment 2 showed also slight nitrate reduction.
Degraded leaves of Phragmites sp. on the soil (A2) and the green Arundo donax leaves (CV) showed
more than 50% and complete elimination of nitrate respectively. In all the samples (except in CV) final
ammonium decreased when compared with initial values (from the lixiviates and inoculum). This means
that DNRA is present in the experiment with CV, probably because samples were not filtered during the
batch experiment and level of DOC was very high, conditions that impose a higher reduction potential able
to promote ammonium instead nitrogen gases. [6]
4. CONCLUSIONS
Most of the leachates of biomass materials and soils sampled in February in Llobregat river basin sites
were not able to leach important amounts of DOC and could not denitrify or reduce nitrate. Soil samples
showed the presence of UV-absorbing compounds that could be associated to complex organic matter as
fulvic acids. These types of compounds are formed by complex organic matter that cannot participate in
the reduction of nitrates. Degraded leaves of Phragmites sp. of the soil (A2) were able to eliminate 50% of
the initial nitrate.
Arundo donax green leaves sampled in May (CV) allowed obtaining a leachate with high DOC able to
perform nitrate reduction but with a partial DNRA mechanism. In these vegetal materials initial nitrogen is
important and is associated to chlorophyll. This material and some soils with degraded reeds could
contribute also to a small load of nitrate or release ammonium.
5. ACKNOWLEDGEMENTS
The present research work has been funded by the Spanish Ministry of Economy and Competitivity
through project ATTENUATION (CGL2011-29975-C04-03, Natural and Induced Attenuation of
groundwater pollution from agricultural and industrial sources).
6. REFERENCES
[1] Bosch C., Calderer M., Jubany I., Mart V., Gibert O., Springael D., Aamand J., Natural attenuation
potential of a groundwater-fed Danish Wetland Impacted by diffuse nitrate contamination, Pster 11th
International UFZ-Deltares/TNO Conference, on Management of Soil, Groundwater and Sediments,
Salzburg, ustria, 2010
[2] Jrgensen P. R., Urup J., Helstrup T., Jensen M. B., Eiland F., Vinther F. P.. Transport and reduction
of nitrate in clayey till underneath forest and arable land. Journal of Contaminant Hydrology, 73 (2004),
207-226.
[3] Rtting T., Boeckx P., Mller C. and Klemedtsson L., Assessment of the importance of dissimilatory
nitrate reduction to ammonium for the terrestrial nitrogen cycle, Biogeosciences Discuss., 8 (2011) 1169-
1196.
[4] Calderer M., Gibert O., Mart V., Rovira M., de Pablo J., Jordana S., Duro L., Guimer J. and Bruno J.,
Denitrification in presence of acetate and glucose for bioremediation of nitrate-contaminated groundwater,
Environ. Technol., 31 (2010) 799-814.
[5] Sgouridis, F., Heppell, C. M., Wharton, G., Lansdown, K., & Trimmer, M.. Denitrification and
dissimilatory nitrate reduction to ammonium (DNRA) in a temperate re-connected floodplain. Water
research, 45 (2011), 4909-4922.
[6] Krafta B., Strousa M. and Tegetmeyera H. E., Microbial nitrate respiration Genes, enzymes and
environmental distribution, J. Biotechnol., 155 (2011) 104-117.
[7] American Public Health Association (APHA). 1998. Standard methods for the examination of water and
wastewater, 20th Edition. American Public Health Association , American Water Works Association and
Water Environment Federation. Washington, DC, USA. p. 5-73.
1c. Remediation technologies and approaches
Introduction
Chromium content in soils is largely dependent on parent materials. Anthropogenic Cr sources include
industrial processes such as electroplating, leather tanning and mineral extraction; besides, Cr is a
common ingredient used in protective coatings, pigments, and stainless steel (Choppala et al., 2013).
Chromium trivalent (Cr(III)) and hexavalent (Cr(VI)) are predominant in the environment (Adriano,
2001). Chromium(III) is a micronutrient for sugar and lipid metabolism in humans (Lukaski, 1999). On
the other hand, Cr(VI) is highly toxic, teratogenic and mutagenic (De Flora, 2000; Costa and Klein,
2006). Regarding plants, the two species of Cr are nonessential and toxic to the plant at cellular levels
(Zayed and Terry, 2003). Since Cr(VI) poses a great threat to humans and plants, cleaning up this
contaminant from soil and water is critical. In this study a greenhouse experiment was performed to
evaluate the effectiveness of nZVI to immobilize Cr(VI) in a contaminated soil as well as the stability of
the immobilization process after a barley crop. The impact of the nanoremediation process on the
develop of the barley plants and on soil properties was also analyzed.
Soil
Soil samples (0-30 cm) from an agricultural field located in Alcal de Henares (Madrid). The soil
samples were firstly air-dried and then sieved (<4 mm). The soil properties were analyzed according to
the Spanish official methodology for soil analysis (MAPA 1994a). In brief, electrical conductivity (EC)
and pH were measured in a 1:2.5 soil-to-water ratio; the organic matter (OM) and total nitrogen
contents were determined using the WalkleyBlack and Kjeldahl methods, respectively; the
percentage of carbonates was measured using a calcimeter; the available phosphorus was evaluated
using sodium bicarbonate at pH 8.5 as an extractant; and available nutrients were extracted with 0.1 N
NH4Ac and assessed using flame atomic absorption spectrometry (FAAS) (AA240FS, Varian, Victoria,
Australia). The soil texture was analyzed using a Bouyoucos densimeter.
Experimental design
Soil samples were spiked with K2Cr2O7 (Panreac, Barcelona, Spain) at two doses of Cr, 73 and 230
mg/kg, soil moisture was brought to 60% of field capacity and kept for 30 days to soil consolidate. Pots
were filled with polluted soil (3.8 kg) and treated with nZVI solution (5%). Four pots per condition were
included. The samples were called Cr-1, Cr-2 (untreated soil samples), Cr-1-NP, Cr-2-NP (nZVI
treated soil samples). The commercial nZVI NANOFER 25S from NANO IRON (Rajhrad, Czech
Republic) was used. After 72 h of interaction soil-nZVI, barley plants (Hordeum vulgare L., cultivar
CB502) at stage 14 according to the Zadoks scale (Zadoks et al., 1974), were transplanted into the
pots (two plants per pot). The barley plants were maintained in a greenhouse and harvested after four
months, when they had completed their growing period. Plant height and dry weight were measured.
Dried grains, shoots and roots were powdered, weighted and mineralized according to the Spanish
official methodology for plant analysis (MAPA 1994b). Briefly, plant samples were ashed at 580 C
overnight; the ashes were digested with a mixture of HCl 37% (1 mL) and Milli-Q water (5 mL) for 30
minutes in boiling, and then filtrated and diluted with Milli-Q water. Chromium, Fe and macronutrients
were measured in these extracts by FAAS (AA 240 FS).
Soil samples were collected after 72 h interaction soil-nZVI and after plants harvest and were
analyzed according to a classic sequential extraction procedure (Tessier et al., 1979) to determine the
effectiveness and the stability of the Cr immobilization using nZVI. Extractions with solutions of
increasing strengths were sequentially added to the soil sample. The fraction and the extractant are as
follow: EX, exchangeable (1 M MgCl2, pH 7); CB, bound to carbonates (1 M CH3COONa, pH 5);
bound to Fe and Mn oxides (NH2OH.HCl 0.04 M in acetic acid 25%); OM, bound to organic matter
(30% H2O2 + HNO3 0.01 M); RS, residual fraction, obtained after acid digestion of the soil in a
microwave reaction system (Multiwave 3000), using HNO3 (69% purity) + HCl (37% purity).
The physico-chemical properties of soils from pot experiment were also analyzed following the
methodology previously described. The Fe in EX fraction using Tessiers procedure was also
determined to evaluate the impact of nZVI treatment on Fe availability in soils.
Statistical analyses
Statistical analysis of data was made using the SPSS statistical package for Windows, release
19.0.0.1 (SPSS Inc., IBM Company). One-way analysis of variance (ANOVA) and Tukeys test at a
significance level of p<0.05 were used.
Table 1 lists the mean values of the main physico-chemical properties of the soil used in the present
study. It is a loam soil with alkaline pH mainly due to the high carbonate percentage (8.6%), moderate
content in organic matter (OM) and low levels of nitrogen and phosphorous.
Immobilization results
Figure 1 shows the results of the sequential extraction procedure for soil samples collected after 72
hours of interaction soil-nZVI and after the plants harvest (4 months). The application of nZVI resulted
effective for the Cr immobilization at the experimental conditions, because of a significant reduction of
Cr in the most available fraction and a concomitant increase of Cr in less available fractions (OX) was
observed. In soil samples collected after 72 h, the Cr associated to EX fraction decreased nearly 90%
and the OX fraction increased from 2 to 48 mg/kg at the low dose and from 12 to 163 mg/kg at the
high dose of contamination. The Cr in OM fraction increased nearly 50% for both doses of
contamination. However, no significant changes were observed in RS fraction after the application of
nZVI. The interaction mechanism between nZVI and Cr(VI) could involve a reduction of Cr(VI) to
Cr(III), which shows lower solubility, as previous studies performed in aqueous medium have
concluded by high-resolution X-ray photoelectron spectroscopy (Li et al., 2008). The strong increase
of Cr in OX fraction might be attributed to the precipitation of Cr(OH)3 or Cr(III) with oxides/hydroxides
of iron (Li et al., 2008; Di Palma et al., 2014; Wang et al., 2014). Comparing with the soil samples
collected after plants harvest, untreated soil samples showed much lower Cr in EX fraction than
untreated ones collected after 72 h, which can be due to losses from leaching phenomena favored by
continuous irrigation during the barley crop. The OX fraction was also quantitatively the most important
fraction in samples collected at longer time. Wang et al. (2014), in a experiment with Cr-contaminated
soil and treated with nZVI, also found that after the treatment, the Cr in EX fraction was completely
converted to the OX and OM fractions.
Impact on barley plants
Figure 2 shows the barley plants grown in Cr-contaminated soils treated and untreated with nZVI at
the end of the experiment. The plants from Cr-polluted soils quickly showed symptoms of toxicity; at
the highest Cr dose, the plants died before one week. In contrast, the plants grown in soil treated with
nZVI did not show symptoms of toxicity and they completed their growing period. This is due to the
reduction of Cr(VI) by nZVI to Cr(III) which precipitates, reducing Cr availability, thus, soil phytotoxicity.
Plants from Cr-1-NP and Cr-2-NP showed similar height; however, the dry weight for root, shoot and
grain was higher in plants from the lowest dose of Cr. Table 2 lists the mean concentration of Cr in the
different part of the plant (root, shoot and grain) according to the treatment. Plants from Cr-2 have not
been included because they did not grow and died before one week. The application of nZVI
significantly reduced the accumulation of Cr both in root and shoot because of the decrease of Cr
bioavailability in soil. In fact at the low dose of Cr, a reduction in Cr uptake nearly of 97% was obtained
after the treatment with nZVI. Chromium was mainly accumulated in root. The Cr distribution in plants
from nZVI-treated soils was similar, regardless the degree of contamination. Thus, the applied dose of
nZVI was effective in a wide range of Cr contamination (73 to 230 mg/kg).
The results obtained for the concentration of Fe and macronutrients (Ca, Mg, Na and K) in the barley
plants are shown in Figure 3. Plants from nZVI-treated soils showed a similar behavior for all this
elements, regardless the dose of Cr contamination. Iron was mainly accumulated in root; the treatment
with nZVI did not produce a higher concentration of this element in plant tissue. In fact, the
concentrations of Ca, Mg and Fe were higher in plants from untreated soils, although the total content
per pot was higher in plants from nZVI-treated soils. In this sense, Wang et al. (2014) found that the
Fe uptake by two plants species (Rape and Chinese cabbage) was significantly reduced after 72 h of
interaction soil-nZVI. As previously explained in Materials and Methods section in the present
experiment, the barley plants were transplanted in the nZVI-treated soils after 72 h of interaction soil-
nZVI. In general, the application of nZVI did not affect negatively to the accumulation of Fe, Ca, Mg,
Na and K in barley plants.
Different physico-chemical properties were analyzed in the soil samples collected after plant harvest
(Table 3) to evaluate the impact of the nanoremediation treatment on soil properties. No negative
effects were observed in soils after the application of nZVI at the assayed experimental conditions.
Treated soils showed pH values slightly higher than those found in untreated ones. It can be due to
the high pH of the commercial nanoparticle suspension (pH=12, Gil-Daz et at., 2014a, 2014b).
Regarding available P, nZVI-treated soils showed slightly lower content that untreated ones, which
3-
can be explained by the fact that nZVI can reduce the availability of PO4 , as other studies have
concluded (Liu et al., 2013; Wu et al., 2013). The Fe content in EX fraction was also included to
evaluate if the treatment with nZVI might increase the Fe availability in soil. However, an increase of
Fe in this soil fraction in treated samples was not observed, in fact, untreated soil samples showed
slightly higher concentration of EX Fe, probably due to the lower pH values. Thus, the application of
nZVI did not produce an increase of the availability of Fe at the experimental conditions.
Conclusions
The present study demonstrated that stabilized nZVI can be used for the immobilization of Cr(VI) in an
alkaline soil in a wide range of Cr-contamination (73-230 mg/kg). The decrease of the Cr availability
conduced to a reduction of soil phytotoxicity which led a better development of the barley plants, even
though they completed their growing period. Consequently, the Cr uptake by the barley plants was
strongly reduced, nearly of 97% for the low dose of Cr. The application of nZVI on soil did not induce a
increase of Fe concentration on plants. An increase of Fe availability in soils was not observed either.
The barley plants from nZVI-treated soils showed a similar behaviour for the content of Fe, Ca, Mg,
Na, K. No negative effects on physico-chemical soil properties were observed after the treatment with
stabilized nZVI at the experimental conditions. Available P was slightly reduced in nZVI-treated soils
due to immobilization processes. Taking into account these results we can concluded that the use of
nZVI is a promising strategy to remediate Cr polluted soils, and barley plants could be grown in these
nZVI-treated soils.
Acknowledgements.
The authors thank IMIDRA (Community of Madrid) for supporting Project FP 13 CONT, and Ministerio
de Educacin y Ciencia for supporting Project CTM 2010-20617-C02-02.
References
Adriano DC. Trace elements in terrestrial environments: Biogeomistry, bioavailability and risk of
metals. 2nd Edition. Springer-Verlang New York Berlin Heidelberg 2001.
Choppala G, Bolan N, Park JH. Chromium Contamination and Its Risk Management in Complex
Environmental Settings. Adv Agron 2013, 120, 129-172.
Costa M, Klein CB. Toxicity and carcinogenicity of chromium compounds in humans. CRC Crit Rev
Toxicol 2006, 36, 155163.
Gil-Daz M, Alonso J, Rodrguez-Valds E, Pinilla P, Lobo MC. Reducing the mobility of arsenic in
brownfield soil using stabilised zero-valent iron nanoparticles. J. Environ Sci Health A 2014b, 49,
1361369.
Gil-Daz M, Prez-Sanz A, Vicente MA, Lobo MC. Immobilisation of Pb and Zn in Soils Using
Stabilised Zero-valent Iron Nanoparticles: Effects on Soil Properties. Clean-Soil, Air, Water 2014a, 42,
1776-1784.
Gonzlez V, Garca I, Del Moral F, Simn M. Effectiveness of amendments on the spread and
phytotoxicity of contaminants in metalarsenic polluted soil. J Hazard Mater 2012, 205-206, 72-80.
Karn B, Kuiken T, Otto M. Nanotechnology and in Situ Remediation: A Review of the Benefits and
Potential Risks. Environ Health Perspect 2009, 117, 823-1831.
Li X, Cao J, Zhang W. Stoichiometry of Cr(VI) Immobilization Using Nanoscale Zerovalent Iron (nZVI):
A Study with High-Resolution X-Ray Photoelectron Spectroscopy (HR-XPS). Ind Eng Chem Res 2008,
47, 21312139.
Li X, Zhang W. Sequestration of Metal Cations with Zerovalent Iron NanoparticlesA Study with High
Resolution X-ray Photoelectron Spectroscopy (HR-XPS). J Phys Chem C 2007, 111, 6939-6946.
Liu H, Chen T, Zou X, Xie Q, Qing C, Chen D, Frost RL. Removal of phosphorus using NZVI derived
from reducing natural goethite. Chem Eng J 2013, 234, 8087.
Lukaski HC. Chromium as a supplement 1. Annu Rev Nutr 1999, 19, 279302.
MAPA 1994a. Mtodos Oficiales de Anlisis, vol. III, Secretara General Tcnica Ministerio de
Agricultura, Pesca y Alimentacin: Madrid; 1994. p. 219-324.
MAPA 1994b. Mtodos Oficiales de Anlisis, vol. III, Secretara General Tcnica Ministerio de
Agricultura, Pesca y Alimentacin: Madrid; 1994. p. 21-38.
Mueller NC, Braun J, Bruns J, ernk M, Rissing P, Rickerby D, Nowack B. Application of nanoscale
zero valent iron (nZVI) for groundwater remediation in Europe. Environ Sci Pollut Res 2012, 19: 550
558.
Nmeek J, Lhotsk O, Cajthaml T. Nanoscale zero-valent iron application for in situ reduction of
hexavalent chromium and its effects on indigenous microorganism populations. Sci Total Environ
2014, 485-486, 739747.
OCarroll D, Sleep B, Krol M, Boparai H, Kocur C. Nanoscale zero valent iron and bimetallic particles
for contaminated site remediation. Adv Water Resour 2013, 51, 104-122.
Singh R, Misra V, Singh RP. Removal of Cr(VI) by Nanoscale Zero-valent Iron (nZVI) From Soil
Contaminated with Tannery Wastes, Bull Environ Contam Toxicol 2012, 88, 210-214.
Tessier A, Campbell PGC, Bisson M. Oxides, hydroxides, and aluminosilicates. Sequential extraction
procedure for the speciation of particulate trace metals. Anal Chem 1979, 51, 844850.
Zayed AM, Terry N. Chromium in the environment: factors affecting biological remediation. Plant Soil
2003, 249, 139156.
Zhang MY, Wang Y, Zhao DY, Pan G. Immobilization of arsenic in soils by stabilized nanoscale zero-
valent iron, iron sulfide (FeS), and magnetite (Fe3O4) particles. Chinese Sci Bull 2010, 55, 365-372.
Table 1. Physico-chemical properties of the soil used in the pot experiment (meanstandard deviation,
n=3).
Characteristics value
pH 8.10.2
EC (dS/m) 0.260.01
CaCO3 (%) 8.60.9
N (%) 0.080.04
OM (%) 1.080.04
P (mg/kg) 233
Ca (mg/kg) 294664
Mg (mg/kg) 44621
Na (mg/kg) 9830
K (mg/kg) 40871
Cd (mg/kg) nd
Cr (mg/kg) 152.8
Zn (mg/kg) 6022
Fe (g/kg) 19.90.2
Fe oxides (mg/kg) 570251
Mn oxides (mg/kg) 18713
Al oxides (mg/kg) 52358
Sand (%) 461.5
Silt (%) 313.2
Clay (%) 232.2
Table 2. Mean concentration of Cr (mg/kg) in root, shoot and grain of barley plants.
Cr-1 8.22a 0.73a 0.11a 1.41b 26b 3264a 417a 181a 504a 0.75b
Cr-1-NP 8.47b 0.54a 0.11a 1.51b 19a 3206a 410a 222a 505a 0.22a
Cr-2 8.27a 0.44a 0.10a 1.22a 25b 3000a 398a 171a 513a 1.22b
Cr-2-NP 8.44b 0.42a 0.10a 1.51b 18a 2994a 385a 175a 525a 0.23a
(1) (2) (3) (4) -1
dS/m; %; mg/kg; Fe in EX fraction according to Tessier fractionation, mg kg . Different
letters mean significant differences (p<0.05; Tukey test).
Figure 1. Results of the sequential extraction procedure for soil samples collected after 72 hours of
interaction soil-nZVI (A) and after the plants harvest (B).
B
Figure 2. Photograph of barley plants at dose 1 (A) and dose 2 (B) of Cr two months later that they
were transplanted.
Cr-NP-1
Figure 3. Mean concentration (mg/kg) of Fe (A), Ca (B), Mg (C), Na (D) and K (E) in root, shoot and
grain of the barley plants according to the treatment.
THEME 1C
*URS Italy (AECOM Group) - Via Watt, 27 20143 Milan +39 024225561
**URS Spain (AECOM Group)- Calle Juan de Mariana 17B, 2 Planta 28045, Madrid: +34 915064730
Abstract
In 2014, the Italian Ministry of the Environment approved a groundwater remediation project designed by
URS based on the phyto-remediation approach.
The project must address high level of groundwater pollution surrounding an abandoned landfill.
Main contamination of groundwater is due to metals (such as Arsenic, Mercury, Beryllium and Selenium),
TPH, Aromatic Hydrocarbons, PAHs, Chlorobenzenes. Groundwater discharges into Mediterranean sea
or into a coastal lagoon.
URS predict that after two years the woods can reduce the pumped water in a period between 3 and 4
years, the woods will allow complete shutdown of the pump & treat system with real economic and
environmental benefits.
Groundwater circulates in silty-sand and silt aquifer, backed by a basal level of clay. Water table is few
meters below the ground surface and aquifer saturated thickness is more than 20 m. During rain seasons
the area in study is also subject to the formation of wetlands due to the outcrop of water table.
Groundwater flow and the hydraulic gradient are locally conditioned by the presence of the landfill (Figure
1), natural discharges are the Adriatic sea and the lagoon bordering the landfill.
Site assessment
Since the early 60s, the area was used as waste disposal of industrial productions. For several years here
were discharged sludge of calcium hydroxides, which forming an outcrop with 3m thickness and an huge
landfill (50 Ha).
Up to the early 70's in same area were discharged other residue from industrial processes, primarily
chlorinated and secondarily, aromatic hydrocarbons and PHA's. Since the mid-70s, stopped the
uncontrolled spill of chlorinated. but, for some years there were discharged highly acidic waste-waters
which, in contact with carbonate deposits, created large cracks and swallow holes
The absence of any protection of landfill has meant that rainwater has leached the contaminated soil and
heavily contaminated the groundwater surrounding the landfill.
The project have been carried out in southern Italy, and concerns one site extended about 94 hectares:
50 hectares of a large industrial landfill; and 44 hectares of uncontaminated land peripheral to the
landfill. The site falls in coastal area with lagoons and is near to Regional Nature Reserve which includes
a lagoon bordering the landfill.
The remedial strategy is focused on the implement of permanent safety measures for soils, waste and
groundwater, permanent safety measures are:
channel deviation;
capping of waste disposal;
physical confinement;
pump&treat;
phyto GW containment.
Rainfall and run-off on the surface of the capping will be channeled into a system of gutters, and then
discharged directly to the sea.
For this project Italian Environmental Ministry has requested to maintain water table inside the perimeter
of slurry wall, continuously lower than the water table outside the wall itself.
Therefore, contaminated groundwater within the system of physical confinement, will be pumped by 32
3
wells and sent to an appropriate treatment plant, water draft is estimated in 300 m /day.
Pump & treat will be integrated by the reforestation of area inside the physical confinement, in order to
allow the progressive reduction of groundwater recharge from rainfall taking advantage from natural
evapotranspiration of plant species.
The phyto-extraction takes into account that in regions where soils have moisture deficit throughout the
hydrologic year (such as the object site), the plants withdraw from the aquifer the water necessary to their
biological activity.
GW phyto -extraction may have many strengths: it is able to draw groundwater for the biological activities
of the plants; it has a low impact on the environment because it does not consume any power; CO2 can be
removed from the trees; it has low cost of installation and can allow to reduce O&M costs due to
groundwater treatment.
Selecting most suitable plants for areas object of study, three parameters have been considered: a) the
resistance to salinity, b) the resistance to water stagnation and c) the preference toward indigenous
species.
The project includes to use halophiles or salt-tolerant transitional halophiles, because their ability to
uptake salts in the soil, in mixture with species more tolerant of water stagnation.
For the structure of the forest (Figure 4), it was decided to use the willow (Salix fragilis / alba) and alder
(Alnus glutinosa). These colonizer plants are characterized by a very rapid growth, stand up well to both
water stagnation and pollution. Those plants are found all over Italian territory.
Other plants were combined to willow and alder, with slower growth, but can reach larger size and
capacity to adsorb greater quantities of water (Figure 5).
During the first few years, to enhance the vegetation land-cover will be planted also shrub and
herbaceous essences. These will provide good coverage from second year, without interfering with the
growth of trees.
To maximize the GW phyto extraction function, also adult tree will be planted (4-6 years old), in figure 6 is
shown a rendering of what it looks like the woods.
A B
The evaluation of consumption potential evapotranspiration was based on the application of the equation
of Heargraves (56 notebook FAO) and the meteorological data daily (maximum and minimum temperature
of the air).
The estimate evapotranspiration was converted into potential evapotranspiration using of an appropriate
crop coefficient.
In Table 1 is shown the potential evapotranspiration based on monthly average (period 1993 to 2012) for
the first three years and in long term conditions, compared with the monthly rainfall calculated over the
same period.
Table 1 - Potential evapotranspiration of projected forest (in light red values above monthly
rainfall)
The study shows that a forest adequately designed, after three years of implantation, requires an amount
of water for its biological function greater than rainfall.
This results was deepened using GW modeling in order to simulated the effect of the water stress on the
aquifer physically confined.
Modeling showed the possibility to gradually reduce the pumping from the wells after two years, and turn
off the P&T when all the plants already grown.
According to the simulation that period would be reached four years from In the forest implant (Figure 7)
Piezometric surface inside the
physical confinement
The study demonstrates suitability of some plants to integrate the traditional systems of P & T, taking
advantage of their ability to adsorb water from the roots to their biological functions.
In the specific case, in a highly contaminated area, located in the south of Italy, 44 Ha forest has been
designed to withstand the typical Mediterranean climatic conditions.
During design phase, in order to select the most appropriate plants, three parameters have been
considered:
resistance to salinity;
resistance to waterlogging;
preference for native plants.
The study provides for the use of halophyte plants, appropriate to the scope for their capacity of absorbing
the salts present in water, mixed with other plants more tolerant of wetland conditions.
For the structure of the tree patterns it was decided to use mainly Willow (Salix fragilis / alba) and Ader
(Alnus glutinosa), both are colonizer plants, characterized by fast growth.
Other plants have been identified, with slower growth, but the trunk higher, with the ability to assume a
larger amount of water, such:
Eucaliptus
Taxodium disticum.
Quercus palustris and Quercus robur.
Tamarix gallica
Groundwater modeling showed the forest will have CO2 emissions per Kwh of electrical power
consumed by P&T system
an immediate growing up of phytoextraction function
and, after four six years, it will be able to fully 1.800
1.561
1.200 1.147
The economic benefits will be considerable: it is
Cumulative tons
1.009
log scale
1.000
possible to estimate a savings between three and 871
800
four million Euros in 10 years post implant of the 595
733
forest. 600
457
400 319
References
Brownfields Technology Primer: Selecting and using Phytoremediation for Site cleanup. US
environmental Protection Agency. Office of solid waste and emergency response. Technology
innovation office. Washington DC 20460
J-Field phytoremediation project. Field events and activities through July 31, 2000 Aberdeen proving
ground edgewood, Maryland.US EPA Work assignment N: 0-034. Lockheed Martin Work Order n:
R1A00034. USEPA contract N: 68-C99-223
Guidance for using Bioswales, vegetative buffers and constructed wetlands for reducing, minimizing
or eliminating pollutant discharges to waters. State of Oregon, Department of environmental Quality,
Dennis Jurries PE, Storm water Engineer. January 2003
Maintaining Hydraulic Control Using Deep-Rooted Tree Systems. A. Ferro M. Gefell R. Kjelgren
D. S. Lipson N. Zollinger S. Jackson. Advances in Biochemical Engineering/ Biotechnology, Vol. 78
Managing Editor: T. Scheper Springer-Verlag Berlin Heidelberg 2003
Emerging Technologies for the Remediation og Metals in Soils Phytoremediation. Interstate
Technology and Regulatory Cooperation (ITRC) Work Group, December 1997.
Newsletter and Technical Publications Freshwater Management Series No. 2: Phytoremediation: An
Environmentally Sound Technology for Pollution Prevention, Control and Redmediation- An
Introductory Guide To Decision-Makers
REMEDIATION OF TOXIC METALS CONTAMINATED SOIL USING EDTA SOIL WASHING
a b a
Erika Jez , Neza Finzgar , Domen Lestan
a
Biotechnical Faculty, University of Ljubljana, Jamnikarjeva 101, 1000 Ljubljana, Slovenia
tel.: +386 (01) 320 31 62
fax.: +386 (01) 256 57 82
E-mail address: domen.lestan@bf.uni-lj.si
b
Envit Ltd., Vojkova cesta 63, 1000 Ljubljana, Slovenia
Abstract
Geostatistical technique was used to predicted spatial reduction of lead contamination after applying
EDTA-based soil remediation. Soil samples from 268 locations in the Meza Valley, Slovenia with Pb
1
up to 8955 mg kg were extracted with the chelating agent ethylenediamine tetraacetate (EDTA). On
average, 63% of Pb was removed with washing solution containing 60 mmol EDTA per kg of soil.
Geostatistical simulations showed that soil remediation has the potential to reduce the area with Pb
above the critical regulatory threshold limit by 91%. The potential of remediation with EDTA soil
washing to mitigate the risk from Pb poisoning was investigated by applying the Integrated Exposure
Uptake Bio-Kinetic (IEUBK) model. The IEUBK model predicted that, after soil remediation, the
number of locations at which the expected blood Pb level in children was higher than the stipulated
1
10 g dL would decrease by 90, 38 and 91% in the towns of Mezica, Zerjav and Crna, respectively.
In a field experiment on metal contaminated and EDTA-remediated soil we studied plant performance
and prospects of potential re-use of remediated soil as a garden substrate. Two experimental plots of
4 1 0.3 m were filled, one with remediated and the other with original contaminated soil. Selected
cultivars were rotated over the course of 16 months. Remediation reduced Pb concentrations in roots,
green parts and fruits in most of the plants below the concentration stipulated by legislation. The
results confirmed the feasibility of soil washing with EDTA as an efficient remediation measure in
Mezica and Crna and advice for soil capping/removal for the most polluted town of Zerjav.
Introduction
Soil contamination, caused by the large amounts of man-made pollutants and chemicals that are being
anchored into agricultural and urban soils every day, is becoming a major problem. Toxic metals are
one of the main causes of concern, since they are persistent in soils and are difficult to remove. Soil
degradation processes are increasing the need to use contaminated land, such as former landmine
sites and urban areas, for several purposes. For example, urban horticulture is currently booming in
the world, fulfilling a variety of functions, including food production. Effective soil remediation is
therefore urgently needed. Soil washing using the chelating agent ethylenediamine tetraacetate
(EDTA) has been found to be effective. In this regard, we have demonstrated the technical and
economic feasibility of chelant soil washing with EDTA solution for soils in the Meza Valley [1]. No
waste-water is generated. Solid wastes are efficiently bitumen-stabilized before disposal. The cost of
soil remediation is driven by the amount of soil to be treated (the economy of scale). The cost of our
technology was evaluated for a demonstrational, 6-ton of soil per day remediation plant which is under
construction in the Pb contaminated Meza Valley, Slovenia (financed in part from the EU Life+
-1
project). The cost amounted to 299 ton of soil [2] which is favourable compared to un-sustainable
soil dig and dump. The physics and chemistry of the novel technology is robust and allows
remediation of soils with different properties and contamination level.
One of the most affected locations in Slovenia is the Meza Valley, historically contaminated from Pb
mining and smelting. The site is home to more than 6000 people in three main residential
communities: the towns of Mezica, Crna and Zerjav. Living in contaminated areas entails being in
contact with contaminated soils in everyday life; indeed, hands to mouth behaviour with children,
accidental soil ingestion and soil dust inhalation are the most significant pathways of human Pb
uptake. In 2007, an on-going action program for improving the quality of the environment and reducing
the health hazard for people in the Meza Valley was introduced a goal of program is 95% of children
1
with levels less than 10 g dL . But annual surveys of blood Pb concentration indicated that the
1
number of children with levels higher than 10 g dL did not change over the 6-year (2007-2013)
duration of the action program.
In on-going study we predict spatial feasibility of EDTA soil washing as remediation solution for Meza
Valley. Second we investigate whether EDTA soil washing has the potential to mitigate the risk to
children from Pb poisoning in the Meza Valley. Soils near towns of Mezica, Zerjav and Crna were
remediated and data obtained were used in the IEUBK model. Remediated soil were also tested as an
agricultural or garden substrate in a four-crop rotation experiment, where several plant species were
cultivated and their metal and nutrient uptake, were analyzed.
The IEUBK model (IEUBKwind32 Lead Model Version 1.1 Build 11) was used to calculate geometric
mean blood lead levels in children aged 3 years old. The model was fed with Meza Valley soil
contamination, remediation data - experimentally obtained data on total and bio-accessible Pb soil
concentrations (calculated to ABA [3]) and data on the Pb concentration in drinking water and in air
particles from external sources were fed into the IEUBK model. All other values were set as default.
For on location pilot soil remediation was done [2] and two experimental plots on raised garden beds
4 m 1 m 0.3 m were filled with EDTA-remediated and original soil up to 0.15 m in height. The
garden beds were exposed to environmental conditions from April 2011 to September 2012. After this
four-crop rotation experiment was. The first crop rotation (April 2011) was represented by onion
(Allium cepa L.), pea (Pisum sativum L.), spinach (Spinacia oleracea L.) and cauliflower (Brassica
oleracea L., Botrytis), the second rotation (August 2011) by spinach and Chinese cabbage (Brassica
rapa L., Pekinensis), the third rotation (April 2012) by carrot (Daucus carota L.), lettuce (Lactuca sativa
L.) and spinach and the fourth rotation (August 2012) by basil (Ocimum basilicum L.) and bell pepper
(Capsicum annuum L.). Seeds and seedlings (treated in Klassman Deilman substrates) were planted
directly in remediated and original soils.
Figure 1: Prediction map of Pb concentrations in soils of the Meza Valley before and after
remediation, with critical Pb soil concentrations indicated by a broken line.
The geometric mean blood Pb levels in 3-years old children from Mezica, Zerjav and Crna before and
after remediation are presented in Fig. 2. The obtained results indicate that soil washing with EDTA is
a feasible remediation option in the towns of Mezica and Crna; a few isolated locations at which higher
than targeted blood Pb levels were predicted could be treated by more conventional methods, for
example by soil capping. Although the IEUBK model showed that EDTA-based soil remediation would
significantly mitigate the health risk in Zerjav (Fig. 2D), the target blood Pb level was reached in less
than half of the locations tested, mainly in the urban periphery. Although the IEUBK is continuously
updated and improved, it does not fully comprehend the diversity of children's behavioral and
physiological responses to environmental Pb. The background in vitro Pb bio-accessibility and bio-
availability methods also rely on simulations of the human gastro-intestinal tract and on animal
models. The results of IEUBK modeling, therefore, need to be taken into the account with caution.
Figure 2: IEUBK predicted blood Pb concentration in children for three urban areas in Meza Valley,
Slovenia, before and after soil remediation.
Remediation technologies often concentrate only on contaminant removal efficiency and overlook the
treated soil's potential use after remediation. EDTA soil washing to some extent deteriorates soil
physical and biological properties. Revitalization of remediated soil by providing missing structure,
nutrients and microbial activity to a large extent restored soil functions. Assessing plant physiological
responses indicated that the applied remedial technology did not limit the use of processed soil in
terms of reducing the plant fitness. Vegetables grown on remediated soil did not exceed Pb limits set
by EU regulations on contaminants in the foodstuffs (Table 1). In fact, Pb was not transported to the
edible parts of any vegetable tested on remediated soil, except for sparse amounts found in spinach
[4]. The decision on remediation and particularly to use remediated soil for food production will
probably depend on the case by case evaluation of the remedial efficiency, selections of cultivars and
soil and site specifics.
Table 1. Concentrations of Pb in edible parts of vegetables and green parts of horticultural plants
cultivated on original and EDTA remediated soil. Data are given as means and standard deviations of
at least three replicates.
-1
Pb (mg kg )
Plant
Original Remediated
Chinese cabagge (Brassica rapa) 168 178
a
b
Chinese cabagge (Brassica rapa) 6.61.3 LOQ
c
Spinach (Spinacia oleracea) 295 114
Spinach Spinacia oleracea) 13.51.4 LOQ
b
a -1 -1
[5]; soil with 1378 mg kg Pb, treated with 60 mmol kg EDTA; 77% of Pb removed; pot experiment.
b -1 -1
[4]; soil with 1585 mg kg Pb, treated with 60 mmol kg EDTA; 80% of Pb removed; field experiment.
c -1 -1
Jelusic et al.(in press); soil with 4037 mg kg Pb, treated with 120 mmol kg EDTA; 70% of Pb
removed; pot experiment.
d -1 -1
[6]; soil with 1585 mg kg Pb, treated with 60 mmol kg EDTA; 80% of Pb removed; pot experiment
e -1
Unpublished results; soil with 4037 mg kg-1 Pb, treated with 120 mmol kg EDTA; 70% of Pb
removed; field experiment.
LOQ: limit of quantification of instrument (atomic absorption spectrophotometer).
Conclusion
Soil remediation is usually the most costly element in the management and restoration of Pb
contaminated sites. Investigation tools capable of reliable prediction of the efficiency of proposed and
planned soil remediation activities are therefore of significant importance. In our study we presented
that using geostatistical simulation can be applied to map the probability of exceeding the regulatory
threshold (critical) limit for Pb in soils of the Meza Valley, Slovenia. Soil extractions with EDTA
2 2
indicated a reduction from 19.4 km to 1.80 km of area in the Meza Valley critically contaminated with
toxic metals after remediation. Another tool used in study was the IEUBK model to predict the
reduction of blood Pb levels in children living in contaminated urban areas in the Meza Valley,
Slovenia. The geometric mean blood Pb levels in 3-years old children from Mezica, Zerjav and Crna
before and after remediation indicated that soil washing with EDTA is a feasible remediation option in
the towns of Mezica and Crna. In the current stage of development, our technology also promises safe
use of Pb remediated soil as a substrate for food production. In addition, revitalization of remediated
soil by providing the missing structure, nutrients and microbial activity is essential and is the subject of
our on-going research activities.
Reference
[1] N. Finzgar, E. Jez, D. Voglar, D. Lestan, Spatial distribution of metal contamination before and
after remediation in the Meza Valley, Slovenia, Geoderma. 217-218 (2014) 135143.
doi:10.1016/j.geoderma.2013.11.011.
[2] D. Voglar, D. Lestan, Pilot-scale washing of Pb, Zn and Cd contaminated soil using EDTA and
process water recycling., Chemosphere. 91 (2013) 7682.
doi:10.1016/j.chemosphere.2012.12.016.
[3] E. Jez, D. Lestan, Prediction of blood lead levels in children before and after remediation of soil
samples in the upper Meza Valley, Slovenia, J. Hazard. Mater. (2015).
doi:10.1016/j.jhazmat.2015.04.049.
[4] M. Jelusic, D. Vodnik, I. Macek, D. Lestan, Effect of EDTA washing of metal polluted garden
soils. Part II: Can remediated soil be used as a plant substrate?, Sci. Total Environ. 475 (2014)
14252. doi:10.1016/j.scitotenv.2013.11.111.
[6] V. Zupanc, D. Kastelec, D. Lestan, H. Grcman, Soil physical characteristics after EDTA
washing and amendment with inorganic and organic additives., Environ. Pollut. 186 (2014) 56
62. doi:10.1016/j.envpol.2013.11.027.
AquaConSoil 2015 13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil, Sediment and Water
Resources: Abstract submission
June 9-12 2015
1 1 2 3 4
Otaegi, N. , Cagigal, E. , Cernik, M. , Slunsk, J. , Bosch, J.
1 TECNALIA Research & Innovation, Parque Tecnolgico de Bizkaia C/Geldo, Edificio 700 E-48160 Derio (Spain);
+34667178879; +34667119800; nerea.otaegi@tecnalia.com; ekain.cagigal@tecnalia.com
2 Technicka Univerzita v Liberci, Liberec, Czech republic;
3 NANOIRON S.R.O., Rajhrad, Czech republic;
4 University Of Duisburg-Essen, Duisburg-Essen, Germany.
INTRODUCTION
However, up to date their practical application has been scarce because of barriers related to the
technical performance of available solutions, concerns over cost and sustainability, fate and risks of
engineered nanoparticles (NPs) and a lack of validated fieldscale performance data that addresses these
key regulatory and market concerns.
Within the European NANOREM project (FP7, Grant Agreement 309517), the main objective of
performing field test sites is just to provide the opportunity for a positive proof of nanotechnologies
application, i.e. a closed mass balance on site, addressing other current limiting factors such as their cost,
sustainability and environmental safety.
The present study describes the preliminary works undertaken and first results obtained from one of the
field test sites supported by the NANOREM project: the Spanish field test, characterized by the presence
of high heavy metals levels both in soil and groundwater.
SITE DESCRIPTION
The Spanish test site is an abandoned 20 Has brownfield in Asturias (Northern Spain), used historically
for the production of fertilizers during 48 years (1950 1998).
AquaConSoil 2015 13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil, Sediment and Water
Resources: Abstract submission
June 9-12 2015
Contaminants include predominantly heavy metals both in soil and groundwater, being arsenic (As), lead
(Pb) and copper (Cu) the most significant ones (above reference values for soil and groundwater). Metals
are associated to the madeground.
Considering the mass of contaminants in groundwater is a mixture of heavy metals, it has been decided to
focus the research on a single metal treatment, thereby facilitating the interpretation of results and the
technology validation. Arsenic has been selected as the target contaminant in this particular case, given
its high groundwater concentration (>5000 g/l) and the lack of pilot test or field cases in the literature that
pay attention to this element, which makes this field experiment particularly groundbreaking.
The main source of arsenic is related to the sulphuric acid production by arsenopyrite (FeAsS) burning.
Pyrite ashes rich in As in the form of AsO4Fe resulting from this burning process were then uncontrolledly
dumped on site as madeground. Still, the study will also consider the effects of treatment on the other
metals and compounds like organic matter and dissolved anions such as nitrates, sulphates and
phosphates. As a result, the effect of NPs injection on natural biogeochemical processes of the soil-
groundwater interface and therefore the sustainability of the treatment with NPs will be better understood.
In order to optimize NPs delivery, preliminary site and laboratory works have been undertaken. Field
works on site have included sites hydrogeological characterization, including Lefranc permeability tests,
continuous groundwater level control by means of automatic dip meters and manual measurements, the
soil and groundwater quality determination and contaminants distribution definition in order to delineate
the exact location and design of the pilot test area.
According to field tests, the main groundwater movement follows a NE-SW direction, towards the adjacent
river, with negligible horizontal and/or vertical fluctuations. The hydraulic permeability of the aquifer unit is
about 1m/day and there is a good hydraulic connection between the different aquifer materials.
Heavy metals are the main type of contaminants in terms of their extension and concentration levels, both
in soil and groundwater media. In order to level site's topography, several waste materials from the sites
industrial activity itself and adjacent industries (iron works, coke batteries, coal washing, etc.) were
dumped and flattened on site as madeground, making it the main source of contaminants (see Figure 1).
SOIL (mg/kg)
As 300
Pb 2400
Zn 3500
Hg 20
Cu 1600
Cd 17 c)
a) b)
Figure 1.a) Diamond core drilling works, 1.b) pyrite ashes in madeground and 1.c) heavy metal content
AquaConSoil 2015 13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil, Sediment and Water
Resources: Abstract submission
June 9-12 2015
In groundwater, leaching processes from contaminated madeground result in high levels of dissolved
heavy metals, among which the most significant one is the arsenic: 5527.2 ppb maximum, 1378,5 ppb
mean concentration detected, which occurs mainly in a pentavalent oxidation state. Other metals include
lead and copper: 1278.4 ppb and 1000.0 ppb maximum concentration respectively.
In order to provide field evidence of the safe and effective deployment of NPs to treat contaminated
groundwater, a conceptual site model has also been completed, including a pre-deployment risk
assessment where the potential side effects of using NPs have been evaluated.
In the laboratory, five different types of NPs have been tested with the Spanish material in the Technicka
Univerzita v LibercI (Czech Republic) laboratories including the commercially available products based on
nano zero valent iron (NZVI), NANOFER 25s and STAR and milled particles (KKM2, A2), carboiron and
goethites (Nano Iron SRO and the University of Duisburg-Essen provided NPs). Main results from the lab
reactivity tests are summarized in following diagrams (Figures 2, 3).
According to these results, it is possible to assess the effect of the different NPs on the arsenic content:
Concentration of solved arsenic decreases significantly in the treatments the University of
Duisburg-Essen University of Duisburg-Essen provided goethites;
Concentration of solved As is also reduced when reacting with NZVI (activated NF Star).
Apparently reactivity with zerovalent iron is stronger (more efficient) than with iron oxides (goethites) in
terms of the presence of arsenic in solution.
According to literature, arsenates - main species presence in the Spanish site - and arsenites can be fully
transformed into As(0) by reaction with zerovalent iron. As(0) is insoluble, thermodynamically stable, and
this reaction is not expected to be reversible.
Additionally, on a theoretical basis, the chemical reaction between arsenite/arsenate and NZVI is
preferred to adsorptive interactions occurred with iron oxides NPs, being the latter weaker bonds and they
can then be more easily desorbed and restored to groundwater.
Figure 2. Laboratory reactivity test results with non-activated NanoFer STAR and Goethites
AquaConSoil 2015 13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil, Sediment and Water
Resources: Abstract submission
June 9-12 2015
There can be some uncertainties on discerning what kind of interactions are happening when using NZVI;
i.e., if arsenic species are actually reacting with Fe(0) or being adsorbed by the stabilizer of NF STAR. In
order to clarify this point, the pH factor and oxidation reduction potential (ORP) trends can provide
valuable information.
From the figures below (Figures 4, 5), it can be observed that ORP and the pH correlates with
concentrations of arsenic in solution, which lead to think that once the Fe(0) has entirely
(stoichiometrically) reacted with As(V)/As(III) (concentration of As below detection limit), corrosion
mechanisms related to an excess of Fe(0) starts affecting ORP (evolving towards negative values) and
increasing pH as a results of Fe oxidation.
Figure 4. ORP evolution following NF STAR treatment Figure 5. pH factor evolution following NF STAR
treatment
In summary, activated NanoFer STAR (NZVI) has been selected as the most suitable product to be
injected in the Spanish site. At the moment, surface modification methods are being tested on this type of
NPs in order to improve their mobility and environmental safety properties.
AquaConSoil 2015 13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil, Sediment and Water
Resources: Abstract submission
June 9-12 2015
A total of eight (8) monitoring wells will be installed down into the aquifer and the NPs injection will be
done in two (2) wells so that a good NPs distribution in the underlying aquifer is assured. The injection will
be performed by Geoplano Drill, S.A. over the entire aquifers column, at low pressure, in open boreholes
previously fitted with packers and automatic sampling systems.
The monitoring strategy wont be focused only on the arsenic removal and NPs efficiency determination
but also on most of the key challenges associated to nanotechnologies application that currently hamper
their real and extended use in the European environmental market. Thus, cost, sustainability and
environmental safety issues will also be addressed in the Spanish field test, engaging with some other
work packages defined in the NANOREM project and offering a complete and realistic approach to
nanotechnologies use.
With this respect, NPs effect on naturally occurring biota, other metals and compounds like organic matter
and dissolved anions such as nitrates, sulphates and phosphates will be controlled in order to address
NPs effect on natural biogeochemical processes of the soil-groundwater interface and their environmental
impact. Long-term fate of NPs will also be monitored.
Additionally a cost evaluation will be performed by inferring the costs of the field test to a whole
remediation project and setting these costs against those of a conventional treatment.
The starting of the injection works is planned by June 2015, following the design of the monitoring strategy
and the implementation of the necessary equipment in the field.
CONCLUSIONS
The European NANOREM project (FP7, Grant Agreement 309517), is designed to unlock the potential of
nanoremediation and so support both the appropriate use of nanotechnology in restoring land and aquifer
resources and the development of the knowledge-based economy at a world leading level for the benefit
of a wide range of users in the EU environmental sector.
Accordingly, a series of field tests have been defined in the project, among which the Spanish field test
aims at validating nanotechnologies application for arsenic removal by addressing the key and most
important challenges associated to this type of technologies.
Following a series of field and laboratory tests, sites hydrogeological and geochemical background have
been characterized (this is still an ongoing task). Thus, the main groundwater movement follows a NE-SW
direction, towards the adjacent river, with negligible horizontal and/or vertical fluctuations. Heavy metals
are the main type of contaminants in terms of their extension and concentration levels, both in soil and
groundwater media, among which the most significant one is the arsenic.
According to results from the laboratory tests, activated NanoFer STAR (NZVI) has been selected as the
most suitable product to be injected in the Spanish site.
At the moment, NPs surface modification methods are being tested in order to improve their mobility and
environmental safety properties.
Injection works starting date is due in June 2015, once the final design of the monitoring strategy and the
installation of all the monitoring equipment are completed.
APPLICATION OF HYDROCYCLONE SEPARATION SYSTEM FOR CONTAMINATED SOIL
TREATMENT AND FLOW SIMULATION
Yi-Kuo Chang1,*, Cheng-Di Dong2, Chiu-Wen Chen2, Chih-Feng Chen2, Zhen-Wei Hong1, Yi-Liang
Shu1, Chia-Ying Li1
1 Dept. Safety Health and Environmental Engineering, Central Taiwan University of Science and
Technology, Taiwan. *Corresponding author.
Address: 666 Po-Tze Road, Taichung city 406, Taiwan
Phone: +886-4-22391647 ext. 6858, Fax: +886-4-22399934
E-mail: ykchang@ctust.edu.tw
2
Department of Marine Environmental Engineering, National Kaohsiung Marine University,
Kaohsiung, Taiwan
ABSTRACT. In this work, the basic characteristics and the geochemistry of metals in the soils were
studied. In addition, the separation processes, hydrocyclone treatment, is designed to separate coarse
and fine-grained particles by high settling velocity of particles. Furthermore, the FLUENT flow
simulations are help to calculate the cut-off size and fractional efficiencies in hydricyclone separators.
The demonstration of liquid-solid flow and the simulation results are feedback to the hydrocyclone
operation parameters and the new hydrocyclone dimensions design. The effects of flows, particles and
geometrical parameters are also considers for calculations. Collection efficiency curves and cut-off size
values predicted by the FLUNET might match with experiments over a wide range of inlet velocities for
different types of hydrocyclones. The aim of present work is to develop new feeding system, to predict
the cut-off size and collection efficiency, which are applicable in big or small hydrocyclones.
1. INTRODUCTION
The rapid industrial development and rising populations have resulted in the increase of industrial
and agricultural investments in Taiwan, which unfortunately has led to the increasing potential risk of
waste contaminated soils and pollution of the major and minor rivers in Taiwan. This not only directly or
indirectly influences the natural load-bearing capacity of the environment, but also causes disruption of
the ecological balance, dissemination of diseases, lowering of land values, and gravely affects the
population's health. The irrigation water quality of certain agricultural regions in Taiwan has been
polluted by the waste drainage from up-stream industrial areas or illegal factories. Polluted water is
then channeled into farm lands from irrigation ditches, contaminating the soils, and leading to many
incidents of agricultural soil contamination in Taiwan. Accordingly, the objectives of this study are (1) to
study the basic characteristics of the soils, (2) to evaluate the effectiveness of hydrocyclone separation.
2. EXPERIMENTAL
The heavy metal containing soils used in this research came from various sections of farmland
from Central Taiwan, which were exceeded than the regulation of Taiwan EPA(2010) (Fig. 1). After
collection, the samples were vacuum dried, and then sealed for storage. The samples were rich in
organic matters and composed of mixtures of large and fine particles.
After soil collection, the soils were vacuum-dried, and stored in plastic boxes prior to analyses.
Particle size was determined with a particle size counter (Coulter LS 100, Particle Size Analyzer) and
vibration sieves. Metal content (e.g., Pb, Cd, Cr, Cu and Zn) in digested soil samples (digested with
High performance microwave digestion unit, Milestone 1200 mega) were determined by a flame
atomic absorption spectrophotometer (Hitachi Z-6000), and inductively coupled plasma optical
emission spectrometer (ICP-OES: PerkinElmer, Optima 2000 DV). Furthermore, X-ray fluorescence
spectrometer (XRF: Specttro, XEPOS) and Scanning Electron Microscope/Energy-Dispersive
Spectrometer (SEM/EDS, Hitachi, Model SEM S-3000N) are help to analyses the microstructure and
the element content.
The content of heavy metals and organic contaminants are primarily governed by the grain size. The
finer the particles and the higher the content of heavy metals and organic matter in the soils. Moreover,
Klamer et al. (1990) recommended analysing metal concentration in a <63 m fraction for grain size.
Thus, the hydrocyclone was applied to separate the particle sizes physically. The experimental setup
as shown in Fig. 2 has been used for the studies of performance characteristic of hydrocyclone
consisting of a vertical cylindrical portion.
overflow
underflow
70
>300 m
300-150 m
60 150-63 m
63-37 m
< 37 m
50
Content (%)
40
30
20
10
0
underflow overflow underflow overflow
CONCLUSIONS
The hydrocyclone was effective for separating the coarse-fine particles. The result of laboratory
research showed that the content of heavy metals contaminants is primarily governed by the grain size.
The optimum separation process was approached through hydrocyclone separation under the
experimental conditions of water/soil ratio of 30, pressure of 14 psi. The FLUNET (ANSYS) was help to
understand flow conditions, which could feedback to hydrocyclone operating conditions and design.
ACKNOWLEDGEMENT
The authors would like to thanks Ministry of Science and Technology (MOST 103-2622-E-166-002-CC3)
for financial support.
REFERENCE
Klamer, J. C., Hegeman, W. J. M. and Smedes, F., 1990. Comparison of grain size correction
procedures from organic micropollutants and heavy metals in marine sediments. Hydrobiologia, 208,
213-220.
Purves, D., 1985. Trace-Element Contamination of the Environment. Elsevier, Amsterdam, The
Netherlands.
Taiwan EPA, 2010. Soil and Groundwater Pollution Remediation Act.
Topic area :
1c. Remediation technologies and approaches
Title :
FRENCH NATIONAL APPROACH FOR DELINEATION OF CONTAMINANT SOURCE ZONES :
COMBINING MASS-BALANCE, MASS-FLUX AND MASS-DISCHARGE METHODS
Authors :
F. Lion, S. Colombano, N. Aubert, G. Boissard
Contact :
BRGM, 3 avenue Claude Guillemin BP 36009 45060 Orlans Cedex 2, France
+33 2 38 64 34 87
f.lion@brgm.fr ; s.colombano@brgm.fr ; n.aubert@brgm.fr ; g.boissard@brgm.fr
Keywords :
delineation, source zones, mass-balance, mass-flux, mass-discharge, Pareto principle
Costs
A
D
Optimization B
Pollution
ii. Use of the results to delineate the spatial contamination distribution, horizontally as well
vertically, as shown in Figure 2. This modelling allows defining several concentration intervals
and drawing several iso-concentration curves, which are usable for a given soil thickness
(generally 0,5 to 1 m thick) ;
Figure 2 : Example of isoconcentration curves
iii. Calculation of the volume of soils Vsoil included in each couple of isoconcentration curves ;
=
2
where S : surface between 2 isoconcentration curves (m )
h : thickness of the concerned stratum (m)
Table 2 shows all the parameters that can be calculated from the results of the investigations.
Table 3 shows all the parameters that can be calculated from the results of the investigations.
%
Concentration Remediation Pollutant mass % Cumulative
Cumulative
intervals cut-off (mpol) for each Pollutant pollutant
pollutant
(mg/kg) (mg/kg) interval (t) mass mass (t)
mass
0 - 0,5k 0 40 3,3% 1230 100,0%
0,5k - 2,5k 500 130 10,6% 1190 96,7%
2,5k - 5k 2 500 180 14,6% 1060 86,2%
5k - 10k 5 000 310 25,2% 880 71,5%
10k 20k 10 000 340 27,6% 570 46,3%
20k - 35k 20 000 200 16,3% 230 18,7%
35k - 50k 35 000 30 2,4% 30 2,4%
v. Study of the evolution of the pollutant mass and the soil volume as a function of the
concentration intervals : determination of the remediation cut-off
By plotting the percentage of pollutant mass and the percentage of soil volume evolutions as a
function of the couples of iso-concentration intervals, as proposed in the Figure 3, one can easily see
that it is technically relevant to treat soils which are contaminated by a concentration greater than
5 000 mg/kg (ratio %mass of pollutant / %soil volume > 1).
%
Ratio =
%
Farthest
curves
Figure 6 : Calculations of the total mass discharge (adapted from SERDP, ESTCP)
After calculating the mass discharge in each polygon and the area of each polygon, the Pareto
principal can be used to determine the area which contributes the most to the discharge, as proposed
in the Figure 7. In this case, we can see that a polygon, whose the area is only 2% of the total area of
the transect, contributes to about 39 % of the discharge. Moreover, a 18% area zone contributes to 90
% of the total mass discharge.
Polygons area / total transect area (%)
Distance to one end of the transect (m)
18%
0 21,9 34,7 47,9 60,4 71,9 86 104,9
3,05 1% 1% 1% 1% 1% 1% 1% 1%
5,18 2% 3% 2% 2% 2% 2% 3% 2%
2% 2% 2% 2% 2% 2% 2% 1%
Depth (m)
7,62
9,14 1% 2% 1% 1% 1% 1% 2% 1%
10,7 2% 3% 2% 2% 2% 2% 3% 2%
13,7 3% 5% 3% 3% 3% 3% 4% 3%
18,3 3% 2% 2% 2% 2% 3%
Mass discharge per polygon / total mass flux (%)
90% Distance to one end of the transect (m)
0 21,9 34,7 47,9 60,4 71,9 86 104,9
3,05 0% 0% 0% 1% 13% 3% 0% 0%
5,18 0% 0% 0% 1% 1% 0% 0% 0%
Depth (m)
7,62 0% 0% 0% 1% 5% 39% 0% 2%
9,14 0% 0% 0% 17% 3% 7% 1% 0%
10,7 0% 0% 0% 1% 1% 0% 1% 0%
13,7 0% 0% 1% 1% 1% 0% 0% 0%
18,3 0% 0% 0% 0% 0% 0%
8. References
Colombano, S., Lprond, H., Hub, D., (2012) Journe technique du 9 octobre 2012 : Du plan de
gestion lachvement des travaux de dpollution. Le Plan de Gestion : des textes de 2007
leur mise en uvre
Erickson, M., Mayer, A., Horn, J. (2002) - Multi-objective optimal design of groundwater remediation
systems: application of the niched Pareto genetic algorithm (NPGA). Adv. Water Resour. 25 (1),
5165.
Falta, R.W., P.S. Rao, and N. Basu. (2005a) - Assessing the impacts of partial mass depletion in
DNAPL source zones: I. Analytical modeling of source strength functions and plume response.
Journal of Contaminant Hydrology 78, no. 4: 259280.
Falta, R.W., N. Basu, and P.S. Rao. (2005b) - Assessing the impacts of partial mass depletion in
DNAPL source zones: II. Coupling source strength functions to plume evolution. Journal of
Contaminant Hydrology 79, no. 1: 4566.
Farhat SK., Newell PhD. and CJ., PE, DEE, Groundwater services Inc. Texas, Nichols EM., LFR
Levine-Fricke New Hampshire, (2006) - Mass flux toolkit, to evaluate groundwater impacts,
attenuation, and remediation alternatives, version 1.0, 1-10.
GeoSiPol (2005) - Gostatistique applique aux sites et sols pollus - Manuel mthodologique et
exemples dapplications, 139 p.
Huntley, D., Beckett, G.D., (2002) - Persistence of LNAPL sources: relationship between risk
reduction and LNAPL recovery. J. Contam. Hydrol. 59, 3 26.
ITRC (Interstate Technology & Regulatory Council) (2010) - Use and Measurement of Mass Flux
and Mass Discharge. 89p., 5 ann
Kavanaugh, M.C., Rao, S.C., Abriola, L., Cherry, J., Destouni, G., Falta, R., Major, D., Mercer, J.,
Newell, C., Sale, T., Shoemaker, S., Siegrist, R., Teutsch, G., Udell, K. (2003) - The DNAPL
Remediation Challenge: Is There a Case for Source Depletion? U.S. EPA/600/R-03/143. Ada,
OK: National Risk Management Research Laboratory, U.S. EPA.
Mariner, P.E., Jin, M. and Jackson, R.E. (1997) - An algorithm for the Estimation of NAPL Saturation
and Composition from Typical Soil Chemical Analyses. Ground Water Monitoring and
Remediation, 3:122-129.
McDade, J.M., McGuire, T.M., and Newell. C.J. (2005) - Analysis of DNAPL source-depletion costs
at 36 field sites. Remediation Journal 15, no. 2: 918.
McGuire, T., McDade, J., Newell, C. (2006) - Performance of DNAPL Source Depletion Technologies
at 59 Chlorinated Solvent-Impacted Sites. Ground Water Monitoring and Remediation 26, n1,
p 7384.
Newell, C.J., and Adamson. D.T., (2005) - Planning-level source decay models to evaluate impact of
source depletion on remediation timeframe. Remediation 15, no. 4: 2747.
Rabotyagov, S., Campbell, T., Gassman, P., Jha, M. and Kling, C. - Searching for Solutions: Cost-
Efficient Pollution Control on a Watershed Scale.
http://www.swcs.org/documents/resources/u_rabotyagov_using_genetic_algorithms.pdf
RECORD (2013) - Retour dexprience critique sur lutilisation de mthodes gostatistiques pour la
caractrisation des sites et sols pollus, 135 pages, n11-0514/1A
SERDP, ESTCP - Mass flux and mass discharge : uses and measurement, power point presentation
Yang, Y., Wu, Jia., Sun, X., Wu, Jic., Zheng, C., (2013) A niched Pareto tabu search for multi-
objective optimal design of groundwater remediation systems Journal of Hydrology 490, 56-
73.
Isolation and Identification of Cultivable Anthracene Degrading Fungi from PAH
a
Department of Soil Microbiology and Symbiotic Systems, Estacin Experimental del
b
Department of Environmental Microbiology, Water Research Institute, University of
Abstract
Fungi represent the dominant living biomass in soil1. The role of these microorganisms
in bioremediation processes hasnt been fully exploited yet, and nowadays there are
many taxa that still remain unstudied2. In addition to their ability to resists and
hydrocarbons (PAHs), fungi play an important role in the functional ecology of soil.
Among their functions, some are known to influence the development of mycorrhizal
phytoremediation since their symbiotic interaction increases the root system of plants,
which augments the potential of dissipation in soil. The aim of this study was to find
phytoextraction. In this study, a total of 24 fungal cultivable species have been isolated
from 5 sampled sites in a historically pyrogenic PAH polluted soil. The isolated fungi
Among the 24 screened fungal species, 9 of them were able to convert anthracene in
submerged cultivation and in a rich carbon source medium, leading to the formation of
isolated fungi for pollutant removal. Fomes and Trichoderma fungi were used in a
Our results show that a selective plant-associated fungus may constitute a promising
phytoremediation.
Introduction
The investigation of new remediation technologies and methods is still very important trend in the field of site
decontamination. The NANOBIOWAT (Environmentally friendly nanotechnologies and biotechnologies in water and
soil treatment) is one of the projects dealing with this problem. In the framework of this project a new type of
surface stabilized nano zero valent iron (nZVI) NANOFER STAR was tested on a site contaminated by hexavalent
chromium and chlorinated hydrocarbons (mainly TCE). Chromium naturally occurs mainly in trivalent Cr (III) and
hexavalent Cr (VI) forms. While trivalent chromium is a bio-genic element, hexavalent chromium has significant
harmful effects on the environment because of its solubility, mobility, high oxidation potential and toxicity in general.
The in-situ pilot test was performed at an industrial area, where chromium was formerly used for plating. The
-1
concentration of dissolved chromium in groundwater was up to 60 mgl and the concentration of chlorinated
-1
hydrocarbons ranged up to 10 mgl .
NANOFER STAR is a new type of nZVI, which is stabilized by its thin inorganic layer. Thanks to that, the nZVI
powder is stable in the air. The advantages are better storage, transportation and handling because the suspension
can be prepared directly on site. On the other hand it still keeps its high reactivity. The new product is Surface
stabilized, Transportable, Air-stable and Reactive.
Application of nZVI
Six 9-meter-deep monitoring wells were installed onsite. One inflow well, three in-line wells approximately 2 m from
the nZVI injection points (first monitoring line wells) and two wells about 5 m further down the groundwater stream
(second monitoring line wells) were utilized for monitoring of the groundwater throughout the test. nZVI was
injected via direct push into 5 boreholes situated in the middle of the monitoring area.
Results
The physical and chemical parameters of the groundwater were monitored during a 3-month period. The redox
potential in the first monitoring line had significantly decreased (up to -400 mV) a day after the application. The
decrease in the second monitoring line was noticed a few days delay. A similar trend was observed for pH, which
slightly increased after the injection.
The migration of nZVI was confirmed by analysis of sediments from the first monitoring line wells, which showed
a high concentration of iron in comparison with the initial analysis before injection. The concentration of Cr (VI) in
the groundwater decreased significantly in the first monitoring line wells (see the Figure 2). Approximately 2 months
after the injection the reductive capacity of nZVI was probably exhausted and the concentrations of Cr (VI) started
to increase slowly. Nevertheless the efficiency of Cr (VI) removal was still 90% at the end of the monitoring process
(after 3 months).
Figure 2: The concentrations of Cr (VI) in the groundwater in monitoring wells (HV-2, MV-2, MV-3, MV-4, MV-5)
and in an inflow well MV-1
The application of NANOFER STAR had also a positive effect on chlorinated hydrocarbons degradation and
probably it also stimulated biological dechlorination (see the Figure 3). The process of reductive dechlorination
(probably both abiotic and biotic) was confirmed with the rise of nontoxic products - ethane and ethene (see the
Figure 4). The toxic intermediate vinyl chloride had not been detected.
The sulphate concentrations werent significantly influenced. The short-term increase of iron was observed after the
nZVI application, which also caused a partial mobility of manganese. The microbiological analyses confirmed
significant increase of psychrophilic bacteria in the groundwater after the nZVI injection.
Figure 3: The concentrations of chlorinated hydrocarbons in groundwater in monitoring wells (HV-2, MV-2, MV-3,
MV-4, MV-5) and in inflow well MV-1
Figure 4: The concentrations of ethane ane ethene in groundwater in monitoring wells (MV-2, MV-3, MV-4)
Conclusion
The new system of monitoring wells MV-1 to MV-5 was installed at said site in December 2013. The direct push
3
method was used for NANOFER STAR injection into the groundwater at the beginning of 2014. In total 20 m nZVI
-1
at the concentration of 1 g.l was injected. The site was monitored for three months until the capacity of nZVI was
exhausted. The site monitoring confirmed a decrease of Cr (VI) and chlorinated hydrocarbons, a positive effect of
NANOFER STAR. The concentration of Cr (VI) in groundwater was below detection limit in the first monitoring line.
The concentration of chlorinated hydrocarbons (especially TCE) had dropped under 50% immediately after the
nZVI injection.
The results of the pilot test showed, the surface stabilized nZVI NANOFER STAR is comparable to other nZVI
products available on the market, but due to its stability, it is much easier to handle and store.
Acknowledgement
The project was supported by the Technology agency of the Czech Republic (project No. TE01020218).
Coordinated approach for treatment of dredged sediment from small harbours
Sara Dastoli 1, Elena Romano 1, Alessandra Polettini 2, Raffaella Pomi 2, Aldo Muntoni 3, Barbara
Villani 4, Antonello Zucca 5
1
ISPRA (Istituto Superiore per la Protezione e la Ricerca Ambientale), Via Vitaliano Brancati 60,
00144 Rome, Italy - email: sara.dastoli@isprambiente.it - elena.romano@isprambiente.it
2
Sapienza, Dipartimento di Ingegneria Civile, Edile e Ambientale. Via Eudossiana 18, 00184 Rome,
Italy - email: alessandra.polettini@uniroma1.it, raffaella.pomi@uniroma1.it
3
Universit di Cagliari, Dipartimento di Ingegneria Civile, Ambientale e Architettura. Via Marengo 2,
09123 Cagliari, Italy - email: amuntoni@unica.it
4
ARPA (Agenzia Regionale per la Protezione Ambientale) Emilia Romagna, Via F. Rocchi 19, 40138
Bologna, Italy email: bvillani@arpa.emr.it
5
Consiglio Nazionale delle Ricerche, Istituto di Geologia Ambientale e Geoingegneria. Piazza dArmi
19, 09123 Cagliari, Italy email: azucca@unica.it
Abstract
The preservation of marine and coastal areas requires a sustainable and environmentally friendly
management of polluted sediments. In small harbours with low financial capabilities and limited
technical and operational capacity, the application of full-scale treatment and valorisation of dredged
sediments is often hampered by high investment and operational costs. The COAST BEST project
CO-ordinated Approach for Sediment Treatment and BEneficial reuse in Small harbours neTworks,
funded in the framework of the LIFE+ 08 Environment European call, has been focused on the actions
to promote sediments utilization as a substitute of natural raw materials, so as to reduce the
consumption of natural resources. A quite flexible treatment process was thus investigated both at
laboratory and pilot scale, based on physical and mechanical washing processes to isolate the
contaminated fractions and promote the recovery and utilization of sand. During the project, the
sediments of small harbours located in the Emilia Romagna Region (Adriatic Sea) were characterized
so as to evaluate the optimal conditions and separation efficiency that can be attained under different
sediment treatment sequences. Simple physical separation treatments (including sieving, hydraulic
separation, attrition, washing) have been combined in a pilot scale multi-stage process and the each
size class (sand, gravel and silt/clay) have been characterized in order to evaluate the chance to
recover the sandy fraction.
1. Introduction
Dredging and management operations of contaminated sediments are generally related to harbour
areas maintenance. Dredged sediments utilization in civil engineering applications is however
hindered by the presence of multiple contaminants, which in turn make the application of remediation
treatment a complex task. The uncertainties in predicting the treatment process yields. The lack of a
market for the recovered materials as well as the poor regulatory framework for sediments utilization,
limit as well the full-scale implementation of sediment treatment and recovery options. Moreover,
harbours of local relevance often experience the scarcity of technical and financial resources, which
magnifies the burden of such constraints. Nevertheless, the significant presence of small harbours
along the European and Italian coast, makes the identification of environmentally, technically and
economically sustainable alternatives for the management of dredging sediments an urgent task. In
this framework, the European Commission funded, within the LIFE+ Environment program, the project
CO-ordinated Approach for Sediment Treatment and BEneficial reuse in Small harbours neTworks
(COAST BEST), aimed to create an optimized system for dredging activities management, by 1)
integrating and harmonizing the different phases of procedures, such as sediment characterization,
dredging, treatment and valorization, and 2) interconnecting small harbours into an integrated network
based on geographical proximity and harbour characteristics criteria. The project also included a pilot
scale demonstration campaign to assess the feasibility of a simple sediment washing process aimed
to separate the valuable sandy fraction (for reusing in civil and environmental applications such as i.e.
beach nourishment), from the fine and (likely) more contaminated material.
The <20 m size fraction belonging to Poces 1, Poces 2, Bellaria 1 and Bellaria 2 samples was further
analyzed through a specific laser device (SEDI-GRAPH).
The <20 m Poces 2 fraction was furtherly subjected to a hydrocycloning treatment to evaluate where
contamination should be concentrated. The cycloning treatment was carried out using a lab scale
system (Mozley), adopting an operative pressure of 2.5 bar, and recirculating the underflow. The
hydrocycloning treatment produces 2 fractions: the sum of overflow 1 and overflow 2 (particle size <10
m) and the underflow (particle size >10 m), which were analysed by SEDI-GRAPH device in order
to assess the efficiency of the separation.
Table 3. Physical-chemical characterization of Porto Garibaldi sediments and its different fractions.
Values higher than regulation limits for urban areas are in orange bold, values higher than regulation limits for industrial areas are in red bold
Parameters (used unit) Fractions National law limits
(D.Lgs. 152/06)
Porto >5000 <5000 <2000 <1000 <500 <63 m Residential Industrial
Gariba m >2000 >1000 m >500 m >63 m use use
ldi m
Mass fraction (% w/w) 1.06 0.66 0.80 0.79 92.06 4.63
Grain size (Shepard, 1954) Sand
Grain size (Nota, 1989) Sand
VS (%TS) 3.44 1.48
TC (%TS) 2.89 2.45
IC (%TS) 1.36 2.03
TOC (%TS) 1.53 0.42
Zn (mg/kg d.w.) 127.70 1110.42 73.30 171.40 150 1500
Ni (mg/kg d.w.) 74.60 540.64 50.60 98.60 120 500
Cd (mg/kg d.w.) 0.71 12.80 0.13 0.25 2 15
Pb (mg/kg d.w.) 25.90 n.d. 28.30 33.20 100 1000
Cr (mg/kg d.w.) 90.60 544.408 66.60 133.20 150 800
As (mg/kg d.w.) 7.31 13.64 6.76 14.37 20 50
Cu (mg/kg d.w.) n.d. n.d. n.d. n.d. 120 600
V (mg/kg d.w.) n.d. n.d. n.d. n.d. 90 250
Hg (mg/kg d.w.) 0.031 n.d. 0.018 n.d. 1 5
Benzo(a)anthracene (mg/kg d.w.) 0.016 1.572 0.003 0.013 0.5 10
Benzo(a)pyrene (mg/kg d.w.) 0.011 1.089 0.003 0.013 0.1 10
Benzo(b)fluoranthene (mg/kg d.w.) 0.012 0.858 0.005 0.022 0.5 10
Benzo(k)fluoranthene (mg/kg d.w.) 0.005 0.526 0.001 0.007 0.5 10
Benzo(g,h,i)perylene (mg/kg d.w.) 0.006 0.382 0.003 0.012 0.1 10
Chrysene (mg/kg d.w.) 0.021 1.449 0.007 0.020 5 50
Dibenzo(a,h)anthracene (mg/kg 0.001 0.136 n.d. 0.001 0.1 10
d.w.)
Indeno(1,2,3,c,d,)pyrene (mg/kg 0.006 0.593 0.001 0.010 0.1 5
d.w.)
Pyrene (mg/kg d.w.) 0.030 2.234 0.010 0.041 5 50
Total PAHs (mg/kg d.w.) 0.165 14.269 0.056 0.229 10 100
Total PCBs (mg/kg d.w.) 0.006 0.029 0.005 0.017 0.06 5
Hydrocarbons C>12 (mg/kg d.w.) 13.00 67.00 12.00 34.00 50 750
Tributyltin (g/kg d.w.) 5.6 12.7 9.2 9.2 5 (D.M. 56/09)
The fine sand fraction (between 63 and 500 m) was found to be potentially suitable for utilization, as
was the < 63 m undersize fraction (although this accounted for 5% only of the original material).
Conversely, in the coarse sand fraction was found contaminants (Heavy metals, total PAHs and most
of the single PAHs) in excess of the threshold limits established by the Italian regulation on soil.
3.2.2 Cervia
Also the CERVIA sample would be suitable for the beach nourishment, since the sand content is close
to 90% of the overall mass of material and the content of fine fractions is quite limited. Following table
reports the physical-chemical characterization of Cervia sediments and its different fractions.
Both the analyzed fractions (between 500 and 63 m and <63 m) were observed to meet the above
mentioned regulatory criteria, and were thus both judged suitable for utilization. Although the technical
guidelines (ICRAM-APAT, 2007), due to low pollutant concentrations and low content (< 10%) of fine
fraction, in the reuse of marine sediments permit the utilization of the whole sediment sample given
the a separation of coarse fraction (> 500 m), for esthetic reasons, should be preferred.
3.2.3 Cesenatico
Cesenatico sediments are not suitable for the sand recovery. Following tables report the physical-
chemical characterization of two different samples (Poces 1 and Poces 2) and their different fractions.
The content of fine particles is remarkable for Poces 1 (Table 5) and Poces 2 (Table 6):, the <20 m
fraction is higher than 40% of total mass and the <63 m size fraction is about 64-68% of total mass.
So, is clearly evident that the recovery of this materials is not conceivable. In addition, the particle size
distribution of the <20 m size fraction analyzed by SEDI-GRAPH confirms that roughly 80% of this
fraction is <10 m and 60% <5 m.
Poces 2 sediment sample is subjected to an hydrocycloning treatment on the <20 m fraction in order
to separate the fraction <10 m and assess if the contamination may be further concentrated in the
<10 m size fraction. The overflow of hydrocycloning treatment content about 96.5% size particles is
<10 m and about 90% size particles is <5 m.
Samples Poces 1 and Poces 2 are very similar in terms of original size distribution and contaminant
content in the separated fractions. The potentially critical pollutants (Cd, As and C>12 hydrocarbons)
were found to be associated to the fine (< 20 m) fraction, while Zn was detected with appreciable
concentrations in all the investigated fractions. Size separation was able to concentrate some
contaminants in selected size fractions, but not to produce individual sediment fractions complying
with the Italian criteria for soil reuse in residential sites (D. Lgs. 152/06).
Table 8.. Physical-chemical characterization of Bellaria 2 sediment and its different fractions.
Values higher than regulation limits for urban areas are in orange bold, values higher than regulation limits for industrial areas are in red bold
Parameters Fractions Law limits
(D.Lgs. 152/06)
Bellaria 2 +250m -250 +125 -125 -63 +20 -20m Urban Industrial
m +63m m areas areas
Mass fraction (% w/w) 1.21 2.15 5.78 14.34 76.52
Grain size (Shepard, 1954) Clayey
silt
Grain size (Nota, 1989) Pelite
VS (%TS) 10.67
TC (%TS) 2.75
IC (%TS) 1.23
TOC (%TS) 1.52
Zn (mg/kg d.w.) 136.60 105.90 118.26 132.54 150 1500
Ni (mg/kg d.w.) 69.90 57.95 61.93 71.86 120 500
Cd (mg/kg d.w.) n.d. n.d. n.d. n.d. 2 15
Pb (mg/kg d.w.) 29.97 59.95 51.94 33.93 100 1000
Cr (mg/kg d.w.) 41.94 43.96 13.95 45.91 150 800
As (mg/kg d.w.) 13.10 15.15 13.34 14.21 20 50
Cu (mg/kg d.w.) n.d. n.d. n.d. n.d. 120 600
V (mg/kg d.w.) n.d. n.d. n.d. n.d. 90 250
Hg (mg/kg d.w.) n.d. n.d. n.d. n.d. 1 5
Benzo(a)anthracene (mg/kg d.w.) 0.004 n.d. 0.004 0.006 0.5 10
Benzo(a)pyrene (mg/kg d.w.) 0.009 n.d. 0.012 0.006 0.1 10
Benzo(b)fluoranthene (mg/kg d.w.) 0.041 n.d. 0.046 0.041 0.5 10
Benzo(k)fluoranthene (mg/kg d.w.) 0.004 n.d. 0.005 0.003 0.5 10
Benzo(g,h,i)perylene (mg/kg d.w.) 0.016 n.d. 0.020 0.015 0.1 10
Chrysene (mg/kg d.w.) 0.025 n.d. 0.026 0.018 5 50
Dibenzo(a,h)anthracene (mg/kg d.w.) 0.002 n.d. 0.002 0.001 0.1 10
Indeno(1,2,3,c,d,)pyrene (mg/kg d.w.) 0.005 n.d. 0.007 0.004 0.1 5
Pyrene (mg/kg d.w.) 0.021 n.d. 0.030 0.021 5 50
Total PAHs (mg/kg d.w.) 0.212 n.d. 0.254 0.199 10 100
Total PCBs (mg/kg d.w.) 0.0026 n.d. 0.0033 0.0037 0.06 5
Hydrocarbons C>12 (mg/kg d.w.) 28.52 n.d. 41.38 26.52 50 750
Tributyltin (g/kg d.w.) 4.90 n.d. 11.90 3.30 5 (D.M. 56/09)
The fine fractions contain critical contaminants at levels higher the threshold limits, therefore they
would be suitable for a number of reuse applications; potential alternatives include e.g., reuse as a
fine aggregate for concrete production, as a low-permeability material for landfill base liner and
impermeable cover construction, or as a source of Al, Si and Ca for cement clinker or ceramics
production.
Table 9. Mass balance and particle size analysis of Porto Garibaldi sediment. Pilot plant configuration: sieving + cycloning
Mass fraction (% w/w) Organic matter and gravel Sand Silt Clay
Feed flow 100 4.98 87.88 4.74 2.40
Sieve oversize 6.31 74.40 24.50 1.10 0.00
Sieve undersize 93.69 0.30 92.15 4.99 2.56
Cyclone under (recovered sand) 91.05 0.00 94.40 3.80 1.80
The content of fraction separated by the sieve is roughly 5%. The material recovered is 91.05% of the
total sediment mass, it contains 94.40% of sand particles and a negligible content of fine ones (3.80%
of silt and 1.80% of clay). In table 10 are reported main results of chemical characterization.
Table 10. Chemical characterization of Porto Garibaldi sediment separated during the pilot treatment.
Values higher than regulation limits for urban areas are in orange bold.
Analytes Fractions Law Limits
(D.Lgs. 152/06)
Input Oversize Undersize Oversize Undersize Urban Industrial
(screen) (screen) (cyclone) (cyclone) areas areas
Total PCBs (mg/kg d.w.) 0.0062 0.0072 0.0046 0.0039 0.0349 0.006 5
Total PAHs(mg/kg d.w.) 0.307 0.307 0.247 0.366 0.725 10 100
Hydrocarbons C <12(mg/kg d.w.) <0.01 1.89 0.13 <0.01 0.08 10 250
Hydrocarbons C >12 (mg/kg d.w.) 21 71 74 16 23 50 750
Al (mg/kg d.w.) 16180 16905 16190 14590 55470
As (mg/kg d.w.) 5.0 5.0 5.0 5.0 16.0 20 50
Cd (mg/kg d.w.) 0.3 0.1 0.3 0.3 0.9 2 15
Cr tot (mg/kg d.w.) 56.2 7.6 56.7 51.1 140 150 800
Cu (mg/kg d.w.) 10.9 10.4 8.8 8.5 47.0 120 600
Ni (mg/kg d.w.) 42.6 8.2 36.8 35.8 94.8 120 500
Pb (mg/kg d.w.) 4.0 2.0 3.0 3.0 21.0 100 1000
V (mg/kg d.w.) 26.8 5.4 26.0 24.1 98.5 90 250
Zn (mg/kg d.w.) 50.5 15.9 49.3 46.4 193.0 150 1500
In order to use the recovered sand in a beach nourishment intervention is necessary to provide a
further treatment stage to achieve an additional removal of inorganic compounds because the limits
concentration Zn and V for urban area was slightly exceeded.
3.3.2 Cervia
The pilot plant test was performed considering these main operating parameters:
- slurry feed flow rate: 3060 kg/h;
- L/S = 7.9.
Table 11 reports the mass balance of the samples of slurry collected during the running of the pilot
plant. The size analysis of each single flow is also reported.
Table 11. Mass balance and particle size analysis of Cervia sediment. Pilot plant configuration: sieving + cycloning
Mass fraction (% w/w) Organic matter and gravel Sand Silt Clay
Feed flow 100 1.13 87.36 8.38 3.12
Sieve oversize 4.46 25.38 65.54 6.40 2.57
Sieve undersize 95.54 0.00 88.38 8.47 3.15
Cyclone under (recovered sand) 91.91 0.00 92.31 5.87 1.81
Also for Cervia sediment, like Porto Garibaldi sediment, the content of fraction separated by the sieve
is roughly 5%.
The material to be reused, recovered by cyclone under, is about 91.91% of the total sediment mass.
In terms of particle size, the recovered material contains 92.31% of particles classifiable as sand and
less than 10% of fine particles (5.87% of silt and 1.81% of clay), limit for reuse in beach nourishment.
Table 12 shows the main results of chemical characterization.
Table 12. Chemical characterization of Cervia sediment separated during the pilot treatment
Analytes Fractions Law Limits
(D.Lgs. 152/06)
Input Oversize Undersize Oversize Undersize Urban Industrial
(screen) (screen) (cyclone) (cyclone) areas areas
Total PCBs (mg/kg d.w.) 0.0021 0.0042 0.0016 0.0001 0.001 0.006 5
Total PAHs (mg/kg d.w.) 0.105 0.341 0.410 0.032 0.062 10 100
Hydrocarbons C <12 (mg/kg 0.31 0.4 0.04 0.02 0.03 10 250
d.w.)
Hydrocarbons C >12 (mg/kg 12 100 11 9 19 50 750
d.w.)
Al (mg/kg d.w.) 8832 8226 9598 6786 9245
As (mg/kg d.w.) 6 5 6 5 6 20 50
Cd (mg/kg d.w.) 0.2 0.1 0.2 0.1 0.2 2 15
Cr tot (mg/kg d.w.) 28.3 11.3 31.8 22.8 33.1 150 800
Cu (mg/kg d.w.) 9.0 7.4 9.4 4.3 7.8 120 600
Ni (mg/kg d.w.) 21.9 9.7 23.0 18.4 24.7 120 500
Pb (mg/kg d.w.) 3.0 1.0 3.0 1.0 3.0 100 1000
V (mg/kg d.w.) 18.9 8.4 20.0 14.0 19.6 90 250
Zn (mg/kg d.w.) 36.7 29.5 38.7 26.7 38.1 150 1500
It can be concluded that the adopted configuration of the pilot plant, led to an excellent result,
confirming the data obtained at laboratory scale.
Concentration values of sediment and its fractions are below the regulatory threshold. So, a simple
sediment treatment (sieving and cycloning units) improves the quality of the sand fraction reducing
fines content and valuing aesthetic properties, even if is considered suitable for sand recovery
dredged sediment not subjected to treatment.
3.3.3 Cesenatico
The pilot plant test was performed considering these main operating parameters:
- slurry feed flow rate: 4298.60 kg/h;
- L/S = 3.6.
Table 13 reports the mass balance of the samples of slurry collected during the running of the pilot
plant. The size analysis of each single flow is also reported.
Table 13. Mass balance and particle size analysis of Cesenatico sediment. Pilot plant configuration: complete
Mass fraction (% Organic matter and Sand Silt Clay
w/w) gravel
Feed flow 100 9.60 44.43 30.24 15.73
Sieve oversize 11.09 64.90 19.00 11.30 4.80
Sieve undersize 2.70 47.60 32.60 17.10 2.70
Cyclone under (recovered sand under spiral classifier 80.13 0.25 58.67 34.65 6.43
and under second cyclone)
Cyclone 2 over (fine fractions) 8.78 0.00 5.00 52.20 42.80
The performance of the plant in its complete configuration was not fully satisfactory: the recovered
material under the spiral classifier and under the second hydrocyclone is about 80% of the total
sediment mass, but the effectiveness of the separation processes was rather low, because the
recovered product contains only 58.67% of particles classifiable as sand and the presence of
undesired fine particles is really remarkable (34.65% of silt and 6.43% of clay). So, this material could
not be reused for beach nourishment.
Table 14 shows the main results of chemical characterization.
Table 14. Chemical characterization of Cesenatico sediment separated during the pilot treatment.
Values higher than regulation limits for urban areas are in orange bold.
Analytes Fractions Law Limits
(D.Lgs. 152/06)
Input Oversize Undersize Oversize Undersize Undersize Oversize Undersize Urban Industrial
(screen) (screen) (screw) (screw) (cyclone 1) (cyclone 2) (cyclone 2) areas areas
Total PCBs (mg/kg 0.0139 0.0171 0.0223 0.0028 0.00294 0.152 0.0182 0.0168 0.006 5
d.w.)
Total PAHs (mg/kg 2.78 2.76 2.51 0.69 4.64 0.39 6.52 0.68 10 100
d.w.)
Hydrocarbons C <12 0.04 0.37 0.25 0.04 - 0.11 0.03 0.08 10 250
(mg/kg d.w.)
Hydrocarbons C >12 159 125 68 71 - 173 30 80 50 750
(mg/kg d.w.)
Al (mg/kg d.w.) 33710 7899 35180 38433 39387 30980 34390 40570
As (mg/kg d.w.) 20 9 21 8 33 9 25 12 20 50
Cd (mg/kg d.w.) 0.6 0.19 0.55 0.23 0.54 0.56 0.58 0.68 2 15
Cr tot (mg/kg d.w.) 78.50 23.30 82.80 31.60 77.80 87.80 79.60 108 150 800
Cu (mg/kg d.w.) 45.60 15.30 48.10 14.30 42.80 53.4 40.4 71.9 120 600
Ni (mg/kg d.w.) 58.70 19.00 59.70 25.10 57.0 62.40 64.90 62.10 120 500
Pb (mg/kg d.w.) 15.00 4.00 19.00 5.00 26.01 12.00 16.00 17.00 100 1000
V (mg/kg d.w.) 60.70 18.60 61.60 19.40 55.09 68.10 57.70 87.90 90 250
Zn (mg/kg d.w.) 119.0 39.4 122.0 53.9 119.0 125.0 115.0 162.0 150 1500
Analysis of hydrocarbons C >12 show some inconsistencies: the overflow concentration of cyclone is
lower than that detected in the underflow. As value is higher of law limits for urban areas.
Table 15. Mass balance and particle size analysis of Bellaria-Igea-Marina sediment. Pilot plant configuration: complete.
Mass fraction Organic matter and Sand Silt Clay
(% w/w) gravel
Feed flow 100 1.23 89.10 5.84 3.83
Sieve oversize 1.73 42.50 43.50 7.90 6.10
Sieve undersize 98.27 0.50 89.90 5.80 3.80
Cyclone under (recovered sand under spiral classifier and 49.80 1.47 95.28 3.25 0.00
under second cyclone)
Cyclone 2 over (fine fractions) 48.48 0.00 92.20 4.10 3.70
Despite the relevant sand content (>88%) recovery of the sandy material is very unsatisfactory and
approximately equal to 50 %. Table 16 shows the main results of chemical characterization.
Concentration values of sediment and its fractions are below the regulatory threshold.
Table 16. Chemical characterization of Bellaria-Igea-Marina sediment separated during the pilot treatment
Values higher than regulation limits for urban areas are in orange bold.
Analytes Fractions Law Limits
(D.Lgs. 152/06)
Input Oversize Undersize Oversize Undersize Undersize Oversize Undersize Urban Industrial
(screen) (screen) (screw) (screw) (cyclone 1) (cyclone 2) (cyclone 2) areas areas
Total PCBs (mg/kg 0.0010 0.0022 0.0013 0.0004 0.0007 0.0116 0.0005 0.0015 0.006 5
d.w.)
Total PAHs (mg/kg 0.25 0.43 0.04 0.04 0.04 0.29 0.11 0.14 10 100
d.w.)
Hydrocarbons C <12 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.03 10 250
(mg/kg d.w.)
Hydrocarbons C >12 11 11 11 11 14 10 10 18 50 750
(mg/kg d.w.)
Al (mg/kg d.w.) 12880 10370 10170 6456 7262 44910 5901 8558
As (mg/kg d.w.) 8 6 7 5 6 20 5 6 20 50
Cd (mg/kg d.w.) 0.29 0.23 0.27 0.22 0.27 0.66 0.23 0.25 2 15
Cr tot (mg/kg d.w.) 43.00 30.50 41.80 29.30 32.40 108.00 37.50 37.30 150 800
Cu (mg/kg d.w.) 9.80 11.0 8.10 7.60 6.20 37.00 5.50 6.80 120 600
Ni (mg/kg d.w.) 23.10 19.10 22.20 18.10 18.50 57.60 17.70 21.50 120 500
Pb (mg/kg d.w.) 3.00 3.00 3.00 3.00 2.00 13.00 2.00 3.00 100 1000
V (mg/kg d.w.) 25.80 21.30 21.10 14.20 15.70 96.60 13.30 18.70 90 250
Zn (mg/kg d.w.) 41.10 37.20 38.60 39.70 38.40 148.00 29.40 35.40 150 1500
CONCLUSIONS
Laboratory results show that Porto Cesenatico and Bellaria-Igea Marina sediments are characterized
by an undesirable fine fraction (< 63 m), between ~8 and ~33% and between ~64 and ~91%,
respectively. According with their size distribution, they show concentration values of some pollutants
higher than the Italian regulatory limits for residential use.
Many contaminants are associated with fine-grained and/or rich of natural organic matter material, so
Porto Garibaldi and Cervia sediments, being relatively poor in fine particles (fine fractions < 10%),
content pollutants below the regulatory limits. So, these sandy sediments (contents of sand > 80%)
could be directly reused after dewatering. However, a treatment train aimed at separating the fine
fraction is proposed in order to further improve the aesthetic and environmental characteristics.
For the reason above mentioned, on the sediments coming from Porto Garibaldi and Cervia harbours
the treatment tests were performed with the pilot plant set in reduced configuration (sieving and
hydrocycloning) and the results achieved can be considered positive: both the overall mass balance
and the particle size distribution of the separated sand were found to be suitable for such utilization
option. Instead, sediments coming from Cesenatico and Bellaria-Igea Marina harbours were treated
using the pilot plant set in full configuration but the performance was rather unsatisfactory, because of
content of undesired fine fractions (about 40%) and As concentration over regulatory limit for urban
areas for Cesenatico sediment and because of amount of recovered product (only 49.80%) for
Bellaria-Igea-Marina sediment.
Table 17 summarizes the separation efficiency achieved by running the pilot plant
Table 17. Summary of the separation efficiency achieved running the pilot plant
Sediment Pilot plant Recovered product (% Sand content in the recovered Undesired fine fractions content in the
configuration of total mass) product (% of total mass) recovered product (% of total mass)
Porto Garibaldi reduced 91.91 92.31 7.68
Cervia reduced 91.05 94.40 5.60
Cesenatico complete 49.80 95.28 3.25
Bellaria-Igea- complete 80.13 58.67 41.08
Marina
The results from the present study indicate that dredged harbor sediments from the pilot area have the
potential of providing materials of good quality for different purposes, including sand for beach
nourishment interventions and fine fractions for use as either fine aggregate or raw material for various
applications. In several cases, the original sediment should be processed before subsequent reuse, to
produce fractions of suitable quality from the viewpoint of both esthetic characteristics and chemical
composition.
REFERENCES
[1] ICRAM-APAT, 2007. Manual for the handling of marine sediments.
[2] Nota D.J.G., 1958. Sediments of the western Guyana shelf. Report of Orinoco shelf expedition, 2,
Mendedel. Landbomvhogedrool, Wegeningera, 98.
[3] Romano E., Bergamin L., Ausili A., Pierfranceschi G., Maggi C., Sesta G., Gabellini M., 2009.
The impact of the Bagnoli industrial site (Naples, Italy), on sea-bottom environment. Chemical
and textural features of sediments and the related response of benthic foraminifera, Mar. Pollut.
Bull., 59, 245-256.
[4] Shepard F.P., 1954. Nomenclature based on sandsiltclay ratios. J. Sed. Petr., 24, 151-158.
[5] Hyde T. I., Ash P. W., Boyd D. A., Randlshofer G., Rothenbacher K., Sankar G., 2011. X-Ray
Absorption Spectroscopic Studies of Platinum Speciation in Fresh and Road Aged Light-Duty
Diesel Vehicle Emission Control Catalysts, Platinum Metals Rev., 55(4), 233.
[6] ICRAM, 2001. Metodologie analitiche di riferimento. ICRAM - Ministero dellAmbiente e della
Tutela del Territorio.
[7] Decreto Legislativo n. 152, 3 aprile 2006. Norme in materia ambientale.
[8] Romano E., Dastoli S., Polettini A., Pomi R., Muntoni A., Villani B., Zucca A., 2013. Analytical
characterization and lab-scale treatment of dredged sediments from small italian harbours.
th
Proceedings of AquaConSoil 2013 - 12 International UFZ-Deltares Conference on Groundwater-
Soil-Systems and Water Resource Management.
[9] Polettini A., Pomi R., De Gioannis G., Muntoni A., Dastoli S., Romano E., Villani B., Morelli M.,
Peretti R., Zucca A. 2013. Lab- and pilot-scale treatment of dredged sediment by means of
enhanced washing separation. A.K. Bullard and E.A. Stern (Conference Chairs), Remediation of
Contaminated Sediments 2013. 7th International Conference on Remediation of Contaminated
Sediments (Dallas, TX, February 4-7, 2013).
Colloidal Fe-zeolites - A novel material for sorption-supported in-situ chemical oxidation
(ISCO)
Introduction:
Fe-containing zeolites are a promising material for the removal of organic contaminants from
groundwater, since they can be tailored for an optimal combination of two functions: I) Zeolites with
appropriate channel structure and SiO2/Al2O3 ratio have excellent adsorption efficiencies for small
organic molecules such as MTBE, BTEX or chlorinated solvents. II) The ion exchange sites located in
the pore channels and cages of zeolites allow dispersion and stabilization of isolated iron ions, which
are highly active in redox reactions. Thus, Fe-loaded zeolites have been shown to function as
heterogeneous Fenton-like catalysts over a wide pH range including neutral conditions [1-3].
With respect to ex-situ treatment of contaminated groundwater, Fe-zeolites can be an interesting
alternative to activated carbon since Fe-zeolite adsorbers offer the option of easy on-site regeneration
by flushing with H2O2 solution [1]. In the framework of the EU project NanoRem a novel concept for in-
situ chemical oxidation based on the application of colloidal Fe-zeolites is developed (Figure 1). The
basic idea is to apply a solid adsorbent and catalyst for Fenton-like oxidation in the form of a
suspension, which can be injected into the aquifer in an initial step, separate from the subsequent
addition of H2O2. The colloidal particles are transported over a certain distance and deposited on the
aquifer sediment, where they form an active zone in the preferred groundwater flow paths. Using a
stationary solid catalyst has the advantage to allow a mixing with the oxidant in the subsurface, i.e. at
the location of the contamination.
By this means, the injection of mixtures of catalyst and H 2O2, causing vigorous reactions and
thus safety issues known from conventional Fenton-based ISCO, could be avoided. In addition, prior
to oxidant injection the Fe-zeolite zone could be initially used as a sorption barrier. By this means
aqueous phase concentrations of contaminants are reduced and further spreading of plumes is
prevented. At the same time, the sorption barrier can enrich contaminants from a larger volume of
water before injecting the oxidant into it. This would correspond to an increased radius of influence of
the ISCO process and a more efficient utilization of H2O2.
Particle injection
Sorption zone
H2O2 injection
Oxidation
Results:
This contribution summarizes results from lab experiments on the selection and optimization of
Fe-zeolites with respect to transport and distribution in saturated porous media as well as adsorption
and catalytic oxidation of various groundwater contaminants (MTBE, trichloroethene (TCE), 1,2-
dichloroethane (DCA) and toluene). Structure-property correlations were derived by screening various
zeolite types differing in framework type and SiO 2/Al2O3 ratio for adsorption of model contaminants.
Channel diameter (determined by the framework type) and surface hydrophobicity (determined by
SiO2/Al2O3 ratio) are the most influential factors in this process. Even though high-silica zeolites have
a low ion-exchange capacity and thus can take up only limited amounts of iron ions, it was possible to
obtain sufficiently active catalysts for oxidation of adsorbed contaminants by H2O2 [3].
Soluble (modified) biopolymers were applied in order to obtain appropriate suspension stability of
the colloidal zeolites (Figure 2). Due to the fact that these soluble polymers are excluded from the
inner pore volume of the zeolites by virtue of their size, no significant adverse effects on contaminant
adsorption and catalytic performance of the Fe-zeolites are observed.
-1
Figure 2: Left: Percentage of particles remaining in suspension after 24 h (c Zeolite = 2.5 g L , 10 mM
KNO3) determined by turbidity measurement at a height of 19 mm below water table; Right: Powder
and suspension of Fe-BEA-35
Particle mobility was studied in 1D-column experiments using standard materials (porous media
and water) and protocols developed in the NanoRem project. For stabilized Fe-BEA-35 (the first
prototype Fe-zeolite selected), promising results on mobility were obtained (Figure 3), showing
breakthrough of 85% particle mass concentration from a 20 cm column (washed quartz sand 0.3 - 0.8
-1 -1 -1
mm, soft water, u = 10 m d , cparticle,in = 1 g L , cstabilizer = 1.5 g L ).
1.2
1
c(Fe-BEA-35)/c(0) [ ]
0.8
0.6
0.4
0.2
0
0 1 2 3 4 5
pore volume [ ]
Figure 4: Sorbed fraction (Left) and kinetics of degradation (Right) of various model contaminants at
-1 -1 -1 -1
cFe-BEA-35 = 50 g L , csubstrate = 25 mg L each, cstabilizer = 5 g L , start of reaction: cH2O2 = 8 g L H2O2,
pH = 7
In addition, column experiments simulating the cycle of catalyst infiltration and immobilization,
contaminant adsorption and degradation were conducted using MTBE as model contaminant. Fe-
BEA-35 which was loaded on washed quartz sand at a mass fraction of 1 wt% showed stable
adsorption and catalytic properties over three cycles of infiltration of MTBE-contaminated water (10 mg
-1 -1 -1
L MTBE in very hard water, u = 1 m d ) with intermittent regeneration by H2O2 infiltration (10 g L
-1
H2O2 in very hard water, u = 1 m d ).
Acknowledgements: This work was supported by funding from European Union within the NanoRem
project.
References:
[1] A. Georgi, R. Gonzalez-Olmos, R. Khler, F.-D. Kopinke, Separation Science and Technology, 45
(2010) 1579.
[2] R. Gonzalez-Olmos, F. Holzer, F.-D. Kopinke, A. Georgi, Applied Catalysis A: General, 398 (2011),
44.
[3] R. Gonzalez-Olmos, K. Mackenzie, F.-D. Kopinke, A. Georgi, Environmental Science and
Technology, 47 (2013), 2353.
[4] US EPA 2002. Methods for Measuring the Acute Toxicity of Effluents and Receiving Waters to
Freshwater and Marine Organisms, Fifth Edition. EPA-821-R-02-012. U.S. EPA (4303T),
Washington, DC, USA
[5] R.P. Schwarzenbach, P.M. Gschwend, D.M. Imboden, Environmental Organic Chemistry, John
Wiley and Sons, New York 1993.
[6] NDRL/NIST Solution Kinetics Database on the Web (http://kinetics.nist.gov/solution)
1c. Remediation technologies and approaches
1 2 1 1
Pilar Garca-Gonzalo , Ana Elena Pradas del Real , M. Carmen Lobo , Araceli Prez-Sanz
1
Department of Agro-Environmental Research, IMIDRA, Alcal de Henares, 28800, Spain.
email : garcia.gonzalo@madrid.org, carmen.lobo@madrid.org, araceli.perez@madrid.org
Telf.:+34 91 8879418
2
Geochimie 4D Group, ISTerre, Universit Grenoble I, 38041 Grenoble Cedex 9, France.
email: pradasda@ujf-grenoble.fr
Introduction
Chromium (Cr) pollution is due to large number of industrial processes such as chrome
plating, wood preserving, pigmenting, pulp and paper, textile dyeing, tanning and leather
processing. (Dhal et al., 2013). Its widespread use has converted Cr in a serious pollutant of air,
soil and water (Zayed and Terry, 2003). Alternative techniques for the clean-up of polluted soil
and water, such as the cost-effective and less disruptive phytoremediation, have gained
acceptance in recent years (Pilon-Smits, 2005). Plant species used for phytoremediation of
heavy-metal pollution may promote the growth of microorganisms, which in turn have capacity
to detoxify metals efficiently by transforming them into insoluble salts or relatively nontoxic
oxidation states (Kuiper et al., 2004). Improvement of the beneficial associations between
microorganisms and plants, particularly in the rhizosphere, is an area of research of global
interest.
Silene vulgaris natural populations were collected throughout Madrid Province and, after two
growing seasons, rhizome cuttings were used to propagate individual plants. Genetically
uniform clones were developed and vegetative propagated on permanent field plots at the El
Encn agricultural experiment station (Alcal de Henares, Spain: 4031N, 319W). Two
genotypes from two different populations were selected: genotype SV21 (Rozas de Puerto
Real; tolerant) and genotype SV38 (Valdemaqueda; sensitive), according to the different
response to Cr(VI) (Pradas del Real et al., 2013).
A pot experiment was conducted under greenhouse conditions (humidity 50-65%, average
air temperature 21C and natural light). Plastic pots were filled with 17 kg of agricultural soil
from El Escorial (Madrid), collected from the top layer (0-30 cm).The soil was air-dried, ground
and sieved (< 5 mm). Soil characteristics were analyzed according to the official soil analysis in
Spain (MAPA, 1994). Basically,: pH 6.62, carbonates of 0.6%, electrical conductivity of 0.65
-1 -1
dS.m , total nitrogen content of 0.13%, available P of 55 mg.kg and organic matter of 2.67%.
The experiment was performed in a randomized block design with three replicates and a 2x3
factorial arrangement with two levels of pollution: i) no pollution, ii) K2Cr2O7 to simulate a Cr(VI)
spilling; and three vegetations treatments: i) no plant, ii) S. vulgaris genotype SV21 and iii) S.
vulgaris genotype SV38.
-1
Cr pollution was simulated spilling K2Cr2O7 solution (1000mg.Cr.L ) to reach a final
-1
concentration of Cr in pots of 100 mg.kg soil. The soil was bought to 60% water holding
capacity and maintained by addition of deionized water for three months (from September to
November). After soil consolidation, 10 cuttings of the corresponding genotype were
transplanted to pots and grown for five months. The total amount of water applied in each
irrigation cycle was based on the normal rainfall in Mediterranean conditions.
Rhizosphere soil was collected at flowering time from each of the three replicates. From each
pot three soil samples were bulked into a single sample. After the plants were uprooted and
shaken, rhizosphere samples were collected from soil that remained gently adhered to plan
roots, collected into sterile Petri dishes and store at 4C until DNA was extracted. Soil DNA was
extracted using the UltraClean Soil DNA Isolation Kit (Mo Bio Laboratories, Inc.). Bacterial 16S
rRNA gene fragments were amplified with the primer set 341F with a GC clamp (40-nucleotide
GC-rich sequence, 5-CCT ACG GGA GGC AGC AG-3) and 907R (5 CCG TCA ATT CMT
TTG AGT TT-3) specific for the domain Bacteria (Schfer and Muyzer, 2001). Amplicons were
analyzed by DGGE using 6% (w/v) acrylamide/bisacrilamide (37.5:1) gels containing a 40%-
60% linear gradient of formamide and urea (100% denaturing solution contained 40% (v/v)
formamide and 7M urea). The wells were loaded with roughly equal amounts of DNA (about 500
ng), and electrophoresis was carried out in 1 TAE buffer. The electrophoresis was run for 14h
at 100V and a constant temperature of 60C, using the Ingeny Phor-U system (Ingeny
International, Goes, The Netherlands). Gels were stained for 90 min in 1 TAE buffer with
SYBR Gold Nucleic Acid Gel Stain (1:10000; Invitrogen, Breda, The Netherlands). Images of
the gels were obtained by the Gel Documention System GENi (Syngene, Frederick, MS, USA)
and analysed using the software GeneTools v. 4.01. Bands were excised from the gels, eluted
in 20 l of nuclease-free water and incubated at 37C for 60 min. Aliquots (2 l) were re-
amplified by PCR for sequencing using the bacterial primers 341F and 907R without the GC-
clamp in the forward primer. The PCR products were purified using the NucleoSpin Gel and
PCR Clean up Kit (Macherey-Nagel).
Statistical analysis
The genetic diversity of each of the soil bacterial communities was determined by DGGE
profile data. Each detected band was defined as an operational taxonomic unit (OTU), and the
number of bands was defined as the species richness (S) of each sample (Bell et al., 2005). To
determine the diversity of bacterial communities, the Shannon-Wiener index of diversity (H)
(Shannon and Weaver, 1963) and Pielous evenness index (EH)(Pielou, 1966) were calculated
according to the formulas: H=- piln(pi) and EH =H/Hmax =H/lnS. Where, pi is the proportion
of the bands in the track (it was calculated as follows: pi = ni/N; where ni is the intensity of band
i in the densitometric curve, and N is the sum of the intensities of all bands) and S is the number
of bands in the track.
Similarity among DGGE profiles was estimated using the Dice correlation coefficient to
compute the distance matrix. The unweighted pair group with mathematical averages method
(UPGMA) was used to build the similarity dendrogram from the distance matrix using the
NTSYS-pc software (Exeter Software, New York, NY).
Results
Bacterial DGGE profiles generated from the universal bacterial primers (341F and 907 RM)
revealed the structural composition of communities in rhizosphere soil samples (Fig. 1). For the
analysis, 16S rRNA fragments of Silene vulgaris genotypes SV-21 and SV-38 on untreated and
Cr treated triplicates pots were compared by running the reaction products in gel. Repeated
DGGE runs of the same PCR product as well as repeated PCR amplification of the same DNA
extract followed by DGGE produced similar banding profiles, indicating a high degree of gel
reproducibility. Each of the distinguishable bands in the separation pattern represents an
individual bacterial species (Luca et al., 2002).
Generally, DGGE patterns of the three replicates for each genotype and treatment were very
similar, indicating the reproducibility of the results. The number of prominent DGGE bands
ranged from 20 to 21 and 23 to 26 for SV-21 and SV-38 untreated samples respectively,
revealing the presence of a high number of bacterial taxa. Cr treated samples showed 18 visible
bands on genotype SV-21 and from 17 to 21 bands on genotype SV-38 with a decrease of
11.4% and 21.6% respectively, compared with untreated samples. This decrease in band
number was dependent on the contamination with Cr.
The DGGE patterns exhibited many dissimilarities between both genotypes on untreated and
Cr treated soils. In this regard few prominent bands were unique to genotype SV-21 (6,13) and
several number of bands were specific to genotype SV-38 (4,5,9,10,11,12,14,17). Conversely,
few common intensely stained bands were present in almost all the lanes ( 1, 2, 3, 7, 8, 15, 16).
Changes in relative intensity of individual DGGE bands were observed between treatments.
Thus, in Cr treated samples, enhanced intensity of bands 8 and 15 in genotype SV-21 and
bands 5, 7 and 17 in genotype SV-38, were observed.
Fig. 1. PCR-DGGE profiles of rhizosphere bacterial communities of genotypes SV-21 and SV-38. Samples
designations are indicate above each DGGE lane. The bands excised for sequencing are numbered from 1 to 17.
Clustering analysis of the DGGE profiles revealed the differences among the rhizosphere soil
samples (Fig. 2). A binary matrix showing the presence or absence of identified bands was
made for all gel lanes and this comparison did not take into account shifts in band intensities.
The DGGE profiles were separated into two major clusters with a similarity about 47%, profiles
of samples from genotype SV-21 cluster into one group and profiles from genotype SV-38 into
another group, showing that major differences were related to genotype effect. In addition
changes in bacterial communities were observed between Cr treated soils and controls in both
S. vulgaris genotypes. Thus, rhizosphere soil samples from both treatments cluster with a
similarity of 70% in SV-21 and about 65% in SV-38. Conversely, slight changes in bacterial
communities were found among repeats with similarities ranging from 78% to 100% in both
genotypes.
Fig. 2. Dendrogram of hierarchical cluster analysis of the DGGE fingerprint based on UPGMA and Jaccard coefficient.
Bacterial diversity of the microbial community was analyze by DGGE using the Shannon-
Wiener index (H). By computerized image analysis, H was calculated on the basis of the
bands on the gel tracks and their corresponding relative abundance indicated by band intensity
(Boon et al., 2002). The results in Table 1 showed the bacterial diversity indices of all the
rhizosphere soils sampled. For the untreated soils, H ranged from 1.57 to 1.72 in SV-21 and
from 2.10 to 2.78 in SV-38, whereas on chromium treated soils H ranged from 1.19 to 1.31 in
SV-21 and from 1.73 to 2.19 in SV-38. Regarding soil treatments, these results indicate a higher
bacterial diversity in rhizosphere of both genotypes on untreated soil samples and, regarding
plant genotypes, this results indicate a higher bacterial diversity in rhizosphere of genotype SV-
38 in both treatments. Data on species evenness on untreated and treated Cr soils respectively,
ranged from 0.52 to 0.57 and from 0.41 to 0.46 for genotype SV-21 and from 0.67 to 0.86 and,
from 0.61 to 0.73 for genotype SV-38. These results reflect differences in species abundance
that lead to an unequal representation of species in the bacterial community, being more
uneven in genotype SV-21 on both treatments.
Samples H S EH
SV-21 R1 1,722 21 0,57
SV-21 R2 1,604 20 0,54
SV-21 R3 1,577 21 0,52
SV-38 R1 2,104 23 0,67
SV-38 R2 2,789 26 0,86
SV-38 R3 2,461 25 0,76
SV-21 Cr R1 1,194 18 0,41
SV-21 Cr R2 1,293 18 0,45
SV-21 Cr R3 1,316 18 0,46
SV-38 Cr R1 2,091 21 0,69
SV-38 Cr R2 1,739 17 0,61
SV-38 Cr R3 2,197 20 0,73
Table 1. Shannon-Weiner index (H), Richness (S) and Evenness (EH) calculated from the DGGE banding pattern
Twenty prominent bands from the gel were excised and sequenced. Bands that were in the
same position in the DGGE gel were excised from different lanes and sequenced to confirm that
the same position corresponded to the same sequence. The position of the sequenced bands
are shown in Fig.1. Most of the excised bands produced legible DNA sequences but three
bands were discarded due to ambiguous chimerical nature. Sequence similarity values
compared to previously reported sequences were more than 98%. The phylogenetic analysis
(Fig. 3) showed that the sequences retrieved in this study were assigned to 16 different genera
and classified into five major class namely Bacilli (Bacillus and Cohnella), Beta-proteobacteria
(Acidovorax, Azohydromonas, Massilia, Ramlibacter and Variovorax), Gamma-proteobacteria
(Arenimonas, Dyella, Lysobacter, and Stenotrophomonas), Flavobacteria (Flavobacterium) and
Sphingobacteria (Chitinofaga, Pedobacter, Pontibacter and Terrimonas). All dominant groups
were present in the rhizosphere of both plant genotypes, with the exception of class
Flavobacteria only present in SV-38 .The 53% of sequences were classified into the beta and
gamma subdivisions of Proteobacteria, the 23,5% into the Sphingobacteria class, the 17,6%
into the Bacilli class and the 5,8% into the Flavobacteria class. At the species level ,differences
in the bacterial community composition were found between the two plant genotypes. In this
regard, eight sequences were specific to genotype SV-38 and were related to Flavobacterium
sp (DGGE-4), Terrimonas ferruginea (DGEE-9), Pontibacter sp (DGEE-5), Conhella sp (DGEE-
14), Lysobacter sp (DGEE-12), Stenotrophomonas maltophilia (DGEE-11), Acidovorax ebreus
(DGEE-10) and Ramlibacter sp (DGEE-17). Similarly, two sequences were specific to genotype
SV-21 and were related to Massilia timonae (DGEE-6) and Azohydromonas lata (DGEE-13).
Furthermore, increasing intensity of bands related to Chitinofaga niastensis (DGGE-7),
Pontibacter sp (DGGE-5) and Ramlibacter sp. (DGGE-17) were observed in Cr treated
genotype SV-38 and, Variovorax sp (DGGE-15) in Cr treated genotype SV-21. These data
indicated a shift in the relative abundance of these sequences of each Cr treated genotype.
Fig. 3. Phylogenetic tree of partial 16S rRNA gene sequences in the rhizosphere of S. vulgaris. Numbers at branch
points indicate bootstrap percentages, and branches have been collapsed for values < 50%.
Discussion
Biotic and abiotic factors such as soil type, seasons, heavy metal pollution, plant
development stage, root architecture, plant species, cultivars and genotypes can affect the
structure of microbial communities in the rhizosphere (Marschner et al., 2001; Wieland et al.,
2001; Berg and Smalla, 2009). The PCR-DGGE cluster analysis of 16S rDNA fragments
showed that the community structure was affected by plant genotype and, to a lesser extent, Cr
contamination, the first being the most influential factor shaping the rhizosphere microbiome.
The bacterial DGGE profiles showed the highest dissimilarity (47%) between samples from
different plant genotypes, being these results in accordance with the different bacterial
community composition observed between SV-21 and SV-38, after sequencing DGGE bands.
The impact of plant species on rhizobacterial communities has been well documented but
recent studies have shown genotype-dependent influence on the entire rhizobacterial
community (Berg and Smalla, 2009; Micallef et al., 2009; Marques et al., 2014). This
phenomenon has been linked to plant genetic constitution (Neal et al., 1973) possibly through
differences in composition and concentration of root exudates compounds (Griffiths et al.,
1999). Furthermore, due to the allogamous nature of this species (Sloan et al., 2012), a high
level of allelic variation among clones is expected.
The microbial community was more diverse in the rhizosphere of both untreated genotypes
than in Cr contaminated samples, based on an inspection of diversity indices. Shannon diversity
index (H), richness (S) and eveness (EH), as calculated from pooled DGGE data, were lower in
Cr treated samples of SV-21 and SV-38 suggesting that Cr contamination changed the structure
of microbial community to a certain extent. The effect of heavy metal toxicity on soil bacterial
communities varies according to the particular heavy metal involved, with Cr being the most
toxic on soil microbial metabolic process (Wang et al., 2010). According to previous reports,
heavy metals affects the qualitative and quantitative structure of soil microbial communities,
resulting in decreased metabolic activity and diversity (Giller et al., 1998). Under heavy metal
stress many microbial species could not cope with Cr (VI) pollution and competitive species can
predominate resulting in a lack of diversity and therefore, the decreases in diversity may reflect
changes in the community composition. However, when comparing diversity indices data from
the two plant genotypes, a higher diversity was found on sensitive genotype SV-38. Considering
that bacterial communities characterized by strong species diversity can better face challenging
environment perturbations than those of low species diversity (Degens et al., 2001), these
results can be explained by assuming that only a portion of the soil microbial pool is in a
metabolic active state (Pennanen et al., 2004). Furthermore, it has been shown that rhizosphere
microbial communities can be affected by root architecture (Kuske et al., 2002; Marschner et al.,
2002). Between the genotypes used in this study, differences in root system architecture were
strongly evident (data not shown), including characteristics as primary root length and degree of
lateral root formation.
References
Abou-Shanab RAI, van Berkum P, Angle JS (2007) Heavy metal resistance and genotypic
analysis of metal resistance genes in gram-positive and gram-negative bacteria present in Ni-
rich serpentine soil and in the rhizosphere of Alyssum murale. Chemosphere 68:360-367
Abou-Shanab RA, Ghanem K, Ghanem N, Al-Kolaibe A (2008) The role of bacteria on heavy
metal extraction and uptake by plants growing on multi-metal contaminated soils. World J
Microbiol Biotechnol 24:253-62
Anderson J, Hooper M, Zak J, Cox S (2009) Molecular and functional assessment of bacterial
community convergence in metal-amended soils. Microb Ecol 58:10-22
Bell T, Ager D, Song JI, Newman JA, Thompson IP, Lilley AK (2005) The contribution of species
richness and composition to bacterial services. Nature 436:1157-1160
Berg G, Smalla K (2009) Plant species and soil type cooperatively shape the structure and
function of microbial communities in the rhizosphere. FEMS Microbiol Ecol 68:1-13
Degens BP, Schipper LA, Sparling GP, Duncan LC (2001) Is the microbial community in a soil
with reduced catabolic diversity less resistant to stress or disturbance?. Soil Biol Biochem
33:1143-1153
Dhal B, Thatoi HN, Das NN, Pandey BD (2013) Chemical and microbial remediation of
hexavalent chromium from contaminated soil and mining/metallurgical solid waste: A review. J
Hazard Mater 250-251:272-291
Ernst WHO, Nelissen HJM (2000) Life-cycle phases of a zinc and cadmium resistant ecotype of
Silene vulgaris in risk assessment of polymetallic soils. Environ Pollut 107:329-338
Giller KE, Witter E, McGrath SP (1998) Toxicity of heavy metals to microorganisms and
microbial processes in agricultural soils: a review. Soil Biol Biochem 30:1389-1414
Griffiths BS, Ritz K, Ebblewhite N, Dobson G (1999) Soil microbial community structure: effects
of substrate loading rates. Soil Biol Biochem 31:145-153
Kuffner M, De Maria S, Puschenreiter M, Fallmann K, Wieshammer G, Gorfer M, Strauss J,
Rivelli AR, Sessitsch A (2010) Culturable bacteria from Zn- and Cd-accumulating Salix caprea
with differential effects on plant growth and heavy metal availability. J Appl Microbiol 108:1471-
1484
Kuiper I, Lagendijk EL, Bloemberg GV, Lugtenberg BJJ (2004) Rhizoremediation: a beneficial
plant-microbe interaction. Mol Plant Microbe Interact 17:6-15
Kuske C R, Ticknor LO, Miller ME, Dunbar JM, Davis, JA, Barns SM, Belnap J (2002)
Comparison of soil bacterial communities in rhizospheres of three plant species and the
interspaces in an arid grassland. Appl Environ Microbiol 68:1854-1863
MAPA (Ministerio de Agricultura), 1994. Mtodos Oficiales de Anlisis, Ed: Secretara General
Tcnica. Vol III, Spain.
Marques J.M., da Silva,T.F., Vollu, R.E., Blank, A.F., Ding, G.C., Seldin, L., Smalla, K., 2014.
Plant age and genotype affect the bacterial community composition in the tuber rhizosphere of
field-grown sweet potato plant. FEMS Microbiol. Ecol. 88, 424-435.
Marsden-Jones, E.M., Turrill, W.B., 1957. The Bladder Campions (Silene maritima and S.
vulgaris). The Ray Society, London.
Marschner, P., Yang, C.H., Lieberei, R., Crowley, D.E., 2001. Soil and plant specific effects on
bacterial community composition in the rhizosphere. Soil Biol. Biochem. 33, 1437-1445.
Marschner, P., Gnter, N., Angelika, K., Laure, W., Reinhard, L., 2002. Spatial and temporal
dynamics of the microbial community structure in the rhizosphere of cluster roots of white lupin
(Lupinus albus L.). Plant Soil 246, 167-174.
Martnez-Iigo, M.J., Prez-Sanz, A., Ortiz, I., Alonso, J., Alarcn, R., Garca, P., Lobo, M.C.,
2009. Bulk soil and rhizosphere bacterial community PCR-DGGE profiles and -galactosidase
activity as indicators of biological quality is soils contaminated by heavy metals and cultivated
with Silene vulgaris (Moench) Garcke. Chemosphere 75, 1376-1381.
Micallef, S. A., Shiaris, M. P., Colon-Carmona, A., 2009. Influence of Arabidopsis thaliana
accessions on rhizobacterial communities and natural variation in root exudates. J. Exp. Bot.
60, 1729-1742.
Neal, J., Ruby, J., Atkinson, T., Larson, R., 1973. Changes in rhizosphere populations of
selected physiological groups of bacteria related to substitution of specific pairs of
chromosomes in spring wheat. Plant and Soil 39, 209-212.
Odum, E.P., 1985. Trends Expected in Stressed Ecosystems. Bioscience 35, 419-422.
Paliouris, G., Hutchinson, T.C., 1991. Arsenic, cobalt and nickel tolerances in two populations of
Silene vulgaris (Moench) Garcke from Ontario, Canada. New Phytol. 117, 449-459.
Pennanen, T., Caul, S., Daniell, T.J., Griffiths, B.S., Ritz, K., Wheatley, R.E., 2004. Community-
level responses of metabolically-active soil microorganisms to the quantity and quality of
substrate inputs. Soil Biol. Biochem. 36, 841-848.
Pielou, E. C., 1966. The measurement of diversity in different types of biological collections. J.
Theor. Biol. 13, 131-144.
Pilon-Smits, E., 2005. Phytoremediation. Annual Review of Plant Biology 56, 15-39.
Pradas-del-Real, A.E., Garca-Gonzalo, P., Alarcn, R., Gonzlez-Rodrguez, A., Lobo, M.C.,
Prez-Sanz, A., 2013. Effect of genotype, Cr(III) and Cr(VI) on plant growth and micronutrient
status in Silene vulgaris (Moench.). Span. J. Agric. Res. 11, 685-694.
Schfer, H., Muyzer, G. S., 2001. Denaturing gradient gel electrophoresis in marine microbial
ecology. In: Paul J.H. (Ed.), Methods in Microbiology. Academic Press, London, pp. 425-468.
Schat, H., Vooijs, R., Kuiper, E., 1996. Identical major gene loci for heavy metal tolerances that
have independently evolved in different local populations and subspecies of Silene vulgaris.
Evolution 50, 1888-1895.
Shannon, C.E., Weaver, W., 1963. The mathematical theory of communication. University of
Illinois Press, Urbana, Illinois.
Sheik, C.S., Mitchell, T.W., Rizvi, F.Z., Rehman, Y., Faisal, M., Hasnain, S., McInerney, M.J.,
Krumholz, L.R., 2012. Exposure of soil microbial communities to chromium and arsenic alters
their diversity and structure. PLoS ONE 7: e40059.
Sloan, D.B., Keller, S.R., Berardi, A.E., Sanderson, B.J., Karpovich, J.F., Taylor, D.R., 2012. De
novo transcriptome assembly and polymorphism detection in the flowering plant Silene vulgaris
(Caryophyllaceae). Mol. Ecol. Resour. 12, 333-343
Smalla, K., Wieland, G., Buchner, A., Zock, A., Parzy, J., Kaiser, S., Roskot, N., Heuer, H.,
Berg, G., 2001. Bulk and rhizosphere soil bacterial communities studied by denaturing gradient
gel electrophoresis: plant-dependent enrichment and seasonal shifts revealed. Appl. Environ.
Microbiol. 67, 4742-4751.
Sderberg, K.H., Probanza, A., Jumpponen, A., Bth, E., 2004. The microbial community in
the rhizosphere determined by community-level physiological profiles (CLPP) and direct soil-
and cfu-PLFA techniques. Appl. Soil Ecol. 25, 135-145.
Wang F, Yao J, Si Y, Chen H, Russel M, Chen K, Qian Y, Zaray G, Bramanti E (2010) Short-
time effect of heavy metals upon microbial community activity. J Hazard Mater 173:510-516
Wieland, G., Neumann, R., Backhaus, H., 2001. Variation of microbial communities in soil,
rhizosphere, and rhizoplane in response to crop species, soil type, and crop development. Appl.
Environ. Microbiol. 67, 5849-5854
Xu, Z.Y., Tang, M., Chen, H., Ban, Y.H., Zhang, H.H., 2012. Microbial community structure in
the rhizosphere of Sophora viciifolia grown at a lead and zinc mine of northwest China. Sci.
Total Environ. 435-436, 453-464
Zayed, A.M., Terry, N., 2003. Chromium in environment: factors affecting biological remediation.
Plant Soil 249, 139-156.
Zhang, W., Huang, Z., He, L., Sheng, X., 2012. Assessment of bacterial communities and
characterization of lead-resistant bacteria in the rhizosphere soils of metal-tolerant
Chenopodium ambrosioides grown on lead-zinc mine tailings. Chemosphere 87, 1171-1178.
REMEDIATION OF WATER AND SOIL CONTAMINATED WITH AMOXICILLIN USING GREEN
ZERO VALENT IRON NANOPARTICLES
REQUIMTE, LAQV/Instituto Superior de Engenharia do Porto, Rua Dr. Antnio Bernardino de Almeida
431, 4200-072 Porto, Portugal. Phone: +351 228 340 537. Email: jtva@isep.ipp.pt
Abstract
Pharmaceutical products have been used for treatment of human and animal diseases and,
because of inappropriate handling and disposal it has been introduced in the environment (namely in
waters and soils), causing a potential risk for the ecosystems and for the human health. Most of the
pharmaceutical products are not completely removed by wastewater treatment plants originating soil
and water contamination being necessary to develop remediation technologies to reduce their
concentrations in the environment [1].
The use of zero valent iron nanoparticles (nZVI) represents one of the first nanoscale
technologies for environmental remediation. The nanoscale dimensions of zero valent iron (ZVI)
provide higher surface areas and therefore improved reactivity [2].
The nZVI can be produce through the traditional synthesis method that uses a toxic product
(sodium borohydride) or through a recent green synthesis method that are presenting promising
results to reduce/eliminate distinct substances hazardous from the environment. This green method
uses extracts from plants (leaves) that contain high concentrations of polyphenols (reducing agents)
that instantly react with iron (III) forming nZVI [3, 4]. The efficiency of ZVI (granular and as
nanomaterial) has been proven in studies focusing on the removal of amoxicillin and ibuprofen from
water [5, 6].
The work aims of this study are to investigate the degradability of amoxicillin present in water and
soil using green zero-valent iron nanoparticles produced by oak leaves. The results indicate that using
green nZVI can reduce the concentrations of amoxicillin.
Introduction
Several pharmaceutical products have been used for treatment of human and animal diseases,
specifically antibiotics. The pharmaceutical products are not totally absorbed by organisms, in some
cases they are completely excreted [1] as metabolites through urine and feces, being subsequently
incorporated into the environment in residual concentrations [7], dispersing and allowing an increase
in bacterial resistance in different settings [8, 9]. Some of these products have been found in several
environmental matrices such as surface, ground and drinking waters, aquatic sediments and soils [1,
10-14]. The inadequate disposal of these compounds can contaminate soils and waters that represent
direct or indirectly sources of food for humans and animals. Most pharmaceutical products are not fully
removed by wastewater treatment; thus it is necessary to apply preventive (such as reduction of use
or inadequate disposal) and corrective measures (reducing its concentrations in wastewater and soils
through remediation actions). One of the remediation technologies that are commonly used for this
type of contaminations is the chemical oxidation/reduction. There is, in the literature, several works
that report examples of this application [15-22], being the chemical nanoremediation a recent
example.
Nanomaterials and nanostructures have nanoscale dimensions that range from 1 to 100nm.
These structures have received a considerable attention in the environmental technologies [23] due
the fact that their developed nanomaterials have properties that allow chemical reduction, as well as
catalysis to degrade/transform the contaminants [24]. The dimensions of nanoscale increase the
surface area and the reactivity of iron particles [24, 25]. The iron nanoparticles represents one of the
first nanoscale technologies for environmental remediation [26]. The use of granular ZVI (the particle
s
size of ZVI goes from mm to cm scale) as a remediation started in the 90 and it was first employed in
permeable reactive barrier (PRB) to remediate contaminated sites [27]. However the progress of
nanotechnologies brought to environmental remediation the use of nanomaterials, leading to the
development of methods of production of new nanoremediation agents. The traditional methods for
the production of nZVIs involve either aggressive chemical reactions with toxic reagents (such as
sodium borohydride) in which hydrogen gas is produced (liquid method) or expensive mechanical
grinding (attrition method) [26]. The chemical procedure creates a large amount of toxic products, and
consequently, greener production methods have been developed and are presenting promising results
to reduce or eliminate hazardous substances from the environment [28]. These green methods use
extracts of plant parts (leaves, peels, fruits or branches) that contain high concentrations of
polyphenols. When these polyphenols (reducing agents) react with iron (III) nZVIs are produced [3, 4,
29-31]. Recently, the successful synthesis of green nanoparticles of silver and gold utilizing green
method using black tea leaves [32]; leaf extract (Hibiscus rosa sinensis) [33]; aleo vera plant extract
[34]has been reported. Other studies focus on the production of iron nanoparticles using green
method using green tea extract [3, 31]; black tea, grape marc, vine leaves extract [6]; eucalyptus leaf
extracts [35]; oolong tea extract [36]. Green synthesis of iron nanoparticles is evolving as a method
that would impart steric stabilization of iron nanoparticles against aggregation, and help overcome the
concerns related with using sodium borohydride as a reducing agent in the routine synthesis reported
so far.
The iron nanoparticles have several sizes and morphologies (spherical, platelets, nanorods,
triangular, ect) formed after the mixture of the extract of plants with iron (III) solution [29]. In recent
years nanoparticles (nZVI) have been used to the removal of various contaminants such as
halogenated organic compounds [26, 37], polycyclic aromatic hydrocarbons [38], organchlorine
pesticides [39] and metal ions [40].
The work aims of this study are to investigate the degradability of amoxicillin present in water and
soil and its degradation by green zero-valent iron nanoparticles is highlighted.
Experimental
Materials
Amoxicillin (900 g per mg) and iron (III) chloride hexahydrate (99%) were obtained from Sigma-
Aldrich. Type II deionizied water (resistivity = 15.0 M cm) and was further purified (resistivity=18.2
M cm) using a Simplicity 185 system (Millipore) was used throughout the study.
Figure 1 Results of the remediation tests of amoxicillin in aqueous solution using nZVI production
with oak leaves extract with different amounts.
This Figure shows that even using different nZVI concentrations almost complete degradation
efficiencies were attained, even with distinct kinetics. Attending to the results, using a nZVI
concentration of 0.2 mM is enough to completely degrade the amoxicillin in the aqueous solutions
(amoxicillin:iron (III) molar ration of 1:140).
This clearly shows that these nanomaterials have the capacity to completely degrade the amoxicillin
representing a possible alternative to other remediation technologies. Further work is being developed
in order to optimize the process and to use the green showednZVI as catalyst for the Fenton reaction.
These results show that the utilization of nZVI produced using green production methods have
enormous potential to remediate soils contaminated with pharmaceutical products. The values here
obtained show that after 300 min the contaminant reduced 50%. This indicates that the nZVI injected
to the soil sample reacted not only with the amoxicillin but probably also with soil components
reducing the reactive nZVI available to degrade the contaminant. This indicates that the soil chemical
and physical properties would influence the adequate amount of nZVI that is required to achieve the
degradation goals.
Conclusions
The results obtaining in this work allowed concluding that the nZVIs produced from oak leaves
extract degraded more than 95%of the amoxicillin in aqueous solutions in 66 hours showing that this
nanomaterial have high capacity to degrade amoxicillin. The soil experiment had to lower
degradations probable due to parallel reaction of nZVI with soil constituents or due to limitation on the
dispersion of the nZVI in all soil matrixes that could have left sure soil zones with no nZVI. Using a
greener remediation agent (green nZVI) and supplying better conditions for a post bioremediation
phase through the addition of a nutrient source (the natural extract).
Acknowledgments
This work received financial support from the European Union (FEDER funds through
COMPETE) and National Funds (FCT, Fundao para a Cincia e a Tecnologia) through projects
Pest-C/EQB/LA0006/2013 and PTDC/AAG-TEC/2692/2012.
References
E. Burcu zkaraova Gngr, Ayse Kuleyin, Feryal Akbal, M. Oya Orkun, Dilek Gm
Abstract
The Fenton Process is based on the production of the highly reactive hydroxyl radical (OH ), which is
resulting from the reaction between hydrogen peroxide (H 2O2) and ferrous ions Fe(II) under acidic
conditions. The oxidation system, which relies on Fentons reagent, can be employed to treat various
types of waters and wastewaters containing a range of organic pollutants like phenols, polycyclic
aromatic hydrocarbons, pesticides, formaldehyde, wood preservatives, plastic additives and rubber
chemicals. The treatment of polluted waters using Fenton process results in reduction of toxicity,
improvement in biodegradability, odour and colour removal. The Fenton process can also be
employed as a post- or a pre-treatment step in the treatment of high strength polluted waters with
extremely toxic and refractory nature. In the Fenton process, iron and hydrogen peroxide are the two
major chemicals that determine not only the operation costs but also the treatment efficacy (Zhang et
al., 2009). The objective of the present investigation was to examine the effectiveness of Fenton
Process in treating polluted waters of different origins. The performance of Fenton oxidation employed
in the treatment of soil washing solution, landfill leachate and phenolic water was investigated with an
aim of determining their optimum reaction conditions. For this purpose, all experiments were
performed in the batch system. The influence of H2O2, FeSO4 concentrations and reaction time on the
removal efficiency were investigated. The pH of reaction mixture was adjusted at the start of the
reaction. Required amounts of FeSO4 and H2O2 were added simultaneously into the solution and then
the mixture was shaken using a mechanical shaker. The progress of reaction was followed by
monitoring the disappearance of the contaminant and chemical oxygen demand (COD). The results
indicated that the Fenton process was successful in the treatment of polluted waters. Organic
pollutants (e.g. phenol, fluorene, etc) were efficiently removed by the Fenton process. Removal
efficiency depended on the reaction time and Fe(II) and H2O2 concentrations. 83% of phenol was
2+
degraded and 60% of COD was removed at conditions of H 2O2 500 mg/L, Fe 30 mg/L, phenol 250
mg/L and pH 3.0. Similarly, in the treatment of soil washing solution 40% of COD was removed under
optimum conditions, which were 2 hours reaction time, 2% H2O2 concentration and 1/50 Fe/H2O2 ratio.
In the treatment of landfill leachate, a COD removal of 66% was obtained for 5000 mg/L hydrogen
peroxide and 30 min reaction time.
Introduction
The Fenton Process is based on the production of the highly reactive hydroxyl radical (OH ), which is
resulting from the reaction between hydrogen peroxide (H 2O2) and ferrous ions (Fe(II)) under acidic
conditions. The oxidation system, which relies on Fentons reagent, can be employed to treat various
types of wastewaters containing a range of organic pollutants like phenols, polycyclic aromatic
hydrocarbons, pesticides, formaldehyde, wood preservatives, plastic additives and rubber chemicals.
The wastewater treatment using Fenton process results in reduction of toxicity, improvement in
biodegradability, odour and colour removal (Bagal and Gogate, 2014). The efficiency of process
depends on temperature, pH, reaction time, hydrogen peroxide and ferrous ion concentration (Benatti
and Tavares, 2012).
Fenton oxidation process is a catalytic reaction of H2O2 with iron ions that predominantly produces
OH radicals as the highly reactive oxidizing species (Rahim Pouran et al, 2014) (1). The reaction of
ferric ions (Fe(III)) with H2O2 generates other radicals with lower oxidation potentials (2) (Garrido-
Ramrez et al., 2010; Gallard and De Laat, 2001; Walling, 1975).
2+ 3+ -1 -1 -1
H2O2 + Fe Fe + OH + OH Ea= 39.5 kj mol k1 =76 M s (1)
3+ 2+ + -1 -1 -1
Fe + H2O2 Fe + HO2 +H Ea= 126 kj mol k2 =0.001-0.01 M s (2)
Various simultaneous reactions occur among the radicals and chemicals in the solute, that cause to
the scavenging of radicals as well. Hydroxyl radicals also react with the organics in solute starting a
chain reaction (Benatti and Tavares, 2012).
The treatment of high strength wastewaters with extremely toxic and refractory nature, is mostly an
arduous process. The Fenton process can be employed as a post- or a pre-treatment step. In the
Fenton process, iron and hydrogen peroxide are the two major chemicals that determine not only the
operation costs but also the treatment efficacy (Zhang et al., 2009). The objective of the present
investigation was to examine the effectiveness of Fenton Process in treating wastewaters of different
origins. The performance of Fenton oxidation employed in the treatment of soil washing solution,
landfill leachate and phenolic wastewater was investigated with an aim of determining their optimum
reaction conditions.
Materials and Methods
Experiments were performed in the batch system. The influence of H 2O2, FeSO4 concentrations and
reaction time on the removal efficiency were investigated. For the all experiments the pH of reaction
mixture was adjusted at the start of the reaction. Required amounts of FeSO 4 and H2O2 were added
simultaneously into the wastewater/soil washing solution/leachate and then the mixture was shaken
using a mechanical shaker. The progress of reaction was followed by monitoring the disappearance of
the contaminant and chemical oxygen demand (COD).
Each day more land on Earth is getting contaminated with different pollutants at different levels.
However, it is known that to determine the extent of pollution on contaminated land and to eliminate
pollution requires more time and major costs. Among the most important of these pollutants, polycyclic
aromatic hydrocarbons were discussed. PAHs, petroleum and petroleum derivatives, are compounds
which are released into environment by omissions, petrol spills and incomplete combustion of fossil
fuels. In the present study, soil contaminated with fluorene which is one of the 16 PAH compounds
was initially extracted with a non-ionic aliphatic surfactant. The soil washing solution was later on
treated with the Fenton process. The influence of treatment time, H2O2 concentration and Fe(II)/H2O2
ratio on the removal efficiency was investigated.
In the kinetic study the H2O2 concentration and Fe(II)/H2O2 ratio was kept constant, which was 2% and
1/50, respectively. Results show that the COD removal from the washing solution increased with a
decreasing trend and reached a maximum of 81% in 4 hours (Figure 1). The optimum treatment time
was selected as 2 hours.
100
Removal Efficiency (%)
80
60
40
20
0
0 100 200 300 400 500 600
Reaction Time (min)
Figure 1a. Effect of reaction time on COD removal efficiency
The influence of H2O2 concentration on COD removal was investigated with 0.5-10% H2O2, a
treatment time of 2 hours and a Fe(II)/H2O2 ratio of 1/50. Removal efficiencies were insignificant (>5%)
for H2O2 concentrations below 1%. It was interesting to notice that the COD removal approached 45%
with an increase to 2% H2O2. Almost all of the COD was removed with 10% H2O2 in 2 hours (Figure
1b).
100
Removal Efficiency (%)
80
60
40
20
0
0 2 4 6 8 10 12
H2O2 Concentration (%)
Figure 1b. Effect of different H2O2 concentrations on COD removal efficiency
The influence of Fe(II)/H2O2 ratio on COD removal was studied with Fe(II)/H2O2 ratios of 1/15, 1/30,
1/50, 1/100 and 1/200. Removal efficiency increased with higher ferrous iron dosages, which is
reflected by greater ratios. About 69% COD removal was achieved within 2 hours with 2% H 2O2 and a
ratio of 1/15 (Figure 1c).
100
Removal Efficiency (%)
80
60
40
20
0
0 1/50 1/25 3/50 2/25 1/10
Fe/H2O2
Recently, the applications of the electrochemical method in Fenton process, named electro-Fenton (E-
Fenton) method, have been reported (Zhang, et al., 2006). These studies can be divided into various
categories. In the first category, ferrous ion is externally applied, and both hydrogen peroxide and
ferrous ion are concurrently generated at the cathode. In the second category, hydrogen peroxide is
externally applied while a sacrificial iron anode is used as ferrous ion source (Lin and Chang, 2000). In
the third, hydrogen peroxide is externally applied and ferrous ion is electrogenerated via the reduction
of ferric ion or ferric hydroxide sludge (Chou et al., 1999). In the fourth category, both ferrous ion and
hydrogen peroxide are electrogenerated at sacrificial anode and cathode via the two-electro reduction
of sparged oxygen, respectively (Brillas and Casado, 2002).
In the remaining part of this article, the E-Fenton method is presented in which hydrogen peroxide is
externally applied while a sacrificial iron anode is used as ferrous ion source. This method is used for
the treatment of high COD strength landfill leachate. The effect of reaction time and added hydrogen
peroxide dosage on the efficacy of COD removal from the landfill leachate is investigated.
The effect of different hydrogen peroxide dosages on the COD removal efficiency is shown in Figure
2a. The COD removal efficiency was investigated at seven different values ranging between 250-
2
10000 ppm at 20 mA/cm current densities and 30 min electrocoagulation time. It was observed that
the COD removal efficiency increased with increased hydrogen peroxide dosage. However, the
removal efficiencies were approximately obtained at 5000 and 10000 ppm hydrogen peroxide dosage.
The optimum COD removal efficiency of 66% was obtained at 5000 ppm hydrogen peroxide dosage.
At this hydrogen peroxide dosage, the energy and electrode consumptions were 0.930 kWh/kg KO
and 2.011 kg COD/kg Fe, respectively (Figure 2b and Figure 2c). For this reason, the optimum
hydrogen peroxide dosage was determined as 5000 ppm and the electrocoagulation experiment was
carried out at four different values ranging between 15-60 min at this hydrogen peroxide dosage. The
effect of electrocoagulation time on COD removal efficiency at 5000 ppm hydrogen peroxide dosage is
shown in Fig. 2d. It can be seen that COD removal efficiency of 56.25% was obtained at 15 min, which
increased to 66%, 70%, 74.21% at 30, 45, 60 min, respectively. In this process, the organic
compounds removed two steps with electrocoagulation and coagulation (Bidga,1995).
100
COD removal efficiency (%)
80
60
40
20
0
0 2000 4000 6000 8000 10000
1,6
KWh/Kg COD
1,2
0,8
0,4
0,0
0 2000 4000 6000 8000 10000
H2O2 dosage (ppm)
3,0
2,5
Kg COD/Kg Fe
2,0
1,5
1,0
0,5
0,0
10 20 30 40 50 60
Time (min)
100
COD removal efficiency (%)
80
60
40
20
0
0 10 20 30 40 50 60
Time (min)
Industrial processes generate a variety of pollutants that may have negative impacts for ecosystems
and humans (toxicity, carcinogenic and mutagenic properties). Phenol is one of the most common
organic water pollutants, because it is toxic even at low concentrations, and also its presence in
natural waters can lead further to the formation of substituted compounds during disinfection and
oxidation processes (Busca et al., 2008).
In this study optimum reaction conditions of Fenton process for phenol removal were determined. The
effect of Fe(II) concentration on phenol degradation and COD removal was investigated within the
range of 15 and 60 mg/L at phenol concentration of 250 mg/L, H2O2 dosage of 500 mg/L, and pH 3.0.
It was found that the removal rate increased with increasing initial Fe(II) concentration. It was obtained
that phenol degradation efficiencies were 81.06%, 83.02%, 84.49% and 84.33% when Fe(II)
concentration was 15, 30, 45 and 60 mg/L respectively. It was obtained that COD removal efficiencies
were 58.75%, 60%, 63.75% and 64.38% respectively. It was also found that no further mineralization
occurred after adding more than a 30 mg/L Fe(II) dosage. Therefore in the next experiments 30 mg/L
constant amount of Fe(II) was used. Usually the rate of degradation increases with an increase in the
concentration of ferrous ion. Also, an enormous increase in the ferrous ions will lead to an increase in
the unutilized quantity of iron salts, which will contribute to an increase in the total dissolved solids
content of the effluent stream and this is not permitted (Babuponnusami and Muthukumar, 2014).
100
80
Removal efficiency (%)
60
40
Phenol COD
20
0
15 30 45 60
Fe (II) concentration (mg/L)
Figure 3a. Phenol and COD removal efficiency at different Fe(II) concentrations
The effect of H2O2 concentration on phenol degradation and COD removal was investigated at phenol
concentration of 250 mg/L, Fe(II) dosage of 30 mg/L and pH 3. It was obtained that phenol
degradation efficiencies were 74.35%, 83.02%, 83.92% and 88.74% when H 2O2 concentration was
250, 500, 750 and 1000 mg/L respectively. It was obtained that COD removal efficiencies were 55%,
60%, 60.63% and 65.63% respectively. The increase in H 2O2 concentration increased the phenol
degradation and COD removal.
Several studies show that the degradation of the pollutants increases with an increase in the dosage
of hydrogen peroxide. However, care should be taken while selecting the operating oxidant dosage.
The unused portion of hydrogen peroxide during the Fenton process contributes to COD and hence
excess amount is not recommended. Also, the presence of hydrogen peroxide is harmful to many of
the organisms and will affect the overall degradation efficiency significantly, where Fenton oxidation is
used as a pretreatment to biological oxidation. Another negative effect of hydrogen peroxide is the
scavenging of generated hydroxyl radicals, which occurs at large quantities of hydrogen peroxide
(Babuponnusami and Muthukumar, 2014). Thus, the dosage of hydrogen peroxide should be
optimised.
100
Removal efficiency (%)
80
60
40
Phenol COD
20
0
250 500 750 1000
Figure 3b. Phenol and COD removal efficiency at different H2O2 concentrations
Conclusion
The results indicated that the Fenton process was successful in the treatment of different wastewaters.
Organic pollutants (e.g. phenol, fluorene, etc) were efficiently removed by the Fenton process.
Removal efficiency depended on the reaction time and Fe(II) and H2O2 concentrations. 83% of phenol
2+
was degraded and 60% of COD was removed at conditions of H2O2 500 mg/L, Fe 30 mg/L, phenol
250 mg/L and pH 3.0. Similarly, in the treatment of soil washing solution 40% of COD was removed
under optimum conditions, which were 2 hours reaction time, 2% H2O2 concentration and 1/50
Fe/H2O2 ratio. In the treatment of landfill leachate, a COD removal of 66% was obtained for 5000 mg/L
hydrogen peroxide and 30 min reaction time.
References
Babuponnusami, A., Muthukumar, K., A review on Fenton and improvements to the Fenton process
for wastewater treatment, Journal of Environmental Ch emical Engineering, 2014, 2, 557572.
Benatti, C.T., Tavares, C.R.G., Fentons process for the treatment of mixed waste chemicals, in T.
Puzyn and A. Mostrag-Szlichtyng (Eds), Organic Pollutants Ten Years After the Stockholm Convention
- Environmental and Analytical Update, ISBN: 978-953-307-917-2, InTech, 2012, 247-270.
Bidga, R.J., Consider Fenton's chemistry for wastewater treatment. Chemical Engineering Progress,
1995, 91, 6266.
Brillas, E., Casado, J., Aniline degradation by Electro-Fenton and peroxi-coagulation processes
using a flow reactor for wastewater treatment. Chemosphere, 2002, 47, 241248.
Busca, G., Berardinelli, S., Resini, C., Arrighi, L., Technologies for the removal of phenol from fluid
streams: A short review of recent developments, Journal of Hazardous Materials, 2008, 160, 265288.
Chou, S.S., Huang, Y.H., Lee, S.N., Huang, G.H., Huang, C.P. Treatment of high strength hexamine-
containing wastewater by electro-Fenton method, Water Research, 1999, 33, 751759.
Gallard, H., De Laat, J., Kinetics of oxidation of chlorobenzenes and phenyl-ureas by Fe(II)/H2O2 and
Fe(III)/H2O2. Evidence of reduction and oxidation reactions of intermediates by Fe(II) or Fe(III).
Chemosphere, 2001, 42, 405413, ISSN 0045-6535.
Garrido-Ramrez, E.G., Theng, B.K.G, Mora, M.L., Clays and oxide minerals as catalysts and
nanocatalysts in Fenton-like reactions A review. Applied Clay Science, 2010, 47, 182192, ISSN
0169-1317.
Lin, S.H., Chang, C.C., Treatment of landfill leachate by combined electro-Fenton oxidation and
sequencing batch reactor method, Water Research, 2000, 34, 42434249.
Manisha V., Bagal, P.R,. Gogate Wastewater treatment using hybrid treatment schemes based on
cavitation and Fenton chemistry: A review, Ultrasonics Sonochemistry, 2014, 21,1-14.
Pouran S.R., Abdul Aziz A.R., Wan Daud, M.A.W., Review on the main advances in photo-Fenton
oxidation system for recalcitrant wastewaters, Journal of Industrial and Engineering Chemistry, 2014,
in press.
Walling, C., Fentons reagent revisited. Accounts of Chemical Research, 1975, 8, 125131, ISSN
0001-4842.
Zhang, H., Zhang, D., Zhou, J., Removal of COD from landfill leachate by electro-Fenton method.
Journal of Hazardous Materials B, 2006, 135, 106111.
Zhang, H., Choi, H.J., Canazo, P., Huang, C.P., Multivariate approach to the Fenton process for the
treatment of landfill leachate, Journal of Hazardous Materials, 2009, 161, 1306-1312.
The OHT process
Adsorption-based
The novel and innovative process developed by
technologies for innovative, OHT to extract the Hydrocarbons (HC) from their
An extensive test program has proven the OHT It must be emphasized that the OHT process
process efficiency on all types of HC studied and has two innovative characteristics:
therefore this process can be envisaged virtually in
In the OHT process, which is not based on
all the situations where HC have to be removed
a gravimetric principle, water is not the
from their environment. It also confirms that the
vector of the HC separation.
OHT process operates at ambient temperatures
The OHT process does not require large
and needs limited quantities of water.
quantities of energy and can be carried
OHT has been demonstrated on sites at pilot out at ambient temperature.
level. A pilot for produced water treatment was
tested in Germany for 3 months with success.
1
Adaptability of the OHT process to 800,0
HC concentration (mg/l)
600,0
In order to study the versatility of the OHT After treatement
500,0 458,1
process, several tests have been performed on a 421,6 421,1
366,0 373,3
400,0
large range of samples, from produced water to oil 323,3
Process advantages
Test parameters:
2
ThS 1D.2 Large scale inventories and strategies for dealing with contamination
Flowers 4 Brabant:
excellent projects for an area oriented approach in the Province of Noord-Brabant
waterboards
municipalities
provinces
ministry
Private companies
Public etc
rain
Surface water
Ground water
ABSTRACT:
The traditional way of dealing with polluted soil and groundwater does not work in all cases. Cities
are expanding, former gasworks, small scale industries are within present day city boundaries and
left there polluted footprint behind. Many historical groundwater pollutions are intermingled and the
polluter pays principle does not work in those cases, or they ask for endless legal procedures. The
polluter is either bankrupt or buried in the graveyard. As a result in urban environments with
complex contaminations spatial developments are at a standstill and neighborhoods at a social
decline. An area oriented approach can be used to tackle the problems.
Groundwater is everywhere
Dont move
In the province of North Brabant in the south of the Netherlands policy makers from a different
background try to look over there sectorial walls. In the project Flowers 4 Brabant specialists and
pioneers who work on an area oriented approach are brought together and with the support of the
ministry of Infrastructure and Environment they are sharing knowledge to stimulate an area oriented
approach.
Area oriented groundwater approach is a policy framework in which the integral and sustainable
design of the management of groundwater is performed within a restricted area. Groundwater
activities and interventions in the groundwater system are linked to environmental objectives,
(recovery) nature and spatial and economic developments are implemented and integrated in the
groundwater system for the long term. The policy framework for area oriented groundwater
approach will be determined by the relevant authorities (provinces, municipalities and/or
waterboards) through an administrative arrangement. From that point, the governing bodies will
manage the area oriented groundwater approach in the form of an implementation plan. This plan
also operate as a reference for the authorisation and decisions within the existing laws and
regulations aimed at soil, groundwater and the use of geothermal energy.
EU guideliness
must be met Therefore your
rights will be
very expensive
The Dutch soil protection act has been amended on 1 July 2012 to make an area oriented
groundwater approach legally possible. The area oriented approach involves the polluted
groundwater being allowed to freely move around inside the system area, providing that it does not
engender risks to people or the ecology and does not spread outside the designated area. The area
oriented approach therefore contributes in a general way to the realization of the objectives of the
Groundwater Directive. According to article 5 plumes has to be monitored and article 6 prohibits the
introduction of pollutants into the groundwater. The area oriented approach is par excellence made
to prevent spreading of plumes outside the contaminated area and also addresses the pollution
source as an essential part of the approach.
Sustainable urban development and soil quality are connected.
A poor groundwater quality can form a risk and a limiting factor for desired urban developments. A
badly designed urban development can form a long term risk for sustainable use of the subsoil, the
environment and the health of its inhabitants. But the other way around urban re-developments are
important triggers and a chance to restore soil quality. Also a good sustainable urban design concept
can draw new customers and increase value and profit. So, groundwater quality, sustainability and
urban development are connected. Can a win-win situation be reached? An area oriented approach
which allows for more synergies through early connecting and overlapping different disciplines in
urban planning is thought to be the most effective approach. But it is complex matter and actors
involved have to collaborate and will need guidance to achieve an successful area oriented approach.
In general local governments will take the role of director and organize the public private
partnerships.
1. The province sees that many projects are bogged down because there is much unclear about
finances, responsibility of the different actors (public and private), and who should pay for what.
2. An area oriented approach will achieve more than just solving a groundwater problem. The
province, municipalities and other actors see advantages for sustainable re development.
The province of North Brabant recognises area oriented approach therefore as an effective tool to
address soil and groundwater contamination, and it is also a tool to (re)start social and spatial
developments which where stagnated due to groundwater contamination. Therefore the province
developed an administrative Implementation Strategy for the area oriented approach.
Confidentiality and cooperation with other actors is an important key. Transparency and balancing
the different interests, which leads to a broad-based approach for all actors is the foundation of the
Implementation Strategy of Brabant. By placing assignments in a broader context, to restore
confidence and link budgets new initiatives germinate. This can been seen at various places in the
province where the area oriented approach is applied and solutions are developed to tackle long
term stagnations and economic growth flourish. Flowers 4 Brabant has demonstrated that an area
oriented approach is still human work. Enthusiastic pioneers are needed, people who have the
courage to step out of there comfort zone.
Biological treatment of micropollutants in drinking water
resources
Marlea Wagelmans1, Janneke Wittebol1
1. Introduction
Contamination of drinking water resources is becoming a threat that is particularly widespread. Nowadays
even in European countries clean drinking water is at risk. Pharmaceuticals, pesticides and other
micropollutants are emerging substances in surface and groundwater causing contamination of drinking
water resources and ultimately to closing down groundwater abstraction wells.
2. Background
Pesticides and other micropollutants in groundwater and surface water are more and more present in very
low concentrations, in the nanogram to microgram per litre range. Concentrations already exceed the EU
limit value for pesticides (or other micropollutants) in drinking water recourses. Closure of the groundwater
abstraction wells or the entailing treatment is costly. It is important to find a sustainable and cost-effective
remediation technique since it remains unknown if there are long term cumulative dose-additive or
synergistic effects of low concentrations of substances occurring as a mixture. Knowledge of degradation
processes at such low concentrations is limited and novel approaches are needed to develop biological
treatment technologies that are efficient at these low concentrations. BIOTREAT is a European project in
which urgently needed sustainable biotechnologies are developed for remediation of drinking water
resources contaminated with micropollutants such as pesticides and pharmaceuticals and their metabolites.
The basis of the technologies is bioaugmentation, which in this case is the introduction of specific degrading
microorganisms or microbial consortia into existing sand filters at waterworks or mobile filters placed close to
the (groundwater) abstraction well. A strategy using metabolic degradation pathways was used to target
water pollution cases where metabolically degrading bacteria are available. In addition a strategy using co-
metabolic processes has been investigated to target pollutants for which suitable and specific degrading
bacteria have not yet been described (not described in this abstract).
3. Methods
Many compounds are exceeding EU standards. The compound BAM (2,6 Dichlorobenzamide) has been
chosen as model compound for metabolic biological degradation. BAM is a metabolite of the broadly used
herbicide dichlorobenzonitrile or dichlobenil. The bacterium Aminobacter sp. MSH1 is found to be capable to
mineralize BAM.
A strategy using metabolic degradation pathways was used to target water-pollution cases where
metabolically degrading bacteria are available. First degradation tests were performed at lab scale. In order
to simulate drinking water production at waterworks we have up scaled the lab-scale batch experiments to a
sand filter column experiment in our lab and finally to a medium scale sand filter column that is used as
experimental column at a drinking water well. The BAM mineralizing bacterium was found capable to
degrade the metabolite in batch culture and sandfilter column experiments at low concentrations. Pilot scale
sand columns have been designed by Bioclear based on these results which were operated by project
partners in Belgium and Denmark.
Figuur 1: The principle of metabolic and cometabolic processes being exploited within BIOTREAT. A strategy for
exploiting metabolic degradation pathways was used to target water pollutions cases where metabolically
degrading bacteria are available, while a strategy using cometabolic processes was used to target pollutants for
which suitable degrading bacteria have not yet been described.
A Life Cycle Analyses (LCA, not described in this abstract) and a cost benefit analyses (CBA) have been
performed in order to gain insight in the sustainability and cost effectiveness of the developed technologies.
For these analyses four scenarios were taken into account:
- Bioaugmentation: addition of bacteria to existing sand filters
- Bioaugmentation with carriers: use of carriers (clay pebbles coated with alginate containing bacteria)
to add bacteria to existing sand filters
- Granulated active carbon (reference)
- Well relocation (reference)
4. Results
Based on the results of the laboratory tests, a total of six submerged sandfilter columns (see figure 2) were
operated, of which two none-inoculated controls and two columns that were exposed to 1 g/L BAM and two
to 30 g/L BAM (all inoculated with MSH1). Start-up retention time was 120 hours and the operation time of
the experiment was 240 hours. No back-flushing was applied. Results show that in the 1 g/L BAM exposed
columns, BAM degradation up to 72% was established. In the 30 g/L BAM exposed columns, 99%
degradation was achieved.
Figure 2: Schematic view and picture of lab-scale sandfilter columns used at Bioclear
In a subsequent experiment, BAM removal was investigated at relevant retention times (20 minutes to 1
hour) during 75 days. Also the effect of additional carbon source (acetate) was investigated. In this
experiment 10 g/L BAM and 1 mg/L BAM were tested. From this experiment (see figure 3) it was concluded
that:
- That degradation of BAM was not improved by providing acetate as an additional C-source.
- Degradation was only seen in the first couple of hours.
- Cell loss due to wash out was high.
Figure 3. Results of sandfilter column experiment performed at Bioclear within the operation period with a
retention time of 1 hour.
High densities of bacteria need to be and remain present in the sand filters in order to continuously remove
BAM from the raw water. Bioaugmentation of the sand filters might therefore be insufficient to remain the
right conditions. Continuous dosing of bacteria might solve this problem, as well as dosing of bacteria via so
called carrier systems. These carriers are clay pebbles, coated with alginate containing the bacteria. This
prevents the bacteria from quick wash out while they remain present in the sand filters in order to
continuously degrade BAM.
Based on the available data it is concluded that the well relocation scenario is the cheapest scenario. But
because the length of the transportation pipe has a large effect on the costs of the well replacing scenario
costs, the well replacement scenario is only cheaper if the new well is placed within a few kilometers from the
3
water production plant (3.5 kilometers for the 800,000 m per year scenario). If clean water is more than a
few kilometers away, implementing a BAM removal strategy becomes the best solution.
For the larger drinking water production volumes the GAC filter is the cheapest BAM removal technique. The
BIOTREAT technology, especially the BIOTREAT bio-augmentation technology + carrier is the most
attractive technology for small scale drinking water production plants due to the relatively low investment
costs.
Compared to traditional physicochemical technologies, biological technologies have the advantage that:
- Natural resources are used (normal expanded clay, alginate, bacteria).
- No harmful chemicals are used, use of contaminated groundwater for drinking water production is
made possible which otherwise might remain unused.
- The application is low cost compared to traditional physicochemical technologies.
- The technology is plug and play. No additional infrastructure is needed at the waterworks.
- Compound specific removal is possible.
3
Based on a production facility of 800.000 m per year the operating costs for these technologies,
3 3
bioaugmentation and bioaugmentation combined with carriers, are 0.20/m and 0.11/m respectively.
5. Biotreat partners
GEUS, DTU, University Leuven, EAWAG, University Gent, Bundesanstalt fr Gewsserkunde, Avecom, LCA
2.0, Bioclear
6. References
[1] Benner J. D.A, Helbling, H-P. E. Kohler, J. Wittebol, E. Kaiser, C. Prasse, T.A. Ternes, C.N. Albers, J.
Aamand, B. Horemans, D. Springael, E. Walravens, N. Boon. Is Biological treatment a viable alternative
for micropollutant removal in drinking water treatment processes?, Water Research 47, 5955-5976, 2013.
SESSION 1D.2: REGIONALLY APPROACHED GROUNDWATER MANAGEMENT IN ZWOLLE:
PREVENTING RISKS AND UTILIZING OPPORTUNITIES
Keywords: groundwater management, groundwater quality, risked based, urban and industrially
contaminated land, sustainable development, integrated solution
The approach of the contaminated groundwater in the surroundings of the railway station in
Zwolle was the first regionally approached groundwater management in the Netherlands.
Abroad there is quite some scepsis about this approach, which is used in the Netherlands
more and more. With this project we would like to demonstrate that this approach is much
more than letting the pollution flow. Its about making well considered, responsible decisions
to prevent risks and exploit opportunities in this case integral measures with energy and green
urban development.
The soil and the water in the surrounding of the railway station in the city centre of Zwolle are
contaminated because of many industrial activities in the past. The various pollutions were too
complex and too large to investigate and remediate each location separately. In 2003 a cooperation
agreement between public and private partners was formulated, in order to protect drinking water and
to remediate the existing groundwater pollution.
Measures, monitoring and results
This approach has been implemented over the past decade. The measures included the relocation of
some drinking water wells, the interception of the contaminated groundwater and remediation of the
soil. For the interception of the contaminated groundwater, existing infrastructure of the former
drinking water extraction are used. In the polluted source area, contaminated groundwater is
extracted, thereby preventing it from flowing to the drinking water wells. The heat from the extracted
groundwater is used as an energy source for office units nearby. After extracting the heat, the
pumped groundwater is aerated and drained to an ditch. Because of the purifying effect of reed
plants (constructed wetland) in this ditch, the pollution will be further biodegraded.
Goals achieved
Now, after 10 years, the remediation goals are achieved. The groundwater quality has significantly
been improved and the drinking water wells are partly relocated , by which the supply of clean
drinking water for the inhabitants of Zwolle can be guaranteed.
By making use of this approach, the municipality of Zwolle aimed to contribute to the sustainable
development of the city centre. This approach turned out to be the most sustainable solution: the
optimum for people, planet and profit.
The contamination
Companies take care Contaminated ground-
was found in a
of source removal in water is used in this
drinking water
the contact layer. building as a source
extraction well at the
Zwolle controls the of energy (heat/ cold).
Engelse Werk.
groundwater plumes
City of Stuttgart, Department for Environmental Protection, Gaisburgstr. 4, 70182 Stuttgart, Germany,
+4971121688725, sandra.vasin@stuttgart.de, hermann.josef.kirchholtes@stuttgart.de
Stuttgart is one of the few large cities in Europe, where mineral water springs and is widely used for
spa and medical purposes. Nineteen springs discharge about 44 million liters of highly mineralized and
carbogaseous water each day. Mineral springs of Stuttgart are, just after Budapest, the second largest
European sources. Due to their vulnerability, they are protected as economical, natural and cultural
assets [Ufrecht 1999].
The usage of volatile chlorinated hydrocarbons (CHC) in Stuttgart over several decades has caused a
severe groundwater contamination as well as a contamination of the mineral springs. As a
consequence this has initiated the intensive site specific investigation and remediation measures,
which have started thirty years ago. These measures have resulted in removal of about 25,000 kg
CHC in the inner city of Stuttgart [Landeshauptstadt Stuttgart 2003]. Despite these tremendous efforts,
the contamination of the deeper aquifers has not changed significantly. Since 1988 no sustained
decline in CHC concentration in the mineral springs could be observed and they are still affected at a
low level. This indicated that the mechanisms of lateral and vertical CHC migration between the
complex structrued layers of the Keuper and the minearal water-bearing layer of the upper
Muschelkalk were not fully understood. Therefore, the applied practices of environmental management
and measures for mitigation of contamination sources were not sufficient and had to be adjusted.
Building on these premises, the management plan for groundwater remediation applies the concept of
integral management of contaminated sites. It focuses on the spatiotemporal analysis of CHC
migration and the associated degradaton and transformation processes between the contamination
sources and the mineral springs. Furthermore, the management plan is considering the fact that the
pollution might originate from several sources. It gives the authorities a good basis for the future work
in protection of groundwater, especially for controlling the contamination in the mineral springs and
their water-bearing aquifer. The groundwater management plan will become binding through the
involvement of the responsible boards and agencies.
The groundwater management plan merge results of integral groundwater and contaminated site
investigations including necessary measures for securing a good groundwater quality in Stuttgart, with
the goal to define an integral remediation plan for the downtown area, focusing on the key sources of
CHC relevant for the mineral springs [Kirchholtes & Ufrecht 2015]. The management plan (Fig. 1) is
generally consisting of:
Figure1.Strategyandelementsofthegroundwatermanagemntplan.
Conventional investigation methods, such as drillings, drill core analysis, organic carbon in rock,
Packertests, water level measurements, water sampling, chemical analysis and integral pumping tests
were performed. Due to the complex hydrogeological conditions, new and advanced investigation
strategies and methods were applied, which in combination allowed for multiple lines of proof. These
innovative methods applied were: isotopic analysis, micro element analysis, microbiological analysis
and tracer tests (with fluorescence, frigene, sulfur hexafluoride).
The hydrogeological conditions and chemical mechanisms derived from this approach, as well as from
extensive already existing data sets, were combined in a conceptual model. The conceptual model
consists of two modules, i.e. (i) the conceptual aquifer model describing the hydrogeological system
and (ii) the conceptual contaminant model describing the chemical processes in the system.
Furthermore, important elements of the conceptual contaminant model are so-called profiles with
substance specific information on 19 sites, as presented in Fig. 2 (red dots). These sites have been
identified to have a significant influence on the CHC contamination of the deeper aquifers.
Based on the conceptual hydrogeological model, an unsteady numerical model with in total 17 layers
was built to simulate the CHC-species transport in groundwater, accounting for degradation and
transformation processes. The calibration of flow and transport was performed for the period from
1960 to 2010 in an iterative process by varying the relevant parameters. Finally after calibration, the
numerical model was validated with data from environmental tracers (sulfur hexafluoride) and isotopes
(tritium).
The aim of the project was to identify the key sources, which have influence on the deeper layers and
thus the mineral water aquifer. By eliminating the sites with no significant influence on deeper
groundwater contamination, nine contaminated sites with major impact on deeper layers have been
identified for further investigations (contaminated sites no. 1, 2, 3, 6, 7, 8, 9, 10 and 19 in Fig. 2). The
initial situation of these sites has been comprehensively investigated and described by conceptual and
numerical model.
In order to understand and discuss the results of conceptual and numerical model, a geo-referenced
3D-visualisation tool (MAG-IS) has been developed [Lang et al. 2014]. The CHC concentrations for
each layer and each species, hydrographs for comparison with measured values, cross-sections,
parameters, flow velocities and geological structures can be visualized by applying this usefull tool.
With MAG-IS it is possible to understand the relation between groundwater flow, contamination input,
transport, sorption (or retention), degradation, transformation processes as well as output of
contamination. MAG-IS has proved itself to be an efficient tool for interdisciplinary team work and
communication between partners involved. It supports the iterative-adaptive approach by its on-line
availability to each partner and easy-to-use structure.
Figure2.Investigatedareaandcontaminatedsites.
With an assumption that the emission of contaminants is constant in the period from 2010 to 2060,
different scenarios were simulated for each of nine key contaminated sites:
Scenario a) no remediation: all ongoing remediations at contaminated sites were turned-off. The
contaminated sites will emit constantly the maximal concentration load.
Scenario b) emission 2010: all ongoing remediations were unmodified continued. Only the
contamination, which has not been included in remediations will be emitted.
Scenario c) 50 % of emission 2010: the ongoing remediations were optimized, so that 50 % of
emission will be emitted, corresponding to b).
Scenario d) no emission: emission of contamination was completely reduced (100 % of
remediation, which is considered as no realistic scenario).
All scenarios-data are summarized in Table 1. The numerical simulation was performed for the
relevant contaminated sites.
Table1.Emittedloadsofcontaminatsfortherelevantcontaminatedsitesfordifferentscenarios
SCENARIO A) NO SCENARIO B) SCENARIO C) 50 % SCENARIO D) NO
NO. CONTAMINATED SITE REMEDIATION EMISSION 2010 EMISSION 2010 EMISSION
Furthermore, for the relevant contaminated sites the following scenarios were calculated:
Scenario prediction run ALL: all contaminated sites were active and emitted contamination
Scenarios prediction run 1, 2, 3, 6, 7, 8, 9,10 and 19: each relevant site was separately
investigated by switching off all other contaminated sites. Thus, each site specific emission could
be considered with its influence on a receptor
As an example, the results for the contaminated site no. 9 (Rmelinstrae 24-30) are presented in
Figures 3 to 5. The site was used for storage of chemicals and mineral oil products from 1912 to 1992.
The CHCs were handeld from 1939 (TCE) and from 1950 (PCE) onwards. At the site there are two
main sources of contamination, where CHCs were deeply inflitrated and thus affected the
groundwater. The geology of the site itself is characterised by many geological faults, which influence
and complicate the hydrogeological conditions. The contaminated site location and its calculated TCE-
plume toward the mineral springs are shown in Fig. 3.
In Fig. 4 the development of the CHC concentrations are presented for the period form 2010 to 2060
in mineral spring Berger Nordquelle (receptor) for the scenarios a) to d), with the scenario prediction
run ALL (all contaminated sites emit pollution). It was known that the mineral spring Berger
Nordquelle was contaminated with a TCE-concentration of 1.7 g/l in 2010. All other species could be
neglected due to their concentration less then 0.1 g/l. Analog in Fig. 5 the development of CHC-
concentrations for the contaminated site no. 9 Rmelinstrae 34-30 (the scenario prediction run 9)
are presented.
Figure3.Contaminatedsiteno.9(Rmelinstrae2430)anditsTCEplume.
Comparison of Figures 4 and 5 indicates that both remediation scenarios lead to similar CHC
concentrations in the investigated receptor. This underlines the assumption that TCE-contamination in
the mineral springs originates mainly from the contaminated site no. 9 Rmelinstrae 24-30. Fig. 5d)
shows that even if the remediation of all other contaminated sites continues as in 2010 (scenario b)),
the complete reduction of the emission by the contaminated site no. 9 would result in the required
remediation of the mineral springs.
Since the mineral springs have a great significance in Stuttgart, a special focus was on securing a
good quality of the carstic groundwater aquifer (upper Muschelkalk) [LUBW 2012]. Furthermore, the
prioritization was influenced by requirements of mineral springs protection law. The aim of prioritization
was to reach the target values in foreseeable time for CHC concentration in three receptors:
mineral springs
upper Muschelkalk aquifer
groundwater above Muschelkalk
All three receptors are subjected to groundwater law. This has resulted in the definition of priorities,
considering that not all necessary measures can be launched at the same time. The groundwater
management plan provides time sequence of the necessary activities and remediation target values
[A.A. 2001a, A.A. 2001b, A.A. 1984, Kss & Kss 2008, Gilles et al 2011]. These values are
presented in Table 2.
Figure4.CHCconcentrationsinmineralspringBergerNordquelleforthescenariosa)tod)basedon
emissionbyallcontaminatedsites(predictionrunALL).
Figure5.CHCconcentraionsinmineralspringBergerNordquelleforthescenariosa)tod)basedon
emissiononlybythecontaminatedsiteno.9Rmelinstrae2430(predictionrun9).
Table2.Determinationofprioritiesandtargetvalues
PRIORITY RECEPTOR TARGET VALUE
1 mineral springs contaminated site influences mineral springs
remediation target is complete prevention of
contamination input in Muschelkalk aquifer
2 groundwater in Muschelkalk aquifer concentration in receptor
CHC < 5 g/l
CHC-species < 1 g/l
3 groundwater above Muschelkalk aquifer emission and immission of a contaminated site
CHC < 20 g/d
CHC < 10 g/l
Based on scenario simulations it was possible to identify and characterise contaminated sites with
high priority and determine their influence on a receptor. It was shown that remediation efforts, despite
generally good local effectivness (at site), are not sufficient for reaching the requested decrease of
contamination in mineral springs. Therefore, at particular contaminated sites additional remediation
measures are necessary to be considered:
contaminated site no. 9 (Rmelinstrae 24-30): This site has the highest priority 1. The current
remediation has no impact on the contaminated plume toward the mineral springs. Therefore, an
additional remediation is necessary, which shall effectively prevent the CHC input at the
contaminated site in the Muschelkalk aquifer.
contaminated site no. 1 (Dornhaldenstrae 5): This site has the priority 2. This contaminated site
has influence on the Muschelkalk well Tbinger Strae, which is used as process water of a
brewery . Also, here is necessary to improve the remediation and to extend it up to the deeper
aquifer of Muschelkalk. This estimation can also be applied for the contaminated site no. 2
(Rotebhlstrae 171).
contaminated site no. 7 (Nesenbachstrae 48): This site has the priority 2. The remediation has to
be optimized to capture the deeper aquifer of Muschelkalk.
In this way the target values for each contaminated site were setup, as presented in Table 3.
Table3.Remediationrequirementfortherelevantcontaminatedsites
SCENARIO A) NO SCENARIO B)
NO. CONTAMINATED SITE REMEDIATION EMISSION 2010 EVALUATION
PCE- EQUIVALENT [g/d]
1 Dornhaldenstrae 5 50 16 priority 2: targeted decrease input in mo
2 Rotebhlstrae 171 160 30 priority 2: targeted decrease input in mo
3 Johannesstrae 60 53 23 priority 3: site specific evaluation
6 Rotebhlplatz 19 16 16 priority 3: site specific evaluation
7 Nesenbachstrae 48 412 5 priority 2: targeted decrease input in mo
8 Wolframstrae 36 15 2.4 priority 3: site specific evaluation
9 Rmelinstrae 24-30 130 56 priority 1: targeted decrease input in mo
in 5 years will be priority 3: site specific
10 Mittnachtstrae 21-25 111 13
evaluation
19 Prag-/Lwentorstrae 129 37 to be further investigated
mo: Muschelkalk
The remediation feasibility study describes the measures, that have to be considered in order to reach
the remediation targets at contaminated sites with the priority 1 and 2. It is developed for those
contaminated sites, for which the scenario modelling indicated the neccessity for furhter investigation
and remediation i.e.:
the contaminated site no. 9 with priority 1
the contaminated sites no. 1, 2, and 7 with prioriy 2.
It is important to emphasise that the remediation feasibility study complements (not replaces) the
single site investigation and remediation. Furthermore, the feasibility study has determined, which
technical options for each contaminated site are possible. Various technically applicable options were
evaluated and compared regarding 23 non-financial (technical, organisational and influence on the
environment) and the financial criterias, with the cost estimation for the alternative remediation
methods (LfU 1994). Finally, the optimal solution for each contaminated site was recommended. The
results of the evaluation matrix are presented in Table 4.
Table4.Summaryoftheremediationfeasibilitystudyforcontaminatedsiteswithpriority1and2
9 1 2 7
NO. / DESCRIPTION
RMELINSTRAE 24-30 DORNHALDENSTRAE 5 ROTEBHLSTRAE 171 NESENBACHSTRAE 48
PRIORITY 1 2 2 2
CONTAMINATION
2.4 0.9 2.9 7.5
CAPACITY [T]
EMISSION OF CHC-
130 50 160 412
EQUIVALENT [g/d]
A suitability and acceptance of the proposed remediation methods have to be tested by a pilot test at
each of four contaminated sites.
In order to control and verify the effectiveness and success of the remediation feasibility study, it is
necessary to establish a monitoring programme. The numerical model and the scenario analysis
enable the determiation of the contamination spreading at chosen measuring wells. If the calculated
contamination spreading for the given remediation scenario is observed by measurements, it can be
concluded that the system is well understood at this particular area and that the applied remediation
measures are appropriate. If not, the optimization of the remediation has to be performed and further
measures for the improvement of the understanding the system have to be defined.
The following three criterias were used in order to define the monitoring network:
the emissions of contamination, which have influence on the deeper layer, i.e. upper Muschelkalk
aquifer are prior to be monitored
each aquifer with emission higher than allowed should be controlled by at least one measuring well
new measuring wells should be drilled only in case that the existing measuring wells are not
suitable or not existing
In such a way a long-term monitoring network, comprising 27 monitoring wells of four aquifers, was
proposed. Fig. 3 indicate an exapmple of the monitoring network for the contaminated site no. 9
Rmelinstrae 24-30 with its emitted contaminated plume calculated by the corresponding numerical
model.
The success of the groundwater management plan can be measured by the improvement of the
groundwater quality in Stuttgart. It is therefore necessary to ensure the implementation of the identifed
investigation and remediation measures as well as the corresponding monitoring programme. In order
to secure the implementation, the following requirements have to be fulfilled:
agreement on the investigation and remediation measures with the responsible water authority
establishment of the action and financing plan for the next 5 years
clarification of the responsibilities by realisation of the necessary measures
application at possible sponsorships and securing the funding
the action and financing plan as well as the approach have to be summarised in a report and
submitted to the city council.
Public relation is considered to be an important part of the groundwater management plan. It attends
to inform the affected public in an early stage about the contamination situation, the needed measures
for protection of mineral springs and groundwater, status and results of the applied measures. Since
the problem of mineral springs and groundwater contamination is a sensible topic in Stuttgart, it is
important to design the communication strategy in the way to reach the relevant audience (community
members, joung people, experts, scientists, media, industry etc.). In this way the aceptance for the
needed measures can be achived, awareness and interest for the protection of groundwater can be
raised. Furthermore, information about the practical results, methods applied and allocation of tools for
experts can be provided to the relevant audience.
Beside the project homepage, various media were applied such as: newsletters, vitreous aquifer,
online-microsite quiz for teens, films, laymans report etc.
3. Brief recap
The new tool groundwater management plan summarizes the results of (i) an integral investigation of
the groundwater quality and contaminated sites and (ii) the measures required to ensure a good
groundwater status. It provides a good basis for an optimal treatment of CHC contamination,
consisting of numerous reports, maps, data bases, GIS-based visualization tool, implementation and
communication strategies.
A remediation feasibility study, as a part of the groundwater management plan, describes the
measures to be taken in order to achieve the remediation targets. It is designed for those sites, where
numerical model predictions have indicated the need for additional investigation and remediation
efforts. In order to verify the effectiveness, impact and success of the remediation concept,
remediation measures have to be controlled in a monitoring program. The groundwater management
plan will become binding through the involvement of the responsible boards and agencies.
2
For the large-scale investigated area of 26,6 km and eight aquifers, extended data and information
were available. At the begining of the project, 182 possible contaminated sites with a significant CHC
concentration were identified. The integral investigation had indicated that 21 sites had an impact on
the deeper layers. Therefore, a detailed investigation of those sites was performed and the results
were summarized in 19 so-called profiles. Furthermore, the numerical model indicated that 9
contaminated sites affected the upper Muschelkalk aquifer, i.e. mineral water-bearing aquifer. Finally,
the priorization characterised 4 contaminated sites with high priority and need for optimized or
additional remediation efforts. For those four contaminated sites feasibility studies were performed and
resulted in recommendation of remediation measures with total costs of more than 12.5 million euros.
The proposed strategy and approach are suitable for multiple sources of contamination. Only in this
way, the contributions of single contaminated sites to the total groundwater contamination can be
identifed and local remediation measures with their spatial impact simulated. Due to very complex
geological conditions, technically there is no alternative to this strategy in order to achieve the CHC
purity in the mineral springs. Stuttgart is appoaching very closely to this goal.
4. Literature
A.A. (1984) Verordnung ber natrliches Mineralwaser, Quellwasser und Tafelwasser (Mineral- und Tafelwasser-
Verordnung) vom 01.08.1994, Bundesgesetzblatt I
A.A. (2001a) Allgemeine Verwaltungsvorschrift ber die Anerkennung und Nutzungsgenehmigung von
natrlichem Mineralwasser vom 9. Mrz 2001, Bundesanzeiger Nr. 56
A.A. (2001b) Verordnung ber die Qualitt von Wasser fr den menschlichen Gebrauch (Trinkwasserverordnung
TrinkkwV 2001) vom 21.05.2001 in der Fassung der Bekanntmachung vom 2. August 2013, Bundesgesetzblatt
Gilles, C., Hartmann, B., Jendritzky, G., Kleinschmidt, J., Stoyke, B. & Vogt, I. (2011) Kommentierte Fassung der
Begriffsbestimmungen- Qualittsstandarrds fr die Prdikatsierung von Kurorten, Erholungsorten und
Heilbrunnen. Deutscher Heilbderverband e.V., 12. Aufl.
Kirchholtes, H.J. & Ufrecht, W., eds. (2015) Chlorierte Kohlenwasserstoffe im Grundwasser,
Untersuchungsmethoden, Modelle und ein Managementplan fr Stuttgart, Springer Vieweg
Landeshauptstadt Stuttgart (2003) Kommunaler Umweltbericht: Das Grundwasser in Stuttgart. Schriftenreihe des
Amtes fr Umweltschutz, 1/2003, Stuttgart
Lang, U., Paul, T. & Mirbach, S. (2014) MAGPlan- MAG-IS Benutzerhandbuch Bericht Ingenieurgesellschaft
Prof. Kobus und Partner im Projekt MAGPlan, Stuttgart
LUBW (2014) Integrales Altlastenmanagement. Leitfaden und Handlungshilfe zur integralen Untersuchung und
Sanierung von Altlasten. Landesanstalt fr Umwelt, Messungen und Naturschutz Baden-Wrttemberg, Karlsruhe
LfU (1994) Eingehende Erkundung fr Sanierungsmanahmen / Sanierungsvorplanung. Texte und Berichte zur
Altlastensanierung, Landesanstalt fr Umweltschutz Baden-Wrttemberg, 10/94, Karlsruhe
Ufrecht, W. (1999) Geologie und Hydrogelogie des Nechakrals zwischen Stuttgart-Mnster und Stuttgart-
Untertrkheim. Schriftenreihe des Amtes fr Umweltschutz, 4/1999, Stuttgart
Acknowledgements
This study was performed within MAGPlan Project (management plan to prevent threats from point
sources on the good chemical status of groundwater in urban areas, http://www.magplan-life.eu). The
project is financially supported by the program LIFE+2008 environment of European Union from
01/01/2010 to 30/09/2015. Responsible project partners are the City of Stuttgart, Department for
Environmental Protection as well as the Baden-Wrttemburg State Institute for Environment,
Measurements and Nature Conservation (LUBW).
HOW TO GET A CAMEL TO GO THROUGH THE EYE OF A NEEDLE:
SUCCESSFUL SITE REMEDIATION OF A FORMER EXPLOSIVES PRODUCTION SITE:
SAFE HOUSING, WORKING AND DRINKING WATER PRODUCTION ON A LONG-TERM BASIS
Session:
ThS 1D.3 Risk mitigation and intervention measures
Authors:
Christian Weingran
Hans Georg Meiners
Institutions:
HIM GmbH ASG, Stadtallendorf
www.him.de, +49 6428 9235-11, asg.weingran@t-online.de
ahu AG Wasser Boden Geomatik, Aachen
www.ahu.de, +49 241 900011-0, g.meiners@ahu.de
Keywords:
Explosive-specific compounds, contaminated munitions sites, project management, citizen participa-
tion, remediation agreements, drinking water
1. The challenge
Anyone visiting Stadtallendorf in the State of Hesse will find a more or less typical German small town
with some 22,000 inhabitants, its commercial, industrial and residential areas. It is home to a few large
business enterprises, many small businesses and, in between all this, houses where people live. For
visitors it might seem strange that many of the houses have enormously thick concrete flat roofs,
many of which are covered with trees and greenery (Fig. 1).
Instead of this booming industrial estate there could actually be flourishing forest. After the end of
World War II it was uncertain for a long time what was going to happen to this area. The problem was /
is that during the war this was the largest munitions production site in Europe. Here, under strict se-
crecy, the two armaments manufacturers, WASAG and DAG produced explosives to arm bombs and
grenades on site. The two ammunition factories covered an area of some 1,000 ha and, as camou-
flage against air reconnaissance, were located in the midst of the forest. More than 15,000 forced
laborers had to work here in the arms factories.
During the Second World War the site was untouched by Allied air raids. It wasnt until after the war
that the Americans dismantled and blew up parts of the armament plants. And this resulted in one of
the worst cases of environmental damage in Germany: soil, streams and groundwater in a wide sur-
rounding area were all contaminated with highly toxic substances from the explosives. At that point
there was no talk of investigations, not to mention remediation. In the post-war period buildings which
were half-way intact were put to use despite the contamination and new buildings constructed on
site. It wasnt until the 1980s and a change in our environmental awareness made it clear just how
great the extent of contamination was and its risk to human health and environment. However, since
neither the technology nor the finances for a remediation were in place, the retrogression to a forestal
area closed to the public was in real terms the option for a long time.
But things turned out differently: among the local people and the town itself there was a total lack of
understanding and little support for the option Forest. They had been living and working here too
long and simply could not imagine such a hard cut. The realization of this option would have meant the
relocation of all the people living there and all the business enterprises plus the abandonment of the
waterworks. Since this option was neither enforceable nor politically desired, a remediation process
was started, although at this point no one really had an idea of how this remediation could work. What
about the contaminants from the explosives in the soil and groundwater? Would it be possible to re-
move the contaminated soils without evacuating parts of the area or indeed the entire area? Would
the excavation not mean an enormous increase in the groundwater pollution more than ever before?
And who was going to assume responsibility for such work whose consequences cannot be foreseen
because no one has any experience in this field. These were just a few of the very many questions
and issues that concerned and moved the people at that time. For most people remediating the site
seemed as impossible as getting the proverbial camel through the eye of a needle.
It was clear from the beginning that, if a remediation were to be taken into consideration, an accurate
picture of the actual pollution had to be made. This meant answering the following questions: Where is
there soil pollution, what kind of pollution and how deep does this go? How far had the substances
already been transported by the groundwater, and again, how deep? It was also necessary to answer
the question as to the risks for health and environment connected to the pollution. There was only very
little scientific background available on such work. At that time there was a fierce debate on the extent
of the danger involved.
The first orienting investigations and an historical examination were carried out from 1988 to 1991. In
the subsequent years until 1997 a complete risk assessment with thousands of drilling and probing
activities were carried out.
2,4,6-trinitrotoluene (TNT)
4-amino-2,6-and 2-amino-4,6-dinitrotoluene (A-DNT)
2,4-and 2,6-dinitrotoluene (DNT)
1,3,5-trinitrobenzene (TNB)
2-,3,and 4-mononitrotoluene (MNT)
2-amino-4-nitrotoluoene (A-NT), 2-amino-6-nitrotoluoene (A-NT)
In addition contamination with RDX and hexyl as well as with contaminants, which are not explosive-
specific, in particular PAH had been detected.
The contaminants were found in the soil, soil vapour, room air, building structure, organic materials
(e.g. leaves, fruit, fungi, fish), sewage system and groundwater.
The distribution of contaminants was heterogeneous, contamination was both scattered and concen-
trated in hot spots. Sporadically crystalline pure TNT was found (Fig. 2). The contamination in some
areas was so severe that a risk to human health could not be excluded. People could come into con-
tact with hazardous substances in particular via the soil (consumption of home grown food). The data
from groundwater monitoring showed that the aquifer was contaminated.
A site-specific and target-specific risk assessment had been elaborated. Soil action values could be
deduced considering toxicity data, tolerable body doses, relevant exposure scenarios and the rele-
vance and importance of the individual exposure pathways. Evaluation of single substances would
have led to an underestimation of the total risk. The evaluation was done by a weighted summation of
individual substances, meaning that the resulting tolerable soil values are given in mg TNT-
equivalents (TNT-TE). Toxicity coefficients were developed for relevant contaminants and for short,
medium and long term exposure to evaluate the toxicity.
On this basis intervention values for the sum of 10 aromatic nitro-compounds as well as for four indi-
vidual substances were determined, differentiating use and exposure time.
At the beginning of the remediation process the many stakeholders (e.g. The State of Hesse, munici-
pality, water suppliers, property owners, residents) all had very different interests and objectives.
However, in a long discussion process a consensus of shared objectives could be reached:
The overall objective was the successful, i.e. the economically and ecologically effective and so-
cially compatible realisation of a use-specific remediation.
The remediation should eliminate or significantly reduce the impairments caused by soil and
groundwater pollution to the land use and to the general public.
The location should be kept as a residential and business location and interventions in green areas
should be minimised.
Drinking water production from the Stadtallendorf waterworks should be kept on a long-term basis.
The soil decontamination should not pose a risk to drinking water production.
At the outset of the work it was critical that among the many parties involved the objectives and the
appropriate measures could be defined. The discussion was carried out using an overall concept of
remediation which, bit-by-bit in the many discussions, developed into an overall concept.
The various people involved were asked the same question: From your point of view, what has to
happen to ensure that the site has a secure future? Here the similarities and differences among the
various ideas were weighed up and systematically dealt with. The overall concept of remediation
evolving from this summarized all elements of the remediation strategy in a poster. This clear visuali-
zation turned out to be extremely helpful in the discussions with inter-disciplinary experts (e.g. engi-
neers, chemists, geologists, urban planners, politicians) as well as in discussions with the general
public. The contexts and interrelationships of the individual remediation elements were comprehensi-
ble and the people involved could see a common goal. This clearly understandable overall concept of
remediation, which was shown at every meeting over the years, was at the end of the day one of the
reasons for the high level of acceptance of the planned measures.
Going by the overall objectives, the following issues were included in the overall concept of remedia-
tion and detailed measures carried out.
Soil remediation
In open spaces and gardens soil excavation was carried out up to a depth of 3 metres sometimes
even deeper. The crucial factor here was that the use-specific intervention levels were exceeded (soil
excavation up to 1 m) and a soil remediation on a groundwater basis (up to 3 m and deeper) (Fig. 4).
The entire excavation volume amounted to some 220,000 tonnes and the removal of explosive-typical
substances approx. 125,000 kg. The proceedings were carried out under the highest precaution
measures (under the aspect of work safety and protection of the local residents). For instance, the
excavation work was mostly carried out under the protection of tent constructions. The excavated soil
was transferred to an interim storage and thermally treated in an incineration plant. After the soil re-
mediation there is still a diffuse residual contamination in the soil and aquifers. The remediation of
these would be almost impossible or only possible with unreasonable expenditure and effort. The re-
sidual pollutant potential is estimated at approx. 30 - 40 t of explosive specific compounds.
Sewer renovation
From 1996 to 2004 a sewer system covering some 70 km was investigated. Contaminated sewer
lines were cleaned with high pressure, closed down, backfilled or in some places removed. All in all, a
total of 3 t TNT was removed from the pipes or from the contaminated flushing water.
Groundwater protection
The waterworks of the explosives production continued to be used after the end of the war and has
provided the public within a wide radius with drinking water (up to 13 million m/a).
In the development of the overall concept of remediation maintaining the water extraction and supply
was always a basic requirement. In the beginning there was great concern that the soil remediation
would increase the pollution of groundwater and pose an additional hazard for the drinking water sup-
ply.
To protect the drinking water supply a hydraulic safeguard was planned and went into operation in
1995. In the immediate proximity of the contaminated areas this comprises nine wells, which prevent
the inflow of contaminated groundwater into the extraction wells for the drinking water supply and min-
imises the contaminated groundwater downstream from the former explosives production site. It was
only with this protection that the soil remediation could be carried out. The water collected by the hy-
draulic safeguard is purified using wet activated carbon in a special water treatment facility. This hy-
draulic safeguard has been in operation for 20 years now without any failures or the permissible val-
ues being exceeded. The volume of groundwater extracted from 1995 to 2013 amounted to 8.5 mil-
lion m; the load of explosive-specific compounds removed in the same period was approx. 1,700 kg.
The hydraulic measures have to date always been accompanied by extensive groundwater monitoring
programmes. The some 160 remediation wells, drinking-water wells and groundwater monitoring wells
are analysed on a regular basis. This includes examining whether there are any changes in the quality
of the groundwater or deviations from the prognosis, whether there is a spread of pollutants beyond
the protected area and whether the threshold values for drinking-water wells are observed.
After ten years remediation work the remediation was successfully completed in 2010. Fig. 6 shows
the remediation result of the individual measures. In a cost/benefit balance these items are on the cost
side. The expenditure can be offset by the benefit: as far as ground and water are concerned, basic
needs for life have been secured for the people living and working there and, to a large extent, a pro-
tected environment and a tenable industrial location important for the region and securing more than
8,500 jobs.
Soil excavation volume 200,000 t
Explosives 24,000 kg
Area 420 ha
5. Factors of success
In retrospect, the following factors were decisive in the successful completion of the remediation and
in getting the camel through the eye of the needle:
(1) The common will for the remediation and the will of the State of Hesse to bear the remedia-
tion costs
These were the basic prerequisites for starting the project and being able to pursue it from more than
25 years. One thing that was particularly noteworthy was the political support received, which despite
changing political majorities and directions, was always in place. Once the remediation had started it
was never seriously questioned or disputed in all that time, even if major individual measures such as
the remediation of TRI-Halde with more than 50 million remediation and disposal costs truly were
milestones. Remediation agreements made between property owners and the State of Hesse ex-
empted the property owners from any costs for the remediation of explosive-specific compounds with-
out a time limit. This was the basis for a remediation without any legal disputes.
(6) Courage, patience and understanding on the part of the many people involved at all levels,
i.e. politicians, local administration, residents and employees
Without this, such a project would have been inconceivable. When money is tight, publicly putting up a
case for such a costly remediation measure and tolerating soil excavation work with heavy equipment
in your own backyard shows just how convinced the people involved were of the benefits of the pro-
ject.
With the soil remediation completed in 2005 the idea was that the project had been wound up. On
both the administrative and operative side funds personnel could indeed be reduced, and the public
interest the project had enjoyed over the years diminished.
At non-accessible levels or where contaminants have not been removed for functional and strategic
reasons, it is still necessary to continue with the hydraulic safeguard and controlled handling and
treatment of soil during construction projects. The handling of these residues also calls for a careful-
ly planned approach.
For this reason the State of Hesse will have to allocate personnel and financial resources for the fol-
low-up for several decades. And this will remain a topic on the agenda for the municipality and the
property owners.
The objective of the follow-up work is that the residual contaminants after the remediation pose no
danger to human safety and health or risk to the environment in connection with the existing and
planned use of the land. Fig. 7 shows the fields of the follow-up activities.
Fig. 7: Follow-up model for Stadtallendorf
7. Concluding remark
For some time now there have been no considerations of abandoning the location and any fears that a
remediation of this location would be impossible have been clearly refuted. To the contrary, the work
carried out in Stadtallendorf have played a substantial part in showing that nowadays there are no
insurmountable technical challenges in the remediation of inhabited former armaments sites neither
in the exploration nor in the evaluation or remediation. All instruments have been tried and tested and
are available. Remediation is possible! The camel got through the eye of the needle.
Bibliography
Hessian Ministry of the Environment, Rural Affairs and Consumer Protection (Edit.) (2006): Boden gut
gemacht. Die Sanierung des Rstungsaltstandortes Stadtallendorf.
Hessian Ministry of the Environment, Rural Affairs and Consumer Protection (Edit.) (2009): Prognose
und Kontrolle des natrlichen Rckhalts und Abbaus von Nitroaromaten im Festgestein am Rstungs-
altstandort Stadtallendorf, Abschlussbericht F+E-Vorhaben FKZ 0330508.
Session 1D.2 - Large scale inventories and strategies for dealing with contamination,
Arne Alphenaar, TTE Consultants, alphenaar@engineers.nl, +31 570-665878, Keizerstaat 16 7411 HH Deventer,
the Netherlands
Reinder Slager, 3 Dimensies, reinder@3dimensies.nl, +31 6-24235462, Schellerpad 26 8017 AM Zwolle, the
Netherlands
Frank Swartjes, National Institute for Public Health and the Environment, frank.swartjes@rivm.nl, + 31 30-
2743356, PO box 1, 3720 BA Bilthoven, the Netherlands
Piet Otte, National Institute for Public Health and the Environment, piet.otte@rivm.nl, +31 30-2743965, PO box 1,
3720 BA Bilthoven, the Netherlands
Introduction
Problem framing
In most countries, the approach for the management of contaminated groundwater is particularly
designed for the one source-one plume case in which each case is assessed and managed
separately. In most countries in the world, however, the groundwater quality in urban areas is often
affected by (former) industrial and commercial activities. This has resulted in a series of often
intermingling groundwater plumes in urban areas. Remediation of each individual groundwater plume
would be impractical and inefficient. Moreover, since these intermingling groundwater plumes have
moved over long distances, it proved to be difficult to link individual groundwater plumes to specific
aboveground activities and responsible parties. Consequently, remediation of individual groundwater
plumes often appeared to be a fruitless activity, despite of the many available groundwater
remediation technologies, due to legal, technical, and financial difficulties.
Area-wide perspective
Therefore, the so-called area-wide groundwater management approach (AwGm, in Dutch abbreviated
as GGB, Gebiedsgericht GrondwaterBeheer) was developed as an alternative to tackle complex
groundwater contamination situations. According to the Dutch Soil Protection Act, area-wide
approaches are permitted in situations where overlapping plumes hindered an effective approach to
individual groundwater plumes. This guidance relates to the delineation of the domain, the
management of sources for groundwater contamination, procedures for monitoring, (risk-based)
assessment of the groundwater quality.
1
AGGb is a combination of Witteveen+Bos, Bioclear, 3Dimensies and TTE Consultants
problems. To overcome limitations of the conventional remediation approach, the pioneers need to
force a breakthrough of a 'paradigm shift'. The challenge now is to translate the success of the
innovators into support to the (early) adapters.
In our study, we examine the success factors for early adopters. The main question is whether area-
wide groundwater management can help to achieve their objective(s), and if so, how to do this in an
efficient way.
A way of thinking
Although the approach is not the Columbus egg, but AwGm is anno 2015 widely accepted in the
Netherlands as an effective strategy. Most AwGm projects are related to a cost effective approach of
complex groundwater contamination. However the main motivation often is stimulation of spatial
development or business expansion. AwGm is encouraged as a paradigm to handle a wide range of
complex groundwater problems, sometimes even seen as a goal by itself. However in essence, its
neither tool nor goal. It's a way of thinking.
AwGm includes delineation of a specific area, stakeholder involvement, risk assessment and
monitoring. It has evolved into the Swiss army knife of the groundwater remediation sector. A general
purpose tool to fix a wide range of groundwater interests and groundwater related problems. In
essence, however, it is rather a way of thinking than a practical tool. The risk based area-wide
approach is not primarily focused to contaminant removal, but on a cost-effective removal of
bottlenecks caused by groundwater contamination.
Policy implementation
The Ministry of Infrastructure and the Environment defined the AwGm as: "a policy framework in which
the integral and sustainable design of groundwater management is performed within a restricted area.
Such that groundwater activities and interventions in the groundwater linked to environmental, nature
(recovery) and the spatial and economic developments are implemented and integrated system
geared for the long term"
Since January 2012, AwGm of soil and groundwater contamination is implemented in the Soil
Protection Act. Placing the (approach) of groundwater contamination in a wider perspective means a
significant move in the Dutch approach on groundwater management. As a consequence, the focus
changes from remediation towards risk reduction and areal function.
At the same time, separating source and plume allows a more efficient response to dynamics of
redevelopment. Permanent groundwater management is picked up by the government, source
removal is the responsibility of the site owners.
The ambition
Groundwater-related problems often diminish by placing them literally in a broader context. This,
however, also increases the complexity of the process. The number of parties involved (public and
private actors) and their sometimes opposite interests increase.
Striving for joint value and the willingness to cooperate are prerequisites for the success of AwGm.
Perfection can be the enemy of the good. The challenge is to find natural synergy and coherence by
combining task, goals and objectives of all stakeholders.
The power of AwGm is to reduce costs and increase added value: from an affordable solution for a
groundwater complex problem to the stimulation of spatial development. Since a (limited) spreading of
contamination in groundwater is inherent to the concept, sceptics consider AwGm as a justification to
uncontrolled spreading.
Decentralization
The introduction of AwGm in the Netherlands runs parallel with shifting the responsibilities from
national to local governments. In the framework of area-wide groundwater management, this
development has huge advantages: local authorities recognize the spatial interests and are willing and
able to make an integrated consideration of all spatial interests.
Decentralization also has some pitfalls. Smaller organizations have knowledge gaps to handle
complex considerations. This involves not only technical knowledge (underground structure,
hydrogeology, risk assessment, etc.), but for example also knowledge and experiences with effective
negotiation.
The task
Different stakeholders have different tasks and interests. It is important to guarantee that the approach
has advantages for all participating parties. When governmental parties (area managers) take over the
responsibility for the groundwater plume, private parties are willing to invest in source remediation.
Good cooperation is a key success factor. In addition to a transparent process and shared objectives,
personal skills are important. The recognition and acknowledgment of roles, responsibilities and
interests is needed to understand each other and to overcome conflicting interests. It is important that
all interlocutors have sufficient mandate within their own organization.
Conclusion
Since the first initiatives in the early 2000's, area-wide groundwater management became a valuable
alternative for addressing complex groundwater contamination compared to a plume-oriented
approach. AwGm got a statutory basis in the new soil covenant and a developed a certain level of
maturity. Idealism and positivism make way for healthy pragmatism.
Due to the complexity and uniqueness of each groundwater pollution, there is neither an optimal way
to implement AwGm, nor a good possibility to standardize the approach. AwGm is not a formula or a
tool, it is a way of thinking. Dont stop with that!
Literature
Een gids voor duurzaam omgaan met de gebiedsgerichte aanpak van verontreinigd grondwater,
2012.Slenders, H., R. Jansen, L. Bakker, C Roovers, T. Verschoor, P. Otte, F. Swartjes. DOG. SKB
rapport
Duurzaam gebruik van de ondergrond: gereedschap voor structuur en visie: TCB, R22(2012), maart
2012.
ThS 2.3 Redevelopment of brownfields 1
Ministry of Infrastructure and Environment Rijkswaterstaat/Soil+, P.O. Box 2232, 3500 GE Utrecht,
+31-6-46372877, michiel.gadella@rws.nl
Abstract
This paper describes the Dutch focus on remediating all urgent sites with unacceptable risks for
human health, ecology or migration of contaminants to the groundwater before the year 2020: the final
countdown. When this countdown is completed an era of 40 years of soil remediation will come to an
end. This countdown clears the way for the Soil Protection Act to be withdrawn and to be incorporated
into the new Environment and Planning Act. The remaining historically contaminated sites will be
addressed at a natural moment of redevelopment of brownfields and urban areas. This paper ends
with lessons learned over a period of 35 years. Ambition, endurance, collaboration of all stakeholders,
focus on urgent sites and political support and sufficient funding are essential for solving a major
historical environmental problem.
Keywords: remediation policies, soil protection act, integrated legislation, remediation program, urgent
sites, governance
1. Introduction
The Netherlands is situated in the delta of the rivers Rhine and Meuse in the North-Western part of
Europe. The Netherlands is a very densely populated country, with 17 million inhabitants. The
population density is 400 inhabitants/km2, whereas, for comparison, the USA has a population density
of 31 inhabitants/km2. So the pressure on land is high, especially in urban areas. The same pressures
on land occur in other countries with densely populated urban areas. The Netherlands is a
prosperous, highly developed country. Administratively, The Netherlands is divided into 12 provinces,
393 municipalities and 24 regional water authorities. All these regional and local authorities are
competent in (different areas of) spatial planning, water management and soil and sediment
management. The Dutch society is of a social-liberal nature with increasing public awareness for
sustainable development and the preservation of natural areas.
Land use, especially in urban areas, is dense (about 90% of the population live in cities). This caused,
in earlier times both local (severe) contaminated sites and diffuse (light) contamination. In the topsoil,
the groundwater and aquatic sediments, a large variety of contaminants have accumulated. The
Netherlands are faced with approximately 60.000 sites that require remediation or some form of
management (e.g. containment, capping). Of these 60.000 sites, approximately 2.000 sites pose
unacceptable risks and require urgent remediation (approximately 400 for human health, 200 for
ecology and 1400 for migration of contamination into groundwater).
The policy on contaminated land has its origin in the early 80s of the 20th century when the first
disgrace in soil contamination became apparent. In the small town of Lekkerkerk a residential area
had been built upon a former industrial dump site and indoor air concentration of Benzenes caused
health problems. This triggered public awareness and political commitment, inventory programs and
soil protective policies.
The policy on contaminated land evaluated from a strict preventative policy and a foresight of total
multifunctional clean-up of all contaminated sites in the 80s towards a more realistic policy which
remained strictly preventative but amended the clean-up ambitions towards functional remediation of
heavily contaminated sites. This resulted in the remediation of sites that had to be remediated for other
reasons than the environment (i.e. the redevelopment for housing on former industrial sites) and
therefore left urgent sites abandoned and not remediated. The soil policy makes a clear distinction
between historical contamination and new contamination. The distinction is juridical marked by the
year 1987 when the Soil Protection Act came into force. Because of the aim of preventing further soil
contamination, polluters have to immediately clean-up any soil contamination which is caused after
1987. No right on pollution! All historically contaminated sites need to be remediated, the urgency
depends on the risks. When the risks are unacceptable the remediation should start as soon as
possible. In other cases it is preferable to remediate at a natural (redevelopment) moment.
The soil policy and remediation operation is fully decentralized to the 12 provinces and 29 largest
municipalities, so in total 41 authorities are competent. The main reason for this decentralized
operation is the fact that the solution for soil contamination is often found in spatial planning at a local
level. Remediation in combination with building activities is more cost efficient and more accepted by
the public.
Funding of the remediation operation is essential. The Dutch government has funded the operation
with on average an annual budget of 100 million euro over the last two decades (it is an extensive
budget but at the same time as few as 6 euros or 3 glasses of beer per citizen!). This budget is shared
among those 41 competent authorities. Furthermore private companies who remediate their
historically contaminated sites get a contribution from the government for their remediation costs at an
average of 35 % and a maximum of 70%. This budget has survived different political debates over the
last years because of a clear foresight of start of the remediation of urgent sites before the year 2020.
A new transition in the soil policy became apparent in the first decade of this century. From a political
perspective there was a need to focus on remediating urgent sites. It is difficult to explain to the
society that so many years after the pollution was caused there still are unacceptable risks at
numerous sites. At the same time on a broader perspective there is a trend to integrate sectorial
legislation on the environment and spatial planning. Environmental legislation consists of dozens of
laws and hundreds of regulations for land use, residential areas, infrastructure, the environment,
nature and water. Each has its own starting points, procedures and requirements. This makes the
execution and enforcement of the legislation too complex. Consequently, it takes too long to get
projects started.
The Dutch government wants to make a legal framework with more flexibility for the competent
authorities and will combine the laws on the environment and planning in a single Environment &
Planning Act (foreseen in 2018). The Act will replace 23 existing laws, including the Water Act, the
Spatial Planning Act and the Soil Protection Act.
Through this new Environment & Planning Act the Government wants to improve links between:
different projects and activities (in the fields of spatial planning, the environment and nature);
sustainable developments such as locations for wind farms;
the various regions.
This Act will result in fewer regulations and will reduce the burden of conducting studies. At the same
time, decisions on projects and activities can be made better and more quickly. Moreover, the Act is
more in line with the new European policy and allows more room for private initiatives.
The aim is that remediation of contaminated sites is fully integrated in the development of brownfields
and residential areas. One of the most important preconditions for transferring the Soil Protection Act
into the integrated Environment and Planning Act is the approach and remediation of urgent sites.
When all urgent sites are remediated sectorial legislation on soil remediation can be withdrawn. After
remediating these urgent sites, soil contamination will still be an issue but will be just one of the
aspects that have to be taken into account when an initiative for a development is taken. A set of
general rules for dealing with contaminated soils in building projects will make a development more
predictable in advance.
This paper describes the approach on remediating all urgent sites in the Netherlands in the last 5
years and the coming 5 years (2009-2015 and 2016-2020): the final countdown. Furthermore this
paper describes the outline of the soil legislation within the Environment and Planning Act. The paper
ends with lessons learned over the past 35 years of dealing with contaminated land.
2. Program 2009-2015i
In 2009 an agreement was signed by the Ministry of Infrastructure and Environment, the provinces,
municipalities and water boards. This agreement marked the start of an unprecedented effort to clean-
up all urgent sites in the Netherlands. This agreement aimed at the remediation of all urgent sites with
unacceptable human risks by the year of 2015. Furthermore the aim was to identify all urgent sites
with unacceptable risks for ecology or migration of contaminants into groundwater also by the year
2015. In terms of finance a budget of 700 million Euro for the period 2010-2015 was shared among
the 41 competent authorities for soil remediation.
The agreement did not only focus on action on the urgent sites, it also had the aim to stimulate
competent authorities to develop and implement policies for the regulation and stimulus of ecosystem
services such as subsurface cold and heat storage in aquifers. This paper only addresses the subject
of remediation of urgent sites.
In 2009 a project started called Focus. This project started with the above mentioned workload. Later
on all sites were selected with an above average chance to be severely contaminated. For example
sites with an historical use as dry cleaner or gasworks were selected and investigated. Other sites,
such as sites for storage of building materials, were not selected for investigation because of the fact
that there is a small chance at detecting unacceptable risks at such a site. For asbestos (for example
in road constructions) specific programs have been drawn up.
Since provinces and larger municipalities are the competent authorities for soil remediation and private
owners have obligations to conduct investigations and remediate their sites, if necessary, close
collaboration of all involved authorities is needed to complete an inventory of this magnitude. A team
of experts of the government and the different competent authorities was formed to stimulate and
support provinces and municipalities in their inventory program and to monitor the progress. This team
is called the Program Management Team.
This resulted in 2011 in a combined list of approximately 400 urgent sites with unacceptable risks for
human health. Unacceptable risks for humans occur when indoor air concentrations of contaminants
exceed limit values or when humans are over-exposed to contaminants as a result of ingestion (hand-
mouth behaviour) and food consumption. Contaminated groundwater used for human consumption is
defined under the risks for migration of the contaminants in groundwater.
Figure 1 depicts the situation of all locations with human risks in the Netherlands in 2011.
Furthermore the inventory also resulted in 2013 in a list of approximately 1600 urgent sites with
unacceptable risks for the ecology (200) or migration of contaminants in groundwater (1400).
Prioritizing the sites with risks for humans the final countdown
The 41 competent authorities for soil remediation have drawn up programs for the remediation of sites
with unacceptable risks for humans. About 50 % of these locations the authorities have to finance
these remediations for themselves because of public ownership or because the owner has no means
of funding the remediation. For the other half of these 400 locations existing private owners/polluters
have to pay for the remediation. For these sites private owners can be forced to remediate their
location within a period of 4 years. Private owners can get a contribution from the government of about
on average 35% (maximum 70%) of the total costs for remediation.
The Program Management Team has been monitoring the progress of these 41 different programs in
order to assess the progress of the overall goal to remediate all urgent sites with unacceptable risks
for humans by the end of 2015. Table 1 depicts the progress over the period 2011-2014.
Table 1 might suggest that progress is insufficient (50-100/year) to address all urgent sites with
unacceptable risks for humans before the end of 2015. More detailed information over the remaining
181 locations indicates that for 90 locations remediation is already in progress and will be ended
before the end of 2015. For just 24 locations remediation before the end of 2015 is uncertain. For
these locations financial or legal issues are the bases of delay until 2016.
Overall the program is a success. Sites with unacceptable risks are prioritized and the countdown is
almost completed for sites with unacceptable risks for humans. This success has to be attributed to
the following key-factors:
Decentralized operation. The local approach towards businesses and landowners is more
successful than a central approach because of the existing network structure. Decentralized
operation also builds capacity.
Funding. The count-down is conducted in a period of financial-crisis. Building activity was low,
lifting along with urban development was very hard, therefore funding of remediation projects
was essential for the progress.
Collaboration, knowledge transfer and monitoring. A decentralized operation breeds
specialists in different organisations. Train the trainer programs have proven to be very
successful. It is also evident that not every program of each individual competent authority is
conducted at the same speed and success. Collaboration in terms of knowledge transfer,
monitoring and benchmarking is needed to keep everybody focussed on the goal of
remediating these locations. The formation of the Program Management Team has proven to
be a strongpoint in the operation.
3. Program 2016-2020iii
The inventory of other urgent sites resulted in a list of 1600 urgent sites with unacceptable risks for
ecology (200) or migration of contaminants to the groundwater (1400). So human risks with top-priority
are dealt with, groundwater contamination still is a top-priority due to the implementation of the EU
Water Framework Directive and Groundwater Directive. Groundwater contamination is more complex
and often too costly. Therefore the approach in remediating the source and managing the plume of the
contaminants on an area based approach is often advised. An area based integrated approach
requires also to incorporate other interest such as heat and cold storage in the aquifer. Groundwater
as a resource for drinking water and (food and drink-)industry is a very important economic factor,
therefore the 2009-2015 program needs to be extended.
The Dutch government, provinces, water boards, municipalities have drawn up a new agreement and
a 5 year program with the aim of dealing with these 1600 locations and establishing new legislation for
the period after this final countdown. These parties involved are aware that the soil and the subsoil
make an increasingly important contribution to the achievement of societal goals (the provision of
energy, the provision of drinking water, groundwater reserves, agriculture, cultural history, nature and
climate mitigation and adaptation). They want to increase this contribution as much as possible in an
innovative manner, taking into account the natural qualities of the soil and water system and the
(targeted) surface and underground functions. The parties involved intend to work towards a further
development to achieve sustainable management and use of the soil and subsoil.
For the period 2016-2020 a budget of 550 million euro is provided by the government and shared
among the 41 competent authorities .
Given these criteria approximately 1400 locations have been selected as urgent for unacceptable risks
for migration of the contaminants in groundwater by the 41 competent authorities.
A large part of these locations will not be remediated on the bases of a site by site approach. Due to
the fact that it are all historical contaminations, a lot of these locations are situated in the vicinity of
each other and plumes of contaminants have mixed in the saturated zone, especially in urban areas.
Therefore a more area based and integrated approach is foreseen for a lot of these locations. An area
based and integrated approach is more complex because more stakeholders are involved. It takes a
lot of effort from local authorities to interest all stakeholders to participate in an area based approach.
The common ground however is that collaboration within a logical area will be more cost-effective. A
lot of knowledge and experience on this point has been built up for example in the CityChlor project.
The aim of the new agreements is that all sites with urgent (unacceptable) risks for migration of
contaminants into the groundwater will be remediated or under control by the end of 2020. Competent
authorities face the challenge to decide whether an integrated area based approach is suitable for
their problem sites. Moreover the competent authorities have to stimulate private owners and
companies to start remediating before 2020. All competent authorities are drawing up plans for the
upcoming period in order to comply to the new agreement. A new Program Management Team will be
formed to stimulate and help competent authorities and to monitor the progress of the goals of the
agreement.
Private companies have contributed to the remediation of a lot of contaminated sites over the last 35
years. Special programs have been carried out for filling stations, drycleaners, railway stations and
gasworks. Further and intensive collaboration with private companies is foreseen for the coming
period. The possibilities for receiving a contribution from the government for private investment in soil
remediation will be modernized. Some administrative burdens will be dissolved and the possibilities for
receiving a contribution for an area based approach will be widened. Private companies also have a
great interest in clean groundwater as a source for their production. The parties involved in the
agreement and private parties will investigate the possibilities of a contribution from companies who
have interest in clean groundwater in the costs for managing the groundwater plumes.
4. New legislation
When all urgent sites with unacceptable risks are remediated or under control, soil remediation of the
other less contaminated sites can be integrated in the regular spatial planning processes. In the future
remediation of contaminated land in the Netherlands will no longer be a stand-alone process but will
only occur as a side effect of developments or in case the ecosystem functions of the groundwater are
under pressure. Therefore the Soil protection Act will be withdrawn in 2018 and legislation on
contaminated land will be incorporated in the new Environment and Planning Act.
The Environment and Planning Act is now under construction, but some broad outlines for soil already
exist:
Most important is the fact that the new legislation will not alter the way soil protection is
regulated. There is no right on pollution so that new contamination will remain forbidden. As
much as possible generic rules for soil protection will be incorporated into the Environment
and Planning Act.
For the remaining historical contamination the site specific approach will be transformed into a
development based approach. Combine different interests! As long as there is no
development no actions have to be taken. If a location is historically contaminated the
competent authority can consider to advise the owner of the location to take specific
precautions (for example no consumption of vegetables from the garden).
If a development is planned the initiator needs to access the quality of the soil. The initiator will
be facilitated with a register of conducted investigations and soil quality maps in order to
reduce the burden of investigations. If the soil at the location of development meets the
standards for the future function, no further (soil) action is needed.
When the soil quality does not fit for the future function remediation is needed. A set with
generic rules about risk assessment and remediation will be incorporated into the legislation.
No specific permits are needed for remediating. Therefore remediation will be carried out as a
natural part of the development.
The Environment and Planning Act provides the competent authorities with integrated
instruments. For example an integrated long term perspective for the area drawn in a plan.
With these instruments the contribution of soil and subsoil to societal goals can be addressed.
5. Lessons learned
Since Lekkerkerk 35 years have passed. In these years a lot of sites have been remediated and a lot
of knowledge has been generated. The end of an era of remediation is in sight, the countdown is
nearly completed. What can we learn from the Dutch approach? Some key success factors are
already mentioned. The most important lessons learned are:
A long term political and financial commitment is needed to solve the problem of contaminated
land
A strict preventative policy and legislation is essential. This is also beneficial for the industry
because prevention measures are cheaper than cleaning up.
It takes more than a generation to build up knowledge and experience and complete a
thorough inventory of the magnitude of the problem of contamination land.
Decentralized operation is an essential factor in the Netherlands, because of the buildup of
capacity, the relation to spatial planning and the close connection of the authorities with the
owners of contaminated land
Focus is needed on the countdown of top-priority locations in order to incorporate soil
remediation in the normal process of redevelopment of urban areas and brownfields.
The need for integrated legislation on the Environment and spatial planning has been a great
stimulus for the focus on remediating the contaminated locations with urgent, unacceptable
risks.
.
The Netherlands have a broad base of experts on dealing with contaminated land. For the upcoming 5
years these experts do have a home market and will contribute to the final countdown. After the period
2015-2020 these experts will find their way in multidisciplinary teams dealing with issues on
environment and planning. For foreign countries who still have a large task in dealing with
contaminated land these experts could be a welcome source of knowledge.
i
Convenant Bodemontwikkelingsbeleid en aanpak spoedlocaties 2010-2015
ii
Midterm review 2013, Doorpakken
iii
Convenant Bodem en ondergrond 2016-2020
Session: ThS 2.5 Reuse of contaminated soil and sediments 1
Objectives: To illustrate how contaminated sediments can transform in situ landfills into
a viable engineering structure without transportation and in an economical way.
A growing global population, climate change, rising wealth, increased consumption, increased
awareness of good health and sanitation practices, more safety regulations and legislation are just
some of the many factors that explain our increased water consumption. Water is becoming scarce
and an expensive commodity in many places on the globe. Meanwhile, sludge volumes have
increased and despite the regulations, the volume of contaminated sludge is not decreasing either.
This evolution has stimulated industrialists and researchers to develop all kinds of ways to separate
solids and liquids in a faster and more economical manner, and sand separation and fine-particle
separation are outcomes of this development. The actual use and re-use of contaminated sediments is
still limited because of the cost involved. Since the owners of the larger sludge volumes are public,
local, regional or national governmental bodies with limited funds, the backlog of sludge that needs to
be dredged from waterways and harbors is still increasing. At the same time, regulations are being
tightened, potential landfill sites are ever fewer, and transport distances are increasing, making the
financial picture worse.
As long as 50 years ago, TenCate developed a technology that is still regarded as innovative by many
today. Originally used as simple sand/water separation units in maritime projects, the tubes functioned
mainly as geotechnical elements in the landscape. It was not until the seventies that the first attempts
were made to dewater sludge and slurries containing organics. More advanced technology and better
control over the dosing of flocculants and coagulants have made it possible since to clean up many
contaminated waterways and harbors in a safe and environmentally sound way. A combination of the
different tube characteristics, namely the ability to perform as a geotechnical element in the landscape
and its filtering and containment properties, now offers solutions which are practicable even for certain
contaminated sludge sites.
We will illustrate the various possibilities TenCate Geotube offers to re-use sludge by discussing
some of the thousands of projects we have been involved with.. Although the complexity of works
presented rises in a linear manner, the order in which the works are presented is not in fact
chronological
The dewatered silt trapped and strengthened in these Geotube systems was used as fill to construct
a sports field parking lot. A total of 34.000m of liquid mud was removed from the harbor by a local
dredging company early 2013, to find a function in nearly 900 linear meters of drainage tubes placed
on a waterproof membrane and covered with a waterproofing membrane to completely confine the
sediments. A layer of fill material was then brought on to create the parking lot and the adjacent sports
field. Neither ground transportation nor storage of contaminated sediments had been necessary since
the sludge had been pumped to the site hydraulically from the dredge. This solution using Geotube
systems has resulted in significant savings compared to conventional landfill as well as in reduced
cost compared to conventional dewatering treatments. It also represented the most positive carbon
footprint.
A different way of incorporating dewatered sludge into the landscape was used in Herne in Germany
in 2009. The moat surrounding Herne castle was filled with leaves and other organics; in summer,
these were the cause of unacceptable odor problems. It was clear for the job owners, the city of Herne
that simply draining the moat and digging out the trash mechanically could endanger the timber-pile
foundations of the castle.
A small hydraulic dredge pumped the dirty mass at 150m/hr out of the moat into dewatering tubes. A
specialist company from the Netherlands, controlled the inline flocculation.
The tubes were filled in an excavated area in a low-lying part of the park lawns. The effluent that
streamed out of the tubes was pumped back into the moat, keeping the water level intact. After a few
weeks, the sediments in the filled TenCate Geotube units had consolidated sufficiently and the tubes
were covered with the original soil. The fabric of these particular tubes is made of polypropylene, and,
being sheltered from UV influences, there will be no biodegradation. This working method allowed
haulage of the wet sludge, the dewatered sludge and of the used tubes to be avoided completely. It
was fast, economical and ecologically friendly. The problem was solved without leaving a trace and a
previously wet area of the park is now just as dry as the surrounding lawns.
In Brazil, a bulk cargo terminal was build using dredged sediments out of the surrounding area. To
construct this 848.000m2 area, more than 1.5 million m3 of fill was required to achieve the platform of
+3,5 meters. The import of the fill material had a major financial impact for the owners of the project.
Besides of importing fill material, the surrounding area had to be dredged for 600.000m3 of
contaminated sediments. Originally it was foreseen that the contaminated dredged material was
disposed in an offsite facility at extra cost. This initial plan proved incompatible financial model of this
development. With the help of TenCate engineers a solution was provided where the dredged
sediments where used/reused in the foundation of the area.
The solution was to use large geotextile tubes to contain and dewater the contaminated sediment
within onsite dewatering cells. These cells where placed under the container storage area, over which
final fill and pavement are placed.
The idea, being totally new, asked for major design considerations:
- The ability of the geotextile tube system to contain and dewater contaminated sediments in a
secure environment.
- To treat all of the effluent and to return this effluent back to the native environment without any
impact.
- To provide a stable platform on which to store ocean containers stacked up to 7 layers high
- To construct a site with a 40 plus year design life
The engineers of the project owner, together with TenCate engineers, came up with a scheme of
testing and trials before the final project could start.
The end conclusion of this first project on this scale is:
- That there are Geotechnical/Environmental Standards of practice and modeling tools available
to predict the field results of Geotextile tube technology.
- Geotextile tubes can be used to contain, dewater and consolidate contaminated sediments in
significant quantities and to allow for use as structural fill
- In this case, TenCate Geotube technology created a saving for construction of the Embraport
project of approximately $50 million, making the project financial/economical feasible.
References
CUR. (2009) Hoogwaardig bouwen met baggerspecie in geotextiele tubes, CUR-publication 222,
Gouda, The Netherlands
Ed Trainer, Mike Watts. Disposal of Coal Mine Slurry Waste using Geotextile Containers at the North
River Mine, Chevron Mining Inc. Geosynthetics 2009, February 25-27, 2009, Salt Lake City, Utah,
439-448.
Koffler, A., Van den Berg G. (2008). A new approach to Dewatering contaminated sediments, I2SM,
Lille, France
Yee T.W., Lawson C.R., Wang Z.Y, Ding L., Liu Y. (2011). Geotextile tube dewatering of contaminated
sediments, Tianjin Eco-City, China. Geotextiles and Geomembranes (2011),
doi:10.1016/j.geotexmem2011.07.005, Elsevier Ltd.
T. Stephens, L.C.Q.C. Melo, M.M. Futai (2013) Beneficial Use of Dredged Contaminated Sediments
using Geotextile Tube technology at a container port in Santos, Brazil
Theme 1. Dealing with contamination of soil, groundwater and sediment
1c. remediation technologies and approaches
KEYWORDS: Construction team, Fixed budget, Time pressure, Chemical oxidation, Stimulated
anaerobic degradation
ABSTRACT:
Overview
An old decaying industrial site, a bankrupt galvanizing company, old town, dilapidated buildings with
asbestos, extremely high concentrations chlorinated ethenes in soil and groundwater, a cocktail of
other substances, heterogeneous soil with lots of clay, intersecting liabilities, no dynamics and no
money. That is in a nutshell the situation as it was found fifteen years ago. After many discussions,
ideas, initiatives, studies and many years of continued last year launched the reorganization and with
amazing results.
Preparation phase
Twenty reputable contractors, often accompanied by consulting engineers saw a big challenge in this
job. They could sign up through a rigorous pre-qualification. This resulted in the top five forwards.
These five contractors were allowed to submit a plan to address how to tackle the problem. One of the
main requirements was the budget of 1.5 million Euros.
Based on the remediation targets the conceptual model was updated through additional research. The
pure product was present in a much larger area which led to a greater demand for a robust approach.
Through lab experiments and feasibility testing, followed by a pilot in an extremely short period of four
months, a final clean-up design was drafted based on chemical oxidation and stimulated anaerobic
biodegradation. In the presentation we will take you in the process in which despite great pressure of
time thoughtful and creative choices were made regarding technique, performance but also the
formation of contracts in which one issue was maintained, namely the budget of 1.5 million Euro.
The pressure of time took a lot from all the actors, especially at times of changes, such as the
discovery of new spots with pure product. In the presentation, we want to show how the construction
phase has taken place and that it is possible also to perform the most robust techniques without
compromising developments or infrastructure.
USING INNOVATIVE GEOTEXTILE CONTSTRUCTIONS AS AN IN-SITU BIOREMEDIATION
TECHNIQUE TO REMEDIATE CONTAMINATED SEDIMENTS AND TO IMPROVE WATER
QUALITY OF SHALLOW LAKES
F.M. (Chiel) Lauwerijssen (1), P.J. (Paul) Stook(2), G.A.R. (Gustav) Egbring (3), M.H.P. (Maarten)
Jansen (4) and M. (Marcel) Klinge (5)
Abstract
Many shallow lakes suffer from accumulation of fine sediment. This sediment build-up is caused by
different factors like shoreline erosion, peat degradation, internal organic production and external
inflow or run-off. Fine sediment is kept in suspension, thereby inhibiting growth of higher order aquatic
plants and natural water quality improvement. In a protected wetland area innovative and sustainable
measures have been undertaken to reduce hydrodynamics, sediment resuspension and turbidity in
order to remediate contaminated sediment, enhance biodiversity and improve water quality.
Two types of light weight geotextile constructions were built to steer fines and to store sediment: the
Sediment Settler and Sediment Storer. Due to the Sediment Settler fines transport decreased by 20-
40 %. One year after installing the Sediment Settler a few decimeters thick sediment layer had
already been formed at its lee side. Also about 15.000 cubic meters of slightly contaminated sediment
was dredged and transported to Sediment Stores in order to alter hydrodynamics and induce nature
development. By doing so, maintenance costs were minimized and sediment had been beneficially re-
used. By the end of the first growing season the sediment top layer was covered by a pioneer
vegetation of reedmace. Even species that have disappeared 30 years ago returned.
This pilot study has shown that it is possible to control sediment settling and transport and improving
recreational and ecological quality of this wetland area. Nautical bottlenecks were reduced and in lee
areas vegetation develops. So, synergies were realized between remediation, nature restoration and
(at some locations) shore line protection. Dredged material can easily be kept and stored in
submerged basins.
Keywords: Geotextile construction, sediment storage, fine transport, water quality, shoreline
protection
1. Introduction
Surface waters including streams, rivers, lakes, estuaries and coastal waters are under increasing
ecological stress due to anthropogenic activities worldwide (UNEP, 1999), leading to the enforcement
of major environmental policies, including the European Water Framework Directive (WFD) 2000/60
(EC, 2000). Nutrient over-enrichment (eutrophication) is one of the stressors, which has altered
ecosystems completely.
The Wormer- and Jisperveld area is the largest uninterrupted peaty grassland in Western Europe and
one of the most important areas for wetland birds in the Netherlands (Niedekker, 2006). It is under
European Natura2000 protection and under the WFD classified as surface water body. This nature
area is located 20 kilometers north of the city of Amsterdam and has a surface area of about 2500
hectare. There are three shallow lakes that are connected to each other by small channels. Most
ditches are manmade and left overs of peat excavation in the past. Peat was harvested and spread
over rectangular shaped parcels to dry characterizing peat lands in the Netherlands. In figure 1 an
Figure 1: Aerial photograph of Wormer- and Jisperveld area with is shallow lakes and small channels.
As a result of industrial activities in the 17th and 18th century heavy metals (copper, lead and mercury)
have accumulated in the sediment. Wind mills were used for paper and paint industry and industrial
wastewater was discharged directly into the surface water. The shallow lakes formed a natural sink
for this pollutants. Some sediments still contain high amounts of heavy metal concentrations (Tauw,
2013).
Dutch policy and legislation stimulates re-use of soil and sediments. The application possibilities are
mainly determined by the concentration of pollutants in the sediment. Usually in this type of areas a
small hopper excavates the dredged material and transports it to a field depot for ripening. However
sediments within the shallow lakes are slightly contaminated. Due to high heavy metal concentrations
it is prohibited to spread it out over nearby parcels. Transporting it by ships or trucks, on the other
hand, will result in high costs.
In many shallow lakes, inorganic sediment particles, but also algal cells go through a rapid cycle of
sedimentation and resuspension. Often, resuspension is mainly due to wave action, but also fish
searching for food in the bottom can stir up considerable amounts of sediment in some situations
(Scheffer, 2004). The water quality of the Wormer- and Jisperveld is poor. It is rich of nutrients and
due to water movement fine sediment particles are kept in suspension and impede sunlight
penetration to the bottom (Witteveen+Bos, 2013a). These fines inhibit the growth of higher order
aquatic plants and natural water quality improvement. Shorelines and channels erode, a process that
is catalysed by the poor water quality, and fine suspended particles settle down in lee areas where
the water currents are slow. Since all waterways are connected fine particles are easily transported by
channels. Near harbors and channel mouths sediments settle and nautical bottlenecks arise. Since
drastically changing the land-use, the constant water levels and the trophic status of the water is not
possible without severe societal and financial consequences for the area, we looked for other ways to
make a change in this negative pattern.
Ecological recovery can be possible if a number of conditions are met (Jansse et al 2008). The most
important condition is the nutrient load of a lake (see figure 2). The nutrient load must be reduced to a
certain level. If this level is reached other measures are possible, to alter hydromorphologial
characteristics in order to increase the resilience of a water system. Steering fines is a promising
measure, since it removes nutrients from the system. Thereby making lakes less susceptible for
external (nutrient) loads compared to the same systems where fines are suspended. This is due to
the effects on water transparency and the reduction of the internal nutrient load (Smolders et al,
2006). Steering fines will only contribute to ecological recovery if nutrient load is low enough, if the
system is changed in a net eroding lake to a net sedimentation lake and if the food web is adapted to
the new conditions. In many cases it will be needed to take additional measures to force the process
of adaption and fully restore the lake.
So, a good solution to reduce or solve the problems with fine sediment in a cost effective manner is to
steer the transport of fine sediment with relative small measures. Fine sediment could be caught by
the creation of lee areas, where silt can settle. Depending on water depth and transparency
submerged and/or emerged vegetation may develop, thus increasing further settlement conditions
and local biodiversity recovery. Dieter (1990) has shown that zones of emergent vegetation help to
reduce wind resuspension in shallow lakes.
Figure 2. Different types of water management measures depending on nutrient loading relative to
critical nutrient loading, modified from Jansse et al 2008.
A good morphodynamic numerical model describes the water levels, currents, waves and sediment
transport in lakes and the exchange of fines with the surrounding area. Every lake has its own
hydrodynamic and morphologic behavior. Therefore the kind of numerical modelling is different for
each lake. In long shaped lakes the wave growth in the length of the lake is important (focus on
waves), while in round lakes the wind induced flow is more important (focus on currents and waves).
In deep lakes waves are less important than currents (focus on currents), while in shallow lakes both
waves and currents are more important (Soulsby et al., 1993). In lakes with a varying depth, density
flows can be important and three-dimensional hydrodynamic modelling is necessary. In order to get
insight in wave pattern and existing transportation patterns, a numerical hydrodynamic model
(Witteveen+Bos, 2013b) was created (programs Delft3D and SWAN).
This model made it possible to visualize water movement and overall transportation routes through
channels and lakes for the most important wind directions (Jansen et al., in prep 2015). Special
attention was paid to the grid size in the lakes and the channels. In the lakes a grid size of 10 m was
applied. Also the orientation of the grid referring to the channels was improved, to avoid irregular
boundaries (Weare, 1979).
The method presented in Dusseljee et al. (2012) has been extended to get in the wave attack and
present transport patterns of fines. Wind data from KNMI, location Berkhout was used to derive the
local wind climate at Wormer- and Jisperveld. Wind direction 210N is the most important wind
direction.
Based on this wind climate the main wind directions and wind speeds with occurrence of 10%, 20%
and 50% were selected and used to calculate the flow and waves in the Wormer- and Jisperveld.
These calculations show the most important flows and waves in the area.
The model also shows the sources of fines. The most important sources are the south-west oriented
banks and the channels.
Figure 3: Wind-induced and wave-induced flow currents for the reference situation during wind from
210N at about 20 m/s wind speed.
Sediment removal is often unsuccessful in controlling internal nutrient loads sufficiently and reducing
algal biomass on the long term (Peterson, 1982; Welch & Cooke, 2005). Furthermore, dredging is an
expensive and disruptive control measure. Besides, in this area it is impossible to use heavy
13th International UFZ-Deltares Conference on Sustainable Use and Management of Soil,
Sediment and Water Resources, 9-12 June 2015, Copenhagen, Denmark.
machinery to build constructions, to excavate dredged material or to transport heavy materials. Since
the area is under European protection as well the impact of dredging activities must be minimized.
So there was a need for a light weight construction that was easy to install and to remove without
damaging the environment. In order to store sediment a screen was developed that was strong
enough to withstand pressure and filter sediment form water. This made gravity and buoyancy the
decisive elements in the design.
Over the last decade geotextiles (Geotubes) are used for dewatering sediment under pressure.
These materials have proven themselves to be strong and to have a long lifespan. Together with
TenCate Geosynthetics B.V. we developed a new type of construction (see figure 4). A modified
prefabricated Geotube unit was linked to
a geotextile screen and a floater. The small
Geotube is filled with sand and used as
an anchor to keep the structure in place. A
floater and propylene lines are used for
tensioning the screen to wooden anchor
poles. The enclosed water surface behind
the construction could then be used to
store sediment. The water depth after filling
determines the kind of vegetation that will
develop over time (emergent or submersed
vegetation)
First, a pilot was conducted to store 500 cubic meters of sediment behind a Sediment Storer of 50
meters in length (Stook et al., 2013). Several weeks after installation of the Sediment Storer in March
(2013) vegetation started to develop. By the end of the growing season the sediment top layer was
totally covered by a pioneer vegetation of reedmace (Typha), thereby isolating the contaminated
sediment underneath and forming a natural cap over the years.
Roots provide structure and induce dewatering of sediments by plant transpiration. Also, water birds
and small mammals found their habitat in this wetland. In winter time the Sediment Storer withstands
ice coverage and storm surges (wind speeds of 140 km/h).
Monitoring the sediment layer around the construction has shown that the whole construction is
strong enough to face the expectations. However, aspects that needed attention were the floater and
the screen. The PVC-floater was not easy to handle and could not resist frost well. Besides, some of
the fines seemed to pass through the screen, since the water level had risen within the enclosed
Further improvement of de Sediment Storer resulted in the innovative design of the Sediment Settler.
The Sediment Settler now reduces the intensity of wave action (like a groyne) and prevents sediment
resuspension. Besides the construction should be save for water users if collision occurs. This time,
the floater was not made from PVC but Styrofoam
(fabricated by DOW Chemicals), which is stronger.
Since vegetation rapidly developed behind the
Sediment Storer two floaters were used to create a
basket that could keep sediment inside. Testing
several geotextiles and non woven materials under
extreme stress in lab experiments, resulted in a
suitable thin non-woven layer that was both
permeable by water and able to maintain fine silt
particles. After filling the basket with sediment, the
construction was covered by a mesh.
In order to define the most effective positions to reduce eddies and to create lee areas, both
constructions were fit in the numerical model to examine the effects. To do so, several positions were
examined at different wind directions. The combination of measures seemed to most effective. For
example, in the northern lake large lee areas were created by Sediment Storers at its west bank. At
the east bank, smaller Sediment Stores were installed to restore damaged shorelines. Near the centre
of the lake Sediment Settlers were installed at the location where strong eddies occurred. Figure 8
shows the influence of these measures on hydrodynamics (Witteveen+Bos, 2013b).
In the southern lake Sediment Stores are able to reduce the width of the mouth of channels to prevent
further erosion and reduce water currents. Sediment Settlers seemed to be most effective in reducing
hydrodynamics at locations where strongest water currents are present.
The bottom shear stress gives a good indication of the effects of measures on erosion and
sedimentation. Figure 9 gives the relative effect of measures in relation to the reference situation. It
shows a decrease of 20% to 50% in the northern lake and in the northern part of the southern lake. In
the middle of the southern lake, in the navigation channel the shear stresses will increase, which will
result in erosion.
Figure 9: Difference in maximum bed shear stress between the design variant and the reference
situation for a wind speed of about 20 m/s from 210N.
In the winter of 2014/2015 more Sediment Storers have been installed to remediate slightly
contaminated sediment. Due to pilot results small modifications were made. The same non-woven
material of the Sediment Settler was now
placed behind the geotextile screen to prevent
sediment from passing through the screen.
Also, the floater was covered by UV-resistant
material so a long lifespan can be guaranteed.
8 CONCLUSION
Erosion and sedimentation of fines are ongoing processes and many shallow lakes and harbors suffer
from accumulation of fine sediment. Moreover, high amounts of sediment negatively influence water
quality and recreation. This study has shown that it is possible to steer sediment settling and therefore
recreational and ecological quality of this wetland area can be improved. At nautical bottlenecks draft
depths are kept and in lee areas vegetation develops. Transportation of fines has decreased
considerably and in some lee areas emerged vegetation developed, demonstrating the potential of
this approach to solve problems with fines and to contribute to ecological recovery of the lakes.
The light weight geotextile constructions designed for this project were easy to handle and to install.
Despite their low weight they have proven to be both resilient and strong. In winter time the
constructions withstand ice coverage and storm surges (wind speeds of 140 km/h). Small
modifications have been made in order to improve its filtering capacity and enlarge its lifespan. Due to
these constructions slightly contaminated sediment could be stored under water, thereby minimizing
maintenance costs and CO2 emissions. Surprisingly, by the end of the first growing season the
sediment top layer was covered by a pioneer vegetation of reedmace. Even species that had
disappeared 30 years ago returned.
So, synergies were realized between remediation, nature restoration and shore line protection.
Dredged material can easily be kept and stored within the shallow lakes. It is expected that this
geotextile construction can also be used for shoreline protection or to prevent harmful algal blooms
reaching shores.
References
1. Legal Framework
OVAM stands for Openbare Vlaamse Afvalstoffenmaatschappij (Public Waste Agency of Flanders)
and is responsible for waste management and soil remediation in Flanders. It is a public Flemish
Institution, established after the decree of July 2nd, 1981. In 1995, Flanders got a specific legislation
on soil. The soil remediation decree provides the Flemish government with a powerful instrument to
fight and to remediate soil pollution. The Decree introduced a strict none-retroactive liability rule and
channeled the liability for future soil pollution. To avoid new pollution by transporting excavated
polluted soil, a new chapter was introduced in 2004.
In numerous construction projects, soil is excavated, removed from the site and used elsewhere. In
doing so, there is often a danger of contaminated soil being managed in a hazardous way. Initially the
regulation for the re-use of excavate soils solemnly focused on the prevention of creating soil pollution.
More recently the regulation additionally aims at visualizing the potential of excavated soil as
substitute to primary minerals. Through sustainable materials management it seeks to reduce the use
of natural resources throughout the life-cycle of materials.
Excavated soil is not to be considered as waste if it is used in accordance with the regulations of the
Soil Decree and its executory directives (VLAREBO). The regulation for the re-use of the excavated
soil is based on the stand-still principle. There cannot be any deterioration in the current environmental
condition and any increase in health or environmental risks must be avoided. In order to meet the
predefined targets, the regulation on the use of excavated soil imposes a soil quality survey. The soil
quality has to be determined in a technical report in the case of the excavation of hazardous soils or in
the case of building projects where more than 250 m unsuspected soil is excavated. The regulation
fixes the conditions for the use of the excavated soil. Depending on the degree of contamination with
possible pollutants, the use of the excavated soil is more and more restricted. The amount of
pollutants that may be present depend on threshold levels (background level, soil remediation
standards (SRS) that differ depending on land use type) of the Soil Decree. Non-contaminated soil
can freely be re-used as "soil". Somewhat polluted soil can be re-used as "soil" if the receiving land is
already more polluted of it can be re-used for building purposes in specific constructions. If the
contamination surpasses certain levels, the soil is considered waste and cannot be re-used. If the
contaminated soil cannot be treated, the soil is dumped on a landfill.
The Flemish government wants to discourage dumping of excavated polluted soils originating from
remediation operations and excavation projects. It promotes recycling of polluted soils by means of
environmental tax. Only soil that cannot be treated by biological, physico-chemical, or thermal means
can be dumped on a landfill at lower environmental tax levels. Dumping soil that can be treated is
discouraged with high taxes. On yearly basis about 800,000 tonnes of soil is treated.
Table 1: Tonnes of treated soil and amount of polluted soil dumped at reduced taxation level.
The protection of the environment but also the juridical protection of different actors (liability) involved
is ensured by the traceability procedure. Firstly the soil quality has to be determined in technical report
in the case of the excavation of hazardous soils or in the case of building projects where more than
250 m unsuspected soil is excavated. The traceability procedure designates further responsibilities
and prescribes rules for excavation, transport and the use on the location of re-use. The procedure is
supervised by a soil management organization that will attest the traceability of the soil by a soil
management report. The soil management organization is accredited by the Public Waste Agency of
Flanders (OVAM). OVAM represents the Flemish government and has the authority to check every
part of the chain, and if necessary take corrective measures. The soil management report stand for
the correct use of the excavated soil with respecting of the prescribed rules. Table 2 gives the
evolution of the number of technical reports and soil management reports that are treated by the soil
management organization. Table 2 also shows that although the mayor part of the soil that are
investigated can be re-used freely, a large part of the soil that is excavated is restricted in use due to
the presence of contaminants.
Table 2: Technical reports, soil management reports and the soil quality.
The data for assessing the potential of excavated soil as substitute to primary minerals is obtained in
the monitoring system for a sustainable surface mineral resources policy in cooperation with the
Land and Soil Protection, Subsoil, and Natural Resources Division (ALBON) and the Flemish Institute
for Technological Research (VITO). The monitoring system provides an overall picture of market
developments on an annual basis. Table 3 and figure 1 give the overall use of primary mineral surface
material and their alternatives during the year 2011. Half of the total need of surface mineral materials
is covered by alternatives such as excavated soil , recycled granulates.
Although there can be substantial fluctuations about 10,000,000 m excavated soil is re-used in
construction projects, either as soil or for building purposes. Table 4 and Figure 2 details the use of
the excavated soil. Most of the soil is re-used as soil for the raising of terrains. The use of soil for
structural purposes in landfills, dikes, barriers and infrastructural works is also important. About
300,000 tonnes is used in high grade substitute in concretes (coarse sands) or for the production of
bricks (clay).
4. Conclusion
The Flemish legal framework on the use of excavated soil define rules and methods that ensure
human health and environmental protection. Nevertheless, the availability of reliable basic data, such
as the total needs for primary mineral resources and the potential of possible alternatives as
substitutes to primary minerals is needed to take the step towards a sustainable management of
excavated soil.
The annual report of the monitoring system for a sustainable surface mineral resources policy
provides an overall picture of market developments on an annual basis. Sustainable materials
management seeks to reduce the use of natural resources throughout the life-cycle of materials
(independent of whether they are resources, finished products, or waste) and as such the monitoring
tool becomes an important subject in Flemish environmental policy.
Paper 4732_ Session 2.6 Reuse of contaminated soil and sediments 2
Introduction
Soil was one of the seven thematic strategies that were to be addressed in the 6th EU Environment
Action Plan (EAP) as an essentially non-renewable source due the time period over which soil is
formed. The value of soil as a resource is many faceted, performing a number of vital functions
including but not? limited to production, filtration and carbon storage. In fact to raise awareness of the
importance of soil to human existence this year, 2015, has been declared by the 68th UN General
Assembly as the International Year of Soils (IYS) (http://www.fao.org/soils-2015/about/en/). This
declaration also aims to:
support effective policies and actions for the sustainable management and protection of soil
resources; and
promote investment in sustainable soil management activities to develop and maintain healthy soils
for different land users and population groups.
The construction industry is often considered resource hungry with little regard historically to
minimising waste or looking for opportunities for reuse materials with issues regarding programme and
long term liability over performance of recycled materials being a concern (50% of the UKs waste
came from construction and demolition sector according to the UK Statistics on waste -210 -2012
released by DEFRA). It has certainly operated as a linear economy "take-make-consume and
dispose" model, which is based on the assumption that resources are abundant, available and cheap
to dispose of. A model which is becoming increasingly apparent is not sustainable and does not make
economic sense. By comparison, a circular economy seeks to preserve the value of the
product/resource for as long as possible and virtually eliminates waste. It retains the resources within
the economy when a product has reached the end of its life, so that they remain in productive use and
create further value.
For a circular economy in soil to work it is important to understand the pressures that result in impact
on the soil resource by the construction industry which includes:
covering soil with impermeable materials, effectively sealing it and resulting in significant
detrimental impacts on soils physical, chemical and biological properties, including drainage and
carbon storage capacity characteristics;
impairing soil quality by:
o over compacting through excessive trafficking by heavy plant;
o contaminating through spillages of chemicals and/or fuels;
o poor storage and handling resulting in loss of nutrients, soil structure, geotechnical properties
(soil becomes too wet to place without treatment);
o poor management practices resulting in mixing of soil types for example subsoil and topsoil;
disposing of surplus soils to landfill.
In essence, soil in construction historically has not been valued with loss due to sealing and landfilling
of surplus soils whether contaminated or not. Whilst sealing, in circumstances, may be unavoidable
landfilling often is not. Over the years, weve all witnessed material going to landfill that could be
reused on another site, so what are the barriers to such intuitively cost effective and sustainable
solutions. In line with the EU 7th EAP whose second action area concerns the conditions that will help
transform the EU into a resource-efficient, low-carbon economy AECOM has been seeking to promote
greater resource efficiency and has firsthand experience of the challenges to soil reuse even in an
economy where landfill tax has increased the drive for more sustainable solutions.
To encourage the regeneration of brownfield sites for a number of years it was possible to get a
Landfill Tax exemption for contaminated soils that needed remediation; a technique known as dig and
dump. The removal of this exemption at similar time as developing in England a protocol (The
CL:AIRE Definition of Waste:Development Industry Code of Practice) to make it easier to reuse soils
has encouraged investment in on-site remediation techniques and a network of soil treatment facilities.
So what does sustainable soil use look like in construction and how can the waste hierarchy be
applied. In terms of prevention we need to not only be adopting normal good practices such as cut
and fill balance, storing soils appropriately so that do not deteriorate ( and become unusable) and
seeking reuse opportunities on site through landscaping and/or noise mitigation but we should be
challenging the unsuitable for use assumptions. Is the soil really hazardous? What is the cost benefit
of further land contamination risk assessment? Questioning which materials should be retained (cost
of disposal versus resale value)?
Recycling arisings is considered to be some form of intervention that significantly changes the
product and could include
stabilisation can be geotechnical or geo-environmental driven;
topsoil manufacture; and
generation of recycled aggregate.
The application of such techniques ultimately depend on whether suitability for use can be achieved
subject to logistics, programme and economics or whether it simply ends up in the to difficult to do
box as discussed below in considering what prevents soil reuse.
treated to make them suitable for a particular purpose but at what point in the process does that soil
cease to be a waste.
In the UK we have a contaminated land framework which is risk based which increases the
opportunity for soil reuse: a soil that may not be of suitable chemical quality for a residential
development with gardens may be suitable for a site in a commercial setting. However would you
naturally call a contaminated land specialist to discuss your infill requirements?
Added Complications
Data
What every project has and often simultaneously fails to havesoils testing for waste disposal
purposes versus soils testing for reuse.
For construction arisings in the UK not to be considered waste they need to:
be suitable for use;
have certainty of use;
use only the quantity necessary and no more; and
be protective of human health and the environment (not contravening the waste framework
directive).
Suitability for use is not just understanding the soil quality in terms of chemistry but if it is to be used in
construction its engineering properties too. So often when commissioning a site investigation, Clients
simply ask for one number site investigation and the lowest cost wins. To win site investigation
contracts mean minimal scope and if not asked for, these contracts do not automatically include
geotechnical testing for reuse. Whereas for disposal the primary interest is the chemistry of the soil
and usually comprises testing to understand the properties of the waste and in particular whether they
are hazardous or not. Some waste soils are listed as Inert and do not require further testing unless
suspected of being impacted by anthropogenic activity. Consequently when this happens there is
insufficient information to meet the criteria without further testing which inevitably causes delays and
depending on programme and space on site may result in the expedient solution of disposing to
landfill.
Another flaw at the data stage is failing to identify the volumes of the different soil types that a donor
has available. In order to maximise value and increase the chances of finding an alternative receiver
site to a landfill facility then for each soil type there should be a volume. Soil types essentially are
classified on their particle size, whether they are granular or cohesive in nature and their
contamination status.
Ecology
In short, stockpiled material can inadvertently become the home of both protected and non-native
invasive species, both flora and fauna. For example, amounts of rubble and other loose material on
site (including stone gabions, piles of wood and broken concrete) can provide very suitable temporary
summer and permanent winter shelter.
Soil can become vegetated in a matter of weeks/ months, thus increasing its attractiveness as a
prospective habitat. The degree of risk associated with this eventuality is highly variable and
dependent upon many factors, such as:
The proximity of the site to locations high in biodiversity, whether formally recognised (e.g.
designated Sites of Specific Scientific Interest) or not;
The manner in which material is stockpiled (tightly packed or loosely spread? fenced? covered?)
The types of material being retained (how permeable and durable?);
Proximity of site to water bodies; and
Time of year that the stockpiles are going to be dismantled amongst others.
Considering the species side of things more specifically from a UK legislation perspective, the primary
concerns include the following:
Badgers - Protected from damage and destruction of setts, obstruction of access to setts and
disturbance of any resident badgers should they be present, under the Protection of Badgers Act
1992. Whether or not badger setts are found during ecological surveys of stockpiled materials, if a
habitat is potentially suitable, setts may be dug at anytime and it cannot be assumed that there none
present, just because they are not noted during survey works. Closure of setts only permitted
between 30th June and 1st December under Natural England Licence.
Common reptiles e.g. grass snake Natrix natrix common lizard Lacerta vivipara, slow worm Anguis
fragilis and adder Viperus berus. Protected from killing and injury under the Wildlife and Countryside
Act 1981 (as amended). Reptiles could use stockpiled mounds as hibernation sites during the winter
months or temporary shelter at night or during the hottest parts of the day during the summer months,
which would delay clearance or removal activities due to unavoidable disturbance caused. Any
destruction without translocation could likely lead to direct killing/ injury and therefore be in breach of
the legal requirements.
Smooth snakes and sand lizards - Protected from disturbance, killing and injury and their places of
rest and shelter are protected from destruction, damage or obstruction by the Conservation of
Habitats and Species Regulations 2010 (as amended) and by the Wildlife and Countryside Act 1981
(as amended). If these species are utilising the stockpiles or any other works are carried out that could
affect their habitat/place of rest and shelter, this would result in the need for a Natural England
Development licence, possibly with the requirement for similar replacement habitat and a translocation
of any smooth snakes or sand lizards prior to works. Undertaking winter removal of stockpiles would
result in infringement of the WCA 1981 (as amended) and Conservation of Habitats and Species
Regulations 2010 (as amended).
Breeding birds All breeding birds are protected while nesting under the WCA 1981 (as amended).
There is opportunity for both tree and ground nesting birds on and surrounding stockpiles/ mounds,
which can in turn affect the ability to operate heavy machinery in the vicinity.
Great Crested Newts - protected from disturbance, killing and injury and their places of rest and
shelter are protected from destruction, damage or obstruction by the Conservation of Habitats and
Species Regulations 2010 (as amended) and by the Wildlife and Countryside Act 1981(as amended).
Any suitable GCN habitat within 250m of a pond used by GCN is considered to be a place of rest and
shelter and therefore legally protected from destruction, damage or obstruction.
Mammals even if not specifically protected by designated ruling or act of law, there is a general
protection against any unlawful killing for all mammals under the Wild Mammals (Protection) Act 1996
whereby habitat destruction resulting in the inhumane killing or injury of wild mammals constitutes a
breach of law. As a result, even fox dens, rabbit warrens and the like cannot be simply disregarded.
Japanese Knotweed is perhaps the most notorious of the non-native invasive species that requires
specialist handling and disposal methods by law. It is one of the many species on Schedule 9 of the
WCA 1981 for which it is an offence to grow or cause to grow in the wild. Dispersal is interpreted
under this law as causing to grow in the wild.
As you can see, there is quite a lot to consider that many do not normally even think about when
undertaking earthworks activities, and more significantly, the legislation is extensive in its scope and
reach
Case Study 1
When is stockpile not a stockpile ? When it is a habitat....
Such an approach not only provided the Client with both cost savings on the ecology work but a more
engaged workforce Over 1000 animals were translocated successfully within one season and only
seven reptiles were subsequently found during our comprehensive destructive search; a testament to
3
the excellent translocation. Even with the added ecological complications at the site 20,000m of
material sorted and relocated to a land restoration site through AECOMs Waste Harmony Service,
saving the client over 600k compared to landfill;
Logistics
The logistics of making a match can actually affect the viability of a match. Not only is there issue of
synchronicity between programmes but soil does not appear all in one go. Can synchronicity be
achieved through storage? Is there space to store on the donor site? Or the receiver site? What is the
additional cost of double or even triple handling material? How will this effect programme? And
ultimately budget?
Culture
Developing a common language is key and has proved to be one of the more difficult aspects. Even
the simplest of words can wrap consultants and regulators up in knots. In navigating the myriad of
legislation, guidelines and protocols for reusing soil arisings, there are significant phrases and words
that lack clarity or do not join up. Consider 'inactive: which is how landfill tax guidance refers to
material eligible for the lower rate of tax. Or 'inert' as a classification for waste acceptance for
innocuous materials. Another example is the use of 'clean' within the CL:AIRE Code of Practice as a
core feature of soils potentially eligible for direct transfer between sites. Unsurprisingly, these words
are often used interchangeably by the less informed who do not appreciate they are not the same and
that all could be relevant when evaluating alternative disposal and reuse options.
A consequence of this is frustration and perhaps additional time and costs as consultants piece their
way through the options to find a best value, legally compliant solution. This situation can be further
compounded when regulators take a different interpretation, requiring extra effort on both sides that
would not be necessary for a simple landfill disposal.
Through AECOMs Waste Harmony Scheme a potential match was identified for a city centre
donor site where the donor advise they had 50,000m3 clean material. However, when the
report arrived 40% of the soil was made ground and although not contaminated from the donors
land use perspective it did contain contamination at concentrations above a residential with plant
uptake land use which could not be taken by the receiver site.
Financial Considerations
Costs savings are the most important driver of soils reuse. However, imperfect information and
understanding again create blockages and inhibit soil trading. Landfill tax is a classic example. How
many understand that non-hazardous waste classification does not necessarily mean the higher rate
of landfill tax applies? Certainly awareness is far from great among cost managers and engineers
involved in early-stage cost planning. Consequently, if the tender cost estimate is not accurate to start
with, how many clients are motivated to understand exactly where their soils are going? ~ In most
cases, if the contract price comes within budget and on programme there is no need to consider the
matter further, even if improvements could be made.
A second example of commercial viability that often takes centre stage is in prioritising leads for
matches between donors and receivers .If there are five leads on the table that could be suitable. how
will these be prioritised? Clearly haulage will be a major cost factor, but how far should the net be cast
in seeking an alternative before it becomes less cost effective than landfill? An obvious question
perhaps, but in practice the process of following up leads to establish commercial viability can be
laborious and time consuming. Furthermore, the market is constantly changing in line with supply and
demand in the local geography. If a project programme slips by a month, it could mean the difference
between paying for disposal and being paid for the material. Local fluctuations in supply and demand
can greatly affect the economics and value of material in a given area.
There are benefits to planning for arising at the early stages of project plan which in effect move the
potential value of any cost savings up the chain too. Ultimately the realising of these savings depend
on the Clients contractual arrangement and if you pass on all the risk in the ground then that usually
means you have passed up on the rewards of any costs savings! If clients were more directly seeing
the benefit of such savings then this may help to drive the sustainable soils agenda.
The frustration of our clients at the costs involved with importing and exporting soils provided the
inspiration for our new soils management service called Waste Harmony. Changes in legislation and
guidance mean that the concept of a soils dating agency is now a genuine reality. The name was how
we initially, somewhat jokingly, referred to the idea, but it has really captured peoples imaginations
and so weve decided to build on the analogy. Where someone has surplus soil arisings, this neednt
be a waste, but rather can be a valuable asset for someone who is looking for that material. Using our
client networks, it was clear that we could set up an exchange that provided a truly cost-effective and
sustainable alternative management solution.
With huge potential across the industry, this innovative idea has been developed with support from
AECOMs Creative and Technical Excellence Council. As can be seen from the discussions above
there are a myriad of challenges posed by what in essence is a simple concept but is more often than
not implemented for a range of reasons. The two key in-house tools that we have developed to
overcome some of these challenges are the financial model for understanding the economic viability of
matches and, in keeping with the dating agency theme, an online portal.
Assessing Viability
In order to be able to turn the market intelligence in to commercial opportunity, understanding the
associated costs and what this means in terms of how far material will travel cost effectively is vital.
It is also quite complex because of the interplay of numerous factors such as volume, material types,
level of treatment required, rules about how material can move between sites, amongst others.
AECOM has built a detailed technical model built on the analysis of market leading cost data produced
by Programme, Cost, Consultancy business line and technical soil disposal and management data
from the networks AECOM operates and interacts with as part of its core business. The AECOM
model enables us to very quickly estimate and compare the costs of landfill with alternatives sourced
through the Waste Harmony network. Questions of potential viability and savings are far more
readily at our finger tips, which means that we are able to react and pass on opportunities more
readily.
If there are no suitable matches, our portal will hold the registration open for a defined period of time
and will notify the manager, should a suitable match be registered at a later date.
Currently, access to the portal is for the AECOM Waste Harmony team only, so clients registering
with AECOM do so through a dedicated point of contact. However, in designing the portal, we have
allowed for the development of a client specific version and some clients have already seen the
benefit of helping to make matches within their own portfolio of sites.
Delivery
Another key aspects of the Waste Harmony Service Offering is an in-house team of dedicated
specialist brought together to meet our Clients needs whether they be:
A straight forward brokerage deal;
Consultancy to unlocked a potential deal; or
A whole service offering including undertaking soil treatment and/or movement.
Finding Client solutions requires technical expertise, legal understanding and financial assessment.
Supportedbyinhouse/externallegalandITservices
Conclusions
As can be seen from the above, there is still much to be gained by applying the principles of The
Circular Economy by pursuing soil trading and promoting reuse options, but there is a long way to go
in turning a simple vision into conscious common practice. It is clear that to have a successful match
between a donor and receiver, all the interfaces between the technical, legal and commercial aspects
need to be brought together and considered as a whole.
In conclusion:
Dont under estimate the human factor! Changes in working practice are often challenging;
The earlier you think about earthworks balance the more opportunity there is for achieving
sustainable soil use, leaving it until construction limits options and contractors have not got it all
resolved;
Understanding the economics (national and regional) is key sustainable soil use only works when
it is economically viable;
Inappropriate soil storage may decrease value; and
Contaminated land expertise required for infilling.
Useful Links
www.claire.co.uk
https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/416471/UK_Statistical_r
elease_UPDATEv6_19_03_2015.pdf
www.wasteharmony.com
www.wrap.oorg.uk
References
CL:AIRE(2011) Definition of waste: development industry code of practice, Contaminated Land:
Applications in Real Environments, London(ISBN:978-1-90504-623-2)
DEFRA (2009) Construction Code of Practice for the Sustainable Use of Soils on Construction Sites,
Department of Environment and Rural Affairs, London
DEFRA(2011) Guidance on applying the Waste Hierarchy, Department of Environment and Rural
Affairs, London
DEFRA (2015) UK Statistics on Waste 2010 to 2012, released 25 March 2015
Metin Turan1, Fikrettin ahin1, Nurgl Ktr1, Adem Gne2, Medine Gllce3, Gleray
Aar3, Hatice tc4
1
Department of Genetic and Bioengineering, Faculty of Engineering and Architecture,
Yeditepe University, Istanbul, Turkey
2
Erciyes University, Faculty of Agriculture, Department of Soil Science, Kayseri,
Turkey
3
Atatrk University, Department of Biology,Erzurum,Turkey
4
Ahi Evran University, Department of Biology, Krehir, Turkey
This study was conducted on aridisols, a soil ordo widely exists in Eastern Anatolia
Region. The trial was conducted in 45 pots with an experimental design of 9 x 5
factorial, 1 plant (wheat), control and 8 microorganisms (Bacillus megaterium M3,
Bacillus subtilis Osu-142, Bacillus pumilus C26, Paenibacillus polymxa, Azospirillum
brasilense Sp-245, Burkholderia cepacia BA-7, Burkholderia cepacia F + AMP,
Raoutella terrigena). Each treatment was five replicate. Plant and soil samples were
taken at the end of the growing period (90 days). Some soil enzymes such as acid and
alkaline phosphate, urease and dehydrogenase, amino acid exhausted from soil and plant
roots, and plant macro and micronutrients elements were determined. The results
obtained have shown that amino acids, plant and soil enzymes exhausted from root parts
significantly affected the wheat plant growing. The highest wheat dry matter was
obtained by Bacillus subtilis Osu-142 PGPR application. But application of PGPR
decreased amino acid exhausted from plant roots. The lowest amino acid exhausted
obtained Paenibacillus polymxa PGPR. Positive correlations were determined between
the PGPR and wheat plant nutrition and dry matter. On the contrary, exhausted amino
acids had negative correlations with PGPR application. The plant enzymes such as
catalase (CAT), peroxidase (POD), super oxide dismutase (SOD) were increased by
PGPR application and the highest values were obtained by application Paenibacillus
polymxa PGPR.
INTRODUCTION
Wheat, on the world among the plants cultivated crop in the field is the first place, while
of great importance in human nutrition is cultivated plants (Anonymous, 2005). The rapid
increase in world population, and vegetable products in a sufficient increase in lack of
hunger, the problem becomes more apparent. This is a problem in eliminating grains and
cereals for wheat has a very important role.
Plant growth regulation (PGPR) of the mechanisms involved has not been revealed,
although the existing literature within the framework of direct and/or indirect describes
two groups. Direct mechanisms, biological nitrogen fixation, auxins, gibberalin effects of
plant hormones, such as the production of ACC deaminase enzyme activity through the
synthesis of ethylene prevention, environmental stress reduction, bacteria-plant
adaptation in the relationship, inorganic phosphorus solubility and increasing the organic
phosphorus compounds of the mineralization, siderophor through the manufacture of iron
to meet the increasing and some other trace elements to provide a rate increase, vitamin
synthesis, root permeability, the effect of increasing their. PGPR have effects of
indirectly with antibiotic production diseases biocontrol agents to reduce the role they
play as the various organic compounds in the soil washing blocker ksenobiyotik break
down the plant for the protection (Elsheikh and Elzidany 1997; Rodriguez and Fraga
1999).
Used in the experiment soil samples, in the examination of specimens calciorthid soil
great group of aridisol is located at the Erzurum, Turkey. The plant material as bread
wheat (Triticum aestivum L. spp.) varieties were used.
This study was conducted on aridisols, a soil ordo widely exists in Eastern Anatolia
Region. The trial was conducted in 45 pots with an experimental design of 9 x 5 factorial,
1 plant (wheat), control and 8 microorganisms (Bacillus megaterium M3, Bacillus subtilis
Osu-142, Bacillus pumilus C26, Paenibacillus polymxa, Azospirillum brasilense Sp-245,
Burkholderia cepacia BA-7, Burkholderia cepacia F + AMP, Raoutella terrigena). Each
treatment was five replicate. Plant and soil samples were taken at the end of the growing
period (90 days). The microorganism-containing solution in the seeds held for a period of
2 hours drying on paper, dried, and 180 kg ha account were planted in pots. The
greenhouse conditions of the temperature and humidity datalogger values (testo 175-H2
V01.10) measured using different moisture level in the watering intervals, and
determining the amount of irrigation.
RESULTS and DISCUSSION
Trial some of the soil chemical analysis results are given in Table 1.
Table 1. Used in the experiment soil samples belonging to some of the chemical analysis
results (n=5)
Mean
pH (1:2.5) 7,760,05
CaCO3 25,661,25
Organic matter % 2,2190,30
Total Nitrogen 0,00110,0002
NH4-N 9,500,4
NO3-N mg kg-1 10,40,5
P 12,101,1
CEC 33,191,4
K 2,710,1
Ca cmol kg-1 12,421,4
Mg 2,890,02
Na 1,010,05
Fe 3,330,02
Cu 1,850,01
Mn mg kg-1 5,280,10
Zn 1,240,02
B 0,320,01
In the experiment, the microbiological properties of soil samples in order to examine the
soil samples taken on the result of the analysis, 162 x 106 bacteria, the amount of total C
of 8.90 mg C m2 h-1, and the total amount of CO2 32,63 mg CO2 m2 h-1 was determined
as (Table 2).
Table 2. Used in the experiment soil samples belonging to some of the microbiological
analysis results (n=5)
Mean
Initial Bacteria Count (x106) 162,003,60
Initial Fungal Count (x104) 185,004,20
Soil C amount (mg C m2 h-1 ) 8,901,22
Soil CO2 amount (mg CO2 m2 h-1 ) 32,632,60
the number of soil microorganisms (106 ) 1794,42
The Root Rhizosphere Region of Amounts Total Amino Acids By The Wheat Plant
Roots
Graph 2. Effects of PGPR on total amino acids content by exhausted wheat plant roots
CONCLUSONS
In this study,the high lime content of wheat grown in soils bitkisininverim in order to
improve different features Pgpr. Applied to different pgpr of wheat plants of the plant
effect on weight compared with the control group, the weight of the plant B. subtilis Osu-
142 with the application of PGPR control application according to the 19% increase was
determined.
Applied to different pgpr of the most effective identified as P. polymxa PGPR of the case
of the application of soil by roots, root rhizosphere region of the amino acid secreted in
the amount of a 34 % decrease in the rate was determined.
Application of P. polymxa significantly increased plant enzymes that CAT, POD and
SOD to control groups as CAT 16%, POD, SOD 21% respectively.
REFERENCES
Anonymous, 2005. FAO Statistical Databases. (www.fao.org)
Elsheikh, E.A.E., Elzidany, A.A., 1997. Effects of Rhizobium noculation, Organic and
Chemical Fertilizers on Yield and Physical Properties of Faba Bean Seeds. Plant
Foods for Human Nutrition, 51, 137-144.
Rodriguez, H.,Fraga, R., 1999. Phosphate solubilizing bacteria and their role in plant
growth promotion. Biotechnology Advantages, 17, 319-339
Theme 2. Soil, groundwater and sediment in the biobased, circular economy
Viktria Feigl, Mnika Molnr, va Ujaczki, Orsolya Klebercz, Ildik Fekete-Kertsz, Mria Tolner,
Emese Vaszita, Katalin Gruiz
Budapest University of Technology and Economics, Department of Applied Biotechnology and Food
Science, Address: 1111 Budapest Megyetem Rkp. 3., Hungary, E-mail: vfeigl@mail.bme.hu
Abstract
Biochars are produced from organic materials by pyrolysis and are used as soil amendment.
According to the feedstock type and the conditions of pyrolysis the properties of the biochars can differ
widely. If we apply biochars derived from organic waste materials as soil ameliorant it is important to
assess the hazards and the risks of their application. In this study we assessed the physical, chemical,
biological and ecotoxicological properties of thirteen biochars from two producers. Our aim was to
assess their applicability as soil amendment prior to microcosm and field trials and to choose the best
biochars able to improve the quality of degraded soils. The biochars were produced in a PYREG type
pyrolyzer at temperatures between 450 and 700 C during 15-20 min residence time. The feedstock
were grain husks, paper fibre sludge, digestate, wood screenings, miscanthus, vine, black cherry,
natural biomass, straw, hazelnut shells, olive stones and meadow. Some biochars were post-treated
with compost, organic liquid or stone powder. To assess the potential benefits and risks of their
application to soil we measured their water holding capacity, pH, toxic metal content, microbial activity
and toxicity to plants (Sinapis alba and Triticum aestivum) and animals (Folsomia candida). The
biochars from mixed organic wastes with minerals and the biochar made from vine had elevated toxic
metal content that indicates the potential risks of their application. The biochars from straw and natural
biomass were toxic both to plants and animals. The grain husk biochars with fibre sludge, the black
cherry biochar and the biochar from wood screenings seemed to be the most promising soil
ameliorants as they ensured favourable conditions to plants, bacteria and soil living animals.
1. Introduction
Biochar is the product of thermal degradation of organic materials in the absence of air (pyrolysis), and
is distinguished from charcoal by its use as a soil amendment (Lehmann and Joseph, 2009). Biochar
has been described as a possible means to improve soil fertility as well as other ecosystem services
and sequester carbon to mitigate climate change (Lehmann et al., 2006; Sohi et al., 2010).
The European Biochar Initiative (EBC, 2012) has recommended the biomass categories for production
of biochar, including garden waste, agricultural and forestry waste, vegetable production waste, animal
by-products, paper production waste etc. The feedstock affects several biochar properties with
agronomic implications, including ash content (affects the soil mineral content), the H/C ratio
(approximates aromaticity of the biochar and is an indication of its ability to be mineralized), pH
(increases soil pH of acidic soils and affects mobility of ions in the soil), surface area (helps predict
CEC and possible sorption of GHGs) and cation and anion exchange (may determine the potential for
+
NH4 and NO3 retention in N cycling) (Singh et al., 2010; Liang et al., 2006, Lehmann and Joseph,
2009, Krull et al., 2009). The chemical characteristics of biochar produced from a feedstock depend
considerably on pyrolysis conditions, especially temperature (Antal and Gronli, 2003). At low
temperature, biochar chemical composition is closer to the original feedstock while high temperature
biochar is closer to graphite (Masiello, 2004).
Benefits of biochar as a soil amendment may vary with its properties, time after its application, and in
relation to soil texture and mineralogy. The observed effects on soil fertility have been explained
mainly by a pH increase in acid soils (Van Zwieten et al., 2010) or improved nutrient retention through
cation adsorption (Liang et al., 2006). However, biochar has also been shown to change soil biological
community composition and abundance (Grossman et al., 2010) having a variety of effects on the soil
biota which may be associated with its impacts on C and N cycling. In spite of the positive effect that
biochar may have on the soil, increasing attention is paid on biochar contamination with polycyclic
aromatic hydrocarbons (PAHs) and trace metals (Freddo et al., 2012), therefore posing a potential
threat to the environment, a fact confirmed by ecotoxicity tests (Oleszczuk et al., 2013).
This paper has the aim to assess the physical, chemical, biological and ecotoxicological properties of
13 biochars from two producers and to recommend the best fitting biochar for improvement of the
quality of sandy soils.
Thirteen biochars produced from various organic waste materials were investigated (see Table 1). The
biochars originated from two biochar producers (marked with A and B) and were produced in a
PYREG type pyrolyzer. Type A biochars were produced at 450500 C pyrolyzis temperature with 20
minutes residence time and type B biochars were produced at 600700 C pyrolyzis temperature with
15 minutes residence time.
To test the possible applicability of the biochars as soil amendment the following properties were
measured: pH (Hungarian Standard 21470/2-81:1981), water holding capacity (hlinger, 1995), loss
on ignition (ash content) (Sluiter et al., 2008), toxic metal content (portable NITON XL3 Analyzer),
aerobic heterotrophic colony forming units (Lorch et al., 1995), common wheat (Triticum aestivum) and
white mustard (Sinapis alba) root and shoot growth (Leitgib et al., 2007), Collembola (Folsomia
candida) mortality (Wiles et al., 1998). Inhibition percentage to plant growth and Collembola mortality
was calculated compared to sandy soil. Selected soil properties were also measured for biochars
mixed into sandy soil (from Hungary) in 10 weight%.
To select the best biochar(s) for the improvement of sandy soils an evaluation system based on
scores (from -5 as the least ideal to 5 as the most ideal) were created. The system took into
consideration the ideal properties and effects of a soil amendment for acidic and sandy soil, eg. good
water holding capacity, alkaline pH, low ash content, support for microbes, low toxic metal content and
no toxic effect.
The results for the thirteen biochars are summarized in Table 1 and 2. It is clear that the properties of
biochars from different feedstock vary. However, nearly all of the biochars had alkaline pH and low ash
content (high loss on ignition), except for the ones containing stone powder. Some contained high
amount of toxic metals (eg. Zn and Cr) compared to the Hungarian Quality Criteria for soil. Most of the
biochars have good water holding capacity as they are able to hold more than 100% water compared
to their dry mass.
Table 1: Physical and chemical properties of the biochars
Ignition
Name Feedstock WHC pH loss
Mo* Zn Cu Cr
% % mg/kg
A1 grain husks and paper fibre sludge 169 8.8 60 8 34 15 <DL
A1 post treated with stone powder
A2 105 6.8 32 2 33 <DL <DL
and compost
A1, digestate and different minerals
A3 116 8.3 37 14 775 123 59
like stone powders, P and Fe
A4 A3 post treated with organic liquid 114 8.0 40 12 863 133 24
B1 wood screenings 151 9.3 80 12 181 26 31
B2 miscanthus 268 9.2 88 15 29 14 <DL
B3 vine 179 9.8 84 13 316 114 59
B4 black cherry 169 8.5 95 14 46 16 <DL
B5 straw 312 10.0 82 14 274 23 33
B6 hazelnut shells 69 9.6 96 13 63 39 200
B7 meadow 197 9.0 93 22 82 43 154
B8 natural biomass 135 9.8 86 11 30 <DL <DL
B9 spelts mixed with paper (2:1) 107 9.0 69 14 227 86 19
* The Hungarian Quality Criteria for soil based on KvVM-EM-FVM Joint Decree No. 6/2009. are 7
mg/kg for Mo, 200 mg/kg for Zn, 75 mg/kg for Cu and Cr. DL: detection limit
The heterotrophic aerobic colony forming number of bacteria and fungi shows that biochars may have
been colonized by microorganisms during storage in the open air (Table 2.). This indicates that
biochars may ensure a habitat for a variety of microorganisms in soil. The toxicity tests show that most
biochars represent a favourable environment for plants, however, five biochars have strongly inhibited
(more than 80% inhibition) plant growth. The biochars have moderate effect on Collembolas: most of
them caused 2040% mortality.
To evaluate the possible toxic or beneficial effect of biochars when applied as soil ameliorants,
biochars were mixed into sandy soil at 10 weight% (at 25 times higher biochar rate compared to
usual application rate). The water holding capacity of the sandy soil increased 1.41.7 times in most
cases (Table 3.). The same biochars showed toxic (from vine, straw and natural biomass) or
stimulating (from grain husks and paper fibre sludge) effect to plants similarly to the tests without
mixing them into soil. The results of the Collembola tests, however, are contradictory. It proves the
importance of microcosm and field experiments before large scale applications.
Table 3: WHC and toxicity of biochars mixed into sandy soil in 10%
Mustard Mustard Wheat
Name WHC* Wheat root Collembola
root shoot shoot
% Inhibition %
A1 50 -15 -15 12 -26 35
A2 39 28 57 22 46 60
A3 41 15 65 31 21 20
A4 40 33 77 -10 47 5
B1 45 15 29 -1 -14 10
B2 - -23 -36 15 64 5
B3 - 69 66 89 95 0
B4 48 -13 11 0 -17 10
B5 - 64 55 84 93 13
B6 - -18 21 75 91 13
B7 47 7 41 -203 28 70
B8 46 69 60 53 68 15
B9 42 15 34 18 -22 15
*WHC for sandy soil was 27%. -: not measured
A scoring system was created to evaluate the applicability of biochars as soil amendment. The results
are shown in Figure 1. Larger total score indicates better performance. From the point of view of
further application for the treatment of sandy soils the best biochars were: A1 from grain husks and
paper fibre sludge, B1 from wood screenings and B4 from black cherry. However, the treated or post-
treated biochars (A2, A3 and A4) seem to be less effective for soil improvement compared to the
untreated biochar (A1) based on the selected properties. Vine and straw biochars did not perform well
according to this evaluation.
30
Total score
20
10
Figure 1: Total score for the tested biochars according to feedstock type
4. Conclusions
In this study selected properties of 13 biochars were tested to assess their future applicability as soil
ameliorant. It can be concluded that physical, chemical, biological and ecotoxicological properties of
the assessed biochars depend greatly on the pyrolyzed feedstock type. However, the effect of
pyrolyzis temperature, residence time, age, storage etc. could also change the biochars properties
and could be further evaluated. We concluded that biochars from grain husk and paper fibre sludge,
wood screenings and black cherry had the most favourable properties regarding their application as
soil ameliorant in sandy soils. Therefore as the next step of our experiments we will apply these
biochars in soil microcosms and field plots for the improvement of sandy soils (see Molnr et al.,
2015).
5. Acknowledgement
The work was carried out in the frame of the Terra Preta project, registration number HU09-0029-A1-
2013 supported by the EEA Grants and the Norway Grants within the Green Industry Innovation
Program of the Norwegian Financial Mechanism 20092014.
References
Antal, M.J., Gronli, M., 2003. The art, science, and technology of charcoal production. Ind. Eng. Chem.
Res. 42, 16191640
Grossman, J.M., ONeill, B.E., Tsai, S.M., Liang, B., Neves, E., Lehmann, J., Thies, J.E., 2010.
Amazonian anthrosols support similar microbial communities that differ distinctly from those extant in
adjacent, unmodified soils of the same mineralogy. Microbial Ecology 60, 192205
EBC, 2012. European Biochar Certificate Guidelines for a Sustainable Production of Biochar.
European Biochar Foundation, Arbaz, Switzerland, http://www.european-
biochar.org/biochar/media/doc/ebc-guidelines.pdf, (accessed on 25.11.14)
Freddo, A. Cai, C. Reid, B.J. 2012. Environmental contextualisation of potential toxic elements and
polycyclic aromatic hydrocarbons in biochar, Environ. Pollut. 171, 1824
HS 214702:1981, Environment soil tests. Soil sample preparation, moisture content, pH and
electrical conductivity determination. (In Hungarian) Hungarian Standards Institution, Budapest
Krull, E. S.; Baldock, J. A.; Skjemstad, J. O.; Smernik, R. J. 2009. Characteristics of biochar: organo-
chemical properties. In Biochar for Environmental Management: Science and Technology; Lehmann,
J., Joseph, S., Eds.; Earthscan: London, UK
Lehmann, J., Gaunt, J., Rondon, M., 2006. Biochar sequestration in terrestrial ecosystems - a review.
Mitigation and Adaptation Strategies for Global Change 11, 403427
Lehmann J. and Joseph, S. 2009. Biochar for Environmental Management: Science and Technology,
J. Lehmann, S. Joseph (Eds.), Earthscan, London (2009), pp. 112
Leitgib, L., Klmn, J., Gruiz, K. 2007. Comparison of bioassays by testing whole soil and their water
extract from contaminated sites. Chemosphere, 66, 3, 428434
Liang, B., Lehmann, J., Solomon, D., Kinyangi, J., Grossman, J., ONeill, B., Skjemstad, J.O., Thies,
J., Luizo, F.J., Petersen, J., Neves, E.G., 2006. Black carbon increases cation exchange capacity in
soils. Soil Science Society of America Journal 70, 17191730
Lorch, H.J., Benckieser, G., Ottow, J.C.G., 1995. Basic methods for counting microorganisms in soil
and water. In: Alef, K., Nannipieri, P., (Eds.), Methods in Applied Soil Microbiology and Biochemistry.
Academic Press, London. pp. 146161
Masiello, C.A., 2004. New directions in black carbon organic geochemistry. Mar. Chem. 92, 201213
Molnr, M., Feigl, V., Farkas, ., Fekete-Kertsz, I., Ujaczki, ., Klebercz, O., Tolner, M., Vaszita, E.,
Gruiz K. 2015. Soil improvement with biochar - Microcosms for characterization of the effects of
biochar on acidic sandy soil, Conference Proceedings of AquaConSoil 2015, Theme: 2.6. Reuse of
contaminated soil and sediments
Oleszczuk P, Josko I, Kusmierz M. 2013. Biochar properties regarding to contaminants content and
ecotoxicological assessment. Journal of Hazardous Materials, 260, 375382
hlinger, H. 1995. Maximum water-holding capacity. In: Schinner, F., hlinger, R., Kandeler, E.,
Margesin, R. (Eds.), Methods in Soil Biology. Springer-Verlag, Berlin Heidelberg, 385386
Singh, B.; Singh, B. P.; Cowie, A. 2010. Characterisation and evaluation of biochars for their
application as a soil amendment. Aust. J. Soil Res., 48, 516525
Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton D. 2008. Determination of Ash in
Biomass Laboratory Analytical Procedure (LAP), Technical Report, NREL/TP-510-42622
Sohi, S., Krull, E., Lopez-Capel, E., Bol, R., 2010. A review of biochar and its use and function in soil.
Advances in Agronomy 105, 4782
Van Zwieten, L., Kimber, S., Morris, S., Chan, K.Y., Downie, A., Rust, J., Joseph, S., Cowie, A., 2010.
Effects of biochar from slow pyrolysis of papermill waste on agronomic performance and soil fertility.
Plant and Soil 327, 235246
Wiles, J.A., & Krogh, P.H., 1998. Tests with the Collembolans Isotoma viridis, Folsomia candida and
Folsomia fimetaria. In: Lokke H. & Van Gestel C. A. M. (Eds.), Handbook of soil invertebrate toxicity
tests, Chichester: John Wiley & Sons, 131156
LETS MAKE GROUNDWATER STRONGER A
WATERSYSTEM-BASED APPROACH
TOWARDS 3D SPATIAL DEVELOPMENT.
Almer Bolman1 and Reinier Romijn2
Spatial development of the subsurface is only feasible when the connective function of
groundwater is taken into account. To accomplish this in The Netherlands, cooperation
and sharing of knowledge between authorities is crucial because our groundwater
management is splintered.
Introduction
Groundwater management in The Netherlands
The way water is managed in The Netherlands was praised in the 2012 OECD report on financing
water resources management [OECD, 2012] and again in 2014 [OECD, 2014]. We would be the
last to argue this laudatory conclusion but there is a side effect to a durative strong organisation.
On the one hand the specialisation that accomponies it leads to better solutions in specialised fields
of work. On the other hand, it hampers effective cooperation across specialisations, as it becomes
harder to communicate on the same level.
In the same year the OECD report [OECD, 2012] was received, the Dutch Water Authorities
launched a report that concluded groundwater management is splintered over multiple policies and
agencies [Sterk Consulting and Colibri Advies, 2012]. Figure 1 depicts the interrelationships on a
legislative level. Combined with a high level of specialisation, the effectiveness of the well
organised water management is under pressure from multifaceted challenges like adaptation to
climate change and spatial planning in the subsurface. This urges for new solutions without spoiling
proven practices.
1
Vallei en Veluwe Waterboard, Apeldoorn, The Netherlands. abolman@vallei-veluwe.nl
0031555272911
2
Dutch Water Authorities, Den Haag, The Netherlands. rromijn@uvw.nl 0031703519751
Both from the notion that formal responsibilities for groundwater management are splintered and
from the growing interest to use the subsurface, the Dutch Water Authorities are urged to appeal
for a central role for groundwater in the development of the subsurface.
Figure 1. Relationships of groundwater to the legislative landscape on the national level and on the
European level. Adaptation from [Sterk Consulting en Colibri Advies, 2012].
STRONG is to bring order in subsurface planning. The primary focus is on national issues, i.e.
mining activities and drinking water abstraction. Mining activities are of economic importance while
potable water is a basic need of life. Both have a large impact on the surroundings and are strongly
conflicting activities. Mining activities have a vast potential for groundwater pollution whilst
drinking water requires large areas of groundwater protection. In The Netherlands, this conflict
explicitly arises with shalegas exploration: two concessions for exploration of shalegas have been
granted under the Mining Act. However without considering environmental effects, as this Act only
considers economic aspects of activities. It is now under revision to incorporate environmental
effects after public debate broke loose and conflicts between authorities arose.
Regional and local uses of the soil and subsurface are in their own way both potentially conflicting
and vulnerable. Lowering of groundwater levels, for instance, to accommodate agriculture needs or
urban expansion has, in the second half of the last century, led to dessication of groundwater-
dependent nature. Soil subsidence, due to lowering groundwater levels is a big threat to built-up
areas on peaty soils.
More recently, aquifer thermal energy storage became popular. Installations may unintentionally
influence one another, so many local governments have drawn up master plans to enable optimal
use. In open heatflux systems (which physically abstract and reinfiltrate groundwater) it is possible
to combine heat storage systems with remediation measures. This is applied in area directed
groundwater management. The Dutch Water Authorities appeal to widen this approach to area
oriented planning to involve quantitative aspects as well.
To enable this approach, a general view of all aspects concerned is essential. This is very difficult to
achieve, as legislation and knowledge are disperged over a number of layers of government. The
fields of knowledge are too wide to incorporate into one organisation, so reorganisation of tasks
seems not to be the best answer. Regional cooperation is the key to success. To make it work it is
essential to share information at a basic level, with clear paths to deeper knowledge and/or
resource persons.
Different activities in the underground can influence both groundwater flow and groundwater
regimes, thus introducing pressure on land use not only at the site of the activity but at a distance
as well. Groundwater abstraction can divert groundwater flow and drop groundwaterlevels in the
surroundings. Some effects of activities in the subsurface have only been found by (trial and) error.
Underground parkings for instance can block groundwater flow, in a similar manner weirs may
block the flow of surface water. The effect: unexpected problems with flooding basements in the
neighbourhood.
Recent and past mistakes
Dutch past has brought forward many examples of unforeseen effects, such as desiccation of
natural habitats and dry rot: affection of the wooden piles underneath building foundations by rot
as a result of low groundwater levels.
The Dutch Technical Commission for the Soil (TCB 2014) has collected a number of anecdotes of
unexpected effects as a result of underground activities. In many instances, the effects on
groundwater flow have not been sufficiently taken into account. The effect can be severe and it can
take many years for the problems to emerge.
A lack of knowledge on the subsurface can be an important aspect in these failures. In Gronau
(Germany), a recent plan to store mineral oil in salt caverns failed dramatically, as it turned out
these caverns where not sealed. Mineral oil seeped up at a farm site. With increasing depth from
the soil surface, our data and thus our information certainty diminishes at a fast pace, as it is quite
expensive to perform deep measurements. This can lead to wrong decisions as this example has
proved.
The lessons learned from past mistakes and the concern to prevent new ones are the moving
forces for the Dutch Water Authorities to emphasize the need to take groundwater effects into
account in spatial planning for the underground. This takes effect in the Authorities input for the
STRONG, which is focused on the sustainable and efficient usage of the subsurface, based on
common challenges addressing subsurface and groundwater issues.
In figure 2, two different approaches towards solving problems in the subsurface are depicted. The
classical approach, liability based, would be for the parties involved to first determine what their
legal obligations and rights in the matter are. This would take up a lot of time, depending on the
severeness of the case and prospected cost of measures. A solution based approach sees the
parties involved step out of their legal context to first focus on the best solution: high effectiveness
at lowest cost. In the next step, dividing input is easier because of the prior process. The allocation
can be based on the necessary competencies in the particular matter. Also see the example The
sulphate plume that didnt get away in the examples chapter.
A A
Domain of org. A
problem Domain of dispute
Domain of org. B A and B agree on best
solution, then bid for domain solution
B B
Figure 2. Difference between two approaches towards problems in the subsurface involving
multiple authorities (A and B in this case): liability based versus solution based. The latter leads to
the best value for money from the taxpayers point of view. See text.
Many initiatives help to further establish consistency in the complex matter of groundwater and
subsurface management. Consistency between scales is an important issue. The European
Groundwater Directive has helped to promote the awareness of interdependency of groundwater
quality on a regional scale. The link to general water management has been made more evident.
Recent shifts in legislative responsibilities concerning groundwater management, along with the
arising awareness of climate change and the discussion on the exploration of shale gas deposits,
have raised the general attention to groundwater management.
A relatively new initiative is the so called area oriented groundwater management. At first devised
to help solve complex groundwater contamination issues, it also is very useful for considering the
bigger picture of groundwater related issues.
The design of the National Structural Vision on the Subsurface (in Dutch abbreviated to STRONG)
has furthermore accommodated the awareness of the connective capacity of groundwater in the
subsurface. In Dutch spatial planning, groundwater management traditionally was a mere aspect of
different fields of legislation: soil management, water management, water management. Now all
authorities are joined in the design of STRONG, which focuses on a sustainable and efficient usage
of the subsurface, based on common challenges addressing subsurface and groundwater issues.
The Dutch Water Authorities emphasized the fact that groundwater management is complex in The
Netherlands. It is an aspect of different fields of public management and splintered over different
authorities and jurisdiction. Groundwater management is emerging as an object of its own and
simplification of legislation is helping to achieve that. Now, the different authorities need to
combine their competencies.
The organisation of groundwater management in The Netherlands has been evaluated [Ellen et al,
2014] and the final concept of the evaluation of the relatively new Water Act has recently been
published [Ministerie van Infrastructuur en Milieu, 2015]. Both conclude that groundwater
management is complexly organized in The Netherlands, yet revision of legal responsibilities is not
considered to be the best solution as yet. Cooperation is promoted as the key. Authorities should
define common goals and strive to find operational uniformity. Sharing knowledge is an important
aspect. By cooperation, letting go of the formal bounds of legislative control, the best organisatory
fit will become apparent and that would be the time to discuss reorganisation.
Sharing knowledge
A groundwater based approach to subsurface planning requires specific knowledge. This knowledge
is available in The Netherlands, but spread over many authorities and institutions. Therefore it is
hard to produce a comprehensive overview. Besides this, it is still not common practice to take into
account the combined effect of various activities. It is essential to join forces and share knowledge
to enforce this. Therefore, the national and regional authorities are cooperating in establishing a 3D
spatial planning for the subsurface. Also at different regional scales many joint efforts arise to
establish sustainable use of the subsurface.
Figure 3. Scheme for sharing knowledge in subsurface developments. In the common knowledge base,
essential knowledge is available at a base level. When aspects are relevant, specialists need to be
involved.
As an example: The Netherlands have made groundwaterflow models publicly accessible. This has
started with the Netherlands Hydrological Instrument [NHI website and De Lange et al, 2014]) and
has amongst other initiatives been adopted by the consortium of water authorities behind the
model database of AZURE [AZURE website and Hekman et al. 2014]. The basic idea: a smal group
of owners make their model available to all possible users and consider improvements brought
forward by these users. In that way, the model database grows in quality. And, also very
important, use is made of a common knowledge base, which for instance facilitates the process of
granting permits for planned activities.
Examples of a cooperative, groundwater based approach
The sulphate plume that didnt get away
Under the city of Ede, an enormous pollution is slowly moving towards the central valley. The front
consists of highly concentrated sulphate, slowly but steadily followed bij Zinc and other pollutants.
A complicating factor is the fact that there is no formal standard for sulphates, yet it is known to be
lethal for cows (via drinking) at concentrations well below those of the plume. In the central
valley nearby, Natura2000 areas are to be protected and enhanced concentrations of sulphate will
Figure 4. 3D overview of the sulphate plume near Ede. From the plant at A, sulphate has leached
to plume B and threatens nature area C. Within the tract, deep wells at site D will draw part of the
pollution to the city's (E) and regional (F) water systems.
give rise to mobilisation of phosphates, which is devestating to the determined nature type. To
make things even more complicated, deep drainage wells in the city are likely to shortcircuit part of
the plume to surface waters in the town where it may lead to deterioration of concrete structures,
formation of unpleasant smells (hydrosulfide) and pollution of surface waters. Formally there was
no party responsible for the remediation of this pollution. However, the province, the city and the
waterboard have joined forces to divert the plume to the River system, where dilution renders the
pollutive agents harmless.
Even though millions of euros and various uncertainties as to the duration of controlling measures
were at stake, these governmental organisations rose over their formal jurisdictions to facilitate the
best remedy. The alternative may well have been a path of assertive actions arising from
jurisdictions, institutional conflicts, and many difficulties for water management, nature
conservation and urban livability.
Natural discharge of groundwater contamination to surface waters
Many contaminations slowly move with the flow of groundwater towards surface waters. This may
lead to unwanted concentrations of contaminants conflicting with the EU Water Directive. In the
example of the sulphate plume, it is clear that critical standards would be passed. The risk however
is not always as clear as in this case. Yet, measures may well be very costly. That is why the Vallei
en Veluwe Waterboard is collaborating with the municipality of Amersfoort, which is responsible for
soil protection, to devise an early warning system based on the natural qualities of the ground- and
surfacewatersystems involved. Knowledge of groundwaterflow, chemical aspects, background
Figure 5. Scheme for regulating "natural spills" where polluted groundwater is slowly entering
surface waters. Remediation is not always necessary. Combination of groundwater models with in
situ samples from monitoring wells could be an efficient monitoring system. Remediation is only
necessary if the load of an agent is likely to surpass the acceptable load of the receiving system.
concentrations, flow rate, et cetera are combined. This should lead to a cost effective monitoring to
identify the moment intervention is inevitable, avoiding both unnecessary measures and
unnecessary contamination of the receiving surface waters.
Conclusion
In this article, we have shown that:
Spatial planning for the subsurface is very complex and groundwater is an important connective
agent between activities.
The complexity and various fields of expertise urge for cooperation and in order to achieve that it is
necessary to exchange knowledge at a basic level.
Past examples show it is very easy to cause unforeseen effects when groundwater flow is not taken
into account. Groundwater management should therefore be more central to subsurface planning,
considering not only 3D space, but the temporal aspect as well.
Recent examples show, that when cooperation between authorities is based on finding the best
solution rather than focus on formal jurisdiction, more efficiency and effectivenes is achieved.
AZURE, http://www.azuremodel.nl
De Lange, W.J., Prinsen, G.F., et al, 2014. An operational, multi-scale, multi-model system for
consensus-based, integrated water management and policy analysis: The Netherlands Hydrological
Instrument, Environmental Modelling & Software, vol. 59, September 2014.
Ellen, G.J., Nijburg, C., De Putter, P., 2014. Provincies en Waterschappen: Samenwerking in
grondwater. Water Governance Centre, december 2014.
Hekman, A., Van Maanen, H., De Lange, W.J., 2014. AZURE: Innoveren in de gouden driehoek,
H2O, January 2014. Available on: http://www.vakbladh2o.nl/index.php/h2o-online/recente-
artikelen/entry/azure-innoveren-in-de-gouden-driehoek
Ministerie van Infrastructuur en Milieu, DG Ruimte en Water, 2015. Evaluatie van de praktijk van
het grondwaterbeheer eindconcept 23 maart 2015.
OECD (2012), A Framework for Financing Water Resources Management, OECD Studies on Water,
OECD Publishing
OECD (2014), Water Governance in the Netherlands Fit for the Future? OECD Studies on Water,
OECD Publishing
Oomes, J, 2015, Management of the subsurface in spatial planning, in writing for the Proceedings
of AquaConsoil 2015.
TCB (2014) Onverwachte gebeurtenissen in de ondergrond gevolgen van ons handelen in beeld,
TCB R23
Aquifer Thermal Energy Systems (ATES)
- in areas of drinking water and groundwater pollutions -
1 Introduction
As the need and interest of renewable energy-source increases, so does the focus on sustainability of our use
of our energy resources, from where new and improved technology arises.
Modern buildings and facilities in Northern latitudes most often needs cooling during summer and heating
during winter. This means that heat must be removed from the building or facility for a part of the year while it
must be added for another part of the year. While both heating and cooling traditionally has been achieved
with fossil fuels, directly or indirectly, the interest for more sustainable and cost-effective solutions has in-
creased, and Aquifer Thermal Energy Systems provide such a solution.
Aquifer Thermal Energy Systems (ATES) utilize a combination of wells, pumps, heat exchangers and heat
pumps to move heat from a building or facility into an aquifer during warm periods and to move heat from the
aquifer into the building or facility during cold periods. Thus, an ATES system facilitate the storage of heat in
times of excess and retrieval in times of deficit, the only necessary energy input being the energy needed to
drive the pumps.
ATES systems are open loop systems, which means that a sets of wells, with some inter-distance, are used.
During cooling periods, water is abstracted from the aquifer through the cold well, heat is transferred to the
water via a heat exchanger at the building or facility, and it is reinjected through the other well, the warm well.
During heating periods, the flow is reversed and heat is now pumped from the warm well, heat is abstracted
at the building or facility and the cold water is reinjected into the aquifer through the cold well. The concept of
an ATES-facility is shown in Figure 1.
Another reason for the slow implementation is that all drinking water supply in Denmark is based on ground-
water extraction, which entails that the protection of the drinking water resource has the highest priority. An
ATES-facility can in a worst-case scenario disturb a drinking-water well by e.g. increasing the groundwater
temperature or changing the hydrological conditions in the vicinity of the well.
Furthermore, an active ATES-facility can have an effect on point-source pollutions. A change in the hydrolog-
ical conditions in a specific area may change the direction and velocity off the groundwater flow and may
mobilise pollutions.
Before an ATES facility is built, the potential impacts on the surroundings must therefore be evaluated and
ocumented. In the following, a feasibility study done for an ATES system located at a site with both heavily
polluted areas and drinking water abstraction wells, is presented.
2.1 Introduction
The company InterXion, which provides cloud data, runs a server park in the municipality of Ballerup, west of
Copenhagen. Thus, InterXion have a large number of servers, which needs cooling, especially in the summer
period. In winter, there is still an excess of heat at the server park and therefore has no need for heating. But
in order to maintain thermal equilibrium in the aquifer, heat is nonetheless abstracted from the aquifer and
simply released to the atmosphere in winter.
The need for cooling at the server park during summer is equivalent to a maximum of 500 kW. With a steady
pumping rate of 60 m3/h this corresponds to a temperature increase of 7 to 10C of the abstracted water during
the cooling period. In winter, when flow is reversed, heat has to be removed from the water at a similar rate
and thus the temperature of the abstracted water has to be reduced by 7 to 10C. The actual temperature
difference is adjusted depending on the individual monthly needs for cooling and heating.
The ATES-facility at InterXion consists of one warm well and one cold well and the 60 m3/h pumped from
one well is injected in the other well. The locations and IDs of the wells are shown on Figure 2.
2/10
Figure 2. Location of InterXion.
The hydrogeology in the area consists of a pre-quaternary limestone layer overlain by quaternary layers con-
sisting of sandy and clayey sediments. The limestone layer, and the quaternary layers located directly above
it, acts as the primary aquifer in the area. The hydraulic head in the primary aquifer is located above the top
of the aquifer, hence it acts as a confined aquifer. In Table 1 the lithology as seen in the well with the ID
DGU.nr. 200.6556 is shown.
Table 1. Lithology in the area of interest based on well DGU nr. 200.6556.
In Figure 3 a cross-section of the hydrogeology in the area is seen. It is seen that the layers are consistent
with no lenses.
3/10
Figure 3. A cross-section of the hydrogeology at InterXion.
The primary aquifer has a transmissivity in the range of 3 12 x 10-3 m2/s. The transmissivity is based on
pumping tests conducted in the area.
As seen on Figure 4 the hydraulic head in the area is located at approximately 17 m DVR90 at InterXion. The
natural groundwater flow in the primary aquifer has a southerly direction with a hydraulic gradient of 0.1 %.
Eight drinking water wells are located within a radius of 1500 meters from InterXion. These wells are abstract-
ing water from the primary aquifer. The location of the abstraction wells is shown in Figure 4.
Figure 4. Location of abstraction wells for drinking water in the area. Furthermore, the hydraulic head in the primary aquifer is shown.
4/10
2.2.3 Land-use and point-source pollution
The InterXion server park is situated in an industrial area and several expected (V1 mapped) and documented
(V2 mapped) point-source pollutions are present in the vicinity. The nearest residential area is located 300
meters to the west.
As seen in Figure 5, ten suspected (V1-mapped) and four known (V2-mapped) point-source pollutions are
located in very close proximity to InterXion. Especially the pollution located on Brydehusvej 21, see Figure 5,
is very well-documented and sees a strong interest from the authorities.
At Brydehusvej 21, very high concentrations of chlorinated solvents are found. These are located in the upper
fill and clay sediments. There has been no indications of the chlorinated solvents having reached the primary
aquifer. There is an on-going remediation at Brydehusvej 21, which includes drainage of the chlorinated sol-
vents from the clay sediments.
The other pollutions in the area consist of oil and gasoline-components and are not as big a threat to the
primary aquifer as the pollution on Brydehusvej 21, nor as well-documented.
Figure 5. Location of documented and possible soil point pollutions in the area.
5/10
3 Analysis
In order to be able to evaluate the potential impacts of the ATES system on pollutions and existing groundwater
abstraction, a 3D hydro-thermal model was set up. A calibrated and validated 3D hydro-thermal model can
provide detailed information on expected groundwater and heat flow during the operation of an ATES system.
The 3D hydro-thermal model was set up with Feflow 6.2. Feflow is a 3D finite element groundwater model
software, which can handle coupled groundwater and heat transport. The model was calibrated as a stationary
model against measurements of hydraulic head and the model was validated transiently against a pumping
test performed in well DGU. Nr. 200.6556. The model was neither calibrated nor validated with regard to its
thermal parameters. Thermal parameter values were taken from literature.
The model were set up to simulate 10 years of operation of the ATES-facility with pumping rates of 60 m3/t
and a heat addition and reduction of 7-10 C via the heat exchanger.
In the following, it will be analyzed if the implementation of an ATES system at InterXion will have any effect
regarding the abstraction wells for drinking water, which are located in the area.
6/10
3.2.1 Change in hydraulic head
The predicted change in hydraulic head in the primary aquifer, due to the pumping and reinfiltration at the
active ATES-facility, is seen in Figure 7. The drawdown and increase of the hydraulic head, seen in Figure 7,
corresponds to the situation at the end of a winter-period after running the ATES-facility for 10 years.
It is seen, that the change in hydraulic head, caused by the ATES facility, is less than 0.25 m outside a radius
of around 200 meter from the warm and the cold well.
Figure 7. Lowering and increasing of the hydraulic head during the active ATES-facility.
The drawdown caused by the active ATES system does thus not interfere with the existing drinking water
wells.
The injected heated and cooled water can potentially effect the temperature of the drinking water abstracted
from the abstraction wells.
The predicted temperature in the aquifer, at the top of the limestone, after 10 years of operation, is shown in
Figure 8. The figure to the left shows the situation at the end of the summer period while the figure to the right
shows the situation at the end of the winter period. It can be seen, that water from both the cold and the warm
storage is drawn southwards due to the general groundwater gradient in the area. However, the maximum
7/10
extent of a 0.5 C change after 10 years of operation is around 200 m in a southern direction and thus not a
problem for existing drinking water wells.
Figure 8. The spatially distribution of the temperature after 10 years of operation. At the left, it corresponds the end of summer. At the right,
it corresponds to the end of winter.
As described, a number of point source pollutions near the proposed ATES system may be influenced by the
operation of the ATES wells. Especially the pollution at Brydehusvej 21 is of concern.
3.3.1 Brydehusvej 21
As mentioned, the chlorinated solvents at Brydehusvej 21 have not yet reached the primary aquifer, and are
only found in the fill and clay layers. Remediation is on-going at the site and occurrence of the chlorinated
solvent in the primary aquifer is not expected.
As seen in Figure 7, the active ATES-facility will cause some changes in the hydraulic head in the primary
aquifer. The hydraulic head will increase around the injection well and decrease around the abstraction well.
In the area where the hydraulic head in the primary aquifer is lowered, leakage from the fill and clay layers to
the primary aquifer could increase, but it is expected that the existing remediation facility, with its pumping in
the upper aquifer, will counteract the formation of a downward gradient. Thus, increased leakage to the primary
aquifer is not expected.
However, to estimate the transport of chlorinated solvents, that potentially might reach the primary aquifer, the
model was used to produce particle pathlines, where the particles were released in the upper part of the
primary aquifer. Two sets of path lines were produced, one with background conditions for 10 years (No ATES
system), and one with the ATES system implemented for 10 years, see Figure 9.
8/10
Figure 9. 10 year particle pathlines without (left) and with (right) an active ATES-facility. The particles are originating form the site at Bryde-
husvej 21.
It can be seen that well with ID DGU.nr. 200.6986, is abstracting all particles originating from the site Bryde-
husvej 21, with an active ATES-facility.
The additional V1 and V2-mapped locations, and the active ATES systems effect on them, has also been
analyzed. Again particle pathlines has been produced for a 10 year period. The results are seen in Figure 10.
Figure 10. The remaining point-source pollution without (left) and with (right) an existing ATES-facility.
It can be seen that the ATES system will have an significant effect on the pollutions located closest the two
wells, where many of the pollutants will enter the ATES system
4 Conclusion
The model runs show that the operation of an ATES system at the InterXion server park for ten years will most
likely not affect the existing drinking water wells. Neither with hydraulic head changes nor temperature
9/10
changes. However, the model runs show that if existing pollutions in the vicinity do reach the primary aquifer,
they will be drawn in by the ATES system.
Therefore, an arrangement has been made with the environmental authorities responsible for the clean-up of
these pollutions, that if sampling shows increasing levels of pollutants in the pumped water, a water treatment
plant will be set up in connection with the ATES facility and water pumped up will be cleaned before reinfiltra-
tion.
ATES was thus found to be a very suitable solution for the cooling of the InterXion server park in spite of both
nearby drinking water wells and pollutions.
10/10
3.5 - Ecosystems services and combined approaches, Managing multiple functions of the subsurface
ECOSYSTEM SERVICES OF THE GROUNDWATER AND THE SUBSURFACE; FILLING THE
KNOWLEDGE GAP
Johannes P.A. Lijzen1, Sophie Vermooten2, Suzanne van der Meulen2, Michiel Rutgers1, Hans Peter
Broers3
1. National Institute for Public Health and the Environment RIVM, P.O. Box 1, 3720 BA Bilthoven, The
Netherlands, Phone: +31 30 2749111, E-mail: johannes.lijzen@rivm.nl; michiel.rutgers@rivm.n
2. Deltares, P.O. Box 85467, 3508 AL Utrecht, The Netherlands, Phone +31 (0)88 335 8273. E-mail
sophie.vermooten@deltares.nl; suzanne.vandermeulen@deltares.nl
3. TNO Geological Survey of the Netherlands, P.O. Box 80015, 3508 TA Utrecht, The Netherlands, Phone
+3188 866 4798; E-mail: hanspeter.broers@tno.nl
1. Introduction
2. Goal and framework
3. Descriptions of ecosystem services
4. Relevant human activities and factsheets
5. Technical decision support Framework
6. Example of elements in the framework
7. Practical application of the framework in current policy developments
1. Introduction
In densely populated areas, the use of groundwater and the subsurface for functions such as groundwater
extraction, aquifer thermal energy storage and infrastructure is increasing. This results in a need for
subsurface spatial planning and careful consideration of the use of groundwater for several (economic)
activities. Since 2013, in the Netherlands special attention is given to ecosystem services (ES) of the
groundwater and subsurface in order to assess how human activities make use of ES and influence ES. A
technical assessment framework for the sustainable use of the groundwater and subsurface was
developed based on the ES concept. This study was commissioned by the Dutch Ministry of Infrastructure
and the Environment. Beside the fact that the broadly available knowledge of the subsurface system was
lagging behind compared to the top soil, it was also seen as a problem that information on groundwater
and subsurface was not readily available to support decision makers in local management and spatial
planning. In this paper, the goals of the framework are described (2), followed by the identified eleven ES
(3) and the thirty-one human activities influencing the ES (4) and the relations between both (5). An
example of elements in the framework is presented (6) and the practical application of ES in policy is
described (7).
The defined ES for groundwater and the subsurface fit in these definitions. Depending on International
developments and the goal of an assessment can be a reason to make changes in the future. Depending
on the goal also modifications can be needed.
Figure 1 Main components of the Natural Capital according to the European Environmental Agency (EEA):
the ecological capital (all ES and ecological assets), and the abiotic capital (renewable and non-renewable
abiotic resources). The groundwater system has elements in all components of this scheme (source:
Maes et al., 2013).
Eleven ES were defined for the groundwater and subsurface compartment (after Broers and Lijzen, 2014,
see Table 1):
Provisioning services
1. Availability of sufficient water with specific quality. Due to the fact that undisturbed groundwater has a
constant temperature and composition, it is a reliable resource. This ES demands a chemical, physical
and biological quality that meets the criteria of the different uses that are mentioned (agriculture, industry,
domestic).
2. Energy content. At several kilometers depth hot and salt groundwater is present. Ata depth of 100-300
meter the temperature of the groundwater equals the yearly average of 10-12C). In more shallow layers
there is more fluctuation in the temperature. From all these layers thermal energy can be withdrawn or
stored for later use.
Regulating services
3. Attenuation and purification capacity of the subsurface. This ES contributes to the (partial) removal of
chemical or biological contamination from the groundwater. Different natural physical, chemical or
biological processes can play a role (physical filtration, adsorption, degradation etc.).
4. bearing capacity of the soil (for buildings and infrastructure). It is defined as the extent to which the soil
can support objects (like machines, buildings, roads, people, trees).. Besides the natural bearing capacity,
extra bearing capacity can be realised for example by piling on deeper sand layers.
5. Storage capacity This ES includes the opportunity to use the subsurface to (temporarily) store rainwater
and surface water, but also resources like CO2, oil, gas, heat and cold, and waste.
6. Bio-geochemical cycles (material and water cycles). This concerns the cycles of carbon, nitrogen,
phosphorus and other elements. The groundwater flow contributes to the transport of these elements
, besides transport and processes in air, vegetation, oceans etc. The subsurface system is also
temporarily- sink of compounds.
7. Temperature regulation. This ES originates from the constant temperature and high heat capacity of
the groundwater and subsurface. Due to evaporation of water by vegetation the temperature of the soil
surface can also be lowered.
8. Providing surface water baseflow and surface water quality. Groundwater contributes to the surface
water baseflow in streams and rivers and therefore supports aquatic systems and the navigability of rivers
for shipping. The baseflow of groundwater has an important mitigating and buffering function in periods of
drought, as streams will not dry out and remain flowing with water of sufficient quality. 9. Upward seepage
to groundwater dependent nature reserves. This ES is defined as the seepage of sufficient groundwater of
good quality so that dependant natural ecosystems can function normally. In the higher grounds and the
lower parts in the Netherlands the processes are different.
Cultural services
10. Culturalhistorical and experience values. This ES concerns value of the subsurface from the cultural
historical perspective, for examplethe maintenance of knowledge and objects of human activities in the
past. The experience values concern the relation of people with the groundwater and subsurface itself.
11. Biodiversity and habitat. This ES can be seen as the habitat for an unique collection of organisms that
support essential processes. The biodiversity of the subsurface is a multidimensional concept in which
size, diversity and stability (against changes) are all important aspects.
These ES were described extensively in 2014 in factsheets per ecosystem service,the main question
being what conditions would allow for a sustainable functioning of the ES. The factsheets are giving
answers to the following questions;:
- How do we define the ES and what processes determine the performance of the ES?
- How is the ES used?
- Is the ES positively or negatively influenced by an anthropogenic activity?
- What measures can we take to optimize the ES?
- What future trends (supply and demand) have influence on the ES?
- What data is available to measure, quantify and map the ES?
Table 1 Ecosystem services defined for the subsurface related to CICES (2013)
Type of ecosystem Ecosystem service of the groundwater Example of economic activity using ES
service and subsurface
Provisioning 1. Availability of sufficient water with Abstraction for drinking water supply
services specific quality;
2. Energetic content Thermal energy production
Regulating services 3. Attenuation capacity of the subsurface In-situ remediation of historical
contamination
4. Soil bearing capacity Water level management for construction
purposes
5. Storage capacity Aquifer Storage and recovery (ASR)
6. Bio-geochemical cycles (material and Buffering excess CO2 in groundwater
water cycles) resources
7. Temperature regulation Aquifer Thermal Energy Storage
8. Providing surface water baseflow and Recreational fishing and bathing in natural
surface water quality streams
9. Upward seepage to groundwater Recreation in wetlands
dependent nature reserves
Cultural services 10.cultural-historical values Recreation related to subsurface
conservation of cultural heritage (springs)
11. Biodiversity and habitat Recreation related to biodiversity (e.g. bird
watching)
A central element in the framework is how we can assess and weigh different human activities using the
groundwater and subsurface (management, planning, trade-offs) and make optimal use of the ES. This is
about win-win situations and trade-offs between human activities and about the impacts of human
activities and values of the subsurface. For this analysis, inn a workshop we distinguished 31 activities
within 7 main categories. These activities rely on ES or can have substantial impact on these ES. This list
is intended to cover most relevant activities in the Netherlands and can be extended or reduced when
applying the method to a different country or region. Table 2 shows these activities.
Abstraction of water. Different types of use were distinguished (irrigation, drinking water, industrial water).
The requirements for the quantity and quality of the water and the periods of abstraction show large
differences between these types of use.
Storage of water and other compounds. Storage means that water or other substances temporarily are
stored and might be used later.
Reservations. Mostly these reservations are implicitly made, but are not being formalized.
Extraction of other resources than water. These extractions can have a high impact on the groundwater
system and subsurface and are often irreversible.
Subsurface spatial occupation. This category contains activities that have a spatial claim in the subsurface
soil, without storing or extracting elements as described in the activities 7 to 14. The spatial claim for the
management of groundwater contamination for example consists of the size of the contaminant plume for
which management or monitoring is needed.
Water level management. Management of surface water in the Netherlands has a large impact on the
groundwater flow and load bearing capacity of the subsurface and soil. It therefore strongly affects the ES
of the groundwater and subsurfacein deep polders in the western of the Netherlands. Water level
management includes drainage of agricultural land and dewatering of polders as well as the inlet of
surface water from outside the area that might infiltrate in de groundwater.
Aboveground activities. Aboveground activities can influence the ES of the groundwater and subsurface
andcan affect other activities we want to perform in the subsurface. An example is the leaching of manure
and pesticides that can influence the deep groundwater, leading to problems with the quality of the
abstracted drinking water. Five aboveground activities were identified.
For nine of the 31 activities defined, detailed factsheets were developed (underlined in table 2). This set of
factsheets summarizes the available knowledge about the mentioned relations and forms a preliminary
guidance for local and regional authorities for decision-making.
Table 2. Human activities in groundwater and subsurface. The underlined human activities were described
extensively in factsheets.
Category Human activities
1. Abstraction of 1. Public and private drinking water abstraction from groundwater
groundwater 2. Irrigation of agricultural crops
3. Process water for food industry
4. Cooling water
5. Building, e.g. dewatering.
6. Management of groundwater remediation.
2. Storage of water 7. Storage of rainwater for process water
and other substances 8. Aquifer Thermal Energy Storage (ATES)
9. Storage of sediment soil or waste in former sand pits
10. Subsurface storage of radioactive waste
11.Subsurface CO2-storage
12. Abstraction of brackish groundwater combined with injection of
brine for production of irrigation water
13. Artificial recharge for drinking water production from surface water
(ASR, MAR).
14. Natural Water Retention Measures; temporal storage to guarantee
summer baseflow and prevent flooding
3. Reservation 15.Preservation of biodiversity and habitat of the subsurface
16. Preservation of cultural historical land archaeological values.
17. Reservation of strategic groundwater resources
4. Extraction of 18. Extraction of gravel, sand and clay
resources (other than 19. Extraction of salt
water) 20. Extraction of shale gas (including groundwater use)
21. Oil and gas extraction
22. Geothermal energy (deep)
5. Subsurface spatial 23.Subsurface infrastructure and buildings
occupation 24. Management of groundwater contamination, including on-site
remediation
6. Groundwater level 25. Groundwater level management of polder systems in lower
management Netherlands
26. Groundwater level management of freely draining areas in the
higher Netherlands
7. Aboveground 27. Nature conservation measures in terrestrial or aquatic ecosystems
activities 28. Recreation (fishing, nature, watersports)
29. Soil sealing
30. Application of manure and pesticides
31. Diffuse inputs to soil in rural areas
5 Technical decision support Framework
The framework is based on three cross tables that give insight into the relation between human activities
and ES. Table 3 (first cross) shows whether an ES is used by a human activity. The term using here
means that without the existence of the ES the activity would not be possible.
9.Upwardseepagetogroundwaterdependentnaturereserves
8.Providingsurfacewaterbaseflowandsurfacewaterquality
1.Availabilityofsufficientwaterwithspecificquality;
6.Biogeochemicalcycles(materialandwatercycles)
3.Attenuationcapacityofthesubsoil
10.culturalhistoricalvalues
11.Biodiversityandhabitat
7.Temperatureregulation
4.Soilbearingcapacity
2.Energeticcontent
5.Storagecapacity
Anthropogenicactivity specificactivity
Abstractionofgroundwater 1.Drinkingwaterabstractionfromgroundwater J N J N N J J N N N N
2.Irrigationwithgroundwaterforagriculturalcrops J N J N N N N N N N N
3.Processwaterforfoodindustry J N J N N J J N N N N
4.Coolingwater J J N N N J J N N N N
5.Building,e.g.dewatering. N N N J N N N N N N N
6.Abstractionsformanagementofgroundwaterremediation.1. N N N N N N N N N N N
Storage 7.Storageofrainwaterforprocesswater N N N N J N N N N N N
8.Aquiferthermalenergystorage(ATES) N J N N J N J N N N N
9.Storageofsedimentsoilorwasteinformersandpits N N J N J N N N N N N
10.Subsurfacestorageofradioactivewaste N N N N J N J N N N N
11.SubsoilCO2storage N N N N J N J N N N N
12.Abstractionofsaltwatercombinedwithinjectionofbrineforproduc J N N N J N N N N N N
13.Artificialinfiltrationfordrinkingwaterproductionfromsurfacewate J N J N J N J J N N N
14.NaturalWaterRetentionMeasures;
Reservations 15.Preservationofbiodiversityandhabitatofthesubsoil N N N N N J J J J N J
16.Preservationofculturalhistoricallandarchaeologicalvalues. N N N N N J J J N J N
17.Reservationofstrategicgroundwaterresources J N J N J J J N N N N
Extractionofresources 18.Extractionofgravel,sandandclay N N N J N N N N N N N
19.Extractionofsalt N N N N N N N N N N N
20.Shalegasextraction(includinggroundwateruse) J N N N N N N N N N N
21.Oilandgasextraction N N N N N N N N N N N
22.Geothermalenergy(deep) J J N N N N J N N N N
Spatialoccupation 23.Subsoilinfrastructureandbuildings N N N J N N N N N N N
24.Managementofgroundwatercontamination,includingonsitereme N N J N N J N J N N N
Groundwaterlevelmanagement 25.GroundwaterlevelmanagementoflowerNetherlands N N N N J N N J N N N
26.GroundwaterlevelmanagementofhigherNetherlands J N N N J N N N N N N
Activitiesonthesoilsurface 27.Natureconservationmeasuresinterrestrialoraquaticecosystems J N J N J J J J J N J
28.Recreation(fishing,nature,watersports) J N N N N N J J J J J
29.Soilsealing N N N J N N N N N N N
30.Applicationofmanureandpesticides N N J J N J N N N N N
31.Diffuseinputsinsoilinruralareas N N J N N J N N N N N
Legenda J activityusestheecosystemservice
N activitydoesnotusetheecosystemservice
? unclearifactivityusesecosystemservice
Table 4 (second cross table) shows how the ES are influenced by the activities in the subsurface and
aboveground using 5 different scores. The influence can be positive, negative, positive as well as
negative, neutral of unknown. It was scored negative when part of the ES is taken and is not available for
future use or if the value of the ES is decreased. When for example an Aquifer Thermal Energy Storage
claims capacity in the subsurface, this cannot be used by other human activities. The influence of an
activity can be positive for maintaining the ES, or it can be negative. Positive and negative impact also can
occur both, depending on the time scale or approach.
Table 4 Influence of human activities on ES
Ecosystemserviceofthesubsurfaceandgroundwater
9.Upwardseepagetogroundwaterdependentnaturereserves
8.Providingsurfacewaterbaseflowandsurfacewaterquality
6.Biogeochemicalcycles(materialandwatercycles)
1.Availabilityofsufficientwaterwithspecificquality;
3.Attenuationcapacityofthesubsoil
10.culturalhistoricalvalues
11.Biodiversityandhabitat
7.Temperatureregulation
4.Soilbearingcapacity
2.Energeticcontent
5.Storagecapacity
Anthropogenicactivities Specificactivity
Abstractions 1.Drinkingwaterabstractionfromgroundwater o o o o o o o
2.Irrigationwithgroundwaterforagriculturalcrops o o o +/ o
3.Processwaterforfoodindustry o o o o o o o
4.Coolingwater o o o o +/ o o
5.Building,e.g.dewatering. o o +/ o o o
6.Abstractionsformanagementofgroundwaterremediation.1. o o o o o o +/
Storage 7.Storageofrainwaterforprocesswater o o o o o + o o o
8.Aquiferthermalenergystorage(ATES) +/ ? o o +/ o o o ?
9.Storageofsedimentsoilorwasteinformersandpits o o o o o o o o
10.Subsurfacestorageofradioactivewaste o o o o o o o o
11.SubsoilCO2storage ? o o o o o o o o o
12.Abstractionofsaltwatercombinedwithinjectionofbrineforproduction o o o o o o o o
13.Artificialinfiltrationfordrinkingwaterproductionfromsurfacewater. +/ o o o o + o o o
Preservations 14.Preservationofbiodiversityandhabitatofthesubsoil o o + o + o o o + +
15.Preservationofculturalhistoricallandarchaeologicalvalues. o o o o o o o o + o
16.Preservationofstrategicgroundwaterresources + o o o o o + + o +
Extractions 17.Extractionofgravel,sandandclay o o +/ o o o o
18.lignintemining(Germantype) +/ o o
19.Extractionofsalt o o o +/ o o o o o o
20.Shalegasextraction(includinggroundwateruse) ? o o o o o
21.Oilandgasextraction o +/ o o o o o o
22.Geothermalenergy(deep) o o o o o o o o o o
Spatialoccupation 23.Subsoilinfrastructureandbuildings o o o o o o o +/ o
24.Managementofgroundwatercontamination,includingonsiteremediati +/ o o o o o +/ o o
waterlevelmanagement 25.GroundwaterlevelmanagementoflowerNetherlands +/ o o +/ +/ +/ +/ o +/ +/ o
26.GroundwaterlevelmanagementofhigherNetherlands +/ o o +/ +/ o +/ +/ +/ +/ o
Abovegroundactivities 27.Natureconservationmeasuresinterrestrialoraquaticecosystems + o o o o + + + + o +
28.Recreation(fishing,nature,watersports) o o o o o o o o o o o
29.Soilsealing o o o o o o o
30.Applicationofmanureandpesticides o o o o o
31.Diffuseinputsinsoilinurbanareas o o o o o
+ positiveeffectofactivityonthemaintainanceorstrenghteningtheESS
negativeinfluenceofactivityonESS(decreasingofvalue)
+/ positiveaswellasnegativeinfluence,dependingontimescaleandpointofview
0 nopositiveand/ornegativeinfluenceofactivity
? unknown,orwhenbehindothersign(?)theeffectisuncertain
Table 5 (third cross table with activities versus activities) visualizes whether human activities are or are not
compatible with each other. Five labels were chosen: activities exclude each other (U), Activities are
known to have significant negative impact on each other (B), activities are likely to have a negative
impact on each other (expert judgement), (W), activities likely do not have a negative impact on each
other (N), or the mutual potential impact is unknown (?).
The resulting cross table gives an overview of activities for which a more extensive assessment is needed
when these are planed near to each other.
Table 5Influencesbetweenhumanactivities
12.Abstractionofsaltwatercombinedwithinjectionofbrineforproductionofirrig
24.Managementofgroundwatercontamination,includingonsiteremediation
13.Artificialinfiltrationfordrinkingwaterproductionfromsurfacewater.
27.Natureconservationmeasuresinterrestrialoraquaticecosystems
6.Abstractionsformanagementofgroundwaterremediation.1.
15.Preservationofculturalhistoricallandarchaeologicalvalues.
26.GroundwaterlevelmanagementofhigherNetherlands
25.GroundwaterlevelmanagementoflowerNetherlands
14.Preservationofbiodiversityandhabitatofthesubsoil
9.Storageofsedimentsoilorwasteinformersandpits
20.Shalegasextraction(includinggroundwateruse)
2.Irrigationwithgroundwaterforagriculturalcrops
16.Reservationofstrategicgroundwaterresources
1.Drinkingwaterabstractionfromgroundwater
28.Recreation(fishing,nature,watersports)
10.Subsurfacestorageofradioactivewaste
7.Storageofrainwaterforprocesswater
30.Applicationofmanureandpesticides
8.Aquiferthermalenergystorage(ATES)
23.Subsoilinfrastructureandbuildings
31.Diffuseinputsinsoilinurbanareas
17.Extractionofgravel,sandandclay
3.Processwaterforfoodindustry
22.Geothermalenergy(deep)
5.Building,e.g.dewatering.
21.Oilandgasextraction
11.SubsoilCO2storage
19.Extractionofsalt
4.Coolingwater
29.Soilsealing
Intendedactivity Specificactivity
Abstractions 1.Drinkingwaterabstractionfromgroundwater W W W N W W B W W W N N N N W W W W N N W W B B B W W B B
2.Irrigationwithgroundwaterforagriculturalcrops W W W N W W W N N N N N N W U N N N N N N W N B B B N B N
3.Processwaterforfoodindustry
4.Coolingwater
5.Building,e.g.dewatering.
6.Abstractionsformanagementofgroundwaterremediation.1.
Storage 7.Storageofrainwaterforprocesswater
8.Aquiferthermalenergystorage(ATES) B W W W N W W N N N W W W N U N N N N N N W N N N N N N N
9.Storageofsedimentsoilorwasteinformersandpits
10.Subsurfacestorageofradioactivewaste B N N N N N N N N W N N N N W N U W W W N N N N N N N N N
11.SubsoilCO2storage
12.AbstractionofsaltwatercombinedwithinjectionofbrineforproduW W W W W W W W W W W W W U W W W W W W W W W W W W N W
13.Artificialinfiltrationfordrinkingwaterproductionfromsurfacewater.
Reservations 14.Preservationofbiodiversityandhabitatofthesubsoil
15.Preservationofculturalhistoricallandarchaeologicalvalues.
16.Reservationofstrategicgroundwaterresources W U W W N W W W N W N N N N N N W N N N W N W N N N W W
Extractionofresources 17.Extractionofgravel,sandandclay
19.Extractionofsalt
20.Shalegasextraction(includinggroundwateruse)
21.Oilandgasextraction
22.Geothermalenergy(deep)
Spatialoccupation 23.Subsoilinfrastructureandbuildings
24.Managementofgroundwatercontamination,includingonsiterem B W W N W N U W N N N N U W W U N N N N N W N N W W N N W
waterlevelmanagement 25.GroundwaterlevelmanagementoflowerNetherlands N N N N N N N W W N N W W N B N W N N N N B W N B B B B B
26.GroundwaterlevelmanagementofhigherNetherlands
Abovegroundactivities 27.Natureconservationmeasuresinterrestrialoraquaticecosystems B B W W W N N N W N N N B N N N W N N N N B W B B N W B N
28.Recreation(fishing,nature,watersports)
29.Soilsealing
30.Applicationofmanureandpesticides B W W N N N N N N N N N B W N B N N N N N N N B B B W N N
31.Diffuseinputsinsoilinurbanareas
Figure2:AuthorizedopenATESsystemsintheNetherlandsin2010.Source:Land+Water11,November2011
Ambition
The city of Utrecht, the fourth largest city of the Netherlands, has the ambition to be a
healthy and attractive place for its citizens, sustainable towards the environment and a
resilient city towards future demands. A positive side effect for Utrecht is that a healthy and
sustainable urban design can
draw new investors towards
areas like former brownfields,
social setback neighbourhoods
and increase value (people &
planet) and profit. Different policy
fields can enhance or hinder
each other when it comes to
reaching this goal. Therefore
Utrecht needs to manage its
natural urban capital by optimal
use of the urban soil water
green energy system. Knowing
various ecosystem services is the
starting point.
Examples
In the opinion of Utrecht, this asks for another operating procedure, in which spatial planning
and the environmental, societal and
economic disciplines work closely
together at an early stage of urban the
(re)development and consider different
scales, both in space as in time. A good
example is the way Utrecht used soil
energy systems (ATES) to remediate
soil pollution. At first it seemed that soil
energy was a threat to the soil quality
(spreading pollution to groundwater) but
it turned out to be a chance to
remediate pollution nd lower CO2
emissions at the same time. Another project focuses on strengthening the green qualities of
an urban area on district level. To realize a sustainable quality of this district, an ecosystem
services approach is being applied. Hence, the emphasis lies on a partnership with residents,
users and stakeholders to analyze the opportunities and threats of the green quality of the
area and to include parameters such as ecosystem services and biodiversity into this
process. With regard to the ecosystem services, important functions such as water storage,
district cooling, carbon storage and cleaning capacity of the subsurface in relation to the
implementation of green projects are examined. By smart combining the features and
functions of an area we can enhance the quality of a district and can ensure that multiple
stakeholders and users can address the benefits. We are actively searching for more smart
solutions and combinations like that.
Conclusions
A positive side effect for Utrecht is that a good and sustainable urban design can draw new
customers towards areas like former
brownfields, social setback neighbourhoods
and increase value (people & planet) and
profit. So, soil quality, sustainability and urban
development are interconnected. This asks for
another operating procedure, in which spatial
planning and the environmental, societal and
economic disciplines work closely together at
an early stage of urban (re)development and
consider different scales, both in space as in
time. But it is also complex matter to deal with.
This asks for another operating procedure, in which spatial planning and the environmental,
societal and economic disciplines work closely together at an early stage of urban
(re)development and considers different scales, both in space as in time.
Utrecht strives towards a healthy urban environment for its inhabitants. To reach this goal an
integrated approach where stakeholders at an early stage are connected is needed. This
starts within the different sectors of the municipality itself. In the past every sector, i.e. soil,
water, transport, energy, spatial planning, worked on its own policies. To reach Utrechts
sustainability goals it is necessary to detect opportunities for sustainability by combining
different policy fields. Public private partnerships within the region contribute also towards
sustainability, as natural capital does not stop at the borders of the municipality. Many
ecosystem services contribute in various ways and also influence each other. When these
ecosystem services are quantified and monitored it gives Utrecht the opportunity to manage
and steer towards a healthy urban environment.
MODELLING URBAN SYSTEMS USING THE FP7 HOMBRE BR2 TOOL
An urban area, with its interacting network components of population, industry, commerce, transport,
utilities and green and blue infrastructure may be regarded as a complex system that may be studied
at a range of scales: from the neighbourhood scale, through the city, regional up to national or
international scale. Interaction Matrix systems techniques have been applied to a variety of complex
systems. Binary or semi quantitative techniques may be especially useful in cases where precise
mathematical descriptions of component relationships are not available to describe the relationships
between components.
The ReMIT/ReSPONSE (or RES) Interaction Matrix technique was proposed and developed by
Imperial Colleges Professor John Hudson in the 1990s as a response to the increasingly complex
problems presented in rock engineering, attempting to move away from traditional models which dealt
in detail with specific individual aspects of a problem without addressing the system as a whole so
producing a narrow, disjointed view. The role of this approach in urban land management was
discussed by Paul Nathanail as part of the work of the CABERNET brownfield regeneration network.
Tony Leneys 2008 University of Nottingham PhD thesis developed the Brownfield ReMIT-
ReSPONSE (BR2) technique to assess the impact of a brownfield redevelopment on the surrounding
urban by considering how a dynamic and interactive urban system may change once perturbed by the
new use for the site.
The work under the FP7 HOMBRE project developed BR2 to produce a user friendly spreadsheet
based tool to guide the user through the steps to be taken and the decisions to be made when
applying the technique to a brownfield site and ultimately collates responses and presents a range of
outputs for analysis. The tool was tested on a real-world redevelopment site: The Markham Vale
Employment Zone in the East midlands of the UK in order to assess two options considered during
the early planning stages of the redevelopment the site.
Relationships between components are coded in a clockwise manner. First horizontally, the cause
which describes the influence of the first component over the second, vertical component (effect).
Urban system components may include aspects of policy, regulation, finance (public, private,
individual), population, built environment or natural environment depending on the context site in
question. However each component may be considered a system in itself, and each system a
component of a larger system, meaning that setting the boundaries defining each component and the
system is subject to the study objectives.
Once relationships have been coded, scores in each row and column are summed to give total Cause
(Row) and Effect (Column) scores for each component (Error! Reference source not found.). These
responses are then analysed using a C:E diagram of Cause versus Effect which offers information
on how each component influences, and is influenced by the system in general.
Due to the clockwise coding method, the sum of each row of the matrix represents that components
effect on the system, termed CAUSE, whereas the sum of each column represents the systems
effect on the component, termed EFFECT. These component values are then plotted to give the C:E
diagram.
The position of each data point on the plot and the constellation of points as a whole can be
analysed to give information about the system. In general, the further along the C=E line a component
lies, the more interactive it is i.e. the more connections to other components in the system, in this
case Pi is the most interactive component and Pn the least interactive. The mean interactivity can be
used to make general observations on the whole system and compare the overall interactivity of
different scenarios. The Interactive Intensity (C+E) may be used to illustrate and compare the
individual component interactivities, for example by plotting them on a bar chart.
A component plotting beneath (and parallel to) the C=E line indicates that it is dominant, i.e. it
influences the system in general more than the system influences it while components plotting above
the C=E line indicates the component is subordinate, i.e. the system in general influences that
component more than it influences the system.
A BR2 analysis can provide a deeper insight in how a specific urban system behaves and how it will
be changed by proposed changes in land use. Such insight can help inform better decision making
and contribute to sustainable urban land management.
AUTHORS
Department of Civil Building and Environmental Engineering (DICEA), Padova University, Via
Francesco Marzolo 9, 35151 Padova
carlo.astival@gmail.com
+39 049 8275484
Raul Berto
Department of Engineering and Architecture (DIA), Universit di Trieste, Via Alfonso Valerio 6/1,
34127 Trieste
raul.berto@phd.units.it
+39 040 5583481
Giovanni Cechet
Department of Engineering and Architecture (DIA), Universit di Trieste, Via Alfonso Valerio 6/1,
34127 Trieste
giovanni.cechet@gmail.com
+39 040 5583481
KEYWORDS
INTRODUCTION
This paper is the result of a research carried out by a workgroup coordinated by prof. Paolo Rosato
and prof. Edino Valcovich and composed by PhD eng. Raul Berto, eng. Giovanni Cechet and PhD
eng. Carlo Antonio Stival. The Research Fund of the Trieste University (FRA 2012) grants this
research project.
The decision to develop the topic of green roofs is born from the recognition these solutions as an
expression of the sustainability concept. A complete development of this topic is complex, in fact is
more frequently to treat separately the different aspects of sustainability, studying individually a
solution, an idea, a system as an expression of it. The green roofs produce effects in the
environmental, technological, social and economic spheres, so it is possible to analyse each of these
effects in accordance with a comprehensive vision.
The scientific production, that is about the green roofs topic, reveals the excellence of these solutions
according to the various aspects debated. However, based on the idea that in this historical moment is
very important to increase the value of the existing buildings, we realized the need to evaluate the
potential implementation of these kinds of top closures on existing buildings. For this reason, insights
on the technological aspects developed with this main goal.
1. STATE OF ART
The overall performance offered by green roof solutions are still under discussion: the variety of
effects brought from greening the top covers of buildings in different areas of performance, and the
difficult in scientific proof of such effects, make complex the measurement of these effects in order to
demonstrate the environmental, social and economical advantage.
However, in scientific research, is a widespread belief that the green roof technologies really increase
the users well-being, in particular because the compensation between the green surface positioned
over the roofing and the natural ground replaced by buildings is consider how a positive effects.
In recent decades, the need to reduce negative effects generated by ground anthropization and
increasing of energy consumption related to the technical services of the buildings, has led to the
implementation in building envelopes of technical and architectural forms tipically of natural
environments. We can speak, in different scene, of Green Architecture and Green Technology.
With Green Architecture meaning the inclusion of specific planning strategies from urban scale to
building shell of technical elements able to reduce energy consumption from non-renewable energy
sources, emissions of carbon dioxide in atmosphere and environmental impacts in general, the last
ones descend from using of determinates construction materials.
With Green Technology, instead, we consider the mitigation of architectural and aesthetic impact. In
particular, we consider the beneficial effects of some design solutions on technological scale, using,
for example, an actual version of vernacular architecture, turning this one into an integration of build
volume in a context of high natural connotation.
Greenery establishes the path to restore at least in part the natural value of original environmental
system in particular location, allowing camouflage of constructions, as a valid alternative to hypogeum
buildings. The continuity of green surface, compared with natural ground, can create specific habitat
for plants and small animals.
2. METODOLOGY
Evaluating of the potential in relation to place a green roof in a new building or to implement it in an
existing construction, as replacement or as an integration of the previous coverage, this is the matter
of this research work and it is handled in this scientific paper.
In particular, the performances have been analized in terms of contribution to contain energy
consumption, to control local microclimate and regimentation of rainwater. Also evaluating the
possibility to install green roof over existing building from a structural point of view: outlining critical
issues, describing relationship with the Italian legislation concerning construction and, finally,
proposing a critical analysis in relations to the most common building typologies on the Italian territory.
3. RESULTS
As part of an assessment of the energy performance, a solution of green roof is a highly complex
system. It consists, in fact, of heterogeneous functional layers, in which it changes the moisture
content, coherently with the functions carried out by a natural soil, which is the subsistence from the
nutrients, the water storage, the breathability of the vegetable and cultivation layers.
Consequently, the performance of thermal insulation and thermal inertia are highly dependent, over a
specific period, from climatic factors such as intensity of precipitation, solar radiation and outside air
temperature.
To define the energy behaviour of a green roof, basing on the performances of the individual
functional layers in the worst conditions of humidity, leads to assign performances significantly lower
than those inferred from the experiments do.
Referring only to the thermal conductivity values of the materials forming the solutions of green roof, ir
can be disclosed how the performance levels established by the regulations for top closures, in terms
of thermal resistance, are reachable with a culture layer thickness above 100 cm. These conditions
considerably reduce the chances to implement a green roof, especially in existing buildings, because
we are in presence of high static load.
MATERIALS thermal Reference Minimum
conductivity thermal resistance thickness
[W / mk] [m2 k / w] [cm]
Expanded cork panels 0,040 0,045 3,00 12 14
Expanded polystyrene panels 0,036 0,040 10 12
Rock wool panels 0,034 0,038 10 12
Fir wood (1) 0,100 0,120 30 35
Cultivation layer of green roofs (2) 0,400 0,900 120 300
It assumes an average moisture content in the wood equal to 15%; each percentage point more of
(1)
moisture increases the thermal conductivity of a value equal to 1.2% from the base value.
(2) It assumes a degree of saturation of the soil equal to 50
Table 1 - Thermal insulation properties comparison of some materials related to a generic cultivation
layer of a green roof.
From above, it results wrong to refer only to the parameter of thermal conductivity of the cultivation
layer to define the properties of heat transmission of a green roof in the winter season. It is also
necessary to underline that, unlike roof traditional technological solutions, the calculation of energy
performance indicators in stationary regime is approximate, as the functional layers of a green roof
varying considerably its physical technical characteristics during the year, especially in terms of
moisture content.
A green roof, in substance, is a complex system not describable through schematization or simplified
models, in particular because it is a system capable of self-regulation as a function of external climate
conditions. We can assume the combination of plant essences, cultivation layer, insulation and
structural support as a layer of non-homogeneous and non-isotropic material, which operate as a
thermal mass for evaporative cooling according physic - technical variables (Moody et al., 2013).
In a formulation of technological performances, a green roof consists in a complex system of
functional layers, in which the significant effects from the point of view of energy performances are
explained, essentially, by the following:
- vegetal layer;
- cultivation layer;
- thermal isolation layer, warm roof solutions;
- structural layer.
The mutual interaction of these layers defines the complex mechanisms of heat exchange into the
different levels of a green roof: conductive, convective, radiative and latent. The methods of heat
exchange are otherwise attributable to functional layers mentioned above, influencing the design
parameters needed to describe the energy performances of a green roof.
Table 2 Relationships between functional layers responsible of heat transfer in a green roof,
different type of heat transfer and relative significant parameters.
The design of the vegetal layer, as the principal final value from the viewpoint of architectural and
aesthetic of a green roof, also considers the different types of heat exchange in relations with these:
- radiative heat exchanges, related to the control of solar radiation coming from the outside
(depending on the vegetal essence planted) and the absorption within the layer itself;
- convective exchanges between the vegetal layer, the soil and the external environment
through air interposed;
- heat transfer for evapotranspiration of the leaves and the vegetation at ground level,
depending on the parameter of stomatal resistance.
The main parameter that influences the performances offered by a vegetal layer is the Leaf Area Index
(LAI), the ratio of the leaf area projected onto the surface and the surface itself expressed in [m2 / m2].
This parameter is an excellent synthesis to measure the effect of the vegetal layer on the overall
performance of the green roof (Arlunno, 2011), specifically on evapotranspiration exchange and solar
radiation control. The values of the LAI for green roofs are usually in the range 0.5 to 5.0. The
percentage of coverage area protected by the leaves is a parameter directly related to, but not
identical with the LAI (Sailor, 2008), and it affects the radiative heat flows on the surface of vegetal
layer.
Even the vertical dimension of the vegetal layer influences the heat flow crossing the green roof
throughout the year; in particular, in the winter season, the presence of the vegetal layer reduces the
dispersion of heat for extra radiative flow towards the celestial vault (Palomo del Barrio, 1998).
The main energy effects related to the vegetal layer are inherent to the performance of roofing in
summer, as they relate to the thermal inertia guaranteed by the technical element. It has been
measured that the proportion of solar radiation reflected by the vegetal layer deviates significantly from
that offered by a traditional roofing: the values of albedo are respectively equal to 0.23 and 0.10, with
a consequent reduction of the quota transmitted to the massive layers of roof (Lazzarini et al., 2005).
The climatic conditions, in this case, play a decisive role in the entity of the inertial performances: the
average porosity of cultivation and the variation of the voids saturation degree increase the thermal
conductivity of the layer, because the water is a better thermal conductor than air, and reduce its
thermal diffusivity. For the last reason it is possible to verify a reduction into the heat flow through the
roofing in order to increase the volumetric moisture content (Palomo del Barrio, 1998).
For these reasons, to assume constant physical - techniques characteristics of the cultivation layer is
incorrect: in calculations of stationary regime (peculiar of the winter season), for small thickness layers
(less than 6 8 cm) it is acceptable even neglecting the thermal resistance (Arlunno, 2011).
It is possible to derive empirical formulas to express the thermal conductivity [Wm-2K-1] and
specific heating cp of a cultivation layer [kJkg-1K-1] in function of the volumetric content of moisture
in the same, expressed in [m3m-3], according to a linear dependence (Moody et al., 2013):
= 1,13 + 0,10 [ ] = 4,75 + 0,76 [ ]
These empirical formulas are obtained from seasonal average values derived from measurements
of a specimen of ground, initially dried in a furnace, to which subsequently controlled amounts of water
have been added.
Finally, the greater incidence parameters to establish the summer performances of a green roof are
the LAI and the average thickness of cultivation layer. Also specific parameters of the dry soil, such as
density, thermal conductivity, the moisture content and the specific heat affect significantly.
WINTER SUMMER
vegetal reducing convective heat transfer reduction of the solar radiation incident on
the roofing produced by shading (LAI with
increase of transpiration phenomena
3)
with consequent increase of heat flows
heat dispersion for evapotranspiration
phenomena
cultivation increase of thermal resistance heat flow reduction during the input than
compared to non-insulated roofing non-insulated roofing
thermal insulation does not exceed the reduction of the temperature at the base
isolated shells of the layer respect to a traditional roofing
thermal conductivity variable as a need of higher water content to dissipate
function of moisture content latent energy
negligible contribution for thicknesses
and significant only for intensive green
roofs
thermo insulated necessary to obtain adequate thermal significant performance depending on the
resistance values material that constitutes it
structural variable performance as a function of effects of thermal inertia as a function of
the material constituent surface mass
Table 3 Summary statement about the effects of green roofs typical functional layers in the
energy performance determination, for summer as for winter, in mediterranean climate.
About the microclimate mitigation effects in the urbanized contexts, the roof greening are configured
as technological systems able to ensure the continuity of complex ecological system. This is
constituted by flows of energy, materials and information related to the man-made and natural
component, even with significant changes to the scale of individual buildings as to the larger urban
settlement.
In particular, cities are an energy-intensive system that leads to some negative effects on the natural
component, such as the increase of the temperature of the urban foliage, waterproofing of surfaces
and reducing the ecosystem biodiversity.
The different morphological characteristics of urban soils (slope class, roughness, surface color), the
mutated heat capacity of an anthropogenic object related to pre-existing conditions of natural soil,
pollution agents emitted into the atmosphere by human activities (primarily traffic, conditioning and
ventilation systems, industrial installations, etc.) result as alterations of radiative and mass fluxes that
characterise an intact atmosphere.
In the proximity of a large urban settlement, you may experience a discontinuity in the atmospheric
layer in direct contact with the Earth surface defined Urban Boundary Layer: in its interior, the layer
enveloping the built volumes is defined Urban Canopy layer. This phenomenon also identified by the
Planetary Boundary Layer, that is the vertical portion of the atmosphere in which activities are
concentrated and the resulting changes induced by them. These features of this air layer depend on
the energy shares of solar radiation reflected and absorbed by surfaces in it included, from the shape
of urban spaces in relation to accessibility solar and widespread human activities.
The anthropogenic surfaces tend to reduce the component of solar radiation reflected, the albedo,
condition that involves a greater share of energy absorption. The albedo of a vegetal surface, usually
equal to 0,20 0,30, is lower than the albedo of some surface materials amply diffused in contexts of
high-density population and employed for the realization of roofing finishes. Among these materials we
can identify, for example, the asphalt (albedo: 0.04 0.10), pebbles and crushed stones (albedo: 0.10
0.15), tiles and brick tiles (albedo: 0.10 0.30), metal sheet (albedo: 0.10 0, 15).
Comparing the heat balance of an urban location with heat balance of a rural location, both similar in
size and geographic position, there is a drastic reduction of the heat exchanges of latent heat through
evapotranspiration and an increase of the energy stored in the building (Cleugh et al., 1986).
Considering the plant in its entirety, the interaction among leaves in terms of geometry and orientation,
makes the definition of the energy balance a great complexity operation. It is however possible to
affirm that the global albedo of vegetation decreases with increasing height of its development and
with the apparent height of the sun in the sky.
If the presence of green areas in urban context, also placed on the roof of a building, allows improving
the energy balance, the contribution is not exhaustive, in fact it should be integrated with the larger
scales, also with an optimized landscape design.
The performance more efficacious due to the green roofs is the control of rainwater downflow, in
particular:
- superficial flow control;
- flow rate reduction of rainwater conduct toward disposal system;
- hydrograph peak wave reduction, that is maximum instant water flow over roofing;
- delay on the confluence of peak wave to the disposal system (Ciaponi et al., 2014).
The water systems management of a green roof is allowed by processes of water storage that
regulate the dispersion in the atmosphere, by the evapotranspiration phenomena for a quota that can
exceed 60% of the volume precipitate (Bengtsson, 2005). The drainage system captures a portion of
the remaining meteoric water volume, this quota depends on the inclination of the roofing itself:
- the effects of detention are measured by determining the ratio between the hydrograph peak
output from a waterproof roofing and the output from the green roof in exam;
- the effects of retention are related to the volume of water held.
Essential parameter for the quantification of these effects is the outflow coefficient , defined as the
ratio between the volume of water going out from the roofing and the volume of water gravitating on it,
in the same time interval. Green cover, in comparison to a traditional waterproofed roofing, presents
an outflow coefficient significantly lower: related to an impermeable surface, for which = 0,95 1,00,
an extensive green, characterized by a layer of a thickness of culture to 10 cm, it reduces the outflow
coefficient to the value = 0.40 0.45. Cultural layers of 40 to 50 cm, adapted to an intensive
greening, show a value = 0.10 0.15 that is closed to the one of the fallow soil.
This performance is of great interest when is compared to the analogue features in a traditional high-
energy performance roofing. Therefore, the flow coefficient is then the most effective parameter to
highlight the contribution of green roofs with the appropriate demands to the environmental protection
class of requirements, even if for arrangement of any public incentive mechanisms.
The green roofs, if they are used on a larger scale of the single building, or rather to neighborhood
scale or more properly of scale to drainage basin, represent an effective strategy for the control of
atmospheric precipitation, even if of considerable intensity. Green roofs prevent possible phenomena
of flooding and restoring the hydrological cycle in contexts with high-density building: this fact
underlines the potential of a diffused conversion to green roofs in an urban area, in particular if this
process is disciplinated by town planning.
The partial transformation of a cover by greening, subordinated to the structural layer residual bearing
capacity, consists in the replacement of the original functional layers with new layers that lead to a
greater reduction in energy consumption and increase in the environmental performance of the
building shell.
The structural performances resulting from the behavior of the technical element of coverage in
existing buildings are referred to two different scales:
Local: it requires that the coverage bearing structure is capable to withstand the stresses
arising from the overload, induced by the functional layers;
Global: it requires new performances verification of the overall structural subsystem.
Operating on existing buildings, to determine the extent of overloading results an applicable action
essential in the redaction of design specifications, which can lead to reduce the ranks of technological
solutions applicable or to exclude its implementation, unless significant interventions of structural
reinforcement and structural stiffening. In existent structural system, extensive green roofs typology is
previleged, because it has a smaller thickness and, consequentially, induces a minor overload. In new
building, the cost of realization of green roof is the binding condition.
From the analyzes carried out, we can say that the need to proceed with the evaluation of structure
safety of an existing building, as intervention object for the realization of green roofs, and its possible
structural adjustment, strongly depend on the construction typology of building itself.
For buildings with a large number of stories and a reduced floor area, in fact, the need to evaluate the
safety and the adaption of buildings is less likely than for the ones having a limited number of plans
and a big floor area. Also the type of structure affects in this sense: in case of heavy structures, the
increase of loads on roofings will be less incident than the whole of the loads acting on the foundation;
it is the opposite for buildings with lightweight structures.
Table 4: Probability of needing to perform an assessment of the sutructural safety for the realization of
a green roof in function of the building typology.
4. FUTURE PROSPECTS
The research work summarized in this paper tries to address some of the main issues that may
concern the green roofs also in order to lay the foundations for future works, which will examine more
in depth the matters discussed.
The need to debate what results in this work by the assessments over one or plus existings buildings
choosed as case of study, rapresents an aim that we want to pursue. Moreover, it results really to
treath also the topic of vertical green, intended as element of vertical opaque enclousure, that can be
integrated into building shell in order to complete green roofing solutions.
The research of a greater continuity between natural environment and antropized environment is a
fascinating challenge for all stakeholders of architecture ad technical discipline, from a point of view of
archievment of environmental preservation objectives.
Definitively, we consider that there a lot of possibilities in the research about this topic and that the
diffusion of this technical solutions of roofing represents the key into the necessary process of building
and urban requalification, one of the greatest challenge of our times.
5. REFERENCES
Arlunno R. (2011), Ambiente ed energia, In: Fiori M., edited by, Coperture a Verde, Milano, Hoepli
Editore, p. 89
Bengtsson L. (2005), Peak flows from thin sedum-moss roof, in Nordic Hydrology, Vol. 36, pp.
269-280
Ciaponi C., Papiri S., Sanfilippo U., Todeschini S. edited by (2014), Acque di prima pioggia nei
sistemi di fognatura, Milano, Hoepli Editore
Cleugh H. A., Oke T. R. (1986), Suburban rural energy balance comparison in summer for
Vancouver, B. C., in Building Layer Meteorology Vol. 36(4), pp. 351-369
Lazzarin R. M., Castellotti F., Busato F. (2005), Experimental measurements and numerical
modelling of a green roof, in Energy and Buildings, Vol. 37(12), pp. 1260-1267
Moody S. S., Sailor D. J. (2013), Development and application of a building energy performance
metric for green roof systems, in Energy and buildings, Vol. 60, pp. 262-269
Palomo del Barrio E. (1998), Analysis of the green roofs cooling potential in buildings, in Energy
and buildings, Vol. 27(2), pp. 179-193
Sailor D. J. (2008), A green roof model or building energy simulation programs, in Energy and
Buildings, Vol. 40(8), pp. 1466-1478
APPLICATION OF TRAP & TREAT TECHNOLOGY FOR ACHIEVING SUSTAINABLE
REMEDIATION OF CONTRASTING CONTAMINANT PLUMES
Introduction
In recent years remediation practitioners within the contaminated land sector have responded to the
continued growth and development of sustainable initiatives within the environmental industry by
bringing to market remedial solutions and concepts that score highly in social, economic and
environmental criteria. One such remedial technology is Trap and Trap (T&T), a patented US-based
technology that has been applied at a small number of sites around Europe in recent years.
Based on the T&T principples Aecom (former URS) developed a sustainable, safe, cost effective and
innovative remedial strategy to provide effective, long-term capture and treatment of a chlorinated
solvent (CHC) plume and Light Non-Aqueous Phase Liquids (LNAPL), at impacted areas of an
operational bulk storage terminal in London The first site in the UK applying the T&T method.
T&T relies upon high-energy, low-volume sub-surface injections of proprietary water-based formulas of
granular activated carbon (GAC) for hydrocarbon absorption (The Trap mechanism). The T&T activated
carbon are impregnated with amendments to chemically or biologically degrade contaminants (The
Treat mechanism).
BOS200 is a prepared formulation of GAC (powder form) inoculated with a selected blend of naturally
occurring micro-organisms and gypsum (calcium sulphate) to support the continued trapping and
biodegradation of petroleum hydrocarbons on the GAC surface.
The concept of respiration is important in that the organism literally breaths nitrate or sulfate while
oxidizing hydrocarbons. The energy derived decreases as one moves down from nitrate toward
methanogenic respiration. It is clear that if oxygen is available, it will be the preferred electron acceptor.
The catch is that organisms must be present that can take advantage of prevailing conditions at any
given time. When the material is first injected, it is saturated with oxygen. Consequently, no matter what
the prevailing condition is within the plume, the prevailing condition within the BOS 200 is initially
aerobic. Once the oxygen is consumed, nitrate will become the next favored electron acceptor, finally
settling into sulfate reduction along with some methanogenic respiration.
T&T does not typically require follow up injection events (as is common place with ISCO for example) to
provide continued protection from an upgradient source, making it a high-value option with a low carbon
footprint and financial burden, especially where upgradient contaminant sources are inaccessible for
direct treatment or removal. As is the case with other injection-based technologies such as chemical
oxidation, T&T injection recipes can be amended in response to ground conditions and contaminant
load to meet a given set of performance objectives. However, where T&T differs is that once installed
the technology needs little to no on-going maintenance to provide continued capture and treatment, with
the GAC surface regenerating via secondary treatment mechanisms (chemical and/or biological
degradation), thus allowing continued trapping to occur. Furthermore, T&T does not generate solid or
liquid wastes for above ground treatment; it does not rely on hydraulic control of the groundwater to
provide containment or treatment of the plume and it requires no active systems, power, or the
installation of engineering measures or pipework to run.
Pilot Test - Remediation of Chlorinated Solvent and Petroleum Hydrocarbon Plumes at a Bulk
Storage Terminal, London - T&T evaluation.
To demonstrate the efficacy of the T&T technology to deliver a tailored treatment regime to meet site
specific down gradient performance standards at a Chlorinated Solvent Plume and a Petroleum
Hydrocarbon Plume, BOS 100 and BOS200 were installed in the shallow impacted part of the saturated
soils in two defined areas.
The scope of the Pilot test was to design and implement a remedial solution to the off-site migration of
an existing dissolved phase CHC plume and LNAPL mobility within saturated Made Ground source
areas at the site. The CHC plume (cis-1,2-dichloroethene (DCE), chloroethane (CA), and vinyl chloride
(VC)) extended across a large portion of the site, emanating from diffuse sources within the Made
Ground, directly below operational infrastructure (such as tanks, utility service corridors and above-
ground process pipework/gantries). In other locations, the shallow groundwater and Made Ground had
become impacted with free phase hydrocarbons below buildings and tanks (up to 770mm thickness)
which represented significant potential on-going liability.
A remedial options appraisal for the site identified the Trap and Treat (T&T) technology as the most
suitable solution. T&T is a patented in-situ technology developed by Remediation Products Inc (RPI)
with which Aecom had gained notable success across North America and Europe. T&T provided the
opportunity to deliver rapid and sustained treatment in a cost effective manner. It could be
implemented at full-scale with minimal disturbance to site operations and with little to no requirement for
on-going maintenance, satisfying conditions for planned site infrastructure upgrades.
The varying nature of contaminant distribution and ground conditions, combined with site operational
constraints, favoured a passive solution that could deliver a tailored treatment regime to meet a set of
downgradient performance standards.
Both BOS100 and BOS200 were applied at the site to demonstrate the effectiveness of T&T (for
example, conceptual design, safety, remedial performance, and implementation practicalities); to
provide detailed full-scale design and cost information (such as injection spacing, loading and recipe);
and to enable the principles of sustainability to be applied to the area-specific remedial performance
targets and works scope. The BOS100 trial objective was to demonstrate the ability of the technology
to provide plume treatment and capture, whilst the BOS200 trial was aimed at showing how the
technology can reduce (L)NAPL mobility and promote biodegradation of the source term.
Geoprobe rig (towing mixing trailer into position) Top-down interwoven injections of BOS product using high pressure
used for MIP, installation of monitoring wells, pump to provide targeted installation of the product.
recovery of soil cores and injection of BOS product.
A detailed characterisation in each trial area were undertaken to define the design and provide context
to the remedial monitoring programme. To achieve this the following work were conducted: high
resolution Membrane Interface Probe (MIP) study, recoverery of soil cores/samples and installation of
nested groundwater monitoring wells positioned along the direction of groundwater flow. Based on the
achieve results the contaminant geometries and aquifer properties were assessed to provide a final
T&T injection design for each area and a set of baseline conditions prior to injection.
Groundwater
flow direction
Capture and treatment were achieved in each trial area by injecting overlapping volumes of BOS in tree
rows perpendicular to the direction of groundwater flow from 0,5 m above the groundwater surface
(approx. 1.5m bgl), through to the base of the Made Ground horizon (6,0 m bgl). This created a
permeable reactive barrier to strip and degrade CHC from the groundwater (BOS100) and facilitate
LNAPL capture and treatment (BOS200) at source. The volume of BOS to be injected and the spacing
and number of rows were determined by data retrieved during the early characterisation works.
Significant reduction of the pollution rather that transferring it
BOS100 Trial
BOS200 Trial
5,000 litres of BOS slurry mixed with gypsum and 21 litres of concentrated bacterial inoculum were
injected across a grid of 20 locations within the Made Ground to provide capture and treatment of the
observable LNAPL. LNAPL thicknesses were decrease from a maximum of 770mm to 16mm by week
4, remaining stable by week 16 (20mm). Mean reductions in LNAPL thickness within the injection area
were 93.4% by week 16, demonstrating excellent stable control of the LNAPL plume, achieving the
project objective.
BOS100 was shipped and stored within airtight barrels until Injection of BOS100 using Geoprobe
immediately prior to use, it was then saturated with water and injection tool connected via high-pressure
pumped directly into the injection mixing vessel to prevent hose to the mixing trailer.
oxidation of the metallic iron and heat generation. BOS200 was
transferred directly from 25kg sacks into the mixing vessel using a vacuum line (Eductor) to minimise
manual handling. No additional H&S measures or arrangements related to the technology were
required over and above the usual provisions for working on contaminated land within a construction
setting. No H&S incidents occurred and no facility activities were compromised during the three-week
project programme - site operations functioned as normal.
Best assessment of the remedial options and considerations of sustainability to include wider
stakeholder impacts on and off-site
T&T was considered the most suitable approach because it was considered to provide the most
effective and robust means of reducing the contaminant burden, whilst minimising site disruption during
installation and operational periods. Aecom undertook a qualitative sustainability assessment during
remedial selection to score shortlisted remedial technologies against a range of economic,
environmental and social factors. A Sustainability Action Plan (SAP) were developed for the T&T works
to identify and action aspects that would maximise sustainability during the project. Examples of the
actions implemented included work phasing to reduce supervision costs and project length; the use of
local suppliers (where possible); employing low-flow groundwater sampling techniques; field screening
to minimise waste generation and electronic reporting.
Aecom (Consultant) and Ejlskov A/S (Contractor) would like to sincerely acknowledge the following stakeholders
who provided key support to this collaborative project: -
Client: Confidential
Remediation Products Inc (RPI), Golden, Colorado, USA; Technology Holder (www.trapandtreat.com)
AST Environmental Inc, Springboro, Ohio, USA; Technical Installation Support (www.astenvironmental.com)
Author
James Wilson, Associate Remediation Specialist, Aecom, St Georges House, 5 St Georges Road, Wimbledon,
UK, SW19 4DR. Email: james.x.wilson@aecom.com
Claus lund, Head of Dep. Ejlskov A/S, Jens Olsens Vej 3, 8200 Aarhus N, Denmark. Email: coe@ejlskov.com
Keywords
Sustainable remediation
Trap & Treat
BOS100
BOS200
LNAPL reduction
CHC Plume capture