Sulfur-based autotrophic denitrification in static beds has proven successful for treating nitrate in groundwater, landfill leachate,andwastewatersKoenigandLiu,1996;Leeetal.,2008; Shaoetal
Trang 1j ou rn a l h o m epa g e : w w w e l s e v i e r c o m / l o c a t e / a q u a - o n l i n e
Laura Christiansona,∗, Christine Lepinea, Scott Tsukudaa, Keiko Saitob,
Steven Summerfelta
a The Conservation Fund, Freshwater Institute, 1098 Turner Road, Shepherdstown, WV 25443, USA
b University of Maryland Baltimore County and Institute of Marine and Environmental Technology, 701 East Pratt St., Baltimore, MD 21202, USA
a r t i c l e i n f o
Article history:
Received 6 April 2015
Received in revised form 13 July 2015
Accepted 17 July 2015
Available online 21 July 2015
Keywords:
Denitrification
Autotrophic
Mixotrophic
Sulfur
Fluidized biofilter
Recirculating aquaculture
a b s t r a c t
Thereisaneedtodeveloppracticalmethodstoreducenitrate–nitrogenloadsfromrecirculating aqua-culturesystemstofacilitateincreasedfoodproteinproductionsimultaneouslywithattainmentofwater qualitygoals.Themostcommonwastewaterdenitrificationtreatmentsystemsutilizemethanol-fueled heterotrophs,butsulfur-basedautotrophicdenitrificationmayallowashiftawayfrompotentially expen-sivecarbonsources.Theobjectiveofthisworkwastoassessthenitrate-reductionpotentialoffluidized sulfur-basedbiofiltersfortreatmentofaquaculturewastewater.Threefluidizedbiofilters(height3.9m, diameter0.31m;operationalvolume0.206m3)werefilledwithsulfurparticles(0.30mmeffective par-ticlesize;staticbeddepthapproximately 0.9m)andoperatedintriplicatemode (PhaseI:37–39% expansion;3.2–3.3minhydraulicretentiontime;860–888L/(m2min)hydraulicloadingrate)and inde-pendentlytoachievearangeofhydraulicretentiontimes(PhaseII:42–13%expansion;3.2–4.8min hydraulicretentiontime).DuringPhaseI,despiteonlyremoving1.57±0.15and1.82±0.32mgNO3–N/L eachpassthroughthebiofilter,removalrateswerethehighestreportedforsulfur-baseddenitrification systems(0.71±0.07and0.80±0.15gNremoved/(Lbioreactor-d)).Lowerthanexpectedsulfate pro-ductionandalkalinityconsumptionindicatedsomeofthenitrateremovalwasduetoheterotrophic denitrification,andthusdenitrificationwasmixotrophic.Microbialanalysisindicatedthepresenceof Thiobacillusdenitrificans,awidelyknownautotrophicdenitrifier,inadditiontoseveralheterotrophic den-itrifiers.PhaseIIshowedthatlongerretentiontimestendedtoresultinmorenitrateremovalandsulfate production,butincreasingtheretentiontimethroughflowratemanipulationmaycreatefluidization challengesforthesesulfurparticles
©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense
1 Introduction
Theglobal demand for foodprotein must bebalanced with
increasedconcernfortheenvironmentalimpactcausedbythese
productionsystems.Land-basedclosed-containmentaquaculture
usingrecirculatingaquaculturesystems(RAS)areuniquelypoised
toproducehighlydesirableandvaluablefoodproductswhilealso
maintainingasmallenvironmentalfootprint.However,whilemost
RASaredesignedtoremovesolidsandrecyclewaterbacktothefish
culturetanks(SummerfeltandVinci,2008;TimmonsandEbeling,
2010),theinabilityofthesesystemstoremovenitrate–nitrogen
fromthewatersignificantlyarreststhisindustry’sultimate
eco-nomic and environmental sustainability For these aquaculture
∗ Corresponding author Tel.: +1 304 870 2241; fax: +1 304 870 2208.
E-mail address: L.Christianson@freshwaterinstitute.org (L Christianson).
systemstomorecompletelyaddressenvironmentalissues,itisnow criticalthateffortsfocusuponthereductionofnitrogenspeciesin effluentwaters.Importantly,theabilitytoconfidentlyand con-sistentlyremove nitratenitrogen fromRAS effluentmay allow expansionofthisindustryintolocalescurrentlyboundby strin-gentwaterqualitystandardsandmaypotentiallyallowincreased reuseoftreatedeffluents.Thereisacrucialneedtodevelop practi-calandcosteffectivemethodstoreduceRASnitrate–nitrogenloads
toallowtheirmaintainedorincreasedproductivitysimultaneously withattainmentofwaterqualitygoalsandgood environmental stewardship
Themostcommonwastewaterdenitrificationsystemsarebased
on heterotrophicdenitrification with the addition of methanol (Payne,1973).However,sulfur-basedautotrophicdenitrification, wherea reducedformofsulfur (e.g.,thiosulfate,elemental sul-fur) serves as the electron donor rather than organic carbon, presentsseveraluniquebenefits(Eq.(1)).Comparedtoheterotopic http://dx.doi.org/10.1016/j.aquaeng.2015.07.002
0144-8609/© 2015 The Authors Published by Elsevier B.V This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ).
Trang 2denitrification, an autotrophic process does not require any
additional potentially expensive carbon source, and produces
less bacterial sludge thus simplifyingtreatment (Batchelor and
Lawrence,1978;KoenigandLiu,1996;ZhangandLampe,1999)
Elementalsulfurisapromisingsubstrateforautotrophic
denitrifi-cationasitisgenerallyinexpensiveandnon-toxic(Batchelorand
Lawrence,1978;SahinkayaandKilic,2014;Sahinkayaetal.,2014)
NO−3+1.10S+0.40CO2+0.76H2O+0.08NH+
4 →0.08C5H7O2N +0.50N2+1.10SO2 −
Amajordisadvantageofthisprocessisthatautotrophsgenerally
growata slowerratethanheterotrophs,thushavelower
deni-trificationrates(SahinkayaandKilic,2014).Majorby-productsof
concernfromsulfur-basedautotrophicdenitrificationaresulfate
andacidity(SahinkayaandKilic,2014)with4.57mgCaCO3
alka-linityconsumedand7.54mgsulfateproducedforeachmgNO3–N
reduced(Sahinkayaetal.,2014).Manysulfur-baseddenitrification
studiesuseamixofsulfurandlimestoneordolomitetobufferpH
andalkalinitydecreases(SahinkayaandKilic,2014;USEPA,1978)
Thepresenceofsulfurinthesystemcombinedwithlow-oxygen
conditionscouldalsoleadtosulfideproduction,thoughnotasa
resultofEq.(1).Anadditionalchallengeisthatelementalsulfur
isrelativelywater insoluble,meaningithasalimited microbial
availabilityatroomtemperature.BatchelorandLawrence(1978)
outlinedthatforelementalsulfur-baseddenitrificationtoproceed,
threestepswerenecessary:(1)thesulfurmustbesolubilized,and
(2)nitratemustbetransportedfromsolutiontothebiofilm
sur-face,where(3)itcanbetransportedthroughthefilmsoitcanbe
denitrified
Sulfur-based autotrophic denitrification in static beds has
proven successful for treating nitrate in groundwater, landfill
leachate,andwastewaters(KoenigandLiu,1996;Leeetal.,2008;
Shaoetal.,2010),andthisapproachpresentsauniqueoptionfor
treatmentof aquacultureeffluents (Sheret al., 2008) Nitrogen
removalratesfrompreviouslaboratorystudiesaregenerallyonthe
orderof0.1–0.4gN/(Ld)(LampeandZhang,1996;Sahinkayaand
Kilic,2014;Sahinkayaetal.,2014).Nitrogen(N)removal
perfor-mancemaybelimitedbyNloadingwithKimetal.(2004)observing
adeclineinNremovalbeyondloadingratesof2.5kgNO3–N/(m3
-d),andKoenigandLiu(1996)notingthatarealbasedloadingrates
(gN/m2-d) weretheirlimitingfactor ina packedsulfur bed.In
anaquacultureapplication,Sheretal.(2008)reportedtheuseof
autotrophicdenitrificationprovidedadualbenefitforrecirculated
waters;not onlywerenitratelevelsbroughtundercontrol,but
theoxidationofsulfideintheanaerobicallydigestedsludgehelped
safeguardagainstsulfidetoxicitywithinthesystem
Fluidizedbedreactorsareaprovenaquaculturewatertreatment
technologyduetotheirpluggingprevention,easeofmaintenance,
lowcostandefficienttreatment(Summerfelt,2006).Because
flu-idizedsandbiofiltersarecommoninthisindustry,theirapplication
asfluidizedsulfurautotrophicdenitrificationreactorscouldbea
naturalextensionofthetechnology.Fluidizedsulfurbiofiltershave
beenresearchedatthelabscale,withKimetal.(2004)showing
higherNremovalratesfromfluidizedsulfurbedsthanpacked
sul-furbeds.Thiswasduetotheabsenceofcloggingandgoodnitrate
transfertothesulfursurfaceinthefluidizedsystem.Inprevious
work,ChristiansonandSummerfelt(2014)determinedfluidization
velocitiesofcommercially-availablesulfurflakes,grains,and
pow-der,andconcludedthegrainsprovidedthemostrealisticoptionfor
full-scaletestingofafluidizedsulfur-baseddenitrificationbiofilter
Theobjectiveofthisworkwastoassessthenitratereduction
poten-tialoffluidizedsulfur-basedbiofiltersfortreatmentofaquaculture
wastewater
2 Methods and materials
2.1 Fluidizedsulfurbiofilterexperimentalset-up Threefluidizedsulfurbiofilters(285L,height3.9m,diameter 0.31m;Fig.1)wereoperatedatTheConservationFund’s Fresh-waterInstitute(Shepherdstown,WestVirginia,USA)for253days
toquantifynitrateremovalfromaquaculturewastewater(Phase I: 225days, 13 March 2014 to 23 October 2014; Phase II: 28 days, 24 October 2014 to 20 November2014) During Phase I, thethree biofilterswereoperatedintriplicatefashion,each flu-idizedat37–39%expansionwithahydraulicretentiontime(HRT)
of3.2–3.3minandahydraulicloadingrateof860–888L/(m2min) basedonthemeanflow rateof63–65L/min.TwoPhaseIstudy periods of relatively consistent influent nitrate concentrations wereselectedforanalysis;Periods1and2allowedevaluationat influentconcentrationsof2.0–5.0and7.6–17mgNO3–N/L, respec-tively(days57–92and190–225,respectively;sixsampleevents each).Phase IIutilized a differentflow ratein each biofilterto assess the impact of HRT on nitrate removal (i.e., no replica-tion;42–13%expansion;3.2–4.8minHRT;67–43L/minflowrate; influent8.5–15mgNO3–N/L)
The waste and wastewater treatment system and biofilter design has beenpreviously described byTsukuda et al.(2015) (Fig.2).In short,wastesludge fromtheproduction of rainbow trout(Oncorhynchusmykiss)andAtlanticsalmon(Salmosalar),was concentratedviamicroscreendrumfiltersandradialflowsettlers andwaspumpedtoaseriesofgravitythickeningsettlingcones.A holdingtankforthesupernatantoverflowfromthesettlingcones fedthethreefluidizeddenitrificationbiofilters.Overflowfromeach biofilterwastreatedusingaradialflowsettler.Biofilterbedheight
Trang 3Fig 2. Process flow diagram for units involved with biofilter denitrification research
(Modified from Tsukuda et al., 2015 ); triplicate replication of settling cones,
biofil-ters, and radial flow settlers not shown Settling cones and radial flow settlers
drained every four weeks and bi-weekly, respectively, to prevent sludge
accumula-tion (off-site sludge disposal).
(2.82m;biofiltervolume0.206m3)wascontrolledwitha
shear-ingpumpatthetopofeachbiofilter.Thestaticsulfurbeddepth
wasapproximately0.9m,althoughthesulfurgrainsinallthree
biofilterswerereplenishedondays181and198followingan
unde-tectedwash-out(68kgorapproximately0.75mSperbiofiltertotal
replenished).Biofilterinfluentnitratelevelsweremanipulatedby
dosinga concentrated sodiumnitrate(NaNO3; 34.0gNO3–N/L)
solutionintothesupernatantholdingtank.Thespringwater
feed-ingtheRASwasnaturallyalkaline(≈275mgCaCO3/L),resultingin
highalkalinityofflows
Thesulfur grains had effective and calculating sizesof 0.30
(D10) and 1.31mm (D90), respectively, and a uniformity
coeffi-cientof3.1(GeorgiaGulfSulfur,CustomerCode1660,distributed
byPrinceAgri-Products,Inc., Quincy, Illinois,USA;Christianson
andSummerfelt,2014).Thisissmallerthanreportedparticlesize
rangesforothersulfur-baseddenitrificationstudiesasmosthave
usedgrainsrangingfrom2to16mm(KoenigandLiu,1996;Oh
etal.,2003; Sahinkayaet al.,2014).Sahinkaya and Kilic(2014)
reportedusingthemostcomparablesize(0.5–1.0mmgrains)ina
packedcolumnstudy,andintheonlyreportedfluidizedbedstudy,
Kimetal.(2004)used2.0–3.35mmsulfurgrains.Thesmallergrain
sizeusedhereprovidedadesirablehighspecificsurfacearea(SSA
bed:4110m2/m3)relativeto,forexample,a4.4mmmean
parti-clesizesulfurproductthathada1363m2/m3SSA(KoenigandLiu,
1996).Elementalsulfurpowderwasinitiallyusedinthefluidized
biofilters,butwasdiscontinuedduetofluidizationandwash-out
challenges.LampeandZhang(1996)similarlyreporteddifficultly
withpowderedsulfurinabatchreactor(i.e.,uniformmixingwas
problematic)
2.2 Waterqualityparametersandanalysis
Waterqualitysampleswerecollectedfromasamplingvalve
locatedat the backof each of thethree biofilters and directly
fromtheinfluent supernatant tank(i.e.,effluent samplevalues
werepooledasreplicatesduringPhaseI,n=3;influentsamples,
n=1) Water chemistry was analyzed weekly onsite, and both
studyphasesfollowedthesamesamplingroutine.Sampleswere
analyzed for chemical oxygen demand (COD), carbonaceous
biochemical oxygen demand (cBOD5), total ammonia nitrogen
(TAN),nitrite–nitrogen(NO2–N),nitrate–nitrogen(NO3–N),total
nitrogen(TN),alkalinity,pH,sulfate (SO4 − sulfide(S2 − total
suspended solids (TSS), total phosphorus (TP), and dissolved
reactivephosphorus(DRP)usingmethodsfromAPHA(2005)and HachCompany(2003).Temperature,dissolvedoxygen(DO),and oxidationreductionpotential(ORP)wereassessedatleasttwice weekly.Measurementsweremadedirectlyfromthesupernatant tank(influent)andtheopenbiofiltertops(effluent)utilizingboth inlineand handheldprobes(HACH HQ40dPortablemeter with either HACH IntelliCAL LDO101 or MTC101 ORP/redox probe; HACHpHDscDifferentialORPsensorwithHACHsc100controller; HACHAdvancedLDOProcessDissolvedOxygenProbewithHACH sc200controller).Flowratewasmeasuredalongtheinfluentpipes
toeachbiofilterandadjustedatleasttwiceweeklyconcurrently withtemperature,DO,andORPreadings,aswellaspriortothe weeklywaterchemistrysamplingevent(DynasonicsDXNPortable UltrasonicMeasurementSystem)
Nitrate–N, sulfide, and alkalinity removal rates were based upon:
Removal rate
=(influenttotalconcentrationexpanded−effluentbiofilter concentration)volume of 206×Lflow rate
(2) with sulfate production rates calculated similarly except the influentconcentrationwassubtractedfromtheeffluent.Statistical analysisconsistedoft-testingtoascertainsignificantdifferences betweeninfluent and effluentparameterconcentrations during bothstudyperiods,orinthecaseofnon-normallydistributeddata
asmostoftheconcentrationdataturnedouttobe,Mann–Whitney RankSumtestswereused(˛=0.05;SigmaPlot12.5).Nitrate–N removalefficiencywascalculatedas:
Removal efficiency
=(influent concentrationinfluent concentration−effluent concentration)×100%
(3) 2.3 CollectionandextractionofDNA
Samplesforscreeningthepotentialdenitrificationcommunity werecollectedfromallthreebiofiltersonthefinaldayofPhase
IPeriod2(day225) Biofilmattachedtothesulfur mediawere detachedbyvigorouslyvortexingasampleofsulfurmedia/biofilter waterin50mLsterileplasticconicaltubesfor5min.Theresulting suspensions of detached surface layer (SL) biofilms were cen-trifugedat10,000×gat4◦Cfor20minpriortoDNAextraction FollowingSLbiofilm detachment,themediaweredirectlyused forDNAextractionofinnerlayer(IL)biofilm.ThegenomicDNA wereextractedfromeachreactor’sSLandILbiofilmusinga Pow-erSoilDNAExtractionKit(MOBIOLaboratories,Inc.,Carlsbad,CA) followingthemanufacturer’sprotocol.Theconcentrationand qual-ityofextractedDNAweredeterminedbyabsorbanceat260nm and260/280nmratio,respectively(NanoDrop2000cUV-Vis spec-trophotometer; ThermoFisher Scientific, Inc., Wilmington, DE) IsolatedDNAwasstoredat−20◦C.
2.3.1 PCRamplification MicrobialcommunityDNAextractedfromthreebiofilterswere pooled in equal quantity and used to amplify nosZ fragments which encode the catalytic subunit Z of nitrous oxide reduc-tase.PrimersnosZ-F(5-CGYTGTTCMTCGACAGCCAG-3)andnosZ-R (5-CATGTGCAGNGCRTGGCAGAA-3)yieldingapprox.700bp frag-ments(Röschetal.,2002)wereused.PCRreactionmixtureswere preparedtocontain2×TaqPCRMasterMix(QIAGEN,Gaithersburg, MD),6pmolofeachforwardandreverseprimers,and100ngof genomicDNAinafinalvolumeof20L.ThePCRamplificationwas
Trang 4onecycleat94◦Cfor20s(denaturation),at65◦Cfor30s
(anneal-ing),andat72◦Cfor40s(elongation);twocyclesat94◦Cfor20s
(denaturation);at62◦Cfor30s(annealing);72◦Cfor40s
(elonga-tion);threecyclesat94◦Cfor20s(denaturation);at59◦Cfor30s
(annealing);at72◦Cfor40s(elongation);fivecyclesat94◦Cfor20s
(denaturation);at57◦Cfor30s(annealing);at72◦Cfor40s
(elon-gation);twentyfourcyclesat94◦Cfor20s(denaturation);at55◦C
for30s(annealing);at72◦Cfor40s(elongation);andthen,final
extensionat72◦Cfor10mininaPTC-200PeltierThermalCycler
(MJResearch,Watertown,MA).Anegativecontrolprepared
with-outDNAwasincludedineveryPCRreactionperformedtotestfor
falsepositivescausedbycontamination.PCRproductswere
sepa-ratedandvisualizedbyelectrophoresisin1.2%agarosegelstained
withEtBr,andwerepurifiedfromexcisedgelslices(about700bp
sizeband)usingtheQIAquickGelExtractionKit(QIAGEN,Valencia,
CA)
2.3.2 Cloningandsequencing
PurifiedSLandILnosZampliconswereligatedintopCR4TOPO
vector, and vector with insert were transformed into OneShot
TOP10chemicallycompetentEscherichiacolicellsusingTOPOTA
CloningKitfollowingthemanufacturer’sinstructions(Invitrogen
LifeTechnologies,Carlsbad,CA).Ninety-sixtotalcloneswere
ran-domlyselectedfromeachSLandILnosZlibraryandwerecultured
forplasmidpreparation.PlasmidDNAswerepurified(Agencourt
SprintPrep384 HCKit, Agencourt Bioscience,Beverly, MA) and
sequencingwasperformedusinganABIPRISMgeneticanalyzer
(AppliedBiosystems,FosterCity,CA)withT7andT3primers
pro-vided in the cloning kit (Invitrogen) at the Biological Analysis
ServiceLaboratory,InstituteofMarineandEnvironmental
Technol-ogy(Baltimore,MD).Sequenceswereeditedandassembledusing
Sequenchersoftware(GeneCodeCorp.,AnnArbor,MI,USA),were
analyzedusing theBasicLocal Alignment SearchTool(BLASTn;
http://www.ncbi.nlm.nih.gov/BLAST), and were compared with
availablesequencesintheGenBankdatabasetocreateneighbor
joiningphylogenetictreestoaidtheselectionoftheclosest
refer-encesequences
2.3.3 Nucleotidesequenceaccessionnumbers
The41partialnosZsequencesthatweregeneratedinthisstudy
havebeendepositedinGenBankdatabaseunderaccession
num-bersKT252910toKT252950
3 Results and discussion
3.1 PhaseI:HighandlownitrateloadingataconsistentHRT
Nitrate reduction wasobserved duringboth Phase Iperiods
(Fig.3a).Althoughdifferencesbetweeninfluentandeffluentnitrate
concentrations were relatively small (Table 1; 1.57±0.15 and
1.82±0.32mgNO3–N/Lforthetwoperiods,respectively),thehigh
flowratesandcompactbiofiltervolumeresultedinmeanremoval
rates of 0.71±0.07 and 0.80±0.15g N removed/(L
bioreactor-d) for the two periods, respectively This is much higher than
the previously reported range of 0.1 to 0.4gN/(Ld) for
sulfur-baseddenitrification(LampeandZhang,1996;SahinkayaandKilic,
2014;Sahinkayaet al.,2014), butsimilartothelow endofthe
rangeforfluidizedsandbiofilterheterotrophicNremovalratesof
0.86–1.74gN/(Ld)(or35.8–72.6mgNO3–N/(Lh);reviewedbyvan
Rijnetal.,2006).Relativetopreviousexperimentswiththese
biofil-ters,Tsukudaetal.(2015)reportedremovalratesof0.4gN/(Ld)
whentheywereoperatedwithfluidizedsand.Christiansonand
Summerfelt(2014)reportedsandwasmuchlessexpensivethan
sulfurproductsforfluidizedbiofiltersonbothavolumetric and
surface area basis ($70–$200/m3 vs >$1000/m3, respectively;
Fig 3. Influent and effluent NO 3 –N (a), sulfate (b), and sulfide (c) concentrations during fluidized sulfur denitrification biofilter operation during Phase I (effluent
n = 3; mean ± standard error).
$0.02/m2 surfaceareavs.≈$0.30/m2 surfacearea, respectively), thoughafluidizedsandbiofilterwouldalsorequirepurchaseofa carbonsourcetofueldenitrification.Influentloadingaveraged1.46 and5.82gN/(Ld)forPeriods1and2,respectively.Nitrateremoval efficienciesaveraged50±4.6%and16±3.2%forthetwoPhaseI studyperiods,withtherelativelyhighefficiencyforPeriod1due
tothelowinfluentnitrateconcentration
Theoretically,theproductionofsulfateisproportionaltothe extentofautotrophicdenitrification,thussulfateproductionmay
bethebestindicatorofthisprocess(Ohetal.,2003;Sahinkayaetal.,
2014).BasedonEq.(1)andaverageremovalsof1.57and1.82mg
NO3–N/L, Periods 1and 2 shouldhave produced anaverageof 11.8and13.7mgSO4 −/L.However,only2.7±2.0and6.1±1.6mg
SO4 −/Lwereproducedduringthesetwoperiods,withno statis-ticallysignificantdifferencebetweeninfluentandeffluentsulfate concentrationsforeitherPeriods1or2(Table1;Fig.3b).Thisis
anindicationthatsomeoftheNremovalwaspotentiallydueto heterotrophicdenitrificationinadditiontoautotrophic.Justasthe elementalsulfur wasconvertedtosulfate, somesulfidepresent
in solution wasalsooxidized (Fig.3c; Table 1; mean removal: 6.19±1.82and8.64±1.04gS2 −/L).Sheretal.(2008)observed
thataRASsludgedigestionbasinalsoprovidedautotrophic deni-trification treatment with sulfide as the electron donor Dual functionalityofnitrateandsulfideremovalwouldbeamore signif-icantbenefitforRASwatersbeingrecirculatedtofishculturetanks
ascomparedtothetreatmentofeffluentwatershere
Reducedalkalinity,anotherindicatorofautotrophic denitrifica-tion,wasobservedherewithaveragedecreasesof16and12mg CaCO3/Lfrom the two Phase I study periods (Table 2) Others havereportedsignificantdropsinalkalinityduringsulfur-based denitrification studies(Koenigand Liu, 1996), and this may be
Trang 5Table 1
Mean ± standard error influent and effluent parameter concentrations during two study periods of Phase I operation of a fluidized sulfur denitrification biofilter experiment where columns were run in triplicate; influent n = 6, effluent n = 18 with the exception of Period 2 cBOD 5 where influent n = 5, effluent n = 15; concentrations in mg/L except sulfide in g S2−/L.
a Abbreviations: Total ammonia nitrogen (TAN); total nitrogen (TN); chemical oxygen demand (COD); carbonaceous biochemical oxygen demand (cBOD 5 ); total phosphorus (TP); dissolved reactive phosphorus (DRP).
b A statistically significant difference between influent and effluent concentrations existed for Period 1, but not Period 2 (t-test).
c No statistically significant difference between influent and effluent concentration for either Periods 1 or 2 (Mann–Whitney Rank Sum tests; ˛ = 0.05).
Table 2
Mean ± standard error flow rates and influent and effluent alkalinity, pH, dissolved oxygen, oxidation reduction potential, and water temperature during two study periods
of Phase I operation of a fluidized sulfur denitrification biofilter experiment; study Period 1: influent n = 14, effluent n = 42; study Period 2: influent n = 10, effluent n = 30.
a No statistically significant difference between influent and effluent concentration for either Periods 1 or 2 (Mann–Whitney Rank Sum tests; ˛ = 0.05).
thelargestoperationalchallengeofsuchasystem(KimandBae,
2000).Thenaturallyalkalinespringwaterusedintheon-siteRAS
herewasconsideredwell-bufferedenoughtonotrequirealkalinity
additionasFurumaietal.(1996)reportedtheoptimum
alkalin-ityforsulfur-basedautotrophicdenitrificationwas150–240mg/L
BasedonEq (1), removalof 1.57and 1.82mg NO3–N/Lshould
haveresulted in alkalinityconsumption of only 7.2and 8.3mg
CaCO3/L for Periods 1 and 2, respectively Likewise, based on
Nremoval rate(0.71and 0.80gN/(Ld)),alkalinityconsumption
shouldhavebeen3.2and3.7gCaCO3/(Ld)althoughitaveraged
7.1±2.7and5.3±2.3gCaCO3/(Ld)forthetwoperiods.The
simul-taneousoccurrence of heterotrophicdenitrification would have
reducedalkalinityconsumptionratherthanincreased
consump-tion, and while nitrification can consume alkalinity, there was
noconsistentchangeinTANconcentrationsacrossthebiofilters
Degradationof possibleaccumulatedsludge withinthebiofilter
mayhaveconsumedsomealkalinity,althoughthiscouldnotbe
ver-ified.Thevariabilityinalkalinitystandarderrorcomplicatedfurther
analysis
No major pHchanges were observed with theinfluent and
effluentboth averagingbetween7.33and7.39for bothperiods
(Table 2) Others have observed notable pH decreases (Koenig
andLiu,1996;SahinkayaandKilic,2014)withnitrite
accumula-tionpossibleatpHbelow 7.4(Furumaietal.,1996).Therewas
noaccumulationofnitritehereaslevelsweregenerallyslightly
reducedover thebiofilters(Table1;Fig.4).Watertemperature
betweentheinfluentandeffluentdidnotnotablyvary,although
a seasonal trend was observed Temperatures peaked between
days100–150(20June2014–09August2014)duringthewarmest
timeforthesegreenhouse-runexperiments(Fig.5a).Asexpected,
effluentDOconcentrationswerereducedtolessthan1.0mgDO/L
whenthecolumnswereoperatingasintended(Fig.5b),andto
lessthan0.5mgDO/Lduringbothanalysisperiods(Table2).This
indicatedastrongaerobicand/orfacultativeanaerobiccomponent existedwithinthebiofilters.Facultativeheterotrophicdenitrifiers usefree oxygenastheirelectron acceptorwhile it isavailable, becauseoxygenisamoreenergeticallyfavorableelectron accep-torthannitrate.Thus,heterotrophicdenitrificationandtheuseof nitrateasanelectronacceptorisreducedwhenfreeoxygenisstill present.Autotrophicdenitrificationhasbeendocumentedunder bothaerobicandanaerobicconditions(ZhangandLampe,1999) Thepotentialimpactofthesulfurwash-outwasevidentasearly
asday150wheneffluentDOlevelsincreased;additionalsulfur wasaddedondays181and198.Oxidationreductionpotentials
Fig 4.Influent and effluent nitrogen species concentrations during fluidized sulfur denitrification biofilter Phase I operation (effluent n = 3; mean ± standard error).
Trang 6Fig 5.Influent and effluent temperature (a), dissolved oxygen (b), and oxidation
reduction potential (c) during fluidized sulfur denitrification biofilter operation from
Phase I (effluent n = 3; mean ± standard error).
werehighlyvariablethoughalwaysnegative,andwerereduced
slightlyacrossthebiofilterduringPeriod1butincreasedduring
Period2(Table2;Fig.5c).ThisincreaseinORPacrossthebiofilters
wasmainlyapparentbecausetheinfluentORPwasmorereduced
duringthisperiod;influentwaterqualitythroughoutthe
experi-mentwasvariableandsomewhatuncontrollableduetothenature
ofthisproductionaquaculturefacility’swastestream
Theterm“mixotrophicdenitrification”referstothe
simulta-neousoccurrenceofheterotrophicandautotrophicdenitrification
(Oh etal.,2003;SahinkayaandKilic,2014).Withthisrelatively
highCODand cBOD5 wastewater,it islikely mixotrophic
deni-trificationwasoccurring Oh etal (2003) observedaddition of
avarietyofsolubleorganicsources(methanol,ethanol,acetate)
didnot inhibitautotrophicdenitrification,although
supplemen-tationoforganiccarboninexcessdiddecreasesulfateproduction
Balancingtheautotrophic/heterotrophicreactionscanreducethe
alkalinityrequirementcausedbyautotrophicdenitrificationdue
toalkalinityproducedbyheterotrophs(Kimand Bae,2000;Lee
etal.,2001;Ohetal.,2003).Becauseheterotrophsgrowfasterthan
autotrophs,someorganiccarbonformsmaybepreferentially
uti-lizedbeforesulfurinamixotrophicdenitrificationreactor(Sunand
Nemati,2012).Availabilityoftheelectrondonormayplayarolein
thisaslimiteddissolutionofsolidsulfurparticlescanlimit
deni-trification,especiallyathigherNloadingrates(Kimetal.,2004)
SuitableCOD:NO3–Nratiosforheterotrophicdenitrificationareon
theorderof3:1to6:1(vanRijnetal.,2006),andinfluentvalues
hereaveraged74±7.5and7.2±1.3COD:NO3–NforPeriods1and
2,respectively,morethansufficienttofuelheterotrophic
denitri-fication(cBOD5:NO3–Nof24±4.9and3.3±0.5).However,during
Periods1and2,CODwasonlyreduced8.5±19and2.4±3.0mg
COD/L,respectively,andcBOD5 concentrationswerenotreduced
acrossthebiofilters(Table1).TheCOD:NO3–Nutilizationratios
were5.4and 1.3mgCODconsumed permgNremovedforthe
twoperiods,respectively.TheverylowutilizationratioforPeriod
2potentiallyindicatedrelativelymoreoftheNremovalwasdueto autotrophicvs.heterotrophicdenitrificationcomparedtoPeriod1 TheabsenceofmeasureablecBOD5reductionswaslikelyduetothe extremelyshortHRTs.Whileitislikelythatheterotrophic denitrifi-cationdidaccountforsomeofthenitrateremoval,internalcycling
ofsolidsmayhavecomplicatedtheCODbalances
3.2 PhaseI:Microbiologicalcharacterization Sequenceanalysisof 96randomlyselectedclonesfromeach thebiofilm surfacelayer(SL)and inner layer(IL) nosZlibraries revealedfourteenuniqueoperationaltaxonomicunits(OTUs)for
SL and nine for IL (Table3).The % Clone of similarsequences
in a library werecalculated for SL-nosZ and IL-nosZ,separately (Table3,upperforSL-nosZandlowerforIL-nosZ).ThenosZlibrary clonesintheSLbelongedto:Alphaproteobacteria(19.6%), Betapro-teobacteria(76.5%),andunclassifiedbacterium(4.3%);andintheIL belongedto:Alphaproteobacteria(2.2%),Betaproteobacteria(17.5%), andunclassifiedbacterium(80.4%).Similarlytopreviousfluidized sandbiofilterdenitrificationstudies(Tsukudaetal.,2015),the den-itrifyingmicrobialpopulationcontainingthenosZgeneintheSL wasmorediversethanintheIL.Here,morethan80%ofIL-nosZ cloneswerecloselyrelatedtotheunculturedbacteriumclone2–80 (AccessionJF509076.1).Thislackofdiversitymayhavebeenthe resultoflowerDOandhighersulfuravailability(electrondonor)
intheIL.Unculturedbacteriumclone2–80werealsofoundinthe
SLbiofilm,buttheirmuchhigher%cloneintheIL(4.3vs.80.4%
inSLvs.IL,respectively)maybeanindicationthereare sulfur-utilizingautotrophicdenitrifiersthathavenotyetbeenisolatedor identified
The microbial communities indicated that the encoding key enzymefordenitrification(nosZ)intheSLwaslargelyfrom Azoar-cus,ThaueraandParacoccusspp.,whichareknownasheterotrophic denitrifiers,andtheirpresence suggeststhegeochemical condi-tionsneartheSLweresuitablefor heterotrophicdenitrification compared to conditions in the IL In contrast, nosZ sequences belongingtoThiobacillusdenitrificans,anobligate chemolithoau-totrophicdenitrifier,weredominantintheILdenitrifyingmicrobial communitysuggestingILprovidedasuitablecultivating environ-mentforautotrophs,althoughthis wasonly4.3%oftheIL-nosZ clones(80.4%wereunculturedbacteriumclone2–80).The opti-malgrowthtemperatureofT.denitrificansisbetween28and32◦C (Shaoetal.,2010),andlowerwatertemperatureshere(13–22◦C; Fig.5a)mayhaveinfluencedthisrelativelylowpercentage.Among knownautotrophicdenitrifiers,theobligatechemolithoautotroph,
T.denitrificans, wasthefirsttobeisolatedand characterized,is capableofutilizingthiosulfate,tetrathionate,thiocyanate,sulfide andelementalsulfurastheelectrondonorfordenitrification,and
isthemostcommonlyreportedautotrophicdenitrifier(Parketal.,
2010,2011;Chenetal.,2013;Xuetal.,2014).Thedetectionof nosZgenesfromautotrophicdenitrifiersinbothSLandILbiofilms stronglyindicatedthecapabilityoffluidizedsulfurbiofiltersto cul-tivateandenrichautotrophicdenitrifyingbacteriaforremovalof nitrate,evenunderrelativelyshortHRTscomparedtopacked sul-furreactorstudies.Inaddition,theco-existenceofautotrophicand heterotrophicdenitrifierssuggeststhesereactorsprovided condi-tionstocultivatebothtypesofbacteriawhichcanofferuniqueand efficientmixotrophicnitrateremoval(Ohetal.,2001)
3.3 PhaseII:Hydraulicretentiontimeimpactonautotrophic denitrification
Wheneachbiofilterwasoperated independently,Nremoval and sulfate production showed a weakly increasing trend at increasing HRTs (Fig.6a and b) Based onthe regression slope
Trang 7Table 3
Nearest neighbor of the nitrous-oxide reductase (nosZ) gene clones in the surface layer (SL, upper part of table) and inner layer (IL, bottom part of table) of the biofilm.
nearest neighbor
a The closest matching sequence was identified using Blastn at the NCBI and selected by neighbor joining phylogenetic analysis from Blastn hits.
(−0.0405gN/(Ld) per L/min of flow rate), decreasingthe flow
rateapproximately20L/minwouldprovideanadditional0.81gN
removed/(Ld)whichequatedtoanadditional167gNremoved/d
forthesebiofilters.Removalofsulfidealsotendedtoincreaseat
higherHRTs,thoughthisregressionwasevenlessstrongly
corre-lated(Fig.6a).Forfluidizedsystems,thisreductioninflowrateto
achievealongerHRTisatradeoffresultinginlessfluidizationofthe
sulfurparticles,andthustheHRTwouldneedtobeincreasedviaa largerbiofilter.Arecommended60%expansion,aswasmodeledin ChristiansonandSummerfelt(2014),wouldhaverequiredaflow rateofover80L/minandyieldedanHRTofonly2.5mininthese biofilters.OtherreportedHRTsforpackedorcontinuouslystirred sulfurdenitrificationreactorshavebeenontheorderof3to24h (LampeandZhang,1996;Leeetal.,2001;SahinkayaandKilic,2014;
Fig 6.Nitrate–N and sulfide removal rate (a) and observed and theoretical sulfate production rate, (b) across a range of flow rates, hydraulic retention times, and fluidization
Trang 8HRTfor complete Nreduction dependedupon thesulfur
parti-clesize,andshowedgreaterthan30minwasrequiredfora40%
nitrateremovalefficiencyusingtheirsmallestsulfursizefraction
(2.8–5.6mm)inpackedbeds.Ataloadingof2.2kgN/(m3-d),Kim
andBae(2000)reportedanHRTof2.34hinapackedbedprovided
completedenitrification.LoadingduringPhaseIIwasbetween2.7
and6.7kgN/(m3biofilter-d),thusagreaterHRT,inapackedbedat
least,wouldhavebeenrequiredforcompleteNremoval.Theonly
comparablefluidizedbedstudyreportedanHRTof0.19h(empty
bedcontracttime)andbedexpansionof25–30%(2–3.35mmsulfur
particlesize;Kimetal.,2004).Undertheseconditions,greaterthan
90%removalefficiencywasachievedfromaninfluent
concentra-tionof20mgNO3–N/L.However,Kimetal.(2004)alsoreported
a decline in N removal-performancewhen N loadingexceeded
2.53kgN/(m3-d),asthepresentstudydid.Ideally,thisstudywould
havebeenimprovedifthebiofilterswere2–3mtallerorifaslightly
smaller-sizedsulfurparticlecouldhavebeenidentified,because
bothoptionswouldhaveincreasedtheHRTwithinthe
denitrifica-tionbed
4 Conclusions
Despiteonlyremoving1.57±0.15and1.82±0.32mgNO3–N/L
each pass through the biofilter during Phase I, removal rates
werethehighestreportedforsulfur-baseddenitrificationsystems
(0.71±0.07and0.80±0.15gNremoved/(Lbioreactor-d)).Lower
than expectedsulfate production indicated someof the nitrate
removalwasduetoheterotrophicdenitrificationalthoughthere
wasnostatisticallysignificantdecreaseinCODorcBOD5
concentra-tionsbetweentheinfluentandeffluent.Mixotrophicdenitrification
wasverifiedviathepresenceofbothheterotrophicandautotrophic
denitrifiers.PhaseIItendedtoindicatethatlongerretentiontimes
mayresultin morenitrateremovaland sulfateproduction, but
increasingtheretentiontimethroughflowratemanipulationmay
create fluidization challenges for these sulfur particles
Opera-tionally,thesulfurparticleswilldegradeovertime,andoptimizing
thebalanceoffluidizationvelocityversusHRTmaybechallenging
Acknowledgements
The authors wish to thank theHerrick Foundation (Detroit,
Michigan,USA)fortheirgracioussupport.Thisresearchwas
addi-tionallysupportedbytheUSDAAgriculturalResearchServiceunder
Agreementno.59-1930-0-046.AdebtofgratitudeisduetoShanen
CoganandFredFordfortechnicalassistanceandIssraArifforinitial
assistancewiththetesting
References
APHA, 2005 Standard Methods for the Examination of Water and Wastewater,
21st ed American Public Health Association, Washington, DC.
Batchelor, B., Lawrence, A.W., 1978 A kinetic model for autotrophic denitrification
using elemental sulfur Water Res 12 (12), 1075–1084.
Chen, L.X., Li, J.T., Chen, Y.T., Huang, L.N., Hua, Z.S., Hu, M., Shu, W.S., 2013 Shifts in
microbial community composition and function in the acidification of a
lead/zinc mine tailings Environ Microbiol 15 (9), 2431–2444.
Christianson, L., Summerfelt, S., 2014 Fluidization velocity assessment of
commercially available sulfur particles for use in autotrophic denitrification
biofilters Aquacult Eng 60, 1–5.
Furumai, H., Tagui, H., Fujita, K., 1996 Effects of pH and alkalinity on
sulfur-denitrification in a biological granular filter Water Sci Technol 34
(1–2), 355–362.
Hach Company, 2003 DR/4000 Spectrophotometer Procedures Manual, 11th ed.
Hach Company, USA.
Kim, E., Bae, J., 2000 Alkalinity requirements and the possibility of simultaneous
heterotrophicdenitrification during sulfur-utilizing autotrophic denitrification.
Water Sci Technol 42 (3–4), 233–238.
Kim, H.R., Lee, I.S., Bae, J.H., 2004 Performance of a sulphur-utilizing fluidized bed
Koenig, A., Liu, L.H., 1996 Autotrophic denitrification of landfill leachate using elemental sulphur Water Sci Technol 34 (5–6), 469–476.
Lampe, D.G., Zhang, T.C., 1996 Evaluation of sulfur-based autotrophic denitrification In: Proceedings of the HSRC/WERC Joint Conference on the Environment, Great Plains/Rocky Mountain Hazardous Substance Research Center.
Lee, D.U., Lee, I.S., Choi, Y.D., Bae, J.H., 2001 Effects of external carbon source and empty bed contact time on simultaneous heterotrophic and sulfur-utilizing autotrophic denitrification Process Biochem 36 (12), 1215–1224.
Lee, H.W., Park, Y.K., Choi, E., Lee, J.W., 2008 Bacterial community and biological nitrate removal: comparisons of autotrophic and heterotrophic reactors for denitrification with raw sewage J Microbiol Biotechnol 18 (11), 1826–1835.
Oh, S., Yoo, Y., Young, J., Kim, I., 2001 Effect of organics on sulfur-utilizing autotrophic denitrification under mixotrophic conditions J Biotechnol 92 (1), 1–8.
Oh, S., Bum, M., Yoo, Y., Zubair, A., Kim, I., 2003 Nitrate removal by simultaneous sulfur utilizing autotrophic and heterotrophic denitrification under different organics and alkalinity conditions: batch experiments Water Sci Technol 47 (1), 237–244.
Park, S., Lee, J., Park, J., Byun, I., Park, T., Lee, T., 2010 Characteristics of nitrogen removal and microbial distribution by application of spent sulfidic caustic in pilot scale wastewater treatment plant Water Sci Technol 62 (6), 1440–1447.
Park, S., Yu, J., Byun, I., Cho, S., Park, T., Lee, T., 2011 Microbial community structure and dynamics in a mixotrophic nitrogen removal process using recycled spent caustic under different loading conditions Bioresour Technol.
102 (15), 7265–7271.
Payne, W., 1973 Reduction of nitrogenous oxides by microorganisms Bacteriol Rev 37 (4), 409–452.
Rösch, C., Mergel, A., Bothe, H., 2002 Biodiversity of denitrifying and dinitrogen-fixing bacteria in an acid forest soil Appl Environ Microbiol 68 (8), 3818–3829.
Sahinkaya, E., Kilic, A., 2014 Heterotrophic and elemental-sulfur-based autotrophic denitrification processes for simultaneous nitrate and Cr(VI) reduction Water Res 50, 278–286.
Sahinkaya, E., Kilic, A., Duygulu, B., 2014 Pilot and full scale applications of sulfur-based autotrophic denitrification process for nitrate removal from activated sludge process effluent Water Res 60, 210–217.
Shao, M.F., Zhang, T., Fang, H.H.P., 2010 Sulfur-driven autotrophic denitrification: diversity, biochemistry, and engineering applications Appl Microbiol Biotechnol 88 (5), 1027–1042.
Sher, Y., Schneider, K., Schwermer, C.U., van Rijn, J., 2008 Sulfide-induced nitrate reduction in the sludge of an anaerobic digester of a zero-discharge recirculating mariculture system Water Res 42 (16), 4386–4392.
Summerfelt, S.T., 2006 Design and management of conventional fluidized-sand biofilters Aquacult Eng 34 (3), 275–302.
Summerfelt, S.T., Vinci, B.J., 2008 Better management practices for recirculating aquaculture systems In: Tucker, C.S., Hargreaves, J.A (Eds.), Environmental Best Management Practices for Aquaculture Blackwell Publishing, Ames, IA,
pp 389–426 (Chapter 10).
Sun, Y., Nemati, M., 2012 Evaluation of sulfur-based autotrophic denitrification and denitritation for biological removal of nitrate and nitrite from contaminated waters Bioresour Technol 114, 207–216.
Timmons, M.B., Ebeling, J.M., 2010 Recirculating Aquaculture, second ed Cayuga Aqua Ventures, Ithaca, NY.
Tsukuda, S., Christianson, L., Kolb, A., Saito, K., Summerfelt, S., 2015 Heterotrophic denitrification of aquaculture effluent using fluidized sand biofilters Aquacult Eng 64, 49–59.
USEPA, 1978 Autotrophic Denitrification Using Sulfur Electron Donors (EPA-600/2-78-113) U.S Environmental Protection Agency, Washington, DC.
van Rijn, J., Tal, Y., Schreier, H.J., 2006 Denitrification in recirculating systems: theory and applications Aquacult Eng 34 (3), 364–376.
Xu, X.J., Chen, C., Wang, A.J., Yu, H., Zhou, X., Guo, H.L., Yuan, Y., Lee, D.J., Zhou, J., Ren, N.Q., 2014 Bioreactor performance and functional gene analysis of microbial community in a limited-oxygen fed bioreactor for co-reduction of sulfate and nitrate with high organic input J Hazard Mater 278, 250–257.
Zhang, T.C., Lampe, D.G., 1999 Sulfur: limestone autotrophic denitrification processes for treatment of nitrate-contaminated water: batch experiments Water Res 33 (3), 599–608.
Dr Laura Christiansonhas been a Research Agricultural Engineer for The Conserva-tion Fund’s Freshwater Institute, Shepherdstown, WV since 2013 She finished her Ph.D in Agricultural Engineering (Co-Major: Sustainable Agriculture) at Iowa State University in December 2011 where her dissertation focused on improvement of agricultural drainage water quality through the use of denitrification “woodchip” bioreactors During her Ph.D., she spent a year in New Zealand studying agricultural water quality and denitrification technologies as a Fulbright Fellow Laura previ-ously completed a M.S in Biological and Agricultural Engineering at Kansas State University and a B.S in Biosystems Engineering at Oklahoma State University.
Christine Lepineis a Research Technician for The Conservation Fund’s Freshwater Institute (TCFFI), Shepherdstown, WV She has been with TCFFI since 2014, originally starting as a Research Intern She also recently graduated magna cum laude from Shepherd University with a B.S in Environmental Studies, Concentration of Resource Management.
Scott Tsukudais the Director of Operations at The Conservation Fund’s Freshwater Institute (TCFFI), Shepherdstown, WV, with his focus on energy monitoring and
Trang 9auditing Some of his past work has included MS Excel computer modeling, PLC
programming, denitrification technologies and alternative waste treatment systems
demonstration He is a member of the Instrumentation, Systems and Automation
Society (ISA) and the Institute of Electrical and Electronics Engineers (IEEE) plus
holds a M.S in Agricultural Engineering and a B.S in Agricultural Engineering from
Cornell University Past certifications include Microsoft Certified Systems Engineer
(MCSE) He is currently Certified Energy Manager (CEM).
Dr Keiko Saitohas been a Research Assistant Professor at University of
Mary-land Baltimore County’s Institute of Marine and Environmental Technology since
2010 Her research focuses on aquatic microbial ecology and aquacultural
micro-biology, and on applying molecular approaches to link the critical roles of
microbial community composition, functional diversity, ecosystem processes, and
bio-degradation/remediation She is working toward development and improve-ment of microbially mediated waste treatment technologies for next-generation aquaculture practices.
Dr Steven T Summerfelt,Professional Engineer, is Director of Aquaculture Systems Research at The Conservation Fund’s Freshwater Institute (TCFFI), Shepherdstown,
WV, where he has been an employee since 1992 He is Project Leader on TCFFI’s USDA-ARS project titled, “Development of Sustainable Land-based Aquaculture Pro-duction Systems” and has authored or co-authored of over 60 refereed papers, 9 book chapters, and a book titled “Recirculating Aquaculture Systems” Steve has designed several large private and public fish culture facilities using closed-containment technologies He has B.S., M.S., and Ph.D degrees in the fields of chemical and environmental engineering.