University of Wollongong University of Wollongong

Research Online Research Online

Faculty of Engineering and Information Sciences - Papers: Part A

Faculty of Engineering and Information Sciences

1-1-2015

Selection of forward osmosis draw solutes for subsequent Selection of forward osmosis draw solutes for subsequent

integration with anaerobic treatment to facilitate resource integration with anaerobic treatment to facilitate resource

recovery from wastewater recovery from wastewater

Ashley J. Ansari University of Wollongong, ashleyz@uow.edu.au

Faisal I. Hai University of Wollongong, faisal@uow.edu.au

Wenshan Guo University of Technology Sydney

Hao H. Ngo University of Technology Sydney

William E. Price University of Wollongong, wprice@uow.edu.au

See next page for additional authors

Follow this and additional works at: https://ro.uow.edu.au/eispapers

Part of the Engineering Commons, and the Science and Technology Studies Commons

Recommended Citation Recommended Citation Ansari, Ashley J.; Hai, Faisal I.; Guo, Wenshan; Ngo, Hao H.; Price, William E.; and Nghiem, Long D., "Selection of forward osmosis draw solutes for subsequent integration with anaerobic treatment to facilitate resource recovery from wastewater" (2015). Faculty of Engineering and Information Sciences - Papers: Part A. 3958. https://ro.uow.edu.au/eispapers/3958

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: research-pubs@uow.edu.au

Selection of forward osmosis draw solutes for subsequent integration with Selection of forward osmosis draw solutes for subsequent integration with anaerobic treatment to facilitate resource recovery from wastewater anaerobic treatment to facilitate resource recovery from wastewater

Abstract Abstract Forward osmosis (FO) can be used to extract clean water and pre-concentrate municipal wastewater to make it amenable to anaerobic treatment. A protocol was developed to assess the suitability of FO draw solutes for pre-concentrating wastewater for potential integration with anaerobic treatment to facilitate resource recovery from wastewater. Draw solutes were evaluated in terms of their ability to induce osmotic pressure, water flux, and reverse solute flux. The compatibility of each draw solute with subsequent anaerobic treatment was assessed by biomethane potential analysis. The effect of each draw solute (at concentrations corresponding to the reverse solute flux at ten-fold pre-concentration of wastewater) on methane production was also evaluated. The results show that ionic organic draw solutes (e.g., sodium acetate) were most suitable for FO application and subsequent anaerobic treatment. On the other hand, the reverse solute flux of inorganic draw solutions could inhibit methane production from FO pre-concentrated wastewater.

Disciplines Disciplines Engineering | Science and Technology Studies

Publication Details Publication Details Ansari, A. J., Hai, F. I.., Guo, W., Ngo, H. H., Price, W. E. & Nghiem, L. D. (2015). Selection of forward osmosis draw solutes for subsequent integration with anaerobic treatment to facilitate resource recovery from wastewater. Bioresource Technology, 191 30-36.

Authors Authors Ashley J. Ansari, Faisal I. Hai, Wenshan Guo, Hao H. Ngo, William E. Price, and Long D. Nghiem

This journal article is available at Research Online: https://ro.uow.edu.au/eispapers/3958

Selectionof forward osmosisdraw solutesfor subsequentintegration with

anaerobictreatment to facilitate resourcerecovery from wastewater

RevisedManuscriptSubmittedto

BioresourceTechnology

April 2015

AshleyJ.Ansaria, FaisalI. Hai a, WenshanGuob, HaoH. Ngo b, William E. Pricec, and

LongD. Nghiema,*

aStrategicWaterInfrastructureLaboratory,Schoolof Civil Mining andEnvironmental

Engineering,Universityof Wollongong,Wollongong,NSW 2522,Australia

b Centrefor Technologyin WaterandWastewater,Schoolof Civil andEnvironmental

Engineering,Universityof TechnologySydney,Sydney,NSW 2007,Australia

c StrategicWater InfrastructureLaboratory,Schoolof Chemistry

Universityof Wollongong,Wollongong,NSW 2522,Australia

_______________________

* Correspondingauthor:LongDucNghiem; Email: longn@uow.edu.au;Ph+612 42214590

2

Abstract

Forwardosmosis(FO) canbeusedto extractcleanwaterandpre-concentratemunicipal

wastewaterto makeit amenableto anaerobictreatment.A protocolwasdevelopedto assessthe

suitabilityof FOdrawsolutesfor pre-concentratingwastewaterfor potentialintegrationwith

anaerobictreatmentto facilitateresourcerecoveryfrom wastewater. Draw soluteswere

evaluatedin termsof theirability to induceosmoticpressure,waterflux, andreversesoluteflux.

Thecompatibilityof eachdrawsolutewith subsequentanaerobictreatmentwasassessedby

biomethanepotentialanalysis.Theeffectof eachdrawsolute(at concentrations correspondingto

thereversesoluteflux at ten-fold pre-concentrationof wastewater) onmethaneproductionwas

alsoevaluated. Theresultsshowthatorganic-baseddrawsolutes(e.g.,sodiumacetate) were

mostsuitablefor FOapplicationandsubsequentanaerobictreatment.On theotherhand, the

reversesoluteflux of inorganicdrawsolutionscouldinhibit methaneproductionfromFOpre-

concentratedwastewater.

Keywords:Forwardosmosis;reversesolute flux; biomethanepotential (BMP) analysis; draw

soluteselection;sewermining.

3

1. Introduction

Therecentlyrecognisedvalueof cleanwater,energy, andnutrientsin municipalwastewater

hasled to a paradigmshift in urbanwatermanagement,towardamodernframeworkthat

incorporatesresourcerecoverywith thetraditionalsanitationmandate.Thevalueof these

resourcesgoesbeyondshort-termeconomic outcomes, becauselong-termhumanhealthand

environmentalbenefitscanplayan evengreaterrole in wastewatermanagementdecisions.

Waterscarcityandenvironmentalpollution havedrivenwaterreuseto becomeanintegral

functionof modernwastewatertreatmentplants(Shannonet al.,2008). Furtherefforts to

includeenergyandnutrientrecoveryarejustifiedby therelationshipbetweenthestringency

of effluentregulationsandenergyconsumption(Iranpouret al.,1999), aswell asconcerns

for worldwidephosphorussecurity(Koppelaar& Weikard,2013).

Cleanwaterreclamationfrom municipalwastewateris well established. However,a greater

focusis requiredto furtherdevelopenergyandnutrientrecoverypractices.Thedilute nature

of municipalwastewateris a majorobstaclehindering energyandnutrientrecovery. Thus, it

is necessaryto pre-concentratemunicipal wastewaterby five to ten-fold to achievethe

requiredstrengthin termsof chemicaloxygendemand(COD) for subsequentanaerobic

treatment(Verstraete& Vlaeminck,2011), throughwhich energyandnutrients canbe

recoveredin theform of biogas(Burn etal., 2013; Nghiemet al.,2014a) andstruvite

(MgNH4PO4·6H2O) (Garcia-Belinchónetal., 2013; Xie et al.,2014), respectively. Themost

commonandeffectivetechniqueto recovernutrientsafteranaerobictreatmentis via struvite

precipitation. In thisprocess,magnesiumsaltadditionis requiredfor struvite

(MgNH4PO4·6H2O) formation.However,becauseof thelow ammoniaandphosphate

concentrations in municipalwastewater, magnesiumsaltmustbeaddedto obtaina

concentrationwell abovethestoichiometricratio to facilitatestruviteprecipitation. In this

instance,thepre-concentrationof wastewaterwill lower themagnesiumrequirementfor

struviteformation(McCartyet al., 2011; Xie et al., 2014), thussignificantlyimprovingthe

economicsof nutrientrecovery(Garcia-Belinchónetal., 2013). Thedeploymentof

innovativetechnologiessuchasforwardosmosis(FO) to pre-concentrateorganicmatterand

nutrientscanfacilitateanaerobictreatment, thusallowing resourcerecoveryto become

economicallyviable.

4

FOis a promisingtechnologyfor thepre-concentrationof wastewaterandhasrecently

demonstratedpotentialfor directsewermining (Lutchmiahetal., 2011; Xie et al.,2013).

Whenapplieddirectly for wastewatertreatment,this concentrationdrivenprocesshasseveral

significantadvantages, includingahigh rejectionof contaminantsandlow fouling propensity

comparedto pressuredrivenmicrofiltration.Therefore, FOcanconcentratetheorganic

matterandnutrientsin wastewaterto asmallvolumefor potentialintegrationwith anaerobic

treatmentto facilitateresourcerecovery. Furthermore,FOprovides robustpre-treatmentfor

reverseosmosis(Hancocket al.,2013) or membranedistillation (Xie etal., 2013) for clean

waterproduction.

Reversesoluteflux is aninherentphenomenonin FO.Whenintegrating FO with a bioreactor,

a majortechnicalchallengeis themigrationof drawsoluteinto themixedliquor. This can

severelyaffectthebiologicalperformance, particularlyof theanaerobictreatmentprocessas

inhibitory substancesareoftenthemajorcauseof instability andfailureof anaerobic

treatmentsystems(Chenetal., 2008). Inorganicsaltsarewidely usedasdrawsolutesfor FO,

sincetheyareusuallyinexpensive,capableof generatinghighosmoticpressures,andareless

likely to inducesignificantinternalconcentrationpolarization(ICP).ICPassociatedwith

inorganicsalts is smallbecauseof their smallsolutesizeandrapiddiffusion;however,these

propertiesoftenpromotea high reversesoluteflux (Shafferet al.,2015). For example,

sodiumchloridehasa high reversesoluteflux, andthereforesodiumconcentrationsarelikely

to exceedthevalueknownto inhibit anaerobictreatment(3 g Na/L) (Feijooetal., 1995)

duringwastewaterpre-concentration.

Severaldrawsoluteshavebeeninvestigatedwith theintentionof avoidingor reducingthe

effectsof reversesoluteflux on subsequentbiologicaltreatment.Lutchmiahetal. (2014)

demonstratedthatzwitterioniccompounds, suchasglycine, havea lower reversesoluteflux

comparedto sodiumchlorideandthepotentialto increasethemethaneyield of concentrated

wastewaterdueto theirosmoprotectantproperties.Bowdenet al. (2012)proposedorganic

ionic saltsassubstitutedrawsolutes in osmoticmembranebioreactors(OMBRs), whereby

saltaccumulationhasdetrimentaleffectson biologicalperformance.Otherapproaches

involve comparingthemicrobial toxicity of drawsolutions(Nawazetal., 2013) or thelong-

termoperationof alternativedrawsolutionsin OMBRsto evaluateeffects(Tang& Ng,

2014). Nonetheless,nostudieshaveevaluated thepotentialimpactof reversesoluteflux on

subsequentanaerobictreatment.This is despitetheavailabilityof thewell-established

5

biomethanepotential(BMP) test,which canbeusedto simulatetheanaerobictreatment

processin batchmodeto assessthemethaneproductionfrom differentsubstrates(Koch et al.,

2015; Mayeretal., 2014; Nghiemetal., 2014b).

In this study,adrawsoluteselectionprotocolwasdevelopedfor FOsystemswhich are

integratedwith anaerobictreatment.FO flux performancewasassessedbasedon waterflux

andreversesoluteflux. Theeffectof reversesoluteflux onanaerobictreatmentwas

evaluatedby BMP analysisof drawsolute-impactedsubstrate.

2. Materials and methods

2.1Preliminarydraw solutionselectionprotocol

A literaturereviewof previousFOstudiesto pre-concentratewastewaterwasconductedto

selecttendrawsolutionsto undergoexperimentalassessment. Firstly, organic-baseddraw

solutionsthathavedemonstrateda suitablyhighwaterflux andtheexpectationto have

negligibleimpacton anaerobictreatmentwereconsidered. Secondly,inorganicdraw

solutionswith low reversesoluteflux wereconsideredandsodiumchloridewasselectedasa

reference.OLI StreamAnalyzer(OLI Systems,Inc.,Morris Plains, New Jersey, USA) was

thenusedto simulateosmoticpressureasa functionof drawsolutionconcentration, to verify

thesuitability for furtherFOexperimental assessmentandbiologicalscreening.

2.2Materialsandchemicals

Cellulosetriacetate(CTA) membranewith embeddedpolyesterscreensupportwasacquired

from HydrationTechnologiesInnovation(HTI) (Albany,Oregon, USA). Digestedsludge

wasobtainedfrom afull -scalewastewatertreatmentplant(Wollongong,Australia)andwas

usedas inoculumfor theBMP measurements.All drawsolutesusedin this studywereof

analyticalgrade.

2.3Forward osmosissystem

FOexperimentswereconductedusinga lab-scale,cross-flow FOmembranesystem

(SupplementaryData,Figure.S1). TheFOmembranecell consistedof two symmetricflow

channelseachwith length,width, andheightof 130,95, and2 mm, respectively,andan

effectivemembraneareaof 123.5cm2.

6

Thefeedanddrawsolutionswerecirculatedby two variablespeedgearpumps(Micropump,

Vancouver, Washington,USA) at 1 L/min (correspondingto a cross-flow velocityof 9 cm/s)

andwasregulatedby two rotameters.Theworkingvolumes of thefeedanddrawsolution

reservoirswere3 and2 L, respectively.Thedrawsolutionreservoirwaspositionedona

digital balance(Mettler-ToledoInc.,Hightstown,NewJersey,USA) andweightchanges

wererecordedto determinepermeatewaterflux. For ionic drawsolutions,a reservoir

containingahighly concentratedsolutionwasalsoplacedon thedigital balanceand was

intermittentlydosedinto thedrawsolutionto maintainconstantosmoticpressure.The

conductivityof thedrawsolutionwascontinuouslymeasuredby aconductivityprobe(Cole-

Parmer,VernonHills, Illinois, USA), whichwasconnectedto a controller(controlaccuracy

of � 0.1mS/cm)andaperistalticpumpto automaticallyregulatethedrawsolution

concentration.For thecovalentorganicdrawsolution, concentrationwasmanuallycontrolled

by addingthecorrectvolumeof highly concentratedsolutionevery2 h.

2.4Forward osmosisassessment

Theflux performanceof eachdrawsolutionwasevaluatedby usingthelab-scale,cross-flow

FOsystemto determinewaterflux ( wJ ) andreversesoluteflux ( sJ ). FOexperimentswere

conductedaccordingto thestandardprocedurepreviouslydescribedby Cathet al. (2013).

Analytical gradesolutesweredissolvedin DI waterat concentrationscorrespondingto an

osmoticpressureof 30bar.This osmoticpressurewasselectedfor two reasons.Firstly,

seawaterhasanapproximateosmoticpressureof 30barandcouldbeusedasareadily

availableandinexpensiveNaClsolution.Secondly,higherosmoticpressureswerenot

investigateddueto thecorrespondingincreasein drawsoluteviscosity(particularly for

organicand/orhigh molecularweightsolutes)andtheexpectedexacerbationof internal

concentrationpolarization.Eachdrawsolutionwastestedin FOmode(activelayerfacingthe

feedsolution)with DI waterasthefeedsolution.Conductivity, pH, andtemperatureof the

feedsolutionweremonitoredhourly. For thecovalentorganicdrawsolutions, a20 mL

samplewaswithdrawnfrom thefeedsolutionevery2 h for subsequenttotal organiccarbon

(TOC) analysis.All FOexperimentswereconducted in duplicateandlastedfor at least6 h.

Reversesoluteflux selectivity(RSFS)describesthevolumeof permeatewaterpergramof

solutethathasdiffusedfrom thedrawsolutionto thefeedsolutionandcanbeexpressedas

sw JJ / . RSFSis importantfor drawsolutionselectionin termsof replenishmentcosts,yet

7

this parametercanmoreimportantlygiveanindicationof theexpectedsoluteconcentration

in FOconcentrate.Thedrawsoluteconcentrationin thepre-concentratedwastewater(Cf )

wascalculatedusingEquation1:

� � RR

JJC

swf �

��1/

1 (1)

where Jw / Js is theRSFSobservedduringtheFO performanceexperiments,and R is the

assumedFOsystemwaterrecovery.Equation1 is basedon thepremisethatflux decline(due

to membranefouling or anincreasein feedsolutionosmoticpressure)is negligibleandthat

RSFSis constant.A systemwaterrecoveryof 90%wasusedto representa ten-fold increase

in thestrengthof municipalwastewaterby FOpre-concentration.Thisconditioncanalsobe

usedto representtheworst-casescenariowith respectto theimpactof drawsoluteson

potentialanaerobictreatmentof thepre-concentratedwastewater.

2.5Biomethanepotentialapparatusandprotocol

BMP measurementswereconductedto indicatetheeffectof eachdrawsoluteon methane

productionduringanaerobicdigestion.TheBMP apparatuscouldsimultaneouslydeployup

to 16 fermentationbottles,whichweresubmergedin a waterbath(RatekInstruments,

Boronia,Victoria, Australia) andconnectedto a biogascollectiongallery(Supplementary

Data,Figure.S2). Thefermentationbottles (WiltronicsResearch, Ballarat,Victoria,

Australia) weresealedwith a rubberbungandsubmergedin thewaterbath to maintaina

temperatureof 35.0� 0.1°C. An S-shapedair lock andflexible plastictubingwereusedto

collectthebiogas.Thebiogascollectiongalleryconsistedof anarrayof inverted1000mL

plasticmeasuringcylinders,which wereinitially filled with a NaOHsolution(1 M). As

biogaswasintroducedto thecylinder,CO2 andH2Sweresequesteredby theNaOHsolution,

andtheremainingCH4 gasdisplacedthesolutioninsidethecylinder.Thevolumeof NaOH

displacedby CH4 gaswasrecordedeveryday.Furtherdetailsof theBMP testingapparatus

aregivenelsewhere(Nghiemetal., 2014b).

Equation1 wasusedto calculatetheamountof eachdrawsoluteto beaddedto thedigested

sludge, to simulatethereversesoluteflux accumulationat90%waterrecoveryfrom pre-

concentratedwastewater. Thecalculatedamountof drawsolutewasfirstly dissolvedin 50

mL of DI waterandthenmixedwith 700mL of digestedsludge.In thecontrolBMP bottles,

8

50 mL of DI waterwasaddedto thesameamount of digestedsludge. TheBMP bottlewas

purgedwith nitrogengas,sealed, andconnectedto thegascollectiongallery.All BMP

experimentsincludingthecontrolwereconductedin duplicate.Thesubstratein eachbottle

wascharacterizedbeforeandaftertheBMP experimentin termsof total solids(TS),volatile

solids(VS), pH, alkalinity, total chemicaloxygendemand(CODt), andsolublechemical

oxygendemand(CODs).

2.6Analyticalmethods

Temperature,pH, andelectricalconductivityweremeasuredusinganOrion 4-StarPlus

pH/conductivitymeter(ThermoScientific,Waltham,Massachusetts,USA). A Shimadzu

TOC analyser(TOC-VCSH) wasusedto determinethereversesoluteflux of covalentorganic

drawsolutions.

For digestedsludgecharacterisation,TS,VS, andalkalinity weremeasuredusingstandard

methods(Eatonet al.,2005). CODwasdeterminedusinga HachDBR200CODReactorand

HatchDR/2000spectrophotometer(programnumber435CODHR) following theUS-EPA

StandardMethod5220D. For CODs, thesludgesupernatantwasfiltered througha 1 µm

filter paperandthefiltrate wasthenanalysed,whilst CODt wasmeasuredby directdilution

of thehomogenisedsludge.

3. Resultsand discussion

3.1Preliminarydraw solutionselection

Tendrawsoluteswereselectedfor experimental assessmentto representa rangeof inorganic

andorganiccompounds. Sodiumchloridewasselectedasareferenceandmagnesiumsulfate

wasselecteddueto its reportedlow reversesoluteflux throughFOmembrane(towardsthe

bioreactorside), causingpotentiallyminimal impacton anaerobictreatment(Achilli et al.,

2010). Organicionic drawsolutions, namelysodiumacetate,magnesiumacetate, andsodium

formate, werealsoselecteddueto their exhibitionof acompetitivewaterflux andpotential

benefitswhencombinedwith biologicalsystems(Bowdenet al.,2012). EDTA disodiumsalt

hasbeenpreviouslystudiedby Hauet al. (2014)for theconcentrationof wastewatersludge.

Neutralorganic-baseddrawsolutes, includingglucose,glycine,glycerol, andurea, were

selectedbasedon theirmoderatewaterflux andtheanticipationfor negligibleeffects on

anaerobictreatment,independentof themagnitudeof reversesoluteflux (Yong et al.,2012).

Glycine hasrecentlybeeninvestigatedandfoundto behighly compatiblewith anaerobic

9

digestion(Lutchmiahetal., 2014). Eachdrawsolutehadpreviouslyshownpotentialfor use

asa FOdrawsolutefor wastewaterapplications.

Themolarconcentrationrequiredto generate30barof osmoticpressurevariessignificantly

betweenthetenselecteddrawsolutes(Table1). Overall,therequiredmolarconcentration

variesfrom 0.3M (EDTA disodiumsalt) to ashighas1.3M (glycine).

[Table 1]

3.2Forward osmosisflux performance

3.2.1 Waterandreversesoluteflux

Thedrawsolutions exhibitedquitediverseflux performancedespitebeingevaluatedat the

sameosmoticpressureof 30bar(Figure1). Glycerolandureacouldproduceamoderate

waterflux (3.09and1.37 L/m2h, respectively)but thereversesolutefluxeswereextremely

high (15.2and106.3 g/m2h, respectively).Thesetwo drawsolutionswereeliminatedfrom

furtheranalysisbecausethehigh reversesoluteflux would resultin excessiveaccumulation

in pre-concentratedwastewater,aswell asunsustainableFOoperation. Theremainingdraw

solutionsexhibitedawaterflux in therangeof 2.18to 4.11L/m2h. Theobservedvariationin

waterflux at thesamedrawsolutionosmoticpressurecouldbeattributedto theextentof ICP

experiencedby eachsolute(Achilli etal., 2010; Bowdenet al.,2012; Zhao& Zou,2011).

ICPdescribesthedilution of thedrawsolutionin themembranesupportlayerwhichreduces

theeffectiveosmoticdriving forceandis affectedby thedrawsolutekinetic characteristics

includingdiffusivity, viscosity, andion or moleculesize(McCutcheon& Elimelech,2006).

[Figure 1]

Drawsolutediffusivity stronglyaffectedwaterflux (Figure2a)andreversesoluteflux

(Figure2b). Waterflux waslinearlycorrelatedto diffusioncoefficientandclearly

representedtheextentof dilutive ICP for eachsolute. Soluteswith low diffusivity

experienced severeICPandweremorelikely to displaya low waterflux. On theotherhand,

highly-mobilesolutescouldreducetheeffectsof ICP, andthushada highwaterflux. This

resultis in goodagreementwith ICP theory, aswithin therelevantrange,soluteswith higher

diffusioncoefficients canproducea largerwaterflux at aconstantbulk drawsolution

osmoticpressure(McCutcheon& Elimelech,2006; Shafferet al., 2015). Theresultsalso

showthatreversesoluteflux tendedto increaseexponentiallyfor soluteswith higher

10

diffusioncoefficients(Figure2b).Thus,a trade-off existsbetweenselectinghighly diffusive

drawsolutes to maximisewaterflux andthosewhich showlow reversesoluteflux.

[Figure 2]

3.2.2 Reversesoluteflux selectivity

In termsof drawsolutionreplenishmentcostandsustainableFOoperation,ahighRSFSis

desirable.However,drawsolutionsthatexhibitedhighRSFSgenerallyhada correspondingly

low waterflux dueto theeffectsof ICP(Figure3). Forexample,magnesiumsulfatehadthe

highestRSFSof 9.01, but waterflux waslow (2.18L/m2h). Interestingly, for mostdraw

solutesinvestigatedhere,similar to thecorrelationbetweenreversesoluteflux anddiffusion

coefficientshownin Figure2b, thewaterflux alsodecreasedexponentiallyastheRSFS

increased(Figure3). Sodiumacetateandmagnesiumacetatearetheonly two exceptionsand

their flux behaviourappearedto divergefrom thetrendof theothersix drawsolutes.Both

solutesdisplayeda sufficientlyhigh waterflux (>3 L/m2h) butcouldalsodemonstrate

suitablyhighRSFSvalues.Onenoticeabledifferencein behaviourbetweenthesetwo solutes

wasthatmagnesiumacetatehada largerRSFSthanthatof sodiumacetatedueto a lower

reversesoluteflux. Thiscouldbeattributedto thelargersizeof themagnesiumcation,since

bothsolutessharethesameanion(Achilli et al., 2010). Furthermore,theuseof organicionic

drawsolutesappearedto benefitFO flux performance,particularlyin thecaseof theacetate

anion.

[Figure 3]

Theexpectedconcentrationof drawsolutewithin thepre-concentratedwastewaterwas

estimatedusingEquation1 (Table2). Sincethesimulatedconcentrationonly dependson

RSFS, soluteswith a low RSFSresultin largerconcentrations,andalternatively, high RSFS

ideally lowerstheexpectedconcentration.Nonetheless,inorganicsaltsareknownto inhibit

anaerobictreatmentevenat low concentrations(Chenetal., 2008).

[Table 2]

3.3Effectof reversedrawsoluteflux on anaerobictreatment

Eachdrawsolutehada noticeableimpacton methaneproductionoverthe25 dayobservation

period(Figure4). ThesubstratecharacteristicsbeforeandaftertheBMP experimentare

11

shownin theSupplementaryData(TableS1).Theorganic-baseddrawsolutes, namely,

glycine,glucose, andtheacetatesdisplayedhighercumulativemethaneproductioncompared

to thecontrol(noaddeddrawsolute), possiblybecausetheyarereadilybiodegradable.

Glycineoutperformedall otherdrawsolutes. This mightbeattributedto its osmoprotectant

properties, which canreduceosmoticstresscausedby inhibitory constituentspresentin the

digestedsludge(Ohetal., 2008). However,dueto thelow salineenvironment,enhanced

methaneproductionwasmostlikely a resultof therelativelyhigh concentrationof glycine

dosed(3.46g/L). Similarly, evenat a lowerconcentration(1.48g/L), glucosepromoted

methaneproductionby providingadditionalorganicsubstrate.Sodiumacetatepresenteda

similar methaneproductionto glucose,andonly slightly higherthanmagnesiumacetate.The

presenceof thesodiumor magnesiumcationappearednot to affectacetateconversion;

howeversodiumacetate(2.41g/L) wasdosedata higherconcentrationthanmagnesium

acetate(1.65g/L). Theresultssuggestthatthesedrawsoluteshavea positiveeffecton

methaneproductionandwouldbesuitablewhenintegratingFO with anaerobictreatment.

[Figure 4]

EDTA disodiumsaltandsodiumformateexhibitedasimilar cumulativemethaneproduction

to thecontrol. EDTA disodiumsaltwasexpectedto enhancemethaneproductionby

increasingthebioavailabilityof essentialelements(Vintiloiu et al.,2013); however, no

additionalmethaneproductionwasobserved, possiblybecausetheconcentrationusedin this

studywassignificantlyhigherthanthatfoundto bebeneficialby Vintiloiu etal. (2013). The

methaneproductionof sodiumformatewasstable,but occurredat aslowerratecomparedto

thatof thecontrol.Thiscouldbeattributedto thehighsodiumconcentrationof 1.8 g Na/L,

particularlywhencomparedwith sodiumacetatewhichcontainedonly 0.7 g Na/L.

Additionally, theCOD contributionof acetate(1.07g COD/g)is muchlargerthanformate

(0.34g COD/g)andwould havepromoteda fasterandmoreconsistentrateof methane

production(Grobicki & Stuckey,1989).

Inorganicdrawsoluteshadanegativeeffectonmethaneproductionovertheobservation

period. Sodiumchloridehadonly slight negativeeffects onmethaneproduction, mostlikely

causedby thedehydrationof bacterialcellsdueto osmoticpressure(Chenetal., 2008). This

inhibition observedat2.3g Na/L is slightly lower thanthe3 g Na/L reportedto betoxic to

methanogenicbacteria(Feijooet al.,1995). Therefore,thepresenceof sodiumchloridein

12

pre-concentratedwastewaterby reversedrawsoluteflux is expectedto havea smallbut

discernibleeffecton anaerobictreatment.Inhibitioncaused by magnesiumsulfateat1.06g/L

wasfoundto bemoreprominentthansodiumchloride. Thelow methaneyield observedfor

magnesiumsulfatewaslikely dueto thecompetition for substratebetweensulfatereducing

andmethaneproducingbacteria. Inhibitory concentrationsfor methanogenshavebeen

reportedto beaslow as1.4 g SO42- /L (Silesetal., 2010) which is in goodagreementwith

theresultsof this study. Despiteotheradvantagesassociatedwith theuseof inorganicsalts,

theuseof thesedrawsolutesis not recommendedwhenintegrating FOwith anaerobic

treatment,with currentFO membranes.

Independentof thedrawsolutionandreversesoluteflux, elevatedsaltconcentrationswould

beexpectedin pre-concentratedwastewaterdueto retentionby theFOmembrane.In thecase

of inorganicdrawsolutions, furtherinhibition of methaneproductionandinefficienciesin the

anaerobicdigestionprocesscouldbeexpected.For theorganicdrawsolutionsdemonstrated

to bebeneficialfor anaerobictreatment,thesalinity of thepre-concentratedwastewater

would notbesignificantlyexacerbatedby reversesoluteflux. Furthermore,significantly

higherCODconcentrationswouldbeachievedduringwastewaterpre-concentrationasa

resultof thecontributionof reversesoluteflux, allowing theopportunityto operateat a lower

concentrationfactor.

3.4Draw solutesuitability for anaerobictreatment

Sodiumacetateandmagnesiumacetateweretwo drawsolutionsthatrankedhigh in termsof

FOflux performance.Both exhibiteda slightly lower waterflux when comparedwith sodium

chloride; however, their reversesoluteflux wassignificantlylower. In FO wastewater

applications,a low reversesoluteflux is crucialfor maintainingflux sustainability,lowering

replenishmentcosts, andreducingsalinity build-up. In terms of BMP, glycinedemonstrated

significantpotentialfor anaerobictreatment.Glucose,sodiumacetate, andmagnesium

acetatewerealsosuitable, astheirpresencein pre-concentratedwastewatercould enhance

methaneproduction.Overall,sodiumacetaterankedhighly in termsof FOflux performance

andsuitability for anaerobictreatment, aswell asprovidingcostadvantagesovermagnesium

acetatein termsof specificcost(Bowdenet al.,2012).

Ionic organicdrawsoluteswerefoundto bethemostsuitableandthereforefurther

implicationsexist.Wastewaterpre-concentrationusinganorganic-baseddrawsolutionis yet

13

to bedemonstrated. Detailedinvestigationsinto flux sustainability,thepotentialaggravation

of organicfouling andcompatibilitywith reconcentrationprocessesarerequired. As

previouslymentioned, theretentionandaccumulationof feedsalinity alsorequirefurther

examination,in termsof reducingtheosmoticdriving forceandalsothecompatibilitywith

anaerobictreatment.

4. Conclusion

This studyassesseddrawsolutionflux performanceandtheimpactof reversesoluteflux on

theanaerobictreatmentof FOpre-concentratedwastewater.Theresultsshow thatorganic-

baseddrawsolutionssuchassodiumacetatearemostsuitablefor this application,dueto the

acceptableflux performanceandbenefitstowardsmethaneproduction.Theeffectsof

inorganicsaltson anaerobictreatmentwerealsodemonstrated.Thereversesoluteflux of

sodiumchlorideonly exerteda smallbutdiscernibleinhibitory effecton methaneproduction.

TheBMP testcouldbeareliablescreeningtool for assessing drawsolutioncompatibility

with anaerobicdigestion.

5. Acknowledgements

This researchwassupportedundertheAustralianResearchCouncil’sDiscoveryProject

fundingscheme(projectDP140103864).Scholarshipsupportto AshleyAnsariby the

Universityof Wollongongis gratefullyacknowledged.

6. References

[1] Achilli, A., Cath,T.Y., Childress,A.E. 2010.Selectionof inorganic-baseddrawsolutions

for forwardosmosisapplications.J.Membr.Sci.,364,233-241.

[2] Bowden,K.S.,Achilli, A., Childress,A.E. 2012.Organicionic saltdrawsolutionsfor

osmoticmembranebioreactors.Bioresour.Technol.,122,207-216.

[3] Burn,S.,Muster,T., Kaksonen,A., Tjandraatmadja,G. 2013.ResourceRecoveryfrom

Wastewater:A ResearchAgenda.WaterEnviron.Res.Found.

[4] Cath,T.Y., Elimelech,M., McCutcheon,J.R.,McGinnis,R.L., Achilli, A., Anastasio,D.,

Brady,A.R., Childress,A.E., Farr,I.V., Hancock,N.T., Lampi,J.,Nghiem,L.D., Xie, M.,

14

Yip, N.Y. 2013.StandardMethodologyfor EvaluatingMembranePerformancein

OsmoticallyDrivenMembraneProcesses.Desalination,312,31-38.

[5] Chen,Y., Cheng,J.J.,Creamer,K.S.2008.Inhibition of anaerobicdigestionprocess:A

review.Bioresour.Technol.,99,4044-4064.

[6] Eaton,A.D., Franson,M.A.H., Association,A.P.H.,Association,A.W.W., Federation,

W.E.2005.StandardMethodsfor theExaminationof Water& Wastewater.AmericanPublic

HealthAssociation.

[7] Feijoo,G., Soto,M., Méndez,R., Lema,J.M. 1995.Sodiuminhibition in theanaerobic

digestionprocess:Antagonismandadaptationphenomena.EnzymeMicrob. Technol.,17,

180-188.

[8] Garcia-Belinchón,C., Garcia-Belinchón,C., Rieck,T., Bouchy,L., Galí,A. 2013.

Struviterecovery:Pilot-scaleresultsandeconomicassessmentof differentscenarios.Water

PracticeTechnol.,8, 119-130.

[9] Grobicki, A., Stuckey,D.C. 1989.Therole of formatein theanaerobicbaffledreactor.

WaterRes.,23,1599-1602.

[10] Hancock,N.T., Xu, P.,Roby,M.J.,Gomez,J.D.,Cath,T.Y. 2013.Towardsdirect

potablereusewith forwardosmosis:Technicalassessmentof long-termprocessperformance

at thepilot scale.J.Membr.Sci.,445,34-46.

[11] Hau,N.T., Chen,S.-S.,Nguyen,N.C.,Huang,K.Z., Ngo,H.H., Guo,W. 2014.

Explorationof EDTA sodiumsaltasnoveldrawsolutionin forwardosmosisprocessfor

dewateringof high nutrientsludge.J.Membr.Sci.,455,305-311.

[12] Iranpour,R., Stenstrom,M.K., Tchobanoglous,G., Miller, D., Wright, J.,Vossoughi,M.

1999.Environmentalengineering:Energyvalueof replacingwastedisposalwith resource

recovery.Science,285,706-711.

[13] Koch,K., Helmreich,B., Drewes,J.E.2015.Co-digestionof foodwastein municipal

wastewatertreatmentplants:Effectof differentmixturesonmethaneyield andhydrolysis

rateconstant.Appl. Energ.,137,250-255.

15

[14] Koppelaar,R.H.E.M.,Weikard,H.P.2013.Assessingphosphaterockdepletionand

phosphorusrecyclingoptions.GlobalEnviron.Chang.,23,1454-1466.

[15] Lutchmiah,K., Cornelissen,E.R.,Harmsen,D.J.H.,Post,J.W.,Lampi,K., Ramaekers,

H., Rietveld,L.C.,Roest,K. 2011.Waterrecoveryfrom sewageusingforwardosmosis.

WaterSci.Technol.,64,1443-1449.

[16] Lutchmiah,K., Lauber,L., Roest,K., Harmsen,D.J.H.,Post,J.W.,Rietveld,L.C., van

Lier, J.B.,Cornelissen,E.R.2014.Zwitterionsasalternativedrawsolutionsin forward

osmosisfor applicationin wastewaterreclamation.J.Membr.Sci.,460,82-90.

[17] Mayer,F.,Gerin,P.A.,Noo,A., Foucart,G., Flammang,J.,Lemaigre,S.,Sinnaeve,G.,

Dardenne,P.,Delfosse,P.2014.Assessmentof factorsinfluencingthebiomethaneyield of

maizesilages.Bioresour.Technol.,153,260-268.

[18] McCarty,P.L.,Bae,J.,Kim, J.2011.DomesticWastewaterTreatmentasaNet Energy

Producer-CanThisbeAchieved?Environ.Sci.Technol.,45,7100-7106.

[19] McCutcheon,J.R.,Elimelech,M. 2006.Influenceof concentrativeanddilutive internal

concentrationpolarizationon flux behaviorin forwardosmosis.J.Membr.Sci.,284,237-

247.

[20] Nawaz,M.S.,Gadelha,G.,Khan,S.J.,Hankins,N. 2013.Microbial toxicity effectsof

reversetransporteddrawsolutein theforwardosmosismembranebioreactor(FO-MBR). J.

Membr.Sci.,429,323-329.

[21] Nghiem,L.D., Manassa,P.,Dawson,M., Fitzgerald,S.K. 2014a.Oxidationreduction

potentialasaparameterto regulatemicro-oxygeninjectioninto anaerobicdigesterfor

reducinghydrogensulphideconcentrationin biogas.Bioresour.Technol.,173,443-447.

[22] Nghiem,L.D., Nguyen,T.T., Manassa,P.,Fitzgerald,S.K.,Dawson,M., Vierboom,S.

2014b.Co-digestionof sewagesludgeandcrudeglycerol for on-demandbiogasproduction.

Int. Biodeterior.Biodegradation.

[23] Oh,G.,Zhang,L., Jahng,D. 2008.Osmoprotectantsenhancemethaneproductionfrom

theanaerobicdigestionof foodwastescontainingahigh contentof salt.J.Chem.Technol.

Biotechnol.,83,1204-1210.

16

[24] Shaffer,D.L., Werber,J.R.,Jaramillo,H., Lin, S.,Elimelech,M. 2015.Forward

osmosis:Wherearewe now?Desalination,356,271-284.

[25] Shannon,M.A., Bohn,P.W.,Elimelech,M., Georgiadis,J.G.,Marinas,B.J.,Mayes,

A.M. 2008.Scienceandtechnologyfor waterpurificationin thecomingdecades.Nature,

452,301-310.

[26] Siles,J.A.,Brekelmans,J.,Martín,M.A., Chica,A.F., Martín,A. 2010.Impactof

ammoniaandsulphateconcentrationon thermophilicanaerobicdigestion.Bioresour.

Technol.,101,9040-9048.

[27] Tang,M.K.Y., Ng, H.Y. 2014.Impactsof differentdrawsolutionson a novelanaerobic

forwardosmosismembranebioreactor(AnFOMBR).WaterSci.Technol.,69,2036-2042.

[28] Verstraete,W., Vlaeminck,S.E.2011.ZeroWasteWater:short-cyclingof wastewater

resourcesfor sustainablecitiesof thefuture.Int. J.Sust.Dev.World, 18,253-264.

[29] Vintiloiu, A., Boxriker,M., Lemmer,A., Oechsner,H., Jungbluth,T., Mathies,E.,

Ramhold,D. 2013.Effectof ethylenediaminetetraaceticacid(EDTA) on thebioavailability

of traceelementsduringanaerobicdigestion.Chem.Eng.J.,223,436-441.

[30] Xie, M., Nghiem,L.D., Price,W.E.,Elimelech,M. 2013.A ForwardOsmosis-

MembraneDistillation Hybrid Processfor Direct SewerMining: SystemPerformanceand

Limitations.Environ.Sci.Technol.,47,13486-13493.

[31] Xie, M., Nghiem,L.D., Price,W.E.,Elimelech,M. 2014.TowardResourceRecovery

from Wastewater:Extractionof Phosphorusfrom DigestedSludgeUsinga Hybrid Forward

Osmosis–MembraneDistillation Process.Environ.Sci.Technol.Lett., 1, 191-195.

[32] Yong,J.S.,Phillip, W.A., Elimelech,M. 2012.Coupledreversedrawsolutepermeation

andwaterflux in forwardosmosiswith neutraldrawsolutes.J.Membr.Sci.,392–393,9-17.

[33] Zhao,S.A.F.,Zou,L.D. 2011.Relatingsolutionphysicochemicalpropertiesto internal

concentrationpolarizationin forwardosmosis.J.Membr.Sci.,379,459-467.

17

List of Tables

Table 1: Molar concentrationrequiredto generate30barof osmoticpressureandsolute

diffusioncoefficients. ConcentrationswerecalculatedusingOLI StreamAnalyzer.

Draw solutesConcentration

(M)

Diffusioncoefficient

(m2/s)Reference

InorganicSodiumchloride 0.65 91047.1 �� (Achilli etal., 2010)Magnesiumsulfate 1.24 10107.3 �� (Achilli etal., 2010)

Organic(ionic)

Sodiumacetate 0.72 91044.1 �� (Bowdenetal., 2012)Magnesiumacetate 0.84 91014.1 �� (Bowdenetal., 2012)Sodiumformate 0.72 91059.1 �� (Bowdenetal., 2012)EDTA disodiumsalt 0.30 101083.5 �� (Lide & Kehiaian,1994)

Organic(covalent)

Glucose 1.13 10107.6 �� (Yong etal., 2012)Glycine 1.31 91006.1 �� (Lutchmiahetal., 2014)Glycerol 1.27 10103.9 �� (Hayduk& Laudie,1974)Urea 1.26 91038.1 �� (Hayduk& Laudie,1974)

Table 2: Expecteddrawsoluteconcentration(by reversesoluteflux) in pre-concentrated

wastewaterassuming90%FOsystemwaterrecovery.

Draw solute Concentration (g/L)Sodiumchloride 5.78Sodiumformate 5.45Glycine 3.46Sodiumacetate 2.41Magnesiumacetate 1.65EDTA disodiumsalt 1.52Glucose 1.48Magnesiumsulfate 1.06

18

List of Figure Captions

Figure 1: Averagewaterflux, reversesoluteflux andRSFSatanosmoticpressureof 30bar.

Errorbarsrepresentthestandarddeviationof duplicateexperiments.

Figure 2: Relationshipbetweendiffusioncoefficientand(a)waterflux (linearregressionR2

= 0.90); (b) reversesoluteflux (exponentialregressionR2 = 0.81). Experimentswere

conductedat constantosmoticpressure(30bar);errorbarsrepresentthestandarddeviationof

duplicatetests.

Figure 3: Variationof waterflux with RSFS.Experimentswereconductedatconstant

osmoticpressure(30bar);errorbarsrepresentthestandarddeviationof duplicatetests.

Figure 4: Cumulativemethaneproductionof digestedsludgewith doseddrawsoluteat

concentrationscorrespondingto FO RSFSand90%systemrecovery;errorbarsrepresentthe

standarddeviationof duplicateexperiments.

19

Mag

nesiu

msu

lfate

Glucos

e

EDTAdis

odium

salt

Mag

nesiu

mac

etat

e

Sodium

acet

ate

Glyc

ine

Sodium

form

ate

Sodium

chlor

ide

0

1

2

3

4

5

6

WaterFlux(L/m

2 h)

ReverseSoluteFlux(g/m

2 h)

ReverseSoluteFluxSelectivity

(RSFS)(L/g)

Water Flux (Jw)Reverse Solute Flux (Js)

0

2

4

6

8

10

12

RSFS (Jw/Js)

Figure 1

20

5.0x10-10 1.0x10-9 1.5x10-90

1

2

3

4

5

6(b)(a)

WaterFlux(L/m

2 h)

Diffusion Coefficient (m2/s)

5.0x10-10 1.0x10-9 1.5x10-90

1

2

3

4

ReverseSoluteFlux(g/m

2 h)

Diffusion Coefficient (m2/s)

Figure 2

21

0 2 4 6 8 100

1

2

3

4

5

Sodium acetateMagnesium acetateOther DS

Exponential regression (R2=1.00)WaterFlux(L/m

2 h)

RSFS (Jw/Js) (L/g)

Figure 3

22

0 5 10 15 20 250

500

1000

1500

Cumulativemethaneproduction(mLCH

4)

Time (d)

ControlGlycine (3.46 g/L)Glucose (1.48 g/L)Sodium acetate (2.41 g/L)Magnesiumacetate (1.65 g/L)EDTA disodium salt (1.52 g/L)Sodium formate (5.45 g/L)Sodium chloride (5.78 g/L)Magnesiumsulfate (1.06 g/L)

Figure 4