Membrane bioreactor (MBR) for wastewater treatment

Separation and Purification Technology 71 (2010) 200–204Contents lists available at ScienceDirectSeparation and PurificationTechnologyj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /s e p p urMembrane bioreactor (MBR)for wastewater treatment:Filtration performance evaluation of low cost polymeric and ceramic membranesP.K.Tewari,R.K.Singh,V.S.Batra,M.Balakrishnan ∗The Energy &Resources Institute (TERI),Darbari Seth Block,India Habitat Center,Lodhi Road,New Delhi 110003,Indiaa r t i c l e i n f o Article history:Received 25September 2009Received in revised form 20November 2009Accepted 23November 2009Keywords:Membrane bioreactor Wastewater treatment Nylon mesh Ash filter Critical fluxa b s t r a c tThis work compares the filtration performance of low cost filters for wastewater treatment in a sub-merged MBR.Two filters mercial nylon mesh of 30␮m pore size and 2–6␮m macroporous ceramic filters prepared from bagasse fly ash were investigated.Critical flux was measured by flux stepping method,and the MBR filtration performance was monitored in terms of flux,transmembrane pressure (TMP)build-up and permeate quality.It was observed that both the nylon mesh and the ash filter are good solid–liquid separators with low TMP build-up.However,the ash filter displayed higher critical flux that was comparable to commercial microporous membranes;it also does not require the formation of a secondary filtration layer to ensure consistently high suspended solids retention.© 2009 Elsevier B.V. All rights reserved.1.IntroductionMembrane bioreactors (MBRs)have emerged as one of the major treatment solutions for domestic wastewater,with over 1500[1]installations operational worldwide.The development and adop-tion of this technology is being driven by both fresh water shortage and anticipated stricter environmental regulations.MBRs offer the advantages of a compact system with better solids removal,disin-fection and an overall superior quality of treated effluent.The main limitations are the high capital investment as well as membrane fouling that lowers productivity and requires eventual membrane replacement.Thus,in the context of developing countries,MBRs continue to be an expensive option.To overcome the limitations of high cost of microporous mem-branes,several workers have explored alternatives such as woven fabric and mesh in submerged MBRs [2–4].Except for a lower total coliform reduction,comparable performance has been reported with different pore size non-woven polypropylene (NWPP)mem-branes and polysulfone (PS)membrane in a submerged MBR treating domestic wastewater [5].Thus the NWPP membranes can be considered where disinfection is not required.Similar results have been obtained in a pilot test employing 20␮m non-woven fabric to treat an influent with chemical oxygen demand (COD)between 800and 1800mg/L [3].The transmembrane pressure was maintained below 5kPa at a permeation flux of 0.18m 3/m 2d and∗Corresponding author.Tel.:+911124682100/41504900;fax:+911124682144/2145.E-mail address:malinib@teri.res.in (M.Balakrishnan).stable operation was possible.Nylon and dacron meshes have been investigated in laboratory scale MBRs for treating both synthetic and actual municipal wastewater [4,6];also plasma surface mod-ified non-woven polypropylene in the pore size 13–38.8␮m have been successfully tested for synthetic wastewater treatment [7].In comparison to polymeric membranes,ceramic (inorganic)membranes offer better chemical and physical stability,enabling longer life and more rugged cleaning methods.Since ceramic membranes are considerably more expensive in comparison to polymeric membranes,low cost MF/UF ceramic membranes have been prepared from naturally occurring materials such as clays [8,9],mixed oxides like cordierite and from waste materials like bagasse fly ash [10].UF membranes of 3kD cutoff prepared using cordierite macroporous supports can effectively reject dyes like methylene blue and heavy metal salts containing Cr(III),Pb(II),Cd(II)[11].Similar properties have been observed with UF mem-branes of 4.5kD cutoff made from low cost mineral support of Moroccan clay [12].The aim of this study was to compare the filtration perfor-mance of two low cost filters viz.a commercial polymeric mesh and a ceramic filter prepared in-house from bagasse fly ash in a submerged aerobic MBR.2.Material and methods 2.1.FiltersNylon mesh (pore size 30␮m)was obtained commercially (Trace Scientific Corporation,Uttar Pradesh,India).The ceramic1383-5866/$–see front matter © 2009 Elsevier B.V. All rights reserved.doi:10.1016/j.seppur.2009.11.022P.K.Tewari et al./Separation and Purification Technology71 (2010) 200–204201Fig.1.Schematic of the experimental setup.filters(pore size2–6␮m)were prepared in-house by tape-casting and sintering[10].The raw material for ceramicfilters wasfly ash. This is a waste material generated by the combustion of bagasse, which is the solid residue obtained after crushing sugarcane.The fly ash was obtained from a sugar factory in Northern India.2.2.Experimental setup and reactor operationThe schematic of the experimental setup is shown in Fig.1.The bioreactor consisted of an acrylic tank with working volume of 12L equipped with an air diffuser(Southern Cogen Systems Pvt. Ltd.,Chennai,India),located under thefilter module,and a sludge drain.The submergedflat sheet typefilter module had an effec-tivefiltration area of0.05m2.The inoculum was activated sludge obtained from the local municipal sewage treatment plant.The starting MLSS concentration was0.8g/L and it was allowed to build up to10–12g/L over the course of the operation.The aeration rate was between2.6and4.5L min−1and the dissolved oxygen concentration was in the range of2–5mg/L.The reactor tempera-ture was in the25–28◦C range as per the controlled air-conditions maintained in the laboratory.The reactor was operated with syn-thetic wastewater,prepared according to the composition reported by Shim et al.[13](Table1).The wastewater was fed through a peristaltic pump(Enertech ENPD-100Optima,Pragati Biomedical, Mumbai,India)synchronized to permeate suction.The permeate was withdrawn intermittently through a second identical peri-Table1Composition of synthetic wastewater[13].Chemical Concentration(g/L)Glucose 1.08Glutamic acid0.46CH3COONH40.35NaHCO33NH4Cl0.053KH2PO40.06K2HPO40.08MgSO4·7H2O0.033MnSO4·H2O0.01FeCl3·6H2O0.002CaCl2·2H2O0.02NaCl0.025staltic pump in a10min cycle(8.35min on and1.65min off).A pressure gauge(Japsin Industrial Instrumentation,Delhi,India)was mounted between thefilter and permeate pump to monitor the transmembrane pressure(TMP).The reactor was operated at a pre-setflow rate(corresponding to afixed hydraulic retention time) that was controlled by adjusting the rotation speed of the peristaltic pumps.Theflux was measured volumetrically in a measuring cylin-der by collecting the permeate over a known time period.Filter cleaning,if required to maintain the imposed permeateflux,was performed by rinsing with tap water.A limited number of chemi-cal cleaning cycles,with0.1–0.5N NaOH were also done.The critical flux was determined by increasing theflux in a step-wise fashion while monitoring the corresponding TMP[14].2.3.Analytical methodsCOD,MLSS and volatile suspended solids(VSS)were determined according to Standard Methods[15].Dissolved oxygen was mea-sured using DO meter(OXI330i/SET,Germany).Bound and soluble extracellular polymeric substances(EPS)were measured according to the method described by Chang and Lee[16].Protein content was determined by Lowry method with bovine serum albumin as the standard[17].The phenol–sulfuric acid method of Dubois et al.[18]was used for polysaccharide determination using glucose as a standard.The particle size of the sludge was measured using Sympatec HELOS BF(Germany)particle size analyzer.The sludge samples were centrifuged and dried in a desiccator prior to analysis.3.Results and discussion3.1.CriticalfluxOperation of MBRs under sub-criticalflux conditions is preferred since the rate of fouling below criticalflux is much more sustainable [19,20].In addition to feed properties(e.g.particles concentration, size distribution),hydrodynamic conditions(e.g.aeration rate), criticalflux value depends upon membrane characteristics(e.g. material,pore size,new/used membrane).Fig.2a shows the results obtained for the nylon mesh and ashfil-ter with different imposedflux.The experiments were conducted at a high MLSS(9.2–9.3g/L).Each run started at a low constant flux,which was maintained for15–30min.If there was no obvi-ous increase in the TMP,the constantflux was stepped up to the next level by changing the suction pump speed.The procedure was repeated till a significant variation in TMP was observed.For both thefilters,at low imposedflux,the increase in TMP was small. For the nylon mesh,significant change in TMP occurred just below 10L/m2h(LMH).In contrast,the rate of TMP increase was still low for the ashfilter at16.7LMH.Thus the ashfilter displayed a rel-atively higher criticalflux(>16LMH)under these conditions.It is established that membranes with larger pore size sieve more par-ticles initially causing quick and irreversible fouling[21,22].This appears to be the cause for the quicker increase in TMP for the nylon mesh in Fig.2a.However,operation at a lower MLSS(1.6g/L) significantly improved the criticalflux,as shown for the nylon mesh (criticalflux of50LMH)(Fig.2b).The criticalflux values obtained in this work is well within the 10–50LMH range reported in the literature for submerged MBRs treating synthetic/municipal wastewater(Table2).The variation of criticalflux with MLSS is in agreement with the observation of Madaeni et al.[26]who concluded that criticalflux was inversely related to MLSS concentration in the0–10g/L range.This is how-ever not a universal observation and other workers have reported contrastingfindings.For instance,Rosenberger and Kraume[27] reported that MLSS concentrations between2and24g/L had little202P.K.Tewari et al./Separation and Purification Technology71 (2010) 200–204Fig.2.Criticalflux.(a)Criticalflux for ash and nylon meshfilter(MLSS9.2–9.3g/L);(b)Nylon mesh at different MLSS.influence onfilterability;similarly Le-Clech et al.[14]observed no change in criticalflux between4and8g/L MLSS but there was a significant increase in criticalflux when the MLSS concentration was increased to12g/L.3.2.Flux and TMP profilesFig.3presents a comparison of theflux profiles at two different imposedfluxes viz.12LMH and8LMH.The experiments at12LMH were performed at medium to high MLSS values(4.7–9.7g/L for nylon mesh and4.4–8.9g/L for ashfilter).It was observed that the ashfilter exhibited stableflux at12LMH over the entire duration of20days.Nofilter cleaning was done during this period.In con-trast,theflux with the nylon mesh varied in the range from4.8 to15LMH over a24-h period and daily washing of the mesh was required to maintain the averageflux at12LMH.At a lower imposed flux of8LMH,both the ashfilter and the nylon mesh displayed no perceptibleflux decline over the20day period.These experiments were conducted at high MLSS value(9–10.9g/L for nylon mesh and 9–10.5g/L for ashfilter)parison offlux profiles for mesh and ashfilters(MLSS range:9–10.9at 8LMH and4.4–9.7at12LMH).Fig.4.Long-term testing of nylon mesh above criticalflux conditions(MLSS range 0.8–10.2g/L).Thefindings are consistent with the criticalflux values(Sec-tion3.1).For the ashfilter(criticalflux>16LMH),both the imposed fluxes were in the sub-critical range;this is expected to lead to lit-tle or no irreversible fouling,contributing towards a uniformflux. For the nylon mesh,with a criticalflux of around10LMH in this MLSS range,operation above the criticalflux(12LMH)resulted in noticeable fouling and a consequentflux decline.It was difficult to maintain theflux at the set value in this case.Fig.4presents the operation with the nylon mesh over a period of50d,wherein the MLSS build-up occurred from0.8to10.2g/L.In spite of daily mem-brane washing(rinsing with tap water to dislodge the particulate deposits on the membrane)and chemical cleaning with0.1–0.5N NaOH between10and20days operation,it was not possible to maintain theflux at12LMH.It was hypothesized that mesh fouling was caused by the extracellular polymeric substances(EPS),which is a known foulant in MBR operation.The soluble EPS content wasTable2Criticalflux in submerged MBRs treating municipal wastewater.Membrane Wastewater MLSS(g/L)Criticalflux(LMH)Reference0.04␮m hollowfibre,Zenon ZeeWeed®500Municipal wastewater10–2115–30[19]0.2␮m tubular membrane module,Millenniumpore Ltd.Synthetic and real sewage310(synthetic feed)18(real feed)[20]0.038␮m polyethersulphoneflat sheet,Huber Municipal(80%domestic+20%leachate+winery wastewater)7.728[23]0.4␮m hollowfiber Zenon ZeeWee TM-10Municipal wastewater(almostcompletely domestic)5and850(8g/L MLSS)[24]0.4␮mflat sheet,Kubota Simulated sewage 6.8–21.1(MLVSS)23(MLVSS17.15g/L)[25]P.K.Tewari et al./Separation and Purification Technology71 (2010) 200–204203parison of TMP profiles for mesh and ashfilters(Imposedflux8LMH; MLSS content9–10.9g/L).significantly higher(average value of172mg VSS/g MLSS over50d operation)as compared to the bound component(average of89mg VSS/g MLSS).Further,the protein component in the soluble EPS was up to9-fold higher than the carbohydrate component.Non-specific protein binding is known to occur with nylon[28]and this could be the reason for mesh fouling.The variation in transmembrane pressure(TMP)over the test duration is presented in Fig.5.The tighter ashfilter displayed a steady TMP increase(2.8–7kPa)whereas the TMP for the nylon meshfilter remained nearly constant(2.3–4kPa).However,the absolute value in both cases was low(below8kPa).Thefiltration resistance calculated using the resistance-in-series model[29]viz.1.09–2.24×1012/m for the nylon mesh and1.61–4.39×1012/m for the ashfilter,were similar to the resistance values reported in the literature for mesh[7]and ceramicfilter[30].3.3.Permeate qualityThe biomass content(in terms of MLSS)along with the per-centage retention is plotted in Fig.6.The volatile suspended solids(VSS)content was also monitored.VSS was observed to be fairly stabilized,accounting for around70%of the total MLSS,which is within the reported limits of70–90%in bio-logical treatment[31];correspondingly,the COD removal was high(>90%).It is to be noted that the mean cell retention time (MCRT)(calculated on the basis of MLSS retention)was not fixed due to regular loss of solids in the permeate as discussed below.The approximate MCRT,calculated by taking into account the average solids loss during the study,was17days for the MBR using meshfilter and36days for the MBR using ashfil-ter.From Fig.6,it is evident that complete solids rejection did not occur with eitherfilter.The volume mean diameter(VMD)of the sludge(136␮m and155␮m for two separate samples taken on38 days and81days of operation respectively)was significantly larger than thefilter pore size(30␮m for the nylon mesh and2–6␮m for the ashfilter).Even so,because of the broad particle size distribu-tion(Fig.7),thefilters allowed the passage of thefine particulate matter.With the nylon mesh,average suspended solids retention of92%was observed over a20day period with a MLSS concen-tration between4.7and9.4g/L.For a similar MLSS concentration range(4.4–8.9g/L),the suspended solids retention with the ash filter was marginally higher at around96%.Furthermore,the ashfil-ter consistently displayed a high retention throughout.In contrast, with a fresh(unused)nylon mesh,in the initial days of operation at low MLSS significant biomass was found to pass into thepermeateFig.6.MLSS growth and retention by mesh and ashfilters.(Fig.8).The rejection increased to80%and above only after10days of operation;also,there were occasionalfluctuations.The average retention over a60day period was around88%.In mesh MBRs, the cake layer formed on the mesh surface(secondaryfilter forma-tion due to the adherence of microbialflocs on the mesh surface) controls the suspended solids rejection;this layer is more pro-nounced at higher suspended solids content.Rejection increases with cake layer build up and once the cake layer is balanced,the rejection becomes constant and can reach100%[32].Thus,high MLSS concentration leads to better SS retention and effluent quality as observed in the latter part of Fig.8.Fig.7.Particle size distribution of sludge.204P.K.Tewari et al./Separation and Purification Technology71 (2010) 200–204Fig.8.Long-term retention of suspended solids with meshfilter.4.ConclusionsThis work compares thefiltration performance of two low cost filters viz.a30␮m nylon mesh and2–6␮m bagassefly ashfilter for wastewater treatment in a submerged MBR.Under the condi-tions of this study,bothfilters demonstrate low TMP buildup,with comparablefiltration resistance.The ashfilter has better proper-ties in terms of a higher criticalflux;also,it does not require the formation of a secondaryfiltration layer to ensure high suspended solids retention.Further investigation on the long-term operation characteristics of thesefilters on treatment of domestic sewage is currently underway.AcknowledgementsThe authors gratefully acknowledge thefinancial support from TERI to conduct this work.The authors also thank Ms N.Ahuja for the EPS analysis and Mr.B.S.Bisht for laboratory assistance during the period of this study.References[1]W.Yang,N.Cicek,J.Ilg,State-of-the-art of membrane bioreactors:Worldwideresearch and commercial applications in North America,J.Membr.Sci.270 (2006)201–211.[2]Y.Kiso,Y.Jung,T.Ichinari,T.Kitao,K.Nishimura,K.Min,Wastewater treatmentperformance of afiltration bio-reactor equipped with a mesh as afilter material, Water Res.34(2000)4143–4150.[3]W.K.Chang,A.Y.J.Hu,R.Y.Horng,W.Y.Tzou,Membrane bioreactor with non-woven fabrics as solid–liquid separation media for wastewater treatment, Desalination202(2007)122–128.[4]Y.Wu,X.Huang,W.Zuo,Effect of mesh pore size on performance of aself-forming dynamic membrane coupled bioreactor for domestic wastewater treatment,in:5th International Membrane Science and Technology Confer-ence,Sydney,November11–13,2003.[5]I.S.Chang,M.Gander,B.Jefferson,S.J.Judd,Low-cost membranes for use in asubmerged MBR,Trans.IChemE(Part B).79(2001)183–188.[6]Y.Wu,X.Huang,W.Zuo,Effect of mesh pore size on performance of self-forming dynamic membrane coupled bioreactor for domestic wastewater treatment,future of urban wastewater systems—decentralization and reuse, in:IWA International Conference,Xi’an China,May18–20,2005.[7]M.C.Chang,R.Y.Horng,H.Shao,Y.J.Hu,Performance andfiltration characteris-tics of non-woven membranes used in a submerged membrane bioreactor for synthetic wastewater treatment,Desalination191(2006)8–15.[8]Q.K.Wang,T.Matsuura,C.Y.Feng,M.R.Weir,C.Detellier,E.Rutadinka,R.L.Van Mao,The sepiolite membrane for ultrafiltration,J.Membr.Sci.184(2001) 153–163.[9]S.Rakib,M.Sghyar,M.Rafiq,rbot,L.Cot,New porous ceramics for tangen-tialfiltration,Sep.Purif.Technol.25(2001)385–390.[10]V.S.Batra,P.K.Tewari,A process for the preparation of inorganic membranefilters for solid–liquid and solid–gas separations,Indian patent applied(2006).[11]N.Saffaj,S.A.Younsi,A.Albizane,A.Messouadi,M.Bouhria,M.Persin,M.Cretin,rbot,Elaboration and properties of TiO2–ZnAl2O4ultrafiltration membranes deposited on cordierite support,Sep.Purif.Technol.36(2004) 107–114.[12]N.Saffaj,S.A.Younsi,M.Persin,M.Cretin,A.Albizane,rbot,Processingand characterization of TiO2/ZnAl2O4ultrafiltration membranes deposited on tubular support prepared from Moroccan clay,Ceram.Int.31(2005)205–210.[13]J.K.Shim,I.K.Yoo,Y.M.Lee,Design and operation considerations for wastewatertreatment using aflat submerged membrane bioreactor,Process Biochem.38 (2002)279–285.[14]P.Le.Clech,B.Jefferson,I.S.Chang,S.J.Judd,Criticalflux determination by theflux-step method in a submerged membrane bioreactor,J.Membr.Sci.227 (2003)81–93.[15]L.S.Clesceri,A.E.Greenberg,A.D.Eaton,Standard Methods for the Examina-tion of Water and Wastewater,20th ed.,American Public Health Association, Washington,1998.[16]I.S.Chang,C.H.Lee,Membranefiltration characteristics in membrane coupledactivated sludge system—the effect of physiological states of activated sludge on membrane fouling,Desalination120(1998)221–233.[17]O.H.Lowry,N.J.Rosebrough,L.Farr,R.J.Randall,Protein measurement with thefolin phenol reagent,J.Biol.Chem.193(1951)265–275.[18]M.Dubois,K.A.Gilles,J.K.Hamilton,P.A.Rebers,F.Smith,Colorimetric methodfor determination of sugars and related substances,Anal.Chem.28(1956) 350–356.[19]F.Fan,H.Zhou,H.Husain,Identification of wastewater sludge characteristics topredict criticalflux for membrane bioreactor processes,Water Res.40(2006) 205–212.[20]P.Le-Clech,B.Jefferson,S.J.Judd,Impact of aeration,solids concentration andmembrane characteristics on the hydraulic performance of a membrane biore-actor,J.Membr.Sci.218(2003)117–129.[21]P.Le-Clech,V.Chen,T.A.G.Fane,Fouling in membrane bioreactor used inwastewater treatment,J.Membr.Sci.284(2006)17–53.[22]F.Meng,S.R.Chae,A.Drews,M.Kraume,H.S.Shin,F.Yang,Recent advancesin membrane bioreactors(MBRs):Membrane fouling and membrane material, Water Res.43(2009)1489–1512.[23]G.Guglielmi,D.P.Saroj,D.Chiarani,G.Andreottola,Sub-critical fouling ina membrane bioreactor for municipal wastewater treatment:Experimentalinvestigation and mathematical modeling,Water Res.41(2007)3903–3914.[24]M.Matoˇs i´c,M.Vukovi´c,M.ˇCurlin,I.Mijatovi´c,Fouling of a hollowfibre sub-merged membrane during long termfiltration of activated sludge,Desalination 219(2008)57–65.[25]J.A.Howell,H.C.Chua,T.C.Arnot,In situ manipulation of criticalflux in asubmerged membrane bioreactor using variable aeration rates,and effects of membrane history,J.Membr.Sci.242(2004)13–19.[26]S.S.Madaeni,A.G.Fane,D.E.Wiley,Factors influencing criticalflux in membranefiltration of activated sludge,J.Chem.Technol.Biotechnol.74(1999)539–543.[27]S.Rosenberger,M.Kraume,Filterability of activated sludge in membrane biore-actor,Desalination146(2002)373–379.[28]A.M.Pitt,The nonspecific protein binding of polymeric microporous mem-branes,J.Parenter.Sci.Technol.41(1987)110–113.[29]I.S.Chang,S.N.Kim,Wastewater treatment using membranefiltration—effectof biosolids concentration on cake resistance,Process Biochem.40(2005) 1307–1314.[30]L.Defrance,M.Y.Jaffrin,B.Gupta,P.Paullier,V.Geaugey,Contribution of variousconstituents of activated sludge to membrane bioreactor fouling,Bioresour.Technol.73(2000)105–112.[31]S.Rosenberger,U.Kruger,R.Witzig,W.Manz,U.Szewzyk,M.Kraume,Perfor-mance of a bioreactor with submerged membranes for aerobic treatment of municipal wastewater,Water Res.36(2002)413–420.[32]B.Fan,X.Huang,Characteristics of a self-forming dynamic membrane coupledwith a bioreactor for municipal wastewater treatment,Environ.Sci.Technol.36(2002)5245–5251.。