Biological nitrogen removal using a vertically moving biofilm system
- 格式：pdf
- 大小：275.01 KB
- 文档页数：7
Biological nitrogen removal using a vertically moving biofilm systemMichael Rodgers,Xin-Min Zhan*Department of Civil Engineering,National University of Ireland,Galway,IrelandReceived 15July 2003;received in revised form 13August 2003;accepted 10September 2003AbstractIn this study,a biological nitrogen removal process using a vertically moving biofilm system was used to treat synthetic wastewater.The process consisted of two pre-denitrification units,one combined carbonaceous removal/nitrification unit and three nitrification units.Each unit employed biofilm growth on a plastic module.In the anoxic units,the modules were vertically moved,while always submerged,in the bulk fluid;in the aerobic units,they were moved vertically up into the air and down into the wastewater.Three small-scale experiments,having different recirculation ratios and influent loadings,were conducted at a controlled temperature of 11°C.In this system,the carbonaceous removal efficiency was in the range of 94–96%and the total nitrogen removal efficiency was 77–82%.In the anoxic units,the denitrification efficiency was 94–98%and the areal denitrification rates,based on the surface area of the biofilm modules,were 2.9–3.8g NO 3-N/(m 2Æd).The nitrification efficiency occurring in the aerobic tanks was up to 95%and the maximum areal ammonium removal rates were 1.3–1.8g NH 4-N/(m 2Æd).Ó2003Elsevier Ltd.All rights reserved.Keywords:Biological nitrogen removal (BNR);Denitrification;Nitrification;Vertically moving biofilm system;Wastewater treatment1.IntroductionHigh concentrations of nitrates in water supplies have led to cases of infant methaemoglobinaemia.Nitrates can also contribute to the development of eutrophica-tion in receiving water bodies.The European Council Urban Wastewater Directive (91/271/EEC)gives an impetus to reduce nitrates in wastewaters and the European Council Nitrate Directive (91/676/EEC)aims to reduce nitrate inputs from agricultural fertilizers.In Ireland,under the Environmental Protection Act (1992),the effluent from urban wastewater treatment plants discharging to sensitive areas––for populations in the range of 10,000–100,000person equivalent (PE)––should either have total nitrogen concentrations not exceeding 15mg N/l or 70–80%reduction of nitrogen influent values.Biological nitrogen removal (BNR)has become a common wastewater treatment practice.Many process configurations for BNR are available.A traditional BNR system is either a separate-stage de-nitrification system or a single-sludge nitrification–denitrification system (USEPA,1993;Tchobanoglous et al.,2003).In the separate-stage denitrification system,whether a combined carbonaceous oxidation/nitrifica-tion unit process or a separate stage nitrification unit process is used,denitrification is accomplished in a separate unit process following carbonaceous removal and nitrification.An external carbon source is added to the denitrification unit to provide readily available car-bonaceous matter for denitrification.The single-sludge nitrification–denitrification system,including the modi-fied Ludzack–Ettinger process,A 2/O process,UCT process and Bardenpho process,combines carbonaceous removal,nitrification and denitrification in the same process and the carbon source present in the wastewater is used to sustain denitrification.In practice,the suspended growth activated sludge and the attached growth biofilm are both used for BNR.In the present study,a BNR process using a vertically moving biofilm system (VMBS)was used to treat synthetic wastewater.This process comprised two pre-denitrification units,a combined carbonaceous oxida-tion/nitrification unit and three nitrification units,all in sequence.Nitrate produced in the nitrification units was recirculated to the first pre-denitrification unit.Each*Corresponding author.Tel.:+353-91-524411x2762;fax:+353-91-750507.E-mail address:xinmin.zhan@nuigalway.ie (X.-M.Zhan).0960-8524/$-see front matter Ó2003Elsevier Ltd.All rights reserved.doi:10.1016/j.biortech.2003.09.017Bioresource Technology 93(2004)313–319unit process employed attached-growth micro-organ-isms in a biofilm on the surface of a plastic biomedia module,which was vertically moved repeatedly in the bulk wastewater in the pre-denitrification units and into and out of the bulkfluid in the carbonaceous oxidation and nitrification units.In comparison with other fre-quently used biofilm systems,including tricklingfilters, biological aeratedfilters(BAF),fluidized bed reactors (FBR)and moving bed reactors(MBR),this VMBS has a number of advantages(Rodgers,1999):no back-washing is required to prevent clogging;no complicated air supply system is required;and maintenance and operation are cheap and simple.These advantages have been demonstrated in a pilot wastewater treatment plant using a VMBS system(Rodgers et al.,2003).The specific objective of this study was to investigate the performance of the small-scale BNR process using the VMBS system in nitrogen removal from a synthetic wastewater.2.Methods2.1.Experimental systemThe small-scale BNR system was located in a room with controlled temperature at11°C and consisted of the following(Fig.1):six polypropylene tanks in series; six biofilm modules––one for each tank;a wastewater feed mixing tank;three peristaltic pumps,one each for the feed,dilution water and the recirculation of the nitrified wastewater;and a pneumatic system complete with limit switches and delay controllers that powered compressed air cylinders to lift and lower the plastic biofilm modules.The six tanks were fabricated from polypropylene sheets,with a square internal base of0.4m side and a height of0.6m.The tanks were connected,in series, with50mm diameter polypropylene piping.The outlets from each tank were arranged so thatflow took place from Tank1through to Tank6.Tank1and2were anoxic and the remaining four tanks were aerobic.In Tanks1and2,a cube of corrugated polyvinyl chloride (PVC)sheets––with a side dimension of0.3m and a specific surface area of150m2/m3––was repeatedly moved vertically up and down,using pneumatic pistons and limit switches,and was always submerged in the bulkfluid;the piston stroke was80mm and the module movement was at the rate of22cycles per minute.In Tanks3–6,higher density modules of the same overall dimensions but with a specific surface area of240m2/m3, providing a module surface area of6.48m2,were sup-ported on a frame and moved by a pneumatic piston,a total distance of0.4m,into and out of the bulkfluid in a cycle typically consisting of4s in thefluid,2s coming out of thefluid,4s out of thefluid and2s going into the fluid.The feed,along with dilution tap water,was pumped into Tank1.Nitrified wastewater from Tank6was re-circulated into Tank 1.Denitrification occurred in Tanks1and2.Any remaining carbonaceous oxidation and nitrification took place in Tanks3–6.Some of the treated wastewater was discharged from Tank6to the public sewer and the remainder was recirculated to Tank 1.The feed was made up daily in the feed mixing tank and was composed of glucose,yeast extract,dried milk, urea,NH4Cl,Na2HPO4Æ12H2O,KHCO3,NaHCO3, MgSO4Æ7H2O,FeSO4Æ7H2O,MnSO4ÆH2O,CaCl2Æ6H2O and bentonite,with total COD of6019mg/l,fil-tered COD of3927mg/l andfiltered BOD5of2555mg/l.This study lasted seven months and included three serial experiments,Experiments1,2and3,with different recirculation ratios and different carbon and nitrogen influent loadings(Table1).In sequence,Experiment1 lasted for3months,Experiment2for1month and Experiment3for3months.2.2.AnalysisSamples were taken on a nearly daily basis from Monday to Friday every week.Total andfiltered chem-ical oxygen demand(COD)was measured in accordance with the standard APHA methods(APHA,1998).Fil-tered samples were obtained byfiltering the wastewater through a Whatman GF/C glass microfiberfilter paper (pore size1.2l m).Dissolved oxygen(DO)was mea-sured in situ with an electrochemical membrane elec-trode(WTW cellOx325,Wissenschaftlich-Technische Werkstatten GmbH&Co.KG,Germany)and a digital DO meter.NH4-N and NO3-N were measured using WTW electrodes.All electrodes were calibrated in accordance with the manufacturers’procedures. Total Kjeldahl nitrogen(TKN)was carried out using equipment supplied by Buchi Laboratoriums––Technik AG.314M.Rodgers,X.-M.Zhan/Bioresource Technology93(2004)313–3193.Results and discussion3.1.Overall performance of the experimental system After Experiments 1,2and 3reached the pseudo-steady states,8sets of samples were tested for each experiment.The mean values of parameters were cal-culated and are discussed below.Figs.2–4present the profiles of NH 4-N,NO 3-N and filtered COD (COD f )along the wastewater flow in the three experimental cases.It appears in these graphs that COD f was de-graded in Tanks 1–3and nitrification took place in Tanks 3–6and was nearly completed in Tank 6.Since Tanks 4–6did not make a significant contribution to carbonaceous removal,the three tanks could be re-garded as single-sludge nitrification units.Removal efficiency of COD f and total nitrogen (TN)are shown in Table 2.Using this system,carbonaceous COD removal was nearly complete,94–96%,and TN removal was 77–82%.TN in the effluent (TNe)tended to increase with increasing TKN loading into this system (Fig.5).WhenTable 1Flow regimes and substrate inflow concentrations Experiment no.Q I (m 3/d)Q T (m 3/d)Q F (m 3/d)Q R (m 3/d)R TKN ininflow (mg/l)COD f in inflow (mg/l)BOD f in inflow (mg/l)10.3970.0490.348 1.010 2.547548531520.4040.0650.3390.990 2.459863241130.3360.0750.2610.5171.54136877570Note :Q I ,total inflow,Q I ¼Q T þQ F ;Q T ,tap water inflow;Q F ,synthetic wastewater inflow;Q R ,return flow;R ,recirculation ratio,R ¼Q R =Q I ;TKN,total Kjeldahl nitrogen;COD f ,filtered COD;BOD f ,filtered BOD 5.M.Rodgers,X.-M.Zhan /Bioresource Technology 93(2004)313–319315the recirculation of nitrified water was considered,TKN concentration entering Tank1(TKN1i)was equal to TKN0/ðRþ1Þ,where TKN0was the concentration of TKN in inflow and R the recirculation ratio;NH4-N in the effluent was negligible in comparison with TKN in inflow.Fig.5shows the relationship between effluent TN(TNe)and TKN1i.TNe significantly relied on TKN1i and can be expressed as TNe¼0.6TKN1i(The regression coefficient,R2¼0:99).Hence,the TN re-moval efficiencyðrÞis expressed as r¼1À0:6=ðRþ1Þand can be improved by raising the recirculation ratio, R.Meanwhile,since d2r=d R2<0,indicating that the increment of the TN removal efficiency decreased with increasing the recirculation rate,a tradeoffbetween r and the operational cost should be considered when raising R to improve the TN removal efficiency.The operation of this system was very simple,com-pared with other biofilters.Clogging,which often occurs in biofilters did not take place during the seven month experimental period,so that backwashing was not necessary.Since the modules moved quickly,0.2m/s in the aerobic tanks(Tanks4–6),the biofilm was kept thin due to the hydraulic shear forces.Trulear and Char-acklis(1982)found that biofilm detachment rate occur-ring on an annular reactor increased with rotational speed.Similarly,Cheng et al.(1997)observed that in a three phase draft-tubefluidized bed using granular acti-vated carbon(GAC),while the mean liquid velocity increased from0.12to0.16m/s,the biomass attached onto the GAC decreased from30.4to15.6mg VSS/g GAC.3.2.Pre-denitrification efficiencyDenitrification was completed in the two anoxic tanks,Tanks1and2.Since there was no nitrate in the feed and dilution water,volumetric denitrification rates in Tank1(DNV1)and Tank2(DNV2)were calculated using the following equations,respectively:DNV1¼ðRþ1ÞÂððNO3-NÞ1iÀðNO3-NÞ1ÞÂQ I=Vð1ÞDNV2¼ðRþ1ÞÂððNO3-NÞ1ÀðNO3-NÞ2ÞÂQ I=Vð2Þwhere,ðNO3-NÞ1i¼RÂðNO3-NÞ6=ðRþ1Þ,the con-centration of nitrate entering Tank1;(NO3-N)1,(NO3-N)2and(NO3-N)6represent the concentration of nitrate in Tanks1,2and6,respectively;Q I is the inflow and V is the volume of the bulkfluid in each tank.The areal nitrate removal rates,based on the surface area of the biofilm substratum,DNS1in Tank1and DNS2in Tank2,were calculated with the above two equations by replacing V with the surface area of the biofilm modules,4.05m2,in the denitrification tanks.Table3lists the denitrification rates in the small-scale system.It is clear from the table that denitrification was nearly complete in Tank1because95%,93%and99%of the total nitrate removal took place in Tank1in Experiments1,2and3,respectively.The nitrate con-centrations entering Tank2were so low that the driving forces of denitrification kinetics were small.Conse-quently,the denitrification rates in Tank2were much lower than in Tank1.The surface nitrate removal rates in Tank1were generally higher than those of upflowfluidized-bed systems cited by the USEPA(1993),0.78–3.4g NO x-N/ (m2Æd)and downflow packed-bed systems,0.29–1.6gTable2Overall performance of the BNR systemExperiment no.COD f in Tank6(mg/l)TN in Tank6(mg/l)COD f removal efficiency(%)TN removal efficiency(%) 127.5(7.5)a13.6(1.6)94%(1%)82%(2%)227.1(7.7)18.6(1.8)96%(1%)81%(2%)341.0(4.9)32.0(2.4)95%(1%)77%(2%)a Data in the brackets are the standard deviations.316M.Rodgers,X.-M.Zhan/Bioresource Technology93(2004)313–319NO x-N/(m2Æd).The volumetric nitrate removal rates were lower than the average denitrification rates of240g N/(m3Æd)of a moving-bed system with acetate as a substrate(Maurer et al.,2001).The reason was the low specific surface area of the biofilm compared to the volume of the bulkfluid,42m2/m3.This specific surface area could be increased by using denser media,provided that clogging could be controlled.Entering Tank1,the(COD f)1i/(NO3-N)1i ratios were equal to17,16and9.3in Experiments1,2and3, respectively,where(COD f)1i was COD f entering Tank 1.The three ratios were more than necessary for deni-trification.Aesoy et al.(1998)found that the required COD/NO3-N ratio was close to4.5g COD/g NO3-N with ethanol as the carbon source and8–10g COD/g NO3-N with hydrolysate from sludge and solid organic waste;Garrido et al.(2001)found3.5g COD/g NO3-N was needed using formaldehyde as the carbon source. Along with nitrate removal,COD f removal also oc-curred in the anoxic tanks.Of the total COD f removal occurring in the experimental systems,62%,52%and 42%was completed in Tank1with the mean areal COD removal rates of27,31and29g COD f/(m2Æd)in Experiments1,2and3,respectively.DO concentrations in the bulkfluid in thefirst anoxic tank,Tank1,were up to0.5mg/l.As a result,aerobic heterotophic growth could have taken place alongside the anoxic heterotro-phic growth.However,the high DO concentration would adversely affect the denitrification rate.Oh and Silverstein(1999)reported more than a35%decrease in the specific denitrification rate at a DO concentration of only0.09mg/l.It is recommended that minimal oxygen should be introduced by the influent and recycleflows or by surface transfer(USEPA,1993).As a result of the high DO,the nitrate removal rates obtained in the present study were less than the maximum rates achieved by a rotating biological contactor treating a similar wastewater(Ødegaard and Rusten,1980).3.3.Nitrification performance in the aerobic tanksThe areal ammonium removal rates(Table4)in Tank3were lower than in the following tanks though the influent ammonium nitrogen concentrations were highest.This resulted from the organic substrate inhi-bition on nitrification.Fdz-Polanco et al.(2000)found that the COD:NH4-N ratio greater than4may result in losing nitrification efficiency in a submerged aerated filter,where two spatial zones appeared,one with high TOC removal rate but low ammonium removal rate, and the other with high ammonium removal rate but low TOC removal rate.The reason for organic substrate inhibition on nitrification is the competition between nitrifiers and heterotrophs for dissolved oxygen and space in the biofilm(Zhang et al.,1995;Okabe et al., 1996).In this system,particularly in Tanks4and5,areal ammonium removal rates up to1.8g NH4-N/(m2Æd) were achieved.The nitrification efficiency compared well with those achieved by other researchers using different biofilm nitrification systems:0.83g NH4-N/(m2Æd)in a moving bed biofilm reactor(MBBR)by Johnson et al. (2000), 1.0g NH4-N/(m2Æd)in a MBBR system by Andreottola et al.(2000),0.84g NH4-N/(m2Æd)infixed-bed submerged biofilters by Canziani et al.(1999)and 0.6–1.6g NH4-N/(m2Æd)in a nitrifying tricklingfilter by Thorn et al.(1996).Fig.6shows that NH4-N in Tanks3–6decreased linearly in Experiments2and3,indicating that the nitrification can be expressed by zero-order kinetics.In Experiment1,the relationship between NH4-N and tank number was not as linear as in the other two cases because the concentrations of ammonium in Tanks5 and6were very low,around1mg/l and were similar to typical values of the half-saturation constant,K N, used in the Monod equation,e.g.,0.3–0.7mg NH4-N/l (Henze,1997).When K N was low with respect toTable3Denitrification performance of the anoxic tanksExperiment no.(NO3-N)1i(mg/l)(NO3-N)1(mg/l)(NO3-N)2(mg/l)NO3-NremovalDNV1(g/(m3Æd))DNV2(g/(m3Æd))DNS1(g/(m2Æd))DNS2(g/(m2Æd))19.2(1.2)a0.9(0.4)0.5(0.1)95%(1%)120.8 6.5 2.870.15212.5(1.2) 1.6(0.4)0.7(0.4)94%(4%)158.9 3.2 3.770.31318.2(1.3)0.4(0.1)0.3(0.1)98%(1%)158.60.1 3.760.01a Data in the brackets are the standard deviations.Table4Areal ammonium removal rates(g NH4-N/(m2Æd))Experiment no.Tank3Tank4Tank5Tank610.7(11.8)a 1.3(8.6)0.6(3.7)0.1(1.2)20.4(16.8) 1.2(15.2) 1.4(9.7)0.5(3.3)30.8(39.3) 1.1(33.1) 1.8(24.5) 1.1(10.6)a Data in the brackets are concentrations of ammonium nitrogen in the preceding tanks(mg/l).M.Rodgers,X.-M.Zhan/Bioresource Technology93(2004)313–319317ammonium concentrations in the bulkfluid,with-out considering other factors,the ammonium remo-val rates were independent of the concentrations of ammonium.The advantage of the system for nitrification is the high oxygen transfer capacity of the biofilm modules that were lifted into the air and lowered into the bulk fluid.The oxygen transfer mechanisms are considered to be similar to those occurring in rotating biological contactors and include:oxygen absorption at the liquid film over the biofilm’s surface when the modules were in the air;direct oxygen absorption by the micro-organ-isms during the air exposure;and direct oxygen transfer happening at the air–water interface caused by the tur-bulence created by the movement of the biofilm modules (Grady and Lim,1980).The oxygen transfer capacity of clean modules depended on the vertical movement cycle rates of the modules and was up to0.001–0.0027sÀ1. The mean dissolved oxygen concentrations in the bulk fluid in Tanks3–6were up to6–9mg/l.Cecen and Orak (1996)found that in a submerged aeratedfilter nitrifying fertilizer wastewater,the maximum nitrification rate was strongly dependent on dissolved oxygen(DO).When DO increased from3.2–3.5mg/l to4.9mg/l,the maxi-mum ammonium removal rate was increased from0.17 to0.41kg NH4-N/(m3Æd).However,in this study,the high DO in Tank6should have had an adverse effect on the pre-denitrification process.Separate pneumatic pis-tons for each module would have given better control and efficiency.4.ConclusionAfter studying the operation of a small-scale6-tank BNR process using a VMBS,the following results were obtained:(1)Thefirst two tanks in the system had very low dis-solved oxygen concentrations and denitrification took place in these two tanks.The remaining four tanks were aerobic and combined carbonaceous oxi-dation/nitrification occurred in the third tank and nitrification alone in the last three tanks.(2)In this treatment system,the carbonaceous removalefficiency was94–96%and the TN removal efficiency was77–82%.Effluent TN significantly depended on the TKN entering Tank1.(3)In the anoxic units,the denitrification efficiency wasin the range of94–98%and the areal denitrification rates were2.9–3.8g NO3-N/(m2Æd).However,the DO in thefirst anoxic tank possibly had an adverse effect on denitrification.(4)The nitrification efficiency occurring in the aerobictanks was up to95%and the areal ammonium re-moval rates were in the range of1.3–1.8g NH4-N/ (m2Æd).(5)No clogging of the biofilm modules occurred duringthe seven month study due to the hydraulic shearing forces resulting from the vertical movement of the biofilm modules.ReferencesAesoy,A.,Ødegaard,H.,Bach,K.,Pujol,R.,Hamon,M.,1998.Denitrification in a packed bed biofilm reactor(BIOFOR)-exper-iments with different carbon sources.Water Res.32,1463–1470.Andreottola,G.,Foladori,P.,Ragazzi,M.,Tatano, F.,2000.Experimental comparison between MBBR and activated sludge system for the treatment of municipal wastewater.Water Sci.Technol.41,375–382.APHA(American Public Health Association),1998.Standard Meth-ods for the Examination of Water and Wastewater.American Public Health Association,Washington,DC.Canziani,R.,Vismara,R.,Basilico,D.,Zinni,L.,1999.Nitrogen removal infixed-bed submerged biofilters without backwashing.Water Sci.Technol.40,145–152.Cecen,F.,Orak,E.,1996.Nitrification of fertilizer wastewaters in a biofilm reactor.J.Chem.Technol.Biotechnol.65,229–238. Cheng,S.-S.,Chen,W.-C.,Hwang,H.-H.,1997.Biofilm formation: the effects of hydrodynamic and substrate feeding patterns in three phase draft-tubefluidized bed for nitrification process.Water Sci.Technol.36,83–90.Fdz-Polanco,F.,Mendez,E.,Uruena,M.A.,Villaverde,S.,Garcia, P.A.,2000.Spatial distribution of heterotrophs and nitrifiers ina submerged biofilter for nitrification.Water Res.34,4081–4089.Garrido,J.M.,Mendez,R.,Lema,J.M.,2001.Simultaneous urea hydrolysis,formaldehyde removal and denitrification in a multifed upflowfilter under anoxic and anaerobic conditions.Water Res.35,691–698.318M.Rodgers,X.-M.Zhan/Bioresource Technology93(2004)313–319Grady Jr.,C.P.L.,Lim,H.C.,1980.Biological Wastewater Treatment: Theory and Applications.Marcel Dekker,New York.pp.755–756. Henze,M.,1997.Basic biological processes.In:Henze,M.,Harrem-oes,P.,Jansen, C.,Arvin,E.(Eds.),Wastewater Treatment: Biological and Chemical Processes.Springer Verlag,Berlin,Hei-delberg,Germany,p.85.Johnson,C.H.,Page,M.W.,Blaha,L.,2000.Full scale moving bed biofilm reactor results from refinery and slaughter house treatment facilities.Water Sci.Technol.41,401–407.Maurer,M.,Fux, C.,Graff,M.,Siegrist,H.,2001.Moving-bed biological treatment(MBBT)of municipal wastewater:denitrifica-tion.Water Sci.Technol.43,337–344.Ødegaard,H.,Rusten, B.,1980.Nitrogen removal in rotating biological contactors without the use of external carbonaceous source.In:Proceedings of the First National Symposium/Work-shop on Rotating Biological Contactor Technology.Champion, Pennsylvania,vol.1,pp.1301–1317.Oh,J.,Silverstein,J.,1999.Oxygen inhibition of activated sludge denitrification.Water Res.33,1925–1937.Okabe,S.,Oozawa,Y.,Hirata,K.,Watanabe,Y.,1996.Relationship between population dynamics of nitrifiers in biofilms and reactor performance at various C:N ratios.Water Res.30,1563–1572. Rodgers,M.,anic carbon removal using a new biofilm reactor.Water Res.33,1495–1499.Rodgers,M.,Zhan,X.-M.,Gallagher,B.,2003.A pilot plant study using a vertically moving biofilm process to treat municipal wastewater.Bioresource Technol.89,139–143. Tchobanoglous,G.,Burton,F.L.,Stensel,H.D.,2003.Wastewater Engineering:Treatment and Reuse.Metcalf&Eddy,New York.pp.749–789.Thorn,M.,Mattsson,A.,Sorensson,F.,1996.Biofilm development ina nitrifying tricklingfilter.Water Sci.Technol.34,83–89. Trulear,M.G.,Characklis,W.G.,1982.Dynamics of biofilm pro-cesses.J.Water Pollut.Control Fed.54,1288–1301.USEPA,1993.Nitrogen Control.Technomic Publishing Company, Pennsylvania,USA,pp.211–311.Zhang,T.C.,Fu,Y.C.,Bishop,P.L.,petition for substrate and space in biofilms.Water Environ.Res.67,992–1003.M.Rodgers,X.-M.Zhan/Bioresource Technology93(2004)313–319319。
