Experimental study on zero excess sludge production process with ozonation unit in MBR process

Journal of Harbin Insitute of Technology ( New Series), Vol. 17, No. 1, 2010 .Experimental study on zero excess sludge production processwith ozonation unit in MBR processWANG Zheng'，WANG Lin2 , WANG Bao-zhen' , LIU Shuo' , JIANG Yifeng'王正王琳，王宝贞，刘硕蒋轶锋(1. School of Munieipal and Environmental Engineering, Harbin Institute of Technology, Harbin 150090, China, mezwang@ 163. com;2. School of Environmental Science and Engineering, Ocean University of China, Qingdao 266003, China;3. College of Biological and Environmental Engineering, Zhejiang Universily of Technology , Hangzhou 310032 , China)Abstract: In order to investigate the infuence factors of zero excess activated sludge ( EAS) process by ozona-tion, a 100 L membrane bioreactor coupled with a sludge ozonation unit ( MBR-SO) was performed for 80 dwithout EAS wasting. Some mathematical models were developed to elucidate the relationship between processparameters and the operating efect. It is considered that the sludge lysing ratio (ξ) , produced COD per unitmass lysed MLSS (η) , observed sludge yield cofficient ( Yon ) and intrinsic yield coffcient for COD producedby lysed sludge ( Y2 ) significanly afet the flowrate to ozonation unit (q). When q is 0. 0067 times of influentflowrate (Q) and ξ is about 0. 72 for each batch ozonation, a relatively stable MLSS concentration of 8168 mg/L and zero Ym are achieved in the MBR-SO system. The calculation of sludge disintegration number indicatesthat the high ξ can apparently decrease the sludge amount needed for ozonation. At the same ozone dose, thelow input ozone concentration and high flowrate can enhance the sludge lysing efcts and a low energy consump-tion of 0.041 Yuan/m' wastewater is obtained.Key words: excess activated sludge; ozonation; MBR ; modeling ;economic analysisCLC number: X703. 1Document code: AArticle ID: 1005-9113(2010)01 0042-05The biological processes are the most widely adaptedin Fig. 1. In order to reach an equilibrium state of con-for domestic and industrial wastewater treatment, which dai-stant sludge concentration , the sludge mixed liquid wasly produce a considerable amount of excess activated sludgecontinuously transferred into sludge ozonation unit for(EAS). For the pupose of reducing the volume and massdisintegration, and after ozonation, the ozonated mixedof the EAS, the ozone oxidation technique was developedliquid was recycled to the MBR for biodegradation. Dueand applied in pilot-scale and full-scale sludge reductionto sterilization of ozone gas, almost all the returmedprocessIn practice, some literatures'- reported thatsludge (Xo,a) was destroyed and became inactive cells.ozone dose (OD) and sludge disintegration number (Nso)Although some can be hydrolyzed and degraded by mi-has been considered as the vital operational parameters and,croorganisms in aeration basin, there still exist certaingenerally, high concentration ozone was used for minimizingconcentration inactive sludge in MBR (6 I . Therefore ,sludge in their studies. In this paper, it was dfferent thattotal sludge concentration (X) was composed of activa-low oczone concentration and high input flow-rate would beted sludge concentration (X) and inactive sludge con-favorable to improve the efct of ozonation 51. It is alsocentration (XJ).known that，under the same influent condition, long sludgeBase on the above depiction, the destroyed MLSSretention time process with a lower sludge yield cofficientconcentration (Xd,R) and total produced COD (S,) inwas relatively effective, though it was not perfect, to reducerecycling sludge can be expressed by:the EAS production. Therefore, in order to obtain a eco-X=XL+X(1)nomically and completely elimination of EAS, a membraneX.n = (1 -ξ)(X[ + X)(2)bioreactor process combined with a sludge ozonation units,=S,+ξη(X.+x)(3)(MBR-SO system) was proposed to investigate this specificwhere s, is COD concentration in MBR basin, mg/L; ξ .issue by dosing certain amount of ozone.is sludge lysing ratio which is defined as the ratio ofMLSS。are ozonation unit toMaterials and MethodsMIS中国煤化Insionless; ηis the1.1 Process Design Foundationmeal:LHCNMH GAucorulg uU 1ass valalluo UMBR, the fllowingThe MBR-SO process used in this study was shownReceived 2007 -11 - 16..42.Journal of Harbin Institue of Technology (New Series)， Vol. 17, No. 1, 2010Influent 一Equalization 0.S.Aenation basinPumpI Q.S.。Efluentbasin. Flat Membrane1931NutrientswBPump! OzonizedFor defoaming IL SludgeSludge storage| basin 1NaHPO.0Of-gas adsonption andmcaurement instrumentOrzonaion Input ga aoption andinstrumentnit 「hasin 2血面因口咽。:pumpOzone generator Air pumpFig. 1 Schematics of the MBR-SO processequations can be derived:(S。-S)YdnvA =Q(S。-S.)Y, -bVX[ +q(S, -S,)Y2- .丹<ξX(1 - nY2)(9)dtIn the MBR-SO system, flowrateqis a key factorqXL + ykVXx.Y3(4)for the process design and stable operation. Sev=q(Xen-x) -k,vX。(5)一(S。 -S.)Ydm, and 9 can be determined byl“ξX(1一ηY2)where Q(S。-S.)Y - bVX, is the observed inereasingthe maximum of q/Q. Therefore， all parameters ap-rate of sludge in MBR; b is sludge decay constant,peared in right-hand of Eq. q/Q maximum should bed~';q is flowrate of ozonated sludge, Ld, V is theobtained before the operation of MBR-SO process.volume of MBR, L; γ is the cefficient of COD pro-1.2 MBR-SO Processduced by hydrolysis of per unit mass ozonated sludge，In this study, a 100 L aeration basin and 8 piecesdimensionless; hp is hydrolysis rate constant, d-'; Y,of flat microfiltration membranes with pore size ofY2 and Y3 is the activated sludge (X[) yield cofficient0.4 μm were adopted for MBR process. The mem-of COD in influent, COD released by ozonolyzed sludgebranes were made from polyethylene and the porosityand COD produced by ozonated sludge ，respectively,percentage was around 40% - 50% . The infuent flow-dimensionless; After long term operation, the systemrate was 300 L/d, and the hydraulic retention timereached an equilibrium state and the equation of stead-( HRT), permeate flux of membrane and dissolved ox-y-state condition can be expressed as follows:ygen (DO) was8h, 0.31 m'/(m2●d) and2 mg/L,.vd=vdXL+vdXg=vdAL=vdxs=0respectively. Sludge was ozonated in a 1 L cylindricalJtltddt、glass contactor(φ = 35 mm;H= 1 700 mm). The 0~(6)zone gas flowrate was 0.1 m'/h. The input oz0ne con-Inserting Eqs. (1)-(5) to(6):centration was 2.2 -2.4 mg/L and ozone transfer effi-Q(So -S.)Yn -q5X(1 - nY2) - h。VX(1 -ciency was at 95%. Intermittent backwashing proce-γY;) =0dure was performed according to the TMP variationswhere the Yon is the observed sludge yield coefficient inmonitored in order to minimize the membrane fouling.aeration basin without ozone treatment, dimensionless.and an average TMP of 13.5 kPa was obtained duringThe Eq. (7) can be rewritten as the following equa-normal operation stage.tion:The seed sludge was taken from the Harbin Wen-Q(S。-S.)Ydm -kgVX.(1 -γ%)(8)chang Wastewater Treatment Plant. The influent wasQξX(1 - nY)supe中国煤化工and sometimes hIt is known that the mass of new generated biosol-calcYHCNMH(NH,CI，K,HPO4,id with the produced COD by XJ is less than the mass ofNa2were added to meetXg before ozonolysis, thush, vxg(1 -yY3) > 0 and thethe required influent quality. The influent characteris-Eq. (8) can be transformned as:tics of MBR-SO process were listed in Tab.1.●43●Journal of Harbin Instiute of Technology (New Series), Vol. 17, No. 1, 2010sludge concentration of 2230 mg/L and SCOD concen-Tab. 1 Influent characteristics of MBR-SO processtration of 5 832 mg/L were obtained at the OD ofInluent charactersticsConcentration0. 05 g/g MLSS.COD/(mg.L"')300-310oPCOD SCOD. MI.SSAmmonium nitrogen/(mg. 1L"I)32 -36Toal phosphorous/(mg.L")9-11至6Temperature/C20 -22After 30 d cultivation, the sludge concentration inMBR increased up to around 8000 mg/L. In order tomaintain the MISS concentration at 8000 mg/L,0.01 0.02 0.03 0.04 0.05sludge ozonation was conducted in semi-batch modeOzone dosage/ (mg*mg MIS51and no sludge was wasted during operation period.Fig.3 Characteristic curve of sludge ozonation unitl.3 Determination of Yobn and Y,According to the definitions of ξ and η, the Equa-In order to determine the Ydbn and Y2 of Eq. (9)simultaneously, a 10 L MBR with the same initialtions for ξ and η could be expresses as :MLSS concentration was performed in comparison withξ=PMISS- PMIsS8000 - 2230.=0.72.the 100 L experiment. The operation conditions of 10 LPMISSs8000MBR were identical to those of 100 L MBR system in1=PscoD, - PScoDo5832 - 14order that there is high contrastive significance betweenPMLss * ξ8000 x0.72them. The substrate adopted in 10 L MBR for Y; deter-1 mg SCOD/ mgMLSSmination was the supematant of sludge mixed liquid af-where PMIsS is the initial MLSS concentration in ozona-ter ozonation, which was diluted by tap water to reachtion unit, mg/L; Pscon and Pscoo, was the SCOD in ini-an identical COD concentration with that of 100 L MBRsystem. The EAS generated in two MBR units weretial and ozonated sludge mixed liquid， respectively,wasted daily to keep the MLSS concentration aroundmg/L. After 20 d operation, a low PscoDo, of 14 mg/L8000 mg/L. After 20 d operation, the two slopes inwas obtained.Fig. 2 which represented the Yan andY2 was0.15 g Ss/1.5 Determination of q/Qg COD and 0.17 g SS/g COD, respectively.Based on the above data and the Eq. (9), themaximum q/Q could be calculated when ML sS concen-tration of the MBR system was preset at 8 000 mg/L:y=0.1723xq(S。-S.)Y出R2=0.9952 !maξX(1 - nY2)◆W astewater0.72 x8000 x(1 -1 x0. 17)(305-14) x0. 15. = 0.0091y=0.1509xSetq/Q = 0.0067,i.e. q = 0.0067x300 L/d=示1.R20.99352 L/d, and the ozonation period was 12 h.旨0.In order to investigate the operation performance020of this designed MBR-SO process, total 100 d experi-Cumulativement could be divided into two stages: the frst stageFig. 2 Relationship between the cumulative sludge wastewas the previous 20 d, which was called control stageand COD removal of wastewater and ozonatedand carried out at MLSS concentration aroundsludge8000 mg/L with continuous sludge wasting for the de-termination of Yahn. The SRT in this stage was around1.4 Characteristics of Ozonation and Determina-60d. The second stage was the following 80 d andtion ofξ and η .called test stage, in which, according to the designedIn order to determine ξ and η, a series of batchflowrate q, the excess sludge was continuously wastedstudies were carried out to understand the characteristicsand destroyed in ozonation unit batch-wise. Finally,of sludge microbial cell oxidation and solubilization un-nneratin nerformance were estab-der the OD of0.05 g/g MLSS (4 ( as shown in Fig.3).lishe中国煤化工It was found that, as regards each batch operation1.6in so unit, the ozonated sludge remained the relativelyTYHC NMH G,one day and thestable composition and characteritics at the same inputozonated sludge were regularly collcted as samples.ozone concentration and OD. The average residualNon-filtered samples were analyzed for MLSS and total●44●Journal of Harbin Instiue of Technology (New Series)，Vol. 17, No. 1, 2010COD (TCOD). Filtered samples with 1 μm glass fibrecording to definition of Nsp and Eq. (9), the Nsp andfiter were analyzed for soluble COD ( SCOD) and par-maximum of Ngp(Nspms ) could be calculated:ticle COD ( PCOD). All analyses were performned inaccordance with Standard Methods'7l. The ozone con-Nsp =Q(S。- S.)Ydncentration was measured by iodometric titration.2 x 8168300x(305 -14) x0.15 =1.252 Results and DiscussionNsDmuξ(1←η2)=2.1 Variation of MLSS Concentration= 1.67In order to evaluate the efect of sludge reduction,0.72x(1-1 x0.17)the sludge concentration and Ydn in both two stages wasBased on the formula of Nspnmss , it could be conclu-plotted in Fig. 4 and an average Ydb of about zero inded that the limiting factors for Nsp wereξ, η and Y2.test stage was obtained. The experiment data indicatedthat, whenq is 0.0067 times of Q and after 80 d con-，TCOD.土TCOD。米TC0D roronal efi100tinuous operation, the MLSS concentration in MBR wasControlstable around 8 100 - 8500 mg/L and the average of230098MISS was 8168 mg/L, which was basically identical to首200%X对9the designed value of 8000 mg/L before the experi-ment.)290007 0,200601曾80015225503Fig.5 Infuent and efMluent COD in MBR-SO processavcpage for control stagg十0.27000 wwThe relationship between ξ and NspDmexs was plotted十0.1in Fig. 6. It could be concluded that, with the in-弓6000crease ofξ, the NsDmx decreased constantly, and, thehigher ξ could decrease the ozonated sludge amount,500oZero lieof Ya10which was very important for maintaining the sludge activity of MBR-SO system. In Lee' s experiment (21 ,Fig.4 MLSS concentration and Yon in both two stagesonly 0.3 of sludge lysing ratio was obtained under theconditions of high MISS concentration ( larger tharIt could be seen that a complete reduction could8000 mg/L)，high ozone concentration (50 mg/L)be achieved at the 0D of 0.05 mg/ mg MLSS and aq ofand lower flowrate (30 mL/d)，which was much lower0. 0067Q was necessary to keep the sludge concentra-than 0.72 of this study under the conditions of lowtion around 8000 mg/L.MLSS concentration ( around 8000 mg/L), low ozone2.2 COD Concentrationconcentration (2. 232 mg/L) and higher ozone flowrateAs shown in Fig. 5, the experiment results indica-(0.1 m'/h). Therefore, it could be concluded thatted that the recycling of ozonated sludge hardly affectedthe lower MLSS concentration and lower input ozonethe COD removal efficiency of MBR process, and aconcentration were favorable to the increase of sludgehigh COD removal fficiency of 94. 5% - 96. 5% inlysing ratio ξ and could decrease the energy consump-both stages was observed. The effluent COD concentra-tion.tion was stable around 14 mg/L. It was a noticeableresult that lower influent COD concentration and betterbiodegradability of released organic matters from oz0n-●n=1.0,Y=0.17ated sludge were the main causes of better COD remov-al eficiency in efluent. Although there was higherCOD concentration in ozonated sludge mixed liquid,6-y=1.2xr+the lower recycling flowrate ( only 0.0067 times of in-fluent) and relatively long HRT of 8 h were contribu-tive to ensure a strong anti-overload ability for MBR中国煤化工system.2.3 The Sludge Disintegration Number (Nsn)MYHCNMHGξ0.81.0In this experiment, a low concentration ozone gasFig.6 NsDms curve as function of ξwith higher input gas flowrate was applied, and, ac-●45●Journal of Harbin Instiue of Technology (New Series)， Vol. 17, No. 1, 20102.4 Economic Evaluationration could treat both the wastewater and activatedThe energy consumption cost of about 15 Yuansludge cost-effectively, and an additional ozonation op-(RMB) is needed to produce 1 kg of ozone gas witherating cost of 0. 041 Yuan/m’ wastewater for EASair. Base on the experimental data, the operating costtreatment was obtained. On the other hand, as com-of sludge ozonation could be calculated. Taking thepared with real wastewater, in which there is largerozonated sludge flowrate q = 0. 0067Q，0D =variation both in wastewater quality and flowrate, it is0.05 kg/kg MLSS, and ozone transfer efficiency ofnecessary to properly control the q/Q ratio for achieving95% as an example, the sludge ozonation operatingstable operation.cost per unit volume sludge is equal to:(15 Yuan/kg O3 x8. 168 kg MLSS/m' x 0.05References:kg/kg MLSS)/95% =6. 13 Yuan/ m'sludge[1] Yasui H, Nakamura K, Sakuma s, et al. A fll-scale oper-and the additional ozonation cost for wastewateration of a novel activated sludge process without excesssludge production. Water Science Technology, 1996, 34treatment is ;(3/4) :395 - 404.6.13 Yuan/m'sludge x (0. 0067Q sludge/ Q[2] LeeJW, Cha HY, Park K Y, et al. Operational strategieswastewater) =0. 041 Yuan/ m' wastewaterfor an activated sludge process in conjunction with ozoneIn China， as regard a conventional wastewateroxidation for zero excess sludge production during wintertreatment plant, about 0.6 Yuan is required for trea-season. Water Research, 2005 ,39:1199 - 1204.ting 1 m' domestic wastewater. Whereas, in this MBR-3] Yoon S H. Operational strategies for an activated sludgeso system, the cost for treating EAS generated by 1 m'process in conjunction with ozone oxidation for zero excesssludge production during winter season. Water Research,domestic wastewater accounts for 6. 83% of wastewater2005, 39 :5274 - 5276.treatment cost.[4] Yasui H, Shibata M. An innovative approach to reduce ex-cess sludge production in the activated sludge process. Wa-3 Conclusionter Science Technology, 1994, 30(9):11 -20.[5] WangZ, Wang L, Wang B z, et al. Variation of parame-In order to realize zero EAS production inters in mixed liquor during sludge reduction by ozonation.wastewater treatment process, a 100 L MBR-SO systemUrban Environment and Urban Ecology, 2007, 20(3): 11was performed with the optimized ratio of q/Q whichhad strong relationship withξ, η, Yohm and Y2. When q[6] Yoon S H. Important operational parameters of membranebioreactor-sludge disintegration ( MBR-SD) system for zerois0.006 7 times ofQ andξ is about 0.72 for eachexcess sludge production. Water Research, 2003 , 37 :1921batch ozonation, the MBR maintained a relatively sta-一1931.ble MLSS concentration. The ozonated sludge amount[7 ] State Environmental Prolection Administration of Chinacould be dropped apparently with a higherξ. Low input( SEPA). Standard Methods for the Examination of Waterozone concentration and high flowrate could enhanceand Wastewater. 3rd ed. Bejing: China Environmental Sci-the sludge lysing effects at the same OD. Economic e-ence Press, 1989.valuation indicated that this MBR-SO process configu-中国煤化工MYHCNMHG