Phosphorus limitation in biofiltration for drinking water treatment

Journal of Fnuronmenral Srienes Vol. I5,No.4.pp .494- -499 .2003Article ID: 1001-0742( 2003 04-0494-06CLC number: X703 Docunent code: APhosphorus limitationinbiofiltration for drinkingwatertreatmentYU Xin'* ，ZHANG Xia-jan' ，LIU Xia-ling , ZHAO Xiao-dong ，W ANG Zhan sheng'(1. Depurtmerol of Eniroumental Science and Engincering, Tsinghua Unirersity，Reiig 000804.0 Chin. E muil: yu xin99 @ mails. linglu .edu.cn; 2. Xuzhou Water Supply Company. Xuzhou 221004. China)Abstract: Bauterial growh potential( BGP) metbod and two prallel plol-acale biofilere were used to invetigte pbonphorus lnitarion and itsfrect oan the renxwal of onganic malles in bililtrain for drinking waler treatment. Addition of plbosphorus can subeantilly increase the D6Psof the samples. Ilb elet was equivalent to that of aditin of a mistue of various iorganie rutients ineludig phosphorus. The bifiter withphosphate added ino is influent peforned a higher biological sabiliy of the efluent and a higher C0Dy.. renoval than the control fiter. Theseresulhe suggested that pboephorus was the limiting nutrient in the biofilration and the renoval eficiency of organie malters coud lhe inproved byadding phosphate into the influent.Keywords: drinking water: hiologicad reatmrnt;: biofiler: lnmiting nutient; phosphorus; carbonIntroductionPollutionof raw water is a conumon problem for the waler plants in China and organie matters areusually among the main pollulants( Wang, 1999). Biofitration is often undertaken to solve this problem asan enhanced process。Besides filtration, the primary mechanism of this process to remove the pollulanls isthe biochenical oxidation achieved by the microbes in the bioflm altached Lo the media in the filters.However, the removal of this process is ofen limited by the low levels and birnegreudbility of the organicmatters. If (12onation is applied proceeding biofiltration， the removal rale of organic matters will besignificanly increased( Cipparone, 1997), but most waler plants in China can not aford is high expensenow, s0 new ways in low cost should be found.Recent research revealed lhat phosphonus might play a role of limilting nutrient. Mietinen et al.(Mietinen, 1996; 1997) foundat natural waters in the northerm hemisphere generally had a high conlentof organic carbon and the microbial growth in drinking waler in Finland was highly regulated nut only byorganie carbon but also by the availability of phosphorus. Sathasivan et al . ( Sathasivan，1999) reportedthe bacterial regrowth in the distribution syscems of Tokyo was evidently limited by phosphorus. Theseresults suggested the micrmbial growth and metabolism in the biofilers could be promoted by the addition ofphosphorus thus the removal rate of organic mattes might be improved.Bul u to now, there is serious lack of the references on the direct application of phosphornus linitationto improve the removal of the organic matters. The studies were restricted in using artificial cockailsubstrates or in a jar-test-scale Nihijima, 1997). In this study, the limiting efet of phosphorus in thebiofiltration proc:rss , which produced nearly 7 m' waler/ d and whose influent was the actual water in field,was testified and analyzed using the bacterial gruwth polential( BCP) method, which was simple and eouldfind the limiting nutient quickly. It was al8o tstified that the additio of phosphate, which is echeap andeasily available, into the inluenl, could increase the CODymn removal of the biofilter, thus a new access to .improve the remnoval eficiency of organic mallers of the biological procsses was reached.1 Materials and methods1.1 Raw waterThe study was caried out in a surfaece waler plant localed in Huai River Basin in the east of China. AFoundation item: The Key Projet of“985" Foundation of Tsinghue Lnivesity; * Coreponding aubhor中国煤化工MYHCNMHGPhosphorus lnitation in biofitrtion for drinking watcr treatnent495conventional Irealntent Irain, i. e. coagulation. sedimentation， Glination and disinfection was used in thislant. The efMuent of the fll-scale sedinentation lank of the plant was used as the raw water of thebiofilters( Table l).Table 1Quality of the raw water( 2001.7.3- 2001. 10.12)CODun.NH: -N.NO:-N. NO: -N. P0:- P. Turhidity.Tenperalur.D0.plI10g/1.mg/Lm/LntuMaximum5.400.300.070.3692.005.9431.508.668.20Minimumu2.250.000.11s, 831.7319. 704.017.74Arerage3.880.180.195 01. 3.5127.156.757.911.2 Analytical methodsAll the indexes mentioned in Tabl: I were measured using Chinese national sluandard methods( Ministry of Health, R. P. China, 2001; China Standard Pres. 1996)1.3 Bacterial growth potential( BGP) methodBCP was measured mainly as the melthod described by Sathasivan and Ohgaki( Sathasivan, 1999). Intheir method, an expensive instument, epifuorescence microscope, was used for the direct tolal mierobialcouning, while in this study, the heterolrophic plate counting( HPC) was taken for the microbial counting.thus the mrasuremnent could be accomplished under luasic laboratory equipmen1. Another change was thedose of phoasphate in demining BGP( AI) and BC;P(P). in their methods. about 300 1ug/L KL,PO,_P(1320 1/L KH,PO,) was required. while this dose was only 50 pug/L KH2 P0,-P in ours, for the ltterwas high enough to reveal the phosphorus lnitation.1.3.1 Glassware preparationFor the levels of available nutrients for the microbes in the samples were very low. all the glaswareshould be lreated 幽the fllowin to be fre from pllution: (1) dipping in delergent ovemight; (2)washing by tap waler and pure water, reperctively; (3) dipping in diluted acid (HCI) solution overigh;(4) washing by tap waler and pure waler, respectively; (5) heating overnight al 210C .1.3.2 Sample preparationA 1000 nl sampling botle was used to clle enough waler sample, diferen inoganie nutrien(s)were added into the botle at the doses shown in Table 2 acording to dfferent requirements.BGP(AI) meant the bacterial growth potential of the sample int which various inonpanic nutrientsincluding phasphons were added . BCP(P) meunt the baclerial growth polentials of the sarmples into which50 pug/L KH,PO.-P were added. BCP(n) meant the bacteial growth potential of the sample into which nonoutrient was added .Table 2 Nutrients added when dfferent BGPs were measuredBCPConpoundCnncentraion. 1/LI50 mm test ubes, 10 m/ube. For each result,KNO,1011.0al Jeast one parallel was necessary. All the testKH.PO( KH:PO.P )219.4 (50.0)tubes were sterilized by autoclaving at 121 C，15Nn,S0450.0min and then stored at 4C .CaCl2 .2H20185.0MxcCly .61,0415.01.3.3 Inoculation, incubation and microbialBCP(AI) FoLCLh .6H:O245.0countingCoCl, .6H2020.4The nw water was cllected into several 18CuCl, ●212027.1mmx 150 mm test tubes al 10 m/tube as theMnS0, .5H201109.5inoculum simultaneous with sampling. The lestZnCl,(NH. )。Mo,O21 *4H20ubes were lhen incubated al 20C for 5 days loBGP(P)"_ KHPO,(KH,PO.P)ensure thal the limiting nutrients in the water hadbeen reduced to the minimum, other nutrients hadExcepriong here thaee when the dose-rreaction relationshipsphonsphorus or carbon to BCP% weoe determnined中国煤化工MYHCNMHGVal. 1596YU Xin etal. .been reduced to the minimum possible and the microbes had been in the stationary phase and adopted tothe waler .The incubaled inoculum was lransferred into the test tubes with waler samples at a dose of 2 ml/lube .The inoculaterd lubes were incubated at 20C for 5 days，then the biomass in the samples were counted withheterotrophie plale counting( HPC) and BGP was expressed as the result of HPC with CFU/ ml s the unit.1.4 The pilot-scale biofltersTwo parallel GAC sand dual media biofilters( BFI13and BF2) were set up in the walter plant(Fig.1). Both .of the filters were cylinders of 3.0 m in height and 0.15m in intemal dtianeler. The waler levels were 2.5 mfrom the bottom of the reactors . The media in thern were10.9 m high in total with 0.55 m granular activalerdcarbon( GAC) and 0.35 m sand， respectively. Ahydraulic loading of 7.5- -8.5 m/h, correponding toBaw-z14昌an empty bed contact time (EBCT) of 6. 4--7.2 min,and a backwashing cycle of 24 h were performed. Theig.1 Schemne of the pl-scale bifilersaddition of phosphate at interested concentration to BFI1. bifller "( BHI)2.lifiler 2(F):3.wter dribuor: was realized by preparing a KH2 PO, solution of a certain4. semnpling portis. granular activated carbon;6. sand;7. gravelconcentration in a tunk( Fig. 1(12)) and controling itslayer;8. orerDlow pipe;9. rugnetice lifting punup; 10. dstibulionlow，the ruw water added with phosphate could bepipe: 11. valve; 12. phosphate udding tank; 13. ronsan-lereltank; 14. efluell collector: 15. backwashing water lanksampled from a sampling port (Fig. 1(4)) and the exactPO"-P concentration could be determined.Before the study, this pair of biofilters had been in performance for several months and the stableremnovals of various pollutants by them were performed( unpublished data) .2 Results and discussion2.1 Phosphorus limitationEfects of phosphorus and the other inorganic elements on the BCPs of the raw waler were compared(Fig.2). BGP(AII) of the raw waler could rise from4.07 x 10' CFU/ml (BGP(n) ) to6.40x 10' and6.78x 10' CFU/ml. However, BCP(P) of the aw waler can rise lo6.67x 10' CFU/ml. There was nosignificant difference between BCP( All) and BCP( P), which suggested that the levels of all the inorganieelements except phosphorus were high enough in the raw waler and no one but phosphorus should havelimiting eflct on the microbial growth and metabolism in the biological process treating this raw waler.10.00 r2001-7-140 BGIP(n).8.00}6.67■BGP(p)6.67 6.40 6.78 |.00 t6.005.ss |4.4.073.484.04 3.89 4.18002.54厘2.80BGP() BGPOP) BAI)-1 BGAPAI)2”7-14 8-12 9-15 9-21 9-23 10-12Samplet, dateFig.2 Eftect of phosphorus and otherFig.3 liming efeet of phosphorusinnrganic elements on the BCPs of tbe raN water中国煤化工MYHCNMHGNo.4Phosphorus limitation in biofiltration for drinking wuter treatment497The details of phosphorus limitation could be seen in Fig.3. During the six determinations from Jul .2001 to Oet. 2001，BCP(P) was 54% higher than BGP(n) in avernge with the highest inerease of 64%and the lowest increase of 41 %，respectively.10.000Fig. 4 showed the dose- reaction relationship of phosphorus to2001-8-12BGP. The uddition of phosphate could afeet BGP even when the8.00dose was up to 1000 y/I.. BGP would rise as the level of PO:-P in .6.00the samples ruse. The shape of the dose- reaction curve was well in. 4.00agreement with Monod Equation which demonstrates the microbialspecific growth rate ( Raledge, 2001): in the lower concentration2.0area, BCP was very sensitive to the increase of phosphorus level;0200400 6001000000 while. in the higher concentration area, the sensitivity decreased,SoubleP in samples.gLi.e. the lower the phosphorus level was， the higher the BGPFig. 4 The doe-reartion rlatioutip of increase per unit amount of phosphale was.plophorus lo BGP[n other studies, the phosphorus levels were much lower,which seldom exceedled 5 prg/L (Mielinen. 1997; Markku, 1999; Nishijima, 1997; Sathasivan, 199),80 i seemned more“ reasonable" that phosphorus had limiting efet on the microbial growth andmelbolism. While, in this sludy, the average phosphorus level of the inluernt was a much higher value.15.01 1ug/L(Table 1). The reasons thal phosphorus sill played a limiling role under such a highconcentration might be mainly due to the higher level of organic mallers. In other studies mentioned above,TOC concentration was between0.7- 3.1 mg/L and CODa concentration was under 2.0 mg/L. While.in this study, the average CODx. concentation was 3.88 mg/I. and the maximum was 5.40 mg/L. So thehigher concentration made the sarmples had a higher level of biovailable orgunice mallers, which ied to ahigher phosphonus demand since a certain raio of C:P was requred in microbial growth and metabolism.2.2 lmpact of the addition of phosphorus on the removal of organic matters of the biofiter andthe biological stability of the effluentPhosphate could be added into the inluent lo promote the microbial metabolism in the biofiler andimprove the removal eficiency of the organic matters of the reactor, since phosphorus was the limitingnutrient in the rew water(Fig.5, Table 3). When nothing was added into both inluenls, there was nolsignificant diferenrce belween the removal of organic malters of BF1 and BF2. The removal curves of thelwo reaclors were interwoven; each of the10.8.120.00average CODwa removal of BF1 or BF26.was about 15%，while that of BF2 was a4.00liule higher. Ater the adii of曾200-20.000 82001-12001-7-212018-101 201830~ 201000phosphate to BFI, the C0ODR remnoval ofB' was inereased. All the points in the .10.0040.00removal curve of BF1 but only one werehigher lhan those of BF2. The average-20.00CODsn rermoval of BF1 exceeded 20%，-4000which was 6.02 perventage points higher2001-9-15 2001-9-222001-9-29 2001-106 2001-10-13,datethan that of the control one. BF2.士Baw waorar士BFI elp女BF2 efluantRemnoval fate of BF1r Removal rate of BF2The BGP(n)s of the raw waler andthe. Lwo eluents were measured Pig.s Elre of poporonon on dhe remonal od omanie matesr by Ihe biflesimultaneously for several times w Uup: when nothing was added into both inluent; Duwn: when KH2PO, was addedinvestigate the efect of phosphorus imlo the inent of BFI中国煤化工MYHCNMHG98YLXinnal.Vol.1Saddition on the biological sabiliy of lhe efluent(Fig .6) . The sequence of the BCP(n) averages was theraw waler(4.26x 10' CFU/ml) > BF2 efluet (2.08x10’CFL/m) > BFI cfluent (1.11xI0CFU/ ml).CODy.PO: P, ng/LDatenluent， ELuent of Remowal rate fuent of Rmowal nute Inluem Elben Inluet Efuemtmg/L BFI. mg/1. adfBFI. 呢BF2, mg/L o[BP2. % ofBF] odfBFI dBF2 of RF22001.7.3 2001.9.153.402.9114. 132.8516.49 20.70 9.33 20.70 11..200.9. 16-2001.1.12”..1.243.3620.56 .3.6114.54 27.45 12.88 13.69 5.90，: allthe data in this table were avenges;。+ ; during this period. nwhing was added into both inluents;，: during this period.KH.PO, was addrd into the inuenl of BFIBecause the difference of C0Ds. removal eficiency of the lwo.00 r6.153reactors was resulted from the adition of phosphale， the organic ..0mallers BFI removed more than BF2 were in fact those that could not4.00be utilized by the microbes for phosphorus limitation. Ilowever, by the3.003.06 2.88addition of phosphate into the infuent， the phosphorus limitation was岳200134 1215grealy alleviated and carbon would becore the precdominant limiling1.00nutrient，so the utilization of carbon would approach its maximum9-21 10-6. 10-9 10-12possible. thus the bioavailable organic mallers in drinking waler would口Raw wt..■ BfMuant of BFI be remarkably reduced, i.e. the level of he organic: mallens that the口EfMluent of BF2.microbes could utilize in the efluent of BF1 was less than that in theFig.6 Efer。poporor wu be efluentl of BF2, the polenial to support the bacterial growth of thehiolngiral sailiy of the eflucnt of the efluent of BFI was reduced, namely, the biological stability of thebiofiltpreffluen! of BFI was increased.The addition dose of phusphale in our experiment was carefully controlled, less han 14 14g/LKH2P02-P in average was added to the raw water of BF'I, thus the evenlual concentration of PO2- -P in itsnfluent did not exceed 28 pug/L in avenage. After the treatment of BFI， the average concentration ofPO;- -P in the efluent was 12. 88 ug/L, a ltte lower than 13 .69 ug/L, the uverage concentration of theraw water, so there would nol be the concem of phosphonus pollution.3 ConclusionsAddition of phosphorus alone could subslantially increase BGPs of the samnples. BCP(P) was 54%higher than BGP(n) in average. This effet was equivalent lo that of addition of a misture of inorganicnutrients including phosphorus , which suggested that phosphorus was the limiting nutrient for the microbialgrowth and metabolism in the biofitration treating the raw walter.Based on the above resuls, phosphate could be added into the inluent of the biofilter to improve theremoval fficieney of organic matters and the biological stability of the efluent. When nothing was addedinto the infuents, both CODk removals of the parallel biofiters wer: about 15%; however, whenphosphate was added into its influent, the CODR。 removal of the biofiler BF1 (20.56%) was 6.02percentage points higher than that of the control filter BF2(14.54%).References :Reardsley M L, CoffeyJ M. 1985. Bicaugmentation: optinizing biological wistewater trealment[J]. Pllution Engineering. 17(6):30- 33.Chins Sianderd Pres, 1996. Cllection of nationd sandurd merhnds for water quality analysis[ M]. Brjing: China Standerd Press.Mondigan M T, Maninko J M. Parcker J, 1997. Brock biology of mircobiology[ M]. Upper Sadde River, New Jesey, LSA: Prenice Hall中国煤化工MYHCNMHGNo.4.Phosphonus linitation in biofiltration for drinking water tratment499Markku J L, Mietiorn IT. Vartinen T at al.. 1999. A nrw sensirive bioaseay for detemination of mirobually availble phosphonus in waler| . Appl Eniom Microbiol, 65(5):2032- 2034.MiettinenI T. Vartisinen T. MarikninenP J 1996. Contamination of drinking maler|J。Natore. 381(20): 654- -655 .MitinenI T. Vntiainen T. Martikainen P J 1997. Phaphonus and harteinl gromth in drinking waterl JI. Appl Eniron Micobin. 63(8):3242- -3245.Minsiny of Consnetion, P.R. China. 1995. Straley on sustainable derelopmen lur Chnre water indusry in 21x rentuyUJl. Wiater andWwastewaler. 21(5):31- -35.Ministry of Hel, P.R. China, 2001. National suniny standards for dinking welter R]. Ministv of Health, P. R. Chinm. Reijig.Nihjinae w. Sbntn E. Okada w, 1997. Impovcovent of bdrgrndeton of orgunie subsrance by adition of pospors in boiopiral ativaedRalledge C, Kristianen B, 2001. Bnsie botechnohog M). Cambridge. U. K. : Canbrdge Lnivrsity Pres.Surhasivan A. OhgakiS, 1999 plicaion of new becteridl regmwh potenial mrthud for water disuibuion sytem- - a clear pridence ofposphorues lnitation.J]. Waler Resarch, 33(1): 137- - 144.ServaisP. Bilen G, Haret M c, 1987. Dermination of the bidegadablel frartion of dssvred organic matter in wateris. WalerSunR Y.1992. Pinciple of zologicad rologyl M]. Beiing; Bejng Normal Inivenits Pres.Techbanoglous G, BurtonF L, 1991. Wateter engincering teament. disposal and reuse[ M'. MrCraw lill series in water reoures andenionmetal rginerin(3rd Edin). New Yarck: Meteal and Eddy ine., MaucGraw Hil.van der Kooi D, Viser A. Hin WAM.1982. Delemining the conentruion of casils asiniablel organie eubon in drinking waer[J].」AWWWA. 74( 10):540 -545.Wang zs. linw J, 199 Drinking warr tratmet wih mrpoelued sur! M]. Bejing; Chin Atireture Idusrial Pes.(Received lur reriew May 14, 2002. Acplr June 21, 2002)中国煤化工MYHCNMHG