EFFECTS OF COPPER AND ZINC ON THE ACTIVATED SLUDGE BACTERIA GROWTH KINETICS ALBERTO CABRERO*,SARA FERNANDEZ,FERNANDO MIRADA and
JULIAN GARCIA
Chemical Engineering Department,Faculty of Chemistry,Compluten University of Madrid,28040
Madrid,Spain
(First received July1996;accepted in revid form September1997)
杜仲提取物
AbstractÐA simple experimental t-up has been propod to study the in¯uence of Cu(II)and Zn(II) on the activated sludge growth kinetics by following the rate of change on biomass concentration during batch growth experiments.Data on biomass were®tted to a sigmoidal equation providing us the main biokinetics parameters(Y x s,and m m).Inocula eded to the system was obtained from a wastewater treatment plant operating at14days cell residence time.A synthetic feed solution contain-ing2000mg/l gelatin(corresponding to1630mg/l COD)rved as source of carbon.Di erent concen-trations of Cu(II)and Zn(II)(1,5,10and20mg/l)were introduced singly in the reactors keeping all environmental parameters constant(pH,T a,basic nutrients).The combined e ects of Cu(II)and Zn(II) were determined
by mixing the metallic ions(5/5,5/10,and10/5mg/l of Cu(II)/Zn(II)respectively). Experimental data showed that Zn(II)was less toxic than Cu(II),as expresd by a slight stimulating e ect for1mg/l Zn(II).Moreover,biokinetic parameters were not adverly a ected by the prence of Cu(II)up to a concentration of5mg/l.However,a concentration of10mg/l and higher,caud rious upts in the system.Combined e ects of copper and zinc on the activated sludge growth kinetics,indi-cated that the two heavy metals acted neither synergistically nor antagonistically.The sigmoidal model very well®ts the experimental data and could be ud in a simulation study.#1998Elvier Science Ltd.All rights rerved
Key wordsÐactivated sludge,growth kinetics,heavy metals,wastewater,sigmoidal equation
NOMENCLATURE
MLSS=mixed liquor suspended solids,g/l
COD=chemical oxygen demand,mg/l
C x(t)=population density at each moment,g/l
C x m=maximum population density,g/l
C x0=initial population density,g/l
S(t)=substrate concentration with time,mg/l
S0=initial substrate concentration,mg/l
t m=half time to reach this speci®c growth rate,h回看射雕处
r max=maximum growth rate of biomass,g/lÁh
Y x s=biomass yield coe cient
m m=maximum speci®c growth rate,hÀ1.
INTRODUCTION
Even though complex chemical manufacturing process can be operated e ectively,consistent suc-cessful wastewater treatment has been slow to develop,and in some respect has not kept up with advances in manufacturing technology.
For a long time,operational parameters for wastewater systems were bad on empirical optim-izatio
n.But before improving the biodegradation process either by better designing treatment plants or engineering more e cient microorganisms,the limits of microbial degradation must be identi®ed.
Since microorganisms are key components for de-composition of organic matter,metal toxicity towards microorganisms has received special atten-tion in recent years,not only becau of the en-vironmental incidence that heavy metal discharges into a receiving water cour can generate,but also becau an important reduction in e ciency of bio-logical wastewater treatment can occur.In natural ecosystems,a variety of factors probably alter the shape of growth curves.The factors may include predation by protozoa,the time for induction of the active organisms to the accumulation of toxins produced by other microorganisms depletion of inorganic nutrients or growth factors,the prence of other substrates that may repress utilization of the compound of interest,and binding of the com-pound to colloidal matter.The impacts or inter-actions of such potentially important factors may make it di cult to predict the activated sludge growth kinetics.
Although the mechanisms by which heavy metals a ect biological treatment process are not well de®ned,the general respon to varying concen-trations is well documented(McCarty,1964; Neufeld and Hermann,1975;Bagby and Sherrad, 1981;Sujarittanonta and Sherrard,1981;Chang et al.,1986).
Wat.Res.Vol.32,No.5,pp.1355±1362,1998
#1998Elvier Science Ltd.All rights rerved
Printed in Great Britain
0043-1354/98$19.00+
0.00
PII:S0043-1354(97)00366-7
*Author to whom all correspondence should be addresd
[Fax:+34-1-3944243].
1355
McDermott et al.(1963)studied extensively the e ect of copper on the activated sludge system.In this
way,a copper sulphate solution was added to the system to provided10to25mg/l Cu(II)concen-trations.In every ca,reduction in removal e -ciency of the BOD or COD was less than4%. Moulton and Shumate(1963)emphasized the im-portance of acclimation and indicated that an accli-mated activated sludge was not adverly a ected by45mg/l Cu(II).Similarly,Neufeld and Hermann (1975),demonstrated that it was possible to achieve a satisfactory level of e uent water quality by vary-ing the sludge age in an activated sludge system that has been previously acclimated to heavy metals.
A study of the e ect of organic load on the tox-icity of copper to the activated sludge process was developed by Salotto et al.(1964).In this study,a relationship between organic load and copper tox-icity in the activated sludge process was estab-lished.They reported that moderate variations of organic loading did not markedly alter the copper toxicity.
In1965,the results of an extensive bench and ®eld study of the activated sludge process with chromium,copper,nickel and zinc were published by the U.S.Public Health Service.In the studies, changes in chemical oxygen demand(COD),bio-chemical oxygen demand(BOD),suspended solids (SS)and turbidity were ud as indications of the e ects of heavy metal toxicity.The report indicated that concentrations of the above metals either singly or in combinations of up to10mg/l in in¯uent wa
stewater would produce at most a5%reduction in treatment e ciency in a continuous activated sludge plant.
Dilek et al.(1991)reported that the substrate removal e ciency was not adverly a ected when di erent copper concentrations(0.5to10mg/l) were dod to a chemostat unit treating a synthetic wastewater.Moreover,a slight increa in the bio-mass yield can be detected up to a concentration of 10mg/l Cu(II).
The discusd rearches studying metal toxicity on activated sludge are related to the acclimatiz-ation period of microorganisms to heavy metals. However,wage contaminated by heavy metals come from a wide variety of industrial sources,and their complex mixtures varying along the time. Changes in the metal concentration make it di cult to achieve a complete acclimation of the microor-ganisms to the new medium.
Other important aspect that has to be considered for an optimal operation of a waste treatment plant receiving heavy metal solutions is the period of time necessary to the attainment of the biokinetic par-ameters.In this way,a quick determination of the parameters is desirable in order to achieve an opti-mal operation of an aerobic biological process. However,in most of the rearch discusd abov
e,the period of time necessary to gather all data required for the experiments varied from1to 4weeks.
In this paper,a simple experimental t-up was ud to determine the cau by which heavy metals copper and zinc either singly or in combination a ect microbial batch-growth curves in respon to shock dod metal concentration in the medium. Moreover,no more than3days were spent for gathering all data required in the experiments.The ®tting of biomass production data to the sigmoidal equation in prence of di erent metal concen-trations allowed the determination of the main bio-kinetic parameters.The system ud in this rearch was constituted by two independent units that oper-ated under de®ned environmental conditions(pH and temperature)and at di erent metal concen-trations.Data obtained were recorded and com-pared to a reference reactor charged with the same synthetic wastewater but without copper and zinc.
Theoretical considerations
Several formulations have been purpod to express the growth dynamics of a microbial popu-lation that is only limited by a particular substrate (Powell,1967;Koch,1982),but the most popular kinetic expression ud today is the Monod equation.This relation is bad on the assumption that only the total amount of the biomass in the culture is su cient to specify the activities of the microorganisms.T
his model can be simpli®ed to others bad on various assumptions found in the literature(Monod,1950;Ener et al.,1983; Robinson and Tiedje,1983).However,Monod equation is referred to a single type of microorgan-ism which grows in a well-de®ned medium where all nutrients are in excess except one(carbon source) which acts as limiting substrate.Investigators typi-cally®tted biomass formation data to the integrated form of the Monod equation describing biomass production using a systematic procedure,which involved numerical adjustment of the biokinetic parameter,until a good®t could be obtained. However,in practice,Monod model cannot provide an adequate description of microbial growth beha-vior for this particular ca becau of the mixed culture character of the activated sludge in which a variety of factors probably alter the shape of mi-crobial biomass curves.Also,a simple sigmoidal equation1had to be ud to accurately describe the di erent steps in the microbial growth(lag pha, exponential pha and stationary pha):
C x t
C x0ÀC x m
1 t a t m m mÁ2t m
C x m 1
where C x(t)is the population density at each moment(MLSS),C x0is the initial population den-sity,C x m is the maximum population density achieve,m m is the maximum speci®c growth rate
Alberto Cabrero et al. 1356
and t m is the half time necessary to reach this speci®c growth rate.
Initial estimates of m m and t m were entered into a nonlinear regression computer program that ®ts data obtained to equation (1)by the Marquardt method.The program estimated,additionally,the standard errors of parameters.Then,true yield coe cient Y x s ,may be obtaining from equation (2):
Y x s t
C x t ÀC x 0S 0ÀS t
2
equation (2)can be simpli®ed to equation (3)when the substrate concentration is nearly exhausted (S (t )I 0):
Y max
C x m ÀC x 0
S 0
烟花绚烂3
where Y max is the maximum true yield,C x m is the maximum biomass concentration and S 0is the in-itial substrate concentration (referred to COD).Moreover,values for maximum growth rate (r max )were estimated from numerical di erentiation of the curves to allow us extract additional information.
MATERIALS AND METHODS
如什么似什么的成语
Experimental batch-growth system
The schematic diagram of the batch growth system ud in the inhibition kinetics of the activated sludge process due to the prence of copper and zinc is shown in Fig.1.The equipment was constituted of 10Erlenmeyer ¯asks acting as growth reactors.The reactors were placed into a thermostatic bath of plexiglass material to remain oper-
gly是什么氨基酸ating temperature constant at 258C.The operating volume for each reactor was of 2litres.
Compresd air was ud to supply oxygen to each ¯ask through porous di ur stones.The air also rved to pro-vide good mixing of suspended solids.An adequate volu-metric air ¯ow rate was supply in order to maintain saturated oxygen conditions in the medium.There was and additionally magnetic stirrer module that achieved an e ective mix in the reactors.
Bacterial growth medium
The same synthetic wastewater was ud throughout the experiments.Components of wastewater ud in this inves-tigation are listed in Table 1.In this solution,gelatin rved as the main carbon and energy source.Phosphate bu er was added as a phosphorous source,as well as to maintain a stable pH.A solution of NH 4Cl was added to the medium to provide the source of nitrogen.All other micro-nutrients were added in su cient quantities to make carbon the growth-limiting substrate.Metals studied (cop-per and zinc)were added as sulphate salts to provide a concentration of 1,5,10and 20mg/l in the reactors.The combinations of the metals were 5/5,5/10and 10/5mg/l of copper and zinc respectively.All inorganic salts were of analytical reagent grade.
Table 1.Composition of wastewater solution fed to the reactors Constituent2岁宝宝早餐
Concentration (mg/l)
Gelatin from porcine skin a 2000K 2HPO 41750KH 2PO 4250NH 4Cl
250Other micronutrients
Ðb
a COD of waste =1630mg/l.
b
Micronutrients were supplied using tap water as
diluent.
Fig.1.Schematic diagram of the batch-growth system.
新加坡辣椒蟹
Activated sludge bacteria growth kinetics 1357
Analytical techniques
Microbial growth evolution was carried out by monitor-ing the MLSS concentration along the time in a VARIAN-CARY 1E spectrophotometer.During batch-growth,liquid samples were withdrawn and centrifuged immediately to parate bacteria from the liquid medium.Depending on turbidity,samples were diluted with water by a factor of 2±8and then centrifuged at 4000rpm for 20min to precipitate the bacteria.After centrifugation,the supernatant was removed,cell pellet was resuspended into water and cell density was then determined by optical den-sity at 450nm.Optical density measurements of the acti-vated sludge culture were converted into MLSS concentration by means of this calibration curve:[MLSS]=À2.118(20.941)+548.74(24.68)ÁABS,where ABS is the absorbance value and [MLSS]is the mixed liquor suspended solids concentration (expresd as mg biomass/l).The coe cient of correlation for this linear re-gression was r 2=0.998.Copper and zinc ion concen-trations were determined using an atomic absorption spectrophotometer.Experimental procedures
Activated sludge inocula was obtained from a biological treatment pilot plant.This plant was operated under steady-state conditions with a MLSS concentration of 2200mg/l.Steady-state was identi®ed by invariant total biomass concentrations over a month.A 350ml/d purge was done from the aerated reactor to maintain a desired mean cell residence time value of 14days.This cell resi-dence time was ®xed according to the studies performed by Bagby and Sherrad (1981).Components of the waste-water fed to the biological treatment pilot-plant are listed in Table 2.
In order to achieve optimal experimental conditions in the medium,some batch growth experiments were made at di erent pH,temperature,suspended solids'inocula con-centration (MLSS)and limiting substrate concentration (gelatine).
Once the optimal conditions in the medium were achieved (pH =7.5;T a =258C;MLSS inocula =0.025g/l;limiting substrate =2g/l),metal toxicity studies were per-formed by measuring the total suspended solids concen-tration along the time.During the experiment,the sludge was suddenly shock dod to 1,5,10and 20mg/l of the appropriate heavy metal added in a soluble sulfate form.Immediately after dosing of the sludge and for about 3days thereafter,MLSS measurement was made by sampling (4ml)the mixed content of the di erent ¯asks at intervals of 2h.The pH measurements in the reactors were made at the beginning and at the end of the exper-iments.It was
controlled at 720.2with a phosphate buf-fer solution.
Accordingly to other studies,(Sierp and Franmeier,1933;Pettet,1956)an increa in the growth medium tur-bidity was obrved after 2days.An explanation of this phenomenon was given by Lombran a et al.(1993).In this study,discrepancies between data for biomass yield and speci®c rate of substrate removal (referred to COD)em to be caud by the fact that not all of the MLSS increa
is a real and active biomass increa.They consider that a considerable amount of soluble and suspended COD is trapped in the ¯ocks but not digested,producing an apparent biomass growth.The problem described was solved by using growth solution supernatant obtained after centrifugation of the suspended solids as a blank.
RESULTS AND DISCUSSION
Biomass production curve and biokinetic parameters Due to the mixed culture character of the acti-vated sludge,they prent a great variance in toler-ance towards the di erent metal studied.In this study,it has been found that relatively lowered con-centrations of certain heavy metals could promote a stimulating e ect on the microbial population growth,as re¯ected by an increa in biomass yield.T
his phenomenon is consistent with obrvations made by other authors and will be discusd later.Converly,for higher metal concentrations (5mg/l),microbial growth began to be inhibited,as re¯ected by a decrea in biokinetic parameters values.
The study allowed to make qualitative and mi-quantitative asssments of the in¯uence of copper and zinc,singly or in combination,on activated sludge growth kinetic parameters by ®tting data obtained to the previously established sigmoidal equation.Figures 2±4show comparisons between the data and the ®tted curves (drawn as solid lines).The estimations for the parameters of the logistic model at di erent metal concentrations are sum-marized in Table 3.The adequate ®tness of the data obtained to the logistic equation con®rms a good choice of the model,which states that the prence of metals a ected both growth rate and biomass yield.
The study of the individual and combined e ect of copper and zinc on the activated sludge growth phenomena was carried out under the predeter-mined conditions described above.All the exper-iments were repeated three times.Included below is a prentation of the most signi®cant results of the experiments.
Control.The experimental t-up was divided into two identical units,each one was constituted of ®v
e Erlenmeyer ¯asks.One of them was charged with the same constituents that the others but void of metals,operating at the same pH and temperature as the others.Control reactor rved to obtain ba-line data to compare the toxic e ect produced by the addition of di erent amounts of heavy metals.Microbial growth curve was characterized by a non-exponential increa for the ®rst 20h.There-after,the culture in the reactor was considered ac-climatized to the new medium.In this way 48±50h were necessary to reach the stationary pha,in which the nutrients were nearly exhausted.It was taken about three days to obrve the ®rst evidence of endogenous metabolism accordingly with that
Table 2.Composition of wastewater solution fed to the pilot-plant Constituent
Concentration (mg/l)
D -(+)-gluco a 500K 2HPO 41750KH 2PO 4250NH 4Cl
250Other micronutrients
Ðb
a COD of waste =430mg/l.
b
Micronutrients were supplied using tap water as diluent.
Alberto Cabrero et al.
1358
found by Herbert (1958).It should be noted on Figs 2±4that the sigmoidal equation ud in this study did not describe the last part of the growth curve (declining pha),however,declining growth pha was not decisive for the attainment of the kinetic parameters.Values of kinetic parameters for the control reactor are shown in Table 3.The results of biomass yield,Y x s ,obtained were in agreement with tho found by Bagby and Sherrad (1981).不斩相思不忍顾
E ect of zinc.The time cour data at four di er-ent zinc concentrations ranging from 1to 20mg/l are shown in Fig.2.This ®gure showed that there was no signi®cant change neither in the acclimation pha nor in the exponential pha for all the metal concentrations studied if compared with the refer-ence reactor.Thus,in comparison to the control reactor,a lower MLSS concentration was obrved for all the experiments except that carried out with 1mg/l of zinc,where a slight stimulating
e ect was detected.When the amount of zinc was incread to 5mg/l a small drop in biomass yield occurred.For 10mg/l Zn(II)a sharply decrea on biomass pro-duction (15%)was obrved,while higher amounts of zinc (20mg/l)caud similar e ects.This ems to indicate that 10mg/l of zinc are more than enough to inhibit sludge growth,however,up to this metal concentration,a precipitation
and/or
Fig.2.Batch-growth experiments in prence of Zn(II).Symbols show the data and curves show model
predictions.
Fig.3.Batch-growth experiments in prence of Cu(II).Symbols show the data and curves show model
predictions.
Activated sludge bacteria growth kinetics 1359
