Water Research 38 (2004) 1081–1088

ARTICLE IN PRESS

*Correspondi

System Engine

Kamitomioka 1

E-mail addre

0043-1354/$ - see

doi:10.1016/S004

Physico-chemical factors affecting the E: coli removal in arotating biological contactor (RBC) treating UASB effluent

Ahmed Tawfika,b,*, Bram Klapwijkb, Joost Van Buurenb,Fatma El- Goharya, Gatze Lettingab

aWater Pollution Control Department, National Research Centre, El-tahrir St., Dokki, Giza, EgyptbAgrotechnology and Food Sciences Department, Sub-department of Environmental Technology, Wageningen University and Research

Centre, P.O. Box 8129, 6700EV Wageningen, The Netherlands

Received 20 April 2001; received in revised form 13 May 2003; accepted 23 May 2003

Abstract

The removal mechanism of E. coli from UASB effluent using a Rotating Biological Contractor (RBC) has been

investigated. Preliminary batch experiments in a RBC indicate a first-order removal kinetics. Variation in the dissolved

oxygen concentration and E. coli counts over the depth of the RBC has been recorded and indicates that the RBC is not

a completely mixed reactor. Therefore batch experiments were carried out in a beaker where the different operating

conditions can be controlled.

Factors affecting the removal of E. coli via a biofilm system as stirring, dissolved oxygen concentration, pH, and

addition of cationic polymer were investigated. The results obtained indicated that the most important removal

mechanism of E. coli in the biofilm is the adsorption process, followed by sedimentation. Die-off is a relatively minor

removal mechanism in an RBC system. Higher removal rate of E. coli was observed in an aerobic compared to an

anaerobic biofilm system. Variation of dissolved oxygen concentration (3.3–8.7mg l�1) and pH-values between 6.5 and

9.3 did not exert any significant effect on the removal rate of the E. coli by the heterotrophic biofilm. A rapid adsorption

of E. coli to the biofilm occurred during the first days after adding the cationic polymer, after which the adsorption

slowed down.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: UASB effluent; RBC; E. coli; Removal rate; Die-off; Adsorption

1. Introduction

During recent years significant attention has been

focused on post-treatment of effluent of Up-flow

Anaerobic Sludge Blanket (UASB) reactor using a series

of algal ponds. Special emphasis has been devoted to

factors affecting the removal of pathogens. E. coli is

used as indicator of these pathogens in the pond systems

[1]. Much less attention has been given to the factors

ng author. Department of Environmental

ering, Nagaoka University of Technology,

603-1, Nagaoka, Niigata, 940-2188, Japan.

ss: tawfik8@hotmail.com (A. Tawfik).

front matter r 2003 Elsevier Ltd. All rights reserve

3-1354(03)00345-2

affecting the pathogen removal in other post-treatment

technologies such as the RBC system. In earlier research,

we found that the RBC is effective for the reduction of

COD, ammonia and E. coli from UASB-effluent [2].

Since the removal of pathogenic bacteria is an important

objective in post-treatment, optimization of the RBC

performance for this purpose is of prime importance.

To optimize the removal of E. coli from UASB

effluent using a RBC system, adequate understanding

of factors affecting the removal mechanism of E. coli is

essential.

The removal of E. coli in an RBC unit can be seen as

the sum of the following mechanisms: die-off, sedimen-

tation, adsorption and filter feeding. According to Crane

d.

ARTICLE IN PRESSA. Tawfik et al. / Water Research 38 (2004) 1081–10881082

and Moore [3] the die-off of E. coli can be described by

the following exponential equation:

Ntt ¼ N0 exp� kdt;

where N0 is the influent E. coli count (x/100ml), Nt the

effluent E. coli count (x/100ml), Kd the first-order

removal rate constant (d�1) and T is the time or

detention time (d�1).

In our research, we found that the total removal of E.

coli in an RBC can also described by an exponential

equation:

Ntt ¼ N0 exp� krt

In which Kr is the first-order removal constant.

Generally, the die-off rate of pathogens is influenced

by factors such as, dissolved oxygen concentration [4],

depth [5] and mixing in the treatment reactor [6].

Polprasert et al. [7] found for an algal pond that

increasing the retention time not only led to decrease of

bacterial concentrations, but also resulted in changes in

the pond environment, such as variation of dispersion

number, biomass concentration, pH and nutrients

availability, which all influence the die-off process. Mills

et al. [8] also pointed out that the rate of bacterial die-off

increases significantly at pH values exceeding 9.0.

Saqqar and Pescod [9] found that the faecal coliform

die-off rate increases with increasing temperature, pH

and with decreasing total BOD5, soluble BOD5 and

surface organic loading rate.

Earlier studies carried out by us [2] indicated that

suspended E. coli (>4.4 mm) in the UASB effluent

constitute 11–49% of total suspended solids. Therefore,

it was concluded that one of the removal mechanisms in

the RBC is sedimentation.

Pathogenic bacteria can be considered as living

colloidal particles. Usually they have a net negative

surface charge in the pH range of natural waters. Also

the biofilm is negatively charged in this pH range.

Despite that it is still possible that the pathogenic

bacteria can adsorb to the biofilm. The Extra-cellular

Polymeric Substances (EPS) provides the possibilities

for adsorption of pathogenic bacteria to the biofilm.

When a pathogenic bacteria approaches a biofilm, one

and the same polymer molecule may attach to both the

surfaces of the organism and the biofilm, thereby

forming a ‘‘bridge’’. Banks and Bryers [10] reported

that a biofilm on different media such as glass, poly

carbonate, and granular activated carbon surfaces

enhanced removal of pathogenic bacteria. A large

variety of different heterotrophic bacteria (including

potentially pathogenic bacteria) have been isolated from

biofilms [11]. According to Cunningham et al. [12]

increasing the thickness of the biofilm, lead to an

increase in pathogenic bacteria removal.

Filter feeding is the process of removing particles by

protozoa and metazoa. Several researchers suggested

that flocculation and/or predation by protozoa

represent important factors for the removal of E. coli.

Heukelekian and Roudolfs [13] already presented

circumstantial evidences, suggesting the responsibility

of ciliated protozoa for the removal of E. coli. Curds

and Fey [14] demonstrated that protozoa play an

important role in the removal of E. coli in an

activated sludge system. Kinner and Curds [16]

showed for RBCs that, there was an overall increase in

numbers of protozoa and metazoa as the loading rate

decreased.

We focused in our study on the removal mechanism of

E. coli in RBCs with a fairly high loading rate and as a

result of the overall number of potozoa and metazoa

was low indictaing that the filter feeding mechanism

was of minor importance. Therefore the aim of this

research was to study the importance of the different

removal mechanism, as die off, sedimentation and

adsorption. Beside that, we investigated the effect of

dissolved oxygen concentrations, pH-values, and catio-

nic polymer addition on the E. coli removal rate

constant.

2. Material and methods

Two batch experiments were conducted in this

investigation: (1) RBC experiments and (2) Beaker

experiments. The effluent of a pilot scale UASB

reactor originally used by Grin et al. [17] which was

also used for this study. The UASB reactor was fed

with domestic sewage (COD=500mg/l, TKN= 65mg/l

and E. coli =4.4� 106/100ml) collected in the

combined sewer system of the village Bennekom, The

Netherlands.

2.1. Rotating biological contactor (RBC)

The schematic diagram of the pilot RBC is shown in

Fig. 1. The RBC system with a working volume of 60 l

was equipped with 10 polystyrene foam disks with a

total effective surface area of 6.5m2. The disk diameter

was 0.6m with a thickness of 0.02m and the discs were

spaced at 0.02m distance and operated at 5 rpm. The

submerged surface amounted to 40%. The disks were

mounted on a steel shaft.

Prior to the batch experiments, the RBC has been

continuously fed with UASB effluent for a period of 9

months. The system was operated at an organic loading

rate of 51 gCOD/m2 d and hydraulic retention time of

1.25 h. The thickness of the biofilm amounted to 2mm.

There was no protozoa in the 1st stage of the two stage

system, while dominating in the 2nd stage of the two

stage system (Trachelophylum Pusillum, Paramecium

Candatum and Vorticella).

ARTICLE IN PRESS

Fig. 1. Schematic diagram for RBC and beaker batch experiment.

Table 1

Batch experiments to elucidate the mechanism of E. coli

removal

Exp.

no.

Conditions in beaker Main presumed removal

mechanism

1a Stirring, no bio-disc Die-off

1b No stirring, no bio disc Die-off+sedimentation

1c Stirring, with bio-disc

(0.13� 0.13m2)Die-off+adsorption

A. Tawfik et al. / Water Research 38 (2004) 1081–1088 1083

2.2. RBC batch experiment

In the first series of RBC batch experiments, we

investigated the effect of the reaction time on E. coli

removal. The system was filled at t=0 with 60 l UASB

effluent. Samples were taken at t=0, 1, 2, 3, 4 and 5 h.

All samples were taken at a depth of 0.01m after filling

the RBC at a constant water level.

The second series of RBC batch experiments, the E.

coli and dissolved oxygen concentrations were measured

at water depths of 0.01, 0.1, 0.2, 0.28, 0.35 and 0.45m in

the pilot-plant after a detention time of 2 h.

Both experiments were repeated four times.

2.3. Beaker batch experiment

The beaker batch experiments were conducted in a 2-l

beaker which enables a better control of the different

operating variables.

Beakers of 2.0 l capacity were used for this set of

experiments (Fig. 1). The beakers were filled with

1900ml of UASB effluent, isolated from light, controlled

at a temperature of 20�C and stirred at a constant speed

of 5.0 rpm. Segments of the polystyrene foam disk from

the RBC (0.13� 0.13m) were placed in the beakers filledwith UASB effluent. In some of the control experiments

no segment disks were used.

1. The mechanism of E. coli removal: To identify the

mechanism of the removal of E. coli by the biofilm, the

following experiments (Table 1) were carried out for a

7.0 days experimental period. The experiment was

repeated two times.

2. Factors affecting die-off and adsorption of E. coli:

Both the die-off and adsorption of E. coli on the biofilm

were investigated as a function of time (Table 2). Other

variables investigated included the effect of the dissolved

oxygen, pH, and addition of cationic polymer.

3. Analysis: Assessment of E. coli was performed

according to the method described by Havelaar et al.

[18]. Statistical analysis was done according to Snedecor

and Cochran [19]. The results of each experiment was

assumed to be independent with different variance.

3. Results and discussion

3.1. RBC batch experiments

Effect of reaction time on E:coli removal: Results of

the residual E. coli count as a function of reaction time

in the RBC system of one of the experiments are

presented in Fig. 2. The reduction in E. coli content can

be described by an exponential equation. The removal

constants Kr for E. coli for the four experiments are

presented in Table 3. The average removal rate constant

is 0.77 d�1 (70.5).The distribution of E:coli and dissolved oxygen in the

RBC system: The E. coli counts and dissolved oxygen

concentration along the depth of the RBC system after a

constant reaction time of 2 h are presented in Fig 3.

From the data it can be seen that dissolved oxygen

concentration is lower at the bottom, whereas the E. coli

count was higher at the bottom of the reactor as

ARTICLE IN PRESS

Table 2

Batch experiments for study of factors influencing die-off and adsorption

Exp. no. Experimental conditions Main presumed removal

mechanism

2a Effect of dissolved oxygen on E. coli removal through adsorption and die-off at a constant pH of 7.070.5

Four stirred beakers with bio-disc and kept under controlled aeration at D.O of

3.3, 6.2, 7.3, and 8.7mg/l

Adsorption+die-off under

aerobic conditions

Stirred beaker, without bio-disc under anaerobic conditions Die-off under anaerobic

conditions

Stirred beaker, with bio-disc under anaerobic conditions (the bio-disc was

maintained under anaerobic conditions for two weeks before starting the

experiment)

Die-off+adsorption under

anaerobic conditions

2b Effect of pH on E. coli removal through adsorption and die-off at constant dissolved oxygen of 7.3mg/l

Four beakers equipped with bio-discs and stirring devices were aerated to control

D.O. concentration at 7.3mg/l throughout the experimental period of 5.0 days at

different pH-values of 6.5, 7.5, 8.4 and 9.3 (using 10% conc. NaOH and 10% conc.

HCl at intervals of 8.0 h)

Die-off+adsorption

2c Effect of the addition of different doses of cationic polymer (HMW 492). All beakers were equipped with bio-discs and stirred.

D.O concentration and pH-value were kept constant at 3.3mg/l and 6.9, respectively

Beaker without polymer addition (blank) Die-off+adsorption

Beaker with 1mg cationic polymer Die-off+adsorption+coagulation

Beaker with 2mg cationic polymer Die-off+adsorption+coagulation

Beaker with 3mg cationic polymer Die-off+adsorption+coagulation

Fig. 2. The effect of reaction time on the removal of E. coli concentration in a batch experiment (D.O=2.1, pH=7.1 and rotating disc

speed=5 rpm)

Table 3

Removal constants of E. coli ðKrÞ

Run no. Run 1 Run 2 Run 3 Run 4 Average

Kr (d�1) 0.46 0.78 1.5 0.36 0.7770.5

A. Tawfik et al. / Water Research 38 (2004) 1081–10881084

compared to samples collected from the top. These

results indicate that the RBC system is not a completely

mixed system. Therefore, we decided to use a completely

mixed beaker for further experiments.

3.2. Beaker batch experiments

These series of experiments were divided into two

groups, the first was devoted to assess the mechanism of

E. coli removal and the second to identify factors

affecting E. coli removal.

3.2.1. Die-off, sedimentation and adsorption of E: coli to

the segment disk

The results presented in Table 4 reveal that the Krvalues for E. coli in the beaker tests equipped with a

ARTICLE IN PRESS

Table 4

Effect of different conditions on the E. coli reduction

Experimental conditions Kr (d�1) (day 3–7) Kr (d

�1) (day 0–7) Possible mechanism

Stirring, no bio-disc 0.2970.15 0.4870.34 Die-off

No stirring and no bio-disc 0.3670.2 0.8670.4 Die-off +sedimentation

Stirring +bio-disc 0.7870.22 1.170.48 Die-off +adsorption

Fig. 3. E. coli concentrations and dissolved oxygen at different depths in the RBC.

Fig. 4. Effect of dissolved oxygen on the E. coli removal rate in a batch experiment with a section of a bio-disc.

A. Tawfik et al. / Water Research 38 (2004) 1081–1088 1085

stirring device and a disk segment of the RBC had the

highest values, whereas the beaker with stirring and no

disk segment had the lowest. The Kr in the beaker with

no stirring and no disk segment was in between.

Based on these data, it could be concluded, that the

most important removal mechanism is adsorption

followed by sedimentation. Apparently die-off appears

to be of relatively minor importance in the removal of E.

coli in the RBC system.

3.2.2. Factors affecting the E: coli removal by a

heterotrophic biofilm

Effect of dissolved oxygen at constant pH-value of

7.070.5: The E. coli removal constant was very low

(0.1 d�1) for the beaker with anaerobic conditions,

whereas in the aerobic beakers with DO from 3.3 to

8.7mg/l the removal constant was significantly higher at

level 5% (1–1.5 d�1) (Fig. 4).

Hanes et al. [20] reported that the die-off of

streptococci and coliforms was higher at oxygen

concentrations corresponding to those of a normal

non-polluted water course (7.6–8.0mg/ml) than at lower

dissolved oxygen concentrations. According to Pearson

et al. [21] these exists a negative correlation between

dissolved oxygen and the number of faecal coliforms in

treated wastewater.

Effect of pH on the E:coli removal: The effect of pH on

the E. coli removal follows from the results is presented

in Fig. 5. The DO was 7mg/l during that experiment.

Apparently the optimum pH-value for the removal of E.

coli is around 7.4. However the removal rates at higher

(till 9.4) and lower (till 6.5) are not significantly lower.

Thus, the pH has no effect in the range of 6.5–9.4.

Effect of cationic polymer addition (HMW 492): The

effect of addition of a cationic polymer at concentra-

tions ranging from 0 to 3mg l�1 to the stirred beakers

with a disk segment on E. coli removal is shown in

Table 5. From the available data it can be seen that the

addition of cationic polymer indeed improves the

removal constant of E. coli during the period from 0

to 4 days. With 3mg/l cationic polymer, the kr value of

2.41 d�1 which is higher than that in the beaker without

polymer addition. Agglomerated particles (attached

with E. coli), formed as a result of coagulation, can

easily be adsorbed by the heterotrophic biofilm. How-

ever, the removal constant over the period from 4 to 7

days indicate a clear decline in the removal rate at all

doses of cationic polymer.

ARTICLE IN PRESS

Fig. 5. Effect of pH on the E. coli removal in batch experiments with sections of bio-discs (measuring during 0–5 days).

Table 5

Effect of cationic polymer (492 HMW) addition on the E. coli

removal in a batch experiments with bio-disc sections at neutral

pH and D.O of 3.3mg l�1.

Cationic polymer dose

(492 HMW)

Kr (d�1) (day

0–4)

Kr (d�1) (day

0–7)

Bio-disc without polymer

addition

1.36 1.06

1mg l�1 1.46 1.22

2mg l�1 1.73 1.29

3mg l�1 2.41 0.94

A. Tawfik et al. / Water Research 38 (2004) 1081–10881086

4. General discussion

The mechanism of E. coli. removal in an RBC system

at a high loading rate is a combination of physical and

biological processes. Physical processes include adsorp-

tion, sedimentation and the biological removal mechan-

isms antibiosis, predation, attack by lytic bacteria and

natural die-off. Tawfik et al. [22] found that the 1st stage

of the three stage RBC system achieved 91.8% removal

for E. coli as a result of sedimentation and adsorption of

E. coli attached on the suspended solids. There was no

protozoa and metazoa in the 1st stage. Whereas in the

2nd and 3rd stage system, the protozoa and metazoa

was dominated such as (Trachelophylum Pusillum,

Paramecium Candatum, Vorticella and Nematoda).

Therefore, they concluded that filter feeding is the most

relevant mechanism in removal of E. coli in the 2nd and

3rd stage of the RBC system, where freely colloidal E.

coli are easily predated by these organisms. Stevik et al.

[23] found that the physical factors are the most

important for the removal of E. coli in the biological

filters for wastewater purification. Reduced grain size,

hydraulic loading rate and increased specific surface area

of the grains significantly reduced transport of E. coli.

Chemical factors such as pH, cation exchange capacity

and wastewater ionic strength showed less significant

effects. The biological processes such as filter feeding is

mainly affected by the applied OLR. Curds [25]

demonstrated that ciliates play an important role in

the removal of dispersed bacterial growth by predation

in an RBC system. Predation includes the bacteriovor-

ous activity of nematodes, rotifers and protozoa and was

considered by Green et al. [26] as an important factor for

the removal of bacteria from wastewater’s in con-

structed wetlands. In this study, we focused on physical

processes.

Results of the present study indicate that increasing

the reaction time has a clear positive impact on E. coli.

removal (Fig. 2). In a previous investigation [2] we found

that increasing the HRT from 2.5 to 10 h in a continuous

RBC system, treating UASB effluent, increased the E.

coli removal from 89% to 99.5%. It was also found that

die-off in the aqueous phase of an RBC plays only a

minor role.

The Kr values of the samples which, were allowed to

settle were significantly higher than the stirred samples

(Table 4). This indicates that sedimentation is one of the

mechanisms responsible for the removal of E. coli from

anaerobically pre-treated sewage. Previously we found

[2] that the major part of E. coli are present in suspended

particles >4.4mm. These are removed by sedimentationor adsorption in the biofilm already in the first stage of a

RBC (99.66%). The colloidal E. coli present in the range

of smaller particles (o4.4 >0.45mm) become adsorbedin the 2nd stage of RBC (99.78%). Also, Milne et al. [27]

found that the survival of E. coli is related to suspended

solids. The higher removal rate of E. coli in the presence

of a heterotrophic biofilm can be attributed to enmesh-

ment in, and/or adsorption of E. coli to the biofilm. The

adsorbed E. coli cells in heterotrophic biofilm may

become degraded by lytic processes and by predation

through protozoa. The relatively long cell residence time

and the aerobic conditions prevailing in the RBC could

then cause further die-off of the attached E. coli.

According to Raman and Chakladar [28] this could

take place even under anaerobic conditions. Ueda and

Horan [29] found that a membrane alone demonstrated

a poor bacteriophage removal efficiency, but removal

efficiency increased as a good biofilm had developed on

the membrane. These researchers proved that the biofilm

accumulating on the surface of a membrane contributed

significantly to phage removal. There are a number of

explanations for the role of the biofilm in pathogenic

ARTICLE IN PRESSA. Tawfik et al. / Water Research 38 (2004) 1081–1088 1087

bacteria removal. The physico-chemical effect of the

biofilm on pathogenic bacteria removal could be due to

adsorption or entrapment to bacterial cells and extra-

cellular polymeric substances. Subsequently, there will

be biological predation of pathogenic bacteria by other

microorganisms. Van der Drift et al. [30] demonstrated

that the removal of E. coli from wastewater treated with

activated sludge is a bi-phasic process. First a rapid

sorption of bacteria to the sludge flocs takes place,

followed by a slower elimination of bacteria, which is

presumably due to predation by ciliated protozoa. The

results obtained by Omura et al. [31] indicate that the

removal of coliform bacteria, enterococcus bacteria, and

coliphages in the activated sludge process and trickling

filter were due to adsorption on the activated sludge

flocs and on the slime in the trickling filter.

Chemical factors affecting E. coli removal by a biofilm

system have been assessed in this study, viz. dissolved

oxygen, pH, and cationic polymer addition. As expected

E. coli removal under anaerobic conditions is signifi-

cantly lower than that under aerobic conditions. We

found this also in previous experiments [32] viz. a E. coli

removal of 94.3% in the aerobic RBC system as

compared to only 43% in an anaerobic RBC treating

UASB effluent under further the same operating

conditions. Barzily et al. [33] found that RBC achieved

a Salmonella tym reduction of about six orders of

magnitude in 6 days, where, oxidation ponds accom-

plished a similar level of Salmonella tym reduction in 14

days. The main difference between Salmonella tym

behavior in dialysis bags attached to the RBC drum

operation and in oxidation pond can be attributed to the

changes in DO concentrations and the higher die-off.

The dissolved oxygen concentration in the oxidation

ponds are high at noon (rise to 20.0mg/l) and low at

night 0.0mg/l while the dissolved oxygen concentration

in the RBC was constant during a day and night.

Influence of pH on the adsorption depends on the

nature of the bacterial surfaces and ionic strength of the

solution [34]. pH affects bacterial surface zeta potential

due to dissociation of carboxylic and amino groups

located on the bacterial cell wall [35]. The effect of pH

on removal depends also on the iso-electric point of the

bacterial species. Since the pH of domestic wastewater

often is close to 7.0 [36], the pH will probably have a

minimal influence on the bacterial removal. In the

present study effects of pH variations within the

investigated range of 6.5–9.4 were found to be insignif-

icant.

The adsorption of E. coli to the biofilm was found to

become greatly enhanced for a short time period of by

cationic polymer additives as shown in Table 5. This is

because cationic polymers interact directly with specific

ionizable groups on the protein surface coat of the E.

coli, thus achieving a surface charge redistribution

favorable to the adsorption of E. coli to negatively

charged biofilm. According to Gambrill et al. [37]

chemical treatment can achieve Faecal coliforms and

Salmonella removal values of 99.999% and 99%,

respectively, by using of lime and Clari-floc as coagulant

and coagulant aid.

5. Conclusions

From the above-mentioned results and discussion, the

following can be concluded.

* In a RBC system at high loading rates, adsorption is

the main E. coli removal mechanism followed by

sedimentation. Die-off has a relatively minor role for

removal of E. coli in the RBC system.* The removal rate of E. coli under aerobic conditions

is significantly higher than under anaerobic condi-

tions.* The adsorption of the E. coli to heterotrophic biofilm

is not influenced by the pH in a range of 6.5–9.4.* A significant improvement in the removal rate of E.

coli can be achieved when cationic polymer is added

for a short period. However, on the long run,

polymer addition exerted almost no improvement in

the removal rate.

Acknowledgements

The first author would like to express their gratitude

to Dutch government (SAIL-IOP/SPP project) for

financial support of this research and to Dr. Ir Jules

Van Lier, director of the SAIL project, for help. The first

author would like to thank R.E. Roersma, B. Will-

emsen, and S. Hobma for technical support.

References

[1] Van der Steen P, Brenner A, Van Buuren J, Oron G. Post-

treatment of UASB reactor effluent in an integrated

duckweed and stabilisation pond system. Water Res

1999;33(3):615–20.

[2] Tawfik A, Klapwijk B, El-Gohary F, Lettinga G.

Treatment of anaerobically pre-treated domestic sewage

by a rotating biological contactor. Water Res 2001;36(1):

147–55.

[3] Crane SR, Moore JA. Modelling enteric bacterial die-off: a

review. Water Air Soil Pollut 1986;27:411–39.

[4] Parhad N, Rao NV. Effect of pH on survival of E. coli.

J Water Pollut Control Fed 1974;34:149–61.

[5] Mayo AW. Effect of pond depth on bacterial mortality

rate. J Environ Eng 1989;115:965–77.

[6] Moeller JR, Calkins J. Bactericidal agents in wastewater

lagoons and lagoon design. J Water Pollut Control Fed

1980;52:2442–51.

ARTICLE IN PRESSA. Tawfik et al. / Water Research 38 (2004) 1081–10881088

[7] Polprasert C, Dissanayake MG, Thanh NC. Bacterial die-

off kinetics in waste stabilization ponds. J Water Pollut

Control Fed 1983;55:285–96.

[8] Mills SW, Alabaster GP, Mara DD, Pearson HW, Thitai

WN. Efficiency of faecal bacterial removal in waste

stabilization ponds in Kenya. Water Sci Technol

1992;26(7–8):1739–48.

[9] Saqqar MM, Pescod MB. Modelling coliform reduction in

wastewater stabilization ponds. Water Sci Technol

1992;26(8):1667–748.

[10] Banks MK, Bryers JD. Microbial deposition rates onto

clean glass and pure culture bacterial biofilm surfaces.

Biofouling 1992;6:81–6.

[11] Le Chevallier MW, Babcock TM, Lee RG. Examination

and characterization of distribution system biofilms. Appl

Environ Microbiol 1987;53(12):2714–24.

[12] Cunningham AB, Bouwer EJ, Characklis WG. Biofilm in

porous media. In: Characklis WG, Marshall KC, editors.

In biofilms. New York: Wiley; 1990. p. 697–732.

[13] Heukelekian H, Rudolfs W. Effect of aeration and

protozoa on E. coli in sewage. Sewage Wks J 1929;1:561–7.

[14] Cuds CR, Fey GI. The effect of ciliated protozoa on the

fate of E. coli in the activated sludge process. Water Res

1969;3:853–67.

[16] Kinner NE, Curds CR. Development of protozoan and

metasoan communities in Rotating Biological Contactor

Biofilms. Water Sci Technol 1987;21(4):481–90.

[17] Grin PC, Roersma RE, Lettinga G. Anaerobic treatment

of raw sewage in UASB reactors at temperatures from

9�C to 20�C. In: Proceedings of the Seminar/

Workshop Anaerobic Treatment of Sewage, Amherst,

1985. p. 109–24.

[18] Havelaar AH. During M on behalf of a working group.

Evaluation of the Anderson Baird–Parker direct plating

method for enumerating Escherichia Coli in water. J Appl

Bacteriol 1988;64:89–98.

[19] Snedecor GW, Cochran WG. Statistical methods. Ames,

Iowa, USA: The Iowa State University Press; 1980.

[20] Hanes NB, Sarles WB, Rohlich GA. Dissolved oxygen and

survival of coliform organisms and enterococci. J Am

Water Wks Assoc 1964; 441–6.

[21] Pearson HW, Mara DD, Mills SW, Smallman D.

Physical–chemical parameters influencing faecal bacterial

survival in waste stabilization ponds. Water Sci Technol

1987;19:145–52.

[22] Tawfik A, Klapwijk B, El-Gohary F, Lettinga G.

Treatment of anaerobically treated domestic wastewater

using rotating biological contactor. Water Sci Technol

2002;45(10):371–6.

[23] Stevik TK, Ausland G, Hanssen JF, Jenssen PD. The

influence of physical and chemical factors on the transport

of E. coli through biological filters wastewater purification.

Water Res 1999;33(18):3701–6.

[25] Curds CR. Protozoa and the water industry. Cambridge,

UK: Cambridge University Press; 1992.

[26] Green MB, Griffin P, Seabridge JK, Dhobie D. Removal

of bacteria in subsurface flow wet-lands. Water Sci

Technol 1997;35(5):109–16.

[27] Milne DP, Curran JC, Findlay JS, Crowther JM, Wallis

SG. The effect of estuary type suspended solids on survival

of E. coli in saline waters. Water Sci Technol 1989;

21(3):61–5.

[28] Raman N, Chaklader N. Up-flow filters for septic tank

effluents. J Water Pollut Control Fed 1972;44:1552.

[29] Ueda T, Horan NJ. Fate of indigenous bacteriophage in a

membrane bioreactor. Water Res 2000;34(7):2151–9.

[30] Van der Drift C, Seggelen EV, Stumm C, Hol W, Tuinte J.

Removal of Escherichia coli in wastewater by activated

sludge. Appl Environ Microbiol 1977;34(3):315–9.

[31] Omura T, Onuma H, Aizawa J, Umita T, Yagi T. Removal

efficiencies of indicator microorganisms in sewage treat-

ment plants. Water Sci Technol 1989;21(3):119–24.

[32] Tawfik A. The biorotor system for post-treatment of

anaerobically treated domestic sewage. PhD thesis, Wa-

geningen University and Research Centre, Sub-department

of Environmental Technology, The Netherlands, 2001.

[33] Barzily A, Cavari BZ, Kott Y. Survival of various

Salmonella Typhimurium strains in adverse environments.

Proceedings of the UK Symposium on Health Related

Water Microbiology, University of Strathclyde Glasgow,

IAWPRC, 3–5 September, 1991.

[34] Harvey RW. Parameters involved in modeling movement

of groundwater. In: Hurst CJ, editor. Modelling the

environmental fate of micro-organisms. Washington, DC:

American Society for Microbiology; 1991. p. 89–114.

[35] Gannon JT, Tan Y, Baveye P, Alexander M. Effect of

sodium chloride on transport of bacteria in a saturated

aquifer material. Appl Environ Microbiol 1991;57:

2497–501.

[36] Canter LW. Septic tank systems. Effects on Groundwater.

Knox R.C. Quality. Boca Raton: Lewis Publishers; 1985.

[37] Gambrill MP, Mara DD, Oragui JI, Silva SA. Wastewater

treatment for effluent reuse: lime induced removal of

excreted pathogens. Water Sci Technol 1989;21(3):79–84.