Zero net growth in a membrane bioreactor with complete sludge retention

ARTICLE IN PRESS

0043-1354/$ - se

doi:10.1016/j.w

�Correspondfax: +39080 53

E-mail addr

Water Research 39 (2005) 5241–5249

www.elsevier.com/locate/watres

Zero net growth in a membrane bioreactor withcomplete sludge retention

G. Laera, A. Pollice�, D. Saturno, C. Giordano, A. Lopez

CNR IRSA, Via F. De Blasio 5, 70123 Bari, Italy

Received 9 February 2005; received in revised form 10 October 2005; accepted 10 October 2005

Abstract

A bench-scale membrane bioreactor was operated with complete sludge retention in order to evaluate biological

processes and biomass characteristics over the long term. The investigation was carried out by feeding a bench-scale

plant with real sewage under constant volumetric loading rate (VLR ¼ 1.2 gCODLreact�1 h�1). Biological processes were

monitored by measuring substrate removal efficiencies and biomass-related parameters. The latter included bacterial

activity as determined through respirometric tests specifically aimed at investigating long term heterotrophic and

nitrifying activity. After about 180 days under the imposed operating conditions, the system reached equilibrium

conditions with constant VSS concentration of 16–18 gL�1, organic loading rate (OLR) below 0.1 gCODgVSS�1 d�1

and specific respiration rates of 2–3mgO2 gVSS�1 h�1. These conditions were maintained for more than 150 days,

confirming that an equilibrium had been achieved between biomass growth, endogenous metabolism, and solubilization

of inorganic materials.

r 2005 Elsevier Ltd. All rights reserved.

Keywords: Membrane bioreactors; Biomass activity; Complete sludge retention; Municipal wastewater

1. Introduction

The main advantages of membrane bioreactors

(MBR) with respect to traditional activated sludge

systems were summarized as (i) longer sludge retention

times (SRT) independent of the hydraulic retention time

(HRT), (ii) smaller footprint, (iii) complete removal of

solids and nearly complete removal of effluent micro-

organisms, (iv) high removal ratios for most contami-

nants, (v) reduced sludge production, and (vi)

rapid start-up of biological processes (Stephenson

et al., 2000; Visvanathan et al., 2000). All these aspects

are currently under investigation, and a debate is still

e front matter r 2005 Elsevier Ltd. All rights reserve

atres.2005.10.010

ing author. Tel.: +39080 5020511;

13365.

ess: [email protected] (A. Pollice).

open on the most appropriate operating procedures to

maximize treatment performance with respect to opera-

tional costs. The latter mainly relate to membrane

cleaning, sludge wastage, and aeration, and these are all

affected by the operational concentration and features

of the sludge within MBR. A possible strategy for

operational cost limitation is reduction of sludge with-

drawal, despite increased aeration costs (Yoon et al.,

2004).

Theoretical investigations have evidenced that bio-

mass production can be limited in MBR by appropriate

operational strategies (Lu et al., 2001; Xing et al., 2003).

Some authors suggested that these systems could be

operated at high sludge concentrations (15–25 gSS/L) by

limiting biomass withdrawal, thus minimizing bacterial

growth. The main drawbacks of this operating proce-

dure were indicated as oxygen transfer limitations and

d.

ARTICLE IN PRESSG. Laera et al. / Water Research 39 (2005) 5241–52495242

increased membrane fouling tendency (Muller et al.,

1995; Lubbeke et al., 1995; Cote et al., 1998; Rosenber-

ger et al., 2000).

Besides operational effects, long SRT were also

investigated for their effects on the biomass, both in

terms of accumulation rates and activity. Lower enzy-

matic activity, and smaller yield and decay coefficients

were reported for the higher sludge ages, suggesting

lower flexibility to xenobiotic compounds due to slower

biomass renovation (Chaize and Huyard, 1991; Huang

et al., 2001).

The possibility of operating MBR without

sludge withdrawal was explored by several researchers,

some of whom worked on relatively short timeframes

(Yamamoto et al., 1989; Benıtez et al., 1995;

Pollice et al., 2004). Others provided results of long-

term experimental tests mainly focused on removal

efficiencies and operational aspects (Chiemchaisri et al.,

1992; Muller et al., 1995; Wagner and Rosenwinkel,

2000; Rosenberger et al., 2002). All these authors

stated the biological applicability of complete sludge

retention, reporting high and stable degradation

rates and very limited sludge production or zero net

growth. However, equilibrium conditions are experi-

mentally difficult to reach in systems having high

or indefinite SRT, and most of these findings were

obtained under non-steady state conditions, with vari-

able loads and sometimes with synthetic wastewater.

Only few works were focused on biomass-related

aspects such as microbiology and bacterial activities,

and maintenance metabolism was often related to

complete sludge retention (Witzig et al., 2002; Wagner

and Rosenwinkel, 2000). Furthermore, starvation

was recently addressed as a means to maintain

the biomass in a state not allowing cell division, but

still able to participate in the degradation processes

to satisfy maintenance energy requirements (Lobos

et al., 2005). Finally, reported results are not always

consistent with respect to the accumulation of

organic inerts and inorganic material within the biomass

under complete sludge retention (Muller et al., 1995;

Lu et al., 2001; Rosenberger et al., 2002; Pollice

et al., 2004). These last two aspects of sludge biology

and fate of inorganics under equilibrium conditions are

especially relevant for long-term MBR practical applica-

tion and modelling.

Purpose of this experimental work was to verify that

in a MBR with complete sludge retention fed on real

municipal wastewater:

�

of treatment performance,

�

�

�

2. Materials and methods

The bench-scale submerged MBR had 6L operating

volume, and a Zenon hollow fiber membrane module

having a surface of 0.047m2 was sunk in the reactor and

operated out-in. A scheme of the bench-scale experi-

mental plant was reported elsewhere (Pollice et al.,

2004).

The permeate was extracted by imposing on the

membrane a negative pressure that never exceeded

0.5 bar. Operational cycles lasted 6min and included

extraction of the permeate (5.5min) and backwash

(0.5min). An air pump provided a constant air flow

through the membrane module to limit fouling and cake

formation. A second air pump was regulated to

maintain residual dissolved oxygen (DO) concentrations

between 1 and 3mgL�1 to favour COD removal and

nitrification. The experimental plant was continuously

monitored for DO, temperature, pH, and transmem-

brane pressure (TMP).

Biological processes were operated by maintaining a

constant volumetric loading rate (VLR), defined as the

amount of COD daily fed per liter of reactor’s volume.

This was obtained by keeping a constant permeate flux

and influent concentration. The system was continu-

ously fed on pre-settled municipal wastewater screened

at 1mm, stored at 4 1C and stirred. The feed concentra-

tion was maintained within a constant range by diluting

the real sewage with tap water when needed.

The plant was started up without any biomass

inoculum, and during the whole experimental period

no sludge was intentionally removed from the reactor

except for measurements of suspended solids. Biomass

samples were withdrawn from the reactor once or twice

per week and analysed for total suspended solids (TSS)

and volatile suspended solids (VSS) according to

standard methods (Standard Methods, 1995). Very

limited amounts of sludge were lost during on-site

membrane cleaning and these were accounted for in the

evaluation of biomass growth and the other sludge-

related parameters. As a general rule of this investiga-

tion, biomass samples were always returned to the

reactor after non-destructive determinations. The lim-

ited amount of biomass that tended to stick to those

parts of the reactor close to the surface level was daily

removed and returned to the sludge bulk. Sludge

management practices during the experiment are sum-

marized in Table 1.

The volume of sludge removed and not returned to

the reactor was estimated to be about 35mL/week on

average, corresponding to 0.83 gTSS/week.

Average values of the biomass yield were calculated

over periods of 3 weeks by taking into account the

sludge losses. In these periods the yield was obtained by

adding the amount of biomass lost to the discrete

integration of the area below the growth curve, and

ARTICLE IN PRESS

Table 2

Operational parameters of the bench-scale complete retention

MBR

Parameter Unit Average Std.

Dev.

Min Max

Flux Lm�2 h�1 16.4 1.2 9.4 17.6

TMP mbar 154 106 15 500

DO mgO2L�1 2.4 1.1 0.4 7.5

Temperature 1C 20.9 4.5 12.1 29.6

pH 6.8 0.3 5.4 7.7

Table 1

Sludge extracted from the reactor during the experiment

Event Unit Amount Frequency Returned

to plant

TSS+VSS determination mL 20 Weekly No

Sludge characterizationa mL 250 Weekly Yes

Membrane cleaning g 1.8 —b No

Respirometry mL 300 Weekly Yes

aNon-destructive viscosity and filterability tests were regularly performed on sludge samples.bOn-site module rinsing was performed 7 times during the whole experimental period, with a frequency that depended on the

patterns of pressure growth (see following paragraph).

G. Laera et al. / Water Research 39 (2005) 5241–5249 5243

dividing by the cumulative COD removed in the same

period.

Respirometry was performed after aerating the sludge

samples overnight to obtain endogenous respiration

conditions. The respirometer adopted was a ‘‘static

gas–static liquid’’ system operated by measurement of

the liquid phase DO concentrations, and its description

was provided elsewhere (Spanjers et al., 1998; Pollice

et al., 2004).

The influent municipal wastewater was sampled three

times per week and analysed for TSS and VSS, total

COD, N-NH4, TKN, N-NO2, and N-NO3. The perme-

ate was sampled daily and analysed for the same

parameters. All analyses were performed according to

standard methods (Standard Methods, 1995). The ion

chromatographic method for the determination of

nitrate and nitrite was modified adopting an UV

detector (at 220 nm) instead of the conductivity detector.

3. Results

3.1. Plant performance

The bench-scale MBR was operated for 336 days with

a constant volumetric loading rate of 1.2 gCODLreact�1 d�1

and process performance, sludge accumulation, and

biomass features were monitored over time. The main

operational parameters are provided in Table 2, and the

average characteristics of the influent wastewater and the

permeate are reported in Table 3.

The filtration performance of the membrane module is

described in Fig. 1, where the evolution over time of flux

and TMP are reported. The normal cleaning procedure

of the module was on-site jet rinsing of the fibers with

moderately pressurized tap water and was performed

when the TMP approached 500mbar. This procedure

was effective in restoring good permeability, and the

membrane was chemically cleaned only when the

positive effect of rinsing tended to last too short. This

happened just twice in 1 year of operation (days 63 and

296, Fig. 1), and a good efficiency of chemical cleaning

in restoring the membrane’s permeability was observed.

A sub-critical flux behaviour is also evidenced, with a

critical TMP value of 150mbar for this system that was

independent of the sludge concentration, and above

which the fouling rate rapidly increased. Similar patterns

were observed in other research works, although their

interpretation with respect to flux sustainability is still

debated (Ognier et al., 2001; Cho and Fane, 2002;

Guglielmi, 2002; Le Clech et al., 2003; Pollice et al.,

2005).

3.2. Biomass accumulation

Fig. 2 shows biomass accumulation in the reactor

under constant VLR, and compares these data with

previous results obtained in short-term tests (Pollice

et al., 2004). This comparison evidenced similar sludge

accumulation patterns for three different VLRs, with

higher initial growth rates for higher VLR. Moreover,

sludge accumulation was very limited during the last 150

days of operation and the system reached an equilibrium

sludge concentration of 17.371.1 gVSSL�1. The VSS/

TSS ratio was rather stable over the whole experimental

period, and an average value of 7572% was maintained

after the initial 50 days.

Further evidence that the equilibrium was reached in

the long-term was provided by the calculation of the

organic loading rate (OLR, gCODgVSS�1 d�1). Fig. 3

ARTICLE IN PRESS

0

200

400

600

0 30 60 90 120 150 180 210 240 270 300 330 360time (days)

TM

P (

mba

r)

0

6

12

18

Flu

x (L

m-2

h-1

)

TMPFlux

chemical cleaning

Fig. 1. TMP and flux variations during the experiment (336 days).

0

4

8

12

16

20

0 30 60 90 120 150 180 210 240 270 300 330 360time (days)

slud

ge c

once

ntra

tion

(gV

SS

L-1

)

0.8 gCOD / Lh

1.2 gCOD / Lh

1.7 gCOD / Lh

Fig. 2. Sludge accumulation in the reactor (K) as compared with previous results of short term tests (� and n) (Pollice et al., 2004).

Table 3

Main average characteristics of the investigated influent primary sewage and permeate

Parameter Unit Influent Permeate Efficiency

Avg Std. Dev. Avg Std. Dev.

TSS mgL�1 200 70 n.d. n.d. 499.9%

VSS mgL�1 160 50 n.d. n.d. 499.9%

COD mgL�1 400 106 57 34 86%

NH4+ mgNL�1 36.6 7.6 0.2 0.2 99%

NO3� mgNL�1 0.7 1.7 40.6 16.9 —

TKN mgNL�1 49.3 9.5 0.8 0.3 98%

n.d. ¼ below the detection limit.

G. Laera et al. / Water Research 39 (2005) 5241–52495244

shows that during the last 150 days of the experiment a

constant OLR was maintained (0.0770.01 gCOD

gVSS�1 d�1). The figure also shows that these data well

agree with previous short-term results (Pollice et al.,

2004).

3.3. Respiration rates

The long-term effects of complete sludge retention on

biomass activity were investigated by periodically mea-

suring the oxygen uptake rate (OUR) of heterotrophic

ARTICLE IN PRESS

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 30 60 90 120 150 180 210 240 270 300 330 360time (days)

OLR

(gC

OD

gV

SS

-1 d

-1)

0.8 gCOD / Lh

1.2 gCOD / Lh

1.7 gCOD / Lh

Fig. 3. Organic loading rate over time as compared with previous data from short-term tests (Pollice et al., 2004).

0

4

8

12

16

20

24

spec

ific

OU

R(m

g O

2 gV

SS

-1 h

-1)

Acetate

Endogenous

0

4

8

12

16

20

24

0 30 60 90 120 150 180 210 240 270 300 330 360time (days)

spec

ific

OU

R(m

g O

2 gV

SS

-1 h

-1)

AmmoniumNitrite

(a)

(b)

Fig. 4. Maximum specific respiration rates of heterotrophic (a) and autotrophic (b) populations.

G. Laera et al. / Water Research 39 (2005) 5241–5249 5245

and autotrophic populations with respirometric tests.

The two bacterial groups were examined by adopting

sodium acetate, sodium nitrite, and ammonium chloride

as synthetic substrates, and measuring the respiration

rates after dosing concentrations of these substrates in

excess with respect to the metabolic needs.

The specific maximum respiration rates with respect

to the VSS (i.e. specific OUR, mgO2 gVSS�1 h�1) are

reported in Fig. 4, where well known differences can be

observed in the behaviour of the two bacterial groups.

Autotrophic nitrifiers are slower than the heterotrophs

in reaching their maximum specific activity. However,

after about 100 days of operation the specific respiration

rates were not very different and both tended to decrease

in the longer term. Stability was reached after about 180

days of operation.

This tendency of the specific activity to decrease

towards a constant value was counterbalanced by the

corresponding increase of the microrganism concentra-

tion towards a constant value over the long term. The

combination of these two effects resulted in the

maintenance of the overall biomass activity, with

consequent constancy of the respiration rates (i.e.

OUR, mgO2 h�1) over the long term.

Endogenous and maximum OUR of both microbial

populations showed similar trends, i.e. after having

reached a maximum between 30 and 90 days of

operation they slowly declined towards an equilibrium

value between 20% and 50% of the maximum. During

the last 150 days of operation, the average endogenous

OUR was 1373mgO2 h�1, and the average maximum

respiration rates recorded for acetate, ammonium and

ARTICLE IN PRESSG. Laera et al. / Water Research 39 (2005) 5241–52495246

nitrite were 2279, 46717 and 673mgO2 h�1, respec-

tively.

An occasional drop of the autotrophic activity was

observed around day 260, when a higher effluent

ammonium concentration was also observed (4.0mgN-

NH4+L�1 against an average 0.2mgN-NH4

+L�1), sug-

gesting the occurrence of an inhibitor in the influent

wastewater. After this, the biomass promptly recovered

its previous activity.

4. Discussion

The experimental results reported above show that the

system had reached an equilibrium after about 180 days

of operation with negligible sludge withdrawal. This

equilibrium was maintained over the long term. Opera-

tion of the plant with complete sludge retention does not

allow for the definition of the system’s sludge age.

However, if the limited volumes of sludge sampled and

not returned to the reactor are taken into account, a

sludge age of about 1200 days can be estimated.

Sludge accumulation in the reactor showed the same

pattern observed in short-term tests (Pollice et al., 2004),

i.e. the initially high sludge growth rate strongly

decreased after about 30 days, and then slowly declined

resulting in stabilization of the average biomass

concentration during the last 150 days (Fig. 2). Also

other relevant process parameters such as the VSS/TSS

ratio, the OLR and the specific biomass activities

reached constant values in the same period (Figs. 3

and 4). This tendency of the OLR and the specific

activities towards constant (minimized) values and the

stabilization of the VSS and the VSS/TSS ratio suggest

that the system had reached and maintained equilibrium

conditions.

Fig. 3 also shows that in MBRs operated under

complete sludge retention the OLR evolves towards a

minimized value that is independent of the VLR. This

suggests that under equilibrium conditions the biomass

concentration is proportional to the volumetric loading

rate.

The absence of biomass net growth observed during

the final 150 days can be explained with the occurrence

of maintenance metabolism. Moreover, the stabilization

of the VSS and the specific activities means that no

accumulation of ‘‘inert’’ material occurred. This is

consistent with a recently proposed interpretation

stating that accumulation of inerts inversely depends

on the system’s SRT, and that they can be degraded by

slow growing bacteria (van Loosdrecht and Henze,

1999).

The MBR showed good removal efficiencies of

suspended solids, COD, ammonia and total nitrogen

(Table 3). These performances suggested that the

biomass viability was maintained over a long period.

Respirometric tests also confirmed the persistence of

biomass viability and no tendency towards the miner-

alization of biosolids, even over long timeframes.

During the last 150 days of operation the OUR values

were below their maximum values, but respiration rates

ranging between 20% and 50% of the maximum were

maintained until the end of the experiment. Moreover,

an occasional drop of the autotrophic activity was

promptly recovered and the OUR values measured

before the inhibition were immediately restored, show-

ing good flexibility of the biomass towards possible

changes occurring in the influent.

In order to estimate biomass growth and decay rates

over the long term, the active fraction needs to be

quantified. Available literature data on this ‘‘classical’’

issue of biological wastewater treatments are extremely

limited and case-specific, and mostly refer to conven-

tional-activated sludge. MBRs were reported to have an

active fraction between 4% and 7% with respect to the

VSS in systems with SRT around 35–50 days (Guglielmi,

2002). This fraction refers to steady state conditions, and

its application to complete retention MBR has to take

into account the evolution over time of the biomass (and

of the active fraction). In the present study, constant

OLR, VSS, and endogenous respiration rate after 180

days of operation suggested that the concentration of

active biomass was approximately constant. Exact

determination of this fraction would have required

sludge sampling that, at the experimental scale adopted,

was not compatible with the assumption of complete

sludge retention.

However, some indications on the growth and decay

of the biomass can also be obtained without information

on the active fraction. Biomass build-up in biological

systems is commonly represented by the Monod model

where a decay term is introduced (Henze et al., 1987):

dX

dt¼ mX � bX ¼ mmax

S

kS þ SX � bX ,

where X is the active fraction of the biomass, m is the

growth rate, b the decay rate, S the substrate concentra-

tion, kS the half-saturation constant.

According to this model, absence of biomass growth

can be expresses as follows:

dX

dt¼ 0) m ¼ b3

mb¼ 1.

Oxygen consumption during biological processes

depends on the utilization external substrate and the

endogenous biomass decay. The OUR can be defined

accordingly, by separately considering these two con-

tributions

OUR ¼ OURext þOURend ¼ m1� Y 0

Y 0

� �X þ bX ,

ARTICLE IN PRESSG. Laera et al. / Water Research 39 (2005) 5241–5249 5247

where Y 0 is the yield of substrate utilization for biomass

production.

The maximum OUR is obtained for maximum growth

rate (i.e. when SbkS) and can be expressed as follows:

OURmax ¼ mmax

1� Y 0

Y 0

� �X þ bX .

When all the available external substrate has been

consumed, the OUR is represented by the decay term bX

(i.e. the oxygen requirement for endogenous respiration)

OURend ¼ bX .

The ratio between OURmax and OURend leads to the

following expression:

OURmax

OURend� 1

� �Y 0

1� Y 0

� �¼

mmax

b

According to this expression, the ratio between mmax and

b is independent of the active fraction X . Therefore,

when OURmax and OURend have comparable values, the

maximum growth is similar to the biomass decay.

The experimental ratio OURmax/OURend at the

equilibrium was calculated to be 1.770.3. The adoption

of a standard yield value of 0.67mgCODbiommgCODsub�1

(Henze et al., 1987) resulted in mmax/b ¼ 1.570.6. Since

the actual growth rate m is smaller than the mmax, it could

be assumed that the ratio m=b approaches one. The

above considerations explain the absence of net biomass

growth experimentally observed and confirm previous

model predictions (Wintgens et al., 2003).

The plot of the observed sludge yield as a function of

the OLR over the whole experimental period evidenced

a tendency of the yield to decrease for decreasing OLRs

(Fig. 5). The experimental data showed that minimized

OLR resulted in extremely limited yields, and that zero

biomass yield could be reached, as previously observed

under non-steady state conditions (van Houten and

Eikelboom, 1997; Rosenberger et al., 2002). Although

extremely low values of OLR are likely to take very long

0.00

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.00 0.02 0.04 0.06 0

OLR (gCOD

slud

ge y

ield

(gV

SS

gC

OD

-1)

Fig. 5. Observed sludge yield

time to be reached, negligible yield can already be

assumed for the operational loading rates around

0.07 gCODgVSS�1 d�1 experimentally maintained, also

taking into account the limited sludge losses occurred

during the experiment.

Absence of sludge accumulation means zero net

biomass growth and also no accumulation of inorganic

material. The amount of inorganics accumulated in the

reactor, determined as the difference between the TSS

and the VSS in the sludge, was compared with the

cumulative inorganic material in the influent wastewater

(Fig. 6). The ratio between these two non-volatile

fractions was found to be rather constant (about 30%)

for the initial 100 days, in accordance with previous

findings on conventional activated sludge and non-

steady state MBR (Wentzel et al., 2002; Wagner and

Rosenwinkel, 2000). Then the accumulation rate was

observed to decrease, and finally resulted in a constant

concentration of inorganics in the reactor over the long

term. Also in this case, as for the other process

parameters, the accumulation of inorganic material

ceased after about 180 days, confirming that the studied

system had reached equilibrium conditions.

Constancy of the VSS/TSS ratio together with zero

net growth provides an experimental evidence of

previously proposed theories predicting the absence of

excess sludge production under sufficiently long sludge

retention times (Xing et al., 2003). No discernible

accumulation of inorganics was already suggested only

based on the observation of constant VSS/TSS ratio,

although never experimentally demonstrated under

steady state conditions (Muller et al., 1995).

At the end of the experiment, the reactor was emptied

to check for possible localized accumulation of solids

due to mixing problems. None of such deposits was

observed, confirming that the system was completely

stirred and that the constant concentration of inorganics

was not caused by ‘‘hidden’’ materials as in previous

experiments (Rosenberger et al., 2002). Therefore, when

.08 0.10 0.12 0.14 0.16

gVSS-1 d-1)

vs. organic loading rate.

ARTICLE IN PRESS

0

30

60

90

120

150

180

210

240

270

0 30 60 90 120 150 180 210 240 270 300 330 360time (days)

influ

ent i

norg

anic

s (g

)

0

9

18

27

36

45

54

63

72

81

slud

ge in

orga

nics

(g)

cumulative influent inorganics

sludge inorganics

Fig. 6. Inorganic fractions in the influent wastewater and accumulated in the reactor.

G. Laera et al. / Water Research 39 (2005) 5241–52495248

high sludge concentrations were maintained, the results

showed a limit value above which no accumulation of

inorganic materials occurred in the reactor. This

observation can be explained by considering that at

sludge concentrations above 15 gVSSL�1 the influent

wastewater contributes for about 5–10% to the total

amount of inorganics in the reactor, and this fraction

could be involved in processes leading to solubilization.

5. Conclusions

Long-term experiments with a complete retention

MBR fed on real sewage at constant volumeric loading

rate showed that equilibrium conditions were reached

after about 180 days of operation and maintained

for more than 150 days. Equilibrium was characterized

by constancy of three key process parameters, i.e.

biomass (VSS) and inorganic material concentrations

and the specific bacterial activity. As a consequence,

the organic loading rate reached values around

0.07 gCODgVSS�1 d�1 in the long term, resulting in

sludge yield values close to zero. This was confirmed by

the results of the activity tests, that showed comparable

values for growth and decay rates (mmax/b ¼ 1.570.6).

No accumulation of inorganic material within the

reactor was observed once the equilibrium was reached.

Complete nitrification and good COD removal

efficiencies were maintained over the whole experimental

period. Moreover, adopting a typical flux and a simple

membrane-cleaning procedure, the initial biomass

growth and long-term operations at high sludge

concentrations caused no significant variations of

membrane permeability and cleaning needs.

The experimental work showed the feasibility of

operating MBR with complete sludge retention (or with

SRT41000 d) from the standpoint of biological pro-

cesses. Moreover, the reported data show that the

biomass evolves towards a minimized maintenance

OLR, resulting in proportionality between the VLR

and the equilibrium biomass concentration. When

designing new plants, this feature allows for defining

the biomass concentration based on the VLR.

Larger scale studies are needed to evaluate all cost-

related issues, including aeration, membrane cleaning

and sludge treatability.

References

Benıtez, J., Rodrıguez, A., Malaver, R., 1995. Stabilization and

dewatering of wastewater using hollow fiber membranes.

Water Res. 29 (10), 2281–2286.

Chaize, S., Huyard, A., 1991. Membrane bioreactor on domestic

wastewater treatment. Sludge production and modeling

approach. Water Sci. Technol. 23 (Kyoto), 1591–1600.

Chiemchaisri, C., Wong, Y.K., Urase, T., Yamamoto, K., 1992.

Organic stabilization and nitrogen removal in membrane

separation bioreactor for domestic wastewater treatment.

Water Sci. Technol. 25 (10), 231–240.

Cho, B.D., Fane, A.G., 2002. Fouling transients in nominally

sub-critical flux operation of a membrane bioreactor.

J. Membr. Sci. 209, 391–403.

Cote, P., Buisson, H., Praderie, M., 1998. Immersed mem-

branes activated sludge process applied to the treatment of

municipal wastewater. Water Sci. Technol. 38 (4–5),

437–442.

Guglielmi, G., 2002. Membrane bioreactors for municipal

wastewater treatment. Ph.D. Thesis, Universita’ di Trento,

Italy.

Henze, M., Grady Jr., C.P.L., Gujer, W., Marais, G.V.R.,

Matsuo, T., 1987. Activated sludge model no. 1. IAWPRC

Task Group on Mathematical Modelling for Design and

Operation of Biological Wastewater Treatment. IAWPRC

scientific and technical reports no. 1. IAWQ, London UK.

Huang, X., Gui, P., Qian, Y., 2001. Effect of sludge retention

time on microbial behaviour in a submerged membrane

bioreactor. Proc. Biochem. 36 (10), 1001–1006.

ARTICLE IN PRESSG. Laera et al. / Water Research 39 (2005) 5241–5249 5249

Le Clech, P., Jefferson, B., Chang, I.S., Judd, S., 2003. Critical

flux determination by the flux-step method in a submerged

membrane bioreactor. J. Membr. Sci. 227, 81–93.

Lobos, J., Wisniewski, C., Heran, M., Gramsick, A., 2005.

Effects of starvation conditions on biomass behaviour for

minimization of sludge production in membrane bioreac-

tors. Water Sci. Technol. 51 (6–7), 35–44.

Lu, S.G., Imai, T., Ukita, M., Sekine, M., Higuchi, T.,

Fugakawa, M., 2001. A model for membrane bioreactor

process based on the concept of formation and degradation

of soluble microbial products. Water Res. 35 (8),

2038–2048.

Lubbeke, S., Vogelpohl, A., Dewjanin, W., 1995. Wastewater

treatment in a biological high-performance system with high

biomass concentration. Water Res. 29 (3), 793–802.

Muller, E.B., Stouthamer, A.H., van Verseveld, H.W., Eikel-

boom, D.H., 1995. Aerobic domestic waste water treatment

in a pilot plant with complete sludge retention by cross-flow

filtration. Water Res. 29 (4), 1179–1189.

Ognier, S., Wisniewski, C., Grasmick, A., 2001. Biofouling in

MBRs: phenomenon analysis and modeling. Proceedings of

MBR3 Third International Meeting on Membrane Bior-

eactors for Wastewater Treatment, Cranfield University,

16th May 2001, Cranfield, UK.

Pollice, A., Laera, G., Blonda, M., 2004. Biomass growth and

activity in a membrane bioreactor with complete sludge

retention. Water Res. 38 (7), 1799–1808.

Pollice, A., Brookes, A., Jefferson, B., Judd, S., 2005. Sub-

critical flux fouling in membrane bioreactors—a review of

recent literature. Desalination 174 (3), 221–230.

Rosenberger, S., Witzig, R., Manz, W., Szewzyk, U., Kraume,

M., 2000. Operation of different membrane bioreactors:

experimental results and physiological state of the micro-

organisms. Water Sci. Technol. 41 (10–11), 269–277.

Rosenberger, S., Kruger, U., Witzig, R., Manz, W., Szewzyk,

U., Kraume, M., 2002. Performance of a bioreactor with

submerged membranes for aerobic treatment of municipal

waste water. Water Res. 36 (2), 413–420.

Spanjers, H., Vanrolleghem, P.A., Olsson, G., Dold, P.L., 1998.

Respirometry in control of the activated sludge process:

principles. IAWQ Scientific and Technical Report no. 7,

IAWQ Publishing, London, UK.

Standard Methods, 1995. Standard Methods for the Examina-

tion of Water and Wastewater, 19th ed. American Public

Health Association/American Water Works Association/

Water Environment Federation, Washington DC.

Stephenson, T., Judd, S., Jefferson, B., Brindle, K., 2000.

Membrane Bioreactors for Wastewater Treatment. IWA

Publishing, London, UK.

van Houten, R.T., Eikelboom, D.H., 1997. High performance

membrane bioreactors: a physiological approach. Proceed-

ings of MBR1–First International Meeting on Membrane

Bioreactors for Wastewater Treatment. Cranfield Univer-

sity, 5th March 1997, Cranfield, UK.

van Loosdrecht, M., Henze, M., 1999. Maintenance, endogen-

ous respiration, lysis, decay and predation. Water Sci.

Technol. 39 (1), 107–117.

Visvanathan, C., Ben Aim, R., Parmeshwaran, K., 2000.

Membrane separation bioreactors for wastewater treatment.

Crit. Rev. Environ. Sci. Technol. 30 (1), 1–48.

Wagner, J., Rosenwinkel, K.-H., 2000. Sludge production in

membrane bioreactors under different conditions. Water

Sci. Technol. 41 (10–11), 251–258.

Wentzel, M.C., Ubisi, M.F., Lakay, M.T., Ekama, G.A., 2002.

Incorporation of inorganic material in anoxic/aerobic

activated sludge system mixed liquor. Water Res. 36 (20),

5074–5082.

Wintgens, T., Rosen, J., Melin, T., Brepols, C., Drensla, K.,

Engelhardt, N., 2003. Modelling of a membrane bioreactor

for municipal wastewater treatment. J. Membr. Sci. 216,

55–65.

Witzig, R., Manz, W., Rosenberger, S., Kruger, U., Kraume,

M., Szewzyk, U., 2002. Microbiological aspects of a

bioreactor with submerged membranes for aerobic treat-

ment of municipal wastewater. Water Res. 36 (2), 394–402.

Xing, C.-H., Wu, W.-Z., Quian, Y., Tardieu, E., 2003. Excess

sludge production in membrane bioreactors: a theoretical

investigation. J. Environ. Eng. ASCE 129 (4), 291–297.

Yamamoto, K., Hiasa, M., Mahmood, T., Matsuo, T., 1989.

Direct solid–liquid separation using hollow fiber membrane

in an activated sludge aeration tank. Water Sci. Technol. 21

(Brighton), 43–54.

Yoon, T.I., Lee, H.S., Kim, C.G., 2004. Comparison of

pilot scale performances between membrane bioreactor

and hybrid conventional wastewater treatment system.

J. Membr. Sci. 242, 5–12.