中国科技论文在线 http://www.paper.edu.

cn

Process Biochemistry 41 (2006) 87–95

Comparison between a sequencing batch membrane

bioreactor and a conventional membrane bioreactor
Han-Min Zhang *, Jing-Ni Xiao, Ying-Jun Cheng,
Li-Fen Liu, Xing-Wen Zhang, Feng-Lin Yang
School of Environmental and Biological Science and Technology, Dalian University of Technology, Dalian 116024, PR China
Received 1 July 2004; received in revised form 26 January 2005; accepted 19 March 2005

Abstract

An operation of a membrane bioreactor in sequencing batch mode named a sequencing batch membrane bioreactor (SBMBR) was
investigated for enhancing nitrogen and phosphorus removal. Its performance was compared with a conventional membrane bioreactor
(CMBR) at various influent COD/TN ratios of 3.4–28.2. The operational parameters were optimized to increase the treatment efficiency. COD
removal averaged at 94.9 and 97.7%, respectively, for SBMBR and CMBR during the 8 months experimental period. The SBMBR system
demonstrated good performance on nitrogen and phosphorus removal at different COD/TN ratios. When COD/TN was 6.3 and the total
nitrogen (TN) load was 0.22 kg/(m3 days), the TN and ammonium nitrogen removals of the SBMBR were maintained over 65 and 90%,
respectively. Total phosphorus (TP) removal of the SBMBR was approximately 90% during most of the experimental time. In comparison, the
CMBR did not perform so well. Its effluent TN concentration was close to that in the influent at COD/TN = 6.3 and TP removal was not stable.
The specific nitrification rate test showed that pH value affected the activity of nitrifiers but no irreversible harm was induced. Furthermore, the
sequencing batch mode operation of MBR retarded membrane fouling according to the monitoring of trans-membrane pressure (TMP).
# 2005 Elsevier Ltd. All rights reserved.

Keywords: Sequencing batch membrane bioreactor (SBMBR); Conventional membrane bioreactor (CMBR); Nitrogen and phosphorus removal; Membrane
fouling; Specific nitrification rate

1. Introduction However, SBR has a potential deficit in that poor

clarification and a turbid effluent are associated with it.
The key nutrients causing eutrophication in waterways To overcome these drawbacks and improve system
are excess phosphorus and nitrogen concentrations in performance, membrane filtration technology has been
effluents from municipal or industrial plants discharged in introduced in biological wastewater treatment. The filtration
the environment. Therefore, it is necessary to removal these function of membrane enables a complete solid–liquid
from wastewater at their sources. A number of biological separation in membrane bioreactor (MBR), eliminating the
nutrient removal (BNR) processes had been developed. need for the formation of flocs or aggregates [2–5]. The
Among these processes, the sequencing batch reactor (SBR) combination of MBR have been used for treating various
has been applied as one alternative BNR technology since its types of wastewater, which demonstrate good performance
process is simple to operate and very flexible for combining for removal of organic matter and biological sludge
nitrogen and phosphorus removal. Its cycle format can be separation [6–8].
easily modified at any time to offset changes in process Recent studies have introduced alternating aerobic and
conditions, influent characteristics or effluent objectives [1]. anaerobic conditions in a submerged MBR by intermittent
aeration for simultaneous removal of carbon and nitrogen.
Yeom et al. [9] reported that an intermittently aerated
* Corresponding author. membrane bioreactor with a submerged fibre hollow
E-mail address: zhhanmin@126.com (H.-M. Zhang). membrane was capable of achieving 96% COD and 83%

1359-5113/$ – see front matter # 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.procbio.2005.03.072

转载
中国科技论文在线 http://www.paper.edu.cn
88 H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95

TN removals at 8–15 h HRT and a very long SRT. Ueda and 2. Materials and methods
Hata [10] introduced an intermittent suction method in a
membrane bioreactor for treatment of domestic sewage from 2.1. Reactors set-up and operation
rural settlements. Ng et al. [11] investigated the effect of
MLSS concentration in a sequencing batch membrane Fig. 1 shows a schematic diagram of experimental
bioreactor on COD and nitrogen removal. Bae et al. [12] apparatus consisting of two 11 L Plexiglas reactors.
evaluated the biological nutrient removal performance of a Temperature was controlled at 25
1 8C using thermostats.
membrane separation process coupled to a SBR in the A balance-box with a float-ball valve was used to control
treatment of dairy industry wastewater. All these studies water level of the two reactors. The membrane module used
demonstrated that MBR operated in an intermittent way can in the two systems was a bundle of U-shaped hollow fibre
obtain a high organic substrate and nitrogen removal. The membranes made of polyethylene (DAIKI, Japan) with a
removal of phosphorus remains inadequate, however, and pore size of 0.1 mm and a filtration area of 0.15 m2.
should be optimized to increase phosphorus removal. SBMBR and CMBR were in continuous operation in
The objective of this study was to enhance nitrogen and parallel for about 8 months. The variation of operational
phosphorus removal by sequential operation of a membrane parameter is listed in Table 1. The exchange ratio f B is
bioreactor in alternating aerobic and anoxic/anaerobic defined as the ratio between the volume discharged per cycle
condition. This system was named a sequencing batch and the entire working volume of the reactor. The water fed
membrane bioreactor (SBMBR). Besides the advantages of rate of the two systems was the same and the SRT was 60
no sedimentation phase and thus a shorter cycle time, the days.
combination may provide a suitable environment for both In the SBMBR system, the water was fed to the reactor
nitrobacteria and phosphorus accumulating organisms during the anaerobic phase and the discharging of water
(PAO) to survive under the appropriate operational condi- occurred during the aerobic phase. To make it work this way
tions. This combination can also make it possible that the a timer-controlled power supply system was used and a Y-
solids retention time (SRT) is independent of hydraulic shaped filter was installed before the electromagnetic valve
retention time (HRT). Furthermore, the intermittent suction that connected to the influent pipeline. An agitator was used
and the introduction of air diffusers near the membrane had for mixing during the anaerobic phase. The alternating
been reported as the effective methods to reduce fouling and aerobic and anaerobic conditions in the SBMBR was created
prolong membrane operation time and lifespan [13–16]. It by the on and off control of the air pump. The air pump and
also allowed good growth of strains with different doubling peristaltic pump worked simultaneously during the aerobic
times. phase, while the electromagnetic valve placed in the influent
The performance of the SBMBR and a conventional pipeline was closed. This resulted in discharge of water and
membrane bioreactor (CMBR) on nutrient removal was the descending of water level in the reactor. For the anoxic/
compared at different influent COD/TN ratios from 3.4 to anaerobic phase, the air pump and peristaltic pump were
28.2. The suitable operation parameters under each closed and the agitator started to work to keep a complete
condition were selected and membrane fouling monitored mixing of sludge and water. The feed water flowed into
at the same time. the SBMBR by gravitational force. Its level was controlled

Fig. 1. Schematic diagram of experimental process.

中国科技论文在线 http://www.paper.edu.cn
H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95 89

Table 1
Operation parameters in the experiment
Phase Operational days (days) TNin (mg/L) COD/TN HRT (h) fB Cycle time of SBMBR (min)
Anaerobic Aerobic
I 1–12 16.5 (2.4) 28.2 (3.1) 22 9.1 60 60
II 13–30 22.9 (1.8) 19.4 (1.5) 22 9.1 60 60
III 31–72 47.1 (6.7) 9.4 (2.7) 22 9.1 60 60
IV 73–110 75.6 (8.1) 6.3 (1.4) 22 9.1 60 60
V 111–168 75.6 (8.1) 6.3 (1.4) 22 12.2 40 120
VI 169–175 75.6 (8.1) 6.3 (1.4) 11 24.2 40 120
VII 176–209 122.8 (7.2) 3.4 (1.4) 11 24.2 40 120
VIII 210–236 122.8 (7.2) 3.4 (1.4) 7.3 36.4 40 120
Standard deviation is given between parentheses; TNin, influent TN concentration.

by the balance-box during this time. When this anaerobic removal of nitrogen and phosphorus. For the SBMBR, the
phase was over the next cycle began with the aerobic phase. effluent of one cycle was collected in a beaker and the
The CMBR was operated continuously in aerobic mixture used for analysis, while the instantaneous effluent
environment through the whole experimental period. The was used for CMBR.
influent and the effluent were not interrupted. Other
conditions were the same as the SBMBR. 2.5.2. Cyclic studies
This test was carried out when the performance of
2.2. Feed medium SBMBR was stable. It was applied to the SBMBR for
optimizing nutrient removal, which includes a series of
Synthetic wastewater fed to the reactors consisted of sampling and analysis during one cycle. t = 0 min was set for
sucrose, NH4Cl, KH2PO4 and mineral solution containing the starting of the anaerobic phase. The sampling was done
MgSO47H2O, CaCl22H2O, FeSO42H2O and NaCl. The by manual adjustment peristaltic pump during the anaerobic
initial complete influent contained 400 mg COD/L, 20 mg phase. Cyclic studies could clarify the transformation of
NH4+-N/L and 4 mg PO4+-P/L. Holding the COD and TP various pollutions during the cycle time and provide
influent concentration, a gradual increase of TN concentra- information for adjusting the operation parameters.
tion was adopted during the experiment (Table 1).

2.3. Inocula 2.5.3. Specific nitrification rate test

The specific nitrification rate (mmol NH4+-N/g VSS) was
Two reactors were seeded with sludge from the feedback determined using two 0.5 L reactors in duplicate for the
sludge tank of the municipal wastewater treatment plant sludge-nitrification performance in SBMBR and CMBR.
located in Chun-liu, Dalian, pre-incubated before being Fifty millilitres mixed liquor was centrifuged at 3500 rpm
equally divided into the above two reactors. The sludge for 10 min and the liquid phase was decanted. The
concentration in both reactors was 2 g/L MLSS to ensure the concentrated sludge was resuspended in 400 mL synthetic
same behavior of the sludge. wastewater with 50 mL pH 7.0 phosphoric acid buffer
solution added. The temperature was constant at 25 8C. The
2.4. Analyses air was injected. At intervals of 30 min samples were taken
and immediately filtered. The content of NH4+-N was
Ammonia nitrogen (NH4+-N), nitrate nitrogen (NO3-N), measured. According to the NH4+-N decreasing rate,
nitrite nitrogen (NO2-N), total phosphorus (TP) and COD specific nitrification rate could be calculated.
were analyzed according to the standard methods for the
examination of water and wastewater (APHA) [17]. The
total organic carbon (TOC) and total nitrogen (TN) contents 3. Results
were determined with a Total Organic Carbon analyzer
(TOC-VCPH, SHIMADZU). Extracellular polymeric sub- 3.1. Performance of the SBMBR and CMBR
stances (EPS) were determined using the method reported
by Zhang et al. [18]. 3.1.1. Organic substance removal
Fig. 2 presents the varieties of COD concentration and its
2.5. Experiments carried out removal efficiency of the two systems during the whole
operation period. Both SBMBR or CMBR showed good
2.5.1. Comparison experiments performance in organic carbon removal. The COD removals
These experiments were conducted for the comparison averaged at 97.7 and 94.9%, respectively. These results
between the capacities of SBMBR and CMBR in the indicated that change of COD/TN ratio in influent did not
中国科技论文在线 http://www.paper.edu.cn
90 H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95

Fig. 2. Comparison of COD concentrations and removal efficiencies in the two systems. Symbols: (*) CMBR effluent; (~) SBMBR effluent; (*) CMBR
COD removal; (~) SBMBR COD removal.

affect COD removal efficiency significantly for both efficiency climbed to 93.1%, was not effect by load increase
reactors. (phase VI). Since the influent COD/TN ratio decreased
further to 3.4 (TN load was 0.33 – 0.36 kg TN/(m3 days)) on
3.1.2. Nitrogen and phosphorus removal the 176th day, an even lower NH4+-N and TN removal
Fig. 3 illustrates the profile of NH4+-N, TN and TP efficiency was obtained, averaged at 75.2 and 20.7%,
concentrations as well as their removal efficiencies in the respectively (phase VII). On the 210th day, the f B was raised
two systems during the whole experiment. It can be seen that again to enhance TN removal (TN load was 0.50 –
in SBMBR, when COD/TN >19.4 (phases I and II) a high 0.60 kg TN/(m3 days)). TN removal was enhanced to
NH4+-N removal efficiency (>90%) has been achieved. TN 36.4%, while a low removal efficiency of NH4+-N
and TP removal in this period averaged at 91.5 and 93.3%, (<50%) occurred, and the effluent average concentration
respectively. NH4+-N removal efficiency decreased to about reached 70.1 mg/L (phase VIII). This result indicates that a
80% when the influent COD/TN ratio reduced to 9.4 (phase favorable nitrogen removal could not be reached under this
III). On increasing the air flow rate from 100 to 150 L/h, the low COD/TN ratio.
NH4+-N removal increased rapidly to over 90% and the After lengthening the aerobic time on the 111th day, the
effluent NH4+-N concentration decreased to below 1 mg/L. SBMBR phosphorus removal worsened and the average
This indicates that the higher TN load was the more drastic removal efficiency was 49.9%. Since the higher f B was adopted
competition for dissolve oxygen between heterotrophic on the 169th day and organic carbon supply increased and
bacteria with nitrifying bacteria. The increased intensity of the phosphorus removal recovered. Except phase V, the
injected air enhanced the penetrability of oxygen into SBMBR phosphorus removal efficiency was kept at about 90%
biological flocs, with benefits for both types of organism. TN and the average removal efficiency was 91.4%.
removal efficiency declined temporarily but soon recovered. As far as the CMBR was concerned, the TN and TP
The removal of TP did not make significant change during removal efficiency was 30.5 and 35.3%, respectively, when
this period. the influent COD/TN ratio was 19.4 and 28.2. During this
It was after 73 days when the influent COD/TN ratio time the NH4+-N average removal efficiency was 93.3%.
tuned down to 6.3 that the NH4+-N removal efficiency of This result indicated that the limiting step of nitrogen
SBMBR started a sharp decrease and the maximum effluent removal in the CMBR was denitrification. TN concentration
NH4+-N concentration had reached to 76.1 mg/L (phase IV). in effluent was close to the level in the influent at COD/
On the 111th day, the SBMBR cycle time was adjusted to TN = 6.3 and TP removal efficiency fluctuated from 14 to
40 min anaerobic and 120 min aerobic (total 160 min) from 95% under influent experiment conditions.
60 min anaerobic and 60 min aerobic (total 120 min)
(Table 1). This step was to enhance nitrification. Nitrifica- 3.2. Cyclic studies of SBMBR
tion improved and the NH4+-N removal efficiency increased
to 95% (Fig. 3(a)). But the average TN removal efficiency During phases I–III a higher removal efficiency of NH4+-
was only 31.9% in this period as shown in Fig. 3(b) (phase N, TP and TN persisted in the SBMBR. The DO cyclic test
V). In order to enhance denitrification and increase the of SBMBR on the 28th day (Fig. 4(a)) showed that after the
organic carbon supply, a higher f B (24.2%) was adopted on aerobic period the massive high volume influent can
the 169th day. The TN load was 0.22 kg TN/(m3 days). decrease the DO concentration quickly in reactor, thus
Denitrification was enhanced and TN removal efficiency sufficient anaerobic time was ensured, which contributed to
was improved to 67.6%. As to NH4+-N, the average removal higher nitrogen and phosphorus removal in the SBMBR.
中国科技论文在线 http://www.paper.edu.cn
H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95 91

Fig. 3. The comparison of nitrogen and phosphorus concentrations and removal efficiencies in the two systems: (a) NH4+-N, (b) TN and (c) TP. Symbols: (—
*—) CMBR effluent; (—~—) SBMBR effluent; (—*—) CMBR removal efficiency; (—~—) SBMBR removal efficiency.

The cyclic test on the 109th day (Fig. 4(b)) showed that anaerobic and 120 min aerobic from 60 min anaerobic and
the time was not long enough for NH4+-N removal when 60 min aerobic.
COD/TN ratio was 6.3 with the aerobic time-controlled at This and the cyclic test on the 167th day (Fig. 4(c))
1 h (i.e. nitrification under this condition was not complete). indicates that nitrification is complete within 120 min
The anaerobic time 60 min has a margin for denitrifying and aerobic time but denitrification was restrained. In order to
phosphorus release. A shorter anaerobic time was therefore enhance denitrification without changing anaerobic time, a
adopted in the following days to avoid the secondary higher f B (24.2%) was adopted to increase the organic
phosphorus release. The cycle time was adjusted to 40 min carbon supply.
中国科技论文在线 http://www.paper.edu.cn
92 H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95

Fig. 5. Ammonia nitrogen degradation rate.

decrease pH value. The existence of an anoxic phase in the

SBMBR provides a condition for denitrification, which
counteracted the alkalinity consumption of nitrification. The
effluent pH value of this reactor remained 7.0–8.0 during the
whole experiment, which was just suitable for the
nitrification reactions. The case in the CMBR was different.
It suited the nitrification reaction but hardly any denitrifica-
tion occurred. The higher influent TN load the more
alkalinity was consumed. An obvious decline of effluent pH
value was observed and a low pH value (5.0–6.0) appeared
even at the COD/TN ratio of 9.4. Therefore, the nitrification
activity was inhibited which led to a lower NH4+-N removal.
The specific nitrification rate test demonstrated that pH
value could affect the activity of nitrobacteria but not be the
destructive factor.

3.4. Sludge characterization

Sludge in the two reactors appeared different because of

the different operation mode. The colour of the sludge was
khaki in the CMBR and filemot in the SBMBR. Microscopic
examination revealed a different community structure
between the two systems. Categories of microbes were
plentiful in the CMBR. Filamentous fungi and protozoa
organisms, including Ciliates (mainly Aspidisca and
Fig. 4. DO profile on the 28th (a) and nutrient profiles on the 109th (b) and Lionotus), Amoebae (mainly Amoeba radiosa) and Rotifers
167th (c) day, respectively, during cyclic tests. Symbols: (—*—) TP; (—
were observed abundantly. Metazoans including Aeoloosma
~—) NH4+-N; (—5—) NO3-N.
hemprichii also appeared in this system. In the SBMBR, the
dominant community of bacteria existed as Epistylis and
3.3. Specific nitrification rate test Vorticella. A small quantity of Aspidisca and Rotifers was
found in this ecosystem. Fig. 6 displays the microscopy
In Fig. 5, the NH4+-N concentration versus time profile photographs of sludge in both reactors.
from SNR batch test is shown. For CMBR and SBMBR, the
specific nitrification rate were 0.56 and 0.40 mmol NH4+-N/ 3.5. Membrane fouling
(g VSS days), respectively. This indicates that in compar-
ison with SBMBR, the sludge in the CMBR possesses a Membrane fouling is one of the main problems of MBR
higher nitrification activity. systems. The avoidance of fouling or reduction of its severity
As the pH value was considered, the optimal pH range for has been pursued for a long time. Researchers have reported
the nitrobacteria is 7.0–8.0 and the result of nitrification is to that intermittent suction and low trans-membrane pressure
中国科技论文在线 http://www.paper.edu.cn
H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95 93

Fig. 7. Trans-membrane pressure (TMP) profiles.

(TMP) helped to prevent irreversible damage to the

membranes [16,19]. It has also been reported that
extracellular polymeric substances (EPS) in MBR can
increase the viscosity of the mixed liquid and the membrane
filtration resistance, which would accelerate membrane
fouling [20–23].
In this work, two parameters, TMP and EPS, were
introduced to reflect the membrane fouling of the two
experimental systems. As slight difference of TMP
(TMP < 1 kPa) was observed between CMBR and SBMBR
until a higher f B (Table 1) was conducted in phase VIII. It
can be seen from Fig. 7 that a rapid increase of TMP
occurred in the CMBR from less than 10 to about 60 kPa.
Hydrocleaning was carried out on the 225th day and then the
membrane filtration performance recovered. There is no
significant increase of TMP value observed in the SBMBR
before the 225th day. The sharp increase of TMP 17–54 kPa
between the 226th and 227th day was caused by the
malfunction of the air-injection system. When the malfunc-
tion was overcame and the membrane was cleaned with tap
water, the membrane filtration performance recovered. This
demonstrated that the membrane-fouling rate in the CMBR
is greater than in the SBMBR, although the membrane

Fig. 6. Microscopically photographs of sludge in both reactors.

Fig. 8. The comparison of EPS concentration in mixed liquid.

中国科技论文在线 http://www.paper.edu.cn
94 H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95

flux of the SBMBR was more than 1.33–2 times of the is ensured. Third, f B and cycles of the SBMBR can be
CMBR. adjusted easily according to the nitrogen and phosphorus
At the beginning of experiment, the MLSS was about concentration in the influent.
2 g/L in both systems. The activated sludge was discharged Attached Ciliates were always the dominant community
according to the same SBR during all experiments. At the in the SBMBR, which indicated that the sequencing batch
end of the experiment, the MLSS was almost the same in the operation is suitable for growth of attached Ciliates. The
SBMBR and the CMBR, which reached to around 5.6 g/L. sludge floc would be bigger and tighter when the attached
The EPS concentrations in both sludge systems varied Ciliates are dominant. The liquid filter capacity is better and
between 340 and 500 mg/L as shown in Fig. 8. A higher this is one reason why the membrane-fouling rate of the
concentration persisted in the SBMBR than in the CMBR SBMBR is lower than that of the CMBR. Another reason
during the whole experimental examination. These results might be attributed to the sequencing batch operation mode.
did not agree with the literature that there is a direct Agitation was introduced into the anaerobic phase when the
correlation between EPS concentration and membrane suction was discontinued. As a result, the aggregation of
fouling. substance onto the membrane surface was reduced but the
diffusion of adsorbents from the membrane surface back to
the liquid was enhanced. Therefore, EPS in the SBMBR was
4. Discussion always higher than that in the CMBR, but the membrane-
fouling rate of the SBMBR was lower than that in the
The CMBR and SBMBR have higher organics removal, CMBR.
the average COD removal were 94.9 and 97.7%, respec-
tively. COD removal decreased to about 80% as the
experiment was close to the end (Fig. 2). The influent of the 5. Conclusions
SBMBR is high and of high flow at the beginning of
anaerobic time and fresh nutrients was not put into the Nutrient removal from synthetic wastewater was studied
reactor during one cycle. The microorganism in the SBMBR using membrane bioreactors in both sequencing batch and
experienced following the log phase, stationary phase and continuous operations. The main findings from this study are
death phase. The influent of the CMBR was continuous. as follows:
Along with the MLSS increasing and F:M decreasing, the
microorganism of CMBR would get into death phase for (1) Both SBMBR and CMBR showed good performance on
long time for lack of nutrients, need to rely on themselves for organic substance removal during the whole experi-
obtaining energy, the products of metabolism were mental period. The average COD removals were 97.7
accumulated and led to worsening of the quality of the and 94.9%, respectively. Change of COD/TN ratio in
final effluent in CMBR. influent did not affect COD removal efficiency
After lengthening the aerobic time on the 111th day, significantly in either reactor.
enhanced nitrification resulted in a high nitrate concentra- (2) The SBMBR demonstrated a good performance on
tion in the anaerobic phase. Consequently, there is a nitrogen and phosphorus removal at different influent
competition for the organic carbon between denitrifying COD/TN ratios. Even running at low COD/TN ratio
bacteria and PAOs. The denitrifying bacteria using nitrate as (COD/TN = 6.3) with 0.22 kg TN/(m3 days), in sui-
electron acceptor can obtain more energy and were more table operation conditions, TN and NH4+-N removal
competitive, thus organic carbon was not converted into efficiencies were maintained above 65 and 90%,
VFAs with inhibition of phosphorus release of PAOs and respectively. As to TP, the removal efficiency was
formation of PHB. During phase V, the phosphorus removal maintained at approximately 90% all through the
worsened and the average removal efficiency decreased to whole experiment except for the starved organic
49.9%. The sequencing batch operation of the SBMBR in substrate feeding.
this experiment can increase organic carbon supply of one (3) The variation of influent COD/TN ratio affected the
cycle by increase f B and, does not require an exogenous CMBR more extensively than the SBMBR. TP removal
supply of organic carbon. After adjusting f B, on 169th day, efficiency fluctuated between 14 and 95% under influent
the TP average removal efficiency increased to about 90% experiment conditions and the TN concentration in the
and TN removal improved. effluent was close to that in the influent at the feed ratio
The CMBR showed some nitrogen and phosphorus COD/TN = 6.3.
removal by the existing anaerobic microenvironment, but (4) Specific nitrification rate tests demonstrated that the pH
the capacity was lower than that of the SBMBR. The reasons value could affect the activity of nitrifying bacteria with
are: first, the sequencing batch operation of the SBMBR can no irreversible harm.
offer alternative anaerobic/aerobic environments; second, (5) The sequencing batch operation mode can reduce
the influent was high and of high volume at the beginning of membrane fouling although a higher EPS existed in the
anaerobic time in the SBMBR and sufficient anaerobic time SBMBR system.
中国科技论文在线 http://www.paper.edu.cn
H.-M. Zhang et al. / Process Biochemistry 41 (2006) 87–95 95

Acknowledgements [10] Ueda T, Hata K. Domestic wastewater treatment by a submerged

membrane bioreactor with gravitational filtration. Water Res 1999;
33(12):2888–92.
This research was partly supported by a grant from [11] Ng WJ, Ong SL, Gomez MJ, Hu JY, Fan XJ. Study on a sequencing
Chinese National Science Foundation Committee (Grant batch membrane bioreactor for wastewater treatment. Water Sci
No. 50308004). Special thanks go to Dr. Chu LiBing for her Technol 2000;41(10–11):227–34.
kind help and valuable discussion. [12] Bae T-H, Han S-S, Tak T-M. Membrane sequencing batch reactor
system for the treatment of dairy industry wastewater. Process Bio-
chem 2003;39(2):221–31.
[13] Ueda T, Hata K, Kikuoka Y. Treatment of domestic sewage from rural
References settlements by a membrane bioreactor. Water Sci Technol 1996;34(9):
189–96.
[1] Pochana K, Keller J. Study of factors affecting simultaneous nitrifica- [14] Benı́tez J, Rodrı́guez A, Malaver R. Stabilization and dewatering of
tion and denitrification (SND). Water Sci Technol 1999;39(6):61–8. wastewater using hollow fiber membranes. Water Res 1995;29(10):
[2] Cicek N, Winnen H, Suidan MT, Wrenn BE, Urbain V, Manem J. 2281–6.
Effectiveness of the membrane bioreactor in the biodegradation of [15] Chiemchaisri C, Yamamoto K. Performance of membrane separation
high molecular weight compounds. Water Res 1998;32(5):1553–63. bioreactor at various temperatures for domestic wastewater treatment.
[3] Konopka A, Zakharova T, Oliver L, Camp D, Turco RF. Biodegrada- J Membr Sci 1994;87:119–29.
tion of organic wastes containing surfactants in a biomass recycle [16] Krampe J, Krauth K. Sequencing batch reactor with submerged hollow
reactor. Appl Environ Microbiol 1996;62:3292–7. fibre membranes for the biomass separation. Water Sci Technol
[4] LaPara TM, Konopka A, Nakastu CH, Alleman JE. Thermophilic 2001;43(3):195–9.
aerobic treatment of a synthetic wastewater in a membrane-coupled [17] APHA/AWWA/AWEF. Standard methods for the examination of
bioreactor. J Ind Microbiol Biotechnol 2001;26:203–9. water and wastewater, 18th ed. Washington, DC; 1992.
[5] Muller EB, Stouthamer AH, van Verseveld HW, Eikelboom MDH. [18] Zhang X, Bishop PL, Kinkle BK. Comparison of extraction methods
Aerobic domestic wastewater treatment in a pilot plant with complete for quantifying extracellular polymers in biofilms. Water Sci Technol
sludge retention by cross-flow filtration. Water Res 1995;29(4): 1999;39(7):211–8.
1179–89. [19] Yamamoto K, Hiasa M, Mahmood T, Matsuo T. Direct solid–liquid
[6] Chiemchaisri C, Yamamoto K, Vigneswaran S. Household membrane separation using hollow fiber membrane in an activated sludge aera-
bioreactor in domestic wastewater treatment. Water Sci Technol tion tank. Water Sci Technol 1989;21(4):43–54.
1993;27(1):171–8. [20] Chang I-S, Lee C-H. Membrane filtration characteristics in mem-
[7] Knoblock MD, Sutton PN, Gupta K, Janson A. Membrane biological brane-coupled activated sludge system—the effect of physiological
reactor system for the treatment of oily wastewaters. Water Environ states of activated sludge on membrane fouling. Desalination 1998;
Res 1994;66(2):133–9. 120:221–33.
[8] Machenbach I. Membrane technology for dyehouse effluent treatment. [21] Cho BD, Fane AG. Fouling transients in nominally sub-critical flux
Membr Technol 1998;96:7–10. operation of a membrane bioreactor. J Membr Sci 2002;209:391–403.
[9] Yeom I-T, Nah Y-M, Ahn K-H. Treatment of household wastewater [22] Nagaoka H, Yamanishi S, Miya A. Modeling of biofouling by extra-
using an intermittently aerated membrane bioreactor. Desalination cellular polymers in a membrane separation activated sludge system.
1999;124(1–3):193–204. Water Sci Technol 1998;38(4–5):497–504.