Membrane Separation Bioreactors For Wastewater Treatment

Critical Reviews in Environmental Science and
Technology
ISSN: 1064-3389 (Print) 1547-6537 (Online) Journal homepage: http://www.tandfonline.com/loi/best20
Membrane Separation Bioreactors for Wastewater

Treatment
C. Visvanathan , R. Ben Aim & K. Parameshwaran
To cite this article: C. Visvanathan , R. Ben Aim & K. Parameshwaran (2000) Membrane
Separation Bioreactors for Wastewater Treatment, Critical Reviews in Environmental Science and
Technology, 30:1, 1-48, DOI: 10.1080/10643380091184165
To link to this article: http://dx.doi.org/10.1080/10643380091184165
Published online: 03 Jun 2010.
Submit your article to this journal
Article views: 4877
View related articles
Citing articles: 196 View citing articles
Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=best20
Download by: [121.54.32.98] Date: 19 June 2017, At: 07:26

Critical Reviews in Environmental Science and Technology, 30(1):148 (2000)
Membrane Separation Bioreactors for

Wastewater Treatment
C. Visvanathan,1 R. Ben Aim,2 and K. Parameshwaran 3
1Environmental Engineering Program, Asian Institute of Technology, P.O. Box 4, Klong
Luang, Pathumthani 12120, Thailand; Email: visu@ait.ac.th; 2Institute National des

Sciences Appliques de Toulouse, Complexe Scientifique de, Rangueil 31077, Toulouse
Cedex, France; 3Center for Membrane Science and Technology, The University of
New South Wales, Sydney 2052, Australia
ABSTRACT: With continuing depletion of fresh water resources, focus has shifted more toward
water recovery, reuse, and recycling, which require an extension of conventional wastewater treat-
ment technologies. Downstream external factors like stricter compliance requirements for wastewater
discharge, rising treatment costs, and spatial constraints necessitate renewed investigation of alter-
native technologies. Coupled with biological treatment processes, membrane technology has gained
considerable attention due to its wide range of applicability and the performance characteristics of
membrane systems that have been established by various investigations and innovations during the
last decade. This article summarizes research efforts and presents a review of the how and why of
their development and applications. The focus is on appraising and comparing technologies on the
basis of their relative merits and demerits. Additional facts and figures, especially regarding process
parameters and effluent quality, are used to evaluate primary findings on these technologies. Key
factors such as loading rates, retention time, cross-flow velocities, membrane types, membrane
fouling, and backwashing, etc. are some of the aspects covered. Membrane applications in various
aerobic and anaerobic schemes are discussed at length. However, the emphasis is on the use of
membranes as a solid/liquid separator, a key in achieving desired effluent quality. Further, technol-
ogy development directions and possibilities are also explored. The review concludes with an
economic assessment of the technologies because one of the key technology selection criteria is
financial viability.
KEY WORDS: membrane bioreactor, membrane technology, solid/liquid separation, membrane air
diffusers, membrane fouling, backwashing, micro-porous membranes.
I. INTRODUCTION
The use of biological treatment can be traced back to the late nineteenth
century. By the 1930s, it was a standard method of wastewater treatment (Rittmann,
1987). Since then, both aerobic and anaerobic biological treatment methods have
been commonly used to treat domestic and industrial wastewater. During the
course of these processes, organic matter, mainly in soluble form, is converted into
Copyright 2000, CRC Press LLC Files may be downloaded for personal use only. Reproduction
of this material without the consent of the publisher is prohibited.
1
H2O, CO2, NH4+, CH4, NO2, NO3 and biological cells. The end products differ
depending on the presence or absence of oxygen. Nevertheless, biological cells are
always an end product, although their quantity varies depending on whether it is
an aerobic or anaerobic process. After removal of the soluble biodegradable matter
in the biological process, any biomass formed must be separated from the liquid
stream to produce the required effluent quality. A secondary settling tank is used
for the solid/liquid separation and this clarification is often the limiting factor in
effluent quality (Benefield and Randall, 1980).
In recent years, effluent standards have become more stringent in an effort to
preserve existing water resources. Recycling and reuse of wastewater for second-
ary purposes is on the rise due to dwindling natural resources, increasing water
consumption, and the capacity limitations of existing water and wastewater con-
veyance systems. In both cases, achieving a high level of treatment efficiency is
imperative.
The quality of the final effluent from conventional biological treatment sys-
tems is highly dependent on the hydrodynamic conditions in the sedimentation
tank and the settling characteristics of the sludge. Consequently, large volume
sedimentation tanks offering several hours of residence time are required to obtain
adequate solid/liquid separation (Fane et al., 1978). At the same time, close control
of the biological treatment unit is necessary to avoid conditions that lead to poor
settleability and/or bulking of sludge. Very often, however, economic constraints
limit such options. Even with such controls, further treatment such as filtration,
carbon adsorption, etc. are needed for most applications of wastewater reuse.
Therefore, a solid/liquid separation method different from conventional methods
is necessary.
Application of membrane separation (micro- or ultrafiltration) techniques for
biosolid separation can overcome the disadvantages of the sedimentation tank and
biological treatment steps. The membrane offers a complete barrier to suspended
solids and yields higher quality effluent. Although the concept of an activated
sludge process coupled with ultrafiltration was commercialized in the late 1960s
by Dorr-Oliver (Smith et al., 1969), the application has only recently started to
attract serious attention (Figure 1), and there has been considerable development
and application of membrane processes in combination with biological treatment
over the last 10 years.
This emerging technology, known as a membrane bioreactor (MBR), offers
several advantages over the conventional processes currently available. These
include excellent quality of treated water, which can be reused for industrial
processes or for many secondary household purposes, small footprint size of the
treatment plant, and reduced sludge production and better process reliability.
The purpose of this monograph is to provide a comprehensive review of
membrane bioreactor technology. The application of membranes in different stages
of biological treatment processes, the historical development of membrane bioreators,
2
FIGURE 1. Number of studies published on MBR.
and factors affecting the design and performance of MBR processes are discussed.
A number of case studies for each type of major MBR application along with some
cost information on MBR processes is also presented.
II. FEATURES OF MEMBRANE APPLICATION IN BIOLOGICAL

WASTEWATER TREATMENT
As our understanding of membrane technology grows, they are being applied

to a wider range of industrial applications and are used in many new ways for
wastewater treatment. Membrane applications for wastewater treatment can be
grouped into three major categories (Figure 2): (1) biosolid separation, (2) biomass
aeration, and (3) extraction of selected pollutants. Biosolid separation is, however,
the most widely studied and has found full-scale applications in many countries
(Table 1). Use of combined night-soil treatment and wastewater reclamation at
plant scale operations in buildings in Japan are examples of some successful
applications, and in these cases membrane-coupled technology is considered a
standard process (Yamamoto and Urase, 1997). Solid/liquid separation bioreactors
employ micro- or ultrafiltration modules for the retention of biomass for this
purpose. The membranes can be placed in the external circuit of the bioreactor or
they can be submerged directly into the bioreactor (Figure 2a).
Asymmetric membranes consist of a very dense top layer or skin with a
thickness of 0.1 to 0.5 m, supported by a thicker sublayer. The skin can be placed
either on the outside or inside of the membrane, and this layer eventually defines
the characterization of membrane separation.
3
TABLE 1
Commercial Scale Solid/liquid Separation MBR Plants
Commercial Number Capacity

Company name Country Type of Waste of Plants (m3/d) Ref.
Rhone Poulenc-TechSep UBIS France Domestic >40 <400 Roullet, 1989

Dorr Oliver MSTS USA Domestic 1 13.6 Smith et al., 1969
Thetfort Syst Cycle-LET USA Domestic >30 <200 Irwin, 1990
Kubota Kubota Japan Domestic 8 10110 Ishida et al., 1993
Kubota UK Domestic 1 96 Brindle and Stephenson, 1997
Mitsui Petrochemical Industries ASMEX Japan Human excreta >40 Lambert, 1983
Zenon Env Inc. Zenogem Canada Industrial 1 116 Knoblock et al., 1994
Dorr Oliver MARS USA Industrial 1 38 Li et al., 1984
Membratek ADUF RSA Industrial 2 80/500 Ross and Strohwald, 1994
4
SITA/lyonnaise des Eaux France Landfill leachate 3 1050 Trouve et al., 1994a
Membratek S.Africa Industrial 2 100500 Brindle and Stephenson, 1997
Grantmij Germany Landfill leachate 3 1050 Brindle and Stephenson, 1997
Degrement France Industrial 1 500 Brindle and Stephenson, 1997

A submerged membrane should be outer-skinned. In general, permeate is
extracted by suction or, less commonly, by pressurizing the bioreactor. In the
external circuit, the membrane can be either outer- or inner-skinned, and the
permeate is extracted by circulating the mixed liquor at high pressure along the
membrane surface. In the later case, the concentrated mixed liquor at the feed side
is recycled back to the aeration tank.
FIGURE 2. Features of membrane application in biological wastewater treatment. (B,

bioreactor; M, membrane module; I, influent; E, effluent.) (Adapted from Brindle and
Application in Wastewater Treatment.)
5
Gas-permeable porous membranes can be used to aerate the mixed liquor in the
aeration tank by bubbleless oxygen mass transfer (Yasuda and Lamaze, 1972). At
the same time, they can be used for fine bubble aeration (Semmens, 1989; Matsuoka
et al., 1992). In certain cases, the membrane can act as support for biofilm
development, with direct oxygen transfer through the membrane wall in one
direction and nutrient diffusion from the bulk liquid phases into the biofilm in the
other direction (Brindle and Stephenson, 1996). Because the membranes can form
bubble-free or fine-bubble mass transfer, the efficiency is very high.
Conventional membrane modules can be used in either a flow-through or dead-
end mode as presented in Figure 2b. In the flow-through mode, the air or oxygen
is continuously pumped through the hollow fibers and gas is vented to keep the
partial pressure of oxygen high along the membrane. In the dead-end mode, the
membrane is pressurized with air or oxygen by sealing one end of the fibers or by
sending the gas from both ends. Most studies reported to date have focussed on the
flow-through mode, and researchers argue that the dead-end mode should be
avoided because it significantly reduces performance and may result in water vapor
condensation inside the membrane fibers. However, because air or oxygen is
vented out in the flow-through system, part of the pumped gas is wasted, and thus
the gas transfer efficiency is reduced. In addition, volatile organic compounds
(VOCs) can diffuse across the membrane into the air stream (Semmens, 1989),
VOCs in wastewater can be very effectively stripped and vented off to the atmo-
sphere. Both these problems can be overcome in the dead-end mode. Also, as the
total amount of air/oxygen supplied should diffuse through the membrane module,
the efficiency is improved and VOCs stripped off can be minimized if not com-
pletely reduced.
An extractive membrane bioreactor was developed to extract (by dialysis)
toxic organic pollutants present in industrial wastewater to a bio-medium for
subsequent degradation (Livingston, 1994). In dialysis mode, organisms can be
maintained in an optimal growth environment through nutrient supplementation
while at the same time digesting inhibitory or recalcitrant compounds that diffuse
across the membrane. Mass transfer of the pollutants across the membrane is
driven by a concentration gradient, because the bio-medium passing on the mem-
brane walls acts as a sink. Although these three applications are described sepa-
rately, they are not mutually exclusive, and they may be coupled together to
achieve added advantages for each process (Brindle and Stephenson, 1997). For
example, a study by the authors to use hollow fiber membrane for solid/liquid
separation and aeration in alternate cycles indicates such coupling (Parameshwaran
et al., 1998).
III. DEVELOPMENT OF MEMBRANE BIOREACTORS
Membranes have been finding wide application in water and wastewater

treatment ever since the early 1960s when Loeb and Sourirajan invented an
6
asymmetric cellulose acetate membrane for reverse osmosis. Many combinations
of membrane solid/liquid separators in biological treatment processes have been
studied since. The trends that led to the development of todays MBR are depicted
in Figure 3. When the need for wastewater reuse first arose, the conventional
approach was to use advanced treatment processes (Figure 3a). For irrigation, this
treatment may be limited to filtration and disinfection, whereas for building reuse
or ground water recharge it may also include reverse osmosis (RO). For example,
Water Factory 21 in Orange Country (California, USA) uses a treatment process
that consists of lime softening, air stripping, recarbonation, sand filtration, carbon
adsorption, and RO for biologically treated effluent (Mills, 1996). The treated
water is used to recharge the ground water. This scheme is relatively complex and
produces large amounts of chemical sludge.
The progress of membrane manufacturing technology and its applications
could lead to the eventual replacement of tertiary treatment steps by microfiltration
or ultrafiltration and this simplified method is being evaluated at Water Factory 21
in the U.S. Parallel to this development, microfiltration or ultrafiltration was used
for solid/liquid separation in the biological treatment process and the sedimenta-
FIGURE 3. Trends in MBR development.
7
tion step could also be eliminated. By pumping the mixed liquor at a high pressure
into the membrane unit, the permeate passes through the membrane and the
concentrate is returned to the bioreactor (Hardt et al., 1970; Arika et al., 1977;
Krauth and Staab, 1988; Muller et al., 1995). However, higher energy costs to
maintain the crossflow velocity led to the next stage of development submerg-
ing the membranes in the reactor itself and withdrawing the treated water through
membranes (Yamamoto et al., 1989; Kayawake et al., 1991; Chiemchaisiri et al.,
1993; Visvanathan et al., 1997). In this development, membranes were suspended
in the reactor above the air diffusers. The diffusers provided the oxygen necessary
for treatment to take place and scour the surface of the membrane to remove
deposited solids. In a parallel attempt to save energy in membrane coupled
bioreactors, the use of jet aeration in the bioreactor has been investigated (Yamagiwa
et al., 1991). The main feature is that the membrane module is incorporated into
the liquid recirculation line for the formation of the liquid jet such that aeration and
filtration can be accomplished with only one pump. Jet aeration works on the
principle that a liquid jet, after passing through a gas layer, plunges into a liquid
bath entraining a considerable amount of air. The limited amount of oxygen
transfer possible with this technique restricts this process to small-scale applica-
tions. However, using only one pump makes it mechanically simpler and therefore
useful to small communities. The invention of air back-washing techniques for
membrane declogging led to the development of using the membrane itself as both
clarifier and air diffuser (Parameshwaran et al., 1998). In this approach, two sets
of membrane modules are submerged in the aeration tank. While the permeate is
extracted through one set, the other is supplied with compressed air for back-
washing. The cycle is repeated alternatively, and there is a continuous airflow into
the aeration tank, which is sufficient to aerate the mixed liquor.
A. Advantages of MBR
There are many advantages in using a MBR process, the prime ones being the
treated water quality, the small footprint of the plant, and less sludge production
and flexibility of operation.
1. Treated Water Quality
The major problem of conventional activated sludge processes is the settling

of sludge. This is caused by poor flocculation of microfloras or the proliferation of
filamentous bacteria. Because solids and colloids are totally eliminated through
membrane separation, settlement has no effect on the quality of treated water.
Consequently, the system is easy to operate and maintain. This is important with
industrial wastewater, because a lack of nutrients leads to excessive growth of
filamentous organisms resulting in poor settlement. Because the final effluent does
8
not contain suspended matter, this enables the direct discharge of the final effluent
into the surface water and the reuse of effluent for cooling, toilet flushing, lawn
watering, or, with further polishing, as process water.
2. Flexibility in Operation
In a MBR, sludge retention time (SRT) can be controlled completely indepen-

dently from hydraulic retention time (HRT). Therefore, a very long SRT can be
maintained resulting in the complete retention of slow-growing microorganisms
such as nitrifying or methanogenic bacteria and this results in greater flexibility of
operation.
3. Compact Plant Size
Volumetric capacities are typically high because a high sludge concentration

can be maintained independently of settling qualities. HRTs as low as 2 h have
been satisfactorily applied (Chaize and Huyard, 1991), and fluctuations on volu-
metric loading have no effect on the treated water quality (Chiemchaisri et al.,
1993). For example, sludge concentrations between 25 and 30 kg/m3 have been
achieved regularly as opposed to the more common 4 to 6 kg/m3 in the conven-
tional aerobic process (Yamamoto and Win, 1991). Moreover, the higher turbu-
lence maintained within the mixed liquor to prevent the membrane from fouling
also prevents the flocculation of biosolids and keeps them highly dispersed. An
analysis on the floc size distribution of MBR sludge and conventional activated
sludge indicates that the floc size in the MBR (a number of samples from different
MBR plants were analyzed) are very much smaller than 100 m and concentrated
within a small range. On the other hand, floc size from conventional activated
sludge processes varied from 0.5 to 1000 m (Zhang et al., 1997). The smaller
flocs from MBRs could stimulate a higher oxygen and/or carbon substrate
mass transfer and thus higher activity levels in the system. Zhang and co-workers
(1997) also found that nitrification activities in MBR processes averaged 2.28 g
NH4N/kg MLSS.h, which was greater than in conventional processes (0.95 g
NH4N/kg MLSS.h). Also, there is an enormous saving in space with MBRs
because there is no need for secondary settling devices and post-treatment to
achieve reusable quality.
4. High Rate Decomposition
Treatment efficiency is also improved by preventing leakage of undecomposed

polymer substances. If these polymer substances are biodegradable, they can be
broken down with a reduction in the accumulation of substances within the
9
treatment process. On the other hand, dissolved organic substances with low
molecular weights, which cannot be eliminated by membrane separation alone, can
be broken down and gasified by microorganisms or converted into polymers as
constituents of bacterial cells, thereby raising the quality of the treated water. For
example, the permeate from microfiltration of screened raw sewage (feed average
BOD5 = 230 mg/l) had an average BOD5 of 93 mg/l. This was mainly the soluble
portion of the influent BOD5, although it showed 99% removal of suspended solids
and 5.8 log removal of fecal coliforms (Johnson et al., 1996). In contrast, most
MBR studies indicate the effluent BOD5 is below 5 mg/l (Parameshwaran and
Visvanathan, 1998; Buisson et al., 1997; Trouve et al., 1994). Due to the high
biomass concentration and the fact that bio-oxidation is an exothermic process,
temperature increase can be maintained at the maximum activity temperature level.
Maximum growth rates are about five times higher than the activity commonly
observed in activated sludge systems. Based on cubic meter of reactor volume,
combining high activity with high biomass concentration results in conversion
rates 10 to 15 times higher than conventional conversion rates (Buisson et al.,
1997), an especially useful feature in cold climates.
5. Low Rate Sludge Production
Studies on MBR indicate that the sludge production rate is very low (Table 2).
Chaize and Huyard (1991) have shown that for treatment of domestic wastewater,
sludge production is greatly reduced if the age is between 50 and 100 days. Low
F/M ratio and longer sludge age in the reactor is generally used to explain this low
production rate.
Praderie (1996) demonstrated that the viscosity of sludge increases with age,
eventually limiting the oxygen transfer in the MBR system. Therefore, he recom-
mends limiting the MLSS concentrate to 15 to 20 g/l for effective oxygen transfer.
It was also noted that with increased age there was greater difficulty in sludge
dewaterability, which could be attributed to excess amount of cellular polymer
formation ( Parameshwaran, 1997; Erikson et al., 1992).
It is also anticipated that micrological activity can be modified with increased
sludge age, but little published information is available on the subject. The initial
microscopic observation (Praderier, 1996; Pliankarn, 1996) on microorganism
population indicates that with increased sludge age, reduction in filamentous
bacteria increased rotifers and nematodes.
6. Disinfection and Odor Control
In this membrane filtration process, the removal of bacteria and viruses can be
achieved without any chemical addition (Pouet et al., 1994; Langlais et al., 1992;
Kolega et al., 1991). Because all the process equipment can be tightly closed, no
odor dispersion occurs. Comparison of conventional biological processes and
MBR is shown in Table 3 and depicts the advantages discussed above.
10
TABLE 2
Comparison of Sludge Production in Conventional Activated Sludge
Process (ASP) and MBR Process Treating Domestic Wastewater
Type of Sludge
Process SRT (d) production Ref.
ASP 1020 0.71 kg MLSS/kg BOD5 Hsu and Wilson, 1992

ASP 14 0.7 kg MLSS/kg BOD5 E.I.A, 1994
ASP 33 0.6 kg MLSS/kg BOD5 E.I.A, 1994
MBR 25 0.53 kg MLVSS/kg BOD5 Trouv et al., 1994a
MBR 25 0.26 kg SS/kg BOD5 Trouv et al., 1994b
MBR 50 0.22 kg MLSS/kg BOD5 Takeuchi et al., 1990
With the exception of wastewater reuse, membrane separation activated sludge

processes have not been widely used. Obstacles to more widespread use include:
High capital and operating costs

Current regulatory standards can be achieved by conventional treatment pro-
cess
Limited experience in use of membranes in these application areas
Lack of interest by the membrane manufacturers
Membranes will only find greater application in the wastewater industry if they
can achieve the required regulatory standards or better at the same or less cost
TABLE 3
Comparison of Operating Data for Conventional, Extended Aeration ASP,
and AS/UF Treatment Processes
Processes
Extended
ASP/UF Conventional aeration
Parameters Unit ASP ASP
System reactor volume 1 2,663 3,423 13,694

Influent BOD mg/l 250 250 250
System MLSS mg/l 10,000 2,500 3,500
Organic loading rate kg BOD/kg. 0.12 0.200.70 0.100.15
MLSS.d
Volumetric loading rate kg BOD /m3.d 1.35 0.59 0.27
Reactor dissolved oxygen mg/l 1.50 1.50 1.50
Sludge retention time d Infinite 20 11
Re-circulation ratio % 240 25 50100
Hydraulic retention time h 5 6 1224
From Smith et al., 1969.
11
compared with present processes, or if regulations were to tighten further such that
conventional processes can no longer achieve the desired effluent quality.
IV. FACTORS AFFECTING THE MBR PROCESS PERFORMANCE
The main aim of membrane-coupled bioreactors is to improve the efficiency

of the biological process step such that high-quality effluent is obtained. Because
biological treatment and membrane separation are rather distinct processes, the
combined MBR process is relatively complex. To optimize the MBR process,
many parameters have to be considered. These include solid concentrations, sludge
age, and the hydraulic retention time (HRT) in the biological step as well as the flux
rate, material costs, and the energy cost of the membrane separation. The treatment
and disposal of the waste sludge also needs to be considered. Comparisons made
on the waste sludge properties of the conventional activated sludge process and the
MBR process indicates that dewatering of MBR waste sludge is difficult compared
with the conventional process. This has been attributed to higher organic matter
content and excess production of extracellular polymers (Parameshwaran, 1997).
As all these parameters are interrelated, optimization is complicated. For example,
an increase in sludge concentration can enhance the biological stage. However,
when sludge concentration exceeds a certain limit, the permeation flux rapidly
declines due to a dramatic rise in the viscosity of the sludge mixture (Praderie,
1996). An increase in sludge concentration can also affect the gas transfer effi-
ciency, and the energy requirements for the aeration therefore increase will (Praderie,
1996).
Permeation flux of membrane filtration is affected by the raw materials of the
membrane and its pore size as well as operational conditions such as the pressure
driving force, the liquid velocity/turbulence, and the physical properties of the
mixed liquor being filtered (Tables 4 to 6).
A. Type of Membrane
Selection of the membrane module plays an important role on the membrane

flux achieved. Membranes can be categorized according to the materials used
(organic or ceramic), membrane type (microfiltration or ultrafiltration), module
type (plate and frame or tubular or hollow fiber), filtration surface (inner skin or
outer skin), as well as the module status (static or dynamic membranes). All are
being tested and many combinations have been considered. There are, however,
overlaps and omissions in the combinations considered largely due to poor com-
munication among international researchers.
The flux will vary depending on the combination considered. For example,
submerged hollow fiber membrane modules (external skin) show the lowest flux
of 3.5 l/m2.h, while ceramic microfilters show the highest of 100 l/m2.h (Tables 4
to 6). Smooth surface membranes (ceramic) offer more resistance to cake layer
12
TABLE 4
Characteristics and Operating Conditions of Aerobic MBR Process (Membrane in External Circuit)
Wastewater type Domestic Synthetic
Membrane configuration UF MF UF MF MF/UF MF/UF MF UF UF

(plate and (hollow (plate and (hollow fiber) (tubular) (hollow (spiral (tubular)
frame) fiber) frame) fiber) wound)
Membrane material Noncellulose Polyvinyl Polysulfone/ Ceramic Polysulfone/ Polyester Polysul- Polysul-
organic acetate cellulose acrylic fone/ fone/
Pore size (Dalton/m) 50,000 0.1 0.1/50,000/ 200,000 50,000 0.01
800,000
Filtration area (m2) 266 0.42 1.1 0.00385 0.1 0.1
Cross flow velocity (m/s) 1.5 15 2.23.6 0.5 4.5
Transmembrane pressure 152186 100200 100 150400 2080 200250 100 135260
(kPa)
Temperature (C) 20 29 20 20 30 25 27
MLSSa (kg/m3) 15 810 3.7 0.8 540 412 640
Flux (L/m2.h) 25 1090b 80100 4.811.4 20 29.2 45
Frequency of cleaning 1/h 1/month
13
Reference Smith et al., 1969 Audic, Chaize and Trouve Muller Suwa Bailey Ishiguro, Lbbecke
1969 1986 Huyard, 1991 et al., 1994c et al.,1995 et al., 1992 1994 1993 et al., 1995
Sour vegetable
Wastewater type Industrial canning Ice cream
Membrane configuration UF UF plate UF UF Tubular

hollow fiber and frame Tubular Tubular

Membrane material Organic Polysulfone Polysulfone Polysulfone Ceramic
Pore size (Dalton/m) 0.04 0.01 0.2
Filtration area (m2) 2 2.17 0.22 0.551.1 0.06
Cross flow velocity (m/s) 2 2.53
Transmembrane pressure (kPa) 140 190390 275 250 10
Temperature (C) 30 3038 31.5 25
MLSSa (kg/m3) 7.512.4 2028 11 47
Flux (l/m2.h) 50 2370 66 40 24
Frequency of cleaning 1/h
Reference Hare et al., Sato and Ishi, Krauth and Stab, Lbbecke Scott and Smith,
1990 1991 1993 et al., 1995 1997
a
Mixed-liquor suspended solids.
b
Unit (l/m2.h.bar).
TABLE 5
Characteristics and Operating Conditions of Aerobic MBR Process (Submerged Membrane)
Wastewater type Synthetic Domestic Industrial Synthetic Synthetic Industrial Domestic
Membrane configuration MF MF MF MF MF MF MF MF
Hollow fiber Hollow fiber Hollow fiber Hollow fiber Hollow fiber Hollow fiber Hollow fiber Hollow fiber
Membrane material Polyethylene Polyethylene Polyethylene Polyethylene Polyethylene Polyethylene Polyethylene Polyethylene
Pore size (m) 0.1 0.1 0.10.2 0.1 0.1 0.1 0.1 0.2
Filtration area (m2) 0.9 0.3 410 0.27 0.6 0.6 0.3 1
Transmembrane 40 13 8 27 80 40 2080 20/44/96
pressure (kPa)
Temperature (C) 23-24 1622 16.6 2530 5 25 29-31
MLSSa (kg/m3) 1011 716 8.3 10.918.2 4 2.5 4.5 1214
Flux (l/m2.h) 9 6 5.5 6.73.5 8.33 12.5 18 6/14/27
Frequency of cleaning
Reference Yamamoto Takeuchi Yamamoto Chiemchaisri, Chiemchaisri Benitez et al., Parameshwaran
14
et al., 1989 et al., 1990 et al., 1991 et al., 1992, 1993 et al., 1992, 1993 et al., 1995 et al., 1998

TABLE 6
Characteristics and Operating Conditions of Anaerobic MBR Process
Pulp High
and strength SS
Wastewater type Brewery Wheat starch paper Distillery Synthetic Industrial High strength
Membrane MF UF UF MF MF MF UF UF MF UF UF
configuration plate and (tubular) (hollow (P and F) (P and F) (tubular)
frame fiber/tubular)
Membrane material Organic Polyethersulfone Polysulfone PVDF
Pore size (Dalton/m) 0.45 40,000 10,000 0.1 2 1 06 2 106 3 106 0.1 2 106 20,000 10,000
Filtration area (m2) 0.012 0.44 - 54 20 12 0.02 0.22
Cross flow 2 1.5 0.9 1.0 0.8 1.52
velocity (m/s)
Transmembrane 150 160 50 40 49 100
pressure (kPa)
Temperature (C) 3540 37 35 35 37
MLSSa (kg/m3) 15.8 30 3138 16.9 15 37.5113.3 15b 7.6
15
Flux (l/m2.h) 30 28 16.25 12.5 3545
Frequency of cleaning 25s/67 min. 1/23 weeks
Reference Anderson, Strohwald Fakhrul- Kimura, Kimura, Nagano Harada Seyfrid and Miami Kitamura, Hall
1984 and Ross, Razi, 1991 1991 et al., 1992 et al., 1994 Broockmann, et al., 1991 1994 et al., 1995
1992 1994 1995
a Mixed-liquor suspended solids.
b Mixed-liquor volatile suspended solids, MLVSS.

adhesion, thus the flux rate is higher. Long lifetime, ability to withstand rigorous
chemical cleaning, and high operating pressures are some of the advantages of
ceramic membranes. The comparisons provided in Tables 4 to 6 indicate that
membrane flux depends not only on the selection of membrane (type, geometry,
etc.), but also on operating conditions such as transmembrane pressure (TMP),
crossflow velocity, etc. Therefore, membrane selection (investment and replace-
ment cost) and operating conditions (energy, cleaning requirements, etc.) dictate
the cost of filtration.
1. Transmembrane Pressure
Membrane filtration performance is affected by the resistance of the membrane

itself and the resistance created by the fluid under going filtration. The resistance
model is illustrated in Figure 4. Based on this model, the permeation flux is
determined by flow resistance in series as in Equation 1.
P
J= (1)
Rt
where J, permeate flux (m3/m2.s); P, transmembrane pressure (Pa); , viscosity

of the permeate (Pa.s); Rt, total resistance for filtration (l/m).
FIGURE 4. Resistance in series model. Rm, membrane resistance; Ref, external fouling
resistance; Rp, polarization layer resistance; Rif, internal fouling resistance.
16
As depicted in Figure 4, total resistance is a function of the intrinsic membrane
resistance (Rm), the polarization layer resistance caused by the concentration
gradient (Rp), external fouling resistance formed by deposited cake layer from
physicochemical interactions of solids with the membrane (Ref), and internal
resistance due to materials absorbed into the pores (Rif). Equation 1 becomes:
P
J= (2)
(
Rm + RP + Ref + Rif )
Because it is difficult to clearly discriminate between Rp and Ref, these two
terms are combined into a single term, external resistance (Re). Thus
P
J= (3)
(
Rm + Re + Rif )
Pressure applied for filtration influences cake compressibility and thus the
resistance caused by the cake layer Re. Therefore, Re can be written as a function
of transmembrane pressure (Re = P). Where is a function of the cake layers
mass transfer properties. Equation 3 becomes:
P
J= (4)
(
Rm + Rif + P )
For a given fluid, the permeate flux is a function of transmembrane pressure,
and Equation 4 shows two distinct pressure-dependent (at low pressure) and
pressure-independent (at increased pressure) regimes. In the pressure dependent
part, permeate flux is more or less proportional to applied pressure. In the pressure-
independent zone, permeate flux is mainly dictated by cake layer resistance (Fig-
ure 5). In continuous operation, permeate flux at low transmembrane pressure is
higher than that at higher transmembrane pressure. This seems to indicate that the
specific resistance of the solids boundary layer is a strong function of the applied
suction pressure, and an increase in resistance to filtration more than offsets
increased driving force (Bentez et al., 1995; Parameshwaran et al., 1998). This
emphasizes that filtration should be carried out at a low transmembrane pressure.
B. Crossflow Velocity
As shown in Figure 5, increasing the transmembrane pressure above the

pressure-independent part is useless because the flux does not increase any further.
17
FIGURE 5. Schematic representation of flux vs. transmembrane pressure.
However, in this region flux is mainly dictated by the cake layer formation and any
force that disturbs it can influence the flux rate. By increasing the crossflow
velocity (in external circuit membranes) or agitation around the membranes (in
submerged membranes), the cake layer-forming materials can be swept away. A
linear relationship between the flux and the crossflow velocity at constant TMP for
the filtration of biomass of an anaerobic digester treating brewery effluent has been
reported (Strohwald and Ross, 1992). With increased crossflow velocity, an im-
provement in flux and shift in the pressure-independent zone (Figure 6) was also
reported (Magara and Itoh, 1991). However, a study by Ghyoot and Verstruete
(1998) with anaerobically digested sludge indicated that an increase in crossflow
velocity had only a minor affect on the permeate flux.
Shear caused by high crossflow velocity may lead to floc rupture and higher
dispersion of biomass. Therefore, enhanced organic and oxygen mass transfer
within the biomass can help increase process efficiency. Improved efficiency of an
anaerobic digester was reported by Pillay et al. (1994). In this study, the solid
concentration in the digester was increased from 2.6 to 5.5%, and the HRT
decreased from 26 to 14 h, while the SRT was maintained at 26 days. Observations
reported by other researchers contradict this observation (Ghyoot and Verstruete,
1998; Brockman and Seyfried, 1993). According to these investigations, the excess
mechanical stress caused during crossflow damaged the interaction between differ-
ent species in the anaerobic consortia, particularly the symbiosis in the acidogenic
and methanogenic stages. However, these authors agree that the use of a centrifu-
gal pump for crossflow would have destroyed the sludge structure faster than any
other kind of pump. It can be concluded from these observations that although
18
FIGURE 6. Relationship between permeate flux and driving pressure at various
velocities.
crossflow can help improve process performance, equipment selection is an impor-

tant factor in alleviating any negative impact on the process.
C. Effects of Aeration on Flux
Although aeration basically serves the purpose of providing the air required for
biodegradation and keeping the biomass dispersed throughout the reactor, it serves
another important role in submerged MBRs. Turbulence induced by aeration
creates crossflow velocity in the vicinity of the membrane module. Therefore, it
can be anticipated that by augmenting the air flow rate or increasing aeration
intensity (airflow rate per unit area) by concentrating the membrane modules over
a smaller floor area, cake removal efficiency can be improved and with it the flux.
However, Udeda et al. (1997) found that although an increase in the air flow rate
partially stimulated cake removal efficiency, there was a critical value beyond
which an increase in air flow rate had virtually no effect.
19
D. Membrane Fouling
In all practical membrane filtration applications, as the resistance increases the

flux will decline. This increase in resistance may be due to changes in Rm, Rif, Re
or all three (Equation 4). If the flux decline is not reversible by simply altering
operating conditions, it is termed fouling (Fane et al., 1989). This is a major
problem in membrane filtration because it reduces productivity, shortens mem-
brane life (often due to the need for aggressive cleaning agents), and impairs the
fractionation capability of the membrane. Membrane fouling can result from the
precipitation of less soluble inorganic species (scaling), adsorption of organic
substances (organic fouling), and adhesion and growth of microbial cells at the
membrane surface (bio fouling). Studies with anaerobic sludge filtration show
almost no cake layer observed on the surface of the inorganic membrane, but
inorganic fouling inside the pores (Rif = 0.9 Rt) plays a key role in limiting flux
(Kang et al., 1996). For organic membranes, a 13-m cake layer formed due to its
much rougher surface structure. Howver, the amount of inorganic foulant depos-
ited was one-fourth that observed in a ceramic membrane (Kang et al., 1996).
Because MBRs use organic membranes, biofouling is a major concern.
Biofouling includes (Winfield, 1979):
The adsorption of macro molecules, which lead to a conditioning film on the

membrane (humic substances, lipopolysaccharides, and other products of
microbial turnover)
Primary adhesion by fast-adhering cells from the micro flora of the mixed
liquor
colonization and growth of bacteria with subsequent adhesion of a number
of different species, exertion of extracellular polymers (slime), and the
development of a biofilm. The application of continuous high filtration
force (TMP) may protect the cake layer against hydraulic forces.
To understand the early phase of biofouling, consider a system of three

components: microorganism (species, composition of mixed population, hydro-
phobicity, surface charge, etc.), membrane surface (chemical composition, surface
charge, hydrophobicity, roughness, porosity, pore size, etc.), and mixed liquor
(suspended matter and colloids, viscosity, pressure, shear forces, boundary layer,
flux rate, etc.).
Typical methods used to reduce fouling effects include control over operating
conditions (low pressure, high turbulence, and intermittent filtration), backwash
(with permeate or air or both), and chemical cleaning. However, when fouling
becomes too extensive and membrane productivity is lowered beyond compro-
mise, the membrane must be replaced.
Biomass filtration with an evenly distributed pore on an isoporous and highly
porous, hydrophilic smooth membrane surface operated at low to modest pressure
20
can reduce fouling (Fane et al., 1989). It has also been suggested that the membrane
should be negatively charged or neutral to limit biomass adsorption (Shimuzu et
al., 1989).
E. Mode of Operation
Flux decline can be reduced by various means of operating control. Instead of

continuously filtering, intermittent filtration can reduce the compression of the
cake layer, thus resistance is reduced and better flux maintained (Yamamoto et al.,
1989; Cheiemchaisri et al., 1992). In certain cases, permeate backwashing has also
been advocated (Trouve et al., 1994 a). Extensive studies on intermittent operation
with air backwashing indicate improved flux rate compared with continuous
operation (Visvanathan et al., 1997; Parameshwaran, 1997). The results of inter-
mittent filtration with air backwashing are shown in Figure 7.
F. Module Arrangement
Module arrangement influences turbulence at the membrane surface. Due to

their high packing density, hollow fiber membranes are most likely to be used in
submerged membrane bioreactors. In hollow fiber membranes, the main reason for
FIGURE 7. Comparison of net cumulative permeate volume (after 8 h) in different modes

of operation (15* = 15 min without air diffusion).
21
flux decline is the accumulation of particles in the spaces between fibers. Solid
accumulation under different packing density of fibers indicates that there are four
conditions: (1) totally dispersed, (2) partially accumulated, (3) completely accu-
mulated, and (4) surface clogging. Particle accumulation between fiber spaces can
be avoided if a packing density lower than a critical value is used (Kiat et al., 1992).
G. Viscosity
Viscosity of permeate and mixed liquor affects permeation flux. An increase

in permeate viscosity directly affects the filtration flux, as shown in Equation 4.
Because the permeate is essentially water, flux is mainly affected by operating
temperature. However, sludge viscosity also indirectly influences flux (Praderie,
1996). The degree of turbulence (indexed by the Reynolds number) in the vicinity
of the membrane surface and the velocity gradient along the membrane surface
during crossflow can be affected by sludge viscosity, which is in turn affected by
concentration. Krauth and Staab (1994) suggest an exponential dependence be-
tween sludge concentration and viscosity for aerobic sludge. Similar results for
anaerobic sludge have also been reported (Ross et al., 1990).
Magara and Itoh (1991) found that a semilogarithmic relationship can be used
to describe the influence of biomass concentration on the limiting flux.
J = 1.571 log(MLSS) + 7.84 (5)
where J, permeate flux (m/d); MLSS mixed liquor suspended solid concentration
(mg/l).
The relationship in Equation 5 was developed for a concentration ranging
between 5 and 15 g/l. In contrast is an observation by Ross et al. (1990) on
anaerobic sludge filtered with ultrafiltration. They found that flux was relatively
stable up to a concentration of 40 g/l, but then decreased sharply and stabilized at
about 60 g/l.
V. APPLICATION OF MEMBRANE SEPARATION BIOREACTORS IN

AEROBIC WASTEWATER TREATMENT
The activated sludge process was invented in 1914 by Edward Arden and
William Lockett of Manchester Corporation in England (Arden and Lockett, 1914)
and is now the most commonly used biological wastewater treatment system.
Capture of biological solids is essential to achieve low-effluent BOD and to control
the accumulation of biomass. Significant improvements in solids separation have
had a noticeable impact on the efficiency and reliability of biological processes.
For nearly 4 decades researchers have studied various aspects of membrane-
22
coupled bioreactors. Yamamoto et al. (1989) were the first to introduce submerged
membranes in an aeration tank for solid/liquid separation. Prior to this, researchers
concentrated on crossflow membrane filtration in external circuits. These early
studies (e.g., Smith et al., 1969) used ultrafiltration membranes in the external
membrane circuit type and, although there are still studies with ultrafilters (e.g.,
Muller et al., 1995; Chaize and Huyard, 1991), the present trend is toward
microfiltration.
A. Bioreactors with Membrane in External Circuit
A pilot plant with a completely mixed biological reactor (4.5 l) connected to

an ultrafiltration module was used by Chaize and Huyard (1991) to treat domestic
wastewater. Air was supplied to the reactor at a rate of 80 l/h and the mixer
operated at 800 to 1200 rpm. The ultrafiltration membrane module (molecular
weight cut off 50,000 D) was operated with a cross-flow velocity of 1.5 m/s and
1 to 2 bar transmembrane pressure. Concentrate from the membrane module was
returned to the reactor.
Experiments were carried out in two runs. The first run was carried out for
160 days with an HRT of 8 h and an SRT of 100 days. After 25 days of operation,
biomass concentration reached a steady state with MLSS concentration varying
between 8000 and 10,000 mg/l. The effluent COD was reported to be less than
30 mg/l, while the influent COD was in the range of 250 to 550 mg/l. The effluent
TKN was less then 10 mg/l, while the influent TKN was in the range of 65 to
150 mg/l. This nitrification phenomenon was observed 14 days after start up. The
F/M ratio varied between 0.06 and 0.1 kg COD/kg MLSS.d.
For the second run, the HRT was varied 8, 4, and 2 h, with corresponding SRTs
of 100, 100, and 50 days in order to study the steady state of biomass concentration
under different operating conditions. Over 200 days of operation, the biomass
concentration increased slowly up to 30 days after which membrane permeability
decreased. Changes in operating conditions (HRT and SRT) had no effect on
carbonaceous removal. However, at the start of the operation, TOC accumulated
in the bioreactor. As the biomass concentration increased, the TOC decreased;
however, accumulation of TOC again increased when HRT was reduced. This
could be due to the accumulation of either bacterial products or raw water compo-
nents. Under different operating conditions, this phenomenon was observed by
several other researchers (refer to Table 7). Although changes in operating condi-
tions disturbed the removal of nitrogen compounds, after a period of adaptation the
removal was maximum for each HRT studied.
Studies by other researchers on the operating conditions and performance
results for membrane coupled bioreactors with the membrane in an external circuit
are presented in Table 7.
23
24
B. Bioreactors with Submerged Membranes
A series of laboratory scale experiments were carried out by Yamamoto et al.

(1989) to determine the feasibility of direct membrane separation in an activated
sludge aeration tank. The study was carried out with a 0.1-m-pore-size hollow
fiber membrane. Short-term experiments with constant MLSS concentration re-
vealed high suction pressure (80 kPa) led to rapid reduction in flux. During long-
term experiments, continuous suction caused severe clogging of the membrane
module with an increasing pressure difference up to 100 kPa. A stable flux was
observed for 120 days at a volumetric loading of 1.5 kg COD/m3.d using intermit-
tent suction at a low pressure of 13 kPa. COD removal of more than 90% was
reported despite the intermittent aeration. However, nitrate removal varied consid-
erably above 80% and the denitrification efficiency ranged from 20 to 60%.
Intermittent aeration did not change the denitrification efficiency, indicating that
the dissolved oxygen could not be depleted in such a short nonaeration time. From
the analysis of the reactor supernatant and effluent, it was found that the membrane
acts as a separator removing a certain amount of dissolved and colloidal COD.
During steady state operation, the F/M ratio was 0.1 kg COD/kg MLSS.d and the
critical organic loading was estimated at 3 to 4 kg COD/m3.d to maintain both
stable flux and an aerobic condition. Absence of a recirculation pump led to a very
low power consumption of 0.007 kWh/m3.
Chiemchaisri (1990) investigated an activated sludge process with a 0.1-m
hollow fiber membrane module for solid-liquid separation to treat low-strength
domestic wastewater. A comparison was made of the performance of the mem-
brane bioreactor under different operating conditions, such as nonaerated and
aerated, with different initial hydraulic retention times of 1, 3, and 6 h (with
corresponding permeate flux of 4.17, 1.38, and 0.7 L/m2.h). The process was
operated at intermittent extraction, 10:10 min operating time. The magnetic stirrer
was employed in nonaerated condition to keep MLSS in suspension. The nonaerated
bioreactor had an advantage over the aerated condition at initial HRTs of 3 and 6
h, because lower energy was required to give similar effluent quality and process
stability. However, at an HRT of 1 h (or higher permeate flux, 4.17 L/m2.h)
aeration was required to prevent membrane clogging. Flux of 4.17 L/m2.h seems
to be a critical value between severe clogging and nonclogging conditions.
At lower flux, no clogging was observed under nonaerated and aerated condi-
tions. The quality of the permeate in terms of COD was independent of the low
volumetric organic loading in the range of 0.2 to 2 kg COD/m3.d. Because of the
long solid retention time (SRT), the process was stable and steady, and COD
removal efficiency was similar in all experimental conditions.
Performance of the 0.03-m pore size with 9 m2 surface area of hollow fiber
membrane was also investigated in a pilot-scale unit (Chiemchaisri et al., 1993)
Two hollow fiber membrane modules were immersed in an aeration tank that was
fed diurnally with domestic wastewater. A suction pump was used at 10:10 min
intermittent operation to extract the permeate through the membrane. The effect of
25
a jet aeration period (0.5 and 1 h) and a jet aeration pattern (15 min two times a
day and 30 min once a day) was investigated. The jet aeration flow rate was 20 l/
min.
The settling of solids to the bottom of the bioreactor and the creation of
anaerobic conditions resulted in the division of the bioreactor into two zones,
aerobic and anaerobic. This resulted in low MLSS in the aerobic zone that could
have reduced membrane clogging. The mean hydraulic retention time (HRT)was
determined after the permeate flux reached a steady state. The average flux was
around 4.17 l/m2.h, corresponding to an average HRT of 1 day under diurnally
varied loading. The diurnal variation had a minor effect on the nitrification process
because more than 80% nitrification was observed throughout the experiment.
The MLSS in the bioreactor was affected by the air flow rate. Optimum air
flow rate in this experiment was taken as 7.5 l/min, which provided sufficient
oxygen for the microorganisms and maintained low MLSS in the aerobic zone.
Direct membrane separation using hollow fiber membranes in an activated
sludge process was investigated on a pilot scale by Chiemchaisri et al. (1992). The
experimental set up used for this study is shown in Figure 8. The system consists
of two parts, the main bioreactor and a separation unit. A separation unit of 10 l
volume was immersed in the main bioreactor, which had a 62-l volume. Two
hollow fiber membrane modules 0.03 and 0.1 m pore size) of 0.3 m2 surface area
each were placed in the separation unit. Paddles driven by a motor provided a
cross-flow of mixed liquor across the membrane surface at a speed of 290 rpm in
10-s cycles in alternate directions. By providing highly turbulent conditions within
the separation zone in conjunction with jet aeration inside the membrane module,
sludge accumulation on the membrane surface and inside the module was reduced.
The permeate flux obtained after 330 days of operation was 8.33 l/m2.h (0.2 m3/
m2d) under intermittent suction. A high degree of organic matter reduction (> 85%)
was observed with 20.8 and 16.5 mg/l of COD in the effluent during continuous
and intermittent aeration modes, respectively. The degree of nitrification and
denitrification was above 90% during intermittent aeration (90 min aeration and 90
min rest) at a dissolved oxygen level of 4 to 5 mg/l. However, at similar intervals
of intermittent aeration at a low dissolved oxygen level (1.5 to 2 mg/l), a reduction
in nitrification and denitrification efficiency (80%) resulted in 4.9 mg/l of total
nitrogen in the effluent. In addition, a virus reduction of 4 to 6 log number was
reported.
Buissson et al. (1998) reported on the concept of immersed membranes for
upgrading wastewater treatment plants. The pilot plant consisted of two reactors in
series, the first one under anoxic conditions and the second one with aeration. In
this denitrification-nitrification mode, the unit was fed with municipal wastewater.
The hollow fiber modules were directly immersed in this aerated tank. Aerators
were positioned at the bottom of the membrane modules to provide the required
amount of air to promote turbulence and increase mass transfer in the vicinity of
the membranes. The membranes were backwashed with treated water at regular
intervals with permeate to remove the filter cake. The researchers reported a sludge
26
FIGURE 8. Schematic of experimental system. (From Chiemchaisri, 1990.)
27
concentration of 15 g MLSS/l and a volumetric loading of 1.2 kg COD/m3/d with
96% COD and 95% total Kjeldahl nitrogen (TKN) removal.
In parallel with the development of hollow fiber submerged membrane bioreactor
systems, Japanese researchers developed Kuboto submerged flat plate membrane
systems. This process has been tested on a pilot scale in the U.K. (Churchouse,
1998; McCann, 1998). In this system (shown in Figure 9), flat plate microfiltration
membrane panels housed within a rectangular box are submerged in an activated
sludge tank and a coarse bubble aeration system is placed at the bottom of
the membrane module. The aerated sludge that rises between the panels provides
sufficient recirculation of sludge at the membrane surface. It also generates an
upward cross-flow over the membrane surface and keeps surface fouling to a
minimum. The permeate from the system is obtained either by a low-pressure
(0.1 bar) suction pump or by gravity. The module arrangement and the aeration
system require very little membrane cleaning. Using a mechanical water jet, the
modules are washed once a year. In addition, this system needs in situ cleaning two
to four times per year for flux improvement. The system was tested in degritted
sewage with a denitrification and nitrification arrangement with wide variation in
effluent quality (30 to 2100 mg/l BOD and 100 to 4000 mg/l COD) (Churchouse,
1998). The final effluent from this process consistently met a 5:1:1 (BOD:SS: NH3-
N) standard with 96% removal achieved of COD and BOD. The average sludge
production was 0.3 kg/kg BOD, which is approximately 40% of the normal
activated sludge process production rate.
A comparison of operating conditions and corresponding performances of
several submerged membrane bioreactors is presented in Table 8.
FIGURE 9. Schematic of the Kubota membrane unit operation and panel construction.
28
29
C. Membrane as Air Diffuser and Clarifier
An alternative method to improve performance of membrane bioreactor sys-

tems uses the membrane in alternate cycles to filter and aerate with both external
circuit and submerged membranes in the bioreactor.
Scott and Smith (1997) studied the alternate functioning of aeration and
separation with a microfiltration ceramic membrane (0.2 m) in an external circuit
of a membrane coupled bioreactor for remediating food process industry wastewa-
ter (Figure 10). It was concluded that a 0.2-m membrane provided markedly
superior fine bubble aeration to traditional aerators and kept permeate suspended
solids below 50 mg/l, despite an input of 1500 to 3800 mg/l. At 25C, influent
COD loading of 13,330 mg/l and BOD of 6500 mg/l, removal efficiencies in excess
of 95% were obtained. In addition, ammonia nitrogen was reduced from 9 to 10.2
mg/l to < 0.3 mg/l and total phosphorous from 20 to 21.2 mg/l to < 4 mg/l. This
study suggested the need to determine optimum air supply rate and bioreactor
HRT. Furthermore, although successful runs were conducted without nutrient
addition, nutrient limitation appears likely in most industrial wastes. Nutrient
FIGURE 10. Schematic of ceramic membrane-based bioreactor.
30
addition therefore may provide further enhancement of degradation rates, although
any improvement would have to be offset against the costs of additives.
Visvanathan et al. (1997) used a 0.1 m hollow fiber membrane module
directly in the reactor for solid-liquid separation of domestic wastewater with
permeate flow outside-to-inside arrangement. During short-term experiments, the
effect of transmembrane pressures, intermittent mode of operation and duration of
air diffusion were investigated to determine the optimum conditions corresponding
to a high and constant flux obtained. Variations in transmembrane pressure were
studied with values of 13.3, 21.3, 32.0, and 41.0 kPa. It was found that a transmem-
brane pressure of 13 kPa was the limiting pressure for all experiments. The
different operating modes studied by varying duration of effluent filtration and air
diffusion through the membrane were 5:5, 10:10, 15:15, 30:30, 60:60, and 15:15*
(15* = 15 min without air).
The results show that an operational mode of 15:15 min provides the best
results. Although cyclic operation with air diffusion could not completely elimi-
nate clogging, the air diffusion through the membrane backwash technique in this
mode of experiment could improve the flux by up to 370% compared with a
continuous operation.
These experimental results also indicated that the transmembrane pressure
increased according to cake formation on the membrane. Steep increases in trans-
membrane pressure were observed even under air diffusion conditions. Periodic
chemical cleaning was needed in order to recover the permeate flux, which
exhibited good quality in terms of very low SS and BOD.
During the experimental runs the sludge was not removed from the reactor, and
more than 90% reduction in COD with influent concentration of 200 to 300 mg/l
and effluent concentration below 20 mg/l was achieved in all runs. The TKN
removal was more than 90%, and total phosphate removal was approximately 50%
in all runs. The MLVSS/MLSS ratio in the bioreactor was in the order of 20 to
30%. Solid matter mass balance calculation indicated a steady accumulation of
inorganic components within the reactor. The lower fraction of active microorgan-
isms in the bioreactor did not show any significant effect on the process efficiency.
Nevertheless, it is anticipated that in longer runs it might, and periodic sludge
draining therefore would be advisable.
The possibility of using microfiltration hollow fiber membrane modules as air
diffusers and solid/liquid separators in alternating cycles for activated sludge
process treating domestic wastewater was reported by Parameshwaran et al. (1998).
The activated sludge system consisted of an anoxic and oxic zone for better
nitrogen removal. Two hollow fiber microfiltration modules with a pore size of
0.2 m were immersed in an 80-l oxic tank (MBR) to effect the direct solid/liquid
separation (Figure 11). Filtration and high-pressure air backwashing were em-
ployed in alternating cycles to improve the flux rate. Backwashing the membrane
with air was used to aerate the activated sludge in the reactor. This achieved
distinct advantages of membrane declogging and aerating reactor contents simul-
taneously.
31
FIGURE 11. Membrane as air diffuser and solid/liquid separator.
In short-term experimental runs, it was found that membrane module are better
air diffusers than stone air diffusers. It was also found that increased backwash air
pressure leads to improved flux rate.
The efficiency of membrane coupled bioreactor systems at different hydraulic
retention times of 15, 10, 6, and 3 h was studied in long-term experiments.
Wastewater was fed to the anoxic tank that passed into the MBR. The contents of
the MBR were recycled to the anoxic tank to effect denitrification. The average
MLSS concentration in the system varied between 12,000 to 14,000 mg/l, and the
sludge age in the MBR was maintained at 50 days throughout the study. The
desired HRT could be maintained at 15, 10, and 6 h with a moderate transmem-
brane pressure (< 42 kPa), but not at 3 h.
Irrespective of the operating conditions, in all experiments COD, BOD, TKN
and total nitrogen removal of more than 95, 98, 95, and 80%, respectively, was
achieved. This study established that using a hollow fiber membrane capable of air
backwashing to solid/liquid separation leads to aeration of mixed liquor and
declogging of membrane modules simultaneously and that conventional aerators
could be eliminated. By using anoxic/oxic systems, efficient total nitrogen removal
can also be achieved.
VI. APPLICATION OF MEMBRANE BIOREACTORS IN ANAEROBIC

WASTEWATER TREATMENT
Since its introduction more than 4 decades ago, anaerobic treatment has
become a well-established process for concentrated industrial wastewater and
32
municipal sludge management. In recent years, there has been a better understand-
ing of the microbiology of this process and improved reactor design has made it
possible to consider them for treatment of dilute low-strength wastewater.
The major difficulty in anaerobic wastewater treatment processes is the reten-
tion of a sufficient quantity of an active biomass due to their slow net growth rates.
As a result, anaerobic systems require a longer minimum solids retention time
(SRT). Operating below this minimum SRT would result in microorganisms being
washed out of the system at a faster rate than the growth rate leading to system
failure. Anaerobic treatment is not popular due to its low efficiency and inability
to meet discharge regulations despite the advantages of generating methane fuel
gas and low sludge yield. Also, these systems are unable to cope with shock
loading and wide fluctuations in influent flow.
The loading rates in anaerobic wastewater treatment systems are mostly dic-
tated by the biomass retention in the reactor. High biomass retention will give good
reactor performance, leading to better gas yields and better-quality effluent. A
lower retention capability will lead to a longer hydraulic retention time (HRT), thus
requiring reactors of larger volume and higher cost. These problems could be
overcome if the biomass in the reactor can be retained longer than the minimum
SRT, thus increasing biomass concentration.
For dispersed growth anaerobic systems, a high concentration of biomass
retention was achieved through improved separator performance by the use of
chemical coagulants and modifications in processes and component design. Apart
from this, various anaerobic reactor configurations, such as anaerobic filters,
fluidized bed reactors, upflow anaerobic sludge blanket reactors (UASB), rotating
contactors, etc. are advocated with a view to longer sludge retention times and
shorter hydraulic retention times. Although all these systems needed less space
compared with the suspended growth anaerobic systems. Except UASB all other
systems use part of their space to accommodate the media on which biomass
attached. However, the problems associated with UASBs are related to the require-
ment for a suitable seed sludge that can be granulized and maintenance of an
appropriate organic loading rate and close control of environmental conditions. In
this situation, the application of membrane separation for anaerobic wastewater has
some distinct advantages:
Increasing the biomass concentration without increasing the reactor volume

irrespective of the granulation of sludge
Colloidal/suspended solid-free, good-quality effluent. A number of mem-
brane anaerobic systems were tested mostly with industrial wastewater to
determine their performance
Improved treatment efficiency that can easily meet standards since colloidal
matters and macromolecules are retained in the bioreactors by the mem-
branes and completely degraded
Intermediate toxin removal at a faster rate to prevent system failure, because
all the microorganisms, including slow-growing methonogens, are retained
33
Some studies have indicated that a higher methane production rate was
obtained in anaerobic MBR
A. Treatment of Domestic Wastewater in Membrane Coupled

Anaerobic Process
Grethlein (1978) demonstrated how a septic tank-membrane system is feasible

for the treatment of domestic wastewater. The experimental set up used in this
study is shown in Figure 12. A two-compartment rectangular tank with a total
volume of 106 l was constructed, and a minimum flow rate of 2.65 l/min was
applied. Operating pressure was in the range of 345 to 1030 kPa (50 to 150 psi).
By using the surge capacity of the septic tank, the average volumetric flow rate of
the inlet and outlet were matched. Two types of membrane module were used (a
flat sheet membrane module and a helicore reverse osmosis unit). In order to
control concentration polarization at the membrane surface and to mix the contents
of the septic tank, the circulation flow (cross-flow) over the membrane surface was
kept much higher than the permeate rate.
It appears that with a 2:2 operating cycle (2 min ON, 2 min OFF), it was
possible to maintain a steady flux of about 20.42 l/m2h with a cross-flow velocity
as low as 22.9 cm/s. During the off period of the cycle (during the off period, feed
side pressure is released by opening the bypass valve) when backflow occurred,
some noticeable solids were floated off, but not enough to clean the entire mem-
brane surface. The spontaneous break-up of the gel layer, which results in cleaning
FIGURE 12. Experimental setup for septic tank effluent treatment.
34
the surface of the membrane, is a self-cleaning method unique to the cyclic
operation procedure, and in the long run is probably the most important mechanism
for achieving long-term practical flux levels. The water quality of the treated
effluent is similar, if not better, than that of secondary treatment effluent. For
example, E.coli and turbidity are below the detection limit, while BOD is reduced
by 85 to 95% for an influent concentration of 270 mg/l. A particular feature of this
system is its ability to reduce the initial feed nitrate concentration of 3.5 mg/l by
about 75%. The anaerobic rate of digestion of organic carbon in the septic tank is
enhanced by a factor of 3 to 4 because of the increased concentration of microor-
ganisms and substrate caused by membrane. The pH stability of the digester is
excellent (6.5 to 7.2), even with intermittent loading. The sludge accumulation was
less than in an ordinary septic tank.
B. Treatment of Industrial Wastewater in Membrane Coupled

Anaerobic Process
A comprehensive study using ADUF( (anaerobic digestion-ultrafiltration) pro-

cesses for the treatment of different industrial wastewater was carried out in South
Africa (Ross and Strohwald, 1994). In this process, a 9-mm tubular polyethersulfone
membrane at a pressure of 500 kPa was used. The ADUF process consists of two
main unit operations: an anaerobic digester coupled with an external pressure
driven ultrafiltration unit. Some of the studies carried out with this unit are
described below.
1. Wine Distillery Wastewater
A pilot-scale anaerobic digester with an external ultrafiltration membrane to

treat wine distillery wastewater (37,000 mg COD/l) was reported by Ross et al.
(1990). A pilot plant with a digester and external UF unit is shown in Figure 13.
The 2.4 m3 digester operated at a MLSS concentration of 30,000 mg/l, and prior
to the installation of the UF module could only be fed with wine distillery waste
at a volumetric loading rate of some 4 kg COD/m3.d at 35C. A UF module with
a total membrane area of 1.75 m2 was used for the test at an inlet pressure of 400
kPa. The high rate of sludge recirculation through the UF unit to comply with a
cross-flow velocity of 2 m/s resulted in a permeate volume (2.4 m3/d) well in
excess of that of the substrate feed rate (0.3 m3/d) to the digester. The oversized
UF unit was accommodated by wasting a volume of permeate equivalent to the
daily feed volume, and by recycling the excess permeate back to the digester. All
the sludge concentrate was returned to the digester. The experimental results
obtained on the ADUF system include the following:
35
An extremely high initial permeation flux of 62.5 l/m2.h at 400 kPa inlet
pressure that gradually decreased to 37.5 l/m2.h after 7 months continuous
operation. Temporary substitution of the original module by a new one also
gave a flux of 37.5 l/m2.h, indicating that the flux decline was not caused
by membrane fouling, but by changes in the digester contents, for example,
the SS in the digester increased over the 7 months to 50 kg/m3 from an initial
concentration of 30 kg/m3
Building up of biomass occurred in the digester, with a concomitant in-
crease in the permissible volumetric loading rate notwithstanding the poor
settleability of the sludge. During the study period, the load rate increased
from 4 to 12 kg COD/m3.d
The operating flux was successfully maintained for a period of several
weeks before membrane cleaning was necessary
The degree of COD removal was 93%, based on a feed and effluent
concentration of 37.0 and 2.6 kg/m3, respectively.
Table 9 presents the information on principle operating conditions and results

of various reported pilot- and full-scale ADUF plants.
Other studies have looked at the possibility of using membrane coupled
anaerobic digesters for industrial wastewater treatment. Ultrafiltration membrane
separation for anaerobic industrial wastewater treatment was studied by Fakhrul-
Razi (1994). The study was conducted at a temperature of about 35C, a pressure
range of 100 to 300 kPa and with minimal pH control (the pH ranges from 6.8 to
7.4 throughout the study). The reactor was subject to different organic loading rates
FIGURE 13. Diagram of pilot-scale ADUF process applied for anaerobic treatment.
36
TABLE 9
Mean Operating Criteria of ADUF Plants Treating Various Industrial Effluents
Wine Egg Maize

Parameters Unit Brewery distillery Malting process process
Volume of digester m3 0.05 2.4 3.0 80 2610

Operational period month 3 18 5 8 36
Feed COD kg/l 6.7 37 3.5 8 415
Permeate COD kg/l 0.18 0.26 0.8 0.35 0.3
COD removal % 97 93 77 95 97
Space load rates kg COD / m3.d 17.0 12.0 5.0 6.0 3.0
Sludge load rate kg COD/kg VSS.d 0.7 0.58 0.5 0.33 0.24
HRT day 0.8 3.3 0.8 1.3 5.2
Temperature C 35 35 35 30 35
MLSS kg/m3 3050 50 10 1030 23
37
Membrane area m2 0.44 1.75 9.6 200 800
Flux L/m2.h 1040 4080 2040 1530 1070
Inlet pressure kPa 340 400 500 500 600
Crossflow velocity m/s 1.5 2.0 1.8 1.8 1.6
Tube diameter mm 9.0 12.7 9.0 12.7 9.0

From Ross and Strohwald, 1994.
(OLR) ranging from 12 to 20 kg COD/m3.d. The SRT was achieved by deliberately
wasting reactor sludge from the system. The SRT range was between 83 to 53 days.
Six steady states were attained over a MLSS range of about 31,000 to 38,000 mg/l.
It was found that rate of gas production in the reactor increased with increasing
OLR. The methane content of the reactor gas declined from 65.3 to 57.1%.
Methane yield ranged from 0.26 to 0. 29 m3 CH4/kg COD. However, no sudden
increase in volatile acids concentration were recorded during the study, indicating
that the anaerobic system coped quite well with the increasing loading rates. The
reduced methane content could be due to the much higher influent COD, resulting
in a higher rate of carbon dioxide formation by the acid-forming bacteria in the
reactor.
The treatment efficiency of the membrane anaerobic system showed a consis-
tent COD removal of above 96%. The highest influent COD applied to the system
was 84,010 mg/l, and the corresponding effluent COD for this was 3100 mg/l,
which represents a COD removal of 96.3%. It was further observed that MLSS and
MLVSS increased with increasing OLR.
The rise in MLVSS concentration indicates that the bacterial population,
including methanogenic bacteria, increased with the higher influent COD applied.
Further increases in the OLR were possible in order to achieve further bacterial
growth leading to a much higher MLVSS concentration in the reactor. During the
experimental study, the SRT was reduced from 83.3 to 58.8 days due to the
increase in the deliberate sludge wastage rate from the reactor. This was necessary
in order to offset the drop in membrane flux from the UF module over the period
of the study. The results indicate that a membrane anaerobic treatment system is
suitable when higher biomass retention and its treatment efficiency is expected.
The study showed that the incorporation of membranes in a system enable it to
retain active bacterial populations and produce a clear effluent as final permeate.
The system also showed that it was capable of higher loading rates and has yet to
achieve its maximum treatment capability. This was a result of good control of
bacterial populations in the reactor provided by the UF membranes. Throughout
the study there was negligible biomass loss through the effluent/permeate.
In an experiment by Harada et al. (1994), the permeate flux decreased signifi-
cantly as the cultivation time elapsed. This may have been brought about by the
formation of a gel layer on the membrane surface. This phenomenon was caused
by insolubilization of high molecular organic substances. In addition, the increase
in MLSS enhanced the deposition of cells to the gel layer matrix and accelerated
the deterioration of the flux. Although washing with water eliminated some of the
gel layer, the permeate flux was not restored to 100% of the initial level after each
wash. In the membrane bioreactor the major portion of the organics present were
substances with molecular weights (MW) as large as 106, while the permeate
consisted of substances with MWs less than 1500. This large difference in molecu-
lar size between the concentrate and the permeate implies that the membrane was
38
capable of not only separating suspended or colloidal matters but also fractionating
soluble organics by molecular size. The exclusion of higher molecular size mate-
rials from the permeate is attributable to the formation of gel layers on the
membrane surface. The formation of a gel layer is responsible for the decline in
permeate flux, although it is useful for production of high-quality effluent.
A comparison of different studies with anaerobic bioreactors followed by
membrane separation is presented in Table 10.
VII. ECONOMICS OF THE MBR PROCESS
The success of any process is ultimately determined by its economic benefits

and higher investment and operating costs, which hinder the progress of MBR
processes. However, cost of potable water and the disposal cost of wastewater
resulting from increasingly stringent legislation has led to the full-scale application
of MBRs in Japan (Aya, 1994). In the MBR process, the cost of energy needed to
maintain the cross-flow velocity and the filtration pressure is almost 90% of the
total operating cost. However, using submerged membranes reduces the pumping
energy requirement to 0.007 kwh/m3 of permeate (Yamamoto et al., 1989) com-
pared with more than 3 kwh/m3 permeate required for the cross-flow mode. Table
11 presents a cost comparison of submerged MBR and conventional activated
sludge processes (ASP). Based on this comparison for a small-scale wastewater
treatment process, MBR systems seem to be more attractive than an ASP in terms
of land and space requirements and energy consumption.
Table 12 presents three different types of membrane bioreactors currently
operating on municipal wastewater at semiindustrial scale plants. Here, the Renovexx
process operates on settled sewage with a tabular membrane module in an external
arrangement. Memcore process operates on coagulated settled sewage whereas the
Kubota process works on screened sewage with a submerged membrane module
system as described in Section VI.
It is interesting to note that these three different MBR designs and membrane
element configurations lead to the same order of magnitude of operational cost per
cubic meter of treated wastewater, whereas both membrane cost and surface area
requirement per cubic meter of treated wastewater is relatively large for the Kubota
system.
A. Cost Aspects of ADUF Process
Strohwald (1994), cited in Ross and Strohwald (1994), carried out a capital and
operating cost estimate for a 1500 m3/d ADUF system with energy recovery from
the generated biogas. The design incorporated a total digester volume of 1800 m3
39
Operating Conditions and Performances of Anaerobic MBR Process
TABLE 10
40
TABLE 11
Comparison of Membrane Bioreactor and Conventional Activated Sludge
Process
MBR (submerged type) Activated sludge process
Plant area (m2) Flow control tank 13.4 Flow control tank 13.4
ASA tanka 20.0 ASA tankb 66.7
Sedimentation tank 66.7
Presediment. tank 5.0
Sedimentation tank 10.0
Sludge 13.5
concentration tank
Total 33.4 Total 100.3

Electric power (kw) Fine screen 0.10 Fine screen 0.10
Flow control 0.25 Flow control pump 0.25
Flow control blower 0.40 Flow control blower 0.40
Blower for aeration 3.70 Blower for aeration 5.50
Suction pump 0.20
Total elect. power 4.65 Total 6.25

Sludge (m3/day) 0.069 0.963
Running costc ($/day) 8.37 11.25
Sludge treatmentc ($/day) 34.65 48.3
Running cost 72% 100%
Space 30% 100%
a Activated sludge aeration tank (load 2 kg/m3.day).

b Activated sludge aeration tank (load 0.6 kg/m3.day).
c Price for electricity at US$ .075 (Exchange rate of 40B/$).
From MRC, 1997.
and total MEMTUF membrane area of 1400 m2. The specific cost calculations are
summarized in Table 13, assuming a 5-year depreciation and interest at 20%, based
on a total capital cost for the system of 385,000. The redemption of capital is the
largest cost contributing factor expected. With respect to the operating cost break-
down, a conservative 2-year membrane life was assumed for membrane replace-
ment. The combined specific capital and operating cost of 0.04 per m3 effluent
treated for a complete ADUF plant compares favorably with the cost for direct
disposal of untreated effluent to municipal treatment works. The added benefit of
water conservation by possible permeate increases the economic attractiveness of
on site ADUF treatment.
41
TABLE 12
Comparison of Operational Variables and Cost Parameters for Three Different
Membrane Bioreactor Systems
Parameters Kubota Renovexx Memcor
Location Kingston Seymour, Newton Aycliffe, Aberporth, U.K.

U.K. U.K.
Function TSS, COD/BOD TSS, precipitated TSS/BOD removal
removal, and and coagulated and disinfection
(de)nitrification solids removal
Membrane material Coated fiber Dynamic Al floc Polypropylene
on substrate
Membrane configuration Flat plate Tube HFF
Membrane pore size (m) 0.4 (nominal) 10 (nominal) 0.20.3
Feed water specification Screened raw Screened settled Coagulated settle
sewage sewage sewage
Footprint 6m9m 115 m2
(including tanks)
Transmembrane pressure 0.1 1 (inlet pressure 2) 0.2
(bar)
Flux (l/m2.h) 2732 110140 100
Production rate (m3/day) 125 4000 3750
Operation/regen. Continuous 180/25a 45/2
(cycle, min)
Cleaning
Mechanical Water jet, 1/year Backwash only with Compressed air
water (1% of total at 6 bar
water produced)
Chemical 2/years Every 34 d
Total membrane area 1.92 0.292 0.72
(m2/m3)
Membrane life (years) ~7 1.52 ~5
(w. brush cleaning)
Membrane cost 80 35 3040
(/m2)
Operating cost (/m3)
Replacement
Chemical <10 24 24
Other 5 24
10 10 24
Total <20 1721b 12
Capital cost (/MLD plant) 1.11.4/1 0.5/4 3.75
a Includes membrane formation.

b Includes upstream chemical cost.
From Judd, S. (1998).
42
TABLE 13
Specific Cost Summary of a Full-Scale ADUF Plant
Item Subtotal (/m3) Total (/m3)
Capital cost 0.282 0.282

Investment
Operating cost
Electricity 0.023
Membrane 0.042
Maintenance 0.026
Man power 0.021 0.112
Total cost 0.394
Ross and Strohwald, 1994.
VIII. CONCLUSION
A number of full-scale MBR applications exist mostly in Japan followed by the

U.S., France, the U.K., Germany, and the Netherlands. At present, the use of MBRs
is reserved to situations that require good treated water quality for reuse or when
stringent discharge standards cannot be technically and economically achieved by
conventional effluent treatment systems (Manem and Sanderson, 1996). In the
past, both capital and operational cost of membrane modules seemed to be the
major hindrance to industrial-scale application of this technology. More recently,
due to increased membrane production and improvement of automated systems,
there has been a significant reduction in the cost of membrane systems during the
last few years. Meanwhile, increasing interest in reuse of treated water and the need
for more compact systems coupled with progress in research is expected to lead to
more widespread application of the MBR process. Among those processes, the
submerged configuration seems likely to see the widest application (Mccann,
1998). Recently, a submerged membrane system was used to upgrade existing
conventional processes in two ways: (1) to increase plant capacity and (2) to
improve treated water quality (Buisson et al., 1997). Given these developments and
considering that, in general, membrane costs are decreasing, it can be expected that
the MBR technology will become more economically competitive with conven-
tional processes and find increasingly wider applications in wastewater treatment.
ACKNOWLEDGMENT
The authors thank Terry Clayton, Centre for Language and Education Tech-
nology, Asian Institute of Technology, Thailand, for his kind assistance in review-
ing and editing the final manuscript of this paper.
43
REFERENCES
Ahmed, T. and Semmens, M. J. (1992a). Use of sealed end hollow fibers for bubbleless membrane
aeration: experimental studies. J. Membrane Sci. 69, 110.
Ahmed, T. and Semmens, M. J. (1992b). The use of independently sealed microporous hollow fiber
membranes for oxygenation of water: model development. J. Membrane Sci. 69, 1120.
Anderson, G. K., James, A., Saw, C. B., and Le, M. S. (1984). Crossflow microfiltration. A
membrane process for biomass retention in anaerobic digestion. 4th World Filtration Congress,
Ostend, Belgique, April 11, 1984. 7584.
Ardern, E. and Lockett, W. T. (1914a). The oxidation of sewage without the aid of filters. II. Soc.
Chem. Ind. J. 33(23), 11221124.
Ardern, E. and Lockett, W. T. (1914b). Experiments on the oxidation of sewage without the aid of
filters. Soc. Chem. Ind. J. 33(10), 523539.
Arika, M., Kobayashi, H., and Hihara, H. (1997). Pilot plant test of an activated sludge combined
process for domestic wastewater reclamation. Desalination 48, 299319.
Audic, J. M., Fujita, Y., and Faup, G. M. (1986). Le Coulage Boues Actives-membranes. Une
Ralit au Japon, TSM, June 1986 . 297300.
Aya, H. (1994). Modular membrane for self-contained reuse systems, Water Qual. Int., No. 4.
Bailey, A. D., Hansford,G. S., and Dold (1994). The use of crossflow microfiltration to enhance the
performance of activated sludge reactor. Water Res. 28(2), 297301.
Benefield, L. D. and Randall, C. W. (1980). Biological Process Design for Wastewater Treatment,
Prentice-Hall, New Jersey.
Bentez, J., Rodrguez, A., and Malaver, R. (1995). Stabilization and dewatering of wastewater using
hollow fiber membranes. Water Res. 29(10), 22812286.
Brindle, K. and Stephenson, T. (1996). Mini review the application of membrane boilogical
reators for the treatment of wastewater. Biotechnol. Bioeng., 49, 601610
Brindle, K. and Stephenson, T. (1997). Membrane bioreactors todays effluent treatment process
for tomorrow. UTA Int., 2/97, 108111.
Brockmann, M. and Seyfried, C. F. (1993). Sludge activity and cross-flow microfiltration a non-
beneficial relationship. Water Sci. Technol. 34, 205213.
Buisson, H., Cote, P., Praderie, M., and Paillard, H. (1997). The use of membranes for upgrading
wastewater treatment plants, IAWQ Conference on Upgrading of Water and Wastewater System,
May 2528, 1997, Kalmar.
Chaize, S. and Huyard, A. (1991). Membrane bioreactor on domestic wastewater treatment sludge
production and modeling approach. Water Sci. Technol. 23, 15911600.
Chiemchaisri, C. (1990). Design consideration of membrane bioreactor in domestic wastewater
treatment, Masters Thesis, EV 90-97, Asian Institute of Technology, Bangkok, Thailand.
Chiemchaisri, C., Wong, Y. K., Urase, T., and Yamamoto, K. (1992). Organic stabilization and
nitrogen removal in membrane separation bioreactor for domestic wastewater treatment. Water
Sci. Technol. 25(10), 231240.
Chiemchaisri, C. and Yamamoto, K. (1994). Performance of membrane separation bioreactor at
various temperature for domestic wastewater treatment. J. Membrane Sci. 87, 119129.
Chiemchaisri, C., Yamamoto, K., and Vigneswaran, S. (1993). Household membrane bioreactor in
domestic wastewater Treatment. Water Sci. Technol. 27(1), 171178.
Choo, K. and Lee, C. (1996). Membrane fouling mechanisms in the membrane-coupled anaerobic
biorector. Water Res. 30(8), 17711780.
Churchouse, S., (1998). Operating experiences with Kubota submerged membrane activated sludge
processes, School of Water Sciences. Lecture notes, Cranfield University, U.K.
Ct, P. L., Bersillon, J. L., Huyard, A., and Faup, J. M. (1988). Bubble-free aeration using
membranes: process analysis. J. WPCF 60(11), 19861992.
44
EIA. (1994). Lassainissement des Agglomrations: Techniques dpurations Actuelles et volutions,
Etude Inter Agences, No. 27, Agence de lEau Artois Picaedie.
Erikson, L., Steen, I., and Tendaj, M. (1992). Evaluation of sludge properties at an activated sludge
plant. Water Sci. Technol. 30(12), 251265.
Fakhrul-Razi, A. (1994). Ultrafiltration membrane separation for anaerobic wastewater treatment.
Water Sci. Technol. 30(12), 321327.
Fane, A. G., Fell, C. J. D., Hodgson, P. H., Leslie, G., and Marshall, K. C. (1989). Microfiltrarion
of biomass and biofluids: effects of membrane morphology and operating conditions, Proc. of
the Vth World Filtration Congress, Nice, France. 320329.
Fane, A.G., Fell, C. J. D., and Nor, M. T. (1978). Wastewater treatment using a combined activated
sludge and ultrafiltration system, Proceedings of Asian Pacific Confederation of Chemical
Engineering (APCCHE), Jakarta, November 1978. 117129.
Fane, A. G. and Fell, C. J. D. (1987). A review of fouling and fouling control in ultrafiltration.
Desalination 62, 117-136.
Ghyoot, W. R. and Verstraete, W. H. (1978). Coupling membrane filtration to anaerobic primary
sludge digestion. Environ. Technol. 18, 569580.
Grethlein, H. E. (1978). Anaerobic digestion and membrane separation of domestic wastewater. J.
Water Pollut. Control Fed. 50(3), 754763.
Hall, E. R., Onysko, K. A., and Parker, W. J. (1995). Enhancement of bleached kraft organochlorine
removal by coupling membrane filtration and anaerobic treatment. Environ. Technol.. 16, 115
126.
Harada, H., Momonoi, K., Yamazaki, S., and Takizawa, S. (1994). Application of anaerobic-UF
membrane reactor for treatment of a wastewater containing high strength particualte oragnics.
Water Sci. Technol. 30(12), 307319.
Hare, R. W., Sutton, P. M., Misher, N. P., and Janson, A. (1990). Membrane enhanced biological
treatment of oily wastewater. WPCF conference, October 1990, Washington, USA.
Hardt, F. W., Clesceri, L. S., Nemerow, N. L., and Washington, D. R. (1970). Solid separaion by
ultrafiltration for concentrated activated sludge. J. Water Pollut. Control Fed. 42(12): 2135,
1970.
Hsu, M. and Wilson, T. E. (1992). Activated sludge treatment of municipal wastewater-USA
practice. Water Quality Management Library, Technomic Publication, Vol. 1. 3768.
Irwin, J. (1990). On-site wastewater reclamation and recycling. Water Environ. Technol. 9091.
Ishida, H., Yamada, Y., Tsuboi, M., and Matsumura, S. (1993). Submerged membrane activated
sludge process (KSMASP) its application into activated sludge process with high concentra-
tion of MLSS. Proceedings of the 2nd International Conference on Advances in Water and
Effluent Treatment.
Ishiguro, K., Imai, K., and Sawada, S. (1993). Effect of biological treatment conditions on permeate
flux of UF membrane in a membrane/activated-sludge wastewater treatment system. IDA
Congress, Yokohama, Japan, Nov. 36, 1993. 307314.
Johnson, W. T., Phelps, R. W., and Beatson P. J. (1996).Water Mining using membranes. Proceed-
ings of the Water Reuse for the community and Industry Latest Developments and Future
Directions Symposium, University of South Wales, Sydney, Australia, August 1,1996.
Judd, S. (1998). Membrane bioreactors: why bother school of water sciences. Lecture notes, Cranfield
University, UK.
Kang, I. J., Yoon, S. H., and Lee, C. H. (1996). Comparison of Fouling Characteristics between
Inorganic and Organic Membrane in Membrane Coupled Anaerobic Bioreactor (MCAB),
Membrane conference proceeding, Tokyo, Japan, 1996, 910911.
Kayawake, E., Narukami, Y., and Yamagata, M. (1991). Anaerobic Digestion by a Ceramic Mem-
brane Enclosed Reactor. J. Ferment. Bioeng. 71, 122125.
Kiat, W. Y., Yamamoto, K., and Ohgaki, S. (1992). Optimal fiber spacing in externally pressurized
hollow fiber module for solid liquid separation. Wat. Sci. Tech. 26(5-6), 12451254.
45
Kimura. S. (1991). Japans Aqua Renaissance 90 Project, Water Science and Technology 23, 1573
1592.
Kitamura, Y., Maekawa, T., and Hayashi, H. (1994). Shouchu Distillery Wastewater Treatment
Using Membrane Fermenter Incorporating Ultrafiltration Membrane Separator. J. Soc. Food
Sci. Technol. 41(1), 5864.
Knoblock, M. D., Sutton, P. M., Mishra, P. N., Gupta, K., and Janson, A. (1994). Membrane
biological reactor system for treatment of oily wastewaters. Water Environ. Res. 66, 133139.
Kolega, M., Gorhmann, G. S., Chiew, R. F., and Day, A. W. (1991). Disinfection and clarification
of treated sewage by advanced microfiltration. Water Sci. Technol. 23, 16091618.
Krauth, K. H. and Staab, K. F. (1988). Substitution of final clarifier by membrane filtration within
the activated sludge process with increased pressure: initial findings. Desalination 68, 179189.
Krauth, K. and Stab, K. F. (1993). Pressurised bioreactor with membrane filteration for wastewater
treatment. Water Res. 27, 405411.
Krauth, K. H. and Staab, K. F. (1988). Pressurized biomembrane rector for wastewater treatment.
Hydrotop, 94, 555562.
Lambert, S. (1983). LUltrafiltration: application aux eaux rsiduaries industrielles et au recyclage
des eaux dimmeubles. leau lindustrie les nuisances. 74, 3538
Langlais, B., Denis, Ph., Triballeau, S., Faivre, M., and Bourbigot, M. M. (1992). Test on microfiltration
as a tertiary treatment downstream of fixed bacteria filtration. Water Sci. Technol. 25(10), 219
230.
Li, A., Kothari, D. and Corrado, J.J. (1984). Application of membrane anaerobic system for the
treatment of industrial wastewaters. Proceedings of the 39th Annual Purdue Industrial Waste
Conference, Purdue University, West Lafayette, Indiana, USA, 1984. 627636.
Livingston, A. G. (1994). Extractive membrane biorectors: a new process technology for detoxifying
chemical industry wastewaters. J. Chem. Technol. Biotechnol. 60, 117124.
Lbbecke, S., Vogelpohl, A., and Dewjanin. W. (1995). Wastewater treatment in a biological high-
performance system with high biomass concentration. Water Res. 29(3), 793802.
Magara, Y., and Itoh, M. (1991). The effect of operational factors on solid/liquid separation by ultra-
membrane filtration in a biological denitrification system for collected human excreta treatment
plants. Water Sci. Technol. 23, 15831590.
Manem, J. and Sanderson, R. (1996). Membrane bioreactors. Water Treat. Membrane Process,
McGraw-Hill, Chap. 17.
Matsuoka, H., Fukada, S., and Toda, K. (1992). High oxygen transfer rate in a new aeration system
using hollow fiber membrane. Biotechnol. Bioeng. 40, 346352, 1992.
Mccann, (1998). Membrane Milestone in UK. IAWQ Year book, 1999, 1415.
Mills, W. R. (1996). Update on Water Reuse in Southern California, Symposium on Water Reuse for
the Community and Industry: Latest Developments and Future Directions, University of New
South Wales, Sydney Australia, August 1, 1996. 19.
MRC Asia (Thailand) Ltd, (1997). Bangkok, personnel communication.
Muller, E. B., Stouthamer, A. H., Verseveld, H. W. van, and Eikelboom, D. H. (1995). Aerobic
domestic wastewater treatment in a pilot plant with complete sludge retention by cross-flow
filtration. Water Res. 29(4), 11791189.
Nagano, A., Arikawa, A., and Kobayashi, H. (1992). The treatment of liquor wastewater containing
high-strength susspended solids by membrane bioreactor system. Water Sci. Technol. 26(3-4),
887895.
Onishi, H., Numazawa, R., and Takeda, H. (1980). Process and apparatus for wastewater treatment,
U.S. Patent, 4181604, Jan. 1, 1980.
Parameshwaran, K. (1997). Membrane as air diffuser and solid/liquid separator in a bioreactor
treating domestic wastewater. Masters Thesis. Asian Institute of Technology, Bangkok, Thai-
land.
46
Parameshwaran, K. and Visvanathan, C. (1998). Recent developments in membrane technology for
wastewater reuse, 2nd International Conference on Advanced Wastewater Treatment recycling
and Reuse 98, Italy, Accepted for oral presentation.
Parameshwaran, K., Visvanathan, C., and Ben Aim, R. (1998). Membrane as solid/liquid separator
and air diffuser in a bioreactor. J. Environ. Eng. (under consideration).
Pillay, V. L., Townsend, B., and Buckley, C. S. (1994). Improving the performance of anaerobic
digesters at wastewater treatment works: the coupled cross-flow microfiltration/digester pro-
cess. Proc of the 7th Int. Symp on Anaerobic Digestion, Jan. 2327, 1994, Cape Town, South
Africa. RSA Litho (Pty) Ltd. 577586.
Pliankarom, S. (1996). Application of air backflushing technique in membrane bioreactor for septic
wastewater treatment. Masters Thesis, EV 96-36, Asian Institute of Technology, Bangkok,
Thailand.
Pouet, M-F., Grasmick, A., Homer, G., Nauleau, F., and Cornier, J. C., Tertiary treatment of urban
wastewater by cross-flow microfiltration. Water Sci. Technol. 30(4), 133-139, 1994.
Praderie, M. (1996). Contribution a letudie du traitment des eaux Residuaire Urbaines par boues
actives a membrane immergees. Doctoral dissertation, INP Toulouse, France.
Rittmann, B. E. (1987). Aerobic biological treatment. Environ. Sci. Technol. 21(2), 128136.
Robb, W. L. (1965). Thin silicon membranes their permeation properties and some applications.
Ann. N.Y. Acd. Sci. 119135.
Ross, B. and Strohwald, H. (1994). Membrane add edge to old technology. Water Qual. Int. No. 4,
1994.
Ross, W. R., Barnard, J. P., le Roux, J., and de Villiers, H. A. (1990). Application of ultrafiltration
membranes for solids-liquid separation in anaerobic digestion system: the ADUF process. Water
SA 16(2), 8591.
Roullet, R. (1989). The treatment of wastewater using an activated sludge bioreactor coupled with
ultrafiltration module. Proceedings of Workshop on Selected Topics on Clean Technology.
Vigneswaran et al., Eds., Asian Institute of Technology, Bangkok, Thailand, 171179.
Sato, T. and Ishi, Y. (1991). Effects of activated sludge properties on water flux of ultrafiltration
membrane used for Human excretment treatment. Water Sci. Technol. 23, 16011605.
Scott, J. A. and Smith, K. L. (1997). A bioreactor coupled to a membrane to provide aeration and
filtration in ice cream factory wastewater remediation. Water Res. 31(1), 6974.
Semmens, M. J. (1989). Using micro-porous membranes to separate VOCs from water. Am. Water
Works Assoc. 81, 162167.
Seyfried, C. E. and Brockmann, M. E. (1995). Membranes in wastewater treatment biological aspects
of the separation of biomass with a microfiltration unit. New and Emerging Environmental
Technologies and Products for Wastewater Treatment and Storm Water Collection, 1995,
Toronto, Ontario, Canada, June 47.
Shimizu, Y., Rokudai, M., Tohya, S., Kayaware, E., Yazawa, T., Tanaka, H., and Eguchi, K. (1989).
Filtration characteristics of charged alumina membranes for methanogenic waste. J. Chem. Eng.
Japan 22, 635641.
Smith, C. W., Di Gregorio, D., and Talcott, R. M. (1969). The use of ultrafiltration membrane for
activated sludge separation. Proceedings of the 24th Annual Purdue Industrial Waste Confer-
ence, Purdue University, West Lafayette, Indiana, USA. 13001310.
Strohwald, N. K. H. and Ross, W. R. (1992). Application of the ADUF process to brewery effluent
on a laboratory scale. Water Sci. Technol. 25, 95105.
Suwa, Y., Suzuki, T., Toyohara. H., Yamagishi, T., and Urushigawa, Y. (1992). Single-stage, single-
sludge nitrogen removal by an activated sludge process with cross-flow filtration. Water Res.
26(9), 11491157.
Takeuchi, K., Futumura, O., and Kojima, R. (1990). Integrated Type Membrane Separation Activated
Sludge Process for Small Scale Sewage Treatment Plants. Ebara Infilco Ltd, Japan.
47
Talat, M. (1988). Application of Direct Membrane Separation to Activated Sludge Process. Masters
Thesis. EV 88-25, Asian Institute of Technology, Bangkok, Thailand.
Trouve, E., Urabin, V., and Manem (1994a). Treatment of municipal wastewater by a membrane
bioreactor: results of a semi-industrial pilot-scale study. Water Sci. Technol. 30(4), 151157.
Trouve, E., Dupont, C., Huyard, A., and Manem (1994b). Cost of anaerobic treatment of wastewaters
on membrane bioreactors: optimal filtration device. Proceedings of the 7th International Sym-
posium on Anaerobic Digestion, Cape Town, South Africa. 567576
Trouve, E., Mandra, V., Manem, J., Urbain, V., and Anselme, C. (1994c). Procds membrane pour
la reutilisation des eaux residuaire urbaines Hydrotop April, 1215 Marseille, France, . 563
569
Udeda, T., Hata, K., Kikuoka, Y., and Seino, O. (1997). Effects of aeration on suction pressure in a
submerged membrane bioreactor. Water Res. 31(3), 489494.
Urbain, V., Trouve, E., and Manem, J. (1994). Membrane bioreactors for municipal wastewater
treatment and recycling. Proceeding of 67th Annual Conference and Exposition of Water and
Environment Federation, Vol. 1. 317327
Visvanathan. C and Ben Aim, R. (1989). Studies on colloidal membrane fouling mechanisms in
cross-flow microfiltration, J. Membrane Sci. 45, 315.
Visvanathan, C., Byung-Soo Yang, Muttamara, S., and Maythanukhraw, R. (1997). Application of
air backflushing technique in membrane bioreactor. Water Sci. Technol. 36(12), 259266.
Winfield, B. A. (1979). The treatment of sewage effluents by reverse osmosis- pH based studies of
the fouling layer and its removal. Water Res. 13, 561564.
Winfield, B. A. (1979). A study of the factors affecting the rate of fouling of reverse osmosis
membranes treating secondary sewage effluents. Water Res. 13, 565570.
Yamagiwa, K., Ohmae, Y., Dahlan, M. H., and Ohkawa, A. (1991). Activated sludge treatment of
small-scale wastewater by a plunging liquid jet bioreactor with cross-flow filtration. Bioresource
Technol. 37, 215222.
Yamamoto, K., Hiasa, H., Talat, M., and Matsuo, T. (1989). Direct solid liquid separation using
hollow fiber membranes in activated sludge aeration tank. Water Sci. Technol. 21, 4354.
Yamamoto, K. and Win, K. A. (1991). Tannery wastewater treatment using sequencing batch
membrane reactor. Water Sci. Technol. 22, 16391648
Yamamoto, K. and Urase, T. (1997). Membrane potential. Water Qual. Int. 3334, January/February.
Yasuda, N. and Lamaze, C. E. (1972). Transfer of gas to dissolved oxygen in water via porous and
nonporous polymer membranes. J. Appl. Polym. Sci. 16, 595601.
Zhang, B., Yamamoto, K., Ohgaki, S., and Kamiko, N. (1997). Floc size distribution and bacterial
activities in membrane separation activated sludge processes for small-scale wastewater treat-
ment/reclamation. Water Sci. Technol. 35(6), 3744.
48

Membrane Separation Bioreactors For Wastewater Treatment

Transféré par

Informations du document

Titre original

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Membrane Separation Bioreactors For Wastewater Treatment

Transféré par

Droits d'auteur :

Formats disponibles

Critical Reviews in Environmental Science and

ISSN: 1064-3389 (Print) 1547-6537 (Online) Journal homepage: http://www.tandfonline.com/loi/best20

Membrane Separation Bioreactors for Wastewater

C. Visvanathan , R. Ben Aim & K. Parameshwaran

To link to this article: http://dx.doi.org/10.1080/10643380091184165

Published online: 03 Jun 2010.

Submit your article to this journal

Article views: 4877

View related articles

Citing articles: 196 View citing articles

Full Terms & Conditions of access and use can be found at

Download by: [121.54.32.98] Date: 19 June 2017, At: 07:26

Membrane Separation Bioreactors for

Luang, Pathumthani 12120, Thailand; Email: visu@ait.ac.th; 2Institute National des

II. FEATURES OF MEMBRANE APPLICATION IN BIOLOGICAL

As our understanding of membrane technology grows, they are being applied

Commercial Number Capacity

Rhone Poulenc-TechSep UBIS France Domestic >40 <400 Roullet, 1989

of this material without the consent of the publisher is prohibited.

FIGURE 2. Features of membrane application in biological wastewater treatment. (B,

III. DEVELOPMENT OF MEMBRANE BIOREACTORS

Membranes have been finding wide application in water and wastewater

FIGURE 3. Trends in MBR development.

1. Treated Water Quality

The major problem of conventional activated sludge processes is the settling

In a MBR, sludge retention time (SRT) can be controlled completely indepen-

3. Compact Plant Size

Volumetric capacities are typically high because a high sludge concentration

4. High Rate Decomposition

Treatment efficiency is also improved by preventing leakage of undecomposed

5. Low Rate Sludge Production

6. Disinfection and Odor Control

ASP 1020 0.71 kg MLSS/kg BOD5 Hsu and Wilson, 1992

With the exception of wastewater reuse, membrane separation activated sludge

High capital and operating costs

System reactor volume 1 2,663 3,423 13,694

IV. FACTORS AFFECTING THE MBR PROCESS PERFORMANCE

The main aim of membrane-coupled bioreactors is to improve the efficiency

Selection of the membrane module plays an important role on the membrane

Wastewater type Domestic Synthetic

Membrane configuration UF MF UF MF MF/UF MF/UF MF UF UF

Membrane configuration UF UF plate UF UF Tubular

of this material without the consent of the publisher is prohibited.

Wastewater type Synthetic Domestic Industrial Synthetic Synthetic Industrial Domestic

of this material without the consent of the publisher is prohibited.

of this material without the consent of the publisher is prohibited.

Membrane filtration performance is affected by the resistance of the membrane

where J, permeate flux (m3/m2.s); P, transmembrane pressure (Pa); , viscosity

As shown in Figure 5, increasing the transmembrane pressure above the

crossflow can help improve process performance, equipment selection is an impor-

C. Effects of Aeration on Flux

In all practical membrane filtration applications, as the resistance increases the

The adsorption of macro molecules, which lead to a conditioning film on the

To understand the early phase of biofouling, consider a system of three

Flux decline can be reduced by various means of operating control. Instead of

Module arrangement influences turbulence at the membrane surface. Due to

FIGURE 7. Comparison of net cumulative permeate volume (after 8 h) in different modes

Viscosity of permeate and mixed liquor affects permeation flux. An increase

J = 1.571 log(MLSS) + 7.84 (5)

V. APPLICATION OF MEMBRANE SEPARATION BIOREACTORS IN

A. Bioreactors with Membrane in External Circuit

A pilot plant with a completely mixed biological reactor (4.5 l) connected to

A series of laboratory scale experiments were carried out by Yamamoto et al.