Natural Resources, 2011, 2, 71-74
doi:10.4236/nr.2011.22009 Published Online June 2011 (http://www.scirp.org/journal/nr)
71
Treatment and Recycling of Wastewater by
Submerged Hollow Fiber Membrane
Dewen He, Huangnian Zhou, Lei Liu, Dingmin Liang, Lu Du
School of Metallurgical Science & Engineerin, Central South University, Changsha, China.
Received October 10th, 2010; revised April 11th, 2011; accepted April 20th, 2011.
ABSTRACT
In this study, the effects of experimental conditions including the MBR equipped novel device and different operating
modes on permeate flux were studied. The results show that the MBR equipped novel device can reduce the resistance
and enhance the flux, decreasing the total resistance (t = 9.649) to 5.962 and increasing the permeate flux to 15-20
l/m2 hr. The permeate flux of intermittent operating mode is more than that of continuous operation and the value of the
permeate flux is between 15 l/m2 hr and 20 l/m2 hr. The MBR equipped novel device which adopting intermittent oper-
ating mode is most effective in this study and the value of permeate flux is between 20 l/m2 hr and 25 l/m2 hr.
R
Keywords: Membrane Bioreactor, Separation, Fouling, Hollow Fiber Membrane
1. Introduction
Membrane bioreactor (MBR) is an effective technology
for wastewater treatment and recycling, in which gravity
settling of the activated sludge system is replaced by a
membrane separation process such as micro-filtration
(MF) or ultra-filtration (UF) [1-2]. The resulting high-
quality and disinfected effluent implies that MBR proc-
esses can be especially suitable for reuse and recycling of
wastewater [3]. A major factor against the application of
membrane micro-filtration or ultra-filtration with the
activated sludge process is the flux decline due to mem-
brane fouling [4-5].
Membrane fouling is often referred to as any form of
flux decline [6]. This includes reversible effects (such as
cake formation), as well as irreversible effects (such as
adsorption). Concentration polarization, though it does
cause a decline in membrane flux, is a phenomenon
which occurs in solution, and will therefore not be con-
sidered a form of fouling. However, when speaking of
reversible flux decline, or reversible hydraulic resistance,
concentration polarization will be categorized with cake
formation, since all of their effects can be reversed by the
hydraulic back-flushing of the membrane. Irreversible
flux decline, which is always a form of fouling, is caused
primarily by the adsorption of natural organic matter
(NOM) onto the surface or inside the pores of the mem-
brane [7-8]. Though this fouling can be reversed by
chemical treatment, it is not considered “reversible”
since it cannot be reduced by standard physical proc-
esses.
During a long term operation, membranes are periodi-
cally back-washed by pumping a fraction of permeate
back through the membranes [9-12]. Backwashing is
used for removing the cake layer (reversible fouling).
Concentration polarization is reduced by non-suction as
well as reverse flow. Thus backwashing can remove re-
versible hydraulic resistance due to cake formation and
concentration polarization. Backwashing entails reversal
of the permeate flow and power consumption of back-
washing pump.
In this paper, a novel device was introduced during the
filtration. This device could be used to counteract fouling
during filtration (especially for cake formation) without
reverse flow. The behavior of MBR under the combina-
tion of this device and intermittent permeation was then
observed.
2. Materials and Methods
2.1. Activated Sludge MBR Operation
The configuration of the activated sludge MBR system
used in this study is schematically described in Figure 1.
This system consisted of hollow fiber (HF) flat-plat
membrane module and bioreactor, similar to typical
MBR [13]. The bioreactor has a working volume of 30 l.

72 Treatment and Recycling of Wastewater by Submerged Hollow Fiber Membrane
(1. feed tank, 2. electromagnetic valve, 3. level controller, 4. level probe 5.
vacuum gauge, 6. pump, 7. fiber membrane, 8. pump, 9. flow meter, 10.
aerator, 11. bioreactor solenoid e valve)
Figure 1. Configurations of activated sludge MBR system.
Table 1. Characteristics of the wastewater.
COD (mg·L−1) 20~80
NH3-N (mg·L−1) 4~18
TP (mg·L−1) 0.3~2.0
The hydraulic retention time of the bioreactor was set to
6 h for the majority of experiments.
The wastewater used in the study was pumped from a
local drain ditch. Characteristics of the wastewater are
shown in Table 1.
2.2. Membranes and Module Configuration
The micro-filtration membranes were polypropylene
hollow fibers with a pore size of 0.1-0.2 μm. An effective
filtration area was 0.25 m2.
In order to investigate the reversible resistance re-
moval, the membranes has been continuously operated
for two months prior to this study. Irreversible resistance
(adsorption fouling) during this study reached steady and
kept constant. To ensure that the condition and perform-
ance of the membrane module was almost the same in all
experiments, backwashing was performed after every
experiment to removing the reversible resistance.
2.3. Membrane Performance Assessment
The rate and extent of membrane fouling were quantified
by measuring permeate flux at constant suction pressure,
which is defined as permeate vol/unit membrane surface
area and filtration time. The variation of membrane pro-
ductivity was evaluated by plotting permeate flux against
filtration time.
To characterize membrane fouling, the water flux was
initially measured with tap water. Also, water flux was
measured at each step of the cleaning procedure, that is,
after taking out and putting the membrane module in tap
water and then after flushing and backwashing the mem-
brane module. The fluxes and filtration resistances were
measured and calculated collectively.
3. Results and Discussion
3.1. Continuous Operation
The permeate flux decline with time during the mem-
brane filtration of activated sludge was first studied with
a polluted membrane. The permeate flux decreased
sharply after the start of filtration, and reached a pseudo-
steady state after 2-3 h of operation.
The initial sharp decline in permeate flux observed
was mainly due to concentration polarization and cake
layer formation on the membrane surface. Further de-
crease in permeate flux over time was caused by the in-
creased thickness of the cake layer due to additional fou-
lant transport to the membrane surface, cake compaction
and concentration polarization.
3.2. Continuous Operation with the Novel
Device
In order to maintain a steady permeate flux during opera-
tion, the cake layer deposited on the membrane surface
needs to be removed by a shear, which can be created by
various means.
A novel device was employed to control concentration
polarization and cake layer formation on the membrane
surface. The fibers through this device made to-and-fro
motion along the direction vertical to two pipes, and the
continuous deformation of the fibers themselves created
a shear to the cake layer on the membrane surface, and
the mutual rubbing between the fibers due to the con-
tinuous deformation of the fibers also created a shear to
the cake layer on the hollow fiber membrane. In addition,
the continuous movement of the fibers in water enhanced
the stirring of water on the membrane surface, and de-
creased concentration polarization.
Compared to continuous operation in Figure 2, con-
tinuous operation with the device reduced the rate and
extent of membrane fouling [14-15].
The resistance-in-series model was applied to evaluate
the characteristics of membrane fouling. According to
this model, the permeation flux (J) takes the following
form.
t
p
J
R


 (1)
tmprfi
RR RR R
rf

  (2)
where p

is the transmembrane pressure;

is the
dynamic viscosity of the permeate; t is the total re-
sistance; is the intrinsic membrane resistance;
R
m
R
p
R
C opyright © 2011 SciRes. NR

Treatment and Recycling of Wastewater by Submerged Hollow Fiber Membrane 73
is the polarization layer resistance caused by the concen-
tration gradient; rf is the reversible fouling resistance
formed by a strongly deposited cake layer; irf is the
irreversible fouling resistance due to some irreversible
adsorption. Here, the
R
R
p
R
R
term experimentally defined
as the portion of the total resistance dislodged only by
water dropping of the membrane module and then im-
merging in tap water.
These equations and the flux data at the end of con-
tinuous operation with the novel device have been used
to calculate the values of each resistance term given in
Table 2.
Figure 2 shows that rf (the percent of total resis-
tance) was very small during continuous operation with
the novel device and cake layer were almost removed
due to this device.
3.3. Non-Continuous Operation
In general, the membrane coupled with bioreactors is
continuously operated for the effective utilization of
given membranes. Continuous membrane filtration,
however, tends to increase fouling, thus decreasing per-
meate flux through the membrane over time. In this study,
non-continuous suction operation was introduced as a
fouling alleviating alternative. Figure 3 compares the
continuous and non-continuous suction operation modes.
Change in permeate flux was plotted in terms of time (28
Figure 2. Flux variation with respect to time / 26kPa suction
pressure.
Table 2. A series of resistances for the membrane at the end of
continuous operation and continuous operation with the novel
device.
with this device without this device
Resistance Va l u e
(1012 m−1)
Percentage*
(%)
Va l u e
(1012 m−1)
Percentage
(%)
mirf
R
R
 4.123 69.2 4.123 42.7
p
R
1.707 28.6 3.687 38.2
rf
R
0.132 2.2 1.839 19.1
t
R
5.962 100.0 9.649 100.0
minutes’ suction of 2 minutes’ suspension). The results
clearly showed that membrane performance was im-
proved with intermittent suction. This finding can be
explained by the enhanced foulant back transport under
pressure relaxation. From Figure 3, as soon as trans-
membrane (suction) pressure was reduced to zero at
non-suction period, the foulants not irreversibly attached
to the membrane surface, diffused away from the mem-
brane surface because of the concentration gradient. As a
result, foulant accumulation near the membrane surface
was lessened and the rate of membrane fouling was re-
duced.
Flux recovery by pressure relaxation has been re-
ported by Hong et al. [10], who studied the ul-
tra-filtration of activated sludge. Like their study, the
permeate flux was only partially recovered, indicating
that reversible hydraulic resistance due to cake forma-
tion and concentration polarization existed.
3.4. Non-Continuous Operation with the Device
With operation of the device, the cake layer was almost
removed. After two minutes’ pressure relaxation, con-
centration polarization was removed. As shown in Fig-
ure 4, almost 100% flux recovery was observed under
the intermittent suction with the device. In consideration
of long-term operation of fibers, the fibers kept loose and
cannot tension during to-and-fro motion. Thus a portion
of fibers close to catchments pipes hardly made continu-
ous deformation and no movement of fibers. The cake
layer on the membrane surface of this portion of fibers
was not eliminated.
4. Conclusions
The MBR equipped novel device can improve the per-
meate flux. The value of the MBR equipped novel device
is between 15 l/m2 hr and 20 l/m2 hr. It is more than the
Figure 3. Flux variation with respect to time with a device /
26kPa suction pressure.
C opyright © 2011 SciRes. NR

Treatment and Recycling of Wastewater by Submerged Hollow Fiber Membrane
74
[5] X. C. Wang and J. Wang, “Kinetic Study of Membrane
Fouling under Cross-Flow Ultrafiltration Opteration,”
Environmental Chemistry, Vol. 21, No. 6, 2002, pp.
552-558.
[6] G. E. Wetterau, M. M. Clark and C. Anselme, “A dynamic
model for predicting fouling effects during the ultrafiltra-
tion of a groundwater,” Journal of Membrane Science,
Vol. 109, No. 2, 1996, pp. 185-204.
doi:10.1016/0376-7388(95)00200-6
[7] M. Cheryan, “Ultrafiltration Handbook,” Technomic,
Lancaster, 1986.
[8] S. Kunikane, Y. Magara, M. Itoh and O. Tanaka, “A
Comparative Study on the Application of Membrane
Technology to the Public Water Supply,” Journal of
membrane science, Vol. 102, 1995, pp. 149-154.
doi:10.1016/0376-7388(94)00292-7
Figure 4. Flux variation with respect to time / 26kPa suction
pressure; non-continuous suction operating mode with the
device was not eliminated.
[9] T. Jiang, M. D. Kennedy, G. J. Walter, van der Meer, P. A.
Vanrolleghem and J. C. Schippers, “The Role of Blocking
and Cake Filtration in MBR Fouling,” Desalination, Vol.
157, 2003, pp. 335-343.
doi:10.1016/S0011-9164(03)00414-4
value of MBR. (10-15 l/m2 hr)
The permeate flux of intermittent operating mode is
more than that of continuous operation and the value of
the permeate flux is between 15 l/m2 hr and 20 l/m2 hr. [10] E. H. Bouhabila, R. B. Aïm and B. Hervé, “Fouling Char-
acterization in Membrane Bioreactors,” Separation and
Purification Technology, Vol. 22-23, 2001, pp. 123-132.
doi:10.1016/S1383-5866(00)00156-8
The MBR equipped novel device which adopting in-
termittent operating mode is most effective in this study
and the value of permeate flux is between 20 l/m2 hr and
25 l/m2 hr. [11] C. Albasi, Y. Bessiere, S. Desclaux and J. C. Remigy,
“Filtration of Biological Sludge by Immersed Hol-
low-Fiber Membranes: Influence of Initial Permeability
Choice of Operating Conditions,” Desalination, Vol. 146,
2002, pp. 427-431. doi:10.1016/S0011-9164(02)00527-1
5. Acknowledgements
The authors acknowledge the financial support of Na-
tional 863 Project (2007AA06Z374).
[12] C. Albasi, Y. Bessiere, S. Desclaux and J. C. Remigy,
“Filtration of Biological Sludge by Immersed Hol-
low-Fiber Membranes: Influence of Initial Permeability
Choice of Operating Conditions,” Desalination, Vol. 146,
2002, pp. 427-431. doi:10.1016/S0011-9164(02)00527-1
REFERENCES
[1] A. S. Malik, O. Boyko, N. Atkar and W. F. Young, “A
Comparative Study of MR Imaging Profile of Titanium
Pedicle Screws,” Acta Radiologica, Vol. 42, No. 3, 2001,
pp. 291-293. doi:10.1080/028418501127346846
[13] C. Albasi, Y. Bessiere, S. Desclaux and J. C. Remigy,
“Filtration of biological sludge by immersed hollow-fiber
membranes: influence of initial permeability choice of
operating conditions,” Desalination, Vol. 146, 2002, pp.
427-431. doi:10.1016/S0011-9164(02)00527-1
[2] D. W. He, Y. T. Xiao and X. Li, “Treatment of Oil/Water
Emulsion by Polyethylene Glycol Ultrafiltration Mem-
brane,” Journal of Central south University, Vol. 12, No.
5, 2005, pp. 542-545. [14] K. J. Kim, A. G. Fane, C. J. D. Fell and D.C. Joy, “Foul-
ing Mechanisms of Membranes during Protein Ultrafil-
tration,” Journal of membrane science, Vol. 68, No. 1,
1992, pp. 79-91. doi:10.1016/0376-7388(92)80151-9
[3] N . Cicek, H. Winna, M.T. Suidan, B. E. Wrenn, V. Ur-
bain and J. Manem, “Effectiveness of the Membrane Bio-
reactor in the Biodegradation of High Molecular Weight
Compounds,” Water Research, Vol. 32, No. 5, 1998, pp.
1553-1563. doi:10.1016/S0043-1354(97)00350-3
[15] M. R. Wiesner and S. Chelam, “The promise of the mem-
brane technology,” Environment Science and Technology,
Vol. 33, No. 17, 1999, pp. 360-366.
doi:10.1021/es993008l
[4] T. Mukai, K. Takimoto, T. Kohno and M. Okada, “Ul-
trafiltration Behaviour of Extracellular and Metabolic
Products in Activated Sludge System with UF Separation
Process,” Water Research, Vol. 34, No. 3, 2000, pp.
902-908. doi:10.1016/S0043-1354(99)00208-0