Vous êtes sur la page 1sur 5

G Model

JWPE-23; No. of Pages 5

Journal of Water Process Engineering xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Journal of Water Process Engineering

journal homepage: www.elsevier.com/locate/jwpe

Treatment of restaurant wastewater using ultraltration and

nanoltration membranes
S. Zulaikha, W.J. Lau *, A.F. Ismail, J. Jaafar
Advanced Membrane Technology Research Centre (AMTEC), Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia



Article history:
Received 28 December 2013
Received in revised form 4 May 2014
Accepted 6 May 2014
Available online xxx

In this work, two commercial ultraltration (UF) membranes, i.e. UF PES-10 kDa and UF PVDF-100 kDa
and two nanoltration (NF) membranes, i.e. NF-90 and NF-270 were used to treat restaurant wastewater
discharged from local medium-sized restaurant which served typical Malay halal foods. The wastewater
which was subject to membrane treatment process was collected at the point of discharge without
undergoing any pre-treatment process. The separation performances of membranes were characterized
with respect to BOD5, COD, turbidity and conductivity removal while the membrane sustainability was
evaluated based on the water ux recovery. Promising results of COD and turbidity removal (between
97.8 and 99.9%) were able to achieve regardless of membrane type, indicating the size of pollutants
contributing to COD and turbidity values are signicantly larger than 100 kDa. Removal rates of BOD5
and conductivity were found to be strongly dependent on the membrane pore size. As NF-90 exhibited
the smallest pore structure among the membranes studied, it showed the highest BOD5 (86.8%) and
conductivity removal (82.3%). In terms of sustainability, it is found that the original water ux of NF-90
was able to be retrieved by >50% compared to between 15 and 38% reported in the UF membranes, after
only a simple rinsing process. This indicated that NF-90 is less susceptible to fouling and is more suitable
and reliable to be employed for restaurant wastewater treatment.
2014 Elsevier Ltd. All rights reserved.

Restaurant wastewater

1. Introduction
Over the past thirty years, membrane-based separation process
has emerged as one of the most promising techniques in the water
and wastewater treatment eld owing to the signicant evolution
of the membrane technology [1,2]. However, the potential of
membrane as treatment process has not been examined for all
types of wastewaters. For instance, the exploration of using
membrane for restaurant wastewater is paid little attention and
this can be reected by the limited number of technical papers
available in the literature [3,4].
Restaurant wastewater is water that has been used for cleaning
meats and vegetables, washing dishes and cooking utensils, or
cleaning the oor. The efuent is usually heavily loaded with
organic matters from the leftovers of food and soup which are
made of oily avorings such as soy sauce, seasoning, spice, etc.
Unlike most of the developed countries where the restaurant
efuent is discharged into foul sewers leading to public sewage

* Corresponding author. Tel.: +60 75535926.

E-mail addresses: lwoeijye@utm.my, lau_woeijye@yahoo.com,
lau.woeijye@hotmail.com (W.J. Lau).

treatment plants, the efuent produced by local restaurant is

mostly discharged into storm drains without proper on-site
treatment process. This situation gets even worse with the lack
of awareness by the general public in particular restaurant owner
of the wastewater management issues. Furthermore, the presence
of oil and grease in the wastewater also tends to foul the storm
drain and generates unpleasant odor.
Although aerobic and/or anaerobic treatment of restaurant
wastewater has been proposed in removing oil and grease from
wastewater, the reactor size and operation cost make the process
unacceptable for most small city restaurants [5]. Chemical
coagulation on the other hand is found to be low efciency in
removing light and nely dispersed oil particles and possible
contamination of foods by chemicals [6]. It is also reported that the
removal efciencies of electrocoagulation and electrootation
processes for oil and grease and suspended solids are not very high
under the general operating conditions, making them unable to
produce efuent that meets the acceptable level especially when
the daily average inuent COD of wastewater are too high [7]. In
view of this, it is realized that membrane technology is very
potential to overcome the limitations encountered by the abovementioned treatment processes, owing to the several major
advantages such as superior pollutant removal, consistent efuent

2214-7144/ 2014 Elsevier Ltd. All rights reserved.

Please cite this article in press as: S. Zulaikha, et al., Treatment of restaurant wastewater using ultraltration and nanoltration
membranes, J. Water Process Eng. (2014), http://dx.doi.org/10.1016/j.jwpe.2014.05.001

G Model

JWPE-23; No. of Pages 5

S. Zulaikha et al. / Journal of Water Process Engineering xxx (2014) xxxxxx

quality, more compact footprint and often much simpler operation

[8,9]. Moreover, no chemicals addition is required and the power
consumption is very low, due to the low operating pressure applied
(i.e. several bars) during treatment process.
The main objective of this work is to study the efciency of the
membrane-based technology in removing major pollutants exist in
the wastewater discharged by medium-sized restaurants serving
typical Malay halal food for local people, particularly Muslim
people. Four types of commercial available ultraltration (UF) and
nanoltration (NF) membranes were selected in this study to
investigate their efciencies in reducing the levels of efuent, in
particular BOD5 and COD so as the treated wastewater could
comply with local discharge standard. It is expected that the
outcomes of this work can provide instructive information on the
potential of using membrane technology for restaurant wastewater treatment process in an effort to reduce the environmental
impact and possibly to recycle the treated efuent for non-potable
reuse in the restaurant.
2. Materials and methods
2.1. Local restaurant wastewater
Table 1 shows the characteristics (detected ranges) of raw
restaurant efuents collected from two local medium-sized
restaurants situated in Taman Universiti, Skudai district (coordinate: 1.533448 N 103.616788 E) at four different business hours,
i.e. 0700, 1200, 1600 and 1800 h, together with the limits of
efuent discharged. Both restaurants are halal, serving customers
with traditional Malay foods. But, since the food served for
breakfast, lunch and dinner are different; the wastewater
composition is varied from time-to-time, even it is discharged
by the same restaurant. In this study, the raw inuent sample was
collected at the point of discharge (i.e. entrance of the drain)
without undergoing any treatment process. Efuents were
produced either from the food preparation (e.g. vegetable/meat
washing), leftovers of wastes (e.g. oily avoring, beverages and
soup) or cleaning activities (e.g. dish/utensil washing and kitchen
oor cleaning). As can be seen from the table, the quality of the
efuent with respect to pH, BOD5 and COD failed to meet the local
government discharge standard. The extremely high values of
BOD5 and COD of the restaurant efuent could be attributed to the
bloody efuent resulted from chicken/beef washing as well as the
oily avoring such as soy sauce, seasoning and spice used in food
preparation. Although turbidity and conductivity are not included
in the regulation, it would be still interesting to include both

Table 1
Typical local restaurant efuent characteristics and parameter limits of efuent
(Standard B) of environmental quality (Sewage and Industrial Efuents) regulation

BOD5 (mg/L)
COD (mg/L)
Turbidity (NTU)


Limits of
(bStandard B)

Zam-Zam Corner

Nor Cafe




The range of value shown is obtained from four different measurements at
different periods.
Standard B is the standard parameter limits discharge into all inland waters,
except inland waters within catchment areas.

parameters during sample analysis. It is mainly because both

parameters are widely considered in many wastewater treatment
2.2. UF/NF membranes and ltration experiments
All membranes used in the UF and NF tests are commercially
available membranes. Table 2 shows the information on the four
commercial membranes used for the treatment of restaurant
wastewater. All the ltration experiments were conducted in a
self-stirred membrane permeation cell (Model: HP4750, STERLITECHTM) under nitrogen atmosphere. During experiment, the cell
which containing a circular-shape at membrane (with 14.6 cm2
effective area) was placed on a magnetic stirrer in which a Teoncoated magnetic bar was tted into the cell and stirred at
constant rotation of 250 rpm, providing agitation to minimize
concentration polarization effect during ltration. The temperature and pressure of the ltration experiments were controlled
at 25 8C and 24 bar, respectively. The membrane water ux J
(L/m2 h) was calculated by dividing the permeate volume Q (L) by
the product of effective membrane area A (m2) and the sampling
time t (h).



The efciencies of membrane in removing specic pollutants

from restaurant wastewater R (%) were determined by dividing the
difference between initial feed value Cf and permeate value Cp with
initial feed value, as expressed in Eq. (2).

R 1
Unit used for each parameter, i.e. BOD5, COD, turbidity and
conductivity is mg/L, mg/L, NTU and mS/m, respectively.
In order to obtain the water ux recovery of the membrane, the
fouled membrane after restaurant wastewater treatment experiments was rinsed by letting DI water ow thoroughly on the top
surface of membrane for 15 min. Then, the pure water ux of the
membrane was re-evaluated and compared with its initial water
2.3. Analytical methods
Analysis on the COD concentration of the feed and permeate
samples was performed using spectrophotometer (Model:
DR5000, Hach) by following the procedures handbook provided.
BOD5 analysis was performed with the use of handheld optical
dissolved oxygen meter (Model: ProODO, YSI Inc.). The removal
efciencies of the membranes for conductivity were determined
using a bench conductivity meter (Model: 4520, Jenway).
Characterization on the sample quality with respect to turbidity
was conducted using a portable turbidimeter (Model: 2100Q,

Table 2
UF and NF membranes used in this work.




UF PES-10 kDa
UF PVDF-100 kDa

Alfa Laval
Amfor Inc.

Polyvinylidene diuoride
Polyamide (aMPD-TMC)
Polyamide (bPIP-TMC)

Tight UF
Loose UF


MPD-TMC 1,3-phenylene diamine-benzenetricarbonyl trichloride.

PIP-TMC piperazine-benzenetricarbonyl trichloride.

Please cite this article in press as: S. Zulaikha, et al., Treatment of restaurant wastewater using ultraltration and nanoltration
membranes, J. Water Process Eng. (2014), http://dx.doi.org/10.1016/j.jwpe.2014.05.001

G Model

JWPE-23; No. of Pages 5

S. Zulaikha et al. / Journal of Water Process Engineering xxx (2014) xxxxxx

3. Results and discussion

3.1. Properties of UF/NF membranes
The properties of UF and NF membranes with respect to contact
angle and pure water ux are summarized in Table 3. The pure
water ux of membranes is found to be governed mainly by the
membrane pore size and its surface hydrophilicity. Of the
membranes tested, it is reported that the UF PVDF-100 kDa
membrane exhibited the highest water ux mainly because of its
loose membrane structure, offering minimum transport resistance
for water molecules to permeate. The lowest water ux experienced by NF-90 membrane is greatly attributed to its smallest pore
dimension. Although UF PES-10 kDa membrane demonstrated
much larger pore size than the NF membrane, its water ux was
reported to be relatively lower compared to the NF-270 membrane
when tested under same condition. This phenomenon can be
explained by the fact that NF-270 which is made of piperazine (PIP)
and benzenetricarbonyl trichloride (TMC) monomer is superhydrophilic (i.e. low contact angle value), thus showing very high
afnity for water during ltration process and as a consequence,
better water ux is experienced.
3.2. Separation efciency of membranes
In order to prove the membrane-based technology treatment
system is able to handle signicant peaks in loading, the efuent
with high loadings of COD and BOD5 (discharged by Zam-Zam
Corner Restaurant) was used in the following experiments. Since
the structure of UF PVDF-100 kDa membrane is very porous, the
operating pressure of ltration process was controlled at 2 bar
throughout the experiments in comparison to 4 bar applied for the
rest of the membranes. In normal practice, it is recommended that
the membrane ux measurement should be carried out at same
operating pressure [10,11], but the highly porous structure of
PVDF-100 kDa membrane used in this work has made the
measurement almost impossible to carry out at 4 bar, owing to
the extremely large amount of permeate produced per small area
of membrane surface (i.e. 14.2 cm2). Because of this, the operating
pressure of this membrane was xed at 2 bar. With respect to
permeate quality assessment, the analysis on the oil and grease
level is not included in this work. Instead, COD analysis is
performed as it is a measure that includes oil and grease as well.
Furthermore, COD is a good indicator to assess the quality of
restaurant wastewater which is heavily loaded with organic
matter and other reducing substances.
Fig. 1 shows the separation efciency of UF and NF membranes
in treating local restaurant wastewater. For the efuent treated by
UF membranes, it is found that there was not signicant difference
in the removal efciency, even though UF PES-10 kDa membrane
exhibited much smaller pore structure than that of UF PVDF100 kDa membrane. The promising removal rate achieved by both
UF membranes implied that most of the pollutants contributing to
COD and turbidity are having particle sizes larger than 100 kDa or
Table 3
Properties of UF and NF membranes used in this work.


Pure water
uxb (L/m2 h)

Mean contact
angle (8C)

UF PES-10 kDa
UF PVDF-100 kDa






MWCO, molecular weight cut-off.

All the membrane water uxes were measured at operating pressure of 4 bar,
except UF PVDF-100 kDa which was measured at 2 bar.

Fig. 1. Separation efciency of membranes in treating restaurant wastewater, (a) UF

membranes and (b) NF membranes (feed characteristics of sample collected from
Zam-Zam Corner Restaurant: BOD5: 1097 mg/L, COD: 16,443.33 mg/L, turbidity:
1208 NTU and conductivity: 1027.33 mS/m).

equivalent to 0.04 mm, according to the ltration and separation

spectrum provided by GE Power & Water [12]. A recent work
conducted by Yang et al. [3] for simulated restaurant wastewater
treatment process also reported that high COD removal (between
98.3 and 99.1%) was able to achieve using submerged membrane
bioreactor. Nevertheless, it must be pointed out that the COD
loading of this simulated wastewater was very low, i.e. 358
762 mg/L compared to 16,443 mg/L reported in this work.
Generally speaking, membrane-based technology is very efcient
in eliminating COD, irrespective of the COD concentration of
restaurant wastewater. In contrast to the good removal rates for
BOD5, COD and turbidity, the experimental results showed that
both UF membranes were unable to reduce efciently the
conductivity of efuent. Both UF membranes were reported to
remove only about 5% of the efuent conductivity. The reason may
arise from the inefciency of UF membranes in eliminating
dissolved ions present in the efuent. This observation is very
common for UF membranes as the size of ions is much smaller in
comparison to the pore size of UF, causing the easy passage of ions
through membrane structure.
On the other hand, it is reported that the use of NF membranes
was able to further improve the quality of the treated efuent. The
results showed that the BOD5 removal was enhanced from around
80% as found in the UF membranes to 86.7 and 82.2% for NF-90 and
NF-270, respectively. Besides showing higher BOD5 removal,
almost complete elimination of COD (at least 98.4%) and turbidity
(at least 99.9%) was also reported in NF membranes which were
tested at low operating pressure as UF PES-10 kDa membrane. The
distinct advantage of using NF over UF in treating restaurant
wastewater is the greater efciency in removing efuent
conductivity. Although both NF membranes used in this study
are made of a polyamide layer coated over microporous substrate,

Please cite this article in press as: S. Zulaikha, et al., Treatment of restaurant wastewater using ultraltration and nanoltration
membranes, J. Water Process Eng. (2014), http://dx.doi.org/10.1016/j.jwpe.2014.05.001

G Model

JWPE-23; No. of Pages 5

S. Zulaikha et al. / Journal of Water Process Engineering xxx (2014) xxxxxx

the difference in the characteristics of polyamide layer (possibly

due to different monomers and fabrication conditions) has resulted
in a wide variation in conductivity removal. However, it is believed
that the higher efciency of NF-90 for conductivity removal
compared to NF-270 is mainly due to its relatively smaller pore
structure which restricts the passage of dissolved ions present in
the wastewater. These results are consistent with the results
previously published in which NF-90 membrane exhibited greater
NaCl and Na2SO4 rejection than that of NF-270 when tested under
same conditions [13]. In certain cases, higher surface charge
density of NF may play a role in enhancing ions removal rate, but
this separation mechanism is ruled out in this work because NF-90
possesses lower surface charge than that of NF-270 [14].
The water quality of the efuent treated by membrane
technology with respect to BOD5 and COD concentration is
presented in Fig. 2. It is noted that the BOD5 and COD concentration
of all the treated efuent was above the standard of discharge,
except the COD of NF-90 membrane. For the BOD5 concentration of
the treated efuent, the highest value was 229 mg/L (in UF PVDF100 kDa) while its lowest value was 145 mg/L (in NF-90). For the
COD concentration, the highest value and the lowest value were
355 mg/L (in UF PES-10 kDa) and 92 mg/L (in NF-90), respectively.
Considering the wastewater samples used in this work were highly
loaded with organic matters and other reducing substances (BOD5:
1097 mg/L, COD: 16,443 mg/L), the signicant reduction in BOD5
and COD concentration was considered to be very promising. In
comparison to Hong Kong where the kitchen wastewater is to be
discharged into public sewers for further treatment [6], most of the
restaurants discharge pipe in Malaysia is not connected to any
sewage treatment process. Because of this, on-site treatment
process is urgently needed to avoid the wastewater discharged
directly into storm drain and further causes environmental
pollution. Of these four membranes studied in this work, it is
realized that NF-90 is the best performing membrane in treating
local restaurant wastewater although its BOD5 and COD of treated
wastewater are still slightly higher than the local permission
standards for efuent discharged. Nevertheless, it must be pointed
out that with the use of membrane technology for wastewater
treatment, restaurant owner could enjoy the benets of reusing
treated efuent for non-potable purposes such as kitchen oor
cleaning, toilet cleaning, etc. and further minimizing fresh water
3.3. Membrane water ux behavior
The variations in membrane water ux as a function of ltration
time are illustrated in Fig. 3. As can be clearly seen, the UF PVDF100 kDa membrane suffered remarkable ux decline during 3-h

Fig. 2. Comparison between permeate quality of UF and NF membranes and

discharged standard (feed properties of sample collected from Zam-Zam Corner
Restaurant: BOD5: 1097 mg/L and COD: 16,443.33 mg/L).

Fig. 3. The behavior of membrane ux as a function of ltration time in treating

Zam-Zam Corner Restaurants wastewater, (a) UF membranes and (b) NF

operation time. Its water ux decreased sharply from initial around

45 to <15 L/m2 h at the end of experiment, recording more than
65% reduction in water ux. Comparing with the ux behavior of
UF PVDF-100 kDa membrane, it is found that UF PES-10 kDa
membrane exhibited quite consistent water ux throughout
experiment periods, producing an average water ux of 6.5 L/
m2 h. The signicant ux deterioration of PVDF membrane can
be explained by the serious membrane fouling such as membrane
adsorption and pore plugging owing to the presence of large
number of oil molecules in the restaurant wastewater. It is
generally agreed that the hydrophobic surface of PVDF membrane
is very susceptible to foulant attachment (i.e. oil and grease) which
forms additional transport resistance for water molecules to pass
through [15].
Fig. 3 also shows the water ux behavior of NF membranes
which were tested under same conditions. It is experienced that
the water uxes of both NF-90 and NF-270 membranes were
decreased gradually during experiment periods in which the water
ux of NF-90 decreased from initial 6.9 to 3.0 L/m2 h while NF-270
decreased from 7.6 to 4.2 L/m2 h after 3-h experiment. However, it
must be pointed out that the ux decline occurred in NF
membranes was not as high as the ux decline experienced by
UF PVDF-100 kDa membrane. In NF experiments, the ux
reduction was reported to be 57% and 45% for NF-90 and NF270 membranes, respectively. Fouling problem is unavoidable in
all types of pressure-driven membranes for water and wastewater
separation processes. In general, the ux decline in UF and NF
membrane is caused by the concentration polarization as well as
reversible and irreversible fouling. In certain cases where the
membrane ux reduction is due to reversible fouling, a high water
ux recovery could be retrieved through a simple cleaning process.

Please cite this article in press as: S. Zulaikha, et al., Treatment of restaurant wastewater using ultraltration and nanoltration
membranes, J. Water Process Eng. (2014), http://dx.doi.org/10.1016/j.jwpe.2014.05.001

G Model

JWPE-23; No. of Pages 5

S. Zulaikha et al. / Journal of Water Process Engineering xxx (2014) xxxxxx

Fig. 4. Water ux recovery and pure water uxes of membranes before and after
treatment process using wastewater sample collected from Zam-Zam Corner

In the following section, all the membranes were subject to a

cleaning process by rinsing the membrane surface with DI water in
order to assess the sustainability of membrane for local restaurant
wastewater treatment process.

that all the membranes studied provided very similar COD and
turbidity removal (97.899.4% and 99.9%, respectively), but
different BOD5 and conductivity removal. The excellent separation
efciency of membranes for COD and turbidity could be explained
by the fact that those pollutants contributing to COD and turbidity
values were having particle sizes that were larger than the pore
size of membranes used. With respect to BOD5 and conductivity
removal, it is found that NF membranes demonstrated better
results than those of UF membranes and the smaller the pore size
of membrane, the higher the removal rates for BOD5 and
conductivity, resulting in higher quality of treated efuent.
Because of this, NF-90 achieved 86.8% BOD5 removal and 82.3%
conductivity removal in comparison to 82.2% and 8%, respectively
shown by NF-270. Besides showing higher separation rates for
pollutants, NF-90 is also reported to be less susceptible to fouling
owing to its high water ux recovery rate upon a simple rinsing
process without using any chemical cleaning agent. The research
ndings showed that the small footprint membrane technology
could be potentially implemented in local restaurants, treating not
only the efuent heavily loaded with organic matters but also
offering benets to restaurants owner to reuse the treated efuent
for non-potable purposes.

3.4. Membrane water ux recovery


Fig. 4 presents the pure water ux of each membrane before and

after wastewater treatment process along with the ux recovery
percentage. Clearly, the ux recovery of UF PVDF-100 kDa
membrane was the lowest (15%) among the membranes studied,
indicating water rinsing process is ineffective in retrieving its
original water ux. Chemical cleaning process is very likely to
further recover membrane water ux, but the risk is the alteration
of membrane properties, leading to undesirable damage to
membrane performance in long run. Compared to the UF PES10 kDa membrane which showed only 38% ux recovery, both NF90 and NF-270 membranes demonstrated greater ux recovery
percentage, recording around 52% and 55%, respectively. The
results obtained from the simple membrane surface rinsing
process revealed that NF membranes are less susceptible to
fouling, thus are more suitable and reliable to be used in restaurant
wastewater treatment process. However, further investigations on
the long term stability of membrane performance and cleaning
protocol are of importance prior to full-scale application in

[1] N. Hilal, M. Khayet, C.J. Wright (Eds.), Membrane Modication: Technology and
Applications, CRC Press, Boca Raton, 2012.
[2] N.N. Li, A.G. Fane, W.S.W. Ho, T. Matsuura (Eds.), Advanced Membrane Technology
and Applications, John Wiley & Sons Inc., Hoboken, NJ, 2008.
[3] B. Yang, G. Chen, G. Chen, Submerged membrane bioreactor in treatment of
simulated restaurant wastewater, Sep. Purif. Technol. 88 (2012) 184190.
[4] X. Yu, Z. Zhong, W. Xing, Treatment of vegetable oily wastewater using an
integrated microltration-reverse osmosis system, Water Sci. Technol. 61 (2)
(2010) 455462.
[5] W. Chu, W. Hsu, Pollution source identication and waste loading reduction at
chinese fast food restaurants, Environ. Int. 25 (1999) 97107.
[6] X. Chen, G. Chen, P.L. Yue, Separation of pollutants from restaurant wastewater by
electrocoagulation, Sep. Purif. Technol. 19 (12) (2000) 6576.
[7] X. Chen, G. Chen, P.L. Yue, Novel electrode system for electrootation of wastewater, Environ. Sci. Technol. 36 (4) (2002) 778783.
[8] V. Diez, C. Ramos, J.L. Cabezas, Treating wastewater with high oil and grease
content using an Anaerobic Membrane Bioreactor (AnMBR). Filtration and cleaning assays, Water Sci. Technol. 65 (10) (2012) 18471853.
[9] A. Salahi, T. Mohammadi, Experimental investigation of oily wastewater treatment using combined membrane systems, Water Sci. Technol. 62 (2) (2010)
[10] M. Manttari, L. Puro, J. Nuortila-Jokinen, M. Nystrom, Fouling effects of polysaccharides and humic acid in nanoltration, J. Membr. Sci. 165 (2000) 117.
[11] M. Manttari, M. Nystrom, Critical ux in NF of high molar mass polysaccharides
and efuents from the paper industry, J. Membr. Sci. 170 (2000) 257273.
[12] GE Power & Water, 2013 retrieved from www.ge.com (accessed 03.06.13).
[13] C.S. Ong, W.J. Lau, A.F. Ismail, Treatment of dyeing solution by NF membrane for
decolorization and salt reduction and salt reduction, Desalin. Water Treat. 50
(2012) 245253.
[14] Y. Kwon, K. Shih, C. Tang, J.O. Leckie, Adsorption of peruorinated compounds on
thin-lm composite polyamide membranes, J. Appl. Polym. Sci. 124 (2012) 1042
[15] W.J. Lau, A.F. Ismail, S. Firdaus, Car wash industry in Malaysia: treatment of car
wash efuent using ultraltration and nanoltration membranes, Sep. Purif.
Technol. 104 (2013) 2631.

4. Conclusions
The effectiveness of two different commercially available UF
membranes UF PES-10 kDa and UF PVDF-100 kDa, and two
different NF membranes NF-90 and NF-270, were evaluated in
this study for treating restaurant wastewater discharged from a
medium-sized Malay restaurant. The preliminary results showed

Please cite this article in press as: S. Zulaikha, et al., Treatment of restaurant wastewater using ultraltration and nanoltration
membranes, J. Water Process Eng. (2014), http://dx.doi.org/10.1016/j.jwpe.2014.05.001