Desalination 202 (2007) 411–417

Sustainable reclamation and reuse of industrial

wastewater including membrane bioreactor
technologies: case studies

N.I. Galil*, Y. Levinsky

Department of Civil and Environmental Engineering, Technion — Israel Institute of Technology,
Haifa 3200, Israel
Tel. +972 4 8292645; Fax +972 4 8293629; email: galilno@tx.technion.ac.il
Received 31 July 2005; accepted 23 December 2005

Abstract
One of the most crucial and difficult elements of the bioprocess is its ability to separate between the biosolids
and the liquid effluent phase. The objectives of this study were to evaluate practical possibilities to upgrade
existing wastewater treatment facilities by operating aerobic treatment based on MBR technology, in order to
obtain high quality effluent for sustainable reclamation and reuse of industrial wastewater. Three different types
of industrial wastewaters have been biologically treated by MBR working on hollow fiber technology: (a) paper
mill; (b) food production; (c) fuel port facilities. The MBR received preliminarily treated effluent by anaerobic,
chemical and physical processes, respectively. The experimental work in this study indicated that biological
treatment of industrial wastewater containing contaminants characterized by hydrophobicity and/or by low
biodegradability would require the adaptation of the MBR operation conditions, by lowering cell residence time
and MLVSS in the bioreactor and by increasing the amounts of excess biosolids accordingly. The effluent was of
high quality and could be considered for reuse in paper mill and food production.
Keywords: Membrane bioreactor; Industrial wastewater; Reuse; Reclamation

1. Introduction stringent regulation [1,2]. The biological treat-

Existing wastewater treatment facilities have ment of wastewater from industrial facilities is
to improve operating performance and provide reported as experiencing difficulties of poor bio-
effluent of higher quality, conforming to more solids separation and voluminous biological
sludge. Bioflocculation is essential for obtaining
a clear effluent, as the end product of activated
*Corresponding author.

Presented at the conference on Wastewater Reclamation and Reuse for Sustainability (WWRS2005), November
8–11, 2005, Jeju, Korea. Organized by the International Water Association (IWA) and the Gwangju Institute of
Science and Technology (GIST).

0011-9164/06/$– See front matter © 2006 Published by Elsevier B.V.

doi:10.1016/j.desal.2005.12.081
412 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417

sludge process [3]. Bioflocculation is basically Yao-po et al. [12] studied MBR treatment of
an aggregation process, in which cells, as well petrochemical wastewater and reported removal
as organic and inorganic colloids, are closely efficiencies of 91% COD, 92% suspended sol-
bound together, mostly by exocellular biopoly- ids, 99% turbidity, 82% phosphorous and 85%
mers, creating a stable biofloc structure [4]. ammonia. Brindle and Stephenson [13] worked
The bioflocculation mechanism is highly with domestic and different types of industrial
sensitive to potential external disturbances cre- wastewater. They reported 85% removal of
ated during the treatment process [5,6]. These COD. The use of MBR could enhance the
disturbances may be of physical and chemical removal of microorganisms, chlorinated aro-
nature, such as temperature, sudden changes in matics, enzymes of cellulose, oil and grease as
pH or organic loading rate, and were found to be well as methanogenic bacteria, originating from
detrimental to bioflocculation, as well as to the anaerobic treatment preceding the aerobic pro-
bioprocess, in anaerobic systems [7]. Phenolic cess. Saung-Goo and Hak-Sung [14] and
and many other organic compounds of aromatic Peys et al. [15] indicated that MBR effluent
nature were reported as potentially causing dam- could achieve superior quality levels: BOD
age to the bioflocculation process, when acti- 5 mg/L and total suspended solids as low as
vated sludge systems were shock loaded with a 1 mg/L.
mixture of phenolic compounds in an integrated The objectives of this study were to evalu-
oil refinery wastewater treatment plant [8]. Sim- ate practical possibilities to upgrade existing
ilarly Rozich et al. [9] reported that when an wastewater treatment facilities by operating
activated sludge process was shock loaded with aerobic treatment based on MBR technology,
2 g/L of phenol, the biofloc changed its appear- in order to obtain high quality effluent for sus-
ance to a swollen, viscous and clustered form tainable reclamation and reuse of industrial
and, moreover, considerable amount of dis- wastewater.
persed material could be noticed.
The high sensitivity of the bioprocess to
some chemical compounds, which may be 2. Methods
found in the influent, often results in effluent Three different industries were included in
characterized by high turbidity, high concentra- this study: (a) paper mill factory which operates
tion of suspended solids, reducing the amount raw solids separators followed by anaerobic
of active biomass in the bioreactor finally lead- treatment; (b) food production plant which oper-
ing to a complete failure of the treatment pro- ates oil and grease separators, chemical floccu-
cess. One of the most crucial and difficult lation and dissolved air flotation; (c) port fuel
elements of the bioprocess is its ability to sepa- facilities which include gravity oil separators.
rate between the biosolids and the liquid efflu- In all the above cases the biological treatment
ent phase. The use of membrane separation was exposed to different problems, which
technologies has been adopted and successfully included: (a) possible presence of deflocculat-
implemented also in the biosolids separation, ing materials such as starch and biocides, in
replacing the conventional sedimentation (gravi- the case of the paper mill; (b) residual hydro-
tational) process. The biosolids separation by phobic compounds which may affect diffusion
membrane bioreactors (MBRs), which are basi- through biosolids surfaces in the food plant,
cally MF and UF processes, can thus remove par- and (c) organic matter characterized by low bio-
ticles in the range of 0.5–10 and 0.005–0.5 mm, degradability (hydrocarbons) in the case of the
respectively [10,11]. fuel port facilities.
 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417 413

3. Results and discussion was reduced from 21.4 to 2.1 mg/L (90%
removal) and ammonia from 9.4 to less than
3.1. Paper mill
1 mg/L (90% removal); (3) the TSS in the efflu-
The full scale treatment plant in a paper mill ent was always lower than 5 mg/L with an aver-
in Israel, includes equalization, raw solids sep- age of 2.5 mg/L, therefore all quality parameters
aration by straining, anaerobic biotreatment reported for total values are very close to the
followed by activated sludge. The operation of soluble values; (4) the bioreactor could maintain
the existing activated sludge is characterized high levels of MLVSS (11,000 mg/L on aver-
by often disturbances, mainly bad settling, age) resulting cell residence times in the range
voluminous bioflocs, followed by wash-out of of 20–25 days.
the biosolids. In order to improve the aerobic The MBR could save the need for further fil-
biotreatment, it was suggested to upgrade the tration. The high effluent quality has already
activated sludge by adopting a membrane biore- promoted a project for the reuse of the effluent
actor (MBR). A pilot plant based on hollow within the paper mill for various production pro-
fiber membrane (Zee Weed 10 supplied by cesses. This pilot was followed by the operation
Zenon) with a capacity of 500 L/day was oper- of two MBR demonstration units in parallel,
ated during 4 months. The flux was tested in the each working on 50 cum/day. One unit was
range of 13–25 L/sqm/h. Usually 15 L/sqm/h based on hollow fiber (Zenon) and the other on
enabled good operation of the MBR system. The flat membranes (Kubota). These units have been
results are summarized in Table 1. operated for about 6 months in order to accumu-
The monitoring of the MBR working on the late the required data for the design of the
effluent of the anaerobic treatment stage indi- full scale treatment unit (8000 cum/day). After
cated the following conclusions (based on aver- about 4 months of operation it could be observed
age values): (1) COD reduction was from 960 to that both MBR demonstration units are produc-
130 mg/L and BOD from 363 to 7 mg/L, remov- ing effluents of similar quality. The main dif-
als of 86 and 98% respectively; (2) substantial ferences, as compared to the first pilot MBR
removals of nitrogen could be observed, TKN unit (500 L/day), indicated lower operational

Table 1
MBR performance of a pilot plant operated at a paper mill

Parameter Influent Effluent Removal (%)

Average St. dev. Average St. dev.
pH 6.8 0.2 7.8 0.2 –
COD total mg/L 960 764 129 30 86
BOD total mg/L 363 323 7.1 7.0 98
TSS mg/L 294 407 2.5 2.1 –
VSS mg/L 261 349 1.1 1.2 –
TKN mg/L 21.4 5.4 2.1 0.9 90
NH4-N mg/L 9.4 8.8 0.98 2.90 90
Turbidity NTU 171 65 0.8 0.6 99
Oil total mg/L 29.5 5.3 10.3 2.2 65
414 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417

MLVSS concentrations, around 8000–9000 mg/L flocculation and dissolved air flotation. The next
on average, and scaling problems caused by the treatment stage will include biological treat-
relatively high hardness of the wastewater. ment, which will be based on MBR technology.
In case of the paper mill factory, the main For this reason a pilot study was carried out dur-
concern was with the high hardness which has ing a period of 4 months. Since one of the raw
created scaling problems during the operation of wastewater streams in this factory is character-
the two demonstration units. The ratio VSS/TSS ized by high concentrations of organic matter,
here indicated 0.78, based on average values. especially oil, the pilot was operated with and
Biosolids dewatering tests, following chemical without including this concentrated stream. The
conditioning by a cationic polymer indicated results are summarized in Table 2.
that MBR biosolids were characterized by The operation of the MBR without the addi-
higher conditioner demand, as compared to con- tion of the concentrated wastewater stream (par-
ventional activated sludge. At this stage of the tial) enabled very high removal of BOD, from
study it could be concluded that the polymer 417 to 4.6 mg/L (99%), oil removal from 20.9 to
addition had to be higher by 50–80% for 7.2 mg/L (66%). The turbidity of the effluent in
MBR biosolids, as compared to conventional these conditions was 1.4 ± 0.8 NTU. The addi-
activated sludge. tion of the concentrated “oily” wastewater to the
MBR influent (full) has substantially increased
the organic loading on the bioprocess also
3.2. Food production affecting the effluent quality. The accumulation
A food factory is located in the Haifa Bay of over 4% oil on the biosolids surfaces
industrial area, specializing in the production of was likely the reason for the release of some
margarine and soups. The existing wastewater colloidal matter, apparently smaller than the
treatment includes gravity separators, chemical ultrafiltration membrane system. The residual

Table 2
MBR performance of a pilot plant operated at food industry

Parameter Partial Full

Influent Effluent Removal (%) Influent Effluent Removal (%)
Average St. Average St. Average St. Average St.
dev. dev. dev. dev
pH 6.6 0.5 7.6 0.3 – 6.1 0.5 7.6 0.14 –
COD total mg/L 797 203 74 14 91 2912 1165 342 169 88
BOD total mg/L 417 106 4.6 4.1 99 1591 634 45 51 97
TSS mg/L 130 151 1.8 0.9 – 1006 947 10.8 8.4 –
VSS mg/L 116 142 0.8 0.6 – 521 546 2.3 1.1 –
TKN mg/L 18.3 6.6 3.1 4.3 (*) 45.2 28.4 7.1 9.8 (*)
NH4-N mg/L 2.2 2.2 1.4 3.1 (*) 11 2.0 4.4 9.9 (*)
Turbidity NTU 21 11 1.4 0.8 93 160 225 4.4 8.4 97
Oil total mg/L 21 14 7.2 6.0 66 264 223 71 122 73

(*) Nitrogen was added.

 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417 415

organic matter and the oil concentrations in the wastewater includes ballast, bilge and run-off
effluent were substantially higher following the water in the rain season. The wastewater has to
addition of concentrated “oily” wastewater and be treated and the effluent is discharged to the
this could explain the increased turbidity of the sea, according to the governmental regulation.
effluent, 4.4 ± 8.4 NTU. The examination of wastewater treatment by
The increased amounts of edible oil in the chemical–physical procedures by flocculation
case of wastewater from food production indi- and dissolved air flotation indicated some
cated that the concentration of these compounds improvement of wastewater quality, however
in the bioreactor could interfere likely by sorption this processes did not provide the required reli-
onto biosolids surfaces, resulting in deterioration ability for achieving the effluent quality, espe-
of the effluent quality. This problem could be cially in cases when gasoline fractions in raw
solved by pretreatment for preliminary removal wastewater increased.
of the oily material or by reducing the cell resi- In order to prevent environmental viola-
dence time in the bioreactor from 10 to 5 days. In tions it was decided to adopt biological treat-
the second case some of the most important ment. For this purpose the MBR process has
advantages of the MBR might be limited, since many advantages since it could handle with
the VSS/TSS ratio and the amounts of excess bio- a wide range of organic contaminants, includ-
solids will be somehow similar to those reported ing soluble and volatile substances. Ballast
for conventional activated sludge. and bilge wastewater have been separately
treated for three months each, by a 500 L/day
3.3. Petrochemical wastewater MBR pilot based on hollow fiber membranes,
Oil and Energy Infrastructure operates working on 15 L/sqm/h. The results are
the Fuel Division in the Port of Haifa. The presented in Table 3.

Table 3
MBR performance of a pilot plant operated at petrochemical industry

Parameter Balast Bilge

Influent Effluent Removal (%) Influent Effluent Removal (%)
Average St. Average St. Average St. Average St.
dev. dev. dev. dev.
pH 7.4 0.3 7.9 0.2 – 7.9 0.2 8.0 0.3 –
COD total mg/L 738 152 460 197 38 485 493 385 105 56
BOD total mg/L 48 20 2.7 1.9 94 60 29 2.1 1.7 97
TOC total mg/L 61 13 25 5.7 59 78 12 31 0.7 60
TSS mg/L 102 110 8.8 6.5 – 272 884 9.0 4.8 –
VSS mg/L 74 97 3.7 4.2 – 193 283 2.9 2.1 –
TKN mg/L 2.4 1.8 18 26 (*) 3.9 1.5 3.1 2.3 (*)
NH4-N mg/L 1.8 1.9 11 16 (*) 3.8 2.3 2.7 1.7 (*)
Turbidity NTU 46 33 0.8 0.9 98 93 107 0.8 0.5 99
Oil mineral mg/L 2.8 – 0.4 – 86 – – – – –
Oil total mg/L – – – – – 4.3 3.4 0.15 0.06 96

(*) Nitrogen was added.

416 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417

The existing treatment stages include gravity Acknowledgements

oil separators with long residence time. The sep- The study was carried out at the Laboratory
arated oily phase is transferred to the neighbor- for Industrial Wastewater Treatment and Water
ing petrochemical industrial complex for reuse Renovation at the Faculty of Civil and Environ-
in production. The influent to the biological mental Engineering, Technion — Israel Institute
treatment contains relatively low concentrations of Technology, Haifa 32000, Israel. The study
of organic matter (TOC around 50–80, BOD was supported by grants from the Israeli Minis-
around 50–60 mg/L and mineral oil 2.8–4.3 mg/L, try of Science, American-Israeli Paper Mill
all based on average values. The bioprocess (AIPM) Hedera, Israeli Food Industries and Oil
required controlled additions of nitrogen and and Energy Infrastructure, Israel. The experi-
phosphorous. mental work was performed in cooperation with
The results indicated in both cases good Protection and Environmental Technologies,
removals of organic matter and oil. The efflu- Israel.
ents were characterized by low suspended solids
and low turbidity. BTEX concentrations were References
reduced from 2 to 4 mg/L to less than 0.01 mg/L
in ballast and less than 0.05 in bilge water. Since [1] A.F. Rozich, C.D. Hahn, W. Wujcik, C. Evans
the MBR configuration was closed and included and C. Bialecki, Implementation of MBR tech-
nology at an industrial biological wastewater
recycling of air and VOC’s, it is likely that
treatment plant, WEFTEC, Conference Proceed-
BTEX were removed by biodegradation. ings, Annual Technical Exhibition & Conference,
The concentrations of MLVSS in the reactor 75th, Chicago, US, September 2002, p. 142.
were usually lower than 1000 mg/L. This could [2] A.P. Togna, A. Bohner, T. Webster, W. Guarini,
be explained by the low organic loading in the P.M. Sutton, M. Murray, S. Robinson and
influent. The cell residence time was 10–12 C. Presgrave, Membrane biological reactor system
days. The oil accumulated on the biosolids was development for treatment of high-strength indus-
in the range of 2%. The VSS/TSS ratio was 0.73 trial wastewater, WEFTEC, Conference Proceed-
ings, Annual Technical Exhibition & Conference,
and no special effects which could be attributed 75th, Chicago, US, September 2002, p. 108.
to oil accumulation have been observed. [3] Metcalf and Eddy, Wastewater Engineering,
Treatment, Disposal and Reuse, Third edn.,
McGraw-Hill Inc., 1991.
4. Conclusions [4] J.L. Pavoni and M.W. Tenney, Bacterial exocellu-
lar polymers and biological flocculation, J. Water
The experimental work in this study indicated Pollut. Control Fed., 44 (1972) 414.
that biological treatment of industrial wastewater [5] N.J. Horan and C.R. Eccles, Purification and
containing contaminants characterized by hydro- characterization of extracellular polysaccharide
phobicity and/or by low biodegradability would from activated sludge, Water Res., 20 (1986)
require the adaptation of the MBR operation 1427.
conditions, by lowering cell residence time and [6] J.W. Morgan, C.F. Forster and L. Evison, Com-
parative study of the nature of biopolymers
MLVSS in the bioreactor and by increasing the
extracted from anaerobic and activated sludge,
amounts of excess biosolids accordingly. The Water Res., 24 (1990) 743.
obtained effluent was of high quality and could [7] R. Blasczyc, D. Gardner and N. Kosarik,
be considered for sustainable reuse in the case of Response and recovery of anaerobic granules
paper mill and food industry. from shock loading, Water Res., 28 (1994) 675.
 N.I. Galil, Y. Levinsky / Desalination 202 (2007) 411–417 417

[8] M. Rebhun and N. Galil, Inhibition by hazardous a petrochemical complex, J. Environ. Sci., 10
compounds in an integral oil refinery, J. Water (1998) 1.
Pollut. Control Fed., 60 (1988), 1953. [13] K. Brindle and T. Stephenson, Detoxification of
[9] A.F. Rozich, A.F. Gaudy and P.C. D’Adamo, industrial wastewaters in an extractive membrane
Selection of growth rate model for activated bioreactor, Water Sci. Technol., 33 (1996) 1.
sludge treating phenol, Water Res., 19 (1985) 481. [14] L. Saung-Goo and K. Hak-Sung, Optimal operat-
[10] M. Cheryan, Ultrafiltration and Microfiltration ing policy of the ultrafiltration membrane bioreac-
Handbook, Technion Pub. Co., 1998. tor for enzymatic hydrolysis of cellulose,
[11] T.K. Chen, C.H. Ni, J.N. Chen and J. Lin, High- Biotechnol. Bioeng., 42 (1993) 737.
strength nitrogen removal of opto-electronic [15] K. Peys, L. Diels, R. Leysen and
industrial wastewater in membrane bioreactor — a C. Vandecasteele, Development of a mem-
pilot study, Water Sci. Technol., 48 (2003) 191. brane biofilm reactor for the degradation of
[12] F. Yao-po, W. Ju-si and J. Zhao-chun, Test of chlorinated aromatics, Water Sci. Technol., 36
membrane bioreactor for wastewater treatment in (1997) 205.