Journal of Hazardous Materials 286 (2015) 416424

Contents lists available at ScienceDirect

Journal of Hazardous Materials

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

Organics and nitrogen removal from textile auxiliaries wastewater

with A2 O-MBR in a pilot-scale
Faqian Sun a , Bin Sun a,b , Jian Hu a , Yangyang He a , Weixiang Wu a,
a
b

Institute of Environmental Science and Technology, Zhejiang University, Hangzhou 310058, China
Shanghai Electric Group Co. Ltd. Central Academe, Shanghai 200070, China

h i g h l i g h t s

A pilot-scale A2 O-MBR system treating textile auxiliaries wastewater was assessed.

Organic matter and recycle ratio strongly affected the performance of the system.
GC/MS analysis found some refractory organics in the MBR permeate.
Combination of organic foulants and inorganic compounds caused membrane fouling.

a r t i c l e

i n f o

Article history:
Received 12 October 2014
Received in revised form 8 December 2014
Accepted 10 January 2015
Available online 13 January 2015
Keywords:
Textile auxiliaries wastewater
Anaerobicanoxicaerobic
Membrane bioreactor
Recycle ratio
Refractory organics

a b s t r a c t
The removal of organic compounds and nitrogen in an anaerobicanoxicaerobic membrane bioreactor
process (A2 O-MBR) for treatment of textile auxiliaries (TA) wastewater was investigated. The results
show that the average efuent concentrations of chemical oxygen demand (COD), ammonium nitrogen (NH4 + N) and total nitrogen (TN) were about 119, 3 and 48 mg/L under an internal recycle ratio of
1.5. The average removal efciency of COD, NH4 + N and TN were 87%, 96% and 55%, respectively. Gas
chromatographmass spectrometer analysis indicated that, although as much as 121 different types of
organic compounds were present in the TA wastewater, only 20 kinds of refractory organic compounds
were found in the MBR efuent, which could be used as indicators of efuents from this kind of industrial
wastewater. Scanning electron microscopy analysis revealed that bacterial foulants were signicant contributors to membrane fouling. An examination of foulants components by wavelength dispersive X-ray
uorescence showed that the combination of organic foulants and inorganic compounds enhanced the
formation of gel layer and thus caused membrane fouling. The results will provide valuable information
for optimizing the design and operation of wastewater treatment system in the textile industry.
2015 Elsevier B.V. All rights reserved.

1. Introduction
As one of the largest industries in the world, the textile industry consumes large quantities of textile auxiliaries (TA), which
include more than 100 kinds of specialty chemicals, such as softening agent, phosphates, polyamide resins, acrylic chelating agents,
polyurethane coating agents and stiffening agents. In the TA production, considerable amounts of TA wastewater are generated. The
TA wastewater is very chemical-intensive and known to contain
high concentrations of organic matter, non-biodegradable matter,
toxic substances and ammonia [13]. These compounds produce
long-term environmental impacts, it is therefore important to

Corresponding author. Tel.: +86 571 88982020; fax: +86 571 88902020.
E-mail address: weixiang@zju.edu.cn (W. Wu).
http://dx.doi.org/10.1016/j.jhazmat.2015.01.031
0304-3894/ 2015 Elsevier B.V. All rights reserved.

remove these organics and ammonia from the TA wastewater for

reducing their harm to the environment.
TA wastewater could be treated by physicochemical
and biological methods or suitable combinations of them.
Physicochemical methods involve adsorption, ion exchange,
coagulationocculation, as well as advanced oxidation processes
(such as Fenton oxidation, ozonation, photocatalytic oxidation
and electrochemical oxidation) [46]. However, methods such
as coagulation, ion exchange, and adsorption only transfer the
organic pollutants from one phase to another. Advanced oxidation
processes, such as Fenton oxidation, ozonation, photocatalytic
oxidation and electrochemical oxidation, are very efcient and
the fastest way for destruction of organic compounds, but they
are expensive and could not be adopted commercially. Compared
with physicochemical methods, biological treatment is often the
most economical alternative for pollutant removal [7].

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

Due to the low biodegradability and high toxicity of many

textile chemicals, a conventional activated sludge system is inadequate in removing high concentrations of refractory organics and
ammonia. In recent years, the membrane bioreactor (MBR), because
of complete biomass retention, has been successfully integrated
with an anaerobicanoxicaerobic (A2 /O) system for industrial
wastewater treatment [8]. The combined system was efcient
and cost-effective in removing refractory pollutants and ammonia, especially at high and varying loading rates [9]. The use of
anaerobic process as a pretreatment process to partially convert
refractory organics to intermediates that are more readily degradable, could be attractive for refractory wastewater treatment. The
anoxicaerobic process is a good option for achieving biological
nitrogen removal via pre-denitrication and aerobic nitrication.
MBR can keep a long sludge retention time (SRT), which allows the
system to keep a sufcient amount of slow-growing bacteria, such
as ammonia oxidizing bacteria and those specializing in degrading
refractory compounds [10]. Therefore, the A2 O-MBR system is very
attractive for chemical-intensive industrial wastewater treatment.
The efciencies of organic compounds and nitrogen removal in
the A2 O-MBR system are inuenced by many factors, such as chemical characteristics of wastewater, internal recycle ratio, hydraulic
retention time (HRT) and substrate loading rate. [1113]. The chemical characteristics of TA wastewater determined the biodegrading
capacity of the system. Internal recycle ratio was found to have an
important role in nitrogen removal performance of the system. In
addition, membrane fouling was inevitable in MBR systems. In the
previous studies reported in the literature, there was little information on the nitrogen removal and behaviors of organic compounds
in a A2 O-MBR system for TA wastewater treatment particularly at
pilot-scale operation.
The objectives of the present research were: (1) to investigate
the pollutant removal performance in a A2 O-MBR system, (2) to
characterize in detail organic chemical composition in the raw
wastewater and the efuents, (3) to compare the performance of
two kinds of membrane modules and examine the membrane fouling behaviors.
2. Materials and methods
2.1. Experimental set-up and operating conditions
The pilot-scale experiments were carried out in a sequential system (in Fig. 1) of an anaerobic reactor (A1), an anoxic reactor (A2),
followed by an aerobic membrane bioreactor (O-MBR). The working volumes of the three reactors were 3.8 m3 , 7.5 m3 and 5.6 m3 ,
respectively. Reactor A1 was packed with bamboo carbon, which
was generated at 600 C under a slow pyrolysis process with a diameter of 35 mm. In the reactor A2, a mechanical stirrer was installed
to agitate its content. The O-MBR could divide into buffer zone
and reaction zone. An internal pump was installed in the buffer
zone and mixed liquor from the bottom was continuously pumped
to the reaction zone. Two different kinds of membrane modules
were immersed and symmetrically placed in the reaction zone. One
was a hollow ber (HF) membrane module made of polyvinylidene
uoride (PVDF) with a nominal pore size of 0.1 m and a total membrane surface area of 12.5 m2 (SMM-1013, Memstar Technology
Co. Ltd., Singapore). The other one was a at-sheet (FS) membrane

417

module with a nominal pore size of 0.1 m and a total membrane

surface area of 10.4 m2 (DF80, Jiangsu Dafu Membrane Technology Co. Ltd., China). Air diffusers were installed underneath the
membrane module to provide dissolved oxygen (DO) as well as
to control membrane fouling and clogging. DO was maintained at
35 mg/L in the MBR during the experiments. Membrane ltration
through a suction pump was carried out in an intermittent suction mode with 9 min of suction followed by a 2 min release. In situ
membrane chemical cleaning was performed to reduce membrane
fouling by 0.5% NaClO for 1 h when the trans-membrane pressure
(TMP) increased to 20 kPa.
The sequential system was operated in a pre-denitrication
mode, which consisted of three different steps. Firstly, TA
wastewater was continuously pumped into reactor A1 for hydrolysis/acidication. Secondly, the water passed to reactor A2, where
denitrication took place and a fraction of organic matter was
degraded by heterotrophic bacteria. Thirdly, the water passed
to the O-MBR where nitrication and degradation of remaining
organic matter occurred, and the permeate water was separated
by membrane modules while nitrate was recycled to A2 by partial recirculation. The inoculating sludge was drawn from the
anaerobic sludge in a conventional TA wastewater treatment plant
(Hangzhou, China). During the study, different internal mixed
liquor recycle ratios from O-MBR to A2 were employed to assess
chemical oxygen demand (COD) and nitrogen compounds removal
of the system. Recycle ratios ranged from 0.5 to 2.5 of the inuent ow rate. At the start-up period (Phase I), the inuent ow
rate was kept at 250 L/h, and the recycle ratio was controlled at
about 0.5. The system was operated stably for 20 days. Afterwards,
the recycle ratio was changed to about 1.5 and the system was
then operated in this condition for 74 days (day 2194, Phase II).
From day 95 onward, the recycle ratio was maintained at 2.5 (Phase
III) and methanol at a equivalent concentration of about 240 mg/L
COD was supplemented as an organic source in the A1 efuent.
Because the recycle ratio was high at all times, mixed liquor suspended solids (MLSS) concentrations in the reactors were similar.
During the whole period of the study, no sludge was removed
from the plant except for some incidents with accidental sludge
loss. MLSS concentration in the O-MBR and the A2 uctuated at
35005000 mg/L. The experiment was conducted under an ambient
temperature of 2025 C. Furthermore, parameters such as temperature, level of the tanks, TMP, ow rates of partial recirculation
and the permeate inside the O-MBR were measured automatically
and registered continuously in a database with the aid of Realinfo
software (Realinfo., China).

2.2. Characteristics of TA wastewater

TA wastewater was collected from Transfar Chemicals Co. Ltd.,
one of the largest TA manufacturing factories located in the southeast of China. It was rstly pumped into an intermediate tank before
being continuously pumped into the bioreactor. A summary of the
inuent characteristics at different experimental phases are shown
in Table 1. The raw wastewater had a highly variable composition. It can be seen that the average COD concentrations in the
three phases were 657, 944 and 828 mg/L, while total nitrogen (TN)
concentrations were 115, 106 and 121 mg/L, respectively.

Table 1
Characteristics of TA wastewater observed during the experiments.
Phase

pH

COD (mg/L)

NH4 + N (mg/L)

TN (mg/L)

COD/TN

Number of data points

I
II
III

7.299.37
7.428.71
7.508.74

657 127
944 163
828 216

87 5
90 19
94 21

115 7
106 19
121 31

6.9 1.8
8.0 2.0
6.0 1.3

10
35
23

418

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

Fig. 1. Schematic diagram of the pilot-scale A2 O-MBR system.

2.3. Analytical methods

Ammonium nitrogen (NH4 + N), nitrite nitrogen (NO2 N) and
nitrate nitrogen (NO3 N) were analyzed according to the standard methods for the examination of water and wastewater (APHA,
1998). COD and TN were measured by the Hach COD kits and Hach
TN kits, respectively (Hach, USA). The pH was measured with a pH
meter (SG2, Mettler-Toledo, Greifensee, Switzerland) and DO by an
oxygen probe (YSI 550A, YSI, Ohio, USA). MLSS was determined by
vacuum ltration of 10 mL of sludge and then dried at 105 C for
2 h.
Gas chromatographmass spectrometer (GCMS) was used to
analyze of the samples with liquidliquid extraction pretreatment
using CH2 Cl2 (chromatogram pure grade, Tianjin Chemical Factory,
Tianjin, China). Subsequently a 200 mL sample was extracted using
10 mL of CH2 Cl2 (chromatogram pure grade, Fisher, USA) three

times at neutral, alkaline and acidic conditions, respectively [14].

Then, a 1 mL pretreated sample was analyzed by 6890 N/5975B
GCMS system (Agilent Corporation, USA). A DB-5MS capillary column with an inner diameter of 0.25 mm and a length of 30 m was
used in the separation system. The GC column was operated in a
temperature programmed mode by maintaining the temperature
at 40 C for 4 min, then increasing to 300 C with an increment of
8 C/min, and total run time was 46 min. The electron impact energy
was set at 70 eV. Organic compounds analysis was undertaken with
reference to the instrument library (NIST 05. L) database.
At the end of ltration, a piece of membrane was cut from
the middle of the membrane modules. The sample was xed with
3.0% glutaraldehyde in 0.1 M phosphate buffer at pH 7.2. The sample, dehydrated with ethanol and silver-coated by a sputter, was
observed using the scanning electron microscopy (SEM) (JEOL JSM5600LV, Tokyo, Japan). Inorganic elementary analysis was carried

Fig. 2. Variation of COD concentrations in the pilot-scale A2 O-MBR system during the 3 phases of operation.

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

out on an Elementary Vario El (Germany) system. Wavelength dispersive X-ray uorescence (WD-XRF) data were obtained from a
Bruker S4 Explorer spectrometer.
3. Results and discussion
3.1. COD removal
The COD removal of the pilot-scale A2 O-MBR during the whole
operation period is presented in Fig. 2. It shows that the A2 O-MBR
performed well in organic matter removal. COD concentration in
the TA wastewater was in the range of 3721291 mg/L with an
average value of 879 mg/L. In the reactor A1, about 100 mg/L COD
was steadily removed. The relatively low COD removal efciency
implied organic compounds with macromolecular structure and
large molecules in the raw wastewater were partially converted
into small-molecule substrates through hydrolysis-acidication
processes in an anaerobic environment. A drop in mean pH of 0.7
unit appears to support this (data not shown). In the reactor A2, COD
concentration was diluted by mixed liquor recirculation. The average COD concentration of A2 efuent was 320 mg/L. In the O-MBR,
most of the COD was removed. Despite the uctuations of inuent
COD concentrations, the average efuent COD concentration was
131 mg/L, with an average total COD removal of 85%. Moreover, it
was also observed that efuent COD concentration was not affected
when the recycle ratio was increased from about 0.5 to 2.5. These
results indicated the system could maintain a consistently good
performance of COD removal. The efuent could meet the national
discharge standard of water pollutants for industrial wastewater in
China (COD < 180 mg/L), and was discharged to municipal sewage
treatment plant for further treatment.
3.2. Nitrogen removal
An overview of internal recycle ratio from aerobic reactor to
anoxic reactor is given in Fig. 3a. The whole operation period was
divided into three phases according to different internal recycle
ratios. NH4 + N and TN concentration along the reactors in three
phases were also measured to gain a better insight into nitrogen
removal (Fig. 3b and c). During the rst 20 days, the internal recycle
ratio was maintained around 0.5 (Phase I). After 10 days of start-up,
the stable performance of NH4 + N and TN removal from A2 O-MBR
was observed. As shown in Fig. 3, when NH4 + N concentration of
raw wastewater was 87 5 mg/L, the efuent concentration from
anoxic reactor was only 57 2 mg/L because of the dilution of the
recycled mixed liquor from aerobic reactor. Nitrication occurred
in the aerobic reactor and the efuent NH4 + N concentration was
9 4 mg/L, which demonstrated that 90% of NH4 + N was oxidized.
During Phase I operation, with an average inuent TN concentration of 115 mg/L, TN concentration in the efuent from anaerobic
reactor, anoxic reactor and aerobic reactor was 103 8, 67 4 and
54 8 mg/L, respectively, indicating that TN was mainly removed in
anoxic reactor and aerobic reactor. The total TN removal efciency
of the A2 O-MBR system was about 53%.
To improve TN removal, internal recycle ratio was increased
from 0.5 to 1.5 through increasing the nitrate load supplied to the
anoxic reactor in Phase II (days 2194). The NH4 + N concentration
in the inuent ranged from 70 to 93 mg/L, except for shock loading
of NH4 + N (days 5470). During stable operation, the nal efuent concentration of NH4 + N in the aerobic reactor remained less
than 5 mg/L, while the NO3 N concentration averaged 42 mg/L
(Table 2). However, it is interesting to observe that denitrication
was not promoted with increase in the internal recycle ratio. TN
concentration in the efuent from anoxic reactor and aerobic reactor was 54 11 mg/L and 48 9 mg/L, respectively. In the Phase
II, although the internal recycle ratio was increased to 1.5, the TN

419

removal efciency was around 55%. In the Phase III (days 95140),
the internal recycle ratio was further increased to around 2.5. As
shown in Table 2, the efuent NH4 + N concentration in the O-MBR
was 3 1 mg/L, whereas, the TN concentration in the nal efuent was 50 11 mg/L. The results indicated that increase of recycle
ratio from 1.5 to 2.5 did not result in an increase of the percentage
of TN removal.
It is well-known that, for a pre-denitrication system, the internal mixed liquor recycle ow from the aerobic to the anoxic zones
plays an important role that affects the TN removal efciency. Theoretically, the TN removal should be 33%, 60% and 71% with the
internal recycle ratio of 0.5, 1.5 and 2.5 when raw wastewater contains enough biodegradable organic matter. In this study, it was
calculated that TN removal in the anoxic reactor was 28% and 43%
for the internal recycle ratio of 0.5 and 1.5. As the total TN removal
with the internal recycle ratio of 2.5 was 58%, TN removal in the
anoxic reactor could not exceed 50%. The results indicated that an
increase of recycle ratio from 0.5 to 1.5 had a greater effect than an
increase from 1.5 to 2.5. Several researchers have reported that TN
removal at internal recycle ratios lower than 5 was very similar to
the theoretical value. Tan and Ng [15] found that TN removal efciencies of domestic wastewater were 57%, 74%, 72% and 61% for
recycle ratio of 1, 3, 5 and 10, respectively. Baeza et al. [16] observed
that TN removal efciencies were 67%, 72% and 80% for the recycle
ratio of 0, 2 and 5, respectively. The maximum TN removal efciency of 58% in this study was much lower than the theoretical
value, indicating that the denitrication capacity was inhibited.
Under the anoxic conditions, the performance of denitrication could be enhanced by the availability of easily biodegradable
organic matter, enough HRT and low DO concentration in the recycle ow. In the three phases, HRT in the anoxic reactor (20 h, 12 h
and 8.6 h) was enough for denitrication in the anoxic zone [17].
Apparently, organic matter and DO concentration were the main
factors responsible for the limited denitrication capacity. A wide
range of COD/TN ratios have been reported in the literature for complete denitrication. Tseng et al. [18] have observed that nearly
complete denitrication was achieved when the inuent COD/TN
ratios of domestic wastewater was 5.3 or greater. Sun et al. [19]
proposed that the COD/NO3 N ratio required to achieve rapid
denitrication of 500 mg NO3 N should be higher than 6.0 when
the organic source is from landll leachate. Henze et al. [20] have
proposed that the corresponding COD/NO3 N ratio of 611 could
achieve good nitrogen removal in single sludge systems for predenitrication process. As shown in Table 1, COD/TN ratio in the
present study ranged from 4.7 to 10, which could be dened as a low
C/N ratio wastewater [21], and TA wastewater contained refractory
organics more than 15% of inuent COD (Fig. 2), resulting in a poor
TN removal much lower than the theoretical value. Moreover, in
Phase III, an extra addition of methanol in the anoxic reactor did
Table 2
Concentrations of nitrogen compounds in the inuent and the efuent of each
reactor.
A1 (mg/L)

A2 (mg/L)

O-MBR (mg/L)

Removal efciency (%)

Phase I
NH4 + N
NO3 N
TN

87 4
<1
103 8

57 2
67
67 4

94
33 6
54 8

90 5

53 5

Phase II
NH4 + N
NO3 N
TN

94 21
<1
110 23

38 9
87
54 10

32
42 10
48 9

96 3

55 7

31

50 11

97 2

58 5

Phase III
NH4 + N
NO3 N
TN

420

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

Fig. 3. Proles of internal recycle ratio, ammonia and TN concentration in the A2 O-MBR system.

not promote the increase in nitrogen removal. These data show that
the limited denitrication capacity is attributable to the oxidation
of organic matter in the anoxic reactor with the oxygen of the internal recycle at high ratios. High DO concentration (above 5 mg O2 /L)
present in the recycle mixed liquor deteriorated TN removal efciency [15,22]. Therefore, it could be concluded that a recycle ratio
around 1.5 should be recommended to provide a proper comprise
between removal efciency and cost under normal conditions.
3.3. Behavior of organic compounds
In order to gain more insight into the organic composition of TA
wastewater and the removal mechanisms for these organics during
the treatment process, GCMS analysis was carried out. The chromatograms of the raw wastewater and the efuent from each unit
of A2 O-MBR are shown in Fig. S1. It could be seen that, at least 121
types of organic components were present in the TA wastewater,
which contained 92 kinds of alkanes and 29 other kinds of organics belonging to chemical additives and their derivatives. After
treatment by anaerobic reactor, the categories of organic compounds decreased to 90. These organics were more readily

degraded in the subsequent anoxic stage with the categories further decreasing to 55. At last, only 20 different kinds of organic
compounds were found in the permeate from MBR.
The details of organic compounds are summarized in Table 3 .
It appears that the alkanes and butylated hydroxytoluene (BHT),
corresponding to 47.6% and 26.6% of the total integration area,
respectively, dominated the organic matter in the TA wastewater.
BHT, a common antioxidant, was likely to come from the raw chemicals used for TA production [23]. Certain compounds, such as pyridine, tetramethylbutanedinitrile (TMSN), butoxytrimethylSilane,
p-xylene and 1,3-dimethylbenzene, which each covered more
than 1% of integration area, were the sub-dominant organic species,
whereas other compounds were the minor species. These organic
compounds in the inuent are generally composed of synthesisrelated chemicals and their derivatives. After the treatment by
anaerobic tank, most of organic compounds have decreased
slightly, such as alkanes, BHT, pyridine and butoxytrimethylSilane. Compared to the alkanes with higher molecular weights,
alkanes with low molecular weights were removed with relatively high efciency. Meanwhile, some new intermediates were
produced. For instance, several alkanes were probably converted to

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

421

Table 3
Main organic compounds present in the inuent and the efuent by GCMS analysis.
No

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
a

Organic compounds

Alkanes
Butylated hydroxytoluene
Pyridine
Tetramethylbutanedinitrile
Silane, butoxytrimethyl1,2-Benzenedicarboxylic acid, mono(2-ethylhexyl) ester
p-Xylene
Benzene, 1,3-dimethylAcetamide, 2,2,2-triuoro-N-(trimethylsilyl)Ethylbenzene
Acetic acid, (trimethylsilyl)Methane, dimethoxySulfur
Benzene, 1,2,4,5-tetramethylNaphthalene, decahydro-2-methylStyrene
Cyclotetrasiloxane, octamethylBis(trimethylsilyl) triuoroacetamide
Phosphorous acid, tris(decyl) este
2H-Benzotriazole, 2-ethylHexa(methoxymethyl) melamine
Silanamine, N-methoxy-1,1,1-trimethyl-N-(trimethylsilyl)Naphthalene, decahydro-2,6-dimethylMorpholine, 4-methyl3-Methyl-2-(2-oxopropyl) furan
Formamide, N,N-dimethylToluene
Propanoic acid, 2-methyl-, trimethylsilyl ester
1,4-Dioxane
d-Limonene
Bicyclo[410]hept-2-ene, 3,7,7-trimethylHexanoic acid, 2-ethylHydrazine, 1-methyl-1-(2-propenyl)
Benzene, 1-methyl-2-(1-methylethyl)Benzene, 1-ethyl-2,4-dimethylCyclohexene, 1-methyl-4-(1-methylethylidene)6-Tetradecanesulfonic acid, butyl ester
Tridecanol, 2-ethyl-2-methyl2-Ethylhexanoic acid, trimethylsilyl ester
Phenol, 3-(1,1-dimethylethyl)-4-methoxyDodecane, 4,6-dimethylDecane, 3,7-dimethylPentadecane
3-tert-Butyl-4-hydroxyanisole
2,4-Dimethoxy-N,N-dimethylbenzylamine
Others
Total

Total removal efciencya (%)

Relative mass percentage (%)

TA wastewater

A1 efuent

A2 efuent

MBR permeate

47.63
26.57
4.81
4.61
2.95
1.72
1.45
1.11
0.94
0.88
0.83
0.77
0.64
0.51
0.44
0.41
0.4
0.39
0.37
0.36
0.32
0.3
0.3
0.26
0.25
0.21
0.2
0.19
0.18

100

40.23
16.73
0.77
5.1
0.52
1.52
2.06
1.69

1.9
0.28
1.05
0.47
0.33

0.79

0.43
0.43

0.72

0.51
0.21

0.15
7.16
5.57
3.28
2.09
2.05
0.81
0.64
0.6
0.56
0.3

1.05
100

44.17
8.8

15.59

3.18
0.43

0.68

0.36

0.61
0.92

0.24

0.95

10.1

1.07

1.86
0.65
0.56
0.5
1.86
5.28
1.09
100

8.44
6.05

34.02

2.06
1.87

0.57

3.65

1.22
1.94

0.87

1.63

17.3

3.46
14.6
2.32
100

98
98
100
19
100
87
86
100
100
93
100
100
100
100
100
2
100
100
100
63
34
100
100
100
100
100
52
100
1

Total removal efciency was calculated according to the peak area under the same condition.

2-ethyl-hexanoic acid, and 6-tetradecanesulfonic acid, butyl ester,

and consistent with the general decrease of pH in the anaerobic
tank due to hydrolysis/acidication.
After the treatment in the anoxic tank, it can be seen from Table 3
that a considerable amount of benzene series organics and alkanes, which were hardly biodegraded at the anaerobic stage, were
degraded. Pyridine, butoxytrimethylSilane, 1,3-dimethyl benzene,
were completely removed. Furthermore, most of new intermediates produced from the anaerobic tank were consumed at
the anoxic stage, probably leading to a satisfactory denitrication performance. By contrast, some new compounds, such
as 3-(1,1-dimethylethyl)-4-methoxy-phenol, 2,4-dimethoxy-N,Ndimethylbenzylamine etc. were produced. Therefore, in order to
treat TA wastewater or other refractory wastewater, it is desirable to have an anaerobic unit as a pretreatment method for better
denitrication performance [24].
Most of the residual compounds in the anoxic efuents were
further degraded at the aerobic stage. In order to make a
quantitative comparison of the relative removal of organic
compounds from the TA wastewater, the areas of the peaks under

the chromatograms were determined under the same conditions.

It can be observed that, alkanes, the main components in TA
wastewater, were reduced by 98%, and only 6 types of alkanes
could be detected in the MBR efuent. Similarly, BHT, the subdominant component, was reduced by 98%, which is better than
that reported in municipal wastewater treatment using a combined system of UASB and two constructed wetlands (1030%) [25].
Although the majority of the organic pollutants were removed,
there were still a few refractory contaminants present in the nal
efuent. It seems that some compounds, such as TMSN, styrene and
hexa-methoxymethylmelamine, were quite stable in the A2 O-MBR
system and often used as organic indicators of industrial efuents
from certain chemical production industries [26]. In addition, 1,4Dioxane, a well-known hazardous pollutant, was hardly removed
in the system. Despite the fact that 1,4-Dioxane biodegradation
was achieved through bioaugmentation and the use of enriched
microbial pure cultures [27], this compound could not be oxidized effectively by conventional biological process due to its high
resistance to biotransformation, and could only be removed effectively by the advanced oxidation processes [28,29]. Suh and

422

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

Fig. 4. TMP evolution during 88-d continuous ltration operation in the MBR.

Mohseni [29] found that O3 /H2 O2 could be used to eliminate

1,4-Dioxane, and the oxidation rate was rst order at the dioxane concentrations lower than 50 mg/L. Therefore, much attention
should be paid attention to some refractory contaminants remaining in the nal efuent.

3.4. Membrane fouling behavior

The membrane fouling of two different kinds of membrane modules (HF and FS) were observed by monitoring TMP throughout the
operation of MBR. The variations of permeation ux and TMP are

Fig. 5. SEM micrographs showing the surfaces of clean membranes and fouled membranes. (a) New FS membrane surface, (b) fouled FS membrane surface, (c) new HF
membrane surface, (b) fouled HF membrane surface (FS = at sheet, HF = hollow ber).

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

423

Table 4
Components of membrane foulants measured by WD-XRF.
Element

Atomic ratio (%)

Clean FS
Fouled FS
Clean HF
Fouled HF

C
61.17
73.41
44.21
49.60

F
38.74
5.01
41.48
27.61

Si
0.01
0.46
0.01
0.07

S
0.01
0.19
0.01
0.10

Cl
0.03
0.06
8.24
0.02

Ca
0.01
0.12
5.01
0.12

presented in Fig. 4. During the operation, the average permeation

ux of HF and FS in the MBR system were 8.52 and 9.74 L/(m2 h),
respectively. As shown in Fig. 4, noticeable membrane fouling of
HF was rst observed on day 56 when the fouling was removed
by on-line chemical cleaning. Subsequently, the membrane fouling
of HF was observed on days 87 and 102. The trend of membrane
fouling of FS was very similar to that of HF. However, the TMP of
HF was always much higher than that of FS. On day 102, the TMP of
FS decreased almost to its initial level after the chemical cleaning,
implying that the fouling in the FS was smaller and more reversible
than fouling observed in the HF.
The cake layer was obviously observed on FS membrane surface,
while it was not evident for HF. After ushing the cake layer with tap
water, the gel layer was seen on both membranes. SEM images of
clean and fouled membranes were taken to determine the deposit
morphology of the gel surface. The membrane pores were clearly
seen on the new membranes (Fig. 5a and c), whereas, gel layers
were observed blocking the membrane pores on the fouled membranes and seemed to be more foulants accumulation (Fig. 5b and
d).
WD-XRF was also used to measure the chemical components
on the clean and fouled membrane surface. Table 4 shows that the
relative contents of C and N on the fouled membrane surfaces (HF
and FS) were much higher than those on clean membrane surfaces,
indicating that membrane fouling was mainly governed by organic
fouling. Moreover, several inorganic elements, such as Al, P, S, Si,
Ca and Mg, were found on the fouled membrane. It is well known
that metal ions may bridge the deposited cells and biopolymers,
resulting in a highly fouling layer and substantial TMP increase [30].
Although the relative contents of these elements were small, they
played an important role in fouled layer formation [31]. The comparison of chemical components between fouled HF and FS implied
that FS exhibited more foulants accumulation than HF on membrane surface. In fact, SEM of longitudinal direction of membrane
showed that the gel layer of FS and HF were about 60 m and 5 m
in thickness (Fig. S2). It could be deduced that gel layer and cake
layer were easily formed on FS membrane, and the cake layer could
protect the gel layer from being removed by aeration from below.
Fan and Huang [32] reported that the gel layer was indispensable
and played a key role in the dynamic MBR rejection capacity of
the ne particles. It was observed that small colloids and macromolecules may adsorb on the thin gel layer of the membrane and
result in severely clogged pores [33]. These might be the reason a
higher ux under lower TMP could be obtained in the FS membrane
module. Therefore, the chemical cleaning method using a combination of acids and oxidizing agents is recommended for fouled
membrane caused by organic foulants and inorganic compounds.

4. Conclusions
Results suggest that the A2 O-MBR system is very efcient in
simultaneous organic matter and nitrogen removal in the chemicalintensive TA wastewater treatment. An increase of internal recycle
ratio from 0.5 to 1.5 had a greater effect than an increase from
1.5 to 2.5 in the nitrogen removal performance. Some refractory compounds present in the MBR permeate, such as TMSN,

Ti
0.03
0.04

0.52

0.34

19.28

21.65

Na

0.10
1.02
0.21

Mg

0.12

0.04

Al

0.62

0.12

0.02
0.02
0.01

Fe

0.05

0.11

styrene, hexa-methoxymethylmelamine and 1,4-Dioxane, should

be discharged into municipal sewage treatment plant for further
treatment. The bridging between organic foulants and inorganic
compounds was a signicant contributor to membrane fouling,
and the combination of acids and oxidizing agents could be a good
option for chemical cleaning of fouled membrane.
Acknowledgements
This work was nancially supported by Fuji Electric Co. Ltd., and
by Open Project of Zhejiang Province Key Laboratory of Environmental Pollution Control Technology. We thank Ms. Lijuan Mao
the technician of 985-Institute of Agrobiology and Environmental Sciences of Zhejiang University, for the assistance in using
6890N/5975B GCMS system. The authors would like to thank
emeritus professor Goen Ho of Murdoch University for his contributions to improve the manuscript.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.jhazmat.
2015.01.031.
References
[1] B. Smith, Future Pollution Prevention Opportunities and Needs in the Textile
Industry, 1994.
[2] A.F. Bertea, R. Butnaru, Polyamide dyeing wastewater recycling after
Fenton-like oxidative treatment, Ind. Textila 63 (2012) 322326.
[3] S.F.A. Shahf, A.K. Shahf, A. Mehdi, A.A. Memon, K. Harijan, Z.M. Ali, Analysis
and treatment of industrial wastewater through chemical
coagulationadsorption process a case study of Clariant Pakistan Limited,
AIP Conf. Proc. 1453 (2012) 353358.
[4] Y.L. Pang, A.Z. Abdullah, Current status of textile industry wastewater
management and research progress in Malaysia: a review, Clean-Soil Air
Water 41 (2013) 751764.
[5] A. Chiavola, Textiles, Water Environ. Res. 81 (2009) 16961730.
[6] R. Aplin, T.D. Waite, Comparison of three advanced oxidation processes for
degradation of textile dyes, Water Sci. Technol. 42 (2000) 345354.
[7] A. Demirbas, Agricultural based activated carbons for the removal of dyes
from aqueous solutions: a review, J. Hazard. Mater. 167 (2009) 19.
[8] W.T. Zhao, X. Huang, D.J. Lee, X.H. Wang, Y.X. Shen, Use of submerged
anaerobicanoxicoxic membrane bioreactor to treat highly toxic coke
wastewater with complete sludge retention, J. Membr. Sci. 330 (2009) 5764.
[9] W.T. Zhao, Y.X. Shen, K. Xiao, X. Huang, Fouling characteristics in a membrane
bioreactor coupled with anaerobicanoxicoxic process for coke wastewater
treatment, Bioresour. Technol. 101 (2010) 38763883.
[10] T. Yu, R. Qi, D. Li, Y. Zhang, M. Yang, Nitrier characteristics in submerged
membrane bioreactors under different sludge retention times, Water Res. 44
(2010) 28232830.
[11] X. Zhou, Y. Li, Y. Zhao, X. Yue, Pilot-scale anaerobic/anoxic/oxic/oxic biolm
process treating coking wastewater, J. Chem. Technol. Biotechnol. 88 (2012)
305310.
[12] Y. Chen, C. Peng, J. Wang, L. Ye, L. Zhang, Y. Peng, Effect of nitrate recycling
ratio on simultaneous biological nutrient removal in a novel
anaerobic/anoxic/oxic (A2 /O)-biological aerated lter (BAF) system,
Bioresour. Technol. 102 (2011) 57225727.
[13] K.G. Song, J. Cho, K.H. Ahn, Effects of internal recycling time mode and
hydraulic retention time on biological nitrogen and phosphorus removal in a
sequencing anoxic/anaerobic membrane bioreactor process, Bioprocess
Biosyst. Eng. 32 (2009) 135142.
[14] Y.M. Li, G.W. Gu, J.F. Zhao, H.Q. Yu, Y.L. Qiu, Y.Z. Peng, Treatment of coke-plant
wastewater by biolm systems for removal of organic compounds and
nitrogen, Chemosphere 52 (2003) 9971005.

424

F. Sun et al. / Journal of Hazardous Materials 286 (2015) 416424

[15] T.W. Tan, H.Y. Ng, Inuence of mixed liquor recycle ratio and dissolved
oxygen on performance of pre-denitrication submerged membrane
bioreactors, Water Res. 42 (2008) 11221132.
[16] J.A. Baeza, D. Gabriel, J. Lafuente, Effect of internal recycle on the nitrogen
removal efciency of an anaerobic/anoxic/oxic (A2 /O) wastewater treatment
plant (WWTP), Process Biochem. 39 (2004) 16151624.
[17] Y.M. Kim, D. Park, C.O. Jeon, D.S. Lee, J.M. Park, Effect of HRT on the biological
pre-denitrication process for the simultaneous removal of toxic pollutants
from cokes wastewater, Bioresour. Technol. 99 (2008) 88248832.
[18] C.C. Tseng, T.G. Potter, B. Koopman, Effect of inuent chemical oxygen
demand to nitrogen ratio on a partial nitrication/complete denitrication
process, Water Res. 32 (1998) 165173.
[19] F.Q. Sun, S.W. Wu, J.J. Liu, B. Li, Y.X. Chen, W.X. Wu, Denitrication capacity of
a landlled refuse in response to the variations of COD/NO3 N in the
injected leachate, Bioresour. Technol. 103 (2012) 109115.
[20] M. Henze, G.H. Kristensen, R. Strube, Rate-capacity characterization of
wastewater for nutrient removal processes, Water Sci. Technol. 29 (1994)
101107.
[21] S.P. Sun, C.P.I. Nacher, B. Merkey, Q. Zhou, S.Q. Xia, D.H. Yang, J.H. Sun, B.F.
Smets, Effective biological nitrogen removal treatment processes for domestic
wastewaters with low C/N ratios: a review, Environ. Eng. Sci. 27 (2010)
111126.
[22] J. Carrera, T. Vicent, J. Lafuente, Effect of inuent COD/N ratio on biological
nitrogen removal (BNR) from high-strength ammonium industrial
wastewater, Process Biochem. 39 (2004) 20352041.
[23] I. Holme, Dyes and nishes for synthetic bres, Int. Dyer 189 (2004) 911.
[24] J. Wang, Z.J. Zhang, L.N. Chi, X.L. Qiao, H.X. Zhu, M.C. Long, Z.F. Zhang,
Performance of anaerobic process on toxicity reduction during treating
printing and dyeing wastewater, Bull. Environ. Contam. Toxicol. 78 (2007)
531534.

[25] C. Reyes-Contreras, V. Matamoros, I. Ruiz, M. Soto, J.M. Bayona, Evaluation of

PPCPs removal in a combined anaerobic digester-constructed wetland pilot
plant treating urban wastewater, Chemosphere 84 (2011)
12001207.
[26] O. Botalova, J. Schwarzbauer, N.a Sandouk, Identication and chemical
characterization of specic organic indicators in the efuents from chemical
production sites, Water Res. 45 (2011) 36533664.
[27] H.L. Huang, D.S. Shen, N. Li, D. Shan, J.L. Shentu, Y.Y. Zhou, Biodegradation of
1,4-Dioxane by a novel strain and its biodegradation pathway, Water Air Soil
Pollut. 225 (2014) 2135.
[28] J.Y. Choi, Y.J. Lee, J. Shin, J.W. Yang, Anodic oxidation of 1,4-dioxane on
boron-doped diamond electrodes for wastewater treatment, J. Hazard. Mater.
179 (2010) 762768.
[29] J.H. Suh, M. Mohseni, A study on the relationship between biodegradability
enhancement and oxidation of 1,4-dioxane using ozone and hydrogen
peroxide, Water Res. 38 (2004) 25962604.
[30] Q. Li, M. Elimelech, Organic fouling and chemical cleaning of nanoltration
membranes: measurements and mechanisms, Environ. Sci. Technol. 38
(2004) 46834693.
[31] F. Meng, H. Zhang, F. Yang, L. Liu, Characterization of cake layer in submerged
membrane bioreactor, Environ. Sci. Technol. 41 (2007)
40654070.
[32] B. Fan, X. Huang, Characteristics of a self-forming dynamic membrane
coupled with a bioreactor for municipal wastewater treatment, Environ. Sci.
Technol. 36 (2002) 52455251.
[33] I. Chang, S. Kim, Wastewater treatment using membrane ltrationeffect of
biosolids concentration on cake resistance, Process Biochem. 40 (2005)
13071314.