Available online at www.sciencedirect.

com

Journal of Hazardous Materials 152 (2008) 316–323

A membrane-based co-treatment strategy for the recovery

of print- and beck-dyeing textile effluents
G. Capar a,∗ , L. Yilmaz b , U. Yetis a
a Environmental Engineering Department, Middle East Technical University, 06531 Ankara, Turkey
b Chemical Engineering Department, Middle East Technical University, 06531 Ankara, Turkey

Received 16 April 2007; received in revised form 27 June 2007; accepted 28 June 2007
Available online 3 July 2007

Abstract
This paper describes the final part of a study on the recovery of print- and beck-dyeing wastewaters of the carpet manufacturing industry by
membrane processes. These wastewaters had been previously treated separately where the print dyeing wastewaters were recovered by chemical
precipitation followed by nanofiltration (NF) and beck-dyeing wastewaters were subjected to microfiltration (MF) and pH neutralization prior to
NF. In this study, a co-treatment scheme after separate pre-treatment stages was adopted to simplify the overall process. The effect of mixing ratio
on membrane fouling was also investigated. The co-treatment strategy was found advantageous since the number of NF units was minimized and
the pH neutralization step in separate treatment of beck-dyeing wastewaters was eliminated, providing a reduction of chemical usage.
© 2007 Elsevier B.V. All rights reserved.

Keywords: Carpet wastewater; Membrane; Wastewater mixing; Water reuse

1. Introduction end, membrane technology emerges as a unique solution and

has been gaining wider acceptance, as evidenced from the fact
Wastewater reclamation and reuse is part of an effective that it is one of the suggested treatment methods for the reclama-
environmental management program where advanced treatment tion of textile effluents in the Best Available Techniques (BAT)
methods such as filtration, UV treatment, carbon adsorption, Reference Document published by the European Union within
and membrane processes are adopted. Textile is the one of the framework of Integrated Pollution Prevention and Control
the most polluting industries, where non-renewable resources (IPPC) Directive (96/61/EC). The technical feasibility of mem-
are extensively used in wet processes such as dyeing, printing, brane technology has been proved; however the problem of flux
degumming and finishing. Textile production is characterized decline remains, which requires the application of a suitable pre-
by high consumption of fresh water with an average value of treatment method. In pre-treatment step, conventional methods
160 m3 t−1 of finished product and peaks above 500 m3 t−1 in like chemical precipitation, sand filtration and ozonation can
multi-stage processes [1]. In addition, the wastewaters generated be used as well as microfiltration (MF) and ultrafiltration (UF)
are quite complex and variable in nature due to the diversity [2,3]. After passing through one or more of these stages, the
of manufacturing processes. In textile industry, recycling of wastewater is further purified to the degree of reuse quality
the reclaimed water back into the most demanding processes in other pressure-driven membrane processes, i.e., nanofiltra-
requires that the wastewater is highly purified since the actual tion (NF) or reverse osmosis (RO). Hence, most approaches
process water is free of any color, turbidity and hardness due consist of two or more step processes. The required quality
to the filtration and softening processes applied on-site. To this for the reclaimed process water depends on the quality of the
final product, making it difficult to set some common qual-
ity criteria. Although a general guideline is provided for the
∗ Corresponding author at: Department of Engineering Sciences, Middle East desired reuse criteria for textile processes (Table 1), the water
Technical University, Inonu Bulvari, 06531 Ankara, Turkey.
Tel.: +90 312 2102389; fax: +90 312 2104462.
quality required for each particular type of textile process is
E-mail addresses: [email protected], currently not clear and needs to be established [4,5]. Regarding
[email protected] (G. Capar). the existence of several sub-sectors where wastewaters of highly

0304-3894/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jhazmat.2007.06.100
 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323 317

Table 1 for combined biological treatment in a fluidized bed where an

Water reuse criteria for textile processes in general improved biodegradation via co-treatment was observed. Bajza
Parameter Reuse criteria and Vrcek [8] also studied the mixing of wastewaters from
COD (mg L−1 ) 60 [5]–80 [4]
chrome tanning process and liming of bovine leather, where
Color (Pt–Co) none [5] they concluded that co-treatment provided the advantage of sat-
Turbidity (NTU) 1 [4] isfying the ecological parameters, eliminating the pre-treatment
Suspended solids (mg L−1 ) 5 [4] steps, and reduce the material costs for neutralization of wastew-
Dissolved solids (mg L−1 ) 500 [6] aters. On the other hand, Chmiel et al. [9] has drawn attention to
Total hardness (mg L−1 as CaCO3 ) 25 [4]–50 [5]
Conductivity (␮S cm−1 ) 1000 [4]
possible problems in biological treatment due to the mixing of
pH 6–8 [4] spent process waters with high and low strengths in food and bev-
erage industry. They offered two different treatment strategies;
first, separate treatment of partial streams where product recov-
variable flow rates and characteristics are generated, each partic- ery was achieved by treatment of the highly polluted pre-rinsing
ular textile effluent has to be handled individually. In addition, water in margarine production, and second, combined treatment
the requirement of longer process trains points out the disad- of different spent process waters of a fruit juice company for
vantage of increased treatment costs. To this end, there is a water recovery through two-stage NF. The same approach can
need to determine the simplest and most energy efficient process be used in water reclamation from textile industry, where the
combinations. wastewaters from different dyeing and rinsing process streams
In searching for the simplest and most efficient process com- may be treated in separate or combined routes depending on
binations, mixing of wastewaters for co-treatment may be one the performance of the treatment process applied and the reuse
solution, which would minimize the number of units required, quality desired. In adopting membrane technology for the recla-
thus enabling cost effectiveness. It may also create a synergy in mation of textile wastewaters, both approaches have been used,
terms of improved treatment performance. In one study, a syn- i.e., membrane processes either follow the existing biological
ergy was sought by Ochieng et al. [7] by mixing wastewaters treatment plants for the wastewater mixtures [10] or they are
from two different industries, namely petroleum and brewery, integrated into individual wastewater streams, enabling direct

Fig. 1. Carpet manufacturing process scheme (Q represents the daily wastewater flow rate, dashed lines show wastewater streams).
318 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323

Table 2
Ingredients of print-and beck-dyeing wastewaters
Aim of use Dyestuff/auxiliary chemical

Print-dyeing wastewater Beck-dyeing wastewater

Adjust pH Citric acid Acetic acid

Adjust viscosity Tanaprint ST 160 Conc. (anionic ammonium salt, high –
electrolyte-resistant synthetic thickener)
Increase wetting Tanasperse CJ (non-ionic penetrant for producing –
homogeneous paste)
Control foaming – Foamaster WWT
Give color Yellow 2R, Yellow 4GN, Blue 2RA (Acid Blue 781), Blue Yellow 2G (Acid Yellow 169), Red 2B (Acid Red 361), Red
5G (Acid Blue 526), Black B (Acid Black 1051), Red G 599, Orange 4G, Blue 4R (Acid Blue 277), Blue 608b
(Acid Red 78), Gray G, Navy R (Acid Blue 176)a
a CIBA Lanaset (metal-complex) dyes.
b CIBA Tectilon (acid) dyes.

recycling of water into the process itself [11]. In this regard, it Although PDW and BDW were recovered successfully in
is found meaningful to study the effect of mixing wastewaters separate routes, the combined treatment of these two wastew-
from different streams on the determination of the best process aters was adopted in order to further optimise the water reuse
train for the recovery of textile effluents. To this end, the aim scheme. In this combined treatment scheme, PDW and BDW
of this study is to evaluate the effect of mixing of wastewaters were mixed in the NF stage after passing through separate pre-
generated from different textile dyeing processes for the opti- treatment steps. Two mixing ratios were applied, i.e., PDW and
mization of the water reuse scheme towards the requirement of BDW were mixed at a volumetric ratio of 4/1 and 1/1, which are
least number of units. For this, technical evaluation and compari- expressed as “PDW/BDW = 4/1” and “PDW/BDW = 1/1”. The
son of two alternative membrane-based treatment schemes were mixing ratio scenario was based on the fact that the actual volu-
performed for the dye-house effluents of a carpet manufacturing metric ratio of PDW to BDW in the carpet manufacturing plant
plant, which represent a significant sub-sector of textile industry is 4/1 (Fig. 1), and 1/1 is the highest possible ratio in the plant
in Turkey. In addition, the effect of mixing ratio on membrane since the capacity of printing is higher than the capacity of beck
fouling was also investigated in order to evaluate the efficiency dyeing, leading to the generation of PDW with higher flow rates.
of co-treatment. The effect of pH on NF separation performance was also studied
since the wastewater pH had previously been observed to highly
2. Materials and methods affect the NF separation performance where the organic matter
rejection significantly improved for BDW at neutral pH [14].
2.1. Treatment strategies Therefore, in this study, the rejection performances were eval-
uated at two pH levels; first at the original pH of the mixtures,
The major wastewater generating activities in carpet dye- which were 6.7–6.9 when chemically precipitated PDW having
house of interest are briefly summarized in Fig. 1. Although pH 7.2 was mixed with pre-filtered BDW having pH 5.3–5.9.
the wastewater generation rates are highly variable with time Second, the pH of the mixtures was reduced to pH 5.7, i.e.,
due to changing demands, the given wastewater generation rates the original pH of BDW. In this way, the possibility of having
represent average values. The print- and beck-dyeing processes increased volumetric ratios of BDW in the mixture was tried
generate the highest fractions of wastewaters in the plant, where to be simulated, because increasing the fraction of BDW in the
the flow rate of print-dyeing wastewater (PDW) is four times mixture would decrease the pH of the mixture towards its acidic
higher than that of beck-dyeing wastewater (BDW) (Fig. 1). In pH. The combined process train provided the advantage of omit-
addition, PDW and BDW contain different dyes and auxiliary ting the pH neutralization step for BDW, which was required in
chemicals (Table 2). separate treatment route, and therefore helped minimize chemi-
The PDW and BDW had been exposed to separate process cal usage in the plant. Because PDW already had neutral pH, it
trains at the points of wastewater generation (Fig. 2(a)). Alter- was thought that the acidic pH of the BDW would naturally rise
natively a combined process train was adopted (Fig. 2(b)). In to neutral level if mixed with PDW. Therefore, the co-treatment
separate process trains, PDW was first chemically precipitated of BDW and PDW in the same process train was considered
with an alum dose of 250 mg L−1 and then treated in single NF to be very practical, as it would provide the recovery of these
up to the degree of reuse quality, whereas BDW was pre-filtered wastewaters in the simplest process train possible.
through a microfilter (MF) having a pore size of 1.0 ␮m and
then recovered in single NF after pH neutralization. These sep- 2.2. Sample
arate process trains had been developed by evaluating several
alternatives, the details of which were presented in our previous The print-dyeing wastewater stream was sampled four times
publications [12–15]. and a mixture was formed for PDW, which contained wastew-
BDW 2.
mixed with BDW 1, and in 4/1 mixture, PDW was mixed with
PDW and BDW are shown in Table 3. In 1/1 mixture, PDW was
teristics of the new wastewater samples obtained by mixing
The characteristics of the pre-treated samples and the charac-
observed in the samples kept for the longest period of time.
period of the study due to bacterial/fungal growth, which was
and BDW 1 increased gradually with time during the time
kept at 278 K in a cold room. However the pH of the PDW
ual sample belonging to a single carpet. All the samples were
wastewaters from different carpets and BDW 2 is an individ-
ples were obtained, where BDW 1 is the mixture of two
other hand, two different BDW (BDW 1 and BDW 2) sam-
aters from the washing step of four different carpets. On the

Fig. 2. Water reuse schemes for (a) separate treatment, (b) combined treatment.

G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323

Table 3
Characteristics of raw, pre-treated, and mixed samples after pre-treatment
Parameter Mean value ± standard deviation (Percent removal)*

Quality of raw wastewaters Quality of pre-treated wastewaters Quality of wastewaters mixed after pre-treatment

PDW BDW 1 BDW 2 PDW BDW 1 BDW 2 PDW/BDW 2 = 4/1 PDW/BDW 1 = 1/1

pH 7.2 ± 0.0 4.7 ± 0.0 5.3 ± 0.0 6.8 ± 0.0 5.0 ± 0.0 5.3 ± 0.0 6.9 ± 0.0 5.7 ± 0.0a 6.7 ± 0.0 5.7 ± 0.0b
COD (mg L−1 ) 852 ± 10 1407 ± 22 1928 ± 30 423 ± 10 (50) 1252 ± 5 (11) 1832 ± 7 (5) 782 ± 6 691 ± 32 816 ± 9 877 ± 13
Color (Pt–Co) 706 ± 0 105 ± 0 50 ± 0 73 ± 2 (90) 43 ± 1 (59) 36 ± 1 (28) 44 ± 6 97 ± 1 35 ± 3 34 ± 4
Turbidity (NTU) 72.0 ± 0.7 5.4 ± 0.0 3.3 ± 0.0 7.1 ± 0.1 (90) 1.4 ± 0.1 (74) 3.2 ± 0.0 (3) 5.2 ± 0.1 9.5 ± 0.1 1.8 ± 0.4 1.8 ± 0.4
Total solids (mg L−1 ) 678 ± 9 1019 ± 4 1323 ± 23 595 ± 7 (12) 968 ± 6 (5) 1204 ± 88 (9) 663 ± 7 751 ± 18 765 ± 7 797 ± 35
Total hardness (mg L−1 as CaCO3 ) 33 ± 1.4 44 ± 0 18 ± 0 33 ± 0 (0) 40 ± 0 (9) 18 ± 0 (0) 40 ± 0 34 ± 0 40 ± 0 40 ± 0
Conductivity (␮S cm−1 ) 697 770 899 793 (-) 762 (1) 806 (10) 883 844 755 777
* Values in parenthesis represent the percent removals in the pre-treatment stage.
a pH of feed was reduced from 6.9 to 5.7 and kept constant at 5.7 by adding HNO3 .
b pH of feed was initially reduced from 6.7 to 5.7 by adding HNO3 .

319
320 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323

2.3. Nanofiltration

0.1 ± 0.0 (94)

235 ± 16 (73)
0 ± 0 (100)

146 ± 14 (82)
0 ± 0 (100)
Feed pH 5.7e
NF experiments were carried out by a lab-scale plate and

205 (74)
6.0
frame module, LabStak M20 (product of DSS Company) in
cross-flow operation where NFT-50 membranes were used. The

PDW/BDW 1 = 1/1
experiments were carried out in total recycle mode of filtra-

0.1 ± 0.0 (94)

tion where permeate and retentate were both returned to the

0 ± 0 (100)

36 ± 11 (95)
0 ± 0 (100)
Feed pH 6.7

17 ± 5 (98)
feed tank. The NFT-50 is a thin film composite membrane con-

71 (91)
sisting of three layers: an ultra-thin polyamide barrier layer, a

6.4
microporous polysulfone interlayer and a high strength polyester
support. It is a hydrophilic membrane with a contact angle of 40◦

0.2 ± 0.0 (98)

0 ± 0 (100)

197 ± 16 (74)
0 ± 0 (100)
(Wilhemy Method). The molecular weight cut off (MWCO) of

5.7d

68 ± 1 (90)
the NFT-50 membrane is not reported by the manufacturer how-

Feed pH

405 (52)
ever its MgSO4 rejection is ≥99%. Four NF membranes, each

5.8
having 0.018 m2 of effective area, were used in two pairs, mak-

Combined treatment

PDW/BDW 2 = 4/1
ing a total effective membrane area of 0.072 m2 . The NF test

0.3 ± 0.0 (96)

The original pH of PDW and BDW 1 increased from 6.8 to 7.2 and from 5.0 to 5.9, respectively due to fungal/bacterial growth during storage at 4 ◦ C.
was carried out at a trans-membrane pressure (TMP) of 590 kPa

0 ± 0 (100)
25 ± 4 (97)
3 ± 1 (93)

136 ± 3 (79)
Feed pH 6.9
at an operation temperature of 291 K. The samples were fed to

188 (79)
the system at a flow rate of 360 L h−1 . Permeates were collected

6.6
when the flux declines reached steady state and analysed for
their constituents.

0.2 ± 0 (94)
0 ± 0 (100)

106 ± 17 (91)
0 ± 0 (100)
7.1c

55 ± 2 (97)
2.4. Flux decline analysis

Feed pH

100 (88)
6.7

Since there are no suspended solids in NF permeate samples, total solids measurements also represent dissolved solids.
The flux measurements were performed initially with the
clean membrane (Jcwi ), then with wastewater (Jww ) at steady

0.2 ± 0.0 (94)

state conditions, followed by the clean water flux of the fouled

0 ± 0 (100)

535 ± 10 (56)
Permeate quality (mean value ± standard deviation (Percent removal)* )

415 ± 8 (77)

2 ± 0 (89)
Feed pH 5.3
membrane (Jcwf ). Finally the clean water flux of the membrane

530 (34)
BDW 2

was measured after chemical cleaning (Jcwc ). The details of flux 5.4

Feed pH was initially decreased from 6.7 to 5.7 by HNO3 and was not kept constant during filtration.
analysis are described elsewhere [14].

Feed pH was decreased from 6.9 to 5.7 and kept constant at 5.7 by constant addition of HNO3 .
2.5. Membrane chemical cleaning
7.2b

0.2 ± 0 (86)
0 ± 0 (100)

60 ± 31 (94)
0 ± 0 (100)
36 ± 4 (97)
Feed pH

The NFT-50 membranes were cleaned by clean-in-place 72 (91)

method before the first use and after the experiments. Two solu-
6.9

tions, namely HNO3 solution at pH 3 and NaOH solution at pH

9–10 were circulated in the system consecutively under a low
0.1 ± 0.0 (93)
272 ± 23 (78)
0 ± 0 (100)

0 ± 0 (100)
5.9a

TMP of 180 kPa (gauge). Each cleaning solution was circulated

247 ± 7 (74)

for 30 min in order to remove the organic and inorganic precip-

Feed pH

226 (70)
BDW 1

6.4

itates from the surface of the membranes. Flushing with clean

water between the two cycles was also carried out. The retentate
NF performance in separate and combined treatment schemes

Feed pH was increased from 5.9 to 7.2 by adding NaOH.

Feed pH was increased from 5.3 to 7.1 by adding NaOH.
Separate treatment

and permeate were continuously discarded. The cleaned mem-

0.2 ± 0.0 (97)

Values in parenthesis represent the percent removals.

0 ± 0 (100)

75 ± 18 (87)
0 ± 0 (100)

branes were always kept wet in 0.25% sodium bisulfite solution

7.2a

19 ± 2 (96)

at 277 K in order to avoid bacterial growth on the membranes.

Feed pH

106 (87)
8.7
PDW

2.6. Analytical techniques

All the analyses were performed according to the Standard

Total hardness (mg L−1 as CaCO3 )

Methods [16], except COD, which was measured following

USEPA approved HACH Method 8000. COD analysis was per-
formed by using HACH DR-2000 Model spectrophotometer
Conductivity (␮S cm−1 )
Total solidsf (mg L−1 )

at wavelengths of 620 (high range) and 420 nm (low range),

respectively. Color measurements were performed by the same
Turbidity (NTU)
COD (mg L−1 )
pH (permeate)

Color (Pt–Co)

instrument, which was already calibrated for color measurement

Parameter

in terms of Pt–Co at a wavelength of 455 nm. Turbidity was mea-

Table 4

sured with a HACH Model 2100A turbidimeter. Total solids

*
a
b
c
d
e
f

content of the samples were determined by gravimetric analy-

 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323 321

Table 5
Ranking of all the permeate qualities for COD, total solids and conductivity
Treatment scheme Feed pH Ranking with respect to permeate quality

COD Total solids Conductivity

Separate PDW 7.2 2 3 4

BDW 1 5.9 8 8 7
BDW 1 7.2 4 2 2
BDW 2 5.3 9 9 9
BDW 2 7.1 5 4 3
Combined PDW/BDW = 4/1 6.9 3 5 5
PDW/BDW = 4/1 5.7 6 7 8
PDW/BDW = 1/1 6.7 1 1 1
PDW/BDW = 1/1 5.7 7 6 6

sis. The pH of the NF feed samples was monitored by a WTW parameter of interest. Color, turbidity and total hardness were
pH 330 Model portable pH meter. not included in the comparison since they were highly removed
in all cases. Although the ranking changed among the pollutant
3. Results and discussion parameters and the feed types, a general trend was observed such
that the permeate quality for all parameters was the best for the
3.1. Effect of wastewater mixing on the co-treatment mixture of PDW and BDW at a volumetric ratio of 1/1 and pH
efficiency of 6.7 in the combined treatment scheme. On the other hand,
the lowest ranking belonged to BDW, which was not exposed to
In order to evaluate the performance of co-treatment, the pH neutralization, and hence treated at its original acidic pH in
separate and combined treatment of PDW and BDW were com- the separate treatment scheme. In the second best place, there
pared in Table 4. In separate treatment of these wastewaters, it was PDW and BDW exposed to pH neutralization in separate
was observed that the permeate of print dyeing wastewaters had treatment schemes. The combined treatment scheme at lowered
quite good quality and satisfied the reuse criteria (Table 1). Sim- pH resulted in the second worst permeate quality. These ranking
ilarly, pH neutralization provided significant improvement in results obviously reveal that combined treatment at neutral pH
the rejection performance for COD and total solids for the beck- produces the highest permeate qualities. The choice of the water
dyeing wastewaters to reach the specified reuse criteria (Table 1). reuse scheme might be done taking into account the purpose of
The effect of pH on membrane performance has been studied reuse at the factory of interest, such as rinsing, dyeing of dark col-
[17–19]. Zeta potential, which is used to quantify the membrane ors, or the very light colors, since each process requires different
surface charge, has been observed to become increasingly more water qualities. However, the optimized scheme offered for the
negative for most membranes as pH is increased [20]. Accord- combined treatment of PDW and BDW at neutral pH, provides
ingly, increasing the surface charge of the membrane would the production of the highest quality water, thus ensures the sat-
result in increased electrostatic repulsion between a negatively isfaction of all quality requirements. Furthermore, the combined
charged solute and membrane. Hence, the improved membrane treatment scheme suggested elimination of the pH neutralization
rejection via pH neutralization for BDW was attributed to a step for separate treatment of BDW, which in turn helped mini-
possible change of the membrane surface charge. At acidic pH mize the number of units in the process train (Fig. 2). This result
values, NF membranes are expected to lose their surface charge needed to be verified in terms of flux declines. Therefore, the
and become almost neutral. Thus, low COD rejection at acidic effect of combining the treatment schemes on the flux declines
pH can be speculated as the passage of acetate ions, which were was also investigated (Table 6).
present due to the addition of acetic acid at the dyeing stage,
to the permeate site possibly by sieving mechanism. When the 3.2. Effect of co-treatment on the membrane flux declines
pH was neutralized, the membrane surface became negatively
charged causing higher rejection of acetic acid probably due to The relative fluxes (Jww /Jcwi ) for separate and combined
the repulsive forces. treatment schemes are depicted in Figs. 3 and 4, respectively.
Color, turbidity and total hardness were all removed very effi- In separate treatment, the flux declines at the original pH levels
ciently for all the feed types in both treatment schemes. All the of the wastewaters were 19%, 20.4% and 15.4% for PDW, BDW
permeate samples were very soft, free of any color and had negli- 1 and BDW 2, respectively (Fig. 3). Although the pH neutraliza-
gible turbidity. In order to evaluate these nine possible treatment tion did not affect the flux decline of BDW 1, which remained at
schemes, all the permeate qualities were ranked from 1 to 9 with 20%, it caused an increased flux decline, i.e., 24.5% for BDW 2.
respect to COD, total solids and conductivity removal efficien- All the flux declines were reversible in separate treatment, where
cies for separate and combined treatment schemes (Table 5). In 100–108% flux recovery was achieved via chemical cleaning
other words, the treatment scheme, which provided the best per- (Table 6). The recovery ratio being greater than 100% for the
meate quality, was ranked as 1st with respect to each specific neutralized BDW samples may indicate opening of the mem-
322 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323

(Jcwi − Jcwf )/Jcwi

Total

12.7

17.8
4.5

12.3
6.4

16.1
15.9

10.6
14.8
(Jcwi − Jcwc )/Jcwi
Irreversible

3.1

5.0

10.2
0

0
a

a
(Jcwc − Jcwf )/Jcwc
Reversible
Fouling

12.7

17.8

9.5

11.6
15.9

5.2
a

Fig. 3. Relative flux decline for PDW and BDW at the original pH values.

brane pores due to chemical exposure when NaOH was added

Concentration polarization

for pH adjustment.
In combined treatment, the same flux decline of 24.3% was
(Jcwf − Jww )/Jcwf

observed for mixing ratios of 4/1 and 1/1 at the original pH of the
mixtures, which were 6.9 and 6.7, respectively. Decreased flux
6.1

3.2
16.2

3.5
19.3

9.8
3.8

15.4
9.0

declines were observed at lowered pH, i.e., 19% for 4/1 mixture
and 22.5% for 1/1 mixture at pH 5.7 (Fig. 4). Although not
complete, the flux declines observed in combined treatment were
(Jcwi − Jww )/Jcwi
Flux decline (%)

recovered to a great extent, i.e., 90–97% (Table 6). Complete

a These data could not be calculated since the clean water fluxes obtained after cleaning were greater than the initial values (due to opening of the pores).

restoration of fluxes may be achieved by increasing the duration

of cleaning. Indeed, the flux declines slightly varied, i.e., from
Total

19.0

20.4
20.0

15.4
24.5

24.3
19.0

24.3
22.5

15.4% to 24.5% in all cases, which implies that co-treatment of

these wastewaters did not cause significant deterioration of the
flux declines.
Flux recovery (%)

The flux declines caused by concentration polarization and

(Jcwc /Jcwi )

fouling were also compared for all cases (Table 6). In separate
treatment, fouling was dominant for the flux declines of PDW
100

100
108

97
107

95
100

108
90

and BDW at acidic pH, however it was reversible to a great

extent. On the other hand, the effect of concentration polariza-
tion on the flux decline was dominant for BDW at neutralized pH.
Wastewater

In combined treatment, fouling was dominant for mixing ratio

(Jww )

15.3

18.3
17.6

16.5
14.2

16.5
15.3

16.5
18.3

of 4/1 at both pH levels as in the case of separate treatment of

PDW. This result is expected since the volumetric ratio of PDW
in the mixture is four times higher than that of BDW, and there-
Cleaned
(Jcwc )
18.9

23.0
23.8

18.9
20.1

20.7
18.9

23.6
21.2
Effect of combining the treatment schemes on flux decline

(Jcwf )
Final

16.5

18.9
21.0

17.1
17.6

18.3
15.9

19.5
20.1
Flux (L m−2 h)
Clean water

(Jcwi )
Initial

18.9

23.0
22.0

19.5
18.8

21.8
18.9

21.8
23.6
7.2

5.9
7.2

5.3
7.1

6.9
5.7

6.7
5.7
pH

PDW/BDW2 = 4/1

PDW/BDW1 = 1/1
Table 6

BDW 1

BDW 2
Sample

PDW

Fig. 4. Effect of pH on flux decline for mixtures of PDW and BDW.

 G. Capar et al. / Journal of Hazardous Materials 152 (2008) 316–323 323

fore PDW characteristics is more pronounced. The reversible [7] A. Ochieng, J.O. Odiyo, M. Mutsago, Biological treatment o mixed indus-
fraction of fouling was higher than its irreversible fraction, and trial wastewaters in a fluidized bed reactor, J. Hazard. Mater. B96 (2003)
furthermore, all the fouling became reversible at pH 5.7, which 79–90.
[8] Z. Bajza, I.V. Vrcek, Water quality analysis of mixtures obtained from
was also confirmed with the effectiveness of cleaning. In the tannery waste effluents, Ecotoxicol. Environ. Saf. Environ. Res. Sec. B 50
case of 1/1 mixing ratio, concentration polarization was higher (2001) 15–18.
at pH 6.7 than at pH 5.7, which means fouling became dominant [9] H. Chmiel, M. Kaschek, C. Blöcher, M. Noronha, V. Mavrov, Concepts for
at acidic pH. These results resembled those obtained for BDW the treatment of spent process water in the food and beverage industries,
alone, i.e., the effect of concentration polarization became much Desalination 152 (2002) 307–314.
[10] A. Bes-Pia, J.A. Mendoza-Roca, A. Roig-Alcoyer, A. Iborra-Clar, M.I.
higher at neutral pH. This is also expected since the BDW char- Iborra-Clar, M.I. Alcaina-Miranda, Comparison between nanofiltration and
acteristics became more pronounced when it was mixed with ozonation of biologically treated textile wastewater for its reuse in the
PDW at equal volumes. industry, Desalination 157 (2003) 81–86.
[11] B. van der Bruggen, G. Cornelis, C. Vandecasteele, I. Devreese,
4. Conclusions Fouling of nanofiltration and ultrafiltration membranes applied for
wastewater regeneration in the textile industry, Desalination 175 (2005)
111–119.
The co-treatment strategy was applied for the dye-house [12] G. Capar, U. Yetis, L. Yilmaz, Membrane based strategies for the pre-
effluents of the carpet manufacturing industry. The separate treatment of acid dye bath wastewaters, J. Hazard. Mater. B135 (2006)
water reuse schemes offered for print-dyeing and beck-dyeing 423–430.
wastewaters were combined into a single route by mixing these [13] G. Capar, U. Yetis, L. Yilmaz, Reclamation of printing effluents of a carpet
manufacturing industry by membrane processes, J. Membr. Sci. 277 (1–2)
wastewaters after separate pre-treatment stages. The comparison (2006) 120–128.
of permeate qualities and the flux declines revealed that print- [14] G. Capar, L. Yilmaz, U. Yetis, Reclamation of acid dye bath wastewater:
dyeing wastewaters and beck-dyeing wastewaters can be treated effect of pH on nanofiltration performance, J. Membr. Sci. 281 (2006)
together to the degree of reuse quality in a single NF unit even 560–569.
when mixed at equal volumes. This conclusion is considered [15] G. Capar, U. Yetis, L. Yilmaz, The most effective pre-treatment to nanofil-
tration for the recovery of print dyeing wastewaters, Desalination 212
to be significant in terms of the management of these wastewa- (2007) 103–113.
ters, since the co-treatment is expected to be less costly than the [16] A.D. Eaton, L.S. Clesceri, A.E. Greenberg (Eds.), APHA/AWWA/WEF
separate treatment due to the requirement of less NF units and Standard Methods for the Examination of Water and Wastewater, 19th ed.,
elimination of the pH neutralization step, which is required in American Public Health Association, Washington DC, USA, 1995.
the separate treatment of beck-dyeing wastewaters. [17] A.E. Childress, M. Elimelech, Relating nanofiltration membrane perfor-
mance to membrane charge (electrokinetic) characteristics, Environ. Sci.
Technol. 34 (2000) 3710–3716.
Acknowledgement [18] H. Ozaki, H. Li, Rejection of organic compounds by ultra low pressure
reverse osmosis membrane, Water Res. 36 (1) (2002) 123–130.
TUBITAK (The Scientific and Technological Research [19] J. Qin, M.H. Oo, H. Lee, B. Coniglio, Effect of feed pH on permeate pH
Council of Turkey) is gratefully acknowledged for the financial and ion rejection under acidic conditions in NF process, J. Membr. Sci. 232
(2004) 153–159.
support via Grant no. MISAG-214. [20] S. Lee, G. Park, G. Amy, S.K. Hong, S.H. Moon, D.H. Lee, J. Cho, Deter-
mination of membrane pore size distribution using the fractional rejection
References of non-ionic and charged macromolecules, J. Membr. Sci. 201 (2002)
191–201.
[1] EPA Report. Best management practices for pollution prevention in the
textile industry. EPA Manual 625/R-96-004, 1996.
[2] A. Bes-Pia, J.A. Mendoza-Roca, M.I. Alcaina-Miranda, A. Iborra-Clar, Glossary
M.I. Iborra-Clar, Combination of physico-chemical treatment and nanofil-
tration to reuse wastewater of a printing, dyeing and finishing textile
BDW: Beck dyeing wastewater
industry, Desalination 157 (2003) 73–80.
COD: Chemical oxygen demand (mg L−1 )
[3] M. Marcucci, I. Ciabatti, A. Matteucci, G. Vernaglione, Membrane tech-
Jcwc : Clean water flux of cleaned membrane (L m−2 h)
nologies applied to textile wastewater treatment, Ann. N.Y. Acad. Sci. 984
Jcwf : Clean water flux of fouled membrane (L m−2 h)
(2003) 53–64.
Jcwi : Clean water flux of clean membrane (L m−2 h)
[4] L. De Florio, A. Giordano, D. Mattioli, Nanofiltration of low-contaminated
textile rinsing effluents for on-site treatment and reuse, Desalination 181 Jww : Wastewater flux (L m−2 h)
MF: Microfiltration
(2005) 283–292.
MWCO: The molecular weight cut off
[5] B. van der Bruggen, K. Boussu, I.D. De Vreese, G. van Baelen, F. Willemse,
NF: Nanofiltration
D. Goedeme, W. Colen, Industrial process water recycling: principles and
PDW: Print dyeing wastewater
examples, Environ. Prog. 24 (4) (2005) 417–425.
RO: Reverse osmosis
[6] BTTG, British Textile Technology Group, Report 5: Waste Mini-
TMP: Trans membrane pressure (kPa)
mization and Best Practice, British Textile Technology Group, 1999.
UF: Ultrafiltration
http://www.e4s.org.uk/textilesonline/content/6library/fr library.htm.