WEFTEC 2014

Effect of Filtration using Sand Filter or MBR on Ozone Requirement for Color Removal in
Textile Wastewater

*Malcolm Fabiyi1, Karen A. Connery1, Andre Rambor2

1
Praxair, Inc. 7000 High Grove Boulevard, Burr Ridge, IL, USA.
2
White Martins Gases Industriais (WMGI), Sao Paulo, SP, Brazil

*Corresponding Author: Malcolm_fabiyi@praxair.com

Abstract
The effect of filtration on ozone requirement for color removal in biologically treated textile
wastewater effluent was investigated using two separate filtration methods. Filtered effluents
were obtained from an ultrafiltration membrane and a sand filtration system. Ozonation was
carried out in bubble column reactor tests to compare the efficiency of ozone tertiary color
removal for the base case of conventional activated sludge (CAS) gravity clarified effluent, to
sand filtered CAS effluent and immersed membrane bioreactor (iMBR) effluent. The ozone
requirement to achieve the 200 mg/L platinum-cobalt color units (CU) target for the iMBR
system was 26% that of the CAS system. The requirement for the sand filtered effluent was 42%
of the CAS effluent. When reuse of the ozone treated water within the facility was not
considered, no technology provided a clear cost advantage relative to the base case. Given that
the membrane treated effluent is the only one capable of being used for any grade of water reuse,
the inclusion of water recovery and reuse makes the membrane filtered system a compelling
alternative.

Key words: Textile, Dye; High purity oxygen; Ozonation

Introduction
Ozonation has been shown to be an effective tertiary treatment for residual color in biologically
treated textile wastewater (Adams, 2003, Vandevivere et al, 1998). The activated sludge
wastewater treatment process is effective for the treatment of textile wastewater. While the
method can provide up to 90% removal of chemical oxygen demand (COD), only about 40% -
60% of the color removal is attributed to biological treatment (Rambor & Fabiyi, 2012; Avsar &
Batibay, 2010). Cost savings can be substantial compared to physical/chemical separation of
color due to a reduction in chemical costs and the elimination of the resulting chemical sludge
(Rambor & Fabiyi, 2012).

The cost of ozone tertiary treatment includes a capital investment in ozone generation and
contacting equipment and operating costs of power and oxygen. The magnitude of these costs is
dependent on the amount of ozone that is required. The ozone demand is in turn dependent on
the quantities of suspended bacterial solids as well as colloidal or dissolved organic material
present in the effluent. Effluents from the secondary clarifier of conventional activated sludge
treatment systems typically still have some residual suspended solids (10-50 mg/L). In the case
of wastewater with high turbidity, higher ozone dosages may be required due to quenching of
ozone with substances in the wastewater other than the desired chromophore target.

Copyright ©2014 Water Environment Federation

3053
 WEFTEC 2014

MBR systems are becoming more prevalent due to the fact that they produce a high effluent
quality and occupy a relatively small footprint, compared to gravity settling based alternatives.
With pore sizes in the range of 3 µm to 0.02 µm, micro and ultra-filtration systems are superior
to gravity clarification in their purification capacity, providing water that is both disinfected and
reduced in turbidity. Full scale systems have been implemented for the treatment and
purification of textile wastewater (Judd, 2011).

The appearance of color in water is due to absorption of light in the visible range and occurs for
molecules containing conjugated structures such as aromatic rings (Langlais, 1991). This is the
case for reactive dyes, an example of which is Remazol Black B as shown in Figure 1. The
decolorizing action of Ozone mainly occurs via the oxidative cleavage of these chromophores
(Matsui, 1996).

Figure 1. Example reactive dye, CAS 17095-24-8 (SciFinder, 2013)

Objectives
This study compared two filtration techniques on biologically treated textile wastewater to
determine the required ozone dose and subsequent treatment cost. The two techniques used were
sand filtration following conventional activated sludge treatment and effluent from an immersed
membrane bioreactor (iMBR).

Figure 2: Schematic of the conventional Activated Sludge System with location of clarified
effluent and sand Filtered Effluent samples

Copyright ©2014 Water Environment Federation

3054
 WEFTEC 2014

Figure 3: Schematic of the membrane bioreactor treated wastewater process

Methodology
Tests were conducted with effluent from an existing 200 m3/hr textile wastewater treatment plant
employing gravity clarification (see schematic in Figure 2) and effluent from a 0.5m3/hr iMBR
Pilot system (Figure 4) located at the same facility. The full scale activated sludge process
operates at 3 g/L MLSS and uses surface oxygenators to supply pure oxygen for aerobic
treatment. A pilot scale sand filter was applied to the gravity settled CAS effluent for
comparative ozonation tests. Both biological systems are fed with equalized textile mill effluent
and employ High Purity Oxygen (HPO) activated sludge biological treatment. Table 1 contains
a summary of key process conditions in the Equalization tank effluent, CAS reactor, CAS Sand
Filtered (SF) effluent, and the membrane bioreactor effluent. It should be noted that the plant
uses ozone for tertiary color removal and the vent gas is recirculated to the surface oxygenators.
Therefore, treatment efficiency in the CAS process is partially attributed to the residual ozone.

The effluent from the secondary clarifier of the CAS process was further filtered using a custom
sand filtration system. No additional treatment was applied to the effluent from the membrane
filtration system. For the sand filter test, a pilot scale sand filter was constructed. A 280 mm
pipe was closed at one end with cloth and filled with 300 mm of 6-25 mm pebbles, 97 mm of
0.6-1.0 mm medium sand, and 240 mm of 0.1-0.3 mm fine sand (Figure 5). Since the cloth is
also a filter, a test was conducted with the cloth alone and found that about 23% of the apparent
color and 20% of the turbidity are removed in the absence of sand. Color measurements are
reported in mg/L CU, a unit of color equivalent to that produced by 1 mg/L platinum in the form
of the chloroplatinate ion (Standard Methods, 2005). Spectroscopic measurements were taken
using a Hach 550 Spectrophotometer with measurements at 455 or 465 nm. Turbidity was
measured using the nephelometric method (Standard Methods, 2005, 2130B) and reported in
NTU. In the economic analysis, the ozone requirement for sand filtration is increased to account
for the color removal from the cloth. During tests, effluent from the sand filter had a pH of 6.9,
apparent color of 828 mg/L CU and turbidity of 12 NTU.

Bubble column ozonation tests were conducted with measurements of ozone consumption,
apparent color, Chemical Oxygen Demand (COD), Total Organic Carbon (TOC) and turbidity
taken at regular intervals throughout each test. Unfiltered spectroscopic color measurements
were measured and reported in mg/L CU, a unit of color equivalent to that produced by 1 mg/L
platinum in the form of the chloroplatinate ion (Standard Methods, 2005). This enabled the
collation of critical process variables that would be utilized for evaluating the capital and
operating costs of the two major alternative filtration methods.

Copyright ©2014 Water Environment Federation

3055
 WEFTEC 2014

The bubble column used in this study is shown in Figure 5. The column diameter is 0.19 m with
a height of 2 m. Batch ozonation tests were conducted by filling the column with 36 L of either
Clarified effluent, Sand Filtered or iMBR effluent. A mixture of ozone and oxygen was passed
through the column using a fine bubble diffuser at the base. Throughout the test period, the
gaseous effluent ozone concentration was measured and 100 mL liquid samples were withdrawn
for analysis. The ozone generator and gas phase ozone monitors are from PCI Ozone and
Control Systems (generator model GL1, series SO 11020/92; monitor model HC 400, series
465). The mixture of ozone and pure oxygen is injected using a fine bubble diffuser at the base
of the column.

The influent to the pilot iMBR system is passed through a 2 mm screen before it enters into the
pilot bioreactor. The iMBR utilizes a Koch hollow fibre submerged ultrafiltration membrane
module with 0.03 micron pore size and surface area of 37 square meters. Separate tanks were
used for biological treatment and membrane filtration. Intermittent coarse bubble air scouring
was applied to the membrane tank for fouling and clogging control. With a permeate flow of 0.5
m3/h, the gross flux was 13.5 L/m2/h (LMH). The system was operated on a seven minute
filtration cycle with one minute of backflow at 1 m3/h for a net flux (filtration – backflow) of 7.7
LMH per cycle. Coarse bubble scouring air was operated intermittently at 10 m3/h with 40
seconds on and 20 seconds off. During testing, effluent from this process had a pH of 7.6,
apparent color of 1,112 mg/L CU and turbidity of 1.9 NTU as shown in Table 1.
Batch ozonation tests were conducted by filling the bubble column with 36 L of either
CAS or iMBR effluent. A mixture of ozone and oxygen was passed through the column at a
flow of 0.75 Nm3/h. The concentration of ozone varied between 36.2-36.5 gO 2 /m3. Throughout
the test period, the effective ozone utilization was determined by measuring the gaseous effluent
ozone concentration to establish the ozone utilization efficiency in the bubble column (Figure 7).

Table 1. Effluent characteristics of the equalization basin (influent to activated sludge process),
Conventional Activated Sludge (CAS) and immersed Membrane Bioreactor (iMBR)

Equalization CAS CAS-SF iMBR

pH 8.3 6.4 6.9 7.6
Apparent color mg/L CU 2122 1316 828 1112
Turbidity NTU 210 71 12 1.9

Copyright ©2014 Water Environment Federation

3056
 WEFTEC 2014

Figure 4. Pilot immersed membrane bioreactor system

Figure 5. Pilot ozonation system

Copyright ©2014 Water Environment Federation

3057
 WEFTEC 2014

Figure 6. Pilot scale sand filter used with gravity settled CAS effluent

Figure 7. Photographs of bubble column taken at 0, 2 and 10 minutes during iMBR test

Results and Discussion

Table 1 shows the initial conditions for the test as well as values of equalized effluent, before
biological treatment. The pH of CAS was 6.4, Sand Filtered Effluent 6.9 and MBR effluent 7.6.
Initial color was 1316 CU for CAS, 828 CU for SF and 1112 CU for MBR. Turbidity was
reduced from 71 NTU for CAS effluent to 12 NTU after sand filtration. Turbidity of the MBR
effluent was 1.9 NTU.

Results from the ozone bubble column tests are shown in Table 2 and Figure 8. Tests were
conducted at 10-15oC, with an ozone dose of 36.2-36.5 g O 3 /m3. Decolorization effects were
observed, and the data could be fit to pseudo-first order rate constants (see Table 2).

Copyright ©2014 Water Environment Federation

3058
 WEFTEC 2014

ln[C] = ln[Co] – kt [1]

where:
C = apparent color (mg/L CU)
Co = initial apparent color (mg/L CU)
k = first order rate constant (1/s)
t = time (s)

For the same ozone feed rate and concentration, the rates of color removal were higher for iMBR
(5.17E-3 1/s) and CAS+SF (3.88E-3 1/s) relative to CAS (1.33E-3 1/s). The estimated ozone
requirement for the MBR system is 26% of the O3 dosage required for an unfiltered CAS
effluent . For sand filtered CAS effluent with cloth, the nominal estimated O3 dosage without
correcting for the filtration effect of the cloth is 34%. Correcting for the presence of the cloth
(increasing the Co by 23%) results is an increase in ozone requirement of 19% to achieve the
same final apparent color. Thus, the expected ozone requirement for a sand filter without cloth is
42% that of an unfiltered CAS effluent.

The kinetics of color removal with ozone has been shown to be affected by pH, ozone
concentration and initial dye concentration (Adams, 2003; Sevimli, 2002). While efforts were
made to assure all tests were conducted at the same gas phase ozone concentration, initial dye
concentration and pH were intentionally not adjusted so that the comparative economics could be
developed for the current operation of the plant.

Table 2. First order rate constants for color removal (gas phase ozone concentration is similar
for each test)

CAS CAS+SF CAS+SF, iMBR

with cloth no cloth

Rate Constant 1/s 1.33E-3 3.88E-3 3.88E-3 5.17E-3

Co (fit) mg/L CU 1290 782 1015 960
pH 6.4 6.9 7.6
Relative to CAS % 34% 42% 26%

Copyright ©2014 Water Environment Federation

3059
 WEFTEC 2014

Figure 8. Results from bubble column ozonation tests demonstrating faster kinetics of filtered
effluent samples, and a reduced ozone dosage for iMBR effluent and sand filtered (CAS+SF)
effluent compared to CAS effluent

The operating costs were estimated by considering the power and oxygen requirements for each
system as shown in Table 3. Using a power cost of 0.7 kWh/m3 for CAS and 1.0 kWh/m3 for
iMBR treating municipal effluent (based on Young, 2012) it is assumed that biological aeration
accounts for 60% of this power and a correction was applied for the increased COD level of the
industrial effluent. The base case is for a CAS system that uses air for biological aeration and
pure oxygen for ozone generation. The CAS, CAS+SF and iMBR cases use pure oxygen for
both ozone generation and biological aeration. A power savings is applied for pure oxygen
aeration since blowers are not needed (excluding the power of oxygenators used to deliver pure
oxygen). There is also a credit applied for oxygen that is recycled to the biological treatment
system. In the case of CAS, 92% of the biological oxygenation requirement is met by the vent
gas recycle. In the case of iMBR, vent gas meets 22% of the requirement. For the sand filter,
recovery and reuse of the vent gas meets 37% of the oxygenation requirement in the biological
basin. The base CAS case represent the scenario in which the oxygen used for the ozone
generation is not recovered and reused in the biological basin. Hence no offsets to the blower
demand for basin aeration is observed for this case.

Copyright ©2014 Water Environment Federation

3060
 WEFTEC 2014

Table 3. Comparison of costs for ozone tertiary treatment of CAS and iMBR effluent.

CAS, base CAS CAS, SF iMBR

Power* 0.92 0.92 0.92 1.22 kWh/m3
Power credit, pure oxygen 0.00 -0.52 -0.60 -0.51 kWh/m3
Ozone power 0.70 0.70 0.30 0.18 kWh/m3
Total power 1.62 1.10 0.63 0.89 kWh/m3
Power cost $0.16 $0.11 $0.06 $0.09 $/m3

Oxygen use, ozone 0.70 0.70 0.30 0.18 kg O2/m3

Oxygen use, biological 0.00 0.05 0.37 0.47 kg O2/m3
Total oxygen use 0.70 0.75 0.67 0.65 kg O2/m3
Oxygen cost $0.07 $0.08 $0.07 $0.06 $/m3
Operating costs, misc** $0.02 $0.02 $0.03 $0.04 $/m3
Total operating cost $0.25 $0.20 $0.16 $0.20 $/m3

Capital, base* $1.26 $1.26 $1.26 $1.58 $M/MLD

Capital, ozone system $0.12 $0.12 $0.06 $0.06 $M/MLD
Capital sand filter $0.35 $M/MLD
Total capital $1.38 $1.38 $1.68 $1.65 $M/MLD
Capital, 4.8 MLD $6.63 $6.63 $8.05 $7.90 $M
Capital, annualized $0.58 $0.58 $0.70 $0.69 $M/yr

Total annualized capital cost $0.33 $0.33 $0.40 $0.39 $/m3

Total cost $0.58 $0.53 $0.56 $0.59 $/m3

Basis: *Based on Young, 2012; **Cost for replacement or maintenance of diffusers, oxygenator,
ozone generator, membrane, sand filter. Assumptions : 10 kWh/kg O3, 10% ozone generation
and dissolution efficiency, 70 g O 3 /m 3 CAS ozone use, iMBR ozone use = 26% CAS, CAS+SF
ozone use = 42% CAS (without cloth), $100/ton oxygen cost, $0.10/kWh power cost, Capital
annualized over 20 years at 6% rate, labor not included, sludge disposal costs not included

For the full scale plant the power requirement of ozone generation is 10 kWh/kg O 3 and the
ozone requirement is 70 gO 3 /m3. Assuming that ozone of 10% (w/w) is generated, and that it is
completely utilized within the ozone contactor, an iMBR ozone requirement that is 26% of CAS,
and CAS+SF ozone requirement that is 42% of CAS is obtained. The corresponding ozone
generation power requirement is estimated to be 1.10 kWh/m3 for CAS, 0.89 kWh/m3 for iMBR
and 0.63 kWh/m3 for CAS+SF compared to a base of 1.62 kWh/m3. Oxygen use for the base
case is 0.7 kg O 2 /m3 and for iMBR 0.65 kg O 2 /m3 with CAS system is slightly higher at 0.75 kg
O 2 /m3. CAS+SF oxygen requirement is 0.67 kg O 2 /m3. Considering $100/ton for oxygen, the
total operating cost for CAS and iMBR is $0.20/m3 vs. $0.16/m3 for CAS+SF relative to
$0.25/m3 for the base case.

Copyright ©2014 Water Environment Federation

3061
 WEFTEC 2014

The operating costs for iMBR and CAS are 20% lower than for the base case and CAS+SF is
40% lower. Assuming $1,260,000/MLD for CAS system, an iMBR system is expected to be
about 25% higher in cost (Young, 2012) to account for the addition of membranes and the
associated infrastructure. For the scale of system in this study, the capital costs for an ozonation
system are estimated to be $120,000/MLD for CAS and about 50% less for iMBR. This includes
ozone generator, contactors, cooling system and accessories. The cost of a sand filtration with
automatic back-flush is estimated to be $60,660/MLD. For a 4.8 MLD plant, amortized at 6%
over 20 years, the annual capital cost for CAS systems would be $0.33/m3 and $0.39/m3 for
iMBR.

The estimated annual cost of operation including amortized capital over 20 years, is $0.53/m3 for
CAS (with vent gas reuse) and $0.59/m3 for iMBR relative to a base case of $0.58/m3 (Figure 9).
The lowest cost to operate is CAS+SF at $0.53/m3. Additional benefits offered by MBR systems
such as improved effluent quality and reduced footprint have not been quantified in this analysis.

Figure 9. Annualized capital (20 years) and operating costs normalized for 4.8 MLD plant

Based on full-scale experience with tertiary ozone treatment of textile wastewater, assumptions
were made regarding capital and operating costs for the ozone system and CAS systems.
Estimates for the MBR system economics were based on Young et al. (2012). Sand filtration
was based on gravity, single filtration (OECD, 2013). Figure 9 shows estimated annualized
capital and operating costs over 20 years for CAS (unfiltered), Sand Filtered and MBR 4.8 MLD
plant. In each of these cases, pure oxygen is used for biological treatment and the vent gas from
the ozonation process is collected and recycled back to the biological treatment basin producing
a credit for biological oxidation.

Based on assumptions used in this case, the lowest cost of ownership is the unfiltered CAS
system with the lowest capital cost ($0.33/m3) and an operating cost similar to MBR ($0.20/m3).
The sand filtration system has the lowest operating cost at $0.16/m3, however it also has the
highest capital cost ($0.40/m3). The MBR capital cost was similar at $0.39/m3.

Copyright ©2014 Water Environment Federation

3062
 WEFTEC 2014

Conclusions
Pilot tests were conducted at a full-scale 200 m3/h textile plant that currently uses ozone tertiary
treatment to achieve effluent color target levels of 200 mg/l CU. This is a reduction from 2122
mg/l CU at the influent of the biological treatment process. The efficiency of tertiary color
removal with ozone was compared between CAS effluent, sand filtered CAS effluent and
percolate from an immersed membrane bioreactor system. Although the iMBR system required
the least amount of ozone to achieve target color levels, the sand filtration was determined to be
the lowest cost to operate due to lower power costs and lower maintenance. Comparing
annualized capital and operating costs, the membrane system is similar to the base case of CAS
with gravity separation and air-based biological treatment.

However, it should be noted that the cost of ownership of the membrane system is not much
higher than the CAS system at $0.59/m3 vs. $0.53/m3. In addition, implementation of the
membrane system would provide additional benefits such as reduced footprint and improved
effluent quality for installations considering reuse. These benefits were not quantified in this
study. Incorporating considerations for water reuse can significantly alter the economics,
potentially making the iMBR system with its high effluent quality, and the potential for
eliminating or reducing water intake and discharge fees, the most cost effective option.

References

1. Adams, C., Gorg, S. 2003. Effect of pH and Gas-Phase Ozone Concentration on the
Decolorization of Common Textile Dyes. Journal of Environmental Engineering, 128 (3),
293-298.
2. Avsar, Y., & Batibay, A. Ozone Application as an Alternative Method to te Chemical
Treatment Technique for Textile Wastewater. Fresenius Environmental Bulletin. PS Volume
19, No 12, 2010.
3. Judd, S.; Judd, C. 2011 MBR Book - Principles and Applications of Membrane Bioreactors
for Water and Wastewater Treatment (2nd Edition), Elsevier.
4. Langlais, B., Reckhow, D., Brink, D. 1991. Ozone in Water Treatment Application and
Engineering, American Water Works Association and Lewis Publishers.
5. Matsui, M. 1996. Ozonation. Chapter 3. Environmental Chemistry of Dyes and Pigments.
John Wiley & Sons, Inc. NY, Eds. Reife A., & Freeman, H.S.
6. Rambor, A., & Fabiyi, M. 2012 Full Scale Application of Ozone for Color Removal and
Biosolids Reduction in a Textile Plant, Proceedings from the annual meeting of the Water
Environment Federation (WEFTEC), New Orleans, 5556-5563.
7. SciFinder (2013) by CAS, a division of the American Chemical Society
https://www.cas.org/products/scifinder (accessed August 14, 2013).
8. Sevimli, M.F., Sarikaya, H.Z., (2002), “Ozone treatment of textile effluents and dyes: effect
of applied ozone dose, pH and dye concentration,” J. Chem. Technol. Biotechnol., 77(7),
842-850.

Copyright ©2014 Water Environment Federation

3063
 WEFTEC 2014

9. Standard Methods for the Examination of Water and Wastewater 2005. 21st edition,
American Public Health Association/American Water Works Association/Water
Environment Federation, Washington DC, USA.
10. Vandevivere, P.C., Bianchi, R., & Verstraete, W. Treatment and Reuse of Wastewater from
the Textile Wet-Processing Industry: Review of Emerging Technologies. J. Chem. Technol.
Biotechnol. 1998, 72, 289-302
11. Young, T., Muftugil, M., Smoot, S., Peeters, J. 2012 “Capital and Operating Cost Evaluation
of CAS vs. MBR Treatment,” Paper presented at the annual meeting of the Water
Environment Federation (WEFTEC), New Orleans, p. 4087-4097.

Copyright ©2014 Water Environment Federation

3064