Two-Stage Anaerobic Fluidized-Bed Membrane Bioreactor
Treatment of Settled Domestic Wastewater
J. Bae*, R. Yoo*, E. Lee* and P. L. McCarty*, **
* Department of Environmental Engineering, Inha University, Namgu, Yonghyun dong 253, Incheon, Republic of Korea
(E-mail: [email protected]; [email protected]; [email protected])
** Department of Civil and Environmental Engineering, Stanford University, Stanford, CA 94305, USA
(E-mail: [email protected])
Abstract
A two-stage anaerobic fluidized-bed membrane bioreactor (SAF-MBR) system was applied for the treatment of
primary-settled domestic wastewater that was further pre-treated by either 10 µm filtration or 1 mm screening.
While the different pre-treatment options resulted in different influent qualities, the effluent qualities were quite
similar. In both cases at a total hydraulic retention time (HRT) of 2.3 hr and 25oC, COD and BOD5 removals were
84%-91% and 92%-94%, with effluent concentrations lower than 25 and 7 mg/L, respectively. With a membrane
flux of 6-12 L/m2/h, trans-membrane pressure (TMP) remained below 0.2 bar during 310 days of continuous
operation without need for membrane chemical cleaning or backwashing. Biosolids production was estimated to be
0.028-0.049 g VSS/g BOD5, which is far less than that with comparable aerobic processes. Electrical energy
production from combined heat and power utilization of the total methane produced (gaseous and dissolved) was
estimated to be more than sufficient for total system operation.
Keywords: anaerobic fluidized bed bioreactor; membrane; sewage treatment
INTRODUCTION
Anaerobic treatment of sewage would be a good alternative to aerobic processes with the
advantages of energy recovery and low sludge production. Although poor effluent quality has often
been found to be a problem with anaerobic treatment of domestic wastewater (Gomec, 2010;
Seghezzo et al., 2007), this can be overcome using a membrane bioreactor, which can capture
effluent suspended solids and maintain a long solids retention time (SRT). However, membrane
fouling has been a major barrier to its application with high operating and energy costs (Smith et.
al., 2012). Many attempts have been made to reduce membrane fouling (Alan et al., 2010; Huang et
al., 2011; Martinez-Sosa et al., 2011). Kim et al. (2011) recently proposed a novel staged fluidized
bed membrane bioreactor (SAF-MBR) system consisting of an anaerobic fluidized bed reactor
(AFBR) and an anaerobic fluidized membrane bioreactor (AFMBR) in series. When continuously
fed 500 mg/L COD synthetic wastewater at 35oC with a total hydraulic retention time (HRT) of 5.0
hr, overall COD removal was 99%, with permeate COD of 7±4 mg/L. The clogging of the
membrane was successfully controlled by the scouring effect of granular activated carbon (GAC) in
the AFMBR, with less energy consumption compared to a submerged membrane bioreactor using
gas sparging for fouling control. In a subsequent study, Yoo et al. (2012) evaluated this system at
25oC using primary settled domestic wastewater that was subsequently screened with a 10 µm filter,
and found similar treatment results.
The objective of this study was to further evaluate the performance of the SAF-MBR system for
treatment of domestic wastewater at ambient temperature. The study by Yoo et al. (2012) using 10
µm filtered settled domestic wastewater was extended for an additional 100 days using a more
practical settled domestic wastewater that was screened with a 1 mm screen, yielding a total study
time for the SAF-MBR of 310 days with the same membranes. Concern has been expressed over
the impact of GAC fluidization on membrane integrity (Smith et al., 2012), and so evaluation of
long-term membrane operation is of importance. In this study, organic removal efficiency, biosolids
production, and energy balance were evaluated while operating the SAF-MBR system without the
use of membrane chemical cleaning or backwashing.
MATERIALS AND METHODS
Reactor operation
The lab-scale SAF-MBR system illustrated in Figure 1 has been described in detail by Yoo et al.
(2012). The 0.245 L AFBR contained 30 g of 10 x 30 mesh fresh GAC as support medium for
bacterial growth. The AFMBR was the same size, but contained 54 g of fresh GAC and a
submerged membrane module consisting of eight 0.45 m long, polyvinylidenefluoride (PVDF)
hollow fiber membranes, with surface area of 0.0215 m2. Fluidization of GAC was maintained with
a magnetic pump to maintain the desired recirculation flow rate (Table 1).
Gas bag
Gas bag
Overflow
line
Peristaltic
pump
Sludge
wasting
line
100 %
expension
Peristaltic
pump
40 %
expension
Pressure
gauge
Peristaltic
pump
Permeate
Flow meter
Flow meter
Recirculation
pump
- AFBR -
Influent
Recirculation
pump
- AFMBR -
Figure 1. Schematic diagram of the SAF-MBR system (after Yoo et al., 2012)
The SAF-MBR system was operated at 25oC under two different Modes based on the pre-treatment
of feed wastewater as summarized in Table 1. Details of operation under Mode I was reported
previously (Yoo et al., 2012). A summary of that data is also presented here in order to compare
with the new results presented here on Mode II operation. The feed to the treatment system was
primary effluent from a sewage treatment plant in Bucheon, Korea, which was taken once a week
and stored under 4oC refrigeration. Just prior to use, the wastewater was filtered through either a 10
m cartridge filter (Mode I) or a 1 mm screen (Mode II). With 10 m filtration about 60% of the
primary effluent total suspended solids (TSS) were removed. However, with 1 mm screening, TSS
rejection was negligible, and thus the feed wastewater was close in composition to the primary
effluent.
The SAF-MBR system was operated at a set-point permeate flux of 6 to 9 LMH for 98 days before
Mode I, which lasted for 61 days. Between Modes I and II, the SAF-MBR was operated for 51 days
to determine the sustainable flux that could be maintained without producing significant membrane
fouling, which was 11 LMH (Yoo et al., 2012). Then Mode II operation was begun and lasted for
100 days.
For both Modes I and II the HRT of the AFBR and the AFMBR were 1.0 and 1.3 hr, respectively.
The corresponding membrane flux of the AFMBR for these two Modes was maintained at 9 L/m2/h
(LMH). About 280 ml of recycle fluid was withdrawn each week to remove excess suspended
biosolids production for disposal. The organic loading rate (OLR) of the AFBR increased from 3.7
kg COD/m3-d (Mode I) to 5.6 kg COD/m3-d (Mode II), an increase caused by the higher feed VSS
during Mode II.
Table 1. Operating conditions for Modes I and II for the SAF-MBR system
Mode I
Mode II
AFBR
Operating period (days)
Influent pre-treatment
AFBR
AFMBR
61
100
10 ㎛ filtration
1 mm screen
25
25
Temperature (oC)
HRT (h)
AFMBR
1.0
1.3
1.0
1.3
-
9
-
9
OLR (kg COD/m3-d)
3.7
1.3
5.6
1.2
Recirculation flow rate (L/d)
446
1008
504
1080
Membrane flux (LMH)
Analytical procedures
COD, BOD5, total and volatile suspend solids (TSS, VSS), pH and alkalinity were determined
according to procedures in Standard Methods (APHA, 1998). Sulfate was analyzed with an ionchromatograph (DX 500, Dionex, Sunnyvale, CA, USA) and volatile fatty acids (VFA) were
analyzed with a HP 6890 series gas chromatograph with flame ionization detector (FID). Biogas
production was measured using a gas bag (TDC, Japan) for collection. Gas composition was
determined using a GC with thermal conductivity detector (HP6890 series Gas Chromatograph,
Hewlett–Packard, USA). Dissolved methane was also measured with a method described in more
detail elsewhere (Shin et al., 2011).
RESULTS AND DISCUSSION
Organic removal and effluent quality
Table 2 contains a summary of the performance of the AFBR, the AFMBR, and the overall SAFMBR system. In Mode I, the AFBR achieved a COD removal of 56%, while the AFMBR reduced
its influent COD by an additional 63%, providing an overall COD removal for the two-stage system
of 84%. Although influent COD in Mode II was higher than that in Mode I, effluent COD slightly
decreased from 25 mg/L (Mode I) to 22 mg/L (Mode II). As a result, COD removal in the AFBR
increased to 72%, resulting in the overall COD removal of 91%. The average effluent BOD5 values
were below 7 mg/L in both Modes, and corresponding BOD5 removal efficiencies were equal to or
higher than 92%. Effluent volatile fatty acids were 3 mg COD/L or less, indicating almost complete
degradation.
Table 2. Performance of SAF-MBR system
COD
VSS
VFA
154±34
88±23
40±7
37±7
-
Eff. (mg/L)
68±18
25±9
21±7
18±4
4±4
AFMBR
Eff. (mg/L)
25±10
7±2
2±1
2±1
3±3
Overall
Rem. (%)
84
92
94
95
-
Inf. (mg/L)
235±44
105±28
93±27
78±23
-
Eff. (mg/L)
66±12
30±10
21±7
19±7
1±1
AFMBR
Eff. (mg/L)
22±8
6±2
1±0
1±0
1±1
Overall
Rem. (%)
91
94
99
99
-
AFBR
Mode II
TSS
Inf. (mg/L)
AFBR
Mode I
BOD5
Some organics were removed by sulphate reduction. Influent sulphate concentration was relatively
constant throughout the study with an average value of 64 mg/L, complete removal of which is
equivalent to the removal of 42.5 mg COD/L. Thus, sulphate reduction not only reduces methane
production potential but also causes a need for sulphide control in both the liquid and gas phases.
The relatively higher COD removal in Mode II than Mode I was associated both with the increase
in influent volatile suspended solids (VSS) and its subsequent greater removal in the AFBR from 51%
(Mode I) to 76% (Mode II). Reasons for the higher removal during AFBR treatment during Mode II
are not clear. Perhaps this resulted from better attachment of VSS on the AFBR GAC biofilm
during Mode II and build-up or subsequent degradation there, the development of organisms with
better VSS degradation capabilities, or because the larger VSS present were themselves more
readily biodegradable. Effluent VSS concentrations in both Modes were near zero through
membrane filtration. The pH and alkalinity varied over the range of 6.75-7.30 and 225-275 mg/L as
CaCO3, respectively. The above results indicate that the SAF-MBR system can produce effluent
quality meeting some of the stricter organic discharge requirements.
VSS reduction and biosolids production
In order to quantify VSS reduction and biosolids production, changes in total biomass accumulation
within the reactors, including the attached biomass on GAC particles need to be considered.
However, measurement of biomass accumulation on the GAC particles is not easy to perform
without upsetting reactor operation. It was measured once for both reactors, 260 and 239 days after
the start-up of the AFBR and the AFMBR, respectively, time periods ranging between the Mode I
and II periods of operation (Yoo et al., 2012). Estimated total biomass at these later dates on the
GAC particles was 1.11 g VSS in the AFBR and 0.58 g in the AFMBR, corresponding respectively
to accumulations of 0.71 and 0.50 mg VSS/L as listed in Table 3.
VSS destruction summarized in Table 3 indicates that the AFBR provided an increased VSS
destruction capacity with high influent VSS, 19 mg VSS/L in Mode I, to 59 mg VSS/L in Mode II.
On the other hand, VSS destruction in the AFMBR remained constant at 13-15 mg VSS/L.
The removal of excess VSS from the SAF-MBR system was accomplished by weekly removal of
the recycle fluid from the AFMBR (280 mL/week) and by a period removal that occurred during
pump and recycle line cleaning. Based on the measured concentration of VSS in the wasting stream,
wasting from the AFMBR was estimated to be 1.11 mg VSS/L, while that with pump and recycle
line cleaning was 1.64 mg/L. The resulting net total biosolids production estimate including both
biomass accumulation on GAC and VSS wasting was 0.049 g VSS/g BOD5 removed or 0.031 g
VSS/g COD removed (Yoo et al, 2012). However, the biofilm growth could be ruled out from this
calculation as most growth on the GAC probably occurred early during reactor start-up and not
during Mode I. With this assumption, net total biosolids production estimate (Table 3) was 0.034 g
VSS/g BOD5 removed or 0.021 g VSS/g COD removed in Mode I. The true biosoilds production
would lie between the two values. In Mode II, the respective values decreased to 0.028 g VSS/g
BOD5 removed and 0.013 g VSS/g COD removed.
Table 3. Removal, wasting, and destruction of VSS during Modes I and II of operation
Mode I
Mode II
Units : mg/L (unless
otherwise specified)
AFBR AFMBR
Total
AFBR AFMBR
Total
Removal
19
16
35
Bulk solution
1.11
Recirculation line & pump
Totals
59
18
77
1.11
0.84
0.84
1.64
1.64
1.91
1.91
2.75
2.75
2.75
2.75
Wasting
Destruction
19
13
32
59
15
74
% of reactor influent
49
72
84
76
79
95
The above biosolids production is far less than the typical value for aerobic secondary sewage
treatment of 0.42 g VSS/g BOD5 (Rittman and McCarty, 2001), or for an aerobic fluidized bed
reactor of 0.12-0.135 g VSS /g COD (Patel, et al., 2006). For an anaerobic fluidized bed reactor,
Borja et al, (2000) reported a biomass yield of 0.075 g VSS/g COD, while Jovanovic et al. (1986)
indicated a net VSS yield of 0.03 g VSS/g COD with an operating OLR of 20 kg COD/m3-d. The
biosolids production in the SAF-MBR system was slightly lower than these reported anaerobic
values, perhaps because the AFMBR was aided by its provision of a longer SRT.
Membrane fouling control by GAC fluidization
The AFMBR was operated continuously for 310 days without the need for cleaning procedures such
as backwashing or chemical washing, and with low trans-membrane pressure (TMP) as illustrated
in Figure 2. This long-term fouling free operation resulted from the scouring effect of the fluidized
GAC (Kim et al., 2011). Before Mode I, the AFMBR was operated for 98 days at a flux of 6 LMH,
and the TMP remained below 0.1 bar. During Mode I, TMP was maintained in the narrow range of
0.042 and 0.057 bar for 61 days. After Mode I operation, the maximum sustainable flux of the
AFMBR was tested over a 51 day period. This was found to be about 11 LMH using a variety of on
and off procedures. The flux of 9 LMH used throughout Mode I was then selected as the best for
longer-term operation and the Mode II evaluation. With this flux during Mode II an average TMP
of 0.116±0.015 bar was obtained.
Mode I
Mode II
0.4
TMP (bar)
0.3
0.2
0.1
0.0
0
50
100
150
200
250
300
Time (d)
Figure 2. TMP variations for the AFMBR during 310 days of operation.
Energy consideration
Table 4 contains an electrical energy balance for the system, calculated as described by Yoo et al.
(2012). This considers all energy transfer efficiencies between pumps, motors, and generators. The
energy requirement was similar for both Modes. Electrical energy requirement for the pump for
fluidization of GAC was the major energy requirement, and was 0.011 and 0.036 kWh/m3 for the
AFBR and AFMBR, respectively. The three times greater energy requirement for the AFMBR than
for the AFBR was due to the need for full fluidization of GAC to cover the membranes in the
AFMBR. Also, a small additional electrical energy requirement of 0.001 kWh/m3 was needed for
the AFMBR permeate pump. The total estimated electrical energy requirement for system operation
during Modes I and II was 0.048 and 0.053 kWh/m3, respectively.
During Mode II, the estimated electrical energy that could be obtained from gaseous and dissolved
methane was 0.034 and 0.064 kWh/m3, respectively, or somewhat higher than during Mode I.
Therefore in Mode II, with gaseous methane only, about 64% of the energy requirement could be
satisfied. With the recovery of dissolved methane, the SAF-MBR system may become an energy
self-sufficient process. Even if not used as an energy source, it is important that dissolved methane
be captured or otherwise destroyed in some manner as it is a strong greenhouse gas. Commercial
systems such as gas stripping or gas-permeable membrane processes are available and can be used
Table 4. Estimated energy balance for the SAF-MBR system
Units : kWh/m3 (unless
otherwise indicated)
Mode I
Mode II
AFBR
AFMBR
Total
AFBR
AFMBR
Total
0.011
0.036
0.047
0.012
0.037
0.051
0.001
0.001
0.003
0.002
0.037
0.048
0.012
0.040
0.053
0.009
0.030
0.023
0.011
0.034
0.052
0.052
0.064
0.064
0.061
0.082
0.075
0.098
Electrical Energy required
for GAC fluidization
for permeate production
Total
0.011
Energy production potential from Methane
Gaseous
0.021
Dissolved
Total
0.021
0.023
Energy produced/required
With gaseous methane
63%
64%
With total methane
174%
188%
to recover the dissolved methane. The above analysis does not include the electrical energy
potential of the methane produced form primary treatment, which if included should make the SAFMBR system a significant net producer of electrical energy.
CONCLUSIONS
The SAF-MBR system was applied for the treatment of a primary-settled domestic wastewater at
25oC and a total HRT of 2.3 h. The effluent quality was similar whether the settled wastewater was
pre-treated using 10 m filtration or 1 mm screening, as the increased VSS concentration with
screening was mostly removed in the first stage AFBR. Overall COD and BOD5 were 84%-91%
and 92%-94% removed with permeate concentrations lower than 25 and 7 mg/L, respectively. For
310 days of operation, no membrane fouling control, such as chemical cleaning or backwashing,
was needed other than that resulting from the souring effect of the fluidized GAC
particles. Electrical energy that could result from combustion of the total methane produced
(gaseous and dissolved) would be more than sufficient for total system operation. Biosolids
production was estimated to be 0.028-0.049 g VSS/g BOD5, which is far less than that from
comparable aerobic processes. The SAF-MBR system has good potential as a low-energy highefficiency cost-effective domestic wastewater treatment system.
ACKNOWLEDGEMENT
This research was supported by an Inha University Research Grant and the World Class University
Program through the National Research Foundation of Korea, funded by the Ministry of Education,
Science and Technology (grant number R33-2008-000-10043-0).
REFERENCES
Alan, Y. H. and Stuckey, D. C. (2010). Treatment of dilute wastewaters using a novel submerged
anaerobic membrane bioreactor. Journal of Environmental Engineering, 132(2), 190-198.
Standard Methods for the Examination of Water and Wastewater (1998). 20th edn, American
Public Health Association/American Water Works Association/Water Environment Federation,
Washington DC, USA.
Borja, R., González, E., Raposo, F., Millán, F. and Martin, A. (2000). Assessment of kinetic and
macroenergetic parameters for a mesophilic anaerobic fluidised-bed reactor treating wastewater
derived from the production of protein isolates from extracted sunflower flour, Process
Biochemistry, 36(4), 369–375.
Gomec, C. Y. (2010). High-rate anaerobic treatment of domestic wastewater at ambient operating
temperatures: A review on benefits and drawbacks. Journal of Environmental Science and Health,
Part A: Toxic/Hazardous Substances and Environmental Engineering, 45(10), 1169–1184.
Jovanovic, M. K., Murphy, K. L. and Hall, E. R. (1986). Parallel evaluation of high rate anaerobic
treatment processes: retention time and concentration effects, Proceedings EWPCA conference
Anaerobic wastewater treatment: A grown-up technology, pp. 145-158.
Khan, A. A., Gaur, R. Z., Tyagi, V. K., Khursheed, A., Lew, B., Mehrotra, I. and Kazmi, A. A.
(2011). Sustainable options of post treatment of UASB effluent treating sewage: A review.
Resource, Conservation and Recycling, 55(12), 1232-1251.
Kim, J., Kim, K., Ye, H., Lee, E., Shin, C., McCarty, P. L. and Bae, J. (2011). Anaerobic fluidized
bed membrane bioreactor for wastewater treatment. Environmental Science and Technology, 45(2),
576-581.
Patel, A., Zhu, J. and Nakhla, G. 2006 Simultaneous carbon, nitrogen and phosphorous removal
from municipal wastewater in a circulating fluidized bed bioreactor. Chemosphere, 65(7), 1103–
1112.
Rittmann, B. E. and McCarty, P. L. (2001). Environmental Biotechnology: Principles and
Applications, McGraw-Hill, New York.
Seghezzo, L., Zeeman, G., Jules, B. V. L., Hamelers, H. V. M. and Lettinga, G. (1998). A review :
the anaerobic treatment of sewage in UASB and EGSB reactor. Bioresource and Technology, 65(3),
175-190.
Shin, J., Lee, E., McCarty, P. L. and Bae, J. (2011). Effects of influent DO/COD ratio on the
performance of an anaerobic fluidized bed reactor fed low-strength synthetic wastewater.
Bioresource and Technology, 102(21), 9860-9865.
Smith, A. L., Stadler, L. B., Love, N. G., Skerlos, S. J. and Raskin, L. (2012). Perspectives on
anaerobic membrane bioreactor treatment of domestic wastewater: a critical review. Bioresource
and Technology, In Press.
Yoo, R., Kim, J., McCarty, P. L. and Bae, J. (2012). Anaerobic treatment of municipal wastewater
with a staged anaerobic fluidized membrane bioreactor (SAF-MBR) system, submitted for
publication.