Anaerobic Fluidized Bed Membrane Bioreactor for ... - ACS Publications

Dec 15, 2010 - 0.1 bar during the final 40 d of operation. COD removals ... concern, as it increases operating and energy costs (5-9). Membranes are ...
0 downloads 0 Views 511KB Size
Download PDF
Environ. Sci. Technol. 2011, 45, 576–581

Anaerobic Fluidized Bed Membrane Bioreactor for Wastewater Treatment JEONGHWAN KIM,† KIHYUN KIM,† HYOUNGYOUNG YE,† EUNYOUNG LEE,† CHUNGHEON SHIN,† P E R R Y L . M C C A R T Y , ‡ A N D J A E H O B A E * ,† Department of Environmental Engineering, Inha University, Namgu, Yonghyun dong 253, Incheon, Republic of Korea, and Department of Civil and Environmental Engineering, Stanford University, Stanford, California 94305, United States

Received August 9, 2010. Revised manuscript received November 23, 2010. Accepted November 30, 2010.

Anaerobic membrane bioreactors have potential for energyefficient treatment of domestic and other wastewaters, membrane fouling being a major hurdle to application. It was found that fouling can be controlled if membranes are placed directly in contact with the granular activated carbon (GAC) in an anaerobic fluidized bed bioreactor (AFMBR) used here for posttreatment of effluent from another anaerobic reactor treating dilute wastewater. A 120-d continuous-feed evaluation was conducted using this two-stage anaerobic treatment system operated at 35 °C and fed a synthetic wastewater with chemical oxygen demand (COD) averaging 513 mg/L. The first-stage was a similar fluidized-bed bioreactor without membranes (AFBR), operated at 2.0-2.8 h hydraulic retention time (HRT), and was followed by the above AFMBR, operating at 2.2 h HRT. Successful membrane cleaning was practiced twice. After the second cleaning and membrane flux set at 10 L/m2/h, transmembrane pressure increased linearly from 0.075 to only 0.1 bar during the final 40 d of operation. COD removals were 88% and 87% in the respective reactors and 99% overall, with permeate COD of 7 ( 4 mg/L. Total energy required for fluidization for both reactors combined was 0.058 kWh/m3, which could be satisfied by using only 30% of the gaseous methane energy produced. That of the AFMBR alone was 0.028 kWh/ m3, which is significantly less than reported for other submerged membrane bioreactors with gas sparging for fouling control.

Introduction With growing concerns over the impact of fossil fuel usage on climate change, the consideration of wastewater as a source of renewable energy is growing. Toward this end, there are now many applications of anaerobic treatment processes for domestic wastewater that produce energy in the form of methane, rather than aerobic processes that consume energy (1). However, anaerobic treatment of domestic wastewater alone has generally not been sufficient to meet stringent effluent requirements for chemical oxygen demand (COD) and suspended solids (SS), thus often necessitating some type of post-treatment (1-3). An alterna* Corresponding author phone: 82-32-860-7507; fax: 82-32-8651425; e-mail: [email protected]. † Inha University. ‡ Stanford University. 576

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

tive newer approach is the use of anaerobic membrane bioreactors (AMBR), which are capable of achieving high effluent quality (4, 5). However, membrane fouling is a concern, as it increases operating and energy costs (5-9). Membranes are commonly placed external to the bioreactor, where a high cross-flow velocity is used to reduce fouling, or submerged within the reactor itself, where extensive gas scouring is generally used. In an extensive review, Liao et al. (6) summarized that energy usage was 3-7.3 kWh/m3 with external cross-flow membranes and 0.25-1.0 kWh/m3 with internal submerged membranes, quantities that are little different than with their aerobic counterparts. Such highenergy requirements reduce the potential advantage of anaerobic over aerobic systems. For low-strength wastewater, Liao et al. (6) recommended more research on combining membranes with existing highrate reactor configurations already determined to be suitable for dilute wastewaters. Concerning bioreactors for this purpose, the anaerobic fluidized bed bioreactor (AFBR) is particularly advantageous, because it is not as subject as most other reactor types to microorganism washout at short hydraulic retention time (HRT) and has especially good mass transfer characteristics as well (10). In this study, the advantages of the AFBR were evaluated when combined with that of an internal submerged membrane bioreactor, termed the anaerobic fluidized membrane bioreactor (AFMBR). The AFMBR used as post-treatment for effluent polishing was evaluated for its energy requirements as well as its ability to help meet stringent COD and SS effluent requirements. Here, granular activated carbon (GAC) was used as the fluidized medium to support biological growth. We are not aware of any other studies of an anaerobic FBR/MBR combination with GAC as support medium. However, there are numerous reports on use of powdered activated carbon (PAC) in anaerobic MBRs, indicating a positive effect in reducing membrane fouling (6, 11). Park et al. (12) were among the first in 1999, reporting that 5 g/L PAC helped increase the flux in the external membranes used by about 25%, which they indicated may have resulted from scouring as well as the removal of finer particles from solution by PAC. Studying a submerged aerobic MBR, Ng et al. (13) suggested that the major role of PAC in fouling reduction was adsorptive removal of organics and fine colloids rather than scouring. In one of the few reports on the use of GAC, Hu and Stuckey (14) evaluated its effect on membrane fouling in comparison with PAC. In each case, the concentration added to the submerged anaerobic MBR was 1.7 g/L. They reported that PAC provided better membrane flux and reduced transmembrane pressure (TMP) more than did either GAC or the control without either. While results of that study suggest that PAC would be a better alternative than GAC, concentrations used were much lower than that generally associated with fluidized bed reactors as used in the study reported here.

Experimental Section A laboratory AFMBR was constructed and used as a posttreatment or polishing bioreactor to reduce COD and SS concentration contained in the effluent from an upstream anaerobic bioreactor. This was felt to be a useful potential application, since some type of post-treatment is generally required for anaerobic bioreactors treating domestic wastewaters (1, 3). Two different reactor configurations used for the first stage of treatment that preceded the AFMBR are an anaerobic baffled reactor (ABR) (2, 10) and an AFBR. 10.1021/es1027103

 2011 American Chemical Society

Published on Web 12/15/2010

FIGURE 1. Schematic diagram of two-stage anaerobic fluidized bed membrane bioreactor. Short-term testing of the AFMBR was conducted using both ABR and AFBR treated effluents, both of which had been operating independently for several months previously and were at or near steady-state conditions. The short-term studies were to determine fouling potential of the effluents and the ability of the fluidized bed system to reduce fouling. A long-term test of the AFMBR was conducted with the twostage AFBR/AFMBR system. Reactors. Figure 1 is a schematic diagram of the twostage AFBR/AFMBR system used in the long-term study. The first reactor in the series was an AFBR and the second for post-treatment, the AFMBR itself. The feed flow rate to the AFBR was somewhat higher than to the AFMBR to ensure that sufficient flow would always be available for the AFMBR, the excess being sent to waste. The AFBR was a 3.93 L reactor with two attached settling chambers at the top to catch carryover GAC. The reactor consisted of a 2.0 m long by 50 mm diameter acrylic tube containing 450 g of 10 × 30 mesh GAC (MRX-M, Calgon Carbon Corp., Pittsburgh, PA). The two joined settlers were made from a 300 mm long by 100 mm diameter tube and had a total volume of 4.71 L. A magnetic pump (Pan World magnet pump, NH-100PX-Z, Korea) was used for recirculation to maintain fluidization of the GAC. The AFMBR was a 2.0 L reactor similar in design to the AFBR but with only one settler at the top. It consisted of a 1.0 m long by 50 mm diameter acrylic tube containing the same amount and kind of GAC as in the AFBR. Additionally, it contained a submerged membrane module with sixteen 0.95-m long, polyvinylidenefluoride (PVDF) hollow-fiber membranes (Kolon Inc.) with inside diameter of 1.9 mm, nominal pore size of 0.1 µm, and representing a total membrane surface area of 0.091 m2. Both AFBR and AFMBR were operated at 35 °C. During the long-term study with the two-stage AFBR/ AFMBR system, the feed to the AFBR mainly consisted of sodium acetate and sodium propionate at equal COD concentrations of 250 mg/L. Yeast extract and oleic acid were added throughout most of the study, each with a COD concentration up to 10 mg/L. The influent TCOD averaged 513 mg/L. NH4Cl (0.191 g/L) and 10 mL/L of filtered anaerobic digester supernatant (COD of about 5 mg/L) were added as

sources of nitrogen and micronutrients. The AFBR was operated for the first 43 days at 2.8 ( 0.2 h HRT, representing an organic loading rate (OLR) of 4.4 ( 0.3 kg COD/m3/d; changing after that to an HRT of 2.0 ( 0.2 h and OLR of 6.2 kg COD/m3/d. A recirculation flow rate of 2.0 L/min resulted in GAC bed expansion of 90-100% to a height of 0.9-1.0 m. The AFBR effluent was delivered to the AFMBR with a peristaltic pump (Masterflex, Model No. 7520-57) at a flow rate automatically controlled to maintain a constant water level at the top of the AFMBR where GAC fluidization was maintained by fluid recirculation at 1.4 L/min with a magnetic pump using a flow-rate controller (Blue-white, F-450). The top open sections of the membrane fibers in the AFMBR were connected to a peristaltic pump (as above) set to pump at 15.4 or 21.6 L/d to achieve membrane fluxes of 7 or 10 L/m2/h and HRT of 3.1 or 2.2 h, respectively. TMP required to maintain either flow was monitored with a vacuum pressure meter (EW-68604-00, ColeParmer). In short-term tests of the AFMBR system, effluents from an ABR as well as from the AFBR were used. The ABR was operated at 24 ( 1 °C, had a total volume of 3.4 L, and contained eight equal volume compartments, each separated by a vertical baffle to direct flow from the top of one chamber to the bottom center of the next. The feed was similar to the equal COD acetate/propionate mixture fed to the AFBR, except that it contained no yeast extract or oleic acid. The feed COD concentration and HRT were varied over the course of the study while a constant organic loading rate of 2 kg/m3/d was maintained. More specific details are provided in the Results and Discussion for each short-term test. Analytical Procedures. COD and total and volatile suspended solids (TSS and VSS) were determined according to Standard Methods (15). Biogas produced was collected in a Tedlar bag (Dupont Corp.) and measured volumetrically with a MilliGascounter (Ritter). Biogas composition was analyzed with an HP 6890 series gas chromatograph TCD with Alltech-CTR I column. Volatile fatty acids (VFA) were analyzed with a HP 6890 series gas chromatograph FID equipped with a 0.5 mm i.d. HP-INNOWAX graphite column (Agilent). VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

577

FIGURE 2. Effect of GAC fluidization on TMP in the anaerobic fluidized bed membrane bioreactor (AFMBR) at a permeate flux of 10 L/m2/h and recirculation rate of 1.4 L/min.

FIGURE 3. Comparative TMP with GAC fluidization, with GAC pretreatment, and without GAC at a permeate flux of 15 L/m2/h.

Results and Discussion Short-Term Test: Effect of GAC Fluidization on Membrane Fouling. In the first short-term test, three AFMBRs were operated in parallel, one was filled with distilled water and no GAC, the other two were filled with AFBR effluent, but only one contained GAC. At this time the AFBR influent had a COD of 2825 mg/L, and the effluent used contained 167 mg/L SCOD and 170 mg/L TSS. Fluid was recirculated in each reactor at a rate of 1.4 L/min, which fluidized the GAC to column height. Permeate flux from each reactor was set at 10 L/m2/h, with permeate being recycled back into the reactor to maintain reactor liquid volume. The effect with time on TMP is illustrated in Figure 2. With distilled water TMP remained at 0.04 bar, while it increased to about 0.3 bar within 6 h in the reactor with AFBR effluent and no GAC. In the reactor with GAC, fluidization was induced for the first 3 h with TMP remaining just slightly above that with distilled water. Recirculation was then stopped to let the GAC settle, following which TMP increased to 0.24 bar at 7 h when recirculation and GAC fluidization were again started. The TMP then dropped rapidly so that within two hours it returned to just above that with distilled water. The AFBR effluent was here demonstrated to cause significant membrane fouling. GAC fluidization, rather than GAC alone without fluidization, was demonstrated to have a major impact in reducing membrane fouling. ABR effluent produced similar behavior of membrane fouling when GAC fluidization was absent (data not shown). A subsequent short-term test was conducted using continuous feed of distilled or ABR effluent to the reactors at a permeate flux of 15 L/m2/h, resulting in an HRT of 1.5 h. During this test, the ABR influent COD was 245 mg/L, the effluent then containing 15.9 ( 3.0 mg/L TCOD, 9.6 ( 1.6 mg/L SCOD, 4.0 ( 1.3 mg/L TSS, and 3.8 ( 1.3 mg/L VSS. In order to see how GAC by itself might reduce fouling, a fourth reactor set was used. Here, the ABR effluent was passed first through an equivalent bed of GAC before entering an AFMBR that itself contained no GAC. The results are illustrated in Figure 3. TMP to maintain the flux increased continuously over the 30 h of operation without GAC. In the reactor with GAC fluidization, fouling was essentially eliminated with TMP remaining near that of distilled water. While GAC treatment before entering the AFMBR helped reduce the TMP somewhat, such GAC treatment alone was clearly not as beneficial as GAC fluidization within the membrane bioreactor. In order to determine the relationship between membrane fouling and the quantity of GAC used, a short-term test of TMP change with time in the AFMBR was conducted for 578

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

FIGURE 4. Combined effect of recirculation rate and total weight of GAC on TMP at a permeate flux of 15 L/m2/h. three different reactor GAC concentrations (225, 300, and 450 g). ABR effluent was continuously fed to the reactors while the permeate flux was set at 15 L/m2/hr. For this test, the ABR influent COD was 535 mg/L, and the effluent contained 23.1 ( 5.9 mg/L TCOD, 15.2 ( 3.4 mg/L SCOD, 6.1 ( 1.4 mg/L TSS, and 5.2 ( 1.5 mg/L VSS. In order both to prevent the GAC particles from leaving the reactor and yet remain suspended throughout the reactor to cover the membranes, the recirculation rate had to be set higher for the lower GAC concentrations, and for this reason it was set at 2.4, 1.8, and 1.4 L/min for each, respectively. Three controls without GAC had the same set membrane flux and the same different recirculation rates in order to determine the effect of hydraulic shear alone. The results illustrated in Figure 4 indicate that all GAC levels significantly reduced membrane fouling below that of their respective recirculation controls. However, the control results illustrate that fouling can be reduced to some extent by the higher shear rate induced by greater recirculation alone. With the three GAC reactors, a higher GAC concentration resulted in a marginally lower suction pressure, but higher concentrations come at higher chemical cost. The energy and cost relationships for the different alternatives evaluated here are complex, but note that the energy for GAC fluidization is related to the product of the GAC mass and the recirculation ratio, a product that was little different in the three GAC cases. More study on these relationships is appropriate. Long-Term Test: Continuous Two-Stage System Operation. The short-term tests indicated that GAC fluidization is effective in reducing membrane fouling in the AFMBR.

FIGURE 5. Effect of GAC fluidization on TMP during continuous operation of the two-stage system consisting of the AFBR and AFMBR connected in series. Operation over a longer time period was then conducted to observe the sustainability of operation. Here, the two-stage AFBR/AFMBR system (Figure 1) was operated continuously for 120 days, using a permeate flux of 7 L/m2/h during the first 4 days, increasing to 10 L/m2/h thereafter. Included for comparison was a short test at 10 L/m2/h to indicate TMP change with the two-stage system, but without GAC fluidization. Results of TMP analyses for long-term operation are illustrated in Figure 5. Without GAC fluidization, TMP increased rapidly to 0.32 bar within 0.5 d. However, with GAC fluidization such fouling did not occur, and with the initial flux rate of 7 L/m2/h, TMP stabilized at 0.03 bar. It then increased to 0.05 bar as the permeate flux was increased to 10 L/m2/h. TMP ranged from 0.05 to 0.07 bar during the first 20 days. Because of an operational error at 20 days, some of the GAC particles were broken by the recirculation pump when they overflowed into the recirculation line shown in Figure 1. Following this event the TMP increased somewhat rapidly to 0.18 bar after 40 days of operation. Backwashing with membrane permeate was performed at 40 days, but this did not alleviate the membrane fouling. A significant rather sudden increase in TMP occurred at day 52. The membrane module was removed from the reactor for chemical cleaning at day 54, during which the top section of the reactor was sealed with an acrylic cover to prevent oxygen entry. The membranes were soaked in 1000 ppm NaOCl for 1 h and then in 500 ppm NaOH solution for 1 h after rinsing with DI water. After the chemical cleaning, TMP decreased to 0.08 bar, but increased rapidly to 0.25 bar within 5 day operation following a similar incidence of GAC particles overflowing into the recirculation line. In order to slowly remove the finer material believed to cause fouling as well as to remove excess suspended solids, 80 mL of liquid (equivalent to about 4% of the AFMBR reactor volume) was taken daily from the recirculation line beginning on day 65. Chemical cleaning was again performed at day 87. Subsequently, the TMP stabilized at less than 0.1 bar for the final 40 days of operation without further chemical cleaning being required. Figure 5 shows clearly that daily removal of TSS from the recirculation line helped reduce membrane fouling significantly. The concentration of TSS in the 80 mL of liquid removed from the recirculation line each day averaged 290 mg/L, which is significantly less than the 11 600 mg/L found attached and suspended within the AFMBR itself at the end of this study. Table 1 contains a summary of the performance of the AFBR, the AFMBR, and the overall two-stage AFBR/AFMBR system during the 120 days of operation. The AFBR achieved an average total COD removal of 88%, while the AFMBR reduced its influent TCOD by an additional 87%, providing an overall COD removal for the two-stage system of 99%.

VFA removal by the AFBR alone was 93%, nearly 100% overall. Excess biosolids production increased the effluent TSS and VSS concentrations over the influent values for the AFBR, but both concentrations were reduced to near zero in the AFMBR effluent. During the last 77 d with operation at 2 h HRT, the combined methane production from the gas and liquid phases for the AFBR was 5.88 mol CH4/m3 wastewater, which is equivalent to 83% of the removed COD (21.2 g COD/day) and 73% of the added COD (24.1 g COD/day). Of the produced methane, 70% exited with the gas phase and the rest remained dissolved in the AFBR effluent (see Supporting Information). In the AFMBR, no gaseous methane production was observed, as the influent COD was so low. Figure 6 illustrates the variability in the AFBR effluent SCOD and TSS, which served as influent feed to the AFMBR. The high influent SCOD over days 10-13 and 42-45 resulted from inadvertent excess COD feeding to the AFBR. The AFMBR responded very well as a polishing unit, removing this excess COD, allowing only a small increase in its effluent SCOD. Results show clearly that using an AFMBR for posttreatment of effluent from a first-stage anaerobic system treating a dilute wastewater can play a significant polishing role to produce high effluent COD and TSS qualities. Energy Requirements. Energy requirements for reducing membrane fouling is an important aspect of AMBR operation to understand. Energy usage for the AFBR/AFMBR system was evaluated using the pump power requirement equation (16) P)

QγE 1000

(1)

where P is power requirement (kW), Q is flow rate (m3/s), γ is 9800 N/m3, and E represents the hydraulic pressure head (m). For the AFMBR, Q equals the reactor recycle rate of 1.4 L/min (2.33 × 10-5 m3/s), and the measured hydraulic pressure head loss through the system was 0.098 m, yielding a power requirement of 2.24 × 10-5 kW. Dividing by the permeate flow rate of 0.91 L/h (9.1 × 10-4 m3/h) yields a pumping energy requirement of 0.025 kWh/m3. Added to this is the average energy loss of permeate flow through the membrane of 0.1 bar, equivalent to a hydraulic head loss E of 1.0 m and an energy requirement of 0.003 kWh/m3, giving a total for the AFMBR of 0.028 kWh/m3. This is only a small fraction of the reported 0.25-1.0 kWh/m3 energy required in AMBRs that use gas sparging to prevent membrane fouling (6). The potential energy advantage of the AFMBR is apparent, even before optimization for further energy reduction. Here, the first stage AFBR also requires energy. Comparable values are Q of 2.0 L/min, E of 0.18 m, and influent flow rate of 1.96 L/h, yielding a pumping energy requirement of 0.030 kWh/ m3. For the combined two-reactor system, the pumping energy requirement was 0.058 kWh/m3. Gaseous methane produced in the AFBR alone was 4.11 mol CH4/m3 wastewater. With an energy yield through methane combustion of 800 kJ/mol, this represents 3290 kJ/m3 or 0.91 kWh/m3, about 16 times the pumping requirement for the combined AFBR/AFMBR system. However, energy conversion efficiencies need to be considered. About 33% of the methane energy can be converted into electrical energy, electric motors to drive axial pumps are about 80% efficient, pump impeller about 85% efficient, and inverter about 95% efficient, giving an overall conversion efficiency of about 21%. Thus, from the gaseous methane recovered, 0.19 kWh/m3 of energy for fluidization could be produced. The total fluidization energy requirement of 0.058 kWh/m3 is only 30% of this amount. Optimization of operation should be able to reduce this energy utilization further. In summary, the AFMBR used for post-treatment of effluent from an AFBR produced an excellent polished VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

579

TABLE 1. Performance of the AFBR, AFMBR, and Overall Two-Stage AFBR/AFMBR System (total 120 days of operation) AFBR analysis TCOD SCOD TSS VSS VFA as COD gas production (mol CH4/m3)a CH4 composition (%)a a

influent (mg/L)

effluent (mg/L)

removal (%)

effluent (mg/L)

removal (%)

overall removal (%)

513((36) 500((41) 14((5) 12((4) 500

59((31) 42((24) 15((6) 14((5) 36((33)

88((2) 91((5) 93((2)

7((4) 7((4) 0.0 0.0 97

99((1) 99((1) 100 100 >99((0)

4.11((0.2) 86((7)

Methane production and composition are for last 77 days of operation.

FIGURE 6. Removal behavior of SCOD and TSS in the AFMBR component of the two-stage AFBR/AFMBR system at a permeate flux of 10 L/m2/h. effluent with continuous operation over 120 days of operation, providing about 99% overall COD reduction. The AFMBR energy requirement was only 0.028 kWh/m3, much less than reported for AMBRs using internal membranes and gas sparging and only a small fraction of the methane energy produced. The effluent from the AFBR that preceded the AFMBR was demonstrated to cause rapid and serious fouling of the membranes without the action of the fluidized GAC particles against the membrane. The major benefit of the fluidized GAC appears to be scouring action on the membrane surface, although more detailed studies are needed to fully understand the degree of significance of other factors such as hydraulic shear and the removal of colloidal or other potentially fouling materials from solution by the GAC (13, 14). To be noted is that this was an initial feasibility study only to evaluate whether fluidized bed operation in conjunction with membrane filtration had potential energy benefits for reducing membrane fouling. It appears from the results that it has. Further study is required to optimize AFMBR performance for reducing energy usage, which might be accomplished by increasing the number of membrane fibers from the 16 used here to perhaps several times that or allowing 580

two-stage (AFBR/AFMBR) system

AFMBR

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 45, NO. 2, 2011

greater permeate flux with no increase in HRT and little increase in reactor energy cost. Particulate materials other than GAC with lower specific gravity would have less energy requirements and thus offer other possibilities if they also prove effective at reducing fouling. The AFMBR was used for polishing of the AFBR effluent, reducing the soluble COD from 30 to 60 mg/L down to 5 to 10 mg/L. The TSS was reduced to near zero. Biological action, sorption onto the GAC, and membrane filtration were probably each involved to some extent in the removals obtained. Soluble COD removal essentially started in the AFMBR with no lag time, microorganisms carried over from the AFBR perhaps serving as adequate seed for rapid startup of the biological component of the membrane system. Thus, addition of an AFMBR to an existing anaerobic treatment system for post-treatment might result in the rapid development of a good polishing system. Also, the AFMBR by itself may be sufficient as a stand-alone treatment system, something not yet attempted. An additional benefit of a fluidized bed reactor with GAC is the potential to remove refractory trace contaminants of concern through sorption, as well as biodegradation of the readily biodegradable organic materials, an aspect yet to be studied. With dilute wastewaters, excessive energy would be required to increase water temperature from ambient to 35 °C, the temperature used here, and doing so would not be practical. Also, complex wastes are likely to have a somewhat lower methane yield per unit of BOD treated than acetate and propionate because of greater cell yield so that the energy benefits might be somewhat less than shown here. Future research with the AFMBR to address many questions that arise is thus needed before practical application can be undertaken for dilute wastewaters containing more complex substrates and at ambient temperatures. Potential energy cost savings need to be balanced by capital and operating costs for GAC. A better fundamental understanding of how GAC reduces membrane fouling is needed, as is the longterm impact of GAC fluidization on membrane structure, strength, and effectiveness.

Acknowledgments This research was supported by the World Class University Project through the National Research Foundation of Korea, funded by the Ministry of Education, Science and Technology (grant number R33-2008-000-10043-0). Membrane materials were provided by KOLON Inc.

Supporting Information Available A table listing detailed performance of the AFBR at the two detention times used over the 120 day period of continuous operation. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Foresti, E.; Zaiat, M.; Vallero, M. V. G. Anaerobic processes as the core technology for sustainable domestic wastewater treatment: Consolidated applications, new trends, perspectives, and challenges. Rev. Environ. Sci. Bio/Technol. 2006, 5, 3–19. (2) Langenhoff, A. A. M.; Intrachandra, N.; Stuckey, D. C. Treatment of dilute soluble and colloidal wastewater using an anaerobic baffled reactor: Influence of hydraulic retention time. Water Res. 2000, 34 (4), 1307–1317. (3) Chernicharo, C. A. L. Post-treatment options for the anaerobic treatment of domestic wastewater. Rev. Environ. Sci. Bio/ Technol. 2006, 5, 73–92. (4) Fang, H. H. P.; Shi, X. L.; Zhang, T. Effect of activated carbon on fouling of activated sludge filtration. Desalination 2006, 189 (1-3), 193–199. (5) Berube, P. R.; Hall, E. R.; Sutton, P. M. Parameters governing permeate flux in an anaerobic membrane bioreactor treating low-strength municipal wastewaters: A literature review. Water Environ. Res. 2006, 78 (8), 887–896. (6) Liao, B. Q.; Kraemer, J. T.; Bagley, D. M. Anaerobic membrane bioreactors: Applications and research directions. Crit. Rev. Environ. Sci. Technol. 2006, 36 (6), 489–530. (7) Meng, F. G.; Chae, S. R.; Drews, A.; Kraume, M.; Shin, H. S.; Yang, F. Recent advances in membrane bioreactors (MBRs): Membrane fouling and membrane material. Water Res. 2009, 43 (6), 1489–1512. (8) Ramesh, A.; Lee, D. J.; Wang, M. L.; Hsu, J. P.; Juang, R. S.; Hwang, K. J.; Liu, J. C.; Tseng, S. J. Biofouling in membrane bioreactor. Sep. Sci. Technol. 2006, 41 (7), 1345–1370.

(9) Yang, Q. Y.; Chen, J. H.; Zhang, F. Membrane fouling control in a submerged membrane bioreactor with porous, flexible suspended carriers. Desalination 2006, 189 (1-3), 292302. (10) Rittmann, B. E.; McCarty, P. L. Environmental Biotechnology: Principles and Applications; McGraw-Hill: New York, 2001. (11) Le-Clech, P.; Chen, V.; Fane, T. A. G. Fouling in membrane bioreactors used in wastewater treatment. Membr. Sci. 2006, 284 (1-2), 17–53. (12) Park, H.; Choo, K. H.; Lee, C. H. Flux enhancement with powdered activated carbon addition in the membrane anaerobic bioreactor. Sep. Sci. Technol. 1999, 34 (14), 2781–2792. (13) Ng, C. A.; Sun, D.; Fane, A. G. Operation of membrane bioreactor with powdered activated carbon addition. Sep. Sci. Technol. 2006, 41 (7), 1447–1466. (14) Hu, A. Y.; Stuckey, D. C. Activated carbon addition to a submerged anaerobic membrane bioreactor: Effect on performance, transmembrane pressure, and flux. J. Environ. Eng. 2007, 133 (1), 73–80. (15) Clesceri, L. S.; Greenberg, A. E.; Eaton, A. D. Standard Methods for the Examination of Water and Wastewater, 20 ed.; American Public Health Association: Washington, DC, 1998. (16) Vennard, J. K.; Street, R. L. Elementary Fluid Mechanics; John Wiley & Sons: New York, 1982.

ES1027103

VOL. 45, NO. 2, 2011 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

581