Effects of internal recycling time mode and hydraulic ... - Springer Link

Bioprocess Biosyst Eng (2009) 32:135–142 DOI 10.1007/s00449-008-0232-6

ORIGINAL PAPER

Effects of internal recycling time mode and hydraulic retention time on biological nitrogen and phosphorus removal in a sequencing anoxic/anaerobic membrane bioreactor process Kyung-Guen Song Æ Jinwoo Cho Æ Kyu-Hong Ahn

Received: 29 January 2008 / Accepted: 1 May 2008 / Published online: 31 May 2008 Ó Springer-Verlag 2008

Abstract This study investigated the effects of internal recycling time mode and hydraulic retention time (HRT) on nutrient removal in the sequencing anoxic/anaerobic membrane bioreactor process. Denitrification and phosphorus release were reciprocally dependent on the anoxic/ anaerobic time ratio (Ax/An). As Ax/An increased, nitrogen removal rate increased but phosphorus removal rate decreased. The increasing Ax/An provided the longer denitrification period so that the organic substrate were consumed more for denitrification rather than phosphorus release in the limited condition of readily biodegradable substrate. Decreasing HRT increased both nitrogen and phosphorus removal efficiency because as HRT decreased, food-to-microorganism loading ratio increased and thus enhanced the biological capacity and activity of denitrifying bacteria. This could be verified from the observation mixed liquor suspended solids concentration and specific denitrification rate. The change of Ax/An and HRT affected phosphorus removal more than nitrogen removal due to the limitation of favourable carbon source for phosphorus accumulating organisms. Keywords Sequencing anoxic/anaerobic membrane bioreactor Phosphorus removal Nitrogen removal Internal recycling time mode Hydraulic retention time

K.-G. Song (&) J. Cho K.-H. Ahn Center for Environmental Technology Research, Korea Institute of Science and Technology, PO BOX 131, Cheongryang, Seoul 130-650, South Korea e-mail: [email protected]

Introduction Membrane bioreactor (MBR) is considered to be the most efficient process for wastewater treatment because of its performance and compactness. In MBR, complete separation of biomass by membrane filtration makes it possible to maintain high concentration of mixed liquor suspended solids (MLSS) in the reactor, which causes long sludge retention time (SRT), reduced sludge production, and enhanced nitrification [1–7]. With the strict regulation on the nitrogen and phosphorus in discharge, these nutrient removal has become the focus of recent studies on MBR [8, 9]. These researches have introduced the alternating anoxic and oxic conditions in a submerged MBR by intermittent aeration for simultaneous removal of carbon and nitrogen without phosphorus removal. Enhanced biological phosphorus removal (EBPR) in MBR was attempted by several researchers, and showed efficient phosphorus removal [10– 12]. However, the system configuration of MBR for EBPR was complex. Ahn et al. [13] developed an innovative simple configuration for both nitrogen and phosphorus removal, called sequencing anoxic/anaerobic membrane bioreactor (SAM) and demonstrated its capability to remove the nitrogen and phosphorus in domestic wastewater, compared to the modified Ludzack–Ettinger type MBR. Because of organic substrate competition between phosphorus accumulating organisms (PAOs) and denitrifying bacteria, it has been recommended that the anaerobic and anoxic conditions should be spatially compartmentalized to achieve an efficient removal of nitrogen and phosphorus in conventional EBPR process [14]. However, in the SAM process, the anaerobic and anoxic conditions are induced in one reactor by introducing the intermittent

123

136

Bioprocess Biosyst Eng (2009) 32:135–142

recycling of the mixed liquor containing nitrate from aerobic reactor (AR) to sequencing anoxic/anaerobic reactor (SAAR). For biological nutrients removal processes, hydraulic retention time (HRT) in each reactor is an important factor for reliable denitrification and phosphorus release. Wareham and Lee [15] observed that the reduced retention time in anaerobic reactor caused deterioration of phosphorus removal in a lab scale UCT process. In the SAM process, the lengths of anaerobic and/or anoxic time period represent the same meaning with the retention time in the anaerobic and anoxic reactors of conventional EBPR process. Therefore, in this study, ratio of anoxic to anaerobic time (Ax/An) is considered as one of the most important parameters for removal of nitrogen and phosphorus. Because the MBR process is operated with higher biomass concentration than a conventional activated sludge processes, the MBR is competent for a higher substrate loading rate. If substrate concentration of influent is constant, substrate loading rate can be controlled by changing HRT. Increase of HRT might yield excellent nutrient removal and low food-to-microorganism loading (F/M) ratio by increased biomass concentration [16, 17]. However, increased HRT generates less substrate gradient for denitrification and phosphorus release under substrate limited condition. Therefore, HRT should be considered an important factor in achieving better nutrient removal efficiency in the SAM process. In this study, two important parameters were the ratio of anoxic to anaerobic time (Ax/An) induced by alternating internal recycling operation and HRT. The main objective of this study is to investigate the effect of these parameters on the performance of SAM process.

continuously fed into the SAAR. The distinctive feature of the SAM is the intermittent internal recycle of the mixed liquor directly from the AR to SAAR. This way of recycling induces different conditions at the SAAR. Recycling causes nitrate accumulation in the SAAR, and then denitrification takes place under the anoxic condition. However, no recycling caused oxygen deficiency as well as nitrate in the SAAR to induce the anaerobic condition. This condition leads PAOs to release phosphorus. In the AR, nitrification by nitrifying bacteria, such as nitrosomonas and nitrobactor, and phosphorous uptake by PAOs takes place. Therefore, simultaneous nitrogen and phosphorus removal in SAM system are achieved. The volume of the SAAR and AR was 4 and 6 L, respectively. The MF membrane module (Kubota, Japan) was fully immersed inside the AR. The membrane has an effective filtration area of 0.1 m2 and its nominal pore size is 0.2 lm. The airlift was installed underneath the membrane module to provide the aerobic condition and a crossflow effect near the membrane surface to minimize membrane fouling. The membrane was operated intermittently to minimize the membrane fouling; 8 min suction and 2 min rest. The constant flux was maintained and the transmembrane pressure (TMP) was continuously monitored. The main purpose of TMP monitoring is simply for cleaning the fouled membrane on time to keep the influent and effluent discharging rate constant. When the TMP exceeded 15 kPa, chemical cleaning using NaOCl 8% was performed. Domestic wastewater from an apartment complex is used for the influent. Table 1 presents the characteristics of influent used for each runs. Internal recycling time mode

As shown in Fig. 1, the SAM system was composed of SAAR and AR where the flat-sheet microfiltration membrane module was immersed. The influent was

The objective of this experiment is to investigate the nitrogen and phosphorus removal efficiency with the different internal recycling configurations. Three lab scale SAMs are operated at the same condition except the recycling time mode. The recycling time mode determines the ratio of anoxic to anaerobic period of SAAR. In this experiment, three different recycling time modes are

Fig. 1 Operation of the SAM process at anoxic phase (a) and anaerobic phase (b)

Effluent

Materials and methods Sequencing anoxic/anaerobic membrane bioreactor

(a)

(b)

Sequencing Reactor

Influent

Influent

Anoxic

Effluent

Anaerobic An a e ro b ic M e m b ra n e Membrane

Internal recycle PPump ump

Blo w e r Blower

Me m b ra n e Membrane No Internal recycle

Blo w e r Blower

Aerobic Reactor

Sludge Wasting

123


137

the influent flow rate, which have effects on the membrane fouling and the substrates loading rate on the system. The recycling time mode was fixed to have the anoxic condition for 3 h and the anaerobic for 1 h. The recycling flow rate is 2.67Q during anoxic period. Thus, the averaged flow rate of recycling is 2Q for a day.

Table 1 Characteristics of the influent Run # 1–3 (Ax/An)

Run #4–6 (HRT)

COD (mg/L)

178 (56)

196 (63)

SCOD (mg/L)

114 (46)

128 (52)

TN (mg/L)

35 (13)

25 (2)

TP (mg/L)

3.6 (1.2)

3.2 (1.0)

The values in parentheses are standard deviations

Analytical methods

designed as described in Table 2. In run #1, the anoxic condition in SAAR lasts for 2 h in every 3 h with internal recycling. No internal recycling induced the anaerobic condition for 1 h in every 3 h. The condition of SAAR changes eight times for a day. Therefore, run #1 totally has the anoxic condition in SAAR for 16 h a day and the anaerobic for 8 h a day. The operation of run #1 focuses on nitrogen removal rather than phosphorus. While run #3 is designed to induce high phosphorus release in the SAAR by operating the system with the anaerobic period three times longer than the anoxic period as shown in Table 2. The recycling flow rates for the anoxic time of run #1, #2, and #3 are maintained to be three, four, and eight times of the influent flow rate (3Q, 4Q, and 8Q), respectively. Considering no recycling in the anaerobic condition, the averaged flow rate of each reactor is as same as 2Q. Therefore, the total mass of recycled mixed liquor for a day is same, even though the recycling time is different from each other reactors. Solids retention time is 78 days. This long SRT enables the high level of MLSS in the reactor. The averaged MLSS concentration of run #1, #2, and #3 was maintained at about 8,400, 8,800, and 8,000 mg/L, respectively.

Ammonia was measured using the Auto Analyzer 3 (Bran Luebbe, Germany). Nitrite, nitrate and ortho-phosphate were determined using ion-chromatograph system (IC DX– 120, Dionex, USA). Total nitrogen, total phosphate and COD were measured using the Hach Digestion Vials (Hach, Co., USA) and the Beckman Spectrophotometer (DU 520; Germany). Dissolved oxygen concentration was measured using the DO meter (Model 58; YSI, USA). pH was determined using a glass electrode pH meter (Model 525A; Orion, USA).

Results and discussion The effect of internal recycling time mode on nutrients removal The experiments were conducted at three different internal recycling time modes. As shown in Fig. 2, average COD removal efficiencies for all runs were above 90% regardless of anoxic/anaerobic time mode. However, nitrogen and phosphorus removal efficiencies were dependent on Ax/An ratio, reciprocally. As Ax/An ratio increased, nitrogen removal efficiency increased, whereas phosphorus removal efficiency decreased. The higher Ax/An allowed more time for denitrification and for supplying substrate for denitrification and thus enhanced nitrogen removal [9]. On the other hand, for phosphorus removal, the higher Ax/An resulted in relatively reduced retention time under anaerobic condition and caused deterioration of phosphorus release and removal [15]. It is agreed with the finding of

Hydraulic retention time This experiment was conducted to show the influence of the substrates loading rate on the nitrogen and phosphorus removal efficiency. The overall HRT varies at 3, 5, and 8 h for run #4, #5, and #6, respectively, as presented in Table 3. The increases of HRT results in the higher permeate flux and

Table 2 Experiment conditions for the effect of internal recycling time mode on nutrient removal Run #1

Run #2

Run #3

Influent flow rate, Q

30 L/day

30 L/day

30 L/day

HRT (total) Internal recycling time mode (anoxic:anaerobic)

8h 2 h:1 h

8h 2 h:2 h

8h 1 h:3 h

Ax/An ratio

2.0

1.0

0.3

Averaged recycling flow ratea

2Q (3Qb)

2Q (4Qb)

2Q (8Qb)

SRT

78 days

78 days

78 days

a b

anoxic time Averaged recycling flow rate = recycling flow rate anoxic time þanaerobic time Recycling flow rate during anoxic time period

123

138


Table 3 Experiment conditions for the effect of HRT on nutrients removal Run #4

Run #5

Run #6

Influent flow rate

78 L/day

52 L/day

30 L/day

Permeate flux

30 L/m2/h

18 L/m2/h

11 L/m2/h

HRT (total)

3h

5h

8h

HRT of AR

2h

3h

5h

HRT of SAAR

1h

2h

3h

Internal recycling time mode (anoxic:anaerobic)

3 h:1 h

3 h:1 h

3 h:1 h

Averaged recycling flow ratea

2Q (2.67Qb)

2Q (2.67Qb)

2Q (2.67Qb)

SRT

78 days

78 days

78 days

a b

anoxic time Averaged recycling flow rate ¼ recycling flowrate anoxic time þanaerobic time Recycling flow rate during anoxic time period

Concentration (mg/L)

100 Ax/An = 2.0 Ax/An = 1.0 Ax/An = 0.3

(a)

Anoxic

Anaerobic

Phosphate Ammonia Nitrate

15 10 5 0 0

30

60

90

120

150

180

60


Removal Efficiency (%)

80

20

40

20

T-N

T-P

Fig. 2 Removal efficiencies of COD, TN, and TP at various Ax/An ratios. Twenty-four samples were collected for 69 days of operation duration. Error bar represents standard deviation

Randall that increasing in anaerobic HRT induced increased EBPR performance for COD limiting conditions [16]. However, in this result, the interesting thing was that Ax/An ratio was more influential on phosphorus removal than on nitrogen removal. As shown in Fig. 2, when Ax/An decreased from 2 to 0.3, the change in phosphorus removal efficiency was 24% (from 57 to 33%) but that in nitrogen removal efficiency was only 12% (from 56 to 44%). Unlike most other microorganisms including denitrifying bacteria, PAOs thrive on short chain fatty acids like acetate and propionate [17, 18]. The influent used in this study had lower SCOD concentration (shown in Table 1), comparing with the influent in previous study [13] since for the previous study, glucose and acetic acid were added to the influent to provide additional readily biodegradable organic substrate. It presumed that the influent of this study


COD

Anaerobic

5 0 0

0

Anoxic

10

20

123

(b)

15

20

30

60

90

120

150

180

210

240

180

210

240

Time (min) Anaerobic

(c) Anoxic

15 10 5 0 0

30

60

90

120

150

Time (min)

Fig. 3 Variation of the phosphate, ammonia and nitrate concentration in the sequencing anoxic/anaerobic reactor of the SAM process for run #1 (a), run #2 (b), and run #3 (c). Samples for track study were taken from SAAR at 62nd day

contained relatively insufficient amount of readily biodegradable organic substrate. The limitation of favourable carbon source for PAOs could make phosphorus removal more sensitive on the condition changes than nitrogen removal. Figure 3 shows the changes of phosphate, ammonia and nitrate concentration in the SAAR for a cycle time on run #1, #2, and #3. In the case of run #1 emphasizing nitrogen removal, the phosphorus release did not happen during the anaerobic period because denitrification was not completed


for 1 h of anaerobic period (no recycling mode). For the other runs, however, anaerobic period (no recycling time) was long enough for the remaining nitrate to be exhausted by complete denitrification, then the phosphorus release occurred throughout for the rest of anaerobic time. In case of run #2, the nitrate concentration became 0 at 210 min, and then the anaerobic condition was induced, and phosphorus release took place for 30 min. For run #3, denitrification was finished at 170 min and then phosphorus release happened during the rest of anaerobic period (100 min). Therefore, the run #3, which had the longest time remained for phosphorus release after completion of denitrification, showed the highest final concentration of released phosphorus at the end of the anaerobic period and the most excellent phosphorus removal efficiency. This clearly shows that the presence of nitrate inhibited the phosphorus release [19]. Nitrate is usually considered to be inhibitive to biological phosphorus removal activity, since the nitrate introduced to the anaerobic zone can be denitrified in this zone, thereby reducing the supply of organic substrates available for PAOs. This reduces the amount of substrate available for PAOs, and hence causes the reduction of phosphorus release [20, 21]. If influent contains significant amounts of fermented product such as acetate, both denitrification and phosphorus release take place simultaneously [22], whereas for less presence of fermented substrate, phosphorus release takes place after exhausting nitrate [19]. As stated above, the influent used in this study may have relatively insufficient amount of readily biodegradable organic substrate. Hence, the insufficiency of fermented organic substrate in the influent could interfere with the simultaneous denitrification and phosphorus release.

139

phosphorus release. Oppositely, the ‘long’ HRT condition is expected to contribute the good nutrient removal efficiency. However, the results from this experiment shows the removal efficiencies of nitrogen and phosphorus increase with the decrease of HRT, as presented in Fig. 4. Cho et al. [26] also reported that increasing HRT by 30% resulted in a decrease of nitrogen and phosphorus removal efficiency by 10% in a pilot scale experiment applied to municipal wastewater treatment. The short HRT of run #4 implies higher F/M ratio to the system, compared to the other runs. The F/M ratios of run #4, #5, and #6 were 0.102, 0.078, 0.053 g COD/g MLSSday, respectively at steady state. This condition keeps the yield of biomass being high, resulting in the large amount of MLSS at the same SRT. The total amounts of MLSS in run #4, #5, and #6 are measured around 150, 130, and 110 g, respectively at steady state. Therefore, the biological capacity of denitrification and phosphorus release would be improved, because high MLSS can contain a larger number of microorganisms, and the bioactivity of biomass in high yield condition is greater than in endogenous phase in substrate deficient condition. Burdick et al. [27] indicated that the specific rate of denitrification in anoxic zone was related to the F/M ratio on the anoxic zone and specific rates of denitrification in anoxic zone typically ranged from 0.05 to 0.15 gNOx-N/g MLSS day. In order to investigate the relationship between F/M ratio (related with HRT) and bioactivity on the denitrification, a batch denitrification test to calculate the maximum specific denitrification rate (SDNR) was

100 HRT = 3 hr HRT = 5 hr HRT = 8 hr

The effect of HRT on nutrients removal

Removal Efficiency (%)

80

Figure 4 illustrates the overall performance on organic substances and nutrients removal. The removal efficiency of COD is 97% regardless of HRT. It appeared that the MBR process showed a significant contribution to COD removal due to the complete retention of all particulate and macromolecular COD components by the membrane [23, 24]. However, in the biological aspect, HRT was closely associated with biological removal efficiencies. In general, the biological removal efficiency seemed to decrease with short HRT because of both the substrate loading rate and F/ M ratio increased, and insufficient time to oxidize substrates with bacteria [23]. Chae et al. [25] also reported that short HRT increased ammonium-nitrogen and nitratenitrogen concentration in the effluent because of its insufficiency for complete nitrification and denitrification. Additionally, phosphorus removal efficiency become low because remaining nitrate-nitrogen will interfere with

60

40

20

0 COD

T-N

T-P

Fig. 4 Removal efficiencies of COD, TN, and TP at various HRTs. Nineteen samples were collected for 63 days of operation duration. Error bar represents standard deviation

123

140


Table 4 Specific denitrification rate for various HRTs SDNR (gNO3-N/gVSS/day)

20

(a)


Anaerobic

Phosphate Nitrate Nitrite

10 5 0 0

30

60

90

120

150

180

210

240

20

(b)

Anoxic

Run #6 (HRT = 8 h)

0.18

0.13

0.06

denitrification is less than run #4. The nitrate becomes 0 around at 220 min. Thus, the duration of anaerobic condition of run #5 is slightly less than run #4 resulting in less phosphorus release and the removal efficiency of phosphorus also becomes lower than run #4. It is evident that the higher F/M ratio induced by shorter HRT increase denitrification rate by enhancing the biological capacity and activity of denitrifying bacteria. Hence, the increased denitrification rate lowered the concentration of nitrate in the reactor and shortened the time for complete denitrification, resulting in higher nitrogen and phosphorus removal efficiency, consequently. Membrane performance Figure 6 presents the variation of TMP in each condition during the experiments. As mentioned above, the main purpose of TMP monitoring was simply for cleaning the fouled membrane on time to keep the influent and effluent discharging rate constant. Thus, the effect of internal recycling time mode and HRT on the nutrients removal could be investigated independently from the membrane fouling. The cleaning time, cleaned run, and operation stop due to accidents from a mechanical problem are illustrated in Fig. 6. However, it can be noted that the TMP of each reactor shows little fluctuation and similar trend below 15 kPa throughout the experiments. Previous researches have investigated the effects of operational parameters such as MLSS, SRT, and HRT on membrane fouling [28, 29]. Meng et al. [24] reported that

Anaerobic

15 80

10

Run #1 Run #2 Run #3

5 0 0


Anoxic

15

Run #5 (HRT = 5 h)

30

60

90

120

150

180

210

240

20

(c)

Anoxic

Anaerobic

15 10 5 0 0

30

60

90

120

150

180

210

Transmembrane Pressure (kPa)


performed. Table 4 summarized the maximum SDNR for each run. The maximum SDNR proportionally increased with increasing F/M ratio. It is evidently implied that the slower denitrification rate is due to the lower substrate loading rate, induced by the longer HRT, resulting in lower MLSS and biological capacity under substrate deficient condition. This reasoning was also supported by track studies. Figure 5 shows the results of track study on run #4, #5, and #6 from the beginning of anoxic condition to the end of anaerobic condition. In these figures, it is clear that the SAAR of run #4 and #5 show the phosphorus release during the anaerobic condition, while run #6 cannot. In case of run #6, the denitrification still takes place even at the end of the anaerobic period (240 min). This means that the anaerobic condition cannot be induced by the inhibition of remaining nitrate, so that no phosphorus release happens. However, in run #4, nitrate start decreasing rapidly right after the end of anoxic time (180 min). At around 195 min, the nitrate concentration becomes 0, and then the anaerobic condition is induced. In run #5, the rate of

Run #4 (HRT = 3 h)

60 Operation suspended

40 Cleaning #1, 2, 3

20

240

Time (min)

0 20

Fig. 5 Variation of the phosphate, nitrate and nitrite concentration in the sequencing anoxic/anaerobic reactor of the SAM process for run #4 (a), run #5 (b), and run #6 (c). Samples for track study were taken from SAAR at 57th day

123

Cleaning #1 Cleaning #2, 3

30

40

50

60

70

80

90

100

Time (day)

Fig. 6 Variation of transmembrane pressure in each condition during the experiments


HRT had no direct effect on membrane fouling but had impacts on the change of biomass characteristics because of its close relation with DO and oxygen uptake rate of biomass. In this study, however, the impact of these factors on membrane fouling cannot be discussed as mentioned above. Further study is required for evaluating the effect of internal recycling time mode and HRT on membrane fouling.

Conclusions This study showed that denitrification and phosphorus release were dependent on Ax/An ratio. As Ax/An ratio representing relative time length of anoxic period to anaerobic period increased, nitrogen removal efficiency increased because of providing extended time for denitrification, while phosphorus removal efficiency decreased because phosphorus release was delayed by remaining nitrate during the anaerobic time, and thus uptake of phosphorus in the following aerobic stage also reduced, when the readily biodegradable COD content of influent was limited. In the SAM process, the higher F/M ratio induced by the shorter HRT leaded the increase of biomass yield as well as biomass activity, especially in denitrification rate, which resulted in improving removal efficiency of nitrogen and phosphorus also. The effect of change in both of Ax/An and HRT was greater on phosphorus than on nitrogen due to the limitation of favourable carbon source for PAOs. However, COD removal was higher than 90% regardless of change of operation conditions including Ax/An and HRT because the membrane significantly contributed to remove COD by the complete retention of all particulate and macromolecular COD components. Acknowledgments This study was supported by the Korea Institute of Science and Technology (KIST) in Korea.

References 1. Smith CV, Gregorio DD Jr, Talcott RM (1969) The use of ultrafiltration membrane for activated sludge separation. In: Proceedings of 24th annual Purdue industrial waste conference, Purdue University, West Lafayette 2. Chaize S, Huyard A (1991) Membrane bioreactor on domestic wastewater treatment: sludge production and modelling approach. Water Sci Technol 27:171–178 3. Ahn K-H, Cha H-Y, Song K-G (1999) Retrofitting municipal sewage treatment plant using an innovative membrane-bioreactor system. Desalination 124:279–286

141 4. Bouillot P, Canales A, Pareilleux A, Huyard A, Goma G (1990) Membrane bioreactor for evaluation of maintenance phenomena in wastewater treatment. J Ferment Bioeng 69(3):178–183 5. Stephenson T, Judd S, Jefferson B, Brindle K (2000) Membrane bioreactors for wastewater treatment. IWA publishing, London 6. Cicek N, Franco JP, Suidan MT, Urbain V, Manem J (1999) Characterization and comparison of a membrane bioreactor and a conventional activated-sludge system in the treatment of wastewater containing high-molecular-weigh compounds. Water Environ Res 71(1):64–70 7. Cicek N (2003) A review of membrane bioreactors and their potential application in the treatment of agricultural wastewater. Can Biosyst Eng 45:6.37–6.49 8. Ueda T, Hata K (1999) Domestic wastewater treatment by a submerged membrane bioreactor with gravitational filtration. Water Res 33:2888–2892 9. Nah YM, Ahn KH, Yeom IT (2000) Nitrogen removal in household wastewater treatment using an intermittently aerated membrane bioreactor. Environ Technol 21:107–114 10. Adam C, Gnirss R, Lesjean B, Buisson H, Kraume M (2005) Enhanced biological phosphorus removal in membrane bioreactors. Water Sci Technol 46(4–5):281–286 11. Yoon TI, Lee HS, Kim CG (2004) Comparison of pilot scale performances between membrane bioreactor and hybrid conventional wastewater treatment systems. J Memb Sci 242:5–12 12. Van der Roest HF, Lawrence DP, van Bentem AGN (2002) Membrane bioreactor for municipal wastewater treatment. Water and wastewater practitioner series: STOWA Repor. IWA Publishing, London 13. Ahn K-H, Song K-G, Cho E, Cho J, Yun H, Lee S, Kim J (2003) Enhanced biological phosphorous and nitrogen removal using a sequencing anoxic/anaerobic membrane bioreactor. Desalination 157:345–352 14. WEF (1998) Biological and chemical system for nutrient removal. Water Environment Federation, Alexandria, pp 240–244 15. Wareham DG, Lee NP (1996) Effect of shortened anaerobic HRT on phosphate release and uptake in the UCT Bio-P process. J Environ Technol 17(2):125–136 16. Randall CW, Stensel HD, Barnard JL (1992) Design and retrofit of wastewater treatment plants for biological nutrient removal— water quality management library, vol 5. Technomic Pub, Lancaster PA 17. Mino T, Van Loosdrecht MCM, Heijnen JJ (1998) Microbiology and biochemistry of the enhanced biological phosphate removal process. Water Res 32(11):3193–3207 18. Oehmen A, Lemos PC, Carvalho G, Yuan Z, Keller J, Blackall LL, Reis MAM (2007) Advances in enhanced biological phosphorus removal: From micro to macro scale. Water Res 41:2271– 2300 19. Kazmi AA, Fujita M, Furumai H (2001) Modeling effect of remaining nitrate on phosphorus removal in SBR. Water Sci Technol 43(3):175–182 20. Meinhold J, Pedersen H, Arnold E, Isaacs S, Henze M (1998) Effect of continuous addition of an organic substrate to the anoxic phase on biological phosphorus removal. Water Sci Technol 38(1):97–105 21. Zou H, Du G-C, Ruan W-Q, Chen J (2006) Role of nitrate in biological phosphorus removal in a sequencing batch reactor. World J Microbiol Biotechnol 22(7):701–706 22. Tam NFY, Wong YS, Leung GL (1992) Effect of exogenous carbon sources on removal of inorganic nutrient by the nitrification–denitrification process. Water Res 26(9):1229–1236 23. Sun DD, Hay CT, Khor SL (2006) Effects of hydraulic retention time on behaviour of start-up submerged membrane bioreactor with prolonged sludge retention time. Desalination 195:209–225

123

142 24. Meng F, Shi B, Yang F, Zhang H (2007) Effect of hydraulic retention time on membrane fouling and biomass characteristics in submerged membrane bioreactors. Bioprocess Biosyst Eng 30(5):359–367 25. Chae SR, Kang ST, Watanabe Y, Shin HS (2006) Development of an innovative vertical submerged membrane bioreactor (VSMBR) for simultaneous removal of organic matter and nutrients. Water Res 40:2161–2167 26. Cho J, Song K-G, Yun H, Ahn K-H (2005) Sequencing anoxic/ anaerobic membrane bioreactor (SAM) pilot plant for advanced wastewater treatment. Desalination 178:219–225

123

Bioprocess Biosyst Eng (2009) 32:135–142 27. Burdick CR, Refling DR, Stensel HD (1982) Advanced biological treatment to achieve nutrient removal. J Water Pollut Control Fed 54:1078–1086 28. Ahmed Z, Cho J, Lim B-R, Song K-G, Ahn K-H (2007) Effects of sludge retention time on membrane fouling and microbial community structure in a membrane bioreactor. J Memb Sci 287:211–218 29. Cho J, Song K-G, Ahn K-H (2005) The activated sludge and microbial substances influences on membrane fouling in submerged membrane bioreactor: unstirred batch cell test. Desalination 183:425–429