Đang tải... (xem toàn văn)
ABSTRACT In this study, the formation characteristics of aerobic granular sludge in a continuous-flow reactor were investigated under several experimental conditions. Both surface loading rate (equal to liquid linear velocity at a sludge settling zone) and aeration rate strongly affected the selection of well-settling sludge in the same manner as sludge settling time in a sequencing batch reactor. By setting and controlling adequate surface loading and aeration rates, small particles were effectively washed out, and well-settling sludge selectively remained in the reactor. As a result, aerobic granular sludge was effectively formed. On the other hand, feeding pattern, i.e., continuous and intermittent feeding, did not affect the aerobic granulation when completely inorganic wastewater was fed. These findings will contribute to the dissemination of aerobic granular sludge technology because the information on the formation of aerobic granular sludge in a continuous-flow reactor is limited.
Journal of Water and Environment Technology, Vol. 8, No.3, 2010 Address correspondence to Satoshi Tsuneda, Department of Life Science and Medical Bioscience, Waseda University, Email: stsuneda@waseda.jp Received April 12, 2010, Accepted July 21, 2010. - 251 - Formation of Aerobic Granular Sludge in a Continuous-Flow Reactor – Control Strategy for the Selection of Well-Settling Granular Sludge - Naohiro KISHIDA*, Atsushi KONO**, Yutaka YAMASHITA***, Satoshi TSUNEDA*** *Department of Water Supply Engineering, National Institute of Public Health, 2-3-6 Minami, Wako, Saitama 351-0197, Japan **Department of Chemical Engineering, Waseda University, 3-4-1 Okubo, Shinjuku-ku, Tokyo 169-8555, Japan ***Department of Life Science and Medical Bioscience, Waseda University, 2-2 Wakamatsu-cho, Shinjuku-ku, Tokyo 162-8480, Japan ABSTRACT In this study, the formation characteristics of aerobic granular sludge in a continuous-flow reactor were investigated under several experimental conditions. Both surface loading rate (equal to liquid linear velocity at a sludge settling zone) and aeration rate strongly affected the selection of well-settling sludge in the same manner as sludge settling time in a sequencing batch reactor. By setting and controlling adequate surface loading and aeration rates, small particles were effectively washed out, and well-settling sludge selectively remained in the reactor. As a result, aerobic granular sludge was effectively formed. On the other hand, feeding pattern, i.e., continuous and intermittent feeding, did not affect the aerobic granulation when completely inorganic wastewater was fed. These findings will contribute to the dissemination of aerobic granular sludge technology because the information on the formation of aerobic granular sludge in a continuous-flow reactor is limited. Keywords: aerobic granular sludge, continuous-flow reactor, nitrification. INTRODUCTION In recent years, aerobic granular sludge technology has received much attention because of its specific characteristics (de Kreuk et al., 2007). Aerobic granular sludge has some advantages over conventional bioflocs, such as excellent settleability and high biomass retention (Liu et al., 2004). Aerobic granular sludge technology is applicable to the removal of not only organic carbon but also nitrogen and phosphorus (Kishida et al., 2006). Aerobic granulation of suspended growth aggregates is a phenomenon that has been most frequently observed in systems applying the sequencing batch reactor (SBR) concept (Wilderer and McSwain, 2004), and most of the previous studies succeeded in the formation of aerobic granular sludge using SBRs (Dulekgurgen et al., 2003; Mosquera-Corral et al., 2005; Kishida et al., 2008). However, in our previous studies, aerobic granulation was observed in a continuous-flow aerobic upflow fluidized bed (AUFB) reactor when completely inorganic wastewater containing high concentration of ammonia (hereinafter called “ammonia-rich wastewater”) was fed (Tsuneda et al., 2003). Campos et al. (2000) has also reported aerobic granulation using a similar continuous-flow reactor although they inoculated the sludge coming from the effluent of a nitrifying biofilm airlift reactor using carrier materials. This phenomenon is quite unique. To disseminate aerobic granular sludge technology, establishment of a - 252 - formation method in a continuous-flow reactor is very important because continuous-flow reactors are more popular than SBRs all over the world. However, there is very little information on factors influencing aerobic granulation in a continuous-flow reactor, and therefore, it takes a long time to start-up aerobic granular sludge processes using a continuous-flow reactor at present. Actually, it took approximately 100 days to form granular sludge with a diameter of 200 m in our previous study (Tsuneda et al., 2003), which makes practical application difficult. The specific objective of this study is to reveal some factors greatly influencing aerobic granulation in a continuous-flow reactor. We focused our attention on the controllable factors influencing the selection of well-settling sludge because the sludge selection is one of the most important operational strategies that are essential to successful formation of aerobic granular sludge in an SBR (Li and Li, 2009). Sludge settling time is known as an important factor influencing the sludge selection in an SBR (McSwain et al., 2004a). Generally, surface loading rate (equal to liquid linear velocity at a sludge settling zone) and aeration rate in continuous upflow reactors govern the washout of sludge from the reactor, and therefore, we investigated the effect of surface loading rate and aeration rate on the selection of well-settling sludge and the formation of aerobic granular sludge. To effectively achieve our goal, sludge selection (washout and retention) characteristics under different surface loading and aeration rates were preliminarily investigated using aerobic granular sludge that had already been formed in our previous studies. Based on the preliminary experimental results, we decided the control strategy for the selection of well-settling sludge. Then, we applied the strategy to the formation of aerobic granular sludge in a continuous experiment. In addition, the effect of feeding pattern, i.e. continuous flow and sequencing batch flow modes, on the formation of aerobic granular sludge was investigated. MATERIALS AND METHODS Reactor Configuration In this study, AUFB reactors with a working volume of 13.8 L (including a gas-solid separator), an internal diameter of 5 cm, and a height of 3.2 m, were used for all experiments. The schematic illustration of the AUFB reactor is shown in Fig. 1. Although the AUFB reactor was originally designed to operate as a continuous-flow reactor, the reactors were used not only as continuous-flow reactors but also as SBR to investigate the effect of feeding patterns on the formation of aerobic granular sludge. Preliminary Experiment for Characterizing Sludge Washout and Retention Sludge selection characteristics (washout and retention) in the AUFB reactor were investigated under different surface loading and aeration rate conditions. Firstly, aerobic granular sludge and suspended biomass were mixed. Secondly, the particle distribution of the mixture was measured using different sieves with opening sizes of 125, 250 and 420 m. Mixed liquor samples were sieved with them, and dry weight of each fraction was measured in the same manner as suspended solids (SS). Thirdly, the mixture was poured into the AUFB reactor. Then, feeding and aeration were done, and the effluent was continuously sampled. Particle size distribution in the effluent solution was also measured. Finally, sludge washout characteristics were evaluated using sludge washout index (SWI) defined by the following equation. - 253 - SWI (%/d) = DWEF eff (g/d) / DWEF in (g) × 100 (1) in which DWEF eff is the dry weight of each fraction separated by sieving the effluent solution, DWEF in is the dry weight of each fraction of initial biomass filled to the AUFB reactor. If biomass growth is ignored, multiplicative inverse of SWI multiplied by 100 ((1/SWI) × 100) is equal to sludge retention time of each sludge separated by sieving. Surface loading rates ranged from 0.6 to 3 m 3 /m 2 /d, while aeration rates ranged from 0.3 to 0.8 L/min. The pH was set at 6.8 to 7.0 by adding NaHCO 3 , and the water temperature was maintained at 22 ± 2°C. Reactor Setup and Operation for the Formation of Aerobic Granular Sludge Two AUFB reactors were used for the formation of aerobic granular sludge. Air was introduced from the bottom of the reactor, and aeration rate was gradually increased from 0.3 to 0.8 L/min based on the results from the preliminary experiments for characterizing sludge washout and retention. Dissolved oxygen concentration at the middle part of reactors (approximately 1.5 m from the bottom of reactors) was over 2 mg/L through the entire experiment. The initial concentration of mixed-liquor volatile suspended solids (MLVSS) was set at 1000 mg/L and the volumetric nitrogen loading rate was increased from 0.15 to 0.23 kg N/m 3 /d on day 22. The pH was set at 6.8 to 7.0 by adding NaHCO 3 and the water temperature was maintained at 22 ± 2°C. The difference between Runs 1 and 2 is influent feeding mode. In Run 1, ammonia-rich wastewater was continuously fed from the bottom of the reactor, while the wastewater was intermittently fed (sequencing batch mode) in Run 2. Hydraulic retention time (HRT) in Run 1 was initially set at 2 d (surface loading rate: 1.2 m 3 /m 2 /d) and decreased to 1.33 d (surface loading rate: 1.8 m 3 /m 2 /d) on day 22, while HRT in Run 2 was fixed at 1.5 d. In Run 2, time lengths of influent feeding, aeration, sludge settling and effluent Fig. 1 - Schematic diagram of the AUFB reactor. Gas flow Influent feeding Air Liquid flow Granular biomass Effluent discharge (Continuous flow mode) Effluent discharge (Sequencing batch operational mode) Gas-solid separator Effluent Sludge settling zone Vertical sectional view Overhead view 5cm 20cm 4.7cm 13cm 25cm Effective volume: 7.5L Sludge settling zone - 254 - discharge phases were fixed at 10 min, 21.75 h, 2 h and 5 min (a total of 24 h), respectively. Volumetric exchange ratio was fixed at 0.67. Wastewater Composition and Seed Sludge Ammonia-rich wastewater was synthesized based on the discharge from a semiconductor manufacturing plant. Table 1 shows the chemical composition of this synthetic wastewater, which contains numerous ammonium ions (NH 4 -N: 230-341 mg/L) and other inorganic ions (Ca: 500 mg/L) peculiar to industrial wastewater. The NH 4 -N concentration at the continuous flow operational mode including the preliminary experiment was constant (300 mg/L), while the concentration at the sequencing batch operational mode (Run 2) was changed from 230 to 341 mg/L on day 22 to increase the NH 4 -N loading rate. Iron in the form of FeSO 4 7H 2 O and phosphorus in the form of KH 2 PO 4 were added to the influent as a trace metal and nutrient, respectively. Aerobic granular sludge used for characterization of sludge washout and retention was obtained from another AUFB reactor that had been operated using ammonia-rich wastewater. Details of the granular sludge cultivation methods employed are described in a previous paper (Tsuneda et al., 2003). Seed sludge (suspended biomass) to form aerobic granular sludge for all the experiments was obtained from an aerobic basin of a municipal wastewater treatment plant (Tokyo, Japan). Table 1 - Wastewater composition Analytical Methods MLVSS, SS and sludge volume index (SVI) were analyzed in accordance with standard methods (APHA, 1998). Ammonia-nitrogen (NH 4 -N) was determined using ion chromatography (DX 120, Dionex, Japan). Both NO 2 -N and NO 3 -N were measured by ion chromatography (IC 2001, Tosoh, Japan). RESULTS AND DISCUSSION Effect of Surface Loading and Aeration Rates on the Selection of Well-Settling Sludge As we expected, the increase in both surface loading and aeration rates strongly enhanced washout of small particles (dispersed sludge), while these effects were hardly observed in well-settling large particles (granular sludge), as shown in Fig. 2. Therefore, it must be possible to select only well-settling sludge, which is essential for the formation of aerobic granular sludge when setting adequate surface loading and aeration rates in a sequencing batch mode. Although selection of well-settling granular sludge could be carried out most effectively under the highest surface loading and aeration rates condition, if we set the highest conditions at the startup period, almost all the Concentration (mg/L) Component Continuous flow mode Sequencing batch mode NH 4 Cl N: 300 N: 230 – 341 CaCl 2 Ca: 500 KH 2 PO 4 P: 3 FeSO 4 7H 2 O Fe: 5 - 255 - biomass might be washed out. Hence, both parameters should be gradually increased for the promotion of aerobic granulation. Fig. 2 - Sludge washout index under different surface loading and aeration rates. Formation of Aerobic Granular Sludge by Controlling Surface Loading and Aeration Rates Surface loading and aeration rates were initially set at 1.2 m 3 /m 2 /d and 0.30 L/min, respectively, which is equivalent to 4.3%/d of SWI of small particles (<125 m) in accordance with the preliminary experiment (Fig. 2). If SWI is below 5.0%/d, it is empirically known in our previous studies that nitrifying bacteria can be retained in the reactor. Actually, biomass was not excessively washed out in this study, and therefore, nitrification performance was stable, as shown in Figs. 3 and 4. The value of SVI gradually decreased due to aerobic granulation, as shown in Figs. 5 and 6. As sludge settling ability increased, surface loading and aeration rates were gradually increased, and finally set at 1.8 m 3 /m 2 /d (HRT: 1.33 d) and 0.80 L/min, respectively, which is equivalent to 16.4%/d of SWI of small particles (<125 m). Under this experimental condition, small particles were effectively washed out, and well-settling granular sludge selectively remained in the reactor. As a result, aerobic granulation successfully proceeded, and aerobic granular sludge with an average diameter of 226 m was observed on day 65. This formation period is much shorter than that employed in our previous study without control strategy for the selection of well-settling granular sludge (Tsuneda et al., 2003). Because specific surface area of small particles is much greater than that of well-settling large particles, large particles are normally outcompeted by small particles due to competition for substrate (ammonia) uptake. Therefore, small particles should be preferentially washed out for the growth of large particles by controlling surface loading and aeration rates, which promotes the formation of aerobic granular sludge. 0 5 10 15 20 25 0.3 0.4 0.5 0.6 0.7 0.8 Ａeration rate (L/min) SWI (%/d) 0 5 10 15 20 25 0.3 0.4 0.5 0.6 0.7 0.8 Ａeration rate (L/min) SWI (%/d) (a) Surface loading rate: 0.6 m 3 /m 2 /d (b) Surface loading rate: 1.2 m 3 /m 2 /d (c) Surface loading rate: 1.8 m 3 /m 2 /d (d) Surface loading rate: 3.0 m 3 /m 2 /d 0 5 10 15 20 25 0.3 0.4 0.5 0.6 0.7 0.8 Ａ eration rate (L/min) SWI (%/d) 0 5 10 15 20 25 0.3 0.4 0.5 0.6 0.7 0.8 Ａ eration rate (L/min) SWI (%/d) <125 m 125-250 m 250-420 m >420 m - 256 - Effect of Feeding Pattern on the Formation of Aerobic Granular Sludge As shown in Figs. 3 and 4, nitrification performance and sludge washout characteristics at continuous flow and sequencing batch modes (Runs 1 and 2) were almost the same. Under this experimental condition, sludge settling and aerobic granulation 0 0.10 0.20 0.30 0 1 2 3 NH 4 + -N removal rate [kg-N/m 3 /day] HRT [day] 0.30 0.50 0.80 Aeration rate [L/min] 0 50 100 0 20 40 60 80 100 NH 4 + -N removal efficiency [%] Time [day] 0 0.10 0.20 0.30 0 1 2 3 NH 4 + -N removal rate [kg-N/m 3 /day] HRT [day] 0.800.50 0 50 100 0 20 40 60 80 100 NH 4 + -N removal efficiency [%] Time [day] Aeration rate [L/min] 0.30 (a) Continuous flow mode (Run 1) (b) Sequencing batch mode (Run2) Fig. 3 - Nitrification performance at two operational modes. Red line shows time course of HRT. 0 50 100 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 Time [day] MLVSS [mg/L] To p Bottom Aeration rate [L/min] 0.30 0.50 0.80 Aeration rate [L/min] 0 50 100 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 Time [day] MLVSS [mg/L] Aeration rate [L/min] 0.30 0.50 0.80 To p Bottom Fig. 4 - Time courses of MLVSS at two operational modes. (a) Continuous flow mode (Run 1) (b) Sequencing batch mode (Run2) 0 50 100 0 100 200 300 400 500 Time [day] SVI [mL/g-MLSS] SVI [1 min] SVI [30 min] Aeration rate [L/min] 0.30 0.50 0.80 0 50 100 0 100 200 300 400 500 Time [day] SVI [mL/g-MLSS] SVI [1 min] SVI [30 min] Aeration rate [L/min] 0.30 0.50 0.80 (a) Continuous flow mode (Run 1) (b) Sequencing batch mode (Run2) Fig. 5 - Time courses of SVI at two operational modes. (a) Continuous flow mode (Run 1) (b) Sequence batch mode (Run 2) Inoculum day 27 day 65 Inoculum day 27 day 65 Fig. 6 - Evolution of aerobic granular sludge at two operational modes (bar = 200 µm). - 257 - characteristics in Run 1 were not apparently different from those in Run 2, as shown in Figs. 5 and 6. These results contradicted the results in previous studies (McSwain et al., 2004b). This is because the characteristics of the influent wastewater are different. When organic wastewater is fed, reportedly, intermittent feeding in the SBR produces feast-famine conditions that suppress filamentous organisms which strongly decrease sludge settling ability (Wilderer and McSwain, 2004). However, when completely inorganic wastewater is fed in the same manner as this study, filamentous organisms are normally unable to survive regardless of the feeding pattern. Therefore, the apparent effect of feeding pattern on the aerobic granulation was not observed in this study. CONCLUSIONS This study primarily reveals factors influencing aerobic granulation in a continuous-flow reactor. Both surface loading and aeration rates affect the selection of well-settling sludge and the formation of aerobic granular sludge. By setting and controlling adequate surface loading and aeration rates, small particles were effectively washed out, and well-settling sludge selectively remained in the reactor. As a result, quicker granulation was successfully attained. On the other hand, feeding pattern is not an important factor in aerobic granulation when completely inorganic wastewater is fed. REFERENCES Campos J. L., Mendez R. and Lema J. M. (2000). Operation of a nitrifying activated sludge airlift (NASA) reactor without biomass carrier, Water Sci. Technol., 41(4-5), 113-120. De Kreuk M. K., Kishida N. and van Loosdrecht M. C. M. (2007). Aerobic granular sludge - State of the art, Water Sci. Technol., 55(8-9), 75-81. Dulekgurgen E., Ovez S., Artan N. and Orhon D. (2003). Enhanced biological phosphate removal by granular sludge in a sequencing batch reactor, Biotechnol. Lett., 25(9), 687-693. Kishida N., Kim J. H., Tsuneda S. and Sudo R. (2006). Anaerobic/oxic/anoxic granular sludge process as an effective nutrient removal process utilizing denitrifying polyphosphate-accumulating organisms, Water Re s., 40(12), 2303–2310. Kishida N., Tsuneda S., Sakakibara Y., Kim J. H. and Sudo R. (2008). Real-time control strategy for simultaneous nitrogen and phosphorus removal using aerobic granular sludge, Water Sci. Technol., 58(2), 445–450. Li A. J. and Li X. Y. (2009). Selective sludge discharge as the determining factor in SBR aerobic granulation: Numerical modelling and experimental verification, Water Res., 43(14), 3387-3396. Liu Y., Yang S. F. and Tay J. H. (2004). Improved stability of aerobic granules by selecting slow-growing nitrifying bacteria, J. Biotechnol., 108(2), 161-169. McSwain B. S., Irvine R. L. and Wilderer P. A. (2004a). The influence of-settling time on the formation of aerobic granules, Water Sci. Technol., 50(10), 195-202. McSwain B. S., Irvine R. L. and Wilderer P. A. (2004b). The effect of intermittent feeding on aerobic granule structure, Water Sci. Technol., 49(11–12), 19–25. Mosquera-Corral A., de Kreuk M. K., Heijnen J. J. and van Loosdrecht M. C. M. (2005). Effects of oxygen concentration on N-removal in an aerobic granular sludge reactor, Water Res., 39(12), 2676-2686. Standard Methods for the Examination of Water and Wastewater (1998). 20th edn, - 258 - American Public Health Association/American Water Works Association/Water Environment Federation, Washington DC, USA. Tsuneda S., Nagano T., Hoshino T., Ejiri Y., Noda N. and Hirata A. (2003). Characterization of nitrifying granules produced in an aerobic upflow fluidized bed reactor, Water Res., 37(20), 4965-4973. Wilderer P. A. and McSwain B. S. (2004). The SBR and its biofilm application potentials, Water Sci. Technol., 50(10), 1-10. . Shinjuku-ku, Tokyo 169 -85 55, Japan ***Department of Life Science and Medical Bioscience, Waseda University, 2-2 Wakamatsu-cho, Shinjuku-ku, Tokyo 162 -84 80, Japan ABSTRACT. reactor, Water Res., 39(12), 2676-2 686 . Standard Methods for the Examination of Water and Wastewater (19 98) . 20th edn, - 2 58 - American Public Health Association/American