The effects of ozonation conditions on the performance of a continuous anaerobic/oxic/anoxic (A/O/A) process with sludge ozonation and phosphorus adsorption were investigated. In this system, excess sludge was ozonated by microbubble ozonation, and then the supernatant of the ozonated sludge was flowed into a phosphorus adsorption column packed with zirconium-ferrite adsorbent. The effluent from the column and the settlings of the ozonated sludge were recirculated in the A/O/A process. Long-term operation of a lab-scale system treating rural wastewater showed that ozonation affected not only the sludge reduction efficiency but also the nitrogen removal efficiency. When the amount of sludge to be ozonated was set at 16% of total MLSS per day, no excess sludge was withdrawn, but the nitrogen removal efficiency was deteriorated. Decreasing the amount of sludge to be ozonated (to 9.4% of total MLSS per day) resulted in efficient nitrogen removal, but the MLSS concentration increased slightly. Phosphorus accumulated in the sludge was re-solubilized by ozonation, and a large part of the solubilized phosphorus consisted of Pi. Almost all Pi was recovered in the phosphorus adsorption column.
Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 135 - Improvement of nutrient removal and phosphorus recovery in the anaerobic/oxic/anoxic process combined with sludge ozonation and phosphorus adsorption Takashi KONDO*, Satoshi TSUNEDA**, Yoshitaka EBIE*, Yuhei INAMORI***, and Kaiqin XU* * Research Center for Material Cycles and Waste Management, National Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-8506, Japan ** Department of Life Science and Medical Bioscience, Waseda University, 2-2 Wakamatsu-cho, Shinjuku-ku, Tokyo 162-8480, Japan *** Faculty of Symbiotic Systems Science, Fukushima University, 1 Kanayagawa, Fukushima, Fukushima 960-1296, Japan ABSTRACT The effects of ozonation conditions on the performance of a continuous anaerobic/oxic/anoxic (A/O/A) process with sludge ozonation and phosphorus adsorption were investigated. In this system, excess sludge was ozonated by microbubble ozonation, and then the supernatant of the ozonated sludge was flowed into a phosphorus adsorption column packed with zirconium-ferrite adsorbent. The effluent from the column and the settlings of the ozonated sludge were recirculated in the A/O/A process. Long-term operation of a lab-scale system treating rural wastewater showed that ozonation affected not only the sludge reduction efficiency but also the nitrogen removal efficiency. When the amount of sludge to be ozonated was set at 16% of total MLSS per day, no excess sludge was withdrawn, but the nitrogen removal efficiency was deteriorated. Decreasing the amount of sludge to be ozonated (to 9.4% of total MLSS per day) resulted in efficient nitrogen removal, but the MLSS concentration increased slightly. Phosphorus accumulated in the sludge was re-solubilized by ozonation, and a large part of the solubilized phosphorus consisted of Pi. Almost all Pi was recovered in the phosphorus adsorption column. Keywords: Ozonation; phosphorus recovery; sludge reduction INTRODUCTION In recent years, biological nutrient removal processes, such as the anaerobic/anoxic/oxic (A/A/O) process, have been widely introduced in wastewater treatment plants (WWTPs). In these systems, nitrogen is removed by nitrifying and denitrifying bacteria, and phosphorus is removed by polyphosphate-accumulating organisms (PAOs). Although the A/A/O process achieves effective nutrient removal, competition for energy sources between PAOs and denitrifying bacteria critically affects the nutrient removal efficiency. To avoid this competition, some new systems employing denitrifying PAOs (DNPAOs), which are capable of utilizing nitrate/nitrite as electron acceptors unlike PAOs, have been proposed (Ahn et al., 2002a, b; Kuba et al., 1996, 1997; Soejima et al., 2006; Tsuneda et al., 2006). Previously, Tsuneda et al. (2006) described an anaerobic/oxic/anoxic (A/O/A) process and succeeded in causing DNPAOs to take an active part in simultaneous nitrogen and phosphorus removal in an acetate-fed sequencing batch reactor (SBR) in which additional acetate was required for inhibition of phosphorus uptake in the oxic period. Treatment and disposal of excess sludge have been serious problems in WWTPs; the Address correspondence to Satoshi Tsuneda, Department of Life Science and Medical Bioscience, Waseda University, Email: stsuneda@waseda.jp Received January 30th, 2009, Accepted May 20th, 2009 Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 136 - treatment of excess sludge may account for as much as 25% to 65% of total plant operating costs (Liu, 2003). One method to reduce excess sludge is solubilization of the excess sludge by using ozone, which is a strong oxidant, and recirculation of the solubilized excess sludge, containing readily biodegradable carbon, in a biological treatment system (Chu et al., 2008; Cui and Jahng, 2004; Kamiya and Hirotsuji, 1998; Nagare et al., 2008; Sakai et al., 1997; Saktaywin et al., 2005; Suzuki et al., 2006; Yasui et al., 1996; Yasui and Shibata, 1994). Introduction of an ozonation system for excess sludge reduction is advantageous in terms of total energy consumption in WWTPs (Nagare et al., 2008). Additionally, Chu et al. (2008) reported that the efficiencies of ozone utilization and sludge solubilization were improved by using a microbubble ozonation system. In this system, microbubbles with a diameter less than several tens of micrometers are generated by a mixing pump that mixed gas and liquid at high speed (Chu et al. 2008). In another system, phosphorus accumulated in excess sludge is solubilized by ozonation, and then the solubilized phosphorus can be easily recovered by crystallization or with a phosphorus adsorbent (Saktaywin et al., 2005; Suzuki et al., 2006). Against this background, we previously proposed an advanced system involving a continuous A/O/A process combined with sludge reduction by ozonation and phosphorus recovery by phosphorus adsorbent (Suzuki et al., 2006) (Fig. 1A). In that system, nitrogen and phosphorus were removed effectively without any sludge production; however, the TOC removal efficiency was deteriorated due to the circulation of ozonated sludge. Furthermore, the ozonated sludge was not suitable as an additional carbon source to inhibit oxic phosphorus uptake. In the present study, to achieve efficient nutrient removal, the previous system was improved by introducing a microbubble ozonation system and by changing the ozonated sludge recirculation lines (Fig. 1B). During 152 days operation of a lab-scale reactor, the operational conditions were optimized by changing the amount of sludge to be ozonated and the recirculation ratio of the residual liquid from the physico-chemical processes to the anaerobic and oxic tanks. Figure 1 Schematic diagrams of the A/O/A process with ozonation and phosphorus adsorption: A, previous study (Suzuki et al., 2006); B, this study. Influent Anaerobic tank Oxic tank Anoxic tank Effluent Sludge return Ozonation process Phosphorus adsorption column Influent Anaerobic tank Oxic tank Anoxic tank Effluent Sludge return Phosphorus adsorption column Microbubble ozonation process Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 137 - MATERIALS AND METHODS A lab-scale continuous A/O/A process with a working volume of 36 L (anaerobic tank, 10.3 L; oxic tank, 10.3 L; anoxic tank, 15.4 L), which was the same as that used in our previous study (Suzuki et al., 2006), was operated for 152 days. The microbubble ozonation system and the phosphorus adsorption column were introduced to the A/O/A process (Fig. 1B). The microbubble ozonation system was a cylindrical reactor with an internal diameter of 0.2 m and a height of 0.8 m (effective volume of 20 L). The ozonation system received sludge withdrawn continuously from the end of the anoxic tank. When the amount of sludge reached approximately 10 L (once in 1 to 3 days), ozonation was conducted under specified ozonation conditions, as shown in Table 1. Ozone gas was generated by an ozone generator (PO-10; Fuji Electric, Japan), and the applied ozone concentration was monitored with a UV ozone monitor (PG-620HA; Ebara Jitsugyo, Japan). Ozone gas and sludge from the anoxic tank were mixed in a turbulent flow by a turbine pump (Nikuni swirling current pump M15NPD02S; Nikuni Co., Japan), and then the mixture of microbubble ozone and sludge was circulated back into the reactor. After ozonation, the supernatant with unsettled microsolids was flowed into the phosphorus adsorption column (Fig. 1B). The phosphorus adsorption column (internal diameter, 75 mm; height, 0.7 m; effective volume, 2 L) was filled with 1.5 L of spherical zirconium-ferrite (ZrFe 2 (OH) 8 ) adsorbent with an effective diameter of 0.7 mm (Japan Enviro Chemicals, Japan). The flow rate of the supernatant was set at 25 mL/min. After phosphorus adsorption, the column was backwashed with 15 L of tap water to remove residual suspended solids (SS). Then, the residual liquid from the physico-chemical processes, which was the mixture of the effluent from the phosphorus adsorption column and the backwash water, was circulated back to the anaerobic tank and the oxic tank at an appropriate rate (Table 1). Table 1 Operational conditions. The hydraulic retention time (HRT) for influent wastewater was adjusted to 10 h. The sludge return ratio from the settling tank to the anaerobic tank was controlled at 80%. The reactor was inoculated with activated sludge (initial MLSS was adjusted to 4,000 mg/L), which was collected from a WWTP (A/A/O process) with efficient biological phosphorus removal. Raw rural wastewater, which was collected from a rural sewage treatment plant daily, was flowed into the system. The characteristics of the influent wastewater were as follows: 160–200 mg/L of SS, 55–80 mg/L of TOC, 45–55 mg/L of T-N, and 4.0–5.5 mg/L of T-P. In this study, four different operational conditions were used to determine the most appropriate ones (Table 1). The nutrient recovery efficiency in each phase, except for Phase 1, was evaluated more than 25 days after changes to prevent effects of the previous operational conditions. Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 138 - MLSS was measured according to the Standard Methods (1995). To determine soluble TOC (S-TOC), NH 4 -N, NO 2+3 -N, NO 2 -N, and PO 4 -P, water samples were filtered using a glass-fiber filter (GF/C, Whatman Japan KK, Japan). TOC and S-TOC were measured with a SHIMADZU TOC-VSCH (Shimadzu, Japan). Total nitrogen (T-N), total phosphorus (T-P), soluble T-N (ST-N), soluble T-P (ST-P), NH 4 -N, NO 2+3 -N, NO 2 -N, and PO 4 -P were measured with a TRAACS 2000 (Bran+Luebbe, Japan). RESULTS AND DISCUSSION The A/O/A process combined with the microbubble ozonation system and the phosphorus adsorption column was operated for 152 days. Water quality in the effluent and MLSS concentration in the reactor were affected by the ozonation conditions (Figure 2). In Phase 1, 23% of total MLSS per day was withdrawn, and the sludge was ozonated every day. This operational condition resulted in a dramatic decrease in MLSS concentration. On day 23, the amount of sludge to be ozonated was reduced from 23% to 16% of total MLSS per day (Phase 2), and as a result, MLSS concentration was maintained at around 3,000 mg/L. Figure 2 Time course of TOC, T-N, and T-P concentrations in the effluent and MLSS in the reactor. Figure 3 shows the water quality profiles at day 49 (Phase 2), day 78 (Phase 3), and day 147 (Phase 4). In Phase 2, the effluent T-N concentration increased to approximately 20 mg/L, whereas no excess sludge was withdrawn in Phase 2 (Fig. 3). Effluent T-N was mostly composed of NH 4 -N, indicating deterioration of nitrification performance. Considering that organic carbon generally induces dissolved oxygen competition between nitrifying bacteria and heterotrophic bacteria, organic carbon loading by circulation of ozonated sludge might have been the cause of the deteriorated nitrification performance. Then, to reduce the organic carbon loading to the oxic tank, the circulation ratio of the residual liquid from the physico-chemical processes was changed (Phase 3). However, nitrification performance was not improved. In Phase 4, to maintain the population density of slow-growing nitrifying bacteria, the amount of sludge to be ozonated was reduced to 9.4 % of total MLSS per day (Table 1). As a result, the effluent T-N concentration decreased to around 10 mg/L. This efficient nitrogen removal was achieved over 2 months with a slight increase of MLSS concentration (Fig. 2). The Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 139 - sludge yield was roughly estimated from the slope in Fig. 2 to be 17 mg-MLSS/L/day. This was 50 % of the sludge yield in the A/O/A process without the physico-chemical processes, operated under the same experimental conditions (data not shown). Figure 3 Water quality profiles at day 49 (Phase 2), day 78 (Phase 3), and day 147 (Phase 4). In our previous study, effluent TOC was deteriorated when an ozonation system was introduced to the A/O/A process (Suzuki et al., 2006). This deterioration might be due to slowly biodegradable materials derived from ozonated sludge (Yasui and Shibata, 1994). In this study, TOC was effectively removed and deterioration was not observed (Fig. 2). Chu et al. (2007, 2008) reported that microbubble ozonation improved the mass transfer of ozone, and the high inner pressure in the bubbles could accelerate the formation of hydroxyl radicals. Therefore, some of the biorefractory and/or slowly biodegradable materials might be oxidized to easily biodegradable materials by changing the ozonation system. Phosphorus was removed effectively in all phases (Fig. 3). The phosphorus concentration increased in the anaerobic phase due to phosphorus release from PAOs. Part of the released phosphorus was then accumulated in the sludge in the subsequent oxic tank by normal oxygen-utilizing PAOs. The residual phosphorus was removed in the final anoxic tank without oxygen. Especially in Phase 4, phosphorus was removed without a decrease in TOC concentration in the anoxic tank, indicating that phosphorus was accumulated by DNPAOs, like previous studies (Ahn et al., 2002a, b; Kuba et al., 1996, 1997; Soejima et al., 2006; Tsuneda et al., 2006). Therefore, it was suggested that DNPAOs contributed to not only phosphorus removal but also denitrification. In the previous system, most of the phosphorus was removed in the oxic tank by the PAOs even though ozonated sludge was added to the oxic tank to inhibit oxic phosphorus uptake (Suzuki et al., 2006). Therefore, the residual liquid from the physico-chemical processes might contain an organic carbon source that was available for the PAOs. Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 140 - Phosphorus accumulated in the sludge was re-solubilized by ozonation. The phosphorus concentrations in the influent (after ozonation) and effluent of the phosphorus adsorption column are shown in Fig. 4. During the operation, about 70% of the phosphorus in the sludge was solubilized by ozonation, and a large part of the solubilized phosphorus consisted of PO 4 -P. Over 90% of the solubilized phosphorus was absorbed, and the effluent PO 4 -P concentration was maintained at less than 1 mg/L until day 119 (Fig. 4). After 119 days operation, the effluent PO 4 -P concentration reached 1 mg/L (Fig. 4). Then, the phosphorus adsorbent was subjected to phosphorus desorption by using an alkali solution and subsequent reactivation by using an acid solution, according to a method described in the literature (Ebie et al., 2008). The reactivated adsorbent was packed into the column again, and phosphorus was effectively adsorbed. Figure 4 Concentration of each type of phosphorus in the influent and effluent of the phosphorus adsorption column. The small graph at the right shows the breakthrough point (PO 4 -P concentration: 1 mg/L) of the adsorbent at day 119. As the amount of excess sludge increases, regulations for its disposal have become increasingly stringent (Liu, 2003; Ødegaard, 2004). Many attempts have been made to reduce excess sludge production. In this study, a microbubble ozonation system was combined with the A/O/A process. In Phase 1, the MLSS concentration in the reactor was dramatically decreased due to excess sludge reduction. MLSS concentration was maintained at around 3,000 mg/L in Phases 2 and 3; however, the nitrification efficiency was deteriorated. Subsequently, reducing the amount of sludge to be ozonated (9.4% of total MLSS in the reactor) improved the nitrification efficiency in Phase 4. Ozonated sludge also acts as nutrient loading in the A/O/A process. Cui and Jahng (2004) reported that nitrogen derived from ozonated sludge was not completely reduced to nitrogen gas by energy sources originating from the ozonated sludge. In the present study, although nitrogen components in the ozonated sludge were not analyzed, deterioration of the nutrient removal efficiency was not observed in Phase 4 (Figs. 2 and 3). Reduction of growth yield potential of microorganisms in the sludge is also important for excess sludge reduction. Kuba et al. (1996) reported that sludge production efficiency could be decreased by using DNPAOs. In the present study, phosphorus was Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 141 - removed both in the oxic and anoxic tanks, suggesting that both PAOs and DNPAOs contributed to phosphorus removal. Therefore, DNPAOs in the reactor might contribute to both nutrient removal and sludge reduction. It is important for practical applications to estimate the benefits in terms of reduced energy consumption (operating costs); however, they were not estimated in this study. Chu et al. (2008) reported that effective ozone utilization and sludge solubilization could be achieved. However, operation of an additional turbine pump that is required would increase the energy consumption. Additionally, energy consumption is likely to be increased compared with conventional WWTPs because of the introduction of the phosphorus adsorption column. Therefore, energy balance and life cycle assessment analyses will be necessary in a future study. CONCLUSIONS In order to meet the increasingly stringent requirements for environmental protection and phosphorus resource recovery, an A/O/A process involving microbubble ozonation and phosphorus adsorption was tested under various ozonation conditions. The main conclusions are as follows. (1) The amount of sludge to be ozonated affected the nitrogen removal efficiency, whereas TOC and phosphorus removal were not affected by ozonation. Under the optimum ozonation conditions (9.4% of total MLSS per day), efficient nutrient removal was achieved with a slight increase in MLSS concentration. (2) Phosphorus in the sludge was accumulated by not only normal oxygen-utilizing PAOs in the oxic tank but also by DNPAOs in the anoxic tank. It was suggested that DNPAOs contributed to both phosphorus removal and denitrification. (3) Most of the phosphorus in the sludge was solubilized by ozonation, and a large part of the solubilized phosphorus consisted of PO 4 -P. Over 90% of the PO 4 -P was absorbed in the phosphorus adsorption column. REFERENCES Ahn, J., Daidou, T., Tsuneda, S. and Hirata, A. (2002a). Characterization of denitrifying phosphate-accumulating organisms cultivated under different electron acceptor conditions using polymerase chain reaction-denaturing gradient gel electrophoresis assay, Water Res., 36(2), 403-412. Ahn, J., Daidou, T., Tsuneda, S. and Hirata, A. (2002b). Transformation of phosphorus and relevant intracellular compounds by a phosphorus-accumulating enrichment culture in the presence of both the electron acceptor and electron donor, Biotechnol. Bioeng., 79(1), 83-93. Chu, L.B., Xing, X.H., Yu, A.F., Zhou, Y.N., Xun, X.L. and Jurcik, B. (2007). Enhanced ozonation of dyestuff wastewater by microbubble technology, Chemosphere, 68(10), 1854-1860. Chu, L.B., Yan, S.T., Xing, X.H., Yu, A.F., Sun, X.L. and Jurcik, B. (2008). Enhanced sludge solubilization by microbubble ozonation, Chemosphere, 72(2), 205-212. Cui, R. and Jahng, D. (2004). Nitrogen control in AO process with recirculation of solubilized excess sludge, Water Res., 38(5), 1159-1172. Ebie, Y., Kondo, T., Kadoya, N., Mouri, M., Maruyama, O., Noritake, S., Inamori, Y. Journal of Water and Environment Technology, Vol. 7, No. 2, 2009 - 142 - and Xu, K. (2008). Recovery oriented phosphorus adsorption process in decentralized advanced Johkasou, Water Sci. Technol., 57(12), 1977-1981. Kamiya, T. and Hirotsuji, J. (1998). New combined system of biological process and intermittent ozonation for advanced wastewater treatment, Water Sci. Technol., 38(8- 9), 145-153. Kuba, T., van Loosdrecht, M.C.M. and Heijnen, J.J. (1996). Phosphorus and nitrogen removal with minimal COD requirement by integration of denitrifying dephosphatation and nitrification in a two-sludge system, Water Res., 30(7), 1702- 1710. Kuba, T., van Loosdrecht, M.C.M., Brandse, F.A. and Heijnen, J.J. (1997). Occurrence of denitrifying phosphorus removing bacteria in modified UCT-type wastewater treatment plants, Water Res., 31(4), 777-786. Liu, Y. (2003). Chemically reduced excess sludge production in the activated sludge process, Chemosphere, 50(1), 1-7. Nagare, H, Tsuno, H., Saktaywin, W. and Soyama, T. (2008). Sludge ozonation and its application to a new advanced wastewater treatment process with sludge disintegration, Ozone: Science & Engineering, 30, 136-144. Ødegaard, H. (2004). Sludge minimization technologies - an overview, Water Sci. Technol., 49(10), 31-40. Sakai, Y., Fukase, T., Yasui, H. and Shibata, M. (1997). An activated sludge process without excess sludge production, Water Sci. Technol., 36(11), 163-170. Saktaywin, W., Tsuno, H., Nagare, H., Soyama, T. and Weerapakkaroon, J. (2005). Advanced sewage treatment process with excess sludge reduction and phosphorus recovery, Water Res., 39(5), 902-910. Soejima. K., Oki, K., Terada, A., Tsuneda, S. and Hirata, A. (2006). Effects of acetate and nitrite addition on fraction of denitrifying phosphate-accumulating organisms and nutrient removal efficiency in anaerobic/aerobic/anoxic process, Biopro. Biosys. Eng., 29(5-6), 305-313. Standard Methods for the Examination of Water and Wastewater (1995). 19th edn, American Public Health Association/American Water Works Association/Water Environment Federation, Washington DC, USA. Suzuki, Y., Kondo, T., Nakagawa, K., Tsuneda, S., Hirata, A., Shimizu, Y. and Inamori, Y. (2006). Evaluation of sludge reduction and phosphorus recovery efficiencies in a new advanced wastewater treatment system using denitrifying polyphosphate accumulating organisms, Water Sci. Technol., 53(6), 107-113. Tsuneda, S., Ohno, T., Soejima, K. and Hirata, A. (2006). Simultaneous nitrogen and phosphorus removal using denitrifying phosphate-accumulating organisms in a sequencing batch reactor, Biochem. Eng. J., 27(3), 191-196. Yasui, H. and Shibata, M. (1994). An innovative approach to reduce excess sludge production in the activated sludge process, Water Sci. Technol., 30(9), 11-20. Yasui, H., Nakamura, K., Sakuma, S., Iwasaki, M. and Sakai, Y. (1996). A full-scale operation of a novel activated sludge process without excess sludge production, Water Sci. Technol., 34(3-4), 395-404. . Res., 31(4), 77 7 -78 6. Liu, Y. (2003). Chemically reduced excess sludge production in the activated sludge process, Chemosphere, 50(1), 1 -7. Nagare, H,. two-sludge system, Water Res., 30 (7) , 170 2- 171 0. Kuba, T., van Loosdrecht, M.C.M., Brandse, F.A. and Heijnen, J.J. (19 97) . Occurrence of denitrifying phosphorus