A number of BNR processes have been developed, in addition to various biochemical models proposed in the past decade (Comeau et al., 1986; Mino et al., 1995; Smolders et al., 1995; Wentzel et al., 1986; Jeon and Park, 2000). Nevertheless, the biological mechanism and effecting factors for nutrient removal at the anoxic stage are not completely understood.
In this study, the interrelationship and the characteristics of microorganisms as the effecting factors of phosphorus release, such as organic matter and nitrate under anoxic state were investigated. In addition, effects of nitrate and organic matter on phosphorus uptake in the anoxic state were investigated; especially a possible role for denitrifying phosphorus removing bacteria (DPB). Next, the fraction of DPB in the anoxic stage was measured and the species of microorganisms in the anoxic stage were identified using the PCR-DGGE method. Finally, the Daewoo nutrient removal (DNR) process was modelled using ASM No. 2d.
From the results of respirometric measurements, it could be concluded that fraction of readily biodegradable substrate, slowly biodegradable substrate, inert soluble substrate and inert particular substrate were about 10-16%, 32-50%, 1-8% and 2-47%, respectively. The biosorption efficiency of the influent sewage with previous composition was 70-85% within 30-min anaerobic contact. The first COD sequestration rate by the physico-chemical reaction before 5 minutes was 4.8 to 9.6 $sec^{-1}$, while the second reaction rate by the biological reaction after 5 minutes was in the range of 1.68 to 2.16 $sec^{-1}$. However, after 30min, SCOD in the batch reactor increased and it was supposed that desorption of a part of the colloidal organic matter pre-sorbed on the activated sludge flocs was evidenced. As the dissolved organic carbon (DOC) of classified molecular weight with contact time, DOC smaller than 1kDa, which is 80% of total DOC, decreased from 76.5 to 5.1 mg/l before 30 minute but, after 30 min, DOC increased to 6.8 mg/l. Thus, after 30 min, it was considered that particle and large molecules might be degraded or desorbed by microorganism. In denitrification test using the sequestered COD, denitrification efficiency was 60% of influent nitrate. This shows that microorganisms involved in denitrification can use the preserved COD on the activated sludge flocs during the anaerobic contact times.
From the DPB fraction test we could get the followings:
Phosphorus release and denitrification occurred simultaneously under anoxic condition and anoxic phosphorus uptake by DPBs occurred at COD/N ratios lower than 4.2. Also, even the nitrite increased until 20 mg/l, it was not influence to denitrification because the concentration of microorganism was higher than 2000 mg/l as Beccari et al. (1983) suggested. The nitrate removal efficiencies by denitrifiers and dephosphatation by DPBs in the batch tests were 76 % and 24 % of total nitrate removal efficiency, respectively.
In the research of lab-scale DNR process, as the COD/N ratio increased from 5 to 10, T-N and T-P removal rate increased. As the HRT decreased from 8hr to 4hr, nitrogen removal rates decreased from 74% to 20% due to the low nitrification rate as the reduced aeration time. Also, when the recycle ratio increased from 1 to 3, the amounts of anoxic P-uptake and nitrate reduction rate increased due probably to the improved activity of DPBs in the anoxic state.
It was verified that the COD/N ratio was influence on the species of microorganism. As the COD/N ratio increased to 7.5, the numbers of nitrifiers decreased. On the other hand the number of denitrifiers increased at the COD/N ratio of 7.5 because the nitrifiers should compete with heterotrophs for dissolved oxygen occupied. When the COD/N ratios were over 7.5, the number of nitrifiers and denitrifiers was similar. Therefore, this result suggested that if the COD/N ratio was over 7.5, the species of the microorganisms capable of nitrification and denitrification might be similar.
Most of PAO species from the DNR sludge using synthetic wastewater were affiliated with the beta subclass of the Proteobacteria. The species belong to two genera: Dechloromonas and Rhodocyclus. However, species representing Acinetobacter sp, Aeromonas sp. and Pseudomonas sp., which historically have been considered dominant in phosphorus-removing sludge communities, were not present. In this system, the major DPBs found were Rhodocyclus and Dechloromonas.
In the Lab scale DNR process using domestic sewage, nutrients and organic matter were removed effectively at low C/N ratios. T-N removal efficiency of this system was stable in spite of the fluctuation of influent COD. However, the phosphorus removal rate was relatively lower than nitrogen removal rate because of the deficiency of readily biodegradable COD (RBDCOD). Thus, the methods for degradation of nonbiodegradable COD should be studied. In the anoxic state, the nitrate removal rate was higher than the theoretical value. As a result, the dephosphatation was preceded by DPB which took up phosphate using nitrate as an electron acceptor. Also, in the full scale DNR process using domestic sewage, nitrate removal rate increased about 12% than theoretical value due to the anoxic P-uptake mechanism.
As the organic matter fed, the glycogen content of the bio-P bacteria declined and concentration of PHAs increased at the anaerobic state. After that, PHAs were reduced inside PAO whereas the glycogen content was found to increase slightly at the anoxic and aerobic stage. Also, the degraded poly-P at the anaerobic stage was resynthesized within the cells of the PAO for phosphorus uptake at anoxic and aerobic stage. The bacterial species composition of the full scale DNR system clones are within the beta subclass, it was the group called DPB existed in the anoxic state, meaning that nitrate removal by denitrifying dephosphatation was conducted.
A DNR model using the suggested parameters and stoichiometry from ASM2d was developed. The value of each pollutant suggested by the model simulation was similar to the measured value and mean that the simulation of this model was good for reproduction.
窒素와 燐을 除去하기 위해 많은 BNR 工程이 開發되었으며 이에 대한 生化學的 모델들이 提示되어 왔으나, 無酸素槽에서의 窒素와 燐 除去에 대한 影響因子나 메커니즘에 대해서는 明確하게 밝혀지지 않았다. 本 硏究에서는 燐 放出에 影響을 주는 微生物의 特性과 有機物, 窒酸性 窒素와의 相關關係를 밝히고, 無酸素 條件에서 有機物이 燐 吸收에 미치는 影響과 denitrifying phosphorus removing bacteria (DPB) 들의 活性슬러지 工程 내 分率 및 PCR-DGGE를 이용한 種 分析이 修行되었으며, 最終的으로 ASM2d를 利用하여 DNR 工程에 대한 모델을 開發하였다. 呼吸率 測定法을 利用하여 流入下水內의 有機物質을 分析한 結果, 쉽게 分解되는 物質이 全體 有機物의 10-16 %, 천천히 分解되는 物質이 32-50 %로 나타났다. 30分 동안의 嫌氣性 接觸에 의한 微生物 吸着率은 약 75-80 %였으며, 60分 경에는 SCOD 濃度가 약간 增加하였다. 이는 천천히 分解되는 有機物質이 微生物들의 extracellular enzyme에 의해 分解되기 때문에 나타나는 現狀으로 判斷되며 分子量 分布 結果를 통해서도 確認할 수 있었다. 嫌氣性 狀態에서의 有機物 吸着은 反應 始作 後 初期 5分 以內에는 物理化學的 反應 그리고 5分 以後에는 生物學的 反應에 의해 일어났다. TCOD/N 比가 7.9인 경우, 嫌氣性 接觸에 의해 吸着된 有機物을 利用한 脫窒率 實驗結果 注入된 窒酸性窒素의 60%가 除去되었으며, 이는 微生物들이 吸着된 有機物을 利用하여 脫窒을 效果的으로 進行한다는 것을 나타낸 것이다. 無酸素狀態에서 有機物 및 窒酸性窒素의 變化에 따른 影響 實驗에서 DPB들에 의해 燐 放出과 脫窒이 同時에 일어남을 確認하였으며, COD/N 比가 낮은 경우에 無酸素性 燐 吸收 效率이 높은 것으로 나타났다. 脫窒 微生物 및 DPB에 의한 脫窒率은 각각 全體 脫窒率의 76%와 24%로 나타났다. COD/N 比 및 滯留時間이 增加하는 동안 窒素 및 燐 除去率이 增加하였으며, DNR 工程을 效果的으로 運轉하기 위해서는 好氣槽 滯留時間을 4時間 동안 維持해야 할 것으로 判斷된다. 또한 內部返送率을 100-300 %까지 變化시킬 경우 全體 窒素 및 燐 除去率은 變化가 없었으나, 無酸素性 燐 攝取效率은 增加하였다. 內部返送率이 200% 인 경우 전체 脫窒量의 12-13%가 DPB에 의해 發生하는 것으로 나타났다. COD/N 比가 增加하는 동안 窒酸化 微生物은 減少하였으나 脫窒 및 燐 除去 微生物은 增加하였다. 대부분의 phosphorus accumulating organism (PAO)은 Proteobacteria의 beta subclass에 속하였으며 Dechloromonas와 Rhodocyclus가 주로 나타났다. 반면 一般的으로 알려진 Acinetobacter sp., Aeromonas sp., Pseudomonas sp. 등의 燐 除去 微生物들은 나타나지 않았다. 本 硏究에서 나타난 代表的인 DPB는 Rhodocyclus sp. Dechlorosoma sp. 였으며 이들에 의해 無酸素性 燐 吸收 및 脫窒이 發生하였다. Full-scale DNR 工程에 대해 評價한 結果 주로 發見된 DPBs는 Rhodocyclus sp. (92% identity)였다. 또한 燐除去 微生物의 活性을 評價하기 위해 細胞內 物質을 分析한 結果 PHA의 主性分은 3HV였으며 嫌氣槽에서 poly-P 分解 및 PHA 合成이 이루어졌고, 生成된 PHA의 51 % 가 無酸素槽에서 分解되어 poly-P가 合成되는 것을 確認할 수 있었다. 즉, DPB들에 의해 無酸素性 燐 攝取에 따른 脫窒이 일어남을 確認할 수 있었다. DNR 工程에 대해 ASM2d를 利用하여 모델링을 修行한 結果 전체적인 傾向은 類似하게 나타났으나, ASM2d 自體의 不確實性으로 인해 窒素를 除外한 다른 項目들은 각 單位工程別로 正確性이 다소 떨어졌지만 流出水 水質은 類似하게 나타났다. 따라서 追加的으로 파라미터들에 대해 遺傳子 알고리즘 및 實驗을 통해 補完하여야 할 것으로 判斷된다.
