| 摘要: | 本研究建構一實驗室薄膜生物反應槽(Membrane bioreactor，MBR)，利用好/缺氧操作模式針對ABS樹脂廢水處理進行研究，研究期間共174天不排泥。反應槽槽體分為兩部分，一為生物處理槽2.3L，另一為薄膜槽3.5L。本研究探討各項水質參數之去除成效，如BOD、COD、TOC、有機氮、氨氮、亞硝酸氮與硝酸氮等，並評估MLSS、MLVSS、食微比以及體積負荷，最後經由流通量與多弁鈺蓬y式電子顯微鏡(Multi-Function Scanning Electron Microscope，簡稱M-SEM)於不同操作條件下探討其薄膜阻塞與流通量。
本研究共分為三階段，其中包含不同水力停留時間(Hydraulic Retention Time，HRT)以及不同操作模式：
Stage 1：HRT為18小時，生物處理槽為持續好氧，薄膜槽進行好氧與缺氧交替，好氧與缺氧之時間比為1.5：1小時，整體上來看，於一個HRT循環下，總好氧時間為13.6小時，總缺氧時間為4.4小時。
Stage 2：HRT為22.4小時，生物處理槽持續好氧，薄膜槽進行好氧與缺氧交替，好氧與缺氧之時間比為2.1：1小時，整體上來看，於一個HRT循環下，總好氧時間為18小時，總缺氧時間為4.4小時。
Stage 3：HRT為22.4小時，生物處理槽進行好氧與缺氧交替，好氧與缺氧之時間比為1：1小時，薄膜槽持續好氧，整體上來看，於一個HRT循環下，總好氧時間為18小時，總缺氧時間為4.4小時。
研究結果得知，HRT對於MLSS濃度有顯著的影響，HRT=18小時下之生物濃度高於HRT=22.4小時。然而，第一階段操作模式下，生物處理槽之MLSS濃度為8000~23300mg/L，第三階段操作模式下，生物處理槽之MLSS濃度降至為3000~5000mg/L，薄膜槽之生物濃度變化趨勢具有與生物處理槽相同的趨勢，但是第三階段操作模式下薄膜槽之生物濃度略高於生物處理槽；在MLVSS/MLSS方面，在三階段操作模式下MLVSS/MLSS介於0.4~0.8，絕大部分為0.5。
BOD之去除效率方面，其去除效率穩定且可高達99.2％，COD與TOC去除效率最高分別為88.8％、83.6％。系統中各槽之BOD體積負荷方面，生物處理槽介於3.03~5.42kg BOD5/m3 day，薄膜槽介於0.02~0.06kg BOD5/m3 day，BOD食微比方面，生物處理槽介於0.9~2.3 kg BOD5/kg MLVSS-day，薄膜槽介於0.01~0.02 kg BOD5/kg MLVSS-day；系統中各槽之COD體積負荷方面，生物處理槽介於6.7~10.8kg COD/m3 day，薄膜槽介於0.28~0.68kg COD/m3 day，COD食微比方面，生物處理槽介於3~4 kg COD/kg MLVSS-day，薄膜槽介於0.1~0.37kg COD/kg MLVSS-day。其值較一般傳統活性污泥程序為高。
在TKN去除方面，Stage 1之去除效率為42±13.9％，Stage 2與Stage 3之去除效果分別為48.9±13.7％ 與 54.4±13.7％。然而，TKN之去除效率為40~50％，但出流水亞硝酸與硝酸之濃度非常低，其可表示此系統中脫硝作用與硝酸化作用(nitratation，NO2-→ NO3-)顯著，但其亞硝酸化作用(nitritation，NH4+ → NO2-)受到抑制。
本研究於操作期間薄膜共進行三次化學清洗，分別為第24、56與78天，但在第99天利用化學清洗薄膜已無法恢復通量，因此更換新的薄膜，此薄膜材質與孔徑皆與原薄膜相同，於第99天後不再進行化學清洗，但其清水清洗次數變為頻繁，以約一星期清洗一次改為約兩天清洗一次。本研究並針對新的薄膜與阻塞之薄膜，利用SEM掃描結果，進行其阻塞與過濾特性之描述。
In this study, the performances of a membrane bioreactor for anoxic/aerobic treatment of continuous-flow ABS resin manufacturing wastewater were evaluated by laboratory-scale experiments with no sludge withdrawn for 174 days. This system consists of two reactors, the first reactor is a biological treatment tank (2.3 L) and the second one is a membrane bioreactor (3.5 L). The removal efficiencies of carbon and nitrogen were examined in terms of BOD, COD, TOC, Org-N, NH4+-N, NO2--N and NO3--N. Furthermore, the mixed liquid suspended solids (MLSS), mixed liquid volatile suspended solids (MLVSS), food to microorganism ratios (F/M) and volumetric loadings were measured in this study. Finally, the permeate quantity and fouling of membrane according to different experimental conditions were estimated by flux and SEM images.
Tree stages include different hydraulic retention time (HRT) and operation modes were carried out in this study as follows.
Stage 1: The total HRT of this system was 18 hours. The biological treatment tank was continuous aerobic, the membrane bioreactor was operated at the anoxic/aerobic cycle for 1/1.5 hours. Total aerobic time of this system was 13.6 hours and the anoxic time was 4.4 hours for each HRT.
Stage 2: The total HRT of this system was 22.4 hours. The biological treatment tank was continuous aerobic, the membrane bioreactor was operated at the anoxic/aerobic cycle for 1/2.1 hours. Total aerobic time of this system was 18 hours and the anoxic time was 4.4 hours for each HRT.
Stage 3: The total HRT of this system was 22.4 hours. The biological treatment tank was operated at the anoxic/aerobic cycle for 1/1 hour. The membrane bioreactor was continuous aerobic. Total aerobic time of this system was 18 hours and the anoxic time was 4.4 hours for each HRT.
The results indicated that the HRT seems significantly to affect the concentration of MLSS. The HRT of 18 hours gave the higher biomass concentration than HRT of 22.4 hours. However, the MLSS concentration of the first stage in biological treatment tank was in a range of 8000 to 23300 mg/L. Consequently, decreased and maintained at a range from 3000 to 5000 mg/L for stage 3. A similar MLSS concentration variation was observed in membrane tank. However, the MLSS concentrations of membrane tank were higher than that of biological treatment tank expect for stage 3. The MLVSS/MLSS ratios were also calculated for both the biological and membrane tanks. In generally, the MLVSS/MLSS ratios are in a range of 0.4 to 0.8, but most of the values were below 0.5.
Generally, the effluent concentrations of BOD were remarkably stable and with a highest removal efficiency of 99.2%. COD and TOC removal was 88.8% and 83.6%, respectively. The BOD volumetric loadings of biological treatment tank were in the range of 3.03~5.42kg BOD5/m3 day and in the range of 0.02~0.06kg BOD5/m3 day for membrane tank. The BOD F/M of biological treatment tank was in the range of 0.9~2.3 kg BOD5/kg MLVSS-day and in the range of 0.01~0.02 kg BOD5/kg MLVSS-day for membrane tank. The COD volumetric loadings of biological treatment tank were in the range of 6.7~10.8kg COD/m3 day and in the range of 0.28~0.68kg COD/m3 day for membrane tank. The COD F/M of biological treatment tank was in the range of 3~4kg COD/m3 day and in the range of 0.1~0.37 kg COD/kg MLVSS-day for membrane tank. Compared to the data obtained from conventional aerobic processes, those data obtained from this study were very high.
The removal efficiency of TKN for stage 1 was 42±13.9%, for stage 2 and stage 3 was 48.9±13.7% and 54.4±13.7%, respectively. However, the results showed that low nitrite and nitrate concentrations in the effluent and only 40% to 50% TKN removed. It implied that denitrification can be accomplished and nitrification of the nitritation (NH4+ → NO2-) is inhibited but the nitratation (NO2-→ NO3-) is significant.
In the first 78 days, tree times of chemical cleanings were carried out. It was 24, 56 and 78 day, respectively. On days 99, the flux can be not recovered even the chemical cleaning. A new hollow fiber membrane module with the same material and pore size replace the fouled one and no more chemical cleaning after days 99. However, the water cleaning was more frequent from once a week to once two days after the membrane changed. Images of the new, fouled and cleaned membrane using SEM provide a means for qualitative description of fouling in this study. |
