CFSBR水質推估公式之發展與建立

Development of Water Quality Evalute Formulas for CFSBR

尤世雄(Se-Siong Yu)

指導教授 ： 廖述良

國立中央大學/工學院/環境工程研究所/碩士(2006年)

摘要

連續流批次式生物處理系統(Continuous-Flow Sequencing Batch Reactor, CFSBR)在自動監控系統的發展上，主要是利用監測pH及ORP的折點來決定相位的改變，已可有效地提升去除效率、放流水質的穩定性及降低成本，但是由於部分的水質項目，如COD、氨氮、硝酸氮、亞硝酸氮及正磷酸等尚無有效的連續自動監測技術，致厭氧相無法準確掌握有機氮的脫氨作用與微生物釋磷作用等操控需求，而影響後續好氧相的硝化與攝磷的作用，造成去除率不佳與放流水質不穩定等結果。
雖然現有的線上即時水質水量自動監測通常無法作到全面的監測，但對一個「半開放」的CFSBR而言，其水質特性相互之間存在一定的關係，是故，本研究利用類神經網路來建立水質推估公式，以利用即時的監測項目資料來同步推估未監測的水質項目資料。爲了能有效地掌握水質的變化趨勢，本研究嘗試於輸入變數中加入初始値以及反應狀態變數。由研究結果指出，於輸入變數加入初始值後的類神經網路模式，確實能更有效地推估磷酸鹽、亞硝酸氮、硝酸氮與氨氮的變化趨勢。

關鍵字

連續流批次式生物處理系統 ； 自動監測控制 ； 類神經網路

並列摘要

The continuous-flow sequencing batch reactor (CFSBR) mainly utilizes and monitors the bend of pH and ORP to determine the change of the phase in the development of the real-time control system, the stability of water quality and reduction of cost can already improve and get rid of efficiency. However, some water-quality characteristics, for instance COD, NH4+-N, NO2--N, NO3--N and PO43--P, etc. do not have effectively automatic monitoring technology yet. It makes the anaerobic phase unable to grasp an ammonification and bio-phosphate release, and influence nitrification and bio-phosphate up-take in the follow-up aerobic phase. Thus it leads the bad removal rate and the unstable effluent quality.
Though the existing on-line sensors are usually unable to accomplish overall monitoring, these variables are determined with a significant time delay. However, CFSBR is a half-opening system; its water-quality characteristic has certain relation with each other. In this study, the water quality evaluate formulas was developed using the network approach, which can simultaneously utilize on-line information to evaluate water quality. In monitoring and controlling CFSBR, the information of nutrient dynamics is very important. In this reason, this study tries to join the initial value and the state variable into input data. And the results show this method can evaluate NH4+-N, NO2--N, NO3--N and PO43--P concentrations and trends well.

並列關鍵字

Automation real-time control ； Continuous-Flow Sequencing Batch Reactor(CFSBR) ； Neural network

參考文獻

3.Charpentier J., Godart H., Martin G. and Mogno Y., “Oxidation-Reduction Potential as a Way to Optimize Aeration and C, N and P Removal: Experimental Basis and Various Full-Scale Examples,” Wat. Sci. Tech., Brighton, 21, 1209-1223. (1989).

4.Chuang S.H., Ouyang C.F., Yuang H.C. and You S.J., “Phosphorus and polyhydroxyalkanoates variation in a combined process with activated sludge and bilfilm,” Wat. Sci. Tech. 37(4-5), 593-597(1997).

5.Dae Sung Lee, Jong Moon Park, “Neural network modeling for on-line estimation of nutrient dynamics in a sequentially-operated batch reactor,” Journal of Biotechnology, 75, 229-239. (1999).

6.Dong-Jin Choi and Heekyung Park, “A hybrid artificial neural network as a software sensor for optimal control of a wastewater treatment process,” Pergamon, Wat. Res. Vol. 35, NO. 16, 3959-3967(2001).

7.Eckenfelder W. W. and Hood J. W., “The Application of Potential to Biological Waste Treatment Process Control,” Proceeding of 6th Purdue Industrial Waste Conference, Feb. 21-23, Purdue University, West Lafayette, Indiana, 221-238. (1951).