Study Of Trade-offs Between Energy Conservation And Emission Reduction For Activated Sludge Process Based On Multi-objective Optimization

Activated sluge process (ASP) is the most important method for wastewater treatment. After 100 years development, the research emphasis has been focused from achieving stable operation at the beginning, followed by satisfying the effluent standard, and so far to the stage of energy conservation and emission reduction.In China, one important target of energy conservation and emission reduction for wastewater treatment plant (WWTP) is that effluent violation should be minimized with limited operation cost, and then effluent quality should be satisfied with the discharge standard eventually.Lots of studies show that treatment processes and equipment may be added or the operation parameters may be adjusted during the upgrading reconstruction of WWTP, which means extra operation cost is needed. Therefore relations between energy conservation and emission reduction are conflicting sometimes, where the trade-offs between these two objectives should be evaluated based on the actual situation.Trade-offs between energy conservation and emission reduction in three processes:the Benchmark simulation process (BSM1), the Five-tank integrative activated sludge process (FTIASP) and the Six-tank integrative activated sludge process (STIASP) developmented by Southeast University, were analyzed by multi-objective optimization method in this study, where three weather influent conditions of Dry, Rain and Storm, and two control methods of open-loop and closed-loop were used.Multi-objective optimization models for ASP were built. The objevtives were:effluent quality (EQ), operation cost (OC), total volume and total suspended solid; the decision variables were:five tank volumes, three oxygen transfer coefficients, mixed liquor recycle, sludge recycle, excess sludge wasting rate and step operating time, etc.; the constrains were the ranges of decision variables and EQ. The multi-objective optimization models were solved by the Non-dominated Sorting Genetic Algorithms Ⅱ (NSGA-Ⅱ) inherited in MATLAB.In order to obtain concentrations of pollutants in the multi-objective optimization models, mathematical models for the three ASPs were built according to the mass balance theory. Reaction rates of pollutants were calculated based on the ASMs (activated sludge models) proposed by IWA (International Water Association). In order to calibrate the model, characteristics of influent COD were determined by using methods of gas chromatography, the respirometry test and mass balance analysis; stoichiometric coefficients were determined by lab experiments; and the kinetic parameters were calibrated by using an iterative calculation procedure, where reaction rate tests and simulations, sensitivity analysis and optimization method were employed. Results showed that satisfying simulation results could be obtained, and the parameter identifiability problem could be solved.Pareto solutions obtained from BSM1 showed that different settings for constrains or objectives had obvious influence on the results. The trade-offs between different objectives could be established through the Pareto solutions, where lots of solutions could be calculated and could be provided for decision makers (DM). Hence multi-objective optimization could separate the tasks between the optimizer and DM:the optimizer was responsible for the optimization; while the DM made the final dicision. Pareto solutions obtained from the model considering four objectives could show the DM a more comprehensive understanding of BSM1 strategies than the models considering two or three objectives.Pareto solutions obtained from FTIASP showed that the trade-offs between EQ’and OC’ were nonlinear, but could be separated into three parts after linear fitting. In the first part |△OC’/△EQ’|=1.24, the second part|△OC’/△EQ’|=0.15 and the third part|△OC’/△EQ’|=0.03, which meant energy amount needed in the first part was 41 times than that in the third part,8 times than that in the second part when achieving the same emission reduction. It indicated that there were lots of options when evaluating the trade-offs between nergy conservation and emission reduction during the upgrading reconstruction of WWTP.One solution was selected from results of FTIASP, and was compared with the default strategy from reaction amount of different processes, different steps and different tanks. Results showed that the selected solution was better than the default. The closed-loop strategies of automatic control of step operating time and the modified FTIASP by adjusting the aeration time of tank3 could achieve much better performances of energy conservation and emission reduction.Using the multi-objective optimization method to evaluate trade-offs between EQ" and OC" in STIASP, where one open-loop fixed time operation (FTO) and three closed-loop time controller strategies (TCNO3, TCNH4 and TCPO4) were analyzed. Results showed that, the optimized FTO strategy could achieve better EQ" and OC" that that of the default strategy, and moreover, TCNO3 strategy based on SNO3 and dSNO3/dt could make EQ" and OC" even better.