| Shenyang Xiannvhe Wastewater Treatment Plant (400000m3/d) are troubling with process controlling parameters being changeless in many years, high operation cost, the lack of intelligent control system for operation process. In this dissertation, Variance and least significant difference, principal components analysis (PCA), orthogonal test and BP neural network optimized by genetic algorithm are used to optimize and modify the key process parameters of BAF and the original aeration system have been reconstructed. The research achieved the aim of reducing operational cost, improving economical and environmental benefit. The results can set a scientific and technical basis for the management of other BAF wastewater treatment plant.Through regulating the influent water flow and controlling hydraulic load, we can ascertain the pollutant load of BAF and the maximum pollutant removal amount daily. Adopting different tactics according to the above conclusion, the study resolved the problem of water quantity resulting from bigger design load. The water quality can be improved from 97 percent to 100 percent and power consumption reduced from 0.27kWh/m3to 0.21kWh/m3.Through dividing the timing of backwashing into different sessions and detecting the SS concentration of each interval, we can discover the law and characteristics of backwashing in C/N filter, N filter and material clogging filter. By means of measuring and anatomizing of filter material, we analyzed the mechanism of air wash, air-water wash and water wash. The study broadened and deeped the theory of backwashing in BAF.The new method of combining the significant difference test with SS concentration to decide the timing of backwashing is developed. We optimized the former parameters of backwashing. After modifying, the removal rate of pollutant is improved, the effect of backwashing is bettered, and filter clogging is solved, and it saved 561500kWh power in a year.By utilizing analysis of Variance and least significant difference, we found the right concentration of DO and reconstructed the original aeration system. The aeration system was changed from one blower of one filter to one blower of two filters. After optimization on aeration system, the removal rate of pollutant was steady and 5702400kWh power was saved in a year.We used principal components analysis (PCA) to ascertain the 3 key factors of NH4+-N removal rate and had do a detailed study on each factor. Taking operation strategies according to the above conclusion, the water quality is improved insignificantly and NH4+-N removal rate is going up from 21 percent to 70 percent.The capability of simultaneous nitrification and denitrification (SND) in BAF had been studied. The further study indicated that the removal rate of SND can be affected by DO, hydraulic load and COD load. Adjusting the new tactics according to the above conclusion, the removal rate of TN was improved, the water quality of outlet was bettered, and the potential of BAF was developed. The result indicated that removal rate of TN was improved from 18.8 percent to 29.2 percent.To reduce the cost of disposal sludge in BAF, we using orthogonal test to find the combination of centrifuge torque, sludge flow rate and the amount of PAM. After carrying out the above optimization strategy, we solved the problem of PAM large amount usage and the PAM cost economized 120000 yuan. A new method of controlling power distribution in BAF was developed. Using databank, we set up artificial neural network predicting model, then optimized the BP neural network by means of genetic algorithm. After training the parameters of ANN and testing of model, we discover that the predicting value has no significant difference with the reality value, the predicting value is 0.92kWh/kgCOD, and the reality value is 0.94kWh/kgCOD,but the electricity energy consumption treated 0.99 kWh/kgCOD before optimization. This method has actual value, and it can direct the management of production and operation, it can provide a new thinking for power cost control. |
PDF Full Text Request