Service Level Research Of Water Quality In Water Distribution Svstems Based On Residual Chlorine And THMs

To evaluate and optimize the water quality in water distribution systems, extensional residual chlorine-I/O model and THMs-I/O model were studied and used as the calculation basis to build the multi-objective optimization model beween service level of water quality in water distribution systems and chlorination costs. In addition, there are errors when water quality monitoring stations respond the real service level of water quality in water distribution systems. Therefore, the service level of water quality in water distribution system was corrected in this study.Firstly, in order to make the residual chlorine-I/O models suitable in residual chlorine slow and fast reaction coexistence and improve computational efficiency of optimization model, extensional chlorine-I/O model and THMs-I/O model were derivated. The experiments were conducted to investigate chlorine decay with tap water at the end of the pipe network and filtrated water of water plant. The results showed that fast reaction of chlorine decay basically finish when initial chlorine mass concentration is above some value. Therefore, the chlorine mass concentration of water quality monitoring stations is linearly proportional relationship with initial mass concentration in slow reaction on chlorination boosters. Moreover, extensional chlorine-I/O model was presented in the pilot-scale water distribution system. The results showed that the extensional chlorine-I/O model is better than traditional chlorine-I/O model. In addition, to match generation of THMs, formation model of THMs based on chlorine consumption was proposed. The effect of three factors including temperature, initial chlorine concentration and bromine ion concentration, on proportional coefficient of the linear model were studied. Experiment results showed that bromine ion concentration had the most effect on the proportional coefficient of the linear model, and the proportional coefficient was independent on temperature and initial chlorine concentration. Moreover, the THMs-I/O model was proved in the pilot-scale water distribution systems and the simulation results showed that the error between real value and predictive value is below9%.Secondly, to study an evaluation model of water quality in water distribution systems, concept of residual chlorine servicel level was proposed to study the evaluation function of residual service level and improve the optimization model in rechlotination. By reference to chlorine performance and "Standards for Drinking Water Quality", service level of each component of THMs was established. Three kinds of expressions, which named weighted average value, standard deviation and multiplication, were used to measure service level of water quality. Multi-objective optimization model of service level of water quality and rechlorination cost was proposed. Chlorine decay and THMs generation were simulated using EPANETMSX. An improved non-dominated sorted genetic algorithm-Ⅱ with integer code was adopted to optimize the model, and pareto fronts could be obtained under different conditions. Results showed that the weighted average expression was the optimal way to correlate residual chlorine and THMs formation. It was enough that boosters be located on outlet of water tower and end of water distribution systems. Too much number of boosters can not significantly improve service level of water quality, but can increase total cost of rechlorination. Therefore, improving water quality and replacement of old pipes was the most feasible method to improve service level of water quality and save total cost of rechlorination.Finally, as there are errors when water quality monitoring stations respond the real service level of water quality in water distribution systems, the service level of water quality in water distribution system was corrected in this study. The results showed that, with the number of water quality monitoring stations increases, the errors tended to decrease and it was not monotonically decrease.The errors would be ignored when the number of water quality monitoring stations was large. In this thesis, chlorine monitoring coverage and ratio between the number of chlorine monitoring stations and node number of water distribution system were used to compare their measurement errors. The results showed that later could be better to measure the errors. By verifying a water distribution system in a large-sized city, the results showed that the corrected service level formula of water quality is closer to the real service level of water quality. The corrected service level formula of water quality was providing a theoretical foundation for running and controlling water distribution systems.This thesis was kindliy supported by National Major Project of Science&Technology Ministry of China (2008ZX07420-004,2009ZX07424-001and2009ZX07424-004) and National Natural Science Foundation of China (No.51208455).