Study On Risk Early-Warning Technology Based On Uncertainty Analysis For Sudden Water Pollution Accidents

In recent years, with the more and more frequent occurrence of sudden water pollution accidents, the lives, property and safety of citizens and the sustainable development of the society are under severe threats. In order to better cope with these sudden water pollution accidents, it is urgent to study the technology of water quality forecasting and early earning, and to establish the risk early-warning model.In this thesis, the main focus is the technology of risk early-warning model against the sudden water pollution accidents. A risk early-warning model based on uncertain methods is proposed to cope with the sudden water pollution accidents affecting downstream areas (denoted as EPs). Then, an actual case is used to verify this model. Moreover, an optimization method is presented based on uncertainty analysis to improve the accuracy of the prediction. The main research work and innovative aspects of the thesis are as follows:(1) In order to reduce the influence of uncertainty on the results of the early warning, and improve the research about the early warning of sudden water pollution accidents, a risk early-warning model (EP-Risk model) has been proposed based on uncertainty analysis and risk assessment method. The risk early-warning is achieved through four steps:the simulation of pollutant, the calculation of pollution accident happening probability, the calculation of the influence of the pollution accident and the confirmation of risk level. Through uncertainty analysis, the diffusion rule of the pollutant can be solved out by a simulation method. Refer to the impact of pollution which was known based on the predication in the river water quality, the risk level of the sudden water pollution accident can be given out with the help of risk matrix method.(2) In order to predict the sudden water pollution accidents more accurately and in time, an EP-Risk model has been established based on Monte Carlo Method, analytic hierarchy process (AHP) method and Risk Matrix method. First, the likelihood of water pollution at the EP is calculated by the Monte Carlo method, which is used for uncertainty analysis of pollutants’transport in rivers. Second, the impact of water pollution at the EP is evaluated by expert knowledge and the results of Monte Carlo simulation based on the Analytic Hierarchy Process. Then, the final risk level of water pollution at the EP is determined by the Risk Matrix method. Finally, the proposed method is applied to a real Chinese spill accident and the warning result is satisfying.(3) In order to reduce the influence of uncertainty on the result of water quality prediction, a dynamic optimization method to predict water quality has been proposed based on generalized likelihood uncertainty estimation (GLUE) method. After the accident happened in upstream of the river, the real-time monitoring data can dynamically update before the pollutant diffuses to downstream area and causes negative influence. Therefore, water quality model parameters can be optimized, thus the accuracy of the warning can be effectively improved.