The Study On Dynamic Warning And Monitoring Modeling For Water Pollution And Risk Analysis For Emergency Disposal Project

The increasing number of water pollution accidents and their corresponding serious harm consequences prompt governments to come up with new requires about the emergency capability construction,which lead to the development of the pollution early warning and response technology and serve for water quality security.This study aims to solve the main technical problems in the process of response to sudden pollution accidents in our country.On the basis of investigation the scientific problems during the sudden pollution emergency response process both at home and abroad,some contents including the abnormal detection warning for sudden pollution accidents,the emergency monitoring for sudden pollution accidents,and the start discrimination and risk analysis for emergency disposal engineering are carried out.In order to solve the problem of early warning for sudden pollution accidents,the surrogate measurement method related to water quality parameters during sudden pollution is set up through the regression analysis and correlation analysis based on routine high frequency continuous water quality monitoring data.The methods of wavelet denoising analysis and neural network prediction are introduced to forecast the water quality and tend to build the abnormal alarm framework by combination of the abnormal threshold judgment standard through comprehensive consideration of the characteristics of sudden pollution accidents and the abnormal threshold of history water quality time series.Both of the two monitoring sites which located in the upstream and downstream of Potomac River respectively are selected as the instance verification.The results show that the correlation of water quality parameters of upstream is more apparent than that in the downstream,especially the correlation coefficients between turbidity and TSS or TP in the upstream,which exhibit more than 0.9.However,they are less than 0.73 in the downstream.The wavelet neural network can improve the accuracy of the abnormal warning by removing single abnormal points and can well identify the abnormal events caused by sudden pollution.The area of receiver operating characteristic curve is above 0.98,which indicates that the method can meet the requirements of abnormal warning.In order to solve the problem of emergency monitoring for early warning of sudden pollution accident,the optimization framework of emergency warning monitoring network based on information entropy is constructed.The hydrodynamic process of pollution migration is simulated by using the historical hydrological and water quality data.The pollutant concentration at different location and monitoring interval in the direction of flow is calculated and the pollutant distribution curve is drawn according to the simulation data of water quality model.The pollutant distribution curves withdifferent locations and time intervals are converted to probability density curves.Then the information entropy and information transfer indexes are calculated on the basis of the probability density curves.Besides,combined with Shannon sampling theorem and Fourier transform,the frameworks of the optimal locations and time intervals of monitoring sites are built.Finally,the framework is verified by the tracer experiment in Truckee River and the nitrobenzene spill in Songhua River.The results exhibit that the simulation results of the water quality model in this framework are well consistent with the actual monitoring results,which means it can be introduced to the analysis of the optimization of monitoring sites.The frameworks,as an effective tool,not only can be used for the design of the emergency monitoring plan during the process of river basin's planning and management,but also for the layout of the emergency response after the sudden pollution accidents.In order to solve the problem of emergency monitoring for source inversion of sudden pollution accident,the source inversion methods reported by the previous studies are analyzed and classified.The typical methods in each classifies are selected to design the optimal monitoring networks.The application scope of different inversion methods and corresponding guidelines of how to design the emergency monitoring network are discussed by combination of tracer experimental data and the method.The results show that each inversion method has different requirements on the monitoring locations and monitoring frequency.The inversion methods are selected by consideration of different emission types and hydraulic conditions.The decision makers should select the appropriate inversion method and set up the corresponding emergency monitoring network in terms of the actual situation of the project.In order to solve the problem of the decision of emergency disposal engineering,a start discrimination model of the disposal project is established on the basis of the spatial-temporal characteristics and threat evaluation of pollution,and there is an overall evaluation about the risk of the pollution accident,the transfer of pollution,and the alarm of spatial-temporal characteristics of the pollution threaten.In addition,the computational procedure of spatial-temporal characteristics of pollution accidents and the evaluation index system of pollutant threat degree are developed and further integrated in this study.The start discrimination technology and decision-making tools are tested by the case of South-to-North Water Diversion Project as well as the successful development of emergency disposal project start judging module.These products are integrated into the disposal of pollution warning framework of South-to-North Water Diversion Project and fulfill the procedural application of emergency control disposal engineering measures.In order to solve the problem of the engineering risk during the implement of emergency disposal engineering,a risk quantitative assessment method for emergencyof water pollution tends to establish.The concept of the emergency disposal engineering risk is proposed and the risk assessment index system of emergency disposal project with 10 sub-indexes is established.The occurrence probability of basic events and minimum cut sets caused by the top event is calculated using fuzzy fault tree,which is based on the actual emergency disposal project about the aniline pollution incident in Zhuozhang River in Shanxi Province.Moreover,the contribution rates of top event for minimum cut sets are ranked,and then the suggestions on how to prevent the occurrence of the top event are given.This method provides the decision base for decision makers how to use the limit emergency resource and offer the effective preventive measures to avoid engineering risk.These results of this dissertation are able to provide reference for the prevention and control of sudden pollution in China's basin,further improve the theoretical foundation of water quality monitoring,and provide technical support for river system management of water system.