Research On Forecasting And Tracing Method Of VOCs Emission Dispersion Of Typcial Chemical Industrial Park

VOCs emissions are one of the most important precursors of ambient PM_2.5 and photochemical smog in air pollution in China.In recent years,the number of chemical industrial parks in China has increased significantly,and a large number of chemical companies have been gathered in various chemical parks,making the chemical industrial park one of the major sources of VOCs emissions in China.A variety of toxic and harmful components,which are contained in VOCs,will seriously affect the ambient air quality and the health of the residents in the surrounding area.This paper focuses on the influence of VOCs emission of chemical industrial park,and the multi-pollutant emission characteristics and diffusion effects of composite pollution sources in chemical industrial park and the key technologies for the impact of VOCs emissions on air quality in surrounding areas were investigated based on the combination study of test analysis and model construction for VOCs emissions from typical chemical industrial park.Firstly,the emission characteristics of VOCs of different types of pollution sources and the application of different VOCs treatment facilities in chemical industial park was discussed,based on the production process investigation and on-site sampling analysis of different enterprises by GC-MS system in the chemical industrial park,in view of the complex multi-pollutant emission of composite pollution sources such as point source,surface source and volume source.The results of five typical chemical company including pharmaceutical factory,dyeware factory,pesticide factory,and printing factory,showed that the main VOCs emission source is the VOCs-contained solvent and the main VOCs components are the Chlorinated hydrocarbon（Methyl chloride,Dichloromethane,Trichloromethane,chlorobenzene,Dichlorobenzene,etc.）,Benzene（benzene,Toluene,Xylene,etc.）,Isopropanol,and etc.Additionally,a cost estimation model for different VOCs emission treatment facility was developed based on cost-benefit analysis method and the investigation and survey of the application and operation of different VOCs emission treatment faciligy in the chemical industrial park.The operating cost of different case in the chemical industrial park was estimated,and the factors which could affect the operating cost and the applicable conditions for different VOCs emission treatment facility werer also presented.Secondly,considering that the enterprise production adjustment could result in the variation of the VOCs emission characteristics,a method to estimate the dynamic VOCs emission characteristics of different pollution source was established combined the standard mass spectrum distribution of different VOCs components in NIST database,which is based on the real-time mass spectrum result of the online magnetic mass spectrometry system in the chemical industrial park and the predetermined VOCs emission characteristics of different pollution source.This method converted the analytical solution solving problem to optimization problem and established a mathematical model,which took the concentration of main VOCs components as the unknow number and the number of VOCs component as the dimension of the solution space.Then an improved swarm intelligence algorithm was applied to solve the optimal solution to get the concentration of main VOCs components.This method could be used to get the time variation characteristic of the main VOCs component of different pollution sources based on the real-time mass spectrum results.Thirdly,in order to investigate the influence of VOCs emissions from the chemical industrial park under different meteorological conditions on the ambient air quality of different receptors,the AERMOD model was used to study the diffusion characteristics of VOCs emissions from the chemical industrial park,based on the meteorological data of the entire year 2016 according to WRF model.In AERMOD model,the dynamic pollutant emission intensity of different pollution sources of 5 typical chemical companies was also considered.The results showed that there was some situation that the pollutant concentration of the receptor was extremely high,although the remarkably low pollutant concentration were common in most cases.Furthermore,only the emission of 5 typical companies were considered in the research,so the influence of the total emission of the chemical industrial park on the ambient air quality should be considered.Fourthly,considering that the conventional meteorological forecast data is insufficient for the application of AERMOD model and the AERMOD model is unavailable for calm wind meteorology situation,an association database which combined the conventional meteorological forecast data,the dynamic emission intensity of different pollution source,and the ambient pollutant concentration of the receptor was established,then a concentration predicting model of the receptor was constructed based on deep neural network and the database.Additionally,a specialized prediction model based on the puff integration method was developed to handle the calm wind situation.The raw training dataset was developed taking the conventional meteorological forecast data and the dynamic emission intensity of different pollution source as input and the ambient pollutant concentration of the receptor as output.Due to the imbalance features of the raw dataset,different oversampling algorithm were deployed and the preprocessing training dataset was developed.Then the pollutant concentration predicting model was developed,which could identify the high pollution situation with an accuracy of 94.7%,and the R-sequared of predicting the pollutant concentration of high pollution situation is 0.977.Finally,a traceability pollutant method for the ambient pollution of receptor was established to analyze the influence of different emission sourcse on the ambient air quality and quantify the contribution of the main emission source under different metorological conditions,based on the air dispersion model and the pollutant concentration predicting model.This method could estimation the contribution percentages of different emission sources based on the dynamic emission intensity of different emission sources and the meteorology data.Then,a method for developing the emission reduction strategy for different emission sources was established,which was based on the current ambient pollutant concentration and the environmental concentration limits.The emergency emission reduction adjustment could be achieved by the method,and the different emission reduction percentage for each emission source could be provided based on the emission intensity of different sources.