Ethylene Process Modeling Based On Strict Superstructure

With the continuous improvement of carbon emission requirements,environmental problems caused by carbon emissions have increasingly become a huge problem facing the petrochemical industry,especially the ethylene industry.Reducing carbon emissions,guaranteeing capacity demand and achieving simultaneous improvement of economy and environment are of great significance to the future development of ethylene industry.Modeling and analysis of ethylene process is one of the effective ways to solve environmental problems caused by industrial production.However,due to the strong non-linear and multi-variable coupling characteristics of ethylene industry,traditional mechanism modeling methods cannot respond to the system behavior of the process,accurately and objectively.However,data-driven model,which is simple and effective,does not require an accurate process model.Data-driven model and mechanism model can complement each other.Therefore,this paper presents a data-driven process simplification model and a comprehensive prediction model.In the process' simplification model,the basic model can be obtained by using the process graph theory(P-graph)framework,and then the structure of P-graph model can be simplified by the carbon emission pinch analysis method;and the optimal and near optimal solution are obtained,simultaneously.By considering the overall and regional product demand and carbon emission constraint in the comprehensive prediction model,the regional raw material distribution in 2005,2020 and 2030 were studied to achieve the overall optimal output structure and supply-demand balance,respectively.The experimental results show that in the process simplification model,the simplification of the maximum structure generation by 60%,and the simplification of feasible decomposition network by 73%.At the same time,the decrease of the carbon emission per unit product by 17%,which plays a better role in simplifying the model structure.In the comprehensive forecasting model,the use of raw materials will meet the regional production demand and carbon emission constraints in 2005,but in 2020 and 2030,the production demand will be reduced by 4700.86t and 15100.86t respectively to meet the overall and regional carbon emission constraints.The comprehensive forecasting model providing a scientific basis for the development of reasonable raw material scheduling.