Operation Optimization Of The Rectisol Based On Data Analysis And Process Simulation

In the chemical industry,the rectisol process is a relatively common acid gas purification unit.Its operation not only affects energy consumption,but also affects the quality of subsequent products and atmospheric environmental issues.However,due to its complex mechanism,numerous variables and the need to consume a lot of utilities,how to fit the adjustable parameters without changing the process and equipment so that the unit is always in an ideal operating state,reducing energy consumption and production costs,it has always been a hot issue in industry and academia.For this reason,this paper uses data analysis and mechanism modeling methods to analyze the characteristics of the rectisol process to model the exhaust gas and energy consumption.The main research is as follows:1.Use data analysis to solve the problem of excessive methane content in exhaust gas.In actual operation,the daily inlet flow will change,which leads to the regulation of methane as a dynamic process,which requires real-time analysis.The data analysis can efficiently and quickly find out the key factors restricting methane content and guide adjustments.In this paper,BP neural network and LSTM network are used for modeling respectively,and finally it is determined that LSTM network is more suitable for this process.Then combined with the historical retrospective method for methane control,the methane emissions can be continuously reduced.2.Use PSRK equation to simulate the rectisol process.The focus is on the H₂S absorption column,CO₂ absorption column,CO₂ flash column,H₂S concentration column,thermal regeneration column and tail gas scrubber,and the simulation results are in good agreement with the actual values.Aiming at the inaccurate problem of the dissolution heat of CO₂ and other substances in methanol,the method of heat compensation was used to correct the problem,and the concentration distribution of the main components in the column was analyzed.3.Use the PSO algorithm combined with Aspen Plus to find the decision variables when the economic benefits are optimal.Through the analysis of the process,the decision variables and their feasible regions are determined.In this paper,six decision variables are selected:lean methanol feed flow,semi-lean methanol feed flow,the flowrate of stripping N₂ into CO₂flash column,The flowrate of stripping N₂ into the H₂S concentration column,the flowrate of methanol into the H₂S concentration column,and the flow of the CO2 flash column into the H₂S concentration tower III.With exhaust gas and purified gas indicators as constraints,the optimized process can reduce the economic cost of 1.6 million yuan per year compared with the original process.