Multi-objective Optimization Of Atmospheric And Vacuum Distillation Unit Operation Based On Oil Blending

Atmospheric and vacuum distillation unit is one of the most important unit inpetrochemical enterprise，its operation optimization directly related to the quality ofthe product, product yield and energy consumption，which has important influence onpetrochemical enterprise’s operating efficiency. Therefore, the research onatmospheric and vacuum distillation unit operation optimization has been praised bythe petrochemical enterprise’s attention. Atmospheric and vacuum distillationunitof5million tons/year are simulated byASPEN sortware，which could obtain the relevant operating data, includingapparatus of material balance, energy balance, the temperature distribution, stressdistribution of the tower, and so on. accurate understanding and grasping of wholedevice running status could be achieved.the results show that the device runningstatus has been reflected by ASPEN simulation.On this basis,An correction method for oil TBP curve is proposed towards thechangeable quality of feed oil, time-consuming and high-cost of oil TBP curvemeasurement. The optimization objective of minimizing quadraticsum error betweensimulation value and operating conditions is performed instead of TBP curvecorrection. The accuracy of simulation is significantly improved by this method.On the premise of the simulation accuracy，taking stripping steam, stripperproduction and mid-pumparound as operating variables, and taking maximum ofatmospheric distillation column yield and minimum of energy consumption asoptimization objectives, multi-objective optimization model of atmospheric andvacuum distillation is established. The optimal solution of atmospheric distillationcolumn yield&energy consumption is obtained by Non-dominated sorting geneticalgorithm with elitist strategy (NSGA-II), which provides a theoretical basis forprocess optimization.A variety of crude oil has been blended by most oil devices for refining, thediversity of the raw material request that the multi-objective optimization of CDUshould be more adapted, Combined with the actual condition of a certain refinery, areduced harmonic model has been set up which will be used in the multi-objectiveoptimization. We have got the optimum mixing ratio of crude oil which will provide the basis for improving the economic benefit.