Control And Optimization Of An Extractive Distillation Process For Separating The Mixture Of Acetonitrile-n-propanol

Process intensification,energy conservation and emission reduction represent a dominant trend in the chemical process engineering,mostly due to increasing awareness of the limited energy resources of the modern society.There are acetonitrile and n-propanol mixed liquid in the production in fine chemical industry and pharmaceutical industry.The goal of this paper is to optimize the extractive distillation process and design reasonable control scheme for the process system in order to get high-purity acetonitrile(99.8%)and n-propanol(99.8%)and save energy.The control scheme could ensure the smooth operation in the whole production process.The mass fraction of acetonitrile and n-propanol in feed stream are 75% and 25%,respectively.And the 1-Methyl-2-pyrrolidinone(NMP)is selected as extractive solvent.The steady-state of two-column extractive process is implemented in Aspen Plus.The reboiler duty and purity of products are as the objective functions.Aspen plus software and generalized regression neural network(GRNN)prediction model are used to optimize the operation parameters of this two-column sequence,respectively.These parameters are theoretical trays,feed location,solvent ratio,reflux ratio and feed temperature.Comparing the results of these two optimized methods,it shows that the average relative error of GRNN prediction model is less than 4%.So GRNN can predict optimization results quickly and accurately.The temperature sensitive trays are obtained by analyzing slope criterion.The operating variables and the controlled variables pairing are selected and designed in the control loop of the extractive distillation process.The results show that cascade control for temperature and component of extractive column can assure the purity of products.The average temperature control loop of solvent recovery column can resist disturbances effectively.The reflux flow and feed flow proportional control loops with time lag elements can reduce the overshoots of controlled variables.In a short time,those control loops can achieve the specified product purity smooth operation and reduce unnecessary energy consumption for feed flowrate and feed composition disturbances.The heat integration process of extractive distillation with heat exchange between cold and hot streams could save energy about 34.9%.The sensitive trays are controlled by changing the steam flowrate in the whole process.The effects of open-loop dynamic response and controllability analysis shows stronger correlation between the process variables and the process is non-linear,asymmetric response and long time lag.The pre-heater is sensitive to feed flowrate and the pressure of solvent recovery vacuum column is sensitive to composition disturbance.So temperature-flowrate cascade control loop and temperature difference control loop are used to solve these problems above.The performance of this control strategy is estimated by adding feed flowrate and composition disturbances.Comparing the results,this optimum control strategy is more sensitive to feed composition disturbance than feed flowrate disturbance,especially in run-back conditions.They can run a new steady-state in the short time and the mass fractions of acetonitrile and n-propanol products are in accordance with 99.8%.In total,the performance of the control strategy is good and the control strategy is feasible and can be used in practice.