Model-based optimization for grade transition in industrial copolymerization process

Abstract

Grade transition operation is essential when several grades of polymers are produced in a continuous stirred-tank reactor via an industrial polymerization process. A dynamic optimization study has been conducted to shorten the continuous copolymerization transition time of SAN, which is a typical general-purpose resin. Model-based dynamic optimization is an effective method of controlling and optimizing chemical processes, especially during transitions in operation. The mathematical model developed considers initiation, propagation, chain transfer, and termination reaction of the copolymerization and was simplified using the pseudo-kinetic rate constant method. And it also includes a moment method designed to predict the polymer properties. Model validation was performed based on actual industrial process data and several constants that including the rate constants that indicate the effect of the chain transfer agent (CTA) were estimated. Model based dynamic optimization was performed. The objective function was defined in order to minimize the grade transition time by manipulating the polymerization temperature and the CTA feed flow rate. The effectiveness of the simulated results was determined by applying the operational trajectories of the two decision variables from the dynamic optimization solution was verified by comparing simulations that used actual operating conditions and the simulated results applying the step change operation. All simulations, parameter estimation and dynamic optimizations were performed using gPROMS process modeling tool.