SIMULTANEOUS DESIGN AND CONTROL OF POLYMERIZATION REACTORS Mariano Asteasuain a , Claudia Sarmoria b , Adriana Brandolin c and Alberto Bandoni d Planta Piloto de Ingeniería Química (UNS – CONICET), Camino La Carrindanga, km. 7, CC 717, Bahía Blanca, Argentina. Tel: (54)-291-4861700. Fax: (54)-291-4861600 a masteasuain@plapiqui.edu.ar c abrandolin@plapiqui.edu.ar b csarmoria@plapiqui.edu.ar d abandoni@plapiqui.edu.ar Abstract: Chemical engineering science has recognized the necessity of integrating process design and control; however, few steps have been taken in this direction in polymer science. In this work, a Mixed-Integer Dynamic Optimization approach is used for the simultaneous design and control of a styrene polymerization reactor. Our goal is to design the process and its control system in order to produce two polymer grades, which are defined in terms of the number average molecular weight (Mn). The process design involves reactor and initiator selection, and the two steady state operating points. The control system consists of a feedforward-feedback control scheme, which is designed to achieve optimal grade transition operation. The control system design includes optimal pairings between controlled and manipulated variables and controller’s tuning parameters for PI feedback controllers, and the best trajectories of the feedforward controllers. Keywords: Simultaneous design and control; grade transition optimization; polymerization reactor; mixed-integer dynamic optimization. INTRODUCTION Polymer synthesis and modification is a very important industry, producing nearly 200 million tons per year of a variety of polymers. Materials produced vary from high- volume resins sold at approximately one dollar per kilogram to high-priced specialty polymers at several thousand dollars per kilogram. During the past years, increased costs of energy, more stringent environmental regulations, and intense worldwide competition have motivated strong interest in optimizing plant designs and operating conditions. Polymers have many application areas, and each area needs different specifications. In order to fulfill these requirements, different grades of the same polymers are usually produced. Continuous plants commonly manufacture these grades in the same equipment by switching the operating points. This operation, called grade transition or grade changeover, may be performed rather frequently so as to satisfy the changing market demands. Therefore, the minimization of off-specification product during grade transition and the transition time is essential for a lucrative process. This is why, in addition to steady- state optimization, intense research has been performed regarding optimal grade transition operation of polymerization reactors. For instance, McAuley and McGregor (1992) developed optimal transition policies among three polyethylene grades in a gas-phase reactor. Using dynamic optimization they calculated the best profiles of the input variables.