Monitoring and control of polyester production process

A real-time monitoring and control strategy is presented for a continuous polyester plant. This process consists of three reaction steps, namely, esterification, prepolymerization, and finishing polymerization. This paper is focused on the prepolymerization reactor. A high-fidelity dynamic model that is efficient and accurate for real-time monitoring and control is developed. As this study is focused on real-time monitoring and control applications, the model development, configuration, and implementation are performed using a real-time dynamic simulation software, SimSuite Pro/sup TM/, running in tandem with the plant distributed control system, a D/3 DCS/sup TM/. By using this model, a multirate control strategy is developed for polyester molecular weight and carboxyl acid concentration control. In addition to the fast measurements, i.e. temperature and pressure, this controller also utilizes slow measurements, i.e. molecular weight and acid concentration. In the absence of slow measurements, the estimates from the dynamic model are used.