John D. Hottovy

Control of polymerization reactions

Abstract For the control of polymerization processes, fundamental models play a very important role. This is due to a number of reasons: the lack of available on-line sensors, the complexity of the polymerizatlon process, and the nunlhieai operating space of batch and semi-batch reactors. Most early research on polymerization reactor control centered around simulation and open-loop optimal control policies using these models. Recently, significant progress in modelling, sensor development, and nonlinear control methods has opened the way for closed-loop control of some fundamental polymer quality variables. This paper will highlight some of the most important past theoretical and practical developments, and will discuss some future directions for research.In a polymerization reactor (11), wherein particulate polymer product material is suspended in a diluent fluid thus forming a reaction slurry, and further wherein the slurry is circulated by a pump 31,33, the flow rate of the diluent fluid introduced into the reactor is manipulated so as to prevent the polymer solids from settling out of the circulating reaction slurry and plugging the reactor 11. This is accomplished by establishing control signals 51,57,59 which typify the flow rate of diluent fluid required to: (a) maintain a minimum velocity for the circulating reaction slurry, (b) maintain a maximum pressure head at a selected point in the reactor and (c) maintain a maximum power level supplied to the circulation pump. The one of signals (a), (b), and (c) above, which requires the greatest flow rate of diluent fluid, is automatically selected to manipulate diluent flow.