Modelling and Experimental Studies of a Continuous Emulsion Polymerization Reactor

Abstract

<p>In continuous emulsion polymerization that commercially important phenomena of sustained oscillations and large start-up transient conversion overshoots have been observed by several investigators. Experimental studies have shown that the conversion, number of polymer particles and all other related properties often oscillate with time. In other words a steady state is never achieved. These oscillations can lead to emulsifier levels too small to adequately cover polymer particles with the result that excessive agglomeration and reactor fouling can occur. Furthermore, excursions to high polymer concentrations, due to the cycling behavior of conversion can result in excessive branching and poor polymer processability. It is evident that this reactor behaviour is highly undesirable since it results in a continuously varying product quality.</p> <p>In this thesis, the complex physical and chemical phenomena occurring in a continuous emulsion polymerization reactor are studied both experimentally and mathematically. Dynamic models suitable for prediction and on-line control studies are developed based on particle age distribution. On-line techniques are developed to measure some states of the control model and to monitor the reactor performance. These techniques are based on turbidity measurements and liquid exclusion chromatography. A multivariable stochastic control algorithm is developed to control the pathological behaviour of continuous emulsion polymerization reactors and ensure a high productivity and product quality.</p>