Inferential model-based control of multicomponent batch distillation

Abstract

Industrial interest in batch distillation has increased significantly in recent years as more batch processing is being used for low-volume high-value specialty chemicals. Most studies have concentrated on the determination of optimum operating strategies: optimum trajectories of reflux ratio and pressure during the batch and processing of the slop cuts that are produced. Little attention has been paid to the problems of control. If instantaneous and perfect analyzers were available, the control of batch distillation would be straightforward once the optimum policies are available. Switching from total-reflux operation to producing product and from this to slop-cut production could be performed when compositions in the reflux drum or in product tankage reached the desired values. However, in most industrial applications of batch distillation, perfect composition measurements are not available. This means that an inferential control system that relies on temperature and flow measurements must be used. Models of the batch distillation process can be used to provide estimates of the required compositions. This paper presents the results of a study of two types of model-based inferential control systems for multicomponent batch distillation. The results showed that both steady-state and dynamic estimators provide good estimation of the distillate compositions using only one temperature measurement.