Self-optimising Control, a Batch Distillation Simulation Study

A dynamic mathematical model of batch distillation columns is formulated using four basic assumptions: binary separation, negligible vapour holdup, constant pressure and constant molar flows. Simulations performed in the modelling tool MATLAB, proved that the model gives satisfactory description of the process behaviour. Simulations studies were then used to apply the theory of self-optimising control to batch distillation columns, in order to provide a systematic procedure for the selection of controlled variables based on operational economics. It was found that the distillate and boilup flows have good self-optimising properties. The study has also shown the unsuitability of the reflux ratio (Rin) and reflux return (LT) to self-optimising control due to their sensitivity to disturbances in batch distillation of the system.