(99c) Multicomponent Batch Distillation Optimal Control and Thermodynamic Uncertainties

Batch distillation is commonly used unit operation in specialty chemicals.  In order to maximize the profitability and improve separation using batch distillation column, optimal reflux operation can be used.  The mathematical and numerical complexities of the optimal control problem get worse when uncertainty is present in the formulation.  However, in specialty chemical industries thermodynamic properties of not all chemical components of a mixture to be separated are known.  These uncertainties in the problem formulation can result in sub-optimal solutions.   These static thermodynamic uncertainties result in dynamic uncertainties due to unsteady state nature of batch distillation operation.  In this work, we derive new optimality conditions based on Ito’s lemma derived from financial literature where time dependent uncertainties are common.    We use stochastic maximum principle for formulation of the resulting stochastic optimal control problem.  This algorithm is implemented in the MultiBatchDS batch distillation process simulator. Finally, a numerical case-study derived from a real world batch distillation separation problem encountered at Firmenich is presented to show the scope and application of the proposed approach.