Robust synthesis of the pressure-swing distillation process under azeotropic feed composition disturbance—Study of the tetrahydrofuran/methanol system

Abstract

Inferential control strategies on the basis of the temperature measurements in the separation process of the tetrahydrofuran/methanol azeotropic mixture using the pressure-swing distillation cannot guarantee the specification of the products. Accordingly, tetrahydrofuran mol% in the feed stream is assumed to be a Gaussian variable with known mean and standard deviation, and a stochastic mixed-integer nonlinear programming optimization framework is developed that is on the basis of the steady-state model of the plant. Proportional-integral controllers are implemented using the “Design Spec/Vary” utility in Aspen Plus and the optimization problem is formulated taking the set-points of the controllers together with the design parameters as decision variables. This leads to a closed-loop stochastic optimization problem in which unscented transform is used as the uncertainty propagation tool. The optimal solution shows more robustness against the imposed feed composition disturbances and handles them more effectively while the desired purity of the products can be maintained.