Global optimization of batch and steady-state recycling chromatography based on the equilibrium model

Abstract

A bi-level approach is presented for the optimal design of batch and steady-state recycling (SSR) chromatography. On the first level, an efficient formulation as mixed-integer problem (MIP) is applied to obtain optimal fractionation times for chromatograms simulated by the equilibrium model. On the second level, the optima are evaluated against a parametrized objective function. The combination of model and optimization method provides a computationally very efficient tool. A comprehensive example study demonstrates that batch chromatography achieves the highest productivities at the cost of limited yields, which is optimal only for vanishing feed costs. In other instances, the more flexible SSR concept provides larger profits. The results reveal also new optimal operating policies for SSR processes with segmented product fractions and waste streams.