Nonlinear Programming Superstructure for Optimal Dynamic Operations of Simulated Moving Bed Processes

Abstract

Over the past decade, many operating modifications have been proposed for simulated moving bed (SMB) processes, including three-zone, VARICOL, and PowerFeed. Nevertheless, it remains a challenge to find the most efficient operating scheme for a particular application. In this study, we generalize the standard SMB operation by considering a superstructure optimization problem with time-variant flow rates. The optimization problem, constrained by a partial differential algebraic equations (PDAEs) model, is fully discretized both in temporal and spatial domains. The resulting nonlinear programming (NLP) problem is handled by IPOPT 3.0, an interior-point solver. In case studies of linear and bi-Langmuir isotherms, our approach successfully finds the optimal operating scheme and parameters. Moreover, through the formulation and solution of a multi-objective optimization problem, we confirm that our approach has the potential to find more advantageous operating schemes than the standard SMB or PowerFeed operating schemes. Finally, we show that these multi-objective solutions can be related to a class of profit maximization problems for SMBs.