Two-level optimization of an existing SMB for p-xylene separation

Abstract

A two-level optimization procedure based on the concept of “separation volume” and equivalent true moving bed modeling strategy is proposed. The mathematical model considers axial dispersion, internal mass transfer described by linear driving force (LDF) model and multicomponent Langmuir isotherm. The procedure is used to optimize an existing simulated moving bed (SMB) unit for p-xylene separation from mixed xylenes. The optimization procedure is of sequential nature and considers single objective function in each level. In the first level, the SMB unit productivity is maximized for given flow rates in Zones 1 and 4. In the second level, the desorbent consumption is minimized in view of the results from the first level of the optimization procedure. The required extract purity and recovery are imposed as optimization problem constraints in both levels. The global solution of the optimization procedure gives the optimal operating conditions leading to maximum SMB unit productivity with a minimum possible desorbent consumption for attainment of that SMB productivity. Three cases with different extract purity and recovery constraints were addressed.