Optimization of batch extractive distillation processes for separating close boiling and azeotropic mixtures

摘要

Operating features and limitations of Batch Extractive Distillation (BED) processes separating close boiling and azeotropic mixtures are studied using dynamic model and optimization techniques. A general Multiperiod Dynamic Optimization (MDO) problem is presented for the separation of binary feed mixtures and the recovery of solvent. Product specifications are imposed as constraints on the operation which allows quantitative assessment of the performance (operation time, productivity, profit, etc.) of BED processes and decomposes the MDO problem into a series of independent Single-period Dynamic Optimization (SDO) problems. Two modes of solvent feeding are discussed: the batch mode and the semi-continuous mode. Also two strategies for the initial feed charge to the reboiler are discussed under the semi-continuous mode of solvent feeding. They are full feed charge and fractional feed charge strategies. Time sequenced reflux ratio and solvent feed rate, although imposing a path constraint on the operation, improve the performance of BED significantly. Using a maximum of two time intervals for each distillation Task (to separate products), conditions have been developed to tackle this constraint efficiently. These considerably ease the solution of the optimization problem. All these features are illustrated using typical close boiling and azeotropic mixtures.