A Generalized Methodology for Optimal Configuratons of Hybrid Distillation-pervaporation Processes

The aim of this work is to illustrate the structural and parametric optimization of continuous hybrid distillation-pervaporation process with different configurations such as series, parallel, and series-parallel arrangement of pervaporation modules in the pervaporation network, and to propose a generalized methodology for difficult separations. A superstructure representation of hybrid process is considered and the process is modeled and optimized using an MINLP approach. The optimization strategy is to obtain the desired degree of either the retentate or the distillate purity without violating the composition constraints of products and heat exchange policy which minimizes the required membrane area by increasing the flux through the membrane. The structural and operating parameters such as number of trays required, feed tray location, reflux ratio, retentate recycle location, permeate recycle location, membrane feed location, number of pervaporation modules required, target composition (which is directly related to membrane area), and membrane selectivity are optimized for each configuration by minimizing the total annual cost (TAC) for the separation system. The optimization studies have been carried out with General Algebraic Modeling System software (GAMS/SBB/CONOPT) and the results of different configurations have been compared on the basis of TAC required for the separation. A total of three industrial case studies have been dealt with. The separation of isopropanol-water as an azeotropic mixture, propylene-propane as a close boiling mixture (system with a low relative volatility) and acetone-water as a tangent pinch mixture have been studied as representative examples.