Evolutionary Algorithm Based Multiobjective Optimization of Vapor Recompressed Batch Extractive Distillation: Assessing Economic Potential and Environmental Impact

Despite generating enormous interest, the full landscape of process intensification and multiobjective optimization (MOO) remains unexplored in the area of batch extractive distillation (BED). This paper is focused on articulating a mixed-integer nonlinear MOO problem to optimize both conventional and vapor recompressed BED, considering both existing and new plant scenarios. Process optimization of BED columns typically involves several technical, economic, and environmental objectives that are conflicting in nature, which lead to many equally good optimal solutions from the perspective of the given objectives. Here, the objective functions considered for BED optimization are minimization of total annual cost and CO2 emissions along with maximization of total annual production of two components that form an azeotropic mixture. To perform this optimization, a unique MOO strategy is adopted, which comprises of three steps. The first step deals with the selection of significant decision variables through Taguchi analysis followed by the formulation of the optimization problem. In the second step, BED is optimized with the help of an elitist nondominated sorting genetic algorithm. The third step comprises of a selection of a solution from the Pareto-optimal front using the TOPSIS (Technique for Order of Preference by Similarity to Ideal Solution) with entropy information for weighting of objectives. The performance of this MOO strategy for BED optimization without and with vapor recompression is illustrated for the separation of acetone-methanol mixture with water as the homogeneous entrainer. Finally, a comprehensive analysis is performed to assess the superiority of vapor recompressed BED over conventional BED.