OPTIMAL DESIGN METHODOLOGY FOR DIVIDING WALL COLUMNS

The dividing wall column, DWC, is a thermally coupled sequence that can perform the separation of fluid mixtures with reduced energy consumption, but also decreasing the capital costs associated. Several authors have developed design procedures for this kind of sequence, using short methodologies or optimization strategies. The use of short procedures is easy and fast, but, in the most of the cases, the resulting design is not optimal. On the other hand, the use of optimization procedures allow obtaining optimal designs, but time and complexity required are elevated. In this work, we propose an optimal and short design methodology for DWC. First, we use a multiobjective genetic algorithm with constraints, coupled to Aspen Plus, to obtain Pareto fronts of DWC for different mixtures, feed compositions and recoveries. The analysis of the Pareto fronts brings useful information about the flows and composition of the interconnection flows, key variables in this methodology, which are incorporated in a short method that generates optimal designs with less computational resources.