An automatic distillation sequence synthesis framework based on a preorder traversal algorithm

Abstract

Distillation energy consumption dominates the process industry; hence, the selection of the distillation sequence will substantially affect the separation energy consumption. We propose a framework for the automatic synthesis and optimization of distillation sequences by integrating the Aspen Plus with the MATLAB programming platform. The framework combines the concept of binary trees in data structures and uses a preorder traversal algorithm to generate a network superstructure in the simulator. In the next step, it automatically formulates and solves a mixed-integer linear programming model based on calculation results; finally, the distillation model is optimized using an improved quadratic interpolation algorithm. Two case studies for C5 alkane and dimethyl carbonate separation showed that the optimal solution obtained by the framework reduced the total annual costs by 13.18 % and 2.88 %, respectively, compared with the processes without systematic optimization. The framework also allowed screening out promising alternatives by rapid evaluation of different process routes.