Dividing Wall Distillation Columns: Optimization and Control Properties

Abstract The optimal design of dividing wall columns is a non‐linear and multivariable problem, and the objective function used as optimization criterion is generally non‐convex with several local optimums. Considering this fact, in this paper, we studied the design of dividing wall columns using as a design tool, a multi‐objective genetic algorithm with restrictions, written in Matlab TM and using the process simulator Aspen Plus TM for the evaluation of the objective function. Numerical performance of this method has been tested in the design of columns with one or two dividing walls and with several mixtures to test the effect of the relative volatilities of the feed mixtures on energy consumption, second law efficiency, total annual cost, and theoretical control properties. In general, the numerical performance shows that this method appears to be robust and suitable for the design of sequences with dividing walls.