Economic and flexible chemical process design. A biobjective optimization problem.

摘要

This work focused on biobjective optimization which includes both costs and flexibility in process design. A design is termed flexible if the design guarantees a feasible region of operation for any bounded realization of the process parameters(input variables). The feasible region of operation is determined by the set of conditions and restrictions imposed on the chemical process under study. An index of design flexibility was developed based on a topology related to the region of realization of the input variables.;To prove whether a feasible region of operation exists or not, a new method using the Lebesgue measure of the feasible region of operation was developed. The new method determines the critical input variables via a sensitivity analysis of the Lebesgue measure of the feasible region of operation. This method of evaluating flexibility seems to be more effective that others that have been proposed in the literature since the method do not need to enumerate vertices nor sets of active constraints. Furthermore, the method determines the critical input variables even if they are not vertices of the polyhedron that represents the region of realization of the input variables.;Two algorithms were developed. One gives the optimal design for a given flexibility index and the other determines the flexibility index for a specific design. Three chemical process example problems are solved to show the application of the new method. One of the results of this work was that the concept of a flexible index in design is only possible if, from an economical point of view, the total cost is more sensitive to changes in the capital (equipment) cost than changes in the operation cost.