Multiobjective optimization of multipurpose batch plants using superequipment class concept

Abstract

We present a novel approach for solving different design problems related to single products in multipurpose batch plants: the selection of one production line out of several available, additional investment into an existing line or plant, and grass-root design of a new plant. Multiple objectives are considered in these design problems. Pareto-optimal solutions are generated by means of a Tabu Search algorithm. In the novel approach the concept of superequipment has been defined as an abstract equipment model, which is capable of performing any physico-chemical batch operation. Each superequipment is transformed into a real equipment unit, for example a reactor, during or after the optimization in order to evaluate performance parameters of a design. This novel concept uses an implicit definition of a superstructure and essentially optimizes on the transfers between different equipment units in a design. On the basis of case studies we demonstrate that the application of the superequipment concept offers a number of advantages for the investigated design problems. For example, in the evaluation of investment into single equipment pieces to be added to existing plants or production lines only the maximum number of additional equipment, each represented as a superequipment, has to be specified instead of a list consisting of a higher number of explicit units. Similar advantages arise for grass-root design problems or for the selection of a production line or plant out of several that are available for the production of a specified chemical. The comparison with optimization results obtained with a conventional Tabu Search algorithm revealed that the superequipment approach is capable of identifying the Pareto-optimal solutions in significantly reduced computation time.