Design and Scheduling of Periodic Multipurpose Batch Plants under Uncertainty

Abstract

This work deals with the design of multipurpose batch plants under uncertainty. The model proposed by Pinto et al. [Comput. Chem. Eng. 2005, 29 (6), 1293−1303] for the detailed design of batch plants is extended to address the problem of uncertainty associated with production demand. Equipment choices, as well as plant topology and associated schedule, are defined simultaneously under an uncertain demand environment. Uncertainty is treated through a two-stage stochastic model leading to a mixed-integer linear programming (MILP) formulation, where profit is maximized and the expected value of perfect information (EVPI) is determined; the latter is estimated as the difference between the values of profit obtained for the wait-and-see and the here-and-now models. A scenario is set up where demand is represented by a discrete probability function and a cyclic operation considered, together with mixed storage policies and sharing of resources. Some illustrative examples are solved to show the model applicability and the EVPI are determined and analyzed.