A novel operational policy, the Process Intermediate Storage (PIS) operational policy, is introduced and used to synthesize, schedule and design multipurpose batch plants. The model is based on the State Sequence Network (SSN) and non-uniform discretization of the time horizon of interest model developed by Majozi&Zhu (2001). Two cases are studied to determine the effectiveness of the operational policy. A plant without dedicated intermediate storage is considered in the first case. In this case the throughput is maximized with and without the use of the PIS operational policy. The use of the PIS operational policy results in a 50% improvement in the throughout. The second case is used to determine the minimum amount of intermediate storage while maintaining the throughput achieved with infinite intermediate storage. This resulted in a 33% reduction in the amount of dedicated intermediate storage. The models developed for both cases are MILP models. A design model is then developed to exploit the attributes of the PIS operational policy. The design model is a MINLP due to the capital cost objective function. This model is applied to a literature example and an industrial case study and in both cases results in improved flowsheets and reduced capital cost.
Dissertation (MEng)--University of Pretoria, 2008.