A switch method framework for process superstructure optimization

Process synthesis has an enduring impact on the economic performance of a chemical process. This study demonstrates a Switch Method Framework (SMF) for superstructure-based process synthesis using an isomerization process for gasoline production as case study. SMF is distinguished from other superstructure representations by modular superstructure elements integrated through switches and distributors, where each element contains unit options selected using methodologies from hierarchical decomposition. As such, SMF combines the best of hierarchical decomposition with the best of superstructure synthesis, blending practical representation of process route combinations with robust exploration of the optionality within the superstructure. SMF uses a genetic algorithm to both integrate unit options between superstructure elements and optimise the operating conditions towards maximal techno-economic return. Applying SMF to the case study, the combination of a Pt/ SO4-ZrO2 catalyst and Simulated Moving Bed (SMB) yielded a maximum Net Present Value (NPV) of $220 million and gasoline research octane number of 95. Compared to a benchmark analogous to the Penex-Molex process, employing a Pt/Al2O3-CCl4 catalyst and SMB, SMF increased the NPV by 18%. Given the improved techno-economics and product quality, SMF shows promise as a hybrid method combining decades of process engineering experience and heuristics with global optimization of the process flowsheet.