Process Synthesis Using Commercial Tools: Feedback from R&D to Industry for Electric Production Applications

Systematic process synthesis methods and tools have been developed for decades together with the advance in computing, with profusive academic activities but with less reports from industry on the use of these approaches.

This study intends providing feedback on superstructure multi-objective optimization experiments conducted together by the EDF R&D and engineering divisions, trying to shift from research approach towards use in real engineering projects.

To do so, the methodology is based on the coupling between a commercial existing process simulation software and an external solver. The process simulator (ProSimPlus) is where the synthesis problem and the exploration space are defined, leading to the definition of a superstructure. The external solver (e.g., MIDACO for multi-objective optimization in this study) is used to solve the underlying mathematical problem, seeking for optimal structural alternatives (flowsheet) and associated degrees of freedom (design and operating parameters) within the given constraints.

So far, studied cases cover various types of industrial processes (e.g., power cycles, separation processes) and stages of industrial projects (technology benchmark, conceptual study, assisting of process design). Experiences will first be presented in the design phases of an industrial project (with examples of power cycle design, such as innovative supercritical CO₂ cycles). Other examples cover usages for technology benchmarks, using multi-objective optimization to provide insights to the process designer on the physical possibilities of each process (with example for the comparison of membrane technologies for the water purification).

Current perspectives can be drawn from these first results, in relation to ongoing challenges in the energy sector with a new landscape for mature and innovative technologies, and to new prospective tools such as the rise of generative approaches for process design.