ABSTRACT

In designing a process for industrial realization, a large number of general and problem-specific design aspects are to be considered. General design aspects include, e.g., total annual cost, reliability, controllability, maintainability, resilience. In addition to these general requirements, the designer has to consider additional specific requirements, most of them are related to implementation, e.g., integration into Total Site, piping issues, safety considerations. Though more and more design aspects can be built into the mathematical model, no solvable mathematical model can cover all aspects of the designer. Consequently, generating the single process flowsheet as a solution to whatever model, this solution does not represent those design aspects that are omitted from the mathematical model. Thus, this single process can hardly be a solid base for the detailed design in the industrial implementation.
To overcome this difficulty, the proposed method in the present work generates all potentially optimal processes for further evaluation by the designer. For this purpose, a rigorous algorithm is proposed, based on a dedicated extension of the P-graph framework.

The list of all potentially optimal processes can be further evaluated by the designers by imposing additional design requirements to select the final design for industrial implementation. The known mathematical optimization based approaches do not provide this service for the design engineer, thus, the proposed approach increases the confidence to the designers by offering more than one solution to select best. The current work is illustrated by HEN synthesis.