Decomposed algorithmic synthesis of reactor-separation-recycle systems

Abstract

A two-level algorithmic approach to synthesis of reaction-separation-recycle systems is presented. On the upper level, reactor conversion and purge fraction are optimized using a simple direct search method, and the flows among subsystems such as the reactor system and separation system are calculated. On the lower level, individual subsystems are optimized in every iteration of the upper level loop. On this level, the structure and operating conditions of the subsystems are determined, and the cost is returned to the upper level. This decomposed approach makes the synthesis of the overall system less complex and more robust than approaches addressing the entire problem simultaneously. The proposed approach is demonstrated on the well-known HDA process. The results are comparable to previously published results using the Hierarchical Decomposition Procedure, and the optimal results are demonstrated to be insensitive to initial estimates.