Simultaneous Optimization of Hydrogen Network with Pressure Swing Absorption Based on Evolutionary Response Surface Method

The simultaneous optimization of complex process units and hydrogen networks is a significant challenge in refinery hydrogen network integration. To address this, an evolutionary response surface-based collaborative optimization method is proposed, enabling the concurrent optimization of pressure swing adsorption (PSA) and the hydrogen network. This method develops a mechanistic model for PSA and alternates between random sampling and evolutionary response surface-based hydrogen network optimization to obtain diverse sampling points and potential optimal solutions. The PSA mechanistic model is then used to compute the accurate output parameters for the sampled points, and these parameters are incorporated into the hydrogen network optimization to obtain precise objective function values. An efficient optimization framework is presented to streamline the process. The proposed method is applied to a refinery hydrogen network integration case study, comprehensively considering both PSA costs and hydrogen utility costs. The results demonstrate that the method is computationally efficient and effectively reduces the refinery's total annual costs. The accuracy of the optimization results is significantly improved compared to traditional methods, providing an effective solution for the collaborative optimization of the refinery hydrogen network and PSA.