Abstract: The integrated design and optimization of hydrogen networks is a significant pathway for refineries to achieve energy conservation and carbon emission reduction. However, the long-term operation of hydrogen networks is subject to various uncertain factors such as external market changes, crude feedstock property fluctuations, and operating condition variations, etc. A novel two-stage stochastic programming modeling method based on flexibile design was used to establish a mathematical programming model, which could integrate optimization strategies to deal with the uncertainties in the production process. Then, the mixed-integer nonlinear programming model with a large number of bilinear terms was constructed, a tailored-made global optimization algorithm based on multiparametric disaggregation technique was proposed to accelerate convergence, and a flexible design and optimization case of hydrogen network in a refinery was presented to verify the universality of the model and the efficiency of the algorithm. The results showed that the integrated optimization model could carry out flexible recovery and cascading matching of hydrogen resources among complex hydrogen network structures, and multi-scenario operation strategies could increase the operational flexibility and maximize the economic benefits of the hydrogen network under uncertain environments.

Key words: hydrogen network, superstructure modeling, stochastic programming, flexibility analysis, global optimization, multiparametric disaggregation technique