Mixed integer linear programming-based optimal capacity configuration of ALK/PEM mixed water electrolysis system

The integration of renewable energy sources in hydrogen production through water electrolysis is pivotal for advancing decarbonization in the energy and chemical industries. However, inappropriate sizing often leads to poor techno-economics of hydrogen production and low rates of renewable energy consumption. This paper proposes a mixed water electrolysis system utilizing both alkaline electrolyzers (ALK) and proton exchange membrane electrolyzers (PEM), coupled with battery storage, to optimize techno-economic performance. A non-convex mixed integer quadratic constraint programming (MIQCP) model is developed to minimize the levelized cost of hydrogen (LCOH) while addressing the complexities of system configuration and scheduling. The model is transformed into a mixed integer linear programming (MILP) problem using segmented linearization and the Big-M method. A K-means clustering algorithm is also used to determine the typical day for resolving the variable explosion problem induced by the sign constraints. A case based on a hydrogen production park proves the effectiveness of the proposed system. Results show that the system achieves an optimal ALK/PEM ratio of 112.5:29, yielding an LCOH of 20.69 ¥/kg, which is 5.78 % and 10.66 % lower than the ALK-only and PEM-only systems, respectively. The internal rate of return for the mixed system is 5.02 %, and power consumption is 56.92 kWh/kg. Sensitivity analyses highlight the impact of PEM electrolyzer cost reductions on system efficiency, underscoring the economic potential of hybrid electrolyzer configurations. The study aims at providing insights into the optimization of renewable energy-driven hydrogen production systems, facilitating better integration and cost-efficiency in green hydrogen projects.