Degradation-aware electro-thermal risk-averse optimization of a grid-connected hydrogen-based multi-energy system with PEM and SOEC electrolyzers

This paper proposes a degradation-aware electro-thermal risk-averse optimization framework for day-ahead scheduling of a hydrogen-based multi-energy system integrating electricity, hydrogen, and thermal energy using Proton Exchange Membrane (PEM) and Solid Oxide Electrolyzer Cell (SOEC) electrolyzers. A mixed-integer linear programming formulation captures PEM overload operation, SOEC thermal inertia, startup/shutdown behavior, heat losses, and waste heat recovery. Temperature-variation penalties suppress aggressive thermal transients and reduce degradation-related stress. Uncertainties in electricity prices, renewable generation, and multi-energy demands are addressed using scenario-based Conditional Value-at-Risk (CVaR) optimization. A robustness metric determines the maximum uncertainty horizon tolerated under deterministic operation. Results show that SOEC systems achieve lower operating costs and emissions due to higher efficiency, whereas PEM systems provide greater flexibility and lower cost sensitivity to risk-averse scheduling. SOEC systems exhibit greater uncertainty absorption capability, while PEM systems remain more adaptable to uncertainty. The results highlight the trade-off between efficiency, flexibility, thermal constraints, and uncertainty absorption.