Distributionally Robust Scheduling of Hybrid Hydrogen-Battery Storage in Microgrids

Abstract

The integration of intermittent renewable energy sources into islanded microgrids necessitates robust and flexible energy management strategies. While Battery Energy Storage Systems (BESS) provide rapid response, their long-term viability is hindered by degradation. Hydrogen Energy Storage Systems (HESS) offer high-capacity, seasonal storage but suffer from lower round-trip efficiencies. This paper proposes a coordinated operational scheduling framework for a hybrid hydrogen-battery energy storage system (HBESS) under deep uncertainty. We formulate a two-stage data-driven distributionally robust optimization (DRO) model using the Wasserstein metric to construct an ambiguity set centered on the empirical distribution of renewable generation and load forecasting errors. To accurately capture system dynamics, we embed linearized battery degradation and electrolyzer-fuel cell thermodynamic constraints directly into the DRO framework. Leveraging strong duality, the infinite-dimensional min-max problem is recast into a computationally tractable mixed-integer linear program (MILP). Numerical simulations on a modified islanded test system demonstrate that the proposed Wasserstein-DRO framework reduces worst-case expected operational costs by up to 14.2% compared to traditional robust optimization, while strictly satisfying reliability requirements and minimizing battery degradation.