Two-Stage Distributionally Robust Scheduling of Port Integrated Energy Systems with Downstream Hydrogen Demand

Abstract

To address wind and photovoltaic uncertainty in port integrated energy systems, this paper develops a two-stage distributionally robust scheduling model that coordinates downstream hydrogen supply and port logistics. The proposed electricity-heat-hydrogen model integrates the main grid, renewable generation, battery energy storage, power-to-hydrogen, hydrogen storage, fuel cells, and combined heat and power units. Berth allocation and quay-crane assignment are jointly optimized to capture logistics-induced energy demand. A Wasserstein ambiguity set is constructed from historical wind-PV forecast errors to represent distributional uncertainty. Day-ahead scheduling and intra-day redispatch are coordinated within a two-stage framework, which is reformulated as a tractable mixed-integer linear program using affine decision rules and strong duality. Case studies show that the proposed method increases the total cost by 6.21% relative to stochastic optimization but reduces it by 45.79% compared with conventional robust optimization, thereby balancing economy and robustness. Sensitivity analysis further shows that increasing downstream hydrogen demand raises system cost and external hydrogen procurement, while coordinated operation of energy devices supports renewable-energy shifting, hydrogen buffering, and waste-heat recovery.