A two-stage robust optimization strategy for microgrids considering the grid-forming capability of energy storage and uncertainties of wind, photovoltaic, and load

The integration of grid-forming (GFM) energy storage can enhance frequency stability in microgrids with high penetration of grid-following (GFL) renewable energy sources. However, the uncertainties associated with wind, photovoltaic, and load pose challenges for the configuration and inertia time constant tuning of GFM energy storage. Therefore, this paper first analyzes the impact of power disturbances on frequency characteristics and formulates constraints for system inertia and reserved power. Then, an uncertainty set is constructed by incorporating typical renewable energy scenario probabilities and load forecast errors. On this basis, a two-stage robust optimization (RO) model is proposed. In the first stage, the rated power, capacity, and operational status of GFM energy storage are determined. In the second stage, the worst-case scenario is identified, and the corresponding optimal scheduling scheme and inertia time constant are derived. The solution algorithm is developed by combining the Karush-Kuhn-Tucker conditions with the Column-and-Constraint Generation method. Based on actual data from a specific region, the proposed strategy yields an energy storage with 97.1 kW rated power and 124.2 kWh capacity, and the optimized inertia time constant can meet frequency requirements. Compared with deterministic optimization, the rated power increases by 68.5 %, while the capacity is reduced by 59.1 %. Compared with the conventional RO strategy, the proposed strategy lowers the daily construction cost of energy storage by 25.4 %. Furthermore, the effects of the initial state of charge on the proposed RO strategy are investigated, and the robustness of the derived configuration scheme is demonstrated under 4 actual scenarios.