Scenario-Aware Scheduling in Microgrids: Look-Ahead Optimization including Frequency Security

Static, single-interval scheduling frequently produces myopic set-points for synthetic inertia and primary frequency support by undervaluing intertemporal constraints and ignoring uncertainty. For frequency-secure microgrid scheduling, dynamic multi-interval and two-stage stochastic-dynamic formulations are developed. Closed-form constraints on rate-ofchange-of-frequency (RoCoF), nadir, and steady-state deviation are enforced each period over a look-ahead horizon. Gridforming inverter resources and an energy storage system are co-optimized to supply virtual inertia and primary frequency support, alongside unit commitment, economic dispatch, and network constraints. Energy storage limits include instantaneous power and multi-period energy envelopes linked by state-ofcharge dynamics and ramping. The method is tested on a 40-bus benchmark system, where the dynamic scheduling framework lowers operating costs by about 3.9% compared with a frequencyaware static dispatch over the full simulation period. The stochastic dynamic version goes a step further by evaluating multiple uncertainty scenarios, making the scheduling decisions more adaptable and helping reduce overall dispatch costs.