Optimal Multi-microgrid Feedback Policies: Guaranteed Demand Fulfillment under Renewable Uncertainty

With increased penetration of Renewable Energy Sources (RES), the conventional distribution grid is advancing towards interconnected multi-microgrid (IMMG) systems supervised by a Distribution Network Operator (DNO). However, the inherent uncertainty of RES poses a challenge in meeting the power demand of critical infrastructures in the microgrids unless sufficient battery storage is maintained. Further, an emerging need is that the power output of the battery must be controlled in such a way that the system is minimally dependent on the main grid, thus safeguarding the system from grid-related contingencies. In this article, we propose a dynamic battery allocation and control strategy to optimize the power provided by the battery reserve while ensuring that critical demand is met with a provable guarantee. Our solution is built upon stochastic feedback control techniques. Under our proposed scheme, the DNO responds to the evolving uncertainty by dynamically balancing the RES and battery resources and eliminates the risk of over or underproduction. We derive battery power capability control strategy for multi-microgrid systems in two settings: when microgrids can (interconnected) and, cannot (individualized) share power amongst each other. We present examples under different scenarios with detailed comparison of performance of the proposed algorithm for individualized and shared settings. In particular, the advantages of interconnecting microgrids through savings in battery power requirements under IMMG over individualized microgrids are quantified. We demonstrate the utility and scalability of our algorithm by instantiating it to a modified IEEE-33 bus network with two microgrids, and a 225-bus network with 10 microgrids, on a real-time OPAL-RT based simulation platform.