The transition toward sustainable, low-carbon energy systems has amplified the need for advanced energy management frameworks that optimize the integration of renewable energy sources (RES), energy storage systems, and modern DC-based loads. Hybrid AC/DC microgrids represent a promising architecture to address these challenges by reducing conversion losses, enhancing grid resilience, and enabling cost-effective renewable integration. This paper presents an extended optimization framework for hybrid AC/DC microgrids, combining simulation-based analysis with real-world validation. A Mixed-Integer Linear Programming (MILP) model is introduced, dynamically allocating AC and DC loads while accounting for converter efficiencies, battery state-of-charge (SOC) constraints, and energy balance equations. The framework was first validated through simulations of the FOSS nanogrid, achieving an optimal balance of 34% AC and 66% DC loads, which reduced grid imports by approximately 20% compared to a full AC configuration. To further demonstrate practical applicability, real-world data from five residential homes in Ireland participating in the StoreNet project were analyzed. Results confirmed that optimized hybrid configurations consistently reduced energy imports compared to both full AC and full DC setups, with optimal DC load proportions ranging between 72% and 84%. These findings underline the scalability and robustness of the proposed methodology, bridging the gap between theoretical models and practical deployments. The proposed framework offers a replicable solution for improving efficiency, reliability, and sustainability in future hybrid microgrid applications. • A scalable MILP-based optimization framework is proposed for hybrid AC/DC microgrids. • Dynamic AC/DC load allocation explicitly accounts for converter losses and battery SOC constraints. • Laboratory validation on the FOSS nanogrid shows up to 20% reduction in grid imports. • Real residential data from five StoreNet homes confirm optimal DC shares between 72% and 84%. • Combined simulation and field validation demonstrates robustness and practical deployability.
Charalambous et al. (Fri,) studied this question.