Hybrid Microgrid Sizing Using Particle Swarm Optimization with Use-Case Objective Functions: A Resilience-Focused Approach

While traditional microgrid optimization focuses primarily on minimizing Net Present Cost (NPC), real-world deployment success depends on multiple operational factors, including system resilience, capacity headroom, and battery health preservation. This paper presents an enhanced Particle Swarm Optimization (PSO) framework coupled with a Multi-Design Optimization (MDO) methodology that incorporates use-case objective functions that take into account several metrics beyond pure economics. Moreover, the enhanced PSO algorithm allowed us to evaluate near-optimal solutions that can help with plant design with constrained industrial choices. The proposed approach extends the established PSO-MDO method by introducing a customized cost function that simultaneously evaluates: (i) economic viability through NPC, (ii) system resilience through capacity factor and headroom utilization, and (iii) battery longevity through Depth of Discharge (DoD) and C-rate constraints. We apply this methodology to the Areza microgrid in Eritrea, a remote PV-BESS-diesel hybrid system that has been operational since 2019, with two primary objectives: first, quantifying design differences between optimization-derived and as-built configurations, and second, assessing system resilience under load growth scenarios. The results demonstrate that the objective function reveals critical trade-offs invisible to traditional NPC-only optimization, particularly regarding battery stress patterns and system capacity margins. The results reveal a broad near-optimal range where NPC changes by about 5%, but CAPEX varies by almost 28%. This indicates how very different combinations of components can achieve similar lifecycle costs. The best-performing design achieves an NPC ≈ EUR 2.02 M and an LCOE ≈ EUR 0.161/kWh, with sizing around PV 973 kW, battery 2236 kWh, inverter 392 kW, and generator 522 kW. The study highlights key trade-offs: reducing diesel use often leads to more curtailment, and systems that rely more on diesel are more sensitive to demand growth. Smaller systems face higher fuel costs when demand rises by 20%.