Design and optimization of hydrogen refueling station powered by hybrid energy system

The rapid transition toward hydrogen-based energy systems necessitates the development of optimized hydrogen refueling station (HRS) configurations that balance economic feasibility, energy security, and sustainability. This study presents a comprehensive sensitivity analysis on the impact of weather conditions on optimized HRS configurations in Makkah, Saudi Arabia, comparing advanced metaheuristic optimization techniques—gray wolf optimizer, constrained particle swarm optimization (CPSO), and flower pollination algorithm (FPA)—against the conventional hybrid optimization of multiple energy resources professional (HOMER Pro) software. The findings reveal that these advanced techniques outperform HOMER Pro in managing renewable intermittency, reducing net present cost (NPC), and achieving a lower levelized cost of hydrogen (LCOH). Under an optimistic scenario (+10% solar and wind), the system achieves a 98% reliability rate and the lowest LCOH of $0.320/kg, reinforcing its economic viability. However, under a pessimistic scenario (−10% solar and wind), NPC increases to $17.30 × 103, reliability drops to 90%, and LCOH rises to $0.340/kg, underscoring the necessity for robust storage solutions and adaptive energy management. The study confirms CPSO's ability to maintain cost-reliability equilibrium, while FPA achieves the lowest LCOH through hybridized resource utilization. Conversely, HOMER Pro exhibits suboptimal performance in fluctuating climatic conditions. These findings align with Saudi Arabia's Vision 2030 goals, advocating artificial intelligence (AI)-integrated real-time optimization frameworks for next-generation hydrogen infrastructures. Future research should focus on hybridizing AI-driven predictive models with metaheuristic techniques to enhance system resilience and scalability in hydrogen economies.