Techno-Economic Optimization of Hydrogen-Integrated Hybrid Microgrids for Rural Electrification Using the Hippopotamus Optimization Algorithm

This study develops a techno-economic sizing and energy-management framework based on the Hippopotamus Optimization Algorithm (HOA) for hydrogen-integrated autonomous Hybrid Renewable Energy Systems (HRES) for rural electrification. A representative remote community in Tabuk, Saudi Arabia, is investigated using 11 years of NASA POWER satellite-derived meteorological data. The modelled community is constructed from a synthesised connected-load inventory representing approximately 600 households and 3000 residents; accordingly, the results represent a simulation-based planning case study rather than a validated design for a specific settlement. Seven configurations combining photovoltaic generation, wind turbines, battery storage, hydrogen production and storage, fuel cells, and diesel generation are evaluated considering Total Net Present Cost, CO2 emissions, and Loss of Power Supply Probability (LPSP), with a Demand Response Management System (DRMS) incorporated into the framework. The three objectives are combined using a weighted-sum scalar formulation, complemented by a hard-constrained formulation for reliability. HOA is benchmarked against Particle Swarm Optimization (PSO), Grey Wolf Optimizer (GWO), Grasshopper Optimization Algorithm (GOA), Walrus Optimizer (WO), and Osprey Optimization Algorithm (OOA) under a common budget of 10,000 objective-function evaluations per run and 10 independent runs. Under the constrained formulation, six of the seven configurations satisfy LPSP ≤ 5% within the investigated sizing bounds, with costs of energy (COE) ranging from $0.1046/kWh for PV/wind/battery to $0.1357/kWh for wind/battery/diesel; the fully renewable PV/wind/hydrogen configuration is feasible at $0.1195/kWh. Only the wind-free configuration fails to satisfy both imposed constraints because of the 30% diesel-energy limit rather than reliability. Evaluation over the eleven individual meteorological years shows that all designs violate the 5% reliability criterion in every year, reaching 2.0–2.9 times the design LPSP because hour-of-year averaging removes prolonged low-resource periods. Re-optimization against the worst observed year increases COE by 33–63% and storage capacity by factors of three to five, with hydrogen storage in the fully renewable configuration increasing from 10 to 75.4 kg. Sensitivity analysis identifies wind availability as the dominant economic parameter, with a 20% wind-speed reduction increasing COE by 36.1%. The DRMS reduces the peak-to-average ratio by 20.0% for the assumed evening-peaking profile, whereas no reduction is obtained for an afternoon-peaking profile consistent with measured Saudi residential demand. These findings demonstrate that meteorological and demand-profile representation materially affects autonomous HRES sizing and should be explicitly considered when interpreting techno-economic optimization results.