Nonlinear operability thresholds in offshore green hydrogen production: Multi-objective system design under metocean uncertainty

Offshore green hydrogen systems depend on sustained wind–wave weather windows for marine transfer, creating nonlinear changes in cost, delivery, and reliability. This study combines 11 years of hourly ERA5 data from sites in the North Atlantic and North Sea with an hourly production-storage-vessel model and tri-objective optimization. Partial-transfer operation, threshold-definition sensitivity, electrolyzer alternatives, interannual variability, and optimization convergence are evaluated. NSGA-II and NSGA-III were each run with 100 individuals, 100 generations, and seven independent seeds. Results show a rapidly changing low-threshold regime followed by diminishing performance gains as transfer capability increases. The North Atlantic Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) design outperformed the North Sea design, achieving a lower cost of $8.53/kg, higher production of 1966 t/year, and a 99.0% transfer-success rate, compared with $11.78/kg, 1427 t/year, and 89.4% for the North Sea. Both compromise designs included approximately 50 t of storage. The framework supports site-specific offshore hydrogen production system design under metocean uncertainty.