Guidelines for reliable predictions of green hydrogen costs by a novel stochastic optimization procedure

This paper presents an innovative stochastic optimization procedure to estimate the production costs of hydrogen and electricity in Hydrogen Supply Chains (HSCs) encompassing solar photovoltaic, electrolyzer technologies, with or without hydrogen storage. The procedure: 1) collects electrolyzer CAPEX and efficiency scenarios, 2) identifies the Probability Density Function (PDF) for CAPEX, 3) sizes HSC components that minimize investment and operational costs, 4) expands the CAPEX database by sampling new scenarios from the best-fitting PDF to predict the Levelized Cost of Hydrogen (LCOH) and Levelized Cost of Electricity (LCOE), and uses Kernel density estimation to provide LCOH and LCOE distributions. Results for Italian cities show that improving electrolyzer efficiency reduces LCOH up to 44%, while hydrogen storage reduces it by 2 €/kgH2 but increases LCOE by lowering grid electricity sales. The procedure generates PDFs of LCOH and LCOE for HSCs and electrolyzer technologies, enabling robust assessment of hydrogen investment strategies under uncertainty.