Giacomo Rubino, Sara Wallinger, Evelina Trutnevyte
Hydrogen could help abate emissions in hard-to-decarbonize sectors as well as aid electricity storage, but its future still faces deep uncertainties. Here, a high-resolution model of coupled hydrogen and electricity systems in Switzerland is used to analyze various scenarios for 2050 in terms of hydrogen demand, net import, and the uptake of electrolysers, fuel cells, hydrogen storage, trucks, and pipelines. Two uncertainty methods are compared: global sensitivity analysis that sets the accuracy benchmark and Plackett-Burman design that is much more computationally efficient. The analysis reveals a very large sensitivity of the hydrogen outcomes, where annualized capital costs of electrolysers, fuel cells, solar PV, batteries, hydrogen storage, and hydrogen pipelines, as well as domestic hydrogen demand are the most influential uncertainties. Different combinations of these uncertainties represent the hydrogen system either as a large-scale, internationally and domestically interconnected system with extensive transport and storage infrastructures, or as a more dispersed system with hydrogen production sites close to the demand with limited transport infrastructures. Plackett-Burman design identifies the key uncertainties and thus can serve as an uncertainty screening tool, if global sensitivity analysis has too high computational costs.