Scenario Based Assessment of Hydrogen Emissions from Road Transport Infrastructure

In this work, hydrogen demand, infrastructure development and emission pathways are quantified for the road transport sector at the European scale. Three scenarios are defined, a business-as-usual, a reference, and a hydrogen-favouring scenario, extending to 2070 and differentiated by assumptions on CO₂ prices, hydrogen prices and infrastructure availability. The resulting hydrogen emission inventories are intended as input to climate chemistry models in a subsequent step to assess the climate implications of road transport hydrogen emissions.

The analysis focuses on the fuel distribution and use stages, where hydrogen refuelling stations (HRS) emerge as the dominant emission source compared to vehicle-level losses (Clark et al. 2025). Emissions at HRS arise from storage venting and boil-off, tanker depressurisation during delivery, compressor leakages, hose venting and purging events, which scale non-linearly with station size and utilisation and depend strongly on station type (compressed or liquid hydrogen). This creates a trade-off between infrastructure scale, utilisation, economic performance and emissions which is an identified gap in literature.

Hydrogen demand needed to be supplied by HRSs in the scenarios is derived for passenger cars, light- and heavy-duty vehicles based on the temporal evolution of transport activity and drivetrain market shares competing with battery electric and conventional technologies. Passenger car activity is modelled using GDP and population dependent motorisation rates as Gompertz functions. Freight activity is estimated based on scenario calculations of the International Transport Forum (ITF). Total passenger and freight transport activity are modelled for reference scenario and is kept constant across the scenarios. Passenger car technology shares and stock evolution in Europe are simulated using the agent-based vehicle choice model VECTOR21 (www.vector21.de). Commercial vehicle stocks are computed using the LAREDO model, informed by expert surveys on hydrogen drivetrain penetration.

Component-level hydrogen emission rates for HRS are compiled from literature (Clark et al. (2025) and others) and distinguished between continuous and event-based releases. Due to the limited number of highly utilised operational stations, reported emission rates span wide ranges. HRSs are simulated by defining component structures, station size and utilisation frequency. To meet spatially resolved hydrogen demand, stations are located using freight and travel demand trip data and clustered to optimise utilisation and scale-dependent costs for compressed and liquid hydrogen supply. Given that liquid hydrogen supply contain high emission processes at lower utilization but are also more cost effective for larger scales, the works thus aims to presents an assessment of trade-offs between costs, utilisation and emissions for HRSs especially across the three scenarios.

Emission inventories are thus created at a 0.1° resolution with global-level analysis, as was done previously (Righi et al. 2025) by authors, and also for other species to provide a consistent input for subsequent climate impact assessments within the project CLEANLIEST.