Abstract
The economic viability of hydrogen fuel-cell trucks (FCETs) depends critically on the levelised cost of hydrogen supply (LCOSH), which falls sharply with station utilization rate. Yet the Alternative Fuels Infrastructure Regulation (AFIR) mandates one refuelling station every 200 km along the TEN-T Core network irrespective of local demand density, risking systematic under-utilisation and inflated hydrogen costs in low-demand corridors. This paper develops a Mixed-Integer Programme (MIP) on the Spanish road network that jointly minimises station investment, hydrogen procurement, and operator detour costs, explicitly capturing real network routing, non-linear economies of scale, and the operating cost of truck operators (OPEX). The model yields optimal networks ranging from 15 large, high-throughput stations at low NFOC to 76 smaller, dispersed ones at high OPEX, with mean LCOSH spanning 8–12 EUR/kg accordingly. Comparing the unconstrained optimum against a greedy AFIR-compliant network (32 stations) and a hybrid scenario that fixes urban TEN-T nodes while optimising the remainder, we find that the pure distance rule imposes a social-cost premium of 5–25% depending on operator running costs, whereas the hybrid scenario consistently recovers most of this loss, holding the premium to 5–12%. Looking ahead, the ongoing transition toward vehicle automation is expected to reduce non-fuel running costs substantially, shifting the efficient network toward fewer, larger, and more concentrated stations, which reinforces the case for demand-driven siting over uniform distance rules. Supplementing AFIR with a minimum-throughput criterion and demand-aggregation instruments would align infrastructure deployment with both current economic realities and the trajectory of an increasingly automated freight sector.
Too many or too few? Evaluating AFIR’s distance rule against demand-driven hydrogen refuelling networks
