How can renewable hydrogen compete with diesel in public transport? Robust design optimization of a hydrogen refueling station under techno-economic and environmental uncertainty

Document Type

Journal Article

Publication Date


Subject Area

place - europe, mode - bus, infrastructure - vehicle, infrastructure - fleet management, technology - alternative fuels, technology - emissions, planning - methods


Renewable hydrogen, Carbon intensity, Hydrogen-powered buses, Robust design optimization, Techno-economic and environmental uncertainty


Heavy-duty transport represents nearly 6% of the greenhouse gas emissions in Europe. Renewable hydrogen is a potential option to decarbonize heavy-duty transport, such as buses. Renewable hydrogen for buses promises excellent environmental performance, at the expense of a higher fuel cost, as opposed to a diesel-powered bus fleet. Despite the inherent uncertainty, feasibility studies in this framework generally assume deterministic techno-economic and environmental parameters, which can lead to a suboptimal performance that is sensitive to the random environment. To provide robust design alternatives, we applied robust design optimization on a wind- and solar-powered hydrogen refueling system and a hydrogen- and diesel-powered bus fleet, to optimize the Levelized Cost Of Driving (LCOD) and Carbon Intensity (CI), subject to technical, economic and environmental uncertainties. A fully diesel-powered bus fleet achieves the optimized LCOD mean of 1.24 €/km, but it results in the worst LCOD standard deviation (0.11 €/km), CI mean (1.33 kg˙CO˙2,eq /km) and CI standard deviation (0.075 kg˙CO˙2,eq /km) among the optimized designs. To reduce the LCOD standard deviation, CI mean and CI standard deviation, part of the diesel-powered bus fleet is converted into hydrogen-powered buses and the renewable-powered hydrogen refueling station is scaled accordingly. Converting 54 % of the diesel-powered bus fleet into hydrogen-powered buses results in a decrease in LCOD standard deviation by 36 %, a decrease in CI mean by 46 % and a decrease in CI standard deviation by 51 %, at the expense of an increase in LCOD mean by only 11 %. Future work will focus on the integration of full-electric buses.


Permission to publish the abstract has been given by Elsevier, copyright remains with them.


Applied Energy Home Page: