Incorporating refuelling behaviour and drivers’ preferences in the design of alternative fuels infrastructure in a city

•We present a model to plan station roll-out in cities where OD data are not available.•It considers traffic covered and average distance of the inhabitants to stations.•Bunching of stations is prevented by a constraint that considers existing gas stations.•Modeling is informed by the results of a survey of more than 200 drivers.•This model can be applied to different cities, capturing refuelling tendencies. The purpose of this article is to present an optimization model to plan the deployment strategy for hydrogen refuelling stations in a city when Origin–Destination (OD) data are not available. This model considers two objectives: to maximize the traffic covered by the selected hydrogen refuelling stations and minimize the average distance of the city’s inhabitants to the nearest hydrogen refuelling station. As OD data are assumed to be unavailable, the clustering of stations in the highest traffic zones is prevented by a new constraint that takes into account information on the distribution of existing conventional refuelling stations. This model is applied to Seville, a city in Southern Spain of about 140km2 with a population of around 700,000. This application uses the results of a survey of more than 200 Sevillian drivers on their current refuelling tendencies, their willingness to use alternative fuel vehicles and their minimum requirements (regarding maximum distance to be travelled to refuel and number of stations in the city) when establishing a network of alternative refuelling stations.