The role of EV fast charging in the urban context: An agent-based model approach

Abstract

Using an agent-based simulation approach, this paper investigates the role of fast-charging infrastructure in urban environments. The simulation model tracks the spatial and temporal behaviours of electric vehicles (EVs), facilitating a comprehensive analysis of the deployment of charging infrastructure. Notably, the model incorporates non-parametric queuing dynamics, information-sharing regarding waiting times, and diverse agent characteristics, deepening insights into the subject matter. Drawing on a large-scale implementation in the municipalities of Frederiksberg and Copenhagen, the study advocates for adopting fast chargers by demonstrating several key points. Firstly, information-sharing significantly reduces waiting times, particularly within the fast-charging network, with potential reductions of up to 30% during peak demand periods. Secondly, larger fast-charging clusters comprising 10–14 outlets outperform smaller clusters, primarily due to reduced waiting times and enhanced prediction accuracy of waiting times, which is a consequence of the information-sharing. Thirdly, placement strategies based on unserved demand metrics yield superior outcomes than those solely driven by observed demand patterns. By effectively monitoring both observed and unmet demand, these strategies tend to better optimize charging infrastructure placement. These insights, which emerge from the sophisticated and heterogeneous nature of the simulation framework, highlight the value of information and unserved demand in this field.