Electric vehicle infrastructure decision support system

Electric vehicles (EVs) need DC fast-charging stations (DCFC) for long-distance trips. DCFCs are costly investments and so charging station companies want to install them in locations where they expect high utilization. Further, government agencies are usually interested in ensuring that DCFCs are available on all roads and adequately spaced so that residents do not feel anxious about EV ownership. DCFC deployment therefore must balance the private and public objectives. This thesis presents a framework, ChargEVal, for simulating charging station deployment scenarios using agent-based modeling (ABM). The ABM utilizes behavioral models for simulating vehicle choice for the trip and charging choice during a trip. ChargEVal supports multiple users to submit multiple simulations simultaneously. ChargEVal also has a dedicated results viewer for viewing the simulation summary statistics and agent state values facilitating detailed insight and simulation comparison. Results from a few sample runs, model verification, and sensitivity analysis are shown. We also answer the question of whether it is more cost-effective to create a new charging station vs upgrading an existing station with more plugs. While the current implementation of ChargEVal is specific to the state of WA, USA; the underlying framework is generic enough to be applied to any geography at any scale.