Probabilistic Assessment of Community-Scale Vehicle Electrification Using GPS-Based Vehicle Mobility Data: A Case Study in Qatar

Abstract

To avoid the operational consequence of thermal rating exceedance and the financial consequence of excessive reinforcement, the impact of domestic charging of electric vehicles (EVs) on power distribution networks must be accurately assessed prior to accepting vehicle electrification at the community-scale. Although driven by routine, charging behaviour patterns are also influenced by geography, meteorological conditions and season, hence will have a localised element to them that could reduce the diversity of charging load profiles. To model this uncertainty, this article develops a probabilistic methodology to quantify EV home charging demands based on vehicle mobility data and underlying trip characteristics. Models articulate the departure time distribution using a mixture of von Mises distributions, and incorporate non-negative conditional distributions of trip durations, distances and parking durations, which in turn generalise localised charging behaviours. The resulting load profiles are used to drive a community electric network model based on a distribution feeder in Qatar, a country with high per km energy consumption, to quantify impact scenarios of high temperature and driving habit in terms of voltage and thermal stability. Results indicate that overnight domestic charging is sufficient to support daily trips and local networks are capable of hosting high EV penetration despite peaks.