Identifying Patterns of Real-World Charging Frequency for a Sample of Plug-in Hybrid Electric Vehicles in North America

Plug-in hybrid electric vehicles (PHEVs) have the capability to effectively utilize electricity from the grid as an energy source for powering an appreciable portion of the total vehicle miles travelled (VMT), thereby reducing greenhouse gas (GHG) emissions, since the Carbon Intensity (CI) of electricity is often less than that of liquid fuels in many parts of the world. Several real-world usage factors can affect the fraction of VMT electrified, with the frequency of charging being one of the most influential factors. Studies in recent years have attempted to characterize the real-world performance of PHEVs based on long-term average fuel consumption and/or other data flags in the readout from vehicle On-Board Diagnostics (OBD), but such approaches are unable to infer accurate estimates for the occurrence of charging events. This paper adopts an approach that relies on analysis of highly granular (trip by trip) information obtained from vehicles equipped with a data communication module (DCM) to infer the occurrence of charging events from change in the battery state of charge (SoC) between trips. Analysis of data obtained from a large sample of PHEVs (one full calendar year for hundreds of vehicles) in the US and Canada reveals three distinct patterns: i) vehicles that are consistently charged, ii) vehicles that are consistently not charged, and iii) vehicles with temporally varying frequency of charging. Unlike some other studies about PHEVs in other parts of the world, results of our sample for PHEVs in North America show that the majority are consistently charged, but with various frequency levels that are regionally dependent.