On the Need for Revisions of Utility Factor Curves for Plug-In Hybrids in the US

Abstract

<div class="section abstract">Plug-in hybrid electric vehicles (PHEVs) have the capability to drive an appreciable fraction of their miles travelled on electric power from the grid, similar to battery-only electric vehicles (BEVs). However, unlike BEVs which cannot drive unless charged, PHEVs can automatically switch to gasoline power and operate similar to a regular (non-plug-in) hybrid electric vehicle (HEV). Though operating similar to HEV is already beneficial in terms of fuel economy, greenhouse gas (GHG) emissions and criteria pollutants compared to conventional internal combustion engine (ICE) vehicles, much of the attractiveness and allure of PHEVs comes from their capability to drive “almost like a BEV”, but without range anxiety about running out of battery charge. The concept of “utility factor” (UF) has been developed as a simple metric for gauging the fraction of total annual distance travelled by a PHEV in charge depletion (CD) mode, in-which electric power from the battery is the primary source of propulsion power for the PHEV. Different standards in different parts of the world have been put in place for UF curves, which are essentially a way of estimating the expected UF of a PHEV as function of its electric driving range. It is important however to keep in mind that UF curves are only as good as how valid the assumptions and data that were used to construct the curves. Unlike the situation in Europe, where several studies and real-world data have challenged the European UF curves, in the US, SAE J2841 standard seemed to hold well compared to real-world public datasets for PHEVs. However, a recent study in 2022 introduced an analysis of two new datasets (Fuelly and BAR) and came to a conclusion that prior datasets for US PHEVs were less relevant. That study seemed to have quickly caught wind with the US Environmental Protection Agency (EPA), who in their notice of proposed rule-making (NPRM) of April 2023 have proposed to reduce the UF curves for corporate average fuel economy (CAFE) compliance, citing a need for an update to SAE J2841 standard, which was last updated in 2010. In this work, we re-analyze the two recent datasets (Fuelly and BAR) cited in the 2022 study. Our findings suggest that for Fuelly dataset, a major portion of the discrepancy between real-world performance of US PHEVs and SAE J2841 could be attributed to a combination of modeling and data cleaning errors. When correct, the results of Fuelly dataset seems to reasonably confirm with SAE J2841 standard, within acceptable margin of error typical to differences between EPA label ratings and the real world for other powertrains besides PHEVs. For the BAR dataset on the other hand, we show how limitations of how the data was collected leads to significant bias, making it unrepresentative of PHEVs in the US at large. While there may be motivation to update the standard for UF curves to keep them closer to real-world performance, it is important to keep in mind the primary categories of reasons for deviations from the standard, which includes deviation from assumed: i) Daily mileage profile, ii) Charging behavior, and iii) actual attained electric range. While analysis of public trave survey data suggests no significant change in daily mileage profile, updating the standard to account for present-day and future-expected charging behaviors and attained electric range in the US requires much more thorough study than what has been conducted to date. </div>