A simplified approach to calculating personalized estimates for electric vehicle charging delays

Abstract

In the past decade, electric vehicles (EVs) have gained traction as a cleaner alternative to internal combustion engine vehicles, commonly referred to as gas-powered vehicles. To promote EV adoption, the government has implemented various regulations and incentives to support the transition to cleaner transportation. However, EV adoption in the United States has progressed more slowly than expected, with EVs accounting for less than 10 percent of new vehicle sales in 2023. Recent surveys indicate that a significant barrier is the perceived inconvenience and uncertainty surrounding EV charging, particularly the additional time required to charge during active use, which we call charging delay. Currently, there exist some models for estimating these charging delays, but these models require users to input a significant amount of information, such as their daily driving schedules, locations of charging stations, and exact distances of trips taken each year, which many users may not even remember. These more complex models are likely to overwhelm users, especially those who may be entirely new to EVs. To fill this gap, this thesis introduces a simplified model for estimating personalized annual EV charging delay using a set of easy-to-provide inputs, including typical driving behavior and access to home and work charging. The model logic captures delay from both routine usage, such as weekly driving patterns or typical trips, and occasional, high-energy long-distance trips, which, while not routine, are still important to account for. For weekly trips, the model considers four scenarios based on combinations of home and work charging access to determine driving and charging schedules. For long-distance travel, the model uses data from the 2022 National Household Travel Survey (NHTS) and performs multiple iterations of bootstrap resampling to create synthetic distributions of long-distance trips within a year. Data related to individual routine vehicle usage and charging delay is unavailable, so we are unable to validate the model’s performance through accuracy calculations. Instead, we performed a one-at-a-time sensitivity analysis to better understand how charging delay is affected by different factors. We found that access to private charging, such as home or work charging, improves charging delay robustness for regular weekly trips, with the exception that relying solely on work charging on workdays can cause stepwise increases in non-workday delays. Additionally, long-distance trip delays are no affected by private charging access and follow a stepwise pattern based on vehicle range. In general, the simplified approach presented in this thesis offers a more accessible way for current and prospective EV owners to clearly understand their own expected experience of EV ownership.