Justine Sears

Abstract:
As the number of electric vehicles (EVs) increase we must consider not only how this fuel switch may affect electrical power infrastructure but also mobility. Specifically, the suitability and charging requirements of these vehicles may differ in rural areas, where the electrical grid may be less robust and miles driven higher. Although other studies have examined issues of regional power requirements of EVs, none have done so in conjunction with the spatial considerations of travel demand. We use three datasets to forecast the future spatial distribution of EVs, as well as these vehicles’ ability to meet current daily travel demand: the National Household Travel Survey (NHTS), geocoded Vermont vehicle fleet data, and an E911 geocoded dataset of every building statewide. We consider spatial patterns in daily travel and home-based tours to identify optimal EV charging locations, as well as any area-types that are unsuited for widespread electric vehicle adoption. We found that hybrid vehicles were more likely to be near other hybrids than conventional vehicles were. This suggestion of clustering of current hybrid vehicles, in both urban and rural areas, suggests that the distribution of future EVs may also cluster in rural areas. Our analysis suggests that between 69 and 84% of the state’s vehicles could be replaced by a 40-mile range EV, depending on the availability of workplace charging. Problematic areas for EV adoption may be suburban areas, where both residential density is high (and potential clustering of hybrids), as well as miles driven. Our results suggest EVs are viable for rural mobility demand but require special consideration for power supply and vehicle charging infrastructure.

Abstract:
The suitability and charging requirements of electric vehicles (EVs) may differ in rural areas, where the electrical grid may be less robust and daily VMT higher. Although other studies have examined issues of regional power requirements of EVs, none have done so in conjunction with the spatial considerations of travel demandand accessibility. We use three datasets to forecast the future spatial distribution of EVs, as well asto assessthese vehicles’ ability to meet current daily travel demand: the National Household Travel Survey (NHTS), geocoded Vermont vehicle fleet data, and an E911 geocoded dataset of every building statewide. We consider spatial patterns in existing daily traveland home-based tours to consider EV charging locations, as well as area-types that are unsuited for widespread electric vehicle adoption. We also consider how built environment attributes, including residential and commercial density and retail accessibility,affect travel demand and thus future EV energy requirements. We found that existing hybrid vehicles were more likely to be located nearother hybrids than conventional vehicles were. This clustering of current hybrid vehicles,in both urban and ruralareas, suggeststhat the distribution of future EVs may alsobeclustered. Our analysis suggests that between 69 and 84% of the state’s vehicles could be replaced by a 40-mile range EV, and 96-99% could be replaced by a 100-mile EV, depending on the availability of workplace charging.We did not find a strong relationship between land-use and travel demand, perhaps due to our low number of urban data points,the highly variable nature of rural travel, and the limitations of using a one-day travel log dataset. Our results suggest EVs are a viable option to serve existing traveldemand by rural residents but may require special consideration for power supply and vehicle charging infrastructure.