5 Conclusions
An analytic process for estimating national PEV non-residential charging requirements within communities and along Interstate corridors has been presented. Scenario analysis was conducted to illustrate EVSE requirements for a range of potential PEV markets. The analysis makes no assumptions regarding which PEV market scenarios are more or less likely. Rather, a range of plausible PEV markets with unique features is developed to explore the relationship between the evolution of the PEV fleet and charging infrastructure requirements.
5.1 Major Conclusions
To facilitate understanding, this report separates PEV charging infrastructure requirements by area served (cities, towns, rural areas, and Interstate corridors) and role during the PEV market growth trajectory (providing coverage to early PEVs versus satisfying demand due to high PEV penetration).
Cities are expected to have the greatest charging infrastructure requirements under both the coverage and demand assessments. About 8,000 DCFC stations would be required to provide a minimum level of coverage nationwide in cities and towns (based on uniform station spacing assuming BEVs are never more than 3 miles from a charging station). Such a network would provide consumer support for long-distance intra-city travel, serve as a safety net for emergency charging situations, and dampen range anxiety concerns.
Demand analysis of community charging demonstrates how utilization of the DCFC coverage network would be expected to grow in a high PEV penetration market. Modeled results for a 15-million PEV market estimate a DCFC plug requirement of 25,000 in U.S. communities (approximately 3.1 plugs per average DCFC station and 3.4 plugs required to support 1,000 BEVs under a home-dominant charging assumption). Demand for nonresidential L2 EVSE (including work and public charging) is estimated at 600,000 plugs necessary to support 15 million PEVs (approximately 40 plugs per 1,000 PEVs).
Sensitivity analysis of the community results for consumer charging demand indicates a strong relationship between the evolution of the PEV and EVSE markets. As this analysis attempts to arrive at charging infrastructure solutions that fill the eVMT gaps between consumer travel patterns and PEV electric ranges, infrastructure requirements are not only proportional to the total number of PEVs in the system, but also inversely proportional to the electric range characteristics of these PEVs. Manufacturer and consumer preferences with respect to electric range, charging power, and utilization of residential EVSE have direct and dramatic consequences on the level of charging demand calculated in this study.
Results suggest that relatively few corridor DCFC stations could enable long-distance BEV travel between U.S. cities, where vehicles are concentrated. Under most scenarios, the number of required stations is similar to the number of corridor DCFC stations already established by Tesla or the number planned by Electrify America within the next two years.
Understanding driving patterns and vehicle characteristics and then prioritizing corridors and setting station spacing accordingly—as illustrated in the network scenarios—could help optimize the utility and economics of early-market corridor-coverage stations. The analysis identifies the majority of consumer long-distance automobile travel as being regional rather than truly cross-country, which emphasizes the importance of multistate DCFC corridor planning (such as the West Coast Electric Highway, corridor planning across Colorado/Utah/Nevada, and the I-95 Fast-Charge ARC led by Nissan and EVgo).
Despite the relatively low number of corridor DCFC stations estimated by this analysis, establishing the financial viability of these stations will be difficult, particularly in the face of low initial utilization and high capital/operating costs. Requirements for the average DCFC complex necessary to support peak traffic volume for 7 million BEVs is estimated at six plugs per station (assuming 70-mile station spacing). Given a national BEV stock of approximately 250,000 through the end of 2016, this implies that the average corridor DCFC station should expect relatively low levels of utilization as the PEV market continues to mature. Utilization expectations for corridor DCFC stations are further tempered when considering the current segmentation of DCFC protocols (Tesla, CHAdeMO, and SAE CCS).
Regardless of geographic scope, organizations planning for charging infrastructure to support consumer adoption of PEVs should be aware of the importance of consumer preferences with respect to electric range and charging behavior. Furthermore, planners are encouraged to focus efforts on providing consumers with adequate charging coverage (particularly DCFC supporting adoption of BEVs) with the expectation to monitor station utilization and grow charging capacity (both in terms of rated power and number of plugs) as the PEV market continues to grow over time.
5.2 Summary of Modeling Limitations
One of the fundamental assumptions of this study is that consumers will attempt to operate their PEVs in the future as they have operated their conventional gasoline vehicles in the past. This assumption places the burden on PEVs that they are able to serve as 1-to-1 replacements for gasoline vehicles in a given consumer’s household fleet of vehicles. In the real world, it is uncertain if consumers will hold PEVs up to this requirement. For instance, National Household Travel Survey (NHTS) data suggest that over 80% of consumer vehicles are owned by multi-vehicle households. Such ownership circumstances may result in corridor charging demand below what is estimated in this analysis as the household gasoline vehicle could be perceived as the more convenient option for long distance travel (based on refueling time, infrastructure availability, or attributes unrelated to driving range).
Similarly, the baseline travel data used to calibrate EVSE estimates in this analysis assume consistent personal mobility patterns out to 2030. In reality, the world of personal mobility is poised to undergo a paradigm shift as the sophistication and adoption of automated driving technology continues to grow. Interactions between evolving mobility patterns and refueling infrastructure supporting advanced technology vehicles are currently being investigated by the consortium of national laboratories participating in the DOE’s SMART Mobility Initiative (DOE 2017c).
The EVI-Pro model used in this analysis assumes charging infrastructure must be sufficient to enable any consumer to maximize eVMT in any PEV. In reality, some degree of consumer self-selection in the new and used PEV markets is likely to reduce the need for non-residential charging as households right-size PEV purchases to meet the daily driving needs of their individual household. While the extent to which consumers are able to successfully right-size PEV purchases is largely unknown, its effect would reduce infrastructure requirements relative to estimates made in this analysis.
EVI-Pro’s fundamental objective of maximizing consumer eVMT enforces no minimum utilization criteria on individual charging stations, likely resulting in a percentage of stations with insufficient revenue potential. Incremental eVMT benefits and utilization of individual stations have been explored in regional simulation studies (Wood et al. 2015a, 2015b, 2017). Detailed financial analysis of the national EVSE networks explored in this study remains an ongoing area of research.
This study is intentionally vague with respect to the percentage of PEVs adopted by residents of MUDs. Inconsistent access to home charging for residents of MUDs is often cited as an infrastructure barrier to increased PEV adoption. Yet even at 15 million PEVs nationally (5% of LDV stock), this analysis is well below the threshold where MUD residents would be required to participate in the PEV market. As such, no distinction is made for MUD residents in this analysis, but sensitivities are explored for portions of PEV owners who adopt non-home-dominant charging behaviors.