Additional contributors: Gabriella Broga, Kirill Tchernyshyov, Athindra Venkatraman

One of the biggest challenges posed by rapid growth in EV adoption today is ensuring our power grid can support this new demand. Today’s grid is not equipped to meet the power required by the anticipated increase in EV charging, at least not in all places. RMI (and many others) have emphasized better forecasting of EV charging needs as a key tool to help utilities and regulators plan for future demand through load management and proactive investment in grid infrastructure.

The key question in planning for EVs is pinpointing when and where charging will be necessary. To help answer these questions, RMI developed GridUp, an open-access tool that identifies where EV load growth is likely to materialize. The tool’s strength lies in its use of data from millions of trips made by today’s (primarily gas-powered) vehicles to model where tomorrow’s EVs are likely to charge, assuming minimal changes to driving behavior. In a recent case study on New York State, we used GridUp to illustrate the magnitude of the problem and present potential solutions to address it. This analysis focused on two New York utilities at the forefront of developing improved EV planning strategies, Consolidated Edison (Con Ed) and National Grid, and highlighted areas in their service territories most likely to require distribution system upgrades — and soon.

When and where will all this EV charging take place?

GridUp’s load forecasts for National Grid’s service territory highlight the magnitude and speed of future EV charging demand (Exhibit 1). We analyzed both an unmanaged charging scenario and a managed charging scenario. The unmanaged scenario assumes that drivers start charging as soon as they arrive at their destination and use the maximum available charging power until their battery is fully charged or they finish their stop. The managed charging scenario assumes that, at locations with extended stop times, drivers take advantage of the entire duration to charge, requiring less power output.

Exhibit 1. Estimated EV Load, National Grid New York Service Territory — Unmanaged Charging Scenario

RMI Graphic. Source: RMI GridUp

This map clearly shows that EV charging needs will be concentrated in certain areas, including parts of Erie, Onondaga, Albany, and Saratoga counties. These growing load “pockets” — especially those with significant fleet activity, such as warehouse districts, transit centers, and ports — often necessitate significant localized grid upgrades, such as adding or upsizing distribution substations and feeders.

Can the distribution system support these loads?

This raises the next question: How will this EV load growth impact the equipment powering the distribution system? Leveraging publicly available data from National Grid, we compared the estimated EV charging load and location data with available grid infrastructure assets, such as substation banks, to calculate the portion of currently available distribution capacity that would be taken up by EV load within the next decade (Exhibit 2).

Exhibit 2: Estimated EV Charging Share of System Peak Load by Substation Bank, National Grid New York Service Territory – Unmanaged Charging Scenario

Source: National Grid NY System Data Portal; RMI GridUp.

This analysis suggests that, by 2030, 7 substation banks out of the 779 within National Grid’s service territory will experience EV load growth at 50 percent or more of their currently available capacity. By 2035, 14 substation banks may be overloaded by EV loads taking up 100 percent of their existing headroom — the extra capacity at a substation which can accommodate additional demand without requiring upgrades. Four of the substations may be overloaded by more than 200 percent, and 16 others may witness EV load growth utilizing 50 percent or more of their current headroom.

This analysis does not account for planned system upgrades that would help meet some of this demand; however, it also does not include non-EV load growth, such as from building electrification, which will be significant. This accelerating growth in electricity demand underscores the need to plan early to build a resilient system that can support it.

Successful load management can meaningfully reduce EV contribution to peak load

The outputs above are based on GridUp’s unmanaged charging scenario. Exhibit 3 shows an illustrative example of the unmanaged and managed charging load shapes from GridUp for a sample census tract within National Grid’s territory. The EV contribution to the system peak (assuming the system peak begins at 5 p.m.) is reduced by approximately 33 percent across the entire service territory in the managed charging scenario — a significant decrease. While this reduction is seen across the board, the magnitude of the effect will vary across different geographies and for the selected peak hour (i.e., local distribution network peak versus total system peak).

Exhibit 3: Illustrative Unmanaged versus Managed Load Shapes for National Grid, NY – 2035

RMI Graphic. Source: RMI GridUp

Comparing the unmanaged and managed charging scenarios’ effect on National Grid’s distribution assets highlights the large potential value of load management. Exhibit 4 illustrates the relative benefits of the managed charging scenario, based on the share of available infrastructure capacity utilized by EV load in 2035. The difference in the two scenarios indicates how effective load management can lower levels of distribution system headroom required to accommodate new EV load growth, potentially deferring or reducing the need for new infrastructure investments.

Exhibit 4: Substation Bank Headroom Allocated to EV Load

Source: RMI Analysis

Achieving 100 percent customer participation in managed charging programs is likely unrealistic. Still, the comparison between entirely unmanaged scenarios with largely managed ones highlights the potential benefits of improved charging load optimization. Furthermore, advanced forms of managed charging, beyond those included above, could provide even greater benefits for distribution infrastructure planning, investments, and system operations. This can and should include both traditional programmatic approaches to managed charging (e.g., time-of-use rates and incentives) and alternative demand-side load-shifting methods, such as various non-wires alternatives and flexible service connections.

EV load forecast data can support proactive planning and demand management for resilient electric service

EV load is expected to grow at a rapid pace in many parts of the United States, with some locations requiring significantly expanded distribution capacity to support it. This suggests the need for long-term, proactive planning for upgrades and investments in the distribution system to ensure reliable and uninterrupted electric service, especially when combined with other forms of load growth. Furthermore, this analysis underscores the key role that various demand-side load management strategies can play in reducing the total amount of new distribution system capacity that will be required.

Access to bottoms-up, localized EV load forecasts can help inform these strategies by right-sizing distribution system upgrades as well as supporting the development of cost-effective operational and programmatic approaches to load management. By balancing the urgent need for certain grid investments with a focus on identifying least-cost opportunities to meet demand, we can create a resilient and reliable electricity system that both provides necessary services and protects affordability for all utility customers.