A Tenth of California’s 2036 EV Fleet Would Match 9 GW of 12-Hour Storage, New Report Finds
- Analysis EVs Infrastructure & Grid Connection
Marin Holt
- 0
- 7 minutes read
- V2G capacity from 10% of EVs: 9 GW
- Energy at 12-hour duration: 108 GWh
- CEC light-duty EV projection for 2036: 9.7 million
- VPP share of California peak demand: more than 15%
- Annual savings, 2024 Brattle study: $550 million
GridLab, Kevala and E3 published a report on August 18 arguing that California should treat customer-owned equipment as grid capacity. The headline number: enrolling 10% of the 9.7 million light-duty electric vehicles the California Energy Commission projects for 2036 in vehicle-to-grid service would supply about 9 GW for 12 hours, or roughly 108 GWh. That is more than a third of the state’s 2036 long-duration storage procurement target. The report, titled “Unlocking California’s Flexible Load,” follows a year of stakeholder roundtables convened by the UC Davis Energy and Efficiency Institute.
The authors put wider virtual power plant potential at more than 15% of California peak demand. They cite a 2024 Brattle Group study for the affordability case, which found $550 million a year in savings from better-designed flexible load programs.
What the report asks California to change
Five recommendations run through the blueprint. Standardize program design so a customer in San Diego meets the same rules as one in Fresno. Pay for verified grid services rather than for enrollment. Open the same pathways to EVs, home batteries and smart commercial buildings. Fix interoperability with consistent frameworks and market signals. Hold compensation below the utility’s avoided cost, so non-participants do not subsidize the people who sign up.
Eric Cutter, a partner at E3, tied that last point to bills directly. Ed Randolph, who ran the California Public Utilities Commission’s Energy Division and now works at Caliber Strategies, argued that customers should not face different rules depending on where they live. Ric O’Connell, GridLab’s executive director, framed the resource as already bought and paid for. “The next generation of grid infrastructure is already sitting in our driveways, homes, and businesses,” he said.
Why It Matters
The 9 GW figure is an accreditation claim, not an engineering one. It holds only if the CPUC and CAISO count a parked vehicle the way they count a grid-scale battery, and today they do not. Aggregated distributed resources face different capacity accreditation and different telemetry requirements than a standalone four-hour battery, and 12-hour duration from a car assumes the car is plugged in, charged and under someone else’s control across the entire evening net peak.
That is where the report is most useful and most demanding. Paying only for verified performance is the right instinct, and it is also the expensive part, because verification at millions of endpoints needs metering and telemetry that most residential chargers do not have today. Developers and aggregators reading the 108 GWh number should read the accreditation recommendations first. Without them the capacity is a modeling result, not a resource the state records.
Critical Perspective
The report puts 9 GW of 12-hour capacity behind a tenth of the 9.7 million light-duty vehicles the CEC projects for 2036. CAISO’s pending change, the one that would let battery aggregators export down to zero net load, is sized by its own advocates at 2 GW, so the accreditation path that exists today reaches roughly a fifth of what this blueprint assumes. A GM-funded study published this month valued vehicle-to-grid participation at $680 to $2,750 per EV a year, which is the range that has to clear against a driver’s willingness to hand over the battery on a summer evening. What share of those 9.7 million drivers signs a contract that lets an aggregator discharge the car during the exact hours they want it charged?
Sources
- E3 (Energy and Environmental Economics)
- GridLab
- pv magazine USA
- Electrek
- PR Newswire (GridLab release)
