California VPP Test Dispatches 535 MW From 100,000 Home Batteries
- C&I Storage Energy Storage Grid Services & Demand Response News
Marin Holt
- 0
- 7 minutes read
- July 29, 2025 test dispatched 535 MW average from 100,000+ residential batteries over two hours from 7 PM to 9 PM Pacific
- Sunrun aggregated two-thirds of participating batteries; Tesla aggregated one-third
- Brattle Group analysis found output was relatively consistent with no significant fluctuations or performance attrition
- Test was the largest virtual power plant dispatch in California history and possibly the world
- Sunrun compensates participating battery owners up to $150 per battery per dispatching season
On July 29, 2025, California’s three largest utilities dispatched 535 MW of electricity from more than 100,000 residential batteries in a coordinated two-hour virtual power plant test. The event, running from 7 PM to 9 PM, produced enough power to supply more than half of San Francisco during evening peak demand.
How the Test Worked
The California Energy Commission, CAISO, and all three investor-owned utilities (PG&E, SCE, and SDG&E) coordinated the dispatch through the state’s Demand-Side Grid Support (DSGS) program. Sunrun and Tesla served as the primary aggregators, enrolling homeowners with rooftop solar and battery systems into a single dispatchable fleet.
Sunrun delivered more than 360 MW during the window, accounting for over two-thirds of total output. The company had already demonstrated 325 MW of dispatch capacity in an earlier event on June 24, 2025. Tesla batteries made up the remainder through its Powerwall virtual power plant network.
Why This Matters for Grid Planning
The Brattle Group, an independent research firm, analyzed the test and found “dependable, planning-grade performance at scale” with consistent output and no major fluctuations or attrition throughout the two-hour window. That finding is significant: it means grid operators now have evidence that residential batteries perform reliably enough to substitute for traditional peaking generation in capacity planning models.
California’s DSGS program has grown to 500 MW of enrolled capacity as of October 2024, with the July 2025 test exceeding that figure. Homeowners receive up to $150 per battery per season for participation. At scale, these payments cost a fraction of building new natural gas peaker plants, which run $1,000 to $1,500 per kW of installed capacity.
National Implications
Texas is building parallel infrastructure through ERCOT’s Aggregate Distributed Energy Resource (ADER) pilot, which expanded to 160 MW of capacity in Phase 3 during 2025. Seven commercial ADERs now participate, up from three at mid-year. The program moved from its initial 80 MW cap after regulators saw residential battery aggregation reduce strain during summer peaks.
PJM, the largest U.S. grid operator, pushed full Order 2222 implementation (which enables distributed resource aggregation in wholesale markets) to February 2028. Illinois responded by launching utility-aggregator VPPs in spring 2026 under the 2025 CRGA Act to address PJM’s high capacity prices directly.
Implementation Realities
The California test proved that 100,000 batteries work as a single resource. The next challenge is permanence. DSGS funding depends on annual legislative authorization, and a 2025 bill to extend the program through 2028 stalled in committee. Without stable funding, aggregators lose the financial signal to keep enrolling customers.
Hardware is not the bottleneck. More than 1.5 million homes in California already have rooftop solar, and battery attachment rates continue climbing. The constraint is regulatory: turning a successful pilot into a permanent, bankable grid resource that utilities and grid operators count on year after year.
Source: Electrek
Critical Analysis
Simultaneous dispatch of 100,000+ residential inverter-based DERs at 535 MW creates a distributed harmonic injection event across California distribution feeders. A 535 MW synchronized dispatch from 100,000 distributed inverters across three IOU territories creates aggregate active power ramps that CAISO must balance in real-time; Brattle Group confirmed planning-grade performance with no major fluctuations during the two-hour window, validating VPP reliability.
5-Year Projection
By 2031, operational data from facilities like this will become the standard requirement for securing interconnection agreements, as ISOs prioritize proven Virtual Power Plant profiles.
Critical Perspective
The article highlights a 535 MW dispatch from 100,000 home batteries. Texas’s ERCOT ADER pilot is comparable but lacks the scale and duration of California’s test. The 2011 Fukushima Daiichi nuclear disaster demonstrated the catastrophic failure of centralized energy infrastructure. Will this distributed model truly eliminate the need for robust, centralized generation when extreme weather events strain the grid?
