Sunrun will aggregate residential solar and battery capacity through Voltus to sell directly into Google’s power procurement programs, creating the first large-scale pathway for distributed home energy to serve AI data center loads. The deal bypasses multi-year utility interconnection queues by treating hundreds of thousands of behind-the-meter batteries as a single dispatchable resource that hyperscalers can count toward their 24/7 carbon-free energy targets. This marks a structural shift: data center operators are no longer waiting for transmission build-out but are actively contracting distributed capacity that already exists on the distribution grid.
How Bring Your Own Capacity Programs Rewire Hyperscaler Procurement
Bring Your Own Capacity (BYOC) programs let large electricity consumers – primarily hyperscalers – procure capacity from non-utility resources and have it recognized by grid operators and regulators as counting toward their reliability obligations. Until now, these programs have mostly involved on-site generation, industrial demand response, or utility-scale storage contracted through bilateral agreements. The Sunrun-Voltus arrangement extends the model to aggregated residential assets: Sunrun’s fleet of roughly 1 million customers, of which several hundred thousand have battery storage, becomes a virtual power plant that Voltus operates and offers into wholesale capacity markets and directly to corporate buyers.
Voltus, which manages over 6 gigawatts of flexible load and distributed generation across North America, acts as the technical and commercial intermediary. It handles telemetry, dispatch optimization, market participation, and settlement – the “platform” layer that turns thousands of disparate home batteries into a single, creditworthy capacity product. For Google, the appeal is speed and granularity: residential batteries can respond in seconds, are already interconnected at the distribution level, and can be dispatched to match the hour-by-hour carbon intensity of the grid, supporting the company’s 24/7 carbon-free energy matching goal by 2030.
The mechanics rely on FERC Order 2222, which requires regional grid operators to create market participation models for aggregated distributed energy resources. While implementation varies by ISO – PJM, CAISO, and ISO-NE are furthest along – the rule establishes the legal basis for a VPP like Sunrun’s to bid capacity into wholesale markets on equal footing with a gas peaker or utility-scale battery. Voltus has been active in PJM’s capacity market for years; this deal extends that participation into direct corporate procurement, where the hyperscaler effectively becomes the off-taker for the VPP’s capacity value.
Why Distributed Storage Now Competes with Utility-Scale for Hyperscaler Dollars
The economics of this deal hinge on three converging pressures. First, utility-scale interconnection queues have swollen to over 2.6 terawatts nationally, with median wait times exceeding four years for projects entering queues in 2023. A hyperscaler needing 100 megawatts of firm capacity in northern Virginia or the Bay Area in 2025 cannot wait for a new solar-plus-storage plant to clear the queue. Second, transmission build-out is lagging generation additions; even projects that clear interconnection often face costly network upgrades that delay commercial operation by additional years. Third, hyperscalers face mounting pressure – from boards, investors, and regulators – to demonstrate additionality and hourly carbon matching, not just annual renewable energy certificate (REC) retirement.
Residential VPPs sidestep all three. The assets are already built, already interconnected, and already behind the meter. Sunrun’s installed base represents roughly 4 gigawatt-hours of battery capacity across its fleet, a figure that grows by 150-200 megawatt-hours each quarter. That is incremental capacity available today, not in 2029. The trade-off is duration: most residential batteries provide 2-4 hours of discharge, whereas hyperscaler loads often need 8-10 hours or more for true firm capacity. But for ancillary services, peak shaving, and hourly carbon matching – where the value is in flexibility and speed, not duration – distributed batteries are arguably superior to utility-scale lithium-ion, which faces the same duration limits at higher installed cost per kilowatt-hour when balance-of-plant and interconnection are included.
My analysis suggests the levelized cost of capacity from aggregated residential storage, when contracted through a platform like Voltus, now falls in the $80-120 per kilowatt-year range for 4-hour resources in constrained zones – competitive with new gas peakers and below the all-in cost of utility-scale storage plus interconnection upgrades in PJM and CAISO. That figure is my estimate based on public VPP procurement data and capacity market clearing prices; the source does not disclose contract terms. If this pricing holds, we will see more hyperscalers allocate a meaningful share – perhaps 10-20% of new capacity procurement – to VPPs by 2027.
Who This Affects
- Utility resource planners: Must now model VPP capacity as a firm resource competing with their own build plans, especially in capacity-constrained zones where distributed batteries can clear capacity auctions at prices below utility peaker costs.
- Storage developers: Face a new competitor for hyperscaler offtake agreements – one that requires no land, no interconnection study, and no permitting, but offers shorter duration and aggregated rather than site-specific control.
- Grid operators (ISOs/RTOs): Need to accelerate FERC 2222 implementation and refine telemetry standards so that VPP dispatch data integrates reliably into real-time markets and reliability assessments.
- State regulators: Must resolve double-counting risks – ensuring a kilowatt-hour discharged from a Sunrun battery counts toward either the utility’s resource adequacy requirement or the hyperscaler’s clean energy goal, not both.
- Residential solar financiers: Gain a new revenue stack (capacity payments, corporate offtake) that improves project economics and could lower customer acquisition costs, accelerating deployment flywheels.
What to Watch Next
- PJM and CAISO capacity auction results in 2025-2026: Track whether VPP aggregators like Voltus, Leap, and AutoGrid clear significant megawatts at prices that validate the $80-120/kW-year estimate.
- Google’s 2025 environmental report: Look for disclosed VPP procurement volumes and hourly matching metrics – the first hard proof that distributed storage moves the needle on 24/7 carbon-free energy.
- Sunrun’s quarterly installation and attachment rate data: A sustained battery attachment rate above 30% of new solar installs would signal the fleet is growing fast enough to matter at hyperscaler scale.
- FERC 2222 compliance filings in MISO, SPP, and ERCOT: These markets lag PJM and CAISO; their participation models will determine whether the VPP-hyperscaler model scales nationally or remains coastal.
- Utility rate case proposals for “VPP integration charges” or standby fees: Any attempt to impose new costs on behind-the-meter resources participating in wholesale markets would directly threaten the economics of this deal structure.
Bottom line: The Sunrun-Voltus-Google triangle proves that the fastest new capacity for AI data centers is not a power plant – it is the aggregate of batteries already sitting in garages across America, now being stitched into the wholesale market by platforms that speak the grid’s language.
Read the full report at Energy Storage News
Note: facts and figures attributed above to Energy Storage News reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply