Sunrun and Voltus have formed a partnership that will aggregate thousands of residential solar-plus-storage systems across the PJM and MISO territories into a virtual power plant selling capacity directly to data centers, with participating homeowners compensated for their contribution. The arrangement feeds Voltus’s “Bring Your Own Capacity” platform, giving hyperscalers a mechanism to secure megawatts of flexible, distributed resources in weeks rather than the years required for traditional generation or transmission build-out. This marks one of the largest commercial deployments to date of a hyperscaler-funded VPP model, turning residential distributed energy resources into a dispatchable capacity product for the most power-intensive segment of the digital economy.
The deal extends a template Voltus established in June with Google, which the companies described as the first hyperscaler-funded VPP aggregating DERs from PJM homes and businesses for data center operations. Data center operators face unprecedented load growth driven by AI workloads, and their procurement timelines are fundamentally misaligned with utility resource planning cycles. By tapping aggregated residential batteries, hyperscalers bypass interconnection queues and gain access to capacity that is already sited, permitted, and interconnected behind the meter. For Voltus, the Sunrun partnership dramatically expands the addressable resource pool; for Sunrun, it creates a new revenue stack for its installed base beyond retail rate arbitrage and wholesale market participation.
Residential aggregation at this scale has long been the theoretical promise of VPPs, but operational friction — customer acquisition, enrollment, telemetry, and settlement — has kept most programs in the pilot phase. Sunrun’s fleet provides a critical advantage: a single contractual counterparty with standardized hardware, communications, and customer relationships across thousands of sites. That uniformity reduces the integration risk that has plagued multi-vendor aggregations. The PJM and MISO focus is also deliberate; both markets have capacity constructs that value seasonal and demand-side resources, and both face resource adequacy concerns that make incremental flexible capacity highly valuable.
The urgency on the demand side is underscored by Oracle’s recent withdrawal of its lawsuit against the Wisconsin Public Service Commission. The tech giant had challenged a collateral requirement imposed due to its credit rating, arguing it would cost over $100 million annually for service from We Energies. Oracle’s willingness to litigate — and then settle — signals how acutely hyperscalers feel the pressure to secure power supply certainty and cost predictability. The Sunrun-Voltus model offers an alternative pathway: capacity that can be contracted bilaterally, deployed rapidly, and scaled incrementally without the balance-sheet exposure of long-term utility service agreements.
For grid operators and regulators, the proliferation of hyperscaler-funded VPPs raises familiar questions about double-counting, measurement and verification, and market power. But it also demonstrates a functional demand response mechanism that aligns commercial incentives with system needs: homeowners earn revenue, hyperscalers secure capacity, and the grid gains flexible resources without new transmission. The next test will be whether this model extends beyond the hyperscaler niche to serve broader load-serving entities, and whether the economics hold as residential battery adoption curves and market rules evolve.
Read the full report at Energy Central.