Shoals and ON.energy Deliver 1.1 GW Battery Project for AI Data Center — Largest in the U.S.
Shoals Technologies Group (Nasdaq: SHLS) and ON.energy have confirmed that manufacturing and deliveries are underway for a 1.1 GW grid-connected battery energy storage system (BESS) at a hyperscale AI data center — a project the companies describe as the largest battery installation in the United States.
What is being built
ON.energy designed the system around its patented medium-voltage AI UPS (PowerHub), a DC electrical architecture built for scalable deployment at AI-grade power densities. Shoals supplies the electrical balance-of-system components, manufactured domestically.
The 1.1 GW figure refers to the full-project scale. Deliveries are structured in phases, with additional phases confirmed as on schedule.
Why battery storage at this scale matters
AI data centers have a power problem that is not just about how many megawatts they can draw from the grid — it is about draw quality and continuity. Large language model training and high-density inference clusters create volatile, non-linear power demand profiles that conventional utility interconnections handle poorly. A 1.1 GW BESS sitting between the grid and the facility provides several things simultaneously: instantaneous response to demand spikes, buffer capacity during grid instability, and the ability to charge during off-peak periods and discharge during peak pricing windows.
That last function has real economics. A facility of this scale can arbitrage grid pricing in ways that meaningfully reduce effective power cost per GPU-hour, which is increasingly the number that determines whether a hyperscaler’s AI compute margin is competitive.
The domestic manufacturing angle
Shoals manufactures in the United States and has emphasized that the project demonstrates domestic production capacity for large-scale AI infrastructure contracts. That matters in the current policy environment: the federal AI infrastructure push has been coupled with explicit preferences for domestically sourced hardware, and vendors who can credibly claim that supply chain are better positioned for the next wave of public-sector and DoD-adjacent data center contracts.
Scale in context
For comparison, the largest previously announced AI-adjacent battery projects in the U.S. peaked below 500 MW. The 1.1 GW number here is more than double that. It also sits within a market where ON.energy’s commercial relationship with Shoals is described as an active pipeline — the announcement notes “additional opportunities” beyond this initial project are in active pursuit.
The project does not publicly name the data center operator or location.