Before construction can begin, the power service agreement must clear the Kentucky Public Service Commission
Decision Focus
On July 29, the U.S. Department of Energy announced a partnership to redevelop its Paducah Site in western Kentucky into a 1.8-gigawatt AI and high-performance computing campus backed by dedicated new power infrastructure. The operational signal for energy heads is not the headline figure. It is the architecture: purpose-built generation that intentionally exceeds campus load, sited on federal land, structured to bypass the interconnection queue entirely.
90-Second Brief
Today, the DOE selected Brookfield to develop and operate the campus and NextEra Energy to build 2GW of new gas-fired generation alongside up to 2.6GW of battery storage, capacity explicitly sized above the 1.8GW campus requirement, with surplus directed to the regional grid under the administration’s Ratepayer Protection Pledge. The investment is privately funded at over $100 billion, with a 2031 completion target. Before construction can begin, the power service agreement must clear the Kentucky Public Service Commission. That regulatory gate is the controlling near-term variable, not the capital commitment.
What Is Really Happening?
The Paducah Site was the Paducah Gaseous Diffusion Plant, operational from 1952 and returned to DOE’s Environmental Management program when commercial enrichment ceased in 2013. Its residual assets—existing transmission capacity, water infrastructure, fiber connectivity, and federally controlled land—remove the greenfield interconnection problem before the first permit is filed. The DOE’s American Energy Hubs initiative is systematically converting former federal industrial sites into AI infrastructure anchors using this asset base as the opening position in negotiations with private capital. Paducah follows the Portsmouth Site partnership; it is a template being replicated, not a one-time transaction.
What distinguishes Paducah architecturally is the generation overage. NextEra’s 2GW gas installation against a 1.8GW campus load means the project is designed to export power, not merely consume it. That inverts the conventional dynamic in which data center operators compete for finite grid capacity. The campus functions as the anchor tenant for new regional generation—a structural role that changes how host utilities, state regulators, and neighboring industrial loads interact with the project. The Ratepayer Protection Pledge introduces a policy overlay that procurement teams evaluating similar co-location structures will need to model explicitly.
Why It Matters for Global Heads of Data Center Energy
The Paducah structure operationalizes what the industry has debated at a conceptual level: owning or anchoring generation rather than queuing for it. With interconnection timelines running multi-year in competitive markets, a project that pairs new dedicated generation with a pre-permitted federal site represents a materially different risk exposure than a standard grid interconnection application. The immediate implication for portfolio planning is not that every operator should pursue a federal land arrangement, but that co-location-with-generation now has a high-visibility reference architecture that regulators, developers, and infrastructure capital can begin to price consistently.
The up-to-2.6GW BESS deployment alongside gas generation merits separate attention. At roughly 1.4GW of storage per 1GW of campus load, the battery layer appears sized to perform capacity firming and grid services concurrently—not to function as backup power alone. That design logic has direct implications for how storage is scoped in future campuses where gas and renewables are mixed in the generation stack, and for how battery procurement contracts are structured relative to offtake obligations.
The Kentucky PSC approval requirement also creates a state-level precedent that adjacent jurisdictions will watch. How the commission rules on wholesale service arrangements, ratepayer protection obligations, and transmission upgrade cost recovery for a campus at this scale will inform regulatory strategy for operators planning similar dedicated-generation structures in other states.
Forward View
If Kentucky PSC approval proceeds without conditions that materially alter the power service terms, the DOE’s American Energy Hubs pipeline accelerates, and the federal site conversion model becomes a credible procurement channel for operators evaluating 500MW-plus build programs. The pace at which DOE issues new request-for-offer processes at other former federal industrial sites is the first signal to monitor.
The generation-surplus design raises a second scenario: whether surplus capacity sales to the regional grid begin to function as a revenue offset against campus energy costs. If regulators in Kentucky and elsewhere treat campus-anchored generation as a quasi-utility function, the financial model for dedicated power infrastructure shifts—changing offtake term structures and capital expenditure sizing across the industry.
A third front is the legislative environment. The introduced “No AI Data Centers on Federal Lands Act” has been assessed as unlikely to pass, but its existence signals organized political pressure against federal land use for AI infrastructure. Even without passage, similar proposals could extend environmental review timelines or constrain future DOE site conversion announcements beyond what current project schedules assume.
What Is Still Uncertain
The Kentucky PSC has not published a ruling timeline, and without that approval the project cannot advance to engineering, permitting, or procurement. The specific terms of the wholesale service arrangement with Big Rivers Electric and the retail structure through Jackson Purchase Energy Cooperative have not been disclosed, making cost-competitiveness against alternative procurement structures impossible to assess at this stage.
The environmental remediation status of the Paducah Site—where DOE continues to oversee cleanup of the former uranium processing facility—has not been characterized in terms of construction sequencing risk. How remediation milestones interact with the 2031 completion target is not confirmed by any public source reviewed as of this writing. It is also unconfirmed whether the 2GW gas installation will face state or federal emissions constraints that could affect capacity factor assumptions or long-term carbon accounting.
One Question for Your Team
If a federal or state authority offered your organization an anchor position in a dedicated-generation campus on a pre-permitted site with existing transmission—contingent on a power service agreement subject to public utility commission approval—how does your current interconnection queue exposure affect the value of that offer, and does your procurement structure allow you to respond within a 90-day window?
Sources
- Constructconnect — $100B Paducah AI Campus Announced with 2GW of New Gas Generation (Link)
