Energy Sourcing Lessons from Utah's Stratos Project

A multi-billion-dollar data center campus was designed to run almost entirely on natural gas. Within weeks, the state's own Republican governor said that would "never" happen. For developers, the lesson is clear: how you plan to power a site is now as consequential as where you put it.

Data center developers have spent the last two years learning that power availability, not capital, sets the pace of a project. The fight unfolding in northern Utah adds a sharper edge to that lesson: it is no longer enough to secure enough electricity. The source of that electricity has become a permitting, political, and social-license question that can stall or reshape a project after it has already been approved.

What's happening in Box Elder County

The Stratos Project, backed by investor Kevin O'Leary and advanced through Utah's Military Installation Development Authority (MIDA), is a proposed data center campus spanning roughly 40,000 acres of privately held, largely vacant land in Hansel Valley, north of the Great Salt Lake. The Box Elder County Commission approved an interlocal agreement for the project in early May 2026, clearing an early hurdle for what backers have described as a phased buildout that could eventually reach 9 gigawatts of capacity.

The Stratos Project portfolio shown on the LandGate platform

The original energy plan was striking in its simplicity: power the complex with on-site natural gas generation, drawing on the Ruby Pipeline that runs through the valley. On-site generation was pitched as a feature and not a compromise, a way to stand up enormous capacity without straining the public grid or shifting costs onto local ratepayers.

That design choice is also what turned the project into a statewide controversy.

The math that changed the conversation

A full, gas-only buildout at the scale proposed would have been one of the largest single sources of new emissions in the state. Independent analyses estimated that running the complex entirely on natural gas could increase Utah's carbon dioxide emissions by roughly half, alongside annual water consumption measured in the billions of gallons for the generating turbines alone, before accounting for the data center's own cooling needs. In a state already contending with a shrinking Great Salt Lake and persistent air-quality concerns, those numbers proved politically combustible.

Public response was swift: large protests at the Capitol, a flood of formal comments, and a grassroots effort in Box Elder County to put the approval to a referendum. Within weeks, Gov. Spencer Cox shifted from forcefully defending the project to publicly declaring it would "never" be powered solely by natural gas. The first phase is now framed as a far smaller commitment — on the order of 1 to 1.5 gigawatts of gas generation with future phases expected to incorporate nuclear, geothermal, solar, and storage.

The state has continued to move. Cox signed an executive order at the end of May establishing a statewide "data center framework" directing agencies to prioritize the health of the Great Salt Lake and protections for air and water. Notably, that order reaches the state's environmental and natural-resource agencies but not MIDA, the independent authority that approved the project, a gap critics have been quick to highlight. Around the same time, the county attorney rejected the referendum applications as administrative rather than legislative actions (a decision opponents are appealing), while nearby Iron County took the opposite tack and imposed a 180-day moratorium on new data center applications.

Why this matters beyond the Stratos Project in Utah

Strip away the local specifics and Stratos is a case study in a structural shift every developer is now navigating: energy sourcing has moved from a back-end engineering decision to a front-end gating factor.

For most of the current build cycle, the dominant constraint has been quantity of power, with interconnection queues stretching four years or more, transformer and breaker lead times, and the scramble for "powered land" with grid headroom. Behind-the-meter and bring-your-own-power models emerged largely as a workaround: if you cannot wait for the grid, generate your own. Stratos was built squarely on that logic.

But the Utah episode shows the workaround carries its own risk profile. On-site generation may sidestep the interconnection queue, yet it pulls the developer directly into air-permitting, water-rights, and emissions scrutiny that a grid-connected load might have diffused across a utility's broader portfolio. 

A few implications stand out for developers evaluating sites today:

An all-gas plan is increasingly fragile, even in friendly jurisdictions. Utah is not a state known for hostility to fossil fuels or industrial development, and the project still cleared its initial approvals. The fuel mix, not the data center itself, drew the backlash. A diversified stack (gas paired with nuclear, geothermal, solar, and storage) is fast becoming the baseline expectation rather than a sustainability talking point.

Water and air are now part of the energy equation. Thermal generation is water-intensive, and in basins already under stress, cooling and turbine water demand can become the binding environmental constraint. Energy sourcing, water availability, and air-quality headroom have to be evaluated together, not in separate workstreams.

Approval is not the finish line. A signed agreement did not insulate Stratos from referendum drives, executive orders, and agency reviews layered on after the fact. The regulatory and community environment can shift faster than a multi-decade buildout timeline — and that risk is highest where the public feels it was not consulted early.

Jurisdiction-level divergence is widening. Box Elder County approved its project while Iron County paused all new applications. Two counties in the same state moved in opposite directions within weeks of each other. Site selection increasingly demands granular, parcel- and jurisdiction-level intelligence about not just zoning and grid access, but the local political and regulatory posture toward large loads.

The takeaway for developers

The Stratos story rewards the developers who do their homework on energy before committing to a site, not after. The most defensible projects will be the ones that can show, from day one, a credible and diversified power plan, a transparent water strategy, and an early, genuine engagement with the communities and agencies that hold de facto veto power.

The land that holds its value is the land where the power can actually be delivered and defended: sites with realistic access to a varied energy mix, environmental headroom for the water and air impacts of generation, and a regulatory environment that won't reverse course after the ribbon-cutting.

At LandGate, that's the kind of due diligence our data is built for, combining parcel-level land intelligence with grid and substation capacity, generation potential across resource types, and environmental context, so developers can pressure-test a site's power story before the politics catch up to it. The Stratos Project is a reminder that the energy question is no longer a detail to solve later. It's the first question to answer.

To learn more about LandGate’s broad, nationwide data offerings for data center site selection and due diligence, book a demo with our dedicated energy & infrastructure team.