When Cardinal News sketched which western Virginia sites might qualify under Gov. Abigail Spanberger’s new data center study, the list read like a map of the region’s industrial past: former coal lands in Buchanan, Dickenson and Wise counties, the old Radford Foundry, Roanoke County’s Smith Gap landfill, and the former Kim-Stan landfill outside Clifton Forge.
The governor’s executive order asks the state to identify brownfields and previously developed sites that could spread data center investment beyond Northern Virginia while limiting the strain on energy, water and land. The study is due in March.
That is a reasonable goal. But a state list does not decide where data centers go. Localities and developers do. The more important question for communities west of the Blue Ridge is what happens after their name appears on a map like this.
If the pattern from the rest of Virginia holds, the answer is that developers arrive first. Land gets optioned quietly. Engineers confirm power and water. A finished application lands before the planning commission, and residents are invited to react to a deal they had no hand in shaping.
Virginians have noticed. In 2023, 69% of voters told the Washington Post–Schar School poll they would be comfortable with a data center in their community. This spring, 35% said so. I read that less as a verdict on the technology than as a verdict on the process.
Rural communities do not have to accept that process. Whoever owns the land can turn it around.
A county, an industrial development authority, a foundation or a nonprofit that controls a developable site does not need to wait for a developer’s proposal. It can write its terms first and then invite developers to compete to meet them. If it leases the land rather than selling it, those terms stop being promises made at a public hearing. They become lease covenants, enforceable for decades.
What should those terms say? Developers should pay for the grid infrastructure their projects require, so local ratepayers don’t. Projects should strengthen the local grid, for example by backing up hospitals, schools and emergency services during outages. Training should start before construction, through the community college, with hiring commitments that can be measured. Water use, wastewater, noise and on-site generation should come with firm limits and public reporting. And the process should be open, with no nondisclosure agreements.
None of this is anti-developer. Serious developers know that projects with community commitments built in get approved, and get approved faster, than projects that have to defend themselves after the fact.
Two cautions matter as much as the terms.
First, confirm the power before anything else. Utility timelines for large new loads now stretch for years. A site that cannot be served will draw no bids, however good its terms. Brownfields also carry their own engineering and liability questions, especially landfills and cleanup sites.
Second, keep the process competitive. The quickest way for a landowner to lose its leverage is to give one familiar developer a first look and never run an open solicitation. With one bidder, there is no comparison and no fallback.
The state’s study gives western Virginia a few months before the calls start. Communities that use that time to decide what they need, and write it down, will not be choosing between a bad deal and no deal. They will be choosing among developers who have already agreed to their terms.
Terry Lee is the founder of Garvey Labs, a Washington, D.C., firm that helps communities set the terms for data center deals so the investment builds local wealth, strengthens the local grid and protects drinking water.
