Brownfield to data center: what transfers with the site

For years the default was greenfield: cheaper land, fewer surprises, easier to build on. That calculus inverted when the binding constraint stopped being land and became power, permitting, and whether a community will tolerate the project.

A retired industrial or power-generation site tends to arrive with a head start on all three. There is existing interconnection infrastructure and often nearby transmission. There is existing water, sewer and road access. There is industrial zoning and industrial history, so the use is less likely to be a category shock to neighbours. In some cases prior electrical studies or interconnection agreements with the grid operator already exist.

Those studies usually need updating rather than starting from nothing, and the timeline is generally shorter than a greenfield start. The conversion advantage is mostly time — which is precisely the currency that is scarce.

This is the section that matters, and it is the same argument as any other power claim, applied where it is least intuitive.

A retired facility may have historic load rights, legacy interconnection equipment, and nearby transmission. None of that automatically converts into data-center-ready capacity. Utilities may require new studies, new upgrades, new protection equipment, and fresh queue positioning. A former plant's connection was engineered for generation *export* or for a different load shape entirely, and a modern AI load is not a like-for-like substitute for what used to sit there.

So the questions are contractual, not archaeological:

• Do the historic rights survive, and did they survive the facility's retirement — or did they lapse with it?
• What did the prior studies actually study, and against what load profile and what era of grid?
• Is a fresh queue position required, and if so, under which process — including whether that process is the one still in force?
• What physically remains, what is serviceable, and what is scrap that merely looks like infrastructure?

The seller's honest answer to these is usually "some of it." The marketing answer is usually "it's already powered."

This is the risk that distinguishes brownfield from every other site type, and it is not a diligence checkbox.

Former industrial properties carry the risk of environmental liability under Superfund and comparable state regimes. The defining feature of that framework is that liability can attach to current owners regardless of who caused the contamination. A buyer can acquire an obligation created decades before they existed.

The practical consequences:

• Investigate before you are asked to. Environmental findings are the most common reason a brownfield deal reprices or dies late, once both sides have spent real money.
• Scope is the variable, not presence. Almost every genuine brownfield has something. What matters is extent, cost, and whether it is bounded.
• Allocation is negotiable and must be negotiated. Indemnities, escrows, insurance and statutory protections all exist. None of them apply by default.
• Demolition is a cost line and a risk line. Structures that appear to be an asset are frequently a liability, particularly where asbestos or legacy fuel systems are involved.

Contamination, demolition and community risk can erase the time advantage that made the site attractive in the first place. The advantage is real; it is just not free.

Brownfield reuse attracts opposition in a specific and predictable way: residents who associate the site with prior industrial harm read a new large-load facility as a continuation of it. Concerns cluster around air quality, noise, water consumption, and electricity prices.

This matters commercially because it does not appear in any document. A site can clear power, environmental and zoning diligence and still stall indefinitely on a discretionary approval that becomes politically impossible. There is no instrument to inspect and no counterparty to negotiate with.

The sites that succeed tend to have industrial history *and* utility access *and* room for phased expansion *and* a community position that has been worked rather than assumed. Treat the last one as a gate with the same standing as the others, and start it at the same time — not after the permits are filed.

Reduced to its core, the trade is: you are buying time and paying for it in complexity.

It tends to be right when the schedule is the binding constraint, when the existing electrical position genuinely shortens the interconnection path rather than merely appearing to, when the environmental scope has been investigated and bounded, and when the community position is workable.

It tends to be wrong when the power advantage turns out to be historic rather than current, when environmental scope is unbounded at the point of commitment, or when the discretionary approval is the real risk and nobody has tested it.

The distinguishing question is the same one that runs through this whole pillar. Not *what did this site once have*, but what does it contractually hold today, and what does that shorten.