How Data Center Siting Impacts Infrastructure Development

The location of a data center significantly impacts not only the facility itself but also every piece of infrastructure connected to it.

This reality is becoming clearer as utility regulators, grid operators, and local governments grapple with an uncomfortable truth: data centers are no longer modest additions to a regional power grid. A single hyperscale campus can draw 500 MW or more — roughly the output of a mid-sized power plant — and the decision about where to site that load affects not just the operator's bottom line. It reshapes transmission investment priorities, strains interconnection queues, and can determine whether surrounding communities receive cleaner or dirtier power for the next two decades.

Advocates at recent siting proceedings have raised pointed questions about what the process would actually look like once regulators start treating data center loads with the same scrutiny applied to large industrial facilities. Those questions deserve serious answers.

Understanding Data Center Siting

Siting, in its simplest form, is the process of selecting where to build. But for data centers, that decision cascades outward in ways most industrial development does not.

A distribution warehouse needs a highway and a labor pool. A data center needs fiber, water, power — and increasingly, *guaranteed* power at scale, around the clock, with near-zero tolerance for interruption. That combination of requirements makes the siting process genuinely complex, explaining why the industry has historically clustered in a handful of markets: Northern Virginia, Phoenix, Dallas, Chicago, and a few others where those inputs converge at acceptable costs.

The clustering pattern that made financial sense for individual operators has created systemic fragility for the grid.

Northern Virginia alone — Loudoun County's "Data Center Alley" — hosts more than 35% of the world's internet traffic. Dominion Energy has publicly acknowledged that data center load growth in that corridor is straining its ability to plan transmission infrastructure on a reasonable timeline. When one geography absorbs that much demand, the grid around it bends.

The current wave of siting decisions is different from what came before. Artificial intelligence workloads require substantially more power per rack than conventional cloud computing — estimates range from 10x to 30x the power density — which means a campus that would have drawn 50 MW five years ago might now need 200 MW or more. Operators are actively hunting for locations where that power is available, affordable, and deliverable without a decade-long interconnection wait.

The Factors That Actually Drive Location Decisions

Power availability dominates. Everything else is secondary.

An operator willing to pay $80/MWh in a constrained market will still struggle if the interconnection queue stretches eight years into the future. That's not hypothetical — MISO and PJM interconnection queues have ballooned to the point where new large loads are routinely looking at 5-to-9-year timelines before they can take service. Proximity to existing transmission capacity, or to planned generation, has become the single most decisive factor in data center location strategy.

Water access is the quiet second constraint. Most large data centers use evaporative cooling, which can consume millions of gallons per day. Operators siting in the American Southwest — drawn by cheap land and favorable tax treatment — are increasingly running into water availability limits that weren't on their radar a decade ago. Mesa, Arizona, and some Nevada markets are already seeing pushback from municipal water authorities.

The regulatory environment matters more than operators typically admit publicly. States that have streamlined permitting, offered property tax abatements, and created clear interconnection pathways have captured disproportionate investment. Virginia's data center tax exemption program has been extraordinarily effective at attracting capital — so effective, in fact, that it has triggered a political backlash as residents and local officials confront the infrastructure burden those facilities create without proportional tax revenue flowing to local governments.

Environmental considerations are entering siting decisions more formally than before, partly driven by corporate sustainability commitments and partly by regulators who are starting to ask hard questions. A data center siting process that triggers environmental review is no longer unusual. What's changing is the depth of that review — questions about air quality impacts from backup diesel generators, water consumption, and the carbon intensity of the regional grid are now standard in contested proceedings.

The Financial Implications Are Bigger Than They Look

Operators who treat siting as a real estate transaction — find cheap land, negotiate a power contract, build — are increasingly getting burned.

The real cost of a siting decision doesn't crystallize until construction is underway or complete. An interconnection study that comes back requiring $150 million in transmission upgrades, costs that the data center operator may bear directly, can demolish the economics of a project that looked compelling on paper. Data centers will either need to purchase additional transmission capacity or wait years in queue — and both options carry costs that compound with time.

Grid upgrade costs passed through to data center operators are becoming a genuine risk factor. Several large operators have disclosed in investor materials that interconnection-related costs are a material uncertainty. For projects financed with debt — which is most of them — that uncertainty creates real problems for underwriters.

On the flip side, operators who site well can capture durable advantages. Access to low-cost renewable energy through long-term power purchase agreements, favorable interconnection positions, and established relationships with regional grid operators can reduce operating costs by 15-25% over a 20-year asset life compared to operators stuck in constrained markets. At data center scale, that's a meaningful number.

What Good Siting Looks Like in Practice

The operators who have navigated this well share a few characteristics.

Microsoft's investment in nuclear-adjacent siting — including its agreement to restart a unit at Three Mile Island specifically to power its data centers — reflects a recognition that firm, carbon-free power is worth paying a premium to secure, and that proximity to that generation source fundamentally changes the interconnection math. It's an extreme example, but the underlying logic applies at smaller scales.

Data center developers who engage grid operators *before* selecting a final site — rather than after — consistently report shorter interconnection timelines and fewer surprise cost allocations. This sounds obvious, but the industry norm has historically been to lock a site, then begin the interconnection process. Reversing that sequence, or running them in parallel, requires organizational discipline but pays off measurably.

Communities where siting has gone poorly share a pattern: local officials approved projects without understanding the infrastructure burden, power and water demands exceeded what the local system could absorb without significant upgrades, and the tax structure meant those upgrades were socialized while the financial benefits were privatized. The lesson isn't that data centers are bad neighbors — it's that the siting process needs to surface those trade-offs explicitly before ground breaks, not after.

Where the Siting Process Is Headed

Two forces are reshaping data center location strategy simultaneously, and they pull in opposite directions.

AI infrastructure demand is pushing operators toward *any* site with available power, loosening constraints around geography, labor, and even connectivity in ways that would have seemed irrational five years ago. Remote, grid-adjacent locations that were previously ignored are now getting serious evaluation — particularly in the Southeast and Mountain West, where transmission infrastructure was built for generation assets that have since retired.

At the same time, regulators are tightening. Siting proceedings that previously moved quickly are now attracting intervenors, environmental review requirements, and public comment processes that add months or years to project timelines. Several states are actively developing data-center-specific siting frameworks that would require applicants to demonstrate grid impact, water impact, and community benefit — essentially treating large data centers more like power plants than like commercial buildings.

The operators who will win the next decade of infrastructure development are the ones who treat siting as a strategic discipline rather than a procurement function.

That means building internal expertise in grid interconnection, engaging regulators and communities proactively, and being willing to pay for sites that look expensive on a per-acre basis but carry lower total infrastructure costs when the full picture is visible. The cheap site with a complicated interconnection isn't cheap. The operators who understand that are already positioning accordingly — and the ones who don't are going to learn it the hard way, one delayed project at a time.

Explore the InfraSale Marketplace for more insights and opportunities.

[INTERNAL LINK: data center trends]

[INTERNAL LINK: infrastructure investment]

[INTERNAL LINK: regulatory challenges]