Data centers, energy and growth in Northern Virginia

The world’s largest data center market

Virginia’s Joint Legislative Audit and Review Commission, or JLARC, describes Northern Virginia as the world’s largest data center market in its 2024 study of data centers in Virginia. The study puts the region at 13% of reported operational capacity globally and 25% of capacity in the Americas. Those are shares of reported operating capacity, rather than shares of all buildings, employment or electricity use. Keeping the denominator attached to the percentage is essential to understanding the claim.

The geography matters too. The market ranking describes Northern Virginia; the study’s energy-demand history and forecast describe Virginia statewide. Read those scales together without treating them as identical. A concentration of one building type can be central to a statewide discussion while still requiring careful local assessment.

Building them

JLARC says an individual data center building usually takes about 12 to 18 months to construct. Data center representatives told the study that approximately 1,500 workers from construction-related industries can be on site at the height of construction. The workforce figure is an approximate peak reported by industry representatives. It is not an average staffing level for the full construction period or a measure of permanent operating employment.

That distinction gives a reader a useful set of questions. Is a project description discussing one building or a wider development? Is the workforce figure a peak, an average or an operating total? Does the schedule cover construction alone? These questions do not resolve a project’s merits, but they keep unlike figures from becoming a misleading comparison.

The power question

The study reports that Virginia’s energy demand was essentially flat from 2006 to 2020: efficiency improvements offset population growth. It also says modern data centers consume substantially more energy than other types of commercial or industrial operations, and identifies the industry’s expansion as a major driver of growing state demand. The historical result and the growth pressure belong in the same discussion.

An independent forecast commissioned by JLARC indicated that unconstrained demand for power in Virginia would double within ten years, driven mainly by data centers. That is the forecast horizon presented in the 2024 study. It is a scenario for unconstrained demand, rather than a measurement of growth already completed. It also concerns demand; the statement by itself does not establish what infrastructure will be delivered or when.

When weighing the forecast, keep its subject, geography, horizon and qualification together. Dropping the word “unconstrained” or turning a forecast into a definite outcome changes the meaning. The useful question is what the scenario implies for planning, while retaining the conditions under which it was presented.

Noise, water and neighbours

Energy is only one part of JLARC’s discussion. The study suggests zoning ordinances with maximum allowable sound levels for both new and existing data centers, so localities can better account for low-frequency noise. It also recommends expressly authorizing local governments to require and consider water-use estimates for proposed developments. These are policy suggestions in the study, rather than a statement that every locality already has those powers or requirements.

For someone reading a proposal, the distinction points toward concrete evidence: what sound limits are being discussed, and what water-use estimate is available? An estimate should be identified as an estimate, just as a proposed rule should remain identified as a proposal. Neither a market ranking nor an electricity forecast answers those local questions on its own.

What it means for efficiency elsewhere

Virginia’s flat-demand period shows why efficiency belongs in the regional story. Population grew while efficiency improvements offset that pressure on energy demand. The subsequent data-center forecast introduces a different scale of growth; it does not make an individual building’s energy record less useful. Evaluating a building’s performance and evaluating total regional demand remain distinct tasks.

Our energy-efficiency guide provides a starting point for thinking about building performance. Read it alongside the regional overview, keeping the questions separate: what does this building use, what is changing across the state, and what evidence supports the claims about each? That approach leaves room to assess efficiency, growth and neighbourhood concerns on their own terms.