JLARC · Data Centers in Virginia ↗
December 9, 2024
State legislative research commission. Statewide sites and buildings: printed p. 5. Economic impact: pp. 14–15. Water use: pp. 61–63. Historical baseline, not a live inventory.
Sources & methods
Follow the evidence.
See the boundaries.
Our public explainers draw on government reports, public agencies, and published research. Open the original sources to examine their methods and assumptions.
Last reviewed September 9, 2026. Source dates and data years differ.
EIA · Virginia electricity growth ↗
May 5, 2026
Federal electricity statistics. Commercial sales: 2019–2025, with 2025 preliminary. Peak demand: the PJM Dominion transmission zone, not a data-center-only or statewide series.
Loudoun County · Data centers by the numbers ↗
County inventory dated March 1, 2026; fiscal figures FY2026–FY2027
County government inventory includes buildings that are built or in construction. FY2027 revenue is a budget projection. These county figures cannot be substituted for statewide totals.
Berkeley Lab · U.S. energy report, 2025 update ↗
Published June 2026
The title says 2025 Update; publication was June 2026. National model for 2030, with reference and uncertainty scenarios. It is not a Virginia forecast.
Berkeley Lab · Water use of data center workloads ↗
June 2025
Peer-reviewed research by Lei, Lu, Shehabi, and Masanet. Explains why hardware, utilization, electricity supply, cooling, and climate change water use per workload.
U.S. DOE · Efficient data center design ↗
July 2024
Federal Energy Management Program design guide covering computing equipment, cooling, electrical systems, and efficiency metrics.
U.S. EPA · Renewable energy certificates ↗
Page updated July 10, 2026
Defines the attributes represented by a renewable energy certificate. A certificate is not a measure of hourly physical electricity delivery.
U.S. EPA · Electricity and unbundled certificates ↗
Page updated June 4, 2026
Explains the separation of a renewable certificate from the physical electricity. Supports the distinction between renewable accounting and physical supply.
Our evidence standards
Useful facts need clear labels.
- Date the measure. A source’s publication date is not necessarily the year its data describe.
- Name the boundary. Virginia, Northern Virginia, Loudoun County, and the PJM Dominion zone are not interchangeable geographies.
- Show the status. Estimates, observed outcomes, preliminary data, and projections are labeled. A project in a pipeline is not assumed to be operating.
- Keep the units. Sites, buildings, floor area, demand, energy, and jobs measure different things.
- Keep proposals identifiable. The initiative’s policy preferences and discussion draft are not represented as findings endorsed by our sources.
How the graphics are calculated
Peak-load bars start at zero and share a 30,000 MW maximum. Job segments are proportional to 59,000 and 15,000. The national forecast uses a zero-to-900 TWh axis with a 521–843 TWh range and 649 TWh reference marker. The example calculator multiplies average MW by 24 or 8,760 hours; annual GWh divides MWh by 1,000.
Graphics use the published numbers and do not imply additional measurement precision. The approximate statewide count is retained as a 2024 baseline; the newer Loudoun inventory has a different geography and includes construction.
Corrections welcome
Found a newer source or a problem with a definition? Send the original reference and the specific claim to the initiative for review.
Editorial independence
This is an independent citizen-led initiative. Links to public agencies and research institutions identify sources; they do not imply endorsement. No nonprofit tax status or introduced legislative sponsorship is claimed.