Eastern Panhandle United

Data center development is encroaching on Eastern Pandhandle farmland and communities without local input or adequate deliberation.

Berkeley and Jefferson Counties lie at the northern end of the Shenandoah Valley, on limestone soils that have been farmed since the eighteenth century. Together the two counties hold 1,369 farms working on 122,507 acres -- about a third of Berkeley County and nearly half of Jefferson.

More than 16,000 acres are under conservation easements and another 30,000+ acres are held as wildlife management areas, nature preserves, parks, scenic corridors, and local land trusts. The network of protected spaces reflects a community committed to balancing rapid regional growth with land conservation.

The limestone that makes this soil productive also makes the water beneath is unusually exposed. Much of the region is karst, or bedrock dissolved over millennia into fissures, conduits, and caves. Water moves through karst quickly and along channels that are difficult to map. A contaminant entering the ground at one point can emerge at a spring miles away in days, rather than filtering slowly through soil and rock. Karst is also why this region is prone to sinkholes.

A karst aquifer is not an evenly satuarated layer that draws down predictably. Rather, it is a network of channels, and pumping from one part of it can affect wells and springs that ordinary modeling may not connect. Most households outside the municipal systems draw from wells and springs.

The Eastern Panhandle received a little over 30 inches of precipitation in 2025, and the region has been recovering from a severe drought since 2024. The areas current drought status ranges from abnormally dry to moderate drought.

Eastern Panhandle Data Center Proposals Overview

Currently, there are two proposed data centers in Berkeley and Jefferson Counties. Both of these sites were selected by the developer without local notice or input. As of September 2026, construction has yet to begin for either project.

Documented

Bedington Project

  • Location: Berkeley County, WV on farmland in Bedington near Scrabble Road and Berkeley Station Road
  • Developer: Penzance (Virginia-based company)
  • Acreage: 638 acres
  • IT Load: over 600 MW
  • Estimated Temporary Jobs: 1000+
  • Estimated Permanent Jobs: 125+ (90k-110k each)

Documented

Kearneysville Project

  • Location: Berkeley and Jeffereson County, WV on an old orchard in Kearneysville off Route 115
  • Developer: QTS
  • Acreage: 300 acres
  • Tenant: QTS (this company builds and occupies their own data centers)

What remains undisclosed

Non-disclosure agreements and single purpose LLCs, often called shell companies, are a common practice in the data center development industry across the United States. West Virginia is no exception, with many project details yet to be disclosed. HB2014 certified data centers, like the Bedington Project, are also except from Freedom of Information Act requests, preventing the public from viewing project details submitted with the application for certification.

Not disclosed

Bedington Project

  • Utility power load study
  • Estimated average and peak water usage
  • Water source and withdrawl location
  • Impact studies and mitigation measures
  • Cooling technology
  • Server base
  • Tenant

Not disclosed

Kearneysville Project

  • Utility power load study
  • Estimated average and peak water usage
  • Water source and withdrawl location
  • Impact studies and mitigation measures
  • IT load
  • Cooling technology
  • Server base
  • Estimated Temporary Jobs
  • Estimated Permanent Jobs
  • I could be wrong, but I think there is a plan that the BC planning commission has seen, but is not public -- need to verify what is and is not known here (not under HB 2014):

What can reasonably be estimated

Where a figure has not been disclosed, it can sometimes be estimated from figures that have been. Currently, only the proposed data center project in Bedington has disclosed IT load (600 MW). Since the project in Kearneysville has not disclosed any information regarding IT load or energy, the following estimations consider a single data center.

Estimated

Annual electricity demand at full build-out.

Estimated energy consumption of the Bedington Project data center
How this was calculated

Definitions:

  • Power is the rate in which energy is generated or consumed and has units of Watts (Power = Energy/Time)
  • Energy is the total amount of work that was done and has units of Watt-hours (Energy = Power x Time)
  • Watt: the international unit of power (1 W = 1 Joule/second)
  • Kilowatt: 1 KW = 1,000 W
  • Megawatt: 1 MW = 1,000,000 W = 1,000 KW
  • Gigawatt: 1 GW = 1,000,000,000 W = 1,000,000 KW = 1,000 MW
  • Terawatt: 1 TW = 1,000,000,000,000 W = 1,000,000,000 KW = 1,000,000 MW = 1,000 GW

The disclosed critical IT capacity, 600 MW, is a power measurement that tells us the rate in which the installed server base will consume energy if operating at full capacity. The total energy the facility will consume depends on:
  • The operating capacity (i.e. how much of the installed server base will be utilized at once)
  • The cooling technology used (NOT included in the 600 MW figure)
  • The amount of time the facility is consuming energy (data centers operate 24/7)
Our calculations assume the facility will operate at a 67.5% utilization:
  • 600 MW x 0.675 = 405 MW
Operating at a rate of 405 MW for 24 hours a day, 365 days a year yeilds:
  • 405 MW x 8760 hours = 3,547,800 MWh = 3.547 TWh (annually)
Note: The 600 MW figure is critical IT capacity only and does not include whole-site power draw. The total facility demand would be this IT load times the facilities Power Usage Effectiveness (PUE), which adds cooling, power conversion losses, and auxillary systems. A typical modern hyperscale PUE is 1.2 to 1.4. This would give the Bedington Project a total power draw between 720 to 840 MW (or 4.26 TWh to 4.97 TWh of energy annually). The estimations here likely understate the energy needs of this facility.

Estimated

Placeholder — water estimate.

Placeholder figure.

How this was calculated

Method goes here.

Where To Go Next