West Virginia Data Centers Expand as Secrecy and Pollution Risks Grow

West Virginia has accelerated its data center push despite unresolved disputes over secrecy, pollution, utility costs, and local control. The state’s 2025 framework shields qualifying projects from many county and municipal requirements. Several large proposals have since moved forward, driven partly by the computing demands of artificial intelligence.

The pitch rests on a straightforward exchange. Developers bring capital, construction work, permanent jobs, and new tax revenue. West Virginia provides land, energy, tax treatment, and a regulatory path designed for unusually large projects.

However, communities cannot evaluate that exchange when essential project details remain confidential or unsettled. Tucker County’s proposed Ridgeline complex shows the problem clearly. Residents have faced conflicting descriptions of the project, redacted permit material, and disagreement among state agencies about its status.

This is not simply a debate about whether data centers are good or bad. It is a test of whether West Virginia can govern industrial projects whose energy needs rival those of cities. The state wants to move faster than neighboring markets while avoiding costs that could fall on residents.

The central conflict is therefore development speed versus public accountability. West Virginia data centers are advancing under rules built to attract investment. Those rules now face pressure from permit appeals, community organizing, federal scrutiny, and the physical demands of AI infrastructure.

West Virginia Data Centers Entered a New Regulatory System

West Virginia changed the balance of power before residents could see the full scale of the projects under discussion.

House Bill 2014 established programs for certified microgrid districts and high-impact data centers. A microgrid is a localized power system that can serve a defined site independently from the wider electric grid.

The Legislature passed the measure in April 2025. Governor Patrick Morrisey signed it later that month, and its principal provisions took effect on July 11, 2025.

The enrolled state law gives the Department of Commerce responsibility for certifying qualifying projects. It also changes how their property is valued and how resulting tax revenue is distributed.

Certified projects receive broad protection from county and municipal regulation. Local governments cannot apply many zoning, permitting, inspection, and licensing requirements that would normally shape industrial development.

That restriction matters because data centers affect much more than the buildings containing servers. Their supporting infrastructure can include transmission lines, substations, water systems, backup generators, gas pipelines, and dedicated power plants.

State environmental permits still apply. Utility contracts and infrastructure decisions can also require regulatory review. Yet those processes examine separate pieces of a project, often under different legal standards.

The law requires data centers to notify the state when they expect to meet its high-impact threshold. The Department of Commerce then determines whether a project qualifies for certification.

At the same time, companies can identify submitted material as confidential business information. The state must maintain records for government use while removing protected identifying information from public versions.

That structure creates a difficult information gap. Officials can know more than residents, while separate agencies may possess different portions of the same project record.

The gap became visible around the proposed Ridgeline facility near Davis and Thomas in Tucker County. Fundamental Data sought an air permit for a natural gas power facility that would support large-scale computing operations.

Early public records did not provide a complete picture of the proposed equipment. Material concerning turbine models, control systems, fuel details, and supporting calculations was withheld or redacted under confidentiality claims.

Residents learned about the proposal through an air-permit notice rather than a conventional local development process. Local officials also said they had little advance knowledge.

An extensive 2025 investigation found that the legislation and project emerged during the same period. Nearly 300 people later attended a community meeting seeking answers about the development.

The timing intensified suspicion, although timing alone does not establish coordination. The more important fact is structural. Residents had limited local authority by the time the project’s possible scale became widely understood.

Fundamental Data has since publicly described Ridgeline as an AI, cloud, and enterprise infrastructure project. Yet its representations to different government bodies have continued to fuel arguments about what is planned and how definite those plans are.

That uncertainty turns routine permitting questions into a larger governance problem. West Virginia created a fast path for data centers before establishing one public process that evaluates their combined effects.

AI Infrastructure Makes the Energy Question Larger

AI changes the data center debate because computing growth becomes an energy-development decision, not merely a real-estate decision.

Traditional data centers already operate continuously and require dependable electricity. AI training and inference add dense clusters of accelerators that consume more power and produce more heat than many earlier workloads.

A developer must therefore secure electricity at sufficient scale, speed, and reliability. Existing utilities may lack the generation or transmission capacity needed for a large campus on the developer’s preferred schedule.

West Virginia’s answer is to let eligible projects pair computing facilities with dedicated microgrids. Those systems can include on-site generation built specifically for a data center.

The approach offers developers greater control over power availability. It can also reduce dependence on upgrades elsewhere in the regional grid, at least when the project truly supplies its own generation and infrastructure.

Governor Morrisey’s August 2026 development plan reinforced that principle. State leaders said developers should build, bring, or procure the energy resources their projects need.

The plan also presented commitments covering water protection, consumer rates, local services, environmental oversight, and workforce development. Those commitments respond directly to the criticisms surrounding the state’s earlier rollout.

The unresolved question is how these principles will operate when a project needs shared infrastructure. A dedicated campus can still rely on pipelines, transmission corridors, municipal water, roads, emergency services, or utility-backed generation.

FirstEnergy affiliates have explored major new generation in West Virginia as data center demand rises. Critics have asked whether existing customers could begin paying for infrastructure before enough new demand becomes certain.

That concern is not abstract. Power plants and transmission facilities require long planning horizons and large capital commitments. A data center proposal can change ownership, scale, equipment, or schedule before its supporting infrastructure is complete.

Regulators can use special contracts, cost-allocation rules, security deposits, and minimum payment commitments to limit exposure. The public must still see enough of those arrangements to judge whether protections are credible.

Secrecy makes that review harder. A developer may reasonably protect customer identities, technical designs, or commercially sensitive negotiations. However, ratepayers need clarity about who bears the cost if projected demand disappears.

AI companies also complicate traditional forecasting. Demand for computing remains strong, but the efficiency of chips and models keeps changing. A future workload might require more total electricity, or it might produce more computation from each megawatt.

Those possibilities can coexist. Efficiency often reduces the cost of using a resource, which can encourage greater total consumption. Energy planners cannot assume that more efficient AI systems will automatically reduce statewide demand.

West Virginia data centers therefore place pressure on several groups at once. Utilities must forecast unfamiliar loads. Regulators must separate private risk from public obligation. Developers must prove that energy commitments survive changes in the AI market.

Residents carry a different risk. They may face higher bills, new industrial infrastructure, or reduced grid reliability if safeguards fail. They receive limited benefit when jobs and tax revenue do not match the scale promised.

The state’s strategy succeeds only if those risks remain with the projects creating them. A general promise that existing customers will be protected is not enough. The protection must appear in enforceable contracts, regulatory orders, and transparent accounting.

Secrecy Weakens the West Virginia Data Centers Pitch

Confidentiality becomes a public risk when residents cannot verify a project’s size, emissions assumptions, water demand, or financial responsibility.

Some secrecy is normal during competitive site selection. Companies often protect prospective customer names, equipment choices, land negotiations, and construction schedules.

The problem begins when confidentiality covers information needed to test environmental or economic claims. Communities cannot assess a project when the public record hides the assumptions behind its permits.

The Ridgeline dispute illustrates that line. Fundamental Data sought permission for a gas-fired facility capable of supplying a large computing development. The company classified substantial portions of its application as protected business information.

The West Virginia Department of Environmental Protection accepted many of those confidentiality claims. It later issued air and construction permits for the facility.

Citizen groups argued that the redactions blocked meaningful review. Their objections focused on information used to calculate expected emissions, including equipment and manufacturer details.

Those groups included Tucker United, the West Virginia Highlands Conservancy, and the Sierra Club. They challenged both the permit and the treatment of confidential material.

The West Virginia Air Quality Board largely upheld the agency’s decision in February 2026. Its permit decision nevertheless required more stack testing after operations begin.

Stack testing measures pollutants directly in exhaust from operating equipment. The board ordered additional testing during the facility’s second and sixth operating years.

That condition matters because the facility received a synthetic minor permit. Such a permit uses enforceable operational limits to keep emissions below thresholds that would trigger stricter major-source requirements.

The board noted that the facility’s physical potential to emit would exceed the permit limits without those restrictions. Continued compliance therefore depends on operating practices, monitoring, and enforcement.

This does not mean the facility will violate its permit. It means the permit’s credibility rests on controls that regulators must verify after construction.

Critics say they need equipment details to evaluate whether the emission calculations are realistic. The developer and regulators have argued that protected technical material can remain confidential without preventing lawful review.

That disagreement has moved into further administrative and court proceedings. Community organizations are seeking broader disclosure and stricter examination of the project’s pollution profile.

A separate dispute has deepened the transparency problem. The Department of Commerce questioned whether Fundamental Data controlled the proposed property and whether its data center plans triggered state notification requirements.

Fundamental Data described the project as conceptual during that exchange. The Department of Commerce then withdrew a violation notice after concluding that the relevant certification requirement had not been activated.

Yet environmental regulators had already relied on evidence of site control when processing the power facility’s permit. The Department of Environmental Protection later reviewed additional ownership documentation.

Reporting on this oversight gap showed how each agency evaluated only its assigned portion. No single regulator reconciled the project’s full land, computing, energy, and environmental record.

Fundamental Data says Ridgeline is privately funded and presents an economic opportunity for West Virginia. The company has also disputed state actions that it believed treated the project unfairly.

Residents see the same sequence differently. They see a developer obtaining a power-plant permit while describing the associated data center as conceptual elsewhere.

Those positions are not necessarily legally incompatible. A company can seek early permits while refining a larger project. However, the difference matters greatly to the public.

A conceptual project should not receive public treatment as a guaranteed economic prize. A sufficiently definite project should not avoid disclosure duties by remaining conceptual in another forum.

West Virginia needs a consistent threshold that separates preliminary exploration from an active development. Otherwise, companies gain flexibility while communities absorb uncertainty.

Pollution Risks Extend Beyond the Server Buildings

The environmental issue is not the server itself, but the industrial system required to power, cool, and protect it around the clock.

Ridgeline’s most immediate pollution dispute concerns natural gas combustion. Turbines can emit nitrogen oxides, carbon monoxide, volatile organic compounds, particulate matter, and greenhouse gases.

Permit limits determine how much a facility can emit legally. Monitoring and stack tests help regulators determine whether actual operations stay within those limits.

A data center may also maintain diesel backup generators. Those generators usually operate during testing, maintenance, or power interruptions, but their local emissions can matter near homes and schools.

Noise creates another layer of concern. Cooling equipment, transformers, generators, turbines, and ventilation systems can produce persistent sound with different frequencies.

A conventional daytime measurement may not capture the experience of low-frequency noise at night. Project reviews need operating assumptions, equipment placement, terrain modeling, and enforceable boundary limits.

Light pollution matters in Tucker County because Davis and Thomas sit near areas valued for dark skies, recreation, and scenic landscapes. Large campuses can affect that character even when they comply with air permits.

Water use remains less certain because cooling designs differ. Evaporative systems can consume substantial water, while air-cooled or closed-loop designs use less water but may require more electricity or capital.

That variation makes project-specific disclosure essential. National averages cannot answer what one West Virginia facility will withdraw, consume, discharge, or recycle.

Developers should disclose the cooling method, expected peak demand, average consumption, drought response, and wastewater route. Without those facts, neither strong criticism nor confident reassurance has a reliable foundation.

State leaders say existing water law and environmental regulation remain in force. That is important, but fragmented permits can still miss cumulative pressure across several facilities.

A water utility might assess whether it can serve one customer. An air regulator examines allowable emissions from one plant. A transportation agency reviews roads, and emergency services evaluate response needs.

The combined project can still change a county’s physical and financial systems. That is why residents continue asking for a comprehensive review beyond individual permits.

West Virginia’s terrain adds another complication. Proposed sites can sit near forests, streams, abandoned mines, karst formations, or communities with limited water infrastructure.

The state should avoid treating environmental questions as generalized opposition to technology. Local objections often concern specific exposure pathways and incomplete documentation.

Developers also deserve predictable standards. Clear disclosure rules can prevent every project from becoming a prolonged argument about which details are public.

One workable distinction separates proprietary designs from impact data. A company might protect a customer name or equipment configuration while disclosing expected energy use, water demand, emissions, and emergency-service needs.

Federal lawmakers have begun examining whether nondisclosure agreements obstruct public understanding of AI infrastructure. A March 2026 Senate inquiry requested information from utility commissions, including West Virginia’s regulator.

The inquiry asked about restrictions involving energy use, water access, infrastructure strain, and environmental or economic effects. It reflects a broader concern that confidentiality can hide costs affecting households.

West Virginia is not alone in facing these questions. Virginia, Texas, Pennsylvania, and other states are reassessing how large computing loads interact with power systems and local communities.

West Virginia’s distinctive choice was to limit local authority while inviting projects that may bring their own industrial power generation. That combination raises the stakes of state-level transparency.

If local zoning cannot govern noise, setbacks, building design, or compatible land use, the state must provide an equivalent venue. Otherwise, important impacts can fall between environmental permits and economic-development decisions.

Jobs and Tax Revenue Need Project-Level Proof

Economic benefits should be measured against the public concessions and infrastructure obligations attached to each project.

Supporters present data centers as a route to diversify an economy long tied to extractive industries. Large campuses can produce construction work, demand for electrical trades, and long-term property revenue.

They can also support suppliers, network infrastructure, and technical training. Communities with declining tax bases understandably take those possibilities seriously.

The difficulty is distinguishing construction activity from permanent employment. A hyperscale campus can require thousands of workers during development but far fewer people after operations begin.

Permanent roles may include technicians, security staff, facilities workers, engineers, and administrators. Their number depends on campus scale, automation, outsourcing, and whether corporate functions remain elsewhere.

Job quality also matters. A smaller number of stable technical positions can still benefit a county, especially when employers fund training and recruit locally.

The public needs specific commitments, however. Announcements should separate direct construction jobs, permanent jobs, contractor roles, and estimated indirect employment.

Tax revenue requires the same discipline. West Virginia’s framework centralizes the valuation of qualifying property and distributes resulting revenue according to statutory formulas.

The host county receives only one portion. Other shares support statewide purposes, including income-tax reduction and designated development or grid funds.

That distribution may spread benefits beyond the project’s location. It can also leave host communities questioning whether their share covers roads, emergency response, water capacity, and other direct burdens.

Tax projections should identify their assumptions. Those include the project’s completion schedule, taxable property, depreciation, exemptions, operating life, and promised investment.

A headline investment figure does not equal immediate taxable value. Projects can unfold in phases, and later buildings may depend on customers or energy capacity that have not been secured.

Berkeley County’s announced project illustrates the scale now under consideration. State officials described a planned investment of $4 billion across approximately 548 acres.

Such an announcement signals serious development interest. It does not settle the project’s final load, construction schedule, tenants, water requirements, or permanent workforce.

Tucker County presents the opposite problem. Its proposed project has attracted intense attention while fundamental details have changed or remained unclear.

Both cases demonstrate why one statewide label cannot substitute for project records. Certification should begin, not end, public examination.

Supporters can strengthen their case by publishing measurable community-benefit commitments. These can cover hiring, apprenticeships, emergency services, infrastructure security, noise limits, water conservation, and decommissioning.

Developers should also provide financial assurance for unfinished or abandoned infrastructure. A project that never reaches full operation should not leave utilities or counties with stranded obligations.

Critics must apply the same evidentiary standard. Not every proposed megawatt will be built, and not every announced campus will operate at maximum load.

Adding every preliminary proposal together can exaggerate near-term demand. Ignoring them until construction begins can leave regulators unprepared.

The appropriate response is scenario planning. Regulators should publish low, expected, and high development cases, along with the investments each case requires.

Each scenario should show who pays if demand arrives and who pays if it does not. It should also identify which costs can be recovered from ordinary utility customers.

This information would turn an ideological debate into a testable public bargain. Residents could compare actual tax receipts, employment, emissions, and infrastructure spending with the original promises.

West Virginia data centers should not receive credit for benefits that remain projections. They also should not be condemned for impacts that permits and operating data show did not occur.

Transparent measurement protects both sides. It rewards projects that meet their commitments and exposes those that depend on secrecy or optimistic forecasts.

Three Signals Will Show Whether the Strategy Works

West Virginia’s data center policy will be judged by enforceable records, not announcements or generalized assurances.

The first signal is the outcome of the Ridgeline litigation and related disclosure disputes. Courts and regulators must decide how much technical information can remain confidential during air-permit review.

A decision requiring broader disclosure would strengthen public scrutiny across future projects. A decision preserving extensive redactions would place more responsibility on regulators to explain their independent verification.

The second signal is the content of utility proceedings for new generation, transmission, and special contracts. These records should reveal whether developers bear the costs created by their demand.

Watch for minimum payment obligations, collateral requirements, exit fees, construction milestones, and protections against stranded assets. Those details matter more than a broad commitment to protect ratepayers.

The strongest evidence would be an order assigning project-specific costs directly to the customer. Weak or heavily redacted safeguards would leave the central financial risk unresolved.

The third signal is how the Department of Commerce applies certification and disclosure rules. The agency needs a consistent way to determine when a conceptual proposal becomes a reportable high-impact data center.

A credible process would publish project names, locations, development stages, expected loads, resource needs, and nonconfidential certification findings. Sensitive customer information could remain protected.

The recent growth of organized opposition will keep these questions visible. In September 2026, community and environmental groups announced the West Virginia Data Center Accountability Project.

The coalition connects residents from several counties facing data center proposals. Its formation shows that project disputes have moved beyond one facility or one region.

This pressure does not guarantee stricter rules. It does ensure that state agencies, legislators, utilities, and developers will face sustained requests for records and public participation.

The industry also has an opportunity. Developers that disclose impacts early can distinguish themselves from projects associated with conflicting claims or extensive redactions.

Early engagement cannot replace enforceable oversight. It can reveal local constraints before they become expensive legal, engineering, or political problems.

For technology buyers, the issue extends beyond West Virginia. AI services depend on physical infrastructure whose energy, water, and permitting risks affect reliability and expansion schedules.

Companies purchasing AI capacity should ask providers where computing facilities obtain power, how they manage water, and whether projects face unresolved litigation. Those factors can influence delivery timelines and reputational exposure.

Investors should examine whether announced capacity has land control, permits, interconnection rights, customers, and financing. A proposed campus is not the same as operating compute.

Residents should focus on documents that create binding obligations. Press releases describe intentions, while permits, utility orders, contracts, and monitoring reports determine actual outcomes.

West Virginia can still build a credible data center sector. Its abundant energy resources and proximity to major eastern markets provide a serious development case.

The state must now show that speed does not require opacity. It must reconcile agency records, disclose measurable impacts, and keep project costs away from households that did not create them.

That standard does not prejudge every facility. It asks each developer to demonstrate that its public benefits exceed its local and statewide burdens.

The question for the next several months is practical: Will West Virginia publish enough evidence for residents to evaluate that bargain before construction makes the most important decisions irreversible?