By Sean O’Leary, Senior Researcher

Recently, ORVI published a paper comparing the regional grid managed by PJM, which relies heavily on base load generating sources, including coal, natural gas, and nuclear, to a more flexible grid design, which integrates more renewable, storage, and demand-side resources and does so based on price and cost-effectiveness. The flexible model is exemplified by the ERCOT grid, which serves Texas and, which over the past decade, has managed to keep electric rates almost unchanged despite large increases in demand.

The ORVI paper suggested that PJM, West Virginia, and its other member states transition to a more flexible model that, like ERCOT’s, shifts financial risk from customers to generators and relies more heavily on market pricing to motivate utilities, generators, and transmission providers to minimize cost. The need to take this approach was recently illustrated by testimony before the West Virginia Public Service Commission on behalf of West Virginia Citizen Action Group, Solar United Neighbors, and Energy Efficient West Virginia.

Chelsea Hotaling and Jim Grevatt of The Energy Futures Group and Catherine Kunkel of the Institute for Energy Economics and Financial Analysis testified in response to an Integrated Resource Plan (IRP), filed by FirstEnergy subsidiaries Potomac Edison and Monongahela Power, in which the companies propose to build a new 1,200 MW natural gas-fired power plant to serve northern West Virginia and the state’s eastern panhandle. The problem as pointed out by Hotaling, Grevatt, and Kunkel is that the proposed power plant is a risky and expensive solution to a problem that may not even exist. But the plant proposal is also a financially rational response by the FirstEnergy companies to perverse incentives, created by West Virginia and PJM, which support base load plant construction while mitigating against flexible, non-base load resources and, which as a consequence, produce unnecessarily high electric rates.

Creating a problem to justify a solution

Although the companies modeled just 243 MW of expected load growth associated with data centers in their IRP, they make reference to “over 1 GW of expected large load additions in the Companies’ service territory that were not modeled in the IRP due to timing considerations.”¹ But both figures are doubtful and, according to Kunkel, neither justifies a 1,200 MW gas plant.

As Hotaling and Kunkel point out, the companies’ efforts to have their forecast of 1 GW or more of data center-related load growth included in PJM’s overall forecast were rebuffed due to the absence of a signed contract or identified customer. The 243 MW that are included in the IRP are similarly lacking. Meanwhile, experts in the field anticipate that as few as 10% of proposed data centers will be built.

If the added data center demand fails to materialize, Kunkel finds that the companies already have sufficient power for the period of the IRP, which extends through 2035. They also fall only slightly short of having sufficient capacity and the shortage doesn’t appear until the last two years of the period.

Table 1: Companies’ projected capacity position without 243 MW data center

The addition of 243 MW of data center load also leaves the companies with significantly less need for generation and capacity than would be provided by a new 1,200 MW plant.

So, why do the companies want to build a power plant that they seem not to need? First, because, in West Virginia’s fully regulated electricity market, utilities earn a guaranteed return on equity for capital expenditures. According to the company’s figures, the proposed plant will cost between $3 billion and $3.7 billion to build, an amount they will be able to fully recover from ratepayers plus a profit regardless whether the plant proves to be necessary.

In other words, the risk falls not on the company and its investors, but on captive customers who have no say in the matter apart from whether the West Virginia Public Service Commission approves the IRP. And the risk is compounded by the fact that, from a cost perspective, this is the worst time in history to build new, gas-fired power plants, the cost of which has tripled in the last three years.

If on the other hand, we were in a market operating on a flexible model and in which power producers are deregulated, this scenario would not arise because investors rather than the public would shoulder the financial risk and, in the face of uncertain demand, it very likely would never be built.

Failing to explore better, less expensive solutions

The companies’ proposal is also noteworthy for all the available energy resources it does not include or, as Grevatt points out, in some cases, even examine. First, it must be noted that the IRP includes a fairly paltry 70 MW of new solar capacity out of the total 1,270 MW that are being proposed. But solar, wind, and battery storage were arbitrarily capped at 200 MW each. And demand-side generating, storage, and energy efficiency resources were not modeled at all.

That’s disappointing because in flexible models, where all resources are allowed to compete based on their ability to deliver power to the grid cost-effectively, batteries, renewables, and demand-side resources are making large inroads because they are often the least expensive resources available. And, in cases such as this one in which utilities are in need of less than 250 MW, it is well within the realm of possibility that a mix of resources, including those that were overlooked can satisfy the need without the requirement of major capital investments.

But, there is the rub. Utilities make money on capital investments and, in West Virginia, are assured of a return on capital. However, many of the technologies that received little or no consideration from the companies, require far less capital than a new power plant, making it not in the utility’s interest to propose them even if they offer power at less cost.

For that reason, it’s not surprising that the FirstEnergy companies don’t offer energy efficiency programs in West Virginia. That said, state law requires utilities to quantitatively assess energy efficiency solutions and the fact that the companies felt they could safely blow off the requirement tells you something about the regulatory environment in which they operate.

On the other hand, flexible grid markets that enable both generating and non-generating resources, both renewable and fossil fuel, to compete on a level playing field in which the ability to deliver service cost effectively is the arbitrator are seeing explosive growth in exactly the resources that the FirstEnergy companies ignored and discouraged.

How much of a difference it would make in the companies’ territories is unknown. As Hotaling put it in her testimony,

“Due to the shortcomings of the Companies’ IRP analysis, we cannot say what resource mix would best address the projected capacity shortfall. As | explained earlier in my testimony, the Companies did not model demand-side resources at all, and the Companies applied cumulative build limits to wind, solar, and storage. If the Companies had given those resources greater consideration, they may have developed a portfolio that included demand response, energy efficiency, battery storage, and/or more renewables, with less reliance on new gas resources. But without the benefit of a more even-handed analysis, we simply do not know.”

Failing to turn a critical eye on existing resources

FirstEnergy is the owner and operator of nearly 3,000 MW of coal-fired capacity at the Fort Martin and Harrison plants in West Virginia. At a time when, despite efforts to counter the trend by the Trump administration, coal plants are falling into disuse, if not retiring, coal plants in West Virginia seem strangely immune to market forces.

Although seven coal plants that are over fifty years old continue to operate, they currently do so at deeply reduced rates of utilization – regularly less than 50% of capacity – and they do so despite the inherent inefficiencies of coal-fueled generation and climbing maintenance costs as the plants age.

The situation is made possible by some analytic sleight of hand in which likely costs associated with the plants are conveniently overlooked. Kunkel explains,

“The modeling conducted in the IRP only reflects only the Companies’ projection of baseline capital and O&M costs, without incorporating the additional costs and risks associated with major repairs/replacements or future environmental compliance. By not adequately reflecting the uncertainty surrounding long-term reliability, major component replacement, and potential environmental compliance costs, the IRP may have understated the costs of continuing to operate the Fort Martin and Harrison plants.”

This is another example of a situation that would not arise in a price and cost-driven market and which results in West Virginia customers paying unnecessarily high prices for electricity. Making changes in the system and the markets upon which it relies is not without complication. Transitioning to a more flexible, market-driven system would require legislative and regulatory changes at the state level and within PJM generally. But, as this real-world case demonstrates, significant rewards are within reach if we have the wisdom, the courage, and the political will to pursue them.

By Nick Messenger, Senior Researcher & Economist

Alarm bells are ringing in Washington, D.C., with the Trump Administration warning that Social Security is projected to become insolvent in 2032. This would trigger an automatic 22% cut to benefits for over 70 million elderly or disabled Americans who rely, at least in part, on the program.

Debate is swirling this week, as two unlikely allies have teamed up with a bipartisan proposal to address the shortfall and prevent looming benefit cuts. U.S. Senators Elizabeth Warren (D-MA) and Bernie Moreno (R-Ohio) have proposed lifting the Social Security payroll tax cap. They estimate this would bring in over $3 trillion to Social Security and keep the program solvent into the next generation.

Currently, in 2026, wealthy Americans pay nothing into Social Security on the portion of their annual incomes that is over $184,500. The vast majority of Americans earn far less. This means that nurses, social workers, non-unionized warehouse workers, construction workers, private school teachers, and anyone who earns less than $184,500 pays a larger share of their income into Social Security than millionaires, billionaires, and wealthier professionals, like lawyers, doctors, and CEOs.

We reviewed county-level Social Security Administration data from December 2024 and calculated that, in Kentucky, Ohio, Pennsylvania, and West Virginia, looming automatic cuts to Social Security would result in the loss of nearly $2.5 billion in monthly benefits for our region. That’s $30 billion per year instantly removed from peoples’ incomes, making disabled Americans and senior citizens less able to afford groceries, transportation, healthcare, and housing. On average, recipients would see a cut of $363 per month, or $4,366 per year.

Appalachians will be heavily impacted by what happens (or doesn’t happen) to Social Security. Data from the Social Security Administration shows that Appalachian counties in Kentucky, Ohio, Pennsylvania, and West Virginia have 13.6% more monthly Social Security recipients per capita than the non-Appalachian counties in the same states. As a whole, Appalachian counties have nearly 8.3% more Social Security recipients per capita than the country.

What’s more, Social Security checks are more important to people in Appalachia and the overall Appalachian economy. Nationally, Social Security provides about 6-7% of peoples’ incomes, but in West Virginia, the state most dependent on Social Security, the figure is nearly two-thirds greater at 10-11%.

Finally, as important as Social Security income is to people in Appalachia, they are likely to get a smaller lifetime benefit than most Americans because life expectancies there are as much as five years less than the national average. In the poorest parts of Appalachia, life expectancy barely reaches age 67, the age at which they become eligible for full benefits.

We acknowledge that there is growing backlash to tax increases. Understandably so, as inflation rose nearly 3.8% last year, which remains well above the Federal Reserve’s 2% target. Everything is more expensive for everyone. And there are many examples of debatable government spending that anyone, of any political persuasion, can point to. But the fact is, raising revenue through lifting the payroll cap remains the most bipartisanly popular approach to addressing Social Security. 75% of surveyed Republicans, 90% of surveyed Democrats, and 80% of surveyed Independent voters supported revenue increases as opposed to benefit cuts.

Critics arguing for “fairness” are misguided.

But Beltway ideologues masquerading as economic experts have hijacked this debate, framing the Warren-Moreno proposal as “unfair.” While at first glance, it might be enticing to argue for “means testing,” where policymakers get to then decide who “deserves” Social Security and who does not, this is a slippery intellectual slope that could lead to more and more taxpayers being denied benefits. What should the cutoff be? The answer is inherently subjective.

Other wonky proposals often center on either masked benefit reductions (such as using chained CPI rather than wages to calculate benefit adjustments or raising the eligibility age) or the privatization of Social Security altogether, which would represent a windfall for private, for-profit investment firms.

Neoliberal and libertarian think-tank proposals peddling benefit reductions are dressed in political, Reagan-era, anti-government rhetoric and intended to further undermine one of the lasting tenets of the social contract established by the New Deal: that each new generation of American workers would pay it forward to help the previous generation retire securely in old age. While Boomers are not faultless in perpetuating some of America’s deep economic problems, it is nevertheless radical to propose that this should be the first generation of workers to abandon that promise. Aside from violating the social contract that enabled a century of American prosperity, the political resistance to the Warren-Moreno payroll tax cap elimination has several problems.

First, Boomers are not “often wealthy.” While many billionaires are coincidentally Baby Boomers, most Boomers are not billionaires. Guys like Bill Gates and Jeff Bezos dramatically skew the distribution of the average. Instead, we should look at the median, which shows us the wealth of the middle-of-the-pack Baby Boomer. Data from the Federal Reserve Survey of Consumer Finances compiled by Fidelity show that the median net worth gap between generations is not nearly as large as the skewed averages.

Despite the political narrative that “all” Baby Boomers have pillaged the American economy and hoarded wealth, in actuality, most Baby Boomers have not. And certainly not Appalachian Baby Boomers.

Second, Millennials may already be on track (on average) to have larger private retirement accounts than Baby Boomers. We can gut check this fairly easily. Data from Fidelity’s “Building Financial Futures” 2025 Q3 report shows average 401(k) balances by generation:

Let’s do some envelope math and assume:

• The national median millennial salary is $59,800

• The current average millennial 401(k) balance of $80,700

• A 10% pre-tax retirement contribution

• A 50% employer match up to 3%

• A modest 6% average investment return (The S&P has averaged around 7% in inflation-adjusted return since 1957, and between 8% and 11% in the last decade)

• 3% annual inflation and a corresponding 3% cost-of-living adjustment to salary

You can plug these basic parameters (or more conservative ones) into a retirement calculator and see that, adjusted for inflation, this puts a millennial worker on track to retire at 65 years old with $326,000 of purchasing power in their 401(k), which is 22% more than the average Baby Boomer 401(k) has today. This allows for apples-to-apples comparisons of generations at the same stage in their economic lives. And, presumably, millennials would still expect to receive the same Social Security benefits as Baby Boomers on top of potentially larger private retirement balances.

Private retirement accounts aren’t a practical, or equitable solution

The debate offered by the pro-privatization and pro-benefits reduction crowd so far assumes that all Americans have equitable access to employee sponsored or private retirement accounts like 401(k)s. Most do not.

The 2020 Census found that only 58.1% of Baby Boomers have a private retirement account, compared with 56.1% of Gen X and 49.5% of Millennial workers. More alarming is the huge racial gap in private retirement account ownership. While 53% of white Americans reported owning a private retirement account, only 36.8% and 28.3% of Black and Hispanic Americans reported ownership.

When the lack of private retirement savings is paired with the decline in defined benefit pensions, maintaining and strengthening Social Security becomes even more important. Between 1980 and 2008, the percentage of Americans with defined benefit pension plans declined from 38% to 20%, according to data from the U.S. government, mirroring the decline in unionization.

In this light, it is easy to see the pushback on lifting the Social Security payroll tax cap for what it is: a modern manifestation of Reagan-era, neoliberal beltway policymaking that preaches that the government can do no good and only markets can solve our retirement woes.

But the promise of Social Security has helped Appalachians. Not enough, to be sure. But as the raw data show, more Appalachians rely on Social Security, and Social Security has delivered them more purchasing power with their benefits, adjusted for regionally lower costs of living, than wealthier areas of the country.

Most elderly Appalachians aren’t hoarding wealth in extravagant houses or giant 401(k) accounts. They have worked hard, paid their payroll taxes, and now their lives depend on their fixed incomes from Social Security. And it is likely that many future Americans without private retirement accounts, including Appalachians and other Americans from marginalized backgrounds, will depend on Social Security too. To risk automatic cuts of over 20% to avoid raising taxes on billionaires, millionaires, and high-income workers earning over $184,500 per year is bad policy.

By Sean O’Leary, Senior Researcher

Last December, I wrote a blog post titled, “The Appalachianization of America’s Economy”, in which I observed that the AI and data center boom, which has recently been a principal driver of overall GDP growth, seems to be giving rise to a recurrence of a phenomenon previously seen in Appalachia’s natural gas regions – a sharp increase in growth as measured by GDP accompanied by stagnation or outright declines in local measures of prosperity, including jobs, incomes, and population.

Recent data bear this out. Since 2005, real GDP in the US has grown by just over 50%, but the Information sector, driven largely by AI and data centers, has grown by almost three times. Yet, employment in the Information sector has actually declined.

This disconnect between GDP growth and growth in jobs not only undercuts the value of GDP as a measure of general prosperity, it also signals another problem – a rising share of income allocated to capital accompanied by a decline in income allocated to labor. This is an inevitable consequence of the growing importance of a capital-intensive and non-labor-intensive sector, such as Information, and it is an engine for growing income and wealth inequality.

Recent pieces by Harold Meyerson and Jared Bernstein, who provides the following chart, discuss this problem in greater detail.

But it’s also important to understand that growing income and wealth inequality doesn’t occur just along class lines. It occurs along geographic lines as well. As hyperscalers seek locations that offer access to abundant and inexpensive energy, communities in rural parts of Ohio, Pennsylvania, Indiana, Texas, and other states are seeing a proliferation of data centers that generate immense amounts of investment and revenue, all of which counts toward GDP, but because data centers are so non-labor-intensive, very little of the money that is invested or the revenue the facilities generate lands in local economies.

The most readily available remedy for this problem is taxes, which can generate large windfalls for host communities. That’s a key point in a recent piece by Jane Flegal, who suggests ways in which we can make data centers good citizens.

However, many communities and states are abating or providing exemptions for sales and property taxes in order to encourage data center development. Tragically, this ensures that data centers, as prolific in generating revenue as they might be, will provide negligible local economic benefit. As a result, many of these communities will likely fall into the same trap that many Appalachian communities did during the natural gas boom, when they experienced skyrocketing GDP growth accompanied by relentless declines in jobs, incomes, population, and overall prosperity.

By Sean O’Leary, Senior Researcher

Do you know the old advice to investors - “Buy low and sell high”? Well, building a new gas-fired power plant right now is doing exactly the opposite. This is the worst time in history to be paying for construction of a new natural gas-fired power plant. It’s buying when the cost of buying has never been higher.

But that’s exactly what the West Virginia Public Service Commission (PSC) is considering. Regulators are currently reviewing integrated resource plans, submitted by the state’s public utilities to determine what kinds of power plants get built, how much customers will pay, and what the state’s electric grid will look like for decades to come.

FirstEnergy and its subsidiary, Mon Power, are asking the West Virginia Public Service Commission to approve the expenditure of roughly $2.5 billion to build a 1,200 MW gas plant that would be capable of generating about the same amount of electricity as West Virginia’s Fort Martin coal-fired power plant. The cost could rise to $3.1 billion, if Mon Power fails to get a subsidized loan from the US Department of Energy. To put the cost in context, a report published in 2023 by the US Energy Information Administration found that the cost to build a gas plant of that size was, at the time, $1.04 billion. In just the last three years, amid feverish concerns about data center-driven electric system load growth, the cost of construction has nearly tripled. That cost will likely be lower in a few years, when load growth fever abates and supply chains catch up with demand.

In the meantime, prudence would dictate that, if we find ourselves in need of additional energy resources, we should thoroughly explore available alternatives before deciding to pay the highest price in history. Those alternatives include resources such as energy efficiency programs, demand response programs, and technologies that would integrate distributed and battery resources into the grid. In addition to being less expensive than adding gas-fired generation, these alternative resources are also more readily available.

Would these distributed resources, perhaps in combination with renewable resources, fully meet the need? Perhaps and perhaps not, but they would, if embraced, at least significantly reduce the need for added gas-fired generation, which would also save money. In any case, a truly thorough and rigorous quantitative analysis of multiple scenarios under a variety of conditions would make clear the most cost-effective way forward.

But that’s not what FirstEnergy and Mon Power are bringing to the table.These utilities have a powerful profit incentive to prefer solutions that require substantial spending over cheaper, more efficient alternatives. They don’t profit from the sale of the actual energy they generate, but rather, from returns on capital investments from building things like power lines, substations, and power plants.The bigger and the more expensive the power plant, the greater the utility’s profits. The only constraint on the utility’s ability to invest and maximize profits is the state Public Service Commission, which must approve all capital expenditures for which the utility will be allowed to bill customers.

The PSC has multiple mandates, including cost-effectiveness. It also cares about grid reliability and resiliency and even the economic impacts of the choices made by the utilities it regulates. That fact should make the challenge of justifying a new gas-fired power plant even greater because, from an economic development perspective, a plant of the size proposed will probably employ fewer than thirty people. At least that’s the case with similarly sized, nearby gas plants in Ohio and Pennsylvania.

But most concerning is the possibility floated by FirstEnergy and Mon Power that the new plant is primarily needed to serve prospective data centers. That’s frightening because in Pennsylvania and Ohio, where data center development is much more advanced, utilities are shifting some of the costs induced by data centers to customers,who are now suffering double-digit rate increases.

This effect can be mitigated if utilities and their regulators adopt special tariffs or rate structures for data centers to ensure that they will pay not only for the energy they use but also for the infrastructure they require. These regulatory measures could also address the risk data centers impose on other ratepayers if they fail to materialize or operate below expected levels. These protections can reduce the need to recover costs from existing customers in the likely scenario that demand for data centers is overstated.

FirstEnergy CEO Brian Tierney told investors last fall that his companies will not seek special tariffs for data centers, claiming that existing customers’ interests will be protected by means of special contracts between FirstEnergy and data center owners. But FirstEnergy is a profit-motivated player, and the consideration of special contracts is a particularly opaque process. That makes the PSC even more reliant than usual on the utility to determine what is fair and unfair.

Pennsylvania’s Public Utility Commission is developing a model tariff for utilities and the legislature is considering requiring utilities to adopt special tariffs for data centers and similar large load customers. In the interests of rate payers and a resilient and reliable grid, West Virginia should do the same.

By Eric Dixon, Senior Researcher

In January Congress passed a “minibus” bill that raided $500 million in previously appropriated coal mine cleanup funds to pay for other federal programs. We’re now seeing the first results of that bill: $45.5 million less in mine cleanup funding every year for the next 11 years. Combined with growing inflation, this means fewer jobs will be supported cleaning up mines and more hazardous coal mining damage won’t be reclaimed in Appalachia and across the country.

When it passed in 2021, the Bipartisan Infrastructure Law provided about $10.9 billion for the reclamation of Abandoned Mine Land (AML) sites across the country in fifteen annual grants to states and tribes. The first four years’ worth of grants were awarded between 2022-2025. The minibus bill cuts $500 million from the total AML funding provided under the Bipartisan Infrastructure Law – but it was unclear at the time of passage if the $500 million would be cut entirely from the last (fifteenth) year of AML grants or equally across the remaining 11 years worth of annual funding. Now we have our answer.

The 2026 AML grants for states and tribes were announced in May and the cuts are here. According to the Office of Surface Mining Reclamation and Enforcement, the $500 million cut “will be applied equally to the remaining 11 grant distribution years, approximately $45.45 million per year.” The figure below shows the annual reduction in funds for each state and tribe, as well as the total cuts that will play out over the next 11 years. Pennsylvania and West Virginia have the largest cuts (by absolute value), at about $15 million and $9 million per year, respectively.

The cuts are reducing the amount of damage states and tribes can reclaim. Inflation – especially recent rises in fuel costs that can drive up the cost of operating construction equipment – is also lowering the spending power of reclamation dollars even from last year, further reducing the amount of reclamation states and tribes can accomplish in 2026.

States and tribes have a five-year window to spend their FY2026 AML grant, and those agencies will now begin planning for fewer dollars by taking steps like selecting fewer reclamation projects or reducing the scope of projects. In Pennsylvania, for example, the cuts are equivalent to the cost of two large abandoned mine drainage treatment systems.

As we explained in a previous post, the extent of AML damage that needs to be cleaned up is likely twice as large as the existing $10.9 billion in funding– even before $500 million was cut.

This is damage to land and water that has lingered since at least the 1970s, and now residents will have to wait even longer for cleanup. If this cut hadn’t occurred, $45 million per year in more mine cleanup would be put to use across the country in the next few years, removing hazards to the local population and supporting more jobs, such as in construction, doing reclamation work primarily in rural areas. Congress should reverse the $500 million reduction, and should protect the program from similar cuts in the future.

Last week, the Department of Energy finally announced the results of its eleven-month review of over 2,200 DOE-funded projects. According to an accompanying spreadsheet shared with Congress by Energy Secretary Wright and first published by Latitude Media, the agency intends to retain or modify its funding agreement with the Appalachian hydrogen hub, also known as ARCH2. No information is available regarding what this designation means but a spokesperson for the hub has since confirmed the news.

Modification of the hub could involve revising the amount of funding awarded to the project or changes to the cooperative agreement outlining the hub’s plans. A recent proposal from the agency points to another potential modification to the hub’s plan, one that is in line with trends in the Ohio River Valley.

Earlier this month, the Department of Energy shared a budget request with Congress that included a proposal to reallocate funds from the Regional Clean Hydrogen Hubs program for use in a new baseload power initiative. This effort seeks to support and expand gas, coal, and nuclear generation to help meet expected demand for data centers and would likely pull from funding previously awarded to three now-canceled projects: ARCHES, the California-based hydrogen hub; PNWH2, the hydrogen hub in the Pacific Northwest; and a demand-side initiative to drive commercial offtake of hydrogen.

While there is no indication that the Appalachian hydrogen hub will be modified to reflect the current push for data centers, four current and former hub developers are already exploring this market. If, on the other hand, the agency decides to retain the funding agreement as currently written, project backers face a difficult road.

The hydrogen hub has all but dissipated

At the time of writing, no more than three percent of the hub’s planned hydrogen production appears to be in active development. Plus, impacts to agency capacity and the uncertainty experienced by prospective hydrogen developers — partly a reflection of the agency’s lukewarm stance on low carbon hydrogen technology — are likely to discourage many developers from making meaningful commitments.

We’ve long maintained that the hydrogen hub was unlikely to materialize, especially at the scale that its supporters promise. Despite the availability of lucrative tax credits and significant political support in our region and, previously, at the federal level, the developers driving the hub would need to overcome high costs and a lack of demand for low carbon hydrogen in order to get a foothold in most of the sectors these projects were targeting. Even the boosters pitching the technology acknowledge that existing subsidies are not enough to make hydrogen viable for most applications, which is why all seven hydrogen hubs asked the Department of Treasury for more funding just four months after receiving their federal designation.

When the application for ARCH2 was approved, eight hydrogen producers were set to produce more than 2,100 tons of low carbon hydrogen per day. However, since the announcement of the award, CNX suspended its project over a year ago, EQT indicated that its project is still under evaluation, and Fidelis New Energy recently asked regulators in West Virginia to suspend review of an air permit application for its hydrogen facility. These three companies likely would have been responsible for an estimated 97% of the hub’s planned production, meaning that almost none of ARCH2’s planned production is in active development.

A fourth producer, Keystate Natural Gas Synthesis, would likely contribute an additional 5%. However, the company has yet to secure enough funding for a final investment decision on its hydrogen facility and it has also pushed back its operational date by five years. The remaining four producers, including the struggling Empire Diversified Energy, would likely produce a negligible amount of hydrogen.

New headwinds for hydrogen

In the fifteen months since the Trump Administration returned to the White House, the future for hydrogen production has only become more challenging. Early executive actions, like the pause on IRA/BIL spending and the termination of equity-related initiatives, significantly affected hub activities. The impact of these decisions can be seen in the federal spending record, which indicates that only about half a percent of the hydrogen hub’s $925 million award has been spent.

Since then, the hydrogen hub has had to contend with a substantial cut to the hydrogen tax credit, reported threats of cancellation, a missed deadline in the hub’s environmental review, and the termination of the office responsible for managing the hydrogen hub program. Meanwhile, other regions have seen numerous natural gas-powered hydrogen projects cancelled or suspended.

Treat everything as if it were a nail

Amid these mounting setbacks, the Appalachian hydrogen hub has not shared a meaningful project update in over a year and a half. In reality, however, what the hub wants to do doesn’t really matter.

As discussed above, the DOE funding is unlikely to meaningfully improve the hub’s viability and political interest has now, regrettably, turned towards advancing large natural gas-powered data centers, a gravity sink that has managed to attract five current and former hub developers: CNX Resources, EQT, Fidelis New Energy, Hog Lick Aggregates, and TransGas Development Systems.

While this shift may reflect a loss for the hydrogen hub, it points to the success of the real underlying project. Whether it was the failed petrochemical renaissance, the Appalachian Storage Hub that never materialized, a hydrogen hub that fell apart almost as soon as it was announced, or the current scramble to build large data centers, these concepts are evidence of a persistent — and pernicious — interest: the continued extraction and use of natural gas in Appalachia. In this way, rather than interrupt the momentum behind the hydrogen hub, the emergence of data centers is only the latest iteration of a long-running effort to preserve the natural gas industry.

As early as sixteen months prior to the launch of the hydrogen hub program, numerous entities were already pitching natural gas-powered hydrogen proposals. This left little room for genuine clean energy alternatives that could reduce industrial carbon emissions — the stated objective of the Appalachian hydrogen hub — while also protecting public health and the environment and minimizing costs to ratepayers and taxpayers.

The inability to look at the Ohio River Valley as anything other than a site of extraction and pollution reflects a profound lack of political will. That the hydrogen hub has all but evaporated should be cause for relief. Instead, it appears that the project’s leadership is set to begin anew, locking us into another cycle of contending with Appalachia’s fossil fuel problem when we could be building a brighter future for the people who live here.

This is the third article in a series about the past, present, and future of ARCH2.

Earlier this month, the Department of Energy announced plans to retain or modify funding for the Appalachian hydrogen hub, also known as ARCH2. Using carbon capture and storage to mitigate emissions in hydrogen derived from natural gas, ARCH2 was intended to jumpstart a regional decarbonization effort spanning Ohio, Pennsylvania, and West Virginia. The Appalachian hydrogen hub was one of seven federally-designated projects allocated funding to support a network of hydrogen producers, end-users, and associated infrastructure and this decision was met with widespread praise from our region’s public officials. Now, a full three years after first being announced, what is there to show for all this pomp and circumstance?

Setting the record straight

Consisting originally of sixteen projects, ARCH2 underwent significant changes in the nine months between being designated a federal hydrogen hub and receiving its first tranche of funds. These changes included the loss of two prospective customers (Chemours and First Mode), one midstream partner (MPLX), and one hydrogen producer (TC Energy). Another developer included in the initial announcement (Clearway Energy Group, likely a hydrogen producer) had already left the hub by the time the Department of Energy approved the project. When the final slate of projects was announced in 2024, ARCH2 consisted of twelve distinct projects: eight hydrogen producers, three potential customers, and one processing/storage facility.

While this may seem like a robust set of projects, in reality, production appears to be mostly concentrated in just three projects, making the hub’s viability contingent on the success of a handful of companies. Now, based on the available information, it appears that the hydrogen hub has all but collapsed. Despite this, however, our region’s leaders continue to champion the project, claiming that the project will bring thousands of jobs to our region.

Core challenges

According to ARCH2’s website, the hydrogen hub intends to produce more than 2,100 metric tons of hydrogen per day, an amount equivalent to roughly eight percent of all the hydrogen produced in the US today. Unlike conventional methods of hydrogen production, ARCH2 plans to capture CO2 emissions during the production process and store the carbon in deep geologic formations for hundreds or possibly thousands of years. The resulting low-carbon hydrogen would then be used to power industrial processes that currently rely on natural gas or other fossil fuels. However, the economics of the project raise serious concerns about the viability of the hydrogen hub.

Tax credits offered by the federal government do not appear to be sufficient to cover the cost of decarbonizing hydrogen and ARCH2 has failed to demonstrate why current and future offtakers would pay more for lower-carbon hydrogen than they would for conventional hydrogen. Unless the federal government enacts a tax on carbon or a similar mandate that raises the price of conventional hydrogen — something that is extremely unlikely — low carbon hydrogen derived from natural gas will not be cost competitive. A proposal from the Biden administration requiring most fossil fuel power plants to switch to hydrogen fuel or employ carbon capture technology might have led to increased adoption of low carbon hydrogen in the power sector but this rule is expected to be repealed.

The hub’s costly hydrogen would also be competing against decarbonization solutions that are not only proven but cheaper. Hydrogen is incredibly uneconomic for most applications when compared to increased energy efficiency, scaling up renewables, and widespread electrification. Technologies like carbon capture and hydrogen could be useful in niche, hard-to-electrify sectors like steelmaking and cement production. However, ARCH2 has floated the use of hydrogen in sectors such as residential heating and power generation where hydrogen is simply unable to compete with alternatives. Any success at building a customer base in these sectors will likely come at the expense of taxpayers, ratepayers, or both and even then these subsidies might not be enough to make these projects viable.

Production

Considering the challenges facing the hydrogen hub, it’s no surprise that its development has been rocky. In the absence of more detailed and updated information from the hub itself, we have been able to piece together a rough approximation of the hydrogen hub’s status using a combination of industry statements, public testimony, and agency records. While the hub’s leadership has sought to dismiss concerns regarding the changes to its slate of projects, it appears that the bulk of the hub’s production is at risk.

As noted earlier, ARCH2’s slate of projects includes eight hydrogen producers. However, only a few of these projects intend to produce meaningful amounts of hydrogen. Neither Enbridge nor Independence Hydrogen have released production goals for their hydrogen facilities but a statement from a hub representative indicates that these two projects will produce a combined one ton of hydrogen per day, or less than a tenth of a percent of ARCH2’s production goal. West Virginia-based Empire Diversified Energy, which has been experiencing financial issues, plans to produce about five tons of hydrogen per day, totaling less than half a percent of the hub’s goal. A fourth producer, KeyState Energy, plans to produce 53 tons per day, or about two and a half percent of the hub’s goal. Few details are available regarding Plug Power’s proposed hydrogen facility in West Virginia but this project’s contributions are also likely to be small. Based on this information, the vast majority of ARCH2’s planned production would have to come from the remaining three producers.

Fidelis New Energy, one of these likely anchor producers, plans to produce 640 tons per day, or thirty percent of the hub’s target production. This leaves potentially two thirds of ARCH2’s hydrogen in the hands of EQT and CNX. While neither company ever disclosed a target production goal, a large-scale hydrogen facility makes sense considering the profile of these companies. Plus, the size of CNX’s original hub project — a partnership with TransGas Development Systems — is consistent with a project of this magnitude.

That so much capacity could be concentrated in just a handful of producers poses a real risk to the hub. The salience of this became apparent after CNX announced that it had suspended its ARCH2 project and is no longer participating in the hub. The impact of this departure was seen a little more than a month later when a hub representative told the Ohio legislature that the hub’s slate at that time would produce “more than 1,000 tons of hydrogen a day,” less than half of the more than 2,100 tons per day initially targeted by the hub. CNX’s footprint was likely to be significant but this sharp drop in production likely also reflects the status of a second anchor producer: EQT never confirmed its hydrogen project and recently disclosed to journalists that it was still evaluating its project.

In short, one key producer dropped out of the hub and the status of a second is still in question, which leaves the bulk of ARCH2’s production capacity in the hands of Fidelis New Energy. Unfortunately for ARCH2, Fidelis New Energy recently asked state regulators to suspend review of the company’s proposed hydrogen production facility in Mason County, West Virginia. Based on the available evidence, it now appears that very little — perhaps three percent — of ARCH2’s planned hydrogen production is in active development.

Offtake

The hub has also experienced issues attracting and retaining prospective customers to the project. A central feature of the hub concept is that it is supposed to be a network of both hydrogen producers and hydrogen consumers. Unfortunately for the hub, two of the five end-users announced as part of the initial slate — Chemours and First Mode — have since left the network. A representative from Chemours noted that the departure was due to “a number of factors” but the company also referenced problems related to the federal hydrogen tax credit, likely referring to the high cost of hydrogen under the existing regulatory scheme.

Of the three remaining customers, Hog Lick Aggregates intends to only use one ton of hydrogen per day, a marginal amount. The second remaining customer, Hope Gas, has not disclosed the amount of hydrogen it plans to purchase but it’s worth noting that Hope Gas intends to blend hydrogen into its residential gas distribution system at levels up to 15%, an application that is widely understood to be cost-ineffective. Plug Power, one of ARCH2’s eight producers, is also listed as having an on-road mobility project associated with two distribution centers in Ohio. Finally, SARTA, the transit authority in Stark County, Ohio, is expected to purchase at least some of the hydrogen produced by Enbridge but for unspecified reasons ARCH2 does not list it as a development partner.

No customers for the hydrogen produced by CNX or EQT were ever announced. Fidelis’s original announcement included plans to develop a data center powered by hydrogen but this project has since been sold and the available permitting information indicates that this facility plans to rely on natural gas.

Conclusion

The hub’s leaders appear to be treating the reinstatement of their funding allocation as a second chance. However, the impacts on ARCH2’s production capacity and the lack of prospective customers means that the Appalachian hydrogen hub has a serious rebuilding effort ahead of it, an endeavor that will take place within a hostile and uncertain market. At the same time, the nearly $1 billion that the federal government has earmarked for ARCH2 represents only about a tenth of the capital required for a project of the scale originally announced.

If our region’s public officials were serious about cleaning up the heavy industry in our region, they would look to cheaper, proven, and more effective methods to reduce carbon emissions, such as energy efficiency, ramping up renewables, and expanding electrification. Investing in these strategies will allow us to apply costly hydrogen and carbon capture to the niche applications where they’re actually needed. Considering the project’s troubled history, it’s difficult to see the case for pursuing the Appalachian hydrogen hub, let alone spending nearly a billion dollars of public funds on its execution. Thankfully, only a small amount of the federal funding — about half a percent — has actually been paid out to the hub, meaning that we can still avoid the worst outcome with little lost other than time.

This is the second article in a series about the past, present, and future of ARCH2.

In 2022, the Department of Energy launched the $8 billion Regional Clean Hydrogen Hub initiative with the goal of accelerating commercial deployment of clean hydrogen. This program later allocated $925 million to the Appalachian hydrogen hub, also known as ARCH2. Almost all of the hydrogen produced by ARCH2, however, would be created from natural gas, a feedstock that is anything but clean. How did this happen?

Innocuous Origins

To understand how we got to this point, we have to go back to the statutory language guiding the creation of the hydrogen hubs. The Regional Clean Hydrogen Hubs program was created by the Infrastructure Investment and Jobs Act, also known as the Bipartisan Infrastructure Law (BIL). Now codified at 42 U.S. Code § 16161a, this statute directs the Department of Energy to establish a program to support the development of at least four clean hydrogen hubs.

These hydrogen hubs were expected to “develop a national clean hydrogen network to facilitate a clean hydrogen economy.” The language in the statute encouraged diversity in the feedstock used to produce this hydrogen, the end-uses planned for this hydrogen, and the regions where these hubs would be located. The statute also called for at least two hubs “in the regions of the United States with the greatest natural gas resources” and allowed for the consideration of hydrogen produced from natural gas in its definition of clean hydrogen. This section of the BIL was “inspired” by H.R. 4904, a bill introduced by Representatives Brian Fitzpatrick (R-PA), Mike Doyle (D-PA), and Conor Lamb (D-PA).

H.R. 4904, titled the Clean Hydrogen Energy Act, likewise called for the creation of a clean hydrogen standard and the development of regional clean hydrogen hubs. However, the bill summary and the accompanying press release both reference the importance of supporting the commercialization of electrolysis and neither directly references the use of natural gas for the production of hydrogen aside from the possibility of “utilizing existing infrastructure and fuel supplies.” Even more notable is the fact that the only direct reference to natural gas in either the bill summary or the press release is through an unfavorable comparison with the combustion of hydrogen, which produces only water and NOx in contrast to the greenhouse gas emissions of burning fossil fuels. This is a curious choice from three members of Congress representing the second-largest natural gas-producing state in the U.S.

Like the BIL, H.R. 4904 incorporates fossil fuels in a number of ways but the presentation of the bill invites a reading where Appalachia, a region dominated by fossil fuels, could receive a massive infusion of federal funding to jumpstart a regional network of clean electricity and clean hydrogen production. It’s possible that the framing of the legislation was not intentional or, if it was, that it was done so in order to present yet another fossil fuel subsidy as something that would fix the climate crisis rather than deepen it. But this is an easy conclusion to draw now, years after the hydrogen hub program launched and our region was set to be saddled with even more pollution infrastructure.

Stacking the Deck

At the time, it was not a foregone conclusion that Appalachia was to receive a natural gas-powered hydrogen hub but many forces in our region were working hard to make it so. Throughout the course of 2021 and 2022, the Ohio River Valley was the subject of growing calls to enlist the region’s vast natural gas and geologic resources in the struggle to decarbonize heavy emitters in the industrial sector:

• Representatives from EQT, AFL-CIO and others met in May 2021 to discuss industrial decarbonization strategies, including “the conversion of natural gas being produced in Appalachia into hydrogen”

• Energy Futures Initiative and the American Federation of Labor and Congress of Industrial Organizations hosted a workshop in July 2021 focused on “the value of hydrogen and CCS market formation” in the Ohio River Valley

• MIT’s Roosevelt Project released a draft report in October 2021 recommending, among other actions, the development of hydrogen and carbon capture and storage “to continue the use of existing fossil fuels in ways consistent with the rising demand for low greenhouse gas production”

• Battelle and Equinor, companies that would later go on to lead competing hydrogen hub proposals, signed a memorandum of understanding in February 2022 “to examine the regional potential for carbon capture and storage”

• EQT, Equinor, Marathon Petroleum, Mitsubishi Power, Shell Polymers, and U.S. Steel — all companies that would later be affiliated with hydrogen hub proposals — formed an “alliance” in February 2022 to pursue “a shared vision for a low-carbon and hydrogen industrial hub” in Ohio, Pennsylvania, and West Virginia

• National Energy Technology Laboratory released a report in March 2022 assessing “the potential for the Appalachian region to develop a hydrogen economy with fossil-derived hydrogen with CCS production”

What these initiatives all had in common was the understanding that the sources of our region’s industrial pollution were not liabilities but assets and that Appalachia’s natural gas industry could continue to operate business-as-usual.

To be sure, industrial activities are a significant source of greenhouse gas emissions in Appalachia and our policymakers have yet to invest in the interventions needed to reduce or eliminate the climate footprint of this sector. However, the scope of responses to this problem has been limited by the misguided expectation that the existing fossil fuel industry is not a central contributor to the problem but the key to solving it.

These boosters and beneficiaries largely sought to use natural gas to produce hydrogen, which could then be substituted in place of fossil fuels to limit carbon emissions in applications as diverse as power generation, heavy industry, transportation, and residential heating. This hydrogen production would be paired with technology that its supporters hope will allow operators to capture their carbon emissions and transport them to facilities where they can be stored in deep geologic formations for hundreds or thousands of years.

Time and time again, the response from policymakers and industry to the immense sources of pollution in the Ohio River Valley and the immiseration resulting from the capital-intensive but not labor-intensive industries we’re saddled with entails only minor adjustments that maintain our region’s landscape of fossil fuel infrastructure and the associated health impacts.

A Headless Blunder

Momentum on its own was not enough to call into existence a proposal like the Appalachian hydrogen hub. These interest groups also required the support of the people whose hands were on the levers, a role these functionaries, it turns out, were more than comfortable playing.

To oversee the hydrogen hub program, the Department of Energy created the Office of Clean Energy Demonstrations, also known as OCED. Founded in 2021 and authorized by the Bipartisan Infrastructure Law, OCED’s mission was to deliver demonstration projects at scale in partnership with the private sector. The $8 billion hydrogen hub program was a significant part of OCED’s $62 billion portfolio.

Through the funding opportunity announcement and the resulting cooperative agreement with each hydrogen hub, OCED had “shared responsibility” for the success of each hub and was expected to have “substantial involvement” in the hubs’ implementation. At the same time, OCED was expected to evaluate the progress of each hub and decide whether the projects could continue to receive further allocations of funding. This placed OCED staff in an awkward position where any interest in disciplining the hubs had to be tempered with the expectation that OCED was as responsible for the success of the project as the hub backers themselves. In this way, OCED was more like a project partner or industry consultant than a regulator, a dynamic that became apparent during conversations with advocates and community groups in Appalachia.

The first signs of trouble could be seen in OCED’s treatment of the statutory language. The section of the Bipartisan Infrastructure Law involving the hydrogen hubs included criteria that called for at least one hub that would produce hydrogen from natural gas. The BIL also called for “at least 2 regional clean hydrogen hubs… in the regions of the United States with the greatest natural gas resources.” However, this criteria was qualified with statutory language that these conditions should be established “to the maximum extent practicable.” In other words, these criteria were intended to be a guide, not a mandate. Similarly, the statute did not suggest that these gas-producing regions should host a hub producing hydrogen from natural gas.

Unfortunately, OCED interpreted this language in the narrowest possible sense and at least four of the seven designated hydrogen hubs relied heavily on hydrogen produced using fossil fuels, including the two hubs located in the largest natural-gas producing regions of the country: HyVelocity in the Gulf South and ARCH2 in Appalachia.

At the same time, OCED failed to take advantage of the expansive discretion it was imbued with through the Bipartisan Infrastructure Law. Section (c)(4) of the authorizing statute allows the agency to “take into consideration other criteria that… are necessary or appropriate to carry out this subchapter.” At first glance, it might seem that OCED’s use of community benefit plans in the hydrogen hub program might be an example of this broad discretion. These plans were intended to “support meaningful community and labor engagement” and “[a]dvance diversity, equity, inclusion, and accessibility.” However, the plans were actually a response to the now-rescinded Executive Order 14008, which sought to ensure that “40% of the overall benefits of certain federal investments flow to disadvantaged communities.”

The community benefit plan for ARCH2, as implemented by OCED, failed to set any meaningful equity, engagement, or environmental justice requirements for the hydrogen hub. Despite the agency’s emphasis on “two-way engagement,” the agency did not mandate that applicants consult with local communities or even disclose information about their projects prior to submitting their applications. This hands-off approach allowed ARCH2 to require that interested parties sign a non-disclosure agreement prior to divulging any project details with seemingly no repercussions.

A more courageous OCED might have evaluated ARCH2 on the basis of the pollution and local harms associated with its planned activities or used the $925 million funding award to discipline the hub. Instead, OCED chose to do the easiest thing, which was to do nothing at all and let events unfold along the well-worn path carved out by the natural gas industry and its enablers.

The Unnatural World

When announcing the slate of designated hydrogen hubs, a key agency leader remarked that federal officials were “once again coming to frontline communities to ask them to bear the burdens of the clean energy transition simply because so much of the existing energy infrastructure already exists in those communities.” This admission mirrored the strained logic of the boosters that spent the months leading up to the passage of the Bipartisan Infrastructure Law making the case for CCS and natural gas-powered hydrogen — the argument that since natural gas is so damaging to the climate we must use it to produce hydrogen.

This sleight of hand, now mimicked by OCED, relies on the abundance of natural gas in our region and the problems created by its continued extraction to naturalize their chosen strategy. These actions foreclose on any alternatives that don’t accept the starting principle that pollution is Appalachia’s status quo and must be maintained at any cost. We cannot forget that these conditions do not arise naturally from our region’s physical or geologic characteristics; instead, it is people that made this so and people that can act to make it otherwise.

This is the first article in a series about the past, present, and future of ARCH2.

By Julia Stone and Eric de Place

This analysis is an expanded and updated version of an earlier presentation to the Ohio River Valley Institute.

A new report describes the exponential rise in the development of gas-fired power in the US, primarily for data centers. According to Global Energy Monitor, the US nearly tripled its gas-fired capacity in development in 2024, totalling almost 252 gigawatts (GW), a volume roughly equivalent to the sum of plans for China, Vietnam, Taiwan, and South Korea. If all in-development plants are built, the US’s existing gas fleet would grow by nearly 50%. More than one-third of gas-fired power generation capacity under development in the US is slated to serve data centers on site, or “behind-the-meter.”

Data center expansion is one of the single-most important drivers of new gas demand in the US, and is evidenced by a new trend of Big Tech entering long-term contracts with gas companies to supply on-site power generation. The top five hyperscalers announced plans to spend $700 billion in 2026, driven by AI and data center development. This development means significant new demand for electricity, much of which will be supplied by gas. Though exact figures on data center expansion are difficult to nail down, several long-term agreements have made national headlines.

Much of this growth is slated for Appalachia: 8.8 GW of on-site (behind-the-meter) gas-fired power generation is at various stages of development in Pennsylvania, all of which would be fueled by Appalachian gas.

Data center growth in other parts of the country could also mean increased gas production in Appalachia. For example, many of the new gas power plants coming online in Virginia, North Carolina, South Carolina, Georgia, and Alabama to replace older plants (including coal-fired ones) could be serviced by gas produced in Appalachia thanks to new pipeline capacity under development (i.e. MVP and several projects along or near Transco and Tennessee Gas Pipeline systems).

Global Energy Monitor’s findings mirror other recent studies, including a new report released by Cleanview in January 2025. According to Cleanview’s study, more than 50 GW of behind-the-meter projects were announced in 2025, including 7.5 GW in Pennsylvania, 2 GW in West Virginia, and 1.1 GW in Ohio.

Long-term growth

According to the Energy Information Agency’s latest Annual Energy Outlook, Appalachian gas production is expected to increase substantially from 37 Bcf/d in 2025 to between 66 Bcf/d and 73 Bcf/d by 2050, between 2024 and 2050. Appalachian gas could account for nearly half of all natural gas produced in the US. This Appalachian gas production would increasingly meet gas demand across the US in the coming decades, including the massive LNG buildout and ensuing demand growth along the Gulf Coast. EIA describes several reasons for this demand growth.

Associated gas production linked to crude oil in the Permian is expected to decline over the same period. The Haynesville basin, where production has grown recently thanks to its proximity to the Gulf’s LNG terminals, is deeper and will be increasingly more expensive to drill, limiting its longer-term outlook.

EIA expects that as Appalachian gas production increases and as gas prices continue to rise along the Gulf, significantly more gas will flow between the Mid-Atlantic and Ohio region (Appalachia production zones) to the Midwest. At the same time, the volume of gas moving from the Midwest region to South Central increases significantly as it becomes more favorable for midstream companies to invest in pipelines and associated infrastructure to transport cheaper Appalachian gas to the Gulf.

This illustrates how Appalachian gas will increasingly flow south through the Midwest to meet LNG demand on the growth, as well as growing gas demand centers in the Southeast. The dramatic increase of gas flows from Appalachia into the Midwest and from the Midwest into South Central will necessitate pipeline development to increase takeaway capacity out of Appalachia.

LNG capacity ramps up

LNG exports are contributing to all-time high residential electricity rates and LNG capacity is set to grow even more. Three new LNG export facilities are driving this year’s growth: Plaquemines LNG, Corpus Christi Stage 3, and Golden Pass LNG. Meanwhile, five additional LNG export projects in the US reached financial investment decisions (FID) and are currently under construction: Golden Pass LNGPort Arthur LNG, Rio Grande LNG, Woodside Louisiana LNG, and CP2 LNG. LNG export capacity could reach as much as 21.1 Bcf/d (25.2 Bcf/d peak) by 2028 from just the first three projects.

At the same time, top Appalachian gas producers, including EQT and Wjlliams, have secured contracts with LNG terminals on the Gulf to buy LNG and sell it directly to customers in Europe and Asia. After buying cargoes from the terminals, the companies can sell them to Europe and Asia at higher prices than they can get in the US. These purchase agreements represent a major expansion for these gas companies, allowing them to control more of the supply chain and to capture the price spread between US domestic gas and international prices.

Capital spending by the 13 largest midstream companies rose 14%in 2024 and 20% in 2025, representing a rebound in midstream investment followed by a period of restraint after 2020, as reported by industry analyst RBN Energy. Most of the gas supply for LNG terminals along the Gulf will come from the Permian and Haynesville basins in the near term; however, Appalachian gas will meet much of the demand growth from power generation and data centers.

Pipeline buildout meets sluggish growth

Appalachian gas production has been hovering between 34 and 36 billion cubic feet per day (Bcf/d) since 2020 due to constraints from limited pipeline takeaway capacity. Since then, however, Appalachian gas production grew 2.1 Bcf/d from 2024 to 2026, largely thanks to the June 2024 startup of the Mountain Valley Pipeline (MVP). MVP directed up to 2.0 Bcf/d of gas from northwest West Virginia toward the Southeast and now two MVP expansion projects are in development: the MVP Boost, which would increase the mainline capacity to 2.6 Bcf/d, and the MVP Southgate project, which would extend the MVP mainline from its terminus in Pittsylvania County, Virginia to new interconnections in Rockingham County, North Carolina.

Pipeline projects are also slated to expand capacity and de-bottlenecking interstate corridors. Examples include a slew of expansions along the Transco interstate pipeline, the Appalachia Reliability project along the Eastern Gas Transmission system, and a series of projects along the Texas Gas Transmission system.

Boardwalk’s Borealis pipeline would extend the Texas Gas Transmission (TGT) pipeline across Ohio, expanding Appalachian gas producers’ access to customers in the Midwest and Gulf Coast. The project is designed to work in concert with Boardwalk’s expansion plans elsewhere on the TGT system, an interstate pipeline that spans eight states from Ohio to Louisiana. FERC is reviewing the Kosci Junction project, a 1.6 Bcf/d capacity pipeline that would extend the TGT system across Mississippi, terminating at Transco Station 85. If Boardwalk completes both the Borealis and the Kosci Junction projects, it would be the first time the Clarington hub in Appalachia would directly connect to the premium Station 85 market, on the border of Mississippi and Alabama.

Near-term uncertainty

Even though national trends indicate that natural gas production in Appalachia is likely to grow in the coming decades, the immediate outlook is a murkier prospect.

Nearly every major player in the Marcellus and Utica gas basin discussed expectations for higher gas demand from new power generation in and near their production areas, with the majority of anticipated power generation being tied to planned data centers.

For example, Appalachia’s largest producer, EQT, will supply 665 MMcf/d of gas to a planned 4.4 GW power station in Homer City, and 800 MMcf/d to a 3.6 GW power plant in Shippingport, PA, each of which will support large data center campuses. NextEra Energy and Google announced plans to partner in multiple sites across the US to scale hyperscale data center campuses with on-site power generation. Likewise, Williams is developing three major gas plants in Ohio to fuel Meta’s hyperscale complexes.

At the same time, EIA predicts Appalachian production will see modest growth in the near term, rising 0.4 Bcf/d in 2026 and 0.7 Bcf/d in 2027. EIA’s latest Short-Term Energy Outlook does not adequately account for the emerging trend of recent data center-related increase in planned gas-fired power capacity but it’s unlikely that all of the proposed data centers in the US will move forward, making it difficult to understand the short-term impact on gas production in Appalachia.

A recent analysis from Aurora Energy Research determined that potentially two thirds of announced data center capacity was “speculative”, meaning that these developers have yet to secure site control, obtain permits, or don’t have active interconnection requests. The bulk of the remaining planned capacity is associated with “newcomers”, or companies with no current operational capacity, which calls into question just how much of the new gas-fired power capacity assessed by the Global Energy Monitor will actually materialize.

By Zane Gustafson, Research Fellow

New data centers are being built quickly in West Virginia. While residents are speaking out and pushing back, many communities don’t have a real say in the decisions that could impact them. These projects would use a staggering amount of fracked gas and could harm their health and environment.

A Massive Project in a Quiet Mountain Town

Tucker County, located just north of the Monongahela National Forest in the heart of the Appalachian Mountains, is a tourist destination for those seeking outdoor recreation and natural scenery. It is also the site of a proposal to build a gigantic, off-grid, gas-powered, hyperscale 1.6 Gigawatt (GW) data center. Called the Ridgeline Facility, this data center would consume more than 10 million Megawatthours (MWh) of electricity each year, the same amount of electricity consumed by all West Virginia residential customers combined.1

This massive data center would be powered by a newly built and equally large 1.6 GW fossil gas power plant. This power plant would provide “behind-the-meter,” or off-grid, electricity to the data center. Unlike most “microgrids” (see the analysis of HB 2014 below) which will have limited connection to the rest of the system, the Ridgeline Facility is expected to be completely separate from the electric grid. Being disconnected from the grid could imperil the viability of the project, because the isolation will limit revenue opportunities and increase costs needed to guarantee independent reliability.

Despite the massive size of the project, most residents of Tucker County were completely unaware of the proposal until March 2025, just as state legislators were fast-tracking a bill to promote data center development. Community members began to organize into the group that became Tucker United, but before they could get answers about the proposed facility, West Virginia’s state legislature passed House Bill 2014.

A New Law Limits Local Control

West Virginia Gov. Morrisey signed HB 2014, also known as the “Power Generation and Consumption Act,” on April 30, 2025, and touted it as making “West Virginia the most attractive state in the country for data centers… Combined with the one-stop shop permitting bill [HB 2002], companies will now be able to quickly build” data centers and their associated power plants.

In summary, HB 2014 legalized “microgrids” that produce and consume electricity separately from the rest of the electric grid. “Microgrids” are allowed to connect to the grid, but no more than 10% of their electricity is permitted to be delivered outside the microgrid district. Section 5B-2-21b of HB 2014 prohibits “[c]ounties and municipalities from imposing or enforcing local laws and ordinances concerning the creation or regulation of any certified microgrid district or certified high impact data center therein.” Further, Section 11-6N-4 requires that the state redistribute just 30% of incremental property tax revenue (new revenue derived from the higher value of the property following the construction of the data center) back to the county in which the data center is sited. Normally, counties in West Virginia retain only slightly less than 100% of the property taxes they collect. Considering the scale of these new data centers, this provision will cause county schools and offices to miss out on millions of dollars in annual revenue they would otherwise receive, according to analysis by the West Virginia Center on Budget and Policy. Moreover, half of the property taxes that data centers pay will be used to fund reductions in West Virginia’s state income tax, which will mostly benefit the state’s highest-income taxpayers.

Health Risks From Increased Pollution

In August 2025, shortly after the passage of HB 2014, the West Virginia Department of Environmental Protection issued a permit authorizing construction of the Ridgeline Facility.

A report by the Harvard T.H. School of Public Health’s Dominici Lab commissioned by Tucker United found that the Ridgeline Facility will expose more than 250,000 people to increased particulate matter pollution across northeastern West Virginia and into northern Virginia (Figure 2). The report found that the health impacts associated with increased exposure, such as heart attacks, asthma attacks, and hospitalizations, will cost $35 million annually.

Figure 1: Increased Particulate Matter (PM_2.5) Pollution from the Ridgeline Facility

Source: Tucker United, Harvard T.H. Chan School of Public Health’s Dominici Lab

Even Bigger Projects Are Planned

Yet the proposed Ridgeline Facility is not even close to the largest proposed data center of its kind in West Virginia. That distinction belongs to the 9.8 GW Adams Fork pair of data center campuses in Mingo County, proposed by TransGas Development Systems, which has struggled to advance other projects at the same site. The second largest is the proposed 8 GW Nscale Monarch Compute Campus, located in Mason County, a few miles north of Point Pleasant. The first phase of the Monarch Compute Campus is already underway, with the facility planning to supply compute capacity to Microsoft and Caterpillar set to supply 2 GW capacity of fossil gas generators. The increase in particulate matter pollution from these sites can be expected to be similarly massive.

It is hard to overstate the size of the Adams Fork and Monarch proposals. If fully built, each of these campuses would be one of the largest in the world. Including the Ridgeline facility and the proposed 600 MW Penzance Bedington Campus in Berkeley County, West Virginia’s 20 GW total of new data center capacity would consume 123 million MWh of electricity each year,2 more than 3.7 times as much as the entire state of West Virginia, which consumed 33 million MWh in 2024.

Powered by gas, these four data centers could combine to burn more than 2.5 billion cubic feet per day, or 914 billion cubic feet per year. This is a staggering amount of gas, equal to 2.7% of all gas consumed nationwide.3

Who Benefits, and Who Bears the Cost?

The examples of the Ridgeline Facility in Tucker County, the Monarch Compute Campus in Mason County, and Adams Fork in Mingo County expose how House Bill 2014 functions at the macro-level: Public officials tout the purported benefits of these data centers while silencing local communities and diverting millions of dollars of likely revenue from the municipalities hosting these projects.

The result so far is a data center buildout that ignores the justified outcry of the communities which bear the brunt of fossil pollution.

By Zane Gustafson, Research Fellow

The war in Iran has introduced major short-term issues in global ammonia supply chains. After the US and Israel launched airstrikes on February 28, 2026, Iran all but shut down the critically important Strait of Hormuz, through which roughly 20% of global oil and gas traversed in 2024 (Figure 1). Three cargo vessels were hit by projectiles on March 10th, and the ongoing threat to vessels is causing insurance rates to skyrocket. Without insurance, vessels do not leave port. As a result, vessel traffic through the Strait of Hormuz has declined by 97% from pre-war levels.

Figure 1: The Strait of Hormuz (EIA)

Five countries in the Middle East (Iran, Saudi Arabia, Qatar, the United Arab Emirates, and Bahrain) collectively supply about one fourth of global ammonia exports. Saudi Arabia and the U.A.E. are able to move some of their product around the Strait by pipeline, but exports from Qatar, Bahrain, and Kuwait must go through the Strait of Hormuz. Ammonia-based fertilizer prices have skyrocketed, with urea prices up nearly 50% since before the war began (from $465/tonne on 2/27 to $672/tonne on 3/23).

While the full impacts of this conflict are still difficult to predict, two long-term factors are driving the proposed ammonia buildout seen in the US in recent years: potential new markets, and government subsidies, specifically the 45Q tax credit for carbon capture and sequestration.

If fully built, proposed ammonia projects in the US would nearly quadruple national capacity from 18 million to 70 million metric tons, potentially costing taxpayers more than $5 billion in annual tax subsidies. The vast majority of this buildout would be “blue” ammonia, which is produced from natural gas using carbon capture and sequestration (CCS) to reduce the resulting greenhouse gas emissions.

The first factor driving this buildout is the industry expectation that new markets for ammonia may grow exponentially in the next few decades. Currently, fertilizers account for 70% of ammonia use globally, and 90% of ammonia use in the United States. The growth in new markets, including the use of ammonia to fuel ships or in power plants, would be a sharp departure from the present day.

The second factor is the availability of massive government subsidies for producing “low-carbon” hydrogen, the key ingredient in the process that transforms natural gas into ammonia. As covered in previous blog posts, the two primary subsidies for ammonia are 45V tax credits, which are awarded based on the amount and carbon intensity of hydrogen produced, and 45Q tax credits, which are granted based on the amount of carbon sequestered. Companies interested in claiming the 45V tax credit must begin construction on their production facilities before January 1, 2028 but those planning to rely on the 45Q tax credit have until January 1, 2033.

While many companies claim that their ammonia facility will capture and store 95% of the carbon emissions, in practice CCS rates rarely exceed 80%. “Blue” ammonia facilities would have a difficult time qualifying for even the lowest tier of 45V tax credit. This, combined with the accelerated timeline for beginning construction, makes it much more feasible for “blue” ammonia facilities to pursue the 45Q tax credit even though it is not as lucrative as 45V.

Of the 35 blue ammonia proposals in the US tracked by the Environmental Integrity Project, nine are currently on hold. Only 18 of the remaining proposals have disclosed their planned capacity (Figure 2). Using the Climate Advocate’s 45V and 45Q Calculator, I calculated how much money each of the proposed “blue” ammonia facilities would be eligible for at CO2 capture rates of 50% and 70%. Combined, the 18 projects could qualify for more than $5 billion in tax credits per year, depending on how successful CCS efforts are.

Figure 2: Locations of proposed “blue” ammonia facilities

Appalachia is home to two of these facilities: the massive Adams Fork project proposal in Mingo County, West Virginia, and the Trillium Piketon project proposal in Pike County, Ohio. If it is fully built and sequesters 70% of its carbon emissions, Adams Fork could receive nearly $1.3 billion in 45Q tax credits per year. Trillium Piketon is much smaller, yet it could still receive nearly $43 million in 45Q tax credits per year if it has a 70% CCS rate (Figure 3).

Because they are several years from operating (assuming they ever do), these facilities are unlikely to mitigate the wild price swings we’re currently seeing due to the war in Iran. Ultimately, it’s possible that a market for all of this planned capacity may never materialize, but if it does it will be taxpayers on the hook for this gas-fired ammonia production.

Figure 3: Proposed “Blue” Ammonia Facilities Could Receive More Than $5 Billion in Annual 45Q Tax Credits

By Ted Boettner, Senior Researcher

This post is part of a presentation Boettner made at the Mercer County Indivisible Town Hall on March 17, 2026, at the Karen Preservati Center in Princeton, West Virginia. The title of the town hall was “Why are West Virginia Power Bills So High?”. This post provides a more detailed analysis of this question along with a brief background on West Virginia’s electricity system and the evolution of retail electricity costs in the state over the last 25 years.

Over the last couple of years, there have been numerous news stories about rising electricity costs in West Virginia. Some residents have to pay more for their monthly electric bill than their mortgage payment. Federal data shows West Virginia’s retail electricity rates have grown faster than most states over the past 25 years. West Virginia residents also use more electricity on average than residents of other states, which compounds the problem. There are several central reasons why West Virginia’s electricity rates have soared recently, including rising fuel costs in the global market. The short answer is coal. The longer answer is below—and more complicated. Before diving into what is driving electricity costs in West Virginia, it is imperative to understand the structure of West Virginia’s electricity system.

West Virginia’s Electricity System, In Brief

There are three main components of the “bulk power system” and electricity costs: generation, transmission, and distribution. Generation includes power plants, transmission includes the high voltage lines that run up and down the hills connected to power plants and to distribution stations, and distribution includes the low voltage lines and wires in your neighborhood that deliver the electricity to your home or business.1

In West Virginia, all three components are vertically integrated (for the most part) under investor-owned utilities that serve captive, in-state customers and that are regulated by the state Public Service Commission (PSC). That’s different than in some “restructured” states, such as neighboring Ohio and Pennsylvania, where there are no state-sanctioned monopolies on generating electricity but there are monopolies on distribution or delivery of electricity. So, customers in Ohio and Pennsylvania choose from competing suppliers of generation, while still relying on local utilities that operate as regulated monopolies for transmission and delivery.

While the public service commission (PSC) in West Virginia regulates and sets retail electricity rates that include all three components of your electricity bill, the Public Utility Commission (PUC) in Pennsylvania approves rates only for the distribution and delivery of electricity. Transmission rates or costs are regulated by the Federal Energy Regulatory Commission since they cross state lines.

West Virginia’s electric utilities, along with 12 other states and the District of Columbia, are members of the PJM Interconnection, a private, non-profit corporation known as a regional transmission organization (RTO) that operates the grid and has exclusive control over electricity transactions and transmission in interstate commerce. PJM does not produce electricity or own transmission lines, but it manages electricity generation and transmission and dispatches electric power through several wholesale electricity markets (energy, capacity, ancillary). PJM’s basic functions are to ensure grid transmission reliability, provide nondiscriminatory interconnection services for generators, and operate electricity markets. Figure 1 below shows the shares of electricity costs by generation, transmission, and distribution for PJM’s Western Region, which includes all of West Virginia.

Figure 1: Generation makes up more than 60% of electricity costs in West Virginia.
Share of electricity costs in PJM’s Western Region, 2024

Source: US Energy Information Administration, 2025 Annual Energy Outlook, PJM West Region, Table 54, Prices by Service Category for 2024 (Projections)

West Virginia is served mainly by four vertically integrated electric utilities owned by American Electric Power (AEP) and FirstEnergy.2 AEP’s Appalachian Power, which covers the southern part of the state, and AEP’s Wheeling Power, which covers Marshall and Ohio counties in the northern panhandle, serve about 458,000 customers in West Virginia.

AEP’s service area (darker blue):

Source: AEP

Meanwhile, First Energy’s Mon Power serves most of northern West Virginia and its subsidiary Potomac Edison serves seven counties in the eastern panhandle. Altogether, Mon Power and Potomac Edison serve about 550,000 customers in West Virginia.

FirstEnergy’s service area (green):

Source: FirstEnergy

West Virginia also has several independent power producers (IPPs), which sell electricity on PJM energy markets to utility companies and others inside and outside of the state.

PJM is the world’s largest regional transmission organization (RTO), but it does not produce electricity or own transmission lines—it manages electricity generation and transmission and dispatches electric power. AEP, for example, gave functional control of its 22,300 miles of transmission lines to PJM in 2004. PJM’s basic functions are to ensure grid transmission reliability, provide nondiscriminatory interconnection services for generators, and operate electricity markets.

West Virginia’s regulated electric utilities use power mostly from their own power plants, but they also purchase electricity from PJM’s energy markets generated by power plants located inside and outside of the state. Figure 2 below shows West Virginia’s total power generation in 2025 by source. Not surprisingly, coal makes up about 87% of power generation, while renewable energy—wind, conventional hydroelectric, and solar—comprises about 7% and natural gas makes up 6%.

Figure 2: Coal makes up nearly nine-tenths of West Virginia’s fuel mix.
West Virginia’s total power generation by source, 2025

Source: US Energy Information Administration, Electricity Data Browser.

West Virginia’s two main regulated electric utilities produced about 81% of the state’s electric power in 2025. Most retail customers in West Virginia pay American Electric Power (AEP) or FirstEnergy for their electricity, which mostly comes from coal plants, such as the John E. Amos Power Plant in Winfield and the Harrison Power Station in Haywood. AEP and FirstEnergy can also buy and sell power on PJM’s wholesale electricity market and engage in bilateral transactions with other load-serving entities to provide power. However, most of their power is self-scheduled and delivered to in-state customers. In 2025, the state’s regulated electric utilities delivered about 33.7 terawatts of power to in-state customers, comprising about 63% of total electricity generated in the state. Residential deliveries made up one-third of sales, while commercial customers accounted for 24% and industrial customers accounted for another 43%.

About 19% of the electricity produced in West Virginia is from independent power producers, such as Longview Coal Plant in Morgantown and NedPower Wind in Mt. Storm. Their electricity is sold on PJM’s spot market or through agreements with industrial customers. While it is unclear how much of the electricity (electrons) produced in West Virginia is consumed in West Virginia, the state is one of the few states that produces more electricity than it consumes.

The Rapid Growth of West Virginia’s Retail Electricity Costs

While electricity rates are complex and are set by a multitude of factors in West Virginia, they are ultimately approved by the West Virginia Public Service Commission (WVPSC), the body that oversees the state’s regulated electricity providers. There are two central components of electricity rates in West Virginia approved by the WVPSC: base rates and rider charges. The diagram below provides a breakdown of my personal residential electric bill from Appalachian Power for December 2025, which shows an average cost of 15.7 cents per kWh. The base rate charges comprise about 55% of my total electric bill, while riders—the biggest of which is Expanded Net Energy Cost (ENEC)—make up 39%. Taxes are 6% of my bill. (Base rates are reviewed every three years by the WVPSC, while ENEC cases are reviewed annually.)

While the above electric bill provides a snapshot and itemization of my residential electricity costs, it is imperative to examine the growth in West Virginia electricity rates over time. Luckily, the US Energy Information Administration (EIA) provides data on the average retail electricity costs per kilowatt hour (kWh) for customers (residential, commercial, and industrial) by state. EIA simply divides the total electricity bills (sales revenue) by electricity usage to arrive at an average cost per kWh. The table below shows the average retail prices in West Virginia, the United States, and bordering states using the latest EIA data (January 2026). For January 2026, West Virginia average electricity rates were lower than all neighboring states and the US average for customers but ranked second lowest for residential rates among bordering states. However, West Virginia’s residential electricity rate was much higher in the summer and early fall (16.14 cents per kWh in October), perhaps reflecting the lower winter rates (as shown in my power bill above).

Table 1: Across sectors, West Virginia’s electricity costs are lower than neighboring states and the US average.
Average Retail Electricity Cost by Sector, Cents per Kilowatt Hour, January 2026

Source: US Energy Information Administration, Monthly Electric Power Monthly, Table 5.3

Looking at annual average electricity rates, Figure 3 below shows electricity rates for residential customers for the United States and West Virginia over the last 25 years. From 2001 to 2025, West Virginia’s average residential electricity rates grew by 146%, compared to 102% nationally. Only four other states experienced larger growth in electricity rate increases over this time. Maryland is the only neighboring state with faster-growing electricity rates over this 25-year period. West Virginia had the lowest residential electricity rates in the country in 2005. However, by 2025, 26 other states had lower rates, including Virginia and Kentucky.

Over the last ten years, West Virginia residential electricity rates have grown by 53%, much faster than neighboring states (eclipsing Ohio’s 32% growth rate, Pennsylvania’s 42% rate, and Virginia’s 34% rate) and the US average. One potential reason is that our neighboring states (except Maryland) have shifted from coal to natural gas. For example, in 2010 Ohio produced about four-fifths (82%) of its power generation from coal, compared to about 21% today. Meanwhile, natural gas has grown from 5% to 60% of electric generation in Ohio over this time period.

Figure 3: West Virginia, once home to the fourth-cheapest retail electricity prices in the US, has fallen to 27th.
Average retail electricity prices, West Virginia and US (including West Virginia’s national electricity price ranking), 2000-2025

* 1 = cheapest.
Source: US Energy Information Administration, Electric Power Monthly

While West Virginia has lower residential electricity rates than the US average and most neighboring states, residential households in West Virginia use more electricity than other states. This is mostly because of the state’s older, less insulated housing stock, which consists mostly of single-family homes. In 2025, the average monthly electricity bill for residents was about $165 (1,073*15.41 cents) compared to a US average of $152 (878*17.3 cents). West Virginia uses about 22% more electricity per residential household than the US average, and consumes more energy (gas and electric) per household than the national average. The average annual electric bill in West Virginia is about 3% ($1,984/$63,150) of median household income, compared to about 2% ($1,823/$83,730) nationally.

Why are West Virginia’s electricity rates growing faster than other states?

There are several reasons why the state has seen above-average growth in retail electricity rates over the last decade. The state has seen a sharp decline in electricity production over the past 20 years, which means the state is exporting less electricity (which helps lower in-state costs). The decline in power generation is the result of aging coal plants that make up the vast majority of generation. These plants have seen capacity decline as other states stopped buying the state’s coal fired electricity. West Virginia has also been hammered by rising coal fuels costs since 2007 as other states shifted away from coal. These rising costs of coal have been mismanaged by one of the state’s regulated electricity utilities as old coal plants have been transferred to in-state customers. Environmental upgrades to old coal plants have also led to higher prices, along with state policies that have mandated the use of more coal plants to serve in-state customers.

1. Producing and Selling Less Power

West Virginia is producing and exporting a lot less electricity today than it was three decades ago. From its 2002 peak through 2024, net electricity production declined by 47%, from 95 terawatt hours to 51 terawatt hours. In 2002, West Virginia consumed about 30% of the electricity generated in the state; in 2024, that number was 65%. So, while West Virginia still produces more electricity than it consumes, the state is producing a lot less electricity today.

Figure 4: West Virginia is producing—and exporting—much less electricity.
West Virginia’s power generation, consumption, and consumption as a share of generation, 1990-2024

Source: US Energy Information Administration, Electric Power Monthly

For example, Appalachian Power/Wheeling Power had about $541 million in off-system sales of electricity it sold to other states (and perhaps some to Mon Power) in 2008, compared to just $46 million in 2024. These off-system sales—mostly power sold out of state—helped keep in-state electricity rates lower. When the state’s regulated electricity utilities can sell power onto the PJM wholesale markets, it has the potential to offset costs for in-state customers by increasing revenue. These off-system sales are credited against increases in annual fuel expenses (ENEC expenses) that can lower rate increases.

The question of why West Virginia is producing less electricity is answered below.

2. Coal Plant Capacity Shrinking, Aging, and Running Less

The sharp decline in generation and in off-system sales in West Virginia is due to several compounding factors. First, over the last several years the state has retired several coal plants. Total generation capacity (net summer) has declined from 16.5 gigawatts (GW) in 2005 to 15.1 GW. The average age of the state’s coal plants is around 50 years. As plants age, they physically degrade and become less efficient. In addition to age, the coal plants are most exposed to being dispatched out of merit order by cheaper gas and renewable generation. The restructuring of electricity markets in the late 1990s has led to more gas and renewables on the grid, which has been hard on coal plants and nuclear plants that are not made to start and stop frequently, causing cycling damage. Operators of these aging coal plants often also scale back investments when the plants get closer to their end of life. So aging plants raise costs and reduce output, worsening the economics in both directions.

This is best seen by looking at the capacity factor of coal plants in West Virginia, which make up about 90% of power generation in the state. The capacity factor, or the amount of electricity produced as a share of potential output (nameplate capacity) in a given year, has plummeted for coal plants in West Virginia. For example, in 2008 coal plants ran 69% of the time compared to just 38% in 2024. While coal plants have aged over this time and have become less efficient, this is not the central reason for the declining capacity factor and generation. One example that highlights this is Longview Power Plant near Morgantown, West Virginia, which is an independent power producer and one the nation’s newest coal power plants, built in 2011. Longview still operates but filed for bankruptcy in 2020, citing “cheap natural gas” in the PJM energy market that depressed power prices.

Figure 5: West Virginia’s coal plants are operating at a lower capacity.
Capacity factor of West Virginia’s coal plants

Source: US Energy Information Administration, Electric Power Monthly

3. Natural gas competition

The shale natural gas revolution that began about 15 years ago in Appalachia has been a larger driver of making West Virginia’s coal-fired power plants economically uncompetitive. When cheap natural gas flooded the market, neighboring states shifted to building more efficient gas plants and closed many coal plants. The new gas-generated electricity that was pushed into the wholesale energy markets on PJM pushed prices downward and coal became less competitive. The irony is that much of that cheap Marcellus gas came from West Virginia.

Coal’s share of generation within PJM fell from roughly 44% in 2013 to under 14% today, replaced almost exclusively by gas combined-cycle plants. As West Virginia’s neighboring states shifted from coal to gas, they no longer relied on West Virginia’s more expensive coal power plants for electricity. As discussed previously, natural gas is replacing coal in PJM markets, which are more conducive to natural gas and renewable energy than baseload coal or nuclear power. The push to lower carbon emissions over the last decade has also made natural gas a more preferred fuel source of utilities as coal plants have had to undergo extensive environmental upgrades to remain federally compliant.

4. Environmental compliance upgrades

In 2021, the WVPSC approved about $448 million in environmental upgrades for three coal plants—Amos, Mountaineer, and Mitchell, owned by Appalachian/Wheeling Power—to comply with EPA requirements regarding wastewater and coal ash (effluent limitation guidelines and coal combustion residue rules), allowing the plants to continue operating through 2040. These upgrades were approved despite Appalachian Power’s own testimony showing that retirement of the Mitchell plant would save ratepayers $27 million per year. (The cost of these upgrades is highlighted in my electric bill above as “environmental compliance surcharge”.)

5. Rising Fuel (Coal) Costs

Large swings in coal costs over the last decade and a half coupled with the mismanagement of coal deliveries by Appalachian/Wheeling Power have led to large electricity rate increases in West Virginia. Nationally, the cost of coal delivered to electric power plants rose sharply from 2007 to 2010, from $35 to $45 per ton. The second large spike in coal prices was in 2022—which was a result of rising energy prices from the fallout of the Ukraine War—where prices rose from $38 per ton in 2021 to $47 per ton by 2023. In West Virginia, the cost rose from $48 a ton in 2018 to $71 a ton in 2025.

The spike in coal prices in 2008 led Appalachian/Wheeling Power to incorporate a rider charge called the “consumer rate relief charge” in 2013, which is included in my electric bill above, and was established through the sale of $380.3 million in bonds approved by the WVPSC. The bond is expected to be paid off by 2028.

The second major increase in electricity charges for Appalachian/Wheeling Power customers in West Virginia was in 2024, when the WV PSC approved $321 million ($39 million per year) in rate recovery from fuel costs from coal. The fuel increases came from the sharp rise in coal prices and the mismanagement of coal contracts by Appalachian/Wheeling Power. The WVPSC determined that $232 million of the $553 million sought by Appalachian/Wheeling Power could have been saved had the company acted prudently.

Conclusion

West Virginia’s retail electric power rates have risen sharply over the past decade, driven by the state’s aging coal plants and the restructuring of electricity markets that have favored natural gas. Compounding this has been large increases in coal fuel costs, which have led to additional charges on customers’ electric bills, along with expensive environmental compliance upgrades to coal plants. The sharp decline of in-state electricity production, which has led to a massive drop in off-site electricity sales out of state, has also contributed to cost increases for in-state consumers.

Going forward, the West Virginia legislature has been pushing for regulated coal plants in the state to produce 69% of the time, despite concerns from Appalachian Power that this isn’t feasible or economical. The push to produce more electricity from in-state coal plants—despite the risk of higher electricity costs—was put in stark relief last month when the chair of the WVPSC, Charlotte Lane, asked state legislators, “if it costs more to run the coal-fired plants than buying off the PJM market, do you want those plants to run and pass those costs to ratepayers?”

This question gets to the heart of the current impasse on growing electricity rates in West Virginia. Are West Virginians okay with paying higher electricity rates, since some of the coal that is mined in the state feeds that state’s coal plants? Or should the state work to ensure that ratepayers pay the lowest rates possible?

If it costs more to run the coal-fired plants than buying off the PJM market, do you want those plants to run and pass those costs to ratepayers?

West Virginia coal production value has shifted from steam to metallurgical (met) coal—met coal is responsible for more than 60% of the total coal value in the state—over the last decade as surrounding states have purchased less West Virginia coal and shifted away from coal-fired power plants. Nearly all of the state’s coal-fired power plants are expected to be retired over the next 15 years. While Governor Morrissey wants to double down and build new coal plants, this strategy is highly unlikely to yield lower electricity costs in the future because coal plants face high capital costs, regulatory uncertainties, and financing challenges. While the Trump administration is aiming to provide $625 million in investments for existing coal plants in the United States, including West Virginia, the plants will eventually have to be retired. The question, then, is what will replace the plants—if anything?

By Sean O’Leary, Senior Researcher

A pair of Ohio River Valley gas plant projects would be the largest, most expensive energy developments in US history. But as the price of gas power soars and demand remains uncertain, there’s a chance that taxpayers and utility customers could get stuck with a big bill.

In announcements on February 19 and March 19, 2026 the White House unveiled plans to build natural gas-fired power plants in Piketon, Ohio and the Monongahela Valley of Pennsylvania that will have a combined capacity of 13.5 Gigawatts (GW) - enough to power roughly 10 million households. The 9.2 GW Piketon facility was said to represent an investment of $33 billion and the Mon Valley facility $17 billion.

At nearly twice the generating capacity of the Grand Coulee Dam in Washington state, the two projects constitute the largest power system development in the nation’s history. And, at a combined cost of $50 billion, the projects are 34% more expensive than the infamous Plant Vogtle nuclear project in Georgia that cost $37 billion and whose impact on customers’ electric bills resulted in Georgia voters throwing out of office the incumbents on Georgia’s Public Service Commission.

But Plant Vogtle isn’t the financial disaster of which the Piketon and Mon Valley projects are most reminiscent. That distinction falls to the Washington Public Power Service System (WPPSS) project to build five nuclear power plants in the Pacific Northwest in the 1970s and early 1980s. The reason WPPSS, which acquired the unfortunate nickname “Whoops”, doesn’t make the list of the largest power projects ever is that it collapsed financially after just one of the five proposed plants was built.

The failure to build the four remaining plants cost municipal bond investors $2.25 billion, at the time the largest municipal bond default in the nation’s history. Because the investors consisted mostly of 88 public utilities, public utility districts, and electric co-ops, the ultimate bill was paid mostly by customers. So, why are the Piketon and Mon Valley projects reminiscent of WPPS? Two reasons, one disturbing and the other worse.

Like WPPSS, Piketon and Mon Valley are responses to expected demand growth. In the case of our region, this anticipated demand is driven by data center expansion whereas, in the case of WPPSS, anticipated population and industrial expansion were expected to drive load growth. The fundamental problem with WPPSS was that the expected demand didn’t materialize. Increases in the Pacific Northwest’s population were almost entirely offset by improvements in energy efficiency and by the region’s evolution away from heavy industry toward a more service-based economy.

Because of the many uncertainties surrounding artificial intelligence and its implications, similar doubts haunt expected data center expansion in our region. If demand for electricity doesn’t grow at the pace or to the degree that’s expected, we may be left with quite a bit of underutilized and very expensive generating capacity.

Equally worrying is the cost of the proposed plants. With a combined price tag of $50 billion, the 13.5 GW of planned generating capacity will cost $3,700/kw, which is an astonishingly high price. Amid the current crush in demand for gas-fired power, developers have recently been paying as much as $2,500/kw for new plants. That means the new projects will cost nearly 50% more than the market rate, which is itself already highly inflated.

As recently as December 2023, the Energy Information Administration reported that the cost of building a new gas-fired power plant was less than $900/kw, making the cost of the Piketon and Mon Valley facilities more than four times that of nearly any gas-fired power plant operating today. This is a consequence of deciding to add new gas-fired capacity at a time when demand for turbines and other equipment required by the new plants is at an all-time high.

Facilities that cost that much to build can never be a source of low-cost energy. Far from reducing electricity rates as Energy Secretary Chris Wright claims, they will raise them. Either that or those operating the plants will lose money. Of course, there is a third, worst case possibility; customers’ rates may rise and the owners may also lose money.

Still, there are hopes for redemption. One is that data center developers and operators will absorb the projects’ entire cost. The possibility is remote however because it would require significant changes in how utilities are regulated and how they allocate costs among customers. It’s true that President Trump recently announced a pledge by hyperscalers to insulate existing customers from cost increases induced by data centers. However, neither utilities nor state regulators were signatories to the pledge, which in any case is unenforceable - a fact of which customers in Pennsylvania and Ohio are constantly reminded as they watch data centers drive up their monthly bills with no mention, at least so far, of refunds or rebates.

There remains one other possible outcome. As with WPPSS and a later tech-fueled power sector boom in the late 1990s and early 2000s, financial reality may come crashing down on the projects before they are completed, leaving investors and/or customers holding a very large bag. In the case of WPPSS the bag takes the shape of the world’s largest tombstone - an immense cooling tower for the never completed Satsop nuclear power station that 50 years later sits abandoned in rural Washington as a monument to the wages of power system miscalculation.

By Nick Messenger, Economist & Senior Researcher

At a gathering last December, West Virginia legislators laid out their economic development priorities for 2026. The centerpiece of their economic development plan was copying and pasting an economic development experiment from Ohio that hasn’t really produced results.

Now, the West Virginia House of Delegates has taken the first major step. Earlier this month, the House of Delegates passed HB4001, which would effectively privatize economic development in West Virginia. “TEAM-WV” would create a statewide organization modeled after JobsOhio, which privatized Ohio’s economic development efforts and has diverted over a billion dollars in potential public revenue to the new organization, whose dealings are not subject to the Ohio Public Records Act (ORC 187.04).

West Virginia legislators say that this is a crucial step for the state. But by centralizing site preparation and corporate partnerships, this kind of heavy-handed state centralization seems directly at odds with House Speaker Roger Hanshaw’s comments about the plan, saying, “We know that the way we best advance West Virginia is by empowering local people in local communities to make decisions about their own economic future.”

Note: As of this post’s publication, the West Virginia Senate failed to pass the current version of HB4001 during the 2026 regular session.

Here’s the biggest problem: despite the hype, I’ve seen firsthand that JobsOhio hasn’t really worked that well in my state, and there’s no compelling evidence that it will work any better for West Virginia—a state that already lacks some of the built-in legacy economic advantages of cities like Columbus, Cleveland, and Cincinnati. Additionally, JobsOhio has not been without its share of misspent funds, recently working to reclaim $60,000 that it contributed to a single podcaster, whose “inappropriate relationship” ultimately led to the resignation of Ohio State University’s President. The episode has led former Ohio Governor and creator of JobsOhio, John Kasich, to question whether the organization has lost sight of its mission.

Taxpayers have forgone over $1 billion for JobsOhio since 2011, yet the state’s economy is barely treading water.

There is plenty of evidence to suggest that JobsOhio has failed to move the needle for Ohio beyond flashy corporate PR announcements, like a big, multibillion-dollar Intel chip factory that is now running more than six years behind schedule. Despite over $1 billion in forgone liquor revenue diverted to JobsOhio for spending on such projects, key indicators of economic well-being, like population, are on the decline. The state’s Department of Development still estimates that Ohio will lose 675,000 people between 2020 and 2025, which is a population decline of 5.7%. The overwhelming majority of economists surveyed by Scioto Analysis recently cited lack of job opportunities as a primary driver of population loss, a problem that has historically plagued West Virginia as well.

Population growth and economic growth are tightly linked. In a recent study by the Urban Institute, researchers found that West Virginia is the only state in the country that did not see an increase in inflation-adjusted median household income in the last fifty years. More tax cuts are highly unlikely to help solve the economic problems in West Virginia if they haven’t helped staunch the population loss in Ohio.

In March 2026, JobsOhio took to social media to brag about its high ranking in this year’s annual “Governor’s Cup” from Site Selection Magazine, which is an economic development trade publication. Every year, the magazine ranks US states based on how much “economic development” they have done, which the magazine measures in the form of large plant or factory projects, rather than incorporating other quality of life-based metrics like education quality, public health, or housing affordability.

This framework conforms with the general belief of most politicians that, if you simply lure more factories to your state with lucrative tax incentives and grants, then you will inherently create more jobs. I like to call this “ribbon-cutting syndrome;” politicians look productive to voters when projects are announced at ceremonies with dirt and shovels. But nobody tends to follow up on whether the promised jobs and salaries actually materialize.

By these surface-level metrics, Ohio has done quite well, with more top-four Governor’s Cup appearances than Ohio State University’s College Football Playoff appearances. Unfortunately, merely “winning” more corporate location battles and cutting the ribbons doesn’t correlate with meaningful, sustained employment.

That may seem surprising, since that’s been the dominant belief in politics for over 40 years. But the chart below illustrates this point. It shows the state-level percentage change in non-farm employment from 2019 to 2022, plotted against the number of large plant facility openings between 2018 and 2020 in each US state from the Governor’s Cup data. And, in short, the line is pretty flat. States with huge facility openings don’t really seem to grow non-farm employment, on net.

There are a few reasons to use this data approach. First, we are testing whether job creation is correlated with past plant openings, which we would expect if the JobsOhio theory of economic development is true. In other words, this chart gives these plants a few years to be built and hire workers, rather than cherry-picking the announcement year data when it would be unreasonable to expect their employment effects to appear in the data. Second, we are examining pre-pandemic jobs as a baseline and extending our analysis into 2022, beyond the COVID-19 pandemic. This actually gives the benefit of the doubt to JobsOhio by allowing any pandemic recovery “job growth” in 2021 and 2022 to be credited favorably to JobsOhio.

Keep in mind that JobsOhio was created in 2011, so it had nearly a decade to usher in job-boosting projects prior to the data I use here. So, any projects that were incentivized by JobsOhio between 2011 and 2019 also didn’t move the needle on employment in the future. This is another common economic development argument: that large plants generate what economists call “agglomeration effects.” For example, Ohio relied on this idea to argue that the billions of taxpayer dollars spent incentivizing the Intel plant would create additional “spillover effects” by attracting other technology and supply chain companies to the region. JobsOhio even dubbed the area “Silicon Heartland,” joining a long list of failed places that thought they, too, would be “the next Silicon Valley.” But recent research by fellow economists Mark Partridge and Carlianne Patrick tests this idea using 20 years of confidential, county-level Census microdata. They found “little evidence that the average highly incentivized large plant generates significant productivity spillovers.”

There is little job-based evidence to support the “smokestack-chasing” version of economic development, as evidenced by the nearly flat trendline in Figure 1. In fact, the US states that experienced the largest job growth have averaged fewer than 25 large facility openings per year, while states with over 100 large plant openings between 2018 and 2020, like Ohio, still generally had near-zero or negative job growth from 2019 to 2022. Clearly, there must be more to economic development than site preparation, grants, and tax incentives to recruit big companies.

But let’s keep digging using even older issues of Site Selection Magazine to examine Ohio’s rankings among US states over time. In the decade following JobsOhio’s creation, the state finished #1 in the nation three times and, impressively, never finished outside of the top four. Ohio’s average annual finish among all states was #2.4, which looks fantastic on paper and to voters.

But, following the assumptions of traditional economic development practitioners, this should mean that all of the investment into JobsOhio is paying off over a decade later. If the traditional economic development paradigm is correct, Ohio should be booming and ranked close to the top of the nation in GDP growth, job growth, and population growth—the traditional measures of “success” preferred by economic development professionals.

Unfortunately, the data still tells a different story.

First, let’s look at real GDP per capita. This is a measure of Ohio’s total economic output value (adjusted for inflation) divided by its population. (We’ll also examine population separately, to make sure that GDP per capita changes aren’t just a reflection of population changes.)

A quick aside: despite its common use, GDP is really not the best measure of economic well-being for a state’s people, since a large portion of the value produced by Ohio workers can actually leave the state. For example, a company that is not headquartered in Ohio redistributes some of the value of goods produced in Ohio as higher salaries to employees or stock returns to shareholders in other states or countries. Still, GDP is a metric that politicians like to discuss, so we’ll take a look at it here.

Ohio’s per-capita GDP performance—compared to the US average—has actually worsened since JobsOhio was created in 2011. Ohio’s GDP per capita was 91.8% of the national average in 2011 and fell to 88.4% in 2024 – a decline of roughly 3%. Simply put, this means that Ohio’s GDP per capita growth has slowed behind the national average.

Ohio has improved its GDP per capita state ranking by just one spot since JobsOhio’s inception, moving from 30th in 2011 among all US states to 29th in 2024. This modest improvement comes despite averaging an annual top-three finish in plant openings and investing hundreds of millions of dollars into JobsOhio’s economic development strategies. If copying JobsOhio generated the same relative results in West Virginia, the state might leap from 49th to 48th in GDP per capita.

Perhaps this result is just a function of population growth? Mathematically, since GDP per capita is a fraction where population is the bottom number, Ohio’s mediocre results could theoretically be because JobsOhio has been so wildly successful that Ohio’s population exploded and rapidly increased the bottom of the GDP per capita fraction. Figure 3 suggests that it is not the reality. Since 2011, the US population has grown faster than Ohio’s, with much of Ohio’s population growth focused in just a handful of urban and suburban counties near Columbus. And, as I noted earlier, the state itself expects to lose population on net over the coming two decades.

Relative to all US states, Ohio ranked 41st in population growth between 2011 and 2024, according to data from the U.S. Census. I note that while West Virginia is ranked dead last in population growth—and thus, trying to be more like Ohio would technically constitute an improvement—emulating a fellow state in the bottom quintile nationally is a questionable strategy to jumpstart West Virginia’s economy.

Let’s look at jobs and incomes. After all, as West Virginia House Speaker Hanshaw noted, “We want to double down on the belief that creating a job, putting men and women in good-paying jobs, is priority number one for the government.” Is there a reason to believe that copying JobsOhio is the best strategy to do that?

Figure 4 plots the change in US, Ohio, and West Virginia total nonfarm employment from 2011 to 2024, normalizing 2011 to 100%. We see that Ohio again lags the national average with a widening employment gap over the 13-year period. While this is better than West Virginia’s negative job growth, it is hardly exemplary performance. Between 2011 and 2024, Ohio ranked 36th nationally in nonfarm job growth.

There’s also little evidence of a so-called “manufacturing renaissance” in Ohio. Figure 5 shows the share of nonfarm employment in manufacturing NAICS/SIC codes from the US Department of Labor for both Ohio and the US.

We can see that after a spike attributable to World War II in the mid-1940s, manufacturing as a share of employment has been on a slow and steady decline both nationwide and in Ohio. Ohio’s manufacturing lead over the national average (the gap between the lines) has actually narrowed. Beginning in 2010, after the Great Recession, we see a relative leveling off, but despite the creation of JobsOhio in 2011, Ohio manufacturing employment at best holds steady, rather than expands. That is to say, there is no spike or other evidence here to suggest that JobsOhio has somehow reversed Ohio’s fortunes in manufacturing employment, despite so many large plant and capital investment announcements. There is little reason to believe that pursuing a private economic development organization would reverse West Virginia’s job fortunes either, and West Virginia politicians should aim to do more than tread water.

In terms of incomes, Ohio improved but still lagged nationally. In 2011, real personal income per capita in Ohio in 2024 dollars was $54,287. In 2024, real personal income per capita was $64,464—a growth rate of roughly 19%. But this lagged the national growth rate of over 23%. Ohioans saw an increase in real income, but by a smaller amount than the average American. In ranking terms, Ohio lost ground, moving from 32nd in 2011 to 38th in 2024 in terms of real personal income per capita among US states. And, as I noted above, this decline doesn’t appear attributable to population growth, since statewide population has grown more slowly and, in most parts of the state, declined.

It appears that by the most common economic development metrics, Ohio has either simply treaded water or declined since the creation of JobsOhio. This is a poor result for a cost of over $1 billion in spending, although the specific costs are hard to verify because JobsOhio was created as a private nonprofit entity and thus is not subject to public records requests, despite the pleasant-sounding nonprofit title. West Virginians would do well to be wary of the same loss of transparency by privatizing their economic development.

We can still try to piece together a solid financial picture of JobsOhio from what we do know. Between 2015 and 2022, one analysis found that JobsOhio had spent over $1 billion on “economic development” in Ohio – money that would have gone into the state treasury otherwise. Over its existence, JobsOhio’s own internal payroll for staff ballooned from $2.5 million to over $19 million, with the average JobsOhio employee earning a compensation package in 2022 worth roughly $180,000—more than double the average Ohioan. The analysis found that JobsOhio spent over $32 million on marketing, over $16 million in “professional services,” and over $210 million in incentives in 2021. JobsOhio’s annual operating expenses were roughly $300 million as of 2021, according to the same analysis. Where does all this money come from?

JobsOhio is funded by Ohio’s liquor monopoly, established in 1933 when the 21st Amendment repealed prohibition. The state actually operated liquor stores but outsourced the retail operations in the 1990s. In 2011, JobsOhio was allowed to lease the liquor franchise for $1.51 billion for 25 years. A report from the Legislative Service Commission in 2022 noted that the state’s liquor franchise generated $338 million in net profit for JobsOhio in fiscal year 2021 alone, making the lease an incredibly favorable deal for JobsOhio at the expense of the State of Ohio, which would presumably otherwise collect those profits directly. These profits far exceeded the $100M that the Legislative Service Commission projected in 2013 JobsOhio would retain after “payments to the state and debt service payments on the economic development bonds.” In early 2025, JobsOhio’s lease was extended by 15 years for zero additional payment, drawing criticism from the Ohio Attorney General that the state was essentially giving away 15 more years of liquor franchise profits for free.

In essence, JobsOhio has shifted hundreds of millions of dollars away from public control and into a private pot of money subject to less scrutiny. Ohio Supreme Court Justice Bill O’Neill dissented on a ruling in favor of JobsOhio’s creation, writing, “Hundreds of millions of dollars in public funds are being funneled into a dark hole to be disbursed without public scrutiny, and the highest court in the land is looking the other way. The Supreme Court of Ohio is the last house on the street, and passing on this case is an abdication of our duty as protectors of the (Ohio) Constitution.”

Indeed, board members of JobsOhio are often corporate officers of large companies that might receive benefits from JobsOhio. In 2014, for example, two board members were advised of a potential conflict of interest due to their corporate affiliation with Marathon Petroleum Corp, which was negotiating an extension of over $3.5 million in tax incentives with JobsOhio. Board members throughout JobsOhio’s existence have included corporate representatives and affiliates of many of Ohio’s largest companies.

West Virginia local officials, school leaders, and residents should take caution as HB4001 moves forward. Table 1 below summarizes the economic trends in Ohio since the state created JobsOhio and diverted billions of dollars in public revenue to the less transparent private entity.

The results, despite over $1 billion in foregone revenue to JobsOhio, suggest quite a poor return on investment for Ohio taxpayers. Ohio’s budgets during this time have still slashed school funding and pinched local governments around the state over the last decade. JobsOhio simply hasn’t delivered a proven jobs boom for Ohioans, and copying it is unlikely to do so for West Virginians.

By Zane Gustafson, Research Fellow

The giant Adams Fork project keeps getting bigger. Back around 2010, the 150-acre site north of Wharncliffe, WV, was slated to become a coal-to-gas facility but construction had barely begun before it was canceled in 2011. The site remained dormant for over a decade until 2023, when TransGas announced new plans to build a massive ammonia plant fueled by natural gas.

As it stands today, Adams Fork has permits to build three massive natural gas facilities. First came a March 2024 permit to build the largest ammonia plant in the country. If fully built — and it’s increasingly doubtful it ever will be — this facility would produce over 13 million metric tons of ammonia, nearly equal to the entire US ammonia output in 2022.

Producing 13 million metric tons of ammonia from fossil gas would consume more than 414 billion cubic feet (bcf) of gas per year.1 This single facility would consume 50 percent more natural gas than the entire rest of the state of West Virginia.

Second, two October 2025 permits to build identical data center campuses outside of Wharncliffe, one at the former Twisted Gun Golfcourse and one at the Harless Industrial Park. Each campus would have the capacity to consume 2,457 Megawatts of electricity. Combined, the two data centers would consume over 29.4 million Megawatthours (MWh) of electricity,2 almost equal to the 33 million MWh electricity consumption of the entire state of West Virginia.

The data centers would be powered by newly constructed gas-fired power plants built solely for their use. These “off-grid” power plants would not provide any electricity to West Virginia ratepayers and were made possible by the passage of HB 2014 in April 2025. Powering them would consume 218 billion cubic feet of fossil gas per year.3

In a May 2025 presentation posted to their website, parent company TransGas Development Systems, LLC stated that it plans to expand data center capacity to 9.8 GW. While not yet permitted, increasing data center capacity from 4.9 GW to 9.8 GW would add another 218 bcf of fossil gas consumption per year.

Between the permitted ammonia plant and the two permitted data center campuses, Adams Fork would consume a staggering 633 bcf of natural gas per year, or 1.73 bcf per day. That is more than double West Virginia’s 299 bcf gas consumption in 2024 and almost as much as the entire capacity of the controversial Mountain Valley Pipeline.

As permitted, the Adams Fork projects would, by themselves, increase gas consumption in the United States by nearly 2 percent.4 And if the data center campuses expand to TransGas’ stated goal of 9.8 GW, gas consumption would reach over 850 bcf per year, more than 2.5 percent of the gas consumption of the entire country.

In addition to being giant consumers of natural gas, these facilities would not produce much value for West Virginians. Producing ammonia from fracked natural gas would increase health risks in nearby communities: cancer, asthma, heart problems, birth defects, respiratory diseases, and more. Ammonia is a corrosive and toxic product in its own right, and there are signs that demand for new ammonia may not live up to the hype.

While the prospects for a low-carbon ammonia economy continue to wane, it is impossible to know what the future holds for data centers. Will growth continue exponentially? Is this a bubble about to pop? Or will the future land somewhere in the middle? Regardless, while data centers may provide a short-term construction boom, they do not provide many local jobs after construction. And, unfortunately, many of the jobs created across the broader digital economy are located far away from Appalachia.

Looking at the big picture, most of the economic benefits of the Adams Fork ammonia plant and data centers will go to big gas companies and out-of-state tech companies, while the negative health impacts of burning all that natural gas will be endured by West Virginians.

By Ted Boettner, Senior Researcher

West Virginia’s long history of oil and gas extraction has left the state saddled with tens of thousands of abandoned wells, many of which are leaking hazardous pollution into the air and ground. Cleaning up these old wells could create thousands of jobs, reduce toxic emissions, and protect nearby communities—but the state’s system for well plugging and regulation isn’t currently equipped to fix the problem, our latest report finds. Click here to view the full report on our website.

Tracing the history of West Virginia’s orphaned well problem

West Virginia has drilled more than 130,000 wells in its 150-year history of oil and gas exploration.

Figure 1: The peaks and valleys of well drilling in West Virginia

Historic Oil and Gas Wells Drilled in West Virginia, 1875-2025

Source: Arnold and Kemnitzer (1931) from 1875 to 1929. Data for wells drilled from 1930 to 1986 are from WV Department of Mines, Office of Oil and Gas, Annual Reports, WV Geological and Economic Survey, The American Institute of Mining, Metallurgical, and Petroleum Engineers, and American Association for Petroleum Geologists. Data from 1987 to 2025 is from the WV Department of Environmental Protection and the Center for Asset Retirement Accountability’s Upstream Energy platform.

Early regulations advised operators to plug old wells with rocks or wooden plugs. More modern plugging methods and bonding were not put into place until the 1960s and 1970s. Regulatory scrutiny peaked in the early 1980s and passage of the Abandoned Well Act (1992) included additional regulations. Today, regulators are failing to stay on top of accelerating well abandonment, though funding from additional severance tax revenues has provided a much-needed boost to enforcement efforts.

Two critical problems have prevented West Virginia from addressing its growing abandoned well problem: there are too few inspectors, and the state’s ‘bonding’ system has failed to adequately fund well plugging efforts.

Too few inspectors

The West Virginia Department of Environmental Protection’s Office of Oil and Gas (OOG), the state agency in charge of inventorying and inspecting abandoned wells and enforcing plugging requirements, is severely under-resourced and understaffed. As of 2025, OOG employs 18 field inspectors—an untenable ratio of one inspector for every 8,700 wells in the state.

Figure 2: Total inspections have fallen sharply since their height in the 1980s.

Total Oil and Gas Inspections, FY 1965-2023

Source: West Virginia Department of Environmental Protection, Annual Reports FY 1965-67, 1969-1973, 1979-1986, 1998-2004, 2010-2019, and 2023.

Note: Years based on availability. Inspections from 1965 to 1985 include drilling and plugging inspections.

The lack of inspectors has led to fewer inspections, fewer identified violations, and an incomplete database of wells:

• Inspection and violation frequency is far below neighboring states and even historical highs. West Virginia conducted 5,929 total well inspections in 2023, well below its peak of roughly 20,000 in 1985, and just a fraction of Pennsylvania’s total 2023 inspections (44,746). Pennsylvania regulators issue a notice of violation for nearly 23% of all inspections; West Virginia regulators issue violations for just over 1% of all inspections.

• A 2012 audit revealed OOG was “not requiring operators to decommission their abandoned wells.” In the last decade, at least 6,000 wells that were legally required to be decommissioned have been abandoned and left unplugged.

• OOG’s abandoned wells database falls tens of thousands of wells short of expert estimates. OOG’s database lists roughly 16,000 abandoned wells, including about 4,500 orphaned wells without a known or solvent operator. Analysts say there are more than 50,000 unplugged abandoned wells in the state—a massive discrepancy, and one due largely to the lack of documentation of wells drilled before 1929 and subsequent shortcomings in regulatory oversight.

The failures of bonds

West Virginia has a surety bond-based system for the financial assurance of oil and gas wells. Operators looking to drill a new well pay an insurer a small annual premium on a modest bond amount set by OOG, and the insurer agrees to pay the state if the operator fails to perform its legal duty to decommission the well at the end of its producing life.

But bonding amounts have failed to keep pace with the cost of proper well decommissioning. Today, a single well bond for a conventional, vertical well is set at $5,000, while a blanket bond covering all of a company’s wells is set at $50,000. The average plugging cost of a conventional well is $95,000.

Figure 3: Bonding amounts cover just a fraction of the cost to plug a single well.

Vertical well bonding amounts vs. average decommissioning cost, 2025

It’s far cheaper for operators to continue paying the small annual bond premium than to pay for the cost of decommissioning a well. And with inspections so infrequent due to insufficient regulatory staffing, many are able to skirt their plugging operations altogether. Enforcing the law—e.g., recovering bonding funds from an insurer or making sure operators cover cleanup costs—can become a lengthy legal process requiring investigations, site visits, and a large amount of staff time. In many cases, OOG doesn’t have the resources to make this happen.

The flaws in West Virginia’s bond-based system have forced regulators to pursue long-term consent agreements with operators to try to address the state’s abandoned well problem. These consent agreements require operators to plug or reactivate a set number of wells within ten years while shielding them from immediate plugging requirements.

Consent agreements have led to some well decommissioning, but they have failed to stem the tide of abandoned wells. From 2013 to 2024, West Virginia signed 88 consent agreements with dozens of operators, ordering about 4,500 wells to be plugged or reactivated in that time. OOG records show just 3,424 wells were plugged in that time, while the documented total of abandoned wells continued to grow by thousands.

Opportunities and recommendations:

Fixing West Virginia’s abandoned well problem presents an opportunity to boost the state’s economy, revitalize oil and gas employment, and protect communities and the environment.

1. Decommissioning the state’s remaining orphaned well inventory could support thousands of jobs and much-needed tax revenue. Remaining funding from the Bipartisan Infrastructure Law could decommission 1,393 orphaned wells through 2031, supporting an estimated 810 total job-years, according to IMPLAN modeling. Decommissioning an additional 4,000 orphaned wells from 2032 to 2041 would create more than 2,000 job-years, or 207 jobs per year over the ten year period. These jobs could help employ hundreds of oil and gas workers who have lost their jobs over the last several years as the drilling boom has subsided.

2. A production fee on oil and gas will help procure enough funding to quickly and completely address West Virginia’s spiraling abandoned well problem while growing the local economy.

3. The federal, bipartisan Orphan Well Prevention Act of 2025 offers a model for West Virginia policymakers. The act outlines a production fee and requires oil and gas operators to set aside funds to decommission wells as they are completed for production.

By Nick Messenger, Senior Researcher & Economist

Cheap energy has always been good politics. Almost every American voter feels the pinch when heating and electric bills rise. Like the snake oil salesmen in the Old West, American politicians have always confidently sold too-good-to-be-true corporate concoctions and foreign policy poultices to cure our energy ailments.

The problem is that political slogans don’t capture the complex reality of reliably delivering energy to hundreds of millions of people on demand. For example, “energy independence” was once an evergreen applause line at the rallies of both major American political parties. As the governor of California in 1974, Ronald Reagan warned that failure to achieve energy independence would lead to “coercion and blackmail” by foreign adversaries. Thirty years later, in 2004, both Senator John Kerry (who would later become the Special Presidential Envoy for Climate in 2021) and his opponent, President George W. Bush, campaigned by promising to deliver energy independence.

It didn’t happen in 1974, nor in 2004. But by tapping into the popular and quintessentially American belief in self-sufficiency, the slogan of “energy independence” conveyed a tantalizing, albeit false, premise: if only America could wean itself off Middle Eastern oil, then our energy prices would automatically come down and our country would be somehow immune to the turmoil of foreign affairs.

Now, nearly half a century since the term was first in vogue, America has, against all odds, theoretically achieved energy independence. (In practice, technical refinery complications mean that we still need to effectively ‘swap’ our oil on the global market). We have become the world’s top oil and gas producer. But instead of getting cheaper, energy prices have stubbornly climbed higher for American households. Why?

Our last election was no different than 1974 or 2004 in this regard. “Unleashing American Energy” joined “energy independence” and “drill, baby, drill” as the latest slogan from the campaign trail. As a candidate in 2024, President Donald Trump promised that his policies would cut Americans’ utility bills in half within one year of his return to the White House. His administration has, uncreatively, worked to fulfill his promise by following a failed playbook for ramping up oil and gas production, even as the U.S. has already hit record levels of output.

But despite the President pushing for even more fossil fuel extraction by fast-tracking oil and gas development on federal lands and proudly unveiling plans for giant new foreign-owned gas power plants, most of us are paying more for our utilities one year later, not less. My own natural gas bill rose by more than 25% in February as an arctic blast of bitter cold descended on much of the U.S., sending temperatures to near zero for several weeks. But beyond just weather, the story is much the same across the Ohio River Valley, where electricity and natural gas prices have soared in recent years under both Presidents Biden and Trump. Even West Virginians, who live in the heart of Appalachian gas and coal country, have had to sacrifice other basic needs to keep their lights and heat on.

How is it that America can be producing more energy than ever before, yet our utility bills continue to rise? And, more importantly, what is the likelihood that this administration will be able to deliver on its lofty campaign promise by sticking with its current policies?

It needs to be said upfront that our energy system is almost infinitely complex. It’s so complex, in fact, that you could spend a lifetime studying it and still not understand all of the intricacies and idiosyncrasies. One of the reasons I enjoy being an economist, though, is that even the most complex systems have to stand up to logical scrutiny and the basic forces that drive our economy. The numbers, so to speak, still have to add up, and incentive structures that politicians create with regulation and tax policy still need to align with real human behavior.

And, from this economist’s point of view, there are several glaring reasons to be doubtful about the ability of Washington’s new energy policies, touted as “unleashing American energy,” to deliver lower bills, and thus more disposable income, for Americans.

1. Exporting more liquified natural gas (LNG) puts Americans in direct competition with foreign consumers for gas, raising prices.

The promise that we can simply drill and frack our way to lower energy prices relies on a bad assumption. While introductory economics suggests that drilling more and increasing the supply of fossil fuels theoretically reduces prices, that is only true if the demand for energy grows more slowly than the supply of gas. These simplified scenarios are illustrated in the figure below, which may give you flashbacks to (or nightmares about) economics class. Nevertheless, their explanatory power is important, and their lesson is critical to debunking this false promise.

In reality, demand is always changing simultaneously with supply. And demand plays a huge role in price determination. This is because 1) demand can (and often does) grow faster than supply chains can ramp up to meet it, and because 2) more drilling costs companies more money, which makes expansion profitable only when demand and prices are already relatively high or rising. This is made even more complex by the fact that oil and gas companies often operate their lowest-cost, highest-output assets first, making additional drilling marginally more expensive when they do have to ramp up to meet more demand. A nuanced, but important point that can be summarized by: drilling more is typically more expensive for companies.

Private energy companies simply aren’t going to voluntarily or preemptively choose Scenario 1 in the image above. If demand were hypothetically constant, “drill, baby, drill” would flood the market for oil and gas, deflate prices, and reduce the company’s own profit margins. Why would they do that?

Such a scenario would be a great windfall for us consumers, but in the real world, production decisions tend to lag price changes. Companies look to increase oil and gas production after demand goes up, which doesn’t really help the average household’s bill. A more realistic example is shown in Scenario 2 above, where demand growth outpaces supply growth and prices still increase, even with more drilling.

For real-world evidence, look no further than this winter, when natural gas prices have spiked despite a general upward trend in gross natural gas withdrawals (supply) over the last 2 years. In the graph below, I chart the daily Henry Hub spot price in blue from January 2023 to January 21, 2026, and then scale and overlay U.S. natural gas withdrawals in red on the right-hand axis.

The dotted trendlines for the data have a positive slope and suggest that both the supply of gas (either from production or storage) and the price of natural gas are increasing. This is precisely the opposite of the “drill, baby, drill” argument in Scenario 1 that politicians make to justify more fracking and drilling. We are pumping more oil and gas than ever before, yet prices have gone up, not down. Why?

Simply put, demand has gone up faster.

Many factors determine the demand for natural gas, but two likely culprits can be identified in the January 2026 short-term energy forecast from the U.S. Energy Information Administration (EIA), which predicts further rising prices and storage declines in the future.

Electricity generation and pipeline exports have ticked up, but the war in Ukraine and subsequent loss of Russian gas for Western European countries have helped to spur a surge in U.S. liquified natural gas (LNG) exports. The same snake oil salesmen who sold the promise of American energy independence are now selling that same American energy outside of America. And simply fracking for more gas won’t solve our affordability problems if the new gas is sold abroad.

During the Biden administration, pauses on new LNG export terminals were enacted, with President Biden explicitly noting at the time that the impact on energy costs for Americans was a factor that would be evaluated by the government during the pause. The negative environmental and climate impacts of the huge LNG terminals, such as in Cameron Parish, Louisiana, where projects have received billions in state and local subsidies, were also a major consideration in the pause. The Trump Administration has since reversed that pause, and LNG exports are expected to continue to rapidly grow. At a basic level, this means that despite U.S. companies producing more natural gas, Americans will find themselves competing for that new supply on the global market, with Europe and Asia poised to import more LNG than ever in 2026, according to Bloomberg News.

Competing in a global market for gas is problematic for U.S. households, since many of the countries buying American gas are already accustomed to paying much higher prices than U.S. consumers are. The figure below shows a sample of average natural gas prices per kilowatt-hour by country from the middle of 2025. If markets proceed toward an equilibrium “global price,” similar to how oil prices work, we would likely see the world price of natural gas settle in the middle between the U.S. price and the higher prices in Europe and the Asia-Pacific.

In other words, consumers in Europe and Asia could still pay more than what we pay in America, on average, and it would still be a price reduction from what they pay right now. Unless we expect private oil and gas companies to turn down customers willing to pay higher prices, Americans may see their costs bid up as export markets grow.

2. Other administration policies, like promoting data centers, enacting tariffs, and opposing clean energy, are creating headwinds for new capacity construction.

First, the proliferation of electricity-guzzling artificial intelligence (AI) data centers across the United States is increasing overall electricity demand and, thus, demand for natural gas used in power generation. Market forecasts have projected that the AI data center sector will continue to grow, with hundreds of millions of dollars in planned data center investment across U.S. communities over the next four years. This is happening even as local communities struggle with the negative consequences and the current lack of a profitable business model for AI companies.

For example, in Homer City, Pennsylvania, a former coal plant is being repurposed into a natural gas-powered data center complex, which has been heralded as an excellent energy investment with no evidence to suggest it will help lower energy prices for residents. And, despite promising over a thousand jobs, the reality is that even the biggest data center projects rarely employ more than one or two-hundred people once they are built. This has led residents to wonder whether the benefits of the Homer City project will actually flow out of state to large technology companies, in much the same way that the largest economic benefits of fracking have generally left the region. It’s worth noting that, if that is the case, all of the Ohio River Valley states currently abate sales and use tax for data center construction and equipment – a significant possible source of revenue for schools, roads, and public services. Abating local tax dollars for projects where the benefits largely leave the state is just not a sound economic development strategy.

Even if data centers are required to generate their own electricity “behind the meter” using private power plants, as the President vaguely announced in his recent State of the Union Address, they would still need to burn natural gas to make that electricity. Thus, even if data centers don’t directly or immediately increase electricity costs on the grid, their additional need for natural gas could still contribute to higher fuel costs for the rest of our power plants. We haven’t seen fuel costs drive up prices significantly yet, with the majority of current electricity price increases being related to transmission and distribution buildouts. But, in time, if the data center buildout materializes at a large scale, we could end up with even higher electricity rates.

Second, even if the future baseline grid load is met by natural gas, renewable energy and storage offers a potential pressure release valve for times of global volatility. While it is a popular political talking point to say that the costs of the renewable energy transition have generally caused higher utility prices, there’s little evidence to support the claim. While generation costs have been relatively flat, transmission and infrastructure buildouts, large capital investments in aging coal plants, geopolitical turmoil, and a lack of supplementation from renewable energy have largely created our region’s energy affordability crisis.

In fact, states with some of the largest shares of their electricity coming from renewable sources have maintained lower consumer prices than areas of the country rich in oil and gas, as shown in the figure below.

I caution that this feat isn’t replicable at the same cost everywhere; the Southwest is much sunnier, and the plains states are much windier than the Midwest, reducing the required capital costs with fewer panels or turbines required to generate the same amount of energy more consistently. A few things stand out, however. First, a few pretty populous states like Texas, Oregon, and Minnesota have achieved below-average residential electricity rates with above-average shares of their electricity coming from wind and solar. And Minnesota and Oregon can hardly be accused of being perpetually sunny places.

I also highlight California on the chart, in the interest of transparency. While it is the most populous state in the country with one of the largest renewable shares of electricity production, it has been far above the average U.S. electricity rate since the 1980s. Explanations for this trend that predate renewable energy adoption include aging infrastructure, wildfire mitigation costs, and pricing by privately owned utilities. It’s also worth noting that Californians also use less electricity than the average American, likely owing to some combination of the temperate climate and energy efficiency measures. I note this because I am not arguing that renewable energy is a “silver bullet” for energy prices or reliability, but it is generally cheaper to build in the short term and a strategic component of long-term price stability.

Renewable energy also has a feature that works to the advantage of consumers in a globalized economy: it is largely consumed locally. Electricity from solar arrays or wind turbines isn’t easily loaded on ships and sold on global markets like natural gas or oil. That means that, when it comes to renewables, Americans could truly be energy independent and wouldn’t directly compete in a global market with foreign consumers. It also means that the price of renewable electricity doesn’t have to factor in the billions of dollars in capital spent by companies to build expensive LNG terminals.

Recent empirical work has also found that increasing renewable energy’s share of power generation in importing countries reduces the overall global demand for LNG, suggesting that they are at least partial substitutes for consumers. This means that investment in renewable energy is another plausible mechanism to reduce overall natural gas demand and lower the price of the gas that America needs at baseline. By phasing out federal funding and tax credits for investments in clean power generation and decarbonization in last year’s One Big Beautiful Bill Act and by promoting the construction of new mega gas-fired power plants, the administration is kicking the ball into its own net, so to speak, when it comes to building an affordable energy future.

The final headwind that the administration has created for itself is the haphazard and broad implementation of tariffs. Tariffs raise prices by either making imports more expensive or by allowing domestic producers to raise their prices to match the new, higher import prices. The Trump Administration’s tariffs went into effect in a staggered fashion, largely in April 2025, May 2025, or August 2025, depending on the industry and country.

I noted earlier that most of the cost of electricity price hikes has been due to transmission and distribution construction—the transformers and power lines that process the electricity generated from power plants and deliver it to homes, businesses, and factories. The graphs below look at the producer price index (prices received by the seller) for three related industries: electricity transmission, control, and distribution, turbine generator sets, and construction material manufacturers.

In all three cases, large price spikes begin after April, May, or August in 2025, which generally corresponds with the administration’s tariffs taking effect. This suggests that the administration’s tariffs have at least partially contributed to our rising utility costs by making the construction of new gas generation and new transmission and distribution infrastructure more expensive. Evidence shows that these higher costs for utility companies are ultimately passed on to household and industrial consumers in the form of higher prices, fees, or riders.

In a sense, while the administration may say that they are pursuing policies to “unleash American energy” with their left hand, they are rapidly attaching new chains with their right hand through a variety of misguided policies. These include:

• Promoting LNG export terminal expansion to sell more natural gas abroad.

• Encouraging data center construction by tech companies, which compete privately for natural gas, even with behind-the-grid electricity generation.

• Ending incentives for investment in less-volatile, cheaper, and locally consumed renewable electricity.

• Enacting global tariffs that raise the costs of the materials and equipment needed to grow and expand our electric grid.

Combined, these policies mean that even if we “drill, baby, drill,” the gas that we produce will very likely either be sold on a more expensive global LNG market or the plants, transformers, and powerlines needed to convert it into usable electricity will cause more prices increases, offsetting the effects of extra production.

The truly bleak scenario is the one where these forces all work simultaneously, perfectly aligning in the future to increase gas costs, raise construction costs, delay timelines, and close the door on alternative, supplemental sources of local clean energy. If you’re a private oil and gas company or a shareholder, this may be just what the doctor ordered, but for the rest of American households and businesses, this policy concoction will likely become more poison than cure, leading to even higher utility bills in the future.