Economics and Retirements of Existing Nuclear Power Reactors

Key points

1. Almost 10 gigawatts of U.S. nuclear power plants have retired in the face of market competition and aging facilities

2. Competition with natural gas in restructured wholesale markets is the primary source of economic pressures for most plants

3. Nuclear plant retirements increase greenhouse gas emissions and other air pollutants from coal and natural gas generation, while also leading to higher energy prices

4. Keeping existing nuclear power plants through at least mid-century provides a foundation for future decarbonized energy grids

5. State and federal governments are introducing policies to prevent the retirements of additional nuclear power plants, in large part to meet climate goals

Amidst a continent-wide energy crisis that is largely caused by insufficient natural gas supplies, Germany continues to rapidly shut down its nuclear power plants. Beyond the economic consequences of growing EU dependence on natural gas imports, dependence is also a growing security threat that Russia may leverage as it seeks political and potential military dominance of its neighbors like Ukraine. Although some European countries are more supportive of nuclear power, others are not and are following Germany’s lead. Germany’s policy decision makes clear the potential consequences for the United States, where many reactors are now retiring due to economics and other factors.

Since 2012, American utilities have permanently shut down 12 nuclear reactors due to a variety of causes with another 3 scheduled to close by 2025. Although there are specific economic and political factors shaping each individual retirement decision, the ultimate effect is shrinking the U.S.’ largest source of carbon-free clean energy generation just as the nation sets its sights on decarbonizing energy grids. In many cases, retiring nuclear plants are replaced by natural gas or even coal, leading to higher energy prices and emissions.

Increasingly, these retirements are becoming national-level politics, with key Senate swing vote Senator Manchin recently expressing support for existing nuclear facilities as part of broader climate-focused legislation.

Despite tremendous growth in solar and wind power the last two decades, they still provide less generation than the nation’s remaining 93 nuclear reactors, which provide about 20% of total U.S. power supply.

This post reviews the causes of retirements, their retirements, and industry and policy actions to prevent them. It highlights the importance of keeping as many reactors as possible online through mid-century to meet national environmental and economic goals.

Retirements of U.S. Nuclear Power Plants

In the last ten years, twelve reactors with a combined capacity of almost 10 gigawatts have retired. Three more (Palisades and Diablo Canyon 1+2) are scheduled to close before 2026.

These retirements are due to several factors. The first wave of retirements, primarily occurring in 2013, were due in large part due to maintenance issues due to aging facilities. The average nuclear plant in the United States is forty years old. The retirements of Crystal River 3 and San Onofre were directly related to botched maintenance to keep the aging facilities running. In both cases additional investment could have probably kept the plants online, but plant owners considered the costs prohibitive.

Economics were moderate factors in the latter retirements in this wave (Vermont Yankee and Fort Calhoun), due in part to the small relative size of these facilities. However, these two plants also featured unique issues, with Vermont Yankee facing significant state-level political opposition and Fort Calhoun suffering severe outages that cut it capacity factor well below 60% in its final years of operation.

Source: CRS

The second wave of retirements, largely occurring between 2018 and 2021, were driven primarily by economic competition with natural gas in wholesale markets. Except for Duane Arnold in Iowa, 5 of these 6 retirements occurred in Eastern competitive markets (PJM, NYISO, and ISO-NE) where Marcellus shale gas greatly reduced electricity prices while greenhouse gas and capacity markets proved insufficient to make up the shortfall.

Several general observations can be made about the twelve retirements to date:

• Many retirements were single unit nuclear power plants, reflecting less favorable O&M costs compared to multi-unit facilities

• Similarly, smaller reactors were more likely to close than larger ones

• Almost all retirements (11 of 12) occurred in restructured wholesale power markets (ISOs), largely reflecting the dominance of short-term economics in market outcomes in those markets

• Renewable energy sources were negligible competitors in almost all retirements, with natural gas often the cause of the retirement and the replacement fuel of choice

Although economics were primary factors for many retirements, political considerations also factored in for some. The initial construction of many nuclear power plants was politically contentious for many communities, and these fraught relationships continued to the present. Such political considerations were key factors in the Vermont Yankee and Indian Point 3 retirement, and are also an underlying factor in the political debate about the planned retirement of Diablo Canyon in California.

Economics of Operating Reactors

With economics a primary factor in plant closures, it is worth reviewing the economics of operating an existing nuclear power plant. The cost profile for a nuclear reactor can be described as high capital costs with low-to-moderate operational costs. Construction of a conventional new reactor is a mega-project, costing well over a billion dollars and taking more than a decade.

Once a reactor is built, it has operations, maintenance, and personnel costs but these are often fixed over the course of year. There are periodic large expenditures, like replacing generators or other large equipment, and these can create decision points for whether to retire or invest.

There are some variable costs for nuclear energy, related primarily to uranium fuel costs. Unlike fossil fuels, whose overall costs are dominated by fuel costs, uranium only plays a small role in plant economics. In most cases, prevailing uranium prices lead to fuel costs of only around $5/MWh for most American nuclear plants (versus $20-100+/MWh for fossil facilities).

Accordingly, in marginal-dispatch energy markets, this cost profile means that whenever a nuclear plant can run, it will run. It has low marginal costs and dispatch markets seeking the lowest energy price in a specific hour will usually dispatch the unit. Only capital-intensive renewable energy sources like wind, solar, and hydro have lower marginal costs than nuclear energy.

In effect, this means that nuclear power plants are price-takers in energy markets. Their revenues are not necessarily dependent on the full social or economic value of nuclear power, but rather market outcomes from short-run marginal competition between natural gas and coal.

In U.S. energy markets, natural gas or coal power is the marginal price setting in almost all hours. With the rise of shale gas, wholesale power prices have collapsed from $50-100/MWh in the late 2000s, to $50/MWh or less for most of the 2010s.

Source: EIA

The use of marginal dispatch wholesale markets for generator compensation can give rise to a “missing money” problem. Essentially, marginal dispatch for what is cheapest in each individual hour does not necessarily ensure system reliability. Since the grid needs to provide service in all demand scenarios, low prices for most hours may not provide enough incentive for new or continued investment in generation capacity. The cost profiles of certain types of energy units, like nuclear energy or renewable energy, are not driven by marginal economics like fuel costs.

This gives rise to capacity markets or other resource adequacy mechanisms. With the noted exception of ERCOT in Texas, all other ISOs in the U.S. have some sort of mechanism designed to maintain reliability with a marginal dispatch market.

However, these mechanisms may not be sufficient to ensure reliability or fully compensate generators for the value of their generation. PJM, ISO-NE, and NYISO all operate capacity markets and yet have seen the lion’s share of nuclear retirements. These capacity markets are administrative constructs, are constantly being redesigned at a regulatory level (undermining market certainty), and suffer general controversy over the year-to-year variations in price outcomes.

Further, all three regions are shedding historic oversupply from when utilities had extra capacity before regional grid integration. In effect, capacity markets have not provided sufficient revenues to stave off nuclear retirements because there is sufficient summertime capacity, assuming all capacity is equal (which is not necessarily the case in extreme market conditions with natural gas).

A final point on market conditions. Generally, renewable energy is not a primary driver in recent or near-term prospective retirements. Many nuclear advocates decry renewable energy subsidies and market competition as unfair and sometimes argue it is a key factor in why nuclear plants are struggling.

Except in a few cases, where transmission constraints are especially severe in regions with excellent wind resources, the economic impact of renewables on nuclear revenues is limited. In Eastern power markets, where most retirements are occurring, economic analysis indicates that lower revenues are almost entirely due to natural gas competition.

Carbon Implications of Nuclear Retirements

During operation nuclear energy does not produce greenhouse gas emissions. Even accounting for the carbon input to build a new facility and emissions across the supply chain, nuclear energy has one of the lowest lifecycle emissions of any energy source. Accordingly, theory and experience in energy markets would indicate nuclear energy closures could increase emissions.

Analysis of state and regional grids confirms that retiring nuclear plants are replaced by fossil generation. In New England, the closure of Vermont Yankee led to region losing 15% of its carbon-free power, which led to increased greenhouse gas emissions as it was primarily replaced by natural gas and oil generation. An analysis by the Breakthrough Institute found that the closure of San Onofre in California was primarily replaced by natural gas, leading to a 37 million metric ton increase in greenhouse gas emissions. A similar analysis from Energy Institute at Haas also found that the generation was primarily replaced by natural gas, and by increasing gas demand led to much higher energy costs for the state.

Source: EIA

Despite anti-nuclear advocates claiming that nuclear plants can be replaced by renewable energy, energy market outcomes the last decade indicate that this is rarely or never the case. Plans and actions by utilities and ISOs all clearly indicate nuclear generation is primarily replaced by natural gas, and sometimes coal and oil. Further, if renewable generation were built with the express purpose of replacing nuclear energy, which has not happened in any retirement to date, it would miss out on the opportunity cost of replacing additional coal or natural gas.

Critically, nuclear units are long-lived electricity infrastructure. The average age of an existing nuclear reactor is about forty years old, most plants now have licenses to operate to at least sixty years and some plants are now seeking an additional extension to eighty years. The Nuclear Regulatory Commission and industry are now investigating whether nuclear plants can safely operate as long as a hundred years.

Given that most nuclear reactors are larger than 1 gigawatt, a single reactor closing before it reaches 100 years could mean the loss of at least 54 gigawatt-years of carbon-free generation (assuming a 90% capacity factor).

This is equivalent to ~473 terawatt-hours between now and 2080. In 2020, wind and solar combined produced about 420 terawatt-hours.

This means that retiring just one nuclear reactor loses more potential clean-energy generation than all existing renewables produce annually.

Even if the nuclear unit only operates until 2050, if it primarily replaces natural gas its operation avoids lifecycle emissions of more than 100 million tons of CO2-equivalent. These are the climate stakes of retiring nuclear plants early.

Three recent analyses illustrate the economic pressures that nuclear units face and their emissions implications:

• A 2017 study by Roth and Jaramillo evaluated the economic competitiveness of existing nuclear plants and whether preventing retirements provides cost-effective carbon mitigation. Finding risks of many retirements by 2040, the study found that the cost of avoided CO2 from keeping a nuclear plant online ranged from $18-97 per metric ton. Notably, the higher numbers here were from the least economic, single-reactor units. Multi-reactor units could be kept online for $18-30 per metric ton.

• In 2018, the Union of Concerned Scientists performed a similar analysis that found that, while many existing plants were profitable, more than a fifth of U.S. capacity was unprofitable or scheduled to close. Their analysis further indicated that if that many nuclear units retired, it would almost completely stall progress on decarbonizing the power sector in the U.S. through at least 2035

• A recent 2021 analysis by S&P Global Platts confirmed that more than 30 gigawatts of nuclear units were at risk of retirement. Their analysis also identified many plants where state level support has been effective in reducing retirement risks.

Source: UCS

Beyond the direct effects from a nuclear retirement, reducing the overall size of the nuclear fleet is likely to increase the nationwide costs of decarbonization. Research has shown that removing technology optionality can increase total decarbonization costs by 11-76%. Portfolio approaches are most cost effective, while also ensuring that we maximize shots on goal in terms of reaching fully decarbonized energy systems as soon as possible. One recent study estimated that keeping the nuclear fleet online for as much as 100-year lifetimes had a net present value to the United States of up to $500 billion because of decarbonization benefits.

Other Implications of Nuclear Retirements

In integrated and diversified energy systems, nuclear power can provide significant benefits across a range of metrics beyond climate performance. Any energy systems analysis or policymaking regarding existing nuclear plants should keep in mind:

• Greenhouse gas emissions are not the only air pollutant that endangers public health – nuclear prevents NOx, SO2, and toxics emissions from coal, oil, and natural gas

• With moderate-to-low operating costs, conventional reactors form the backbone of national energy grids, keeping prices low through a merit order effect (much like that of renewables)

• Due to energy density, nuclear power has low material and land use requirements

• Nuclear power provides good jobs, with some of the highest wages, most unionized workforces, and positions for veterans

• Many countries have adopted nuclear energy for the energy security benefits of not requiring imports of fuels, like natural gas

Beyond greenhouse gas emissions, the prevention of air pollution is worth exploring in detail. A report by the Clean Air Task Force found that closing four nuclear plants in Illinois would cause enough air pollution to cause 1,200-2,700 premature deaths, 140,000 lost work days, and over 30,00 additional asthma attacks. Monetized, these damages would be equivalent to $1-2.4 billion annually, more than enough to justify keeping the plants online irrespective of the carbon benefits.

Given their relatively large size, nuclear plants play a critical role in local economies, many of which are in rural areas and lack ready job alternatives. Losing a nuclear plant means that cities lose tax revenues and local services may be shuttered. Hundreds or even thousands permanent direct or indirect jobs are lost, as are the economic benefits from the thousands of seasonal workers that visit during a once-every-two year refueling.

When Vermont Yankee closed, the town of Vernon lost half its tax base, as well as a core part of its identity. The 2020 closure of Indian Point in New York will lead to the local school district losing one third of its funding by 2024. Lacey, New Jersey lost as much as 42% of its budget and hundreds of jobs from the closure of Oyster Creek.

Beyond tax revenues and jobs, nuclear plants are often good local corporate citizens and communities also lose out on charitable donations and other support. Given the local economic benefits from nuclear plants, local communities are generally supportive of nuclear power (as opposed to anti-nuclear opponents, many of whom are not locals).

Even as national policy needs to focus on supporting coal communities suffering from the downsides of the energy transition, it needs to also support nuclear retirement communities. Better yet, we should avoid the retirements in the first place.

Nuclear plants are also critical for regional reliability. In a 2018 report on the impact of nuclear and coal retirements on reliability, NERC (the entity responsible for North American electric grid reliability) found:

“The key conclusion is that generator retirements are occurring, disproportionately affecting large baseload, solid-fuel generation (coal and nuclear). If these retirements happen faster than the system can respond with replacement generation, including any necessary transmission facilities or replacement fuel infrastructure, significant reliability problems could occur. Therefore, resource planners at the state and provincial level, as well as wholesale electricity market operators, should use their full suite of tools to manage the pace of retirements and ensure replacement infrastructure can be developed and placed in service. Again, ensuring reliability throughout a significant retirement transition will likely include construction of new transmission and fuel infrastructure.”

Renewables are a critical part of the energy transition but unlocking their full potential requires greatly expanded transmission. Perhaps the only thing harder to build than a nuclear facility is a transmission line and the amount of transmission needed for decarbonization scenarios is well beyond what is currently planned.

If nuclear units are retired quicker than replacement infrastructure, the reliability implications could be substantial. This is especially true as much of the replacement capacity for retiring nuclear units is natural gas and we are still struggling with integrated electricity-natural gas reliability planning.

State Policy to Prevent Retirements

Recognizing the environmental, economic, and public health fallout of closing existing nuclear power plants, state legislatures and regulators in six states have taken significant action in the last several years to prevent economics-driven closures. Nuclear energy plays a major role in general and clean energy supply in many states. Twelve states receive more than 30% of their in-state generation from nuclear power.

In sum, a total of 20 reactors with a combined capacity of 19.4 gigawatts have received support from state policies to keep them online.

State support generally takes the form of environmental payments, either through direct programs or expanding state renewable portfolio standards into clean energy standards. States acting include:

• Connecticut

• Illinois

• New Jersey

• New York

• Ohio

• Pennsylvania

Source: CRS. Note that this map does not include the 4 reactors in Illinois (Byron and Dresden) supported by Illinois’ recent clean energy legislation.

There is an interesting economic question about whether such payments are “subsidies” that unfairly distort markets. In my view, no. These payments are just like renewable energy credits. They are compensation to plant owners for the environmental attributes and public health benefits of nuclear energy. Further, these policies are usually structured so that utilities are the compliance entity, meaning the costs fall on ratepayers and not taxpayers.

As currently structured, wholesale markets are administrative constructs. These constructs typically value three things: energy supply in megawatt-hours in hourly auctions, capacity availability in annual auctions, and ancillary services related to the operations of the grid. Federally-regulated markets do not account for the cost of carbon or other environmental externalities. Thus, the state payments to nuclear units are ensuring that plant owners fully realize the value of their facilities for society. They internalize the true economic value of nuclear power.

This view is supported by recent case law. There were initially concerns about whether state support for nuclear energy was legal under state energy policy authorities and whether it unduly interfered with federal regulation of wholesale markets (per the Supreme Court in Hughes v. Talen Energy Marketing).

In the last several years, state policy efforts supporting existing reactors have withstood multiple legal challenges, with courts finding:

1. States’ generally have broad powers to support nuclear power for certain reasons

2. Recent policies do not deal with areas preempted by federal regulation of wholesale power markets

In part, these rulings were based on the design of state support – generally state policies are compensating for the environmental attributes of nuclear power, an area for which FERC has yet to claim or exercise federal jurisdiction. Unlike Talen, where Maryland attempted to subvert PJM’s federally regulated capacity market, Circuit Courts have found that states are acting on the basis of environmental attributes and other state powers.

One interesting development with recent state policy debates around existing nuclear is the emergence of nuclear energy advocacy groups advocating for keeping plants online. A key part of successful advocacy is allying nuclear and renewable energy political interests. Nuclear plants are often owned by dominant utilities in a state, which in many cases historically opposed renewable policies because of concerns about market share.

By combining utility support, union support from nuclear workforces, and environmentalists, successful coalitions can support clean energy action. This was most recently seen in Illinois, where state support for Byron and Dresden was passed as part of a 100% carbon-free by 2050 standard.

That said, just because state support is needed to keep existing reactors online and is legal, it is not always politically the right choice. This has been demonstrated most clearly in Ohio, where financial support for nuclear plants was controversially coupled with support for coal plants in H.B. 6.

Not only does such an approach undermine the climate argument for nuclear energy, it also prevents coalition building across clean energy interests.

Later, it was discovered the utility involved, FirstEnergy, was bribing state legislators to pass the bill, leading to significant criminal charges and fines. Even though H.B. 6 provided funds to both nuclear and coal plants, it was often labelled as a nuclear scandal. In March 2021, the nuclear funding from H.B. 6 were repealed, even as the coal funding remained.

Federal Legislation to Support Existing Reactors

Given the national-level importance of keeping existing reactors online for their carbon and other benefits, there have been growing questions for years about the role of federal policy in preventing nuclear retirements. The role of existing nuclear energy in President Obama’s Clean Power Plan was hotly debated before President Trump repealed the approach.

In 2021, the $1.2 trillion bipartisan infrastructure bill provided the first major federal support to keep reactors online with a $6 billion civil nuclear credit program. Originally part of the American Nuclear Infrastructure Act, the program directs the Department of Energy to run a reverse bid auction to fund nuclear plants that are expected to close due to economic factors.

If a bid is successful, credits are provided for four years. After this initial period, the plant should target economic competitiveness or seek a lower credit amount than originally awarded for an additional four years. The program is set to end by 2031.

Similarly, although its fate remains uncertain, the reconciliation bill passed by the House of Representatives as part of “Build Back Better” in November 2021 contained a “Zero-Emissions Nuclear Power Production Credit.” The specific amount of the credit is based on a relatively complex formula based on plant electricity revenues, state support, and other factors.

Given Senator Manchin’s interest in nuclear noted at the top of this article, this provision or something similar could be expected if any climate version of the Build Back Better agenda passes.

Future of Nuclear Power in the U.S. through Mid-Century

Over the last ten years, the U.S. has lost almost 10 gigawatts of nuclear reactors with another 3 gigawatts scheduled to retire in the next three years. Although some of these retirements (like Crystal River or San Onofre) may have been inevitable due to severe maintenance and aging issues, most of them are due to economics. Retiring facilities were more likely to be in restructured competitive markets, have only one reactor, and be smaller than the remaining fleet.

While remaining operational reactors are likely to be more competitive, keeping them online is critical to the future of U.S. decarbonization goals.

States have changed policies to provide financial support for almost 20 gigawatts of reactors in the most economically competitive states, helping allay the most severe scenarios of nuclear retirements envisioned several years ago. Recent market developments, particularly the highest U.S. natural gas prices in years due to LNG exports, could bolster plants without additional intervention.

Nevertheless, additional federal funding, state policy action, and industry actions to increase competitiveness are needed to keep the existing nuclear fleet providing clean power.

A back-of-hand calculation illustrates the stakes.

If every existing reactor not scheduled for retirement can make it to 100 years of operation, they would collectively produce ~45,000 terrawatt-hours of carbon-free electricity. This is more than 11 times annual power demand in the U.S.

Assuming this generation solely displaces natural gas (even though it also displaces coal), nuclear plants would prevent the emissions of 17.7 billion tons of CO2, which is almost 3 years of total U.S. net annual emissions and 12 years of power sector emissions.

Additional resources

• Congressional Research Service: U.S. Nuclear Power Plant Shutdowns, State Interventions, and Policy Concerns

• Union of Concerned Scientists: The Nuclear Power Dilemma: Declining Profits, Plant Closures, and the Threat of Rising Carbon Emissions

• Carbon Brief: Mapped: The US nuclear power plants ‘at risk’ of shutting down

• Kim, Taiwo, & Dixon: The Carbon Value of Nuclear Power Plant Lifetime Extensions in the United States