The United States has not broken ground on a new nuclear reactor that actually reached completion on schedule in decades. The two units at Plant Vogtle in Georgia, the only new commercial reactors to come online in a generation, took roughly twice as long to build as planned and cost more than $30 billion, nearly double the original estimate. That track record is the backdrop for what the White House announced on May 23, 2025: four executive orders directing federal agencies to quadruple the nation’s nuclear generating capacity, from roughly 100 gigawatts today to 400 gigawatts by 2050.
If achieved, the target would make nuclear power the backbone of the American electricity system. It would require building more new reactor capacity in 25 years than the country constructed during the entire original nuclear era of the 1960s and 1970s. The orders restructure how the Nuclear Regulatory Commission licenses reactors, how the Department of Energy tests advanced designs, and how the Pentagon procures nuclear power for military bases and data centers. Nothing this ambitious has been attempted in U.S. energy policy before, and the gap between the directive on paper and the reality of pouring concrete is where the story gets complicated.
Where the numbers start
The starting point is well established. According to the Energy Information Administration, U.S. utilities operated 94 commercial reactors with a combined net generating capacity of nearly 97 gigawatts as of 2024. The White House and DOE round that to 100 gigawatts in official communications, a small gap that means the actual expansion needed is slightly larger than the 4x ratio suggests.
The 400-gigawatt target is written into the text of the executive order on NRC reform, which gives the commission 18 months to complete a wholesale rewrite of its licensing regulations. The order frames the current regulatory structure as the primary bottleneck, arguing that licensing timelines measured in years rather than months have driven away private capital. The NRC has already begun publishing Federal Register notices tied to the rulemaking, a signal that the agency is treating the deadline as binding.
That NRC overhaul builds on momentum from the bipartisan ADVANCE Act, signed into law in 2024, which had already directed the commission to modernize its licensing framework and reduce review timelines. The executive orders go further and faster, but they did not emerge from a vacuum.
The near-term milestones
A companion order focused on rebuilding the nuclear industrial base sets two targets for 2030: the DOE is to facilitate 5 gigawatts of power uprates at existing reactors, and 10 new large reactors with complete designs are to be under construction. The order also steers DOE Loan Programs Office financing toward reactor projects and emphasizes domestic manufacturing of reactor components and fuel cycle services, tying nuclear expansion to broader industrial policy.
The uprate goal is the more achievable of the two. Uprates involve engineering modifications to existing plants, not new construction, and the NRC has approved hundreds of them over the past three decades. Five gigawatts across a fleet of 94 reactors is aggressive but within the range of what the industry has done before.
Ten new large reactors under construction by 2030 is a different proposition. As of mid-2026, no new large reactor project in the United States has reached the construction permit stage since Vogtle. Site selection, environmental review, design certification, and construction licensing have historically taken a decade or longer. Compressing that into roughly four years would require a pace the U.S. nuclear industry has not demonstrated in the modern regulatory era.
A third order creates a pilot program for reactor construction and operation outside national laboratories under Atomic Energy Act authorities, and it directs the DOE to expedite or eliminate National Environmental Policy Act reviews for reactor work through categorical exclusions. That testing reform directive is aimed at shortening the path from prototype to commercial deployment for advanced designs, including small modular reactors and microreactors. A fourth order designates the Department of Defense as the executive agent for advanced reactor technologies intended to power AI-oriented data centers and defense installations.
Together, the four orders create a structure in which civilian regulators, energy researchers, and defense planners are all expected to treat nuclear expansion as a central mission rather than a niche program.
The feasibility question
Building 300 gigawatts of new nuclear capacity in 25 years would mean averaging 12 gigawatts per year, roughly equivalent to bringing a dozen large reactors online annually for a quarter century. For comparison, during the peak of the original U.S. nuclear buildout in the 1970s and early 1980s, the country was completing roughly 5 to 8 gigawatts of new capacity per year. France, which executed the most successful national nuclear program in history after the 1973 oil crisis, built about 60 gigawatts over two decades. China, currently the fastest nuclear builder in the world, has roughly 25 gigawatts under construction and has been commissioning about 3 to 5 gigawatts annually.
No country has ever sustained the build rate the 400-gigawatt target implies.
The executive orders acknowledge supply-chain constraints in broad terms but do not quantify them. Reactor-grade steel and concrete, large forging capacity, a trained nuclear construction workforce, and enriched uranium supply are all potential chokepoints. The DOE’s FY 2027 budget request repeats the 400-gigawatt goal but does not include an independent cost estimate or a supply-chain capacity analysis. Without a transparent cost range or a breakdown of assumed build rates by decade, utilities, state regulators, and investors have limited ability to judge whether the targets are realistic.
The Vogtle experience looms over every projection. Georgia Power’s two new AP1000 reactors were supposed to cost about $14 billion and take roughly seven years to build. They ended up costing more than $30 billion and took over a decade. The project’s co-owners abandoned their stakes along the way. Advocates for new nuclear argue that Vogtle’s problems were specific to a first-of-a-kind construction effort and that subsequent builds would benefit from lessons learned. Skeptics counter that cost overruns and delays have plagued nuclear construction globally, from Flamanville in France to Olkiluoto in Finland, and that the pattern is structural, not incidental.
Legal and political friction ahead
The NRC’s 18-month rulemaking deadline is one of the most consequential elements of the orders, and one of the most legally vulnerable. A rulemaking of that scope would normally take years. Environmental and safety advocacy groups have historically used the public comment process and federal courts to slow or block regulatory changes they view as weakening oversight. The directive to use categorical exclusions for NEPA reviews at DOE sites is likely to draw particular legal challenge, since categorical exclusions bypass the detailed environmental impact statements that reactor projects have traditionally required.
If courts determine that the exclusions are too broad or that the rulemaking was rushed, key elements of the accelerated siting and testing strategy could be delayed or overturned. That legal risk introduces a layer of uncertainty that the executive orders themselves do not address.
There is also an open question about technology mix. The orders promote small modular reactors and microreactors as central to the buildout, especially for defense applications and data centers, but no SMR design has yet been built and operated commercially in the United States. NuScale Power, the furthest along in the NRC licensing process, canceled its first planned project in Idaho in 2023 after costs escalated and customers withdrew. If advanced designs encounter further technical setbacks or cost overruns, the burden of meeting the 400-gigawatt target falls back on conventional large reactors, which carry their own well-documented history of delays.
What 400 gigawatts would actually mean
If the target were somehow met, the implications for the U.S. electricity system would be transformative. Nuclear plants run at capacity factors above 90%, meaning 400 gigawatts of nuclear capacity would generate roughly 3,150 terawatt-hours of electricity per year. Total U.S. electricity generation in 2024 was about 4,100 terawatt-hours, according to the EIA. Nuclear would go from supplying about 19% of the nation’s electricity to supplying more than three-quarters of it, a shift that would fundamentally reshape the grid, the utility business model, and the country’s carbon emissions profile.
That scenario would also require massive investment in transmission infrastructure to move power from reactor sites to demand centers, a challenge the orders do not directly address. State public utility commissions, which approve the rate increases that fund large power projects, would need to sign off on costs that could run into the hundreds of billions of dollars. How that cost is shared between federal taxpayers, ratepayers, and private investors remains undefined.
None of this means the orders are empty gestures. The legal directives are real. The NRC is moving. The DOE is reorganizing its priorities. Federal loan guarantees and defense procurement dollars create tangible financial incentives. But the distance between signing an executive order and connecting a new reactor to the grid is measured in years of engineering, construction, regulation, and litigation. The 400-gigawatt target is the most ambitious nuclear energy goal any U.S. administration has ever set. Whether it becomes a construction program or remains a policy aspiration depends on decisions that have not yet been made, money that has not yet been committed, and technology that has not yet been proven at scale.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
