Introduction

On March 4, 2026, the United States Nuclear Regulatory Commission (NRC) unanimously voted to grant a construction permit to TerraPower, a nuclear power startup founded by Bill Gates, allowing construction to move forward for a new nuclear power plant in Kemmerer, Wyoming (“Kemmerer Plant”). The plant is expected to come online in 2031 and will provide power to the Rocky Mountain Power Service area which includes Wyoming, Utah, and Idaho. TerraPower hired Bechtel to serve under an Engineering, Procurement, and Construction (EPC) contract where Bechtel holds a single contract with TerraPower covering design, procurement, and construction. TerraPower and Bechtel just broke ground on the Kemmerer Plant on April 23, 2026. 

The Kemmerer Plant faces significant construction risks. The plant will feature a Natrium reactor, which uses liquid sodium rather than water as a coolant, combined with a molten salt energy storage system. This will be the first plant using this technology at a commercial scale in the United States.  History has shown that nuclear construction projects, particularly first-of-a-kind designs like this one, are uniquely susceptible to construction defects, delays, and cost overruns. The consequences of such issues can lead to regulatory shutdowns, legal disputes, and severe safety hazards.

Lessons from Plant Vogtle: A Blueprint for What Can Go Wrong

For comparison, the most recent nuclear power plants constructed in the United States were the Plant Vogtle Units 3 and 4 in Georgia, which opened in 2023 and 2024—the first such plants built from scratch in the United States in over 30 years. Yet Vogtle’s path to completion was defined by continuous issues. The project ran seven years late and more than $17 billion over budget. One estimate found that each month of construction delay cost an additional $90 million in capital costs.

In April 2012, NRC inspectors discovered that installed rebar failed to meet the approved plans. Addressing the issue caused project delays as a plan amendment had to be run back through the U.S. Nuclear Regulatory Commission. Separately, welds on vessel nozzles were found to be defective, requiring excavation and re-welding. Four welds in the containment vessel were later discovered to be cracking, again triggering reconstruction.

Perhaps most consequentially, the NRC in 2021 issued an inspection finding that Southern Nuclear had failed to properly separate safety-related and non-safety-related cables for reactor coolant pumps and shutdown equipment. The result was that most of the plant’s wiring had to be completely redone before the NRC would authorize fuel loading. These are not isolated incidents, but manifestations of the systemic quality assurance challenges that accompany large-scale nuclear construction. Even when discovered and addressed during construction, defects can impact a Project, including by creating delays or increasing costs (for one party or another). Such issues can impact any project but are especially noteworthy in nuclear projects given their scale and heightened regulatory requirements.

TerraPower’s new project, as the first commercial deployment of a new reactor technology, faces all the same risks as Vogtle, and more.

International Precedent: Flamanville 3

Nuclear construction defect problems are not confined to the United States. Flamanville 3 in northern France was planned for completion in 2012 at an estimated cost of €3.3 billion, but ultimately ran 12 years and €20 billion over budget. Électricité de France (EDF) in spring 2018 announced the discovery of numerous faulty welds in the secondary cooling circuit, resulting in 53 welds requiring replacement. Because some of these welds were in hard-to-reach spaces, custom remote-controlled robots had to be created and deployed. This welding defect further delayed the project by three to four years and increased project costs by an estimated €1.5 billion. Flamanville demonstrates that even in nations with deep nuclear expertise, construction defect impacts on a new reactor design are nearly unavoidable.

The First-of-a-Kind Challenge

A First-of-a-Kind project is defined as a project that is essentially different in scope or in detail from previous experience. The Kemmerer Plant’s Natrium reactor is very clearly a first-of-a-kind project: no commercial-scale sodium-cooled fast reactor has ever been built in the United States, and this is the first construction permit ever issued by the NRC for a commercial non-light-water power reactor. First-of-a-Kind projects carry inherent risks, compounding the ordinary challenges of nuclear construction.

One primary concern is workforce inexperience. Nuclear energy technologies require highly specialized workers, and a first-of-a-kind project means nobody has previously built something just like it. As a result, workers and supervisors lack the experiential knowledge of the construction, impacting both pace of construction and ability to identify flaws before they become embedded defects. The lack of direct experience for certain project components may also increase the amount of design assistance and information requests that also impact project pace and increase the need for third-party quality assurance personnel. Consequently, both cost and schedule are more likely to change over the course of a first-of-a-kind project than on other projects.

Supply chain integrity is another pressing issue for the Kemmerer Plant. There is no established supply base for Natrium-specific components. Procurement of first-run parts from new or unproven vendors increases the likelihood of manufacturing defects and quality deviations. Troublingly, according to the NRC Inspector General, counterfeit parts have been found in numerous U.S. nuclear power plants. For a first-of-a-kind project, the lack of an established and pre-qualified supplier network heightens cost, schedule, and quality risks.

Unique Challenges of Sodium-Cooled Reactor Construction

Beyond the problems identified with traditional nuclear construction, The Kemmerer Plant’s Natrium design introduces unique construction and operational hazards associated with liquid sodium as a coolant. Sodium is highly chemically reactive as it ignites on contact with air and explodes violently when it contacts water. These dangerous properties demand extraordinary precision in the creation and installation of coolant system components, and the historical record of sodium-cooled reactors is troubled.

Japan’s Monju Fast Breeder reactor illustrates some of these risks. On December 8, 1995, intense vibration caused a thermowell inside a coolant pipe to break at a potentially defective weld point, allowing several hundred kilograms of liquid sodium to leak. The sodium ignited on contact with air, generating temperatures capable of warping structural steel. The explosion resulted in the reactor being shut down until 2010, and it was permanently decommissioned in 2016 operating for less than one year of its 30-year lifespan.

Monju was not an anomaly. Russia’s BN-600 reactor reported 27 sodium leaks over a 17-year period, 14 of which resulted in sodium fires. France’s Phénix prototype suffered five sodium-water reaction incidents in its steam generators and the UK’s Prototype Fast Reactor experienced 37 sodium leaks. All three of these examples demonstrate how even minor defects in fabrication like an imprecise weld or a vibration-sensitive sensor installation, can result in catastrophic consequences when liquid sodium is the coolant. In a water-cooled reactor, a small leak is typically a maintenance issue; in a sodium-cooled system it can become a fire.

TerraPower is aware of these challenges and has taken steps to address them. The company is building a dedicated Sodium Test and Fill Facility (TFF) at the Kemmerer site that will test the full sodium loop before being introduced to the reactor core. This additional precaution adds complexity and schedule impacts not seen in conventional nuclear builds.

Legal Consequences of Construction Risks in Nuclear Projects

Construction risks in a nuclear facility create a distinctive and particularly complex legal landscape. Multiple parties bear potential exposure, including the general contractor, subcontractors, material and component suppliers, design engineers, and the project owner. The legal framework governing these disputes is multilayered, spanning federal nuclear regulation, state construction law, and contract.

At the federal level, the Price-Anderson Nuclear Industries Indemnity Act channels public liability for nuclear incidents to the licensed operator. Third-party claims arising from a nuclear accident are brought against TerraPower, not against contractors or subcontractors, regardless of which party’s conduct caused the defect. However, this does not immunize contractors from downstream liability. TerraPower could pursue breach of contract, warranty, and indemnification claims against any construction party whose defective work causes loss, delay, or remediation.

As with any large construction project, potential legal issues do not stop with third parties. If the Kemmerer Plant has similar issues to recent projects, the project could have substantial costs related to correction of defects during the project, project delays, and other cost overruns. The question then will be who bears those costs. Answering those questions requires determining who was responsible for the need to correct, who was responsible for the delays, and who was responsible for any over cost overruns. Clear project documentation will be key to resolving any disputes on these issues. Even with clear documentation, complex litigation may still result.

NRC regulatory violations can also present a financial issue. If the NRC issues a Confirmatory Action Letter or a Notice of Violation, like it did repeatedly during the Plant Vogtle build, TerraPower could face civil penalties, mandatory remediation orders, and construction shutdowns. Again, such actions could lead to downstream disputes between TerraPower and the design and construction team.

Conclusion

TerraPower’s Natrium reactor is a genuine technological milestone, and its construction permit represents the first major advance in U.S. commercial nuclear licensing in a generation. The streamlined NRC review, completed 18 months ahead of schedule, reflects real progress in regulatory efficiency. However, obtaining a construction permit is simply the beginning for TerraPower.

The history of nuclear construction, from Vogtle to Monju, demonstrates that first-of-a-kind projects in complex, safety-critical industries are particularly susceptible to construction issues. The resulting legal consequences can lead to complex construction litigation.