The Reactor Meant to Showcase Nuclear's Comeback Keeps Hitting Problems

In March 2026, the U.S. Nuclear Regulatory Commission (NRC) issued the first commercial nuclear reactor construction approval in nearly a decade. The destination: Kemmerer, Wyoming. The company: TerraPower, the world's most famous nuclear startup, co-founded and funded by Bill Gates. The vision was simple, almost poetic — transform an old coal power plant site into an advanced zero-emission nuclear facility, and use it as proof-of-concept for a new energy era.

Six months later, reality tells a different story. TerraPower is losing its primary construction partner (Bechtel), still depends on a fuel that does not exist at commercial scale, and sees its promise of a next-generation reactor bump into the same problems that have plagued the nuclear industry for decades: cost overruns, delays, and execution complexity.

The Bechtel split

In September 2026, the engineering and construction firm Bechtel announced it would have no further involvement in TerraPower's Natrium project. The split was revealed through a Worker Adjustment and Retraining Notification (WARN) filed in Virginia, which listed up to 200 positions at risk due to the decision to "move forward separately."

Bechtel had worked with TerraPower since the project's early stages — including securing the NRC construction permit, completing the exterior enclosure of the sodium test and fill facility, and beginning foundational work for Kemmerer Unit 1. But the company could not reach agreement on the next phase, and is now scaling back its involvement.

TerraPower tried to downplay the impact. "Construction in Kemmerer will maintain its current schedule with exemplary subcontractor teams," the company said in a statement. But the truth is that the full EPC (Engineering, Procurement and Construction) contract will be rebid — meaning new delays, new costs, and more uncertainty about a project that has already taken years to reach this point.

Notably, in August 2026, TerraPower announced that Hyundai Engineering & Construction would become its EPC partner for up to eight future Natrium reactors — but that applies only to future projects, not the Wyoming demonstration reactor, which remains without an established construction contractor.

The HALEU fuel problem

Even if TerraPower resolves its construction problems, another nearly insurmountable obstacle remains: fuel. The Natrium reactor requires high-assay low-enriched uranium (HALEU), enriched to between 5% and 20% in uranium-235 — a material that, to date, does not exist in any commercial production scale anywhere in the world.

Russia is currently the only commercial source of HALEU, and the invasion of Ukraine has dramatically complicated any plans to import it. TerraPower originally planned to use Russian fuel to bring the reactor online by 2028. That target has already been pushed to 2030-2031, and even that depends on a supply chain that does not yet exist.

The U.S. Department of Energy has recognized the problem and begun establishing a HALEU fuel bank, but it is still in early stages. Without fuel, there is no reactor — and without a reactor, there is no proof of concept.

The irony at the center

Here is the central paradox of advanced nuclear energy: TerraPower was chosen to be the face of the "nuclear renaissance" precisely because it has money, talent, and the Gates brand behind it. But the problem of nuclear energy is not a lack of technological innovation — it is construction.

The Vogtle plant in Georgia is the most recent example. Built by Westinghouse, it ran seven years behind schedule and cost twice as much as budgeted. When it finally came online in 2023, it was the first nuclear plant built in the United States in more than three decades.

Companies like Blue Energy, a MIT-born startup, are trying to solve exactly that problem — prefabrication at shipyards, modular assembly, financing like wind and solar projects. CEO Jake Jurewicz is blunt: "The nuclear problem is a construction problem." But even Blue Energy, with its radically different approach, projects three years to build a facility that will initially operate on natural gas.

TerraPower's liquid-sodium reactor and molten-salt energy storage is cutting-edge technology. But technology cannot overcome decades of delays or a missing fuel supply chain. And Wyoming is not going to wait forever.

The question that remains: if the most prominent name in advanced nuclear cannot build its first reactor, who can?

Sources: Nuclear News (ANS), Nuclear News (ANS), Industrial Equipment News

✓ Independent sources cross-checked and verified before publishing