When Thorizon, a nuclear startup headquartered in The Hague, announced that it had chosen Hyundai Engineering as the construction partner for its molten-salt reactor, the pairing raised eyebrows on both sides of the Atlantic. A small Dutch-French venture with roots in European research labs was handing the building job to a South Korean industrial heavyweight better known for petrochemical plants and offshore platforms. The reactor they intend to build would operate at temperatures above 600 °C yet at roughly atmospheric pressure, a combination that eliminates the high-pressure steam systems at the heart of every conventional nuclear plant operating today. If the design works at scale, it could remove the physical conditions that made accidents like Fukushima possible. That is a significant “if.”
The Oak Ridge inheritance
Thorizon’s concept descends directly from the Molten Salt Reactor Experiment (MSRE), which ran at Oak Ridge National Laboratory in Tennessee from 1965 to 1969. During that campaign, engineers circulated hot fluoride salts through a small reactor core and discovered that the chemistry was brutal on hardware. Standard industrial valves corroded or seized. The workaround was a freeze valve: a section of pipe where a plug of solidified salt was kept frozen by a small cooling system. If the reactor overheated or lost power, the plug melted, and gravity drained the fuel salt into a holding tank shaped to prevent any self-sustaining chain reaction. An Oak Ridge assessment of commercial valve hardware confirms that freeze valves were adopted precisely because nothing off the shelf survived the conditions.
The physics that make this possible are not exotic. Molten fluoride and chloride salts boil at temperatures far above their normal operating range, so the reactor vessel never needs to contain pressurized steam or superheated water. A peer-reviewed study in the Annals of Nuclear Energy examining the behavior of a fast-spectrum freeze valve explains that this low-pressure characteristic removes the driving force behind the steam explosions and hydrogen buildup that contributed to the Chernobyl and Fukushima disasters. No high-pressure containment means no catastrophic pressure release.
The freeze plug also doubles as a passive safety system. Lose all electric power and the heater maintaining the plug shuts off. The plug melts. Fuel drains by gravity alone. No operator action, no backup diesel generators, no emergency cooling pumps. The Annals of Nuclear Energy study uses computational models, validated against the original MSRE data, to simulate how quickly the plug responds and how reliably it opens under accident conditions.
Why Hyundai Engineering
Hyundai Engineering brings something Thorizon lacks: experience building large, complex chemical and energy facilities on schedule and at scale. The company’s portfolio spans LNG terminals, refinery units, and modular process plants, all of which share characteristics with a molten-salt reactor. Salt loops, high-temperature heat exchangers, and corrosion-resistant piping are routine in petrochemical work. Thorizon’s reactor, targeting roughly 250 megawatts of thermal output (enough to generate approximately 100 megawatts of electricity), is closer in engineering character to a chemical processing facility than to a traditional nuclear station with its massive concrete containment dome and steam generators.
As of early 2025, Thorizon reported that pre-FEED (front-end engineering and design) work was underway, with Hyundai Engineering contributing to plant layout and constructability studies. No signed full-scope engineering, procurement, and construction (EPC) contract has appeared in public filings, and the precise division of responsibilities between Hyundai and Thorizon’s own engineering staff has not been formally disclosed. Readers should treat the scope of Hyundai’s role as evolving rather than finalized.
Materials: the unsolved middle
The MSRE identified Hastelloy-N, a nickel-molybdenum alloy, as the best available structural material for containing hot fluoride salts. It performed well during the experiment’s four-year run, but four years is not forty. Commercial reactors need materials that hold up for decades under neutron bombardment and corrosive salt flow.
Modern experiments at Oak Ridge’s neutron-scattering beamlines are probing exactly that question, measuring crack growth, swelling, and chemical attack in candidate alloys exposed to salt environments over extended periods. European labs, including facilities at Petten in the Netherlands, are running parallel programs. The data from these campaigns will be essential for any licensing application, because regulators will demand proof that reactor vessels and piping can survive their intended service life without brittle fracture or through-wall corrosion.
This is not a theoretical concern. During the MSRE, engineers observed tellurium-induced grain-boundary cracking in Hastelloy-N, a degradation mode that had not been predicted. Solving that problem for a commercial reactor operating for 40 or 60 years, rather than four, is one of the hardest open engineering challenges in the molten-salt field.
The regulatory gap
No European regulator has ever licensed a molten-salt reactor. France’s Autorité de Sûreté Nucléaire (ASN) and the Netherlands’ Authority for Nuclear Safety and Radiation Protection (ANVS) both operate licensing frameworks built around pressurized-water reactors. Reviewing a design that uses liquid fuel, operates at low pressure, and relies on a freeze-plug drain instead of conventional emergency core cooling will require either adapting existing rules or writing new ones.
Key questions regulators will need to address include how to handle molten fuel during maintenance and refueling, whether online fission-product removal (a feature some MSR designs propose) introduces proliferation or waste-handling concerns, and what off-site emergency planning looks like for a reactor that cannot suffer a high-pressure steam release. Thorizon has indicated it is in early-stage dialogue with Dutch regulators, but no formal licensing application has been submitted as of June 2025.
For comparison, Kairos Power’s Hermes test reactor in Oak Ridge, Tennessee, which uses a fluoride-salt coolant (though with solid fuel, not dissolved fuel), received a construction permit from the U.S. Nuclear Regulatory Commission in late 2023, the first for a non-water-cooled reactor in the United States in decades. That precedent suggests regulators can adapt, but the process took years of pre-application engagement and hundreds of pages of safety analysis.
What the evidence supports and what it does not
Two claims sit at the core of this project, and they deserve separate treatment.
The first is that molten-salt reactors can operate at low pressure with passive drain safety. This is well supported. The MSRE demonstrated it experimentally. Peer-reviewed modeling confirms the underlying physics. Multiple independent research groups have reproduced the key findings. Anyone dismissing the concept as unproven is ignoring six decades of data.
The second claim is that Thorizon and Hyundai Engineering can deliver a licensed, grid-connected plant on a predictable schedule and at a competitive cost. This is not supported by any public evidence. No detailed engineering design has been published. No construction cost estimate has been independently reviewed. No licensing application is on file. The gap between a validated computer simulation and a functioning power station is wide, and it is filled with fabrication challenges, supply-chain constraints, and regulatory reviews that no laboratory paper can shortcut.
Conflating these two claims is the fastest way to either overhype the technology or dismiss it unfairly.
Three signals worth watching
For anyone tracking advanced nuclear energy in Europe, the milestones that will separate ambition from progress are concrete and measurable. First, publication of a reference plant design with specific power output, fuel composition, and cost targets. Second, a formal licensing application submitted to ANVS or ASN, which would trigger a public technical review. Third, independent material test results from ongoing neutron-scattering and corrosion studies confirming that structural alloys can survive multi-decade exposure to hot salt under irradiation.
Until those milestones land, Thorizon’s reactor sits in a familiar but frustrating category: grounded in solid physics, backed by credible research, and still years away from proving it can do what its designers believe it can. The Hyundai partnership adds industrial credibility and construction know-how. Whether that is enough to push a 1960s laboratory success into a 2030s power plant is the question that matters now.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
