What a Thorium Molten Salt Reactor Can and Cannot Deliver

Most debates about the energy transition treat it as a two-way contest: wind and solar on one side, conventional uranium reactors on the other. One study argues that the uranium side may hold about 29 years of assured reserves, while thorium burned in a molten salt reactor could last about 1,300 years. That gap is striking, and it depends heavily on assumptions you should test before accepting it.

The study comes from Nisho, whose bottom-up analysis argues that a full wind-and-solar build-out runs into seasonal storage and mineral limits. It also argues that solid-fuel uranium expands too slowly to fill the gap. This is one study’s scenario result, not a consensus view.

Interest in molten-salt designs is rising regardless. As of October 2026, China operates an experimental thorium reactor, the US Nuclear Regulatory Commission (NRC) is advancing Kairos Power permits, and Copenhagen Atomics is running full-size non-nuclear prototypes.

Here you will find what a thorium molten salt reactor is, what the evidence supports and what it does not, and what all of it means for your copper and nuclear exposure.

Why do wind, solar and conventional uranium both struggle to scale?

The case against a full wind-and-solar build-out

Nisho’s argument starts with a calendar problem. Wind and solar output can dip for long stretches, so the study argues storage must cover months, not the hours that batteries handle well.

The second problem is materials. Nisho’s work, dating from 2018 onward, assumes power generation must a little more than double, and it finds combined known minerals amount to about one fifth of what first-generation technology would need.

Conventional nuclear fares little better in the model.

Nisho’s uranium simulation Building 25 new reactors a year from 2025 would take 75 years to use known uranium and reach only about 60% self-sufficiency, according to Nisho’s simulation.

What critics of that conclusion say

No independently verifiable source for the study, and no named peer-reviewed critique, could be located. Its conclusions belong to Nisho’s assumptions, and those assumptions do a lot of work.

Case for predominantly renewable systems

• Transmission, hydro, demand response and overbuilding can reduce how much storage you actually need
• Recycling, aluminium substitution and lower material intensity can cut copper demand per unit of energy
• The answer changes depending on which technologies a model allows

Case for a firm nuclear or dispatchable component

• Several studies find low wind and solar output can persist for days or weeks across large regions
• Batteries suit hours well but look much less suited to storing months of national demand
• Mining, refining and grid construction are slow and capital-heavy

Mineral intensity is not static either. A thorium reactor still needs steel, nickel alloys and graphite, so it does not escape the materials question.

The fairest reading is that the study stress-tests a restricted-technology scenario. It does not prove renewable-heavy grids fail, so treat its conclusions as a hypothesis to weigh rather than a verdict.

Long-duration chemistries such as iron-air batteries aim to stretch storage from hours toward days, though neither they nor lithium-ion yet address the months-long seasonal gaps Nisho’s model highlights.

How does a thorium molten salt reactor work, and why is liquid fuel the key difference?

On a schematic, the design looks unfamiliar. Instead of solid fuel rods sitting in pressurised water, the fuel is dissolved in hot liquid salt that circulates through the reactor at roughly 600-700°C and at atmospheric pressure.

Thorium, uranium-233 and the fuel cycle

Thorium is fertile, not fissile. A fissile material can sustain a chain reaction on its own; a fertile one cannot, but it can be converted into fissile fuel when it absorbs neutrons.

That means a thorium reactor needs an initial fissile driver, such as uranium-235, plutonium or existing uranium-233 (U-233), to start. Once running, thorium absorbs neutrons and breeds U-233, which then fissions and releases energy.

Because thorium is fertile rather than fissile, the uranium to thorium transition is less a fuel swap than a redesign of the whole reactor, which is why you need a fissile driver such as uranium-235 or plutonium to start the cycle.

Liquid fuel is what makes this cycle practical. Because the fuel circulates, operators can in principle recover bred material and remove contaminants without fabricating new solid fuel rods.

The concept is not new. Oak Ridge National Laboratory’s Molten Salt Reactor Experiment (MSRE) ran from 1965 to 1969 at up to 7.4 MWth (megawatts of thermal output) for more than 13,000 full-power hours.

Why the reserve comparison needs care

The headline figures look decisive. On Nisho’s numbers, using uranium and spent fuel in liquid-fuel reactors would stretch the supply to about 6,200 years.

Read the reserve figures carefully These durations depend on assumed reactor technology, extraction rates, demand, and whether economically recoverable reserves or broader resources are being counted. They are not a like-for-like current reserve comparison.

The latest uranium data complicates the “29 years” framing. The OECD-NEA/IAEA Red Book 2024 puts identified resources recoverable below US$260/kgU at more than 8.1 million tonnes, up 2.1% on the previous edition. The Red Book offers no comparable global thorium total; the IAEA keeps a deposit database instead.

So “1,300 years” is an argument about long-run fuel security under one specific reactor design. It is not a measurement you can set beside today’s uranium inventory.

Where does the technology stand in 2026, and who is actually building it?

If the fuel argument holds only for a specific design, the obvious question is whether anyone has built it. The answer depends on how much proof you require.

Operating evidence and permitted demonstrations

China’s TMSR-LF1 at Wuwei, built by the Shanghai Institute of Applied Physics, is the strongest modern evidence. The 2 MWth experimental reactor reached criticality in October 2023, full power in June 2024, added thorium in October 2024 and reported its first experimental thorium-to-U-233 conversion in November 2025. China is reportedly targeting a 100 MW pilot plant by 2035 and commercial operation around 2040.

Kairos Power is next on the ladder, though its Hermes reactors are fluoride-salt-cooled and not thorium-fuelled. The NRC directed staff this month to issue Hermes 2 construction permits, with operations expected around 2030.

Terrestrial Energy submitted an NRC topical report on 8 October 2026. That is a licensing-support step only, not a construction approval. Flibe Energy and ThorCon have no current dated milestones on record.

Copenhagen Atomics: the factory-built bet

The Danish company wants to build reactors the way you might build shipping containers. Its commercial design is rated at 100 MWth, and in February 2026 it reported two years of continuous operation of a molten-salt pump and test loop.

It has built two full-size non-fission prototypes and is constructing a third. Its BALDER experiment at Switzerland’s Paul Scherrer Institute runs at about one-hundredth of commercial thermal output, with initial operation limited to 30 days.

Nisho’s figures differ from the company’s current targets. He cited a 40 MW container-sized unit and a factory making 10-20 units a year; the company’s longer-term gigafactory claim runs to 15-50 GW a year.

The company expects to file its licence application by the end of 2026, with tests no earlier than 2028 and commercial units in the early 2030s. Nisho’s roughly 2035 first-unit estimate looks more conservative.

Most thorium headlines describe testing and permitting, not power on the grid. Read company timelines as targets, not schedules.

What could still stop thorium molten salt reactors from scaling?

Why the Oak Ridge programme ended

Nisho says molten-salt reactors were dropped for economic reasons. The record suggests something different: the MSRE worked, but the wider US programme lost funding and ended around 1973 as priorities shifted toward fast-breeder designs and military goals.

That matters because it means the idea was not disproved. It was also never proven commercially.

The risk register

The Bulletin of the Atomic Scientists offers the most sceptical recent view, arguing that many historical technical difficulties remain relevant. Its critique bites hardest against promises of near-term, low-cost mass deployment, not against continued demonstration.

What has changed:

• Computer modelling of reactor physics and heat flow
• Nickel alloys, corrosion monitoring, pumps and sensors
• Modular factory manufacturing and policy support
• Private programmes such as Kairos, Terrestrial Energy and Copenhagen Atomics

What has not changed:

• Long-duration materials proof
• Radioactive salt processing
• Commercial economics
• Licensing at scale

The barriers are mostly about proof at commercial scale rather than basic physics. Weigh any announcement by how many items on this list it actually addresses.

For readers wanting to see how designers test salt behaviour without fuel, our deep-dive into fuel-free molten salt loop research explains what Berkeley’s FLiNaK loop can and cannot validate.

What does this mean for copper, mining and nuclear investors?

How tight is copper really?

Nisho framed minerals as a hard ceiling. More recent forecasts point to real tightness, but a narrower gap than a flat “30% shortfall” suggests.

The IEA’s Global Critical Minerals Outlook 2026, released on 11 September 2026, expects copper deficits to persist through 2035 while the gap narrows from about 30% to 25% as projects advance.

These are forecasts, not established shortages. Past performance does not guarantee future results, and projections are subject to market conditions and various risk factors.

For reference, Discovery Alert reported on 13 October 2026 a uranium spot price of US$89.63/lb U₃O₈ and a term price of US$96.50/lb at September month-end, with copper around US$14,540/t in a market snapshot without an exchange timestamp.

Substitution and what to ask of thorium investments

• Aluminium works in many power cables, overhead conductors and transformers
• Copper keeps its edge in connectors, compact equipment, motors and some underground uses
• Graphene remains uncertain as a substitute

Copper demand rests on many routes beyond any single reactor type, and thorium reactors still need steel, nickel alloys, graphite and pumps. Before backing any thorium-linked story, ask:

1. Has it secured, or even filed for, a licence?
2. What long-duration materials data does it have?
3. How will it process radioactive fuel salt?
4. What evidence supports its cost claims?

Thorium is best viewed as a long-dated option on firm power, not a near-term driver of copper or uranium prices.

Mineral intensity is not static, and the critical mineral dependencies of advanced reactors, from nickel alloys to graphite, show that next-generation nuclear still competes for constrained supply chains.

Weighing a long-dated option: what the thorium case does and does not settle

Thorium molten salt reactors look like a credible long-term research pathway for firm power and industrial heat. They are not yet a validated replacement for renewables or conventional nuclear.

The strongest evidence sits in China’s operating experiment, Kairos’ advancing permits and Copenhagen Atomics’ non-nuclear testing. The largest gaps remain licensing, fuel-salt processing, corrosion and component life, first-of-a-kind cost and factory scale.

If you hold mineral or nuclear exposure, the milestones worth tracking run through 2028 and the mid-2030s: licence application outcomes, BALDER results and China’s pilot plant. Each one either narrows the gap between ambition and proof, or widens it.

These statements are speculative and subject to change based on market developments and company performance.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions.