Before anybody gets too excited, it's better to understand what exactly happened.
China ran an experimental reactor that achieved some conversion of thorium into uranium. More precisely, the conversion ratio was 0.1 [1]. This means that for each new fissile atom generated from thorium (i.e. uranium-233) 10 atoms have been burned from the original fissile inventory.
Now, conversion happens in every nuclear reactor. Some new fissile material (generally Pu-239) is generated out of "fertile material" (generally U-238). And, surprisingly, that conversion ratio is quite high: 0.6 for pressurized light water reactors and 0.8 for pressurized heavy water reactors [2].
What China has achieved therefore is well below what is business as usual in regular reactors. The only novelty is that the breeding used thorium, rather than uranium.
Is this useless? No, it is not. In principle increasing the conversion ratio from 0.1 to something higher than 1.0 should be doable. But then, going from 0.8 in heavy water reactors to more than 1.0 should be even easier. Why don't people do it already? Because the investment needed to do all the research is quite significant, and the profits that can be derived from that are quite uncertain and overall the risk adjusted return on investment is not justified. If you are a state, you can ignore that. If China continues the research in thorium breeding, and eventually an economically profitable thorium breeder reactor comes out of that, the entire world will benefit. But the best case scenario is that this would be three decades in the future.
My understanding is that reactors will use that plutonium just fine, so the energy you get from a fresh fuel rod is almost exclusively from uranium fission but, as time goes on, an increasingly large share is from plutonium fission.
In principle, using Thorium would give you the energy from Thorium fission, then Uranium fission, then plutonium fission, which is pretty cool. However, I suspect you might hit an issue here where such a low conversion rate would make the reactor go sub-critical.
Fundamentally the problem is that Uranium is so damn energy dense and abundant enough that there's little need to set up these complicated recycling systems. If we start to run out of Uranium then this technology starts to look appealing, but in the modern day it just doesn't make economic sense.
> Uranium is so damn energy dense and abundant enough that there's little need to set up these complicated recycling systems
Uranium is abundant, but not homogenously so [1]. (China has some. But not a lot. And it's bound up expensively. And it's by their population centres.)
For the Americas, Europe, Australia, southern Africa and Eastern Mediterranean, burning uranium makes sense. For China, it trades the Strait of Malacca for dependence on Russia and Central Asia.
Uranium can be stockpiled for years in advance, relatively easily. Enough to tide over a small war while you're setting up domestic production. And China should have enough low-grade ores for that.
There's not that much Uranium actually that's economically sensible to extract. The NEA says in their 2024 report on Uranium [1]:
> Considering both the low and high nuclear capacity scenarios to 2050 presented in this edition, and assuming their 2050 capacity is maintained for the rest of the century, the quantities of uranium required by the global fleet – based on the current once-through fuel cycle – would likely surpass the currently identified uranium resource base in the highest cost category before the 2110s.
Their "high" scenario assumes having a bit more than double of today's capacity by 2050; today we have about 4-5% supply from nuclear energy worldwide.
We really do. Nuclear waste is less toxic than plenty of trash we just bury. And calling it "waste" is a bit reductive, given it almost certainly becomes valuable to reprocess within another century or two.
The notable thing here is that it's a molten salt reactor design, where the fuel is dissolved in a molten salt (FLiBe). This allows online continuous processing of the fuel, unlike with solid fuel rods sealed inside a pressure vessel.
This unlocks a lot of options for the fuel cycle, including the use of thorium.
This work builds on a previous molten salt reactor experiment at Oak Ridge, decades ago. There's a whole lore about MSRs.
It breeds thorium to fissionable uranium from a starting fissionable uranium starter fuel. It doesn't directly use thorium for fuel.
What people need to understand about the cycle efficiency is that when you mine uranium, the fissionable part of uranium (U-235) is only 1% of that uranium, the rest is nonfissionable U-238.
Thorium is about twice as abundant as Uranium (all isotopes). The MSR uses Thorium to create U-233, a fissionable but not naturally occurring Uranium isotope.
So the "unlimited energy aspect" is that about 200-300x more breedable Thorium exists than fissionable U-235.
A MSR nation could also try to breed U-238 into plutonium, which would provide another 100x more breeding stock, although LFTR never talked about U-238 breeding. IIRC the plutonium may be difficult to handle because of gamma rays, but I don't recall exactly.
While I don't have confidence that even LFTR/MSR reactors can get economical enough to challenge gas peakers, it may be possible to make truly price-competitive MSR electricity with the right modular design. I wish the Chinese the best of luck, because if they do it will spur the rest of the world to adopt this about-as-clean-and-safe-as-it-gets nuclear design.
China has thorium, and while less than others [1], it’s better than they do with uranium [2].
> it may be possible to make truly price-competitive MSR electricity with the right modular design
Yes. But probably not in the near term with thorium. This isn’t designed to be cheaper. It’s designed to be more available to China than being dependent on Russian deposits.
MSRs have some attractive features, but they also have significant drawbacks.
The most pressing is that fissionable material is spread throughout the fluid, so fission and decay of fission products is occurring right up to the edge of the fluid. The walls and pipes containing the molten salt, and anything dipped into the salt, are exposed to unmoderated neutrons. One can shield using (say) graphite, but then damage to that (and soaking up of radioactive materials) become issues.
The Molten Salt Reactor Experiment at Oak Ridge was near the end of its radiation exposure lifetime when the program ended.
Contrast this to light water reactors. These are designed so that no lifetime component sees unmoderated neutrons. There's a thick barrier of water between the fissioning fuel and the reactor vessel wall and the support structures for the fuel bundles. The bundles themselves are exposed, but they are replaced for refueling and are not lifetime components.
To add to this, even with the shielding provided by water in light water reactors, the neutron exposure is _the_ limiting factor for the reactor vessel.
The metric to look for is called "DPA" (displacements per atom), the number of neutron collisions that a material can tolerate before losing enough structural integrity to fall below the acceptable limits. The best modern reactor steels are at 150-180 DPA.
And a lot of potentially cool reactors like TWR (travelling wave reactor) end up being logistically impossible because lifetime-limited components will be exposed to multiple hundreds of DPAs.
Thorium 232 is the thorium in the cycle yes. And all kinds of nonsense is correct for the daughter products. But in general, to actually use do anything with thorium you need excess neutrons.
Even the daughter uranium 233 only produces on average 2.48 neutrons per fission, so it’s very difficult even in a combined lifecycle process to have enough - thorium doesn’t produce uranium 233 immediately (takes almost 30 days), neutron capture with that low a ratio requires a LOT of thorium, which is going to mostly just suck up all neutrons and you won’t have any extra for addition uranium 233 fissions, etc.
It’s quite difficult (impossible?)to have actually work without a source of a large amount of additional neutrons.
> to actually use do anything with thorium you need excess neutrons
Unless 100% of those neutrons is being absorbed by the thorium, this means you'll have neutron flux at the boundary. Which, for a liquid moderator, means all the pipes and tanks and pumps.
> Now, the research team is conducting systematic studies on the key scientific issues related to adding thorium, and aims to completethe construction of a 100-megawatt TMSR demonstration project, and begin operation by 2035.
For comparison: A commercial nuclear power plant is 1 gigawatt, a 10x difference. I assume this would be the next step.
The typical 1 gigawatt rating for a nuclear power reactor is measuring electrical output. Given the various inefficiencies, the actual reactor output (as heat) is something like 3x that amount. Whereas a research reactor will be quoted as thermal output.
That to say, a typical commercial reactor might be 30x the power of a 100 MW research device.
Thorium MSRs don't make sense for the Americas, Europe or Australia. We have plenty of uranium.
Nuclear is receiving solid research backing in both America and China. (India is playing too. Austrlia is having an identity crisis.) Our different geologies mean there will probably be one solution for China, India and North Africa, on one hand, and the rest of the world, on the other hand.
Breeding is a technology looking for a business case.
It's more expensive than just using fresh uranium in current market conditions. It's a way from keeping future uranium shortages from making nuclear power more expensive; it's not a way to make nuclear cheaper than it currently is.
The truth is that nuclear power is not that financially attractive at the present and would the price of uranium rise enough that breeders would become economically viable most countries would just stop bothering with nuclear power altogether.
The cost of nuclear power is almost entirely capex and financing, not opex. Uranium input cost for nuclear power plants is 0.5c/kWh. With breeders you can divide that by about 100.
At least as of a couple years ago nuclear costs just a little more than solar plus storage and that’s not stopping anyone heh.
Capex and financing is still an issue for many countries, and the opex is a non-zero commitment beyond just the fiscal portion. Most countries that pass-over nuclear energy are fairly justified in their decision. The status-quo is still not super psyched about nuclear proliferation.
There is room to change that, but the cards are very heavily stacked in China's favor. America's bad at the financing part, fickle when it comes to enforcement & supply chains, and ostensibly 2 days away from bailing on the IAEA itself. The proliferation-resistance of Thorium reactors gives China an export trump card that America will struggle to match.
> The truth is that nuclear power is not that financially attractive
Let me fix that for you: "The truth is that nuclear power is not that financially attractive in the bureaucratic high cost litigious Anglo-sphere". And that's pretty much all infrastructure these days, unfortunately.
They’re not financially attractive in other parts of the world either. China, a zero litigation single party state, is building some but a tiny % compared to their renewable buildout
"China currently has 58 operable reactors with a total capacity of 56.9 GW. A further 30 reactors, with a total capacity of 34.4 GW are under construction" [1]
So, yes, but...
China installed 256GW of solar in the first 6 months of 2025 [2]. A full year estimate of ~350gw. So, the total of all nuclear under construction is 1/10th of the solar they installed in one year.
Don't get me wrong, its cool to see diversity of non fossil sources, glad they are building some, but its a niche in their overall energy buildout. And they can only build that small niche because they dont have to be market priced, its state subsidized.
It's not the litigiousness that makes it expensive. France was producing nuclear power plants at a cost per watt that nearly matches modern China. In fact, the mind-numbing cost overruns seem unique to the US.
france cant do it any more either. Flamanville was 12 years late and [1] 400% over budget. EPR2 is already delayed and over budget and they havent even started building yet!
They highlight less the advantages from breeding, than other advantages of the molten salt design, like not needing a lot of cooling water, which allows this reactor to operate in the Gobi desert, the possibility of replacing the fuel without halting the reactor and various safety features.
Nuclear reactors don't need a particularly big amount of cooling water.
The thermodynamic cycle needs a cold source though, and it's most commonly water. This doesn't depend on the reactor design and this is equally as true of coal plants.
As long as you are making electricity out of a thermodynamic cycle, you need a heat source (be it a flame or a nuclear reaction) and a cold source.
It also apparently provides a way to make reactors that don’t depend as much on water so they don’t all have to be near the coast.
This would allow Western China to also develop reactors to help underpin their renewable and coal energy.
> The interest in MSR technology and Thorium breeding did not disappear however. China's nuclear power production relies heavily on imported uranium,[10] a strategic vulnerability in the event of i.e. economic sanctions. Additionally, the relative lack of water available for cooling PWRs west of the Hu line is a limiting factor for siting them there.
> also apparently provides a way to make reactors that don’t depend as much on water so they don’t all have to be near the coast
Non-water microreactors broadly fall into two categories: ones using a different moderator, most commonly sodium, a sodium salt or helium; and those using heat pipes. Most microreactor designs don’t use water.
There is no business case for basic research, but if you stop basic research long enough you will have no business. The United States and its allies seem to have completely forgotten this.
It makes sense for big monopolies like Bell, or the CCP. The investment can be justified if the ones investing are confident they will be able to capture the value and not some competitor.
Bell Labs also served to maintain positive perceptions of the monopoly. Unix was famously developed despite the knowledge that AT&T would not be able to offer it as an independent product.
Reducing the energy sector to pure business would probably work in the 1990s, but not now, when countries are afraid of strategic dependence on potentially hostile suppliers.
Uranium isn't as ubiquitous as, say, natural gas, and stockpiling it comes with a big heap of physical problems. I can definitely see countries spending on more expensive technology if it comes with more energy security.
I mean we're already doing that in many avenues. Solar being the most obvious. The only functioning solar manufacturing plants in the US are Chinese-owned and are only here to take advantage of subsidies.
Plenty has been learned by the US/West from copying their approach to agriculture, robotics in factories, mining, drones, etc. Have you seen their electromagnetic catapult technology?? That stuff seems like its from the space-age! There's even plenty of tech that we can't really explain like the all-moving wingtips on the new J-50s. (and yes, I'm avoiding talking about their supersonic cruise missiles)
No country has seriously invested in the thorium fuel cycle because it cannot be used to create weapons. Unfortunately, the technology also began to look most promising as an energy source around the same time the Three Mile Island nuclear accident effectively ended all interest in nuclear energy in the United States.
India has shown some of the most interest to date, due to their lack of domestic uranium reserves. But it's been slow going their fast breeder reactor plans were delayed by like two decades. But it is built and it was loaded with fuel last month [0]
The French interest in breeder reactors and nuclear reprocessing also originates from a similar concern about lack of domestic access to raw uranium. Though Super-phoenix [0] was a more traditional uranium -> plutonium approach and not thorium. They gave up because just using uranium is way, way cheaper than synthesizing your own fissile materials.
Theoretically, perhaps, but I don’t think anyone with a serious interest in weapons would pursue it. From a nonproliferation perspective, I’d guess the infrastructure necessary to remove contaminants from uranium bred through the thorium cycle would be costly and difficult to conceal.
Multiple countries have detonated nuclear bombs using U-233 derived from thorium reactors! [0] Practically I agree with you that thorium is proliferation resistant and if someone is bomb hungry they won't prioritize it, but if you want to set up the bomb and all you have is thorium... The infrastructure wouldn't necessarily be significantly larger or worse than conventional enrichment.
To be fair, these advances are not being made in China due to "free industry". They have something of a command economy for their critical sectors. So it's unfair not to point out that it's easy to make advances if a nation as a whole points to a hill and says, "take that hill". Of course you can do it under those circumstances.
If it's just your company or some trifling consortium trying to develop nuclear energy advances in a "free industry" environment, the guy who is just slapping up windmills, [T Boone Pickens RIP], is just gonna mop the floor with you. There's just no way to compete on moonshots like that.
Meanwhile Germany just decommissioned its last nuclear reactors. Given the challenges of baseload renewable generation, it's frustrating to watch working infrastructure being dismantled while we're still heavily dependent on fossil fuels.
> Comparing those old conventional reactors to MSR is not suitable at all
It is given we're talking about perceptions. I see no evidence Germany's Greens are suddently rational when it comes to modern reactor designs, of which MSRs are one.
To be fair, a lot of nuclear reactors around the world should be shut down just due to age and outdated designs. However they should also be being replaced with modern reactors, which few people have, which makes shutting them down while we are still largely utilizing fossil fuel power and chemical plants really dumb.
By all the doomerism about German and nuclear there is at least Wendelstein 7-x doing frontier work. It's fine to get rid of legacy nuclear if there is a feasible bridge ahead.
Not sure what the point of this comment is. China has its equivalent EAST, France has ITER. Countries can do both fission and fusion research. To me the problem isn't that Germany closed some legacy reactors, but that too little is done into looking into alternative designs.
This type of progress shows China is capable of moving from an economy that’s build on labor arbitrage or copying others to genuine innovation. It’s also further evidence of the extreme competence of the CCP in governance, which I feel should be acknowledged despite their authoritarian negatives.
> This type of progress shows China is capable of moving from an economy that’s build on labor arbitrage or copying others to genuine innovation
China has been genuinely innovating in manufacturing techniques for decades. If anything, that ingenunuity peaked when Xi began his term, and has been degrading as his dictatorial tendencies needlessly hamstrung Chinese industry.
I don't think it makes sense to extrapolate from one particular technical field to governance in general.
The US managed to defeat both Nazi Germany and Japan plus develop nuclear weapons, all in 1941-5. Was it a proof of extreme competence of the US government in general? The some government tolerated abuse of blacks and forced segregation in the South, I would call it a serious governance failure.
China is far better at long term societal planning. Ultimately, I expect they will be the ones who can solve the climate crisis, after being one of the biggest contributors to the problem.
Sure, but most of that is from industrial production, and really should be debited on importing nations’ CO2 accounts. Whereas in the US transportation, heating and construction are the main consumers.
We have basically limitless carbon free energy with the tech we have now: solar, wind, batteries, fission breeders, large power grids that can move power around cheaply, etc. Put all those together and we have incredible energy abundance.
We also have the ability to electrify most transport except maybe long haul trucking and long haul aviation. Aviation (ALL aviation) accounts for less than 5% of global CO2 emissions, which means we could leave that alone and cut elsewhere until we have batteries and other infrastructure good enough for that.
Build all this out and it'll be cheaper and more scalable than what we currently have.
We in the USA choose to stick with ancient technology because we have a sunk cost and an existing political power structure built around it. Meanwhile China is eating our lunch. Make America Great Again! By... pretending it's 1945 and trying to LARP the previous century.
Classic innovators' dilemma at the national level.
I'd love to see as much electrification as possible.
On the aviation note, sadly, aviation bats higher than its C02 accounting. Contrails add another 1-2% on top of contribution from it's C02 emissions. It's entirely avoidable and could be resolved at relatively low cost.
> We in the USA choose to stick with ancient technology because we have a sunk cost and an existing political power structure built around it.
Yes, and also vast oil and gas reserves China doesn't have.
Also there is strong public fear and dislike of nuclear power.
In countries where there are no or little fossil fuels it is mainly this public opinion which has crippled the nuclear industry. Germany is a prime example.
Public opinion is obviously much less important in China.
> Public opinion is obviously much less important in China.
That really isn’t true. The reason Shanghai didn’t expand their maglev to Hangzhou is because residents were worried about electrical magnetic radiation, which I don’t think is really a thing. Nuclear took a long time to get started in China because people thought the government to be inept and corrupt, an image that has only recently faded away in the last decade. Without free elections, public opinion is actually much more important if you want to avoid economically destructive riots.
But this all happens in back rooms, the legal system isn’t very relevant, so if you have an issue but it isn’t a very popular one, you don’t really have any recourse. For example, niche environmental issues, or ones that aren’t widely recognized yet as dangerous…
China has distracted the USA energy focus by dumping cheap solar panels here while continuing to develop advanced nuclear generation capabilities at home.
China is simply betting on all horses: solar, wind, thorium, batteries, coal even, anything to not buy foreign oil and be as independent, self-sufficient as possible.
Seems like it's working too
Yup.. Happens to align somewhat with climate goals too, luckily for the rest of us. Once solar+batteries becomes the cheapest form of generation, the coal usage should also drop, if that isn't already the case
> Once solar+batteries becomes the cheapest form of generation, the coal usage should also drop
Marginal versus bulk. It can make sense, economically, to keep building coal plants even if solar is cheaper if you’re building solar as fast as you can and still need more power.
> I don't see the U.S rushing to adopt either renewables or nuclear. We're just increasing our fossil fuel burning (natural gas)
This is wrong. Natural gas is falling from 42% of U.S. electricity generation in '23 and '24 to 40% in '25E and '26E [1]. Renewables, meanwhile, keep marching from 23% ('24) to 24% ('25E) and 26% ('26E). (Nuclear falls from 19% ('24) to 18% ('25E and '26E).
Before anybody gets too excited, it's better to understand what exactly happened.
China ran an experimental reactor that achieved some conversion of thorium into uranium. More precisely, the conversion ratio was 0.1 [1]. This means that for each new fissile atom generated from thorium (i.e. uranium-233) 10 atoms have been burned from the original fissile inventory.
Now, conversion happens in every nuclear reactor. Some new fissile material (generally Pu-239) is generated out of "fertile material" (generally U-238). And, surprisingly, that conversion ratio is quite high: 0.6 for pressurized light water reactors and 0.8 for pressurized heavy water reactors [2].
What China has achieved therefore is well below what is business as usual in regular reactors. The only novelty is that the breeding used thorium, rather than uranium.
Is this useless? No, it is not. In principle increasing the conversion ratio from 0.1 to something higher than 1.0 should be doable. But then, going from 0.8 in heavy water reactors to more than 1.0 should be even easier. Why don't people do it already? Because the investment needed to do all the research is quite significant, and the profits that can be derived from that are quite uncertain and overall the risk adjusted return on investment is not justified. If you are a state, you can ignore that. If China continues the research in thorium breeding, and eventually an economically profitable thorium breeder reactor comes out of that, the entire world will benefit. But the best case scenario is that this would be three decades in the future.
[1] https://www.world-nuclear-news.org/articles/chinese-msr-achi...
[2] https://en.wikipedia.org/wiki/Breeder_reactor#Conversion_rat...
My understanding is that reactors will use that plutonium just fine, so the energy you get from a fresh fuel rod is almost exclusively from uranium fission but, as time goes on, an increasingly large share is from plutonium fission.
In principle, using Thorium would give you the energy from Thorium fission, then Uranium fission, then plutonium fission, which is pretty cool. However, I suspect you might hit an issue here where such a low conversion rate would make the reactor go sub-critical.
Fundamentally the problem is that Uranium is so damn energy dense and abundant enough that there's little need to set up these complicated recycling systems. If we start to run out of Uranium then this technology starts to look appealing, but in the modern day it just doesn't make economic sense.
> Uranium is so damn energy dense and abundant enough that there's little need to set up these complicated recycling systems
Uranium is abundant, but not homogenously so [1]. (China has some. But not a lot. And it's bound up expensively. And it's by their population centres.)
For the Americas, Europe, Australia, southern Africa and Eastern Mediterranean, burning uranium makes sense. For China, it trades the Strait of Malacca for dependence on Russia and Central Asia.
[1] https://www-pub.iaea.org/MTCD/Publications/PDF/Pub1800.pdf
Uranium can be stockpiled for years in advance, relatively easily. Enough to tide over a small war while you're setting up domestic production. And China should have enough low-grade ores for that.
There's not that much Uranium actually that's economically sensible to extract. The NEA says in their 2024 report on Uranium [1]:
> Considering both the low and high nuclear capacity scenarios to 2050 presented in this edition, and assuming their 2050 capacity is maintained for the rest of the century, the quantities of uranium required by the global fleet – based on the current once-through fuel cycle – would likely surpass the currently identified uranium resource base in the highest cost category before the 2110s.
Their "high" scenario assumes having a bit more than double of today's capacity by 2050; today we have about 4-5% supply from nuclear energy worldwide.
[1] https://www.oecd-nea.org/jcms/pl_103179/uranium-2024-resourc...
Out of curioosity, do they forecast at what point it becomes cheaper to breed than mine?
and we still don't know where to store the trash. Thorium seems better (but my knowledge is close to zero here, I must admit:-) )
> we still don't know where to store the trash
We really do. Nuclear waste is less toxic than plenty of trash we just bury. And calling it "waste" is a bit reductive, given it almost certainly becomes valuable to reprocess within another century or two.
The notable thing here is that it's a molten salt reactor design, where the fuel is dissolved in a molten salt (FLiBe). This allows online continuous processing of the fuel, unlike with solid fuel rods sealed inside a pressure vessel.
This unlocks a lot of options for the fuel cycle, including the use of thorium.
This work builds on a previous molten salt reactor experiment at Oak Ridge, decades ago. There's a whole lore about MSRs.
> notable thing here is that it's a molten salt reactor design
Notable, but not unique. The unique bit is it burns thorium.
It breeds thorium to fissionable uranium from a starting fissionable uranium starter fuel. It doesn't directly use thorium for fuel.
What people need to understand about the cycle efficiency is that when you mine uranium, the fissionable part of uranium (U-235) is only 1% of that uranium, the rest is nonfissionable U-238.
Thorium is about twice as abundant as Uranium (all isotopes). The MSR uses Thorium to create U-233, a fissionable but not naturally occurring Uranium isotope.
So the "unlimited energy aspect" is that about 200-300x more breedable Thorium exists than fissionable U-235.
A MSR nation could also try to breed U-238 into plutonium, which would provide another 100x more breeding stock, although LFTR never talked about U-238 breeding. IIRC the plutonium may be difficult to handle because of gamma rays, but I don't recall exactly.
While I don't have confidence that even LFTR/MSR reactors can get economical enough to challenge gas peakers, it may be possible to make truly price-competitive MSR electricity with the right modular design. I wish the Chinese the best of luck, because if they do it will spur the rest of the world to adopt this about-as-clean-and-safe-as-it-gets nuclear design.
> Thorium is about twice as abundant as Uranium
China has thorium, and while less than others [1], it’s better than they do with uranium [2].
> it may be possible to make truly price-competitive MSR electricity with the right modular design
Yes. But probably not in the near term with thorium. This isn’t designed to be cheaper. It’s designed to be more available to China than being dependent on Russian deposits.
[1] https://www.nature.com/articles/492031a
[2] https://www-pub.iaea.org/MTCD/Publications/PDF/Pub1800.pdf
That's what you learn playing factorio
Not really, the US bred several tons of U-233 from thorium in the 60s.
https://en.wikipedia.org/wiki/Thorium-based_nuclear_power
MSRs have some attractive features, but they also have significant drawbacks.
The most pressing is that fissionable material is spread throughout the fluid, so fission and decay of fission products is occurring right up to the edge of the fluid. The walls and pipes containing the molten salt, and anything dipped into the salt, are exposed to unmoderated neutrons. One can shield using (say) graphite, but then damage to that (and soaking up of radioactive materials) become issues.
The Molten Salt Reactor Experiment at Oak Ridge was near the end of its radiation exposure lifetime when the program ended.
Contrast this to light water reactors. These are designed so that no lifetime component sees unmoderated neutrons. There's a thick barrier of water between the fissioning fuel and the reactor vessel wall and the support structures for the fuel bundles. The bundles themselves are exposed, but they are replaced for refueling and are not lifetime components.
To add to this, even with the shielding provided by water in light water reactors, the neutron exposure is _the_ limiting factor for the reactor vessel.
The metric to look for is called "DPA" (displacements per atom), the number of neutron collisions that a material can tolerate before losing enough structural integrity to fall below the acceptable limits. The best modern reactor steels are at 150-180 DPA.
And a lot of potentially cool reactors like TWR (travelling wave reactor) end up being logistically impossible because lifetime-limited components will be exposed to multiple hundreds of DPAs.
What absorbs the neutrons then?
The thorium cycle is generally neutron negative.
> thorium cycle is generally neutron negative
Source for the fuel cycle?
Thorium 232 -> 233 is neutron negative. But after that you get all kinds of nonsense.
Thorium 232 is the thorium in the cycle yes. And all kinds of nonsense is correct for the daughter products. But in general, to actually use do anything with thorium you need excess neutrons.
Even the daughter uranium 233 only produces on average 2.48 neutrons per fission, so it’s very difficult even in a combined lifecycle process to have enough - thorium doesn’t produce uranium 233 immediately (takes almost 30 days), neutron capture with that low a ratio requires a LOT of thorium, which is going to mostly just suck up all neutrons and you won’t have any extra for addition uranium 233 fissions, etc.
It’s quite difficult (impossible?)to have actually work without a source of a large amount of additional neutrons.
> to actually use do anything with thorium you need excess neutrons
Unless 100% of those neutrons is being absorbed by the thorium, this means you'll have neutron flux at the boundary. Which, for a liquid moderator, means all the pipes and tanks and pumps.
A better explanation of the significance:
https://www.stdaily.com/web/English/2025-11/17/content_43298...
Thanks! We've put that link in the toptext as well.
> Now, the research team is conducting systematic studies on the key scientific issues related to adding thorium, and aims to completethe construction of a 100-megawatt TMSR demonstration project, and begin operation by 2035.
For comparison: A commercial nuclear power plant is 1 gigawatt, a 10x difference. I assume this would be the next step.
The typical 1 gigawatt rating for a nuclear power reactor is measuring electrical output. Given the various inefficiencies, the actual reactor output (as heat) is something like 3x that amount. Whereas a research reactor will be quoted as thermal output.
That to say, a typical commercial reactor might be 30x the power of a 100 MW research device.
This post is just an excerpt—it's the first 4 paragraphs of a 29-paragraph article,
https://archive.is/DQpXM ("China reaches energy independence milestone by ‘breeding’ uranium from thorium"–SCMP)
Thanks! - since you've found a readable link, I've swapped https://humanprogress.org/china-reaches-energy-milestone-by-... out for the original source at the top, and put the archive link in the top text.
This came up several times the last few weeks, but never stayed on the front for long. Also no comments.
I guess soon the west has to copy chinas tech.
> soon the west has to copy chinas tech
Thorium MSRs don't make sense for the Americas, Europe or Australia. We have plenty of uranium.
Nuclear is receiving solid research backing in both America and China. (India is playing too. Austrlia is having an identity crisis.) Our different geologies mean there will probably be one solution for China, India and North Africa, on one hand, and the rest of the world, on the other hand.
Breeding is a technology looking for a business case.
It's more expensive than just using fresh uranium in current market conditions. It's a way from keeping future uranium shortages from making nuclear power more expensive; it's not a way to make nuclear cheaper than it currently is.
Emphasis on current market conditions. Relations with uranium mining countries and environmental opposition to uranium mining could shift conditions.
The truth is that nuclear power is not that financially attractive at the present and would the price of uranium rise enough that breeders would become economically viable most countries would just stop bothering with nuclear power altogether.
The cost of nuclear power is almost entirely capex and financing, not opex. Uranium input cost for nuclear power plants is 0.5c/kWh. With breeders you can divide that by about 100.
At least as of a couple years ago nuclear costs just a little more than solar plus storage and that’s not stopping anyone heh.
With recent price drops of solar and storage the difference is now multiples.
Capex and financing is still an issue for many countries, and the opex is a non-zero commitment beyond just the fiscal portion. Most countries that pass-over nuclear energy are fairly justified in their decision. The status-quo is still not super psyched about nuclear proliferation.
There is room to change that, but the cards are very heavily stacked in China's favor. America's bad at the financing part, fickle when it comes to enforcement & supply chains, and ostensibly 2 days away from bailing on the IAEA itself. The proliferation-resistance of Thorium reactors gives China an export trump card that America will struggle to match.
> The truth is that nuclear power is not that financially attractive
Let me fix that for you: "The truth is that nuclear power is not that financially attractive in the bureaucratic high cost litigious Anglo-sphere". And that's pretty much all infrastructure these days, unfortunately.
They’re not financially attractive in other parts of the world either. China, a zero litigation single party state, is building some but a tiny % compared to their renewable buildout
They need a lot of energy from a variety of sources. China has 30 or so reactors under construction (half or so of all active projects).
"China currently has 58 operable reactors with a total capacity of 56.9 GW. A further 30 reactors, with a total capacity of 34.4 GW are under construction" [1]
So, yes, but...
China installed 256GW of solar in the first 6 months of 2025 [2]. A full year estimate of ~350gw. So, the total of all nuclear under construction is 1/10th of the solar they installed in one year.
Don't get me wrong, its cool to see diversity of non fossil sources, glad they are building some, but its a niche in their overall energy buildout. And they can only build that small niche because they dont have to be market priced, its state subsidized.
[1] https://www.world-nuclear-news.org/articles/ten-new-reactors... [2] https://ember-energy.org/latest-updates/global-solar-install...
It's not the litigiousness that makes it expensive. France was producing nuclear power plants at a cost per watt that nearly matches modern China. In fact, the mind-numbing cost overruns seem unique to the US.
Here's a Nature article about it:
https://archive.ph/Tpe0j
Seems to me like it's more of a story of corruption than of over-regulation
france cant do it any more either. Flamanville was 12 years late and [1] 400% over budget. EPR2 is already delayed and over budget and they havent even started building yet!
UK cant do it either, see hinkley point c [2]
[1] https://www.nucnet.org/news/long-delayed-nuclear-plant-conne... [2] https://www.world-nuclear-news.org/articles/edf-announces-hi...
China has more uranium reserves and less thorium reserves than the US though
Most thorium: India, Brazil, Australia, US, Turkey
Most uranium: Australia, Kazakhstan, Canada, Russia, Namibia
They highlight less the advantages from breeding, than other advantages of the molten salt design, like not needing a lot of cooling water, which allows this reactor to operate in the Gobi desert, the possibility of replacing the fuel without halting the reactor and various safety features.
> other advantages of the molten salt design, like not needing a lot of cooling water
This advantage is conserved by all non-water moderated reactor designs.
Nuclear reactors don't need a particularly big amount of cooling water.
The thermodynamic cycle needs a cold source though, and it's most commonly water. This doesn't depend on the reactor design and this is equally as true of coal plants.
As long as you are making electricity out of a thermodynamic cycle, you need a heat source (be it a flame or a nuclear reaction) and a cold source.
It also apparently provides a way to make reactors that don’t depend as much on water so they don’t all have to be near the coast.
This would allow Western China to also develop reactors to help underpin their renewable and coal energy.
> The interest in MSR technology and Thorium breeding did not disappear however. China's nuclear power production relies heavily on imported uranium,[10] a strategic vulnerability in the event of i.e. economic sanctions. Additionally, the relative lack of water available for cooling PWRs west of the Hu line is a limiting factor for siting them there.
https://en.wikipedia.org/wiki/TMSR-LF1?wprov=sfti1#History
> also apparently provides a way to make reactors that don’t depend as much on water so they don’t all have to be near the coast
Non-water microreactors broadly fall into two categories: ones using a different moderator, most commonly sodium, a sodium salt or helium; and those using heat pipes. Most microreactor designs don’t use water.
Nuclear plants don't need more water than a coal plant of the same power, they both use the same steam turbine with water as cold source.
There is no business case for basic research, but if you stop basic research long enough you will have no business. The United States and its allies seem to have completely forgotten this.
It makes sense for big monopolies like Bell, or the CCP. The investment can be justified if the ones investing are confident they will be able to capture the value and not some competitor.
Bell Labs also served to maintain positive perceptions of the monopoly. Unix was famously developed despite the knowledge that AT&T would not be able to offer it as an independent product.
> in current market conditions.
That is, as long as we don't build more nuclear power plants.
If you want to increase nuclear power adoption, then you're not going to stay in “current market conditions” for long.
Reducing the energy sector to pure business would probably work in the 1990s, but not now, when countries are afraid of strategic dependence on potentially hostile suppliers.
Uranium isn't as ubiquitous as, say, natural gas, and stockpiling it comes with a big heap of physical problems. I can definitely see countries spending on more expensive technology if it comes with more energy security.
I mean we're already doing that in many avenues. Solar being the most obvious. The only functioning solar manufacturing plants in the US are Chinese-owned and are only here to take advantage of subsidies.
Plenty has been learned by the US/West from copying their approach to agriculture, robotics in factories, mining, drones, etc. Have you seen their electromagnetic catapult technology?? That stuff seems like its from the space-age! There's even plenty of tech that we can't really explain like the all-moving wingtips on the new J-50s. (and yes, I'm avoiding talking about their supersonic cruise missiles)
I think I read recently that this was a US idea that was abandoned that China took up and made it work. Is that accurate?
No country has seriously invested in the thorium fuel cycle because it cannot be used to create weapons. Unfortunately, the technology also began to look most promising as an energy source around the same time the Three Mile Island nuclear accident effectively ended all interest in nuclear energy in the United States.
India has shown some of the most interest to date, due to their lack of domestic uranium reserves. But it's been slow going their fast breeder reactor plans were delayed by like two decades. But it is built and it was loaded with fuel last month [0]
The French interest in breeder reactors and nuclear reprocessing also originates from a similar concern about lack of domestic access to raw uranium. Though Super-phoenix [0] was a more traditional uranium -> plutonium approach and not thorium. They gave up because just using uranium is way, way cheaper than synthesizing your own fissile materials.
[0] https://www.world-nuclear-news.org/articles/indias-prototype...
[1] https://en.wikipedia.org/wiki/Superph%C3%A9nix
Thorium can be used to make weapons via the breeding cycle. It's much less convenient and straightforward than uranium/plutonium, but it is possible.
Theoretically, perhaps, but I don’t think anyone with a serious interest in weapons would pursue it. From a nonproliferation perspective, I’d guess the infrastructure necessary to remove contaminants from uranium bred through the thorium cycle would be costly and difficult to conceal.
Multiple countries have detonated nuclear bombs using U-233 derived from thorium reactors! [0] Practically I agree with you that thorium is proliferation resistant and if someone is bomb hungry they won't prioritize it, but if you want to set up the bomb and all you have is thorium... The infrastructure wouldn't necessarily be significantly larger or worse than conventional enrichment.
0 - https://en.wikipedia.org/wiki/Uranium-233
Seems presence of U-232 is more manageable than I thought.
Also, it’s only energy positive under some specific carefully managed conditions, and is a real pain to make work.
If you have easy access to uranium, you just use it directly instead.
Well, fible energy is trying to do lots. Gates invested in MSTR (molten salt thorium reactor).
But regulation, while it has its purposes, stifles many things. At the same time time it’s not even doing what they were meant for.
There are a number of countries being run far better than the US or the EU
> There are a number of countries being run far better than the US or the EU
It will be funny if China is what convinces the US to be more open to free industry. Opposite day vs the 1970s
To be fair, these advances are not being made in China due to "free industry". They have something of a command economy for their critical sectors. So it's unfair not to point out that it's easy to make advances if a nation as a whole points to a hill and says, "take that hill". Of course you can do it under those circumstances.
If it's just your company or some trifling consortium trying to develop nuclear energy advances in a "free industry" environment, the guy who is just slapping up windmills, [T Boone Pickens RIP], is just gonna mop the floor with you. There's just no way to compete on moonshots like that.
China has many capital controls but generally supports industrial activity in the state interest; the US does the opposite.
Yes: https://en.wikipedia.org/wiki/Thorium-based_nuclear_power#Hi...
Historical experiments with alternative fuel cycles, not serious development attempts. A serious development attempt happened in India though.
Meanwhile Germany just decommissioned its last nuclear reactors. Given the challenges of baseload renewable generation, it's frustrating to watch working infrastructure being dismantled while we're still heavily dependent on fossil fuels.
Comparing those old conventional reactors to MSR is not suitable at all. And they were not fully functional past their expiry date.
> Comparing those old conventional reactors to MSR is not suitable at all
It is given we're talking about perceptions. I see no evidence Germany's Greens are suddently rational when it comes to modern reactor designs, of which MSRs are one.
To be fair, a lot of nuclear reactors around the world should be shut down just due to age and outdated designs. However they should also be being replaced with modern reactors, which few people have, which makes shutting them down while we are still largely utilizing fossil fuel power and chemical plants really dumb.
By all the doomerism about German and nuclear there is at least Wendelstein 7-x doing frontier work. It's fine to get rid of legacy nuclear if there is a feasible bridge ahead.
Not sure what the point of this comment is. China has its equivalent EAST, France has ITER. Countries can do both fission and fusion research. To me the problem isn't that Germany closed some legacy reactors, but that too little is done into looking into alternative designs.
There's a danish company building modular container sized molten salt reactors.
https://www.copenhagenatomics.com/
This type of progress shows China is capable of moving from an economy that’s build on labor arbitrage or copying others to genuine innovation. It’s also further evidence of the extreme competence of the CCP in governance, which I feel should be acknowledged despite their authoritarian negatives.
> This type of progress shows China is capable of moving from an economy that’s build on labor arbitrage or copying others to genuine innovation
China has been genuinely innovating in manufacturing techniques for decades. If anything, that ingenunuity peaked when Xi began his term, and has been degrading as his dictatorial tendencies needlessly hamstrung Chinese industry.
"extreme competence of the CCP in governance"
I don't think it makes sense to extrapolate from one particular technical field to governance in general.
The US managed to defeat both Nazi Germany and Japan plus develop nuclear weapons, all in 1941-5. Was it a proof of extreme competence of the US government in general? The some government tolerated abuse of blacks and forced segregation in the South, I would call it a serious governance failure.
Yea but afaik China doesn’t have that kind of issue. They do have an issue with anticommunists but I’m not sympathetic to their cause.
They very much d have that sort of issue and worse. Uighurs and other minorities, treatment of gays....
Thorium is abundant in Sri Lanka’s mineral sands. Mined with dredgers at shallow depths 10-100m off the western coast.
A detailed explanation of the Thorium Fuel Cycle [1].
I'm glad China is doing this even though I'm skeptical about nuclear power ever being commercially viable. At least they're trying different things.
[1]: https://www.youtube.com/watch?v=2IqcRl849R0
I wonder if people would think China copied this from the West.
China is far better at long term societal planning. Ultimately, I expect they will be the ones who can solve the climate crisis, after being one of the biggest contributors to the problem.
> after being one of the biggest contributors to the problem.
How many "Made in China" products do you have at home right now? Who is contributing to the problem?
Actually by far the biggest, in terms of total greenhouse emissions (30% of the world). Though other countries emit a lot more per capita.
Sure, but most of that is from industrial production, and really should be debited on importing nations’ CO2 accounts. Whereas in the US transportation, heating and construction are the main consumers.
[flagged]
We have basically limitless carbon free energy with the tech we have now: solar, wind, batteries, fission breeders, large power grids that can move power around cheaply, etc. Put all those together and we have incredible energy abundance.
We also have the ability to electrify most transport except maybe long haul trucking and long haul aviation. Aviation (ALL aviation) accounts for less than 5% of global CO2 emissions, which means we could leave that alone and cut elsewhere until we have batteries and other infrastructure good enough for that.
Build all this out and it'll be cheaper and more scalable than what we currently have.
We in the USA choose to stick with ancient technology because we have a sunk cost and an existing political power structure built around it. Meanwhile China is eating our lunch. Make America Great Again! By... pretending it's 1945 and trying to LARP the previous century.
Classic innovators' dilemma at the national level.
I'd love to see as much electrification as possible.
On the aviation note, sadly, aviation bats higher than its C02 accounting. Contrails add another 1-2% on top of contribution from it's C02 emissions. It's entirely avoidable and could be resolved at relatively low cost.
https://contrails.org/faq/#how-are-contrails-contributing-to...
> We in the USA choose to stick with ancient technology because we have a sunk cost and an existing political power structure built around it.
Yes, and also vast oil and gas reserves China doesn't have.
Also there is strong public fear and dislike of nuclear power.
In countries where there are no or little fossil fuels it is mainly this public opinion which has crippled the nuclear industry. Germany is a prime example.
Public opinion is obviously much less important in China.
> Public opinion is obviously much less important in China.
That really isn’t true. The reason Shanghai didn’t expand their maglev to Hangzhou is because residents were worried about electrical magnetic radiation, which I don’t think is really a thing. Nuclear took a long time to get started in China because people thought the government to be inept and corrupt, an image that has only recently faded away in the last decade. Without free elections, public opinion is actually much more important if you want to avoid economically destructive riots.
But this all happens in back rooms, the legal system isn’t very relevant, so if you have an issue but it isn’t a very popular one, you don’t really have any recourse. For example, niche environmental issues, or ones that aren’t widely recognized yet as dangerous…
"Public opinion is obviously much less important in China."
In the US public opinion doesn't really matter either. It's the oligarchs' opinions that matter
Public opinion here appears to count for nearly jack squat so I don’t buy this explanation at all.
If grid energy was cheap enough, synthetic fuel for aircraft and trucks would be competitive.
> We in the USA choose to stick with ancient technology because we have a sunk cost and an existing political power structure built around it.
You don't want to discount the cultural attachment people have to what their parents did and their childhood.
China has distracted the USA energy focus by dumping cheap solar panels here while continuing to develop advanced nuclear generation capabilities at home.
China is simply betting on all horses: solar, wind, thorium, batteries, coal even, anything to not buy foreign oil and be as independent, self-sufficient as possible. Seems like it's working too
Seems like a wise thing to do too.
Yup.. Happens to align somewhat with climate goals too, luckily for the rest of us. Once solar+batteries becomes the cheapest form of generation, the coal usage should also drop, if that isn't already the case
> Once solar+batteries becomes the cheapest form of generation, the coal usage should also drop
Marginal versus bulk. It can make sense, economically, to keep building coal plants even if solar is cheaper if you’re building solar as fast as you can and still need more power.
How exactly has it distracted the U.S?
I don't see the U.S rushing to adopt either renewables or nuclear. We're just increasing our fossil fuel burning (natural gas).
> I don't see the U.S rushing to adopt either renewables or nuclear. We're just increasing our fossil fuel burning (natural gas)
This is wrong. Natural gas is falling from 42% of U.S. electricity generation in '23 and '24 to 40% in '25E and '26E [1]. Renewables, meanwhile, keep marching from 23% ('24) to 24% ('25E) and 26% ('26E). (Nuclear falls from 19% ('24) to 18% ('25E and '26E).
[1] https://www.eia.gov/outlooks/steo/