Dear Internet, we need to have a talk about Thorium. It has many good attributes as a nuclear fuel, but the things being said on the internet have become largely misleading, if not all-out inaccurate. Every internet person I meet in real life who finds out that I am a nuclear engineer asks me why we aren’t using the end-all, be-all that is thorium. Every post regarding nuclear energy on reddit is packed full of comments claiming that Thorium will end all concerns about nuclear energy and that Uranium is only in use due to some dark dark conspiracy.
Some places on the internet have become echo-chambers for this kind of thing, and while it’s great to spread awareness of thorium, blatant disregard of the associated challenges is a detriment to civilization’s energy debate. Besides, taking a moderate viewpoint lends credibility to any cause. This page will try to point people in the right direction if they get lost, using things like references and whatnot. And we’ll make a wall of shame where anyone who perpetuates a misconception will get to be displayed.
To learn more about Thorium, we feature a page about Thorium as nuclear fuel, as well as a big page about the fluid fueled molten salt reactors (MSRs) that are good at using it. If you think we’re too negative-nancy here, go check out those pages.
We love Thorium and think it has a bright future, both in solid and fluid fueled reactors. I personally have studied it a huge amount and many years ago considered getting a THORIUM vanity plate. As we claim elsewhere and throughout comment posts abound, we just think that people need to remain calm and accurate when discussing its merits and demerits.
Quite False. Not only can they be used to make bombs (see Misconception #3), but they also were not canceled for any weapons-related reason. One of the most lucid descriptions of what happened to molten salt reactors like the LFTR can be found on page 49 of WASH-1222 . There, they describe a few privately-funded working group studies of the MSBR, including the Molten Salt Breeder Reactor Associates (consisting of the engineering firm Black & Veatch and five midwestern utilities) and the Molten Salt Group, headed by Ebasco Services, Inc. (with 5 other industrial firms and fifteen utilities involved). These groups concluded that the MSBR (basically the LFTR) is attractive and potentially cheaper than LWRs. They said that a demonstration plant is warranted, but the performance cannot be predicted with confidence. Then, a list of factors that limit industrial involvement is given. They include (verbatim):
Weapons were produced with graphite or heavy-water moderated production reactors and with gas centrifuge enrichment. Oh, and thermonuclear weapons require tritium as well, which is something that many Thorium MSR designs excel in producing (darn that lithium!). The commercial LWRs had nothing to do with making bomb material. Stop the nonsense.
To be fair, you can rightly argue that U-Pu-fueled reactors got developed in the first place (in the Manhattan project of the 1940s) for weapons reasons. Back then (before enrichment), Th-fueled reactors couldn’t even go critical, much less make bombs. Natural uranium reactors were the only way to go. This gave them the technical head start that has arguably led to their dominance. However, when MSRs were finally given their chance in the 1950s and 60s, their (non-existant) inability to make bombs was not to blame for the cancellation.
Misleading at best. When people say this, they tend to imply that Th-fueled reactors are the only reactors that never need enrichment, which isn’t true. The nice thing about any breeder reactor (using Th-U or U-Pu) is that eventually they can become fissile self-sufficient, meaning they breed more (or equal) fissile material than they consume. The first electricity-producing reactor in the world (EBR-I in Idaho, 1951) was created to demonstrate that breeding was possible (in a liquid-metal cooled fast breeder reactor, or LMFBR). Any breeder reactor concept on the planet can run without additional enrichment (or some other external source of fissile material) after their initial startup by breeding fissile material out of fertile material like Th-232 or U-238. But you have to start your reactor up with fissile material from somewhere. If you take a vat of Thorium and try to turn it on, you'll be sorely disappointed because it cannot possibly sustain a chain reaction, under any circumstances. So you start it up with denatured bombs or enriched U-235 and then it becomes self-sufficient on Th-232 or U-238. I occasionally read misleading things that say Thorium will just fire right up. Alas.
It should be noted, however, that the key advantage of Th fuel is that it allows thermal breeding. This means that you can start up a Th-based breeder with substantially (between 3 and 10 times) less fissile material than you need to start an equivalent-powered fast breeder reactor. Once started, the fast breeder will make far more fissile material (because they make have a better breeding neutron economy), but the amount of fissile in fast spectrum reactors is always more than in thermal reactors.
TL;DR: They do to start up, and U-Pu breeders like the LMFBR can do the same so it’s not Thorium specific.
False! They can indeed make bombs. Thorium reactors work by breeding Th-232 through Protactinium-233 (27.4 day half life) and into Uranium-233, which is fissile. Pa-233 is a pretty strong neutron absorber, so the MSBR (basically the LFTR) has to extract it from the core once it is produced and let it decay to U-233 away from the neutrons. Once the U-233 is created, it gets fed back into the reactor. Well, if you went rogue, you could build up a little excess reactivity (maybe add some low-enriched U235?) and then divert the freshly-bred U-233 into a weapons stream to make U-233 nuclear bombs. It may be difficult to do this several times without going subcritical, but it certainly could be done. A U-233-filled bomb has been tested before, and it worked just fine.
Here’s a quote from a Frank von Hippel paper on the subject :
"On the one hand, gamma radiation from U-232 makes the U-233 from high- burnup U-233-thorium fuel cycles more of a radiation hazard than plutonium. On the other hand, because of its low rate of spontaneous-neutron emission, U-233 can, unlike plutonium, be used in simple gun-type fission-weapon designs without significant danger of the yield being reduced by premature initiation of the fission chain reaction"
And another (also ):
"In the case of the molten-salt U-233 breeder reactor, it was proposed to have continual chemical processing of a stream of liquid fuel. Such an arrangement also offers a way to completely bypass the U-232 contamination problem because 27-day half-life Pa- 233 could be separated out before it decays into U-233."
Options to make bomb-making less favorable include fostering substantial U-232 contamination in the reactor and denaturing the U-233 with U-238 that keeps the in-reactor inventory safe. Both of these options can conceptually be bypassed in the Pa separation route though. Besides, U-232 isn’t releasing the gammas, its decay products are, and it has a 70 year half-life. So you can just chemically purify your stolen goods and then make the bomb anytime within the next decade or so.
There are about a dozen other ways people try to amp up the proliferation resistance of various fuel cycles. But they always forget that the owner of such a plant can secretly install a chemical cell that does Pa separation. Really, most civilian power to bombs proliferation paths are mythical, in any reactor! But since the consequences of proliferation are so dire, nuclear power plants need to have baseline proliferation safeguards in place. Thorium-powered reactors, whether fluid fueled or not, are no exception.
This one is mostly true, but also partially false. The average crustal concentration of Thorium is 0.00060%, compared with 0.00018% for Uranium . But, the oceanic abundance of Th is 4x10-12%, compared with 3.3x10-7% (mass percent). Considering that the oceans contain 1.4x1021 kg of water, that amounts to 56,000 tonnes of Th and 4.62 billion tonnes of Uranium. Moreover, mining the entire crust is difficult, whereas the ocean delivers to you. While seawater extraction of uranium is not yet competitive with traditional mining (it’s hovering around 4x more expensive), it is possible and may become economical in the near future. So while Misconception 4 is correct with respect to the crust, it’s not necessarily relevant from a global resource perspective, and there may very well be more accessible Uranium available to us. The crust is estimated to weight around 1.0x1022 kg, so overall, there is actually more Th. If you want to get very technical and start including asteroid and star mining, the abundance of Th in the universe is estimated at about 2x that of Uranium.
If you’re the Indian government, however, you’ll note that you have hundreds of thousands of tonnes of Th but basically zero U. So you guys might want some Th-power to secure a domestic supply! China has about an estimated 50% more known U than Th [4,5].
Another point, if you look at the known reserves of economically extractable Thorium vs. Uranium [4,5], you’ll find that they are both nearly identical (though many people argue that we can economically extract Th from lots of common sands). And remember, if we close the fuel cycle (whether using Th-U or U-Pu), the fuel resources are a non-issue for millenia.
Undenatured Thorium cycles certainly produce fewer transuranic elements (Np, Pu, Am, Cm,+), which are the major dangerous nuclides in nuclear waste in the 10,000+ year timeframe. In fact, the long-term decay heat from Thorium-MSRs can be orders of magnitude lower than that from traditional reactors. However, this same capability exists in many other reactor concepts, including U-Pu fueled fast reactors with reprocessing. So, if someone says that MSR/LFTR waste is better than traditional LWR waste, they are correct. If they say Thorium is the only game in town that can reduce waste like this, then they are not correct.
On one hand you can choose between a Th-U fuel cycle and a U-Pu fuel cycle. On the other hand, you can choose between a fluid fueled reactor (like a MSR) or a solid fueled reactor (like a LWR or a sodium-cooled fast reactor). Now, the Th-U cycle works really really well with MSRs, and that’s why they are often discussed together. There’s nothing wrong with this, but it’s nice to know what benefits come from which choice. The Th-U fuel cycle can be (and has been) used in solid fueled reactors and the U-Pu fuel cycle can be (and has been) used in MSRs.
The attributes of a system that come from choosing a fluid fueled reactor include: the ability to have passive safety by draining the fuel into cooled storage tanks, online fission product removal, low/zero fabrication cost, low fissile requirement, low excess reactivity (since you refuel online) 
The attributes that come from choosing the Th-U fuel cycle over the U-Pu cycle include: the possibility of thermal breeding (as demonstrated in the Shippingport LWR), the reduced production of minor actinides (see Misconception #5), allowing nuclear waste to be safer without aggressive reprocessing, and the ability to use the Thorium mineral base instead of the Uranium minerals (useful if your country has Th but no U. See Misconception #4).
Here we get to publicly shame those who propagate the myths. You can do better!
Update:The American Nuclear Society has issued Nuclear Technology entitled "The Reemergence of the Thorium Fuel Cycle" that is meant to "provide an even-handed description of its inherent attributes, and identify some of the data gaps that have yet to be resolved."
The papers therein cover a wide variety of the things discussed on this page.
Questions? Corrections? Comments? See someone who needs to be added to the wall of shame? Send us a note.
REDDIT ALERT A fairly lively discussion of the content of this page happened on Reddit in this thread.
Swing by our main page to nuclear energy in general.