|Fig. 1: Monazite, the mineral that is the world's primary thorium source. (Source: Wikimedia Commons)|
Thorium reactors have been the subject of optimistic hype in recent years, with claims that they would be safer, produce less hazardous waste, be easier to find fuel for, and be harder to use for nuclear proliferation than traditional light-water uranium reactors. All of these claims should be examined before the new technology is pursued, so that anyone who chooses to build a thorium reactor will be aware of the risks that have been obscured by the hype. Like all nuclear reactors, thorium reactors do pose safety and proliferation risks (in addition to technological challenges), and precautions to minimize these risks should be taken if thorium reactors are to be used.
Thorium and uranium reactors use different fission reactions, which means the reactions have different products and properties. Uranium reactors use uranium-235 as their fissile fuel, as well as the plutonium-239 that is produced from the uranium-238 that is also present in the fuel, and this reaction produces enough neutrons to perpetuate the reaction until the fuel is used up or the control rods are used to absorb the neutrons and slow the reaction. [1-3] Not all the plutonium produced by this reaction is broken down, and significant quantities of it remain in the waste. [2,3] In thorium reactors, thorium-232 (the most common isotope naturally found) becomes thorium-233 when it absorbs a neutron, and this unstable isotope decays into protactinium-233 and then quickly into uranium-233, which is the fissile fuel. [2,4] This production of fissile fuel from non-fissile but fertile fuel is called breeding, so the thorium reactor is a type of breeder reactor, with the same risks of instability that are found in uranium-based breeder reactors.  However, the fission of uranium-233 doesn't release enough neutrons to perpetuate the reaction, so more neutrons must be pumped into the reactor to keep it going. [2,4,5] Also unlike the uranium reactor, no plutonium is produced in the thorium reactor, and if plutonium is added to the reaction, it is broken down, becoming a neutron source for the reaction, and does not remain in the waste in significant quantities. [2,3]
Plans for thorium reactors often depict them as molten salt reactors (only the plans exist, since no commercial thorium reactors have been built). [1,4,5] The fuel is a liquid that contains the isotopes involved in the fission chain reaction as well as a salt such as a molten fluoride, and proponents of this design argue that this fuel is expected to be stable at higher temperatures than uranium fuel rods. [1,6] Unlike the solid fuel rods of a uranium reactor, where the concentration of neutrons and the different isotopes can vary from point to point, the liquid fuel can circulate and mix, so the composition is more or less constant throughout.  The liquid fuel could also flow through a unit that extracts the fuel-poisoning (in other words, neutron-absorbing) fission products continuously, whereas the fission products stay in the solid-core uranium reactor until the spent fuel rods are removed. 
Solid-core thorium reactors have also been suggested, with the idea that thorium fuel rods could replace the fuel rods in existing uranium reactors with minimal modifications required.  The neutron source in this case would be a second type of fuel rod that would be present next to the thorium rods and would contain uranium-235 and/or plutonium recovered from uranium reactor waste, which would release the necessary neutrons during fission.  Such a reactor would be a breeder that produces enough neutrons to perpetuate its own reaction, and would therefore be at risk of a runaway reaction the same as uranium reactors are.
One of the main arguments in favor of thorium reactors is that the nature of the thorium reaction, especially as liquid fuel, make it resistant to meltdowns, and therefore safer. This argument doesn't address the full picture, however, since the breeding process still comes with dangers of instability, and the waste is a major health hazard. Proponents argue that since thorium reactors operate subcritically, they can't sustain their own fission chain reactions, so in theory the runaway chain reactions that cause nuclear meltdowns would not occur, and if there were a power failure, a reaction supplied with neutrons by a particle accelerator would stop on its own. [2,4] Additionally, the uniform isotope concentrations found in liquid fuel make it more predictable: it's less likely that one part of the fuel will suddenly overheat, compared to a solid fuel rod with a potentially damaged, non-uniform crystal structure.  However, not all thorium reactors use easy-to-turn-off accelerators to provide the necessary neutrons, and problems similar to those of uranium-based breeder reactors arise when the thorium reaction does have enough neutrons to continue. The thorium reactor that India is working on uses a plutonium core to provide the neutrons, so it is a fast-breeder reactor, and would be more difficult to shut down if something went wrong. [5,7] It is also a solid-fuel heavy water reactor, so it lacks whatever high-temperature stability benefits a molten salt reactor might have.  Even if the reaction's stability were never a problem, the waste still would be, since a side product of the thorium reaction is uranium-232, which decays into a string of isotopes that are strong gamma ray emitters. [3,4] Gamma rays are a serious health risk and difficult to shield against, so the waste produced by thorium reactors would still be an issue. [3,4]
Supporters of thorium reactors argue that, since the thorium chain reaction does not produce plutonium, and its waste contains uranium-232 and is therefore difficult to handle, thorium would be a difficult and inefficient way to produce nuclear weapons. [1,3] Again, this statement is misleading. Fissile plutonium is not the only substance that is considered a proliferation risk: just 8 kilograms of uranium-233 is considered enough to construct a bomb, and uranium-233 is exactly what thorium reactors breed.  The issue of uranium-232 is also not a dealbreaker for nations that wish to make nuclear weapons, since there are well-known ways to chemically separate protactinium-233 from the thorium reaction, so that it can decay into uranium-233 separately, with minimal uranium-232 contamination.  The protactinium-233 would produce considerable heat as it decays - 50 Watts per gram - so it would be difficult to handle large quantities, but if small amounts were processed in parallel, this process could be manageable.  For a nation that already has thorium reactors for civilian power, those reactors could easily be used to irradiate thorium into extractable protactinium while still producing electricity, making it difficult for other nations to notice the proliferation risk.  This is especially true if the the power plant normally reprocesses its own fuel, since the separation of uranium-232 from the spent fuel is a normal step in reprocessing.  A small laboratory is all that would be required in addition to the thorium reactor itself, so uranium-233 weapon production would certainly be within reach for a government, if not a terrorist group. 
Although there is currently no way to gauge exactly how much thorium or uranium is in the Earth's crust, existing geological data implies that there is about three to four times as much thorium as there is uranium, by weight.  The largest known deposits are in Australia, Brazil, Turkey, Norway, China, India and the U.S.  China, India, the U.S., and the U.K. are considering thorium reactors, so three of these four would be able to acquire this fuel without needing to depend on foreign nations. [4,5,7] Nonetheless, the Earth's uranium supplies have not been used up to such an extent that its scarcity is driving us to thorium for cost reasons just yet: it may even be possible to make Earth's uranium fuel resources last for thousands of years if closed-cycle reactors are used, although there would still be the proliferation and instability risks of breeder reactors.  The technology readiness level of thorium reactors is far enough behind that of the uranium reactors that it could take decades to develop a working thorium reactor, so the cost of developing the technology may not be worth the change in fuel type, at least from a financial perspective. [5,6] Molten salt reactors would be especially expensive to develop, since new reactors would need to be constructed instead of refitting old uranium reactors to use thorium. [2,3]
Thorium has its share of risks as a nuclear fuel, and it isn't obvious that it would be a significant improvement over uranium in terms of safety, proliferation prevention, or cost. If an easy-to-switch-off accelerator is used to supply neutrons, and a molten salt reactor is used, then some (though not all) of the main failure modes of uranium reactors could become less likely, but at the cost of developing new technology and constructing new facilities, while also risking hard-to-detect proliferation. If India or the other nations considering thorium do develop thorium reactors, they should be aware of the risks that come with the new technology, and the the additional vigilance that would be needed to minimize nuclear proliferation by way of thorium.
© Ashley Micks. The author grants permission to copy, distribute and display this work in unaltered form, with attribution to the author, for noncommercial purposes only. All other rights, including commercial rights, are reserved to the author.
 M. M. Waldrop, "Nuclear Energy: Radical Reactors," Nature 492, 26 (2012).
 T. Dean, "New Age Nuclear," Cosmos, 13 Jun 06.
 P. Fairley, "Cleaner Nuclear Power?" Technology Review, 27 Nov 07.
 S. F. Ashley et al., "Nuclear energy: Thorium Fuel Has Risks," Nature 492, 31 (2012).
 L. Pham, "Considering an Alternative Fuel for Nuclear Energy," New York Times, 19 Oct 09.
 P. McKenna, "Is the 'Superfuel' Thorium Riskier Than We Thought?" Popular Mechanics, 5 Dec 12.
 H. Hodson, ""India's Thorium-Based Nuclear Dream Inches Closer," New Scientist, 9 Nov 12.