Ah, thorium. Thorium will power the future, resulting in endless electricity for all… oh wait, or was that fusion power? Anyway, every man and his dog seems to be pushing to use thorium, instead of uranium or plutonium, with a seemingly endless stream of articles about how great it’s going to be. Apparently, it’s safer, more economic, it will never run out, and you can’t make bombs of out. Sadly, I think this is mostly hype.
There have been experimental reactors fueled by thorium, there will be more, but large scale, commercially viable reactors? In less than ten years? No chance.
Wrong. We have made bombs from thorium. The delightfully named Operation Teapot MET in 1955 was the first, I’ve seen it claimed that there were several more.
The material used in these bombs is uranium-233, a different isotope to the U-235 that’s generally used. U-233 is frighteningly trivial to make from thorium nuclear fuel. It’s also trivial to separate using chemical methods with no need for the huge cascades of centrifuges that the Iranians are having Stuxnet-related problems with. Ease of production and ease of separation make controlling the proliferation of nuclear bombs from thorium into a bit of a bugger. Thorium-based bombs do have one nasty disadvantage that’s stopped super-powers from deploying them – the U-233 decays into materials that kick out really nasty gamma rays. Standard military-grade Cold War nuclear bombs will happily sit on a shelf for decades without a best-before date. Seriously, all US bombs out there are at least twenty years old and they’ve expected to work first time, guaranteed, or your money back. They’re also safe to handle because they put out very little radiation, your biggest worry would be dropping one on your foot.
Thorium-based bombs won’t do that. You wouldn’t want to be standing next to them, you wouldn’t want complicated electronics near them, you’d want to use them fairly quickly, but they are still nuclear fucking weapons. If you’re some tin-pot dictator who’s down on his luck and is convinced he’s got his back to a wall, then these will do in a pinch, where “do” means you get to burn whole cities to the ground.
Personally, I think that’s a bad thing.
The chances of making thorium reactors cheap, safe and reliable – weird materials:
What’s top of the list of thorium reactor designs? The Liquid Fluoride Thorium Reactor.
I used to be a materials scientist. I’ve worked with assorted nasty chemicals, especially when we were trying to stain stainless steel. Boiling mixed acids? Piece of cake. The only stuff that scared us was anything with fluorine in it. So, what coolant do Liquid Fluoride Thorium Reactors plan to use? Liquid fucking fluoride salts. What’s the number one cause of coolant leaks? Corrosion. What’s really fucking corrosive? Liquid fucking fluoride salts.
If you’re working with pipes full of really evil stuff, what metal do you make the pipes out of? If bog-standard stainless steel won’t do the job, you go for fancier stainless steels. If they won’t do the job, you go for one of the Hastelloys. If you’re dumb enough to mess around with hot fluorine salts, then your main choice is Hastelloy N.
Oh, and on top of the corrosion is the radiation flux and the high temperature, which all tend to act synergistically to give you demon-possessed corrosion. And if it leaks, then any contact between the fluoride salds with moisture means you’ve got hydrofluoric acid, which is nasty enough to eat through glass.
Yes, we’ve made a molten-salt reactors before using fluoride salts and yes, the corrosion was reported as negligible. However, negligible for five years isn’t a very high standard. The UK’s Magnox reactors worked just fine for six, before anyone noticed the breakaway corrosion that no-one had predicted. We want these reactors to run without problem for decades, so we need to know that they will not be affected by corrosion, or all the factors that can complicate corrosion, such as erosion-corrosion at bends in pipes, corrosion at welds, corrosion near welds, crevice corrosion, corrosion at junctions with dissimilar materials, corrosion under massive neutron fluxes, and corrosion due to the complicated mixture of elements that are generated by all the nuclear reactions in the fuel.
The MSRE was made from Hastelloy N, a nickel-based alloy which is used for holding molten fluoride salts and pretty much nothing else. Because we don’t use it for much else, we don’t have a great deal of experience using it, nor have we worked out all the pitfalls. For instance, nuclear reactions create tellurium. The MSRE experience discovered that tellurium causes cracking in Hastelloy N. Whoops. What else is going to go wrong? Probably quite a lot, coz that’s just materials corrosion and we haven’t even got on to fatigue yet, let alone all the complicated nuclear stuff. Oh, and there’s only one company in the world that has made Hastelloy N and they’ve replaced it with a different alloy, with even less practical experience resulting in different and unknown pitfalls.
To simplify this argument:
1) A nuclear reactor needs to be very well understood to be affordable and safe.
2) Thorium reactors are made out of weird and nasty materials under weird and nasty conditions. Those materials will do surprising things, certainly expensive, probably dangerous.
We can add (1) and (2) to conclude that you’d have to be really desperate to invest tens of billions of dollars and decades of effort into having a chance of this being viable.
(In this case, it seems like India is sufficiently desperate. They’ve far more thorium ore than uranium and they had a desire to be independent of overseas fuel suppliers. However, their thorium programme was started in decades ago, before uranium was freely traded. Now India has much easier access to global supplies of uranium, my impression is that the thorium research is mostly powered by inertia and national pride. Good luck with that.)