There are two distinct uses of molten salt in nuclear technology. In one, it is used as a coolant for a reactor with solid fuel cores. In the other, like LFTR, the fuel is dissolved in the molten salt coolant.

The big appeals of dissolving the fuel in molten salt are that, first, they can operate at near atmospheric pressure, so they can be built with less expensive and much easier to reproduce weldments, second, they operate at high temperature, for better thermodynamic efficiency of generated power, third, fission products tend to combine chemically with the molten salt, so the risk of a atmospheric release is much lower and fourth, molten salts are less reactive at higher temperatures, so an appropriate mix will intrinsically dampen reactions before reaching its boiling point.

These points combined support a passively safe reactor design ~ a reactor design that does not have explosive release of radioactive products or threat of meltdown as outcomes of failure of the components of the reactor. There is still a threat of leaks, but the big Fukashima / Chernobyl risks are not present.

When molten salt reactors have the fuel mixed in, they require constant processing inline ~ that is the process that is dramatically different from light water reactor technology. The constant inline processing gives the bias to thorium fuel cycles for this type of design, because the creation of plutonium is six reaction stages away from the original thorium, so thorium fuel mixes in operation will have a far, far lower concentration of plutonium than uranium versions of a molten salt reactor.

It is in the processing stage that one difference in approach appears for thorium fuel cycles. The thorium fuel cycle include protactinium-233 which is a neutron absorber. It can be removed and allowed to decay into U233 for bomb making material. Or the quantity of thorium in the breeding mix can be increased, to reduce the concentration of of Pa232 to the point where the processing takes out the U233 directly.

Its the Pa233 source of the U233 that is responsible for one anti-proliferation property ~ some of that Pa233 decays instead into U232, which has a short half life and hard gamma emitters in its decay chain, and so difficult to mask from detectors. But U233 and U232 are not possible to separate chemically, so separating out the U232, either to make something easier to transport or for the bomb making itself, is difficult.

The accelerator approach is a different approach to the continuous processing approach. There are two distinct thorium fuel cycles. The Thorium-Plutonium cycle is not focused on power production but rather on the production of pure U233, and is to proposed operate alongside existing light water reactors to process the plutonium they produce on site.

The U233 is used in a U233-Thorium breeder reactor.

The Wikipedia machine says that the accelerator is a medical grade accelerator, so it would not seem to be the main issue.

I'm not sure I understand why this approach eliminates the requirement for inline processing, but it may be due to allowing a lower concentration of fissile products without quenching the reaction, so that the decay products dissolved in the salt transition down their decay chains before reaching concentrations that interfere with the performance of the reactor.


I've been accused of being a Marxist, yet while Harpo's my favourite, it's Groucho I'm always quoting. Odd, that.