Imagine a bathtub full of water, all the way to the edge of the tub. You may have the water temperature just the way you want it, but the water is cooling. Eventually you want to add hot water to the tub, but you can’t do so without it spilling over the edge. So you must withdraw water from the tub before you can add more hot water, and the hot water you add must be quite a bit hotter than the rest of the water in the tub in order to increase its overall temperature. Thus you’re throwing out water from the tub that is still fairly warm so that you can replace it with water that is a lot hotter.

This is analogous to what has to happen in a uranium-fueled MSR. The fuel is like the water in the tub. The temperature of the fuel is like the enrichment of the uranium. Operation of the reactor burns up the uranium-235 in the fuel, reducing its enrichment, just as the water in the tub is cooling. You must increase the enrichment of the fuel in order to keep the reactor operating, but to do so you cannot just add more uranium fuel at a higher-than-average enrichment. You have to take some of the fuel out of the reactor. But that fuel still has a fairly significant enrichment in it. You’re going to lose the potential to “burn-out” the uranium-235 from that fuel. And as soon as you add fuel with higher-than-average enrichment, it mixes with the existing fuel and increases the average enrichment, but it’s still lower than what it came into the reactor as.

This is a situation where a basic advantage of the molten-salt reactor (continuous mixing) turns into a big disadvantage. In a solid-fueled reactor, you can add fuel assemblies with higher-than-average enrichment and move around assemblies with lower-than-average enrichment so as to achieve an overall average enrichment. That way, when it’s time to pull a fuel assembly out of the reactor, you will have been able to “burn-out” most of the uranium-235. Indeed, this is what is done in PWRs and CANDU reactors. But it’s not possible in a molten-salt reactor, just like you can’t heat up the bathtub without taking some water out first.

Now what about LFTRs and the thorium fuel cycle, do they have this problem? No, because the fuel they’re adding to the core is of a different chemical species (uranium) than the fertile material (thorium). In a uranium-235-fueled MSR, the fuel and the fertile material are both uranium. Making power in a LFTR involves consuming uranium-233 fuel, and it is replaced at the same rate it is consumed. In a uranium-235-fueled MSR, uranium-235 is being consumed much much faster than uranium-238 is being turned into plutonium, thus one must add new uranium-235. But it’s not that simple, even if one was going to the theoretical extreme of adding pure uranium-235 to the reactor, that would involve an enormous loss of the work of separation that had taken place to create that uranium-235 in the first place.

This is why we are convinced that attempting to design and build a uranium-235-fueled MSR will lead to even poorer outcomes, from a fuel performance perspective, than even today’s wasteful uranium-fueled PWRs and CANDUs. At least they can “burn-out” the U-235 content in their fuels. Uranium-235-fueled MSRs will not be able to do that efficiently.

Molten-salt reactors need to implement the thorium fuel cycle. The thorium fuel cycle needs to be implemented in molten-salt reactors. Both pieces require the other. Attempting to separate them from one another leads to bad designs and poor performance.

The implications of this realization are rather significant. It strongly implies that the first stage of nuclear development had to have been implemented in solid-fueled reactors. There really was no option in the post-World-War-II era to proceed immediately to molten-salt reactors, even if the technology had been available, because the molten-salt reactors would have needed either plutonium or uranium-233 that was not available.

Nature fixed the first stage of nuclear development by virtue of the fact that uranium-235 was the only naturally-occurring fissile material on Earth, and the fact that it was such a small fraction of the overall uranium resources.

I must admit that this realization really altered my entire perspective on the early era of nuclear development. Realizing that molten-salt reactors just really weren’t an option at the outset for nations that wanted nuclear energy was a bit sobering, and certainly toned down some of the frustration that I have felt for many years about how things moved forward. The United States, the Soviet Union, Britain, France, and China all built their first generation of nuclear reactors to produce plutonium for nuclear weapons, but even if this had not been their objective, they still would have had to build nuclear reactors based on uranium that would have produced plutonium. It is simply an inevitability of the fact that uranium-235 is the only fissile material one can use at the beginning.

At this stage of nuclear development, we have been operating at this “first stage”, consuming uranium-235 in the presence of uranium-238, for many decades, and have accumulated substantial inventories of plutonium, as well as smaller inventories of even-heavier transuranics such as americium and curium. There is no further need to make any more transuranics, even if a molten-salt reactor could effectively undertake that task. But in light of the strong evidence that a molten-salt reactor would be marginal in that mission, and may even be worse than solid-fueled reactors from a fuel consumption perspective, it really is clear that molten-salt reactors should be directed towards the second and third stages of development.