As someone that often works for multiple years on pilot and poc projects, can we stop calling those “toys”.

Sorry we don’t have madscientist money here.

That reactor is 2MWt, which is still somewhere about 1000x smaller than actual production reactors. But this is not the issue here, because in MSR the reactor is not the hard part, it’s its entire fuel cycle.

The entire point of having fuel as a solution instead of hard, nonreactive ceramic pellets put in tubes made of refractory metal is that there could be perhaps a way to extract fission products from coolant/fuel, which would prevent neutron capture by these fission products, which makes in turn better use of neutrons, so more fissile material can be bred. Benefit of this is that if that online recycling process can be made to work (big if - unsolved for now) then reactor works always like it’s been freshly refueled. The hard part here is not reactor, it’s the cleaning of fuel while reactor is still online. This has not been demonstrated, instead only new fuel was added, which is something that can be done with CANDU and some other designs where reactor is divided into channels

First attempts at something like this used heavy water acidified solution of uranium nitrate, but this proved too corrosive and also water needed to be pressurized, and also it decomposes when subjected to radiation in this way. Today what is used is FLiBe, which is low-melting salt that doesn’t decompose in this manner, but also is more corrosive and in different ways than water as used in PWRs. If that was the only problem, we would have MSRs left and right, but there are three other big problems

Recovery of excess bred 233U or removal of neutron-absorbing fission products from FLiBe is hard, because you can’t use normal methods used in nuclear reprocessing. There’s no extraction like in PUREX, there’s no ion exchange resin that can survive it, there’s only fluoride volatility and some electrochemical methods, and it all would require significant research before anything close to viable comes up. The salt also probably has to be kept anhydrous at all times. This is the first problem. Maybe this reactor will be used for it, maybe it’ll fail, but there’s a related Problem that doesn’t appear in more conventional reactors. In normal case, you can just leave fuel elements in water until the spiciest isotopes decay so that you don’t have to deal with them. Here, we intentionally work with freshly irradiated, so ridiculously spicy fuel, and intentionally concentrate the most radiotoxic isotopes that are out there. Worse than that, all these fission products are not in form of chemically inert ceramic, these are in form of water soluble fluoride salts and this means that if anything of this gets into soil, it’ll dissolve meaning that either fuel leak or waste stream leak would have much more severe consequences than if it was in conventional form. If you’re trying to say that MSRs are safer for some reason, i’d have some serious reservations.

The other problem is that FLiBe is a good moderator, meaning that any MSR reactor design using this salt is thermal reactor, and we already have this figured out in form of PWRs where we can use water instead. Look up India’s plans for thorium power - they want to use PWR reactors for breeding 233U, with heavy water or not, because this already works and there’s no actual reason for use of this highly experimental and uncertain technology. Keeping fuel rods in reactor for longer time is not an actual showstopper like it was expected in 60s when this concept first surfaced, in fact with advancement of nuclear technology burnup only goes up, i think it already is 2x or 3x what it used to be in early commercial power reactors. If MSR was the only way to make breeding work, we’d probably take effort to manage ridiculous radiotoxicity of this fuel mix, but because both chemical engineering to do so is not there and alternatives that don’t have this problems exist, we don’t. Charitably i’d could describe MSR fuel cycle idea as an highly experimental but promising while also requiring significant research expense. Less charitably, looking at all those years of research yielding nothing, i could also describe it as a dead end grift. You decide

Note that all these problems come up with use of MSR, not thorium. Thorium for nuclear power is fine, but requires reprocessing, and some countries don’t want to do this for diplomatic reasons (americans specifically) (tho i suspect it’s masking the actual reason: some bean counter at westinghouse calculated it’s cheaper to use fresh uranium instead - reprocessing is a lot of dangerous, well-paid, complicated work - in countries where labour costs are lower, or where govt is willing to pay up to have reserve of nuclear material, which amounts to all other countries that have sufficiently advanced nuclear industry, reprocessing does happen. french, chinese, russians, indians, japanese, koreans, and probably a couple more do reprocess their fuel. there’s a couple of countries that send their fuel to manufacturer, and some just discard it underground without reprocessing) (this is also why yucca mountain filling up is a problem of entirely american making, and the only thing that is lacking in order to solve it is political will)

The article makes it very clear its running continuously, which is what they are celebrating

i think you’ve read different article

Chinese scientists have achieved a milestone in clean energy technology by successfully adding fresh fuel to an operational thorium molten salt reactor, according to state media reports.