@Andydaws
"A thorium-u233 makes roughly the same quantities of waste as a "classic" u 238-Pu239/240 cycle - thorium protagonists have a nasty habit of failing to compare like with like - they compare a "once through" cycle for uranium fuelled plant with a recycling-based approach for the thorium plant."
What *kind* of molten salt reactor are you talking about?
Molten salt fueled/cooled or molten salt moderated? AFAIK the Gen IV "Molten Salt" proposal just uses a molten salt as a coolant with all the usual paraphernalia of a conventional reactor.
I'll declare my hand - I think the Thorium-U233 cycle has potential
I'm glad you made that clear. It's pretty well hidden otherwise.
Let's be clear - making a thorium cycle work is hard, even if you use a throrium cycle in "conventional" (and that's stretching the normal use of the word) fast reactors.
True. The MSRE did *not* incorporate breeding of Thorium.
"More usually, the enthusiasts go further - they argue for a cycle based on thermal breeding in a molten-salt system."
Not to do so eliminates some of the *key* advantages of a molten salt reactor (IE one where the salt is both coolant and *fuel*).
"The neutron economy of such a system is utterly marginal - a 1% variation in the ability to extract fission products from the salt makes the difference between it producing surplus fuel, and needing continual top-ups."
The key one of which seems to be Xe135. The ability to purge the reactor of this was *the* key reason for designing a molten salt reactor (as part of the nuclear powered bomber programme) in the first place. And AFAIK the MSRE did *not* require top ups of Uranium. However as it did not including the breeding function this merely says a non breeder can be designed well enough *not* to need a top up.
"Worse, making a molten salt system works requires, a large scale and complex chemical processing plant to be added on to a reactor. A Molten Salt Breeder Reactor (MSBR) won't work if you let just a few percent of the siffion products (like Xenon) stay in the fuel - you need 95% plus efficient extraction on every circuit of the fuel."
I looked for "siffion" products but found nothing so I'll assume you mean fission products.
MSRE indicated both Xe and Kr poisons could be efficiently stripped by spraying the salt in a small chamber with an atmosphere of Helium. Both came *readily* out of the salt mix for later absorption onto carbon bed filters to be retained till their decay products could be released into the atmosphere.
" Worse, you HAVE to get out 90% or more of the intermediate between thorium and uranium on every cycle - protactinium. And to get that out will involve delights like passing 800C flouride-uranium salts through a column of molten bismuth, then extracting the protactinium frojm the bismuth somehow."
AFAIK already worked out. The description I'm aware of was a 1GW station would need A 4m (14 feet) high column. my experience of the chemical industry is that a 14 foot high column at probably a few atm (I'd expect it to operate near the same pressure as the core) is not *that* big a deal *despite* the temperature, which I would expect to be *lower* than the core in any case. This hardly compare to reactors for Ammonia or Nitric acid, typically running at 300-400atm, c300c and maybe 8-12 feet wide and 50-100feet high.
"Here's one for the enthusiasts - once the uranium is "bred" in the fuel/salt mixture (and leaving aside the delights of managing two such circuits, one for the fuel, and one for the breeder blanket), you have to get it out."
This *is* a reasonable concern and something which has not been demonstrated.Its complexity is the reason the MSRE did not include the breeding process. *However* the separation process is *well* understood using a combination of adding fluorine and distillation, *provided* you do have 2 separate circuits. The unavailability of a material with sufficient hot strength and chemical and radiation resistance at the preferred temperature was what stopped the MSRE testing this. Metallurgy has improved a bit in 40 years.
" And to do that, you have to bubble fluorine - that well known non-reactive and benign gas - through that same 800C molten uranium salt, then capture the resulting uranium hexaflouride. "Hex" is not only "hot" both thermally, and radiologically, but it's venomously corrosive - the separation membranes in enrichment plants have to be made of pure (99.9% plus) nickel to withstand it, and even then last only a few years."
As you point out this is *known* technology from the Uranium enrichment industry. Agressive certainly but *well* within the state of the art. Note also the "freeze valves" developed for MSRE would allow parallel processing columns (if needed) to be isolated for maintenance and replacement
And yes, it has virtues - thorium abundance, and potentially, it can be "drained down" in an accident. Bit that still means you have to remove decay heat from a couple of thousand tonnes of fuel mixture (more than in a conventional reactor), and have secure cooling and storage for the chemical plant and fission product inventory.
I'd doubt that. The MSRE salt mix density was 2300 Kg/m^3. Given that's lower than Aluminum and the *entire* volume of the MSRE (with no space taken up by the actual graphite moderator) would come to about 5500 Kg, with a maximum thermal output of 8MW I'd say 100x bigger would give c550 tonnes, so a 1GW (common size of power plant) reactor would have to be *very* badly designed to need more than a 1000 tonnes of salt .
As I said, it might have potential - but compared to something like a lead-cooled fast reactor, which have already been built in considerable numbers - the Soviets used them to power the "Alfa" class subs - doesn it look like an obvious route? Hardly...
You are aware that the design you're describing actually uses a lead/*bismuth* alloy?
It's major features being an ability to operate with natural convection in a "stealth" mode which is handy on a naval submarine. and will fail "badly" if the coolant freezes in the tubes.It's got *all* the issues conventional reactors have with fuel element design/certifcation and Xe135 and Kr reactor poisons, *without* the ability to irradiate decay products to *much* shorter lived elements, possibly *the* key benefit of this design if you want to have a nuclear fuel *cycle* instead of the burn/store arrangements most countries seem to have at present.
On the subject of the Alpha reactors for re-processing didn't the USSR just *dump* the cores at the end of life?
I'm not arguing, merely commenting that the phrase "Molten salt reactor" if used loosely has more than 1 meaning.
In case you haven't seen it this is the description of the work written by one of the team shortly after the MSRE. worth reading for the good, the bad and the could have been better.
www.energyfromthorium.com/pdf/NAT_MSREexperience.pdf