The reprocessing part is much more difficult: in parts the salt is a lot more corrosive due to higher fluorine content and other parts may need higher temperature or exposure to molten metals like bismuth. So I would expect parts of the reprocessing unit to last much shorter, with regular replacement. How and if the reprocessing works is one of the big open points – the reactor is easy and proven to work, reprocessing has several open questions.

@MartinBurkle:
1. Fluoride salts are the fuel of the reactor. The salt tested in the MSRE is FLiBe, a mixture of Lithium Fluoride, Beryllium FLuoride and contaminants such as Uranium 233 Fluoride, the fuel. Besides the temperatures and intense radioactivity, fluorides are chemically poisonous. Beryllium is poisonous. Also, Lithium salts are psychoactive drugs. You don’t want to eat these things. However, they are inside five levels of containment because they are a reactor core. At normal temperatures, they resemble rock: The fluorides are solids that don’t dissolve well in water. As for plant personnel, most sane reactor designs pump the salts through a heat exchanger, to move the heat to a less-poisonous medium that turns a turbine to make power (Kirk favors Helium).

The salts are very safe within a reactor: They are immune to the three main damage mechanisms in a reactor core: phase change (melting or boiling), corrosion, and heat damage. 1a. They have very low vapor pressures, even at high temperatures, so they do not boil, and build up pressure. At high temperatures, the neutrons get above the fission resonance of U233, so the reactor can’t run away. 1b, Molten ionic salts are immune to radiation damage: After a particle or something hits the chemical bond, the ions just bounce around and reform in the fluid. 1c, Since they are chemically stable fluorides, very little or nothing can break the chemical bonds of the salt. I.e. it can’t corrode.

@MartinBurkle,@JasonRibeiro: In air or water, FLiBe will cool to a rock-like substance. Fluorine is more chemically reactive than oxygen, carbon, hydrogen or calcium, so these can’t break the bonds of the salt. Fluoride salts do not chemically interact with air, water or concrete.

@MartinBurkle: Water doesn’t dissolve most fluorides, including heavy metal fluorides. Water might entrain particles of salt. The whole point of a five-level containment is to keep the particles of salt inside.

@MartinBurkle: Concrete contains water. Hot salt will cause it to spall. The salt will have no chemical reaction with it.

: The corrosion problems in the MSRE are all known and fixable. See WASH 1097 and ORNL-4829, referenced above.

@ChrisGentile: The ARE had a specific power of 80Kw/liter of salt.

@Ebenus, @RobMcMillin:
Agree with David; also China does not have the NRC issues, and is building one.
Another issue is that commercial reactor companies currently sell reactors at cost,
and make their profits by selling solid fuel assemblies. Salt maintenance
provides a nearly identical comemrcial opportunity, but this is not well known.

@SteveBurrows: Current large LFTR designs look like large, rectangular industrial buildings. The giant dome of light-water reactors is needed because in an emergency, they have to be able to passively cool steam from the core.

@Ken: Chlorine is a larger, emptier atom than Fluorine. It slows neutrons a lot less. It’s practical to make a fast-neutron reactor using chlorine salts, and therefore a chlorine salt reactor can fission Uranium 238, and the inconvenient even-numbered transuranic elements that make spent nuclear fuel hazardous for thousands of years. That is, it converts these into fission-products, radioactive waste that is dangerous only for about 300 years.

@David: Yes, U233 is a nuclear fuel usable in a properly designed trigger.
However, diverting a critical mass from a reactor will shut it down,
if the reactor is designed for this.
Monitoring should be able to detect a shutdown.
A LFTR is a breeder, but the best theoretical reactors achieve
109% per year and practical reactors can easily be less.
Also, U233 isn’t quite as good as Pu239:
The critical mass is a bit larger, and the pit will emit a lot of dangerous gamma
radiation as it decays. Skilled weapon designers could cope, but if a sovereign
power wants nuclear weapons, they can probably get them anyway.

@BillW: India is focused on conventional solid-fuel reactors. According
to a respondent here (Jagdish), they are making steady progress, and are
reluctant to adopt a technology abandoned by the U.S.
They are actually trying to sell a partially-thorium-powered CANDU-style reactor
on the international market.

@SteveRappolee: 1) Orbital reactors that -start- very radioactive would be very
dangerous during the launch phase (imagine a crash…)
(All orbital reactors so far have had low-radioactive uranium isotopes at launch)
I think this proposal would not pass licensing. I wouldn’t want it to pass, myself.
2) The engineering issue with a nuclear-thermal LFTR is the salt-to-H2
heat-exchanger, not the power density. There are wonderful micro-machined ceramic
heat-exchangers now. The system is more complex than a NERVA.
3) Yes. The Radkowsky fuel concept is designed for LWRs.
4,5) Probably not; The even-numbered actinides need fast neutrons to fission.

@PaulChamberlain: It looks to me like the upper limit is concerned with
engineering costs: Larger units are -much- more expensive to prototype.
The industry has settled on 1-1.5GWe units as being the most economical.
E.g. this is the maximum size of commercial turbines.
LFTRs have lifetimes comparable to LWRs. The issue with LFTRs is the
moderator lifetime. David LeBlanc has a simplified design that may solve this.

@MicheleAlessandrini: Fission products’ radioactivity
decays below the level of natural uranium in about 300 years.
Also, about 60% of fission products have possible value in industry:
The noble metals (Platinum, Palladium and Ruthenium, 20%) are useful as catalysts.
Catalyst cost currently limits the size of fuel refineries.
The noble gases (Xenon and Krypton, 40%) are useful in anesthesiology,
lamps and insulating windows.
The salt-loving fission products (e.g. Cesium, Iodine, 40%) are not valuable and
probably have to be buried.

@Michael: Special radioisotopes? Pu238 is from the reactor-catalyzed radiodecay
chemistry of the Thorium fuel Cycle.

Everyone should be aware that as of 25 Jan 2011 the Chinese Academy of Science (CAS) announced a development
program for Thorium-Fuel Molten Salt Reactors (TFMSRs). Reps from the CAS visited Oak Ridge Labs last Fall (2010) to make a reality check, and have now decided to eat our collective lunch by going after the IP and patent rights to molten salt reactors. This is a true “Sputnik Moment” for U.S. energy development.

The rest of the world can go their merry way, boiling water, risking explosions, and straining to create reactor designs using solid-fuel uranium or thorium. Flibe Energy [...] http://flibe-energy.com [...] will create a better way to “burn” all the HEU, spent fuel rods, Pu239, and 99% of the TFMSR liquid fuel, while reducing the nuclear waste storage/disposal problems by a
factor of 1,000, and max storage time to 300 years. Think Boeing-style U.S. factories producing small (100MWe), modular, standardized TFMSRs for clean nuclear energy.

So far the biggest Chinese energy investment is in 200 Westinghouse AP-1000 solid-fuel reactors to be built over the next ten years. Construction of each pre-fab AP-1000 will go from groundbreaking to on-line power production in 3 years (vs. NEVER for the US). CAS estimates 10 years to take the TFMSR from prototype to commercial reality.

My estimate is 5 years for China to have a Boeing-style, pre-fab factory for TFMSR’s producing one 100MWe generating station each week (Holton’s generating capacity is 20MWe). Expect the AP-1000′s to be cancelled in favor of TFMSR’s which take at least 50% less money and land area to build, and they can be air-cooled or used for industrial heat.

China has stockpiled enough thorium from their rare-earth mining operatons to keep ALL those TFMSR’s running for
centuries. Meantime the US considers thorium a toxic nuclear waste (which keeps US rare-earth mines closed, and DoD at China’s mercy) and has buried 3,200 tons of thorium “waste” in the Nevada desert. Plans are in place to spend $500 million to destroy the thorium stash in Nevada. Better to sell the 3,200 tons to China, OR let Flibe Energy jump-start TFMSR production in the US. Think jobs!