Robert Maxwell wrote:
How do you propose handling the corrosion problems of thorium-cycle molten salt reactors? Isn't that the main thing that makes them impractical for long-term use?
It's not a horribly difficult problem, and one that has some pretty good solutions.
Here's a 2010 report from Idaho National Labs on the issue, which is covered on pages 5 to 10. (Highly recommended, easy to read with lots of graphs)
It looks like chromium content in a molten salt with graphite causes most problem with favored alloys, with a chromium carbide intermediary and chromium plating out on the graphite. Nickel is largely unaffected, and cobalt and molybdenum should likewise vastly improve things.
Some of the higher chromium alloys had a corrosion rate rangine up to 1 mm/year with graphite, but Hastelloy N was 0.045 mm/year. Incoloy 800H without graphite only had a corrosion rate of 0.0033 mm/year (300 years/mm).
The report also says at a nickel coating stops the corrosion. They tried spray on moly and diamond coatings but they spalled, which should be a simple surface bonding/structural issue that could be fixed (getting Teflon to stick to a frying pan wasn't easy, either). Silicon carbide coatings also seem to eliminate the issue.
On approach I'd at least take a look at is using sacrificial anodes, perhaps aluminum or even zinc, depending on which element offers protection and is easily removed and reprocessed from the fluid. And of course since molten salts are so conductive, they could try putting an electric charge between the vessel walls and the graphite.
I've also read that some of the reaction products (like gold), plate out, making recovery for reprocessing difficult. But that could also be an advantage, if the reactor is coating its plumbing instead of corroding it (like hard water protecting copper pipes with a thin layer of scale).
So, as is true in most applications, you can pick the wrong alloy and run into corrosion trouble, or switch to an alloy that largely avoids the problem, while developing better alloys and coatings based on further experience and experiments.
If we just gave up on simple metallurgy and coatings problems, jet engine turbine temperatures would still be stuck at early 1940's levels.