George and I were swapping emails about liquid fuel in Molten Salt Reactors. He suggested I post this, so here goes:
As I see it, liquid fuel is the key, and not an indirect safety feature.
When solid fuel heats up, it comes together (corium) and gets even hotter. Corium quickly reaches 2800°C, hot enough to eventually melt through any reactor vessel, unless it is cooled by emergency measures.
But liquid fuel will never get that hot, because it expands and cools, or melts the freeze plug and drains and cools, or expands and bursts a pipe or a seam and spills out and cools. Any of these events would prevent the fuel from ever getting hot enough to melt the reactor. It would drain out, before the reactor vessel got anywhere close to melting.
The only way I can see liquid fuel melting an MSR is if the designers built a burst-proof, drain-proof reactor vessel without a freeze plug. And there would be no rational reason for doing so. On the contrary, MSRs would be designed to burst and drain in case of damage or sabotage. Not doing so would be like building a car without any brakes. It just wouldn’t make sense.
The problem with solid fuel is that it can’t get away from itself and cool off. That’s the fatal flaw in every solid fuel design, and that’s what all the safety features and cooling systems and redundancy is there to contend with.
Liquid fuel can get away from itself, and cool off naturally. That’s the beauty and genius of liquid fuel. In my opinion, Weinberg, Wigner, and their colleagues should have been awarded the Nobel Prize for this alone. It’s that big of a deal. It changes everything.
Because of this inherent safety feature (heat = expansion = cooling), I would contend that the liquidity of the fuel is a primary, and not a secondary, safety feature.
It’s the non-killer app of MSR.
A more important question is whether the molten salt containing about half the elements can eat its way through the reactor wall. Or maybe a more exact question is how can the metal be examined while the reactor is running?
It’s not “about half the elements”, Wikipedia has the specific isotopes of each element in the uranium fission decay chain. We know exactly how much U235, Pu239, U233 fission generates, to make 1 gigawatt-year electricity.
Salts are extremely chemically stable, and the fuel and most fission products chemically bind to the salt. The reactor wall material would be selected to minimally corrode with the specific salt used (FLiBe salt is only one of several good options, of fluoride salts or chloride salts).
All we need is one reactor material that corrodes only minimally during the reactor lifetime, or lasts long enough to for a simple replacement procedure. We have had materials tested and certified since the 1960s. We have new materials that should last longer, still to be certified. Materials testing would include radiation damage and chemical corrosion. Then each component would be thoroughly inspected before being installed.
Since the molten fuel salt is transparent, internal camera inspection can be done continually. Aircraft are frequently inspected with equipment that can detect early stress fractures in metal; similar equipment could inspect reactor walls from outside. Since there isn’t high pressure, sensors can be installed throughout the reactor (LWR reactor vessels have to be extremely carefully cast to withstand the pressure).
Remember also, since there is no high-pressure water like LWR has, there is no chance of pressure explosions. There is no way for propelling radioactive material into the environment. If someone fired a rifle into the reactor wall, the salt would puddle on the floor and cool to solid, chemically trapping most of the radioactive material. Drain the fuel to storage tanks, fix the hole, start the reactor up again.