> The article does not talk at all about how, or whether, the type of reactor it discusses would prevent a meltdown of the second kind, the kind that happened a

> The article does not talk at all about how, or whether, the type of reactor it discusses would prevent a meltdown of the second kind, the kind that happened at Fukushima. It's immune to meltdown. The first important safety feature is the fact that it simply doesn't melt. The fissile material is contained in a ball of graphite (which is theorized to sublimate directly to carbon vapor somewhere north of 7000F) and silicon carbide. (decomposes to silicon and carbon somewhere north of 5000F) They tested the fuel at 3200F and none of it failed. The zirconium cladding on the fuel at Chernobyl, Three Mile Island, Fukushima, etc all failed somewhere in the ballpark of 2000F. The second thing is called Doppler broadening; once the uranium fissile material gets that hot, it ceases to be able to sustain a nuclear reaction. The uranium nuclei are zipping around too quickly for thermal neutrons to effectively induce fission. Normal zirconium cladding fails before this effect can become significant. The net result of these two effects is that you don't need coolant. As the temperature increases, the rate of decay decreases to the point where the device will passively cool itself without coolant. This equilibrium point is really stinkin' hot, but the entire facility will have been designed around this fact. The reactor design itself is fairly interesting in its uninterestingness: the only interesting dial an operator has is the coolant pump. If you pump more coolant in, you get more power out; that's it. Turning the reactor "off" just means turning the coolant pumps off, which in most other types of reactors is the thing that causes the disaster. There aren't control rods, I don't think it even has a SCRAM facility.