Postby ladajo » Tue Mar 29, 2011 12:46 am
Dan,
You are referencing the SL-1 accident. The unit went prompt critical during a procedure where one of the operators (the navy guy) pulled the control rod out too far too fast. The resulting prompt critical burst flashed over the core, and the resulting steam explosion impaled him to the roof with a component of the vessel head. One operator died from the balst and ejection from the top of the vessel. The other operator sustained mortal injuries in the balst and fall as well, but survived long enough that the investigating team found him. He then died in the ambulance is my understanding.
I have seen the sight in Idaho, another geometric deadspot in the desert.
There was a facility at INEL called PCF, or Prompt Criticality Facility. The used to build scaled down reactor vessels and insert live scaled down cores in them. They would pop the rods, and test the reactor vessels for live accident containment. What a job that must have been.
Again, to level the terminology playing field: Sub-critical = power decreasing, neutron population not self sustainable at current level. Critical = self sustaining neutron population at current power. Super Critical = power increasing, neutron population growing. Prompt Critical = rapid "Holy shoot the chicken" neutron population burst growth which effectively instantly cooks off the fuel mass within the affected geometry. The prompt neutrons, or those produced directly from fission of fuel, have a 1.0 or greater proability of fission, due to extreme population density, and/or due to high density fuel (enriched or a fast fuel). Note that U-235 is considered a slow fuel. But at higher enrichment (density levels) it can cook off with fast neutrons, especially if saturated with them.
In a normally operating power plant, U-235 fissions an avg. of 2.34 neutrons, which in turn pin ball down to slow enough energies to create another fission event. The plant also gains some neutron population from fission products decay, and fission, but the majority come from the fuel itself.
Fast Fuel prefers fast neutrons to fission, slow fuel prefers slow neutrons to fission. A slow fuel can burn via fast neutrons in a saturation event, ie metric butt-ton of fast neutrons in an dense geometry. Essentially, shoot enough hot neutrons at it and it will burn (prompt criticality can do this.)
Fast fuel weapons work better and are easier to pop than slow fuel weapons. The reason the pit accident did not pop was because the was no neutron moderator or reflector to drive the neutron density high enough to get the pit to go prompt. So instead of dying painlessly in a spectacular plasma ball fashion, the dude cooked himself, and went out the painful way. If he had not of had the presence of mind to pull the pit apart, odds are several more folks would have cooked too, and the building would have burnt down, leaving an ugly mess to clean up (if possible).
Edit: I suck at spelling again.