From: John De Armond
Subject: Breeders (was Re: FAQ)
Date: Mon, 19 Apr 93 15:10:14 GMT
jmc@SAIL.Stanford.EDU (John McCarthy) writes:
>A qualification may be required for fast breeders, although the possible
>explosions of some possible designs still don't (I'm told) correspond
>to bombs. As for nuclear power, I only posted, because it seemed
>necessary, and it would be better if someone professionally concerned
>with it drafted the FAQ answer.
Yackadamn's vivid imagination notwithstanding, there is no concern that
a fast breeder could explode. The design-basis accident for a breeder
is called a "hypothetical core disassembly accident" or HCDA. In this
scenario, a prompt criticality transient occurs. Enough energy is
released to melt the core and (if present) boil the coolant. There
would not be a nuclear explosion for the same reasons real nuclear
explosions are so difficult to initiate. Many factors such as neutron
leakage, thermal doppler shift and the physical disruption of the core
geometry mitigate against a sustained reaction. Post-transient, there
are two major concerns. One, that the melted fuel could puddle and
achieve criticality again. This is prevented by designing the reactor
vessel so that the molten fuel flows into several non-critical sumps.
The second concern is the escape of plutonium vapor. For safety
analysis purposes It is assume that a significant portion of the Pu in
the fuel is vaporized and is available first as a gas and then as an
aerosol of either solid Pu or Pu oxide particles. The typical LWR
containment leak rate of 0.05 percent volume per day would be inadequate
to contain this mist to a degree necessary for public protection.
Most designs therefore propose multiple containments.
A good general reference is "Nuclear Reactor Engineering" by Glasstone and
Sesonkse. The bibliography points toward more specific references.