- Reactor containment buildings are not designed to contain a meltdown,
and can rupture in a matter of hours.
- In 2002, the NRC's own inspector general concluded:
"NRC appears to have informally established an unreasonably high burden of requiring absolute proof of a safety problem, versus lack of a reasonable assurance of maintaining public health and safety."
http://thebulletin.org/web-edition/features/the-bulletin-archives-containment-of-reactor-meltdownFrom the Bulletin Archives: Containment of a reactor meltdown
By Frank von Hippel | 16 March 2011
Article Highlights
* A filtered vent system could relieve the pressures inside a dangerously pressurized containment building by releasing some of its radioactive gases through a large filter system.
* The industry is concerned that accident mitigation techniques could be interpreted as tacit admissions that serious accidents can happen.
<snip>
Reactor containment buildings today are not designed to contain a reactor core meltdown accident, however. Their "design basis accident" is a loss-of-coolant accident in which large amounts of volatile radioisotopes are released from a temporarily over-heated core, but in which the uncontrolled release of energy from the core into the containment atmosphere is terminated by a flood of emergency core cooling water before an actual meltdown occurs. This is essentially what happened during the accident at Three Mile Island although, due to various errors, the core remained only partially cooled for a period of hours.
The threat of overpressurization. If for any reason the emergency core cooling system were not effective and a core meltdown occurred, the build-up of internal pressure in a sealed reactor containment building could rupture it within a matter of hours. The threat would come from steam, hydrogen and other gases.
<snip>
The above was originally published in 1982, but was republished because it was referenced in a new article by von Hippel:
http://thebulletin.org/web-edition/features/second-chances-containment-of-reactor-meltdownSecond chances: Containment of a reactor meltdown
By Frank von Hippel | 14 March 2011
Article Highlights
* The hydrogen explosions at the Fukushima Nuclear Power Plants are a startling reminder of the unlearned lessons since similar events at Three Mile Island in 1979.
* The problem of containment over-pressurization and the potential need to vent is not new: It has been a long-term issue in the nuclear reactor safety community.
* The time is now to question what is "safe enough" in the nuclear industry.
<snip>
In 1982, after the accident at Three Mile Island, Jan Beyea and I wrote an article on filtered vents in the Bulletin of the Atomic Scientists in 1982 (page 52). As we reported in that article, the US Nuclear Regulatory Commission (NRC) was negative about the idea. They had a number of arguments but none of them seemed very strong to us. Those interested can read our summary of the arguments pro and con in the 1982 article.
The unspoken argument against requiring that US nuclear power plants be retrofitted with filtered vents was that the industry thought that they were already safe enough and that the expense would be wasteful. And, as today, the commission did not want to force the industry to do more than it was willing to do.
In 2002, the NRC, despite alarming evidence that a pressure vessel had almost corroded through, refused to force an owner to shutdown the reactor for inspection before its regular refueling shutdown. After a review, the NRC's own inspector general concluded:
"NRC appears to have informally established an unreasonably high burden of requiring absolute proof of a safety problem, versus lack of a reasonable assurance of maintaining public health and safety."
We failed after Three Mile Island in 1979 to reform the Nuclear Regulatory Commission or force improved containment designs. The tragedy in Japan may have given us another opportunity.
That article has a bad link (extra space in the url),
the correct link to the NRC inspector general report is
http://www.nrc.gov/reading-rm/doc-collections/insp-gen/2003/02-03s.pdf