Nuclear Info Kept From Congress and the Public

Thermite writes "On March 6, 2006 an accident occurred at Nuclear Fuel Services in Erwin, Tennessee. According to reports, almost 9 gallons of highly enriched uranium in solution spilled and nearly went into a chain reaction. Before the accident in 2004, the NRC and The Office of Naval Reactors had changed the terms of the company's license so that any correspondence with Nuclear Fuel Services would be marked 'official use only.' From the article: 'While reviewing the commission's public Web page in 2004, the Department of Energy's Office of Naval Reactors found what it considered protected information about Nuclear Fuel Service's work for the Navy. The commission responded by sealing every document related to Nuclear Fuel Services and BWX Technologies in Lynchburg, Va., the only two companies licensed by the agency to manufacture, possess and store highly enriched uranium.' The result was that the public and Congress were both left in the dark for 13 months regarding this accident and other issues at the facility."

Comments

Second most serious violation

[dontthink]

FTA:

McIntyre defended the commission's decision not to fine Nuclear Fuel Services, even though the agency rated the uranium leak last year as its second most-serious violation.

(Emphasis mine) Personally, I would be interested to know what the most serious violation was...

That's the Risk of Privatization

[mpapet]

In the mad rush to privatize government, the broader issue of a serious lack of oversight will become quite common.

http://www.vanityfair.com/politics/features/2007/0 6/murphy200706

Fissionables in solution are weird.

[Tackhead]

> A lot of nuclear materials can under-go a chain reaction if a significant mass is accumulated. It has to do with production versus escape of neutrons and scales as volume-to-area. So, if two sub-critical masses were combined, they could become critical. I am somewhat leary of a "spill" making something go critical, unless the mass was over-critical and the container provided some damping effect.

Actually, the "spill" makes it more likely, not less likely.

Fissionables in solution are tricky things to deal with. Consider the following four cases:

1) Homer Simpson drops a subcritical hunk of a water-soluble U235 salt into a swimming pool. No big deal. It's a single subcritical mass of U235, and the neutrons fly straight out of it and into the surrounding water, and not enough bounce back into the mass to present a problem. Homer reaches in with a net, and pulls the chunk of salt out of the net. "No problemo."

2) A little while later, as the harmless chunk dissolves into the huge pool, there will be localized spots near the chunk, with both sufficiently-high concentration of fissionable materials and the right amount of moderating material between them for a criticality incident. "D'OH!"

3) "Aha! I'm smart! I'll prevent that scenario by dissolving it, a bit at a time, by adding it to the pool by using a salt shaker near the pump intake!" Congrats! The U235 atoms are, at all times, sufficiently widely-dispersed, that there is no criticality risk. "Woohoo!"

4) A few weeks after your swim, the place is shut down and everyone gets fired. The maintenance guy forgets to drain the pool. The water gradually evaporates, and concentrations in the remaining water begin to rise... and a few years later, some guy spraying graffiti by the abandoned poolhouse wonders WTF that blue flash was. "D'OH!" again.

I'm on a roll here, so I may as well close off the "security by obscurity" issue. There are places where security by obscurity works, and this is one of them.

The deal here is that criticality incidents, especially involving fissionables in solution are a function of degree of enrichment (in the case of uranium as the solute), nuclear properties of the solvent, local concentrations of the ions in solution, and a whole boatload of other things, in order to build cool toys, you often have to deal with them all, simultaneously. I'm not in the building-of-cool-toys industry, and have mercifully I've never had a need to know.

Some of these things are public domain, but others (particularly those things pertaining to the design of shipborne Naval reactors, which use HEU because there simply isn't enough space on all types of ships to permit the use of LEU-based designs) are classified. Given a description of an incident, however, it may be possible to place upper and lower bounds on some of the classified parameters - bounds that are narrower than the published numbers, and there are plenty of adversaries who would be delighted to deduce things about our Naval capabilities (a lot more interesting/useful than even our bomb designs), given just a few more missing puzzle pieces. The math is hard, and denying adversaries the pieces of the puzzle that they can use to derive the whole picture isn't security by obscurity, it's just good security practice.