Tuesday, February 15, 2005

Normalized SECURITY statistics, courtesy NNSA

From Anonymous:

To get an idea of security violations (Level 1 and Level 2, the most serious and moderately serious incidents affecting national security) over the 3 years (2002-2004) for LANL, LLNL, and SNL (ABQ and CA combined), I took the total incidents at each site, divided by 3, and multiplied by 200,000 person-hours over the total person-hours at the three sites (from the DOE safety CAIRS website for 2003). This gives a number that is comparable to the TRC safety numbers (~2), but now for security.


Let me make sure I understand this. LANL has the _lowest_ normalized number of security incidents? And, low is good, and high is bad? And, LANL was shut down for bad security? And LLNL was not, nor SNL? And, these numbers came from NNSA? And, NNSA "fully suports" Nanos' decision to have shut LANL down?

Have I got the facts roughtly straight here?
I'm sure the support of Nanos has nothing to do with the fat pension fund that NNSA assures us they have no intention of raiding...

Or the Navy buddy Nanos has at NNSA in Linton Brooks...
Whatever you do, don't forget our OTHER Admiral, Bob Foley, Vice President of the University of California, in charge of the national laboratories (LANL, LLNL, and LBNL). We got Navy oozing out our portholes, swabby! ...er, Labbie!

Low is, indeed, good; and high is, indeed, bad:

LLNL is high; SNL is medium; and LANL is low:
LLNL Bad; SNL OK; and LANL Good.

Simple, even in Navy terms...I hope that Northrup-Grumman is watching...8 bells and all's well.
Well, just one more itty-bitty thing, while we're at it.

LLNL and SNL got shut down for the CREM investigation, as ordered by DOE/NNSA.

For one or two weeks, as I recall.

Not three months, like LANL.

(Just to keep the pecking order in order.

Seven months, and counting.

But who's counting?
The Level 3 resumption schedule for DX Hydros looks to be in a constant-time-to-completion regime.
Thank you for finding and publishing the security violation measure.

A zero-order correction would be to take account of the fact that LANL has done (almost) no work that could result in security violations for about 6 months. The 36-month average TRC Security Number of 1.3 for LANL would then become
1.3 x 36/30 = 1.56 = 1.6
for the 30 months of actual work. This is a less-than-significant change given the scale of the other numbers.

The other issue to be considered might be the annual rate and whether it exhibits a trend. Of course, the smaller statistics lend a greater chance of apparent variation. In any case it would be interesting to see the apparent trend lines.
LANL may have fewer (normalized)problems but some managers and staff have a knack for making mountains out of molehills. Credit SNL and LLNL for keeping page 30 stories from getting onto page 1, while LANL personnel do the opposite by the way they are handled, such as telling the LA Times that the problems are very serious before they even know the details, or by behaving as if they are in denial.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?