Best Values of W/e

[Chris Alston <alstonc@odrge.odr.georgetown.edu>]

>Last week Doug Simpkin asked about W/e values for dry air, and people
uniformly
>recommended the 33.97+-0.06 J/C value of Boutilon and Perroche-Roux.
>This is still the recommended international value.
>
>However a few years back I re-analyzed the data that Boutillon
>and Peroche-Roux had used and found that using more up-to-date values
>of parameters such as half-lives and more accurate estimates of
>the uncertainties on graphite to air stoppingr- power ratios led
>to a value of 33.82+-0.13 which is 3 sigma lower than the earlier
>value (but still pretty much 34!).  However, for standards labs
>work this is a big change.
>
>I never published the work outside internal reports because I still
>don't believe the analysis technique that both BPR and I used.  In the
>covariance analysis, it turns out that ignoring just on of the 13 or
>so initial data values leads to a substantial change in the final value.
>What makes no sense to me is that I removed one of the values well
>below the average value and the average decreased!  The math is correct
>and some people swear it all makes sense - but I am not sure it isn't all
>nonesense!
>
>The value of W/e is central in air-kerma primary standards but if we
>just switch over to using absorbed dose standards, then the problem goes
>away.  Just one more reason for using TG-51.
>
>Ciao
>
>Dave Rogers
>
>
>
>====================================================================
>D.W.O. Rogers                           phone: 613 993-2715
>IRS/INMS Bldg M-35                      fax:   613 952-9865
>National Research Council of Canada     e-mail dave@irs.phy.nrc.ca
>Montreal Rd, Ottawa K1A 0R6
>http://www.irs.inms.nrc.ca/inms/irs/irs.html

************************************************************************
The RADSAFE Frequently Asked Questions list, archives and subscription
information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html