Re: Source Expiration Dates-Statistics

[Michael Kay <makay@teleport.com>]

One reason an expiration date is placed on a source (even if physical
integrity and ingrowth of daughters is not an issue) is that the calibration
has a specified uncertainty associated with it. The number you use as the
calibration number is the center of a confidence interval. The percentage
uncertainty is usually quite low when the source is calibrated. For sources
with a very long half life, decay during a lifetime of use is generally
negligible. But for sources where decay may be significant, this percentage
uncertainty will increase over time. This is simply because (Square root of
number of counts per unit time)/(number of counts per unit time)  the
fractional uncertainty due to Poisson statistics, increases as the number of
counts per unit time decreases. This number needs to be adjusted by the
original uncertainty to give the overall uncertainty at the time the source
is used as a calibration source.

This may seem pedantic for the calibration source, but the overall error in
a measurement where another radioactive source is compared to the
calibration source includes the terms for counting uncertainty, and the
uncertainty in the calibration of the calibration source, and the time the
calibration source has decayed (if it is appreciable)--as well as the
counting uncertainty in the material being measured.

Michael A. Kay, ScD, CHMM
makay@teleport.com

Alan Jackson wrote:

> Ben and RADSAFE,
>
> Be careful with "expired" sources.  If you want to use them past the
> expiration date you will need to document that the source integrity is
> unaffected and that the source functions properly.  Regulators have
> been known to issue violations for using expired sources without this
> documentation.  Many counting laboratories avoid this by using only
> unexpired sources.
>
> Some sources can be degraded over time.  Important lines may be lost
> due to decay, ingrowth may occur or physical integrity may become
> compromised.  The cynic in me believes that another reason for a
> manufacturer to expire sources is you might forgo the trouble of
> certifying a perfectly usable source and buy more sources.  I think that
> this wastes money and increases your radioactive waste volume.  If you
> buy new sources, and it is appropriate, specify with the manufacturer
> that no expiration date be placed on the sources or the certificates.
>
> I consulted an ex NBS/NIST scientist about these issues in the past and
> he believes that old sources are often better than new ones.  The key is
> to pay attention to the nuclide production method (trace nuclides
> included), decay chains and source integrity.
>
> -Alan Jackson
> ************************************************************************
> The RADSAFE Frequently Asked Questions list, archives and subscription
> information can be accessed at http://www.ehs.uiuc.edu/~rad/radsafe.html

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