[ RadSafe ] RE: TLD neutron calibration

[Neill Stanford]

Mike,
This is a great question. There are, unfortunately, two goals: worker
dosimetry and accreditation. You must, of course make sure that you pass the
DOELAP (in your case) and should therefore have a calibration factor for
either the unmoderated or D2O moderated Cf-252. You should pick which one
according to "typical" facility conditions. This is much more straight
forward in a nuclear power plant where the neutron field is fairly
homogeneous, and for which the D2O configuration is the accepted standard.
When you are dealing with discrete sources of neutrons, Pu for example, in
different worker-source geometries, the standard single calibration factor
becomes less accurate. Good Health Physics dictates that you use a
conservative (bare Cf252) factor and explain to management why that is OK.
You could also use a building or job specific correction factor, but this is
a logistical nightmare and difficult as it is often impossible to have a
single accurate correction factor for situations where the field conditions
may change even in a single building, depending on the geometry etc.

OK, so the solution is to use a dosimeter that can detect the different
conditions and apply the appropriate correction factor. Several DOE
facilities use 8 element dosimeters to do just this. By using an albedo and
anti-albedo element, you can get an indication of the field, how moderated
it is. This is not a neutron spectrometer, just an indicator. From this
indicator you can generate a family (or define a function) of correction
factors that depends on the ratio of the albedo (hard) to anti-albedo
(thermal) components. Sounds too simple, but it works. At Rocky Flats and
Pantex, they have used this type with great success. We defined a function
that fit Cf-252 bare to 9 cm of moderation within 5%. When tested in the
field against some very challenging Pu configurations, the algorithm was
within about 15% relative to a TEPC detector. I have a demo algorithm
package on my website
http://www.stanforddosimetry.com/algorithm_downloads.htm that includes the 8
element design. You can play with this and "deliver" doses using the various
neutron sources. Pure or mixed, including photons and/or betas, it performs
very well.

If I am correct, LANL is using a very similar approach with your 8 element
dosimeter.

Hope this helps, feel free to call or email for more.


Neill Stanford, CHP
-----------------------------------------------------
Stanford Dosimetry, LLC
PO Box 935
921 S. Fourth St.
Suite B
La Conner, WA 98257
 
www.stanforddosimetry.com
360 466-1090 (voice/fax) NEW NUMBER & ADDRESS AUG 1
360 770-7778 (cell)
------------------------------------------------------

-----Original Message-----
From: Michael McNaughton [mailto:mcnaught at lanl.gov] 
Sent: Friday, August 26, 2005 6:41 AM
To: Sandy Perle; radsafe at radlab.nl; Neill Stanford
Subject: TLD neutron calibration

Sandy, Neill, and others

What neutron spectrum is usually used to calibrate albedo neutron TLDs? Is
it always 252Cf in a 30-cm-diameter D2O sphere? Are other spectra commonly
used?

Thanks, mike
Mike McNaughton
Los Alamos National Lab.
email: mcnaught at LANL.gov or mcnaughton at LANL.gov
phone: 505-667-6130; page: 505-664-7733