Re: Electronic Personal Dosimeters

[bflood@SLAC.Stanford.EDU (Bob Flood)]

In my nuclear power days, I worked with 1,000's of the MGs. To add to
Sandy's comments:

>     MGP Electronic Dosimeter Evaluation:
>     
>     1. Accuracy is only as good as the source calibration and energy you 
>     calibrate them to. 

Accuracy is valid **only** at the calibration energy. MGs have a distinct
energy dependence for photons, and accuracy changes with photon energy.
Compromise settings are necessary.
>     
>     2. Precision is excellent.

Indeed.
>     
>     4. From item 4 you can see that the EPD highly under-responds to 
>     low-energy photons, and under 51 keV there is essentially NO response 
>     whatsoever.
>     
Agreed, and there is variation by model. Some "go blind" below 50 keV,
others below 70 keV. Also, there is a fall-off in response at high energies,
such the over 3 MeV photons from N-16. This a generic silicon diode problem.

>     5. Angular dependence Category IX was passed without any difficulty.

Be very careful here. There is good response when tested in a +/- 85 degree
arc in the vertical and horizontal, but the unit under-responds when
irradiated from behind (assuming it is calibrated in a frontal geometry).
This is important because users tend to want to wear the unit turned
backwards in a shirt or coverall pocket. Part of the calibration compromise
(could it be part of an ALARA improvement?). This is not a generic silicon
diode problem; it's a MG design issue.
     


Note the 1/3 failure rate. Saw this each year! 100% failure over three years
is a significant operational problem. It has a strong impact on the
inventory a site has to maintain. When the warranty runs out, it also
becomes a significant annual maintenance cost.
>     
>     8. Moisture has caused problems by getting inside the case ...
>     
Indeed.

>     9. Speaker issues where the wires are not well manufactured, 
>     therefore, the alarms are not heard.. Need a program to ensure that 
>     the speaker is functional, that the battery is ok .. etc.

Most common failure - the speaker detaches. Since the alarm function is a
major selling point, ??????????
>     
>
One more item, a productivity issue, to think about. None of the EDs that
I've seen lend themselves to any kind of automated testing or calibration.
Given the need to check for problems (based on the established failure
rate), this becomes a serious labor problem. In a program where TLD process
for multiple power plants was a 4 man-year job, ED calibration and
maintenance required 6 man-years, and repairs were done by the manufacturer.
And each person was issued an individual TLD, but the EDs were shared (a
much smaller inventory)!


Opinion: Electronic dosimeters are a major help for compliance with the
special dosimetry requirements for entry into high radiation and locked high
radiation areas. The dose and dose rate measurements plus alarm feature
clearly meet the license/reg requirements. They also make electronic access
control a reality in well-confined facilities like nuclear power plants. The
measurement shortcomings make accreditation difficult and perhaps
impossible, depending on your particular radiological environment. As a
replacement for TLDs for dose-of-record measurements, EDs carry the burden
of being an inferior measurement system, all radiations considered, and
require more of the organization's resources (high initial investment, high
labor cost, high failure rate). In short, I'd buy EDs only for use as
secondary dosimeters, and only if I need the features they offer that TLDs
cannot perform. Otherwise, I'd use pocket chambers - they have their
problems, but they are reasonable secondary dosimeters at reasonable cost.
Bob Flood
Unless otherwise noted, all opinions are mine alone.
(415) 926-3793
bflood@slac.stanford.edu