Answer to Question #12326 Submitted to "Ask the Experts"

Can a G-M detector (such as a submersible model like the AMP-100), which has been calibrated by the manufacturer in air, accurately measure, under water, the dose rate from material contaminated with ¹³⁷Cs, or must a correction factor be applied?

The short answer to your question is a qualified "Yes." The qualifications relate to whether the conditions under which the measurements are to be made are consistent with those for which the detector is designed and for which it provides acceptably accurate readings. The major concerns in this case would be whether the dose rates to be measured are within design specifications, whether the photon energy distribution producing dose is within the acceptable range for the detector, and possibly whether there is an acceptable directional response.

I checked the specifications for the detector, and the acceptable dose rate range is from about 50 µSv h^-1 to 10 Sv h^-1, which is likely sufficient for most cases of underwater measurements of concern. According to specifications, the acceptable photon energy distribution, measurable with ±20% accuracy, relative to ¹³⁷Cs, covers the range from 70 keV to 2 MeV. A major difference between the water environment and the air environment is that the water produces much more photon scattering than does the air for a particular source-detector separation. The increased scattering results in more low-energy photons being incident on the detector. Depending on the source-detector geometry and the energies of the original photons, some of the scattered photons may fall below 70 keV, which might result in them not be detected. If the detector is relatively close to the source(s), the response will tend to be weighted by primary photons and forward-scattered photons, whose energies would likely be >70 keV.

Regarding directional response, the specifications note that the angular dependence is "Less than ±20% for 45° from centerline indicator." This is somewhat difficult to interpret, but I assume it means that the angular response does not change by more than 20% when the direction of incident photons are along a line that varies up to 45^o from a perpendicular line to the center point of the longitudinal centerline of the detector. This interpretation would lead to the inference that photons entering the detector at greater angles, with the most severe cases being entrance into the ends of the detector (±90^o), would suffer degradation in response greater than 20%, although it is not possible to tell how large an effect it might be. If the longitudinal axis of the detector is facing the source(s), and the detector is rather close to the source(s), the end effect will be reduced.

I cannot provide much more definitive information at this point. If you require more quantitative analysis it may well demand a probabilistic type of analysis of the kind represented by Monte Carlo calculations. I wish you well in your continued work.

George Chabot, PhD, CHP