Radioassay of a portion of the main radioactive ventilation header at the Radiochemical Development Facility using in situ gamma spectrometry methods for intact fluid systems

There exists a need for a method to radioassay components and fluid systems internally contaminated with radioactive material. Often operating constraints require these analyses to be performed without system penetration. The results of an accurate analysis of this type could be used for radiological planning and waste generation estimates. Gamma spectrometry methods are a valid choice, but have a large associated uncertainty due to unknown source distribution, shielding composition, and background radiation contribution. The most common approach to the problem is to survey the component with a field survey meter and, given process knowledge of the isotope of concern, convert the exposure rate to an activity using a thumbrule. Some use of portable gamma spectrometry equipment has been made in this application, but the assumptions made to compensate for shielding and correct for unrepresentative efficiency calibration involve large margins of uncertainty. More often than not, the uncertainty in any of these methods is not well defined, even though it could approach several orders of magnitude. A procedure was developed to use portable gamma spectrometry equipment to characterize the type and amount of radioactivity in an intact fluid system without system penetration. The background radiation contribution was eliminated using partial shielding. The source and shielding distribution effects on the efficiency calibration were corrected by comparing fluence rate ratios to detector efficiency. This efficiency correction factor was based on fluence rates generated by both a point-kernel shielding software application and a Monte Carlo simulation software template run out of a spreadsheet application. The accuracy of the analysis was then measured by performing an uncertainty analysis on the entire process using numerical and analytical propagation of error techniques. The entire portion of the header analyzed was determined to contain 872 Mbq ± 40 % of ¹³⁷Cs(^137mBa). The header was analyzed piece-wise, with an associated activity and uncertainty for each of the fifteen segments analyzed. The overall analysis uncertainty was based on the activity estimate and uncertainty analysis developed fi-om the Monte Carlo simulation software template, which was found to be the more applicable software for this procedure.