Development of a Low-Level Ar 39 Calibration Standard Quantified by Absolute Gas Counting Measurements

Argon-39 is an attractive environmental radiotracer for monitoring phenomena related to ground-water transport. With a half-life of 269 years, ³⁹Ar provides continuity and overlap with other more commonly used radiotracers such as ³H and ¹⁴C. The age of a water sample is determined by comparing the ³⁹Ar specific activity of the gas separated from ground- water with equilibrium ³⁹Ar atmospheric levels (1.01 Bq/kg-Ar_NAT or 1.80×10−6 Bq/cc-Ar_STP ). Such measurements require the use of low-level gas counting techniques which have been calibrated using well characterized ³⁹Ar standards. This presentation describes the generation of ³⁹Ar, via reactor irradiation of potassium carbonate, followed by quantitative analysis, based on length-compensated proportional counting, to yield standards that are approximately 60 and 3 times atmospheric background levels of ³⁹Ar. Multiple measurements of the 60× standard, at various pressures, were performed in Pacific Northwest National Laboratory’s shallow underground counting laboratory in order to study the effect of gas density on beta-transport within the counters. In order to estimate the specific activity of the standard from measurements based on length-compensated proportional counting it is necessary to account for disintegrations that do not deposit sufficient energy in the counter to register above threshold. These losses are commonly referred to as the so-called Wall-Effect (betas reaching the wall before depositing sufficient energy for detection) and Threshold-Effect (total deposited energy below detection threshold). For this study both the Wall- and Threshold-Effect have been estimated using Monte-Carlo simulations and applied to the experimental measurements. An uncertainty model of the measurements and data analysis has been developed in accordance to the Guide to the Expression of Uncertainty in Measurements (GUM). The most challenging source of uncertainty to quantify is that from the Monte-Carlo simulations. The total expanded uncertainty (K=2) result for the 60×-background ³⁹Ar standard, less any uncertainty contribution from the Monte-Carlo simulation, is 1.3% (approximately 95% confidence). Efforts to estimate the magnitude of the uncertainty from the simulation are discussed along with future directions for improved simulations.

This presentation was used for the Low-Radioactivity Underground Argon Workshop held at Pacific Northwest National Laboratory in Richland, Washington on March 19 - 20, 2018.