Estimation of the uncertainty components associated with the measurement of radionuclides in air filters using γ-ray spectrometry

Abstract

This paper discusses and assesses the major sources of uncertainty arising in measurement of radionuclide activity concentrations in air. Besides counting statistics, the main contributors to the overall uncertainty of the activity concentration of long-lived radionuclides are the heterogeneities of the calibration source and the test samples (approximately 4%). In addition, in the case of radionuclides with half-lives of the order of days or shorter, a significant uncertainty results from the variability of the “source term” because the peak concentration of the radionuclide might occur either at the beginning or at the end of the sampling period. For¹³¹I (T_1/2≈8 days), for instance, the relative standard uncertainty was estimated as 17% owing to this effect. In addition, the uncertainty introduced by the software used for evaluation of the γ-ray spectra is discussed. Details of the uncertainty components, including their assessed probability distributions and degrees of freedom, and the conversion into standard uncertainties are reported in this paper, following the ISO Guide to the expression of uncertainty in measurement. A standard form for documenting and reporting the uncertainty budged is proposed and illustrated with an example.