Speaker
Description
Authors: 1. M.W. van Rooy (NMISA, South Africa), 2. M. Herranz (UPV/EHU, SPAIN),
3. R. Idoeta (UPV/EHU, SPAIN), 4. L. Verheyen (SCK CEN, Belgium), 5. M. Bruggeman (SCK CEN, Belgium), 6. M. Lepy (LNE-LNHB, France), 7. S. Pierre (LNE-LNHB, France), 8. P. Saganowski (NCNR RC POLATOM, Poland), 9. A. Luca (IFIN-HH, Romania)
Metrologically sound analyses of radioactivity of samples requires that the measurement method is fit for purpose and validated. One of the parameters to consider in the validation, especially when dealing with low-level radioactivity analysis, is the detection limit of the method or for a specific measurement condition of the method. In 2010, with an update in 2019, the concepts of the computation of detection limits and characteristic limits for measurements of ionizing radiation have been published in the ISO 11929 standard. Since most laboratories rely on commercial software to perform gamma-ray spectrometry analyses, verification of this software with respect to the computation of characteristic limits may be required to obtain the proof of compatibility with the ISO standard. Moreover, a verification is useful in order to know and understand the conditions (e.g. critical settings) used in the software to calculate the result. To help the community of radioactivity analysis laboratories, the Gamma-ray spectrometry working group of ICRM started an exercise to investigate the problem of verification of characteristic limits in gamma-ray spectrometry. For that purpose, well defined gamma-ray spectra together with specific instructions fixing key parameters in the computation of detection limits were sent to several laboratories for analysis, evaluation and reporting. The results obtained by different participating laboratories were then compared and were also verified by manual computation. It was observed that different results were obtained by different software packages and that generally no consistent values are directly obtained even when fixing the key parameters. The main causes of the differences in results could be explained by considering the specific approaches and conditions used in the software.
This paper will report on the results obtained in the exercise and will highlight and discuss the impact of important parameter settings of the different software packages. Guidance will be provided on how best to set these parameters to obtain consistent values between software packages used to calculate the decision threshold and detection limit of a specific spectrum.
