Improving Accuracy and Precision in Gross Alpha and Beta Counting with Proportional Detectors

Following ambiguous results of an intercomparison exercise in 2004, we have reevaluated the sample preparation, measurement procedures and data treatment for gross alpha and beta counting. It was found that the Portuguese standards NP 4332:1996 and NP 4330:1996 do not provide enough background for correct treatment of data. This paper describes the improvements made in gross alpha and beta activities measurement with proportional detectors in simultaneous mode. These improvements include a crosstalk correction both on counting and in the minimum detectable activity; the use of adequate statistical tools for data analysis; the use of QC charts to control the stability of the detectors and the background and, attention to the traceability of results. The calibration curves were computed with an appropriate number of replicates and with a throughout evaluation of the uncertainty budget. Based on this model, we established new criteria to report results