Detection of radioactive hot particles in environmental samples using a Marinelli-beaker measuring geometry

Abstract

The presence of one or more hot particles in environmental samples (e.g., soil) in a Marinelli beaker on a HPGe detector can be detected by repeated mixing and counting, and subsequent application of a statistical test to examine whether the observed counts belong to the same Poisson distribution. The method is, however, only efficient if the photons from different random positions of a point source in the Marinelli beaker produce sufficiently different count rates. For a detector with a Be-window, as used here, this is the case for the low-energy photons of 241Am and to a somewhat smaller extent for the high energy photons of 137Cs. As a result, even in the presence of a considerable background activity of the sample matrix, the presence of up to 5 hot particles can be detected by five repeated mixing and counting processes with a probability of >95% for 241Am, but only with a probability of 86% for 137Cs (assuming that about 1000 counts are recorded each time). The easiest way to increase this probability substantially is to increase the counting time or increase the number of mixings. If self adsorption of the photons within the sample occurs, or if the hot particles have different activities, the probability to detect their presence by this method is also increased. An efficiency calibration for the counting geometry used is not required.