The uncertainties of the analysis results of
134Cs,
137Cs and
40K in the Brown Rice certified reference material by gamma-ray spectrometry using a Ge detector have been evaluated. The massic activities of
134Cs,
137Cs and
40K in the CRM were determined by a simple comparison and efficiency curve methods using standard gamma-ray sources of the analyzing radionuclides and a multi-nuclides mixed standard source, respectively, for gamma-ray peak efficiency calibration. Standard uncertainties of constituted parameters of the model equations of the analysis methods were determined from close examination of each uncertainty component of the parameters. Although it is not contained in the model equations, a parameter of repeatability due to a variation of the sample volume, of the geometrical condition and of the sample preparation were evaluated by an experiment using radionuclide standard solutions of
137Cs and
152Eu. The evaluated combined standard uncertainties of each parameter of the model equations were further combined following the uncertainty propagation law and obtained expanded uncertainties multiplying by a coverage factor (
k = 2). Gamma-ray peak efficiency and net peak count were the largest uncertainty components in the simple comparison method, and those of the other components were negligibly small. As compared to the simple comparison method, due to the additional component of true coincidence summing (TCS) correction and gamma-ray emission probability, expanded uncertainty of the efficiency curve method was larger than that of the simple comparison method. Due to lack of appropriate information on uncertainties for TCS correction of cascade gamma-rays from
134Cs, 5 % relative standard uncertainty, which was obtained from radioactivity analysis of the
134Cs standard source by the efficiency curve method, was assigned for the efficiency curve method of
134Cs. The utilization of a low background gamma-ray spectrometer and the development of an accurate correction method of TCS are required to obtain a smaller uncertainty of the analysis methods.
View full abstract