Uncertainty analysis in environmental radioactivity measurements using the Monte Carlo code MCNP5

Autores UPV
Año
CONGRESO Uncertainty analysis in environmental radioactivity measurements using the Monte Carlo code MCNP5

Abstract

In environmental radioactivity measurements, High Purity Germanium (HPGe) detectors are widely used due to its excellent energy resolution. Monte Carlo (MC) codes are a useful tool to complement experimental measurements in the calibration procedures of the laboratory. However, the efficiency curve of the detector is affected by some uncertainties associated to different aspects of the measurements. These uncertainties can be classified into some categories: geometrical features of the measurement (distance source-detector, volume of the source, etc), properties of the radiation source (activity of each radionuclide, branching ratio), and detector characteristics (Ge dead layer, active volume, end cap thickness). The Monte Carlo simulation is also affected by other kind of uncertainties mainly related to cross sections and to the calculation itself. Normally all these uncertainties are not well known and it is required a deep analysis to determine their effect on the detector efficiency simulated. In this work, the Noether-Wilks formula is used to carry out the uncertainty analysis. A probability density function (PDF) is assigned to each variable before the sampling process. The size of the sampling is determined from the characteristics of the tolerance intervals by applying the Noether¿Wilks formula and according to the degree of precision desired for uncertainty measures. Results of the analysis transform the efficiency curve into a region of possible values inside the tolerance intervals. By means of a sensitivity analysis, the influence of each variable can be determined. Results show that the most important variables affecting the efficiency depend on the detector features.