Uncertainties in gamma-ray spectrometry

Abstract

High resolution gamma-ray spectrometry is a well-established metrological technique that can be applied to a large number of photon-emitting radionuclides, activity levels and sample shapes and compositions. Three kinds of quantitative information can be derived using this technique: detection efficiency calibration, radionuclide activity and photon emission intensities. In contrast to other radionuclide measurement techniques gamma-ray spectrometry provides unambiguous identification of gamma-ray emitting radionuclides in addition to activity values. This extra information comes at a cost of increased complexity and inherently higher uncertainties when compared with other secondary techniques. The relative combined standard uncertainty associated with any result obtained by gamma-ray spectrometry depends not only on the uncertainties of the main input parameters but also on different correction factors. To reduce the uncertainties, the experimental conditions must be optimized in terms of the signal processing electronics and the physical parameters of the measured sample should be accurately characterized. Measurement results and detailed examination of the associated uncertainties are presented with a specific focus on the efficiency calibration, peak area determination and correction factors. It must be noted that some of the input values used in quantitative analysis calculation can be correlated, which should be taken into account in fitting procedures or calculation of the uncertainties associated with quantitative results. It is shown that relative combined standard uncertainties are rarely lower than 1% in gamma-ray spectrometry measurements.

Evidence against solar influence on nuclear decay constants

“…The measurement techniques employed are as follows: ionisation current measurements in a re-entrant ionisation chamber (IC) or a hospital calibrator (HIC) [31,32], net area analysis of full-energy γ-ray peaks (and integral spectrum counting) by γ-ray spectrometry with a HPGe detector (HPGe) [33], particle counting in a planar silicon detector in quasi-2πconfiguration (PIPS) [34], X-ray counting at a small defined solid angle with a gas-filled proportional counter (PC) [35,36], live-timed β–γ anti-coincidence counting (LTAC) [37], triple-to-double coincidence counting with a liquid scintillation vial and three photodetectors (TDCR) [38], liquid scintillation counting (LSC) [38], particle and photon counting in a sandwich CsI (Tl) spectrometer (CsI) [39], internal gas counting (IGC) [40], and α-particle counting at a small defined solid angle with a large planar silicon detector (αDSA) [35,36]. An overview of standardisation techniques and their sources of error can be found in the special issues 44(4) and 52(3) of Metrologia [41,42] and references in [25,28].…”

60Co in cast steel matrix: A European interlaboratory comparison for the characterisation of new activity standards for calibration of gamma-ray spectrometers in metallurgy

“…From Table 4 it can be observed that the combined standard uncertainty values varied by an order of magnitude among the participating laboratories. To further optimize the estimation of uncertainties in gamma-ray spectrometry measurements, dedicated work was been recently published by Cassette et al (2015) and by Lepy et al (2015).…”

A new approach to characterize very-low-level radioactive waste produced at hadron accelerators

“…A rigorous treatment of uncertainties in γ-ray spectrometry can be found in Lépy et al (2015). Complementary information can be found in Sima and Arnold (2009).…”

“…(5) is simplified because assumes independence between the input variables but also because real spectra often present multiple interfering peaks from which the activity is evaluated using numerical methods. A complete formulation of the uncertainty associated to the activity of γ-emitters, including the evaluation of the covariance terms, can be found in Lépy et al (2015).…”

“…To take these effects into account, we generated multiple efficiency calibration functions by making the input parameters vary within given intervals. Nevertheless, such stochastic variations play a minor role compared with the relative uncertainties affecting the total efficiency (5-10% according to Lépy et al, 2015).…”