skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Representing full-energy peak gamma-ray efficiency surfaces in energy and density when the calibration data is correlated

Abstract

The non intrusive quantification of gamma-emitters in waste containers calls for a correction to be made for the perturbation introduced by the container contents relative to the configuration used for the absolute calibration of the measurement system. There are several potential ways of achieving this including the use of an external transmission beam, and, the exploitation of the differential attenuation between different energy lines from the same nuclide. The former requires additional hardware while the second is not always applicable. A third method, which overcomes both these objections and is also commonly applied as the method of choice on systems that do not have the capability to (axially) scan the item, is termed the Multi-Curve approach. When applying the Multi-Curve method the density of the waste matrix inside the item under study is estimated from the net weight and the fill height estimate and the efficiency at the assay energies of interest is obtained by navigating the efficiency-energy-density surface using the (interpolation) scheme developed for the model parameters via the calibration procedure. In addition to the nominal efficiency values an uncertainty estimate for each is made using an independent analysis engine which is designed to incorporate reasonable deviations in themore » properties of the item from the ideal conditions of calibration. Prominent amongst these would be deviations of fill matrix homogeneity, deviations from a uniform activity distribution and deviation in the atomic composition of the materials present from those used to make the calibration items (this is of concern below about 200 keV where the photoelectric effect, which has a strong Z-dependence, comes into play). The Multi-Curve approach has proven to be robust and reliable. However what one finds is that the uncertainties assigned to the traditional Multi-Curve interpolation scheme are underestimated. This is because correlations in the input data are neglected and also because the fitting procedure is a multi-step process which does not properly propagate the raw uncertainty information. Correlations occur because, for example, {sup 152}Eu emits several lines so that if the activity of {sup 152}Eu in the mixed nuclide source is high, relative to the rest, than all of the {sup 152}Eu lines will be affected in the same, systematic rather than random, direction. The propagation of uncertainties is incomplete because fitted results are used as starting points for secondary fits and again correlations are not included. It is appreciated that in representing the efficiency by way of empirical forms which has no physical basis that the basic assumption of curve fitting (namely that the fitted function is the correct one) has been violated so that a case can be made that the accuracy in any given region of the surface is best assessed by a subject matter expert from the scatter in the calibration data from the fitted behavior. This kind of ad hoc treatment can be accommodated by terms in the separate Total Measurement Uncertainty engine applied to the final assay result. However, from a purists stand point it is of interest to conduct a more thorough assessment of the impact of correlations in the calibration data. As noted although not generally important for the assay of actual unknown waste items correlations may be consequential for certain testing and validation need. This work addresses how correlations in the calibration data can be included and propagated through the Multi-Curve scheme and assesses how important they are in a typical scenario. (authors)« less

Authors:
; ; ;  [1]
  1. Canberra Industries Inc., Meriden, CT (United States)
Publication Date:
Research Org.:
WM Symposia, 1628 E. Southern Avenue, Suite 9 - 332, Tempe, AZ 85282 (United States)
OSTI Identifier:
21294698
Report Number(s):
INIS-US-09-WM-07325
TRN: US10V0224040888
Resource Type:
Conference
Resource Relation:
Conference: WM'07: 2007 Waste Management Symposium - Global Accomplishments in Environmental and Radioactive Waste Management: Education and Opportunity for the Next Generation of Waste Management Professionals, Tucson, AZ (United States), 25 Feb - 1 Mar 2007; Other Information: Country of input: France; 10 refs
Country of Publication:
United States
Language:
English
Subject:
46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND TECHNOLOGY; 42 ENGINEERING; CALIBRATION; CONTAINERS; CORRELATIONS; ENERGY DENSITY; EUROPIUM 152; GAMMA RADIATION; INTERPOLATION; KEV RANGE 100-1000; PEAKS; PHOTOELECTRIC EFFECT; RADIATION DETECTION; RADIATION DETECTORS; RADIOACTIVE WASTES; SURFACES; VALIDATION

Citation Formats

Henry, M, Croft, S, Zhu, H, and Villani, M F. Representing full-energy peak gamma-ray efficiency surfaces in energy and density when the calibration data is correlated. United States: N. p., 2007. Web.
Henry, M, Croft, S, Zhu, H, & Villani, M F. Representing full-energy peak gamma-ray efficiency surfaces in energy and density when the calibration data is correlated. United States.
Henry, M, Croft, S, Zhu, H, and Villani, M F. Sun . "Representing full-energy peak gamma-ray efficiency surfaces in energy and density when the calibration data is correlated". United States.
@article{osti_21294698,
title = {Representing full-energy peak gamma-ray efficiency surfaces in energy and density when the calibration data is correlated},
author = {Henry, M and Croft, S and Zhu, H and Villani, M F},
abstractNote = {The non intrusive quantification of gamma-emitters in waste containers calls for a correction to be made for the perturbation introduced by the container contents relative to the configuration used for the absolute calibration of the measurement system. There are several potential ways of achieving this including the use of an external transmission beam, and, the exploitation of the differential attenuation between different energy lines from the same nuclide. The former requires additional hardware while the second is not always applicable. A third method, which overcomes both these objections and is also commonly applied as the method of choice on systems that do not have the capability to (axially) scan the item, is termed the Multi-Curve approach. When applying the Multi-Curve method the density of the waste matrix inside the item under study is estimated from the net weight and the fill height estimate and the efficiency at the assay energies of interest is obtained by navigating the efficiency-energy-density surface using the (interpolation) scheme developed for the model parameters via the calibration procedure. In addition to the nominal efficiency values an uncertainty estimate for each is made using an independent analysis engine which is designed to incorporate reasonable deviations in the properties of the item from the ideal conditions of calibration. Prominent amongst these would be deviations of fill matrix homogeneity, deviations from a uniform activity distribution and deviation in the atomic composition of the materials present from those used to make the calibration items (this is of concern below about 200 keV where the photoelectric effect, which has a strong Z-dependence, comes into play). The Multi-Curve approach has proven to be robust and reliable. However what one finds is that the uncertainties assigned to the traditional Multi-Curve interpolation scheme are underestimated. This is because correlations in the input data are neglected and also because the fitting procedure is a multi-step process which does not properly propagate the raw uncertainty information. Correlations occur because, for example, {sup 152}Eu emits several lines so that if the activity of {sup 152}Eu in the mixed nuclide source is high, relative to the rest, than all of the {sup 152}Eu lines will be affected in the same, systematic rather than random, direction. The propagation of uncertainties is incomplete because fitted results are used as starting points for secondary fits and again correlations are not included. It is appreciated that in representing the efficiency by way of empirical forms which has no physical basis that the basic assumption of curve fitting (namely that the fitted function is the correct one) has been violated so that a case can be made that the accuracy in any given region of the surface is best assessed by a subject matter expert from the scatter in the calibration data from the fitted behavior. This kind of ad hoc treatment can be accommodated by terms in the separate Total Measurement Uncertainty engine applied to the final assay result. However, from a purists stand point it is of interest to conduct a more thorough assessment of the impact of correlations in the calibration data. As noted although not generally important for the assay of actual unknown waste items correlations may be consequential for certain testing and validation need. This work addresses how correlations in the calibration data can be included and propagated through the Multi-Curve scheme and assesses how important they are in a typical scenario. (authors)},
doi = {},
url = {https://www.osti.gov/biblio/21294698}, journal = {},
number = ,
volume = ,
place = {United States},
year = {2007},
month = {7}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share: