Traceable to what and where - or whom to believe, Experience with Certified Reference Materials at SAL and ITU

Year
1999
Author(s)
R. Fiedler - IAEA Safeguards Analytical Laboratory
H. Aigner - IAEA Safeguards Analytical Laboratory
S. Deron - IAEA Safeguards Analytical Laboratory
K. Mayer - Institute For Transuranium Elements
Abstract
Accountability measurements of large inventories or high throughputs of nuclear materials and their verifications for safeguards purposes should be performed to the latest standards in measurements, hence with the highest possible accuracy. Therefore reference materials certified to an uncertainty of 0.05% or better must be available to calibrate or validate the analytical methods to the desirable level of accuracy. Similar needs are evident for the analyses of nuclear signatures in environmental samples for safeguards purposes. The IAEA Safeguards Analytical Laboratory (SAL) and the Institute for Transuranium Elements (ITU) are in the fortunate situation to have access to a wide range of Certified Reference Materials (CRM) of various compositions and origins. Experience obtained with the use of CRMs for the calibration and quality control of element assay by potentiometric titration (New Brunswick Laboratory [NBL], of CETAMA and AEAT Harwell) and of isotopic analysis by thermal ionising mass spectrometry (Institute for Reference Materials and Measurements [IRMM], the Khlopin Radium Institute in St Petersburg, NBL and Harwell) will be presented. The method of total sample evaporation and the precision achievable with modern mass spectrometers allow to observe smallest though measurable biases between the measured and the certified values. The significance of these differences will be discussed taking into account the uncertainty statements of the various CRM and a careful estimate of the total uncertainty budget of the measurements. The study points out possibilities of improving further the accuracy of accountability and safeguards analyses. It reiterates the very high responsibility which producers of CRM have in this process. Users of CRM should be provided with comprehensive information on the procedure of preparation and certification of the CRM. Users of CRMs of different origins should be encouraged to publish similar comparisons to the benefit of the measurement community.