Uncertainty Estimation for Destructive Analysis Techniques

Accurate and traceable measurement values constitute an important component of the nuclear materials control and accountability system. The uncertainties associated with measurements must be determined and propagated to determine the uncertainty of SNM inventory at a facility. Facilities use Certified Reference Materials (CRMs) for calibration to establish and confirm traceability to national or international measurement standards. It will therefore be beneficial to indicate the sources and estimate of measurement uncertainties for CRMs. Also, recording the uncertainty budget enables one to identify major sources that contribute to the total uncertainty. This paper describes how NBL certifies measurement values by identifying individual uncertainty components, estimating their contributions, and combining these values into a total measurement uncertainty. NBL uses destructive analysis techniques such as titration, coulometry and mass spectrometry for the certification of nuclear reference materials. These methods are reliable and reproducible, and have been proven to yield highly accurate values for the measured properties with well-quantified combined uncertainties. Each of these measurement methods can be completely described by equations that link the property being measured to the International System of Units (SI). Properly designed measurement protocols are essential for obtaining quality data that can be used to identify sources of uncertainty and evaluate their relative contributions. The sampling and analysis plans that NBL develops to certify nuclear reference materials are designed to identify these uncertainty components, minimize sources of uncertainty, and provide traceability to the national measurement base. We use the equations that describe the measurement methods in estimating the different uncertainty components in the certification of reference materials. NBL also develops and implements computer-based applications to evaluate measurement uncertainties. These developments allow us to better document the traceability chain to the SI and provide a better estimate of the total uncertainty.