The Need and Opportunity for Improved Uncertainty Quantification and Reporting for Nondestructive Assay

: In order to be useful a physical measurement or calculation must be assigned a justifiable statement of the associated total measurement uncertainty (TMU) which is fit for purpose. Uncertainty quantification (UQ) is the process of placing probabilistic bounds on the confidence we have in our state of knowledge or understanding, reporting the bounds, and explaining how the bounds were obtained. UQ is a vital aspect in quality nondestructive assays (NDA). Guidance available to the NDA community on formulating and expressing uncertainties is scattered throughout technique-specific reports. It is difficult to find, often not current, and can be confusing because of the variety of notations and terminology that have been used over the years. ASTM sub-committee C26.10 is responsible for writing and maintaining guides and standards on a diverse range of commonly used NDA methods for nuclear safety, safeguards, security, and waste. Constructing and verifying total measurement uncertainty budgets that are suitable for the end user has been recognized as a complex and specialist task. Statistical methods and tools have advanced substantially in the past decade rendering certain previous assumptions and simplifications unnecessary. The opportunity to create a rational and consistent UQ framework across techniques is also evident and will promote better understanding and good practice. To address these issues, preparation of a consensus guide for the determination of TMU in NDA measurement results has begun. This is intended to be a valuable resource for, amongst others, subject matter experts, instrument operators, instrument designers, facility managers, instructors, and end users of NDA data. In this paper we discuss the importance of this activity to the NDA measurement community and solicit input to the process. We illustrate the discussion with a pertinent few examples.