Nuclear Data Uncertainties in 2004: A Perspective

Abstract

Interest in nuclear data uncertainties is growing robustly after having languished for several years. Renewed attention to this topic is being motivated by the practical need for assuring that nuclear systems will be safe, reliable, and cost effective, according to the individual requirements of each specific nuclear technology. Furthermore, applications are emerging in certain areas of basic nuclear science, e.g., in astrophysics, where, until recently, attention has focused mainly on understanding basic concepts and physics principles rather than on dealing with detailed quantitative information. The availability of fast computers and the concurrent development of sophisticated software enable nuclear data uncertainty information to be used more effectively than ever before. For example, data uncertainties and associated methodologies play useful roles in advanced data measurement, analysis, and evaluation procedures. Unfortunately, the current inventory of requisite uncertainty information is rather limited when measured against these evolving demands. Consequently, there is a real need to generate more comprehensive and reasonable nuclear data uncertainty information, and to make this available relatively soon in suitable form for use in the computer codes employed for nuclear analyses and the development of advanced nuclear energy systems. This conference contribution discusses several conceptual and technical issues that need to be addressed in meeting this demand during the next few years. The role of data uncertainties in several areas of nuclear science will also be mentioned briefly. Finally, the opportunities that ultimately will be afforded by the availability of more extensive and reasonable uncertainty information, and some technical challenges to master, will also be explored in this paper.