Abstract

This manuscript addresses a key technical gap in cross-section adjustment and uncertainty analysis research employed in support of neutronic model validation and data assimilation. The key assumption in this body of research is that cross-sections, in condensed group format, are responsible for the key sources of uncertainties in responses of interest, e.g., multiplication factor, flux and power distribution, etc. In reality, the condensed group cross-sections are contaminated not only by epistemic uncertainties, propagated from ENDF, but also by the various modeling assumptions and approximations, typically customizing the condensed cross-sections to the reactor type and flux spectra. Thus, any reliable procedure employed to identify or adjust for key sources of uncertainties must be able to distinguish between epistemic uncertainties and modeling errors. This manuscript proposes a new algorithm that adds new types of constraints to conventional sensitivity analysis procedure to ensure that uncertainty analysis and cross-section adjustment studies are not contaminated by the modeling errors. The proposed approach will directly benefit research and development efforts in many areas of analysis relying on uncertainty and sensitivity methods, such as criticality safety, core monitoring, and analysis of first-of-a-kind reactor and fuel concept technologies. Demonstrative examples using a CANDU-6 core model during steady state and transient conditions are employed to exemplify application of the proposed methodology.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call