Validating nuclear data uncertainties obtained from a statistical analysis of experimental data with the “Physical Uncertainty Bounds” method

Abstract

Concerns within the nuclear data community led to substantial increases of Neutron Data Standards (NDS) uncertainties from its previous to the current version. For example, those associated with the NDS reference cross section 239Pu(n,f) increased from 0.6–1.6% to 1.3–1.7% from 0.1–20 MeV. These cross sections, among others, were adopted, e.g., by ENDF/B-VII.1 (previous NDS) and ENDF/B-VIII.0 (current NDS). There has been a strong desire to be able to validate these increases based on objective criteria given their impact on our understanding of various application uncertainties. Here, the “Physical Uncertainty Bounds” method (PUBs) by Vaughan et al. is applied to validate evaluated uncertainties obtained by a statistical analysis of experimental data. We investigate with PUBs whether ENDF/B-VII.1 or ENDF/B-VIII.0 239Pu(n,f) cross-section uncertainties are more realistic given the information content used for the actual evaluation. It is shown that the associated conservative (1.5–1.8%) and minimal realistic (1.1–1.3%) uncertainty bounds obtained by PUBs enclose ENDF/B-VIII.0 uncertainties and indicate that ENDF/B-VII.1 uncertainties are underestimated.

Unrecognized Sources of Uncertainties (USU) in Experimental Nuclear Data

“…However, the estimated uncertainty of 0.13% is practically equal to the value that would be obtained if we neglect all correlations between experiments, which could be an indicator of missing inter-experiment correlations. In addition an uncertainty boundary quantification of the same experiments using the PUBs methodology [74,83] summarized in Appendix VI C clearly shows that the 0.13% uncertainty is significantly lower than the estimated minimum realistic uncertainty bound of 0.23%. Such a difference also indicates that it is very likely that there are missing uncertainties of single experiments and correlations between several measurements.…”

“…While it remains an open area of research, an emerging consensus view in the nuclear data community is that this effect is primarily related to deficiencies in the input data rather than to evaluation methods, provided that evaluators incorporate the available capabilities provided by these methods properly. This phenomenon can be attributed to under-estimation of experimental data uncertainties (for various reasons) and, in many instances, also to inadequate consideration of uncertainty data correlations [74,75].…”

“…Even though it has been demonstrated that inclusion of energy dependent USU values results in an added need to iterate the generalized least-squares (GLS) procedure in performing an evaluation [78], it is worth noting that the increased uncertainties derived for quantities like the 239 Pu fission cross sections generally have been justified by using independent methods (e.g., the PUBs estimate [74]).…”

“…One of the first nuclear-data-specific examples is shown in Ref. [74]. It investigates whether the 239 Pu(n,f) cross-section uncertainties that were produced in the Neutron Data Standards evaluation increased with USU or without consideration of USU are more realistic.…”

“…This is a straightforward demonstration of the PPP effect resulting from employing this artificially constructed covariance matrix in the present least-squares evaluation process. A more realistic case would include an improved UQ leading to a realistic estimation of the covariance matrix, possibly using correlations, for example, based on the PUBs approach [74]. It is also clear that for an actual evaluation of 252 Cf(sf) ν tot , not only the covariances must be carefully analyzed and quantified, but also the original mean values, and in particular the quoted uncertainties, must be reassessed.…”