On the effects of oxygen cross-sections in the fast neutron fluence analysis for a reactor pressure vessel scraping test

Studies in support of the assessment of aging structural materials in pressurized water reactors are being performed at the Paul Scherrer Institut. To that aim, a state-of-the-art methodology based on applying a CASMO-4/SIMULATE-3/MCNPX calculation scheme has been developed. In the frame of the methodology validation, an investigation is currently reported pertaining to the sensitivity of the calculated results, for a specific reactor pressure vessel scraping test, to the nuclear data used with the Monte Carlo code. Thus, the MCNPX-2.4.0 calculations have been carried out using three different data libraries, based on JEF-2.2, ENDF/B-VI.8 and JENDL-3.3 evaluations, respectively. A significant discrepancy ( 10%) in fast neutron flux (E > 1 MeV) estimations has been observed between the results obtained with JENDL-3.3, versus the two other libraries, while noting that the later both provide very satisfactory agreement (within ±5%) with the reference results based on the experimental data. Subsequent analysis has indicated that the observed discrepancy can be attributed mostly to differences in the oxygen data. The discussed cross-section differences could potentially lead to more sizable discrepancies for other applications, and thus need to be brought to the attention of neutron data evaluators and users.