Predictive Bias and Sensitivity in NRC Fuel Performance Codes

Recent Updates to NRC Fuel Performance Codes and Plans for Future Improvements

“…In order to facilitate these reviews, capabilities to perform uncertainty analyses have been added to FRAPCON-3.4a. Recently, an uncertainty study was performed to determine the models and manufacturing parameters that have the largest impact on the FRAPCON-3.3 outputs of regulatory interest (maximum fuel centerline temperature, maximum rod internal pressure, and maximum strain increment) [16]. Eight models were identified as those necessary in a bounding design calculation in FRAPCON-3.3.…”

“…Prior to the release of FRAPCON-3.4a, an assessment of FRAPCON-3.3 was performed on the individual model predictions such as fuel thermal conductivity and other material properties and the integral model predictions such as centerline temperature and rod internal pressure to address concerns that FRAPCON-3 is conservatively biased in its predictions [16]. It has been noted that the original fuel performance codes used by the NRC, the predecessors of FRAPCON-3 and FRAPTRAN were intrinsically conservative [16].…”

“…It has been noted that the original fuel performance codes used by the NRC, the predecessors of FRAPCON-3 and FRAPTRAN were intrinsically conservative [16]. In fact, the current version of FRAPCON-3 continues to provide conservative options (although these options are no longer used or recommended by PNNL).…”

“…The assessment performed on FRAPCON-3.3 to determine if the code predictions are conservatively biased was also performed on FRAPTRAN 1.3 [16]. The conclusion of this assessment was that both FRAPCON-3.3 and FRAPTRAN-1.3 provide best-estimate predictions of fuel and cladding temperatures and cladding stress and strain [16].…”

“…The conclusion of this assessment was that both FRAPCON-3.3 and FRAPTRAN-1.3 provide best-estimate predictions of fuel and cladding temperatures and cladding stress and strain [16]. Several areas were identified as areas that should be re-examined and new models developed to improve the model's predictive capability.…”