Speaker
Description
In the last few decades, the field of nuclear data has been increasingly more diligent in the generation and use of uncertainty information related to nuclear data. This increased interest is fueled by the more wide-spread request for characterization of uncertainties, perhaps for quantifying performance margins or for identifying the best use of experimental resources. Use cases range, for instance, from setting criticality safety limits [1], to nuclear data validation [2], to estimating uncertainties on performance of reactor systems [3]. However, despite these needs, processed covariances are not readily available for use in LANL’s application codes such as MCNP. As the covariance user community continues to expand, we must deliver a general purpose processed covariance data library in a similar manner to the widely available continuous energy ACE libraries or internally available multigroup libraries, rather than continuing to rely on ad-hoc processing on an as-needed basis. Thus, an effort is now underway to create such a covariance library or series of libraries based on ENDF/B-VIII.0 covariances.
The creation of this library is not as simple as executing the processing code NJOY [4]. There are known deficiencies in evaluated covariance data, undesirable effects of processing, and difficulties in interpretation of the data. Furthermore, there are challenges in validation and verification, as there is no internal library to use as a baseline (although external libraries do exist) and there is no widely accepted validation test suite. Finally, the very concept of a general-purpose covariance library is difficult, due to the limited set of experimental data used in covariance evaluation.
These challenges have motivated the development of covariance testing tools to be integrated into a formal testing suite. The tests include mathematical checks (e.g., positive definiteness), physics checks (e.g., comparing to experimental uncertainty templates [5]), and “gut-check” validation results on critical benchmarks. Both via this test suite and throughout the processing workflow, deficiencies in both evaluated covariances and the processing tools have been identified, and an approach to correcting these deficiencies is suggested. Lastly, this work includes considerations regarding the delivery of the data, including a usable and modern format and documentation/communication.
References
1. Brian C. Kiedrowski, Forrest B. Brown, Jeremy L. Conlin, Jeffrey A. Favorite, Albert C. Kahler, Alyssa R. Kersting, D. Kent Parsons & Jessie L. Walker, Whisper: Sensitivity/Uncertainty-Based Computational Methods and Software for Determining Baseline Upper Subcritical Limits, Nuclear Science and Engineering, 181:1, 17-47, (2015).
2. D. Neudecker, M. Grosskopf, M. Herman, W. Haeck, P. Grechanuk, S. Vander Wiel, M.E. Rising, A.C. Kahler, N. Sly, P. Talou, Enhancing nuclear data validation analysis by using machine learning, Nuclear Data Sheets, 167, 36-60, (2020).
3. Alexander Aures, Wolfgang Bernnat, Friederike Bostelmann, Jérémy Bousquet, Bernard Krzykacz-Hausmann, Andreas Pautz, Kiril Velkov, Winfried Zwermann, Reactor simulations with nuclear data uncertainties, Nuclear Engineering and Design, 355, 110313, (2019).
4. Macfarlane, Robert, Muir, Douglas W., Boicourt, R. M., Kahler, III, Albert Comstock, & Conlin, Jeremy Lloyd, The NJOY Nuclear Data Processing System, Version 2016. https://doi.org/10.2172/1338791
5. Neudecker, Denise, Hejnal, Brooke Ellen, Tovesson, Fredrik, White, Morgan Curtis, Smith, Donald L., Vaughan, Diane Elizabeth, & Capote, R., Template for estimating uncertainties of measured neutron-induced fission cross-sections, EPJ Nuclear Sciences and Technologies, 4, 21, (2018).
LA-UR-21-29495