Answers are grounded in this paper and cite their sources.
There is a notable lack of easily reproducible, open-source, modern simulation benchmarks for nuclear reactors that can serve as credible references for verification and validation. To be effective, benchmarks should be supported by well-maintained models, comprehensive documentation, and clearly defined reference results to ensure long-term usability. In practice, however, this is rarely the case. The continual evolution of reactor physics simulation tools, including updates to geometry models, nuclear data libraries, and neutronics codes erodes the consistency and comparability of benchmark results over time.This work presents a systematic reevaluation of the BEAVRS Hot Zero Power benchmark, originally released in 2013, by adopting ENDF/B-VIII.1 as the new reference cross-section library in place of the decade-old ENDF/B-VII.1 baseline. Differences between nuclear data libraries and their processing pipelines are quantified and shown to exceed Monte Carlo statistical uncertainties by approximately 40 pcm. The treatment of intermediate temperatures is examined in detail, revealing a discontinuity in the “nearest” temperature method caused by mismatched temperature grids between base and thermal scattering libraries.The result is an updated, fully reproducible set of hot zero power reference values that improves agreement with measured control rod worths relative to both the original 2013 benchmark and the interim PHYSOR 2026 reevaluation, along with a quantified uncertainty budget attributable to nuclear data library selection.
No takes yet. Share an insight, caveat, or question.
Oteri et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: