Nuclear Technology / Volume 185 / Number 1 / January 2014 / Pages 57-70
Technical Paper / Fuel Cycle and Management / dx.doi.org/10.13182/NT13-28
The Electric Power Research Institute (EPRI) has sponsored the development of a set of benchmarks that can be used to quantify the bias and uncertainty in computed reactivity decrements due to burnup. The bias and uncertainty covers imprecision in both the nuclide inventory and cross sections. The EPRI benchmarks are a function of enrichment, operating conditions (such as soluble boron concentration, burnable absorbers, and specific power), and storage rack conditions. The benchmarks are analyzed using SCALE 6.1 with both ENDF/B-V and ENDF/B-VII cross-section libraries. The depletion analyses are performed using the TRITON module, and the criticality calculations are performed with KENO-V.a and MCNP. The analysis shows that SCALE 6.1 with the ENDF/B-VII 238-group cross-section library supports the use of a depletion bias of only 0.0015 in Δk, where k represents the neutron multiplication factor, at peak reactivity after discharge from the core. This peak reactivity occurs after 100 h of cooling. If credit is taken for more cooling, the bias should be increased to 0.0025. The depletion uncertainty is 0.0064. Using MCNP for the criticality calculations rather than KENO-V.a produces essentially the same results if the same ENDF/B cross-section library is used. Reliance on the ENDF/B-V cross-section library produces much larger disagreement with the benchmarks. The analysis covers numerous combinations of depletion and criticality options. In all cases, the historical uncertainty of 5% of the Δk of depletion (“Kopp memo”) was shown to be conservative for fuel with >30 GWd/T burnup. However, the Kopp memo's uncertainty may be exceeded at low burnups where the absolute magnitude of the uncertainty is small.