Reanalysis of graphite nuclear data limits for AGR dosimetry with updated carbon scattering data and Magnox plant measurements

Abstract

The Advanced Gas-cooled Reactors (AGR) operating in the UK are, like the Magnox reactors that preceded them, graphite moderated and cooled by pressurised CO2 gas. As such, the nuclear data for carbon are of primary importance in neutron dosimetry and nuclear heating calculations. There is evidence from benchmark data and plant measurements of underprediction of neutron fluxes at the lower end of the fast neutron range for thicknesses of graphite greater than ~50 cm. Furthermore, the range of validation is limited to ~70 cm of graphite, with components beyond this being allocated bounding damage rates. Following the previous ISRD, opportunities were taken to: Look again at the carbon scattering cross-sections by incorporating the ENDF/B-VIII.0 Beta 4 carbon data to assess the impact on the analysis of benchmark data. Reassess validation data from the earlier Magnox reactors to put them on a more appropriate equivalent graphite thickness scale to the AGRs. The investigation into graphite nuclear data libraries concluded that the incorporation of ENDF/B-VIII Beta 4 carbon data into the JEF2.2 and ENDF/B-VII.1 data libraries with BINGO collision processing had only a limited effect on the calculated responses. The UKNDL data, with DICE collision processing, currently used in dosimetry calculations for EDF Energy’s AGRs produces the closest agreement with measurement for all calculated responses for graphite thicknesses greater than 10 cm. For reactor plant validation, applying the sensitivity of the 93mNb(n,n’) response to graphite density resulted in a 15 cm increase to the validation database when compared to a simple “straight-line” distance. However, caution is advised in applying the benchmark equivalent distance derived from locations within a graphite column, to ex-core plant measurements.