Kinetic Parameter Measurements in the MINERVE Reactor

Abstract

In the framework of an international collaboration, teams of the PSI and CEA research institutes measure the critical decay constant ( $\alpha _{0}=\beta /\Lambda $ ), delayed neutron fraction ( $\beta $ ) and generation time ( $\Lambda $ ) of the Minerve reactor using the Feynman- $\alpha $ , Power Spectral Density and Rossi- $\alpha $ neutron noise measurement techniques. These measurements contribute to the experimental database of kinetic parameters used to improve nuclear data files and validate modern methods in Monte Carlo codes. Minerve is a zero-power pool reactor composed of a central experimental test lattice surrounded by a large aluminum buffer and four high-enriched driver regions. Measurements are performed in three slightly subcritical configurations (−2 $\text {c} \!\! | $ to −30 $\text {c} \!\! | $ ) using two high-efficiency 235U fission chambers in the driver regions. Measurement of $\alpha _{0}$ and $\beta $ obtained by the two institutes and with the different techniques are consistent for the configurations envisaged. Slight increases of the $\beta $ values are observed with the subcriticality level. Best estimate values are obtained with the Cross-Power Spectral Density technique at −2 $\text {c} \!\! | $ , and are worth: $\beta =716.9\pm 9.0$ pcm, $\alpha _{0}=79.0\pm 0.6\,\,\text {s}^{-1}$ and $\Lambda =90.7\pm 1.4\,\,\mu \text {s}$ . The kinetic parameters are predicted with MCNP5-v1.6 and TRIPOLI4.9 and the JEFF-3.1/3.1.1 and ENDF/B-VII.1 nuclear data libraries. The predictions for $\beta $ and $\alpha _{0}$ overestimate the experimental results by 3-5% and 10-12%, respectively; that for $\Lambda $ underestimate the experimental result by 6-7%. The discrepancies are suspected to come from the driven system nature of Minerve and the location of the detectors in the driver regions, which prevent accounting for the full reactor.