Influence of the delayed neutron group parameters on reactivity estimation by rod drop analysis

Recent efforts have been undertaken in the reevaluation of the delayed neutron group parameters, namely, the half-life periods and the delayed neutron fractions Indeed, 8-group parameters are now proposed instead of the standard Tuttle 6-group. Moreover, several laboratories among the nuclear engineering community are conducting researches to provide a standard set of group time constants that can suit a lot of fissile isotopes. Since, delayed neutron parameters are widely used in order to estimate reactivity by conducting dynamic experiments, it is worth investigating the influence of the group model (i.e. the number of groups and the number of fitted parameters) on reactivity estimates. This paper focuses on two issues. First, various sets of parameters obtained from JEFF-3.1 precursors data fits are compared. Using simulated rod drop transients, we discuss the uncertainty and bias on reactivity estimation. It is shown that the sets of parameters with fewer fitted parameters and with a greater number of groups give very satisfactory results. Second, the influence of the neutron spectrum is quantified. Biases up to 14% are obtained. Standard statistical tests are done in order to identify such biases on rod drop reactivity estimates.