A Feynman-α analysis ranging over time constants of delayed neutrons based on time-sequence counting data consecutively acquired for a very long time

Abstract

ABSTRACT Time-sequence counting data of a neutron counter were acquired for 17 consecutive days at shutdown state of UTR-KINKI reactor and then a Feynman-α analysis over a very wide gate width range was carried out to give light on a delayed-neutron contribution of all precursors, where statistical error of the correlation amplitude Y was estimated using the bootstrap method. A first plateau of the Y could be observed in a gate width range of several hundred milliseconds. With an enlargement of gate width beyond a time constant of prompt neutrons, the Y monotonically increased above the first plateau of prompt neutrons. Enlarging gate width beyond 100 s, the Y seemed to approach a second plateau of delayed neutrons. The second plateau originated from the delayed neutrons released from the longest-life 87Br. In the gate width range beyond 100 s, however, the statistical error increased sharply. The increasing error indicates a statistical limit resulting from the present measurement time. A value Y p of the first plateau of prompt neutrons was determined from another Feynman- analysis confined to a short gate width range and the Y was divided by the plateau’s value Y p to eliminate an unknown detection efficiency. The Y-to-Y p ratio was calculated using several sets of delayed-neutron group parameters for thermal fission of 235U nuclide, to validate these sets by comparing with the measured ratio to which all precursors contributed. The ratios calculated using Keepin’s 6-group and Spriggs’ 8-group parameters agreed very well with the measured ratio in a gate width range up to 100 s and consequently these parameters could be validated. However, the ratios calculated using the other sets were in a very poor agreement with the measured ratio in the gate width range.