The use of computed neutron coincidence counting with time interval analysis for the analysis of Fork-measurements on a fresh MOX-LWR fuel assembly under water

The objective of this study was to investigate the influence of different important parameters on measurement results for various fork-detectors. Computed Neutron Coincidence Counting (CNCC) with Time Interval Analysis (TIA) was used for this study. The performance of the electronics for the different fork-detectors was studied by investigating the deadtime perturbed zone of the Rossi-alpha distribution in TIA. The measurement revealed anomalies in the performance of the electronics of the IAEA BWR and LANL fork-detector. The IAEA PWR fork-detector functioned well and the deadtime parameter was calculated. The optimal setting for the pre delay was investigated and it was found that a pre delay of 10 micro seconds should be considered as an optimum between excluding from analysis data in the deadtime perturbed zone and keeping a high signal-to-noise ratio. For the shift register electronics used with the fork-detectors, a pre delay of only 4.5 micro seconds was used. The study of the pre delay and the deadtime showed that the calculated triples-rate is strongly dependent on these parameters. An accurate determination of the triple-rate in this type of measurements has proven to be quite difficult and requires proper operation of the electronics, a correct pre delay and an accurate deadtime correction formalism. By varying the boron concentration in water, the change of the decay time of the Rossi-alpha distribution was clearly observed. This change is due to the variation of the thermal multiplication. The variation of this decay time with the boron concentration proves that Boehnel's model for fast neutron multiplication is not valid under these measurement conditions and that a model for fast and thermal multiplication should be used in order to obtain unbiased measurement results. CNCC with TIA has proved to be a valuable tool in which parameter settings can be varied a posterori and the optimal setting can be determined for each measurement. Moreover, the display of the time intervals allows a simple control of the proper system performance