Year
2001
Abstract
The heavy metal loading of fuel assemblies modifies the response of the detector when neutron coincidence counting is used. A correction for this effect has been in use for more than 10 years in the case of active interrogation of LEU fuel assemblies. The same effect can be seen in the case of passive measurements of MOX assemblies. Until now it has only been possible to make good measurements by using calibration curves that are specific to one fixed fuel assembly design. This paper describes a heavy metal correction applied to passive coincidence data that allows a single calibration curve to be applied for a wide range of assembly designs and heavy metal loadings. This is important to safeguards for two reasons. Firstly, because it ensures consistency over a wide range of fuel assembly types, giving greater independence from operator supplied information. Secondly, establishing multiple calibrations is an inefficient use of resources when a better alternative exists. A set of data from 632 MOX fuel assemblies was collected, using a Neutron Coincidence Collar with automated data acquisition, and analysed. Assemblies ranged from BWR types that include separate MOX and low enriched uranium fuel rods (8x8, 9x9, and 10x10), to PWR types that have only MOX fuel rods using either depleted or natural uranium (14x14, 15x15, 16x16, 17x17, and 18x18). The heavy metal loading of the assemblies varied from 46 to 137 g/mm. Relative differences between measured and declared 240Pu effective values were between -2.4 and 2.2 % and were independent of the heavy metal loading. Uncertainties are less than 1%. This correction may also be useful for other situations where the geometry of the sample changes with mass, for example neutron coincidence counter measurements of boxes containing different number of fuel pins.