Characterization of neutron reference fields at US Department of Energy calibration fields.

The Health Physics Measurements Group at the Los Alamos National Laboratory (LANL) has initiated a study of neutron reference fields at selected US Department of Energy (DOE) calibration facilities. To date, field characterisation has been completed at five facilities. These fields are traceable to the National Institute for Standards and Technology (NIST) through either a primary calibration of the source emission rate or through the use of a secondary standard. However, neutron spectral variation is caused by factors such as room return, scatter from positioning tables and fixtures, source anisotropy and spectral degradation due to source rabbits and guide tubes. Perturbations from the ideal isotropic point source field may impact the accuracy of instrument calibrations. In particular, the thermal neutron component of the spectrum, while contributing only a small fraction of the conventionally true dose, can contribute a significant fraction of a dosemeter's response with the result that the calibration becomes facility-specific. A protocol has been developed to characterise neutron fields that relies primarily on spectral measurements with the Bubble Technology Industries (BTI) rotating neutron spectrometer (ROSPEC) and the LANL Bonner sphere spectrometer. The ROSPEC measurements were supplemented at several sites by the BTI Simple Scintillation Spectrometer probe, which is designed to extend the ROSPEC upper energy range from 5 to 15 MeV. In addition, measurements were performed with several rem meters and neutron dosemeters. Detailed simulations were performed using the LANL MCNPX Monte Carlo code to calculate the magnitude of source anisotropy and scatter factors.

Multisphere default spectra--solution spectrum and dosemeter response implications.

Initial calibration of a multisphere spectroscopy system has been completed at Los Alamos National Laboratory using four standard calibration scenarios. Spectrum unfolding was performed using three methods of constructing the default spectrum: simple parameter models, Monte Carlo calculations and physical measurement. Comparisons of the resulting spectra for each solution method are presented. Implications of the spectral solutions upon dosemeter characterisation are addressed.