Empirical comparison of neutron activation sample analysis methods
Date
2011
Authors
Gillenwalters, Elizabeth, author
Johnson, Thomas E. (Thomas Edward), 1964-, advisor
Pinder, John E., committee member
Kearney, Philip D., committee member
Journal Title
Journal ISSN
Volume Title
Abstract
The U.S. Geological Survey (USGS) operates a research reactor used mainly for neutron activation of samples, which are then shipped to industrial customers. Accurate nuclide identification and activity determination are crucial to remain in compliance with Code of Federal Regulations guidelines. This facility utilized a Canberra high purity germanium detector (HPGe) coupled with Canberra Genie™ 2000(G2K) software for gamma spectroscopy. This study analyzed the current method of nuclide identification and activity determination of neutron activated materials utilized by the USGS reactor staff and made recommendations to improve the method. Additionally, analysis of attenuators, effect of detector dead time on nuclide identification, and validity of activity determination assumptions were investigated. The current method of activity determination utilized the G2K software to obtain ratio of activity per nuclide identified. This determination was performed without the use of geometrically appropriate efficiency calibration curves. The ratio of activity per nuclide was used in conjunction with an overall exposure rate in mR/h obtained via a Fluke Biomedical hand-held ion chamber. The overall exposure rate was divided into individual nuclide amounts based on the G2K nuclide ratios. A gamma energy of 1 MeV and a gamma yield of 100% was assumed for all samples. Utilizing the gamma assumption and nuclide ratios, a calculation was performed to determine total sample activity in μCi (microCuries). An alternative method was proposed, which would eliminate the use of exposure rate and rely solely on the G2K software capabilities. The G2K software was energy and efficiency calibrated with efficiency curves developed for multiple geometries. The USGS reactor staff were trained to load appropriate calibration data into the G2K software prior to sample analysis. Comparison of the current method and proposed method demonstrated that the activity value calculated with the 1 MeV assumption could be as much as 3-4 orders of magnitude higher than the activity value established with the G2K software. The exposure rate calculation was also performed for each sample using actual gamma energies and yields to verify accuracy of the G2K software calibration. Facility specifications for detector dead time during sample analysis was stated to be 10% or less. Investigation of the effect of detector dead time on nuclide identification was performed. It was demonstrated that accurate nuclide identification could be performed with a detector dead time as high as 86.08% and a keV tolerance range of 1.5. A shielded lead cave was created to allow for greater source-to-detector distance. Additionally, an attenuator system was developed to aid in the reduction of detector dead time to meet facility specifications of less than 10%.