Study of Compton Suppression For Use in Spent Nuclear Fuel Assay

Open Access
Bender, Sarah Elizabeth
Graduate Program:
Nuclear Engineering
Doctor of Philosophy
Document Type:
Date of Defense:
December 06, 2013
Committee Members:
  • Kenan Unlu, Dissertation Advisor
  • Jack Brenizer Jr., Committee Chair
  • Igor Jovanovic, Committee Member
  • Joshua Alexander Robinson, Committee Member
  • Christopher Orton, Special Member
  • Jon Schwantes, Special Member
  • Spent fuel
  • Compton suppression
  • gamma spectroscopy
  • lanthanum bromide
  • burnup
  • Monte Carlo
  • Geant4
Nuclear material accountancy is of continuous concern for the regulatory, safeguards, and verification communities. In particular, spent nuclear fuel reprocessing facilities pose one of the most difficult accountancy challenges: continuously monitoring large volumes of highly radioactive, fluid sample streams. Current accountancy methods for nuclear fuel reprocessing facilities are resource intensive and time-consuming. The adaptation of passive gamma-ray detection coupled with multivariate analysis techniques could reduce the man-power requirements and processing time of samples. However, in measured gamma-ray spectra from spent nuclear fuel, the Compton continuum from the dominant 661.7 keV 137Cs fission product gamma-ray photon peak obscures lower energy lines. The application of Compton suppression to gamma-ray measurements of spent fuel will reduce the high continuum from nuclides like 137Cs and may allow other less intense, lower energy peaks to be detected, potentially improving the accuracy of multivariate analysis algorithms. There has been previous investigation into the use of room temperature detectors for gamma-ray spectroscopy of spent nuclear fuel. Interest has also been expressed in the application of Compton suppression to room temperature detectors for similar applications. Therefore, the focus of this study has been to assess Compton suppressed gamma-ray detection systems for the multivariate analysis of spent nuclear fuel. This objective has been achieved using direct measurement of samples of irradiated fuel elements in two geometrical configurations with Compton suppression systems. This allowed for the quantification of the number of additionally resolvable peaks through the application of Compton suppression and enabled the analysis of the effect of Compton suppressed gamma-ray detection in the presence high radiation field. A novel Compton suppressed detector model for the simulation of spent fuel measurements was developed. In order to address the objective to quantify the number of additionally resolvable photopeaks, direct Compton suppressed spectroscopic measurements of spent nuclear fuel in two configurations were performed: as intact fuel elements and as dissolved feed solutions. These measurements directly assessed and quantified the differences in measured gamma-ray spectrum from the application of Compton suppression. Several irradiated fuel elements of varying cooling time from the Penn State Breazeale Reactor spent fuel inventory were measured using three Compton suppression systems that utilized different primary detectors: HPGe, LaBr3, and NaI(Tl). The application of Compton suppression using a LaBr3 primary detector to the measurement of the current core fuel element, which presented the highest count rate, allowed four additional spectral features to be resolved. In comparison, the HPGe-CSS was able to resolve eight additional photopeaks as compared to the standalone HPGe measurement. Measurements with the NaI(Tl) primary detector were unable to resolve any additional peaks, due to its relatively low resolution. Samples of Approved Test Material (ATM) commercial fuel elements were obtained from Pacific Northwest National Laboratory. The samples had been processed using the beginning stages of the PUREX method and represented the unseparated feed solution from a reprocessing facility. Compton suppressed measurements of the ATM fuel samples were recorded inside the guard detector annulus, to simulate the siphoning of small quantities from the main process stream for long dwell measurement periods. Photopeak losses were observed in the measurements of the dissolved ATM fuel samples because the spectra was recorded from the source in very close proximity to the detector and surrounded by the guard annulus, so the detection probability is very high. Though this configuration is optimal for a Compton suppression system for the measurement of low count rate samples, measurement of high count rate samples in the enclosed arrangement leads to sum peaks in both the suppressed and unsuppressed spectra and losses to photopeak counts in the suppressed spectra. No additional photopeaks were detected using Compton suppression with this geometry. A detector model was constructed that can accurately simulate a Compton suppressed spectral measurement of radiation from spent nuclear fuel using HPGe or LaBr3 detectors. This is the first detector model capable of such an accomplishment. The model uses the Geant4 toolkit coupled with the RadSrc application and it accepts spent fuel composition data in list form. The model has been validated using dissolved ATM fuel samples in the standard, enclosed geometry of the PSU HPGe-CSS. The model showed generally good agreement with both the unsuppressed and suppressed measured fuel sample spectra, however the simulation is more appropriate for the generation of gamma-ray spectra in the beam source configuration. Photopeak losses due to cascade decay emissions in the Compton suppressed spectra were not appropriately managed by the simulation. Compton suppression would be a beneficial addition to NDA process monitoring systems if oriented such that the gamma-ray photons are collimated to impinge the primary detector face as a beam. The analysis has shown that peak losses through accidental coincidences are minimal and the reduction in the Compton continuum allows additional peaks to be resolved.