Multivariate analysis of gamma spectra to characterize used nuclear fuel
Abstract
The Multi-Isotope Process (MIP) Monitor provides an efficient means to monitor the process conditions in used nuclear fuel reprocessing facilities to support process verification and validation. The MIP Monitor applies multivariate analysis to gamma spectroscopy of key stages in the reprocessing stream in order to detect small changes in the gamma spectrum, which may indicate changes in process conditions. This research extends the MIP Monitor by characterizing a used fuel sample after initial dissolution according to the type of reactor of origin (pressurized or boiling water reactor; PWR and BWR, respectively), initial enrichment, burn up, and cooling time. Simulated gamma spectra were used in this paper to develop and test three fuel characterization algorithms. The classification and estimation models employed are based on the partial least squares regression (PLS) algorithm. A PLS discriminate analysis model was developed which perfectly classified reactor type for the three PWR and three BWR reactor designs studied. Locally weighted PLS models were fitted on-the-fly to estimate the remaining fuel characteristics. For the simulated gamma spectra considered, burn up was predicted with 0.1% root mean squared percent error (RMSPE) and both cooling time and initial enrichment with approximately 2% RMSPE. Finally, this approach to automated fuelmore »
- Authors:
-
- Univ. of Tennessee, Knoxville, TN (United States). Dept. of Nuclear Engineering
- Pacific Northwest National Lab. (PNNL), Richland, WA (United States)
- Publication Date:
- Research Org.:
- Pacific Northwest National Laboratory (PNNL), Richland, WA (United States); Univ. of Tennessee, Knoxville, TN (United States)
- Sponsoring Org.:
- USDOE Office of Nuclear Energy (NE), Fuel Cycle Technologies (NE-5); USDOE
- OSTI Identifier:
- 1342225
- Alternate Identifier(s):
- OSTI ID: 1416187
- Grant/Contract Number:
- AC05-76RL01830
- Resource Type:
- Journal Article: Accepted Manuscript
- Journal Name:
- Nuclear Instruments and Methods in Physics Research. Section A, Accelerators, Spectrometers, Detectors and Associated Equipment
- Additional Journal Information:
- Journal Volume: 850; Journal ID: ISSN 0168-9002
- Publisher:
- Elsevier
- Country of Publication:
- United States
- Language:
- English
- Subject:
- 46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND TECHNOLOGY; Used nuclear fuel; Fuel characterization; Multivariate analysis; Gamma spectroscopy
Citation Formats
Coble, Jamie, Orton, Christopher, and Schwantes, Jon. Multivariate analysis of gamma spectra to characterize used nuclear fuel. United States: N. p., 2017.
Web. doi:10.1016/J.NIMA.2017.01.030.
Coble, Jamie, Orton, Christopher, & Schwantes, Jon. Multivariate analysis of gamma spectra to characterize used nuclear fuel. United States. https://doi.org/10.1016/J.NIMA.2017.01.030
Coble, Jamie, Orton, Christopher, and Schwantes, Jon. 2017.
"Multivariate analysis of gamma spectra to characterize used nuclear fuel". United States. https://doi.org/10.1016/J.NIMA.2017.01.030. https://www.osti.gov/servlets/purl/1342225.
@article{osti_1342225,
title = {Multivariate analysis of gamma spectra to characterize used nuclear fuel},
author = {Coble, Jamie and Orton, Christopher and Schwantes, Jon},
abstractNote = {The Multi-Isotope Process (MIP) Monitor provides an efficient means to monitor the process conditions in used nuclear fuel reprocessing facilities to support process verification and validation. The MIP Monitor applies multivariate analysis to gamma spectroscopy of key stages in the reprocessing stream in order to detect small changes in the gamma spectrum, which may indicate changes in process conditions. This research extends the MIP Monitor by characterizing a used fuel sample after initial dissolution according to the type of reactor of origin (pressurized or boiling water reactor; PWR and BWR, respectively), initial enrichment, burn up, and cooling time. Simulated gamma spectra were used in this paper to develop and test three fuel characterization algorithms. The classification and estimation models employed are based on the partial least squares regression (PLS) algorithm. A PLS discriminate analysis model was developed which perfectly classified reactor type for the three PWR and three BWR reactor designs studied. Locally weighted PLS models were fitted on-the-fly to estimate the remaining fuel characteristics. For the simulated gamma spectra considered, burn up was predicted with 0.1% root mean squared percent error (RMSPE) and both cooling time and initial enrichment with approximately 2% RMSPE. Finally, this approach to automated fuel characterization can be used to independently verify operator declarations of used fuel characteristics and to inform the MIP Monitor anomaly detection routines at later stages of the fuel reprocessing stream to improve sensitivity to changes in operational parameters that may indicate issues with operational control or malicious activities.},
doi = {10.1016/J.NIMA.2017.01.030},
url = {https://www.osti.gov/biblio/1342225},
journal = {Nuclear Instruments and Methods in Physics Research. Section A, Accelerators, Spectrometers, Detectors and Associated Equipment},
issn = {0168-9002},
number = ,
volume = 850,
place = {United States},
year = {Tue Jan 17 00:00:00 EST 2017},
month = {Tue Jan 17 00:00:00 EST 2017}
}
Web of Science
Figures / Tables:
Figures / Tables found in this record: