Technical Review of On-Line Monitoring Techniques for Performance Assessment: Limiting Case Studies (NUREG/CR-6895, ORNL/TM-2007/188, Volume 3)
On this page:
Download complete document
Manuscript Completed: October 2007
Date Published: August 2008
J.W. Hines, J. Garvey, D.R. Garvey, and R. Seibert
Department of Nuclear Engineering
The University of Tennessee-Knoxville
Knoxville, TN 37966-2210
Oak Ridge National Laboratory
Managed by UT-Battelle LLC
Oak Ridge, TN 37831-6156
S.A. Arndt, NRC Project Manager
Division of Engineering Technology
Office of Nuclear Regulatory Research
U.S. Nuclear Regulatory Commission
Washington, DC 20555-0001
NRC Job Code N6080
Traditionally, the calibration of safety-critical nuclear instrumentation has been performed at each refueling cycle. However, many nuclear plants have expressed a desire to move toward condition-directed rather than time-directed calibration. This condition-directed calibration is accomplished through the use of on-line monitoring (OLM).
With a sound OLM system in place, nuclear plants may be able to extend the required calibration interval. Only recently have nuclear plants come to the point where they are ready to apply for license amendments to extend their calibration frequency. To help support the regulatory review of these amendments, researchers from The University of Tennessee's Nuclear Engineering department were contracted to draft a NUREG/CR series. The goal of this entire NUREG/CR series is to provide guidance to the regulatory review process of OLM. The volumes present the technical background of OLM and explain most of the theory behind the OLM concept. Additionally, they discuss and analyze specific issues regarding the application of OLM in a nuclear power plant. The first volume in this series, NUREG/CR-6895, entitled Technical Review of On-line Monitoring Techniques for Performance Assessment, Volume 1: State-of-the-Art, was completed in March 2005 and published in January 2006. Volume 1 offers a general overview of current sensor calibration monitoring technologies and their uncertainty analysis, a review of the supporting information necessary for assessing these techniques, and a cross-reference between the literature and the requirements listed in the SER. Volume 2: Theoretical Issues presents an in-depth theoretical study and independent review of the most widely used OLM techniques It includes a presentation of the theory and further explanation of the assumptions inherent in the empirical models and the uncertainty quantification techniques.
This third and final volume summarizes seven case studies investigating the effects of model development and assumptions on model performance. Two case studies concern the effect of not meeting model assumptions: evaluating query data outside the training region and training with faulty data. Recommendations are given for identifying and correcting the problems caused by not meeting these important data assumptions. The third and fourth case studies investigate the effects of high noise levels on model performance and compare different methods of data denoising, respectively. The remaining three case studies examine different features of model development by comparing vector selection methods, different numbers of memory vectors, and robust distance measures. Methodologies to determine the appropriate model development parameters for each of these cases are outlined. Finally, a section is included that highlights the special considerations needed for redundant-sensor model architectures. Although this study is not an exhaustive review of the many issues in OLM system development, it provides a base set of considerations that must be accounted for and a method for testing these considerations with other model architectures.