Loss detection results on simulated tank data modified by realistic effects

Burr, T., Hamada, M.S., Howell, J. and Suzuki, M. (2012) Loss detection results on simulated tank data modified by realistic effects. Journal of Nuclear Science and Technology, 49(2), pp. 209-221. (doi: 10.1080/00223131.2011.649076)

Abstract

Solution monitoring (SM) is a type of process monitoring (PM) intended to improve nuclear safeguards in large commercial facilities that contain solutions. Typically, masses (M) and volumes (V) are estimated from frequent in-process measurements. Transfers between tanks can be identified in these data, segments of which can then be compared to generate transfer differences (TDs). A safeguards concern might then be raised if either these TDs or deviations in M or V data during “wait” modes become significant. Average M and V TDs should be 0 (perhaps following a bias adjustment) to within a historical limit that is a multiple of the standard deviation of the M or V TD, as should deviations during “wait” modes. Statistical test options can be compared on the basis of their estimated probabilities to detect various material loss scenarios. Multivariate statistical PM options have previously been applied to residuals produced from simulated SM data that had no process variation, only random and systematic measurement errors. This article examines how detection probabilities might be estimated with real data. To do this, realistic effects such as pump carryover, evaporation, condensation, and mixing/sparging, are included in simulated data. In real facilities false alarms are a concern, particularly when data need to be evaluated regularly, on a day-to-day basis. The need to widen control limits to avoid alarming on innocent process variation effects is discussed, and the consequential reduction in DPs is illustrated with numerical examples.

• Item Type: Articles
• Status: Published
• Refereed: Yes
• Glasgow Author(s) Enlighten ID: Howell, Dr John
• Authors: Burr, T., Hamada, M.S., Howell, J., and Suzuki, M.
• College/School: College of Science and Engineering > School of Engineering > Systems Power and Energy
• Journal Name: Journal of Nuclear Science and Technology
• ISSN: 0022-3131
• ISSN (Online): 1881-1248
• Published Online: 02 February 2012