Interim Report on Systematic Errors in Nuclear Power Plants; Seismic Safety Margins Research Program

摘要

This report is an investigation of systematic errors in nuclear power plants. A review of the Licensee Event Reports (LERs) for Zion-1 and 2 (1974-April 1979) and a study of 100 design errors compiled by the Oak Ridge National Laboratory (ORNL) for the period 1977-78 are carried out. Two general classes of errors, random and systematic, are defined and the latter is analyzed in detail throughout the study. Systematic errors are studied from various points of view, namely, their causes, effects, mode of discovery, and their relation to fragility curves. A review of the data reveals that wrong analysis, specification errors and unforeseen environmental conditions (subcategories of design errors) and operational procedure errors which affect the performance of components (component systematic errors) dominate and they are mostly discovered by tests. Systematic errors affecting system function and reliability (systemic systematic errors) are mostly found by reviews of design and analysis. This study shows that a large portion of systemic errors are common to both Zion -1 and 2 units and they are mainly caused by wrong analysis. The data also show that systematic errors which affect the fragility curve constantly or the lower part of the fragility (Category I systematic errors) dominate. Systematic errors which cause failure under abnormal conditions and affect the rising part and the upper tail of the fragility curve (Category II systematic errors) are mainly caused by design errors and they are systemic in nature.