An analysis of check valve performance characteristics based on valve design

Although it had been theorized by nuclear industry valve experts that the two most significant factors in assessing check valve performance were valve type (or design) and operating conditions, until recently, no data was available to support their assumptions. In co-operation with the Nuclear Industry Check Valve Group (NIC), Oak Ridge National Laboratory (ORNL) undertook a review and analysis of check valve failures recorded in the Institute of Nuclear Power Operations' (INPO) Nuclear Plant Reliability Data System (NPRDS). This study involved the characterization of failures according ro several parameters, including valve design (e.g. swing check, lift check). Since the valve design is not inherently included within the NPRDS engineering record for each component in the database, ORNL relied on input from NIC, valve manufacturers and catalogs to supply the missing information. As a result, nearly 60% of the 21 000 check valves listed in the NPRDS component database and 85% of the 838 failures occurring during 1991-1992 were identified according to valve design. This data provided the basis to perform previously unavailable cross-correlations between parameters such as valve design versus failure mode, valve design versus failure discovery method, population/failure distributions by valve design, etc. Performance assessments and predictions based on more specific sets of parameters (as opposed to generic check valve failure rates obtained from standard reference sources that generally ignore the valve design) should result in a significant impact on future nuclear plant operations, including inservice testing (IST) practices, maintenance, and probabilistic risk assessments (PRAs) by providing a means to calculate more appropriate relative (and ultimately absolute) failure rates for check valves. (C) 2000 Elsevier Science S.A. All rights reserved.