THE EFFECTS OF USING UNSCREENED INDEPENDENT EVENTS ON PLANT SPECIFIC COMMON CAUSE FAILURE PROBABILITIES

摘要

It is generally very difficult and labor intensive to perform screening of independent events for a plant specific common cause failure (CCF) analysis because the huge number of events and lack of detailed information from currently available CCF data bases. In this study, the effects of using unscreened independent events on plant specific CCF probabilities were qualitatively estimated based on three plant specific CCF analyses. Plant specific CCF analyses for major components such as essential service water pumps, residual heat removal pumps, high pressure injection pumps, motor driven auxiliary feedwater pumps, emergency diesel generators, motor operated valves, air operated valves, check valves, batteries, and battery chargers were investigated. Based on the results of this study, it was concluded that the effects of using unscreened independent events on CCF probabilities would not be significant for those major components because of one or combination of the following reasons: (1) Most of the screened out CCF events would contribute little to independent events because their causes would always fail multiple components, (2) There was little need of screening independent event because few or no CCF events were screened out, (3) Screening of independent was not necessary because double counted events or mis-classified events were removed, (4) Independent events should not be screened out because the root cause, even though CCF events were screened out due to lack of coupling mechanism to cause multiple failures, still existed and could cause independent events, and (5) The screening criteria for the plant specific CCF analyses prevented CCF events due to causes which could contribute also to independent events from being screened out. For motor driven auxiliary feedwater (AFW) pump, even though CCF event screening itself fell into the above 5 cases, the independent events needed to be adjusted because it was not possible to separate out failures of mechanical pump parts of turbine driven AFW pumps from those motor driven AFW pumps. All mechanical pump parts independent events were conservatively excluded and only the pump driver (motor) independent events were used in estimating motor driven AFW pump CCF quantification parameters. Finally, it is recommended to perform qualitative assessment, like those performed in this study, for the effects of using unscreened independent event as a minimum in a plant specific CCF analysis when screening of independent events is very difficult to perform.