skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: State variable method of fault tree analysis

Technical Report ·
OSTI ID:5265359

The current technique of Fault Tree Analysis (FTA) generally employs computer codes that calculate the minimal cut sets of the Boolean function, where each cut set comprises basic initiator events (roots) whose intersection implies the occurrence of a TOP (system failure) event. Because the number of calculations can be very large for typical fault trees, the importance of any given cut set is assessed by qualitative algorithms that may include the number of basic events in the cut set, and quantitative importance algorithms that involve probabilistic upper and lower bound estimates, and the sets are culled before quantitative probability calculations are made. The assumption of statistical independence of all events is often imposed as a requirement for quantititative analysis. The question addressed in this paper is: can a tractable mathematical model be found that can perform quantitative calculations without the need of upper or lower bound simplifications and include within its structure the capability of handling common cause/common mode statistical dependence, failure mode coupling interdependence, and sequential failure time dependence. The Failure Mode State Variable (FMSV) model developed here is the affirmative answer to this question. The model is applied to simplified fault trees representing some nuclear power system components and subsystems considered for space electric power generation.

Research Organization:
Los Alamos National Lab., NM (USA)
DOE Contract Number:
W-7405-ENG-36
OSTI ID:
5265359
Report Number(s):
LA-UR-84-53; CONF-840113-5; ON: DE84006007
Resource Relation:
Conference: Symposium on space nuclear power systems, Albuquerque, NM, USA, 10 Jan 1984; Other Information: Portions are illegible in microfiche products
Country of Publication:
United States
Language:
English