Explicit time-dependent analysis in Monte Carlo reliability simulations

Conference · Wed Dec 31 23:00:00 EST 1986 · Trans. Am. Nucl. Soc.; (United States)

OSTI ID:7118648

Lewis, E E; Tu, Z; Hald, U P; Boehm, F

Markov Monte Carlo methods have been employed for estimating reliability and availability of highly redundant systems with complex dependencies between component failure rates. There is, however, a major drawback with such Monte Carlo simulation when compared with analytical solutions or the numerical integration of the Markov equations. The result of a Monte Carlo simulation is a single number, the mission unreliability or unavailability (and an associated confidence interval). In contrast, deterministic methods provide curves of unreliability or unavailability versus time, allowing time-dependent trends to be observed and permitting a better and more general picture of system behavior to be obtained. To alleviate this drawback the authors have developed graphic algorithms that allow one to plot system unreliability and/or unavailability and the associated confidence intervals versus mission time. The procedures, moreover, add < 10% to the computation time required for a single Monte Carlo simulation to the maximum mission time to be considered.

• Research Organization: Northwestern Univ., Chicago, IL (USA)
• OSTI ID: 7118648
• Report Number(s): CONF-8711195-
• Journal Information: Trans. Am. Nucl. Soc.; (United States), Journal Name: Trans. Am. Nucl. Soc.; (United States) Vol. 55; ISSN TANSA
• Country of Publication: United States
• Language: English

Related Subjects

22 GENERAL STUDIES OF NUCLEAR REACTORS
220900* -- Nuclear Reactor Technology-- Reactor Safety
AGING
ALGORITHMS
AVAILABILITY
FAILURES
FUNCTIONS
LAGRANGIAN FUNCTION
MATHEMATICAL LOGIC
MONTE CARLO METHOD
NUCLEAR FACILITIES
NUCLEAR POWER PLANTS
POWER PLANTS
PROBABILITY
REACTOR COMPONENTS
REACTOR SAFETY
RELIABILITY
RISK ASSESSMENT
SAFETY
SAMPLING
THERMAL POWER PLANTS
TIME DEPENDENCE