A general method dealing with correlations in uncertainty propagation in fault trees

https://doi.org/10.1016/0951-8320(89)90013-6Get rights and content

Abstract

This paper deals with the correlations among the failure probabilities (frequencies) of not only the identical basic events but also other basic events in a fault tree. It presents a general and simple method to include these correlations in uncertainty propagation. Two examples illustrate this method and show that neglecting these correlations results in large underestimation of the top event failure probability (frequency).

References (15)

There are more references available in the full text version of this article.

Cited by (30)

  • Designing a new procedure for participation of prosumers in day-ahead local flexibility market

    2023, International Journal of Electrical Power and Energy Systems
  • Optimizing resource allocations to improve system reliability via the propagation of statistical moments through fault trees

    2023, Reliability Engineering and System Safety
    Citation Excerpt :

    A recent study concerning the uncertainty analysis and the propagation of the uncertainty in FTs defines different stages of knowledge where the aleatory and the epistemic uncertainties are represented by frequentist probability and uncertainty theory [40]. From a quantitative perspective, fault trees are investigated for the influence of sequential failure logic [41], repairable components and mission phases [42], mutually exclusive events [43], mutually dependent events [44,45], rare events [46], dependability analysis via bayesian networks [47], and correlations in uncertainty propagation [33]. Some recent literature about the reliability analysis of FTs include the irrelevance coverage models for dynamic fault trees via algebraic structure functions [48].

  • Development of network-based probabilistic safety assessment: A tool for risk analyst for nuclear facilities

    2019, Progress in Nuclear Energy
    Citation Excerpt :

    Some of the limitations of PSAs which adopt the standard FT analysis can be overcome by combining the FT with other techniques, or by employing an alternative approach. For example, the statistical dependencies between events can be handled by introducing correlation coefficients within the FT (Zhang, 1989; Fleming and Mikschl, 1999; Ebisawa et al., 2015). Alternatively, the concept of a Bayesian Network (BN) can also be employed.

  • Application of Bayesian statistics to seismic probabilistic safety assessment for research reactor

    2018, Nuclear Engineering and Design
    Citation Excerpt :

    For example, the uncertainties in basic event probabilities are considered by implementing an FTA along with Monte-Carlo simulation (USNRC, 1975), Latin hypercube sampling (Ellingwood, 1990; Kim et al., 2011; Kwag and Ok, 2013) or Fuzzy set theory (Tanaka et al., 1983; Singer, 1990; etc.). The statistical dependencies between events are handled by introducing the correlation coefficients (Zhang 1989; Fleming and Mikschl, 1999; Ebisawa et al., 2015) or utilizing a binary decision diagram or using a common cause failure analysis within a fault tree. The concept of a Bayesian network can manage not only the logic gate relationships but also various statistical dependencies that cannot otherwise be represented by the standard fault tree.

  • Probabilistic energy consumption analysis in buildings using point estimate method

    2018, Energy
    Citation Excerpt :

    In various applications, several approximate methods have been used to model uncertainties. Truncated Taylor series expansion method [32]; the discretization method [33]; the common uncertain source method [34]; the first-order second-moment method (FOSMM) [35]; the Cumulant method [36]; and the point estimate method (PEM) [37,38]. The main idea behind these methods is to use approximate principles for calculating the statistical moments of a random quantity that is a function of random variables, opposing to the computationally demanding MCS approach.

View all citing articles on Scopus

This work was carried out whilst the author was at JBF Associates, Inc., Knoxville, Tennessee.

The author is a visiting scholar at Mechanical, Aerospace and Nuclear Engineering Department, University of California, Los Angeles.

View full text