Evaluation and Comparison of Estimation Methods for Failure Rates and Probabilities

  • Jussi Vaurio
  • Kalle Jänkälä
Conference paper


This paper compares some features of Bayesian estimation techniques developed recently to assimilate data from multiple units to obtain effective estimators of failure rates for individual units. Several problems have been identified recently in various“two-stage” super-population methods [1, 2]. Hofer et al.[3] noticed that the results obtained with the method are sensitive to the order and ranges of multidimensional integration, and in the limit the results behave as if the components (or plants) were completely identical. Such lack of variation was also noticed in a common-cause failure study [4]. Becker and Schubert obtained suspicious results with the 2-stage method [5]. Meyer and Hennings [6] studied the impacts of different forms of the improper hyper-priors and integration limits. Hofer and Peschke re-formulated the method so that the population variability is better accounted for [7]. Bunea et al [8] claim that this approach still has some mathematical problems, and the choice of super-population and integration ranges are not always uniquely established. One problem with all these 2-stage-methods is that numerical methods are needed to calculate multidimensional integrals. This is a burden with hundreds of components in realistic risk assessment studies (PSA).


Standard Deviation Dirichlet Distribution Bias Term Gamma Density Multidimensional Integral 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Shultis J K et al, Report NUREG/CR-2374, USNRC, 1981Google Scholar
  2. 2.
    Kaplan S, IEEE Trans. Power App. Syst. 102, 195–202, 1983CrossRefGoogle Scholar
  3. 3.
    Hofer E et al., Risk Analysis 17(2) 249–252, 1997CrossRefGoogle Scholar
  4. 4.
    Bidwell D A, In YH, Proceedings PSA′99, 305–309, 1999Google Scholar
  5. 5.
    Becker G, Schubert B, Proceedings ESREL′98, 487-494, 1998Google Scholar
  6. 6.
    Meyer W, Hennings W, Proceedings ESREL′99, 893-898, 1999Google Scholar
  7. 7.
    Hofer E, Peschke J, Proceedings ESREL′99, 881-886, 1999Google Scholar
  8. 8.
    Bunea C et al., Two-stage Bayesian models — application to ZEDB project. Proceedings of ESREL 2003, vol. 1, 321-329. A.A. BALKEMA, Lisse, 2003Google Scholar
  9. 9.
    Bunea C et al., A non-parametric two-stage Bayesian model using Dirichlet distribution. Proceedings of ESREL 2003, vol. 1, 331-337. A.A. BALKEMA, Lisse, 2003Google Scholar
  10. 10.
    Vaurio J K, Risk Analysis, Vol. 7, No. 3, 329–338, 1987CrossRefGoogle Scholar
  11. 11.
    Jänkälä K E, Vaurio J K, Proceedings PSA′87, 281-286, 1987Google Scholar
  12. 12.
    Vaurio J K, Jänkälä K E, Effective Empirical Parametric Estimation of Failure Rates and Event Frequencies. Proceedings of PSAM5, vol. 4, 2143–2148. Universal Academy Press, Tokyo, 2000Google Scholar
  13. 13.
    Vaurio J K, Jänkälä K E, James-Stein Estimators for Failure Rates and Probabilities. Reliability Engineering and System Safety 36, 35–39, 1992 (also in SRE Symposium, Nyköping, 8-10 October 1990)CrossRefGoogle Scholar
  14. 14.
    Jänkälä K E, Vaurio J K, Proceedings PSA′93, 804-810, 1993Google Scholar
  15. 15.
    Lehto M, Jänkälä K, Mohsen B, Vaurio JK, Proceedings PSA′96, 507-514, 1996Google Scholar

Copyright information

© Springer-Verlag London 2004

Authors and Affiliations

  • Jussi Vaurio
    • 1
  • Kalle Jänkälä
    • 2
  1. 1.Fortum Power and Heat OyLoviisaFinland
  2. 2.Fortum Nuclear ServicesFortumFinland

Personalised recommendations