Efficiency of Analytical Methodologies in Uncertainty Analysis of Seismic Core Damage Frequency

Abstract

Fault Tree and Event Tree analysis is almost exclusively relied upon in the assessments of seismic Core Damage Frequency (CDF). In this approach, Direct Quantification of Fault tree using Monte Carlo simulation (DQFM) method, or simply called Monte Carlo (MC) method, and Binary Decision Diagram (BDD) method were introduced as alternatives for a traditional approximation method, namely Minimal Cut Set (MCS) method. However, there is still no agreement as to which method should be used in a risk assessment of seismic CDF, especially for uncertainty analysis. The purpose of this study is to examine the efficiencies of the three methods in uncertainty analysis as well as in point estimation so that the decision of selecting a proper method can be made effectively. The results show that the most efficient method would be BDD method in terms of accuracy and computational time. However, it will be discussed that BDD method is not always applicable to PSA models while MC method is so in theory. In turn, MC method was confirmed to agree with the exact solution obtained by BDD method, but it took a large amount of time, in particular for uncertainty analysis. On the other hand, it was shown that the approximation error of MCS method may not be as bad in uncertainty analysis as it is in point estimation. Based on these results and previous works, this paper will propose a scheme to select an appropriate analytical method for a seismic PSA study. Throughout this study, SECOM2-DQFM code was expanded to be able to utilize BDD method and to conduct uncertainty analysis with both MC and BDD method.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 15 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.