Fuzzy Reliability Analyzer: Quantitative Assessment of Risk Events in the Construction Industry Using Fuzzy Fault-Tree Analysis

Abstract

Fault trees are deductive techniques constructed by taking a system failure event and deconstructing it into its root causes (basic events, gate events). Fault trees can be solved qualitatively, by determining minimal cut sets, and quantitatively, by calculating the probability of occurrence of the risk event. In conventional fault-tree analysis (FTA), the probability of occurrence for all basic events must be assessed in order to allow for quantitative fault-tree analysis. However, conducting quantitative fault-tree analysis, especially in construction projects, entails several difficulties owing to the lack of sufficient data, leading to an approximation of the probability of occurrence for some basic events. Assuming probabilities for any basic event will add further uncertainty to the analysis, resulting in a potentially questionable end result. To overcome the challenge of assessing probabilities, this paper presents a comprehensive framework in which experts can use linguistic terms rather than numerals to assess the probability of occurrence of basic events. Fuzzy arithmetic operations are used to perform quantitative fault-tree analysis. Fuzzy Reliability Analyzer (FRA) was developed to automate both qualitative and quantitative FTA. The method presented is demonstrated via a case study to quantify the probability of failure of horizontal directional drilling (HDD) in meeting project objectives. Fourteen minimal cut sets were identified and the fuzzy probability (FPro) of the top event (TE) was calculated. The proposed approach offers the advantage of allowing experts to express themselves linguistically to assess the probability of occurrence of basic events, which is more appropriate for the construction domain. In addition, the proposed method offers the risk analyst the advantage of ranking basic events according to their level of contribution to the probability of the risk event, which can help in establishing more effective risk response strategies.