Fault tree analysis as a quantitative hazard analysis with a novel method for estimating the fault probability of microbial contamination: A model food case study

Abstract

A novel method was developed to estimate the fault probability of basic events in fault tree analysis (FTA) for microbial contamination. A simple food preparation process was used for the analysis. The process consisted of mixing and cutting. The fault probability (P-F) was for the occasion at excess level of E. coli compared to a threshold level. The basic events were identified as faults in sanitary handling of the hands, mixing bowl, knife, and cutting board. The top event was the ultimate fault, i.e. contamination of a model food (cotton cloth). P-F of the basic events was estimated, using normal distribution functions established on the basis of contamination levels, which were determined by microbiological analysis. The area under the normal distribution function to the right of a threshold level, which was the contamination level required to cause a hazard, was regarded as P-F. Finally, the P-F of the top event could be estimated from P-F of the basic events by FTA quantitative analysis. The estimated P-F of the model food proved to agree with that measured experimentally.