Fault Tree Analysis and Risk Assessment of Unbound Granular Pavement Material Performance

This paper describes the methodology issues considered in using Fault Tree An alysis to develop a risk assessment tool for unbound granular pavement material performance in Queensland. Example results from the tool are also presented. Granular pavements form a major geotechnical asset for Queensland Department of Main Roads (QDMR), and material underperformance is a significant concern. The uncertainty and economic consequences surrounding underperformance of granular pavements make them a prime candidate for structured risk assessment. The model was built using GoldSim, a simulation program that allows modelling of complex systems and quantitative probabilistic simulation. GoldSim is used to model a Fault Tree describing pavement performance factors and possible failure modes. The model uses semi-quantitative input ratings to describe the project scenario and the consequences for each failure mode, and outputs semi quantitative risk rankings for each failure mode. The model provides a relative measure of risk and is useful for comparing project scenarios and identifying high-risk areas for specific scenarios. The paper presents the model outputs for a particular scenario and compares them against actual performance.