Abstract
Fragility functions express the probability that an asset exceeds some serviceability or limit state for a given level of environmental perturbation or other loadings, to which the asset is subjected. They are important components in the quantitative risk analysis of infrastructure exposed to natural hazards and they have typically been derived for structural assets. It is relatively difficult to derive fragility functions for geotechnical assets, such as highway or railway slopes and embankments, due to their inherent heterogeneity. In this paper, a generic granular highway embankment is modelled using the finite element method, considering various groundwater profiles and scour depths at the toe to quantify the deformation of the road surface. A probabilistic assessment of the magnitude of deformation and the groundwater level and scour depth is undertaken to derive fragility functions for the prediction of damage to assets exposed to these multiple hazards. The process of fragility function derivation is explained, uncertainty values are derived, and various regression methods are undertaken. This study is a first attempt to provide a basis for the prediction of slope deformation, and hence of damage, due to moisture ingress and scour, which can be aggravated by climate change. This can be used for the assessment of existing assets, and the design of new ones in the pursuit of more resilient transport networks, as well as for other assets such as levees, dams and other similar earthworks, with some limitations.