System reliability in slope stability

System reliability in slope stabilityLast month’s issue of Geotechnique contained a paper by Reale, Xue and Gavin titled “System reliability of slopes using multimodal optimisation”

[http://www.icevirtuallibrary.com/doi/10.1680/jgeot.15.P.142] which we thought was excellent. The author’s present an algorithm for examining many modes of failure of slopes simultaneously stating, correctly, that for a given slope, two potential failure mechanisms might exist with similar probabilities of failure.

At AGEC we’ve been involved in many back-analyses of landslides and slope failures (check out some useful papers in our downloads section). Invariably, there are many processes involved in a slope stability failure but it is common that a localised specific mechanism will be governing. It is for this reason that we, at AGEC, while fans of probabilistic oversight on slope stability for asset maintenance, always recommend a site walkover and geomorphological study of the asset by an experienced engineer.

The experienced eye will pick up on the localised factor that contributes to instability which may be missed in a probabilistic or deterministic analysis despite best efforts. Appropriately, the paper above goes on to state that designers should consider the stability of the slope as a whole system rather than examining the critical slip plane in isolation –this is to account, for example, for a rainfall-induced reduction in near-surface suctions giving rise to a preferential shallow slip.

In summary, we love the detailed mathematics and novel approach outlined in the paper which would have real applications in practice, particularly for extensive geotechnical asset management but we would always advocate putting experienced eyes on the asset.