Slope stability analysis using deterministic and probabilistic approaches for poorly defined stratigraphies
Ghadrdan, M and Dyson, AP and Shaghaghi, T and Tolooiyan, A, Slope stability analysis using deterministic and probabilistic approaches for poorly defined stratigraphies, Geomechanics and Geophysics for Geo-Energy and Geo-Resources, 7 Article 4. ISSN 2363-8419 (2021) [Refereed Article]
The stability of slopes directly affects human lives, the environment, and the economy. Inaccurate geological profiles within numerical slope stability models can lead to potentially catastrophic consequences when model conditions do not appropriately reflect real-life stratigraphy. In cases where localised deposits are prevalent, probabilistic methods are often necessary to accommodate for unknown or poorly defined stratigraphies. Currently, there are no commercial geotechnical software packages that simulate probabilistic constitutive behaviour of materials within finite element methods for large-scale stratigraphic analysis. Instead, commercially available probabilistic methods such as the random limit equilibrium method are incapable of incorporating non-linear constitutive soil behaviour. For this reason, advanced constitutive models are seldom coupled with probabilistically varying soil layers or spatially variable soil parameters. The objective of this research is the implementation of a simplified method for probabilistic stratigraphic analysis within a commercially available FE environment, providing a technique to assess the effects of stratigraphic uncertainty on slope stability. The proposed method is presented, highlighting the impact of localised thin layers of soft material as well as their frequency and location on the slope of an operational open-pit mine. The significance of these stratigraphic effects is presented through a case study of Australia’s second-largest open-pit mine, at which the stability of a collapsed coal slope is analysed. To improve the reliability of the finite element method for slope stability assessment, the Monte Carlo approach has been incorporated to consider varying shear strength distributions for models incorporating advanced constitutive behaviour. Thicker probabilistically generated deposits of silty material resulted in increased slope Factors of Safety. Similarly, greater proportions of silty deposits within a predominantly clayey interseam produced larger safety factors than slopes without localised thin silty layers. Stratigraphic analysis indicated that the Factor of Safety was most sensitive to localised silt layers at depths greater than 83 m below ground level.
coal, slope stability, deterministic analysis, random field analysis, factor of safety