Bayesian Approach To Improve The Confidence Of The Estimation Of The Shear Strength Of Coarse Mine Waste Using Barton’s Empirical Criterion

Abstract

The evaluation of the shear strength of waste rock is required for the verification of the stability of high waste dumps, especially those that reach hundreds of meters in height. Mine waste rock material in open pit mining contains particles of metric scale which precludes the utilisation of commercial laboratory testing equipment. To overcome testing limitations, the shear strength of waste rock is frequently estimated using the empirical criterion of Barton-Kjærnsli. This criterion takes into consideration the nonlinearity of the shear strength envelope, characterising the behaviour of very coarse granular materials submitted to high loads. In the criterion, a stress-dependent structural component of the shear strength is parametrised with the equivalent roughness (R) and equivalent strength (S) and the structural component is added to the basic friction angle (φb) of the parental rock to determine the shear strength of the waste rock material. This paper demonstrates the use of Bayesian inference to determine the best set of parameters φb, R and S that satisfied both: large-scale laboratory testing results characterising a waste rock material, and reconciliation data from observations of stability of the waste dumps. The methodology allows the estimation of project-specific model parameters that honour both, laboratory data and site performance information. This objective is achieved through the estimation of correction factors to downgrade the strength from laboratory to field scale.