Velarde, G & Macciotta, R 2023, 'Probabilistic pit slope stability analysis targeting a reliability-based design acceptance criteria: a parametric study', in PM Dight (ed.), SSIM 2023: Third International Slope Stability in Mining Conference, Australian Centre for Geomechanics, Perth, pp. 161-184, https://doi.org/10.36487/ACG_repo/2335_08
(https://papers.acg.uwa.edu.au/p/2335_08_Velarde/)
Abstract: Design of open pit slopes is a decision-making process which aims to maximise the ore recovery while minimising the excavation volumes. The current practice of designing open pit slopes adopts the widely accepted Guidelines for Open Pit Slope Design by Read & Stacey (2009). An optimum design should satisfy a design acceptance criteria (DAC). However, designing open pit slopes is a complex process that involves inherent risks and uncertainties. As a result, reliability analyses are becoming increasingly important for performance-based slope designs. In a reliability approach, the amount of information on the slope materials and behaviour would reflect the reliability of a slope design. This paper presents a parametric study defined by the uncertainties of the rock mass strength properties and the slope geometric configurations at three different design reliability levels targeting a reliability-based DAC (RBDAC). The reliability assessment is performed using probabilistic analysis adopting the two-dimensional limit equilibrium method and Monte Carlo simulations. The input variables for the rock mass strength are defined through probability density functions (PDF) that capture the natural variability, while the input variables of geological structures are defined through kinematic assessments.
The PDFs of the rock mass strength properties were modelled based on the generalised Hoek–Brown criterion using the mean, coefficient of variation (COV) and dependence between quantitative properties of the criterion. Results show that most of the resultant pairs of Factor of Safety (FoS) and probability of failure and associated COV of the resulting FoS (COVFoS) are consistent with the RBDAC. Based on this, a redesign is proposed showing the applicability of the RBDAC and comparing it to the current DAC. This approach has significant implications for slope optimisation or mitigation plans for future pushbacks in case of instabilities. 

Keywords: reliability-based design acceptance criteria, coefficient of variation, probability density functions