Slope stability modelling of slurry dams through cone penetration data interpretation

This research entailed slope stability analysis of Orapa Mine Slurry Dam 2, located at Orapa in Botswana. The Finite Element (FE) Strength Reduction Technique was used and findings were compared with results from three Limit Equilibrium analysis techniques, namely Bishop Simplified, Spencer’s and Morgenstern-Price methods. The FE analysis was executed using the geotechnical engineering Rocscience software often referred to as Phase 2 or RS2 and the LE analysis was executed using the geotechnical engineering Rocscience software called Slide. The main purpose was to assess the geotechnical performance of the slurry dam. The LE and FE analysis results were observed to be in agreement with an average difference of 5.28%. The overall probability of failure of the slurry dam was zero and the overall reliability of the slurry dam has proven to be satisfactory with factors of safety for all slurry dam walls averaging at 1.65. The most important governing parameter in the stability of the dam is the friction angle. Furthermore, this study involved a regression analysis through supervised machine learning using MATLAB. This was to investigate any possible ways in which Cone Penetration Test (CPTu) measured variables (Cone Tip Resistance, Sleeve Friction, and Pore Pressure) could be predicted using depth as input into a function in a MATLAB workspace. The Gaussian Process Regression algorithm has responded well to the slurry dam data and produced good models.