Reliability Analysis of Tailings Dams: A Case Study in Jiangxi Province, China

A tailings dam is an artificial embankment made of special soil materials derived mainly from mine tailings. Because of the artificial and unique distribution of the soil layers with significant uncertainties of physical-mechanical properties, stability assessments by routine deterministic methods are inadequate. To better assess the stability of tailings dams, a new reliability analysis model combined with support vector machine (SVM) and shear strength reduction (SSR) techniques is proposed. First, an orthogonal design is applied to generate high-quality training samples according to the variability of soil layers. Second, the factors of safety for these samples are obtained by SSR techniques based on the numerical modeling code, FLAC3D. Third, the SVM is applied to deduce the explicit solution of the response surface for reliability analysis. Fourth, a Monte Carlo Simulation (MCS) is performed based on the explicit performance function. The Yongpin tailings dam in Jiangxi Province, China is used as a case history to demonstrate the effectiveness and promise of the proposed methodology.