Technical Session 2.3B - Analysis & Design 4
Abstract
In recent years, the use of filtered dry stacked tailings has steadily increased in the mining industry since it is an innovative solution for the construction of tailings storage facilities (TSF). This technology increases the safety of tailings dams, enhances the recovery of water and enables the construction of co-disposal techniques mixing filtered tailings and waste rock. However, unlike conventional tailings dams, the run-out distance in case of a breach cannot be evaluated with inundation methods because the material is partially saturated and will not flow as it is assumed in most hydraulic models. Hence, other techniques must be used to assess the run-out distance of filtered tailings facilities. The material point method (MPM) is a particle-based method that represents material as a collection of material points. This approach is particularly useful in the context of large deformations where mesh distortions often produce unrealistic or undesirable results. The run-out distance in case a filtered tailings facility fails involve very large deformations and therefore the MPM is suitable to solve the problem. The MPM technique is implemented in the software Anura 3D and was applied to an existing filtered dry stack in the Peruvian Andes to predict run-out distances in a 2D framework. Geotechnical characterization was carried out from advanced laboratory and in-situ testing for the tailings and the foundation materials which was used to estimate Mohr-Coulomb parameters. Results show the applicability of MPM for run-out problems for different failure scenarios giving results from 20 m to 70 m. These distances can be applied for geohazard risk management of filtered tailings facilities.
Authors
- Brahian Roman, Staff Consultant (Geotechnical Engineering) | SRK Peru
- Anderson Barreto, Staff Consultant (Geotechnical Engineering) | SRK Peru
- Martín Villanueva, Senior Civil Engineer, Practice Leader | SRK Peru
Click here to see other SRK presentations at this event.
