Abstract
Traditional dam-break flood models were developed to predict flooding caused by water reservoir failures, using approaches such as physical erosion models, weir-based formulations, empirical breach equations, and comparative case analysis (USBR 1988). Although not originally developed for tailings dams, these methods are still widely used, often with limited adaptation, in tailings dam breach analysis (TDBA), now required under the global mining standard (ICMM 2020) following several catastrophic failures.
In 2021, the Canadian Dam Association (CDA) published a Technical Bulletin on Tailings Dam Breach Analysis (CDA 2021) to provide guidance for estimating the consequences of tailings dam failures. Prior to its release, limited guidance existed for conducting such analyses. The bulletin defines four representative TDBA cases based on the presence of supernatant water and the liquefaction potential of the tailings. These are grouped into two post-failure flow regimes: water-laden tailings flows (Cases 1A and 1B), characterized by hydraulic entrainment, and solids-dominated releases (Cases 2A and 2B), governed primarily by the mechanical failure of tailings.
While software tools such as HEC-RAS, FLOW-3D, or MADFLOW are mentioned in CDA (2021) as commonly used for TDBA, this paper proposes the Material Point Method (MPM), a numerical method suited for large-deformation problems, as an alternative for estimating runout distances in solids-dominated scenarios. MPM is based on continuum mechanics and is gaining attention in the geotechnical community for its ability to simulate post-failure behavior. Methodology is illustrated through three applications, including the runout assessment of a waste rock dump facility.
