A method for assessing the long-term integrity of tailings dams

Abstract

Mine tailings are a by-product of the processing of minerals. At most mines they are a waste product that needs to be managed. Tailings composition and properties vary widely and are in most cases highly erodible due to their fine particle size and can contain elevated concentrations of unwanted minerals and process chemicals. Therefore, if released to the environment they can be a significant environmental problem. A common management strategy is to store them in 'tailings dams' where they will remain in perpetuity. Little work has been done to assess the long-term erosional behaviour of tailings dams. Computer based Landscape Evolution Models (LEMs) provide information on erosion rates, type of erosion and where erosion is likely to occur. They can therefore provide guidance on long-term behaviour which allows designs to be tested and improved. Here a LEM, SIBERIA, is used to assess two hypothetical tailings dam designs using different surface covers and climates. The results suggest that a tailings dam that can capture rainfall can erode less than a capped design that must shed any runoff. An embankment with a small and steep catchment has minimal erosion potential and any material eroded from the internal wall of the embankment is deposited internally and provides erosion protection. If the external embankment is maintained then there is potential for long-term encapsulation of tailings. The single biggest issue for the employment of LEMs is that of parameterisation and here assumes (1) a uniform and consistent armour or (2) a consistent and self-sustaining vegetation cover. The modelling and methods here provide a template for tailings dam assessment at other sites globally, and will improve tailings dam design and reduce environmental risk.