Tailings Dam Performance : Modeling and Safety Analysis of a Tailings dam

Abstract: Storage and management of mine waste are both needed in the mining industry. After mineral extraction of the ore, there are generally leftovers with insufficient economical value that need to be taken care of. The finest grained fractions are referred to as tailings. Since every mine site and every tailings impoundment is unique, there is unfortunately not an universal answer to proper management that can be applied everywhere. Even though local guidelines and regulations can be considered to give a best practice in terms of design, there is correspondingly a need for dam safety stewardship on an operational level. Without such stewardship, not even the best designed dams or facilities would be fully controlled in terms of safety. Conversely, even badly designed dams can be operated in safe manners with good stewardship and surveillance programs. The coupling between design and stewardship is therefore important in order to reach proper tailingsmanagement.In the design of tailings dams, a certain value of the factor of safety for slopes of the dams is normally striven for to secure stability. The value is generally based on national regulations and/or guidelines. In Sweden the factor of safety should not be lower than 1.5 under normal conditions. In the guidelines, recommendations are often given on dam surveillance and field measurements of e.g. pore water pressure, deformations and seepage. Field measurements are taken, but are generally assessed in terms of trends (change with time) and not by comparison to anticipated performance.In this study, numerical modeling has been used for stability analyses and dam performance, as predictions of deformations and pore water pressure levels. An upstream tailings dam located in northern Sweden has been used as a case. The granular materials being part of the model based were described based upon geotechnical investigations (field and lab). The tailings material was modeled, on a constitutive level, by the Hardening Soil model. Good agreement betweensimulated behavior and laboratory tests was achieved. Other dam materials were simulated by the Mohr-Coulomb model.The model was built as a staged construction model where historical events between 1992 and 2013 were simulated. The historical events included dam raises, increased beach elevations, remedial works etc. The simulations of historical events were used for facilitating comparison with field measurements. By means of inclinometer data, horizontal deformations were measured and evaluated for a period of six years. These deformations were accurately simulated, which was considered to verify the numerical model. By this verification, the model is considered accurate enough to be used for simulating future events. Both stability and corresponding dam performance were computed, by simulating a period of 10 years. The stability analyses were used for the set-up of rockfill support plans, i.e. where and when remedial works are needed in order to maintain a certain safety level. The corresponding values in deformations and pore water pressures were used for the set-up of alert levels for each measuring unit. These alert levels will help the engineer in field with data interpretation, where the simulated values are compared with field measurements taken. The proposed methodology is recommended for tailings dams in general, which reduces the gap between design and stewardship. Hence, one step closer to proper tailings management is taken.