Abstract

The new legislation in Brazil has brought more attention to the technique of stacking filtered iron ore tailings for mining waste disposal. To minimize the risk of liquefaction, it's important to maintain the compacted structure in an unsaturated condition. This requires knowledge of the hydraulic characteristics of the material in order to evaluate the long-term performance of the embankment and design an adequate monitoring system. One way to track the material's behavior in the field is through electrical resistivity techniques. Using an electrical resistivity measurement system at various frequencies and current levels, we evaluated the behaviour of compacted iron ore waste. The examined tailings have been deposited in filtered tailings piles, which are intended to exceed a height of 100 meters. This study aims to investigate the electrical response of the material to different levels of water content, degree of saturation and suction. It was possible to establish good correlations between electrical resistivity and factors such as suction, degree of saturation, and water content. It was also identified an association between the electrical resistivity and the optimum water content

Abstract

Chile is the third country with most tailing dams worldwide (764, of which 15% active, 62% inactive, and 23% abandoned). Today, one of the main challenges of the mining industry is to ensure environmental sustainability. To achieve this, it is not only necessary to have regulations guaranteeing correct design, maintenance, closure stages, but also supervision and suitable technology enabling rapid, reliable, and cost-effective diagnosis of the overall tailings dam stability. For the last 20 years, French instrumented dynamic cone penetrometer, P.A.N.D.A®, has been used for in-depth quality control of tailings deposits, mainly due to its rapidity, portability, easy-to-use, cost-effective, and environmentally friendly testing which is based on standardized methods and able to produce immediate results. The last few years, different studies have been carried out to characterize tailings dams from a geotechnical perspective as well as to assess their spatial and temporal variability. This article presents a summary of the 20 years of Chilean experience in compaction quality control and geotechnical characterization of mine waste facilities using P.A.N.D.A® to assess the overall stability, slope analysis stability, liquefaction triggering, spatial variability of material properties and evaluation of some geotechnical parameters. Indeed, given the scale of the problem these structures represent for the community and the environment, it seemed necessary to bring together most of the work carried out in Chile to propose an innovative alternative to the rest of the mining community.

Abstract

The state of the practice to evaluate the dynamic liquefaction potential of a soil column entails the use of simplified methods that compares the cyclic stress ratio with the cyclic resistance ratio. One of the most used methods is the Boulanger & Idriss 2014, which relies on cone penetration test data to estimate in-situ cyclic resistance ratio and cyclic stress ratio distributions, considering corrections factors for soils with fine contents. Saye, Olson and Franke in 2021 presented a novel method to assess liquefaction susceptibility using cone penetration test data on soils ranging from non-sensitive clays to clean sands. The procedure was developed using +400 documented case records of liquefaction and non-liquefaction in clean sands, silty sands, sandy silts, and low plasticity fine grained soils. Although promising, this method is not widely used in the industry yet. This paper presents a comparison between both methods for tailings. It uses cone penetration test soundings performed in tailings with a wide range of mineralogy and fine contents, combined with variations in peak ground accelerations and magnitudes. This analysis shows that Saye, Olson and Franke’s method is generally more robust, particularly for the analysis of fine tailings, as Boulanger & Idriss’ method relies on site specific data or the engineer’s judgement to define an Ic cut-off value that screens out clay-like soils from the liquefaction analysis.

Abstract

Athree-dimensional (3D) ground model was developed to design the stabilisation of a dam founded on both weak and liquefiable units up to about 18 m below ground surface. The ground model covers a linear extent of approximately 800 m and was developed from five separate site investigations completed over a four-year period and digitisation/georeferencing of historic drawings/plans. Combined, the investigation comprised 206 cone penetration tests (CPTs), 37 boreholes and 36 test pits, including several vane shear tests, ball penetrometer tests and sampling. CPT data was processed to identify different material behaviours, generally based on the following features: corrected tip resistance; sleeve friction resistance; pore water pressure ratio; state parameter; and the soil behaviour type index. Each CPT interpretation was compared with information from the nearest borehole using a purpose-built python code. This information was reviewed manually in an iterative process to delineate the various geotechnical unit based on CPT response and the physical logs. This process identified a continuous weak organic layer across the site which had not been previously picked-up by the Engineer of Record (EoR). The works identified nine separate geotechnical units, with one of these subdividable based on its CPT response/grain size. The 3D ground model was built in Seequent Leapfrog Geo using the following information: unit levels specified from each CPT, digitised historic drawings/plans and a topographic survey. This paper describes the process of development and presents the full 3D ground model used as critical input to the stabilisation design of the dam