The high demand for cobalt in the global energy transition is driving the mining industry to explore alternative and more ethical cobalt sources. Reprocessing of copper flotation tailings is a potential circular economy approach to managing mine waste and supplementing global critical minerals supply. Tracking cobalt deportment from primary ore, through the concentrator and into tailings is a critical first step in assessing the valorisation potential. The mineralogy, particle liberation properties and mineral associations strongly influence the recoverability of cobalt from tailings. This research integrated Micro-X-Ray Fluorescence (mu XRF), chemical assay, Mineral Liberation Analysis (MLA) and Advanced Mineral Identification and Characterization System (AMICS) to geometallurgically track cobalt deportment throughout an operating copper flotation circuit and into the fresh tailings. Cobalt was found to be hosted in several mineral phases, including cobaltiferous sulphides and pyrite. Detailed geometallurgical characterisation and deportment studies will assist in identifying mineral processing pathways to recovering value from copper tailings.

As high-grade ore deposits become depleted, there is an increasing demand to process marginal low-grade ores with complex metallurgical challenges. Orebodies can be rich in pyrite, which is the most abundant sulphide gangue mineral in ore deposits. Pyrite has been depressed to mining tailings for decades. Although these mine tails are an ongoing environmental liability, they also present an opportunity as they may contain valuable metals such as Au, Ag, Co and Ni.Pyrite-rich orebodies present challenges to current mining prospects as they tend to be non-responsive to traditional methods for pyrite depression. If not depressed, pyrite floats along with valuable minerals producing a low-quality concentrate that causes significant problems for smelting operations. Furthermore, the necessity for higher pyrite rejection causes projects and operations to resort to complicated and often inefficient processing flowsheets. These process developments can be economically, operationally, and environmentally challenging. Thus, understanding why pyrite's flotation performance varies within orebodies and the features affecting its floatability is critical for an efficient depression and possible reclamation of pyrite.Natural pyrite can exhibit different electrochemical properties, chemical composition, and texture depending on the ore genesis. These variations have been reported to affect pyrite's flotation response at different levels, though the mechanisms are not well understood. Different pyrite textures may influence flotation by altering the electrochemical properties of pyrite and the ways it interacts with the pulp, or they can introduce naturally floatable species within pyrite grains. This review examines why pyrite variations show differentiated flotation behaviour and identifies knowledge gaps from experimental and observational studies. The paper focuses on the effect of texture and composition of pyrite on electrochemical properties and floatability. The work investigates pyrite textures at Mount Isa Mines and their observed effect on flotation performance.

Cobalt production is essential in supporting energy storage and electrification initiatives in the global transition to a low-carbon economy. The main aspects of cobalt production are examined through assessment of global demand and supply, the common geological settings and mineral processing routes, and the associated situated environmental, social and governance (ESG) risk factors. The supply risk of cobalt is high, owing to its primary extraction as a by-product during copper and nickel production from several deposit types. Concurrent ESG risks create additional complexity at the development and operational stages. The immediate forecast trend is an increase in global demand for cobalt accommodated by a corresponding increase in the production. However, reducing the supply risk and mitigating the complexity of the ESG dimensions of currently undeveloped ore sources is key for sustainable metal production. Future cobalt projects’ concurrent and innate interaction with different risk factors can only be managed if changes to mining operations are made, together with regulation of artisanal mining and reprocessing of mine waste streams for cobalt.

Advances in ore characterisation methods, such as automated mineralogy and core scanning technology, as well as in the breakage characterisation field, suggest that the relationship between the ore characteristics and its comminution behaviour can be thought of differently. This study focuses on developing a relationship that allows the transformation of the quantitative characterisation of rock particles (i.e., mineralogical and textural mea-surements) into particle strength. This study characterised ten different ore types using quantitative methods (i. e., SEM-based technologies and image analysis). At the same time, the strength of particles was collected from physical specimens tested in the SILC over a limited range of particle sizes (3 to 16 mm). The analysis of the databases shows that none of the characteristics has total control over the strength of particles. However, the shape of the grains and the contrast of the physical properties of minerals that compose the particles seem to be more significant in predicting the particles' strength of these rocks. In addition, the model was more successful in predicting particle strength for most rocks with more regular grain sizes, indicating that this parameter requires further comprehension. This research shows that ore characterisation data has the potential to be used to predict the strength of rocks' particles up to a millimetre scale, which is relevant to comminution, which would allow a direct link to be established between the plant and ore characteristics at the particle scale. However, more experimental and numerical studies are needed to develop a predictive and generalised model from quantitative ore characteri-sation data that may allow improving the modelling of the comminution stage by incorporating more resolution to the system and reducing the laborious work of comminution testing, helping to optimise the whole size reduction process.

In developing a geometallurgical characterisation for comminution, the geological information, usually captured from drill core with a resolution of down to micrometre scale, is compared with the results of comminution tests. These tests require several kilograms of material and are empirical tests that do not focus on understanding the effect of ore characteristics separately from the machine functioning. The resultant particles from the comminution process depend upon the operating conditions of the comminution devices (i.e., ball size and load, rotational speed, mill specifications) and the heterogeneity of the rock characteristics (i.e., texture and mineralogy) that influence the mechanical properties of particles which are being reduced in scale from metres to micrometres. A newer approach for comminution modelling and characterisation considers the measurement of the particle strength and minimum required energy to break them at different sizes and its application as a parameter to characterise the breakage behaviour of the ores. However, the relationship between particle strength and minimum energy with the ore characteristics that may link this information to ore' comminution potential has not been explored yet. This study focuses on selecting variables that have been previously identified as relevant for strength at the core scale (i.e., mineralogy, mineral association, grain size, grain shape and porosity) and quantifying their variation as the particle size decreases from 30 to 3 mm. The results show that for a set of ten different rock types, the median value of ore characteristics such as modal mineralogy, mineral association and grain shape does not change with particle size. However, as the particle size is reduced over that size range, these characteristics do become more variable. This behaviour is similar to the variation observed in the strength values of samples, which shows size independence within the studied particle size range, but it is not clear which characteristic has more relevance to the strength results. Detailed geological characterisation information is rarely used to predict how the geological variability may affect the plant's operation. The results from this work indicate that mineralogical and textural features might be controlling the strength of particles formed during comminution. By understanding the physics of the fracture process, these characteristics may be identified. So, a novel link between the detailed mineralogical and textural characterisation at different particle scales and the strength of particles could be developed, which would be helpful in the connection of geological data with the newest developments in comminution research.

The characterisation technologies have advanced rapidly in the last decade. From a qualitative observation of minerals with optical microscopy, more quantitative techniques have emerged. Examples are the SEM-based technologies that focus on mineralogical identification at the microscale and the X-ray microtomography systems that allow identifying rock features in three dimensions. Features such as rock texture and mineralogy have a degree of control on how the rock behaves in the processing plant and thus can affect the project's economic feasibility. None of the available measurement devices is currently capable of identifying all the aspects of rock characteristics that are of interest in linking mineralogy and texture to process response in a single measurement. However, through the integrated use of the techniques in a complementary approach is possible to generate the required suite of information about the mineralogical composition and mineral grain size and shape in a given sample. A multisource method for rock characterisation has been developed in this work. This method includes: • A multistage imaging process that uses 2D and 3D microscopes • An object-segmentation technique to separate mineral grains in the photomicrographs for the quantification of mineralogical and textural properties. • A segmentation technique was developed to create particles of different sizes from a larger image.

The heuristic of a constant liberation by size independent of the degree of comminution has been established as a robust assumption over time. Therefore, when assessing the comminution circuit performance, achieving a final target P80 size at the lowest possible energy and cost seems to be the only concern. The introduction of higher capacity fine screening to replace hydrocyclones challenges this fixed approach due to the inherent difference in classification principles between the technologies. The impact of classification efficiency and mode on the liberation properties of a polymetallic ore ground in a production ball mill circuit are presented. It was found that the liberation properties remained constant across the comminution device but were significantly different across the circuit. The circuit closed with screens was found to improve the coarse particle liberation in comparison to the hydrocyclones. The liberation data is presented and implications for circuit performance are discussed. It is postulated that the assessment of comminution circuit design and performance should be based on the ultimate objective, recovery of the valuable minerals, and thus be based on the ore- and process-specific recovery window, not on a fixed 'grind size'.

The primary breakage properties of rocks such a strength (sigma) and elastic moduli (i.e. Young's modulus (E) and Poisson's ratio (nu)) are not usually characterised at comminution scale. Traditional comminution tests such as the Bond Work Index grinding test and the Drop Weight Test are used for describing the breakage behaviour of different ores. Nevertheless, these tests do not decouple the ore characteristics from the mode of breakage. The primary breakage properties control the initiation and propagation of fractures in rocks and exert control over the resultant progeny of breakage events when the rock is stressed. However, how these properties change as the particle size decreases in comminution is not clear yet. This work explores the change of primary breakage properties and the derived quantities estimated using the Short Impact Load Cell (SILC) (e.g. apparent stiffness and mass-specific fracture energy) within a particle size relevant to ball milling, using a suite of real ores. The results of this work show that the median values of strength in populations of cylindrical (minicore) particles do not vary with size within the range size studied (3-30 mm) for all the ore types studied. In contrast, the stiffness and the mass-specific fracture energy derived from the primary breakage properties data show a size dependency for all of the ore types in agreement with Weibull's weakest link theory. The degree of variation of the primary breakage properties has also been studied over a range of particle sizes. The results show that this degree of variation tends to increase for some of the rocks as the particle size decreases, but the variation is size independent for other rock types. These results align with previous results in the literature that indicate the size independency of tensile strength in small samples but are contrary to earlier researchers' findings that suggest that the variance decreases with size. The ore dependent nature of the primary breakage properties found in this work requires further research. Understanding the behaviour of the primary breakage properties in a comminution context might help to characterise the comminution potential of ores better, as well as to develop characterisation and modelling techniques that are more responsive to the ore characteristics. (c) 2021 Elsevier B.V. All rights reserved.

Current tests for characterising rock breakage in comminution do not allow decoupling of rock breakage properties from the machine environment. Furthermore, it is difficult to predict the comminution behaviour of different ores or how they might behave in another breakage device. Consequently, this decoupling process is essential for comminution modelling of variable ores and blends, which interact with different breakage equipment. Previous studies have shown that the Short Impact Load Cell (SILC) is a versatile tool for characterising the primary ore breakage properties such as strength, apparent stiffness and mass-specific fracture energy, regardless of breakage environment. However, its potential as an ore characterisation technique to relate these mechanical properties to the ore characteristics (e.g. mineralogy, texture) has not been investigated in detail. In this first paper, the results from breaking cylindrical and irregular particles from different ore types in the SILC and Slow Compression Testing demonstrate the SILC's ability to accurately measure strength, stiffness and specific energy of different ore types at different particle sizes which is relevant either for comminution modelling or for rock mechanics testing. The results of breaking cylindrical particles of nine different rock types show that the SILC is capable of matching the results of the slow compression machine for Indirect Tensile Strength from the Brazilian Test and the apparent stiffness results from the Uniaxial Compression test, allowing the accurate calculation of the mass-specific fracture energy distribution for each rock at different sizes. In addition to the cylindrical samples, sets of irregular particles were produced for three rock types and were broken with the SILC. The results show that the median values of strength and stiffness do not change significantly with the shape of the particle, but the standard deviation does increase when irregular particles are tested. If these results are compared with the cylindrical particle results, it can be seen that the geological characteristics of the rock control the median values of the parameters. When the variation around median value is considered it is apparent that particle shape exerts the major control, but geological characteristics still contribute up to 30% of the variation in strength, 25% in stiffness and 25% in the mass-specific fracture energy. The results indicate that the SILC is a suitable device that allows measurement of the primary breakage properties relevant to comminution, which can be easily associated with the geomechanical properties of the rock mass and geological characteristics. The further development of the proposed testing methodology can provide a more comprehensive and less empirical comminution testing for future modelling, incorporating geological and geomechanical properties of the rock.

A number of previous researchers have noticed that the degree of mineral liberation in a given size fraction is the same regardless of where in a comminution circuit a sample is collected. This behaviour has been observed for a range of ore types, and was found to be independent of the mode of breakage. This provides a useful heuristic for modelling the particle characteristics in mineral comminution and separation circuits. However, the published research does not explicitly consider the effect on the liberation behaviour of minerals of the amount of energy applied to break the ore. In this study, a gold-bearing pyrite ore and a copper sulphide ore were comminuted over a wide range of energy levels, using impact breakage. The liberation properties of the product particles were characterised using a Mineral Liberation Analyser (MLA). It was found that that the amount of impact energy applied did not significantly affect the degree of liberation of the minerals in a given size fraction, measured using particle sections. This provides sound experimental evidence to support the heuristic model.

The powerful modern toolbox of hybrid Process Mineralogy for flowsheet development uses best practice sampling as one of its tools. In this paper, the three key components of best practice sampling are reviewed with case studies. These three components are: 1. Minimum sample mass. 2. Rules of unbiassed sampled extraction. 3. The safety line. These excellent models and rules are not commonly taught in undergraduate programmes. In this review paper, which is intended as an introductory reference for those practitioners in Process Mineralogy who have not had exposure to the sampling models, simple and practical explanations are presented for reference. It is shown that finer particle size distributions lead to smaller minimum sample mass requirements. While sampling theory allows us to estimate the error involved obtaining a mass of sample for mineralogical analysis it is also useful to account for errors in the process mineralogy measurements themselves. Examples of the confidence intervals on liberation measurements made on highand low-grade samples are provided to illustrate the importance of sample size—specifically measuring sufficient numbers of particles—in these analyses.

Mineral liberation is a key step in many mineral processing flow sheets and is achieved by breaking large pieces of ore into smaller particles which are suitable for the subsequent separation process. The ore being broken consists of mineral grains which may exhibit a range of properties affecting how they break. In the literature on mineral liberation the breakage of ore is frequently described as random or non-random, with several types of non-random breakage identified. With numerous researchers investigating this topic over the years, a variety of definitions of random and non-random breakage in mineral liberation have been presented in the literature. This paper examines the published work in this area and provides a comprehensive review of random and non-random breakage published in the literature on mineral liberation. It does not aim to revise the definition of these terms but to review the wide range of descriptors used by researchers in this field and identify common approaches to defining random breakage. The definition could be summarised as random liberation being the independence of breakage from both ore properties and mechanical properties during comminution.

•Bootstrap resampling can calculate confidence intervals for particle composition distribution.•It is performed for each individual particle composition class.•It can quantify the required no. of measured particles for the accuracy needed.•It can quantify if there is a benefit in combining particle composition classes.•It can assess if 2 particle composition distributions are statistically different.

Recent advances at JKMRC in modelling mineral liberation required the development of a method to measure ore textural characteristics such as the mineral grain size distribution and the spatial location of mineral grains in three dimensions.This paper describes the development of this method which uses X-ray micro-tomography to provide three-dimensional (3D) measurement of the grain size distribution and spatial location of the minerals in an ore. The methodology was developed on a gold-bearing pyrite ore and used to quantify the textural characteristics which are key inputs for modelling mineral liberation in this ore during comminution. The measurement and image processing steps required to quantify the grain size distribution of the mineral of interest, pyrite, are described together with the image analysis method. The quantification of 3D mineral characteristics together with corresponding two-dimensional (2D) sections through the same particle, provided the opportunity to explore the relationship between 2D and 3D grain size data. (C) 2015 Elsevier Ltd. All rights reserved.

The compositional distribution of a population of particles is often referred to as the liberation spectrum. The challenge is to sensibly represent the vast amount of information contained in the liberation spectrum in a format that can be interpreted in a processing context. This paper describes one approach that can be used to represent the liberation spectrum, from compositional data obtained from automated SEM-based systems, and discusses how ore characteristics, such as mineral grade and grain size and distribution, can be interpreted. Ores from five different locations are compared using pyrite as the mineral of interest to demonstrate the approach. How this data might be used in a mineral processing context is also discussed, specifically with reference to how this approach may be used to determine grinding targets for separation processes. (C) 2014 Published by Elsevier Ltd.