Search Documents

By T. Mützea, L. Schützenmeister, G. Kacheb

High-Pressure Grinding Rolls (HPGR) are employed in cement plants worldwide. Although the technology has been mastered in the majority of applications, design-, scale-up- and operationalchallenges still exist today. In addition to modelling throughput, energy requirement and comminution success, the prediction and avoidance of malfunctions is of great importance in current research. The performance of the mill is highly influenced by the roll speed and the fineness of the feed material. Another parameter influencing the composition of mill feed is the circulating load, which is determined by the grinding circuit configuration, the target fineness and the separation efficiency of the classifier. In order to meet this circumstance and simultaneously maintain a reasonable level of material input and testing effort, a complete grinding circuit on a small technical scale was set up in the research centre of Thyssenkrupp Industrial Solutions AG. The present study varied the roll speed (0.5-2 m/s) and the proportion of fines in the feed (0-75 %) in order to investigate the influence of both parameters on the finish grinding of a Portland cement with a target fineness of 4000 cm²/g according to Blain. The experimental results are compared with model concepts and critically examined with regard to their transferability to industrial plants. Keywords: Cement grinding, grinding circuit, HPGR

Jan 1, 2020

By D. Rankin, H. J. Stucke, W. K. B. Loftus

SYNOPSIS Planning for the Finsch opencast mine is described with particular reference to the proposed stages of mining, the method used to determine the limits of surface waste rock stripping and the derivation of the waste rock mining rate. Actual mining practice is described and the possibility of installing a skipway into the open mine at a later date is discussed. In addition, the waste dumping policy is referred to: this includes the cutting of slots through the hill on the west side of the pipe to increase tramming speeds. The crushing, washing and recovery sections of the treatment plant are described together with the various changes being made to the circuit. The most important of these are the installation of X-ray sorters for the final recovery of + 7 mesh diamonds, and the vibrating grease belt for the recovery of - 7 mesh diamonds. INTRODUCTION The Finsch Mine, a subsidiary of De Beers Consolidated Mines, Limited, is situated approximately 90 miles west of Kimberley, and some two miles from the village of Lime Acres, in the Northern Cape. Lime Acres is a company village shared with the Northern Lime Company which operates a limestone quarry and works near the township. The mine is near the southern end of a well mineralized belt producing limestone, asbestos, diamonds, manganese and iron ore, and the whole area has developed rapidly in the last 20 years. Power for mining operations is drawn from the Electricity Supply Commission. The mine is conveniently close to the main Kimberley-Postmasburg railway line. Water is drawn from the Vaal River through a 57 mile State-owned pipeline at present temporarily operated by mine staff. DISCOVERY AND PROSPECTING OPERATIONS The diamondiferous orebody is situated on the farm Brits, which is State-owned ground that has been leased to the Company for the period of mining operations. There are indications of interest in the area by prospectors (Richter, de Bruin and others) over a period of some 30 years and the discovery would possibly have been made years earlier but for a law preventing prospecting for precious stones on State-owned land. Although the pipe formed a clear depression on the hill where it was situated, it was overlain by some 5 to 40 ft of banded ironstone rubble which was washed into the depression from the hillside above the pipe and obscured the deposit. In 1958, Mr Allister Thornton Fincham applied for, and was granted, the right to prospect for base minerals on the farm Brits. Asbestos is known to occur in the area and, initially, was Mr Fincham's primary interest. However, during the prospecting operations, garnets were discovered and it was realised that diamonds might be present. In 1960, when the Precious Stones Amendment Act was passed, a company

By V. Sheikhzadeh, M. Saghaeian, N. Esmaeili, R. Hejazi, A. D. Ahmadabadi

The main challenges in most low grade iron ore deposits are finding the optimum flowsheet for beneficiation to achieve a final concentrate from both technical and economic viewpoints. Dry preconcentration techniques are being used to assist in reducing the amount of ROM feed to the plant. In this paper the benefits of utilising a dry cobbing magnetic separation technique for upgrading the product of HPGR in a closed circuit with dry screen 4 and 1.4 mm by performing a pilot scale test work was investigated. The experimental test work was done on two sample types of Yogi BIF magnetite iron ore located in Western Australia. Prior to magnetic separation studies, detailed characterisation of the feed materials including physical, chemical, mineralogical and magnetic properties were carried out. Furthermore, the effects of process parameters: drum rotation speed (1.5, 2.5 and 3.5 m/s), magnetic field intensity (1100 and 3800 gauss) and feed size (minus 4 and 1.4 mm) were investigated. The results revealed that sample type1 with 33.90% Fe could be upgraded to 41.51% Fe by using a stage of 1100 gauss dry magnetic separator with 69.12% weight recovery and 84.36% Fe recovery at a size of 4 mm. Also for sample type2, it is possible to increase Fe grade from 29.84% to above 36.68% with 75.58% weight recovery and 90.32% Fe recovery by applying one stage of 1100 gauss dry magnetic separator to a 4 mm size sample. Keywords: Banded iron formation (BIF), pre-concentration, dry cobbing technique, magnetic separator

Jan 1, 2020

By J. C. D. Gush

Paper written on project work carried out in partial fulfilment of Laboratory-scale testwork was undertaken to recover oxidic copper-cobalt minerals by froth flotation, which requires maintaining a certain sulphide level in the pulp. To improve control of the sulphide concentration, an electronic closed-loop system was designed that responds to the potential output of an ion-selective electrode. The PID control was improved by adding a pre-emptive control and a self-calibration feature; a data logging facility is also included. An inexpensive valve design was incorporated using locally available parts. The operating results of this control design are presented.

Jan 1, 2005

Mining executives often have a difficult task determining what the strategic objective of the business should be as this can be impacted by the prevailing market conditions. In addition, they have no mechanism to quantitatively 'test' the impact of this strategic decision on the business and understand the underlying dynamics. During the commodities bull run of 2003 to 2008 the strategic objective may have been to grow the long term value of the business (NPV) through increased tonnage, acquisition and finding new reserves, which all came with an increasing fixed cost base. Now with the financial crisis upon us and the collapse of commodity prices and demand, executives have adjusted their strategies as 'cash is king' and short-term cash flow, in some instances at the expense of long terms value, is the order of the day. For many mining companies, mine closures, reductions in production and cost cutting exercisers are now the focus. In many instances, management do not have an ability to rapidly test different strategic alternatives to 'test' the impact on value, unit costs, reserves and profitability at the operational level and optimize the underlying trade-off variables. Economic modelling of the complete business value chain is a means of linking the operational 'reality' and strategic choices, so that the full impact can be assessed. This paper describes some of the challenges facing mining executives and how economic modelling can be applied to make decision making more rigorous.

Jan 1, 2010

By G. Sánchez, S. Seiler, B. Klein, T. Cebeci, P. Bradshaw

The Decar Nickel District is a greenfield discovery by FPX Nickel Corp., of nickel mineralisation located in central British Columbia (BC), Canada. The Baptiste deposit is part of the Decar Nickel District and is a unique deposit containing a nickel-iron alloy mineral called awaruite, with over 2-billion tonnes of inferred resource grading 0.12% Davis Tube Recoverable (DTR) nickel. This deposit represents a significant nickel resource. Awaruite is a high density and ferromagnetic mineral, which could be recovered, along with magnetite that is also present in the samples, using magnetic separation. However, there is still a lack of understanding of the properties of the awaruite and its metallurgical response. This information is imperative for the set-up of the variables involved in its concentration. The aim of this paper is to show and discuss the effect of magnetic field strength for the recovery of nickel from Baptiste deposit samples. Magnetic separation at bench and pilot scale with different magnetic field strengths is evaluated as a concentration method for awaruite. Mineralogical information is provided to support the results and also to briefly assess the variability of DTR nickel within the deposit. The Baptiste deposit is unique and will benefit from a sound scientific understanding of the process solution as well as other ongoing work. If successful, the outcome will lead to the development of a mine. With greater than 2-billion tonnes of inferred resource, the mine would represent a significant new source of nickel globally. Keywords: Nickel ore processing, magnetic separation , awaruite, magnetite

Jan 1, 2020

By S. Banisi, A. Ghasemi, E. Nematollahi

Discrete element method (DEM) has been used as a popular simulation method to verify designs by visualising how material will flow through complex equipment geometries. Although DEM simulation is a powerful design tool, finding a DEM model which includes real material properties (e.g., shape, cohesion and friction coefficients) is not computationally feasible. In order to obtain more realistic results, particle energy loss due to rolling friction has been highlighted by many researchers. To implement a reverse torque to account for rolling friction, various models have been proposed. On account of the complexity of the problem, there is no unique model for all applications (i.e. dynamic and pseudo-static regimes). In this paper, an in-house developed DEM software (KMPCDEM ©) was used to assess the robustness of three commonly used models by comparing the repose angle of iron ore pellets obtained through the draw down test in a laboratory setup and then the elastic–plastic spring dashpot model was modified based on considering the individual parameters instead of relative parameters of two contact entities. Results showed that the modified model could produce higher repose angle. The modified model was used to calibrate DEM simulation of repose angle of iron ore pellets in a laboratory setup of the draw down test. Comparison of the calibrated DEM simulation and laboratory results showed 2° differences between the predicted and measured angle of repose. This difference was attributed to considering the particle shape effects by assuming a spherical shape for particles. Keywords: DEM, rolling friction,particle shape, repose angle

Jan 1, 2020

By J. R. Percival, L. Ip, M. D. Piggott, S. J. Neethling

Component wear due to particle-laden flow is a significant challenge in the minerals processing industry in the handling and transportation of abrasive mineral slurries. Studying wear through experimentation is time-consuming and expensive; therefore possessing the ability to accurately simulate, understand and predict wear in these complex environments is highly desirable. The ability to predict wear is important in both the design of equipment in order to maximise the performance over the wear life of the components, as well in minimising potentially costly and dangerous failures and unscheduled maintenance.. Particle-induced wear is a complex problem, requring an understanding of coupled fluid-solid motion, geometry deformation in response to fluid-solid behaviour, moving geometries and wear modelling. Although the literature demonstrates that rebounding particles and changes to surface profiles can have a significant effect on subsequent wear rates and patterns, the challenge of implementing a model to account for this effect in complex geometries has been addressed by only a few authors in the literature. This paper describes a new wear simulator, leveraging fluidity, an open-source, massively parallel, finite element based CFD simulator capable of unstructured adaptive meshing. The wear simulator includes the ability to model particle-laden flow using representative Lagrangian particles and dynamic remeshing of geometry in response to wear and component movement. These capabilities are demonstrated using wear and the associated dynamically deforming boundaries within a Coriolis tester arm, with the results comparing favourably with the experimentally observed wear patterns. Keywords: CFD, wear, two-phase flows, anisotropic mesh adaptivity, boundary deformation

Jan 1, 2020

By P. G. Van Der Walt

This paper deals with the reticulation of deslimed tailings from surface to the stope face underground. Reference is made to the experience gained at the pilot plants operating on four of the gold mines administered by Gold Fields of South Africa, as well as at the full-scale system at East Driefontein which was commissioned recently. Mine-wide backfill systems are currently being installed on five other mines administered by Gold Fields. These six full-scale systems are capable of handling more than 270 000 tons of backfill solids per month. The type of reticulation system chosen, the selection of flowrates and the composition of the slurry, as well as the selection of the appropriate pipes and ancillary equipment are discussed. Both theoretical and practical aspects are taken into consideration. Other aspects covered include the design of underground storage dams and communication systems.

Jan 1, 1988

By J. E. Dickinson, R. Campbell, M. J. Cole, K. P. Galvin, M. C. Cole

It is common practice in the beneficiation of iron ore ultrafines to apply reverse flotation, and hence float the gangue minerals, leaving behind a high-grade iron ore product. Given the difficulty in floating particles finer than 20 μm, especially those finer than 10 μm, it is necessary to subject the feed to desliming ahead of flotation. There are two problems with this approach. Firstly, given the known inefficiencies of cyclones, desliming needs to be aggressively applied to particle sizes upwards of 45 μm to ensure all the offending ultrafine gangue is removed. Secondly, desliming also removes a significant portion of the valuable iron ore. Thus, a significant portion of the ultrafines, including the iron ore, is lost to the tailings. This paper examines the direct flotation of iron ore fines using the Reflux Flotation Cell. The RFC operates using a concentrated bubbly zone, in the absence of a conventional froth zone, for delivering strong counter current washing through an inverted fluidised bed of bubbles ahead of the product discharge. A system of parallel inclined channels in the lower part of the cell provides enhanced segregation of the bubbles from the tailings, preventing the loss of air bubbles into the tailings. Experiments conducted under semi-batch conditions produced encouraging results with concentrates of ~65% Fe grade in a single flotation stage. Comparatively, up to five successive batch flotation stages were required when using a laboratory scale batch mechanical flotation cell to achieve similar Fe grades. For a given grade, lower Fe recoveries were typically achieved by the RFC. This was attributed to the effective prevention of ultrafine particles succumbing to hydraulic entrainment under a positive bias flux, as well as an increasing feed volume involved in the semi-batch operation using inverted fluidisation water for cleaning. Overall, similar grade-recovery curves were produced, however, significantly higher grades were obtained by the RFC and with fewer separation stages. Keywords: Desliming flotation, iron ore, modified release analysis, reflux flotation cell

Jan 1, 2020

By O. Dahl, B. Musuku, N. Schreithofer

Recovery of fine pentlandite is proving to be a challenging task in the Kevitsa sequential flotation operations, with over 60% of the nickel losses reporting into the -25μm size fraction. Since commissioning, the Kevitsa process has undergone major changes to address throughput and flotation performance. Major changes were implemented in the configuration of the milling and flotation circuits as well as in the operating conditions and reagent regime. Some of these changes improved nickel flotation while others deteriorated flotation performance. Most of the changes have had a direct impact on nickel flotation. This paper reviews the systematic analysis, laboratory testing and salient changes done to the process with special focus on the impact on the fine pentlandite flotation. Keywords: Blending, Flotation, Chalcopyrite, Pentlandite, Fines

Jan 1, 2020

By V. Sheikhzadeh, M. Saghaeian, N. Esmaeili, R. Hejazi, A. Dehghani Ahmadabadi

High Pressure Grinding Rolls (HPGR) are mostly used prior to pelletising processes to increase the specific surface area (SSA) of iron ore concentrate. For expressing SSA, the Blaine index is utilised as one of the most important dimensional properties of iron ore pellet feed. ISO 21283 is a standard test method that explains the manual of Blaine air-permeability apparatus for determination of this parameter. In this standard, the disaggregation method is based on passing the minus of 1 mm samples from the 0.150 mm sieve. The disaggregation process has an effective role on the results of Blaine airpermeability test especially for regrinding HPGR products. In this research, firstly the Blaine index of samples was measured without disaggregation. Following this, disaggregation was done by passing the minus of 1 mm samples through three sieves with different dimensions, 0.150, 0.090 and 0.045 mm, separately, and the Blaine index was calculated for each one. Finally, lab-mill with rubber balls, called Jar mill, was investigated as another option. Screen analysis and image analysis with SEM were performed to evaluate the effects of the sample's disaggregation on the size distribution and agglomeration of samples. The results showed that the disaggregation method with sieve size 0.150 mm, according to ISO 21283, was not applicable for HPGR products because the particles were still stuck together. In addition, the disaggregation method with the sieve size 0.045 mm or with Jar mill offered reasonable results. Keywords: Disaggregation, Blaine index, specific surface area, regrinding HPGR, ISO 21283, Jar mill

Jan 1, 2020

By Pieter de Waal

The beneficiation equipment under the control of plant metallurgists exploits the differences in the physical properties between valuable and waste mineral components. In many cases, laboratory results are too late for process control and can only serve the purpose of "post" reports, the calculation of a mass balance and the issue of a quality certificate. As a result, experienced operators regularly adjust separation equipment based on the colour and texture of the various product streams in the beneficiation plant. Metallurgists quickly learn to recognise the different minerals in their ore body. The stereo microscope has been a valuable tool used by operators to quantify known mineral species. For many years it was the only practical method available for the purpose of process control and even for quality contro1. The modem spectroscope is far superior to the human eye and can rapidly detect a change in mineral composition. New generation optical scanners and digital signal processors can provide useful signals, related to mineral composition that can readily be applied for the purpose of process control. Successful case studies, covering dry in-line and slurry applications as well as in-plant fast spot analysers are presented.

Jan 1, 2014

"Chairman: Dr DeV. HarrisRapporteur: Mr A.H. MunroPapers:Corporate exploration strategies by MacKenzie, B.W.The Engel Simulator and the search for uranium by Griffiths, J.C. and Singer, D.A.The Chairman opened the session by tracing the changes which have taken place in the exploration field since the days of the lone prospector who relied upon his individual skills and to a large extent on luck. The deposits which now remain to be discovered are better covered and are of lower grade. A much greater team effort, involving geoscientists and engineers, is required. The gamble is thus much greater -not only are the probabilities of failure still significant but the consequences are more severe. A recent Canadian survey indicated that expenditures of between two and nine million dollars on regional exploration programs are required to locate one significant deposit.The Chairman went on to say that these facts have led companies to regard exploration as another form of investment and to treat it accordingly. This had placed exploration in the context of corporate strategics and structures, which was the subject of the first paper of the session.In introducing his paper, Dr B. W. MacKenzie said that the paper suggested a decision framework for exploration planning. Initially the environment is analyzed to define observable and relevant economic characteristics. The terms of reference for exploration planning and the appropriate allocation of capital and skills are then established in the light of corporate objectives and growth alternatives. With the aid of this information and in the light of profit and corporate risk criteria, exploration strategies can be selected.Dr MacKenzie went on to say that the usefulness of the concept and criteria outlined in his paper depended critically upon three characteristics of any particular economic environment, namely, the return resulting from an economic discovery, the unit discovery Cost and the probability of making an economic discovery. He then showed by means of a case history relative to the Canadian Shield and Cordilleran areas that these three characteristics could, in fact, be evaluated for regions with reasonable histories of mineral activity.The first contributor to the discussion of the paper, Mr Alfred Weiss, mentioned that his group had been involved in exploration modelling for some ten years and were thus particularly interested in the paper. He went on to say that they had experienced greatest difficulty with reconnaissance programs and that he sensed that these difficulties were also present with the methodology which had been outlined. Mr Weiss asked Dr MacKenzie how he would deal with such situations when no information was available but where there was the potential for both exploration and discovery. Dr Mackenzie agreed that the lack of information was a very severe limitation and said that at the present time his methodology probably did not provide a basis for breaking down the broad geological regions he considered into areas for further assessment."

Jan 1, 2014

By C. B. Carelse, M. Manuel, D. Chetty

Automated mineralogical analysis of platinum group minerals (PGM), using the mineral liberation analyser, provides information used to understand flotation efficiency, primarily from a PGM liberation and association standpoint. To date, however, there is no consistent understanding regarding PGM speciation and its impact on floatability. Previous studies of PGM from UG2 ores concluded that a larger dataset is needed from which the various parameters can be modelled. Parameters on telluride, arsenide, sulfurarsenide and alloy species need to be integrated to fully understand the impact of PGM speciation on flotation, and for this, ores other than PGE-sulfide-rich UG2 must be assessed. Flotation rougher rate tests were performed on feed samples from two Platreef ores to determine the flotation response of individual liberated PGM species. This was achieved through automated mineralogical analysis of feed, rougher concentrates and tails. Based on 1529 liberated PGM grains assessed over the circuit for two Platreef ores, preliminary results show that PGM speciation, species grain size and competition from gangue appear to influence floatability in an ‘optimised’ laboratory environment. Keywords: PGM, PGM species, flotation, liberation

Jan 1, 2020

The optimal design of production phases and ultimate pit limit for an open pit mining operation may be generated using conventional or stochastic approaches. Unlike the conventional approach, the stochastic framework accounts for expected variability and uncertainty in metal content by considering a set of equally probable realizations (models) of the orebody. This paper evaluates the performance of a new stochastic network flow approach for the development of optimal phase design and ultimate pit limit using a gold deposit as the case study. The stochastic and conventional frameworks as considered here utilize the maximum flow and Lerchs-Grossman (LG) algorithms, respectively. The LG algorithm is restricted to considering an estimated (average-type) orebody model, while the stochastic maximum flow algorithm is developed to simultaneously use a set of simulated orebody realizations as an input. The case study demonstrates that, when compared to the conventional LG algorithm as used in the industry, the stochastic approach generates a 30 per cent increase in discounted cash flow, a 21 per cent larger ultimate pit limit, and about 7 per cent more metal, while it maintains a consistency in phase size.

Jan 1, 2012

The optimal design of production phases and ultimate pit limit for an open pit mining operation may be generated using conventional or stochastic approaches. Unlike the conventional approach, the stochastic framework accounts for expected variability and uncertainty in metal content by considering a set of equally probable realizations (models) of the orebody. This paper evaluates the performance of a new stochastic network flow approach for the development of optimal phase design and ultimate pit limit using a gold deposit as the case study. The stochastic and conventional frameworks as considered here utilize the maximum flow and Lerchs-Grossman (LG) algorithms, respectively. The LG algorithm is restricted to considering an estimated (average-type) orebody model, while the stochastic maximum flow algorithm is developed to simultaneously use a set of simulated orebody realizations as an input. The case study demonstrates that, when compared to the conventional LG algorithm as used in the industry, the stochastic approach generates a 30 per cent increase in discounted cash flow, a21 per cent larger ultimate pit limit, and about 7 per cent more metal, while it maintains a consistency in phase size.

Jan 1, 2012

By X. W. Pan

It has been found that the MnO content of a slag decreases sharply when direct current is applied to a 1T1£tal-Slagsystem consisting of liquid Mn-Si-Ccsat metals and MnO-SiO2-CaO-AI203 slag in graphite crucibles. This paper reports a preliminary investigation into the electrochemical effects of direct current on the metal-slag system, the silicon content of the metals, and the MnO content of the slag. The results show that, when a direct current of 7 A is maintained, the MnO content of a metal containing 10 to 11 per cent silicon by weight is reduced rapidly from 26,5 to 17,2 per cent by weight. This reduction seems to proceed further when the process is continued. When the direct current and the voltage are kept constant at TA and 4 V respectively, the MnO content of a metal containing 1 to 2 per cent silicon by weight is reduced to 1 from 25 per cent by weight, which results in a manganese yield of 96 per cent.

Jan 1, 1995

By M. B. Shongwe

Pt-based alloys are being developed for high-temperature applications with the aim of replacing some of the currently used Ni-based superalloys (NBSAs) in the highest temperature applications. The Pt-based alloys have a similar structure to the NBSAs, and since Pt is more chemically inert than nickel and has a higher melting point, they can potentially be used at higher temperatures, up to 1300°C, and in more aggressive environments. Several experimental Pt-based alloys were studied at Mintek, and an optimum composition was found to be Pt84:Ahi:Ru2:Cr3 (at.%). On the basis of hardness and microstructure, a later study identified a new optimum: Ptr8:A111:Ru5:Cr6 (at.%). There are at least two Pt3A1 allotropes, and the high-temperature cubic structure has better properties than the lower temperature tetragonal form, and so needs to be stabilized. This work describes the latest results obtained in transmission electron microscopy (TEM) studies of the quaternary Pt-based superalloys. These results are both interesting and important, because the samples have a higher precipitate density compared to those from earlier work. The precipitate morphology is mainly cubic, with minor areas having irregular-shaped precipitates. The high volume fraction of the precipitates is a major breakthrough, since the objective of this work is to improve the alloys. A prior disadvantage was that the proportion of the precipitates was too low in samples before this work, especially compared with the work from Germany on Pt-Al-Cr-Ni¬based alloys as well as the NBSAs. TEM -Pt3A1 diffraction patterns were studied, and for each diffraction pattern, many possible lattice point combinations were tried, with the angle between the lattice spots as well as interplanar spacings being calculated for each phase (cubic or tetragonal) to match the measured results. An overall analysis of the diffraction results indicates that the cubic phase fitted the experimental lattice points with much lower errors compared to the tetragonal phase. Thus, with the close match achieved with the cubic structure, the structure of -Pt3A1 precipitates is likely to be cubic. X-ray diffraction has been carried out on selected samples, and the results confirmed the presence of cubic -Pt3A1 and (Pt). Different X-ray diffractometers were used to further verify the results, and the results were also compared with those from TEM.

Jan 1, 2012

By M. Kumral, S. Hmoud

Optimising the configuration of infill drill holes is an important step during mineral exploration campaigns. However, this optimisation process encompassing all possible drilling configurations within a mineral deposit is highly combinatorial and hard to solve in a reasonable time. This paper proposes using spatial clustering algorithms to divide the orebody into contiguous zones and target the best zones with infill drilling in order to reduce the size of this optimisation problem. Several clustering algorithms are assessed in a spatial context on hypothetical datasets. This evaluation shows that the hierarchical density-based spatial clustering application with noise (HDBSCAN) algorithm can be used in combination with the uniform manifold approximation and projection (UMAP) dimensionality reduction algorithm to identify contiguous zones in the orebody.

Sep 1, 2021