Search Documents

By F. Wang, P. Rosario, A. Vazirizadeh

Advances in data science, machine learning, and artificial intelligence are transforming mining and mineral processing, and making them more algorithm intensive. The paradigm is shifting from one of detection and control to one of prediction and optimisation. To build up reliable prediction and optimisation models these steps should be passed: collect, aggregate, cleanse, and process large amounts of complex, structured, and unstructured data as a first stage. Next, the right tools must be used to extract meaningful features from the operational data for the build-up of a prediction model. A grinding circuit is one of the challenging areas in terms of determining its overall efficiency and the prediction of product quality. In a processing plant, some of the KPIs are typically unmeasurable or infrequently measured for real-time application, e.g. ore grindability and grind size. However, these variables are required to calculate the grinding circuit efficiency and to model the operation performance. In this paper, advanced analytic tools have been used to extract one selected KPI (cyclone overall P80 or final grind size) and its relationship with measured variables for an industrial dataset. The results illustrate that the performance of the grinding circuit can be predicted for real-time monitoring by means of measured data. Keywords: Grinding circuit, data-driven, machine learning, process modelling, hybrid modelling

Jan 1, 2020

By S. Krishnannair

Singular Spectrum Analysis (SSA) is an effective data-adaptive multimodal tool for monitoring chemical processes. However, monitoring multiscale signals using principal component analysis obtained from the multilevel decomposition of process signals using basic SSA assume the upper control limits of monitoring charts (" ) to have a Gaussian distribution. This assumption deteriorates the performance of basic SSA when applied to the monitoring of nonlinear processes. To address the issue in this paper, the kernel density estimation (KDE) was used to estimate upper control limits of monitoring charts for basic SSA based nonlinear process monitoring techniques. The monitoring performance of the resulting basic SSA-KDE approach was then compared with a basic SSA approach of which upper control limits of the monitoring charts were determined based on a Gaussian distribution. Application on nonlinear dynamic simulated process and base metal flotation plant chemical processes showed that the SSA-KDE approach provided better overall performance than the SSA approach with Gaussian assumption-based upper control limits for its monitoring charts. Keywords: Fault detection, singular value decomposition, singular spectrum analysis, kernel density estimation

Jan 1, 2020

By A. Marek

The safe and optimal installation of roof support remains one of the biggest challenges in underground mining today. Traditional hand-held, roofbolt installation in conventional mining has been replaced, in a lot of mining operations, by mechanized equipment. This was a first step to moving the operators out of the unsafe working environment and a start on the trend toward a changed support regime in mechanized mining. However, the challenge of handling of the drill steel itself, as well as inserting resin cartridges and bolts, remains a major hazard and the cause of many lost-time injuries. The paper describes and discusses developments in this area, of mechanized bolting and on-board automation, presenting solutions to known problems, achievements to date, and future developments. The solutions presented will identify with addressing customer needs while touching on possibilities for further technology changes, keeping focus on the safety aspect for operators and the need to bring the operators even further away from the unsafe working environment. The-state-of-the-art technology in mechanized bolting equipment is presented and a brief outlook on the past, current, and future developments is given.

Jan 1, 2012

By M. Fener, S. Kahraman, O. Gunaydin, M. Alber

Falling into the weak rocks category, fault breccias have extremely poor engineering properties. These pebbles typically cause issues with slopes, subterranean construction, and building projects. Professionals will benefit from the creation of some predictive models for fault breccia triaxial strength, as smooth specimen preparation is typically challenging and timeconsuming. The purpose of this study is to develop some predictive models for the differential stress (Δσ) based on physical and textural properties. Artificial neural networks were used to analyse data related to Misis fault breccia. Initially, models with moderate (noticeable, but not good) correlation coefficients were created using multiple regression analysis. After that, the regression models and three distinct artificial neuron network models were contrasted. Regression models are weaker and less trustworthy than artificial neuron network models, as demonstrated by this comparison. Pointed out is the practicality and ease of use of the artificial neuron network model with S-wave velocity and volumetric block proportion. Ultimately, it can be concluded that artificial neuron networks analysis provides a reliable indirect method for predicting the differential stress of Misis fault breccia.

Sep 1, 2025

By E. R. Leeman

INTRODUCTION The first references in the literature to the so-called C.S.I.R. 'doorstopper' strain cell were made in 1964 1,2,3,4. These were the climax of nearly two years of experimental and developmental work with the instruments and the operating techniques involved. Since that time numerous field measurements have been made with 'doorstoppers' and they have been developed to a stage where they are in commercial production and are being supplied to users throughout the world. No detailed description of the latest 'doorstopper' equipment has, however, appeared in print and this paper is written to bring the literature up to date in this respect. The triaxial strain cell was developed three years ago 5, 6, 7 and was first described at the First Congress of the International Society for Rock Mechanics in Lisbon in 1966 8. Since that date laboratory tests and-particularly important-field tests 7' 9 have demonstrated the validity and practicality of the instruments and techniques. These instruments have also been developed to a stage where they have been released for commercial production. In this form they are very different from those described two years ago and a full description of them at this stage seems to be justified. THE INFORMATION SUPPLIED BY 'DOORSTOPPER' AND TRIAXIAL STRAIN CELLS Both the 'doorstopper' and triaxial strain cells were designed specifically to determine the absolute stress in rock using an overcoring stress relieving technique. They were not designed to measure changes in stress. They may, however, be used for the latter purpose provided the glue which is used to stick the strain gauges to the rock is known to possess sufficiently stable strain-time characteristics as not to affect the accuracy of the results during the time the measurements are made. The 'doorstopper' was designed primarily to measure the major principal stress in situations where its direction and those of the two other principal stresses were either known or could be assumed. For example, at great depth in undisturbed virgin ground, it is reasonable to assume that the maximum principal stress is vertical and that the other two lie in a horizontal plane. The borehole in which the 'doorstopper' is installed would thus be drilled in the horizontal plane parallel to one of the minor principal stresses, the direction of which would either have to be assumed or determined as described later. The stresses measured would be the vertical major principal stress and the principal stress acting in a horizontal plane normal to the borehole axis. It is possible, as will be shown later, to obtained the complete state of stress using 'doorstoppers'. In this event, it is necessary to make measurements in three boreholes drilled in any three known directions to each other and relative to a system of three orthogonal axes. The triaxial strain cell was designed to determine the complete state of stress in a single borehole drilled in any direction in any stress field. At first glance it would appear to supersede the 'doorstopper'. This is not necessarily true since there are

By W. C. Joughin

The science of soil mechanics is 100 years old, and rock mechanics is about 80 years old. While methods of analysis and design have been developed and have evolved over time, these are relatively young sciences. The rapid increases in computing power and new technologies have enabled more sophisticated modelling and monitoring. However, there are still many aspects of soil and rock mechanics that are not well understood. Geotechnical failures, which have major consequences, still occur. These consequences may include environmental damage, major production holdups and associated loss of revenue, damage to infrastructure, and loss of life. High-consequence events, which are rare, are more difficult to anticipate and to design for, because by their nature they involve extraordinary circumstances or conditions, often geological in nature. The risks are usually mitigated by conservative designs and monitoring. Detailed geotechnical investigations help us to understand the natural variability of soil and rock masses and identify unusual or unexpected conditions. Investigating and researching major geotechnical failures is essential to enable these unusual circumstances to be anticipated. In the past, severe unanticipated events may have been treated as natural events or ‘acts of God’. However, society now has much greater expectations and it is essential to have policies and procedures in place that enable appropriate management of these rare, high-consequence risks. A good example is the Global Industry Standard on Tailings Management (GISTM), which was introduced after the catastrophic dam collapse at Vale’s Corrego de Feijao mine in Brumadinho, Brazil. The address will explain the concepts of uncertainty and variability, and how they should be taken into account in geotechnical design. The challenges facing geotechnical engineers, mine owners, and managers will be discussed. referencing a number of real case studies.

Sep 6, 2023

By A. K. Raji, E. F. Orumwense, M. O. Ojumu

This research paper introduces a novel passive method for ocean mining using an electricpowered robotic subsea dredging crawler designed to operate in various subsea conditions, including unstructured and rough terrain. The electric-powered robotic subsea dredging crawler is equipped with two driving tracks and features a complex assembly of tank tread chain link, tank tread idler wheels, tank tread sprockets, bronze bushings, motor hub, motor mount, and a 12-volt direct current motor. To evaluate the performance of the electric-powered robotic subsea dredging crawler, it was compared with existing ocean hydraulic exploration crawlers, such as MK3 ROST, in terms of drive performance on an inclined ramp with five angles of inclination (0°, 15°, 20°, 25°, 30°). The experiment aimed to determine the maximum velocity, torque, and efficiency of the crawler at different angles of inclination. The collected data were analysed and plotted using Python, and the effect of the electric-powered robotic subsea dredging crawler’s weight on its driving performance was also considered. The results of this study demonstrate a quadratic relationship between driving performance and inclination angle, which can be used to optimise the electric-powered robotic subsea dredging crawler’s performance for future ocean mining robot drive applications.

Oct 2, 2025

By D. V. J. W. Groenewald, T. M. Louw, L. Auret, S. M. Bradshaw, A. V. Cherkaev

Unexpected equipment failures lead to financial losses due to production disruption. Time-based maintenance, based on checks over regular time periods, has been employed to minimise disruption. However, the improvements it can offer are limited as it may lead to unnecessary service of healthy equipment or missed failures occurring between service intervals. Condition-based maintenance (CBM) provides a further improvement to equipment reliability by detecting faults and predicting remaining useful life (RUL) based on equipment health monitoring. Its application in the mineral processing industry is hindered due to the high cost of health monitoring equipment. In this study, two stochastic process models, the adaptive auto-regressive model and the gamma process, are compared in their ability to forecast the degradation state and to provide RUL prediction. The gamma process model was found to be effective in both tasks. Due to its simplicity and minimum hyperparameters requirements, it is proposed to use the gamma process as a baseline for comparison with other models. Keywords: Condition-based maintenance, remaining useful life, slurry pump, adaptive autoregressive mode, gamma process

Jan 1, 2020

By M. M. Masood, G. Y. Raju, T. Verma

The inverse velocity method has proven to be an effective approach for predicting slope failures in surface mines by analysing displacement monitoring data. However, the accuracy of inverse velocity method predictions is significantly affected by instrumental noise and natural environmental variations, which influence the identification of different deformation stages. To enhance predictive accuracy, this study applies and evaluates three filtering techniques to velocity time series data: Exponential smoothing filter, short-term smoothing filter, long-term smoothing filter and also compares it to raw data (no filtering). A refined prediction framework, that is, filtered inverse velocity method analyser, is proposed to improve slope failure forecasting in surface mining operations. The results demonstrate that filter selection plays a crucial role in optimising failure time predictions, offering valuable insights for geotechnical monitoring and early warning systems in surface mines.

Sep 1, 2025

By C Smit

SYNOPSIS A prototype collimated radiometric face scanner was tested in the Harmony Gold Mine. The results obtained during the pilot study indicate that in situ radiometric uranium assays are statistically indistinguishable from those obtained conventionally from channel chip samples. In addition, the study demonstrated that reasonably reliable gold estimates can be deduced from the radiometric measurements, by use of the ratio of gold to uranium within a mine. The instrumentation, calibration procedures, and background determination are described briefly.

Jan 1, 1985

By K. Grossauer, W. Schubert

For tunnelling in poor and heterogeneous ground a reliable on-site short-term prediction is required to allow a safe and economical construction. With the use of absolute displacement monitoring data and advanced software for evaluation of the data in combination with a routine update of the geological model a considerable increase in the reliability of the short-term prediction could be achieved. The paper deals with theoretical considerations, numerical studies and case histories from alpine tunnels in heterogeneous rock. Tunnelbau in schlechtem und heterogenem Gebirge erfordert auf der Baustelle eine verlaessliche Kurzzeitprognose, um ein sicheres und wirtschaftliches Bauwerk gewaehrleisten zu können. Die Verwendung von absoluten Verschiebungsmessdaten und deren Auswertung mit Hilfe von hoch entwickelten Softwareprogrammen in Kombination mit einer laufenden Aktualisierung des Gebirgsmodells ermöglicht eine erhebliche Erhöhung der Verlässlichkeit von Kurzzeitprognosen. Dieser Beitrag beschäftigt sich mit theoretischen Ansätzen, numerischen Modellen und Fallstudien von Tunnelvortrieben in heterogenem Gebirge. La construction de tunnels dans la montagne hétérogène de faible résistance exige un pronostic fiable à court terme sur le chantier, afin de pouvoir garantir un ouvrage sûr et économique. L'utilisation des données de mesure de décalage absolues et de leur évaluation à l'aide de logiciels sophistiqués en combinaison avec une mise à jour courante du modèle de la montagne permet une augmentation considérable de la fiabilité des pronostics à court terme. Cette contribution présente des approches théoriques, des modèles numériques et de cas pratiques de progressions de tunnels dans des montagnes hétérogènes.

Jan 1, 2003

By M. Kamali

In this research, in order to predict the peak particle velocity (PPV)(as vibration indicator) caused by blasting projects in the excavations of the Karoun III power plant and dam, three techniques including statistical, empirical, and neural network were used and their results were interpreted and compared. First, multivariate regression analysis (MVRA) was used as statistical approach. Next, PPV was predicted using some widely used empirical models. Lastly, an artificial neural network was used. In the ANN model, maximum charge per delay, total charge per round, distance from blast site, direction of firing, blasthole length, number of blastholes, total delay in milliseconds, number of delay intervals, and average specific charge were taken into consideration as input parameters and consequently the PPV as output parameter. The results of the techniques were interpreted from two points of view. Firstly, the correlation between the observed data and predicted ones, secondly the total error between observed data and predicted ones. The MVRA had a satisfactory correlation but its error of estimation was comparatively very high. The empirical model had reliable correlation and a small error of estimation; in total the results of empirical method were more reliable than those of MVRA. Generally, the ANN approach showed very high correlation and a very small error. The results of this research indicated that the ANN model is the best predicting model for PPV in comparison with other approaches.

Jan 1, 2010

By J. R. L. Read, A. N. Ogden

This paper describes mining industry funded research that is being undertaken to address the critical gaps and uncertainties in our knowledge so that we can develop new approaches and create more effective ways for predicting the reliability of rock slopes in large open pit mines. It is envisaged that this research will not only lead to more effective pit slope design but further reduce the potential for unforseen geotechnical risks by enabling the geotechnician to communicate the risks more effectively with management and to optimise site specific control measures. To achieve this, several research tasks are in hand. The first is to accurately represent the structured rock mass with a 3D computer model (Siromodel) that creates polyhedral blocks from structural distribution data characterised by position, orientation and persistence. Siromodel is being coupled with particle flow and other numerical codes to evaluate how the structured rock mass deforms in the stress environment of an open pit and how a candidate failure surface may propagate through the dilating rock mass along the path of the least shear and/or tension resistance. A beta version of the Siromodel software has been distributed to 20 Sponsor mine sites for technical evaluation. The end state is to develop methods of analysis that characterise the reliability of a described candidate surface, and the probability of the candidate surface propagating to failure, recognising that the rock mass strength may well be defined in a different way than before.

Jan 1, 2006

By M. A. Mahboob, T. Celik, B. Genc

Mineral Resources estimation plays a crucial role in the profitability of the future of mining operations. The conventional geostatistical methods used for grade estimation require expertise, understanding and knowledge of the spatial statistics, resource modelling, geology, mining engineering as well as clean validated data to build accurate block models. However, the geostatistical models are sensitive to changes in data and would have to be rebuilt on newly acquired data with different characteristics, which has proved to be a time-consuming process. Machine learning methods have in recent years been proposed as an alternative to the geostatistical methods to alleviate the problems these might suffer from in Mineral Resource estimation. In this paper, a systematic literature review of machine learning methods used in Mineral Resource estimation is presented. This has been conducted on such studies published during the period 1990 to 2019. The types, performances, and capabilities, of several machine learning methods have been evaluated and compared against each other, and against the conventional geostatistical methods. The results, based on 31 research studies, show that the machine learningbased methods have outperformed the conventional grade estimation modelling methods. The review also shows there is active research on applying machine learning to grade estimation from exploration through to exploitation. Further improvements can be expected if advanced machine learning techniques are to be used.

Nov 30, 2022

By L. Uotinen, M. Rinne, R. Mishra

Underground mining gives rise to geotechnical hazards. A formal geotechnical risk assessment can help to forecast and mitigate these hazards. Frequentist probability methods can be used when the hazard does not have many variables and a lot of data is available. However, often there is not enough data for probability distributions, such as in the case of new projects. The risk assessment is often subjective and qualitative, based on expert judgement. The purpose of this research is to present the use of Bayesian networks (BNs) as an alternative to existing risk assessment methods in underground mines by combining expert knowledge with data as it becomes available. Roof fall frequency forecasting using parameter learning is demonstrated with 1141 sets of roof fall data across 12 coal mines in the USA. The prediction is nearly identical for individual mines, but when multiple mines are evaluated it is difficult to find a single best fit distribution for annual roof fall frequency. The BN approach with TNormal distribution was twice as likely to fit the observed data compared to the Poisson distribution assumed in the past. A hybrid approach using BN combining multiple probability distribution curves from historical data to predict annual roof fall is proposed. The BN models can account for variability for multiple parameters without increasing the complexity of the calculation. BNs can work with varying amounts of data, which makes them a good tool for real-time risk assessment in mines.

Jun 1, 2021

By J. van der Walt, W. Spiteri

Flyrock remains a significant threat to the health and safety of mine employees and integrity of infrastructure, as well as to the safety of the neighbouring communities and their property. This investigation was motivated by the general lack of fundamental research and mathematically quantifiable data in the literature regarding the relationship between blast design parameters and their impact on flyrock. The focus was to develop a concept that can be used to mathematically quantify the trajectory of flyrock resulting from a blast, which can be used for future research. The ultimate goal for this technique, once it has been fully developed, is to: ➤ Enable mining operations to generate a database with accurate historical flyrock measurements resulting from their blasting operations ➤ Allow research teams to conduct scientific investigations into flyrock and the impact of various blast design parameters ➤ Generate point-cloud data to visualize blasts and flyrock in a virtual reality environment for training and education purposes. This paper summarizes a conceptual technique and preliminary fieldwork that was carried out to determine the technique’s feasibility and motivate further development. The results show conclusively that a modified photogrammetric technique is capable of capturing flyrock data for further processing and analysis. The data acquisition procedure can, at this point, be used to meet the first aim of the project, namely to gather a field database of historical flyrock generation. Further development of the technique is ongoing and it is envisioned that the scientific-based technique will provide a method whereby future flyrock studies will be comparable and that assumptions will be limited.

Apr 5, 2023

By T. Mapeta, S. M. Rupprecht

Discussions on reasonable prospects for eventual economic extraction (RPEEE) include the CRIRSCO 2019 reporting template, JORC 2012, SAMREC 2016, CIM 2019 Guideline, and Regulation SK-1300. These reporting codes all conform to the principles of reasonable prospects for eventual economic extraction. To demonstrate that all mineral resources align with the RPEEE principles, the Competent Person is responsible for endorsing the mineral resource estimate. Moreover, establishing mineral resources and RPEEE requires technical input from numerous technical specialists, such as mining and geotechnical engineers, metallurgical engineers, social and environmental specialists, financial and governmental practitioners, and more. The aim is to improve the transparency of the disclosure of mineral resources and guarantee the investor or potential investor that the reporting of mineral resources is based on reliable technical, economic, environmental, social, and other relevant factors. This paper discusses the important protocols and definitions in some international reporting codes for the purposes of developing a guide memoire that will assist the Competent Person to interpret, establish and apply RPEEE appropriately and easily.

Jun 27, 2025

By M. I. Mabala, B. P. Watson, M. J. Kanda

A failed and stable Upper Group 2 chromitite pillar database was collected in 2007 as part of the PlatMine 1.2 pillar strength project. The database consisted of 177 pillars, of which 33 had failed. Most of the data were collected from various intermediate-depth mines near the town of Thabazimbi. Spans between rows of chain pillars were about 30 m. Pillar loads were estimated using pseudo-3D, elastic, boundary element modelling software. The pillar-load data were evaluated using a maximum likelihood regression analysis to determine strength parameters, following a similar approach to historically acceptable practice. The original formula was named the PlatMine Formula. This paper describes re-evaluations of the database, using a linear formula and several variations of the power formula. The formula providing the lowest standard deviation to the data in the database was considered the best representation of the pillars in the database. Through this rigorous process, a new formula for peak pillar strength was developed for Upper Group 2 chromitite pillars, which is unaffected by volume. The small standard deviation provided by this formula suggests a reliable relationship between strength and w/h ratio, in the w/h ratio range between 1.5 and 4.7. Accuracy, precision, and recall analysis confirmed the efficacy of this equation to the data in the database. Importantly, the new formula provided in the paper does not predict a significantly different strength to the current PlatMine formula, within the confines of the database.

Nov 27, 2024

The AMIRA Code of Practice and Guidelines for Metal Accounting was developed in response to demand from a number of sponsor companies, including one major platinum producer. Since its publication in 2007, the Code has become widely accepted in the mining industry world-wide as a guide for best practice in metal accounting. With the changes in the structure of the platinum mining industry in South Africa over the last 10 ? 15 years, in particular, the starting up of a number of small platinum producers who must have their concentrate smelted and the contained valuable metals refined by third parties, the need for metal accounting systems that are accurate and have an acceptable level of precision has become even more critical to the success of a platinum-producing operation. The current state of metal accounting in the industry is discussed and possible areas for improved accounting and associated metallurgical efficiencies are identified.

Jan 1, 2012

By B. Maybee, J. Venter, J. Wesseloo

The selection of open-pit slope angles involves high-value decisions with small changes in slope angle, often representing significant changes in net present value. The existing methods used to define design acceptance criteria have evolved from factor of safety to probability of failure (P[F]) to risk consequence to risk frontiers, which represents current state of the art. Current design acceptance measures are based on references to published tables representing the experience and judgement of their authors, who did not specify the purpose of the measures contained therein. As the purpose of the published design acceptance criteria is not specified (i.e., definite stability, marginal stability, optimization, etc.), such tables cannot be used to achieve optimum slope angles that maximize profitability. This paper develops a design acceptance criterion that maximizes profitability by defining a formulation for optimum risk that balances expected risk and reward. The model developed is titled the Mining Risk Model (MRM), and is applied to open-pit slope angle selection through an equation for optimum probability of downside (P[D]O) that balances the upside or opportunity, the P[F], and the downside impact. This formulation for optimum risk is unique, as many authors have presented objective functions for their risk models that can be optimized, but none of the sources reviewed contained a formulation for optimum risk. The MRM is sufficiently flexible to allow the design performance measure that drives P[F], and hence P[D]O, to be selected based on the intended goal. Furthermore, the essential information that must be known to quantify optimum risk is defined. This allows users to determine what information to collect for optimum risk decisions. A further benefit of the MRM is that slope angle decisions and pit shells can be ranked to select the best option, and a threshold is provided that separates acceptable from unacceptable decisions. Finally, the workflow and information required to determine optimum risk are presented.

Sep 9, 2024