Search Documents

By Claude Gosselin

En 1995, la Compagnie minière Québec Cartier a mis de l'avant un projet de recherche sur la modélisation de l'écoulement des gaz pour ses fours de frittage à l'usine de bouletage de Port-Cartier. Cette démarche s'inscrivait dans un large processus de collaboration avec l'Université Laval et le Consortium de Recherche Minérale (COREM). L'usine de bouletage possède deux fours de frittage d'une capacité moyenne de 600 tonnes/heure pour chacun des fours. Les fours de frittage sont des systèmes complexes d'échange de chaleur qui fonctionnent à contre-courant (le débit de boulettes versus les débits de gaz). Pour rencontrer les spécifications reliées aux propriétés mécaniques des boulettes, le processus de fabrication requiert des mesures et un contrôle des systèmes très précis. L'information fournie par plus de 70 capteurs (pression, température, puissance, vitesse, hauteur de lit, position de volet....) est utilisée pour manipuler une vitesse de grille, cinq ventilateurs et neuf zones de brûleurs. Malheureusement, les variables qui influencent fortement les mécanismes de frittage, c'est-à-dire les débits de gaz dans les différentes zones du four ainsi que la masse de boulettes qui circulent ne sont pas mesurées. Une méthode est donc proposée pour une estimation en ligne de ces variables. La méthode est basée sur un arrangement de 65 équations reliant pressions, températures et débits de gaz dans les conduites, les ventilateurs, les sections du lit et les fuites (Gosselin, 1998). Pour des raisons de simplification de cet article, seul un sous-système du four sera présenté. Les techniques de réconciliation des données reliées à la conservation de la matière et de l'énergie montrent que la mesure en ligne du débit de boulettes mène à l'estimation des débits de gaz et apporte des avantages intéressants pour le contrôle des variables stratégiques des fours.

May 1, 2001

By Jr. Blanchette, Claude Bazin, Paul Blatter

"Reporting of the gold production during a month implies to estimate the mass of gold that is processed by a mill, recovered as dorés, lost to the tailings and the change of gold inventory within the plant equipments. Carbon in pulp and/or carbon in leach tanks may contain gold inventory as large as 72 000 g (2 200 troy oz) of gold which needs to be estimated from measurements obtained from sampling the material contained in a tank. Sampling and analysis introduce measurement errors that could be systematic and/or random. This paper examines the effect of measurement errors on the estimation of the gold inventory and on the variation of the inventory in the carbon in leach tanks of the Agnico-Eagle Mines Limited Lapa concentrator. Results of the analysis demonstrate that the inventory is measured with ±10% reproducibility, while the variation of inventory used for mass balance is known with ±300% reproducibility.INTRODUCTIONThe preparation of a production report implies to reconcile measured data with the constraint of material balance over the considered period of production. The sampling and data reconciliation problems were recently reviewed by Morrison (2008) and Jansen (2009a; 2009b). Typically a production report is prepared to cover a period over which the mass conservation hypothesis can be considered valid. In gold processing plants, the change of inventory can seldom be neglected and plays a significant role in material balance calculations. This is particularly true for circuits using carbon adsorption, where the complete cycle of carbon is measured in days.In a 2000 t/d gold ore processing plant, the inventory can be as large as 140 000 g of gold in gravity concentrator bins, carbon in pulp tanks and in the electrolysis circuit. This is equivalent to the mass of gold that fed the circuit during 10 days of operation with 7 g/t feed assays. The change of inventory may therefore strongly impact the production figures and especially the plant recovery at the end of a week or a month. Despite the financial impact associated with gold inventory, few publications are devoted to the subject. Bruington (1991) and Morrison (2008) discussed the problem and stressed the need for the development of reliable methods to estimate the gold inventory in carbon in pulp or carbon in leach circuits."

Jan 1, 2011

Unrelieved residual strains can influence the states of stress in rock masses. If cheap, stable structures are to be constructed in Tock in which large pressures are exterted by unrelieved residual strains, the pressures must be either utilized or neutralized. A photoelastic technique has been used to show that the release of residual strains in rocks is influenced by the mechanical properties of anisotropic rock fabrics. The relationships between directions of residual strain release and the attitudes of minor geological structures have been determined. For engineering design purposes it is possible to estimate the principal stress directions in rock containing unrelieved residual strains if the attitudes of such minor geological structures as foliation planes, lineations, and joints are known.INTRODUCTIONQuarry workers have long been aware that near-surface rocks may be highly stressed. In many cases the stresses are too large to be the result of gravitational loads. Dale (1923, pp. 34-35) gives many instances of deformation produced by the violent release of residual strains in quarry rocks. Coates (1963) has described powerful rock bursts and cases of rock deformation underground implying the existence of much larger stresses than are likely to result from present-day gravitational loads. The only explanation is that pressures of up to several thousands of pounds per square inch may be caused by unrelieved residual strains in rock.Instruments have been developed for measuring states of stress in rock masses. Most stress determinations in rock have been made on the assumption that rock behaves as a simple elastic material. It is now known that rock is a...

Jan 1, 1964

By P. A. Cornejo

Blastings always generates induced vibrations through seismic wave, these are lost energy from the blasting and therefore causing the disturbance of slopes and structures near the blast source. Most of the components of an open pit mine have to reach a physical stability, especially during its closure process. Aiming to develop a vibration prediction model using the historical records from an open pit mine located on Peru, this research has two main focuses depending on the distance from the blast source to the structure: to protect the final slopes and to protect the villagers houses. Both are based on the Devine model of the Square Root Scaled Distance (SRSD) to predict the Peak Particle Velocity (PPV). The Mid field model, where is found mainly final slopes from the pit, limited by the failure criteria of McKenzie based on the geo mechanical properties of the rock mass. The Far field model ranges distances where are located the most community households, limited by the German rule DIN 4150 for highly sensitive structures. Both scenarios resulted on a distribution abacus to limit the weight of used explosives to get the best operational advantage and structure safety.

Mar 2, 2022

By M. Hokkanen, M. Hagfors

The Finnish Defence Forces has destroyed old and spoilt munitions in open surface mass detonations at Hukkakero fell in municipality of Kittilä, Lapland, since 1988. The technique was learned from Sweden. The destruction camp is organized every year at autumn and it lasts 4-5 weeks. During the disposal period one or two charges are exploded per day. The gross weight of a typical charge varies from 150-250 tons including 15-25 tons of explosives. They mainly consist of TNT, but may also contain smaller amounts of amatol and hexotol. As a booster there have been used 2-3 tons of emulsion explosives.

Jan 1, 2015

By Donald Leroux

For the last few years, the grinding model first proposed by Andrews and Mika (1975) has regained attention from the scientific community. Though sophisticated identification algorithms have been developed, the number of its parameters that have to be estimated remains huge. Most of these constants (often more than 90%) are for the breakage function. This paper describes a new method for reducing efficiently the number of the breakage function parameters of the Andrews-Mika grinding model. It makes use of second-order statistics and of a stereostochastic model for liberation (for instance, BooKing).

Jan 1, 1992

By P Ireland, S Mitra, G Evans, A Wang, M M. Hoque

This work presents an experimental study of flow dynamics in a bubble column system (crosssection: 100 mm × 100 mm) using a non-intrusive optical technique with targeted application in a novel REFLUX™ Flotation Cell. The study had two aims – measurement of the bubble size distribution (BSD) and local energy dissipation rate in the system due to bubble motions over a low range of gas flux varied from 0.02 to 0.08 cm/s without impairing the visibility of the system. A central bubble plume was generated by injecting compressed air through a sintered glass distributor in quiescent liquid. Highspeed shadowgraphy imaging was performed to determine the BSD in the presence of surfactant (Sodium Dodecyl Sulfate (SDS) concentration ~ 10 ppm). The BSD exhibited a unimodal shape at different locations of the column for all gas flux cases. A non-intrusive particle image velocimetry (PIV) technique was used to quantify the flow field generated by the bubble motions, which was used to estimate the local energy dissipation rate. The local energy dissipation rate exhibited an increasing power-law dependency on the gas flux.

Aug 24, 2022

By Y. Fu, C. Aldrich

This study investigates the use of dense convolutional neural networks to estimate bubble size distributions in flotation froths. 20 videographically captured images of froths from an industrial platinum flotation plant in South Africa were labelled semantically with an image annotation tool, LabelMe (https://github.com/wkentaro/labelme). As a result, pixels in the images could be categorized into three classes, namely ‘bubbles’ (class 0), ‘bubble boundaries’ (class 1) and ‘background’ (class 2). A U-net architecture with a ResNet34 backbone was trained to segment the images via classification of the pixels in the images. 5 of the 20 images were used to validate the performance of the neural network. Explicit inclusion of the bubble boundaries as a class proved to be significantly more effective than training the network with two classes only (i.e. bubbles and background). However, the thickness of the boundary layer regions were critical. If too thin, it was difficult to distinguish between some contiguous bubbles and if too thick, some fine bubbles may not be identified. The results were promising, although there is still significant scope for improvement, among other by making use of more images for training and validation. Keywords: Flotation Froths, Convolutional Neural Networks, Deep Learning, Bubble size Analysis

Jan 1, 2020

By G. J. Arseneau

Mineral resource statements are reported as tonnage and grades. While great care is placed on generating a reliable estimation of grade, the determination of bulk density for resource estimation is often overlooked. Often mineral resource statements are prepared using a simple average value of bulk density. While a simple average estimate of bulk density can be appropriate for some deposit types, an over simplistic estimation of bulk density can result in serious errors in tonnage estimates and consequently to an over or under estimation of metal content. The construction of a reliable bulk density model requires an adequate bulk density database in order to develop a good understanding of the variability of bulk density with grade and geology. A reliable bulk density database is also essential in determining the appropriate modeling technique to minimize the errors in mineral resource statements caused by incorrect bulk density determinations. This paper summarizes bulk density measuring techniques and draws on examples from various deposit types examining the basic principles of bulk density modeling and domain construction for each deposit type. Examples are presented of simple bulk density domains and more complex grade variable bulk density domains with a discussion of best approaches for minimizing errors in density estimation.

Aug 1, 2013

By B. Oraee

Accurate estimation of equipment costs is a key factor in feasibility study and evaluation of design alternatives of mining projects. In this paper, capital and operation costs of backhoe loaders are estimated using multiple linear regression (MLR), based on principle component analysis (PCA). These cost functions are consisted of five independent variables; bucket size, digging depth, dump height, weight and horse power. The MLR is conducted in two steps. At the first correlation between independent variables is omitted using PCA technique. Thereafter, MLR functions are established using selected significant PCs and total cost functions are constituted as functions of initial variables. At the end, accuracy of functions are evaluated using mean absolute error rate method.

Jan 1, 2011

Careful monitoring of the position of the cave front is essential for well-managed block cave mining. Seismic monitoring systems yield 3-D information about the location of fracturing ahead of the cave. However, because the a seismic gap between cave surface and seismogenic zone is of unknown extent, the cluster of micro-seismicity can only yield an upper bound for the cave height. The large velocity contrast between the fractured rock in the aseismic gap and the broken loose material within the cave can be used by an inversion procedure to find a few simple geometric parameters (such as height of the top of the cave) if a simple 3-D shape?for example a paraboloid?is assumed for the cave. Ray-tracing must be used as the large velocity contrasts cause seismic waves to bend around the cave, but the inversion takes the form of a minimization problem in only a few dimensions and so the computational time is feasible. This technique has been successfully tested on realistic synthetic data.

Jan 1, 2007

By F. Nakhaei, M. Le Roux, Q. P. Campbell, S. Botlhoko

The results of coal sink and float tests are plotted as washability curves. Analysis of these curves indicates the most effective method by which the coal can be cleaned. Sink and float experiments are samplingdependent, destructive, and use toxic chemicals. The present research introduces an alternative method based on a 3D imaging system called RhoVol, in which the density of individual particles is determined from mass and volume measurements. A quantitative error analysis was conducted on coal sample density measurements obtained by RhoVol, and the main factors influencing the errors identified and investigated. The results show that the drawback of this technique is related to its not being able to detect hidden concavities in the particle, resulting in an overestimation of particle volume. Furthermore, it is difficult to capture the impact of porosity and cracks on the volume of coal particles obtained from 3D silhouette images. This mismatch between the RhoVol data and sink and float test results has led to attempts to apply neural network (NN) and linear regression techniques to produce a fully reliable model for correction and estimation of coal density. The NN approach offers superior predictive capability over linear regression, and the estimated density distribution is in line with the sink and float analysis. The root mean square error in estimation of density using the NN model was less than 0.05 g/cm³.

Aug 2, 2022

By Tiago Webber

Mineral resources estimation comprises basically two stages: (i) data collection, analysis and validation, and (ii) the estimate of the grades and tones of the deposit. The most applied method for sampling survey in mineral exploration is rotary drilling with total core recovery. High drilling costs make borehole geophysical logging attractive to obtain the maximum possible information from each borehole helping to understand ore intersections, grades variability, limits of the deposits and continuity of the mineralization. In this paper, borehole geophysical logging was used to enhance the sampling survey and to derive grades indirectly (without lab analysis) with a certain amount of uncertainty. Indicator kriging (IK) was the estimation method used, as it is adequate to incorporate the uncertainty associated with the geophysical derived grade values. The results were compared to the conventional procedures using grades determined via core analysis and ordinary kriging. Borehole geophysical logging proved to be efficient to define the essential features of ore deposit and IK provided results with acceptable precision and accuracy if compared to the model provided by ordinary kriging.

Oct 1, 2009

By Lenin T., Madhav M. R., Padmavathi V.

Use of PVDs and preloading on the ground can significantly reduce the time taken for ground to consolidate. Many soils in situ are lightly overconsolidated, with overconsolidation ratios (OCR) ranging from 1.0 to 2.0 due to various reasons such as glaciation, ground water level fluctuation, aging, cyclic loading, creep, erosion, cementation, etc. The effects of overconsolidation ratio are analysed in this study. Compression index and consolidation coefficient are important parameters in determining time – settlement response of soft soils under preloading and with PVDs. These parameters are typically estimated using oedometer tests on samples, with the assumption that the results represent the entire deposit. The in-situ time-settlement plots at different depths are used in the back analysis for estimating compressibility and consolidation parameters of in situ ground, resulting in more representative values that reflect the deposit's overall behavior. From observed time versus settlements plots, variations of compression index and coefficient of consolidation with depth of a deposit are presented. Estimation of compression index with depth is possible based on different OCR values as shown in this study. Barron (1948), Asaoka methods are used to estimate the coefficient of consolidation with flow in the radial direction.

Sep 1, 2022

By Lizbeth A. Escobar, Phol M. Llanos, Maritza Y. Avila

The present research develops the estimation of critical parameters such as the distances and times between holes to avoid failure due to dynamic pressure in electronic detonators based on the near-field ground vibration model of Holmberg & Persson and the hydrodynamic model of Neyman.

Feb 1, 2020

By R. A. Haney

The Mine Safety and Health Administration has developed a model to estimate diesel particulate (dp) exposures in underground mines. The estimate uses either in-mine measured dp concentrations or engine manufacturers' standardized emission data and through a series of calculations estimates full shift dp concentrations by applying standard engineering principles of ventilation, control technology and work cycle. This model, in the form of a computer spreadsheet, provides the mining industry with a method to estimate diesel particulate levels and determine the impact of various diesel particulate controls on occupational exposure to dp in underground mines. A detailed description of the estimator and several examples demonstrating the results are provided.

Jan 1, 1998

By G. P. Saseen, R. A. Haney

The US Mine Safety and Health Administration (MSHA) developed a model to estimate diesel particu¬late (dp) exposures in underground mines. The estimator uses in-mine measured dp concentrations or engine manufacturers' standardized emission data. Through a series of calculations, full-shift dp concentrations are estimated by applying standard engineering principles of ventilation, control technology and work cycle. This model, in the form of a computer spreadsheet, provides the mining industry with a method for estimating dp levels and determining the impact of various diesel particulate controls on occupational exposures to dp in underground mines. A detailed description of the estimator and several examples demonstrating the results are provided.

Jan 1, 2000

By Tetevi B. Wilson

Analysis of the distribution of ore-bodies in space can be useful generally in the estimation of mineral resources or the management of exploration. This study is di¬rected to analysis of the undiscovered potential of well known areas of mining district size, referred to as metallogenic units (MUs). The analysis employs an effort¬ adjusted and truncated probability model for number of occurrences within a subdivision (quadrat) of a MU and Monte Carlo sampling to generate an approximation to the probability distribution for number of occurrences and number of mines within an MU when it is totally explored. Exploration potential for Monitor, Bodie, Au¬rora and Camp Douglas MUs (Walker Lake quadrangle of Nevada and California) are estimated to be 9, 4, 7 and 4 mines, respectively.

Jan 1, 1992

By Firat Atalay

Estimating the spatial distribution of ore grade is one of the most critical steps to proceed with an investment decision on a deposit. Kriging is the most widely used method to estimate the ore grade, but alternative techniques are being developed. Machine learning algorithms can be used as alternative methods to classical kriging. In this paper, the iron (Fe) grade of a deposit is estimated with the XGBoost algorithm, and the results are compared with kriging estimation results. For the estimations, samples collected from drillholes were used. Due to the different natures of the estimation methods, different steps were taken to perform the estimations. The results show that XGBoost produced higher-range estimates, which is a desired result in ore grade estimation, and the minimum and maximum of the estimates were lower and higher than the kriging estimates, respectively. However, like kriging, the estimation results were smoother than composites, and the variance of the XGBoost estimates was lower than the variance of composites. This means that even though estimation with XGBoost mitigates the smoothing effect, the estimation results suffer from smoothing like kriging.

Dec 1, 2024

By Kwangmin Kim, Junnhyeok Park

In the hard-rock mining industry, the generation of fines during blasting is an important issue because it has a significant effect on production processes. Although various approaches have been used to estimate fines generation in recent decades, as a practical matter, measuring the amount of fines is nearly impossible given the technical limitations and the scale of mine blasts. In this study, a dynamic PPV damagemodel was developed and applied to predict fines generation as a result of blasting. This modified version of the crushed zone model (CZM) considers dynamic rock properties and peak particle velocity (PPV), which represents the induced stress in blasting. Dynamic rock-fracturing tests were conducted using a split Hopkinson pressure bar (SHPB) to determine the parameters that represent the relationships between rock strength and strain rate as well as the relationships between fines amount and strain rate in the model. A case study was conducted that entailed determining the radius of the crushed zone for given blast designs and rock properties and then measuring the fines amount within this crushed zone. This case study indicated that the dynamic PPV model can provide a rough estimate of fines generation for given dynamic rock properties and blast design parameters and that this method has the potential to be a useful indicator, or index, of fines generation.

May 16, 2020