Search Documents

By S. M. Spottiswoode, J. A. L. Napier, M. Kataka

The move away from the use of strike stabilising pillars at depth and the adoption of the more flexible system of mining towards final dip remnant pillars brings new opportunities for optimising mine design. In the initial planning stage, the size and spacing of the final pillars is designed using classical design criteria, through limiting modelled values of Energy Release Rate (ERR), Average Pillar Stress (APS) and Excess Shear Stress (ESS). In reality, mines often leave pillars that are larger than those originally designed. One of the reasons for leaving more ground is an unexpectedly high rate of seismicity. In this paper, we present developments in integration of modelling and seismicity that are aimed at developing design criteria that are more responsive to local conditions.

Jan 1, 2003

By M. Tangstad, B. Monsen, H. Midtgaard

The increased attention towards reducing emissions of fossil CO2 due to its potential global warming effect has resulted in a focus on renewable resources, as charcoal. Since charcoal is produced from growing wood it does not contribute to global warming. This paper gives a brief review of consequences of and reasons for using charcoal in the silicomanganese (SiMn) production. Results are presented from practical SiMn smelting experiments, on a pilot scale, using charcoal as reduction material and comparing with coke. As slag- and CO2-reactivity are important for furnace performance these features have also been measured.

Jan 1, 2004

By D. G. Krige

For the purpose of determining confidence limits for ore grade estimates, an extensive simulated data base was used, which included 1728 ore blocks with ‘actual’ grades and the corresponding estimates based on ordinary and simple kriging. The practical results show that conditional unbiasedness is essential and that acceptable confidence limits can then be set using Normal theory for the simulated error distribution based on the logarithms of the ratios of ‘actuals’/estimates with a variance related to the kriging variance. The mean of the data used to prepare the variogram and the estimated block grade are to be used properly in this approach. A practical analysis was also done of the actual point and block data from a very large mined out Carbon Leader area. The distributions of the actual grades for two levels of block sizes with different levels of variability were fitted using three models, ie. 2-. 3- and compound-lognormal. The trends of tonnages, grades and relative profits for a range of pay limits, as well as the usual tonnage-grade curves, show a practical insensitivity for the choice between the three lognormal models considered.

Jan 1, 2003

By B. E. Ydstie, B. F. Lund, K. R. Løvåsen, B. A. Foss

The paper presents sensitivity analysis for a dynamic model of submerged arc silicon furnaces. The model we study, called ?Simod?, was developed by Elkem in cooperation with NTNU and Sintef. It is today a much used tool in Elkem for operator training and analysis of the dynamic behaviour of the silicon production process. The aim of our work is to bring the simulator tool closer to industrial practice, and construct a mechanism to adapt the model to the behaviour of a running process, online. The paper presents the results of the sensitivity analysis that has been undertaken for important parameters in the Simod model. The parameters associated with the process inputs are electric power loss, loss of fixed C and reactivity of carbon. The output signals that have been investigated are fed quartz, tapped silicon and silica. The paper also interprets the results of the sensitivity analysis in a process operation context.

Jan 1, 2004

By D. R. Hack, N. M. Sirakov, M. L. Feves, M. Iwanowski

This paper presents a new approach capable of volume calculation of 3D reconstructed geological objects. The method is based on a well-known concept that states the volume of an object can be computed as the limit of the sum of the area of infinite number of sections, which cut the 3Dobject. Since shortage of input data about the geological objects usually exists, Mathematical Morphology and polynomial interpolation are employed to generate a set of intermediate sections between the set of initial sections. A Mathematical Morphology tool is applied to compute the area of the two of sets initial and generated 2D sections. Regularities together with Delaunay Triangulation are employed to perform 3D reconstruction. A formula is developed on purpose of volume calculation of the reconstructed objects. A set of ten theoretical, electrical resistivity cross-sections were generated and used to perform an experiment to show the capability of our approach to estimate the volume of complex 3D geological object in case of sparse 2D sections. The volume determined using the morphological approach presented in this paper show good correlation with the volume of the original input object. Advantages and disadvantages of our volume calculation approach are discussed in the conclusion of this paper. Keywords: electrical resistivity, sparse 2D sections, geological objects, smooth 3D visualization and volume calculation.

Jan 1, 2003

By C. F. Vos, J. D. Miller, J. C. Davidtz

Small replacements of C12-trithiocarbonate (C12-TTC) for dithiocarbonate (DTC)/dithiophosphate(DTP) in a reagent suite for the flotation of platinum group metals (PGM) have significantly improved the rejection of gangue. Coupled to this is an improvement in kinetics, grade and recovery of PGMs. In a pilot-plant trial, after a 5 mole per cent replacement of the current DTC/DTP collector combination by C12-TTC, large opaque bubbles were replaced by small clear ones. In addition, a similar flotation response was realized with one-third of the current depressant dosage. During a preliminary plant trial on underground UG2 ore, the initial rougher concentrates were characterized by higher PGM concentrate grades and this was in part due to improved chromite rejection. The concentrate produced from the TTC trial also exhibited a much finer particle size distribution. The overall circuit performance, which included secondary regrind and flotation, showed an increase in concentrate grade of approximately 44% while the tailings grade remained unchanged

Jan 1, 2006

By W. Robert, J. Wirth, G. Grasselli

ABSTRACT: Strength, deformability and fluid flow properties of rock joints are to a great extent controlled by the surface roughness. The hydraulic conductivity of a rock fracture depends on the aperture distribution, the surface roughness and the contact area, each of which depend on the stresses that act on the fracture plane. However, despite much recent work, a complete understanding of the relationship between void space geometry, contact areas, applied stress, and hydraulic conductivity has yet to be achieved. Moreover, although in situ rock joints are subjected to both normal and shear loading, few studies have been conducted to investigate the effect of shear displacements on fluid flow through a single rock fracture, and which have addressed the issue of flow anisotropy induced by shearing. Experimental flow tests show that with increasing shear displacement, the fracture becomes heterogeneous and anisotropic, and becomes more permeable in the direction parallel to the displacement (Yeo et al., 1998). On the other hand, dry shear tests show that the surface fails in areas facing the shear direction (Grasselli and Egger, 2003). Moreover, the experimental results confirm that the shear strength of rock joints is direction-dependent. In this paper, a three-dimensional morphological description of several a natural fractures that have undergone shear displacement were generated using the Automated Topographic System at the University of Basel. The data analysis method of Grasselli (Grasselli et al., 2002a) is used to determine the surface parameters that are known to control the shear strength. Simulations of fluid flow through this fracture, in various directions relative to the direction of shear, are conducted using an FEM flow simulator. The results constitute the most detailed analysis yet undertaken of the relationship between shear displacement, shear strength, and fracture transmissivity. RESUME: La rugosité des joints en roche joue un rôle primaire sur la réponse hydromécanique du massif fracturé. En effet, la conductivité hydraulique des discontinuités en roche est influencée par la rugosité, et par la distribution des ouvertures et des zones de contact, dont chacun dépend de l’effort normal et du cisaillement qui agissent sur la surface du joint. Les essais d'écoulement en laboratoire montrent que si cisaillée, la fracture devient anisotrope, et plus perméable dans la direction parallèle au déplacement (Yeo et al., 1998). D'autre part, les essais de cisaillement indiquent que seules les zones de la surface qui font face à la direction de cisaillement sont impliquées dans le cisaillement (Grasselli et Egger, 2003). D'ailleurs, les résultats expérimentaux confirment que la résistance au cisaillement des joints de roche est dépendante de la direction. Dans cet article, plusieurs fractures en roche ont été analysées en utilisant un système de mesure topographique à l'Université de Bâle. La méthode d'analyse de données proposée par Grasselli (Grasselli et al., 2002a) a été employée pour calculer les paramètres de la surface qui déterminent la résistance au cisaillement. Plusieurs simulations du coulage, dans diverses directions relativement à la direction du cisaillement, ont été faites à l’aide d’un programme de calcul à éléments finis. Les résultats constituent une des analyses les plus détaillé sur le rapport entre le déplacement pendant le cisaillement, la résistance au cisaillement, et la transmissivité spécifique de la fracture. ZUSAMMENFASSUMG: Die Festigkeit, Verformbarkeit und Hydromechanik von geklüftetem Fels wird hauptsächlich von der Oberflächenrauhigkeit der Bruchflächen bestimmt. Die hydraulische Leitfähigkeit einer Kluft hängt von der Verteilung der Rissbreiten, der Oberflächenrauhigkeit und den Kontaktflächen ab, die ihrerseits von dem auf die Bruchfläche wirkenden Druck beeinflusst werden. Obgleich in letzter Zeit viel in dieser Richtung geforscht wurde, kann der Zusammenhang zwischen Porenraumgeometrie, Kontaktflächen, wirkendem Druck und hydraulischer Leitfähigkeit noch nicht vollständig erklärt werden. Obwohl Klüfte in situ sowohl Normal- als auch Schubspannungen ausgesetzt sind, wurde der Einfluss von Scherverschiebungen auf den Kluftwasserfluss und die daraus resultierende Fliessanisotropie in einer einzelnen Kluft bisher kaum untersucht. Experimentelle Untersuchungen des Durchflusses zeigen, dass die Kluft für größer werdende Scherverschiebungen heterogen und anisotrop wird wobei die Permeabilität parallel zur Verschiebungsrichtung ansteigt (Yeo et al., 1998). Andererseits zeigen trockene Scherversuche, dass die Oberfläche an Stellen versagt, die gegen die Verschiebungsrichtung anstehen (Grasselli und Egger, 2003). Des weiteren bestätigen die experimentellen Ergebnisse eine Richtungsabhängigkeit der Kluftscherfestigkeit. In diesem Artikel wird eine dreidimensionale morphologische Beschreibung mehrerer natürlicher, durch Scherspannungen verschobener Klüfte vorgestellt, die vom Automated Topographic System der Universität von Basel erstellt wurde. Das von Grasselli (Grasselli et al., 2002a) entwickelte Datenanalyseverfahren wird zur Feststellung der die Scherfestigkeit bestimmenden Oberflächenparameter herangezogen. Mittels einer FEM-Simulation wird der Durchfluss durch eine solche Kluft in verschiedenen Richtungen simuliert. Die Ergebnisse stellen die bisher ausführlichste Analyse des Zusammenhangs zwischen Scherverschiebung, Scherfestigkeit und Durchlässigkeit (Transmissivität) einer Kluft dar.

Jan 1, 2003

By P. J. Brereton-Stiles, M. S. Rennie, N. Singh

Ferroalloy producers are large-scale users of electrical energy, which in many countries is in short supply and increasingly expensive. Preheating and/or pre-reduction using natural gas, recycled furnace gas or coal fines have lowered the specific use of electricity, but the high energy concentration necessary for most reduction processes can be achieved only by the use of high electric currents. Most countries have insufficient electricity to supply the demands of industry and the domestic consumer, particularly during certain periods of the day and certain times of the year. The generation and distribution of electricity is a large-scale expensive business, which is financed and managed directly or indirectly by government bodies in most countries. Both economic and political imperatives determine how much electrical energy is made available to whom and at what price. In South Africa demand is set to outstrip supply within the next few years, and the local supplier Eskom has moved from a traditional demand tariff pricing structure to a time-of-use tariff pricing so as to limit peak demand usage. This will mean a substantial increase in electricity cost to those users unable or unwilling to reduce consumption during peak periods. The optimal use of the available electrical energy depends both on the process, the furnace size, the configuration and infrastructure of each plant and its cost structure. As the cost and availability of electricity is set to change dynamically, the management of the situation becomes a real time control and optimisation problem. This paper discusses some the process and operational constraints associated with low- or no-load conditions and how to manage the situation. Electrode management will be one of the key issues.

Jan 1, 2004

By S. Ganguly, R. Kononov

Carbothermal reduction of manganese oxides and manganese ores in the solid state was studied in hydrogen, helium and argon at different temperatures, ore compositions and carbon to manganese oxide ratios. Wesselsore (South Africa) and two grades of the Groote Eylandt ore (Australia) with different level of impurities were examined. Ores were characterised by XRD, X-ray fluorescence, optical microscopy and scanning electronmicroscopy. Isothermal and temperature programmed carbothermal reduction experiments were conducted in a fixed bed reactor in a vertical tube furnace, with on-line monitoring of gas composition by the CO-CO2 infrared sensor. The reduced samples were characterised by XRD, SEM and LECO analyses. Extent of reduction was calculated using data on the off-gas composition, and LECO carbon and oxygen contents in the reduced sample. Manganese oxides in the ore were reduced to a-Mn and carbides Mn23C6 and Mn7C3 depending on the carbon to ore ratio. The reduction rate of manganese ores in hydrogen was higher than in helium and argon. Key words: Manganese, Carbothermal Reduction, Hydrogen, Argon, Carbon Monoxide, Helium, Kinetics.

Jan 1, 2007

By C. E. Doran

Since the early 1990s the underground mining industry has embarked on a vigorous program of improving process and eliminating costs. A leading development in this program has been the implementation of leaky feeder radio systems as an enabling infrastructure for future automation. El-Equip Inc. has led this development with more than 175 mines implementing mine-wide radio infrastructures for voice and data. In this paper, we describe the implementation of a digital radio network which supports existing installations through to an automation infrastructure employing full digital cellular communications with mobile video roaming capabilities. We outline the evolutionary path from existing low speed data communications links through to 155 Mbps and beyond.

Sep 1, 2001

The modelling of lithological domains is a critical step in mineral resource and reserve evaluation. Deterministic models rely on an interpretation of the available drill hole data and expert knowledge, but do not account for the uncertainty in the spatial extent of the lithological domains. Instead, probabilistic models based on conditional simulation can be used to map the probabilities of occurrence of the domains within the deposit, which reflect the uncertainty in their presence or absence at unsampled locations. The mineral grade model is then obtained by weighting the predicted grades associated with each lithological domain by the probability of occurrence of this domain. Such an approach accounts for the continuity of the grades proper to each lithological domain and for the uncertainty in the spatial extent of these domains within the deposit. An application to a porphyry copper deposit is presented. Keywords: spatial uncertainty; plurigaussian simulation; geological control; kriging; geostatistics.

Jan 1, 2007

By M. C. Bétournay

Strainbursts are common in overstressed brittle rock around deep mine openings. Such failures are characterized by a biaxial state of stress in the plane of the opening boundary. A zone of compressive stresses imposes initial failures near or at the surface of the opening that results in spalling of thin slabs of rock roughly parallel to the free face. While the failure of rock under biaxial loading conditions is increasingly common in deep underground mines, only a few representative laboratory studies have been reported in the literature, without emphasis on rockbursting aspects, and none in the last decade. This paper presents the results of a testing program on a granite rock type. Biaxial testing, conducted on 100 mm cubic samples, was carried out to examine the influence of lateral confinement and loading rate on the rock’s ultimate strength. The paper reports the test results in terms of normalized biaxial failure envelope s1/sc-s2/sc. A discussion of the failure mechanisms observed is presented. Les coups de toits de déformations sont choses communes dans le rock cassant surcontraint en place autour des ouvertures minières profondes. De telles ruptures sont caractérisées par un état de contraintes biaxiales dans le plan de la périphérie de l’ouverture. Une zone de contraintes compressives impose des ruptures initiales à ou près de la surface de l’ouverture, qui produit des relâchements de minces plaques de roc plus ou moins parallel à la surface. Bien que les ruptures de roc sous des conditions biaxiales sont de plus en plus communes dans les mines à grandes profondeurs, il n’y a eu que quelques études représentatives en laboratoire de publiées, sans emphase sur des aspects de coups de terrains, et aucunes publiées depuis la dernière décennie. L’article présente les résultats d’une campagne d’essais appliquée à un type de granit. La campagne d’essais biaxiaux, sur des échantillons cubiques de 100 mm, avait pour objectif d’examiner l’influence du confinement latéral et le taux de chargement sur la résistance ultime du roc. Cet article fait part des résultats en terme de l’enveloppe de rupture biaxiale normalisée s1/sc-s2/sc. Une discussion sur les mécanismes de ruptures observés est présentée. Dehnungsbrüche findet man häufig in übergespannten spröden Felsen um die Öffnung von tiefen Minen. Solche Brüche sind von einer zweiachsigen Spannung in der Ebene der Grenze der Öffnung charakterisiert. Eine Zone von Druckspannungen verursacht initiale Brüche, entweder an der Öberflache der Öffnung oder in der Nähe. Davon scheiden sich dünne Felsplatten mehr oder weniger parallele mit der freien Wand. Obwohl Felsbrüche unter zweiachsiger Beanspruchung immer mehr häufig in tiefen Minen beobachtet sind, hat man über nur einige repräsentative Laborversuche in der Literatur berichtet. Diese Studien diskutieren das Thema ohne Betonung auf Felsbrüchen, und keine sind in den letzten zehn Jahren erschienen. Die vorliegene Arbeit gibt die Ergebnisse eines Testprogramms auf einem Typ von Granitfels. Durch zweiachsige Prüfungen von kubischen Proben mit Seitenlangen von 100 mm sollte man den Einfluss von seitlicher Ermittlung und der Beanspruchungsrate auf die Bruchfestigkeit des Felsen studieren. Man findet auch die Prüfergebnisse in bezug auf das normalisierte, zweiachsige Bruchkriterium s1/sc-s2/sc. Die beobachteten Bruchmechanismen sind auch diskutiert.

Jan 1, 2003

By P. Dennis, T. Pieters, A. De Kievit, S. Ganguly

TEMCO has been producing manganese ferroalloys in submerged arc electric furnaces for over 40 years. The furnace linings all along have been of conventional insulation type until 2001 when a freeze lining of UCAR concept was installed for a 20 MW Furnace. Continuous improvement of furnace lining performance has been a key component in improving the business competitiveness of TEMCO. The expected minimum campaign life of the freeze lining is fifteen years as compared to the ten years of the conventional linings. The success of the freeze lining is very critical to the quanta and direction of heat flow through the lining that control the freeze protection. Under freeze protection can lead to reduced lining life and over freeze protection, on the other hand, can lead to operating difficulties and loss of smelting efficiency. It was, therefore, necessary to develop a computer based system to monitor and control the freeze protection. The computer based monitoring and control system uses the basic principles of heat transfer, activeX object and heat flux from dual thermocouples and that is reviewed.

Jan 1, 2004

By Jeremy Bosman, Johan de Korte

Traditionally the feed to coal preparation plants was de-slimed at 0,5 mm. The +0,5 mm size fraction was processed via dense-medium cyclones and the -0,5 mm size fraction was beneficiated with froth flotation, water-only cyclones or spirals. With the introduction of large diameter cyclones, the aperture sizes of de-sliming screens have progressively increased to about 3 mm in the belief that the larger cyclones are not able to effectively beneficiate coal which has a particle size below the ?breakaway? size. The efficiency of separation in the large dense-medium cyclones is improved by presenting a coarser feed to the units. At the same time, the capacity of the de-sliming screens is increased due to the larger screen apertures and in effect, the throughput capacity of the dense-medium section is increased. The 3 x 0,5 mm size fraction, which was previously processed in the dense-medium cyclones, is now added to the fine coal size fraction and processed by water-only methods such as Teetered Bed Separators (TBS) and spirals. Whether this practice is the most economical way to process the 3 x 0,5 mm size fraction requires further thought since the best processing route for the 3 x 0,5 mm size fraction will depend on a number of factors, amongst which are: o The washability of the coal and the amount of near-density material o Processing efficiency o Specification of the required product o The price paid by the customer for the product o The cost of processing the coal Breakaway size The ?breakaway? size is defined as the particle size below which the Epm of the smaller particles starts to decrease significantly. Bosman1 provides an approximate breakaway size versus cyclone diameter (Figure 1). It should be kept in mind though that the Epm of the small particles is a function not only of the cyclone diameter but also of feed pressure, medium density, medium viscosity, particle shape and density, top-size of particles in the feed, etc.

Jan 1, 2006

By P.J. Terbrugge

Open pit slope design has conventionally been effected as a bottom-up function utilizing available geotechnical information. This results in a decision criterion based on probability of failure and factor of Safety with a risk assessment being carried out on the proposed design slopes. The design approach recommended reverses the process using fault event tree decision methodology. The consequences of this approach are that acceptable risk criteria have to be determined by mine owners. Slopes are then designed by the technical staff to achieve these corporate goals. Benefits that arise from the process are that the owners take a proactive decision on the risk-benefit relationship allowing the technical staff to optimize the geotechnical exploration programme and design.

Jan 1, 2006

By F. M. G. Egerton

Mechanisation of mining operations is not a new concept in the world of mining. Machinery has been developed, manufactured and used for many decades in massive orebodies and coal mines, both locally and abroad. Because of their size and the huge tonnages available, massive ore bodies lend themselves very easily to mechanisation. However, little progress in mechanisation has been made in narrow tabular orebodies in South Africa. The equipment available for narrow tabular orebodies requires excavations of a size that introduce large volumes of waste dilution. Nevertheless, it is essential for South African miners to experiment with mechanisation and to find a solution to the problem. This address looks at current conventional mining methods and compares them to bord-and-pillar type mining methods. Off-reef mechanisation is also touched on. It then looks at different hybridised mining methods that are being used, i.e. a mixture of conventional mining and mechanisation. It finally looks at future prospects for mechanisation in narrow tabular orebodies.

Jan 1, 2004

After the closure of collieries, they naturally start to fill up with water. As a result, hydraulic gradients develop between them and different hydraulic pressures are exerted onto peripheral areas or compartments within mines. This results in water flow between mines, or onto the surface. This flow is referred to as intermine flow. The collieries in the Witbank Coalfield have geometries such that there are several areas where this intermine flow is possible. Since the Department of Water Affairs and Forestry has declared intermine flow as one of its greatest concerns for granting closure to South African mines, much research into this phenomenon is required. The challenges in determining intermine flow are numerous, and attention has focused on identification of areas where these flows can take place. The quantification of these flows is problematic due to the uncertainties in exact geometric configurations and the variation in site-specific hydraulic properties of thecoal and overlying lithilogical layers. During this study, use has been made of numerical flow modelling and several analytic solutions to test the applicability of the flow models as well as to predict groundwater flow directions, filling times of voids and flow volumes. The numerical modelling methodology entailed a downscaling approach starting with a broad regional model covering the entire area, followed by modelling the interactions between interconnected mines, and finally looking in detail at the major areas of interaction.

Jan 1, 2005

By H. Hurkchand

Background: A cross-sectional study was conducted in November2001 to establish the prevalence rates of Human Immunodeficiency Virus (HIV) and sexually transmitted infections (STIs) (Chlamydia trachomatis and Neisseria gonorrhoea) in a coalmining town in Mpumalanga. Methods: 155 adults were recruited (79 males and 76 females), by holding community meetings, and by door-to-door recruitment. Saliva and urine specimens were collected for the analysis of HIV and STI prevalence, respectively. Results: The mean age for males was 31.35 years (min = 19 yrs and max = 50 yrs) and the mean age of the female participants was31.90 years (min = 17 yrs, max = 49 yrs). The prevalence of HIV was found to be 22.78% [95% CI (14.09?33.59)] in the males vs. 47.37%[95% CI (35.79?55.16)] in the females, [p = 0.001, OR = 3.05, 95%CI(1.49?6.26)]. The prevalence of C. trachomatis was 7.90% [95% CI(2.95?16.39)] in the males vs. 12.86% [95% CI (6.05?23.00)] in the females, p = 0.324. The prevalence of N. gonorrhoea was 3.79%[95% CI (0.79?1.08)] in the males vs. 9.86% [95% CI (4.06?19.26)]in the females, p = 0.137. HIV was significantly associated with C. trachomatis, [p = 0.032, OR = 3.18, 95% CI (1.04?9.72)], but was not significant when stratified by gender. HIV was not significantly associated with N. gonorrhoea (p = 0.072) but it was significantly associated when stratified by gender, p = 0.001 for females. There were no significant associations between HIV and age (p = 0.408), even upon stratification by gender. Conclusions: HIV prevalence in this community is high, and corresponds to national figures. However, the prevalence of STIs is surprisingly low as it would be expected to be high in a migrant population. Although communities are exposed to expanding peer-education activities that encourage behaviour change, the prevalence of HIV in this group. There is an urgent need for interventions designed to treat or prevent HIV infection in women generally and in women at high risk. Keywords: HIV prevalence, STI prevalence, antenatal, mining

Jan 1, 2005

By L. Archambeault

This paper describes preliminary research on the application of Markov Decision Processes (MDP) to Real Option Valuation (ROV) and the optimisation of mine scheduling. The MDP framework is a novel approach to option valuation and scheduling in mining operations. A learning agent is introduced into the valuation process of an open pit, where prices and block ore grades have probabilistic values. The prices are modelled using a mean reverting diffusion process and the block grades using sequential Gaussian simulation. The agent is asked to learn which production parameters should be used in order to maximise the overall value of the project. The introduction of the agent permits a real option approach to mine valuation, such that the value associated with the robustness of a design to uncertainty can be measured. A simulated example is used in which there are ten blocks to be extracted under conditions of grade and price uncertainty. Using policy iteration, an optimal policy is generated and the value of production options is found. The potential financial gains from applying MDPs to mine valuation and optimisation are substantial and warrant further investigation.

Jan 1, 2007

By M. Maleki

Variogram analysis and kriging lack robustness in the presence of outliers and data with long-tailed distributions, which often arises when estimating grades in precious metal deposits. The capping technique, consisting of truncating the data to some top-cut grade, is widely used in order to mitigate the influence of the values in the upper tail of the distribution. However, this procedure omits part of the grade variability and is likely to provoke a bias in the estimates. To avoid these issues, a recently proposed approach is to decompose the grade of interest into three components (the truncated grade, a weighted indicator above the top-cut grade, and a zero-mean residual) and jointly estimate the truncated grade and the indicator by cokriging. This approach is attractive as it provides unbiased grade estimates, allows choosing the ?optimal? top-cut value, and essentially works with truncated and indicator data, thus avoiding the use of outlying values for calculating sample variograms and performing spatial interpolation. This work presents an application of this approach to a disseminated gold deposit that has been identified through exploration drilling. The effect of using an indicator covariate is assessed through leave-one-out cross-validation, by comparing the grade estimates with the true grades and with the grade estimates obtained with the conventional capping approach, which considers only the truncated grade as the variable of interest. As a result, cokriging the truncated grade and the indicator above top-cut grade outperforms the conventional capping approach, yielding significantly more accurate estimates. A few complementary guidelines are provided for validating the model hypotheses and for the implementation of cokriging.

Jan 1, 2014