Search Documents

By Andre Zingano

An underground mine using room and pillar and retreat mining methods was developed during the 1950?s. In 2009 an open cast mine was opened in the same reserve with the objective of mining five coal seams including the one previously mined. Recently, a series of slope failures were detected at the open cast mine. The surface mine map was overlain on the underground mine map, and it was observed that the slope failures were located above a retreat mined area where the overburden had caved and the ground subsided. The deformation and displacement attributable to subsidence caused a loss of strength in the overburden strata. The objective of this paper is to study the causes of slope sliding (or failure) due to the ground subsidence of rock layers above oldmined out areas. A map was created to overlay the two mines -- the abandoned underground mine and the active surface mining. Two- and three-dimension numerical modeling was conducted to simulate the slope behavior and to verify whether the mine subsidence was the cause of slope instability. This study indicated that the subsided rock layers were the cause of slope instability and sliding.

Jan 1, 2013

By Xianyang Qiu, Xiuzhi Shi, Chen Xie, Xiaofeng Huo, Yanhai Wang

In upwards medium-deep hole cut blasting, many confinement effects exist due to the small number of free surfaces, which is mainly provided by an empty hole. If the cut blasting effect cannot meet expectations, then subsequent blasting construction will be seriously affected, and large security risks to operators in the stope can be induced. The number of free surfaces is directly related to the layout of the cutting borehole. Therefore, it is necessary to optimize the layout of the first round of cutting boreholes. Theoretical analysis, numerical simulations, and field tests were used to study the formation process of cut blasting and optimize the layout of the first round of cutting boreholes. Based on the blasting parameters in the field, LSDYNA software was used to establish cut blasting models with four kinds of borehole layouts in the first round. The RHT constitutive model was used to simulate the rock-breaking effect. The numerical results show that the cut blasting effect is optimal when Nos. 3, 5, 7, and 9 are charge holes and Nos. 1, 2, 4, 6, and 8 are empty holes. In addition, the greater the number of empty holes, the larger the release space and the better the cut blasting effect. The uniformity of the empty hole distribution, namely, the uniformity of the release space, also affects the cut blasting effect. The field results indicate that a good cut blasting effect can be achieved by optimizing the borehole layout.

Sep 7, 2023

By Chunlin He, Xinsheng Li, Feng Yang, Lu Han, Yuezhou Wei, Feng Gao

The solvent extraction of vanadium (V) from Na2CO3–NaHCO3 solution was attempted using Aliquat-336 (methyltrioctylammonium chloride) extractant dissolved in sulfonated kerosene as diluent with 2-octanol as phase modifier. The influence factors, such as Na2CO3 concentration, NaHCO3 concentration, temperature, extractant concentration and impurity P, on vanadiumextraction were investigated. The results revealed that the single stage extraction efficiency of vanadium reached above 89%from 20 g/L Na2CO3 – 20 g/L NaHCO3 solution under the following conditions: extractant concentration of 15%, room temperature of 20–30 °C, 2-octanol of 6%, phase ratio (O/A) of 1:1, and oscillation time of 1 min. The addition of NaHCO3 into Na2CO3 solution could lower the pH value of the Na2CO3 solution, improve extraction efficiency of vanadium, and not introduce other impurities. The phosphate anion did not interact with the vanadate anion in the Na2CO3–NaHCO3 solution. The P in the Na2CO3–NaHCO3 solution had little influence on vanadium extraction.

Jul 30, 2020

By CH. Liu, H. X. Jiang, X. Xiao

"A study has been conducted to characterize the fatigue life in 7050-T7451 thick plate by the comparison between an experimental alloy and commercial alloy. The fracture origins were specified and their sizes were measured by scanning electron microscope (SEM). Additionally, the size, number density and spatial distribution of porosity in the experimental material and commercial material were characterized and compared, using X-ray Computed Tomography(XCT). The results showed that microporosity was the predominant site for crack nucleation and subsequent failure of the experimental alloy. However, as for the commercial alloy, fatigue cracks were found to originate from constituent particles. The presence of microporosity in 7050-T7451 thick plate reduced fatigue life of an order of magnitude or more compared to the commercial alloy. The relationships between size of origin and fatigue life were proposed. It was showed that the fatigue endurance was reduced when the size of porosity increased.INTRODUCTION 7xxx series aluminum alloys, due to their combination of high strength, high resistance to stress corrosion cracking, and good fracture toughness, are widely used in structural engineering applications where high-strength and low-density characteristics are fundamental design requirements (Williams& Starke, 2003). Every designer is striving to find an optimal combination of properties including strength, corrosion resistance and fatigue resistance (Wanget al., 2016). Based on the design principles of aircraft, a higher fatigue resistance of aluminum alloys are required. Thus, many investigations have been carried out on the microstructures and its effect on fatigue properties of the 7xxx series aluminum alloy (Robson, 2004; Schubbe, 2009).Microporosity, which is a common defect in casting materials, can lead to a significant deterioration of mechanical performance and fatigue resistance, especially in the high alloyed 7xxx aluminum alloys. The actual valid loading area decreases due to microporosity, which results in detrimental effect on the mechanical properties. Many experimental studies have been conducted in this research field. Mayer et al.(2003) investigated the fatigue properties of magnesium alloys produced by high-pressure die-casting. The results show that porosity in magnesium alloys greater than0.18mm diameter may be considered as the most important material defect for high cycle fatigue and fully reversed loading conditions. Nadot et al.(2004) found that the near surface defects are much more dangerous than the internal defects. Wang et al.(2001a) used the Weibull statistics method to analysis the influence of the microporosity, oxides and particles on the A356-T6 alloy fatigue life. The results show that, the porosity is more detrimental to fatigue life than oxide films and particles. With regard to fatigue resistance, although the deleterious effect of microporosity has been generally acknowledged, a quantitative understanding between casting microporosity and fatigue life in 7050-T7451 plate is not well established."

Jan 1, 2018

By Larry F. Drolet

Charleston, SC is anything but boring. The city was founded in 1670 and has played a significant role throughout the history of the United States. Located along the coast of the Atlantic Ocean in the southeastern United States, Charleston was a vital city in America?s early existence. Charleston?s coastal location has helped the city become the fourth largest cargo port in the United States and earn the 2011 designation as the number one travel destination in the United States by the Conde Nast Traveler Magazine. Part of Charleston?s charm and beauty can be attributed to its historic structures and churches that line nearly every sidewalk, roadways that were laid out based on the width of horse carriages, great weather, southern hospitality and coastal living. Now imagine the challenges of installing new, or re-placing aging infrastructure, in this setting. Every shovel full of dirt, every stick of pipe and every roadway sub-surface has its own historic story to tell. For engineering minded individuals, Charleston?s infrastructure has almost as fascinating a history as its more visible surface structures. Charleston was among one of the first American cities to construct separate sanitary and storm water drainage infrastructure. This concept has roots dating back to the city?s founding and is a credit to the foresight of its founding fathers who understood the basic health differences between sanitary and storm water.

Jan 1, 2013

By R. V. Ramani, P. M. T. White, D. Sutton

1.1 General The examination and evaluation of a mineral deposit to determine whether or not it can be profitably exploited entails many factors unrelated to the physical nature and intrinsic worth of that deposit. Nevertheless, geology and geologically related factors have great bearing on the economic potential of the deposit. This is only natural because the deposit owes its existence to these factors. This is even more so in underground mines where the whole system is a geologically enclosed one. Geology plays a major role in at least four phases of a mineral venture. Firstly, it acts as an indicator of the existence of parti¬cular minerals. For example, it is inconceivable that a sedimentary coal deposit would exist in a granitic batholith or stock, or that a sandstone would host primary copper or nickel sulphides. Further, the positional relationship of different rock types may be significant in locating deposits such as replacement zinc and copper sulphides in limestones which have been intruded by igneous rocks. Thus, geology is important in the exploration phase of a mineral deposit. Secondly, once potentially valuable minerals have been discovered, knowledge of geology plays a significant role in determining the dis¬tribution, grade and extent of the mineralization. Controls on mineralization include one or all of those related to structure, rock type, sedimentary features, post-depositional redistribution, etc. Con¬sequently, in order to effectively evaluate a deposit in terms of

Jan 1, 1974

By S. Viswanathan, Q. Han, A. S. Sabau

"In spite of recent advances in the mathematical modeling of fluid dynamics for materials processing applications, no significant advances have been made in the numerical discretization of these equations. In this work, the application of two-step projection methods for the numerical simulation of interdendritic flows is discussed. Unlike previous methods, the methods presented here are constructed for the exact equations which are characterized by variable density and volumetric fraction of the liquid. The drag terms, which describe the momentum loss due to the flow around and through the dendrite structures, are treated implicitly. Numerical examples for shrinkage-induced flow during solidification of an Al-4.5% Cu alloy bar is used to illustrate the effectiveness of the proposed algorithm.IntroductionLiquid contraction during solidification plays an important role in the formation of casting defects such as shrinkage voids, microporosity, and macrosegregation. Owing to their importance, the study of transport processes during solidification have received considerable attention during the past years [1, 2]. By the use of volume-averaging techniques, the mass and momentum equations for flow through a stationary dendritic structure were derived [3]. By using the volume-averaging techniques for multi-phase systems, two-phase transport equations for solidification problems were derived [4]."

Jan 1, 1999

By A. Coulombe

"RésuméL’analyse des séquences d’exploitation, actuellement en usage pour les gisements très inclinés, a démontré qu’elles ne sont pas toujours optimales du point de vue du contrôle des terrains et de la concentration des travaux d’abattage (nombre de chantiers simultanément en opération). Effectivement plusieurs séquences sont très propices à la création des nombreuses zones de hautes contraintes, qui sont à l’origine des effets de pilier. L’étude de cas de la mine Lac Shortt a démontré que certaines séquences d’exploitation, actuellement en usage, ne sont surtout pas adéquates dans le cas d’un état élevé des contraintes naturelles. En effet la mine a eu de sérieuses problèmes avec le contrôle des terrains (coups de terrains et les effondrements des épontes). Ces effondrements étaient responsables de la dilution excessive et des pertes importantes des réserves minières. Les deux nouvelles séquences proposées, visent la meilleure redistribution des contraintes d’exploitation en vue d’éliminer la création des effets de pilier. En plus elles permettent une concentration plus élevée des travaux d’abattage.À notre avis les séquences proposées pourraient améliorer substantiellement la stabilité des chantiers d’abattage et par la suite les performances technique et économique des opérations souterraines. ABSTRACTThe planning of underground mining sequences represents an important aspect of the mining operation. The analysis of mining sequences actually in use for very steep orebodies has demonstrated that they are not always optimal, considering ground control and the number of stopes simultaneously in extraction. These sequences are conducive to the creation in the rockmass of several stressed zones caused by pillar effects. The case study of Lac Shortt mine showed that the sequences of exploitation currently in use are not adequate especially in the case of a high state of the natural constraints. Indeed, the mine had serious problems with ground control (rock bursts and collapse of walls). These collapses were responsible for excessive dilution and significant losses of mining reserves. The two new sequences suggested aim for the best redistribution of exploitation constraints in order to eliminate creation of pillar effects. Moreover, they allow a higher number of stopes to be extracted simultaneously. In the opinion of the authors, the sequences suggested could substantially improve the stability of working areas and, thereafter, technical and economical performances of the underground operations."

Jan 1, 1998

By L. P. Purtell

The performance of ten anemometers of several types has been measured at low air speeds. Multiple calibration runs provided sufficient data for measuring random error about the mean calibration curve in addition to the difference between indicated and true speeds. Measurements were also performed of the minimum operating speeds of the anemometers.

Jan 1, 1979

By Thomas L. Vandergrift

Multiple-scam mining is becoming more common in the United States, especially in the East, where primary reserves arc being depleted, forcing mining companies to develop secondary seams, Although mine engineers and planners are aware of the basic effects of seam interaction and generally follow guidelines for columnizing entries and avoiding load abutments, mine design earn greatly benefit bum more quantitative estimates of scam interaction effects. This paper presents case studies from two eastern mines subject to seam interaction effects where pre-mining stresses were quantified: the Mingo Logan Coal Company's Mountaineer Alma A Mine, and Peabody Group's Lightfoot No. 2 Mine. In each of these mines. mining is being carried out within 60 to 90 ft of overlying longwall and pillar panels under maximum cover of about 1,100 ft As part of overall mine planning and design efforts al these mines. NSA Engineering performed numerical modeling analyses using the non-linear boundary-element code MULSIM/NL Input parameters were calibrated by modeling areas where ground conditions resulting from seam interaction were well documented. By incorporating actual cover loading based on topography and seam dip, and the integrity of remnant, gate road, arid harrier pillars in the overlying mines, the pre-mining vertical stress in the underlying scants was estimated based on the pre-mining stress, plans for mains, gate road, bleeder, and pillar panels were developed. Although mining experience in these areas has beers somewhat limited to date, she designs developed as a result of these analyses have so far proven successful.

Jan 1, 2000

By W. M. McKewan

The U.S. Bureau of Mines has established a wire rope research laboratory to examine the factors that affect the life of wire rope. A 2-in-diameter 6x25 fiber core rope was degraded on a bending fatigue machine. This was the first in a series of tests on rope of this construction and size. Baseline testing at the laboratory was reported in an earlier publication. Tensile and nondestructive tests were performed on samples of the rope to determine the relationship between rope deterioration and rope breaking strength. The tests indicated that as a wire rope nears the end of its useful life, deterioration and the consequent loss of rope strength increase at an accelerated rate.

Jan 1, 2010

By Robert F. Chaiken

An absorption-desorption model of particle reactivity was employed to describe the observed relationship between two apparently diverse phenomena which have been the subject of considerable study at the U.S. Bureau of Mines: (1) coal-rock dust reflectance and (2) coal-rock dust explosibility. The model accounts for heterogeneities in particle-size and material properties; however, in the present report, only particle-size distribution is considered. A mean particle-size parameter, appropriate to the explosion extinction process, is derived, and then taken as appropriate to describing optical reflectance. The derived form-to-function relationships for the two processes, when compared with experimental data, suggest that a problem exists with previously derived theory relating optical reflectance to rock dust concentration. A reexamination of that theory leads to a new theoretical expression, changed by a factor of 2, for the variation of rock dust concentration with reflectance which is in quite good agreement with experimental results. The current work offers justification for measuring and establishing rock dust requirements for specific coal dust size distribution(s) found in the mine-distributions that undoubtedly depend on the mining method and the type of coalbed.

Jan 1, 2010

Mineral Resource Throughput Management is based on the extraction of the maximum potential value from an ore reserve with the emphasis to generate maximum ROI and ROR (return on reserve). A steel producing company in India implemented MRTM to significantly improve their ROI and ROR. The value chain of this company consists of their own iron ore and coal mines, ore washing plants, sinter plants, steel plants, and sales and marketing division. The target for the mines was to produce at a stable and predictable rate and at specific quality requirements. These requirements were set to improve productivity and ROI of the steel plant. The implementation constituted the alignment of the reserve potential, total production chain, human resources and management culture at the mines with the requirements of the steel plant. Many typical mining practices were challenged and changed. Production targets regarding lump vs. fines ratio, consistent product tons, quality targets such as Fe, Al content and moist, and standard deviation on the quality targets were not achieved. A reverse distribution model with release mechanisms to satisfy the demand were developed and implemented at the mines. A stable and predictable flow of mine product with quality standards well within the required standard deviation targets were achieved. The contribution from the mines to significantly improve the Net Profit, ROI and ROR of the company were reached.

Jan 1, 2007

By M. Matabane, T. Khati

"Accurate delineation of the contact between a kimberlite pipe and country rock at production level depths is a challenge due to limited geological data. Geological information is obtained from widely spaced diamond core boreholes which are drilled either from surface or from higher mining levels within the pipe. Kimberlite pipe/country rock contacts are notoriously irregular and variable, further reducing the confidence in contact positions defined by the drill-holes. At Finsch Diamond Mine (FDM), the opportunity arose to further improve the confidence in the contact positions relative to the planned slot (end) positions of each sublevel cave tunnel during the development stage of these tunnels. As a result, the accuracy of the 3D geological model has improved.The use of diamond drill core for this purpose is expensive due to site establishment requirements. The lengthy time taken during site establishment also delays the development of tunnels and support cycles, thereby extending the completion dates. FDM has reduced delays during development by adopting percussion drilling, in conjunction with gamma ray logging. The S36 drill rig is mounted on a moveable platform and does not require a costly and lengthy site establishment. The holes are generally drilled (0°/flat) on grade elevation, and these holes could also be drilled from the rim tunnels (developed in waste) into the kimberlite pipe. A single-boom production drill rig is normally used to drill holes about 20 m in length.On completion of the contact delineation drilling, gamma logging of the holes is conducted using the GeoVista geophysical sonde (or probe) to log the natural gamma signature of the dolomite/ kimberlite contact. The advantage of this tool is that the readings are continuous within centimetre intervals, and due to contrasting characteristics between kimberlite (rich in clay minerals) and dolomite, the contact position can be determined accurately. The better definition of contact positions also adds value to tunnel stopping distance in terms of developing the tunnel’s slot at the optimum distance from the contact (easier blasting of longhole rings, avoidance of contact overbreak and premature waste ingress, and other matters relating to extraction of ore from these tunnels). This method is highly successful and has reduced development costs (on-time completion), improved definition of the pipe’s contact position for geological modelling, improved blast design, and mitigated early waste ingress by maintaining the contact’s integrity."

Feb 1, 2019

By S. D. D. Welsby

"A pilot-scale flotation column was used to evaluate four polyglycol frothers at Teck’s Highland Valley Copper Mine. The column was fed with flotation plant feed from the grinding circuit cyclone overflow. Frother dosage was changed by controlled addition to the column feed. Pulp, froth and metallurgical performance were assessed using gas dispersion measurements, water recovery and copper, molybdenum and gangue grades and recoveries. Ranking the frothers based on pulp and froth phase performance using gas dispersion measurements and water recovery yielded contradictory rankings. All of the frothers changed the froth structure in a similar manner such that, at the same structural conditions, they all gave the same degree of separation between water and entrained solids. The only difference between the frothers was what dosage was required to achieve a certain structure. Changes in pulp structure appeared to be related to the critical coalescence concentration (CCC) for each frother. Changes in froth structure were not completely described by the CCC and involved other factors, such as the hydrophobicity of particles in the froth. All of the frothers were capable of achieving the same metallurgical result, just at different dosages. Optimal dosages for the frothers were anywhere from the CCC, where copper and molybdenum grades were maximised, to 2 to 3 or 4 to 7 times the CCC, where copper and molybdenum recoveries, respectively, were maximised.INTRODUCTION Background Due to supply issues, Teck’s Highland Valley Copper (HVC) Mine replaced Dowfroth 250C, their auxiliary frother, with Huntsman Polyfroth W31. The primary criterion for selection was that the new frother gave the same performance as Dowfroth 250C, minimising disruption to the circuit. Frothers play a dual role in flotation, altering the bubble size in the pulp zone and the stability of the froth. A stronger or weaker frother, or a blend of both, will alter the balance between pulp and froth effects and, ultimately, metallurgy. Dowfroth 250C and Polyfroth W31 may not necessarily be the best frothers for HVC from a metallurgical standpoint. There may be a different frother or frother blend that could deliver higher recovery in the bulk roughers."

Jan 1, 2014

By S. P. Chong, Peter V. Avotins

The iron and steel industry continues to operate at depressed production levels due to the low level of demand for steel and foundry products. Crude steel production in the Western world dropped by 13.7% from 459.3 Gt (506 billion st) in 1981 to 396.4 Gt (437 billion st). The reduction was particularly severe in North America where the industry suffered a loss of 37.3% from 123.6 to 77.5 Gt (136 to 85 billion st). This reduced level of production severely affected iron ore and agglomerates output. Substantial curtailment was made in both sintering and pelletizing plants to meet reduced demand. The pressure on these plant operators shifted from quantity to quality. With energy cost becoming an ever increasing part of production costs, conservation was a major area of emphasis. Iron Making Sinter has continued to be the favored burden for blast furnaces overseas. As hot metal production declined and excess sintering capacity developed, many of the overseas operators that have been using pellets found it advantageous on hot metal production cost to increase the ratio of sinter at the expense of pellets. This was the case despite the depressed price level of pellets in the world market. In North America where pellets have been the predominant burden material-mainly because it is necessary to pelletize the fine grain iron ores available domestically-there was also increased interest in sinter usage. Bethlehem Steel Corporation's 13.6-m (44.5-ft) furnace at Sparrows Point experimented with a burden with more than 70% of sinter during the second half of the year and indicated some favorable results as compared to its previous 50% super flux sinter and 50% acid pellet burden. The blast furnace operators continued to press for improved burden quality. In addition to the emphasis on the importance of burden chemistry and quality, uniformity, much emphasis was placed on improving reducibilitiy and high temperature metallurgical qualities. These burden properties were recognized to hold further promise in reducing fuels rates. Sintering With reduced consumption caused by the cutback in hot metal production, many sinter plants sat idle while others operated at greatly reduced rates. The deemphasis on production in many plants allowed them to experiment with various means to reduce operating costs. The most notable action was the introduction of low cost South American pellet feed fines into the sinter ore mix. These feed fines, which are finer than 150 µm (100 mesh) ore, were previously thought to be too fine for sintering. However, when the full production rate is not required the use of up to 40% in the ore mix was reported to be successful with the aid of burnt lime and adequate premixing and granulation of the sinter mix before sintering. The emphasis on quality and tighter cost control also has resulted in a further breakthrough in sintering technology. It was recognized that good feed preparation-including selection of compatible ores, good blending, good pseudo-particle formation, additive size control increased bed depth, and closed control of the sintering bed heat pattern-could allow the production of high quality sinters with much lower heat input and FeO level. Bed depth of more than 600 mm (23.6 in.) was practiced in a number of operations. Total heat input of better than 1.3 GJ/t (290,300 kcal per st) has been achieved. Kawasaki Steel's Chiba Works has also achieved a 5.3% SiO2 in the sinter with good quality. The low heat input approach allowed production

Jan 5, 1983

By A. G. Ulsoy

In recent years there has been an Increasing Interest In the application of automatic control theory to mineral processing systems, and numerous applications of mineral process control have appeared in the literature. A general approach to the synthesis of mineral process control systems is presented. The first phase of the control system design consists of specifying the control objectives, process variables and their interactions; these topics will be discussed with specific reference to mineral processing. The second phase Is con¬cerned with selection of control strategies algorithms, instrumenta¬tion, and other components of the control system and is discussed at length. These concepts are then illustrated by a flotation control case study.

Jan 1, 1981

By Bruno Pinheiro Silva, João José dos Santos Carvalho, Flávio Túlio Busatto, Leonardo Passos Perdigão

Sinter has high participation in the blast furnaces ferrous burden in ArcelorMittal Tubarão, being therefore of paramount importance the constant control of its physical, chemical and metallurgical qualities, aiming at the fulfillment of the client required quality. With the great challenges faced by market conditions, steelmaking companies seek operational sustainability, aiming at stability and reducing costs. One of the main steps to obtain these results in the Blast Furnaces is the raw materials quality stabilization, in order to guarantee the perfect burden distribution control and a better use of the reductant gas inside the reactor, aiming a lower fuels consumption. It is known that the smaller a ferrous particle is, better its reducibility, since a larger surface area provides greater contact with the reductant agent. Based on this concept, the objective of this work is to quantify the effect of the sinter size produced in ArcelorMittal Tubarão Sinter Plant in its reducibility, and in this way enable the definition of a better sinter particle size distribution aiming at a balance between the permeability effects and gaseous yield in the Blast Furnaces. In addition, the reducibility effects on sinter size reduction were evaluated, where the 5 to 10 mm fraction was analyzed in two sinter samples, the first obtained from the conventional route and the second after crushing of the fraction over 50 mm, in order to simulate the effect of the use of a Secondary Crusher in Sintering.

Oct 1, 2019

By S. M. Shelton, F. D. DeVaney, M. M. Fine

Ferrograde manganese ore or its equivalent is essential to the manufacture of every ton of steel produced ; although it is well -known that there are large deposits of low - grade ore in the United States , the production of substantial amounts of ferrograde material from domestic sources presents a problem that becomes more and more critical as the stockpiled ore is depleted through inability to obtain adequate supplies from foreign sources . Consequently, a program of laboratory and pilot - plant investigations designed to solve the problem of obtaining a domestic supply of manganese has been undertaken by the Metallurgical Division, Bureau of Mines . This paper is one of a new series that will include reports of results obtained in the various investigations conducted under the manganese program . The scope of the program and the organization that was set up to direct and supervise the work were discussed in the foreword of the first paper of this series , Report of Investigations 3606 , Manganese Investigations 1. Concentration of Manganese Ores from the Drum Mountain District, Utah, by S. R. Zimmerley, J. D. Vincent , and C. H. Schack . Other reports will be published as the work progresses .

Feb 1, 1942

By R. Minnitt, C. Morgan, D. Fourie

The problems associated with outliers and their adverse effects on resource estimation in deep level South African gold mines are presented, and solutions suggested in available literature are reviewed. Traditional Witwatersrand type outlier capping uses the 99th percentile grade as the upper limit, a somewhat blunt approach. The occurrence of high and low grade areas in mining operations and the estimation of resources using geostatistical techniques require that the assumption of stationarity be upheld. Honouring the stationarity of data requires that mining areas be subdivided into appropriately sized and spaced domains. In order to overcome high-grade conditionally biased block estimates, the overly strong influence of extreme grades is mitigated by using naturally occurring grade sub-domains, by identifying outliers within the grade sub-domains, allocating relevant capping limits, and weighting them in a post-estimation process to. Rather than applying a capping value across a domain it applies a maximum cap based on the extreme value and its range of influence; the method also ensures that relevant capping occurs even in lower grade sub-domains. The proposed method for capping outliers recognises that metal contained in high grade chip samples should not be over-capped, and based on the grade of surrounding samples. The post-processing method of capping outliers does not interfere with the kriging equations. The method is simple, elegant, and can be easily applied. An advantage is that the output from the kriging process is not overwritten, allowing for comparison between original kriging results and post-processing results.

Sep 1, 2021