Search Documents

By Debasis Deb, Ankur Narang, G. P. Raja Sekhar, Kamal C. Das

"The stability of underground structures made (especially) in jointed rock mass is of the utmost important to designers, engineers and operators. Rock bolting is generally being practised to reinforce excavation walls and roofs by minimizing the movement of rock joints. This study proposes a new analytical model for the prediction of displacements, axial load and shear stress of a fully grouted rock bolt intersected by single or multiple rock joint(s). The governing equations for rock bolt are derived based on the assumption that shear force generated by slippage along the bolt-grout interface has a linear relationship with respect to the relative displacement at the interface. The proposed model is novel compared to the existing literature in the sense that it accounts for rock kinematics as an enriched sum of a continuous part (elastic) and a discontunous (jump) part due to the presence of joint planes. The complete solution of the bolt displacement, axial load and shear stress is derived as a function of the rock displacements, and, it comprises of Heaviside and/or Dirac delta functions in the solution space due to the jump at the intersection point(s) of bolt(s) and joint(s).Our novel solution is able to predict the elastic behavior of a grouted bolt, subjected to opening of the joint planes in pull-out test and also in excavated tunnel. These two cases also serve as benchmarks for grouted rock bolts with multiple joints and can be used for validation of numerical and experimental models. Further, the solution is applied to analyze the behaviour of grouted rock bolt as well as cable bolt installed in circular tunnel, with variations in rock joint(s) opening displacement, shear stiffness in the boltgrout interface and locations of the joint planes. Results indicate that for a specified grout stiffness, the maximum axial load developed in the bolt for discontinuous distribution of rock mass displacement is significantly higher than that of the continuous distribution of rock displacement. This paper provides an analytical interpretation of bolt beaviour considering multiple joint planes with variable joint opening displacements. The results obtained via the analytical method are in line with the corresponding investigation done by other authors. Thus, our analytical treatment can be used to understand bolt behaviour in realistic situations."

Jan 1, 2015

By D. H. C. Wilton, W. A. Sears

"Abstract-The 682.8 ± 1.6 Ma Tickle Point Group consists mainly of rhyolite flows and tuffs with- in the Avalon Zone of the Appalachian Orogen in Newfoundland. Stratabound massive sulfide mineralization, consisting of pyrite, sphalerite, chalcopyrite and galena, occurs at the Winter Hill prospect, where a lower pyritic, Cu-rich zone and an upper Zn-Pb-rich zone are hosted by a carbonate-calc-silicate lens. At the Frenchman Head prospect, stringers of pyrite, minor sphalerite, chalcopyrite and rare galena are hosted by felsic volcanic rocks of the Tickle Point Group .Base metal ratios, metal zonations and the overall stratigraphic framework suggest that these sulfide occurrences represent volcanogenic massive to semi-massive sulfide mineralization (stringer type at Frenchman Head). As such these prospects represent the only known VMS occurrences within the Avalon Zone of Newfoundland. Pb isotope data indicate that galena from the Winter Hill prospect is less radiogenic and older than galena from Dunnage Zone occurrences, and suggest that the Winter Hill mineralization formed in an environment tectonically distinct (exotic) from the Dunnage Zone."

Jan 1, 1996

By C. A. Biley

"A dedicated iron removal stage was installed at Anglo American Platinum’s Rustenburg Base Metals Refiners (RBMR) during an expansion project which began execution in 2010. The Pressure Iron Removal (PIR) process aims to precipitate iron from a nickel leach solution under medium temperature (145 °C) oxidative conditions using caustic for neutralisation. Since commissioning, the process has been robust in removing iron from the refinery but has shown significant variability in the residues produced. These residues, which are typically comprised of ferrihydrite, hematite or natrojarosite, have varying iron contents and different physical properties which have implications for their filtration and ultimate removal from the refinery. Since over 95% of nickel plated at RBMR passes through the PIR section, volumetric constraints caused by poor filtration performance directly influence nickel production and stabilisation of this process remains central to the continued ramp-up of the refinery to full nameplate capacity. This paper details the development of the PIR process and the journey from concept to implementation. Recent laboratory and plant scale tests have highlighted the importance of the initial phase formed in the autoclave for the ultimate residue requiring filtration and how the feed acid concentration is a critical control parameter. On this basis, a new caustic dosing philosophy was implemented which improved pH control in the autoclave. Further optimisation of the PIR process is on-going.INTRODUCTIONIn 2010, an expansion of Rustenburg Base Metal Refiners (RBMR) to increase nickel production from 21 to 33 ktpa began execution. The management of iron was pivotal in the ultimate configuration in the new circuit and one of the primary differences from the original circuit was the addition of a dedicated iron removal stage: Pressure Iron Removal (PIR). A significant drive for this new unit operation was to facilitate the separation of iron from other leach residues. Not only did this result in a significant decrease in the amount of leach residue but, without the constraints associated with iron precipitation during leaching, has permitted more aggressive leaching conditions and, correspondingly, greater throughput and efficiencies.However, as with many commercial implementations of even mature technologies, the journey from development to implementation has presented challenges for the continued ramp-up of the refinery to full nameplate capacity. Specifically, since all nickel bearing leach solution must pass through the PIR section prior to the nickel purification and recovery circuit, throughput difficulties in the PIR section can directly impact supply of feed to the nickel tank house. Since commissioning, the PIR autoclave has effectively removed iron from the refinery, but has shown significant variation in the solids phases produced in the autoclave. This variation presents difficulties for the filtration of the discharge slurry and represents a key throughput limitation which periodically impacts production. Additionally, the increasing production rates associated with the ramp-up of the refinery have resulted in a significant increase in the required volumetric throughput through the PIR process. This has placed even greater emphasis on improving the stability of the solid phase produced in the autoclave, ultimately targeting an easily filterable residue with high iron content."

Jan 1, 2016

By K. Soldenhoff, D. Wilkins, E. Ho

In the processing of rare earth ores, a pure, mixed rare earth chloride solution is usually produced and used as the feed to solvent extraction circuits to separate the individual rare earths. Cerium(III) may be separated from the other rare earths by oxidising it to cerium(IV), which precipitates as cerium(IV) hydroxide in chloride liquors. This oxidation step may be inserted either before rare earth separation by solvent extraction or between solvent extraction circuits. Although oxidation of cerium by sodium hypochlorite has been known for many years, detailed conditions for oxidation are not readily available. The effect of hypochlorite addition, temperature, pH and cerium concentration on oxidation of cerium(III) from synthetic solutions was investigated

Jan 1, 2014

By J. A. Maritz

"The Hedley and Grant formulation for pillar strength has been widely adopted by the South Africa hard rock mining industry. Today, some forty year later, this formula still provides the foundation for pillar strength calculations when designing for stability and utilising pillars in mine layouts. Various mine specific back analyses were conducted to adjust the equation parameters based on underground observations of failed pillars and comparing to the pillar stresses calculated using numerical models. In all these instances, the influence of shear stresses was ignored. The key to more representative pillar strength estimation is a better understanding of the pillar loading environment.In this paper, special mining cases are evaluated and the existence of shear stresses on the pillar loading environment is illustrated. By revisiting some back analysis data for optimisation of the equation parameters, recommendations were made in respect of the influence of the shear stress component and on how this should be considered in determining the pillar strength.INTRODUCTIONCommon practice on flat dipping, shallow and intermediate depth ore deposits in the South African hard rock and coal mining industry is to exploit the mineral resource using a room-and-pillar layout. The design methodology is mainly based on a Factor of Safety (FoS) approach, where the strength of the pillar is compared to the effective load on the pillar. In these environments, FoS values of 1.6 are applied as it was found to correspond to a probability of failure in coal mines of 0.001. As discussed by Ryder and Jager (2002) page 279: “For lack of anything else more substantial, the figure of 1.6 is generally used in hard rock pillar design in South Africa”. Dependant on the designed purpose of the pillar, the FoS might be selected as high as 2.0 where long term stability is required. The factors directly influencing the FoS include the pillar height (which is linked to the mining height), effective pillar width, a rock-strength factor and extraction ratio."

Jan 1, 2015

By H. K. Olsen, B. Thomassen, H. K. Schønwandt, L. Grahl-Madsen, H. Stendal

Abstract-The central part of the early Proterozoic Ketilidian orogen in South Greenland consists of an I-type granitoid batholith transected by four regional NE-SW-trending shear zones. The batholith was probably emplaced during subduction, from the south toward the Archaean craton in the north, in a sinistral transpressional system. A regional sediment sampling program covering 5000 km2 shows a close connection between a gold-mineralized trend and one of the major shear zones within the epizonal facies of the batholith. The gold mineralization is not directly related to the major shear zones, but occurs in quartz-veined and silicified second-order structures. High-grade (up to 114 ppm Au over 6 m) low-tonnage mineralization is found within these second-order structures cutting both amphibolites and granitoids. The mineralization is characterized by negligible or small alteration halos. Gold is associated with various element combinations of enhanced As-Bi-W- Cu, which indicate that the mineralization is mesothermal in origin.

Jan 1, 1995

By H. K. Olsen, B. Thomassen, H. K. Schønwandt, L. Grahl-Madsen, H. Stendal

The central part of the early Proterozoic Ketilidian orogen in South Greenland consists of an I-type granitoid batholith transected by four regional NE-SW-trending shear zones. Thebatholith was probably emplaced during subduction, from the south toward the Archaean craton in the north, in a sinistral transpressional system. A regional sediment sampling program covering5000km2shows a close connection between a gold-mineralized trend and one of the major shearzones within the epizonal facies of the batholith. The gold mineralization is not directly related to the major shear zones, but occurs in quartz-veined and silicified second-order structures. High-grade(up to 114 ppm Au over 6 m) low-tonnage mineralization is found within these second-order structures cutting both amphibolites and granitoids. The mineralization is characterized by negligible or small alteration halos. Gold is associated with various element combinations of enhanced As-Bi-W-Cu, which indicate that the mineralization is mesothermal in origin.

Jan 1, 1995

By J. Y. Liu

In Comparison to the application of HPGR (high-pressure grinding rolls) in the cement industry and in iron ore processing, the introduction of this energy-efficient comminution technology to processing of nonferrous base-metal ores has progressed slowly as yet. In the field of minerals industry, challenges lies not only in the large variety of the feed material with respect to particle size, density, hardness, moisture and mineralogy of the material but also in complexity of the processing flowsheet, which need to be addressed from case to case based on result of HPGR comminution test work. Whereas conventional comminution test with HPGR in lab-scale or pilot scale needs quite a large amount of feed material, a compressive stressing test of particle-bed in piston-die device provides an effective way to study interparticle breakage behavior of ore material with much less feed. In the present study the effects of interparticle breakage on comminution of a copper ore has been investigated by compressive stressing of particle bed in a piston-die device. The feeds to the particle-bed stressing tests were four narrow size fractions in the size range of 0.3~3.35 mm and a broad size fraction of 0/3.35 mm. The pressures applied were in the range between 30 and 180 MPa. The results showed that the specific energy consumption increases linearly with the pressure; the product fineness increases also with the pressure, but the magnitude of this increase declines as the pressure goes up. Under the same stressing intensity, the product size distribution curve becomes steeper as the particle size decreases, so there is no self-similarity between product size distributions of different feed sizes. The suitability of a material-characterization model for interparticle breakage was verified for this ore. In addition, the effect of interparticle breakage on comminution of this ore with respect to its influence on the downstream flotation were analyzed in terms of size-metal distribution and mineral liberation.

Jan 1, 2016

By B. MohammadSadeghi, S. Bakhshivash, M. Haghshenas, F. Rahimi

"Spring-Back/Forward in sheet metals is defined as recovery of elastic strain upon unloading in a deformed sheet specimen. Since this phenomenon can cause alternation in the dimensions of a part, leading to irrecoverable compensations, the operative parameters need to be investigated, thoroughly. To assess spring-forward and effective bending parameters, V-shaped bending dies are the most commonly used experimental approach. In the present research, the effect of bending angle and punch tip radius on the spring-forward behavior of an Al-Mg-Si alloy sheet specimen is studied. To this end, a V-shaped die with four bending angles (i.e. 30°, 45°, 60°, and 90°) was designed and fabricated. Specimens with thickness of 2 mm were cut with dimensions of 50 mm×10 mm. Results show that by increasing the bending angle, the spring-forward angle decreases. In addition, smaller punch tip radiuses lead to further spring-forward. The effect of these parameters on the amount of spring-forward angle, strain distribution and homogeneity along sheet thickness was assessed using finite element analysis (FEA). Simulation results show that as the bending angle increases, the strain is developed more homogenously. Finally, micro-hardness test was used to study the effect of strain hardening caused by bending operation on strength evolution. Results show higher hardness trends for the zones in which the maximum plastic strain occurs.INTRODUCTIONPrecise production plays a significant role in the durability of a part. When it comes to sheet metal forming, this issue should be taken into account as an important factor. In order to predict the alternations, bending dies with special geometric parameters (i.e. V-shaped, U-shaped and L-shaped dies) are commonly used (Wagoner, Lim, & Lee, 2013)."

Jan 1, 2016

By C. M. Kleemann

The paper describes the present milling practice at Western Mining Corporation Limited's new nickel mine at Kambalda in Western Australia. It details the crushing, autogenous grinding, flotation and spray drying circuits for the production of a nickel concentrate. It also indicates some of the metallurgical problems which have been encountered in concentrating this ore.

Jan 1, 1973

By Guichao Wang, Geoffrey M. Evans, Graeme J. Jameson

"In this paper we present a new method for studying the detachment of particles from bubbles in a rotating turbulent eddy. The eddy is formed in a wall cavity in a two-dimensional water tunnel with transparent walls. When water flows through the tunnel, a vortical flow field develops in the cavity. The properties of the eddy can be modulated by changing the free-stream velocity of the water in the tunnel. Bubbles are pre-loaded with one or more particles in a fluidized bed flotation device located beneath the vortex cavity. Loaded bubbles are released one at a time into the cavity, and the motion of the bubbleparticle aggregate is studied using a high-speed video camera. The diameters of the particles and the bubbles, and the number of particles initially attached to the bubble, can be varied. The trajectories taken by the bubbles are quite complicated. In some cases, the bubble moves to the centre of the eddy, and particles rotate around its axis. If the rotational speed is sufficient, particles may detach due to centrifugal force. However, other modes were observed, including inertial detachment due to rapid changes in direction of the surface of the bubble, because of changes in trajectory of the bubble as a whole, or because of pulsations and oscillations of the bubble surface. Clusters of bubbles held together by particles were seen to form and reform. In the traditional explanation for the detachment of particles in flotation cells, it is assumed that particles detach from bubbles rotating in an eddy due to centrifugal force (Schulze, 1977). Although the conditions assumed in Schulze's theory may exist, it is only one of a range of phenomena that can lead to the detachment of particles from bubbles in a turbulent vortex.INTRODUCTIONThe recovery of particles by flotation methods is dependent on the particle size. The recovery in the existing flotation machines reaches a peak for particle size, typically in the range of 20 to 100 µm. This size range varies as a function of the particle’s properties (type and composites). For coarser particles, the recovery drops very quickly (Gaudin, Groh & Henderson, 1931; Jowett, 1980). If the upper size limit can be increased for flotation, dramatic savings in the grinding energy can be achieved."

Jan 1, 2016

By T. Krumins, F. C. Ruff, M. Eichhorn, L. Zunti

Hatch Ltd. has recently completed a project to increase refractory gold ore processing capacity at Newmont Mining Corporation’s Twin Creeks facility. The objective of the project was to increase throughput by approximately 10% by adding a third autoclave feed pump to each of the two autoclave circuits. To minimize project capital cost, autoclave feed pumps from Newmont’s decommissioned Lone Tree facility were evaluated and found to be suitable for this new service. The existing pressure oxidation circuit was reviewed in a debottlenecking study to determine other upgrades required to facilitate the increased throughput. Interruption to operations was minimized with a detailed construction plan, supplemented with detailed 3D layout model sequencing studies. The focus of this paper is project development, multi-discipline design considerations, and implementation results.

Jan 1, 2014

By K. Krishnapisharody, S. Clark, M. Barati

Different gas purging methods to metallurgical baths have been applied to date, among which are top lancing and bottom blowing through tuyeres or porous plugs. Among these, the porous plug bottom blowing technique has shown higher efficiency in gas utilization than tuyeres and top blowing, due to smaller bubble size and lower liquid velocity. However, the studies related to bottom stirring of metallurgical baths are limited to slag-less or thin slag-covered melts. In particular, there is no examination of stirring ability in conditions similar to Falconbridge's Slag Make Converter, where the bath is covered with a thick slag layer and a stream of solids is charged on the bath from top. This study was undertaken to investigate the application of gas purging through a porous plug in stirring in and melting of the solid feeds. The spreading behaviour and melting time of a solid layer forming on a plume created by stirring were calculated using the existing models. The results show that formation of a thick solid layer, specifically in areas beyond the plume region is very likely. Nevertheless, the thickness of the layer and melting time do not appear unreasonably high. It is predicted that by proper controlling of gas flowrate and plug configurations, successful solid charging through the mouth can be practiced.

Jan 1, 2007

By J. Q. Chen

The elastic-plastic behavior and the fracture mechanisms of two 2198 Al-Cu-Li thin sheets are investigated. The first material (2 mm thick) is in a recrystallized state after cold rolling. The second material (6 mm thick) is hot rolled and has a fibrous grain structure. The two materials are studied in both T351 and T851 conditions. Tensile tests are carried out along different sheet directions using flat and notched specimens. Fracture resistance was characterized using Kahn tear specimens and M(T) panels. Macroscopic fracture surface of smooth tensile specimens depends not only on testing material but also on the loading direction. For Kahn tear specimens, fracture is initiated in a small flat triangle whereas slant fracture is observed during crack propagation. Fracture mechanisms were investigated using scanning electron microscopy (SEM) and X-ray 3D tomography.

Jan 1, 2011

"A method is described for determining the viability of hard rock orebody extraction by block caving. Using a very limited amount of geotechnical data, the procedure enhances Laubscher’s empirical stability graph method by using the complex structural modelling system (ChaSM) to assess cavability and block cave tonnages.IntroductionIn considering the selection of suitable mining methods, block caving has gained popularity over other extraction systems in recent years (Butcher, 1999). This has been ascribed to the fact that block caving is the lowest cost per ton mining method, with orebodies that were in the past considered unsuitable for caving, now being successfully mined. However, even though all orebodies will cave, a problem exists in the evaluation of economic viability — it is necessary to be able to predict both when the orebody will cave and what will be the sizes of the initial ore fragments (and the associated tonnages) that report to the drawpoints. This problem is compounded by the fact that, during the scoping study phase of a project, only limited geotechnical data may exist on which to base an assessment of whether block caving is viable or not. Empirical and numerical methods of cavability assessment do not account for the full effects of geological discontinuity trace lengths, nor for the stress regime. Those require significant geotechnical data to be available. The result is that uncertainty may exist as to whether block caving is an economic option. This uncertainty is complicated by the fact that limited experience exists with the caving of more competent orebodies.This paper describes a method for assessing cavability, and for predicting the size of initial ore fragments and the tonnages reporting to block cave drawpoints, using only the level of data normally available during a scoping study. Fragment size and tonnages are assessed by identifying potentially unstable blocks and wedges in the undercut back using statistical joint data. The effects of lateral confining stresses are accounted for by comparing the in situ stress orientations with the joint orientations. To evaluate the accuracy of the method, a back analysis exercise was conducted to predict cavability and initial tonnage for a mine in which the actual behaviour is known. The method is then applied to cavability and tonnage predictions for a hard rock orebody."

Jan 1, 2002

By F. Khomamizadeh

In this study, deep cryogenic treatment was applied to A319 cast aluminum samples. Although the cryogenic treatment didn't have any significant effect on mechanical properties of the well degassed and fluxed samples, but the tensile strength of the samples containing a considerable amount of gas and shrinkage defects was improved surprisingly, This improvement was attributed to the strengthening of the a-aluminum matrix which slows down the propagation of the existing defects. There was no significant increase in the hardness after cryogenic treatment, The abrasion resistance of the alloy was improved after cryogenic treatment, by developing a homogeneous protector shield, composed of abrasive grains penetrated into the soft aluminum matrix and kept tight.

Jan 1, 2005

By P. Su

Friction stir spot welding of a range of Mg-alloys is investigated (AZ91, AZ31, AM50 and AM60). The peak temperatures during friction stir spot welding of all Mg-alloys are close to their solidus temperatures. Also, the apparent material viscosity of Mg-alloy sheet is lower than that when Al-alloy sheet material is spot welded. This is important since it means that the widths of the stir zones forming during spot welding of Mg-alloy sheet are smaller than those when Al-alloy sheet is welded (when identical welding parameter settings are applied). These lower apparent viscosity values must be taken into account during numerical modeling of the Mg-alloy spot welding process. The efficiency of energy utilization is very low when Mg-alloy sheet is spot welded and is of the same order as that when fabricating Al-alloy sheet. The energy generation contributions of the rotating pin and the tool shoulder are investigated. When the spot welding operation is carried out without the application of a dwell period, the contribution made by the rotating pin vis-à-vis energy generation is around 66%. The contribution made by the tool shoulder increases to about 65% when a dwell period of four seconds is incorporated during the spot welding operation. This increase results from a longer contact time between the tool shoulder and material within the stir zone, The failure load during overlap shear testing increases when the bonded area and the energy input increase during spot welding. Sample failure initiates from unbonded regions and from discontinuities at the edge of the stir zone. The formation tendency of discontinuities is increased when Mg-Alloy sheet is spot welded because of the limited deformation characteristics of the base material at lower temperatures.

Jan 1, 2006

By J. J. Jonas, G. R. Stewart, A. M. Elwazri

Two important parameters for dynamic recrystallization can be derived from the changes in the strain hardening rate: the critical strain for initiation of dynamic recrystallization and the point of maximum softening. In this work, these values are determined from the flow stress-strain data obtained in continuous compression testing in the range 900-1100°C. The resulting strains are used to calculate a kinetic model of dynamic recrystallization for two materials: 304 austenitic stainless steel and a hypereutectoid plain carbon steel. The mechanical parameters used to define the proposed model are confirmed with quantitative metallographic analysis of the recrystallized microstructure.

Jan 1, 2004

By C Jincheng, L. Liang, L. Wenjun

"Considering overgrinding in a grinding process, the grinding characteristics of comminuted hematite products using a HPGR and a conventional CC were comparatively studied. The major properties including particle size, grinding kinetics and grinding technical efficiency (Et) were investigated. Individual size distributions (ISDs) of -2.0mm in a HPGR product showed a more significant “double-peak” characteristic than a CC product. As a result, HPGR products were finer and had a wider full-size distribution because of the greater proportion of fines. The parameters m and zi in m-th order grinding kinetics were analyzed by the partial-derivative method and the critical time tc(tc = exp (1/ zi)) was obtained. zi mainly affected the breakage rate when t < tc; otherwise, m mainly affected the breakage rate. In these tests, tc was considerably higher than the grinding times, and HPGR products yielded a higher zi-value than the CC product for the same sieve size. This indicated that HPGR products had a higher breakage rate and needed a shorter grinding time for a certain grinding fineness. Et was calculated by simulating the size distributions of batch grinding products. HPGR products gave a lower Et at a decreasing rate than the CC products for -2.0 mm feed and -0.5mm feed, and then showed a slightly higher Et for -2.0+0.5 mm feed. This indicated that overginding in HPGR full-size products was more serious under the condition of coarse grinding, whereas the efficiency was higher and overgrinding weakened significantly in HPGR coarse products because of screening-out fines. Based on the above research, a new process of “ultra-crushing in HPGR – classified grinding” was advanced, reducing ores tonnage into first stage ball mill significantly, improving grinding efficient and weakening overgrinding in the whole grinding process.INTRODUCTIONIn mineral processing, comminution accounts for a substantial portion (approximately 60~70%) of the energy consumed in plants. As an ultra-fine crushing equipment,the application of the HPGR for industrial comminution has been paid close attention because of its low specific energy, low steel consumption, high capacity and high ratio of reduction, among others (Schonert, 1998; Norgate & Weller, 1994; Fuerstenau & Kapur, 1995). HPGRs comprise two counter-rotating rolls that draw feed material into the roll gap. One roll is fixed to the mainframe, while the other movable roll is mounted on slides and positioned by hydraulic rams. The specific pressing force (SPF) (i.e., total force divided by projected roller area) is typically in the range of 3~9 N/mm2 depending on the application, while pressures of up to 1000 N/mm2 have been measured within the operating gap (Maxton , Morley, & Bearman , 2003)."

Jan 1, 2016

By Carlos Alberto Spier

The Águas Claras and Pico mines are two world-class, iron-ore mines hosted within Lower- Proterozoic banded iron-formations (locally known as itabirites) of the Minas Supergroup located in the Quadrilátero Ferrífero District, Minas Gerais, Brazil. The Águas Claras orebody consists of a 2,500 m long roughly tabular-shaped lens hosted within the dolomitic itabirite of the Cauê Formation. Dolomitic itabirite is the protore of the soft high-grade iron ore, which is the main ore type of the Águas Claras orebody, and represents more than 85% of the 284 Mt mined since 1973. Hematite is the main constituent of the iron ores. It occurs as martite, granular hematite and, locally, as specularite. Magnetite appears subordinately as relicts within martite and hematite crystals. Gangue minerals are very rare. These consist of dolomite, chlorite, talc and apatite and are especially common close to the contact with the protore. This virtual absence of gangue minerals is reflected in the chemistry of ores which are characterized by very high Fe contents (an average of 68.2% Fe). The Pico orebody is a continuous ~3,000 m long body of lenticular shape hosted within quartz itabirite, which is the protore of the soft high- and low-grade ores at the Pico mine. The soft high-grade ores, together with the low-grade ores, called iron-rich itabirite, are the main types of ores at the Pico mine, respectively representing approximately 51 % and 29 % of the reserves within the pit. The iron oxide mineralogy is the same as that of the Águas Claras mine, but in different proportions. Gangue minerals are very rare in the high-grade ores, but are slightly more common in the iron-rich itabirite. Quartz is the dominant gangue mineral, and is found with minor quantities of chlorite. The chemistry of the high-grade ores is characterized by high Fe contents (an average of 67.0 %) and low P, Al2O3, and SiO2 which are concentrated in the fines. Iron rich itabirites average 58.6% Fe and 13.5% SiO2.

May 1, 2003