Search Documents

By C A. Locke, A B. Christie, J Cassidy, J L. Mauk

New and existing gravity and aeromagnetic data are used to determine the structure of the Waihi region and to investigate any relationship between structure, hydrothermal alteration and epithermal mineralisation. The 24 mGal, 10 km diameter, negative Bouguer gravity anomaly which occurs south of Waihi has been modelled as a I.5 km deep asymmetrical basin. This basin, the Waihi Depression, is interpreted to be infilled with low-density volcanic rocks of the Uretara Formation and Whitianga Group, overlying older andesites of the Waipupu Formation. To the north and west, the margins of the depression are defined by the Waihi and Mangakino Faults, which have been modelled with dips of 45 - 60¦ to the south and west respectively. The southern and eastern boundaries of the Waihi Depression are defined by smooth, shallow dipping (15 - 20¦) boundaries. The Waihi Depression is interpreted as a fault-angle depression, resulting from southeast extension in the early Pliocene. The Waihi Fault is modelled as an approximately E-W feature, located to the south of Waihi. Small 2 - 4 mGal positive Bouguer gravity anomalies occur north and west of the margins of the depression, and correlate with known areas of hydrothermal alteration and areas of low magnetic gradient.

Jan 1, 2003

The prospect of utilisation of phosphate ores which contain high P2O5 values associated with Al and Fe has produced a great deal of interest over many years. Processing options have generally been based on the utilisation of by-products which were required to offset the processing costs.   This paper presents a range of processing options which have been investigated at laboratory or technical scale on phosphate ores which predominantly consist of crandallite, CaAl3(PO4)2(OH)5.3H2O and millisite, Ca(NaK)Al6(PO4)4(OH)9.3H2O. Christmas Island (Indian Ocean) and Mt Weld (Western Australia) phosphate ores are good examples of such feedstocks. It is expected that these processes may also be applicable to such phosphate ores from other locations.

Jan 1, 1990

By T Graham

The invention of the traditional belt and idler conveyor system is attributed to Thomas Robins in 1892. This basic concept, ubiquitous in most mining operations, has not changed markedly in 125 years and still suffers from five forms of energy losses. Firstly, Wheeler, Roberts and Jones (2004) showed that indentation rolling losses, a consequence of the indentation a roller inflicts on the belt, consume 60 per cent of all energy. Secondly, Wheeler (2006) explained that bulk solid flexure resistance resulting from material interaction and movement absorbs 20 per cent. The remaining energy is lost by idler rotating resistance (5 per cent), belt flexure resistance (5 per cent) and secondary resistances (10 per cent).Traditional conveyor systems are continuous as opposed to batch systems, implying both positive and negative features. On one hand, continuous systems have high availability (available infers ‘ready to run’), but, on the other, utilisation of availability is often low; for example, when a conveyor is running empty or below capacity. This converts into reduced system OEE (overall effective efficiency). Many attempts have been made to mitigate energy losses and low OEE issues, including development of rail-based systems. The key drawbacks have been higher capital cost, low flexibility, poor availability and lack of market acceptance.A recent conceptual development known as the ‘autonomous rail conveyor’ (ARC) promises to address issues of energy loss and poor availability and provide increased flexibility by offering solutions for difficult applications. The ARC uses a series of interconnected skeletal bogies covered with a conventional conveyor belt as a load bed running on relocatable light rail. The bogeys have two axles, each driven by integrated variable speed-geared motor and brake systems, and are connected to previous/next bogeys by specialised couplers. Power is provided by a generator (located in a dedicated bogey) and ‘daisy chain’ plug in cables. Centralised redundant control is provided at a preferred point. Although the system is batch in nature, the number of carriages and trains is flexible, meaning that the system is effectively continuous. Loading is achieved by feeding the ARC from a perpendicularly oriented apron feeder, while material off-loading is accomplished by a twist-dump operation in which the belt is twisted through an angle of 50° below the horizontal, causing material to flow from the belt into a receiving conveyor.The ARC system can run multiple trains over any distance, although benefits are more significant in distances of over two kilometres. Capital and operating costs are predicted to be lower than competitive products, with planned maintenance being simple to perform.CITATION:Graham, T, 2017. The autonomous rail conveyor significantly reduces ore haulage costs, in Proceedings Iron Ore 2017, pp 467–472 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Jul 24, 2017

By D N. B Skinner, R L. Brathwaite, M S. Rattenbury

The Institute of Geological and Nuclear Sciences Ltd has carried out a mineral resource assessment of the Coromandel region, one of the most mineral-rich regions of New Zealand. The assessment focussed on metallic minerals and clays, because these are the main commodities that have potential for increased production and export from the region (and from New Zealand).   Three steps were used in the assessment: 1) development of mineral deposit models, 2) comparison of the models with regional geology, structure, geochemistry and past mineral production to estimate the number of undiscovered mineral deposits, and 3) application of grade-tonnage models to determine the quantity and value of the known and undiscovered resources. In step 1, descriptive and grade-tonnage mineral deposit models were developed for 11 different types of mineral deposits: ò     Low-sulphidation andesite-hosted epithermal Au-Ag ò     Low-sulphidation rhyolite-hosted epithermal Au-Ag ò     High-sulphidation quartz-alunite epithermal Au ò     Porphyry Cu-Au-Mo ò     Andesite-hosted polymetallic veins adjacent to mineralised porphyry stocks ò     Zn-Pb-Ag skarn ò     Cu-Ag skarn ò     Hot spring Hg ò     Shoreline placer Ti ò     Rhyolite-hosted halloysite clay ò     Volcanic-related kaolinite clay   Step 2 was aided by a geographic information system (GIS) assembled from a number of digital spatial data sets including: a digital elevation model, an aeromagnetic map (from open file mining company data), a geological map (from the GNS QMAP project), location of areas of hydrothermal alteration, locations of mineral deposits (from the GERM database), stream sediment geochemical analyses (from the REGCHEM database), and a lineament map derived from satellite remote sensing data and the digital elevation model.   Based on May 2001 commodity prices, potential resources estimated for the 11 mineral deposit and grade-tonnage models total NZ$22 806 million for the metallic minerals, NZ$1000 million for halloysite clay deposits and NZ$57 million for kaolinite clay deposits.

Jan 1, 2001

By D E. Gadd

Since October 1972, New Zealand Steel Mining Ltd has loaded around ten million tonnes of ironsand concentrate from single-point mooring buoys located off-shore at Taharoa on the west coast of the North Island of New Zealand. Before the first off-shore loading system was installed a considerable research effort was applied to proving the slurry loading technique. Subsequently the system has been expanded and a new buoy, pipelines and pumping system were installed in early 1978. Since then 110 000 DWT class ore carriers have been loaded at rates of around 2600 tonnes per hour through twin submarine pipelines. This paper reviews the operation of the ship loading system and attempts to answer the question: 'If the Taharoa operation were to be planned in 1980, would the same research spending be necessary or have published research results since, made such practical tests unnecessary?'

Jan 1, 1980

By E Jak, P Hayes, J Chen, M Shevchenko, E Nekhoroshev, D Shishin

Recent decades have seen significant advancements in analytical and experimental techniques, thermodynamic theory, and computational capabilities in high temperature research which are particularly timely to address the challenges of increasingly complex and variable feedstocks in both primary and secondary pyrometallurgical processes, the need to optimise and improve them. Currently a research program is in progress on the development of the 20-component multi-phase thermodynamic database for pyrometallurgical smelting, refining and recycling systems describing thermodynamics and phase equilibria of molten slag, matte, speiss, salts, alloys and associated solids. The integrated experimental and modelling program has been supported for nearly two decades by several consortia of 14 major international companies with over 30 operations around the world, integrates many components, incorporates many in-house developments, and brings together many different groups of professionals of different levels from junior and mid-career to senior from both industry and university sectors. The aims of the paper are:  to outline the many components and issues of the overall program including analytical, experimental and thermodynamic modelling research as well as implementation of the results in industrial practice, professional education, planning and organisation issues.  to highlight the opportunities, challenges, and possible solutions.  to present these to all different groups of professionals involved in the current program.  and thus, importantly, to facilitate possible future further developments and collaborations in the field of phase equilibria and thermodynamics of complex high-temperature systems.

Jun 19, 2024

A belt of serpentinized ultrabasic rocks intrudes a moderately folded geosynclinal sequence of fine to coarse sediments, acid, intermediate, and basic volcanics as well as a batholithic mass of granitic rocks. The ultrabasic body contains small included masses of melanocratic and leucocratic rocks, together with chromium, copper, zinc, and nickel mineralization.

Jan 1, 1969

By D G. Randall, R Jivanji, J Nathoo, T Reddy

Paradoxically, the increased emphasis on water treatment in mining has generated a new problem û hypersaline brines. Currently brines are either disposed of into ash systems, lined evaporation ponds or hazardous waste sites. Evaporative crystallisation (EC) is an alternative, but the energy requirements can be prohibitive and the salt product is still a waste that must be disposed of. Eutectic freeze crystallisation (EFC) is an alternative technology that is capable of separating aqueous solutions into pure water (as ice) and pure, solidified solutes. It is highly energy efficient and does not depend on the introduction of any solvents. In addition, the simultaneous production of pure ice and potentially saleable pure salt(s) is a major advantage. Although EFC has been shown to be effective in separating a single salt and water, it is still in the early stages of being applied to the multi-component hypersaline brines that are typical of reverse osmosis retentates generated by mining and extractive metallurgy.This paper presents an EFC brine treatment protocol developed by the Crystallisation and Precipitation Unit at the University of Cape Town. The protocol incorporates brine analysis, thermodynamic modelling to establish which salts will freeze out and at which temperatures, followed by experimental studies to determine the kinetic parameters, such as nucleation temperatures, salt yields and crystal purity.Case studies showing the application of EFC for the purification of typical brines containing high levels of sodium, chlorine, sulfate and ammonia (Brine 1) and a mixture of sodium and magnesium sulfate (Brine 2) are covered. Modelling tools were applied to describe the phase behaviour of the multi-component saline systems under different process conditions. Experimental validation was carried out using synthetic brines of different (and increasingly complex) compositions. The results showed that EFC could be used to selectively recover the sodium as a sodium sulfate salt from both different brines. The simulation tools were useful in the analysis and understanding of the process.A preliminary economic analysis showed that the operating costs for EC were approximately nine times those for EFC, but the capital costs were much less- equipment for EC cost about half that of EFC. However, it is important to note that the capital cost calculations for EC are based on a technology that is already well established and with only relatively marginal future equipment cost savings expected. In contrast, EFC is a new technology, with significant future technology improvements expected.

Jan 1, 2009

By J Fisher, M Castro, A Hariyadi

The train haulage system being designed for PT Freeport Indonesia’s Grasberg Block Cave (GBC) will be located within the mine at an elevation of almost 3000 m above sea level on the west side of the island of Papua, Indonesia. The developing block cave is beneath the active Grasberg Open Pit and will transport 160 000 t of ore per day from multiple chute galleries to the underground crushing plants. The design entails 28 km of rail and 119 ore chutes, which will be brought online over the course of ten years.The train haulage system will be driverless and fully automated. Sequentially, the system includes remote controlled rock breakers at the top of the chute grizzlies, followed by remote controlled chute loading of the trains, then autonomous train haulage and autonomous unloading. An overhead catenary system (OCS) will power the locomotives, which are equipped with on-board batteries to drive through the loading and unloading sections. Service locomotives with drivers move material-laden freight cars in and out of the service shaft, destined for workshops, the magazine and batch plants. Service locomotives are also used for haulage level maintenance. Highly reliable and redundant control and communications systems are required to manage 12 trains in simultaneous operation.This paper provides an overview of the train haulage system’s technical complexities and the associated support systems interactions. The paper also highlights the project management issues and successes of this globally sourced project.CITATION:Hariyadi, A, Castro, M and Fisher, J, 2016. Automated train ore transport, in Proceedings Seventh International Conference and Exhibition on Mass Mining (MassMin 2016), pp 563–570 (The Australasian Institute of Mining and Metallurgy: Melbourne).

May 9, 2016

By S C. Labram, A Stabile

"Modern train load out stations are becoming faster and have the ability to load a 100 wagon train in less than three hours. At 70 metric tonnes per wagon, that is an average fill rate of 2333 t/h and considering the train is continuously moving, the instantaneous feed rate into the wagon is in the region of 4200 t/h.The main feed conveyor that maintains the level in the surge hopper, above the final weigh hopper, operates in excess of 5000 t/h to ensure uninterrupted charging of the weigh hopper to not disrupt the train loading operation.To avoid demurrage costs (a charge by the Rail Operating Company for exceeding the contractually agreed loading time) the train has to be handled speedily and efficiently and spend the absolute minimum time at the loading facility. Everything else has a much lower priority, including sampling.Thermo Fisher Scientific’s brief was an engineering, procurement and construction management (EPCM) contract to provide a sampling scheme adjacent to the prepurchased and predesigned load-out facility, but which interfaced seamlessly with the loading system of the trains and provided the following: an instantaneous particle size distribution (PSD) result for each truck loaded, and a final PSD for the total train prior to the train leaving the loading facility a crushed (-3 mm) composite sample for chemical analysis for the whole train a composite physical analysis sample for the whole train.The design presented some interesting challenges to our team, but the end result is a system that not only ‘stands alone’ but interfaces seamlessly with the loading system. This paper discusses the design itself and potential improvements to future systems.CITATION:Labram, S C and Stabile, A, 2017. An automatic sampling methodology to determine contractual particle size distribution in rapid train load out stations, in Proceedings Eighth World Conference on Sampling and Blending, pp 229–234 (The Australasian Institute of Mining and Metallurgy: Melbourne)."

May 9, 2017

By Bagley A. S

The accuracy and precision of rutile analyses are discussed. Comparison is made between the wet chemical and X-ray fluorescence spectrographic methods of. estimating Ti02 in commercial rutile. Various allotropes of 'I'i02 are identified by diffractometry and a comparative analysis of the coloured varieties of rutile is given. For X-ray fiuorescence . analysis, specimen preparation is critical and sample grinding and pressures. for pellet making are discussed in detail.Cellulose bonded and backed pellets are shown to be subject to relaxation with increasing age.INTRODUCTIONTo date much of the analytical work on commercial samples has been done by wet chemical methods.This led to various discrepancies in results, which are of particular importance when certificates must be issued for overseas shipments. The present work reviews the nature of rutile and compares the results of Ti02 estimation by wet chemical methods and those obtained by X-ray fluorescence spectrographic analysis.In the rutile industry the results from different analysts and even those from the same analyst are not always consistent. To the analytical chemist this should not be surprising.Fairbairn (1951) showed the results obtained from 34 different analysts on a specially prepared sample of granitic rock. It is of some interest to note that the Ti02 content as determined by the 34 analysts using unspecified methods ranged from 0-17 per cent to 0·30 per cent. Fairbairn (1953) found that1. replicate analyses from the same analyst give a high order of precision,2. analyses by different analysts using the same procedures give reasonable precision, and...

Jan 1, 1965

By I Coombes, M Trim, Y Sun, M McRae, A Spiegl, H Lagger, A Amon, P Barraclough, M Villeneuve

The Brisbane Airport Link Project (AirportlinkM7) includes the design and construction of more than 12.5 km of tunnels, with the Kedron interchange including four ramp tunnels that intersect with mainline tunnels and result in caverns with spans of up to 26 m. These caverns are in highly variable ground conditions with less than 35 m of cover underlying residential neighbourhoods and commercial properties. The Kedron caverns were constructed using the sequential excavation method (SEM). Owing to complex geometry and excavation sequence, the caverns were designed based on numerical modelling methods, including both two- and three-dimensional models. These numerical methods were used to develop a suite of support types with excavation sequences for the expected range of ground conditions. A combination of methods that included in-tunnel monitoring (convergence monitoring and extensometers), probe holes, endoscope probes, face mapping and ground settlement measurements were used to verify performance of the tunnel support systems and to select and refine the appropriate support system for each section of the caverns. The construction monitoring data indicated that numerical modelling can provide a good basis for predicting the performance of large-span underground openings in highly variable ground conditions. This paper presents a comparison of the performance of the caverns predicted by state-of-the-art numerical models and the actual measured performance.CITATION:McRae, M, Villeneuve, M, Trim, M, Sun, Y, Barraclough, P, Spiegl, A, Amon, A, Lagger, H and Coombes, I, 2014. Brisbane Airport Link Project – design and construction of the 26 m span Kedron Caverns: comparing monitoring data with the design, in Proceedings 15th Australasian Tunnelling Conference 2014 , pp 295–306 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 17, 2014

It is no longer satisfactory to assess the value of a coking coal on its intrinsic properties alone; its value depends upon its behaviour as a component in a blend. No single test can be used to make this assessment, it can only be made by carrying out a suite of tests. The quality parameters which are considered desirable, and which have widespread acceptance, have been discussed and their application to the blending of coals from New South Wales and New Zealand has been reported and discussed. Blending can be utilised to offset deficiencies, or capitalise on advantages, of the individual coals making up a blend. Coals can be blended, which alone are poorly, Coking, but which have compensating features. Blending also allows better exploitation of coking coal reserves. Several applications of blending have been reviewed with regard to coal producers and exporters, as well as coal users. The possibility of extracting a coking coal fraction from fuel coals is discussed. Blending systems and the coking of coal are also briefly reviewed.

Jan 1, 1971

By A B. Christie, F P. Bierlein

Lode gold deposits of the Western and Eastern provinces, South Island, New Zealand , formed as a part of the Cordilleran-style tectonism during Tasman/Tahua and Rangitata orogenesis along the Paleozoic-middle Mesozoic southern Gondwanan margin. These deposits, hosted by lower grade greenschist facies sedimentary rock sequences of the Buller, Takaka, Torlesse, and Caples terranes, universally show geological features that are consistent with the orogenic gold deposit model, and genetic features are generally suggestive of a metamorphic origin, with the possible exception of a magmatic fluid source for the mineralisation at Sams Creek. However, ore fluid chemistry in Mesozoic deposits of the Eastern province, particularly the low concentration of non-aqueous volatiles and the very dilute nature of the fluids, is uncharacteristic of most orogenic gold deposits; this must be explained by either a unique fluid source or fluid type, which we believe unlikely, or an unrecognised significance of relatively rare gas-rich fluid inclusions in the deposits. Absolute age data for gold deposits in both provinces remain equivocal, thus hindering definition of the exact temporal relationship between ore formation and terrane accretion, regional metamorphism, and magmatic arc emplacement. Mercury and antimony occurrences in subgreenschist grade parts of the Haast Schist may target the upper levels of unrecognised orogenic gold deposits. If low-grade, high-tonnage intrusion-related gold systems occur on the South Island, then they are most likely to be discovered in association with the small molybdenite-, scheelite-, or cassiterite-bearing quartz stockworks spatially related to the Karamea and Median batholiths.

Jan 1, 2005

Modelling and Optimisation of Direct Reduction of Iron Ore by Coal in a Rotary Kiln

Sep 13, 2010

By P Lotz

With the recent increase in environmental awareness, gold mining companies are fast becoming signatories of the International Cyanide Management Code (ICMI, 2007). The objective of the Code is to improve the management of cyanide used in gold mining and to assist in the protection of human health and the reduction of environmental impacts associated with the use of cyanide. In particular the code specifies the following cyanide concentration limits: 50 ppm WAD cyanide in a tailings storage facility, 0.5 ppm WAD cyanide for effluents, and 0.022 ppm free cyanide or 0.022 ppm WAD cyanide as surrogate in the post mixing zone of river discharge. Weak acid dissociable (WAD) cyanide refers to those cyanide species liberated at a moderate pH of 4.5, such as HCN(aq) and CN-, and the majority of Zn, Cd, Cu, Ni and Ag complexes with low dissociation constants. It is potentially toxic to humans and animals. Gold plants using cyanide should ideally measure WAD cyanide online and destroy WAD cyanide in excess of the applicable limits to ensure compliance with the Code. Mintek has recently developed an online WAD cyanide analyser (WAD Cynoprobe). The WAD cyanide analysis technique is based on the use of a ligand exchange reagent that liberates WAD cyanide bound to metals such as Zn, Cd, Cu, Ni and Ag. The ligand exchange reaction is represented by the following chemical equation: Metal(CN)X + yLigand ↔ Metal(Ligand)y + xCN- The plant filtrate is treated with a ligand exchange reagent, to liberate WAD cyanide, before measuring the free cyanide concentration via amperometry. The WAD Cynoprobe measures both free and WAD cyanide concentration. Its fast analysis time, approximately five minutes, makes it suitable for use in the control and optimisation of the cyanide destruction and recovery process. The instrument has proven accurate and reliable in the analysis of synthetic copper, nickel and zinc cyanide solutions made up in the laboratory as well as under industrial gold plant conditions during various trials. The WAD Cynoprobe has gained widespread commercial acceptance and is currently implemented on nine gold plants around the world.

Jan 1, 2007

By D Craw, G Eis

Late Cretaceous Blue Spur Conglomerate preserved as isolated remnants at Gabriels Gully, Weatherstones, Forsyth, Waitahuna Gully and Adams Flat was deposited during the intiation and early stages of extensional movement on the Tuapeka Fault zone at each locality. The deposits represent successive courses of a single fluvial system flowing across the Tuapeka Fault zone from the up-thrown (northeast) block. Each course, beginning with the channel at Gabriels Gully, was abandoned as a result of uplift-induced defeated antecedence in favour of a new course farther to the southeast around the southeast-propagating tip of the Tuapeka Fault zone. Schist and quartz in the conglomerate was derived ultimately from schistose rocks northeast of the Tuapeka Fault zone. Caples Terrane greywacke detritus in the middle and upper beds of the conglomerate was probably recycled from pre-existing sedimentary cover on the up-thrown block, following initiation of uplift on the Tuapeka Fault zone at each locality. Most of the gold is derived from vein sources in Otago Schist rocks to the north, and/or associated placers. Hydrothermal veins at Waipori, Hindon, Barewood, and Macraes are the most likely primary sources. There is no evidence to support previous interpretations of deposition by river systems flowing from the southwest, or axially along the Tuapeka Fault zone from northwest to southeast. High gold grades within Blue Spur Conglomerate should have a linear trend more or less similar to the orientation of the host channels, but are likely to drop off rapidly to the southwest. Extensions of the conglomerate-filled channels are unlikely to be preserved, but there is a strong probability that younger placers recycled from the former channels exist.

Jan 1, 1999

From time to time lead, which is normally considered to be a soft ductile metal, fails by cracking. The cracks which form appear to be of two distinct types-(a) those which follow a zig-zag course around the boundaries of crystals and (b) those which follow a more direct path, frequently two systems of cracks appearing at right angles.The former appear definitely to be associated primarily with the properties of the crystal boundaries, whilst the latter appear rather to be determined by the system of stresses applied. In both cases the lead appears to be a brittle rather than a ductile metal, for the failure occurs without the large amount of deformation usually associated with the tensile breaking of lead. The latter is, however, a function of time and when lead is deformed by long applications of stress, the metal appears to- be less ductile than when the stress is applied suddenly. Thus, if a given test piece of commercial lead is broken by a stress of 1100 lb. per sq. in. in several hours, it may extend 40 to 50%before failing, whereas when broken by a stress of 500 lb. per sq. in. acting for 12 months or more it willshow an extension of only 15 to 25 %. With lower stresses still, failure would take much longer and undoubtedly lower extensions would be obtained.As the stress is lowered it is found that the marked local contraction characteristic of ductile metals becomes less noticeable, and failure, instead of being by extension of *Lecture delivered to the First 'Ordinary Meeting of the Australasian Institute of Mining and Metallurgy, 21st August, 1935.

Jan 1, 1935

Predicting the future is a task that should always be delegated. If that is not possible, then stating your views confidently, but retrospectively, is preferable. Hindsight encourages clarity of thought and sharpens perception. When all else fails one can look back into history and see the trends, evaluate the current situation and project it a little way into the future - in this case 10 years. In exploration the rate of change has been increasing over the past 20 years. With expected developments in technology, legislation, and their sociological impacts, we will see such rapid change over the next decade as to make us uncomfortable, unless we are prepared for it.   The social obligations of our industry also have changed enormously over the past 20 years and are still changing. This will lead to further changes in the way we conduct our business if we wish to stay in business. As exploration technology improves, a further dimension is added by the expectations of the communities in which we operate. This extra dimension needs to be quantified so that we not only keep up with current developments but also get ahead of the game. This paper is based on the New Zealand experience, some of which will be familiar to mineral explorers in other countries.

Jan 1, 1997

The development of a robust belt sampling strategy and protocol inunderground gold mines is a fundamental requirement for effective grade quality control, ore accounting and mine-mill reconciliations. As part of the ore management system, a belt sampling system has been developed by the author, which is used to predict monthly gold grade and its error for underground/belt/mill reconciliation and grade control. This paper presents the step-by-step development of the belt samplingstrategy and sample preparation protocol at Target, an underground gold mine in South Africa. The belt sampling system at Target consists of an underground belt transporting run-of-mine material into an underground transfer point, a weightometer, a two stage crushing system, a tube splitter and a catchment device. The design allows run-of-mine ore on the belt to be sampled automatically by a go-belt sampler. GyÆs sampling theory of particulate materials forms the basicframework of the belt sampling strategy. The theory is used to quantify the optimum sample mass of broken material to be taken from the belt and to predict its reliability. It is also employed to determine the Fundamental Sampling Error (FSE), time interval between samples and particle size requirements of the final underground belt sample. Further, GyÆs sampling theory provides the means of analysing the crushing processes, splitting, segregation error, analytical variance, and the best sample preparation protocol. Although the belt sampling strategy developed in this study was forWitwaterstrand gold ore, the process is entirely general and may be very successfully used in any underground or surface mine.

Jan 1, 1999