CONCLUDING REMARKS The volume and pervasiveness of the volcanism in the Gregory Rift of northern Tanzania is almost without paral el. The fal -out of radioactive ash associated with S Oldupai Gorge showing E the basal lava (black) the younger, nephelinite and carbonatite volcanism, it which is overlain by L has been suggested, may account for rapid speciation fossil-rich Beds I and II of both cichlid fish (e.g., in Lake Victoria) and of Great of the Oldupai sequence. IC Apes and Hominids. T R Photographs from visits in 995, 998, and 2005. A graphite petrography Introduction Industrial minerals such as graphite have recently Graphite purity is particularly important for the higher become the focus of attention for many listed exploration value end uses like lithium-ion batteries and is a key companies, particularly due to developments in battery determinant in saleability of the product. Producing technologies and new product opportunities such as high purity graphite may also adversely affect the cost graphene. Consequently the race has been on to report of production, as additional processing to make the larger tonnage exploration targets and resources, with product saleable will increase the operating cost. certain projects being described, for example, as world class or highest grade. Graphite flake size distribution is one of the more debated project factors. However a number of facts Although resource tonnes and graphitic carbon content about flake size are currently true; firstly, the larger the (grade) are important metrics, the evaluation of graphite flake the higher the purity of the graphite product is projects (as with other industrial minerals) is more likely to be and secondly, the larger and purer the flake complex; key attributes in addition to deposit size and size the higher is the sel ing the price. grade, are product flake size distribution and purity (Scogings, 20 4; Scogings and Chesters, 20 4). Graphite prices related to flake size and purity Graphite type Purity Size Size Low High (% Carbon) (mesh) (Microns) (US$, CIF) (US$, CIF) Flake 94 to 97 +80 + 80 ,050 , 50 Flake 94 to 97 + 00 -80 + 50 - 80 900 ,000 Flake 94 to 97 - 00 - 50 + 200 750 800 Flake 90 +80 + 80 750 850 Flake 90 + 00 -80 + 50 - 80 700 800 Flake 90 - 00 - 50 + 200 600 650 Flake 85 to 87 + 00 -80 + 50 - 80 550 600 Amorphous 80 to 85 -200 -75 430 480 Amorphous 70 to 75 500 (ex-works) 550 SOURCE: Industrial Minerals Magazine 30th April 20 5 www.indmin.com
30 geobulletin JUNE 2015
Volume 58 ~ Number TWO ~ JUNE 2015 SAMREC 2009 Table 1, Section 5.5 Treatment / Processing S ASSESSMENT CRITERION: T 5.5 Treatment / Processing E L EXPLORATION RESULTS (A) IC (i) Describe any obvious processing factors that could have a significant effect on the prospects of any possible T exploration target or deposit R A MINERAL RESOURCES (B) (i) Discuss the level of study, possible processing methods and any processing factors that could have a material effect on the likelihood of eventual economic extraction. (i ) The basis for assumptions or predictions regarding metal urgical amenability and any preliminary metal urgical test work should already be carried out. (i i) It may not be possible to make assumptions regarding metal urgical processes and parameters when reporting Mineral Resources. Where no assumptions have been made, this should be explained. Source: The SAMREC Code 2007, as amended July 2009 The current edition of the SAMREC Code for public in addition to basic assay tests for graphitic carbon reporting of Exploration Results, Mineral Resources content. In this regard, the question is often raised and Ore Reserves includes Table which is a high- about how to test graphite flake size across a deposit; level checklist of assessment and reporting criteria. given that relatively costly and time-consuming lab Although not prescriptive, it is important for the flotation procedures are usual y required to separate Competent Person (“CP”) to “report all matters that graphite from gangue minerals. might material y affect a reader’s understanding or interpretation of the results or estimates being reported” The author suggests that petrographic examination of (SAMREC, page 28). The Code goes further and states polished thin sections is a relatively affordable and that the CP has the responsibility to consider all criteria quick option to estimate the in-situ graphite flake size listed and which additional criteria should apply to the distribution and likely liberation characteristics. particular project. Back to basics – take a look at the rocks The author’s intention is to address some of the criteria listed under ‘Treatment / Processing’ of Section 5.5 of The microscopic investigation of rocks in transmitted SAMREC Table and to provide examples from graphite and reflected light is one of the classic mineralogical exploration projects related to the issue of flake size methods of analysis. Polarized-light microscopy and liberation characteristics. These criteria include provides a non-destructive way to identify minerals, as the description of any “obvious processing factors that they can be studied within their textural framework. could have a significant effect on the prospects of any This method provides clues to the history of formation possible exploration target or deposit” and “the basis of the material, using specific textural characteristics for assumptions or predictions regarding metal urgical such as structural fabric, mineral assemblages and amenability and any preliminary metal urgical test relationships and has distinct advantages over bulk- work”. analytical methods that use sample powders for mineral identification or chemical composition such Given that industrial minerals such as graphite are as XRD and XRF. It is recommended that polarized- normal y produced and sold according to size, purity light microscopy, complemented by methods such as and / or performance specifications, the responsibility SEM (Scanning Electron Microscope), QEMSCAN falls on the CP to ensure that samples are tested for (Quantitative Evaluation of Minerals by Scanning appropriate parameters such as flake size and purity, Electron Microscopy) and MLA (Mineral Liberation
Analyser, or automated SEM) should be used when evaluating a graphite project. • Petrographic studies to define geological domains (e.g. lithologies, mineralogy and textural Graphite explorers are encouraged to ‘get back to characteristics); S basics’ and use thin section petrography as a primary • Selection of samples according to geological E tool to evaluate and compare prospective targets domains, for metal urgical testwork; L (Scogings, 20 5). Thin section petrography helps with • Laboratory-scale test work to determine the IC the geometal urgical domaining of graphite deposits response to mineral processing methods; T and selection of composites for metal urgical testing, R in addition to explaining subsequent metal urgical test In order to define geometal urgical domains, a suite A results. of samples may be prepared, representing the main lithologies from which two thin sections can be made Geometal urgy is multi-disciplinary approach that for each sample, one perpendicular to the graphite combines geology and mineralogy with extractive flakes and a second approximately paral el to the metal urgy, to create a predictive model that assists with flakes. In the case of RC chips, these can be cast into selecting appropriate mineral processing for a deposit. resin and made into polished thin sections in which It is used to reduce risk during mineral processing plant case the sample orientation is random. design and can also assist with production scheduling. Mining and processing based solely on grade (e.g. Graphite populations graphite content) may not be sufficient, as seemingly low grade mineralisation may result in a high quality An example of how thin section microscopy can help concentrate if processed appropriately. understand metal urgical results is where graphite recoveries are lower than anticipated. Thin section There are several steps that may be used in developing examination highlighted that there were two graphite a geometal urgical model for a graphite deposit, populations, with the majority as coarse flakes but with including: a second population of very small flakes within large crystals (porphyroblasts) of K feldspar. The small flakes Core samples representative of rock types on a specific project. Approximate graphite contents indicated as percentages.
Volume 58 ~ Number TWO ~ JUNE 2015 S E L IC T R A Polished thin sections made from core or outcrop (two samples on the left) and RC chips (two samples on the right)
Photomicrograph illustrating two graphite populations within one Photomicrograph of large ‘clean’ graphite flakes with one minor sample: large flakes in the general rock matrix, compared with fine sulphide inclusion flakes within K feldspar were not being liberated at the coarse crushing size used to liberate large flakes, hence were not being recovered. Mineral impurities Sulphide minerals such as pyrite and pyrrhotite are common impurities in graphite deposits and thin section petrography can help define areas or specific lithologies where sulphides may be absent, present as discrete grains or interleaved within graphite flakes Photomicrograph of graphite flakes with discrete pyrite crystals and thus more difficult to liberate. and pyrrhotite blebs
Flake size A second example of the benefits of thin section petrography may be where flake size varies across an S individual deposit, or between prospects within a region E and where the explorer wishes to select an appropriate L target. In this particular case, the explorer identified one target as having a population of very small flakes IC T in a retrograde sericitic assemblage (Figure 7) and a R second target as containing coarse flakes in a medium to high grade metamorphic assemblage (Figure 8) and A elected to fol ow up on the second target. Conclusions and recommendations Photomicrograph of pyrrhotite interleaved along cleavage planes in graphite • Graphite explorers are urged to ‘get back to basics’ and use thin section petrography as a basic tool to help address treatment / processing aspects of industrial mineral resources according to SAMREC 2009 requirements. • It is suggested that petrographic examination of polished thin sections be done early on in the project and during the subsequent resource dril ing phase. Polished thin sections are relatively inexpensive and can be used to estimate the size and shape of in-situ graphite flake populations, relationships with other minerals including contaminants such as sulphide minerals, and for estimating likely liberation size. Photomicrograph of small graphite flakes generally less than 75 microns in length, inter- However it should be borne in mind that in situ flake grown with sericite size estimations don’t necessarily translate directly to flake sizes produced by metal urgical processes such as gravity separation or froth flotation. • Core dril ing is the preferred technique for graphite exploration, as this provides undisturbed samples for thin sections and for metal urgical tests. Reverse Circulation (RC) drill chips may also be used to make thin sections; however RC chips are not suitable for metal urgical tests due to the grinding action of the drill bit, which reduces flake size. Caution should be exercised when selecting RC chips, as there may have been preferential grinding of softer (possibly graphite-rich) rocks. •Once appropriate metal urgical samples have been Photomicrograph of coarse graphite flake more than 1mm in length, associated with biotite selected, some cost-effective and fairly quick tests
Volume 58 ~ Number TWO ~ JUNE 2015 such as i) assay by size and i ) heavy liquid separation (SG 2.3) may be considered as precursors to more Scogings, A.J. and Chesters, J., 20 4. Graphite: The six S detailed flotation tests. steps to striking success. Industrial Minerals Magazine, E L December 20 4, 37-44. Industrial Minerals, Euromoney • Knowing this type of information means being Institutional Investor PLC, Nestor House, Playhouse IC smarter early on in the project and can guide to Yard, London EC4V 5EX, United Kingdom. T more intel igent and informed selection of composite R drill samples for metal urgical testing, in addition Scogings, A.J., 20 5. Graphite: Where size matters. A to benefitting mine planning and metal urgical Australia’s Paydirt, March 20 5, p.78-79. Paydirt processing further down the line. Media Pty Ltd, Hay Street, West Perth, Western Australia 6005. www.paydirt.com.au References Acknowledgements SAMREC, 2009. The South African code for the reporting of exploration results, mineral resources Sincere thanks are extended to CSA Global Pty Ltd for and mineral reserves. Prepared by the South African their support in publishing this review. Mineral Resource Committee (SAMREC) Working Group. www.samcode.co.za Andrew Scogings PhD, MAIG, MAusIMM, MGSSA, RP Geo. (Industrial Scogings, A.J., 20 4. Industrial Minerals – Reporting Minerals) Resources according to Clause 49 of JORC 20 2. Principal Consultant, CSA Global Pty Ltd Geobul etin, June 20 4, Vol. 57, No. 2, 28-33. Industrial Minerals Magazine Consultant Geological Society of South Africa Quarterly News Perth, Australia Bul etin. densityIndustrial Minerals - Introduction Why measure Bulk Density? Bulk density is often a relatively neglected parameter Geological resources are general y modelled as volumes during industrial mineral exploration and general y in three-dimensional space, after which the estimated doesn’t receive the attention devoted to other measures volume must be converted to mass using density value/s, such as i) sample width in borehole intersections, i ) thus the measurement of density should be an integral chemical analysis and i i) product performance testing part of the resource estimation process. (Scogings, 20 5b). The author’s intention is to address certain aspects As noted by Lipton and Horton (20 4, p.97) “There concerning bulk density listed in Section 2.4 of the are three fundamental inputs to any Mineral Resource SAMREC code Table . The first part of this review estimate: grade, volume and bulk density”; they also describes bulk density and some of the methods most state that “The estimation of density commonly receives commonly used for measuring the density of rocks and less attention than is paid to geochemical data and materials; this is supported by case studies from Minerals may be based on fewer data points derived from less Technologies Inc. (“MTI”) mines in Australia and South control ed measurement practices”. Africa, where the author was previously involved.
SAMREC 2009 Table , Section 2.4 Specific gravity and bulk tonnage data ASSESSMENT CRITERION: T 2.4 Specific gravity and bulk tonnage data S EXPLORATION RESULTS (A) E L (i) If target tonnage ranges are reported then the preliminary estimates or basis of assumptions made for bulk density or specific gravity(s) must be stated. IC (i ) Specific gravity samples must be representative of the material for which a grade rage is reported. T R MINERAL RESOURCES (B) A (i) Describe the method of bulk-density / specific-gravity determination with reference to the frequency of measurements, the size, nature and representativeness of the samples. (i ) The bulk density must have been measured by methods that adequately account for void spaces (vugs, porosity, etc), moisture and differences between rock and alteration zones within the deposit. (i i) Direct assumptions for bulk density estimates used in the evaluation process of the different materials. Source: The SAMREC Code 2007, as amended July 2009 The current edition of the SAMREC code for public What is Bulk Density? reporting of Exploration Results, Mineral Resources and Ore Reserves includes Table which is a high- Bulk density is a measure of mass per unit volume of level checklist of assessment and reporting criteria ( rock and in South Africa it is general y expressed as metric tonnes per cubic metre. Density is determined Table ). Although not prescriptive, it is important for by measuring the mass of a sample and dividing this the Competent Person (“CP”) to “report all matters that by its volume. General y the dry mass is obtained by might material y affect a reader’s understanding or drying the sample and then weighing it, which is the interpretation of the results or estimates being reported” easy part. The chal enging bit happens when trying to (SAMREC, page 28). The Code goes further and states determine the volume of a rock sample especial y when that the CP has the responsibility to consider all criteria specimens have irregular shapes, are friable, soft and listed and which additional criteria should apply to the / or porous. particular project. Density may be defined in a number of ways (Table 2) The author’s intention is to address some of the and it is important to ensure that the appropriate density criteria listed under Section 2.4 ‘Specific gravity measurement is used for any specific project. Assays and bulk tonnage data’ of SAMREC Table and to for constituents such as Cr2O3 (in a chromitite seam), provide examples from industrial mineral exploration MgO (in magnesite or dolomite) or Graphitic Carbon and mining projects related to the issue of drill core, (in a graphite schist) are normal y reported on a dry stockpile and product density. These criteria include the weight basis and therefore in such cases the ‘dry bulk description of “the method of bulk-density / specific- density’ (DBD) is applicable. Lipton and Horton (20 4) gravity determination with reference to the frequency of define DBD as the mass per unit volume, including pore measurements, the size, nature and representativeness spaces but excluding natural water content. of the samples” and that “the bulk density must have been measured by methods that adequately account The ‘in situ bulk density’ (ISBD) includes natural water for void spaces (vugs, porosity, etc), moisture and content and according to Lipton and Horton (20 4) differences between rock and alteration zones within should be applied when estimating tonnages of the deposit”. material to be mined. The use of ISBD would apply
Volume 58 ~ Number TWO ~ JUNE 2015 S E L IC T R A 2 Paraffin wax coated Bentonite core sealed inside plastic to preserve in situ moisture friable chromitite content. Source: MTI sample being weighed in to a commodity such as bentonite, which may contain air. Source: MTI 25 to 35% moisture before being mined and core should therefore be sealed immediately after dril ing to preserve in situ moisture content. Specific Gravity (SG) is commonly used to describe density but caution should be exercised, as SG (also known as Relative Density) is often measured using pulverised samples in equipment known as a 3 Archimedes (water pycnometer. This method does not take into account displacement) equipment. Competent, porosity or natural water content, which is a limitation of non-porous chromitite the method for use in geological resource estimation. core sample being Determining bulk density from small samples weighed in water. Source: MTI The geologist frequently has only drill core samples sample mass in air and in water). Samples should be to use for density measurement and there are several competent and not absorb water; if porous they should practical methods available, essential y based around be waterproofed with substances such as paraffin wax the issue of measuring volume. Each density method or beeswax which melt at ~60oC, spray lacquer or has its own potential source of error and it is useful to verify the results of one method against a second if at all possible. It is important to ensure that rock / mineralised samples are representative and that a 4 Paraffin wax particular type of rock is not sampled preferential y, blocks. Source: Protea e.g. hard material relative to soft material (Lipton and Chemicals, Wadeville Horton, 20 4). Water displacement method: there are several methods which rely on displacement of water to estimate sample volume and are described in detail by Lipton and Horton (20 4, pages 99- 0 ) who list six water displacement methods. One of the most common methods for exploration samples is based on the Archimedes principle in which the sample is first weighed in air, after which it is weighed in water. The density is calculated as the mass of the sample in air, divided by the volume (difference between the
S Coating a friable E chromitite core sample L with molten paraffin wax. IC Source: MTI T R A Pyroxenite core sample being measured using a calliper. Source: MTI results (Lipton and Horton, 20 4). The cal iper method may also be used on half core samples, such as might be encountered when evaluating an older project with previously cut core, or where core has been sampled prior to measuring density. Errors will arise if the core was not aligned correctly when cutting and this should be verified before proceeding with this method. The fol owing parameters should be measured on half core namely i) core length; i ) core segment width and; i i) core segment height (Lipton and Horton, 20 4). Vacuum-packed hairspray, vacuum packed in plastic or wrapped in pyroxenite core. ‘cling wrap’ film to help prevent ingress of water. Source: MTI Pulp sample method: density of competent rocks that have very low porosity and low natural water content may be measured using a gas pycnometer and rock pulp samples (finely mil ed rock) but this method is not suitable for porous rocks, as the fabric is destroyed by the mil ing process. The gas pycnometer method determines volume within the sample chamber from which the gas is excluded. The pycnometer will accurately give volumes for samples weighed into plastic vials which are in turn dropped into the sample chamber. Best precision is obtained from the largest Pyroxenite core covered Cal iper method: this is applicable for drill core samples with cling wrap, possible volume of sample which is typical y around 30 illustrating water that can be trimmed at right angles to form a regular grams. SG data derived from a gas pycnometer may ingress. Source: MTI cylinder. A pair of cal ipers is used to measure the form a useful part of the density database and in the core diameter at several points to estimate an average author’s experience such SG data can be a valuable result, while the core length is determined using a tape QC tool. measure or ruler. The core is weighed and the density determined simply by using the formula of weight / Stoichiometric method: there may be an obvious volume. The cal iper method has the advantage of correlation between SG or bulk density and rock simplicity, but it is cautioned that using small diameter chemistry, such as with relatively simple mineral core or short core lengths may result in unreliable assemblages such as some barite and chromite ores.
Volume 58 ~ Number TWO ~ JUNE 2015 the mineral resource or ore reserve is also a good S check, not only of the three-dimensional geological E model, but also of bulk density. L Operating mines general y measure raw material IC stockpile volumes for audit and reconciliation purposes, T but the question arises of selecting an appropriate R bulk density for conversion of volume to mass. Bulk A density (BD) values for free-flowing powders and granular materials can vary significantly according to particle size distribution and on how closely the particles are packed. Since powders and granular materials are composed of particles and voids, the volume of a given mass of particles depends on how Gas pycnometer. Source: Intertek, Perth closely they are packed. In practical terms, the bulk density of a powder tends to increase the more it is Assuming that a barite ore consists of discrete barite subjected to tapping, vibration or other action which (BaSO4) and quartz (SiO2) or that chromitite ore causes particles to become better packed, with less consists essential y of chromite and pyroxene minerals, void space between larger particles; this is known as it should be possible to estimate bulk density based the ‘tapped bulk density’. Bulk density of free-flowing on XRF ‘whole rock’ analyses. An example is barite powders / granular materials can be determined by ore in which pure barite has a density of ~4.5 g/ml fil ing a container of known volume, at which stage compared with quartz which has a much lower density the material is weighed and the ‘loose bulk density’ of ~2.7 g/ml. can be estimated. The container is then tapped and refil ed until the material stops settling at which stage Seeing that density is expressed in terms of volume the tapped bulk density can be estimated. and that XRF whole-rock analyses are expressed on a weight percentage basis, the calculated density must QAQC be based on mineral volumes in order to maintain a constant volume. The relationship between whole- QAQC methods commonly applied to other factors in rock chemistry and density is non-linear, which is an exploration program such as equipment calibration, especial y obvious when there is a marked difference duplicates, standards and external laboratory tests in SG between the different mineral phases (Lipton and should also apply to density measurements. Horton, 20 4).
Determining bulk density from larger samples Bulk samples may be obtained if trial mining or Calibration of a density production is already in progress at a site. The in-situ balance using standard volume of bulk samples can be estimated by surveying weights. Source: MTI an excavated void (for example an extracted bentonite or chromitite seam) or by surveying a stockpile before and after removal. The sample mass may be determined by directly measuring truckloads across a weighbridge; however sub- samples will have to be taken to determine moisture content as it is impractical to measure the moisture of an entire stockpile or run of mine material. Reconciliation of tonnage mined against
Case study – chromitite in South Africa method (uncoated, vacuum packed and paraffin wax). However the ‘cling wrap’ method proved to The main aim of this paper is to give practical be unreliable as it entrained air (reducing the density) examples of density measurement and the first example and was not waterproof. The author has observed S is of drill core from the Batlhako mine at Ruighoek; this significant differences when using cling wrap on other E MTI operation produces a range of premium-grade industrial mineral projects and recommends avoiding L chromite sands for foundry, chemical, metal urgical this method. IC and refractory applications. T The density of a second sample of un-weathered pyroxenite R In the first example the LG6 chromitite is ‘fresh’ or un- core ‘SG 0’ was measured by various methods, after A weathered competent rock consisting predominantly of which the core was cut in half and the cal iper method chromite and pyroxene, hence the Archimedes water used to determine volume. This yielded similar results to displacement method was deemed suitable. Given the various immersion methods (Table 4). that the chromitite seams in this particular example were unweathered, non-porous and competent, a set of Fol owing the initial tests on control sample SG6, a mil ed samples was also analysed by gas pycnometer range of pyroxenites and friable chromitites were as a check; this data set demonstrated acceptable tested, which il ustrated that densities were general y correlation between methods. within to 3% of the cal iper method. The significantly lower DBD of weathered material (e.g. sample SG2) Scatterplot comparing highlighted the need to test density across a range of SG (pycnometer) weathering domains within a mineral deposit. with DBD (Archimedes method) for chromitite and chromiferous It was concluded that for competent, non-porous core pyroxenite. samples at the chromite mine the fol owing methods Source: MTI were suitable: i) cal iper; i ) water immersion and i i) gas pycnometer, while porous core sample densities are best measured using: i) cal iper and i ) wax-coated, spray lacquered or vacuum-packed water displacement methods. Case study – bentonite in Australia Measuring the ISBD of sodium bentonite presents a whole set of chal enges related to the fact that such material absorbs water and swells; therefore A further example from Ruighoek concerns pyroxenite direct immersion in water cannot be used with much drill core from the chromitite hangingwal , which confidence. the mine planners wished to evaluate for an open pit situation. In this case the pyroxenite ranged from MTI’s sodium bentonite mine is located approximately weathered (friable and porous) to fresh (competent 350km inland of Brisbane in Queensland, eastern and non-porous) hence there were several options, Australia. The bentonite beds were deposited within including water displacement of sealed samples and a high energy fluvial/lacustrine environment of Upper the cal iper method. An un-weathered pyroxenite Jurassic to Lower Cretaceous age. Several bentonite core sample ‘SG6’ was chosen as a control and beds have been identified on the property; these range density was estimated using the cal iper and various up to ~ 4m in thickness and consist of dioctahedral water displacement methods (Table 3). The cal iper smectite (montmoril onite) with accessory minerals method yielded comparable results to the Archimedes including feldspar, kaolinite, quartz and zeolite. The
Volume 58 ~ Number TWO ~ JUNE 2015 Table 3 Chromitite and pyroxenite drill core DBD estimated by various methods. Source: MTI S Method Diameter Length Volume Mass Mass Sealant Sealant Sealant Density Difference E
in air in water mass density volume vs. Cal iper L
cm cm cm3 g g g g/cm3 cm3 g/cm3
SG2 oxidised pyroxenite T Cal iper 6.3 .35 353.95 880.25
2.49 R Vacuum pack
887.85 530.75 7.6 0.9 8.26 2.52 .5% A Paraffin Wax
887.35 536.9 7. 0.9 7.89 2.57 3.3%
SG6 competent pyroxenite Cal iper 4.76 9.6 348.93 45.85
94.45 7 .85 0.4 0.9 0.44 4.24 Table 4 Fresh pyroxenite ‘half’ core BD estimated using the cal iper method. Source: MTI Sample ID Width Height Length Radius Area Volume Mass in Air Density SG 0B cm cm cm cm cm2 cm3 g g/cm3 (‘half’ core)
064. 3.35 beds are capped by volcaniclastic rocks identified RAB drill chips petrographical y as either tuff or ignimbrite in addition at the Australian to volcanogenic sandstone which is often cross-bedded. bentonite mine. Silicified fossil wood is fairly common in sandstones Source: MTI and conglomerates above the bentonite. In the case of the Australian bentonite example, all exploration dril ing was carried out by a method known as Rotary Air Blast (RAB) using a bladed bit, which results in small drill chips unsuitable for water immersion or the cal iper method. An alternative dril ing method had to be considered in order to measure ISBD and after discussion with the contractor,
mineralogical composition, degree of weathering, 10 Bentonite core trimmed for calliper moisture content and overburden thickness. A further method. The core benefit of reconciling actual volume and tonnes ‘shavings’ were used mined against the estimated mineral resource volume for moisture analysis. S Source: MTI and tonnes is to verify the geology model. In this E particular case the surveyed volume was within 3% of L the model ed volume, indicating that the exploration IC and model ing methods were applicable for this style T of bedded mineralisation. R the RAB rig was modified to drill core (without water) A at several strategic locations. On reclaiming the cores, Another example from the Australian mine addresses all samples were sealed in plastic bags to retain in- the estimation of bulk density of sun-dried (granular) situ moisture before estimating density. The core bentonite stockpiles. As with surveying the volume of samples were then trimmed with a hacksaw to yield bentonite mined from a pit, an option for stockpiles is regular cylindrical shapes from which volumes could to measure the stockpile before and after shipment and be estimated using the cal iper method, and moisture estimate the volume removed. An alternative is to extract content derived from the ‘shavings’. Density values of some material from the stockpile and fill a container of between .72 and .84 t/m3 were obtained (Table 5) known volume, which can then be weighed. This latter and it was elected to use .8 t/m3 for estimation of in- procedure was adopted at the Australian bentonite situ ‘wet’ bentonite resources. Table 5 Bentonite ISBD estimated using the cal iper method. Source: MTI Bentonite Length Diameter Volume Mass Moisture ISBD cm cm cm3 g % tonnes
/m3 5D 4.6 6.4 48.0 254.6 27.3 1.72 5D 2.3 6.4 74.0 35.9 27.0 1.84 Once a mine is in operation, it is advisable to verify densities that were estimated during the exploration ‘Dried and crushed’ bentonite about to be tipped into cubic metre phase of the project. This can be achieved by surveying box. Source: MTI (above and below) the volume of an excavated void, for example an extracted bentonite seam in an opencast pit and using this in conjunction with truckloads of mined material measured on a weighbridge (Table 6). This procedure was adopted at the Australian mine and verified that an ISBD range of .74 to .8 t/m3 is probably applicable to this type of bentonite (~27% moisture; ~ 80% montmoril onite). It is to be expected that ISBD would vary across such a deposit according to Table 6 Bentonite ISBD estimated from a surveyed open pit, Australia. Source: MTI Description Bentonite Tonnes hauled Volume m3 Volume m3 Difference ISBD (over weighbridge) (surveyed) (geology model) in volume (calculated t/m3) NP Block 3 5D 0,0 7 5,773 5,947 3% 1.74
Volume 58 ~ Number TWO ~ JUNE 2015 Table 7 Bentonite stockpile BD measured using a cubic metre box. Source: MTI S E Bentonite Location Untapped Mass Tapped Mass Volume Moisture Density Untapped Density Tapped L (tonnes) (tonnes) (m3) (%) t/m3 t/m3 IC 5B OP Dry Stockpile .253 .4 0.4 .25 1.41 T 5D OPW Dry Stockpile .306 .43 .9 .3 1.43 R A mine and it was estimated from fil ing a box of one that going to an even lower 4.0 SG standard would cubic metre volume that; i) loose (untapped) density release more barite into the market, but could cause was ~ .3 t/m3 and ; i ) that tapped density is ~ .4 problems and increased costs for dril ing fluids and t/m3 (Table 7). waste management (Scogings, 20 5a). Case study - Barite The author has calculated SG for a theoretical series
of barite-quartz compositions between SG 2.7 and Given the recent trend towards the use of lower SG SG 4.5 in an attempt to quantify the effect of dilution barite for oil dril ing applications, this study evaluated by ‘silica contaminants’. In addition, a series of barite- the stoichiometric method of estimating density using quartz dilutions (by mass) were prepared and measured a series of barite-silicate blends. The American by gas pycnometer at Intertek in Perth, Australia. The Petroleum Institute standard for oil dril ing barite was pycnometer results appear to verify i) the non-linear SG 4.2 until 20 0, during that year a new lower SG relationship between whole-rock chemistry and density 4. product was accepted as an alternative standard. and; i ) that a barite product with density of 4. could Over the past few years some experts in the oil industry have as much as 23% silicate mineral by volume, rising expressed reservations that, as lower SG is related to to ~30% when SG is decreased to 4.0. This latter dilution of barite by abrasive contaminants such as value for mineral impurities at SG 4.0 is higher than ‘silica’ (quartz or chert); this would result in increased when assuming a straight-line relationship between mill wear, increased tonnage required to be mil ed chemistry and SG (Table 8). and more abrasive dril ing mud. The consensus was Table 8 Barite product: SG related to dilution by SiO2 impurities BaSO4 SiO2 SG (calculated) Barite (calculated) Silicate (calculated) SG (calculated) SG (pycnometer) (% by mass) (% by mass) (g/mL by mass) (% by volume) (% by volume) (g/mL by volume) (g/mL lab. blend) 00 0 4.50 00 0 4.50 4.50 95 5 4.4 92 8 4.35 4.38 90 0 4.32 84 6 4.22 4.22 85 5 4.23 77 23 4.09 4.04 80 20 4. 4 7 29 3.97 4.0 75 25 4.05 64 36 3.86 3.78 50 50 3.60 38 63 3.38 3.46 25 75 3. 5 7 83 3.00 3.02 0 00 2.70 0 00 2.70 2.7
• The method/s chosen should take into account physical and chemical variations across the deposit such as weathering, porosity, mineralogy and moisture content S E • QAQC methods commonly applied to other factors L in an exploration program should also apply to IC density measurements. T R Acknowledgements A The author sincerely thanks Mineral Technologies Inc. for permission to use exploration data from its mines in Australia and South Africa. The assistance of Frazer Fal ens and Ann Evers at Intertek Minerals Australia in Perth is grateful y acknowledged, as is the support provided by CSA Global Pty Ltd. Stoichiometric density Conclusions estimates for barite- References quartz blends. Source: Industrial Minerals • Mineral resource estimations rely on three main Research, Intertek Perth inputs: i) grade, i ) volume and i i) bulk density, of Lipton, I.T. and Horton, J.A., 20 4. Measurement of bulk which the latter is often relatively neglected during density for resource estimation – methods, guidelines mineral exploration and quality control. Mineral Resource and Ore Reserve Estimation – The AusIMM Guide to Good Practice. 2nd • The SAMREC 2009 code requires that the methods edition, Monograph 30. and assumptions of estimating bulk density be described when reporting Mineral Resources and SAMREC, 2009. The South African code for the Reserves reporting of exploration results, mineral resources and mineral reserves. Prepared by the South African • Poor quality bulk density measurements result in Mineral Resource Committee (SAMREC) Working unreliable tonnage estimates and impact negatively Group. www.samcode.co.za on mine scheduling and reconciliation of mineral production against reserves Scogings, A.J. (20 5a). Dril ing grade barite. Supply, Demand & Market. Industrial Minerals Research, • Determination of sample mass is the ‘easy part’ of January 20 5. 226 pp. Industrial Minerals, Euromoney estimating density. The difficult step general y lies in Institutional Investor PLC, Nestor House, Playhouse trying to determine the volume of a sample Yard, London EC4V 5EX, United Kingdom. • There are several methods for estimating the volume Scogings, A. J. (20 5b). Bulk Density: neglected but of rocks and materials, each of which has practical essential. Industrial Minerals Magazine, April 20 5, limitations and it is suggested that more than one 60-62. Industrial Minerals, Euromoney Institutional method be used, as an internal check Investor PLC, Nestor House, Playhouse Yard, London EC4V 5EX, United Kingdom. • The use of ‘cling wrap’ film to seal samples should be avoided, as entrapped air can lead to significantly low density results compared with other methods Andrew Scogings