Assessing the diamond potential of kimberlites from discovery to evaluation bulk sampling

Please download to get full document.

View again

of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 0 | Pages: 9

Extension: PDF | Download: 0

Related documents
Mineralium Deposita (2003) 38: DOI /s ARTICLE Luc Rombouts Assessing the diamond potential of kimberlites from discovery to evaluation bulk sampling Received: 19 February
Mineralium Deposita (2003) 38: DOI /s ARTICLE Luc Rombouts Assessing the diamond potential of kimberlites from discovery to evaluation bulk sampling Received: 19 February 2002 / Accepted: 19 June 2002 / Published online: 3 September 2002 Ó Springer-Verlag 2002 Abstract The economic evaluation of diamond-bearing kimberlites is usually carried out in four stages. Expenditure tends to increase by an order of magnitude at each successive stage. At the end of each stage, the sample results should be critically appraised before deciding to proceed to the next phase. In the first stage, even before individual kimberlite bodies have been discovered, the indicator mineral geochemistry will give a first rough idea of the diamond potential. The relative abundance of harzburgitic pyropes (subcalcic chrome-rich) is often directly correlated with the diamond grade. In the next stage, when the kimberlite body has been discovered, a relatively small sample of a few hundred kilograms will be enough to recover sufficient microdiamonds to allow an extrapolation of the size distribution towards the commercial-sized diamonds and a rough estimate of their grade. If positive, the third stage should be a limited bulk sampling programme (order of 200 tonnes) to determine the commercial-sized diamond grade, expressed as carats per tonne. The aim of the final stage is to obtain a parcel of the order of 1,000 carats to estimate the average commercial value of the diamonds. The robustness and reliability of the grade and value estimates can be verified with extreme value analysis and by obtaining the confidence limits with bootstrapping. Keywords Diamond Æ Evaluation Æ Extreme value analysis Æ Microdiamond Æ Size distribution The economic context The value of rough uncut natural diamond production is of the order of US$7 billion per year. With existing mines continuously being depleted, the incentive for Editorial handling: B. Lehmann L. Rombouts Terraconsult, Oosterveldlaan 273, Mortsel, Belgium finding new diamond deposits is strong. In recent years, between US$350 to 500 million is spent annually on diamond exploration worldwide. Mining or exploration companies often spend US$1 to 10 million each on specific projects. The main diamond exploration target is kimberlite. As a rough appreciation, one could state that, worldwide, only 10% of the kimberlites are diamond-bearing and only 1% contain diamonds in economic quantities. Kimberlite provinces may contain 100 bodies within a km area, with only a few of them having economic diamond grades. The cost of testing the diamond content of kimberlites may run into the millions of dollars. With so many potential targets to test, it is imperative to find efficient means to discriminate as early as possible between barren or poor-grade kimberlites and kimberlites with true diamond potential. This note discusses the optimal techniques to assess the diamond potential at each stage in the evaluation process and to prioritize kimberlite targets for further costly testing. The discussion on the evaluation techniques is valid for any volcanic source rock of diamonds. While for convenience the term kimberlite is used, the techniques are equally applicable to lamproites and other lamprophyric diamond-bearing rocks. Exploration stage: the use of indicator mineral chemistry Kimberlite is the dominant primary source rock of diamonds at the Earth s surface. Diamonds are formed at depths of more than 150 km (more than 50 kbar lithostatic pressure) in a relatively cool environment (950 to 1,250 C). The Earth s mantle at depths of 150 km or more is, however, in most places too hot to have diamonds crystallized out from carbon. Temperatures above 1,250 C at 150 km result in graphite being formed. The Archean cratons, not disturbed since 2.5 billion years by any major thermal tectonic event, have deep and cool lithospheric keels surrounded and underlain at depth by the hotter asthenosphere. The 497 Archean cratons tend, therefore, to have low geothermal gradients, resulting in the right pressure and temperature conditions for diamond formation at depths of 150 km or more. Kimberlite volcanism, originating from the base of the lithosphere, can bring the diamonds in a sudden explosive event to the surface. Kimberlite volcanism is a rare event, seemingly restricted to well-defined time periods. An important worldwide event occurred for instance about 1,100 million years ago (Premier in South Africa, Madhya and Andra Pradesh kimberlites in India, Mali, Argyle lamproite in Australia); another example are the Cretaceous kimberlites of West Africa, Southern Africa and Central Africa. The kimberlite event may be related to plumes rising from great depths in the lower mantle into the base of the lithosphere. Highly volatile magmas are generated, and the resulting kimberlite volcanism brings not only diamonds, but many other, much more common minerals from great depths to the surface. At depths of 150 km, the mantle, or base of the lithosphere, is mainly composed of lherzolite, a rock containing olivine, clinopyroxene and orthopyroxene. CO 2, H 2 O and other volatiles, generated at greater depths and rising through the base of the lithosphere, react with the lherzolite, depleting it in calcium. The resulting metasomatized lherzolite has a harzburgite composition, containing mainly olivine and orthopyroxenes. A common accessory mineral of both lherzolite and harzburgite is pyrope garnet. Pyropes, being much more abundant than diamonds, are more easy to recognize in stream sediments and other erosion products of kimberlites. The sampling of sediments for pyrope garnets, therefore, is an efficient exploration tool for kimberlites. The peridotitic mantle pyropes are rich in chrome and magnesium. The harzburgitic pyropes are, just as their harzburgite mother rock, depleted in calcium, relative to the lherzolitic pyropes (Sobolev 1977). The lherzolitic pyropes show a positive correlation between their chrome and calcium content: both element abundances increasing with depth (Gurney 1984). Other useful indicator minerals of kimberlite are eclogitic garnets with Na 2 O contents higher than 0.06 wt% (Gurney 1984), magnesium-rich ilmenite, chrome-diopside and chromite. The chemical composition of the indicator minerals can be used as an early evaluation tool of the diamond potential of kimberlites, even before they are actually discovered. Sobolev (1977) in Russia, and Gurney (1984) in Southern Africa, found a positive correlation between the relative abundance of harzburgitic pyropes in kimberlites and their diamond content. The following formula can be derived from Gurney s observation that 85% of the pyrope inclusions in diamonds have a chrome-content greater than: Cr 2 O 3 ¼ 11:637 þ 3:606 CaO This linear function, displayed on a Cr 2 O 3 versus CaO scatter plot (Fig. 1), is used by Gurney to separate harzburgitic (chrome-rich, calcium-poor with good diamond potential) from lherzolitic garnets. If the Cr 2 O 3 content is less than 0.05 wt% and the CaO content less than 3.5 wt%, the garnets are not of mantle origin, but crustal, and of no relevance to diamonds. Apart from lherzolite and harzburgite, also eclogite is a common rock type at the base of the lithosphere. Eclogite, composed of clinopyroxene and pyrope almandine garnets, is a possible source rock for diamonds, which can be sampled by the ascending Fig. 1 The relative abundance of harzburgitic garnets compared with lherzolitic garnets for a kimberlite in Angola with a diamond grade of 0.12 carats/ t, as discernible on a Cr 2 O 3 versus CaO plot for peridotitic garnets 498 kimberlite magma. Eclogitic garnets with a Na 2 Ocontent greater than 0.06 wt% are considered good indicators for eclogitic diamonds (Gurney 1984). The high sodium content in garnet is considered to indicate equilibration at pressures high enough to be compatible with the presence of diamond. Alternative indicators for the diamond potential of a kimberlite or a lamproite are magnesium-rich chromites. Chromites with a Cr 2 O 3 content in the wt% range, TiO 2 below 0.6 wt% and MgO in the wt% range are common as inclusions in diamonds (Meyer and Boyd 1972; Sobolev 1977; Daniels 1991). During the ascent of the kimberlite magma to the surface, diamonds are transiting during a short period of time in an oxidizing environment at lower pressure and temperature, and may become resorbed. The oxidation state of the kimberlite magma is reflected by the composition of ilmenite. Ilmenite with MgO content below 5 wt% indicates a highly oxidizing environment with no preservation of diamonds. Ilmenite with MgO content higher than 12 wt% on the other hand indicates excellent diamond preservation conditions. The chemical analyses required on the indicator minerals can be done by electron microprobe. Therefore, they are a cheap (order of US$10 per grain) and quick method to assess the diamond potential of a kimberlite at an early stage during exploration. Griffin et al. (1989) developed a new technique based on the nickel content of the peridotitic garnets to assess the diamond potential of kimberlites and lamproites. Determination of the nickel content in garnets requires the use of a proton microprobe. Only few proton microprobes are geared for this type of analysis, and the nickel technique, therefore, is more costly and less practical. A recent and exhaustive review on the use of indicator mineral geochemistry in diamond exploration is given by Fipke et al. (1995). Discovery stage: the use of microdiamonds The chemical composition of the indicator minerals, which led to the discovery of a kimberlite body, can give a qualitative idea of the diamond potential. The method, however, is not without exceptions. Attempts have been made to make even rather precise predictions on the diamond grade based on the chemistry of the indicator minerals. The latter, however, can never quantify the degree of dilution by country rock fragments, a major factor in grade variations within a kimberlite body. For instance, it is not uncommon for a kimberlite pipe to have a diamond grade of 0.2 carats/tonne (t) in the tuffs of the crater facies, 0.8 carats/t in the diatreme breccia and 1.5 carats/t in the underlying hypabyssal kimberlite. The different facies may have a similar indicator mineral geochemistry, but with diamond grade variation largely influenced by the mechanics of kimberlite explosion and degree of mixing with barren country rock fragments. Therefore, the indicator mineral geochemistry should be combined with a parameter directly related to the amount of country rock dilution in the crater or diatreme. A useful indicator of dilution is the magnesium and chromium content of the kimberlite, which should normally be directly correlated with the pure kimberlite component of the rock. Commercial-sized uncut diamonds for the gem industry are larger than 1 mm. Smaller diamonds cannot be cut and polished at a profit. Some Russian mines recover finer diamonds, down to 0.4 mm, to be used for industrial purposes as diamond powder. Many mines recover diamonds only down to 1.5 mm, as the value content of the smaller diamonds is negligible. Kimberlite bodies with economic grades have commercial stone densities in the range of 1 to 100 stones per tonne. Assessing the grade of commercial-sized diamonds, therefore, requires large samples. Typically, a bulk sampling programme costs several US$ 100,000, if not several US$ millions. Before embarking on this costly exercise, the microdiamond content of smaller samples (on the order of a few hundred kilograms) may give a semiquantitative indication of the commercial-sized diamond grade. Microdiamond analysis allows targets to be prioritized for bulk sampling and is an important argument in justifying continuing expenditure on evaluation sampling. By dissolving the kimberlite rock in hydrofluoric acid or by caustic fusion, the microdiamonds are liberated. Microdiamonds are usually recovered down to 0.1 mm (Fig. 2). The frequency of the microdiamonds increases exponentially with decreasing size. The microdiamond stone density in economic kimberlite bodies is often in the range of 1 to 10 stones per kg. Samples of a few tens of kilograms, taken in a spatially representative way from the kimberlite, combining to a total sample size of several 100 kg, may yield enough microdiamonds and diamonds in the range 0.1 to 2 mm to allow a reliable extrapolation of the size distribution of the microdiamonds and to roughly estimate the grade of the commercial-sized diamonds. Fig. 2 A set of microdiamonds from the Theunissen kimberlite dyke, South Africa 499 Microdiamond recovery techniques The most common technique for recovering microdiamonds is by caustic fusion of the kimberlite core or rock samples, as for instance used by Lakefield Research of Canada, a commercial laboratory. Sample bags with kimberlite rock pieces, not exceeding 10 cm in dimension, possibly weighing 5 to 10 kg, are put, without any further handling except weighing, together with NaOH granules into a kiln, and heated overnight at 450 C to dissolve the kimberlite rock by caustic fusion. If the rock is rich in carbonate, the kiln charge has to be reduced by half as the carbonate reacts vigorously with the caustic soda. The residue of the caustic fusion is cooled. After cooling, the residue passes over a 0.1-mm screen (if recovery is down to 0.1 mm) and is thoroughly cleaned with warm water for a day. The caustic soda remnants are further removed by adding a small amount of HCl, if necessary. The caustic fusion and the washing on the 0.1-mm screen usually reduces the original sample from say 10 kg to a few grams. The residue is dried in an oven at 110 C on its 0.1-mm screen. After drying, the residue is screened at say 2 mm and the coarse fragments returned for inspection. If the kimberlite rock is rich in silica, the residue may contain large fragments (e.g. granite xenoliths). If the large fragments are indeed country rock xenoliths, they can be discarded. If they are silicified kimberlite (as occurs in hot climates), they may require a new caustic fusion attack. A strong hand magnet can be used to separate the magnetic minerals from the dry residue. This should be done with great care to avoid the sticking of diamonds on magnetic minerals. After this first magnetic pass, the residue may pass through a magnetic separator, with magnetic separation set at 20,000 gauss. Both the magnetic and non-magnetic fractions are sent to the observation laboratory, where trained mineral observers pick out the microdiamonds. Sometimes clear spinels or zircons may resemble diamonds. If in doubt, the grain can be studied under a scanning electron microscope, where the diamond diagnosis is done by elimination (carbon is not directly measured, but the absence of Si, Al, Mg or other common elements, points to diamond). Each sample and fraction should at least be picked twice by two independent observers. The microdiamonds can be weighed if a highprecision balance is available (required precision is mg to weigh microdiamonds with 0.1 mm diameter), else they can be measured under the microscope in their x, y, z directions and a theoretical weight attached to them. The theoretical weights can be corrected by weighing 20 microdiamonds together and by backcalculating the correction over each grain. An alternative method, such as used by the laboratory of Scientific Services in South Africa, is the chemical dissolution of the kimberlite core or rock samples with hydrofluoric acid. The kimberlite rock sample is crushed down to minus 10 mm, for instance in a roller crusher. Samples of say 10 kg are mixed with 10 l of HCl acid, diluted with 4 l of water in a large plastic container, agitated for several hours, diluted again with water and allowed to settle overnight. Next day, the sample is de-slimed over a 0.1-mm screen. The residue is mixed with about 15 l of HF acid, with the exothermic reaction raising the temperature to about 80 C, and agitated for several hours. At the end of the day the mixture is diluted with plenty of water and allowed to settle overnight and de-slimed over a 0.1- mm screen the next morning. The sample is mixed with 4 l of HF for several hours, diluted with water and de-slimed over a 0.1-mm screen. The residue is dried. The HF leach residue is mixed with NaOH grains, the temperature is raised so as to melt the NaOH granules (caustic fusion at C) and, as a result, the Ca- and Mg-fluoride grains and coatings are dissolved. After the caustic fusion, the residue is washed and screened on a 0.1-mm screen, the same screen as used before (the screen will be discarded after treating each sample to avoid contamination). The caustic fusion residue is mixed with HF, if mainly silicates remain, and heated to boiling temperature for 4 h. If the remaining material consists of mainly oxides, the third HF attack may be replaced by a HCl attack. The residue is screened and dried in a nickel crucible. Depending on the results of the three acid attacks, the residue may be treated for a second time by caustic fusion if many Ca or Mg fluorides are present. The next step is to fuse ( C) bifluoride powder (KHF 2 ) with the residue for about 30 min to dissolve the most refractory silicates such as zircons, to let the mixture cool, dissolve in water and screen at 0.1 mm. Final clean-up of the remaining grains is done by boiling in aqua regia for 2 h. After dilution with water and screening, the concentrate (usually 1 g or less from an initial 10 kg sample) is dried and sent for picking by the mineral observers, using similar techniques as described for the caustic fusion method. The cost of recovering the microdiamonds by caustic fusion or by hydrofluoric acid is similar and of the order of US$350 per 10 kg of kimberlite. Other methods are sometimes used to liberate and recover microdiamonds from the kimberlite rock. Attrition milling in stages can liberate finer and finer diamonds without breaking them. The fine diamonds can be recovered from the kimberlite gangue by dense media separation and by picking under the binocular microscope. Microdiamond size distribution Kimberlites with economic diamond grades contain between 1 to 10 microdiamonds (larger than 0.1 mm) per kg. Microdiamond counts can be misleading though. Several examples are known of kimberlites or lamproites with high microdiamond counts (order of 1 per kg), but with uneconomic macrodiamond (larger than 1 mm) grades. In those cases, the size distribution of the microdiamonds shows that the frequency of stone occurrences drops at such a rate with size that, by the time one reaches the macrodiamond sizes, the frequency of occurrence has dropped to uneconomic grades. At the smallest sizes recovered, close to the 0.1 mm bottom screen, the frequency may drop with decreasing size due to poor recovery efficiencies of the finest microdiamonds. The finest microdiamond tend to float in water and may be lost during the treatment process; they are also harder to recognize under the binocular microscope. The drop in frequency at the smallest sizes and the exponential decrease in frequency with increasing size result in a microdiamond size distribution with a similar skewness as the lognormal distribution. The logarithmic variances of the microdiamond size distribution are high: they can vary from 3 to 9. Due to their lognormal-like appearance, the microdiamond size distributions are often nearly linear on a cumulative lognormal graph. Deviation from the 2-parameter lognormal model can often be corrected by adding a third parameter s. The lognormal frequency density distribution is: 1 f ðxþ ¼pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi exp 1 # logðx þ sþ f 2 2prðx þ sþ 2 r with x the stone size in carats, r the logarithmic standard deviation and n the logarithmic mean. Grade prediction based on mic
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!