ultra fine grinding technology
This coal mining project is an open pit mine located in Nigeria, announced by mining company - Western Goldfields - that it has discovered 62,400,000 tonnes of proven reserves of coal deposits worth US$1.2 billion which could be used for the generation of electric power...
minerals engineering conferences - ultrafine grinding 06
What is ultrafine grinding and what role can ultrafine grinding play in the minerals industry now or in the future? This theme attracted representatives from equipment manufacturers, suppliers of grinding media, mining companies, consultants and academics to a conference in the seaside resort of Falmouth in Cornwall, UK (12-13th June 2006). The conference was attended by 64 delegates from 24 countries spread across all five continents.
While the concept of grinding minerals hardly requires an introduction, it emerged that the definition of ultrafine particles depends on the application: it can relate both to particles in the low m size range and to particles which have been considerably reduced in size. Traditional barriers to ultrafine grinding are the relatively high energy requirement and a perceived reduction in efficiency during further processing. However, technological advances in ultrafine grinding may be welcomed by the mining sector, where the necessity to process increasingly complex ores is presently coupled to buoyant commodity prices.
While conventional milling techniques may produce an ultrafine grind, the required energy rises sharply as the product size, often characterized in terms of the 80 % passing size, d80, decreases. Ultrafine grinding techniques are those techniques which are more energy-efficient than conventional milling techniques in the sub 100 m range. Other advantages over conventional techniques are the option to avoid contamination of the product with ferrous chips, for example from steel balls used as grinding media, lower wear for ultrafine applications, the ability to contain dust and manage heat generation. On the downside, D. Yan states that ultrafine grinding techniques are less predictable than conventional ball mills in terms of energy requirement and grinding performance while throughput may be limited.
During ultrafine grinding, particle breakage occurs by familiar mechanisms: impact or attrition by shear or a combination of these. Classic examples of impact techniques are jetmills where particles are either accelerated against a surface or against other particles moving in the opposite direction. While impact techniques are not suitable for breaking relatively tough materials such as quartz, modified jetmills may be useful for some tough applications. J. Roth (PMT-Jetmill) presented the spiral jetmill as an alternative to classic jetmill designs. The feed is injected into the milling chamber through tangential nozzles while the product is recovered through the classifier rotor located in the middle of the milling chamber. Intense shear between the particles in the milling chamber can produce delamination, preserving or even increasing the aspect ratio of particles. This is beneficial, for example, when talc is ground to enable separation of quartz prior to its application as a reinforcing filler in plastics. In another example, B.G. Kim added that it is attractive to preserve the flaky shape of graphite particles during ultrafine grinding in view of their high conductance.
A. Mc Veigh (Hicom International) discussed application of the Hicom mill for enhancing the recovery of kaolin from china clay waste minerals (silica and mica). Kaolin is commonly used a filler in paper, with platy kaolin enabling the production of lighter and smoother paper. In the Hicom mill, grinding is induced by the inability of the feed to follow the nutating motion of the milling chamber. Product particles escape through openings in the grinding chamber. The Hicom mill can be operated with or without the addition of grinding media. Using relatively fine grinding media, the Hicom mill achieved better delamination than a sand mill currently used by a china clay producer.
Evidence of high-intensity grinding in a Hicom mill was also observed by M. Thornhill, who ground potassium feldspar concentrates with a view to increasing the availability of potassium and assessing their potential as a fertilizer. Trials suggested that ultrafine grinding irreversibly increased the leachability of potassium, although a lower limit to the particle size was observed: very fine particles tended to agglomerate, reducing the specific surface area and the associated leachability of potassium. Comparison with ultrafine grinding of a potassium-rich nepheline syenite concentrate suggest that the effect of mechanical activation achieved by ultrafine grinding is mineral-specific. Although mechanical activation is a difficult concept, E. van der Ven pointed out that ultrafine metal powders can be hazardous on account of their extreme reactivity.
A. Mizitov (Oy Microworld) described the grinding action in a uniRim mill as the equivalent of a pestle-and-mortar. Grinding in the uniRim occurs between the cooled mantle of the grinding vessel and grinding blocks attached to a central rotor. Grinding also occurs well away from the rotor in a planetary mill, which consists of independently-rotating canisters located at the end of arms extending from a central rotor. The combination of centrifugal acceleration of particles towards the rim of the canisters and the even higher angular velocity of the canisters creates a high-intensity grinding environment. E. Kuznetzov (Cyclotec) provided a video to suggest that previous issues such as heat and dust generation, mechanical integrity and discontinuous discharge had been resolved. Scale-up to 100 t/hr is expected to be feasible.
M.K. Abd El-Rahman (CMRDI) studied ultrafine grinding of a variety of minerals in a planetary mill as a function of the rotation speed of the canisters and the presence and size of grinding media in the canister. The advantage of higher rotation speed levelled off at higher speeds, possibly due to cake formation on the canister mantle, while degradation of grinding media increased. Relatively fine grinding media initially led to reduced particle size reduction but, for prolonged grinding, enabled attainment of a smaller product size. This is thought to be due to the smaller interstitial space between media particles, which are generally much larger than the size of product particles. Perhaps surprisingly, a relatively low grinding media-to-feed ratio appeared to increase the initial rate of particle size reduction. For further improvement of grinding performance, the use of chemical additives acting as dispersants was deemed promising.
While stirred media mills stem from a fairly basic design, they have been subject to extensive further development and refinement. For example, the MaxxMill developed by Maschinenfabrik Gustav Eirich consists of a vertical rotating vessel which contains one or more eccentrically-positioned rotors and flow deflectors. Because the MaxxMill operates in dry mode, the smallest particles can be sucked out and classified in an auxilliary cyclone. Particles above a maximum tolerated size are returned to the mill, enabling control of the product size. S. Gerl (Gustav Eirich) presented options to integrate the MaxxMill into processes.
Maelgwyn Minerals Services' M. Battersby introduced the Deswik Turbomicronizer (TM) Mill, a high-speed stirred media mill with a series of impellers along the central shaft. Having evolved from a horizontal design, the vertical Deswik TM Mill is less sensitive to failures of bearing seals and blocking of screens. Feed material is introduced as a slurry at the bottom of the mill, moving upwards in a helix-type flow pattern along a hard-wearing polyurethane resin mill lining. Grinding media are recycled internally to the base of the mill.
Given the technological advances, the focus is turning on the reliability and availability of ultrafine grinding technology in a mining environment. Stirred media mills are starting to feature in large-scale commercial mining applications. Following K. Barns' introduction of the five key process technologies marketed by Xstrata Technology, the application of IsaMill for ultrafine grinding was discussed by D. Curry. Lead/zinc ore at the Mc Arthur River mine in Northern Territories, Australia, requires grinding the ore to below 7 ?m in order to achieve sufficient liberation of galena and sphalerite. Compared to conventional grinding, ultrafine grinding was reported to increase the zinc recovery by 10 %. The IsaMill, which has been developed in partnership with Netzsch Feinmahltechnik, is a high-intensity (up to 300 kW/m3) horizontal disk mill with 8 grinding chambers in series. While media are present to enhance the grinding process, media is retained in the mill without the use of screens. The product is reported to have a narrow particle size distribution, which is considered to be a key factor in maintaining efficient downstream flotation. Using d80 as a parameter, IsaMill may be scaled up, as witnessed by a 3 MW unit being commissioned for AngloPlatinum at Potgietersrust, South Africa.
Anglo Platinum already operates a 2.6 MW unit at Rustenburg, South Africa, for tailings re-treatment, aiming to recover Platinum Group Metals (PGM). C. Rule (Anglo Platinum) explained that ultimate recovery of PGMs locked up in silicates depends on enhanced liberation brought about by ultrafine grinding. A notable improvement in PGM recovery is observed when grinding more tailings below 75 ?m. Anglo Platimum's experience with IsaMill suggests that the wear of mill lining and media consumption were lower than predicted for this application and that the product size distribution and energy consumption were within specification.
The generally advanced nature of stirred media mills was confirmed by the laboratory-scale investigation of two ultrafine grinding techniques by J. Parry. Aiming to liberate fine galena in coarse sphalerite present in ore from Red Dog mine, both the IsaMill and the Stirred Media Detritor (from Metso Minerals) achieved comparable results, as measured with the Mineral Liberation Analyser (from JKTech).
Besides developments in the design and scale-up of stirred media mills, a significant technological breakthrough was achieved with the introduction of a new class of grinding media. Previously, typical grinding media, such as the ore itself, slag, silica sand or river pebbles, suffered from lack of quality, notably inconsistent size and competence. According to P. Hassall (St Gobain-Zirpro), the advent of manufactured media lead to substantial improvement of the media quality through a uniform chemical composition and a high hardness, sphericity, roundness, density, and competency. As a result, the grinding efficiency increases and the energy consumption is lowered. The modern media are typically manufactured ceramic materials such as aluminium oxide, yttrium-, cerium- and magnesium-zirconia oxides and various silicates. B. Clermont (Magotteaux International) described that Keramax MT1 media, containing a mixture of oxides, lowered the energy consumption because there was less sliding friction between hard, spherical media particles. G.A. Graves (Zircoa) emphasised that the smooth surface of media particles, a narrow media size distribution, a high hardness and fracture toughness are vital to achieve energy consumption reduction. Image analysis may be used to monitor the shape and size of media particles. It should be noted that determination of the optimum size of media particles for a given application requires careful consideration. While an application may require media with various media particle sizes, there may be a case for staged milling and ultrafine grinding with optimized media sizes.
While ultrafine grinding can improve recovery and reduce downstream reagent requirements, the effect of extra particle size reduction should be balanced by the cost of additional grinding energy. For example, the Deswik TM Mill reduced a feed with d90 of 89 m to 20 m with 10 kWh/t or 12 m with 16 kWh/t. Ultimately, the economics will decide whether the application of ultrafine grinding will continue to grow.
ultrafine grinding - an overview | sciencedirect topics
UFG of pyrite concentrates for subsequent leaching is used in ores, where refractoriness to direct cyanidation arises from fine to ultrafine (<20, >0.02m) gold mineral inclusions in the pyrite and/or arsenopyrite. By grinding to 80% passing 10mm a significant fraction of the colloidal size (<0.5m) gold is also being exposed and rendered amenable to cyanidation. On the downside, the huge increase in surface area of pyrite that is created by UFG magnifies 10-fold any preg-borrowing effects, probably assisted by free cyanide, consumption by adsorption onto pyrite surfaces and the formation of thiocyanate (SCN). If there is carbonaceous matter in the UFG concentrate to be leached, it can contribute to significant losses, because of its relatively huge surface and the irreversible sorption of gold (preg-robbing). More details on this technique may be found in Chapter 17.
Ultrafine grinding (UFG) has continued to evolve in terms of equipment development. A number of specialist machines are commercially available including Xstrata's IsaMill, Metso's Vertimill, Outotec's High Intensity Grinding (HIG) mill, and the Metprotech mill. UFG equipment has been developed with installed powers of up to 5MW.
Compared with conventional ball or pebble milling, the specialist machines are significantly more energy efficient and can economically grind to 10m or lower, whereas the economical limit on conventional regrind mills was generally considered to be around 30m. Coupled with improvements in downstream flotation and oxidation processes, the rise of UFG has enabled treatment of more finely grained refractory ores due to a higher degree of liberation in the case of flotation or enhanced oxidation due to the generation of higher surface areas.
In 1993, the Salsigne Gold Mine was reopened. Salsigne treated a gold-bearing pyrite/arsenopyrite ore by flotation, with the flotation tails treated in a CIL circuit and the concentrate reground in a conventional mill to approximately 2530m. The oxygen demand for reground concentrate was high and the rate of oxidation was slow. The concentrate was initially oxidized for approximately 6h using oxygen injection via a Filblast aerator before cyanidation. Additional oxygen was added in in the second CIL stage and hydrogen peroxide was added into the fourth unit to maintain dissolved oxygen concentrations of >10ppm.
Goldcorp have commenced operations at the Elenore Gold Project in Quebec, Canada. The mineralogy of the ore and hence the circuit selection show similarities to those at Salsigne. The main sulfides are arsenopyrite, pyrite, and pyrhottite. The ore is floated, with the flotation tails passing to a tails CIL circuit and the flotation concentrate reground before passing to the concentrate CIL circuit via preaeration tanks designed to achieve 18-h contact with oxygen. The main difference between the Salsigne and Elenore projects is that the Elenore concentrate is ground to 10m and oxidation of the sulfides is substantially complete before cyanidation.
Ultrafine grinding is used to liberate gold finely disseminated in metallic sulfides. KCGM is the first gold mine using ultrafine grinding followed by cyanidation (Ellis and Gao, 2002). The gold sulfide concentrate is grind with an IsaMill to a P80 of 1012m. In the first 3years of operation, high consumption of cyanide and high gold content in leach residues were experienced with difficulty overcoming these issues (Deschnes etal., 2005).
Eleonore Mine is the first commercial application of the IsaMill in Canada (Deschnes and Fulton, 2013). Ultrafine grinding is applied to liberate gold finally disseminated in sulfide minerals. The pyrrhotite concentrate produced by flotation was used to determine the leaching strategy at the laboratory scale. The concentrate grind at a P80 of 10m contained 65% gangue minerals, 23% pyrrhotite, 2.2% pyrite, 9.6% arsenopyrite, 75.5g/t Au, and 5.0g/t Ag. Gold was present as native gold and electrum. A test conducted at 2000ppm NaCN and pH 11.0 produced 97.0% extraction of gold in 72h. It was found that an efficient leaching required only 35ppm DO. However, the presence of reactive pyrrhotite resulted in a cyanide consumption of 31.5kg/t NaCN.
Three key features were identified to optimize the process: the use of a pretreatment, the addition of oxygen and the addition of lead nitrate. A 16-h duration was the optimum retention time for the pretreatment (Figure26.15). In the conditions below, the lowest consumption of cyanide (6.0kg/t NaCN) was associated with a high extraction of gold (97.5% Au; leach residue at 1.92g/t Au). The 8-h pretreatment did not passivate the sulfides as much as the 16-h one and the cyanide consumption increased to 6.6kg/t while the gold extraction slightly decreased to 96.9% (leach residue at 2.36g/t Au). A 24-h pretreatment decreased the gold extraction and increased the cyanide consumption.
Figure26.15. Effect of duration of pretreatment on gold extraction from the Eleonore flotation concentrates. Pretreatment: 0.25L/min/kg oxygen, pH 11.0; cyanidation: 2000ppm NaCN, pH 11.0, DO 35ppm, 20C, 35% pulp density (Deschnes and Fulton, 2013).
For the addition of oxygen in the pretreatment, it was found that increasing the flow rate of oxygen addition from 0.13L/min/kg to 0.25L/min/kg reduced the cyanide consumption from 6.4kg/t to 6.0kg/t (Figure26.16), as well as the gold content of the leach residue from 2.52 to 1.92g/t Au. When the oxygen addition was increased to 0.53L/min, the gold content of the leach residue increased to 2.46, while the cyanide consumption went to 5.5kg/t NaCN (Figure26.16).
Figure26.16. Effect of oxygen flow in the pretreatment on gold extraction from the Eleonore flotation concentrate. Pretreatment: 2.00kg/t lead nitrate, pH11.0, 16h; cyanidation: 2000ppm NaCN, pH 11.0, DO 35ppm, 20C, 35% pulp density (Deschnes and Fulton, 2013).
It was found that increasing the lead nitrate addition from 1.0kg/t to 6.0kg/t in the pretreatment resulted in a decrease in the consumption of cyanide from 7.2kg/t to 1.2kg/t (Figure26.17). At 6kg/t lead nitrate, the pyrrhotite passivation reached a point where the oxygen consumed by the concentrate was significantly reduced. While the DO was usually below 0.5ppm during the entire pretreatment, at 6kg/t lead nitrate, the DO increased beyond 4ppm after 2h. According to the trend, the gold extraction obtained with 23kg/t lead nitrate appears not to be biased and would probably have to be in the range of 2.2g/t Au. At dosages above 4kg/t lead nitrate, the gold content of the leach residue was observed to slightly increase; considering the 0.1g/t variation in fire assay. The gold extraction is not overly sensitive to lead nitrate beyond 2kg/t addition.
Figure26.17. Effect of lead nitrate in the pretreatment on gold extraction from the Eleonore flotation concentrate. Pretreatment: 0.25L/min/kg oxygen, pH 11.0, 16h; cyanidation: 2000ppm NaCN, pH 11.0, DO 35ppm, 20C, 35% pulp density (Deschnes and Fulton, 2013).
With the right conditions, free cyanide concentration in leaching can be significantly reduced, as shown in Figure26.18. A reduction of free cyanide from 2000ppm to 800ppm NaCN had a minor effect on the gold content of the leach residue (which varied by 0.2g/t). For this series, the gold extraction showed an average of 97.0%. A 6.8% decrease in gold extraction occurred when the cyanide concentration was reduced to 550ppm (gold extraction of 92.2%). This is expressed by a sharp increase of the gold content of the leach residue. The cyanide consumption was 1.2kg/t NaCN for the experiment, using a dosage of 800ppm NaCN.
Figure26.18. Effect of cyanide concentration on gold extraction from the Eleonore flotation concentrate. Pretreatment: 0.25L/min/kg oxygen, pH 11.0, 16h, 5.5kg/t lead nitrate; cyanidation: 0.5kg/t lead nitrate, pH 11.0, DO 35ppm, 20C, 35% pulp density (Deschnes and Fulton, 2013).
It is critical to be aware that lead nitrate plays a major role in the composition of the solution that will be processed for cyanide destruction. Figure26.19 shows the variation of solution composition in terms of thiocyanate, iron, copper, and cyanate concentrations as a function of lead nitrate added. The main contributor to cyanide consumption is the formation of thiocyanate, which decreased from 2877 to 843mg/L. The iron concentration also decreased significantly (from 156 to 27mg/L). Because the iron cyanide is a very strong complex, this becomes an issue when its concentration is too high in the barren solution. Difficulties will be encountered in striving to meet the criteria for total cyanide after destruction. The dispersion of thiocyanate concentration indicates some variation in the surface conditions of pyrrhotite prior to cyanidation. In this example, about 50% of the cyanide was unaccounted for. The unaccounted cyanide consumed is probably related to the precipitation of iron cyanide hydroxide species.
Figure26.19. Variation of CNS, Fe, Cu, and CNO concentrations as a function of lead nitrate addition during cyanidation of Eleonore flotation concentrate. Pretreatment: 0.25L/min/kg oxygen, pH 11.0, 16h; cyanidation: 2000ppm NaCN, pH 11.0, DO 35ppm, 20C, 35% pulp density (Deschnes and Fulton, 2013).
Optimization of the plant by the project owner continues after plant commissioning with the aim of maximizing plant throughput within the limitations of ore supply and the maximum capacity of the high capital cost unit processes. This is typically the comminution circuit or downstream concentrate treatment process (e.g.,POX or bio-oxidation plant) for refractory gold ores. Recovery and operating costs are other targets for optimization.
The Macraes Gold Project has treated sulfide ore, oxide ore (in campaigns) and retreated some tailings. The expansion in the throughput of the Macraes Gold Project since plant commissioning in 1990 is illustrated in Figure11.3. The periodic high treatment rates for oxide ore represent periods when the main grinding circuit was used to process oxide ore, during 1991, just prior to plant upgrade in 1999, and twice to process stockpiled oxide ore, in 2001 and 2003. The low rate of continuous treatment of oxide ore through a new mill can be seen post May2003.
In late 1994, the plant was expanded by the addition of flotation and fine grinding capacity. The primary grind size was increased and ultra-fine grinding of flotation concentrate installed to improve gold recovery from flotation concentrate in the subsequent CIL circuit.
In late 1999, a further expansion of the plant was undertaken with the addition of a ball mill (ML350) to reduce the primary grind size and allow a further increase in throughput. In addition, a POX plant was installed to improve refractory gold recovery through oxidation of the flotation concentrate.
In late 2001, a retreatment flotation circuit was installed to recover gold from old sulfide tailings and in mid-2003, a further single-stage SAG mill was installed to allow for the parallel treatment of oxide ore or additional sulfideore.
Each plant expansion was followed by a steady increase in plant throughput. The only element of the plant that became redundant over this period was the ultra-fine regrind facility that was shut down in 1995 due to higher-than-expected operating costs.
The term refractory ore is used to classify different gold ores that are not amenable to the traditional cyanidation processes. Gold in refractory sulfide ores occurs as fine inclusions or in solid solution typically within pyrite, marcasite, and arsenopyrite grains. In this type of ore, because the gold is encapsulated, the interaction with cyanide to form the soluble metal complex is inhibited (Marsden and House, 2006; La Brooy etal., 1994).
Grinding has been used to enhance gold liberation from the host sulfide-containing mineral or for particle-size optimization before oxidative pretreatments; however, when gold is encapsulated in a sulfide matrix, conventional grinding (P80=75m) without a pretreatment usually does not result in improved gold recoveries in the leaching process. Ultra-fine grinding (P8010m) has been proposed as an alternative to unlock gold from sulfidic refractory ores; however, the cost of size reduction to this level makes it applicable only for concentrates (Corrans and Angove, 1991; Gonzalez-Anaya etal., 2011), as is practiced at Kalgoorlie Consolidated Gold Mines, a joint venture between Barrick and Newmont.
Pressure oxidation is one of the main methods used in the industry to improve gold recovery from sulfidic ores. During the autoclaving process, iron sulfides present in the ore are oxidized to ferric sulfate or other oxidized solid compounds such as hematite, thus liberating the gold particles and making them available for leaching. Roasting is also widely used as an oxidative pretreatment for sulfidic refractory ores, in which iron sulfides are oxidized to hematite by oxygen at high temperatures; the porous characteristics of hematite allow for penetration of the cyanide leach solution. Another known method for gold liberation from the sulfide matrix is biological mineral oxidation. In this process, bacteria act as a catalyst for the conversion of iron sulfides in the ore to soluble ferric iron (Fraser etal., 1991). These refractory oxidation processes may be carried out on either a sulfide concentrate or whole-ore feed.
The presence of carbonaceous matter in the ore has also been identified as a cause for the refractory behavior of some gold ores. Like activated carbon, the carbonaceous matter can adsorb the gold cyanide complex, thus reducing the recovery of gold in a cyanide-based process. During the leaching step, once the gold-cyanide complex is formed, it is readily adsorbed by the carbonaceous matter in the ore, thus reporting to the tailings instead of the pregnant solution or activated carbon. The carbon components of the ore responsible for this preg-robbing effect have been identified mainly as native or organic carbon (or noncarbonate carbon), often referred to as total carbonaceous matter (TCM). Encapsulation of gold by carbonaceous ores has also been associated with lower gold recoveries from these types of ores (Dunne etal., 2013) (see also Chapter 49).
The use of adsorbent media such as activated carbon (in carbon-in-leach (CIL)) or resin (in resin-in-leach (RIL)), to compete with the TCM in the ore for gold adsorption from the cyanide leach solution, allows for acceptable gold recoveries from ores with mild preg-robbing characteristics. In the presence of ores with higher TCM content, surface modification or oxidative pretreatments are required (Afenya, 1991).
Traditional methods for pretreatment of carbonaceous ores include flotation, the use of blinding agents, roasting, and chlorination. Flotation has been used to remove carbonaceous material from ores; however, this method is applicable if only most of the gold particles are not associated with the carbonaceous components (Dunne etal., 2013). Blinding is based on selective adsorption of certain chemicals on the carbon surface. Diesel oils, kerosene, paraffin wax, and different surfactants have been used to coat the carbon surface effectively, thus reducing the preg-robbing characteristics of the ore (Zhou etal., 2013). Successful examples of this practice include the Stawell gold mine in Australia (CIL) and the Penjom mine in Malaysia (RIL).
Roasting is the most suitable process to ensure the most complete oxidation of carbonaceous material; high temperature (500600C) and oxygen or air are used to convert carbon to carbon dioxide. Gaseous chlorine (Cl2), hypochlorite (OCl), and ozone (O3) may be used to deactivate the surface of the carbonaceous material in the ore. The mechanism of carbon deactivation is not well understood; it has been suggested that upon exposure to chlorine, the carbon surface is modified, forming carboxyl-type groups that passivate the adsorption sites or change the surface charge, thereby repelling the gold-cyanide ions (see Chapter 49).
Ores in which gold is associated with sulfides and with high carbonaceous matter content are considered double refractory. The northern Nevada region in the United States is a main area where this type of ore is found. Ores in the Carlin Trend District are characterized by the presence of submicroscopic gold (invisible gold) finely disseminated within a sulfide-rich matrix (mainly pyrite, arsenian pyrite, marcasite, and arsenopyrite) within carbonaceous material. Sulfide concentrations vary from 0.5% to 3.5% whereas the carbonaceous matter content can range from 0.5% to 4% total organic carbon (Zhou, 2013). Figure50.1 shows typical gold deportment of two double-refractory ores (a and b) compared with a nonrefractory oxide ore (c). The main gold carrier in the ore shown in Figure50.1(a) is pyrite; the ore has a total sulfide content of 1.43% and a TCM content of 0.94%. In the ore shown in Figure50.1(b), gold is present as native gold, associated with pyrite and arsenopyrite and also as surface gold or gold adsorbed onto the TCM surface. This ore has lower TCM content (0.58%); however, a larger portion of the gold is present as surface gold. This is just a small illustration of the variability of ore components and gold carriers, key factors on selecting the processing method for a given ore.
The elution or stripping properties of a resin is as important as its loading performance. The functional groups on resins have different properties, including the strength of the ionic bond; elution methods therefore have to be developed and optimized for each resin type.
Recognized methods of elution of strong-base resins use zinc cyanide, ammonium thiocyanate, or acidic thiourea. Although each method has advantages and disadvantages (Fleming and Cromberge, 1984), no single method seems to be appropriate for all cases. For example, Fleming (1989) used the zinc-cyanide method for a conventional strong-base resin (A161RIP) used at the Golden Jubilee RIP plant, while for a more selective resin such as Minix, the thiourea method is more effective. Advantages of this stripping procedure are that no regeneration of the resin is necessary and elution is faster, as illustrated in Figure32.2.
The most effective elution technique for Minix makes use of acidic thiourea after an acid wash (1M) to remove nickel, zinc, and some copper. The gold-selective adsorption characteristics of Minix, particularly its high selectivity against cobalt and iron, make it possible to use its advantages (fast kinetics and circuit simplicity) without fear of poisoning the resin (Fleming and Hancock, 1979), as is the case with other commercially available strong-base resins. Elution can be improved by an increase in the thiourea concentration, the acid concentration or the temperature. With an eluant containing 1M thiourea and 0.5MH2SO4 at 60C, elution efficiencies of more than 99.6% can be effected within four or five bed volumes of eluant. Using this elution strategy, the resin is easily stripped to below 100g/t of residual gold.
Electrowinning has proved to be an effective and simple method for recovering gold from the eluate. Electrowinning is done simultaneously with the elution. The barren eluate is recycled to the eluant tank and is recycled for the next elution after the reagents have been made up to their relevant concentrations. Two operating plants, namely Penjom Gold Mine and Barbrook Gold Mine, have now demonstrated the feasibility of this elution strategy for Minix with simultaneous electrowinning.
[Ed.: Caledonia Resources announced5 in January 2006 the commissioning of Barbrook Gold Mine plant expansion, including ultrafine grinding, oxidation, and RIL circuits, and in November 2006 was placed on care and maintenance as a result of damage caused by employees of a labor brokerage company during an illegal industrial action.6 Barbrook was sold7,8 to Vantage Goldfields in 2008.]
Elution of AM-2B resin is similar to that of Minix; but a longer elution time seems to be required (Bolinsky and Shirley, 1996). The process includes a series of elutions, contacting the loaded resin with different solutions. The elution flowsheet differs from plant to plant, and some of the steps described below are optional. Usually the following steps are carried out:
All of the stripping operations are carried out at 5560C and atmospheric pressure. The original design of the stripping section required up to 288h (12days) for elution before achieving gold recovery. In the 1980s, a new stripping technique was developed, which takes from 12 to 24h, a considerable improvement over the original design. Gold from the pregnant thiourea solution is recovered by transferring the solution from the elution section to holding tanks for an electrowinning section. During electrowinning, the solution is circulated between the holding tanks and electrowinning cells. Residual gold on the stripped resin is expected to be below 100g/t.
Combining the elution and electrowinning operations enables elution of AuRIX 100 to be achieved by continuous electroelution using an eluate solution based on 1M sodium hydroxide at 60C (Mackenzie, 1993). In this process, the eluate is passed through a bed of resin and is continually recycled through an electrowinning cell back to the resin bed. In some instances, an improvement to the elution rate can be obtained when the alkaline eluant contains a low concentration of alkali metal cyanide salt and an alkaline salt of a carboxylic acid such as sodium benzoate (Virnig, 1996).
The elution of AuRIX 100 resin was studied exhaustively in a pilot plant in Mexico (Fisher etal., 2000). Initially, resin elution was performed with an eluate composition of 40g/L NaOH, 70g/L sodium benzoate, and 100mg/L free CN. Total eluate volume in the circuit was 64L. Sodium benzoate was found, in bench-scale testing, to accelerate elution kinetics. Elution was carried out at 60C for 6h. At termination of elution, the difference between the gold concentration in the eluant entering and the eluate exiting the column was less than 1mg/L Au. Elution profiles (eluate exiting the column) typical of the sodium benzoate-containing eluate are shown in Figure32.3. Loaded and eluted resin analysis for a sample using this eluate is given in Table32.1.
Elutions were also carried out without sodium benzoate. Figure32.4 shows the effect of sodium benzoate in altering the elution profile. At the end of 6h, the same differential between the eluant entering and eluate exiting the column (less than 1mg/L Au) was obtained. Therefore, sodium benzoate was not a necessary addition for elution of AuRIX 100. Eluate produced by this type of resin is suitable for conventional gold electrowinning with single-pass efficiencies of 66.591.7% and overall gold recovery in electrowinning of 94.599.8%.
A conventional strong-base resin (Duolite A161L) was used at the Golden Jubilee Mine for the recovery of gold, primarily due to the fact that no gold-selective resin was commercially available in the western world at that stage. The zinccyanide elution process is suitable for all nonselective strong-base resins. It is a reversible reaction and therefore, in order for the reaction to proceed to completion, it is important that the concentration of gold in the eluate should be kept as low as possible. This can be done either by the use of a very large volume of eluant, which is pumped in a single pass through the elution column (this is clearly an impractical and expensive approach), or by the use of the electroelution method, in which eluate solution is recirculated continuously between an electrowinning cell and the elution column. The latter approach was adopted at Golden Jubilee (Fleming and Seymore, 1989) and made use of the Mintek-designed electrowinning cell.
In the Golden Jubilee elution, 1 bed-volume of solution containing approximately 0.6M Zn(CN)42 was recirculated from a surge tank through a heating box, elution column, and an electrowinning cell, back to the surge tank. After about 6h of elution at 60C, zinc oxide and sodium cyanide were added to the eluate to compensate for the zinc cyanide that had loaded onto the resin and to restore the concentration in solution to the starting value of 0.6M.
The major disadvantage of the zinccyanide electroelution process is that it is slow; for the first 6months, each elution cycle was continued for a period of 45days. However, since only one elution was necessary per week, the slow elution did not create a bottleneck in the process. Moreover, it was possible to systematically reduce the elution time over 12months of operation, mainly as a result of improvements to the efficiency of the electrowinning process, to attain an ultimate elution period of 48h.
After elution, the resin was washed with water to remove entrained zinc cyanide solution and the resin was then regenerated with 1M sulfuric acid solution. The spent regenerant solution, containing zinc sulfate saturated with hydrogen cyanide gas, was pumped directly into a stirred lime slurry. The zinc cyanide that was produced during this neutralization reaction could be recycled to elution.
Vitrokele is a generic name given to the specialist technology that has been developed for the recovery of cyanide from precious-metal plant process streams (Signet Engineering, 1996) and the recycling of cyanide within the circuit. The resin that has been developed for gold-processing applications, Vitrokele 912, is manufactured by Rohm & Haas in France [Ed.:now part of Dow Chemical Company]. It is understood that the technology was commercially applied at the Connemara plant in Zimbabwe for gold recovery by means of resin-in-solution (RIS) and that significant capital cost savings were realized by use of the resin in place of the conventional activated-carbon process. The flowsheet given for the process suggests that similar chemistry to that used at the Golden Jubilee plant (Fleming, 1989) was employed. Therefore, it is thought that the resin used was probably a strong-base anion exchanger similar to the A161L resin.
Saint-Gobain Zirpro has been instrumental in the development of ceramic media for ultra-fine grinding applications in stirred mills. The history and experience are long, dating back to the mid nineteen seventies. The company was in-fact established in 1971 to refine Zirconium Oxide to meet the Saint-Gobain Glass requirements for high performance refractories. The history of the Glass division is somewhat longer and it celebrated its three hundred and fiftieth anniversary last year.
Zirconia (Zirconium Oxide) and Zircon (Zirconium Silicate) turned out to be significant raw materials necessary to manufacture high quality ceramic beads. Initially beads were produced by a fusion process; the operation required high temperatures and the beads were formed in a molten state. The resulting product (ER120) was based on Zircon and had a density of 4.0g/cc. The beads were round and non-abrasive and were ideal to replace the glass and natural sand products used in the burgeoning stirred or bead mill applications. A comparative increase in density of over 50% was the important factor; greatly enhancing the productivity of the mills. Higher density media, such as steel (7.5g/cc) for example, were largely discounted due to increased abrasion rates and contamination. Therefore ceramic technology was readily adopted by major industries, including pigments, paints and agrochemicals. The products proved to be extremely successful and remained at the forefront of the technology for over twenty years. Increased demands on milling technology were however inevitable and eventually faster, finer, more precise targets were expected. The industry responded with the development of new mill designs which operated at higher speeds, with higher energy and higher throughput rates. A new type of bead was required which would be tough and could withstand the new operating conditions. The result was the evolution of sintered beads, initially based on the same chemistry and having the same density as the fused products. These beads required low temperature forming before densification (sintering) at high temperature. The products proved to be suitable in many of the new applications, providing tougher beads with extended bead lifetimes. Zirpro again developed a class leading product (RIMAX) which gave exemplary performance in many varied fields for example cosmetics, inks and automotive coatings. Today the evolution has continued with the general acceptance of high density (6.0g/cc) stabilized Zirconia beads as the media of choice (Hassall and Nonnet, 2007). The higher density provides the potential for superior and economic grinding and although initially expensive, the controlled wear provides an overall cost effective solution. Zirpro developed a premier product ZIRMIL now widely adopted and widely used in the processing of the most demanding applications, such as pharmaceuticals and ceramics.
Mill design has also evolved to meet the ever increasing demands for ever more specialized materials and applications. At the forefront of these developments are two extremely different projects; the first is nano grinding of electronic materials and the second the ultra-fine grinding of ore bodies in the mining industry. For nano grinding, beads sizes of approximately 100m are required and potentially bead densities increasing beyond of 10g/cc. For mining the environment is severe and beads must withstand high impacts from hard and large feed materials in dilute slurries.
In the nano application Zirpro has launched a derivative of the ZIRMIL range. Advanced ceramic technology and process engineering have enabled the production of 100 and 200m beads in industrial quantities at economic price levels. The material is currently under evaluation in extended customer trials. For mining a composite material MINERAX has been developed and successfully launched into the industry (Fig. 1). It is a tough composite material with a classic density of 3.9g/cc and fully competent in all ultra-fine mining applications.
The Albion technology, schematized in Fig. 14, was developed by Xstrata Plc to treat concentrates produced from refractory base and precious metal ores. The technology is a sulfate based process employing ultrafine grinding (P80 of 1015m) at temperatures of around 8590C, atmospheric pressure to accelerate the kinetics and increase copper recovery level from chalcopyrite, in conventional agitated tanks with corrosion resistant alloy steel shells (Nazari et al., 2012a; Kowalczuk and Chmielewski, 2008; Ellis et al., 2008). The Albion process is an auto-thermal operation, i.e. the leach slurry temperature is set by the amount of heat released in the leaching reaction.
Both the Albion and ActivOx processes make use of ultrafine grinding to achieve sulfide dissolution (enhanced matrix attack) at lower temperature and pressures than required by conventional high pressure oxidation (Ellis et al., 2008). Note that fine grinding produces particles with P100 of <38m while ultrafine grinding produces particles sized within 1 and 20m range (La Brooy et al., 1994). Ultrafine grinding performs the same function as roasting, pressure oxidation, bio- and chemical oxidation which is to break down the sulfide matrix to liberate precious metals locked in silicates or other minerals (Flatman et al., 2010).
According to Hourn et al. (2005), ultrafine grinding of sulfide minerals to particle size of 80% passing 812m will eliminate mineral passivation by sulfur precipitates, as the leached mineral will disintegrate prior to the precipitate layer becoming thick enough to passivate it. The oxygen used for oxidation is injected into the base of the Albion leach reactor at supersonic velocity to achieve the required mass transfer and leaching rate. Chalcopyrite is acid leached through ferrous ion oxidation (Fe3+ being the main oxidizing agent) by oxygen according to the mechanism suggested by Hiroyoshi et al. (2001) in Eqs. (17) and (18). Ferrous oxidation by oxygen takes place as in Eq. (19).
Copper is extracted via SX-EW to produce copper cathodes (Kowalczuk and Chmielewski, 2008). Excess sulfide sulfur in chalcopyrite leaching is present in the residue as elemental sulfur. This makes precious metals recovery difficult as S0 can form a protective coating on the mineral particles. Once present, the coating may hinder the leaching process or even stop it completely. Jeffrey and Anderson (2003) and Lu et al. (2000) have suggested non-cyanide leaching methods such as sodium hydroxide to overcome elemental sulfur issues.
Oraby and Eksteen (2013) have shown that one can leach copper sulfides (including chalcopyrite), oxides and native copper effectively from a copper mineralprecious metal concentrate using an alkaline glycine solution at pH of 1011 with hydrogen peroxide as oxidant. The copper glycinate solution can be treated for copper recovery by a number of conventional technologies such as precipitation using NaSH or solvent extraction. Elemental sulfur formation is prevented by performing the oxidation in alkaline glycine solution.
Hourn et al. (2005) have reported that Albion leach process can operate under either acidic or alkaline conditions (see Fig. 14). In the first case (acid leach), base metals are extracted along with precious metals as by-product, while in the second case; precious metals encapsulated in pyrite, arsenopyrite, selenide or telluride ores are alkaline leached with no requirement of recovering base metals. The alkaline leach process of refractory precious metal bearing sulfides such as pyrite progresses through pyrite dissolution (Eq. (20)) to finally expose the precious metals for subsequent cyanidation.
The Albion process is commercially operational at two plants treating zinc sulfide concentrates that are located in Spain and Germany, while a third Albion process plant operating in the Dominican Republic is treating refractory gold/silver concentrates (Turner and Hourn, 2013).
Uwadiale (1990a) used selective oil agglomeration to upgrade Agbaja oolitic iron ore from a feed value of 45.6% Fe to a concentrate containing 65% Fe at 89.3% Fe recovery. The ore was very fine grained and ultrafine grinding (<5m) was required for liberation. The main iron mineral was goethite with minor hematite and maghemite. Reagent additions were 5mL of oleic acid, 5mL of 10% NaOH and 0.07g of sodium silicate added to the grinding charge of 50g. The pH of the ground pulp was adjusted to 9 and 7mL of kerosene added before agitating the charge in a blender. At the end of the agglomeration process the slurry was screened at 38m to separate the agglomerates from the siliceous gangue. At pH 11, no agglomerates were formed.