Baumgartner Et Al 2016 AusIMM

Comprehensive Mineralogical Characterisation at the Cerro Corona Cu-Au Porphyry Mine – the Fundamental Key for Geometallurgical Applications R Baumgartner1, P Gomez2 and G Escobar3 ABSTRACT The Cerro Corona mine consists of a Cu-Au porphyry deposit located in the Andes of northern Peru. The deposit comprises a typical Cu-Au multiphase porphyry mineralisation that has been overprinted by an epithermal event. This has generated a complex pattern of argillic alteration facies, which was accompanied by a locally structural-controlled Au-Cu mineralisation with the presence of gold and limited Cu suldes and sulfosalts. These features, along with the porphyrystyle mineralisation, produced a multifaceted occurrence of gold. Since the beginning of operations at Cerro Corona in 2008 there has been a continuous improvement of the geological and mineralogical knowledge. All mineralogical studies have mainly focused on host rock types and, to a lesser extent, alteration minerals, but little work has been done on the mineralisation itself. As a consequence, limited mineralogical information was available for use in the process optimisation initiatives. In the last two years, a series of mineralogical characterisation studies were undertaken with two main goals. First, to gain an understanding of the mineralisation genesis and second, to obtain a detailed characterisation of copper and gold minerals to support ongoing process optimisation. The rst goal was achieved by conventional optical microscopy while the second required electronic microscopy and other microanalyses due to the necessity of qu antifying several minerals of interest for the processing. The results were used to rene pre-existing geometallurgical domains, explain copper and gold losses in the otation process and to some extent, improve copper and gold recoveries. The mineralogical characterisation is an ongoing work. Mineralogy provides the basis for any optimisation exercise since it looks in detail at the material to be treated and therefore, should be part of all geometallurgical programs.

INTRODUCTION Geometallurgy is now embedded in a large number of operations around the world. A constant increasing number of people and companies are applying geometallurgy to improve the extraction of their resource. The roots of geometallurgy lie in the mineralogy and there are numerous papers showing that a detailed mineralogical characterisation permits to optimise the process (Williams, 2013; Olson Hoal, 2013; Brissette and Roman, 2012; Rocha, Ulloa and Diaz, 2012; Lamberg, 2011). The Cerro Corona mine has been in operation since mid2008 and has estimated resources of 3.05 Moz of gold at 0.81 g/t and 108 Mt copper at 0.42 per cent and reserves of

1.75 Moz of gold at 0.9 g/t and 60.5 Mt copper at 0.47 per cent (Gold Fields, 201). The mineralogical knowledge of the deposit has improved since the beginning of operation through mineralogical studies. However, these studies were qualitative and therefore, a quantitative study was necessary to be able to obtain additional information than only the mineral species and texture of a limited number of samples. For this reason, a detailed mineralogical study has been conducted in order to characterise the current mineral domains used at the mine, through quantitative mineralogy and particle analysis. The results of the characterisation are presented in this paper.

1.

MAusIMM(CP), Senior Geoscientist, Gold Fields La Cima SA, Santiago de Surco, Lima 33, Peru. Email: regina.baumgartner@goldfields.com.pe regina.baumgartner@goldfields.com.pe

2.

Geology and Exploration Manager, Gold Fields La Cima SA, Santiago de Surco, Lima 33, Peru. Email: paul.gomez@goldfields.com.pe paul.gomez@goldfields.com.pe

3.

Geometallurgy Superintendent, Gold Fields La Cima SA, Santiago de Surco, Lima 33, Peru. Email: gustavo.escobar@goldfields.co gustavo.escobar@goldfields.com.pe m.pe

THE THIRD AUS IMM INT ERNATIONAL G EOMETALLURG Y CONFERE NCE / PERTH, WA, 15�16 JUN E 2016

221

R BAUMGARTNER, P GOMEZ AND G ESCOBAR

GEOLOGICAL OVERVIEW Cerro Corona is a typical Cu-Au porphyry deposit and is located in the northern Andes of Peru. The deposit geology consists of a series of Miocene porphyritic diorite stocks intruding a Tertiary limestone sequence (Figure 1). Dioritic, monzodioritic and granodioritic porphy intrusive rocks occur also in the region (Figure 1). The Cerro Corona porphyritic diorite is composed of a total of six stocks within the Cerro Corona diorite and only four of them generated hydrothermal alteration within the adjacent stocks or in the adjacent limestone sequence with Cu-Au mineralisation. The largest and earliest stock altered adjacent limestone to a marble facies but does not appear to have precipitated economic mineralisation. Only scarce occurrence of skarn mineralisation in the limestone has been observed to date. The alteration facies observed in the altered and mineralised stocks are typical of porphyry-style alteration and show potassic, scare phyllic and propylitic alteration. Argillic and silicication alteration overprints the previous alterations. Advanced argillic alteration occurs in the upper parts of the deposit, is structurally controlled and overprints part of the porphyry-style alteration. Porphyry-style alteration and mineralisation have been therefore overprinted to some extent, resulting in gradational boundaries between alteration types as well as remnants of typical porphyry-style alteration. The argillic alteration is the main challenge since there are subtle mineralogical variations which deserve a subalteration facies classication (ARG1, ARG2 and ARG3). The main difference of the subalterations is the type and quantity of clay minerals. Table 1 shows the main alteration present at Cerro Corona and their characteristics. The main economic copper minerals are chalcopyrite and to a lesser extent chalcocite and co vellite. Gold minerals include mainly native gold and smaller amounts of electrum (up to ten per cent). Pyrite is present in all parts of the deposit and

does not show an occurrence zonation as it is the case in other porphyry copper-gold deposits.

MINERALOGY STUDIES In order to characterise the overall modal mineralogy, grain size as well as liberation of the principal copper mineral, chalcopyrite, a series of techniques have been used. In order to obtain representative samples from the different alteration types and spatial locations, samples from different sources have been collected.

Samples Samples have been collected from three different sources: 1. drill hole samples (diamond drill hole (DDH), 6 m adjacent intervals) 2. pit wall samples (10 kg channel sample) 3. production polygons (split of blastholes composite). It is clear that the third type of sample represents a large tonnage of rock and therefore, may show smoothed results, as opposed to a more local sample such as the rst and second type. In each type of samples, the different alterations are represented depending on the total tonnage of each alteration in the life-of-mine (LOM) mining plan. In Table 2, the details of the sample types and alteration are shown. All samples were selected in order to cover the most important rock types by tonnages and ounces equivalent. The spatial distribution has been taken into account whenever possible due to the availability of mined polygons at the time of the sample collection, DDH coarse rejects and access to pit walls. Figure 2 shows the spatial distribution of all samples collected. The copper and gold head grades of the samples collected were distributed among the distribution curve of the LOM resource model in order to cover most representative

FIG 1 – Regional geological map of the surroundings of the Cerro Corona mine. 222

THE THIRD AUS IMM INT ERNATIONAL GEOME TALLURGY CONFE RENCE / PERTH, WA, 15�16 JUNE 2016

COMPREHENSIVE MINERALOGICAL CHARACTERISATION AT THE CERRO CORONA CU�AU PORPHYRY MINE

TABLE 1 Key minerals and principal sulfide assemblages in the different alteration types. Alteration type

Key minerals

Principal sulfide assemblages

Occurrence

Silicification (SIL)

Quartz, Fe oxides-hydroxides

Fe sulfides-chalcopyrite

As intense stockwork around barren cores

Argillic 1 (ARG1)

Quartz, kaolinite, muscovite (illite), smectite

Fe sulfides-chalcopyrite (± chalcocite+covellite)

Upper parts of the deposit

Argillic 2 (ARG2)

Quartz, feldspars, montmorillonite, kaolinite, illite, carbonates

Chalcopyrite-(Fe sulfides)

Replacing locally K alteration

Argillic 3 (ARG3)

Quartz, montmorillonite muscovite (illite), chlorite


Potassic (K)

Quartz, K-feldspar, biotite, Fe Oxides ( magnetite)


Abundant in deep parts, and core of the quartz-diorite stocks

Supergene enrichment (SSH, weathering facies rather than alteration)

Quartz, feldspars, kaolinite, montmorillonite

Chalcopyrite-chalcocitecovellite, (Fe sulfides)

Supergene weathering, upper part of the deposit

TABLE 2 Summary of the sample collected for the mineralogical study. Alteration

Polygons

Pit wall

DDH

Total samples

SIL (silicification)

9

11

6

26

ARG1

3

14

6

23

ARG3

11

6

5

22

K

4

0

11

15

ARG2

2

1

5

8

SSH (supergene)

4

0

0

4

Total

33

32

33

98

DDH – diamond drill hole.

head grades. Figure 3 shows the copper and go ld head grades by alteration type for the collected samples.

Analytical methods and test work Complementary mineralogy studies were conducted on the collected samples as well as rougher otation test works. Mineralogical studies included QEMSCAN analysis (using a Quanta 650F QEMSCAN) to obtain modal mineralogy for all minerals as well as grain size, associations and liberation of chalcopyrite and pyrite. A detailed semi-quantitative X-ray diffraction (XRD) study has been completed on all samples with the aim of distinguishing between the different clay minerals present at Cerro Corona. For gold mineralogy, a preconcentration using heavy liquid separation (using sodium polytungstate at a specic gravity of 3.0) has been conducted on a selected number of samples (80) to increase the chance of analysing a higher number of gold grains. These were subsequently analysed using a Quanta 650F QEMSCAN in order to have similar information as for the other minerals as described above. A total of 666 gold grains were analysed. Laboratory rougher otation test work has been conducted on all samples, using standard conditions used at the mine site laboratory for comparison purposes.

RESULTS AND INTERPRETATION The quantitative mineralogical study on the 98 samples identied the presence of two types of pyrite. These include a common pyrite and an amorphous-type pyrite. These are generally accompanied by marcasite (Figure 4).

FIG 2 – Spatial distribution of the samples collected for the mineralogical study. The colours correspond to the type of alteration and the symbol to the type of sample (square: pit wall samples; flat disk: DDH samples; string: polygon samples). Current pit (dark grey) and final pit (light grey) are also shown.

FIG 3 – Copper and gold head grades by alteration type from the collected samples in a logarithmic scale. For abbreviations see Table 1.

THE THIRD AUS IMM INT ERNATIONAL G EOMETALLURG Y CONFER ENCE / PERT H, WA, 15�16 JUNE 2016

223


The amorphous-type pyrite appears to be the results of a recrystallisation of previous pyrite and chalcopyrite. Microprobe analysis on this type of pyrite has shown that it can contain up to 1.7 wt per cent copper. This type of pyrite is called amorphous because it not identied by the XRD analysis but as an amorphous material, due to its poor crystallisation. The distribution of amorphous pyrite in the different alterations is shown in Figure 5. The silicication and ARG1

FIG 4 – Occurrence of two types of pyrite, generally accompanied by marcasite. Note the porous texture of the amorphous-type pyrite.

alterations contain the highest quantity of amorphous pyrite while the potassic alteration contains less. Pyrite occurs in lower quantities than amorphous pyrite.

Modal mineralogy The overall modal mineralogy has been obtained and classied by alteration type, which is shown in Figure 6. Alteration ARG2, contain a higher content of carbonates and chlorite (as clinochlore), indicating that this alteration has overprinted part of the propylitic alteration. This is also shown by its occurrence in the distal portions of the deposit. The silicied portion of the deposit, consisting of quartz-veining stockwork and classied as SIL, contains the highest quartz contents. It is also the alteration with t he highest contents of Fe oxides and hydroxide in the form of magnetite and hematite. As expected, it can be seen that secondary copper minerals such as covellite and chalcocite (shown under the name ‘other Cu suldes in Figure 6) are the most abundant in the SSH type of alteration. The pyrite content varies from one alteration to another and is the most abundant in SIL and ARG1.

FIG 5 – Box plot of the quantity of pyrite and amorphous pyrite by alteration type.

FIG 6 – Average modal mineralogy for each type of alteration. Minerals from the legend occur from left to right in the bar chart. 224



Clay mineralogy Clay mineralogy was determined by quantitative XRD in all samples. In some of them, it was possible to distinguish between intercalated montmorillonite-illite and montmorillonite. If it was not possible, they were grouped together and reported as such. The following clays have been determined: • •

•

•

•

kaolinite montmorillonite 1 – corresponds to an interstratied montmorillonite-illite clay (expansive) montmorillonite 2 – corresponds to the mineral phase montmorillonite montmorillonite 1–2 – includes both montmorillonite 1 and 2 illite (reported as muscovite).

Figure 7 shows the abundance of the different clay minerals by alteration type. The argillic type alteration (ARG1, ARG2 and ARG3) contain more clays than the other alterations. The ARG1 alteration contains the highest amount of illite of all alteration types (18.4 per cent on average). Since XRD analysis can become expensive for a large number of samples, not all samples from drill hole campaigns or exploitation polygons can be analysed. However, a quantitative estimation of clays is possible using short wave infra-red (SWIR) through a Terraspec © instrument and corresponding software. The SWIR spectra was measured on coarse reject samples from each drill hole sample, as well as larger samples such as polygon samples. With the XRD and SWIR information, it is possible to build a quantitative calibration model that is used to quantify clays. This provides a quantitative clay data to be used for the interpretation of metallurgical recovery results. The resolution of this quantication is by no means similar to XRD but provides information on an acceptable resolution for prediction purposes, such as recovery.

Microtextures Information on the texture can be obtained with the QEMSCAN through the particle analysis. This permits to have a visual control on the microtextures of unliberated chalcopyrite grains. Table 3 shows examples of various textures at Cerro Corona that can impact processing based on the work of Amstutz (1961). This information can be used to predict the behaviour of the different particles depending on

the type of intergrowth in a otation circuit. It is clear that an extensive database of the dominant type o f texture in each alteration is necessary in order to make prediction. Currently, the textures shown in Table 3 have been identied and the future work will provide additional data within the different types of alteration and spatially. The simple intergrowth type with rectilinear or slightly curved boundaries (Type 1a, Table 3), the vein-like type (type 3a, Table 3), and rimmed textures belonging to type 2a (Table 3) are microtextures in which chalcopyrite could be partially or totally liberated with a ner milling and therefore, increase the copper extraction. In the microtexture 1b (Table 3), the chalcopyrite will remain locked and will be probably lost in the tails. Therefore, with a simple texture analysis, it is possible to predict the behaviour of chalcopyrite in the otation circuit but it also permits to identify the causes of copper losses in the tails. As mentioned above, in some cases, ner milling would increase liberation and reduce the losses of copper in the tails. In other cases, a solution may be more challenging. The silicication and potassic alterations appear to contain more simple intergrowths of type 1a and 3a while argillic alterations (ARG1–ARG3) contain higher proportions of types 1b and 2a intergrowths.

Grain size of chalcopyrite and gold The grain size of chalcopyrite and gold has been measured in each sample and in selected samples, respectively. Chalcopyrite is the main copper carrier and native gold is the main gold carrier.

Chalcopyrite The grain size of chalcopyrite has been measured in the 98 samples and the results are shown in Figure 8 as average grain size by alteration type. It can be seen that the average P80 of chalcopyrite is the highest in ARG2 and K alteration (82 μm) and the lowest in ARG1 (65 μm). This information needs to be coupled with the liberation and association information since it is the latter that will provide information on the theoretical extraction.

Gold The size of the gold grains has been measured in 80 samples. Figure 9 reports the distribution of the gold grain size by alteration type. The ARG1 alteration contains the smallest

FIG 7 – Clay mineralogy according to alteration type. Minerals from the legend occur from left to right in the bar chart. THE THIRD AUS IMM INT ERNATIONAL G EOMETALLURG Y CONFER ENCE / PERT H, WA, 15�16 JUNE 2016

225


TABLE 3 Examples of different textures at Cerro Corona that may impact the process (modified from Amstutz, 1961). Colours: orange = chalcopyrite; green = amorphous pyrite; yellow = pyrite; pink = quartz. Images from QEMSCAN particle analysis or backscattered electron. Type

Typical microtextures

1a

Description •

•

•

1b

•

•

2a

3a

•

•

•

Simple intergrowth simple or locking type Rectilinear or slightly curved boundaries Most common type

Mottles, spotty or amoeba-type locking or intergrowth Simple common pattern

Coated, mantled, enveloped, corona-, rim-, ring-, shell- or atoll-like

Vein-like, stringer-like type Common examples

gold grains (5 μm) while the ARG2 alteration shows the highest average gold grain size (18 μm). It is clear that there are different mineralogical factors that will inuence the copper and gold recovery by otation, but there are also operational factors.

Liberation and associations The liberation and association information is key for prediction or interpretation of copper and gold recoveries. In fact, it provides information on the occurrence of the mineral of interest. If it is measured on a material that will enter the plant, it can provide predictive information. When it is measured on tails, it can inform on the losses of the valuable mineral (due to the material or to operational factors). At Cerro Corona, it has been used to predict the theoretical behaviour of chalcopyrite using theoretical grade-recovery curves deduced from the mineralogical data and subsequently compared to laboratory otation results. The liberation of chalcopyrite on average and by alteration type is shown in Figure 10 while its associations are shown in Figure 11. It can be noted that most of the chalcopyrite has ternary/complex association, ie more than two or more minerals associated to chalcopyrite. In this case, the minerals associated to chalcopyrite are pyrite, amorphous pyrite, quartz, muscovite (as illite), K-feldspar and/or minor magnetite/hematite and goethite. Most of the 226

Examples at Cerro Corona

Occurrence and extraction solutions Abundant

Finer milling.

Abundant

No direct solution, will float if chalcopyrite is associated with pyrite. If associated with gangue, it may likely be depressed.

Scarce

Finer milling

Scarce

Finer mill size to increase liberation.

binary associations consist mainly of chalcopyrite and quartz, amorphous pyrite, pyrite or muscovite (probably illite) to a lesser extent. Theoretical grade-recovery curves have been obtained for each sample and can be compared with the rougher otation tests conducted in the laboratory for the corresponding samples. One example is given here for the theoretical recovery-grade curves for ARG3 samples. Recoveries obtained from rougher otation are highlighted for two selected samples in Figure 12. The copper recovery of 84 per cent corresponds to a curve that is atter than the curves with higher recoveries (eg 93 per cent copper recovery). This is due to the fact that the liberation and grade size inuenced the performance of the recovery. Figure 13 shows that the sample returning lower copper recovery contains smaller liberated chalcopyrite grains and scarce liberated binary particles (Figure 13a) than in the sample returning a higher copper recovery (Figure 13b). The latter shows larger chalcopyrite liberated grains as well as a larger amount of liberated binary particles with larger grain sizes than the other sample (Figure 13b). Therefore, grain size, liberation and association permit to predict the copper recovery using theoretical grade-recovery curves. The liberation and association of native gold has also been determined in order to determine gold losses in the obtained



FIG 8 – Box plot showing the average P80 for chalcopyrite grains by alteration type.

FIG 9 – Box plot showing the average P80 for gold grains by alteration type. The numbers in each box correspond to the number of analysed samples. The number of gold grains analysed by alteration type was: ARG1: 75, ARG2: 44, ARG3: 171, K: 146, SIL: 202 and SSH: 28.

FIG 10 – Liberation grade of chalcopyrite by alteration type. For reference, < = 100 per cent indicates that 91 to 100 per cent of the surface of the chalcopyrite is liberated. Liberation classes occur from left to right in the bar chart. THE THIRD AUS IMM INT ERNATIONAL G EOMETALLURG Y CONFER ENCE / PERT H, WA, 15�16 JUNE 2016

227

R BAUMGARTNER, P GOMEZ AND G ES COBAR

FIG 11 – Associations of chalcopyrite classified as free chalcopyrite, binary and ternary/complex associations. The latter is in general a combination of pyrite, amorphous pyrite, quartz, muscovite (as illite), K-feldspar and/or minor magnetite/hematite and goethite.

FIG 12 – Theoretical grade-recovery curves calculated from mineralogy results for the ARG3 alteration (150 μm). The two coloured curves (red – 84 per cent; orange – 93 per cent) were also obtained from mineralogy but the laboratory flotation results are shown in the text box for comparison. gold recoveries. Samples with low gold recoveries contain in general gold locked in pyrite or quartz. Gold locked in amorphous pyrite and pyrite (Figure 14) is in general lost in the cleaner circuit and to a lesser extent in the rougher circuit while gold locked in quartz is lost in the rougher otation. Gold locked in chalcopyrite will report to the nal concentrate. In Figure 14a, the gold is mainly locked in chalcopyrite and as expected, the gold rougher otation recovery for this sample returns 95.2 per cent while Figure 14b, the gold rougher recovery for this sample is 89.9 per cent because pyrite has been depressed already in the rougher stage (total gold grains analysed: 15). 228

Relations of recovery and mineralogy As shown above, the recovery is closely related to the mineralogy of the valuable elements as well as the gangue minerals. In a plant, operational factors have to be taken into account. However, information on the mineralogy permits to predict the theoretical recovery and plan in accordance to the obtained results. The comparison of the otation data with the mineralogical data is best done with plant data since it is the real recovery obtained from a material. In fact, laboratory otation results, although they are standardised, contain still variability that can be attributed to the otation personnel when not automatised.



A

B

FIG 13 – Particle maps for chalcopyrite grains. (A) Particle map for the lower copper recovery sample (P9), showing smaller grain sizes and scarce binary liberated particles. (B) Particle map for the higher copper recovery sample (GDF192/64-70) showing a larger size of the liberated chalcopyrite grains as well as more liberated binary particles than in (A). For colour legend, see Figure 14.

A

B

FIG 14 – Particles with gold encapsulated (blue and black circles). (A) Gold encapsulated in mainly chalcopyrite and to a lesser extent in amorphous pyrite and pyrite. (B) Gold encapsulated in amorphous pyrite and pyrite. Colour legend: orange = chalcopyrite, yellow = pyrite, green = amorphous pyrite, pink = quartz, light orange = K-feldspar. A plant example for gold recovery illustrates the occurrence of gold discussed above. During a day of operation, argillic alteration material (ARG1 and ARG3) was processed. The gold recovery in the rougher otation circuit was 70 per cent, but this recovery decreases drastically in the cleaner circuit to a gold recovery of 51 per cent. The head grade of the material processed was 0.51 per cent copper and 0.72 g/t gold. Looking in details at the occurrence of gold, the mineralogical results show that most of the observed gold is ne grained and locked in amorphous pyrite, pyrite and

to a lesser extent chalcopyrite (Figure 15). This can justify the losses of gold in the cleaner tails. It is not discarded that losses due to operational factors occurred. The others losses that might have been overlooked are the eventual gold grains locked in quartz. However, most of those would have been lost in the rougher otation circuit. Some ne-grained free gold might have also been lost in the rougher otation if the size was not large enough to be able to oat.

FIG 15 – Mineral particles with encapsulated gold (black circles). Total gold grains: 15. For colour legend, see Figure 14. THE THIRD AUS IMM INT ERNATIONAL G EOMETALLURG Y CONFER ENCE / PERT H, WA, 15�16 JUNE 2016

229


CONCLUSIONS The mineralogical information obtained for this study provides useful and applicable information for the ore characterisation but also recovery predictions, within errors. Since the extraction process treats a combination of minerals, it is important to have the information of the valuable mineral (in this case chalcopyrite and gold) as well as the gangue minerals. These minerals can be associated to the valuable mineral and impact indirectly the recovery or can be present in the rock (such as clays) and have a direct impact on the extraction process by operational problems such as otation problems and/or milling concerns. If all major alterations are characterised mineralogically and spatial location is taken into account, it provides a powerful tool for extraction predictions and exibility in the mine planning and processing. Of course, this implies a large amount of work and related costs but it can provide an insight at each alteration type and their performance in the processing plant. This might conduct to a different scheme of treatment for each type of alteration, which has been previously treated the same way. The study also permitted to characterise in detail the different argillic alterations (ARG1–ARG3) and show that their distinction is important. This had a direct impact in the blending strategy at the mine. This is a rst step in the geometallurgical ore characterisation program at Cerro Corona. The second step is to obtain mineralogical information from daily plant products to be able to monitor the evolution of the performance of the plant by identifying if the copper (and gold) losses are due to the nature of the material or due to operational factors.

ACKNOWLEDGEMENTS Gold Fields La Cima SA is acknowledged for permission to publish this paper and being supportive. The entire

230

geometallurgy team at Cerro Corona is thanked, including Flor Nuñez, Anthony Julian, Valentin Yapo and Julio Castro. The geology team is thanked for helping in the sample selection. Mahdi Ghobadi is also thanked for extensive discussions and ndings during sample analysis.

REFERENCES Amstutz, G C, 1961. Microscopy applied to mineral dressing, Quarterly of the Colorado School of Mines , 56:445-481. Brissette, M and Roman, E, 2012. Concentrate grade and regrind size improvements on the east pit hypogene ore at Kemess Mine, in Proceedings 44th Annual Meeting of Canadian Mineral Processors ( eds: E Legault-Seguin and J Zinck), pp 27–38 (Canadian Mineral Processors: Ottawa). Gold Fields Ltd, 2015. Mineral Resource and Mineral Reserve supplement to the integrated annual report 2015, 146 p. Available from: . Lamberg, P, 2011. Particles – the bridge between geology and metallurgy, in Proceedings Conference on Minerals Engineering (ed: J Atalo), pp 1–16 (Luleå Takniska Universitet). Olson Hoal, K, 2013. The importance of mineralogical input into geometallurgy programs, in Proceedings The Second AusIMM International Geometallurgy Conference (ed: S Dominy), pp 17–25 (The Australasian Institute of Mining and Metallurgy: Melbourne). Rocha, M, Ulloa, C and Diaz, M, 2012. Geometallurgical modelling in the Los Bronces Mine, in Proceedings International Seminar on Geometallurgy 2012 (eds: C Barahona, R Kuyvenhoven and K Pinto) , p 34 (Gecamin: Santiago de Chile). Williams, S, 2013. A historical perspective of the application and success of geometallurgical methodologies, in Proceedings The Second AusIMM International Geometallurgy Conference (ed: S Dominy), pp 37–47 (The Australasian Institute of Mining and Metallurgy: Melbourne).


Baumgartner Et Al 2016 AusIMM

Recommend Documents