• Author / Uploaded
• rolandoh1

Citation preview

Rheological Investigation of the Flotation Performance of a High Clay Containing Gold Ore from Carlin Trend S Farrokhpay1, D Bradshaw2 and R Dunne3 ABSTRACT It is well known that clays affect the flotation performance, and many of the refractory gold ores around the world contain clay minerals. Therefore, an ore sample from the Carlin Trend was selected for this study as these ores contain large quantities of clays. The presence of clay minerals in an ore makes it susceptible to a range of variables, such as the solution chemistry determined by pH and the chemistry of the rheology modifiers. The viscosity of the ore slurry was measured as a function of solid concentration and particle size distribution, in addition to using different rheology modifiers and pH regulators. The results were correlated with the ore flotation performance. It was found that various reagents known commonly as dispersants or rheology modifiers have different effects on the ore slurry viscosity. This study revealed that the gold flotation recovery can be improved by changing the pulp rheology.

INTRODUCTION The Carlin–type deposits represent some of the largest hydrothermal gold deposits in the world, and the Carlin trend is a well-known major refractory gold deposit in Nevada, USA. The mineralisation of the deposit is divided into oxide, sulfide, refractory, or carbonaceous sulfide ores. The refractory sulfide consists of finely disseminated pyrite, arsenopyrite and arsenian pyrite, all of which must undergo a refractory ore treatment process (such as pressure oxidation or roasting) to make the gold available for cyanidation. Carlin-type ores are characterised as relatively high-grade gold and/or silver deposits with arsenic enrichment. The gold is extremely fine-grained, submicron size and hosted almost entirely within pyrite. Some gold is also found in arsenopyrite (Castor, 2001). The underlying rocks, out of which the minerals are dissolved, are normally silty carbonates, though silicates and other sediments are also possible. The material in the deposit is altered in a way that the carbonate minerals are either dissolved or converted to silicates by silicate rich hydrothermal water. Another alteration is the formation of clay minerals by the interaction of water and feldspar. The absences of base metal sulfides, and the even distribution of pyrite and arsenopyrite in the host rock, are the most obvious differences with the other sulfide deposits (Arehart, 1996). Previous studies on Carlin composites at Newmont has shown that the composites containing higher quantities of carbonate minerals have worse liberation properties with regards to pyrite and arsenopyrite, and that the arsenian pyrite contains the majority of the gold values (Kappes, Brosnahan and Gathje, 2010).

It is well known that clays have a major impact on mineral processing in various ways. They have process implications for gold, nickel, copper, lead/zinc and iron ore processing. The extremely small particle size of clays, along with their high surface area makes clay minerals hard to manage in a wide range of unit operations; from crushing/screening to filtering. Clay minerals result in higher flotation reagent consumption and poorer selectivity of valuable minerals. They can also affect froth stability (Farrokhpay and Bradshaw, 2012; Farrokhpay, 2011) as the overall flotation performance. Furthermore, the rheological behaviour of slurries depends mainly on the viscosity of the dispersion medium, pulp density, particle size and shape, and particle-particle and particle-dispersion medium interactions (Shaw, 1992). For mineral slurries the dispersion medium is water and being a Newtonian fluid, water has a fixed viscosity (at a fixed temperature). However, the water chemistry and the type/ amount of ions present in water (for example, in sea water) can affect the rheology of mineral suspensions (Farrokhpay and Zanin, 2012). The particle-particle and particle-dispersion medium interactions are also important and depend on both particle properties and solution physical chemistry; they depend on factors such as pH, ionic strength, and chemical reagents (such as dispersant and flocculants) (Cruz et al, 2013; Prestidge, 1997). Many of the refractory gold ores around the world contain large quantities of clay minerals, and flotation is one method used ahead of pre-treatment and cyanide leaching. These high clay ores are excellent candidates for assessing the impact of clay mineralogy on rheology and the flotation process.

1. MAusIMM, Senior Research Fellow, Julius Kruttschnitt Mineral Research Centre, The University of Queensland, 40 Isles Road, Indooroopilly Qld 4068. Email: [email protected] 2. MAusIMM, Professor Research Fellow, Julius Kruttschnitt Mineral Research Centre, The University of Queensland, 40 Isles Road, Indooroopilly Qld 4068. Email: [email protected] 3. Fellow Metallurgy, Newmont Mining Corporation, 10101 East Dry Creek Road, Englewood Co 80112, USA. Email: [email protected] World gold CONFERENCE / Brisbane, qld, 26 - 29 september 2013

333

S Farrokhpay, D Bradshaw and R Dunne

Therefore, an ore sample from the Carlin Trend was selected for the study. The rheology of the ore slurry was monitored, as a function of solid concentration and particle size distribution, in addition to using different rheology modifiers and pH regulators. The Carlin type ores usually have quartz, kaolinite, illite and muscovite. A comparison of the rheological properties (yield stress and viscosity) of suspensions of different clays including illite and kaolinite relative to quartz has been recently published as shown in Figure 1 (Ndlovu et al, 2013). Such a comparison provides a preliminary indication of the rheological effects of the major gangue minerals present in the ore sample, demonstrating the rheological dominance of clay minerals relative to quartz. Although quartz occurs in a higher proportion than other clay minerals in the ore, it can be seen that at equivalent concentrations, illite and kaolinite suspensions are characterised by comparatively higher suspension yield stresses and viscosities. Although the rheological behaviour of illite and kaolinite may change when present in a more complex ore system due to interaction with other minerals, it can be seen from this comparison that the combined effects of kaolinite and illite do indeed contribute significantly towards the overall flow behaviour and process performance of the ore suspensions. Further investigations using more controlled systems would be required in order to obtain conclusive results, and to determine the rheological effects of each clay mineral.

were adjusted at 12 L/min and 1000 rev/min, respectively. The total flotation time was 15 minutes. The reagents added were potassium amyl xanthate (PAX) as collector (125 g/t added at three different stages), MIBC as frother (40 ppm), and copper sulfate (125 g/t) as activator. The standard flotation condition was at the ore slurry natural pH (ie pH 7) and pulp density of 25 per cent solid. The flotation was also conducted at different pulp densities of 18 per cent, 32 per cent and 38 per cent. All per cent solids are per cent weight (w/w) unless otherwise stated. All recalculated head assays for these samples were crosschecked to determine both the reproducibility of the sampling technique and the consistency of the ore feed being used.

Rheology measurements The viscosity of the ore slurry samples was measured using an Anton Paar DR301 rheometer with vane geometry. Measurements were made at a strain rate of 40 s-1, at a measuring time of 60 seconds. Different rheology modifiers were chosen from those commonly known in the mineral processing industry as ‘dispersants’ or ‘rheology modifiers’: Cyquest 40E and 3223 (Cytec), BorreFlo D-919 (a sodium lignosulfonate from Borrogaard), Nalco 71307, a high cationic charged latex and Freevis 903, anionic polymer (both from Nalco). Soda ash (Na2CO3), caustic soda (NaOH), and lime (CaO) were used as pH regulators. The scope of the rheology test work is to investigate the impact of physical parameters as well as chemical additives on the rheology behaviour (viscosity) of the ore slurry, and consequently, its flotation performance.

RESULTS AND DISCUSSION Ore characterisation The chemical assay of the ore using inductively coupled plasma (ICP) chemical analysis and the ore mineralogy determined by X-ray diffraction (XRD) are presented in Table 1.

TABLE 1 Chemical assay and X-ray diffraction results. Chemical assays (inductively coupled plasma)

FIG 1 - A comparison of the viscosity of different clays including the major gangue minerals usually present in the Carlin ore.

MATERIAL AND EXPERIMENTAL METHODS The ore used in this study was obtained from the Carlin Trend in Nevada which was a gold ore with high content of clay minerals. The density of the ore was found to be 2.735 g/cm3. Grinding calibration tests were conducted to establish the time required to achieve the targeted particle size (P80) values. The ore samples were ground (in Brisbane tap water) to a P80 of 53, 106 and 126 µ, for a predetermined time.

Mineralogy (X-ray diffraction)

Au

1.04 ppm

Anatase

0.6%

As

0.33%

Calcite

15.6%

C

2.39%

Dolomite

1.7%

S

2.88%

Gypsum

4.0%

Al

5.3%

Illite + Muscovite

21.7%

Ca

8.4%

Kaolinite

7.6%

Fe

3.4%

Marcasite + Pyrite

4.6%

K

2%

Quartz

44.1%

Mg

0.7%

Si

22%

A bottom driven 5 L flotation cell was used to carry out the flotation tests. The gas flow rate and impeller rotational speed

The XRD results show that the ore is comprised predominantly of quartz. Marcasite and pyrite were identified as the major sulfide minerals. Dolomite, gypsum, siderite and trace amounts of anatase are also present. Other minerals detected, albeit in very small concentrations, were elemental sulfur, sphalerite, jarosite and alunite. Perhaps of

334

World gold CONFERENCE / Brisbane, qld, 26 - 29 september 2013

Flotation experiments

Rheological Investigation of the Flotation Performance of a High Clay Containing Gold Ore from Carlin Trend

most significance is the clay content. Illite/muscovite and kaolinite were identified as the major clay gangue minerals similar to what has been previously reported for Carlin type ores (Wells and Mullens, 1973). These minerals belong to the phyllosilicate group of minerals which are synonymous with several potential processing issues including bubble coalescence and low grinding efficiency, all of which are related to medium viscosity properties. It is known that the rheological behaviour of mineral slurries can dramatically affect separation processes such as flotation. The rheological behaviour of mineral slurries indicates the level of interparticle interaction or aggregation, and therefore, it can be used as a useful analytical tool to understand processing control parameters (Farrokhpay, 2012). Preliminary studies linking the mineralogical content and rheological response have identified phyllosilicate gangue minerals as major contributors towards ore slurry flow behaviour (Burdukova et al, 2008). However, there is no clear understanding of the process performance effect specific to each mineral belonging to this group. Moreover, there is a large degree of variability in the rheological characteristics of minerals belonging to this group, with some exhibiting extreme complex rheological behaviour, while others do not deviate significantly from non-phyllosilicate minerals such as quartz. Owing to its wide use in the ceramics industry, kaolinite has been studied extensively, relative to other phyllosilicate minerals. However, most of these studies have been fundamental and not directly related to the gangue mineralogy effects in minerals processing. Despite the relative abundance of illite in many ores, an understanding of the specific effects it has on process performance is often speculative. It should be highlighted that clay minerals are inherently small and fine particle size has been related to: changes in rheology (Adeyinka et al, 2009) increased froth stability increased entrainment of fine gangue to the concentrate increased potential for the formation of slimes coatings (Arnold and Aplan, 1986) •• increased residence time. •• •• •• ••

0

2

4

When clay minerals occur in an orebody they will enter the processing circuit because their small size makes them virtually unavoidable. However, the effect of particle size is often combined with the surface chemistry and ion exchange capacity of the minerals. Celik, Hancer and Miller (2002) found that in situ sonication during conditioning seems to dislodge clay slimes from the surface of the valuable minerals.

Impact of physical parameters on the ore slurry rheology and its flotation performance The effect of physical parameters (per cent solid and particle size distribution) on the ore slurry rheological behaviour was investigated. Preshearing of the samples ensured homogeneity and complete dispersion of the solids prior to measurement. The solid concentration data was converted from per cent weight (w/w) to per cent vol (v/v) using the ore density and the apparent viscosity values were plotted as a function of both solid concentrations. The viscosity values of the ore slurry as a function of pulp density for different P80 values of 53, 106 and 125 µ are presented in Figure 2. The results show a characteristic increase in the apparent viscosity with solid concentration. A critical solid concentration was observed at 25 per cent (w/w), the point at which the viscosity of the ore suspension dramatically increased. This signifies the concentration below which the flow behaviour of the ore suspension can be more easily controlled. At this concentration the particle size effects also become significant, with apparent viscosities increasing in the order of 125 µm < 106 µm < 53 µm. Such behaviour is expected, since it has been well established that viscous effects are much more pronounced at ultra-fine particle sizes (Tangsathitkulchai, 2003; Prestidge, 1997; Mikulášek, Wakeman and Marchant, 1997). For comparing flotation performance in different conditions, it is often useful to consider both the grade and the recovery simultaneously, using a ‘grade/recovery curve’. This is a graph of the recovery of the valuable metal achieved versus the product grade at that recovery, and is particularly useful for comparing separations where both the grade and the

Pulp density (%), v/v 8 10

6

12

14

16

18

160

Apparent viscosity (cP)

140 120 100 80 60 40 20 0 0

5

10

15

20 25 Pulp density  (%), w/w

P80‐53 µm

P80‐106 µm

30

35

40

P80‐125 µm

FIG 2 - Apparent viscosity of the ore as a function of pulp density for different P80 values. World gold CONFERENCE / Brisbane, qld, 26 - 29 september 2013

335

6.0

1.4

5.0

1.2

4.0

As  Grade (%)

Au Grade (ppm)

S Farrokhpay, D Bradshaw and R Dunne

3.0 2.0

0.8 0.6 0.4

1.0

0.2

0.0 0.0

20.0

40.0

60.0

80.0

0.0

100.0

0.0

Au Recovery (%)

20.0

40.0

60.0

80.0

100.0

As Recovery (%)

30.0

25.0

25.0

20.0

20.0

Fe  Grade (%)

S Grade (%)

1.0

15.0 10.0

15.0 10.0 5.0

5.0 0.0 0.0

20.0

40.0

60.0

80.0

100.0

S Recovery (%)

0.0 0.0

20.0

40.0

60.0

80.0

100.0

Fe Recovery (%)

FIG 3 - Effect of  pulp density on the ore flotation response (18 per cent ,  25 per cent ■ , 32 per cent ● and 38 per cent ▲). recovery are  varying. Figure 3 shows the grade/recovery relationship for gold, sulfur, arsenic and iron, at different flotation pulp densities. As gold appears in pyrite and arsenopyrite, a good agreement is observed among Au, As, S and Fe assays. Figure 3 shows that the highest recovery of Au, As, S and Fe is for the highest pulp density of 38 per cent (Au, As, S and Fe recoveries at different pulp densities follow the order of 18 ≈ 25 per cent