Grinding The Primary Conditioner

C J Greet1 and P Steinier2 1. 2. MAusIMM, Magotteaux Australia Pty Ltd, Suite 4, 83 Havelock Street, West Perth WA 6005. E-mail: christopher.greet@magotteaux.com Magotteaux Australia Pty Ltd, Suite 4, 83 Havelock Street, West Perth WA 6005. E-mail: philippe.steinier@magotteaux.com ABSTRACT Most operations, when considering their comminution circuit, are primarily concerned with achieving the desired particle size distribution to obtain adequate liberation for the separation process, at minimum cost. Invariably, the effect the grinding media has on the subsequent separation process (ie flotation, leaching) is not taken into account when selecting the media. This choice is driven by cost rather than metallurgical outcomes. Maximising valuable mineral recovery, with optimum selectivity against gangue minerals, is about good particle preparation. And, particle preparation is strongly related to achieving adequate mineral liberation with the correct pulp chemical conditions. However, the pulp chemistry of a system is largely ignored and relies heavily on pH control and the addition of appropriate reagents (collectors, frothers, activators, depressants). The changes effected by the addition of the various reagents are obtained through conditioning the ground pulp prior to or during the concentration stage, subsequent to grinding. Frequently, reagents are added during grinding with good effect, but the impact of the grinding media on down stream processing is usually overlooked. A number of case studies examining the effect of grinding media on pulp chemistry and subsequent flotation performance were completed on several different sulfide mineral systems (for example, lead/zinc, copper, refractory gold). The evidence suggests that a change in grinding media type from forged to high chrome (ie conditioning the pulp during grinding) resulted in a shift in pulp Eh to less reducing conditions, an increase in the dissolved oxygen content of the pulp and a reduction in the iron species present. The changes have a positive impact on flotation behaviour. The results of the case studies are discussed. INTRODUCTION The key to a successful separation in mineral processing is the preparation of particles with adequate liberation under the correct pulp chemical conditions. While the importance of liberation on flotation separations is generally understood and well documented in the literature (Johnson, 1987; Jackson et al, 1989; Young et al, 1997; Greet and Freeman, 2000), the importance of pulp chemistry is more nebulous, particularly with regard to the impact of grinding media. Extensive work examining the electrochemical interactions between grinding media and sulfide minerals has been completed (for example, Iwasaki et al, 1983; Natarajan and Iwasaki, 1984; Yelloji Rao and Natarajan, 1989a; Yelloji Rao and Natarajan, 1989b). Broadly, these studies indicate that most sulfide minerals are more noble than the grinding media used during comminution, therefore a galvanic couple between the media and the sulfide mineral(s) exists, which increases the corrosion rate of the

grinding media. The corrosion products of the grinding media, iron oxy-hydroxide species, invariably precipitate on to the surfaces of the sulfide minerals thereby affecting their floatability (Johnson, 2002). Cullinan et al (1999) completed laboratory experiments examining the effect of different ferrous based grinding media on galena flotation performance in the lead circuit of the Mount Isa Mines Lead/Zinc Concentrator. On a size-by-size basis, their work indicated that grinding with high chrome grinding media increased the maximum galena recovery of the -3 micron fraction from 63 per cent (for forged steel grinding media) to 79 per cent (Figure 1). This improvement could not be attributed to entrainment, since slightly higher water recovery values were obtained when grinding with forged steel grinding media. Improvements in selectivity for galena against sphalerite and iron sulfides were also observed when high chrome grinding media was employed. Pulp chemical measurements recorded during these tests indicated that grinding with forged steel resulted in significantly more reducing conditions than those observed when high chrome grinding media was used (Table 1). The other feature of the pulp chemistry was the elevated levels of oxidised iron species (as measured using an EDTA extraction technique (Rumball and Richmond, 1996; Cullinan et al, 1999; Greet and Smart, 2002)) recorded for the forged steel case, which were 1.56 times greater than the values reported for the high chrome alloy. The elevated levels of EDTA extractable iron for the forged steel grinding media were the result of increased corrosion between this media type and the sulfide minerals in the ore. It was postulated that the improvement in fine (-3 micron) galena flotation response when the ore was ground with high chrome grinding media was due to lower levels of oxidised iron species within that system.
100.0 90.0 80.0 70.0

Rmax, %

60.0 50.0 40.0 30.0 20.0 10.0 0.0 1.0 10.0 100.0

Size, microns
Forged steel high chrome

FIG 1 - Maximum galena recovery versus particle size for Mount Isa lead/zinc ore ground with different grinding media types (Cullinan et al, 1999).

Cullinan (1999) completed more fundamental work examining the effect of the grinding environment on galena flotation. In these tests, 100 grams of Rapid Bay galena was ground in different mills: mild steel mill using forged steel balls; a stainless steel mill using high chrome balls; and a ceramic mill using ceramic balls. The resultant recovery versus time curves are provided in Figure 2, and illustrate that the galena recovery increased markedly as the media type changed from forged to high chrome, to ceramic. The pulp chemistry measurements taken for these experiments are listed in Table 2. While the variations in pH and Eh were comparatively minor, the amount of EDTA extractable iron decreased dramatically as the grinding media type was changed to more inert materials. This corresponded to an increase in the galena flotation response, again suggesting that the corrosion products on the grinding media impact on flotation response. TABLE 1 Pulp chemical measurements for Mount Isa lead/zinc ore ground with different grinding media types (Cullinan et al, 1999). Media type Forged steel High chrome
100.0 90.0 80.0

pH 7.8 8.2

Eh, mV (SHE) 72 275

Per cent EDTA extractable Pb Zn Fe 1.7 0.053 1.11 3.1 0.071 0.71

Mass recovery, %

70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0

Flotation time, minutes

Forged steel High chrome Ceramic

FIG 2 - Mass recovery versus flotation time curves for rougher flotation tests completed on 100 grams of Rapid Bay galena ground to a P50 of nine microns using forged steel, high chrome, and ceramic media. The tests were completed in demineralised water (pH 7), and employed 1000 grams per tonne of sodium ethyl xanthate collector.

Samples of galena were collected after grinding in the three different milling environments, and their surfaces examined using x-ray photoelectron spectroscopy (XPS). The atomic concentrations of oxygen, lead, iron and sulfur for galena particles ground using the three different grinding media are listed in Table 3. There is a marked difference in the surface concentration of iron. That is, it decreased from 16.6 per cent for the forged steel case to below the detection limit for ceramic grinding media. The XPS spectral data (Cullinan, 1999) suggested that the iron present on the surface occurred as oxy-hydroxide species (ie Fe(OH)3, FeOOH, Fe2O3, and Fe3O4). When the sample was ion etched, to determine the thickness of these oxidised iron surface layers, there was no appreciable alteration in the Fe2p spectra or the atomic concentration, which suggested that these layers were relatively thick. The surface concentration of these species decreased as the grinding media became less reactive, and resulted in increasing exposure of the lead sulfide surface. These findings were consistent with the EDTA extraction data provided in Table 2. TABLE 2 Pulp chemical measurements for Rapid Bay galena ground with different grinding media types (Cullinan, 1999). Media type Forged steel High chrome Ceramic pH 5.5 5.5 5.0 Eh, mV (SHE) 300 300 370 Per cent EDTA extractable Fe 1.24 0.16 0.02

TABLE 3 Composition, determined via XPS, of the un-etched surfaces of Rapid Bay galena ground with forged steel, high chrome and ceramic grinding media (Cullinan, 1999). (Note: The data was normalised to remove the percentage of surface carbon.) Media type O Forged steel High chrome Ceramic 53.1 50.0 33.6 Atomic composition, % Pb Fe 15.6 16.6 20.6 10.2 32.0 <0.1 S 14.7 19.2 34.4

These results suggest that the iron debris from forged steel grinding media can have a significant deleterious effect on the flotation performance of fine (-3 micron) galena. Other work (Matthews, 2002) confirms this, and clearly showed that ferric hydroxide has a depressing effect on galena flotation. By changing the media type to an inert material (ie ceramic grinding media) the galena flotation response can be markedly improved. It is postulated that similar effects would be noted for other sulfide minerals. Cullinans work (1999) and the efforts of Peng (2003) provide strong evidence that the iron contamination on the surfaces of minerals ground with forged steel emanate from the grinding media. Further, the corrosion products from the galvanic couple between the forged steel grinding media and the sulfide minerals within the ore drive the pulp chemistry of the system. Therefore, changing the grinding media to a less reactive material should have positive implications for pulp chemistry and flotation behaviour. That is, conditioning the particle surfaces during grinding, by using the most appropriate grinding media may have significant positive ramifications on down stream processing.

METHODOLOGY In the past, many laboratory studies have not used plant-operating conditions during testing. This has led to a suspicion of laboratory results. To avoid this complication a new tool, the Magotteaux Mill, has been developed (Greet, et al; in press). The Magotteaux Mill (Figure 3) allows the researcher to generate a product in the laboratory that has nominally the same physical properties (particle size distribution) and pulp chemical properties (Eh, pH, dissolved oxygen, oxygen demand and EDTA extractable iron) as an equivalent sample taken from the plant. This is achieved by grinding an appropriate sample to achieve the particle size distribution of the flotation feed, and manipulating the pulp chemistry, by purging the system with gas, so that it matches the grinding mill discharge.

FIG 3 - Schematic representation of the Magotteaux Mill. The experimental strategy adopted to achieve the desired outcomes is completed in three phases: Phase 1 - Plant data collection: The collection of plant data is vital to the success of the test program, for this data forms the basis of the calibration process by defining the target parameters. This initial step involves: o The completion of a pulp chemical and EDTA survey of the grinding and adjoining flotation circuit; o Determination of the oxygen demand at strategic points within the circuit; o The completion of a metallurgical survey; o The collection of conditioned flotation feed for laboratory testing; and o The collection of a bulk sample of the grinding circuit feed for further testing. It is important to note that the metallurgical survey must include both a down-thebank survey of the flotation stage immediately following grinding, and a block survey of the plant to determine overall metallurgical performance.

Phase 2 Magotteaux Mill calibration: The data collected in Phase 1 essentially describes the circuit under consideration, and provides targets for the Magotteaux Mill calibration. The calibration process uses the same grinding media as the operating plant. The objective of the calibration process is to produce a laboratory mill discharge which has the same particle size distribution as the conditioned flotation feed, and the pulp chemistry of the plant grinding mill discharge. To achieve this match involves careful manipulation of the dissolved oxygen, Eh and grinding time, such that all the measured parameters line up when grinding the bulk sample collected during the metallurgical survey. This task is not trivial. Once the Magotteaux Mill is calibrated, oxygen demand and flotation tests are completed on the ground ore. These data are compared with the results of tests conducted on the conditioned flotation feed. If they are the same, a match is achieved.

Phase 3 Media testing: With the Magotteaux Mill calibrated, alternative grinding media are substituted into the mill for testing. The procedure determined during the calibration process for the current grinding media is then applied while grinding the bulk sample employing the alternative grinding media. In this way, it is possible to measure changes in pulp chemistry and flotation response. The changes observed are attributed to the variations in grinding media composition, as the only intentional parameter being changed in the test is the grinding media.

When the flotation results of tests completed on the bulk sample prepared in the Magotteaux Mill using the current grinding media are correlated with the metallurgical survey of the plant, it is possible to determine the scale-up between the laboratory and the plant. Knowing this relationship (scale-up), and the laboratory flotation response of the bulk sample when ground with high chrome grinding media, a prediction of the plant performance can be made if the alternative grinding media were to be used. Thus, this approach provides a robust laboratory methodology that firstly provides a link between the laboratory and the plant using the existing grinding media and secondly investigates changes to the pulp chemistry and metallurgical outcome when high chrome grinding media is substituted into the grinding circuit. Further, it provides a means of estimating plant performance should the best high chrome grinding media be installed in the plant. If used correctly, this strategy will provide an excellent screening procedure that will yield valuable information about the best grinding media for the operation. This should focus pilot plant and plant trials on only the most promising of grinding media types. CASE STUDIES A considerable amount of work has been completed in recent times employing the above approach. Three case studies have been selected to illustrate the success of this methodology. Lead/zinc Work was completed at a lead/zinc mine examining the effect of high chrome grinding media, employed during lead regrinding, on pulp chemistry and flotation performance. Plant data was collected, and subsequently used to calibrate the Magotteaux Mill using fresh lead

regrind circuit feed. The target parameters and the results achieved using the laboratory mill are listed in Table 4. With the laboratory mill calibrated a series of tests were performed using forged, 15, 21, and 32 per cent chrome grinding media. The pulp chemical data of the Magotteaux Mill discharge and flotation feed for each grinding media type are listed in Table 5. An Eh-pH diagram comparing the effect of grinding media on pulp chemistry changes during laboratory flotation tests is given in Figure 4. TABLE 4 Magotteaux Mill calibration data for a lead/zinc ore: targets and results.

Parameter Size distribution P80 %-38 microns Pulp chemistry pH Eh, mV (SHE) DO, ppm % EDTA Fe
400

Plant NA 94.1 7.85 -141 0.61 0.85

Range NA 1.5 0.50 30 0.20 0.15

Magotteaux Mill NA 94.9 8.30 -145 0.00 0.87

Match Yes Yes Yes No Yes

300 3 200 Eh, mV (SHE) 2 3 100 1

0 7.0 -100 7.5 1. Mill Discharge 2. Pb Clr Feed 3. Pb Clr Tail 8.0 1 2 8.5 9.0 9.5 10.0

-200

-300 pH Forged Average 15% Cr Average 21% Cr Average 32% Cr Average

FIG 4 - Eh-pH curves for laboratory grinding and flotation tests conducted on fresh lead regrind circuit feed ground with forged, 15, 21 and 32 per cent chrome grinding media. Table 5 and Figure 4 indicate that changing the grinding media from forged steel to high chrome resulted in an increase in Eh to more oxidising potentials, and a shift in pH to more basic levels. The dissolved oxygen content of the pulp also increased, which corresponded to

a decrease in the percentage EDTA extractable iron. There were subtle differences in the pulp chemistry for each of the high chrome alloys tested, with the pulp becoming more oxidising as the chrome content is increased. It is postulated that these changes in pulp chemistry are strongly related to the corrosion mechanisms encountered for each grinding media type. The Eh-pH curves profiling the grinding and flotation process of the laboratory tests are shown in Figure 4, and provide an excellent indication of where reactions are occurring. From the Nernst Equation 1 there is a dependence of redox potential on pH:

E = Eo +

0.059 log10 n

aRe ac tan ts a Pr oducts

(1)

TABLE 5 Pulp chemical data for Magotteaux Mill discharge and flotation feed for laboratory tests conducted on fresh lead regrind circuit feed samples ground with forged, 15, 21 and 32 per cent chrome grinding media.
Media Magotteaux Mill discharge pH Eh, mV (SHE) DO, ppm 0.00 -145 8.30 3.41 189 8.89 4.22 187 8.82 1.60 194 8.65 Flotation feed pH Eh, mV (SHE) DO, ppm EDTA Fe 0.87 0.00 -182 7.92 0.44 2.91 192 8.38 0.44 3.38 191 8.33 0.40 0.55 190 8.14

Forged 15% Cr 21% Cr 32% Cr

Applying the Nernst equation to water results in a Pourbaix diagram that describes three domains, separated by lines of equilibria. The upper most of these is the water-oxygen line (Equation 2), above which water decomposes and oxygen is evolved, and below which water is stable:
EO2 = + 1.23 + 0.015 log10 pO2 0.059 pH

(2)

This can be simplified further (Johnson, 1988; Natarajan and Iwasaki, 1973) for an oxygenated aqueous solution with no well defined redox couples to (Equation 3):
EO2 = + 0.9 0.059 pH

(3)

What does this mean in terms of chemical reactions that occur in dilute aqueous solutions? In broad terms, if the changes in Eh and pH result in a line parallel to the water-oxygen line this means that water equilibria is being maintained. That is, any change in Eh is directly proportional to a change in pH with a similar relationship to that expressed in Equation 3. If the changes in Eh and pH result in a line that is perpendicular to the water-oxygen line then the evidence suggests that oxidative reactions are occurring. In the case of the forged steel grinding media the Eh-pH curve (Figure 4) is perpendicular to the water-oxygen line, and indicating that this regime is very reactive. It is assumed that the dominant reactions occurring are the corrosion of the forged grinding media (as evident by the higher percentage EDTA extractable iron value), and the oxidation of sulfide minerals, probably pyrite (due to the reduction in pH to more acid levels). The Eh-pH curves for the

three high chrome alloys tested are approximately parallel to the water-oxygen line, suggesting that these systems are comparatively inert with few oxidative reactions occurring. Standard laboratory rougher, rate flotation tests were completed on fresh lead regrind circuit feed ground in the Magotteaux Mill with forged, 15, 21, and 32 per cent chrome grinding media. The lead grade/recovery curves for theses tests are provided in Figure 5. The lead and diluent recovery, at 85 per cent lead recovery, are given in Table 6. The pulp chemical changes noted above had a positive impact on galena flotation response. That is, at 85 per cent lead recovery, there is an increase of 1.5 per cent lead grade between forged and 21 per cent chrome media. The increased concentrate grade can be attributed to improved selectivity for galena against chalcopyrite and non-sulfide gangue (Table 6). It is also possible to realise an improvement in lead recovery through the use of high chrome grinding media. That is, at 50 per cent lead grade, there is an increase in lead recovery of over four per cent when using 21 per cent chrome grinding media compared to forged steel (Figure 5).
60.0 55.0 50.0 45.0 40.0 35.0 30.0 50.0

Grade - Pb(%)

Forged 3 15% Cr 1 21% Cr 2 32% Cr 2 60.0 70.0 80.0 90.0 100.0

Recovery - Pb(% )

FIG 5 - Lead grade/recovery curves for laboratory flotation tests conducted on fresh lead regrind circuit feed samples ground with forged, 15, 21 and 32 per cent grinding media. TABLE 6 Lead grade and diluent recovery, at 85 per cent lead recovery, for laboratory flotation tests completed on fresh lead regrind circuit feed samples ground with forged, 15, 21 and 32 per cent grinding media. Media Pb grade, % Ag 83.20 84.17 85.36 85.88 Diluent recovery, % Cu Zn 41.44 74.28 48.27 81.32 43.92 74.32 43.04 82.89 IS 32.54 53.25 32.45 35.52 NSG 28.25 25.54 27.62 14.66

Forged 15% Cr 21% Cr 32% Cr

51.55 48.95 53.07 51.63

These data suggest that a change from forged to high chrome grinding media changed the pulp chemistry of the system such that the Eh was shifted to more oxidising potential, the dissolved oxygen content of the pulp increased and the levels of EDTA extractable iron were decreased significantly. These changes had a positive impact on galena flotation behaviour, with a positive shift in the lead grade/recovery curve, particularly for the 21 per cent chrome alloy.
Copper

A similar study was completed at a copper/gold mine examining the effect of high chrome grinding media, employed during primary grinding, on pulp chemistry and flotation performance. Plant data was collected, and subsequently used to calibrate the Magotteaux Mill using SAG mill feed. The target parameters and the results achieved using the laboratory mill are listed in Table 7.
TABLE 7 Magotteaux Mill calibration data for a copper ore: targets and results. Parameter Size distribution P80 %-38 microns Pulp chemistry pH Eh, mV (SHE) DO, ppm % EDTA Fe Plant Range Magotteaux Mill Match

195 33 9.5 -75 0.5 1.2

5 2 0.2 20 0.5 0.1

195 31 9.3 -65 0.3 1.3

Yes Yes Yes Yes Yes Yes

With the laboratory mill calibrated, a series of tests were performed using forged, 15, 21, and 32 per cent chrome grinding media. The pulp chemical data of the Magotteaux Mill discharge and flotation feed for each grinding media type are listed in Table 8. An Eh-pH diagram comparing the effect of grinding media on pulp chemistry changes during laboratory flotation tests is given in Figure 6.
TABLE 8 Pulp chemical data for Magotteaux Mill discharge and flotation feed for laboratory tests conducted on SAG mill feed samples ground with forged, 15, 21 and 32 per cent chrome grinding media.
Media Magotteaux Mill discharge pH Eh, mV (SHE) DO, ppm 0.2 -65 9.3 1.2 174 9.0 1.2 205 9.1 1.7 240 9.4 pH 9.5 9.7 9.5 9.4 Flotation feed Eh, mV (SHE) DO, ppm EDTA Fe 1.3 5.0 138 0.2 6.6 195 0.1 6.5 225 0.1 6.1 260

Forged 15% Cr 21% Cr 32% Cr

Table 8 and Figure 6 indicate that changing the grinding media from forged steel to high chrome resulted in an increase in Eh to more oxidising potentials, and the pH was comparatively stable. The dissolved oxygen content of the pulp also increased, which corresponded to a decrease in the percentage EDTA extractable iron. There were subtle

differences in the pulp chemistry for each of the high chrome alloys tested, with the pulp becoming more oxidising as the chrome content is increased. Again, the forged steel grinding media Eh-pH curve (Figure 6) is perpendicular to the wateroxygen line, suggesting that this system is very reactive. It is assumed that the dominant reactions occurring are the corrosion of the forged grinding media (as evident by the higher percentage EDTA extractable iron value), and the oxidation of sulfide minerals. The Eh-pH curves for the three high chrome alloys tested are approximately parallel to the water-oxygen line, suggesting that these systems are comparatively inert with few oxidative reactions occurring.

300
3 2 3 1 1 2

250

200
1

150
Eh, mV (SHE)

100

1. Mill discharge 2. Cu rougher feed 3. Cu rougher tailing

50

0 7.0 -50
1

7.5

8.0

8.5

9.0

9.5

10.0

-100
pH
Forged 15% Cr 21% Cr 32% Cr

FIG 6 - Eh-pH curves for laboratory grinding and flotation tests conducted on SAG mill feed ground with forged, 15, 21 and 32 per cent chrome grinding media.

Standard laboratory rougher, rate flotation tests were completed on SAG mill feed ground in the Magotteaux Mill with forged, 15, 21, and 32 per cent chrome grinding media. The copper grade/recovery curves for theses tests are provided in Figure 7. The copper and gold grades and gold recovery, at 90 per cent copper recovery, are given in Table 9. The pulp chemical changes observed had a positive impact on both copper and gold flotation response. That is, at 90 per cent copper recovery, there is an increase of 1.5 per cent copper grade between forged and 21 per cent chrome media. The increased copper concentrate grade can be attributed to improved selectivity for chalcopyrite against iron sulfides and non-sulfide gangue. The change to high chrome grinding media also had a marked positive influence on gold recovery to copper concentrate (Table 9).

TABLE 9 Copper and gold grades, and gold recovery, at 90 per cent copper recovery, for laboratory flotation tests completed on SAG mill feed samples ground with forged, 15, 21 and 32 per cent grinding media. Media Cu, % Grade Au, ppm Au recovery, %

Forged 15% Cr 21% Cr 32% Cr

23.1 24.5 24.6 15.7

41.5 42.6 41.5 29.4

80.5 87.0 88.6 90.0

However, it is worth noting that the flotation performance of the 32 per cent chrome alloy was inferior to the other grinding media types tested. It is postulated that the poor performance of this alloy may be related to the system being either over oxidised (ie too high an Eh), or over reagentised. It is important to realise that increasing the chrome content of the media further than the optimum grade does not automatically translate into better metallurgy. Such a course of action may in fact result in increased levels of oxygen within the pulp, which may oxidise the sulfide minerals and retard their flotation. The choice of alloy is very much driven by the mineralogy of the system under investigation. Alternatively, it must be recognised that by using high chrome grinding media, the surface chemistry of the particles of interest have been dramatically changed. Therefore, the reagent regime employed when grinding with forged media may no longer be adequate. At the very least, this means a reduction in collector addition. However, it is more likely that another collector may be more appropriate.

30.00

25.00

Grade - Cu(%)

20.00

15.00

10.00

5.00

0.00 40.00

50.00

60.00

70.00

80.00

90.00

100.00

Recovery - Cu(%)
Forged pH 9.5 - Average 15% Cr pH 9.5 - Average 21% Cr pH 9.5 - Average 32% Cr pH 9.5 - Average

FIG 7 - Copper grade/recovery curves for laboratory flotation tests conducted on SAG mill feed samples ground with forged, 15, 21 and 32 per cent grinding media.

Again, these data suggest that a change from forged to high chrome grinding media changed the pulp chemistry of the system such that the Eh was shifted to less reducing potential, the dissolved oxygen content of the pulp increased and the levels of EDTA extractable iron were decreased significantly. These changes had a positive impact on both copper and gold flotation behaviour, with a positive shift in the copper grade/recovery curve, particularly for the 21 per cent chrome alloy.
Refractory gold

A study examining the effect of high chrome grinding media used in primary grinding on pulp chemistry and flotation performance was completed at a refractory gold operation. Plant data was collected, and subsequently used to calibrate the Magotteaux Mill using SAG mill feed. The target parameters and the results achieved using the laboratory mill are listed in Table 10.
TABLE 10 Magotteaux Mill calibration data for a refractory gold ore: targets and results. Parameter Size distribution P80 %-38 microns Pulp chemistry pH Eh, mV (SHE) DO, ppm % EDTA Fe Plant Range Magotteaux Mill Match

204 36 7.0 -215 0.0 1.3

5 2 0.2 20 0.5 0.1

201 35 7.1 -235 0.0 2.3

Yes Yes Yes Yes Yes No

Once the Magotteaux Mill was calibrated a series of tests were performed using forged, 15, 21, and 32 per cent chrome grinding media. The pulp chemical data of the Magotteaux Mill discharge and flotation feed for each grinding media type are listed in Table 11. An Eh-pH diagram comparing the effect of grinding media on pulp chemistry changes during laboratory flotation tests is given in Figure 8.
TABLE 11 Pulp chemical data for Magotteaux Mill discharge and flotation feed for laboratory tests conducted on SAG mill feed samples ground with forged, 15, 21 and 32 per cent chrome grinding media.
Media Magotteaux Mill discharge pH Eh, mV (SHE) DO, ppm 0.0 -235 7.1 0.5 70 7.0 0.5 120 7.0 0.2 105 7.0 Flotation feed pH Eh, mV (SHE) DO, ppm EDTA Fe 2.3 0.1 -28 6.8 1.3 0.7 145 6.6 1.2 0.8 140 6.6 1.2 1.3 145 6.5

Forged 15% Cr 21% Cr 32% Cr

Table 11 and Figure 8 indicate that changing the grinding media from forged steel to high chrome resulted in an increase in Eh to more oxidising potentials, and a marginal shift in pH to more acidic levels. The dissolved oxygen content of the pulp also increased, which corresponded to a decrease in the percentage EDTA extractable iron. There were subtle differences in the pulp chemistry for each of the high chrome alloys tested, with the pulp becoming more oxidising as the chrome content is increased.

In the case of the forged steel grinding media the Eh-pH curve (Figure 8) was again perpendicular to the water-oxygen line, and indicating that this regime is very reactive. It is assumed that the dominant reactions occurring are the corrosion of the forged grinding media (as evident by the higher percentage EDTA extractable iron value), and the oxidation of sulfide minerals, probably pyrite (due to the reduction in pH to more acid levels). The Eh-pH curves for the three high chrome alloys tested are approximately parallel to the water-oxygen line, suggesting that these systems are comparatively inert with few oxidative reactions occurring. Standard laboratory rougher, rate flotation tests were completed on SAG mill feed ground in the Magotteaux Mill with forged, 15, 21, and 32 per cent chrome grinding media. The gold grade/recovery curves for theses tests are displayed in Figure 9. The gold and sulfur grades, and diluent recovery, at 70 per cent gold recovery, are given in Table 12.

300

200
2

3 3 1 1 1

100
Eh, mV (SHE)

0 5.0 5.5 6.0 6.5

2

7.0

7.5

8.0

-100 1. Mill discharge 2. Rougher feed 3. Rougher tailing -200

1

-300
pH
Forged 15% Cr 21% Cr 32% Cr

FIG 8 - Eh-pH curves for laboratory grinding and flotation tests conducted on SAG mill feed ground with forged, 15, 21 and 32 per cent chrome grinding media.

The pulp chemical changes noted above had a positive impact on gold flotation response. That is, at 70 per cent gold recovery, there is an increase of 29 grams per tonne gold grade between forged and 15 per cent chrome media. The increased gold concentrate grade can be attributed to improved recovery of sulfur, and better selectivity for pyrite against non-sulfide gangue (Table 12). It is also possible to realise an improvement in gold recovery through the use of high chrome grinding media. That is, at 200 grams per tonne gold grade, there is an increase in gold recovery of over 11 per cent when using 15 per cent chrome grinding media compared to forged steel (Figure 9).

The 32 per cent chrome alloy again showed inferior flotation performance compared to the 15 and 21 per cent chrome grinding media. The arguments detailed in the copper ore case study apply equally to this ore type. Once more the data suggest that a change from forged to high chrome grinding media changed the pulp chemistry of the system such that the Eh was shifted to less reducing potential, the dissolved oxygen content of the pulp increased and the levels of EDTA extractable iron were decreased significantly. These changes had a positive impact on gold flotation behaviour, with a positive shift in the gold grade/recovery curve, particularly for the 15 per cent chrome alloy.
TABLE 12 Gold and sulfur grades, and diluent recoveries, at 70 per cent gold recovery, for laboratory flotation tests completed on SAG mill feed samples ground with forged, 15, 21 and 32 per cent grinding media. Media Au, ppm Grade S, % Diluent recovery, % S NSG 50.33 55.03 53.53 69.97

Forged 15% Cr 21% Cr 32% Cr

300.00

210 239 213 144

22.46 27.82 26.15 16.07

0.89 0.65 0.77 1.97

250.00

Grade - Au(ppm)

200.00

150.00

100.00

50.00

0.00 20.00

30.00

40.00

50.00

60.00

70.00

80.00

90.00

100.00

Recovery - Au(ppm)
Forged 15% Cr 21% Cr 32% Cr

FIG 9 - Gold grade/recovery curves for laboratory flotation tests conducted on SAG mill feed samples ground with forged, 15, 21 and 32 per cent grinding media. Other sulfide minerals

Similar work has been completed on ores containing platinum group metals and nickel. All demonstrate similar trends to the data presented above.

DISCUSSION

The data clearly demonstrate that for the same ore, treated in the same way, changing the grinding media had a significant impact on the pulp chemistry of the system. That is, the Eh shifted to more oxidising potentials, the oxygen content of the pulp increased and the level of iron species in the pulp decreased. These changes in pulp chemistry had a positive impact on the flotation response of the valuable mineral in each of the case studies. The reason for this positive outcome is directly related to the reactivity of the grinding media, and how the corrosion products of the grinding media interact with the sulfide minerals present in the ore. Forged steel grinding media, which has a rest potential considerably lower than the high chrome grinding media and sulfide minerals, behaves as the anode and corrodes, releasing iron oxy-hydroxy species into the pulp. As discussed by Johnson (2002), these species do alter the flotation response of sulfide minerals. Changing the grinding media to something more electrochemically inert minimises media corrosion, and subsequently reduces the levels of iron oxy-hydroxy contaminating the system. This reduction in contamination improves flotation behaviour as illustrated in the case studies above. Thus, by changing the grinding media it is possible to condition the pulp during grinding to achieve the best pulp chemistry for flotation. Importantly, this work remains applicable should high pressure grinding rolls (HPGR) supersede SAG milling, as the foremost candidate for this approach is the primary ball mill following either of these unit processes.
CONCLUSIONS

In all instances, changing from forged steel to high chrome grinding media had an impact on the pulp chemistry by: Shifting the Eh to more oxidising potentials, Increasing the level of oxygen in the pulp, and Reducing the amount of EDTA extractable iron in the pulp.

These phenomena had a positive impact on the flotation behaviour of the valuable minerals, although it must be noted that the highest chrome content of the grinding media does not always give rise to the best metallurgy. The optimum alloy for an ore is dependant on the mineralogy of the system.
FURTHER WORK

Based on results of this nature, a number of plant trials investigating the effect of high chrome grinding media on pulp chemistry and flotation response are currently underway. It is Magotteauxs intention to monitor the progress of these trials both in terms of pulp chemistry and metallurgical performance. With these data, it is anticipated that a more thorough evaluation of the economic performance of high chrome grinding media can be obtained, that is, the impact on grinding media wear, metallurgical performance and reagent consumption.

ACKNOWLEDGEMENT

The authors wish to thank Magotteaux for granting permission to publish this paper.
REFERENCES

Cullinan, V J, 1999. Improving the Flotation Response of Fine Galena, PhD Thesis, University of South Australia. Cullinan, V J, Grano, S R, Greet, C J, Johnson, N W and Ralston, J, 1999. Investigating Fine Galena Recovery Problems in the Lead Circuit of Mount Isa Mines Lead/Zinc Concentrator, Part 1: Grinding Media Effects, Minerals Engineering, 12(2):147-163. Greet, C J and Freeman, M, 2000. Back to basics The way forward at Pasminco Rosebery Mine, in Proceedings Seventh Mill Operators Conference, pp 115-122 (The Australasian Institute of Mining and Metallurgy: Melbourne). Greet, C J and Smart, R StC, 2002. Diagnostic Leaching of Galena and its Oxidation Products with EDTA, Minerals Engineering, 15:515-522. Greet, C J, Small, G L, Steinier, P and Grano, S R, (in press). The Magotteaux Mill: Investigating the Effect of Grinding Media on Pulp Chemistry and Flotation Performance, Minerals Engineering. Iwasaki, I, Reid, K J, Lex, H A and Smith, K A, 1983. Effect of Autogenous and Ball Mill Grinding on Sulphide Flotation, Mining Engineering, pp 1184-1190. Jackson, B R, Sutherland, D N, Gottlieb, P and Quittner, V H, 1989. QEM*SEM: The Mineralogical Assessment System for the Metallurgist, Base Metals Conference India, pp 471-483. Johnson, N W, 1987. Applications of Kinetics and Liberation Data to Analysis of an Industrial Flotation Process, CIM Bulletin, 80(899):113-117. Johnson, N W, 1988. Application of Electrochemical Concepts to Four Sulphide Flotation Separations, in Proceedings Electrochemistry in Mineral and Metal Processing II, pp 131149. Johnson, N W, 2002. Practical Aspects of the Effect of Electrochemical Conditions in Grinding Mills on the Flotation Process, in Proceedings Flotation and Flocculation: From Fundamentals to Applications (Eds: J Ralston, J Miller and J Rubio), University of South Australia, pp 287-294. Matthews, E, 2002. Control of Iron Hydroxide Slime Coatings in Galena and Sphalerite Mineral Systems, Master of Applied Science Thesis, University of South Australia. Natarajan, K A and Iwasaki, I 1973. Practical Implications of Eh Measurements in Sulphide Flotation Circuits, AIME Transactions, 256:323-328.

Natarajan, K A and Iwasaki, I, 1984. Electrochemical Aspects of Grinding Media-Mineral Interactions in Magnetite Ore Grinding, International Journal of Mineral Processing, 13:5371. Peng, Y, 2003. Grinding environment Studies in the Control of Oxidation and Interactions Between Sulphide Minerals and Grinding Media, PhD Thesis, University of South Australia. Rumball, J A and Richmond, G D, 1996. Measurement of Oxidation in a Base Metal Flotation Circuit by Selective Leaching with EDTA, International Journal of Mineral Processing, 48:1-20. Yelloji Rao, M K and Natarajan, K A, 1989a. Effect of Galvanic Interaction between Grinding Media and Minerals on Sphalerite Flotation, International Journal of Mineral Processing, 27:95-109. Yelloji Rao, M K and Natarajan, K A, 1989b. Effect of Electrochemical Interactions Among Sulphide Minerals and Grinding Media on Chalcopyrite Flotation, Minerals and Metallurgical Processing, pp 146-151. Young, M F, Pease, J D, Johnson, N W and Munro, P D, 1997. Developments in Milling Practice at the Lead/Zinc Concentrator of Mount Isa Mines Limited From 1990, in Proceedings Sixth Mill Operators Conference (Ed: D Lauder) pp 3-12 (The Australasian Institute of Mining and Metallurgy: Melbourne).