Is Measuring pH Enough? C J Greet1, J Kinal2 and I Mitchell3 1. MAusIMM, Principal Metallurgist, Magotteaux Australia Pty Ltd, Suite 4, 83 Havelock Street, West Perth WA 6005. Email: [email protected] 2. Senior Metallurgist, Magotteaux Australia Pty Ltd, Suite 4, 83 Havelock Street, West Perth WA 6005. Email: [email protected] 3. MAusIMM, Metallurgy Manager, Lionore Lake Johnston Operations. Email: [email protected]

ABSTRACT Many flotation plants use pH to adjust the pulp chemistry of their system to achieve optimum separation. For example, in a sequential lead/zinc circuit the pH during sphalerite flotation is increased to 10.5 or greater to depress galena and pyrite, thereby improving selectivity for sphalerite. However, maintaining a pulp at a particular pH value does not necessarily provide the operator with an abundance of information regarding the chemistry changes occurring within a plant. This paper provides the reader with a number of pulp chemical observations in several plants, showing that the same trends are noted over time; however, the magnitude of the measured values do change. Further, the changes in magnitude are invariably associated with a change in mineralogy. Thus, is measuring pH enough? Can the Eh, dissolved oxygen and temperature be measured at the industrial scale reliably? And, are these measurements useful? INTRODUCTION Most plants operate at a ‘constant’ pH, and the control strategies used to achieve this are many and varied. However in its simplest form, a pH modifier (usually lime) is added to the pulp at some point within the circuit upstream from the pH probe which is linked to a control valve. A pH value, determined experimentally, is selected as the set point. If the measured pH drops below the set point, the control valve opens and more lime is added until the desired pH value is achieved. The control valve then closes until the pH decreases again, when the cycle is repeated. The sole objective of this control strategy is to maintain the pH at a particular value. Unless the pH modifier consumption rate is monitored, the operator does not have any information that would provide some guidance on any changes in the pulp chemistry of the system. For example, if the mineral system under consideration contained varying amounts of pyrrhotite, pH modifier consumption rates would increase as the pyrrhotite content increased because of pyrrhotite oxidation (an acidic reaction). Such a change in the mineralogy, and associated changes in pulp chemistry, would go undetected using the above pH control strategy thereby making it difficult to determine the probable reasons for the deterioration in metallurgical performance. The measurement and control of other pulp chemical parameters (Eh, dissolved oxygen and temperature) in base metal flotation plants is somewhat limited. Examples of instances where instruments have been installed to measure these parameters are: Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

1

• •

•

At Kagara Zinc’s Mount Garnet plant it was proposed to install Eh probes to improve the controlled potential sulfidisation step when treating copper supergene ore (Whittering, 2005). In Pasminco’s Elura plant (now Consolidated Broken Hill’s Endeavour Mine) the dissolved oxygen content of the pulp was measured in the second of two conditioning tanks used to oxygenate the pulp prior to lead rougher flotation to oxidise the iron sulfides (Bojcevski, 2003). In Pasminco’s Rosebery plant (now Zinifex Rosebery Mine) the pulp temperature was monitored during conditioning of copper rougher concentrate with stream prior to copper cleaner flotation. The pulp was heated to nominally 65ºC in order to destroy the collector and deactivate galena prior to copper cleaner flotation (Mwaba, 1998).

In all cases, however, the pulp was treated in some way to achieve a desired set point as recorded by a suitable probe positioned at an appropriate location in the circuit. Further, these examples are rare, and it is not common practice to consider using these parameters as part of a plant operating strategy. It has been recognised for some time that the measurement of Eh may impart considerable information about the sulfide mineral system under investigation. Natarajan and Iwasaki (1973) completed work showing that Eh could be measured reliably in both laboratory and plant slurries. Further, they provided a method for its interpretation. Woods (1976) provides strong argument that all reactions on the surfaces of sulfide minerals are electrochemical in nature and numerous other works (Winter and Woods, 1973; Woods et al, 1990; Woods et al, 1992, etc) have demonstrated that it is possible to use Eh to observe mineral oxidation as well as the adsorption of collector on to sulfide mineral surfaces. However, Woods (1976) does offer a warning note regarding the use of Eh measurements in flotation plants, in that the reading obtained is a mixed potential, and should be viewed with caution. Ralston (1991) provides a review of Eh in understanding sulfide mineral flotation, and cites the example of the dependence on Eh in the collectorless flotation of chalcopyrite (Heyes and Trahar, 1977). Further, it has been recognised that it is possible to separate various copper sulfides (ie chalcocite, bornite and chalcopyrite) and pyrite by operating under different Eh regimes (Richardson and Walker, 1985). However, much of this work remains in the realm of the academic researcher. Woodcock and Jones (1970) were perhaps the first researchers to measure and compare pulp chemical parameters in a number of base metal sulfide concentrators. Essentially they observed very similar trends through the grinding and flotation circuits of the six Australian lead-zinc concentrators studied. That is, the pulp potential became less reducing as the pulp passed through the plant, and the dissolved oxygen content increased from low levels in the grinding circuit to values near saturation in the flotation circuit. Grano et al (1993) noted very similar trends in the grinding/lead rougher-scavenger circuit of the Hilton Concentrator near Mount Isa. Another observation worthy of note relates to the effect of temperature on pyrite recovery in South Africa (O’Connor et al, 1984). This data shows in both the laboratory and the plant that as the temperature of the pulp increased, the recovery of pyrite increased. While the authors’ Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

2

explanations of why pyrite recovery decreased at low temperatures are overly complicated, the observation is remarkable. While the work of many of these authors has been dedicated to determining the optimum pulp chemical conditions for the flotation of various sulfide minerals, perhaps a more immediate and useful step would be to measure these parameters as they occur in their current environment (ie controlled pH, air set points, pulp potentials and dissolved oxygen concentrations). Among the questions to be asked are the following: • • •

Do measurements of Eh, dissolved oxygen and temperature through a circuit provide any meaningful results that could be of practical benefit to an operation? Does measurement of these pulp chemical parameters vary with time? Are there differences in the profiles of these parameters with ore type? EXPERIMENTAL

Pulp chemical measurements A sample of slurry was ‘cut’ from the process stream of interest, and poured into a small beaker. The sample was then stirred gently with the probes in the beaker for nominally two minutes until equilibrium readings were obtained. The Eh, pH, dissolved oxygen, and temperature data were then logged using a TPS 90-FLMV data logger. The logged data was downloaded from the TPS 90-FLMV to a laptop computer where it was manipulated. The Eh was measured using an Ionode platinum redox electrode Model IJ64. Prior to use, the electrode was calibrated using Zobell solution (1:1 solution of Part A and B) to give 231 mV, and all Eh values are expressed relative to the standard hydrogen electrode. The platinum electrode was cleaned periodically to maintain it in a bright condition. The pH was measured using an Ionode pH electrode Model IJ63 combined glass electrode, and was calibrated using standard buffer solutions at pH 7.0 and 10.0. The dissolved oxygen content of the pulp was measured using a YSI-5739 dissolved oxygen sensor. The instrument was calibrated in a 20 g/l solution of sodium sulfite for the zero calibration, and in air. A thermocouple appropriately calibrated was used to record the pulp temperature. EDTA extractions EDTA extractions were completed on the same process streams as the pulp chemistry. Each stream was ‘cut’ and the sample poured into a small beaker. The beaker containing the sample was taken back to the laboratory and the wet weight recorded. EDTA extractions were then performed as follows: • •

A 25 millilitre aliquot of slurry was syringed from each of the plant samples, which had been homogenised prior to sampling. The syringed sample was weighed.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

3

• • • • •

Contents of the syringe were injected into a 400 millilitre beaker containing 250 millilitres of three per cent (by weight) EDTA solution, pH modified to 7.5 with sodium hydroxide. The slurry and EDTA solution were thoroughly mixed for five minutes using a magnetic stirrer. The EDTA/slurry mixture was coarse filtered using a Whatman number 40 filter paper. The filtrate from the coarse filtering was fine filtered using a 0.2 micron millipore filter. The filtered EDTA solution was submitted for assay.

The remainder of the pulp sample was pressure filtered, and the solids dried. The dry solids were weighed and submitted for assay. The percentage of EDTA extractable metal ion calculations follow the methodology developed by Rumball and Richmond (1996). Pulp chemical surveys Pulp chemical surveys of operating plants are conducted while on site to determine the pulp chemical conditions of the circuit. The Eh, pH, dissolved oxygen, and temperature of the following process streams were measured: • • • • •

Cyclone underflow. Ball mill discharge. Cyclone overflow. Conditioned flotation feed. Scavenger tailing.

A field trip to Lionore’s Lake Johnston Operation was completed from 25 January to 7 February 2005. In this instance, pulp chemical surveys were completed daily. One additional process stream was included in these surveys, the flash flotation tailing. In the Lake Johnston circuit, the cyclone underflow feeds flash flotation, and the flash flotation tailing discharges in to the ball mill for further grinding. Assays Assaying of solution and solids samples is generally completed by the on site laboratory. In the Lake Johnston case, the elements assayed were nickel, copper, iron, sulfur and MgO. Element to mineral conversions The following assumptions were made when considering these conversions: • • •

All the nickel present occurred as pentlandite (Pn). All the copper present occurred as chalcopyrite (Ch). The sulfur not associated with pentlandite and chalcopyrite, occurred as iron sulfides (IS).

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

4

• •

The iron sulfides were present as pyrite (Py) and pyrrhotite (Po) - one third and two thirds respectively (Clay, 2005). All remaining minerals were lumped into one category called non-sulfide gangue (NSG).

Thus, the percentage of nickel sulfide is given by: Nickel sulfide (%) = % Ni x 2.7176. The amount of sulfur attributed by the nickel sulfides is given by: S in nickel sulfide = % Ni x 2.7176 x 0.3323. The percentage of copper sulfide is given by: Copper sulfide (%) = % Cu x 2.8879. The amount of sulfur attributed by the copper sulfides is given by: S in copper sulfide = % Cu x 2.8879 x0.3494. The amount of iron sulfide (one third pyrite, two thirds pyrrhotite) present is calculated using: Iron sulfide (%) = 2.4517 x (% S – (SPn + SCh)). The non-sulfide gangue assay is determined using: Non-sulfide gangue (%) = 100 – (% Pn + % Ch + % IS). It should be noted that a separate MgO assay was provided, which could be used to determine the percentage of non-sulfide gangue that was not associated with MgO. RESULTS Eh profile through grinding and rougher flotation circuits Examples of Eh profiles through copper/gold, lead/zinc, nickel and PGM grinding and rougher flotation circuits are provided in Figures 1 to 4.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

5

250

Eh, mV (SHE)

200 150 100 50

R ou gh er Fe ed

O ve rfl ow C yc lo ne

D /C Ba ll M ill

C yc lo ne

U nd er flo w

D /C M ill SA G

R ou gh er Ta il

November 2004

0

Circuit position Figure 1: The Eh profile through primary grinding and rougher flotation circuits of a copper/gold operation. 250

1999

200

Jun-03

Oct-05

Eh, mV (SHE)

150 100 50 0 -50 -100 -150 -200

2º Pb

R

od

m

ill

di s

ch

ar g

e

ro ug he rf Pb ee d sc av ta Zn ilin ro g ug he rf Zn ee d sc av ta ilin g

-250

Circuit position Figure 2: Eh profiles through primary grinding and rougher flotation circuits of a lead/zinc operation. Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

6

350

Eh, mV (SHE)

300 250 200 150 100 50

Sc av en ge rt ai lin g

C

on di ti

on ed

flo at

fe ed

ov er flo w yc lo ne

m il l ba ll

1º c

di sc ha rg e

ta i flo at 1º

1º c

yc

Fl as h

lo ne

un de rfl o

lin g

w

0

Circuit position Survey 27 April 2006

Figure 3: The Eh profile through primary grinding and rougher flotation circuits of a nickel operation. 250

150 100 50

Cylpebs Jun-03 Balls Jun-03 Fin al tai lin g

F lo tat ion fee d

rflo w Cy clo ne ov e

ha rge mi ll d isc SA G

cl o ne un de rflo w

0

Cy

Eh, mV (SHE)

200

Circuit position Figure 4: The Eh profile through primary grinding and rougher flotation circuits of a PGM operation. Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

7

In general, the Eh profiles in Figures 1 to 4 exhibit very similar trends despite coming from very different sources. That is, the Eh decreased by between 100 and 150 mV, becoming more reducing as the pulp passed from the primary cyclone underflow to the primary ball mill discharge. The decrease in Eh can be attributed to corrosion of the forged grinding media. The Eh then increased (becoming less reducing) in the cyclone overflow, and continued to increase to more oxidising pulp potentials in the rougher flotation circuit. The increase in Eh across the rougher flotation circuit can be attributed to the use of air as the flotation gas. These trends are nominally the same as those observed by Woodcock and Jones (1970) and Grano et al (1993). While the Eh profiles are remarkably similar it is abundantly clear that there are significant differences in the magnitude of the Eh values. For example, the ball mill discharge Eh for the copper/gold ore was 75 mV (SHE) compared with -35 mV (SHE) for the lead/zinc ore measured in October 2005 (which compares with -215 mV (SHE) for the 1999 data). These variations in the magnitude of the Eh can be explained in terms of mineralogical differences, and operating pH values. The total sulfide content of the copper/gold ore is about five per cent compared with nominally 20 per cent for the lead/zinc ore, thereby making the lead/zinc ore more reactive. Further, the grinding circuit for the copper/gold ore is operated at pH 12, a pH regime where grinding media corrosion is significantly retarded, while the lead/zinc operation has a considerably lower pH of 8.5 where corrosion occurs readily. These differences contribute strongly to the observed variations in the magnitude of the Eh values. The relationship between Eh and dissolved oxygen The dissolved oxygen profile through the grinding and rougher flotation circuits of the copper/gold operation is given in Figure 5. It is interesting to compare the Eh profile in Figure 1 with the dissolved oxygen data collected from the same process streams (Figure 5). The dissolved oxygen data portrays a very similar trend. As with the Eh profile, the dissolved oxygen content of the pulp decreases as the pulp passed from the primary cyclone underflow to the primary ball mill discharge. Again, the decrease can be attributed to corrosion of the forged grinding media. The dissolved oxygen concentration then increases in the cyclone overflow, and continued to increase during rougher flotation. The increase in dissolved oxygen levels in the rougher flotation circuit can be attributed to the use of air as the flotation gas. Similar observations were made for the lead/zinc, nickel and PGM operations described above. Further, these trends are similar to those noted by Woodcock and Jones (1970) and Grano et al (1993).

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

8

8.0 7.0

DO, ppm

6.0 5.0 4.0 3.0 2.0 1.0

November 2004 ai l

ee d

ou gh er T R

ve r O ne C yc lo

R ou gh er F

flo w

/C D ill Ba ll M

U ne yc lo C

SA G

M

ill

nd er flo w

D /C

0.0

Circuit position Figure 5: The dissolved oxygen profile through the grinding and rougher flotation circuits of a copper/gold operation. The Lake Johnston concentrator Lionore owns and operates two mines, Maggie Hayes and Emily Ann, at its Lake Johnston Operations. The ore from each mine is campaign treated through the concentrator. The run of mine ore is stage crushed, before reporting to a fine ore bin. The fine ore is ground in a primary ball mill in closed circuit with cyclones to the desired P80 (85 microns for Maggie Hayes ore and 120 microns for Emily Ann ore). Lime is added with the fresh feed to the primary ball mill to achieve a pH of 8.5 in the flotation feed. The cyclone underflow passes through a flash flotation unit to scalp out the coarse liberated pentlandite before discharging the flash flotation tailing back to the primary ball mill. At the time this work was conducted, the flash flotation concentrate reported to final concentrate. The cyclone overflow is conditioned with collector and guar (to depress MgO minerals) prior to rougher, middling and scavenger flotation. The rougher/middling/scavenger concentrates are cleaned in two stages to produce a recleaner concentrate which is combined with the flash flotation concentrate to form the final concentrate. Scavenger tailings report to the tailings dam. At the time of surveying the plant was treating Emily Ann ore, producing a 14 per cent nickel grade at 74 per cent nickel recovery.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

9

The Eh profile through the Lake Johnston grinding rougher flotation circuit The Eh profiles through the primary grinding/rougher flotation circuit for the survey completed in January/February 2005 are displayed in Figure 6. Essentially, the Eh trend was the same as that described above. That is, the Eh decreases as the pulp is ground in the primary ball mill, then is increased through flotation. The increase in Eh can be attributed to the use of air as the flotation gas. However, the magnitude of the Eh varied considerably, ranging between -110 and 85 mV (SHE) in the primary ball mill discharge. The data in Figure 6 can be simplified by plotting the Eh of the ball mill discharge against time (Figure 7). This suggests that the Eh in the ball mill discharge decreased with time. Also appearing in Figure 7 is the pH of the ball mill discharge for the same period of time. It is apparent that while the pH did change slightly for this period (varying between 8.0 and 8.5), it did not shift dramatically away from the target. Therefore, the change in Eh was not due to a change in the pH.

250

Eh, mV (SHE)

200 150 100 50 0 -50 -100

g Sc

on ed

av e

flo

ne rt ai lin

at fe ed

rfl ow on di ti C

C yc lo

ne

di s

ov e

ch ar ge

g lin ta i as h Fl

Ba ll m ill

C

yc lo ne

un de rfl

ow

-150

Circuit position 27.01.2005

28.01.2005

29.01.2005

30.01.2005

31.01.2005

02.02.2005

03.02.2005

04.02.2005

05.02.2005

Average

01.02.2005

Figure 6: The Eh profiles through the primary grinding/rougher flotation circuit of the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

10

100

9.0

50

8.5

0

8.0

-50

7.5

-100

7.0

-150

6.5

pH

9.5

27 .0 1. 20 05 28 .0 1. 20 05 29 .0 1. 20 05 30 .0 1. 20 05 31 .0 1. 20 05 01 .0 2. 20 05 02 .0 2. 20 05 03 .0 2. 20 05 04 .0 2. 20 05 05 .0 2. 20 05

Eh, mV (SHE)

150

Time, day Eh - Ball mill discharge

pH - Ball mill discharge

Figure 7: The Eh and pH of the ball mill discharge versus time for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005. The Eh-pH curve of the grinding and rougher flotation circuit for 27 January and 5 February (the two extremes in the current data set) are presented in Figure 8 to determine where the 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   aPr oducts

  

(1)

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 and Natarajan and Iwasaki, 1973) for an oxygenated aqueous solution with no well defined redox couples (Equation 3): EO2 = + 0.9 − 0.059 pH

(3)

What does this mean in terms of chemical reactions that occur in dilute aqueous solutions?

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

11

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. 300 250 6

Eh, mV (SHE)

200

2

150 6

100 50 1. 2. 3. 4. 5. 6.

0 -50 -100

1

5

3

4

5

Cyclone underflow; Flash flotation tailing; Ball mill discharge; Cyclone overflow; Rougher feed; and Scavenger tailing.

4 2 1

3

-150 7.0

7.5

8.0

8.5

9.0

9.5

10.0

pH 27.01.2005

05.02.2005

Figure 8: The Eh-pH curves for the grinding and rougher flotation circuits for 27 January and 5 February when the Lake Johnston concentrator was treating Emily Ann ore.

An examination of this data (Figure 8) suggests that the ore treated at the beginning of the sampling campaign (27 January) was different from that processed at the end (5 February). In gross terms, through the primary grinding circuit (points 1 to 4) the changes in Eh and pH are almost parallel to the water-oxygen line, indicating maintenance of water equilibrium for the survey completed on 27 January. For the 5 February data (points 1 to 4), the changes in Eh and pH are perpendicular to the water-oxygen line suggesting that oxidative reactions are occurring. It is speculated that these oxidative reactions are grinding media corrosion and sulfide mineral oxidation. This data suggests that the reactivity of the ore during grinding has changed over the sampling period. However, it was interesting to note that during the flotation stage the changes in Eh and pH (points 5 and 6) for both data sets were nominally parallel to the water-oxygen line indicating maintenance of water equilibrium. Thus, it is suggested that most of the significant reactions are occurring during grinding. Simultaneously to collecting the pulp chemical data, samples of the same process streams were collected from the plant and EDTA extractions completed. The percentage of EDTA extractable iron (calculated) in the cyclone overflow is plotted against time in Figure 9 for the January/February 2005 period. Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

12

EDTA extractable Fe, %

1.5

1.2

0.9

0.6

0.3

05 20 .0 2.

05

04

.0 2.

20

20 03

.0 2.

20 .0 2. 02

05

05

05

05 20 .0 2.

01

.0 1. 31

.0 1.

20

20

05

05

05 30

29

.0 1.

20

20 .0 1. 28

27

.0 1.

20

05

05

0.0

Time, days Figure 9: The percentage EDTA extractable in the cyclone overflow versus time for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005.

The EDTA extractable iron calculated data provides an excellent measure of the corrosion of the grinding media. These data indicate that the percentage of EDTA extractable iron produced in the primary grinding circuit increased over the sampling period. At the beginning of the sampling period (27 January) the percentage EDTA extractable iron was 0.49 per cent which increased to 0.84 per cent by 5 February. The decrease in Eh and the increase in the amount of EDTA extractable iron suggests that the pulp became more reactive during the sampling period. One possible reason for this may be the elevated iron sulfide content of the ore during this period. Figure 10 plots the iron sulfide level in the cyclone underflow for the sampling period. During the first seven days of the test, the iron sulfide content of this process stream averaged 49 per cent while over the last three days of testing, the iron sulfide percentage increased to 69 per cent. This significant increase in iron sulfide concentration would result in elevated grinding media corrosion, and lead to a decrease in the Eh and an increase in the EDTA extractable iron (Figure 11).

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

13

Iron sulphide content, %

90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0

27 .0 1. 20 05 28 .0 1. 20 05 29 .0 1. 20 05 30 .0 1. 20 05 31 .0 1. 20 05 01 .0 2. 20 05 02 .0 2. 20 05 03 .0 2. 20 05 04 .0 2. 20 05 05 .0 2. 20 05

0.0

Time, days

150

1.4

100

1.2

50

0.9

0

0.7

-50

0.5

-100

0.2

-150 0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

80.0

EDTA extractable Fe, %

Eh, mV (SHE)

Figure 10: The iron sulfide content in the cyclone underflow versus time for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005.

0.0 90.0

Iron sulphide content, % Eh data

EDTA data

Figure 11: The Eh and percentage EDTA extractable iron versus iron sulfide content in the cyclone underflow data for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

14

14.0

50.0

13.5

40.0

13.0

30.0

12.5

05

.0 2

.0 5 .0 2 04

.0 2

.0 2

.0 5 03

01

.0 2

.0 1 31

02

.0 5

.0 5 30

29

.0 1

.0 1

.0 5 .0 1 28

.0 1 27

Ni grade, %

60.0

.0 5

14.5

.0 5

70.0

.0 5

15.0

.0 5

80.0

.0 5

Ni recovery, %

How this change affected metallurgical performance was interesting. An examination of the nickel recovery data (Figure 12) for the sampling period indicated that it increased with time. It is likely that the increase in nickel recovery is related to the recovery of nickeliferous iron sulfides (Clay, 2005). However, as the nickel recovery increased, the nickel concentrate grade decreased. The decrease in nickel concentrate grade can be attributed to an increase in the iron content (Figure 13).

Time, days Ni recovery, %

Ni grade, %

12.5

24.0

Fe grade, %

26.0

05 .0 2. 05

13.0

04 .0 2. 05

28.0

03 .0 2. 05

13.5

02 .0 2. 05

30.0

01 .0 2. 05

14.0

31 .0 1. 05

32.0

30 .0 1. 05

14.5

29 .0 1. 05

34.0

28 .0 1. 05

15.0

27 .0 1. 05

Ni grade, %

Figure 12: The nickel recovery and concentrate grade versus time for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005.

Time, days Ni recovery, %

Ni grade, %

Figure 13: The nickel and iron grades in final concentrate versus time for the Lake Johnston concentrator treating Emily Ann ore from 27 January to 5 February 2005. Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

15

DISCUSSION

While the data set used in this paper is comparatively small and more needs to be collected to confirm the observations made, it is apparent that trends do exist. Further, judicious measurements of Eh in appropriate process streams can reveal useful information about the feed mineralogy as well as chemical reactions that are occurring within the process. While it may be argued that simply measuring the Eh of a process stream is a retro-step, the authors believe that it is a step in the right direction in gaining more control over the separation process. The measurement of Eh, dissolved oxygen and temperature continuously in a base metal sulfide flotation plant are not common practice, but the instrumentation does exist. Yokogawa and Emerson Process Management (to name but two) manufacture these instruments, and they are readily available. The issue then becomes one of where to put it, and what supporting data is needed to make this information work for the plant metallurgist. In this work, the cyclone underflow, ball mill discharge and cyclone overflow have been examined in several plants treating different ore types to gather various pieces of the puzzle. It is recommended that the Eh of the ball mill discharge be recorded on line, as it is in the grinding mill that many chemical reactions occur. However, it is not just a matter of putting a probe in to the ball mill discharge and all operating problems disappear, as there is a need to marry this information with pH measurements, mineralogy (actual or calculated), and EDTA extractable metal ion analyses. Plus, the practitioner should have knowledge of the chemistry of their plant system so that they can appropriately interpret the observations made. Once these relationships have been developed, it should be possible to detect changes in pulp chemistry, relate them to changes in mineralogy and from there, predict potential impacts on plant metallurgy. CONCLUSIONS

Firstly, the pulp chemical trends observed through primary grinding and rougher flotation in several plants are approximately the same regardless of the system under investigation. The differences in the magnitude of the various pulp chemical parameters (pH, Eh, dissolved oxygen, and temperature) are the result of variations in the mineralogy and processes employed. Secondly, the Eh and dissolved oxygen profiles within the primary grinding and rougher flotation circuits tend to mirror one another reasonably closely. Thirdly, the Eh profile measured over a ten-day period in the Lake Johnston concentrator while treating Emily Ann ore showed the same trend, however the Eh varied considerably over the sampling period, ie the Eh became more reducing over the sampling period. The reduction in Eh could be traced to an increase in the iron sulfide content of the ore, which resulted in increased corrosion of the forged steel grinding media. Metallurgically, this resulted in an increase in nickel recovery through the recovery of nickeliferous iron sulfides, and a decrease in nickel concentrate grade. The reduction in nickel concentrate grade was accompanied by an increase in the iron grade in the concentrate.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

16

Finally, commercially available instruments are currently on the market, so completing such an analysis of plant operations on a routine basis is possible. However to be successful, it is necessary to develop models that include use of the relationships between changes in pulp chemistry and changes in mineralogy in order to predict potential impacts on plant metallurgy. ACKNOWLEDGEMENTS

The authors gratefully acknowledge the support of Lionore and Magotteaux in providing them with the opportunity to complete this work, and allowing it to be published. Special thanks are reserved for Djoko Julianto who collected the pulp chemical data during his first visit to Australia. REFERENCES Bojcevski D (2003), personal communication. Clay P (2005), personal communication. Grano S R, Lauder D W, Johnson N W, Sobieraj S, Smart R StC, and Ralston J (1993), Surface Analysis as a Tool for Problem Solving in Sulphide Ore Flotation – A Case Study of the Hilton Concentrator of Mount Isa Mines Limited, in the Proceedings of the Symposium on the Polymetallic Sulphides of the Iberian Pyrite Belt. Heyes G W and Trahar W J (1977), The Natural Floatability of Chalcopyrite, International Journal of Mineral Processing, 4, pp 317 to 344. Johnson N W (1988), Application of Electrochemical Concepts to Four Sulphide Flotation Separations, in the Proceedings of the Electrochemistry in Mineral and Metal Processing II, pp 131 to 149. Mwaba C (1998), personal communication. Natarajan K A and Iwasaki I (1973), Practical Implications of Eh Measurements in Sulphide Flotation Circuits, In AIME Transactions, 256, pp 323 to 328. O’Connor C T, Dunne R C and Botelho de Sousa A M R (1984), The Effect of Temperature on the Flotation of Pyrite, Journal of the South African Institute of Mining and Metallurgy, 84 (12), pp 389 to 394. Ralston J (1991), Eh and Its Consequences in Sulphide Mineral Flotation, Minerals Engineering, 4 (7), pp 859 to 878. Richardson P E and Walker G W (1985), Proceedings of the 15th International Mineral Processing Congress, 11, pp 198. 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 Minerals Processing, 48, pp 1 to 20. Whittering R (2005), personal communication. Winter G and Woods R (1973), The Relation of Collector Redox Potential to Flotation Efficiency: Monothiocarbonates, Separation Science, 8 (3), pp 261 to 267. Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

17

Woodcock J T and Jones M M (1970), Chemical Environment in Australian Lead-zinc Flotation Plant Pulps: I – pH, Redox Potentials and Oxygen Concentration, in the Proceedings of the Australasian Institute of Mining and Metallurgy, 235, pp 45 to 60. Woods R (1976), Electrochemistry of Sulphide Flotation, in Flotation (Edited by: M C Fuerstenau), pp 298 to 333 (AIME: New York). Woods R, Young A C and Yoon R H (1990), Ethyl Xanthate Chemisorption Isotherms and Eh-pH Diagrams for the Copper/Water/Xanthate and Chalcocite/Water/Xanthate systems, International Journal of Mineral Processing, 30, pp 17 to 33. Woods R, Basilio C I, Kim D S, and Yoon R H (1992), Ethyl Xanthate Chemisorption Isotherms and Eh-pH Diagrams for the Silver + Water + Ethyl Xanthate System, Journal Electroanalytical Chemistry, 328, pp 179 to 194.

Metallurgical Plant Design and Operating Strategies (MetPlant 2006) 18 - 19 September 2006 Perth, WA

18