Tuesday, 25 November 2008

Laboratory and Miniplant Studies on Cu/Ni Separation

M. Q. Xu, P. F. Wells

Vale Inco Technical Services Limited, Mississauga, Ontario, Canada

Email: manqiu.xu@valeinco.com



ABSTRACT


During the flowsheet development for a potential commercial Cu/Ni separation plant at the Inco ‘s Sudbury operation, extensive laboratory and miniplant studies were carried out. There are two potential feed streams for Cu/Ni separation at the existing operation. It is shown that Cu/Ni separation from the rougher concentrate is relatively simple without need of regrind. However, Cu/Ni separation can be difficult from the bulk concentrate because of the presence of excess flotation reagents in the bulk concentrate slurry and locked particles. Excess flotation reagents created excess froth increasing entrainment. By replacing the slurry water with tap water and with addition of some hydrogen peroxide, entrainment is minimized and good Cu/Ni separation is achieved on the bulk concentrate. Fine regrind is required to increase liberation of the locked particles for ultimately high copper recovery at the minimal nickel content.



Keywords: Chalcopyrite; Pentlandite; Pyrrhotite; Flotation; Depression









INTRODUCTION


During the flowsheet development for a potential commercial Cu/Ni separation plant at the Inco’s Sudbury operation (Xu, et a!., 2004), a number of parameters that could influence the Cu/Ni separation efficiency were investigated in laboratory and miniplant testwork. These included the Cu/Ni separation circuit configurations, the use of flotation columns as copper cleaner/recleaner in comparison to mechanical cells, and the use of cyanide, TETA or dextrin as Po depressant.



Cu/Ni separation from a concentrate containing chalcopyrite (Cp), pentlandite (Pn), pyrrhotite (Po) and some minor amount of nonsuiphide gangue is to float chalcopyrite while depressing pentlandite, pyrrhotite and gangue minerals (Agar, 1991). No collector is required to float chalcopyrite, but aeration and high lime conditioning is required to depress pentlandite. A number of reagents was evaluated for pyrrhotite depression, including cyanide, triethylenetetramine (TETA) and dextrin. It was found that TETA in combination with sodium sulfite or sodium metabisulfite is very effective in depressing pyrrhotite at high pH. The present paper describes the results from the laboratory study on two concentrate streams obtained from Inco’s Clarabelle mill and some miniplant resuits on composite ores are also included.







EXPERIMENTAL


Feed materials


Three types of feed materials were used in the present study. As shown in Table 1, the first two materials were rougher and bulk concentrates taken from Inco’s Clarabelle mill for batch testwork and the third feed was a Clara-belle mill composite ore for continuous mini- plant testwork. There are two potential feed streams for Cu/Ni separation at the existing operation. Which concentrate stream is to be used for the commercial plant will be dependent on how much copper recovery is desired, since the rougher concentrate contains about 80% of total copper from ore, while the bulk concentrate contains up to 95% of the copper. The composite ore is used in miniplant work to confirm the laboratory batch results.





Laboratory flowsheet


The laboratory flowsheet is shown in Fig. 1. It consists of the five stages: high lime aeration conditioning, copper rougher (sometimes called Cu/Ni separator) in mechanical cells, regrind of copper rougher concentrate, and copper cleaner! recleaner in columns. Flotation columns for Cu/Ni separation has been successfully applied at the Falconbridge’s Strathcona mill (Heinrich, et al., 1995; Kelebek, et al., 1996; Kelebek, et al., 1997). The regrind of copper rougher concentrate is not necessary for Cu/Ni separateon on rougher concentrate with a good degree of liberation, while it is required on bulk concentrate which contains a considerable amount of locked Cp/Pn and Cp/Po particles (Agar, 1989). In the case of bulk concentrate, the regrind target is to produce particle size 80% passing 40um.





Fig. 1 Laboratory flowsheet for Cu/Ni separation

Miniplant flowsheet


The miniplant Cu/Ni separation flowsheet is shown in Fig. 2. The flowsheet is part of a whole flowsheet starting with composite ore to produce final copper and nickel concentrates, and rock and Po tailings (Xu, et al., 2003). In this case, Cu/Ni separation was done on rougher cleaner concentrate. As compared to the laboratory flowsheet in Fig. 1, the miniplant flowsheet employed an additional copper scavenger stage which allowed high copper recovery possible. Attempts were made by recycling the copper cleaner tails to the Cu/Ni separator feed in order to simplify the flowsheet but it was found that this made the Cu/Ni separator less effective.





Fig. 2 Miniplant Cu/Ni separation flowsheet



RESULTS


Lab results on rougher concentrate


For the laboratory testwork on rougher concentrate, the first set of tests was to determine the time required for the high lime/aeration conditioning. The pH was adjusted to 11.812.0 using lime during aeration conditioning. Copper recovery as a function of nickel grade in the copper concentrate is shown in Fig. 3 for three aeration conditioning times. Aeration conditioning time up to 15 minutes was beneficial but not much beyond 30 minutes conditioning time.





Fig. 3 Cu recovery vs Ni grade in the Cu rougher stage





Once the aeration time was determined, incremental flotation tests were carried out in the copper cleaner stage after 30 minutes aeration conditioning and batch rougher flotation. The main diluent to the copper rougher concentrate was Po and three reagent schemes were tested for Po depression in the copper cleaner. The copper recovery vs Ni grade in copper concentrate is shown in Fig. 4. These tests were conducted in a 2.2 L Denver cell. Copper recovery vs Po recovery is shown in Fig. 5. It is evident that high cyanide dosage is required to effectively depress pyrrhotite. As a consequence, Cp was initially depressed. TETA in combination with sodium metabisulfite (SMBS) or sodium sulfite (Na2SO3) was found most effective in depressing Po at high pH among the three reagents evaluated.





Fig. 4 Cu recovery vs Ni grade in Cu cleaner stage







Fig. 5 Cu recovery vs Po recovery in Cu cleaner stage





Tests were also carried in a column cell and compared to the tests in Denver mechanical cell in Fig. 6. The column with some wash water addition produced better copper concentrate than the mechanical cell.





Fig. 6 Cu recovery vs Ni grade: comparison between column and Denver cells in Cu cleaner stage





Copper recovery as a function of nickel grade in the final concentrate is shown in Fig. 7 for the three flotation stages. With the rougher concentrate of 11.9% Cu, 11.7% Ni, over 99% copper recovery is obtained with the nickel assay below 2% in the copper rougher stage. Copper recovery in the column cleaner is over 95% and the nickel assay is reduced to below

0.8%. Copper recovery in the column recleaner is above 86% with 0.35% Ni in the final copper concentrate. Since the rougher concentrate contains only about 80% of total copper from the ore, the copper recovery to the final copper concentrate from the ore is slightly less 70%.





Fig. 7 Cu recovery vs Ni grade for three flotation stages with rougher concentrate

Lab results on bulk concentrate


Similar tests to rougher concentrate were conducted on bulk concentrate. As described earlier, Cu/Ni separation on bulk concentrate allows more copper removal from the bulk concentrate. While testing the bulk concentrate, it was immediately noted that the water from the bulk concentrate slurry was very frothy. As shown in Fig. 8, replacing the process water with tap water improved Cu/Ni separation. Replacing the process water with tap water and with 1.0 g/kg addition of hydrogen peroxide further improved the Cu/Ni separation in the copper rougher stage.





Fig. 8 Cu recovery vs Ni grade in Cu rougher stage for bulk concentrate



Copper recovery as a function of nickel grade in the concentrate is shown in Fig. 9 for the three flotation stages with bulk concentrate, following a 30 minutes aeration conditioning and regrind of copper rougher concentrate. With the bulk concentrate of 10.5% Cu, 10.9% Ni, over 99% copper recovery is obtained with the nickel assay of 2.9% in the Cu rougher stage. Copper recovery in the column cleaner is over 95% and the nickel assay is about 1.6%. Copper recovery in the column recleaner is above 88.8% with 0.56% Ni in the final copper concentrate. Since the bulk concentrate contains about 95% of total Cu from the ore, the copper recovery to the final copper concentrate from the ore is calculated at 84% at 0.56% Ni grade.





Fig. 9 Cu recovery vs Ni grade for three flotation stage with bulk concentrate



Miniplant results


To further demonstrate the feasibility of Cu/Ni separation, a miniplant campaign was carried out using the Cu/Ni separation flowsheet shown in Fig. 2. In this case, the feed to the Cu/Ni separator was the rougher cleaner concentrate, which contains between 90% and 95% of the total copper in the ore. Copper recovery as a function of nickel content in the copper concentrate from the miniplant results is shown in Fig. 10. The laboratory batch data is also included. Increasing copper recovery from 70% to 85% would be expected to increase the nickel content from 0.3% to 0.55% based on this set of data. The Falconbridge Strathcona operating point fits almost exactly on this curve.





Fig. 10 Copper recovery vs nickel content in the copper concentrate for laboratory and miniplant data



CONCLUSIONS


Cu/Ni separation in the complex sulfide ores usually requires the separation of chalcopyrite, pentlandite and pyrrhotite from each other. The methodology to achieve this separation has been demonstrated where chalcopyrite and pyrrhotite are recovered in a flotation concentrate while depressing pentlandite in the Cu/Ni separator stage. This is accomplished by oxidative conditioning with high lime addition and intensive aeration. The oxidative conditionings removes and decomposes xanthate from mineral surfaces. The chalcopyrite and pentlandite separation in the first stage can be very sensitive to the residual chemical reagents in the process water.



Too much collector addition in the upstream bulk rougher will have a detrimental effect on Cu/Ni separation. This detrimental impact can be minimized with hydrogen peroxide addition into the high lime/aeration conditioning stage but with an overall increase in operating costs.



The copper cleaner, a second separation stage, is mainly to separate chalcopyrite from pyrrhotite. Fine regrind of the separator concentrate is found to be necessary for improving both Cp/Pn and Cp/Po liberation for bulk concentrate. For below 70% copper recovery, rougher concentrate from the existing mill can be used and regrind is not needed. While sodium cyanide can be very effective in depressing Po, its use is not acceptable because of environmental concerns. The use of TETA in combination with sodium sulfite or sodium metabisulfite in this case is found to be more effective in depressing Po. Column flotation in this stage is superior to mechanical cells due to the reduced entrainment of fine particles. The copper re-cleaner stage is required to produce a final copper concentrate with minimal nickel content, which is a critical factor in the sale of the final copper concentrate.



The distribution of the precious group metal (Au, Ag, Pt and Pd) and impurities such as Zn, Pb and As is a concern. Some of PMG are recovered along with the copper concentrate. If the copper concentrate is to be sold, this implies potential loss of revenues. Zn and Pb are about equally split between copper and nickel concentrates according to the mass ratio, while more As remains with nickel concentrate.



Acknowledgements The authors thank the management of Inco Technical Services Ltd. for permission to publish this work. They also wish to thank the ITSL Minerals personnel for the testwork, and A. Lee for the optical photos.



REFERENCES

Agar, GE, 1989. The Separation of Chalcopyrite from Pyrrhotite, in Processing of Complex Ores (editors: GS Dobby and SR Rao), Pergamon, New York, NY, 227-234.

Agar, GE, 1991. Flotation of Chalcopyrite, Pentlandite, Pyrrhotite Ores, International Journal of Mineral Processing,

33:1-19.

Heinrich, GW, Wells, PF, Kelebek, 5, and Whittaker, PJ, 1995. Copper-Nickel Separation Pilot Plant Evaluation of Flow- sheet Options, in Proceedings of 27th Annual Meeting of the Canadian Mineral Processors, Ottawa, 101-118.

Kelebek, 5, Beauchamp, JR. Murphy, DA, Marrs, G, and Macnamara, DD, 1996. Column Flotation Practice in Cu-Ni Separation at Falconbridge, in Column ‘96 (editors CO Gomez and JAFinch), CIM, 119-133.

Kelebek, 5, Wells, PF, Macnamara, DD, Mars, GG, and Heinrich, GW, 1997. Recent Improvements in Column Flotation Circuits at Falconbridge, In Processing of Complex Ores (editors JA Finch, SR Rao and J Holubec), CIM, 33-47.

Xu, M, Quinn, P. and Wells, PF, 2003. A Movable Mineral Processing Miniplant, in Proceedings of 35th annual meeting of the Canadian Mineral Processors, Ottawa, 317-329.

Xu, M and Wells, PF, 2004. Development of Cu/Ni Separation at Inco, in Proceedings of 36th annual meeting of the Canadian Mineral Processors, Ottawa, 15-28.










No comments: