You are on page 1of 19

Improving Fine (<20m) Copper Sulphide Mineral Recovery by Magnetic Agglomeration Measuring the Degree of Agglomeration

Rebecca J. Fleming

Rio Tinto Technology & Innovation, Australia

Glenn C. Wood
Barry G. Lumsden

Northparkes Mines, Australia


Centre for Multiphase Processes, University of Newcastle, Australia

Presentation Overview
Introduction to Northparkes Mine Fine Copper Recovery at Northparkes Magnetic Agglomeration Experimental Design & Analysis

Study Conclusions Questions

Northparkes Mine
Copper & gold mine Joint venture 80% Rio Tinto & 20% Sumitomo Group Located ~ 390 km west of Sydney Underground and open cut ore sources

Northparkes Mine Sydney

Facts about Northparkes Mine


Copper Minerals: predominately chalcopyrite (CuFeS2) & bornite (Cu5FeS4) Head grade: 0.83% Cu, 0.44 g/t Au Recovery targets: 89.7% Cu, 79.0% Au. Grade targets: 34.4% Cu, 16.8 g/t Au

Northparkes Flowsheet
Ore processed through 2 parallel modules. Planned milling rates: Module 1 245 t/h, Module 2 423 t/h

CV12 CV21 Oversize Crusher CR03 Ball Charger BC02

Metal Detector

Ore Ore Stockpile Stockpile 2

Weightometer Optional

Apron Feeder AF01 Feeders FE06-09 SV06 Svedala (2400x6100) CV10 Weightometer Magnet (MA02) CV11 PP091 (Frother)

Tromm TL06

ML03 (8500x4300) To Flotation Circuit

NC NO From Grinding Circuit HP41

HP13

CY03 A-D (20" CAVEX) PP064/65 (12/10 F-AH)

CY04 A-O (15" CAVEX)

CV330 Ball Charger Hopper Splitter

CV329

FT45 Flash Float Rougher (Skimair) Flash Float Cleaner (Outokumpu 5) To HP41

MODULE 2 RECLAIM (321) & GRINDING (330) CIRCUIT

CV334 ML04 (5500x9400)

Trommel TL04

FT46

CV332 SCATS

M FLOT

To Primary Hopper PP068 SUMP (65VD-GPS)

To Con Pump

Flash Float Con

HP15

PP066/67 (14/12 FF-AH)

To Concentrate

MODULE 2 CONCENTRATE THICKENING (355) &

From Concentrate Pump PP134 SUMP (65 VDS-GP)

Northparkes Flowsheet
Ore processed through 2 parallel modules. Planned milling rates: Module 1 245 t/h, Module 2 423 t/h
TK50 Pre-Float TO FLOAT CELLS PP115 (NaHS) PP69 AG21 NO NC AG22
Proflote Units

CV715

PP089/90 (Frother) PP110 (Collector) PP103 (Promoter) 1st Roughers (Dorr Oliver) FT15-16 2nd Roughers (Dorr Oliver) FT17-18 PP112 (Collector) BL01-03 Blowers

Optional

CY06 A-H 15" Krebbs TK36 Conditioner 1 Trommel TL06 ML06 Optional

CV450

CV451

1st Scavengers (Dorr Oliver) FT19-20

CV452

2nd Scavengers (Dorr Oliver) FT21-22

CV453

TK37 Conditioner 2

NC NO To Flotation Circuit From Grinding Circuit HP41 PP222 (12/10 F-AH)

NO NC PP080 (8/6 E-AH) HP25 AG30/31 PP079 (12/10 E-M)

PP075 (8/6 E--AH) FT27 (8 Downcomers)

TK57/58 Cleaner-Scavenger Feed Conditioner

HP12

NC Cleaner PP223 SUMP (65VD-GPS) To HP41 Jameson Cell Cleaner-Scavengers (Dorr Oliver) FT23-26

NO
To Hopper 14 CY05 Feed

CV454

CV457

FT28 (3 Downcomers) Optional Recleaner Jameson Cell

MODULE 2 FLOTATION (340) CIRCUIT


To Con Pump Flash Float Con

Optional HP26

HP11 PP076 (6/4 D-AH) CV459

PP081 (6/4 D-AH)

PP082 (4RV-AF)

To Tailings Thickener

To Concentrate Thickener

Mod 1 Concentrate Thickener O/F From Concentrate Pump

PP105 (Floc) From Mod 1 Tails Pump

From Mod 2 Tails Pump

Tailings Thickener

Fine Copper Recovery


Significant copper and mass losses in <11 Module 1 m Magnetic conditioning installed to target these ultrafine losses
40 35 30
% Retained

25 20 15 10 5 0
-11 +11 +20 +38 +53
Cu Distribution Mass Distribution

+75

Size Fraction (m)

+106

+150

Mean Copper and Size Distribution of Northparkes Module 1 final tail (2009)

Magnetic Agglomeration (1 of 2)
ProFloteTM installed at Northparkes Selective agglomeration of paramagnetic minerals Paramagnetic minerals become magnetised when exposed to a strong magnetic field Agglomeration of magnetised paramagnetic minerals occurs when there is sufficient magnetic attraction

Vt = Total energy of attraction

Vt Va Vr Vm

Va= London van der Waals energy (attractive) Vr = Electrostatic energy (repulsive) Vm= Magnetic energy (attractive)

*Sourced from: Svoboda (1987) Magnetic Methods for the Treatment of Minerals, Elsevier Ltd., Amsterdam, Netherlands.

Magnetic Agglomeration (2 of 2)
Agglomeration (magnetic attraction energy) dependent on: mineral magnetic susceptibility, mineral particle size, and distance between minerals and magnetic induction.
Mineral Reported magnetic susceptibility (M3kg-1x10-9)* 1596.0 101.0 -5.7 1.0 5.0 -0.15

Chalcopyrite Bornite Quartz Pyrite Gold

*Sourced from: Svoboda (1987) Magnetic Methods for the Treatment of Minerals, Elsevier Ltd., Amsterdam, Netherlands.

Magnetic Conditioning Previous Results*


ProFloteTM installed in Module 2 flotation feed stream in 2005 Randomised block statistical plant trial to assess its effectiveness ~6 months to complete trial No difference in copper concentrate grade with magnetic conditioning
Size Fraction <20 m 20 m to 38 m >38 m Total sample Change in Cu Recovery (%) with magnetic conditioning +2.1 +1.0 +1.2 +1.4 Level of Statistical Confidence (%) 98.9 99.8 Very low 96.0

*Sourced from: Rivett, T., Wood, G. and Lumsden, B. (2007) Improving Fine Copper and Gold Flotation Recovery A Plant

Evaluation. Proceedings of the Ninth Mill Operators Conference, The Australasian Institute of Mining and Metallurgy, Melbour ne, Australia. pp 223 228.

Measuring Agglomeration
Unrelated flotation study at Northparkes (December 2009) discovered agglomeration could be observed using laser diffraction Laser diffraction is an optical size distribution technique

Different sized particles scatter light, from a laser beam, at different angles and intensities

Images sourced from: Malvern Instruments Inc, 2010. Available from http://www.malvern.com/processeng/processes/classification

5.0

4.5 4.0
Volume % In Size Class

Measuring Agglomeration Stage 1


Magnetic Conditioning
5.0

3.5 3.0
2.5

No Magnetic Conditioning
4.5 4.0
3.5

5.0

4.5 4.0
Volume % In Size Class

3.5 3.0
2.5

2.0
Volume % In Size Class

1.5

3.0
2.5

2.0
1.5

1.0 0.5 0.0 0.1


0.1 1.0

2.0 1.5
1.0

1.0 0.5 0.0

0.5

1.0
10.0 Particle Size ( m) 100.0

0.0 10.0 0.1 1000.0 Particle Size ( m)

100.0
1.0 10.0

1000.0
100.0 1000.0

Particle Size ( m)
Cleaner Feed (before sonication) average - without magnetic conditioning

Cleaner Feed (before sonication) average - with Magnetic Conditioning


Cleaner Feed (after sonication) average - with Magnetic Conditioning

Cleaner Feed (before sonication) average

Cleaner Feed (after sonication) average

Cleaner Feed (after sonication) average - without magnetic conditioning

Natural agglomeration % agglomeration estimated as change in % volume (of a size fraction) before and after sonication

Measuring Agglomeration Stage 2


ProFloteTM installed in Module 1 flotation feed stream early 2010 Rather than traditional plant trial, laser diffraction technique used to estimate degree of agglomeration. Test work completed within 1 week. Paired t-test analysis showed statistically significant increase in particle size with magnetic conditioning

Mean Results Size Fraction Copper Minerals (%) Agglomeration (%)*

<11 m
<20 m <38 m

26.4
27.9 28.1

13.8
14.6 14.8

*Natural agglomeration removed (found with magnetic conditioning off)

Correlating Magnetic Agglomeration with Copper Recovery


Traditional on/off plant trial not conducted unstable feed conditions during study
Previous trial showed 2.1% increase copper recovery for <20 m with magnetic conditioning Using plant recovery & copper distribution results, 2% increase <20 m copper recovery could be achieved with 20% agglomeration in <11m fraction <11 m fraction accounts for over 80% of the <20 m fraction
Copper Distribution Size Fraction 11 m to 20 m <11 m Total <20 m Copper Recovery (%) 96 84 Before Magnetic Agglomeration (%) 22 78 86.6 After 20% Magnetic Agglomeration (%) 38 62 88.6

2% increase

Conclusions
Agglomeration of paramagnetic copper minerals observed in cleaner feed stream following magnetic conditioning of flotation feed Agglomeration occurring for copper minerals <11 m <11 m is the most difficult fraction to recover by flotation at Northparkes Degree of copper agglomeration can be estimated by laser diffraction, coupled with sieve sizing, copper assays and mineralogy Laser diffraction method is a promising approach to substantiating magnetic conditioning

Paper Amendments
Page 224, Procemin 2010 Seminar Proceedings Correlation of Magnetic Agglomeration with Copper Recovery section
Rivett et al [8] showed a 2.1% increase in copper flotation recovery for the <20 m fraction with magnetic conditioning. Based on plant recovery and the copper distribution results obtained in this recent test work, a 2% increase in <20 m copper recovery could be achieved with 20% agglomeration in the <11 m fraction. This is shown in Table 7. This 20% agglomeration is considered reasonable as it falls within the agglomeration range observed in this study which spanned from 12.9% to 23.3% agglomeration in the <11 m fraction. The focus on the <11 m fraction is due to it accounting for over 80% of the <20 m fraction. The 2% increase estimation is comparable to the 2.1% increase measured in the plant trial [8], suggesting that the laser diffraction analysis is a promising approach for assessing magnetic conditioning. However, to quantitatively correlate the degree of agglomeration with copper recovery, a statistical plant trial would need to be performed in conjunction with the laser diffraction analysis.

Questions

Acknowledgements
David Rahal, Technical Director, Knelson (Deswik) Milling Solutions Inc

Northparkes Mines, Rio Tinto

Professor G Jameson

Magnetic Agglomeration (2 of 2)
Agglomeration (magnetic attraction energy) dependent on: mineral magnetic susceptibility, mineral particle size, and distance between minerals and magnetic induction.
Vm = magnetic energy (attraction) o = magnetic permeability of a vacuum 1, 2 = volume magnetic susceptibilities of the particles b1, b2 = radius of the particles H = magnetic field strength h = distance between the surface of the particles

3 3 (8 o 1 2 b1 b2 H 2 ) * Vm 9(h b1 b2 ) 3

*Sourced from: Svoboda (1987) Magnetic Methods for the Treatment of Minerals, Elsevier Ltd., Amsterdam, Netherlands.