Modelling of Macraes POX Circuit

1
Modelling of Macraes POX Circuit

• May 2006

2
Acknowledgements
OceanaGold GRD Minproc Brent Hill Tony
Frater David King Quenton Johnston Nevin
Scagliotta Adrian Marin
3
Presentation Outline

• Background
• Macraes POX circuit
• Integration of Reefton concentrates
• Modelling
• Metsim model calibration
• Model prediction of increased throughput
• Conclusion/Recommendation

4
Belo Horizonte Office
Johannesburg Office
5
(No Transcript)
6
(No Transcript)
7
Macraes Processing Background

• Historical Processing
• Small scale operation from 1862 until 1950
• 15 000 oz gold and 100 t scheelite recovered
• Modern Processing (Since 1990)
• Crush / Grind / Flotation / CIL
• Crush / Grind / Flotation / Fine grind / CIL
• Crush / Grind / Flotation / Fine grind / POX / CIL

8
Modern Project History

• Major Projects
• 1.5 Mt/a sulphide treatment plant 1990
• 3.0 Mt/a expansion 1994
• MREP 4.5 Mt/a - 1999
• Increase for sulphide and oxide capacity
• Newmont POX technology
• 170 t/d BOC cryogenic oxygen plant
• Smaller Projects
• Unit cell installation
• Reclaim circuit
• 0.5 Mt/a oxide mill
• Autoclave optimisation
• Current capacity approximately 6 Mt/a

9
Macraes Processing Issues 1

• Massive sulphide orebody hosting FeS2 / FeAsS
• Muscovite / quartz/ chlorite / siderite in gangue
• Presence of organic carbon, double refractory
• Variability. Low and high preg-robbing ore types
• 50 to 80 CIL recovery without POX
• Poor recovery with conventional POX

10
Macraes Processing Issues 2

• Newmont technology required for controlled POX
• Limestone for free acid control
• Washing for chlorides
• Scale formation in autoclave

11
Macraes POX Circuit Design

• Concentrate grade 8 - 12 S
• 3.5 m dia. x 12.6 m
• 21 semi-elliptical ends
• 4 agitator, 3 compartment vessel
• 225C and 3,140 kPag
• Koch Pyroflex membrane and AP302

12
(No Transcript)
13
Autoclave
Scaled Agitator
14
Reefton Processing

• Orebody
• Native gold with minor sulphides in quartz veins
• Gold in FeS2, FeAsS, Sb2S3
• Processing
• Crush / Grind / Flotation / Filtration /
  Transport
• Concentrate at 17.1 S
• No organic carbon
• Highly refractory, complete oxidation required

15
Reefton / Macraes Integration

• Additional S oxidation requirement
• Oxygen plant constraint
• Autoclave retention time constraint
• Differing POX conditions
• Requirement for modelling to optimise capacity

16
History of Macraes POX Modelling

• Spreadsheet POX model developed and verified
• Single-compartment Metsim model developed
• Three-compartment Metsim model developed
• POX chemistry modified based on XRD results
• Thermodynamic data sources consolidated

17
Plant Trials and Model Calibration

• Plant trial in March 04 generated 23 data sets
• Solids and solution assays recorded
• Operating conditions recorded
• Autoclave Pressure
• Temperatures in C1, C2 and C3
• Cooling water to C1 ,C2 and C3
• Oxygen flow rate and purity
• Overall oxidation from feed and discharge assays
• Compartment oxidation inferred from heat balance

18
Sulphur Analysis Discrepancy

• Trial datafor 98 oxidation, 20 t/h CW added
• Model resultsfor 98 oxidation, 16 t/h CW added
• Site assay 10 of the total S (TS) is sulphate S
• No TS reported for the trial data
• No free acid in discharge reported
• Can not do overall S balance calculation

19
MLA Mineralogy Investigation

• MLA used for quantitative mineralogy
  investigation
• MLA results 2 of TS is sulfate S
• Site assay 15 sulfate S for the same sample
• Revised S and gangue mineralogy according to MLA

20
Plant Trials in 10/04 and 01/05

• Updated trial data collection template
• Additional data for heat/mass balance
• Updated mineralogy data used
• Good correlation between models and assays
• No heat adjustment factor required

21
Plant Trials in 10/04 and 01/05
22
Plant High Throughput Trials in 07/05

• In July 2005 eight plant trials run
• Four data sets from scaled autoclave and
• Four sets from clean autoclave
• Scaled agitators show poorer oxygen dispersion
• Scaled sets average oxygen utilisation is 79
• Clean sets average oxygen utilisation is 85

23
Plant Trials in 07/05
24
Model to Predict Various Scenarios

• Plot leach kinetics for all plant trials
• Use average kinetic curve for further modelling
• Scenarios modelled
• Grade 10, 12 and 14 total S
• Throughput 2.7, 2.8, 2.9, 3.0, 3.1 and 3.3 t/h
  TS
• Constant oxygen partial pressure
• Oxygen 7 t/h

25
The Final Kinetic Curve Used for Scenario
Modelling
26
Scenario Modelling Results

• For 10 S and 12 S
• - C1 temp drops with higher throughput
• For 14 S
• - C1 maintains 225C for all scenarios modelled

27
Scenario Modelling Results
28
Scenario Modelling Results

• Above 2.7 t/h TS, oxygen constrained
• Increasing throughput, decreases RT for 12 S
• Increasing throughput, increases RT for 14 S
• For 14 S the RT is over 50 mins
• The autoclave is not constrained by RT at 14

29
(No Transcript)
30
Conclusions

• Metsim a useful framework for plant
  optimisation/design
• Careful selection of chemistry and thermodynamic
  data
• Plant trial data for model calibration
• Modelling can assist in plant optimisation and
  future design