Modern Flowsheeting Technology with Process Mineralogy

Norman O. Lotter

Challenge 1: Declining Ore Reserves, Declining Grades

Declining Head Grades

Source: Xstrata plc Annual Report 2009

Challenge 2: More Complex Orebodies

Upper Footwall Mineralization

Nickel Rim South, Sudbury

Implies more complex processing

Challenge 2: More Complex Orebodies

• Strong gold futures• Project pipelines developing well• New orebodies are challenging• New flowsheets are a combination of

gravity/leach gold and flotation

Implies more complex processing

Implications

– “Get it Right First Time” essential, not optional

– View Concentrator as a Revenue Stream, not as a Cost Centre

• Higher Recoveries = Higher NPV

– McNulty Startup Models• Want a Type I Startup or Better

– Modern Process Mineralogy can Help• Integrated use of Sampling, Mineralogy, Mineral

Processing

Structure

Modern Process Mineralogy 1983-2010– Enabling Technologies Developed and Commercialised

– Associated Methods Developed and Validated

– Hybrid Discipline

• Sampling• Quantitative and Qualitative Mineralogy

• Mineral Processing

– Integrated Practice is Powerful

• Better Ore Characterisation

• Clearer Process Implications• Better Flowsheet Formulation

Synergy

Sampling

Quantita

tive Mineralogy

Mineral Processing

Process Mineralogy

Modern Process Mineralogy

Sampling – Provides a Quantitative Structure for True Samples

Understand the orebody

Quantitative Mineralogy –Provides Clear Information on What we are Processing and How to Process it

Mineral Processing – Uses the Mineralogy and Advanced Mineral Processing Methods to Develop Optimum Flowsheets

Wanted: Reliable Test Data

Start with Representative SamplesUse appropriate sampling models and methods

Samples, not SpecimensUnderstand the Orebody

Gy’s Fundamental Variance

Use appropriate sampling models and methods

( ) ( )

−

∑

=

−

−

−n

iiis

vMaa

aM

vg

M 11

22

2 /.

.1

vf =

∑ =

= n

i iMdM

g13

.1 di

3

Gy’s Fundamental Variance

vf

= K/Ms

Sufficiently Small Variance

0

50

100

150

200

250

5 10 15 20 25 30 35 40 45 50Fundamental Variance %

Ms,

kg

5-10%

Geomet Units

A Geomet Unit is a true ore type, or group of ore types, that has a unique set of textural and compositional properties from which it may be predicted that it or the group will have similar metallurgical performance

Massive Sulphides

Net-Textured Sulphides

Disseminated Sulphides

Raglan Geomet Units

Representative Sampling

Statistical Benchmark SurveyingThis procedure extracts a representative suite of flowsheet samples from an operating concentrator at the 95% confidence level

Samples, not Specimens

Wanted: Reliable Test Data

Characterise the Samples Minerals, not Assays

MicroprobeQEMSCAN

Wanted: Reliable Test Data

Use High-Confidence Flotation Testing (HCFT)

Minimises Errors; Produces Reliable, Reproducible Results

High Confidence Flotation Testing

Sampling– Representative sample fv

< 5%

– Gy’s Ms model

– Crushing and Blending

– Spin-Riffling to Test Lots

– External Refdist with QC

Flotation Testing– Replicate Tests ea 2 kg ore

– First Concentrate Quality Control to < 5% RSD

– Total Mass Balance

– Relative Standard Deviation of Reconciled Heads (Internal Refdist) <5%

±3.27% error

95% Confidence Level

Impact on Grade/Recovery Curve

20

0

10

50 10060 70 80 90

HCFT Tightens the Confidence Limits on the Grade/Recovery Curve

Reduces Scale-up Risk

Micro-Flotation Testing

Developed by Bradshaw and O’Connor 1993-1997, Cape Town, South Africa

Performs specific flotation tests on 50-100 g samples of pure minerals

First used to demonstrate synergy of mixed collectors [Bradshaw, 1997]

Micro-Flotation: Understanding the Interactions in Flotation

Mini-Pilot Plant

Macro-Flotation: Demonstrating the Optimised Flowsheet

Continuous operation at 10-15 kg/hr

Mineral Surface Analysis – TOF-SIMS

Tof-SIMS What’s on the Mineral

Surface…?

• Identifies activators, collectors, frothers adsorbed on the mineral surface

• Provides key information in building theory on flotation mechanisms

Understanding the Interactions in Flotation

Montcalm

Timmins

Montcalm – Drill Core to Flotation Tests

Ore Characterisation : Geomet Units

Disseminated

Sulphides

Net-Textured

Sulphides

Massive

Sulphides

Pentlandite Chalcopyrite PyrrhotiteSilicates

Ore Characterisation : Problematic Ores

Pentlandite Chalcopyrite PyrrhotiteSilicates Pyrite

Pyrite…. Fine-grained Chalcopyrite…..

Bimodal grain sizes of Pentlandite…….

Process Implications

Processing Implications

• Pyrite.. Variable.. Dilutes the concentrate• Pentlandite… Bimodal size distribution.. Need a

staged grind

Constraints

• Flowsheet frozen by earlier ownership and fast-track project schedule

Montcalm – StartUp

Rougher-Scavenger Flotation

1st Cleaner Flotation

2nd Cleaner Flotation

Concentrate

To Cu/Ni Separation

2nd Clnr Tails

1st Clnr Tails

Scav Tails

B

A

C

DB

C

Rougher-Scavenger Flotation

1st Cleaner Flotation

2nd Cleaner Flotation

Concentrate

To Cu/Ni Separation

2nd Clnr Tails

1st Clnr Tails

Scav Tails

B

A

C

DB

C

Lime

Case Study: Montcalm Project, North Ontario

Montcalm 2004– Type I StartupComparison of Montcalm Start-up Curve with McNulty Curves

0

20

40

60

80

100

120

0 2 4 6 8 10 12 14

Quarter after start-up

Ni o

utp

ut

in N

i an

d C

u c

on

c,%

of

des

ign

Type 4

Type 3

Type 2

Type 1 Montcalm start-up

Montcalm start-up - October 2004

Case Study: Montcalm Project, North Ontario

Montcalm – Type I Startup

0

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

HCFT Act

Grade % Ni

Gra

de %

Ni

80

81

82

83

84

85

86

87

88

89

90

HCFT Act

Rec % Ni

Rec

over

y %

Ni

Post-Commissioning

Statistical Benchmark Survey

QEMSCAN Measurements

Final Tailings

+10675+53CS1CS2CS3CS4-5CS6CS7

Liberated

Middling

Locked0

5

10

15

20

25

% P

entl

and

ite

Pentlandite Liberation in Final Tail

Pentlandite

+10

675+53

CS

1

CS

2

CS

3

CS

4-5

CS

6

CS

7

Liberated

Middling

Locked0

3

6

9

12

15

18

21

24

% C

hal

copy

rite

Chalcopyrite Liberation in Final Tail

Chalcopyrite

Ultrafine Liberated Losses

Coarse Locked Losses

Grinding Circuit

Post-Commissioning Remedies

• Rearrange the grinding circuit– Increase circulating load

– Tighten size distribution• Less ultrafines• Less topsize

• Process response– 25% gain in milling capacity

– 2.57% gain in nickel recovery

– 1.83% gain in copper recovery

Eland Platinum – the Opportunity

Eland Commissioning Nov 2007

Commissioned Nov 2007

• Standard Design for Bushveld UG2 Ore Type

• Commissioned on SNPX, changed to SIBX 2008

TAILC1C2

C3

M1: 45% -75 microns

M2: 80% - 75 microns

F1

F2

TAILC1C2

C3

M1: 45% -75 microns

- 75 microns

F1

F2

TAILC1C2

C3

M1: 45% -75 microns

M2: 80% - 75 microns

F1

F2

TAILC1C2

C3

M1: 45% -75 microns

- 75 microns

F1

F2

Eland – the Objective

Specific Objective

• Eland’s Concentrator is a generic design– No tailoring for this specific orebody

– Design and construction were fast-tracked

• Operations management prepared for post-commissioning improvements– Wanted improved concentrate grade and recovery of PGE– Process Mineralogy were requested to assist

Eland Operations

Understand the Orebody

Eland – Mineral Compositions

EPMA - Characterise Mineral Compositions

Eland – Composition of Discrete PGM

0%

20%

40%

60%

80%

100%

1 2 3 4 5

Pt

Pd

Rh

Cu

Ni

Fe

S

Composition of the Discrete PGMs varies widely…Electrochemical Implications…?

Distribution of Platinum Group Minerals

Eland UG2

30%

6%

23%

16%

13%

3%

1%

4%

1%

1%

0% 10% 20% 30% 40%

Pt Sulphide

Pt(PdNi) Sulphide

PtCu Sulphide

Ru(Rh) Sulphide

PtRh Sulphide

Pd Sulphide

Pt Arsenide

Pt Telluride

PtPd Telluride

Au-Ag

Quantitative Measurement

Discrete PGM Sulphides

Discrete PGM Amphoterics

Electrum

Eland Platinum - Implications

Implications

• The suite of discrete PGM minerals is too diverse and complex for a single xanthate collector– Need an optimised mixed collector system

High-Confidence Flotation Testing

Xanthate 1

Xan

that

e 2

(+,+,+)

(+,-,-)

(-,+,+)

(-,+,-)

(-,-,+)

(+,-,+)

DTC

(+,+,-)

Replicated Factorial at High Confidence

Mixed Collector Suites Improve Flotation Performance

New Mixed Collector Suite Formulated

Using “Reagent Sudoku”

Case Study: Eland Platinum

Successful Plant TrialsDemonstrate the Value

Successful Plant Trials Result from Proper Statistical Designs that only test Promising Laboratory Work

Successful Plant Trials

Plant Trial Format

• Uses on-off switching between standard SIBX and Mixed Collector Exp 820

• Run on West Pit “Normal” Ore• Reference Distributions collected• Statistical Test Design

Successful Plant Trials

Plant Trial FormatBlock 1

Baseline SIBX

Block 2

Mixed Collector

Days

2 weeks 2 weeks 2 weeks

Legend

Changeover

Data capture

Plant operating on ‘other’

ore or plant standing

(variable block length)

Block 3

Baseline SIBX

Block 4

Mixed Collector

2 weeks

Block 1

Baseline SIBX

Block 2

Mixed Collector

Days

2 weeks 2 weeks 2 weeks

Legend

Changeover

Data capture

Plant operating on ‘other’

ore or plant standing

(variable block length)

Block 3

Baseline SIBX

Block 4

Mixed Collector

2 weeks

Eland Plant Trial

Grade Bin g/t 4E

Frequency %

0

5

10

15

20

25

30

0.5 0.6 0.7 0.8 0.9 1 1.1 1.2 1.3 1.4 1.5

SIBX

Exp 820

Final Tailings Grades Reduced

0.86

0.88

0.9

0.92

0.94

0.96

0.98

1

Ave

. Tai

lings

G

rade

g/t

4E

Eland Plant Trial

0

5

10

15

20

25

30

35

100 125 150 175 200 225 250 275 300

Grade Bin g/t 4E

Frequency %

Final Concentrate Grades Increased

145

150

155

160

165

170

175

180

185

190

Ave

. Con

c.

Gra

de g

/t 4E

SIBX

Exp 820

Eland Plant Trial

Final Concentrate Quality Increased

• Optimised Mixed Collector

Reduces Entrainment of

Chrome

1

1.2

1.4

1.6

1.8

2

2.2

Ave

. Con

c G

rade

% C

r 2O

3

SIBX

Exp 820

Case Study: Eland Platinum

Successful Plant TrialsDemonstrate the Value

Mixed Collector Project at Eland Platinum Gains 2.48% PGE Recovery and 16.6% Higher Concentrate Grade

Differences Demonstrated at > 90% confidence level

Accurate Scale-up from Lab Work

Rec

ove

ry G

ain

% P

GE

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

4E PGE

Lab

Plant

Conclusions

Conclusions

• Modern Process Mineralogy now has the toolbox– New mines, type I startups

– Current operations, improving performance

– Better flowsheeting

– Clearer demonstration of value

– Less risk

Acknowledgements

Acknowledgements

The author would like to thank the CIM for their sponsorship of this lecture series, and Xstrata for their support and permission to publish certain sections of this lecture from earlier XNi and XAlloys project work.

Additionally the author would like to thank his many colleagues for their support in the development of Process Mineralogy.