Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
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.