Council for Mineral Technology

Presentation toMintek 75 – A Celebration of Technology

Automated SEM study of PGM distribution across a UG2 concentrate bank

5 June 2009

Deshenthree ChettyHead – Process Mineralogy

• Introduction• Automated SEMs• Parameters considered• Data quality: uncertainties• Parameter variation down the concentrate

bank• Preliminary conclusions and further work

Outline

• PGM characterisation routinely carried out, what are the implications for floatability?

• Previously, batch flotation tests (Penberthy, 2001) determined floatability characteristics. Can these be related to a dynamic plant-scale system?

• Part of a larger study that critically assesses the BMS proxy, and PGM identification by automated SEM at sub-10 um levels, along with statistical assessment of data from automated SEM.

Introduction

• Therefore, sampled a UG2 primary rougher circuit at a SA concentrator (more PGMs!)

• Aim: Can the floatability parameters we routinely use adequately resolve differences in PGM-bearing particle behaviour in the ten concentrate cells of the rougher circuit? Test of robustness of the PGM-bearing particle parameters, as determined by auto-SEM

• First results of ongoing investigation

Introduction

Automated SEMs

Arrived June 2008

Automated SEMs

Automated SEMs

• PGM species

• PGM size distribution

• PGM mode of occurrence-liberation characteristics -mineral association (BMS and gangue)

• Floatability

Parameters considered

Data quality

Sample Number of grains PRF 198 PRC1 784 PRC2 465 PRC3 338 PRC4 456 PRC5 360 PRC6 414 PRC7 321 PRC8 288 PRC9 410 PRC10 363 PRT 116 TOTAL 4513

• Statistically valid interpretations?

• Comparing numbers, are they the same or different? We require uncertainties.

• Resampling statistics to obtain std deviations on mean values of different parameters, e.g.modalabundance of species, size classes of PGMs sampled.

6-9 micron

0.0

10.0

20.0

30.0

40.0

50.0

0 50 100 150 200

Aliquot size

Are

a %

0.0

10.0

20.0

30.0

40.0

50.0

Std

dev

10% value 90% value "True" value Std Dev

Data quality

Sampling distributions 6-9 micron category

(25.21)

0

5

10

15

20

25

30

35

40

0-5 5-10 10-15 15-20 20-25 25-30 30-35 35-40 40-45 45-50 50-55 55-60 60-65 65-70 70-75 75-80 80-85 85-90 90-95 95-100

Category

Perc

ent

20 grains 50 grains 100 grains 200 grains

Data quality

PGM grouping Modal abundance (%) Uncertainty (±%) PGE Sulfides 93.65 3.34 PGE Alloys 4.24 3.13 PGE Arsenides 1.13 1.06 PGE Tellurides 0.75 0.84 PGE Bismuthotellurides 0.23 0.17

Size class relative abundance (vol%) Uncertainty (±%)

0-3 5.75 1.08 3-6 30.76 4.96 6-9 25.83 4.96

9-12 14.24 4.74 12-15 6.29 3.67 15-18 4.84 4.05 18-21 4.87 4.87 21-24 2.33 4.15 24-27 0.00 0.00 27-30 0.00 0.00 >30 5.10 8.61

PRC1: 784 grains, resampled with 200 grain aliquots, 1000 simulations. Mean and uncertainty at 90% confidence level

• Automation, but time-consuming for what we want to achieve…

• … = 6 months. Add: 4 months for sample insertions and vacuum pumping, standards file establishment, offline processing, reporting and interpretation, at instrument efficiency of 80%, this is ~11 months continuous work

Data quality

grainsPolished sections

run time (hours)

checking time (hours)

Total time (hours)

500 129 386 64 4511000 8 38 31 691000 15 75 60 1351000 38 173 154 3271000 29 128 57 1851000 36 163 72 2351000 46 207 92 2981000 53 212 106 3181000 83 333 167 5001000 71 283 141 4241000 116 463 231 694500 272 815 68 883

hours 1243 4519days 155 188

Sample grades

0

10

20

30

40

50

60

70

80

PRF PRC1 PRC2 PRC3 PRC4 PRC5 PRC6 PRC7 PRC8 PRC9 PRC10 PRT

Con

cent

ratio

n (g

/t)

Pt Pd Rh Ru Ir

1.61PRT

14PRC10

20PRC9

23PRC8

18PRC7

25PRC6

26PRC5

38PRC4

45PRC3

102PRC2

199PRC1

5.41PRF

PGE (g/t)Sample

UG2 Feed composition

PdAsSb

0.6PdAsSn PGE Others

PtPdAsSn

PtPdBiTe

0.9PdBiTe PGE Bismuthotellurides

PtBiTe

PtPdTe2.5PGE Tellurides

PtTe

PtRuAs

0.1PtAs PGE Arsenides

PdAs

PdPb

2.6PdHg PGE Alloys

PtFe

RuS

PtPdRhCuS

PtPdAsS

PdS

93PtRhS

PtRhAsS

PtRhCuS

PtS

PGE Sulfides

PtPdNiS

Size distribution

Size distribution

0%10%20%30%40%50%60%70%80%90%

100%P

RF

PR

C1

PR

C2

PR

C3

PR

C4

PR

C5

PR

C6

PR

C7

PR

C8

PR

C9

PR

C10

PR

T

Primary rougher cells

Volu

me

%

0 - 3 3 - 9 9 - 15 15 - 21 >21

Mode of occurrenceL

PGM = BMS = Silicate/Oxide Gangue =

SL AG SAG SG G

L = Liberated PGM

SL = PGM assoc. with liberated BMS

AG = PGM attached to gangue

SAG = BMS attached to gangue

SG = Locked BMS

G = Locked PGM

Mode of occurrence

Mode of occurrence

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

PRF PRC1 PRC2 PRC3 PRC4 PRC5 PRC6 PRC7 PRC8 PRC9 PRC10 PRT

Volu

me

%

L SL SG AG SAG G

• Liberation index = Area of the valuable mineral/total area of particle• Liberated PGM or PGM-BMS particle = 1• Small, locked PGM-BMS in gangue ~0

Mode of occurrence

Liberation Index

0%

20%

40%

60%

80%

100%PR

F

PRC

1

PRC

2

PRC

3

PRC

4

PRC

5

PRC

6

PRC

7

PRC

8

PRC

9

PRC

10

PRT

Volu

me

%

<0.2 0.2 - 0.4 0.4 - 0.6 0.6 - 0.8 0.8 - 1.0

• Floatability index:

Mode of occurrence

liberated PGMs >3um ECD Fast Floating liberated BMS >10um ECD liberated PGMs <3um ECD liberated BMS <10um ECD

PGMs >3um ECD attached to gangue Slow Floating 1

BMS >10um ECD attached to gangue PGMs <3um ECD attached to gangue Slow Floating 2 BMS <10um ECD attached to gangue

Non-floating PGMs and/or BMS locked in gangue

Mode of occurrence

Floatability Index

0%10%20%30%40%50%60%70%80%90%

100%

PRF

PRC

1

PRC

2

PRC

3

PRC

4

PRC

5

PRC

6

PRC

7

PRC

8

PRC

9

PRC

10

PRT

Volu

me

%

Fast Floating Slow Floating 1 Slow Floating 2 Non-floating

Associated gangue (majority of SAG class):• Orthopyroxene/talc – naturally floatable gangue• Chlorite – naturally floatable gangue• Amphibole - ?• Plagioclase – non-floating (BMS-driven?)

Mode of occurrence

Gangue AssociationsLiberation Characteristic

0

5

10

15

20

25

30

35

40

45

50

Chlorite

Plagioc

lase

Zirco

n

Chromite

Orthop

yroxe

ne

Phlogo

pite

Amphibo

le

Quartz

Clinop

yroxe

ne

Sphen

eRutile

Calcite

# gr

ains

GSGSAGAGSL

PRC3

• Smaller data sets

Liberated PGMs…

PGE sulfides

75.00

77.00

79.00

81.00

83.00

85.00

87.00

89.00

91.00

93.00

95.00

PRF PRC1 PRC2 PRC3 PRC4 PRC5 PRC6 PRC7 PRC8 PRC9 PRC10 PRT

Volu

me

%

Liberated PGMs…

0.00

2.00

4.00

6.00

8.00

10.00

12.00

PRF PRC1 PRC2 PRC3 PRC4 PRC5 PRC6 PRC7 PRC8 PRC9 PRC10 PRT

PGE Alloys PGE Arsenides PGE Bismuthotellurides PGE Tellurides PGE Others

• Floatable vs non-floatable gangue; can the floatability index be made more robust by accounting for floatable gangue in a numeric descriptor? Which plant conditions favour such floatability?

• Data are not enough, especially when finer details are sought!

• Is there really a difference expected with such similar grades from PRC 5 to PRC 10?

Preliminary conclusions

• Colleagues and students:-Lelanie Gryffenberg-Thelma Lekgetho (UFS)-Itumeleng Molebale (UP)-Charles Bushell-Dr Chris McLaren

• MPD for simulations, flotation response information and ongoing discussions

Acknowledgements

www.mintek.co.za

Thank you