Teck Carmen de Andacollo

Comminution Ore Testing

9 February 2012

Dr Toni Kojovic
Managing Director, SimSAGe Pty Ltd
Research Associate, JKMRC
 Rock Characterisation
Why do we need the tests?
1. Given a breakage process, estimate the
– Product size for a given ore feed size & hardness.
– Energy to generate a particular product size.
2. Circuit design and optimisation
– Equipment sizing and selection
3. Geometallurgy
– Variability testing and modelling
 Issues
• Ores vary in hardness and density
• Power consumption depends on the duty and ore
hardness/size/density
• Ore hardness is a difficult property to define since
relationship with the machine is not always clear….
– AG mill treating soft ore may have a low capacity due to lack of
hard and coarse ore media
– Cone crusher will produce a similar product size for a wide range
of ores, but will draw different powers in doing so.
• Can be expensive, so need to define purpose
 Summary of Available Tests
• Tumbling mill tests
– Bond Mill Index (BMWI)
– Rod Mill Index (RMWI)
– SPI (Mins) & SAGDesign (Revs)
• Compression / impact tests
– Bond crushing work index (BCWI)
– UCS (Uniaxial Compressive Strength)
– Point Load (useful to estimate UCS)
– JK Drop Weight (and ore abrasion) (A, b and ta)
– SMC (predicted A, b and DWI)
– JKRBT (A, b and Emin)
• Proxy tests
– EQUOtip (Ls)
– GeM Comminution Index (Ci)
 Comminution Test
UCS PLT EQUOtip GeM Ci
Application Map
In-Pit Pri Crusher Grinding Final
Product Circuit Product

Pri
Crusher SAG mill Ball Mill
Ore bench
SAG
stockpile

RoM

drill holes

Bond CWi PLT SPI SAGdes DWT/SMC RBT Bond Wi Bond Ai

 Test Summary
Test Sample Requirements Test Application/Use Cost per
Output sample
Density 30 pieces, 31.5x26.5mm SG Input in AG/SAG models, slurry $50
(wet/dry or (~0.75kg) pulp density calculations
He)
Point Load 50 pieces, 75x50mm Is(50) Crushing hardness, proxy for UCS, $50
fragments or core pieces blasting model input, variability
(~5kg) testing
GeM Ci 5 kg of half core, or sized rock Ci-CRU, Proxy for crushing and grinding $80
fragments (eg. -26.5+22.4mm) Ci-GRD hardness, variability testing,
lithology + alteration mapping
EQUOtip Half or full core, need cut Ls Lithology + alteration mapping, $10-20 per
surface surface hardness, proxy for UCS, p- m
wave velocity (rock mechanics) depending
on volume
Bond Ball 10kg of rock or drill core BMWi Industry standard for estimating $1,200
Mill Work specific energy (kWh/t) , mill design
Index + optimisation, JKSimMet model
input
 Test Summary / 2
Test Sample Requirements Test Application/Use Cost per
Output sample
DWT 75kg of ore, sized into SG Crushing and AG/SAG $4,500
63x53mm, 45x37.5mm, A, b and hardness, mill design +
31.5x26.5mm, 22.4x19mm and A*b; App optimisation, JKSimMet
16x13.2mm Fns; ta model inputs
3kg of ore 53x45mm (abrasion)
SMC 100 pieces of cut core pieces or A, b and Crushing and AG/SAG $1,500
narrowly sized rock fragments A*b; hardness, variability testing,
(~5kg depending on rock size) DWi mill design + optimisation
RBT 50kg of ore, sized into SG Crushing and AG/SAG $4,350
45x37.5mm, 31.5x26.5mm, A, b and hardness, mill design +
22.4x19mm and 16x13.2mm A*b; App optimisation, JKSimMet
3kg of ore 53x45mm (abrasion) Fns; ta model inputs
RBT Lite 200 particles in 11.2x9.5mm A, b and Crushing and AG/SAG $300
range (~ 0.5kg total) A*b hardness, variability testing,
proxy for full DWT A*b
RBT 50 sized rock fragments (eg. A*b Crushing and AG/SAG $100
Express -11.5+9.5mm, ~0.1kg) hardness, variability testing,
proxy for A*b
 Test Summary / 3
Test Sample Requirements Test Application/Use Cost per
Output sample
SPI 2kg of coarse rejects from drill Mins AG/SAG specific energy $500
core or RVC cuttings (-19 mm (kWh/t), mill design +
material) – prepared by crushing optimisation, variability
to 100% -19 mm (80% -13 mm). testing
SAG ~10kg of ½ core crushed to 80% Revs AG/SAG mill specific power $3,000
design -19mm. required to treat a target tph
or estimate the expected tph
given an existing mill.
Bond Ai 1600g of -19+12.7mm rocks Ai Relative liner and media $400
wear index
Bond CWI 20 rock specimens in the size CWi Estimate specific crushing $500
range -76+50 mm are broken power (kWh/t)
under the impact of twin
pendulums (~10kg).
UCS Cylindrical core specimens with a UCS Crusher design, drill and $300 per
L:W ratio of 2.5-3.0, and a blast, geotechnical strength sample
diameter of about 50 mm.
 JK Drop Weight
• Drop height  input energy
(kWh or J)
• specific energy = kWh/particle
mass (kWh/t)
• Break rocks at different specific
energies and size products
– map that relates specific
energy to product size
distribution (t10)

JKMRC Drop-Weight
Apparatus
 What does the JKDWT measure?
• Quantify the degree of breakage at a given energy

A T10
T10(%)
(%)
38.1
38.1

19.8
19.8

10.4
10.4

4.4
4.4

t10  A 1  e b. Ecs 
• A*b – ore specific hardness index A*b A*b=23
=23(hard
(hardore)
ore)

• A and b parameters 0.2

0.2used in 0.5
JKSimMet
0.5 model
1.0
1.0 2.0
2.0
A*b ~ slope Energy
Energy(kWh/t)
(kWh/t)

%fines per kWh/t

 Standard Drop Weight Test
• 5 size fractions each at 3 energy levels
• 1 energy level for all 5 size fractions

Size Range
Energy (kWh/t)
(mm)
63 x 53 0.10 0.25 0.40
45 - 37.5 0.10 0.25 1.00
31.5 - 26.5 0.25 1.00 2.50
22.4 - 19 0.25 1.00 2.50
16 - 13.2 0.25 1.00 2.50
 JK Ore Abrasion Test
• Resistance to Abrasion Breakage
• ta parameter
– Smaller values of ta indicate more resistance
– 3kg of 55x38mm rocks for 10 mins @ 70% critical
speed in 305 x 305mm lab. mill
• ta = t10 /10
– 0.2-0.3 hard ore (AG, ABC)
– 0.3-0.6 medium hard ore (SAG, SABC)
– 0.8 to 1.5 friable ore (SAG)
 Crusher Appearance Function
• Crusher is assumed to break by impact only
• Appearance function generated from the
t10 v Cum. % Passing map for the specific
ore
• t10 is a function of crusher type, ore
hardness and crusher operation
 Crusher Appearance Function -
Example
Appearance Function Data, tn (%)
t10 (%) t75 t50 t25 t4 t2
10 3.00 3.60 5.30 23.50 53.30
20 6.00 7.40 10.80 44.80 82.90
30 9.20 11.40 16.50 63.20 95.70

Size Reduction/Specific Comminution Energy, Ecs (kWh/t)

Initial Particle Size (mm)
t10 (%) 14.5 28.9 41.1
10 0.39 0.34 0.27
20 0.84 0.71 0.58
30 1.34 1.14 0.93
 DWT Energy vs Size Reduction

100
T10 (%) 90 Product Size
38.1 Distribution
80

70
19.8
X 60 T2

Cum % Passing
50
10.4
40

30

ore)   b  Ecs 
T4
4.4
tA*b  (hard
10=23 A  exp 20

T10
10
T25
T50
0 T75
0.2 0.5 1.0 2.0
0.01 0.10 1.00 10.00 100.00
Energy (kWh/t) Particle Size (mm)
Xo=32.5mm

Rock Size,
PROCESS
X=32.5mm
MODEL
 Example of Ore Assessment for
AG/SAG Mills
Comparison of size reduction expected in a SAG and AG mill for
two ores with different A*b
Mill Diameter (m) 9.4   Mass rock (kg) 0.173 0.173
Particle Size (mm) 50   Mass ball/rock (kg) 2.71 8.03
Ball/Rock Size (mm) 125  
SG rock 2.65  
SG ball 7.85  
 
Breakage Parameters Hard Ore Soft Ore
A 55.2 68.9
b 0.66 1.42
A*b 36.4 97.8

Breakage Hard Ore Soft Ore

Parameter AG SAG AG SAG
Ecs (kWh/t) 0.40 1.19 0.40 1.19
t10 (%) 12.8 30.0 29.9 56.1
Example Ore Comparison

A=55.2

Hard Soft
A 55.2 68.9
b 0.66 1.42
A*b 36.4 97.8
 Breakage in JKRBT vs DWT
RBT Low Energy
48 m/s, 0.3 kWh/t

DWT - High Energy 2.5 kWh/t

M=15kg, H=32cm, m=5g

RBT High Energy

121 m/s, 2.0 kWh/t
 18 640 640

PLT Trends 17

16
600

560
60

560

SAG kWh/t
15 520
520

G
h
p
A
S
t
14 480
140
480
13
120 40
Drop Weight Parameter - Axb/SG

440
12
100
40
4.0 5.0 6.0 7.0 8.0 9.0
400 4.0 5.0 6.0 7.0 8.0 9.0
IS(50) - MPa
80

60

10
40

20 8

SMC Index DWI (kWh/cu.m)

0
0 2 4 6 8 10 12 14 16 6
Point Load Rock Strength (MPa)

Database Power (Database)

4

0
0 2 4 6 8 10 12
Point Load Strength (MPa)
Batu Hijau
Ore Characterisation
TPH Modelling
 SMC
DWT test at 1 size x 5 specific energy levels (20 core pieces or sized
fragments)

DWI  100 x SG / A * b
AG/SAG Specific Energy  K.f(F80, DWI, BL, SPEED, L/D,TromX)
 SMC Mi Values
• Mia, Mic and Mih derived from DWT test results on 1 size x
5 energy SMC test.
• Mib calculated from Bond BM Work Index
• Used in universal specific power model to estimate total
specific energy to reduce the primary crusher product to
final product


W  K 4 M i P 80 f  p80   F 80 f  f 80  
K takes into account circuits with recycle pebble crusher
 SABC Example
• WT = Wa+Wb+Wc
where
– Wa = specific energy to grind coarser particles in
tumbling mills
– Wb = specific energy to grind finer particles in
tumbling mills
– Wc = specific energy for conventional crushing
 Effect of Size on A*b
• Throughput models rely on DWT equivalent A*b
parameters.
• Hence critical step in using SMC and RBT Lite is
to define size scale-up between test and
average DWT size.
• Scale-up is ore-specific (1.2 to 2.7).
• Find that mineralogy and degree of alteration
explains the scale-up behaviour well.
 Effect of Size on A*b
70
83mm
60 42mm
coarser rocks appear
50 20mm
t10 ‘softer’ 10mm
40
5mm
(%)
30

20

10

0
0.01 0.1 1 10

Specific Input Energy Ecs (kWh/t)

 A*b versus Size Trends (SMC)
1.8

1.6

1.4

1.2
A*b Normalised

1.0

0.8

0.6

0.4
min gradient
-
0.2 mean gradient
average particle size for DWT = 32.6mm max gradient
0
0 10 20 30 40 50 60 70

Particle Size (mm)

A*b Terminology can be confusing!
• Standard is Drop Weight Test
– true A and b parameters, embedded in JKSimMet AG/SAG Model
• Full JKRBT Test can deliver statistically similar A and b parameters
– based on 3-year validation study
• All other tests estimate the full DWT parameters
– SMC (1 x size, 5 x Ecs)
– RBT Lite (1 x size, 4 x Ecs)
– RBT Express (1 x size, 1 x Ecs)
• Testing as close to average DWT size as possible reduces the
uncertainty in the A*b estimate.
 GeM Ci Test
• A rapid, low-cost comminution test for comparative
ranking of performance.
• It was noted that large volumes of core were being crushed
and pulverized for assays - opportunity to add value.
• Ci test is based on constrained lab crushing with analysis of
resulting size distribution - simple idea.
• After sizing, the sieved fractions can be returned for
assaying - minimal sample loss.
• Power meter provides additional value – quantify energy
consumed per unit mass.
 Examples of Ci Site Specific Models
Very strong correlations with Bond BMWi and A*b
when combined with mineralogy/XRF assays
Model with Ci Data + Mineralogy Model with Ci Data + XRF Assays
80
20.0

70

Predicted A*B (DWT equv)

18.0
BMWi pred (kWh/t)

16.0 60

14.0 50

12.0 40

10.0 30

Rel Error=±3.1% Rel Error=±6.0%

8.0 20
8.0 10.0 12.0 14.0 16.0 18.0 20.0 20 30 40 50 60 70 80

BMWi meas (kWh/t) Measured A*b (DWT equiv)

 RBT Express
The A*b Express test offers a simple and rapid methodology:
• One size fraction (typically ~10mm)
• One energy (0.2 kWh/t)
• T10 is calculated from the product using 1-point screening
• The slope (T10/Ecs) is the A*b Express index

38.1

19.8

T10 (%)
10.4
A*b =23
(hard ore)

4.4

Single
Single low
low energy
energy

0.2 0.5 1.0 2.0

Energy (kWh/t)

A*b – ore specific hardness index ( slope at zero energy)

 RBT Express Concept
Feed = M kg

T10
A *b 
0.2
M+1mm
Oversize
(+1mm) M1mm
JKRBT (0.2 kWh/t) T10  100  %
Mfeed

Undersize
(-1mm)
M-1mm
1mm Sieve
 RBT Express is Valid
Andacollo
Andacollo--SAG
SAGFeed
Feed Andacollo
Andacollo--Core
Core
80
80 80
80

70
70 70
70

60
60 60
60

50
50 50
50

Lite
Lite

A*bLite
A*bLite

40
40 40
40

A*b
A*b

30
30 30
30

20
20 20
20

10
10 10
10

00 00
00 10
10 20
20 30
30 40
40 50
50 60
60 70
70 80
80 00 10
10 20
20 30
30 40
40 50
50 60
60 70
70 80
80
A*b
A*bExpress
Express A*b
A*bExpress
Express

A*b Express test can provide a reliable A*b estimate using

a single test at 0.2 kWh/t
 Thank You

Questions ?