HOME

DEVELOPMENT OF NEW COMMINUTION TESTING

METHODOLOGIES FOR GEOMETALLURGICAL
MAPPING OF ORE HARDNESS AND THROUGHPUT
T Kojovic1, S Michaux2 and S Walters3

ABSTRACT
The emerging discipline of ‘geometallurgy’ is becoming increasingly recognised as a discrete and
high-value activity that reflects an ongoing trend towards more effective mine site integration and
optimisation. Constrained sampling that reflects and defines inherent ore body variability is a key
geometallurgical requirement. This requires use of larger numbers of low-cost physical testing which
can be applied to small sample volumes suitable for defining natural variability.
The AMIRA P843 ‘GeMIII’ project (Geometallurgical Mapping and Mine Modelling) is a major
industry-supported research initiative designed to develop new tools, methods and protocols to
support geometallurgical integration. As part of this integrated research a new more rapid low-cost
comminution test (GeM Comminution index) has been developed which can be employed as a front
line tool for geometallurgical mapping purposes and predictive throughput modelling. The test has
been designed to be inserted into routine assay sample preparation and is based on constrained
jaw crushing protocols linked to analysis of resultant size distributions. Extensive validation and
modelling has shown the GeM Comminution Index (Ci) is correlated with the Drop Weight index
A*b and Bond Mill Work Index (BMWi). Large scale trials have been conducted within a commercial
assay laboratory to demonstrate and optimise incorporation of the Ci test into routine sample
preparation protocols.
While Ci based estimates of A*b and BMWi are not as precise compared to larger volume more
expensive test-work, the ability to undertake large numbers of tests typically in a systematic down-
hole manner, provides a high level of data support through adjacency. A Ci based approach is highly
suitable for variability mapping, domaining and selecting large composite samples for more precise
testing. Within the context of the AMIRA P843 project this forms part of an integrated work flow
designed to support geometallurgical integration.
The development of a comminution test linked to routine assay sample preparation represents
significant value adding to a process that in most cases is already going to be carried out. This paper
introduces the Ci test concept and application to geometallurgical testing for throughput modelling.
Keywords: geometallurgy, comminution testing, crushing

INTRODUCTION
A key aspect of the AMIRA P843 GeMIII research project is development of a suite of new or
modified measurement and testing methodologies that can be used for early-stage geometallurgical
characterisation typically at core scale during feasibility. The aim is to provide more automated
‘machine vision’ capabilities to supplement or replace visual core logging approaches; develop more
flexible and cost-effective capabilities for automated mineral mapping and microscopy; and deliver
a suite of integrated small-scale and lower-cost physical tests for measuring processing performance
parameters. This approach must be capable of application to shared sample volumes ideally at the
scale of geochemical assays (Walters, 2009). For geometallurgical mapping and modelling large
data sets related to small sample volumes is a more effective statistical approach to defining natural
variability than a small number of ‘more precise’ data points. The end result is a multi-tiered sampling

1. MAusIMM, Research Consultant, Julius Kruttschnitt Mineral Research Centre, Sustainable Minerals Institute, The University of Queensland, 40 Isles Road, Indooroopilly Qld 4068, Australia.
Email: t.kojovic@jkmrc.uq.edu.au
2. Senior Research Fellow, Julius Kruttschnitt Mineral Research Centre, Sustainable Minerals Institute, The University of Queensland, 40 Isles Road, Indooroopilly Qld 4068, Australia.
Email: f.shi@uq.edu.au
3. Professor, Julius Kruttschnitt Mineral Research Centre, Sustainable Minerals Institute, The University of Queensland, 40 Isles Road, Indooroopilly Qld 4068, Australia.
Email: s.walters2@uq.edu.au

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 891
 T KOJOVIC, S MICHAUX AND S WALTERS

100.0
E H512_236 C I=1.61

P ercent R etained on S ieve (%)

E H512_290 C I=3.17
C E 107_364 C I=4.29
10.0
C E 107_176 C I=5.88

1.0

0.1
0.01 0.1 1 10 100
S iz e F raction (mm)

FIG 1 - Examples of Ci product size distributions and their corresponding Ci.

and testing strategy, with large numbers of relatively low-cost ‘comparative’ tests used to define
variability followed by small numbers of high precision ‘bankable’ tests representative of variability.
The GeM Comminution index test has been designed to support this approach through provision of
early comparative information on comminution performance suitable for geometallurgical mapping
(Walters and Kojovic, 2006).This paper outlines the concept behind the Ci test, its development
history and application for geometallurgical mapping of ore hardness and throughput.

DEVELOPMENT OF THE CI TEST

Jaw crushers are routinely used for size reduction as a rapid and low-cost preparation step for other
metallurgical tests or routine assay sample preparation. The Ci was designed to add value to routine
sample preparation for assays though use of constrained jaw crushing and modelling of resultant size
distributions (Figure 1). This data provides important information on breakage patterns and fines
generation. After sizing the sieved fractions can be returned for continued testing such as assays with
minimal sample loss. The Ci is a classic example of adding value to a process that in most cases is
already going to be carried out.
The primary purpose of the Ci is to provide large volumes of data on comparative comminution
behaviour suitable for spatial modelling and mapping. This typically involves relatively close spaced
down hole sampling which ultimately can be extended to match assay scale (Figure 2). This facilitates
statistical analysis of variability and trend analysis which can be used to define predicted processing
performance domains such as throughput.

FIG 2 - Example of continuous down hole Ci profile.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 892
NEW COMMINUTION TESTING METHODOLOGIES FOR GEOMETALLURGICAL MAPPING OF ORE HARDNESS AND THROUGHPUT

FIG 3 - Typical trend between Ci and standard ore hardness parameters A*b and BMWi.

The test has a quick turnaround (~10min/test) and has proven to be repeatable and robust across
a range of core sizes and geometries. Natural rock fragments can also be used as feed. The Ci has
undergone preliminary trials within a major commercial laboratory involving 13,000 samples as
part of routine assay sample preparation.
The resulting size distribution from the Ci test is used to derive two parameters: Ci (CRU) which is
related to the JKMRC Drop Weight Test A*b impact breakage parameter - range 0.0-6.0; and the Ci
(GRD) which has a good relationship with the Bond Ball Mill Work Index (BMWi) – range 0.0 to 1.0.
The typical trends are illustrated in Figure 3.
The circumstances where the Ci test protocol has proven to show limitations include:
• Friable core (pieces not the same size in each sample)
• High clay content in rock
• Variable feeding (choke vs single piece at a time)
• Crusher gap CSS is left unchecked to drift
• Natural fragments that have large variation in shape
An orientation study is recommended to determine the optimal application of the Ci testing protocol
for individual sites and sample mediums.
The Ci test carries four levels of QA/QC protocols relevant to the sieving required in the procedure
and sample response which are currently carried out in Excel software (Figure 4). These QA/QC

C S S S elec tion

P roc edure
C rus her T es t R es ults

Import Data
R eport

FIG 4 - Ci Excel software linked to LIMS system.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 893
 T KOJOVIC, S MICHAUX AND S WALTERS

tests have proven to be very effective as borne out in the laboratory Ci trial, identifying several cases
where the data was inadequate. These cases were typically related to holes in sieves, technicians
overlooking to tare the mass balance, and unusual samples that were outside the normal ore domain.
Ci repeatability
In order to asses the repeatability of the Ci method 164 samples were tested as part of the initial
development phase. The samples consisted of a number of ore and feed types that varied in size,
geometry, volume and fragment shape (Table 1). Drill core sizes tested included PQ, HQ, NQ and
BQ. Drill core shapes tested included whole, half and quarter core. A range of natural fragments
were tested according to stringent sample preparation - fragments were from defined size fractions
in the range of -63+13.2mm as produced using the 4th√2 sieve series. Results of the extensive testing
indicate that the Ci is a robust test that is relatively insensitive to sample geometry (Figure 5).
TABLE 1
Summary of comminution index development data base.

Origin Core Class Size (mm) Number of Tests

EH512 PQ 85 31
CE107 PQ 85 20
CE056 HQ 60 5
CE100 HQ 60 9
Diorite Non-Std 54 27
Diorite Fragment -31.5+26.5 6
Marble Non-Std 54, 28 17
Marble Fragment -31.5+26.5, husk 7
Granite PQ/HQ/NQ/AQ proxy 95, 60, 45, 30 22
Basalt Fragment Various Fractions 15
Phylite Fragment -63+53 5
TOTAL 164

The repeatability of the Ci was also tested in a separate experimental study. The reproducibility of
this test was excellent with a Coefficient of Variation within the same sample feed geometry (eg all
HQ half core, or all NQ half core) of 2.6 per cent (Figure 6). If the sample feed geometry was changed
within a sample (for example PQ core to BQ core to natural fragments) the Coefficient of Variation
increased to 10 per cent.

FIG 5 - Ci (CRU) results for a range of sample feed geometries.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 894
NEW COMMINUTION TESTING METHODOLOGIES FOR GEOMETALLURGICAL MAPPING OF ORE HARDNESS AND THROUGHPUT

FIG 6 - Repeatability of the comminution index (Ci-CRU).

RELATIONSHIP OF CI TO INDUSTRY ACCEPTED COMMINUTION PARAMETERS

Results indicate that the Ci-GRD offers a reliable index for estimating BMWi, with 8.7 per cent
average relative error across 190 samples from 7 mine sites and 10 development rock types (Figure 7).
The universal correlation with BMWi is very promising, but integration with supporting information
is also important at some sites for example where the mineralogy range is very wide or the core is
heavily broken.
The universal correlation of Ci-CRU with A*b shows lower level of correlation compared to BMWi
estimates (Figure 8). It is also apparent that for some rock types with low A*b (ie hard crush) the Ci
shows a greater variation than A*b. For example, in Figure 8 at a Ci value greater than 3.5, the A*b
values have a limited range between approximately 20 and 50, however the Ci shows a much greater
range. The significance of this relationship is still under investigation.
Outcome of work to-date indicates that universal correlation of Ci-CRU with A*b is clearly less
accurate than correlation with BMWi. Work is underway in the AMIRA P843A extension project to
improve the correlation accuracy of the correlation. One option under investigation is the use of a
power meter to quantify the energy consumed in crushing. Deposit specific calibrations using RBT or

30

25

20
P red B MWi (kWh/t)

15

10 Boddington
S udbury
Cadia Eas t
Ernes t Henry
A qqaluk
5 Res olution
KUC
Develop1
Develop2
ideal
0
0 5 10 15 20 25 30
Me a s B MW I (kW h/t)

FIG 7 - Quality of the prediction of bond mill work index BMWi from Ci-GRD.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 895
 T KOJOVIC, S MICHAUX AND S WALTERS

400
Cadia East
350
Ernest Henry

Aqqaluk
300
Boddington
A*b (DWT Normalised)

250 Sudbury

Bingham
200

150

100

50

0
1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5
JKCi CRU

FIG 8 - Relationship between A*b (normalised by SG) and Ci-CRU.

SMC/DWT tests are therefore required to improve the accuracy of the correlation. The best accuracy
possible with site based calibrations, results in an average relative error of 16.2 per cent across 422
samples from 4 mine sites (Figure 9).
Experience has shown that universal calibrations are unlikely to be valid and each deposit has
different geological factors. As such scoping and orientation studies for Ci testing are required for
individual sites. This typically involves 20 - 30 samples for calibration which need to be carefully
selected in consultation with site personnel. The samples should represent an expected dynamic
range and must be integrated with other geometallurgical assessment criteria. If possible, sufficient
sample mass should be acquired for repeat tests.
Linking Ci testing to throughput modelling
To illustrate the application of the Ci in geometallurgical mapping of throughput a large scale
comparative study was carried out at a P843 sponsor site. This involved comparing the results of a
large underground bulk sample extracted from a drive and tested using conventional methods for
A*b and BMWi with the results of three drill holes drilled along the drive. 443 Ci tests were conducted
at two meter assay intervals along the drill core and used to estimate BMWi and A*b for each sample.

300
Ernes t Henry
A qqaluk
250 Boddington
Cadia Eas t
ideal
200
P red A *b

150

100

50

0
0 50 100 150 200 250 300

Meas A*b

FIG 9 - Quality of the current universal prediction of A*b from Ci-CRU.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 896
NEW COMMINUTION TESTING METHODOLOGIES FOR GEOMETALLURGICAL MAPPING OF ORE HARDNESS AND THROUGHPUT

The mill throughput was estimated using a set of engineering equations developed from 30 operating
AG/SAG mills (closed and open circuit), covering a range of diameters, lengths, speeds, ball loads,
grate designs, ore characteristics and feed size distributions. These models are similar to the models
developed by Morrell (2004) for scale-up, design and optimisation. The equations have the form:

SP(kWh/t) = SAG(kW)/TPH = f´(F80,T80,SG,DWAB,BMWi,BL,CS,D,D/L)

where:
SP = AG/SAG specific power required for given set of operating conditions
F80 = SAG mill feed 80 per cent passing size
T80 = SAG Mill product 80 per cent passing size (or transfer size)
SG = ore specific gravity
DWAB = Drop Weight specific power required to break feed to transfer size, which is dependent on
ore A and b parameters, feed size distribution and trommel size.
BMWi = Bond ball mill grindability index
BL = ball charge (per cent)
CS = mill speed (per cent of critical)
D = mill diameter
D/L = aspect ratio = mill diameter/mill length (EGL)
The above equations are combined with the Bond ball mill design equation (1961) to complete
the circuit throughput model, providing a link between the SAG and ball milling stages. That is, the
throughput capacity is dependant on both the SAG and ball mill performance as they interact in
practice (see Figure 10).

BallMill(Kw)
TPH =
⎛ 1 1 ⎞
10 × BMWi × (EF1 × EF2 × EF3 ….) × ⎜ – ⎟
⎝ P80 T80 ⎠

where:
TPH = expected feed capacity (tph)
PBM = installed ball mill power (kW)
EF1, EF2 … = Bond’s efficiency factors
T80 = SAG circuit transfer size
P80 = final ball mill circuit product size.
A *b + B MWi

P ebble
crus her tph
P 80

tph S AG
(kW)
F 80

BM
tph (kW)
T 80
B MWi

FIG 10 - Conceptual interaction of SAG and Ball Mill circuits

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 897
 T KOJOVIC, S MICHAUX AND S WALTERS

FROM
CRUSHING

TO FLOTATION

Pebble
Crusher
COARSE ORE STOCKPILE MP1000
2 x 750 kW
Ball Mill
22 ft. x 34 ft.
9.0 MW TO FLOTATION

Ball Mill
SAG Mill 22 ft. x 34 ft.
40 ft. x 22 ft. 9.0 MW
20.0 MW

FIG 11 - SABC flowsheet used in GEMIII throughput modelling

The maximum throughput may be SAG or Ball Mill limited, depending on ore conditions.
The sensitivity of the Excel model to A*b was fine tuned using JKSimMet (Wiseman and Richardson,
1991) to better capture the effect of recycle load and pebble crushing on transfer size to the ball mill.
This modification provides a correction to the SAG transfer size T80 (from the baseline) when A*b
changes at constant BMWi and increasing recycle fraction.
The effect of A*b and feed size F80 is included in the DWAB term. However, ideally the complete
feed size distribution is required to determine DWAB. The simplification presented here assumes
the shape of the distribution remains constant when the F80 changes as part of a comparative
geometallurgical mapping type approach.
The estimated throughput is based on the existing SABC circuit, consisting of an open circuit SAG
followed by two ball mills in closed-circuit with hydrocyclones (Figure 11).
The results for the bulk ore sample treated through the existing SABC plant are compared with the
Ci/RBT-based estimates in Table 2.
TABLE 2
Comparison of actual and forecast ore parameters and plant throughput.

BMWi Bulk Ore GeMIII Plant

Sample A*b
(kWh/t) Plant Throughput TPH Forecast
Bulk Ore Sample (ave) 28.9 20.8 1400 1417
Ci tested Hole 1 – ave 29.7 20.2 - 1458
Ci tested Hole 2 – ave 32.0 20.3 - 1460
Ci tested Hole 3 – ave 30.8 20.4 - 1451

The Ci/RBT based throughput estimates were based on the estimated ore hardness parameters
(A*b and BMWi), actual mill operating powers during the plant trial, and surveyed grind size (P80)
and SAG feed size (F80). As can be seen (Table 2) the mill throughput estimates are consistent
with the plant trial on the bulk ore sample. This is an encouraging validation which illustrates the
potential for embedding Ci-derived throughput predictions into the entire resource model.

CONCLUSIONS
The Ci has proven to be a fast and reproducible comparative comminution test designed specifically
to support large-scale geometallurgical mapping and modelling. As such it is not designed as a
replacement for existing higher precision tests used for ‘bankable’ engineering design. Insertion of
the Ci test into routine assay preparation as shown through a large scale laboratory trial provides an
opportunity to collect the data during feasibility studies. This represents significant value adding to
routine drill core testing. Estimates of BMWi from extensive Ci development and validation studies

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 898
NEW COMMINUTION TESTING METHODOLOGIES FOR GEOMETALLURGICAL MAPPING OF ORE HARDNESS AND THROUGHPUT

show a high level of correlation suitable for geometallurgical domaining. Estimates of A*b show a
lower level of correlation which is currently undergoing further research to improve the models.
Within the emerging field of geometallurgy rapid ‘front line’ testing methods such as the Ci provide
the opportunity to support a new approach to definition of inherent geological variability in terms
of processing performance parameters. This is part of an integrated methodology being developed
within the AMIRA P843 GeMIII research project in close collaboration with the global minerals
industry.

ACKNOWLEDGEMENTS
The authors acknowledge financial support and permission to publish from industry sponsors of
the AMIRA International P843 GEMIII Project – Anglo Gold Ashanti, Anglo Platinum, Barrick, BHP
Billiton, Codelco, Datamine, Golder Associates, GEOTEK, ioGlobal, Metso Minerals, Newcrest,
Newmont, Oz Minerals, Peñoles, Rio Tinto, Teck Cominco, Vale, Vale Inco and Xstrata Copper.

REFERENCES
Bond, F C, 1961. Crushing and Grinding Calculations, Allis-Chalmers publication, No. O7R9235B.
Kojovic, T, Fengnian, S, Larbi-Bram, S and Manlapig, E, 2008. JKRBT, Any Ore, Any Mine, MetPlant 2008
Conference, Perth, Australia, 1 - 14.
Morrell, S, 2004. Predicting the Specific Energy of Autogenous and Semi-Autogenous Mills from Small
Diameter Drill Core Samples, Minerals Engineering, Vol 17/3, pp 447 - 451.
Napier-Munn, T J, Morrell, S, Morrison, R and Kojovic, T, 1996. Mineral Comminution Circuits: Their
Operation and Optimisation: Brisbane Australia, JKMRC University of Queensland, JKMRC Monograph
Series in Mining and Mineral Processing 2, 413 pages.
Walters, S G, 2009. New Research Initiatives in Geometallurgical Integration – Moving Towards a Common
Operating Language. In proceedings of the 7th International Mining Geology Conference, 17 - 18th August,
Perth – AUSIMM.
Walters, S and Kojovic, T, 2006. Geometallurgical Mapping and Mine Modelling (GeMIII) –The way of the
Future. Proc.SAG2006 Conference, Vancouver, Vol IV, 411 - 425.
Wiseman, D M and Richardson, J M, 1991. JKSimMet - the mineral processing simulator. Proceedings 2nd
Can Conf on Comp Applications in the Min Ind, (Eds Paulin, Pakalnis and Mular), Univ B.C. and CIM,
Vol II, pp427 - 438.

XXV INTERNATIONAL MINERAL PROCESSING CONGRESS (IMPC) 2010 PROCEEDINGS / BRISBANE, QLD, AUSTRALIA / 6 - 10 SEPTEMBER 2010 899