You are on page 1of 8




ABSTRACT production forecast. The main sources chemistry parameters. A key to practical
of error are in the flotation test itself, the geometallurgical modelling is having a
The key to successful flotation plant geostatistical error (related to sample proven and affordable test that provides
design, production planning and mine/ density), and the model calibration error. the required data from small amounts of
mill optimisation is a solid understanding The latter is determined through plant available exploration drill core. This 3
of the resource to be processed. benchmarking. makes it possible to conduct enough
As advocated by the authors of this tests to attain metallurgical validity for
paper, the main components of INTRODUCTION insertion in the geological model. The
geometallurgical modelling of an orebody first section of the paper covers the
and its associated flotation plant are: Many mineral extraction operations are various modes for carrying out the
an ore sampling program and challenged on a daily basis with feed ore MinnovEX Flotation Test (MFT) and the
subsequent laboratory testing of these variability that impacts the throughput of information attained from these tests.
samples in order to extract process the grinding section and the subsequent
model parameters, concentrate grade and recovery that can Furthermore, the authors recognise that
geostatistical distribution throughout be produced. A key to maximising the a flotation circuit does not operate in
the resource model of the process value of a mining and processing facility isolation, but rather is strongly influenced
parameters, is to develop a prior understanding both by the feed ore delivered to the
calibration of the plant via of the characteristics of the feed ore mill and by the manner in which the
benchmarking (for existing operations), over the lifecycle of the operation from preceding grinding circuit processed that
and design and financial evaluation of the ore. In other words, a link is required
plant simulation using a system of project, through to monthly and yearly through geometallurgically enabled
process models and the distributed production forecasts and on-going circuit comminution and flotation simulation
metallurgical parameters (from step b) troubleshooting and optimisation. This tools. Considering just the flotation
as the data set. is achieved through the application of characteristics of the feed ore presents
geometallurgical modelling of an ore body. a partial picture; also of significant
It is important that the grinding and importance is knowledge of how the
flotation models are linked. For example, The main components of geometallurgical rock behaved in the grinding circuit from
a laboratory test conducted on a drill modelling are: a throughput and grind perspective.
core sample, intended to represent a an ore sampling program and Section 2 of the paper demonstrates this
portion of the orebody, is conducted at subsequent laboratory testing of these concept through a step-bystep example
a specific grind (represented by a P80). samples in order to extract process that tracks various blocks of ore from pit
However, when that ore is actually model parameters, through grinding to flotation feed.
processed through the plant it may geostatistical distribution throughout
well be (and in many cases most likely the resource model of the process The third of the paper section briefly
will be) at another P80-. Additionally, the parameters, discusses precision analysis. A benefit
flotation plant residence time will often calibration of the plant via of a statistical approach to geometallurgical
be determined solely by the 2 grinding benchmarking (for existing operations), modelling is that the precision of
circuit capacity and feed slurry density. and a design or production plan can be
In the case of SAG mill ball mill circuits, plant simulation using a system
the fluctuations in tonnage and grind of proven process models and the
are known to be high. The modelling distributed metallurgical parameters
approach described in this paper allows (from step b) as the geometallurgical
for changes in the measured flotation data set.
kinetic parameters in order to reflect the
expected grind as determined by the This paper will focus on the application
comminution process. of geometallurgical modelling concepts
to flotation circuit design and production
A valuable aspect of the geometallurgical forecasting.
modelling approach is the ability to
identify and quantify sources of error The first section of this paper describes
to obtain the precision of a design or the method of quantifying pulp

quantified through error analyses. With program that is simple, quick and can The pulp kinetic parameters are
modelling tools, laboratory tests and be done at the lowest possible cost determined from the MFT results by
geostatistics, there are quantifiable therefore allowing as many tests as applying a comprehensive parameter
sources of error at each step. This possible to be performed on individual extraction methodology that decouples
section of the paper provides an overview drill core samples to characterise the true flotation and entrainment, and
of error sources in each step of the flotation properties of the ore. then models the MFT according to
geometallurgical modelling process fundamental flotation principles [Dobby,
[Lozano, Bennett, 2003]. The objective of the MFT is to measure Savassi, 2005]. The analysis of each MFT
the pulp kinetics for each of the yields the following information, for each
bench-scale flotation tests mineral species in the ore at a set (pre- mineral (or minor element) of interest:
determined) reagent suite. The MFT is The maximum recovery (Rmax) and
Geometallurgical modelling can be designed to determine the kinetics cumulative frequency distribution of
defined as an approach that measures of mineral separation in the pulp phase rate constants for each mineral species
metallurgical variability within an ore exclusively, while froth effects in the at the test grind (described by Kavg and
body and quantifies the effect of this plant are later accounted for through alpha, the latter being a descriptor of
variability on the comminution and FLEET (Flotation Economic Evaluation the spread of rate constants).
metallurgical response of the ore, as it Tool) modelling. Further details on the The quantitative effect of grind on Kavg
pertains to full-scale production. The MFT can be found elsewhare [Dobby, and Rmax.
geometallurgical approach surpasses the Kosick, Amelunxen, 2002; Dobby, A standard set of Rmax and Kavg values
resource model as it encompasses not Savassi, 2005]. at a single grind-size2 common for the
only the variability of alteration, lithology full set of drill core samples, together
and metal grades, but also metallurgical In order to perform multiple tests in a with Rmax-slope parameters, all of which
process specific characteristics that quick and cost-effective manner, the may be distributed across the mine
influence the grinding and mineral MFTs are carried out in two forms: the block model (Rmax-slope is the change in
recovery characteristics of the ore full MFT and the mapping MFT. Once Rmax per one micron change in P80).
[Bennett, Lozano, 2004]. sufficient information about the kinetics
of the mineral components in the ore Table 1 summarises an example set
Since the success of the approach body is known (through completion of of pulp kinetic parameters that would
hinges on how well the variability of the sufficient full MFTs), then the method be determined for a typical copper
ore is represented, it is obvious that, for can be simplified for mapping purposes. porphyry material. A typical set of pulp
geometallurgical modelling purposes, The simplification resides mainly in kinetic parameters derived from an MFT
many data points from small scale reduced degree of screen analysis, performed on a drill core sample ground
tests are more valuable than a small which thereby reduces the number to 128m is shown on the left side of
number of detailed investigations. of samples submitted for chemical the table. The kinetic parameters from
Therefore, a standard low-cost test analyses. This can significantly lower the each MFT need to be reconciled to a
that can be performed on a small mass overall test cost, with little compromise standard grind before the parameters
of drill core material is essential. In in the quality of the results. Hence, more are distributed across a mine resource
4 order to maintain the integrity of drill core samples can be tested at the model. Therefore, the terms Rmax-slope
the geometallurgical model, it is also same project cost - ultimately leading to and Kavg-slope are used to correct the set
important that the parameters from each overall better accuracy in both design and of test kinetic parameters to a set of
test are primary1 characteristics that can production forecasting. standard kinetic parameters, as shown
be distributed across an ore block model on the right side of the table.
using standard geostatistical methods. At the commencement of a drill core test
Describing the floatability of an ore program, a decision would be made on
through a set of pulp kinetic parameters the proportion of full MFTs and mapping
for each of the mineral component of MFTs. This decision is based on the
the ore means that both the concentrate prerequisite that sufficient full tests are
grade and recovery can be calculated performed per mineralogy/ore type, in
simultaneously using fundamental order to calibrate the key interpolation
flotation modelling methods. relationships needed for the mapping
MFT parameter extraction. When the
The MinnovEX Flotation Test (MFT) has mineralogy/ore type is simple, the split
been developed to satisfy both of the is typically around 15 % full MFTs to
above requirements. It is a standard 85 % mapping MFTs. In more complex
bench-scale test that is used to measure ore-bodies, with more variable lithology, 1
Primary data is defined as ore-specific parameters,
which in the case of flotation describe the inherent
the primary floatability characteristics of alteration and/or mineralogy, the floatability of the ore regardless of the plant
proportion of full tests to mapping tests flowsheet, equipment or operating targets and
a sample of ore. The philosophy behind conditions.
MFT work for samples is to have a test has to be increased. 2
Described by P80 and the Rosin-Rammler slope, or
m value

testwork grind m = 0.93 95.8standard grind m = 0.70

P80 = 128 for fleet simulation P80 = 100
kinetics at testwork conditions kinetics at standard grind

CuSulph 91.7 1.5 1.3 CuSulph 92.9 1.5 1.3 -0.041 -0.004
Pyrite 96.0 2.8 3.2 Pyrite 95.8 2.8 3.2 0.002 0.004
Au 71.5 1.0 1.2 Au 73.6 1.0 1.2 -0.058 -0.002
Ag 80.0 0.6 4.3 Ag 81.9 0.6 4.3 -0.065 -0.001
NSG 3.9 0.3 10.0 NSG 2.8 0.3 10.0 0.023 0.001

Table 1. Typical set of pulp kinetic parameters derived from an MFT

Selecting the Drillcore samples for Metallurgical testwork

Communition and Flotation testwork

Extraction of ore-specific primary parameters for
geostatistical distribution

Population of the mine block model using Geostatistics to
distribute the extracted parameter

CEET simulation to predict TPH FLEET simulation to predict final
and P80 on a block-by-block basis concentrate grade and recovery
on a block-by-block basis

Mine / Plant
Production planning /
Design Mode Optimisation
forecasting Mode

Figure 1. Creating a Geometallurgical Model


CIRCUIT SIMULATION been selected, it can be seen that the
actual grindsizes (P80) at which the three
The Rmax-slope and Kavg-slope parameters presented in Table 1 provide the link between tests were performed were 65, 108 and
drill core flotation testwork results and FLEET flotation circuit simulation on a block-by- 128m, respectively. DC-1 is the softest
block basis for design and/or production forecasting studies. of the three samples and as a result, the
laboratory mill over-ground this sample.
A geometallurgical model is created by following the steps outlined in Figure 1 Likewise, DC-3 is a much harder sample,
[Dobby, Bennett, Bulled, et al, 2004]. and the desired grind was not attained in
the specified grinding time. Even when
The importance of accounting for the effect of grind-size on recovery from the initial attempts are made to account for the
grind at which the MFT test is performed to the ultimate prediction of flotation circuit Bond value in selecting the lab grind time,
recovery and grade per ore block is illustrated in a step-by-step example. these differences often occur. However,
a unique aspect of the MFT is that the
STEP 1 The drill core samples for metallurgical testwork are selected over the region kinetic parameters measured at each test
of interest within the resource. A grid pattern over the area of interest (for example, grind can be reconciled to the standard
the first five years of operation) is preferred, ensuring, as a minimum, that all major grind through the Rmax_slope function.
geological features of the resource are represented. More complex geological conditions
will dictate a greater number of samples. Clearly, the joint involvement of geological, STEP 4 The grindability parameters and
mineralogical and metallurgical staff in sample selection is required. floatability parameters at standard grind
are distributed throughout the blocks
STEP 2 Bench-scale laboratory testwork is carried out on each of the drill core samples. in the region of interest in the mine
For SAG hardness tests (MinnovEX SPI) typically 2 kg of drill core is required. For ball resource model using geostatistical
mill hardness measurements, 1.2 kg for MinnovEX Modified Bond tests or 10 kg for full techniques (involving the consideration
Bond tests is required [Kosick, Bennett, 1999]. Since full Bond tests are only carried out of sample location, ore type and grade).
on 5 to 10 % of the samples, with Modified bond tests performed on the remaining An outcome is illustrated in Figure 2.9
samples, the core requirement is 12 kg for up to 10 % of the samples and only about Early in a project development, often
4 kg for the other 90 %. there will be insufficient samples tested
to develop a geostatistical distribution.
The MFT requires no additional drill core as this test is performed on the products from In this situation, the distribution of
the SPI and Bond dry grinding tests. At the outset of the testwork program, a standard parameters is made along either a
grind-size for the feed for the MFT is selected. Based on initial hardness information regional basis or ore type basis.
for each drill core sample (SPI and BWi), an estimate is made of the required grinding
time to attain the standard grind-size for the MFT, and each sample is wet ground in Figure 2 is a sample region of a mine
a laboratory mill for the 8 specified length of time. MFTs are then carried out on all resource model, demonstrating areas
samples, with product streams analysed either according to the full MFT or mapping of higher and lower theoretical recovery
MFT protocols. of copper sulphide minerals, after
geostatistical distribution of the
STEP 3 The primary ore parameters, shown in Table 1, which describe fundamental Rmax_cusulph parameter across the
grindability and floatability of the ore samples, are extracted from the lab data according resource model.
to model fitting techniques [Dobby and Savassi, 2005]. Table 2 provides a Case Study
selection of typical parameters measured for three random drill core samples (we will
follow the three samples through the next few steps of the process). Note that this table
does not show the complete set of parameters, but only a selection for illustrative
purposes - the full set of floatability parameters per sample would be as per Table 1.

Drill SPI BWi MFT Rmax_ Rmax_ Rmax_

core [min] [KWh/t] grind cusulf cusulf cusulf
LABEL (at Test (at Std Slope
grind) grind) #
[%] [%]
DC-1 28 12.4 65 81.0 77.6 -0.090

DC-2 35 11.7 108 94.3 94.6 -0.023

Figure 2: Sample region of a mine resource model
DC-3 91 15.0 128 91.7 92.9 -0.041

Table 2: A selection of typical grindability and floatability parameters for three random drill core samples.
Note: A standard target grind of 100 m had been selected.

Step 5 At this stage, each block in the geometallurgical model has been attibuted the The kinetic parameters at standard grind
interpolated ore-specific parameters, necessary for grinding simulation. Within (100 m) are distributed through the
a web-based simulation platform, the grinding simulation software mines the mine block model. However these values
resource model on a blockby-block basis, calculating the grinding circuit throughput are adjusted on a block-by-block basis
(TPH) and product grind (P80) for each block. Example grinding simulation results for according to their Rmaxslope values and
the three blocks of ore (samples) listed in Table 2 are given in Table 3. (In this table is is the P80 value calculated for that block.
assumed that DC-1 was located in Block 1, DC-2 in Block 2 and DC-3 in Block 3, and a The slope values are also distributed
typical SAG ball mill circuit was used for the simulation.) across the block model and are therefore
available on a block-by-block basis. The
Spatial spi bwi tph P80 case note actual Rmax values used in the FLEET
location [min] [microns] simulation are adjusted according to the
difference in P80 between the standard
BLOCK 1 x1 y1 z1 28 12.4 5185 120 Ball-mill (i) grind and the block grind and the slope
limited for that block.
BLOCK 2 x2 y2 z2 35 11.7 5250 109 Down (ii)
It is important to make the correction for
the difference in grind-size between that
produced in the MFT to that attributed
BLOCK 3 x3 y3 z3 91 15.0 4436 89 SAG (iii) to each block of ore, to achieve accurate
Limited design and production forecasting.
Table 3: Typical CEET simulation results for three blocks of ore Table 5 highlights this by showing the
original test grind for a drill core sample
Note the following: compared to its associated block grind,
i. In this scenario, Block 1 comprises soft ore, therefore affording the opportunity and the impact that this has on the
to maximise tons by producing a coarser grind. A limit of 120 m (based on coarse prediction of ultimate recovery (Rmax)
fraction results of flotation tests) has been set on how coarse the grind is allowed for that block. Modelling is conducted
to become, thereby limiting the throughput for Block 1 to 5185 tph. by utilising the kinetic parameters of all
ii. For Block 2, neither the SAG mill nor the Ball mill has become the limiting element minerals.
due to this blocks lower BWi. Throughput therefore can be maximised up to a limit
typically set by some constraint external to the grinding circuit (typically tailings
pumps or concentrate filter capacity dictates this limit on maximum throughput). At
this maximum throughput of 5250 tph the grinding circuit can produce a grind of
109 m.
iii. Block 3 comprises the hardest ore and the grinding circuit is SAG mill limited. The
maximum tonnage that can be treated through the SAG mill is only 4436 tph before
the mill would overload. At this lower tonnage, the ball mill circuit produces the finer
grind of 89 m.

STEP 6 The tonnage and P80 results from the grinding simulation become input values
to the flotation circuit simulation. Similar to Step 5, the FLEET software tool calculates
final concentrate grade and recovery on a block-by-block basis based on the tonnage
and grind feeding the flotation circuit for the block, and the floatability characteristics
attributed to that block. Table 4 provides the set of feed parameters for the same three
blocks of ore discussed in Step 5.

Spatial tph P80 Rmax_ RMAX_ Actual Rmax_

location [microns] CuSulf CuSulf CuSulf
(at Std SLOPE used in FLEET
grind) simulation
BLOCK x1 y1 z1 5185 120 77.6 -0.090 75.8
BLOCK x2 y2 z2 5250 109 94.6 -0.023 94.3
BLOCK x3 y3 z3 4436 89 92.9 -0.041 93.4
Table 4: Typical FLEET feed values for three blocks of ore

Grind at which MFT Rmax_CuSulf value Grind-size attributed to the Actual Rmax_CuSulf
test was performed extracted from MFT block of ore after CEET value for that block,
[microns] test results [%] simulation [microns] used in FLEET simulations
BLOCK 1 65 81.0 120 75.8
BLOCK 2 108 94.3 109 94.3
BLOCK 3 125 91.7 89 93.4
Table 5: The actual Rmax_cusulph used in FLEET simulations, compared to the values established from the MFT test

PRECISION ESTIMATION FOR RIGOROUS sample of copper ore using 3 trained

Ore type Std error % in CuSulf
GEOMETALLURGICAL MODELLING operators is shown in Table 6.
Standard error on the CuSulph Rmax is
A statistical approach to geometallurgical less than 1% and on the NSG Rmax R max k avg
modelling presents the user with an about 0.5%. This could obviously 1 2.3 0.9
opportunity estimate precision on plant improve with operator experience. 2 1.3 0.9
circuit design or production forecasts. 3 1.7 0.8
For the model to be used for production
OPERATOR 1 2 3 Table 7: Standard errors in block parameter estimates
planning and/or design purposes, it can
only be considered rigorous and CuSulf
definitive if accompanied by an Rmax 90.5 91.2 89.5 These figures can be reduced by
associated precision model.
Kavg 1.6 1.6 1.6 applying geostatistical methods, such
ALPHA 2.4 2.2 1.9 as Kriging, for distribution of data
A complete geometallurgical modelling
NSG across the mine blocks. Such methods
study comprises the following main
would involve the determination of
components [Dobby et al, 2004]: Rmax 6.5 6.9 5.7 variograms to establish the range of
A bench-scale laboratory testwork Kavg 0.4 0.3 0.3 influence of each sample to
program, neighbouring mine blocks. Any
ALPHA 10 10 10
Geostatistical distribution over the significant reduction in the number of
resource model of the parameters Table 6: Reproducibility of MFT parameters
samples or increase in the variability of
extracted from the bench-scale test, the ore will naturally result in an
Plant benchmarking for model increase in the estimated standard
calibration (for existing operations), errors. Distribution of data across
In the study of a copper porphyry ore
and the mine blocks is usually the greatest
body where full MFTs were conducted
Circuit simulation using the source of error in forecasting due to
on 15% of the samples, with mapping
distributed metallurgical parameters insufficient samples and testing.
MFTs on the remainder, the
and a system of process models, The precision of the process model
estimation of CuSulph kinetics for the
calibrated according to plant is obtained from plant calibration
mapping tests was found to introduce
measurements. work. The model error is the difference
a standard error of 1.5% for Rmax and
0.1% for Kavg. This could be improved between the measured values of
Each of these steps presents a source plant concentrate grade and recovery
by the development of the estimation
of error that can be accounted for and and the forecast values determined
techniques used in the mapping MFTs.
can be summarized as: by application of the model to the MFT
160 MFTs were conducted on drill core
1. Measured (test) precision and parameters of the feed. The standard
samples to represent the approximately
kinetics estimation precision error has been shown to be within 1%
1.5 million tons of ore to be treated
2. Distribution (geostatistical) precision on large copper operations. The higher
in 1 year in a large copper plant. The
3. Model calibration precision the number of benchmark calibration
kinetics data was distributed across
the mining blocks with respect to feed surveys carried out, the more accurate
The authors have considered the the process calibration becomes.
grade and ore type of the blocks. The
precision associated with each of these
standard error in estimating the
steps. Selected examples follow: The total standard error on the kinetics
individual block values of the
A comparison of the extracted parameter estimation can be determined
distributed data for CuSulph is shown
parameters from testing a single by normal statistical methods for error
in Table 7.

accumulation and the FLEET model is The success of CEET led to the CONCLUSIONS
then run as a Monte Carlo simulation to development of FLEET and the MFT
estimate precision on the projected plant for the application of geometallurgy It is the authors belief that the
results. Model precision must then be to industrial flotation. The final step to geometallurgical modelling approach
accounted for to give an overall picture of link SPI, Bond, CEET, FLEET, the MFT for flotation circuits, as described in this
the precision of the technique. and future process models, was an paper, is the leading scientific method
Internet portal that could host a common to quantify the impact of ore variability
Finally, however precise, the process geometallurgical dataset for any on the metallurgical performance of
model must be geometallurgically particular project or resource. This was industrial flotation circuits. The
enabled in order to generate meaningful important from two perspectives. First, quantification of ore variability linked
results, i.e. it must be capable of inputs from a common geometallurgical with geometallurgically enabled process
automatically processing all the blocks dataset for a particular project would be models opens the opportunity for more
in the resource model and therefore used by all geometallurgically enabled accurate circuit design and the most
able to optimise designs or forecast models. Also, the output from any reliable production forecasting method
results based on the complete resource particular run of one process model that is currently available. Finally, there is
or geometallurgical model. The process would form part of the input for the considerable opportunity to apply these
model must also be capable of process model for the subsequent geometallurgical technologies to
accommodating variations in grind downstream process, and the data optimisation studies and advanced
and tonnage for each block in the transfer and integrity had to be handled process control to maximize the value
geometallurgical model. If the process through this common dataset. Second, from mine to mill.
model is not geometallurgically enabled, by configuring this interface and dataset
then it becomes relatively ineffective handling ability through the Internet, on- REFERENCES
for industrial design or production line support worldwide could be made
forecasting. available continually and inexpensively. Bennett, C, and Lozano, C, 2004.
This Internet portal and interface was The Architecture of the Geometallurgical
PROGRESS OF GEOMETALLURGICAL developed in 2002 for the mining Model, in Proceedings Procemin, 2004,
MODELLING FOR FLOTATION industry and is called Process Access pp T 1-8, 2004, Santiago
[Kosick, Bennett and Dobby, 2002].
The first holistic use of geometallurgical It is anticipated that as more mining Dobby G, Bennett C, Bulled D, and
modelling began in 1999 when the companies become aware of Process Kosick G, 2004 Geometallurgical
process model CEET was developed. Access, it will become a standard Modelling The New Approach to Plant
Prior to CEET there was no mechanism interface method for facilitating Design and Production Forecasting/
for exploiting the block comminution data geometallurgical work worldwide. Planning, and Mine/Mill Optimization,
in a resource model for optimisation of in Proceedings 36th Annual Meeting of
plant design or production forecasting Since the completion of FLEET in 2002 the Canadian Mineral Processors, 2004,
and planning. CEET was developed using and until the time of writing this paper pp 227-240 (The Canadian Institute
a geometallurgically enabled structure FLEET geometallurgical modelling of Mining, Metallurgy and Petroleum:
in order to utilize the SPI and Bond studies have found wide application in a Ottawa)
data block by block throughout the range of mineral processing plants and
resource model to design and optimise projects world-wide for copper, lead, Dobby G, Bennett C, and Kosick, G, 2001
comminution circuits. zinc, nickel, gold, coal, molybdenum and Advances in SAG Circuit Design and
iron ore flotation. FLEET has been used Simulation Applied to the Mine Block
The development of a geometallurgically for circuit design, flotation plant Model, in Proceedings SAG 2001, pp
enabled comminution process model troubleshooting, optimisation, and IV 221-234 (Department of Mining and
was one of the critical steps required to production forecasting. Over 100 Mineral Process Engineering, University
facilitate the development of an accurate calibration-benchmarking surveys have of British Columbia: Vancouver)
geometallurgically enabled process been carried out in 17 concentrators.
model for flotation. Knowledge of the Dobby G, Kosick G, and Amelunxen
throughput and grind parameters per Due to the limitation on paper length, R, 2002 A Focus on Variability within
block (which translate to residence time case studies using FLEET will be the Ore body for Improved Design
and liberation in the flotation circuit) is presented in a subsequent paper. of Flotation Plants, in Proceedings
very important if the geometallurgically 34th Annual Meeting of the Canadian
enabled flotation process model is to Mineral Processors, 2002, pp 77-90 (The
be accurate enough to use for design Canadian Institute of Mining, Metallurgy
or production forecasting (which is and Petroleum: Ottawa)
the ultimate test of a geometallurgical

Dobby G, and Savassi O, 2005 An Lozano, C, and Bennett, C, 2003.

Advanced Modelling Technique for Geometallurgical Modelling Applied to
Scale-up of Batch Flotation Results to Production Forecasting, Plant Design and
Plant Metallurgical Performance, in Optimization, in Proceedings Procemin,
Proceedings Centenary of Flotation 2003, pp T11 1-11, 2003 (Universidad de
Symposium, 2005, in press (The Chile: Antofagasta)
Australasian Institute of Mining and
Metallurgy: Melbourne)

Kosick, G, and Bennett, C, 1999 The

Value of Orebody Power Requirement
Profiles for SAG Circuit Design, in
Proceeding 31st Annual Meeting of the
Canadian Mineral Processors,1999, (The
Canadian Institute of Mining, Metallurgy
and Petroleum: Ottawa)

Kosick, G, Bennett, C, and Dobby,

G, 2002 Managing Company Risk by
incorporating the Mine Resource Model
into Design and Optimization of Mineral
Processing Plants, in Proceeding 104th
Annual General Meeting of the CIM,
2002, (The Canadian Institute of Mining,
Metallurgy and Petroleum: Vancouver)

CONTACT INFORMATION 2011 SGS. All rights reserved. The information contained
herein is provided as is and SGS does not warrant that it
will be error-free or will meet any particular criteria of per-
formance or quality. Do not quote or refer any information
Email us at herein without SGS prior written consent. Any unautho-
rized alteration, forgery or falsification of the content or appearance of this document is unlawful and offenders
may be prosecuted to the fullest extent of the law.
SGS Technical Paper #2005-03