Flotation data for the design of process plants

Part 1 – testing and design procedures

R. C. Dunne*1, G. S. Lane2, G. D. Richmond3 and J. Dioses2
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

This paper discusses the methods used in the design of flotation plants, including benchscale
batch and locked cycle tests and pilot plant trials. The methods used to establish appropriate
flotation circuits as well as the interpretation of the test work data are also discussed in this paper.
Careful and appropriate sample selection must be followed by equally carefully designed and
executed flotation test work if a successful outcome is to be achieved. These steps are detailed
and the correct use of each type of test and the information which can be obtained is described.
Practical design considerations such as flotation time, type, number and size of flotation cells, and
the means of froth transport are all important and appropriate test work can guide decisions on all
these factors.
Keywords: Base metals, Flotation, Test work, Recovery, Mineralogy, Liberation, Flowsheet

This paper is part of a special issue on Metallurgical Plant Design and Operating Strategies

Introduction Sample selection

The first stage of process design is a thorough review of
This paper discusses the steps associated with the
the geology and mineralogy of the deposit. Mining
development of a flowsheet, from mineralogy through
methods and mine plans/schedules need to be assessed
test work, equipment selection to plant design.
when selecting samples for the various phases of the
Irrespective of the project, the development of the
flotation test work programme. Ore selection must take
optimum flotation process is reliant on four key factors:
account of the fact that orebodies are not homogenous
N understanding of the geology and associate miner- and consequently flotation response will vary within the
alogical variations that impact on metallurgical deposit. Changes in head grade, value and gangue
performance mineralogy and redox profile will lead to changes in
N definition of the mining method and the relationship flotation response. The basis of sample selection includes
between the ore reserve and mine production the following:
N evaluation of mineral and water chemistry which can
N rock type and rock alteration
affect the flotation process
N completion of adequate metallurgical and process test
N mineralogical characteristics
work to allow the design of flowsheet, reagent scheme
N ore depth and redox profile (oxide, transitional,
supergene hypergene)
and process control requirements.
N geological model and ore deposition theory
N chemical assay, or grade
Steps in the development and design of N mine schedule.
a flotation circuit Sample types
Background Three major sample types are typically chosen.
There are many papers in the literature that discuss the Samples are selected for initial reagent definition to
development and design of flotation circuits for mineral achieve the desired mineral separation. Separation may
concentrators. These range from texts of ‘biblical status’ be a simple bulk float (i.e. pyritic gold ores, some copper
such as Taggart1 to research oriented texts such as ore with negligible pyrite) or as is the case for
MacDonald and Brison2 to more applied papers such as polymetallic ores, a complex multistage sequential
MacDonald et al.3 and Bulatovic.4 This paper describes separation (i.e. copper, lead, zinc).
the outcomes of the authors’ experiences in the Variability samples are selected to assess orebody
development of base metal projects in Australia. variability on the basis of known physical and chemical
characteristics. As an absolute minimum, samples
should be selected on the basis of spatial distribution
in the orebody if no other characteristics that indicate
1
Newmont Mining Corporation, Greenwood Village, CO, USA variable flotation response are apparent.
2
Ausenco, South Brisbane, Qld, Australia Bulk samples are usually selected for pilot plant trials
3
Minerals and Metals Group, Southbank, Vic., Australia or large scale flotation tests. The extent to which the
*Corresponding author, email [email protected] sample is representative of the ore requires careful

ß 2010 Australasian Institute of Mining and Metallurgy

Published by Maney on behalf of the Institute and The AusIMM
Received 15 April 2002; accepted 1 September 2010 Mineral Processing and Extractive
DOI 10.1179/037195510X12816242170898 Metallurgy (Trans. Inst. Min Metall. C) 2010 VOL 119 NO 4 199
 Dunne et al. Flotation data for the design of process plants: Part 1

consideration.5 The ‘representative sample’ is useful for type of cell, collectors, frothers depressants, pH intensity
development of a general flotation procedure, but of agitation and time. Other variables may include
additional testing may be desirable on individual desliming, in continuous or cycle tests, circulating loads
samples from various areas or depths of the deposit to in grinding and flotation, and recirculation of water.
establish the optimum conditions in each case and to
obtain mill design data valid over the expected range of Laboratory batch flotation test work
ore variation. Obtaining only a single bulk sample from The open circuit batch flotation test work is generally
a large drill hole, winze, trial pit or development face is the first step in an investigation. While the conditions
an option. However, unless the relationship between the employed in batch testing should be readily transferable
metallurgy, mineralogy and geology of the bulk sample to plant-scale operations, this is generally less important
can be related to the variability test work then it is not than the use of a procedure that can be closely
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

recommended.6 This is particularly important with duplicated.8

massive sulphide orebodies where deleterious penalty An important rule in flotation process development is
elements may only be found in particular structures/ to aim for simplicity. A process with unnecessary steps
areas and not observed in bulk samples used to provide or reagents is difficult to investigate and analyse. There
concentrate samples for market assessment. is a tendency to add new reagents or steps without
Furthermore, ore types that perform poorly on their proper evidence that such changes really provide an
own may be ‘masked’ if combined with ores having good improvement.
flotation response. Many different makes and types of laboratory
Fresh diamond core or mined rock (occasionally flotation machines have been used over the years. In
available in ‘mature’ orebodies) is preferred for labora- recent times the preferred makes are those supplied by
tory flotation test work. The use of reverse circulation Agitair, Denver or Wemco. For a test work laboratory
drill chips should be avoided due to the oxidation of treating a wide variety of ores, the most useful
minerals, stratification of heavy particles, loss of softer laboratory flotation machine is probably one with cells
components and fines and overproduction of fines. of 500, 1000 and 2000 g capacities, with a different
Generally, reverse circulation drill samples provide diameter impeller for each cell, with stainless steel or
unrepresentative metallurgy. Ore samples that are prone rubber covered impellers, glass or plastic cell bodies, and
to oxidation (i.e. massive sulphides and pyrrhotite ores), requiring an outside source of low pressure air. A
and require storage for future test work should be stored rotameter installed in the air line allows for control and
at low temperature (,4uC). duplication of the airflow from test to test.
The initial assessment usually involves grinding the
ore sample (drill core samples) to various particle sizes
Flotation test work and flowsheet and applying a reagent regime that is known to give the
development desired mineral(s) separation. Following a review of the
initial flotation test work results alterations are made to
Test work definition
grind size (coarser or finer), reagent system and flotation
The purpose of a flotation test work programme is either times to further enhance mineral recovery. Recovery and
diagnostic (in the initial evaluation and development selectivity problems can be anticipated by conducting a
phase), to develop design parameters (for an engineering detailed mineralogical examination before (or in some
phase), or for modelling purposes (for geometallurgical cases due to pragmatism, in parallel to) the flotation
block modelling or performance forecasting). As such programme. Optimisation of the cleaning stage or
the test work aims to define one or more the following: stages, that may include further reagent addition and
N valuable mineral kinetics and separation response regrinding, follows once acceptable recoveries are
(from gangue minerals) attained in the rougher float.
N valuable mineral recovery and grade of the saleable Several ingredients go into planning a good test
concentrate programme. Characteristics of the ore, empirical knowl-
N type and quantity of deleterious elements (e.g. edge of related flotation separations, and economic
mercury, arsenic) in the concentrate that will incur considerations are all important factors in the selection
financial penalties from a smelter or create down- of the reagents and conditions that should be studied.
stream processing problems An understanding of experimental methods, that is, how
N operating costs (power, reagents, etc.) to conduct the laboratory work efficiently, is a basic
N capital costs (flowsheet, residence time, etc.) requirement. The evaluation of results as the work
N flowsheet flexibility for treating the mine output. proceeds is fed back to the planning phase. Most
Many factors that affect flotation are largely beyond the experimental programmes in flotation necessarily consist
control of the investigator. These include characteristics of a series of planning, experimental and evaluation
of the ore such as fineness of mineral dissemination, the phases.
presence of surface active gangue minerals, degree of The relationship between laboratory and plant metal-
oxidation, and presence of soluble constituents. The lurgy is often complex. The flotation feed preparation
quality and quantity of water are also in this category process (grinding and conditioning) produces different
although these factors sometimes can be controlled by size distributions and chemical environments. These
treatment and by recirculation. relationships must be understood by the metallurgist.7,9
Variables that are subject to control include: grinding
factors such as fineness of grind, pulp density, type of Laboratory locked cycle test work
grinding medium and chemical additives;7 conditioning A locked cycle test is conducted when the open circuit
factors such as collectors, depressants, pH, time, batch test procedure is established.10 Analysis of the locked
intensity and temperature; and flotation factors such as cycle test stability is critical in assessing the metallurgical

200 Mineral Processing and Extractive Metallurgy (Trans. Inst. Min. Metall. C) 2010 VOL 119 NO 4
 Dunne et al. Flotation data for the design of process plants: Part 1

recoveries. It is important to ensure that the product mass between the mini pilot plant and the full scale plant
and metal flow is equivalent to the feed to each cycle thus relate to grind size distribution (often broader in the
ensuring that equilibrium is reached and recoveries and plant and high specific gravity minerals are finer) and
concentrate grades are not overstated. energy input (plant flotation cells have a lower energy
A locked cycle test is a repetitive batch test used to intensity).
simulate a continuous circuit. The basic procedure has a
complete batch test preformed in the first cycle which is Pilot plant operation
then followed by similar batch tests which have Pilot plants should generally only be undertaken to
‘intermediate’ material from the previous cycle added validate the outcomes of the benchscale test work
to the appropriate location in the current cycle. These programme and not to develop or conduct preliminary
batch tests, or cycles, are continued in this manner for evaluations of flowsheets. Some exceptions occur when
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

an arbitrary number of cycles (usually more than six). only a continuous process will allow flowsheet definition
The final products from each cycle, i.e. final concentrate and, even in these cases, parallel benchscale test work
and final tailings, are filtered and removed for further should be undertaken in parallel as a control.
processing. At the end of the test, all the products, final The level of control over the process diminishes from
and intermediate, are dried, weighed and subjected to bench scale to pilot scale (even when on-stream analysis
chemical analysis. The test is balanced and a metallur- is included in the pilot plant) and the amount of effort,
gical projection is made. resources and cost increase dramatically. However, pilot
The main objectives of the locked cycle test are to plants offer the opportunity to produce concentrates for
simulate plant operation with regard to the build up of dewatering, marketing and downstream processing test
fines (typically gangue) or composites; recirculating work and tailings streams for dewatering, rheology, and
loads, water quality, reagents and soluble metal species. downstream test work.
Even where separation by batch test is good, locked
cycle tests may be used to determine the impact of Variability assessment
recycled reagents in process water for multistage The selected flowsheet should be assessed and optimised
flotation. for the variability samples. Failure of the base flowsheet
Locked cycle analysis is time consuming but provides may occur during this evaluation due to a number of
the best simulation of plant conditions at benchscale. reasons, including the oxidised nature of the sample,
However, ores that yield very low middlings deportment changes in mineralogy or head grade, and changes in
of values (very clean separation of value minerals from liberation requirements.
gangue) may not benefit significantly from locked cycle Locked cycle tests may be conducted on variability
test work for determination of the metallurgical balance samples if the outcomes of the batch test work are
or flowsheet design. The number of stages required to significantly different to those observed in the composite
achieve optimum separation of the middlings defines the test work programme.
number of cycles required in the lock cycle test. An outcome of the variability test work programme
Six cycles is generally considered the minimum, with may be the need to campaign or blend particular ore
the tailings of each subsequent stage of processing types to maximise revenue from the plant. This may
recycled to the feed of the next cycle. Problematic ores impact on the mine plan and the optimum project
may not reach equilibrium even after 10 stages, and the concept.
assessment of water quality may take significantly more
effort. Flotation circuit design
The assessment of equilibrium conditions and the
prediction of metallurgy from locked cycle tests is a Parallel or series
matter of much discussion among practitioners. Simple A flotation circuit is a combination of flotation cells and
methods, such as averaging the recovery and grade of auxiliary equipment arranged to deliver the optimum
the last two or three cycles may be adequate for simple results from an ore following grinding and reagent
process routes. For complex separations, a detailed treatment on a continuous basis. The circuit is designed
assessment of the equilibrium conditions for mass, from the results of laboratory results and pilot plant
values and gangue is required by determining a detailed testing of ore samples. The arrangement of flotation cells
balance for each cycle and considering the build-up or is usually for both series and parallel flow. Some
reduction in middlings mass and metal deportment. designers prefer parallel lines for flexibility and operat-
These methods vary from a simple check that concen- ing availability, others maintain that operating avail-
trate and tailings deportment equate to feed inputs, to ability is a function of the design of the processing lines
complex statistical assessments based on in-house and of the ease and type of their maintenance and not of
databases. their number. Single line designs can have significant
advantages in conserving floor space and building
Mini pilot plant volume, in reducing operating labour and in reducing
The availability in recent years of mini flotation pilot the hardware for measurement and control.
plants5 provides an alternative method for predicting the
effects of circulating loads and changes in the solution Flotation circuit complexity
chemistry using drill core samples. The mini pilot plant Flotation circuit details and complexity depend upon a
allows great operational flexibility and therefore permits number of factors of which the numbers of commodities
optimisation of operating conditions during its execu- to be recovered, their values, and the ore texture are the
tion. Recent work has shown the results obtained from most important. The simplest flotation circuits are found
the mini pilot plant can be equivalent to that of a with a single, low value commodity (i.e. coal). Slightly
conventional scale pilot plant. The key differences more complex circuits have a rougher and two stages of

Mineral Processing and Extractive Metallurgy (Trans. Inst. Min. Metall. C) 2010 VOL 119 NO 4 201
 Dunne et al. Flotation data for the design of process plants: Part 1

cleaning. Porphyry copper ores represent a further Regarding the question of the appropriate number of
degree of complexity that involves roughing, with or flotation cells in series, the ultimate performance will be
without scavenging, regrinding of the rougher concen- a function of the flotation kinetics, the degree of ‘short
trate followed by two or more stage cleaning. More circuiting’ due to the number of cells in series, and the
complexity is involved with ores that either require very required total installed volume. A cell is considered as
high grade concentrates because of rigorous specifica- having no or limited backflow to the previous stage. The
tions due to downstream processing requirements, or more cells-in-series the better the ‘plug flow’ character-
with very finely disseminated complex sulphide ores istics of the installation and the less short circuiting
containing copper, lead and zinc or platinum group which will occur; a condition which must be considered
metals. Fineness of dissemination of minerals, and the favourable for most flotation systems. A minimum of six
required concentrate purity all contribute to the degree cells in series is usually recommended, but some circuits
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

of elaboration of the flotation flowsheet. The flotability have been installed with as few as three or four.
(measure of how fast or slow minerals float) of each
mineral component will dictate where these particles
Scale-up criteria
report in the flotation circuit. For example on cleaning, Up until the early nineties, the air flow number was the
the finer, coarser and partially liberated particles are less most commonly used approach for scale-up criterion for
likely to re-float and thus report to the recycle stream. similar metallurgical performance between cells of differ-
Excessive build-up of particles in recycle streams creates ent sizes.11,12 More recent work at the University of
instability in flotation circuits. Queensland (Julius Krutschnitt Mineral Research Centre)
has shown the bubble surface area flux, a parameter
Flotation time and number of flotation cells which is determined by bubble size and superficial gas
The final design of a flotation circuit is only determined velocity, has been identified as an alternative criterion for
from the result of laboratory and pilot plant testing, flotation scale-up from a metallurgical point of view.13
guided by experiences gained from other operations with Investigations have demonstrated a strong correlation
similar types of ores and ore textures. Estimation of between the flotation rate constant and bubble surface
area flux Sb. Scale-up test work has shown that Sb could
flotation cell requirements as to size (volume) and
be used to scale-up from a 250 L cell to a 3 m3 at similar
numbers, once the tonnages, time requirements, and
froth residence times.14 Work conducted at the Pasminco
safety factors have been decided is more routine.
Broken Hill concentrator indicated the Sb was able to
Scale-up rules are driven by consideration of hydro-
scale-up from a 60 L pilot cell to a 100 m3 Outokumpu
dynamics (superficial air flowrate and energy intensity),
tank cell. Their results have further demonstrated the
flotation kinetics, short circuiting (tanks in series model-
importance of froth residence time in determining the
ling), froth recovery characteristics and experience.
overall kinetics of flotation.
An example of the estimation of a rougher circuit’s
flotation cell requirements is given in the Denver Froth transport in a flotation circuit
Bulletin (M75125). This recommends ratios of plant Froth transfer is generally by gravity where possible due
circuit retention times to laboratory batch times in the to the lower flows when compared with tailings streams.
range 1?6 to 2?6 with an average of 2?15.11 While there is In some flow sheets, for example the reverse flotation of
a considerable variation in the magnitude of the safety iron ores, the silicate rich concentrate may be transferred
factors, the following considerations enter into their by gravity in preference to the iron rich flotation tailings.
determination. One option is to provide a greater The transport of froth (synonymous with concentrate
rougher circuit volume than required by the design transport) has its own unique set of problems. Froth is a
criteria to allow for feedrate variances originating in the three-phase system or air, water and solids. The ‘froth
grinding circuit, as well as from spillages and recycle factor’ is a measure of the air contained in the froth and
streams. A 25% increase has been common, but where quantified by filling either a measuring cylinder or bucket,
the variability in grinding circuit product is likely to be of known volume, with froth and measuring the froth
unusually large, this may increased to 50%. column. After air dissipation the remaining water and
Wemco’s scale-up flotation time factors are based on solids volume is measured. The original volume of froth
a scale-down from continuous cells to a 7 L batch cell divided by the combined volume of water and solids is the
and considering all aspects of hydrodynamics, including ‘froth factor’. Measured ‘froth factor’ values are not
mechanical froth removal. Plant installation surveys employed by the flotation cell or pump designers. They
determined that froth removal rate was the key to are modified based on experience and application.
equating batch and continuous test results. In a batch Froth transfer from flotation cell to cell requires either
cell test, pulp level is at a relatively high level for good gravity flow or pumping. The characteristics of the froth
froth removal. In a plant, the pulp level can be varied. depend to a large degree on the type of ore being treated,
For those installations where the pulp level in the plant the fineness of the particles, the solids concentration, the
was relatively high, the scale-up factor was 2?5. For amount of air in the froth slurry and the type of reagents
plants with low pulp level, the scale-up factor was used.
substantially greater. This scale-up factor is related to Froth factors for launders and sumps are dependant
the ‘froth factor,’ measured as the percentage of the on reagent addition rates and reagent type, the ability to
froth over the lip of the weir, and then related the ‘froth add spray water, the nature of the sprays, the design of
factor’ to the ratio of rate constant of laboratory batch the launder or sump and the fineness and nature of the
cell to that of the plant. For cells operating with froth particles. Froth factors for up to 8 to 10 have been
factors of 55–60%, the scale-up factor is y2?5. For measured/reported for ores with a stable froth structure
lower pulp levels in the cells, the difference between the (but should never be used for pump calculations). These
batch and continuous cells is greater. froths require high volume launders, energy to release

202 Mineral Processing and Extractive Metallurgy (Trans. Inst. Min. Metall. C) 2010 VOL 119 NO 4
 Dunne et al. Flotation data for the design of process plants: Part 1

air and effective addition of spray water (to avoid Column cells are no longer cheaper (capex/opex basis)
excessive dilution). Open launders assist in effective on a capacity basis, although they do offer reduced
froth transport and air/slurry separation. A minimum footprint and short circuiting per stage when compared
width of 300 to 400 mm is recommended. with conventional cells.
Froth can vary from brittle (easily broken down and the The selection of columns, including Jameson Cells is
bubbles are generally large) to tenacious (the air is tightly appropriate when the following apply:
bound in the froth and will remain a froth for many hours N replacement of multistage cleaning applications
– bubbles tend to be very fine). Froth characteristics will where froth wash water is advantageous
vary from day to day and even hour to hour depending on N non-sulphide and non-floating gangue is the principal
these many parameters, so it is important to select froth component to be removed
pumps for the worst pumping scenario. The main N the value component is fast floating, and/or
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

requirement is to select the correct size pump and impeller N the flotation kinetics are not significantly impacted by
type for the type of froth and then the correct pump speed power input.
(Warman Bulletin, no. 28). Over-speeding is one of the The acceptance of column cells in Australia has diminished
parameters that will affect pump performance dramati- to such a degree that they are now rarely considered in
cally and deleteriously when handling froths. As a guide, mineral flotation being restricted predominantly to coal,
vertical pumps are good for pumping brittle type froths. final cleaning stages or molybdenum flotation.
For medium type froths to very tenacious froths, The evaluation of column circuits is typically con-
horizontal froth pumps are generally superior. ducted at pilot scale (.5 m height60?1 m diameter)
Recommended froth factors are shown in Table 1. after initial benchscale test work to establish appropriate
kinetics or the need for multistage cleaning.
Flotation equipment
Conventional flotation cells
Conclusions
Convention flotation equipment has changed radically The design and selection of the required flotation
in the last decade and very large cells up to 300 m3 are process for the recovery of valuable minerals must
now available15 and larger cells in design. Cell volumes follow an established and methodical approach. This
range from a couple of square metres up to 300 m3. will include, but is not limited to the following:
Flotation cells come in varies shapes, from square and N thorough review of the deposit’s geology and miner-
rectangular (,30 m3) to U shape and circular (tank alogy
cells) for the very large flotation cells. Smaller cells have N determination of the best method of sample selection,
external or peripheral froth launders whilst the larger taking into account mine life and schedule, chemical
cells have both internal and external launders. Internal assay or grade
launders promote froth removal and thus improve single N a mineralogical investigation in concert with the
cell mineral recovery, promoting coarse particle flotation test work
(.100 mm) removal. Improper design of internal laun- N definition of the grade and recovery of the valuable
ders will lead to bubble coalescence, below the launder, minerals, the extent and effects of deleterious minerals
and hence froth collapse. Furthermore, insufficient N quantifying all the controllable flotation variables
launder width and slope will result in poor froth N sequencing the flotation test work into a defined
transport and lead to decreased cell performance. programme, starting with laboratory batch tests and
Impeller speed and design are important factors in on to locked cycle tests and finally, if required, a pilot
ensuring proper slurry mixing, air dispersion and a plant test work on a representative bulk sample
quiescent pulp/froth interface, to prevent froth collapse. N assessment of the selected flotation process through
test work on variability samples
Column flotation cells N consideration of the impact of the laboratory and
Column flotation underwent a revival in the late 1980s pilot plant grinding circuit product size distribution
and early 1990s due to its capability to replace multi- and the impact that plant cyclone classification may
stage cleaning and low capital cost. In Australia, there have on flotation performance
was a trend to installing columns in rougher flotation N use of established and appropriate scale-up factors
applications as well as cleaner flotation driven in the based on experience and knowledge gained during the
main by the cheaper installation costs of column circuits flotation test work
when compared with conventional cells. N the selection of appropriate flotation cells based on
The general consensus is that these columns worked well the observed flotation behaviour.
where the value minerals are fast floating but failed to yield
optimum metallurgy where the fast float components had Acknowledgements
been removed in preceding flash flotation, or the flotation The authors would like to thank the following people
kinetics were affected significantly by power input and pre- who provided data for the original paper published by
flotation conditioning was inadequate. the AusIMM: Peter Wassel, Peter Bourke, Ian
Arbuthnot, Steve Hughes and Asa Weber. Thanks are
Table 1 Froth factors for pumps
also extended to Lawrence Ruwoko of Ausenco for
Type Froth factor assistance in editing the original paper. This paper has
been reproduced with the kind permission of the
Brittle froth 1.1–1.25 Australasian Institute for Mining and Metallurgy from
Medium froth 1.25–1.5 Metallurgical Plant Design and Operations Strategies
Tenacious froth 1.5–1.6
(MetPlant) 2002, 15–16 April 2002, Sydney, Australia.

Mineral Processing and Extractive Metallurgy (Trans. Inst. Min. Metall. C) 2010 VOL 119 NO 4 203
 Dunne et al. Flotation data for the design of process plants: Part 1

8. R. L. Coleman: ‘Metallurgical testing procedures’, in ‘Mineral

References processing plant design’, (ed. A. L. Mular and R. B. Bhappu), 144–
1. A. F. Taggart: in ‘Handbook of mineral dressing’, 19–181; 1945, 182; 1978, New York, SME.
New York, John Wiley and Sons. 9. G. Lane, J. Fleay and A. Marin: ‘Examining the effects of
2. R. D. MacDonald and R. J. Brison: ‘Applied research in flotation’, comminution on downstream flotation’, Proc. Conf. Crushing
in ‘Froth flotation’, (ed. D. W. Fuerstenau), Chap. 12; 1962, New and Grinding ’99, Perth, WA, Australia, March 1999,
York, SME. IRR.
3. R. D. MacDonald, W. C. Hellyer and R. W. Harper: ‘Process 10. G. E. Agar and W. B. Kipkie: ‘Predicting locked cycle flotation test
development testing’, in ‘SME mineral processing handbook’, (ed. results from batch data’, CIM Bull., 1978, 71, 140–147.
N. L. Weiss), 30-116–30-120; 1985, New York, AIME. 11. N. Arbiter: ‘Development and scale-up of large flotation cells’,
4. S. Bulatovic: ‘Handbook of flotation reagents: chemistry, theory Min. Eng., 2000, 52, 28–33.
and practise’; 2007, Elsevier. 12. N. Arbiter, H. Cooper, M. C. Fuerstenau, C. C. Harris, M. C.
5. V. L. Andrade, N. A. Santos, K. L. C. Goncalves and Kuhn, J. D. Miller and R. F. Yap: ‘Flotation’, in ‘SME mineral
Published by Maney Publishing (c) IOM Communications Ltd and the Australasian Institute of Mining and Metallurgy

H. Wyslousil: ‘Obtaining metallurgical data from drill core samples processing handbook’, (ed. N. L. Weiss), 5–93; 1985, New York,
using a mini pilot plant’, Proc. 34th Ann. Meet. of Canadian AIME.
Mineral Processors, Ottawa, Ont., Australia, January 2002, 13. B. K. Gorain, F. Burgess, F. P. Franzides and E. V. Manalpig:
Canadian Mineral Processors, Paper 16. ‘Bubble surface area flux, a new criterion for flotation scale-up’,
6. D. Lunt, I. Ritchie and J. Fleay: ‘Metallurgical process develop- Proc. AusIMM 6th Mill Operators’ Conf., Madang, Papua New
ment and plant design’, Proc. Conf. Mindev ’97, 111; 1997, Sydney, Guinea, October 1997, AusIMM, 141–148.
AusIMM. 14. J. E. Nesset: ‘The application of residence time distributions for
7. J. Fleay and G. Lane: ‘Grinding environment effects in the flotation and mixing circuits’, CIM Bull., 1988, 81, 75–83.
flotation of Kambalda nickel ores’, Proc. 5th Mill Operators Conf., 15. K. R. Suttell: ‘Large sizes still dominate flotation cell applications’,
43–50; 1994, Roxby Downs, SA, AusIMM. E&MJ, Jan. 1989.

204 Mineral Processing and Extractive Metallurgy (Trans. Inst. Min. Metall. C) 2010 VOL 119 NO 4