More than a century of progress - min-eng.com machine (formally Bateman BQR) (inset) here at Messina...
Transcript of More than a century of progress - min-eng.com machine (formally Bateman BQR) (inset) here at Messina...
Flotation has been at the heart of the
mineral processing industry for over 100
years, addressing the ‘sulphide’ problem of
the early 1900s, and continues to provide one of
the most important tools in mineral separation
today. The realisation of the effect of a minerals
hydrophobicity on flotation all those years ago
has allowed us to treat oxides, sulphides and
carbonates, coals and industrial minerals
economically, and will continue to do so in the
future.
There have been a number of important
changes in the industry over the years as flotation
technology and equipment have advanced.
Xstrata Technology considers “the most noticeable
has been the increase in sizes of the flotation
machines, from the multiple small square cells
that were initially used, to the 300 m³ round cells
used today that are the norm in large scale plants.
“Other changes have been more subtle, but
equally as important. One of these has been the
design of the flotation circuit to make the most
of the liberation and surface chemistry effects of
the minerals. In a lot of these situations it is not
a matter of ‘bigger is better’, that will make the
process work, but being smarter in the
application of flotation technology.”
Xstrata Technology is one company that
believes the smarter use of flotation machines
can deliver big improvements in plant
performance. Through its use of the naturally
aspirated Jameson Cell, Xstrata Technology has
been making inroads into the processing of
more complex ores. Having a small footprint,
and using the high intensity mixing environment
of slurry and naturally induced air in a simple
downcomer, the Jameson Cell provides an ideal
environment for the separation of hydrophobic
particles and gangue, it says. The small
footprint of the cells also makes them ideal to
retrofit into a circuit especially where space is
tight.
While the cell has been included in some
flotation applications as the only flotation
technology such as coal and SX-EW, the main
applications in base metals have seen the cell
operating in conjunction with conventional cells.
The combination of the two technologies
enables the Jameson Cell to target the quicker
floating material, while the conventional cells
target the slower floating material. “Such a
combination provides a superior overall grade
recovery response for the whole circuit, than
just one technology type on its own,” Xstrata
Technology says. Below are some of the duties
for which the Jameson Cell can be used.
Jameson Cells in a scalping operation target
fast floating liberated minerals, and produce a
final grade concentrate from them. The wash
water added to the Jameson Cell assists in
obtaining the required concentrate grade due to
washing out the entrained gangue. Scalping can
be done at the head of the cleaner (also known
as pre cleaning), or at the head of the rougher
(also known as pre roughing), and minimises
the downstream flotation capacity using
conventional cells needed to recover the slower
floating minerals.
Sometimes deleterious elements found in the
orebody are naturally highly hydrophobic, and
need to be removed at the start of flotation,
otherwise they will report with the valuable
minerals to the concentrate and effect
concentrate grade. Mineral species such as talc,
carbon and carbon associated minerals, such as
carbonaceous pyrite, can all be difficult to
depress in a flotation circuit. On the other hand,
floating them off in a prefloat circuit before the
rougher is an easier way to handle them.
Jameson Cells acting as a prefloat cell at the
head of a rougher circuit, or treating the
hydrophobic gangue as a prefloat rougher
cleaner, is an ideal way to produce a ‘throw
away’ product before flotation of the valuable
minerals, minimising reagent use and circulating
loads.
Jameson Cells can be used in cleaning circuits
to produce consistent final grade concentrates.
The ability of the cell to keep a constant pulp
level, even with up stream disturbances or loss
of feed, enables a constant grade to be
obtained.
Xstrata Technology concludes: “Importantly in
a lot of these circuits, it is not the selection of
one type of technology that produces the
FLOTATION
24 International Mining | NOVEMBER 2011
More than a century of progressJohn Chadwick examines new technologies and applications from some of the keyplayers in mineral flotation, a technique that is so important to the global industry
Stawell gold mine in co-operation with OutotecServices completed a flotation circuit upgradeon time and on budget last year that, instead ofthe projected 3.5% improvement, resulted in anincrease of 4.5% since the project wascompleted. Payback was also impressive,occurring within less than four months.
FLOTATION
required grade and recovery, but the selection of
several technologies to get the best results. The
interaction of slow floating and fast floating
minerals, entrainment, hydrophobic gangue and
a myriad of other variables make it rare for just
one type of technology to prevail, but the
combination of different flotation machines can
achieve the required outcome more efficiently,
as well as make the circuit robust enough to
handle variations in feed quality.”
The Jameson Cell has benefitted from over 20
years of continuous development. Early this
year, the 300th cell was sold into Capcoal’s Lake
Lindsay coal operation in the Bowen Basin of
Australia. Around this time there were a number
of coal projects taking in new Jameson Cells,
including expansion projects for Wesfarmers’
Curragh and Gloucester Coal’s Stratford
operations (both in Australia), Riversdales’
Benga project in Mozambique and Energy
Resources’ Ukhaa Khudag coking coal project in
Mongolia.
Le Huynh, Jameson Cell Manager, said the
interest for coal preparation plants has
remained strong, where operators needed
dependable and reliable technology to treat fine
coal, an important source of revenue. During
2010, the Jameson Cell business also found
success in other applications, including
recovering organic from a copper raffinate
stream at Xstrata-Anglo American’s Collahuasi
copper SX-EW plant in Chile.
Le said the consistent generation of very fine
bubbles and the high intensity mixing in the
Jameson Cell, was ideal for recovering very low
concentrations of organic from raffinate
streams, typically less than several hundred
ppm. High throughput in a small footprint,
simple operation and extremely low
maintenance due to no moving parts in the cell
are distinct advantages in this application.
The cell is designed with features specific to
suit such hydrometallurgy applications including
specialist materials, a flat-bottomed flotation
tank with integrated pump box and tailings
recycle system, and large downcomers. The
Collahuasi cell was the first of its type in Chile,
though there are many other large cells installed
in SX-EW plants in Mexico, USA and Australia to
treat both raffinate and electrolyte streams.
Dominic Fragomeni, Manager Process
Mineralogy, Xstrata Process Support (XPS),
notes that accurate, rapid development of a
milling and flotation flowsheet for a new
orebody is key to successful mine development.
Time-honoured conventional practice has
typically favoured the extraction of a bulk
sample of up to several hundred tonnes for
conventional pilot plant campaigns which could
operate at several hundred kilograms per hour.
Where sample extraction is limited, much
reliance has been placed on locked cycle tests
alone to produce design basis criteria. These
approaches can be lengthy, expensive, carry
scale up risk, and have seen a wide range of
successes and failures at commissioning and
during life of a mine.
XPS has miniaturised the pilot plant process.
At the same time, it has improved the
representativeness of results from the pilot
plant campaign by using exploration drill core to
formulate the pilot plant sample. This Flotation
Mini Pilot Plant (MPP) was developed in
collaboration with Eriez subsidiary Canadian
Processing Technologies (CPT) and operates in
fully continuous mode either around the clock or
can be made to demonstrate unit operations on
a shift basis. The feed samples are in the range
of 0.5-5 t and can consist of exploration ½ NQ
drill core which improves the sample
representativeness. The MPP operates in the
range of 7-20 kg/h, an order of magnitude lower
in sample mass and typically at a lower cost
when compared to conventional pilot plants.
XPS has developed and validated a
representative sampling strategy, an appropriate
quality control model for metallurgical results
and has accurately demonstrated operations
results using Raglan and Strathcona ores and
flowsheets. These validation campaigns, in
‘scale down’ mode from the full scale plants,
have produced actual mill recoveries to within
0.5% at the same concentrate grade with
internal material balance consistent with the
plant.
When designing a plant to recover copper,
Scott Kay, Process Engineer with METS suggests
(in METS Gazette, issue 32, October 2011) it
would be prudent “to perform some
mineralogical analysis test work such as
QEMSCAN (Quantitative Evaluation of Mineral
by Scanning electron microscopy) to provide
some knowledge on the proportion of sulphide
and oxide minerals present, the grain sizes of
each mineral and a suggested grind size before
jumping into the bulk of the beneficiation test
work.
“Ideally, the characteristics of the copper
bearing minerals should suggest an appropriate
grinding circuit P80 of between 100 and 200 μm
(0.1 and 0.2 mm), which can be controlled by
cyclones, or in some cases fi ne screens.
“Flotation reagent selection is paramount and
test work is necessary to ensure the optimum
reagent suite is utilised. If the ore contains a low
amount of iron sulphides, xanthate collectors
are often suitable to float copper sulphide
26 International Mining | NOVEMBER 2011
Jameson Cell in a cleanerscalping duty at Phu Kham,Laos, producing final gradecopper concentrate prior toconventional cleaning circuit(flowsheet presented in Mayat Austmine 2011, Brisbane)
minerals. If native gold is present,
dithiophosphates can be used which are less
selective to iron sulphides. Increasing and
controlling the pH within the flotation vessel
to between 10 and 12 causes the process to
become more selective, away from iron sulphide
gangue minerals such as pyrite to produce a
cleaner copper mineral concentrate. Depending
on the ore mineralogy, activators and
depressants may be required to achieve the
optimum reagent suite.
“Recovery of copper oxide minerals can be
achieved with flotation by sulphidising the ore.
In essence, this creates a thin layer of copper
sulphide (chalcocite) on the oxide grains which
can then be activated and collected in the froth.
When employed, this occurs after the sulphide
flotation stage, however, this is not commonly
used as other beneficiation processes, such as
leaching and SX-EW are often more cost
effective for copper oxide minerals.
“A common flotation circuit usually includes a
rougher/scavenger and a cleaner stage. As most
copper orebodies exhibit an in-situ grade of less
than 1% Cu, the mass pull to the rougher froth is
often low. This means that the throughput of the
cleaner stage is significantly less than the
throughput of the rougher stage which imparts a
relatively low capital and operating cost to the
flotation circuit.
“To counteract the possible absence of a
scavenger stage, a slightly higher mass pull to
the rougher froth is targeted (although still low
overall) to increase overall copper recovery. The
rougher froth can then be reground to increase
the liberation of the copper sulphides from the
iron sulphides before being fed to the cleaner
flotation vessels. This results in a significant
upgrade in copper in the cleaner froth whilst still
achieving a high copper recovery. The final
flotation concentrate usually contains between
25 and 40% Cu.”
28 International Mining | NOVEMBER 2011
The Delkor BQRflotation machine(formally Bateman BQR)(inset) here at Messina MowanaCopper mine in Botswana. 15 x 50BQR and 16 x 200BQR flotation cells for Copper roughing,cleaning & re-cleaning. Oxide / Sulphide
FLOTATION
Alain Kabemba, Flotation Process Specialist
at Delkor notes the major trend to treating
lower-grade and more finely disseminated ores
and lately the re-treatment of tailings. He also
points to the broad applicability of size to below
10 μm.
Real systems do not fulfil ideal conditions,
mainly because of feed variation or
disturbances. “Before considering disturbances
to flotation specifically,” Kabemba says “it is
important to emphasise the interlock between
grinding and flotation, not only with respect to
particle size effects, but equally to flotation feed
rate variations. The grinding circuit is usually
designed to produce the optimum size
distribution established in testing and given in
the design criteria. When the product size alters
from this optimum, control requires either
changing feed tonnage to the circuit or changing
product volume, with either causing changes in
flotation feed rates.
“While grindability changes due to the
variation in ore properties are disturbances to
the grinding circuit, they generate feed rate
changes as disturbances to the flotation circuit.
The variations in ore properties which affect
flotation from those assumed in the design
criteria must therefore necessarily include
grindability changes.
“This reflects important differences in
flotation machine characteristics between the
two processes. Grinding circuits are built and
designed with fixed total mill volumes and
energy input, so the grinding intensity is not a
controllable variable, instead grinding retention
time is changed by variation of feed rates. In
contrast, the flotation circuit is provided both
with adjustable froth and pulp volume for
variation of flotation intensity by aeration rate or
hydrodynamic adjustment. Reagent levels and
dosages provide a further means for intensity
control.”
One recent trend has been towards larger,
metallurgical efficient and more cost effective
machines. These depart from the simpler
tank/mechanism combination towards design
which segregates and directs flow and towards
providing an external air supply for types which
had been self aerating and towards the
application of hydrodynamic principles to cell
design, like the Delkor BQR range of flotation
machines, initially the Bateman BQR Float Cells.
Bateman has steadily developed the BQR
flotation cells which have been in application for
the past 30 years, and with its acquisition of
Delkor in 2008, decided to rebrand the
equipment into the Delkor equipment range.
Kabemba explains that BQR cell capacities
range from 0.5 to 150 m3 currently installed, and
can be used in any application as roughers,
scavengers and in cleaning and re-cleaning
circuits.
“The main objectives of the BQR design were
to achieve the following core hydrodynamic
functions:
■ Provide good contact between solid particles
and air bubbles
■ Maintain a stable froth/pulp interface
■ Adequately suspend the solid particles in the
slurry
■ Provide sufficient froth removal capacity
■ Provide adequate retention time to allow the
desired recovery of valuable constituent.
“This led to the following benefits
■ Highest possible effective volume and
reduced the froth travel distance
■ Improved metallurgical performances in
terms of grade recovery and reduced capital
and operating costs based on reduced
fabrication material and ease of maintenance
Kabemba says “there are not many
differences in terms of design between BQR
Flotation cells; however, from the BQR1000
upwards, the flotation cells have internal
launders to maintain the design objectives and
benefits highlighted.
“Operating variables, such as impeller speed,
air rate, pulp and froth depths have to be
adjustable over a sufficient range to provide
optimum results with a given ore, grind and
chemical treatment, but adjustment should not
extend beyond the hydrodynamic regime in
which good flotation is possible.”
FLOTATION
NOVEMBER 2011 | International Mining 29
Proven PerformanceLarox Flowsys is now Flowrox
FLOWROX VALVES AND PUMPS for abrasive, corrosive and other demanding shut-off, control, pumping and dosing applications.
EXPERTISE WITH INSIGHT• Over 30 years of experience• More than 100 000 installed products worldwide
PARTICIPATING PARTNER• Long-term partnerships with the customers• Close cooperation between experts in all disciplines
FACILITATING PERFORMANCE• The lowest total costs of ownership • Improved efficiency and extended service intervals
Flowrox OyMarssitie 1P.O. Box 338FI-53101 LappeenrantaFinland
Tel. +358 201 113 311Fax +358 201 113 [email protected]
Meet us at
Flotation ’11
Conference
The largest current BQR flotation machine is
shown in the table below. In the near future the
BQR2000 (200 m3) and BQR3000 (300 m3) will
be available to the market.
Kabemba also explained that “circular cells
reduce the amount of dead volume when
compared to square cells. This enables a much
higher effective pulp volume, hence increasing
the effective energy input into the flotation cell.
In addition ‘tank type’ cells offer enhanced froth
removal due to the uniform shape of the circular
launders.”
He concluded that “fully automated flotation
cells are becoming more and more common with
the aid of smart control and advances in
software in the marketplace.”
Better fine particlerecoveryFLSmidth’s flotation team
notes that fundamental
flotation models suggest
that a relationship exists
between fine particle
recovery and turbulent
dissipation energy1.
Conversely, increased
turbulence in the rotor-
stator region is theoretically related to higher
detachment rates of the coarser size range2.
Conceptually, the suggested modes of recovery
for the extreme size distribution regions appear
to be diametrically opposed.
Industrial applications have previously
confirmed that imparting greater power to
flotation slurries yields significant
improvements in fine particle recovery.
However, recovery of the coarser size class
favours an opposing approach, The FLSmidth
team believe. An improvement in the kinetics of
the fine and coarse size classes, provided there
is no adverse metallurgical influence on the
intermediate size ranges, is obviously beneficial
to the overall recovery response. Managing the
local energy dissipation, and hence the power
imparted to the slurry, offers the benefit of
targeting the particle size ranges exhibiting
slower kinetics.
New concept, Hybrid Energy Flotation™
(HEF™), using phenomena described above was
recently introduced by FLSmidth. In principle it
decouples regimes where fine and coarse
particles are preferentially floated. HEF includes
three sections:
1. Standard flotation machines (energy, rpm,
rotor size) at the beginning of the row, where
Designations BQR 1500
Nameplate Volume (m^3) 150
Diameter (m) 6.1
Height (m) 5.98
Total surface Area (m^2) 29.22
Effective surface Area (m^2) 15.28
Total volume (m^3) 174.7
Volume in internal launders 4.08
Volume lost in stator tube 1.06
Crowder diameter at surface 2.5
Volume lost in Crowder 2.05
Effective volume - no air 167.5
Effective Volume – aerated 150.8
Copper andmolybdenite recovery of-20 μm fraction in HEFcleaner duty
FLOTATION
30 International Mining | NOVEMBER 2011
If you need MORE growth or want to generate LEAN efficiencies, contact SGS. Daily, the SGS global team delivers technical expertise, effective services and sustainable solutions that can enhance your project.
SGS IS THE WORLD’S LEADING INSPECTION, VERIFICATION, TESTING AND CERTIFICATION COMPANY
www.sgs.com/[email protected]
FOCUSED + BEST PRACTICES
TESTING + EXPERIENCE
TROUBLESHOOTING + EFFICIENCY
PRODUCTIVITY + SUSTAINABILITY
TECHNOLOGIES + KNOW HOW
EXPERTISE + TECHNICAL DEPTHTRANSPARENCY + RISK MITIGATION
MARKET INSIGHT + COST REDUCTIONSDEVELOPMENT + STREAMLINED PROCESSES
CONFIDENCE + EFFICIENT LOGISTICSSPEED-TO-MARKET + GLOBAL NETWORK
COMPETITIVE ADVANTAGE + PRODUCTIVITY
MORE + LEANGROWTH EFFICIENCIES
World Class Solutions for the Global Minerals Industry
JKTech is the world leader in resource characterisation and mining process improvement
JKTech’s innovative expertise offers tailored solutions across the entire life of mine:
Maximise Profitability | Increase Throughput | Increase Recovery
Reduce Costs | Improve Sustainability | Manage Risk
Comminution, Flotation, Mining, Geometallurgy, Sustainability, Risk Management and Process Statistics
flotation is froth phase limited and
operational and set-up parameters have small
influence on the recovery
2. Higher power flotation machines (high rpm,
standard rotor size) at the end of the row to
increase recovery of fine particles
3. Lower power flotation machines (low rpm,
larger rotor) at the end of the row (mixed with
higher energy cells) to increase recovery of
coarse particles.
This concept was successfully implemented
at the Mineral Park concentrator in Arizona and
will be expanded at various mines in the nearest
future.
This subject will be expanded upon at the 5th
International Flotation Conference (Flotation '11)
in Cape Town, South Africa. The fundamental
parameters that influence fine and coarse
particle recovery will be reviewed. The potential
dual recovery benefit is then presented in terms
of its practical implementation in a scavenging
application. HEF is proposed as the preferred
methodology of recovering these ‘slow-floating’
size ranges; a method that opposes the
traditional approach of residence time
compensation.
Eriez® Flotation Group introduced the
StackCell® flotation concept in 2009. This
innovative technology recovers fine particles
more efficiently than mechanical flotation cells.
“We’ve taken the inherent
advantages of mechanical flotation
and adapted them to a new
design that is significantly
smaller and requires less
energy,” explained Eriez Vice
President Mike Mankosa. “We
focused on reducing the
retention time and energy
consumption by implementing a
completely different approach
to the flotation process. This new
approach provides all the
performance advantages of column
flotation while greatly reducing capital,
installation and operation costs.”
At the core of the StackCell technology is a
proprietary feed aeration system that
concentrates the energy used to generate
bubbles and provides bubble/particle
contacting in a relatively small volume. An
impeller in the aeration chamber located in the
centre of the cell shears the air into extremely
fine bubbles in the presence of feed slurry,
thereby promoting bubble/particle contact.
Unlike conventional, mechanically agitated
flotation cells, the energy imparted to the slurry
is used solely to generate bubbles rather than to
maintain particles in suspension. This leads to
reduced mixing in the cell and shorter residence
time requirements.
The StackCell sparging system operates with
low pressure, energy efficient blowers that
decrease power consumption by 50% compared
to air compressors or multi-stage blowers used
in other flotation devices.
The low-profile StackCell design features an
adjustable water system for froth washing and
also takes advantage of a cell-to-cell
configuration to minimise short-circuiting and
improve recovery rates. Space requirements for
StackCell offers column-likeperformance in a substantially
smaller footprint thanconventional cells.
These compact,stackable units
provideconsiderable
savings for newinstallations and
are ideal forexpanding
capacity in anexisting plant
NOVEMBER 2011 | International Mining 31
FLOTATION
the StackCell design are approximately half of
equivalent column circuits, with corresponding
reductions in weight leading to reductions in
installation costs. Units can be shipped fully
assembled and lifted into place without the
need for field fabrication.
This technology can provide recoveries and
product qualities comparable to column
flotation systems while using a low profile
design. Not intended to replace the need for
column flotation, it does provide an alternative
method to column-like performance where
space and/or capital is limited. The small size
and low weight of the new StackCell makes
possible lower cost upgrades where a single cell
or series of cells may be placed into a currently
overloaded flotation circuit with minimal retrofit
costs.
Steve Flatman, General Manager of Maelgwyn
Mineral Services (MMS) also comments on the
“trend of moving towards a finer grind to
improve mineral liberation. Unfortunately
conventional tank flotation cells are relatively
inefficient in recovering these metal fines below
30 μm and very inefficient at the ultra fine grind
sizes below 15 μm. The incorporation of regrind
mills on rougher concentrates has further
exacerbated this problem. To date the
conventional flotation tank cell manufacturers
have attempted to counter this fall off in
recovery of fine particles by inputting increasing
amounts of energy (bigger agitation motors)
into the system to improve bubble particle
contact. Unfortunately this tends to compromise
coarse particle recovery.”
He says the solution is MMS’s “Imhoflot
pneumatic flotation technology and specifically
the Imhoflot G-Cell. Recent pilot plant test work
at a nickel operation with a three stage Imhoflot
G-Cell pilot plant enabled an additional 30%
nickel to be recovered from the conventional
flotation tank cell final plant tails. The recovery
was predominantly associated with the minus-11
μm fraction indicating that this improved
recovery was not just related to additional
residence time. The above results are in line
with an earlier pilot plant trial using G-Cells on a
zinc operation where an additional 10-20% zinc
was recovered from cleaner tailings this time
being associated with minus 7 μm material.
“It is postulated that the above
improvements are related to the order of
magnitude increase in terms of air rate
(m³/min/m³ pulp)for the G-Cells due to their
principle of operation where forced bubble
particle contact takes place in the aeration
chamber rather than the cell itself with the cell
merely acting as a froth separation chamber.
Typically in percentage terms the G-Cell air rates
are five to ten times that of conventional
flotation although the overall or total air usage
is approximately half.
“When this additional targeted energy input
is combined with the centrifugal action of the G-
Cell and small bubbles benefits are obtained in
both the flotation rate (kinetics) and overall
recovery. The improved kinetics results in a
much lower residence time than conventional
flotation facilitating a double benefit of both
reduced footprint and improved recovery.”
A technical paper will be presented at the
MEI Flotation 11 Conference in South Africa
providing more detail on the specific case
studies.
Metso notes a main drawback of column
cells being low recovery performance, typically
resulting in bigger circulating loads. Its CISA
sparger is derived from the patented Microcel™
technology and enhances metallurgical
performance by allowing flexibility on the grade-
recovery curve. Metso Cisa says the main
advantages of its column technology include:
■ Improved recovery and optimised grade
■ Increased throughput
■ Enhanced bubble particle contact
■ No plugging
■ On-line replacement and reduced wear and
maintenance
■ Unique sparger Technology.
At the bottom of the column, the sparger
system raises mineral recovery by increased
carrying capacity due to finer bubble sizes. This
maximises the bubble surface area flux which is
a standard parameter in evaluating flotation
device performance. It also provides maximum
particle-bubble contacts within the static mixers
and effective reagent activation from the
mechanical operation of the pump.
It is well known that coarse particles behave
poorly in a conventional flotation cell and were
previously regarded as ‘non-floatable’. However,
recent laboratory work demonstrates that
Fluidised Bed Froth (FBF) flotation extends the
upper size limit of flotation recovery by a factor
of 2-3 resulting in significant concentrator
performance benefits. AMIRA’s P1047 project,
Improved Coarse Particle Recovery by FBF
Flotation, is expected to commence in 2012, and
will be structured in two phases.
Some of the benefits for FBF technology are:
■ Early rejection of gangue with minimum
mineral loss
Potential for significant increase in
concentrator throughput or significant
improvement in capital efficiency
■ Reduced energy consumption
Independent modelling predicts that if particles
of 1 mm can be floated, comminution energy
consumption will be lowered by at least 20%.
■ Better management of water requirement
FBF cells can take product straight from the
milling circuit without dilution, and the feed to
the FBF cell could be up to 80% w/w solids,
which could lead to significant savings in
process water demand.
■ Improve recovery of metallic and other dense
minerals.
In a continuous FBF Cell, dense mineral
particles will tend to sink to the bottom and
accumulate in the cell, thus they can be
recovered in a concentrated form by emptying
the cell periodically. This could be a significant
benefit where the concentration of the heavy
metallic material is too low to warrant a
separate treatment plant to recover them.
Upgraded servicesIn Australia, Northgate Minerals’ Stawell gold
mine recently completed a project through
which it aimed to increase recoveries by 3.5% by
upgrading the flotation plant. This upgrade was
implemented after Stawell changed its
production profile to process lower grade ore at
higher throughput rates.
Instead of the projected 3.5% improvement,
Imhoflot stage G1.2 pilot plant
FLOTATION
32 International Mining | NOVEMBER 2011
the upgrade from Outotec Services has resulted
in an increase of 4.5% since the project was
completed on time and on budget last year,
despite the wettest seasonal weather in recorded
history. Payback was also impressive, occurring
within less than four months. “The projected
payback was 5.5 months, so it was a pleasant
surprise when it happened so soon” explains
Jodie Hendy, senior metallurgist at Stawell.
The project has also achieved payback in less
than four months and has delivered further
ongoing benefits, including easier operation and
reduced maintenance costs, says Outotec
Services, which worked in close partnership
with Stawell Gold to ensure the site remained
fully operational during the upgrade.
The mine, which has produced more than 2
Moz in its 26-year history, previously employed
a flotation circuit consisting of a bank of eight
mechanical trough cells in the rougher circuit,
followed by two banks of 2 x OK3 Outotec cells
as cleaners. The feed rate to the cells was
between 90-105 t/h, at 50-55% solids. The
overall flotation circuit was not performing at
optimal rate due to entrainment problems in the
rougher cells when feed density increased from
45% to 55% solids, typically at 105 t/h.
In anticipation of future production levels and
as part of Stawell’s focus on operational
excellence, it was decided to upgrade the
flotation circuit. Following a site audit from
Outotec Services, a 2 x TankCell® -20 configuration
equipped with larger TankCell -30 mechanisms
was proposed to help optimise flotation. The
larger mechanisms would allow operation at
very high percent solids (50% and over).
The TankCell design also allows a much
deeper froth depth and better concentrate grade
through optimised launder lip length and
surface area. These cells known for great
performance, ease of operation and reduced
power and air consumption. Outotec Services
was commissioned to handle the complete
turnkey solution of the new rougher circuit,
including design, supply, installation and
commissioning.
The schedule was demanding but achievable,
in just 30 weeks. It was decided to adopt the
partnering approach between Stawell and
Outotec Services, because this collaborative
method ensured open communication, with all
parties having greater ownership of the project
and its aims. This close teamwork resulted in
meticulous planning and site remaining fully
operational at all times. Pipework and electrical
easement ducts, for example, were rerouted
early in the project. Tie-in points for new cells
and rerouting of pipework were also planned
upfront in the project and all disruptive work
was completed during shutdowns.
The project overcame a number of challenges,
including an extremely limited footprint, which
was adjacent to a gabion wall, close to the run-
of-mine pad and also close to a reagents shed,
which could not be moved. Additionally, existing
process requirements at Stawell required specific
elevations for the new TankCells. Structural
stability was the main issue when designing the
tank support structure due to the height of the
tanks and the limited footprint. Sufficient
stiffness was required such that the operation
frequencies of the TankCells would not interfere
with the natural frequency of the tank support
structure. Through FE modelling of the structure,
section sizes and bracing orientations were
optimised to produce the required stiffness.
Despite the challenges, the turnkey
installation of the new rougher circuit, along
with blowers for the complete flotation circuit,
was completed within deadlines. Because all
tie-in points had been already carefully planned
upfront, commissioning was a seamless exercise.
Designed to cope with projected increases in
production and considerably more operator
friendly than its predecessor, the new TankCell -
20 cells have quickly proved their worth at site.
The air demand for the old rougher cells, for
example, was estimated at over 3,000 Am3/h,
whereas the estimated air demand on the
Outotec TankCells is a maximum of 992 Am3/h.
The Outotec FloatForce® rotor-stator
mechanism, with its unique design, delivers
enhanced flotation cell hydrodynamics and
improved wear life and maintenance.
“Maintenance on the Outotec TankCells has
also been minimal since the upgrade, Hendy
commented. “Basically we check the cells
during shutdowns but there has been no
maintenance required in the nine months since
commissioning. “The TankCells have really
delivered on their reputation. Basically, they do
exactly what they are supposed to do.”
Smarter reagent useTurning to flotation reagents, Frank Cappuccitti,
President of Flottec explains that Flottec and
Cidra are “working very hard jointly on
developing instruments that will measure
hydrodynamics in the flotation cell and circuit in
a bid for better flotation control. This would be a
great step forward in using a combination of
reagents and sensors to optimise flotation
systems. It brings together the knowledge we
have developed in both how reagents effect
hydrodynamics and measuring the
hydrodynamics to maintain optimum conditions.
He explains that back in the 1990s, when he
worked at a well-known mining chemicals
supplier, “we spent most of our research on
trying to find the best collectors. The thinking
was that we could try to develop collectors with
absolute specificity. In other words, we could
develop a collector that would float only specific
minerals and provide clients with an almost
perfect flotation separation. This was our
approach to flotation optimisation.
Unfortunately, we discovered that there was no
such thing as absolute specificity. In fact, we
had trouble measuring any improvements in the
circuits because they were multi-variant and
highly complex. Every change made was always
a trade off between grade, recovery and cost.
Changing one thing in the circuit seemed to
improve something but always got a negative
response in some other variable. It was also
very hard to measure the performance of the
flotation circuit because the only real
parameters you could measure on line were
concentrate grades and tails of the circuits,
which were always after the fact. There was little
ability and understanding about what real time
FLOTATION
34 International Mining | NOVEMBER 2011
Fluidised Bed Flotation concept cell
measurements we could take other than air
rates, cell levels and flow rates. So even if we
got an improvement or a response to a change,
we never knew if that was a response to a
change or a natural variation in the system.
Every test needed long term statistical trials to
get some confidence in any real change.
“So, I wrote a paper in the 1990s that
basically said that until we could measure the
real time variables in a flotation system and
learned to really understand and control the
system, we were limited in our ability to work on
continuous improvement in reagent
optimisation. We needed new sensors that could
measure the performance of the flotation circuit
so we could learn to control it. Once we got
this, then we could actually measure
improvements and use this to develop reagents.
“Fortunately, with the advent of strong
computing power and software, we have moved
forward tremendously in the last decade in
understanding the flotation circuit. Froth
cameras that tried to measure froth grade and
velocity were one of the first new sensors
developed to assist in optimising circuits.
Through the work of universities such as McGill
and organisations like JKtech, new sensors have
been developed that could actually measure
reliably and in real time the hydrodynamic
parameters in the flotation cell. Flotation cell
hydrodynamics (gas dispersion parameters) is
critical to the performance of the cell. When we
talk about these parameters, we are talking
about measuring what is happening in a
flotation cell. Flotation is really about making
bubbles and using the surface area of the
bubble to do the work of transporting
hydrophobic minerals to the froth. In flotation
cells, we add air, create bubbles of a certain size
and speed that provide the surface area to do
the flotation. The more bubbles and the smaller
the bubble, the more surface area we have to do
the work. This surface area we create is known
as the surface bubble flux (Sb) and controls the
kinetics of flotation. Now that we have
instruments that can measure the air into a cell
(known as Jg), measure the size of the bubble
diameter (Db) and the gas hold up (Eg), we can
figure out how the relationship between these
parameters and how they affect the Sb and
flotation circuit performance. We can also now
do research on how reagents can be used to
control these parameters as well.
“Research of the last few years has shown
that frothers actually play a much more
important role in flotation hydrodynamics than
once thought. Frothers perform two major
functions. They create and maintain small
bubbles in the pulp to transport the minerals
and they create the froth on top of the cell to
hold the minerals until they can be recovered.
The froth is created because frothers allow a
film of water to form on the bubbles which
makes them stable enough not to break when
they reach the surface of the cell. Fortunately,
the water drains over a short period of time and
the froth will eventually break down. Froth
breakdown is essential for cleaning and
transporting the concentrates. Small bubbles
are essential in making flotation efficient. For
the same volume of air in a cell, smaller bubbles
give much higher surface area, which in turn
gives much higher kinetics.
“We now know that as you increase the
concentration of frothers to the cell, the bubble
size gets smaller, and the film of water on the
bubble gets bigger. But bubble size does not
keep getting smaller forever. The frother will
reduce the bubble down to a certain size, which
is about the same for all frothers in the same set
of conditions. The concentration of frother
where the bubble is at a minimum is known as
the critical coalescence concentration or CCC.
Each frother has a different CCC. Each frother
also has a different ability to add water to the
bubble and hence provides different froth
stability. This also changes with concentration.
We have learned in the last few years that each
frother has a hydrodynamic curve which relates
the bubble size with the froth stability. Strong
frothers give very high froth stability at the CCC,
while weak frothers give very low stability of the
froth at the CCC.
“This new understanding of how frothers
affect flotation cell hydrodynamics has lead to
new methodologies to optimise flotation
circuits. Flottec has worked on an optimisation
system where a frother is added to a circuit at
the CCC (which guarantees maximum kinetics or
maximum Sb) and the performance is measured.
Then frothers of different strength are added
(always at the CCC) until the right strength for
maximum performance is determined. Adding
the frother at the CCC is the critical optimisation
difference. By doing this you are always
guaranteed to have maximum kinetics. If the
frother used is too strong, the dosage will have
to be cut back below the CCC or the froth will be
too persistent. This lowers flotation kinetics. If
the frother is too weak, too much has to be
added to get the froth strength and this
increases cost and likely reduces recovery.
Flottec has been conducting research with
McGill University to develop the hydrodynamic
curves and CCC for all families of frothers in
order to implement the new methodology of
frother optimisation in plants.
“The next step in this research is to be able to
use new sensor technology to measure and
control the flotation system by controlling the
hydrodynamics in the cell. With our current
knowledge of how air rate, cell levels, and
frother addition affect bubble size, water
recovery and gas hold up, we can use these
control variables to maintain the optimum
hydrodynamics in the cell resulting in the
optimum flotation circuit performance. Flottec is
working with companies like Cidra to develop
new sensors that can provide real time
information on cell hydrodynamics (gas
dispersion parameters) and on froth stability
properties in order for us to optimise the
reagents and operating strategies used in a
plant. This will bring flotation performance to
the next level.”
Clariant Mining Solutions business is
investing considerably in mining chemicals. It
has opened a new laboratory at its US
headquarters in Houston, Texas, dedicated to
the development and optimisation of chemical
solutions for North American customers. The
laboratory is part of a planned multi-million
dollar investment into Clariant’s global Mining
Solutions business, which includes establishing
several new Mining Solutions laboratories
around the world. This network is intended to
enable the business to better support customer
needs and address regional challenges. The new
laboratories will complement existing facilities
in Europe and Latin America.
FLOTATION
36 International Mining | NOVEMBER 2011
Flottec frothers – hydrodynamic profiles - foamheight versus gas holdup, 5 litres/min
“Mining is a strategic focus area for Clariant,”
said Christopher Oversby, Global Head of
Clariant’s Oil & Mining Services business unit.
“This investment further demonstrates Clariant’s
ongoing commitment to providing innovative
technologies and solutions for our mining
customers around the world.”
The Houston laboratory will process ore
samples from customers in the USA and Canada.
These samples were previously handled in
Clariant’s mining laboratories located in South
America and at the company’s global research
facility in Frankfurt, Germany. “We are very
excited about the new mining laboratory and the
opportunity it provides us for offering our North
American mineral processing customers even
more localised services and attention,” said
Paul Gould, Global Head of Marketing and
Application Development for Clariant Mining
Solutions. “The Houston lab will allow Clariant
technicians to more efficiently develop
optimised reagent solutions for our US and
Canadian customers.”
Additionally, Clariant is in the process of
developing a new Clariant Innovation Center in
Frankfurt at a cost of €50 million. Employing
nearly 500 people and covering 30,000 m2, the
facility will focus on customers using an integrated
multidisciplinary approach to problem solving.
Clariant says “an open innovation approach on
joint ventures with external partners will ensure
the acceleration of the ‘idea-to-market’ process.
Mining research and development will also be
part of this facility.”
Axis House has been developing reagent
technologies for the past 10 years, at its
flotation laboratory in Cape Town, South Africa
and more recently at it metallurgical labs in
Sydney and Melbourne. These were acquired
when Axis House bought the oxide flotation
reagent technology from Ausmelt Chemicals. A
practical application technology strategy was
followed with Axis House providing a
complimentary suite selection and optimisation
service to its clients, who were then mainly
interested in the Axis developed technology of
combining fatty acids, hydroxamates and
sulphidisation suites to effectively and
economically float oxide minerals.
Early on the focus was on developing
reagents to float complex ores which contained
multiple minerals with varying flotation kinetics
Often the limiting factor was not only the
sluggish flotation kinetics of the minerals but
the process plant’s own equipment limitations,
like flotation and conditioning times. Developing
a reagent that floated a certain mineral was
simply not enough. The solution was to develop
suites of reagents which could function
synergistically. By altering the types of
collectors and the dosages, the company could
optimise both the use of the processing
equipment and the collecting power. It says
“this approach has successfully been applied to
various types of base metal oxide ores.”
It is now taking this innovative approach into
the field of rare earth element (REE) flotation.
This fits into the Axis House business plan as
the chemistries are quite similar to what is in
existence at Axis already. Of course some
tweaks will have to be made to the reagents as
well as the laboratories – this process has
already started, with the first batch of REE test
material having arrived at Cape Town, and new
reagent samples at the ready. There are a large
number of REE projects coming online in the
next few years. Most of these orebodies have
not been previously treated at industrial level
and so will face difficulties when scaling up.
REO (Rare Earth Oxides) are often difficult to
float and the development of multiple collector
systems for these ore types would help increase
the viability of these projects.
Jerry Sullivan, Global Marketing Manager-
Mineral Processing, Cytec Industries Inc,
discussed collectors, which contain mineral-
selective functional groups. “They have a
hydrophobic hydrocarbon tail. Changing the
molecule’s functional group changes the
preference for what mineral it will adsorb on to.
Changing the length of the hydrocarbon chain
changes the hydrophobicity of the molecule.
This is related to the strength of the collector.
“Within the collector molecule, there are
donor atoms whose goal is to form bonds with
acceptor atoms within the ore. Nitrogen,
oxygen, and sulphur are the most important
donor atoms in all reagent chemistry. Sulphur
is the most important donor in sulphide
collectors. Nitrogen and oxygen are additional
donor atoms. Phosphorous and carbon are
central atoms carrying the donors. They only
have indirect participation in interactions.” He
noted the general characteristics of sulphide
collectors to be:
■ Ionic collectors are stronger and less
selective
■ Neutral, ‘oily’ collectors are weaker, more
selective
■ Higher homologues (more carbons) are
stronger than lower homologues (fewer
carbons)
■ Cytec’s NCPs are very selective collectors
■ Selectivities of collectors have been
extensively studied, and are well established
in terms of:
■ Mineralogical preferences
■ pH effects.
“There is a strong case for formulated
products (or blends),” he continued “That is
because mineralogy is complex. Plant
performance is also inherently variable.
Mineralogy changes routinely. In addition,
different minerals have different affinities for
reagents. Various minerals will compete for a
given reagent. Modifiers used will also influence
reagent partitioning. Particle size distribution
will also affect recoveries (recovery losses in
coarse and fine size range). A single collector
will not be sufficiently robust. Indeed, most
plants use two or more collectors. The goal is to
pick reagents that will get to the right minerals.
Utilising a collector blend can balance cost and
performance.
“Cytec has multiple collectors and collector
blends that are continuously being developed to
tailor to the customer’s application.” A few of
the collector families that have recently been
introduced to the market include the new XR
Series Xanthate Replacement Collectors,
developed to meet the desire to replace
xanthates. “This new series of collectors are
cost competitive with xanthates and are strong
collectors but with high selectivity. In addition,
they are safer and vastly improves handling and
level of toxic exposure of the personnel to
product, stock safety management and
simplifies plant operations.
The XD 5002 blends were developed to
operate in a broad pH range 8-12 and be highly
selective in Cu/Mo, Cu/Au sulphide ores,
enhance Mo recovery in Cu/Mo bulk float and
enhance Au recovery in Cu/Au ores. The
MAXGOLD™ blends were introduced to float
primary Au ores; auriferous pyrite, arsenopyrite,
and tellurides and are also capable of enhancing
recovery in Cu/Au ores.
Monitoring and controlIt is now possible to use measurement devices
based on impedance tomography to create real-
time 3D images. The technology opens up
entirely new possibilities in controlling flotation
processes. “With Flotation Watch the operator
can see what takes place underneath the
surface. Flotation Watch measures several
parameters at the same time, on-line. The
sensor can measure the stiffness of the froth,
the thickness of the froth, analyse the interface
area between the froth and the slurry and it can
analyse the slurry too depending on the
customer needs,” says Jukka Hakola, Numcore’s
Vice President of Sales and Marketing.
With Numcore measurement devices, the size
and quantity of air bubbles and the solid matter
content of the froth bed can be monitored by
means of electric conductivity distribution.
“With Flotation Watch the stiffness of the
flotation froth can be measured and this helps
to keep the recovery in higher level. The signals
for the production failures, such as hardening
and collapse of the froth bed, can be seen
beforehand and avoided. This way we can help
FLOTATION
38 International Mining | NOVEMBER 2011
to minimise the losses in the flotation process,”
says Hakola.
Real-time characteristics are a key in this
technology; in other words, the system
continuously provides the operator with factual
data on what is happening in the flotation cells,
for example the location of minerals and the
bottom surface of the froth bed. “Because it has
not been possible to look inside tanks,
controlling a mineral concentration process has
largely been based on experience-derived
knowhow. Now that operators can ‘look’ inside
the process, it is possible for them to maintain
an optimal mix all the time,” says Hakola.
Numcore has, in close co-operation with a few
key customers, developed measurement
technology to better serve everyday work. “We
have now delivered several Flotation Watch
sensors to flotation cells in several markets and
for different metals such as copper, zinc and
gold. One of the main benefits is that
contamination of the probe is taken into account
in mathematical formula and the measurement
probe does not need to be cleaned. Our sensor
has been in a zinc rougher flotation cell for nine
months and is giving accurate results to the
operator. We can now offer automated control
for flotation process with Flotation Watch and
see that this can bring new benefits for our
customers,” he promises.
Numcore’s measurement technology is
currently in test use at Inmet’s Pyhäsalmi
copper-zinc mine (IM, April 2010,
pp10-18), among others.
According to Seppo Lähteenmäki,
Processing Mill Manager, the
system has provided accurate
information on the condition of
the froth bed, and the technology
has functioned reliably. “We have
tested the device for a few
months, and it has provided clear
benefits for those operators who
have received operator training for
it and actively monitored the data
provided by the system. The
device appears to be so useful, in
fact, that we are seriously
considering buying it after the test
period,” he says.
Mettler Toledo notes that pH greatly
determines the efficiency of the flotation, which
minerals will float, or even if there will be any
flotation at all. The critical pH value for efficient
flotation depends on the mineral and the
collector. Below this value the mineral will float,
above it, it will not (or, in some cases, vice versa).
In a recent white paper www.mt.com/pro-ph-
flotation, the company says “in order to
overcome difficulties with the hostile
environment in flotation cells, sensor
manufacturers are very creative in their choice of
sensor design. It is possible to find pH
electrodes with a ceramic, plastic, rubber or
even a wood reference diaphragm. Still, their
performance can be severely limited as the
colloidal particles and sulphides interfere
almost instantly with the reference system. The
sensor’s maintenance requirement is therefore
high, requiring very frequent cleaning and
calibration, and usually sensor life is short.”
Mettler Toledo has acknowledged this issue
and to combat it has designed the InPro 4260i
pH electrode with Xerolyt® Extra solid polymer
electrolyte. The InPro 4260i does not have a
diaphragm and instead features an open
Need a better way to free your froth?
Worldwide Slurry Pumping Solutions
Install the HVF (High Volume Froth Pump.) Our pump removes air from the impeller eye while running, which allows continuous operation and improved efficiency. This pump won’t stop you up.
Designed for air-entrained slurries, the HVF pump can be used in phosphate, hard rock mining and oil sands applications.
HVF Pump
Got a froth problem we can solve? GIW offers expert slurry pumping solutions around the globe. Minerals are everywhere. So is GIW.
GIW Industries, Inc.
Relation between collector and critical pH value
FLOTATION
NOVEMBER 2011 | International Mining 39
junction, which is an opening that allows direct
contact between the process medium and the
electrolyte. Contrary to the miniscule capillaries
of any other type of diaphragm in conventional
pH electrodes, the diameter of the open junction
is extremely large and much less susceptible to
clogging or fouling. Another significant
difference is in the choice of polymer electrolyte.
Xerolyt® Extra was designed specifically for
service in tough environments to provide a strong
and lasting barrier against sulphide poisoning.
The company’s Intelligent Sensor
Management (ISM) is a platform based on
sensors with embedded digital technology for
better pH management. The integrated system
consists of a digital sensor and
ISM-compatible transmitter. The
key to the technology is a
microprocessor which is contained
within the sensor head and is
powered by and read through the
transmitter. Critical sensor
information such as identification,
calibration data, time in operation
and process environment exposure
are all recorded and used to
continuously monitor the health of
the sensor.
By constantly keeping track of process pH
value, temperature and operating hours, ISM
calculates when sensor calibration, cleaning or
replacement will be needed. Any need for
maintenance is recognised at an early stage.
In recent years, researchers at Imperial
College have been focusing on measuring air
recovery in industrial flotation cells and have
found that a peak in metallurgical performance
(improvements in both grade and recovery)
corresponds well with a peak in air recovery.
Major platinum and copper operations have
already observed the benefits of using this
methodology as developed by the researchers.
JKTech is now licensed by Imperial Innovations
to commercially provide this methodology and
associated benefits to the global minerals industry.
The PAR technique comprises two stages –
evaluation and implementation. The evaluation
stage involves determining the effect of the
technology at a mine site, typically determining
the peak air recovery for a bank (or banks) of
flotation cells and evaluating the resultant
metallurgical performance. The implementation
stage involves setting the air rates to those that
maximise the air and/or metal recovery, and
support and training of site personnel including
operating manuals. The implementation stage
requires an end-user license to be obtained by
the sites through Imperial Innovations.
Pumping frothGIW Industries has launched its new High
Volume Froth (HVF) pump. Unlike any other
pump on the market, GIW says, the HVF pump
can pump froth without airlocks. It provides
continuous operation without shutdown or
operator intervention. The new hydraulic design
actually removes air from the impeller eye while
the pump is running, so you can keep your
process moving and improve efficiency.
The GIW HVF can be retrofit into many
existing froth applications. The pump's de-
aeration system includes a patent-pending
vented impeller and airlock venting. This helps
to eliminate sump overflow due to pump airlock;
reduce downtime; and allow water use to be
restricted to the bare minimum. Fewer pumps
are required for less capital expense, requiring
less water and power usage.
The HVF pump has been fully tested on froth
and viscous liquids. The pump exceeded
expectations at a large phosphate company in
Finland. The company's existing pumps were not
able to provide the required flow and were
airlocking at only one-third of process design
capacity. After installing an HVF pump, the
company achieved a flow of 415 m3/h.
Traditional slurry pumps are prone to airlock
when working with slurries that incorporate
froth. A pump works by pulling in a liquid at a
certain pressure and adding mechanical force to
expel the liquid at a higher pressure. The air in
the froth does not want to move to a higher-
pressure zone, and it is prone to build up at the
lower-pressure pump entrance. The
accumulation of air can eventually block the
pump entrance completely, leading to airlock,
which requires pump shutdown or operator
intervention to avoid sump overflow.
How is GIW’s HVF pump different? The main
innovation is in the impeller design. Typically, air
bubbles gather at the centre of the impeller as
the heavier fluids are spun to the outer edges.
The HVF pump's de-aeration system includes
the vented impeller and airlock venting. In the
HVF pump, small holes in the centre of the
impeller allow air bubbles to pass through to a
separate port. The port vents air up and out of
the pump to normal atmospheric pressure.
Any liquid that passes through the port is
returned to the process tank. Air is no longer
building up at the impeller eye or pump
entrance, so airlock is avoided. IM
References1. Schubert, H. "On the optimization of hydrodynamics
in fine particle flotation." Minerals Engineering 21,
2008: 930-936.
2. Jameson, G. J. "New directions in flotation machine
design." Minerals Engineering 23, 2010: 835-841.
Designed for air-entrained slurries, the pumpcan be used in phosphate mining, hard rockmining and oil sands. The pump offersimproved efficiency and is environmentallyfriendly and cost-effective, GIW reports
40 International Mining | NOVEMBER 2011
FLOTATION