Int. J. Miner. Process.

91 (2009) 19–27

Contents lists available at ScienceDirect

Int. J. Miner. Process.

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i j m i n p r o

Quantifying contributions to froth stability in porphyry copper plants

M. Zanin a,⁎, E. Wightman b, S.R. Grano a, J.-P. Franzidis b
a
Ian Wark Research Institute, University of South Australia, (The ARC Special Research Centre for Particle and Material Interfaces), Mawson Lakes Campus,
Adelaide, South Australia 5095, Australia
b
Julius Kruttschnitt Mineral Research Centre, University of Queensland, Isles Road, Indooropilly 4068, Australia

a r t i c l e i n f o a b s t r a c t

Article history: A method to describe froth stability in industrial flotation cells, and its link with froth recovery, is discussed
Received 15 January 2008 in this paper. The direct measurement of froth stability in industrial flotation cells is still controversial, and
Received in revised form 12 November 2008 different techniques have been developed. In this paper, the froth half-life time and bubble size on the top of
Accepted 14 November 2008
the froth are discussed as indicators for froth stability. Experiences at two flotation concentrators treating
Available online 27 November 2008
porphyry copper ore are compared. The relative contributions to froth stability from frother in solution and
solids loading in the froth were investigated in both concentrators, and links were found between both the
Keywords:
Froth flotation
froth half-life and bubble size on the top of the froth and the amount and nature (hydrophobicity) of
Flotation froths particles in the concentrate. The comparison also suggests how two significantly different operating
Process instrumentation strategies for the froth phase may lead to similar outcomes in terms of froth recovery.
© 2008 Elsevier B.V. All rights reserved.

1. Introduction and collapse rate, which were calculated by automated image

In the flotation process it is clearly recognised that there are two
distinct particulate systems: a pulp phase and a froth phase. Only
relatively recently (in the past two decades) has the froth zone been
recognised as a region which contributes significantly to the overall
flotation performance. Froth is a complex three phase system and the
structure and stability of the froth are critical in determining its ability
to transport particles into the concentrate launder.
Froth stability is dependent not only on the type and concentration
of frother used but also on the nature and concentration of particles
present in the system, in particular the effects of particle hydro-
phobicity and size (Dippenaar, 1982; Johansson and Pugh, 1992;
Schwarz and Grano, 2005). Ata et al. (2003) investigated changes in
bubble size in the froth phase of a flotation column and found bubble
growth rate to be sensitive to the degree of hydrophobicity of particles
present in the froth zone. Maximum froth stability was found to occur
with moderately hydrophobic particles, although a high coalescence
rate was not necessarily related to low flotation recovery since highly
floatable particles may reattach in the froth (Ata et al., 2003).
To date, much of the work investigating the role of particles in
froths has focussed on idealised systems, with particles whose
properties are known. In the recent years, however, the emphasis has
shifted to understanding the role of particles in controlling froth
stability in an industrial setting (Subrahmanyam and Forssberg, 1988;
Barbian et al., 2005; Hadler et al., 2006; Morar et al., 2006). In the
work by Runge et al. (2007), a correlation was found between copper
grade of the concentrate produced by plant cells and froth velocity
⁎ Corresponding author. Tel.: +61 8 8302 3263; fax: +61 8 8302 3683.
E-mail address: massimiliano.zanin@unisa.edu.au (M. Zanin).
analysis of the top of the froth. A prediction of concentrate grade
down the bank was formulated, based on froth velocity and collapse
rate values. In this paper, measurements of froth stability (half-life)
and froth appearance (bubble size on the top of the froth) in selected
cells down banks of cells are correlated to the composition of the
froth in terms of both copper grade and percent solids. It is believed
that the amount of suspended hydrophobic particles (which is
proportional to the product of copper grade and percent solids) is
more realistically linked to froth stability than copper grade alone.
Contributions to the change in froth stability from changes in the
content of hydrophobic particles in the froth are probed at two
different porphyry copper concentrators. The comparison between
the two plants also suggests that optimum metallurgical performance
of the froth is a trade off between a number of froth parameters so
that different operating strategies may lead to similar outcomes in
terms of froth recovery.
2. Theoretical background
2.1. Froth recovery and froth stability in mineral flotation
During flotation, only a fraction of the particles that are attached to
air bubbles in the collection zone may be effectively transported to the
product launder. The recovery of particles attached to bubbles across
the froth phase is referred to as froth recovery. Savassi et al. (1997)
gave a definition of froth recovery as:
Mass rate of particles reporting to the concentrate via true flotation
Rf =
Mass rate of attached particles at the pulp − froth interface
ð1Þ

0301-7516/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.minpro.2008.11.003
20 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27
Direct measurement of froth recovery (as defined in Eq. (1)) is still
problematic, although some techniques have been proposed (Alexander
et al., 2003). An indirect method was developed by Vera et al. (1999),
which related froth recovery to the first order flotation rate constants.
Several indicators have been proposed to characterise an industrial
froth. One of these is the froth retention time (FRT), defined by
Laplante et al. (1983) as:
Hf
FRT =
Jg
where Hf is the froth depth and Jg is the superficial gas velocity. The
froth retention time is physically a measure of the average lifetime of
an air bubble in the froth system (Malysa et al., 1987). For a given froth
type, the longer the froth retention time FRT, the greater the
probability for an attached particle to be dropped back into the slurry.
Increasing the froth depth increases water drainage and reduces
entrainment, which is often desirable. On the other hand, the froth
recovery of the valuable minerals may also be reduced. A more
accurate definition of the mean froth retention (or residence) time
was given by Zheng et al. (2006), who considered that: a) not all of the
volume of the froth is occupied by air, and b) due to bubble bursting on
the surface, only a fraction of the total air flow introduced to the cell
reports to the concentrate. For practical use, however, Eq. (2) is used in
this work to express the froth retention time at plant scale.
Froth stability is another critical factor determining froth recovery
but as yet there is no unique measure, or precise definition, of froth
stability. Over the years many measures of froth stability have been
employed, including air recovery (Moys, 1984; Morar et al., 2006),
bubble size and solids loading on bubbles at the top of the froth
(Ventura-Medina et al., 2003), bubble growth rate (Ata et al., 2003),
froth rise velocity (Barbian et al., 2005) and froth decay or persistence
(Iglesias et al., 1995). In the latter method, a characteristic time is used
to describe froth stability (half-life time, t1/2), which is the time taken
for the foam to drop to one half of its initial height (Hf max) when the
gas flow is discontinued. Following this approach, the froth decay
process can be expressed in the following form:
! slurries, which has been demonstrated on a series of common frothers
Hf 1 t
= − α  ln
Hf max 2 t1 = 2
where Hf is the height of the froth at time t and α is an empirical
dimensionless constant. Tsatouhas et al. (2006) used the static test
and the half-life time to characterize the stability of mineral flotation
froths, both in laboratory and plant flotation cells, finding good
agreement.
2.2. Relationship between froth recovery and froth stability
Froth recovery is obviously related to the froth stability: stable
froths ensure high froth recoveries, while froths showing rapid bubble
collapse and volume variation are responsible for lower froth
recoveries. Gorain et al. (1998) found an inverse relation between
froth recovery and froth retention time FRT:
− βFRT
Rf = e
where β is a dimensionless constant, related to the physical and
chemical properties of the froth. Tsatouhas et al. (2006) proposed a
modification of Eq. (4) in which the froth half-life time is considered
as a measure of froth stability:
− t FRT
Rf = e 1=2
The relation allows froth recovery to be liked to two measurable
indicators of the froth phase performance (froth retention time and
froth half-life time). Nevertheless it should not be forgotten that
Eq. (5) is also empirical, and that factors such as frother concentration
and suspended solids are not explicitly included.
3. Experimental
In the context of the AMIRA research project P541B “Optimising
Froth Zone Performance in Mineral Flotation”, an analysis and
comparison has been made of the froth zone performance in two
ð2Þ
concentrators, both treating porphyry copper ore. The two ores are
however different in composition: chalcocite (Cu2S) and covellite
(CuS), with lower chalcopyrite (CuFeS2), are the main copper bearing
minerals in the ore treated by concentrator A (South America), while
chalcopyrite and bornite (Cu5FeS4) are the main copper bearing
minerals treated by concentrator B (North America). Furthermore,
concentrator A treats a relatively high grade ore (averaging 1.5% Cu
head grade), while concentrator B treats a low grade ore (averaging
0.6% Cu head grade). Both concentrators employ crushing, SAG milling
and flotation circuits.
3.1. Tools and methods
In each of the two concentrators, the assessment of the froth phase
was performed by a combination of measurements, comprising froth
visual analysis as well as the determination of the residual frother
concentration in solution, the froth half-life (in-situ of the flotation
cells), and the froth retention time, FRT. These measurements were
undertaken in the context of plant metallurgical surveys, in which
metallurgical samples were collected from the lip of each individual
cell down the rougher/scavenger lines in which the froth phase was
assessed.
Frother concentration in solution was measured on samples of
slurry collected down the banks. The measurement of frother
concentration in solution is often complicated by possible degradation
of the surfactant during handling and storage prior to the analysis.
Gelinas and Finch (2005) have developed a colorimetric technique
allowing rapid, on-site, determination of frother concentration in
with alcoholic behaviour. In this work, however, the analysis was
ð3Þ
carried out off-site; therefore, proper sample preservation procedure
was developed. This involved filtering the slurry samples and storing
the filtrates at low temperature (4 °C), in closed opaque plastic bottles,
prior to the analysis. The pH in the bottles was also raised to 12 to
prevent microbial degradation of the frother during storage. Frother in
solution was measured at the Ian Wark Research Institute, in Adelaide,
by gas chromatography with flame ionisation detection (GC-FID),
following the method of Huang et al. (2002). In this method, the
chromatographs of plant samples are compared with reference
chromatographs generated with clean solutions of the frother of
interest. In this way, the interference of any other species in solution
(background signal) is removed. A database of commercial frothers
was built within a previous research project (AMIRA P541A), and
calibration curves were generated for the GC-FID analysis.
The froth half-life was measured in-situ of the flotation cells by
means of a modified froth stability column (Fig. 1). This has the same
ð4Þ general design as the column of Barbian et al. (2005), but it is also
equipped with a bottom valve, which is closed when the equilibrium
froth height is reached in the column, shutting off the gas flow and
allowing the measurement of the half-life of the froth. In each test, the
column was inserted downward through the froth in the cell to a
location below the froth/slurry interface. The bottom valve was kept
closed while the column was set in position. Once the column position
ð5Þ was set, the valve was opened, allowing froth formation inside the
column. When the equilibrium froth height was reached, the bottom
valve was closed again, at which point the rate of froth collapse and
the forth half-life time, t1/2, were measured.
 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27 21

Table 1
Summary of the general metallurgical performance of the two concentrators in the
plant surveys undertaken.

Concentrator A B
Rougher feed tonnage (to one line) [tph]: 865 1380
Cells number and volume 9 × 160 m3 11 × 85 m3
Total residence time [min]: 26 12
Rougher feed solids [%]: 31 30
Rougher feed d80 [µm]: 150 220
Rougher feed Cu grade [%]: 1.4 0.5
Rougher feed Fe grade [%]: 1.8 2.3
Cu/Fe ratio: 0.78 0.22
Cu sol./Cu T: 0.085 n.a.
Rghr/Scav Cu recovery [%]: 88 86
Rghr/Scav Cu grade [%]: 12.0 11.9
Enrichment ratio: 9:1 24:1
Rghr/Scav mass yield [%]: 9.6 3.6
Rghr/Scav water recovery [%]: 12.5 7.8
Frother [g/t]: 10 (blended) 25 (MIBC)
Collector [g/t]: 6 20
NaSH [g/t]: 6 –
pH: 10.2 10.0

differences between the two operations: the volume of the cells in

concentrator A is double that in concentrator B, as is the total
residence time down the rougher/scavenger banks. The size distribu-
tion of the feed to the two concentrators was also different, being
much coarser in concentrator B. The overall metallurgical perfor-
mance of the rougher/scavenger banks was however very similar:

Fig. 1. Device for the measurement of the froth half-life in situ of plant flotation cells.

At the same time as the froth half-life, froth depth (FD) and
superficial gas velocity (Jg) were measured in situ of the cells. Froth
depth was measured manually, while the superficial gas velocity was
measured using a JKMRC Jg probe (Gorain et al., 1996). The froth
retention time FRT was calculated, by means of Eq. (2).
Continuous images of the top of the froth were collected from each
cell using a Canon digital video camera mounted vertically above the lip
of the cell. Images were captured for approximately 5 min, and
subsequently processed using Visiofroth™ software, developed by
Metso Minerals Cisa, to extract a number of parameters including
froth velocity and bubble size distribution on top of the froth. A full
description of the Visiofroth™ system and the derivation of the
parameters may be found in Runge et al. (2007). Several thousand
bubbles were analysed to obtain information on the bubble frequency
distribution, number mean (average) bubble diameter (d10) and Sauter
mean bubble diameter (d32). In most flotation studies the interfacial
area is a control parameter hence the Sauter mean is the most
appropriate average bubble diameter to represent the bubble size
distribution (Ata et al., 2003). The Sauter mean bubble diameter is most
often used to represent the bubble size distribution in the collection
zone of the pulp; in this work, it has also been used to represent the
bubble size distribution on top of the froth. It is defined as the volume to
surface mean bubble diameter and is described by Eq. (6).

3 2
d32 = Σni di = Σni di ð6Þ

4. Results and discussion

4.1. Summary of the general metallurgical performance of the two

concentrators

A summary of the general metallurgical performance of the two Fig. 2. Size-by-size recovery of copper down the rougher/scavenger banks of
concentrators is reported in Table 1. There are several major concentrator A (a) and concentrator B (b), at the time of investigation.
22 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27

total Cu recovery was about 87%, at 12% Cu grade, for both

concentrators.
On a size-by-size basis, the cumulative recovery of copper down the
bank was also similar (Fig. 2), with recoveries N90% for the −38 µm
and −106 + 38 µm fractions, decreasing to about 70% for the +106 µm
fraction. The low recovery for the coarse particles may be, in part, related
to the froth phase performance. The coarse particles may have a higher
probability of detachment from the froth phase, and thus require more
stable froths to be recovered efficiently. Detailed mineralogical and
liberation analysis (not available at the time of investigation) would be
necessary to confirm these interpretations.
Other notable differences between the two concentrators, which
would be expected to affect the overall metallurgical performance, are
that concentrator A uses a blended, relatively strong, frother (at a
concentration of about 10 g/t, added at the head of the roughers) in
combination with a deep froth in the cells (18–25 cm), while
concentrator B uses MIBC (a weaker frother, which is stage added at
the head of the rougher cells and in the scavenger cells, at a total
concentration of about 25 g/t) in combination with a shallow froth in
the cells (5–10 cm). The implications of the two different operating
strategies will be discussed further below.

4.2. Froth stability down the bank

Typical froth formation and decay curves, as measured in-situ of

rougher cell 1, are presented in Figs. 3 and 4 for concentrators A and B,

Fig. 4. Froth formation (a) and froth decay (b) profiles measured in concentrator B, cell
1. The equations of the regression curves are also reported. (⁎) The R-squared value for
the froth decay regression considers only the region t N 5 s.

respectively. According to Eq. (3), the regression curve for froth decay
tends to +∞ for t → 0, while the experimental data obviously tend to 1.
The model generally fits the data well in the region t N t1/2, while it
diverges for the first few experimental points, at t bb t1/2. Therefore,
the R-squared factor for the regression was calculated considering
only the experimental points for t N 5 s.
Measurements were repeated in selected cells down the rougher/
scavenger banks in both concentrators A and B, and froth half-life
times are reported in Fig. 5, together with solids percent and copper
grade in the froth, for each of the cells in which measurements were
made.
It can be seen that in concentrator A the froth half-life time was 57 s
in the first cell, while in concentrator B it was only 18 s. It would appear
that the difference in froth stability between the two plants is in part due
to the different frothers (stronger frother in concentrator A), and in part
due to the different solids percent and copper grade in the feed (both
higher in concentrator A). A significant decrease in froth stability was
measured down the bank, and in both concentrators, with half-life times
of the order of 16 s and 5 s in the last cells of concentrator A and
concentrator B, respectively. This decrease in froth stability correlates
well with the decreasing mineralisation of the froth, while decreasing
frother concentration in solution is also possible. This latter possibility is
explored below.

4.3. Deportment of frother in solution

Fig. 3. Froth formation (a) and froth decay (b) profiles measured in concentrator A, cell
1. The equations of the regression curves are also reported. (⁎) The R-squared value for Concentrators A and B use frothers which have very different
the froth decay regression considers only the region t N 5 s. surfactant properties. Concentrator A uses a blended frother, which
 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27 23

Fig. 5. Froth half-life times (s) measured in selected cells down the rougher/scavenger banks in concentrators A and B. Solids percent and copper grade of the different streams are
also reported as a reference.

was shown in the laboratory to be relatively stronger than MIBC,
which is used in concentrator B. MIBC is well known as a weak frother,
and it is mainly used where selectivity is sought. In order to draw a
link between froth stability and the effect of particles and frother
concentration, a balance of the frother in solution in the flotation
products was attempted. The situation was significantly different in
the two concentrators. The balance for concentrator A is presented in
Fig. 6.
Frother concentration in solution in the feed to the first cell of the
rougher/scavenger bank (7.5 ppm) was not much greater than the
concentration in the last cell (6.9 ppm), suggesting that differences in
froth stability were probably more related to the suspended solids
than to frother concentration in solution. The blended frother
appeared to have high solubility in most of its components, and
frother deportment down the circuit was mainly driven by water
recovery, with limited enrichment in the concentrates. About 80% of
the total frother in solution reported to the tailings stream. The
recirculated process water contained about 3.5 ppm of frother,
contributing roughly to one half (54%) of the total frother in solution
at the head of the roughers (the remaining 46% being fresh reagent
added). Process water was tested in the laboratory, showing
pronounced frothing properties.

Fig. 6. Balance of frother in solution in concentrator A. The water flows are reported, together with the measured frother concentration in solution. The distribution of frother in the
flotation products is reported as a percentage of the total frother in the feed.
24 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27
Fig. 7. Measured froth half-life time versus predicted half-life in situ of the cells. Down
the bank analysis at concentrator A (a) and concentrator B (b). Linear regression
equation and R-squared values are also reported.
In concentrator B, measurements of MIBC concentration down the
bank revealed that the frother was disappearing from solution rapidly,
leading to low residual concentration. A balance of frother deportment
was not possible. It is believed that this was due to the high volatility
of MIBC, although the possibility of reagent adsorption on the mineral
surfaces is presently under investigation. Stage addition of frother is
practiced in concentrator B, in order to maintain a reasonable
concentration in solution in the scavengers (concentrations around
3 ppm were measured). Because of the rapid disappearance of MIBC
from solution, process water in concentrator B did not contain any
residual frother in solution, having no frothing properties as a result.
4.4. Prediction of froth stability down the bank
The comparative analysis of froth stability and froth composition,
as reported in Fig. 5, suggests a link between froth stability and frother
type and concentration, and particles in the froth. Although the
decrease of frother concentration down the bank may contribute to
the corresponding decrease in froth stability, it is unlikely that the
large difference in froth half-life measured between first and last cell
of the banks is driven only by differences in frother concentration. The
contribution from the particles present in the froth appears evident.
The following linear regression was used to fit the froth half-life time
measured in each of the selected plant cells:
t1 = 2 = a + b  solidsk  grade
where solids% represents the amount (percent by weight) of
suspended particles in the froth, grade represents the fraction of
hydrophobic particles within the suspended solids (copper grade is
used in the calculations), and a and b are the fitting parameters. Eq. (7)
links the froth half-life time to the amount of hydrophobic particles
suspended in the froth (measured by the product solids% · grade).
Fig. 7 presents the predicted half-life times versus the measured
values, for both concentrators A and B.
The experimental points show some scatter, caused both by the
fluctuations in plant conditions with time and the standard deviation
of a single half-life time measurement. The regression gave R-squared
values of 0.89 for concentrator A and 0.90 for concentrator B.
The parameters a and b of the regression equation can be used to
characterise the frothing response of a system, which is particularly
useful when the effect of different frothers on the resulting froth
stability needs to be predicted. For a given mineral system, a and b are
characteristic of the frother type (and concentration, to a lesser
extent). The y intercept, a, represents the half-life time of the froth in
the absence of particles, and the slope, b, gives a measure of the
dependence of the froth half-life on the stabilising effect of particles.
Tests carried out at the Wark with different frothers indicate that
frothers with high glycol content tend to give higher a values and
lower b values (denoting high froth stability regardless of the
contribution of particles), while alcohol based frothers tend to give
lower a values and higher b values (denoting low stability in the
absence of particles and the higher stabilising effect of particles).
Within the presented study, the comparison between the regression
equations for concentrators A and B confirms the findings of the
laboratory scale tests:
- a stronger frother is used in concentrator A (a = 8.8) compared to
concentrator B (a = 4.5).
- the presence of solids has a greater influence on froth stability in
concentrator A (b = 3.7) than in concentrator B (b = 1.9).
It has to be said that, at this stage, the prediction of froth stability in
the plant still needs refinement. Eq. (7) does not explicitly consider
other factors which may have an impact on the froth half-life time,
such as particle size and liberation. Research is currently underway to
improve the model, to include the effect of particle size, in the first
instance.
4.5. Bubble size distribution on top of the froth down a bank of cells
Considering the mechanisms involved, froth stability is a reflection
of drainage, bubble coalescence within the froth and bubble bursting
on the top of the froth (Ventura-Medina and Cilliers, 2002). Therefore,
it is expected that changes in froth stability measured by the half-life
of the froth should also be reflected in changes in bubble size down
the bank in some manner.
Bubble size distribution on the top of the froth was found to
change significantly down the bank at both concentrators. In
concentrator A (Fig. 8a), the bubble size on the top of the froth
was generally smaller at the end of the bank than at the head of the
bank. Starting from the head of the bank (cell 1), bubbles on the top
of the froth became larger and more polydisperse (cells 2 to 5), and
then become progressively smaller and more uniform towards the
end of the bank. Bubble size data collected down the rougher
scavenger bank in concentrator B are shown in Fig. 8b. Trends in
down the bank behaviour are generally consistent with those for
concentrator A.
Examples of the digital images of the top of the froth are shown in
Fig. 9, which illustrate that the bubbles were generally uniform in the
first cell of a bank, with bubble size tending to become more
polydisperse and bimodal progressing down the bank, with the last
ð7Þ few cells in the bank (where the froth is most barren) having bubbles
much smaller and more uniform in size. This is consistent with the
findings of Hadler et al. (2006), who found that the air recovery
(fraction of air reporting to the weir as unbroken bubbles) generally
 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27 25

Fig. 8. Bubble size distribution on top of the froth down the rougher bank in concentrator A (a) and concentrator B (b).

decreases down the bank, due to increased bubble bursting. The
loading of hydrophobic particles on the bubble lamellae has an impact
on bubble stability. Highly mineralised bubbles may become larger
with coalescence before bursting occurs, and this is reflected by the
bubble size profile down the bank (Fig. 8).
In Eq. (7), the froth half-life time was related to the amount of
hydrophobic particles in the froth. Using a similar approach and
following on from the work of Nesset et al. (2006), bubble size on top
of the froth can also be predicted using an empirically derived model.
As in Eq. (7) the amount of hydrophobic solids present in the froth is
represented by the product (% solids · Cu grade). The residence time of
bubbles in the froth is also expected to influence the extent of bubble
coalescence in the froth and is a function of the superficial gas velocity,
froth depth and the physical characteristics of the cell (e.g. presence
of internal launders). Since insufficient data is available for all cells
measured in this investigation, these parameters are lumped in the
constant C in Eq. (8).
d32 = d0 + C  ðsolidsk  Cu gradeÞ ð8Þ
Fitting this empirical model to the data presented in Table 2 results
in reasonable agreement between the measured and predicted bubble
size on the top of the froth as shown in Fig. 10.
4.6. Relationship between operating conditions and froth recovery
Eq. (5) provides a relationship between froth recovery and two
indicators of froth performance, namely the froth retention time FRT
and the half-life time t1/2. The equation is empirical, and the individual
contributions of frother concentration and feed composition are not
26 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27

Fig. 9. Digital images of the top of the froth showing the progression from uniform bubbles in the first cell to more polydisperse bubbles partway down the bank and small, uniform
bubbles at the end of the bank.

quantified. Because of these limitations, the use of Eq. (5) to predict needed to ensure adequate froth recovery. By decreasing froth depth
froth recovery is problematic. However, some useful qualitative the froth retention time is shortened, increasing the probability of a
indications can be obtained. The ratio FRT/t1/2 is an indicator of the particle in the froth reporting to the launder. This is reflected in the
magnitude of froth recovery: the lower the ratio, the higher froth FRT/t1/2 ratio, which is, in both concentrators, low in the first cells of
recovery. According to this relation, froth recovery can be increased by: the roughers (from 0.2 to 0.9), and increases progressively down the
a) increasing the froth half-life time, or b) decreasing the froth bank (up to 2.0–2.3). According to Eq. (5), these values would lead to a
retention time. The latter can be achieved by acting on process theoretical froth recovery (Rf) between 60% and 80% in the first cells of
variables such as froth depth (FD) and superficial gas velocity (Jg), the rougher. On the contrary, in the last cells of the bank, the FRT/t1/2
while the froth half-life time can be increased only by acting on the ratio shows values between 1.5 and 2.5, corresponding to theoretical
frother chemistry and, where possible, on the solids load in the froth. froth recoveries between 10% and 25%. Within this investigation, no
Tables 2 and 3 present a summary of the down the bank
measurements in concentrators A and B, respectively, collected in
the context of plant metallurgical surveys.
The operating conditions in the two concentrators were quite
different. Froth depth in concentrator A was in the range of 18–25 cm,
while in concentrator B it was in the range of 4–10 cm. This is a direct
consequence of the different frother used: in concentrator B, where
MIBC was used, the froth was much less stable (as indicated by the
half-life times measured in the cells), and shallow froth depths were

Table 2
Main parameters of the froth phase characterisation in concentrator A.
(⁎)
Cell FD Hf max Jg FRT t1/2 FRT/t1/2 Rf
# [cm] [cm] [cm/s] [s] [s] [%]
1 18 90 1.1 17 57 0.3 74
2 25 100 1.1 24 28 0.9 44
6 25 58 0.6 42 18 2.3 10
9 25 50 0.8 31 16 1.7 15

Hf max = maximum height of the froth in the instrument, at equilibrium. (⁎) Froth Fig. 10. Predicted versus measured d32 in concentrators A and B using (d0 = 7.7 mm,
recovery is estimated according to Eq. (5). C = 0.007).
 M. Zanin et al. / Int. J. Miner. Process. 91 (2009) 19–27
Table 3
Main parameters of the froth phase characterisation in concentrator B.
Cell FD Hf max Jg FRT t1/2 FRT/t1/2
# [cm] [cm] [cm/s] [s] [s]
1 5 45 1.4 4 18 0.2
3 8 43 1.1 8 13 0.6
5 10 19 1.1 9 8 1.2
11 9 29 1.1 8 5 1.8
Hf max = maximum height of the froth in the instrument, at equilibrium. (⁎) Froth
recovery is estimated according to Eq. (5).
direct measurements of froth recovery were available to validate this
relation. However, separate measurement of froth recovery has
confirmed the correlation to model froth recovery for one concentrator
(B), at different operating conditions (Zanin et al., 2008). It is
interesting to observe how two plants operating very differently
(weak frother in conjunction with shallow froth in concentrator B
versus strong frother in conjunction with deep froth in concentrator A)
can give similar values of the ratio FRT/t1/2, and similar froth recovery.
The overall metallurgical performance of the rougher/scavenger banks
of the two plants, which is in part determined by the froth phase
performance, was also the same.
5. Conclusions
The complexity that exists in industrial flotation circuits is well
known, and the stochastic nature of flotation makes it difficult to
closely control the characteristics of the feed including ore, system
chemistry and cell operating conditions. The data presented in this
paper provide information on the important role that particles play in
determining the stability and appearance of the froth in an industrial
setting. These results also provide insights into what additional
measures are needed to further understand the froth such as the
residence time of particles and air in the froth, the composition of
those particles and their ability to rupture or stabilise thin liquid films.
The present work confirms how, in flotation, different components
(frother type and concentration, amount and nature of the solids in
the froth) contribute to the stability of the froth phase as a whole.
Semi-empirical models, relating both the froth half-life and the bubble
size on the top of the froth to the amount of hydrophobic material
present in the froth, have been developed. Data collected at two
concentrators processing porphyry copper ore have been fitted using
the models and reasonable agreement between measured and
predicted values were found.
The measurements of froth stability in two different concentrators
have been correlated to froth recovery, using the approach by
Tsatouhas et al. (2006). Despite the very different froth operating
strategies at the two concentrators, (weak frother in conjunction with
shallow froth (concentrator B) and strong frother in conjunction with
deep froth (concentrator A)), very similar froth recoveries were
obtained, reflecting the similar overall metallurgical performance of
the rougher/scavenger banks in the two plants.
Acknowledgements
The authors would like to acknowledge the financial support of the
sponsors of the AMIRA P541B project and the Australian Research
Council (ARC). Support from site personnel during each campaign is
also gratefully acknowledged.
27
References
(⁎) Alexander, D.J., Franzidis, J-P., Manlapig, E.V., 2003. Froth recovery measurement in
Rf plant scale flotation cells. Miner. Eng. 16, 1197–1203.
[%] Ata, S., Ahmed, N., Jameson, G.J., 2003. A study of bubble coalescence in flotation froths.
Int. J. Miner. Process. 72, 255–266.
82
Barbian, N., Hadler, K., Ventura-Medina, E., Cilliers, J.J., 2005. The froth stability column:
55
linking froth stability and flotation performance. Miner. Eng. 18 (3), 317–324.
30
Dippenaar, A., 1982. The destabilisation of froths by solids. 1. The mechanism of film
16 rupture. Int. J. Miner. Process. 9, 1–14.
Gelinas, S., Finch, J.A., 2005. Colorimetric determination of common industrial frothers.
Miner. Eng. 18, 263–266.
Gorain, B.K., Franzidis, J.-P., Manlapig, E.V., 1996. Studies on impeller type, impeller
speed and air flow rate in an industrial scale flotation cell. Part 3: effect on
superficial gas velocity. Miner. Eng. 9 (6), 639–654.
Gorain, B.K., Napier-Munn, T.J., Franzidis, J-P., Manlapig, E.V., 1998. Studies on impeller
type, impeller speed and air flow rate in an industrial scale flotation cell, Part 5:
validation of K–Sb relationship and effect of froth depth. Miner. Eng. 11, 615.
Hadler, K., Barbian, N., Cilliers, J.J., 2006. The relationship between froth stability and
flotation performance down a bank of cells. Proceedings of the XXIII International
Minerals Processing Congress, Istanbul, Turkey, pp. 516–523.
Huang, C., Su, Y., Hsieh, Y., 2002. Optimization of the head-space solid-phase
microextraction for determination of glycolethers by orthogonal array design.
J. Chromatogr. A 977 (1), 9–16.
Iglesias, E., Anderez, J., Forgiarini, A., Salager, J., 1995. A new method to estimate the
stability of short life-time foams. Colloids Surf. 98, 167.
Johansson, G., Pugh, R.J., 1992. The influence of particle size and hydrophobicity on the
stability of mineralised froths. Int. J. Miner. Process. 34, 1–21.
Laplante, A.R., Toguri, J.M., Smith, H.W., 1983. The effect of airflow rate on the kinetics of
flotation — Part 1: The transfer of material from the slurry to the froth. Int. J. Miner.
Process. 11, 203.
Malysa, E., Malysa, K., Czarnecki, J., 1987. Method of comparison of the frothing and
collecting properties of frothers. Colloids Surf. 23 (1–2), 29–39.
Morar, S.H., Hatfield, S.P., Barbian, N., Bradshaw, D.J., Cilliers, J.J., Triffett, B., 2006. A
comparison of froth stability measurements and their use in the prediction of
concentrate grade. XXIII International Mineral Processing Conference, Istanbul,
Turkey, 3–8 September, 2006.
Moys, M.H., 1984. Residence time distributions and mass transport in the froth phase of
the flotation process. Int. J. Miner. Process. 13, 117–142.
Nesset, J.E., Hernandez-Aguilar, J.R., Acuna, C., Gomez, C.O., Finch, J.A., 2006. Some gas
dispersion characteristics of mechanical flotation machines. Miner. Eng. 19,
807–815.
Runge, K., McMaster, J., Wortley, M., La Rosa, D., Guyot, O., 2007. A correlation between
Visiofroth™ measurements and the performance of a flotation cell. In: Nelson, M.,
Smith, R. (Eds.), Proc. 9th Mill Ops. Conf., Aust. Inst. Min. Metall. Publ., Melbourne,
pp. 79–86.
Savassi, O.N., Alexander, D.J., Johnson, N.W., Manlapig, E.V., Franzidis, J-P., 1997.
Measurement of froth recovery of attached particles in industrial cells. Proceedings
of the Sixth Mill Operators Conference. In: Lauder, D. (Ed.), Aust. Inst Min. Metall.
Publ., Melbourne, pp. 149–155.
Schwarz, S., Grano, S., 2005. Effect of particle hydrophobicity on particle and water
transport across a flotation froth. Colloids Surf., A Physicochem. Eng. Asp. 256,
157–164.
Subrahmanyam, T.V., Forssberg, E., 1988. Froth characteristics and grade–recovery
relationships in the flotation of lead–zinc and copper ores. Miner. Eng. 1 (1), 41–52.
Tsatouhas, G., Grano, S.R., Vera, M., 2006. Case studies on the performance and
characterisation of the froth phase in industrial flotation circuits. Miner. Eng. 19,
774–783.
Ventura-Medina, E., Cilliers, J.J., 2002. A model to describe flotation performance based
on physics of foams and froth image analysis. Int. J. Miner. Process. 67 (1–4), 79–99.
Ventura-Medina, E., Barbian, N., Cilliers, J.J., 2003. Froth stability and flotation
performance. In: Lorenzen, L., Bradshaw, D.J. (Eds.), Proc. XXII Int. Miner. Proc.
Congr., Cape Town, pp. 937–945.
Vera, M.A., Franzidis, J-P., Manlapig, E.V., 1999. Simultaneous determination of
collection zone rate constant and froth zone recovery in a mechanical flotation
environment. Miner. Eng. 12, 1163.
Zanin, M., Gredelj, S., Grano, S.R., 2008. Factors affecting froth stability in mineral flotation
and implications on minerals recovery: a case study. In: Kuyvenhoven, R., Gomez, C.,
Casali, A. (Eds.), Proceedings of Procemin 2008, Santiago (Chile), pp. 197–206.
Zheng, X., Franzidis, J-P., Johnson, N.W., 2006. An evaluation of different models of water
recovery in flotation. Miner. Eng. 19, 871–882.