Minerals Engineering 19 (2006) 774–783

This article is also available online at:

www.elsevier.com/locate/mineng

Case studies on the performance and characterisation of the froth

phase in industrial flotation circuits
G. Tsatouhas c, S.R. Grano a,*
, M. Vera 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
Oil Sands Technology Group—Extraction and Separation, Calgary Research Centre, Shell Canada Limited, 3655-36th St, NW Calgary,
Alberta, Canada 2L 1Y8
c
Huntsman Australia, Performance Products, Gate 2 Newsom Street, Ascot Vale Victoria 3032, Australia

Received 15 July 2005; accepted 9 September 2005

Available online 26 October 2005

Abstract

This paper deals with two separate case studies investigating the froth phase performance and characterisation of two industrial
rougher/scavenger flotation circuits. Froth phase performance was quantified using a mass balance approach to estimate froth zone
recovery. Measured characteristics of the froth phase included frother solution concentration determined by gas chromatography,
and the time taken for an equilibrium froth sample to decay to one-half of its original froth height. The latter measurement is referred
to as the Ôfroth half-lifeÕ and is strongly linked to froth stability. Special methods and techniques developed to preserve frother in solution
and to measure froth half-life are briefly described. The frother type in the first case study was a mixture of straight and branched alco-
hols, whilst the frother type in the second case study was a mixture of alcohols, aldehydes and triethoxybutane. The first case study
focussed on a flotation circuit treating a low grade ore containing only a small fraction of floatable copper sulphide minerals, while
the second case study focussed on a flotation circuit treating a higher grade complex sulphide ore containing significant quantities of
chalcopyrite, galena, sphalerite and pyrite.
It was found that froth zone recovery of valuable mineral generally decreased down-the-bank of the two industrial rougher/scavenger
circuits. Moreover, decreases in froth zone recovery significantly limit the overall cell recovery of valuable mineral achievable from the
plant scavenger cells. However, the decrease in froth zone recovery could not be linked to the removal of frother from the pulp solution
to the concentrate product in the preceding rougher flotation stages. Measurements of residual frother in solution suggested that, approx-
imately, only 5–10% of the added frother was removed into the rougher/scavenger concentrate, with the remainder appearing in the scav-
enger tailings. This finding suggested there was apparently adequate frother in solution in the scavenger stages.
There was, however, a correlation to the froth half-life, with the froth half-life also generally decreasing down-the-bank. A simple,
empirical model, based on the froth half-life and froth residence time of gas, is proposed here to predict froth zone recovery. Further,
it is proposed that the froth stability, as measured by the froth half-life, is strongly linked to the presence of particles in the froth, with
poorly mineralised scavenger froth characterised by a short half-life and, potentially, a low froth zone recovery. The importance of par-
ticles on froth stability was confirmed in separately conducted laboratory experiments. These experiments also demonstrated the wide
variation in froth stability behaviour between different frother types.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: Flotation froths; Flotation frothers; Sulphide ores

*
1. Introduction
Corresponding author.
E-mail addresses: george_tsatouhas@huntsman.com (G. Tsatouhas),
stephen.grano@unisa.edu.au (S.R. Grano), marco.vera@shell.com In flotation, the transport of hydrophobic particles
(M. Vera). attached to bubbles from the collection zone of the cell,

0892-6875/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mineng.2005.09.033
 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783
where particle collection by bubbles occurs, to the concen-
trate launder is of key importance. Transport takes place
via a froth phase in froth flotation. Frother surfactants are
principally added to a flotation pulp so that a transient, min-
eralised froth is formed at the top of the flotation cell, form-
ing the so-called froth phase. A transient froth is a
prerequisite for the successful transport of mineral laden
bubbles from the pulp zone/froth zone interface to the con-
centrate launder. It is desirable that the mineralised froth
breaks down (i.e., the bubbles collapse or coalesce), once
the froth reports to the concentrate launder. A froth which
is overly unstable may cause significant loss of valuable min-
eral from the froth to the pulp phase and, subsequently, to
the cell tailing stream. Conversely, an overly stable froth will
cause problems in pumping, and may adversely impact min-
eral separation in subsequent flotation stages. Our research
here is focussed on controlling froth stability to allow parti-
cle transport to take place, while also ensuring the froth
breaks down once it enters the concentrate launder.
The recovery of particles attached to bubbles across the
froth phase is referred to as froth recovery. Froth recovery
can be defined in several ways. However, it is now accepted
that the recovery of attached particles across the froth zone
is best expressed in terms of the first order rate constants
(Savassi et al., 1997):
k has apparently low stability. In contrast, froth recovery in
Rf ¼ ð1Þ
kc
In Eq. (1), Rf is the fractional froth recovery, k is the over-
all flotation cell first order rate constant, treating the collec-
tion and froth zones as a single phase, and kc is the
collection zone rate constant. It is noted that entrained par-
ticles are included in this particular definition. Entrainment
is the phenomena of the recovery of mineral particles,
which are not attached to air bubbles, by their unselective
recovery with water (Johnson et al., 1974). However, in the
context of this current paper, which principally focuses on
hydrophobic value minerals, the contribution by entrain-
ment is expected to be only minor.
This definition also provides an experimental approach
to the measurement of froth recovery in which the overall
flotation cell rate constant is determined as a function of
froth height, or the froth residence time (FRT) of gas (Lap-
lante et al., 1983; Vera et al., 1999). The froth residence
time of gas is defined as
FD
FRT ¼ ð2Þ
Jg
In Eq. (2), FD is the froth depth (cm) and Jg is the super-
ficial gas velocity (cm/s), which is the gas flow per unit
cross-sectional area of the flotation cell. In this approach,
the overall flotation cell first order rate constant is mea-
sured as a function of froth residence time, with the collec-
tion zone rate constant, kc, determined by extrapolation to
zero froth depth. The ratio of the overall flotation cell first
order rate constant, at any froth depth (FD), to the collec-
tion zone rate constant defines the froth recovery at that
775
froth depth according to Eq. (1). This particular method,
termed the mass balance approach, is used in the current
study to determine froth recovery of the plant cells. It is
recognised that in this approach, the froth recovery at zero
froth height is assumed to be 100%. This may not be the
case due to bubble bursting at the top of the froth, or inef-
ficiencies in the lateral transport of particles in the froth,
even at very shallow froth depths. However, for industrial
cells it is an expedient approach, in the first instance, to
obtain an estimate of froth recovery.
Both the physical scale of the froth (defined here by FRT)
and froth type, exemplified by the froths stability, may con-
trol the froth recovery of attached particles. Froth residence
time and froth height characterise the physical scale of the
froth, whilst the frother type, frother concentration as well
as the particle characteristics in the froth (e.g., particle size,
particle contact angle, particle loading in the froth) may all
control froth stability (Schwarz et al., 2002). A decrease in
the carrying capacity of the froth, and the possibility that
some mineral particles may become detached and be lost
from the froth, has been observed from froth recovery mea-
surements conducted at plant scale where there is continuous
removal of froth in successive flotation cells (Savassi et al.,
1997). The decrease in froth recovery seems particularly
important in the scavenger flotation stages, where the froth
the initial stages of rougher flotation of a continuous separa-
tor is usually higher, and the froths apparently more stable.
This particular aspect of the operation of flotation froths is
the principal focus of this paper.
It is the purpose of this paper to determine if decreases
in froth recovery in a continuous, industrial flotation sepa-
rator may be linked to changes in froth stability, and to
determine reasons for the change in froth stability with flo-
tation cell number. Froth stability is used here to describe
the general phenomena of thin film rupture, bubble coales-
cence and loss of froth volume. In this work, both the effect
of particles contained within the froth, as well as frother
type and concentration on froth stability has been investi-
gated. In recent years there has been strong interest in mod-
elling froth zone performance in laboratory flotation cells,
in scale-up to a continuous flotation process, and in froth
phase assessment in plant cells. Froth phase performance
is better understood here in terms of froth recovery, Rf,
and hence a specific objective of this work is to identify
how physical factors, such as froth height, and chemical
factors, such as frother concentration in solution and froth
solids loading, control froth recovery. The exponential
decay of froth recovery with an increase in froth retention
(residence) time is well known and useful in terms of froth
modelling (Gorain et al., 1998):
Rf ¼ 100  expðbFRTÞ ð3Þ
In Eq. (3), b is a parameter, which is arguably related to the
rate at which bubbles coalesce and burst, or simply a param-
eter, which could depend on the physical and chemical froth
factors. An expansion of the Rf–FRT relationship to account
776 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783
for the influence of froth stability is used here to describe a
semi-empirical model to predict froth phase performance
in plant cells.
2. Experimental
2.1. Sampling and froth recovery measurements
A series of comprehensive sampling surveys were
performed at two different major sulphide mineral concen-
trators located in Australia (Cu circuit—Circuit A) (Cu–
Pb–Zn circuit—Circuit B). In the feed to Circuit A, copper
occurred in bornite and chalcopyrite while there was only
minor pyrite (Table 1). In the feed to Circuit B, copper
occurred in chalcopyrite, lead in galena and zinc in sphal-
erite. The two flotation circuits treat ores with significantly
different head mineralogy (Table 1), which has implications
for the stability of the froth in both plants. It is noted that
copper recovery and rate constant were used to define froth
recovery in Circuit A, whilst lead recovery and rate con-
stant were used to define froth recovery in Circuit B.
In each case, the objective was to assess froth zone per-
formance, with the sampling conducted in Circuit A in
2001 and Circuit B in 2002, respectively. Briefly, plant flota-
tion cells in the rougher and scavenger banks were operated
at different froth depths classified as shallow, intermediate
(baseline) and deep, corresponding to short, intermediate
and long froth residence times, respectively, through Eq.
(2). In some cases, and at specific froth depths, the flotation
cells were also operated at different airflow rates. However,
the flotation rate constant and froth recovery data pre-
sented here are at a fixed airflow rate for simplicity sake.
In addition to the comprehensive sampling surveys, the
metallurgical performance of a single flotation cell in the
lead/copper rougher flotation bank of Circuit B was also
investigated. Three different froth depths, three different
air flow rates and three different frother concentrations
were employed in the flotation bank in these investigations.
The frother concentration in Circuit B was varied from the
baseline plant frother addition rate to high and low addi-
tions, in order to determine the overall effect of frother con-
centration on froth recovery. The flotation rate constant, k
(min1), and froth recovery, Rf, were calculated from the
recovery of copper (Circuit A) and lead (Circuit B) for each
of the cell operating conditions used in the test programme,
assuming perfect mixing in the flotation cell:
R
k¼
t  ð1  RÞ
Table 1
Head mineralogy of the feeds to Circuits A and B
Circuit Mineral
Chalcopyrite Bornite Galena
A 1.2 1.8 – –
B 0.7 – 1.3
Note: Non-sulphide gangue includes silicates and alumino-silicates.
In Eq. (4), R is the fractional recovery of copper or lead,
and t is the effective mean residence time (minutes) of the
pulp in the flotation cell.
The objective of the flotation bank surveys was to char-
acterise froth performance down-the-bank under standard
operating conditions, and to obtain a data set for a single
flotation cell under different operating conditions in Circuit
B only. The data is used to explore relationships between
performance and conditions, i.e., between froth recovery
and froth depth, air flow rate and frother concentration.
The surveys involved collection of samples from all the
major streams down the respective rougher and scavenger
flotation banks. Sample collection was performed over a
2-h period, during which four increments were taken at each
sampling point. The samples collected from each cell during
each survey, included: (a) the feed to the flotation bank (i.e.,
feed to rougher cell 1), (b) the combined concentrate of the
flotation bank, (c) the tailings from the flotation bank, (d)
timed lip concentrates from individual cells in the bank,
and (e) tailings from each cell in the tested bank.
Pulp chemical surveys were also performed in parallel
with the main metallurgical surveys to determine frother
concentrations in solution in major streams, as well as the
distribution (or recovery) of frother down-the-bank. This
information is used to complement the metallurgical assess-
ment. Surveys involved collection of samples from all the
major streams, which was also extended to both the rougher
and cleaner banks of both circuits. Frother in solution in
aqueous samples was measured using gas chromatography
with flame ionisation detection. Frother balances were
made using all the measured chemical assay data, % solids,
measured frother assays and, taking as the reference, the
feed to the flotation bank. Flowrates of the various slurry
streams were obtained by computation. Individual adjust-
ment of the standard deviations of frother assays was neces-
sary to aid the balance and to check the quality of the data.
Preliminary work showed that the relevant frothers were
stable in solution after sampling, provided the samples were
stored in the absence of oxygen and at temperatures less
than 3 °C. This study used three different frothers—plant
frother A (Circuit A), a mixture of low molecular weight
alcohols, aldehydes and esters; plant frother B (Circuit B),
a triethoxybutane and alcohol mixture, and MIBC (methyl
isobutyl carbinol). All frothers were used as received.
2.2. Froth stability methods
ð4Þ Froth stability generally refers to the foam decay pro-
cess, during which time there is no further generation of
Sphalerite Pyrite Non-sulphide gangue
0.9 96.1
21.7 29.4 46.9
 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783 777

gas bubbles. Froth decay is indicated by some visible
occurrence such as froth volume variation, bubble collapse,
or change in the bubble size. In order to quantify the global
change of the froth, froth volume variation is recorded as a
function of time. The faster the change, the less stable the
foam. The simplest method of estimating the stability of
a foam is to measure the change in volume (or height,
Hf), for constant cross-sectional area with time, t. The ini-
tial time is taken when there is negligible further increase in
froth volume, and the gas flow is then discontinued. The
froth formed was then allowed to decay and the change
in froth height noted as a function of decay time. A differ-
ential equation was then fitted to describe the rate of foam
collapse. According to Iglesias et al. (1995):
dH f k
¼ ð5Þ
dt t 40 ppm of the plant frother concentration, as well as with
In Eq. (5), k is an empirical constant. Iglesias et al. (1995)
noted that plotting the froth height against log (t) yielded a
series of straight lines for a range of surfactant types and
foam stabilities, justifying in part Eq. (5).
 Furthermore,

Iglesias et al. (1995) plotted HH of against log t
, which nec-
t1=2
essarily caused all the straight-line relationships to con-
verge to HH of ¼ 0:5 at t = t1/2. In the current study, a
similar approach was adopted using Eq. (6):
 
Hf t tion rates and during the survey period. The overall cell rate
¼ a log ð6Þ
Ho t1=2
In Eq. (6) a is dimensionless empirical constant. The half-
life is based on the height value and the reference is taken
as the time, t1/2, for the foam to decrease to one-half of its
original height, Ho.
A series of experiments was designed in both plant and
laboratory flotation cells. Plant tests were performed
in situ of different rougher cells (cells 1 and 4) of the lead/cop-
per rougher flotation bank in Circuit B to highlight differ-
ences in froth stability, due to differences in particle type
and solids loading in the froth. A 500-mL cylindrical column
(1.5 m long, 80 mm diameter) was immersed in each cell
between the froth surface and just below the pulp/froth inter-
face and the froth collected over 2 min using plant air condi-
tions. The equilibrium foam height was recorded and the gas
flow discontinued by means of a sliding partition at the base
of the column. The froth was then allowed to decay and the
change in froth height noted as a function of time.
Froth decay experiments were also carried out in a grad-
uated multi-holed sparger column (1.5 m long, 2.5 cm i.d.)
(Schwarz et al., 2002), using plant flotation feed slurry for
Circuit B only. The slurry sample was collected prior to
reagent addition in the plant. The particle surface chemis-
try was controlled using reagent (collector and frother)
additions in the laboratory with the slurry entering the col-
umn diluted to 10% solids weight/slurry weight. The exper-
imental approach described previously was used to
investigate the decay process at controlled air (400 mL/
min) and feed flow rates, which were adjusted to optimise
the froth residence time. Dynamic measurements were
made, both with and without particles, at 8, 20 and
MIBC used as the reference surfactant.
3. Results
3.1. Plant froth zone recovery
3.1.1. Circuit A
Table 2 presents a summary of parameters measured in
individual cells of Circuit A, under the same reagent addi-
constant decreased down-the-bank due to both decreases in
the collection zone rate constant and froth zone recovery.
While our attention here is principally on froth zone recov-
ery, it is worthwhile discussing the significance of the collec-
tion zone rate constant at this point. The collection zone
rate constant decreases down-the-bank due to the removal
of fast floating components in the initial stages of flotation,
the rate constant being an undistributed rate constant of all
floating components. In the scavenger stages only the slow
floating components remain, being the less hydrophobic
value mineral such as coarse composites. This is reflected
in the decreasing copper assay of the froth down-the-bank
(Table 2). Both the decreasing grade (hydrophobicity) of
the particles entering the froth, as well as the decreasing sol-
ids loading in the froth (Table 2), have consequences on the
stability of the froth as discussed further below.

Table 2
Cell characterisation summary and froth parameters measured in Circuit A
Plant cell Superficial gas Rate constant at Froth recovery at Froth (gas) residence Solids loading Copper grade
velocity (cm/s) FD = 13 cm (min1) FD = 13 cm (%) time at FD = 13 cm (s) in froth (%) in froth (%)
Roughers 1 0.99 0.39 62 13.1 48.2 42.5
2 1.19 0.44 54 10.9 49.4 39.7
3 0.84 0.18 53 15.5 48.0 31.1
4 0.96 0.20 52 14.1 40.0 18.8
Scavengers 1 0.64 0.06 29 20.3 26.8 24.1
2 0.72 0.12 25 18.8 26.5 7.4
3 0.56 0.09 28 24.1 27.9 3.3
4 0.68 0.06 23 21.3 23.0 2.6
Note: Circuit A consisted of four rougher cells, followed by four scavenger cells. Rate constant and froth recovery at 13 cm froth depth are only shown.
778 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783

The superficial gas velocity values are also shown for ref- 80
erence in Table 2, and it is expected that for higher air rates
in the cell there would be higher froth zone recovery for a

Froth Recovery (%)

60 Scavenger Cells
given froth stability. Higher air rates will reduce the resi- Rf ~ 20-35%
dence time of the gas in the froth before reaching the froth ~27% solids in froth
surface. Note that changes in superficial gas velocity from 40 Rougher Cells
~8 % Cu in froth
cell to cell may be accounted for by considering froth resi- Rf ~ 50-65%
dence time, FRT. The froth residence time values presented ~48% solids in froth
20 ~42% Cu in froth
in Table 2 refer to values calculated at a fixed froth depth,
however, during the campaign in Circuit A, the froth depth (a)
0
was varied under different survey conditions in order to
assess the effect of froth depth on froth recovery. Ro1 Ro2 Ro3 Ro4 Scv 1 Scv 2 Scv 3 Scv 4 Cell
The copper flotation rate is plotted as a function of froth 100
depth for rougher cell 1 and scavenger cell 3 in Fig. 1, from Rougher Cells
which the value of the collection zone rate constant, kc, 80 Scavenger Cells

Froth Recovery (%)

may be determined by extrapolation to zero froth depth.
This allows calculation of froth recovery, Rf, at the exper- Ro1
60 Ro2
imental conditions and at any froth depth, using the defini-
tion of Rf as the ratio of overall rate constant, k, and the Ro4 Ro3
40
collection zone constant, kc. The froth recovery values Scv 1
shown in Table 2 and Fig. 2a are at a froth depth of Scv 3
20 Scv 2 Scv 4
13 cm. At the head of the rougher bank (in cells 1 and 2),
the froth recovery is approximately 50–65%, decreasing (b)
0
down-the-bank to the last scavenger cells where the froth
0 5 10 15 20 25 30 35
recovery is estimated to be only 23%. This trend confirms
other measurements of froth recovery in industrial flota- Froth (Gas) Residence Time (s)
tion cells (Savassi et al., 1997). Fig. 2. Froth zone characterisation for the rougher (Ro cells 1–4) and
The decrease in froth recovery may be linked to the par- scavenger (Scv cells 1–4) cells in Circuit A: (a) froth recovery of copper as
ticle type and concentration that reports to the froth phase. a function of cell number, and (b) froth recovery of copper as a function of
froth residence time, FRT. The continuous lines shown in b are calculated
There is a decrease in solids loading in the froth from
using Eq. (7) and the parameters shown in Table 4.
48% w/w in the rougher cells to 27% w/w in the scaven-
ger cells, i.e., particle floatability decreases down-the-bank
(Table 2). There is also a decrease in the copper composi-
tion of the particles (i.e., the proportion of hydrophobic 70
Predicted Froth Recovery (%)

particles) in the froth with increasing cell number (Table 60

2). To further illustrate the importance of the particles that
50 Rougher
report to the froth phase on froth recovery, the results of a
simple, empirically based, regression is shown in Fig. 3. 40 Cells
Scavenger
Here a predicted froth recovery is calculated based on the Cells
30
% of solids in the slurry in the concentrate, as well as the
20
10 Predicted Froth Recovery % =
1.2 %Solids in Conc. + 0.2*%Cu of Solids in Conc.
Flotation Rate Constant (min-1)

0
Rougher Cell 1 0 10 220 300 40 50
5 60 70
1.0
Scavenger Cell 3 Measured Froth Recovery (%)
0.8
kc = 0.65 Fig. 3. Predicted froth recovery of copper as a function of measured froth
0.6 recovery for cells in Circuit A. The predicted froth recovery is calculated
using the regression equation shown.
0.4 kc = 0.32

0.2 % of copper in the concentrate solids. Using the same

regression coefficients for each cell number shows remark-
0.0 able agreement between the measured and predicted froth
0 20 40 60 80 100 120 140 160 180 recovery, calculated using the regression equation shown
Froth Depth (mm) in Fig. 3. Evidently, the % of solids in the slurry in the con-
Fig. 1. Copper flotation rate constant as a function of froth depth for centrate is an apparently more important determinant on
rougher cell 1 and scavenger cell 3 at a fixed aeration rate in Circuit A. froth recovery than % copper in the concentrate.
 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783 779

Froth recovery (Rf) is plotted as a function of froth res- 10

100
idence time, calculated according to Eq. (2), in Fig. 2b. The Decrease FD 7.5 ppm Frother B
nature of the froth changes down-the-bank, i.e., the com- Survey 7 10 ppm
pp mFrother
F rother B
B
80
position and nature of the particles, as well as the froth 5.4 ppm
ppm Frother
F rot he
rBB

Froth Recovery (%)

structure. Visual observation confirmed that the bubbles
in the froth of the scavenger cells were larger and bursted
more frequently in comparison to the bubbles in the froth
of the rougher cells. Fig. 2b suggests an exponential
decrease in froth recovery with an increase in froth reten-
tion time. A longer froth residence time leads to a higher
probability of particle drop-back within the froth, resulting
in low froth recovery. This is the case for the weakly stable
froths encountered in the scavenger cells, compared to the
rougher cells in Circuit A. In this case, bubble coalescence
causes the particles to be detached from the bubbles. The
data for the rougher and scavenger cells apparently fall
on different fitted lines of Eq. (3), with different values of
b apparently applicable. Most likely, there is a significant
difference in froth stability between the two stages, and
the large decrease in froth recovery in the scavenger cells
is due to the weakly structured and, hence, very different
froth compared with the rougher cells. It is also noted that
cell 1 of the scavengers shows a marked decrease in froth
recovery that is most strongly correlated to the decreased
solids loading in the froth, rather than to a decrease in cop-
per composition of the particles (Table 2). The significance
of solids loading in the froth on froth stability is discussed
further below.
3.1.2. Circuit B
In a similar approach, the froth recovery–froth residence
time relationship was explored in the lead/copper rougher
flotation bank of Circuit B. In this circuit, the focus was
on assessing the effect of frother B concentration and froth
depth on lead recovery in a set of surveys in a single cell
(cell 3) of the first rougher bank. Table 3 gives a summary
of the measured parameters from surveys conducted in Cir-
cuit B. Values of froth recovery of lead were found to vary
from 38% at low frother concentration (5.4 ppm in the
solution phase) and intermediate froth depth (15 cm), to
75% at intermediate frother concentration (7.5 ppm) and
shallow froth depth (10 cm). Fig. 4 shows the froth recov-
ery of lead plotted against froth residence time (FRT),
under the different survey conditions. While there is a lack
of data points for tests conducted at different frother con-
centrations, it is clear that froth recovery is not only a func-
60 Survey 4
Survey 3
40
Survey 6
Increase FD
Survey 5
20
0
0 10 20 30 40 50 60
Froth (Gas) Retention Time (sec)
Fig. 4. Froth zone recovery of lead as a function of froth residence time,
FRT, and frother B concentration, measured in cell 3 of the rougher bank
in Circuit B. The continuous lines are calculated using Eq. (7) and the
parameters shown in Table 4.
tion of froth residence time but also froth type, which may
be controlled by frother addition.
Changes in froth depth, but without an accompanying
change in frother concentration (in surveys 3, 6, and 7),
gave changes in froth recovery that were explicable in terms
of froth residence time changes only (Fig. 4). This conclu-
sion could be drawn because these data points fit a Rf–FRT
relationship defined by Eq. (3), using a single value of b.
The general trend for the change in froth depth tests fol-
lows the expected exponential decay of Rf with an increase
in froth residence time. The results demonstrate that other
factors, such as froth stability, altered through changes in
frother concentration, also controls froth recovery in addi-
tion to froth residence time. In contrast, changes in frother
concentration in surveys 4 and 5 gave changes in froth
recovery that could not be accounted for by simple changes
in froth residence time (Fig. 4). In these cases, different val-
ues of the parameter b where required to fit the experimen-
tal data points indicating that the stability of the froth had
changed (Fig. 4). Unfortunately, the stability of the plant
froth was not measured in parallel with the changes in
frother concentration in these particular tests. However,
it was measured in separate tests on the rougher flotation
bank in Circuit B, enabling a correlation to be made
between froth stability, as measured by the froth half-life,
to froth recovery. This correlation is examined further
below.

Table 3
Cell characterisation summary and froth parameters measured in Circuit B
Survey no. Cell feed Pb Frother Superficial Froth Rate Solids loading Lead grade Overall cell Froth Froth (gas)
grade (%) concentration gas velocity depth constant in froth (%) in froth (%) recovery (%) recovery (%) residence
(ppm) (cm/s) (cm) (min1) time (s)
3 0.60 7.5 0.61 18 0.14 45.6 15.7 22.7 60 29.5
4 0.64 10 0.63 25 0.23 46.0 15.5 31.3 54 39.7
5 0.67 5.4 0.63 15 0.11 46.1 17.3 27.3 38 23.8
6 0.64 7.5 0.63 25 0.12 37.4 17.3 23.0 49 39.7
7 0.74 7.5 0.63 10 0.17 54.5 21.5 22.9 75 15.9
780 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783
3.2. Plant froth stability measurements
The preceding discussion implies that both chemical
(e.g., frother type and concentration) and physical condi-
tions (froth depth, aeration rate, etc.) may be manipulated
in flotation systems to achieve a desired froth zone recov-
ery. However, the particle characteristics and their influ-
ence on drainage, stability and frothing phenomena, need
also be considered as they play a significant role in control-
ling froth zone recovery.
The stability of froths generated in cells 1 and 4 of the
rougher bank in Circuit B is shown in Fig. 5. The froth
in cell 1 in Circuit B was heavily mineralised (Table 3),
and the froth formed in the rougher stages was relatively
stable. Froth phase stability behaviour is quantified here
in terms of the froth half-life parameter, t1/2, estimated
from the froth decay profiles. The froths generated in the
two cells are significantly different, both in terms of their
different froth half-lives and their different solids loading.
While the froth formed in cell 1 is significantly more stable
than the froth in cell 4, both froths are relatively stable
compared to the more unstable and dynamic froths pro-
duced in the rougher/scavenger cells of Circuit A. Differ-
ences in froth phase behaviour are likely to be influenced
by the composition (i.e., hydrophobicity) and solids load-
ing of particles in the froth, with Circuit A froth having
much less particles contained within it due to its lower head
grade of valuable and gangue sulphide minerals (Table 1).
Reasons for low froth recovery, particularly in the scaven-
ger stages of Circuit A, are discussed further below in terms
of the distribution of frother amongst the products of
flotation.
3.3. Plant frother in solution measurements
Frother type and concentration are likely to control the
foaming properties in the presence of particles through a
35
Rougher Cell 4
30
Rougher Cell 1
Froth Height (cm)
25
20
15
10
5
0 presence of frother in solution in the scavenger stages.
0 10 20 30 40 50 60 70 80
Time (sec)
Fig. 5. In situ decay measurements of plant froths generated at plant
reagent and gas conditions in the Rougher 1 bank in Circuit B. Rougher
cell 1 froth contained 66.8% solids, 48.7% galena, t1/2  130 s; Rougher
cell 4 froth contained 45.6% solids, 18% galena, t1/2 = 35 s.
number of mechanisms. In cases of low froth recovery
and stability, it had not, to this point, been established
whether there was sufficient frother in solution to create a
stable froth and effectively transport particles from the col-
lection zone to the concentrate product. A pulp chemical
survey was performed in both Circuits A and B under nor-
mal operating conditions (i.e., at normal plant frother con-
centration), to determine the distribution of frother across
the plant cells. Chemical sampling of plant cells was made
from key sample points, with feed, concentrate and tailings
samples collected from each major unit stream. The deport-
ment of frother to the flotation products was subsequently
determined using the mass balanced flows of water.
A summary of the mass balance of residual frother in
solution as a percentage of the frother recovered in the flo-
tation products of each unit is shown in Fig. 6. In each cir-
cuit, as frother is consumed down-the-bank, the majority
of frother presented to each unit is found to deport to
the respective tailing product. In Circuit A, for example,
only 4% of frother A added to the plant rougher feed
reported to the combined rougher and scavenger concen-
trate stream (froth product), the remainder reporting to
the tailing stream. This behaviour in frother deportment
was also observed for laboratory scale batch tests (not
reported here) treating Circuit A rougher feed. A key find-
ing was that frother recovery was linked strongly to the
recovery of water, and clearly the largest amount of frother
predominantly deports with the largest water flow, i.e., the
tailings product. An important practical finding was that
the frother in Circuit B largely decomposed in the water
recovery circuit designed to recycle water from the tailings
stream to the process (Fig. 6b).
The percentage of water that reports to the concentrate
may be calculated from Fig. 6(a) in the case of Circuit A.
The water recovery is 3.1% in the roughers and 1.4% in
the scavengers, compared with 2.6% and 1.5%, respectively,
for the frother itself. It could be expected that frother
recovery may be similar to water recovery for a froth in
which the main mechanism of frother recovery to the con-
centrate is in the bulk liquid phase. This appears to be the
case for the scavengers, suggesting that little of the frother
in the scavenger concentrate is reporting as frother
adsorbed at the gas/liquid interface. Further experimental
work would be required to verify this statement unequivo-
cally. The decrease in froth zone recovery could not be
linked to the removal of frother from the pulp solution
to the concentrate product. This is particularly pertinent
to the situation that apparently exists in the scavenger
froths of Circuit A. In this case, it appears that froth stabil-
ity is poor and froth recovery low in spite of the continuing
90 100
3.4. Effect of particles and frother concentration on
froth stability at laboratory scale
Fig. 7 shows the effect of frother concentration, frother
type (MIBC and frother B) and hydrophobic particles (i.e.,
 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783 781

573 tph
h
Frother A 40 % w/w
3.3 ppm
Feed Tailing
97.4% Fr A
Rougher Scavenger
590 tph Key
40% w/w
40 18.1 tph 8.4 tph
Water flow (tph)
3.25 ppm 33% w/w % Solids w/w %)
(% 22% w/w
2.8 ppm Frother Concentration (ppm) 3.4 ppm
100% Fr A Frother Recovery (%)
2..6% Fr A 1.5% Fr A
Concentrate Concentrate
(a) Circuit A

Recycled Water 2 ppm Fr B

Frother B
7..1 ppm 6.7 ppm
Feed 97.1% Fr B 98..8 %Fr B
Rougher 1 Rougher 2
Tailing
7.. 6 ppm
100%Fr B 9.. 0 ppm
Concentrate 1.2% B
1.2%Fr Thickener
8.0 ppm
Fr B Cleaners
Key 4.4 ppm
7.9 ppm Fr B (b) Circuit B
Frother Concentration (ppm)
Frother Recovery (%) Fr B
Final Concentrate
Fig. 6. Mass balance of residual frother in solution, shown as the percent recovery of frother in the flotation products from down-the-bank chemical
surveys: (a) Circuit A, frother A; (b) Circuit B, frother B. Note: Frother addition points, measured frother concentrations, water flows and frother
recoveries (distribution) are also shown for reference.

45 i.e., greater t1/2 values. Hydrophobic particles appear to

40 give added stability to the froth, especially in the case of
35 MIBC, where the increase in concentration from 10 to
40 ppm shows a greater change in froth stability in the
30
Half Life (sec)

presence of particles than that observed with frother B.

25 Frother B no particles MIBC gives lower t1/2 values and, hence, weaker froths.
20 Frother B with particles These tests have highlighted important differences between
MIBC no particles frother types and particle effects on froth stability, and how
15
MIBC with particles the froth half-life may be used to represent froth stability.
10
This has implications for its significance and usefulness in
5 froth phase modelling, which is explored further below.
0
0 10 20 30 40 50 60 70
3.5. Froth recovery and froth residence time correlation
Frother Concentration (ppm) at plant scale
Fig. 7. Froth half-life (t1/2), as a function of frother concentration
determined from froth decay measurements in a laboratory sparger The exponential decay of froth recovery with an increase
column. Notes: Froths were generated in the absence and presence of in froth retention (residence) time is well known and useful
particles (rougher feed—Circuit B) using frother B (plant frother) and in terms of froth modelling (Gorain et al., 1998). Based on
MIBC.
the data shown in Figs. 2b and 4, it is reasonable to sup-
pose that the froth zone recovery of particles that arrive
treating Circuit B rougher feed) on froth stability, as at the pulp–froth interface may be described in terms of
defined by the froth half-life in a laboratory sparger col- froth residence time by a simple exponential function such
umn. The surfactant type clearly gives differences in froth as Eq. (3). An alternative interpretation to the b parameter
structure, with weaker frothers such as MIBC requiring in Eq. (3) is proposed here, which attempts to represent the
higher concentrations to give moderately stable froths, interaction between particle transport in the froth transfer
782 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783
Table 4
Froth parameters useful for describing froth zone performance measured during surveys in Circuits A and B
Plant conditions Rf (range) (%) FRT (range) (s)
Circuit A—rougher cells
3.5 ppm frother A 65–50 10–30
Circuit A—scavenger cells
3.5 ppm frother A 30–15 30–50
Circuit B—rougher cells
5.5 ppm frother B 40–35 25
7.5 ppm frother B 75–50 15–30
10 ppm frother B 50 40
and froth stability in the Rf–FRT relationship through con-
sideration of the froth half-life.
  
FRT
Rf ¼ 100  exp  ð7Þ
t1=2
By taking into consideration differences in froth stability
it is possible that different frother types and/or concentra-
tions may show similar dependencies of froth recovery on
the ratio of the froth residence time to froth half-life. For
each survey condition, the exponential model according
to Eq. (7) was fitted to the data points in Fig. 2b (Circuit
A) and Fig. 4 (Circuit B), respectively. Table 4 shows the
froth half-life values that were used to fit lines using Eq.
(7) to the plant data shown in Figs. 2b and 4, respectively.
Table 4 also summarises useful parameters for describing
froth zone performance in the flotation systems. In partic-
ular, it shows that the frother predominantly reports to the
tailing stream over the range of frother concentrations
examined in Circuit B, in agreement with previous com-
ments made in relation to Fig. 6. It should be noted that
Eq. (7) is an empirical relationship only and has no direct
physical meaning. It simply states that the decrease in froth
recovery with froth residence time may also be related to
the stability of the froth. For a more stable froth, the froth
recovery decreases to a lesser extent for the same froth res-
idence time.
It is apparent from the different values of froth half-life
in Table 4 that the relationship, according to Eq. (7), is
strongly dependent on frother concentration and feed com-
position. While the data points show scatter, the interac-
tion of froth recovery with froth stability and froth
retention time appear to be useful for interpreting froth
performance in flotation at an industrial scale. Froth stabil-
ity appears to serve as a good indication of the significance
of the sub-processes occurring within the froth. To ensure a
froth zone recovery greater than 70%, the FRTt1=2
ratio should
be <0.3.
4. Conclusions
This work has shown that the froth phase has a pro-
nounced influence on the overall flotation rate in industrial
flotation circuits. The overall cell rate constant can be
Froth half-life t1/2 (s) Recovery of frother
to tailing stream (%)
35 97.4
15 95.9
25 96.4
54 95.9
65 96.5
reduced to less than 25% of the collection zone rate con-
stant due to inefficiencies of the froth phase. This is partic-
ularly pronounced for scavenger cell froths in which the
low stability of the froth apparently limits froth recovery
and the overall cell rate constant. Because of the complex-
ity and interactive nature of processes occurring within the
froth phase, an important first step towards optimising the
froth phase recovery depends upon determining practically
useful froth parameters, such as the froth residence time
and froth stability, which may then be correlated with froth
recovery.
Two separate industrial case studies were targeted to
investigate froth phase performance and quantify the mea-
sured characteristics of the froth phase. Froth phase per-
formance was defined by the froth zone recovery and the
measured characteristics of the froth phase including
frother concentration and froth stability. Froth zone recov-
ery was found to generally decrease down-the-bank in Cir-
cuit A, with decreases in froth zone recovery limiting the
recovery achievable in the plant scavenger cells. The
decrease in froth zone recovery could not be linked to the
removal of frother from the pulp solution to the concen-
trate product. However, there was a correlation of froth
recovery to froth stability, as measured by the froth half-
life. This parameter is strongly linked to the presence of
particles in the froth, with a wide variation in froth stability
between cells and feed types, which may help to explain the
low froth recovery of poorly mineralised scavenger froths.
These particular froths are characterised by short half-lives
and low froth recovery. A simple exponential model, based
on the froth half-life and froth residence time of gas, is pro-
posed to predict froth zone recovery, however its signifi-
cance or usefulness has not yet been completely defined.
Further research is continuing to determine if observa-
tions regarding the relationship between froth recovery
and froth stability may be applied generically to all plants.
It is suspected that froth instability contributes to low
recovery in scavenger stages generally, particularly for
feeds containing low quantities of the mineral to be floated.
Methods to counter froth instability are also being
researched. These aspects are being studied in the new
AMIRA International project, P541B ‘‘Optimising Froth
Zone Performance in Mineral Flotation’’.
 G. Tsatouhas et al. / Minerals Engineering 19 (2006) 774–783
Acknowledgements
The authors wish to gratefully acknowledge the spon-
sors of the AMIRA International project P541A ‘‘Particle
and Frother Interactions in Flotation Froths’’ including
Anglo Platinum, Cytec, Newmont (Golden Grove Opera-
tions), Ok Tedi Mining, Rio Tinto, Teck Cominco and
the Australian Research Council. The authors also wish
to thank the P9 project ‘‘Mineral Processing’’ for the froth
recovery measurements in Circuit A. This paper was first
published in Proceedings Centenary of Flotation Sympo-
sium, Brisbane, Australia, 5–9 June 2005, by The Austral-
asian Institute of Mining and Metallurgy.
References
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
783
industrial scale flotation cell, Part 5: Validation of k–Sb relationship
and effect of froth depth. Miner. Eng. 11, 615.
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.
Johnson, N.W., McKee, D.J., Lynch, A.J., 1974. Flotation rates of non-
sulphide minerals in chalcopyrite flotation processes. Trans. AIME 256.
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.
Savassi, O.N., Alexander, D.J., Johnson, N.W., Manlapig, E.V., Franzi-
dis, J.-P., 1997. Measurement of froth recovery of attached particles in
industrial cells. In: Lauder, D. (Ed.), Proceedings of the Sixth Mill
Operators Conference. Aust. Inst. Min. Metall. Publ., Melbourne, pp.
149–155.
Schwarz, S., Grano, S., Fornasiero, D., Ralston, J., 2002. Froth behaviour
in the presence and absence of model quartz particles. In: Proceedings
of WCPT4—World Congress on Particle Technology, Sydney, Paper
290.
Vera, M.A., Franzidis, J.-P., Manlapig, E.V., 1999. Simultaneous deter-
mination of collection zone rate constant and froth zone recovery in a
mechanical flotation environment. Miner. Eng. 12, 1163.