Vous êtes sur la page 1sur 6

Minerals Engineering 92 (2016) 242247

Contents lists available at ScienceDirect

Minerals Engineering
journal homepage: www.elsevier.com/locate/mineng

Short communication

A new method for flotation rate characterization using top-of-froth


grades and the froth discharge velocity
J. Yianatos , I. Panire, L. Vinnett
Automation and Supervision Centre for Mining Industry, CASIM, Department of Chemical and Environmental Eng., Federico Santa Maria Technical University, Chile

a r t i c l e

i n f o

Article history:
Received 9 October 2015
Revised 6 February 2016
Accepted 24 March 2016

Keywords:
Flotation rate characterization
Top-of-froth grade
Froth discharge velocity

a b s t r a c t
A new methodology is proposed for flotation characterization in industrial operations. The approach considers the mineral recovery to be proportional to both the top-of-froth (TOF) grade and the froth discharge velocity down a bank of cells. The procedure allows for the identification of the fractional
recovery profile from the discharge velocities and the TOF grades. In addition, if the total recovery of
the bank is available, the cell recoveries can be estimated by scaling the fractional recoveries. For this purpose, a single parameter was used to scale the recoveries for each sampling survey in order to obtain the
kinetic response along the flotation banks. Industrial tests were performed in two rougher banks; one
bank consisted of six 250 m3 self-aerated cells in a 1-1-1-1-1-1 arrangement, and the other bank consisted of nine 130 m3 self-aerated cells in a 1-2-2-2-2 arrangement. The results showed good agreement
with the recovery profiles obtained from the cell-by-cell mass balances along two industrial flotation
banks.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
The metallurgical characterization of a flotation bank is typically performed using the feed, the concentrate and the tail grades.
All of the key streams are sampled, and repeats are conducted to
improve the data reliability (Di Sandro et al., 2013). In addition,
data reconciliation is performed if an excess of information is available (Wills and Napier-Munn, 2006). From the mass balance
results, kinetic models can be fitted to the mineral recovery data
to evaluate and to compare flotation banks. The model analysis
allows for the evaluation of the flotation process and detection of
opportunities for improvement in the design, operation and control of the circuit (Yianatos et al., 2006).
The sampling of the different streams is subject to access limitations, sampling constraints (e.g., circuit layout, personnel) and
measurement uncertainties (Di Sandro et al., 2013). Therefore,
the mineral recovery estimations are sensitive to errors in both
the sampling procedures and the grade analyses. For example, feed
and tail grades in rougher circuits have low values and are susceptible to measurement errors, which may lead to biased results in
the mineral recovery.
Despite the usefulness of metallurgical evaluations, sampling
surveys are not frequently conducted due to their high preparation
requirements, time of analysis, and high costs (Wills and
Corresponding author at: P.O. Box 110-V, Valparaso, Chile.
E-mail address: juan.yianatos@usm.cl (J. Yianatos).
http://dx.doi.org/10.1016/j.mineng.2016.03.026
0892-6875/ 2016 Elsevier Ltd. All rights reserved.

Napier-Munn, 2006). Yianatos and Henrquez (2006) proposed a


methodology for metallurgical characterization that consists of
two mass balances: one around the first cell and the other around
the overall flotation bank. In this approach, only five streams must
be sampled to characterize the flotation kinetic of an arrangement
of cells in series.
The relationship between the mineral recovery and the air
recovery has been studied in the literature based on vision systems
in industrial flotation machines. Air recovery has been found to be
a suitable estimate of the mineral recovery in rougher flotation
banks (Smith et al., 2010; Hadler and Cilliers, 2009; Neethling
and Cilliers, 2008). Air recovery was estimated from the inlet air
flowrate to the froth, the froth height over the lip, the weir lip
length and the froth discharge velocity. The latter was obtained
using visual techniques in the froth surface image. The potential
to optimize the metallurgical results by using changes in the aeration distribution along the flotation bank was observed. Hadler
et al. (2012) related the air recovery to the froth depth and
observed a peak in the air recovery as the froth depth increased.
Some authors have used the bubble load measurement to study
the flow of particles collected by true flotation that enters the froth
zone. Yianatos et al. (2008) and Falutsu and Dobby (1992) developed devices to sample collected minerals in flotation machines.
Yianatos et al. (2008) used the bubble load measurement and the
total mass balance in an industrial flotation cell to estimate the
froth and collection zone recoveries. More recently, Yianatos
et al. (2014) studied the collection of valuable minerals in rougher

243

J. Yianatos et al. / Minerals Engineering 92 (2016) 242247

flotation circuits using the top of froth (TOF) as an alternative


approach to the bubble load measurement. This methodology considered that, for non-selective froths, the TOF material retains the
characteristics (grade and particle size) of the mineral entering
the froth via true flotation. Compared with the bubble load, the
TOF measurement allows a complete flotation bank to be sampled
with sufficient mass for size analysis and chemical assays. Therefore, the evolution of the collected mineral along a flotation bank
(non-selective froths) can be determined. This information motivates the development of a procedure to estimate the kinetic
response using TOF measurements.
In this short communication, a new methodology to estimate
the fractional recovery profiles along rougher flotation banks is
presented. Data on the TOF grade, the froth discharge velocity
and the total recovery of the bank are required to obtain the recovery profiles. For this purpose, the TOF grade and the froth discharge
velocity measurements were performed at industrial scale. The
fresh feed, the final concentrate and the final tail were also measured to obtain the total recovery of the flotation bank. This total
recovery was used to scale the fractional recoveries and for estimating the kinetic response along the flotation banks.

which was located at the beginning of the rougher flotation bank,


using a slurry sampler. In bank A, the PR cell tailing was sampled
using a Grindex submergible pump located inside the cell near
the bottom. The use of this type of submergible pumps was evaluated in previous work reported by Concha (2009), where simultaneous samples of industrial cell tailings were collected using a
submergible pump and a manual in depth sampler. A good agreement between both results was obtained. The tailing samples of
the remaining cells of bank A were collected using auxiliary gate
valves. These auxiliary valves are located in an emergency exit tube
connected to the tailing pipelines of each cell. Only the first cell
and the overall bank were sampled in bank B. Cell-by-cell and
short-cut methods were used to characterize both banks
(Yianatos and Henrquez, 2006). The method used in each bank is
shown in Table 1. The samples were assayed for Cu to estimate
the cell-by-cell recovery. The grade data were reconciled to satisfy
the total and component mass balances in each cell of the flotation
banks. Table 1 also summarizes the feed characteristics, the range
of the operating conditions and the metallurgical results around
the rougher flotation circuits A and B. The Cu recoveries were
approximately 9092% for bank A and 84% for bank B.

2. Plant description and experimental procedure

2.2. Top-of-froth measurements

2.1. Plant description

The top-of-froth measurements (TOF) were conducted along the


rougher banks in parallel with metallurgical samplings. Each measurement corresponded to a sample of approximately 500 mL from
the froth surface of each cell. Fig. 2a shows the TOF sampler with
the typical sampling location (Fig. 2b). The sampling point was
located between the froth crowder and the peripheral concentrate
launder. This location minimizes contamination by entrainment
and avoids stagnant zones (Yianatos et al., 2014). Using the TOF
measurements, the characteristics of the mineral (grade and particle size) that overcame the drainage and coalescence processes
were determined.
Repeats of the TOF measurements in cells PR, C1, C2 and C6
were conducted in the sampling campaign 1. From these datasets,
the estimate for the standard deviation of the TOF grades was 1.3%,
which shows an adequate repeatability and reproducibility of the
measurement procedure.

Experimental tests were performed in an industrial copper


flotation plant fed by approximately 8000 tph from SAG grinding.
The tests focused on two of the eight banks of the rougher circuit:
banks A and B. As shown in Fig. 1a, bank A consists of six 250 m3
self-aerated cells that are arranged in a 1-1-1-1-1-1 configuration.
This circuit includes a cell at the beginning of the bank that operates as a pre-rougher (PR). The concentrate from the PR cell is
mixed with the concentrate from cell 1 (C1), and the tail from
the PR cell feeds cell 1. The rougher circuit B consists of nine
130 m3 cells arranged in a 1-2-2-2-2 configuration, as shown in
Fig. 1b. The cell pairs in the rougher banks B (i.e., 23, 45, 67,
and 89) are connected by an open section (communicating vessels). Level control is implemented at the end of each pair of cells.
In each flotation cell shown in Fig. 1 (both banks A and B), the froth
discharge velocity is measured using VisioFroth cameras (Metso
Minerals, 2006) and recorded using the PI (process information)
system. In these rougher circuits, all reagents were added in the
conditioning tank located at the head of the flotation banks.
Metallurgical samplings for mass balances were performed in
each cell of the rougher flotation banks. Three sampling campaigns
were conducted in bank A, and one sampling campaign was conducted in bank B. The feed, tail and concentrate were sampled
per circuit during the surveys of approximately 23 h, depending
on availability. Feed samples were collected from the feed box,

2.3. Froth discharge velocity measurements


The froth discharge velocity was measured in each rougher cell
using the installed VisioFroth cameras. Fig. 2b shows the sampling
point location. The VisioFroth system includes software that computes the froth discharge velocity, the bubble size distribution, the
froth stability (bubble collapse rate), the froth colorimetry, and
changes in the froth texture amongst other properties. The froth
velocity was obtained from the displacement between consecutive

(a)
PR

C1

C2

C3

C4

C5

C6

C6

C7

C8

(b)
C1

C2

C3

C4

C5

Fig. 1. Arrangement of cells in series, rougher flotation (a) bank A and (b) bank B.

C9

244

J. Yianatos et al. / Minerals Engineering 92 (2016) 242247

Table 1
Cu recoveries and operating conditions in the rougher flotation circuits A and B.
Sampling campaign
Bank A

Bank B

Mass balance method

1
Cell-by-cell

2
Cell-by-cell

3
Short-cut

4
Short-cut

Cu feed grade, %
Feed tonnage, tph
Cu recovery, %
JG, cm/s
hF, cm
xTOF of Cu, %
vD, cm/s

0.6
781
92.6
0.630.89
511
3.833.9
0.811.9

0.56
798
91.6
0.450.74
339
2.033.1
0.513.1

0.65
732
89.7
Not available
440
0.830.4
3.09.2

0.64
798
83.9
0.731.28
430
3.328.3
1.37.2

images using the modified Fast Fourier Transform (Runge et al.,


2007).
Fig. 3 shows profiles of the froth discharge velocity
(Fig. 3a and c) and the TOF grade of Cu (Fig. 3b and d) along the
rougher banks A and B, respectively. Table 2 shows the froth
depths along the rougher flotation bank. A relatively sharp
decrease in the froth discharge velocity was observed along bank
A (Fig. 3a). In Fig. 3b, a decrease in the TOF grade of Cu is also
observed, which is mainly related to the decrease of the floatable
mineral liberation down the rougher bank A. Froth depths in bank
A were relatively shallow (314 cm) except in the pre-rougher cell,
where froth depths of 1140 cm were observed. For bank B
(Fig. 3c and d), the TOF grade and the froth discharge velocity oscillate along the circuit because of the differences in the froth depth
of the cells from the same bank, as shown in Table 2 (after cell 1
each bank consists of two paired cells: C2C3, C4C5, C6C7,
C8C9). The first cell of each paired cells showed froth depths that
were significantly lower than those observed in the second cell.
The lower froth depth in the first cell means that a larger pressure
difference was required for the transport of pulp from the first to
the second cell of each pair. This effect can be attributed to design
constraints for the current pulp flowrates between the paired cells.
Additionally, partial solid deposits on the bottom of the cells (e.g.,
coarser particles) contribute to increase the pressure difference
required for the pulp transport. The cells in rougher bank B are

self-aerated; therefore, differences in the apparent pulp density,


due to gas holdup changes, are less likely to affect consistently
all the paired cells.
Fig. 3c and d shows an inverse correlation between the froth
discharge velocity and the TOF grades for bank B, from the third
cell onwards. This correlation can be explained in terms of the
decrease in the froth residence time due to the froth depth
decreasing, whereas the froth discharge velocity increases at lower
froth depths. Low values for the froth residence times involve nonselective froths and lower TOF grades, compared with higher froth
residence times (selective froths). Yianatos et al. (2015) reported a
more detailed analysis of this subject.
3. New method for flotation characterization
A new approach for the Cu recovery estimation along rougher
flotation banks is proposed; this approach is based on two nonconventional measurements. The methodology assumes that the
mineral recovery in a flotation cell is proportional to the TOF grade
and to the froth discharge velocity, as shown Eq. (1). The TOF grade
represents the mass fraction of the valuable mineral recovered to
the concentrate, whereas the froth discharge velocity is proportional to the mass flowrate entering the concentrate stream.

RCu;i a  v disch;i  xTOF;i ;

(a)

i cell number

(b)
External
Launder
Froth
Crowder

Sample
Froth

Pulp

TOF sampling
point

Internal
Launder

Discharge froth velocity


sampling point

Fig. 2. (a) Top-of-froth measurement and (b) sampling points.

245

J. Yianatos et al. / Minerals Engineering 92 (2016) 242247

(b)
Sampling Campaign 1
Sampling Campaign 2
Sampling Campaign 3

10

40

Sampling Campaign 1
Sampling Campaign 2
Sampling Campaign 3

35

TOF grade of Cu, %

Froth discharge velocity, cm/s

(a)
15

30
25
20
15
10
5
0

0
PR

C1

C2

C3

C4

C5

PR

C6

C1

C3

C4

C5

C6

Cell number

Cell number

(c)

(d)
40

15

35

TOF grade of Cu, %

Froth discharge velocity, cm/s

C2

10

30
25
20
15
10
5
0

0
C1 C2 C3 C4 C5 C6 C7 C8 C9

C1 C2 C3 C4 C5 C6 C7 C8 C9

Cell number

Cell number

Fig. 3. Profiles of (a) the froth discharge velocity, bank A, (b) the TOF grade of Cu, bank A, (c) the froth discharge velocity, bank B and (d) the TOF grade of Cu, bank B.

Table 2
Froth depth in each cell of the rougher flotation circuits A and B.
Cell

Sampling campaign
Bank A

PR
C1
C2
C3
C4
C5
C6
C7
C8
C9

Bank B

11 cm
6 cm
10 cm
6 cm
6 cm
5 cm
6 cm

39 cm
13 cm
7 cm
8 cm
14 cm
3 cm
4 cm

40 cm
5 cm
8 cm
5 cm
6 cm
4 cm
4 cm

10 cm
7 cm
23 cm
4 cm
30 cm
6 cm
30 cm
5 cm
11 cm

In Eq. (1), a is a function of the froth properties, such as the mineral size and density, the gas holdup in the froth, the flow area to
the concentrate stream, the solid percentage in the froth and a factor related to the momentum variation between the sampling
points (i.e., sampling points of TOF and froth discharge velocity,
Fig. 2b). It should be noted that the a values depend upon the operating conditions, which cause changes in the froth characteristic
down the bank. These changes are mainly related to the profiles
of the superficial gas rate and the froth depth. In this sense, the a
parameter would change cell by cell. From a practical point of
view, however, it has been observed that a single value of a
(weighted-average value) can be adequately used to describe the
kinetics of the whole bank. Therefore, assuming that a is the

weighted-average value for the whole bank, the total rougher


recovery can be obtained from Eq. (2). This approach implies that
changes in the froth properties and the flow area to the concentrate stream are non-significant in the first cells, where the main
fractional recovery occurs. In addition, the incremental cell recovery becomes less significant down the bank and has a less significant impact on the flotation rate characterization.

Rbank;Cu a 

6
X
v disch;i  xTOF;i ;

i cell number

i1

3.1. Summary of the calculation method


(i) Characterize the circuit by measuring the froth discharge
velocities and the TOF grades in each cell of the rougher
bank (Fig. 3).
(ii) Perform the total mass balance for the flotation bank to
obtain the overall recovery of the valuable mineral (Table 1).
(iii) Determine a from Eq. (2) using the TOF grades, the froth discharge velocities and the overall recovery determined in
step (ii).
(iv) Obtain a mineral recovery profile using Eq. (1) and a estimation. The latter is obtained in step (iii).
3.2. Dimensionless approach
The requirement to sample the flotation feed, the tailings and
the concentrate of the bank of cells to determine the total recovery

246

J. Yianatos et al. / Minerals Engineering 92 (2016) 242247

(a)

(b)

0.6
0.5

0.5

RCu,i /RCu,bank

0.4

RCu,i /RCu,bank

0.6

Sampling Campaign 1
Sampling Campaign 2
Sampling Campaign 3

0.3
0.2
0.1

0.4
0.3
0.2
0.1

0.0

0.0
PR

C1

C2

C3

C4

C5

C6

Cell number

C1 C2 C3 C4 C5 C6 C7 C8 C9

Cell number

Fig. 4. Ratio between Cu recovery per cell and the total Cu recovery of the bank (fractional recovery profile). (a) bank A, (b) bank B.

may introduce errors during sampling, assaying, analysing,


amongst others. Another way of looking at the flotation rate characterization, which uses the top-of-froth grades and the froth discharge velocities, is by a dimensionless approach for estimating the
fractional recovery profile in an industrial flotation bank. In this
case, dividing Eq. (1) by Eqs. (2) and (3) is obtained:

Ri;Cu
v disch;i  xTOF;i
P6
;
Rbank;Cu
i1 v disch;i  xTOF;i

i cell number

Eq. (3) gives the fractional recovery profile for each cell from the
discharge velocities and the TOF grades. If the total recovery of the
bank is available, this can be multiplied by the fractional recoveries
to obtain the actual cell recovery. Fig. 4a and b shows the fractional
recovery profiles along the rougher flotation bank A and B, respectively. In both banks, the first cell recovered approximately 4050%
of the total Cu recovered in each circuit. The fractional recovery
profiles are smoothed versions of the TOF grade and the froth discharge velocity profiles shown in Fig. 3a and b.
4. Results and discussions
The froth discharge velocities (Fig. 3a), the TOF grades of Cu
(Fig. 3b) and the rougher recovery of Cu (Table 1) were used to

Cu recovery estimation, %

100
90
80
70
60
PR

Sampling 1
Sampling 2
Sampling 3
Sampling 4
y=x
10%

50
40
30
30

40

50

60

70

80

90

100

Copper recovery from mass balance, %


Fig. 5. Cu recovery estimated from the new approach (Eq. (2)) versus the Cu
recovery obtained from the mass balances (cell-by-cell) in rougher banks A and B.

obtain the weighted average a values for each sampling campaign


from Eq. (2). The values obtained for a were 0.12, 0.10 and 0.13 for
bank A (sampling campaigns 1, 2 and 3, respectively) and 0.19 for
bank B (sampling campaign 4). The differences in the a values of
each bank are attributed to equipment characteristics, i.e., equipment size and design, as well as differences in the mineral characteristics and operating conditions. Fig. 5 shows a comparison of the
cumulative Cu recovery from the mass balance and the predicted
Cu recovery using the a values along with Eq. (1), for sampling
campaigns 14. The new approach for Cu recovery estimation
showed a good agreement with the recovery values obtained from
the mass balance (cell-by-cell), in which more than 85% of the data
were included in a range of 10% relative error. A higher deviation
was observed in the PR cell where a difference of less than 16% was
obtained. The greater differences were found in sampling campaign 1. Fig. 6a shows the kinetic response along the rougher bank
A for sampling campaign 2 using the a values and the cumulative
results of Eq. (1). The Cu recoveries from the mass balance are
shown for comparison. For sampling campaign 2, the Cu recovery
for cell 2 was smaller when using the new approach than when
obtained from the cell-by-cell mass balance. A maximum relative
difference close to 6% was obtained in the second cell. This difference may be caused by the variability of the TOF grade between
cells (Fig. 3a) or by the uncertainties in the recovery results
obtained from the mass balances. Fig. 6b shows a comparison
between the Cu recovery obtained from the mass balances and
the Cu recovery obtained from the new approach in the rougher
bank B. The available mass balance information only included the
first cell and the overall rougher circuit (short-cut method). The
recovery profile obtained from the new approach shows a relatively linear increase rather than a first-order behaviour because
of the significant differences in the froth discharge velocity of the
cells (Fig. 3c). These differences may negatively influence the overall mineral recovery of the bank. A relative difference close to 12%
was observed in the first cell of the bank B. It should be noted the
new approach allowed low incremental recoveries to be
determined.
Based on the froth discharge velocities and the TOF grades, the a
parameters that satisfy the Cu recovery results for different banks
and surveys (Table 1) were obtained, which allowed the Cu recovery profiles along two rougher flotation banks to be estimated. The
operating conditions for the different sampling campaigns in bank
A were stable for the same flotation banks; consequently, the
resulting a parameters were similar.

247

J. Yianatos et al. / Minerals Engineering 92 (2016) 242247

(b)
100

90

90

80

80

Copper recovery, %

Copper recovery, %

(a)
100

70
60
50
40
30
20
Data from mass balance
New approach (=0.10)

10
0

70
60
50
40
30
20
Data from mass balance
New approach (=0.19)

10
0

PR

C1

C2

C3

C4

C5

C6

C1

Cell number

C2

C3

C4

C5

C6

C7

C8

C9

Cell number

Fig. 6. Comparison between the Cu recoveries obtained from the mass balances (cell-by-cell) and from the new approach: (a) rougher bank A, sampling campaign 2 and (b)
rougher bank B, sampling campaign 4.

5. Conclusions

References

A new approach for metallurgical characterization in rougher


flotation banks has been proposed. The mineral recovery in a
flotation cell was considered proportional to both the top-offroth (TOF) grade and the froth discharge velocity. The proportionality parameter (a) was obtained from the total rougher recovery
and allowed the kinetic response to be reconstructed.
The weighted average a values were approximately 0.1 and 0.2
for banks A and B, respectively. These differences in the a values for
each bank were attributed to the equipment characteristics and
the different operating conditions. The reconstructed Cu recovery
profiles showed good agreement with the recovery profiles
obtained from the reconciled mass balances (cell-by-cell) in two
different industrial flotation banks. The new methodology allows
the kinetic response of the flotation banks to be characterized
using two non-conventional measurements: the top-of-froth grade
and the froth discharge velocity.
A dimensionless approach for estimating the fractional recovery
in an industrial flotation bank was derived. This recovery profile
was obtained from the discharge velocities and the TOF grades
with no additional metallurgical data. This approach will be useful
when the total recovery is not available.
The proposed methodology is a powerful tool for the characterization of flotation circuits at industrial scale using two simple and
reliable measurement systems: the TOF grade and the froth discharge velocity. This approach has the potential to reduce the
effort and expense of metallurgical audits and surveys for plants
with installed froth discharge velocity sensors and especially when
the collection of samples is complex (e.g., tails samples).

Concha, J., 2009. Balance de materiales y caracterizacin cintica de la flotacin


Chemical Engineer Thesis. Dept. of Chemical and Environmental Eng.,
Universidad Tcnica Federico Santa Mara.
Di Sandro, D., Laskowski, D., Schwarz, S., 2013. Current practice on structuring,
conducting and interpreting a flotation survey. In: 10th International Mineral
Processing Conference, Procemin 2013, Santiago, Chile, 1518 October.
Falutsu, M., Dobby, G., 1992. Froth performance in commercial sized flotation
columns. Miner. Eng. 5, 12071223.
Hadler, K., Cilliers, J., 2009. The relationship between the peak in air recovery and
flotation bank performance. Miner. Eng. 22 (5), 451455.
Hadler, K., Greyling, M., Plint, N., Cilliers, J., 2012. The effect of froth depth on air
recovery and flotation performance. Miner. Eng. 3638, 248253.
Metso Minerals, 2006. Gua de instalacin y del usuario de OCS VisioFrothTM. Metso
Minerals, http://www.metso.com/MetsoMinerals.
Neethling, S., Cilliers, J., 2008. Predicting air recovery in flotation cells. Miner. Eng.
21 (1214), 937943.
Runge, K., McMaster, J., Wortley, M., La Rosa, D., Guyot, O., 2007. A correlation
between VisiofrothTM measurements and the performance of a flotation cell. In:
Ninth Mill Operators Conference, Fremantle, Western Australia, 1921 March.
Smith, C., Hadler, K., Cilliers, J., 2010. Flotation bank air addition and distribution for
optimal performance. Miner. Eng. 23 (1113), 10231029.
Wills, B., Napier-Munn, T., 2006. Mineral Processing Technology. Elsevier Science &
Technology Books.
Yianatos, J., Bergh, L., Vinnett, L., Panire, I., Iriarte, V., 2015. Correlation between top
of froth grade and operational variables in rougher flotation circuits. In:
Proceedings of Flotation15, Cape Town, South Africa.
Yianatos, J., Bergh, L., Vinnett, L., Rojas, I., 2014. On the collection of valuable
minerals along rougher flotation banks. Miner. Eng. 6668, 202206.
Yianatos, J., Henrquez, F., 2006. Short-cut method for flotation rates modelling of
industrial flotation banks. Miner. Eng. 19, 13361340.
Yianatos, J., Henrquez, F., Oroz, A., 2006. Characterization of large size flotation
cells. Miner. Eng. 19, 531538.
Yianatos, J., Moys, M., Contreras, F., Villanueva, A., 2008. Froth recovery of industrial
flotation cells. Miner. Eng. 21 (1214), 817825.

Acknowledgements
Funding for process modelling and control research was provided by CONICYT, Project Fondecyt 1160547, and by Federico
Santa Maria Technical University, Project 116.27.2.

Vous aimerez peut-être aussi