2712 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
and simplicity of the method compared to the methods of
measuring froth recovery. The froth stability method could
be made an online measurement with relative ease through
the insertion of a small column below the pulp-froth inter-
face, using a laser measurement of froth height.
To test the methodology described above, the beta
value for the Cu/Ni ore was determined by only a single
froth recovery test. Since this involves testing at three dif-
ferent froth heights, it resulted in 3 froth recovery values at
3 different FRT/FS values and gave the beta values shown
in Figure 3. These were used to calculate froth recoveries for
11 metallurgical tests done at different conditions. Since
froth recoveries were not measured by the variable froth
depth method for these other conditions as they were in the
iron ore example, another method was required to deter-
mine froth recoveries. A batch flotation test was performed
at very shallow froth depth, where it was assumed that the
froth recovery was 100% and the multicomponent model
of Gorain et al. 1997 was applied:
k PS R
b f =(5)
where k =first order rate constant, P =ore floatability
parameter, Sb Rf =froth recovery. Since Rf is assumed to
be 1 in this case, and k and Sb are known, the ore float-
ability, P, can be calculated. This value was then applied
to the data from the hybrid cell and the froth recoveries
calculated. Figure 4 shows that the correlation between the
two methods is not as good as when the froth recoveries
were determined using the variable froth depth method for
every condition. This may be due to a number of factors.
Firstly, it is known that froth recoveries are not 100% when
using the batch flotation method at very shallow froths
(Amelunxen et al., 2014). This would result in a smaller
rate constant and, therefore, a smaller floatability param-
eter. When this floatability parameter was applied to the
hybrid tests, it would result in the calculation of a higher
than expected froth recovery, which is what is observed in
Figure 4. Secondly, since only three data points were used
to calculate the beta value, there is a possibility that there is
error in this value.
However, notwithstanding the uncertainty in the mea-
surements, this is still a viable methodology for the intro-
duction of an online froth recovery measurement to allow
online flotation models to run since currently there is no
such measure available.
CONCLUSIONS
This paper has shown a simple method for estimating froth
recoveries using measured froth stabilities and the froth
retention time that is easily available in any process control
environment. It was shown that the initial very large vari-
ability in the relationship between froth recovery and the
froth retention time was due to differences in froth stability
that were created by changing process parameters such as
particle size or reagent dosage. Once this was accounted
for by dividing the froth retention time by the froth stabil-
ity, the data all fell onto a common decaying exponential
relationship with a fitted β-value. This value is the same
for an ore type irrespective of the metal or mineral being
recovered since it is a property of the froth phase. Different
ore types will have different β-values, but once they are
known, they can be used to calculate the froth recoveries
that can be used in online process control in conjunction
with the froth retention time, an easily available parameter,
and the froth stability, which has the potential to be easily
measured online.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.8
0.9
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Cu Froth recovery (using Beta)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.8
0.9
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Ni Froth recovery (using Beta)
Figure 5. Comparison of froth recoveries calculated using Equation 3 and Equation 2 for the Cu, Ni ore (a) copper, (b) nickel
Cu
Froth
recovery
(using
k/PS)
b
Ni
Froth
recovery
(using
k/PS)
b
and simplicity of the method compared to the methods of
measuring froth recovery. The froth stability method could
be made an online measurement with relative ease through
the insertion of a small column below the pulp-froth inter-
face, using a laser measurement of froth height.
To test the methodology described above, the beta
value for the Cu/Ni ore was determined by only a single
froth recovery test. Since this involves testing at three dif-
ferent froth heights, it resulted in 3 froth recovery values at
3 different FRT/FS values and gave the beta values shown
in Figure 3. These were used to calculate froth recoveries for
11 metallurgical tests done at different conditions. Since
froth recoveries were not measured by the variable froth
depth method for these other conditions as they were in the
iron ore example, another method was required to deter-
mine froth recoveries. A batch flotation test was performed
at very shallow froth depth, where it was assumed that the
froth recovery was 100% and the multicomponent model
of Gorain et al. 1997 was applied:
k PS R
b f =(5)
where k =first order rate constant, P =ore floatability
parameter, Sb Rf =froth recovery. Since Rf is assumed to
be 1 in this case, and k and Sb are known, the ore float-
ability, P, can be calculated. This value was then applied
to the data from the hybrid cell and the froth recoveries
calculated. Figure 4 shows that the correlation between the
two methods is not as good as when the froth recoveries
were determined using the variable froth depth method for
every condition. This may be due to a number of factors.
Firstly, it is known that froth recoveries are not 100% when
using the batch flotation method at very shallow froths
(Amelunxen et al., 2014). This would result in a smaller
rate constant and, therefore, a smaller floatability param-
eter. When this floatability parameter was applied to the
hybrid tests, it would result in the calculation of a higher
than expected froth recovery, which is what is observed in
Figure 4. Secondly, since only three data points were used
to calculate the beta value, there is a possibility that there is
error in this value.
However, notwithstanding the uncertainty in the mea-
surements, this is still a viable methodology for the intro-
duction of an online froth recovery measurement to allow
online flotation models to run since currently there is no
such measure available.
CONCLUSIONS
This paper has shown a simple method for estimating froth
recoveries using measured froth stabilities and the froth
retention time that is easily available in any process control
environment. It was shown that the initial very large vari-
ability in the relationship between froth recovery and the
froth retention time was due to differences in froth stability
that were created by changing process parameters such as
particle size or reagent dosage. Once this was accounted
for by dividing the froth retention time by the froth stabil-
ity, the data all fell onto a common decaying exponential
relationship with a fitted β-value. This value is the same
for an ore type irrespective of the metal or mineral being
recovered since it is a property of the froth phase. Different
ore types will have different β-values, but once they are
known, they can be used to calculate the froth recoveries
that can be used in online process control in conjunction
with the froth retention time, an easily available parameter,
and the froth stability, which has the potential to be easily
measured online.
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.8
0.9
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Cu Froth recovery (using Beta)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.8
0.9
1.0
0.0 0.2 0.4 0.6 0.8 1.0
Ni Froth recovery (using Beta)
Figure 5. Comparison of froth recoveries calculated using Equation 3 and Equation 2 for the Cu, Ni ore (a) copper, (b) nickel
Cu
Froth
recovery
(using
k/PS)
b
Ni
Froth
recovery
(using
k/PS)
b