2906 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
did not complete model validation on an independently
acquired data set.
Effect of Air Flow Rate on Copper Recovery
The results of the effect of air flow rate on copper recovery
when water rate was varied are presented in Figure 7. As
alluded to in the model, air flow rate showed a non-linear
relationship to copper recovery, with recovery increasing
with an increase in air flow rate before plateauing beyond a
critical value. This behavior was observed for all the water
rates tested. Figure 7 also shows that increasing the water
flow rate resulted in higher copper recovery, possibly due
to the increase in upward forces that are likely to transport
more particles into the concentrate stream.
Previous work by Demir et al. (2022) conducted on
a pilot-scale HydroFloat ® reported a similar relationship
between copper recovery and air flow rate. Increasing the
air flow rate increases the number of bubbles generated in
the cell per unit volume, increasing gas holdup (Liu et al.,
2020). This increases the probability of bubble-particle col-
lision and attachment, increasing copper recovery. However,
copper recovery plateaued above a certain critical air flow
rate, indicating the presence of another mechanism in the
high air flow region, which decreased recovery. An example
of such mechanisms could be turbulence at higher air flow
rates, the turbulence within the bed increases, which could
cause particles to detach from bubbles.
While not extensively discussed in this manuscript, it
is noteworthy to mention that in the small-scale tests, the
bed height was not maintained constant throughout each
experiment. Given that this is a batch system, the height
of the fluidised bed decreased as concentrate was removed
via true flotation and hydraulic elutriation. Additionally,
it was noted that the reduction in bed height was more
Table 3. Regression model statistics
Standard Error R-sq R-sq(adj)
3.7 85.2 78.8
Table 4. Statistics associated with the regression analysis coefficients
Term Coefficients Standard Error t Stat P-Value
Intercept 43.43 5.48 7.93 9.66E-05
Jg 340.44 103.75 3.28 1.35E-02
SWV 19.51 4.55 4.29 3.60E-03
Jg2 –1257.57 508.83 –2.47 4.27E-02
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80 90 100
Predicted Cu Recovery, %
R2 =85.2
R2(adj) =78.8
Figure 6. Measured copper recovery versus predicted copper recovery for the small-scale tests
Measured
CuRecovery,
%
did not complete model validation on an independently
acquired data set.
Effect of Air Flow Rate on Copper Recovery
The results of the effect of air flow rate on copper recovery
when water rate was varied are presented in Figure 7. As
alluded to in the model, air flow rate showed a non-linear
relationship to copper recovery, with recovery increasing
with an increase in air flow rate before plateauing beyond a
critical value. This behavior was observed for all the water
rates tested. Figure 7 also shows that increasing the water
flow rate resulted in higher copper recovery, possibly due
to the increase in upward forces that are likely to transport
more particles into the concentrate stream.
Previous work by Demir et al. (2022) conducted on
a pilot-scale HydroFloat ® reported a similar relationship
between copper recovery and air flow rate. Increasing the
air flow rate increases the number of bubbles generated in
the cell per unit volume, increasing gas holdup (Liu et al.,
2020). This increases the probability of bubble-particle col-
lision and attachment, increasing copper recovery. However,
copper recovery plateaued above a certain critical air flow
rate, indicating the presence of another mechanism in the
high air flow region, which decreased recovery. An example
of such mechanisms could be turbulence at higher air flow
rates, the turbulence within the bed increases, which could
cause particles to detach from bubbles.
While not extensively discussed in this manuscript, it
is noteworthy to mention that in the small-scale tests, the
bed height was not maintained constant throughout each
experiment. Given that this is a batch system, the height
of the fluidised bed decreased as concentrate was removed
via true flotation and hydraulic elutriation. Additionally,
it was noted that the reduction in bed height was more
Table 3. Regression model statistics
Standard Error R-sq R-sq(adj)
3.7 85.2 78.8
Table 4. Statistics associated with the regression analysis coefficients
Term Coefficients Standard Error t Stat P-Value
Intercept 43.43 5.48 7.93 9.66E-05
Jg 340.44 103.75 3.28 1.35E-02
SWV 19.51 4.55 4.29 3.60E-03
Jg2 –1257.57 508.83 –2.47 4.27E-02
0
10
20
30
40
50
60
70
80
90
100
0 10 20 30 40 50 60 70 80 90 100
Predicted Cu Recovery, %
R2 =85.2
R2(adj) =78.8
Figure 6. Measured copper recovery versus predicted copper recovery for the small-scale tests
Measured
CuRecovery,
%