XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 2855
interaction between size and shape variables and main-
mineral is used in the model to capture a singular rela-
tion between recovery and particle size and shape for each
mineral (as in an analysis of covariance framework, a.k.a
ANCOVA). After training the logistic regression, model
intercepts are adjusted to reflect the actual particle balance
of the process (i.e., prior adjustment).
RESULTS AND DISCUSSIONS
Model Mineralogy
Figure 3 displays the mineral composition of the calcu-
lated feed sample of four different pilot trials. It shows
that the feed composition is quite stable throughout the
pilot test campaign. The feed contains chalcocite (the main
Cu-bearing mineral) and chalcopyrite, carbonate gangue
(mainly dolomite, calcite, and ankerite), quartz, horn-
blende, orthoclase, and others.
Pilot Flotation Results
Figure 4 shows the recovery of all main minerals, i.e., chal-
cocite, chalcopyrite, dolomite, pyrite, and quartz increased
with a decrease in froth height. The chalcocite recovery
increased by approximately two times, i.e., 42% to 83%
when operated at shallow froth compared to the deep froth.
In contrast, Cu-bearing minerals grade is decreased with a
decrease in froth height, so there exists a trade-off between
those parameters to achieve balance in terms of Cu grade
and recovery. This finding could be applied in the plant
operation for different commodities or flotation duties, for
example, shallow froth could be used for the rougher stage
where the main target is to achieve a high recovery while
the deep froth in the cleaning stage helps to improve the
concentrate grade. Deep froth flotation also helps to reduce
the entrained of fine unwanted particles, less water pulls
and drop back of fine gangue particles to the pulp.
Figure 4 also displays the effect of recirculation load on
grade and recovery, test “mid froth” without recirculation
and with 6 m3/h recycled (approx. 30% to the feed flow
to the cell). As expected, a recirculation load of 6 m3/h led
to an increase in chalcocite recovery by approx.15%. The
chalcocite grade is slightly reduced with recycling.
Effect of Size and Liberation on the Recovery
Figure 5 shows the effect of the liberation degree on the
chalcocite recovery. For all the tests, the recovery of highly
liberated chalcocite particles is higher than that of poorly
liberated ones. Also, recovery increased with an increase in
particle size. Figure 5 also indicates that pneumatic cell can
recover the ultrafine (says below 3 -5 µm) liberated chal-
cocite. As seen in the grade and recovery plots (Figure 4),
the recirculation load improved the overall recovery. Results
in Figure 5 indicate that this is achieved by increasing the
recovery of both liberated and non-liberated particles slow
floating particles, due to the increase in residence time.
Such conditions are surely favourable in a scavenger setting,
where metal losses should be avoided.
Effect of Mineral Types and Size on Recovery
Figure 6 shows the effect of froth height, recirculation load,
mineral hydrophobicity, and size on the recovery of the lib-
erated minerals. The concentrate contains mostly fine liber-
ated sulfide-bearing particles i.e., chalcocite, chalcopyrite,
and pyrite below 30 µm. Due to this stream being the scav-
enger, coarser liberated Cu-bearing minerals already floated
at flash and rougher flotation and went to the cleaner stage.
Figure 4. Effect of froth height and recirculation load on the grade and recovery of the main minerals
interaction between size and shape variables and main-
mineral is used in the model to capture a singular rela-
tion between recovery and particle size and shape for each
mineral (as in an analysis of covariance framework, a.k.a
ANCOVA). After training the logistic regression, model
intercepts are adjusted to reflect the actual particle balance
of the process (i.e., prior adjustment).
RESULTS AND DISCUSSIONS
Model Mineralogy
Figure 3 displays the mineral composition of the calcu-
lated feed sample of four different pilot trials. It shows
that the feed composition is quite stable throughout the
pilot test campaign. The feed contains chalcocite (the main
Cu-bearing mineral) and chalcopyrite, carbonate gangue
(mainly dolomite, calcite, and ankerite), quartz, horn-
blende, orthoclase, and others.
Pilot Flotation Results
Figure 4 shows the recovery of all main minerals, i.e., chal-
cocite, chalcopyrite, dolomite, pyrite, and quartz increased
with a decrease in froth height. The chalcocite recovery
increased by approximately two times, i.e., 42% to 83%
when operated at shallow froth compared to the deep froth.
In contrast, Cu-bearing minerals grade is decreased with a
decrease in froth height, so there exists a trade-off between
those parameters to achieve balance in terms of Cu grade
and recovery. This finding could be applied in the plant
operation for different commodities or flotation duties, for
example, shallow froth could be used for the rougher stage
where the main target is to achieve a high recovery while
the deep froth in the cleaning stage helps to improve the
concentrate grade. Deep froth flotation also helps to reduce
the entrained of fine unwanted particles, less water pulls
and drop back of fine gangue particles to the pulp.
Figure 4 also displays the effect of recirculation load on
grade and recovery, test “mid froth” without recirculation
and with 6 m3/h recycled (approx. 30% to the feed flow
to the cell). As expected, a recirculation load of 6 m3/h led
to an increase in chalcocite recovery by approx.15%. The
chalcocite grade is slightly reduced with recycling.
Effect of Size and Liberation on the Recovery
Figure 5 shows the effect of the liberation degree on the
chalcocite recovery. For all the tests, the recovery of highly
liberated chalcocite particles is higher than that of poorly
liberated ones. Also, recovery increased with an increase in
particle size. Figure 5 also indicates that pneumatic cell can
recover the ultrafine (says below 3 -5 µm) liberated chal-
cocite. As seen in the grade and recovery plots (Figure 4),
the recirculation load improved the overall recovery. Results
in Figure 5 indicate that this is achieved by increasing the
recovery of both liberated and non-liberated particles slow
floating particles, due to the increase in residence time.
Such conditions are surely favourable in a scavenger setting,
where metal losses should be avoided.
Effect of Mineral Types and Size on Recovery
Figure 6 shows the effect of froth height, recirculation load,
mineral hydrophobicity, and size on the recovery of the lib-
erated minerals. The concentrate contains mostly fine liber-
ated sulfide-bearing particles i.e., chalcocite, chalcopyrite,
and pyrite below 30 µm. Due to this stream being the scav-
enger, coarser liberated Cu-bearing minerals already floated
at flash and rougher flotation and went to the cleaner stage.
Figure 4. Effect of froth height and recirculation load on the grade and recovery of the main minerals