2958 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
The upper section is made up of inclined channels the
lower section is the aeration and fluidization zone. The feed
is not deslimed. The rejects are discharged from the bottom
of the system and are mostly coarse gangue particles. The
overflow products are fines and coarse hydrophobic par-
ticles. The overflow must be classified into fines and coarse
particles. The fines go to flotation and the coarse particles
to regrinding. The only results published (Crompton et al.
2022) are based on a lab-scale tests, which showed 90 %
recovery for 500-µm chalcopyrite particles. A schematic of
the coarseAIR ™ machine is shown in Figure 16. The simi-
larity to the RFC flotation cell is obvious.
DISCUSSION
There is very little information on the metallurgical per-
formance of large industrial flotation cells, especially as
related to particle size recovery. The literature is full of
papers on batch and pilot plant tests of performance vs.
particle size, probably because testing and data collection
for these tests are substantially easier than doing a similar
evaluation on large and very large flotation cells, where the
volumetric flotation feed rates can vary from 2,000–10,000
m3/h. Because of these difficulties, laboratory recoveries are
downgraded by 1% to 3% as a ‘rule of thumb’ estimate of
expected results at continuous industrial scale. With com-
plex ore mineralogy, high variability, and a complex treat-
ment circuit be complex recoveries up to 5% lower may be
expected.
Furthermore, laboratory and pilot flotation testing are
performed under steady or semi-steady state conditions,
where feed grades and flows, particle size, slurry densities,
and flotation reagents are constant. This is not the norm
for large scale concentrators dealing with high tonnages of
low-grade ores, where there are continuing perturbations
in feed grade, particle size, volumetric flow, and slurry den-
sity. In addition, full-scale, in-plant tests are expensive and
time consuming. A side-by-side test of two or more types
of tank cells will cost each supplier $5 to $10 million, and
expenses to the plant owner will be similar. A testing period
of 6 to 12 months will be required to achieve statistically
significant results.
Consequently, performance measurements in large flo-
tation cells have been focused on indirect hydrodynamic
and air dispersion measurements. Attempts to measure
metallurgical performance have been extremely difficult
given the variability in feed parameters. Most tests are lim-
ited to attempts at measuring concentrate grade and overall
recovery. The results are clouded by the feed variability and
lag times in a typical flotation typical circuit, with rough-
ers, scavengers, and cleaners interacting. Statistical methods
are required to determine if the measured improvements
are meaningful. Published information indicates that
improvements in concentrate grade are in general statisti-
cally valid, but this is not the case for ‘possible’ improve-
ments in recovery. For very large cells, grade improvements
of 1% have been measured and statistically substantiated.
The challenges of in-plant testing of large cells are described
in Nelson and Redden (1997), Grau et al. (2018), and
Yianatos et al. (2006)
Many publications demonstrate the improvements in
particle size recovery with pneumatic type machines. These
are usually pilot units evaluating the machine performance
on a particular flotation steam in a flotation cleaner cir-
cuit. When improvements in recovery, concentrate grade,
and particle size recovery are noted, one must ask how
these outcomes translate to overall performance increases
in grade and recovery. Typically, the overall benefits are
far less than the larger benefits seen at the pilot level on a
particular stream—usually 1–3% for recovery and 1% for
concentrate grade. Higher numbers may be attained, but
this is unusual.
While pneumatic machines can make improvements
in the recovery of coarse and fine particles, these machines
Figure 16. Schematic of the coarseAIR™ Flotation Cell
(Crompton et al. 2022)
The upper section is made up of inclined channels the
lower section is the aeration and fluidization zone. The feed
is not deslimed. The rejects are discharged from the bottom
of the system and are mostly coarse gangue particles. The
overflow products are fines and coarse hydrophobic par-
ticles. The overflow must be classified into fines and coarse
particles. The fines go to flotation and the coarse particles
to regrinding. The only results published (Crompton et al.
2022) are based on a lab-scale tests, which showed 90 %
recovery for 500-µm chalcopyrite particles. A schematic of
the coarseAIR ™ machine is shown in Figure 16. The simi-
larity to the RFC flotation cell is obvious.
DISCUSSION
There is very little information on the metallurgical per-
formance of large industrial flotation cells, especially as
related to particle size recovery. The literature is full of
papers on batch and pilot plant tests of performance vs.
particle size, probably because testing and data collection
for these tests are substantially easier than doing a similar
evaluation on large and very large flotation cells, where the
volumetric flotation feed rates can vary from 2,000–10,000
m3/h. Because of these difficulties, laboratory recoveries are
downgraded by 1% to 3% as a ‘rule of thumb’ estimate of
expected results at continuous industrial scale. With com-
plex ore mineralogy, high variability, and a complex treat-
ment circuit be complex recoveries up to 5% lower may be
expected.
Furthermore, laboratory and pilot flotation testing are
performed under steady or semi-steady state conditions,
where feed grades and flows, particle size, slurry densities,
and flotation reagents are constant. This is not the norm
for large scale concentrators dealing with high tonnages of
low-grade ores, where there are continuing perturbations
in feed grade, particle size, volumetric flow, and slurry den-
sity. In addition, full-scale, in-plant tests are expensive and
time consuming. A side-by-side test of two or more types
of tank cells will cost each supplier $5 to $10 million, and
expenses to the plant owner will be similar. A testing period
of 6 to 12 months will be required to achieve statistically
significant results.
Consequently, performance measurements in large flo-
tation cells have been focused on indirect hydrodynamic
and air dispersion measurements. Attempts to measure
metallurgical performance have been extremely difficult
given the variability in feed parameters. Most tests are lim-
ited to attempts at measuring concentrate grade and overall
recovery. The results are clouded by the feed variability and
lag times in a typical flotation typical circuit, with rough-
ers, scavengers, and cleaners interacting. Statistical methods
are required to determine if the measured improvements
are meaningful. Published information indicates that
improvements in concentrate grade are in general statisti-
cally valid, but this is not the case for ‘possible’ improve-
ments in recovery. For very large cells, grade improvements
of 1% have been measured and statistically substantiated.
The challenges of in-plant testing of large cells are described
in Nelson and Redden (1997), Grau et al. (2018), and
Yianatos et al. (2006)
Many publications demonstrate the improvements in
particle size recovery with pneumatic type machines. These
are usually pilot units evaluating the machine performance
on a particular flotation steam in a flotation cleaner cir-
cuit. When improvements in recovery, concentrate grade,
and particle size recovery are noted, one must ask how
these outcomes translate to overall performance increases
in grade and recovery. Typically, the overall benefits are
far less than the larger benefits seen at the pilot level on a
particular stream—usually 1–3% for recovery and 1% for
concentrate grade. Higher numbers may be attained, but
this is unusual.
While pneumatic machines can make improvements
in the recovery of coarse and fine particles, these machines
Figure 16. Schematic of the coarseAIR™ Flotation Cell
(Crompton et al. 2022)