2926 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
a body of earlier studies points to at least three other mech-
anisms that impact on the overall efficiency of flotation.
These are a) the cleaning of particle surfaces, b) increased
liberation of valuable minerals, as well as c) an increased
gas holdup and bubble flux. The impact of each of these
mechanisms is a function of the type of ore being treated,
the grind and hence the liberation, and also the extent to
which valuable mineral surfaces are passivated.
Cleaning of Particle Surfaces
The feed slurry to an HCD such as the Mach Reactor is
exposed to very high energy dissipation rates and shear,
especially when compared to the conditions in conven-
tional mechanical flotation cells. These Reactors therefore
can play a major role in situations where, for instance, sul-
phide minerals are passivated by oxidised layers, such as
typically experienced in the case where old tailings dams
are retreated, or when freshly liberated sulphide mineral
surfaces are coated by hydroxides or gangue slimes. The
intense inter-particle attritioning taking place when the
slurry is recirculated through the Reactor aids the removal
of passivating layers and enhancing the flotation rate as a
result, leading to significantly increased flotation kinetics
across a range of minerals and applications.
Increased Liberation
A further element of the cleaning of particle surfaces due to
the repeated and intensive inter-particle interaction is that
partly liberated valuable particles could get exposed increas-
ingly as the gangue matrix is systematically being worn
away. In certain cases, such as where PGMs are associated
with chromite particles on the grain boundaries, this could
lead to direct liberation and thus increased kinetics of the
valuables. The results of earlier tests seem to support such a
mechanism, showing a shift in the particle size distribution
after the slurry was processed through the Mach.
Increased Gas Holdup and Bubble Flux
With additional air being introduced into the feed to a flo-
tation cell, in the form of very fine bubbles, the bubble size
distribution in the pulp is shifted to a smaller average. This
also causes an increase in the gas holdup in the pulp phase
in earlier cases figures of as high as 50% were measured
which is a significant increase on the levels of around 10
to 15% that are typical of mechanically agitated flotation
cells. An important impact of these is that the degree of
entrainment is significantly reduced in many cases, lead-
ing to increased selectivity between the valuables and the
gangue and thus an increase in the grade of the concentrate.
This paper describes the salient results of a comprehen-
sive test programme that was carried out at the SMC con-
centrator to assess the likely impact that the introduction
of the Mach Reactor technology would have on the metal-
lurgical performance of the flotation plant.
METHODOLOGY
Samples
Slurry samples were collected by manual cross-stream sam-
plers from the Rougher feed, Rougher Scavenger feed,
Cleaner feed and Final Tailings streams in the SMC process-
ing plant to conduct the Mach Reactor flotation tests. The
average of the measured 4E grades for the Rougher feed,
Scavenger feed and Cleaner feed were 3.3 g/t, 4.2 g/t and
8.1 g/t, respectively, with the average of the Final Tailings
being 0.7 g/t. It should be stressed however that these grades
are not necessarily indicative of plant averages, as these were
essentially one-off grab samples and for instance not com-
posited. Each sample was a total volume of 50 litres, which
was transferred to an agitated slurry holding tank in the
Mach Reactor test rig (Figure 1).
Preconditioning and Flotation
Preconditioning of the various samples were done by circu-
lation of the slurry through the Mach Reactor for a required
number of ‘passes’ (i.e., one pass equates to a volume of
50 L being circulated). In early set-up testwork, it was
observed that addition of air into the Mach Reactor dur-
ing preconditioning resulted in flotation occurring inside
the mixing tank, with a mineralised froth phase building.
As a result, the data obtained using the standard operating
procedure turned out to be unexpectedly poor, for instance
characterized by very low concentrates grade and lower
than expected feed grades.
The suspicion was that, not only did the addition of
air cause flotation in the mixing tank, but also had a nega-
tive impact on the performance by over-oxidation of the
sulphide minerals and thereby impeding the nucleation of
nano-bubbles (NBs) on the mineral surfaces. As a result,
for all the tests reported here, no air was introduced to the
Mach Reactor during the preconditioning, and the sam-
ple to be floated was collected at the Mach Reactor dis-
charge pipe, rather than at the drain valve on the pump
suction line.
After preconditioning, a batch of slurry of the required
volume was extracted as an aliquot from the mixing tank
and transferred into the Denver D12 cell, with a nomi-
nal volume of 5 L. The various reagents were then added
in the cell, as shown in Table 1, with CuSO4 as activator,
sodium isobutyl xanthate (SIBX) as collector, FinFix 300 as
a body of earlier studies points to at least three other mech-
anisms that impact on the overall efficiency of flotation.
These are a) the cleaning of particle surfaces, b) increased
liberation of valuable minerals, as well as c) an increased
gas holdup and bubble flux. The impact of each of these
mechanisms is a function of the type of ore being treated,
the grind and hence the liberation, and also the extent to
which valuable mineral surfaces are passivated.
Cleaning of Particle Surfaces
The feed slurry to an HCD such as the Mach Reactor is
exposed to very high energy dissipation rates and shear,
especially when compared to the conditions in conven-
tional mechanical flotation cells. These Reactors therefore
can play a major role in situations where, for instance, sul-
phide minerals are passivated by oxidised layers, such as
typically experienced in the case where old tailings dams
are retreated, or when freshly liberated sulphide mineral
surfaces are coated by hydroxides or gangue slimes. The
intense inter-particle attritioning taking place when the
slurry is recirculated through the Reactor aids the removal
of passivating layers and enhancing the flotation rate as a
result, leading to significantly increased flotation kinetics
across a range of minerals and applications.
Increased Liberation
A further element of the cleaning of particle surfaces due to
the repeated and intensive inter-particle interaction is that
partly liberated valuable particles could get exposed increas-
ingly as the gangue matrix is systematically being worn
away. In certain cases, such as where PGMs are associated
with chromite particles on the grain boundaries, this could
lead to direct liberation and thus increased kinetics of the
valuables. The results of earlier tests seem to support such a
mechanism, showing a shift in the particle size distribution
after the slurry was processed through the Mach.
Increased Gas Holdup and Bubble Flux
With additional air being introduced into the feed to a flo-
tation cell, in the form of very fine bubbles, the bubble size
distribution in the pulp is shifted to a smaller average. This
also causes an increase in the gas holdup in the pulp phase
in earlier cases figures of as high as 50% were measured
which is a significant increase on the levels of around 10
to 15% that are typical of mechanically agitated flotation
cells. An important impact of these is that the degree of
entrainment is significantly reduced in many cases, lead-
ing to increased selectivity between the valuables and the
gangue and thus an increase in the grade of the concentrate.
This paper describes the salient results of a comprehen-
sive test programme that was carried out at the SMC con-
centrator to assess the likely impact that the introduction
of the Mach Reactor technology would have on the metal-
lurgical performance of the flotation plant.
METHODOLOGY
Samples
Slurry samples were collected by manual cross-stream sam-
plers from the Rougher feed, Rougher Scavenger feed,
Cleaner feed and Final Tailings streams in the SMC process-
ing plant to conduct the Mach Reactor flotation tests. The
average of the measured 4E grades for the Rougher feed,
Scavenger feed and Cleaner feed were 3.3 g/t, 4.2 g/t and
8.1 g/t, respectively, with the average of the Final Tailings
being 0.7 g/t. It should be stressed however that these grades
are not necessarily indicative of plant averages, as these were
essentially one-off grab samples and for instance not com-
posited. Each sample was a total volume of 50 litres, which
was transferred to an agitated slurry holding tank in the
Mach Reactor test rig (Figure 1).
Preconditioning and Flotation
Preconditioning of the various samples were done by circu-
lation of the slurry through the Mach Reactor for a required
number of ‘passes’ (i.e., one pass equates to a volume of
50 L being circulated). In early set-up testwork, it was
observed that addition of air into the Mach Reactor dur-
ing preconditioning resulted in flotation occurring inside
the mixing tank, with a mineralised froth phase building.
As a result, the data obtained using the standard operating
procedure turned out to be unexpectedly poor, for instance
characterized by very low concentrates grade and lower
than expected feed grades.
The suspicion was that, not only did the addition of
air cause flotation in the mixing tank, but also had a nega-
tive impact on the performance by over-oxidation of the
sulphide minerals and thereby impeding the nucleation of
nano-bubbles (NBs) on the mineral surfaces. As a result,
for all the tests reported here, no air was introduced to the
Mach Reactor during the preconditioning, and the sam-
ple to be floated was collected at the Mach Reactor dis-
charge pipe, rather than at the drain valve on the pump
suction line.
After preconditioning, a batch of slurry of the required
volume was extracted as an aliquot from the mixing tank
and transferred into the Denver D12 cell, with a nomi-
nal volume of 5 L. The various reagents were then added
in the cell, as shown in Table 1, with CuSO4 as activator,
sodium isobutyl xanthate (SIBX) as collector, FinFix 300 as
