XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 2029
INTRODUCTION
The dense medium cyclone (DMC) separates solids pri-
marily by density. It is a high-tonnage device that has been
used widely especially in the coal industry to upgrade run-
of-mine material in separating gangue from product coal.
It has also shown application in mineral processing in the
treatment of iron ore, diamonds, lead-zinc ores (Udaya et
al., 2005). The flow in a DMC is very complicated with
the presence of vorticity turbulence, an air core and seg-
regation medium. It involves multiple phases: fluid (dense
medium), air and particles of different sizes, densities and
other properties (Wang et al., 2009).
Its general working principle has been well documented
in the literature (Vakamalla and Mangadoddy, 2023). The
feed enters tangentially near the top of the cylindrical sec-
tion, thus forming a strong swirling flow. Centrifugal forces
cause the denser or larger particles to move towards the
wall, where the axial velocity points predominantly down-
ward, and to discharge through the spigot. The less dense
and smaller particles move towards the longitudinal axis of
the DMC, where there is usually an air core, and the axial
velocity tends upward and transports this material through
the vortex finder. Separation by DMCs has been proved to
be effective, and today this process is employed in increas-
ingly wider areas of application.
High quality reserves of many ore bodies are being
rapidly depleted in many regions, including Africa. This
research was thus initiated with the objective of investigat-
ing the application of the DMC as a means of upgrading
more complex and lower grade ore bodies (Singh, 2022).
Sedimentary phosphate (P2O5) was selected for this
endeavor, for which downstream applications (phosphoric
acid or direct fertilizer) should meet market specification of
31% P2O5.
The results from this study highlighted the challenges
of gravity separation as an upgrade route for phosphate
due to a marginal differential density between the phos-
phate bearing mineral and quartz gangue. However, it was
demonstrated that lower grade material was amenable to
upgrade via Dense Medium Separation (DMS) although
there was a large amount of near density material around
narrow cut points. The recommended methodology is con-
firmed by HLS modelling data, DMS pilot testing and
CFD modelling of the dense medium cyclone.
METHODOLOGY
Gravity Separation
Phosphate bearing ores are rarely upgraded via gravity
separation due to the marginal density difference between
valuable mineral and gangue. However, dense medium
separation was undertaken in these circumstances whereby
the composition of quartz contained within an intermedi-
ate enriched fraction showed a liberated pebble like forma-
tion. The detailed methodology followed for the test work
is described in Singh et al. (2022).
The scrubber product underwent gravity separation to
produce a concentrate at a saleable P2O5 grade, by reduc-
ing the quartz content. To determine the feasibility for
pilot DMS testwork, washability data was modelled. Based
on the positive results, pilot scale DMS testwork was per-
formed. Results were compared to those obtained via labo-
ratory and simulations.
Laboratory Heavy Liquid Separation Testwork (HLS)
Laboratory heavy liquid separation (HLS) tests were ini-
tially carried out to benchmark the performance of a pilot
DMS operation. The results from laboratory testing were
modelled to predict pilot performance which was compared
against actual pilot data with tests having been conducted.
Heavy Liquid Separation (HLS) is a laboratory tech-
nique used to evaluate the feasibility of gravity separation
via Dense Medium Separation (DMS) on a specific ore (de
Wet and Singleton, 2008). It also assists in determining
a suitable separating density and assessing the likely effi-
ciency of a dense media circuit by performing tests on the
sink and float products.
HLS aims to separate an ore into a series of fractions
based on density difference of minerals present in the ore.
During the process, the ore is added to a solution of a
certain specific gravity, allowing minerals to be classified
into floats and sinks due to different specific gravities. The
solution is prepared using mixtures of Tetrabromoethane
(TBE), Acetone and FerroSilicon (FeSi), depending on
the required specific gravity of the solution. For a specific
gravity below 2.96, TBE and acetone are mixed together,
however, for a specific gravity of above 2.96, TBE and FeSi
are mixed together. In this investigation, the predominant
gangue present in the phosphate was quartz, thus, densi-
ties below 3.0 g/cm3 were tested this included tests at 2.3,
2.5, 2.6, 2.65, 2.7 and 2.96g/cm3. The HLS procedure is
outlined in Figure 1.
Modelling Laboratory HLS Data to Predict Pilot Dense
Medium Separation (DMS) operation
The separation efficiency of gravity concentrators is based
on particle density. However, reduced particle size results
in decreased efficiency as revealed by size-based effi-
ciency curves (Galvin and Iverson, 2022). The Weibull
Distribution Model is a continuous probability distribution
that is mainly related to particle size. To adapt the model
INTRODUCTION
The dense medium cyclone (DMC) separates solids pri-
marily by density. It is a high-tonnage device that has been
used widely especially in the coal industry to upgrade run-
of-mine material in separating gangue from product coal.
It has also shown application in mineral processing in the
treatment of iron ore, diamonds, lead-zinc ores (Udaya et
al., 2005). The flow in a DMC is very complicated with
the presence of vorticity turbulence, an air core and seg-
regation medium. It involves multiple phases: fluid (dense
medium), air and particles of different sizes, densities and
other properties (Wang et al., 2009).
Its general working principle has been well documented
in the literature (Vakamalla and Mangadoddy, 2023). The
feed enters tangentially near the top of the cylindrical sec-
tion, thus forming a strong swirling flow. Centrifugal forces
cause the denser or larger particles to move towards the
wall, where the axial velocity points predominantly down-
ward, and to discharge through the spigot. The less dense
and smaller particles move towards the longitudinal axis of
the DMC, where there is usually an air core, and the axial
velocity tends upward and transports this material through
the vortex finder. Separation by DMCs has been proved to
be effective, and today this process is employed in increas-
ingly wider areas of application.
High quality reserves of many ore bodies are being
rapidly depleted in many regions, including Africa. This
research was thus initiated with the objective of investigat-
ing the application of the DMC as a means of upgrading
more complex and lower grade ore bodies (Singh, 2022).
Sedimentary phosphate (P2O5) was selected for this
endeavor, for which downstream applications (phosphoric
acid or direct fertilizer) should meet market specification of
31% P2O5.
The results from this study highlighted the challenges
of gravity separation as an upgrade route for phosphate
due to a marginal differential density between the phos-
phate bearing mineral and quartz gangue. However, it was
demonstrated that lower grade material was amenable to
upgrade via Dense Medium Separation (DMS) although
there was a large amount of near density material around
narrow cut points. The recommended methodology is con-
firmed by HLS modelling data, DMS pilot testing and
CFD modelling of the dense medium cyclone.
METHODOLOGY
Gravity Separation
Phosphate bearing ores are rarely upgraded via gravity
separation due to the marginal density difference between
valuable mineral and gangue. However, dense medium
separation was undertaken in these circumstances whereby
the composition of quartz contained within an intermedi-
ate enriched fraction showed a liberated pebble like forma-
tion. The detailed methodology followed for the test work
is described in Singh et al. (2022).
The scrubber product underwent gravity separation to
produce a concentrate at a saleable P2O5 grade, by reduc-
ing the quartz content. To determine the feasibility for
pilot DMS testwork, washability data was modelled. Based
on the positive results, pilot scale DMS testwork was per-
formed. Results were compared to those obtained via labo-
ratory and simulations.
Laboratory Heavy Liquid Separation Testwork (HLS)
Laboratory heavy liquid separation (HLS) tests were ini-
tially carried out to benchmark the performance of a pilot
DMS operation. The results from laboratory testing were
modelled to predict pilot performance which was compared
against actual pilot data with tests having been conducted.
Heavy Liquid Separation (HLS) is a laboratory tech-
nique used to evaluate the feasibility of gravity separation
via Dense Medium Separation (DMS) on a specific ore (de
Wet and Singleton, 2008). It also assists in determining
a suitable separating density and assessing the likely effi-
ciency of a dense media circuit by performing tests on the
sink and float products.
HLS aims to separate an ore into a series of fractions
based on density difference of minerals present in the ore.
During the process, the ore is added to a solution of a
certain specific gravity, allowing minerals to be classified
into floats and sinks due to different specific gravities. The
solution is prepared using mixtures of Tetrabromoethane
(TBE), Acetone and FerroSilicon (FeSi), depending on
the required specific gravity of the solution. For a specific
gravity below 2.96, TBE and acetone are mixed together,
however, for a specific gravity of above 2.96, TBE and FeSi
are mixed together. In this investigation, the predominant
gangue present in the phosphate was quartz, thus, densi-
ties below 3.0 g/cm3 were tested this included tests at 2.3,
2.5, 2.6, 2.65, 2.7 and 2.96g/cm3. The HLS procedure is
outlined in Figure 1.
Modelling Laboratory HLS Data to Predict Pilot Dense
Medium Separation (DMS) operation
The separation efficiency of gravity concentrators is based
on particle density. However, reduced particle size results
in decreased efficiency as revealed by size-based effi-
ciency curves (Galvin and Iverson, 2022). The Weibull
Distribution Model is a continuous probability distribution
that is mainly related to particle size. To adapt the model