40 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
sources. Two case studies are described in this paper to
showcase the key role of process mineralogy and geometal-
lurgy, one focusing on Zn and the other on the REE. Both
case studies are based on research at the University of Cape
Town.
ZINC CASE STUDY
Zn is a critical metal that finds application in green energy
battery storage, as well as longstanding use as a key com-
ponent in galvanizing steel thereby extending product life
span (Wang et al., 2021 Watari et al., 2021). The primary
source of zinc is from sphalerite which is recovered during
flotation to produce a concentrate before further refining,
most commonly through the roast leach electrowinning
process. The Gamsberg zinc deposit is a sediment-hosted,
strata-bound and metamorphosed sedimentary exhalative
Zn-Pb-Ag deposit in the Northern Cape Province in South
Africa. The first concentrate was produced from this ore in
2018 from the Gamsberg concentrator (Price et al., 2023).
Unlike most zinc ores, Gamsberg ore is characterized by
extensive Mn enrichment that is hosted in a variety of sili-
cate (garnet, pyroxene, pyroxmangite), oxide (pyrolusite)
and sulphide minerals (sphalerite, alabandite) (Schouwstra
et al., 2010). The distribution of Mn into this extensive
suite of minerals is consistent with its occurrence as a geo-
chemically abundant metal (Figure 2a) compared to zinc,
as a geochemically scarce metal (Figure 2b). The occurrence
of Mn substituting for Zn in the crystal lattice of sphaler-
ite, varying from almost pure sphalerite to Fe-Mn-rich vari-
ants with up to 20 wt.% Fe +Mn (McClung and Viljoen,
2011), represents an ongoing challenge at Gamsberg. This
case study describes a program of process mineralogical and
geometallurgical research focusing on finding solutions to
manage this challenge, some of which are described more
extensively in Price et al. (2023).
For any process mineralogy study, one of the main
objectives is to determine the characteristics of the valu-
able minerals, including understanding the distribution
of the valuable elements into these minerals, and their
associations whether in a run-of-mine ore, concentrate or
tailings. Three mineralized horizons, locally known as mag-
netite pyroxene (MPO), pyrrhotite-dominated pelitic ore
(PEO_PO) and pyrite-dominated pelitic ore (PEO_PY)
are combined to form the run of mine ore at Gamsberg.
An example of the mineralogy of these three ores is given
in Figure 4a. Sphalerite, pyrrhotite, pyrite, and quartz are
the dominant minerals in all three ore types. The MPO ore
also has significant garnet, pyroxene, pyroxenite and mag-
netite, whereas the pelitic ores have significant mica and
chlorite. As described by the nomenclature, the PEO_PO
ores have more pyrrhotite than pyrite, and vice versa for the
PEO_PY ores. The Mn distribution in these ores, as illus-
trated in Figure 4b, was determined through a combination
of auto-SEM-EDS (QEMSCAN) analysis to determine the
mineral grades and electron probe microanalysis (EPMA)
analysis to accurately determine the sphalerite composition.
The majority of the Mn in the MPO is hosted in garnet,
pyroxene and pyroxenoid (silicate minerals) with only neg-
ligible Mn hosted in sphalerite. Coupled with the higher
sphalerite grades in the MPO, this ore could be considered
the best ore. For the pelitic ores, however, up to 90% of the
Mn can be hosted in sphalerite with the balance of the Mn
occurring in garnet.
On flotation of this Zn ore, the final concentrate domi-
nantly comprises sphalerite, with minor galena, pyrite and
silicate minerals. The Mn grade of the final flotation con-
centrate, however, is a concern, since if it exceeds a cer-
tain threshold, refineries may impose treatment penalties
(the generally accepted range for Mn in zinc concentrate
is 0.02 to 0.8 wt.% Sinclair, 2005). Although the concen-
trate may also contain Fe derived from both sphalerite and
the Fe-sulphides (pyrite and pyrrhotite), the concentrate
quality specifications tend to be more lenient (the generally
accepted range for Fe in zinc concentrate is 1.5 to 10 wt.%
Sinclair, 2005). Because Mn occurs in the crystal lattice of
sphalerite, substituting for Zn through isomorphous sub-
stitution, it cannot be physically separated from a sphalerite
concentrate. Optimising the flotation circuit to minimize
the entrainment of Mn-bearing silicate minerals would
have negligible effect on the Mn grade of the zinc concen-
trate, especially for the pelitic ore where the majority of the
Mn is hosted in sphalerite (Figure 4b). EPMA composi-
tional analyses of sphalerite have also shown the occurrence
of different sphalerite populations at Gamsberg ore com-
prising Mn-rich and Mn-poor variants, and therefore the
potential strategy to manage concentrate quality is through
ROM ore blending – one of the focuses of geometallurgy.
To proactively plan ore blending strategies up front
knowledge of the ore variability across the extent of the
deposit is required. In this case, prior knowledge of the
sphalerite composition is needed. High quality compo-
sitional data on sphalerite composition can be generated
through EPMA analyses, but this can be a costly exer-
cise when applied to an entire drill hole database. Initial
attempts to calibrate the QEMSCAN to accurately mea-
sure Mn in sphalerite, did not generate the required con-
fidence levels, and so a strategy of applying data analysis
and machine learning was conceived (Price et al., 2023).
This strategy required a large training dataset, representa-
tive of the three main ore types of the Gamsberg North
sources. Two case studies are described in this paper to
showcase the key role of process mineralogy and geometal-
lurgy, one focusing on Zn and the other on the REE. Both
case studies are based on research at the University of Cape
Town.
ZINC CASE STUDY
Zn is a critical metal that finds application in green energy
battery storage, as well as longstanding use as a key com-
ponent in galvanizing steel thereby extending product life
span (Wang et al., 2021 Watari et al., 2021). The primary
source of zinc is from sphalerite which is recovered during
flotation to produce a concentrate before further refining,
most commonly through the roast leach electrowinning
process. The Gamsberg zinc deposit is a sediment-hosted,
strata-bound and metamorphosed sedimentary exhalative
Zn-Pb-Ag deposit in the Northern Cape Province in South
Africa. The first concentrate was produced from this ore in
2018 from the Gamsberg concentrator (Price et al., 2023).
Unlike most zinc ores, Gamsberg ore is characterized by
extensive Mn enrichment that is hosted in a variety of sili-
cate (garnet, pyroxene, pyroxmangite), oxide (pyrolusite)
and sulphide minerals (sphalerite, alabandite) (Schouwstra
et al., 2010). The distribution of Mn into this extensive
suite of minerals is consistent with its occurrence as a geo-
chemically abundant metal (Figure 2a) compared to zinc,
as a geochemically scarce metal (Figure 2b). The occurrence
of Mn substituting for Zn in the crystal lattice of sphaler-
ite, varying from almost pure sphalerite to Fe-Mn-rich vari-
ants with up to 20 wt.% Fe +Mn (McClung and Viljoen,
2011), represents an ongoing challenge at Gamsberg. This
case study describes a program of process mineralogical and
geometallurgical research focusing on finding solutions to
manage this challenge, some of which are described more
extensively in Price et al. (2023).
For any process mineralogy study, one of the main
objectives is to determine the characteristics of the valu-
able minerals, including understanding the distribution
of the valuable elements into these minerals, and their
associations whether in a run-of-mine ore, concentrate or
tailings. Three mineralized horizons, locally known as mag-
netite pyroxene (MPO), pyrrhotite-dominated pelitic ore
(PEO_PO) and pyrite-dominated pelitic ore (PEO_PY)
are combined to form the run of mine ore at Gamsberg.
An example of the mineralogy of these three ores is given
in Figure 4a. Sphalerite, pyrrhotite, pyrite, and quartz are
the dominant minerals in all three ore types. The MPO ore
also has significant garnet, pyroxene, pyroxenite and mag-
netite, whereas the pelitic ores have significant mica and
chlorite. As described by the nomenclature, the PEO_PO
ores have more pyrrhotite than pyrite, and vice versa for the
PEO_PY ores. The Mn distribution in these ores, as illus-
trated in Figure 4b, was determined through a combination
of auto-SEM-EDS (QEMSCAN) analysis to determine the
mineral grades and electron probe microanalysis (EPMA)
analysis to accurately determine the sphalerite composition.
The majority of the Mn in the MPO is hosted in garnet,
pyroxene and pyroxenoid (silicate minerals) with only neg-
ligible Mn hosted in sphalerite. Coupled with the higher
sphalerite grades in the MPO, this ore could be considered
the best ore. For the pelitic ores, however, up to 90% of the
Mn can be hosted in sphalerite with the balance of the Mn
occurring in garnet.
On flotation of this Zn ore, the final concentrate domi-
nantly comprises sphalerite, with minor galena, pyrite and
silicate minerals. The Mn grade of the final flotation con-
centrate, however, is a concern, since if it exceeds a cer-
tain threshold, refineries may impose treatment penalties
(the generally accepted range for Mn in zinc concentrate
is 0.02 to 0.8 wt.% Sinclair, 2005). Although the concen-
trate may also contain Fe derived from both sphalerite and
the Fe-sulphides (pyrite and pyrrhotite), the concentrate
quality specifications tend to be more lenient (the generally
accepted range for Fe in zinc concentrate is 1.5 to 10 wt.%
Sinclair, 2005). Because Mn occurs in the crystal lattice of
sphalerite, substituting for Zn through isomorphous sub-
stitution, it cannot be physically separated from a sphalerite
concentrate. Optimising the flotation circuit to minimize
the entrainment of Mn-bearing silicate minerals would
have negligible effect on the Mn grade of the zinc concen-
trate, especially for the pelitic ore where the majority of the
Mn is hosted in sphalerite (Figure 4b). EPMA composi-
tional analyses of sphalerite have also shown the occurrence
of different sphalerite populations at Gamsberg ore com-
prising Mn-rich and Mn-poor variants, and therefore the
potential strategy to manage concentrate quality is through
ROM ore blending – one of the focuses of geometallurgy.
To proactively plan ore blending strategies up front
knowledge of the ore variability across the extent of the
deposit is required. In this case, prior knowledge of the
sphalerite composition is needed. High quality compo-
sitional data on sphalerite composition can be generated
through EPMA analyses, but this can be a costly exer-
cise when applied to an entire drill hole database. Initial
attempts to calibrate the QEMSCAN to accurately mea-
sure Mn in sphalerite, did not generate the required con-
fidence levels, and so a strategy of applying data analysis
and machine learning was conceived (Price et al., 2023).
This strategy required a large training dataset, representa-
tive of the three main ore types of the Gamsberg North