448 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
METHODOLOGY
Ore Samples Studied
The lithium (Li) ore sample, originating from a spodumene
(mineral containing lithium) bearing pegmatite hard-rock
deposit and provided by Sayona Lithium (Sayona, 2023),
comprised approximately 300 kg of run-of-mine material.
This sample was processed by crushing to a size of 100% pass-
ing 4 inches and then classified using 1½- and ½-inch screens.
The resulting size fractions, ranging from 1½ to 4 inches
(–4.0/+1.5) and from ½ to 1½ inches (–1.5/+0.5), were
selected for the ore sorting tests.
For the scandium ore sample, sourced from
Imperial Mining Group Ltd.’s Crater Lake proj-
ect where the primary rock type is a massive syenite
(Imperial Mining Group, 2023), three drums of ore
material were received. These were initially sorted into
size fractions of nominally –4.0/+2.0, –2.0/+1.0, and
–1.0/+0.375 inches. The largest fraction (–4.0/+2.0) was
further screened to separate into +4.0 and –4.0/+2.0
sizes. Similarly, the smallest fraction (–1.0/+0.375)
underwent additional screening to produce –1.0/+0.5,
–0.5/+0.375, and –0.375-inch sizes. These steps were nec-
essary to match the ore sorter’s capability, which can sort
material with a top-size of approximately 4.0 inches and
a bottom-size of around 0.5 inch. Subsequently, the larger
fractions (+4.0 and –4.0/+2.0) were crushed and screened
to increase the quantity of material in the –2.0/+1.0 and
–1.0/+0.5 size ranges for use in sorting production tests.
Lastly, the REE ore sample was obtained from carbon-
atite rock at Commerce Resources Corp.’s Ashram deposit
(Commerce Resources Corp., 2023). Similar to the lithium
sample, about 300 kg of this ore sample was crushed to
100% passing 4 inches and then sorted using 3-, 1½-, and
½-inch screens. The size fractions selected for particle sort-
ing tests were 3 to 4 inches (–4.0/+3.0), 1½ to 3 inches
(–3.0/+1.5), and ½ to 1½ inches (–1.5/+0.5).
Particle Ore Sorting System
The particle ore sorter used in this study, housed at Corem,
is the Comex ® OCXR–1000 integrated X-ray and opti-
cal sorting system (Comex, 2023). This sorter is capable
of capturing images of rock particles transported on a belt
conveyor that is one meter wide, utilizing an array of four
sensors. The suite of sensors includes a dual-energy X-ray
transmission (DE-XRT) analyzer, a high-resolution vis-
ible light optical camera that discerns color using the RGB
spectrum at each pixel, and hyperspectral imaging (HSI)
facilitated by two infrared cameras covering the visible to
near-infrared (VIS-NIR 400–1000 nm range) and short-
wave infrared (SWIR 930–1700 nm range).
Designed for precise rock-by-rock separation, the
sorter is equipped with 110 air jet nozzles, which can effi-
ciently sort rock particles down to approximately 12.5 mm
(0.5 in.) in size, with the system accommodating top rock
sizes generally between 4 to 8 inches, depending on sensor
and ore type. The system allows for two distinct modes of
operation: static and dynamic tests. Static tests involve the
scanning of rock or core fragments without actual separa-
tion, aimed at assessing the heterogeneity of the samples
and formulating a calibration algorithm for each sensor.
Conversely, dynamic tests involve the active separation of
rocks utilizing the air jet nozzles on bulk rock samples. The
quantity of material for these tests can vary, ranging from
a few tens of rocks to large-scale production tests involving
30 to 5,000 kilograms of feed material.
Ore Sorting Tests
In the case studies presented, static imaging tests using the
four sensors available were first carried out on approxi-
mately 30 to 100 rocks, selected from different size fractions
of the ore samples prepared for sorting tests. Calibration
algorithms specific to a given sensor were developed for
each feed material, employing the Comex software pro-
vided with the sorter. Dynamic sorting tests were then
performed, wherein rocks were ejected using various com-
binations of sensors and measurement parameters. These
dynamic tests aimed to refine the sorting strategies using
the most appropriate sensors, for later application in pro-
duction tests undertaken with ~50–100 kg of feed material
from the selected size fractions.
For each ore sample, production tests were performed
on selected size fractions, ensuring the size ratio did not
exceed 3:1 between the largest and smallest particles within
any fraction. A pre-set series of successive separation runs
was executed during each production test, resulting in the
division of the original size fraction into various sorting
products, each corresponding to a specific set of conditions
and parameters of the sorting protocol employed.
Elemental and Mineralogical Analysis of Sorting
Products
Representative samples of each sorting product were
subjected to X-ray fluorescence (XRF) spectrometry or
inductively coupled plasma optical emission spectroscopy
(ICP-OES) for elemental composition analysis. Total con-
tents of carbon and sulfur were measured by a combustion
method with a LECO instrument (combustion at 1380 °C,
infrared detector). Rare earth elements (15 REEs: Ce, Dy,
Er, Eu, Gd, Ho, La, Lu, Nd, Pr, Sm, Tb, Tm, Y, Yb) concen-
trations were determined with inductively coupled plasma
METHODOLOGY
Ore Samples Studied
The lithium (Li) ore sample, originating from a spodumene
(mineral containing lithium) bearing pegmatite hard-rock
deposit and provided by Sayona Lithium (Sayona, 2023),
comprised approximately 300 kg of run-of-mine material.
This sample was processed by crushing to a size of 100% pass-
ing 4 inches and then classified using 1½- and ½-inch screens.
The resulting size fractions, ranging from 1½ to 4 inches
(–4.0/+1.5) and from ½ to 1½ inches (–1.5/+0.5), were
selected for the ore sorting tests.
For the scandium ore sample, sourced from
Imperial Mining Group Ltd.’s Crater Lake proj-
ect where the primary rock type is a massive syenite
(Imperial Mining Group, 2023), three drums of ore
material were received. These were initially sorted into
size fractions of nominally –4.0/+2.0, –2.0/+1.0, and
–1.0/+0.375 inches. The largest fraction (–4.0/+2.0) was
further screened to separate into +4.0 and –4.0/+2.0
sizes. Similarly, the smallest fraction (–1.0/+0.375)
underwent additional screening to produce –1.0/+0.5,
–0.5/+0.375, and –0.375-inch sizes. These steps were nec-
essary to match the ore sorter’s capability, which can sort
material with a top-size of approximately 4.0 inches and
a bottom-size of around 0.5 inch. Subsequently, the larger
fractions (+4.0 and –4.0/+2.0) were crushed and screened
to increase the quantity of material in the –2.0/+1.0 and
–1.0/+0.5 size ranges for use in sorting production tests.
Lastly, the REE ore sample was obtained from carbon-
atite rock at Commerce Resources Corp.’s Ashram deposit
(Commerce Resources Corp., 2023). Similar to the lithium
sample, about 300 kg of this ore sample was crushed to
100% passing 4 inches and then sorted using 3-, 1½-, and
½-inch screens. The size fractions selected for particle sort-
ing tests were 3 to 4 inches (–4.0/+3.0), 1½ to 3 inches
(–3.0/+1.5), and ½ to 1½ inches (–1.5/+0.5).
Particle Ore Sorting System
The particle ore sorter used in this study, housed at Corem,
is the Comex ® OCXR–1000 integrated X-ray and opti-
cal sorting system (Comex, 2023). This sorter is capable
of capturing images of rock particles transported on a belt
conveyor that is one meter wide, utilizing an array of four
sensors. The suite of sensors includes a dual-energy X-ray
transmission (DE-XRT) analyzer, a high-resolution vis-
ible light optical camera that discerns color using the RGB
spectrum at each pixel, and hyperspectral imaging (HSI)
facilitated by two infrared cameras covering the visible to
near-infrared (VIS-NIR 400–1000 nm range) and short-
wave infrared (SWIR 930–1700 nm range).
Designed for precise rock-by-rock separation, the
sorter is equipped with 110 air jet nozzles, which can effi-
ciently sort rock particles down to approximately 12.5 mm
(0.5 in.) in size, with the system accommodating top rock
sizes generally between 4 to 8 inches, depending on sensor
and ore type. The system allows for two distinct modes of
operation: static and dynamic tests. Static tests involve the
scanning of rock or core fragments without actual separa-
tion, aimed at assessing the heterogeneity of the samples
and formulating a calibration algorithm for each sensor.
Conversely, dynamic tests involve the active separation of
rocks utilizing the air jet nozzles on bulk rock samples. The
quantity of material for these tests can vary, ranging from
a few tens of rocks to large-scale production tests involving
30 to 5,000 kilograms of feed material.
Ore Sorting Tests
In the case studies presented, static imaging tests using the
four sensors available were first carried out on approxi-
mately 30 to 100 rocks, selected from different size fractions
of the ore samples prepared for sorting tests. Calibration
algorithms specific to a given sensor were developed for
each feed material, employing the Comex software pro-
vided with the sorter. Dynamic sorting tests were then
performed, wherein rocks were ejected using various com-
binations of sensors and measurement parameters. These
dynamic tests aimed to refine the sorting strategies using
the most appropriate sensors, for later application in pro-
duction tests undertaken with ~50–100 kg of feed material
from the selected size fractions.
For each ore sample, production tests were performed
on selected size fractions, ensuring the size ratio did not
exceed 3:1 between the largest and smallest particles within
any fraction. A pre-set series of successive separation runs
was executed during each production test, resulting in the
division of the original size fraction into various sorting
products, each corresponding to a specific set of conditions
and parameters of the sorting protocol employed.
Elemental and Mineralogical Analysis of Sorting
Products
Representative samples of each sorting product were
subjected to X-ray fluorescence (XRF) spectrometry or
inductively coupled plasma optical emission spectroscopy
(ICP-OES) for elemental composition analysis. Total con-
tents of carbon and sulfur were measured by a combustion
method with a LECO instrument (combustion at 1380 °C,
infrared detector). Rare earth elements (15 REEs: Ce, Dy,
Er, Eu, Gd, Ho, La, Lu, Nd, Pr, Sm, Tb, Tm, Y, Yb) concen-
trations were determined with inductively coupled plasma