XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3383
finer grinding (decreasing P80), no significant improve-
ment in recovery was observed. One potential explanation
for this lies in the viscosity of the slurry. Serpentine minerals
are prone to overgrinding, generating substantial fines that
significantly increase slurry viscosity (Senior and Thomas
2005, Uddin, et al. 2012). This viscosity increase, evident
by the paste behavior of the slurry during flotation, could
hinder awaruite flotation efficiency, limiting further recov-
ery gains despite finer grinding. While viscosity seems like a
strong factor, which has been reported for other ultramafic
deposits, exploring other potential contributors is impor-
tant to develop a better understanding of factors that affect
recovery. A 20% feed solids concentration was used for all
these experiments, without the addition of any dispersant.
In terms of nickel grade, the cumulative grade of the four
concentrates was over 1%, indicating an upgrade ratio of
approximately 5 (feed grade was 0.22%). The recoveries
achieved for this sample (roughly 60%) align well with
awaruite liberation studies presented by the authors in pre-
vious studies(Seiler, Sánchez and Bradshaw, et al. 2022). A
cleaner flotation test performed on the rougher concentrate
for the 90 mm (P80) particle size feed showed that by using
only flotation, it is possible to generate a concentrate with
a nickel grade of over 40%. It is noteworthy that the con-
centrate contained low concentrations of penalty elements
(Seiler, Sánchez and Pawlik, et al. 2023).
As previously mentioned, the significant drawback of
these weakly acidic conditions is the high acid consump-
tion, reaching 34.8 kg/t for the P80 of 60 mm grind size.
This represents an acid consumption that is 28% higher
compared to the 150 mm grind size (27.2 kg/t). The acid
consumptions seem to directly correlate to the increased
surface area of serpentine minerals available for acid reac-
tion at finer grinds (Uddin, et al. 2012). While finer grind-
ing did not significantly improve nickel recovery (Figure 2),
it comes at a substantial cost. The measured acid consump-
tion for the 90 mm grind size was 31.7 kg/t, and the extra
acid consumption on the cleaner test for this sample was
1.7 kg/t.
Figure 3 shows the nickel recovery and grade obtained
for Samples B and C for different grind sizes. Similar size
distributions were achieved for both samples. As a refer-
ence, P80 values of 140, 70, and 49 mm were measured for
10, and 20, and 30 min grind times, respectively. Nickel
recoveries of over 80% were achieved for Sample B, mean-
ing that most of the nickel that is recovered in the magnetic
separation is floatable, highlighting the potential synergy
between these separation processes. Figure 3 also dem-
onstrates that a simple magnetic separation and flotation
flowsheet achieves significantly higher nickel grades than
typically reported for nickel sulfide deposits. For instance,
the cleaner flotation concentrate of Sample B at a P80 grind
Figure 3. Nickel grade and recovery for Samples B (LIMS) and C (MIMS) at different
grinding times
finer grinding (decreasing P80), no significant improve-
ment in recovery was observed. One potential explanation
for this lies in the viscosity of the slurry. Serpentine minerals
are prone to overgrinding, generating substantial fines that
significantly increase slurry viscosity (Senior and Thomas
2005, Uddin, et al. 2012). This viscosity increase, evident
by the paste behavior of the slurry during flotation, could
hinder awaruite flotation efficiency, limiting further recov-
ery gains despite finer grinding. While viscosity seems like a
strong factor, which has been reported for other ultramafic
deposits, exploring other potential contributors is impor-
tant to develop a better understanding of factors that affect
recovery. A 20% feed solids concentration was used for all
these experiments, without the addition of any dispersant.
In terms of nickel grade, the cumulative grade of the four
concentrates was over 1%, indicating an upgrade ratio of
approximately 5 (feed grade was 0.22%). The recoveries
achieved for this sample (roughly 60%) align well with
awaruite liberation studies presented by the authors in pre-
vious studies(Seiler, Sánchez and Bradshaw, et al. 2022). A
cleaner flotation test performed on the rougher concentrate
for the 90 mm (P80) particle size feed showed that by using
only flotation, it is possible to generate a concentrate with
a nickel grade of over 40%. It is noteworthy that the con-
centrate contained low concentrations of penalty elements
(Seiler, Sánchez and Pawlik, et al. 2023).
As previously mentioned, the significant drawback of
these weakly acidic conditions is the high acid consump-
tion, reaching 34.8 kg/t for the P80 of 60 mm grind size.
This represents an acid consumption that is 28% higher
compared to the 150 mm grind size (27.2 kg/t). The acid
consumptions seem to directly correlate to the increased
surface area of serpentine minerals available for acid reac-
tion at finer grinds (Uddin, et al. 2012). While finer grind-
ing did not significantly improve nickel recovery (Figure 2),
it comes at a substantial cost. The measured acid consump-
tion for the 90 mm grind size was 31.7 kg/t, and the extra
acid consumption on the cleaner test for this sample was
1.7 kg/t.
Figure 3 shows the nickel recovery and grade obtained
for Samples B and C for different grind sizes. Similar size
distributions were achieved for both samples. As a refer-
ence, P80 values of 140, 70, and 49 mm were measured for
10, and 20, and 30 min grind times, respectively. Nickel
recoveries of over 80% were achieved for Sample B, mean-
ing that most of the nickel that is recovered in the magnetic
separation is floatable, highlighting the potential synergy
between these separation processes. Figure 3 also dem-
onstrates that a simple magnetic separation and flotation
flowsheet achieves significantly higher nickel grades than
typically reported for nickel sulfide deposits. For instance,
the cleaner flotation concentrate of Sample B at a P80 grind
Figure 3. Nickel grade and recovery for Samples B (LIMS) and C (MIMS) at different
grinding times