1234 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
rates for 200 °C without causing damage but sug-
gested heating the feed up to higher temperatures
to flip the hastingsite to the conductor state, cool
and pass the material through the HTR at a lower
temperature, preferably 100 to 150 °C to stay within
normal operating limits.
• Treating more fines—significant REE values are
lost in the minus 45 µm fraction (28.3%), but this
is not practical to recover with electrostatic separa-
tion due to materials handling difficulties. However,
if it is possible to scalp at 20 µm whilst still reject-
ing the troublesome ultra fines this should result in
a recovery improvement. Conversely, grinding to a
coarser primary particle size of 80% passing 250 to
500 µm will reduce fines content and allow higher
allanite recovery. If sufficient hastingsite is rejected in
this stage, the non-conductors can then be reground
and WHIMS used for final cleaning. This approach
would be ideal for small scale operations, coupled
with dry primary HPGR grinding, to avoid having
to dry the HTR feed first. However, for the likely
larger scale operation needed to achieve good proj-
ect economics, HTR would likely need to follow
WHIMS to reduce the mass being treated, discussed
in the next section.
FLOWSHEET DEVELOPMENT
A flowsheet is proposed which exploits the coarse liberation
characteristics of the ore whilst addressing the issue of co-
enrichment of hastingsite with allanite.:
demonstrating superior Ti grade and recovery at 140 °C
compared with the baseline of 100 °C.
A sub-sample of cleaner WHIMS magnetics was dry
screened at 45 µm fraction and the coarse fraction sub-
jected to testing in a High Tension Roll (HTR) at 100 and
200 °C feed temperatures to determine if an effective sepa-
ration of allanite from hastingsite could be achieved. Both
minerals are non-conductors at lower temperatures but
it was thought that hastingsite may switch to conductive
behaviour at higher temperatures. Middlings were retreated
through the machine twice, which yielded a non-conductor
value stream TREE+Y grade of 5.0% for a recovery
of 27.1% relative to WHIMS magnetics. Final mid-
dlings graded lower than stage feed at 1.21% and conduc-
tors (rejects) were 0.73%, comprising 40.1% of test feed.
Photographs of the various products for the three pass
HTR test are presented as Figure 13.
The colour difference between the non-conductors
(brown) and conductors (black) were very distinct, indicat-
ing a significant mineral separation has occurred. Middlings
were still black in colour though partially receptive to sepa-
ration as conductors and may benefit from regrinding to
liberate the allanite content and allow it to pass to the non-
conductors product.
These results were seen as extremely encouraging, with
further optimisation work planned:
• Operation at a higher temperature—Testing at
higher feed temperatures will help identify if further
improvements in separation are possible. Mineral
Technologies advised that their HTR units are only
Figure 13. Electrostatic Testing products for three pass run
rates for 200 °C without causing damage but sug-
gested heating the feed up to higher temperatures
to flip the hastingsite to the conductor state, cool
and pass the material through the HTR at a lower
temperature, preferably 100 to 150 °C to stay within
normal operating limits.
• Treating more fines—significant REE values are
lost in the minus 45 µm fraction (28.3%), but this
is not practical to recover with electrostatic separa-
tion due to materials handling difficulties. However,
if it is possible to scalp at 20 µm whilst still reject-
ing the troublesome ultra fines this should result in
a recovery improvement. Conversely, grinding to a
coarser primary particle size of 80% passing 250 to
500 µm will reduce fines content and allow higher
allanite recovery. If sufficient hastingsite is rejected in
this stage, the non-conductors can then be reground
and WHIMS used for final cleaning. This approach
would be ideal for small scale operations, coupled
with dry primary HPGR grinding, to avoid having
to dry the HTR feed first. However, for the likely
larger scale operation needed to achieve good proj-
ect economics, HTR would likely need to follow
WHIMS to reduce the mass being treated, discussed
in the next section.
FLOWSHEET DEVELOPMENT
A flowsheet is proposed which exploits the coarse liberation
characteristics of the ore whilst addressing the issue of co-
enrichment of hastingsite with allanite.:
demonstrating superior Ti grade and recovery at 140 °C
compared with the baseline of 100 °C.
A sub-sample of cleaner WHIMS magnetics was dry
screened at 45 µm fraction and the coarse fraction sub-
jected to testing in a High Tension Roll (HTR) at 100 and
200 °C feed temperatures to determine if an effective sepa-
ration of allanite from hastingsite could be achieved. Both
minerals are non-conductors at lower temperatures but
it was thought that hastingsite may switch to conductive
behaviour at higher temperatures. Middlings were retreated
through the machine twice, which yielded a non-conductor
value stream TREE+Y grade of 5.0% for a recovery
of 27.1% relative to WHIMS magnetics. Final mid-
dlings graded lower than stage feed at 1.21% and conduc-
tors (rejects) were 0.73%, comprising 40.1% of test feed.
Photographs of the various products for the three pass
HTR test are presented as Figure 13.
The colour difference between the non-conductors
(brown) and conductors (black) were very distinct, indicat-
ing a significant mineral separation has occurred. Middlings
were still black in colour though partially receptive to sepa-
ration as conductors and may benefit from regrinding to
liberate the allanite content and allow it to pass to the non-
conductors product.
These results were seen as extremely encouraging, with
further optimisation work planned:
• Operation at a higher temperature—Testing at
higher feed temperatures will help identify if further
improvements in separation are possible. Mineral
Technologies advised that their HTR units are only
Figure 13. Electrostatic Testing products for three pass run