XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3739
size used. He simply states that Sylvanite showed decreased
power consumption per unit of –200 mesh produced. His
comments about Lake Shore include that they use a “com-
plicated method” to reach conclusions concerning power
versus capacity increases. He also states that International
Nickel increased power and capacity equally by going to
grate from overflow, and that McClelland increased circu-
lating load and decreased overgrinding. His overall conclu-
sions on the debate between high- and low-level discharges
are: 1. low discharges increase power draw at the same ball
loading (and grates can also facilitate higher ball charge
level) 2. despite a number of reports of efficiency benefits,
oddly, he does not concede that there is an efficiency ben-
efit, at least not in all cases.
Operating data from Lake Shore (Lake Shore Staff,
1940) included data on comparisons of the performance of
“tube” ball mills provided with “center” discharges and con-
verted to “low” grate discharges. Converting the 5' × 16'
tube mill from center (high level) to low discharge increased
power draw an average of 12% and “grinding capacity rat-
ing” by an average of 23%. For the 6' × 16' tube mill, power
draw increased 13% and capacity rating by 35%. The
ratios of capacity over power increases equate to efficiency
increases of 9% and 19% respectively. Size distribution data
provided was not readily useful for further in-depth analy-
sis. Because charge levels were tried to be maintained, but
not exactly monitored, it is not known exactly what part of
the power draw increases are due to the change in type of
discharge. As well, their mills were operated in a variety of
arrangements (as far as number of mills and discharge type
of primary and secondary units), so their “grinding capacity
rating” calculations are somewhat ambiguous.
The Sylvanite Gold plant in Ontario (Taggart, 1945)
reported testing of 3 secondary grinding “tube mills” in
parallel, no. 1 open-ended with a full grate (no trunnion),
but that end of the mill on rollers no. 2 overflow (trun-
nion) discharge and no. 3 with a grate discharge and pulp
lifters through a trunnion opening. The data are summa-
rized in Table 1.
These results are striking. The ground product totals
are born out by feeds and product size distributions of the
combined parallel circuits. However, it is odd that there
is no explanation how they assigned the relative circuit
tonnages -we must presume that was done rigorously via
mass flow measurements.
Latchireddi et al. (2015) ran comparative (overflow
vs. open grate discharge) wet, limestone grinding tests in a
0.41m diameter by 0.56m long (16" × 22") ball mill with a
30% (107 kg) graded ball charge. Various %solids and mill
speeds were tested. The main conclusions were as follows.
• Grate discharge is more grinding efficient.
• The slurry hold-up in the grate mill was about a fifth
of that in the overflow mill. Compared to the sta-
tionary ball charge voids, the grate mill slurry filling
varied from about 50% to 100%, while the overflow
slurry filling was about 270%.
• From liquid tracing tests, there is a lesser degree of
back mixing in the grate discharge mill, but not
strikingly so. Percent solids was a significant factor
regarding mixing. Short-circuiting of feed to dis-
charge is less with the grate discharge.
• The grate discharge grinding performance is more
sensitive to percent solids and speed than is that of
the overflow mill.
Recently, Latchireddi et al. (2023) presented successful
operation of numerous ball mills installed at copper, iron,
and gold operations after converting overflow ball mills to
open-ended grate mills.
A summary of quantified plant efficiency comparison
is given in Table 2.
BATCH SLURRY LOAD TESTING
In order to test the effects that operating at different slurry
levels within the ball mill has on mill power draw and grind-
ing performance, a series of tests were conducted on splits
of a ball mill feed sample using the Metcom torque mill
at Midland research laboratory. The torque mill, shown in
Figure 1, is 24" diameter and 8" long, and operates at 65%
of critical speed.
The initial (baseline) test, from a series of media sizing
tests, was run for 6 minutes with a slurry filling of 100%
as it occupies 100% of the voids in a stationary ball charge.
Six additional tests were run for the same period of time
(6.0 minutes) on the same ratios of (74.7%) solids and
water, but at volumes from 37.5% to 300% of the initial
Table 1. Data from Sylvanite Gold plant
No.1: Open Grate No.2: Overflow No3 Grate/Trunnion
Tons through 200m per HP 0.59 0.47 0.53
CSE @200m 40% 50% 42%
Mill Gr. Rate t new 200m/HP-h 1.50 0.96 1.26
size used. He simply states that Sylvanite showed decreased
power consumption per unit of –200 mesh produced. His
comments about Lake Shore include that they use a “com-
plicated method” to reach conclusions concerning power
versus capacity increases. He also states that International
Nickel increased power and capacity equally by going to
grate from overflow, and that McClelland increased circu-
lating load and decreased overgrinding. His overall conclu-
sions on the debate between high- and low-level discharges
are: 1. low discharges increase power draw at the same ball
loading (and grates can also facilitate higher ball charge
level) 2. despite a number of reports of efficiency benefits,
oddly, he does not concede that there is an efficiency ben-
efit, at least not in all cases.
Operating data from Lake Shore (Lake Shore Staff,
1940) included data on comparisons of the performance of
“tube” ball mills provided with “center” discharges and con-
verted to “low” grate discharges. Converting the 5' × 16'
tube mill from center (high level) to low discharge increased
power draw an average of 12% and “grinding capacity rat-
ing” by an average of 23%. For the 6' × 16' tube mill, power
draw increased 13% and capacity rating by 35%. The
ratios of capacity over power increases equate to efficiency
increases of 9% and 19% respectively. Size distribution data
provided was not readily useful for further in-depth analy-
sis. Because charge levels were tried to be maintained, but
not exactly monitored, it is not known exactly what part of
the power draw increases are due to the change in type of
discharge. As well, their mills were operated in a variety of
arrangements (as far as number of mills and discharge type
of primary and secondary units), so their “grinding capacity
rating” calculations are somewhat ambiguous.
The Sylvanite Gold plant in Ontario (Taggart, 1945)
reported testing of 3 secondary grinding “tube mills” in
parallel, no. 1 open-ended with a full grate (no trunnion),
but that end of the mill on rollers no. 2 overflow (trun-
nion) discharge and no. 3 with a grate discharge and pulp
lifters through a trunnion opening. The data are summa-
rized in Table 1.
These results are striking. The ground product totals
are born out by feeds and product size distributions of the
combined parallel circuits. However, it is odd that there
is no explanation how they assigned the relative circuit
tonnages -we must presume that was done rigorously via
mass flow measurements.
Latchireddi et al. (2015) ran comparative (overflow
vs. open grate discharge) wet, limestone grinding tests in a
0.41m diameter by 0.56m long (16" × 22") ball mill with a
30% (107 kg) graded ball charge. Various %solids and mill
speeds were tested. The main conclusions were as follows.
• Grate discharge is more grinding efficient.
• The slurry hold-up in the grate mill was about a fifth
of that in the overflow mill. Compared to the sta-
tionary ball charge voids, the grate mill slurry filling
varied from about 50% to 100%, while the overflow
slurry filling was about 270%.
• From liquid tracing tests, there is a lesser degree of
back mixing in the grate discharge mill, but not
strikingly so. Percent solids was a significant factor
regarding mixing. Short-circuiting of feed to dis-
charge is less with the grate discharge.
• The grate discharge grinding performance is more
sensitive to percent solids and speed than is that of
the overflow mill.
Recently, Latchireddi et al. (2023) presented successful
operation of numerous ball mills installed at copper, iron,
and gold operations after converting overflow ball mills to
open-ended grate mills.
A summary of quantified plant efficiency comparison
is given in Table 2.
BATCH SLURRY LOAD TESTING
In order to test the effects that operating at different slurry
levels within the ball mill has on mill power draw and grind-
ing performance, a series of tests were conducted on splits
of a ball mill feed sample using the Metcom torque mill
at Midland research laboratory. The torque mill, shown in
Figure 1, is 24" diameter and 8" long, and operates at 65%
of critical speed.
The initial (baseline) test, from a series of media sizing
tests, was run for 6 minutes with a slurry filling of 100%
as it occupies 100% of the voids in a stationary ball charge.
Six additional tests were run for the same period of time
(6.0 minutes) on the same ratios of (74.7%) solids and
water, but at volumes from 37.5% to 300% of the initial
Table 1. Data from Sylvanite Gold plant
No.1: Open Grate No.2: Overflow No3 Grate/Trunnion
Tons through 200m per HP 0.59 0.47 0.53
CSE @200m 40% 50% 42%
Mill Gr. Rate t new 200m/HP-h 1.50 0.96 1.26