3708 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
concentrator in Peru (Medina et al. 2023): “The ball mills
draw very high power, with both mills operating above
90% of the total available power. However, the Line 1 ball
mill is drawing higher power than Line 2 ball mill. Both
mills were operating with a wide variation in power draw,
which could be a result of variation of ball charge levels,
given that the ball mills, unlike SAG mills, are operated with
fixed speed drive of 75% critical.”
Furthermore, the paper states that “The current 80%
passing (P80) product from the grinding circuit is between
110 and 190 microns (μm), with a design of P80 106 μm
and a current target of P80 160 μm..” This coarser ball mill
circuit product causes a loss of recovery as stated in the
same paper: “Preliminary analysis indicates that, on aver-
age, rougher copper recovery drops by around 0.3% for every
10 μm increase in P80.” Totally about 1.6% recovery loss,
only because the target P80 cannot be achieved. A con-
stant maximum ball mill power draw would have been
beneficial for the achieved copper recovery. In case the
ball mills would have been variable speed driven, then by
continuously adapting the mill speed slightly within a cer-
tain operating range, the mill power draw could be kept
more constant at maximum level, even with fluctuating
ball charge and independent of the mill liners status (new,
half-worn, worn). Consequently, the average ball mill cir-
cuit P80 value would have been slightly finer resulting in a
higher recovery and lower CO2 emissions per ton of con-
centrate. Unfortunately, the Constancia ball mills are fixed
speed, meaning that an enormous amount of energy has
been applied to reduce large pieces of ore-rock in several
steps to fine particles with a P80 of only 160 μm, instead
of the original optimal design P80 of 106 μm. Therefore, a
certain fraction of coarse particles containing non-liberated
target minerals are not recovered in the rougher flotation
and sent to the tailings (recovery loss).
At these high throughput rates, variable speed on the
ball mills would not have been able to achieve the orig-
inal target P80 of 106 μm, since the installed power on
the mills is too low. Interestingly the original design phi-
losophy for Constantia considered that 16 MW installed
power for each ball mill would soon become the bottle-
neck (Lane et al., 2015): “The optimum design had a slight
bias towards more power on the ball mills than for the SAG
mills, but 16 MW mills were selected for the SAG and ball
mills, thus standardizing the motors to the same 8 MW
model that would facilitate spares and interchange ability
to reduce working capital and risk.” Often the installed ball
mill power is around 5 to 10% higher than the required
mill power draw at the pinions for design conditions. But
Klohn, Stephenson and Granados (2016) reported that
the Constancia ball mill specific energy is 9.7 kWh/t for
the target P80 of 106 μm, resulting into a copper recov-
ery of 90%. So based upon the original design throughput
capacity of 76’000 tpd a mill power draw at the pinions of
each ball mill of 15’358 kW would be required. With two
16’000 kW gearless mill drives this would still have been
just enough with only a small 4% margin, but also a bit
more installed power would have been possible. However,
fixed high-speed dual pinion motors were selected and with
Source: Medina et al. 2023
Figure 8. Wide range of ball mill power draw at Constancia
concentrator in Peru (Medina et al. 2023): “The ball mills
draw very high power, with both mills operating above
90% of the total available power. However, the Line 1 ball
mill is drawing higher power than Line 2 ball mill. Both
mills were operating with a wide variation in power draw,
which could be a result of variation of ball charge levels,
given that the ball mills, unlike SAG mills, are operated with
fixed speed drive of 75% critical.”
Furthermore, the paper states that “The current 80%
passing (P80) product from the grinding circuit is between
110 and 190 microns (μm), with a design of P80 106 μm
and a current target of P80 160 μm..” This coarser ball mill
circuit product causes a loss of recovery as stated in the
same paper: “Preliminary analysis indicates that, on aver-
age, rougher copper recovery drops by around 0.3% for every
10 μm increase in P80.” Totally about 1.6% recovery loss,
only because the target P80 cannot be achieved. A con-
stant maximum ball mill power draw would have been
beneficial for the achieved copper recovery. In case the
ball mills would have been variable speed driven, then by
continuously adapting the mill speed slightly within a cer-
tain operating range, the mill power draw could be kept
more constant at maximum level, even with fluctuating
ball charge and independent of the mill liners status (new,
half-worn, worn). Consequently, the average ball mill cir-
cuit P80 value would have been slightly finer resulting in a
higher recovery and lower CO2 emissions per ton of con-
centrate. Unfortunately, the Constancia ball mills are fixed
speed, meaning that an enormous amount of energy has
been applied to reduce large pieces of ore-rock in several
steps to fine particles with a P80 of only 160 μm, instead
of the original optimal design P80 of 106 μm. Therefore, a
certain fraction of coarse particles containing non-liberated
target minerals are not recovered in the rougher flotation
and sent to the tailings (recovery loss).
At these high throughput rates, variable speed on the
ball mills would not have been able to achieve the orig-
inal target P80 of 106 μm, since the installed power on
the mills is too low. Interestingly the original design phi-
losophy for Constantia considered that 16 MW installed
power for each ball mill would soon become the bottle-
neck (Lane et al., 2015): “The optimum design had a slight
bias towards more power on the ball mills than for the SAG
mills, but 16 MW mills were selected for the SAG and ball
mills, thus standardizing the motors to the same 8 MW
model that would facilitate spares and interchange ability
to reduce working capital and risk.” Often the installed ball
mill power is around 5 to 10% higher than the required
mill power draw at the pinions for design conditions. But
Klohn, Stephenson and Granados (2016) reported that
the Constancia ball mill specific energy is 9.7 kWh/t for
the target P80 of 106 μm, resulting into a copper recov-
ery of 90%. So based upon the original design throughput
capacity of 76’000 tpd a mill power draw at the pinions of
each ball mill of 15’358 kW would be required. With two
16’000 kW gearless mill drives this would still have been
just enough with only a small 4% margin, but also a bit
more installed power would have been possible. However,
fixed high-speed dual pinion motors were selected and with
Source: Medina et al. 2023
Figure 8. Wide range of ball mill power draw at Constancia