11
how liner wear affects IA overtime. Figure 9, 38 ft. SAG
mill, illustrates how IA can be used to observe the overall
effectiveness of increasing mill speed from new to worn-out
liners. Specifically, IA indicates when the liners are worn-
out, there no longer throw and the mill has reached a point
where increasing mill speed does not improve grinding and
instead wastes energy. Figure 10, a fixed-speed 30 ft. AG
mill, shows how IA is a function of slurry volumetric fill
level and liner wear. This allows for the opportunity to opti-
mize the mill by aligning the toe (FTA) of the slurry with
the actual impact angle (IA), as these are both measured
AAM signals. Figure 11, 32 ft. SAG, illustrates how the
Impact Angle can be maintained automatically for optimal
grind using mill speed. It also illustrates the point where
the mill has reached maximum speed and therefore can no
longer be used to keep IA at the optimal real-time operat-
ing point.
Advanced Analytics Measurements (AAM) can also
be used for real-time detection and measurement of both
optimal and sub-efficient operating cases. Figure 4 provides
examples of this with three different cases: Grinding media
overthrowing the slurry, media underthrowing the slurry
and optimal operation. Figure 4 also shows these three
modes of operation in relation with the power curves.
Of the two cases, overthrowing and underthrowing,
overthrowing is the absolute worst operating condition
for SAG and BM as it can significantly increase the bro-
ken liner &grinding media occurrences, peening, peg-
ging, and blinding. Thus, to reduce the amount of time
in overthrowing operation and maintain optimal mill
operation, Advanced Analytics Measurements known as
Liner Damage Level (LDL) and Impact on Toe (IOT)
are derived and monitored in real-time. It has been found
that overthrowing typically occurs when Liner Damage
Level (LDL) 70 and Impact on Toe (IOT) 15 as shown
in Figure 13. Also shown is this figure are times when the
mill is overfilled or overloaded resulting from underthrow-
ing operation. For this condition, LDL 30 and IOT –15
is typically observed. Thus, LDL and IOT can be used in
real-time to operate the mill in a more optimal manner in-
between overthrowing and underthrowing conditions. By
monitoring these two AAM variables, an operator or con-
trol system can quickly change course to keep LDL ~50
and IOT~0. By staying out of overthrowing operation, we
provide a safer environment for maintenance crews, reduce
shutdown time and increase the actual amount of time in
normal production. By reducing the amount of time in
underthrowing operation, we avoid sudden throughput
reductions (to avoid further problems) and improve power
consumption, safety and refrain from operating the mill in
an overfilled or overloaded manner. Optimal operation has
the benefit both achieving maximum throughput at low
energy consumption rates, and it is the safest operating
point for both personnel and equipment.
AAM can also be used to generate a reliable method
for estimating the relationship between the ball filling level
(Jb) and the total charge (Jc). Specifically, the AAM signal
Jb/Jc shown in Figure 14, can be used to estimate the right
amount of ore relative to the current mill ball charge fur-
ther optimizing a mill’s performance and efficiency.
In closing, the 1967 British statistician George Box
wrote, “All models are wrong, but some are useful.” Thus,
while Advanced Analytics Measurements (FLEN, LDL, IA,
FTA, IOT, and Jb/Jc) are not perfect models, they are very
useful as they represent a real-time snapshot of the process
and enable the implementation of improved process con-
trol strategies. These strategies include better methods of
manual mill operation as well as automatic operation with
throughput, mill speed, and/or water control to pave the
path for achieving Operational Excellence on every rota-
tion of the mill.
REFERENCES
CIM Academy — Nunez, E. 2022. Ten Advanced Process
Control (APC) Lessons from the Plant. https://academy
.cim.org/cim/2021/webinars/350690/eduardo.nunez
.ten.advanced.process.control.28apc29.lessons.from
.the.plant.html?f=listing%3D4%2Abrowseby%3D
8%2Asortby%3D2%2Amedia%3D1%2Aspeaker
%3D761488.
Gugel, K. 2015. Optimal SAG mill control using vibra-
tion &digital signal processing techniques, SAG
Conference, Vancouver, Canada.
Nunez, E and Baron, J. 2022. Teck’s HVC Real Time
Vibration Profile to Optimize SAG Mill Performance,
54th Annual Canadian Mineral Processors Operator
Conference &CIM Academy, ID367337, Vancouver,
Canada.
Powell, M. and Mainza, A. 2006. Extended grinding curves
are essential to the comparison of milling performance.
Minerals Engineering 19(2006):1487–1494.
how liner wear affects IA overtime. Figure 9, 38 ft. SAG
mill, illustrates how IA can be used to observe the overall
effectiveness of increasing mill speed from new to worn-out
liners. Specifically, IA indicates when the liners are worn-
out, there no longer throw and the mill has reached a point
where increasing mill speed does not improve grinding and
instead wastes energy. Figure 10, a fixed-speed 30 ft. AG
mill, shows how IA is a function of slurry volumetric fill
level and liner wear. This allows for the opportunity to opti-
mize the mill by aligning the toe (FTA) of the slurry with
the actual impact angle (IA), as these are both measured
AAM signals. Figure 11, 32 ft. SAG, illustrates how the
Impact Angle can be maintained automatically for optimal
grind using mill speed. It also illustrates the point where
the mill has reached maximum speed and therefore can no
longer be used to keep IA at the optimal real-time operat-
ing point.
Advanced Analytics Measurements (AAM) can also
be used for real-time detection and measurement of both
optimal and sub-efficient operating cases. Figure 4 provides
examples of this with three different cases: Grinding media
overthrowing the slurry, media underthrowing the slurry
and optimal operation. Figure 4 also shows these three
modes of operation in relation with the power curves.
Of the two cases, overthrowing and underthrowing,
overthrowing is the absolute worst operating condition
for SAG and BM as it can significantly increase the bro-
ken liner &grinding media occurrences, peening, peg-
ging, and blinding. Thus, to reduce the amount of time
in overthrowing operation and maintain optimal mill
operation, Advanced Analytics Measurements known as
Liner Damage Level (LDL) and Impact on Toe (IOT)
are derived and monitored in real-time. It has been found
that overthrowing typically occurs when Liner Damage
Level (LDL) 70 and Impact on Toe (IOT) 15 as shown
in Figure 13. Also shown is this figure are times when the
mill is overfilled or overloaded resulting from underthrow-
ing operation. For this condition, LDL 30 and IOT –15
is typically observed. Thus, LDL and IOT can be used in
real-time to operate the mill in a more optimal manner in-
between overthrowing and underthrowing conditions. By
monitoring these two AAM variables, an operator or con-
trol system can quickly change course to keep LDL ~50
and IOT~0. By staying out of overthrowing operation, we
provide a safer environment for maintenance crews, reduce
shutdown time and increase the actual amount of time in
normal production. By reducing the amount of time in
underthrowing operation, we avoid sudden throughput
reductions (to avoid further problems) and improve power
consumption, safety and refrain from operating the mill in
an overfilled or overloaded manner. Optimal operation has
the benefit both achieving maximum throughput at low
energy consumption rates, and it is the safest operating
point for both personnel and equipment.
AAM can also be used to generate a reliable method
for estimating the relationship between the ball filling level
(Jb) and the total charge (Jc). Specifically, the AAM signal
Jb/Jc shown in Figure 14, can be used to estimate the right
amount of ore relative to the current mill ball charge fur-
ther optimizing a mill’s performance and efficiency.
In closing, the 1967 British statistician George Box
wrote, “All models are wrong, but some are useful.” Thus,
while Advanced Analytics Measurements (FLEN, LDL, IA,
FTA, IOT, and Jb/Jc) are not perfect models, they are very
useful as they represent a real-time snapshot of the process
and enable the implementation of improved process con-
trol strategies. These strategies include better methods of
manual mill operation as well as automatic operation with
throughput, mill speed, and/or water control to pave the
path for achieving Operational Excellence on every rota-
tion of the mill.
REFERENCES
CIM Academy — Nunez, E. 2022. Ten Advanced Process
Control (APC) Lessons from the Plant. https://academy
.cim.org/cim/2021/webinars/350690/eduardo.nunez
.ten.advanced.process.control.28apc29.lessons.from
.the.plant.html?f=listing%3D4%2Abrowseby%3D
8%2Asortby%3D2%2Amedia%3D1%2Aspeaker
%3D761488.
Gugel, K. 2015. Optimal SAG mill control using vibra-
tion &digital signal processing techniques, SAG
Conference, Vancouver, Canada.
Nunez, E and Baron, J. 2022. Teck’s HVC Real Time
Vibration Profile to Optimize SAG Mill Performance,
54th Annual Canadian Mineral Processors Operator
Conference &CIM Academy, ID367337, Vancouver,
Canada.
Powell, M. and Mainza, A. 2006. Extended grinding curves
are essential to the comparison of milling performance.
Minerals Engineering 19(2006):1487–1494.