XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3897
Crusher and HPGR Model Performance
Figure 2 presents the mapped capacities for the two crush-
ers (CR01 and CR02) and the HPGR based on the equip-
ment data. Two-point mapping was performed to represent
the performance regions for the equipment. The crushers’
CSS set-points can be adjusted within the mapped perfor-
mance regions. The HPGR capacity limits are changed by
changing the speed and adjusting the gap using the two
settings shown in Figure 2. Since the HPGR is placed near
the exit of the final product, the capacity is limited by the
supplied capacity from the undersize of the screen S01. The
model performance shown in Figure 2 is limited by the val-
idation of the actual equipment performance and is subject
to change with changing equipment type and size.
Comminution Plant
B01
CV02
SP1
S -Screen
CR -Crusher
HPGR -High Pressure Grinding Rolls
CV -Conveyor
B -Bin
F -Feeder
SP – Stockpile
ES – Edge Splitter
CT – Control
Material
Feed
CV01
CV03
CV04
CV05
CV06
CV07
CV08
B02 B03 B04
S01
SP2
F1
CR01 CR02
HPGR01
CV09 CV10
CV11 CV12
CV13
CV14
CV15
CV16
CV17
ES01
CT01
CT02 CT03 CT04
Figure 1. Comminution plant for optimizing the bin sizing using dynamic process simulation
(a) (b)
5 10 15 20 25 30 35 40 45 50 55
Setting [mm]
400
500
600
700
800
900
Model Capacities for Settings
CR01 Capacity vs CSS
CR02 Capacity vs CSS
HPGR01 Capacity vs Gap
1 2 4 8 16 32 64 128 250
Sieve Size [mm]
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Model PSD Response for Settings
CR01 Feed
CR01 Prod CSS 51
CR01 Prod CSS 46
CR02 Feed
CR02 Prod CSS 22
CR02 Prod CSS 19
HPGR01 Feed1
HPGR01 Prod1
HPGR01 Feed2
HPGR01 Prod2
Figure 2. Mapped a) capacities and b) PSDs for two crushers (CR01 and CR02) and HPGR based on selected equipment data
Model
Throughput
[tph]
Cummulative
Fraction
Passing
[-]

Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3897
Crusher and HPGR Model Performance
Figure 2 presents the mapped capacities for the two crush-
ers (CR01 and CR02) and the HPGR based on the equip-
ment data. Two-point mapping was performed to represent
the performance regions for the equipment. The crushers’
CSS set-points can be adjusted within the mapped perfor-
mance regions. The HPGR capacity limits are changed by
changing the speed and adjusting the gap using the two
settings shown in Figure 2. Since the HPGR is placed near
the exit of the final product, the capacity is limited by the
supplied capacity from the undersize of the screen S01. The
model performance shown in Figure 2 is limited by the val-
idation of the actual equipment performance and is subject
to change with changing equipment type and size.
Comminution Plant
B01
CV02
SP1
S -Screen
CR -Crusher
HPGR -High Pressure Grinding Rolls
CV -Conveyor
B -Bin
F -Feeder
SP – Stockpile
ES – Edge Splitter
CT – Control
Material
Feed
CV01
CV03
CV04
CV05
CV06
CV07
CV08
B02 B03 B04
S01
SP2
F1
CR01 CR02
HPGR01
CV09 CV10
CV11 CV12
CV13
CV14
CV15
CV16
CV17
ES01
CT01
CT02 CT03 CT04
Figure 1. Comminution plant for optimizing the bin sizing using dynamic process simulation
(a) (b)
5 10 15 20 25 30 35 40 45 50 55
Setting [mm]
400
500
600
700
800
900
Model Capacities for Settings
CR01 Capacity vs CSS
CR02 Capacity vs CSS
HPGR01 Capacity vs Gap
1 2 4 8 16 32 64 128 250
Sieve Size [mm]
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Model PSD Response for Settings
CR01 Feed
CR01 Prod CSS 51
CR01 Prod CSS 46
CR02 Feed
CR02 Prod CSS 22
CR02 Prod CSS 19
HPGR01 Feed1
HPGR01 Prod1
HPGR01 Feed2
HPGR01 Prod2
Figure 2. Mapped a) capacities and b) PSDs for two crushers (CR01 and CR02) and HPGR based on selected equipment data
Model
Throughput
[tph]
Cummulative
Fraction
Passing
[-]

Help

Destination page number Search scope Search Text

loading

3898 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
Baseline Process Evaluation
A baseline process simulation is generated to establish
the nominal performance of the circuit. Table 1 pres-
ents the conditioning of the baseline circuit simula-
tion, wherein the four bins (B01-B04) capacities are
referred to as unit values representing the nominal value.
All the mass flows in the circuit are normalized to refer-
ence values. Figure 3 presents the baseline simulation
results for mass flow at different points of the circuit
with both crushers operating at upper limits of CSSs
(50 mm and 20 mm) and the screen deck-1 is set to
40 mm. The process is simulated for 24 hours of opera-
tions considering one startup sequence of the process and
all initial conditions of bins as empty at the material level.
After the initial startup sequence, the process reaches into a
dynamic production operation. It can be observed that in
this simulation the CR01 is fully maximized, while CR02
is under-utilized due to the insufficient supply of mate-
rial caused by the interlocks in process operation. Due to
the varying operation of the two crushers together with
the control system response for fresh feed adjustment, the
screen experiences fluctuating loading during the opera-
tion. These fluctuations in the undersize of the screening
process create the varying feed for HPGR.
Figure 4 presents different bin levels and corresponding
interlock signals generated for SP1. In this design setup, the
capacity of bins B01 and B04 are not bottlenecks, mean-
ing, they do not trigger the interlock signal for SP1. The
first bottleneck signal was observed to be dependent on the
crusher CR01 mass flow as the level in the B02 overshoots
periodically, causing the interlock in SP1. On the other
hand, bin B03 is underutilized as the crusher CR02 does
not receive the required mass flow to run at its maximum
utilization. This also results in the bin level in B03 staying
Table 1. Baseline process variable setup for the bin sizing study
B01 Level SP 50% S01 Deck-1 40–50 mm
B02 Level SP 70% S01 Deck-2 20 mm
B03 Level SP 70% CR01 CSS 45–50 mm
B04 Level SP 70% CR02 CSS 16–20 mm
CT01-CT04 PI controller Interlock SP1 80% in any bin (On/Off)
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
Circuit Feed, Avg =0.79 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
S01 Feed, Avg =0.72 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
CR01 Product, Avg =0.97 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
CR02 Product, Avg =0.82 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
HPGR01 Product, Avg =0.88 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
Edge Splitter Recirculation, Avg =0.43 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5 Circuit Product, Avg =0.93 [-]
Figure 3. Normalized reference simulation results for mass flow at different points of the circuit
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]

Previous Page Next Page

Extracted Text (may have errors)

3898 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
Baseline Process Evaluation
A baseline process simulation is generated to establish
the nominal performance of the circuit. Table 1 pres-
ents the conditioning of the baseline circuit simula-
tion, wherein the four bins (B01-B04) capacities are
referred to as unit values representing the nominal value.
All the mass flows in the circuit are normalized to refer-
ence values. Figure 3 presents the baseline simulation
results for mass flow at different points of the circuit
with both crushers operating at upper limits of CSSs
(50 mm and 20 mm) and the screen deck-1 is set to
40 mm. The process is simulated for 24 hours of opera-
tions considering one startup sequence of the process and
all initial conditions of bins as empty at the material level.
After the initial startup sequence, the process reaches into a
dynamic production operation. It can be observed that in
this simulation the CR01 is fully maximized, while CR02
is under-utilized due to the insufficient supply of mate-
rial caused by the interlocks in process operation. Due to
the varying operation of the two crushers together with
the control system response for fresh feed adjustment, the
screen experiences fluctuating loading during the opera-
tion. These fluctuations in the undersize of the screening
process create the varying feed for HPGR.
Figure 4 presents different bin levels and corresponding
interlock signals generated for SP1. In this design setup, the
capacity of bins B01 and B04 are not bottlenecks, mean-
ing, they do not trigger the interlock signal for SP1. The
first bottleneck signal was observed to be dependent on the
crusher CR01 mass flow as the level in the B02 overshoots
periodically, causing the interlock in SP1. On the other
hand, bin B03 is underutilized as the crusher CR02 does
not receive the required mass flow to run at its maximum
utilization. This also results in the bin level in B03 staying
Table 1. Baseline process variable setup for the bin sizing study
B01 Level SP 50% S01 Deck-1 40–50 mm
B02 Level SP 70% S01 Deck-2 20 mm
B03 Level SP 70% CR01 CSS 45–50 mm
B04 Level SP 70% CR02 CSS 16–20 mm
CT01-CT04 PI controller Interlock SP1 80% in any bin (On/Off)
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
Circuit Feed, Avg =0.79 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
S01 Feed, Avg =0.72 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
CR01 Product, Avg =0.97 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
CR02 Product, Avg =0.82 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
HPGR01 Product, Avg =0.88 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5
Edge Splitter Recirculation, Avg =0.43 [-]
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Time [s] 10 4
0
0.5
1
1.5 Circuit Product, Avg =0.93 [-]
Figure 3. Normalized reference simulation results for mass flow at different points of the circuit
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]
Mass
Flow
[-]

3896 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
change in the time response of the buffer capacity such as
bin. These are time-varying operational issues which occur
in many circuits but often are limited at the design phase
of the comminution circuit using steady-state simulation.
In contrast, dynamic simulation capability provides the
opportunity to study the simultaneous effects of equipment
selection and design, buffer choice and control system
response for designing a new circuit layout. In particular,
this paper presents a methodology of integrating dynamic
process simulation for the design of a suitable size bin for
complex circuit configuration. The goal of the circuit is to
maximize overall process performance while minimizing
the size of bin choice in the circuit. A case study consisting
of a conceptual circuit with multiple crushers and HPGR
is used to demonstrate the design process.
DYNAMIC SIMULATION AND
MODELLING OF CRUSHING PLANT
The dynamic process modelling is capable of simulating
time-dependent interaction and effects of a crushing plant,
mimicking the real-time behaviour of the circuit perfor-
mance in production. The dynamic process modelling
approach can capture discrete and gradual changes occur-
ring in plant operation, such as delays in material flow,
start-up sequence, discrete events, and wear. Each equip-
ment model is based on the mathematical description of
mass m and properties γ in a derivative form with respect to
time, as shown in Equations 1 and 2. (Asbjörnsson, 2015
Sbárbaro &del Villar, 2010)
((t) (t)) dt
dm(t) m m
,in j,out =-o o (1)
(
(
()
(
dt
d
m t)
m t)
t
t)
where ,in
,in
n
h
c1(t)W
c
c(t) c c(t)), c(t) =-=
o
R
T
S
S
S
V
X
W
W
(2)
The dynamic simulation used in the research is based on
the previous development in the MATLAB/Simulink envi-
ronment (Asbjörnsson et al., 2013). The material flow is
regulated using interlocks and regulatory controllers (e.g.,
On/Off, PI and PID.). Bins, splitters, and stockpiles are
modelled using the perfect-mix principle while the con-
veyors are modelled using the state-space model represent-
ing the material delay units in dynamic process simulation
(Asbjörnsson, 2015 Bhadani et al., 2021). For a system
with multiple parallel flows, recirculating mass, larger lag
times, partial integration of different stages and a mismatch
between manipulated variables and control variables, sta-
bility can be harder to achieve (Asbjörnsson et al., 2022
Dorf &Bishop, 2016). The product size distribution
(PSD) of the crusher and HPGR model is backfitted using
equipment manufacturers’ data to a simplified population
balance model using the Whiten selection function and
Reid–Stewart breakage function (Bhadani et al., 2023
Duarte et al., 2021 Napier-Munn et al., 1996). The capaci-
ties are also mapped to the manufacturers’ data. The screen
model and edge splitter model are based on the Whiten
expression using a reduced partition curve (Bhadani et al.,
2021 Napier-Munn et al., 1996).
CASE STUDY FOR BIN SIZING
APPLICATION
The selection of equipment, control philosophy, and bin
sizing simultaneously play a role in generating suitable
conceptual design. Figure 1 presents the layout of the
comminution plant under study using dynamic process
simulations. The process represents the segment after the
products from the primary crusher in stockpile 1 (SP1).
The SP1 feed together with the products of the two crush-
ers (CR01 – secondary crusher), tertiary crusher (CR02)
and recirculation by edge splitter (ES01) are fed to the bin
(B01). The bin (B01) is regulated to maintain the level set-
point to feed the two parallel screens (S01) using a PI con-
troller. The oversize of the screen (+40 mm) is conveyed to
bin B02 as secondary crusher feed. The middle fraction size
(20–40 mm) is fed to bin B03 as a tertiary crusher feed. The
undersize fraction (0–20 mm) is conveyed to the bin (B04)
as HPGR feed. The process circuit is inspired by previous
research and industrial relevance (Asbjörnsson et al., 2016).
The two crushers CR01 and CR02 have the opera-
tional setpoints of closed side setting (CSS) 50 mm and
CSS 20 mm, respectively. The bins B02, B03 and B04 are
controlled using PI controllers. A safety interlock is added
to the stockpile SP1 in connection to the four bin levels
(B01-B04) exceeding 80% of the total capacity of any bins.
The nominal capacity of the 4 bins is based on the reference
design (nominal capacity referred to a value of 1) and the
challenge here is to study the plant performance based on
changing bin capacities. The key performance indicator for
this study is mass flow (throughput) at different points in
the circuit, crushers and HPGR utilization, control triggers
due to interlocks and dynamic behaviour. The models used
for KPIs are based on the previous development (Bhadani
et al., 2020). The process simulation for this case does not
include the effects of the discrete changes in the system
such as discrete failure, run-of-mine truck frequency, mate-
rial characteristics change, capacity influence of conveyors,
and wear in crushers and screens.

Previous Page Next Page

Extracted Text (may have errors)

3896 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
change in the time response of the buffer capacity such as
bin. These are time-varying operational issues which occur
in many circuits but often are limited at the design phase
of the comminution circuit using steady-state simulation.
In contrast, dynamic simulation capability provides the
opportunity to study the simultaneous effects of equipment
selection and design, buffer choice and control system
response for designing a new circuit layout. In particular,
this paper presents a methodology of integrating dynamic
process simulation for the design of a suitable size bin for
complex circuit configuration. The goal of the circuit is to
maximize overall process performance while minimizing
the size of bin choice in the circuit. A case study consisting
of a conceptual circuit with multiple crushers and HPGR
is used to demonstrate the design process.
DYNAMIC SIMULATION AND
MODELLING OF CRUSHING PLANT
The dynamic process modelling is capable of simulating
time-dependent interaction and effects of a crushing plant,
mimicking the real-time behaviour of the circuit perfor-
mance in production. The dynamic process modelling
approach can capture discrete and gradual changes occur-
ring in plant operation, such as delays in material flow,
start-up sequence, discrete events, and wear. Each equip-
ment model is based on the mathematical description of
mass m and properties γ in a derivative form with respect to
time, as shown in Equations 1 and 2. (Asbjörnsson, 2015
Sbárbaro &del Villar, 2010)
((t) (t)) dt
dm(t) m m
,in j,out =-o o (1)
(
(
()
(
dt
d
m t)
m t)
t
t)
where ,in
,in
n
h
c1(t)W
c
c(t) c c(t)), c(t) =-=
o
R
T
S
S
S
V
X
W
W
(2)
The dynamic simulation used in the research is based on
the previous development in the MATLAB/Simulink envi-
ronment (Asbjörnsson et al., 2013). The material flow is
regulated using interlocks and regulatory controllers (e.g.,
On/Off, PI and PID.). Bins, splitters, and stockpiles are
modelled using the perfect-mix principle while the con-
veyors are modelled using the state-space model represent-
ing the material delay units in dynamic process simulation
(Asbjörnsson, 2015 Bhadani et al., 2021). For a system
with multiple parallel flows, recirculating mass, larger lag
times, partial integration of different stages and a mismatch
between manipulated variables and control variables, sta-
bility can be harder to achieve (Asbjörnsson et al., 2022
Dorf &Bishop, 2016). The product size distribution
(PSD) of the crusher and HPGR model is backfitted using
equipment manufacturers’ data to a simplified population
balance model using the Whiten selection function and
Reid–Stewart breakage function (Bhadani et al., 2023
Duarte et al., 2021 Napier-Munn et al., 1996). The capaci-
ties are also mapped to the manufacturers’ data. The screen
model and edge splitter model are based on the Whiten
expression using a reduced partition curve (Bhadani et al.,
2021 Napier-Munn et al., 1996).
CASE STUDY FOR BIN SIZING
APPLICATION
The selection of equipment, control philosophy, and bin
sizing simultaneously play a role in generating suitable
conceptual design. Figure 1 presents the layout of the
comminution plant under study using dynamic process
simulations. The process represents the segment after the
products from the primary crusher in stockpile 1 (SP1).
The SP1 feed together with the products of the two crush-
ers (CR01 – secondary crusher), tertiary crusher (CR02)
and recirculation by edge splitter (ES01) are fed to the bin
(B01). The bin (B01) is regulated to maintain the level set-
point to feed the two parallel screens (S01) using a PI con-
troller. The oversize of the screen (+40 mm) is conveyed to
bin B02 as secondary crusher feed. The middle fraction size
(20–40 mm) is fed to bin B03 as a tertiary crusher feed. The
undersize fraction (0–20 mm) is conveyed to the bin (B04)
as HPGR feed. The process circuit is inspired by previous
research and industrial relevance (Asbjörnsson et al., 2016).
The two crushers CR01 and CR02 have the opera-
tional setpoints of closed side setting (CSS) 50 mm and
CSS 20 mm, respectively. The bins B02, B03 and B04 are
controlled using PI controllers. A safety interlock is added
to the stockpile SP1 in connection to the four bin levels
(B01-B04) exceeding 80% of the total capacity of any bins.
The nominal capacity of the 4 bins is based on the reference
design (nominal capacity referred to a value of 1) and the
challenge here is to study the plant performance based on
changing bin capacities. The key performance indicator for
this study is mass flow (throughput) at different points in
the circuit, crushers and HPGR utilization, control triggers
due to interlocks and dynamic behaviour. The models used
for KPIs are based on the previous development (Bhadani
et al., 2020). The process simulation for this case does not
include the effects of the discrete changes in the system
such as discrete failure, run-of-mine truck frequency, mate-
rial characteristics change, capacity influence of conveyors,
and wear in crushers and screens.