1038 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
illustrates that the machine learning-based flotation model
achieves a close prediction of Cu recovery compared to the
actual values.
Flow Sheet Constraint Manager
IES offers the capability for users to optimise flow sheet
simulation results according to their defined parameters.
This functionality allows users to easily define optimisation
boundaries by dragging, dropping, and resizing a subset
box on the flow sheet. The equipment and streams within
the subset can provide attributes as parameters that can be
adjusted, as well as constraints that can be applied to the
optimisation function.
The Constraint Manager within IES enhances simula-
tion accuracy by allowing users to define operating ranges
for equipment settings as parameters and establish process-
ing targets as constraints. In the integrated flow sheet of the
mining operation, the Constraint Manager was utilised to
manage throughput potential. By setting limits on the SAG
mill load, cyclone pressure, and final grind size, the maxi-
mum achievable throughput was calculated under these
constrained conditions. This approach utilised the fine-
tuned flow sheet to deliver more realistic simulation results.
Flow Sheet Fine-Tuning
The flow sheet fine-tuning was carried out to address the
process variability observed in the Copper mining opera-
tion and develop a predictive flow sheet to handle future
variability. The variabilities in the operation are illustrated
in Table 2, showcasing the range of operational data. This
variability serves as a justification for fine-tuning the equip-
ment models within the flow sheet. By refining the models,
it was possible to enhance the predictability and accuracy of
the simulation result.
Figure 3 depicts the comparison between the pre-
dicted throughput generated by IES and the actual values
over fourteen months period. It is important to note that
plant throughput is influenced by various factors, some of
-20%
0%
20%
40%
60%
80%
100%
6000
6500
7000
7500
8000
8500
Error Actual IES
Figure 3. Actual and IES predicted plant throughput from Jan 2021 to Mar 2022
Table 1. Variation in operation data
Variable Min Max 15th Percentile PI Data 85th Percentile PI Data
Estimated SAG DWi Value 6.3 7.3 6.4 7.2
Estimated SAG Axb Value 38.0 43.7 38.4 43.0
BWi (kWh/t) 11.0 11.4 11.2 11.4
SAG Mill Feed %Solids 69.5 74.6 70.5 74.0
SAG Mill Fractional Speed 0.68 0.76 0.72 0.75
Throughput (t/h) 6358 7813 6611 7724
SAG Power (kW) 12034 14067 12966 13976
Total Ball Mill Power (kW) 48499 50141 48550 49859
Flotation Feed P80 (µm) 187 222 191 214
Error
(%)Throughput
(tph)
illustrates that the machine learning-based flotation model
achieves a close prediction of Cu recovery compared to the
actual values.
Flow Sheet Constraint Manager
IES offers the capability for users to optimise flow sheet
simulation results according to their defined parameters.
This functionality allows users to easily define optimisation
boundaries by dragging, dropping, and resizing a subset
box on the flow sheet. The equipment and streams within
the subset can provide attributes as parameters that can be
adjusted, as well as constraints that can be applied to the
optimisation function.
The Constraint Manager within IES enhances simula-
tion accuracy by allowing users to define operating ranges
for equipment settings as parameters and establish process-
ing targets as constraints. In the integrated flow sheet of the
mining operation, the Constraint Manager was utilised to
manage throughput potential. By setting limits on the SAG
mill load, cyclone pressure, and final grind size, the maxi-
mum achievable throughput was calculated under these
constrained conditions. This approach utilised the fine-
tuned flow sheet to deliver more realistic simulation results.
Flow Sheet Fine-Tuning
The flow sheet fine-tuning was carried out to address the
process variability observed in the Copper mining opera-
tion and develop a predictive flow sheet to handle future
variability. The variabilities in the operation are illustrated
in Table 2, showcasing the range of operational data. This
variability serves as a justification for fine-tuning the equip-
ment models within the flow sheet. By refining the models,
it was possible to enhance the predictability and accuracy of
the simulation result.
Figure 3 depicts the comparison between the pre-
dicted throughput generated by IES and the actual values
over fourteen months period. It is important to note that
plant throughput is influenced by various factors, some of
-20%
0%
20%
40%
60%
80%
100%
6000
6500
7000
7500
8000
8500
Error Actual IES
Figure 3. Actual and IES predicted plant throughput from Jan 2021 to Mar 2022
Table 1. Variation in operation data
Variable Min Max 15th Percentile PI Data 85th Percentile PI Data
Estimated SAG DWi Value 6.3 7.3 6.4 7.2
Estimated SAG Axb Value 38.0 43.7 38.4 43.0
BWi (kWh/t) 11.0 11.4 11.2 11.4
SAG Mill Feed %Solids 69.5 74.6 70.5 74.0
SAG Mill Fractional Speed 0.68 0.76 0.72 0.75
Throughput (t/h) 6358 7813 6611 7724
SAG Power (kW) 12034 14067 12966 13976
Total Ball Mill Power (kW) 48499 50141 48550 49859
Flotation Feed P80 (µm) 187 222 191 214
Error
(%)Throughput
(tph)