XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1129
consumption, but the focus and strategies can vary by mill
and process.
THE PIPELINE DIGITAL TWIN
This paper focuses on an application of a Digital Twin for
a copper mine’s tailings pipeline, where two pumping sta-
tions each have two parallel pipelines of multiple pumps in
series, which is represented by a block diagram in Figure 1.
Several fixed speed pumps and variable speed pumps
operate to pump the slurry from a thickener underflow
sump to a booster station, then from the booster station
to varying destinations around a tailings pond. The total
energy to pump a volume of slurry to a destination remains
the same, but the efficiency of the pumps can vary the
amount of power required. The control system has a loop
controlling the pump speeds to maintain a level setpoint in
the tanks, but splitting the load demanded from the pumps
to achieve that objective is set based on a bias adjusting the
speed settings. The Digital Twin strives to adjust that load
split for each set of parallel pipelines and minimize operator
intervention in adjusting this bias manually to achieve the
minimum power requirement for pumps.
The Digital Twin of the pumping system is developed
to provide the optimal speed setpoints to the pumping sys-
tem. The solution is developed in three different layers. The
first layer of the solution is developing a simulation model
using the existing equipment information. The second
layer is incorporating historical data to account for years of
wear and to follow the performance of the physical pumps
closely. The third layer optimizes the operation to minimize
power consumption.
IDEAS SIMULATION CAPABILITIES IN
THIS APPLICATION
A screenshot of a section of the model is shown below in
Figure 2 for further discussion.
The simulation capabilities needed for this application
are the high-fidelity pump and pipe network, and the solids
slurry characteristics.
The IDEAS model with a fully dynamic pump and
pipe network is defined by engineering data from a pro-
cess plant. Key parameters include but are not limited to:
elevations pipe dimensions, roughness, and detailed fitting
resistance pump flow versus head curves, power curves,
Figure 1. Problem statement block diagram
Figure 2. Screenshot of model section
consumption, but the focus and strategies can vary by mill
and process.
THE PIPELINE DIGITAL TWIN
This paper focuses on an application of a Digital Twin for
a copper mine’s tailings pipeline, where two pumping sta-
tions each have two parallel pipelines of multiple pumps in
series, which is represented by a block diagram in Figure 1.
Several fixed speed pumps and variable speed pumps
operate to pump the slurry from a thickener underflow
sump to a booster station, then from the booster station
to varying destinations around a tailings pond. The total
energy to pump a volume of slurry to a destination remains
the same, but the efficiency of the pumps can vary the
amount of power required. The control system has a loop
controlling the pump speeds to maintain a level setpoint in
the tanks, but splitting the load demanded from the pumps
to achieve that objective is set based on a bias adjusting the
speed settings. The Digital Twin strives to adjust that load
split for each set of parallel pipelines and minimize operator
intervention in adjusting this bias manually to achieve the
minimum power requirement for pumps.
The Digital Twin of the pumping system is developed
to provide the optimal speed setpoints to the pumping sys-
tem. The solution is developed in three different layers. The
first layer of the solution is developing a simulation model
using the existing equipment information. The second
layer is incorporating historical data to account for years of
wear and to follow the performance of the physical pumps
closely. The third layer optimizes the operation to minimize
power consumption.
IDEAS SIMULATION CAPABILITIES IN
THIS APPLICATION
A screenshot of a section of the model is shown below in
Figure 2 for further discussion.
The simulation capabilities needed for this application
are the high-fidelity pump and pipe network, and the solids
slurry characteristics.
The IDEAS model with a fully dynamic pump and
pipe network is defined by engineering data from a pro-
cess plant. Key parameters include but are not limited to:
elevations pipe dimensions, roughness, and detailed fitting
resistance pump flow versus head curves, power curves,
Figure 1. Problem statement block diagram
Figure 2. Screenshot of model section