XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3851
control hardware and software with a simulation model
of the power portion of the system and was configured to
enable real-time based testing of the existing configuration
at Detour Lake Mine. A MATLAB/SIMULINK based
model simulates the grid, the ACS6000 power portion and
the motors. The inputs of the model are the firing com-
mands to the IGCTs and the Main Circuit Breaker (MCB)
commands (close, open, trip). The outputs of the model are
the network voltage, the ARU and INU currents, encoder
feedback (speed and position), MCB feedback and the DC
link voltage.
The first step was to configure the simulator to match
the configuration of the drive at Detour Lake Mine and
to replicate the failure mode of the drive system during a
network transient event. The drive behaviour on site was
reproduced by changing the amplitude of the supply volt-
age (Dr. Leitz, D. 2020). The simulation results fit quite
well with the actual behaviour of the drive on site and the
results can be seen in Figure 7.
After setting up the simulator, it was possible to rapidly
simulate various settings and to visualize the reaction of the
drive system. Based on several tests, it was determined that
during a network transient, the motors continued to out-
put full power during a ridethrough event, and this demand
caused a reduction in the DC bus voltage that resulted in
an overcurrent situation when the network voltage would
overshoot during the recovery phase. This was confirmed
on site using the onboard dataloggers (Figure 8).
Fine Tuning the Drive System Using Simulator Results
and Implementation On-site
Based on multiple tests, the failure of the drive during cer-
tain network disturbance events could be attributed to a
combination of several factors. The following optimizations
were proposed and implemented to improve the perfor-
mance of the drive. Owing to the use of the simulator and
the ability to quickly prototype, deploy and observe the
effect of the changes, it was possible to implement several
changes at the same time. This was previously not possible
because the effect of every change had to be observed dur-
ing a transient event, which was time consuming due to the
random nature of these events.
1. The torque reference to the motors during a net-
work transient event was reduced which helped
reduce the stress on the intermediate DC bus volt-
age. The inertia of the mill would be used to coast
through such an event and in extreme cases, the
drive can brake the mill to keep the DC link stable
using regeneration.
2. The Proportional Integral Derivative (PID) con-
trollers of the ARU were tuned based on the results
from the simulators. Previously, the PID control-
lers were tuned to be slow which is desirable to
prevent the overreaction of the ARU that might
cause further network instability, but the slow
PID controller was not sufficiently fast enough
Figure 6. Transformer and ARU arrangements
control hardware and software with a simulation model
of the power portion of the system and was configured to
enable real-time based testing of the existing configuration
at Detour Lake Mine. A MATLAB/SIMULINK based
model simulates the grid, the ACS6000 power portion and
the motors. The inputs of the model are the firing com-
mands to the IGCTs and the Main Circuit Breaker (MCB)
commands (close, open, trip). The outputs of the model are
the network voltage, the ARU and INU currents, encoder
feedback (speed and position), MCB feedback and the DC
link voltage.
The first step was to configure the simulator to match
the configuration of the drive at Detour Lake Mine and
to replicate the failure mode of the drive system during a
network transient event. The drive behaviour on site was
reproduced by changing the amplitude of the supply volt-
age (Dr. Leitz, D. 2020). The simulation results fit quite
well with the actual behaviour of the drive on site and the
results can be seen in Figure 7.
After setting up the simulator, it was possible to rapidly
simulate various settings and to visualize the reaction of the
drive system. Based on several tests, it was determined that
during a network transient, the motors continued to out-
put full power during a ridethrough event, and this demand
caused a reduction in the DC bus voltage that resulted in
an overcurrent situation when the network voltage would
overshoot during the recovery phase. This was confirmed
on site using the onboard dataloggers (Figure 8).
Fine Tuning the Drive System Using Simulator Results
and Implementation On-site
Based on multiple tests, the failure of the drive during cer-
tain network disturbance events could be attributed to a
combination of several factors. The following optimizations
were proposed and implemented to improve the perfor-
mance of the drive. Owing to the use of the simulator and
the ability to quickly prototype, deploy and observe the
effect of the changes, it was possible to implement several
changes at the same time. This was previously not possible
because the effect of every change had to be observed dur-
ing a transient event, which was time consuming due to the
random nature of these events.
1. The torque reference to the motors during a net-
work transient event was reduced which helped
reduce the stress on the intermediate DC bus volt-
age. The inertia of the mill would be used to coast
through such an event and in extreme cases, the
drive can brake the mill to keep the DC link stable
using regeneration.
2. The Proportional Integral Derivative (PID) con-
trollers of the ARU were tuned based on the results
from the simulators. Previously, the PID control-
lers were tuned to be slow which is desirable to
prevent the overreaction of the ARU that might
cause further network instability, but the slow
PID controller was not sufficiently fast enough
Figure 6. Transformer and ARU arrangements