1128
Pipeline Digital Twin with Power Optimization
Renee Roesel, Timothy Towers, Tanaji Mali
ANDRITZ Inc, Alpharetta, GA
ABSTRACT: The online Digital Twin Technology in the industrial arena has evolved as an important building
block of today’s Digital Transformation, capturing the combined power of process simulations and real-time
plant data to create a virtual industrial plant environment. The Digital Twin represents a digital replica of the
physical process using asset-specific engineering information. The Digital Twin models run parallel to the actual
plant, evaluating the current state of the process and providing valuable process insights, helping operations
with proactive decision-making and improved operational visibility. The online Digital Twin provides many
real-world benefits, including virtual instruments (soft sensors), data validation, anomaly detection, what-if
analysis, and setpoints for optimal operation.
This paper details an application of a Digital Twin for a copper mine’s tailings pipeline. There are two pump-
ing stations at this mining operation, each with two parallel pipelines utilizing pumps in series. This Digital
Twin can predict the power consumption of each pump based on the running conditions (flow rate, %solids,
pressure, pump speed, etc.) and has an optimizing layer to evaluate power consumption at various running
conditions. Before the Digital Twin, the distribution of flow between the parallel lines was generally set at a
constant load split which was occasionally changed by control room operators. The addition of the Digital Twin
minimizes the power consumption by varying the load distribution between the two lines.
INTRODUCTION
The implementation of digital solutions to guide or actively
seek optimal operating conditions is on the rise through all
industries, and the mining industry has plenty of challenges
that could be overcome using such solutions. The IDEAS ™
simulator from ANDRITZ brings years of experience
simulating mining processes, with a number of papers pro-
viding evidence on IDEAS’ capabilities for mining (from
Parathasarathi, Szaruga, and Szatkowski 2009 to Schug
and McGarry 2016 and Tripathi, et al 2016). During the
development of high-fidelity mining simulations and a
long history of connecting to an offline copy of the con-
trol system for realistic operator training came the desire
to use the insight built into the simulation model to aid
operations. For the IDEAS Digital Twins at ANDRITZ,
the term Digital Twin is not just a simulation model that
closely predicts the operation’s performance, but one that
directly uses real-time plant data in a process model and
leverages the IDEAS model’s information and decision-
making capabilities. A number of discussions of Digital
Twins using IDEAS software are available, discussing vir-
tual instruments which use the simulation model to pro-
vide process values instead of physical transmitters (Schug,
et al 2019 and Sepulveda, et al 2019), to different Digital
Twin strategies for the mining industry (Cristoffanini, et al
2020 and Nazari and McGarry 2018).
However, each process can have different priorities
that lead to the development of different Digital Twins.
One common goal for most mills is minimizing power
Pipeline Digital Twin with Power Optimization
Renee Roesel, Timothy Towers, Tanaji Mali
ANDRITZ Inc, Alpharetta, GA
ABSTRACT: The online Digital Twin Technology in the industrial arena has evolved as an important building
block of today’s Digital Transformation, capturing the combined power of process simulations and real-time
plant data to create a virtual industrial plant environment. The Digital Twin represents a digital replica of the
physical process using asset-specific engineering information. The Digital Twin models run parallel to the actual
plant, evaluating the current state of the process and providing valuable process insights, helping operations
with proactive decision-making and improved operational visibility. The online Digital Twin provides many
real-world benefits, including virtual instruments (soft sensors), data validation, anomaly detection, what-if
analysis, and setpoints for optimal operation.
This paper details an application of a Digital Twin for a copper mine’s tailings pipeline. There are two pump-
ing stations at this mining operation, each with two parallel pipelines utilizing pumps in series. This Digital
Twin can predict the power consumption of each pump based on the running conditions (flow rate, %solids,
pressure, pump speed, etc.) and has an optimizing layer to evaluate power consumption at various running
conditions. Before the Digital Twin, the distribution of flow between the parallel lines was generally set at a
constant load split which was occasionally changed by control room operators. The addition of the Digital Twin
minimizes the power consumption by varying the load distribution between the two lines.
INTRODUCTION
The implementation of digital solutions to guide or actively
seek optimal operating conditions is on the rise through all
industries, and the mining industry has plenty of challenges
that could be overcome using such solutions. The IDEAS ™
simulator from ANDRITZ brings years of experience
simulating mining processes, with a number of papers pro-
viding evidence on IDEAS’ capabilities for mining (from
Parathasarathi, Szaruga, and Szatkowski 2009 to Schug
and McGarry 2016 and Tripathi, et al 2016). During the
development of high-fidelity mining simulations and a
long history of connecting to an offline copy of the con-
trol system for realistic operator training came the desire
to use the insight built into the simulation model to aid
operations. For the IDEAS Digital Twins at ANDRITZ,
the term Digital Twin is not just a simulation model that
closely predicts the operation’s performance, but one that
directly uses real-time plant data in a process model and
leverages the IDEAS model’s information and decision-
making capabilities. A number of discussions of Digital
Twins using IDEAS software are available, discussing vir-
tual instruments which use the simulation model to pro-
vide process values instead of physical transmitters (Schug,
et al 2019 and Sepulveda, et al 2019), to different Digital
Twin strategies for the mining industry (Cristoffanini, et al
2020 and Nazari and McGarry 2018).
However, each process can have different priorities
that lead to the development of different Digital Twins.
One common goal for most mills is minimizing power