1116
Maximizing Copper Production by Proper Water Management
Using a Digital Twin
Osvaldo A. Bascur
OSB Digital, LLC.
ABSTRACT: Low grade ores mineral processing plants require large amounts of energy and water to operate
in a sustainable and profitable state. These ores present large variations in their mineralogy, metal content and
hardness. These low-grade ore plants are processing mostly rock in the first part of the process, followed by the
traditional mineral processing and water recovery systems. Currently, mineral processing plants operate in silos
and lack the necessary integration of data from mining, grinding, classification, flotation, thickening and tailings
processing. The lack of process information with the right degree of detailed are missing for understanding the
integrated process. Today, tight coordination between the mining product, grinding, classification, flotation and
water recovery processing is a must.
A novel strategy using a Digital Twin was designed to increase the necessary water recovery from the thickeners
and tailing ponds to maximize the copper production rate in a low-grade ore industrial plant. The correct shape
of the grinding particle size distribution are monitored to improve both for flotation metal production rate and
flocculation of the tails produced in the rougher flotation. The plant data model consists of a Rock Processing,
a Water Processing which integrates the plant to find the best operating conditions that optimize the copper
production rate base on the plant schedule.
The implementation of the Digital Twin results are: a 40% increase in water recovery for a maximization of cop-
per production rate of 32%. These savings are very significant based on a zero-capital investment requirements
with using their actual process historian data infrastructure. The OSB Digital Twin is based on the implementa-
tion of an integrated mineral processing plant model built using the PI System and OSB Grinding, Flotation
and Thickening dynamic simulators. The critical process operational modes are calculated based on the plant
business plan and current data to transform the timeseries raw data into information to build the necessary
machine learning models. These models are used to understand the integrated behavior of the plant and to avoid
violating costly process constraints in the grinding, classification, flotation and thickening processes.
Keywords: Digital Twins, Maximizing Net Metal Production Rate, Integrated Plant Operating Modeling,
Water Management, Predictive Analytics, Cloud Computing, Dynamic Process Models
Maximizing Copper Production by Proper Water Management
Using a Digital Twin
Osvaldo A. Bascur
OSB Digital, LLC.
ABSTRACT: Low grade ores mineral processing plants require large amounts of energy and water to operate
in a sustainable and profitable state. These ores present large variations in their mineralogy, metal content and
hardness. These low-grade ore plants are processing mostly rock in the first part of the process, followed by the
traditional mineral processing and water recovery systems. Currently, mineral processing plants operate in silos
and lack the necessary integration of data from mining, grinding, classification, flotation, thickening and tailings
processing. The lack of process information with the right degree of detailed are missing for understanding the
integrated process. Today, tight coordination between the mining product, grinding, classification, flotation and
water recovery processing is a must.
A novel strategy using a Digital Twin was designed to increase the necessary water recovery from the thickeners
and tailing ponds to maximize the copper production rate in a low-grade ore industrial plant. The correct shape
of the grinding particle size distribution are monitored to improve both for flotation metal production rate and
flocculation of the tails produced in the rougher flotation. The plant data model consists of a Rock Processing,
a Water Processing which integrates the plant to find the best operating conditions that optimize the copper
production rate base on the plant schedule.
The implementation of the Digital Twin results are: a 40% increase in water recovery for a maximization of cop-
per production rate of 32%. These savings are very significant based on a zero-capital investment requirements
with using their actual process historian data infrastructure. The OSB Digital Twin is based on the implementa-
tion of an integrated mineral processing plant model built using the PI System and OSB Grinding, Flotation
and Thickening dynamic simulators. The critical process operational modes are calculated based on the plant
business plan and current data to transform the timeseries raw data into information to build the necessary
machine learning models. These models are used to understand the integrated behavior of the plant and to avoid
violating costly process constraints in the grinding, classification, flotation and thickening processes.
Keywords: Digital Twins, Maximizing Net Metal Production Rate, Integrated Plant Operating Modeling,
Water Management, Predictive Analytics, Cloud Computing, Dynamic Process Models