Abstract:

Ores are becoming extremely variable with mineralogy and hardness disturbing all processes. The low grade plants require large amounts of Energy and Water. The design production targets of a plant are hard to reach due to high variability of the ore. The synchronization between grinding/classification/flotation/water recovery process is a must. These low grade ores mineral processing plants need to run like a clock. A novel approach of using a Digital Twin was designed to maximize copper production and water recovery. Instead of the traditional metal recovery, a Net Metal Production Rate operational strategy is presented to determine the optimal mill throughput. Process data together with Machine Learning algorithms using Dynamic Grinding, Flotation, Tailing’s water separation models were implemented to allow for a synchronized processing of the ore. The rheological characteristics of the pulp enables the prediction of the pulp/paste boundaries to be respected. The new capabilities provided by the computational capabilities for advanced analytics running in Cloud enable remote access. This is novel approach will be described to process low grade ores with increase water recovery in a profitable and sustainable manner

Keywords: Energy and Water Management, Copper Production Maximization, Digital Twin Mineral Processing Plant, Machine Learning, Digital Transformation

Menu

Mr. Freddy Retamal