Abstract
In the quest for operational efficiency and maximized recovery rates, mines are increasingly utilizing sensors throughout the material flow to further increase mine-to-mill optimization. By leveraging advancements in AI and machine learning, real-time material characterization data—such as that generated by X-Ray fluorescence sensors deployed in digger buckets at the mine face or on conveyor belts—enables new, data-driven blending strategies. These strategies reduce grade variability, enhance grade targets, and optimize flotation processes, ultimately boosting overall recovery.
Material classification sensors deployed in-bucket on loading equipment play a crucial role by providing real-time, precise grade measurements directly at the mine face on a per-bucket scale. Traditional ore control methods rely heavily on estimations based on mine polygons, which often result in discrepancies between expected and actual ore grades due to blast movement and mixing, problems in positioning systems, and poor mine network conditions affecting fleet management connectivity. By obtaining a grade prediction for the material loaded into each truck, in-bucket XRF sensing enables more accurate and informed decision-making regarding material routing and stockpile management.
Complementing initial bulk ore sorting at the mine face, sensors installed on conveyor belts offer continuous monitoring of the material conveyed from the crusher throughout the grinding circuit. This real-time data is not just collected, but immediately fed back to the mine, allowing for immediate adjustments in truck routing and extraction points to ensure grade stability in the material sent to the plant. The instant feedback loop established by the conveyor belt ore composition sensors is essential for mitigating the inherent variability in mining operations and for fine-tuning blending strategies on the fly, giving immediate control over the process.
Improved blending is achieved by strategically utilizing high-grade stockpiles. Rather than rapidly feeding high-grade material, which can cause undesired high variability in the process and consequent metal losses, the material is introduced slowly and consistently. This approach minimizes fluctuations and maintains a steady feed grade for the plant. As a result, the grade targets can be met more reliably and with less variability, and the flotation process operates under optimal conditions with minimum perturbations, becoming more stable and efficient. This enhanced efficiency directly correlates with higher recovery rates, as the process can be better optimized to the consistent feed characteristics made possible through advanced blending strategies.
Implementing these technologies within mining operations involves integrating data and establishing feedback mechanisms between the mine and the mill. Case studies will demonstrate the tangible benefits achieved, underscoring the critical role of real-time data and precise material characterization in modern mining practices. The findings highlight how MineSense's technologies are pivotal in enhancing operational efficiency and maximizing resource recovery.