Simulation of sensor-based sorting based on drill core analysis

Sensor-based sorting has been integrated into more and more applications in mineral processing over the past decades. Various sensor types, such as line scan cameras, near infrared sensors and X-ray transmission sensors, allow the separation of a vast variety of raw materials. Sensor-based sorting has several economic benefits which can make the difference between an uneconomic operation and a profitable mine. Sensor-based sorters can be applied to fulfill various sorting tasks, such as waste rejection, product quality improvement and ore type diversion, and have a significant impact on the layout of the processing plant. Therefore, it is important to include sensor-based sorting into the planning of a processing plant at an early stage. To evaluate the applicability of sensor-based sorting for any specific ore type, it is currently necessary to provide several hundred kilograms of material for subsequent bulk tests. In exploration, ore samples are usually rare and only available in form of drill cores. The acquisition of drill cores is time-consuming and expensive especially for deposits in great depth. Hence, it is unfavorable to spoil valuable drill cores for bulk tests. This paper describes a methodology to evaluate the applicability of sensor-based sorting based on drill core analysis. That methodology offers the possibility to calculate the achievable performances of different sensor types for various grain sizes without comminution of drill cores. The general methodology is shown on basis of an exemplary test work on drill cores from a currently conducted exploration.