It is generally acknowledged that bulk ore sorting is an attractive prospect for the mining industry. However, uncertainties relating to the operational performance of sensor technologies, the grade heterogeneity present in analysed material and the physical separation performance of the coupled system complicate the evaluation and deployment of bulk ore sorting strategies at operations.
To assist in the quantification of these uncertainties, the Cooperative Research Centre for Optimising Resource Extraction (CRC ORE) and Anglo American have implemented a systematic methodology to evaluate the deployment of sensor technologies for bulk ore sorting. This methodology was implemented at a bulk ore sorting demonstration plant using penetrative prompt gamma neutron activation analysis (PGNAA) and surface x-ray fluorescence (XRF) measurements of base metals, as a proxy for precious metals, at Mogalakwena PGE Mine.
The methodology comprises of three phases. The initial phase involves the preliminary investigation of sensor technologies and analysis to reduce identified risks prior to commissioning of the bulk ore sorting plant. The second phase executes a series of experiments to prove the sensing technologies in the operational environment and provide confidence the sensors can drive separation. The final phase demonstrates the production scale bulk ore sorting performance of the plant and validates its performance by campaign of the sorted products at the processing plant.
Throughout the operational deployment and evaluation of bulk ore sorting at Mogalakwena, several key lessons were learned. These learnings relate to the calibration requirements of the sensor technologies, the application of surface vs penetrative sensor techniques and the preservation of grade heterogeneity delivered to the sensors.
This paper presents the methodology used, and lessons learned, in the operational deployment and evaluation of PGNAA and XRF sensors for bulk ore sorting at Mogalakwena PGE Mine.