Computer Science · Artificial Intelligence
Mineral Identification and Domain Characterisation using Two Automated Hyperspectral Core Logging Systems, Los Bronces Cu-Mo Porphyry Deposit
M. Quigley, Carsten Laukamp, Bans G. Yildirim
See on the mapAbstract
Acquiring adequate representivity of the mineralogical variability of an ore deposit is a recognised challenge for geometallurgical modelling. Automated hyperspectral core logging overcomes this issue through the rapid collection of contiguous infrared reflectance measurements for the objective and consistent logging of mineralogy in drill core or chips. This paper presents preliminary results from two automated hyperspectral core logging systems, the CSIRO-developed HyLogger-3 and Corescan's Hyperspectral Core Imager Mark III (HCI-3). Thirty-three metres of selected intervals sampled from multiple diamond drill holes across the Los Bronces Cu-Mo porphyry deposit, Chile, were assembled as a 'composite virtual drill hole' and scanned with both systems. Samples were selected to represent five alteration-related domains with predicted varying comminution/breakage attributes. Minerals identified from the shortwave-infrared (SWIR) data, collected by both core logging systems and confirmed by XRD analysis, include: muscovite (sericite), tourmaline, gypsum, chlorite, phlogopite/biotite and trace kaolinite and montmorillonite. Quartz and feldspar were also logged from the HyLogger-3 thermal-infrared (TIR) data. Domains were characterised based on the spatial distribution and intensity of one or more key minerals. Future work will correlate the spectro-mineralogical results with breakage tests to investigate whether a fundamental relationship between rock hardness and drill core mineralogy can be established using hyperspectral core logging. Determining the hardness variability of a deposit reduces the risk and uncertainty of mill throughput and energy usage, leading to more efficient and profitable ore recovery.