Mineralogical domains by XRD multivariate statistical analysis and SEM-based image analysis of a nickel sulfide deposit, Sta Rita mine, Brazil

摘要

This study was carried out on samples from a disseminated Ni and Cu sulfide ore zone emplaced in a mafic-ultramafic layered intrusion at Bahia State, Brazil. This intrusion was latter metamorphosed into granulite-facies. Industrial plant benchmark data have showed a poor relationship between ore chemical composition and its mineral processing behaviour. Therefore, the knowledge of the mineralogical variability in the deposit is crucial for mine planning. More than 300 samples from drill core samples from the feasibility studies and short term mining plan were grouped into geological domains by multivariate statistical analysis (MSA) of X-ray diffraction data (XRD). Representative samples of the main mineralogical domains pre-defined by XRD-MSA ore-types were later subject to detailed mineralogical studies by scanning electron microscopy (SEM) and automated image analysis (MLA - Mineral Liberation Analyser). These results validated the pre-defined mineralogical domains, concerning about sulfides mineralogy and their associations, which are key features for the geometallurgical model and mineral processing. It is possible to conclude that studied area from the Santa Rita deposit region can be classified into seven mineralogical domains based on the main minerals (serpentine, olivine, and pyroxene) content.