Mineralogical Fingerprint of Iron Ore Tailings in Paraopeba River Bedload Sediments after the B1 Dam Failure in Brumadinho, MG (Brazil)

Abstract

The study presents SEM-based automated mineralogy to distinguish between natural sediments and iron ore tailings deposits from the Paraopeba River, after the failure of B1 Dam in Brumadinho, Minas Gerais, Brazil. Samples were obtained from borehole cores drilled over channel bars and banks eight months after the failure. After preliminary facies description, sediments from 54 chosen intervals were subjected to density measurement, X-ray diffraction (XRD), SEM-based automated mineralogy (QEMSCAN) analysis and determination of geochemical major components. Hierarchical clustering analysis (HCA) and principal component analysis (PCA) revealed six main mineral associations governed by different contents and ratios of quartz, kaolinite and hematite. Natural sediments are predominantly composed of mineral associations containing kaolinite, quartz and quartz + hematite with density values ranging from 2.5 to 3.3 g/cm3. Tailings deposits have density values higher than 3.5 g/cm3 and are mainly composed of hematite with occasional occurrences of kaolinite + hematite. Because of geological complexity and historical terrain occupa-tion and usage, geochemical anomalies are common in the Paraopeba River sediments. Our data suggests that mineralogical oriented studies should precede detailed geochemical investigations, to enhance the understanding of the source of such anomalies and the environmental jeopardy associated to the occurrence. In this sense, SEM-based mineralogy has an enormous potential in environment studies.