Avaliação do efeito da porosidade nas etapas de beneficiamento de minério de ferro da Samarco

Abstract

It was investigated if there is a significant difference in the porosity parameters of the ores that feed the Samarco Concentrator and if this possible disagreement causes different performances in the desliming and flotation processes. Two ore samples from the Alegria 3/4/5 pit and one from the Alegria 9 pit (samples 5/14, 5/4 and 9/4) with a porous phases percentage (porous hematite and goethite) of around 50% and one ore sample from the Alegria 3/4/5 pit (sample 5/35) with a much lower porous phases percentage (5.6%) were used in order to check the sensibility of the porosity determination methods. The run of mine samples were submitted to grinding tests to produce samples to the desliming and flotation tests, following the Samarco procedures. The grinding product, desliming underflow and overflow, flotation concentrate and tailing of the 5/35, 5/14, 5/4 and 9/4 samples were analysed to obtain a chemical, size and mineralogical characterization in the Samarco labs. The mercury intrusion porosimetry (MIP) and gas adsorption- condensation porosimetry (GACP) were used to characterize the porosity parameters, supported by scanning electron microscopy. There is a significant difference of porosity between the investigated samples. The 5/35 sample shows smaller porosity (by GACP and MIP total volume of pore minor than 1,0mm and 100nm) and smaller specific surface area (by BET), which is consistent with the minor quantity of porous phases. The 5/4 sample, despite its mineralogical composition, shows larger specific surface area, larger pore volume, smaller mean diameter of mesopores and it was the only sample that shows micropore volume. The porosity data, from the two methods (MIP and GACP), and the specific surface area data to the other samples 9/4 and 5/14 indicate similar properties. Besides the detection boundary of gas adsorption (100nm maximum), the GACP was more effective in distinguishing the samples than mercury intrusion porosimetry. This latter method showed problems during the analysis execution (low equipment availability, difficulty in maintenance and acquisition of spares and penetrometers and poor repeatability). However in the value range where both methods are simultaneously applicable (3.0nm to 100.0nm), consistent values of pores volume were found (BJH GACP and MIP) in the samples. The desliming and flotation processes results for the four samples were consistent with the chemical, particle size and mineralogical data. There is no evidence that the porosity had caused any influence in those performance results.