Abstract

The studied copper sulfide deposit is located in the northern region of the Carajás Mineral Province, in the state of Pará, Brazil. The ore is hosted in granite rocks of three units: Xingu Complex, Plaquê Suite and Gelado Granite structurally associated with the Cinzento Transcurrent System and the Pojuca Fault of the Carajás Transcurrent System.

The granitic host-rocks reveal hydrothermal alteration with amphibole, albite, biotite, and garnet. The mineralization shows chalcopyrite as the main Cu-bearing mineral, but chalcopyrite-bornite associations are locally observed. The Cu-sulfide mineral association can be identified by the Cu/S ratio obtained by chemical analyzes of the geological database.

For process variability studies, 50 representative ore samples from drill cores were selected. The mineral processing samples vary from 0.2% of copper grade and 4.9% of copper grade. Flotation locked cycle tests (LCT) were performed and Cu recovery (%) ranged between 84% and 99% with an average copper recovery of 96%. The copper grade in the concentrate varies from 25% to 53% and shows an average of 34% of copper grade. The minimum and maximum values for copper grade in the concentrate are explained respectively by the presence of pyrite and bornite as the main Cu-bearing mineral in specific regions of the deposit.

The geometallurgical model for copper recovery and copper grade in the concentrate was developed in two stages: i) estimation of geometallurgical parameters through multiple regression equations; and ii) application of the transfer functions to the block model.

Variability results of copper recovery were described by the multiple regression equation that considered Cu grade and the Cu/S and Cu/Fe ratios as independent variables. The model showed 70% of adjusted determination coefficient (adj R2) and a normal distribution of the residuals, without bias. For the copper grade in the concentrate, the variability are described by the Cu/S ratio, showing 96% of R2.

To assign values of copper recovery and copper grade in the concentrate to the block model, Cu along with Fe and S grades were estimated. The interquartile value for copper recovery in the block model is 95% to 97% and average copper recovery is 96%. For the copper grade in the concentrate, the interquartile value of the blocks is 30% to 33% and the average is 32%. The regression equations for copper recovery and copper grade in the concentrate are, respectively, 98% and 96% compliant with the block model.

By analyzing the assigned values, we identified distinct regions within the deposit exhibiting varying geometallurgical behaviors. To assess geometallurgical uncertainty, qualitatively, we adapted resource classification criteria, considering the distance between each block and the nearest tested sample. Our findings revealed that: 49% of the blocks exhibited low to moderate geometallurgical uncertainty; 33% displayed high uncertainty and 18% had very high uncertainty.