The possibility of using limestone instead of lime as an agent precipitating impurities from cobalt-rich solutions from the hydrometallurgical plants of Ruashi Mining in R.D. Congo was studied. These solutions come from the copper solvent extraction circuit and are intended for the production of cobalt hydrates. Precipitation tests performed in the laboratory have shown that limestone competes with lime in terms of precipitation of Fe, Mn and Cu impurities and co-precipitation of Co. The parameters which were exploited are: Eh, pH, time, % Solid and the particle size. Reduced to the equivalent quantities which enter into a precipitation reaction with 1.5 as the evaluated value of the limestone/lime ratio, the consumption of these two precipitating agents is almost the same. Co-precipitation of cobalt was acceptable and was limited to a yield of 3.0 % Co. Under these same conditions, copper was also removed with yields of around 40 %. The iron precipitated in good yields reaching over 90 %. On the other hand, Mn showed too low precipitation yields of less than 40 % by the fact that it required real oxidizing conditions which could bring it back to its insoluble form of MnO2. Finally, with the considerations on the cost of acquisition on the Ruashi Mining site of 160 and 50 USD respectively for the ton of lime and limestone, without integrating the cost linked to the fragmentation of the limestone to have the correct grain size, it is technically possible to substitute lime by limestone.

The flotation by surface sulphidisation of the oxidized copper-cobalt-bearing ore from Kimpe (1.97% Cu; 0.66% Co) was studied in order to evaluate its behavior by the analysis of its most influential parameters and their possible interactions. Tests limited to a simple roughing were planned and carried out according to a factorial approach using the Taguchi method. The results were found on the basis of the Signal-to-Noise (S/N) analysis and the analysis of variance (ANOVA) by considering the experience plan of the orthogonal matrix L16 (45). The work consisted in studying the doses of the main collector potassium amyl xanthate (PAX), of the sulphiting (NaHS), of the dispersant (Na2SiO3) and of the mixture (Gasoil-Rinkalore 840) as well as the particle size of the former miner at float. The predictive model of Taguchi's analysis predicts recovery yields of 88.44% Cu and 84.90% Co under optimal copper recovery conditions and 88.70% Co and 86.00% Cu under optimal conditions cobalt recovery. The optimal relative levels of the parameters studied are: 300 g/t PAX, 200 g/t Mixture, 4000 g/t NaHS, 200 g/t Na2SiO3 and a particle size of 20% +75 µm for Cu and 400 g/t PAX, 200 g/t Mixture, 4000 g/t NaHS, 300 g/t Na2SiO3 and a particle size of 20% + 75 µm for cobalt. It also emerges from the results obtained that the dose of PAX and that of NaHS are the influencing parameters. In terms of parametric interactions, for the recovery of copper, the mixture and the dose of the dispersant are in interaction considered to be weak. While the dosing of the collector and the particle size are in strong interaction given the differences in parallelism noted. The interaction between the dose of silicate and the dose of sulphiting can be considered weak for the recovery of cobalt.