Otimização dos processos de moagem e flotação utilizando o modelo do balanço populacional por classes de tamanho, mineral e liberação do minério de grafita

Abstract

Currently, only 60% of all Syrah's Fines (SyrFin) production has a carbon content that meets market specifications. This has generated a loss of revenue and an increase in costs. The adequacy of the product to market specifications necessarily involves the optimization of the grinding and concentration process. The optimization of milling and concentration processes by flotation is a challenge. In the grinding stage, mass transport takes place between size classes, mineral and liberation. The mass transport between size classes, mineral and liberation depends on the characteristics of the minerals present, as well as the operational conditions of the process. In concentration by flotation there is no significant mass transport between the classes, but the flow is divided, according to the characteristics of size, mineral and liberation and the operational conditions of the process, composing the concentrate and tailings flows. To determine the most suitable grinding degree for the process, it is necessary to know the impact of grinding on concentration operations. When the project involves the optimization of more than one step, as is the case, the problem becomes infinitely greater, requiring a safer and less expensive approach than the simple direct test done in laboratory. Simulation presented itself as the best alternative for optimization. With the application of the population balance model by size classes and liberation, it was possible to determine that the hydrophobicity of graphite is sufficient to explain the gangue mineral mass transport in the coarser classes to the concentrate. In other words, the gangue mineral is led to the concentrate associated with the hydrophobic part of the graphite. By ignoring the liberation, the conclusion would be that there is an excess of collector. When in fact it was a grinding problem. The correct approach to the problem allowed, in the optimized test, a graphite recovery above 93% and carbon contents above 91%.