Objective: By evaluating operational filtration conditions, chemical, mineralogical, and particle size properties of the tailings, the study aimed to identify critical variables affecting porosity and provide predictive models for optimizing dewatering operations. Theoretical Framework: This research builds upon existing theories of solid-liquid separation and dewatering processes in mineral processing. Key references include classical works on filtration dynamics, particle size distribution, and cake porosity characterization. The study addresses gaps in literature regarding the relationship between tailings composition and filtration results. Method: The Leaf Test method was employed on 33 fresh slurry samples from the Brucutu plant to simulate industrial filtration conditions. Filtration cycle parameters such as cake formation and drying times were standardized. Porosity was calculated using Grace's equation and correlated with characterization data, including mineralogical composition, true density, and particle size. Statistical methods such as clustering, regression, and Random Forest modeling identified key predictors of porosity. Results and Discussion: The results indicated that porosity correlates strongly with silica content and certain mineralogical attributes, such as the presence of martitic hematite and quartz. Cluster analysis revealed two sample groups with distinct filtration characteristics. While operational parameters showed limited impact on porosity, the statistical models highlighted the significance of ore composition. Research Implications: This research provides a foundation for optimizing iron ore tailings filtration by identifying the key variables influencing porosity. The findings support more efficient dewatering techniques, contributing to sustainable tailings management. Further, the development of predictive models aids industrial operations in minimizing risks associated with tailings disposal. Originality/Value: This study is among the first to integrate statistical modeling and mineralogical characterization in exploring filter cake porosity. The results offer novel insights into the optimization of solid-liquid separation processes in the mining industry.
Read full abstract