Statistical Modeling for Tailings Consolidation Using Index Properties

Abstract

The extraction of mineral resources from the earth generates large volume of mine tailings which are usually disposed hydraulically in tailings pond. To minimize the environmental concerns associated with deposition of mine wastes and reclamation of the containment ponds, the tailings need to undergo an efficient consolidation. The main objective of this research was to develop statistical models for tailings consolidation using index properties based on data in the literature thereby, capturing physicochemical interactions. The first model was developed for groups of tailings (sedimentary clays, residual soils and oil sand tailings) and the second model was developed for all tailings. Results indicated that void ratio varied significantly at low effective stress (< 5 kPa) and high hydraulic conductivity (> 10 [negative superscript 6] cm/s) for sedimentary clays and residual soils, because physicochemical interactions are maximum under these conditions. For oil sand tailings, effective stress (σ') and hydraulic conductivity (k) were strongly correlated with void ratio due to relative homogeneity in material composition. The effect of process conditions on consolidation behavior was found to be negligible for this class of material. By converting void ratio to liquidity index and normalized void ratio, volume compressibility and hydraulic conductivity data in the literature merged for materials of each group even at low σ' and high k. This means that the newly developed relationships can better describe tailings behavior even at low σ' and high k. The volume compressibility coefficients (A, B) varied linearly with plasticity index (I[subscript p]) whereas the hydraulic conductivity coefficients (C, D) followed decreasing power law functions of I[subscript p]. Strong agreements between modeled data and measured data (published in the literature) were obtained for both void ratio (R[superscript2] = 0.98) and hydraulic conductivity (R[superscript2] = 0.99).