Stochastic representation of the agglomeration process: implications on the saturation variability in a dynamic heap leach

Abstract

Liquid saturation within a heap leach is a critical factor in the heap’s metallurgic performance. This saturation is affected by the agglomerate quality used to build the heap. In this context, our study investigates how the spatial distribution of the agglomerate quality within the heap affects its fluid saturation. The heterogeneity of the heap’s fluid saturation is quantified using the saturation nonuniformity. The differences in agglomerate quality, represented by the permeability of the media, are generated by imperfections in the agglomeration process, which are represented as a stochastic process modeled with Markov chains. The agglomerates produced by this process are stacked orderly in narrow diagonal stages to build a heap. The fluid flow within this test heap is represented as a multiphase flow in a porous media and simulated in OpenFOAM, and the results are analyzed at steady state. Our results show that the volumetric water content distribution within the simulated heap was similar to that of a real copper leaching heap. Four key factors were studied, three of them regarding the physical structure of the heap (agglomeration quality, slope of the heap and percentage of well-agglomerated ore) and a fourth operational parameter (irrigation flux). The agglomeration quality had a significant impact on the heap’s saturation nonuniformity due to the diagonal layers of low permeability along the heap’s angle of repose. On the other hand, heap slope variations between 32° and 38° did not have a significant impact on the average saturation nonuniformity. Also, even when the heap’s percentage of well-agglomerated ore was kept constant, it had more saturation nonuniformity than a uniformly distributed heap with the same percentage of well-agglomerated ore. Finally, the saturation nonuniformity remained constant when the heap was operated with different irrigation fluxes, suggesting that it is a property inherent to the heap’s physical structure and not an operation parameter. The main limitation of this study is the inability of the developed model to represent spatial changes in the capillarity characteristics of the agglomerates. A sensitivity analysis including capillarity characteristic changes revealed that its impact is as relevant as the spatial distribution, which must be further explored in the future