Optimizing flotation bank performance through froth depth profiling: Revisited

Abstract

This communication revisits previously reported results on froth depth profiling along a rougher flotation bank. The optimization problem is reformulated as to maximize the overall bank Cu recovery subject to a lower bound constraint on the overall Cu concentrate grade. This formulation differs from that originally proposed in Maldonado et al. (2007) where the sum of the squared Cu tailing grade of each cell group was minimized for a given target bank Cu concentrate grade. A semi-empirical steady-state mathematical model of a bank of cells previously validated using industrial flotation data from Los Pelambres mine in Chile was used to simulate the process. In order to improve resolution a genetic algorithm was implemented to search for the optimal froth depth profile as opposed to the discrete dynamic programming technique originally implemented. Results show that optimal froth depth profiling resulting from solving the reformulated optimization problem produces an increase in the overall bank recovery for a given target Cu concentrate grade compared to that obtained when solving the original formulation. Moreover, the resulting optimal mass-pull profile tends to be more balance in the first cells which partially agrees with recent observations.