Using the Rosin-Rammler method to correlate recovery with particle size distribution and content in a sulfidic gold ore

Abstract

As gold ores become more challenging to treat, it is increasingly important that the impact of refractory behaviors on recovery is understood. For example, gold extraction and recovery faces many challenges when associated with, or occluded by, sulfides, as the gangue material can produce cyanicidic complexes or even passivating films on the gold surface. By using a methodology developed by Claude Bazin in 1994 in his research, the correlation of mass:gold, mass:sulfur and sulfur:gold by particle size distribution on recovery was developed for a low-sulfide gold ore from the Waihi Favona mine. This represents a unique application of Bazin’s method as previous work had been done on sulfide ore flotation. Overall, gold recovery values increase as P80 decreases and the higher-grade drum material samples tended to have higher overall recovery values. The lowest recovery in both the bucket and drum materials samples occurred in the coarsest samples, which automated mineralogy revealed as showing silica encapsulation of gold particles. There was no pyrite encapsulation found in the bucket material samples, and total pyrite encapsulation was rare in the drum material samples. In this research, overall gold recovery values ranged from 75 to 92%. Recoveries in the mid-80s can be routinely achieved in samples that had a P80 of either 200 or 270 mesh, and in the low 80s in the samples with a 150 mesh P80. This suggests that the mill could reduce fine grinding energy costs by as much as 34%, or nearly NZ$700,000 annually. It is also conceivable that as a consequence of a coarser required grind, plant throughput could be increased, thereby reducing operating costs further. The total sulfur content did not change significantly as a result of cyanide leaching. Plotting the overall change in sulfur content against overall recovery did not show any strong correlation. Examining gold recovery by size fraction shows that recovery does increase as the head sulfur value increases, but this is believed to be a result of increasing liberation with continued grinding, since the size fraction with the highest sulfur content is also in the finest size of -400 mesh. The Rosin-Rammler analysis shows that particle distribution of a sample does not change greatly as a function of leaching. The Rosin-Rammler analysis of the mass:sulfur:gold ratios do show some change, most noticeably in the coarsest size fraction of each sample. Mass:gold ratios showed a drop in particle size and an increase in mass retained. Mass:sulfur ratios did not show as much of a change in particle size, but percent retained did increase after leaching. Sulfur:gold particle size decreased, but the percent retained decreased. Creating a model of the ratios showed trends that can be used to predict ore-specific gold recovery. The best predictor is the mass:gold distribution, followed by the mass:sulfur distribution, which both describe gold liberation. Sulfur:gold was not found to be a good predictor of extraxtion for these gold ores.