The Gravity Recoverable Gold Test and Flash Flotation

"The role of flash flotation in precious metal recovery is reviewed. Interacting flash flotation and gravity recovery is discussed. The GRG test can be used, when linked to the behaviour of gold in grinding and classification units, to predict the recovery of GRG by flash flotation. The Cadia mill in New South Wales is used as a case study of circuit design, performance and simulation. It is shown that high throughput coarse grind applications such as Cadia require very effective fine GRG recovery systems in the primary grinding loop, which are more consistent with flash flotation than batch centrifuge concentrators (BCCs). BCCs can effectively recover a significant fraction of the GRG from flash concentrates, as predicted by laboratory testing and confirmed at a growing number of mill sites.INTRODUCTIONThe GRG test was designed to characterize Gravity Recoverable Gold (GRG) ten years ago (Woodcock and Laplante, 1993; Laplante et al, 2000), in response to the need to predict how much gold could be recovered from the main circulating load of grinding circuits using centrifuge concentration. The amount of GRG is not itself a prediction of gravity recovery: the performance of a gravity circuit nested in a circulating load also depends on the type and size of unit used, how much of the circulating load is treated, and how many times GRG is recycled and presented to the recovery circuit. The GRG test was first used in a simulator to predict gravity recovery (Laplante et al, 1995), and indeed most of the early tests targeted ores or prospective ores for which gravity recovery was an obvious match. As the test became more accepted, more and more samples returned a verdict that precluded the use of gravity recovery, or made it marginal. Typically, such ores are either copper-gold or base metal ores in which the GRG content is often fine (largely below 105 um) and ranges from 8 to 33%. Extensive simulation and plant sampling tests, either published (Laplante et al, 1997) or proprietary, show that full scale gravity recovery for these applications is generally below lo%, and can be as low as 2%. Not in all cases are such low recoveries economically unjustifiable, since there is documented evidence (Darnton et al, 1992) that a significant proportion of the gold recovered, at least in the case of base metal ores, would not have reported to the copper or zinc concentrate and received NSR credits."