A Simple Gravity-Recoverable Gold Test

"The gravity-recoverable gold (GRG) ore characterization test was developed 15 years ago and has been applied to hundreds of ore samples. The test accurately determines the size distribution of GRG, at which grind it is recoverable, and the extent of its liberation above 150 µm. This contribution shows how the test can be simplified, what are the likely problems to occur, and what information is then lost. Two simplified versions of the test are presented, of which the simplest, grinding a 20-kg sample to a P80 of 75 µm, is retained as the best one. A database of 18 comparative tests is presented. The choice of the full GRG test vs. the simple test is discussed, as is the deportment of the GRG recovered in the standard test but not in the simplest test.INTRODUCTIONThe GRG test was developed in 1990 to characterize the progressive liberation of gold in representative samples of ores, with the primary objective of generating data for a gold gravity recovery model (Woodcock and Laplante, 1993; Laplante et al., 2000). Characterization includes the size distribution of GRG as well as the grind size at which it becomes recoverable. This is achieved in a 3-stage protocol of liberation and recovery using a laboratory centrifuge unit, a Laboratory Knelson Concentrator 3 inches in diameter (KC MD3). This protocol is shown in Figure 1. The first stage takes place at a F100 of 850 ?m (F80 of ~550 ?m), the second at F80 of 180 ?m, and the third at a P80 of 75 ?m. The rationale for the three cycles is the desire to preserve the original size distribution of the gold grains in the ore and to obtain a recovery grind size curve. The feed is a representative sample of ore that can be extracted from ore samples (e.g. drill core reject) or stream products of the grinding circuit other than circulating loads or product having undergone hydraulic classification (e.g. rod mill discharge or SAG screen undersize). Ideally the samples should weigh between 60 and 100 kg, depending on feed grade and gold top size. It is beyond the scope of this paper to fully discuss measures of representativity in the ore sample, for example Gy’s minimum sample mass model (Gy, 1979). However, further work in this regard may be an opportunity. The large initial weight facilitates representativity and minimizes weight and sulphide recovery to the concentrate (since concentrate weight is relatively constant at ~100 g), which limits the recovery of gold in nonsulphide gangue or gold carriers (i.e. sulphides). Thus, the 100 g concentrate carries principally discrete gold species, such as electrum, and native gold. The concentrate and a 600-g representative sample of gangue of each stage are screened down to 20 ?m and each size class is fire assayed (the concentrates to extinction, the tails up to 30 g per size class). The five coarsest size fractions of stage-1 concentrate are individually upgraded using a hydrosizer and the concentrates examined with an optical microscope to assess the degree of liberation of the gold phase. Scanning electron microscope work has established that the GRG has an excellent degree of liberation below 150 ?m in stage 1 and generally a good degree of liberation in stages 2 and 3 (Guerney et al., 2003); as a result the size distribution of the GRG has been found to approach closely that of gold grains (Nesset et al. 2005)."