Technical note The potential for dry processing using a Knelson Concentrator M. Greenwood, R. Langlois, K.E. Waters ⇑ Department of Mining and Materials Engineering, McGill University, MH Wong Building, 3610 University, Montreal, Quebec, Canada H3A 0C5 article info Article history: Received 24 September 2012 Accepted 17 January 2013 Available online 27 February 2013 Keywords: Gravity separation Knelson Concentrator Dry processing of minerals Gold processing Synthetic ore abstract Centrifugal gravity concentrators currently operate on a wet basis, with slurry feed and fluidising water used to enhance the separation based upon density differences. This work investigated the potential for running a laboratory scale 3 00 Knelson Concentrator on a dry basis. Air was used as the fluidising medium in order to separate tungsten from silica in a synthetic ore (1% w/w tungsten), and compared to an opti- mised wet process. The wet processing attained a mean tungsten recovery of 94.92% (tungsten grade: 30.96%). The dry processing at two different fluidising air pressures attained a recovery of 78.53% (tung- sten grade: 6.32%) and 69.90% (tungsten grade: 15.57%) at 2 psi and 3 psi respectively. This preliminary work shows that it is feasible to separate minerals on a dry basis in a Knelson Concentrator. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction Gravity concentrators utilise the differences in density of minerals in order to facilitate separation. These concentrators may employ natural gravity, or 1 G, such as spirals. In a spiral concentrator, the denser materials flow down along the inner section of the trough, and the less dense material flows along the outer section. In order to enhance the separation at fine particle sizes, or low concentrations of high value material (such as gold), a centrifugal force may be imparted on the slurry, increasing the gravitational force on the particles and enhancing the separation. The two main centrifugal gravity separators that are in opera- tion are the Knelson (1992) and the Falcon (Honaker et al., 1996; Lins et al., 1992). The Knelson Concentrator was at the forefront of utilising centrifugal forces in the gravity concentration of gold, with LaPlante et al. instrumental in developing the Gravity Recov- erable Gold (GRG) Test which is still in use today (Laplante et al., 1995a, 1995b; Laplante et al., 1996). Centrifugal separators are not limited to gold (Gul et al., 2012), but research is also being carried out in coal (Uslu et al., 2012) and hematite (Chen et al., 2008) processing, as well for processing fine material (Honaker et al., 1996; Kroll-Rabotin et al., 2010; Oruç et al., 2010). However, currently potential limitations include the fact that a large volume of water is required in their operation. The feed enters the concentrator as slurry, and fluidising water is used to enhance the substitution of light material for heavy mate- rial in the riffles of the concentrator. With environmental costs becoming more apparent, investigating dry processing is of para- mount importance. This paper introduces the possibility of config- uring a laboratory scale 3 00 Knelson Concentrator to operate on a dry basis. This has the potential to be utilised in a number of fields, including reducing the nuggeting effect in gold processing and pro- cessing heavy mineral sands in conjunction with dry magnetic and electrostatic separators. 2. Materials and methodology 2.1. Materials Synthetic ore was used to mimic the composition of a gold ore. It has been shown that synthetic ore can accurately simulate gold ore under gravity test conditions (Laplante et al., 1995a, 1995b). Tungsten was used to simulate gold in the synthetic ore. Tungsten was chosen due to having a density similar to that of gold (19.25 g cm 3 and 19.30 g cm 3 respectively) and therefore will behave the same way in the centrifugal separator. Grey polyhedral tungsten particles (Zhuzhou Cemented Carbide Work of China) were used in this study. The particles have a tungsten content of 99.92% and a density of 17.98 g cm 3 . Silica (Unimin Canada Ltd.) was used as the low-density gangue (2.65 g cm 3 ). Fig. 1 shows the size distributions of the silica and tungsten used in the tests, determined by Ling (1998) to be ideal for wet processing using a 3 00 Knelson Concentrator. Wet process- ing was taken as the baseline as a comparison for the dry process- ing. The tungsten and silica were pulverized using a LM2-P pulverizing mill (Labtechnics, Australia) and screened to achieve the required size fractions. 0892-6875/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mineng.2013.01.014 ⇑ Corresponding author. Tel.: +1 514 398 1454; fax: +1 514 398 4492. E-mail address: Kristian.waters@mcgill.ca (K.E. Waters). Minerals Engineering 45 (2013) 44–46 Contents lists available at SciVerse ScienceDirect Minerals Engineering journal homepage: www.elsevier.com/locate/mineng