On the Role of Bubble Size in Column Flotation

"Three models to predict the effect of bubble size (Db) on metallurgical response were tested in a pilot-scale flotation column at Teck Highland Valley Copper mine. The column was fed with slurry collected from the Cu/Mo separation circuit. The first model was the one developed by Gorain et al. (1997), which assumes that the collection zone rate constant (kc) is inversely proportional to Db. The second model was the one proposed by Hernandez-Aguilar et al. (2010), which considers kc as an inverse square function of Db. The third model was the one developed by Yoon and Luttrell (1989) which provides expressions to describe the efficiency of particle collection (E) based on first principles. The results showed a significant bias between experimental and predicted data using the model of Gorain et al. (1997), possibly reflecting the model’s inability to explain the impact of bubble size, bubble rise velocity and flow regime on E. The model of Hernandez-Aguilar et al. (2010) well described the experimental evidence. The model of Yoon and Luttrell gave accurate and unbiased predictions. However, Yoon and Luttrell’s approach calls for a criterion to determine whether the flow regime is potential or intermediate, a subject that requires more fundamental research.INTRODUCTIONParticle collection, i.e., bubble-particle collision followed by attachment, is the basis of flotation. In a quiescent environment (such as the one observed in a column, or away from the impeller in a mechanical cell), a collision event can take place as long as the “degree of turbulence” caused by the rising bubble does not prevent particle and bubble to flow in close proximity. Upon collision, the particle slides on the bubble surface for a period of time. If the particle is hydrophobic, attachment occurs when the “sliding time” is longer than the “induction time”, which is the time required to thin and rupture the intervening liquid (usually water) film forming a three-phase contact angle. While the induction time is largely influenced by the degree of hydrophobicity of the particle, the degree of turbulence is governed by the bubble size. Generally, direct control of bubble size is not exercised by most mineral processing practitioners, in spite of its obvious importance. Studies to model the effect of bubble size on particle collection using fundamental approaches have been conducted (e.g., Anfruns and Kitchener, 1977; Webber and Paddock, 1983; Yoon and Luttrell, 1989; Dai et al., 1999). However, these models have been validated under highly idealized laboratory conditions (micro-scale experiments using single minerals and uni-sized bubbles), far from resembling those existing in practical industrial applications. Fundamental studies to interpret the effect of bubble size on particle collection in industrial environments are largely absent in the literature. Partly because of this lack of practical knowledge, many mineral processing practitioners perceive laboratory “bubble size” studies to be valuable academic exercises, but not something that can be directly applied to daily production issues."