Analysis of flotation rate distributions to assess erratic performances from size-by-size kinetic tests. (1st April 2020)

• Record Type: Journal Article
• Authors: Vinnett, L.; Marion, C.; Grammatikopoulos, T.; Waters, K.E.
• Is Part Of: Minerals engineering. Volume 149(2020)
• Publication Date: 2020-04-01
• Subjects: Flotation rate distribution -- Flotation kinetics -- Gamma distribution -- Analysis of losses; Mines and mineral resources -- Periodicals; Ressources minérales -- Périodiques; Mines and mineral resources; Periodicals; Electronic journals; 622
• DOI: 10.1016/j.mineng.2020.106229
• Languages: English
• ISSNs: 0892-6875

Abstract

Highlights: A concave trend for the mean rate constant was observed with particle size. However, the Gamma f ( k ) showed different shapes in the fine and coarse classes. J-shaped distributions were obtained in the −20 μm class. Mound-shaped distributions were observed in the +75/−150 μm class. An analysis of losses showed the recovery in the coarse class was controlled by R ∞ . The recovery in the fine class was also defined by very slow floating fractions. Abstract: This paper presents a size-by-size kinetic characterization for copper-lead separation from a complex ore. Batch tests are run under extended flotation times to obtain approximately steady recoveries (unscreened samples) at the end of the tests. The size-by-size time-recovery curves are fitted to the Gamma model to obtain estimates for the maximum recovery R ∞ and the flotation rate distribution f ( k ). From reverse J-shaped distributions to normal probability density functions are representable from a Gamma f ( k ). Four size classes are studied: −20 μm, +20/−38 μm, +38/−75 μm and +75/−150 μm. The concave relationship between recovery and particle size is obtained up to a limiting flotation time; thereafter, a decreasing trend is observed for Pb and Cu. The finest (−20 μm) and coarsest (+75/−150 μm) classes present slower average rate constants. However, the different f ( k ) shapes (reverse J-shaped versus mound-shaped distributions) indicate that the use of average rate constants leads to limited conclusions in terms of mineral losses. Therefore, better kinetic interpretations can be obtained from the overall analysis of: the R ∞ estimates, a location parameter of f ( k ) and the metal losses in a low f ( k ) percentile ( e.g., 5%). From the proposed methodology, the poor performance of any erratic component can then be characterized as slow-rate-limited, R ∞ -limited or both together.