The interactive effects of chemical and process parameters on the flotation performance of a UG2 ore B. McFadzean ⇑ , S. Pani, J. Wiese, C.T. O’Connor Centre for Minerals Research, University of Cape Town, Private Bag X3, Rondebosch 7701, South Africa article info Article history: Received 20 June 2014 Accepted 26 August 2014 Available online 29 September 2014 Keywords: Interactive effects Flotation performance Factorial design abstract A major challenge in the processing of PGM-bearing UG2 ore is the fact that it is comprised largely of chromite (60–90% by volume). Chromite is generally not naturally floatable but significant quantities report to the concentrate by entrainment. This is a serious problem for the downstream smelting process, which is generally constrained to a grade of about 3% chromite. Four of the most accessible levers for the control of flotation performance which are available to a plant metallurgist are frother and depressant dosage, air flow rate and froth height. The effects of each of these factors on the flotation performance have been extensively studied, but interactive effects are not well understood. This study presents the results of an investigation into the interactive effects of these parameters using a factorial experimental design approach. The flotation tests were carried out using a 2 m high continuous column flotation cell which facilitated the generation of a deep froth, since this was the phase of most interest. The results were analysed using statistical analysis software that fitted a regression model to the data for each of the outputs. These results highlight the interactive effects of these variables on the flotation performance and illustrate the complexity that is required in process control. Ó 2014 Elsevier Ltd. All rights reserved. 1. Introduction South Africa is the largest producer of platinum in the world (Cawthorn, 1999). About 75% of the world’s platinum reserves are present in the Bushveld igneous complex, which consists of three main ore bodies, viz. the Merensky reef, the UG2 reef and the Platreef. The Merensky reef is comprised largely of orthopyrox- ene (60%), plagioclase feldspar (20%), pyroxene (15%), phlogopite (5%) and some olivine. The base metal sulfides and associated platinum group minerals are contained within these gangue minerals. The Platreef consists of a complex collection of pyroxenites, serpentinites and calc-silicates. The base metal mineralisation and PGE concentrations are irregular in value and distribution (Schouwstra et al., 2000). The mineral composition of UG2 ore is different from the other ores in that it is dominated by hydrophilic chromite gangue (60– 90% by volume), with some silicate minerals (5–30% pyroxene and 1–10% plagioclase). The chromite reports to the concentrate by mechanical entrainment rather than true flotation and this results in a linear relationship with water recovery (Engelbrecht and Woodburn, 1975; Bishop and White, 1976). Large amounts of chromite in the concentrate have severe implications in PGM smelting because of spinel formation and decreased capacity in the furnace (Ekmekci et al., 2003). Process parameters such as froth height, superficial air velocity, frother concentration and depres- sant dosage all play major roles in the recovery of chromite to the concentrate (Yoon and Luterell, 1989; Harris et al., 1963; Jowett, 1966; Sadler III, 1973; Ekmekci et al., 2003; Valenta, 2007; Hay and Roy, 2010). There have been a number of studies conducted on the effects of individual parameters on the flotation performance of a UG2 ore (Ekmekci et al., 2003; Valenta, 2007; Hay and Roy, 2010; Hay, 2010). It is well known that an increase in froth height will lead to drainage of water from the plateau borders and bubble thin films, leading to a lower degree of entrainment (Savassi et al., 1998; Ekmekci et al., 2003). Ekmekci et al. (2003) found that an increase in froth height decreased the grade of Cr 2 O 3 in the concen- trate. However, they pointed out that this would also affect the PGM recovery – a response that they did not investigate. The effect of froth height on PGM recovery is not well studied. Alvarez-Silva et al. (2014) showed that differences in froth height had little effect on PGM recovery. However, the two froth heights that were tested were both very deep and the critical point in PGM losses may have been expected at shallower froth depths. A meticulous plant-scale CCRD study conducted by Harris et al. (2013) showed that PGM http://dx.doi.org/10.1016/j.mineng.2014.08.016 0892-6875/Ó 2014 Elsevier Ltd. All rights reserved. ⇑ Corresponding. Tel.: +27 216505528. E-mail address: Belinda.mcfadzean@uct.ac.za (B. McFadzean). Minerals Engineering 70 (2015) 92–98 Contents lists available at ScienceDirect Minerals Engineering journal homepage: www.elsevier.com/locate/mineng