Abstract
Platinum group elements (PGE) are found at the Bushveld complex (BC) in South Africa.
Three PGE-rich layers of economic importance exist at the BC. These layers are Merensky,
Upper Group 2 (UG2), and Platreef. These layers have been mined for several decades and
mine sludge was produced. Due to the heavy nature of mechanised mining machinery piping
failure may lead to oil spillage, resulting in contamination of the mine sludge. Therefore, this
study aimed to decouple the effects of the removal of hydrocarbon contaminants from mine
sludge, and further establish the impact that contaminants removal has on the flotation
performance of PGE-bearing ores.
Two ores (pristine ore and contaminated mine sludge) were utilised for this study. The ores
were characterised using Fourier transform infrared spectroscopy (FTIR), Gas
chromatography-mass spectroscopy (GC-MS), X-ray fluorescence (XRF), X-ray diffraction
(XRD), and Scanning electron microscopy- Energy diffraction X-ray spectroscopy (SEMEDS).
These analytical equipment were used to determine the existence of hydrocarbons in the
samples, elemental and mineralogical composition. A fire assay with an Inductively coupled
plasma-optical emission spectroscopy (ICP-OES) finish was used to determine the feed grade
of the two ores. The ozonation process was then used to pretreat the contaminated mine sludge.
The impact of pH and washing time was investigated using the response surface methodology
(RSM) approach. Baseline batch flotation tests were conducted for pristine ore, contaminated
mine sludge and ozone-pretreated mine sludge. The samples were all floated at a particle size
distribution (PSD) of 53% passing 75μm.
The results indicated that the mine sludge contained C-H bonds, which indicated the presence
of hydrocarbon contaminants. GC-MS qualitative analysis indicated that the mine sludge was
dominated by alkane types of hydrocarbon contaminants, with a fraction of polycyclic
aromatic, aromatic, and cyclodecyne. Other types of hydrocarbon contaminants were also
detected. These components match with the presence of diesel, kerosene, biojet, heavy
hydrocarbons, and petrol as contaminants. The contaminated mine sludge was found to have a
grade of 9.93 g/t and a pristine ore grade of 4.74 g/t. Bulk mineralogical analysis indicated that
both samples under study consisted of base metal sulphides (millerite, chalcopyrite, pentlandite
and pyrrhotite). Enstatite, Anorthite, and magnesium chromite were also detected in both
samples. The surface morphology results showed that both samples were partially liberated.
The particle size distribution (PSD) of the mine sludge was finer than that of pristine ore. The
maximum removal of hydrocarbon contaminants by ozonation was 98.40%, which was
achieved at a pH of 10.5 at 90 minutes of washing time. Flotation results showed that there was
vii
a decrease in recovery with increased grade after the removal of the contaminants from the
mine sludge. The mass pull of mine sludge was 58.41% before the ozonation process and that
of pristine was 8.03% After ozonation, the mass pull decreased to 32.89% The recovery from
pristine ore of Ni, Cu and Fe were18.639,42.369 and 0.922%, respectively. The recoveries of
Ni, Cu and Fe from the pretreated mine sludge decreased from 45.63 to 21.02, 55.15 to 22.90
and 52.72 to 21.30%, respectively. The final concentrate grade for Cu increased from 0.020 to
0.028% after the ozonation process. Ni and Fe concentrate grades slightly decreased from 0.116
to 0.083 and 0.075 to 0.046, respectively. Ni recovery (21.052%) after ozonation proved to be
the closest to pristine ore recovery (18.639%Ni).