A Minerology Based Simulation of Dry Magnetic Separation of an Iron Ore

"Mineral processing unit models require information on particulate properties such as particle size, solids flowrate and mineralogical composition to model their operations. Process models that describe the separation process can be classified based on the level of information required from the ore, i.e. the feed stream to the separating unit. The most comprehensive level of mineral processing models is the particle-based modelling approach. Particle based modelling describes a modelling approach where mineral composition, mineral liberation and textural information of particles making up the process streams is used in modelling. The advantage of this approach is that all model streams carry particle level information (size, liberation, shape). This information is obtained by applying automated mineralogy techniques such as Quantitative Evaluation of Minerals by Scanning Electron Microscopy (QEMSCAN) analysis which provide accurate and detailed information on process stream particle properties. Currently the use of liberation information in modelling and simulation is rare, due to the limited number of separation models that liberation data as input. The objective of this study is to develop a particle based model for a dry magnetic separator, where mineralogical information is used directly to simulate a dry magnetic separator operation. Magnetic separation tests were carried out on high, medium and low grade magnetite ores using a laboratory magnetic separator. Streams from the separator were sampled and sized for chemical and liberation analyses. A sized mass and elemental balance around the magnetic separator was carried out. The elemental assays of each stream were converted to mineral quantities and a sized mineral balance was done. QEMSCAN measured liberation data was reconciled to the mineral balance based on mineral quantities obtained from chemical assays. This was followed by the development of a particle based model that calculates the trajectory of a single particle based on its mineral composition and textural properties. The model was tested on experimental data and was able to predict the effect of variation in process feed stream mineralogical attributes (liberation, grade, particle size) on magnetic separation performance. Results of this study highlight the benefits of the particle based modelling approach. INTRODUCTION The extraction of mineral values from ore requires liberation of the value minerals from the gangue followed by their separation. Liberation is achieved by size reduction operations (Desmond, 2008; Fuerstenau and Abouzeid, 2002). Separation of gangue material from value mineral takes place in various mineral upgrading stages such as classification, magnetic separation and flotation. In these concentration stages value minerals are separated from gangue minerals based on the physical properties of the ore particles making up the feed such as size, liberation, volume of magnetic material and density. The success of separation is related to the liberation of the feed particles, the performance of the separation technology applied, as well as the inherent characteristics of the feed material (Leißner et al., 2013)."