Unravelling the Factors Impacting on Concentrate Quality by Geometallurgical Data Analysis at Fortescue Metals Group's Iron Bridge Magnetite Mine

摘要

The Iron Bridge magnetite mine and infrastructure project is currently in development as a joint venture between Fortescue Metals Group, Formosa and BaosteeL It is supported by an extensive database of reverse circulation drill hole samples analysed by X-ray fluorescence (XRF), with a large subset tested for concentrate mass recovery, concentrate and tail grades by Davis tube recovery (DTR) testing. A limited number of rock physical testing results, including unconfined/ uniaxial compressive strength and crushing work index, completes the database. By combining all available information - assays, DTR results, geology, stratigraphy, mineralogy and rock physical results - and using multivariate statistical analysis techniques, the impact of the critical factors that control concentrate quality, mass recovery and grindability characteristics have been outlined. A set of geochemical groups was created by using a cluster analysis technique from the combined XRF-DTR data. These groups allow optimal separation of material with consistently high concentrate Fe grades and mass recovery. This approach is superior to domaining based on stratigraphic logging or mass recovery as it is commonly used in standard magnetite estimation models. Using the results of a normative mineralogy study, the main geochemical groups were given a mineralogical profile fitting their geochemical characteristics. These findings further refined the understanding of the link between the process characteristics of the samples and their geochemical signature and mineralogy. The geochemical grouping defined on head grades and DTR concentrate mass recovery was generalised into a 'geomet' code based solely on the routinely assayed head grades, thus bypassing the need for systematic DTR testing. The geomet code identifies spatial domains with homogeneous geochemistry, mineralogy and process characteristics in both oxide and fresh material that form the basis of the current Iron Bridge resource model. The allocation of geometallurgical characteristics in terms of concentration amenability and concentrate characteristics on a block-by-block basis over the whole project allows for an optimum mine schedule and plant feed. This ensures that the desired concentrate characteristics and mass recovery can be better predicted over the life-of-mine, thereby improving the mine's economics and stability by derisking the mine schedule and optimising the performance of the concentrators. A key benefit of the geomet code devised for the Iron Bridge mine is that it can be attributed in an automated manner to all samples analysed for a routine XRF analytical suite, from exploration to blasthole drilling. This allows each sample to be allocated to a geometallurgical material type with predictable process response in terms of concentrate mass recovery and concentrate quality without the need to complete systematic DTR testing. A standard multivariate analysis coupled with a cluster analysis has allowed the linking of interdisciplinary results between geology, mineralogy, geochemistry, process and metallurgy to better understand, predict and ultimately mine and process the Iron Bridge deposits. This allows plant operations and product quality to be optimised through each stage thanks to the predictable ore characteristics embedded in the mining model.