A Novel Geometallurgical Approach to Tailings Storage Facility Characterisation and Evaluation

Abstract

Millions of tonnes of sulphide tailings are produced each year from base metal and gold flotation operations. Of these tailings a substantial amount is stored in Tailings Storage Facilities (TSFs). In addition to the environmental benefits of removing sulphide components from tailings there is the potential of additional economic benefit through recovery of valuable components. In an iron oxide/copper/gold operation such as Ernest Henry Mining (EHM) magnetite, cobalt, copper and gold could be recovered as economic products from the TSF. The EHM TSF was used to develop the TSF characterisation method described in this thesis. Geometallurgy has become accepted practice with major mining companies for in situ ore characterisation. If a TSF is seen as a potential new orebody (or resource) the implication is that geometallurgical methods should also be used in the evaluation of a TSF. However the traditional approach to tailings evaluation has been predominantly based on assays and limited metallurgical testwork, if any. Variations in tailings mineralisation (and thus grade and recovery potential) are not taken into account. Despite an extensive literature review no evidence was found of published literature on geometallurgical characterisation of TSFs and specifically no published evidence of any 3D block modelling of TSFs that incorporates metallurgical variables. This thesis presents the first such published case study. There are several likely advantages in applying geometallurgical characterisation principles to a TSF. A geometallurgical approach will identify and quantify variations in tailings mineralisation (and thus grade and recovery potential). Project value and risk are defined more accurately if a systematic sampling and testing approach is employed, enabling the use of metallurgical attributes in the establishment of resource domains. Key aspects of the TSF characterisation process investigated in this work include the following: • A well-designed drilling, sampling and assay program. • A combined Principal Component Analysis (PCA) and k-means cluster analysis approach to improve the identification of different domains in the TSF having the same assay characteristics. • Mineralogical investigation of spatially different samples using optical microscopy, MLA and quantitative XRD. • Small-scale physical tests to identify and quantify metallurgical processing differences between domains. The laboratory mineral processing testwork in this work consisted of 40 batch flotation tests plus replicates to determine the full grade recovery curves for the recovery of sulphur, copper, pyrite and cobalt, 41 diagnostic leach tests plus replicates to determine CN recoverable and sulphide-locked gold, and 42 Davis Tube magnetic separation tests plus replicates to determine magnetite recovery. • The use of statistical methods to develop processing proxy models to allow the prediction of value recovery throughout the deposit. • Block modelling that quantifies the spatial distribution of metal grades and recoveries. • Conditional simulation to generate a distribution of grade and recovery outcomes (in this particular study gold). Conditional simulation allows for a better understanding of risk involved in the potential reprocessing of a TSF by incorporating model uncertainties. The key outcomes of this work were: • The first application of geometallurgical principles to tailings storage facility characterisation, including block modelling for resource evaluation. The method applied may be generic and applicable to other tailings dams. • The first use of a combination of PCA and k-means clustering in class-based modelling – this represents an advancement of the approach used by Keeney (2010). • The first published 3D geometallurgical model of a TSF. • The first published conditional simulation of TSF inferred resource data. • Novel conditional simulation of metallurgical attributes for a TSF. • First understanding of the gold, magnetite, cobalt and REE deportment data for an iron oxide Cu-Au deposit tailings resource aimed at the potential reprocessing of the EHM TSF. • The inferred resource value for gold, magnetite, cobalt and copper for the EHM TSF.