Caciagli N. (2016) Multielement Geochemical Modelling for Mine Planning: Case Study from an Epithermal Gold Deposit. In: Martín-Fernández J., Thió-Henestrosa S. (eds) Compositional Data Analysis. CoDaWork 2015. Springer Proceedings in Mathematics & Statistics, vol 187. Springer, Cham
Mineralisation and alteration processes will result in zones with distinct geochemical characteristics within an orebody. To visualise the mine scale variability that arises as a result of these processes, geochemical domains are defined using a k-means clustering algorithm to analyse multielement data. The fact that the chemical values can be grouped in a defined 3D location clearly suggests that the clusters have meaning in terms of geological process. Principal component analysis (PCA) of these clusters can further improve the understanding of this variability. These clusters form the basis of the geochemical domains which have direct implications for characterization and proportional sampling of geometallurgical and waste rock domains. In the case study presented, pre-mining geochemical characterisations were undertaken at an epithermal gold deposit to support metallurgical sampling and mine planning. k-means cluster analysis and principal component analysis of the geochemical clusters was used to support the metallurgical sampling programme by identifying domains for variability testing. The geochemical clusters identified were used to define the oxide, sulphide and transition zones, a critical factor for mineral processing and recoveries and a key variable in the economics of the project. The R software environment for statistical computing was used for exploratory data analysis (e.g. PCA; zCompositions, robCompositions) and k-means analysis (fpc).
Compositional data analysisPCACluster analysis geologyMining