Data-Driven Synthesis of a Geometallurgical Model for a Copper Deposit

Geometallurgy integrates aspects of geology, metallurgy, and mine planning in order to improve decision making mining schedules. A geometallurgical model is a 3D space that typically synthesized from early-stage small-scale samples composed several metallurgical units, or domains. This work explores the synthesis for copper deposit using purely data-driven unsupervised approach. To this end, dataset 1112 drill used, which are clustered different methods, namely, k-means, hierarchical clustering (AGG), self-organizing maps (SOM), DBSCAN. Two cluster validity indices (Silhouette Calinski–Harabasz) used select final model. validate potential proposed approach, simulated economic evaluation conducted. Results demonstrate k-means exhibits better performance terms modeling obtained scheduling increases project’s Net Present Value (NPV) by as much 4%. Based on these results, methodology an appealing alternative generating models within greenfield, brownfield ongoing operations.