Abstract

The incorporation of geological controls is essential for an accurate assessment of the in-situ and recoverable resources in an ore deposit, directly impacting the downstream stages of mining projects. Commonly, the mineral resources evaluation is carried out hierarchically, considering the definition of geological domains first and then predicting or simulating the metal grades within each domain. Nevertheless, this approach assumes a weak correlation between the metal grades across the domain boundaries, which could not be suitable when gradual variations of the grades are observed across these boundaries. To account for the latter scenario known as soft boundaries, we compare three approaches: (i) a hierarchical prediction of the geological domain indicators and the grade within each domain, (ii) a direct prediction of partial grades, defined as the product of the grade and a geological domain indicator, and (iii) a joint simulation of the grade and geological domain indicators. A por-phyry copper deposit in which the copper grade is controlled by mineralogical and rock type domains is used as a case study. When compared with production data, the proposed approaches generate more precise predictions than the traditional approach consisting in accounting for the hard boundaries between mineralogical domains but ignoring the soft boundaries between rock types. The joint simulation approach provides more realistic grade variations across the rock type boundaries, allows for an unbiased prediction of the recoverable resources, and quantifies the uncertainty on these resources based on multiple grade outcomes. Ultimately, we emphasize that identifying the geological controls and the nature (hard or soft) of the geological boundaries, and then defining the proper approaches to account for them is necessary to accurately assess the in-situ and recoverable resources.