Quantifying uncertainty in geological boundary modeling for mineral resource estimation: application to a porphyry copper deposit

Francky Fouedijo
Mineral resource evaluation aims to accurately predict the grades and tonnages of a mineral deposit that will be exploited during a specified time frame. Such a goal requires the definition of geological domains and their associated boundaries. Moreover, the geological boundary locations are uncertain and can be one of the primary sources of risk in a mining project, and because of that, they must be assessed. The definition of geological domains and their associated boundaries is the first-order decision and will influence all subsequent steps in the mineral resource evaluation process. Thus, a poor definition of geological domains and their associated boundaries could deteriorate the accuracy of the resource estimates. This research proposes an uncertainty quantification method to assess and visualize the uncertainty of 3D geological boundaries. The proposed method is fully automated, reproducible, data-driven, geological-driven, variogram modeling-free, and account for the uncertainty in the parameters. It includes many methods from geostatistics, statistics, machine learning, and data science. The proposed method is applied to a porphyry copper deposit.