Implications of Algorithm and Parameter Choice: Impacts of Geological Uncertainty Simulation Methods on Project Decision Making

Abstract

Uncertainty in geological models and grade uncertainty are two major contributors to the total resource uncertainty of a mining project. Previous attempts at determining uncertainty in geological models using methods such as MPS (multiple-point statistics), SIS (sequential indicator simulation), and multiple applications of RBF (radial basis functions) with different parameters have shown that it is nontrivial; the uncertainty profiles are dependent on the method and the parameters selected. Most of the methods tested require additional information in the form of either local probabilities or proportions derived from the existing geological interpretation or a conceptual geological model in the form of a training image. This makes some methods amenable to use in the early stages of a project because the method allows for a more complete testing of different geological concepts. In later stage projects where there is an increased level of confidence (due to the amount of data collected) in the geologic interpretation, methods that achieve ranges of uncertainty around the interpretation likely provide a more realistic assessment of uncertainty. This paper details the continuation of research into geostatistical tools suitable for the evaluation of geological uncertainty in order to further understand the intricacies of the methods and the impact of the technique on the resulting uncertainty profile. Suggestions of which methods to use based on the amount of geological information available are provided.