Geostatistics approach for uncertainty assessment and resource classification
Understand in one day what is really at stake in reporting resources and gather practical information on geostatistical capabilities in that matter
At the different stages of the resource evaluation of mining projects, the best estimates of tonnages and grades from the available drill holes data have to be delivered. It is also required to attach to these estimates indications of their confidence. Geostatistical models are developed in a probabilistic framework, so they are particularly adapted to offer a satisfactory solution to that issue, keeping in mind the following:
- One cannot be satisfied by only using criteria based on kriged blocks estimates, like the kriging variance because of the difficulty of combining them on bigger volumes;
- What shall we do when non-linear estimates like MIK or UC arz used to estimate tonnages and grades above cut-offs but do not provide error variances?
The close link between the resource classification and the support is a key point that cannot be ignored. Additional efforts must be made to obtain meaningful confidence intervals at different scales. Practical solutions to calculate confidence intervals from Gaussian transforms or from conditional simulations are presented and discussed.
Who should attend
The course is aimed at geologists, engineers and project managers who want to deeply understand what is really at stake in reporting resources and gather practical information on geostatistical capabilities in that matter.
This course is focused on mining issues:
- Geostatistical measures of the quality of kriged block estimates;
- Uncertainty on recoverable resources estimates;
- How to report resources according to mining codes;
- Resource Classification from simple criteria to confidence intervals from simulations.