ABSTRACT
Appropriate geological knowledge is a determinant factor in mining profitability. Variability forecasting of geological attributes reduces mining risks. Traditional mine planning does not take into account the uncertainty associated with the geological model. Generally, the model used for planning is built using traditional estimation techniques, which do not provide a measure of local variability, being inadequate for short range mine planning. Opposite to the traditional techniques, geostatistical simulation reproduces the actual variability and spatial continuity (histogram and variogram) of the original data set. Stochastic simulation can be used to measure the uncertainty, which is associated with an estimate. This study presents an algorithm for conditional sequencial simulation and describe a procedure to assess geological uncertainty. Subsequently, it is introduced a framework proposing an efficient way to use this information in mine planning. A case study in a major Brazilian coal mine illustrates the steps involved.
