Application of a new approach for drillholes location optimization

Abstract

ABSTRACT Geostatistics has seen limited application to the problem of determining drillhole locations, and both approaches in use have been successful in improving drilling strategies. The first approach in use is a sequential optimization scheme. The effect of drilling each candidate drillhole is evaluated, and the one yielding the largest decrease in the estimation variance is chosen. The second approach attempts to identify regions of high estimation variance and locates drillholes in these regions. A number of different sets of drillholes can then be evaluated and the best set selected. Here, the computer implementation of a third approach is discussed, and computational results are presented applying all three approaches to the same property. This new approach guarantees that the set of drilling sites chosen will be the best possible set from among the given candidate sites. Two models have been programmed, representing differing definitions of the “best” set of locations. In the first model, the best set of locations is that which minimizes the estimation variance associated with the reserve estimates. This optimization is subject to a budget limitation on the cost of drilling or a maximum number of holes. The second model defines the best set of locations as that which minimizes the drilling cost or number of holes, subject to meeting some maximum limit on the estimation variance. This marriage of geostatistics with traditional operations research techniques is applied to a deposit in the development stage. Comparisons are made between the new approach and the more traditional geostatistical approaches to the drillhole location problem. This paper highlights the technique involved and summarizes the results of the test study.