Optimal Determination Of Dig Limits For Improved Grade Control

Ore and waste cannot always be visually separated. The distinction between ore and waste may need to be made on the basis of blasthole sampling. Hand drawn dig limits or computer-generated contour lines do not account for uncertainty, economics or equipment. Geometric methods for interpreting grade information such as the polygonal or triangulation methods share the same limitations as contouring. Geostatistical techniques exist for the realistic modeling of grade variations and uncertainty. Decision analysis techniques exist for the binary classification of ore, and waste blocks in the presence of this uncertainty. Block-by-block classification is not enough; mining equipment cannot exercise free selection of blocks. There is a need for procedures to determine optimal dig limits that account for (1) the uncertainty in the grades, (2) the economics of risk-qualified decision-making, and (3) the ability of the mining equipment to dig to specific limits. This paper presents a procedure for optimal determination of dig limits that account for these constraints. Simulated annealing (SA) is used with modest computational requirements because the number of variables is relatively small. The solution of the optimization problem is dig limits that maximize profit under multiple constraints. The method is illustrated with different constraints and tested against hand drawn dig limits in a "Dig Limit Challenge," which illustrates the efficacy of the new method.