Automatic Determination Of Dig Limits Subject To Geostatistical, Economic, And Equipment Constraints

Grade control in open pit mining requires the specification of dig limits that account for mineral grades, economic costs, and selectivity of mining equipment. Visual identification of ore rock types is ideal, but this is not always possible in lower grade disseminated deposits where ore/waste contacts are not visually discernible. In this case, conventional grade control practice consists of a two-step procedure (1) create a map of mineral grades at some selective mining unit scale, and (2) determine practical ore-waste boundaries or dig limits on the basis of the gridded block grades or assay information. This procedure is laborious, depends on a subjective assessment of where the boundary should be, and may be economically sub-optimal. We pose the determination of dig limits as an optimization problem and solve that problem with the technique of simulated annealing. Simulated annealing has the unique advantage of being able to combine multiple non-linear constraints into a single objective function. We use maximum profitability and the ability of the equipment to mine the proposed dig limits as constraints in the determination of optimal dig limits. We require a map of expected profit for eah block. Geostatistical techniques are recommended, for mapping expected profit but not necessary. Geostatistics will provide a quantification of the uncertainty in the grades within rock types using all available blasthole samples and exploration drilling. Some variant of L-optimal estimation or kriging can be used to determine the block-by-block classification that is economically optimum. We also need the expected profit for each block. It is unrealistic to assume free selection, that is, each block cannot be extracted independently of its neighbors. The optimal balance of ''accepting dilution'' and ''wasting ore'' is achieved to maximize profit subject to equipment constraints.