FINANCIALLY EFFICIENT ORE SELECTIONS INCORPORATING GRADE UNCERTAINTY

Abstract

Traditional mining selection methods focus on local estimates or loss functions that do not take into account the potential diversification benefits of financial risk that is unique to each location. A constrained efficient set model with a downside risk function is formulated as a solution. Estimates of this nonlinear mixed-integer combinatorial optimization problem are provided by a simulated annealing heuristic. A utility framework that is congruent with the proposed efficiency model is then used to choose the optimal set of local mining selections for a decision-maker with specific risk-averse characteristics. The methodology is demonstrated in a grade control environment. The results show that downside financial risk can be reduced by around 33% while the expected payoff is only reduced by 1% when compared to ore selections generated by traditional cut-off grade techniques.