Open pit mine planning with stockpiling

Abstract

This dissertation consists of three papers; the first is published in European Journal of Operational Research, the second is nearing submission to it Optimization and Engineering, and the third is nearing submission to International Journal of Mining, Reclamation and Environment. These papers apply operations research techniques to open pit mine production scheduling with stockpiling (OPMPS+S). The first paper, "Linear Models for Stockpiling in Open-pit Mine Production Scheduling Problems," reviews existing models to solve OPMPS+S and shows that a nonlinear-integer model provides an exact solution but is intractable even for medium-size data sets. Then, we present an approximation to that nonlinear-integer model, solve the nonlinear-integer and proposed models for multiple data sets and show that the latter model provides solutions very close to those provided by the nonlinear-integer model. By pairing this novel formulation with recently developed linear programming algorithms and heuristics, we demonstrate a dramatic decrease in solution time. The second paper, "Comparison of an Improved Open Pit Mine Stockpiling Model with Commercial Software," introduces an extension to the model presented in the first paper to blend material with multiple grades and a contaminant in the stockpile. Existing state-of-the-art algorithms which exploit the problem structure produce integer solutions with associated net present value at least 18.7% higher than that provided by commercial of-the-shelf software. The third paper, "Open Pit Mine Planning with Degradation Due to Stockpiling," presents three new models that consider degradation due to stockpiling, compare them with models without degradation, and show that degradation has a major impact on the value a stockpile provides.