Previous research related to the optimization of mining operations has predominantly focused on generating a life-of-mine production schedule that maximizes the discounted cash flows of the material extracted and products produced. Stochastic optimization models address the issue of integrating uncertainty into the decision-making, leading to mine designs and production schedules with higher value and better risk management, thus helping to ensure that the mining operation is capable of meeting production targets over time. More recent models address the challenge of stochastic global optimization, which aim to holistically optimize a mining complex, from the production schedule, through to the products created, marketed and sold. Existing stochastic formulations, however, assume that the bottlenecks in the mining complex, such as mine production and milling capacities, have been defined a-priori, thus ignore the impact that the quantity and timing of capital expenditures required to create these capacities may have on the overall profitability of the operation.
This work builds on previous developments in stochastic global optimization for mining complexes and integrates capital expenditure options in order to appropriately design the bottlenecks or constraints in the model. This formulation is solved using a combination of the particle swarm optimization and simulated annealing algorithms. An application for a copper mining complex demonstrates the ability to decide when to invest capital in order to increase the number of both trucks and shovels used. The results indicate that the stochastic optimizer is able to outperform its deterministic-equivalent by significantly reducing the risk associated with materials sent to the mill, in addition to an overall increase in net present value by 5.7%.
Published September 2015 , 26 pages