generation, as well as significantly decreases the size of potentially AMD-generating wastepile. Due to the increased complexity of the optimization problem, the metaheuristic simulated annealing solving technique is utilized. Overall, the case study on a gold mining complex displays a 16% reduction in total waste generated and a 6% increase in NPV while significantly decreasing the risk of not meeting environmental constraints. While the previously mentioned study optimizes only the quantity of PAG waste produced, more recent work by Levinson and Dimitrakopoulos (2024) utilizes a non-linear stochastic formulation that aims to mitigate the risk of AMD at the source. This is done by utilizing a simulation of the geochemical properties of extracted blocks and using this information to blend waste material to achieve a stable mixture. By eliminating future monitoring expenses and the need for costly and potentially hazardous water cover, the long-term risk of environmental damage is mitigated while improving the cost structure of the operation. Additionally, progressive reclamation is scheduled simultaneously to extraction to accelerate reclamation of the waste dumps and enhance environmental performance. The methodology implies co-disposal of PAG and NAG waste to achieve a net production (NPR) ratio between neutralization and acid generation potential of above 2, which is deemed satisfactory to prevent AMD. In addition to the novel formulation, the researchers investigate the use of a machine learning algorithm to reduce computation time. The comparison between the base case and the optimized scenario shows a 52.5% reduction of cells with unsafe NPR ratios. Simultaneously, it is shown that a large part of the reclamation efforts can be finalized during the production period without significant effects on the NPV of the project. Based on the reviewed methodologies, which technical aspects are summarized in Table 1, certain trends can be observed. One of the most notable developments in the field is the increased scope of the evaluation, with earlier studies focused on waste management in single pits based on the simulated grade of the material extracted. A gradual shift can be observed to optimize larger mining complexes with the addition of geochemical variables, leading to more comprehensive methodologies. This, however, results in much more complex optimization problems, greatly extending the computation time. To address this issue, the use of metaheuristics and machine learning solving techniques is becoming more prevalent. Table 1 – Summary of waste management methodologies Author Year Objective function Planning scope Model type Solution methodology/Solver Spleit and Dimitrakopoulos 2017 Maximize NPV, minimize trucking cost, penalize deviations from production targets Longterm SMIP CPLEX, sequential optimization Rimélé et al. 2018 Maximize DCF, penalize deviations from production targets Longterm SMIP Heuristics (sliding time window)
RkJQdWJsaXNoZXIy MTM0Mzk2