Long-term extraction sequence optimization is an essential step in the development of mining projects. However, for dredge mined heavy mineral sands deposits this process still done manually as available software solutions are tailored to conventional open pit mines and are not applicable due to the constraints that come with dredge mining and the heavy mineral sands market. In this research an extraction sequence optimization model was developed that can take into account the mining and market constraints that come with the dredge mining of heavy mineral sands, based on flexible meta-heuristic methods. A comparison between the ant colony optimization and genetic algorithm methods showed better performance for the genetic algorithm. The genetic algorithm was validated using synthetic validation models and applied to a real dataset in a case study. The results show that the developed genetic algorithm model can take into account the necessary objectives of a spatially continuous mining path, maximum capacity of extraction and processing equipment, and fixed product offtake rates. The genetic algorithm model showed good performance and consistency at identifying the mining direction and the optimization of production and stockpiling profiles. However, errors between the optimized and ideal extraction sequences are still present in the validation study and the extraction paths show significant variance between different model runs. The developed model is in its current form a potentially useful tool for the selection of equipment and optimizing operational parameters such as the offtake rate. Through further work the accuracy of the optimization model can be improved and the variance in results be reduced. Furthermore, through extension of the optimization scope a broader optimization model could be created that does not just save time and labour, but also provides additional insights into risk and robustness. Additionally, through the implementation of additional factors a model could be created that not just optimizes the mining operation economically, but simultaneously minimizes the negative impacts of mining.

dataset: https://doi.org/10.4121/1d3885ff-dc9c-4d05-9e4d-2d8f562d9c8d ... Read more

This study investigated the application of co-indicator kriging (Co-IK) in combination with multiple indicator simulation (MIS) for the modeling of the Phase 7 high-sulfidation epithermal gold–silver deposit. This approach was selected in response to the deposit’s pronounced grade variability, intricate alteration zoning, and multi-metal mineralization patterns, which collectively result in highly skewed, non-gaussian grade distributions. By integrating Co-IK and MIS, the objective was to develop probabilistic, domain constrained models that accurately represent the spatial variability of the deposit while mitigating the smoothing bias inherent in traditional kriging methods. ... Read more