Aggregation and Mathematical Programming for Long-Term Open Pit Production Planning

• Author / Creator

  Tabesh, Mohammad

• The objective of this thesis is to develop, implement and verify a theoretical framework based upon aggregation and mathematical programming for solving the long-term open pit production planning problem. The goal is to closely estimate the maximum net present value of the operation by providing an optimum and practical mining, processing and stockpiling schedule for the open pit mining operation while respecting the technical and operational constraints. As stated by many researchers in the area and illustrated in the forth chapter of this thesis, using blocks as units of planning results in over-estimation of the operation’s profitability and flexibility. Thus, we introduced a clustering algorithm along with a mathematical formulation that can result in good production plans that result in high NPV, are practical and do not under- or over-estimate the value of the operation. In this thesis, we introduced, implemented and verified a specifically-designed clustering technique based on agglomerative hierarchical clustering, in order to aggregate blocks into mining-units that are homogenous in rock type and grade, and have mineable shape and size. We designed the algorithm, developed the codes and implemented and tested the algorithm on small test datasets and large real-size deposits to evaluate the performance of the algorithm. We showed that we can balance the clustering control parameters to obtain clusters of blocks aligned with the clustering purpose such as long-term planning units, ore polygons and blast patterns. Moreover, we formulated, implemented and verified a mixed integer linear programming model, for long-term open pit production planning problem, which can use two different sets of units for making mining and processing decisions. Our model is able to simultaneously determine the optimum stockpiling strategy and the optimum mining and processing schedule in reasonable processing time. We implemented our model on a small test dataset as well as real-size deposits to understand the effects of using different units of planning for making mining and processing decisions. We showed that we can obtain practical and optimum production schedules for real-size open pit mines in a reasonable time by benefiting from the clustering technique we introduced. Moreover, we showed how incorporating stockpile optimization in long-term production scheduling can increase the net present value of the operation. We benchmarked our model against commercial scheduling software to illustrate the flexibility and accuracy of our planning approach. The main contributions of this thesis to the mining body of knowledge are (i) introducing a clustering algorithm that creates aggregates of blocks homogenous in rock type and grade, with controlled shape and size, and respects the mining direction as well as other constraints and boundaries, (ii) introducing an MILP formulation for the long-term open pit production planning problem, with dynamic cut-off grade, that maximizes the net present value of the mine, by using two different units for making mining and processing decisions, while respecting operational and technical constraints, and (iii) incorporating stockpiling in the long-term scheduling to simultaneously optimize the mining, processing and stockpiling strategies.

• Subjects / Keywords

  • Mining aggregates
  • Mixed integer linear programming
  • LTOPP
  • Long-term open pit production planning
  • Clustering

• Graduation date

  Fall 2015

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/R3N01050C

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.