Details

Autor(en) / Beteiligte

• Sari, Yuksel Asli

Titel

Novel Optimization Models for Surface and Underground Mine Planning

Ort / Verlag

ProQuest Dissertations & Theses

Erscheinungsjahr

2019

Quelle

ProQuest Dissertations & Theses A&I

Beschreibungen/Notizen

• Mine planning and optimization affect efficiency, profitability and productivity of operations significantly. Low commodity prices, high resource degredation maintenance costs and high fixed infrastructure costs necessitate the use of optimal decision making tools for mining companies to make profit. All mines have different characteristics and planning phases. In this research, different optimization problems that suit various mining techniques and planning stages are studied. In essential, there are two types of mining: surface mining and underground mining. Surface mining operations are generally long-term because overburden must be removed to access the profitable orebody. This requires strategic long-term planning at the feasibility stage. The first publication in the scope of this research focuses on long-term surface mine planning with environmental considerations. The provided solution optimizes the problem using mixed integer linear programming (MILP). When operation starts and bench sectors are mined on a daily basis, the need for short term planning arises. The second publication addresses the dig-limit optimization problem, which is an important part of short-term planning. With the proposed MILP optimization method, the ore-waste boundaries are delineated with the equipment size constraints. Although underground mining also starts with exploration and resource estimation/simulation stages, the problems that need to be addressed are very different from surface mining techniques and it has its own unique challenges. Special focus is given to the sublevel stoping underground mining technique. Stope optimization is a complex problem, comprised of two sub-problems: stope layout optimization and stope sequencing. MILP formulations of stope layout optimization are impractical because of the large size of the problem. In the third and fourth manuscripts, two different heuristic stope layout optimization algorithms are presented where the former uses a clustering heuristic to identify stopes with high grade concentration and the latter uses a greedy heuristic based on dynamic programming to solve the sub-problems and explore the promising stope combinations. Fifth manuscript tailors the greedy heuristic algorithm to poly-metallic mines with pillars. Both heuristic approaches are shown to be near-optimal through comparing with developed novel MILP formulations case studies in smaller problem instances. When the stope layout is finalized, the sequence can be optimized to yield the optimal project value. In the sixth and final manuscript within the scope of this research, the stope sequencing problem is formulated in MILP. To account for risk emerging from geological uncertainties, chance constrained programming is implemented. This approach maximizes the expected net present value of the operation while minimizing the deviations from the expected value due to ore grade uncertainty. It focuses the search on a unique direction based on the specified desired project risk level.