This research presents a comprehensive study that focuses on optimizing the operational aspects of underground mining processes through a sophisticated modeling framework. The study introduces a Stochastic Mixed Integer Linear Programming (SMILP) framework that evolved from an initial Mixed Integer Linear Program (MILP) model. This advanced model considers a wide range of mining operations, including primary (decline) development, ventilation development, operational development, ore pass development, stope design and extraction, as well as backfilling and stockpile management. The primary aim of this research is to enhance the economic efficiency and sustainability of mining projects by optimizing the extraction and management of mineral resources. The research objectives encompass the following key aspects: a) a comprehensive framework has been proposed and developed to address production scheduling optimization while considering primary, ventilation, and operational developments, as well as backfilling, ore pass, and stockpile management; b) accounting for geological uncertainties, the research delves into techniques for incorporating stochastic variables, such as ore grade, into the optimization framework. This addition ensures a more realistic representation of the mining environment and its associated uncertainties; c) the framework has been extended to maximize the Net Present Value (NPV) of an integrated stope design and production scheduling optimization. This approach ensures that the economic viability and profitability of mining projects are at the forefront of decision-making processes. The research encompasses a holistic exploration of these objectives, culminating in the development and validation of the SMILP model.
The successful integration of mathematical modeling and optimization techniques into the mining industry addresses the challenges posed by geological uncertainties, infrastructure development, extraction methodologies, and resource management. The execution of the research was facilitated by the effective implementation of all constraints and objectives using the MATLAB programming language. Complex problem-solving was achieved through the utilization of IBM ILOG CPLEX, a powerful optimization solver.
To investigate the practicality and performance of the proposed model, six comprehensive case studies were conducted. These case studies aimed to analyze and compare the economic and operational outcomes when employing the SMILP model. The primary performance metric used in these case studies was the NPV, a crucial indicator of a project's economic viability. Among the case studies presented, the SMILP model with stockpile management (Case 4) exhibited superior performance, reporting the highest NPV when compared to the other five cases. Case 4 achieved a 25-year mine life production schedule with a generated NPV amounting to $8,077.86 M, while the MILP model with stockpile management (Case 2) yielded $7,801.20 M, indicating a 4% increment in financial gain by the SMILP model. These findings underscore the effectiveness and practical applicability of the developed model in real-world mining scenarios.
In conclusion, this thesis represents a substantial contribution to the field of mining engineering and operations research. It offers a robust and versatile framework for optimizing underground mining processes, considering various uncertainties and constraints. The research outcomes have the potential to significantly enhance the economic prospects of mining projects, making them more sustainable and profitable. This work not only contributes to the advancement of mining science and engineering but also holds promise for improving the efficiency and sustainability of mining operations in the real world.