Thesis Implementación de algoritmos de optimización de leyes de corte para la maximización de van en depósitos mono-metálicos y en bimetálicos con estrategia de acopio de mineral
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Program
Ingeniería Civil de Minas
Campus
Campus Santiago San Joaquín
Abstract
En una operación minera, el elegir correctamente la ley de corte es fundamental para así asegurar el potencial económico que esta posee. Esta decisión permite distinguir entre el material minado que aporta valor económico de aquel que no lo hace. En 1964, Lane propuso un algoritmo que marcó un hito en este ámbito, al presentar un modelo diseñado para maximizar las ganancias de explotaciones mineras aplicadas a un mineral. No obstante, este enfoque resulta limitado cuando se trata de escenarios más complejos, como la extracción simultánea de varios minerales o la posibilidad de almacenar temporalmente el mineral extraído. Este estudio parte de la hipótesis de que aplicar una política de ley de corte que varíe a lo largo del tiempo, junto con la opción de acopio, puede generar un mayor Valor Actual Neto (VAN) en comparación con estrategias más rígidas que usan leyes de corte constantes. En base a esto, el objetivo principal fue desarrollar e implementar un algoritmo en Python capaz de optimizar el VAN en operaciones mineras que involucran dos minerales, considerando además el uso de acopios Para ello, se llevó a cabo una revisión del modelo de Lane y sus posteriores desarrollos, se seleccionaron casos de estudio relevantes y se implementó el algoritmo tanto en hojas de cálculo como en código Python. Posteriormente, se compararon diferentes estrategias: con y sin uso de acopios, y con leyes de corte fijas frente a leyes variables. La evaluación se centró en los resultados económicos y el impacto en el VAN. Los resultados indican que las estrategias con leyes de corte variables ofrecieron un mejor rendimiento económico, aumentando el VAN en un 6,39 % en el primer caso de estudio y en un 4,43 % en el segundo. Además, se comprobó que el uso de acopios no solo mejora la eficiencia del proyecto, sino que también permite extender la vida útil de la mina.
In a mining operation, correctly selecting the cut-off grade is essential to ensure the project's economic potential. This decision allows for distinguishing between the mined material that adds economic value and that which does not. In 1964, Lane proposed an algorithm that marked a milestone in this field by introducing a model designed to maximize profits from mining operations focused on a single mineral. However, this approach becomes limited when applied to more complex scenarios, such as the simultaneous extraction of multiple minerals or the possibility of temporarily stockpiling extracted ore. This study is based on the hypothesis that applying a time-dependent cut-off grade policy, along with the option to stockpile material, can generate a higher Net Present Value (NPV) compared to more rigid strategies that use constant cut-off grades. Based on this, the main objective was to develop and implement an algorithm in Python capable of optimizing the NPV in mining operations involving two minerals, also considering the use of stockpiles. To achieve this, a review of Lane’s model and its later developments was conducted, relevant case studies were selected, and the algorithm was implemented both in spreadsheets and Python code. Various strategies were then compared: with and without the use of stockpiles, and with fixed versus variable cut-off grades. The evaluation focused on economic outcomes and the impact on NPV. The results indicate that strategies using variable cut-off grades offered better economic performance, increasing the NPV by 6.39% in the first case study and 4.43% in the second. Additionally, it was confirmed that stockpiling not only improves the project's efficiency but also extends the mine’s operational life.
In a mining operation, correctly selecting the cut-off grade is essential to ensure the project's economic potential. This decision allows for distinguishing between the mined material that adds economic value and that which does not. In 1964, Lane proposed an algorithm that marked a milestone in this field by introducing a model designed to maximize profits from mining operations focused on a single mineral. However, this approach becomes limited when applied to more complex scenarios, such as the simultaneous extraction of multiple minerals or the possibility of temporarily stockpiling extracted ore. This study is based on the hypothesis that applying a time-dependent cut-off grade policy, along with the option to stockpile material, can generate a higher Net Present Value (NPV) compared to more rigid strategies that use constant cut-off grades. Based on this, the main objective was to develop and implement an algorithm in Python capable of optimizing the NPV in mining operations involving two minerals, also considering the use of stockpiles. To achieve this, a review of Lane’s model and its later developments was conducted, relevant case studies were selected, and the algorithm was implemented both in spreadsheets and Python code. Various strategies were then compared: with and without the use of stockpiles, and with fixed versus variable cut-off grades. The evaluation focused on economic outcomes and the impact on NPV. The results indicate that strategies using variable cut-off grades offered better economic performance, increasing the NPV by 6.39% in the first case study and 4.43% in the second. Additionally, it was confirmed that stockpiling not only improves the project's efficiency but also extends the mine’s operational life.
Description
Keywords
Ley de corte, Acopio de mineral, Valor Actual Neto (VAN)