Desarrollo de un modelo predictivo a tiempo real para la estimación del nivel de bolas en los molinos de la planta SAG, CODELCO División El Teniente.
Loading...
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universidad de Concepción
Abstract
El Teniente, la mayor mina subterránea de cobre del mundo, ha experimentado un incremento continuo en el procesamiento de minerales. Este aumento no ha sido respaldado adecuadamente por mejoras en el proceso, resultando en un aumento en la granulometría del producto de la molienda, lo que afecta negativamente la eficiencia de la flotación.
El objetivo general de este estudio es desarrollar una herramienta en línea que estime el nivel de bolas en el circuito de molienda SAG utilizando un modelo de potencia. Para lograr esto, se plantearon objetivos específicos como realizar mediciones en terreno para ajustar el modelo y caracterizar el muestreo del año 2020 del molino SAG 401, con el fin de simular la función de selección específica y justificar el control del nivel de bolas.
La propuesta de valor de esta implementación radica en mejorar el control operativo y obtener mejores resultados, además de comprender mejor la eficiencia de la molienda y la fractura bajo diferentes condiciones operativas. Esto también permitirá un mejor control en la carga de bolas. Durante el desarrollo del estudio, se seleccionó el modelo de Hogg & Fuerstenau para la estimación del nivel de bolas. Se realizaron mediciones en terreno y calibraciones para ajustar el modelo, y se incluyeron modelos de desgaste y relaciones metalúrgicas para mejorar la precisión.
Los resultados demostraron que el modelo es efectivo en la estimación del nivel de bolas tanto para los molinos de bolas como en los SAG, alcanzando la mejor precisión de 95,8% de precisión en las mediciones durante las calibraciones correspondientes para el SAG 401 y la peor de 79% en el molino SAG 501, donde en el molino 501 se implementó al modelo, el desgaste del cilindro, demostrando que es un factor crítico en la sensibilidad del modelo. La implementación de un modelo en línea para estimar el nivel de bolas en la planta SAG es viable y puede mejorar significativamente la eficiencia del proceso de molienda. Sin embargo, es crucial realizar calibraciones y ajustes continuos para asegurar la precisión del modelo.
El Teniente, the largest underground copper mine in the world, has experienced a continuous increase in mineral processing. This increase has not been adequately supported by process improvements, resulting in a rise in the granulometry of the milling product, negatively affecting flotation efficiency. The main objective of this study is to develop an online tool to estimate the ball level in the SAG milling circuit using a power model. To achieve this, specific objectives included conducting field measurements to adjust the model and characterizing the 2020 sampling of the SAG 401 mill, in order to simulate the specific selection function and justify ball level control. The value proposition of this implementation lies in improving operational control and achieving better results, as well as gaining a better understanding of milling efficiency and fracture under different operational conditions. This will also enable better control over ball loading. During the development of the study, the Hogg & Fuerstenau model was selected for estimating ball level. Field measurements and calibrations were carried out to adjust the model, and wear models and metallurgical relationships were included to improve accuracy. The results demonstrated that the model is effective in estimating ball levels for both ball mills and SAG mills, achieving the highest accuracy of 95.8% during the corresponding calibrations for SAG 401, and the lowest at 79% for SAG 501, where cylinder wear was implemented in the model, demonstrating that it is a critical factor in the model’s sensitivity. The implementation of an online model to estimate ball level in the SAG plant is feasible and can significantly improve the efficiency of the milling process. However, it is crucial to perform continuous calibrations and adjustments to ensure model accuracy.
El Teniente, the largest underground copper mine in the world, has experienced a continuous increase in mineral processing. This increase has not been adequately supported by process improvements, resulting in a rise in the granulometry of the milling product, negatively affecting flotation efficiency. The main objective of this study is to develop an online tool to estimate the ball level in the SAG milling circuit using a power model. To achieve this, specific objectives included conducting field measurements to adjust the model and characterizing the 2020 sampling of the SAG 401 mill, in order to simulate the specific selection function and justify ball level control. The value proposition of this implementation lies in improving operational control and achieving better results, as well as gaining a better understanding of milling efficiency and fracture under different operational conditions. This will also enable better control over ball loading. During the development of the study, the Hogg & Fuerstenau model was selected for estimating ball level. Field measurements and calibrations were carried out to adjust the model, and wear models and metallurgical relationships were included to improve accuracy. The results demonstrated that the model is effective in estimating ball levels for both ball mills and SAG mills, achieving the highest accuracy of 95.8% during the corresponding calibrations for SAG 401, and the lowest at 79% for SAG 501, where cylinder wear was implemented in the model, demonstrating that it is a critical factor in the model’s sensitivity. The implementation of an online model to estimate ball level in the SAG plant is feasible and can significantly improve the efficiency of the milling process. However, it is crucial to perform continuous calibrations and adjustments to ensure model accuracy.
Description
Tesis presentada para optar al título de Ingeniero Civil Metalúrgico
Keywords
Molinos de bolas, Minería, Aplicaciones web, Mina El Teniente (Chile)