Developing a digital twin for addressing complex mine ventilation problems

Abstract

The ventilation system of a deep-level mine is complex and dynamic, which makes it challenging to identify changes in such a system and even more so to predict the effect of these changes with conventional approaches. An integrated approach is therefore required to analyse such complex scenarios while considering numerous variables. The purpose of this study is to derive a method for the development and application of a digital twin model for a deep-level mine ventilation system using such an integrated approach. Developing a digital twin model for any system is a time-consuming process because of the complexity of the network and the numerous variables that must be considered. The methodology developed in this study allows setting up a model that includes an in-depth explanation of the individual components that will be calibrated and, ultimately, lead to the development of a calibrated digital twin that can be used to identify key problems in a ventilation system. A case study on a deep-level mine was used to evaluate this methodology. In the case study, the main underground booster fans were operating at a higher differential pressure than that of the original blueprint model. This indicated primarily that there were restrictions within the return airways of the mine. Before applying this methodology, these areas were inaccessible due to high air temperatures and deteriorating ground conditions. However, the digital twin model provided helped to identify the restricted areas and predict the potential impact on the entire ventilation system by addressing these restrictions. This information enabled a clear understanding as well as a solution for mining personnel to address the identified restrictions. Furthermore, the model was used to determine a ventilation strategy that enabled mining personnel to access these previously inaccessible areas. As a result, the restrictions were removed, which yielded an increase of 7% in the total airflow through the mining block. Additionally, the digital twin model predicted the improvement in airflow to an accuracy of 96% when compared to measured results. Therefore, this study highlights the value of using a digital-twin model to solve complex problems within deep-level mine ventilation systems. Ultimately, the digital twin model could be used successfully to identify the airway restrictions and managed to predict the total system impact and provide the optimum solution to solve this complex problem. This illustrates that the objectives of this study were achieved by implementing the derived methodology.