Modeling power failures in deep-level mines to assist with emergency dewatering planning

Abstract

A mining complex’s intricate dewatering system removes water that enters the shaft through various means. This system comprises a series of water pumps that require a constant power supply to avoid flooding key areas of the mine. Due to the instability of Eskom’s infrastructure and power supply, mines need to plan for unforeseen power failures. The problem addressed in the study is that mines are unsure of how the dewatering system will respond to a total power failure. It is further not feasible to conduct a real-life test on the system’s response as human lives and expensive equipment will be put at risk. To address the problem, a three-dimensional thermohydraulic simulation tool, namely Process Toolbox, was used to model an integrated dewatering system with dynamic boundaries. The simulation was developed to challenge the existing industry method that excludes dynamic boundaries and only assesses components individually; therefore, it does not integrate a holistic system. A verified, integrated, and dynamic dewatering system was developed in Process Toolbox to replicate Mine A’s operating conditions. The dewatering simulation model was built and calibrated using data obtained during physical investigations (pump tests and ultrasonic flow meter readings) and digital investigations (data) to ensure the simulation components could replicate real-world operating conditions to within a mean absolute error of 5%.

Mine A determined that a 48-hour period would be sufficient for restoring power or implementing alternative plans to avoid flooding. Therefore, a 48-hour requirement with no pumping after a power failure had to be achieved. Critically, the results of the integrated dynamic simulation tool indicated that the lowest pumping level would flood after 22 hours. This opposed the existing method’s result of 67 hours, which was a cause for concern given that the industry method’s error is 67% when compared with the simulation method. The results further illustrated the importance of accurate simulation methodologies. The mine would have lost equipment and put lives in danger if the 67-hour calculation time was supported and an incident occurred. The modelling environment provided a platform for testing contingency planning hypotheses through scenario simulations. Four scenarios were developed through 200 simulations and iterations. Each scenario built on previous scenarios until the 48-hour requirement was met. This type of scenario testing is not possible with the current method. The study objective was met by providing verified, dynamic, and integrated simulation-backed results based on semi-empirical data. The solution provided a reliable platform to aid decision-making, scenario investigations, and process predictions – factors the industry method could not provide.