Abstract

The intrusion of wet muck is an ongoing challenge in underground mines, along with the potential for mud rushes, which pose risks to workers, equipment, and infrastructure. This paper describes the development of a probabilistic model for short-term mine planning that estimates high- and medium-risk zones for mud events. Here, a risk level classification model is proposed based on a logistic regression model and uses operational, environmental, topographic, and lithological variables to identify high and medium-risk zones for mud events in the Diablo Regimiento and Esmeralda mines. Operational variables include uniformity and extraction rate, while environmental and topographic variables include the distance between drawpoints and the surface, moisture at drawpoints, and precipitation. The percentage of broken material-obtained from gravity flow simulations-is also incorporated. The results highlight the precision and predictive capability of this type of model, predicting 86% of mud events at high risk In the Diablo Regimiento sector and 78% of mud events at high risk in the Esmeralda Bloque 1 & 2 sector. Furthermore, this model successfully determined the percentage of the operational footprint in high-risk zones in the Diablo Regimiento sector as well as identifying an increase in high-risk zones in the Esmeralda Bloque 1 & 2 sectors. This risk modeling approach emerges as a fundamental tool for short-term planning and decision-making to minimize risks associated with mud events in Block Caving mines. The robustness and predictive capacity of this type of model support its applicability in future sectors facing similar challenges related to the occurrence of mud events.