Quantifying Wet Muck Entry Risk for Long-term Planning in Block Caving

Abstract

Wet muck entry is one of the major geotechnical risks associated with long-term production goals in cave mining. The objective of this research is to quantify the effect of the main risk variables related to wet muck entry in an effort to prioritize and confront these variables accordingly. Logistic regression modeling was carried out using mine data from Pipa Norte Mine (PNM) and Sur Andes Pipa Mine (SPM), both located at El Teniente Mine. A confusion table was employed to calibrate the model, while a scatter plot and an error relative frequency histogram were used to validate it. The results indicated that ore draw and environmental conditions are the main risk variables associated with wet muck entry. The aforementioned metrics show an acceptable agreement between mine and modeled data. Hence, we used these results to create a statistically significant predictive model, which may be useful for the risk evaluation of numerous long-term plans. Based on optimal calibrated conditions, the predictive model is a powerful instrument not only to identify high-risk zones susceptible to wet muck entry but also to make long-term preventive decisions, which could lead in mitigating the risks of wet muck entry in cave mines.