Temperature Prediction Model in the Main Ventilation System of an Underground Mine

Abstract

A model to forecast the underground temperature in a mine ventilation circuit was developed on the basis of a case study and actual data describing temperature, airflow, and drift length collected over several years. A mathematical model featuring seven variables with interactions provided reliable predicted temperatures, achieving a correlation of R2 = 0.933 with an estimation error of ±2 °C. Its soundness was proven using both the node-to-node analysis and the multi-node approach. The multi-node approach was shown to be an interesting option to model underground mining environments. This model can be very useful to predict the temperature evolution along the main ventilation system, determine the best workplace conditions in terms of temperature, and analyze different planning scenarios of the mine. Moreover, some recommendations are presented for obtaining reliable data when using temperature sensors and the model in a U-shaped ventilation system.