As Luossavaara-Kiirunavaara Aktiebolag advances the development of its underground iron ore mines to depths exceeding 1,600 m, increasing rock temperatures, humidity, and airflow resistance introduce challenges to both worker comfort and operational efficiency. In large undergrou
...
As Luossavaara-Kiirunavaara Aktiebolag advances the development of its underground iron ore mines to depths exceeding 1,600 m, increasing rock temperatures, humidity, and airflow resistance introduce challenges to both worker comfort and operational efficiency. In large underground mines, even small increases in operational efficiency can yield high energy savings. This applies to both heat stress management using cooling and ventilation methods, and air flow quantity reduction according to the planned production tonnage in production levels. This study evaluates potential underground heat stress conditions and develops a predictive framework for temperature forecasting in the Kiruna and Malmberget mines.
The heat stress assessment applied Wet Bulb Globe Temperature (WBGT) measurements using both natural wet-bulb and globe temperature sensors, to quantify thermal comfort under varying metabolic rates. Receiver Operating Characteristic (ROC) analysis revealed that the WBGT index exhibited the strongest discrimination between acceptable and unacceptable temperatures under low metabolic rates, while high metabolic rate conditions achieved optimal balance, detecting well acceptable conditions and avoiding marking unacceptable conditions as acceptable. The operational WBGT thresholds were rounded to zero decimal places.
The temperature forecasting component employed regression modeling using surface conditions and production plans, across multiple mine regions, validated through holdout, k-fold, and leave-one-out cross-validation methods. Model performance varied regionally, with stability closely linked to dataset size and composition. Larger combined datasets produced more reliable predictions, while smaller or imbalanced datasets led to higher variance and lower R² values.
Overall, the results emphasize the necessity of mine-specific thermal comfort thresholds and rigorous model validation procedures. By combining physiological heat stress evaluation with predictive temperature modeling, this research provides Luossavaara-Kiirunavaara Aktiebolag with a practical approach for safer, energy-efficient ventilation management in deep mining operations.