Rapid assessment of fire parameters in coal mines using machine learning

Fires are among the most frequent accidents in coal mines in Russia. Valid reaction choice during a fire requires information not only about the fact itself, but also about the location of the seat of fire, its current intensity and anticipated change of the fire hazards. In this respect, this study explores applicability of neural network models for the joint assessment of the distance to the seat of fire, current heat liberation intensity and short-term prediction of temperature and carbon oxide concentration at the check points in a roadway. The input data of a neural network model were the time windows of previous temperatures and CO concentrations at three tandem check points, and ventilation airflow velocities. The forecast horizon for the temperature and CO concentration was 60 s. In the test sampling, the absolute error was 99.34 kW for the heat liberation capacity, 4.11 m for the distance to the seat of fire, 2.97 °C for the predicted temperature and 12.85 ppm for the predicted CO concentration. The results show that within the assumed range of scenarios in this study, the neural network model with the input data in the form of the time windows of previous temperatures and CO concentrations is usable for the joint assessment of fire parameters and short-term prediction of the test fire hazards at the check points in coal mines. 

Fedotkin I. O. Rapid assessment of fire parameters in coal mines using machine learning. MIAB. Mining Inf. Anal. Bull. 2026;(8):157-166. [In Russ]. DOI: 10.25018/0236_1493_2026_8_0_157.

I.O. Fedotkin, Graduate Student, University of Science and Technology MISIS, 119049, Moscow, Russia, e-mail: fedotkin.iliya@gmail.com, ORCID ID: 0009-0004-2399-480X.

1. Liang Y., Yang Y., Guo S., Tian F., Wang S. Combustion mechanism and control approaches of underground coal fires: a review. International Journal of Coal Science & Technology. 2023, vol. 10, article 24. DOI: 10.1007/s40789-023-00581-w.

2. Balovtsev S. V., Skopintseva O. V., Kulikova E. Yu. Analysis of accidents and development trends in aerological safety of coal mines. MIAB. Mining Inf. Anal. Bull. 2024, no. 12, pp. 135—149. [In Russ]. DOI: 10.25018/0236_1493_2024_12_0_135.

3. Arhipov I. A., Filin A. E. Accident rate analysis in coal mines in Russia. MIAB. Mining Inf. Anal. Bull. 2019, no. 1, pp. 208—215. [In Russ]. DOI: 10.25018/0236-1493-2019-01-0-208-215.

4. Niu Y., Si R., Li Z., Wang L., Huang Z., Jia Q. Experimental study on gas and coal dust explosive overpressure and flame dynamic characteristics in an engineering-level test roadway. Frontiers in Earth Science. 2023, vol. 11, article 1330932. DOI: 10.3389/feart.2023.1330932.

5. Skopintseva O. V., Balovtsev S. V., Rybichev A. A. Investigation of the explosion and fire hazard of coal dust in formations of the middle stage of metamorphism. MIAB. Mining Inf. Anal. Bull. 2026, no. 5, pp. 80—94. [In Russ]. [In Russ]. DOI: 10.25018/0236_1493_2026_5_0_80.

6. Zhikharev S. Ya., Rodionov V. A., Pikhkonen L. V. Innovative methods for investigating technological properties and explosion/fire risk data of coal dust. Gornyi Zhurnal. 2018, no. 6, pp. 45—49. [In Russ]. DOI: 10.17580/gzh.2018.06.09.

7. Babrauskas V., Peacock R. D. Heat release rate: The single most important variable in fire hazard. Fire Safety Journal. 1992, vol. 18, no. 3, pp. 255—272. DOI: 10.1016/0379-7112(92)90019-9.

8. Salami O. B., Kumar A. R., Aamir I., Pushparaj R. I., Xu G. Enhancing fire safety in underground mines: Experimental and large eddy simulation of temperature attenuation, gas evolution, and bifurcation influence for improved emergency response. Process Safety and Environmental Protection. 2024, vol. 183, pp. 260—273. DOI: 10.1016/j.psep.2023.12.056.

9. Zhang M., Li Z. Experiments on a mine system subjected to ascensional airflow fire and countermeasures for mine fire control. Fire. 2024, vol. 7, no. 7, article 223. DOI: 10.3390/fire7070223.

10. Yuan L., Zhou L., Smith A. C. Modeling carbon monoxide spread in underground mine fires. Applied Thermal Engineering. 2016, vol. 100, pp. 1319—1326. DOI: 10.1016/j.applthermaleng.2016.03.007.

11. Li B., Li Y., Sun Y., Zhang W., Li J., Zhang Z., Cui Y., Dong J., Liu H. Study on the influence of forced ventilation on the maximum fire temperature in roadway heading. Scientific Reports. 2025, vol. 15, article 9830. DOI: 10.1038/s41598-025-94169-w.

12. Kopylov N. P., Fedotkin D. V., Karpov A. V., Sushkina E. Yu. Modeling of extinguishing of oil product fires in the tanks using water-based extinguishing agents. Occupational Safety in Industry. 2020, no. 8, pp. 14—22. [In Russ]. DOI: 10.24000/0409-2961-2020-8-14-22.

13. Nematollahi Sarvestani A., Oreste P., Gennaro S. Fire scenarios inside a room-and-pillar underground quarry using numerical modeling to define emergency plans. Applied Sciences. 2023, vol. 13, no. 7, article 4607. DOI: 10.3390/app13074607.

14. Weisenpacher P., Glasa J., Valasek L. Investigation of various fire dynamics simulator approaches to modelling airflow in road tunnel induced by longitudinal ventilation. Fire. 2025, vol. 8, no. 2, article 74. DOI: 10.3390/fire8020074.

15. Fernández-Alaiz F., Castañón A.M., Gómez-Fernández F., Bascompta M. Mine fire behavior under different ventilation conditions: Real-scale tests and CFD modeling. Applied Sciences. 2020, vol. 10, no. 10, article 3380. DOI: 10.3390/app10103380.

16. Haghighat A., Luxbacher K. Tenability analysis for improvement of firefighters’ performance in a methane fire event at a coal mine working face. Journal of Fire Sciences. 2018, vol. 36, no. 3, pp. 256—274. DOI: 10.1177/0734904118767066.

17. Ang C. D., Rein G., Peiró J., Harrison R. Simulating longitudinal ventilation flows in long tunnels: Comparison of full CFD and multi-scale modelling approaches in FDS6. Tunnelling and Underground Space Technology. 2016, vol. 52, pp. 119—126. DOI: 10.1016/j.tust.2015.11.003.

18. Yuan L., Mainiero R. J., Rowland J. H., Thomas R. A., Smith A. C. Numerical and experimental study on flame spread over conveyor belts in a large-scale tunnel. Journal of Loss Prevention in the Process Industries. 2014, vol. 30, pp. 55—62. DOI: 10.1016/j.jlp.2014.05.001.

19. Tan P., Zhang C., Xia J., Fang Q.-Y., Chen G. Estimation of higher heating value of coal based on proximate analysis using support vector regression. Fuel Processing Technology. 2015, vol. 138, pp. 298—304. DOI: 10.1016/j.fuproc.2015.06.013.

20. Mowrer F. W., Williamson R. B. Methods to characterize heat release rate data. Fire Safety Journal. 1990, vol. 16, no. 5, pp. 367—387. DOI: 10.1016/0379-7112(90)90009-4.

21. Ingason H. Design fire curves for tunnels. Fire Safety Journal. 2009, vol. 44, no. 2, pp. 259—265. DOI: 10.1016/j.firesaf.2008.06.009.

22. Dubey S. R., Singh S. K., Chaudhuri B. B. Activation functions in deep learning: A comprehensive survey and benchmark. Neurocomputing. 2022, vol. 503, pp. 92—108. DOI: 10.1016/j.neucom.2022.06.111.

23. Kapoor S., Narayanan A. Leakage and the reproducibility crisis in machine-learning-based science. Patterns. 2023, vol. 4, no. 9, article 100804. DOI: 10.1016/j.patter.2023.100804.