JOURNAL OF ROCK MECHANICS

JOURNAL OF ROCK MECHANICS

Prediction of maximum ground surface Subsidence due to longwall mining: a comparative evaluation of empirical equations and statistical and artificial intelligence models

Document Type : Original Article

Authors
1 PhD Candidate, School of Mining Engineering, College of Engineering, University of Tehran, Tehran, Iran
2 2Professor, School of Mining Engineering, College of Engineering, University of Tehran, Tehran, Iran,
Abstract
Ground surface subsidence caused by longwall coal mining is inevitable, but its prediction and control can help reduce damage to surface and subsurface structures. This study aims to (1) identify the most suitable empirical equation for estimating maximum surface subsidence (Smax) based on key geometric parameters of coal seam thickness (hs), overburden depth (H), and panel width (Lw); and (2) develop and compare two data-driven models: Random Forest (RF) and Linear Multivariate Regression (LMR). To do so, a collected database consist of 85 real-world cases was analyzed, and Smax was computed using each empirical method. Four statistical indicators (R², VAF, MAPE, RMSE) were used to evaluate and rank the empirical models. Then, the database was divided into two sets of train and test. The RF model, trained and tested on randomly split data, achieved the highest predictive accuracy, while the LMR model showed acceptable results. Specifically, the RF model yielded R² = 0.921, VAF = 92.71%, MAPE = 23.34%, and RMSE = 0.287 in the training phase; and R² = 0.939, VAF = 94.11%, MAPE = 13.21%, and RMSE = 0.246 in the testing phase. Finaly, sensitivity analysis revealed that hs has the greatest influence on Smax, whereas H plays the least significant role. Overall, RF proved to be a promising alternative to empirical methods in practical subsidence prediction.
Keywords
Subjects

[1]    Majdi, A., Hassani, F. P., & Nasiri, M. Y. (2012). Prediction of the height of destressed zone above the mined panel roof in longwall coal mining. International Journal of Coal Geology, 98, 62-72.
[2]    Bazaluk, O., Kuchyn, O., Saik, P., Soltabayeva, S., Brui, H., Lozynskyi, V., & Cherniaiev, O. (2023). Impact of ground surface subsidence caused by underground coal mining on natural gas pipeline. Scientific Reports, 13(1), 19327.
[3]    Zhu, H., Wang, H., Gao, R., & Zhao, Y. (2023). Surface Subsidence Characteristics of Mining Panel Layout Configuration with Multi-Seam Longwall Mining. Processes, 11(6), 1590.
[4]    Wang, Z. H., Wu, S. X., Li, J. L., Sun, W. C., Wang, Z. F., & Liu, P. J. (2023). Surface subsidence and its reclamation of a coal mine locating at the high groundwater table, China. International Journal of Environmental Science and Technology, 1-20.
[5]    National Coal Board. (1975). Subsidence Engineering Handbook. National Coal Board Mining Department.
[6]    Karmis, M., Jarosz, A., Schilizzi, P., & Agioutantis, Z. (1987). Surface deformation characteristics above undermined areas: Experience from the Eastern United States. Civil Engineering Transactions, 29(2), 106-114.
[7]    Van der Merwe, J. N. (1991). Subsidence caused by high extraction coal mining in the Sasolburg and Secunda areas: prediction thereof and the mitigation of its effects (Doctoral dissertation, PhD thesis, University of the Witwatersrand).
[8]    Peng, S. S., Luo, Y., & Zhang, Z. M. (1995). Subsidence parameters-their definitions and determination. PREPRINTS-SOCIETY OF MINING ENGINEERS OF AIME.
[9]    Luo, Y., & Peng, S. S. (2000). Prediction of subsurface subsidence for longwall mining operations. In Proceedings of the 19th international conference on ground control in mining (pp. 163-170). West Virginia University, WV.
[10]  Yavuz, H. (2004). An estimation method for cover pressure re-establishment distance and pressure distribution in the goaf of longwall coal mines. International Journal of Rock Mechanics and Mining Sciences, 41(2), 193-205.
[11]  Majdi, A., Ghomghaleh, A., & Arabzadeh, M. H. (2007). Theoretical prediction of ground settlement induced by longwall mining. In ISRM Congress (pp. ISRM-11CONGRESS). ISRM.
[12]  Cheng, J., Liu, F., & Li, S. (2017). Model for the prediction of subsurface strata movement due to underground mining. Journal of Geophysics and Engineering, 14(6), 1608-1623.
[13]  Yan, Y., Li, M., Liu, J., Yan, W., Zhang, J., & Zhou, B. (2021). Ground subsidence evolution from 1000 m deep mining: a case study in Fengfeng mining area. Shock and Vibration, 2021, 1-9.
[14] Cui, X., Zhao, Y., Wang, G., Zhang, B., & Li, C. (2020). Calculation of residual surface subsidence above abandoned longwall coal mining. Sustainability, 12(4), 1528.
[15] Guo, W., Zhao, G., Bai, E., Guo, M., & Wang, Y. (2021). Effect of overburden bending deformation and alluvium mechanical parameters on surface subsidence due to longwall mining. Bulletin of Engineering Geology and the Environment, 80, 2751-2764.
[16] Prakash, A., Kumar, A., Verma, A., Mandal, S. K., & Singh, P. K. (2021). Trait of subsidence under high rate of coal extraction by longwall mining: Some inferences. Sādhanā, 46, 1-8.
[17] Wang, K., Li, J., & Jin, Z. (2022). Influence of the primary key stratum on surface subsidence during longwall mining. Sustainability, 14(22), 15027.
[18] Khanal, M., Qu, Q., Zhu, Y., Xie, J., Zhu, W., Hou, T., & Song, S. (2022). Characterization of overburden deformation and subsidence behavior in a kilometer deep longwall mine. Minerals, 12(5), 543.
[19] Taherdito, A. H., Sulistianto, B., Karian, T., & Widodo, N. P. (2023). Surface Subsidence Prediction Using Numerical Method, Cased Study of Longwall Mining In Indonesia.
[20] Diddle, B., Agioutantis, Z., Maldonado Esguerra, E., Romero Benitez, J. D., & Parra Valencia, M. (2024). Prediction of Dynamic and Final Vertical and Horizontal Movements Due to Longwall Mining. Rock Mechanics and Rock Engineering, 1-18.
[21] Maldonado, E., Benitez, J. R., & Agioutantis, Z. (2024, June). Sensitivity analysis of a dynamic subsidence prediction model for longwall extraction. In ARMA US Rock Mechanics/Geomechanics Symposium (p. D042S056R001). ARMA.
[22] Liu, X., Zhang, Y., Zhang, J., Yang, T., Jia, P., & Guo, R. (2024). Modelling surface subsidence of coal mines using a bonded block numerical method. Geomatics, Natural Hazards and Risk, 15(1), 2336017.
[23] Cheng, G., Liu, H., Li, F., Nie, T., Wang, Q., & Peng, C. (2025). Study on subsidence evolution induced by coal mining under highway based on finite element simulation. Energy Exploration & Exploitation, 01445987241312703.
[24] Wu, S., Wei, W., Li, L., Cheng, B., Zhang, S., & Xie, H. (2025). Effect of alluvial thickness on the surface subsidence characteristics at deeply buried mining area with thin bedrock: A case study. Journal of Mountain Science, 1-17.
[25]  Sutherland, H. J., & Munson, D. E. (1984). Prediction of subsidence using complementary influence functions. In International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts (Vol. 21, No. 4, pp. 195-202). Pergamon.
[26]  Sheorey, P. R., Loui, J. P., Singh, K. B., & Singh, S. K. (2000). Ground subsidence observations and a modified influence function method for complete subsidence prediction. International Journal of Rock Mechanics and Mining Sciences, 37(5), 801-818.
[27]  Hood, M., Ewy, R. T., & Riddle, L. R. (1983). Empirical methods of subsidence prediction—a case study from Illinois. In International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts (Vol. 20, No. 4, pp. 153-170). Pergamon.
[28]  Li, W. X., Liu, L., & Dai, L. F. (2010). Fuzzy probability measures (FPM) based non-symmetric membership function: Engineering examples of ground subsidence due to underground mining. Engineering Applications of Artificial Intelligence, 23(3), 420-431.
[29]  Ma, H., Ji, Y., & Wang, S. (2022). Calculation of Subsidence due to Longwall Mining with Overburden Grouting and Its Application. Geofluids, 2022.
[30]  Pan, R., Li, Y., Wang, H., Chen, J., Xu, Y., Yang, H., & Cao, S. (2021). A new model for the identification of subcritical surface subsidence in deep pillarless mining. Engineering Failure Analysis, 129, 105631.
[31]  Rasouli, H., Shahriar, K., & Madani, S. H. (2021). Mine subsidence prediction using gene expression programming based on multivariable symbolic regression. ITEGAM-JETIA, 7(29), 13-24.
[32]  Xu, J., Zhu, W., Xu, J., Wu, J., & Li, Y. (2021). High-intensity longwall mining-induced ground subsidence in Shendong coalfield, China. International Journal of Rock Mechanics and Mining Sciences, 141, 104730.
[33]  Yin, H., Guo, G., Li, H., Wang, T., & Yuan, Y. (2022). Prediction method and research on characteristics of surface subsidence due to mining deeply buried Jurassic coal seams. Bulletin of Engineering Geology and the Environment, 81(10), 449.
[34]  Zhang, B., Yi, J., Zhang, Z., Xu, L., & Xu, N. (2019). A Comprehensive Method for Subsidence Prediction on Two-Seam Longwall Mining. Energies, 12, 3139.
[35]  Gutierrez, J. J. (2010). Estimating Highway Subsidence due to Longwall Mining (Doctoral dissertation, University of Pittsburgh).
[36]  Munson, D. E., & Eichfeld, W. F. (1980). European empirical methods applied to subsidence in US coal fields (No. SAND-80-1920). Sandia National Labs., Albuquerque, NM (USA).
[37]  He, C., & Xu, J., (2018). Subsidence prediction of overburden strata and surface based on the voussoir beam structure theory. Advances in Civil Engineering, 2018.
[38]  Karmis, M., Triplett, T., Schillizzi, P., & Hasenfus, G. (1984). Ground movement characteristics above mined panels in Appalachia-An empirical approach.
[39]  Rezaei, M., Majdi, A., & Monjezi, M. (2014). An intelligent approach to predict unconfined compressive strength of rock surrounding access tunnels in longwall coal mining. Neural Computing and Applications, 24, 233-241.
[40]  Majdi, A., & Rezaei, M. (2013). Prediction of unconfined compressive strength of rock surrounding a roadway using artificial neural network. Neural Computing and Applications, 23, 381-389.
[41] Yilmaz, I., & Kaynar, O. (2011). Multiple regression, ANN (RBF, MLP) and ANFIS models for prediction of swell potential of clayey soils. Expert System with Application, 38, 5958–5966.
[42] Breiman, L. (2001). Random forests. Machine learning, 45, pp.5-32.
[43] Zhou, J., Shi, X., Du, K., Qiu, X., Li, X., & Mitri, H.S. (2017). Feasibility of random-forest approach for prediction of ground settlements induced by the construction of a shield-driven tunnel. International Journal of Geomechanics, 17(6), p.04016129.
[44] Hastie, T., Tibshirani, R., & Friedman, J. (2017). The elements of statistical learning: data mining, inference, and prediction.
[45] Tang, L. & Na, S., (2021). Comparison of machine learning methods for ground settlement prediction with different tunneling datasets. Journal of Rock Mechanics and Geotechnical Engineering, 13(6), pp.1274-1289.
[46] Asteris, P. G., & Armaghani, D. J. (2025). An empirical-driven machine learning (EDML) approach to predict PPV caused by quarry blasting. Bulletin of Engineering Geology and the Environment, 84(4), 1-17.
[47]  Yang, Y., & Zhang, Q. (1997). A hierarchical analysis for rock engineering using artificial neural networks. Rock mechanics and rock engineering, 30, 207-222.
[48] Matinpour, F., & Majdi, A. (2025). Determining the abutment angle in longwall coal mining using NLMR, GEP and GEO techniques. Earth Science Informatics, 18(2), 358.