An Enhanced Empirical Model for Predicting Continuous Fracturing in Rock Masses

Abstract A reliable empirical prediction of the height of fracturing (HoF) is significant for mining impact assessment and mine sustainability. However, the prediction accuracy is often limited by the parameter and data coverage involved in empirical models. This paper develops a new predictive model that incorporates mining parameters, as well as the proportion and strength of stiff strata, as predictors for enhanced mining impact assessments. A HoF database has been established by sourcing 80 cases from 50 mines in Australia and China. Statistical interpretation identifies positive correlations of HoF to mining height, panel width, and cover depth. Numerical modelling and sensitivity analysis identify that both the proportion and strength of stiff units negatively correlate to HoF and are more significant than bedding plane frequency. The two critical rock parameters are statistically independent and should be considered for enhanced HoF predictions. A predictive model is established by nonlinear regression over the 80 datasets, with both rock parameters involved by mapping to a dimensionless indicator of strata competency considering dimensional consistency. The model is tested with an $${R}^{2}$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:msup> <mml:mrow> <mml:mi>R</mml:mi> </mml:mrow> <mml:mn>2</mml:mn> </mml:msup> </mml:math> of 0.84 and more accurate predictions than several historical models regarding 24 new cases. Prediction bounds with a confidence level of 95% are developed considering the rock parameter variability, providing conservative HoF predictions for cases where significant natural features should be preserved for sustainable mining practices. This paper assesses the influence of rock parameters that have not been considered in historical models on continuous fracturing and provides an enhanced empirical method for mining impact assessment.