Predicting the height of the water-conducting fractured zone using multiple regression analysis and GIS

The water-conducting fractured zone induced by mining under aquifers is channels for water and sand inrushes. A method based on multiple regression analysis and a geographic information system (GIS) is proposed to predict the height of the water-conducting fractured zone in this paper. Five main indicators are found to control the height of the water-conducting fractured zone during fully mechanized caving mining under aquifers in which the thickness of the coal seam is more than 3 m: thickness of the coal seam, proportion of hard rock, length of the panel, mined depth and dip angle. The height of water-conducting fractured zone and the predictive variables from eighteen coal mines in China are investigated. Based on information entropy theory, a nonlinear multiple regression model and the weight of the indicators are considered, and the nonlinear multiple regression model is used in the GIS. Then, the approach is validated with a case study of the Xiegou Coalmine in the Shanxi province of China, in which the thickness of the coal seam which is under aquifers is more than 3 m. Error analysis of the approach is calculated with different hydrogeology classification. The results indicate that this model is a useful tool for predicting and analyzing the height of the water-conducting fractured zone, since the height of the water-conducting fractured zone can be quantitatively calculated and visualized.