Subsurface Fracture Development due to Longwall Mining and its Interpretation using Image Processing Techniques

This paper presents the physical modelling results of subsurface fracture development associated with Iongwall mining operations and an application of image processing techniques to interpretion of the results. Physical modelling results haióbeen obtained by employing a large sand and plaster model loaded purely by gravity as a means of fracturing the subsurface strata above the longwall excavation. The results have shown fracture development and crack propagation as a longwall face advances and demonstrated that the strength of stratified rocks, the presence and position of a weak band-, and the extraction thickness, have significant effects upon the overall fracture patterns above the excavation. Binary images of fracture patterns were scanned from the black and white photographs of the tested models using an image analyser coupled to a video camels. These images were manipulated by a custom written C program which allows the images to be processed in both vertical and horizontal direction. The images were filtered to quantitatively differentiate between vertical and horizontal fracture intensities. The vertical fracture intensity is particularly important as vertical fractures are considered to provide the main avenues for surface and subsurface water inflow into the excavation. The use of the image analysis techniques has shown a promising improvement in image enhancement and in model results interpretation in a quantitative manner. The numerical nature of the image analysis results will allow further statistical comparison with other modelling techniques to be undertaken and significantly increase the scope for use of the modelling technique results.