From a mining perspective, a roof stability prediction model is merely the comparison of stresses that are present in the longwall panel with the geological data of the roof directly above the area of interest. The primary focus of this thesis is the examination of data obtained for the Moranbah North coal mine, recorded during the mining process in 2004. There has been limited work in this field in the past with most mines considering their roof using rating methods such as the CMRR (coal mine roof rating) rather than trying to predict areas of particular instability.
Basic longwall mining concepts have been reviewed in order to provide a basis for the model. Of particular note have been the methods of assessing roof stability and the methods used to control areas of trouble. Rockbolt types and rockbolting methods, which have been found to be the most commonly used method in avoiding roof failure, have been examined in some detail.
The data recording method used for this thesis has been the internal stress recorders on the longwall chocks. The Moranbah North coal mine uses 150 chocks across its longwall panel with each chock taking a stress reading every minute. These readings have been reviewed and checked for any areas where there may be inaccurate recording, and subsequently compacted into a much more “user-friendly”. This data can then be exported into a graphing program, such as Surfer to create a contour map showing areas of higher stress. This allows for visual recognition of areas of possible instability based purely on stress.
The final steps required involve a comparison of the compacted data with geological information for the site. In the event that a correlation is found between geology and stress regions, then unmined areas that are likely to cause trouble will be identified on the mine lease.