Longwall Roof Fall Prediction and Shield Support Recommendations

In the 1990s the German mining industry introduced a new generation of shield supports. The new design of support has a maximum load capacity of 10,000 kN, making these units as strong as the shields used in Australia and in the USA. DMT took more than 3,100 underground observations in order to verify the roof fall frequency by statistical analysis. The results of this work have led to practical recommendations for roof control and the required shield support system on longwall faces. The underground observations have been correlated to Rock Mass Classification, to stress calculations and to the angle between the direction of the fissures and the direction of longwall mining_ The analysis work yielded the following two sets of results: 1. There is a critical distance between the canopy tip and the coal face (Tip to Face TF,n,). The TF,n, is predictable and relates to: the thickness of the first roof layer and its uniaxial compressive strength. The face support should have a TF that is less than the TFcrtt in every underground longwall situation. Exceeding the TF crit call immediately result in a roof fall. 2. Using the obtained regression equation DMT is able to calculate the probability of the roof fall frequency (FF), which describes the roof fall sensitivity. When TFcr;, is exceeded the predicted roof fall FF relates to: the measured support resistance (MSR) of the shield support the calculated vertical stress (p?) the fissure-direction index (DI = angle between main fissure direction and direction of mining) and the distance by which TFcrit is exceeded (ATF). Armed with these results DMT is now able to predict the critical distance between the tip of the canopy and the coal face (TFcrit), as well as the roof fall frequency, for all shield designs. By applying the new calculation method it is now possible to compare alternative longwall layouts and different shield support types under pre-set geological conditions. Mining engineers on site are therefore in a position to make the necessary roof control preparations required to run the longwall operation to maximum efficiency. The results provide a useful basis for making practical recommendations and for selecting the most effective design of shield support. The paper uses practical examples to demonstrate the R&D results and present the various methods now available for calculation and prediction in longwall roof control.