Empirical and Analytic Approaches to Subsidence Prediction

Empirical methods for describing the shape of the subsidence trough over coal mines in Europe are tested against field measurements of subsidence over longwall panels in the United States. The graphical methods developed by the National Coal Board in the United Kingdom do not correlate well with the U.S. measurements; however, the profile functions typically used on the Continent give quite acceptable fits to the data. In the case of two longwall panels mined at Old Ben No. 24 in Illinois, a trigonometric profile function appears the best choice; in the case of two longwall panels mined at York Canyon in New Mexico, an error integral function is somewhat superior. Analyses of subsidence development curves show that rate effects occur. A comparison of fitting parameters establish influence functions as an important methodology for future analysis. In conjunction with the evaluation of empirical methods, an analytic model based on material failure and bulking was developed. A key element of the model is the void-volume distribution function which establishes a mechanism for transfer of significant amounts of the mined volume to the surface. Preliminary calculations demonstrate the features of the model and illustrate close correlation to physical observations in the field. Moreover, centrifuge models simulating the subsidence process have been used successfully to demonstrate the dominant or key failure modes necessary to the analytic model. General features of the centrifuge simulations and model calculations correlate well.