To verify an interpretation of a cross-borehole seismic survey, two- and three- dimensional finite element models were constructed which incorporated elastoplastic material properties with strain softening for simulating nonlinear rock deformation, and interface elements for simulating slip along existing planes of weakness. The seismic survey was conducted to determine the location of an abandoned underground coal mine which is flooded.
Assuming that P-wave velocity in rocks is influenced primarily by the magnitude of stress in the direction of the P-wave propagation, good agreement was found between the velocity anomalies revealed by a difference tomogram from the crosshole seismic survey and evaluation of horizontal stress relaxation and concentration by the finite element model. Excellent agreement was also found between the difference tomogram and the zones of horizontal tensile and compressive strains predicted by the finite element model. On the other hand, poor agreement was found between the difference tomogram and the predicted fractures around the mined-out area suggesting that the water-saturated fracture zones were not the primary factor controlling the P-wave propagation.
This analysis represents a unique large scale demonstration of the stress response around a mine void in a sedimentary environment. The agreement between the seismic data and the model results indicates that carefully executed tomographic surveys can be effective in detecting old mines, and that the velocity anomalies recorded are most representative of the in-situ stress state and horizontal strain distribution around the mine void. The ability of a model based on very few assumptions to mimic the seismic observations demonstrates both the validity of the model and its capability as a tool to accurately predict the effects of mining.
Incorrectly mapped abandoned underground mines pose an extreme safety hazard to any active mines due to the possibility of a sudden inflow of water and methane. For this reason, it is necessary to accurately determine the limits of prior mining activity so that an adequate safety barrier may be maintained between the old and new mines. The most common and reliable method for making this determination is drilling a series of closely-spaced holes from the surface to verify the presence or absence of the old mine in a specific region of interest.
In the case described in this paper, a series of main entries for a new underground mine had been projected for development from west to east in the Illinois No. 6 coal seam approximately m (705 feet) beneath the margin of a shallow lake, as shown in Figure 1 (Cotten and Geldmacher, 1990). These entries would be used to move material and equipment into the eastern section of the mine and remove the mined coal for the rest of the mine life, approximately 30 years. Immediately to the north of the projected mine lies an abandoned mine from which the last coal was removed about 40 years ago. Due to the age of this mine, the accuracy of the available mine survey was questionable, and the exact limits of the old mine were unknown. Therefore, determination of the actual position of the old works well in advance of mining became necessary so that a barrier pillar of sufficient width could be maintained.
If this situation had been confronted beneath a land area, the direct means of addressing it would have been to drill a series of closely svaced vertical holes from the surface on avvroximate 6 m (20 ft) centers to determine the absence or presence of the old mine along a line located a specified distance north of the new entry projections. If a mine void was encountered, then the drilling program
