Incorporating Uncertainty in Coal Seam Depth Determination via Seismic Reflection and Geostatistics
Modelling mineral deposits requires the use of all possible source of information. Traditionally, core samples from borehole are the most used way to access the ore body, however this method is expensive and provides information restricted to a close neighbourhood within the sample location. Continuity between sampled points needs to be inferred in order to infill values among bore hole locations using interpolation techniques. In contrast, geophysical methods including seismic reflection provide data at much closer intervals, thus approximating continuous sampling along a seismic section. These data are then used to infer spatial continuity, for example the fault of a coal seam in between bore holes. Wave travel time along the seams is recorded by seismic survey at a dense grid. Additionally, sonic wave velocity logged along boreholes can be interpolated at a dense grid. Sonic Logging provides direct and continuous measurements of the sonic wave velocity at all seams down the holes logged. Therefore, this logged sonic velocity can be simulated within a dense grid compatible to the time grid. Multiple velocity grids (equally probable models) are generated within the simulation framework. In combining both grids, i.e. velocity and time, seam depth can be obtained. Consequently risk in depth determination for each seam due to velocity uncertainty can be assessed. Both data types (time and sonic) are subject to various sources of error. Currently, velocity is indirectly determined using processed seismic data, which may breed errors in geologic sections interpretation. The present paper will show results from a Sonic Logging velocity simulation and its uncertainty determination, in order to use the results in calculating seam depth via seismic reflection and additionally provided a measure for error in this parameter. A case study in a major coal deposit illustrates the procedure.
Unable to display preview. Download preview PDF.