
Full text loading...
The Coal Measures exhibit a strongly cyclic sequence of bedding, with thin coal seams interleaved with thicker layers of country rock, generally shales or sandstones. The coal itself is an unusual rock type as it has both a low seismic velocity and a low density in comparison with the surrounding rocks. These seismic properties generate a high impedance contrast across any coal-country rock interface and so the reflection coefficient at normal incidence lies between 0.3 and 0.5 with typical values around 0.35. For such a cyclic system with rapidly alternating impedance contrasts, O'Doherty and Anstey (1971) have shown that as the observation point is taken deeper into the layering, the composite transmission effect of the 'peg-leg' multiples in the thin beds can quite quickly become larger than the directly transmitted arrival. Such a mechanism allows a large part of the incident energy to be transmitted through the cyclic stack but with slight delay due to the multiple paths in the thin coal seams. The apparent attenuation of seismic waves in such cyclic structures is frequency dependent (Spencer, Edwards and Sonnad 1977, Schoenberger and Levin 1974,1978). The attenuation will reach a maximum at a frequency which is inversely proportional to the thickness of the layers. Of recent years, there has been considerable interest in the use of seismic methods for coal exploration particularly in the development of mine sites. As a result a number of studies have been made of the seismic response of the coal measures. Dresen and Ullrich (1976) have made scaled down simulations of the coal measures for model seismic experiments. Van Riel (1965) and Rüter and Schepers (1978) have constructed synthetic seismograms for normal incidence plane waves for impedance models of coal sequences. These calculations were performed in the time domain and so the thinnest layers which could be resolved were between 0.75 and 1.00 m. Even for a small number of coal seams, interference effects give very complex reflection seismograms. The restrietion of the minimum thickness of coal seams can be removed by working in the frequency domain. This is the approach adopted by Fertig and MüIler (1978, 1979), who have calculated seismograms at large offsets from a point source with particular interest in conversions from P and S waves. This work allowed the construction of the full response of the Coal Measures including geometrical spreading but the algorithm used by Fertig and Müller did not allow them to analyse the way in which internal multiples contributed to the theoretical seismograms. In this paper, we also make use of theoretical seismograms for offset receivers calculated in the frequency domain, but are able to use the reflection matrix methods discussed by Kennett (1979) to isolate the contribution of the multiples to the final seismograms. The major effect arises from multiples within the coal seams and then the actual disposition of the seams gives rise to complex interference phenomena.