Detailed Geological Characterisation from Seismic Data

The use of seismic reflection surveying continues to grow within Australia's underground coal mining regions of the Sydney and Bowen Basins. 3D seismic surveys aimed at providing detailed information on the structures with the potential to affect longwall operations are now being routinely undertaken. For this project, we have investigated the potential for acoustic impedance inversion to complement the information available from conventional seismic surveys. It was our expectation that acoustic impedances derived from seismic reflection data could be related to geotechnical rock properties.

Acoustic impedance is defined by the product of seismic P-wave velocity and rock density. The contrast in values between different layers gives rise to seismic reflections. Acoustic impedance inversion is the name given to the process whereby the traces observed in seismic sections are converted to the impedance sections that would give rise to the observed seismic data. As a precursor to acoustic impedance inversion, the normal seismic processing procedures are replaced by true amplitude processing designed to preserve the relative amplitudes of the reflections from the different rock layers present.

The methods of seismic inversion have been developed mainly for the investigation of petroleum reservoirs. Commercial software packages are available and for this project we utilised the Hampson and Russell software available at Curtin University of Technology. For the true amplitude processing of the seismic data, the Promax software operated at Velseis Processing was used.

Inversions were undertaken for three 3D seismic surveys and two 2D surveys. The sites were at Grasstree and North Goonyella Mines in the Bowen Basin and at West Cliff and Dendrobium Collieries in the Sydney Basin. To assist with the understanding of the impedance results, we derived an empirical relationship between acoustic impedance and the newly developed Geophysical Strata Rating (GSR). This allows impedance values to be converted into GSR values that have more meaning in geotechnical assessment.

To obtain satisfactory inversions, we used the model based approach. This requires an initial model of acoustic impedances to be derived from acoustic impedance information (density and sonic data) in exploration boreholes. At coal mines, there are usually a sufficient number of exploration boreholes to allow determination of a reasonable starting model, especially for 3D seismic surveys. For 2D seismic surveys, the boreholes need to be on the seismic lines and development of initial models may be more difficult to achieve.

An advantage of model based inversion is that it integrates seismic and borehole information. However, there are no unique solutions and it is necessary to explore a range of possible solutions before deciding on the most likely result. The ambiguity depends on the amount and quality of the available borehole data.