Underground » Geology
Faults cause breaks in continuity of seismic horizons and at these discontinuities, diffraction patterns in seismic waves are generated. This is a well understood phenomenon but modern seismic acquisition and processing techniques are mainly directed towards providing high resolution images of the reflecting layers in the subsurface. Diffraction behaviour is largely ignored.
The potential for a diffraction analysis to reveal the locations of small faults and dykes that can impact on underground coal mine operations is therefore a fertile area for investigation. In this project, the objective was to undertake this investigation and establish whether diffraction analysis would allow identification of the location of small faults within 2D and 3D survey data. The study has involved computer generated synthetic seismic data and actual survey data from high resolution seismic surveys from Australian coal mines. It was anticipated that new processing techniques might need to be developed.
Extensive computer modelling of different geological scenarios encountered in coal mining allowed investigation of diffraction behaviour in both raw shot records and processed sections. Pre‐stack depth migration was shown to collapse diffractions and make the detection of small faults difficult. However, in conventional post stacked sections, diffractions are preserved and can be extracted by utilising relatively straightforward procedures aimed at removing reflection events from coal seams with relatively shallow dips. By comparison with stacked sections produced after the reflection data were removed from the shot records, it was found that normal moveout corrections and stacking do not adversely affect the diffraction behaviour. Diffractions can therefore be studied and extracted from conventional stacked seismic sections without the need for significant reprocessing.
Encouraged by these findings, 2D and 3D seismic survey data from the Sydney and Bowen Basins were studied. As far as possible, these data were selected from sites where mining or other geological information (mainly drilling) provides evidence for the true subsurface geology. It was found that diffractions could be observed in situations where the coal seams are relatively parallel and the coal seam of interest is mainly towards the top of the coal measure sequence. In terms of the size of the structures that were detected, the known faults for most of the examples studied had large throws and could also be detected by conventional reflection interpretation. In one case diffractions from a known fault with a throw of less than 1 m were detected, as were diffractions from dykes less than 4m wide. These were not detectable in the reflection data.
From the computer modelling study of small, 1m throw faults, it was found that diffractions are generated from these structures and that they are present on final stacked sections. In these situations, the reflection data are not able to define the faults and it is only by a diffraction analysis that their presence can be inferred. In addition to the example of actual fault and dyke detection provided above, other real seismic survey data also showed this same behaviour and diffractions were clearly present when there was no displacement evident in the reflection surface. Presumably these diffractions are associated with minor coal‐seam structures.
As far as the differences between 2D and 3D seismic surveying are concerned, it was observed that for both types of survey, diffractions were generated at geological structures. Survey geometry is therefore not an issue for diffraction studies. However, from the 3D survey data studied, it was observed that the dip on the diffraction hyperbola was greatest on lines perpendicular to the strike of the structure. This observation is consistent with expectations and adds a further reason to the generally accepted view that seismic investigations are best conducted with lines perpendicular to the structural trend of the area under investigation.
The computer modelling has also allowed investigation of different dyke scenarios (simple dyke, dyke within a shear zone, dyke and fault within a shear zone), and the effect of random noise on the seismic shot records, the type of seismic source (explosives versus lower frequency Vibroseis), the style of faulting (reverse fault versus normal fault) and the depth of fault. In all cases, diffractions are generated at the locations of the discontinuity in the reflecting surface. The field examples also demonstrate that the choice of seismic source and style of faulting do not present issues. As discussed above, igneous dykes have been shown to be valid targets for this approach.