Aspects of the lateral and vertical resolution of surface electroseismic data with implications for groundwater exploration in fractured Karoo rocks

Abstract

Horizontal bedding-plane fractures are of prime importance for the drilling of successful water-supply boreholes in Karoo rocks, but usually have very localized occurrences and small aperture widths. Geophysical methods conventionally used for groundwater exploration in Karoo rocks are unable to detect these fractures. Electroseismic (ES) methods make use of the fact that electromagnetic waves are radiated when seismic waves cause fluid flow across interfaces between porous media of different electrokinetic properties. Since the electrokinetic contrast between a fluid-filled fracture and its host is likely to be high, ES signal generation may occur at both the top and bottom interfaces of the fracture. The ability of surface ES methods to detect and resolve the fractures depends on the lateral and vertical resolution of ES data. The ES Fresnel zones are larger than their seismic equivalents and the lateral resolution of ES data is correspondingly poorer. Seismic velocity increases with depth have an additional negative impact on the lateral resolution of ES data. Assuming that a fracture has a measurable surface ES response, a simplified model shows that the response from a localized fracture may be observed at surface positions laterally displaced from the fracture. The presence of large localised unfractured areas occurring within fractured areas may not be detected at all. Rayleigh’s vertical resolution criterion applied to electroseismic data requires that an embedded thin layer have a minimum thickness of half the dominant wavelength of the seismic wave. Since the Biot slow wave has a much smaller wavelength than the Biot fast wave, it would seem that the limits of vertical resolution are determined by the dominant wavelength of the slow wave. The dissipative nature of the slow wave, however, may cause the practical limit of vertical resolution to be more dependent on the wavelength of the fast wave. The detection of a thin embedded layer depends on the conditions of constructive and destructive interference of signal generated at its top and bottom interfaces. For both the fast and slow wave a thin bed with a thickness less than a quarter of the dominant wavelength results in a total ES response that is weaker than the response from the top interface alone. For both fast and slow pressure waves this thickness is much larger than the aperture width of the bedding-plane fractures in Karoo rocks, and detection of these fractures with surface ES methods seems unlikely.