Azimuthal Resistivity to Characterize Fractures in a Glacial Till. Mark Boris, University of Saskatchewan Jim Merriam, University of Saskatchewan

Transcription

1 Azimuthal Resistivity to Characterize Fractures in a Glacial Till Mark Boris, University of Saskatchewan Jim Merriam, University of Saskatchewan Abstract Azimuthal resistivity was used to characterize fractures in a glacial till at the King site, near Birsay SK. The King site is a hydrogeological test site where the presence of fractures has been identified, but the fracture orientation was not previously known. Preliminary azimuthal resistivity surveys indicated that the site was electrically heterogeneous. There was no interpretable anisotropy in the azimuthal resistivity data. A revised methodology was developed where the effects of heterogeneity could be removed from the azimuthal resistivity data. A detailed 3D, isotropic, geoelectric model was developed for the site. This model was used to predict the azimuthal resistivity observations from an isotropic, heterogeneous earth. These predictions were removed from the observations, leaving the anisotropy. The process resulted in an interpretable anisotropy over portions of the site where the variations in resistivity varied by approximately a factor of two. However the process did not remove the heterogeneity from portions of the site with larger resistivity contrasts. Introduction Fractures in a geologic medium can greatly influence its hydrogeological characteristics. Fractures can increase the hydraulic conductivity of an otherwise impermeable rock or soil by orders of magnitude in the dominant fracture directions. Therefore knowledge of the presence, extent, intensity, and direction of fractures is desirable for any hydraulic engineering project. The azimuthal resistivity method detects electrical anisotropy, which is related to fracture direction and intensity. The method is applied to the King site, near Birsay SK. The King site is a hydrogeological test site where the hydrogeological and geochemical characteristics have been well studied (Shaw and Hendry, 1998, Hendry et al, 2000). This site consists of a shallow, electrically heterogeneous, fractured till over unfractured till and clay formations. The orientation of the fractures was not known. The heterogeneity of the till obscures the anisotropy, making it necessary to remove this signal before the anisotropy can be investigated. If a robust method of separating the electrical anisotropy from the heterogeneities can be developed, then azimuthal resistivity can be used as a site-investigation tool for fracture characterization at any general site. Background Study Site The King site is located 140 km southwest of Saskatoon, SK., near Birsay, SK. The site was chosen for its hydrogeological characteristics. Specifically, it consists of a thick clay-till aquitard over a thick, low-permeability bedrock layer. Also, the till aquitard consists of only one geological unit, the Battleford Formation. The aquitard is also well characterized at the site

2 (Hendry and Wassenaar, 1999). The site includes an instrumented peizometer well cluster, and is the subject of ongoing geochemical and hydrogeological investigations. The King site is suitable for an azimuthal resistivity study because the upper 4m of the Battleford formation are fractured, providing an anisotropic resistivity target. Figure 1 shows the layout of the site. Piezometer wells and all measurement locations are shown. Also shown is a drainage ditch north and approximately parallel to the image lines. Geology The surface geology at the King site consists of 80m of Battleford Formation till. The Battleford Formation is a clay-rich till that is Quaternary in age and is part of the Saskatoon Group. It is typically soft, massive, unfractured, unstained, and thin. The Battleford till at the King site is anomalous in that it much thicker than it typically is. It also is fractured and oxidized (brown color) through the upper 3-4m. There is a seasonal water level variation to a depth of approximately 6m that suggests that the fractures might continue beyond the oxidized zone (Shaw and Hendry, 1998). These fractures are expected to provide the electrical anisotropy that is the target of this investigation. Azimuthal resistivity Azimuthal resistivity is a modified resistivity method where the magnitude and direction of the electrical anisotropy are determined. An electrode array is rotated about its center so that the apparent resistivity is observed for several directions (Taylor and Fleming, 1988). It is generally assumed that the anisotropy is caused by the presence of fluid-filled fractures in a relatively resistive rock or soil. Figure 1: Site layout and measurement locations

3 Azimuthal resistivity has been used to determine fracture geometry in crystalline rocks and glacial till by Lane et al (1995), Carlson et al (1996 a and b), Hansen and Lane (1996), Chapman and Lane (1996), Lane et al (1996), and Baker and McIntyre (2001). Hansen and Lane (1996), also characterize the porosity and aperture of high-angle fractures using the azimuthal resistivity method. One of the key problems with the azimuthal resistivity method is its sensitivity to lateral heterogeneities. A site exhibiting significant lateral changes in resistivity can produce azimuthal resistivity measurements where the effects of anisotropy are entirely masked by the effects of heterogeneity. Much of the published work based on this method does not account for the possibility of data corruption from lateral changes in resistivity and in fact might be presenting erroneous conclusions (Watson and Barker, 1999). Watson and Barker (1999) and Busby (2000) present arguments that at sites where the electrical anisotropy is to be determined based on single isolated measurements, an offset Wenner electrode array should be used. The offset Wenner array identifies observations that are affected by heterogeneities and ensures that only true electrical anisotropy is interpreted as such. The offset Wenner array does not, however, allow any interpretation of anisotropy at sites dominated by heterogeneity. Methods of Investigation Several azimuthal Schlumberger soundings were made at the site during September and October, They were typically made with electrode spacings (AB/2) of 3m to 25m and were made with the electrodes oriented in six directions. The locations of the soundings are shown in Figure 2. No consistent effect of anisotropy could be determined and it was clear that simple azimuthal resistivity measurements would be insufficient to characterize the electrical anisotropy at this site. The next step in the project was to determine the 3D distribution of resistivity over the site. It was hoped that this information could be used to remove the effect of lateral heterogeneity, but not anisotropy, from the azimuthal measurements. In May 2001, twelve Wenner and dipole-dipole array imaging lines were surveyed across the site. These covered the spatial extent of the peizometer wells and were parallel to the drainage ditch to the north of the site. These lines are shown in Figure 2. Each line was 120m long and had electrode spacings ranging from 2m to 40m. The data was used to generate the 3D model. Six azimuthal Wenner measurements were made in June 2001 (see Figure 2). An electrode spacing of 12m was chosen to provide a depth of investigation of approximately 4m, well into the visibly fractured till. Twelve electrode orientations were used at each station. All modeling was done using a Born approximation technique based on Loke and Barker (1995) and Moller et al (2001). The technique was modified to include three dimensions. A Monte-Carlo integration technique was used to generate the Frechet derivatives. The imaging lines were inverted to provide a 3D geoelectric model for the site. The model consisted of 4 layers (2m, 2m and 4m thick over a basement) of 21x33 4mx4m cells each. A forward model for the six azimuthal Wenner array measurements was generated using the 3D model and the same Born approximation.

4 Results and Discussion Figures 2, 3, and 4 show azimuthal resistivity patterns from the initial azimuthal Schlumberger measurements. The patterns for three electrode spacings are shown and none show consistent magnitude of resistivity or pattern of apparent anisotropy. This prompted the collection and analysis of additional data to determine the 3D distribution of resistivity at the site. Figure 5 shows the top layer of the 3D geoelectric model produced from inverting the 12 imaging lines. Note that there is much variation of the near-surface resistivity. Especially noticeable is the resistivity high at the west end of the site. This feature has been confirmed by drilling as being a gravel channel. The surface resistivity generally ranges from 3 Ohm-m to 6 Ohm-m elsewhere across the site. There is a vertical change in resistivity at 8m (median resistivity of approximately 7 Ohm-m to 8m and approximately 11 Ohm-m below 8m) in the central and eastern parts of the site. This corresponds approximately to the thickness of the oxidized till and the suspected zone of fracturing. Figure 2: Initial azimuthal resistivity. AB/2=5m. See Figure 1 for measurement locations.

5 Figure 3: Initial azimuthal resistivity. AB/2=12.5m. See Figure 1 for measurement locations. Figure 4: Initial azimuthal resistivity. AB/2=25m. See Figure 1 for measurement locations.

6 Figure 5: Top layer of 3D geoelectric model.

7 Figure 6: Azimuthal resistivity. See Figure 1 for station locations. Figure 7: Azimuthal resistivity forward model.

8 Figure 8: Anisotropy. Observation-forward model. Figure 9: Anisotropy of the three eastern stations with fracture set.

9 Figure 6 shows the unprocessed azimuthal resistivity from each of the six azimuthal Wenner measurements. As expected, there is little similarity in pattern or magnitude of the resistivity across the site. Figure 7 shows the forward model of the measurements. Figure 8 shows the anisotropy for each station. The anisotropy was determined by subtracting the forward model from the observations. The three eastern stations (6, 7, and 8) show similar anisotropy (see Figure 9), while the western stations are inconsistent. This is attributed to poor modeling of the gravel channel at the west end of the site. The inversion of the 3D geoelectric model and the forward modeling of the azimuthal measurements are not currently robust enough to remove the effects of such large heterogeneities. Further work includes developing a high-resolution model that might better define the anomaly from the gravel channel and allow it to be better removed. Stations 6, 7, and 8 indicate that there is a fracture set in the 75 to 110 degree direction (see Figure 9). Station 6 also indicates the presence of a fracture set in the 15-degree direction, but this is not evident at the other two stations. The magnitude of the anisotropy is on the order of 20%. Further work includes predicting fracture characteristics (density, width, etc.) from the magnitude of the anisotropy. Conclusion The geology at the King site consists of heterogeneous, fractured till. The lateral heterogeneities dominate the azimuthal resistivity signal, even though the fractured material was at the surface. No consistent azimuthal pattern was found at the site that could be interpreted as fracture sets. A detailed isotropic, 3D geoelectric model was developed for the site and was used to forward model the azimuthal resistivity measurements. The isotropic forward model was subtracted from the azimuthal observations, leaving only the anisotropy. In the portion of the site where the resistivity changes were moderate (ranging only by a factor of approximately two), the method worked well and a strong azimuthal anomaly was identified and interpreted as being a fracture set. In the portion of the site where the surface resistivity was dominated by a resistive gravel channel with a large resistivity contrast, the results were not interpretable in terms of anisotropy. References 1. Busby, J.P. (2000), The effectiveness of azimuthal apparent-resistivity measurements as a method for determining fracture strike direction: Geophysical Prospecting, 48, Carlson, D.A., Taylor, R.W. and Cherkauer, D.S. (1996a), Azimuthal Resistivity as a Tool for Determination of the Orientation of Preferred Hydraulic Transmissivity for a Dolomite Aquifer in Southwestern Wisconsin: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings, Carlson, D.A., Taylor, R.W. and Cherkauer, D.S. (1996 b), Use of Azimuthal Electrical Resistivity Method Used for Determining Joint Orientations in Glacial Tills Around Milwaukee, Wisconsin: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings, Chapman, M.J. and Lane, J.W. (1996), Use of directional borehole radar and azimuthal square-array D.C. resistivity methods to characterize a crystalline-bedrock aquifer: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings,

10 5. Hansen, B.P. and Lane, J.W. (1996), Orientation And Characteristics Of Fractures In Crystalline Bedrock Determined By Surface And Borehole Geophysical Surveys, Millville and Uxbridge, Massachusetts: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings, Hendry, M.J. and Wassenaar, L.I. (1999), Implications of the distribution of deltad in pore waters for groundwater flow and the timing of geologic events in a thick aquitard system: Water Resources Research, 35, Hendry, M.J., Wassenaar, L.I. and Kotzer, T. (2000), Chloride and chlorine isotopes as tracers of solute migration in a thick, clay-rich aquitard system: Water Resources Research, 36, Jansen, J. and Taylor, R. (1996), Determining fracture geometry from azimuthal resistivity data: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings, Lane, J.W., Haeni, F.P., Soloyanis, S., Placzek, G., Williams, J.H., Johnson, C.D., Muursink, M.L., Joesten, P.K. and Knutson, K.D. (1996), Geophysical Characterization of a Fractured-Bedrock Aquifer and Blast-Fractured Contaminant-Recovery Trench: SAGEEP 1996 Conference, Environ. Eng. Geophys. Soc. Proceedings, Loke, M.H. and Barker, R.D. (1995), Least-squared deconvolution of apparent resistivity pseudosections: Geophysics, 60, Moller, I., Jacobsen, B.H. and Chrisiansen, N.B. (2001), Rapid inversion of 2-D geoelectrical data by multichannel deconvolution: Geophysics, 66, Shaw, R.J. and Hendry, M.J. (1998), Hydrogeology of a thick clay till and Cretaceous clay sequence, Saskatchewan, Canada: Can. Geotech. J., 35, Taylor, R.W. and Fleming, A.H. (1988), Characterizing jointed systems by azimuthal resistivity surveys: Ground Water, 26, Watson, K.A. and Barker, R.D. (1999), Differentiating anisotropy and lateral effects using azimuthal resistivity offset Wenner soundings: Geophysics, 64,