Comparison of Acoustic and Electrical Image Logs from the Coso Geothermal Field, CA
Authors: Nicholas C. Davatzes, Steve Hickman
Keywords: FMS, BHTV, fracture, fluid flow
Conference: Stanford Geothermal Workshop Session: Field Studies
Year: 2005 Language: English
Geo Location:
Abstract: Electrical and acoustic image logging tools provide an invaluable opportunity to characterize the fracture populations that typically control fluid flow in geothermal systems. However, these two tools detect fractures by measuring different properties of the borehole wall. Fractures interpreted from electrical image logs are identified by contrasts in resistivity with the surrounding borehole wall. By contrast, acoustic image logs are dominated by the geometry of the borehole wall; fractures are associated with breaks in the continuity of the borehole wall.

We analyzed populations of natural and induced fractures from well 58A-10 in the Coso Geothermal Field using both electrical and acoustic image logs to: (1) determine how fracture populations detected by each technique differ, (2) relate these fracture populations to temperature gradient anomalies in the borehole and the local state of stress, and (3) evaluate the use of these techniques for interpreting fracture-dominated fluid flow in geothermal systems. Consequently, a corollary goal of this on-going study is to combine insights from both techniques to interpret the origin of planar structures imaged by each technique and to refine models of fracture permeability in the Coso Geothermal Field.

Electrical image logs were obtained from the Formation Microscanner tool (FMS) which has been used extensively in the Coso Geothermal Field. The FMS tool images about 40% of the borehole circumference within four equally spaced strips and provides direct caliper measurements of the borehole diameter. The acoustic log used in this study was produced by a new state-of-the-art borehole televiewer (BHTV) being developed specifically for use in geothermal wells up to 300?C under joint funding from the Department of Energy and Navy geothermal programs. The BHTV tool images the complete circumference of the borehole and provides high quality images where the walls are fairly smooth and cylindrical. Where large faults are evident from observations in the FMS log or mud log, or where caliper data suggests widening of the borehole, the BHTV image typically has poor quality usually due to scattering of the signal due to the alignment and roughness of the borehole walls.

Comparison of FMS and BHTV logs from the same interval revealed differences in the populations of natural fractures imaged by the two techniques. The FMS log clearly records open fractures, but also appears to record some closed or potentially healed fractures as well as rock fabrics that include foliations and other banding of minerals. Foliations are characterized by large regions of subparallel planar structures of dark and bright banding which typically have consistent spacing. The boundaries of these regions coincide with fault zones and changes in image log character consistent with changing rock types. In general, fabric elements comprise more than 50% of the planar structures interpretable from the FMS log.

The residual fracture population left by subtracting these fabric elements from the total FMS data set is similar to the fracture population interpreted from the BHTV image log in frequency distribution and orientation; however, the FMS data set still indicates more fractures than the BHTV data set. In addition, those structures interpreted as fabric elements are absent, or occasionally poorly visible in the BHTV image. It seems likely that the BHTV data preferentially records open fractures that are likely to be mechanically and hydrologically important at present. In contrast, FMS data seems to provide a more complete record of the entire fracture population, possibly including healed fractures, and therefore includes more of the deformation history.

Both data sets capture changes in image log character that are associated with small perturbations of the magnetic field (probably associated with the remnant magnetization of minerals in the borehole wall) and fracture distributions. Further analysis in conjunction with the mudlog suggests that changes in image log character are also associated with transitions between rock types and changes in alteration intensity/mineralogy. We found that most fractures are associated with distinct zones of image log character, suggesting that there is an association of fractures with specific rock types or alteration. Similarly, induced fractures and breakouts preferentially occur within zones of distinct image log character suggesting these regions are distinguished by their fracture toughness and stiffness. Some of these zones demonstrate distinct, coherent subpopulations of fracture orientation, but in general fracture populations show a wide, nearly uniform distribution, especially in proximity to fault zones.

The complete azimuthal imaging of the borehole wall by the BHTV allowed the detection of breakouts and large populations of induced tensile fractures that are generally undetectable in the FMS data set. These tensile fractures have the morphology of petal-centerline fractures that are often observed in core and are interpreted to form ahead of the drill bit. These fractures occur in pairs that are systematically related to the orientation of the remote stress state, and provide a new technique for estimating the orientation of principal stresses from borehole image data.

In addition, breakouts were clearly visible as paired patches of data loss 180? apart in the BHTV image logs and were not identifiable in the FMS image log. The width of these breakouts can be used to constrain the reasonable range of stress magnitudes. Analysis of the azimuth of pairs of petal-centerline fractures and breakouts suggest that Shmin is oriented along an azimuth of 97? ± 18?, similar to other measurements from the east flank of the geothermal field and larger scale inversions from seismicity data. This orientation of Shmin is consistent with the local orientation of the Coso Wash normal fault just to the west of the well location. Research relating this stress state to the orientations of natural fractures and temperature gradient anomalies is ongoing.
Download:
File Size: 4337KB