Lawrence Berkeley National Laboratory

Electrical resistivity may contribute to progress in understanding geothermal systems by imaging the geometry, bounds and controlling structures in existing production, and thereby perhaps suggesting new areas for field expansion. To these ends, a dense grid of magnetotelluric (MT) stations plus a single line of contiguous bipole array profiling has been acquired over the east flank of the Coso geothermal system. Acquiring good quality MT data in producing geothermal systems is a challenge due to production related electromagnetic (EM) noise and, in the case of Coso, due to proximity of a regional DC intertie power transmission line. To achieve good results, a remote reference completely outside the influence of the dominant source of EM noise must be established. Experimental results so far indicate that emplacing a reference site in Amargosa Valley, NV, 65 miles from the DC intertie, is still insufficient for noise cancellation much of the time. Even though the DC line EM fields are planar at this distance, they remain coherent with the nonplanar fields in the Coso area hence remote referencing produces incorrect responses. We have successfully unwrapped and applied MT times series from the permanent observatory at Parkfield, CA, and these appear adequate to suppress the interference of the cultural EM noise. The efficacy of this observatory is confirmed by comparison to stations taken using an ultra-distant reference site east of Socorro, NM. Operation of the latter reference was successful by using fast ftp internet communication between Coso Junction and the New Mexico Institute of Mining and Technology, using the University of Utah site as intermediary, and allowed referencing within a few hours of data downloading at Coso. A grid of 102 MT stations was acquired over the Coso geothermal area in 2003 and an additional 23 stations were acquired to augment coverage in the southern flank of the first survey area in 2005. These data have been inverted to a fully three-dimensional conductivity model. Initial analysis of the Coso MT data was carried out using 2D MT imaging. An initial 3D conductivity model was constructed from a series of 2D resistivity images obtained using the inline electric field measurements (Zyx impedance elements) along several measurement transects. This model was then refined through a 3D inversion process. This model shows the controlling geological structures possibly influencing well production at Coso and correlations with mapped surface features such as faults and regional geoelectric strike. The 3D model also illustrates the refinement in positioning of conductivity contacts when compared to isolated 2D inversion transects. The conductivity model has also been correlated with microearthquake locations, well fluid production intervals and most importantly with an acoustic and shear velocity model derived by Wu and Lees (1999). This later correlation shows the near-vertical high conductivity structure on the eastern flank of the producing field is also a zone of increased acoustic velocity and increased Vp/Vs ratio bounded by mapped fault traces. South of the Devil's Kitchen is an area of high geothermal well density, where highly conductive near surface material is interpreted as a clay cap alteration zone manifested from the subsurface geothermal fluids and related geochemistry. Beneath the clay cap, however, the conductivity is nondescript, whereas the Vp/Vs ratio is enhanced over the production intervals. It is recommended that more MT data sites be acquired to the southwest of the Devil's Kitchen area to better refine the conductivity model in that area.