154 Geophysics in Geothermal Exploration However, for designing properly a vertical deep drilling operation, which is risky and costly, there was a strong need to characterize the overlying sedimentary units. Thus, 2D vintage seismic reflection profiles done for oil exploration in 1984 mainly to image the Mesozoic and the Cenozoic formations were reprocessed and reinterpreted for geothermal targets at the sedimentary-basement interface (Munck et al., 1979). Moreover, a series of former oil wells was used for calibrating the reprocessed 2D seismic profiles and stratigraphy. Therefore, the provisional geological profile of the first well GPK-1 between surface and the top basement was very accurate thanks to the high density of old petroleum wells drilled in this area. At regional scale, older public gravity and magnetic measurements (Rotstein et al., 2006) indicated that the Bouguer anomaly was delineating a large area corresponding to a granite batholith already proved by a core taken in the old petroleum well, 4616, located at SsF that reached the top of the granitic basement at 1380 m depth. With the drilling of GPK-1 and its stimulation in 1988, three old petroleum wells surrounding the site (4598, 4601, 4616) were reopened in order to instrument downhole three-directional permanent probes for monitoring induced seismic activity during stimulation operation. Those probes were designed to withstand high temperature (125 °C) and severe corrosion conditions at the bottom of the holes. In parallel to the coring of the HDR exploration well (EPS-1), three old petroleum wells, 4550, 4601 and 4616 were deepened in 1990 to about 1500 m in order to instrument down-hole seismic sensors. The fact that those sensors would have been installed in the deep Triassic sandstone or the Carboniferous basement, allowed enhancing detection of very low magnitude events and reducing the uncertainty on the location of the induced events. An additional peripheral seismic observation well, OPS-4; was drilled in 2000 and located less than 2 km south of GPK-2. It started from surface to 1540 m in the Lower Triassic formation in order to reduce azimuthal bias during the seismic monitoring. In parallel, a permanent network of surface seismic stations was installed and regularly densified according to stimulation and circulation phases. The drilling of the first well GPK-1 to 2000 m and its later deepening to 3600 m were also a good opportunity to use innovative image log tools based on acoustic (BHTV, UBI) and electrical methods (FMS, FMI, ARI). In 1987, FMS tool was used probably for the first time in continental Europe for characterizing the fracture depth, the fracture azimuth and their dip in the granite section. Moreover, drilling induced tensile fractures were also characterized for measuring the orientation of the main horizontal stress. Detailed interpretation of standard geophysical logs in the basement of GPK-1 (Traineau et al., 1991) was achieved as well as detailed comparison between borehole image logs and continuous coring done in EPS-1 (Genter et al., 1997). During this early phase of reconnaissance (1987-1991), VSP (Vertical Seismic Profiling) was carried out to better image the fault structures close to the GPK-1 well and for improving the velocity model allowing an accurate location of seismic events to be recorded during stimulation experiments. Moreover, results of VSP were used to reprocess two vintage reflection seismic lines crossing the SsF area. Based on the interpretation of five 2D seismic lines, a 3D geological
RkJQdWJsaXNoZXIy NjA3NzQ=