SEISMIC WHILE DRILLING

SWD Experiments at Site 1107

While drilling at Hole 1107A, measurements of the accelerations of the drill string and of ground- and water-borne seismic energy were made over a 31-hr interval beginning at 0400 on 24 May and concluding at 1109 on 25 May. The hole was jetted or drilled with a tricone roller bit (no cores) through 373 m of sediments. Drilling rates averaged 30 m/hr for the first 228 m and then slowed to an average of 24 m/hr for the lower sedimentary section. Basement was encountered at 373 m, and drilling rates to the initial target depth of 422 m were 3 to 4 m/hr. Concurrent drill-string accelerometer and two OBS measurements were made over the depth range 170 to 422 m. OBS measurements were also made in drilling from 48 to 170 m and, after setting casing, from 422 to 493 m.

The two OBSs were deployed at ~1330 on 23 May and recovered at ~2000 on 27 May. The data from each OBS were recorded in one continuous (500 and 674 MB) file (200 samples/s, two bytes/sample, three or four channels) for 4 days. The proximal OBS (#A8) recorded three channels consisting of a vertical geophone and two horizontal geophones. The distal one (#A4) recorded a vertical and two horizontal components and a hydrophone. The accelerometer data (400 samples/s, two bytes/ sample, three channels, and 1-s headers) are in 30-min (4.464 MB) files, totaling 277 MB for the 31 hr.

The OBS data concurrent with the accelerometer data were extracted from the two complete OBS data files and converted into 40-s traces (SEGY format) for plotting and correlation purposes. The trace length was arbitrarily chosen. This produced working files consisting of 5598 traces out of the 9224 recorded.

Figure F7 compares samples of drill-string acceleration spectra during drilling in the lower sediment section at 225 m and in basement at 375 m. The principal spectral component at 2.5 Hz appears to be the drill-string rotation rate (51 rpm) times the number of roller cones (three). Similarly in the lower plot, the 3.4-Hz component appears to be the drill-string rotation rate (67 rpm) times the number of roller cones (three). The cutting action of the bit in hard rock produces more sound energy, much of it in the frequency range of 32 to 50 Hz.

In the 30-s sample of drilling in basement shown in Figure F8 , the coherence between comparable channels on the two OBS recordings is not as high as anticipated. The spectrum shows considerable energy in the interval 10 to 40 Hz, indicating that hard-rock drilling produces more and higher frequency energy than the sediment drilling. It was not possible to perform a more detailed study regarding the correlation of OBS and drill-string accelerometer data on board because the accelerometer recordings were not provided in digital form during the leg.

Two-Ship Seismic Experiment at Site 1107

During the Resolution's stay at Site 1107, the German research vessel Sonne was conducting a geophysical survey of the area (Flueh and Reichert, 1998) using 25 ocean-bottom hydrophones and ocean-bottom seismometers. Data files containing shot positions and times of four profiles that were shot by the Sonne while the two OBSs were deployed from the Resolution were transmitted from the Sonne to the Resolution via e-mail.

Based on the shot information (shot positions are shown in Fig. F9; Table T1), OBS data was converted to SEGY-format, 20-s traces with an 8 km/s reduction velocity, and added to the GEOMAR dataset. Figure F10 shows the data for the Profile P8, which passed close by. The noise level of these data is somewhat greater because of the proximity and sound level of the JOIDES Resolution. Crustal P- and S- arrivals are clearly shown, but no refracted or reflected arrivals from the mantle are recorded because of the short length of the profile. The curvature of the crustal phases indicates high-velocity gradients. The different components of the OBS recordings are revealing some information on the nature of the seismic phases (Fig. F10). The amplitude of the P-wave first arrivals is strong on the hydrophone and the vertical geophone and is very weak on the horizontal geophones. The hydrophones show strong water multiples, whereas the horizontal geophones have strong pegleg multiples (arrivals that bounce within the sedimentary layer). The pegleg multiples may be better detected by the horizontal channels because their angle of emergence is low. The vertical component of OBS #A4 has a poor signal-to-noise ratio compared to the other recordings, probably because low frequencies are filtered out by the instrument (see also Figs. F85 and F86, both in the "Hammer Drill Site" chapter). For a further interpretation, these data have to be included in the GEOMAR dataset.

Additional two-ship work had been planned (see "Vertical Seismic Profile" (in "Downhole Measurements") in the "Explanatory Notes" chapter), but time did not allow execution.

Locating the OBSs on the Seafloor

At the third deployment in 1660 m water depth, OBS #A4 was dropped at 300 m east of the hole and landed at 210 m 116º from the hole. OBS #A8 was dropped at 100 m east of the hole and landed at 81 m 165º from the hole. In both cases, this was southwest of the drop locations. Drifts were 6.5% and 5.8% of the water depth, respectively. The surveying procedure was the same as at Site 1105. The estimate precision is 30 m.

NEXT