The processed P-wave slowness logs from both passes are shown in Figure F5. Only those intervals where reliable P-wave slowness is computed are shown, as well as the preliminary shipboard P-wave logs from both passes for comparison. Both passes are generally of high quality over the entire interval. The trends of the shipboard and reprocessed logs agree, and variations in the shipboard log above 180 mbsf are eliminated. Enlarged hole size and switching between the high-energy fluid and P-wave modes affected the shipboard log in this interval. Note that below 260 mbsf, the reprocessed log for pass 1, but not pass 2, is 10-15 ms/ft slower than the shipboard log.
The processed S-wave slowness logs from both passes are shown in Figure F6. Only those intervals where reliable S-wave slowness is computed are shown, as well as the preliminary shipboard logs from both passes for comparison. The S-wave logs are of high quality over the interval from 226 to 350 mbsf for pass 1 and from 65 to 360 mbsf for pass 2. Above 226 mbsf, pass 1 data have poor signal-to-noise ratio and reliable values could not be picked. In the overlapping interval, reprocessed S-wave logs from both passes generally follow the trends of the shipboard logs. Above 226 mbsf, the reprocessed data from pass 2 measure S-wave slowness values up to 1100 ms/ft. This is 50-150 ms/ft higher than the shipboard values over this interval. The pass 2 reprocessed log confirms the presence of the velocity inversion below 350 mbsf, and S-wave slowness values reach 1000 ms/ft (VS = 305 m/s) at their maximum. Use of the low-frequency dipole source enables sufficient signal-to-noise ratio in these intervals to accurately extract the S-wave velocity.
The processed P- and S-wave logs from pass 1 and 2 are compared in Figure F7. The trend and vertical resolution of both P-wave passes agree closely from 65 to 360 mbsf. Core plug measurements (vertical orientation) are systematically 8-12 ms/ft higher than both passes. The offset between the core and log measurement can be attributed to the higher porosity (lower velocity) of core measurements made under ambient surface conditions (e.g., Goldberg et al., 1987). In this case, porosity rebound of the core measurements varies from 1.5% to 6.5%, with the largest difference occurring at ~226 mbsf, perhaps related to lithologic changes at this depth.
The P-wave slowness values from pass 1 are systematically 2-4 ms/ft higher than those from pass 2. The peak P-wave amplitude and frequency also differ between the two passes. Peak frequency is ~1.5 kHz (20%) higher, and amplitudes are three times higher in pass 1 than pass 2. Testing indicates that these differences are not due to filtering prior to processing. The monopole source was nominally operating at the same frequency between passes, and it is unlikely that use of the DSI low-frequency dipole source during pass 2 could affect the monopole transmitter frequency, according to Schlumberger. They have completely different circuits, and line voltage fluctuations will not change the frequency of the source. Since hole conditions did not change, a possible explanation of these offsets is a change in centralization of the tool between passes. Tool eccentralization will cause drastic reduction in amplitude of the monopole signal (e.g., Goldberg et al., 1984), and this reduction could introduce frequency and phase-picking shifts. Further studies may be undertaken to extract the P-wave component of the dipole signal (a low-amplitude precursor to the flexural mode) for comparison. This mode propagates differently in the borehole and could help to explain the observed offsets between the two passes using the monopole source.
The S-wave logs agree well over the interval 226-350 mbsf where the two intervals overlap. The trend and vertical resolution of both passes reproduce closely. As noted above, pass 1 dipole data have poor signal-to-noise ratio in low-velocity intervals. Pass 1 S-wave slowness values are not reliable above 226 mbsf or below 350 mbsf. The low-frequency dipole source used for pass 2 gives reliable slowness values over the entire logged interval and confirms the velocity inversion below 350 mbsf. The S-wave slowness values for pass 2 above 226 mbsf are significantly greater than the shipboard results. Use of the unprocessed S-wave velocities may introduce errors of 20% or more in these shallow sediments.
Figure F8 shows crossplots of P- vs. S-wave slowness from shipboard results and from the current reprocessing. The difference in the range of P- and S-wave slowness is considerable, but both data sets have relatively linear relationships with correlation coefficients >0.80. Data points above P-wave values of 180 ms/ft have significantly greater scatter. Linear regression of the processed results yield the relationship
A similar regression of the shipboard data yields
which has ~33% greater slope than for the reprocessed logs. Core measurements of P- and S-wave velocity should be compared to the reprocessed logs to further confirm the improved accuracy and reliability of these results.