QUALITY OF LWD LOGS

Figures F4 and F5 show the quality control logs for Holes 1173B and 1173C. Target ROPs of 50 m/hr in the interval 120-350 mbsf, 35 m/hr in the interval 350-450 mbsf, and 50 m/hr in the interval 450 mbsf to total depth (TD) were chosen in Hole 1173B. A target ROP of 60 m/hr was chosen in Hole 1173C. To record one sample per 15 cm, the ROP must be maintained at 65 m/hr or lower. This was achieved for 96% of the total section in Hole 1173B and 76% of the total section in Hole 1173C (Figs. F4, F5). Although RAB image resolution would improve with even slower drilling rates, the quality of RAB images is high (~3-cm pixel) and no significant resolution loss is observed between the higher and lower ROP intervals in Hole 1173B.

The differential caliper (DCAL) is the best indicator of borehole conditions. The bulk density correction (DRHO), calculated from the difference between the short- and long-spaced density measurements, varies from 0 to 0.1 g/cm³ (Figs. F4, F5), which shows the good quality of the density measurements in Holes 1173B and 1173C. A standoff of <1 in between the tool and borehole wall indicates high-quality density measurements with an accuracy of ±0.015 g/cm³. The differential caliper values are <1 in over 98% of the total section in Hole 1173B. Only the uppermost 20 m of the hole and several small intervals through the section show minor washouts. The differential caliper values are <1 in over 52% of the total section in Hole 1173C (Fig. F5). The uppermost 75 m was washed out to 1-2 in because of soft sediments, and the bulk density measurements in this interval are not accurate.

Time-after-bit (TAB) measurements are 10-20 min for resistivity and gamma ray logs and 30-70 min for density and neutron porosity logs. TAB values as long as 140 min from 710 mbsf to TD coincide with low ROP while drilling in basement rocks. However, the low values of the DCAL and DRHO logs indicate that the borehole condition was not significantly deteriorated in this interval.

Preliminary waveform processing and evaluation were carried out by Anadrill-Schlumberger in Houston, Texas, during and immediately after Leg 196. Raw waveforms from the four receivers of the ISONIC tool are shown for two depths (262.15 and 397.58 mbsf) in Figure F6. They show, in order of arrival, drilling "noise" phases and a dispersive wavetrain of a leaky P-wave mode traveling in the borehole and along the formation/borehole wall interface.

The formation P-wave velocity may be estimated from the leaky mode wavetrain; at the two depths, the arrival of the leaky mode changes. The contour plots of coherence vs. slowness and time show peaks corresponding to the leaky P-wave arrivals. For the shallow depth (Fig. F6C) the coherence shows a single broad peak at a slowness of 180-190 µs/ft (velocity of 1604-1693 m/s), where it may be difficult to extract the P-wave velocity because of the frequency and slowness of these phases in Holes 1173B and 1173C. For the deeper depth (Fig. F6D), arrivals are more clearly separated, with the earliest one corresponding to a coherence peak at a slowness of ~160 µs/ft (velocity of 1905 m/s). Such separation may improve subsequent analyses by bandpass and wavenumber domain filtering. Processing of the ISONIC waveforms will be conducted postcruise to attempt to extract the P-wave velocity. Preliminary velocity values (Fig. F32) are probably largely unreliable and should be used with a great deal of caution (see "Logs and Physical Properties").

Figure F7 shows a correlation diagram for the overlapping interval between Holes 1173B and 1173C, using the ring and deep resistivity logs. The curves have similar shapes, including three significant peaks that can be correlated between 120 and 175 mbsf. Peak A at 135.25 mbsf, Peak B at 147.59 mbsf, and Peak C at 164.82 mbsf in Hole 1173B correlate to Peak A´ at 130.07 mbsf, Peak B´ at 141.04 mbsf, and Peak C´ at 157.50 mbsf in Hole 1173C, respectively. This suggests Hole 1173C is ~7 m deeper and ~2 m longer than the equivalent section in Hole 1173B. The depth measurement of Hole 1173B is reliable with <0.5 m uncertainty as indicated by the physical depth of the casing pipe (120.9 mbsf) being close to the measured depth of the casing shoe (120.6 mbsf) and the close correlation of the wireline log depth recorded in Hole 1173A with the LWD logs in Hole 1173B. Possible reasons for the depth shift in Hole 1173C include (1) hole deviation, (2) poor drilling control in shallow soft sediment, and (3) depth errors in identifying the seafloor in Hole 1173C.

Based on the correlation of the overlapping interval between Holes 1173B and 1173C, a composite log is generated using the following method:

1. Because the depth recorded in Hole 1173B is reliable, the logs in Hole 1173C are depth shifted and spliced to the logs in Hole 1173B.
2. For bulk density and neutron logs the section below Peak B in Hole 1173B is used in its entirety. In Hole 1173C, these logs are not recorded below Peak B´.
3. In Hole 1173C, the logs from the seafloor to Peak B´ are compressed proportionally to 95.6% of their original depth and spliced in the section from the seafloor to Peak B in Hole 1173B.