IODP Proceedings    Volume contents     Search
iodp logo

doi:10.2204/iodp.proc.304305.103.2006

Downhole measurements

Operations

Hole U1309B

After reaching a depth of 101.8 mbsf, Hole U1309B was conditioned and filled with freshwater and the bit released at the bottom of the hole. The pipe was pulled to 35.7 mbsf, and three tool strings were deployed (Fig. F282A). The first deployment consisted of the triple combo tool string, which contained the Hostile Environment Natural Gamma Ray Sonde (HNGS), the Accelerator Porosity Sonde (APS), the Hostile Environment Litho-Density Sonde (HLDS), the Dual Laterolog (DLL), and the Lamont-Doherty Earth Observatory (LDEO) TAP tool. The APS was not used on the first pass because it would activate the formation during the neutron capture phase as gamma rays are emitted. For the second pass, the pipe was pulled up to 21.6 mbsf. The second pass was logged to the seafloor with the APS minitron on.

The second tool string deployment consisted of the Scintillation Gamma Ray Tool (SGT), the General Purpose Inclinometry Tool (GPIT), the FMS, and the Dipole Sonic Imager (DSI). The first pass proceeded without difficulty, and the FMS arms were closed several meters before the logging head entered pipe. During the second pass, when the bottom of the FMS began coming into the pipe, some overpull was recorded. The tool string was brought to the surface to check for damage. It appeared that the outer cover of the caliper that serves as housing for the pad wiring was broken and the arm expansion springs were fully extended.

The third tool string consisted of the GPIT and the SGT with spacers. This last run was devoted to the Offshore Service Unit-F-model Modular Configuration MAXIS Electrical Capstan Capable (OSU-FMEC) heave compensator tests. For reference, a first pass with the LDEO-Borehole Research Group wireline heave compensator (WHC) at 900 feet/h and a station log was run. Next, the OSU-FMEC was used, and three passes were performed at different speeds (900, 300, and 1800 ft/h), followed by stationary logs. These data will be used to evaluate and improve the new heave compensator system.

All logging passes reached 96 mbsf, or 7 m above the maximum penetration. Borehole conditions were excellent during the three runs, and no ledges or obstructions were encountered. Additionally, the drill pipe was held a depth of 21.6 mbsf before tool string reentry into pipe during each run, except for the first triple combo pass.

Hole U1309D

Three phases of logging operations were performed in Hole U1309D (Fig. F282B). During Expedition 304, Phase 1 (0–400 mbsf) was accomplished. Phases 2 (400–836 mbsf) and 3 (836–1414.5 mbsf) were accomplished during Expedition 305.

Phase 1 (0–400 mbsf)

A first attempt to log Hole U1309D with the triple combo was made on 31 December 2004. The drill pipe was set inside the casing. We could not get past an obstruction at 39 mbsf. A second run was made by using a short tool string (DLL and HNGS), but we could not pass the same obstruction. As we were using an 11 inch bit, there was no way to move the pipe to pass the obstruction in the 9⅞ inch hole. This logging attempt ended on 1 January 2005 at 0600 h.

After doing some shallow spot coring with the 11 inch bit (see “Operations”), a round trip was made to replace the bit with a 9⅞ inch cleanout bit. On 2 January at 0700 h, the second logging attempt started with the triple combo (Fig. F282B). The pipe was set at 64 mbsf so it would be below the known obstruction zone. While exiting the pipe, the triple combo set down at 74 mbsf. The pipe was moved up to 54 mbsf, and, after working the tool string, it went past 74 mbsf. It shortly set down at 79 and 96 mbsf, but finally went to bottom. After we had a problem with the Schlumberger Minimum Configuration Maxis acquisition software while coming off bottom, the computer was rebooted. There was some activation over the first 5 m from the APS on the subsequent logs. A repeat pass was performed at the bottom of the hole to minimize the effect of APS activation.

The second tool string deployed was the FMS-sonic, which went to bottom with no problems. On the first pass, we had some overpull at 129 mbsf and the tool was partially closed before it became free. Calipers were opened again and there was more overpull at 99 mbsf, where we had planned to close the calipers to enter pipe. The second FMS-sonic pass proceeded without difficulty.

After changing over to the accelerometer to test the Schlumberger heave compensator, the tool stopped at 67 mbsf. Working for 30 min, the tool string advanced to ~69 mbsf. At that point, we decided to make sure we could still come back uphole. It appears that a piece of the borehole wall had probably fallen in and wedged the tool in place. Half the tool was in pipe and we could not go up or down. We finally freed the tool after working it for an additional 30 min. Any further attempts to log the hole were cancelled. A 15 min OSU-FMEC heave compensator test in pipe at 900 feet/h up and a stationary test were performed.

During all logging runs in both holes, the WHC was turned on following the exit from pipe and used continuously while the tool strings were in the open hole. Following acquisition, logging data were transmitted to LDEO for depth and environmental correction processing and subsequently returned to the ship. All data presented in this chapter have been processed and depth-shifted as described in “Processing and data quality.”

Phase 2 (400–836 mbsf)

After reaching a depth of 837.4 mbsf, Hole U1309D was conditioned and filled with freshwater for logging operations. The pipe was tripped, and a logging bottom-hole assembly (BHA) with logging bit was run into the hole. To avoid obstructions and to allow an overlap with the previous logging runs from Expedition 304, the pipe was set at 170 mbsf, and a total of five tool strings were deployed (Fig. F282B). The first deployment consisted of the triple combo tool string, which contained the HNGS, the APS, the HLDS, the DLL, and the LDEO TAP. The first pass covered the interval from the bottom of the hole at 836.4 mbsf to the pipe. The pipe was identified in the logs at 170 mbsf. For data quality check, a short repeat pass was run in the lowermost 120 m. The second tool string deployment consisted of the DSI, the SGT, the GPIT, and the FMS. The first pass covered the interval from 836.4 to 350 mbsf, and the second pass logged the entire open hole up to the pipe. Because of time constraints and the need to wait for daylight for marine mammal watch associated with the seismic check shot work, 3 h was devoted to heave compensator tests with the OSU-FMEC. One joint of the pipe was removed, and the pipe was set to a depth of 161 mbsf before tool string reentry. The third tool string consisted of the three-component Well Seismic Tool (WST-3). The tool was lowered to the end of the pipe and halted in a safe position. No marine mammals were sighted during a 1 h watch, so the air gun was soft-started by increasing pressure levels over 0.5 h. After clearance and during the soft-start, the tool was lowered to the bottom of the hole and subsequently clamped. In 50 m intervals, a total of 13 stations were occupied for check shot recording (Table T25). For the two lowermost stations, no clear arrival could be identified in the seismograms, but all other stations had good signal-to-noise ratio for many individual and all stacked seismograms. The fourth run included the Ultrasonic Borehole Imager (UBI), the GPIT, and the SGT. A sinker bar was added to speed up lowering of the tool in the pipe. The tool was not lowered to total depth but only to 824.4 mbsf to protect the sensor at the end of the tool string. The first pass from 824.4 to 724 mbsf was acquired in high-resolution mode at a speed of ~120 m/h. Subsequently, the tool was lowered again to 824.4 mbsf and the normal mode, at a speed of ~260 m/h, was started. However, because of software problems in detecting the first arrival of the signal, no consistent data could be obtained with normal mode. It was decided to abandon this acquisition mode and to conduct additional high-resolution measurements in the depth interval (704–503.2 mbsf) where good hole conditions had been documented by caliper data and where lithology is significantly variable. The GBM was deployed in the fifth run. A low-magnetic sinker bar was added and the tool had no problems reaching total depth. The tool obtained data for the downhole, as well as the uphole pass at a speed of 360 m/h. The tool also measures temperatures within the housing, which are within ±2°C of ambient borehole temperatures. Prior to and after logging, the gyrocompass of the tool was oriented with the vessel heading. Vessel heading parameters at those times were provided by the bridge. The data from the GBM were sent to Goettingen (Germany) for further processing, and preliminary results were returned to the ship. All data presented in this chapter have been processed and depth-shifted as described in “Processing and data quality.”

Phase 3 (836–1414.5 mbsf)

After reaching a final depth of 1415 mbsf, Hole U1309D was conditioned and filled with freshwater for logging operations. The pipe was tripped, and a logging BHA with logging bit was run into the hole. The pipe was set at 194 mbsf and three tool strings were rigged up, but only two were successfully deployed (Fig. F282B). The first deployment consisted of the triple combo tool string. The first pass covered the interval from the bottom of the hole at 1415 mbsf to the pipe. The pipe was identified in the logs at 194 mbsf. For data quality check, a short repeat pass was run in an interval of low recovery (1270–1096 mbsf). The second tool string deployment consisted of the DSI, the SGT, the GPIT, and the FMS. After reaching the bottom of the hole, communication to the lower part of the tool was lost. It was concluded that the joint between transmitter and receiver section in the DSI must have been damaged, and the tool was pulled back to the rig floor. The broken joint was identified, and the DSI was removed. Hence, no sonic data were acquired during this phase. The remaining SGT, the GPIT, and the FMS string were lowered back into the hole. A successful first pass was recorded from total depth to 734 mbsf, and a second pass was run from total depth to 629 mbsf. The third tool string consisted of the WST-3. The tool was lowered to the end of the pipe and halted in a safe position. After 1 h of marine mammal watch, the air gun was soft-started by increasing pressure levels >0.5 h. After clearance and during the soft-start, the tool was lowered to the bottom of the hole. After reaching total depth, communication to the tool was problematic, and high temperatures were suspected to have caused electronic problems. The tool was pulled to shallower depth, but the problems remained. The tool was pulled out of the hole and replaced by the single-component WST. After lowering the tool into the hole, one test position at 328 mbsf was chosen for a tool check. It recorded successfully, and lowering of the tool to total depth was begun. Midway downhole, heave had increased beyond safe deployment conditions, and it was decided to terminate the logging operation.

When pulling back the tool, a sudden increase in tension (1500 lb over normal) was registered and continued to be erratic, indicating a problem with the wireline. When approaching the surface, the tool speed was reduced at ~600 meters below rig floor (mbrf) and approach to rig floor was slowed to prevent any damage to the oil-saver/circulation sub in the top of the drill pipe. At 570 mbrf, the tension increased to 6000 lb and the tool was stopped; tension was released, and the oil-saver was removed. It was decided to recover the wireline with T-bars by pulling it up in ~30 m intervals and cutting it in sections. A series of wireline knots was subsequently recovered. The WST was pulled back to the rig floor without any damage.

During all logging runs, the WHC was turned on following the exit from pipe, and it was used continuously while the tool strings were in open hole. Following acquisition, logging data were transmitted to the LDEO for depth and environmental correction processing (see “Downhole measurements” in the “Methods” chapter); these data were then returned to the ship.

Processing and data quality

Hole U1309B

The original logs were shifted to the seafloor depth given by the drillers (1653.4 mbrf) and depth-matched to the gamma ray log and FMS images from pass 2 of the FMS-GPIT-SGT-DSI tool string. The density log from the triple combo tool string was depth-matched to the FMS images because the extremely low gamma ray values (<10 gAPI) made matching by gamma ray data alone difficult. The seafloor depth could not be determined from gamma ray values, also because of the very low values; therefore, the seafloor depth given by the drillers was used. The quality of the data was assessed by checking against reasonable values for the logged rock types, by repeatability between different passes of the same tool. Hole U1309B diameter was recorded with the caliper on the HLDS (LCAL) and on the FMS (C1 and C2) (Fig. F283) tools. The borehole was in good condition, typically 10 inches wide except for occasional small enlargements, the largest of which is at 61 mbsf, where the calipers reach 15 inches.

The excellent hole conditions resulted in particularly good measurements in the contact tools, such as density, porosity, and FMS. Good repeatability is observed between the two passes. Hole deviation, as measured by the GPIT, reached 7° (hole azimuth 040°–045°) in the lower 40 m of the hole, which contributed to the poor sonic data. The DSI was operated in the following modes: for pass 1, P- and S-wave mode monopole and for pass 2, P- and S-wave mode monopole, lower dipole, and Stoneley mode. The acoustic velocities from pass 1 seem to be in reasonable value range for gabbroic rocks but do not match the patterns in the other physical property logs (density, electrical resistivity, and FMS). Further processing postcruise may improve data usefulness. No TAP data could be recovered from the tool in Hole U1309B, and the reasons for tool failure are unclear. The quality of the FMS images recorded in Hole U1309B is excellent. Flushing with resistive freshwater mud improved FMS response as it reduced conductivity contrast between the borehole fluid and the wallrock. The contact between the FMS pads and the borehole wall is confirmed because maximum aperture measured by the calipers is 15.2 inches, less than the maximum possible aperture (16 inches).

Hole U1309D

Phase 1 (0–400 mbsf)

A selection of logs acquired during Phase 1 is presented in Figure F284A. Their uppermost limit corresponds to the pipe depth, as indicated by the dashed horizontal line. Their lowermost limit represents the deepest depth achieved by the various probes.

In Hole U1309D, FMS images were recorded between 400 and 96 mbsf. For the majority of the borehole, the walls are relatively smooth and uniform with a typical diameter of ~11 inches (~28 cm). More irregular sections of the borehole are located within the upper interval, especially in the intervals between 113 and 145, 222 and 230, and 243 and 249 mbsf. Within these zones, there are several intervals where the diameter of the borehole exceeds the limits of the FMS calipers, and in these locations, image quality is expected to be low. Within the remainder of the borehole, data are continuous and of very high quality. Furthermore, in some intervals (for example, at 96–117, 236–248, and 273–400 mbsf), the two passes did not follow the same path, which increases the wall coverage.

The original logs were shifted to the seafloor depth given by the drillers (1656 mbrf). Caliper data indicate that Hole U1309D is generally in good shape, especially between 286 and 400 mbsf with a gauge hole (C1 and C2) (Fig. F284A). The top section is in good shape but has several irregular sections ranging from 13 to 18 inches, especially in the intervals between 113 and 145, 222 and 230, and 243 and 249 mbsf. The DSI data were good in some sections but very noisy throughout the irregular areas of the hole. The hole deviation, as measured by the GPIT, reaches 1.75° at 125 mbsf and smoothly decreases to <1° from 196 mbsf to the bottom of the hole.

Phase 2 (400–836 mbsf)

Conditions in Hole U1309D continued to be very good, particularly in the deepened part from 400 to 836 mbsf (Fig. F284B). The excellent conditions allowed particularly good measurements for contact tools such as density, porosity, and FMS. No TAP data could be recovered from the tool, and the reason for tool failure was suspected internal connection problems. Comparing the caliper data from the Phase 1 and Phase 2 logging runs of the upper 400 m, progressive enlargement with continued drilling is evident (Fig. F284B). This supports the strategy of implementing intermediate logging runs to ensure good data quality for contact measurements.

The logs were shifted to the seafloor depth given by the drillers (1656 mbrf) and depth-matched to the Phase 1 logging run of the upper 400 m. As in the topmost interval, the resistivity log from the triple combo tool string was depth-matched to the FMS images. The GBM data were depth-matched to the total magnetic field data recorded by the GPIT.

The measurement quality of natural gamma ray data is good; however, the natural radioactivity of the formation measured with this tool is overall very low (<5 gAPI). As with Phase 1 data, potassium, thorium, and uranium concentrations derived from the spectrum are below the detection limit, so quantitative values for these three elements should not be trusted.

Magnetic field intensity and direction were recorded by the GBM and GPIT (here GPIT data are taken from the Expedition 305 FMS-sonic tool string). The GBM setup did not allow direct depth measurement, only magnetic field versus time. We obtained the corresponding logging depth by correlation with the Schlumberger time-depth log output from the logging winch. The GBM tool recorded data during both the downhole and uphole run (Fig. F285). Because of tool rotation, the horizontal field components x and y oscillate around the zero line. The vertical field component z shows a high level of repeatability for the downhole and uphole logs. In addition to the GBM fluxgate sensors, the angular rate of the GBM tool around the x-, y-, and z-spin axes was measured using three fiber-optic gyros. The rotation record of the tool is determined by the accumulation of the angular rate during the logging run. Rotation data will be used for reorientation of the magnetic data. Gyro temperature and telemetry temperature were also recorded in order to allow temperature correction for the gyros and to obtain chassis temperature, which closely relates to borehole temperature. Inclinometer data around x- and y-axes were also recorded. Because there are no centralizers used, the inclinometer data are required to determine what, if any, correction is required for trade-off between horizontal and vertical component.

Onshore processing of the gyro data is required to correct for the orientation of the tool. This will allow determination of the horizontal field components. The magnetic field intensity derived from GBM logging differs slightly from the International Geomagnetic Reference Field expected at this latitude (horizontal field = 27,483 nT; vertical field = 30,313 nT; total field = 40,918 nT; inclination +47.8°) (Fig. F286). The in situ magnetic field inclination (raw data) averages +51.11° (downhole run median = +51.36°) and +52.04° (uphole run median = +51.79°) for the open hole depth interval. The GPIT inclination averages +51.42° (median = +52.13°). The two magnetic tool logs agree well with respect to the sequence of magnetic variations, although they show differences in magnitudes (Fig. F286) and a slight offset in the vertical component.

The FMS and UBI data were processed to correct for acceleration and sticking that occurred during the uphole logging. The two passes of the FMS were depth-matched, with the second pass as reference. Although the FMS images could be corrected with confidence, the UBI images show artifacts of sticking even after speed correction. One possible explanation is that because of the low tool speed, the heave of ~3 m during the data acquisition was not entirely compensated. Depth-matching the UBI data to the FMS passes alleviated some of those effects. All resistivity scan data sets were statically and dynamically normalized during conversion to color images. In most intervals, the coverage of the borehole wall by the FMS is excellent and is complemented by the UBI images.

Phase 3 (836–1414.5 mbsf)

Data recorded during the last logging run have been depth-matched to the upper logging curves. Triple combo data show high data quality and allow an excellent overlap with the previous logging run. Between 630 and 655 mbsf, the triple combo caliper (LCAL) recorded borehole diameter data smaller than the bit size, which may be due to electronic failure within this depth interval because no obstruction could be detected while lowering the triple combo. Two FMS passes were acquired, and overall images repeated very well. Caliper data registered no obstruction in the interval mentioned above. Poor data quality was typical between 1275 and 1180 mbsf within the first pass. The second pass had no problems, and high-quality images were gained. The check shot station recorded by the WST will be discarded because of wireline problems that occurred during this run. Because depth is measured by cable length and there were knots in the wireline, the exact depth determination of this check shot station is unreliable (see “Operations,” above).

In general, logging data and core petrophysical data compare very well (Fig. F287). The density and velocity data are in the same range for core samples and logging measurements. The main difference is between the sample porosity (see “Physical properties”) and the neutron porosity logs. The logging values are much higher than the core sample data; however, they follow similar trends. The reason is that neutron porosity measurement is sensitive to all hydrogen (such as in pore water or bound in minerals) present in the formation (Serra, 1984), whereas sample porosity reflects only pore water. Neutron porosity curves are calibrated to limestone porosity, which may further contribute to some discrepancies.

Results

Natural radioactivity

Formation natural radioactivity was measured during each run with two different tools. The high-resolution curve from the SGT and spectral data from the HNGS are shown in Figure F283. The different gamma ray tools show the same general patterns throughout the logged interval. Total gamma ray values (HSGR) obtained with the triple combo tool string range from 0 to 12 gAPI in both holes. Such very low gamma ray measurements are expected in oceanic crust. Significant increases observed on the total gamma ray log occur at 56.5, 64.1, and 73.4 mbsf in Hole U1309B and correlate with minor tool responses in electrical resistivity, porosity, and density measurements. In Hole U1309D, several significant increases in gamma ray are recorded, for example at 172, 373, 632–638, 742–756, and 866–872 mbsf (Fig. F284A). The spectral gamma ray measurements also show very low values that, in many instances, are below the tool detection limits for Th (0.7 ppm), U (0.35 ppm), and K (0.18 wt%). Potassium values are low, between 0 and 0.1 wt%. Thorium and uranium values are mostly between 0 and 1 ppm.

Neutron porosity and capture cross section

The neutron porosity in the logged section of the hole ranges from 56.6% to 3.2%. Absolute values in the neutron porosity log do not match the discrete laboratory porosity measurements, which are lower. This may be caused by the presence of bound water in the wallrock and/or sampling bias due to the fact that fractured intervals commonly result in poor recovery and the rocks recovered have veins and fractures that are rarely sampled for physical property measurements. In Hole U1309B, neutron porosity maxima occur at 36.8, 41.6, 46.4, 52.6, and 66.9 mbsf (Fig. F283). These high values may correspond to the tool losing contact with the borehole wall, a high degree of fracturing, the presence of hydrous alteration minerals such as clay minerals in veins, or altered groundmass (see “Metamorphic petrology”). These high neutron porosity values generally correlate with lows in the density log and low resistivities. Within the logging interval 57.6–61.5 mbsf, high neutron porosity values may correspond to depths where peridotites were recovered (Fig. F284A). High neutron porosity in this particular interval could be explained by the high content of bound water in the serpentine minerals (10 wt% H2O; see “Geochemistry”). In Hole U1309D, numerous neutron porosity maxima occur, especially in the intervals 100–290, 410–500, 550, and 1100–1150 mbsf, where the borehole shape is the most irregular (Figs. F284B, F284C). However, in some intervals, such as 311–343, 740–755, and 1090–1200 mbsf, where the borehole is in gauge, high neutron porosity values are related to the presence of serpentinized rocks with high contents of bound water in the serpentine minerals.

The capture cross section ranges from 11.8 to 40 capture units (cu). Capture cross section values are ~10–30 cu for most silicate minerals and up to 100 cu for Fe-Ti oxides (Serra, 1984). High values are recorded in diabase, oxide-rich gabbros, and serpentinized rock. However, sigma values can also be caused by large borehole conditions because the tool is sensitive to bad borehole wall contact.

Density and photoelectric factor

Density values range from 1.6 to 3.1 g/cm3 over the entire logged section in Hole U1309B and from 1.2 to 3.7 g/cm3 in Hole U1309D. In both holes, very low density values correspond to borehole washouts or fractures in igneous rocks and generally correlate with high values in the neutron porosity log. Density measured downhole shows excellent correlation with density values from discrete laboratory measurements (see “Physical properties”).

The photoelectric effect (PEFL), which varies from 1.1 to 5.2 b/e in Hole U1309B and from 0.4 to 9.4 b/e in Hole U1309D, is commonly a good indicator of lithologic variations. In Hole U1309B, PEFL values tend to be slightly higher in two intervals (29.4–34.7 and 61.7–69.8 mbsf) than in the rest of the section. These two intervals appear to correspond to the diabases crossed within the logged section. In Hole U1309D, PEFL maxima are recorded in several intervals (675, 880–886, 1182–1183, and 1301–1303 mbsf) corresponding to diabase and oxide-rich gabbros (Fig. F284A).

Electrical resistivity

In Hole U1309B, the electrical resistivity curves range between 3 and 400 Ω·m, most likely due to changes in fracture intensity and alteration. The lowest resistivity intervals are well correlated with the lowest density and highest neutron porosity values (Figs. F283, F284).

In Hole U1309D, the electrical resistivity values are locally extremely high (up to 40,000 Ω·m) between 1190 and 1195 mbsf, most likely in fresh olivine-rich troctolites. The bulk of the data varies between 30 and 3000 Ω·m with lithology (e.g., occurrence of oxides), fracture intensity, and alteration.

Lithological and structural interpretation

The logging data reflect the overall variability of the rock types drilled, comprising diverse gabbroic rocks, diabase, and olivine-rich troctolites (Fig. F284) (see “Igneous petrology”). Logging data also reflect structural changes and alteration modes. Based on density, velocity, PEFL, neutron porosity, and resistivity logs, we are able to distinguish characteristics that may correspond to lithology changes. The FMS images visualize the structural as well as partly textural variation downhole.

In the gabbroic rock intervals, logging bulk density varies between 2.8 and 3.2 g/cm3 and resistivity ranges from 50 to 2000 Ω·m. In general, the PEFL is <4 b/e and it averages ~3.1 b/e. Logged compressional velocity ranges between 5.5 and 6.5 km/s. Gabbros and olivine gabbros are characterized on the FMS images by high resistivities that correspond to light-colored zones (for example, the interval between 360 and 400 mbsf) (Fig. F288).

Most intervals of oxide gabbro, as identified in the visual core descriptions (see “Igneous petrology”), can be recognized in the logging data. They are generally characterized by elevated values of density (3.0–3.2 g/cm3), PEFL (4–8 b/e), and capture cross section (>30 cu) and low electrical resistivity (<100 Ω·m). PEFL is increased because of the presence of Fe in the formation. PEFL is proportional to the atomic radii of the atoms in the formation. Among the common elements, Fe has high PEFL (Serra, 1984). The intervals of oxide gabbro inferred from logs are intervals 191–196, 351–355, 642–645, 672–677, 719–720, 880–886, 989–998, 1182–1184, 1268–1274, and 1300–1302 mbsf. Although oxide gabbros have been described between 235 and 243 mbsf and 400 and 560 mbsf, there are no significant peaks in the PEFL log. The gabbros in these intervals contain disseminated oxides in lower concentrations (see “Igneous petrology”), and low PEFL values, ~4 b/e, match this observation.

Most olivine-rich rocks, such as olivine gabbro, troctolite, or olivine-rich troctolite, show high levels of serpentinization, and they contain more structurally bound H2O than olivine-poor gabbros (see “Igneous petrology” and “Geochemistry”). Based on this relation, we interpret intervals with neutron porosities of <5% as the least altered gabbro. In concert with low neutron porosity are high resistivities (>500 Ω·m). The slightly altered gabbros, on the other hand, show neutron porosities between 5% and 15% and intermediate resistivities (50–500 Ω·m). The olivine-rich troctolite at 689–691 mbsf is highly altered and has H2O contents of 8 wt% (see “Geochemistry”). For this section, neutron porosity averages 20%.

The recovered diabases have varying thicknesses. They are mostly very thin (<1 m) and show log responses reminiscent of oxide gabbros. The reason for that is that most diabases have a Fe2O3 content ranging between 11 and 13 wt%, which is comparable to that of oxide gabbro (see “Geochemistry”).

Structural features like discrete, open faults and fracture zones are portrayed by enlarged borehole diameter (>11 inches), which causes sudden apparent drops in density (1.5–2 g/cm3), resistivity (10–50 Ω·m), and velocity (4–5 km/s) and an increase in neutron porosity to values >40%. In addition, FMS and UBI images (Figs. F281, F290) show structural variations as well as textural variations of gabbroic rocks. FMS sections with patchy appearances correspond to intervals of coarse-grained gabbro (resistive patches) (Fig. F289A) or oxide-rich gabbro (conductive patches) (Fig. F290).

The continuous structural information gained from the FMS images with respect to dip and azimuth of conductive fractures is a crucial contribution to the understanding of the tectonic evolution of Atlantis Massif (see also “Structural geology”). On FMS and UBI images, mapping structural features consists of connecting a perfect plane (sinusoid) through the presumed geological object (mainly conductive features on FMS images or low-amplitude features on UBI images).

In Hole U1309B, 142 planar structures have been manually mapped (Fig. F291; “Supplementary material”). These features correspond to veins, fractures, faults, foliation, or lithologic boundaries. Dip magnitudes range between 5° and close to 85°, with most dips ranging from 20° to 60°. The number of subvertical planes (60°–90°) is much lower than subhorizontal (0°–30°) and intermediate (30°–60°) planes. This difference in detected plane abundance is linked to the borehole verticality, as the probability of encountering a horizontal fracture is higher than the probability of encountering a vertical one. Preliminary structural analyses of features (Fig. F291) show features dipping mostly southeast (mean dip direction = 130.4°). These orientations are consistent with orientations of a majority of veins measured on the core face and rotated into a geographical reference frame using paleomagnetic data (see “Reorientation of structure data using paleomagnetic data”). Within the interval 79.4–80 mbsf, FMS images show conductive rocks with a series of resistive clasts (white round features) that may be correlated to the breccia samples recovered in Section 304-U1309B-16R-1 (Fig. F292). At 58 mbsf, there is a change from a resistive to more conductive formation that can be correlated to the transition from gabbro to peridotite (Fig. F293). Below this contact, many conductive features inferred to be fractures are also present, although the overall resistive nature of the formation is still apparent below 62 mbsf.

Figure F294 shows an example of how consistent the dip of structures is in short depth intervals. Between 772 and 783 mbsf, the shallow structures in the gabbro dip toward the north, whereas the steep structures dip toward the south. The combination of FMS and UBI images is well suited to depiction of steep or complicated structures (Figs. F289B, F295).

Figure F296 compares conventional logging data to results from structural observations of brittle features (see “Structural geology”). The borehole enlargements seen in the caliper data are an expression of the local weakness in the formation caused by intense brittle deformation. Most borehole enlargements and the associated reduction in density, for example, correlate with peaks in the intensity of cataclastic features or large numbers of veins. Interestingly, the peaks in cataclastic and vein intensity at ~690 and ~745 mbsf show only slight decreases in wallrock density and no significant borehole enlargement. The fault zone at ~785 mbsf is not associated with any borehole enlargement. The lower two zones contain fault gouge that has been cemented by fluids (see “Structural geology” and “Metamorphic petrology”).

There is not only a good correlation of logging data with cataclasis and vein occurrence but also with alteration intensity; these are commonly genetically related. Here, it appears that alteration most strongly affects the neutron porosity. High intensities of alteration are associated with three fault zones between 690 and 785 mbsf (Fig. F297). The interval between 788 and 790 mbsf shows distinct logging responses (Fig. F298), indicating alteration. The FMS portrays it as a discrete layer with abrupt upper and lower boundaries and from which core has probably not been recovered. Overall, the most conductive intervals are well correlated with the oxide gabbro, troctolite, and serpentinized lithologic units defined in the core description (Fig. F299).

The interval between 729 and 760 mbsf shows a complex interaction of alteration and structural features. Between 729 and 735 mbsf and 740 and 750 mbsf, neutron porosities as high as 20% indicate intense alteration, probably chloritization (Fig. F300) (see “Metamorphic petrology”). Low values in resistivity (<100 Ω·m) and velocity (<5 km/s), as well as high values of natural gamma ray (~4 gAPI) and capture cross section (~30 cu), not only support this interpretation but indicate fluid flow in the past. In previously drilled ODP holes, an uptake in potassium with increased time of alteration and fluid flow has been observed (e.g., ODP Holes 504B, 395A, and 1137A). Below 765 mbsf, an increase in resistivity (>1000 Ω·m) and a drop in neutron porosity (0%–5%) occur, and values remain more or less constant down to 1092 mbsf. This agrees well with the observation of low core alteration in this interval. Between 1092 and 1170 mbsf and 1185 and 1195 mbsf, neutron porosity sharply increases locally to values >20% and resistivity decreases to <100 Ω·m. These intervals correspond to the olivine-rich troctolites, when highly serpentinized (see “Metamorphic petrology”).

The vertical magnetic field variations recorded by the GBM correlate well with magnetic susceptibility measurements on whole cores (Fig. F301). The differences in the magnetic behavior are observable in intervals containing altered olivine-rich troctolites. Between 300 and 350 mbsf, the magnitude of the vertical component strongly increases. Within this interval, the recorded PEFL remains low. Oxide gabbros are characterized by high magnetic susceptibility, minor to high variation in the magnetic field component, and an increase in PEFL. The magnitude of these anomalies may depend on the presence of disseminated or massive oxides. Gabbroic rock intervals between 400 and 600 mbsf are associated with more or less constant field intensity at ~32,000 nT. In the lower part of the interval logged with the GBM, there is an overall increase in the magnetic field strength that may be caused by changes in lithology. Below 680 mbsf, less olivine gabbro is present but there are more oxide gabbros.

During the uphole logging run, the maximum recorded temperature of the gyros was 71.09°C (743 mbsf), which is close to the instrument operating limit. A second temperature sensor in the telemetry module, which measures close to the ambient borehole temperature, had a maximum of 71.35°C at 808 mbsf. The downhole measurement on the GBM tool is slightly lower than the uphole measurement. This is probably because of time needed for the tool to reach equilibrium in a positive downward gradient.

Seismic stratigraphy

Site U1309 is located 350 m west of MCS Line Meg-4, being closest to common depth point 4136 (Fig. F302). The site is positioned on a slightly higher elevation (20–40 m) than the projected position on the seismic line. The time-depth conversion for Hole U1309D was calculated from a check shot survey. Velocity information was provided by the downhole sonic logging data and, in the upper 53 mbsf where the pipe in the hole prevented downhole measurements, by P-wave velocity measurements on core samples (see “Physical properties”). The resulting traveltime-to-depth curves are shown in Figure F303. Comparing the core-, log-, and check shot–derived relations, one can see an excellent agreement. Because of the high core recovery, the lithologies are adequately sampled (see “Physical properties”).

Average velocities computed from automatic picks on stacked seismograms generally fall in the range 5600–5800 m/s (5.6–5.8 km/s), but interval velocities can vary considerably between stations, particularly between 623 and 673 mbsf with 7971 and 5055 m/s, respectively (Table T25). Two errors might have occurred during check shot data acquisition. First, an error in station depth of a few meters cannot be ruled out. Second, the vertical movement of the air gun due to heave can contribute a small shift in time. However, stacking of several shots made at each station should have averaged the heave uncertainties. Adjusting the depths of the two stations (623 and 673 mbsf) by 2–4 m (which is probably beyond the potential station location error) would change these two interval velocity values; however, the high and low values (>7000 and <6000 m/s, respectively) would still be preserved at these stations. Despite our overriding caution with these initial results, it is possible that a lens of relatively unaltered, high-velocity, relatively olivine rich rock a few tens of meters thick occurs within the Fresnel zone of the vertical seismic profile (VSP) experiment (of the order of several meters).

A reversed polarity, zero-phase wavelet of 150 ms length was statistically extracted from the migrated seismic Line Meg-4 and used for convolution with the downhole impedance curve derived from logging P-wave velocity and density in order to calculate a synthetic seismogram (Fig. F32).

Hole U1309D transected a series of strong reflections seen in the seismic lines (Fig. F302). The lowermost reflection in seismic Line Meg-10 was identified as Reflector D by Canales et al. (2004). One interpretation put forward by those authors was that it coincided with the transition from shallower altered peridotite to underlying unaltered peridotites. In Line Meg-4, this reflector is at ~2.346 s two-way traveltime (TWT). Based on initial drilling results, several explanations for Reflector D are possible. This reflection could coincide with an apparent closure of fractures below ~295 mbsf and can be seen by the disappearance of low density and velocity peaks. Alternatively, density continues to increase monotonically until the base of the olivine-rich troctolite interval (310–360 mbsf), which may cause the reflector by its velocity contrast with underlying units. The lowermost reflection seen at 2.5 s TWT might be associated with the transition from a major fault zone, located at ~745–780 mbsf, to relatively unaltered gabbros below (see “Metamorphic petrology”). This inference is supported by an increase in both velocity and density logging data at 760 mbsf and a reduction in porosity as measured on core samples (see also Fig. F300).

Temperature profile

During the final logging run, the temperature of Hole U1309D was recorded using the TAP tool during the downhole and uphole runs, as well as the main log and the repeat log (Fig. F34). The borehole temperature measured uphole is slightly higher than downhole, and the second pass is higher than the first because the TAP tool sensor takes time to equilibrate. Logging curves show a slight change in the temperature gradient below 375 mbsf and at 720 and 1100 mbsf. These changes are recorded in each pass. The depth intervals coincide with changes in lithology (presence of olivine-rich troctolites) and/or structural events (fault zone). The maximum recorded temperature is 118.9°C at 1415 mbsf.