Previous   |    Next

doi:10.2204/iodp.proc.339.107.2013

Downhole measurements

Logging operations

Two holes were logged at this site, Holes U1389A and U1389E. Hole U1389A was cored with the APC/XCB and had good conditions for logging. Logging the first hole at the site also enabled prediction of lithology of the unrecovered intervals prior to coring Holes U1389B–U1389E. Hole U1389E was cored with the RCB, resulting in a wide borehole and poor conditions for logging at this site.

Hole U1389A

Downhole logging measurements in Hole U1389A were made after completion of XCB coring to a total depth of 354.9 m drilling depth below seafloor (DSF). In preparation for logging, the hole was flushed with a 30 bbl sweep of high-viscosity mud. The hole was displaced with 112 bbl of barite-weighted mud (10.5 ppg), and the pipe was pulled up to 85.7 m DSF. Three tool strings were deployed in Hole U1389A, the triple combo, FMS-sonic, and VSI tool strings (Fig. F38; see also “Downhole measurements” and Table T6 in the “Methods” chapter [Expedition 339 Scientists, 2013b] for definitions of tool acronyms). The Accelerator Porosity Sonde (APS) was left out of the triple combo tool string because the APS is not designed for high-porosity formations and it often overestimates porosity in wide and rugose boreholes, both of which were expected in Hole U1389A based on the previous two logged sites (see Fig. F42 in the “Site U1387” chapter [Expedition 339 Scientists, 2013d]). This also allowed the Hostile Environment Natural Gamma Ray Sonde to be placed at the bottom of the tool string to record intervals with low core recovery at the base of the hole (Fig. F39; see also “Lithostratigraphy”).

At 2240 h on 23 December 2011, the triple combo tool string (resistivity, density, and NGR tools) descended from the rig floor into the pipe (Fig. F40). The downlog proceeded at 1000 m/h and reached the base of the hole at 355 mbsf. A short uplog was run from 355 to 280 mbsf to provide data to cross-check with the main pass, which was logged up from 355 mbsf to seafloor at 275 m/h.

Rig up of the FMS-sonic tool string started at 0330 h on 24 December. A downlog was taken at 730 m/h to the base of the hole. Standard (high)-frequency Dipole Sonic Imager transmitter settings were used, with P-wave window settings of 130–190 µs/ft. Two uphole passes of the FMS-sonic were run, Pass 1 to the base of the pipe and Pass 2 to the seafloor, at 550 m/h. Rig down was completed at 1100 h.

Marine mammal watch for the VSP started at ~1100 h. The Sercel G. GUN parallel cluster (two 250 in³ air guns separated by 1 m) was ramped up in a soft-start procedure. The air gun cluster was positioned on the port side of the JOIDES Resolution at a water depth of ~7 mbsl with a borehole offset of ~45 m. The VSI tool string started its descent down Hole U1389A at 1145 h. Before reaching the bottom of the hole, a short uplog was taken from 269 to 219 mbsf to depth-match the VSI gamma curve to the other logging runs. The survey started at 1300 h, and 10 out of 11 stations recorded good sonic waveforms. Stations were spaced at roughly 25 m intervals, where the borehole was not washed out (Fig. F38). The survey finished at 1500 h, and the tool was brought uphole.

Hole U1389E

Downhole logging measurements in Hole U1389E were made after completion of XCB coring to a total depth of 989.9 m DSF. In preparation for logging, the hole was flushed with a 50 bbl sweep of high-viscosity mud and the bit was released at the bottom of the hole. The hole was displaced with 375 bbl of barite-weighted mud (10.5 ppg), and the pipe was pulled up to a logging depth of 102.2 mbsf. Two tool strings were deployed in Hole U1389E, the triple combo and the FMS-sonic tool strings (Fig. F38; see also “Downhole measurements” and Table T6 in the “Methods” chapter [Expedition 339 Scientists, 2013b] for definitions of tool acronyms). The triple combo tool string included the resistivity, density, porosity, and NGR tools. The High-Resolution Laterolog Array (HRLA) was placed at the top of the tool string in order to obtain density and NGR data closer to the bottom of the hole.

The triple combo tool string started downhole at 0150 h on 2 January 2012 (Fig. F40). A bridge at 525 mbsf was passed, but a second bridge at 567 mbsf blocked further downhole progress; therefore, the bottom 423 m of Hole U1389E remained unlogged. Rig up of the FMS-sonic tool string started at 0730 h, and the first pass reached the same bridge that the triple combo tool string encountered (567 mbsf). The second pass was blocked by a bridge at 525 mbsf. A VSP had been planned for this hole but could not be run because of concerns that air gun shots could weaken the structural integrity of the JOIDES Resolution’s sea chest (salt-water intake) that had been thinned by corrosion.

Differences in the seafloor depth given by the step in the NGR logs are partly accounted for by changes in tide level (Fig. F40). The triple combo tool string downlog and main pass found seafloor at 656 mbrf, and FMS-sonic Passes 1 and 2 found it at 654 mbrf, whereas the driller’s mudline was at 656.2 mbrf. Tide level was high for triple combo logging (+0.5 to +1.4 m), low for FMS-sonic logging (–1.4 to –1.0 m), and +0.1 m for the driller’s mudline. Seas were calm (maximum peak-to-peak heave = 0.7 m), giving little contribution to the offset. The balance of the offset probably comes from the wireline heave compensator being centered slightly differently for each seafloor depth determination. However, a good depth match of the open-hole NGR logs between logging runs with the core data was achieved using a seafloor depth of 656 mbrf for the triple combo main pass (Fig. F41).

In Hole U1389E, the triple combo main pass found seafloor at 654 mbrf and FMS-sonic Pass 2 found it at 653 mbrf (no mudline was taken in this RCB hole). The tide was rising for the triple combo run (–0.5 to +0.1 m) and falling for the FMS-sonic run (+0.4 to –0.4 m).

Log data quality

In Hole U1389A, the quality of the log data is good because the borehole diameter (10–11 inches) was close to the bit size. Few washouts were observed and they generally did not exceed 14 inches (Fig. F39). The widest intervals, exceeding 16 inches, were observed at ~106 and ~190 mbsf. No bridges were found in Hole U1389A. Resistivity, sonic velocity, and NGR logs are generally robust to the moderate changes in hole diameter. The density log correlates well with measurements on cores, although there is some reduction in values in the larger washouts. The photoelectric effect log gave anomalously high values, >10 b/e^–, because of barium in the logging mud. The FMS resistivity images were of good quality because of good contact with the borehole wall in most of the borehole.

In Hole U1389E, log data quality was affected by the large borehole diameter, exceeding 14 inches in 67% of the logged interval, and exceeding the 18 inch limit of the Hostile Environment Litho-Density Sonde (HLDS) caliper arm in 25% of the interval (Fig. F42). Some caves were found (e.g., 222–234, 264–269, and 297–309 mbsf), with a large one from 511 to 517 mbsf associated with very low values in all the logs. The large number of washouts indicates that this RCB hole was in significantly worse condition than the same interval (102–355 mbsf) in XCB Hole U1389A. A large amount of seawater circulation was required to keep the RCB bit clear of clay build-up and to prevent sand from building up around the drill pipe, often resulting in washout of the sand-rich sediment. Resistivity, sonic velocity, and NGR logs were generally robust to changes in hole diameter, although some reduction in values is evident in the larger washouts. The photoelectric effect log gave reasonable values (<5 b/e^–) where good contact was made with the borehole wall; the 375 bbl of barite-weighted mud only filled the lowermost 300 m of the hole, assuming an in-gauge hole. The FMS resistivity images were poor because of bad contact with the borehole wall in most of the borehole.

Comparison of log data quality in the contrasting borehole conditions of Holes U1389A and U1389E

Log data in Holes U1389A and U1389E cover a common interval between 100 and 350 mbsf that permits an assessment of log data quality, particularly in wide boreholes (Fig. F45). Hole U1389A was cored by APC/XCB, and U1389E was cored by RCB. Both coring systems used 9.875 inch diameter bits, but the resulting hole diameter was much wider in Hole U1389E than in Hole U1389A. Differences between the two holes include the number of water jets on the bits, six for the XCB (a Russian polycrystalline diamond compact bit) and four for the RCB (CC-4 model); the absence of roller cones on the XCB bit; the age of the hole, less than 2 days for the XCB hole and more than 4 days for the RCB hole in the common interval; and water circulation, which had to be maintained more vigorously in the deeper RCB hole to prevent the pipe from getting stuck.

Caliper data from the HLDS show that most of Hole U1389A was between 10 and 11 inches in diameter, not much wider than bit size (Fig. F45). In contrast, Hole U1389E was almost always >12 inches wide and was often wider than the maximum extent of the caliper arm, 18 inches. An indication that the hole reaches >25 inches wide (at 305–310 mbsf) is given by the “diameter of invasion” log, which is based on the separation of the deep- and shallow-reading HRLA resistivity logs. Hole U1389E is also “rougher” (more variable in diameter over short scales) and has some bridges (hole narrower than bit size), for example at 319 mbsf. The effect of these contrasts in hole diameter on the other logs is discussed below.

The HSGR logs have lower values in the wider hole, but in general the same trends and patterns are seen in the HSGR logs from both holes. Correction for hole diameter is possible based on the caliper logs. However, in rough holes like Hole U1389E, any small depth discrepancy between the gamma and caliper logs would lead to further errors being introduced into the log, and it is not possible to accurately correct for hole sizes greater than the 18 inch caliper limit.

The bulk density logs offer a stark contrast between the two holes. In the wider parts of Hole U1389E, the sensors were unable to make good contact with the borehole wall and consequently read values as low as 1.15 g/cm³, close to borehole fluid. In Hole U1389A, bulk density values typically match the MAD bulk density values. Where the holes are relatively smooth and not too wide, the log bulk density values overlay very closely (e.g., 195–215 mbsf). Even in Hole U1389A, some of the departures to lower density values correspond to small washouts and may be suspect (e.g., 190 mbsf).

Electrical resistivity is measured by the HRLA at five different depths of investigation. The deepest measurement (RLA5) is least affected by variations in borehole diameter. In general, the RLA5 logs from both holes overlay each other reasonably well, but the Hole U1389E log is noisier and drops to lower values in thin, washed out zones, as well as peaking to higher values where the hole narrows. The HRLA curves also have additional noise close to the metal drill pipe (100–140 mbsf). An additional HRLA log, the “true resistivity” computed from the five deep-to-shallow reading measurements (not shown in the figure), apparently overcorrects, providing resistivity values in Hole U1389E that are higher than in Hole U1389A.

The sonic velocity logs from the two holes give well-matched patterns and values in the upper part of the common section (100–220 mbsf). In the washed-out zones below 220 mbsf in Hole U1389E, velocity is sometimes underestimated, perhaps caused by a longer path for the sonic wave to travel. In the rough and wide 220–320 mbsf section, Hole U1389E velocities are on average 3.2% slower than those in Hole U1389A. Even though the velocity data repeat very well between Passes 1 and 2 in Hole U1389E, this is not necessarily a good indication that they are recording the true formation velocity well, rather that the two passes respond similarly to the hole conditions. Another note of caution is that the washed out zones tend to be sand-rich layers; thus their velocities would be underestimated more than those of other lithologies.

Logging units

The logged intervals at Site U1389 were assigned to one logging unit because no major base-level changes in the log data were discerned (Figs. F39, F42). At the scale of this unit, the NGR signal typically ranges from 20 to 60 gAPI, with few peaks above 70 gAPI below 400 mbsf. The signal is dominated by the radioactivity of potassium and thorium, with uranium contributing a relatively minor component (<2.5 ppm) (Figs. F43, F44). Similar to the previous sites, potassium and thorium curves are closely correlated. As both of these elements are found in clay minerals, which are abundant (see “Lithostratigraphy”), the NGR signal is primarily tracking clay. Minerals like K-feldspar and mica also contribute to the NGR signal. In Hole U1389A, the density and sonic velocity logs increase downhole (Fig. F39), reflecting sediment compaction with depth.

Logging Unit 1 was divided into two subunits, mainly on the basis of changes in mineralogy inferred from the NGR logs and a change in character of the resistivity logs.

Logging Subunit 1A: base of drill pipe to 318 mbsf

Logging Subunit 1A is characterized by medium-amplitude cycles in bulk density, NGR (and its uranium, thorium, and potassium components), resistivity, density, and sonic velocity values. Several orders of cycles are observed, varying from one to several tens of meters in thickness (Fig. F43). The potassium and thorium concentrations co-vary closely, suggesting that clay content controls these elements. The uranium concentrations correlate with the potassium and thorium concentrations at some depths but vary independently at others. As expected from downhole compaction, the resistivity, density, and sonic velocity logs have an increasing downhole trend and are generally well correlated.

Logging Subunit 1A at Site U1389 strongly resembles logging Subunits 1A observed in Holes U1386C and U1387C (see “Downhole measurements” in the “Site U1387” chapter [Expedition 339 Scientists, 2013d]). All of them are Pleistocene in age, although their lower boundaries are diachronous (see “Biostratigraphy” and “Lithostratigraphy” in the “Site U1386” chapter [Expedition 339 Scientists, 2013c]).

Logging Subunit 1B: 318–568 mbsf

Logging Subunit 1B is distinguished from the subunit above by higher amplitude variability in the NGR logs (Figs. F43, F44), which have a succession of sometimes thin peaks. The potassium baseline (~0.6%) is similar to that in Subunit 1A, whereas thorium levels decrease to 2 ppm. This suggests that for Subunit 1B, fewer thorium-bearing minerals are present, and there is perhaps a dominance of low-thorium clays such as illite, which has been used as an indicator of terrestrial rather than marine origin (Rider, 1996). Uranium appears to be closely correlated with the potassium and thorium curves.

Subunit 1B is also characterized by a decrease in resistivity values and a disappearance of higher frequency variability compared to the subunit above (Fig. F42). This decrease is most likely caused by a major increase in interstitial water salinity: sodium and chloride concentrations start to increase downhole from 263.4 mbsf, and reach a brinelike concentration maxima at 533 mbsf (Fig. F33; see also “Geochemistry”). An increase in the amount of ions in the pore water solution lowers the apparent formation resistivity. The sharp decrease in sodium and chloride concentrations from 533 to 757.4 mbsf has not been clearly recorded in the resistivity logs because the deepest resistivity data are at 541 mbsf.

In Subunit 1B, the NGR logs are sometimes well correlated with the caliper log (e.g., 360, 390, and 485–495 mbsf) (Fig. F44), in contrast to the more common anticorrelation between these logs observed for most of the logged intervals on this expedition.

The top of Subunit 1B is equivalent to lithostratigraphic Subunit IB, recognized in cores from 320 to 385 mbsf as dominated by calcareous mud lithologies (see “Lithostratigraphy”). However, <50% of this subunit was recovered in cores, and log data give information on the unrecovered material. In this interval, especially from 318 to 360 mbsf, the NGR logs suggest the presence of meter- to several meter-thick sandy intervals interbedded with clay-rich intervals. Hole U1389A is in-gauge (close to bit diameter) in this interval, so there is good confidence in the log data. Thus, the muddy intervals seem to have been preferentially recovered in cores.

Formation MicroScanner images

Because of the good borehole conditions in Hole U1389A, FMS resistivity images reveal numerous resistive and conductive intervals whose thicknesses range between several centimeters and a few meters. Gradual and sharp transitions between alternations are observed. In Hole U1389A, the image quality is generally poor because of poor contact between the sensors and the borehole wall, although some intervals still provide reasonably good images. In some places, most of these alternations correlate very well with the density, NGR, and sonic curves (Fig. F46). Conductive beds on the FMS images correlate with lower values in the NGR log, lower bulk densities, lower resistivities, lower P-wave velocities (e.g., two conductive intervals from 317 to 320 and 326 to 328.5 mbsf in Fig. F46) and correspond to coarse-grained intervals (sands and silts). These intervals contrast with muddy intervals appearing as less conductive on the FMS images and correlating with higher values in the NGR log, higher bulk densities, higher resistivities, and higher P-wave velocities (e.g., two conductive intervals from 320 to 323 and 325 to 326 mbsf in Fig. F46). The good contrast between sandy and muddy intervals on the FMS images is interpreted as resulting from the presence of brine in the pore spaces, which lead to high conductivity in the more permeable layers (sand). At the scale of the boreholes, the FMS images have a downhole trend of increasing conductivity, especially below 320 mbsf, that we relate to the progressive downhole increase in pore water salinity (see “Geochemistry”).

Vertical seismic profile and sonic velocity

Data provided by VSP in Hole U1389A and sonic velocity logs in Holes U1389A and U1389E help establish conversions between depth and two-way traveltime (TWT) and link the borehole logs to the seismic section.

Good stacks of sonic waveforms were obtained during the VSP survey. It was easier to clamp the geophone to the borehole wall in Hole U1389A than in the much wider holes in the two previous VSP surveys. All 11 of the stations yielded good check shot traveltimes, ranging from 0.990 s TWT at 110.0 mbsf to 1.254 s TWT at the deepest station at 355.3 mbsf (Fig. F47; Table T25). The original traveltimes were corrected to the sea level datum and are based on the first break of the sonic waveform.

Sonic traveltimes were also calculated from sonic log velocities and give TWTs very close to those from the VSP, assuming a velocity of 1.62 km/s just below the seafloor. Sonic velocity logs were taken in Holes U1389A (88–337 mbsf) and U1389E (102–550 mbsf), and velocities increase downhole because of compaction. The TWT at 550 mbsf is 1.460 s, based on the sonic data.

Heat flow

Nine APCT-3 downhole temperature measurements in Holes U1389A, U1389C, and U1389D, ranged from 13.7°C at 19.5 mbsf to 15.29°C at 89.6 mbsf (Table T26), giving a geothermal gradient of 20.9°C/km (Fig. F48). The measurements increase linearly with depth, and the trend line intersects the seafloor at 13.36°C. Some scatter is present, as well as minor differences between the three holes. These differences are probably caused by (1) nonunique fits to the APCT-3 temperature equilibration curves to obtain the formation temperature and (2) lateral heterogeneity in heat flow leading to small temperature differences between the holes. Similar to Site U1386, the bottom water temperature was difficult to determine accurately from the APCT-3 temperature profiles. The intersection of the linear temperature trend with the seafloor probably gives the best estimate here.

Thermal conductivity under in situ conditions was estimated from laboratory-determined thermal conductivity using the method of Hyndman et al. (1974) (see “Physical properties” in the “Methods” chapter [Expedition 339 Scientists, 2013b]). The calculated in situ values are within 0.1% of the measured laboratory values. Thermal resistance was then calculated by integrating the inverse of the in situ thermal conductivity over depth (Fig. F48). A heat flow of 26.6 mW/m² was obtained from the linear fit between temperature and thermal resistance (Pribnow et al., 2000). This value is quite low for the Gulf of Cádiz area (Grevemeyer et al., 2009).

Top of page   |    Previous   |    Next