IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.320321.109.2010

Downhole measurements

Logging operations

Downhole logging of Hole U1337A started after APC/XCB coring to a total depth of 449.8 m DSF ended on 23 May 2009 at 0245 h (all times are U.S. Pacific Daylight Savings Time; UTC – 7 h). While drilling the last three cores, the hole was checked for fill at the bottom and no fill was detected. In preparation for logging, the hole was flushed with a ~50 bbl sweep of sepiolite/attapulgite mud (~9 ppg) followed by a wiper trip up to 80 m DSF and down to the bottom. No tight spots were encountered during the reaming, and no fill was noted at the bottom. A go-devil was then pumped through the drill string to open and lock the lockable flapper valve located above the bit. Finally, the hole was displaced with barite/attapulgite mud (~10 ppg) and the bit was raised to the final logging depth of 83.3 m DSF.

We deployed three downhole tool strings in Hole U1337A: (1) a modified triple combo that did not include a neutron porosity measurement, (2) an FMS-sonic combination, and (3) a VSI tool with a Scintillation Gamma Ray (SGT-N) sonde. For tool and measurement acronyms, see "Downhole measurements" in the "Methods" chapter.

The modified triple combo was lowered into the hole at 1410 h on 23 May, and after some testing and adjustment of the WHC the tool string, reached the bottom of the hole (4885 m wireline log depth below rig floor [WRF]). A first uphole logging pass started at 2020 h (Fig. F41). The tool string was lowered back to the bottom of the hole and a second uphole pass started at 2248 h. This second run ended when the tool string crossed the seafloor, marked by a peak in natural radioactivity clearly visible in the Hostile Environment Natural Gamma Ray Sonde (HNGS) measurement. The seafloor was detected at 4442.5 m WRF, which is significantly different from the drillers seafloor of 4472 m DRF. This discrepancy is mostly due to a wheel axle breaking in the depth measuring mechanism on the wireline winch. This failure occurred and was repaired while the tool was above the open hole interval; therefore, it does not affect the accuracy of the wireline depths relative to the seafloor. The modified triple combo reached the rig floor and was rigged down at 0421 h on 24 May.

The planned VSP at Site U1337 was shot during the daylight hours of 24 May in order to maintain a watch for the presence of marine mammals. The VSI tool was deployed in the second tool string together with a SGT-N tool to detect the seafloor. While the VSI/SGT-N combination was being lowered in the hole, the marine mammal watch started at 0645 h and the air guns began ramping up the shooting a half-hour later. The first VSI station was taken at 1010 h immediately above the base of the hole (4912.8 m WRF). We took as many as five shot records at each station for stacking and moved the tool string uphole by 15 m between stations. VSP operations were interrupted at 1213 h when a hydraulic hose failed on the WHC. Given that the ship's heave was no more than 1 m peak to trough and that a delay would have required cutting short the VSP experiment at sundown, we decided to continue acquiring data and to repair the WHC during the rig-up of the third tool string. We acquired 16 successful VSP stations up to 4688.1 m WRF; three shallower stations were attempted, but the VSI tool could not be coupled to the borehole wall because of an enlarged hole and the soft formation. At 1515 h we decided to end VSP operations and started logging with the SGT-N tool, detecting the seafloor at 4474 m WRF. Rig-down was completed at 1820 h.

The third and final wireline tool string deployed in Hole U1337A was the FMS-sonic combination (Fig. F42). The FMS-sonic tool string was lowered in the hole at 2005 h, and after some testing of the WHC it reached the bottom of the hole at 4914.3 m WRF. The first uphole logging pass started at 0125 h on 25 May and ended at 0235 h. After returning to the bottom of the hole, the second pass started at 0310 h and ended after reaching the seafloor at 4467 m WRF. Rig-down was completed at 0838 h.

Downhole log data quality

Figures F4 and F43 show a summary of the downhole log data acquired in Hole U1337A. These data were processed to convert to depth below seafloor and to match depths between different logging runs, resulting in the WMSF depth scale (see "Downhole measurements" in the "Methods" chapter).

The overall quality of the logging data can be assessed from the repeatability of measurements acquired in different passes and by comparing downhole log data to core measurements. In general, downhole log data acquired in Hole U1337A show excellent repeatability between passes (Figs. F4, F43). A key factor that influences downhole log data quality is the size and irregularity of the borehole, especially for measurements that require a good contact with the borehole wall (e.g., density measured by the Hostile Environment Litho-Density Sonde [HLDS] and resistivity images obtained by the FMS tool). The "hole diameter" track in Figure F4 is measured by a caliper arm on the HNGS and shows a hole generally well above the bit diameter (11.4375 inches), with values greater than the maximum measurable range (~18 inches) above 210 m WMSF. Despite the occasionally large hole, HLDS density data are of high quality, as shown by their close correlation with MAD core measurements on discrete core samples and with the electrical resistivity logs, which are relatively insensitive to hole size. There is only one interval where the densities logged in the two passes differ significantly, between 110 and 140 m WMSF (Fig. F4). This difference is most likely due to poor contact with the formation in the second logging pass, which resulted in anomalously low density measurements; the higher densities measured in the first pass are the more reliable values.

Resistivities obtained by the electrode spherically focused resistivity (SFLU) measurement were lower than those obtained in induction measurements (e.g., medium induction phasor-processed resistivity in Fig. F4), probably because of current loss at the electrodes. The higher induction resistivities are closer to values typically measured in deep-sea sediments, and we applied a simple empirical correction to the SFLU data by multiplying them by a constant factor of 1.6. This correction brings the overall measured resistivity values in close agreement.

P-wave velocities measured by the Dipole Sonic Imager (DSI) in the two logging passes closely agree in the lower part of Hole U1337A, below ~230 m WMSF (Fig. F4). Above this depth, the processing of the sonic log waveforms was unable to distinguish the formation velocities from those of the fluid circulating in the hole (~1500 m/s). Core measurements (see "Physical properties") and results of the VSP experiment (see below) also indicate velocities near or just above 1500 m/s in the top 200 m of the hole.

Figure F43 compares spectral gamma ray logs acquired by the HNGS in the triple combo and FMS-sonic tool strings. The gamma ray measurement is highly attenuated when the tool is inside the bottom hole assembly and the drill pipe (above 77 m WMSF in Hole U1337A), and data in this interval should only be used qualitatively. The total gamma ray log is very similar in the two passes illustrated in Figure F43; in the spectral data, some features are repeatable (e.g., the uranium peaks around 240 m WMSF and at the seafloor), whereas others are less so (e.g., in the thorium log). Spectral gamma ray features that are not repeatable in different passes may not be reliable.

In Hole U1337A, we also acquired FMS electrical resistivity images. The quality of these measurements depends on close contact between the measuring pads on the tool and the borehole wall. The FMS borehole images are of high quality between ~200 m WMSF and the base of the logged interval, where they accurately reproduce sediment layers. Above 200 m WMSF, the hole is significantly enlarged and the images are marred by low-resistivity swaths, which are likely caused by poor contact between the tool pads and the borehole wall.

Finally, hole diameters measured by the HLDS caliper arm (Fig. F4) and by the FMS tool (not shown) were not consistent, with the HLDS caliper, measuring a 3–4 inch larger diameter. The tool calipers were recalibrated after logging Hole U1337A; although the recalibration reduced the differences in the hole diameters, the two measurements are still not the same. This difference may be due to the different shape of the pads at the end of the caliper arms. The larger FMS pads may smooth out small-scale irregularities in the borehole wall and thus measure a smaller hole diameter.

Logging units

Downhole log measurements of bulk density, electrical resistivity, and P-wave velocity in Hole U1337A correlate very well (Fig. F4). The likely reason is that variations in sediment composition result in variations of porosity, and changes in porosity affect similarly bulk density, resistivity, and P-wave velocity. High porosities obviously result in low bulk densities. Also, resistivity variations in water-saturated sediments are controlled by variations in their content of pore water, which is the component of bulk sediment that conducts electricity. Finally, laboratory measurements and rock-physics models show that P-wave velocities in sediments decrease with increasing porosity. Details on these fundamental relationships between porosity and logged properties are given by Hearst and Nelson (1985) and Ellis and Singer (2007).

The downhole logs of gamma ray radioactivity (Fig. F43) are controlled by the sediment content of naturally occurring radioactive elements (potassium, uranium, and thorium) and do not closely depend on porosity. The most significant features in the gamma ray logs are the peaks in uranium content at the seafloor and 240 m WMSF. The seafloor peak is notably large, as it is attenuated because the tool is measuring through the drill pipe (see above) and it matches a peak in natural gamma ray counts made on the cores immediately below the seafloor (see "Physical properties"). The uranium peak at 240 m WMSF corresponds to a chert layer that was only recovered in fragments and that is clearly seen in the FMS images as a ~40 cm thick high-resistivity layer (see next section and Fig. F44).

We divided the logged section into three units on the basis of the overall variation of density, resistivity, and P-wave velocity. Logging Unit 1 (from the base of the drill pipe at 77 m WMSF to 212 m WMSF) has low densities and resistivities that are variable but show no trend with depth and mostly corresponds to lithologic Unit II (see "Lithostratigraphy"). The transition between logging Units 1 and 2 is marked by a distinct increase in logged densities and resistivities and corresponds to the transition between lithologic Units II and III. Whereas density and resistivity do not show a trend with depth in logging Unit 2, Unit 3 (from 339 m WMSF to the base of the logged interval) displays a clear increase with depth of density, resistivity, and P-wave velocity. This increase is probably due to lithification of the pelagic carbonate oozes in lithologic Unit III, which turn in to chalk at the base of the sediment column of Site U1337 (see "Lithostratigraphy").

Core-log correlation

By comparing the continuous wireline logging data to measurements taken on the core we can start piecing together a more complete recovery story and identify where there may be gaps in the core record for Site U1337. Additionally, the logs provide in situ physical property measurement of formations downhole. Here we make a first attempt to correlate core photographs and GRA density data from the core with the continuous downhole logging density measurements (Fig. F44). We focus on an interval of interest where core recovery was particularly poor to determine the amount of core missing between 230 and 250 m WMSF. At 240 m WMSF, a layer showing very high resistivity is observed in the FMS images (Fig. F44C). Above and below this band are contrastingly conductive layers. The resistive layer corresponds to a 40 cm thick chert, which is only recovered in the core as rubble, whereas the enveloping conductive layers are two diatom mat units (160 cm thick above the chert layer and 60 cm thick below). The presence of these two markedly different formations in the FMS images is complemented by the wireline HLDS density measurement, which shows a decrease in density to ~1.4 g/cm3 where the conductive layers (diatom mats) are observed and a peak in density at ~1.8 g/cm3 at the chert unit (Fig. F44B). In contrast, the HNGS (spectral gamma ray) logging data shows a peak of ~15 gAPI at the chert unit (previously discussed) and very little variability either side (Fig. F44A).

We used the location of the top and base of the diatom mats surrounding the chert to line up the core photographs and their associated GRA density measurements with the FMS images. It must be noted that none of the core image/measurements have been distorted from the original to better fit the wireline data; therefore, the fit of the logging and core data sets is by no means perfect. We did not consider changes in original in situ formation thickness caused by expansion and contraction of the core following recovery. In the correlation process we matched broad patterns and changes in density curves rather than small-scale variations.

The core GRA density data fit fairly well to the HLDS density above the chert layer and show intermediate- to high-density values dropping to much lower values in the diatom mat unit (Fig. F44D–F44G). In Hole U1337A, there is a large gap in core recovery below 240 m WMSF, beneath the chert layer; Hole U1337B was terminated at this depth because of a stuck core barrel. However, some remnants of the chert in the form of chert rubble are seen in Cores 321-U1337C-11X and 321-U1337D-27X. The chert layer has likely been fragmented by the drilling process, resulting in disturbed core sections and chert rubble forced further downsection. The chert rubble has elevated the GRA density values taken in the upper part of Core 321-U1337D-27X, so that the data are no longer representative of the formation at this level (Fig. F44G). Additionally, as previously mentioned, diatom mats have been used as "line-up" markers, but in Hole U1337D much lighter colored sediment (not visible in Fig. F44) occurs above the identified diatom mat in Core 321-U1337D-28H, suggesting that the core should perhaps reside lower on the depth scale (see gray dashed lines in Fig. F44). However, even if Core 321-U1337D-28H were moved, there is only ~1 m of section missing from Site U1337. The best fit of the GRA density measurements to the logged density below the chert layer is seen in Core 321-U1337C-12X, in which the close correspondence of the curves gives some confidence on the core location on the WMSF depth scale. Additionally, in the CSF depth scale the base of Core 321-U1337C-11X is at 238.96 m CSF and the top of Core 321-U1337C-12X is at 240.50 m CSF, 1.54 m apart. This is similar to the distance between the images of these cores in our correlation (Fig. F44F). More detailed postexpedition work will refine these core-log correlations.

Vertical seismic profile

Figure F45A shows the stacked waveforms measured at 16 stations (214.1–438.4 m WSF) by the vertical direction geophone in the VSI. The waveforms show two clear events: a direct arrival, where the arrival times (red dots) increase with increasing receiver depths, and a later event with smaller amplitude and opposite polarity that instead arrives earlier at deeper receivers. The latter event is a strong reflection from the top of the basaltic oceanic basement. The basement reflection converges with the first arrival at the base of Hole U1337A, which reaches the top of the basalt.

Table T26 lists the values of the measured and corrected one-way arrival times from VSP in Hole U1337A. The measured traveltimes are the differences between the arrival of the acoustic pulse at a hydrophone located immediately below the air gun source and the arrival at the borehole receiver. The corrected traveltimes are the traveltimes from the sea surface to the borehole receiver and account for the depth of the air guns (7 mbsl) and for the depth of the hydrophone below the air guns (2 m). More details on the VSP measurement procedure are in "Downhole measurements" in the "Methods" chapter.

Figure F45B shows the relationship between depth below seafloor in Hole U1337A and the TWT, which is the traveltime to reflectors in surface seismic sections. To construct this relationship, we start from the TWT from sea level to the seafloor (5.9444 s), computed from the uncorrected seafloor depth measured by the ship's echo sounder (4458.3 m). The difference between this time and the arrival time to the shallowest receiver in the VSP gives an average P-wave velocity of 1514 m/s between the seafloor and 214.1 m WSF. At first approximation, the traveltime-depth relationship in Figure F45 assumes a constant velocity in the interval 0–214.1 m WSF. The dependence of the arrival times on depth in the VSP receiver array can be fit very closely by a second degree polynomial:

t(z) = 5.9414 + (1.4378 × 10–3) z – (4.8984 × 10–7)z2,

where t is TWT (s) and z is depth below seafloor (214.1 ≤ z ≤ 438.8 m WSF). The maximum residual on this fit is 0.45 ms and the root-mean-square residual 0.23 ms.

The variation of P-wave velocity in the depth interval spanned by the VSP receiver array can be obtained from the derivative of the time-depth relationship above, which gives

V(z) = [(0.7189 × 10–3) – (4.8984 × 10–7) z]–1,

where V is P-wave velocity (m/s) and z is depth (214.1 ≤ z ≤ 438.8 m WSF). This relationship is compared to the velocities measured by acoustic logging with the DSI in Figure F4. The acoustic log measures small-scale details in the velocity structure that cannot be resolved by the VSP arrival time data, yet both estimates of velocity give the same trend of increase in velocity with depth.

The traveltime-depth relationship allows for correlating stratigraphic events in Site U1337 to reflections in the surface seismic data. In Figure F46, we correlate the downhole logging results in Hole U1337A to seismic reflection Line 4 of the AMAT-03 site survey in the proposed Site PEAT-7C area, which crosses the location of Site U1337 at Shotpoint 1756. To make an accurate correlation, the TWTs in the seismic reflection line were shifted so that the seafloor reflection time (originally at 5.989 s) matched the more accurate TWT to the seafloor measured by the JOIDES Resolution echo sounder (5.9444 s; see above).

Reflections in the seismic line correspond to fluctuations in bulk density and velocity; for example, the increase in density between logging Units 1 and 2 at 212 m WMSF (corresponding to the boundary between lithologic Units II and III) matches a cluster of reflection events at 6.2 s TWT. This preliminary well to seismic correlation demonstrates that the Site U1337 results can be used to construct an up to date, age-calibrated interpretation of seismic stratigraphy and Neogene sedimentation history in the equatorial Pacific (e.g., Mayer et al., 1985; Mitchell et al., 2003).

Heat flow

Downhole temperature measurements at Site U1337 included eight APCT-3 measurements in Holes U1337A and U1337B and a SET measurement in Hole U1337C (Table T27). Measured temperatures ranged from 3.14°C at 43.5 m DSF to 11.24°C at 298.1 m DSF and closely fit a linear geothermal gradient of 32.4°C/km (Fig. F47). Temperature at the seafloor was 1.634°C, based on the average of the measurements at the mudline in the nine temperature profiles. The thermal conductivity under in situ conditions was estimated from the laboratory-determined thermal conductivity using the method of Hyndman et al. (1974). The estimated in situ thermal conductivities in Figure F47 are as much as 2.3% below the measured laboratory values.

A simple estimate of the heat flow can be obtained from the product of the geothermal gradient times the average in situ thermal conductivity (0.89 W/[m·K]), which gives a value of 29.1 mW/m2. However, thermal conductivity increases with depth, hence a more accurate heat flow value can be obtained from the slope of the temperature measurements plotted versus thermal resistance as in the Bullard method (see Pribnow et al., 2000). The variation of thermal conductivity with depth was estimated by fitting a linear trend, and thermal resistance was calculated as the integral in depth of the inverse thermal conductivity (Pribnow et al., 2000). The fit between temperature and thermal resistance gives a slightly lower heat flow of 28.4 mW/m2, which is similar to values of nearby sites in the global heat flow database (Pollack et al., 1993).