Previous   |   Next

doi:10.2204/iodp.proc.311.105.2006

Downhole logging

Logging while drilling

Operations

Hole U1327A was spudded at 1333 mbrf water depth (drillers depth) at 1830 h on 23 September 2005. Hole U1327A was drilled after finishing Hole U1326A by pulling the drill string clear of the seafloor and moving the ship in dynamic positioning mode. LWD tools included the GeoVISION resistivity tool, the EcoScope tool, the SonicVISION tool, the TeleScope MWD tool, the ProVISION nuclear magnetic resonance tool, and the ADNVISION tool. For details on each tool and the measurements it makes, see "Downhole logging" in the "Methods" chapter.

Similar to the other sites, the uppermost 10 m of the hole was drilled with a rotation rate of 10–15 rpm, a pump flow rate of 190 gpm, and an ROP of 10–15 m/h. We then increased these rates to 40 rpm and 220 gpm until 30 mbsf, and then again to 60 rpm and 270 gpm (to start the MWD telemetry), keeping the instantaneous ROP below 50 m/h with depth. The target depth of 300 mbsf (1633 mbrf) was reached at 0810 h on 24 September 24. Considering the proximity between Holes U1327A and U1328A (1.9 mi apart), it was decided to raise the drill string above the seafloor and move by dynamic positioning to the location of Hole U1328A without tripping back to the surface as originally planned.

Gas monitoring with real-time logging-while-drilling/ measurement-while-drilling data

LWD logs were acquired to plan coring and pressure coring operations in subsequent holes at Site U1327. As Hole U1327A was drilled without coring, the LWD data had to be monitored for safety to detect gas entering the borehole. As explained in "Downhole logging" in the "Methods" chapter, the primary measurement we used for gas monitoring was annular pressure while drilling measured by the EcoScope tool in the borehole annulus. We looked for sudden decreases of >100 psi in the annular pressure, which could be caused by low-density gas entering the borehole. We also monitored pressure increases of the same magnitude, which could be the result of fluid acceleration caused by a gas kick (Aldred et al., 1998).

Figure F54 shows the measured borehole fluid pressure profile in Hole U1327A after subtraction of the best-fit linear trend. The borehole fluid pressure shows only small fluctuations within ±5 psi over the general trend, and the fluid pressure anomalies observed were well below the 100 psi level that would have required preventive action.

We also monitored the coherence of the sonic waveforms acquired by the SonicVISION tool, focusing on the velocity of fluid in the borehole. Loss of coherence in the waveforms and a slower velocity in the drilling fluid indicate the presence of gas. The sonic waveform coherence image in Figure F54 shows a generally well defined fluid arrival with a slowness of ~200 ms/ft, which corresponds to the expected fluid velocity of ~1500 m/s. There are two anomalous intervals at 120–140 and 232–255 mbsf. The upper interval (120–140 mbsf) shows a fluid velocity higher than the general trend, coinciding with a high-resistivity layer interpreted to contain gas hydrate (see "Gas hydrate and free gas occurrence"). The deeper interval (232–255 mbsf) is below the GHSZ at this site, and if excess gas was present in the formation it had to be free gas. The sonic waveform coherence suggests that the fluid velocity decreased in this interval and that the slowness was higher than the upper bound of 240 µs/ft used in the processing. The fluid pressure measurement, however, shows no significant anomaly at this depth, and if gas was present in the drilling fluid at ~232–255 mbsf it had to be in low concentrations. These results suggest that the fluid velocity measured while drilling may indicate layers that contain gas hydrate or free gas and warrant further study.

Logging quality

Figure F54 also shows the quality control logs for Hole U1327A. The ROP was generally 60 m/h or less in the interval from the seafloor to TD. This is sufficient to record one measurement every 4 cm (~25 measurements/m) in the GeoVISION resistivity, and no significant resolution loss was observed with variation in ROP. The ultrasonic caliper log, which is a direct measurement of the borehole diameter recorded by the EcoScope tool, is our best indicator of borehole size. Most of the hole had a diameter slightly >10 inch (25 cm), with larger washouts slightly >11 inch (28 cm) restricted to the uppermost 70 m of the hole. The density correction, calculated from the difference between the short- and long-spaced density measurements, generally varies from 0 to 0.2 g/cm³ (Fig. F54), showing the good quality of the density measurements. Figure F55 is a summary of the LWD gamma ray, density, resistivity, and resistivity image logs with density and porosity measurements from cores from Holes U1327C and U1327D superimposed (see "Physical properties").

The depth relative to the seafloor was fixed for all LWD logs by identifying the step change in the gamma ray log associated with the seafloor. For Hole U1327A, the gamma ray logging pick for the seafloor was at a depth of 1316 mbrf.

Wireline logging

Operations

Wireline logging was conducted in Holes U1327D and U1327E. Hole U1327D was drilled as a dedicated hole for pressure coring and logging, and was completed at 1625 h on 7 October 2005, reaching a TD of 300 mbsf. The ship's heave was consistently >4 m, which was unsafe for the deployment of any logging tool. We therefore had to wait until 2300 h, when the heave decreased to ~3 m, to start rigging up. The hole was displaced with a barite and sepiolite mud mixture, and rigging of the tool string for the first run was completed by 0055 h on 8 October. Wireline logging operations in Hole U1327D began with deployment of the triple combo tool string (resistivity, density, and porosity measurements), which consists of the HNGS, the DIT, the Hostile Environment Litho-Density Tool (HLDT), the Accelerator Porosity Sonde, and the Temperature/Acceleration/Pressure (TAP) tool. For details on the different tools, see "Downhole logging" in the "Methods" chapter. The tool string reached the bottom of the hole at 1610 mbrf (296 mbsf) without difficulty by 0235 h. Logging immediately showed that the hole was severely enlarged in many locations. The ship's heave was ~3–3.5 m, and during the logging run the heave compensator shut down and had to be restarted several times. Just before the caliper tool was to be closed as it entered the drill pipe, with the heave compensator turned off, the swell caused the tool string to descend suddenly, breaking the caliper's arm. This was confirmed at 0335 h when the tool reached the rig floor. Rig down was complete at 0530 h when rigging for the VSP operations started (see "Vertical seismic profile").

Hole U1327E was drilled to complete wireline operations with acoustic logs that were deemed to be critical to the scientific objectives of the cruise. To save the only HLDT left and considering that the heave was ~3.5 m and expected to worsen, it was decided to run an arm-free tool string composed of the DSI, the Scintillation Gamma Ray Tool (SGT), and the DIT. Hole U1327E was completed to a TD of 300 mbsf at 1600 h on 10 October. The hole was displaced with 10.5 ppg barite mud, and rig up of the tool string was completed at 2110 h. Logging started when the tool string reached 1602 mbrf (289 mbsf) at 2240 h. The ship's heave stayed at 3 m or less during the entire logging run, and there was no notable difficulty. The first pass was completed to the seafloor (1313 mbrf) at 2345 h. The tool string was then brought down for two short repeat runs from TD up to 1490 mbrf (177 mbsf); the first was for quality control and the second to test the new heave compensation system developed by Schlumberger, which worked well. The tool string was back on the rig floor at 0155 h on 11 October. Rig down was completed by 0300 h.

Logging quality

Wireline logging data from the triple combo tool string in Hole U1327D are affected by poor hole conditions (Fig. F56) typical of these unconsolidated formations. The hole shows numerous enlargements that are beyond the 16 inch (41 cm) maximum range of the caliper tool in intervals 110–130, 160–210, and 237–277 mbsf. The caliper tool was turned off at depths shallower than 105 mbsf. Outside these intervals, the neutron porosity and density logs give readings that are close to the core measurements. Where the hole was enlarged, however, the densities are too low and the porosities are too high. Induction log resistivities are generally not as affected by poor hole conditions as the nuclear logs. Nevertheless, the logged resistivities seem to be anomalously low and noisy at depths above 105 mbsf, where we have no caliper information and where the hole was presumably enlarged as well.

The acoustic and resistivity data measured in Hole U1327E appear to be of excellent quality. The acoustic waveforms and slowness-time coherence projection are shown in Figure F57. The quality of the recorded waveforms and the high level of coherence in the monopole and upper dipole waveforms suggest that the hole was in very good condition. As a result, the velocity profiles acquired are very robust, and almost no additional processing was required to derive reliable compressional (V_P) and shear (V_S) wave velocities. Because of operational concerns (see "Operations" in "Wireline logging"), we did not run a caliper tool in this hole. The high quality of the wireline logs suggests that Hole U1327E was more in gauge than Hole U1327D, but we have no direct measurements of hole size to confirm this conclusion.

The depth relative to seafloor for all wireline logs was fixed by identifying the step change in the gamma ray log associated with the seafloor. The gamma ray pick for the seafloor was at 1314 mbrf in Hole U1327D and 1313 mbrf in Hole U1327E.

Logging-while-drilling and wireline logging comparison

Figure F58 shows a comparison of LWD (Hole U1327A) and wireline (Hole U1327D) data, using the gamma ray, neutron porosity, density, and resistivity logs. In general, the LWD and wireline data from each hole match relatively well, exhibiting similar curve shapes and absolute logging values. There are, however, some notable differences.

One difference is seen in gamma ray measurements, where the LWD log gives higher values (100 gAPI on average) than the wireline log (50 gAPI on average). Moreover, the LWD and wireline gamma ray curves have different shapes, and there is no obvious correlation between the two. The cause of this discrepancy has not been fully understood. The wireline data, however, were in agreement with data recorded during Leg 146.

The wireline neutron porosity and density logs contain several intervals where the neutron porosity readings are anomalously high and the density readings are anomalously low compared to the LWD logs. Most of these anomalies are in intervals where Hole U1327D was enlarged (see the caliper log in Fig. F56) and are probably caused by poor contact of the tool pad with the borehole wall.

The comparison of resistivities shows a general trend that is very similar between LWD and wireline logs. The low resistivities measured by the wireline logs in the shallow interval above 105 mbsf are probably caused by washouts in Hole U1327D. There are, however, remarkable differences between the resistivity highs in the LWD and wireline logs that seem to indicate a heterogeneous distribution of gas hydrate. LWD resistivities show a broad peak reaching 10 m at 120–140 mbsf and two smaller peaks (as high as ~4 m) at 110–112 and 170–173 mbsf. The wireline resistivities do not contain any of these resistivity highs but instead have a resistivity peak that reaches ~5 m at 155–160 mbsf; this peak is not seen in the LWD resistivity logs.

There are also differences between wireline resistivity logs acquired in different holes. Figure F59 compares the gamma ray and resistivity logs acquired in Holes U1327D and U1327E. Although differences in the gamma ray logs may be the result of having used different logging tools in the two holes (the HNGS in Hole U1327D versus the SGT in Hole U1327E), the resistivity logs were acquired with the same DIT tool. Discounting differences above 105 mbsf, which are most likely caused by a washout in Hole U1327D, the resistivity peak noted earlier at 155–160 mbsf in Hole U1327D is not observed in Hole U1327E. Also, the Hole U1327E resistivity logs show two minor resistivity peaks (3–4 m) at 107–115 and 125–128 mbsf that are not seen in the Hole U1327D logs. On the other hand, the peak at 107–115 mbsf in Hole U1327E resistivity correlates to a peak found at the same depth in the LWD resistivity log collected in Hole U1327A (cf. Figs. F59; F55).

As a quality control measure, we repeated the logging run in the bottom section of Hole U1327E; logs from this repeat run are also shown in Figure F59 for resistivity (285–175 mbsf) and for the acoustic logs (273–167 mbsf). The logs in the first and second runs are practically identical. The only significant difference is the P-wave velocity log between 190 and 230 mbsf, where the automated processing algorithm picked a secondary coherence high in the second pass (Fig. F57), a mistake that can be easily corrected during full postcruise processing.

The difference in resistivities logged by LWD (Hole U1327A) and wireline (Holes U1327D and U1327E) tools deserves further explanation. Here, we concentrate on carefully documenting these differences, and we discuss their origin in "Differences in resistivity logs, instrumental effects, and lateral heterogeneity."

Logging units

The logged section in Holes U1327A, U1327D, and U1327E can be divided into three logging units based on obvious changes in the LWD and wireline gamma ray, density, electrical resistivity, and acoustic measurements (Figs. F55, F56, F57). There is no obvious correlation between the logging units defined here and the lithostratigraphic units defined in "Lithostratigraphy."

Logging Unit 1 (0–120 mbsf) is characterized by a resistivity trend that steadily increases from ~1 m near the seafloor to ~2 m at the bottom of the unit (120 mbsf). This increase in resistivity with depth is matched by an increase in density (from 1.7 g/cm³ near the seafloor to 2 g/cm³ at 120 mbsf) and a decrease in porosity (from 70% near the seafloor to 50% at 120 mbsf). The P-wave velocities (V_P) average ~1550 m/s. This unit shows only a few small resistivity peaks (e.g., at ~110 mbsf in the LWD logs) (Fig. F55) that may be attributed to the presence of gas hydrate.

Logging Unit 2 (120–230 mbsf) is characterized by a constant background resistivity value of ~2 m and relatively high V_P values that average ~1750 m/s over most of the interval. Resistivity and V_P show a distinct bulge over lower values in the units above and below; this bulge is very obvious for V_P (Figs. F56, F57) but more subtle for the resistivity (Figs. F55, F56). Density and porosity are ~2 g/cm³ and 50%, respectively. This unit shows a number of resistivity peaks that can be attributed to the presence of gas hydrate, although the peaks do not consistently correlate between holes, and the relatively high V_P also suggests gas hydrate occurrence. Although the high velocity values above 230 mbsf suggest the presence of some amount of gas hydrate, no significant waveform amplitude loss can be observed. Elsewhere, strong attenuation has been associated with large amounts of gas hydrate (e.g., Guerin and Goldberg, 2002).

Logging Unit 3 (230–300 mbsf) is characterized by a sharp decrease in V_P (Fig. F56). Below 230 mbsf, V_P drops to very low values near the fluid velocity (~1500 m/s), suggesting the presence of small amounts of free gas. This is supported by the very low dipole waveform amplitudes. V_S also decreases. Although the presence of free gas should not affect shear velocity because fluids do not transmit shear energy, the borehole might have crossed a slightly overpressured interval, where the release of free gas led to a loss of cohesiveness of the formation. The rapid contrast between high and low V_P over a short interval at 205–210 mbsf is likely the origin of the BSR, which should be confirmed by the generation of synthetic seismograms from the recorded V_P and density logs. The depth to the BSR is also discussed in more detail in the "Vertical seismic profile" results below.

Logging Unit 3 also displays a small drop in resistivity compared to Unit 2. Whereas resistivity tends to be just above 2 m in Unit 2, it is just below 2 m in Unit 3. This resistivity decrease is subtle but clearly observable in the wireline logs (Fig. F56). Density and porosity do not show an appreciable change from Unit 2 and remain around 2 g/cm³ and 50%, respectively.

Logging-while-drilling borehole images

The GeoVISION, ADNVISION, and EcoScope LWD tools generate high-resolution images of borehole log data. The ADNVISION and EcoScope tools produce images of density and hole radius (computed on the basis of the density correction, which depends on the borehole standoff). The GeoVISION tool produces a gamma ray image and shallow, medium, and deep resistivity images (see "Downhole logging" in the "Methods" chapter).

Figure F60 shows some of the LWD images collected by the EcoScope and GeoVISION tools. It should be noted that the display in Figure F60 is highly compressed in the vertical direction. The unwrapped images are ~90 cm wide (for an 11 inch diameter borehole), and the vertical scale is compressed by a factor of ~37:1. These high-resolution images can be used for detailed sedimentological and structural interpretations and to image gas hydrate distribution in sediments (e.g., in layers, nodules, or fractures). Gas hydrate–bearing sediments exhibit high resistivities without a corresponding high bulk density. Layers with high resistivities and high densities are likely to be low-porosity, compacted, or carbonate-rich sediments. The most striking feature in the images shown in Figure F60 is the high-resistivity, low-density layer between 120 and 140 mbsf, which suggests high porosity and high gas hydrate concentration.

Logging porosities

Sediment porosities can be determined from analyses of recovered cores and from downhole measurements (see "Physical properties" and "Downhole logging," both in the "Methods" chapter). The LWD density and neutron logs from Hole U1327A were used to calculate sediment porosities. Core-derived physical property data, including porosities (see "Physical properties"), were used to both calibrate and evaluate the log-derived sediment porosities.

The LWD log–derived density measurements were used to calculate sediment porosities () using the standard density-porosity relation

= (_g – _b)/(_g – _w).

We used a constant water density (_w) of 1.03 g/cm³ and a grain/matrix density (_g) of 2.75 g/cm³, which is the average grain density measured in the core samples (see "Physical properties"). The density log–derived porosities range from ~60% at 20 mbsf to ~40% at 300 mbsf (Fig. F61).

The LWD neutron porosity log (Fig. F61) yielded sediment porosities ranging from an average value of ~70% at 20 mbsf to ~50% at 300 mbsf. Porosities measured by the neutron log are expected to be higher than those computed from the density log in sediments containing clay, because the neutron log essentially measures hydrogen abundance, and hydrogen in clay minerals is counted as porosity. The EcoScope neutron porosity shown in Figure F61 is the "best thermal neutron porosity," which has been corrected for the effect of clay, so that it is only slightly higher than the density porosity.

The comparison of core- and LWD-derived porosities in Figure F61 reveals that the log-derived porosities agree with the core-derived values throughout the logged interval, with the density porosities being slightly lower and the neutron porosities slightly higher than the core porosities. The core-derived porosities, however, do not show the porosity increase displayed by LWD/MWD porosities at 120–140 mbsf in logging Unit 2.

Gas hydrate and free gas occurrence

As previously discussed (see "Downhole logging" in the "Methods" chapter), the presence of gas hydrate is generally characterized by increases in electrical resistivity and acoustic velocity that are not accompanied by a corresponding decrease in porosity. A decrease in porosity alone in a water-saturated sediment can result in an increase in resistivity and acoustic velocity. Resistivities logged in Holes U1327A, U1327D, and U1327E show a number of positive anomalies over a general increase of resistivity with depth without a corresponding decrease in porosity (Figs. F55, F56), suggesting that there are several intervals where gas hydrate may be present. The presence of gas hydrate is known to increase P-wave velocity and attenuation, and the relatively large V_P values measured in logging Unit 2 (Figs. F56, F57) are in agreement with the general inference of gas hydrate occurrence at this site.

Water saturation from Archie's equation

To estimate the amount of gas hydrate that might be present at Site U1327, we used the Archie relation (e.g., Collett and Ladd, 2000)

S_w = [(a x R_w)/(^m x R_t)]^1/n,

where

• S_w = water saturation,
• a = tortuosity coefficient,
• R_w = formation water resistivity,
• = density porosity computed from ADNVISION enhanced resolution bulk density,
• m = cementation coefficient,
• R_t = GeoVISION high-resolution button deep average resistivity, and
• n = saturation coefficient.

We use the button deep resistivity instead of the ring resistivity because the button deep resistivity matches the resistivities with the greatest depth of investigation measured by the EcoScope tool (except for having a higher resolution), whereas the ring resistivity values are generally lower than the EcoScope deep resistivities.

Gas hydrate saturation (S_h) is the percentage of pore space in sediment occupied by gas hydrate, which is the complement of the water saturation S_w:

S_h = 1 – S_w .

The procedure followed to estimate S_w with Archie's relation is illustrated in Figure F62. We first computed porosity from the density log as described above with a water density _w of 1.03 g/cm³ and a grain/matrix density _g of 2.75 g/cm³ (see "Physical properties"). Gas hydrate, however, is less dense than seawater and assuming, in principle, that water fills the pores may lead to overestimating the porosity. A simple calculation shows that this effect is small and can be neglected. If gas hydrate saturation were 100%, the overestimate of porosity would be

= [(_g – _b)(_w – _h)]/[(_g – _w)(_g – _h)].

Using the values of water and grain/matrix density defined above, a gas hydrate density _h of 0.91 g/cm³ (Sloan, 1998), and a bulk density _b of 1.7 g/cm³ (which is the bulk density measured in the high-resistivity interval 140–160 mbsf of Fig. F55), we find that the overestimate of porosity is only 4%, even when gas hydrate saturation is 100%. At first approximation, we can ignore the density effect of gas hydrate in the calculation of porosity from density. It should be noted that the small density effect of gas hydrate also implies that the bulk density decrease from ~2 to 1.7 g/cm³ observed in the 140–160 mbsf high-resistivity interval in Hole U1327A cannot simply be caused by high gas hydrate saturation. A gas hydrate saturation S_h in a sediment of porosity would cause a decrease in bulk density compared to a fully water-saturated sediment equal to

_b = –S_h(_w – _h).

For 100% gas hydrate saturation, a porosity of 50%, and densities as above, this decrease in bulk density is only 0.06 g/cm³, much less than the observed density change of 0.3 g/cm³. The low density in the 140–160 mbsf high-resistivity interval in Hole U1327A must be mostly caused by a porosity increase.

To continue with our Archie-based procedure, we next estimated the formation water resistivity (R_w) by first constructing a salinity versus depth function based on IW salinity measurements (see "Interstitial water geochemistry"). This salinity versus depth function is a simple exponential decay fitted to the data (with a salinity of 35 at the seafloor decreasing with an exponential decay constant of 55 m to an asymptotic value of 21 at depth). At every logging depth, we combined the salinity value with a formation temperature obtained from the geothermal gradient estimated from the downhole formation temperature measurements (see "In situ temperature profile") and used the formulas of Fofonoff (1985) to obtain the corresponding value of the water resistivity R_w.

To estimate the water saturation S_w, we also need to choose values for the Archie coefficients a and m. One way to do this is to choose a logged interval where the sediments can be assumed to be water saturated and fit a and m to a plot of measured resistivity versus porosity, known as a "Pickett plot" (e.g., Doveton, 1994). In marine sediments, however, the range of porosity is relatively small and it is not possible to obtain a robust estimate of both a and m. We prefer to set a = 1, which is physically the most realistic value because it gives a resistivity equal to the formation water resistivity when the porosity is 100%. We then compute an "estimated m" coefficient by

m_est = –log(F)/log(),

where F = (R_t/R_w) is the formation factor. This m_est curve should give the appropriate value to be used in Archie's law in water-saturated intervals and will give anomalously high values in intervals that contain hydrocarbons. A reasonable value of m = 2.2 can be chosen from the baseline trend of the m_est curve in Figure F62.

The next step is to compute the resistivity R₀ predicted by Archie's equation for a water-saturated formation of a given porosity, which is given by

R₀ = (a x R_w)/^m.

Using Archie coefficients of a = 1 and m = 2.2, we computed an R₀ curve that follows the measured resistivity R_t throughout logging Unit 1 but is significantly lower than R_t in several intervals of logging Units 2 and 3 (Fig. F62). Finally, we computed the water saturation S_w using a saturation coefficient n = 2.

The most notable intervals in logging Unit 2 that show a measured resistivity R_t greater than the resistivity R₀ predicted for water-saturated conditions are at 120–140 mbsf, where the water saturation S_w predicted by Archie's law is 30%–60%, and 185–203 mbsf, where the predicted S_w is as low as 65%. In logging Unit 3, two intervals (235–250 and 255–262 mbsf) also show a predicted water saturation S_w between 60% and 90%. The Unit 2 intervals are in the GHSZ, whereas the Unit 3 intervals are below the GHSZ. This suggests the presence of free gas in the intervals 235–250 and 255–262 mbsf, in good agreement with the sonic log, which shows V_P <1500 m/s in these intervals (Fig. F56).

Vertical seismic profile

Operations

VSP operations in Hole U1327D started with rig up of the WST, which was completed by 0600 h on 8 October 2005. After rig up, the WST was lowered in the drill pipe just above the mudline (~1300 mbrf), awaiting daybreak. The first GI air gun shots were fired at 0820 h, 1 hour after the start of marine mammal observations, ramping up to the desired operational intensity over 30 min. At 0900 h, it was decided to send the tool to TD, with ship heave in the 3–3.5 m range. There were some problems getting below 1415 mbrf (101 mbsf), where the caliper log indicated a ledge. This obstruction was finally cleared at 0930 h, and the tool reached a maximum depth of 1590 mbrf (276 mbsf) at 0945 h. The first shots for the VSP were fired at 0950 h, with a plan to try a station every 5 m and move along if there was no success in obtaining good coupling between the tool and the borehole wall. We recorded 16 successful stations between 1590 and 1495 mbrf (276–181 mbsf), shooting seven times or more at each station for subsequent stacking. Only a few 5 m intervals were missed, where good coupling could not be achieved because of irregularities in the borehole (see the caliper log of Hole U1327D in Fig. F56). Data acquisition was complicated by several interruptions caused by the heave compensator shutting down when the heave reached >4 m. At shallower depths, however, it became impossible to get any good couplings despite a number of attempts up to 1452 mbrf (138 mbsf). We suspected damage to the arms, and at 1245 h we decided to bring the tool to the rig floor for inspection. The WST, however, did not pass through the bit at the end of the drill pipe. It was not clear if the reason was the flapper valve, the torn arms, or a combination of both. After several unsuccessful attempts to clamp the wire, it was decided to try again to pull the tool string. This attempt was eventually successful and the WST was back on the rig floor at 2300 h, with its two arms damaged.

Time versus depth relationship, interval velocity estimation, and depth to the bottom-simulating reflector

The VSP provides a direct measurement of the time versus depth relationship with the first break of the direct compressional wave arrival. Figure F63 shows the stacked waveforms at the 16 stations that were successfully recorded. Stacking was accomplished by taking the median of the seven or more waveforms recorded at each station. The first break times can be easily seen and are marked by red crosses in Figure F63. Table T24 lists the first break times and the times corrected for the ~50 m horizontal offset of the source, the shot depth of 2 mbsl, and the delay between the trigger time and the start of recording. Corrected first break times are traveltimes from the sea surface to each receiver along a vertical path.

The corrected first break times are plotted in Figure F64 versus the depth of the receivers. The error bars show an uncertainty in the picked first break times of ±1 ms, which combines traveltime errors caused by uncertainties in the depth of the receiver and fluctuations in the vertical source position caused by waves (the latter errors are reduced but not eliminated by stacking). In a medium in which the velocity was constant over the vertical range spanned by the receivers, the first break times follow a straight line whose slope gives the velocity. The first break times of Figure F64 clearly show two separate lines; one with a high velocity of 1843 m/s for receivers between 181 and 234 mbsf and one with a low velocity of 1281 m/s for receivers between 251 and 276 mbsf. The two lines fit the first break times well within the estimated errors of ±1 ms. These velocities define a relatively fast interval, possibly containing gas hydrate, above an interval whose velocities are slow enough to require the presence of at least a small amount of free gas. The velocity change between the fast and slow interval defines the location of the ubiquitous BSR. Similar VSP velocity variations near the BSR were observed at Leg 146 Site 889 on the Cascadian margin (MacKay et al., 1994) and at Sites 994, 995, and 997 on the Blake Ridge (Holbrook et al., 1996).

To compare the VSP results to other velocity measurements, it is useful to translate the first break times into interval velocities. In principle, the vertical distance between two receivers divided by the difference between the two first break traveltimes should give an immediate estimate of the interval velocity. In practice, the first break picks are not exact, and small errors in time picking can translate to large errors in the estimated velocities. A reliable estimate of interval velocity requires an inversion that combines the first break measurements and some smoothness constraint. We use here a Bayesian inversion method where the smoothness of the final solution is determined by the data, while the standard deviation of the traveltime measurement errors is fixed to 1 ms (see above). For method details, see Malinverno and Briggs (2004). The inversion is applied to determine the interval velocity in a layered medium with 5 m thick layers.

The results are shown in Figure F65, which compares interval velocities from the VSP in Hole U1327E, P-wave velocity measurements from the sonic log in Hole U1327D (Fig. F56), and interval velocities determined with the same inversion method from the VSP data collected in Hole 889B, which is ~600 m east of Site U1327. In the VSP interval velocity inversions (Fig. F65A, F65C), the solid line shows the best estimate of compressional wave velocity and the dashed lines show its uncertainty defined by the posterior standard deviation. The posterior standard deviation measures the uncertainty in the estimated velocity caused by the uncertainty in the traveltime picks. The depth interval with the steepest velocity decrease, which should correspond to the BSR, is highlighted by gray rectangles in Figure F65. The Hole U1327D VSP shows the deepest BSR location (245–260 mbsf), whereas the Hole U1327E sonic log suggests a shallower BSR (228–243 mbsf). Apart from this depth difference, the velocities obtained by VSP inversion and measured by the sonic log are remarkably similar above and below the BSR at Site U1327. The BSR at Hole 889B is shallower again, with the velocity gradient change at 215–230 mbsf.

Whereas surface seismic reflection lines show as the BSR a continuous single reflector in the area around Site U1327, high-resolution seismic surveys often resolve the BSR into a number of dipping, high-amplitude reflectors that presumably contain free gas. These bright reflectors abruptly terminate at a depth that should correspond to the bottom of the GHSZ (for examples see fig. 4 of Wood et al., 2002). If the free gas is concentrated in layers with coarser grained sediments and the local sedimentary succession is dipping with respect to the bottom of the GHSZ or is otherwise laterally heterogeneous, it is possible that two nearby boreholes may record the transition to sediment containing free gas at different depths. This lateral variability may explain the variation in the depth to the top of the free gas zone shown in Figure F65.

Differences in resistivity logs, instrumental effects, and lateral heterogeneity

Here, we propose some explanations for the differences in LWD (Hole U1327A) and wireline (Holes U1327D and U1327E) resistivities. In principle, we could assign these variations to either instrumental effects or lateral changes in gas hydrate distribution. By "instrumental effects" we mean that different logging tools may measure different resistivities in the same formation. By "lateral heterogeneity" we mean that different results in different wells are caused by the formation and/or the gas hydrate distribution being horizontally heterogeneous at the scale of a few tens of meters.

We focus first on instrumental effects. Two possible reasons for different logging tools measuring different resistivities in the same formation containing gas hydrate are dielectric effects and resistivity anisotropy. Gas hydrate has a dielectric constant much lower than that of formation water (e.g., Sun and Goldberg, 2005), and the dielectric properties of the formation will affect electromagnetic wave propagation in the high-frequency limit. The highest frequency electromagnetic measurements in our logging suite are carried out by the EcoScope tool, which measures resistivity from the attenuation and phase shift of 400 kHz and 2 MHz electromagnetic waves propagating through the formation (for more details, see "Downhole logging" in the "Methods" chapter).

If the dielectric properties of gas hydrate were responsible for the different resistivity readings, the greatest differences would be observed between the high-frequency EcoScope measurements and the lowest frequency measurements, which are the resistivities measured by the GeoVISION tool at frequencies of 1.5 kHz (Bonner et al., 1996). Indeed, high-frequency measurements by the Array Resistivity Compensated tool (similar to the EcoScope) have been found to give resistivities higher by a factor of seven compared to resistivities measured at low frequency by the GeoVISION tool in the same gas hydrate–bearing layer of a well in the Gulf of Mexico (T. Collett, pers. comm.). On the other hand, Boissonnas et al. (2000) found no evidence of dielectric effects in gas hydrate–bearing formations using 2 MHz LWD measurements and noted that 2 MHz is probably too low a frequency for dielectric properties to significantly affect electromagnetic wave propagation. In Hole U1327A there is essentially no difference between the resistivity measured by the high-frequency EcoScope and the low-frequency GeoVISION tools in Hole U1327A (Fig. F55). Therefore, dielectric effects cannot explain the observed variation in measured resistivities at Site U1327.

Another instrumental effect may be the result of resistivity anisotropy. The results of detailed IW analyses (see "Interstitial water geochemistry") suggest that gas hydrate around Site U1327 may be concentrated in thin sand layers (which will have high resistivity) separated by mostly water-saturated, relatively low resistivity, clay-rich layers, which will have low resistivity. If we measured the effective resistance at scales greater than the bed thickness in such a medium, the resistance to electrical currents flowing vertically will be much higher than the resistance to currents flowing horizontally: current flowing vertically must go through the high-resistivity layers, whereas current flowing horizontally will focus in the low-resistivity layers. If different logging tools were affected differently by horizontal and vertical resistivities, they may measure a different resistivity in the same anisotropic layer. For example, wireline induction logs (which are only sensitive to horizontal resistivity) have been reported to measure resistivities lower than those obtained by laterolog devices (which are sensitive to both horizontal and vertical resistivity) in laminated, anisotropic formations (e.g., Chemali et al., 1987). These differences have been exploited to invert logging measurements made by different tools for horizontal and vertical resistivity of gas hydrate–bearing formations in the Mallik 5L-38 well (Collett et al., 2005).

If resistivity anisotropy is responsible for the differences in resistivities observed at Site U1327, the greatest differences should be between the measurements made by the LWD EcoScope tool and the wireline induction tool, which are only sensitive to horizontal resistivity in a vertical well (Chemali et al., 1987; Hagiwara, 1996), and the measurements made by the LWD GeoVISION tool, which focuses current vertically similarly to a laterolog measurement (Bonner et al., 1996). In fact, we observe the opposite pattern: the greatest differences are observed between the EcoScope resistivity measurements in Hole U1327A and the wireline induction measurements in Holes U1327D and U1327E, which should both be sensitive to horizontal resistivities only. The resistivities measured by the EcoScope and GeoVISION tools, which should differ the most if anisotropy is a factor, are essentially identical. Therefore, anisotropy effects cannot explain the observed variation in measured resistivities at Site U1327 either.

If lateral heterogeneity is the reason for the observed differences in measured resistivities, the predicted pattern is simply that the resistivities measured in the same borehole should be the same, whereas resistivities measured in different boreholes may differ significantly. The observations at Site U1327 agree much better with this prediction. To start, as already noted, the measurements made in Hole U1327A by the LWD EcoScope tool, which uses the highest frequencies and is only sensitive to horizontal resistivity in a vertical well, are entirely consistent with the measurements made in the same hole by the LWD GeoVISION tool, which uses the lowest frequencies and may have some sensitivity to vertical resistivity. Moreover, measurements made by the same wireline induction tool in Holes U1327D and U1327E have resistivity peaks at different depths (Fig. F59). Instrumental effects clearly cannot explain these differences.

The IR core images taken on the catwalk, which measure the core liner temperature, give more evidence in favor of lateral heterogeneity in gas hydrate distribution (see "Physical properties"). Figure F66 shows a comparison of the LWD resistivity image in Hole U1327A with the IR images from Hole U1327C. The LWD resistivity image shows high resistivities between 120 and 140 mbsf, and the IR images show a layer of distinct low temperatures whose top is at ~130 mbsf (top of Core 311-U1327C-17X) and bottom is at ~160 mbsf (bottom of Core 19X). Given that core recovery is incomplete, these two layers have approximately the same thickness. They are, however, displaced by ~10 m in depth. This is a significant lateral change, given that Holes U1327A and U1327C are only ~15 m apart. Moreover, the IR images themselves show significant lateral variability on their own; the temperatures in the 133–155 mbsf intervals are different in Holes U1327C and U1327D (Fig. F66).

If there is lateral heterogeneity in gas hydrate distribution, there may be significant differences between gas hydrate occurrence in different holes. On the other hand, observations made in the same hole should agree. This is, in fact, what we observe if we compare the IR images from Hole U1327D to the wireline resistivity logs measured in the same hole. The IR images show a cold interval at 157–162 mbsf that matches a well-defined resistivity peak in the induction log at 155–160 mbsf (Fig. F66). There is also a less prominent match between a thin, cold interval at 224–225 mbsf and a broader resistivity peak at 220–228 mbsf. On the other hand, if we compare the IR images from Hole U1327D to the resistivity logs from Hole U1327E, there are no matches between cold intervals and resistivity peaks (Fig. F66).

In conclusion, we see significant differences between resistivity (and IR) measurements that ought to give the same response for horizontally continuous features in different holes and correspondence between diverse measurements taken in the same hole. Additional independent evidence for lateral variation is provided by differences in the depth to the BSR observed in VSP interval velocities and sonic logs in different holes (Hole U1327D and U1327E, respectively). Magnetic susceptibility measurements also show significant lateral variation between holes, related to a corresponding change in lithology (see "Physical properties"). This combined evidence strongly points to significant small-scale lateral heterogeneity in gas hydrate distribution at Site U1327.

Temperature data

The Lamont-Doherty Earth Observatory (LDEO) TAP tool was deployed on the wireline triple combo tool string in Hole U1327D (Fig. F67). During the process of coring and drilling, cold seawater is circulated in the hole, cooling the formation surrounding the borehole. Once drilling ceases, the temperature of the fluid in the borehole gradually rebounds to the in situ equilibrium formation temperature. Thus, the temperature data from the TAP tool cannot be immediately used to assess the in situ formation temperature. However, the plot of the temperature profile in Figure F67 reveals a few gradient changes that were caused by borehole temperature anomalies. Specifically, the sudden temperature decrease at 280 mbsf during the uphole trip corresponds to a borehole restriction clearly visible on the caliper log (Fig. F56). The more gradual decrease in temperature during the uphole trip around 170 mbsf corresponds to a decrease in borehole radius as seen on the caliper log (Fig. F56). Finally, the large step decrease in temperature at 95 mbsf during the uphole trip is likely to be related to a large washout, although at this depth the caliper log was turned off.

Top of page   |   Previous   |   Next