Previous   |    Next

doi:10.2204/iodp.proc.348.103.2015

Physical properties

During Expedition 348, physical property measurements were performed on core samples from Holes C0002M and C0002P and cuttings samples from Holes C0002N and C0002P. MAD, P-wave velocity, and electrical resistivity measurements were conducted on discrete samples taken from cuttings and the working half of cores. Thermal conductivity was measured on working halves using a half-space probe. The full suite of whole-round multisensor core logger (MSCL-W) and color spectroscopy logger (MSCL-C) measurements were conducted on whole-round core samples and archive halves, respectively. On cuttings, only measurements of natural gamma radiation (NGR) using the MSCL-W system and magnetic susceptibility using a Kappabridge system were conducted. Dielectric permittivity and electrical conductivity were only conducted on cuttings from Hole C0002N.

Moisture and density measurements

Cuttings

MAD measurements were made on 617 cuttings samples from 870.5 to 3058.5 mbsf to provide a detailed characterization of grain density, bulk density, and porosity. A total of 265 samples from both the 1–4 and >4 mm fraction size were measured at 10 m intervals (Table T28; Fig. F87). In addition, 111 intact cuttings, 12 caving samples, and 3 drilling-induced cohesive aggregates (DICAs) and pillow cuttings were separately handpicked from the >4 mm size fraction at various depths from 1222.5 to 3058.5 mbsf (Tables T29, T30, T31). MAD measurements on 1–4 and >4 mm size fractions and handpicked intact cuttings were conducted every 20–40 m between 2105.5 and 2855.5 mbsf in Hole C0002P.

Grain density values for the 1–4 and >4 mm size fractions are in close agreement throughout the cuttings interval (Fig. F87A). The grain density has considerable scatter from 870.5 to 940.5 mbsf with values ranging from 2.24 to 2.70 g/cm³ due to mixing with artificial cement. The scatter diminishes below 940.5 mbsf, and the grain density slightly increases from 2.68 g/cm³ at 950.5 mbsf to 2.72 g/cm³ at 3058.5 mbsf.

In contrast to the grain density values, bulk density and porosity for the two size fractions maintain close agreement only above 1600.5 mbsf. Below this depth, the trends for the two size fractions begin to sharply diverge. Bulk density for the 1–4 mm size fraction is consistently lower (porosity is higher) than that for the >4 mm size fraction while maintaining a similar value with depth to 3058.5 mbsf (Fig. F87B, F87C). The sharp increase in bulk density and decrease in porosity from 870.5 to 970 mbsf are probably caused by cement mixing (see “Lithology”). In general, bulk density of bulk cuttings increases from 1.9 to 2.0 g/cm³ and porosity decreases from 50% to 40% with depth in both size fractions. Exceptionally higher bulk density (and lower porosity) values are observed in intervals 1680–1690 and 2010.5–2090.5 mbsf in Hole C0002N and 2162.5–2220.5 and 2601.5 mbsf in Hole C0002P (see “Effects of operation on MAD measurements”). Taking a closer look at the data, bulk density increases and porosity decreases above 1600 mbsf, followed by a gradual decrease in bulk density and increase in porosity toward ~2000 mbsf. Below 2000 mbsf, bulk density shifts to slightly higher values and porosity to lower values and stays almost constant above 2550 mbsf. Below 2550 mbsf, bulk density values decrease, with an apparent porosity increase, and stay almost constant from 2750 to 3058.5 mbsf.

Handpicked intact cuttings and remolded cuttings

Visual observation suggests that three types of cuttings, intact cuttings, DICAs, and pillow cuttings, are present with different degrees of induration or strength (see “Operations”). Unexpectedly high porosity and low bulk density values seem to be caused by the presence of DICAs and pillow cuttings, which are formation materials remolded during the drilling and recovery process. Therefore, we separately hand picked intact cuttings and DICAs/pillow cuttings from the >4 mm size fractions in addition to MAD measurements on the bulk cuttings (Tables T29, T30).

Grain density, bulk density, and porosity values for handpicked intact samples were measured within the 1222.5–3058.5 mbsf interval (Figs. F87, F88). Some intact samples from Hole C0002P were separated according to lithology when possible (mud or silt/sand) (Table T29). Grain density measurements for the handpicked intact cuttings are consistent with the bulk cuttings, including both the 1–4 and >4 mm size fraction. However, handpicked intact cuttings show a consistent decrease in porosity and increase in bulk density when compared to the bulk cuttings (Fig. F87). Grain density slightly increases from 2.66 to 2.70 g/cm³, and bulk density exponentially increases from 2.05 to 2.41 g/cm³ throughout the entire interval.

Porosity and bulk density values of handpicked intact samples are generally offset from the bulk measurements. Porosity of the intact samples decreases from 37.0% to 26.9% at 1222.5 to 1800.5 mbsf, stays constant at ~27% from 1800.5 to 2000.5 mbsf, and decreases to 22.2% at 2245.5 mbsf in Hole C0002N (Figs. F87C, F88C). Porosity of the intact samples from Hole C0002P is slightly offset from the trend in Hole C0002N and decreases from 22.7% to 20.3% at 1960.5-2162.5 mbsf. Porosity stays constant at ~21.0% to 2660.5 mbsf and slightly decreases to 17.3% at 3058.5 mbsf. No significant differences in porosity between lithologies are observed when compared with the other handpicked intact samples.

Bulk density, porosity, and grain density of the DICAs/pillow cuttings at 1225.5, 1505.5, and 2090.5 mbsf are consistent with the 1–4 and >4 mm size fraction trends (Fig. F87). This suggests that the presence of DICAs/pillow cuttings does not affect the grain density measurements on bulk cuttings, but they do seem to influence the bulk density and porosity measurements reported above. The grain density agreement suggests that the measurements are of good quality, and the differences in porosity and density are real, but values from the bulk samples are affected strongly by artifacts of the drilling process. Thus, the bulk density and porosity data on handpicked intact cuttings better represent formation properties.

Effects of operation on MAD measurements

In some intervals, including 1680–1690 and 2010.5–2090.5 mbsf in Hole C0002N and 2162.5–2220.5 and 2601.5 mbsf in Hole C0002P, bulk cuttings, particularly of the >4 mm size fraction, show sharp increases in bulk density and decreases in porosity close to the handpicked intact cuttings values (Figs. F87, F89). This feature seems to be strongly related to drilling operations in these intervals, including WOW, hole cleaning, mud pump circulation, and RCB coring (Fig. F87). In the interval from 2010.5 to 2090.5 mbsf, for example, bulk cuttings increase sharply in bulk density and decrease in porosity followed by a gradual recovery with depth (Figs. F87, F89A). In this interval, there was a 50 h long WOW at 2008 mbsf and an 8 h mud circulation period for mud loss treatment at 2038 mbsf (see “Operations”). The size of cuttings in this interval is generally larger, and thus the size of >4 mm bulk cuttings was measured by visual observation of the long axis (Fig. F90A). The average cuttings size in the >4 mm fraction quickly increases with depth, reaches a maximum at 2055.5 mbsf, and then gradually decreases with depth. A rough correlation with increasing cuttings size and decreasing porosity is seen throughout this interval (Fig. F90B). The sudden decrease in apparent porosity and increase in the apparent bulk density with larger cuttings size are probably due to WOW and hole cleaning in this interval.

In the coring interval between 2163 and 2218.5 mbsf, cuttings were collected three times during (1) RCB coring, (2) drilling with a polycrystalline diamond compact drill bit, and (3) opening with an underreamer (Fig. F89B, F89C). In this interval, the bulk cuttings in the >4 mm fraction during coring and the subsequent hole opening with the polycrystalline diamond compact drill bit in both size fractions show lower porosity. This implies that the RCB coring BHA produces a larger amount of better quality cuttings.

Furthermore, cuttings sampled during reaming between 2105.5 and 3050.5 mbsf in Hole C0002P are of much better quality with fewer DICAs and no pillow cuttings present. This results in reasonable values of MAD measurements on bulk cuttings. The porosity of bulk cuttings in both size fractions during reaming is also lower and is similar to that of the handpicked intact cuttings (Figs. F87, F89C).

Discrete cores and porosity-depth trend

A total of 67 discrete core samples from Holes C0002M and C0002P were measured for MAD. All MAD data from Expedition 348 cores are summarized in Figure F88 and Table T32. In Hole C0002M between 475.85 and 507.2 mbsf, bulk density ranges from 1.87 to 2.03 g/cm³, grain density ranges from 2.68 to 2.82 g/cm³, and porosity ranges from 41.6% to 49.8%. Porosity, bulk density, and grain density are consistent with results from Holes C0002B and C0002L on core samples in the same depth interval (Expedition 315 Scientists, 2009; Strasser et al., 2014b). The scatter present could correspond to the occurrence of sand layers that yield lower porosity and higher bulk density; these layers become less common with depth.

In Hole C0002P between 2163 and 2217 mbsf, bulk density ranges from 2.16 to 2.42 g/cm³, grain density ranges from 2.68 to 2.77 g/cm³, and porosity ranges from 18.0% to 32.6% (Figs. F87, F89D; Table T32). At 2205 mbsf, a fault zone was observed, and grain density decreases to 2.64 g/cm³, bulk density decreases to 1.96 g/cm³, and porosity sharply increases to 42.6%. Below the fault zone, both grain density and bulk density increase back to ~2.71 and ~2.34 g/cm³ at 2217 mbsf, respectively, whereas porosity decreases back to a trend that is offset from the trend seen above the fault zone. Porosity ranges from 21.7 to 31.5% between 2206 and 2217 mbsf.

Combined with MAD measurements on handpicked cuttings and discrete core samples including data from previous expeditions, porosity generally decreases with depth at Site C0002 (Fig. F88). The porosity data can be fit well by an exponential function (Athy, 1930):

where z is the depth below seafloor, a is an empirical constant (4.26 × 10^–4), and ϕ₀ is reference porosity (59.1%). It should be noted that good data fit by Athy’s model can be coincidental because the model is suitable for locations where the loading condition does not change and is monotonically progressive with the burial depth (e.g., in a sedimentary basin). This fit may not be applicable for an accretionary prism setting where the loading condition and history may be different with depth. Future postexpedition research, including laboratory experiments, borehole breakout analysis, and microstructural analyses will shed light on the evolution of porosity in the inner accretionary wedge.

Electrical conductivity and P-wave velocity measurements on discrete samples and cuttings

Electrical conductivity and P-wave velocity (V_P) measurements were conducted on 8 cubic samples collected from the working half of cores taken from 476–507 mbsf in Hole C0002M. Discrete samples could not be collected from RCB cores in this depth interval in Hole C0002B during Expedition 315 due to sample disturbance. Section 348-C0002M-4R-1, collected with the SD-RCB, shows similar types of disturbance as cores from Hole C0002B, with wavy open cracks and soft-sediment intervals. Overall, the samples were very fissile and tend to crack under load. As a result, the samples exhibited high attenuation associated with the P-wave analysis.

Electrical conductivity of both core and cuttings samples was measured at 10 kHz to allow direct comparison with Expedition 315 measurements. Electrical conductivity values measured during Expedition 338 were recalculated from raw data for 10 kHz and are reported in Table T33.

Electrical conductivities of samples from Hole C0002M appear lower than those recovered at deeper intervals from the same lithologic Unit II (Kumano Basin fill) in Hole C0002B, although porosity decreases with depth. Both chlorinity and salinity display a broad minimum at ~400 mbsf, and the variations in chlorinity and salinity with depth are a possible explanation for the lower conductivity values. In the 476–507 mbsf interval, samples from Hole C0002M decrease in electrical conductivity with depth (Fig. F91). This may be due to decreasing porosity with depth and lithologic variations implied from an apparent increase in NGR and clay content within the same interval. Anisotropy is observed with similar horizontal components and is systematically lower in the vertical component (Fig. F92). The range of vertical plane anisotropy is comparable to that observed deeper in lithologic Unit II. Anisotropy related to clay compaction fabric is expected to increase progressively with depth (Henry et al., 2003). The scatter in anisotropy values and the lack of an obvious trend with depth suggest that the presence of cracks influences the anisotropy of these samples.

P-wave velocity measurements on Hole C0002M cores could not be systematically done because of high attenuation. Five samples show velocities ≤1550 m/s, probably due to the presence of gas/air within the sample (Table T33; Fig. F93). Three samples have relatively high velocities in the horizontal directions, from 1830 to 1950 m/s, and the velocity along the z-axis ranges from 1583 to 1738 m/s. When combined with Expedition 315 data, these measurements define a trend for lithologic Unit II below 475 mbsf:

Vertical-plane anisotropy could be reliably measured on only two samples and appears to be as high as 9% and 16%, which might be expected for samples bearing fractures in the x-y plane (Table T33).

In Holes C0002N and C0002P, electrical conductivity and P-wave velocity were measured on discrete core samples and on cuttings from the corresponding depth interval (Tables T33, T34, T35). Preparation of cubes for P-wave and electrical conductivity anisotropy was difficult in Core 348-C0002P-5R and in the uppermost two sections of Core 6R due to the presence of dense fracturing attributed to the fault in Section 5R-4 (see “Structural geology”).

Electrical conductivity for discrete samples from the 2163–2216.87 mbsf coring interval in Hole C0002P is generally greatest in the z (vertical)-direction. This can be explained by the subvertical strata that generally exceed 45° dip (Fig. F91). It is also observed that when two axes of measurements lie near the stratification plane (e.g., z and y or z and x), the values obtained in these two directions are similar and consistently higher than the other direction. This further suggests that the stratification is the primary source of anisotropy within the samples. The ratio of the maximum and minimum conductivities measured on such samples varies between 1.2 and 2. The least anisotropic sample (348-C0002P-2R-4, 73–77 cm; 2177.46 mbsf) contains high amounts of sand. Systematic variations in the average conductivity are observed with depth, with a minimum at 2182 mbsf and a maximum at 2175 mbsf. Lithologic variations in the samples remain a possible explanation for the differences in electrical conductivity. Measurements on cuttings are compared with values from discrete core samples (Fig. F94; Table T34). Most cuttings in the cored interval lie within the lower range of measurements done on discrete core samples. This may be likely because most cuttings platelets are prepared parallel to stratification, or fissility, over this interval. When all data from this expedition are compared with those from previous expeditions, they define a broad trend of decreasing conductivity with depth (Fig. F94), as may be expected considering that the evolution of conductivity is broadly approximated well by an exponential Archie’s law. However, conductivity on cuttings appears relatively scattered. The range of variations (0.15–0.9 S/m) for both silt/clay and sandstone lithologies is broad in the interval from 2000 to 2020 mbsf and may reflect the large variability of porosity measurements on core samples from Hole C0002P. Below 2020 mbsf, sandstone cuttings display a larger amount of scatter and greater variability with depth when compared to the mudstone cuttings. The sandstone cuttings measured in the upper part of Hole C0002P (2000–2200 mbsf) plot in the same conductivity range as the mudstones, but these sandstones are generally fine grained and silty. The coarser sandstone cuttings sampled below 2400 mbsf have higher conductivity than the fine-grained sediment, but this contrast appears to decrease with depth. The conductivity of the mudstone cuttings shows a small offset between 2650 and 2720 mbsf. Mean conductivity is 0.23 ± 0.05 S/m between 2100 and 2700 mbsf and 0.17 ± 0.04 S/m below 2700 mbsf.

Water content of cuttings samples was obtained from wet and dry weight on all the platelets after impedance measurements, and porosity was estimated assuming a grain density of 2700 kg/m³. For discrete core samples, porosity data are available from MAD measurements on a separate discrete sample collected next to the cubic samples. When conductivity, σ, is plotted versus porosity, ϕ, it appears that most samples plot above Archie’s law:

assuming fluid conductivity, σ_f, is either equal to the fluid used for sample preparation (NaCl solution; 5.28 S/m at 21.5°C) or approximating a possible in situ fluid conductivity (4 S/m, corresponding to a chlorinity of 420 mM; see “Geochemistry”) (Fig. F95). No strong difference between the electrical properties of sand-rich and clay-rich lithologies is observed in this depth interval, with most of the sandstone data plotting in the upper range of conductivity at a given porosity. It appears that the relatively high conductivity of the sandstones in the 2400–2800 mbsf interval results from the combined effect of higher porosity values and a different porosity-conductivity relationship.

In the P-wave velocity measurements, traveltime was first determined by automatic picking in the Geotek system. Systematic manual repicking was needed for cuttings and Hole C0002P core samples, as the automatic picking method is sensitive to variations of signal frequency and appears to overestimate the range of P-wave variations because attenuation of high-frequency P-waves in the samples reduced the central frequency of the transmitted signal. Variations in sample attenuation increase the scatter of the automated picking, and the delay for the zero distance appears to be overestimated when performing the standard calibration with the 500 kHz transducers.

The traveltime was manually determined using the first maximum of the second derivative of the waveform, referenced to the signal recorded with the transducers in contact for delay time. Values obtained on discrete samples show scatter between 1900 and 2900 m/s, but the following features seem to emerge from the data set (Fig. F96; Table T35):

1. Measurements on cuttings appear slightly higher but compatible with those measured on cores, perhaps reflecting the likelihood that the selected cuttings are probably biased toward the stronger parts of the formation.
2. P-wave velocity sharply increases with depth above the cored interval between 1990 and 2150 mbsf and decreases between 2150 and 2250 mbsf throughout the cored interval. These velocity variations appear to be related to differences in formation physical properties above and below the fault zone rather than to the fault zone itself, considering the length scale at which this variation is observed and the absence of measurements within the fault zone.
3. On average, velocity increases below 2250 mbsf. A number of high-velocity outliers (V_P = ~4 km/s or more) are observed and correspond to unusually hard cuttings samples, which is likely the result of the presence of natural cements. A change in the depth trend of velocity toward lower values is observed at ~2700 mbsf, evidence supporting the hypothesis that a lithologic or diagenetic change occurs at this depth interval.

Thermal conductivity

Thermal conductivity was measured on the working half of cores from Holes C0002M and C0002P using a half-space probe. All data are summarized in Figure F97 and Table T36, along with data from Expeditions 315 and 338 (Expedition 315 Scientists, 2009; Strasser et al., 2014b). Thermal conductivity ranges from 1.44 to 1.58 W/(m·K) from 477 to 507 mbsf in Hole C0002M and 1.59 to 1.82 W/(m·K) from 2173.0 to 2214.6 mbsf in Hole C0002P. Thermal conductivity in Hole C0002M is consistent with the data from Expeditions 315 and 338. Overall, thermal conductivity increases with depth; however, its rate of increase is much less in the accretionary prism below ~1000 mbsf when compared to that of the Kumano Basin.

The correlation between thermal conductivity and porosity is shown in Figure F98. The theoretical values of thermal conductivity are calculated for different grain thermal conductivities (k_s) based on a geometric mean mixing model:

where k_w is water thermal conductivity. The relationship between thermal conductivity and porosity of the prism sediment from Hole C0002P and previous expeditions follows the extrapolation from the data of Kumano Basin sediment at Site C0002 or Shikoku Basin sediment at Sites C0011 and C0012 (Expedition 315 Scientists, 2009; Expedition 322 Scientists, 2010a; 2010b; Expedition 333 Scientists, 2012a; 2012b; Strasser et al., 2014b). The grain thermal conductivity ranges from 2.1 to 3.4 W/(m·K) and is centered at 2.6 W/(m·K).

MSCL-W (whole-round cores)

Whole-round SD-RCB cores from Hole C0002M and RCB cores from Hole C0002P were analyzed by the MSCL-W. The results of gamma ray attenuation (GRA) density, magnetic susceptibility, NGR, and electrical resistivity measurements (see “Physical properties” in the “Methods” chapter [Tobin et al., 2015]) on whole-round cores are summarized in Figure F99. The data for Hole C0002M are shown together with Expedition 315 and 338 data (Expedition 315 Scientists, 2009; Strasser et al., 2014b). MSCL-W P-wave measurements are not presented here because they exhibit an extreme amount of noise due to poor contact between liner and sediment with voids. Magnetic susceptibility and electrical resistivity on Core 348-C0002M-3R (493.50–496.06 mbsf) are not available due to the presence of an aluminum core liner.

Overall, all MSCL-W data in Hole C0002M are consistent with the previous expedition results, except for NGR, which shows higher values than the previous expeditions. This is probably caused by the larger diameter, and thus volume, of the SD-RCB cores (inner diameter of 73 mm for SD-RCB core liner and 66 mm for regular RCB core liner) and by the use of a different calibration curve (see discussion in “Natural gamma radiation (cuttings)” for details). In Hole C0002P, all MSCL-W data are almost constant in the coring interval of 2163–2218.5 mbsf. Extremely high magnetic susceptibility is observed at 2183 mbsf (interval 348-C0002P-3R-1, 56–64 cm), where a metal piece derived from the core catcher was found in the core liner. Magnetic susceptibility shows a slight offset across the fault at 2205 mbsf (Section 348-C0002P-5R-4); it is ~2.5 × 10^–4 SI above the fault and ~1.5 × 10^–4 SI below it. Electrical resistivity increases right below the fault in Section 5R-5, which might have been disturbed during pulling out from the core barrel.

Color spectroscopy (archive halves)

The results of color reflectance measurements using the MSCL-C are summarized in Figure F100. In Hole C0002M between 475 and 507.57 mbsf, L* ranges from 19 to 48, a* ranges from –2.3 to 1.4, and b* ranges from –1.3 to 6.0. All values show no significant difference from cores collected during Expeditions 315 and 338 (Expedition 315 Scientists, 2009; Strasser et al., 2014b). In Hole C0002P between 2163.0 and 2217.5 mbsf, L* ranges from 20 to 66, a* ranges from –4.1 to 8.5, and b* ranges from –8.0 to 4.6.

Natural gamma radiation (cuttings)

Unwashed cuttings were collected in the core cutting area and packed in a 12 cm long core liner. The liner filled with cuttings was scanned with the MSCL-W to determine the NGR of the cuttings. The NGR of a liner filled with water was measured as a background reference, yielding a value of 34.8 counts/s.

NGR generally increases with depth (Fig. F101). In Hole C0002N, a sharp increase in NGR from 17.6 counts/s at 875.5 mbsf to ~40 counts/s at 975.5 mbsf is probably caused by a decrease in mixing with borehole cement in the cuttings. NGR slightly increases with depth to ~42 counts/s at 1825.5 mbsf and then is shifted toward a slightly higher value of ~45 counts/s and keeps increasing with depth to ~48 counts/s at the bottom of Hole C0002N (2325.5 mbsf). In Hole C0002P, NGR increases from 38 counts/s at 1965.5 mbsf to 47 counts/s at 2095.5 mbsf, followed by a sharp decrease to 40 counts/s at 2115.5 mbsf. The low NGR values in the upper portion of the hole probably reflect mixing with cement and metal in the cuttings. Below 2115.5 mbsf, NGR increases with depth toward ~50 counts/s at 2445.5 mbsf, shifts slightly to a lower value of 43.5 counts/s, and increases again with depth to ~50 counts/s at the bottom of Hole C0002P.

In addition to the unwashed cuttings, we attempted to measure NGR on lightly washed cuttings from Hole C0002P to understand the effect of mud water. Washed cuttings were collected separately with unwashed cuttings in the core cutting area. NGR values on the lightly washed cuttings show a similar trend to those of unwashed cuttings, but with a greater scatter (Fig. F101B).

It should be noted that NGR values of unwashed cuttings measured during Expedition 348 are higher by ~6 counts/s than those measured during Expedition 338. This is probably caused by the usage of different calibration curves between expeditions, because NGR values of a liner filled with water and granite measured for quality checks during Expedition 348 are consistently higher than those run during Expedition 338. Also, there is no difference in NGR values between mud water collected during Expedition 338 and Expedition 348 when calibrated with the curve used for Expedition 348 (Table T37).

Magnetic susceptibility (cuttings)

Magnetic susceptibility was measured on 148 vacuum-dried cuttings samples from both the 1–4 and >4 mm size fractions. Sample weight varied between the two cuttings sizes, so we calculated the mass magnetic susceptibility (MMS) from measured raw data magnetic susceptibility (bulk susceptibility) by

The MMS in the >4 mm fraction is generally less scattered, ranging from 7.88 × 10^–8 to 1.2 × 10^–6 m³/kg in the 1–4 mm fraction and from 9.69 × 10^–8 to 4.05 × 10^–7 m³/kg in the >4 mm fraction between 875.5 and 3000.5 mbsf (Fig. F102). The depth trend of MMS values closely matches lithology. MMS decreases from 4.52 × 10^–7 to 9.6 × 10^–8 m³/kg between 875.5 and 975.5 mbsf, which corresponds to lithologic Unit III. MMS is almost constant at ~1.2 × 10^–7 m³/kg in Subunit IVA (975.5–1045.5 mbsf) and sharply increases at 1045.5 mbsf (Subunit IVA/IVB boundary). MMS gradually decreases from 5.4 × 10^–7 m³/kg at 1045.5 mbsf to 1.2 × 10^–7 m³/kg at 1215.5 mbsf. High MMS probably corresponds to abundant volcanic glass observed in these intervals (see “Lithology”). MMS is almost constant at ~1.1 × 10^–7 m³/kg below 1215.5 mbsf to the Subunit IVD/IVE boundary and slightly increases with depth to 1.3 × 10^–7 m³/kg at 1765.5 mbsf. Below 1765.5 mbsf, MMS is almost constant at ~1.1 × 10^–7 m³/kg, except for scatter near 2000 and 2200 mbsf, possibly due to contamination of metal from window milling.

Dielectric permittivity and electrical conductivity (washed cuttings)

Approximately 150 cuttings samples were collected from 875 to 2325 mbsf at 10 m intervals in Hole C0002N, and ~120 samples were retrieved from 2325 to 3058 mbsf at 10 m intervals from Hole C0002P. During the expedition, only samples from Hole C0002N were tested because Hole C0002P samples were recovered too late in the expedition and will be used onshore for postexpedition analysis. Real and imaginary relative dielectric permittivity was measured in the frequency range from 300 kHz to 3 GHz, but some of the data were rejected due to experimental errors, instrument performance and calibration, and poor coupling with the sample. An example of the raw dielectric data is provided in Figure F103 for Sample 348-C0002N-312-SMW (2255.5 mbsf). The example dielectric spectra show that the real dielectric permittivity (blue) monotonically declines smoothly at all frequencies above ~30 MHz, but below that frequency the data rapidly vary between high and low values. This occurs because of instrumental errors and is not related to the sample. More conductive samples are typically less affected. We have rejected all data with values <100 MHz for this reason. The imaginary dielectric data presented in Figure F103 are less affected by these issues across the entire frequency range presented and show real relative dielectric values ε′_r = 35–50 in the frequency range from 100 MHz to 3 GHz, which is consistent with wet clays. Pure water has a real relative dielectric permittivity of ε′_r = 80, and ultradry clay would typically be ε′_r = ~5–6.

The 100 and 300 MHz and 1 and 3 GHz data were extracted from the dielectric spectra for each sample. We chose 100 MHz because it is the lowest frequency of the acceptable data that is common to all of the samples and 3 GHz because it is the highest. 300 MHz and 1 GHz data were also selected to provide two additional frequencies evenly distributed in (frequency) log space. In Figure F104, the 300 MHz data are plotted against depth to create a pseudo-log. In addition to the dielectric data, we also provide the conductivity of the water decanted from the paste sample after centrifugation (see “Physical properties” in the “Methods” chapter [Tobin et al., 2015]) and the water content of the residual paste at the time of dielectric analysis. Water content analysis was done on ~55 of the samples because the water content affects the dielectric permittivity of the paste.

A simple appraisal of the dielectric analysis results suggests that there are possible step change boundaries at 1050 and 1250 mbsf. Loss-angle logs at four spot frequencies were created by dividing the imaginary dielectric permittivity by the real dielectric permittivity (Fig. F105). These logs indicate a number of additional boundaries and features that are not as easily identified in the primitive (real and imaginary) logs. One zone containing upper and lower boundaries at 1000 and 1080 mbsf, respectively, is marked in yellow and dark blue. This boundary is more apparent at 100 and 300 MHz than at 3 GHz; however, the reverse is true for another subtle boundary at 2180 mbsf (dark green). A possible feature occurring at 1520 mbsf (light green) consists of two adjacent measurements that are higher than the surrounding values. LWD data recorded downhole included gamma ray and resistivity logs, and these typically correlate favorably with dielectric logs (Hizem et al., 2008). In Figure F106, we present the typical dielectric analysis logs against the downhole logs along with the log units (see “Logging”).

A a step increase in the gamma ray log at ~1650 mbsf indicates that the hole has passed from a less (low gamma radiation) to a more clay-rich unit (high gamma radiation). There is also a trend in the (39 and 27 inch) resistivity logs: these increase with depth from a starting value of ~1 Ωm at 900 mbsf to ~2 Ωm at 1600 mbsf. Below 1600 mbsf, the trend remains at a nearly constant value of ~2 Ωm. The same trend in resistivity logs is observed in the real and imaginary dielectric cuttings logs and in the water conductivity logs.

Top of page   |    Previous   |    Next